Nanocrystals

edited by
Yoshitake Masuda

SCIYO
Nanocrystals
Edited by Yoshitake Masuda

Published by Sciyo
Janeza Trdine 9, 51000 Rijeka, Croatia

Copyright © 2010 Sciyo

All chapters are Open Access articles distributed under the Creative Commons Non Commercial Share
Alike Attribution 3.0 license, which permits to copy, distribute, transmit, and adapt the work in any
medium, so long as the original work is properly cited. After this work has been published by Sciyo,
authors have the right to republish it, in whole or part, in any publication of which they are the author,
and to make other personal use of the work. Any republication, referencing or personal use of the work
must explicitly identify the original source.

Statements and opinions expressed in the chapters are these of the individual contributors and
not necessarily those of the editors or publisher. No responsibility is accepted for the accuracy of
information contained in the published articles. The publisher assumes no responsibility for any
damage or injury to persons or property arising out of the use of any materials, instructions, methods
or ideas contained in the book.

Publishing Process Manager Iva Lipovic

Technical Editor Martina Peric
Cover Designer Martina Sirotic
Image Copyright Frannyanne, 2010. Used under license from Shutterstock.com

First published September 2010

Printed in India

A free online edition of this book is available at www.sciyo.com

Additional hard copies can be obtained from publication@sciyo.com

Nanocrystals, Edited by Yoshitake Masuda

p. cm.
ISBN 978-953-307-126-8
SCIYO.COM
WHERE KNOWLEDGE IS FREE
free online editions of Sciyo
Books, Journals and Videos can
be found at www.sciyo.com
Contents
Preface VII

Chapter 1 Morphology control, self-assembly

and site-selective deposition of metal oxide nanocrystals 1
Yoshitake Masuda

Chapter 2 Growth of undoped and metal doped

ZnO nanaostructures by solution growth 31
Rathinam Chandramohan, Thirukonda Anandamoorthy Vijayan
and Jagannathan Thirumalai

Chapter 3 Synthesis and morphology control

of Eu3+ doped M2O2S [M=Y, Gd] nanostructures 45
Rathinam Chandramohan, Jagannathan Thirumalai
and Thirukonda Anandamoorthy Vijayan

Chapter 4 Emission of semiconductor nanocrystals

in photonic crystal environment 65
Sergei G. Romanov and Ulf Peschel

Chapter 5 A review of high nanoparticles concentration composites:

semiconductor and high refractive index materials 109
Igor Yu. Denisyuk and Mari Iv. Fokina

Chapter 6 Diluted magnetic semiconductor nanocrystals in glass matrix 143

N. O. Dantas, E. S. Freitas Neto and R. S. Silva

Chapter 7 KTiOPO4 single nanocrystal

for second-harmonic generation microscopy 169
Loc Le Xuan, Dominique Chauvat, Abdallah Slablab,
Jean-Francois Roch, Pawel Wnuk and Czeslaw Radzewicz

Chapter 8 Zeolite nanocrystals - synthesis and applications 191

Teruoki Tago and Takao Masuda

Chapter 9 Complex nature of charge trapping

and retention in NC NVM structures 207
A.Nazarov and V. Turchanikov
VI

Chapter 10 Carrier storage in Ge nanocrystals grown

on silicon oxide by a two step dewetting / nucleation process 231
A. El Hdiy, M. Troyon, K. Gacem and Q.T. Doan

Chapter 11 Nano-crystals for quadratic nonlinear imaging:

characterization and applications 249
Sophie Brasselet and Joseph Zyss

Chapter 12 Hybrid colloidal nanocrystal-organics based light emitting diodes 271

Aurora Rizzo, Marco Mazzeo and Giuseppe Gigli

Chapter 13 Seasoning ferroelectric liquid crystal with colloidal

nanoparticles for enhancing application flavors 291
Jung Y. Huang

Chapter 14 Organic nanocrystals for nanomedicine and biophotonics 311

Koichi Baba, Hitoshi Kasai, Kohji Nishida and Hachiro Nakanishi
Preface
Nanocrystals have attracted much attention for the next generation science and technology.
They cover many areas of materials science such as inorganic materials, organic materials,
hybrid materials, biomaterials, and relate to various research fields, such as chemistry,
physics, biology, electronic devices, optic devices, etc.
The nanocrystals provide excellent novel functions by control of crystal growth, morphology,
size, crystallinity, anisotropy, orientation, self-assembly, integration, etc. Each crystal has
a different crystal structure. Thus, they have anisotropy of crystal faces and crystal axis.
Morphology control of nanocrystals can be realized by the use of several crystal growth modes
through the change of crystal growth conditions. For instance, variety of crystal growth modes
can be used for morphology control of nanocrystals in aqueous solution processes by the
change of solution temperature, ion concentration, pH, additives, etc. Novel nanocrystals that
have unique structures and functions were developed using these techniques. Additionally,
self-assembly, orientation, integration and patterning of nanocrystals provide novel functions
and values. Self-assembly and 2D-patterning of nanocrystals, for instance, were developed
for photonic crystals. Self-assembly brought out a novel function from the nanocrystals.
Nanocrystal is one of the most important research fields for the next generation science and
technologies. Nanocrystal science is a cross-cutting research and includes various materials
and research areas. It has ripple effects throughout the materials science. The scientific
knowledge and technologies in this book will contribute to the development of nanocrystal
research and related future applications.
In closing, I wish to express my sincere sense of gratitude to the authors, book manager Ms. Iva
Lipovic, and the publishing staff. I dedicate this book to my parents, Mr. Toshio Masuda and
Ms. Nobuko Masuda, my sisters, Ms. Shinobu Horita and Ms. Satoe Amaya, my children, Ms.
Yuuka Masuda, Ms. Arisa Masuda and Mr. Ikuto Masuda, and my wife, Ms. Yumi Masuda.

Editor

Yoshitake Masuda
National Institute of Advanced Industrial Science and Technology (AIST)
Japan
Morphology control, self-assembly and site-selective deposition of metal oxide nanocrystals 1

X1

Morphology control, self-assembly and site-selective

deposition of metal oxide nanocrystals

Yoshitake Masuda
National Institute of Advanced Industrial Science and Technology (AIST), 2266-98
Anagahora, Shimoshidami, Moriyama-ku, Nagoya 463-8560, Japan

Keywords
Nanocrystals, Metal Oxide, Morphology Control, Self-assembly, Site-selective Deposition, Self-
assembled Monolayer, Liquid Phase Process, Aqueous Phase Process, Aqueous Solution Process

1. Introduction
Nanocrystals of metal oxides have been attracted much attention for future science and
technology. Morphology control, self-assembly and site-selective deposition of nanocrystals
will open a new frontier in materials science. They were realized in this chapter using
aqueous solution processes for metal oxide nanocrystals.
Site-selective deposition (SSD) of metal oxide thin films was developed to fabricate
nano/microstructures of metal oxide such as TiO2, Fe3O4, ZnO, Y2O3:Eu, etc.. Several
conceptual processes for SSD using self-assembled monolayers (SAMs) as templates were
proposed, and nano/micropatterns of ceramic thin films were successfully fabricated.
Molecular recognition of SAMs was effectively used to achieve high site-selectivity. These
processes can be used for the fabrication of various metal oxide devices under
environment-friendly conditions.
We also developed a self-assembly process of particles to fabricate desired patterns of
colloidal crystals. A micropattern of colloidal methanol prepared on a SAM in hexane was
used as a mold for particle patterning, and slow dissolution of methanol into hexane
caused shrinkage of molds to form micropatterns of close-packed particle assemblies. This
result is a step toward the realization of nano/micro periodic structures for next-generation
photonic devices by a self-assembly process.
Furthermore, metal oxides were synthesized in aqueous solutions to form anisotropic
nanostructures. Stand-alone ZnO self-assembled films were, for instance, prepared using
air-liquid interfaces. The ZnO films had sufficiently high strength to free-stand-alone and
showed high c-axis orientation. The films can be pasted onto desired substrates. ZnO
particles having a hexagonal cylinder shape, long ellipse shape or hexagonal symmetry
radial whiskers were also prepared in aqueous solutions. The morphology was controlled
by changing the supersaturation degree. Anatase TiO2 particles with high surface area of
270 m2/g were prepared at 50°C. The particles were assemblies of nano TiO2 crystals covered
with nanorelief surface structures. The crystals grew anisotropically along the c-axis to form
2 Nanocrystals

acicular crystals. TiO2 films consisted of anisotropic acicular crystals were also prepared. The
films showed high c-axis orientation. Acicular BaTiO3 particles were prepared using
morphology control of BaC2O4 • 0.5H2O. They were prepared in aqueous solutions and
annealed with co-precipitated amorphous phase to form acicular BaTiO3 particles.
In this chapter, we will mainly focus on “liquid phase morphology control of metal oxides
nanocrystals” of ZnO1, TiO22 and BaTiO3 particles3, and “liquid phase site-selective
deposition of metal oxide nanocrystals” of TiO24 and Eu:Y2O35.

2. Liquid Phase Morphology Control of Metal Oxide Nanocrystals

2.1. Morphology Control of Stand-alone ZnO Self-assembled Film1
Stand-alone ZnO films were fabricated using aqueous solutions. The films were assemblies
of sheet shaped nanocrystals. They had gradient structure, high c-axis orientation, high
surface area and unique morphology. They were formed at the top of solutions without
substrate. Air-liquid interface was used as a template in this process.
Zinc nitrate hexahydrate (Zn(NO3)2 6H2O) (15 mM) was dissolved in distilled water at
60°C. Ethylenediamine (H2NCH2CH2NH2) (15 mM) was added to the solution to induce the
formation of ZnO1. The solution was kept at 60°C using a water bath for 6 h with no
stirring. The solution was then left to cool for 42 h in the bath. Polyethylene terephthalate
(PET) film, glass (S-1225, Matsunami Glass Ind., Ltd.) and an Si wafer (p-type Si [100], NK
Platz Co., Ltd.) were used as substrates.
The solution color was changed from transparent to white after the addition of
ethylenediamine. ZnO particles were formed by the homogeneous nucleation and growth.
The solution became transparent again after 6h. The supersaturation degree of the solution
was high at the initial stage of the reaction for the first 1 h and decreased as the color of the
solution changed. Ethylenediamine accelerated deposition of ZnO. Zinc-ethylenediamine
complex forms in the solution as shown by eq. 16.
Zn2+ + 3H2NCH2CH2NH2 [Zn(H2NCH2CH2NH2)3]2+ (1)
The chemical equilibrium in eq. 1 moves to the left and the zinc-ethylenediamine complex
decomposes to increase the concentration of Zn2+ at elevated temperature.
OH- concentration increases by the hydrolysis of ethylenediamine as shown by eq. 2.
H2NCH2CH2NH2 + 2H2O H3NCH2CH2NH32+ + 2OH- (2)
ZnO and Zn(OH)2 are thus formed in the aqueous solution as shown by eq. 3.
Zn2+ + 2OH- Zn(OH)2 ZnO + H2O (3)
Films were formed at the top of the solution. Air-liquid interface was used as template. The
films had sufficiently high strength to be obtained as stand-alone films. Additionally, a film was
scooped to past onto a desired substrate such PET film, Si wafer, glass plate or paper, and the
pasted ZnO film was then dried to bond it to the substrate. Both sides of the film can be pasted
on substrate. The film physically adhered to the substrate. The film maintained its adhesion
during immersion in lightly ultrasonicated water, however, it can be easily peeled off again by
strong ultrasonication. The film can be handled easily from substrate to other substrate. It also
Morphology control, self-assembly and site-selective deposition of metal oxide nanocrystals 3

can be attached strongly to substrate by annealing or addition of chemical regents such as silane
coupling agent to form chemical bonds between the film and the substrate.
The film grew to a thickness of about 5 µm after 48 h, i.e., 60°C for 6 h, and was left to cool
for 42 h.
The air side of the stand-alone film had a smooth surface over a wide area due to the flat
air-liquid interface (Fig. 1-a1), whereas the liquid side of the film had a rough surface (Fig.
1-b1). The films consisted of ZnO nano-sheets were clearly observed from the liquid side
(Fig. 1-b2) and the fracture edge-on profile of the film (Fig. 1-c1, 1-c2). The nano-sheets had
a thickness of 5-10 nm and were 1-5 µm in size. They mainly grew forward to the bottom of
the solution, i.e., perpendicular to the air-liquid interface, such that the sheets stood
perpendicular to the air-liquid interface. Thus, the liquid side of the film had many ultra-
fine spaces surrounded by nano-sheet and had a high specific surface area. The air side of
the film, on the other hand, had a flat surface that followed the flat shape of the air-liquid
interface. The air-liquid interface was thus effectively utilized to form the flat surface of the
film. This flatness would contribute to the strong adhesion strength to substrates for
pasting of the film. The air-side surface prepared for 48 h had holes of 100-500 nm in
diameter (Fig. 1-a2), and were hexagonal, rounded hexagonal or round in shape. The air-
side surface prepared for 6 h, in contrast, had no holes on the surface. The air-side surface
was well crystallized to form a dense surface and ZnO crystals would partially grow to a
hexagonal shape because of the hexagonal crystal structure. Well-crystallized ZnO
hexagons were then etched to form holes on the surface by decrease in pH. The growth face
of the film would be liquid side. ZnO nano-sheets would grow to form a large ZnO film by
Zn ion supply from the aqueous solution. Further investigation of the formation
mechanism would contribute to the development of crystallography in the solution system
and the creation of novel ZnO fine structures.
The film showed a very strong 0002 x-ray diffraction peak of hexagonal ZnO at 2θ = 34.04° and
weak 0004 diffraction peak at 2θ = 72.16° with no other diffractions of ZnO (Fig. 2). (0002) planes
and (0004) planes were perpendicular to the c-axis, and the diffraction peak only from (0002)
and (0004) planes indicates high c-axis orientation of ZnO film. The inset figure shows that the
crystal structure of hexagonal ZnO stands on a substrate to make the c-axis perpendicular to the
substrate. Crystallite size parallel to (0002) planes was estimated from the half-maximum full-
width of the 0002 peak to 43 nm. This is similar to the threshold limit value of our XRD
equipment and thus the crystallite size parallel to (0002) planes is estimated to be greater than or
equal to 43 nm. Diffraction peaks from a silicon substrate were observed at 2θ = 68.9° and 2θ =
32.43°. Weak diffractions at 2θ = 12.5°, 24.0°, 27.6°, 30.5°, 32.4° and 57.6° were assigned to co-
precipitated zinc carbonate hydroxide (Zn5(CO3)2(OH)6, JCPDS No. 19-1458).
Stand-alone ZnO film was further evaluated by TEM and electron diffraction. The film was
crushed to sheets and dispersed in an acetone. The sheets at the air-liquid interface were
skimmed by a cupper grid with a carbon supporting film. The sheets were shown to have
uniform thickness (Fig. 2a). They were dense polycrystalline films constructed of ZnO
nanoparticles (Fig. 2b). Lattice image was clearly observed to show high crystallinity of the
particles. The film was shown to be single phase of ZnO by electron diffraction pattern.
These observations were consistent with XRD and SEM evaluations.
4 Nanocrystals

Fig. 1. SEM micrographs of high c-axis oriented stand-alone ZnO self-assembled film. (a1)
Air-side surface of ZnO film. (a2) Magnified area of (a1). (b1) Liquid-side surface of ZnO
film. (b2) Magnified area of (b2). (c1) Fracture cross section of ZnO film from air side. (c2)
Magnified area of (c1).
Morphology control, self-assembly and site-selective deposition of metal oxide nanocrystals 5

Fig. 2. XRD diffraction pattern of high c-axis oriented stand-alone ZnO self-assembled film.
(a) TEM micrograph of ZnO nano-sheets. (b) Magnified area of (a). (Insertion) Electron
diffraction pattern of ZnO.
6 Nanocrystals

The film pasted on a silicon wafer was annealed at 500°C for 1 h in air to evaluate the
details of the films. ZnO film maintained its structure during the annealing (Fig. 3). The air
side of the film showed a smooth surface (Fig. 3-a1) and the liquid side showed a relief
structure having a high specific surface area (Fig. 3-b1, 3-b2). The air side showed the film
consisted of dense packing of small ZnO nanosheets and the size of sheets increased
toward the liquid-side surface (Fig. 3-a2). ZnO sheets would grow from the air side to the
liquid side, i.e., the sheets would nucleate at the liquid-air interface and grow down toward
the bottom of the solution by the supply of Zn ions from the solution. Annealed film
showed X-ray diffractions of ZnO and Si substrate with no additional phases. As-deposited
ZnO nano-sheets were shown to be crystalline ZnO because the sheets maintained their
fine structure during the annealing without any phase transition. High c-axis orientation
was also maintained during the annealing, showing a very strong 0002 diffraction peak.

Fig. 3. SEM micrographs of high c-axis oriented stand-alone ZnO self-assembled film
annealed at 500°C for 1 h in air. (a1) Fracture edge-on profile of ZnO film from air side. (a2)
Cross-section profile of ZnO film from air side. (b1) Fracture edge-on profile of ZnO film
from liquid side. (b2) Cross-section profile of ZnO film from liquid side.
Morphology control, self-assembly and site-selective deposition of metal oxide nanocrystals 7

The solution was further kept at 25°C for 1 month to evaluate the details of the deposition
mechanism. The film prepared at the air-liquid interface for 1 month was not hexagonal
ZnO. The film showed strong X-ray diffractions of zinc carbonate hydroxide single phase.
ZnO would be dissolved by decrease in pH. ZnO would be crystallized at the initial
reaction stage for the first 48 h. ZnO was then gradually etched and dissolved by nitric acid
and zinc carbonate hydroxide was crystallized using Zn ions which were supplied by the
dissolution of crystalline ZnO.
In summary, nano-sheet assembled stand-alone ZnO film was successfully fabricated using
a simple solution process. Air-liquid interface was used as a template to form the films. The
film had high c-axis orientation and showed a strong 0002 diffraction peak and weak 0004
peak. The air side of the film had a flat surface, whereas the liquid side had a rough surface
having many ultra-fine spaces surrounded by ZnO nano-sheets. The rough surface of the
liquid side was suitable for sensors or dye-sensitized solar cells. The film was also pasted
on a desired substrate such as PET films, Si substrate or glass plates. The surface of low
heat-resistant flexible polymer film was modified with high c-axis oriented crystalline ZnO
film without heat treatment. This low-cost, low-temperature technique can be used for a
wide range of applications including sensors, solar cells, electrical devices and optical
devices using the various properties of high c-axis oriented crystalline ZnO.

2.2 Morphology Control of Nanocrystal Assembled TiO2 Particles2

TiO2 particles were prepared in aqueous solutions at ordinary temperature. The particles
were assemblies of nanocrystals that had acicular shape. They had high surface area of 270
m2/g and unique morphology. They are candidate material for dye-sensitized solar cells
and photo catalyst.
Ammonium hexafluorotitanate (12.372 g) and boric acid (11.1852 g) were dissolved in
deionized water (600 mL) at 50°C2. Concentration of them were 0.15 and 0.05 M,
respectively. The solution was kept at 50°C for 30 min using a water bath with no stirring.
The solution was centrifuged at 4000 rpm for 10 min (Model 8920, Kubota Corp.).
Precipitated particles were dried at 60°C for 12 h after removal of supernatant solution.
The solution became clouded about 10 min after mixing ammonium hexafluorotitanate
solution and boric acid solution. The particles were homogeneously nucleated in the
solution, turning the solution white.
X-ray diffraction analysis indicated that the particles were single phase of anatase TiO2. The
peaks were observed at 2θ = 25.1, 37.9, 47.6, 54.2, 62.4, 69.3, 75.1, 82.5 and 94.0°. They were
assigned to the 101, 004, 200, 105 + 211, 204, 116 + 220, 215, 303 + 224 + 312 and 305 + 321
diffraction peaks of anatase TiO2 (JCPSD No. 21-1272, ICSD No. 9852) (Fig. 4).
The 004 diffraction intensity of randomly oriented particles is usually 0.2 times the 101
diffraction intensity as shown in JCPDS data (No. 21-1272). However, the 004 diffraction
intensity of the particles deposited in our process was 0.36 times the 101 diffraction
intensity. Additionally, the integral intensity of the 004 diffraction was 0.18 times the 101
diffraction intensity, indicating the c-axis orientation of the particles. Particles were not
oriented on the glass holder for XRD measurement. Therefore, TiO2 crystals would be an
anisotropic shape in which the crystals were elongated along the c-axis. The crystals would
8 Nanocrystals

have a large number of stacks of c planes such as (001) planes compared to stacks of (101)
planes. The diffraction intensity from the (004) planes would be enhanced compared to that
from the (101) planes.

Fig. 4. XRD diffraction pattern of anatase TiO2 particles. (a): TEM micrograph of anatase
TiO2 particles. (b): Magnified area of (a) showing morphology of acicular crystals. Insertion
in (b): FFT image of (b) anatase TiO2. (c): Magnified area of (a) showing lattice images of
anatase TiO2.

Crystallite size perpendicular to the (101) or (004) planes was estimated from the full-width
half-maximum of the 101 or 004 peak to be 3.9 nm or 6.3 nm, respectively. Elongation of
crystals in the c-axis direction was also suggested by the difference in crystallite size.
The particles were shown to be assemblies of nano TiO2 crystals (Fig. 4a). Particle diameter
was estimated to be 100–200 nm. Relief structures had formed on the surfaces and open
pores had formed inside because the particles were porous assemblies of nanocrystals.
Morphology control, self-assembly and site-selective deposition of metal oxide nanocrystals 9

Nanocrystals were shown to have acicular shapes (Fig. 4b). They were about 5–10 nm in
length. The longer direction of acicular TiO2 is indicated by the black arrow. The inserted FFT
image shows the 101 and 004 diffractions of anatase TiO2. Nanocrystals are assigned to the
single phase of anatase TiO2. It is notable that the diffraction from the (101) planes has a ring
shape due to random orientation but that from the (004) planes was observed only in the
upper right region and lower left region in the FFT image. Anisotropic 004 diffractions
indicated the direction of the c-axis, which was perpendicular to the (004) planes, as shown by
the white arrow. It was roughly parallel to the longer direction of acicular TiO2. These results
suggest that acicular TiO2 grew along the c-axis to enhance the diffraction intensity from the
(004) planes. Crystal growth of anatase TiO2 along the c-axis was previously observed in TiO2
films7. Anisotropic crystal growth is one of the features of liquid phase crystal deposition.
Acicular nanocrystals showed lattice images of anatase TiO2 (Fig. 4c). They were constructed of
anatase TiO2 crystals without amorphous or additional phases. Anatase crystals were not
covered with amorphous or additional phases even at the tips. Bare anatase crystal with
nanosized structure is important to achieve high performance for catalysts and devices.
Crystallization of TiO2 was effectively utilized to form assemblies of acicular nanocrystals in the
process. Open pores and surface relief structures were successfully formed on the particles.
The dried particles were dispersed in water to evaluate zeta potential and particle size
distribution after evaluation of N2 adsorption. The particles had positive zeta potential of 30.2
mV at pH 3.1, which decreased to 5.0, −0.6, −11.3 and −36.3 mV at pH 5.0, 7.0, 9.0 and 11.1,
respectively. The isoelectric point was estimated to be pH 6.7, slightly higher than that of
anatase TiO2 (pH 2.7–6.0)8. Zeta potential is very sensitive to the particle surface conditions,
ions adsorbed on the particle surfaces, and the kind and concentration of ions in the solution.
The variations in zeta potential were likely caused by the difference in the surface conditions
of TiO2 particles, affected by the interaction between particles and ions in the solution.
Mean particle size was estimated to be ~550 nm in diameter with a standard deviation (STD) of
220 nm at pH 3.1. This was larger than that observed by TEM. Slight aggregation occurred at pH
3 because the particles were dried completely prior to measurement. Particle size increased with
pH and showed a maximum of near the isoelectric point (550 nm at pH 3.1, 3150 nm at pH 5,
4300 nm at pH 7, 5500 nm at pH 9 or 2400 nm at pH 11.1). Strong aggregation resulted from the
lack of repulsion force between particles near the isoelectric point.
The particles were generated in the solution at pH 3.8 in this study. It would be suitable to
obtain repulsion force between particles for crystallization without strong aggregation.
TiO2 particles exhibited N2 adsorption-desorption isotherms of Type IV (Fig. 5a). The
desorption isotherm differed from adsorption isotherm in the relative pressure (P/P0)
range from 0.4 to 0.7, showing mesopores in the particles. BET surface area of the particles
was estimated to be 270 m2/g (Fig. 5b). This is higher than that of TiO2 nanoparticles such
as Aeroxide P25 (BET 50 m2/g, 21 nm in diameter, anatase 80% + rutile 20%, Degussa),
Aeroxide P90 (BET 90–100 m2/g, 14 nm in diameter, anatase 90% + rutile 10%, Degussa),
MT-01 (BET 60 m2/g, 10 nm in diameter, rutile, Tayca Corp.) and Altair TiNano (BET 50
m2/g, 30–50 nm in diameter, Altair Nanotechnologies Inc.)9. A high BET surface area
cannot be obtained from particles having a smooth surface even if the particle size is less
10 Nanocrystals

than 100 nm. A high BET surface area would be realized by the unique morphology of TiO2
particles constructed of nanocrystal assemblies.
Total pore volume and average pore diameter were estimated from pores smaller than 230
nm at P/Po = 0.99–0.431 cc/g and 6.4 nm, respectively. They were estimated to be 0.212
cc/g and 3.1 nm, respectively, from pores smaller than 11 nm at P/Po = 0.80. Total pore
volume was also estimated by the BJH method from pores smaller than 154 nm to be 0.428
cc/g. Average pore diameter was estimated to be 6.3 nm using BET surface area.
Pore size distribution was calculated by the BJH method using adsorption isotherms (Fig.
5c). It showed a pore size distribution curve having a peak at ~2.8 nm and pores larger than
10 nm. TiO2 particles would have mesopores of ~2.8 nm surrounded by nanocrystals. Pores
larger than 10 nm are considered to be interparticle spaces. The pore size distribution also
suggested the existence of micropores smaller than 1 nm.
Pore size distribution was further calculated by the DFT/Monte-Carlo method. The model
was in fair agreement with adsorption isotherms (Fig. 5d). Pore size distribution showed a
peak at ~3.6 nm that indicated the existence of mesopores of ~3.6 nm (Fig. 5e). The pore
size calculated by the DFT/Monte-Carlo method was slightly larger than that calculated
from the BJH method because the latter method is considered to have produced an
underestimation10-12. The pore size distribution also suggested the existence of micropores
of ~1 nm, probably resulting from microspaces surrounded by nanocrystals and the uneven
surface structure of nanocrystals.
The particles were shown to have a large surface area as well as micropores of ~1 nm,
mesopores of ~2.8–3.6 nm and pores larger than 10 nm, by N2 adsorption characteristics.
Assembly of acicular nanocrystals resulted in unique features and high surface area.
TiO2 particles were generated in the solutions at 90°C for 1h using an oil bath with no
stirring for comparison. The solutions became clouded after the addition of boric acid
solutions into ammonium hexafluorotitanate solutions. High temperature accelerated
crystal growth of TiO2. Hydrogen chloride of 0.6 ml was added into the solutions of 200ml
to decrease crystallization speed of TiO2. The pH of the solutions was 2.4 one hour after
mixing the solutions. BET surface area of the particles was estimated to 18 m2/g. This is
much lower than that of the particles prepared at 50°C and slightly lower than that
prepared at 90°C for 8 min in our previous work (44 m2/g)13. Formation of TiO2 was
accelerated at high temperature and it decreased surface area. The particles grew in the
solutions to decrease surface area as function of time. Crystallization of TiO2 was shown to
be strongly affected by growth conditions such as solution temperature and growth time.
In summary, anatase TiO2 particles, 100–200 nm in diameter, were successfully fabricated
in aqueous solution. They were assemblies of nanocrystals 5–10 nm that grew
anisotropically along the c-axis to form acicular shapes. The particles thus had nanorelief
surface structures constructed of acicular crystals. They showed c-axis orientation due to
high-intensity X-ray diffraction from the (004) crystal planes. The particles had a high BET
surface area of 270 m2/g. Total pore volume and average pore diameter were estimated
from pores smaller than 230 nm at P/Po = 0.99–0.43 cc/g and 6.4 nm, respectively. They
were also estimated from pores smaller than 11 nm at P/Po = 0.80–0.21 cc/g and 3.1 nm,
respectively. BJH and DFT/Monte-Carlo analysis of adsorption isotherm indicated the
Morphology control, self-assembly and site-selective deposition of metal oxide nanocrystals 11

existence of pores ~2.8 and ~3.6 nm, respectively. Additionally, the analyses suggested the
existence of micropores of ~1 nm. Crystallization and self-assembly of nano TiO2 were
effectively utilized to fabricate nanocrystal assembled TiO2 particles having high surface
area and nanorelief surface structure.

Fig. 5. (a): N2 adsorption-desorption isotherm of anatase TiO2 particles. (b): BET surface area
of anatase TiO2 particles. (c): Pore size distribution calculated from N2 adsorption data of
anatase TiO2 particles using BJH equation. (d): N2 adsorption-desorption isotherm and
DFT/Monte-Carlo fitting curve of anatase TiO2 particles. (e): Pore size distribution
calculated from N2 adsorption data of anatase TiO2 particles using DFT/Monte-Carlo
equation.
12 Nanocrystals

2.3. Morphology Control of Acicular BaTiO3 Particles3

Acicular BaTiO3 crystals were fabricated using solution processes. Morphology control of
them was realized precise control of crystallization in the solutions.
Oxalic acid (252 mg) was dissolved into isopropyl alcohol (4 ml)3. Butyl titanate monomer
(0.122 ml) was mixed with the oxalic acid solution, and the solution was then mixed with
distilled water (100 ml). The pH of the solution was increased to pH = 7 by adding NaOH
(1 M) and distilled water, while the volume of the solution was adjusted to 150 ml by these
additions. The aqueous solution (50 ml) with barium acetate (39.3 mg) was mixed with the
oxalic acid solution. The mixed solution containing barium acetate (0.77 mM), butyl titanate
monomer (2 mM) and oxalic acid (10 mM) was kept at room temperature for several hours
with no stirring, and the solution gradually became cloudy. Stirring causes the collision of
homogeneously nucleated particles and destruction of large grown particles, and so was
avoided in this process. The size of the precipitate was easily controlled from nanometer
order to micrometer order by changing the growth period. Large particles were grown by
immersion for several hours to evaluate the morphology and crystallinity in detail.
Oxalate ions (C2O42-) react with barium ions (Ba2+) to form barium oxalate (BaC2O4
0.5H2O). BaC2O4 0.5H2O is dissolved in weak acetate acid provided by barium acetate
((CH3COO)2Ba), however, it can be deposited at pH 7 which is adjusted by adding NaOH.
BaC2O4 · 0.5H2O was thus successfully precipitated from the solution.
Acicular particles were homogeneously nucleated and precipitated from the solution (Fig.
6a). They were on average 23 µm (ranging from 19 to 27 µm) in width and 167 µm (ranging
from 144 to 189 µm) in length, giving a high aspect ratio of 7.2. They had sharp edges and
clear crystal faces, indicating high crystallinity. A gel-like solid was also coprecipitated
from the solution as a second phase.
XRD diffraction patterns for the mixture of acicular particles and gel-like solid showed
sharp diffraction peaks of crystalline BaC2O4 · 0.5H2O with no additional phase. Acicular
particles were crystalline BaC2O4 · 0.5H2O and the gel-like solid would be an amorphous
phase.
Fortunately, BaC2O4 · 0.5H2O has a triclinic crystal structure as shown by the model
calculated from structure data 14 (Fig. 6b XRD first step) and thus anisotropic crystal growth
was allowed to proceed to produce an acicular shape. Each crystal face has a different
surface energy and surface nature such as zeta potential and surface groups. Anisotropic
crystal growth is induced by minimizing the total surface energy in ideal crystal growth.
Additionally, site-selective adsorption of ions or molecules on specific crystal faces
suppresses crystal growth perpendicular to the faces and so induces anisotropic crystal
growth. These factors would cause anisotropic crystal growth of BaC2O4 · 0.5H2O and
hence allow us to control morphology and fabricate acicular BaC2O4 · 0.5H2O particles. The
positions of diffraction peaks corresponded with that of JCPDS No. 20-0134 (Fig. 6b XRD
third step) and that calculated from crystal structure data 14 (Fig. 6b XRD second step),
however, several diffraction peaks, especially 320 and 201, were enhanced strongly
compared to their relative intensity. The enhancement of diffraction intensity from specific
crystal faces would be related to anisotropic crystal growth; a large crystal size in a specific
Morphology control, self-assembly and site-selective deposition of metal oxide nanocrystals 13

crystal orientation increases the x-ray diffraction intensity for the crystal face perpendicular
to the crystal orientation.
EDX elemental analysis indicated the chemical ratio of the precipitate, which included
acicular particles and gel-like solid, to be about Ba / Ti = 1 to 1.5. The chemical ratio
indicated that the coprecipitated amorphous gel contained Ti ions. Additional Ba ions can
be transformed into BaCO3 by annealing and removed by HCl treatment in the next step.
The ratio was thus controlled to slightly above Ba / Ti = 1 by adjusting the volume ratio of
acicular particles and gel-like solid. Consequently, acicular particles of crystalline BaC2O4 ·
0.5H2O with Ti-containing gel-like solid were successfully fabricated in an aqueous solution
process.
In comparison, isotropic particles of barium titanyl oxalate (BaTiO(C2O4)2 · 4H2O) were
precipitated at pH 2. TiOC2O4 was formed by the following reaction in which the reaction
of oxalic acid (H2C2O4 · 2H2O) with butyl titanate monomer ((C4H9O)4Ti) and hydrolysis
can take place simultaneously 15.
(C4H9O)4Ti + H2C2O4 · 2H2O → TiOC2O4 + 4C4H9OH＋H2O..................(a)
TiO(C2O4) was then converted to oxalotitanic acid (H2TiO(C2O4)2) by the reaction:
TiO(C2O4) + H2C2O4・2H2O → H2TiO(C2O4)2 + 2H2O…………..(b)
Alcoholic solution containing oxalotitanic acid (H2TiO(C2O4)2) formed by reaction (b) was
subjected to the following cation exchange reaction by rapidly adding an aqueous solution
of barium acetate at room temperature:
H2TiO(C2O4)2 + Ba(CH3COO)2 → BaTiO(C2O4)2↓ + 2CH3COOH (c)
BaTiO(C2O4)2 isotropic particles were formed by reaction (c).
On the other hand, neither BaC2O4 · 0.5H2O nor BaTiO(C2O4)2 was precipitated at pH 3 to
pH 6. Gel-like solid was formed in the solution and their XRD spectra showed no
diffraction peaks. The amorphous gel that precipitated at pH = 3 to 6 would be the same as
the amorphous gel coprecipitated at pH 7.
These comparisons show that the crystal growth and morphology control of BaC2O4 ·
0.5H2O are sensitive to the solution conditions.
The precipitate was annealed at 750 °C for 5 h in air. Acicular BaC2O4 · 0.5H2O particles
were reacted with Ti-containing amorphous gel to introduce Ti ions to transform into
crystalline BaTiO3. X-ray diffraction of the annealed precipitate showed crystalline BaTiO3
and an additional barium carbonate phase (BaCO3). Excess precipitation of BaC2O4 · 0.5H2O
caused the generation of barium carbonate phase (BaCO3) as expected.
The annealed precipitate was further immersed in HCl solution (1 M) to dissolve barium
carbonate (BaCO3). Acicular particles of crystalline BaTiO3 were successfully fabricated
with no additional phase. Particles showed acicular shape with 2.8×10×50 µm and x-ray
diffraction of single-phase crystalline BaTiO3 (Fig. 6c). The high aspect ratio of the particles
(17.8 = 50 / 2.8) would be provided by that of BaC2O4 · 0.5H2O particles. The particle size of
acicular BaTiO3 can be easily controlled by the growth period and solution concentration
for BaC2O4 · 0.5H2O precipitation which decides the particle size of BaC2O4 · 0.5H2O.
14 Nanocrystals

Fig. 6. (a) Conceptual process for fabricating acicular BaTiO3 particles. Morphology control
of BaC2O4 · 0.5H2O particles and phase transition to BaTiO3. (b) SEM micrograph and XRD
diffraction pattern of acicular BaC2O4 · 0.5H2O particles precipitated from an aqueous
solution at pH = 7. XRD diffraction measurement data (first step), XRD pattern calculated
from crystal structure data16 (second step) and XRD pattern of JCPDS No. 20-134 (third step)
are shown for triclinic BaC2O4 · 0.5H2O. (c) SEM micrograph and XRD diffraction pattern of
acicular BaTiO3 particles after annealing at 750 °C for 5 h and HCl treatment. XRD
diffraction measurement data (first step) and XRD pattern of JCPDS No. 05-0626 (second
step) are shown for tetragonal BaTiO3.
Morphology control, self-assembly and site-selective deposition of metal oxide nanocrystals 15

BaTiO3 has a cubic crystal structure at high temperature above phase transition and has a
tetragonal crystal structure at room temperature. The cubic crystal structure is completely
isotropic and the tetragonal crystal structure results from stretching a cubic lattice along
one of its lattice vectors. For both of the crystal structures it is difficult to control anisotropic
crystal growth, however, with our newly developed process we could successfully control
the morphology and fabricate acicular particles. This was achieved by controlling the
morphology of triclinic BaC2O4 · 0.5H2O to acicular shape and the phase transition to
BaTiO3 by introducing Ti ions from the coprecipitated amorphous phase. The novel concept
can be applied to a wide variety of morphology control and crystal growth control for
advanced electronic devices composed of crystalline materials.
In summary, a novel process to fabricate acicular BaTiO3 particles was developed.
Morphology control of crystalline BaC2O4 · 0.5H2O to acicular shape was realized in an
aqueous solution. The particles were then transformed into crystalline BaTiO3 by
introducing Ti ions from the coprecipitated amorphous gel phase during the annealing
process. Consequently, acicular particles of tetragonal BaTiO3 were produced by combining
several key technologies. Morphology control in this system has high scientific value for
crystal growth, and acicular particles of crystalline BaTiO3 may have a great impact on
ultra-thin MLCC in future.

3. Liquid Phase Site-selective Deposition of Metal Oxide Nanocrystals

3.1. Site-selective Deposition of Anatase TiO24
Site-selective deposition of anatase TiO2 nanocrystals was achieved in aqueous solutions.
Nucleation and crystal growth of TiO2 were accelerated on super hydrophilic surfaces. It
allowed us to form micro-patterns of TiO2 nanocrystals.
Transparent conductive substrate of F doped SnO2 (FTO, SnO2: F, Asahi Glass Co., Ltd., 9.3-
9.7 Ω/�, 26 × 50 × 1.1 mm) was blown by air to remove dust and was exposed to ultraviolet
light (low-pressure mercury lamp PL16-110, air flow, 100 V, 200 W, SEN Lights Co.) for 10
min through a photomask (Test-chart-No.1-N type, quartz substrate, 1.524 mm thickness,
Toppan Printing Co., Ltd.) (Fig. 7)4. The initial SnO2: F substrate showed a water contact
angle of 96°. The UV-irradiated surface was, however, wetted completely (contact angle 0–
1°). The contact angle decreased with irradiation time (96°, 70°, 54°, 35°, 14°, 5° and 0° for 0
min, 0.5 min, 1 min, 2 min, 3 min, 4 min and 5 min, respectively). This suggests that a small
amount of adsorbed molecules on the SnO2: F substrate was removed completely by UV
irradiation. The surface of the SnO2: F substrate would be covered by hydrophilic OH
groups after irradiation. Consequently, the SnO2: F substrate was modified to have a
patterned surface with hydrophobic regions and super-hydrophilic regions.
Ammonium hexafluorotitanate ([NH4]2TiF6) (2.0620 g) and boric acid (H3BO3) (1.8642 g)
were separately dissolved in deionized water (100 mL) at 50°C. Boric acid solution was
added to ammonium hexafluorotitanate solution at concentrations of 0.15 M and 0.05 M,
respectively. The SnO2: F substrate having a patterned surface with hydrophobic regions
and super-hydrophilic regions was covered by a silicon rubber sponge sheet (Silico-sheet,
SR-SG-S 5mmt RA grade, Shin-etsu Finetech Co., Ltd.) to suppress deposition of TiO2 at the
initial stage. The substrate was immersed perpendicularly in the middle of the solution
16 Nanocrystals

(Fig. 7). The solution was kept at 50°C with no stirring. The silicon rubber sponge sheet was
removed from the SnO2: F substrate after 25 h, then the substrate was kept for a further 2 h
at 50°C. The substrate was covered by the sheet instead of immersion of substrate at 25h to
avoid agitation of the solution.
The solution became clouded in about 10 min after the mixing of ammonium
hexafluorotitanate solution and boric acid solution. The particles were homogeneously
nucleated in the solution and made the solution white. They then gradually precipitated
and fell to the bottom of the vessel, so the solution became transparent over a period of
hours. Ti ions were consumed for crystallization of TiO2 particles, thus decreasing the ion
concentration in the solution. The super-saturation degree of the solution was sufficiently
low to realize slow heterogeneous nucleation without homogeneous nucleation which
forms TiO2 particles. The silicon rubber sponge sheet was removed from the SnO2: F
substrate after 25 h, then the substrate was kept for a further 2 h at 50°C. Consequently, the
patterned surface on the SnO2: F substrate was exposed to the transparent solution
including Ti ions at a low concentration for 2 h. Heterogeneous nucleation and slow
crystallization of TiO2 progressed only on the substrate.
Deposition of anatase TiO2 proceeds by the following mechanisms7:

Equation (a) is described in detail by the following two equations:

Fluorinated titanium complex ions gradually change into titanium hydroxide complex ions
in an aqueous solution as shown in Eq. (c). The increase of F- concentration displaces Eqs.
(a) and (c) to the left, however, the produced F- can be scavenged by H3BO3 (BO33-) as
shown in Eq. (b) to displace Eqs. (a) and (c) to the right. Anatase TiO2 formed from titanium
hydroxide complex ions (Ti(OH)62-) in Eq. (d).
Liquid phase patterning was not realized in the initial solution but realized in the solution
after 25 h. Solution condition was evaluated as function of time to clarify this reason. The
solution was transparent immediately after the mixing of ammonium hexafluorotitanate
solution and boric acid solution, became clouded after 0.5 h and showed maximum
whiteness after 1 h. Anatase TiO2 particles nucleated homogeneously in the solution and
grew to form large particles, which gradually precipitated and made the bottom of the
vessel white. The solution became slightly white after 5 h and transparent after 25 h. The
solutions changed to transparent by the filtrations. Precipitated particles from the residual
solution and particles from the supernatant solution trapped by filters were determined by
XRD evaluation to be a single phase of anatase TiO2.
Morphology control, self-assembly and site-selective deposition of metal oxide nanocrystals 17

Precipitated particles from the residual solution increased as 0 mg, 165.1 mg, 230.9 mg and
424.4 mg at 0 h, 2 h, 5 h and 25 h, respectively. The precipitation increased rapidly at the
initial stage and moderately after 2 h, reflecting the decrease of crystal growth rate.
The weight of particles > 2.5 µm in diameter was estimated to be 2.8 mg, 49.7 mg, 9.5 mg, 0
mg and 0 mg at 0.5 h, 1 h, 2 h, 5 h and 25 h, respectively. Particles were formed at the initial
stage and precipitated, making the bottom of the vessel white. This is consistent with the
color change of the solution.
The precipitate from the filtrate collected by all of the filters was evaluated by XRD. The
white powder contained anatase TiO2, rutile TiO2 and a large amount of boric acid. These
were crystallized from ions in the filtrate during drying. The weight was estimated to be
1455 mg, 1429 mg, 1392 mg, 1345 mg and 1341 mg at 0.5 h, 1 h, 2 h, 5 h and 25 h,
respectively. This indicated that the solution contained a high concentration of ions at the
initial stage, which then decreased as a function of time. Ion concentration would decrease
by the crystallization and precipitation of anatase TiO2. This result is consistent with the
weight variation of precipitated particles, weight variation of suspended particles and
solution color change shown in the photographs.
Liquid phase patterning was not realized in the initial clouded solution but realized in the
transparent solution after 25 h. Evaluation of solution condition as function of time showed
the reason of this phenomenon. TiO2 particles formed at the initial stage around 1 h and
precipitated gradually. Ions were consumed for crystallization of TiO2 and decreased as a
function of time. Heterogeneous nucleation predominantly progressed after 5 h.
Consequently, TiO2 was formed on super hydrophilic regions selectively to realize liquid
phase patterning.
FTO substrate was immersed in the solution for 25 h to form a thick film and ultrasonicated
in water for 20 min. TiO2 film was constructed of two layers. Under layer with 200 nm
thickness was a polycrystalline film of anatase TiO2. Upper layer with 300 nm thickness
was an assembly of acicular TiO2 crystals which grew perpendicular to the substrate. The
film was shown by electron diffraction pattern to be a single phase of anatase TiO2. Electron
diffraction from the 004 plane was stronger than that of the 101, 200, 211 planes, etc to show
anisotropic crystal growth along the c-axis. Additionally, 004 diffractions were strong
perpendicular to the substrate, showing that the c-axis orientation of acicular crystals was
perpendicular to the substrate. The FTO layer was shown to be a single phase of SnO2 with
high crystallinity. Acicular TiO2 crystals had a long shape, being ~ 300 nm in length and 10
– 100 nm in diameter. A lattice image of anatase TiO2 was observed from the crystals.
The film deposited on the substrate was evaluated by XRD analysis. Strong X-ray diffractions
were observed for films deposited on FTO substrates and assigned to SnO2 of FTO films. The
004 diffraction peak of anatase TiO2 was not observed clearly for TiO2 film on FTO substrates
because both of the weak 004 diffraction peak of TiO2 and the strong diffraction peak of FTO
were observed at the same angle. Glass substrates with no FTO coating were immersed in the
solution. Weak X-ray diffraction peaks were observed at 2θ = 25.3, 37.7, 48.0, 53.9, 55.1 and
62.7° for the films deposited on glass substrates. They were assigned to 101, 004, 200, 105, 211
and 204 diffraction peaks of anatase TiO2 (ICSD No. 9852) (Fig. 7). A broad diffraction peak
from the glass substrate was also observed at about 2θ = 25°.
18 Nanocrystals

The intensity of the 004 diffraction peak was stronger than that of the 101 diffraction peak
for the film obtained by the liquid phase crystal deposition method, though the intensity of
101 was stronger than that of 004 for anatase TiO2 powders with no orientation (ICSD No.
9852). The integral intensity or peak height of 004 was 2.6 times or 2.2 times that of 101,
respectively, suggesting high c-axis orientation of anatase TiO2 crystals. Crystallite size
perpendicular to the 101 or 004 planes was estimated from the full-width half-maximum of
the 101 or 004 peak to be 9 nm or 17 nm, respectively. Elongation of crystals in the c-axis
direction was also suggested by the difference of crystallite size. These evaluations were
consistent with high c-axis orientation observed by TEM and electron diffraction.
Crystallite size estimated by XRD was similar to that in TiO2 under layer rather than that of
acicular crystals observed by TEM. TiO2 thin film prepared on a glass would be constructed
of not acicular crystals but polycrystals in under layer.
After having been immersed in the solution, the substrate was rinsed with distilled water and
dried in air (Fig. 7). The initial FTO surface appeared to be blue-green under white light due
to light diffracted from the FTO layer. On the other hand, TiO2 films deposited on the super-
hydrophilic surface appeared to be yellow-green. The color change would be caused by
deposition of transparent TiO2 film which influenced the wavelength of the diffracted light.

Fig. 7. Conceptual process for liquid phase patterning of anatase TiO2 films using super-
hydrophilic surface. XRD diffraction pattern of anatase TiO2 film on a glass substrate.
Morphology control, self-assembly and site-selective deposition of metal oxide nanocrystals 19

The micropattern of TiO2 was shown by SEM evaluation to be successfully fabricated (Fig.
8(Top)). TiO2 deposited on super-hydrophilic regions showed black contrast, while the
initial FTO regions without deposition showed white contrast in Fig. 8(top). The average
line width in Fig. 8(top) is 55 μm. Line edge roughness16, as measured by the standard
deviation of the line width, is ~2.8 μm. This represents a ~5% variation (i.e., 2.8/55) in the
nominal line width, similar to the usual 5% variation afforded by current electronics design
rules. The minimum line width of the pattern depends on the resolution of the photomask
and wavelength of irradiated light (184.9 nm). It would be improved to ~ 1 μm by using a
high-resolution photomask.
The FTO layer was a particulate film having a rough surface (Fig. 8-b1, b2). Edged particles of
100 – 500 nm in diameter were observed on the surface. The micropattern of TiO2 thin film
was covered by an assembly of nano crystals of 10 – 30 nm in diameter (Fig. 8-a1, a2). The
nano crystals would be anatase TiO2 which grew anisotropically. The TiO2 film also had large
structural relief of 100 – 500 nm in diameter. As the thin TiO2 film was deposited on the edged
particulate surface of the FTO layer, the surface of TiO2 had large structural relief.
The morphology of the TiO2 layer and FTO layer was further observed by fracture cross
section profiles (Fig. 9). The polycrystalline FTO layer prepared on a flat glass substrate was
shown to have a thickness of ~ 900 nm, and a high roughness of 100 – 200 nm on the
surface (Fig. 9a). Nano TiO2 crystals were deposited on the super-hydrophilic FTO surface
(Fig. 9a), whereas no deposition was observed on the initial FTO surface. The super-
hydrophilic FTO surface was covered with an array of nano TiO2 crystals (Fig. 9b, c), which
had a long shape of ~ 150 nm in length and ~ 20 nm in diameter. These observations were
consistent with TEM and XRD evaluations. Nano TiO2 crystals would grow along the c-axis
and thus enhance the 004 X-ray diffraction peak and 004 electron diffraction peak. They
formed a long shape having a high aspect ratio of 7.5 (150 nm in length / 20 nm in
diameter) as shown in the SEM fracture cross section profile (Fig. 9b, c) and TEM
micrograph. The orientation of nano TiO2 crystals with their long axis perpendicular to the
FTO layer (Fig. 9b, c) would also enhance the 004 diffraction peak.
In summary, a micropattern of anatase TiO2 thin film was successfully fabricated on an
SnO2: F substrate in an aqueous solution. Crystalline anatase TiO2 was deposited by liquid
phase crystal deposition at 50°C. Nucleation and crystal growth of TiO2 were accelerated on
the super-hydrophilic SnO2: F surface, but were suppressed on the hydrophobic initial
SnO2: F surface. Consequently, liquid phase patterning of anatase TiO2 was achieved on an
SnO2: F substrate. TiO2 crystals were directly deposited on the SnO2: F surface without any
insulating layers which decrease the electrical conductivity between TiO2 and the SnO2: F
substrate. The micropattern of anatase TiO2 on the SnO2: F substrates could be applied to
electrodes of dye-sensitized solar cells or molecular sensors. Additionally, this process can
be used to form a flexible micropattern of anatase TiO2 electrodes on low-heat-resistant
conductive polymer films. This process will contribute to the microfabrication of TiO2
electrodes for dye-sensitized solar cells or molecular sensors.
20 Nanocrystals

Fig. 8. SEM micrographs of a micropattern of anatase TiO2 films on SnO2: F substrates. (Top)
Micropattern of anatase TiO2 films. (a1) Surface of anatase TiO2 films deposited on super-
hydrophilic region. TiO2 was formed on super-hydrophilic region which was cleaned by UV
irradiation before the immersion. (a2) Magnified area of (a1) showing surface morphology
of anatase TiO2 film. (b1) Surface of SnO2: F substrate without TiO2 deposition. TiO2 was not
formed on non-cleaned region. (b2) Magnified area of (b1) showing surface morphology of
SnO2: F substrate.
Morphology control, self-assembly and site-selective deposition of metal oxide nanocrystals 21

Fig. 9. SEM micrographs of anatase TiO2 films on SnO2: F substrates. (a) Fracture cross
section of TiO2 films. (b, c) Magnified area of (a) showing morphology of nano TiO2 crystals.
22 Nanocrystals

5
3.2. Site-selective Deposition of Eu:Y2O3
Micropatterning of Eu doped Y2O3 nanocrystals was also realized using site-selective
deposition in the solutions. Self-assembled monolayer was utilized as template in this system.
Si substrate (p-type [100], 1–50 Ωcm, Newwingo Co., Ltd.) was cleaned ultrasonically in
acetone, ethanol and deionized water for 5 min, respectively, in this order and was exposed
to ultraviolet light and ozone gas for 10 min to remove organic contamination by using a
UV/ozone cleaner (184.9 nm and 253.7 nm) (low-pressure mercury lamp 200 W, PL21-200,
SEN Lights Co., 18 mW/cm2, distance from lamp 30 mm, 24 °C, humidity 73%, air flow 0.52
m3/min, 100 V, 320 W)17-20. APTS (3-Aminopropyltriethoxysilane)-SAM was prepared by
immersing the Si substrate in an anhydrous toluene solution containing 1 vol% APTS for 1
h in N2 atmosphere. The substrate was rinsed with a fresh anhydrous toluene in N2
atmosphere. The substrate with SAM was baked at 120 °C for 5 min to remove residual
solvent and promote chemisorption of the SAM.
APTS-SAM was then irradiated by ultraviolet light (PL21-200) through a photomask (Test-chart-
No.1-N type, quartz substrate, 1.524 mm thickness, guaranteed line width 2 μm ± 0.5 μm,
Toppan Printing Co., Ltd.) for 10 min. UV irradiation modified an amino-terminated silane to a
silanol forming a pattern of amino-terminated silane regions and silanol regions17-20. Patterned
APTS-SAM having amino regions and silanol regions was used as a template for patterning of
yttrium oxide. Initially deposited APTS-SAM showed water contact angles of 48°. The UV-
irradiated surface of SAM was, however, wetted completely (contact angle <5°). This suggests
that SAM of APTS was modified to hydrophilic OH group surfaces by UV irradiation.
The patterned APTS-SAM was immersed in an aqueous solution containing Y(NO3)3 · 6H2O (4
mM), Eu(NO3)3 · 6H2O (0.4 mM) and NH2CONH2 (50 mM) at 25 °C5. The solution was heated to
77 °C gradually as shown in Fig. 10 since urea (NH2CONH2) decomposes to form ammonium
ions (NH4+) above 70 °C (Eq. (a)). The decomposition of urea at elevated temperature plays an
essential role in the deposition of yttrium oxide. The aqueous solution of urea yields ammonium
ions and cyanate ions (OCN-) at temperatures above 70 °C21 (Eq. (a)). Cyanate ions react rapidly
according to Eq. (b). Yttrium ions are weakly hydrolyzed22,23 in water to YOH(H2O)n2+ (Eq. (c)).
The resulting release of protons (H+) and/or hydronium ions (H3O+) accelerates urea
decomposition (Eq. (b)). The precipitation of the amorphous basic yttrium carbonate (Y(OH)CO3
・xH2O, x=1) can take place through the reaction in Eq. (d)24,25. The controlled release of cyanate
ions by urea decomposition causes deposition of basic yttrium carbonate once the critical
supersaturation in terms of reacting component is achieved. Since the decomposition of urea is
quite slow, the amount needed to reach supersaturation within a given period of time must be
considerably higher than the stoichiometric amount of yttrium ions, as revealed by previous
studies of lanthanide compounds26.
The temperature of the solution increased gradually and reached 77 °C in about 80 min5. The
solution was kept at ~ 77 °C during deposition. The pH of the solution increased from 5.2 to
5.8 in about 90 min and then gradually decreased to 5.6. Temperature and pH increased for
the initial 90 min and became stable after 90 min. The average size of particles
homogeneously nucleated in the solution at 100 min was about 227 nm and increased to 262
nm at 150 min, 282 nm at 180 min, 310 nm at 210 min, and 323 nm at 240 min (Fig. 10a).
Particles nucleated and grew after the solution temperature exceeded 70 °C because urea
Morphology control, self-assembly and site-selective deposition of metal oxide nanocrystals 23

decomposes above 70 °C to form carbonate ions21 which causes deposition of basic yttrium
carbonate23-26. The particles grew rapidly at the beginning of the growth period and then their
growth rate decreased exponentially (Fig. 10a). The decrease in growth rate was caused by the
decrease of supersaturation degree influenced by a decrease in solution concentration.

Fig. 10. Conceptual process for site-selective deposition of visible-light emitting Y2O3:Eu thin
films using a self-assembled monolayer. (a) Time variation of particle size distribution. (1-5)
Particle size distribution of yttrium carbonate particles at (1) 100 min, (2) 150 min, (3) 180
min, (4) 210 min or (5) 240 min.
24 Nanocrystals

Yttrium carbonate films were observed to deposit on amino regions of a patterned SAM
after the immersion in an aqueous solution (Fig. 11A, B). Deposits showed white contrast,
while silanol regions without deposition showed black contrast in SEM observation.
Narrow lines of depositions having 10–50 µm width were successfully fabricated in an
aqueous solution. Patterned APTS-SAM showed high ability for site-selective deposition of
yttrium carbonate in solution systems.

Fig. 11. (A) SEM micrograph of patterned Y2O3:Eu thin films and (B) magnified area of (A).
(a) SEM micrograph of patterned Y2O3:Eu thin films. Characteristic X-ray images of Y, O, C
and Si for (a) Y2O3:Eu thin films. (Top right) Elemental analysis of Y2O3:Eu thin films
deposited for 90 min.

Yttrium carbonate films were also deposited on the hydrophobic octadecyl surface of
OTS(octadecyltrichlorosilane)-SAM having water contact angle (WCA) of 116 ° and as-
purchased silicon wafer having WCA of about 20-50 ° which was kept in a plastic case in
air. On the other hand, the films were not deposited on UV irradiated silicon wafer having
WCA < 5 °. The super hydrophilic surface of WCA < 5 ° suppressed film deposition,
whereas the hydrophobic surface and medium surface of WCA > 20-30 ° accelerated film
deposition possibly because of hydrophobic interaction between deposition and substrate
surface. This is consistent with a former study32. Yttrium carbonate was deposited both on
bare single crystal Si wafers, and on Si wafers coated with sulfonate-functionalized organic
self-assembled monolayers.
Yttrium, europium, oxygen and carbon were observed from as-deposited thin films on
amino regions, while silicon and oxygen were detected from non-covered silanol regions by
Morphology control, self-assembly and site-selective deposition of metal oxide nanocrystals 25

EDX (Fig. 11). The molecular ratio of yttrium to europium was determined to be 100 : 8
(Fig. 11). It was close to that of Y(NO3)3 · 6H2O to Eu(NO3)3 · 6H2O, i.e., 100 : 10, in the
solution because the chemistry of Eu(NO3)3 is similar to that of Y(NO3)3 to incorporate
europium in the precipitation. The content of europium was in the range we had expected.
Y2O3:Eu with atomic ratio Y : Eu = 100 : ~ 8 was reported to have strong
photoluminescence27,28. Carbon was detected from yttrium carbonate. Silicon and oxygen
were detected from silicon wafer covered with a natural oxide layer (amorphous SiO2).
Amino regions were covered with thin films composed of many large particles (about 100–
300 nm in diameter) and very high roughness (RMS 25.6 nm) (Fig. 12A). Silanol regions, on
the other hand, showed only nano-sized small particles (about 10–50 nm in diameter) and
very low roughness (RMS 1.7 nm). The high site-selectivity of deposition and the big
difference in surface morphology and roughness were clearly shown by AFM observation.
The thickness of the films was estimated from AFM scans across deposited and
undeposited regions of the substrate. It increased with immersion time after 45 min (0 nm
at 45 min, 60 nm at 70 min and 100 nm at 90 min (Fig. 12A)). The average growth rate (70
nm/h = 100 / 90 min) was higher than that previously reported (2 nm/h = 35 nm / 15 h)25.
An amorphous yttrium basic carbonate film was deposited at 80 °C from aqueous solutions
of YNO3•5H2O and urea on Si wafers coated with sulfonate-functionalized organic self-
assembled monolayers in previous studies. The thickness was then evaluated by TEM after
the treatment with ultrasonication for half an hour in distilled water. The difference of
growth rate was caused mainly by the difference of the substrate treatment by
ultrasonication. Additionally, the thickness of our film was smaller than the particle size in
the solution shown in Fig. 10 (227 nm at 100 min). Heterogeneous nucleation and
attachment of initial particles of yttrium carbonate occurred without the attachment of
aggregated large particles shown in Fig. 10. The yttrium carbonate was then grown on the
substrate to form a film of 100 nm thickness after immersion for 90 min. The particles of
about 100 nm in height were removed by ultrasonication for 30 min and the film of several
nm in height remained as reported25.
Yttrium was not detected by XPS from the substrate immersed for 45 min, however, it was
clearly observed from that immersed for 90 min. This indicates that the deposition began
between 45 and 90 min after immersion. The solution temperature reached 70 °C in ~ 45
min and then the solution began to decompose and release carbonate ions, causing the
deposition of basic yttrium carbonate. The deposition mechanism evaluated by XPS is
consistent with the change of solution temperature, decomposition temperature of urea and
chemical reaction of this system. The binding energy of Y 3d5/2 spectrum from the
deposition (158.2 eV) was higher than that of metal yttrium (155.8 eV)29. The spectrum
shifted to lower binding energy (156.7 eV) after annealing at 800 °C in air for 1 h and is
similar to that of Y2O3 (157.0 eV)30. The binding energies of Y 3d5/2 spectra in as-deposited
films and annealed films were higher than that of metal yttrium possibly due to the
chemical bonds formed between yttrium ions and oxygen ions. The chemical shift of Y
3d5/2 binding energy by annealing is consistent with crystallization of as-deposited films
to crystalline Y2O3. C 1s spectra were detected at 289.7 eV and 284.6 eV from as-deposited
films. The C 1s spectrum at 289.7 eV then disappeared by the annealing. C 1s at 284.6 eV
was assigned to surface contamination and C 1s at 289.7 eV was detected from as-deposited
26 Nanocrystals

yttrium carbonate. The disappearance of C 1s at 289.7 eV is consistent with the phase

transition from yttrium carbonate to Y2O3.
As-deposited film was shown to be an amorphous phase (Fig. 12B-a) by XRD
measurement. The film showed no diffraction peak after annealing at 400 °C for 1 h,
however, it showed 222, 400 and 440 diffraction peaks of crystalline cubic Y2O331 without
any additional phase after annealing at 600 °C for 1 h and the intensities of diffraction
peaks increased further by annealing at 800 °C for 1 h (Fig. 12B-b). The film was shown to
be a polycrystalline Y2O3 film constructed from randomly deposited Y2O3 particles without
crystal-axis orientation. The crystal structure model and diffraction pattern of Y2O3 were
calculated from the crystal structure data of ICSD #23811 as shown in Fig. 12B. The
crystallization by annealing confirmed from XRD measurement is consistent with XPS
evaluation.
We attempted to remove Y2O3 films from the silicon substrate by debonding with scotch
tape or by ultrasonication for 5 min in water. However, the films maintained their bonds
with the substrate, indicating that strong adhesion had formed between films and
substrate.
The thin film annealed at 800 °C for 1 h, i.e., crystalline Y2O3:Eu thin film, was shown to be
excited by 230–250 nm (center: 243 nm) and emit red light photoluminescence centered at
611 nm in the fluorescence excitation spectrum (Fig. 12C-a). Neither the as-deposited film
nor the film annealed at 400 °C for 1 h showed photoluminescence, on the other hand, the
films annealed at 600 °C or 800 °C for 1 h emitted light centered at 617 nm by 250 nm in
fluorescence emission spectra (Fig. 12C-b). The fluorescence intensity of the film annealed
at 800 °C was stronger than that of the film annealed at 600 °C. Fluorescence intensity
increased by the phase transformation from amorphous yttrium carbonate to yttrium oxide
and crystal growth by the heat treatments, and is consistent with the crystallization
observed by XRD32. The spectra are described by the well-known 5D0–7FJ line emissions (J =
0, 1, 2, …) of the Eu3+ ion with the strongest emission for J = 2 at 612 nm. The thin film
annealed at 800 °C produced visible red light photoluminescence by excitation from Nd:
YAG laser (266 nm) (Fig. 12C, inset). The white square shows the edges of the Y2O3:Eu thin
film and the red color shows visible red emission from the irradiated area on the substrate.
In summary, we have proposed a novel process for fabricating visible red light emitting Eu-
doped Y2O3 and its micropattern using a self-assembled monolayer and an aqueous
solution system. The patterned APTS-SAM having amino groups regions and silanol
groups regions achieved site-selective deposition of yttrium oxide in an aqueous solution.
The deposited films were crystallized by annealing at 600 °C or 800 °C for 1 h. Crystalline
Y2O3:Eu produced visible red light photoluminescence centered at 611 nm by excitation
from Nd: YAG laser (266 nm). This study showed the high potential of aqueous solution
systems and self-assembled monolayers for the fabrication of functional metal oxide thin
films and their micropatterns.
Morphology control, self-assembly and site-selective deposition of metal oxide nanocrystals 27

Fig. 12. (A) (a) AFM images and cross-section profile of Y2O3:Eu thin films on NH2 groups
regions. (b) AFM images and cross-section profile of Y2O3:Eu thin films on OH groups
regions. (B) XRD patterns of Y2O3:Eu thin films (a) before and (b) after annealing at 800 °C
for 1 h. (The upper picture) Crystal structure model and diffraction pattern of cubic Y 2O3
calculated from crystal structure data of ICSD #23811. (C) (a) Fluorescence excitation
spectrum (emission: 611 nm) for Y2O3:Eu thin film after annealing at 800 °C for 1 h. (b)
Fluorescence emission spectra (excitation: 250 nm) for Y2O3:Eu thin films before and after
annealing at 400, 600 or 800 °C for 1 h. Inset: Photoluminescence image for Y2O3:Eu thin film
annealed at 800 °C for 1 h (excitation: 266 nm).
28 Nanocrystals

4. Summary
Morphology control, self-assembly and site-selective deposition of metal oxide nanocrystals
were reported in this chapter. The novel nano/micro structures of nanocrystals will create
the next generation of metal oxide devices. Additionally, they were realized in solutions at
ordinary temperature and atmospheric pressure. These environmental-friendly processes
will contribute to green innovations.

5. Acknowledgement
The author thanks Prof. Kunihito Koumoto of Nagoya University, Dr. Tatsuki Ohji and Dr.
Kazumi Kato of AIST, Dr. Masako Ajimi, Dr. Makoto Bekki and Dr. Shuji Sonezaki of
TOTO Ltd. Research Laboratory. Some of works that were reviewed in this chapter were
partially supported by METI, Japan, as part of R&D for High Sensitivity Environment
Sensor Components.

6. References
(1) Masuda, Y.; Kato, K. Cryst. Growth Des. 2008, 8, 275-279.
(2) Masuda, Y.; Kato, K. Cryst. Growth Des. 2008, 8, 3213-3218.
(3) Masuda, Y.; Yamada, T.; Koumoto, K. Cryst. Growth Des. 2008, 8, 169-171.
(4) Masuda, Y.; Kato, K. Chem. Mater. 2008, 20, 1057-1063.
(5) Masuda, Y.; Yamagishi, M.; Koumoto, K. Chem. Mater. 2007, 19, 1002-1008.
(6) Gao, M. D.; Li, M. M.; Yu, W. D. J. Phys. Chem. B 2005, 109, 1155-1161.
(7) Masuda, Y.; Sugiyama, T.; Seo, W. S.; Koumoto, K. Chem. Mater. 2003, 15, 2469-2476.
(8) Furlong, D. N.; Parfitt, G. D. J. Colloid Interface Sci. 1978, 65, 548-554.
(9) Wahi, R. K.; Liu, Y. P.; Falkner, J. C.; Colvin, V. L. J. Colloid Interface Sci. 2006, 302, 530-
536.
(10) Kruk, M.; Jaroniec, M. Chem. Mater. 2001, 13, 3169-3183.
(11) Ravikovitch, P. I.; Odomhnaill, S. C.; Neimark, A. V.; Schuth, F.; Unger, K. K. Langmuir
1995, 11, 4765-4772.
(12) Lastoskie, C.; Gubbins, K. E.; Quirke, N. J. Phys. Chem. 1993, 97, 4786-4796.
(13) Katagiri, K.; Ohno, K.; Masuda, Y.; Koumoto, K. J. Ceram. Soc. Japan 2007, 115, 831-834.
(14) Mutin, J. C.; Dusausoy, Y.; Protas, J. J. Solid State Chem. 1981, 36, 356-364.
(15) Potdar, H. S.; Deshpande, S. B.; Date, S. K. J. Am. Ceram. Soc. 1996, 79, 2795-2797.
(16) Masuda, Y.; Sugiyama, T.; Lin, H.; Seo, W. S.; Koumoto, K. Thin Solid Films 2001, 382,
153-157.
(17) Dressick, W. J.; Calvert, J. M. Jpn. J. Appl. Phys. 1993, 32, 5829-5839.
(18) Collins, R. J.; Shin, H.; De Guire, M. R.; Heuer, A. H.; Sukenik, C. N. Appl. Phys. Lett.
1996, 69, 860-862.
(19) Masuda, Y.; Koumura, T.; Okawa, T.; Koumoto, K. J. Colloid Interface Sci. 2003, 263, 190-
195.
(20) Masuda, Y.; Itoh, M.; Yonezawa, T.; Koumoto, K. Langmuir 2002, 18, 4155-4159.
(21) Shaw, W. H. R.; Bordeaux, J. J. J. Am. Chem. Soc. 1955, 77, 4729-4733.
(22) Ryabchikov, D. E.; Ryabukhin, V. A. Analytical Chemistry of Yttrium and the Lanthanide
Elements; Humphrey Science: Ann Arbor, MI, 1970.
Morphology control, self-assembly and site-selective deposition of metal oxide nanocrystals 29

(23) Baes, C. F.; Mesmer, R. E. The Hydrolysis of Captions; Wiley: New York, 1976.
(24) Aiken, B.; Hsu, W. P.; Matijevic, E. J. Am. Ceram. Soc. 1988, 71, 845-853.
(25) Agarwal, M.; DeGuire, M. R.; Heuer, A. H. Appl. Phys Lett. 1997, 71, 891-893.
(26) Matijevic, E.; Hsu, W. P. J. Colloid Interface Sci. 1987, 118, 506-523.
(27) Sharma, P. K.; Jilavi, M. H.; Nass, R.; Schmidt, H. J. Lumin. 1999, 82, 187-193.
(28) Kwaka, M. G.; Parkb, J. H.; Shon, S. H. Solid State Commun. 2004, 130, 199-201.
(29) Fuggle, J. C.; Martensson, N. J. Electron Spectrosc. Relat. Phenom. 1980, 21, 275.
(30) Wagner, C. D. Practical Surface Analysis; 2 ed.; John Wiley, 1990; Vol. 1.
(31) Paton, M. G.; Maslen, E. N. Acta Crystallographica 1967, 1, 1948-23.

Figure Captions
Figure 1. SEM micrographs of high c-axis oriented stand-alone ZnO self-assembled film.
(a1) Air-side surface of ZnO film. (a2) Magnified area of (a1). (b1) Liquid-side surface of
ZnO film. (b2) Magnified area of (b2). (c1) Fracture cross section of ZnO film from air side.
(c2) Magnified area of (c1). [Reprinted with permission from Ref.1, Y. Masuda and K. Kato,
Cryst. Growth Des. 8, 1, 275, 2008. Copyright @American Chemical Society (2008)]
Figure 2. XRD diffraction pattern of high c-axis oriented stand-alone ZnO self-assembled
film. (a) TEM micrograph of ZnO nano-sheets. (b) Magnified area of (a). (Insertion) Electron
diffraction pattern of ZnO. [Reprinted with permission from Ref.1, Y. Masuda and K. Kato,
Cryst. Growth Des. 8, 1, 275, 2008. Copyright @American Chemical Society (2008)]
Figure 3. SEM micrographs of high c-axis oriented stand-alone ZnO self-assembled film
annealed at 500°C for 1 h in air. (a1) Fracture edge-on profile of ZnO film from air side. (a2)
Cross-section profile of ZnO film from air side. (b1) Fracture edge-on profile of ZnO film
from liquid side. (b2) Cross-section profile of ZnO film from liquid side. [Reprinted with
permission from Ref.1, Y. Masuda and K. Kato, Cryst. Growth Des. 8, 1, 275, 2008.
Copyright @American Chemical Society (2008)]

Figure 4. XRD diffraction pattern of anatase TiO2 particles. (a): TEM micrograph of anatase
TiO2 particles. (b): Magnified area of (a) showing morphology of acicular crystals. Insertion
in (b): FFT image of (b) anatase TiO2. (c): Magnified area of (a) showing lattice images of
anatase TiO2. [Reprinted with permission from Ref.2, Y. Masuda and K. Kato, Crystal
Growth & Design, 8, 9, 3213, 2008. Copyright @American Chemical Society (2008)]
Figure 5. (a): N2 adsorption-desorption isotherm of anatase TiO2 particles. (b): BET surface
area of anatase TiO2 particles. (c): Pore size distribution calculated from N2 adsorption data
of anatase TiO2 particles using BJH equation. (d): N2 adsorption-desorption isotherm and
DFT/Monte-Carlo fitting curve of anatase TiO2 particles. (e): Pore size distribution
calculated from N2 adsorption data of anatase TiO2 particles using DFT/Monte-Carlo
equation. [Reprinted with permission from Ref.2, Y. Masuda and K. Kato, Crystal Growth &
Design, 8, 9, 3213, 2008. Copyright @American Chemical Society (2008)]

Figure 6. (a) Conceptual process for fabricating acicular BaTiO3 particles. Morphology
control of BaC2O4 • 0.5H2O particles and phase transition to BaTiO3. (b) SEM micrograph
and XRD diffraction pattern of acicular BaC2O4 • 0.5H2O particles precipitated from an
aqueous solution at pH = 7. XRD diffraction measurement data (first step), XRD pattern
calculated from crystal structure data16 (second step) and XRD pattern of JCPDS No. 20-
30 Nanocrystals

134 (third step) are shown for triclinic BaC2O4 • 0.5H2O. (c) SEM micrograph and XRD
diffraction pattern of acicular BaTiO3 particles after annealing at 750 °C for 5 h and HCl
treatment. XRD diffraction measurement data (first step) and XRD pattern of JCPDS No. 05-
0626 (second step) are shown for tetragonal BaTiO3. [Reprinted with permission from Ref. 3,
Y. Masuda, T. Yamada and K. Koumoto, Cryst. Growth Des. 8, 169, 2008. Copyright
@American Chemical Society (2008)]

Figure 7. Conceptual process for liquid phase patterning of anatase TiO2 films using super-
hydrophilic surface. XRD diffraction pattern of anatase TiO2 film on a glass substrate.
[Reprinted with permission from Ref.4, Y. Masuda and K. Kato, Chem. Mater., 20, 3, 1057,
2008. Copyright @American Chemical Society (2008)]
Figure 8. SEM micrographs of a micropattern of anatase TiO2 films on SnO2: F substrates.
(Top) Micropattern of anatase TiO2 films. (a1) Surface of anatase TiO2 films deposited on
super-hydrophilic region. TiO2 was formed on super-hydrophilic region which was cleaned
by UV irradiation before the immersion. (a2) Magnified area of (a1) showing surface
morphology of anatase TiO2 film. (b1) Surface of SnO2: F substrate without TiO2 deposition.
TiO2 was not formed on non-cleaned region. (b2) Magnified area of (b1) showing surface
morphology of SnO2: F substrate. [Reprinted with permission from Ref.4, Y. Masuda and K.
Kato, Chem. Mater., 20, 3, 1057, 2008. Copyright @American Chemical Society (2008)]
Figure 9. SEM micrographs of anatase TiO2 films on SnO2: F substrates. (a) Fracture cross
section of TiO2 films. (b, c) Magnified area of (a) showing morphology of nano TiO2
crystals. [Reprinted with permission from Ref.4, Y. Masuda and K. Kato, Chem. Mater., 20,
3, 1057, 2008. Copyright @American Chemical Society (2008)]

Figure 10. Conceptual process for site-selective deposition of visible-light emitting Y2O3:Eu
thin films using a self-assembled monolayer. (a) Time variation of particle size distribution.
(1-5) Particle size distribution of yttrium carbonate particles at (1) 100 min, (2) 150 min, (3)
180 min, (4) 210 min or (5) 240 min. [Reprinted with permission from Ref.5, Y. Masuda, M.
Yamagishi, K. Koumoto, Chem. Mater., 19, 1002, 2007. Copyright @American Chemical
Society (2007)]
Figure 11. (A) SEM micrograph of patterned Y2O3:Eu thin films and (B) magnified area of
(A). (a) SEM micrograph of patterned Y2O3:Eu thin films. Characteristic X-ray images of Y,
O, C and Si for (a) Y2O3:Eu thin films. (Top right) Elemental analysis of Y2O3:Eu thin films
deposited for 90 min. [Reprinted with permission from Ref.5, Y. Masuda, M. Yamagishi, K.
Koumoto, Chem. Mater., 19, 1002, 2007. Copyright @American Chemical Society (2007)]
Figure 12. (A) (a) AFM images and cross-section profile of Y2O3:Eu thin films on NH2
groups regions. (b) AFM images and cross-section profile of Y2O3:Eu thin films on OH
groups regions. (B) XRD patterns of Y2O3:Eu thin films (a) before and (b) after annealing at
800 °C for 1 h. (The upper picture) Crystal structure model and diffraction pattern of cubic
Y2O3 calculated from crystal structure data of ICSD #23811. (C) (a) Fluorescence excitation
spectrum (emission: 611 nm) for Y2O3:Eu thin film after annealing at 800 °C for 1 h. (b)
Fluorescence emission spectra (excitation: 250 nm) for Y2O3:Eu thin films before and after
annealing at 400, 600 or 800 °C for 1 h. Inset: Photoluminescence image for Y2O3:Eu thin
film annealed at 800 °C for 1 h (excitation: 266 nm). [Reprinted with permission from Ref.5,
Y. Masuda, M. Yamagishi, K. Koumoto, Chem. Mater., 19, 1002, 2007. Copyright
@American Chemical Society (2007)]
Growth of undoped and metal doped ZnO nanaostructures by solution growth 31

X2

Growth of undoped and metal doped ZnO

nanostructures by solution growth
Rathinam Chandramohan, Thirukonda Anandamoorthy Vijayan
and Jagannathan Thirumalai
Department of Physics,Sree Sevugan Annamalai College,Devakottai - 630 303
India

1. Introduction
ZnO is one of the most studied materials of the II-VI oxide materials that derive continuous
attention of the researchers worldwide since forties (Bunn 1935). Because of its current and
possible applications in several novel devices, renewed interest has emerged and several
reviews (Liu et al, 2005; Tsukazaki et al, 2005), and conference proceedings are published
exclusively for ZnO nano crystallites similar systems at Singapore (2005), (2009) and
Changchan, China (2006) to explore the feasibility of commercial application for future
devices. Yet the ream of novel devices from this wonderful material is yet to be
accomplished in full (Wellings, et al, 2008). With a wide band gap of 3.2 eV and a large
exciton binding energy of 60 meV at room temperature, ZnO, line GaN, will be important
for blue and ultraviolent optical devices. ZnO has several advantages over GaN in this
applications range however, the most important being its longer exciton binding energy and
the ability to grow single crystal substrates. Other favourable aspects of ZnO include its
broad chemistry leading to many opportunities for wet chemical etching, low power
threshold for optical pumping, radiation hardness and biocompatibility. Together, these
properties of ZnO make it an ideal candidate for a variety of devices ranging from sensors
through to ultra-violet laser devices and nanotechnology based devices such as displays. As
fervent research into ZnO continues, difficulties such as the fabrication of p-type ZnO that
have so far stated that the development of devices had over come (Yang etal, 2008). To give
a quantitative report on the state of art of ZnO nanocrystals is quite difficult and an attempt
has been made to survey the chemical growth of this system in this study. The chemical
solution growth of ZnO nano thin films composed of nano crystallites using a two step
double dip chemical deposition method has been discussed In detail in this chapter. The
growth and characterization of nano structures of ZnO has been reported by Wang (Wang,
2004).
Mitra et al (1998) has prepared Zinc Oxide thin films using chemical deposition technique.
The structural, morphological properties of the prepared films are characterized using X-ray
diffraction and scanning electron microscope. They have used Zn salts as precursor and
successfully synthesized ZnO films. The growth of highly textured Zinc oxide (ZnO) thin
films with a preferred (101) orientation has been prepared by employing chemical bath
32 Nanocrystals

deposition using a sodium zincate bath on glass substrates has been reported by
(Ramamoorthy et al, 2004). The films were characterized by XRD, SEM, EDX, UV-Vis-NIR,
FTIR and PL in order to justify the suitability for commercial device quality. (Natsume et al,
2000) have studied the d.c electrical conductivity and optical properties of zinc oxide film
prepared by a sol-gel spin coating technique. The temperature dependence of the
conductivity indicated that electron transport in the conduction band was due to thermal
execution of donor electrons for temperatures from 250 to 300 K. (Chapparro et al, 2003)
have proposed the spontaneous growth of ZnO thin films from aqueous solutions. An
electroless – chemical process is proposed, consisting in the formation of the super oxide
radical (O2-) followed by chemical reaction of two O2- with Zn (NH3)42+ cations. (Wellings et
al, 2008) have deposited ZnO thin films from aqueous zinc nitrate solution at 80C onto
fluorine doped tin oxide (FTO) coated glass substrates. Structural analysis, surface
morphology, optical studies and electrical conductivity were studied and thickness of the
ZnO films was found to be 0.40 m. (Walter Walter et al, 2007) have studied the
characterization of strontium doped ZnO thin films on love wave filter applications. X-ray
diffraction, scanning electron microscopy and atomic force microscopy studied the
crystalline structure and surface morphology of films. The electrochemical coupling
coefficient, dielectric constant, and temperature coefficient of frequency of filters were then
determined using a network analyzer. (Vijayan et al, 2008, a, b) have reported the
preparation conditions for undoped ZnO using double dip technique and used them for gas
sensor applications. They have also reported the synthesis of Sr doped ZnO using double
dip technique and used them for gas sensor applications. Recently (Chandramohan et al,
2010) have synthesized Mg doped ZnO thin films using double dip chemical growth and
reported the ferromagnetic properties of the films. Tahir Saeed et al. (1995) have deposited
thin films of mono phase crystalline hexagonal ZnO from solutions of zinc acetate in the
presence of ethylenediamine and sodium hydroxide on to glass microscope studies. Two
distinct morphologies of ZnO were observed by scanning electron microscopy. The
deposited films were specular and adherent. (Cheng et al 2006) have fabricated thin films
transistors (TFTs) with active channel layers of zinc oxide using a low – temperature
chemical bath deposition. Current voltage (I-V) properties measured through the gate reveal
that the ZnO channel is n-type. (Sadrnezhaad et al 2006) have studied the effect of addition
of Tiron as a surfactant on the microstructure of chemically deposited zinc oxide. Addition
of tiron charges the surface morphology and causes to form the fine – grained structure. The
obtained results indicate that increasing the number of dipping carves to progress the
deposition process. (Piticescu, et al 2007) have studied the influence of the synthesis
parameters on the chemical and microstructural characteristics of nanophases synthesized
in the two methods. ‘Al’ doping tends to a lower material density and to a smaller gown
size. Zhou et al (2007) have studied microstructure electrical and optical properties of
aluminium doped zinc oxide films. The ZnO:Al thin films are transparent ( 90%) in near
ultraviolet and visible region A. with the annealing temperature increasing from 300C to
500C. The film was oriented more preferentially along the (002) direction, the grain size of
the film increased, the transmittance also became higher and the electrical resistivity
decreased. Joseph et al (2006) have reported the structural, electrical and optical properties
of Al-doped ZnO thin films prepared by chemical spray deposition. XRD studies and SEM
studies revealed that the film was polycrystalline in nature with (002) preferred
concentration and smooth conditions have exhibited a resistivity of 2.45 x 10-4 m with an
Growth of undoped and metal doped ZnO nanaostructures by solution growth 33

optical transmittance of 97% of 550 nm. Oral et al (2007) have studied microstructure and
optical properties of monocrystalline ZnO and ZnO : Li /Al thin films. Crystallized films
had a grain size under 50 nm and showed C-axis grain orientation. All films had a very
smooth surface with RMS and surface roughness values between 0.23 and 0.35 nm. Peiro et
al (2005) have reported microwave-activated chemical bath depositions of zinc oxide thin
films. Scanning electron microscopic characterization suggested that both the shape of the
crystals and the textures of the film were highly influenced by the chemical path
composition. Composition of films grown on bone glass or fluorine – doped tin oxide (SnO2:
F) showed that heterogeneous deposition was favoured on conducting substrates due to the
localized heating. Bulk ZnO is quite expensive and unavailable in large wafers. So, for the
time being, thin films of ZnO are relatively a good choice. Usually, the doped ZnO films
with optimum properties (perfect crystalline structure, good conducting properties, high
transparency, high intensity of luminescence) are obtained when they are grown on heated
substrates and annealed after deposition at high temperature in oxygen atmosphere (Peiro
et al, 2005 Lokhande et al, 2000; Srinivasan et al, 2006; Chou et al, 2005). However, for an
extensive use in the commercial applications pure and doped ZnO films must be prepared
at much lower substrate temperatures. Therefore, it is necessary to develop a low-
temperature deposition technology for the growth of ZnO films. Many works are seen in the
low temperature growth of this interesting ZnO system both undoped and metal doped
(Tang et al, 1998, Cracium et al, 1994; Gorla et al, 1999; Kotlyarchuk et al, 2005) thin films
and nano thin films. Advantages are effectiveness and simplicity of the deposition
equipment, high deposition rates, wide spectrum of deposition parameters for the control
and the optimization of film properties, and film thickness. The sum of all these special
features enables the growth of oxide thin films at low temperature substrates with perfect
crystallinity. The present work is a preparation and characterization of undoped ZnO, Sr-
doped ZnO (SZO) and Al- doped ZnO (AZO) thin films by chemical deposition technique.
In which the influence of solution concentration, solution pH value, film thickness,
annealing temperature and concentration of strontium and aluminium atoms of the grown
films are investigated. In addition it demonstrates that any dopant can be used in principle
along with the precursor to enable them to be included in the system. The technique can be
tuned to get the desired morphology and nanocrystallites of desired sizes distributed over
any type of substrate for various applications.

2. Substrate and its preparation

Thin film requires a substrate to support itself. The substrate provides the necessary
mechanical strength and rigidity needed for the film and it has adequate thermal ability to
ensure at room temperature and withstand at high temperature. The function of the
substrate is to provide the base on to which the thin film circuits are fabricated and various
thin film multilayers are deposited. To form the thin film with defined electrical parameters,
the substrates must be smooth and flat otherwise electrical and optical properties may be
affected. Therefore in choosing a suitable substrate, in addition to considering the need to
provide the mechanical support to the deposits, due consideration must be given to the
possible influence of the substrates on the properties of the deposits. Commonly used
substrate materials for polycrystalline thin film circuits include alumina, glass, silicon and
metals, beryllium oxide based ceramic, aluminium nitride. When the films are deposited
34 Nanocrystals

into glass, electrical and optical measurements are not disturbed by an underlying layer and
are thus easier to interpret. Of all these, glass is found to posses all the requirements and is
economically and widely used. However any type of substrate may be used in this simple
growth method. Substrate cleaning in thin film technology is an important step prior to
deposition. It is necessary to remove the contaminants that would otherwise affect the
properties of the film. Cleaning involves the removal of contaminants without damage to
the substrate. While cleaning, the bond between the substrates is broken and contaminants
are set free from the substrates. The properties that can be affected by the presence of
contaminants include morphology, nucleation electronic properties and the substrate film
interface. Expected contaminants include fingerprints, dust, oil, and lint particles. The
proper cleaning technique depends on the nature of the substrate and nature of the
contaminants. The composition, physical properties such as porosity, thermal expansion,
melting point, conductivity and chemistry of the substrate should be carefully considered in
designing the cleaning operation. The energy required to break those bond could be
supported by chemical, salvation, thermal (or) mechanical process. As the other techniques
this technique also involves rigorous cleaning of the substrates.

3. Growth of nanocrystallites of undoped and doped ZnO

Experimental
A schematic diagram of the shape-selective synthesis of doped and undoped metal oxide
nanostructures via double dip technique is shown in Fig. 1. Preparation of undoped and
doped ZnO nano thin films ZnO thin films were performed using a two-step chemical bath
deposition technique using a solution comprising of high purity zinc sulphate, magnesium
sulphate and sodium hydroxide with a pH value of 9 as first step and a dip in hot water
kept near boiling point as the second step. Before deposition, the glass substrates were
cleaned by chromic acid followed by cleaning with acetone. The well-cleaned substrates
were immersed in the chemical bath for a known standardized time followed by immersion
in hot water for the same time for hydrogenation.

Possible formation mechanism

The process of solution dip (step 1) followed by hot water dipping (step 2) is repeated for
known number of times. According to the following equation, the complex layer deposited
on the substrate during the dipping in sodium zincate bath will be decomposed to ZnO due
to subsequent dipping in hot water. The proposed reaction mechanism for undoped ZnO is
according to the following equations

ZnSO4+ 2 NaOH→ Na2ZnO2 + H2SO4 ↑.... (1)

Na2ZnO2 + H2O→ ZnO + 2 NaOH ....... (2)

Part of the ZnO so formed was deposited onto the substrate as a strongly adherent film and
the remainder formed as a precipitate. The addition of Metal sulphate in the ratio of Zn:
Metal as 100:1 in the first dip solution leads to the formation of MZO films.

ZnO thin films were prepared using double dip technique shown in Fig. 1 by varying
deposition parameters such as solvent medium, solution pH, concentrations, temperature,
Growth of undoped and metal doped ZnO nanaostructures by solution growth 35

number of dippings, etc., The effect of these parameters are studied using various
characterizations and the optimized deposition parameters are arrived for undoped ZnO thin
films. The ‘Al ‘and ‘Sr’ doping were carried out by adding the respective metallc salts in the
solution bath at different proportions (Zn : M as 100 : 1 or 10 : 1 where M =’ Sr’ or ‘Al’.

Water bath Sodium

Zincate bath

Manual Dipping
Thermometer
T=90 to 95 C

Glass plate

Distilled
Water

Precipitate

HOT
PLATE
Fig. 1. Schematic representation of solution dip technique

Opimization of growth conditions

Undoped, Strontium doped ZnO (SZO) and Aluminium doped ZnO (AZO) thin films are
prepared by solution grown double dip technique. The films are annealed in air to improve
the crystallinity and grain sizes. All the synthesized films are characterized for their
structural, optical, surface morphology, surface roughness, compositional analysis, X-ray
photoemission spectroscopy analysis and electrical properties. The doping of Strontium and
Aluminium (0.1mM and 1mM) concentration thin films by solution grown double dip
technique is performed and investigated. The salient features of various studies carried out
and the important findings are presented in this chapter. Zinc Oxide and doped (Sr and Al)
thin films find interesting applications in the filed of gas sensor, solar cells, optoelectronic
36 Nanocrystals

devices etc., Aluminium doped Zinc Oxide thin films are more suitable for gas sensor
applications owing to their band gap and stability. This chapter describes widely the
synthesis ZnO, Sr-doped ZnO (SZO) and Al-doped ZnO (AZO) thin films by solution grown
double dip technique from aqueous solutions of ZnSO4. The films prepared are found to be
compact and homogeneous. The deposition conditions are optimized to obtain uniform, thin
films suitable for gas sensor applications. The optimized deposition conditions to prepare
ZnO films are

Bath composition and deposition conditions

Zinc Sulphate : 0.1M

Sodium hydroxide : 0.2M
Solution pH : 9 ± 0.2
Bath temperature : 90ºC

SZO and AZO thin films are prepared at various molarities (0.1mM and 1mM) under
optimized condition to obtain uniform, thin film suitable for gas sensor applications. The
optimized deposition conditions to prepare SZO and AZO thin films are

Bath composition and deposition conditions

Strontium doped Zinc oxide (SZO) thin films

Zinc sulphate : 0.1M

Sodium hydroxide : 0.2M
Strontium Sulphate : 0.1mM and 1mM
Solution pH : 9 ± 0.2
Hot water temperature : 90º C

Aluminium doped Zinc Oxide (AZO) thin films

Zinc sulphate : 0.1M

Sodium hydroxide : 0.2M
Aluminium Sulphate : 0.1mM and 1mM
Solution pH : 9 ± 0.2
Hot water temperature : 90º C

The growth conditions have been optimized by us for various dopants including non
metals. Several parameters involved in this technique offer wide range of selection of
parameters. It is found that the films deposited at room temperature are found to be smooth
and uniform and compatible with any physical or chemical techniques.

4. The structure and morphology of the nanostructures

To support the discussion on the optimization of growth conditions the SEM studies were
carried out. The studies reveal that by altering the deposition conditions morphologies with
minor variations can be obtained. Figure 2-4 shows the SEM micrographs of ZnO grown
using double dip technique where 40, 80 and 100 respectively.
Growth of undoped and metal doped ZnO nanaostructures by solution growth 37

It is observed that improving the number of dipping yields the better morphology films. It is
observed that the film is quite uniform up to a thickness in the range of few microns with
the variation in the range of 10 nm. While the grains of the film surfaces are uniformly in the
average range of 300 nm, the surface seems to be formed by the stacking of different
nanorods or cylindrical grains whose size varied from 20 to 500 nm.

To support the discussion on the optimization of growth conditons the AFM studies were
carried out. The studies reveal that by altering the deposition conditions morphologies with
minor varitions can be obtained. Figures 5 (a - c) represents 3D AFM micrographs obtained
for ZnO samples grown at room temperature under optimized conditions. It is observed
that the stacking of different nanorods or tubular grains whose size varied from 20 to 500
nm makes the samples topography. Adjusting the deposition parameters may control the
size of the grains. It is interesting that the morphology when further explored using 2D and
3D analysis supported by SEM investigations that morphology is also due to the nearly
spherical nano grains.

Adjusting the deposition parameters gives a control on the growth of nanocrystallites with
various sizes and shapes. It is interesting to know that the dip rate, interval between
successive dippings, the variations in second dip bath, etc produces significant changes in
the morphology inviting more number of researchers in this low cost nanocrystallites
growth. For doping with other metals respective salts may be replaced in the place of
aluminum sulphate. This method of growth facilitates fabricating excellent structures for
future devices at low cost and low temperature. The technique is found to be highly
reproducible and can be extended to large area and large scale fabrication systems.

In summary, the synthesis and optimization of undoped and doped ZnO systems have been
reported. The Morphological studies through AFM and SEM reveal’s excellent features
associated with nanocrystallites of which the structure is made. The SEM reveals
continuously stacked nanorods of diameter ranging from 20nm to few hundred nanometers.
The AFM studies reveal the surface to be of minimum roughness composed of spherical and
hexagonal shaped grains. They are uniformly distributed throughout the surface exhibiting
the superiority of the films. Extensive characterizations on the structure, microstructure
optical and electrical properties have been made and the exotic choice available in this
simple method has paved way for the synthesis of many similar systems by our group like
Fe, Mg and Mn doped ZnO thin films and other TCO systems like CdO, etc. Also the
properties of these thin film nanocrystallites can be tailored to suit variety of applications
like, phosphors, display panels, thermal conduction and opto electronic devices. The
technique is easy for automation and anticorrosive coatings can be coated employing doped
ZnO systems on to various mechanical spares. The potential of this technique is yet to be
exploited in full by the industrial community. The crystallite shape and size control is also
feasible in this excellent method.
38 Nanocrystals

(a)

(b)

(c)

Fig. 2. Surface morphology on (a) pH =8, (b) pH =9 and (c) pH=10 of ZnO thin films.
Growth of undoped and metal doped ZnO nanaostructures by solution growth 39

(a)

(b)

Fig. 3. Surface morphology on (a) Al-doped (0.1mM), (b) Al-doped (1mM) ZnO thin films.
40 Nanocrystals

(a)

(b)

Fig. 4. Surface morphology on (a) Sr-doped (0.1mM), (b) Sr-doped (1mM) ZnO thin films.
Growth of undoped and metal doped ZnO nanaostructures by solution growth 41

(a)

(b)

(c)

Fig. 5. (a-c) Shows 3D AFM micrographs obtained for ZnO, SZO and AZO samples grown at
room temperature under optimized conditions.
42 Nanocrystals

5. References
Bunn, C.W. (1935). The lattice-dimensions of zinc oxide. Proc. Phys. Soc. London., 47., 5., 835-
842, ISSN 0959-5309.
Liu, C.; Yun, F.; & Morkoç, H, (2005). Ferromagnetism of ZnO and GaN: A Review J. Mater.
Sci: Mat in Electronics. 16., 9., 555-597, ISSN: 0957-4522.
Tsukazaki, A.; Ortomo, A.; Onuma, T.; Ohtani, M. & Makino, T.; Smiya, M.; Ohtani, K.;
Chichibu, S.F.; Fuke, S.; Egawa, Y.S.; Ohno, H.; Kainuma, K. & Kawasaki, M. (2005).
Repeated temperature modulation epitaxy for p-type doping and light-emitting
diode based on ZnO. Nat. Mater., 4., 1., (December 2004) 42-46, ISSN: 1476-1122.
Özgür, Ü.; Alivov, Ya. I.; Liu, C.; Teke, A.; Reshchikov, M.A.; Doğan, S.; Avrutin, V.; Cho, S.-
J. & Morkoç, H. (2005). A comprehensive review of ZnO materials and devices. J.
Appl. Phys.98., 4., (August 2005) 041301-041404, ISSN 0021-8979.
Wang, Z.L. (2004). Zinc oxide nanostructures: growth, properties and applications. J.Phys.:
Cond. Matter., 16., 25., (June 2004) R829-R858, ISSN 0953-8984.
Symposium N, ICMAT IUMRS 2005 3-8 July 2005, Singapore.
Symposium D, ICMAT 2009, 28 June -3 July, Singapore.
The 6th International Workshop on Zinc Oxide and Related Materials, 5- 7 Aug 2010 in
Changchun, China.
Wellings, J.S.; Chaure, N.B.; Heavens, S.N. & Dhamadasa, I.M. (2008). Growth and
characterisation of electrodeposited ZnO thin films. Thin solid films, 516., 12., (April
2008) 3893-3898, ISSN: 0040-6090.
Yang, Y.; Tay, B.K.; Sun, X.W.; Han, Z.J.; Shen, Z.X.; Lincoln, C.; & Smith, T, (2008).
Nanoelectronics Conference, INEC 2008 2 nd IEEE International, Nanyang
Technical University, Singapore 24-27 March 2008.
Mitra, P.; Chatterjee, A.P. & Maiti, H.S. (1998). Chemical deposition of ZnO films for gas sensors.
J. Mater. Sci: Mat in Electronics., 9., 6., (December & 1998) 441-445, ISSN: 0957-4522.
Ramamoorthy, K.; Arivanandhan, M.; Sankaranarayanan, K. & Sanjeeviraja, C. (2004). Mater.
Chem. Phys., 85., 2-3., (June & 2004) 257-262, ISSN: 0254-0584.
Natsume, Y. & Sakata, H. (2000). Zinc oxide films prepared by sol-gel spin-coating. Thin solid
films., 372., 1-2., (September & 2000) 30-36, ISSN: 0040-6090.
Chapparro, M.; Maffiotte, C.; Gutierrez, M.T. & Herrero, J. (2003). Study of the spontaneous
growth of ZnO thin films from aqueous solutions.Thin solid films., 431., 1., (May &
2003) 373-377, ISSN: 0040-6090.
Water, W. & Yan, Y– S. (2007). Characteristics of strontium-doped ZnO films on love wave
filter applications. Thin solid films., 515., 17., (June & 2007) 6992-6996, ISSN: 0040-6090.
Vijayan, T. A.; Chandramohan, R.; Valanarasu, S.; Thirumalai, J.; Venkateswaran, S.; Mahalingam,
T. & Srikumar, S.R. (2008). Optimization of growth conditions of ZnO nano thin films
by chemical double dip technique. Sci. Tech. Adv. Mater., 9., 035007, ISSN: 1468-6996 .
Vijayan, T. A.; Chandramohan, R.; Valanarasu, S.; Thirumalai, J.; Subramanian, S. P. (2008).
Comparative investigation on nanocrystal structure, optical, and electrical
properties of ZnO and Sr-doped ZnO thin films using chemical bath deposition
method. J. Mater. Sci., 43., 6., (March & 2008) 1776–1782, ISSN: 0022-2461.
Chandramohan, R.; Thirumalai, J.; Vijayan, T. A.; ElhilVizhian, S.; Srikanth, S.; Valanarasu,
S. & Swaminathan, V. (2010). Nanocrystalline Mg Doped ZnO Dilute Magnetic
Semiconductor Prepared by Chemical Route. Adv. Sci. Lett. 3., 3., (September &
2010) 319-322, ISSN: 1936-6612.
Growth of undoped and metal doped ZnO nanaostructures by solution growth 43

Saeed, T. & Brien, P. O’. (1995). Deposition and characterisation of ZnO thin films grown by
chemical bath deposition. Thin solid films., 271., 1-2., (December & 1995) 35-38, ISSN:
0040-6090.
Cheng, H– C.; Chen, C– F.; Lee, C– C. Thin solid films., 498., 1-2., (March & 2006) 1, ISSN:
0040-6090.
Sadrnezhaad, S.K. & Vaezi, M.R. (2006). The effect of addition of Tiron as a surfactant on the
microstructure of chemically deposited zinc oxide. Mat. Sci. Engg: B., 128., 1-3.,
(March & 2006) 53-57, ISSN: 0921-5107.
Piticescu, R. R. Piticescu, R. M. & Monty, C. J. (2006). Synthesis of Al-doped ZnO
nanomaterials with controlled luminescence. J. Europ. Cer. Soc. 26., 14., (March &
2006) 2979-2983, ISSN: 0955-2219.
H–M Zhou, D–Q Yi, Z– M Yu, L. Rang, Xiao, Jian Li, (2007). Preparation of aluminum doped
zinc oxide films and the study of their microstructure, electrical and optical
properties. Thin solid films., 515., 17., (June 2007) 6909-6914, ISSN: 0040-6090.
Joseph, B.; Manoj, P.K. & Vaidyan, V.K. (2006). Studies on the structural, electrical and optical
properties of Al-doped ZnO thin films prepared by chemical spray deposition.
Ceramics International., 32., 5., (September & 2005) 487-493, ISSN: 0272-8842.
Oral, A.Y.; Batisi, Z.B. & Aslan, M.H. (2007). Microstructure and optical properties of
nanocrystalline ZnO and ZnO:(Li or Al) thin films. Appl. Sur. Sci. 253., 10., (March
& 2007) 4593-4598, ISSN: 0169-4332.
Peiro, A. M.; Ayllon, J. A.; Pearl, J.; Domenech, X. & Domingo C. (2005). Microwave
activated chemical bath deposition (MW-CBD) of zinc oxide: Influence of bath
composition and substrate characteristics. J. Crystal. Growth., 285., 1-2., (November
&2005) 6-16, ISSN: 0022-0248.
Lokhande, B.J. & Uplane, M.D. (2000). Structural, optical and electrical studies on spray
deposited highly oriented ZnO films. Appl. Surf. Sci., 167., 3-4., (October & 2000)
243-246, ISSN: 0169-4332.
Srinivasan, G. & Kumar, J. (2006). Optical and structural characterisation of zinc oxide thin
films prepared by sol-gel process. Cryst. Res. Technol. 41., 9., (September & 2006)
893- 896, ISSN: 0232-1300.
Chou, T– L.; Ting, & J– M. (2005). Deposition and characterization of a novel integrated ZnO
nanorods/thin film structure. Thin solid films. 494., 1-2., (January & 2006) 291-295,
ISSN: 0040-6090.
Tang, Z.K.; Wang, G.K. L.; Yu, P.; Kawaraki, M.; Ohtomo, A.; Koinuma, H. & Segawa, Y.
(1998). Appl. Phys. Lett. 72., 3270, ISSN: 0003-6951.
Cracium, V.; Elders, J.; Gardeniers, J.G.E. & Boyd, L. W. (1994). Appl. Phys. Lett., 65., 2963,
ISSN: 0003-6951.
Look, D.C.; Reynolds, D.C.; W.Litton, C.; Jones, R.L.; Eason, D.B & Cartwell, G. (2002). Appl.
Phy. Lett., 81 1830, ISSN: 0003-6951.
Gorla, C.R.; Emanetoglu, N.W.; Liang, S.; Mayo, W.E.; Cu, Y.; Wraback, M. & Shen, H.
(1999). J. Appl. Phys., 85., 2602, ISSN 0021-8979.
Kotlyarchuk, B.; Sarchuk, V. & Oszwaldowski, M. (2005). Preparation of undoped and
indium doped ZnO thin films by pulsed laser deposition method. Cryst. Res.
Technol., 40., 12., (December & 2005) 1118- 1123, ISSN: 0232-1300.
44 Nanocrystals
Synthesis and morphology control of Eu3+ doped M2O2S [M=Y, Gd] nanostructures 45

X3

Synthesis and morphology control of Eu3+

doped M2O2S [M=Y, Gd] nanostructures
Rathinam Chandramohan, Jagannathan Thirumalai
and Thirukonda Anandamoorthy Vijayan
Department of Physics,Sree Sevugan Annamalai College,Devakottai - 630 303
India

1. Introduction

Global claim for phosphor materials as efficient sources of energy that can supply sustained
competence is growing day by day. The phosphors are facing increased global challenges
including high production of rare earth materials, environmental and recycling issues, and
necessity to supply devices very quickly that may be outdated rapidly due to new
technological developments arising in the industry and market. A number of applications
have emerged in recent years that will change the future of the industry and new
technologies like nanoscale innovations and specialty phosphors are garnering increased
attention. The primary drivers for growth are the expansion of key end-use applications
including solid-state lighting and fluorescent lighting. Current research in nanotechnology
is focused on new materials, novel phenomena, new characterization technique and
fabrication of nano devices.
Y2O2S:Eu3+ and Gd2O2S:Eu3+ are excellent materials of current interest (Lei et al., 2010; Liu &
Kuang, 2010; Li, et al., 2010; Nakkiran et al., 2007; Thirumalai et al., 2007) owing to their
interesting optical and opto-electronic properties. The crystal structure of M2O2S (M = Y, Gd
and including all lanthanides) are discussed in detail (Delgado da Vila et al., 1997; Sabot &
Maestro, 1995; Mikami & Oshiyama, 1998). The crystal symmetry of the above two systems
is trigonal, with the space group P3m1 (D33d), as determined by X-ray diffraction. These
systems are grouped under wide band gap (4.6 – 4.8 eV) semiconductors. Y2O2S:Eu3+ and
Gd2O2S:Eu3+ as a red phosphor, with its sharp emission line for good calorimetric definition
and high luminescence efficiency, is extensively used in the phosphor screen of display
devices, fluorescent lamps used for lighting purposes, television sets used for entertainment
and information gathering, X-ray imaging instruments used in hospitals and laser
instruments used for experimental purposes and, many other electrical and opto-electronic
equipments. They employ luminescent materials for (Nakkiran et al., 2007) electronic portal
imaging devices (EPID), radioisotope distribution and so on (Yeboah & Pistorius, 2000;
Chou et al., 2005). Due to the large size and weight of CRTs, developments of flat-panel
displays (FPDs) are of great interest. Among several FPD technologies, liquid-crystal
displays (LCDs) dominate the FPD market and plasma display panels (PDPs) are now
commercially available in the large area TV market (Yu et al., 2005). New and enhanced
properties are expected due to size confinement in nanoscale dimensions that can
revolutionize the display devices market in future. Commercially available bulk oxysulfides
are quite expensive and are not easily available. So, for the time being, Y2O2S:Eu3+ and
Gd2O2S:Eu3+ nanostructures are relatively a good choice while compared with the bulk
systems. However, for an extensive use in the commercial applications, Y2O2S:Eu3+ and
Gd2O2S:Eu3+ nanocrystals must be prepared at lower temperatures. Therefore, it is necessary
to develop a low-temperature synthesis technology for the growth of oxysulfide
nanophosphors. In this background, this chapter has been devoted to the nanophosphors
development using these two systems. The realm of novel devices from this wonderful
material is yet to be accomplished in full. To give a quantitative report on the state of art of
Y2O2S:Eu3+ and Gd2O2S:Eu3+ is quite difficult and an attempt has been made to give an
account of the synthesis of the nanophosphors in this chapter.
A detailed survey on Y2O2S: Eu3+ and Gd2O2S: Eu3+ nanophosphors discusses various
synthesis techniques adopted by different research groups as follows. Powder phosphors
of (Y1–XREX)2O2S, (Gd1–XREX)2O2S and (La1–XREX)2O2S where RE=Eu3+, Tb3+, or Tm3+ that
were prepared by combustion reactions from mixed metal nitrate reactants and
dithiooxamide (CSNH2)2 with ignition temperatures of 300 – 350 ºC (Bang et al., 2004). The
Y2O2S:Eu3+ red phosphor which was prepared by a new method of decomposing the metal
complexes Y(NO3)3•3(DMSO) and Eu(NO3)3•4(DMSO) in H2S atmosphere at 900 ºC (Guo
et al., 2008). Nanocrystalline Y2O2S:Eu3+ was successfully prepared using a combustion
synthesis method employing conventional sulfur flux (Fu et al., 2008). A method of
preparing red emitting Eu3+:Y2O2S phosphors in which Yttrium sulfite doped Eu is used as
starting material and the Eu-activated oxysulfide is obtained from either directly by
reducing the sulfite with carbon monoxide, or by first oxidizing sulfite and then reducing
the obtained oxysulfate (Koskenlinn et al., 1976). Preparation of spherical Y2O2S and
Y2O2S:Eu particles using a solid–gas reaction of monodispersed precursors with elemental
sulfur vapor under an argon atmosphere has been investigated (Delgado da Vila et al.,
1997). A one-step solvothermal process developed for the preparation of Eu3+ actived
yttrium oxysulfide phosphor in ethylenediamine solvent at 280°C through a reaction of
yttrium oxide, europium oxide, and sulfur powder (Kuang et al., 2005). Luminescent
Y2O2S:Eu3+ nanoceramics prepared through a gel–polymer thermolysis process employing
a urea–formaldehyde resin (Dhanaraj et al., 2003). Nanocrystals of Y2O2S:Eu3+ synthesized
using a two step sol–gel polymer thermolysis method (Dhanaraj et al., 2004). Ensembles of
Y2O2S:Eu3+ widegap semiconductor nano-crystals exhibit ON–OFF fluorescence blinking
phenomenon, which mimic II–VI semiconductor quantum-structures synthesized using
sol-gel polymer thermolysis method (Thirumalai et al., 2007). Trivalent europium-doped
yttrium oxysulfide nanocrystals synthesized using sol–gel thermolysis. A significant blue
shift observed in the fundamental absorption edge for the nanocrystals having an average
crystallite size (f) in the range 9–15nm indicated a strong quantum confinement with a Bohr
exciton radius of 5–13 nm (Thirumalai et al., 2007; Thirumalai et al., 2008). The Y2O2S:Eu3+
nanocrystallines that were prepared by a new ethanol assisted combustion synthesis
method using sulfur contained organic fuel (thioacetamide) in an ethanol-aqueous solution
(Xixian et. al., 2006). The luminescence dynamics of optical centers in nanocrystals
depending critically on the phonon density of states (PDOS) is quite distinct from that of
bulk materials. It is shown that energy transfer (ET) in nanocrystals is confined by discrete
PDOS as well as direct size restriction. For applications, the nanoconfinement effects on ET
Synthesis and morphology control of Eu3+ doped M2O2S [M=Y, Gd] nanostructures 47

significantly reduce the efficiency of sensitized or upconversion luminescence (Chen et. al.,
2003). The Y2O2S:Eu phosphor powders were prepared with a flux fusion method and
electrophoretically deposited on an ITO-coated glass substrate to form a thin layer (Tseng
et al., 1998). The nanostructured yttrium oxysulfide films prepared via vapor phase growth
(V. V. Bakovets et al., 2008). Their first step was the deposition of 50-nm-thick
nanostructured yttria films from yttrium dipivaloylmethanate vapor at 525 ºC. Next, the
films were sulfided in ammonium thiocyanate vapor at temperatures from 800 to 1100 ºC.
Hexagonal yttrium oxysulfide was obtained at 900 ºC and higher temperatures. The
investigations of pseudobinary systems Ln2O2S---La2O2S (Ln = Nd, Sm, Eu, Gd, Dy, Yb, Lu,
and Y) with complete solid solubility only for the systems like Nd2O2S---La2O2S and
Sm2O2S---La2O2S; a two-phase region is found for all other systems (M. Leskelä et al., 1976).
The solid solubility in the isostructural oxysulfide series is discussed in terms of the
differences in ionic radii of the two rare-earth components. The size- (submicrometer-sized)
and morphology- (spherical) controlled composite Gd–Eu oxalate particles were prepared
in an emulsion liquid membrane (water-in-oil-in-water emulsion) system (Hirai et al.,
2002). The oxalate particles thus prepared were calcined in air to obtain Gd2O3:Eu3+
phosphor particles and in sulfur atmosphere to obtain Gd2O2S : Eu3+ phosphor particles.
Usually synthesis of any phosphor material would necessitate rigorous conditions such as
heavy milling, intermediate and very high temperature heat treatment cycles, etc. Whereas
the hydrothermal approach avoids all those difficult and time consuming steps by
modifying the reaction parameters to suitable forms so as to react and produce target
compound(s) at relatively lesser duration with the capacity to develop desired shapes and
dimensions with high crystallanity. The Y2O2S:Eu3+ and Gd2O2S:Eu3+ nanostructures show
optimum properties (perfect crystalline structure, high stability and good morphological)
when they are grown on perfect hydrothermal conditions. In addition to the conventional
vapor-phase method, which includes vapor transport and condensation (Kong et al., 2001),
metal-organic chemical vapor deposition (Zhang et al, 2004), thermal evaporation (Pan et
al, 2001) and solution-phase methods have been developed as alternative ways to
synthesize semiconductor nanostructures with different shapes and dimensions.
Hydrothermal method is a widely used technique that can control the shape and
dimension of nanostructures among all solution-based approaches (Zhang et al, 2002).
Unlike conventional vapor-phase methods, the hydrothermal method can produce various
nanostructures at a relatively low temperature (below 200 °C) using simple equipments;
however, the reaction time required for the growth of nanostructures is too long (usually
from a few hours to several days) (Wang & Li, 2003; Wang et al, 2003; Zhang et al, 2002).
The various low-dimensional nanostructures, such as nanowires, nanotubes, nanosheets
and fullerene like nanoparticles that have been selectively synthesized from rare-earth
compounds (hydroxides, fluorides) based on a facile hydrothermal method (Wang & Li,
2003; Wang et al, 2003). The subsequent dehydration, sulfidation and fluorination
processes lead to the formation of rare-earth oxide, oxysulfide and oxyhalide
nanostructures, which can be functionalized further by doping with other rare-earth ions.
An effective method to synthesize Y2O2S:Eu3+, Mg2+, Ti4+ nanoparticles. Tube-like Y(OH)3
were firstly synthesized by hydrothermal method to serve as the precursor.
Nanocrystalline long-lasting phosphor Y2O2S:Eu3+, Mg2+, Ti4+ was obtained by calcinating
the precursor with co-activators and S powder (Li, et. al., 2010). The afterglow properties of
Eu3+ activated long lasting Gd2O2S phosphor by hydrothermal route. Rod-like Gd(OH)3
were firstly synthesized by hydrothermal method to serve as the precursor. Long lasting
Gd2O2S:Eu3+,Ti4+,Mg2+ phosphor were obtained by calcinating the precursor with co-
activators and S powder (Hang et al., 2008). Hydrothermally prepared Gd(OH)3 nanorod
precursor, codoped with Eu, Ti and Mg, was converted into the desired phosphor by
calcinating the precursor in CS2 atmosphere (Mao et al., 2008). Therefore, the development
of a simple and fast synthetic route that can control the shape of nanostructures under
ambient conditions is the need of the hour and hydrothermal technique has many
advantages and technological possibilities. Earlier (Wang & Li, 2003; Wang et al, 2003;
Hang et al., 2008; Mao et al., 2008) hydrothermal method was used to synthesize oxysulfide
nanotubes and nanorods.
To the best of the author’s knowledge, no systematic study has been reported on other
nanostructures like nanocrystals, nanosheets, nanobelts, nanotubes, nanorods, nanowires
and nanoflowers of this oxysulfide system (Thirumalai et al, 2008; Thirumalai et al, 2009 a,
b). Hence maximum efforts were to be put forth in selecting techniques for the synthesis of
various nansostructures with good crystallinity, so also any other oxysulfide
nanostructures of high research potential activity. Therefore, the present study has been
undertaken with a view to synthesizing uncontaminated, highly crystalline Y2O2S:Eu3+ and
Gd2O2S:Eu3+ adopting two methods of hydrothermal routes owing to the relatively lower
temperature with respect to the bulk counterpart (solid-state reaction method): Template –
free method (two-step synthesis) involving the synthesis of as-prepared Y(OH)3 and
Gd(OH)3, followed by subsequent Eu3+ doping and sulfurization leading to conversion of
Y2O2S:Eu3+ and Gd2O2S:Eu3+ nanostructures, respectively, seemed to be a topotactic
reaction. Template – assisted method (single-step synthesis) involving Anodic Aluminium
Oxide (AAO) membranes used for synthesizing of Y2O2S:Eu3+ and Gd2O2S:Eu3+
nanostructures. Furthermore, being a low temperature and an easy-to-adopt methodology,
this method yields phase pure Y2O2S:Eu3+ and Gd2O2S:Eu3+ nanostructures with better
reproducibility, definite shape and dimensions. Hydrothermal method is basically a
simple, easy and fast synthetic route where the most important synthesis parameters are
the precursors, the solution concentration, solution pH value, solvent, temperature and
time. Also, this processing route provides the basis for a nearly low cost, low temperature
method for the preparation of homogeneous nano-sized ceramics compared to any other
existing methods.

2. Formation, structure, and morphology of the Y2O2S:Eu3+and Gd2O2S:Eu3+

nanostructures
2.1 Template – free Method
A systematic study has been undertaken on other nanostructures like nanocrystals,
nanosheets, nanobelts, nanotubes, nanorods, nanowires and nanoflowers by varying
reaction parameters such as solution concentration, pH, growth temperature and reaction
time and solvent using template-free method. The effect of these parameters are studied
using various characterization studies and the optimized reaction conditions are arrived
for Y2O2S:Eu3+ and Gd2O2S:Eu3+ nanostructures. The detailed synthesis procedure is
discussed already in detail (Thirumalai et al, 2008; Thirumalai et al, 2009 a, b).
Structural studies (XRD) reveal that the products Y2O2S:Eu3+ and Gd2O2S:Eu3+ are pure
hexagonal phase and they are in good agreement with standard JCPDS [(Y2O2S:Eu3+;
Synthesis and morphology control of Eu3+ doped M2O2S [M=Y, Gd] nanostructures 49

JCPDS # 24-1424) and (Gd2O2S:Eu3+; JCPDS # 26-1422)] data (Thirumalai et al, 2008;
Thirumalai et al, 2009 a, b) and they are highly crystalline in nature. It is already stated
that, the conversion from hydroxide to oxysulfide seems to be a topotactic reaction. Firstly,
the hydroxide(s) [Y(OH)3 and Gd(OH)3] of nanocrystals / nanoplates, nanosheets,
nanobelts, nanotubes, nanorods and nanowires are selectively synthesized were based on
the preparation of colloidal hydroxide precipitates at room temperature, and the
subsequent hydrothermal treatment at 100 – 180 ºC for approximately 12 – 48 hours
(Thirumalai et al, 2008; Thirumalai et al, 2009 a, b). The hydrothermal method was shown
to be effective in the synthesis of zero- and one-dimensional nanostructures. By the simple
tuning of factors such as pH, temperature and concentration, the experimental conditions
could be chosen to favor the anisotropic growth of materials. In the present work,
nanostructures of hydroxide(s) were successfully obtained through this precipitation–
hydrothermal synthetic method by properly tuning the temperature, pH and time, the
crystal structures have been found to be responsible for the growth of hydroxide
nanostructures with nearly controllable aspect ratios. The conversion of hydroxide to
oxysulfide seems to be a topotactic reaction (i.e., the morphology does not change while the
phase of the material changes). However in any topotactic reactions, where significant
atomic rearrangement due to chemical changes take place, though the morphology remains
intact. Yet it is possible to change the resulting morphology. With this important criteria
the as-synthesized Y(OH)3 and Gd(OH)3 were converted to Y2O2S:Eu3+ and Gd2O2S:Eu3+
nanostructures. To investigate the optimized growth of the Y2O2S:Eu3+ and Gd2O2S:Eu3+
nanostructures, SEM and TEM (Thirumalai et al, 2008; Thirumalai et al, 2009 a, b)
micrographs of nanocrystals, nanosheets, nanobelts nanotubes, nanorods and nanowires,
respectively, were obtained by varying the reaction parameters (pH ~ 6 – 13, 100 – 180 C,
24 – 48 hrs) and its resulting morphological features and crystallanity are found to be good
and were discussed in Table I (Thirumalai et al, 2009 a, b). The studies reveal that by
altering the reaction conditions (like solution concentration, pH, temperature and time) the
morphology, shape and crystallinity may be perfectly controlled. The SEM and TEM
results are discussed and presented in Fig. 1-4. Fig. 1a, 2a and Fig. 3b, 4b show that growth
is clearly seen indicating the nanocrystals formed due to combination of spherical and
hexagonal-like structure at a pH of 6, 100C, 48 hrs, respectively. Fig. 1b, 2b and Fig. 3c, 4c
show stacked nanosheets of the oxysulfide(s) growth at pH ~ 8, 120C, 12 hrs, respectively.
This may be attributed to the two-dimensional growth tendency of the oxysulfide(s)
nanosheets at lower pH. Nevertheless, Fig. 1c, 2c and Fig. 3d, 4d, show the nanobelts of
Y2O2S:Eu3+ and Gd2O2S:Eu3+ grown at a pH of 8 – 9, 180C, 24 hrs, respectively. These
nanosheets and/or nanobelts curl from the edge, indicating a possible rolling process for
the formation of the nanotubes/rods/wires. Fig. 1(d, e, f), 2(d, e, f) and Fig. 3(e, f, g), 4(e, f,
g) show that the nanotubes, rods, wires were grown at a pH of 12 – 13, 140 – 180C, 24 – 48
hrs and has no template to drive the directional growth of nanotubes, rods and wires. The
images indicate that the samples are single crystalline in nature and uniformly distributed.
Also, based on the above studies it is evident that these nanostructures are stable under
thermal treatment, which may be rather useful for their subsequent applications as
catalysts. Furthermore, the morphology of these nanostructures is likely to be a near-
quantum structure. It is having high surface-to-volume ratio, that plays a major role in the
density of singly ionized oxygen vacancies and the charge state of these defects, that may
be due to the existence of surface depletion.
Fig. 1. SEM images show the optimal experimental conditions of Y2O2S:Eu3+ nanostructures
using template – free method.
Synthesis and morphology control of Eu3+ doped M2O2S [M=Y, Gd] nanostructures 51

Fig. 2. SEM images show the optimal experimental conditions of Gd2O2S:Eu3+

nanostructures using template – free method.
Fig. 3. (a) SEM image of hexagonal-shaped bulk Y2O2S:Eu3+ (A). The TEM images of
Y2O2S:Eu3+ nanostructures (template – free method) from the sample A7 to A12: (b)
Nanocrystals, (c) Nanosheets, (d) Nanobelts, (e) Nanotubes, (f) Nanorods, (g) Nanowires, (h)
a close up of the boxed area in (g) shows a structure of single-nanowire, (i) HRTEM image of
Y2O2S:Eu3+ nanowire. Courtesy: J. Colloid. Interface. Sci. 336., 2., (2009) 889-897.
Synthesis and morphology control of Eu3+ doped M2O2S [M=Y, Gd] nanostructures 53

Fig. 4. (a) SEM image of hexagonal-shaped bulk Gd2O2S:Eu3+ (B). The TEM images of
Gd2O2S:Eu3+ nanostructures (template – free method) from the sample B7 to B12: (b)
Nanocrystals, (c) Nanosheets, (d) Nanobelts, (e) Nanotubes (inset: shows TEM image of the
end portion of a single-nanotube), (f) Nanorods, (g) Nanowires, (h) a close up of the boxed
area in (g) shows a structure of single-nanowire, (i) HRTEM image of Gd2O2S:Eu3+
nanowire. The spacing between two adjacent lattice planes is 0.325 nm, which corresponds
to the separation of the hexagonal phase lattice planes (110), and the inset shows Fast
Fourier transform (FFT) pattern of the corresponding nanowire. Courtesy: J. Mater. Sci. 44.,
14., (2009) 3889-3899.
Fig. 5. A schematic diagram of the shape-selective synthesis of doped oxysulfide
nanostructures via hydrothermal route (template – free method).

2.2 Possible formation mechanism

A schematic diagram of the shape-selective synthesis of doped oxysulfide nanostructures via
template-free hydrothermal route is shown in Fig. 5. Based on these experimental results
and discussions, a possible formation mechanism from nanocrystals to nanowires of
hydroxides might be proposed as follows: When the Y/Gd(NO)3 and NaOH solutions are
mixed together, the resultant hydroxide nanocrystals can accommodate some Na ions and
water molecules between their adjacent surfaces. According to Kerr (Kerr, 1996), these Na
ions and water molecules are then exchanged and fused to form nanosheet or nanobelt
according to the pH and temperature variation as indicated by the schematic diagram in Fig.
5. As the reaction continues the conformational variation inevitably causes torsional
movement within the chains (Colomban, 1999; Mazerolles, 1999), causing the nanosheet /
nanobelt to roll up. Because the thin lamellar nanosheet / nanobelt has very good flexibility,
they curl and buckle with appropriately 45 degree rotation. As the curling continues, they
bind tighter and fuse together at the two edges to form a nanotube with a cylindrical cross
section. Furthermore, nanorods and nanowires are obtained according to reaction time, pH
and temperature following the same formation kinetics.
Synthesis and morphology control of Eu3+ doped M2O2S [M=Y, Gd] nanostructures 55

3. Template-assisted method
In order to synthesize one-dimensional nanomaterial into a device, a fabrication method that
enables well-ordered nanomaterials with uniform diameter and length is important.
Template-directed growth is a nanomaterials fabrication method that uses a template, which
has arrays of nanopores with uniform diameter and length that is needed for a device.
Template-based growth is commonly a solution or colloidal dispersion based process. It is
less expensive and readily scalable to mass production. The diameter, density and length of
nanotubes, nanorods and nanowires are easily controlled independently. It also offers the
advantage of less contamination and is environmentally benign. However, template-based
synthesis slightly suffers from the polycrystalline nature of the resultant nanowires and
nanorods, in addition to the difficulties to find appropriate templates with pore channels of
desired diameter, length and surface chemistry and to remove the template completely
without compromising the integrity of grown nanotubes, nanorods and nanowires and is
another cumbersome work. In this work the Y2O2S:Eu3+ and Gd2O2S:Eu3+ of highly ordered
arrays of super-nanostructures like, nanotubes, nanorods and nanoflowers were synthesized
by hydrothermal synthesis via template assisted synthesis using commercially available
AAO templates [Whatman Nuclepore® Inc., (13 mm diameter, 6 μm thickness)]. In
comparison with the bulk, this template assisted nanosynthesis route offer several advantages
for property study and practical applications. As a whole, to the best of our knowledge, this is
the first attempt to prepare such nanostructures from the zero- to three-dimensional scale
using both template – free and template – assisted method for these two sytems.
We have extended the method to the synthesis of various nanostructures like
nanotubes, nanorods and nanoflowers preparation using template-assisted
hydrothermal technique by varying reaction parameters such as solution pH, solvent
temperature, time, etc. The porous AAO templates were immersed into a required
amount of [Y/Gd(CH3COO)3·4H2O] and [Eu(CH3COO)3·4H2O] solution, respectively,
and evacuated for about 15 min using a vacuum pump to get rid of bubbles within the
nanopores. The template was taken out of the solution, rinsed with distilled water and
dried in air. This procedure was repeated several times at regular intervals.
Subsequently, the template was put into a Teflon-lined stainless steel autoclave, and
the solution of YEu(CH 3COO)3 and GdEu(CH3COO)3, respectively, followed by
addition specific amount of Na 2S. The pH of the reaction solution was adjusted in the
range from 6 to 9. After that the autoclave was tightly sealed, heated around 150 – 200
°C for 20 hrs and then allowed to cool down to room temperature naturally. For the
synthesis of nanoflower-like structures, hexamethylenetetramine (HMT) was used as
surfactant. The as-obtained Y2O2S:Eu3+ and Gd2O2S:Eu3+ template was subsequently
annealed at 300ºC for 1 hr under inert Sulphur atmosphere, respectively, to yield the
final product.
Structural studies (XRD) reveal that the products Y2O2S:Eu3+ and Gd2O2S:Eu3+ are pure
hexagonal phase and they are in good agreement with standard JCPDS [(Y2O2S:Eu3+ : JCPDS
# 24-1424) and (Gd2O2S:Eu3+ : JCPDS # 2 6-1422)] data (Thirumalai et al, 2008; Thirumalai et
al, 2009 a,b) and they are highly single crystalline in nature. The patterns show obviously
broadened diffraction peaks compared with the bulk Y2O2S:Eu3+ and Gd2O2S:Eu3+ systems,
signifying the decrease in size of these crystallites. No peaks attributable to other types of
oxysulfide(s) are observed in the XRD patterns, indicating the high purity of the phases
obtained. The morphology of the resulting samples synthesized by the hydrothermal
method was studied using SEM. Using the commercially available AAO templates the
morphologies of highly oriented nanoarrays of structures like nanotubes, nanorods and
nanoflowers of Y2O2S:Eu3+ and Gd2O2S:Eu3+ were synthesized. The optimal experimental
conditions and resulting morphologies of Y2O2S:Eu3+ and Gd2O2S:Eu3+ nanostructures are
given in Table – I. Fig. 6(a1-c2) & Fig. 7(a1-c2) show the SEM micrographs of the synthesized
Y2O2S:Eu3+ and Gd2O2S:Eu3+ nanotubes, nanorods and nanoflowers, respectively. They were
found to be single-crystalline in nature.
To investigate the optimized growth of the Y2O2S:Eu3+ and Gd2O2S:Eu3+ nanostructures,
SEM micrographs shows highly oriented nanotubes, nanorods and nanoflowers,
respectively, that were obtained by varying the reaction parameters (pH ~ 6 to 9, 150 – 200
C, 20 hrs) and its resulting morphological features and crystallanity are found to be good.
The studies reveal that by altering the reaction conditions (like solution concentration, pH,
temperature and time) the morphology, shape and crystallinity are perfectly controlled. Fig
7(a1, a2) and Fig 8(a1, a2) show, the AAO templates top-view and cross-sections of the as-
synthesized Y2O2S:Eu3+ and Gd2O2S:Eu3+ nanostructures, respectively. A cluster of
nanotube-like arrays of the oxysulfide(s) growth were observed [Fig 7(b1, b2) and Fig 8(b1,
b2)] at a pH of 6, 150 C, 15 hrs. In Fig 7b1 and Fig 8b1 the tube and/or rod-like growth is
clearly seen. The nanotubes and/or rods were grown at a pH of 6 – 9, 160 –180 C, 20 hrs
and the cross-section view reveals that the nanopores are perfectly filled. The nanorods are
straight and have a uniform diameter of about 80 nm, which is basically equal to the pore
size of the AAO template employed. To investigate the effect of solvent concentration on
Y2O2S:Eu3+ and Gd2O2S:Eu3+ nanostructures, SEM micrographs were obtained at different
solvent concentration (using hexamethyltetramine (HMT) as solvent along with water). The
results are discussed and presented in Fig 7, 8(c1 and c2) and Table – I. The sample is
composed of a large number of nearly uniform flower-like nanostructures to the surface of
the AAO template. They are arranged uniformly in a large area. The average size of the as-
obtained nanostructure is about 4-5 μm. A closer inspection reveals that the flower is made
up of many thin petals, the thickness of which is ~ 30 nm. Fig. 6c2 and 8c2 show the cross-
sectional view of SEM image of the Y2O2S:Eu3+ and Gd2O2S:Eu3+ nanorod arrays of length
around hundred nanometers for the flower-like structures. Every petal is curled and thin,
with a smooth surface and a large surface area. We can see a layer of flowers on the surface
of the template. Because of these nanocrystalline sheets like small petals, the flower-like
structure on the alumina template surface formed by these nanocrystalline sheets may be
called nanoflowers. The nanoflower growth is clearly seen and the cross-section reveals that
the nanopores are perfectly filled for concentration of 2 mM. Fig. 6(a2, b2, c2) and Fig. 7(a2,
b2, c2), show a slightly discrete, unattached and dislocation free nanotube/rod structure in
the direction vertical to the AAO template, respectively, which was different from the
shared tube/rod wall between the tubes/rods of the AAO template. The nanotube/rod wall
was also constituted by many deposited particles with an average particle size of 20 nm. It
can be seen that the product obtained by the hydrothermal reaction with the AAO template
had three structural configurations: (i) many fine nanoparticles packed to form tubes/rods,
(ii) many parallel (iii) slightly discrete and completely dislocation free tubes/rods
constituted nanorod-like arrays vertical to the AAO template, respectively. The formation of
1D structure depends greatly on the reaction kinetics. Such 3D structures with high surface
areas can be used relevantly as catalysts, molecular sieves and biosensors (Zhang, 2007).
Synthesis and morphology control of Eu3+ doped M2O2S [M=Y, Gd] nanostructures 57

Experimental Conditions* Resulting morphologies

(pH~6, 150C, 20h) Nanotubes

Y2O2S:Eu3+
(pH~7–8, 200C, 20h) Nanorods

(pH~8–9, 120C, 12h) Nanoflowers

(pH~6, 150C, 20h) Nanotubes

Gd2O2S:Eu3+

(pH~7–8, 200C, 20h) Nanorods

(pH~8–9, 120C, 12h) Nanoflowers

Table 1. Optimized growth conditions of Y2O2S:Eu3+ and Gd2O2S:Eu3+
(Template – assisted Method) nanostructures

*The as-obtained Y2O2S:Eu3+ and Gd2O2S:Eu3+ grown on AAO template was subsequently
annealed at 300 ºC for 1 hr under inert (A2 or N2) / CS2 / Sulphur atmosphere to yield the
final product.

3.1 Possible formation mechanism

Highly oriented single-crystalline Y2O2S:Eu3+ and GdO2S:Eu3+ nanotubes, nanorods and
nanoflowers were synthesized using a template–assisted (AAO and Au coated AAO
templates) hydrothermal technique. While Y/Gd (CH3COO)3.4H2O : Eu(CH3COO)3.4H2O as
precursors act as sources for Y/Gd : Eu ions, during the reaction process, the NaOH not
only plays a role as a solvent for lowering the reaction temperature but also acts as a
reactant and it is used for achieving basic environment through hydrolysis. Here, Na2S is
acting as a sulfurizing agent. Controlling the pH and reaction time is a key factor in
achieving the morphological control. The porous structure of alumina with positive charged
walls attracts the ions having negative charge from the solution leading to a preferential
deposition at the walls that further grow towards the core. Supersaturation in the growth
region is favorable to an anisotropic growth. The shape of a crystal is determined by the
relative specific surface energy of each facet of the crystal. Therefore, the initial deposition of
nanocrystals is critical for the formation of aligned nanotubes and nanorods. It is inferred
that the AAO template may plays an important role in controlling the pH of the reaction
mixture and/or local concentration, which leads to an inhomogeneous concentration
distribution and may affect the shape development. Immersing the alumina template in the
solution containing surfacant (HMT), the solution impregnated through pores to form
nanorods immediately. However, the nanorods were not restricted in the pores anymore
and extends to the surface of the alumina template, due to forces of adhesion. This in turn
may lead to spreading nanorods out of the pore diameter, leading to the possible formation
of nanosheets. The nanosheets so formed over the template surfaces comes in intact with the
extended nanorods and try to form a stable structure. This situation may lead to the
formation of flowery nanostructures. The schematic of the mechanism for nanotube,
nanorod and nanoflower generations are shown in Fig. 8.
In summary, this novel template-free and template-based approach has been developed for
growing high yield and highly oriented single crystal Y2O2S:Eu3+ and Gd2O2S:Eu3+
nanostructures using hydrothermal growth. The pH, growth temperature, reaction time and
surfactant are mainly determining the shape of the nanostructures. The morphological
studies were performed through SEM and TEM for the optimized conditions, which reveals
excellent features and are reported. The morphological studies (SEM and TEM) revealed
that the nanostructures (0D, 1D, 2D and 3D), with various structures like nanocrystals,
nanosheets, nanobelts, nanotubes, nanorods, nanowires and highly oriented nanoarrays
(tubes, rods, flowers) successfully synthesized. The Y2O2S:Eu3+ and Gd2O2S:Eu3+
nanostructures was further examined by HRTEM, electron diffraction and FFT images are
seen to be single-crystalline in nature. Also, the SEM studies of the nanorods over AAO
tempalte (template – assisted method) reveal that they are uniformly distributed throughout
the surface exhibiting the superiority of the structures. The studies agree to a great extent
with the structural studies results. This technique is found to be highly reproducible and can
be extended to large area and large scale fabrication systems. Large-area, non-collapsed and
highly-oriented Y2O2S:Eu3+ and Gd2O2S:Eu3+ nanostructures are expected as ideal functional
components for opto-electronic and nanoscale devices of the next generation.

Fig. 6. The SEM images of Y2O2S:Eu3+ nanostructures (template – assisted method): (a1, a2)
Nanotubes, (b1, b2) Nanorods and (c1, c2) Nanoflowers. The SEM micrographs show the
cross-sectional views of the samples (a2, b2, c2) corresponding to (a1, b1, c1).
Synthesis and morphology control of Eu3+ doped M2O2S [M=Y, Gd] nanostructures 59

Fig. 7. The SEM images of Gd2O2S:Eu3+ nanostructures (template – assisted method): (a1, a2)
Nanotubes, (b1, b2) Nanorods and (c1, c2) Nanoflowers. The SEM micrographs show the
cross-sectional views of the samples (a2, b2, c2) corresponding to (a1, b1, c1).
 Sol Hydrothermal
particles treatment

Solution Nanotube

Nanorod

Nanoflower

Sol contains: Y/Gd (CH3COO)3 : Eu(CH3COO)3

+ Na2S + NaOH (pH ~ 4-8)

Hexamethylenetetramine (HMT)

Fig. 8. Schematic illustration of the growth of nanotube, nanorod and nanflower structures
by a template-assisted hydrothermal technique.

4. References
Bakovets, V. V.; Levashova, T. M.; Filatova, I. Yu.; Maksimovskii E. A. & Kupcha, A. E.
(2008). Vapor phase growth of nanostructured yttrium oxysulfide films. Inorg.
Mater., 44., 1., (February & 2008) 67-69, ISSN: 0020-1685.
Bang, J.; Abboudi, M.; Abrams, B. & Holloway, P. H. (2004) Combustion synthesis of Eu-,
Tb- and Tm- doped Ln2O2S (Ln=Y, La, Gd) phosphors. J. Lumin. 106., 3-4.,
(November & 2003) 177-185, ISSN: 0022-2313.
Chen, X. Y.; Zhuang, H. Z. & Liu G. K.; Li, S. & Niedbala, R. S. (2003). J. Appl. Phys., 94, 1,
ISSN 0021-8979.
Chou, T.W. ; Mylswamy, S.; Liu, R.S. & Chuang, S.Z. (2005). Eu substitution and particle size
control of Y2O2S for the excitation by UV light emitting diodes. Solid. State.Comm.,
136., 4., (August 2005) 205-209, ISSN: 0038-1098.
Synthesis and morphology control of Eu3+ doped M2O2S [M=Y, Gd] nanostructures 61

Colomban, P.; Folch, S. & Gruger, A. (1999). Vibrational Study of Short-Range Order and
Structure of Polyaniline Bases and Salts. Macromolecules. 32., 9., (April & 1999) 3080-
3092, ISSN: 0024-9297.
Delgado da Vila, L.; Stucchi, E. B.; Davolos, M. R. (1997). Preparation and characterization of
uniform, spherical particles of Y2O2S and Y2O2S:Eu. J. Mater. Chem., 7., 10., 2113-
2116, ISSN: 0959-9428.
Dhanaraj, J.; Jagannathan R. & Trivedi, D. C. (2003). Y2O2S:Eu3+ nanocrystals—synthesis and
luminescent properties. J. Mater. Chem., 13., 7., (May & 2003) 1778-1782, ISSN 0959-
9428.
Dhanaraj, J.; Geethalakshmi, M.; Jagannathan, R. and Kutty, T.R.N. (2004). Eu3þ doped
yttrium oxysulfide nanocrystals – crystallite size and luminescence transition(s).
Chem. Phys. Lett., 387., 1-3., (March & 2004) 23-28, ISSN: 0009-2614.
Fu, Z.; Geng, Y.; Chen, H.; Zhou, S.; Yang, H. K. & J. H. Jeong, (2008). Combustion synthesis
and luminescent properties of the Eu3+-doped yttrium oxysulfide nanocrystalline.
Opt.Mater. 31., 1., (September & 2008) 58-62., ISSN: 0925-3467.
Guo, C.; Luan, L.; Chen, C.; Huang D. & Su, Q. (2008). Preparation of Y2O2S:Eu3+ phosphors
by a novel decomposition method. Mater. Lett. 62., 4-5., (February & 2008) 600-602,
ISSN: 0167-577X.
Hang, T.; Liu, Q.; Mao, D. & Chang, C. (2008). Long lasting behavior of Gd2O2S:Eu3+
phosphor synthesized by hydrothermal routine. Mater. Chem. Phys., 107., 1.,
(January & 2008) 142-147, ISSN: 0254-0584.
Hirai, T.; Hirano, T. & Komasawa, I. (2002). Preparation of Gd2O3 : Eu3+ and Gd2O2S : Eu3+
Phosphor Fine Particles Using an Emulsion Liquid Membrane System. J. Colloid.
Inter. Sci. 253., 1., (September & 2002) 62-69, ISSN: 0021-9797.
Kerr, T. A.; Wu, H. & Nazar, L. F. Concurrent Polymerization and Insertion of Aniline in
Molybdenum Trioxide: Formation and Properties of a [Poly(aniline)]0.24MoO3
Nanocomposite. (1996). Chem. Mater. 8., 8., (August 1996) 2005-2015, ISSN: 0897-
4756.
Kong, Y. C.; Yu, D. P.; Zhang, B.; Fang, W. & Feng, S. Q. (2001). Appl. Phys.Lett., 78., 407.,
ISSN 0003-6951.
Koskenlinn, M.; Leskela M. & Niinisto, L. (1976). Synthesis and Luminescence Properties of
Europium-Activated Yttrium Oxysulfide Phosphors. J. Electrochem. Soc. 123., 1.,
(January & 1976) 75-78., ISSN: 0013-4651.
Kuang, J.; Liu, Y. & Yuan, D. Electrochem. (2005). Preparation and Characterization of
Y2O2S:Eu3+ Phosphor via One-Step Solvothermal Process. Solid. State. Lett., 8., 9.,
(July & 2005),H72-H74, ISSN: 1099-0062.
Lei, B.; Liu, Y.; Zhang, J.; Meng, J.; Man, S. & Tan, S. (2010). Persistent luminescence in rare
earth ion-doped gadolinium oxysulfide phosphors. J. Alloys. Comp, 495., 1.,
(February and 2010) 247-253, ISSN: 0925-8388.
Leskelä M. & Niinistö, L. (1976). Solid solutions in the rare-earth oxysulfide series. J. Solid.
State. Chem. 19., 3., (November 1976) 245-250, ISSN: 0022-4596.
Li, W; Liu, Y; & Ai, P. (2010). Synthesis and luminescence properties of red long-lasting
phosphor Y2O2S:Eu3+, Mg2+, Ti4+ nanoparticles. Mater. Chem. Phys., 119, 1-2.,
(January & 2010) 52-56, ISSN: 0254-0584.
Liu, Y. & Kuang, J. (2010). Intense visible luminescence from Nd3+-doped yttrium
oxysulfide. J. Lumin. 130., 3., (March and 2010) 351-354, ISSN: 0022-2313.
Mazerolles, L.; Floch, S. & Colomban, P. (1999). Study of Polyanilines by High-Resolution
Electron Microscopy. Macromolecules. 32., 25., (November & 1999) 8504-8508, ISSN:
0024-9297.
Mao, S.; Liu, Q.; Gu, M.; Mao D. & Chang, C. (2008). Long lasting phosphorescence of
Gd2O2S:Eu,Ti,Mg nanorods via a hydrothermal routine. J. Alloys and Comp., 465., 1-
2., (October & 2008) 367-374, ISSN: 0925-838.
Mikami, M. & Oshiyama, A. (1998). First-principles band-structure calculation of yttrium
oxysulfide.Phys. Rev. B, 57., 15., (April &1998) 8939-8944, ISSN: 1098-0121.
Nakkiran, A.; Thirumalai, J.; & Jagannathan, R. (2007). Luminescence blinking in Eu3+ doped
yttrium oxysulfide (Y2O2S:Eu3+) quantum-dot ensembles: Photo-assisted relaxation
of surface state(s). Chem. Phys. Lett., 436., 1-3., (February and 2007)155-161, ISSN:
0009-2614.
Pan, Z. W.; Dai, Z. R. & Wang, Z. L. (2001). Nanobelts of Semiconducting Oxides. Science.
291., 5510., (March & 2001) 1947-1949, ISSN: 0036-8075.
Sabot, J. L. & Maestro, P. (1995). Kirk-Othmer Encyclopedia of Chemical Technology, Concise, 4th
Edition, John Wiley & Sons, IV-edn, ISBN: 0-471-41961-3.
Tseng, Y.H.; Chiou, B.S.; Peng, C.C. & Ozawa, L. (1998). Spectral properties of Eu3+-activated
yttrium oxysulfide red phosphor. Thin Solid Films., 330., 2., (September & 1998) 173-
177, ISSN: 0040-6090.
Thirumalai, J.; Jagannathan, R. & Trivedi, D.C. (2007). Y2O2S:Eu3+ nanocrystals, a strong
quantum-confined luminescent system. J. Lumin. 126., 2., (October and 2007) 353-
358, ISSN: 0022-2313.
Thirumalai, J.; Chandramohan, R.; Sekar M.& Rajachandrasekar, R. (2008). Eu3+ doped
yttrium oxysulfide quantum structures — structural, optical and electronic
properties. J. Nanopart. Res. 10., 3., (August & 2007) 455-463, ISSN: 1388-0764.
Thirumalai, J.; Chandramohan, R.; Divakar, R.; Mohandas, E.; Sekar, M. & Parameswaran, P.
(2008). Eu3+ doped gadolinium oxysulfide (Gd2O2S) nanostructures —synthesis and
optical and electronic properties. Nanotechnology. 19., 39., (October & 2008) 395703-
5, ISSN: 0957-4484.
Thirumalai, J.; Chandramohan, R.; Auluck, S.; Mahalingam, T. & Srikumar, S. R. Controlled
synthesis, optical and electronic properties of Eu3+ doped yttrium oxysulfide
(Y2O2S) nanostructures. (2009a). J. Colloid. Interface. Sci. 336., 2., (August & 2009)
889-897, ISSN: 0021-9797.
Thirumalai, J.; Chandramohan, R.; Valanarasu, S.; Vijayan T. A.; Somasundaram, R. M.;
Mahalingam, T. & Srikumar, S. R. (2009b). Shape-selective synthesis and opto-
electronic properties of Eu3+-doped gadolinium oxysulfide nanostructures. J. Mater.
Sci. 44., 14., ( July & 2009) 3889-3899, ISSN: 0022-2461.
Wang, X. & Li, Y. (2003). Rare-Earth-Compound Nanowires, Nanotubes, and Fullerene-Like
Nanoparticles: Synthesis, Characterization, and Properties.Chem. Eur. J. 9., 22.,
(November & 2003), 5627-5635, ISSN: 0947-6539.
Wang, X.; Sun, X.; Yu, D.; Zou, B. & Li, Y. D. (2003). Rare earth compound Nanotubes. Adv.
Mater. 15., 17., (September & 2003) 1442-1445, ISSN: 0935-9648.
Xixian, L.; Wanghe, C. & Mingming, X. (2006). Preparation of nano Y2O2S:Eu phosphor by
ethanol assisted combustion synthesis method. J. Rare. Earths., 24., 1., (February &
2006) 20-24, ISSN: 1002-0721.
Synthesis and morphology control of Eu3+ doped M2O2S [M=Y, Gd] nanostructures 63

Yeboah, C. & Pistorius, S. (2000). Monte Carlo studies of the exit photon spectra and dose to
a metal/phosphor portal imaging screen.Med. Phys. 27., 2., (February & 2000) 330-
339, ISSN: 0094-2405.
Yu, T.; Zhu, Y.; Xu, X.; Shen, Z.; Chen, P.; Lim, C-T.; Thong, J.T-L.& Sow, C-H. (2005).
Controlled Growth and Field-Emission Properties of Cobalt Oxide Nanowalls. Adv.
Mater. 17., 13., (July & 2005) 1595-1599, ISSN: 0935-9648.
Zhang, B. P.; Binh, N. T.; Wakatsuki, K.; Segawa, Y.; Yamada, Y.; Usami, N.; Kawasaki, M.
& Koinuma, H. (2004). Appl. Phys. Lett., 84., 4098, ISSN 0003-6951.
Zhang, C.; Tao, F.; Liu, G-Q.; Yao, L- Z. & Cai, W-L. (2008). Hydrothermal synthesis of
oriented MnS nanorods on anodized aluminum oxide template. Mater. Lett. 62., 2.,
(January & 2008) 246-248, ISSN: 0167-577X.
Zhang, D.; Fu, H.; Shi, L.; Fang, J. & Li, Q. (2007). Carbon nanotube assisted synthesis of
CeO2 nanotubes. J. Solid. State. Chem., 180., 2., (February 2007) 654-660, ISSN: 0022-
4596.
Zhang, J.; Sun, L.; Yin, J.; Su, H.; Liao, C. & Yan, C. (2002). Control of ZnO Morphology via a
Simple Solution Route. Chem. Mater. 14., 10., (September & 2002) 4172-4177, ISSN:
0897-4756.
Emission of semiconductor nanocrystals in photonic crystal environment 65

X4

Emission of semiconductor nanocrystals in

photonic crystal environment
Sergei G. Romanov* and Ulf Peschel
Cluster of Excellence “Advanced Engineered Materials”
Institute of Optics, Information and Photonics, University of Erlangen - Nürnberg,
Günther-Scharowsky-Str.1, 91058 Erlangen, Germany
*Ioffe Physical Technical Institute, Polytekhnishesckaya ul., 26, 194021 St. Petersburg,
Russia

1. Introduction
Widespread application of colloidal semiconductor nanocrystals (NCs) [1] in light emitting
devices is guaranteed by their low production cost, broadly variable emission frequency and
relatively long lifetime. In order to exploit the advantages of NCs, the light sources should
be equipped with non-dissipative photon management structures with the aim to form the
spectrum of the source, to manage the emission diagram and to arrange the radiative
recombination of electron excitations in the most purpose-efficient way. Obviously, the
electrically biased light sources are the most application relevant [2], but developing the
principles of photon management architectures can be performed with systems operating
with photoexcitation of the NC emission.
One of the most awaiting realisations of the photon management in light sources is based on
the photonic crystals (PhC). The radiation in such sources is controlled through modification
of the electromagnetic vacuum [3,4,5]. The pre-condition of such control is the modulation
of the photon density of states (DOS) that allows either accelerate or suppress the radiation
rate of light emitters in a pre-defined spectral range. The development of PhC-based sources
begun with investigation of the spectra and directionality changes of the emission from
embedded sources. Nowadays the main aim is to design special defect states that are
capable of purposive shaping the emission characteristics.
In PhC-based devices the emitter should be coupled to the outside world only via PhC
eigenmodes in order to gain a full control upon emission characteristics. This can be
achieved by immersing the emitter in the PhC interior. At the early stage it became clear that
the fabrication of the 3-dimensional PhCs for the visible range of the spectrum took a long
time due to complexity of required nanofabrication [6,7,8]. Meanwhile the nature offers an
easy affordable solution – the opal crystals [9,10], the play of colours of which is based on
principles of PhCs. Opals consist of face centred cubic packed lattice of silica spheres [11],
i.e., the dielectric permittivity inside the opal is periodically modulated in all three
dimensions. The interference of light waves that are scattered in the opal lattice results then
in formation of directions, which are forbidden for propagation, if the wavelength is
comparable to the lattice constant.
66 Nanocrystals

This phenomenon can be formalised in terms of the photonic energy band structure that
looks similar to the electron energy band structure of solids. This energy band structure
represents the dispersion of propagating modes in the energy-wavevector space. The
spectral intervals without propagating modes are called the photonic bandgaps (PBG).
Obviously, the complete absence of modes can be achieved only in infinitely large PhC with
high refractive index contrast between scatterers and surrounding medium, hence, in reality
it is more important to differentiate PhCs with omnidirectional and directional PBG. So far,
no omnidirectional PBG was demonstrated in the visible due to the absence of dielectrics
with high enough index of refraction [12], but this requirement can be fulfilled in the near-
infrared with, e.g., Si or Ge-based inverted opals.
In the visible we ought to deal with directional bandgap crystals. This means that the
internal light source, the emission of which is simultaneously coupled to the all available
modes of a PhC, will not be blocked completely at any single frequency. Thus, there are two
fractions of the energy flow – the one that is modified by the interaction with the structure
and the other that leaks unaffected from the crystal. The aim of designing the photon
management architectures is to tailor the fraction under control in the purposive way and to
maximise the modified fraction of light flow.
The attractiveness of opals for emission manipulation was immediately realised by
researches and the very first publications that considered the opal as a PhC were aimed
primarily at the emission modification [13,14]. Certainly, the bulk opals in use were
structurally very imperfect and the in-void synthesised emitters were randomly distributed
over the PhC volume. However, even in such conditions it became possible to establish
some links between PBG and emission characteristics. Since that time the tremendous
progress was achieved triggered by the invention of artificial opal films of high structural
quality [15]. Another crucial development in late 90s was the infiltration of the opals with
colloidal NCs as a method that allows to preserve the crystal quality and the refractive index
contrast of the opal-based PhC [16]. Nowadays semiconductor NCs are conveniently used in
studies of the PhC light sources [17].
The opals discussed in this chapter are the PhC with the directional PBG. We will consider
the photoluminescence (PL) from opal-embedded NCs possessing the emission band in the
visible under continuous wave (cw) excitation of moderate power. Any time-resolved
characteristics and optical non-linearities are outside the scope of our discussion. The aim of
this chapter is to demonstrate what kind of emission modification one can achieve with NC
that are evenly distributed over the opal-based PhC volume or placed in close vicinity to the
PhC surface. We will describe several tests that allow to identify the emission changes and
estimate the order of magnitude of these changes.

2. General properties of light sources in photonic crystals

Let us introduce some definitions. In the free space the number of electromagnetic (EM)
 3V
field modes with frequency   0 is N ( )  . Correspondingly, the mode density is
3 2c3
dN ( ) 4 2 3  2
D ( )  and D ( )   2 (Fig.1). The dipole emission power is U  d D ( )
d 3V
and the emission rate is   U    D( ) . Using the Golden Fermi rule and following [18]
Emission of semiconductor nanocrystals in photonic crystal environment 67

we consider the 2-level system with the transition frequency 0 that interacts with EM field
of vacuum. The rate of radiative recombination from an excited e to a ground g state
can be written taking into account the field quantization as

 ˆ  2
2  dˆ  E
e g 
2


g ,1kn ( x 0 ) e, vac  (k , n  0 ) (0.1)
k ,n

The summation is taken over all one photon states 1k , n that satisfy the energy
ˆ
conservation law. 2-level atom is located at x0 and its dipole operator is d .

Fig. 1. (а) The spectrum of the optical mode density in a free space. (b) The same spectrum in
a PhC with complete omnidirectional PBG. The defect mode is shown at the mid-frequency
of the bandgap.

In the PhC this general expression is reduced to the density of modes D( ) . Let us expand
  
the operator of electric field over the Bloch modes E k , n ( x) with quasimomentum k and
number n .
ˆ  k , n   
E ( x)   ( E k , n ( x)aˆk , n  h.c.) (0.2)

k n
, 2 0V


where aˆk , n - is the photon annihilation operator in the mode k , n , and the mode field is
normalised to crystal volume
1     2
VV dx ( x) E k .n ( x)  1 (0.3)

Then the emission rate is

 2
0 d  
e g   ( x 0 , u , 0 ) (0.4)
 0

 
where  ( x 0 , u , 0 ) - is the projection of the local density of states (LDOS).

  1     2
 ( x 0 , u , 0 )   u  E k ,n ( x 0)  (k ,n  0 ) ,
V k , n
(0.5)
68 Nanocrystals

 ˆ 
where the matrix element of dipole transition d  g d e and u is the unit vector along

d direction. It is apparent from (0.4) that the photon emission rate is proportional to the
density of photonic modes and the square of the electric field, in particular, the spontaneous
emission is suppressed in the interval of low mode density. The full density of modes is the
the local density of radiating states that is averaged over the dipole orientation in the unit
cell. This is why the PBG in a full DOS D ( )  2   ( 2  k2, n ) leads to the PBG in the local

k ,n
DOS independently on the emitter orientation.
LDOS differs from DOS by taking into account the distribution of the mode intensity in the
primitive lattice cell [19]. In fact, the dipole can add a photon to the emission flux if not only
the state is available in the field but also the field magnitude differs from zero in the dipole
site. Depending on the dipole positioning in the unit cell of a PhC the LDOS changes and,
accordingly, changes the emission rate. For example, the air modes at the upper PBG edge
possess the field maximum in the regions of “light” dielectric. Hence, in order to maximise
the outcome, the emitter should be positioned in the middle of the PhC voids. If this emitter
finds itself in the point of the zero LDOS, only the weak spontaneous emission will be
possible even in the non-zero total DOS. At the PBG edges the spontaneous emission can be
enhanced, because of higher LDOS. But in the infinitely large PhC obeying the spherical
symmetry (the case that cannot be realised experimentally) the Fermi Golden rule fails
because of the DOS discontinuity. In the case of finite size PhC such discontinuity is
replaced by van Hove singularity and disappears in the case of directional PBG.
In order to consider the stimulated emission one has to introduce the impurity atoms with

certain volume density  ( r ) that experience the polarization  by the PhC field and
possess the inverse population of their energy levels Im   0 . Polarization of impurity

atoms that is excited in point (r , t ) by the mode Ek(T, n) can be expressed as

  
Pst(1) ( r , t )   ( r ) Ek(T, n) (r )exp(i (T ) k , n   )t (0.6)
 (2) 
In self consisted approach it is necessary to include the polarization wave P st (r , t ) that is
 (1) 
induced by the field E ( r , t ) :
 (2)    (1) 
P st (r , t )   (r ) E (r , t ) (0.7)
( 2)
The polarization P st (r , t ) induces the electric wave

 (2)  
E (r , t )  1  k2, nl 2 E (kT, n) exp(i   )t (0.8) (0.9)
2

the full self consisted stimulated field is expanded in the series

  ( j )   (T ) 
Ek(T, n) (r )   E (r , t )  E k , n (r )exp(  k , nl  (ik(T, n)   )t ) (0.10)
j 0
Emission of semiconductor nanocrystals in photonic crystal environment 69

On the one hand, the enhancement factor of stimulated emission per unit length
  k(T, n)
 k , n  1/ vg (k , n) is inversely proportional to the group velocity vg ( k , n)  along, e.g.,
k z

ik(T, n) F1 ( k , n)
the z direction. On the other hand,   k , n  , where the effective density of
2vg ( k , n)
 (T ) 
impurities normalised to the mode intensity E k , n (r ) is

 1   (T )  2 
F1 (k , n)    (r ) E k , n (r ) d r (0.11)
VV

 
Finally, the stimulated emission Est ( r , t )  Ek(T,n) (r )exp(  k , nl  (ik(T, n)   )t ) acquires the
enhancement in the range of slow waves.
It is worth noting the more elaborated approach to the dynamics of emission in PhC. For
example, the strongest modification experiences the emitter whose emission band is located
at the edge of the complete PBG due to strong DOS anomaly [20]. The emitter with
transition frequency deeply inside PBG will emit photon, but this photon has no chance to
propagate and will be absorbed by the same emitter. S. John named this situation as the
„dressed” atom [21]. If the transition frequency falls close to the edge, the emission
dynamics will be different from the free space emission owing to limited number of modes.
This last case bears multiple quantum optics consequences, e.g. laser-like collective emission
[22] due to non-Markovian interaction (the interaction that depends on its history) between
the atom and the field.
In the case of directional PBG in finite size PhCs the effect of PBG upon the spontaneous
emission becomes much less pronounced, first of all, because of the shallow variation of the
DOS spectrum in the PBG [23]. Hence, such crystals can be used, primarily, to design PhCs
with pre-defined emission indicatrix and for directional enhancement or suppression of the
emission in a pre-defined spectral range.
For the inefficient light sources with low quantum efficiency   1 , the overall relaxation
rate
tot   NR (1    ( 2 )) (0.12)

is defined by the non-radiative transitions  NR . Then, the emitter power

I  P rad /  NR (0.13)

must be proportional the radiative emission rate and correspondingly, to the LDOS as well
as the pumping power P . This case is applicable to all experiments described below. It
should be emphasised that the power emitted in the narrow homogeneous emission band is
proportional to LDOS only if the emission bandwidth is much narrower compared to the
spectral interval of the LDOS variation. In the opposite case, namely the emitter with the
broad inhomogeneous bandwidth, the LDOS spectrum should be acquired. One way to
70 Nanocrystals

obtain the LDOS spectrum is to compare the emission of two samples, one PhC-based and
one non-PBG reference that possess the same channel of non-radiative recombination [24]:

I PhC PPhC  radPhC D ( )

   K radPhC (0.14)
I ref Pref  rad .ref Drad .ref ( )

This is the true spectrum only in the case of quadratic DOS spectrum in the reference sample
and the emitter localisation in the PhC along the constant LDOS surface. In the opposite and
more frequent case of the broad distribution of emitters over the unit cell volume, this
method produces the correct estimate of specific extrema in LDOS spectrum.

3. Experimental technique
The opal crystals used for impregnation with colloidal NC are the thin film crystals
prepared by either sedimentation or by crystallization in the moving meniscus. This
technology is well established and documented [15]. The current improvement of
crystallization methods aims at eliminating the cracks of the film [25] and better crystallinity
[26]. Typical example of the opal film is shown in Fig.2. The symmetry of the opal lattice is
very close to the face centred cubic (fcc) symmetry (Fig.3). Opal possess the open porosity
that allows embedding the semiconductor NCs in voids between touching spheres
(Fig.3a,b). In order to describe the light propagation in the PhC it is convenient to use the
lattice representation in the reciprocal space (Fig.3c). If the optical data are collected along
the normal to the opal film, this direction corresponds to the L directions in the 1st
Brillouin zone. The dashed line on the surface of the Brillouin zone is the line along which
the data were obtained along the oblique direction to the film normal.

Fig. 2. Scanning electron microscope images. Top view (left panel) shows the (111) plane of
fcc lattice at the film surface and the film cracks on the right hand side and side view (right)
shows the film cross-section. The opal film is crystallised in the moving meniscus from
PMMA spheres of 368 nm in diameter.

Transmission/reflectance spectra of the opal films were measured under white light
illumination from a tungsten lamp. The transmitted/reflected light was collected within a
solid angle of approximately 2o along different directions with respect to the [111] axis.
Emission of semiconductor nanocrystals in photonic crystal environment 71

The energy band structure of the opal film was calculated under assumption of the fcc
lattice symmetry (Fig.3d). This diagram relatively closely corresponds to the PBG structure
revealed by the transmission spectra of the opal film measured at different angles of light
propagation (Fig.3e). Bragg law approximation allows to associate the most pronounced
resonances in these transmission spectra with diffraction at crystal planes.
PL measurements were performed under continuous wave (cw) excitation from an Ar+ gas
laser (Fig.4). Typically, the beam was focused in 0.1 mm in diameter spot. Two schemes
were used for PL excitation/collection – the back window, in which the PL is collected from
the sample side that is opposite with respect to illuminated side, and the front window, in
which the PL signal was measured from the same illuminated side of the sample (Fig.5). The
PL was typically collected from a solid angle   5o . PL spectra were recorded, when the PL
intensity becomes stable after each change of the excitation power. In order to trace the
anisotropy of the emission, PL spectra were measured at different angles  with respect to
the [111] axis of the opal lattice. In the case of an array of randomly oriented dipoles smaller
than the wavelength, the averaging over the sample volume results in an isotropic light
source. In this case, the distortion of the spherical wavefront of an isotropic PhC-embedded
emitter is a direct consequence of the PBG. In what follows, we will refer to the fraction of
the wavefront that is blocked for propagation by the first bandgap as the Bragg cone.

Fig. 3. (a, b) Schematics of the fcc lattice fragment. The {111} family of planes is represented
by a tetrahedron. The interstitial voids between touching spheres are clearly seen. (c)
Brillouin zone of the fcc lattice. Letters show the main symmetry points. Numbers at arrows
name the directions in the reciprocal space in correspondence to the lattice axes in the real
space. (d) The energy band structure of the opal crystal assembled from PMMA spheres of
368 nm in diameter (see Fig.2). The ellipse marks the typical range of this diagram, which is
relevant to the discussion in this chapter. (e) Example of experimental transmission spectra
of the opal film from 368 nm spheres measured under s-polarised light [27]. Numbers are
the Miller indices of fcc crystal planes, the diffraction at which corresponds to the
transmission minima (compare to panel (d)).
72 Nanocrystals

Fig. 4. (a) Schematics of the PL typical measurement set-up – cw excitation and lock-in-
based registration detection. (b) Improved excitation conditions to allow constant size of
illuminated spot to be preserved while changing the detection angle.

Fig. 5. Schematics of PL measurements in transmission (a) and in reflectance mode (b).

4. Directional suppression of CdTe nanocrystal emission in thin opal film

The opal films for this experiment were prepared from 1% aqueous colloidal solution of
latex spheres of D=240 nm in diameter dried on glass slides (1 cm2). Films crystallize in the
randomised fcc lattice, which has the [111] axis as the growth direction. Subsequently, films
with a thickness of 20-30 m were sintered for 2 h at 100oC. CdTe core-shell NCs were
synthesized as described elsewhere [28]. A polymer shell was used to prevent the
agglomeration of colloidal particles. Infiltration of CdTe colloidal NCs into an opal film was
performed by dipping the latter in 0.02 M (referring to Te) CdTe NC aqueous colloidal
suspension for 1 min. As a result of electrical charging of these polymer shells, the CdTe
NCs are attached to the surface of latex spheres (Fig.6a).

(a) (b)
[111]
sphere
laser opal 
beam film
void

LB
NCs l* PL

d
Fig. 6. (а) Schematics of NC CdTe layout in the octahedral opal void. (b) Characteristic
length scales that are relevant to the formation of the emission spectrum of opal-embedded
NC. (c) The spatial distribution of emission from a point source located inside the opal
lattice. View along [111] axis (   0o ). Black holes are the Bragg cones. Courtesy of D.
Chigrin.
Emission of semiconductor nanocrystals in photonic crystal environment 73

The transmission spectrum of the opal film demonstrates the minimum centred at 2.23 eV
(Fig.7a), which manifests the directional (111) bandgap. The position of this minimum
corresponds to the Bragg diffraction resonance at the stack of (111) planes in the fcc lattice
(Fig.3b)  B  2 c /(2  neff  0.816 D) , where neff is the effective index of refraction obtained
from the effective medium approximation to the opal lattice and c is the light velocity. The
relative bandwidth of the transmission minimum E / EB  0.067 exceeds by 20% the
gapwidth calculated for ideally packed opal [29], which is an indication of lattice disorder.
Impregnation of the opal with CdTe NCs leads to the “red” shift of this minimum by 0.05 eV
due to increase of the refractive index and the transmission reduction at   2.5 eV , i.e.,
above the absorption edge of CdTe NCs. The CdTe fraction can be estimated from this shift
as 1 to 2 volume % (for different samples) or up to 4% of the void volume. Important, that
impregnating the opal with NCs does not destroy the optical quality of the opal-based PhC.
The spectral position of the transmission minimum changes rapidly with changing the
incidence angle of the light beam with respect to the film normal according to the Bragg law.
Due to the destructive influence of opal crystal defects the transmission attenuation in (111)
resonance gradually decreases with the angle increase   70o (Fig.3e) [26].
The relatively narrow linewidth of the NC emission (Fig.7b) compared to the PBG width
would not allow for tracing the emission change at different overlaps with the PBG. But the
PL bandwidth of NC in the opal appears much broader due to NC interaction with the inner
opal surface. Moreover, the PL bandwidth in the CdTe-opal was broadened due to partial
degradation of the NC luminescence after intense laser light illumination. On a later stage
the excitation power was kept below 6 mW to avoid further degradation of NC emission.
The PL spectrum of CdTe-opal collected at = 70o represents the emission of CdTe NCs in
the latex opal without influence of the directional PBG (Fig.7b).

1
opal 1.0 CdTe CdTe-opal 1.0 CdTe-opal
1.0
normalised transmission

normalised PL intensity

normalised PL intensity
PL intensity (arn.units)

CdTe-
opal 0.8 2
CdTe- 70o
opal 0.6 3
0.5 x100

0.1 0.5 30o 0.4

opal
0.2 1
(a) (b) 0o (c) (d)
0.0 0.0
2.0 2.2 2.4 2.6 2.0 2.4 2.8 2.0 2.2 2.4 2.6 2.0 2.2 2.4 2.6
energy (eV) energy (eV) energy (eV) energy (eV)

Fig. 7. (а) Transmission spectra at   0o of the latex opal assembled from spheres of
D  240 нм and the opal impregnated with NC CdTe. (b) PL spectrum of NC in water
suspension (dashed line) in comparison to PL spectra of the bare opal (thin line) and CdTe-
opal (thick line) obtained at   70o under excitation by 351 nm line of Ar+-laser with
1.9 mW power in a spot of 0.1 mm in diameter and collected from a 5o wide solid angle. The
PL intensity of bare opal is >100 times weaker in magnitude compared to that of CdTe-opal.
(c) PL spectra of CdTe-opal at different angles of collection. (d) Comparison of the CdTe-
opal PL spectrum obtained at   0o (curve 1) and the reconstructed spectra obtained by
multiplication the transmission spectrum and PL spectrum at   70o (2) and the same as (2)
but with the account taken for the fraction of non-modified emission (3).
74 Nanocrystals

If the bandgap is present, the emission flow is PBG-blocked along the Bragg cones (Fig.6c)
and the central frequency of the PL minimum coincides with the transmission dip. The
angle dependence of PL spectra is clearly seen from comparison of spectra collected at
andFig.7cThis angle dispersion of PL minimum follows that of the transmission
minimum. In particular, in the Brillouin zone, the   30o corresponds to the shift along the
LU line from the L towards U direction (Fig.3c).
One can notice that the intensity contrast for the PL dip is reduced by a factor of two
compared to the 6-fold reduction in the transmission minimum. To understand this we have
to separate the measured emission flux in the ballistic and the diffuse components. The
diffuse background is comprised by photons, which experience scattering at lattice defects
[30]. In thin film opals the mean free path of photons, l*, is about 15 m that is shorter the
film thickness (Fig.6b), i.e. scrambling of trajectories of photons emitted at the distance from
the film edge l>l* is expected at any detection angle. Moreover, the scattering results in
progressively shallower dip at higher angles of detection due to the longer light path.
Another source of unstructured emission is the near-surface emission that comes without
attenuation and fills in the PBG minimum. In the bulk opals the latter contribution can be
eliminated by bleaching the emitters in the near-surface zone [31], whereas in the thin film
opals one can use the photonic hetero-crystal approach (see section 8).
The unstructured contribution to the PL spectrum can be quantitatively estimated using the
spectrum of unmodified emission at   70o and the transmission spectrum at   0o [32].
The result of I PL (  0)  I PL (70)  T (0) is shown by curve 2 in Fig.7d, which overestimates
the emission suppression. More accurate fit (curve 3) can be obtained taking into account
the diffuse light I PL (  0)  3I PL (70)  T (0)  0.25 I PL (70) . This consideration proves the
substantial diffuse fraction in the light detected along the PBG direction.

1 1
normalised relative PL intensity

1
0.1 0.141; 3.17
2 0.9 =30o
normalised transmission
normalised PL intensity

PL intensity (arb.units)

0.8
2
0.01
0.7
1 T
0.1 0.6
1E-3

1 0.1 0.5
1E-4 0.076; 3.97

0.4
(a) (b) 0.043; 4.16 (c)
0.01 1E-5
1.8 2.0 2.2 2.4 2.6 1.8 2.0 2.2 2.4 2.6 1E-3 0.01 0.1 1 10
energy (eV) energy (eV) excitation power (mW)

Fig. 8. (а) PL spectra of CdTe-opal with ~1 vol.% (1) and ~2 vol.% (2) CdTe fraction at
  0o . (b) Relative PL intensity spectra of CdTe-opal at low (1) and high (2) CdTe fraction
in comparison to transmission spectrum. (c) Input-output characteristics of PL intensity
acquired at   30o and   2.5, 2.3 and 2 eV (squares, circles and triangles, respectively).
Points –experiment, lines – two-parametric approximation, numbers - fit parameters I 0 and
P0 , respectively.

Next test was made to check if the NC-opal spectrum depends on the NC concentration.
Obviously, only low volume concentration was explored, because with high fraction of NC
the PBG properties of the samples will be altered dramatically due to changing the refractive
index contrast and the uneven NC distribution over the opal voids. Fig.8a demonstrates the
Emission of semiconductor nanocrystals in photonic crystal environment 75

almost 2-fold increase of the PL intensity followed the doubling of NC concentration.

Moreover, the PBG attenuation in the relative PL spectrum remains almost the same
(Fig.8b). It is worth noting that obtaining the relative PL spectrum, as it was suggested in
early works [33,34], is the very useful method for revealing the PBG effect upon the
emission of PhC-coupled light sources especially in the case of complex spectra and weak
attenuations.

5. Stimulation of CdTe nanocrystal emission in thin opal film

In the case of an externally pumped emitter inside a PhC of finite size, the field of the
photonic mode becomes a superposition of the outgoing waves and the waves reflected
from the PhC boundary. For the PBG frequencies the light intensity decays exponentially
with distance z into the photonic crystal. Such photonic mode has the form of a standing
wave with an envelope function that decays exponentially as exp (−γ z) , where γ is the
extinction coefficient. Thus, this mode in an opal film is a standing wave formed by
interfering evanescent Bloch states inside the crystal and a plane wave in the vacuum.
For a microscopic emitter in the PhC the amplitude of the external field will depend on the
distance from the boundary of the PC and will be determined by (i) the strength of the light
attenuation in the photonic crystal and (ii) the position of the emitter relative to the nodes
and antinodes of the standing evanescent Bloch wave (LDOS). Note that attenuation of light
always leads to a decrease of the field and also of the emission rate while the position of the
emitter relative to the standing wave of the field can decrease (in the node) or increase (in
the antinode) the emission rate. The experimental results show the intensity suppression in
the PBG frequency range. As was shown, within the PBG the PL is composed of light
emitted at different distances from the PhC boundary. It can be concluded from the
transmission spectra that the attenuation length corresponding to the centre of the band gap
is about 10 μm (6 μm for an ideal structure) [35]. Thus, only  1/ 3 of the 30 m thick opal
film contributes to the PL intensity observed externally (  1/ 5 for the ideal lattice). These
estimates are in good agreement with the measured relative PL intensity, shown in Fig.7b,
which exhibits a decrease of PL intensity by a factor of five.
The PL intensity of the NC-opal as a function of the excitation power can be represented by
input-output characteristics (Fig.8c). These characteristics exhibit saturation with increasing
excitation power for all explored frequencies and can be fitted to the expression

I PL  I 0 (1  exp( P / P0 )) , (0.15)

where P is the excitation power. In this fit the pre-factor I0 is the power radiated by a
saturated emitter and the parameter P0 in the emission saturation threshold. Since these
parameters acquire unique values for a given frequency and angle of detection, they can be
represented in a spectral form. The I 0 ( ) spectrum at 70o is a monotonous function of
frequency (Fig.9c). It is measured in PL intensity units and closely resembles the PL
spectrum. The parameter P0 ( ) is given in excitation power units and corresponds to the
projected excitation leading to a complete saturation of the input-output curve.
76 Nanocrystals

0.16 0.06
0.16

1.2 4.0
I0 (arb.units)
0.05 0.24
0.12 0.12

P0 (mW)
=0o =30o =70o
0.04
0.08
0.08
0.8 3.2 0.03 0.20
0.04
0.04
(a) (b) 0.02 (c)
1.8 2.0 2.2 2.4 2.6 2.8 1.8 2.0 2.2 2.4 2.6 2.8 1.8 2.0 2.2 2.4 2.6 2.8
energy (eV) energy (eV) energy (eV)

Fig. 9. Spectra of parameters I 0 (open circles) and P0 (circles) at three detection angles
  0, 30, 70o . PL spectra are shown by lines for comparison.

The I 0 ( ) spectra at = 0o and 30o follow the canvas for that at = 70o with the exception
of the clearly resolved minimum superimposed on the monotonous background (Fig.8a,b).
Moreover, I 0 ( ) spectra closely resemble the PL spectra. By contrast, the spectra of the
saturation threshold P0 ( ) at = 0o and 30o have their maxima in the bandgap. In
particular, P0 peak is located at the low frequency edge of the bandgap, covers the whole
bandgap range and follows the bandgap angular dispersion. The P0 magnitude is nearly
doubled in the gap along the [111] axis, but its resolution becomes worse with increasing
detection angle. Such degradation correlates the decrease of ballistic component in the
detected emission flow because, neglecting mode re-coupling at the opal-air boundary, the
emitter in the ballistic limit radiates in the same mode as detected outside the PhC.
Positioning of emitting NCs along the “heavy” dielectric boundary aligns them with the EM
field distribution in the unit cell of the opal that ensures sampling by all NCs the same
LDOS. In the case of a saturated emitter and in the presence of an effective non-radiative
recombination channel, the radiated power is proportional to the spontaneous emission rate
to a given mode (0.13). Therefore, the I 0 ( , ) is an estimate of the spontaneous emission
rate spectrum along a given direction. The good correlation between I 0 ( ) and the PL
spectrum for a given direction (Fig.9) suggests that the spontaneous emission is the
dominating process in the radiative relaxation in the opal-embedded NCs.
P0 measures the range of the emission response to the pump power increase. For a given
number of NCs, it depends on several factors. One is the probability of an electron transition
between two bands in a NC band structure that is constant in this experiment. Another is
the population of these bands. The others are the probability of coupling the emitted photon
to the optical mode reservoir of the PhC and the balance between radiative and non-
radiative relaxation. At = 70o the P0 ( ) is dominated by processes in the electronic
system of the CdTe NCs (Fig.9c).
The weak variation of the saturation threshold across the emission band corresponds to a
uniform density of electrons as a function of energy. Such distribution is the result of non-
resonant excitation and fast non-radiative relaxation of the electronic excitations. The CdTe
NCs in the opal suffer the surface effects that have an extremely strong influence on the
electronic structure. In particular a wide "impurity" band is formed due to the surface states
(which can be Tamm-like states, surface defects, and impurity atoms localized at the
surfaces). Relaxation of electrons and holes within this large energy band has essentially a
Emission of semiconductor nanocrystals in photonic crystal environment 77

hopping character [36,37,38], that provides quite a uniform energy distribution of carriers
within the band. Roughly speaking, an electron jumps from one localized state to another
with almost the same probability for the states of similar or markedly different energies.
Due to Auger processes [39], which are extremely efficient near surfaces, the hopping
relaxation is likely to happen. Thus the fact that there is similar character of the dependence
of the PL intensity on the pumping for the frequencies below and above the PBG indicate
that the energy distribution of carriers is almost pumping-independent in provided
experimental conditions.
In turn, the sublinear character of the input-output characteristics suggests that the rate of
non-radiative recombination grows super-linearly with the increase of carrier concentration
n. This is typical for some processes, e.g., the Auger recombination rate is proportional to n2
[39]. We can roughly estimate the influence of non-radiative recombination and variation of
radiative lifetime on the dependence of PL intensity on pumping. Assuming for simplicity,
that PL intensity is proportional to the carrier concentration n and can be characterized by
“mean radiative lifetime” τ (the rate of radiative recombination is n/τ), we obtain that at the
equilibrium ¶n / ¶t = 0 the excitation is equal to the relaxation

n
P= + GNR (n ) (0.16)
t

Within the PBG, the PL intensity is influenced by the increase of radiative lifetime τ. We
may then ask how does such increase will modify the dependence of I on P in the system
with spontaneous recombination and a super-linear dependence of GNR (n ) on carrier
concentration? Taking the derivative of the pumping P in eq. (0.16) with respect to the PL
intensity I and noting that I = n / t , we obtain

¶P ¶G ¶I 1
= 1 + t NR or = (0.17)
¶(n / t ) ¶n ¶P ¶G
1 + t NR
¶n

Taking into account that ¶GNR (n )/ ¶n > 0 , we find that the increase of radiative lifetime τ
leads to a slower growth of the PL intensity with increased pumping. In other words, the
saturation threshold for a frequency within the PBG has to be less than that for a frequency
outside the PBG. But this conclusion contradicts the observed PBG-related peak of P0 (w)
(Fig.9). To resolve this conflict one has to assume the presence of the amplified spontaneous
emission in addition to the spontaneous one. The acceleration of the emission rate is a
consequence of applying resonant conditions upon the emitter at a certain frequency. In
what follows we will discuss two realizations of resonant modes in opal PhCs, bearing in
mind that the modes of the allowed band are propagating ones, whereas the modes of the
Bragg cone are the leakage modes of a PBG resonator. Nevertheless, since these resonance
conditions are intrinsic to the PhC, we can postulate the omnipresence of the emission
enhancement in the incomplete PBG.
78 Nanocrystals

5.1. Coupling to slow propagating modes

One source of the emission amplification in PhC can be associated with slowly propagating
modes. The complex topology of iso-frequency surfaces in an incomplete PhC gives rise to
beam steering effects [40,41,42]. As a result, the actual direction of the energy flow inside a
PhC does not necessarily coincide with the mode wavevector k, i.e., photons emitted with
different wavevectors can propagate along the same direction. The iso-energy surface in k-
space of the opal lattice in the PBG range contains eight necks, two per each [111] axis
(Fig.6c, 10a) [43]. When this surface crosses the boundary of the 1st Brillouin zone (BZ), the
normal component of the group velocity vector v vanishes. This means that all eigenmodes,
the wavevectors of which end up at the intersection of the neck with the zone boundary,
have the v pointing along the BZ boundary. Due to the topology of the 1st BZ, some
eigenmodes with wavevectors along the [ 111 ] direction and group velocity pointing in the
[111] direction are expected. Because the direction of energy transport in a non-absorbing
PhC coincides with the group velocity vector, there should always be some energy flux in
the Bragg cone. In what follows, we will refer to modes with wavevectors parallel to the
group velocity vector, as type 1 modes and to other modes as type 2 modes [44].

Fig. 10. (a) Iso-frequency (left) and group velocity (right) contours of opal at the frequency
within the Bragg bandgap. k and v are vectors of type 2 inhomogeneous waves pointing
along the [111] axis. (b) Group velocities of type 1 (solid curves) and type 2 (dash curves)
Bloch modes along the [111] direction. The group velocity is given in units of c.

Fig. 10b shows the calculated group velocities along the [111] axis for wavevectors of the
XULK section of the first Brillouin zone. Calculations were performed using the plane wave
expansion method [45], where the Hellmann-Feynman theorem was used to calculate the
group velocity vectors. The best fit to experimental data was obtained for the opal made of
243 nm diameter spheres of neff = 1.6 . It is instructive to separate contributions to the
energy flux from type 1 and type 2 eigenmodes. For frequencies below the Bragg bandgap,
the flux along the [111] direction is solely formed by type 1 modes of the 1st and 2nd photonic
bands. Within the PBG, only type 2 modes of the 1st and 2nd bands contribute to the flux.
Above the PBG, the flux is composed by the type 1 modes of the 3d and 4th bands as well as
by the type 2 modes of the 1st and 2nd bands. For sake of clarity, contributions of the latter
are omitted in Fig. 10b, because their frequencies are above the Bragg gap.
The group velocity vanishes at the low frequency bandgap edge in the ballistic limit of an
infinite PhC and then grows slowly with increasing frequency, comprising, in average, 1/10
of the group velocity absolute value outside a bandgap. Correspondingly, type 2 modes
traverse the opal slowly and interact with the pumped medium for longer time. In
agreement with this model, the slowing down of the group velocity towards the low
Emission of semiconductor nanocrystals in photonic crystal environment 79

frequency bandedge is the reason for the “red” shift of the P0-spectrum maximum with
respect to the pseudogap centre (Fig. 9a).
The relative number of type 2 modes is proportional to the ratio of the small solid angle in k-
space and the corresponding solid angle in real space:

N ~  d k / d  (0.18)

where the summation is taken over all contributions to the energy flux [43]. Since the type 2
modes originate at the necks of the dispersion surface, the ratio of the solid angles in (0.18) is
below 4% over the bandgap. Correspondingly, the contribution of the stimulated emission
to the total PL signal is small.

5.2. Coupling to defect modes

The available in the Bragg cone modes are the leakage modes of a PBG resonator. An
obvious example of a resonator inside the opal is a lattice defect with its mode in the PBG.
The quality factor of defect modes depends on the correlation between the localization
length lloc and the opal film thickness t , the distance to the opal boundary, the detuning of
the resonant frequency from the PBG centre and the coupling of this resonance to similar
defect modes. The radiation coupled to the eigenmode of a single defect is localized in the
defect vicinity. In this case, the photon occupation number of this mode and the
corresponding strength of the EM field increases under the cw pumping. Consequently, the
radiated power acquires a super-linear component due to the backreaction of the emitted
radiation upon the radiative transition probability. In the directional PBG, this effect also
acquires the related anisotropy. In an opal with relatively high defect concentration, the
defect modes are coupled throughout the film and the resulting quality factor is diminished.
It is instructive to demonstrate the presence of amplified spontaneous emission simply by
plotting the ratio of PL spectra acquired at different excitation power. Such ratio spectra
show a peak that corresponds to a dip in transmission spectra indicating that PL intensity
grows faster with the increase of pumping inside PBG than outside. There is a shift of the
maximum in the ratio spectra to higher photon energy with the detuning of angle of
incidence from [111] (Fig.11).

1
=0o =20o
normalised transmission

4.6

4.4
ratio

4.2

4.0
(a) (b)
0.1
3.8
2.0 2.2 2.4 2.6 2.0 2.2 2.4 2.6
energy (eV) energy (eV)

Fig. 11. The ratio of PL spectra acquired at excitation powers 2.54 and 0.49 mW vs.
transmission spectrum at   0 and 20o in panels (a) and (b), respectively. The
measurements were performed at T=18K in order to reduce the non-radiative relaxation
probability.
80 Nanocrystals

Faster growth of PL intensity inside the PBG is clearly seen also in Fig.12 (a, b). In Fig.12c
one can see that in the PBG spectral range the PL intensity increases much faster with more
intensive pumping than outside the PBG. Moreover, this effect gets stronger with the
increase of the absolute value of pumping, and this indicates that the functions I in / I 1,2
out

possess more complicated dependence on excitation power than a simple power law [35].

1.9
I2.2 0.26 I2.2
6.0
1.8 I2.0 I2.4 0.0943/0.0176=5.35
intensity ratio

0.24
1.7 5.5
0.49/0.0943=5.2
0.22

ratio
1.6 5.0

0.20
1.5
4.5
2.54/0.49=5.18
1.4 0.18

(a) (b) 4.0

(c)
1.3 0.16
0.01 0.1 1 0.01 0.1 1
2.0 2.2 2.4 2.6
excitation power (mW) excitation power (mW) energy (eV)

Fig. 12. Ratios of I in / I 2out and I in / I 1out at (a) frequencies w = 2.2 and 2.0 eV and (b) at
2.2 and 2.4 eV as a function of pumping. (c) Ratios of the PL spectra along [111] direction
obtained at different levels of pumping intensities: 0.943 and 0.0176 mW; 0.49 and 0.
0.943mW; 2.54 and 0.49 mW. The peak magnitude increases with increasing the power for
the same power increment.

Thus apart from spontaneous emission, which is characterized by an increased radiative

lifetime, there is a competing process that is characterized by faster dependence on P within
the PBG. Let us check if the emission coupling into localized photon states of lattice defects
which are located in the PBG can bring such effect. In disordered media a light wave can be
localized due to multiple scattering [46,47], but such localization can be more easily
achieved introducing disorder in otherwise periodic media [6]. In the latter case, the
localization of light leads to the appearance of a non-zero density of photonic states in the
PBG [48,49]. The localized modes, contrary to the evanescent modes, can have large electric
field amplitude in the bulk of PhC, and, as a result, the radiative recombination time in such
modes is much smaller. Consequently, the recombination is channelled into the localized
modes which act as photon reservoirs. According to the Einstein principle, the radiative
recombination in a single localized mode can be written as

n n
I loc   s ( P) (0.19)
 loc  loc

where  loc is the spontaneous radiative recombination time and s is the number of photons
in the mode. The first and second terms in Eq.(0.19) is the spontaneous and stimulated
emission, respectively. In the bulk homogeneous medium at moderate excitations (in case of
lasers – below lasing threshold) the relation s  1 holds, and therefore the stimulated
emission can be neglected. In case of disordered PhC, the photon accumulation in the
localized modes can lead to substantial enhancement of stimulated emission. It is obvious
that the second term in Eq.(0.19) depends on the pumping power superlinearly. Thus, the
experimentally observed increase of parameter P0 or the peak in ratio of PL intensities at
Emission of semiconductor nanocrystals in photonic crystal environment 81

different pumping intensities inside PBG can be explained by stimulated emission through
the localized modes.
On the other hand, when a light-amplifying material is inserted into a PC, each localized
state can be considered as the analogue of a laser mode. For laser modes the dependence of
the PL intensity on the pumping has an abrupt change at threshold. Each laser mode is
characterized by its own decay time (or Q-factor) and spatial distribution of the
electromagnetic field. In the case of localized defect modes, the spatial distribution of the
field can be altered by varying the pumping [50]. Therefore, the threshold intensity is
different for different modes. One can conclude that the dependence of the PL intensity
upon the pumping averaged over different modes should be some superlinear function, and
could explain the faster growth of I ( P) within the PBG.
Another disorder-induced modification of the observed PL intensity is related to the
scattering of the radiation in the localized mode into a propagating mode. Such a process
makes possible the propagation of an emitted photon from the bulk of PhCs to the boundary
without attenuation. Consequently, it leads to an increase in the effective thickness of the
layer contributing to the detected PL intensity.

6. Emission indicatrix
The emission conditions for dipoles change as soon as the pseudogap overlaps with the
emission band (Fig.13a). In the first approximation, the spatial pattern of emission changes
according to the shift of the Bragg cone (Fig.6c). In the angle-resolved measurements, the PL
intensity for the given direction is proportional to the part of the total radiated power, which
is coupled to modes, whose group velocity is within the solid angle selected by the detector
aperture. Following procedure presented in [51], one can introduce the radiated power per
solid angle in a coordinate space dP / d  . This power gives a rate at which the dipole
energy is transferred to modes with a group velocity pointing to the observation direction.
The fact that due to the beam steering phenomenon several eigenmodes with different wave
vectors, kn , can have parallel group velocity vectors, is taken into account by extra
summation,  , over all wave vectors for which xV ˆ nk  0 holds. x̂ is a unit vector of the
observation direction.
Dipole moments of NC are randomly distributed in space. Consequently, the radiative
power should be averaged over the all dipole moment orientations, hence, an ensemble of
NC is equivalent to the set of point sources producing an isotropic distribution of wave
vectors. An angular distribution of radiative power inside a PhC depends on the topology of
the iso-frequency surface of the crystal at the emission frequency. A schematic view of the
iso-frequency surface of an opal PhC is presented in Fig.13b for a frequency inside the Bragg
gap. To plot the iso-frequency surface, one should calculate a PBG structure for all wave
vectors within the irreducible Brillouin zone and then solve the equation nk  0 for a
given frequency 0 . The iso-frequency surface at the Bragg PBG frequency of the opal
deviates from a sphere mostly along the [111] axes, where the Bragg gap openings are
located. In the vicinity to openings the iso-frequency surface forms a neck with the
alternating negative and positive Gaussian curvature separated by parabolic lines with
vanishing curvature. A small Gaussian curvature formally implies bunching of many Bloch
82 Nanocrystals

eigenwaves with different wave vectors travelling in the same direction due to the crystal
anisotropy. Such concentration of radiated power along certain directions is a linear
phenomenon and called the photon focusing [52].

Fig. 13. Fragment of the PBG diagram along ΓLU cross-section. Frequency scale is adjusted
to the discussed sample. (b) Illustration of the 3-dimensional iso-frequency surface of the 1st
optical mode of an opal PhC for a frequency 2.2 eV inside the Bragg PBG inserted in the 1st
BZ of the fcc lattice. (c) Cross-section of the BZ with iso-frequency contours at frequencies
2.08, 2.2 and 2.27 eV (compare to the diagram in panel (a)).

In Fig.13c the iso-frequency contours are presented for three different frequencies. They cut
the Brillouin zone through the symmetry points X, U, L and  One can assume that the
experimentally measured far-field PL intensity represents the signal, which is averaged over
different Brillouin zone cross-sections due to lattice disorder.

Fig. 14. Wave contours corresponded to frequencies 2.08, 2.2 and 2.27 eV (a, b, c
respectively). Only the wave contours corresponded to the first and third bands are shown
in the diagrams (a, b and c), respectively. Light grey region in (b) is the Bragg cone
projection. A dashed circle is a wave contour in vacuum. Group velocity is plotted in the
units of the speed of light in vacuum.

As the frequency remains below the gap, an iso-frequency contour is continuous and almost
circular. The Gaussian curvature does not vanish for any wave vector. This implies a small
anisotropy in the energy flux inside the crystal. It order to obtain the wave contour in
coordinate space, one should plot a ray in the observation direction x̂ starting from the
point source position and having the length of the group velocity Vnk . The calculation of
the group velocity was discussed earlier in relation to schematics in Fig.10. The wave
contour at 2.08 eV is single valued function of observation direction (Fig.14a). Fig.15 shows
Emission of semiconductor nanocrystals in photonic crystal environment 83

the angular distribution of the radiated power for the same frequency. The latter is nearly
isotropic and resembles, reasonably, the Lambert law (dashed line). To calculate the
radiated power, one should sum over all optical modes, which are available at the given
frequency.

Fig. 15. Angular distribution of the radiated power for the same frequencies as in Fig.14. The
radiated power in free space (Lambert law) is shown by dashed line for comparison.

With the frequency increase up to the midpoint of the first PBG, the topology of the iso-
frequency contour abruptly changes. The gap developed along the L direction and the iso-
frequency contour becomes open (Fig.13c). This topological discontinuity results in a
complex contour with alternating regions of different Gaussian curvature. Vanishing
curvature leads to the folds of the wave contour (Fig. 14b). The folds in the wave contours
yield that two Bloch modes are travelling in any observation direction outside the gap.
Contributions from both modes should be taken into account, if the total radiated power is
calculated. In the middle panel of Fig.15 the strongly anisotropic angular distribution of the
radiative power corresponded to the 2.2 eV is presented, featuring zero intensity in the
direction of the gap, and infinitely high intensity spikes in directions of folds of the wave
contours occurring at, approximately,   40o with respect to the [111] axis.
When the frequency crosses the upper boundary of the L Bragg gap, i.e. the gap along L
direction closes, the 3rd and the 4th photonic modes come in consideration. In the Fig. 14c a
wave contour of the 3rd band at 2.27 eV is depicted. It is single valued function elongated to
the [111] axis, which leads to the preferable energy flux in this direction. An angular
distribution of the radiative power at this frequency shows infinite intensity spikes at
  40o and a Gaussian-like lobe centred at   0o (Fig.15, right). An infinite intensity
corresponds to the points of vanishing Gaussian curvature of the 1st and the 2nd photonic
bands, while the central lobe is associated with the 3rd and the 4th bands.
Fig.16a shows experimentally obtained directionality diagrams in the vicinity to the Bragg
gap. In the first approximation, these patterns satisfy the iso-frequency profiles (Fig. 13c)
being the superposition of the isotropically distributed emission from a point source and the
Bragg cones of the fcc lattice (Fig.13b). The next iteration takes into account the light
focusing occurring due to uneven curvature of iso-frequency contours in a sense of
diagrams in Fig.15. In fact, the emission from a thin film opal consists of isotropic
unmodified (scattered and near-surface emission) and PBG-moulded ballistic components.
The isotropic light source provides the background, the intensity of which varies as cos ,
and the ballistic component carries the fingerprint of the Bragg gap. Summing the diagrams
of ballistic and diffuse components, the details of the experimentally obtained emission
indicatrix can be accurately explained [53] (Fig.16b).
84 Nanocrystals

1.0
2.27eV (b)

normalised Intensity
0.5

simulation
experiment
0.0
-90 -60 -30 0 30 60 90
angle 

Fig. 16. (а) Directionality diagrams of the emission at specified frequencies across the PBG
spectral range. Dashed line shows Lambertian distribution of emitted power. (b)
Comparison of calculated and experimental emission directionality diagrams at
  2.27 eV . Simulation includes emission focusing along special directions in the vector
diagram. To match the experimental pattern a PBG-unaffected emission component has
been added to the simulated pattern as the cos background.

7. Emission of nanocrystals in inverted opals

Reduction of the volume fraction of “heavy” dielectric compared to that in opals assembled from
spheres favours the wider PBG opening [54]. The inverted opals prepared by impregnation of
the opal voids with another dielectric and subsequent removing the opal spheres offer such
enhancement of PhC performance. Therefore, it is reasonable to examine the luminescent
properties of the NC CdTe in the inverted opal. Usually, the semiconductors are used for opal
inversion. The close vicinity of NCs to the surface of semiconductor brings another complication
to the radiative energy relaxation due to energy transfer between the NCs and the
semiconductor. In this section we will consider both aspects of NC to PhC interaction.
TiO2 inverted opals were prepared by templating in thin film opals. As a first step, opal
films consisting of approximately 15-20 monolayers of 300 nm diameter monodisperse
polymethyl methacrylate (PMMA) beads were prepared by slow-drying of a sphere
suspension on microscope slides. Then these films were dipped into a solution of TiCl4 in
HCl, followed by a moisture-promoted hydrolysis and heating at 160°C for 1 h. At this
stage, an interconnected TiO2 framework is formed in the voids between the PMMA beads.
After dissolving the polymer beads in tetrahydrofurane, TiO2 inverted opal films were
formed (Fig.17a). The reflectance spectra of the TiO2 inverted opal display a pronounced
(111) diffraction resonance indicating a directional PBG. Its angular dispersion follows the
Bragg law (Fig.17b). The strong deviation from a diffraction on (111) planes occurs around
= 40o due to the multiple wave diffraction at anticrossing of (111) and (200) resonance
dispersions [55]. The relative full width at half maximum (FWHM) of the reflectance peaks
is about 10 % (15% in transmission) compared to 5.6 % in opals [56]. Owing to a broader
gap, the spectral overlap of pseudogaps along different directions of the Bragg resonance in
TiO2 opal is remarkably larger than in PMMA template. Simultaneously, the Bragg cones are
also larger. These factors are particularly important since increasing both, spectral and
spatial PBG dimensions is a pre-condition of stronger modification of emission of PhC-
embedded light sources.
To realize a PhC light source, a small amount of colloidal CdTe NCs stabilized by thin organic
shells was embedded into the replica pores by soaking the films in dilute aqueous colloidal
dispersions. CdTe NCs of 3 nm in diameter were used to fit the emission band to the Bragg
Emission of semiconductor nanocrystals in photonic crystal environment 85

PBG of TiO2 inverted opal. After drying, the CdTe NCs remained attached to the inner surface
of the TiO2 frame through electrostatic interaction with functional groups of thiol stabilizers.

550

(b)
500

 (nm)
450

D=246 nm
400 neff=1.352

200 nm (а) 10 20 30 40 50 60 70
 (degrees)

Fig. 17. (а) SEM image of inverted TiO2-opal. (b) Angle dispersion of the (111) resonance in
TiO2-opal (points) and its approximation by the Bragg law. The diameters of spheres and
the effective index of refraction extracted from the Bragg fit are shown.

PL spectra were measured at T = 300 and 18 K under cw excitation by the 457.9 and 351.1
nm lines of an Ar+ laser. The excitation power was varied between 10-5 and 10-1 W. PL
spectra were recorded in the front window configuration thus allowing the emission to
traverse the film. The emission was collected within a 4° solid angle along the direction
defined by angle of θ with respect to the [111] axis of the opal fcc lattice.
Under 457.9 nm excitation the PL spectrum at T = 300 K exhibits a band with a maximum at
2.38 eV and a FWHM of 0.26 eV. Under 351.1 nm excitation this band becomes broader and
shifts up to 2.44 eV. At T = 18 K this PL band shifts further to 2.47 eV, but its bandwidth is
~1.5 times narrower (Fig.18a). In fact, the low temperature PL band (curve 3) resembles that
observed in the NC suspension, where NCs are relatively isolated from the environment.
Such evolution of NC emission with changing the excitation conditions has been ascribed to
the NC-substrate interaction [57,58]. It is suggested that the TiO2 template provides a
potential relief due to surface defects, which co-ordinates the deposition of NCs.

1.0
(b)
normalised PL intensity

1.0
 = 0° (a)
PL intensity (arb.units)
relative PL intensity

4
0.8
transmission,

0.5 1 2 0.6
1
0.4
3
3
0.0 0.2 2
2.0 2.2 2.4 2.6 2.8 1.8 2.0 2.2 2.4 2.6 2.8
energy (eV) energy (eV)

Fig. 18. (а) PL spectra of the CdTe NC-TiO2 inverted opal at θ = 0° under 457.9 nm and 351.1
nm line excitation at T = 300 K (curves 1, 2, respectively) and at T = 18 K (3). (b) Relative PL
spectrum (curve 1) of the CdTe NCs-TiO2 opal obtained as the ratio of PL spectra at θ = 0°
(2) and 70° (3) compared to transmission spectrum (4).

The TiO2 framework itself has its electronic bandgap at about 3.1 eV thus facilitating the
absorption of the 351 nm radiation and enhancing the energy exchange with CdTe NCs. It is
further helped by the fact that the band edge emission from TiO2 is above the spectral range
of CdTe PL band. On the other hand, the polymer shells of NCs reduce the energy exchange.
Nevertheless, defect states at TiO2-CdTe interface trap the photogenerated electron-hole
86 Nanocrystals

pairs in polymer-capped NCs by, e.g., dissociation of excitons [59]. In turn, the radiative
recombination of trapped carriers leads to broadening of the observed PL band by an
amount comparable to the trap energy range. The narrowing of the NC PL band at low
temperature can be understood as the result of reducing the contribution of trapped excitons
to the emission flux due to increasing their lifetime. This observation emphasises that the
emission of CdTe NCs in TiO2-opal involves multiple-level relaxation of electronic
excitations i.e., this system cannot be treated as a two-level system.
The relative PL spectrum for =0o contains the minimum, the position of which is in good
agreement with the (111) Bragg gap in the transmission spectrum. As the reference, the PL
spectrum of the same sample obtained at θ = 70° was used (Fig. 18b). The slight dissimilarity of
pseudogaps in the transmission and relative PL spectra is due to (i) the difference in coupling
of externally (outside the PhC) and internally (inside it) generated light to eigenmodes of the
PhC and (ii) the unaffected near-surface emission coupled to free space modes.
The input-output characteristics of CdTe-TiO2-opal [60] are similar to that of the CdTe-opal
(section 5). The P0 spectrum shows a maximum at the bandgap frequencies (Fig.18). This
maximum appears more pronounced under UV-excitation because it is no longer
overshadowed by the overlapping minimum, which is blue-shifted following the shift of the
PL maximum. A comparison of P0 -spectra obtained at different angles of detection
demonstrates that the maximum follows the PBG dispersion and vanishes gradually at
larger angles θ of detection, whereas the minimum remains unchanged. The difference with
the CdTe-opal sample (section 5) is almost 2-fold increase of the PBG width.

0.90
(a) 0.95
(b)

PL intensity (arb.units)
parameter P0 (arb.units)

Transmission,
0.85
0.90

0.85
0.80

0.80

0.75
1.8 2.0 2.2 2.4 2.6 2.8 1.8 2.0 2.2 2.4 2.6 2.8
energy (eV) energy (eV)

Fig. 19. (а, b) Parameter P0 ( ) (dots) in comparison to transmission and PL spectra obtained
at   0o for excitation by (a) 351.1 and (b) 457,9 nm lines.

The interplay of the minima and maxima in P0 spectra shows that (i) the minimum
corresponds to the faster saturation of spontaneous emission in the PL band maximum that
is determined by higher matrix element for electron transitions between excited and ground
states in the electronic band structure of the NC, whereas (ii) the maximum results from the
emission stimulation due to stronger NC to optical mode coupling provided either by slow
propagating modes or localized defect modes.
The drawback of studied inverted opal is higher amount of defects compared to the opal
template. These extra defects are added due to inhomogeneous infiltration of the opal voids and
uncontrollable lattice contraction due course of annealing. Moreover, the higher refractive index
contrast in the TiO2 opal replica compared to the opal template leads to stronger distortion of the
EM field distribution around the defect. This leads to emission losses, e.g., the emission initially
radiated to PhC eigenmodes experiences higher chance to be scattered in the Bragg cone [61] and
Emission of semiconductor nanocrystals in photonic crystal environment 87

to be transported by defect modes. Since the stimulation takes place only for modes in the PBG
direction, this effect becomes buried in the diffuse radiation, leading to suppressing the P0 peak
in the PBG interval. Hence, in spite of higher strength of the PBG effect in studied inverted opals,
the emission enhancement becomes counteracted by the influence of disorder.

8. Emission of nanocrystals in hetero-opals

The main idea behind PhC heterostructures is to bring two different PhC in contact (Fig.20a)
[62]. The obvious consequence of heterostructuring is the fingerprints of two [63] or more
PBG structures in the optical properties. Such a structure contains the interface, which
implies a sharp step in the spatial distribution of the EM field. In spite of the fact that the
geometrical “thickness” of the interface is less than the light wavelength, the transition
region is formed in the vicinity of the interface due to the EM field continuity and a need of
light coupling from modes of one PhC to modes of the other. In this region, which extends
over several lattice periods, one can expect special conditions for the light propagation [64].
Hetero-opals can be used for achieving a control upon the emission characteristics [65] if one
of the opals is infilled with light-emitting NC and used to control the NC-to-mode coupling
(opal-source) and another is used to shape the emission spectrum according to its PBG
(opal-filter). It appears that the properties of such hetero-opal differ from the linear
superposition of properties of the source and filter opals due to the interface interaction [66].
However, the light coupling at the interface remains so far largely unexplored issue [67].

3.0
(a)

2.5
energy (eV)

2.0
source filter
D=240nm D=300nm
neff=1.41 neff=1.37

1.5
U L  L U
(b)
wavevector

Fig. 20. (а) Band diagrams of the opal-source and the opal-filter in contact. (b) SEM image of
the hetero-opal assembled successively from two opal films crystallised from latex spheres
of 240 and 300 nm in diameter.

The examined hetero-opal was prepared using the successive assembling of opal films [68].
Opal films studied in this work were grown on glass or quartz slides by crystallization of
polystyrene (PS) beads ranging from 240 to 450 nm in diameter. Aqueous 1-5 vol.% colloidal
suspensions of PS beads were placed in a Teflon cylindrical cell (7 mm inner diameter) and
then the solvent was evaporated under moderate flow of warm air. Typical thickness of opal
films was 5-10 m. The sintering at 90 for 1 h was applied to allow further treatment of films.
Sandwich-type hetero-opals were prepared by self-assembling the top opal film on the
surface of the bottom opal film, which was crystallized previously from beads of another
diameter. Fig. 20b shows SEM image of a hetero-opal with the bottom film consisting of
D=240 nm and the top film – D=300 nm beads. Here and below, we denote this structure as
88 Nanocrystals

the 240/300 nm opal. The abrupt interface between films indicates that the formation of the
top film proceeds independently on the geometrical profile provided by the bottom one.
The layer-by-layer (LbL) deposition technique, which is based on alternating adsorption of
layers of oppositely charged species on the surface, was used for impregnation of opal films
with CdTe NCs. LbL technique was originally developed for positively and negatively
charged polyelectrolyte pairs [69] and then extended to the assembly of polymer-linked NCs
[70]. This method was successfully applied to a variety of substrate materials with flat and
highly curved surfaces [71]. Following LbL procedure, the substrate with an opal film was
immersed for 30 minutes in a 5 mg/ml aqueous solution of
poly(diallyldimethylammonium) chloride (positively charge polyelectrolyte), then
thoroughly washed and immersed for 30 minutes in a 10-3 M aqueous solution of CdTe NCs
capped with thioglycolic acid and, thus, carried a negative charge at appropriate pH [72]. 0.2
M of NaCl was added to both solutions to facilitate a formation of smooth layers. Moderate
stirring was applied to solutions to accelerate the mass transfer. Because opal voids are
either of 0.41 or 0.23D in size and connected via 0.15D channels [11], both polyelectrolyte
molecules and CdTe NCs can easily access internal pores. We assume that the in-void
coating proceeds in the same manner as a deposition on the open surface. The LbL
procedure was repeated several times to increase the amount of deposited CdTe NCs. The
NC-polyelectrolyte layer formed on the opposite side of the slide was finally removed by
washing in acetone and ethanol.
The LbL deposition provides uniformity of the coating thickness and NC environment
(Fig.21a), as compared to a direct infiltration of NCs into opals. No leakage of NCs from the
bottom opal film back to water takes place during the deposition of the second film judging
from the absence of the luminescence of the background solution. Anchoring of NCs is also
required to prevent their diffusion to another film of a hetero-opal. The important condition
is the need for the emission spectrum of NC to overlap both PBGs in the hetero-opal.
Correspondingly, CdTe NC of 2 and 5 nm with respective emission bands centred at 2.29
and 2.08 eV were explored.

Fig. 21. (a) TEM image of 420 nm PS beads, which are LbL-coated by CdTe NCs in one
deposition circle. The NCs appear darker due to the higher contrast. (b) Layout of the PL
measurements.  - the angle of incidence and the angle of detection,  - the solid angle of
the light collection.

A straightforward consequence of the PBG anisotropy in hetero-opals is the spectral and

spatial anisotropy of the PL of NCs embedded therein. In Fig. 22 the source and the filter PL
spectra are shown in comparison to transmission spectra at different angles of the light
detection to demonstrate the correlation between the emission minimum and the
corresponding Bragg gap. It is instructive to emphasize the different physical origin of this
Emission of semiconductor nanocrystals in photonic crystal environment 89

minimum in the source and the filter PL spectra. The directional minimum of the source PL
spectrum is a consequence of the lower mode number available for coupling with emitter
within the PBG bandwidth along a given direction. In contrast, the minimum in the filter PL
spectrum appears due to the back-reflection of the source radiation from a filter film within
the PBG bandwidth of the filter, which does not directly affect the LDOS in the source. In
other words, the origin of the filter PL minimum is the same as that of the transmission
minimum. The diffuse scattering of emitted photons at the interface and in the filter film can
be the reason for the absence of the source PBG minimum in the filter PL spectrum.

1.0
(a) (b) (c)

log (normalised PL intensity)

normalised transmission

normalised PL intensity
1.5
source 0 filter

0o
1.0
0.5

20o 0o
0.5 -1
20o
40o 70o
0o 20o 40o 40o
70o
0.0
0.0
1.6 1.8 2.0 2.2 2.4 1.6 1.8 2.0 2.2 2.4 1.6 1.8 2.0 2.2 2.4
energy (eV) energy (eV) energy (eV)

Fig. 22. (а) Spectra of transmission of the 20 m-thick 240/300 nm source-filter hetero-opal
obtained along different directions that are labelled by the angle. (b,c) Source and filter PL
spectra of a hetero-opal with 2 nm NC in the source film, respectively. Curves in the panels
are shifted vertically for clarity. Dotted line shows The PL spectrum obtained at 70o that is
used as a PBG-unaffected reference.

Transmission minima for the light propagating normally to the film planes are centred at
1.81 and 2.24 eV and the FWHM of these resonances is about 9% in transmission (Fig.22a).
The transmission minima in such hetero-opals can be traced down to 60o due to the limit
applied by enhanced scattering at the interface. The PL spectra of the source and filter films
demonstrate minima in agreement with transmission spectra (Fig.22a). The relative PL
spectra from the source and filter sides of the hetero-opal demonstrate two minima for the
larger NC. These minima can also be noticed in the PL spectra of smaller NC (Fig.23).

2.0
(a) anomalous (c) (d)
1.0 1.0 (b) 1.0 1.0 20o
minimum 1
Normalized PL Intensity
Relative PL Intensity (a.u.)

1.8 40o 2.5

0o
1.6
40o 2.0
0.8
1.4 20o 0.8
0.5 0o 0.5 0o
1.5
1.2

0.6 40o 40o

1.0 1.0
0.6
0o 20o
0.8 20o anomalous
minimum 2
0.0 0.0 0.5
1.6 1.8 2.0 2.2 2.4 1.6 1.8 2.0 2.2 2.4 1.6 1.8 2.0 2.2 2.4 1.6 1.8 2.0 2.2 2.4
energy (eV) energy (eV)
energy (eV) energy (eV)

Fig. 23. Relative PL spectra of the source (a, c) and filter (b, d) at different detection angles
for NC of 2 (a,b) and 5 nm (c,d) with PL bands centred at 2.29 and 2.08 eV, respectively.
Angles are indicated at curves. Dashed PL spectra are obtained at 70o. All spectra are
normalised to achieve the same intensity at the “red” edge. Vertical dashed lines indicate the
position of “anomalous” minima.
90 Nanocrystals

One minimum in relative source and filter spectra moves towards higher energies with
increase of the detection angle in accord with the “blue” shift of the (111) Bragg gap. The
relative midgap suppression of PL intensity at =0o is about 50%. However, the second
minimum marked by a dashed line is a stationary one, which shows no shift with changing
the angle of detection (Fig. 23). Interestingly to note, that the emission attenuation in the
directional and stationary minima are comparable in the case of the 5 nm NCs, the PL
intensity of which is evenly distributed between filter and source PBG minima.
Experimental findings so far can be summarized as followed. (a) The emission from the
hetero-opal acquires the artificial anisotropy due to the difference of directional Bragg gaps.
Important, that this anisotropy is fully under control because the PBG frequencies in a bi-
layer opal directly relate to the sphere diameters. (b) The emission spectra acquire the
additional non-dispersive minimum, the position of which is correlated but does not
coincide with the directional minimum of another opal in a heterostructure. Since such effect
was not observed in single light emitting films, its appearance is reasonable to assigned to
the light coupling at the interface. The fact that the position of non-dispersive minima is not
sensitive to the change of the NC emission band supports this assumption. If so, the
spectrum transformation after internally generated light crossing the interface looks like
loosing the memory about the PBG directionality.
In order to illustrate this conclusion the angle diagrams of the emission intensity obtained
from the source and filter sides of the hetero-opal are compared in Fig.24. The diagrams of
the source change their shape like it was discussed with respect to the single light emitting
opal film (section 6, Fig.16), moreover, they are broader than the Lambertian diagram.
Oppositely, the diagrams of the filter are almost insensitive to changing the frequency and
appear much narrower compared to the Lambertian shape. Such phenomenon can be
considered as the emission focusing. Such focusing should be controlled by the difference in
opal film lattice parameters.

Fig. 24. Emission indicatrices of the normalised to the maximum emission intensity at
different frequencies from (a) the opal source and (b) the opal filter in the vicinity to (111)
resonances in the 240/300 nm hetero-opal. Dashed line shows the Lambertian diagram.

One can raise a question concerning the mechanism of the emission flux formatting. On the
one hand, the absence of memory about the source PBG points to the diffuse character of
photon propagation. However, the scattering in the volume of the source (thickness ~14 m)
is not sufficient for complete randomising of the emission flow as demonstrated by the
source diagrams. In contrast, the light crossing of the hetero-interface is considered as the
sequence of scattering of the incident light, which is transported by the Bloch modes of the
source PhC, and following coupling of scattered light to the Bloch modes of the filter PhC
[64]. The interface scattering occurs due to the symmetry mismatch between two reservoirs
Emission of semiconductor nanocrystals in photonic crystal environment 91

of modes in these PhC lattices. Hence, a fraction of the source radiation propagates to the
detector as coupled to the filter eigenmodes and the other fraction – as the uncoupled
decaying modes. Therefore, the formatting of outgoing flux proceeds accordingly to the
interface coupling conditions, moreover, it is associated with light losses. On the other hand,
the light coupling is controlled by the iso-frequencies of two PhC in contact. Fig.25
illustrates this process. In particular, the light propagation from the source with a spherical
iso-frequency surface is allowed to all but the directions in the Bragg cones of the filter. With
the increase of the emission frequency above the PBG in filter and source crystals, the
corresponding iso-frequency surfaces acquire the complex profile leading to rapid variation
of the coupling conditions with the frequency.

Fig. 25. Schematics of the iso-frequency surfaces in two PhCs with slightly different lattice
parameters and the same effective refractive index. (a) Omnidirectional coupling at
frequencies below the 1st bandgap in both PhCs. (b) Directional coupling of the emission
generated in a source crystal at the frequency below its 1st bandgap to the filter film at the
frequency within its 1st bandgap. Shaded cones are the Bragg cones.

The intensity of scattered radiation increases in the source PBG interval with the increase of
the angle of light incidence at the interface [73], hence, progressively lower flux couples to
the filter. Since scattering equalises the angle distribution of the PL intensity over all
directions bringing its spectrum in agreement to the direction-independent DOS spectrum,
this is observed in the relative PL spectra of the filter as the anomalous non-dispersive
minimum. Thus, effectively, the moving filter PBG minimum squeezes the emission
diagram towards the stationary source-related minimum leading to the efficient
compressing the filter diagram.
It is instructive to represent the emission anisotropy as the filter-to-source intensity ratio
(the anisotropy factor). Such ratio shows a minimum in the filter (111) PBG and the
maximum in the source (111) PBG, the spectral separation of which depends on the lattice
parameters of the hetero-opal crystals (Fig.26). While this ratio quantifies the emission
anisotropy, it is not possible to pin down its absolute value because the substrate introduces
the asymmetry in emission outcoupling in a hetero-opal.

0.2
0.18 1.0
(a) 1.0 (b)
Normalised Transmission

0.16 1.0
IPL filter/IPL source

0.14
0.12
0.8 0.8
0.1
0.08 0.5
0.6 0.6
0.06

0.4
0.04 0.4 0.0
1.8 2.0 2.2 2.4 1.6 1.8 2.0 2.2 2.4 2.6
Energy (eV) Energy (eV)

Fig. 26. Anisotropy factor for the source-filter hetero-opals assembled from (a) 240/269 and
(b) 240/300 nm opals along the [111] direction.
92 Nanocrystals

The shape of the anisotropy factor spectra changes dramatically with the increase of the
excitation power and the detection angle. Along the film normal, the minimum of
anisotropy factor for the filter remains unchanged along the pumping increase, whereas the
maximum for the source PBG monotonously decreases (Fig.27a). The reason of this effect is
the stimulated emission in the source (compare to Fig.12c) and the absence of amplification
in the filter opal. Along the detection angle increase this picture changes. At   20o (i) the
extrema of the anisotropy parameter shift to higher frequencies according to PBG
dispersions, (ii) the filter minimum becomes also pumping-dependent and (iii) the transition
region between resonances becomes distorted. At   40o (i) the maximum of anisotropy
parameter in the source PBG is replaced with the maximum in between filter and source
PBGs and (ii) with the increased pumping the emission anisotropy reduces by three times in
its maximum. At   70o the pumping dependence of anisotropy disappears and the
anisotropy spectrum becomes  1/ I P L ( ) (Fig.27d).

2.2
2.0 (a) (b) (c) (d)
1.8
1.6
Ifilter / Isource

1.4
1.2
1.0
20
0.8 7.7
2.2
0.6
1.2
0.4 0.06
1.6 1.8 2.0 2.2 2.4 1.6 1.8 2.0 2.2 2.4 1.6 1.8 2.0 2.2 2.4 1.6 1.8 2.0 2.2 2.4
energy (eV) energy (eV) energy (eV) energy (eV)

Fig. 27. Evolution of the anisotropy parameter with the increase of the excitation power for
different detection angles. The excitation power is specified in the legend in panel (a). Panels
(a-d) correspond to the detection angles 0, 20, 40 and 70о.
(I20 / I0.6)filter / (I20 / I0.6)source

20 20
Saturation threshold (mW)

0o
(a) (b) (c)
Saturated intensity (a.u.)

0.04 2.5 20o

40o
0.8 2
70o
15 0.03 15
1.5
I20 / I0.6

0.02
0.4 10
10 1

0.01

0.0 5
1.8 2.0 2.2 2.4 2.6 1.8 2.0 2.2 2.4 2.6 1.8 2.0 2.2 2.4 2.6
Energy (eV) Energy (eV) Energy (eV)

Fig. 28. Spectra of the saturation threshold (circles) and the saturated emission intensity (open
circles) for the (a) source and (b) filter opals at   0o . The ratio of the PL spectra obtained at
pumping powers 20 and 0.6 mW are shown by lines for comparison. The arrow in (a) indicates
the (111) source PBG. (c) Spectra of enhancement anisotropy parameter for different detection
angles. Up- and down-arrows shows PBG of the source and the filter, respectively.

Spectra of the emission anisotropy parameter obtained at 40o show that the emission
intensity passed through the interface in the spectral interval between filter and source PBGs
(centred at ~2.1 eV) becomes relatively lower with at the increase of the pumping power.
This observation reveals an important consequence of the interface coupling – the intensity
Emission of semiconductor nanocrystals in photonic crystal environment 93

of the source PL in this range increases much faster compared to that passed through the
filter. This non-linear effect can be associated with the increase of the light path in the gain
medium of the pumped source opal film. We mentioned that the transmittivity of the
interface is much lower for the oblique incident light because of increased interface
scattering with possible trapping of scattered light with PBG frequencies at the interface
[73]. Taking into account the emission spectrum having the maximum at ~2.05 eV, it is
possible to conclude that the scattered radiation is responsible for the source PL increase
and that this increase is correlated with the gain spectrum (Fig.23c).
The estimate of the emission rate changes made by using the emission saturation threshold
agrees the above picture. The peak in P0 ( ) of the source opal coincide the source PBG,
whereas no clear features is observed in the spectrum of this parameter for the filter emission
(Fig.28a,b). Similarly the spectrum of the ratio of PL intensities measured under pumping of
different powers reasonable well corresponds to the P0 ( ) spectrum (section 5.2). Then the
expression  I 20 / I 0.6  filter  I 20 / I 0.6 source allows to quantify the spectrum of the anisotropy of
the emission stimulation (Fig.28c). Since the parameter of enhancement anisotropy is almost 1
for the   70o , where no PBG influence is expected, it produces the absolute value of
enhancement. At   0o and 20o one can see the 1.5 time enhancement in the source PBG and
the emission suppression in the filter PBG. This once again points to the influencing of the NC
emission by filter PBG. At higher angles the influence of PBG becomes indistinguishable
compared to the 2.5 times enhancement of the interface-trapped emission at 2.05 eV. It is worth
noting that the development of the latter band is detectable at lower angles as well, but its
magnitude is comparable to the PBG enhancement. Looking back to Fig. 27, this enhancement
is the reason for the corresponding peak of the anisotropy parameter.

9. Modification of nanocrystal emission by the local field of colloidal crystals

The control over the emission spectrum and directionality of light sources can be achieved
by tailoring their EM environment, as discussed earlier [3,4,7,6]. So far we considered the
emission of sources embedded in opals. Such configuration leads, in the first instance, to the
PBG-related suppression of spontaneous emission due to reduction of the mode density and
pushing emission flow away from the PBG direction. Simultaneously, the emission coupled
to the resonance modes in the same PBG interval experiences stimulation. At the PhC
surface the strong local fields are developed at PBG frequencies. The local field pattern of
colloidal crystals has been visualized using the near-field optical microscopy [74,75].
Although such field exponentially decays with increasing the distance from the PhC surface,
it can promote the emission from NC within the wavelength-scale distance [76]. If so, the
fraction of the enhanced emission will be added on top to the spontaneous emission of NC
in contrast to the case of the PhC-embedded NCs, when the stimulated emission is added on
top of the suppressed spontaneous emission (Fig.29a).
The idea of the spatially separated NC-PhC light source is shown in Fig.29b. In such
configuration the dipole transition probability of an oscillator placed in between a 2D PhC
and a dielectric substrate depends on the NC location with respect to nodes and antinodes
of the EM field, hence, the emission of a light source radiated in the near-field zone of the
PhC can be either enhanced or suppressed, but the enhancement effect prevails [77,78].
94 Nanocrystals

The PBG emission enhancement should not be confused with traditional ways of increasing
the brightness of light emitting devices based on grating couplers and randomly textured
surfaces [79,80]. These techniques increase the external quantum efficiency by extracting
that fraction of emission, which is normally trapped in the structure due to the total internal
reflectance. With the increase of the refractive index contrast, such gratings can be
considered in terms of a slab 2D PhCs, in which the wave optics phenomena co-exist with
geometrical optics [81]. To date, the efficient out-coupling of radiation has been realized by
coupling the radiation to leaky guided modes with frequencies above the light cone in
emitters nanopatterned as 2D PhCs [82].

Fig. 29. Schematics of the emission spectrum modification for the light source inside a 3D
PhC (left) and at the PhC slab surface (right). Note the different origins of the transmission
minima. On the left it occurs due to diffraction at the stack of crystal planes. On the right it
occurs due to excitation of leaky guided modes at the PhC surface. (b) Schematics of the
structure consisted of a glass substrate with LbL-deposited CdTe NC and LB deposited
monolayer of silica spheres. (c) SEM image of a LB monolayer of 519 nm silica spheres on a
glass substrate. Inset: the SEM side view of LB monolayer.

From a technology point of view sandwiched NC-opal structures look simpler compared to
NC-impregnated opals. However, the light source should be thin enough to be
accommodated in the near-field zone of the PhC (Fig.20c) [83].

9.1. Transmission, diffraction and photoluminescence spectra of sandwich sources

Uniform thin films containing 2 nm CdTe NCs [84] were prepared by LbL assembly [85].
The following cyclic procedure was used: (i) dipping of the glass substrate into a solution (5
mg/ml in 0.2 M NaCl) of poly(diallyldimethylammonium chloride) (PDDA) for 10 min, (ii)
rinsing with water for 1 min; (iii) dipping into aqueous suspensions of the negatively
charged NCs (ca. 10-6 M) for 10 min; (iv) rinsing with water again for 1 min. Each cycle of
this procedure results in a ‘bilayer’ consisting of a polymer/NC composite. 20 bi-layers were
deposited in order to increase the brightness of luminescence. The thickness of a 20 bi-layer
film is about 60 nm. Since the volume fraction of NCs is about 27%, the average RI of the PE-
CdTe film is about 1.9. In what follows these films are referred as NC films (NCFs).
Silica spheres of the nominal diameter D=519 nm were hydrophobised with 3-
trimethoxysilylpropyl methacrylate. The Langmuir-Blodgett (LB) technique, with a low
barrier speed of 6 cm2 min-1, was used to compress sphere arrays in hexagonally packed
monolayers (1L) on the surface of doubly distilled deionised water and to transfer them to
Emission of semiconductor nanocrystals in photonic crystal environment 95

the substrate. Using this technique, the monolayer of spheres was deposited at a surface
pressure of 4mN/m [86] on to the LbL NCF substrate (inset, Fig.2). Subsequent monolayers
were deposited after drying the deposited ones. Both mono- and multiple-layers (up to five
layers (5L)) of the SiO2 spheres were prepared. Monolayers of spheres form the LB colloidal
PhC possessing the PBG of (2+1)-dimensionality. Although no 3D lattice is formed due to
the lack of lateral alignment between monolayers, the LB crystal remains ordered within
each monolayer of spheres and as a periodic stack of monolayers [87,88].
The PL spectra of NCFs were excited by the 457.9 nm line of an Ar-ion laser. A laser spot
size of 5 mm in diameter was selected in order to mimic the emission from a large-area
display. The excitation conditions – the laser power and the spot size, were maintained
constant at all angles of the detection in order to eliminate excitation-related variation of the
PL spectra (Fig.4b). The light collection cone was restricted to 6o by an aperture between the
sample and collecting lens. Emission was collected in the forward direction by exciting the
NCF through the substrate and detecting after passing the PhC film and in the backward
direction – similarly excited but not traversed the PhC film.

(a) 1.0 (b) 0ΜL (c) I0/I60 60

I0/I60 relative PL intensity

normalised PL intensity

5ΜL
normalised transmision

0ML

1ML 4.5
40
0o

T (%)
2ML 0.5
0.6
I / I

4.0
5ΜL 0.4 20
0.2
60o
0.0 N
0 1 2 3 4 5
0.0 3.5 0
400 500 600 700 500 520 540 560 580 600 500 520 540 560 580 600

(nm) (nm) (nm)

Fig. 30. (a) Transmission spectra of samples with different number, N, of sphere monolayers
along the LB crystal normal. Inset – relative light attenuation at the transmission minimum
as a function of N. The short wavelength minimum is the result of the resonance splitting in
symmetric and antisymmetric combinations. This splitting decreases along the increase of
the number of monolayers. (b) Normalized PL spectra of bare NCF (thin line) and 5L LB-
coated NCF (thick line) at   0o . (c) Relative PL spectrum in comparison to far-field
transmission spectra obtained at angles   0o and 60o .

The s-polarized far-field transmission spectrum of the 1ML sample at   0o on a glass

substrate (Fig.30a) shows the well-defined minimum of 0.08 relative FWHM centred at
548 nm, which provides 12% attenuation of the incident light. The central wavelength of this
minimum increases slightly with the increasing number of layers, N . For multiple layer
coatings, the transmission significantly decreases towards shorter wavelengths due to light
scattering at package irregularities. Transmission spectra of multilayers demonstrate the
increase of the relative attenuation I / I 0 with increasing N as I / I 0 ~ N 0.9 (inset Fig.30a),
where I 0 is the extrapolated transmission in the absence of the minimum. This, practically,
linear dependence shows that the attenuation in the LB crystals is an additive function per
number of layers. Additional shallow minima were detected at, approximately, 444, 470 and
481 nm for the 1L, 2L and 5L samples, respectively (Fig.30a). The 2L and 5L samples also
possess the transmission minimum centred at 1093 nm due to the Bragg diffraction on a
stack of monolayers [87], which is not shown.
96 Nanocrystals

Fig.31a,b shows the angular dispersion of the transmission minima in a 2L LB-NCF sample.
There are I, II, III, IV branches of minima, which are degenerate at   0o . Calculated
dispersion of leaky guided modes in a 2D grating is superimposed on the transmission
pattern. The correspondence is observed only between I and (1,0) in s-polarised light and
between III and (0,1) branches in p-polarised light.

Fig. 31. (a, b) Transmission spectra of CdTe-1L-LB sample under s- and p-polarised light.
Rome numbers denote the transmission minima. Lines are calculated eigenmodes of 2D slab
PhC possessing the hexagonal lattice of scatterers and the effective refractive index of the
1ML of silica spheres. (c) The PL band of this sample.

The emission band of the NCF assembled from 2 nm NCs is centred at 540 nm (Fig.30b) and
possesses the 0.085 relative FWHM. At   0o the emission band of the 0.09 relative FWHM
of the 5L-LB NCF is centred at 533 nm. The blue shift of the emission band of the LB-NCF
sample with respect to that of the bare NCF is the result of the LB coating as was proved by
the multiplication of the PL spectrum of the bare NCF by the transmission spectrum of the
5L LB that reproduces such bandshift. The broadening of the PL band of LB-NCF sample
could be induced by the chemical modification of the NC surface during the course of the
LB film deposition. PL spectra of the LB-NCFs look similar for the different number of LB
monolayers. The relative PL spectrum I ( )  0 / I ( )  60 shows the maximum centred at
555 nm, which coincides with the minimum of the transmission spectrum (Fig.30c). The
angle-resolved PL spectrum at   60o was used as the reference (Fig.30c).

Fig. 32. (a) Angle-resolved transmission spectra of 2L LB-NCF under s-polarized light.
Numbers show the angles of light incidence. (b) Angle-resolved relative PL spectra.
Numbers show the detection angles. Arrows indicate the correspondence between
transmission minima and PL bands. (c) Angle-resolved PL spectra of the uncoated NCF.
Numbers show the detection angles. (d). Diffraction spectra at angles  shown at curves.
Inset: Schematics of the light transmission and the first order diffraction in LB coating.
Emission of semiconductor nanocrystals in photonic crystal environment 97

The overlap of the PBG with PL band takes place at 0o    20o (Fig.31). It is immediately
seen that the PBG effect upon the emission intensity is small. Over this range the maxima in
the relative PL spectra correlate the minima of the transmission spectra (Fig.32 a, b). It is
worth noting that the PBG-related modulation of the PL intensity is proportional to the
attenuation in the transmission minimum (Fig.32a,b). In the case of the uncoated NCF no
angular dependence of the PL spectrum was detected (Fig.32c)

9.2. Emission indicatrix

70
40 III 60
(a) (b) 400 (c)

diffraction intensity (arb.units)

60
400
PL intensity (arb.units)

60
30 50
300 50
50 300
I

T (%),
40 40
20 I II, III
200
40 200
30
10 30
590nm 550nm 510nm
30 100 100
20
0 10 20 30 40 50 60 0 10 20 30 40 50 60 0 10 20 30 40 50 60
 (degrees)  (degrees)  (degrees)

Fig. 33. Angle diagrams of emission intensity (solid circles) at   590, 550 and 510 nm in
comparison to transmission diagrams (open hexagons). These diagrams are compared to the
angle diagram of diffraction intensity at   550 nm (stars) in panel (b) and the Lambert
5
diagram (line) in panel (c). Shaded areas show the emission power between 0 and 5o - S550
60
(b) and 0 and 60o - S510 (c). Arrows mark the angle position of transmission minima. Rome
numbers at arrows indicate the dispersion branch as indicated in Fig.31 a,b.

The PL angle diagrams, I  ( ) , of a 2L-LB NCF obtained at wavelengths of 510, 550 and
590 nm are shown in Fig.6 together with corresponding transmission diagrams T ( ) . The
transmission diagrams show the minima superimposed on a smooth background. The
angular width of these minimum is less than 10o with respect to the mid-gap direction.
Comparison of emission and transmission diagrams reveals that the PL intensity peaks
along the direction of the transmission minimum [89]. This observation correlates with the
observation of the relative PL maxima at transmission minima (Fig.32 a,b). It is worth
noting that I  ( ) does not follow the Lambert diagram I   cos (Fig.33c, 34a). To
quantify the radiation power emitted by the NCF, we calculated the total area under the
angle diagram S60   I  ( )d , where the angle interval   60o (Fig.33c, 34a). Assuming

the even azimuth distribution of the emission intensity and taking into account that the
emission intensity becomes sufficiently low at   60o , the S60 is proportional to the power
radiated at a given wavelength. Integration of the S60 over the studied spectral range
estimates the total emitted flux.
In order to characterize the effect of PhC coating upon the emission directionality, the S60
area was normalized to that under the Lambert-like diagram, SL   I (  ,  0
o
)cos d ,

98 Nanocrystals

where the intensity I ( ,  0o ) is taken from the experimental data and   60o . Deviation
of the ratio S60 / SL spectrum of NCF-LB from that of bare NCF occurs in the PBG range: the
width of the PL indicatrix is reduced by 14% at PBG wavelengths compared to that outside
the PBG. Similar observations apply to samples with the 1L, 2L and 5L-thick crystals.
The quantity, S5 , which is proportional to the emission flux propagating within the angular
cone  from 0o to 5o at a given wavelength along the film normal was calculated to map
the emission diagram distortion in the angle-wavelength plane. S5 was then normalized to
S60 to obtain the fraction S5 / S60 . The spectrum of S5 / S60 shows that the spectral band of
enhanced emission is centred at 555 nm as a counterpart of the transmission minimum
(Fig.34b). In contrast, the S5 / S60 fraction for the bare NCF increases monotonically across
the same wavelength range.

Fig. 34. (a) The definition of the 5o- and 60o-wide flux fractions – dashed and dotted lines,
correspondingly. Dots show the PL indicatrix at   590 nm. Line is the Lambertian
indicatrix. (b) Spectra of the emission flux fraction propagating along the film normal within
a five degree wide cone for bare NCF (squares) and 2L-LB-NCF (circles) in comparison to
the transmission spectrum of the 2L-LB-NCF (line).

By increasing the principal angle of the 5o-wide probe section the enhancement band follows
the dispersion of the minimum of the far-field transmission spectra. This means that the
radiation enhancement occurs in the direction of the transmission minimum, as it follows
from comparison of Figs.35a and 32a. At higher angles, the enhancement becomes less
pronounced in response to the decreasing attenuation in the corresponding transmission
minimum. It is worth noting the resolution of other transmission minima is too poor to
generate a corresponding flux maximum.
The diffraction in the 2D grating that is conveniently used in lightning devices for extraction
of guided light [90,91]. For example, 1ML of SiO2 spheres was able to diffract the beam of
4% bandwidth from the emission trapped in a glass plate between the light source and the
monolayer of spheres [92]. The diffraction spectra were measured in order to exclude the
diffraction outcoupling as the reason for the PL intensity increase. They were obtained
keeping the sum of the incidence and diffraction angle constant     76o . The dispersion
of these resonances satisfies the expression

m0   ( sin   sin  ) (0.20)

Emission of semiconductor nanocrystals in photonic crystal environment 99

or, in particular, 0  (sin   sin(76o   )) , where the grating period is   456 nm . This is
the half of the period a  3D of the trigonal lattice for the wave vector of incident light
propagating along the K direction in the Brillouin zone of the 2D hexagonal lattice with
the distance between centres of spheres of 526 nm. Such diffraction produces resonances at
shorter wavelengths at   20o compared to the eigenmodes of the 2D PhC (Fig.32d).
Moreover, the PhC-induced band in the relative PL spectra (squares in Fig.35b) follows the
dispersion branch I of transmission minima and does not satisfy the dispersion (0.20).
Therefore, the enhanced PL band cannot be explained by the diffraction at the grating. This
conclusion is also supported by comparison of the angle diagram of the intensity of the
diffracted beam at 0  550 nm (crosses in Fig.33b) and the emission intensity diagram at the
same wavelength. Hence, we can abandon the mechanism [93,94,95] assigning the increase
of the PL intensity to diffractive outcoupling of the emission.

Fig. 35. (а) Spectra of the 5o-wide emission flux fractions propagating along different
directions. (b) The transmission pattern of a 2L LB coating under s-polarized light. Squares and
circles represent the dispersion of the enhancement band in the forward- and backward-
measured relative PL spectra, respectively. Stars show the dispersion of peaks in radiation
directionality spectra of panel (a). Dispersion of diffraction (0.20) is shown by crosses.

Another mechanism, the diffraction of the second order, assumes interaction of counter-
propagating modes with the grating period that gives rise to the out-coupled beam with
0  2neff  / m , where m  1 . This beam propagates within a very narrow radiation cone
along the grating normal [96,97]. However, this diffraction was also not observed.
More adequate explanation takes into account the PBG structure of a 2D lattice of dielectric
spheres [98,99]. This model explains reasonably well the observed far-field transmission
spectra and their angular dependence. Moreover, this theory predicts considerable
broadening of the minima bandwidth due to light leakage to the substrate [100]. The
spectral position of the first bandgap in such PhC corresponds to the sphere diameter and
the next gap takes place at a wavelength, which is lower by a factor of 1.21. The observed
minima at 548 and 444 nm for 1L LB film are in good agreement with this model. As was
experimentally demonstrated, a 2D model remains valid for the description of the PBG
structure at wavelengths   D for LB colloidal multilayers [87].
100 Nanocrystals

Dips in transmission spectra of LB films appear due to light losses for the excitation of
eigenmodes, which propagate in the LB film plane. At the PBG resonance, e.g. for   0o at
 /   1 , the local field is enhanced by one (in the case of a PhC on a substrate [99,101]) or
two orders of magnitude (for a PhC membrane [77]) due to large evanescent components of
the EM field. The strong local field on a scale of the PhC lattice constant speeds up the
radiative recombination rate as compared to that in bare NCF. The apparent independence
of the magnitude of the emission enhancement upon the number of monolayers in multiple-
layer LB films correlates with the surface-related nature of the observed effect.
With this picture in mind, the PL spectrum transformation can be explained as follows. In
the case of the bare NCF, the emission is coupled to the substrate and free space. In the
presence of an LB coating, the formation of “hot spots” of the local field at PBG wavelengths
at PhC surface leads to a higher emission rate of NCs at these spots. Since the relative PL
spectrum demonstrates a peak at PBG wavelengths, one can assume that the acceleration of
the emission rate in “hot” spots prevails over the rate decrease at “cold” spots.
The PBG origin of the enhancement leads to the emission directionality in agreement with
the angular dispersion of the PBG minimum. If NCs emit into eigenmodes of the PhC, then
the outcoupling of this radiation is the reciprocal process with respect to the coupling of
external radiation to PhC modes in a transmission experiment.
Interestingly, the emission indicatrix (Fig.33) in the studied case appears more sensitive to
the PBG directionality compared to the PL spectra (Fig.30b), but the maximum in the
directionality spectra agree the enhancement bands of the relative PL spectra (Fig.35b)
The strong argument in favour of the PBG-related enhancement mechanism is the
correlation between the relative LB-NCF PL enhancement bands measured in the backward
direction, for which the emission does not pass the LB crystal, and the minima in LB crystal
transmission spectra (Fig.36). The structure of the forward and backward relative PL
spectra and the dispersions of these bands are similar, i.e., enhancement reveals itself in
opposite directions simultaneously, although less efficient in the backward direction. This
property correlates the reciprocity of the LB-NCF sample transmission and the leakage
character of guided modes in the LB-NCF sample. These observations point to the fact that
the NCF forms an interacting unit with the LB PhC.

through LB crystal through LB crystal through LB crystal

(a) (b) (c)
mormalised relative PL intensity,

1.0 1.0 1.0

normalised trasnsmission

from
0.9 LB crystal 0.9 0.9
PL intensity,

from LB crystal
from LB crystal
0.8
0.8 0.8

0.7 T 2ΜL T 2ΜL

T 2ΜL 0.7 0.7
0o 5o 10o
500 520 540 560 580 600 500 520 540 560 580 600 500 520 540 560 580 600
 (nm)  (nm)  (nm)

Fig. 36. Normalized angle-resolved transmission (thin solid line) and relative PL spectra at 0
(a), 5 (b) and 10o (c) of a 2L LB-NCF observed in the backward direction through the glass
substrate (thick solid) and the forward direction through the LB crystal (dash-dotted lines).
Arrows indicate the centres of transmission minima.
Emission of semiconductor nanocrystals in photonic crystal environment 101

10. Summary
In this review we demonstrated different aspects of the PBG control on the emission of
semiconductor nanocrystals that is coupled to optical modes of colloidal crystal-based thin
film photonic crystals. We discussed (i) changes of the emission spectra and the emission
directionality, (ii) methods used to recognise and estimate the emission modification, (iii)
physical mechanisms behind the emission control and (iv) energy transfer between the
nanocrystals and the carcass of photonic crystals. This basic information allows to estimate
pros and contras of photonic crystal-integrated light sources and their prospects in the
design of future lightning devices.
The described experiments were limited to the emission control in the spectral range of the
lowest frequency directional bandgap, namely, the (111) bandgap in the opal crystals. The
directional nature of this bandgap reduces dramatically the strength of the PBG effect
because the light flow can escape the photonic crystal using remaining unblocked
propagation directions. Preparing inverted opals from high refractive index dielectrics can
at least partially lift up this problem due to smaller escape angle range.
Exploring the range of high order bandgaps is also in progress, but due to the nature of the
photonic bandgap diagram, the (111) gap in opals is the only true one in terms of the
absence of photon modes. All other high order bandgaps are, in fact, merely diffraction
resonances with contribution of slow propagating modes. This uncertainty blurs up the
physical mechanisms of emission modification. Obviously, in the case of inverted opals with
an omnidirectional bandgap the physical nature of the emission control can reveal itself in a
full extent, but such crystals were not yet reported for the visible for very fundamental
reasons.
We did not also address any works related to optimisation of the NC-to-PhC interaction for
following reasons. First, the PBG effect on emission of embedded sources strongly depends
on the ordering of the crystal lattice. In this sense the photonic crystals prepared by
nanolithography provide the precise control upon the light source positioning and overall
lattice regularity, but at the cost of time-consuming and expensive technologies and small
crystal volumes. Second, the design of light sources with specific functionality requires
corresponding structuring of the photonic crystal, e.g., formation of resonators. Third, the
optimisation usually exploits already established operation principles, but does not lead to
new phenomena. Instead, we concentrated the attention on photonic crystals prepared by
colloidal assembly because their combination with colloidal nanocrystals can benefit from
inexpensive technology of both components, while preserves all possibilities for emission
control offered by other realisations of photonic crystal-enhanced light sources. Overall,
along the progress in improving the colloidal crystal ordering and methods of their
structuring, this approach looks prospective for up-scaled production.

Acknowledgments
Authors gratefully acknowledge valuable contributions of their collaborators and colleagues
N. Gaponik, A. Eychmueller, A.L. Rogach, M. Bardosova, C.M. Sotomayor Torres, D.N.
Chigrin, V.G. Solovyev, R. Kian and R. Zentel.
102 Nanocrystals

11. References

1 Yin Y., A. P. Alivisatos, Colloidal nanocrystal synthesis and the organic–inorganic

interface, Nature, 437 664-670 (2005)
2 Rogach A. L., N. Gaponik, J. M. Lupton, C. Bertoni, D. E. Gallardo, S. Dunn, N. Li Pira,
M. Paderi, P. Repetto, S.G. Romanov, C. O'Dwyer, C. M. Sotomayor Torres, A.
Eychmüller, Light-Emitting Diodes with Semiconductor Nanocrystals, Angewandte
Chemie Int. Edition, 47, iss.35, 6538 – 6549 (2008)
3 Parcell E.M., Spontaneous emission probabilities at radio frequencies, Phys.Rev., 69, 681
(1946)
4 Bykov V. P., Spontaneous emission in a periodic structure, Sov. Phys. JETP 35, 269-273
(1972).
5 Bykov V. P., Spontaneous emission from a medium with a band spectrum, Sov. J. Quant.
Electron. 4, 861-871 (1975).
6 John S., Strong localization of photons in certain disordered dielectric superlattices,
Phys.Rev. Lett. 58(23), 2486-2489 (1987).
7 Yablonovitch E., Inhibited Spontaneous Emission in Solid-State Physics and Electronics,
Phys. Rev. Lett. 58, 2059-2062 (1987).
8 Joannopoulos J. D., S. G. Johnson, J. N. Winn, R. D. Meade, Photonic Crystals: Molding the
Flow of Light, 2nd Edition, Princeton University Press, Princeton and Oxford, 2008,
284p.
9 Sanders J. V. , Colour of Precious Opal, Nature, 2049, 1151-1154 (1964).
10 Darragh P. J. , A. J. Gaskin, B. C. Terrell & J. V. Sanders, Origin of Precious Opal, Nature,
209, 13-16 (1966).
11 Balakirev V.G., V.N.Bogomolov, V.V.Zhuravlev, Y.A.Kumzerov, V.P.Petranovskii,
S.G.Romanov, LA.Samoilovich, Three-dimensional superlattices in the opals,
Crystallography Reports, 38, 348-353 (1993).
12 Lopez C., Material aspects of photonic crystals, Advanced Mater., 15, N20, 1679-1704
(2003).
13 Astratov V.N., V.N.Bogomolov, A.A. Kaplyanskii, A.V. Prokofiev, L.A. Samoilovich, S.M.
Samoilovich, Yu.A. Vlasov, Optical Spectroscopy of Opal Matrices with CdS
Embedded in its Pores: Quantum Confinement and Photonic Band Gap Effects, Il
Nuovo Cimento, 17D, 1349-1354 (1995).
14 Romanov S.G., A.V.Fokin, V.V.Tretiakov, V.Y.Butko, V.I.Alperovich, N.P.Johnson,
C.M.Sotomayor Torres. Optical properties of ordered 3-dimensional arrays of
structurally confined semiconductors. J.Cryst.Growth, 159, 857-860 (1996).
15 Xia Y., B. Gates, Y. Yin, and Y. Lu, Monodispersed Colloidal Spheres: Old Materials with
New Applications, Adv. Mater., 12, No. 10 693-713 (2000).
16 Gaponenko S.V., Bogomolov V.N., Kapitonov A.M., Prokofiev A.V., Eychmueller A.,
Rogach A.L., Electrons and photons in mesoscopic structures: Quantum dots in a
photonic crystal, JETP Letters, 68, №2, 142-147 (1998).
17 Lodahl P., A. F. van Driel, I. S. Nikolaev, A. Irman, K. Overgaag, D. Vanmaekelbergh, W.
L. Vos, Controlling the dynamics of spontaneous emission from quantum dots by
photonic crystals, Nature, 430 , 654-657 (2004).
Emission of semiconductor nanocrystals in photonic crystal environment 103

18 Boedecker G., C. Henkel, Ch. Hermann, O. Hess, Spontaneous emission in photonic

structures: Theory and simulation, pp.23-42 in “Photonic crystals – Advances in
Design, Fabrication and Characterization”, eds. K.Busch, S.Loelkes, R.B.
Wehrspohn,. H.Foell, 354 p Wiley-VCH, Weinheim, 2004.
19 Sprik R., B. A. van Tiggelen, and A. Lagendijk, Optical emission in periodic dielectrics,
Europhys. Lett. 35, 265 (1996).
20 Vats N., S. John, K. Busch, Theory of fluorescence in photonic crystals, Phys.Rev. A, 65,
043808-1-13 (2002).
21 John S.and J. Wang, Quantum optics of localized light in a photonic band gap, Phys. Rev.
B 43, 12 772-12789 (1991).
22 Vats N. and S. John, Non-Markovian quantum fluctuations and superradiance near a
photonic band edge, Phys. Rev. A 58, 4168-4185 (1998).
23 Busch K. and S. John, Photonic band gap formation in certain self-organizing systems,
Phys. Rev. E 58, 3896-3908 (1998).
24 Koenderink A.F.,L.Bechger,A.Lagendijk,W.L.Vos, An experimental study of strongly
modified emission in inverse opal photonic crystals, phys.stat.sol.(a) 197 №.3,648–
661 (2003).
25 Chabanov A. A., Y. Jun, and D. J. Norris, Avoiding Cracks in Self-Assembled Photonic
Band Gap Crystals, Appl. Phys. Lett. 84, 3573 (2004).
26 Khunsin W., G. Kocher, S. G. Romanov, C. M. Sotomayor Torres, Quantitative analysis of
lattice ordering in thin film opal-based photonic crystals, Adv. Funct. Mater., 18,
2471-2479 (2008).
27 Romanov S. G., Specific features of polarization anisotropy in optical reflection and
transmission of colloidal photonic crystals, Phys. Solid State, 52, N 4, 844-854 (2010)
28 Rogach A. L., L. Katsikas, A. Kornowski, D. Su, A. Eychmueller, H. Weller, Synthesis and
Characterization of Thiol-Stabilized CdTe Nanocrystals, Ber. Bunsenges. Phys. Chem.
100, 1772-1778 (1996).
29 Romanov S. G., T. Maka, , C. M. Sotomayor Torres M. Müller, R. Zentel, D. Cassagne, J.
Manzanares-Martinez and C. Jouanin, Diffraction of Light from Thin Film PMMA
Opaline Photonic Crystals, Phys. Rev. E, 63, 056603-1-5 (2001).
30 Nikolaev I. S., P. Lodahl, and W. L. Vos, Quantitative analysis of directional spontaneous
emission spectra from light sources, in photonic crystals, Phys.Rev. A 71, 053813-1-
10 (2005).
31 Koenderink A. F. and W. L. Vos, Optical properties of real photonic crystals: anomalous
diffuse transmission, J. Opt. Soc. Am. B, 22, No. 5 1075-1084 (2005).
32 Romanov S. G., T. Maka, C. M. Sotomayor Torres, M. Müller, R. Zentel, Emission in an
SnS2 Inverted Opaline Photonic Crystal, Appl. Phys. Lett., 79, 731-733 (2001).
33 Romanov S.G., A.V. Fokin, V.I. Alperovich, N.P. Johnson, R.M. De La Rue, The Effect of
the Photonic Stop-Band upon the Photoluminescence of CdS in Opal, Physica Status
Solidi a, 164, 169-173 (1997).
34 Bogomolov V. N., S. V. Gaponenko, I. N. Germanenko, A. M. Kapitonov, E. P. Petrov, N.
V. Gaponenko, A. V. Prokofiev, A. N. Ponyavina, N. I. Silvanovich, S. M.
Samoilovich, Photonic band gap phenomenon and optical properties of artificial
opals, Phys.Rev. E, 55, 7619-7625 (1997).
104 Nanocrystals

35 Romanov S.G., M. A. Kaliteevski, C.M.Sotomayor Torres, J. Manzanares Martinez, D.

Cassagne, J.P. Albert, A.V. Kavokin, V.V. Nikolaev, S.Brand, R.A. Abram, N.
Gaponik, A. Eychmueller, A.L. Rogach, Stimulated emission due to light
localization in the bandgap of disordered opals, Phisica Status Solidi (c), 1, 1522-1530
(2004).
36 Golub J. E., R. Eichmann, and P. Thomas, Dimensional crossover of exciton hopping
relaxation in mesoscopic structures, Europhys. Lett. 47(5), 628-631 (1999).
37 A.A. Kiselev, Hopping-induced energy relaxation with allowance for all possible versions
of intercenter transitions, Semiconductors 32(5), 504-508 (1998).
38 Baranovskii S. D., T. Faber, F. Hensel, and P. Thomas, On the Description of Hopping -
Energy Relaxation and - Transport in Disordered Systems, J. Non-Cryst. Solids 198-
200, 222-225 (1996).
39 Zegria G. G. and V. A. Kharchenko, New mechanism of Auger recombination of
nonequilibrium current carriers in semiconductor heterostructures, Sov.Phys.JETP,
74(1), 173-181 (1992).
40 Yeh P., Optical Waves in Layered Media, John Wiley&Sons, New York, 1998; ISBN 0-471-
82866-1.
41 J Russell P.St., Optics of Floquet-Bloch waves in dielectric gratings, Appl. Phys. B, 39, 231-
246 (1986).
42 Kosaka H., T. Kawashima, A. Tomita, M. Notomi, T. Tamamura, T. Sato, S. Kawakami,
Superprism phenomena in photonic crystals, Phys. Rev. B, 58 R10096-99 (1998).
43 Romanov S.G., D. N. Chigrin, T. Maka, M. Müller, R. Zentel, C. M. Sotomayor
Torres, Directionality of light emission in three-dimensional opal-based photonic
crystals, Proc. SPIE Int. Soc. Opt. Eng. 5825, 160-172 (2005).
44 Romanov S. G., D.N. Chigrin, C. M. Sotomayor Torres, N. Gaponik, A. Eychmüller, A. L.
Rogach, Emission Stimulation in a directional bandgap of a CdTe-loaded opal
photonic crystal, Phys.Rev.E. 69, 046606-1-4 (2004).
45 Johnson S.G., J.D. Joannopoulos, Block-iterative frequency-domain methods for
Maxwell’s equations in a plane-wave basis,Optics Express, 8 (2001) 173-190
46 Wiersma D.S. and A. Lagendijk, Light diffusion with gain and random lasers.Phys. Rev.E.,
54(4), 4256-4265 (1996).
47 John S. and G. Pang, Theory of lasing in a multiple-scattering medium, Phys. Rev.A., 54(4),
3642-3652 (1996).
48 Vlasov Yu. A., M.A. Kaliteevski, and V.V. Nikolaev, Different regimes of light
localization in a disordered photonic crystal, Phys. Rev. B , 60(3), 1555-1562 (1999)
49 Kaliteevski M. A., J. Manzanares Martinez, D. Cassagne, and J. P. Albert, Disorder-
induced modification of the transmission of light in a two-dimensional photonic
crystal, Phys.Rev.B., 66, 113101-1-4 (2002).
50 Soest G. van, F. J. Poelwijk, R. Spirk, and A. Lagendijk, Dynamics of a random laser above
threshold, Phys. Rev. Lett. 86, 1522-1525 (2001).
51 D.N. Chigrin, Radiation pattern of a classical dipole in a photonic crystal: Photon
focusing, Phys. Rev. E 70, 056611-1-12 (2004).
52 Etchegoin P. and R. T. Phillips, Photon focusing, internal diffraction, and surface states in
periodic dielectric structures, Phys. Rev. B 53, 12674-12683 (1996).
Emission of semiconductor nanocrystals in photonic crystal environment 105

53 Romanov S. G., D. N. Chigrin, V. G. Solovyev, T. Maka, N. Gaponik, A. Eychmüller, A. L.

Rogach, and C. M. Sotomayor Torres, Light emission in a directional photonic
bandgap, Phys.stat.sol. (a), 197, 662-672 (2003)
54 Ho K. M., C. T. Chan, and C. M. Soukoulis, Existence of a photonic gap in periodic
dielectric structures, Phys. Rev. Lett. 65, 3152-3155 (1990)
55 S. G. Romanov, U. Peschel, M. Bardosova, S. Essig, K. Busch, Suppression of the critical
angle of diffraction in thin film colloidal photonic crystals, Phys. Rev. B, 82 (2010) in
press.
56 Reynolds A., F. López-Tejeira, D. Cassagne, F. J. García-Vidal, C. Jouanin, and J. Sánchez-
Dehesa, Spectral properties of opal-based photonic crystals having a SiO2
matrixPhys. Rev. B., 60, 11422-11426 (1999).
57 Zhukov E. A., Y. Masumoto, H. M. Yates, M. E. Pemble, E.A. Muljarov, C. M. Sotomayor
Torres and S. G. Romanov, Interface Interactions and the Photoluminescence from
Asbestos-Templated InP Quantum Wires, Microstructures and Superlattices, 27, No.
5/6, pp. 571-576 (2000).
58 Bardosova M., S.G. Romanov, C.M. Sotomayor Torres, N. Gaponik, A. Eychmueller, Y.A.
Kumzerov, J. Bendall, Effect of template defects in radiative energy relaxation of
CdTe nanocrystals in nanotubes of chrysotile asbestos, Micro- and Meso-porous
materials, 107, 212-218 (2007).
59 Mikhailovsky A.A., A.V. Malko, J.A. Hollingsworth, M.G. Bawendi, V.I. Klimov,
Multiparticle interactions and stimulated emission in chemically synthesized
quantum dots, Appl.Phys.Lett., 80, 2380 (2002).
60 Solovyev V. G., S. G. Romanov, C. M. Sotomayor Torres, M. Müller, R. Zentel, N.
Gaponik, A. Eychmüller and A. L. Rogach. Modification of the spontaneous
emission of CdTe nanocrystals in TiO2 inverted opals, J. Appl.Phys., 94, 1205-1210
(2003).
61 Astratov V.N., A.M. Adawi, S. Fricker, M.S. Skolnick, D.M. Whittaker, P.N. Pusey,
Interplay of order and disorder in the optical properties of opal photonic crystals,
Phys. Rev. B, 66, 165215-1-13 (2002)
62 Stefanou N., V. Yannopapas, A. Modino, Heterostructures of photonic crystals: frequency
bands and transmission coefficients, Computer Physics Commun. 113, 49-77 (1998)
63 Romanov S. G., H. M. Yates, M. E. Pemble, R. M De La Rue, Opal-Based Photonic Crystal
with Double Photonic Bandgap Structure, J. Phys.: Cond. Matter 12, 8221-8229 (2000).
64 Istrate E., E. H. Sargent, Photonic crystal heterostructures and interfaces, Rev. Modern
Physics, 78, 455-481 (2006).
65 Gaponik N., A. Eychmüller, A.L. Rogach, V.G. Solovyev, C.M. Sotomayor Torres,
S. G. Romanov, Structure-related optical properties of luminescent hetero-opals, J.
App. Phys., 95, 1029-1035 (2004).
66 Romanov S. G., N. Gaponik, A. Eychmüller, A.L. Rogach V. G. Solovyev, D.N. Chigrin
and C. M. Sotomayor Torres, Light emitting opal-based photonic crystal
heterojunctions, in “Photonic Crystals” eds. K. Busch, S.Lölkes, R. Wehrspohn, H.
Föll, Wiley-VCH, Weinheim, ISBN-13: 978-3-527-40432-2, 132-152 (2004).
67 Romanov S.G., Optical characterization of opal photonic hetero-crystals, in “Frontiers of
Multifunctional Integrated Nanosystems” eds. E. Buzaneva and P. Scharff, Kluwer
Acad. Publ., ISBN978-1-4020-2171-8, 309-330 (2004).
106 Nanocrystals

68 Jiang P., G. N. Ostojic, R.Narat, D. M. Mittleman, V.L. Colvin, The Fabrication and
Bandgap Engineering of Photonic Multilayers, Adv. Mater. 13, No. 6, 389-393 (2001)
69 Decher G., Fuzzy Nanoassemblies: Toward Layered Polymeric Multicomposites. Science,
277 1232-1237 (1997).
70 Rogach A. L., D. S. Koktysh, M. Harrison, N. A. Kotov, Layer-by-Layer Assembled Films of
HgTe nanocrystals with Strong Infrared Emission, Chem. Mater., 12, 1526-1528 (2000).
71 Caruso F., Nanoengineering of Particle Surfaces. Adv. Mater., 13, 11-22 (2001).
72 Gaponik N., D. V. Talapin, A. L. Rogach, K. Hoppe, E. V. Shevchenko, A. Kornowski, A.
Eychmüller, and H. Weller, Thiol-Capping of CdTe Nanocrystals: An Alternative
to Organometallic Synthetic Routes, J. Phys. Chem. B 106, 7177-7185 (2002).
73 Ding B., M. Bardosova, I. Povey, M. E. Pemble, and S.G. Romanov, Engineered Light
Scattering in Colloidal Photonic Heterocrystals, Adv. Funct. Mater., 20, 853-860 (2010).
74 Ohtaka K., H. Miyazaki, T. Ueta, Near-field effect involving photonic bands, Mat. Science
and Engineering B, 48 153- 161 (I 997).
75 Fluck E., N. F. van Hulst, W. L. Vos, and L. Kuipers, Near-field optical investigation of
three-dimensional photonic crystals, Phys.Rev. E 68, 015601-1-4 (2003).
76 Koenderink A. F., M. Kafesaki, C. M. Soukoulis, V. Sandoghdar, Spontaneous emission
rates of dipoles in photonic crystal membranes, J.Opt.Soc.Am.B, 23, 1196-1206 (2006)
77 Inoue M., Enhancement of local field by a two-dimensional array of dielectric spheres
placed on a substrate, Phys.Rev.B, 36, 2852-2862 (1987).
78 Kurokawa Y., H. Miyazaki, Y. Jimba, Optical band structure and near-field intensity of a
periodically arrayed monolayer of dielectric spheres on dielectric substrate of finite
thickness, Phys.Rev B 69, 155117-1-9 (2004).
79 Rigneault H., F.Lemarchand, A.Sentenac, H.Giovannini, Extraction of light from sources
located inside waveguide grating structures, Opt.Lett., 24, 148-150 (1999).
80 Joyce W.B., R.Z. Bachrach, R.W.Dixon, D.A.Sealer, Geometrical properties of random
particles and the extraction of photons from electroluminescent diodes, J.Appl.Phys.,
45, 2229-2254 (1974).
81 Bermel P., C. Luo, L.Zeng, L. C. Kimerling, J. D. Joannopoulos, Improving thin-film
crystalline silicon solar cell efficiencies with photonic crystals, Optics Express, 15,
№25, 16986-17000 (2007)
82 Boroditsky M., T. F. Krauss, R. Coccioli, R. Vrijen, R. Bhat, E. Yablonovitch, Light
extraction from optically pumped light-emitting diode by thin-slab photonic
crystals, Appl. Phys.Lett., 75, 1036-1039 (1999).
83 Romanov S.G., M. Bardosova, M. Pemble, C.M. Sotomayor Torres, N. Gaponik A.
Eychmüller, Emission pattern of planar CdTe nanocrystal light source coated by
two-dimensional Langmuir-Blodgett photonic crystal, Materials Science and
Engineering C, 27, 968-971 (2007) .
84 Gaponik N., A.L. Rogach, Aqueous synthesis of semiconductor nanocrystals in
“Semiconductor Nanocrystal Quantum Dots. Synthesis, Assembly, Spectroscopy
and Applications”, Rogach, A. (Ed.), Springer, Wien, N ew York, 2008, 372 p., ISBN:
978-3-211-75235-7
85 Shavel V., N. Gaponik, A. Eychmüller. The Assembling of Semiconductor Nanocrystals,
Eur. J. Inorg. Chem., 18 3613-3263, (2005).
Emission of semiconductor nanocrystals in photonic crystal environment 107

86 Bardosova M., P. Hodge, V.Smatko R.H. Tredgold and D. Whitehead, A new method of
forming synthetic opals, Acta Phys. Slovaca, 54, 409-415 (2004).
87 Romanov S.G., M. Bardosova, M. Pemble, C.M. Sotomayor Torres, (2+1)-dimensional
photonic crystals from Langmuir-Blodgett colloidal multilayers, Applied Physics
Letters, 89, 43105 (2006).
88 Romanov S.G., M. Bardosova, D.E. Whitehead, I. Povey, M. Pemble, C.M. Sotomayor
Torres, Erasing diffraction orders – opal versus Langmuir-Blodgett colloidal
crystals, Applied Phys. Lett.,90,133101 (2007).
89 Romanov S.G., M. Bardosova, I.M. Povey, M. Pemble, C.M. Sotomayor Torres, N.
Gaponik and A. Eychmüller, Modification of emission of CdTe nanocrystals by the
local field of Langmuir-Blodgett colloidal photonic crystals , J. Appl. Phys. 104,
103118 1-8 (2008).
90 Alferov Zh.I., V.M. Andreev, S.A.Gurevich, R.F.Kazarinov, V.R.Larionov, M.N. Mizerov,
E.L.Portnoi, Semiconductor lasers with the light output through the diffraction grating
on the surface of the waveguide layer, IEEE J. Quant. Electronics, 11, 449-451 (1975).
91 Turnbull G. A., P. Andrew, M. J. Jory, W. L. Barnes, and I. D. W. Samuel, Relationship
between photonic band structure and emission characteristics of a polymer
distributed feedback laser, Phys.Rev. B, 64, 125122-1-6 (2001).
92 Yamasaki T., K. Sumioka, and T. Tsutsui, Appl.Phys.Lett, Organic light-emitting device with an
ordered monolayer of silica microspheres as a scattering medium, 76, 1243-1245 (2000).
93 Dakks M.L., L.Khun, P.F.Heidrich, B.A. Scott, Grating Coupler For Efficient Excitation Of
Optical Guided Waves In Thin Films, Appl.Phys.Lett., 16, 523-525 (1970).
94 Matterson B. J., J. M. Lupton, A.F. Safonov, M. G. Salt, Wi. L. Barnes, I. D. W. Samuel,
Increased Efficiency and Controlled Light Output from a Microstructured Light-
Emitting Diode, Adv. Mater. 13, 123-127 (2001).
95 Ziebarth J. M., A. K. Saafir, S. Fan, M. D. McGehee, Extracting Light from Polymer Light-
Emitting Diodes Using Stamped Bragg Gratings, Adv. Funct. Materials, 14, 451 (2004).
96 Streifer W., D.R. Scifres, R.D. Burnham, Coupled wave analysis of DFB and DBR lasers
IEEE J. Quant. Electronics, 13, 134-141 (1977).
97 Kanskar M., P.Paddon, V.Pacradouni, R. Morin, A. Busch, J.F,Young, S.R. Johnson,
J.MacKenzie, T. Tiedje, Observation of leaky slab modes in an air-bridged
semiconductor waveguide with a two-dimensional photonic lattice, Appl.Phys.Lett.,
70, 1438-1440 (1997).
98 Miyazaki H.T., H. Miyazaki, K.Ohtaka, T.Sato, Photonic band in two-dimensional lattices
of micrometer-sized spheres mechanically arranged under a scanning electron
microscope, J. Appl. Phys., 87, 7152-7159 (2000).
99 Miyazaki H.T., H. Miyazaki, K.Ohtaka, T.Sato, Photonic band in two-dimensional lattices
of micrometer-sized spheres mechanically arranged under a scanning electron
microscope, J. Appl.Phys., 87, 7152-7159 (2000).
100 Kurokawa Y., H.Miyazaki, Y.Jimba, Light scattering from a monolayer of periodically
arrayed dielectric spheres on dielectric substrates, Phys.Rev.B, 65, 201102-1-4 (2002).
101 Kurokawa Y., H. Miyazaki, Y. Jimba, Optical band structure and near-field intensity of a
periodically arrayed monolayer of dielectric spheres on dielectric substrate of finite
thickness, Phys.Rev B 69, 155117-1-9 (2004).
108 Nanocrystals
A review of high nanoparticles concentration
composites: semiconductor and high refractive index materials 109

X5

A review of high nanoparticles concentration

composites: semiconductor and
high refractive index materials
Igor Yu. Denisyuk
Doctor of Science, professor, head of the chair of Quantum Sized Optical Systems of State
University of Information Technologies, Mechanics and Optics
Russia, 197101, Saint-Petersburg, Str. Kronverskii 49

Mari Iv. Fokina

Assistant professor of State University of Information Technologies, Mechanics and Optics
Russia, 197101, Saint-Petersburg, Str. Kronverskii 49

1. Introduction
At the present time homogeneous optical and optoelectronic media such as glasses,
monocrystals and polymer materials are widely used. Each of these media has a specific set
of properties. For example, the polymer materials allow producing flexible and transparent
films, for example film OLED. They are cheap. The technology of polymers treatment is
very simple and suitable for some applications. However the holes and electrons mobility
in polymer materials are many time less in comparison with inorganic well known
semiconductors. Another problem is a relative high exciton decay energy in polymer with
value of 100 meV that result on temperature dependence of photogeneration.
The possibility of combining the different properties into a single material should be rather
useful. It is impossible to solve this problem by traditional ways because the properties
reflect the internal structures of these different materials.
The method of nanostructuring provides the possibility of combining the properties of
polymers and crystals. The resulting nanocomposite is the mechanical mixture of inorganic
semiconductor distributed uniformly in the polymer matrix. Under the condition of uniform
distribution of nanoparticles and if the size of such nanocrystals is small (2-5 nm), they don’t
distort an incident light wave and the light scattering is low. If to use high refractive index
nanoparticles such as ZnS, CdS, ZnO, TiO2 incorporation of nanoparticles into polymer will
result on significant increasing of refractive index of material. Same time these material has
a proprieties of homogenous semiconductor material because of small nanometer size
distance between semiconductor nanoparticles and easy tunnel transportation of charge
carrier. Nanocomposite with high concentration of small size nanocrystals becomes
effectively a homogeneous medium, having semiconductor proprieties of inorganic material
with low scattering and good flexibility and processability of polymer. The set of properties
of this mixture is determined by both components, namely polymer and nanocrystals, and
110 Nanocrystals

by the ratio of concentrations of them. The main efforts of research directed to develop
quasi- homogenous nanocomposite material with nanoparticles and polymer matrix
comparable content for photonics application areas: photoresist for nanolithography,
microoptics, organic solar cell and OLED.

2. Crystalline lattice of small nanoparticles

Physics of semiconductor theory use macroscopic charge carrier statistic parameters of bulk
crystals for all structures types including microns and sub- microns elements in microchips.
It is correct approximation now as the sizes of microchips elements more than 100 nm are
more time larger in comparison to interatomic distances. In contrary, typical nanoparticles,
for example CdS or ZnS with the size of 1 - 3 nm include a few atomic layers and its
interatomic distance can differ essentially from same of bulk crystals. Now are a few of
works where these effects are investigated.
Indeed, if to shrink elements down to the nanometer scale, creating nanodots, nanoparticles,
nanorods and nanotubes a few tens of atoms across, they've found weird and puzzling
behaviors unexpected for bulk and micron sized material.

Fig. 1. Mean Cd-S distances RCd-S as a function of the size of CdS nanocrystals deduced
from EXAFS experiments. Bulk values of the cubic and the hexagonal phase of CdS are also
indicated (dotted).

Certainly at decreasing of nanoparticles sized up to few nanometers that correspond to

some tens atoms its crystalline lattice change essentially from same of bulk crystal and it is
appear instability of crystalline lattice. Large surface of small particle will result on
augmentation of influence of surface states on crystalline lattice of nanoparticles.
There are a lot of work that show size and surface effects of crystalline lattice change in
nanometers size nanoparticles.
A review of high nanoparticles concentration
composites: semiconductor and high refractive index materials 111

In the work [1] it is shown nanoparticles crystalline lattice parameter dependence from
organic substance settled on its surface and its size. Micron size particles, for example CdS
have crystalline lattice similar to bulk crystals at decreasing of nanoparticles sizes down to
nanometers range its lattice change essentially.

In Fig. 1 show Cd-S distance dependence from nanoparticles diameter and organic shell
nature.
A few CdS samples have been investigated, the particles differing in size and crystal
structure:
- three samples of polyphosphate-stabilized nanoparticles with cubic crystal structure
(diameters approximately 3.0-12.0 nm),
- four samples of thioglycerol-stabilized particles (hexagonal and cubic (see below),
diameters from 1.4 to 4.0 nm), and
- three samples of crystallized nanoparticles, the structures and superstructures of which are
known from SC-XRD (diameters from 1.3 to 1.7 nm).
Microcrystalline CdS of hexagonal crystal structure was used as a reference substance. The
smallest “particles” investigated consist of a three-dimensional network of Cd8(SR)16 units
(R) thioglycerol), which may serve as a model for the surface of thiol-stabilized CdS
nanocrystals. The other two crystallized clusters (Cd17S4(SR)26 and Cd32S14(SR)36, R)
mercaptoethanol and 1-mercapto-2-propanol, respectively) may be regarded as fragments of
the cubic (zinc blende) phase of CdS and appear tetrahedrally. Thus, at least for the latter
three samples, “diameter” is to be taken only as a reference point for the “size“ of the
particles. EXAFS spectra have been taken in transmission mode at the Cd K-edge in the
energy range from 26.4 to 29.0 kV at temperatures between 5 and 296 K.
All of the samples clearly showed the Cd-S coordination shell, whereas the Cd-Cd
coordination (second shell) was visible in particles larger than 3.0 nm only. Figure 1 shows
the dependence of the Cd-S bond length as a function of the particle diameter deduced from
the EXAFS analysis. The quality of the data allows us to divide the plot into three regions.
Samples 8-10 exhibit a slight contraction of the Cd-S bond with decreasing particle size,
which is due to the minimization of the surface energy, obviously unhindered by the
ionically bound stabilizers.
In contrast, the covalently bound stabilizers of samples 1-5 expand the Cd-S bond. This
expansion is larger for the smaller particles, and it becomes larger as the steric interaction
among the stabilizers comes into effect. For samples 1-3, the bond lengths determined by
EXAFS match very well those from the SC-XRD analysis. From P-XRD, sample 7 is assigned
to the hexagonal crystal phase, by means of which the larger mean bond length, compared
to samples 8 and 9 (similar size but cubic phase), is explainable. Like sample 7, sample 6 is
prepared at elevated temperatures, which makes it likely that this sample also belongs to the
thermodynamically stable hexagonal crystal phase (P-XRD does not allow an unambiguous
assignment). This guess is corroborated by the “out of order” bond length (Fig. 1) and by the
analysis of the third moments of the pair distribution function (Fig. 1). According to the
surface-to-volume ratio of the nanocrystals, this quantity increases with decreasing particle
size but is divided into two groups: on one hand, all particles clearly belonging to the cubic
structure (four equivalent Cd-S bonds lead to C3) for the largest particles) and on the other
hand, samples 6 and 7, with distinctly elevated anharmonicity (in hexagonal CdS, three Cd-
S bonds are equivalent, and one differs from those). Thus, for sample 6, we are in a position
112 Nanocrystals

to state the assingnment to the hexagonal phase by means of EXAFS spectroscopy, which
has not previously been possible, by applying P-XRD and HRTEM. The Debye temperatures
and static disorders are extractable from the EXAFS data together with the bond lengths and
anharmonicities, as mentioned above. The Debye temperatures that were determined
increased slightly with decreasing particle size, which points to a stiffening of the Cd-S
bonds. When compared to the bulk value, all of the nanoparticles displayed an elevated
static disorder, which, in the first instance, increases with decreasing size according to the
surface-to-volume ratio (samples 10-6). For the very small clusters of samples 1-5, again
slightly reduced static disorders are observed (in good agreement with the SC-XRD of
samples 1-3). Possibly, this finding is a hint toward different regimes of particle growth:
thermodynamically controlled growth leads to a crystallizable species, whereas subsequent
Ostwald ripening yields larger and less specific colloids.
At decreasing of nanoparticles sizes up to 1 nanometers distances in crystalline lattice
between metal atoms will increase that result of moving from crystalline to amorphous state
of material in result. Transformation from crystalline to amorphous form of small particles
well known for Fe2O3, Se, inorganic materials.

3. Methods of preparation and stabilization of semiconductor nanoparticles

3.1 Inorganic nanoparticles based nanocomposit
From 90 years was develop the main methods of nanoparticles synthesis. Few nanoparticles
types now is commercial available from Aldrich and other commercial supplier. Methods of
the nanoparticles synthesis can be divided into three main groups.
1. Synthesis of semiconductor nanoparticles in solutions of the corresponding salts by
controlled addition of anions (or cations) or by hydrolysis [2];
2. Preparation of nanoparticles as a result of phase transformations [3];
3. The synthesis of nanoparticles in aerosols [4].
Preparation of nanocomposite material having both high nanoparticles concentration,
absence of its coagulation and homogeneous optical and semiconductor proprieties is a
mostly difficult problem. Usually it accomplished by preparation of suitable nanoparticles
with modified surface and then to its incorporation in polymer, having surface active
proprieties often.
For consideration of nanoparticles state in nanocomposite, we should assume that
nanoparticles can interact with polymer matrix with formation of ordered polymer layers in
its surface. In the work [5] was made numerical study of state of nanoparticles (fullerene) in
polymer matrix and received interesting results on this bi- phase system elastic proprieties.
The purpose of this work is to investigate the effect of nanoparticle size on elastic properties
of polymeric nanocomposites using MD simulations. For this, molecular models of a
nanocomposite were constructed by reinforcing amorphous polyethylene (PE) matrix with
nano sized buckminister fullerene bucky-ball (or simply bucky-ball). Bucky-balls of three
different diameters (0.7, 1.2 and 1.7 nm, respectively) were utilized to incorporate size effect
in the nanocomposites. To represent them as a generic nanoparticle system, all bucky-balls
were configured as rigid body. This is necessary because a bucky-ball embedded inside the
polymer matrix may deform excessively depending on its size and may overshadow the
composite mechanical properties attributed to filler size. The assumption of rigid bucky-ball
will ensure that the shape of filler does not contribute to variation in elastic properties. The
A review of high nanoparticles concentration
composites: semiconductor and high refractive index materials 113

assumption may be unrealistic for large diameter buckyballs, it is a reasonable assumption

for small bucky-balls and solid nanoparticles. In addition to this shape constraint, the
volume fraction of the filler, matrix characteristics (density, molecular weight, molecular
weight distribution, branch content, degree of crystallinity, etc.) and their force interaction
with the nanoparticle were kept constant in all nanocomposites. Molecular models of the
neat PE matrix were also developed for comparison. Elastic properties of the neat and
nanocomposite systems were then evaluated using four different modes of deformation,
namely, unidirectional tension and compression, and hydrostatic tension and compression,
respectively.
Molecular models of nanocomposites were developed by symmetrically placing a spherical
fullerene bucky-ball in the PE matrix, as shown schematically in Fig. 2. The dashed box in
Figs. 2 a and b indicates the periodic cell or unit cell that was simulated by MD. Three types
of bucky-balls, C60, C180 and C320 (subscripts denote number of carbon atoms), were used
to incorporate the size effect. All bucky-balls were infused in matrix by approximately 4.5
vol%. Periodic boundary conditions were employed to replicate the unit cells in three
dimensions. In nanocomposites, the PE matrix was represented by united atom (UA) –CH2–
units. The initial structure of the matrix was constructed by positioning the bucky-ball at the
center of the unit cell and by randomly generating PE chain(s) on a tetrahedron lattice
surrounding the bucky-ball.

Fig. 2. (a) Schematic diagram of polymer nanocomposites, (b) periodic cells used for MD
simulations.

Numerical simulation shown that elastic properties of nanocomposites are improved

appreciably with the infusion of bucky-balls in PE matrix. The trend shows that with the
increase in filler size, the extent of enhancement in elastic properties is gradually reduced.
The result is somewhat surprising because in all cases the volume fraction was maintained
constant (4.5%).
It can be concluded from this observation that size of the filler has considerable influence on
polymer density even with non-bonded inter-molecular interactions between polymer and
nanoparticle. The effect can be well understood from the radial density distribution of PE for
both neat and nanocomposites, as shown in Fig. 3 in which the distribution is constructed by
measuring local densities of PE at various radial distances starting from the center to the
114 Nanocrystals

half-length of the periodic box. It is interesting to find that local densities are not constant
along the radial distance. A 200–250% increase in polymer density exists for all
nanocomposites at a distance close to the nanoparticle. At further distances, the distribution
fluctuates in a similar manner as in the neat polymer system. The fluctuating character is
inherent because mass needs to be conserved [6]. The collective contributions of these
factors yielded a decreasing trend in polymer bulk density with the increment of filler size.
It appears from the analysis that polymer density distribution plays the foremost role in size
effect. However, it is not elucidated why size difference influence polymer density. The
discernible contribution from filler size can be realized from radial distribution plot as
shown in Fig. 4.

Fig. 3. Radial density distribution of various: (a) neat PE and (b) nanocomposite models.
Space occupied by nanoparticles is schematically shown by the quarter circles.
A review of high nanoparticles concentration
composites: semiconductor and high refractive index materials 115

Fig. 4. PE-Bucky radial distribution functions (RDF) of various nanocomposite models.

It is known that the radial distribution function for any atom pairs gives a measure on how
corresponding atoms are distributed in three-dimensional space due to VDW interactions.
Hence, g(r)PE-Bucky refers to radial distribution of PE atoms with respect to Bucky-ball
atoms. As atomic position of all bucky-balls were fixed, a plot of g(r)PE-Bucky would thus
provide information about the polymer distribution due to interaction with a nanoparticle.
Fig. 4 reveals that the size of Bucky-ball has strong influence on the g(r) plot. It is observed
that the value of g(r) assumed zero from 0 to 3.4 A  for all nanocomposites, then increases
with radial distance. The zero value refers to the VDW thickness h. It is also evident that h
does not depend on filler size. It is quite expected because parameters describing LJ
potentials are identical for all nanocomposites and the nature of the h is known to be
governed by such interactions between nanoparticle and polymer [7]. However, the relative
distribution of polymer atoms towards the nanoparticle, as indicated by the variation in g(r)
at a particular radial distance, is quite different with the change in filler size. It is obvious
from Fig. 4 that more atoms are tending to disseminate across the polymer–nanoparticle
interface as the size of buckyball decreases.
It appears from the above discussion that with the reduction in filler size, the bulk density of
polymer and the attractive interaction energy between polymer and nanoparticle at the
interface increase substantially. Enhancements of these parameters are then translated to
improved elastic moduli.

Preparation of nanocomposite with high nanoparticles content.

According to above numerical simulation and discussion there are strong interaction
between polymer and nanoparticles. At rapprochement of nanoparticles up to distance of 3
– 5 nm nanocomposite behavior change dramatically. In real nanocomposite with
nanoparticles size of 2-3 nm these condition of nanoparticles – nanoparticle distance about 3
nm begin with about 10 vol. % nanoparticles concentration. So, 10 % is a border between
116 Nanocrystals

usual nanocomposite with no nanoparticles interaction and high nanoparticles

concentration composite where nanoparticle – nanoparticle interaction play the main role in
behavior formation of whole composition.

At preparation of nanocomposite with high nanoparticles concentration a contradiction is

appear: to avoid nanoparticles coagulation we need to increase interaction between them
and the polymer matrix, but at the same time those interactions with polymer will result in
hardening of the composites in result of polymer cross-linking over nanoparticles.
In most part research this contradiction has been avoided by use of very fast drying of
material [8]. At fast drying nanoparticles have not enough time for coagulation and solid
material keep good distribution take place in solution. In that works, the dangerous stage of
particles interactions passed fast and the composites kept their transparency. For example, if
a layer is prepared by spin coating, the process of solid coating preparation occupies only
some part of a second. Certainly, this method is not suitable for preparing thick
nanocomposite layers and bulk nanocomposites.
Next two works show possibility to obtain high nanoparticles concentration in material that
suitable to obtain high refractive index homogenous nanocomposite material based on high
concentration of ZnS or CdS semiconductor nanocrystals in polymeric matrix. Same
homogenous nanocomposite with high semiconductor nanoparticles concentration around
20 - 30 vol % are suitable as a homogenous semiconductor materials, so synthesis of these
materials were described in detail here.
One example of high concentration ZnS nanocomposit thin film preparation give the work [8].
In the work was used previously prepared ZnS-monomers prepolymer that was UV-cured
at once after spin coating in thin film. UV curing technology was used to rapidly
prepolymerize the ZnS–macromer system, and then a radical polymerization process was
carried out to complete the polymerization reaction. Another advantage of using this
macromer is that the macromer in solution has a viscosity, which is favorable for spin-
coating to form films. Furthermore, a polymerizable moiety as capping agent has also been
utilized to modify the surface of ZnS nanoparticles in order to immobilize the ZnS particles
into the polymer. Research on the preparation of nanoparticles–polymer composites using
polymerizable surfactants, ligands or capping agents has previously been reported [9, 10].
This approach can effectively avoid the phase separation and results in transparent
composites because the functionalized inorganic particles with polymerizable vinyl groups
can be copolymerized with the monomers to form integrated polymeric materials.
In this work firstly was being synthesized the thiophenol (PhSH)–4-thiomethyl styrene
(TMSt)-capped ZnS nanoparticles with high concentration in DMF. Then a UV curable
urethane-methacrylate macromer (UMM) was introduced into the nano-ZnS containing
DMF solution. The ZnS nanoparticles were immobilized into the polymer matrix via
copolymerization of the macromer (UMM) with 4-thiomethyl styrene (TMSt) bound on the
surface of ZnS particles to synthesize a series of nano-ZnS–poly(urethane-methacrylate
macromer) (PUMM) transparent composite films with high refractive indices. The structure
and composition of the thiolcapped ZnS nanoparticles were characterized via TEM, X-ray,
FTIR and chemical analyses. The thermal properties, optical properties and microstructure
of the nanocomposite films were investigated in detail.
Detailed description of high concentration nanocomposite synthesis method according to
work [8].
A review of high nanoparticles concentration
composites: semiconductor and high refractive index materials 117

3.2 Materials
Anhydrous zinc acetate, N,N-dimethylformamide (DMF), thiophenol (PhSH), thiourea, 2,4-
tolylene diisocyanate (TDI), dibutyltin dilaurate (DBTL), 2-hydroxyethyl methacrylate
(HEMA) and other chemical reagents were of analytical grade and were used without
further purification. 4-Vinylbenzyl chloride (w 95% GC grade, Fluka) and 2,2-dimethyl-2-
hydroxyacetophenone (Darocur 1173 from Ciba Special Chemicals) were used as received.
2,2’-Dimercaptoethyl sulfide (MES) was synthesized as reported previously [11]. Synthesis
of 4-thiomethyl styrene (TMSt) TMSt was synthesized from 4-vinylbenzyl chloride in a
manner similar to that reported in the literature [12]. 15.3 g of 4-vinylbenzyl chloride (0.1
mol), 9.12 g of thiourea (0.12 mol), 200 ml of ethanol and 0.08 g of p-methoxyphenol as
inhibitor were put into a four-necked flask fitted with a reflux condenser. The reaction
mixture was stirred at reflux temperature for 4 h under N2 flow and then was cooled. 75 ml
of 20% solution of sodium hydroxide were added to the above mixture, and the resulting
solution was immediately heated to 80 C and continuously stirred at 80 C for 0.5 h. Finally,
the resulting solution was cooled to room temperature and 100 ml of CHCl3 were added.
The organic phase was separated out and washed with distilled water until neutral. Then
the organic phase was dried over anhydrous Na2SO4 and the product (TMSt) was obtained
by removing the organic solvent under reduced pressure. Yield 75%, n20 d ~ 1.625.
The method for preparing thiol-capped ZnS nanoparticles (TCZnS) was similar to that
reported in ref. [8].

3.3 Preparation of ZnS–PUMM nanocomposite films

UV curable macromers (UMM)–DMF solution of desired weight ratio, containing 2 wt% of
Darocur1173 and 0.5 wt% of AIBN as initiator were mixed with 3 ml of thiol-capped
colloidal ZnS–DMF solution according to the required doping content of ZnS particles in the
films. The mixture solutions were concentrated to a suitable viscosity at room temperature
under vacuum, and the resulting viscous solution was spincoated on silicon wafers or
quartz plates at 1000–3000 rpm.
The coated films were dried under vacuum for 10 min at 45 uC, and then were exposed to the
UV radiation of a medium pressure mercury lamp of 2 kW for 3 min. After UV curing reaction,
the films were cured at 70 uC for 2 h, 100 uC for 1 h, 120 uC for 1 h and treated at 160 uC for 0.5 h.

Fig. 5. TEM micrographs of the nanocomposite films of (a) TCZnS32 and (b) TCZnS79.
118 Nanocrystals

Resulting TEM micrographs of the nanocomposite films TCZnS32 and TCZnS79 are show in
Fig. 5. The ZnS nanoparticles ranging from 2 to 5 nm are uniformly dispersed inside the
polymer matrix and the ZnS nanoparticles remain their original size without aggregation
after immobilization into the polymer matrix, indicating that the thiol capping agents and
polymer play an important role in stabilizing and dispersing nanoparticles.
The main proprieties of nanocomposites with different content of ZnS nanoparticles show in
Table 1.
The WXRD pattern of ZnS nanoparticles synthesized in the work shows broad peaks typical
of samples in the nanosize regime (Fig. 6). The peaks in the diffraction pattern appearing at
2h values of 28.5, 47.5 and 56.3u correspond to (111), (220) and (311) planes of the cubic
structure of sphalerite ZnS. The resulting ZnS crystallite structure is in accordance with that
reported previously.

Fig. 6. Wide-angle X-ray diffraction pattern of PhSH–TMSt-capped ZnS nanoparticles.

The FTIR spectrum of the PhSH–TMSt-capped ZnS nanoparticles is shown in Fig. 7.

Fig. 7. FTIR spectrum of PhSH–TMSt-capped ZnS nanoparticle powder, dried from colloidal
ZnS–DMF solution in vacuum.
A review of high nanoparticles concentration
composites: semiconductor and high refractive index materials 119

The peaks at 2920, 2855 and 1406–1477 cm-1 are assigned to the characteristic vibration of the
methylene groups in TMSt. The peaks assigned to C–C vibrations of benzene rings are
observed at 1603, 1577, 1508 and 690–990 cm -1. The stretching vibration band of vinyl
groups on TMSt located at 1629 cm -1 also appears in Fig. 7 (insert), although its intensity is
very weak. The absorption peak of the S–H vibration at 2550–2565 cm -1 is not observed in
the IR spectrum, indicating that the mercapto groups of TMSt and PhSH molecules were
bound to the ZnS nanoparticle surface. The band of the CLO stretching vibration on residual
DMF molecules is also observed at 1647 cm -1 which is lower than that of free DMF
molecules at 1667 cm -1. This shift reflects that there is a relatively strong interaction between
DMF molecules and the surface of colloidal ZnS nanoparticles.28 In addition, the broad
peak near 3404 cm -1 in the IR spectrum may be from the absorption of traces of moisture or
adsorbed water associated with DMF.
The chemical composition of the thiol-capped ZnS nanoparticles was determined by EA and
ICP-AES analyses. Anal. Found: C, 30.5; H, 2.66; S, 25.30; Zn, 42.90. The relative molar ratio
of S to Zn was calculated to be 1.2 on the basis of the above quantitative analyses. This result
is in good accordance with the EDAX result for ZnS particles, which showed that the ratio of
the number of S to that of Zn is 1.18 within an accuracy of 2%. If all of the capping agents
(RSH) are capped on the ZnS particles, the molar ratio of Zn2+ : S2- : RS- was calculated to be
1 : 0.6 : 0.6, based on the feed ratio (Zn2+ : RSH ~ 1 : 0.6), and the contents of carbon and
hydrogen on the ZnS particles were also calculated to be 31.8 and 2.78, respectively. The
contents of these two elements agree well with the results of chemical analyses (EA and ICP-
AES) within experimental error. Therefore, it can be deduced that almost all of the capping
agents, PhSH and TMSt, were capped to the ZnS particles and the molar ratio of Zn2+ : S2- :
RS-2 for the PhSH–TMSt-capped ZnS particles was determined to be 1 : 0.6 : 0.6.

3.4 Nanocomposite films

FTIR spectra of the ZnS–PUMM nanocomposite films for the TCZnS16–TCZnS79 are shown
in Fig. 8 (X is weight percent of PhSH–TMSt-capped ZnS particles in the films). The
absorption peaks at 3300 and 1657 cm-1 are attributed to N–H and CLO bonds of urethane
linkages and the latter covers the characteristic peaks of CLO of methacrylate groups.

Fig. 8. FTIR spectra of nanocomposite films of TCZnS16–TCZnS79.

120 Nanocrystals

The CLO band gradually decreases in intensity with increasing TCZnS nanoparticles,
compared with the intensity of the characteristic absorption band of phenyl (688–740 and
1600 cm-1). This result indicates that an increasing amount of TCZnS with high phenyl
content is immobilized into the PUMM matrix. The IR absorbances of the CLC double bonds
at 1629–1639 cm-1 for the methacrylate groups and the capping agent (TMSt) disappear,
indicating that they have completely polymerized. Fig. 9 illustrates TGA curves of pure
PUMM, TCZnS16, TCZnS48 and TCZnS86 films at a heating rate of 10 oC min-1 under
nitrogen atmosphere from 50 to 750 oC.

Fig. 9. TGA curves of nanocomposite films of TCZnS16, TCZnS48 and TCZnS86 at a heating
rate of 10 oC min-1 under nitrogen flow.

The nanocomposite films have the initial decomposition temperatures of 201, 204 and 203 oC
for TCZnS16, TCZnS48 and TCZnS86 respectively, and these values relate to the
decomposition temperature of the polymer matrix (PUMM). There are two obvious weight
loss regions: between 200 and 330 oC, and from 550 to 600 oC. The weight loss between 200
and 330 oC is predominantly attributed to the weight loss of the polymer matrix. The
secondary weight loss at 550–600 oC is considered to be the weight loss of another part of the
polymer and the thermal decay of the partial thiol-capped agents on the surface of ZnS
nanoparticles. As shown in Table 1, the residues of the nanocomposite films TCZnS16–86 at
750 oC are in the range of 11.5–56.5% and they increase with increasing TCZnS content in the
films. By and large, these char yields are in agreement with the theoretical weight fraction of
inorganic ZnS contained in the films, indicating that the ZnS particles were successfully
incorporated into the polymer matrices. This result also effectively supports the chemical
analysis results for the TCZnS particles. Fig. 10 shows DSC curves of PUMM, TCZnS16,
TCZnS48 and TCZnS79 films at a heating rate of 10 oC min-1 under nitrogen flow.
A review of high nanoparticles concentration
composites: semiconductor and high refractive index materials 121

Fig. 10. DSC curves of PUMM, TCZnS16, TCZnS48 and TCZnS79 films at a heating rate of 10
oC min-1 under nitrogen flow.

The pure PUMM polymer exhibits a glass transition temperature (Tg) of about 106 oC.
However, no significant thermal transition peaks are observed for the ZnS–PUMM
nanocomposites below 200 oC. Also was used the torsion braid analysis (TBA) to measure
the thermal transition behavior of the nanocomposite samples. The Tg of the polymers still
are not observed. This suggests that nanocomposites have higher rigidity and crosslinking
density due to the incorporation of the thiol-capped ZnS nanoparticles, which restricted the
motion of the polymer chain segments. Thus, it may be that the glass transition temperature
of the polymer is close to the decomposition temperature of it. Thermal analyses indicate
that the ZnS nanoparticles were successfully immobilized into the polymer matrix and the
nanocomposite films exhibit a good thermal stability.
Maximal weight concentration of thiol capped ZnS is around 86% as at higher concentration
nanocomposite films with good mechanical properties cannot be obtained.
Another example of ZnS and CdS high concentration nanocomposite give the work [13, 14].
In the work has involved nanoparticles stabilization primarily by steric barriers. So, each
nanoparticle should be covered by quite a thick shell linked with the surface of nanoparticle.
This was accomplished by having a multi-atom chain connecting the acid group and
aromatic group of the shell molecule.
Was used UV curable monomers with an acid group at one end and a vinyl group at the
other and low viscosity at room temperature to accomplish both bonding to nanoparticle
surface and, same time, possibility to UV-curing. Based on these requirements one suitable
substance has been chosen: 2-carboxyethyl acrylate (CEA). This substance has an acrylic
group for curing and can be used as shell material because of the acid group. This molecule
has only a short distance between the groups, and only a relatively thin shell would be
expected before polymerization. And we know that a thin shell formed at synthesis of
nanoparticles is not enough at high nanoparticles concentration in a thermoplastic matrix.
Certainly, we would expect additional monomer units to add to the shell monomers during
the UV curing reaction.
122 Nanocrystals

The method to introduce the ZnS nanoparticles into CEA found is the following: ZnS
nanoparticles with the shell of 5-Phenylvaleric acid have been put into toluene and heated
for 10 hours at 80 C. This operation is needed to remove residue water from the
nanoparticles surface. After the nanoparticles powder was dried in air at 80 C during 10
minutes, it was put into CEA. Ultrasonic dispersion for about 30 – 40 minutes in apparatus
was done.
After dissolution of ZnS into CEA, a shell of CEA is formed at the surface of each
nanoparticle. As the result ZnS – CEA nanocomposite dispersion (or solution) is stable for a
long time.
UV curing of nanocomposite was made by usual way by addition of photoinitiator and
curing a film with UV light. Experimental conditions: photoinitiator Dimethoxy phenyl
acetophenone 0,1 w%, film thickness 100 um, UV light 365 nm, 5 mW/cm2, room
temperature, time of curing 10 minutes. Maximal ZnS volumetric concentration in the
compositions was 25%. The resulting RI of the UV cured film was 1,65, compared with an
RI of 1,45 for the pure CEA film. Thus, the RI increase is 0,20. Dependence of RI on
nanoparticles concentration is shown in Fig. 11 Maximal nanoparticles concentration has
been limited by viscosity increase up to the point of a non-flowing composition.

Fig. 11. Dependence of nanocomposite RI on nanoparticles concentration.

TEM photo (Fig 12) shows inside structure of nanocomposite. The photo was made of a
nanocomposite cured immediately after preparation. The nanoparticles concentration is 20
vol.%. Note that the distribution of particles is almost uniform, which explains the
transparency.
A review of high nanoparticles concentration
composites: semiconductor and high refractive index materials 123

Fig. 12. TEM photo of ZnS-CEA nanocomposite (nanoparticles concentration is 20 vol. %)

4. Electron and hole transport over disperse nanocomposite systems and

nanolayers
At present time charge transport over disperse semiconductor were investigated mostly for
organic conjugated materials. For high concentration nanocomposite materials are used
same understanding. In this chapter will be considered charge transport over disperse
organic materials as the nearest analog of nanocomposite.
At organic materials when atoms are bonded together to form a molecule, the upper atomic
orbitals interact with each other to form delocalized molecular orbitals while the deep
atomic orbitals are still localized in the atomic potential well (Fig. 13 a) [15].

Fig. 13. Electronic structures of (a) a polyatomic organic molecule or a single chain polymer
and (b) an organic solid.
124 Nanocrystals

When the orbital overlap occurs directly between the nuclei of the atoms, the orbitals form
σ-bond and the sideway overlapping of the orbitals form π-bonds. Materials having π-bonds
orbitals are named as conjugated materials same time. It is the π electrons which mainly
determine the electronic and optical properties of the molecule. In ground state, the π-
electrons form the π-band and the highest energy π-electron level is known as the highest
occupied molecular orbital (HOMO). In excited state, the π-electrons form the π-band and
the lowest energy π-electron level is known as the lowest unoccupied molecular orbital
(LUMO). The HOMO resembles the valence band and the LUMO resembles the conduction
band in the inorganic semiconductor concepts. The energy separation between the HOMO
and the vacuum level corresponds to the gas phase ionization energy (I ) and that between
g
the LUMO and the vacuum level corresponds to the gas phase electron affinity (A ) [2.1]. In
g
an organic solid, the molecules or polymer chains are packed closely together and result in
an electronic structure as shown in Figure 13b. It can be observed that the electronic states
are localized to individual molecules with narrow intermolecular band widths.

4.1 Excitons in organic

When an electron has been excited from the ground state orbital to a higher orbital, it leaves
a hole in the ground state orbital. The resulting bound state of an electron and a hole due to
the Coulombic interaction is called an exciton. In general, excitons can be divided into two
classes. If it is delocalized with radius much larger than the interatomic spacing, it is a Mott-
Wannier type of exciton. On the other hand, if it is localized and tightly bound, it is called a
Frenkel exciton. In organic materials, since the excitations are often localized on either
individual molecules or a few monomeric units of a polymer chain, the excitons are highly
localized and are considered to be Frenkel excitons which usually have large binding energy
of some tenths of an eV or even higher. For example, in the case of Alq , a commonly used
3
OLED material, the exciton binding energy is ~ 1.4 eV.

4.2 Transport
According statement in beginning of this paragraph, properties of disperse systems will be
considered in example of conjugated polymers charge transport theory and abbreviation of
polymer semiconductor is suitable for other disperse systems included nanoparticles in
polymer matrix. According to this in mostly part of recent works nanoparticles like fullerene
C60 are investigated with large organic molecules like phthalocyanine and perylen together
as component of polymeric disperse compositions.[16]
Electronic properties of polymers can be described in terms of semiconductor physics [17].
The particular framework of one dimensional periodic media is well suited to the basic
understanding of an isolated polymer chain [18]. Polymers are bonded by strong covalent
bonds. As π-orbitals overlap is weaker than s-orbitals overlap, the energy spacing (band
gap) between bounding and antibounding molecular orbitals is larger for the π–π∗
difference than for the σ –σ∗ one. One can thus, in a first approach, limit the band study to
the π–π∗ molecular orbitals. Those are respectively the HOMO (for Highest Occupied
Molecular Orbital) and LUMO (for Lowest Unoccupied Molecular Orbital), in terms of
molecular physics. They are also the usual valance (VB) and conduction bands (CB) of
A review of high nanoparticles concentration
composites: semiconductor and high refractive index materials 125

semiconductor physics, respectively σ -bonds then only contribute to the stability of the
molecular structure.
In a real material also, 3-dimensional interactions play a major role in transport properties,
even dominating the transport which becomes an interchain hopping process. Small
molecules are bounded by weak interactions in the condensed state: Van der Waals forces.
There results a weak coupling between them, the resonance integral t1 is thus small (tenth of
eV at most) [19], resulting in narrow flat bands. Mobility is thus a priori smaller in small
molecules, owing to a large effective mass

 
2 2 1
* n  E
m  2 2
k

There can of course be exceptions to such a rule, the interdistance spacing can be small and
molecular materials can in fact possess a rather large mobility. The first electrically pumped
injection organic laser was indeed made from small molecules (a tetracene single crystal)
[20].
Transport and mobility in organic materials require a knowledge of the charged species. A
review of transport properties is given by Schott [21]. Energy levels of the charges are
usually determined by cyclic voltametry for materials in solution. They can be characterized
by XPS or UPS (X-ray and UV photoelectron spectroscopies) for solid materials. In small
molecules, charged species are localized spatially, they are simply the cation (positive) and
anion (negative) radicals. In polymers, the electron–phonon coupling leads to the so-called
polarons which are charges dressed by a reorganization of the lattice [22]. Polarons may be
regarded as defects in conjugated polymer chains. Such defect stabilises the charge which is
thus self-trapped as a consequence of lattice deformation. So in the vast majority of organic
semiconductors, transport bears all characteristics of a hopping process in which the charge
(cation or anion) propagates via side to side oxidation–reduction reactions (Fig. 14.).

Fig. 14. Hopping process between molecules 1 and 2. b- Intra- (full arrows) and
intermolecular (broken arrows) charge-transport.

One must distinguish between intramolecular charge transport along a conjugated polymer
chain and intermolecular charge transport between adjacent molecules or polymer chains
(Fig. 14 b.).
The former which is specific to conjugated polymers is the most efficient. Charge mobility in
organics is field dependent, especially in the law mobility materials in which it usually
follows phenomenologically a Poole–Frenkel law:  exp( E ) [23]. Mobility can be
126 Nanocrystals

experimentally determined by photo-current transients (time of flight) [24], field effect

transistor saturation currents [25], space charge limited currents [26] or impedance
spectroscopy [27]. Mobilities in organic semiconductors are usually rather small: from 10−2
in well ordered conjugated polymers (liquid crystalline polyfluorene), down to 10−8
cm2/(V·s) in guest–host polymer systems (dye doped poly-vinylcarbazole – PVK. Electron
and hole mobilities differ by orders of magnitude in a single material; in small molecules
such as the widely studied tris (8-hydroxyquinolinolato) aluminium – Alq3 – as well as in
conjugated polymers such as the famous poly-paraphenylvinylene – PPV. The lowest
mobilities are usually dispersive, which is the result of a distribution of mobilities [28].
Mobility can increase by up to two decades upon applying a voltage, being eventually very
large above 1 MV/cm in conjugated polymers [29]. Mobility is increased by orders of
magnitude when the molecular packing is improved. This is achieved by molecular
ordering. Single crystals have the best performances, electron mobility in fullerene C60
single crystals is 2.1 cm2/(V·s) [30], but it is reduced by at least 3 orders of magnitude by
imperfect purification and uncontrolled crystallization [31], as well as by oxygen traps.
Charge transport is also improved by purification or deposition conditions; for instance,
mobility becomes non-dispersive in Alq3 upon purification (oxygen induces traps) [32] and
it becomes non-dispersive in soluble PPV derivatives upon selection of the solvent used for
deposition [33]. Mobility is usually low and dispersive in randomly distributed polar
molecules, but it is increased significantly when the dipoles are organized [34]. A record non
dispersive electron mobility of up to 2*10−4 cm2/(V·s) was recently achieved in an air stable
amorphous glassy molecular material [35]. It is important that the mobility always drops by
at least 2 orders of magnitude with impurities or defects (traps).
Light absorption and photogeneration process depend from nature of nanoparticles. If
nanoparticles is inorganic semiconductor like CdS or CdSe, processes of photogeneration
and absorption are similar to inorganic bulk crystals, is nanoparticles are pigment
nanoparticles like phthalocyanine for example, photogeneration and absorption processes
are similar to same in organic.
Charge transport over disperse system depend from concentration of charge traps in
material (Fig. 11.), if concentration is low and distance between it is long, mobility become
low. In nanocomposite (composition of inert polymer material and semiconductor
nanoparticles) condition of charge transport over it take part at high nanoparticles
concentration only. Lowest concentration limit is about 5 volumetric % of nanoparticles in
material, but usually about 30 - 40 vol. % concentration is optimal for obtain charge
transport over disperse nanocomposite material as over homogeneous media. Same time
achievement of 30 vol. % of nanoparticles in material is a difficult task for technology.

5. Application of semiconductor nanostructures: solar cells, OLED

Important area of semiconductor nanoparticles application is n-type semiconductor layers in
organic multilayer thin film structures as addition to usual p-type organic semiconductor.
Typical structure of organic solar cell as well as structure of OLED (organic light emitting
device) consist organic thin layers: p-type for holes injection (OLED) or extraction (solar
cell), n-type - for electrons injection or extraction in corresponding application.
A review of high nanoparticles concentration
composites: semiconductor and high refractive index materials 127

5.1 Principle of operation of solar cell (PV - photo Voltaic) and OLED structures
Before discussing the development of organic PVs the basic principles are outlined. Almost
all organic solar cells have a planar-layered structure, where the organic light-absorbing
layer is sandwiched between two different electrodes. One of the electrodes must be (semi-)
transparent, often Indium–tin-oxide (ITO), but a thin metal layer can also be used. The other
electrode is very often aluminum (calcium, magnesium, gold and others are also used).
Basically, the underlying principle of a light-harvesting organic PV cell (sometimes referred
to as photodetecting diodes) is the reverse of the principle in light emitting diodes (LEDs)
(see Fig. 15) and the development of the two are somewhat related [36].

Fig. 15. A PV device (right) is the reverse of a LED (left). In both cases an organic material is
sandwiched between two electrodes. Typical electrode materials are shown in the figure. In
PVs electrons are collected at the metal electrode and holes are collected at the ITO
electrode. The reverse happens in a LED: electrons are introduced at the metal electrode
(cathode), which recombine with holes introduced at the ITO electrode (anode).

Fig. 16. Energy levels and light harvesting. Upon irradiation an electron is promoted to the
LUMO leaving a hole behind in the HOMO. Electrons are collected at the Al electrode and
holes at the ITO electrode. F: workfunction, c: electron affinity, IP: ionisation potential, Eg:
optical bandgap.
128 Nanocrystals

In LEDs an electron is introduced at the low-workfunction electrode (cathode) with the

balanced introduction of a hole at the high-workfunction electrode (anode). At some point
the electron and the hole meets, and upon recombination light is emitted [37]. The reverse
happens in a PV device. When light is absorbed an electron is promoted from the highest
occupied molecular orbital (HOMO) to the lowest unoccupied molecular orbital (LUMO)
forming an exciton (see Fig. 16).
In a PV device this process must be followed by exciton dissociation. The electron must then
reach one electrode while the hole must reach the other electrode. In order to achieve charge
separation an electrical field is needed, which is provided by the asymmetrical ionisation
energy/workfunctions of the electrodes. This asymmetry is the reason why electron-flow is
more favoured from the low-workfunction electrode to the high workfunction electrode
(forward bias), a phenomenon referred to as rectification. The light harvesting process along
with the positioning of energy levels is depicted in Fig. 16.
In the solid phase, the HOMOs and LUMOs of adjacent molecules may interact and form a
conduction band (CB) and a valance band (VB) respectively (this will be described below).
The shape of the CB and VB changes when the organic material is put into contact with
electrodes (see Fig. 17), depending on the conductance of the polymer and on whether the
electrodes are connected or not.

Fig. 17. The relative energy levels of the electrodes, CB and VB are shown in three situations,
with no external bias. (A) CB and VB are shown along with the low-workfunction electrode
(Al) and the high workfunction electrode (ITO) when isolated from each other. (B and C)
The cell is assembled and short circuited, causing alignment of the electrode potentials. In
(B), an insulating organic material is used. In C a hole-conducting polymer is used forming a
Schottky junction at the high-workfunction electrode.

If the cell is short circuited the Fermi levels of the electrodes align (B and C), and in doing so
the CB and VB are pulled skew. In B the polymer material is an insulator. This gives a field
profile that changes linearly through the cell. In C a hole-conducting (p-type) semiconductor
is used (most polymers are much better hole conductors than electron conductors). If the
material is doped or illuminated charge carriers are generated. Due to the p-conduction
properties, the generated holes are allowed to redistribute freely and they will flatten the
bands approaching the high-workfunction electrode (a Schottky junction). The distance over
which the CB and VB exhibit curvature is called the depletion width. In B the depletion
width extends throughout the material. In C the depletion width is less than half the
material thickness. Under external bias the relative electrode potentials can be changed,
depending on the size and direction (forward or reverse) of the bias.
A review of high nanoparticles concentration
composites: semiconductor and high refractive index materials 129

5.2 Comparison of inorganic and organic PV

In a crystalline inorganic semiconductor with a 3D crystal lattice the individual LUMOs and
HOMOs form a CB and a VB throughout the material. This is fundamentally different from
most organic dye semiconductors where the intermolecular forces are too weak to form 3D
crystal lattices. Consequently the molecular LUMOs and HOMOs do not interact strongly
enough to form a CB and VB. Thus charge transport proceeds by hopping between localised
states, rather than transport within a band. This means that charge carrier mobility in
organic and polymeric semiconductors are generally low compared to inorganic
semiconductors. Also, charge separation is more difficult in organic semiconductors due to
the low dielectric constant. In many inorganic semiconductors photon absorption produces
a free electron and a hole (sometimes called charge carriers), whereas the excited electron is
bound to the hole (at room temperature) in organic semiconductors. Conjugated polymers
lie somewhere between the inorganic semiconductors and organic dyes. In general, excitons
are considered to be localised on specific chain segments. However, there are cases where
excitons seem to be delocalised. In these cases, the excitons are referred to as polarons [38, 39].
In simple PV devices and diodes based on organic semiconductors the primary exciton
dissociation site is at the electrode interface (other sites include defects in the crystal lattice,
absorbed oxygen or impurities) [40]. This limits the effective light harvesting thickness of
the device, since excitons formed in the middle of the organic layer never reaches the
electrode interface if the layer is too thick. Rather they recombine as described above.
Typical exciton diffusion distances are on the order of 10 nm.
In fact inorganic PV have the best proprieties of charge carrier photogeneration and transport,
polymer materials have the worst same proprieties. Organic crystalline dyes is lie between
inorganic semiconductors and polymers. However flexible solar cell can be made in the basis
of polymer films mostly, organic crystalline dyes can be introduced to structure as a thin layer,
but inorganic semiconductors layers can not to be introduced in this structure absolutely.
From this point of view use of polymer based nanocomposite included semiconductor
nanocrystals of different types (inorganic semiconductors - TiO2, CdS; fullerenes; carbon
nanotubes; organic crystalline dyes in nanocrystalline state) is the single way to combine
high semiconductor proprieties of inorganic and crystalline materials with good flexibility
of polymer film.
There are another advantage of inorganic semiconductor nanocrystals for PV and OLED
application:
All semiconductors have a proprieties of production of single atom oxygen at presence of
oxygen molecules and its illumination by light, especially in UV region. Presence of water
vapor made this process more effective. Both PV and OLED device use at presence of light
certainly and oxygen and water from air. Single atom oxygen produced by action of light is
very reaction active chemical. Its action to thin nanometers sized polymeric structure will
result on fast its destruction and degradation.
In the work [41] show results of degradation of PV structure at action of oxygen and the
light. In fact presence of oxygen and the light illumination are standard conditions of PV
and OLED structure application as oxygen and water vapors go easily over polymer film of
flexible structure.
Fig. 18 shows the degradation of the photocurrent of a plastic solar cell under 1000W/cm2
illumination inside the sealed container. Already within 1 h the effect is very pronounced. It
also shows how a temporary exposure of the cell to air leads to an accelerated decrease of
130 Nanocrystals

photoefficiency. The photocurrent seems to become inhomogeneous, suggesting selective

degradation. Higher degrees of degradation around the rim of the cell could be an effect of a
mechanism of lateral degradion through diffusion of oxygen or water vapor.

Fig. 18. Decay of photocurrent (Isc) in Cell as a result of degradation under 1000 W/cm2
illumination in an inert atmosphere, followed by exposure to the atmosphere.

From this point of view use of inorganic nanocrystals, more stable to oxygen action is
preferable too.
In despite of evident advantage of inorganic and organic nanocrystals application in PV and
OLED devices, practical examples are not numerous because of difficulties of thin layer
preparation from new and unknown materials nanocomposites as well as because of
nanocomposites preparation.
Now mostly applicable are: fullerene C60 nanoparticles, carbon nanowires, TiO2 and CdS
nanocrystals: all n-type semiconductors.

5.3 Example of PV structure based on fullerene

Excitons do not dissociate readily in most organic semiconductors. The idea to overcome
this obstacle is use a heterojunction: to use two semiconductor materials with different
electron affinities and ionisation potentials. This will favour exciton dissociation: the
electron will be accepted by the material with the larger electron affinity and the hole by the
material with the lower ionisation potential.
One of the most used acceptors in heterojunction cells is the fullerene C60 nanoparticles (or
large organic molecules in some publication) [42]. Besides having a high electron affinity,
C60 is fairly transparent and also has fair electron conductance (10-4 Scm-1). This makes
fullerenes a good component in PV cells. The first report of a conducting polymer/C60 cell
came in 1993 by Sariciftci et al. [43]. In one study, fullerene was vacuum sublimed onto a
MEH–PPV layer that was spin coated on ITO-covered glass . Au was used as the electron-
collecting electrode (see Fig. 19).
A review of high nanoparticles concentration
composites: semiconductor and high refractive index materials 131

Fig. 19. A two-layer heterojunction photovoltaic cell with C60 fullerene nanoparticles. The
electron accepting C60-layer contacts the Au electrode, while the electron donating MEH-
PPV layer contacts the ITO electrode.

In this example a defect of structure is a low light absorbing area close to border between C60
and MEH-PPV semiconductors. It is clear that exciton dissociation is most effective at the
interface in heterojunction cells, thus the exciton should be formed within the diffusion
length of the interface. Since typically diffusion lengths are in the range of 10 nm, this limits
the effective light-harvesting layer. However, for most organic semiconductors the film
thickness should be more than 100nm in order to absorb most of the light. It follows that
thicker film layers increase light absorption but only a small fraction of the excitons will
reach the interface and dissociate. This problem can be overcome by blending donor and
acceptor, a concept called dispersed (or bulk) heterojunction (see Fig. 20) [41].

Fig. 20. Dispersed heterojunction between a transparent ITO electrode and an Al electrode.

Certainly, this structure can be prepared by mixing of polymer or polymer like materials
only and use of nanocomposites here is without alternative. In 1994 Yu [44] made the first
dispersed polymer heterojunction PV cell by spincoating on ITO covered glass from a
solution of MEH-PPV and C60 in a 10:1 wt-ratio. Finally, Ca was evaporatedonto the organic
layer. The cell showed a photosensitivity of 5.5 mA/W, an order of magnitude larger than
the photosensitivity of the pure polymer.
132 Nanocrystals

One limitation of this approach is the relative low solubility of fullerenes in normal solvents.
This problem was solved when Hummelen et al. [45] synthesiseda number C60-derivatives
with increased solubility in 1995, which allowed the fullerene content to be as high as 80% in
the prepared films. Using a methano-functionalised fullerene derivative Yu et al. [46]
repeated the fabrication procedure with a polymer/fullerene ratio of 20/80, the contacts
were made of ITO and Ca, and the cell had a QE of 29% and a PCE of 2.9% (under
monochromatic light, intensity at 20 mW/cm2). Thus a substantial increase compared to
earlier polymer/fullerene mixtures. In 2000 Shaheen et al. [47] reportedhigh QE values of
85% in a PPV-derivative and fullerene heterojunction cell, with a PCE of 2.5%. All
approaches described above are to use high concentrated nanocomposites based on
fullerene C60 nanoparticles stabilized by linking to polymer chain. Different structures of
fullerenes based nanocomposites materials are show in Fig. 21.
It is clear that the control of morphology in dispersed heterojunction devices is a critical
point. The degree of phase separation and domain size depend on solvent choice, speed of
evaporation, solubility, miscibility of the donor and acceptor etc. One strategy towards
increasing control is to covalently link donor and acceptor. In 2000 Stalmach et al. [48]
synthesized PPV-C60 diblock copolymers through controlled living radical polymerization.
The same year Peeters et al. [49] synthesized a number of p-phenylene vinylene oligomers
(OPV) attachedto C60 and investigated their use in PV devices. Peeters found that charge
separation lifetimes was dependent on the number of repeating oligomer units. Thus, charge
separation lifetimes were much longer for 3–4 units compared to 1–2 repeating units. A cell
consisting of the longest oligomer (4 repeating units) between aluminium and a PEDOT-PSS
covered ITO electrode had an ISC of 235 mA/cm2 and a VOC of 650 mV, but a relatively low
FF of only 0.25. Van Hal et al. [50] made a similar study on fullerene oligio (thiophene)-
fullerene triads varying the number of monomer units. In agreement with Peeters they
found that a certain length is needed in order to observe charge transfer upon excitation.
Thus photoinduced charge transfer was much more pronounced for 6 monomer units,
compared to an oligomer with 3 monomer units. Such model studies are important in
understanding charge transfer and light harvesting in greater detail. In 2003 Krebs et al. [51]
have made an interesting study on a dyad consisting of a poly(terphenylene cyanovinylene)
terminated with an ADOTA dye. The dye is a cation and the assembly thus resembles a soap
molecule and have the ability to form LB films. By spincoating the dyad on an ITO covered
glass substrate, followed by evaporation of Al on the organic layer, the short circuit current
of the dyad was 100 fold larger compared to the pure polymer.
While some control is introduced by covalently linking the donor and acceptor in
polymer/oligomer-C60 assemblies the final morphology may suffer from phase separation
and clustering of the fullerene (or dye) units, which potentially limits efficient charge
separation due to low donor/acceptor interfacial area. Also, increased phase separation may
disrupt the continuity of the phases and reduce the charge transport properties of the
material, due to inefficient hopping between different domains, reducing overall
performance. This may be a critical point, since intramolecular charge recombination might
occur at a fast rate. One way to control a bicontinuous phase separation andinsure a large
interfacial area between donor and acceptor is to covalently graft fullerene moieties onto the
donor-polymer backbone (Fig. 21), so-called double-cable polymers (due to their p/n type
conduction properties).
A review of high nanoparticles concentration
composites: semiconductor and high refractive index materials 133

Fig. 21. Different morphologies of heterojunction cells based on fullerene nanocomposites.

Top, left: Two-layered structur e of fullerenes and polymer chains. Top, right: dispersed
heterojunction. Middle, left: fullerenes with polymer chains attached. Middle, right: self-
assembled layered structure of double-cable polymers. Bottom: self-assembled layered
structure of diblock copolymers. The layered structure of double-cable polymers and
diblock copolymers are expected to facilitate efficient electron and hole transport.

These assemblies have been intensively investigated in recent years as promising

components in PV devices [52], but they are also interesting as components in molecular
electronics. The first reports of polymers bearing fullerene on the side chains came in 1996
by Benincori et al. [53] The polymer was a polythiophene. Cravino et al. [54] synthesiseda
fullerene-thiophene double-cable and used it in a solar cell. However the PV device had
limited cell efficiency due to low levels of fullerene content (C60 was attached to 7% of the
repeating units). It is important that the fullerene content reaches the percolation threshold
to insure efficient electron transport. Thus as the complexity of the designed systems
increase the more critical it becomes to optimize design parameters. Even though the
synthesized double cable polymers have shown some promising results, we have yet to see
the fully optimized double-cable polymers. An alternative to polymer-fullerene double-
cables is block copolymers consisting of a donor and acceptor block. In general block
copolymers are known to phase separate and form ordered domains similar to the double-
cable polymers. In 2003 Krebs et al. [55] synthesiseda block copolymer consisting of an
electron acceptor block and an electron donor block. The backbone was polyacetylene, and
by using phenyl and pentafluorophenyl as side groups the HOMO and LUMO of the
individual blocks could be tuned so that hole or electron conductance is favored.
134 Nanocrystals

5.4 D well ordered bulk heterojunction based on carbon nanotubes

Dispersed heterojunction described above ensure augmentation of effective surface of
junction between n and p- types of semiconductors, but same time irregular structure of n
and p- types areas will result on augmentation of length of charge transportation from n and
p semiconductors border to the electrodes (Fig. 20). Certainly if to make regular 3-D
structure of different phases of heterojunction, it is possible to diminish charge
transportation length up to thickness of the layer. This structure can to resolve problem of
relative high exciton energy dissociation of organic material. Indeed, high exciton energy
dissociation in organic in comparison with inorganic will result on diminish of organic solar
cell efficiency (problem was described in details above of this paragraph). Same time use of
polymers for flexible solar cell thin film have no alternative in classical technique as
polymers materials only allow to form thin flexible film. However, if to use technique of
self- assembly, it is possible to prepare polymer free thin film consist from inorganic
nanoparticles and to obtain inorganic high efficiency and stable solar cell based on
nanoparticles assembly structure.
However, for practical making of these structure are need to prepare 3-D elements with
tens nanometers sizes with artificial distribution of n- and p- types of semiconductor nano-
sized areas that is too difficult at the present time. Now only way to make so is to use self-
assembly technique for preparation of nanometers sized vertical columns from carbon
nanotubes well ordered. The spaces between columns should be filled by another
semiconductor. This technique will allow to produce well ordered heterojunction with
charges transport distances minimized in comparison to structure in Fig. 20. At the present
time investigations of self- assembly carbon nanotubes processes is in beginning only. There
are a lot of both scientific and technological problems. The first practical realization of this
way give a work [56]. In the fallowing text will be considered semiconductor solar cell
strictures based on single carbon nanotubes and its combination with other types of nano-
semiconductors: CdS nanoparticles and porphyrines organic pigments. The principle of self-
assembly structure shown in Fig. 22.

Fig. 22. Illustration of random charge carrier transport in bulk heterojunction (in the left) -
charge transportation length is long and well oriented 3-D nanostructure carbon-nanotube-
directed charge transport in an organized hybrid assembly.

Certainly should be take into account that each nanotubes have diameter of 1,5 nm and it is
very difficult to made really this structure.
A review of high nanoparticles concentration
composites: semiconductor and high refractive index materials 135

Commercially available semiconductor carbon nanotubes (SWNTs) contain both metallic

and semiconducting nanotubes with different chirality. The work function of SWNT
bundles is known to be about -4.8 eV versus absolute vacuum scale (AVS). Carbon
nanotubes possess a bandgap in the range of 0-1.1 eV, depending upon their chirality and
diameter.
Semiconducting carbon nanotubes undergo charge separation when subjected to bandgap
excitation. Two different approaches can be considered for the use of carbon nanotubes in
solar cells (Fig. 23):
(i) direct bandgap excitation of semiconducting nanotubes; or (ii) the use of conducting
tubes as conduits to improve the transport of charge carriers from light-harvesting
nanoassemblies.
The methods employed to deposit carbon nanotubes as thin films on a conducting surface
for use as photoresponsive electrodes in solar cells are discussed in the following sections.
Examples of the two strategies presented in Fig. 23 are also illustrated.

Fig. 23. Strategies to employ carbon nanotubes in photochemical solar cells: (left) by direct
excitation of carbon nanotubes; (right) by excitation of light-harvesting assemblies anchored
on carbon nanotubes. The electrons and holes generated by photoexcitation are referred to
as e and h, respectively. One of these charge carriers is collected at the electrode surface and
the other one is scavenged by the oxidized (O) or reduced (R) form of the redox couple in
the electrolyte.

Photoinduced charge separation in SWNT films An interesting semiconducting property of

SWNTs is their ability to respond to light. For example, the photoresponse of carbon
nanotubes filaments was realized in early years from the elastic response of the aligned
bundles between two metal electrodes [57]. Avouris and coworkers [58] have monitored hot
carrier luminescence from ambipolar carbon nanotube field-effect transistors (FETs). The
holes and electrons injected via an external circuit produce emission resulting from electron-
136 Nanocrystals

hole recombination in the system. The recent report of bandgap fluorescence from a
semiconducting SWNT sample rich in individual nanotubes has made it possible to
correlate optical properties with individual tube species as a result of their well-defined
optical transitions. Spectroscopic studies have demonstrated that the relaxation of electrons
and holes to the fundamental band edge occurs within 100 fs after photoexcitation of the
second van Hove singularity of a specific tube structure [59]. These early studies confirm the
ability of carbon nanotubes to possess a band structure that can undergo electron-hole
charge separation with visible light excitation (Fig. 24).

Fig. 24. Schematic illustrating the density of states of a single carbon nanotube.
Photogenerated holes are captured at the collecting electrode surface, resulting in current
generation in a photoelectrochemical cell. C1 and C2 refer to conduction bands and V1 and
V2 refer to valence bands. e and h refer to the electron and holes generated following
photoexcitation of the SWNTs.

In order to use photogenerated charge carriers for generating electricity, it is important that
they are separated before undergoing recombination. However, spatially confined charge
carriers in the nanotube are bound by Coulombic interactions with the bound pair referred
to as an exciton [60]. Most of these excitons from higher C2 and V2 levels relax via interband
transitions to the low-lying C1 and V1 levels of the fundamental gap to produce a second
sub-bandgap exciton. A small fraction of the excitons are able to dissociate and form
unbound electron-hole (e-h) pairs. The dissociation of excitons to create the charge-
separated state thus becomes an important process to tap them into photocurrent
generation. The charge separation in carbon nanotubes can be probed using femtosecond
laser pump-probe spectroscopy. This technique is useful to investigate the ultrafast
processes that occur following the excitation of carbon nanotubes or semiconductor
materials. In a typical experiment, the absorption changes in the sample are recorded at
different delay times following excitation with a short laser pulse. Difference absorption
spectra at various delay times have been recorded by exciting SWNT suspensions in THF
A review of high nanoparticles concentration
composites: semiconductor and high refractive index materials 137

with a 387 nm laser pulse (pulse width 130 fs). (A Clark MXR-2010 laser system and
Ultrafast Systems detection setup was used for these measurements.) Representative
transient absorption spectra and the decay of absorption at 700 nm are shown in Fig. 25.

Fig. 25. Time-resolved transient absorption spectra of an SWNT suspension in THF (flow
cell) recorded using a 387 nm laser pulse (pulse width 150 fs; Δt = 0 corresponds to the end
of the pulse). The inset shows the bleaching recovery at 700 nm.

The photoexcitation causes the bleaching of SWNT absorption in the red region. The
broadness of the bleaching band essentially arises from the diversity of tube diameters,
chiral angles, and the aggregation of nanotubes.
The bleaching in the visible region, which corresponds to the C2-V2 transition, recovers in
~1 ps as the bound electron-hole pairs or excitons relax to the low-lying C1-V1 state. The
dynamics of the transient bleaching recovery and the decay of the emission in the infrared
arising from charge recombination in the fundamental gap have been studied recently by
Ma et al. [59]. They observed that the electron-hole pairs accumulate in the fundamental gap
(C1-V1) and their lifetime (10-100 ps) is dependent on the excitation intensity. Based on the
difference between the emission decay and transient absorption recovery, these researchers
highlighted the involvement of charge trap states as the additional contributing factors
responsible for electronic transitions. The presence of such surface states are likely to
stabilize the photogenerated charge carriers and contribute to the overall photocurrent
generation. Such enhanced charge separation is crucial for increasing the probability of
charge collection at the electrode surface. The transient bleaching observed following laser
pulse excitation shows that there is a significant number of charge carriers produced in the
SWNTs. The obvious question is whether one can collect the photoinduced charge carriers
generated in SWNTs suitably for photocurrent generation, similar to the photovoltaic
application of other semiconductors.
138 Nanocrystals

5.5 SWNT-semiconductor hybrids

In photoelectrochemical cells based on nanostructured or mesoscopic semiconductor films,
the electron transport across particles is susceptible to recombination loss at the particle
grain boundaries. The use of a nanotube support to anchor light-harvesting assemblies (e.g.
semiconductor particles) provides a convenient way to capture photogenerated charges and
transport them to the electrode surface. An illustration of these two scenarios can be seen in
Fig. 23.
SWNTs are an ideal candidate as conduit for collecting and transporting charges across
light-harvesting assemblies. Of particular interest is a CdS-SWNT composite that is capable
of generating a photocurrent from visible light with unusually high efficiency [61]. The
luminescence of CdS is quenched by SWNTs. Transient absorption experiments have
confirmed the quick deactivation of excited CdS on a SWNT surface, as the transient
bleaching recovers in about 200 ps.
In order to test the hypothesis of electron transfer between excited CdS and SWNT in the
composite film, CdS particles are deposited on SWNT electrodes (referred to as
OTE/SWNT/CdS) [62]. The SWNT film was first deposited on the OTE using the
electrophoretic deposition method described earlier. The electrode was then sequentially
immersed in solutions containing Cd2+ and S2- to form CdS nanocrystallites. The electrode was
thoroughly washed with deionized water between the two immersions so that only adsorbed
Cd2+ ions react with S2-. It may be noted that such an ion adsorption precipitation method is
similar to methods employed for casting nanostructured films of metal chalcogenides on oxide
films [63]. It is interesting to note that Cd2+ ions readily adsorb on SWNTs and react with S2- to
form CdS nanocrystallites with a characteristic onset absorption around 500 nm.
It is employed the OTE/SWNT/CdS electrode in a photoelectrochemical cell containing
acetonitrile solution with 0.1% triethanolamine as a sacrificial electron donor. Triethanolamine
undergoes irreversible oxidation as it scavenges photogenerated holes from the electrode surface.
Photocurrent generation is seen when the CdS-modified SWNT film is irradiated with visible
light (λ > 380 nm). An open circuit voltage of ~200 mV and a short circuit current of 6.2 μA were
recorded. The dependence of the IPCE on the excitation wavelength is shown in Fig. 26.

Fig. 26. IPCE of OTE/SWNT/CdS. Inset: absorbance difference between OTE/SWNT/CdS

and blank SWNT film.
A review of high nanoparticles concentration
composites: semiconductor and high refractive index materials 139

The onset of IPCE is seen at ~500 nm and closely follows the absorbance characteristics of
CdS (see inset of Fig. 26). The observed photocurrent is dominated by the initial excitation of
CdS as is evident from the photocurrent action spectrum. Furthermore, the anodic current
observed with SWNT/CdS films confirms the direction of the electron flow from CdS to the
collecting electrode mediated by the SWNT network. The ability of the CdS-SWNT
nanocomposite system to undergo photoinduced charge separation opens up new ways to
design light-harvesting assemblies.

5.6 TiO2 nanoparticles based PV

In the work [64] was used TiO2 nanoporous film in structure of dye sensitized solar cell.
Nanoporous film is real alternative to high concentration nanocomposite. Preparation of
high concentrated homogenous nanocomposite is a difficult task as nanoparticles tend to
coagulated at high concentration. Same time nanoporous film have uniform distribution of
pores with its high volumetric concentration - up to 60 %. If to fill nanopores by polymer we
obtain stable nanocomposite material (porous TiO2 film - see Fig. 27).

Fig. 27. SEM image of highly porous TiO2 layers consisting of a 3D interconnected network
of anatase crystallites used for device fabrication. Right - Energy diagram of
ITO/TiO2/MEH-PPV/Au.

With the control of the nanostructured morphology, metal oxides are believed to act as
promising alternatives as the electron acceptor and transporter in bulk-heterojunction solar
cells. Among the metal oxides, TiO2 is a very good candidate for this purpose because the
use of nanocrystalline TiO2 as electron accepting electrode for dye-sensitized solar cells has
shown an overall power conversion efficiency as high as 10%.
Mesoporous TiO2 films were deposited on conducting glass ITO or SnO2. By varying the
temperature in the nanoparticle synthesis, the average particle diameter was altered
between 20.5 and 41.5 nm. XRD measurements indicate that the TiO2 particles were pure
anatase within the detection limit of 3%–5%. Also, transmission electron microscopy
indicates that the particles were crystalline. The film porosity was 57.5%±1.5% and was
independent of the average particle size [64]. If to use MEH-PPV polymer as a p-type
semiconductor, reported maximum EQE is 6%, short circuit current density - 0.4 mA/cm2
and power conversion efficiency - 0.17% under 100 mW/cm2 white light illumination [65].
140 Nanocrystals

Takahashi et al. [66] reported short circuit current density of 0.35 mA/cm2 and power
conversion efficiency of 0.13% under the irradiation of AM 1.5 illumination (100 mW/cm2).
The short circuit current could be improved by about 3 times and power conversion
efficiency could reach 0.47% by blending MEH-PPV with [2- [2- [4- (dimethylamino)
phenyl]–ethenyl] -6- methyl -4H- pyran –4– ylidene] propanedinitrile (DCM). All results are
comparable to same received with carbon based nanocomposites described above.

6. Conclusion
At present time application of semiconductor nanoparticles in different devices like solar
sell and light emitting devices is a reality. However there are some problems connected with
low value of charge carrier mobility and high exciton dissipation energy in disperse
nanocomposite material. This problem can be solved by following way:
- augmentation of nanoparticles concentration in nanocomposite for preparation of hybrid
material in which charge carrier transportation go over nanoparticles and not polymer
matrix. Limitation: loss of flow proprieties and hence loss of good processability of material.
Maximal level of nanoparticles concentration is limited in 20 - 25 vol % of nanoparticles
concentration. This value is enough for charge carrier transport.
- formation of nanostructure in material layer to provide directed charge carrier transport
between electrodes with minimal path length. This can be made by nanoimprint or self
assembly methods. At present time there are the first examples of these nano- ordered
structures.
Both problems can be solved and nanocomposite materials go to factory scale use
application. For example, in conference (7-th International Conference ELECTRONIC
PROCESSES IN ORGANIC MATERIALS (ICEPOM-7) Ukraine, Lviv, May 26 – 30, 2008 was
being announced plastic nanostructured solar sell as a new product for commercial
production since 2009: T. Yoshida, M. Matsui, K. Funabiki, H. Miura, Y. Fujishita Plastic
solar cells employing electrodeposited nanostructured zno and organic photosensitizer,
developed by Center of Innovative Photovoltaic Systems (CIPS), Gifu University, Yanagido,
Japan. These results were published in the work [67].

7. References
1. A. Eychmu, J. Phys. Chem. B 104, 6514 (2000)
2. R. Rossetti, J. L. Ellison, J. M. Gibson and L. E. Brus, J. Chem. Phys. 80, 4464 (1984)
3. M. W. Peterson, M. T. Nenadovic, T. Rajh, R. Herak, O. Micic, J. P. Goral and A. J. Nozik,
J. Phys. Chem. 92, 1400 (1988)
4. M. S. El-Shall, W. Slack, W. Vann, D. Kane and D. Hanley, J. Phys. Chem. 98, 3067 (1994),
I. Yu. Denisyuk, T. R. Williams and J. Ed. Burunkova, Mol. Cryst. Liq. Cryst., 497,
(2008.)
5. Ashfaq Adnan, C.T. Sun, Hassan Mahfuz A molecular dynamics simulation study to
investigate the effect of filler size on elastic properties of polymer nanocomposites
// Composites Science and Technology 67 (2007) 348–356
6. Odegard GM, Clancy TC, Gates TS. Modeling of the mechanical properties of
nanoparticle/polymer composites. Polymer 2005; 46(2) p. 553–62.
A review of high nanoparticles concentration
composites: semiconductor and high refractive index materials 141

7. Binder K. Monte carlo and molecular dynamics simulations in polymer science. New York
(USA): Oxford University Press; 1995.
8. C. Lu , Z. Cui, Z. Li, B. Yang and J. Shen, J. Mater. Chem., 13, 526 (2003)
9.J. H. Golden, H. Deng, F. J. DiSalvo, J. M. J. Fre´chet and P. M. Thompson, Science, 268,
1463 (1995)
10. J. H. Golden, F. J. DiSalvo, J. M. J. Fre´chet, J. Silcox, M. Thomas and J. Elman, Science,
273, 782 (1996)
11. C. Gao, B. Yang and J. Shen, J. Appl. Polym. Sci., 75, 1474 (2000)
12. P. Barbaro, C. Bianchini, G. Scapacci, D. Masi and P. Zanello, Inorg. Chem., 33, 3180 (1994)
13. Igor Yu. Denisyuk, Todd R. Williams, Julia E. Burunkova Hybrid optical material based
on high nanoparticles concentration in UV-curable polymers – technology and
proprieties // Mol. Cryst. Liq. Cryst., Vol. 497, pp. 142–153, 2008
14. Todd R. Williams, Igor Yu. Denisyuk, Julia E. Burunkova Filled polymers with high
nanoparticles concentration – synthesis, optical and rheological proprieties //
Journal of Applied Polymer Science, Volume 116 Issue 4, P. 1857 - 1866, 2010
15. H. Ishii, K. Sugiyama, E. Ito, and K. Seki, Adv. Mater. 11, 605 (1999)
16. J. M. Nunzi, C. R. Physique, 3, 523 (2002)
17. C. Kittel, Introduction à la physique de l’état solide, Bordas, Paris, (1972)
18. C. Cojan, G.P. Agrawal and C. Flytzanis, Phys. Rev. B, 15, 909, (1977)
19. J. Lange and H. Bässler, Phys. Stat. Sol. B, 114, 561 (1982)
20. J.H. Schön, C. Kloc, A. Dodabalapur and B. Batlogg, Science, 289, 599 (2000)
21. M. Schott and C. R. Acad. Sci. Paris Sér., 4, 381 (2000)
22. D. Emin, in Handbook of Conducting Polymers, edited T.A. Skotheim, M. Dekker,
(1996), Vol. 2. p.
23. W.D. Gill, in Photoconductivity and Related Phenomena edited J. Mort, D.M. Pai,
Elsevier, (1976) p. 63.
24. R.G. Kepler, P.M. Beeson, S.J. Jacobs, R.A. Anderson, M.B. Sinclair, V.S. Valencia and
P.A. Cahill, Appl. Phys. Lett., 66, 3618 (1995)
25. G. Horowitz, Adv. Mater., 10, 365 (1998).
26. P.W.M. Blom, M.J.M. De Jong and J.J.M. Vleggaar, Appl. Phys. Lett., 68, 3308 (1996)
27. H.C.F. Martens, J.N. Huiberts and P.W.M. Blom, Appl. Phys. Lett., 77, 1852 (2000)
28. H. Scher, in Photoconductivity and Related Phenomena edited J. Mort, D.M. Pai,
Elsevier, (1976) p. 63
29. M.N. Bussac and L. Zuppiroli, Phys. Rev. B, 55, 15587 (1997)
30. J.H. Schön, C. Kloc, R.C. Haddon and B. Batlogg, Science, 288, 656 (2000)
31. J.H. Schön, S. Berg, C. Kloc and B. Batlogg, Science, 287, 1022 (2000)
32. G.G. Malliaras, Y. Shen, D.H. Dunlap, H. Murata and Z.H. Kafafi, Appl. Phys. Lett., 79,
2582 (2001)
33. A.R. Inigo, C.H. Tan, W. Fann, Y.-S. Huang, G.-Y. Perng and S.-A. Chen, Adv. Mater., 13,
504 (2001)
34. C. Sentein, C. Fiorini, A. Lorin and J.M. Nunzi, Adv. Mater., 9, 809 (1997)
35. H. Murata, G.G. Malliaras, M. Uchida, Y. Shen and Z.H. Kafafi, Chem. Phys. Lett., 339,
161 (2001)]
36. H. Spanggaard and F. Krebs, Solar Energy Materials & Solar Cells, 83, 125 (2004)
37. J.H. Burroughes, D.D.C. Bradley, A.R. Brown, R.N. Marks, K. Mackay, R.H. Friend, P.L
Burns and A.B. Holmes, Nature, 34, 539 (1990)
142 Nanocrystals

38. U. Rauscher, H. B.assler, D.D.C. Bradley and M. Hennecke, Phys. Rev. B, 42, 9830 (1990)
39. E.L. Frankevich, A.A. Lymarev, I. Sokolik, F.E. Karasz, S. Blumstengel, R.H. Baughman
and H.H. Horhold, Phys. Rev. B, 46, 9320 (1992)
40. L.J. Rothberg, M. Yan, F. Papadimitrakopolous, M.E. Galvin, E.W. Kwock and T.M.
Miller, Synth. Met., 80, 41 (1996)
41. T. Jerankoa, H. Tributscha, N.S. Sariciftcib and J.C. Hummelen, Solar Energy Materials &
Solar Cells, 83, 247 (2004)
42. Md.K.H. Bhuiyan and T. Mieno, Thin Solid Films, 441, 187 (2003)
43. N.S. Sariciftci, D. Braun, C. Zhang, V.I. Srdanov, A.J. Heeger, G. Stucky and F. Wudl,
Appl. Phys. Lett., 62, 585 (1993)
44. G. Yu, K. Pakbaz and A.J. Heeger, Appl. Phys. Lett., 64, 3422 (1994)
45. J.C. Hummelen, B.W. Knight, F. LePeq, F. Wudl, J. Yao and C.L. Wilkins, J. Org. Chem.,
60, 532 (1995)
46. G. Yu, J. Gao, J.C. Hummelen, F. Wudl and A.J. Heeger, Science, 270, 1789 (1995)
47. S.E. Shaheen, C.J. Brabec, N.S. Sariciftci, F. Padinger, T. Fromherz and J.C. Hummelen,
Appl. Phys. Lett., 78, 841 (2001)
48. U. Stalmach, B.d. Boer, C. Videlot, P.F.v. Hutten and G. Hadziioannou, J. Am. Chem.
Soc., 112, 5464 (2000)
49. E. Peeters, P.A.v. Hal, J. Knol, C.J. Brabec, N.S. Sarciftci, J.C. Hummelen and R.A.J.
Janssen, Phys. Chem. B., 104, 10174 (2000)
50. P.A. van Hal, J. Knol, B.M.W. Langeveld-Voss, S.C.J. Meskers, J.C. Hummelen and R.A.J.
Janssen, J. Phys. Chem. A., 104, 5974 (2000)
51. F.C. Krebs, M. Jørgensen, Macromolecules, 35, 7200 (2002)
52. A. Cravino and N.S. Sariciftci, J. Mater. Chem., 12, 1931 (2002)
53. T. Benincori, E. Brenna, F. Sannicol, L. Trimarco and G. Zotti, Angew. Chem., 108, 718
(1996)
54. A. Cravino, G. Zerza, M. Maggini, S. Bucella, M. Svensson, M.R. Andersson, H.
Neugebauer, C.J. Brabec and N.S. Sariciftci, Monatsh. Chem., 134, 519 (2003)
55. F.C. Krebs and M. Jørgensen, Polym. Bull., 50, 359 (2003)
56. P. V. Kamat, Nanotoday, 4, 315 (2006)
57. Y. Zhang, , and S. Iijima , Phys. Rev. Lett., 82, 3472(1999)
58. M. Freitag, Nano Lett., 4, 1063(2004)
59. Y.-Z. Ma, J. Chem. Phys., 120, 3368 (2004)
60. C. L.Kane, and E. Mele, J., Phys. Rev. Lett., 90, 207401(2003)
61. L. Sheeney-Haj-Ichia, Angew. Chem. Int. Ed. 44, 78 (2004)
62. I. Robel, Adv. Mater., 17, 2458(2005)
63. S. Hotchandani, and P. Kamat, J. Phys. Chem., 96, 6834 (1992)
64. N. Kopidakis, N. R. Neale, K. Zhu, J. van de Lagemaat, and A. J. Frank Appl. Phys. Lett.,
87, 202106 (2005)
65. P. M. Sirimanne, T. Shirata, L. Damodare, Y. Hayashi, T. Soga, T. Jimbo, Solar Energy
Materials and Solar Cells, 77, 15 (2003)
66. K. Takahashi, K. Seto, T. Yamaguchi, J.-I. Nakamura, C. Yokoe, and K. Murata, Chem.
Lett. 33, 1042 (2004)
67. T. Dentania, K. Nagasakaa, K. Funabikia, J. Jinb, T. Yoshidac, H. Minourac and M.
Matsui, Dyes and Pigments 77, Issue 1, (2008), 59-69
Diluted magnetic semiconductor nanocrystals in glass matrix 143

X6

Diluted magnetic semiconductor

nanocrystals in glass matrix
N. O. Dantas1, E. S. Freitas Neto 1 and R. S. Silva 1,2
1Laboratório de Novos Materiais Isolantes e Semicondutores (LNMIS), Instituto de Física,
Universidade Federal de Uberlândia, 38402-902, Uberlândia, Minas Gerais, Brazil
2Instituto de Ciências Exatas e Naturais e Educação (ICENE), Licenciatura em Física,

Universidade Federal do Triângulo Mineiro, 38025-180, Uberaba, Minas Gerais, Brazil

1. Introduction
Diluted magnetic semiconductor (DMS), are semiconductors in which a magnetic impurity
is intentionally introduced; a small fraction of the native atoms in the hosting non-magnetic
semiconductor material is replaced by magnetic atoms. The main characteristic of this new
class of compounds is the possibility of the onset of an exchange interaction between the
hosting electronic subsystem and electrons originating from the partially-filled d or f levels
of the introduced magnetic atom (Erwin et al., 2005; Norris et al., 2008). Once that onset is
reached in the above-mentioned exchange interaction, it enables the control of both the
electronic, and the optical properties of the end material, using external fields in regimes
hardly achieved with other classes of materials. Slightly transitions in metal-doped II-VI and
IV-VI semiconductor, as for instance, Cd1-xMnxS, Pb1-xMnxS, and Pb1-xMnxSe, is a typical
diluted magnetic semiconductor, in which a small amount of Mn2+ is substitutionally
incorporated into the hosting CdS, PbS and PbSe semiconductor crystal structure (Ji et al.,
2003; Silva et al., 2007; Dantas et al., 2008; Dantas et al., 2009).
Quantum confinement effects can be considered with the incorporating of magnetic ions in
semiconductors NCs, modifying the optical, magnetic, and electronic properties in
relationship to semiconductor bulk. The transition metal ion (Mn2+) d-electrons, usually
located in the band gap region of the hosting semiconductor, are available to promote
exchange interactions to the sp-band electrons of the hosting semiconductor (Fudyna, 1988).
The sp-d exchange interaction taking place in II-VI, and IV-VI DMS (as for instance in
Cd1-xMnxS, Pb1-xMnxS and Pb1-xMnxSe) provides a unique interplay between optical
properties and magnetism, which could be strongly-dependent upon the doping mole
fraction (x) (Silva et al., 2007; Dantas et al., 2008; Dantas et al., 2009). By varying the
material’s doping profile (x) a fine tuning of the semiconductor band gap energy can be
achieved. Furthermore, quantum size effects caused by the shrinking in DMS bulk II-VI and
IV-VI, as in nanosized particles, enhance the optical and the magnetic properties, even
further. In addition, in the presence of applied magnetic fields the sp-d interaction involving
electrons, holes, and the hosted magnetic ions is affected, as a result modifying the DMS
properties and providing the material basis for new applications in magneto-optical,
144 Nanocrystals

magneto-transport, spintronics, lasers, and infrared devices. Theoretical models to explain

the incorporation of magnetic impurities in nanocrystals are reported in the literature
(Erwin et al., 2005; Dalpian & Chelikowsky, 2006; Norris et al., 2008), such as the “self-
purification” mechanisms which are explained through energetic arguments. These
mechanisms show that the formation energy of magnectic impurities increases when the
NCs size decreases. Moreover, the binding energy of the impurities in the crystalline faces is
highly dependent on the semiconductor material, such as the crystal structure, and NCs
shape.
Nanocrysttaline structures doped with a small amount of magnetic impurities are obtained
from a controlled process known as thermal diffusion of precursor ions, for NC formation in
conditions of thermodynamic equilibrium. In this context, Cd1-xMnxS, Pb1-xMnxS, and
Pb1-xMnxSe NCs have been synthesized by fusion method in glass matrixes. In this chapter
we report the synthesis process of nanocrystals in a glass matrix, the synthesis routes of
diluted magnetic semiconductor Cd1-xMnxS, Pb1-xMnxS, and Pb1-xMnxSe nanocrystals grown
on a borosilicate glass matrix and their investigation by experimental techniques of optical
absorption (OA), photoluminescence (PL), electron paramagnetic resonance (EPR) spectra,
atomic force microscope (AFM), and x-ray diffraction (XRD).

2. Synthesis of nanocrystals in glass

Nowadays, one of the biggest interests is the nanostructured systems production which
present desired physical properties to technological applications, and that are of low cost.
Among the materials, which satisfy these necessities, there are nanocrystal-doped glasses
(Woggon, 1997; Gaponenko, 1998). These are interesting in the physical property studies of
low-dimensionally structures, and its optical transitions of electrons in quantum
confinement regime (Bányai & Koch, 1993; Woggon, 1997; Gaponenko, 1998). Since the
origin of nanocrystal-doped glasses, the optical fiber-based communication systems, which
before had the entire amplifying process and optical signal processing performed in an
electronically way, begin to amplify and process the optical signals through the use of fully
optical devices, in which the use of those devices considerably enhance the quality in signal
transmission. The first evidences of nanocrystals existence in glasses undergone through
thermal annealing were given by Rocksby at about 1930’s (Woggon, 1997). Since the second
half of 20th century, companies like Corning Glass Industries, Schott Optical Glass, Hoya,
and Toshiba, have been using quantum dot-doped glasses (Gaponenko, 1998). Although, the
semiconductor-doped glasses potential to application in optical devices still be a well
discussed subject in diverse papers, the research continues focused in the direction of a more
primary stage, where the main objective is the understanding of physics involved in this
kind of material.

2.1. Nanocrystals’ growth kinetics

The semiconductor nanocrystals’ growth kinetics in doped glass matrixes is the resulted
precipitation in a supersaturated solid solution by dopers, controlled by the diffusion
process of the solved semiconductor materials in the glass matrix (Woggon, 1997;
Gaponenko, 1998). The solid solution is defined as been constituted by a unique phase in
which more than one atomic specie is inserted, and for which the atom identity, that occupy
one or more sites, in the solution is variable (Zarzycki, 1991). In this solid solution, the
Diluted magnetic semiconductor nanocrystals in glass matrix 145

precursor elements in the glass matrix, which can move, themselves by diffusion, are
considered the solutes the glass matrixes are the solvents while the quantum dots are the
solid phase or the precipitated from the process. For an occurrence of precipitation, the
solution must be supersaturated, i. e., the solute concentration must exceed the saturation
value at a given temperature and pressure. The appearing of a new phase happens by the
discreet particles formation with well-defined, well-arranged interfaces, in an aleatory way,
within the original phase.
These growth kinetic processes, in general, can be divided in three different stages: the
nucleation, the normal growth, and the coalescence or competitive growth (Zarzycki, 1991;
Gaponenko, 1998).

2.1.1. Nucleation
In the temperature, where there is an appreciable atomic mobility, there is also a continuum
rearrangement of atoms in thermal disturbance. If the phase is thermodynamically
unstable, these rearrangement domains have a temporary existence, so then they are
destroyed, and replaced by others. When the phase is metastable, such fluctuations are
potential sources of a stable phase, and don’t become permanent. The fluctuation effects can
produce dots that are different in size, shape, structure, or composition.
In the simplest classical model, which was proposed by Volmer and Weber (Volmer &
Weber, 1926), and Becker and Döring (Becker & Döring, 2006), it is assumed that embryos of
the processes have uniform structure, composition, and identical properties to those of
future phase, and only differ in shape and size. The shape, in question, is the one that results
from the minimum free energy formation, which will be connected to the interface nature. If
it is assumed that, in first approximation, the surface energy is independent of
crystallography orientation, and the elastic deformation energy is negligible, the embryos
will have a spherical shape. The embryos’ size is dependent of the thermodynamic stability
condition. When two phases coexist in different homogeneous regions, a phase transition or
formation of a different phase within another can occur. “Nucleation”, is the process of
forming a new phase within an existing phase, separated by a well-defined surface.
In a supersaturated solid solution there is an excess of solute in the solvent (the solute
concentration exceeds the saturation value at a given temperature and pressure). This excess
can be turned into a precipitate, if the nucleation process happens. The quantum dots-doped
glasses are examples of materials created by nucleation in a supersaturated solid solution,
where there is the coexistence of phases: the solvent (glass matrix), the solute (doper), and
the precipitate (quantum dots).
Assuming that within a determined volume (matrix) a coexistence of disperse atoms
(solvent) with particles forming atoms (doper) occur, and defining gm as a free energy per
disperse atom, and gc as the free energy per crystal atom, it is obtained that the free energy
of the compound of particles varies from an amount ΔG, when the quantum dot nuclei are
formed. This variation can be given by:

 4 R 3   gc - g m 
G  G'     4 R 
2
 (1)
 3  V 
146 Nanocrystals

The term (gc-gm/V) of the Eq. (1), represents the free energy variation per unit of volume, R is
the radius of quantum dot nucleus, V is the volume per particle in the quantum dots, γ is the
surface energy per area unit.
When the matrix is supersaturated by dopers which will form the semiconductor crystalline
phase, the first term of the Eq. (1) is negative, while, the second one is positive. As the terms
are proportional to R3 and R2, respectively, it can be concluded that the second term
influence will be lower when R increases and the curve ΔG versus R will increase until it
reaches a maximum value and after it will decrease, as it is represented in Fig. 1.

Fig. 1. Free energy variation ΔG as a function of the particle radius (R) (Christian, 1965).

The position of this maximum is given by:

G `
0 (2)
R

what leads to a critical radius Rc of the quantum dot nucleus, given by:

2 V
Rc  (3)
( gc - gm )

A particle of radius Rc will be in a instable equilibrium situation. If the radius is lower than
Rc, the particle tends to be re-dissolved, once an increase in radius leads to an increase of
ΔG. If the radius is greater than Rc, the particle tends to grow, once an increase in radius
leads to decrease in ΔG. The particles with R < Rc are called “embryos”, while the ones with
R > Rc are called “nuclei”.
The free energy variation, in a transformation, also depends on quantum dot size, which are
formed in the semiconductor phase. The quantum dot radius depends on the number of
particles that are dispersed in the glass matrix, and also on equilibrium concentration for the
semiconductor phase. Therefore, from Gibss-Thomson’s equation, the free energies can be
related to semiconductor concentration in the glass matrix, as follows:

g m
- g c   KT ln  N ( R) N ( ) (4)
Diluted magnetic semiconductor nanocrystals in glass matrix 147

Here, N(R) is the equilibrium concentration for the semiconductor species in quantum dots
with radius R, N(∞) is the equilibrium concentration for the semiconductor species that are
dispersed in the glass matrix, K is the Boltzmann constant, and T is the temperature.
The critical radius (Rc) for any volume, in terms of this equation, is expressed by:

2 V
Rc  (5)
KT ln  N ( R ) N ( )

which results in:

 2 V 
N ( R )  N (  )exp   (6)
 KTRc 

According to Eq. (6), it is possible to determine the equilibrium concentration for quantum
dots of radius R. In the equilibrium, the quantum dots should not increase or decrease in
size, i. e., the absorbed species rate must be equal to released species rate. In Fig. 2 is
presented a typical curve of these concentrations.
It is observed, in Fig. 2, that the point, where the curve intercepts the line of doping
concentration existing in the matrix, and this point defines the critical radius, from which
the quantum dot nuclei will grow. It is also observed that the quantum dots are completely
re-dissolved when the temperature is increased from T2 to T3. In this case, the dissolution
rate is proportional to the difference between the equilibrium concentration and the existing
concentration in matrix.

Fig. 2. The equilibrium concentration of the atoms dispersed in the matrix as a function of
the quantum dot radius for three different temperatures (Barbosa et al., 1997).
148 Nanocrystals

Thus, the quantum dots with smaller radii will be re-dissolved much faster than those with
larger radii, which would lead to their size dispersion. It is clear that, when N(R) is below of
N(∞), there will be no growth of any quantum dot nucleus. The ratio between N(R) and N(∞)
is used as a supersaturation measure, given by:

N ( R)
 (7)
N ()

A matrix always will be supersaturated when Δ > 1. In terms of supersaturation, the critical
radius can be written as:

2 V
Rc  (8)
kT ln(  )

The supersaturation degree is also defined as follows:

[ N ( R) - N ()]
m    -1 (9)
N ( )

2.1.2. Growth
The theoretical considerations, on crystals growth description, are based on three general
models, on the type of liquid-crystal interface and the nature of active sites for
crystallization (Zarzycki, 1991): normal growth (or continuum growth); growth determined
by processes of bi-dimensional nuclei formation and subsequent increase; and coalescence
or competitive growth. For briefly, it will be considered just the first, and last mechanism.

2.1.2.1. Normal growth

With the decrease in solution supersaturation during the initial stages of nucleation, the so
called normal growth process starts. During this process, the nuclei that reached a critical
radius increases in size, while the others are re-dissolved in the matrix (Zarzycki, 1991).

2.1.2.2. Coalescence or competitive growth

When the supersaturation degree of the matrix decreases a lot, i. e., almost all
semiconductor material is already incorporated in a nucleus, the stage denominated as
coalescence or competitive growth takes place. There is a competition in which the larger
nanocrystals grow from the smaller ones. The study of this process is known as Coarsening
Theory of Lifshitz-Slyozov, and leads to a size distribution with same name (Zarzycki, 1991 ;
Gaponenko, 1998). This distribution has the peculiarity of being asymmetric around its
average values, with an sudden cut to the particles with larger size, and a huge dispersion to
the smaller ones. In practice, these different stages occur simultaneously in the real growth
process, however, it is possible to consider each stage separately for theoretical purposes.
Diluted magnetic semiconductor nanocrystals in glass matrix 149

2.2. Diluted magnetic semiconductor nanocrystals formation: magnetic impurities

incorporation
A study of Mn incorporation in semiconductors II-VI and IV-VI was performed, assuming
that the adsorption of this magnetic ion occurs on surface of the three crystallography faces
((1 1 1), (1 1 0), and (0 0 1)) of these semiconductors. The obtained results, from the density
functional theory, show that the binding energy of Mn, on surface of the crystallography
faces, is dependent of the crystalline structure of the semiconductor materials. For
crystalline structures of zinc-blend type (ZnS, CdS, ZnSe, and CdSe), the binding energy is in
the range of 2 to 7 eV. While, for the crystalline structures of rock-salt type (PbS, and PbSe),
this energy is around of 2 eV (Erwin et al., 2005).
The magnetic impurities incorporation in nanocrystals, produces changes in optical,
magnetic, and structural properties of these materials (Furdyna, 1988; Bacher et al., 2005).
The exchange interactions, between the levels sp of atoms in semiconductor nanocrystals
and the level d of Mn2+ ions, completely modify these nanocrystals properties. Most of the
semiconductor nanocrystals have a diamagnetic phase. However, with the magnetic
impurities incorporation, forming a diluted magnetic semiconductor, this material starts to
present paramagnetic, ferromagnetic, anti-ferromagnetic, or spin glass phases, even more,
they modify the lattice parameter of semiconductor nanocrystals, and Mn-Mn exchange
interactions in closer Mn ions also occur.
The electronic configuration on Mn ions introduced in diluted magnetic semiconductor is A-
(3d5) or Ao (3d5 + h (holes)) (A- is the negatively charged c, and Ao denotes the neutral
center). Studies show that there are three types of Mn centers, when it is incorporated to
semiconductor materials. The first is formed to the Manganese in Mn3+ state, which is found
in 3d4 electronic configuration with the spin in ground state S = 2, considered as a neutral
accepter center Ao (3d4). The second type of center occurs when the Manganese, in Mn3+
state, imprisons an electron and strongly bind it to the d layer, where it starts to show an
electronic configuration 3d5 with S = 5/2, denoted by A-(3d5). This second type of Mn center
become negatively charged, been able to attract and weakly bind a hole, forming a third
center , denoted by A0 (3d5 + h) (Sapega et al., 2002).

2.3. Synthesis of DMS NCs in a glass matrix

In this section, we describe the main synthesis of DMS NCs in glasses, methods which have
been being developed in recent years, by our research group. From the adequate
composition, the masses of powder compounds, that will form the glass matrix as well as
the DMS NCs, are measured, mixed, and homogenized.
The first step of sample preparation consisted of melting powder mixtures in an alumina
crucible at high temperature for a determined time. In the sequence, a quick cooling to room
temperature was undergone to the crucible containing the melted mixture. In the second
step, was carried out a thermal annealing of the previously melted glass matrix at specific
temperature for several hours aiming to enhance the diffusion of precursor ions into the
host matrix.
The DMS NCs, which were synthesized by this methodology, growth kinetics can be
explained based on the described models in sections 2.1 and 2.2. The optical, magnetic, and
structural properties of these DMS NCs will be presented in the following sections.
150 Nanocrystals

2.3.1. Synthesis of Pb1-xMnxS NCs

Pb1−xMnxS NCs embedded in an oxide glass matrix were synthesized by the fusion method.
The synthesis process proceeds as follows. First, the Mn-doped SiO2–Na2CO3–Al2O3–PbO2–
B2O3+S (wt %) powder was melted in an alumina crucible at 1200°C for 30 min. Then, it was
cooled down to room temperature. After that, thermal annealing treatment proceeded at
500°C. Finally, spherically shaped Pb1−xMnxS NCs were formed in the glass matrix. In order
to study the effects of the synthetic process on the magnetic properties of DMS NCs, four
Pb1−xMnxS samples with x-concentration varying from 0% until 40% denominated as
SNABP: Pb1-xMnxS, have been synthesized under different thermal treatments, with
annealing times of 2, 4, 8, and 10 h, respectively (Silva et al., 2007).

2.3.2. Synthesis of Pb1-xMnxSe NCs

The semimagnetic Pb1−xMnxSe samples were synthesized in glass matrix SNABP with
nominal compositions 40SiO2 · 30Na2CO3 · 1Al2O3 · 25B2O3 · 4PbO (mol %) adding 2Se
(wt %), in which the incorporation of Mn2+ ions varies between 0 < x < 5%. The preparation
process consisted of melting the powder mixtures in an alumina crucible at 1200°C for 30
min. In the sequence, the crucible containing the melted mixture underwent quick cooling to
room temperature. In a second step, thermal annealing of this melted and cooled glass
matrix was carried out at 500°C for several hours, in order to enhance the diffusion of Pb2+,
Mn2+, and Se2− species into the host matrix, or produce a rearrangement of ions entering the
formation of the NCs. This annealing process produces good quality Pb1−xMnxSe NCs
showing small size distribution of dots. The same process holds for undoped PbSe quantum
dot formation. Two types of Mn-doped samples were grown: (i) SNABP matrix only doped
with x-content of Mn, and labeled SNABP: xMn; (ii) SNABP templates, labeled SNABP:
Pb1−xMnxSe containing Pb1−xMnxSe NCs with the percentage of manganese-to-lead in the
range 0 < x <5% (Dantas et al., 2009).

2.3.3. Synthesis of Cd1-xMnxS NCs

Cd1−xMnxS NCs were synthesized in a glass matrix SNAB with a nominal composition of
40SiO2 · 30Na2CO3 · 1Al2O3 · 29B2O3 (mol %) adding 2(CdO+S) (wt %), and Mn-doping
concentration (x) varying with respect to Cd-content from 0% to 10%. The first step of
sample preparation consisted of melting powder mixtures in an alumina crucible at 1200°C
for 30 min. Then, the crucible containing the melted mixture underwent quick cooling to
room temperature. In the second step, thermal annealing of the previously melted glass
matrix was carried out at 560°C for 10 hours in order to enhance the diffusion of Cd2+, Mn2+,
and S2− species into the host matrix. As a result of the thermal annealing, Cd1−xMnxS NCs
were formed in the glass template. There were two classes of samples for different Mn
concentrations: (i) doped with Mn and denominated SNAB: xMn and (ii) NC samples
Cd1−xMnxS NCs were denominated SNAB: Cd1−xMnxS (Dantas et al., 2008).

3. Optical Properties of DMS NCs

The optical properties with the Mn2+ ions incorporation in Pb1-xMnxS, Pb1-xMnxSe, and
Cd1-xMnxS NCs, which were grown in glass matrixes through the methodologies described
in section 2.3, were investigated by Optical Absorption (OA), and/or Photoluminescence
spectroscopy techniques. The obtained results will be presented and discussed as follows.
Diluted magnetic semiconductor nanocrystals in glass matrix 151

3.1. Optical Absorption (OA) of Pb1-xMnxS NCs

The optical absorption spectra of samples SNABP: Pb1-xMnxS (x = 0, 0.003, 0.005, and 0.010)
are shown in Fig. 3. Strong blue shift with respect to the optical absorption of bulk PbS
(band gap at 0.28 eV) is clearly observed (see Fig. 3) in all SNABP: Pb1-xMnxS samples,
indicating the quantum confinement effect of carriers within the as-produced PbS NCs.

Fig. 3. Room temperature optical absorption spectra of SNABP: Pb1-xMnxS samples. (Silva,
2008)

Furthermore, with the introduction of the magnetic impurity (Mn2+) in the PbS NC lattice
the optical properties are completely modified due the exchange interaction (sp-d) between
the electronic subsystem of the PbS NC and electrons originated from the partially-filled
Mn2+ ions (Lee et al., 2005). This exchange interaction, scaling with the x-content, responds
for the relative blue shift of the effective band gap observed in the OA spectra of samples
SNABP: Pb1-xMnxS, as shown in Fig. 3. More specifically, effective band gaps of 0.95 eV
(1307 nm), 1.00 eV (1238 nm), 1.07 eV (1158 nm) and 1.12 eV (1111 nm) were observed for x-
contents of 0, 0.003, 0.005 and 0.010, respectively.

3.2. Optical Absorption (OA) of Pb1-xMnxSe NCs

Optical absorption spectra recorded for different samples have provide strong evidences of
Mn2+ ion incorporation into the PbSe NCs samples, labeled as SNABP: Pb1-xMnxSe (x>0).
Due to the exchange interaction (sp-d hybridization) between electronic subsystems, the
incorporation of magnetic ion in NCs modify the confined electronic states and thus, the
optical properties of the quantum dots (Furdyna, 1988; Ohno, 1998). This exchange
152 Nanocrystals

interactions causes the blue shift of optical resonance, proportional to the Mn-concentration
x, between Pb1-xMnxSe and PbSe NCs.
Figure 4 shows this effect for SNABP: Pb1-xMnxSe samples embedded with Pb1-xMnxSe with
x = 0, 0.005, 0.01 and 0.05%. The observed NC blue shift changes from 0.84 eV (1476 nm) for
x=0 to 0.89 eV (1398 nm) for x = 0.05%. Effects associated to the spatial confinement can be
estimate since bulk semiconductor lead-salt Pb1-xMnxSe samples have rock salt crystal
structure with a direct band-gap, at the L-point of the Brillouin zone, with a value ranging
between 0.28 eV, for bulk PbSe, and 3.4 eV, for the bulk MnSe which displays hexagonal
structure. The appearance of well defined subband peaks in the absorption spectra
demonstrates a relatively small size distribution and good quality of these SNABP:
Pb1-xMnxSe samples synthesized by fusion method. The absorption peak observed near
570 nm, for SNABP: xMn sample, is attributed to the presence of 0.05% of Mn2+ ion.

Fig. 4. Room temperature optical absorption spectrum as a function of the wavelength: for a
(a) glass matrix; (b) for glass matrix doped with x = 0.05% Mn (SNABP: xMn) and
Pb1-xMnxSe NCs (SNABP: Pb1-xMnxSe) for Mn-concentration: (c) x = 0; (d) x = 0.005%; (e)
x = 0.01%; and (f) x = 0.05%. (Dantas et al., 2009)

Atomic force microscopy (AFM) images of these samples (shown in section 5.2) confirm the
same average size for both the PbSe NCs and Pb1-xMnxSe NCs. Then, this blue shift in the
OA spectra (Fig. 4), between PbSe and Pb0.95Mn0.05Se NCs, was associated to the
incorporation of Mn2+ ions in the PbSe dot structure.
Diluted magnetic semiconductor nanocrystals in glass matrix 153

3.3. Optical Absorption (OA) of Cd1-xMnxS NCs

The OA spectra of the Cd1-xMnxS NC samples, that were synthesized as described in section
2.3.3, were obtained using a spectrophotometer Varian-500 operating between 175–3300 nm.
OA spectra provided other evidence of Mn2+ ion incorporation in the SNAB: Cd1−xMnxS and
SNAB: xMn samples.(Wang et al., 2004; Levyayb et al., 1998) The introduction of magnetic
impurities in semiconductors modified NCs’ optical properties as a result of sp-d exchange
interactions between electrons confined in dot states and located in the partially filled Mn2+
states, causing a blue shift with increasing x in the band gap of Cd1−xMnxS samples
compared to undoped CdS NCs.

Fig. 5. OA spectra of Cd1−xMnxS NCs embedded in a glass matrix for samples with
concentrations: x = 0, 0.005, 0.01, 0.05, and 0.10. The blue shift is marked by lines and arrows.
(Dantas et al., 2008)

Figure 5 shows this effect in embedded Cd1−xMnxS NCs samples, for x=0, 0.005, 0.01, 0.05,
and 0.10. Note the blue shift of band gap varying from 3.07 eV (403 nm) to 3.22 eV (385 nm).
The band gap of Cd1−xMnxS semiconductor varies between 2.58 eV (CdS bulk) and 3.5 eV
(MnS bulk). The appearance of well defined subband peaks in the absorption spectrum
demonstrates the high quality of the synthesized samples and the relatively small
distribution of the NCs.
154 Nanocrystals

Using a simple confinement model based on effective-mass approximation (Brus, 1984), the
energy of the lowest exciton state in the microcrystallites of radius R smaller than the
exciton Bohr radius aB can be estimated by the following expression:
Econf  Eg   h 2 2 2  R 2  - 1.8  e 2  R  , where Eg is the energy gap of material (bulk),  the reduced
effective mass; e the elementary charge,  the dielectric constant, and the estimated
average radius for CdS NCs was R ~ 2.2 nm.
As all samples were subjected to the same thermal annealing at 560ºC for 10 hours, is
expected that the Cd1-xMnxS NCs have the same size as the corresponding CdS NCs, in
agreement with AFM data which will be shown in section 5.3. Therefore, the OA resonance
blue shift that was observed in Fig. 5, which increases with increasing Mn2+ concentration,
occurred due to the incorporation of Mn2+ ions into CdS NCs.

3.4. Photoluminescence (PL) of Cd1-xMnxS NCs

Through the OA spectra, shown in Fig. 5 (section 3.3), we have confirmed that there was the
formation of Cd1-xMnxS NCs with quantum confinement properties as well as bulk-like
properties. The formation of these NCs can also occur, even in the quick cooling, for
unannealed samples, which we shall call of Cd1-xMnS NCs as-grown samples. The room
temperature PL spectra of the Cd1-xMnxS NCs as-grown samples were acquired with a
spectrofluorometer (NanoLog – Horiba JY).
Figure 6(a) shows the PL spectra at room temperature of Cd1-xMnxS NCs as-grown samples
with different nominal x-concentrations: 0, 0.005, 0.050, and 0.100. The samples were excited
at the absorption band edge of the NCs with bulk-like properties (510 nm), which implies
that the electrons are occupying the bottom of conduction band (CB).
Our results clearly indicate that nonradiative decay paths are present, alongside the
radiative recombination of the electron-hole pairs (Ee-h), which implies that the energy levels
Mn2+ (4T1), trap (1), and trap (2) are occupied by electrons. The recombination aspects are
well described in the diagram depicted in Fig. 6(b) which shows the emission I between the
levels 4T1 and 6A1, characteristic of the d orbital of Mn2+ ion when it is substitutionally
incorporated in semiconductors II-VI (Zhou et al., 2006; Beaulac et al., 2008; Archer et al.,
2007). In these materials the Mn2+ ions can be substitutionally incorporated in two distinct
sites: one in the NC-core (labeled as SI), and other near the NC-surface (labeled as SII).
The traps (1) and (2) are deep defect levels attributed to the CdS (Smyntyna et al., 2007),
whose origins are not quite clear until the moment. Despite of its not so clear origin, we
assume that, these two defect levels are possibility related to the VCd-VS divacancy centers
with different orientations, in analogy to the studies performed for CdSe NCs with
hexagonal wurtzite structures (Babentsov et al., 2005), a material that presents great
similarities with ours. It was shown that there are two energetically different divacancies: an
oriented along the hexagonal c-axis (assigned to trap (1)), and other oriented along the basal
Cd-Se bond directions (assigned to trap (2)). This divacancy model could also be used to
explain the origins of the two deep traps (1) and (2) in CdS NCs with hexagonal wurtzite
structures as well as DMS Cd1-xMnxS NCs. These considerations are quite reasonable since
the wurtzite structure is a common phase for CdS NCs grown in a glass matrix (Cheng et al.,
2006; Xue et al., 2009) as well as for Cd1-xMnxS NCs (0 < x ≤ 0.500) (Jain, 1991).
Diluted magnetic semiconductor nanocrystals in glass matrix 155

Fig. 6. (a) Room temperature PL spectra of as-grown Cd1-xMnxS NCs embedded in the glass
matrix SNAB for samples with x = 0, 0.005, 0.050, and 0.100. (b) Schematic diagram showing the
radiative recombinations I, E1, and E2 by the straightened arrows. The nonradiative transitions
from the level 4T1 to trap-levels (1) and (2) are represented by the wavy arrows. (c) Relative
intensities between the emissions I and E1 (solid line), and the emissions I and E2 (dashed line).

From these two traps, there are the emissions E1 and E2 which can be observed in Fig. 6(a)
for all samples. The PL spectra of CdS NCs (x = 0.000) were fitted using three like-gaussian
components associated to these emissions (E1 and E2) as well as the radiative recombination
of the electron-hole pairs (Ee-h). It is clear that the emissions E1 and E2 are more intense than
the emission Ee-h, showing that the nonradiative processes are dominant from the CB bottom
to trap-levels (1) and (2). Figure 6(a) also shows that an overlap, between the emissions Ee-h
and I for the Cd1-xMnxS NCs samples with x = 0.005, 0.050, and 0.100, takes place.
Undoubtedly, this overlapped band is more intense than the band of the emission Ee-h
observed for CdS NCs samples, which indicates that majority contribution of the
overlapped band can be attributed to emission of M2+ ion (4T1 – 6A1).
Related to I emission was observed that for Cd1-xMnxS NCs grown in a colloidal solution, it
is suppressed due to nonradiative processes when the Mn2+ ions are incorporated in the site
SII (Zhou et al., 2006). Thus, the strong emission I, observed in the PL spectra of the Cd1-
xMnxS NCs samples (see Fig. 6(a)), gives evidence that the Mn2+ ions are substitutionally
located in the core of nanoparticles, i.e., in the site SI. Figure 6(c) shows the behavior of
relative intensity between the emissions I and E1 (I/E1), and the emissions I and E2 (I/E2), for
different x-concentration. The enhancement in relative intensity I/E1 shows that the
156 Nanocrystals

increasing in x-concentration favors the incorporation of Mn2+ ions in the NC-core (SI). On
the other hand, the I/E2 relative intensity curve presents a completely different shape, where
a decrease in the I/E2 is followed by an increase with the augment of x-concentration. Such
kind of behavior can be understood taking into account nonradiative processes, in order that
electron transfers occur from the level 4T1 to the trap-levels (1) and (2), as schematically
depicted in the Fig. 6(b) by the wavy arrows. The values I/E2 are larger than I/E1 for x =
0.005 and 0.100, indicating that, there is a smaller energy transfers (via electrons) to trap (2)
than to trap (1) for these two x-concentrations. While for x = 0.050, the energy transfers from
the level 4T1 to the trap-levels (1) and (2) are practically equals. These results indicate that
possibly there is a competition between the electron transfers for the trap-levels dependent
of the x-concentration.

4. Magnetic properties of DMS NCs

The physical properties of semiconductor nanocrystals change with the incorporation of
magnetic impurities, making them diluted magnetic semiconductors. This phenomenon
occurs due to the sp-d exchange interaction, which involves Mn ions and electrons in
conduction band or holes in the valence band, when there are Mn-doped semiconductors II-
VI or IV-VI. This exchange interaction causes changes in the magnetic properties of the
nanocrystals, influencing the sp-d exchange interactions of these materials.

4.1. Electron Paramagnetic Resonance (EPR)

A study of the Manganese’s magnetic properties in host matrixes can be performed from the
EPR technique. The studies by this technique, to the obtained results, confirm that the
oxidation state of Mn is 2+, in which the Mn2+ ions belong to the group 3d. The free ions of
this group present, in their ground configuration, the 3d incomplete layer, which is
responsible for the paramagnetism. In the presence of a crystalline lattice, the Mn2+ ions start
to have the splitted by the crystalline field. This splitting produces a reduction in the orbital
movement contribution of the magnetic moment, been the magnetism from these ions
fundamentally attributed to the electronic spin (Silva, 2008).
The crystal structure of the hosting semiconductor has a strong influence into the
incorporated Mn2+-ions, as observed in the EPR spectrum recorded from the glass sample
embedded with Pb1-xMnxS, Pb1-xMnxSe, and Cd1-xMnxS NCs, favouring a strong electron
spin-nucleus spin interaction. The experimental spectra of Mn-doped in NC samples can be
modelled by the spin Hamiltonian (Dantas et al., 2008):

 
Hˆ   eSˆ  g e  B  D Sz2 - S  S  1  / 3   E Sx2 - Sy2  ASˆ  Iˆ , (10)


where the first term is represent the Zeeman interaction with  e , ge, and B being the Bohr
magneton, the Lande factor and the applied magnetic field, respectively., in which the
second and third two terms describe the zero-magnetic field fine-structure splitting due to
spin-spin interaction of electrons, which is nonzero only in environments with symmetries
lower than cubic, and the fourth term ( ASˆ  Iˆ ) is stemmed from the hyperfine interaction
between electron and nuclear spins. In a magnetic field the spin degeneracy of Mn2+-ions
Diluted magnetic semiconductor nanocrystals in glass matrix 157

will be lifted by the Zeeman splitting, resulting in six energy levels classified by magnetic
electron spin quantum number MS. Due to hyperfine splitting, each of these transitions will
be split into six hyperfine levels characterized by the magnetic nuclear spin quantum
number MI. The main hyperfine lines in the spectra are due to allowed MS =  1 transitions
with ΔMI = 0, whereas other lines from forbidden transitions (due to breakdown of selection
rule) with nonzero ΔMI may also be observed. Hence, the typical EPR of Mn2+ with electron
(S = 5/2) and nuclear (I = 5/2) spins is composed of 30 lines – five fine structure transitions,
each splitting into six hyperfine lines. The incorporation of metal-transition in NC sites
cause changes in coordination states modifies the crystal field. These changes are analyses
by EPR spectra. Typical EPR of Mn2+ with electron (S = 5/2) and nuclear (I = 5/2) spins is
composed of 30 lines – five fine structure transitions, each splitting into six hyperfine lines.
For the Mn-doped in PbS, PbSe and CdS NCs the hyperfine lines were only observed for the
central MS = 1/2 ↔ -1/2 transition. The location of Mn2+ is presented in two different sites
(surface and core) in NCs can be determinate by EPR measurements. In addition, the
interaction constants A, D, and E depend strongly on the characteristics of the crystal field.
For instance, when a Mn2+ ion is located close to or on the NC surface, a large structural
difference between the NC and the glass matrix results in a larger hyperfine constant A and
larger D and E values. Hence the EPR spectrum varies when the local structure of Mn2+ ion
in the NC changes (Silva et al., 2007; Dantas et al., 2008; Dantas et al, 2009).

4.1.1. EPR of Pb1-xMnxS NCs

Figure 7 shows the EPR spectra for different Mn-concentrations in Pb1-xMnxS NCs, to the
samples that were annealed at 500ºC for 10 hours. The six hyperfine transitions were
observed due to the electron spin-nucleus spin interaction. The lines width comes from the
sum of two contributories, ΔH = ΔHI + ΔHD , where ΔHI is the width which appears in the
spectra cause intermolecular processes, and ΔHD is the width related to “spin-spin”
interaction between the first neighbours of Mn (Hinckley & Morgan, 1965). These “spin-
spin” interactions are proportional to r-3, where r is the average distance between the
Manganese atoms (Mn-Mn). To bigger distances than 55 Å (r > 55 Å) the lines are narrow
and exclusively determined by ΔHI. In this case, the EPR spectra of Mn2+ ion are sixtet
degenerated spin what assures the presence of six hyperfine interaction lines. On the other
hand, to smaller distances than 9 Å (r < 9 Å), the six hyperfine interaction lines become
wider and due to the hyperfine structure collapse, the spectra appear as a single line as
observed for samples with Mn-concentration of x = 0.40. The arrow-indicated structures
which appear around 345 mT, are caused by different Mn2+ ions localizations in PbS
quantum dots (core or surface).
158 Nanocrystals

Fig. 7. EPR spectra of Pb1-xMnxS NCs for x = 0.001, 0.003, 0.005, 0.007, 0.01, 0.05, and 0.40.
(Silva, 2008)

Figure 8(a) shows the EPR spectra for the SNABP: PbS, SNABP: 0.005Mn, and SNABP:
Pb0.995Mn0.005S samples, that were annealed at 500ºC for 10 hours. It can be perceived that the
hyperfine interactions are more clearly observed to SNABP: Pb0.995Mn0.005S samples. These
interactions, caused by the Mn2+ ion presence in the crystalline field from PbS quantum
dots, produce a splitting in this fine structure (electronic transition +1/2 ↔ -1/2) into six
hyperfine transitions due to electron spin-nucleus spin interaction.

o
500 C/10 h SNABP: Pb0.995Mn0.005S SNABP: Pb0.995Mn0.005S
SNABP: Pb0.995Mn0.005S 5.2
SII
EPR Intensity Change (a. u.)

5.0
4.8 H

4.6
4.4
EPR Intensity (a. u.)
EPR Intensity (a. u.)

EPR Intensity (a. u.)

4.2
4.0
SNABP: 0.005 Mn 3.8
3.6
Exp
2 4 6 8 10
Annealing Time (hours)

SNABP: PbS
o
500 C by:
Cal
2 hours
4 hours SI
8 hours
(a)
10 hours
(c)
(b)
310 320 330 340 350 360 370 380
Magnetic Field (mT) 310 320 330 340 350 360 370 380 310 320 330 340 350 360 370 380
Magnetic Field (mT) Magnetic Field (mT)
Fig. 8. (a) EPR spectra for the SNABP: PbS, SNABP: 0.005Mn e SNABP: Pb0.995Mn0.005S
samples (Silva, 2008). (b) Changing in the EPR intensity signal of the SNABP: Pb0.995Mn0.005S
sample that were annealed at 500ºC by increasing times [modified from (Silva et al., 2007)].
(c) Experimental and calculated EPR spectra of Pb0.995Mn0.005S NCs [modified from (Silva et
al., 2007)].
Diluted magnetic semiconductor nanocrystals in glass matrix 159

The EPR spectra intensity modifies with the increasing in the annealing time at 500ºC, as
shown in Fig. 8(b). This increasing in intensity of EPR signal is caused by a greater
incorporation of Mn2+ ions, in PbS NCs which increase their density in the glass matrix, as
observed by Atomic Force Microscopy images in Fig. 12.
For Pb1-xMnxS NCs that were grown in the glass matrix SNABP, the EPR spectra are caused
by the contributions of Mn2+ ions incorporated in the core (Signal SI), and/or on the surface
(Signal SII) of the PbS NCs. The signals SI and SII simulations was performed using the
WINEPR and SINFONIA Brucker’s softwares, being the resulting simulated signal
compared with the obtained EPR spectrum to the SNABP: Pb0,995Mn0,005S sample and shown
in Fig. 8(c). It was observed the signal SII predominance over the signal SI, giving strong
evidences of a higher Mn2+-concentration near to the NC-surface. The average values
obtained to hyperfine constants (A) are of ASI = 8.20 mT and ASII = 9.37 mT, with the electron
g-factor equal to 2.005 (Silva et al., 2007). The different hyperfine constant values between
the Pb1-xMnxS quantum dots and the respective bulk material, are attributed to quantum
confinement of the electrons, promoting a higher interaction between the electron spin and
the Mn2+ ion nucleus (Ji et al., 2003).

4.1.2. EPR of Pb1-xMnxSe NCs

A high sensitivity Bruker ESP-300 spectrometer, operating in the X-band microwave
frequency (9.5 GHz), with swept static field and the usual modulation and phase sensitive
detection techniques, was used to record the EPR spectra of the Pb1-xMnxSe NCs samples
that were synthesized by methods described in section 2.3.

Fig. 9. Left panel: EPR spectra for pure PbSe and for semimagnetic Pb1−xMnxSe NCs, with
x = 0.005, x = 0.01, and x = 0.05. In each panel, lines show the difference between undoped
and doped PbSe samples. Right panel: EPR spectra of Pb0.95Mn0.05Se NCs measured in the X-
band and at room temperature and the computed EPR spectra obtained by a summation of
two spectra with A = 7.3 mT and 7.9 mT, corresponding to Mn2+ sites inside labeled as SI
and on the surfaces labeled as SII of NCs, for a system with S = 5/2, I = 5/2, D = 30 mT,
E = 4 mT, and g = 2.005. (Dantas et al., 2009)
160 Nanocrystals

Figure 9(a) shows EPR spectra of a set of Pb1−xMnxSe NCs. The broad background line, in all
spectra, indicates the magnetic interaction between Mn2+ ions inside the Pb1−xMnxSe
structure.
The presence of the six hyperfine lines, as shown in the inset, confirms the uniform
incorporation of Mn2+ ions into the dot. Figure 9(b) shows fairly good agreement between
experimental (theoretical) spectrum of Pb0.95Mn0.05Se NCs, measured in the X-band and at
room temperature and shown as solid blue (dotted green) line. This EPR simulation is
composed of two spectra: i) A broad line that uses a fine structure constant,
A = 7.3 mT corresponding to Mn2+ ions inside the NC, ii) a stronger fine structure constant
A = 7.9 mT corresponding to ions located close to the surface. The hyperfine parameters are
D = 30 mT, E = 4 mT, and the Mn parameters are SMn = 5/2, I = 5/2, and gMn = 2.005.
Analysis of the EPR spectra (shown in Fig. 9) provided further evidence for the presence or
absence of doped PbSe dots. The incorporation of Mn2+ ions in Pb1−xMnxSe NCs is confirmed
by the presence of the six central hyperfine lines in the EPR spectrum. These hyperfine lines
are not present in glass samples only doped with Mn2+ ions.

4.1.3. EPR of Cd1-xMnxS NCs

The room-temperature modifications in the electronic states induced by Mn2+ ion
incorporation into CdS NCs (forming Cd1-xMnxS NCs) were examined by EPR using a high
sensitivity Bruker ESP-300 spectrometer operating in the X-band microwave frequency (9.5
GHz).
Figure 10(a) shows EPR spectra for the synthesized sample set. The broader background
signal observed in the EPR spectra is due to the spin-spin interaction between Mn2+
electrons when they are incorporated into the Cd1−xMnxS NCs.

Fig. 10. Panel (a) room temperature EPR spectra of Cd1−xMnxS NCs for samples with
concentration: x = 0, 0.005, 0.01, 0.05, and 0.10. Panels (b) and (c) EPR spectra of glass matrix
doped with xMn and Cd1−xMnxS NCs for x=0.05 and 0.10, respectively. (Dantas et al., 2008)
Diluted magnetic semiconductor nanocrystals in glass matrix 161

The six line structure confirms the uniform incorporation of Mn2+ ions into the host CdS NC
structure, whereas the broader EPR line indicates the presence of magnetic exchange
interaction when two Mn2+ ions get close enough and are found in small agglomerate
islands. The difference between the glass samples (SNAB: xMn and SNAB: Cd1−xMnxS NCs)
is observed in the EPR spectra [Figs. 10(b) and 10(c)]. Note that the hyperfine interaction due
to Mn2+ ions results from the presence of a crystalline field in CdS NCs. The hyperfine
interaction observed in the spectrum of glass samples doped with Mn2+ is not as evident as
those observed for Cd1−xMnxS NCs. This difference could possibly be attributed to the
formation of small islands of crystalline phase MnO. (Mukherjee et al., 2006)
It has already been reported that Mn2+ ions are incorporated into two distinct sites of NCs;
when found at the dot core, it produces the EPR signal SI, when located on or near the NCs
surface it produces the EPR signal SII. (Silva et al., 2007; Zhou et al., 2006) This analysis is
strongly supported by good agreement between the experimental EPR spectrum and the
simulated one. In Fig. 11, the EPR spectrum of Cd0.95Mn0.05S NCs, measured in the X-band
and at room temperature, is shown as a solid blue line and the calculated one as a solid
green line.

Fig. 11. Experimental and simulated EPR spectra of Cd0.95Mn0.05S NCs. The simulation uses
superposition of two spectra with A = 7.6 mT and 8.2 mT, corresponding to Mn2+ ions
located inside (labeled as SI) and on the surfaces (labeled as SII) of NCs for a system with
parameters: S = 5/ 2, I = 5/ 2, D = 40 mT, E = 5 mT, and ge = 2.005. (Dantas et al., 2008)

The EPR simulation was obtained by a sum of two spectra with A = 7.6 and 8.2 mT,
corresponding to Mn2+ ions located inside the NC (labeled as SI) and near the NC surface
(labeled as SII) for a dot system with parameters: S = 5/ 2, I = 5/ 2, D = 40 mT, E = 5 mT, and
ge = 2.005. On the other hand, as the annealing time increases, the probability of finding
162 Nanocrystals

magnetic ions inside NCs occupying neighboring lattice sites as well as the number of
antiferromagnetic spin correlated clusters increases. This combination of effects enhances
the dipolar interaction and increases the lattice distortions on the Mn2+ sites. Furthermore,
the accumulation of Mn2+ ions on the NCs surfaces also strengthens Mn–Mn interactions.
(Silva et al., 2007; Zhou et al., 2006; Jian et al., 2003) As a consequence, the intensity of the
broader background peak is increased.
EPR spectra provided evidence that Mn2+ ions are incorporated at two distinct sites: at the
core and/or at the surface of CdS NCs. Influences of CdS NCs crystalline field and the
presence of MnO cluster phases could be confirmed by the presence of six hyperfine lines
assigned to the Mn2+ ions in the samples.

5. Structural properties of DMS NCs

The structural properties of the Pb1-xMnxS, Pb1-xMnxSe, and Cd1-xMnxS nanocrystals that
were grown in glass matrixes, through the methodologies described in section 2.3, their
spatial distribution, size homogeneity, and the nanocrystal shapes were analyzed by Atomic
Force Microscopy using the contact mode. It was also employed the X-Ray Diffraction
technique to investigate the crystalline structure of the Pb1-xMnxSe NCs samples.

5.1. Atomic Force Microscopy (AFM) of Pb1-xMnxS NCs

In the Fig. 12 are shown atomic force microscopy (AFM) images that were obtained for the
SNABP: Pb0.995Mn0.005S NCs samples, which were annealed at 500ºC for 2, 4, 8, and 10 hours.
It can be observed that with the increase in annealing time, the Pb0.995Mn 0.005S NCs density
increases as well as their average size. The obtained values were 4.2 nm (2 hours), 4.3 nm (4
hours), 4.4 nm (8 hours), and 4.8 mn (10 hours).

Fig. 12. AFM images of SNABP: Pb0.995Mn0.005S NCs samples that were annealed at 500ºC for
(a) 2, (b) 4, (c) 8 and (10) hours (Silva, 2008).
Diluted magnetic semiconductor nanocrystals in glass matrix 163

The increase in the Pb0.995Mn 0.005S NCs density within host glass matrix results in the
intensity increase of optical absorption spectra (see Fig. 3), as well as, the increase in EPR
signal intensity (Fig. 8(b)).

5.2. Atomic Force Microscopy (AFM) of Pb1-xMnxSe NCs

To confirm the NCs formation on glass templates we have taken AFM images, as shown in
Fig. 13, for (a) PbSe NCs and (b) Pb0.95Mn0.05Se NCs samples. The 3D and the 2D (inset)
morphologies of Pb1-xMnxSe NCs can be noted. As shown in Figures 13(a) and 13(b), is
possible to estimate the average size for PbSe and Pb1-xMnxSe NCs as 5.2 nm, and with 6%
the size distribution, from these AFM images.

Fig. 13. AFM images of NCs showing size distribution below 6% for: (a) PbSe (b)
Pb0.95Mn0.05Se. AFM image in 3D e 2D (inset) illustrates the morphology of Pb1-xMnxSe NCs,
for x = 0 and 0.05. (Dantas et al., 2009)

These AFM data have confirmed that both PbSe NCs and Pb1-xMnxSe NCs have the same
average size (5.2 nm). Thus, undoubtedly, the observed blue shift in OA spectra of these
samples, shown in Fig. 4 (section 3.2), is related to incorporation of Mn2+ ions into PbSe NCs.

5.3. Atomic Force Microscopy (AFM) of Cd1-xMnxS NCs

Figure 14 shows the AFM images that were recorded for glass samples embedded with CdS
NCs [Fig. 14(a)] and Cd0.95Mn0.05S NCs [Fig. 14(b)] in order to confirm both the NCs
formation and OA data for size distribution, shown in Fig. 5 (section 3.3).
164 Nanocrystals

The average NCs size estimated from these AFM images is at around R ~ 2.3 nm with
corresponding size distributions of 5% and 9%, respectively. The AFM morphology of
isolated NCs is shown in two-dimension (2D) and three-dimensions (3D) images. Thus, is
observed that all samples are single-phase materials with hexagonal wurtzite structure,
which is a very common phase for CdS NCs grown in glass matrix (Cheng et al., 2006; Xue
et al., 2009), as well as for Cd1-xMnxS NCs (0 < x ≤ 0.500) (Jain, 1991).

Fig. 14. AFM images showing nearly 5% size distribution and dot size R ~ 2.3 nm. The 2D
and 3D images illustrate the morphology of (a) CdS NCs and (b) Cd0.95Mn0.05S NCs. (Dantas
et al., 2008)

As it was expected, the AFM data confirm that both CdS NCs and Cd1-xMnxS NCs have the
same average size. Thus, it was proved that the incorporation of Mn2+ ions into CdS NCs
provokes a blue shift in OA spectra of nanoparticles, according to the results shown in
Fig. 5.
Diluted magnetic semiconductor nanocrystals in glass matrix 165

6. X-Ray Diffraction (XRD) of Pb1-xMnxSe NCs

The XRD patterns of the SNABP: Pb1-xMnxSe samples (x ≥ 0) for x = 0 and x = 0.05% are
shown in Fig. 15(a). It is noted that the typical bulk PbSe rock salt crystal structure is
preserved for the Pb1-xMnxSe dot samples having Mn-concentration x < 0.05%. Nevertheless,
the characteristic XRD peaks is shifted towards lower diffraction angle values as the Mn2+
incorporation in the hosting PbSe structure increases, as shown in Fig. 15(b), and this is a
clear indication that a decrease in the lattice constant is occurring. We have estimate the
lattice constant (Cohen Method) of the structure using the (111), (200), and (220) peaks of the
XRD spectrum. Here, the average lattice crystal constant found for Pb1-xMnxSe samples is
6.130 Ǻ for x = 0% (PbSe) and 6.127 Ǻ for 0.05% incorporation of Mn2+ into the NCs. This
monotonic decrease observed in the lattice constant can be attributed to the replacement of
Pb2+-ions, having larger ionic radius (119 pm) in the rock salt PbSe crystal structure, by
Mn2+-ions with smaller ionic radius (83 pm).

Fig. 15. (a) XRD measurements of PbSe (lower line) and Pb0.95Mn0.05Se (top line) NCs
embedded in glass matrix. (b) The effects associated to the incorporation of Mn2+ ions into
PbSe NCs are seen as an intensity increase (left panel) and a shift to higher 2θ diffraction
angle (right panel) of the (111) peak. Probably, the diffusion is enhanced along the (111)
direction. (Dantas et al., 2009)

Once the grown samples have very small Mn-concentration, the crystalline structures for
undoped and doped NCs remained rock salt and showed similar values for the lattice
constants. However, it is expect a decrease in the crystalline quality of samples containing
higher concentration of Mn, with the possible occurrence of MnSe clusters inside the
Pb1-xMnxSe NCs.
166 Nanocrystals

7. Conclusions
In this chapter, it was presented the main advances, which have been obtained by us in last
few years, related to study of Diluted Magnetic Semiconductor Nanocrystals (DMS NCs) in
glass matrixes.
Probably the first time, Pb1-xMnxS, Pb1-xMnxSe, and Cd1-xMnxS DMS NCs were successfully
grown in a glass matrix by the fusion method, when subjected to an adequate thermal
annealing.
Our results, obtained by the experimental techniques, showed that it was possible to control
the optical, magnetic, and structural properties of these DMS NCs, confirming the high
quality of the synthesized samples.
Therefore, we have proved that the use of glass matrixes, as host material for DMS NCs, is
an excellent alternative, since they provide great stability to the nanoparticles, with a
relatively low cost of synthesis.
We believe that this chapter may be useful for further investigations on the optical,
magnetic, and structural properties of Mn-doped nanocrystals.

8. Acknowledgements
The authors gratefully acknowledge the financial support from the Brazilian agencies:
MCT/CNPq, Capes, Fapemig and FUNEPU. We are also grateful to our collaborators: Denis
Rezende de Jesus (D. R. Jesus), Fernando Pelegrini (F. Pelegrini), Gilmar Eugênio Marques
(G. E. Marques), Henry Socrates Lavalle Sullasi (H. S. L. Sullasi), Ilde Guedes da Silva (I.
Guedes), Newton Martins Barbosa Neto (N. M Barbosa Neto), Paulo César de Morais (P. C.
Morais), Qu Fanyao (Qu Fanyao), and Walter Elias Feria Ayta (W. E. F. Ayta).

9. References
Archer, P. I.; Santangelo, S. A. & Gamelin, D. R. (2007). Direct Observation of sp-d Exchange
Interactions in Colloidal Mn2+- and Co2+-Doped CdSe Quantum Dots. Nano Lett.,
Vol. 7, Issue 4 (March 2007), pag. 1037, 7 pages, ISSN: 1530-6984
Babentsov, V.; Riegler, J. Scheneider, J.; Ehlert, O.; Nann, T. & Fiederle, M. (2005). Deep level
defect luminescence in cadmium selenide nano-crystals films. J. Cryst. Growth, Vol.
280, Issue 3-4 (July 2005), pag. 502, 7 pages, ISSN: 0022-0248
Bányai, L. & Koch, S. W. (1993). Semiconductor Quantum Dots, World Scientific Publishing
Co. Pte. Ltd., ISBN: 9810213905, Singapore
Barbosa, L. C.; Reynoso, V. C. S; de Paula, A. M.; de Oliveira, C. R. M.; Alves, O. L.;
Craievich, A. F.; Marotti, R. E.; Brito Cruz, C. H. & Cesar, C. L. (1997). CdTe
quantum dots by melting heat treatement in borosilacate glasses. J. Non-Cryst. Sol.,
Vol. 219, Issue 1 (October 1997), pag. 205, 7 pages, ISSN: 0022-3093
Beaulac, R.; Archer, P. I.; van Rijssel, J.; Meijerink, A. & Gamelin, D. R. (2008). Exciton
Storage by Mn2+ in Colloidal Mn2+-Doped CdSe Quantum Dots. Nano Lett., Vol. 8,
Issue 9 (August 2008), pag. 2949, 5 pages, ISSN: 1530-6984
Bacher, G.; Schömig, H.; Scheibner, M.; Forchel, A.; Maksimov, A. A.; Chernenko, A. V.;
Dorozhkin, P. S.; Kulakovskii, V. D.; Kennedy, T. & Keinecke T. L. (2005). Spin-spin
interactions in magnetic semiconductor quantum dots. Physica E, Vol. 26, Issue 1-4
(February 2005), pag. 37, pages, ISSN: 1386-9477
Diluted magnetic semiconductor nanocrystals in glass matrix 167

Becker, R. & Döring, W. (2006). Kinetische Behandlung der Keimbildung in übersättigten

Dämpfen. Ann. Phys., Vol. 416 (reprinted from Vol. 24 (1935)), Issue 8 (March 2006),
pag. 719, 34 pages, ISSN: 0003-4916
Brus, L. E. (1984). Electron-electron and electron-hole interactions in small semiconductor
crystallites: The size dependence of the lowest excited electronic state. J. Chem.
Phys., Vol. 80, Issue 9 (May 1984), pag. 4403, 7 pages, ISSN: 0021-9606
Cheng, Y; Wang, Y.; Bao, F. & Chen, D. (2006). Shape Control of Monodisperse CdS
Nanocrystals: Hexagon and Pyramid. J. Phys. Chem. B, Vol. 110, Issue 19 (April
2006), pag. 9448, 4 pages, ISSN: 1089-5647
Christian, J. W. (1965). The theory of Transformation in Metal and Alloy, Pergamon Press, ISBN:
0080440193, Oxford
Dalpian, G. M. & Chelikowsky, J. R. (2006). Self-Purification in Semiconductor Nanocrystals.
Phys. Rev. Lett., Vol 96, Issue 22 (June 2006), pag. 226802, 4 pages, ISSN: 0031-9007
Dantas, N. O.; Neto, E. S. F.; Silva, R. S.; Jesus, D. R. & Pelegrini, F. (2008). Evidence of
Cd1-xMnxS nanocrystal growth in a glass matrix by the fusion method. Appl. Phys.
Lett., Vol. 93, Issue 19 (November 2008), pag. 193115, 3 pages, ISSN: 0003-6951
Dantas, N. O.; Silva, R. S.; Pelegrini, F. & Marques, G. E. (2009). Morphology in
semimagnetic Pb1-xMnxSe nanocrystals: Thermal annealing effects. Appl. Phys. Lett.,
Vol. 94, Issue 26 (June 2009), pag. 263103, 3 pages, ISSN: 0003-6951
Erwin, S. C.; Zu, L.; Haftel, M. I.; Efros, A. L., Kennedy, T. A. & Norris, D. J. (2005). Doping
semiconductor nanocrystals. Nature, Vol. 436, Issue 7047 (July 2005), pag. 91, 4
pages, ISSN: 0028-0836
Furdyna, J. K. (1988). Diluted magnetic semiconductors. J. Appl. Phys., Vol. 64, Issue 4
(August 1988), pag. R29, 36 pages, ISSN: 0021-8979
Gaponenko, S. V. (1998). Optical Properties of Semiconductor Nanocrystals, Cambridge
University Press, ISBN: 0521582415, Cambridge
Hinckley, C. C. & Morgan, L. O. (1965). Electron spin resonance linewidths of mangane (II)
ions in concentrated aqueous solutions. J. Chem. Phys., Vol. 44, Issue 3 (1965), pag.
898, 8 pages, ISSN: 0021-9606
Jain, M. K. (1991). Diluted Magnetic Semiconductors, World Scientific Publishing Co. Pte. Ltd.,
ISBN: 9810201761, Singapore
Ji, T.; Jian, W.B.; Fang, J. (2003). The first synthesis of Pb1-xMnxSe nanocrystals. J. Am. Chem.
Soc., Vol. 125, Issue 28 (June 2003), pag. 8448, 2 pages, ISSN: 0002-7863
Jian, W. B.; Fang, J.; Tianhao, J. & He, J. (2003). Quantum-size-effect-enhanced dynamic
interactions among doped spins in Cd1-xMnxSe nanocrystals. Appl. Phys. Lett., Vol.
83, Issue 16 (October 2003), pag. 3377, 3 pages, ISSN: 0003-6951
Lee, S.; Dobrowolska, M. & Furdyna, J. K. (2005). Effect of spin-dependent Mn2+ internal
transitions in CdSeZn1−xMnxSe magnetic semiconductor quantum dot systems
Phys. Rev. B, Vol. 72, Issue 7 (2005), pag. 075320, 5 pages, ISSN: 1098-0121
Mukherjee, S.; Pal, A. K.; Bhattacharya, S. & Raittila, J. (2006). Magnetism of Mn2O3
nanocrystals dispersed in a silica matrix: Size effects and phase transformations.
Phys. Rev. B, Vol. 74, Issue 10 (September 2006), pag. 104413 , 10 pages, ISSN: 1098-
0121
Norris, D. J.; Efros, A. L. & Erwin, S. C. (2008). Doped Nanocrystals. Science, Vol. 319, Issue
5871 (March 2008), pag. 1776, 4 pages, ISSN: 0036-8075
168 Nanocrystals

Ohno, H. (1998). Making Nonmagnetic Semiconductors Ferromagnetic. Science, Vol. 281,

Issue 5379 ( 1998), pag. 951, 6 pages, ISSN: 0036-8075
Sapega, V. F.; Moreno, M.; Ramsteiner, M.; Däweritz, L. & Ploog, K. (2002). Electronic
structure of Mn ions in (Ga,Mn)As diluted magnetic semiconductor. Phys. Rev. B,
Vol. 66, Issue 7 (2002), pag. 075217, 6 pages, ISSN: 1098-0121
Silva, R. S.; Morais, P. C.; Fanyao, Q.; Alcalde, A. M.; Dantas, N. O. & Sullasi, H. S. L. (2007).
Synthesis process controlled magnetic properties of Pb1-xMnxS nanocrystals. Appl.
Phys. Lett., Vol. 90, Issue 25 (June 2007), pag. 253114, 3 pages, ISSN: 0003-6951
Silva, R. S. (2008). Síntese e Estudo das Propriedades Ópticas e Magnéticas de Pontos Quânticos de
Pb1-xMnxS Crescidos em Matrizes Vítreas, Tese (Doutorado em Física)-Universidade
de Brasília, 98, [21]f., Brasília
Smyntyna, V.; Skobeeva, V. & Malushin, N. (2007). The nature of emission centers in CdS
nanocrystals Radiation Measurements, Vol. 42, Issue 4-5 (2007), pag. 693, 4 pages,
ISSN: 1350-4487
Volmer, M. & Weber, A. (1926). Nucleus formation in Supersaturated Systems. Z. Phys.
Chem., Vol. 119, Issue (1926), pag. 277, 25 pages, ISSN: 0942-9352
Woggon, U. (1997). Optical Properties of Semiconductor Quantum Dots, Springer – Verlag,
ISBN: 3540609067, Berlin
Xue, H. T. & Zhao, P. Q. (2009). Synthesis and magnetic properties from Mn-doped
CdS/SiO2 core-shell nanocrystals. J. Phys. D: Appl. Phys., Vol. 42, Issue 1 (January
2009), pag. 015402, 5 pages, ISSN: 0022-3727
Zarzycki, J. (1991). Glasses and the vitreous state, Cambridge University Press, ISBN:
0521355826, Cambridge
Zhou, H.; Hofmann, D. M.; Alves, H. R. & Meyer, B. K. (2006). Correlation of Mn local
structure and photoluminescence from CdS:Mn nanoparticles. J. Appl. Phys., Vol.
99, Issue 10 (May 2006), pag. 103502, 4 pages, ISSN: 0021-8979
KTiOPO4 single nanocrystal for second-harmonic generation microscopy 169

0
7

KTiOPO4 single nanocrystal for second-harmonic

generation microscopy
Loc Le Xuan, Dominique Chauvat, Abdallah Slablab and Jean-Francois Roch
Ecole Normale Superieure de Cachan
France

Pawel Wnuk and Czeslaw Radzewicz

University of Warsaw
Poland

1. Introduction
Nonlinear second-harmonic generation (SHG) microscopy has become a commonly used tech-
nique for investigating interfacial phenomena(Kemnitz et al., 1986; Shen, 1989) and imag-
ing biological samples.(Moreaux et al., 2000) Different non-centrosymmetric nanometric light
sources have been recently studied in this context, e.g. organic nanocrystals.(Shen et al., 2001;
Treussart et al., 2003) For those systems, resonant optical interaction leads to an enhance-
ment of the nonlinear response but also to parasitic effect that is detrimental for practical ap-
plications, namely photobleaching due to two-photon residual absorption.(Patterson et al.,
2000) Conversely, inorganic non-centrosymmetric materials with far-off resonance interac-
tion avoid this limitation.(Johnson et al., 2002; Long et al., 2007) Recent achievements have
been obtained using KNbO3 nanowires as a tunable source for sub-wavelength optical mi-
croscopy(Nakayama et al., 2007) and Fe(IO3 )3 nanocrystallites as promising new SHG-active
particles with potential application in biology.(Bonacina et al., 2007) However, either the di-
mensions of the used crystals are still of the order of the micrometer along one axis,(Nakayama
et al., 2007) or the corresponding bulk material is not easily grown,(Bonacina et al., 2007) so
that the crystal characteristics are not directly available. A complementary approach con-
sists in considering a well-known SHG-active bulk material and investigating its properties in
nanoparticle form. Different materials have been considered, e.g. BaTiO3 (Hsieh et al., 2009)
and ZnO (Kachynski et al., 2008). In this view, we were among the pioneers in this domain,
considering the well-known KTP material. Potassium titanyl phosphate (KTiOPO4 , KTP) is a
widely used nonlinear crystal.(Zumsteg et al., 1976) Studies on this material have focused on
the optimized growth of large-size single crystals, which have found numerous applications
in laser technology for efficient frequency conversion.(Driscoll et al., 1986)
Here we show that KTP particles of nanometric size (nano-KTPs) are an attractive material for
SHG microscopy. Under femtosecond excitation and in ambient conditions, a single nano-KTP
with a size around 60 nm independently determined with atomic force microscopy (AFM),
generates a perfectly stable blinking-free second-harmonic signal which can be easily detected
in the photon-counting regime. Furthermore, we demonstrate that this single nanocrystal can
170 Nanocrystals

be characterized in situ by retrieving the orientation of its 3-axis with respect to the optical
observation axis. This analysis uses both nonlinear polarimetry(Brasselet et al., 2004) and
defocused imaging with a model that has been adapted from similar techniques developed
for single-molecule fluorescence.(Bohmer & Enderlein, 2003; Brokmann et al., 2005; Sandeau
et al., 2007)
Such characterized nanocrystal can have potentially many applications in nano-optics or biol-
ogy, and downscaling their size is a crucial step forward. From the point of view of detection
techniques, we have developed two methods for investigating SHG nanocrystals and decrease
the minimum detectable size for a SHG nanocrystal. In particular both techniques exploit the
coherent characteristics of the second harmonic emission. First, a coherent balanced homo-
dyne detection is associated to SHG microscopy, allowing us to detect nano-objects with high
signal-to-noise ratio and phase sensitivity.(Le Xuan et al., 2006) Second, broadband femtosec-
ond pulses are used to improve the second harmonic signal, with a perspective of spectral
manipulation for coherent control of the nonlinear process.(Wnuk et al., 2009)

2. Photostable second harmonic generation from single KTP nanocrystals

The KTP nanoparticles are extracted from the raw powder that remains in the trough at
the end of the flux-growth process which leads to the synthesis of large sized KTP single
crystals.(Rejmankova et al., 1997) We then apply a size-selection procedure by centrifugation
which was previously worked out for optically active nanodiamonds (Treussart et al., 2006)
and which leads to a colloidal solution of non-aggregated particles.
Analysis with dynamic light scattering (DLS) of this solution reveals a strong peak in the size
distribution centered around 150 nm (Fig.1). Centrifugation at a higher speed or for a longer
duration acts as a low-pass filter on the size distribution, reducing the average peak value
from 150 nm to 80 nm (30 min. at 5000 rpm), and 60 nm (10 min. at 11000 rpm).

Fig. 1. Size selection of the nanoparticles in the polymer solution. DLS spectra for successive
centrifugations (a) 5 min. at 5000 rpm, (b) 30 min. at 5000 rpm, (c) 10 min. at 11000 rpm, and
corresponding to an average particle size of (a) 150 nm, (b) 80 nm, (c) 60 nm.

Since we are interested in the study of a well-isolated single nanocrystal, the sample is pre-
pared so as to enable identification of the same single nano-object under AFM, a classical
white-light optical microscope, and a home-made scanning SHG microscope (Figure 2). We
KTiOPO4 single nanocrystal for second-harmonic generation microscopy 171

designed structured samples by first spin-coating the colloidal solution on a glass cover-slip,
then etching it through a mask to leave square areas with nano-KTPs directly on glass. Fig-
ure 3a shows one such area labeled with photo-written numbers by the focus laser itself at
high mean power (100mW at a repetition rate of 86MHz) and imaged using transmission mi-
croscopy in white-light illumination. Figure 3b corresponds to a 10×10 µm2 zoom inside this
domain. Well-contrasted spots are clearly visible and are attributed to the strong elastic scat-
tering from nano-KTPs resulting from the mismatch between the index of refraction of glass
(n ≈ 1.5 in the visible spectrum) and the average one of KTP (n ≈ 1.8). Figure 3c is an AFM
image of exactly the same domain taken in non-contact operation mode. It indeed reveals a
large number of objects with nanometric size.

AFM

S
O
PBS F DM
SM
APD y

APD x
λ/2
P
CCD2

CCD1

L
Fig. 2. Experimental setup. AFM: atomic force microscope mount on top of the optical set-up.
L: femtosecond laser (86 MHz repetition rate, ≈ 100 fs pulse duration, λ = 986 nm), P: polar-
izer, λ/2 half-wave plate, DM: dichroic mirror, SM: scanning mirror, O: microscope objective
(N.A. = 1.40, × 100), S: sample, CCD1 white-light camera, CCD2 sesitive CCD camera, PBS:
polarization beam splitter, APD: avalanche photodiode functioning in photo-counting regime

KTP is a non-centrosymmetric orthorhombic crystal with large nonlinear coefficients, of

(2)
the order of χzzz = 2d33 ≈ 34 pm/V along its 3-axis, when excited in the near in-
frared.(Sutherland, 1996) We use a SHG microscope, as described in Figure 2, to identify if
the scattering objects identified in Figures. 3a an 3b correspond to optically-active particles
with nonlinear response. Femtosecond infrared light pulses (100 fs) at 86 MHz repetition
rate are tightly focused on the sample with a high numerical aperture microscope objective
(NA=1.4, ×100). Since the nanocrystal size is much smaller than the wavelength, we can ne-
glect any phase shift between the fundamental and second-harmonic fields. Compared to
propagation in a bulk nonlinear crystal, there is no phase-matching requirement, thus allow-
ing us to excite the nanocrystal at any wavelength. Here we choose the excitation wavelength
172 Nanocrystals

Fig. 3. Correspondence of nanoparticles in images taken with different observational tech-

niques. (a) Overview in white-light microscopy of the structured sample. The image displays
a square area with nano-KTPs deposited on the glass surface and surrounded by a grid of thin
polymer film. One square is unambiguously located by photoetched numbers on this grid.
(b) 10×10 µm2 zoom of the white-light image. (c) Corresponding 10×10 µm2 AFM scan. (d)
Corresponding raster-scan SHG image. Circles in (c) pinpoint the nanoparticles in the AFM
image which are SHG-active. Note that a small triangular area in the bottom-left corner is not
scanned by the AFM in (c), thus missing one SHG emitter evidenced in (d).

at 986 nm so that the light-matter interaction is highly non-resonant both at this fundamen-
tal and the corresponding 493-nm second-harmonic wavelength, where KTP remains highly
transparent.(Hansson et al., 2000)
The previously observed domain (see Figures 3b and 3c) is raster scanned and the SHG signal
is collected in the backward direction with the same microscope objective and is detected by
avalanche photodiodes operated in the photon-counting regime (Figure 3d). SHG is detected
from a large fraction of the scatterers revealed by the white-light observation of the sample, as
evidenced by the one-to-one correspondence between the spots that appear in Figure 3b and
Figure 3d. Nearly all these emitters can also be unambiguously identified in the AFM image
(white circles in Figure 3c). The AFM measurement determines the height of the nanocrystals,
a value taken as an estimate of the nano-KTP size. A shape and phase analysis of the AFM
scan indicates that other non-emitting nano-objects observed in the AFM image are most likely
small polymer residues left after the etching process of the polymer during the structuration
procedure.
KTiOPO4 single nanocrystal for second-harmonic generation microscopy 173

Fig. 4. Characterization of an isolated nano-KTP: (a) AFM image (scale bar: 200 nm), (b)
photostability of the second-harmonic emission by the nano-KTP with average incident power
8 mW. Quadratic power dependence of the signal, as expected for SHG, has also been checked
(not shown).

We now focus on the study of an isolated nano-KTP, which corresponds to the spot labeled #1
in Figure 3c and 3d. This single particle is representative of the properties of a large number
of similar nanocrystallites. A zoom of the AFM image shown in Figure 4a reveals a single
particle of almost rectangular shape, with a 60-nm height and a size of 120 nm in its transverse
dimensions. We note that the value of the transverse size is over-estimated, due to convolution
of the nanocrystal shape with the AFM tip and to residual polymer that might have stuck to
the nano-KTP sides. The corresponding SHG image in Figure 3d is a diffraction-limited spot
as expected for an emitter with sub-wavelength size. The emitted photon flux at the center of
the spot leads to 2.5×105 detection counts per second for an average incident power of 8 mW.
It leads to a measured signal-to-background ratio as large as 250. Most remarkably, Figure
4b shows the photostability of the emitter under ambient conditions (room temperature and
direct exposure to oxygen) since a constant detection signal is recorded for more than 120
minutes. This feature is due to the highly non-resonant character of the interaction: indeed,
there is no population in any excited level that could lead to the occurrence of photoinduced
degradation processes.

3. Defocused imaging of second harmonic generation from a single nanocrystal

Under similar experimental conditions, we found that the second-harmonic reflection of a
strongly focused infrared excitation beam at the surface of a macroscopic KTP crystal with
known axis orientation can be accurately modeled by the radiation of a single nonlinearly in-
duced dipole.(Le Xuan et al., 2006) To analyze the second-harmonic emission of a single nano-
KTP with sub-wavelength size, we apply the same model which neglects depolarizing and
propagation effects associated to the dielectric material.(Bohren & Huffman, 2004) Under an
incident electromagnetic field at fundamental frequency ω, the nano-KTP is then equivalent
to a single dipole oscillating at frequency 2ω, with components associated to the second-order
susceptibility tensor coefficients
(2)
Pi2ω = 0 ∑ χijk (−2ω; ω, ω ) Eω ω
j Ek (1)
j,k

where i, j, k span the Cartesian coordinates x, y, z in the laboratory frame. Describing the
orientation of the crystal axes by the Euler angles , the nonlinear coefficients in the labora-
174 Nanocrystals

tory frame are related to the known nonlinear coefficients in its crystallographic X, Y, Z axes
through simple rotations
(2) (2)
χijk = ∑ χ I JK cos(i, I ) cos(j, J ) cos(k, K ) (2)
J,K
The resulting nonlinear dipole radiates a second-harmonic field , which is predicted to have
a specific polarization through equations 1 and 2. Following recently developed polarimetry
techniques for nonlinear microscopy,(Brasselet et al., 2004; Le Floc’h et al., 2003) a polarizing
beamsplitter analyzes the SHG orthogonal field components along x and y laboratory axes,
located in the plane perpendicular to the microscope optical axis z. Each corresponding in-
tensity is recorded as the excitation field is rotated with a half-wave plate in the path of the
incident beam. Qualitatively, the polar graph of Figure 5a displays x and y second-harmonic
emission similar to dipolar radiative patterns. They are related to the projection of the non-
linear emitting dipole in the (x,y) plane, i.e. Euler angle φ. In the case of a single nonlinear
coefficient being non-null in Equations 1 and 2, we would have obtained two homothetic polar
graphs. The different responses along x and y are due to the contribution of the five nonlinear
coefficients in the KTP tensor.(Sutherland, 1996) Here, equations 1 and 2 allow us to retrieve
φ = 70◦ ± 5◦ .
This polarimetry technique has however a low sensitivity to any dipole component along the
longitudinal z-axis. To retrieve the orientation of the dipole in the three dimensions, we apply
a defocused imaging technique.(Bohmer & Enderlein, 2003; Brokmann et al., 2005; Sandeau et
al., 2007) For a given direction of propagation in the far field, the radiated field is proportional
to

E2ω    2ω
FF ∝ k × k × P (3)
while the intensity radiation diagram is given by, Ii2ω (k)
∝ 2ω (k )|2 , where i stands for x, y, z.
|Ei,FF
This radiation diagram is directly measured as a function of excitation polarization orientation
by positioning a highly-sensitive CCD camera slightly out of the conjugated focal observation
plane. The experimental defocused image is displayed in Figure 5b. Qualitatively, the outer
"moon-shaped" structure is characteristic of an emitting dipole out of the (x,y) plane of the
sample.(Bohmer & Enderlein, 2003; Brokmann et al., 2005; Sandeau et al., 2007) Compared
to the case of single-chromophore fluorescence,(Bohmer & Enderlein, 2003; Brokmann et al.,
2005) the image exhibits specific features associated to the nonlinear coupling described by
equation 1 between the nonlinearly induced dipole and the polarization state of the excitation
field . This allows us to determine the full set of Euler angles of the observed nanocrystal with
a good accuracy while relying only on the SHG signal.(Sandeau et al., 2007) For the chosen
nano-KTP #1 of Figure 3c, the model developed in (Sandeau et al., 2007) gives a single set of
angles: Figures 5c and 5d display the numerical simulations corresponding respectively to
Figures 5a and 5b with the determined Euler angles of the crystal axes.
For more insight on the method, we focus in the next part on a more detailed 3D orientation
determination of one typical single KTP nanocrystal. The experimental setup is described in
Figure 2. For defocused imaging, the SHG signal is directed towards a sensitive CCD camera
(CCD2 on the setup figure, having 512×512 pixels, 13×13 µm2 pixel size, thermoelectrically
cooled) which is translated by 15 mm towards the objective, so as to obtain a contrasted image
of about 600×600 µm2 size.
KTiOPO4 single nanocrystal for second-harmonic generation microscopy 175

Fig. 5. 3D orientation of an isolated nano-KTP: (a) polar graph showing the measured SHG
emission of the nanocrystal along x (red points) and y (blue points) axes, as a function of the
direction of the linearly polarized excitation field, (b) corresponding experimental defocused
image (scale bar: 250 µm), (c) and (d) numerical calculation of the polar graph and the de-
focused image respectively for a nano-KTP with Euler angles of the crystalline axis equal to
(70◦ ± 5◦ , 15◦ ± 5◦ , 60◦ ± 5◦ ).

Figure 6 exhibits measured defocused images from a single KTP nanocrystal for different po-
larization orientations of the incoming fundamental field in the sample plane. Different fea-
tures appear on these images. First, they exhibit a symmetry axis which orientation in the
sample plane is related to the nanocrystal in-plane orientation. Second, the structure of the
image is characteristic of a nanocrystal exhibiting out-of-plane orientation with θ = 0◦ . Third,
the symmetry axis of the structure rotates when changing the incident polarization direction.
This is characteristic from a nonlinear coherently induced dipole emission as opposed to a
single fluorescent dipole, since the KTP multipolar symmetry structure allows coherent non-
linear coupling with various incident field polarization directions.
The measured images are interpreted using a vectorial model calculation, in which the KTP
nanocrystal is modeled by a cubic shape sampled by placing one individual dipole every nm3
within this cube. The experimental imaging configuration is taken into account. Starting from
the ring pattern observation for a first estimate of the 3D orientation parameter, the emission
pattern calculation (Fig. 6(f-j)) is then implemented to manually adjust at best the measured
images (Fig. 6(a-e)). This leads to a 3D orientation determination with error margins of ±
5◦ at maximum for θ and φ. This error is much larger for the ψ Euler angle (from ± 10◦
to ± 50◦ ), which is specific to the case of KTP for which there is only a weak difference be-
tween nonlinear coefficients involving the "1" and "2" directions perpendicular to the main
"3" axis.(Vanherzeele & Bierlein, 1992) In order to confirm the values for the obtained orien-
tation angles, they were used as input parameters to calculate the in-plane polarimetric SHG
response of the measured nanocrystal. The corresponding angular dependencies agree well
 metric s
176 Nanocrystals
previou
two KT
(a)
1.2
(b)
3.0
2.5
(c)
6
(d)
6
(e)
2.5
sole obs
2.0 4 2.0
0.8 1.5 4 1.5
0.4 1.0 2 2 1.0
0.5 0.5
0.0 0.0 0 0 0.0 X

(f) (g) (h) (i) (j)

6
1.2 2.0 5 4
6 Y
1.5 4 3
0.8 3 4
1.0 2
0.4 2 2
0.5 1 1
0.0 0.0 0 0 0

x x
0 0
1000 (k) 5000 (l)
45 315 45 315

IX IY
y 90 0 270 y 90 0 270

1000

135 225 135 225

180 180

Fig. 6. (a-e) Experimental defocused images of a KTP nanocrystal for several incident po-
larization directions relative to the x-axis: (a) 0◦ , (b) 60◦ , (c) 90◦ , (d) 120◦ and (e) 150◦ . The
Fig. 4. (a-e)
images sizes areExperimental
600×600 µmdefocused images
2 . The integration timeoffora each
KTPimage
nanocrystal
is 5 s. (f-j)for several inci-
Corresponding
dent polarization ◦ ◦ 60 ◦ 90◦◦± ◦
calculated images,directions
leading to relative to the parameters
the orientation X axis: (a) (θ 0= ,30(b)
± 5◦ , φ, (c)
= 115 , (d)
5◦ , ψ120
=
and ◦
90 (e) ◦
± 20150 ◦
). (k,l) Experimental
. The images sizespolarimetric analysis
are 600×600 µm 2
Ix and Iy of
. The this nanocrystal
integration time (circle mark-
for each im-
ers) and corresponding calculated polarization responses (lines) for the
age is 5 s. (f-j) Corresponding calculated images, leading to the orientation parameters Euler set of angles
(θdeduced
= 30◦ ±5 from
◦ ,φ the defocused
= 115 ◦ ±5◦ ,ψimaging analysis.
= 90◦ ±20 ◦ ). (k,l) Experimental polarimetric analysis I and
X
IY of this nanocrystal (circle markers) and corresponding calculated polarization responses
F
(lines) for experimental
with the the Euler set data
of angles deduced
(Fig. 6(k,l)). from
Note thatthe defocused
a direct imaging of
determination analysis.
the Euler angles
ti
from such polarimetric responses would have been more delicate since the KTP symmetry 8
exhibits several ambiguous situations when projected in the sample plane. This points out the
P
larizationadvantages
direction.ofThis
the defocused imaging technique,
is characteristic which encompasses
from a nonlinear coherently a complete
induced 3Ddipole
couplingemission
as opposed
information in the observed pattern shapes.
to a single fluorescent dipole, since the KTP multipolar symmetry structure allows
r
Performing numerical simulation on nanocrystal of different sizes, we note that that small dif-
coherentferences
nonlinear coupling
occur withthe
but in general various incident
patterns field polarization
of the defocused images havedirections.
similar features, with
(
a symmetry axis along the direction of the nonlinear
Starting from the ring pattern observation for a first estimate of the 3D induced dipole. Theorientation
ring-structuredparameter,
shape appears to be in all cases representative of the tilt angle of the "3" axis of the crystal with
the emission pattern calculation (Fig. 4(f-j)) is then implemented to manually adjust at best the A fin
respect to microscope objective axis z. The rough similarity among the observed pattern with
measuredrespect
images (Fig. 4(a-e)).
to crystal size shows This
thatleads to aimaging
radiation 3D orientation
can be used determination with error
as a robust technique to de-margins
of ±5◦ attermine
maximum for θindependently
orientation and φ . This on the is
error crystal
much size, provided
larger thatψ
for the it lies below
Euler 150-200
angle nm.±10 ◦ to
(from
one of t
Above is
±50◦ ), which this size range,
specific the fine
to the casestructure
of KTPoffor the which
observed ringsisisonly
there seen to clearlydifference
a weak deviate frombetween approac
the single dipole model. For KTP nanocrystal of size smaller than 100 nm in diameter, the SHG
nonlinearemission
coefficients
can be involving the “(1,2)”
modeled correctly directions
in a dipolar perpendicular
approximation. to the
In addition main
to 3D “3” axis [25].
orientation erential
The ψ angle is thus not
determination, suchstrongly relevant
a technique by the
presents nature in property
unique the present study. Ina diagnostics
of providing order to confirm
for the with lar
values forthethe
crystalline
obtained quality of an isolated
orientation nano-object.
angles, they were Indeed
usedtheaspresence of nanocrystals
input parameters to of dif-
calculate the
ferent orientations in a same focus point would strongly affect the defocused image properties
in-plane described
polarimetric SHG response of the measured nanocrystal. The corresponding angular crystalli
above, and allow a direct conclusion on the mono-crystallinity of the observed ob-
dependencies agree well
ject. This is illustrated with the experimental
in Fig. 7, representativedata (Fig. 4(k,l)).
of a nanometric size Note
KTP forthat a direct
which imagesdetermi- curve ca
nation ofarethe
deprived
Euler from
anglesa symmetry
from such axis and deviate from
polarimetric the previously
responses wouldobserved
havering
beenstructures.
more delicate tals orie
since the KTP symmetry exhibits several ambiguous situations when projected infocal
Such images are representative of the presence of at least two KTP nanocrystals in the the sample
plane. This points-out the advantages of the defocused imaging technique, which encompasses
a complete 3D coupling information in the observed pattern shapes. #85722 -
The application of this technique to orientation measurements of more than ten isolated
(C) 2007
nanocrystals shows furthermore that only up to two incident polarization angles are neces-
sary for an unambiguous angular determination. This is illustrated in Fig. 5 for two different
 size KTP for which images are deprived from a symmetry axis and deviate fro
usly observed ring structures. Such images are representative of the presence
KTiOPO4 single nanocrystal for second-harmonic generation microscopy 177
of a
TP nanocrystals in the focus point of the objective, which could not be revealed
volume
servation of theofdiffraction-limited
the microscope objective, which
spotcould
size.not be revealed by the sole observation of
the diffraction-limited spot size.

12
5 X
4 4
8 3 3
2 2
4
1 1
0 0 0
Y

45 3000 315

90 270

135 225

180

Fig. 7. (a-c) Defocused images of a nanocrystal for several incident polarization directions
relative to the x (horizontal) axis: (a) 0◦ , (b) 306◦ and (c) 90◦ . The images sizes are 800×800
µm2 . This nano-object is visibly constituted of several distinct sub-domains, (d) Polarimetric
Fig. 6. (a-c)analysis
Defocused images of a nanocrystal for several incident polarization direc-
of this nano-cluster of nanocrystals. The adjustment of polarimetric responses (black
line)
ions relative tois the
doneX using two nanocrystals
(horizontal) (a) 0◦ , (b)
of respective
axis: 306◦ and
orientations (θ1 =(c) φ1 ◦=. 260
20◦ ,90 The ◦ , ψ = 0◦ )
images
1 sizes are
and ◦ ◦ ◦
2 (θ2 = 45 , φ2 = 350 , ψ2 = 0 ).
800×800µ m . This nano-object is visibly constituted of several distinct sub-domains. (d)
PolarimetricA analysis of this
finer observation nano-cluster
allows us to concludeof thatnanocrystals. The adjustment
in this two-KTP nanocrystals of ofpolarimetric
picture, one
them is likely to lie close to the vertical x axis of the images, since the SHG pattern approaches
responses (black line) is done using two nanocrystals 1 and 2 of respective orientations
a ring structure for a x-polarized excitation. The other KTP nanoparticle is preferentially ex-
(θ1 = 20 ,φcited
◦
1 =for260 ◦ ,ψ = 0◦ ) and (θ =◦45◦ ,φ = 350◦ ,ψ = 0◦ ).
a polarization
1 angle close to
2 45 (relative 2 to x), and therefore
2 manifests with larger
signals in Fig. 7(b,c) as the excitation angle increases. This deviation from a mono-crystalline
structure is confirmed in the polarimetric response of such a nano-object, which curve cannot
be fitted using a single KTP unit-cell model, but rather by two or more nanocrystals orienta-
ner observation can furthermore conclude that in this two-KTP nanocrystals p
tions.
them is likely to lie close to the vertical X axis of the images, since the SHG p
We have demonstrated in this section that defocused imaging of second harmonic generation
radiation from a single KTP nanocrystal can provide a direct retrieval of its 3D orientation.
ches a ring structure for a X-polarized excitation. The other KTP nanoparticle is
This orientation information, based on the knowledge of the nano-object crystalline symme-
◦ excitation-dependent polarization
lly excited for a polarization angle close to 45 (relative to X), and therefore ma
try, can thus be used as a complete characterization of the
state from a nonlinear nanosource. In addition, we show that this technique provides a unique
rger signals in Fig. 6(b,c) as the excitation angle increases. This deviation from a m
possibility to inform on the mono-crystallinity nature of a nano-object. Finally, this analysis
method is a promising tool of analysis of the vectorial electric-field character from an un-
ine structure is confirmed in the polarimetric response of such a nano-object,
known electromagnetic incident radiation, taking advantage of the rich features appearing
annot be fitted using a single KTP unit-cell model, but rather by two or more nan
from the nonlinear coupling between matter at the subwavelength scale and complex optical
polarization components.
entations. The polarimetric responses could indeed be adjusted by the coherent ad

$15.00 USD Received 26 Jul 2007; revised 18 Oct 2007; accepted 18 Oct 2007; published 20 N
7 OSA 26 November 2007 / Vol. 15, No. 24 / OPTICS EXPRESS
178 Nanocrystals

4. Balanced homodyne detection of second harmonic radiation from single KTP

nanocrystal
SHG is a coherent process thus interferometric detection schemes are well suited to the detec-
tion of the emitted field. Among possible techniques, coherent balanced homodyne is known
to exhibits a sensitivity to extremely low photon flux rates (Yuen et al., 1983). Although high
sensitivity with such a technique has been demonstrated to detect surface SHG emission from
an air-semiconductor interface (Chen et al., 1998), to the best of our knowledge it had never
been applied to the study of nano-objects. Since the optical phase of the SHG emitted field
is directly related to the sign of the associated nonlinear susceptibility, measuring directly its
value provides important information about the nonlinear material (Stolle et al., 1996). At
nanometric scale, it allows to determine the absolute orientation of dipoles with nonlinear
hyperpolarisabilities such as organic nanocrystals (Brasselet et al., 2004), to detect the polar-
ity in cell membranes (Moreaux et al., 2000), or to study the boundary between microscopic
crystalline domains (Laurell et al., 1992). Such application prospects strongly motivate the
development of phase-sensitive SHG microscopy with highly sensitive detection compared
to SHG interferometric schemes previously developed (Stolle et al., 1996). In this work co-
herent balanced homodyne detection is associated to SHG microscopy, allowing us to detect
nano-objects with high signal-to-noise ratio and phase sensitivity.
The principle of the experiment is shown in Fig.8. Femtosecond infrared light pulses are
injected in an inverted optical microscope and tightly focused. Backward second-harmonic
emission by a nonlinear crystal placed on a glass cover slide at the microscope focus is col-
lected by the microscope objective and transmitted through a dichroic mirror toward the de-
tection set-up. A small fraction of the fundamental beam is also reflected from the cover-slide
and follows the same optical path, providing the phase reference. Second-harmonic emis-
sion of the crystal, further referred as “signal”, corresponds to an electrical field E(2ω ) =
(2ω ) (2ω )
| E(2ω ) | exp[i Φobj ] , where Φobj is the phase shift of the SHG field emitted by the object as
(ω )
compared to the one of the incident fundamental field Ein on the object. For a macroscopic
(ω ) 2 (2ω )
nonlinear crystal, the SHG field is E(2ω ) ∝ χ(2) Ein . Thus, Φobj reflects the phase of the
nonlinear susceptibility χ (2) .
For a nonlinear crystal in its spectral transparence window, χ(2)
is real with a positive or negative value depending on the absolute orientation of the nonlin-
(2ω )
ear response in respect to the crystal axis. Therefore determination of Φobj allows to infer the
absolute orientation of the nonlinear crystal.
For the coherent optical homodyne detection of the SHG signal, the local oscillator (LO) is
generated by sending part of the incident infrared beam into a bulk BBO nonlinear crystal
(Fig.8). A Glan prism ensures linear polarization of the local oscillator along the x-axis and
x-polarized signal and local oscillator are recombined by a non-polarizing 50/50 beamsplit-
ter cube. The 180◦ out-of-phase interferences at the two output ports of the beamsplitter are
detected by photodetectors D1 and D2 . In balanced detection mode, the two resulting pho-
toelectric signals S1 and S2 are subtracted 1 , thus canceling out their DC component. The
interferometric signal is then
  
(2ω ) (2ω ) (2ω ) (2ω )
∆S(2ω ) = K × 2V PLO Psig cos Φsig − ΦLO (4)

1 We use an autobalanced photoreceiver (Nirvana, Model 2007, New Focus, San Jose CA) which achieves
50 dB common noise rejection
KTiOPO4 single nanocrystal for second-harmonic generation microscopy 179

∆S (2ω) ∆S (ω)
2
L

(V)
P O
0
−2
4

(V)
0
−4
0 1 2 3
t(ms)
BS
F APD
PZT P
BBO

BS
D1
+/−
∆S (2ω)
D2

D3
CC +/−
∆S (ω)
D4

Fig. 8. Schematic of the experimental set-up. L: femtosecond laser (86 MHz repetition rate,
≈ 100 fs pulse duration, λ = 986 nm); BBO: nonlinear χ(2) crystal; P: Glan prism; O: micro-
scope objective (N.A. = 1.40, × 100) leading to a ≈ 300 nm FWHM diameter focal spot; BS:
non-polarizing beamsplitter; CC: corner-cube; PZT: mirror mounted on a piezoelectric trans-
ducer; KTP: macroscopic KTiOPO4 crystal or single sub-wavelength size crystal; F: SHG filter;
APD: avalanche photodiode in photon counting regime; D1 and D2 : p-i-n Si photodetectors
of the balanced receiver recording SHG at 2ω (a SHG filter, not shown, is put in front of the
detectors); D3 and D4 : the same for the balanced receiver at fundamental optical frequency ω
(an IR filter, not shown, is put in front of the detectors). Insert: full lines : detected signals as a
function of time for a forward translation of the mirror. Up: signal ∆S(ω ) at fundamental fre-
quency with a sinusoidal fit at frequency f = 1.9 kHz, and bottom: ∆S(2ω ) at SHG frequency
with a sinusoidal fit at frequency 2 f = 3.8 kHz, dashed line: voltage applied to the PZT on half
a period (vertical scale: a.u.).

(2ω ) (2ω )
where PLO and Psig denote the mean optical power of local oscillator and signal, V is the
fringe visibility, and K includes the quantum efficiency of the photodiodes. The phase shift
(2ω ) (2ω ) (2ω ) (2ω )
of the signal at frequency 2ω , Φsig = Φ1 + Φ2 contains the phase shift Φ1 of the
(2ω )
x-polarized radiation emitted by the nonlinear object and the phase shift Φ2 due to propa-
gation in the interferometer.
(2ω )
In order to measure the phase shift Φsig , temporal interference fringes are created by displac-
ing a mirror mounted on a piezoelectric transducer (PZT) driven with a triangular shape volt-
age. The resulting fringes are shown in the inset of Fig.8. To extract the phase shift informa-
tion, fringes at 2ω are compared to reference fringes resulting from fundamental beam. More
precisely, we perform with detectors D3 and D4 a separate balanced homodyne detection of
the residual fundamental beam reflected by the sample by mixing it with the residual light at ω
180 Nanocrystals

which follows the reference arm. A reference interference signal (see inset of Fig. 8) is then ob-
 
( )   (ω ) (ω ) (ω ) (ω )
tained, equal to ∆S ω = K × 2V PLO Psig cos Φsig − ΦLO with same notations mean-
   
(2ω ) (2ω ) (ω ) (ω )
ing as in Eq. 4. The quantity to be extracted is then ∆Φ = Φsig − ΦLO − Φsig − ΦLO
(2ω )
which can be written as ∆Φ = Φ1 + Φ0 , where Φ0 is an “instrumental” phase shift. Al-
though Φ0 can be measured with a reference crystal its value is not required if we are only
(2ω )
concerned in variation of Φ1 while keeping Φ0 constant. Finally, the value of ∆Φ is ex-
tracted using a numerical procedure.

∆S (2ω) (mV)
(a) KTP
z 600 (c)
Xφ 400
y 200
Y θ
0
x ψ Z cover 0 100 200 300
ψ (degrees)
plate ψ (degrees)
(d)
∆S (2ω) (mV)

80
∆φ (degrees)

(b) 300

180◦
40
200

0 100
0 4 8 0 100 200 300
(ω)
Pin (mW) ψ (degrees)

Fig. 9. Results from the balanced homodyne detection. (a) Laboratory axes (x, y, z) with x-
polarized incident fundamental beam propagating along z-axis. (X, Y, Z) KTP crystal principal
axes (θ = 90◦ and φ = 23.5◦ ). Full lines denote axes in the (x, y) plane. ψ : rotation angle of the
χ(2) nonlinear crystal around the z-axis. (b) Linearity of ∆S(2ω ) as a function of fundamental
incident power. Points : experimental data, full line : best linear dependence fit. (c) : ∆S(2ω )
as a function of ψ (in degrees). (d) : Relative SHG phase shift ∆Φ as a function of ψ.

A bulk KTP crystal with axes as shown in Fig.9 is used to test the setup. The Z-axis is parallel
to the crystal input face. With a strongly focused pump beam, a SHG backward emitted beam
from the interface is observed (Boyd, 2003). We first check in Fig.9b that ∆S(2ω ) associated to
(ω )
this surface SHG grows linearly with the fundamental incident power on the crystal Pin , as
expected from Eq.4. If the crystal is rotated by 180◦ around the microscope optical axis (z-axis),
the nonlinear coefficient χ(2) is transformed in −χ(2) and a 180◦ SHG-phase-shift is expected.
Results are shown in Fig.9c. The amplitude shows a maximum of x-polarized SHG when the
linear polarization is aligned along the Z-axis of the crystal (ψ = 0◦ or 180◦ ) which exhibits
the highest nonlinear coefficient. The full line is the theoretical result assuming an x-polarized
incident plane wave normal to the interface and a single nonlinear emitting dipole which
takes into account all nonlinear coefficients of the KTP χ(2) tensor. It yields good agreement
KTiOPO4 single nanocrystal for second-harmonic generation microscopy 181

with experimental points. The phase-shift ∆Φ (Fig. 9d) remains constant while the nonlinear
dipole has a positive projection along the x-axis (0◦ < ψ < 90◦ ). It then endures a 180a-phase-
˛
shift when it becomes negative (90◦ < ψ < 180◦ ). The small deviation to the constant value is
attributed to drift of the φ0 parameter during the step-by-step rotation of the KTP crystal.

NSHG (cts/20 ms)

(nm)

800
1000

(a) (a)400 (b)

0

0
0 2 4
−1000

(ω)
−1000 0 1000 (nm) Pin (mW)
∆S (2ω) ∆S (ω)

+1
(c)
(V)

0
−1
−2
+4
(mV)

0 (d)
−4
t(ms)
1 2 3 4 5 6

Fig. 10. Application of the balanced homodyne technique to sub-wavelength-size nonlinear

crystals. (a) Raster-scan image of SHG signal from a KTP nano-crystal with an avalanche
photodiode. The FWHM diameter of SHG spot intensity is about 300 nm, very close to the
theoretical two-photon microscope resolution (320 nm FWHM). (b) Number of detected SHG
photons as a function of incident power with a best square-law dependence fit. (c) Funda-
mental interference signal ∆S(ω ) as a function of time for a single nano-crystal and (d) corre-
sponding SHG interference signal ∆S(2ω ) .

We then apply the method to the detection of SHG from KTP nanocrystals. To first locate the
crystals, the sample is raster-scanned in x and y directions with piezoelectric translators and
SHG is detected with the avalanche photodiode as shown in Fig.10. A SHG signal image of
such a nano-crystal is shown in Fig.10a. A quadratic dependence of the detected SHG inten-
(ω )
sity is observed upon varying Pin as expected (see Fig.10b). After positioning the focused
infrared excitation beam at the center of the detected emission spot of a single nano-crystal,
we switch to the coherent balanced homodyne detection set-up. SHG interference fringes
(Fig.10d) associated to fundamental ones (Fig.10c) are clearly visible, giving evidence for the
coherence of the nano-crystal SHG emission.
Since the balanced homodyne detection method consists in the projection of the signal electric
field on the spatio-temporal mode of the local oscillator, mode-matching between signal and
LO beams determines maximal amplitude of the fringes. This can be quantified by the fringe
visibility V of Eq.4 being equal to unity for perfect mode-matching. With a 2 mW LO average
182 Nanocrystals

power, the smallest SHG fringe amplitude measured with 1 s duration averaging is equivalent
(2ω )
to a signal power Psig ≈ (7.2 ± 5.0) × 10−19 W assuming V = 1. The uncertainty evaluated
for 95 % confidence interval is equivalent to 3.2 detected photons/s, close to the shot noise
limit. However a unity fringe visibility is practically difficult to achieve. In an auxiliary ex-
periment using equal powers for signal and LO beams, we measure V = 0.21, leading to an
actual sensitivity of 80 photons/s. Such a reduced value for the visibility factor is attributed
to imperfect mode-matching between signal and LO modes, including polarization mismatch,
wave-front distortion, and frequency chirp on the SHG-emitted femtosecond pulses.
To conclude this section, we have demonstrated a SHG phase-sensitive microscope with co-
herent balanced homodyne detection, showing a sensitivity at the photon/s level. The high-
spatial resolution and the sensitivity of the technique is well-adapted to the study of SHG
from nano-crystals.

5. Coherent nonlinear emission from a single KTP nanoparticle with broadband

femtosecond pulses
As a coherent process, the number of SHG photons emitted by a noncentrosymmetric
nanoparticle scales as the square of the number of "oscillators" in the nanocrystal, i.e., as the
sixth power of the nanoparticle average size. As a two-photon process the SHG intensity is
also expected to scale, for a constant average laser power, as the inverse of the pulse time
duration. It is therefore tempting to reduce the pulse duration from the standard 100 fs to 10
fs available from broadband ultrafast lasers in order to enhance the second harmonic photon
emission rate by an order of magnitude, thus reducing the limit size of detectable nanoparti-
cles. Nevertheless, this downscaling has to be done with care since high-order phase disper-
sion of the microscope objective and other optical components induce temporal aberrations
in the excitation pulse interacting with the nanoparticle at the focus of the microscope (Guild
et al., 1997). Using recently developed techniques for temporal characterization at the micro-
scope focus and careful precompensation (Lozovoy et al., 2004), SHG from large objects has
indeed been shown to scale as the inverse of the pulse time duration (Xi et al., 2008). Ad-
ditionally, the broadband excitation corresponding to a 10-fs pulse offers the possibility of
spectral manipulation for coherent control (Warren et al., 1993; Weiner, 2000) of two-photon
absorption and non-resonant second harmonic generation (Broers et al., 1992; Meshulach &
Silberberg, 1998; Wnuk & Radzewicz, 2007).
While it has been recently shown that a SHG-active nanoparticle of a size about a few hun-
dreds of nanometers can be used for pulse detection in a nanoscopic version of the FROG
technique (Extermann et al., 2008), the effects of manipulation of the excitation beam on SHG
emission, i.e.,decreasing the time duration of a precompensated pulse as well as structuring
its broadband spectrum, have not yet been investigated at the single nanoparticle level with
a size well below the wavelength of light. Moreover in this size range phase matching con-
ditions are automatically fulfilled, improving the coherent emission. Here we have shown
that the use of broadband ultrashort laser pulses, precompensated using an automatic genetic
algorithm,improves the second harmonic emission from a single nanoparticle of size about
100 nm. It results in a contrast enhancement of the SHG image obtained by raster scanning
the sample. In this process, smaller nanoparticles are revealed for a given background. We
also manipulate the broadband incident spectrum in a very simple manner to obtain a non-
degenerate sum-frequency generation from a single nanoparticle.
KTiOPO4 single nanocrystal for second-harmonic generation microscopy 183

The experiment is realized using a titanium doped sapphire (Ti:Sa) femtosecond laser with
a 100 nm bandwidth. The laser spectrum is sufficiently broad to support pulses of approxi-
mately 10 fs duration. A genetic algorithm (Baumert et al., 1997) is used to search for the best
phase correction on the incident pulse. A bulk KTP crystal with one face polished served as
reference sample. Under femtosecond beam illumination, the second harmonic signal emitted
by the crystal and detected with an avalanche photodiode operated in photoncounting regime
served as the feedback information in the genetic algorithm procedure. The bulk crystal, in-
stead of a nanocrystal, was selected because it provides a high signal to noise ratio, which
speeds up the search for the optimal phase of the excitation pulse.
The phase
ratio, correction
which speeds was
up theachieved with
search for thetwo consecutive
optimal phase ofsystems: a two-prism
the excitation compressor
pulse. The phase
used for the rough compensation of (mostly quadratic) phase, and a compact
correction was achieved with two separate systems (Fig. 1.a): (1) a two-prism compressor pulse shaper us-
ingused
a diffraction grating
for the rough and a spatial
compensation light modulator
of (mostly (SLM) and
quadratic) phase, (Weiner, 2000). The
(2) a compact SLM-based
pulse shaper
pulseshaper was used
using a diffraction for the
grating andgenetic algorithm
a spatial search (SLM)
light modulator for the [11].
optimal
The phase.
SLM-basedTo evaluate
pulse
theshaper
resultwasof this
usedsearch
for thewe estimated
genetic the duration
algorithm search forof
thethe excitation
optimal phase.pulse at the focus
To evaluate of the
the result
microscope by recording
of this search the interferometric
we estimated the duration of theautocorrelation of at
excitation pulse thethe
pulse
focususing the
of the second har-
microscope
monic field generated
by recording in the KTP autocorrelation
the interferometric nonlinear crystal. Thepulse
of the result of the
using thegenetic
second algorithm search
harmonic field
leads to a pulse of, approximately, 13 fs duration. Once the optimal pulse shapeleads
generated in the KTP nonlinear crystal. The result of the genetic algorithm search was to
found
a
its pulse
durationof, approximately, 13 fs duration.
can then be increased (see Fig.
to any value 1(b)). Once
by adding, with the
the optimal pulse shapeamount
SLM, a controlled was
found
of the its duration phase.
second-order can thenThebe corresponding
increased to anySHGvaluephoton
by adding, with was
number the SLM, a controlled
then measured as a
amount of the second-order phase.
function of the corresponding pulse duration. The corresponding SHG photon number was then
measured as a function of the corresponding pulse duration.

Fig. 11. Detected rate ofrate

Fig. 2. Detected theofSHG from
the SHG froma a100
100 nm KTPnanocrystal
nm KTP nanocrystal as a function
as a function of pulse time
of pulse time
duration showing the improvement of the SHG signal with the inverse of the pulse duration.
duration showing the improvement of the SHG signal with the inverse of the pulse duration.
For a constant average incident power, the SHG signal is expected to scale as the inverse
Forofathe duration
constant of the pulse
average at the
incident microscope
power, the SHG focus. Indeed,
signal we observed
is expected a linear
to scale increase
as the of of
inverse
thethe SHG signal
duration of thewith
pulsetheatinverse of the time
the microscope duration,
focus. as shown
Indeed, in Fig. a2.linear
we observed However, the of
increase
themeasured
SHG signal average slope
with the is approximately
inverse 0.7 insteadasof
of the time duration, unity.inThis
shown Fig. is
11.due in part the
However, to the
mea-
broadband spectrum which cannot be perfectly compressed by the SLM due to pixelization of
sured average slope is approximately 0.7 instead of unity. This is due in part to the broadband
the correction signal and imperfections in spectral wings compensation. Analysis of the data
spectrum which cannot be perfectly compressed by the SLM due to pixelization of the cor-
shown in Fig. 2 reveals that the slope is indeed close to unity for long pulses, while it
rection signalwith
decreases and the
imperfections in spectral
pulse duration. Such wings compensation.
a behavior is consistentAnalysis of the dataphase
with imperfect shown
compensation – even small phase errors become important when the pulse duration
approaches the Fourier transform limit. Despite this imperfect phase compensation, the SH
count rate is nevertheless improved by almost an order of magnitude, as compared to
excitation by standard 100 fs pulses with the same average intensity.
With precompensated pulses, we recorded maps of the SH signal originating from well
dispersed KTP nanoparticles (nanoKTP) deposited by spin-coating a colloidal solution on a
glass cover-slip [5]. The maps were acquired by raster scanning the sample with a PZT-driven
translation stage. Figure 3 shows maps of the same area of the sample recorded with different
184 Nanocrystals

in Fig. 11 reveals that the slope is indeed close to unity for long pulses, while it decreases
with the pulse duration. Such a behavior is consistent with imperfect phase compensation:
even small phase errors become important when the pulse duration approaches the Fourier
transform limit. Despite this imperfect phase compensation, the SH count rate is neverthe-
less improved by almost an order of magnitude, as compared to excitation by standard 100
fs pulses with the same average intensity. With precompensated pulses, we recorded raster-
scan maps of the SH signal originating from well dispersed KTP nanoparticles deposited by
spin-coating a colloidal solution on a glass cover-slip. Figure 12 shows maps of the same
area of the sample recorded with different pulse durations ranging between 200 fs and 13 fs.
Firstly, for nanoparticles already visible at 200 fs, a clear increase of the signal-to-noise ratio
is observed for shorter pulses. Secondly, Fig.12 reveals that a higher number of nanoparticles
appear in the maps acquired with shorter pulses, since smaller objects can be detected due to
enhanced second harmonic emission. This contrast enhancement is crucial for many practical
applications of nonlinear microscopy.
The pulse autocorrelation measured on an individual KTP nanocrystallite in the focus of the
microscope is close to the Fourier limit, which confirms the efficiency of the phase compen-
sation and gives evidence that all spectral components of the precompensated pulse are con-
verted by the nanoparticle. We note that the bandwidth of the SHG is, for bulk nonlinear
crystal, limited by phase matching conditions (Boyd, 2003). However, because of the sub-
wavelength dimension of the nanoparticle spectral filtering associated to phase-matching be-
comes negligible even for ultrashort femtosecond laser pulse.
The broadband nonlinear response opens the way to the coherent control of the different spec-
tral components of the single incident beam in order to manipulate the second-order nonlinear
emission (Dudovich et al., 2002).
As a proof-of-principle of such spectral manipulation, we used a two-band incident spectrum,
containing well-separated but coherent "red" and "blue" parts (see inset of Fig. 13),in order to
excite and subsequently filter out a non-degenerate sum-frequency (SF) signal. If we simply
assume two average excitation frequencies "R" and "B" in the single incident beam,the SF
signal is due to a nonlinear polarization of the form (Boyd, 2003):
(2)
PωR +ωB = 0 ∑ χijk (−ω R − ω B ; ω R , ω B ) Eω R ωB
j Ek (5)
j,k

where i, j, k stand for x, y, z, the laboratory axes. With two well-separated spectral bands,
Eq.5 shows that we have access to the different polarizations j, k of the driving fields, then
(2)
select a specific nonlinear coefficient χijk , and thus control the direction i of the nonlinear
dipole at the generated sum frequency. This can be applied e.g. to enhance the recognition of
the nanoparticle-marker in a complex environment, and to investigate the shape dependence
of the nonlinear optical response of an individual nanoparticle. Yet another possible future
application concerns the measurement of the polarization properties of a femtosecond pulse in
the focus of a microscope objective. In principle, the spatial resolution of such a measurement
is limited by the minimum size of the nanocrystallite that produces a measurable signal. With
sub 100-nm detection limit demonstrated in our experiment we hope to obtain in the future a
detailed 3-D map of the polarization in the vicinity of the focus.
A simple two-band excitation is performed, as an experimental proof of principle, by blocking
a part of the incident pulse spectrum. A bandpass filter is then placed in the detection path to
transmit ω R + ω B , and reject both 2ω R and 2ω B . We checked that excitation with any single
spectral band, i.e., ω R or ω B , does not produce SF signal. Varying the power of the blue band
 observation confirms the efficiency of the phase compensation and gives evidence that all
spectral components of the precompensated pulse are converted by the nanoparticle. We note
that the bandwidth of the SHG is, for bulk nonlinear crystal, limited by phase matching
conditions [17]. However, because of the sub-wavelength dimension of the nanoparticle
spectral filtering associated to phase-matching becomes negligible even for ultrashort
KTiOPO4 single nanocrystal
femtosecond laser pulse.for second-harmonic generation microscopy 185

Fig. 12. SHGFig. maps

3. SHGof maps
the same area
of the same ofofthe
area nanoKTP
the nanoKTP sample.
sample. The
The maps maps correspond
correspond to excitation to excita-
with pulse durations ! of (a) 200 fs (SNR=2.5), (b) 100 fs (SNR=4), (c) 65 fs (SNR=4.5) and
tion with pulse durations of (a) 200 fs (SNR=2.5), (b) 100 fs (SNR=4), (c) 65 fs (SNR=4.5)
(d) 13 fs (SNR=16). For each pulse duration both a three-dimensional graph of the SH intensity
and(d)
13 fs (SNR=16).as wellFor
as a each pulse
normalized duration both
two-dimensional a shown.
map are three-dimensional graph of the SH intensity
as well as a normalized two-dimensional map are shown.
The broadband nonlinear response opens the way to the coherent control of the different
spectral components of the single incident beam in order to manipulate the second-order
nonlinear emission [18].
of the spectrum while keeping fixed
As a proof-of-principle of suchthespectral
intensity of the red band
manipulation, leadsa to
we used a linearincident
two-band increase of
SFG spectrum,
vs. blue containing
band power (Fig. 13), as
well-separated butexpected
coherent from
“red” Eq. 5. Dueparts
and “blue” to the
(seeintrinsic absence
inset of Fig. 5), of
ensemble averaging,
in order to excite and thissubsequently
result can be considered
filter as an improved
out a non-degenerate version of
sum-frequency (SF)Kurtz powder
signal. If
method (Kurtzassume
we simply & Perry, two1968),
average now appliedfrequencies
excitation on a single "R100-nm
and "Bsize nonlinear
in the nanoparticle.
single incident beam,
To conclude this issection,
the SF signal due to awe have shown
nonlinear that of
polarization phase-compensated
the form [17]: 13-fs infrared pulses gen-
erated through the application of a genetic algorithm allow one to gain about one order of
magnitude (as compared to standard 100-fs pulses with the same average intensity) in the
#106778 - $15.00 USD Received 23 Jan 2009; revised 16 Feb 2009; accepted 16 Feb 2009; published 9 Mar 2009
SHG emission rate of a non-centrosymmetric single KTP nanocrystal. This leads to efficient
(C) 2009 OSA 16 March 2009 / Vol. 17, No. 6 / OPTICS EXPRESS 4656
optical detection of smaller nanocrystals which appear only when shorter pulses are used.
Since there is no phase-matching limitation on the bandwidth of the excitation pulses, even
shorter (sub-10 fs) pulse scan can be used to even further enhance the second harmonic sig-
nal. Optical autocorrelation from a nanocrystal with an AFM measured size of about 100 nm
has been recorded, showing that pulse characterization can be obtained even on SHG-active
nanoparticles of sub-wavelength size.
 186 Nanocrystals

Fig. 13. Coherent sum frequency generation from a single nano-KTP vs the intensity of the
Fig. 5. Coherent sum frequency generation from a single nano-KTP vs the intensity of the blue
blue band. Inset: corresponding pulse spectrum with two distinctive parts "red" and "blue".
band.TheInset: corresponding pulse spectrum with two distinctive parts “red” and “blue”. The
nonlinear response is detected in the sum-frequency band with an interference filter trans-
nonlinear response is detectedshown
mission (spectral transmission in the sum-frequency
in red). Note that in theband
latter with
case theanwavelength
interference
scale filter
transmission
has been (spectral
multipliedtransmission
by two. shown in red). Note that in the latter case the wavelength
scale has been multiplied by two.

Conclusion
6. and prospects
Conclusion
In summary, we have reported the observation and the characterization of nanometric-sized
e have shown that phase-compensated 13-fs infrared pulses generated through t
crystals extracted by centrifugation from KTP powder. For a well-isolated single nanocrys-
plication of
tal, a genetic
in situ algorithm
AFM analysis allow
of its size one of
and analysis toitsgain about one
second-harmonic order
emission of magnitude (
properties
mpared to have
100-fsbeen performed.
pulses with The highly efficientaverage
the same nonlinear response leads to
intensity) inthe
theemission
SHGofemission
a large rate o
number of SHG photons in a photostable and blinking-free manner due to non-resonant co-
ncentrosymmetric single KTP nanocrystal. This leads to efficient optical detection
herent interaction. By recovering the radiation pattern from the recorded defocused images,
maller nanocrystals
we retrieve which
the in situappear only when
three-dimensional shorter
crystal pulses
orientation. are used.
Solution-based Sincesynthe-
chemical there is no pha
sis of KTP nanocrystals with a monodisperse size controlled by
atching limitation on the bandwidth of the excitation pulses, even shorter (sub-10 capping agents,(Biswas et al., fs) puls
2007) now under way, should lead to optimized KTP nanocrystallites and to a more accurate
n be used estimate
to even further enhance the second harmonic signal. Autocorrelation
of the size-detection threshold. It also opens the way to surface functionalization of from
nocrystal with an AFM Fully
these nanocrystals. measured sizenano-KTPs
characterized of about are100 nmforhas
attractive been recorded,
the development of novel showing th
schemes of nonlinear microscopy. Moreover, due to the
lse characterization can be obtained even on SHG-active nanoparticles of non-resonant interaction, they cansub-waveleng
be
envisioned for probing the localized electromagnetic field enhancement that appears at the
ze. This, coupled to amplitude
apex of a metallic andet al.,
tip(Bouhelier phase
2003) measurements basedby on
or is randomly generated nanoparticle
granular metallic nonline
mission [15], opens the way to a full near-field characterization of ultra-short pulses.
The broad spectrum of 10-fs pulses can also be tailored to manipulate the non-resona
nlinear second-order excitation process and extract e.g. a nondegenerate sum-frequen
ixing signal generated in a single nanoparticle. In association with polarization spect
odulation [20], this will lead to direct measurement of each of the nanoparticle nonline
efficients. This work can be extended to the optical nonlinear response of oth
 KTiOPO4 single nanocrystal for second-harmonic generation microscopy 187

structures(Stockman et al., 2004) while avoiding any detrimental quenching effect(Buchler et

al., 2005; Carminati et al., 2006) induced by the metallic interface. The coherent character of the
second harmonic emission from a single KTP nanocrystal have also been demonstrated in a
balanced homodyne detection scheme, which permit nanoparticle detection with high signal-
to-noise ratio and phase sensitivity. Furthermore, broadband femtosecond pulses have been
successfully employed for studying of a single KTP nanoparticle, which open a bright perpec-
tive for spectral manipulation for coherent control of the nonlinear process at the nanoscale.

Acknowledgements
The authors are grateful to Dominique Lupinski and Philippe Villeval from Cristal Laser for
initial nanoKTP material, Thierry Gacoin, Sandrine Perruchas, Cédric Tard and Géraldine
Dantelle for the fabrication of the nanocrystal solution, Joseph Lautru for sample preparation.
We acknowledge Jean-Pierre Madrange, André Clouqueur for technical assistance. This work
is a fruitful collaboration with Nicolas Sandeau, Sophie Brasselet, Véronique Le Floc’h, Joseph
Zyss, Chunyuan Zhou, Yannick de Wilde, Yannick Dumeige, François Treussart, François
Marquier, Jean-Jacques Greffet, Rémi Carminati. This work is supported by AC Nanosience
research program, CNano IdF research.

7. References
T. Baumert, T. Brixner, V. Seyfried, M. Strehle, and G. Gerber, "Femtosecond pulse shaping by
an evolutionary algorithm with feedback," Appl. Phys. B 65, 779-782 (1997).
S. K. Biswas, A. Pathak, P. Pramanik, "Synthesis of Nanocrystalline KTiOPO4 Powder by
Chemical Method" J. Am. Ceram. Soc. 90, 1071 (2007)
M. Bohmer, J. Enderlein, "Orientation imaging of single molecules by wide-field epifluores-
cence microscopy", J. Opt. Soc. Am. B 20, 554 (2003).
C. F. Bohren, D. R. Huffman, Absorption and Scattering of Light by Small Particles ; Wiley-VCH:
Weinheim, Germany, 2004.
L. Bonacina, Y. Mugnier, F. Courvoisier, R. Le Dantec, J. Extermann, Y. Lambert, V. Boutou,
C. Galez, J.-P. Wolf, "Polar Fe(IO3)3 nanocrystals as local probes for nonlinear mi-
croscopy", Appl. Phys. B 87, 399 (2007).
A Bouhelier, M. Berverhuis, and L. Novotny, "Near-Field Second-Harmonic Generation In-
duced by LocalField Enhancement," Phys. Rev. Lett. 90, 013903.1-013903.4 (2003).
R. W. Boyd, in Nonlinear Optics, Academic Press, 2nd ed., (2003).
the S. Brasselet, V. Le Floc’h, F. Treussart, J.-F. Roch, J. Zyss, E. Botzung-Appert, and A. Ibanez, "In
(as situ diagnosticsof the crystalline nature of single organic nanocrystals by nonlinear
microscopy" Phys. Rev. Lett. 92, 207401-204405 (2004).
fa B. Broers, L. D. Noordam, and H. B. van Linden van den Heuvall, "Diffraction and focusing
of of spectralenergy in multiphoton processes," Phys. Rev. A 46, 2749-2756 (1992).
ase X. Brokmann, M.-V. Ehrensperger, J.-P. Hermier, A. Triller M. Dahan, "Orientational imaging
and tracking of single CdSe nanocrystals by defocused", Chem. Phys. Lett. 406, 210
ses (2005).
m a B. C. Buchler, T. Kalkbrenner, C. Hettich, V. Sandoghdar, "Measuring the Quantum Efficiency
of the Optical Emission of Single Radiating Dipoles Using a Scanning Mirror",Phys.
hat Rev. Lett. 95, 063003 (2005).
gth R. Carminati, J.-J. Greffet, C. Henkel, J.-M. Vigoureux, "Radiative and non-radiative decay of
ear a single molecule close to a metallic nanoparticle", Opt. Commun. 261, 368 (2006).

ant
ncy
tral
ear
her
188 Nanocrystals

J. Chen, S. Machida, and Y. Yamanoto, "Simultaneous measurement of amplitude and phase

in surface second-harmonic generation", Opt. Lett. 23, 676 (1998).
T. A. Driscoll, H. J. Hoffman, R. E. Stone, P. E. Perkins, "Efficient second-harmonic generation
in KTP crystals", J. Opt. Soc. Am. B 3, 683 (1986).
N. Dudovich, D. Oron, and Y. Silberberg, "Single-pulse coherently controlled nonlinear Raman
spectroscopy and microscopy," Nature 418, 512-514 (2002).
J. Extermann, L. Bonacina, F. Courvoisier, D. Kiselev, Y. Mugnier, R. Le Dantec, C. Glez, and
J. -P. Wolf,"Nano-FROG: Frequency resolved optical gating by a nanometric object,"
Opt. Express 16, 10405-10411(2008)
J. B. Guild, C. Xu, and W. W. Webb, "Measurement of group delay dispersion of high numerical
apertureobjective lenses using two-photon excited fluorescence," Appl. Opt. 36, 397-
401 (1997).
G. Hansson, H. Karlsson, S.Wang, F. Laurell, "Transmission Measurements in KTP and Iso-
morphic Compounds", Appl. Opt. 39, 5058 (2000).
C. Hsieh, R. Grange, Y. Pu, D. Psaltis, "Three-dimensional harmonic holographic microcopy
using nanoparticles as probes for cell imaging," Opt. Express 17, 2880 (2009)
J. C. Johnson, H. Yan, R. D. Schaller, P. B. Petersen, P. Yang, R. J. Saykally, "Near-Field Imag-
ing of Nonlinear Optical Mixing in Single Zinc Oxide Nanowires", Nano Lett 2, 279
(2002).
A. V. Kachynski, A. N. Kuzmin, M. Nyk, I. Roy and P. N. Prasad, ?Zinc Oxide Nanocrystals
for Nonresonant Nonlinear Optical Microscopy in Biology and Medicine,? J. Phys.
Chem. C, 112, 10721 (2008)
K. Kemnitz, K. Bhattacharyya, J. M. Hicks, G. R. Pinto, K. B. Eisenthal, T. F. Heinz, "The phase
of second-harmonic light generated at an interface and its relation to absolute molec-
ular orientation", Chem. Phys. Lett. 131, 285 (1986).
S. Kurtz and T. Perry, "A powder technique for the evaluation of nonlinear optical materials,"
IEEE J.Quantum Electron. 4, 333-333 (1968).
F. Laurell, M.G. Roelofs, W.Bindloss, H.Hsiung, A. Suna, and J.D. Bierlein, "Detection of fer-
roelectric domain reversal in KTiOPO4 waveguides," J. Appl. Phys. 71, 4664 (1992)
V. Le Floc?h, S. Brasselet, J.-F. Roch, J. Zyss, "In Situ Diagnostics of the Crystalline Nature of
Single Organic Nanocrystals by Nonlinear Microscopy" J. Phys. Chem. B 107, 12403
(2003).
L. Le Xuan, S. Brasselet, F. Treussart, J.-F. Roch, F. Marquier, D. Chauvat, S. Perruchas, C.
Tard, T. Gacoin, "Balanced homodyne detection of second-harmonic generation from
isolated subwavelength emitters", Appl. Phys. Lett. 89, 121118 (2006).
L. Le Xuan, C. Zhou, A. Slablab, D. Chauvat, C. Tard, S. Perruchas, T. Gacoin, P. Villeval,
and J. -F.Roch, "Photostable Second-Harmonic Generation from a Single KTiOPO4
Nanocrystal for NonlinearMicroscopy," Small 4, 1332-1336 (2008).
J. P. Long, B. S. Simpkins, D. J. Rowenhorst, P. E. Pehrsson, "Far-field Imaging of Optical
Second-Harmonic Generation in Single GaN Nanowires", Nano Lett. 7, 831 (2007).
V. V. Lozovoy, I. Pastirk, and M. Dantus, "Multiphoton intrapulse interference. IV. Ultrashort
laser pulsespectral phase characterization and compensation," Opt. Lett. 29, 775-777
(2004).
D. Meshulach and Y. Silberberg, "Coherent quantum control of two-photon transitions by a
femtosecondlaser pulse," Nature 396, 239-242 (1998).
L. Moreaux, O. Sandre, J. Mertz, "Membrane imaging by second-harmonic generation mi-
croscopy", J. Opt. Soc. Am. B 17, 1685 (2000).
KTiOPO4 single nanocrystal for second-harmonic generation microscopy 189

Y. Nakayama, P. J. Pauzauskie, A. Radenovic, R. M. Onorato, R. J. Saykally, J. Liphardt, P.

Yang, "Tunable nanowire nonlinear optical probe", Nature 447, 1098, (2007).
G. H. Patterson, D.W. Piston, "Photobleaching in Two-Photon Excitation Microscopy", Bio-
phys. J. 78, 2159 (2000).
P. Rejmankova, J. Baruchel, P. Villeval, C. Saunal, "Characterization of large KTiOPO4 flux
grown crystals by synchrotron radiation topography", J. of Crystal Growth 180, 85
(1997).
N. Sandeau, L. Le Xuan, C. Zhou, D. Chauvat, J.-F. Roch, S. Brasselet, "Defocused imaging
of second harmonic generation from a single nanocrystal", Opt. Express 15, 16051
(2007).
Y. Shen, P. Markowicz, J.Winiarz, J. Swiatkiewicz, P. N. Prasad, "Nanoscopic study of second-
harmonic generation in organic crystals with collection-mode near-field scanning op-
tical microscopy", Opt. Lett. 26, 725 (2001).
Y. R. Shen, "Surface properties probed by second-harmonic and sum-frequency generation",
Nature 337, 519 (1989).
M. I. Stockman, D. J. Bergman, C. Anceau, S. Brasselet, J. Zyss, "Enhanced Second-Harmonic
Generation by Metal Surfaces with Nanoscale Roughness: Nanoscale Dephasing, De-
polarization, and Correlations", Phys. Rev. Lett. 92, 057402 (2004).
R. Stolle, G. Marowsky, E. Schwarzberg, G. Berkovic, "Phase measurements in nonlinear op-
tics", Appl. Phys. B, 63, 491 (1996) and Refs. therein.
R. L. Sutherland, Handbook of Nonlinear Optics , Dekker, New York, 1996 pp. 267.
F. Treussart, E. Botzung-Appert, N. T. Ha-Duong, A. Ibanez. J.-F. Roch, R. Pansu, "Second
Harmonic Generation and Fluorescence of CMONS Dye Nanocrystals Grown in a
Sol-Gel Thin Film", Chem. Phys. Chem. 4, 757 (2003).
F. Treussart, V. Jacques, E Wu, T. Gacoin, P. Grangier, J.-F. Roch, "Photoluminescence of single
colour defects in 50ănm diamond nanocrystals", Physica B 376, 926 (2006).
H. Vanherzeele, and J. D. Bierlein, "Magnitude of the nonlinear-optical coefficients of
KTiOPO4 ", Opt. Lett. 17,982-985 (1992).
W. S. Warren, R. Rabitz, and M. Daleh, "Coherent Control of Quantum Dynamics: The Dream
Is Alive,"Science 259, 1581-1589 (1993).
A. M. Weiner, "Femtosecond pulse shaping using spatial light modulators," Rev. Sci. Instrum.
71, 1929-1960 (2000).
P. Wnuk and C. Radzewicz, "Coherent control and dark pulses in second harmonic genera-
tion," Opt.Commun. 272, 496-502 (2007).
P. Wnuk, L. Le Xuan, A. Slablab, C. Tard, S. Perruchas, T. Gacoin, J-F. Roch, D. Chauvat, and
C. Radzewicz, "Coherent nonlinear emission from a single KTP nanoparticle with
broadband femtosecond pulses", Optics Express, 17, 4652 (2009)
P. Xi, Y. Andegeko, L. R. Weisel, V. V. Lozovoy, and M. Dantus, "Greater Signal and LessPho-
tobleaching in Two-Photon Microscopy with Ultrabroad Bandwidth Femtosecond
Pulses," Opt.Commun. 281, 1841-1849 (2008).
H. P. Yuen, V. W. S. Chan, "Noise in homodyne and heterodyne detection," Opt. Lett. 8, 177
(1983).
F. C. Zumsteg, J. D. Bierlein, T. E. Gier, "Kx Rb1− x TiOPO4 : A New Nonlinear Optical Material,"
J. Appl. Phys. 47, 4980 (1976).
190 Nanocrystals
Zeolite nanocrystals - synthesis and applications 191

X8

Zeolite nanocrystals - synthesis

and applications
Teruoki Tago and Takao Masuda
Division of Chemical Process Engineering, Faculty of Engineering, Hokkaido University
N13W8, Kita-Ku Sapporo, Hokkaido, 060-8628, Japan

1. Introduction
Zeolites, crystalline aluminosilicate materials, possess 3-dimensionaly connected framework
structures constructed from corner-sharing TO4 tetrahedra, where T is any tetrahedrally-
coordinated cation such as Si and Al. These framework structures are composed of n-rings,
where n is the number of T-atoms in the ring (e.g. 4-, 5-, and 6-rings), and large pore
openings of 8-, 10-, and 12-rings are framed by these small rings. Figure 1 shows the pore
sizes and framework structures of typical zeolites. The sizes of the intracrystalline pores and
nanospaces, depending on the type of zeolite providing the framework, are close to the
molecular diameters of lighter hydrocarbons. Moreover, strong acid sites exist on the
nanopore surfaces, enabling the zeolites to be used as shape-selective catalysts in industrial
applications such as fluid catalytic cracking of heavy oil, isomerization of xylene and
synthesis of ethyl-benzene. However, compared to the sizes of micropores exhibiting a
molecular-sieving effect, the crystal sizes of zeolites are very large, approximately 1–3 µm.
When the zeolite is used as a shape-selective catalyst, the diffusion rates of reactant
molecules within the zeolite crystals are lower than the intrinsic reaction rates. This
resistance to mass transfer limits the reaction rates and low selectivity of intermediates.
Moreover, since effective active sites (acid sites) for catalytic reactions are distributed on the
internal surfaces of the main channels and the external surfaces of the crystal, the pore
mouths are easily plugged due to coke deposition, leading to short lifetimes for the catalysts.

Micropore Mesopore
≈
C10H8
C6H6
C2H6
CO2

0.1 nm 1 nm Mesoporous
Mesoporous silica
silica 10 nm 50 nm
≈

MOR
MOR
LTA
LTA MFI
MFI
FAU
FAU
BEA
BEA
Fig. 1. Zeolite structure, pore size and molecular diameter of hydrocarbons.
192 Nanocrystals

Faster mass transfer is required to avoid these serious problems, and two primary strategies
have been proposed; one is the formation of meso-pores within zeolite crystals (Groen et al.,
20004a, 2004b; Ogura et al., 2001), and the other is the preparation of nano-crystalline
zeolites (Tsapatsis et al., 1996; Mintova et al., 1999, 2002, 2006; Ravishankar et al., 1998;
Grieken et al., 2000; Hincapie et al., 2004; Song et al., 2005; Larlus et al., 2006; Morales-
Pacheco et al., 2007; Kumar et al., 2007).
In nano-crystalline zeolites, the diffusion length for reactant hydrocarbons, assignable to the
crystal size, decreases, and the external surface area of the crystal increases as the crystal
size decreases. The increase in external surface area and the decrease in diffusion resistance
are effective for improving catalytic activity in gas-solid and liquid-solid heterogeneous
catalytic reactions. Because of these favorable properties of nano-crystalline zeolites for
catalytic reactions, the preparation methods for several types of zeolite nanocrystals have
been reported and reviewed (Tosheva & Valtchev, 2005; Larsen, 2007).
In this chapter, a method for preparing nano-crystalline MFI and MOR zeolites in a solution
consisting of a surfactant, organic solvent, and water is introduced (denoted as the emulsion
method hereafter). Nano-crystalline zeolite is expected to be a promising material for
increasing the outer surface area as well as decreasing the diffusion resistance of the organic
reactant within the micropores, thereby improving the catalytic activity and lifetime.

2. Synthesis of nano-crystalline zeolites in water/surfactant/organic solvent

Usually, zeolites are prepared in an alkaline water solution containing Si and Al sources,
and alkaline metal ions (sodium or potassium). In the synthesis of some types of zeolite, an
organic structure directing agent (OSDA), such as an ammonium alkyl cation, is also
necessary to form the zeolite framework. The water solutions containing these inorganics
and OSDA are poured into a Teflon-sealed stainless steel bottle and heated to a desirable
temperature.
In some types of zeolites (e.g. MFI and MOR), crystal nucleation occurs first, followed by the
growth of the zeolite, with the nucleation simultaneously continuing to occur during the
growth stage. One method to prepare nano-crystalline zeolite is the stoppage of the
hydrothermal treatment when the nucleation occurs. However, because unreacted Si and Al
species still remain in the solution, an amorphous phase tends to form on the surface of the
zeolite crystals. Another method is increasing the nucleation rate by increasing the
concentrations of the Si and Al sources. With increasing nucleation rate, the crystal size of
the zeolite decreases. However, the zeolite thus obtained has a broad size distribution due to
simultaneous nucleation and crystal growth during the hydrothermal treatment.
Accordingly, in order to obtain nano-crystalline zeolites, it is important to separate the
nucleation and growth stages. Recently, there has been growing interest in the synthesis of
nano-crystalline zeolites with the addition of surfactants, including
water/surfactant/organic mixture (Kuechl et al., 2010; Naik et al., 2002; Lee et al., 2005; Tago
et al., 2004, 2009a, 2009b) (emulsion method), with research focused on size and morphology
control, because these parameters affect the performance in applications, such as separation
and catalytic reactions. Because surfactant molecules that have hydrophobic and
hydrophilic organic groups in the molecules are adsorbed on a solid surface in a solvent, the
interface energy difference between the solid surface and solvent can be reduced, leading to
Zeolite nanocrystals - synthesis and applications 193

enhancement of nucleation of metal and/or metal oxide nano-particles. In the emulsion

method, the surfactant adsorption effect is applied to prepare nano-crystalline zeolites.
In general, there are four types of surfactants, anionic, cationic, nonionic, and bipolar
surfactants. A nonionic surfactants; Polyoxyethylene(15)oleylether (O-15),
polyoxyethylene(15)nonylphenylether (N-15), and poly(oxyethylene)(15)cethylether (C-15),
and ionic surfactants, sodium bis(2-ethylhexyl) sulfosuccinate (AOT) and cetyltrimethyl
ammonium bromide (CTAB), are employed. (Figure 2) Cyclohexane or 1-hexanol is used as
an organic solvent.

nonionic surfactant
polyoxyethylene(15)nonylphenylether :(NP-15)

C9H19― ―(OC2H4)nOH

Polyoxyethylene(15)oleylether : (O-15)
C9H18= C9H17 －(OC2H4)nOH

poly(oxyethylene)(15)cethylether : (C-15)
C16H33－(OC2H4)nOH

ionic surfactant
cetyltrimethyl ammonium bromide : (CTAB)
C16H33N(CH3)3Br

sodium bis(2-ethylhexyl) sulfosuccinate : (AOT)

H

CH3 (CH2)3CH(C2H5)3CH2OCOCH2C OSO3-Na+

CH3 (CH2)3CH(C2H5)3CH2OC
=

O
Fig. 2. Typical molecular structures of surfactants.

In the emulsion method, two solutions are prepared, one is a water solution containing Si
and Al sources obtained by hydrolyzing metal alkoxide with a dilute OSDA/water solution,
and the other is a surfactant/organic solvent. The concentrations of the Si and Al sources in
the water solution and the molar ratio of Si to the OSDA as well as types of the OSDA are
very important factors in preparing zeolite crystals. In the case of MFI,
tetrapropylammoniumhydroxide (TPAOH) is used as an OSDA, and
tetraethylammoniumhydroxide (TEAOH) is used for the preparation of MOR zeolite.
194 Nanocrystals

Moreover, two additional parameters, the concentration of surfactant in the organic solvent
and the molar ratio of water to surfactant, are important in the emulsion method. These
parameters affect the morphology and crystal sizes.

Surfactant/Organic Solvent
Alkoxide(SiO2, Al2O3)
OSDA/H2O

Stirred at 50℃ for 1h

Hydrothermal treatment
Heated to 100℃～150℃
for 24 h~120 h

Washed, Dry, Calcination

Zeolite Nanocrystal
Fig. 3. Preparation procedure for nano-crystalline zeolite by the emulsion method.

The water solution thus obtained is added to the surfactant-organic solvent, and the mixture is
magnetically stirred at 323 K for 1 h. The water/surfactant/organic solvent mixture is then
placed in a Teflon-sealed stainless steel bottle, heated to 373 ~ 423 K, and held at the desired
temperature for 12 ~ 120 h with stirring to yield zeolite nanocrystals. The precipitate thus
obtained is centrifuged, thoroughly washed with propanol, dried at 100 ºC overnight, and
calcined under an air flow at 500 ºC to remove the surfactant and the OSDA molecules. After
the air calcination, the weight of the sample is measured to calculate the zeolite yield (Figure 3).

2.1 Effect of ionicity of surfactant on preparation of MFI zeolite nanocrystals

The morphologies of the silicalite-1 (MFI zeolite) nanocrystals prepared using various
surfactants were investigated under conditions of TOES concentration of 0.63 or 2.73 mol/L,
hydrothermal temperature of 140 or 100 ºC and concentration of surfactant in organic
solvent (solution B) of 0.50 mol/L. Figures 4 and 5 show X-ray diffraction (XRD) patterns
and FE-SEM images of the obtained samples, respectively.
In the XRD pattern for the sample prepared in the AOT/cyclohexane solution, peaks
corresponding to sodium sulfate rather than silicalite-1 are seen. Moreover, the sample
showed an irregular morphology. In this case, the TPA-OH molecules cannot act as an
OSDA in the synthetic solution. This result is ascribed to the fact that the surfactant of AOT
and the OSDA of TPA-OH have opposite ionic charges.
Zeolite nanocrystals - synthesis and applications 195

When using CTAB/1-hexanol, the silicalite-1 crystals, which are approximately 1.0 µm in
size, are embedded in the amorphous SiO2 as seen from SEM observations. The coexistence
of silicalite-1 crystals and amorphous SiO2 is revealed by the X-ray diffraction analysis.
Since the pH of the synthetic solution is alkaline, the surface of the SiO2 produced by
hydrolysis of TEOS has a negative charge. Accordingly, it is considered that CTAB and
TPA-OH are independently adsorbed on the surface of SiO2 because of their cationic
ionicity. Therefore, the SiO2 species that adsorb TPA-OH and CTAB change into silicalite-1
crystals and amorphous SiO2, respectively.
In the case of the nonionic surfactants, C-15, NP-15 and O-15 (the nonionic
surfactant/cyclohexane system), the XRD patterns of the samples showed peaks
corresponding to pentasile-type zeolite (the reference silicalite-1), and mono-dispersed
silicalite-1 nano-crystals were obtained, as seen from SEM observations. In contrast, the
silicalite-1 crystal prepared in water (without a surfactant) shows a heterogeneous structure
with smaller crystals (diameter of approximately 30 nm) on a larger one (approximately 120
nm), indicating that nucleation, crystallization and crystal growth occurred simultaneously.
These results indicate that the ionicity of the hydrophilic groups in the surfactant molecules
plays an important role in the formation and crystallization processes of the silicalite-1
nanocrystals. Since the aggregation of the silicalite-1 nuclei are inhibited by the adsorbed
surfactants on the surface during hydrothermal treatment, mono-dispersed nanocrystals can
be prepared.

H2O/AOT/C6H12

H2O/CTAB/C6H13OH
intensity

H2O/C-15/C6H12

H2O/NP-15/C6H12

H2O/O-15/C6H12

Reference Silicalite

10 20 30 40 50

2θ / degree
Fig. 4. X-ray diffraction patterns of samples prepared in water/surfactant/organic solvent.
Effects of ionicity of the surfactant on crystallinity of MFI zeolite (Silicalite-1).
196 Nanocrystals

Nano-crystalline zeolite has a high external surface area, which is very important for use as
heterogeneous catalysts and as seed crystals. To evaluate the crystallinity of the nanocrystals,
X-ray diffraction patterns and Raman spectroscopy can be used. The former is based on
elastic scattering of X-rays from structures that have long range order (zeolite crystal
structure, types of zeolite), and the latter method can elucidate middle-range order (types of
Si-O rings). Accordingly, the Raman spectra measurements are useful to evaluate the zeolite
crystallinity in detail (Dutta et al., 1991; Li et al., 2000, 2001). Figure 6 shows the Raman
spectrum of the silicalite-1 nanocrystals. The peaks in the range of 300 – 650 cm−1 are
indicative of the type of silicon-oxygen rings present in the structure of zeolite. The
spectrum also shows peaks around 380, 430 and 470 cm−1, which correspond to the five-, six-
and four-member rings, respectively. These peaks are in good agreement with the peaks of
the reference silicalite-1 prepared in water. In the structure of the MFI-type zeolite,
continuous chains of five-member rings are connected by the four- and six-member rings.
These results indicated that the surfaces of the nanocrystals with a diameter of
approximately 120 and 80 nm are also well-crystallized without amorphous SiO2.

Fig. 5. FE-SEM micrographs of samples prepared in water/surfactant/organic solvent.

Effects of ionicity of the surfactant on crystallinity of MFI zeolite (Silicalite-1).
Zeolite nanocrystals - synthesis and applications 197

6-rings

5-rings 4-rings

Intensity

Silicalite-1 nanocrystals

Reference Silicalite

200 300 400 500 600 700 800

Raman Shift / cm–1
Fig. 6. Raman spectrum of the nanocrystals.

2.2 Preparation of MOR zeolite by the emulsion method

In the preparation of MFI zeolite nanocrystals by the emulsion method, since the non-ionic
surfactant is revealed to be the appropriate surfactant to prepare nano-crystalline zeolite,
this method is applied to prepare MOR zeolite nanocrystals as well. First, the effects of the
hydrothermal time on the crystallinity of MOR zeolite prepared in water/O-15/cyclohexane
are examined. The concentrations of Si in a water solution and a surfactant in cyclohexane
are 0.75 and 0.5 mol/L, respectively, and the hydrothermal temperature is 423 K. Figure 7
shows the X-ray diffraction patterns of MOR zeolite prepared using the emulsion method
with different hydrothermal times. The X-ray diffraction patterns of samples prepared in the
water solution without a surfactant (conventional method) are also shown in the figure for
comparison.
The samples prepared in the water solution show a broad halo pattern of SiO2 at
hydrothermal times of 72 and 96 h, and show peaks corresponding to MOR zeolite as well
as BEA zeolite (beta-type zeolite) at 120 h. In contrast, as compared to the sample prepared
in the water solution, though an amorphous pattern is observed from the sample prepared
at a hydrothermal time of 72 h, the samples at hydrothermal times of 96 and 120 h show
peaks corresponding to MOR zeolite. Moreover, peaks ascribable to other types of zeolite,
e.g. MFI and BEA, cannot be detected. Because hydrolysis of the Si (tetraethylorthosilicate)
and Al (aluminum-tri-isopropoxide) sources should be completed during the preparation of
water solution containing Si, Al, and the OSDA molecules (TEA-OH), the appearance of the
diffraction peaks corresponding to MOR zeolite after 96 h indicates that the MOR zeolite
precursors are prepared in the solution until approximately 72 h, followed by nucleation
and crystal growth of MOR zeolite. Accordingly, MOR and MFI zeolites can be prepared in
water/surfactant/organic solvent. Sizes control by the emulsion method is next examined.
198 Nanocrystals

Prepared in H2O/O-15/C6H12
(emulsion method)

120 h

96 h
72 h
Intensity

( BEA) Prepared in H2O（Conventional method）

120 h
96 h
72 h

Reference MOR

10 20 30 40 50
angle 2 / degree
Fig. 7. X-ray diffraction patterns of MOR zeolite prepared in the emulsion method with
different hydrothermal times.

2.3 Crystal size control

In a catalytic reaction using a zeolite, the reaction proceeds on acid sites located on the outer
surface of the crystal as well as the inside pore surface. The crystalline zeolite affects the
outer surface area and the diffusion length, assignable to the crystal size, for reactant
hydrocarbons within the crystal. Accordingly, it is very important to develop preparation
methods for zeolite nanocrystals with different crystal sizes.
In the emulsion method, it is considered that non-ionic surfactants adsorbed on the surfaces
of zeolite precursors induce the formation of zeolite nuclei, enhancing the nucleation of
zeolite. Moreover, the surfactant concentration in the solution can be easily changed.
Accordingly, the effects of varying surfactant concentration [O-15] on the morphology of
MOR zeolite were examined. Figure 8 shows FE-SEM images of the obtained samples. The
Si/Al ratio is 12.5. The surfactant concentrations were changed in the range from 0.5 to 0.75
Zeolite nanocrystals - synthesis and applications 199

mol/L. Interestingly, the crystal size and morphology depended on the surfactant
concentration, regardless of the same concentrations of the Si and Al sources, and template
in the water solution. MOR zeolite nanocrystals with an average size of approximately 80
nm were obtained at a surfactant concentration of 0.5 mol/L. As the concentration increased
to 0.65 mol/L, growth of MOR zeolite with column-like morphology is observed, and the
crystal size reached ~1.0 μm at 0.75 mol/L. Figure 9 shows NH3–TPD profiles and N2
adsorption–desorption isotherms of the obtained samples, respectively, at surfactant
concentrations of 0.5 and 0.75 mol/L. The NH3–TPD profiles of these sample shows NH3
desorption peaks above 600 K, ascribable to desorption of NH3 from strong acid sites of
MOR zeolites. Moreover, these MOR zeolites show almost the same peak area. The MOR
zeolites also show almost the same N2 adsorption–desorption isotherms, indicating that
these zeolites exhibit almost the same surface area. Accordingly, these MOR zeolites possess
almost the same number of acid sites, regardless of the crystal morphology.
As discussed above, it is revealed that the surfactant in the synthetic solution induces the
nucleation and crystallization of zeolite as compared to the conventional preparation
method (without a surfactant). Moreover, the surfactant concentrations in the solution were
found to influence crystal growth. The crystal morphology changed from the nano–
crystalline to large columnar crystals with increasing surfactant concentration.

Fig. 8. FE-SEM micrographs of MOR zeolite with different crystal sizes. (Tago et al., (2009b))
200 Nanocrystals

15

dq/dT / mol kg− 1 K− 1

[O-15]=0.5 mol/L
10
[O-15]=0.75 mol/L

Si/Al=12.5
0
300 400 500 600 700 800 900
Temperature [K]
150
V / ml g− 1

100
[O-15]=0.5 mol/L
[O-15]=0.75 mol/L
50

Si/Al=12.5
0
0.0 0.2 0.4 0.6 0.8 1.0
P/P0 / −
Fig. 9. NH3–TPD profiles and N2 adsorption–desorption isotherms of the MOR zeolite with
different crystal sizes. (Tago et al., (2009b))

2.4 Mechanism
In order to consider the mechanism of zeolite nanocrystals formation in
water/surfactant/organic solvent, it is necessary to investigate the relationship between the
ionicity of the surfactant and the template molecule, and the ionic charges of the surface of
SiO2 and/or zeolite precursor in detail.
In the method, the cyclohexane used as the organic solvent contributes to stabilization of the
hydrophobic group of the surfactant. The effects of the ionicity of the hydrophilic group of
the surfactant on the morphology and crystallinity of the obtained silicalite-1 zeolite
nanocrystals are described above.
In an anionic surfactant (e.g. sodium bis(2-ethylhexyl) sulfosuccinate, AOT), because the
surfactant of AOT has opposite ionic charges to the OSDA molecules, the OSDA cannot act as a
structure-determining agent in the synthetic solution. Accordingly, in the relationship between
the ionicity of the surfactant and OSDA molecules, a surfactant without electrostatic affinity to
the OSDA molecules is needed for the preparation of zeolite crystals. In a cationic surfactant (e.g.
cetyltrimethyl ammonium bromide, CTA-Br), the molecular structure is similar to that of the
OSDA. Moreover, since the surface charge of the SiO2 and/or zeolite precursor is negative, the
surfactant molecules of CTA+ with cationic hydrophilic group can be adsorbed on the surface,
followed by the formation of amorphous SiO2. In contrast, in the case of nonionic surfactants,
MOR as well as MFI zeolite nanocrystals can be obtained.
Zeolite nanocrystals - synthesis and applications 201

chain）
up p oly -o xyethylene Hydrophilic group
gro （
Hydrophilic
(CH2CH2O) (CH2CHO)
O) (CHCH2O)
(CHCH 2 H

H
H

H
O

O
H OH

O
H
HO H H

H
O H
H
Si O Si Al Si O Si O
O O O O O O
O O
Si Si Al Si
Si Si Si Si
Adsorption on hydrophilic Adsorption on hydrophobic
surface (–Si-OH) hydrating water

up
c gr o
y d r ophili
H
(CH2CHO) (CHCH2O)
H O) H H
C O O
H 2
H Si Si
(C O
O O O O
Si
Si Si Si Si

Direct adsorption on hydrophobic

surface (–Si-O-Si-)
Fig. 10. A schematic showing the possible relationship between the surfactant and the
zeolite surface. (Tago et al., (2009b))

A schematic figure showing the possible relationship between the surfactant and the zeolite
surface is shown in Fig. 10. The hydrophilic group of the surfactant O-15 is composed of
poly-oxyethylene chains, making it a non-ionic surfactant. The surface of the zeolite
precursor and crystal is composed of a hydrophobic surface (-Si-O-Si-) and hydrophilic
silanol groups (-O-Si-OH). The hydrophilic groups of the surfactant can be adsorbed on the
hydrophilic silanol groups, leading to stabilization of the silanol groups. On the
hydrophobic surface, hydrophobic hydration occurs in the conventional preparation
method without a surfactant (Burkett & Davis, 1994, 1995; De Moor, 1999a, 1999b). It is
considered that the hydrophilic groups of the surfactant are adsorbed onto the water
molecules hydrophobically-hydrated on the zeolite surface. Moreover, because the
hydrophilic property of the poly-oxyethylene chains in the surfactant is much smaller than
that in water, it is considered that the hydrophilic poly-oxyethylene chains can be directly
adsorbed onto the zeolite surface due to their higher affinity to the zeolite surface than water
molecules. These adsorbed surfactants can contribute to the stabilization of the surface of
zeolite precursor and crystals, which will influence the nucleation rate of zeolite, so that the
MOR zeolite nanocrystals can be obtained.
202 Nanocrystals

Three-dimensional cylindrical networks of the micellar structure composed of water and

surfactant molecules likely exist in water/O-15/cyclohexane solvent, wherein silica-based
cylindrical (Matsune et al., 2006) and spherical (Tago et al., 2002, 2003; Takenaka et al., 2007)
materials including metal and/or metal-oxide nano-particles are obtained at 323 K. These
networks lead to an increase in the viscosity of the solvent. Though the reaction temperature
of 323 K is different from the hydrothermal temperature, the viscosity of the zeolite
synthetic solution is thought to be high. It is considered that the viscosity of the solvent
affected the nucleation rate and diffusion of silicate species. However, since the yields of
zeolite are above 80%, the diffusion of silicate species in the solution is likely not a rate-
limiting step to form zeolite nanocrystals, and thus the viscosity affects the nucleation rate.
Moreover, as the surfactant concentration is increased, the excessive stabilization of the
precursors and the high viscosity of the solution lead to a decrease in the nucleation rate of
zeolite. Since the growth rate of the crystal is much higher than the nucleation rate,
crystalline growth of the MOR zeolite is dominant at high surfactant concentrations, so that
column-like crystals with a large size are obtained. Accordingly, the surfactant can play an
important role in the nucleation and crystal growth of the MFI and MOR zeolite prepared in
the water/surfactant/organic solvent (emulsion method), and the crystal size and
morphology can be controlled by the surfactant concentration.

3. Applications of nanocrystals
3.1 Heterogeneous catalytic reaction over zeolite nanocrystals
The acid sites in zeolites are located on the external and internal surfaces. In nano-crystalline
zeolite, the external surface area of the crystal increases and the diffusion length for reactant
hydrocarbons, assignable to the crystal size, decreases as the crystal size decreases. Due to
these favorable properties, nano-crystalline zeolites have been used as heterogeneous
catalysts.
Zeolites exhibit a molecular sieving effect for lighter hydrocarbons, hence, the hydrocarbon
molecules with sizes larger than the pore mouth size are mainly adsorbed on the external
surface. Accordingly, an increase in the external surface area is effective for increasing
catalytic activity. Song et al. (2004a, 2004b) reported the preparation of nano-crystalline MFI
zeolites and they evaluated the effects of crystal size on external surface area and adsorption
properties of toluene on the zeolite. Since the molecular size of toluene is almost the same as
the pore mouth size in MFI zeolite, the external surface area, which is a function of the
crystal size, affects the adsorption capacity for toluene. Zeolites with crystal sizes less than
100 nm have a higher adsorption capacity for toluene. Botella et al. (2007) have reported the
Beckmann rearrangement reaction over nanosized BEA zeolite where the reactions mainly
proceed on the external surface of zeolite. Serrano et al. (2005) have reported catalytic
cracking of polyolefins over nano-crystalline MFI zeolite. Because the molecular size of
polyolefins is larger than the pore size, cracking of polyolefins mainly occurs over the acid
sites on the external surface, while the formation of lighter olefins proceeds on the internal
pore surface. Jia et al. (2010) have reported dehydration of glycerol to produce acrolein over
nano-sized MFI zeolite. Since the hydrophilic glycerol molecules are strongly adsorbed on
the external surface of the zeolite, nano-crystalline MFI zeolite is effective for this reaction
due to the high surface area. In these reports listed above, the increase in the external surface
area of the zeolite nanocrystal is important to improve the catalytic activity and lifetime.
Zeolite nanocrystals - synthesis and applications 203

In contrast, Sakthivel et al. (2009) have reported the isomerization/cracking of hexane over
nano-sized BEA zeolite. Since the molecular size of hexane is smaller than the pore size of
MFI zeolite, the cracking reaction proceeds mainly on the acid sites located inside pores.
Moreover, the small crystal size of BEA zeolite leads to low diffusion resistance of the
reactant hexane and the products. Serrano et al. (2010) have reported epoxide rearrangement
reactions over MFI zeolite, where the high external surface area and low diffusion resistance
within the micropores due to the small zeolite crystal size enhance the rearrangement
reaction without diffusion restriction of the epoxide molecules.
MFI zeolites are effective catalysts to synthesize lighter olefins by the methanol-to-olefin
(MTO) reaction and acetone conversion to olefin. In these olefin syntheses, selective
formation of lighter olefins occurs due to the molecular sieving effect of the zeolite.
However, when the reaction of the hydrocarbon occurs over acid sites located on the
external surface of zeolite crystals, this results in a non-shape-selective reaction as well as
coke deposition, leading to short lifetimes for the catalyst. These undesirable phenomena
will be accelerated in nano-crystalline zeolites because of their large external surface areas.
Accordingly, a method for selective deactivation of acid sites located on the external surface
of zeolite crystals is desired in order to prevent these undesirable reactions. Masuda et al.
(2001) have reported a method called the “catalytic cracking of silane method”, in which
SiO2 units are formed selectively on the acid sites of the zeolite using organic silane
compounds. Tago et al. (2009c) have reported that the yields of lighter olefins such as
ethylene and propylene are increased over the selectively de-activated MFI zeolite.

3.2 Structured materials composed of zeolite nanocrystals

Since the nano-crystalline zeolites have large external surface areas, they are potential
candidates for seed crystals from which a structured material can be prepared. Moreover,
the micro- and meso-pores formed in nano-crystalline zeolites are characteristic of the
structured materials. Accordingly, design of the micro- and meso-pores is important.
Serrano et al. have reported the preparation of micro-, meso-, and macroscopic hierarchical
materials composed of nano-crystalline zeolites. They prepared hybrid zeolitic-mesoporous
materials with the properties of both MCM-41 and MFI zeolite using zeolite seeds by
assembling around cetyltrimetyl ammonium bromide micelles. Moreover, they have also
developed other methods to fabricate hierarchical materials using organosilane bonded
zeolite nanocrystals, where the degree of aggregation of the zeolite can be controlled.
Nano-crystalline zeolites have been used as seed crystals for zeolite membranes. Since the
pore spaces and pore sizes in the crystals represent a very important factor, affecting
membrane performances (Hasegawa et al., 2007), nano-crystalline zeolite crystals are useful
as seed crystals. In the preparation of zeolite membranes, after the deposition and/or
seeding of the nano-crystalline zeolites on the base materials, secondary growth of the
zeolite was carried out by a hydrothermal treatment (Kita et al., 1995; Kondo et al., 1997).
Wang et al. (2002) have reported nano-structured zeolite 4A membranes, wherein nano-
crystalline zeolites 4A were seeded on an alumina filter by a dip-coating method. Hasegawa
et al. (2006) and Tago et al. (2008) have reported Silicalite-1 nanocrystal-layered membranes,
where the Silicalite-1 nanocrystals are piled up on an alumina filter, followed by
hydrothermal synthesis to form a Silicalite-1 layer on the nanocrystal layer. In these
membranes, the thickness of the nanocrystal layer (amount of nano-crystalline zeolite used
as seed crystals) can be easily changed, so that the separation properties of these membranes
204 Nanocrystals

can be controlled (Hedlund, et al., 1999; Pera-Titus et al., 2005). Moreover, because the zeolite
crystals are seed crystals for the membranes, the separation properties of the membranes are
improved with decreasing crystal sizes of the zeolite.

4. Conclusions
Zeolites are being used extensively in industrial processes such as heterogeneous catalysts
and adsorbents. Nano-crystalline zeolites have large external surface areas as well as low
diffusion resistance. Accordingly, control of the crystal size and the tailoring the mesopores
and macropores among the crystals are important factors for applications of these
nanocrystals. In order to obtain nano-crystalline zeolites, it is important to separate the
nucleation and growth stages. The addition of a surfactant into synthetic solutions of
zeolites is a promising method to control the crystal size and improve the crystallinity due
to adsorption of the surfactant on the zeolite surface.

5. References
Aguadoa, J.; Serrano. D. P. & Rodriguez, J. M. (2008). Micropor. Mesopor. Mat., 115, 3, 504-513,
1387-1811
Botella, P.; Corma, A., Iborra, S., Montón, R., Rodríguez, I. & Costa, V. (2007). J. Catal., 250, 1,
161-170, 0021-9517
Burkett S.L. & Davis M.E. (1994). J. Phys. Chem., 98, 17, 4647-4653, 0022-3654
Burkett S.L. & Davis M.E. (1995). Chem. Matter., 7, 5, 920-928, 0897-4756
De Moor P. P. E. A.; Beelen T. P. M., Komanschek B. U., Beck L.W., Wagner P., Davis M. E. &
Santen R. A. (1999a). Chem. Eur. J., 5, 7, 2083-2088, 0947-6539
De Moor P. P. E. A.; Beelen T. P. M. & Van Santen, R. A. (1999b). J. Phys. Chem. C., 103, 10,
1639-1650, 1089-5647
Dutta, P. K.; Rao, K. M. & Park, J. Y. (1991). J. Phys. Chem., 95, 17, 6554-6656, 0022-3654
Grieken, R. V.; Sotelo, J. L., Menendez, J. M. & Melero, J. A. (2000). Micropor. Mesopor. Mat.,
39, 1-2, 135-147, 1387-1811
Groen, J. C.; Peffer, L. A. A., Moulijn, J. A. & Perez-Ramirez, J. (2004a). Micropore. Mesopor.
Mat. 69, 1-2, 29-34, 1387-1811
Groen, J. C. ; Jansen, J. C., Moulijn, J. A. & Perez-Ramirez, J. (2004b). J. Phys. Chem. B, 108, 35,
13062-13065, 1520-6106
Hasegawa, Y.; Ikeda, T., Nagase, T., Kiyozumi, Y., Hanaoka, T. & Mizukami, F. (2006). J.
Membr. Sci., 280, 1-2, 397-405, 0376-7388
Hasegawa, Y.; Nagase, T., Kiyozumi, Y. & Mizukami, F. (2007). J. Membr. Sci., 294, 1-2, 186-
195, 0376-7388
Hedlund, J.; Noack, M., Kolsch, P., Creaser, D., Caro, J.& Sterte, J. (1999). J. Membr. Sci., 159,
1-2, 263-273, 0376-7388
Hincapie, B. O. ; Garces, L. J., Zhang, Q., Sacco, A. & Suib, S. L. (2004). Micropor. Mesopor.
Mat., 67, 1, 19-26. 1387-1811
Jia, C. J.; Liu, Y., Schmidt, W., Lu, A. H. & Schuth, F. (2010). J. Catal., 269, 1, 71-79, 0021-9517
Kita, H; Horii, K., Ohtoshi, Y., Tanaka, K. & Okamoto, K. (1995). J. Mat. Sci. Lett., 14, 3, 206-
208, 0261-8028
Zeolite nanocrystals - synthesis and applications 205

Kondo, M.; Komori, M., Kita, H. & Okamoto, K. (1997), J. Membr. Sci., 133, 1, 133-141, 0376-
7388
Kuechl, D. E.; Benin, A. I., Knight, L. M., Abrevaya, H., Wilson, S. T., Sinkler, W., Mezza, T.
M. & Willis, R. R. (2010). Micropor. Mesopor. Mat., 127, 1-2, 104-118, 1387-1811
Kumar, S.; Davis, T. M., Ramanan, H., Penn, R. L. & Tsapatsis, M. (2007). J. Phys. Chem. B,
111, 13, 3393-3403, 1520-6106
Larlus, O.; Mintova, S. & Bein, T. (2006). Micropor. Mesopor. Mat., 96, 1-3, 405-412, 1387-1811
Larsen, S. C. (2007). J. Phys. Chem. C, 111, 50, 18464-18474, 1932-7447
Lee, S.; Carr, C. S. & Shantz, D. F. (2005). Langmuir, 21, 25, 12031-12036, 0743-7463
Li, Q. H.; Mihailova, B., Creaser, D. & Sterte, J. (2001). Micropor. Mesopor. Mat., 43, 1, 51-59,
1387-1811
Li Q. H.; Mihailova, B., Creaser, D. & Sterte, J. (2000). Micropor. Mesopor. Mat., 40, 1-3, 53-62,
1387-1811
Masuda, T.; Fukumoto, N., Kitamura, M., Mukai, S. R., Hashimoto, K., Tanaka, T. &
Funabiki, T. (2001). Micropor. Mesopor. Mat., 48, 1-3, 239-245, 1387-1811
Matsune, H.; Tago, T., Shibata, K., Wakabayashi, K. & Kishida, M. (2006). J. Nanoparticle
Research, 8, 6, 1083-1087, 1388-0764
Mintova, S.; Olson, N. H., Valtchev, V. & Bein, T. (1999). Science, 283, 5405, 958-960, 0036-
8075
Mintva, S. & Valtchev, V. (2002). Micropor. Mesopor. Mat. 55, 2, 171-179, 1387-1811
Mintova, S.; Valtchev, V., Onfroy, T., Marichal, C., Knozinger, H. & Bein, T. (2006). Micropor.
Mesopor. Mat., 90, 1-3, 237-245, 1387-1811
Morales-Pacheco, P.; Alvarez, F., Del Angel, P., Bucio, L. & Dominguez, J. M. (2007). J. Phys.
Chem. C, 111, 6, 2368-2378, 1932-7447
Naik, S.P., Chen, J.C. & Chiang, A. S. T.(2002). Micropor. Mesopor. Mat., 54, 3, 293-303, 1387-
1811
Ogura, M.; Shinomiya, S., Tateno, J., Nara, Y., Nomura, M., Kikuchi, E. & Matsukata, M.
(2001). Appl. Catal. A. Gen., 219, 1-2, 33-43, 0926-860X
Pera-Titus, M.; Llorens, J., Cunill, F., Mallada, R. & Santamaria, J. (1995). Catal. Today, 104, 2-
4, 281-287, 0920-5861
Ravishankar, R.; Kirschhock, C., Schoeman, B. J., Vanoppen, P., Grobet, P. J., Storck, S.,
Maier, W. F., Martens, J. A., DeSchryver, F. C. & Jacobs, P. A. (1998). J. Phys. Chem.
B, 102, 15, 2633-2639, 1089-5647
Sakthivel, A.; Iida, A., Komura, K., Sugi, Y. & Chary K. V. R. (2009). Micropor. Macropor.
Mat., 119, 1-3, 322-330, 1387-1811
Serrano, D. P.; Aguado, J., Escola, J. M. & Rodriguez, J. M. (2005). J. Anal. Appl. Pyrol., 74, 1-2,
353-360, 0165-2370
Serrano, D. P.; Aguado, J., Escola, J. M., Rodriguez, J. M. & Peral, A. (2006). Chem. Mat., 18,
10, 2462-2464, 0897-4756
Serrano, D. P.; Garcia, R. A. & Otero, D. (2009). Appl. Catal. A-Gen., 359, 1-2, 69-78, 0926-860X
Serrano, D. P.; Van Grieken, R., Melero, J. A., Garcia, A. & Vargas, C. (2010). J. Mol. Catal. A:
Chem., 318, 1-2, 68-74, 1381-1169
Song, W.; Justice, R. E., Jones, C. A., Grassian, V. H. & Larsen, S. C. (2004). Langmuir, 20
(2004) 4696
Song, W., Justice, R. E., Jones, C. A., Grassian, V. H. & Larsen, S. C., Langmuir, 20 (2004) 8301
Song, W.; Grassian, V. H. & Larsen, S. C. (2005). Chem. Comm., 23, 2951-2953, 1359-7345
206 Nanocrystals

Tago, T.; Hatsuta, T., Miyajima, K., Kishida, M., Tashiro, S. & Wakabayashi, K. (2002). J. Am.
Ceram. Soc., 85, 9, 2188-2194, 0002-7820
Tago, T.; Tashiro, S., Hashimoto, Y., Wakabayashi, K. & Kishida, M. (2003). J. Nanoparticle
Research, 5, 1-2, 55-60, 1388-0764
Tago, T.; Nishi, M., Kono, Y. & Masuda, T. (2004). Chem. Lett., 33, 8, 1040-1041, 0366-7022
Tago, T.; Nakasaka, Y., Kayoda, A. & Masuda, T. (2008). Micropor. Mesopor. Mat., 115, 1-2,
176-183, 1387-1811
Tago,T.; Iwakai,K., Nishi, N. & Masuda,T. (2009a). J. Nanosci. Nanotechnol., 9, 1, 612-617,
1533-4880
Tago, T.; Aoki, D., Iwakai, K. & Masuda, T. (2009b). Top. Catal., 52, 6-7, 865-871, 1022-5528
Tago, T.; Sakamoto, M., Iwakai, K., Nishihara, H., Mukai, S. R., Tanaka, T. & Masuda, T.
(2009c). J. Chem. Eng. Jpn., 42, 1, 162-167, 0021-9592
Takenaka, S.; Orita, Y., Matsune, H., Tanabe, E. & Kishida, M. (2007). J. Phys. Chem. C, 111,
21, 7748-7756, 1932-7447
Tosheva, L. & Valtchev, V. P. (2005). Chem. Mater., 7, 10, 2494-2513, 0897-4756
Tsapatsis, M.; Lovallo, M. & Davis, M. E. (1996). Micropor. Mat., 5, 6, 381-388, 0927-6513
Wang, H.; Huang, L., Holmberg B. A. & Yan, Y. (2002). Chem. Comm., 16, 1708-1709, 1359-
7345
Complex nature of charge trapping and retention in NC NVM structures 207

X9

Complex nature of charge trapping and

retention in NC NVM structures
A. Nazarov and V. Turchanikov
Lashkaryov Institute of Semiconductor Physics, National Academy of Sciences of Ukraine
Ukraine

1. Introduction
In conventional silicon floating gate (FG) non-volatile memory (NVM) charge responsible
for the information stored is accumulated and retained in a bulk of polysilicon bar
consisting of tens of thousands Si atoms where surface to bulk atoms ratio is negligibly
small and interface nearly perfect due to the high temperature technological processes.
Electrons, usually exploited for charge accumulation, are mainly the same quasiparticles as
in the bulk Si isolated by macroscopic potential barrier on the bar boundaries. Typical
structure of NVM memory transistor used in embedded and standalone systems is shown
on the Fig.1a. Let alone the details of addressing, the main memory structure can be
presented by stacked system of substrate, tunnel SiO2, floating gate, control SiO2 and
program/access gate in split gate devices (Van Houdt et al., 1993) (Fig.1b). In this case
programming is realized by hot-electron injection from drain region of MOS transistor and
erasing by Fowler-Nordhaim (FN) tunnelling of electrons from floating Si-poly gate through
the lowered barrier into silicon. There are a lot of problems that must be solved, but, despite
of the efforts spent on development more and more reliable and capacious NVM elements
and circuits, some fundamental problems persist. Here we mention only two of them.
Firstly, tunnelling oxide through which the floating gate is charged must not be very thin to
prevent excessive random occurrence of leakage channels in dielectric, this, in turn, leads to
SiO2 overstressing by high electric fields needed to reach conditions of FN tunnelling, as
contrasted to the direct tunnelling. To disclose the complexity further let us take into the
account that the more perfect dielectric we create the more abrupt breakdown we obtain.
That is: the transition between safe charging current and irreversible breakdown converges.
As a consequence, more and more intricate cell structures are introduced for
reprogramming substituting, for example, FN electrons tunnelling by hot holes injection (US
Patent, 2004) and embedding additional charge balance controlling circuitry on the chip (US
Patent, 2007).
208 Nanocrystals

Fig. 1. (a) Typical structure Fig. 1. (b) Split gate NVM Fig. 1. (c) NVM capacitance cell
of conventional floating gate cell with voltage sources with nanocluster inclusions
NVM cell needed to Write, Erase
and Read.

Secondly, scaling down of storage elements accompanied by high W/E potentials reaching
20V and more along with multylevel cell (MLC) technology spreading give rise for
inappropriate programming risk through cross-programming forcing special measures to be
taken for protection of adjacent cells (Kuchubatla, 2010).
Recent years activities in the field of research and development of alternative to
conventional floating poly-Si NVM systems take into view the solution of these problems as
a part of a new embedded and standalone NVM systems generation industry deployment.
Here we concentrate on charge trapping and retention in silicon nanoclusters NVM
structures as directly derived from bulk Si floating gate predecessors. Typical NC NVM
structure presented on Fig.1c oddly enough resembles conventional NVM (Fig.1b), except
the gate which is formed by array of nanoinclusions.
Qualitative differences in the electrical properties of conventional and NC systems have its
roots in quantitative difference of Si atoms number in both systems. To illustrate this we
present in Table1 quantity of Si atoms concentrated in a bar 20x50x50 nm3 and nanoclusters
with diameters 3 nm and 2 nm. Mean number of surface atoms is also presented in this
table.

Type/Size Bar 20x50x50 nm3 Sphere of 3 nm Sphere of 2 nm

diameter diameter
Volume 5E-17 cm 3 1,41368E-20 cm 3 4,18867E-21 cm3
No of Si atoms 2500000 706 209
No of Si surface atoms 18420 317 141
Surface to bulk 0.007368 0.449 0.675

Table 1. Atomic comparison of Floating Gate and NanoClusters NVM elements

From this we can see, that under the transition from bulk Si to NC the role of boundary
atoms will grow. Also must grow the role of the whole NC medium including interclusters
and boundary SiO2.
Complex nature of charge trapping and retention in NC NVM structures 209

Effect produced by bulk → NC transition Electrical affected

(1) Bandgap, conduction and valence bands (2)
modifications  Bandgap widening due
to quantum
confinement effect
 Electrons emission
from NC instead of
holes trapping
 Nonexponential,
staircase-like, charge
relaxation (Coulomb
blockade effect)

(3) Simultaneous accumulation of opposite (4)

charges  The possibility of
unipolar
reprogramming
 Charge heterogeneities
generation
 Nonexponential
retention characteristics
 Strong dependence of
retention on the gate
material work function

(5) Stable current flow in pre-breakdown region. (6)

Soft breakdown  Wide range of
reprogramming fields
 Reduced necessity of
charge balance control
 The possibility of
unipolar
reprogramming
 The possibility of
unipolar and antipolar
reprogramming of the
same cell

Table 2. Qualitative effects introduced under conventional NVM to NC NVM transition.

The main feature of the transition is adding or removing electrons to the block only slightly
affects their position in energy bands. Situation radically changes when nanoclusters are
included in a gate dielectric as a storage media. In this case only dozens to hundreds silicon
210 Nanocrystals

atoms participate in charge storage processes of which considerable part are boundary,
forming microscopic traps with properties strongly dependent on local fields in turn,
depending upon their population. Program/Erase procedures are not reduced to simple
electrons injection/extraction but rather to trapping/recombination/emission. Complex
charge interchange processes are involved during every stage of NC NVM cell operation:
write, erase and storage, affecting the whole set of their characteristics, including window
formation, charge relaxation and retention. As an orientation to the classification guidelines
of this work we selected three main streamlines in which most evident divergence between
conventional NC NVM cells occurred. These are tabulated in Table2.

2. Experimental
European Union sponsored consortium “Nanoforum” that provides a comprehensive
source of information on all areas of nanotechnology to the business, scientific and social
communities in its eighth report “Nanometrology”(2006) characterised capacitance
spectroscopy as one of the main nanometrology techniques (Nanometrology, Eighth
Nanoforum Report, 2006). We widely use such kind measurements for full electrical
diagnostics of the NVM capacitance cell, which are first stage fabrication of the NVM
transistor cell, but totally determines memory characteristics of the last device.

Fig. 2. A simplified block diagram of the computer aided nanocluster NVM cell diagnostic
system.

The main characteristic of the W/E ability of the NVM capacitance cell is the window width
under the positive and negative reprogramming biases on the gate. This window we
calculate as a difference:

ܹ‫ ܹܱܦܰܫ‬ൌ ܾܽ‫ ݏ‬ቀܸ௙௕ ሺ൅ሻ െ ܸ௙௕ ሺെሻቁ (1)

Complex nature of charge trapping and retention in NC NVM structures 211

Whhere Vfb(+) and d Vfb(-) are thee flatband volta ages of the cap pacitance-voltagee (CV)
chaaracteristics calcu
ulated for positiv
ve and negative gate
g biases corresspondingly. Resppective
sim
mplified block diagram of the datta measurement and acquisition ssystem is presen nted on
Fig
g. 2:
Maain parameters off the system comp ponents are:
• DACs rang ges are: ±5.12V(±22.56V), ±10.24V and
a ±20.48V with max bias of ±35.884V
• DVM max x speed – 5000 meeas./s
• CV converrter time constantt 0.2 ms
In our develop pment of the exp perimental DAQ Q system on the first stage we fo oresaw
three main measurement
m pro
ocedures (Turcha anikov et al., 20005b): window cycling
c
measurements;; and W/E relaxaation measuremen nts; charge retenttion measuremen nts.

2.11 Window cyclin ng measurementts

Ap pplication specifiic data acquisitio
on system (DAQ
Q), software was specially design
ned for
wiindow testing of o write/erase cy ycles in nanomemory structurees with the algo orithm
illu
ustrated by Fig.3..

(b) (a) (c)

Figg. 3. a Algorith hm of structure biasing under write/erase
w win ndow testing of NVM
strructures with shiffting bias, b. - Ennlarged view of first
f BS series, c. - Window testing
g three
biaasing types initiall potential levels are arbitrary.

Firrst, CV characteriistic of the virgin

n structure is mea
asured. Next, antiipolar symmetriccal bias
is applied
a to the strructure, negative in the even cyclees and positive in
n the odd for fixed time
(in
n our case 20s). There
T were 6 succh bias-stress (BSS) cycles. After tthe first 4 cycles (dead
cyccles) no CV meaasurements were made with the goal to provide transient-free window w
forrmation. After thhe last two cycless CV characteristtics were measurred and stored and Vfb
determined. Then bias was increm mented (in our ca ase by 1V) and aabove described series
reppeated, and so on n up to 10 times.. Thus we got 21 files with CV daata and 2 files with
w Vfb
daata for each bias polarity with sub bsequent determmination of windo ow width (1). Eaach CV
currve was measureed in both directiions: from accum mulation to inverssion and vice versa and
conntains 2048 pointts (1024 for each ramp
r direction). Measurement tim me was near 5 s, the
t Vfb
212 Nanocrystals

was obtained after simple smoothing “on the fly” for the second ramp branch to prevent
distortions caused by discharging of the shallow trapping sites (Nikollian & Brews, 1982).
Next, CV data were smoothed by 80-point Fast Fourier Transform (FFT), the dC/dV values
were extracted and maximum of absolute (dC/dV) determined. The height of this
maximum was used as the criteria of surface potential heterogeneities on the
dielectric/silicon interface, which directly associated with heterogeneous charge trapping in
the nanoclusters in the dielectric. Summarizing graphical presentation of the results
obtained is given on Fig.4 (Turchanikov et al., 2005a).

Fig. 4. Normalized C-V and dC/dV-V characteristics for virgin and subjected BS structures.

In each bias cycle the sample can be illuminated by LED to supply minority carriers in space
charge region (SCR) with the goal to prevent nonequilibrium depletion of Si substrate that
does not take place in the NVM MOSFET cell. Additionally the experimental system can
measure window formation with one fixed bias and second – variable, ascending or
descending (Fig.3.c). This is especially useful when breakdown is asymmetrical or for
experiments with unipolar recharging.

2.2 W/E relaxation measurements

Window cycling gives us information on W/E window formation between bias cycles but
not inside them. The latter may be extremely useful in estimating physical processes of
nanoclusters recharging in the presence of external electric field (Tsoi et al., 2005). Algorithm
of the cell charging for relaxation measurement is presented by Fig.5. In these experiments
we used two identical series of bias time cycles with different biases in the first and second
series. Measuring Vfb after each biasing period we obtain a relaxation dependence of Vfb on
the cumulative time of biasing, tcum, where tcum is defined by:

t cum1  t1 , t cum 2  tcum1  t 2  t cumn  tcumn 1  t n (2)

Complex nature of charge trapping and retention in NC NVM structures 213

a b
Fig packets.
g. 5. Relaxation teesting algorithm: (a) single and (b) multiple pulses p

Sevveral relaxation cycles

c can be glu
ued together, to obtain
o data on noon-stationary proocesses
developing in nan nocluster NVM cells during W/E cycling. It sshould be noted d, that
sommetimes pulses were
w grouped in pulse packets coonsisting of pulsees with identical width
whhich differs from packet to packett. This was introd
duced with the go oal to access max
ximum
ressolution for low injection times and,
a simultaneouusly, overlap wid
de time range avoiding
hu
uge excessive dataa accumulation.

2.33 Charge retention measuremen nts

Wee used successive approximation ns register (SAR) algorithm to meeasure charge rettention
aftter write or erasee pulses. Time fo or SAR to reach desired
d precision
n was ~140 ms an nd the
seqquence of measurement intervals was regular in contrary
c to commmonly used logarrithmic
scaale. This was mad de starting from next
n reasons: 1) time
t needed to mmeasure one Vfb point
p is
sm
mall sufficiently as
a not to interrup pt the whole reten ntion process; 2) resulting data arrray is
commpact; 3) long innterval between measurements
m may lead to misintterpretation of th
he data
obtained. The latterr is illustrated by Fig.6.

Figg. 6. Complex ch
harge retention process
p (points) mistaken
m by simp
ple exponential decay
d
(lin
ne) deduced from
m logarithmic in time sampling (cirrcles).

Onn this figure exp

perimental pointss present compleex charge dissipaation in NC NVM cell
con
nsisting of two interdepending
i p
processes transfo
orming to one annother in the enccircled
reg
gion. Provided we
w have conven ntional sampling in the logarithm mic time scale (ccircled
214 Nanocrystals

points) relaxation process could be misinterpreted as a single exponential decay with a very
good fitting degree following parameters placed in the inserted table.

3. Samples
In our work we used samples prepared on p-Si substrates using next technological processes
for NC formation:
a) Low energy Si implantation in thermal SiO2 (EME, Departament d’Electronica,
Universitat de Barcelona, Barcelona, SPAIN) (Turchanikov et al., 2005c);
b) Ultralow energy Si implantation in thermal SiO2 (Institute of Microelectronics,
National Centre of Scientific Research IMEL/NCSR Demokritos, Athens, GREECE;
Nanomaterials Group, CEMES-CNRS, Toulouse, France) (Coffin et al., 2005);
c) LPCVD amorphous Si deposition with subsequent solid-state recrystallization and
oxidation (Institute of Microelectronics, National Centre of Scientific Research IMEL/NCSR
Demokritos , Athens, GREECE) (Turchanikov et al., 2007);
d) LPCVD poly-Si deposition with RPECVD SiO2 as a control gate dielectric (Institute
of Semiconductor Electronics, RWTH Aachen University, GERMANY) (Turchanikov et al.,
2005d).
The data on these samples are summarized in Table3.
Such variety of NC NVM samples allowed us to perform comparison of properties of the Si
nanoclusters of different size and natures and to extract common features in the
charge/discharge behaviour of such type systems.

Source RWTH, IMEL, Greece. IMEL, Greece. Barcelona Univ.

Germany (c) (b) (a)
(d)
Equivalent 25-35 17 5+2; 10+2; 35+2 43 (nc layer 20nm)
total oxide (calculated)
thickness (nm)
Control Oxide 20 10 3-5 (const) 12
thickness (nm)
Tunnel oxide 23 3.5 2; 7.5; 11
thickness (nm) (good quality (good quality thermal (with small
thermal oxide) oxide) nanocrystals)
Si nanocrystals 510 (TEM) 3.3 2.5; 3 2.70.2
size. Diam.
(nm)
Si nanocrystals (24)x1011 1012 1,7x1012 >1013
density
(cm-2)
N atoms per 3270-26100 940 409; 706 515
NC
Ns atoms on 881-3524 384 220; 317 257
surface of NC
Table 3. Summarized characteristics of the experimental samples used.

XTEM and scanning microscopy pictures of the formed nanoclusters are presented in Fig.7.
Complex nature of charge trapping and retention in NC NVM structures 215

Fig. 7. a - XTEM picture of crystalline Si nanoclusters fabricated by technology (c); b – XTEM

picture of Si nanoclusters fabricated by low-energy ion implantation with technology (a); c
– SM picture of amorphous Si nanoclusters fabricated by technology (d); d – profile of Si
ions implanted in silicon dioxide by technology (a).

4. W/E window formation

4.1 Nanoclusters’ ionization
At the beginning let us devote some attention to conventional NVM. Formation of W/E
window in these cells is realized through adding or subtracting negative electrons to the
floating gate. Processes involved usually are direct tunneling, FN tunneling or hot carriers
injection, sometimes hot holes injection is used to neutralize negative charge. But the latter
process is rarely used due to the heavy damages produced by hot holes in SiO2 and
complexity in implementation (US Patent , 2006). Thus in both states (program and erase)
the MOS transistors are in positive threshold voltages range.
The situation changed with nanoclusters embedded in the gate dielectric. In (Turchanikov,
2005a) we analyzed reprogramming cycles for NC NVM cells fabricated by technology of
ion beam synthesis (IBS) but similar situation was observed on the samples with
nanoclusters fabricated by different technologies. The write-erase cycles duration of 20 s for
negative and positive polarity of start voltage pulses are depicted by Fig.8a and Fig.8b,
correspondingly. It should be noted that during the first voltage pulse only positive charge
build-up in the oxide is observed for both voltage polarities.
216 Nanocrystals

Fig. 8. Window write-erase cycling from virgin samples with first positive (a) and negative
(b) poly-Si bias

On the next step of our study the transient processes of initial charge trapping in the virgin
samples were examined. We applied voltage pulses of different duration from 10 ms to 1 s.
Fig.9 presents the process of charge trapping under the voltage cycling on the gate electrode.
It can be seen that the initial storage process has two stages: first, the positive charge
accumulation and second, the positive charge dissipation. Under the subsequent bias stress
(BS) with positive voltage, two transients had been also observed. It should be noted that
full recovery of the C-V characteristics in the second BS cycle was not achieved. Detailed
analysis of the transient process related to the hole trapping in first BS cycle, with both
positive and negative voltage applied to the gate, reveals two stages of charge build-up in
the oxide (see inset in Fig.9 for negative voltage applied). The transient process can be
described by the following:

 V FB ( t )   V FBfast ( 0 )  exp(  t /  fast )   V FBslow ( 0 )  exp(  t /  slow ), (3)

where VFBfast (0) , VFB

slow
(0) and  fast ,  slow are the initial (t=0) amplitudes and time
constants for fast/slow trapping processes respectively. Density of the trapped charge (Nt)
can be calculated using the following expression:

C ox  V FB (4)
N t 
qS

where Cox is the oxide capacitance; q is the charge of electron; S is the area of the gate
electrode. Parameters of charge storage kinetics for the first BS cycle for both positive and
negative voltage on the gate are presented in Table4.
Complex nature of charge trapping and retention in NC NVM structures 217

Fig. 9. Write-erase cycling starting from hole injection in the dioxide. Inset: experimental
charge build-up modeling for first hole accumulation process with a single exponential
decay.

The important point is the close coincidence of the time constants and the trapped charge
density for the “fast” process for both positive and negative voltage applied during the first
step of the charging. This implies the similar charging process in both cases.

APPLIED “Fast” process “Slow” process

VOLTAGE τ(ms) Nt (1011 cm-2) τ(ms) Nt (1011 cm-2)
- 25.6 V 275 ± 57 4.43 1264 ± 163 10.25
+ 25.6 V 219 ± 22 3.50 2419 ± 235 2.34
Table 4. Relaxation process parameters for Fig. 9.

Together with flat-band voltage measurements after charge injection the slope of the C-V
characteristics was determined. It’s worthy to note that main decrease of the slope is observed
during “fast” stage of hole trapping both for positive and negative applied voltages (see Fig.10).
Special control experiments at liquid nitrogen temperature (similar to (Nikollian & Brews,
1982)) show that the slope decrease of the C-V characteristic is associated with the increase
of fluctuations of the surface potential at the SiO2/Si interface, but not with the surface state
generation.
Taking into account the above results one can conclude that the “fast” process of the positive
charge build-up in the oxide results in increase of the potential fluctuations at the SiO2/Si interface,
that can be associated with electron emission from the silicon nanoclusters located near the SiO2/Si
interface. Fig.11 presents the schematic picture of spatial distribution of the surface potential
in the SiO2/Si interface after the “fast” positive charge trapping process.
218 Nanocrystals

Fig. 10. Vfb and maximum of dC/dV vs. bias stress time for positive (+25.6V) (a) and
negative (-25.6V) (b) bias on poly-Si electrode

The following build-up of the positive charge during the “slow” process smoothes the
potential fluctuations in the SiO2/Si interface at the positive voltage applied (Fig.10a) results
in the restoration of the C-V characteristics slope for negative voltage (Fig.10b). The latter
case can be also related to the positive charge dissipation process, which is observed at the
last stage of the relaxation, i.e. the electrons trapping in the oxide. Indeed, if we apply
positive voltage to the gate just after the first cycle with negative voltage, the total
restoration of the C-V characteristic slope was observed. This is the evidence of nanocrystals
charge neutralization.

Fig. 11. Schematic picture of NCM device charging: initial conditions of the device (a) and
after fast stage of charging (b) starting from virgin sample.

“Slow” process of positive charge trapping has a different set of parameters for negative and
positive applied voltages (see Table4), so it is the argument in favour of different processes
of trap charging at positive and negative voltages. At the same time, along with background
positive charge incorporation in dielectric, there is the question of write/erase window
formation. This process, as illustrated by Fig.9 at the second switching, is characterised by
the following parameters: τ=63.7±3,3ms and trap concentration of Nt= (7.8±0.1)x1011cm-2.
Window formation may be due to two processes: holes or electrons trapping/detrapping in
dielectric. The latter is preferable because of the comparatively short time constant, as
compared to positive charge build-up presented in Table4. Thus, we can conclude that
Complex nature of charge trapping and retention in NC NVM structures 219

inddependently on thet polarity of thee voltage applied d on the first stagge of the charging g there
aree two main proceesses of the charg ge storage in Si nanocrystals
n rich SSiO2: first, the ion
nisation
of the
t Si nanocrystalss in SiO2 and theirr positive charging
g; second, uniformm positive charge trrapping
andd neutralization of the defects in dioxide
d matrix. Thhe evidence of faast recovery of the t Vfb
duuring negative-po ositive bias switchhing proves that the process of thee forming write – erase
wiindow is, presum mably, related to thhe electron trapping/compensatin ng in modified SiO O 2.
Deetailed examination of complex reelaxation processses (Turchanikov et al., 2005c) led d us to
thee conclusion thatt silicon behaviorr in nanoclusterss is similar to inttrinsic silicon – no n free
carrriers, no charge injection.
i But exccept for field ionizzation, there is an
n alternative way y to get
freee carriers in the nanoclusters, naamely, to raise th he temperature o of the sample. In paper
(Tuurchanikov et al., 2005c) series ofo the above desccribed experimen nts under the eleevated
temmperatures in thee range 20 C – 3000 C were perform med. Discharding relaxations inv volving
higgh-field ionizatioon along with in ntricate processees that took placce under simultaaneous
nanoclusters ionizaation and neutralization by electro ons injected fromm electrodes a fam mily of
rellaxation curves presented
p on Fig
g.12 were obtaineed. If we guess ssome analogy beetween
nanoclusters and bulk
b Si, assuming g the model pressented in Fig.13, when in commo on this
pro ocess can be desccribed as followin
ng:

  t 
N p (t , T ) ~ N 0 (T ) 1  exp    , (5)
   (T )  

whhere ΔNp is the charge stored in n nanocluster, N0 is the free elecctrons concentrattion in
nanocluster, τ is thee relaxation time constant. Here we
w must not forgeet, that in real situ
uation,
acccording Table4, thhere are two proccesses and, conseequently, two exp
ponents. Assumin ng that
Si nanocluster can n be considered as intrinsic semiiconductor, N0(T T) can be presen nted as
folllowing:
EG
N 0 (T )  ni (T )  Nc N v exp
p ( ), (6)
2 kT

Fiig. 12. Vfb relaxattion at elevated teemperatures. Fig. 13. Enerrgy diagram for NC
N
NVM structu ure for Si ionizatio
on.
220 Nanocrystals

wh here NC, NV and EG – analogs off the bulk density y of states in con
nductance and valence
v
ban nds and bandgap p correspondingly. Calculated acttivation energy fo or τ was in the raange of
0.005÷0.22 eV.
Thhis is a very low value
v for experimmental temperatuures over 150 0C b but allows us to neglect
n
seccond exponent in n equation (5) and
d estimate Eg of the
t nanoclustes frrom expression (66). The
obtained value of Eg=1.50±0.12 eV iss in the good agreeement with one found for the ban nd gap
forr Si nanoclusters with
w size of near 3.5 nm (Garrido--Fernandez et. al., 2002).

4.2
2 Unipolar windo ow formation in NC NVM structu ures
Ussually window in n NC NVM strucctures is formed by b application off two opposite polarity
pu
ulses (Tsoi et al.,, 2005; Carreras et al., 2005) butt occurrence of eelectron injection n from
nanocluster underr medium electrric field can bee mirrored in terms of the applied a
proogramming bias polarity. Namelly, there is no need to change b bias polarity. Thee same
ressult can be attaineed under the positive to more possitive gate bias sw
witching and vice versa.
Th
his phenomenon was called as “unipolar
“ bias reecharging” (Turcchanikov et al., 2005d).
2
Coorresponding processes are illustraated in Fig.14.

Fig
g. 14. A comparison of the W/E procedure Fig. 15. Window formation undeer the
in NVM cells undeer the usual (antiipolar) and antiipolar charging and symmetricaal bias
unnipolar conditionss. for NVM structuress with small nan nodots.
Thee zone selected for unipolar rech harging
exp
periments, due to o the opposite charge
c
polarity stored in a sstructure via biassing, is
marrked. Charging p pulse duration is 1280
ms

he phenomenon is
Th i observed in alll studied samplles with nanoclu usters floating gaate. On
exaample of the stru
uctures possessing g of Si nanodots with
w size near 5 n nm, which were
obtained by LP CV VD and low-temp perature RPE CV VD silicon dioxidee (Winkler et al., 2004),
somme features of thiis process can be demonstrated. With
W the goal to u understand the un nipolar
reccharging phenom menon, at first, thet NVM structu ures with silicon
n nanodots for charge
c
acccumulation undeer the symmetricaal antipolar biasiing were tested. As it can be seen n from
Fig
g.15, positive charrge accumulation n appears under the
t more positivee bias in a case of
un
nipolar charging,, in contrary to t the antipolarr charging, wheen positive charge is
acccumulated underr the most negativ ve structure biasin
ng.
Complex nature of charge trapping and retention in NC NVM structures 221

Thhe experimental daata for positive biias may be explain ned by the model presented in Fig.16.
Firrst, under the loww fields, negative charge,
c as a part of
o a through electtron flow (not preesented
on
n this figure for the simplicity) is trapped on the limited l number o of sites associatedd with
nan nodots (i), seconnd, with the field d rising, negativee trapped chargee is compensated d by Si
ion
nization in nanod dots (i.e. electron
n injection from the nanodots) ((ii) and, third, aftter the
em
mptying again, eleectron trapping prrevails (iii). The la atter inevitably inv
vokes positive feeedback
wiith respect to the electric field in thhe main bulk of dielectric and sub bsequent breakdo own of
thee available Si sitess. The more Si in nanodots the mo ore is the breakdo own field, so for thhe large
nannoclusters the breakdown field is greaater than for small. This fact we verrified experimentaally by
thee measurement off current-voltage characteristics
c (see Table1).

Fig
g. 16. Model of ch
harge accumulatio
on in NVM NC memory
m cells und
der the positive biias

Unnder the negativee biasing of the NVM

N structure ch
harge accumulatio on processes are rather
diffferent as presennted by Fig.17. Th he main differennce consists in thhe fact, that the charge
c
acccumulation in the region of low (i),( intermediate (ii) and medium field (iii) is due to the
carrriers’ trapping b
by the sites associaated with nanodoots in contrast to ttheir ionization. In
I this

Figg. 17. Model of ch

harge accumulatio
on in NVM nanoccluster memory ccells under the neegative
biaas.
222 Nanocrystals

situ
uation (i ÷ iii) ch
harge accumulatioon depends upon n competitive pro
ocesses of electron flow
thrrough imperfect RPECVD
R control oxide and holes tunneling througgh tunnel oxide. Under
thee high field (iv) nanodots ionizzation, with subssequent positive charge incorporation,
preevails.
Thhe possibility of thhe unipolar rech
harging process iss dictated by two
o competitive proocesses
in the singled out zone of Fig.15 (Tuurchanikov et al., 2005c):
2
1. Negative charge
c trapping under
u medium fieelds
2. Nanodots ionization (eleectron injection from nanodott) – positive charge c
inccorporation.

5. Relaxation
5.11 Relaxations un nder unipolar an nd antipolar charrging
Because of existing of electron injecttion process from m nanoclusters duuring unipolar chharging
it’ss possible to expect differences inn the relaxation processes
p under unipolar and anttipolar
chaarging. Above mentioned
m researcch was performeed in paper (Turcchanikov et al., 2005b),
2
wh here algorithm off the cell chargingg for relaxation measurement
m was also described. It
I must
be noted, that in thhis study cumulaative charging tim me, i.e. the sum o
of biasing time foor each
poositive and negativ ve bias was the saame and equals ~24
~ s.
Figg.18 represents the
t Vfb relaxatio
on process sequeencing, i.e. Vfb eestablishment with
w an
inccrease of cumulaative charging tim me in a series off bias pulses, unnder anti- and un nipolar
biaasing for 7 W/E cycles
c for nanodoots with size near 5 nm.

Figg. 18. a Anti- annd unipolar relax

xation charging process
p cycling in
n the structures with
sm
mall nanodots. Reeprogramming bias b field for antiipolar W/E cycliing is +4.2x106 V/cm
V
(firrst half-cycle in each cycle) andd -4.2x106 V/cm m (second half-cy ycle) , for unipoolar -
+4.2x106 V/cm (firrst) and +6.4x106 V/cm (second)); b - a compariison of the anti-- and
unnipolar relaxationn process (an expaanded view of th
he 7-th W/E cyclee from Fig.18a).

Thhe relaxation proccess of Vfb establiishment obtained

d in paper (Turch
hanikov et al., 20007) was
sim
mulated on the baasis of two relaxaation time constan
nts, discarding veery short relaxatiion (<5
mss), which cannot be
b reliably processsed by used DAQ Q system:
Complex nature of charge trapping and retention in NC NVM structures 223

WINmodel(t) = A1*exp(-τ1/t) + A2* exp(-τ2/t), (7)

where A1, A2, τ1, τ2 are simulated components of slow relaxation processes.
Thus it was experimentally determined the sum W/E window width Wexperimental, by
simulated components of slow relaxation with time constants of τ1, τ2, and it was simulated
maximum window WINmodel including the fast process with τ3 < 5 ms. The difference of
this magnitudes ΔWIN has to characterize the fast component of the relaxation process with
time constant τ3 < 5 ms. It was shown that under the antipolar recharging conditions a
comparison of the experimental data with model parameters demonstrates a difference
(±30%) in the predicted and observed results for positive charge trapping (Table5.). This
difference may be understood on the basis that some fast high field electron emission (with
time constant <5 ms) processes from the nanodots take place. In any case, under the antipolar
recharging, the formation of the W/E window is dictated, in the main part, by slow trapping of the
free carriers, electrons or holes, on the sites associated with nanodots.

In a case of unipolar biasing the situation is rather different. The positive charge
incorporation, more than 95%, via switching to more positive bias potential is much faster
(time constant < 5 ms) than negative charge trapping (time constant > 0.5 s) as illustrated by
Fig.18b and Table5. Thus, under unipolar recharging the formation of the W/E window is
dictated mainly (the switching to more positive polarity) by fast trapping of the free carriers
associated with electron injection from nanodots.

Single pulse duration 80 ms, cumulative charging time 24 s (300 pulses)

ANTIPOLAR UNIPOLAR
Neg. charge Pos. charge Neg. charge accum. Pos. charge
accum. accum. accum.
WINexperiment 1.3 V 0.82 V
al
τ1, s /A1, 4.0/ 7.4/ 4.9/
% of
experiment 46% 31% 54%
τ2, s /A2, 0.33/ 0.15/ 0.56/
% of Σ <3%
experiment 41% 41% 41%
WINmoel(∞)
% of 87% 72% 95%
experiment
τ3, s / <0.005/ <0.005/ <0.005/ >97%
ΔWIN, 13% 28% 5%
% of
experiment
Table denotations are in accordance with (1) and (2)
Table 5. Relaxation parameters for the NVM samples
224 Nanocrystals

5.2. Staircase charging

Fig. 19. Vfb relaxation under the irregular bias pulse sequence of 10 pulse trails with
duration incremented by doubling in the range 20÷10240 ms, (a) summary relaxation curve;
(b) positive bias of +7.52 V; (c) negative bias of -7.52 V; and (d) Vfb as a function of the
number of BS attempts (reshaping of results of Fig.19c).

Performing experiments on charge relaxation with trail sequences of bias impulses (Fig.5) it
was found that this relaxation was of staircase type (V. Turchanikov et al., 2007b). From
Fig.19a, it can be concluded that during the positive BS cycle net negative charge is
incorporated in the NVM structure, while during the negative BS cycle the situation is
reversed, as expected. Detailed examination of the relaxation process, as presented by
Fig.19b and c, reveals a staircase Vfb behavior in equilibrium during BS charging, that
depends mostly on the injection time in each sequence rather than on the cumulative BS
time. To illustrate this, the dependence of Vfb was rebuilt in coordinates of the number of BS
attempts under the same bias time (see Fig.19d). It is evident that the stair appears upon the
transition from one time sequence (10 points) to another. To verify the assumption that the
W/E window width depends on the single bias width rather than on the cumulative
charging time, the experiment under a constant bias with same cumulative BS charging time
but different durations of a single bias shot was performed. These results are presented in
Fig.20 and a summary of the corresponding W/E window is shown in Fig.21.
Using computer simulations, it was demonstrated that for a positive bias the window width
follows the single decay time constant relation:

଴ሺାሻ ି௧
οܸ௙௕ ൌ οܸ௙௕ଵ ‡š’ ൬ ሺశሻ ൰, (8)
ఛభ
Complex nature of charge trapping and retention in NC NVM structures 225

Fig. 20. Relaxation dependence of the Fig. 21. Write/erase window formation
write/erase window with fixed bias of +7.52 under a fixed bias of +7.52V (VfbPos), -7.52V
and -7.52 V, fixed cumulative charging time (VfbNeg) and fixed cumulative charging time
of 24 s and different durations of a single of 24 s, as a function of the duration of single
charging pulse. Numbers correspond to charging pulse. Full triangles correspond to
different single charging pulse durations in positive pulse application, full circles to
a cycle: 1.80, 2.160, 3.240, 4.320, 5.400, 6.480, negative, while full triangles show the width
7.560, 8.640, and 9.720 ms. Cumulative of the memory window.
charging time for each cycle was constant
and equal to 24 s for both positive and
negative cycles.

଴ሺାሻ
where οܸ௙௕ଵ represents the total flat band voltage shift in V for the given process for an
ሺାሻ
applied positive voltage pulse and ߬ଵ is the time constant of the relaxation process, with
ሺାሻ ଴ሺାሻ
mean values of ߬ଵ = 235±35 ms and οܸ௙௕ଵ = 0.716±0.18 V. This means that in every charging
sequence the same single negative charge trapping mechanism prevails (Fig.21). The density of
charge trapping sites was found to be equal to Nt= 1×1012 cm-2.
The process of Vfb relaxation under negative bias is more complex than that occurring under
positive bias. Instead of a single relaxation process, in this case at least two processes take place
ሺିሻ
and a large dispersion of time constants was observed. A fast time constant ሺ߬ଵ ሻ varying from
଴ሺିሻ
200 to 350 ms was measured with οܸ௙௕ଵ changing from 0.2 to 0.3 V, while a slow time
ሺିሻ ଴ሺିሻ
constant ሺ߬ଶ ሻ was also measured in the range of 11–25 s, with οܸ௙௕ଶ in the range of 0.05–
଴ሺାሻ ሺାሻ
0.12 V. If to compare the modelling results for positive ሺοܸ௙௕ ൌ οܸ௙௕ଵ ൎ ͲǤ͹ʹܸሻ and
ሺିሻ ଴ሺିሻ ଴ሺିሻ
negative ሺοܸ௙௕ ൌ οܸ௙௕ଵ ൅ οܸ௙௕ଶ ൎ ሺͲǤʹͷ ോ ͲǤͶʹሻܸሻ biases with the corresponding results
ሺାሻ ሺିሻ
from measurements οܸ௙௕௠௘௦ ൌ ͲǤ͹ʹܸܽ݊݀οܸ௙௕௠௘௦ ൌ ͲǤ͹ʹܸ it is possible to understand that
an additional very fast process of positive charge accumulation (τ<5ms and Nt3≈(3.5×1011÷15
5.8×1011) cm-2) that cannot be detected by our relaxation measurements setup takes place,
226 Nanocrystals

that contributes to NVM sample recharging. Moreover, we have detected a very slow
process of positive charge accumulation under the negative gate bias.
Thus, it was concluded that several single or distributed processes take place under the negative
biasing of the structure (positive charge accumulation), that can be associated either with
hole trapping in the nanocrystals or alternatively with hole trapping in deep traps (slow
process) located in the vicinity of the Si nanocrystals or electron emission from similar
amphoteric defects. Possibly the fastest process of positive charge trapping is associated with
shallow states at the interface of silicon nanocrystals with SiO2 or electron injection from
nanocrystals. Returning to the nature of staircase window formation (Fig.20), a computer
simulation via C–V measurements of charge accumulation under charge leakage from the
nanoclusters was made, from which it was proposed that charge draining from defect states
at the interface of silicon nanoclusters with SiO2 was possibly at the origin of the staircase
characteristic of Vfb as a function of time. However, this process would have ended in the
collapse of the W/E window in a short time (s), which does not happen in our structures, as
it was deduced from charge retention studies (Turchanikov et al., 2005a). On the other hand,
Coulomb blockade effects associated with carrier trapping/detrapping from nanoclusters of
sizes below few nanometers (Carreras et al., 2005), as those involved in our samples, cannot
be neglected. So, finally, a superposition of two effects was proposed for the explanation of the
staircase window formation: charge draining from defects in the vicinity of nanoclusters and
Coulomb blockade effects associated with carrier trapping/detrapping in nanocrystals.

6. Charge retention
The Fig.22 presents results of the retention experiments in the moderate time range (up to
≈3500 s), which was performed with the goal of estimation of the charge dissipation
processes in NVM cells and window width formation possibilities. At the first glance there
is nothing unusual in Vfb relaxation processes except for the facts that both retentions (for +8
V and +14 V biases) are cymbate and charge dissipation after +8 V pulse is very swift. The
latter leads to the narrowing of W/E window from ≈0.75 V at t=6s (first measured point) to
≈0.11 V (t=3500 s) that can be correctly depicted by two exponential decays with time
constants of τ1=75 s (fast) and τ2=1000 s (slow). Taking into account huge reduction of W/E
window width in the course of relaxation the short time retention experiments were
performed with increased time resolution (0.45 s). These results are presented by Fig.23. Let
us describe the retention curve after +8 V pulse more thorough.
First, the W/E windows appears to be even greater then estimated from Fig.22 and exceeds
1.2 V. Second, accumulation of negative charge in NVM cell is obvious, because at the first
moments of relaxation Vfb>0, hence there is no question about simple accumulation and
dissipation of accumulated positive charge. And, third, retention cannot be presented by
simple superposition of exponential decay processes.
Complex nature of charge trapping and retention in NC NVM structures 227

Fig. 22. Moderate time charge retention Fig. 23. Short time charge retention
characteristics of NVM sample with characteristics of NVM sample with
nanodots under unipolar bias conditions. nanodots under unipolar bias conditions.
Vfbi – flatband voltage of the virgin sample. Vfbi – flatband voltage of the virgin sample.

The last approval signifies that the process has to be described the following expression

௧ ఛ೏ష ௧
οܸ௙௕ ሺ‫ݐ‬ሻ ൌ  οܸ௙௕ଵ ሺͲሻ ൈ ‡š’ ቀെ ఛ ቁ ൅ ߜሺ‫ ݐ‬െ ߬ௗ ሻ ൈ οܸ௙௕ଶ ሺ߬ௗ ሻ ൈ ‡š’ሺെ ఛమ
ሻ, (9)
భ

where τd is a “dead” time for the second relaxation process and δ(t-τd) is a delta function.
The process of this type physically means that retention constituents are not independent,
i.e. the first charge dissipation process serves as a trigger turning on the second. In reality
there must exist a third process that prevents the triggering of the second simultaneously
retarding the first.
Basing on the above stated experimental data next model of charge redistribution during
retention in nanodots NVM system was proposed (Fig.24). After positive applied biasing
(small field, long time) two types of charge are accumulated in gate dielectric of NVM
structure, one negative, localized in a plane near Si-SiO2 interface, second – positive
localized deeper in SiO2 closer to metal gate. Naturally, there are distributions of such
charges, we present them as concentrated in planes for the simplicity.
In accordance with our model the whole retention process splits into three stages:
• Stage 1 – under the field E1 electrons are tunneling from the trapping sites through
tunnel SiO2 into the silicon substrate. Direct recombination of positive and negative charges
in dielectric is very slow which follows from the retention curves after +14 V bias (Fig.22,
Fig.23) and does not substantially affect the process. Other possible processes are retarded
either by E1, E2 or E3 fields directions. This process leads to decreasing of the whole set of
electric fields in dielectric. For charge transport on this stage only the reduction of E1 is
substantial. When this field drops to zero flow of electrons in the direction of Si stops – no
field, no current, Stage 2 begins;
• Stage 2 – unstable equilibrium takes place, when field E3 reverse its direction and
electrons emission from metal becomes possible. This emission leads to the recombination of
some positive charge, in turn field E1 drops rising electron emission from the negative plane
and restoring E1 to zero and so on, Vfb is stabilized on zero level – intermediate part on
228 Nanocrystals

Fig.23. This situation will continue up to the whole recombination of positive charge. Then
charge dissipation proceeds to Stage 3. Naturally, the Stage 2 duration depends upon initial

Fig. 24. Three stages model of charge Fig. 25. Retention characteristics of IBS
dissipation in NVM structure with samples obtained by ultra low energy Si
nanocluster inclusions after +8 V (20 s) implantation with gate storage biased.
biasing. Potential distributions induced by
positively (red), negatively (blue) charged
planes and sum distribution (black) as well
as the field and charge flow directions are
shown.

(after biasing) charges distribution in NVM structure. This, in turn depends upon the
charging field, work functions of metal and Si, prehistory of the sample and technological
details (cell structure) of the fabrication process. Stage 2 may be very long when field E1
drops to zero, but field E3 is not reversed and may be altogether absent if negative trapped
charge is small;
• Stage 3 – only captured electrons remain in dielectric determining middle time Vfb
relaxation process.
But the most evident confirmation of the crucial role that can play fields redistribution
during storage period in NC NVM structures we obtained studying charge retention in
samples with nanoclusters obtained by ultralow energy Si ion implantation (Coffin H. et al.
2005) with varying control gate storage potential (Ievtukh V. et al., 2008).
Slightly varying gate potential in the range +0.2 V ÷ -0.8 V we simulated possible gate
material work function changes from Al to Pt (negative potentials are equivalent to work
function increasing and positive to decreasing, gate material was Al) (CRC handbook of
Chemistry and Physics (2008). Obtained results are presented on Fig.25. Thus, application of
the positive voltage to the gate during retention process can considerably increase
information window.
Complex nature of charge trapping and retention in NC NVM structures 229

7. Conclusions
• Reduced quantity of Si atoms in storage media qualitatively affects all the electrical
characteristics set of NVM cells with nanoclusters: window formation, W/E relaxation and
charge retention.
• Specialized DAQ system, hardware and software, for studying the
trapping/retrapping processes in NC NVM cells were designed.
• It was found that positive charge accumulation due to the nanoinclusions
ionization in many cases is preferable for information storage in contrary to conventional
structures.
• This process creates charge heterogeneities in dielectric mirrored in substrate space
charge region that corresponds to local charge trapping in the dielectric.
• Effect of opposed charge accumulation under identical bias sign was found. It was
called “unipolar recharging/reprogramming”.
• In some cases W/E relaxation was of distinctly pronounced discrete (“staircase”)
nature that may be due to the interdependent redistribution of the different set of traps in
dielectric separated by energy and/or space.
 This opens the way for alternative development of Multy Level NVM Cell (MLC)
on the basis of injection duration in contrary to charge control.
• Temperature dependences of relaxation time constant enabled us to determine
band gap in nanoclusters which was 1.50±0.12 eV for samples used.
• Charge retention processes in NC NVM cells are of complex nature, including
parasite short-time specific for structures with nanoinclusions. Peculiar feature – to trap
charge of opposite sign simultaneously – can lead to nonexponential decay, when
completion of one process triggers the beginning of the other.
• Gate material work function variations directly affect retention characteristics of
NC NVM cell and may be used for improvement of storage characteristics.

8. References
Carreras J., Garrido B., Arbiol J., Morante J. R. (2005) Reliable, Fast and Long Retention Si
Nanocrystal Non-Volatile Memories in Materials and Processes for Nonvolatile
Memories, edited by A. Claverie, D. Tsoukalas, T-J. King, and J.M. Slaughter,
pp.229-234, (Mater. Res. Soc. Symp. Proc. 830, Warrendale, PA, 2005)
Coffin H., Bonafos C., Schamm S., Carrada M., Cherkashin M., Ben-Assayag G., Dimitrakis
P., Normand P., Raspaud M., Claverie A. (2005) Si nanocrystals by ultra-low energy
ion implantation for non-volatile memory applications Materials Science and
Engineering B, Vol. 124–125, (2005), pp.499–503, ISSN 0921-5107.
CRC handbook of Chemistry and Physics (2008), ed. D.R. Lide, 89-th edition, 2008-2009, pp.2496,
ISBN-10: 0849304792.
Garrido-Fernandez B, Lopez M, Garsia C, Perez-Rodriguez A, Morante J R, Bonafos C,
Carrada M and Claverie A (2002) Influence of average size and interface
passivation on the spectral emission of Si nanocrystals embedded in SiO2
J.Appl.Phys. Vol.91, No2, 2002, p.p.798-807, ISSN 0021-8979.
230 Nanocrystals

Ievtukh V., Turchanikov V., Nazarov A., Lysenko V., Normand P., Dimitrakis P. (2008)
Processes of window formation and dissipation in nanocluster nonvolatilememory
structures formed by ultra low energy ion implantation, Europhysics Conference
Abstract Volume 32F, ISBN: 2-914771-54-1, 22nd Gen. Conference of the Condensed
Matter Division of the European Physical Society, Rome, Italy 25÷29 Aug. 2008.
Kuchubatla, R. (2010) IMFT 25-nm MLC NAND: technology scaling barriers broken,
Electronic Engineering Times Europe, April, 2010, pp. 16-18.
Nanometrology, Eighth Nanoforum Report (2006), nanoforum.org European Nanotechnology
Gateway, July 2006, p. 60.
Nicollian E.H., Brews J.R.(1982) MOS Physics and Technology Wiley, ISBN 0-471-08500-6,
NewYork,1982
Tsoi E., Normand P., Nassiopoulou A., Ioannou-Soulgeridis V., Salonidou A. and
Giannakopoulos K. (2005) Silicon nanocrystal memories by LPCVD of amorphous
silicon, followed by solid phase crystallization and thermal oxidation, Journal of
Physics: Conference Series, Vol.10, 2005, pp.31–34, ISSN 1742-6588.
Turchanikov V.I., Nazarov A.N., Lysenko V.S., Carreras J., Garrido B. (2005a) Charge storage
peculiarities in poly-Si-SiO2-Si memory devices with Si nanocrystals rich SiO2
Microelectronics Reliability, Vol.45, No. 5-6, May-June 2005, pp. 903-906, ISSN 0026-2714.
Turchanikov, V., Nazarov, A., Lysenko, V., Winkler, O., Spangenberg, B., Kurz, H. (2005b)
Computer as a tool for nanocluster NVM cells diagnostics, EUROCON 2005.The
International Conference on "Computer as a Tool" Serbia & Montenegro, Belgrade,
November 22-24, 2005
Turchanikov V.I., Nazarov A.N., Lysenko V.S., Carreras J., Garrido B. (2005c) Charge
trapping in Si-implanted SiO2-Si memory devices at high electric fields and
elevated temperatures Journal of Physics: Conference Series, Vol.10, 2005, pp.409-412,
ISSN 1742-6588.
Turchanikov V.I., Nazarov A.N., Lysenko V.S., Winkler O., Spangenberg B., Kurz H. (2005d)
Study of the unipolar bias recharging phenomenon in the nonvolatile memory cells
containing silicon nanodots Materials Science and Engineering: B, Vol.124-125, 5
December 2005, pp.517-520, ISSN 0921-5107.
Turchanikov V., Nazarov A., Lysenko V., Ostahov V., Winkler O., Spangenberg B., Kurz H.
(2007a) Charge accumulation in the dielectric of nanocluster NVM MOS structures
under anti- and unipolar W/E window formation Microelectronics reliability, Vol.47,
No4-5, 2007, pp.626-630, ISSN 0026-2714.
Turchanikov V., Nazarov A., Lysenko V., Tsoi E., Salonidou A., Nassiopoulou A. G. (2007b)
Charging/discharging kinetics in LPCVD silicon nanocrystal MOS memory
structures Physica E: Low-dimensional Systems and Nanostructures, Vol.38, No.1-2,
April 2007, pp.89-93, ISSN 1386-9477.
US Patent (2004), No US 6,760,270 B2 Jul. 6, 2004.
US Patent (2006), No US 7,075,828 B2, Jul. 11, 2006.
US Patent (2007), No US 7,164,603 B2 Jan. 16, 2007.
Van Houdt, J., Haspelslag, L., Wellekens, D., Deferm, L., Groeseneken, G. and Maes, H.E.
(1993). HIMOS - A High Efficiency Flash E PROM Cell for Embedded Memory
Applications, IEEE Trans.on Electron Devices, Vol.40, No.12, pp. 2255
Winkler O., Baus M., Spangenberg B., Kurz H. (2004) Floating nano-dot MOS capacitor
memory arrays without cell transistors, Microelectronic Engineering, vol. 73-74, 2004,
pp. 719-724, ISSN: 0167-9317.
Carrier storage in Ge nanocrystals grown on
silicon oxide by a two step dewetting / nucleation process 231

10
X

Carrier storage in Ge nanocrystals grown on

silicon oxide by a two step dewetting /
nucleation process
A. El Hdiy, M. Troyon, K. Gacem* and Q.T. Doan
LMEN, EA 3799, Université de Reims Champagne Ardenne, UFR sciences, 51687 Reims
cedex 2, France
*At the present time, at CEA Saclay, France

1. Introduction
In microelectronics, dimensionality miniaturization theoretically leads to reliability increase.
However, an important concern in silicon technology is the effective reliability of MOS
(metal-oxide semiconductor) devices such as MOSFETs (MOS field-effect transistors) and
memory cells as they are scaled to smaller dimensions. Indeed, as the silicon dioxide (SiO2)
in CMOS technology is thinned below 2 nm for higher density and performance, limitations
associated with the poly-silicon gate become increasingly important. These limitations
include increasing poly-depletion effects, high gate resistance, doping impurity
penetration,…As a consequence, some problems affecting the CMOS devices reliability
appear. Among the main problems, the exponentially increased gate leakage current, the
reduced threshold for dielectric breakdown and oxide charging which results in voltage
shifts. To avoid these difficulties, group IV nanocrystals embedded in a SiO2 matrix have
been studied extensively because of their potential for integrated optoelectronic devices on
silicon substrates. The use of silicon (Si) nanocrystals (NCs) instead of standard
polycrystalline silicon floating gate was proposed and many studies have been made
describing the NC capabilities in different devices (Tiwari et al., 1995; Park et al., 2003;
Conibeer et al., 2006). However, due to their large bandgap variations and their potential for
bringing a strong quantum confinement, germanium NCs have attracted more interest than
Si-NCs as reported in many studies (Choi et al., 1999; Skorupa et al., 2003; Chatterjee et al.,
2008).
For non-volatile memory device application, a long charge retention time at room
temperature is the most important. This is the reason why the study of the NC specific
properties is of principal importance. Indeed, device reliabilities are strongly affected by the
variation of the NC structural parameters; namely, their density, their size, and
consequently the NC spacing. The NC spacing controls the energetic potential recovery of
the NCs and promotes the carrier hopping between them. Some of these effects cannot be
evaluated by the standard techniques, because the evolution of the curves contains a global
232 Nanocrystals

response; all the NCs are involved in the transport and storage, so one obtains just average
information which is affected by the NC size fluctuation and density.
It is interesting to obtain information on an isolated NC (or a limited number of NCs) in a
given structure to optimize the device functionalities. Indeed, electric transport through
NCs and carrier storage in NCs depend on the NC structural properties. It appears that the
use of the atomic force microscopy (AFM) is very appropriate if the properties of individual
NCs or quantum dots (QDs) are to be exploited with nm-scale resolution (Okada et al., 2002;
Stomp et al., 2005; Smaali et al., 2006, Smaali et al., 2010, Gacem et al., 2010). So, in this work,
we present two kinds of results; results obtained with standard methods in which one
brings average information, and results obtained by use of the C-AFM technique. The
standard techniques consist of the use of high frequency (1 MHz) capacitance – voltage
(HFC-V) and current – voltage (I-V) methods usually applied in microelectronic industry.
Note that the kind of carriers involved in the electric transport and the charge storage
depends on the nature of the electrode/sample contacts. With the standard methods, for
which a metallic contact is used, electrons are concerned. However, conduction by holes
dominates when working with a conductive AFM using a p-doped diamond probe, which
leads to formation of an artificial nano-heterojunction at the contact with a large conduction
band discontinuity. Obviously, if the C-AFM probe was a metallic probe (Pt-Ir, Cr-Co,…),
electric conduction could be governed by electrons. Indeed, in a recent study, we studied the
carrier transport through quantum dots by the C-AFM technique using two kinds of
conductive AFM probes (p-doped diamond and metallic probes), and showed that the type
of the carriers flowing through the nanostructure depended on the used AFM probe (Smaali
et al. 2010).
HFC-V and I-V were used to study the carrier storage capabilities of the Ge-NCs in a large
temperature range [77 – 300K]. For these techniques, the studied sample is similar to a MOS
(metal – oxide – silicon) capacitor. This work allowed us to study not only the carrier
transport and storage but also to evidence at room temperature a Coulomb blockade effect
for nanocrystals of 3.5 nm average diameter. The latter appears for increasingly low voltages
when the temperature decreases.
The C-AFM method was used to study carrier storage in a single NC in vacuum and at
room temperature. With this technique, electric images of NCs of any arbitrary positions on
the sample surface can be studied using the same conductive AFM probe. However, this
technique requires that the NCs are not capped with an additional thick layer. In our case,
the studied samples were elaborated in the same conditions than those used for standard
techniques but without layer covering the NCs. The C-AFM brought interesting information
on the NCs, especially on conduction characteristics and carrier storage. This study was
performed with a home made AFM working inside a scanning electron microscope (SEM)
(Troyon et al., 1997).

2. Sample elaboration
The crystalline silicon substrate used to elaborate samples is (100) oriented, boron doped
and thermally oxidized to ~ 5 nm thick. The amorphous germanium layer was deposited by
Molecular Beam Epitaxy (MBE) in ultra-high vacuum (10-11 Torr) at room temperature. The
Ge-NCs were formed after 30 min in situ annealing at 700°C by the combination of
unwetting and crystallization processes (Karmous et al., 2006; Szkutnik et al., 2008). The
Carrier storage in Ge nanocrystals grown on
silicon oxide by a two step dewetting / nucleation process 233

mean diameter (D) of Ge-NCs is uniquely controlled by the nominal thickness (t) of the
amorphous layer following a relation D  t. The thickness was varied between 0.5 and 5 nm
leading to NC mean diameters between 3.5 and ~34 nm. For all samples, Ge-NCs present a
unique aspect ratio (height over diameter) of about 0.8 which strongly differs from the
aspect ratio (about 0.15) of Ge quantum dots in epitaxy on Si substrate (Berbezier et al.,
2002).
In the present work, our samples are labelled A and B. For the A samples, the silicon
substrate was boron doped to ~1015 cm-3, and a 18 nm thick amorphous silicon (a-Si) was
deposited as a capping layer. The use of a-Si layer leads to an electrical conduction between
one electrode and Ge-NCs for very weak biases. This situation could not be obtained if we
had an insulating layer, as in memory devices, where the carrier exchange through the oxide
makes damages and finally leads to an oxide breakdown. From an energetically point of
view, the a-Si bandgap is higher than that of the Ge-NC, which reinforces the spatial
confinement of the carriers in Ge-NCs, and at the same time remains small to have an
electronic transport under biases as weak as those applied in this study. Weak biases avoid
oxide damages in order to have reproducible measurements. In other words, it is not
necessary to apply a high bias to the sample for characterization measurements. The
samples of this series are labelled A3.5, A17, A21 and A35 and characterized by their average
size and density (i.e., 3.5 nm and 2.41012 cm-2 for A3.5, 17 nm and 1.51011 cm-2 for A17, 21
nm and 6.71010 cm-2 for A21, 35 nm and 2.41010 cm-2 for A35). Transmission electron
microscopy (TEM) cross-section image shown in Fig. 1(a), gives a typical example of the
stack layers for Ge-NCs with an average diameter of 3.5 nm. It can be clearly seen that Ge-
NCs are monocrystalline ({111} plans are identified) and free of extended defects as shown
in Fig. 1(b).

Fig. 1. (a) TEM image (cross sectional view) of Ge-NCs formed by annealing of an
amorphous Ge layer (0.5 nm thickness) deposited on an ultrathin SiO2 (5 nm thickness) for
30 min at 700 °C and capped by a 18 nm of amorphous silicon. (b) High resolution TEM
image (cross sectional view) of a Ge-NC where the distance between {111} plans are
evidenced (Szkutnik et al., 2008).

For the B sample, the silicon substrate was n doped to ~51018 cm-3 and the Ge-NCs were
not capped in order to allow electrical AFM measurements. This sample contains 6109 Ge-
NCs per cm2 the average diameter of which is equal to 70 nm.
234 Nanocrystals

3. Experiments
3.1 C-V and I-V techniques
For the study of the A samples, electrical contacts were made using silver bond on the two
faces of each sample. The structure is similar to a MOS capacitor (see Fig. 2). The area of the
silver bond electrode is approximately circular and its diameter is about 2 mm. HFC-V and
I-V measurements are obtained by limiting the gate bias to relatively low values ( 3 V) in
order to prevent the silicon dioxide damaging and are carried out at various temperatures
from 300 K down to 77 K. I-V measurements were made with a HP4140B picoammeter and
HFC-V characteristics were performed with a HP4284A RLC-meter. We only focus here, on
the HFC-V in order to have a direct response of the carrier exchange (trapping/detrapping)
effect. The HFC-V measurements were made from inversion (positive gate bias) to
accumulation (negative gate bias) regimes and immediately back to the inversion regime
with the same voltage sweep rate. The carrier trapping/detrapping process was shown by
the occurrence of the hysteresis cycle in the CV curves.

Fig. 2. A schematic cross section of the studied samples of the A series. V is the bias
polarization: the structure is at inversion regime if V > 0 and at accumulation regime if V < 0.

3.2 Conductive-AFM/SEM technique

The sample B was studied by using a home-made AFM combined to a scanning electron
microscope (SEM) (Troyon et al., 1997) equipped with a field emission gun (LEO Gemini).
Our measurements were made with this AFM/SEM combined instrument because working
inside the SEM allows one to get rid of the influence of the water layer covering the surface
in ambient atmosphere which hinders the electric measurements. Indeed, working in wet
atmosphere may create local oxidation of the surface at the tip/surface contact (Okada et al.,
2001). Furthermore, this hybrid instrument allows obtaining three simultaneous images;
electric, topographic and secondary electron images. For example, the secondary image
shown in Fig. 3 allows the verification of the probe geometry and check the state of the
probe and of the sample surface. Ge-NCs are well seen and are randomly distributed at the
surface of the sample. Note that precise analysis of the Ge-NC structural (average size and
Carrier storage in Ge nanocrystals grown on
silicon oxide by a two step dewetting / nucleation process 235

density) properties, the size distribution and the density have been previously studied and it
was shown that the distribution of the NC size is very narrow (Szkutnik et al., 2008).

Fig. 3. Secondary electron image showing a p-doped diamond AFM tip in contact with Ge-
NCs of 70 nm in diameter.

Another important reason to perform measurements inside the SEM is that this particular
sample requires a preliminary electron beam irradiation in the condition for obtaining an
electron beam induced current (EBIC) image to obtain afterwards (electron beam off) a
measurable conduction current. Indeed, due to the thickness of the SiO2 layer it was
impossible to get a C-AFM image for polarization voltages smaller than  8 V. The reason
why a preliminary electron beam irradiation is needed is explained in section 4.3.

(1) AFM Head.

(2) Preamplifier.
(3) Piezo-tube supporting the sample.
(4) Position of the laser diode.
(5) Photodiode adjustment system.
(6) AFM cantilever support.

Fig. 4. View of the AFM mounted inside the specimen chamber of the SEM.

Fig. 4 gives an image showing the various components of the AFM and Fig 5 gives a
schematic representation of the hybrid system used in the present work. The detected
current remains very low. So, we installed a low noise (~5pA RMS, 1 MHz bandwidth)
current/voltage amplifier (107 gain) close to the AFM tip, inside the SEM, to limit the electric
noise. An ohmic contact was made on the back side of the samples by depositing an
aluminium layer. The sample surface (the top side) was probed in the contact mode while a
sample bias V was applied to the substrate with respect to the tip, which is itself referenced
to the ground through the preamplifier. In our system, the sample is scanned with respect to
236 Nanocrystals

the AFM probe. The electron-beam irradiation of the samples was performed in the
conditions used for obtaining a nano-EBIC image with the e-beam focused in a fixed
position just in front of the AFM probe and as close as possible to it (Smaali et al., 2008); the
electron primary energy was 5 keV, the primary current was 1 nA, the sample bias was V =
0 volt and the images (256256 pixels) were acquired at a line scanning frequency of about
0.5 Hz.

Fig. 5. Schematic representation of the experimental combined AFM/SEM set-up.

4. Results and discussion

4.1 Capacitance – voltage characteristics
HFC-V and I-V measurements cycling on A samples show the hysteresis loops
representative of the carrier storage. The question is: where does the carrier storage occur?
Does it appear in NCs and/or in the oxide bulk? To give an appropriate response, Fig. 6(a)
shows typical HFC-V curves with arrows indicating the sense of the measurement. The same
cycle is obtained whatever the beginning of the measurement direction from positive or
negative gate bias. Note that the silicon substrate is p-doped. The HFC-V curves measured
from positive to negative gate bias are shifted towards negative biases compared to the
HFC-V curves measured in the opposite direction. This highlights the electron
trapping/detrapping in Ge-NCs and clearly indicates that this process is either non-existent
or negligible in oxide.
Let us describe the electron trapping process. In inversion regime (when the silicon
substrate is negatively polarized), electrons in the substrate accumulate close to the Si/SiO2
interface and holes (the majority carriers) move away from this interface towards silicon
bulk. On the other side of the sample, the trapped electrons can be emitted from Ge-NCs to
the gate via a-Si. Hence, the HFC-V curve is shifted to negative bias (this curve is taken as
reference to calculate trapped charge due to HFC-V shift). If the bias is reversed, electrons
can be injected in Ge-NCs and the HFC-V curve is moved to positive bias as shown by the
arrows in Fig. 6(a). These electrons do not come from the substrate because of the electric
Carrier storage in Ge nanocrystals grown on
silicon oxide by a two step dewetting / nucleation process 237

field orientation, but from the other side via the a-Si. The contribution of the current coming
from electron trapping in a-Si layer, which is expected due to the presence of defects (such
as dangling bond and vacancies), is not considered here since carrier emission and capture
in these defects are very fast in comparison to the time measurement. Indeed, since the
energy levels of the defects are close to the conduction and valence bands they produce very
fast conduction paths for electrons and holes. Note that the electron trapping is enhanced at
low temperature.

Fig. 6. (a) Electron storage in nanocrystals and its effects on the capacitance – voltage cycles,
the temperature effect is also shown, (b) electron storage amplitude for three samples (A17,
A21 and A35).

In Fig. 6(b), we present the hysteresis width (V) deduced from the HFC-V cycles for three
samples of the A series as a function of the temperature. V is reduced when the
temperature increases and is also reduced for samples containing NCs with high average
diameter as shown in Fig. 6(b). Note that V represents the global stored charge. The
analytic expression of V is given by Tiwari et al. (Tiwari et al., 1995), which shows that V
is directly related to the Ge-NC density, to the number of electrons stored by NC and to the
NC mean diameter. Two main parameters control the V amount: the Ge-NC size and
density, but their effects can be different. If the NC size controls the trapped charge density
(represented by V) the trapping process will be stronger in NCs with high average
diameter. This seems true at low temperature (< 150 K) as shown in Fig. 6(b); the total
charge trapping is much more efficient in large Ge-NCs than in smaller ones. But at high
temperature (> 150 K), the Ge-NC density controls the stored charge amount. The
temperature plays thus a particular role in Ge-NC size and density effects on the global
stored charge.
As it has been quoted above, eventual carrier storage can occur in the oxide after flowing
through the oxide interfaces barrier height. Nevertheless, this is screened by the effective
charge stored in NCs. Finally, the carrier exchange is made between the Ge-NCs and the
gate as quoted in Fig. 7 giving a schematic and qualitative situation for electron storage in
NCs when the gate is negatively polarized by reference to the Si substrate (on the left of Fig
7) and NCs discharging when the bias polarization is inverted (on the right of Fig. 7).
238 Nanocrystals

Fig. 7. Energetic band diagram showing carrier trapping in NCs and detrapping according
to the bias polarization.

4.2 Current – voltage characteristics

The carrier trapping phenomena are also observed in I-V curves as shown in Fig. 8 where
the measurements are performed from negative (-3 V) to positive (+3 V) and back to
negative gate voltage. Note that the charge storage in standard MOS capacitors and its
effects on the I-V curves is well known (Balland & Barbottin, 1989; El Hdiy et al., 1993). The
shift of the I-V curve along the voltage axis is controlled by the type of the effective oxide
charge; a positive oxide charge causes a negative voltage shift of the I-V curve, while a
negative one leads to a positive voltage shift. Moreover, if the stored charge is located near
the interfaces (gate/oxide or p-Si/oxide), the I-V curves show a distortion without shift (or
with a very weak shift) along the voltage axis (El Hdiy et al., 1993).

Fig. 8. Typical result showing occurrence of the hysteresis in the I-V curves obtained in the
sample containing NCs of 21 nm diameter (A21). The enlargement of the cycle is due to the
thermo-electronic effect reduction.
Carrier storage in Ge nanocrystals grown on
silicon oxide by a two step dewetting / nucleation process 239

Let us remark that only the I-V characteristics in inversion regime present a hysteresis cycle.
And this hysteresis increases when the temperature is reduced indicating the increase of the
carrier storage as previously reported (Kuo & Nominanda, 2006). We also notice that the I-V
curves are not shifted indicating that the trapping/detrapping phenomena do not occur in
the oxide bulk but more in the Ge-NCs as explained in the following.
The measurement first begins at negative bias (-3 V). At negative bias, holes move from the
p-Si valence band to the gate by tunnelling effect, namely Fowler Nordheim tunnelling
(FNT) effect, while electrons move from the gate to Ge-NCs, but their energy and their
density are affected in the a-Si. Hole (majority carriers) density is higher than that of
electrons (minority carriers). It is expected that trapping process (electrons and/or holes
trapping) appears in both oxide bulk and in Ge-NCs; holes in the oxide and electrons in
NCs. Indeed, when the bias voltage increases from -3 V to zero, becomes positive and
reaches + 3V, hole density is reduced but the density of electrons coming from Si substrate
becomes more and more important showing a current increase at positive bias voltage (from
0 to +3 V). When the applied voltage turns back from +3 V to 0 V, the current presents
higher values than the first scan. In this case, one has two possibilities leading to increase of
current; the first process is the hole detrapping from the oxide, and the second process is the
electron detrapping from the Ge-NCs. Both processes can appear during the scan of the bias
from zero to +3 V. However, if the hysteresis was related to the hole trapping/detrapping in
the oxide, the sense of the hysteresis cycle, indicated by the arrows in the figure would be
inverted. Because hole trapping reduces the electron barrier height and the tunnelling
distance, leading to a higher current when the bias voltage varies from 0 to +3 V than when
it varies from +3 to 0 V. I-V measurements were also performed at low temperatures. The
results show that the hysteresis size depends on the temperature; the hysteresis width
expends when the temperature decreases suggesting that the carrier trapping is enhanced as
also shown by the HFC-V technique. Finally the same carrier trapping process is revealed by
both I-V and HFC-V characteristics.
Let us note that the use of the Fowler-Nordheim tunnelling standard expression can
qualitatively lead to determine the density of the carriers trapped in Ge-NCs, or at least it
can inform on the trapping efficiency versus the temperature. This approach was previously
made in the case of the standard MOS capacitor (DiMaria et al., 1993; El Hdiy & Ziane, 2001)
to extract the trapped charge density from I-V measurements. We carefully use it according
to the following assumptions: (a) there is no trapped carriers in the oxide, or the eventual
stored carriers do not affect the initial oxide field E0, (b) the charge contained in the NCs is
supposed to be distributed in a thin layer at the NC/oxide interface and (c) the
supplementary field caused by the carrier storage process remains lower than E0.
The standard FNT expression (Fowler & Nordheim, 1928) is:

I  AE 2 exp B / E  (1)

Where, I is the tunnelling current, A and B are the FNT parameters considered to be
constant. E is the electric field near the injecting electrode (the interface which is negatively
polarized). Note that the expression given in Eq. (1) is valid only for high voltages; at low
voltages, we must take into account the total expression of the tunnelling current (FNT and
direct tunnelling) (Schuegraf & Hu, 1994). The oxide electric field can be expressed as:
240 Nanocrystals

qN
E  E0  (2)
 Ge

Considering that the contribution of the stored electrons in NCs to the electric field E
variations remains relatively negligible, the density of the trapped electrons is given by:

E02 Ge  I 
N T    Ln  (3)
qB  I0 

Where, E0 is the field when the NCs are totally empty, q is the elementary charge, Ge is the
Ge permittivity (Ge is used instead of oxide because the charge is stored in Ge-NCs) and N is
the density of the stored electrons. I0 and I are the current before charge storage (N = 0) and
after carrier storage (N  0), respectively.
During the measurement under negative gate bias, electrons are trapped by NCs because of
their weak energy at the injecting electrode. When the gate bias is inverted, the negative
charge stored in NCs affects the tunnelling current leading to a current I lower than I0. The
use of Eq. (3) to data of Fig. 8, allows us to show that the stored charge density at 77 K is
twice higher than that trapped at 300 K; N(77 K) / N(300 K)  2. On the other hand, N
can be measured by taking the current (I and I0) values at bias voltage corresponding to the
large width of each cycle hysteresis.

Fig. 9. (a) Current-voltage characteristic at a positive gate bias (inversion regime). The steps
observed in the reverse I-V curve are related to electron resonant tunnelling via Ge NC
discrete levels. The inset contains the corresponding conductance as a function of gate
voltage. (b) Evolution of current-voltage characteristics with temperature. Current jumps
appear for increasingly weak voltages when the temperature decreases.

In addition to the carrier trapping evidence, we notice the remarkable behaviour exhibited
by NCs which have a mean diameter of 3.5 nm and an average density of 2.51012 cm-2.
Their I-V curves present steps in the reverse part as shown in Fig. 9(a). The observed jumps
Carrier storage in Ge nanocrystals grown on
silicon oxide by a two step dewetting / nucleation process 241

are not linked to the oxide breakdown for the following reasons. The maximum value of the
oxide field is lower than 6 MV/cm (value from which damage in the oxide can occur), and
to prevent defect creation in the oxide, the bias was scanned rapidly for higher values (from
2 to 3 V). Moreover, to be sure that the observed steps do not correspond to the oxide
breakdown, the I-V characteristics were repeated many times indicating approximate
reproducible results. This little difference in terms of the plateau voltage position and height
can be linked to both the temperature and size dispersion effects. Moreover, other
measurements made at low temperatures have shown that the plateaus became larger.
These jumps are representative of a resonant tunnelling of electrons into the discrete energy
levels of Ge-NCs. Such tunnelling resonance behaviour has been attributed to the Coulomb
blockade effect in Si quantum dots embedded in Si-rich SiO2 deposited in plasma phase
(Kim, 1998). Other recent studies have also shown the Coulomb blockade at room
temperature in roughly spherical nanocrystalline silicon dots the main diameter of which
was 5 nm (Wu et al., 2004). .
Let us now explain this resonant tunnelling current. It is well known that in tunnelling
resonance, electrons accumulate between double barriers (Goldman et al., 1987; Kim, 1998,
Kareva et al., 2002; Vexler et al., 2006). In the case of our sample, electrons can temporarily
accumulate at the Si/SiO2 interface quantum well and between the double barrier formed by
Ge-NCs, a-Si, and oxide. Under a positive bias, electrons are injected from the Si/SiO2
quantum well through the oxide to the subband levels of the Ge-NCs where they fill the
same energy levels as in the Si/SiO2 quantum well. A simple description of the tunnelling
resonance process can then be given: if we consider that electrons have the energy level of
the silicon conduction band edge at the Si/SiO2 interface (Ec0), when the gate is positively
biased, current flows only if Ec0 is swept through the first quantum energy level (E1) of Ge-
NCs resulting in a current peak in the I-V curve. Since current continues to flow after Ec0 is
swept past E1, one observes current steps rather than peaks in the I-V curves. Two main
conditions for observing the Coulomb blockade effect are (a) the tunnelling resistance
between Ge NC and the electric contacts must be higher than the quantum resistance (Rt >>
h/e2  26 k) and (b) single electron charging energy (E=e2/2C)>> kBT, where
C  40aSir = 2.277 aF is the self-capacitance of an approximately spherical Ge NC and
aSi =11.7 (Ref. Orwa et al.’2005) (this expression of C gives an overestimation of the
capacitance value, since the spherical form of the Ge-NC is truncated at the contact with the
oxide). This capacitance gives a single electron charging energy E of about 35 meV,
supporting the existence of the Coulomb blockade effect. The first criterion can be met by
weakly coupling the NC to two electrodes which is the case in our sample. The NCs are
confined between SiO2 and non doped amorphous silicon. The second criterion depends on
temperature. At room temperature, the Coulomb blockade condition occurs when the
capacitance is smaller than the thermal one, Cth = 3.1 aF (3.110−18 F), for a single quantum
dot (or a diameter  5 nm). In the situation investigated here, the capacitance values are
smaller than Cth. Consequently, the charge energy (e2/C) is higher than the thermal energy
(kBT). On the other hand, Fig. 9(b) shows that current jumps appear for increasingly weak
voltages when the temperature decreases (Gacem et al., 2007). Note that this Coulomb
blockade effect appears only for the NCs with the smallest main diameter (3.5 nm), and not
in the other studied samples which probably would require temperatures lower than 77 K.
This is related to the strong quantum confinement in 3.5 nm Ge-NC than in the others.
Indeed, figure 10(a) gives a schematic energetic band diagram of the studied samples
242 Nanocrystals

showing the bandgap of each element and corresponding to a thermal equilibrium situation
characterized by the alignment of the Fermi level. The conduction and valence band
bending are not shown for convenience. The minimum difference between the a-Si and Ge-
NC conduction bands is about 0.15 eV (here a-Si gap is equal to 1.7 eV), because it is well
known that the bandgap of thin amorphous silicon may be substantially increased as
compared to bulk a-Si owing to the confinement effect. This means that the bandgap value
of the a-Si could by higher than 1.7 eV. Fig. 10(b) shows the NC gap as a function of the
mean diameter obtained from calculation developed by Niquet et al. (Niquet et al., 2000) for
spherical Ge nanocrystals. The Ge-NC band gap weakly increases from the sample A35 to A21
and then A17, but takes a high value (~1.55 eV) for the sample A3.5. This behaviour is
remarkable and its consequence on electric characteristics is to lead a stronger quantum
confinement than for the other samples, which is even revealed at room temperature.

Fig. 10. (a) Schematic band diagram of the sample structure. The conduction and valence
band bending are not shown here. (b) Energy bandgap of spherical Ge nanocrystals versus
NC size; the line corresponds to theoretical calculation extracted from (Niquet et al., 2000).

4.3 Conductive-AFM measurements

I-V and HFC-V measurements show that the electron trapping process occurs in Ge-NCs.
But these techniques cannot allow the study of what happened in a single NC. They give
average information which must be completed by supplementary data on single NC. This is
the reason why we used the AFM technique to study the transport and the
trapping/detrapping process in a NC.
The study was done by conductive atomic force microscopy (C-AFM). The measurements
were performed on sample B at room temperature. The tip that we used was n-doped silicon
tip covered with p-doped diamond-like carbon (resistivity  0.003-0.005 Ω cm) and the
nominal cantilever stiffness was 2 N/m. In preliminary experiments, we first collected C-
AFM images in which the measured current was a few pA which is not enough to obtain
good resolution images. For detecting measurable current at a reasonable bias voltage
(lower than  8 V) we were obliged to preliminary irradiate the sample with the electron
Carrier storage in Ge nanocrystals grown on
silicon oxide by a two step dewetting / nucleation process 243

beam in the condition for obtaining a nano-EBIC image. Note that this phenomenon has also
been observed for InAs/GaAs quantum dots (Troyon & Smaali, 2008).
First of all, let us explain what is the nano-EBIC technique. Its principle is schematically
represented in Fig. 11.

Fig. 11. Schematic representation of the set up used for the nano-EBIC technique. The AFM
tip and the electron beam focused near this one are both immobile, whereas the sample is
scanned in the (x, y) plane.

The AFM and EBIC images are simultaneously obtained by scanning the sample with
respect to the fixed AFM probe by using the piezoelectric tube of the AFM. During image
acquisition the electron beam is focused in a fixed position just in front of the AFM probe
and as close as possible to this one. The back side of the sample is connected to the ground
and the current is punctually collected by the conductive AFM tip near the electron probe
during scanning the sample. The tip-sample contact constitutes a nano-heterojunction with
formation of a very small depletion zone which is able to collect the electron beam induced
current. The size of the depletion zone determines the resolution. We have demonstrated
that a resolution of the order of 20 nm can be obtained (Smaali et al., 2008; Troyon & Smaali,
2008).

Fig. 12. Topographic image a), b) electric image obtained at 0 V just after acquisition of an e-
beam induced current image (e-beam off) and c) second electric image obtained two hours
after in the same irradiated zone.
244 Nanocrystals

After electron beam irradiation in the condition for obtaining electron beam induced
current, C-AFM measurements without electric biasing (0 V) were performed at different
times with the electron beam off. Typical results are shown in Fig. 12. This figure gives a
comparison between three kinds of images acquired on the same area of a sample
containing NCs of 70 nm diameter: (a) is an AFM topographic image, (b) is a C-AFM image
obtained without sample bias (e-beam off) just after acquisition of a nano-EBIC image and
(c) is obtained in the same condition as image (b) but 120 minutes later. Although the tip
radius is large (~150 nm), the NCs are well resolved; this can be probably explained by the
presence of small diamond nanocrystals or grains formed at the extremity of the tip during
the fabrication process. Note that even at 0 V, 0.4 nA of current is obtained. The fact that a
current could be collected without bias is explained by the charge transfer resulting from the
alignment of the Fermi levels of the sample and probe during scanning.
The C-AFM images afterwards obtained on the previously irradiated zone show that the
current flowing through the NCs remains of the same order of magnitude as the image
taken two hours before, although the sample is not polarized. This process shows that
charges have been trapped in the NCs during the electron beam irradiation and produce a
current, which is a hole current. Indeed, the contact between the p-doped diamond AFM tip
and the sample surface forms a nano-heterojunction. This kind of nano-contact leads to the
transport and trapping of holes, but prevents electrons to flow due to the high conduction
bands discontinuity. This mechanism is explained by the energetic band diagrams of Fig. 13.
Fig. 13(a) shows the band diagram of the contact between the AFM diamond tip and the
oxide and Fig. 13(b) shows the band diagram when the AFM tip is in contact with the NC. In
both cases, electrons cannot flow because of the large conduction band discontinuity.

Fig. 13. Energy band diagrams of the probe-sample nano-contact when (a) the probe is
positioned in contact with the oxide and (b) when it is positioned above the NC the bandgap
of which (0.7~1.5 eV) is size-dependent.

During scanning, holes can move from the tip to the NC because of the weakness of the
band bending at the tip/NC interface. Coming from the substrate by tunnelling through the
oxide, some of holes are trapped in the quantum well (in the NC) and could accentuate the
NC energy band bending. With no contact between the AFM tip and the sample, holes
remain stored in the NCs because of the presence of a native oxide covering the NCs and
also of the vacuum which creates a barrier confining holes in the NCs. The hole detrapping
Carrier storage in Ge nanocrystals grown on
silicon oxide by a two step dewetting / nucleation process 245

begins when the AFM tip is in contact with the NC in two situations; unbiased and
negatively polarized tip. In the unbiased tip case, a part of stored holes can tunnel through
the little barrier from NCs to the AFM tip. While a negative polarization to the tip (which
does not affect the barrier height at the NC/tip interface) accentuates the band bending at
each interface enhancing the hole current through the structure. Whatever the sign of the
electric polarization, electrons cannot move through the system (tip/NC/oxide/silicon)
because of the strong conduction band pseudo-discontinuity added to the barrier height due
to the presence of the oxide.
As quoted above, electric conduction results from hole detrapping from NCs. This
discharging effect of Ge-NCs has been clarified by the analysis of a series of C-AFM images
recorded on one single NC just after the acquisition of a NF-EBIC image, without polarizing
the sample. Five images have been recorded every 30 min. Between two images the C-AFM
probe is retracted to avoid contact with the sample. The successive electric images show a
progressive reduction of the intensity with a total extinction of the NC at the fifth image
when the NC is completely discharged. Fig. 14 shows the evolution of the maximum current
measured in the NC versus time and on the right part the corresponding current images and
the simultaneously acquired topographic images which were extracted from larger images.
These images reveal a detrapping phenomenon. The C-AFM images obtained without
sample bias, demonstrate the capacity of the NCs to trap charges and also to release charges
after each passage of the C-AFM tip on the NCs in the condition of no sample bias. Losing
charge during acquisition of an electric image reduces the current in the next image. Indeed,
during contact between the AFM tip and an isolated NC under 0 volt bias, the Fermi levels
alignment leads to hole emission from previously charged NC because of the low effective
barrier height at this contact which facilitates the current flow. Obviously, silicon hole
tunnelling effect through the oxide is unlikely because of its thickness (~5 nm). So, after the
acquisition of the following electric images the hole current is more and more reduced.

Fig. 14. Variation of the maximum current flowing through an NC versus time in a series of
five images (V = 0) illustrating a hole detrapping process. The electric images recorded
every 30 min after irradiation, are presented on the right as well as the simultaneously
acquired topographic images.
246 Nanocrystals

The charge retention is an important and crucial parameter which requires attention in both
elaboration procedure and reliability characterization. The use of NCs to replace the
standard floating gate in memory devices is necessary since many problems can be avoided.
It is interesting to make comparison between both characterization techniques used in this
work even if two kinds of nanostructures (containing capped or no capped NCs) have been
differently studied. Both methods give information on carrier transport and storage. They
also underline dominant role of the Ge-NCs in the carrier capture and storage. However, it
is clearly seen that the C-AFM technique is required if information on an isolated NC (or a
limited number of NCs) in a given structure is needed. It would be interesting to study the
stored charge amount as a function of the NC size. This could be performed by Electrostatic
Force Microscopy technique which is our future ambition.

5. Conclusion
Ge nanocrystals grown by a two step dewetting / nucleation process on an oxide/p-(or n-)
doped silicon substrate have been electrically studied by standard methods (capacitance –
voltage and current – voltage) and by conductive atomic force microscopy technique. The
kind of the carriers involved in the electric transport and the charge storage depends on the
nature of the electrode/sample contacts. With the standard methods, for which a metallic
contact was used, electrons were concerned. The study was performed over a wide range of
temperature varying from 300 down to 77 K and results evidenced an electron storage
phenomenon in nanocrystals. The temperature effect in the enhancement of the electron
retention was revealed by the enlargement of the hysteresis cycles obtained on HFC-V and
IV measurements. Indeed, the reduction of the measurement temperature reduces the
thermo-electronic emission from NCs and facilitates carrier storage. Resonant tunnelling
effect through germanium nanocrystals with large voltage gaps was observed at room
temperature in ultradense Ge nanocrystals of ~ 3.5 nm mean size. It appeared for
increasingly low voltages when the temperature decreases. All these results are consistent
with a Coulomb blockade effect in ultrasmall Ge nanocrystals. Conditions for the occurrence
of the Coulomb blockade have been reminded.
On the other hand, the C-AFM technique using a p-doped diamond probe was used to study
uncapped Ge-NCs. The current flowing through the conductive probe was measured for
imaging local conductance, while the deflection of cantilever was optically detected for
revealing geometrical structure. The heterojunction resulting from the AFM tip/sample
contact allowed hole transport through the structure and hole storage in Ge-NCs.
Preliminary electron beam irradiation in the conditions for obtaining a nano-EBIC image
was needed for performing the C-AFM study. Without this e-beam irradiation, no noticeable
current could be detected.

6. Acknowledgments
Isabelle BERBEZIER and her group are acknowledged for providing the samples. This work
was supported in part by the European Community FP7 program in the scope of the FIBLYS
project under grant CP-TP 214042-2.
Carrier storage in Ge nanocrystals grown on
silicon oxide by a two step dewetting / nucleation process 247

7. References
Balland, B. & Barbottin G. (1989). The trapping and detrapping kinetics impact on C(V) and
I(V) curves. In: Instabilities in Silicon Devices Vol.2, Barbottin G. & Vapaille A. (Ed.),
(19-69), North-Holland, Amsterdam.
Berbezier, I.; Ronda, A. & Portavoce. (2002). SiGe nanostructures: new insights into growth
processs. A. J. Phys. Condens. Matter., Vol. 14, (8283-8331).
Conibeer G. et al. (2006). Silicon nanostructures for third generation photovoltaic solar cells.
Thin Sol. Films, Vol.511–512, (654-662).
Chatterjee, S. (2008). Titania-germanium nanocomposite as a photovoltaic material. Solar
Energy Vol. 82, (95-99)
Choi, W. K.; Kanakaraju, S.; Shen, Z. X.; & Li, W. S. (1999). Characterization of Ge
nanocrystals in a-SiO2 synthesized by rapid thermal annealing. Appl. Surf. Sci. Vol.
144-145, (697-701).
DiMaria, D. J.; Cartier, E. & Arnold, D. (1993). Impact ionization, trap creation, degradation,
and breakdown in silicon dioxide films on silicon. J. Appl. Phys. Vol. 73, No.7 (3367-
3384)
El hdiy, A.; Salace, G.; Petit, C.; Jourdain, M. & Vuillaume, D. (1993). Study of defects
induced by high-electric-field stress into a thin gate oxide (11 nm) of metal-oxide-
semiconductor capacitors. J. Appl. Phys. Vol. 74, No.2 (1124-1130).
El Hdiy, A. & Ziane, Dj. (2001). Electric bidirectional stress effects on metal-oxide-silicon
capacitors. J. Appl. Phys. Vol.89, No.2 (1405-1410).
Fowler, R.H. & Nordheim, L. (1928). Electron emission in intense electric fields. Proc. Roy.
Soc. London. Vol. 119, No. (173-181).
Gacem, K.; El Hdiy, A.; Troyon, M.; Berbezier, I.; Szutnik, P.D.; Karmous A. & Ronda A.
(2007). Memory and Coulomb blockade effects in germanium nanocrystals
embedded in amorphous silicon on silicon dioxide. J. Appl. Phys. Vol.102,
(093704/1-4).
Gacem, K.; El Hdiy, A.; Troyon, M.; Berbezier, I. & Ronda, A. (2010). Nanotech. Vol.21,
(065706/1-6).
Goldman, V. J.; Tsui, D. C. & Cunningham, J. E. (1987). Observation of Intrinsic Bistability in
Resonant-Tunneling Structures. Phys. Rev. Lett. Vol.58, No.12 (1256-1259)
Karmous, A.; Berbezier, I. & Ronda A. (2006). Formation and ordering of Ge nanocrystals on
SiO2. Phys. Rev. B Vol. 73, No.7 (075323/1-5).
Kareva, G. G.; Vexler, M. I.; Grekhov, I. V. & Shulekin, A. F. (2002). Electron Tunneling
through a Double Barrier in a Reverse-Biased Metal-Oxide-Silicon Structure.
Semiconductors, Vol.36, No.8 (889-894).
Kim, K. (1998). Visible light emissions and single-electron tunnelling from silicon quantum
dots embedded in Si-rich SiO2 deposited in plasma phase. Phys. Rev. B, Vol.57,
No.20 (13072-13076)
Kuo, Y. & Nominanda, H. (2006). Nonvolatile hydrogenated-amorphous-silicon thin-film-
transistor memory devices. Appl. Phys. Lett. Vol. 89 (173503/1-3)
Niquet, Y. M.; Allan, G.; Delerue, C. & Lannoo, M. (2000). Quantum confinement in
germanium nanocrystals. Appl. Phys. Lett. Vol. 77, No.8, august 2000, (1182-1184).
Okada, Y.; Miyagi, M.; Akahane, K.; Iuchi, Y. & Kawabe, M. (2001). Self-organized InGaAs
quantum dots on GaAs (311)B studied by conductive atomic microscope tip. J. Appl.
Phys. Vol. 90, No.1, (192-196).
248 Nanocrystals

Okada, Y.; Miyagi, M.; Akahane, K.; Kawabe, M. & Shigekawa, H. (2002). Self-organized
InGaAs quantum dots grown on GaAs (311)B substrate studied by conductive
atomic microscope technique. J. Cryst. Grow., Vol.245, (212-218).
Orwa, J. O.; Shannon, J. M.; Gateru, R. G. & Silva, S. R. P. (2005). Effect of ion bombardment
and annealing on the electrical properties of hydrogenated amorphous silicon
metal-semiconductor-metal structures. J. Appl. Phys.,Vol.97, No. (023519/1-5)
Park, N.M.; et al. (2003). Size-dependent charge storage in amorphous silicon quantum dots
embedded in silicon nitride. Appl. Phys. Lett. Vol.83, No.5, august 2003, (1014-1016)
Schuegraf, K. F. & Hu, C. (1994). Hole injection SiO2 breakdown model for very low voltage
lifetime extrapolation. IEEE Trans Electron Dev. Vol. 41, No.5 (761-767)
Skorupa, W.; Rebohl, L. & Gebel, T. (2003). Group-IV nanocluster formation by ion-beam
synthesis. Appl. Phys. A Vol.76, (1049-1059).
Smaali, K.; Troyon, M.; El Hdiy, A. And Molinari, M. (2006). Imaging the electric properties
of InAs/InP(001) quantum dots capped with a thin InP layer by conductive atomic
force microscopy: Evidence of memory effect. Appl. Phys. Lett. Vol.89, (112115/1-3).
Smaali, K; Fauré, J; El Hdiy, A & Troyon, M. (2008). High-resolution scanning near-field
EBIC microscopy: application to the characterisation of a shallow ion implanted
p+n silicon junction. Ultramicroscopy 108 (605-612).
Smaali, K.; El Hdiy, A.; Molinari, M. & Troyon, M. (2010) Band gap determination of the
native oxide capping quantum dots by use of different kinds of conductive AFM
probes: example of InAs/GaAs quantum dots. IEEE. Electron. Dev. To be published
Stomp, R.; Miyahara, Y.; Schaer, S.; Sun, Q.; Guo, H. & Grutter P. (2005). Detection of single-
electron charging i an individual InAs quantum dot by noncontact atomic-force
microscopy. Phys. Rev. Lett. Vol.94, (056802/1-4).
Szkutnik, P. D.; Karmous, A.; Bassani, F.; Ronda, A.; Berbezier, I.; Gacem, K.; El Hdiy, A. &
Troyon, M. (2008) Ge nanocrystals formation on SiO2 by dewetting: application to
memory. Eur. Phys. J. Appl. Phys. Vol. 41, (103-106).
Tiwari, S.; Rana, F.; Chan, K.; Hanafi, H.; Chan, W.; & Buchanan, D. (1995). Volatile and
Non-Volatile Memories in Silicon with Nano-Crystal Storage. IEDM Tech. Dig. Vol.
95, (521-524).
Troyon, M.; Lei, H.N.; Wang, Z. & Shang, G. (1997). A Scanning force microscope combined
with a Scanning electron microscope for multidimensional data analysis. Microanal.
Microsc. Microstr. Vol.8 (393-402).
Troyon, M.; Smaali, K.; Molinari, M.; El Hdiy, A.; Saint-Girons, G.; & Patriarche, G. (2007).
Local electronic transport through InAs/InP(001) quantum dots capped with a thin
InP layer studied by an AFM conductive probe. Semicond. Sci. Technol. Vol.22 (755–
762).
Troyon, M. & Smaali, K. (2008). Influence of electron irradiation on electronic transport
mechanisms during the conductive AFM imaging of InAs/GaAs quantum dots
capped with a thin GaAs layer. Nanotech. Vol. 19, (255709/1-7)
Vexler, M. I.; El Hdiy, A.; Grgec, D.; Tyaginov, S. E.; Khlil, R.; Meinerzhagen, B.; Shulekin, A.
F. & Grekhov, I. V. (2006). Tunnel charge transport within silicon in reversely-
biased MOS tunnel structures. Microelectron. J., Vol.37, (114-120).
Wu, L.; Dai, M.; Huang, X.; Zhang, Y.; Li, W. Xu, J. & Chen, K. (2004). Room temperature
electron tunneling and storage in a nanocrystalline silicon floating gate structure. J.
Non-Cryst. Sol. Vol. 338-340, (318-321).
Nano-crystals for quadratic nonlinear imaging: characterization and applications 249

11
X

Nano-crystals for quadratic nonlinear imaging:

characterization and applications
Sophie Brasselet
Institut Fresnel, Domaine Universitaire St Jérôme, 13397 Marseille Cedex 20
France

Joseph Zyss
Laboratoire de Photonique Quantique et Moléculaire, Institut d’Alembert, ENS Cachan,
61 av. du président Wilson, 94235 Cachan
France

1. Introduction
Nonlinear optics is a well established field today, covering a large and rich range of
applications which expands every year. Its extension to sub-wavelength scale processes is
however highly non trivial, and nano-scale nonlinear optics is only in its infancy.
Downscaling high order light-matter optical interaction is however nowadays accessible
using new technological tools such as near field techniques, high resolution microscopy and
pulsed femtosecond lasers. As a consequence, a large amount of effort has been recently
invested into the invention and engineering of nano-structures that exhibit nonlinear optical
properties and lead to new optical functions. Among the promising routes followed by
nonlinear nano-optics, the development of new nano-sources originating from frequency
mixing processes is particularly successful. In this chapter, we describe how nano-crystals
have been advantageously developed and used in Second Harmonic Generation (SHG) (or
by extension Sum Frequency Generation : SFG), which is the lowest order nonlinear process,
dependent on the square of the incident field (Boyd, 1992). Recent developments have
shown that these nano-structures are potentially key elements in various fields, such as new
nano-probes for bio-imaging, or nano-scale optical fields probing in the ultra-short pulses
regime. They also combine the interesting properties of frequency mixing processes with the
nano-scale regime: unlike the resonant fluorescence process, coherent harmonic generation
is active in the non-resonant regime and therefore free from photo-bleaching which is
otherwise generally considered a major drawback from fluorescent nano-probes today. The
coherent nature of frequency mixing also allows generating different wavelengths and
therefore creating nano-sources of a large range of possible frequencies. An additional
advantage of these nano-sources is to avoid phase-matching constraints since their size is
well below the coherent length of the underlying nonlinear process, at which nonlinear
propagation suffers from destructive interferences between the propagating and induced
nonlinear waves. The availability of a large emission frequency range finally allows imaging
250 Nanocrystals

bright nano-emitters over a dark background in complex environments sometimes

contaminated by auto-fluorescence such as in living cells.
The coherent build up of an induced dipole at twice the frequency of the incident field
demands a material exhibiting a non-centrosymmetric structure, which explains why
crystals featuring such a structure, although not a majority, have been such determining
materials in nonlinear optics to these days, both as a source of inspiration as well as towards
applications (Boyd, 1992; Chemla & Zyss, 1987).

As shall be reviewed in the first section of this chapter, downscaling non-centrysymmetry to

the nano-scale has been achieved following different possible routes, covering both bottom-
up to top-down approaches, in organic as well as in inorganic structures. These nano-
probes, which are the nano-scale counterparts of the traditional nonlinear active bulk
materials, require specific attention with respect to material engineering. Indeed working at
nanometric scales requires revisiting the basic rules of nonlinear optics such as phase
matching and nonlinear coupling optimization, which lead to original light-matter
interaction processes. Very much like in linear optics, working with nano-objects in
nonlinear optics is therefore a research field in itself.
While the study of nonlinear bulk crystals has largely developed along the search for
macroscopic nonlinear coupling directions and phase matching conditions, the investigation
of nonlinear nano-crystals requires a complete different approach, furthermore constrained
by the lack of control of the orientation of the object in the macroscopic framework. In this
context, determination of the orientation of the nanocrystal with respect to the laboratory
framework sets a problem by itself, thus in considerable contrast with the case of bulk
crystals which can be in most cases cut and oriented at will. We will describe in the second
section of this chapter the different techniques which have been advantageously developed,
either based on a randomization of the nano-crystals orientation in solution, or by
immobilization on a substrate by nonlinear microscopy.
Finally, the last section of this chapter will summarize the actual status of applications of
nonlinear active nano-crystals, both for bio-imaging using nonlinear microscopy in living
cells and tissues, and for nanophotonics, in particular for nano-scale optical fields probing.

2. Nano-crystals development for Second Harmonic Generation

The conception and fabrication of crystals with nanometric to sub-micrometric size
exhibiting strong nonlinear quadratic optical responses is a topic of growing interest. In
particular, Second Harmonic Generation (SHG) active nano-crystals offer a broad range of
potential applications from nanoprobes for bio-imaging to nano-scale photonics with
original optical properties. SHG, which provides the ability to transform an exciting optical
frequency into its harmonic, when transposed to the nano-scale, could provide new types of
light sources and labels with (i) an emission wavelength far away from its excitation (in
particular in the infra-red range, which is better adapted to bio-imaging in tissues) (ii) a
lower photobleaching probability due to the possibility to work far from resonances, (iii)
possibility to tune the emission wavelength by tuning the incident one, or by performing
frequency mixing of different excitations, (iv) a less stringent dependence, compared to bulk
crystals, on phase matching conditions.
Nano-crystals for quadratic nonlinear imaging: characterization and applications 251

The first prerequisite condition when engineering a SHG-active nano-object is to obtain a

non-centrosymmetric structure. As seen in the examples below, a large amount of researches
have lead to new routes of fabrications of SHG-active nano-crystals, in order to meet this
requirement: either by using crystal unit-cell structures deprived of center of symmetry in
inorganics, or engineering a non-centrosymmetric packing of non-centrosymmetric
molecules in molecular organic crystals (Chemla & Zyss 1987). As a positive consequence,
the coherent emission from SHG-active nano-crystals is supposed to be weakly dependent
on the shape of the nano-structure, contrary to metallic nanoparticles which rely on the non-
centrosymmetry breaking at their interface. Nevertheless, such shape dependence remains a
research topic of interest for nano-objects of the order of a significant fraction of the
wavelength (typically from a few ten’s to a few hundred’s of nanometers). In addition, these
structures are expected to exhibit a small size-dispersion, to be amenable to manipulation in
solution, and ultimately to be bio-compatible, which imposes low chemical reactivity, low
toxicity, stability in aqueous environment, and possibly surface functionalization. They are
finally also expected to lend themselves via adequate biomineral or bio-organic linkage, to
tethering to some biological material of interest, such as towards the SHG monitoring of
intra or inter cellular traffic of a DNA fragment, a biotine-avidine linkage moiety, a protein,
an antigen-antibody pair or a deactivated virus. As described in this section, various
strategies have been developed to achieve stable and efficient SHG nano-emitters, either
from a top-down or from a bottom-up approach.

2.1 Organic nano-crystals

Molecular based nano-crystals are the first structures developed as SHG-active nano-
materials. Based on molecular nano-assemblies, they have been designed starting from SHG
active asymmetric molecules as individual building blocks, and further grown from a
bottom-up approach. The fabrication of these novel nanomaterials constitutes an alternative
approach to the synthesis of multichromophoric architectures, which has been a trend over
the last decade (Yokoyama et al. 2000, Rekaï et al. 2001, Le Bozec et al. 2001). Molecular-
based assemblies present furthermore the advantage of a large flexibility with respect to the
nanoparticle nature and the active molecule used. The fabrication procedure follows an
essential requirement, which is to conserve the molecular non-centrosymmetry at the
nanometric scale, in addition to a high molecular density packing, which ensures SHG
efficiency. Assuming indeed a nano-object consisting of n molecules of induced nonlinear
dipoles oriented in the same direction, the overall SHG signal is proportional to n2, which
can lead to measurable signals in nanostructure of tens of nanometer sizes (Brasselet & Zyss
2007) (see Section 3).

One of the first reports of second-harmonic emission for nanoscale molecular crystals (Shen
et al., 2000) was referring to the emblematic NPP structure (NPP for
N-(4-nitrophenyl)-(L)-prolinol)) whereby the usual trend towards centrosymetry had been
fought by introduction into the molecular structure of a chiral amino-acid fragment (proline
alcool derivative) (Zyss et al. 1984) and which had marked a milestone in nonlinear
molecular crystals as the first infrared optical parametric oscillator (Josse et al. 1992).
Another early example was based on the non-centrosymmetric CMONS molecule (-[(4’-
methoxyphenyl)methylene]-4-nitro-benzene-acetonitrile, a polar donor– acceptor nonlinear
structure) (Figure 1a), from which nano-crystals have been fabricated by rapid nucleation of
252 Nanocrystals

the molecules, confined in the nano-pores of a silica network sol-gel (1:1 tetramethoxysilane
(TMOS) : methyltrimethoxysilane (MTMOS)) formed by spin-coating (Sanz et al. 2001). The
size of the nano-crystals, ranging from 20 to 100nm with a remarkable monodispersity, was
controlled by the growth kinetics and parameters such as temperature, matrix porosity, and
the CMONS:alkoxides molar fraction (Sanz et al. 2001, Treussart et al. 2003). The study of
individual nano-crystals by SHG microscopy has confirmed the non-centrosymmetric
structure of the nano-objects, which was consistent with the bulk symmetry (Brasselet et al.
2004). Although these structures showed very efficient SHG activity in addition to two-
photon fluorescence, they were embedded in a gel film and therefore not useable for future
labelling and manipulation.

Developing isolated nano-objects which can be manipulated in solution has been achieved
in MnPS3/DAZOP particles of 10nm size, based on the hybrid bottom-up synthesis of an
intercalation compound consisting of push-pull organic chromophores (DAZOP : (4-[2-(4-
dimethylaminophenyl)azo]-1-methylpyridinium) embedded in a layered manganese
hexathiohypodiphosphate (MnPS3) inorganic host lattice (Yi et al. 2004) (Figure 1b). The
DAZOP non-centrosymmetric ferroelectric-like arrangement is obtained within the layered
MnPS3 matrix taking advantage of ionic interactions operating during the particle formation
within nanoreactors provided by inverse micelles (typically using Brij97 as a surfactant).
This strategy has lead to particles with an average second order hyperpolarisability  of the
order of 42.10-27 esu, measured by Hyper Rayleigh Scattering (HRS) (Yi et al. 2005) (see
Section 3). This value is three to four orders of magnitude higher than the  coefficient
typically measured in nonlinear molecules. Studies by SHG microscopy (see Section 3) on
individual particles immobilized in PVA (polyvinylalcohol) spin coated thin films have
shown that the particles tend to form aggregates of larger size (about 100nm), still
preserving a quasi one-dimensional order (Delahaye et al. 2006, Delahaye et al. 2009). The
stability of the SHG emission from these particles was seen to be strongly dependent on the
surfactant used for the synthesis, suggesting an environment dependence of the particle
photo-damage (Delahaye et al. 2009).

(a) (b)
MnPS3
N N
+
N N
N N
N
N
N
+
N
N N
+ DAZOP

MnPS3

20 nm

CMONS
TEM (H20)
Fig. 1. (a) Unit-cell structure of the CMONS macroscopic crystal in its non-centrosymmetric
form. (b) MnPS3/DAZOP nano-crystals : scheme, fabrication procedure and Transmission
Electron Microscopy Image.

Top-down approaches have also been attempted to reach molecular nano-crystals. Crystal
grinding has been tested on the very efficient co-inclusion compound based on the
Nano-crystals for quadratic nonlinear imaging: characterization and applications 253

confinement of guest non-linear active molecules within the host polycyclic hydrocarbon
perhydrotriphenylene (PHTP) channels (Komorowska et al. 2005). The photoactive
molecules DANA (4-Dimethylamino-40-nitroazobenzene), used as guest systems, lead to
typical bulk values of 340 pm/V. Particles of sizes ranging from 400 to 1000 nm (for which
the goal of mono-dispersion had not been pursued at this stage) could be obtained after fine
grinding in a PVA-water solution, followed by sonication. Adequate dilution before spin-
coating allowed obtaining isolated particle active for SHG and Two Photon Fluorescence.
Another interesting structure, building-up on the concept of octupolar symmetry (Zyss &
Ledoux 1994) has been developped based on TTB (1,3,5-tricyano-2,4,6-tris(p-
diethylaminostyryl)benzene), a third-order symmetry molecule forming highly efficient
nonlinear crystals (Le Floc’h et al. 2005). Micro- and nano-crystals have been investigated
(Brasselet & Zyss 2007). 3D Octupolar symmetry evidences in principle an optimized
structure whereby the SHG response does not depend neither on the incident polarization
nor on the orientation of the object (Brasselet & Zyss 1998). However in the case of TTB, the
planar geometry still led to a highly anisotropic nonlinearity (Le Floc’h et al. 2005).
Resorting to a fully 3-D octupolar structure (Zyss & Ledoux 1994) would allow generating
an orientation independant SHG response, which is of considerable interest for nano-crystal
tracking for instance.

Among other possibilities, the fabrication of molecular aggregates from micelles structure
based on the mixing of surfactant and active molecules has also been successful to generate
non negligible SHG signals (Eisenthal 2006). These structures, which size can be easily
engineered, are interesting model systems to study the contribution of surface effects in the
nonlinear radiation from sub- micrometric size objects (Revillod et al. 2008).

2.2 Inorganic nano-crystals

While organic nano-crystals can exhibit a wide range of absorption wavelengths governed
by chemical synthesis, inorganic nanostructures are often transparent from the visible to 12
m. A large amount of studies is currently undertaken in order to fabricate robust,
transparent, mono-disperse functionalizable inorganic nano-crystals, with typical sizes
ranging from 10 to 300 nm. Since the nano-structures described below exhibit very similar
nonlinear coefficients (typically in the range 10-30 pm/V in bulk), the goal of the report
below is not to discuss their relative efficiencies but rather to give an overview of the
fabrication and application routes exploited so far. Efficiency measurement issues are
discussed in Section 3 of this chapter.

Zinc Oxide (ZnO) nano-crystals are among the first inorganic nano-crystals studied for
nonlinear optics (Johnson et al. 2002). Made of a nontoxic and biocompatible semiconductor
material, ZnO nano-crystals were later synthesized in organic media by using a
nonhydrolytic sol−gel process and subsequently dispersed in aqueous media using
phospholipid micelles, leading to successful encapsulation such as used for drug delivery
(Kachynski et al. 2008, Kuo et al. 2009). Further incorporation with the biotargeting molecule
folic acid makes this structure adapted to a wide range of targeted bio-imaging studies. ZnO
nano-crystals of <100nm size have been synthesized, with additional X-ray diffraction and
Transmission Electron Microscopy (TEM) ascertaining for their tetragonal (and pyramid
shape) structures (Kachynski et al. 2008). ZnO exits also in the form of nano-wires of typical
254 Nanocrystals

1 m lenths with hexagonal cross-sections, which have been studied also for bio-imaging
applications (Nakayama et al. 2007).

The non-toxicity of Iron Iodate has motivated the fabrication of Fe(IO3)3 nano-crystals of
typical size 20-300nm, by co-precipitation synthesis (Galez et al. 2006). The Fe(IO3)3 material
belongs to the hexagonal space group P63 with nonlinear coefficients in bulk of about
10pm/V off resonance. Nano-crystals of 30 nm sizes have been seen to aggregate in larger
clusters (80nm) when the solvent is evaporated on a surface (Bonacina et al. 2007).
Single-crystalline KNbO3 nano-wires of rectangular cross-section of 40nm-400nm x 1-20 m
sizes (depending on the reaction time) could be obtained by a hydrothermal method, mixing
at high temperature Potassium hydroxide and niobium pentoxide in deionized water
(Magrez et al. 2006). Structural characterizations showed a conserved orthorhombic phase
(mm2) with the growth axis of the wires parallel to the [011] direction. A rough estimate for
the averaged nonlinear coefficient of KNbO3 nano-wires lead to equivalent bulk values of
about 10-30 pm/V, as expected in KNbO3 crystals. These KNbO3 nano-wires have been
applied to trapping and bio-imaging in living cells (Nakayama et al. 2007).

Aside from these bottom-up techniques, nano-crystals have also been obtained from powders,
with a good size dispersion ascertained by Dynamic Light Scattering (DLS) experiments.
KTiOPO4 (KTP) nano-crystals, isolated from a flux-grown powder extracted from a bulk
crystal, can be commercially provided (Cristal Laser). These nanocrystals are monodisperse
in size (with sizes ranging between 10nm and 80nm). KTP belongs to the orthorhombic
point group mm2 with typical nonlinear coefficients ranging from 2 to 17 pm/V off
resonance. Thorough studies have been undertaken on isolated KTP nano-crystals, in order
to measure small size crystals (Le et al. 2006), infer their three-dimensional orientation on
substrates (Sandeau et al. 2007), and relate SHG and size information using combined
atomic force microscopy and optical techniques (Le et al. 2008).
At last, BaTiO3 nano-crystals of 30nm and 90nm sizes were obtained from commercial dry
powders (Nanoamor (30nm) and Techpower (90nm)). Specific studies have allowed a good
control of the dillution behavior of the particles in solution to further bio-imaging
applications (Hsieh et al. 2009, Hsieh et al. 2010). By dispersion in aminomethylphosphonic
acid and sonication of the colloidal solution, the phosphonic acid reacts on the particle
surface, ending up with an amine group coating, which introduces an electrostatic repulsion
between the particles. BaTiO3 is of tetragonal structure with similar range of nonlinear
coefficients as in the crystals reported above. Measured efficiency cross section on single
nano-crystals has been evaluated and seen to surpass, by orders of magnitude, the typical
fluorescence cross section from traditionally used molecules in two-photon fluorescence
(Rodriguez et al. 2009b, Pu et al. 2008, Hsieh et al. 2009).
In all the examples above, considerable stability and efficiency of the measured signals could be
observed, as compared to traditional nano-probes used in fluorescence. It is therefore not
surprising that many of the previously described structures (ZnO, Fe(IO3)3, BaTiO3 and KNbO3)
have been directly applied to imaging in biological media (see Section 4 of this chapter).

Semi-conductor quantum dots are now the smallest existing structures active for SHG with
typical sizes ranging from 2 to 15 nm. While semiconductor quantum dots are widely
known and used in fluorescence microscopy as bio-markers, their use for nonlinear
Nano-crystals for quadratic nonlinear imaging: characterization and applications 255

radiation has been much less studied. CdS and CdTe cadmium based materials are both
based on a zinc blende crystal structure of cubic Tetrahedral Td symmetry. CdS
nanoparticles of about 5 to 10 nm were first synthesized by co-precipitation reaction
between an aqueous solutions of Cd(NO3)2 and Na2S in the presence of hexametaphosphate
(HMP) as a stabilizer (Fu et al. 2001, Wang et al. 2005). Measured hyperpolarizabilities in
solution by HRS were found to be up to 7.24×10-26 esu, which is four orders of magnitudes
higher than typical nonlinear molecules. This value was found to be higher than the one
expected from a pure bulk response (typically 78pm/V off resonance), indicating possible
surface effects (Fu et al. 2001).
CdTe/CdS nanocrystals with a diameter of 10 to 15 nm have been recently studied by SHG
microscopy, immobilized in a PMMA (Polymetylmetacrylate) thin film (Zielinski et al.
2009). Their synthesis is based on a progressive crystalline growth in a noncoordinating
solvent, by regular injection of cadmium oleate solution and trioctylphosphine/Te followed
by trioctylphosphine/S. Polarization analysis of the second-harmonic emission confirms the
expected zinc blende symmetry (see Section 3). The nonlinear efficiency of these structures is
expected to come predominantly from the core (with a bulk CdTe nonlinear coefficient of
about 150 pm/V), even though the shell is also non-centrosymmetric. In addition to the
possibility to observe very small nano-crystals thanks to a high efficiency, intriguing
wavelength dependence of the SHG efficiency indicated a probe of deep resonant levels
mechanisms (Zielinski et al. 2009).

3. Optical characterization of SHG active nano-crystals

SHG is a multiple field tensorial coupling process, involving both the crystal symmetry and
the incident fundamental as well as outgoing harmonic fields polarizations. This coupling
has been accounted for by introduction of the nonlinear tensor field and its contraction with
the molecular susceptibility tensor, a procedure that can be generalized to any order,
leading to a convenient invariant scalar formulation (Boulanger et al. 1997, Brasselet & Zyss
1998). Furthermore, in objects smaller than the focal excitation volumes, the SHG efficiency
grows with the object size, as expected from the higher number of active molecules/unit
cells (Brasselet 2010). Therefore the analysis of a SHG signal from an isolated object, and in
particular the measure of its efficiency, relies on the knowledge of its orientation, its
symmetry, and its size relative to the excitation focal volume. While size information can
possibly be known on average by techniques such as Dynamic Light Scattering (DLS),
Transmission Electron Microscopy (TEM), or Atomic Force Microscopy (AFM), symmetry
information can be inferred from bulk structures or more rarely from X-ray diffraction
studies on an ensemble of nano-objects. Although these measurements provide averaged
information, they can be still valuable as a first input for SHG investigations from single
nano-crystals. In particular the nano-crystals symmetry will often be considered to be the
same as for the bulk, thus preserved at the nano-scale.

In a planar wave approximation, the overall measured SHG intensity of a single nano-
crystal originates from the coherent sum of n induced nonlinear dipoles, n being the number
of molecules (or crystalline unit-cells) contained in the nano-crystal present in the focal
volume of excitation (Delahaye et al. 2009, Brasselet 2010). The macroscopic components of
256 Nanocrystals

these induced nonlinear dipoles, projected in the (X,Y,Z) macroscopic frame, are
proportional to:

PI2    n  J , K  IJK  E J E K (1)

with (I,J,K) = (X,Y,Z), and  IJK   being the SHG tensorial component of the crystal unit-
cell (or unit molecule) oriented along the Euler set of angles    ,  ,  . E J is the
projection of the incoming field polarization at the fundamental frequency on the J
macroscopic direction. From a microscopic point of view,  IJK   originates from the
projection of the microscopic unit-cell nonlinear hyperpolarizability  ijk ((i,i,k)= (1,2,3) being
the unit cell frame), projected in the macroscopic frame, and can be written (Zyss et al. 1982):

     (2)
 IJK      i , j ,k  i .I      j .J      k .K      ijk

 
with i.I   the cosine projection factor between thei (microscopic) and I (macroscopic)
axes. Note that d ijk  N ijk / 2 are the traditionally used nonlinear coefficients in a bulk
crystal with N the molecular (or unit-cell) density. They can therefore be used as known
numbers in this expression. Eq. (2) is inferred from the oriented gas model (Zyss et al. 1982),
widely used in the understanding of nonlinear responses from organic materials. It assumes
in particular that there is no significant interaction between the molecules which therefore
behave such as uncorrelated nonlinear emitters, which has been a realistic approach in
many examples, subject of course to adequate local field corrections, such as the classical
Lorenz-Lorentz or Onsager terms.

As described in the next Sections, two measurements of the macroscopic information from
the nonlinear induced dipoles radiation can be performed:

- An incoherent summation over a large collection of nnc nano-crystals, all

macroscopic dipoles PI2   being oriented in random directions and positions,
from either time or spatial (or both) fluctuations (see Section 3.1). In this case the
overall intensity analyzed along the I polarization direction is proportional to:

I I2  PI 2     nnc  J ,K n 2  IJK     ILM    EJ EK EL* EM*  nnc .n 2 .  2 . I  

2 2


(3)


With 2 being the averaged measured nonlinear coefficient. The incoherent

nature of the process imposes in this case that the signal is proportional to nnc , the
number of nano-crystals in the focal volume.

- A direct coherent measurement of the macroscopic radiating dipole PI2   from a

single nano-crystal oriented in the  direction (see section 3.2):
Nano-crystals for quadratic nonlinear imaging: characterization and applications 257

I I2     PI 2     n  J , K  IJK   E J E K  n 2 . 2 .  I  
2 2 2
(4)

In both cases the signal is proportional to n2 , with n the number of unit-cells (or molecules) in
the nano-crystal. Since n = N.V with V the nano-crystal volume and N the crystal density, the
measured signals are proportional to the sixth power of the nano-crystal diameter. Note that
rigorously, V is the overlap between the excitation volume and the nano-crystal volume itself.

3.1 Averaging SHG information

In order to measure SHG efficiency from nano-crystals while circumventing the issue of
their orientation knowledge, studies have been performed using orientation-averaging
techniques. Averaging SHG information is typically performed in a powder measurement,
which is an historical way to analyse SHG efficiency from molecular micro-crystals (Kurtz et
al. 1968, Halbout et al. 1984). Powder analyzes have been performed for instance on
ferroelectric Strontium barium niobate nanoparticles of 40 nm size, made from SrxBa1–
xNb2O6 (SBN) (Rodriguez et al. 2009). The drawback of this technique is the lack of
knowledge on the possible formation of macroscopic clusters.
Another traditional technique, Hyper Rayleigh scattering (HRS) (Terhune et al. 1965), relies
on the measurement of the nonlinear scattering of the nano-objects in solution, which is
more amenable to environments where the nano-crystals can be isolated from their
neighbours by electrostatic repulsion. HRS, which has been a determining tool for molecular
engineering in nonlinear optics (Zyss & Ledoux 1994) including molecular based nano-
objects (Le Bozec et al. 2001), is a robust technique to infer averaged nonlinear coefficients
from nano-crystals, using known solvents as references. Since the obtained signal is a result
of time, spatial and orientational averaging over fast diffusing molecules or particles, this
method does not require a preliminary orientation of molecules or particles before study. It
has been applied to CdS particles (Fu et al. 2001, Wang et al. 2005), molecular aggregates
(Eisenthal et al. 2006, Revillod et al. 2008), hybrid nanoparticles (Yi et al. 2005), BaTiO3 and
PbTiO3 nano-crystals (Rodriguez et al. 2009b), and more recently FeTiO3 nano-crystals
(Mugnier et al. Submitted).  2 averaged coefficients up to more than 10-26 esu have been
measured, which is orders of magnitudes above single nonlinear molecules, due to their size
extension into larger scale aggregates (as a comparison a DAZOP molecule efficiency is
typically 220.10-30 esu (Yi et al. 2005)).

While these techniques are powerful to provide an averaged efficiency estimation of nano-
crystals, they are however strongly limited when analysis is needed at the single nano-object
level, which is the relevant one towards nano-crystal engineering. In particular, no
particular knowledge can be inferred on the crystalline quality of isolated crystals, neither
on their possible behaviour distributions.

3.2 SHG imaging of single nano-crystals

SHG microscopy, initially introduced as an imaging tool to visualize the microscopic crystal
structure in polycrystalline ZnSe (Hellwarth, 1974), allows today investigating nano-crystals
one by one, provided they are sufficiently distant from one another (typically by 1 m) on a
glass substrate or in a polymer thin film. SHG microscopy is based on the illumination of a
258 Nanocrystals

nano-object by focussing the light of a pulsed laser (for instance a Ti:Sa laser, tuneable
between 690 nm and 1050 nm (150fs pulse width, 80MHz repetition rate), using a high
numerical aperture objective (NA ranging between 0.9 and 1.4). The focused light forms a
focal volume of typically 300 nm in the lateral dimension and 700 nm in the axial direction.
In a typical SHG microscope, the SHG emission from nano-objects is collected backward
through the same objective as used for the fundamental excitation, then filtered-out
spectrally to possibly remove remaining laser light and fluorescence emission, and focused
on avalanche photodiodes (Figure 2a). The final image of nano-crystals is generally obtained
by scanning either the sample on a piezoelectric mount or the excitation beam by
galvanometric scanners, over typically 20 m to 100 m dimensions (Figure 2a) (Brasselet et
al. 2004, Brasselet & Zyss 2007). Other implementations have been performed based on the
phase information from the SHG signal in single nano-crystals, taking advantage of the
coherent nature of this optical process. Based on interferometric measurements with a
reference signal coming from the SHG field of a macroscopic nonlinear crystal, these
implementations have allowed either a direct phase measurement in a homodyne detection
set-up (Figure 2b, Le et al. 2006), or 3D imaging using digital holographic image
reconstruction (Figure 2c, Pu et al. 2008). The first scheme has allowed measuring KTP nano-
crystals below 30nm size, benefiting from the homodyne detection sensitivity provided by
the local oscillator (Le et al. 2006). The second provides a scheme for a scan-free 3D imaging.
In this case sensitive CCD camera is placed away from the object plane, and the image is
added coherently with a reference SHG plane wave. By digital propagation in the observed
interference pattern, the field is reconstructed at any plane, with a diffraction limited
resolution (Pu et al. 2008).
In the context of quadratic nonlinear microscopy, let us also mention a parent technique,
also resorting to a frequency mixing process, the Pockels linear electro-optic effect that is
leading to an  optical frequency modulation, where and stand respectively for the
higher frequency of the illuminating optical laser field and for the lower one attached to an
externally applied electric voltage (herein ranging from the kHz to a few tens of MHz). In an
interferometric read-out configuration using a sensitive homodyne detection to boost the
modulated signal while minimizing the intensity actually shining the material, this
configuration turns out to be compatible with low cw power (typically 100µW). This effect
has been implemented and applied to poled polymers (Toury et al. 2006), doped artificial
membranes (Hajj et al. 2009), and more recently to KTP nano-crystals down to 100 nm size
(Hajj et al. Submitted).

Obviously a SHG image of many isolated nano-crystals exhibits a wide range of intensities
(Figure 2a), for several reasons:
- This is first due to the size variations between nano-crystals, their efficiency scaling
with the sixth power of their diameter, as detailed above. In addition an estimation
of size variations in the 10-100 nm range is impossible: any size below the
diffraction limit exhibits an image spot of roughly 300 nm size, and only crystals
sizes above this limit can be discriminated.
- Even in a mono-disperse population, intensity variations occur from the wide
range of possible orientations of nano-crystals in the sample, coupling with
different efficiencies to the incident polarization as seen in Eq. (4).
Nano-crystals for quadratic nonlinear imaging: characterization and applications 259

(a) (b)

CCD
Radiation pattern
detection

SHG (exc 900nm) counts / 10ms


Y E
140

120

X 100

80

60

40

20

1m

(c)
APD
IY
APD IX Y 
E
Polarized 
detection
X

Fig. 2. (a) Imaging and polarization resolved nonlinear microscope. A polarizing

beamsplitter separates the signal towards two avalanche photodiodes (APD). SHG
polarimetry consists in recording the SHG emission while the incident linear polarization is
rotated (angle ) on two orthogonal analysis directions (IX and IY are represented in polar
plots in red and green respectively for a given nano-crystal). A typical SHG image (from
Brasselet 2010) is shown from a thin PVA polymer film containing MnPS3 particles
(MnPS3/DAZOP/Brij97, 10-4 mol/L DAZOP). Excitation wavelength : 920nm (a SHG
spectral filter is placed in the detection path at 460 nm - excitation power: 1 mW-SHG signal
scale : counts/50ms). A defocused imaging (above) is shown from a 80 nm size nano-KTP
(Sandeau et al. 2007). (b) (from Le et al. 2006) SHG balanced homodyne detection set-up. (c)
(from Hsieh et al. 2009) Harmonic Holography set-up. On both cases, a reference arm is
created by SHG in a BBO (-barium borate) crystal. While the homodyne set-up restores the
information from interferences fringes observed by modulating a mirror position in the
interferometer, the holography set-up images directly the phase information in a wide field
interference pattern, which is used for further reconstruction imaging.

3.3 Determining orientation and crystallinity information from a single nano-crystal

Measuring a single nano-crystal nonlinear efficiency is therefore not direct and requires
careful investigation of the SHG response. In order to approach this issue, knowledge of the
SHG polarization coupling mechanism in the nano-crystals is required. In addition to SHG
imaging, a full polarimetric analysis has been therefore developed based on the recording of
the SHG emission signal under a continuous rotation of the incident linear polarization in
the sample plane (Brasselet et al. 2004). The polarization rotation is performed by an
achromatic half-wave plate mounted on a rotating step motor, while the signal is recorded
on two polarization directions, using two avalanche photodiodes separated by a polarizing
beamsplitter (Figure 2a). The orientation () and symmetry ( tensor) information in the
nano-crystal are contained in the  IJK  coefficients of Eqs. (2) and (4). The principle of
polarization-resolved SHG is to tune the field contributions E J by rotating a linear
E J
260 Nanocrystals

polarization in the (X,Y) sample plane (Figure 2a). Typical polarization responses projected
on the X and Y directions are illustrated in Figure 2a and Figure 3 on a 1D symmetry nano-
crystal. Very anisotropic polarization responses can be observed in the direction of the nano-
object, as expected, whereas a symmetry containing different orientations, such as in a 1D
nano-object cluster, would lead to more complex polarimetric response (Figure 3b).
It is therefore possible, providing that the nano-crystal symmetry or orientation is known, to
discriminate mono-crystalline structures from non-crystalline ones (Brasselet et al. 2004,
Komorowska et al. 2005, Brasselet & Zyss 2007, Brasselet 2010). A non-crystalline structure
is seen in this case as a collection of nonlinear dipoles oriented in different  directions
adding up incoherently. This information, which is not accessible from an ensemble
measurement, is of paramount importance towards further nano-crystals synthesis. SHG
polarimetry has been successful in investigating molecular ordered materials (Le Floc’h et
al. 2003, Anceau et al. 2005). Recently, demonstrations have shown its potential to provide
orientational and structural behaviour of single isolated CMONS nano-crystals (Brasselet et
al. 2004), as well as hybrid (Delahaye et al. 2006, Delahaye et al. 2009) and inorganic
(Zielinski et al. 2009) nano-crystals.

a) b) E
X

nano-particle
 Y
IX
Y
X  Sample plane IY

Fig. 3. (a) Representation of the Euler set of angles  (,defining the orientation of a
nano-crystal of high symmetry axis designated by the black arrows. In the case of a nano-
object abiding to cylindrical (or higher) symmetry as here, one needs only two Euler angles,
the third  rotation around the symmetry axis of the nano-object being then unnecessary. (b)
Polarization-resolved SHG response from a single 1D symmetry nano-crystal (above) versus
a collection of different orientations of 1D symmetry nano-crystals (below). The polarization
response is represented as a polar graph in the projection sample plane (X,Y), projected
along the X and Y polarization analysis directions.
Several observations must be added to the polarimetric analysis discussed above, including
considerations which are crucial towards a proper analysis of polarization dependent SHG
data from single nanocrystals.

- Instrumentation distortions by reflection optics. In most of the measurements a

polarization analysis requires rotating a linear polarization before reflection on a dichroic
mirror or other mirrors (Le Floc’h et al. 2003; Chou et al. 2008; Schön et al. 2008). It is know
that these optics, although often corrected for a reliable s/p (or X/Y) intensity reflection ratio,
can strongly distort any intermediate polarization angle by applying a phase shift between
the X and Y components. The input polarization, becoming elliptic, needs to be correctly
accounted for in the analysis of Eq. (4). This can be done by preliminary calibrations (Le
Floc’h et al. 2003, Schön et al. 2008).
Nano-crystals for quadratic nonlinear imaging: characterization and applications 261

- Effect of the high aperture excitation/collection. Eq. (4) is written in the plane wave
approximation. Observation of single nano-crystal however requires high numerical
aperture (NA) objectives for both the excitation and collection of the SHG radiation which
requires a more rigorous analysis. Indeed, under such conditions, an incident field
polarization in the (X,Y) plane is known to contain also Z contributions in the focal plane,
that can reach up to 45% of the incident amplitude at the edge of the focussing spot
(Richards & Wolf 1959). Depending on the nano-crystal off-plane orientation, the SHG
signal therefore also contains Z-coupling nonlinear coefficient contributions that can
strongly alter its polarization response (Yew & Sheppard 2006, Schön et al. 2010). Such
contributions are however located spatially at the border of the incoming Gaussian field
amplitude, and may induce extra nonlinear coupling terms only for objects reaching 100 nm
to 150 nm size. The high aperture collection also mixes-up polarization states of the emission
radiation. Both effects can be fully taken into account by a complete model calculating the
nonlinear induced dipole at every point of an object placed at the center of the focal spot of
the objective (Sandeau et al. 2007, Schön et al. 2010). In general nano-crystals of sizes well
below 100 nm will only experience slight modifications of their polarization responses
under high NA conditions.
- Effect of the spatial extension of the nano-crystal. Another limitation to Eq. (4) is that it
considers that the coherently added dipoles are all positioned spatially on the optical axis of
the microscope objective, thus ignoring possible phase retardation effects in the volume of
the object. Extensive studies accounting for the spatial extension, of a nano-crystal, have
shown that for sizes above a few tens of nm, the radiation does not resemble that of a single
macroscopic dipole volume (Sandeau et al. 2007), in particular the backward emission is
considerably reduced comparing to the forward emission, and phase matching conditions
have been re-written accounting for the Gouy phase shift occurring in the excitation volume
(Sandeau et al. 2007). Whereas the “punctual object” assumption might be crucial for
efficiency measurements in nano-objects of unknown size distribution, it does not however
alter the polarization resolved microscopy analyses which are relative measurements.
- Limitation to a 2D projection information. The information provided by polarimetric
SHG is limited to the projection of the nonlinear tensor in the sample plane (X,Y) and is
therefore not a complete 3D information. Indeed the excitation field is defined in this plane,
therefore only X or Y coupling directions are allowed in this geometry, except from high NA
focusing effects as discussed above. Any information related to off-plane coupling requires
complementary analysis. In some cases, as demonstrated for CMONS (Figure 4a), two-
photon fluorescence can provide additional information which, combined with SHG, is
sufficient to deduce such 3D orientation and crystallinity information from the nano-object
(Brasselet et al. 2004). Pure 2D projection analysis in SHG may also lead to the missing 3D
information in nano-crystals of complex symmetry involving non-diagonal coupling
directions (Zielinski et al. 2009, Brasselet 2010). Another method has been proposed to
deduce 3D information from nano-crystals, based on the defocused imaging of the nonlinear
radiation from the nano-object (Figure 2a, 4b). This has lead to successful determination of
3D orientation information based on the image analysis of the projection of the nonlinear
dipole radiation expansion in the (X,Y) sample plane (Sandeau et al. 2007).
262 Nanocrystals

a) b) =0° =60°
SHG TPF 300

IX IY
300

1) IX IY 200
100
1200 200
1000 100
3000
2500

x m
800 2000
Y X Y 0
X

x m
600 0 1500
-100 400
1000
200 -100
-200 500
0 -200
0
-300
-300
-200 0 200
-200 0 200
y m
SHG TPF TPF y m
SHG

=50° =60° =70° 300 300

2) 200
100
800
200
100 1200
600
X

x m
x m
X 0
400
0 800
-100 -100 400
200
Y -200 -200
0 0
Y
-300 -300

SHG SHG TPF TPF

 2
-200 0 200 -200 0 200

Non-crystalline
y m y m
E
X Y
p
80 3
E X
Counts / 30 s

60

40 1
2
20 Y
0
450 500 550 600 650 700 750 =30° =115° =90°
Wavelength (nm)

Fig. 4. (a) SHG and two-photon fluorescence (TPF) polarization dependence from two
different CMONS nano-crystals embedded in a sol-gel film (Brasselet et al. 2004). The SHG
and TPF emission wavelengths can be discriminated and therefore lead to a separated
polarization analysis of the two signals. 1) : mono-crystalline nanoparticle which projection
in the sample plane resembles that of a 1D nano-object. The 3D orientation of the nano-
crystal could be deduced from the simultaneous fit of the X and Y projections of the SHG
and TPF responses. 2) typical nano-crystal which polarization response cannot be fit by a
pure mono-crystalline orientation. (b) Defocused imaging of 80 nm a KTP nano-crystal for
two different input polarizations in the sample plane ( angles). The defocused images,
compared to modelling accounting for the full propagation of nonlinear induced dipoles in
the microscope, allowed retrieving a 3D orientation information, confirmed by polarimetry
(Sandeau et al. 2007).

3.4 Determining the nonlinear efficiency from a nano-crystal

Based on Eq. (4) and assuming that the nano-crystal symmetry and 3D orientation are
known, it is now possible to determine its nonlinear efficiency. This is generally done by
measuring a reference nonlinear response, similarly as in HRS (Yi et al. 2005, Revillod et al.
2008), which is provided here by a macroscopic crystal. In known excitation/detection
polarization coupling directions, the SHG signal from a crystal of known orientation is:
2
I ref 
2 
  ref  
. I
2 2
(5)

which nonlinear coefficient  ref

 2  is taken as a reference. The proportionality coefficient

contains the same collection and scaling factors as in Eq. (4), therefore SHG intensities can
directly be compared, for a same input power, and a given (IJK) set of polarization
conditions:
n 2 . IJK     IJK
2 
 2   ref2  . I ref2 I I2   (6)
2
Nano-crystals for quadratic nonlinear imaging: characterization and applications 263

The macroscopic nonlinear coefficients from a nano-crystal  IJK 2 

  can therefore be
measured for a given orientation. Note that this macroscopic coefficient is generated from
an object of much smaller size than that of a bulk crystal. Since  IJK2 
  is proportional to
the volume of the nano-object (Eq. (4)), and similarly  ref
 2  is proportional to the excitation

volume Vexc in the microscope, the nano-crystal efficiency can be transformed into an
“equivalent bulk” value which is the efficiency that the nano-crystal would exhibit if it were
,bulk     IJK  .Vexc V
filling the whole focal volume:  IJK
2  2  . Typically Vexc ~ 60.106 nm3, and
V ranges between 103 nm3 and 106 nm3. Note that since the excitation takes place over about
300 nm in the sample plane, it is considered here as homogeneous over the nano-crystal size.
More refinements are necessary when the nano-crystal size approaches the diffraction limit
size (Delahaye et al. 2009). Values up to 1pm/V have been measured in 100 nm size aligned
nano-clusters of MnPS3/DAZOP hybrid nanoparticles, which reach up to 1000 pm/V in an
equivalent bulk crystal (Delahaye et al. 2009). In addition of being stable, these nano-crystals
are therefore very bright nano-sources.

3.5 Retrieving size information from a nano-crystal

Assuming that both the orientation  and efficiency  IJK 2 
  of the nano-crystal can be
measured, it is now possible to retrieve size information from this object. For this, the
macroscopic to microscopic relation expressed in Eq. (2) is used and introduced in:

2 
V   IJK   / N . IJK   (7)

Where  IJK   are deduced from the microscopic nonlinear coefficients  ijk (Eq. (2)), known
from the unit-cells crystals values: d ijk  N ijk / 2 and assuming that both bulk nonlinear
efficiencies and density are the same as in the bulk crystal.
Such an analysis has been performed on a large collection of MnPS3/DAZOP hybrid nano-
crystals immobilized in PVA (Figure 5), which efficiency were used to evaluate a lower limit
to the particles size distribution (Delahaye et al. 2009). The lower limit size given here is
imposed by the lack of knowledge of the 3D off plane orientation angle of the nano-crystals
(even though the 2D orientation of the dipoles in the plane could be determined by SHG
polarimetry as described in the previous Section). The population of the deduced sizes was
seen to be consistent with the sizes of the spots measured on the SHG scans, which most
often were above the diffraction limit. This analysis allowed in particular evidencing an
aggregation behaviour of the nano-particles in the polymer film, which lead to 100 nm size
objects. Such statistical analysis, accounting for both SHG polarization dependence and
image size measurements, can therefore provide a complete picture of the structural
behaviour of nano-crystals.
264 Nanocrystals

Occurence
250 4

2
SHG signal (counts/50ms) 200
1m
0
60 80 100 120 140
150 radius (nm)

100

50

70 80 90 100 110 120

Particle radius (nm)

Fig. 5. Histograms of sizes of (MnPS3/DAZOP/Brij97, 10-4 mol/L DAZOP) nano-crystals in

a PVA polymer film. These sizes were deduced from the SHG signal dependence on the
particle radius R (as detailed in the text), depicted in the curve where each marker
represents an isolated nano-crystal. In this analysis the DAZOP molecules were assumed to
have the same hyperpolarizabtility as in solution, namely   220 10 30 esu . The estimated
size could be compared to the size measured for each nano-crystal on a SHG scan, which
some times surpasses the diffraction limit (Delahaye et al. 2009).

4. Applications
4.1 Bio-imaging
While SHG microscopy, together with polarimetric analyses, is today particularly successful
for bio-imaging benefiting from intrinsic responses from bio-molecular assemblies
(Campagnola et al. 2002, Zipfel et al. 2003, Stoller et al. 2002, Strupler et al. 2007), biological
applications still require efficient nano-emitters that can report biological relevant
information at nano-scale in targeted regions of a sample of interest. Several issues such as
toxicity, photo-bleaching, instability, lack of sharp contrast, are still considered as strong
drawbacks for the fluorescence process which is more traditionally used in bio-imaging.
These issues can be advantageously circumvented by the use of SHG active bio-marker.
SHG active nano-crystals are interesting alternatives due to the stability of their emission,
the flexibility in the emission wavelengths and the subsequent possibility to observe them
over a dark background, and at last their potential to no-toxicity. They can be additionally
excited in the IR (700-1500nm), within wavelength ranges that are less scattered by
biological media. They also do not exhibit phase-matching constraints contrary to more
bulky media, being therefore able to efficiently emit SHG regardless of the phase properties
of the excitation field.
ZnO nano-crystals have been successfully used for bio-imaging applications (Kachynski et
al. 2008). Sum frequency, second harmonic, and non-resonant four wave mixing nonlinear
signals have been obtained from stable dispersion of ZnO nanoparticles targeted to live
tumor (KB) cells. Robust intracellular accumulation of the targeted (FA incorporated) ZnO
nanocrystals could be observed without any indication of cytotoxicity. A water dispersion of
well-separated single nanocrystals was obtained from an encapsulation technique, using
Nano-crystals for quadratic nonlinear imaging: characterization and applications 265

phospholipid micelles formed by the self-assembly of polymer-grafted lipids (Kachynski et

al. 2008).
The holographic SHG imaging described in Section 3 allowed 3D imaging of BaTiO3
particles of 30 nm size in HeLa cells, treated by incubation at 37°C with the particles
uptaken by endocytosis (Hsieh et al. 2009). The particles were seen to be randomly
distributed in the cells, possibly as clusters, into vesicles. The cells were seen to be alive after
incubation, and fixed for imaging (Hsieh et al. 2009, Kuo et al. 2009, Hsieh et al. 2010).
At last, KNbO3 was successfully incorporated and optically trapped for imaging in living
cells (Nakayama et al. 2007).

SHG active nano-crystals are also of interest as punctual nano-sources for imaging through
thick tissues, since the wide range of usable wavelengths can be extended to the IR, which is
less scattered by natural media (Bohren & Huffmann, 1983). The wide frequency range
available from SHG active nano-crystals has been exploited using Fe(IO3)3 nanoparticles
illuminated at 800nm and 1550nm, in tissue phantoms made of submicrometric polystyrene
beads, as well as in a section of a fixed mouse liver (Exterman et al. 2009). Using long
working distance objectives (with lower NA objectives down to 0.6), better imaging quality
was observed at large depths (180 m in a phantom) for the larger IR wavelength.
At extensive depths, biological tissues are known to exhibit scattering properties which can
also alter considerably the quality of optical focusing. SHG efficiencies can therefore be
largely decreased by scattering when exploring large depths. A possible method to correct
for the deformation of the optical wavefront brought by the medium is based on phase
conjugation, which however requires a point source inside the medium. Indeed in this
technique a source is used as a reporter of the local phase of the excitation field; its scattered
field is then recorded in amplitude and phase in a hologram, and a phase-conjugated beam,
reconstructed from this field, is sent back into the sample, thus providing automatic
correction of phase distortions. SHG active nano-crystals (300 nm size BaTiO3) have been
used to play the role of the point source (Hsieh et al. 2010). This approach has shown
successful digital phase conjugation in turbid media using off-axis digital holography to
record the scattered SHG field from the nano-crsytals in a turbid medium made of a layer of
parafilm fixed on the particles sample. The measured phase conjugated beam was generated
in a spatial light modulator, and then sent back through the turbid medium. This
observation of a nearly ideal focus on a nano-crystal, would not have been feasible in linear
optics where all possible structures scatter the incident field without any spatial
discrimination (Hsieh et al. 2010).

4.2 Nano-scale optical fields probing

The nonlinear mechanism taking place in SHG couples coherently the emission induced
dipole to the excitation fields, therefore providing a possible way to probe excitation fields
at the nano-scale in phase, polarization and amplitude. Thanks to their sizes, nano-crystals
are unique reporters of optical fields properties at nano-scales. This approach has been
advantageously used to apply Frequency Resolved Optical Gating (FROG) (Trebino et al.
1997) in single Fe(IO3)3 nano-crystals under ultra-short excitation (69 fs), bringing new
prospective to monitor the phase and amplitude spectral distortions of such pulses
(Exterman et al. 2008). This technique, usually performed on bulk samples, is a unique
266 Nanocrystals

demonstration that opens to the exploration of optical fields time properties at any place in a
focal spot, since the size of the used particle is well below the optical resolution. The
additional advantage of the use of nano-crystal is the absence of constraints imposed by
nonlinear phase-matching in bulk crystals, which can be limiting in the case of extreme
ultra-short pulses analysis.
SHG responses under ultra-short pulses are generally strongly dependent on the spectral
and time phase profile of the pulses. In particular the optimum efficiency is reached for
Transform Limited pulses, which exhibit no spectral/time dependence in phase over their
whole bandwidth (“flat phase”). In order to maximize the SHG response from nano-crystals
under short pulses excitations (13 fs to 200 fs), an optimization of the signal has been
performed in KTP nano-crystals by directly adapting adequate optimization generic
algorithms (Baumert et al. 1997) to their emission signals (Wnuk et al. 2009). As expected, an
increase of the emission intensity has been observed when decreasing the duration of the
excitation fs pulses.

5. Conclusion
This chapter has given an overview of the current state of research on SHG active nano-
crystals. While this field is still in constant progress, many achievements have already
allowed revisiting nonlinear optics in order to adapt to the scale of these new nano-objects.
Their applications to both nanophotonics and biophotonics fields have shown their potential
and are probably only at their infancy.

6. Acknowledgements
We thank Véronique Le Floc’h, Kasia Komorowska, Christelle Anceau, Nicolas Sandeau,
Dominique Chauvat, Jean-François Roch, and Isabelle Ledoux for fruitful discussions and
contribution to nonlinear microscopy developments in Laboratoire de Photonique
Quantique et Moléculaire, ENS Cachan, France. Part of this work has been undertaken in
collaboration with Alain Ibanez (Institut Néel, Grenoble, France), Emilie Delahaye and René
Clément (Laboratoire de Chimie Inorganique, Institut de Chimie Moléculaire et des
Matériaux d’Orsay, Université Paris XI, Orsay, France), S. Perruchas, C. Tard and T. Gacoin
(Laboratoire de Physique de la Matière Condensée, Ecole Polytechnique, Palaiseau, France).

7. References
Anceau C., Brasselet S., Zyss J. (2005). Local orientational distribution of molecular
monolayers probed by nonlinear microscopy. Chem. Phys. Lett. 411(1-3), 98-102
Baumert T., Brixner T., Seyfried V., Strehle M. et Gerber G. (1997). Femtosecond pulse
shaping by an evolutionary algorithm with feedback. Appl. Phys. B: Lasers and
Optics, 65, 6
Bohren C. F., and Huffmann D. R. (1983). Absorption and Scattering by Small Particles. Wiley
Science paperback, Wiley & Sons, New York
Bonacina L., Mugnier Y., Courvoisier F., Le Dantec R. , Extermann J., Galez C., Boutou V.,
Wolf J.P. (2007). Polar Fe(IO3)3 nanocrystals as local probes for nonlinear
microscopy. Applied Physics B, Lasers and Optics, 87(3), 399-403
Nano-crystals for quadratic nonlinear imaging: characterization and applications 267

Boulanger B. and Zyss J. (1997). Physical Properties of Crystals, Vol. D of International Tables for
Crystallography, A. Authier, ed. Kluwer, Dordrecht, The Netherlands
Boyd R. W. (1992). Nonlinear Optics. Academic, New York
Brasselet S. and Zyss J. (1998). Multipolar molecules and multipolar fields: probing and
controlling the the tensorial nature of nonlinear molecular media. J.Opt.Soc.Am.B
15(1), 257-288
Brasselet S. and Zyss J. (2007). Nonlinear polarimetry of molecular crystals down to the
nanoscale. C. R. Physique 8, 165-179
Brasselet S. (2010). Second Harmonic Generation microscopy in molecular crystalline nano-
objects. Nonlinear Optics, Quantum Optics 40 (1-4), 83-94
Brasselet S., Le Floc'h V., Treussart F., Roch J. F., Zyss J., Botzung-Appert E., and Ibanez A.
(2004). In situ diagnostics of the crystalline nature of single organic nanocrystals by
nonlinear microscopy. Phys. Rev. Lett. 92(20), 4
Chemla D.S. and Zyss J. (Eds.) (1987) Nonlinear Optical Properties of Organic Molecules and
Crystals (Vol.1). Academic Press, Orlando
Chou C.-K., Chen W.-L., Fwu P. T., Lin S.-J., Lee H.-S., and Dong C.-Y. (2008). Polarization
ellipticity compensation in polarization second-harmonic generation microscopy
without specimen rotation. J. Biomed. Opt. 13(1), 014005
Delahaye E., Sandeau N., Tao Y., Brasselet S., Clément R. (2009). Synthesis and Second
Harmonic Generation microscopy of nonlinear optical efficient hybrids
nanoparticles embedded in polymer films. Evidence for Intra and Inter
nanoparticles orientational synergy. J. Phys. Chem. C, 113 (21) 9092–9100
Delahaye E., Tancrez N., Yi T., Ledoux I., Zyss J., Brasselet S., and Clement R. (2006). Second
harmonic generation from individual hybrid MnPS3-based nanoparticles
investigated by nonlinear microscopy. Chem, Phys. Lett. 429, 533–537
Eisenthal Kenneth B. (2006). Second Harmonic Spectroscopy of Aqueous Nano- and
Microparticle Interfaces. Chem. Rev. 106, 1462-1477
Extermann J., Bonacina L., Courvoisier F., Kiselev D., Mugnier Y., . Le Dantec R, Galez C.,
Wolf J.P. (2008). Nano-FROG: Frequency Resolved Optical Gating by a nanometric
object. Optics Express, 16(14) 10405-10411
Extermann J., Bonacina L., Cuna E., Kasparian C., Mugnier Y., Feurer T., Wolf J.P. (2009).
Nanodoublers as deep imaging markers for multi-photon microscopy. Optics
Express, 17(17) 15342-15349
Fu D., Zhang Y., Liu J., Lu Z. (2001). Hyper-Rayleigh scattering of CdS nanoparticles
stabilized by inorganic heteropolyanions. Mater. Lett. 51, 183-186
Galez C., Mugnier Y., Bouillot J., Lambert Y., and Le Dantec R. (2006). Synthesis and
characterisation of Fe(IO3)3 nanosized powder. J. Alloys Comp. 416, 261-264
Hajj B., de Reguardati S., Hugonin L., Le Pioufle B., Osaki T., Suzuki H., Takeuchi S.,
Mojzisova H., Chauvat D., Zyss J. (2009). Electro-Optical Imaging Microscopy of
Dye-Doped Artificial Lipidic Membranes Biophys. J. 97(11), L29
Hajj B., Gacoin T., Zyss J., Chauvat D. (submitted 2010). Electrooptical Pockels scattering
from a single nanocrystal.
Halbout J.-M., Blit S., Chung T. (1984). Evaluation of the phase-matching properties of
nonlinear optical materials in the powder form. IEEE Journ. of Quant. El. 17(4), 513-
517
268 Nanocrystals

Hellwarth, R. Christensen, P. (1974). Nonlinear optical microscope examination of structures

in polycrystalline ZnSe. Opt. Commun, 12, 318-322
Hsieh C. L., Grange R., Pu Y., and Psaltis D. (2009). Three-dimensional harmonic
holographic microcopy using nanoparticles as probes for cell imaging. Opt. Express
17(4), 2880–2891
Hsieh C. L., Pu Y., Grange R., and Psaltis D. (2010). Digital phase conjugation of second
harmonic radiation emitted by nanoparticles in turbid media. Opt. Express 18(12),
12283-12290
Hsieh C. L., Grange R., Pu Y., and Psaltis D. (2010). Bioconjugation of barium titanate
nanocrystals with immunoglobulin G antibody for second harmonic radiation
imaging probes. Biomaterials 31(8), 2272–2277
Johnson J. C., Yan H. Q., Schaller R. D., Petersen P. B., Yang P. D., and Saykally R. J. (2002).
Near-field imaging of nonlinear optical mixing in single zinc oxide nanowires.
Nano Lett. 2(4), 279–283
Josse D., Dou S.X., Zyss J., Andreaaza P., Périgaud A. (1992). Near-Infrared Optical
parametric Oscillation in a N-(4-Nitrophenyl)-L-Prolinol Molecular Crystal.
Appl.Phys.Lett. 61,121
Kachynski A. V., Kuzmin A. N., Nyk M., Roy I., and Prasad P. N. (2008). Zinc oxide
nanocrystals for nonresonant nonlinear optical microscopy in biology and
medicine. J. Phys. Chem. C 112(29), 10721–10724
Komorowska K., Brasselet S., Zyss J., Pourlsen L., Jazdzyk M., Egelhaaf H.J., Gierschner J.,
Hanack M. (2005). Nanometric scale investigation of the nonlinear efficiency of
perhydrotriphynylene inclusion compounds. Chem. Phys. 318 (1-2), 12-20
Kuo T. R., Wu C. L., Hsu C. T., Lo W., Chiang S. J., Lin S. J., Dong C. Y., and Chen C. C.,
(2009). Chemical enhancer induced changes in the mechanisms of transdermal
delivery of zinc oxide nanoparticles. Biomaterials 30(16), 3002–3008
Kurtz S. K., Perry T. T. (1968). A Powder Technique for the Evaluation of Nonlinear Optical
Materials. J. Appl. Phys., 39, 3798–3813
Le X. L., Zhou C., Slablab A., Chauvat D., Tard C., Perruchas S., Gacoin T., Villeval P., and
Roch J. F. (2008). Photostable second-harmonic generation from a single KTiOPO4
nanocrystal for nonlinear microscopy. Small 4(9), 1332–1336
Le X. L., Brasselet S., Treussart F., Roch J. F., Marquier F., Chauvat D., Perruchas S., Tard C.,
and Gacoin T. (2006). Balanced homodyne detection of second-harmonic generation
from isolated subwavelength emitters. Appl.Phys. Lett. 89(12), 121118
Le Bozec, H., Le Bouder, T., Maury, O., Bondon, A., Ledoux, I., Deveau, S., Zyss, J (2001).
Supramolecular Octupolar Self-organization towards Nonlinear Optics. Adv.
Mater., 13, 1677-1681
Le Floc’h V., Brasselet S., Roch J.F., Zyss J. (2003). Monitoring of orientation in molecular
ensembles by polarization sensitive nonlinear microscopy. J. Phys. Chem. B, 107,
12403-12410
Le Floc’h V., Brasselet S., Zyss J., Chao R., Lee S.H., Jeon S.J., Cho M., Min K.S., Suh M.P.
(2005). High efficiency and quadratic optical properties of a fully optimized 2D
octupolar crystal by nonlinear microscopy. Adv. Mater. 17(2),196-200
Magrez, A. , et al. (2006). Growth of single-crystalline KNbO3 nanostructures. J. Phys. Chem.
B 110, 58–61
Nano-crystals for quadratic nonlinear imaging: characterization and applications 269

Mugnier Y., Houf L., El-Kass M., Le Dantec R. , Hadji R., Vincent B., Djanta G., Badie L.,
Eschbach J., Rouxel D. and Galez C. (submitted 2010). In situ crystallization and
growth dynamics of acentric iron iodate nanocrystals in w/o microemulsions
probed by Hyper-Rayleigh Scattering measurements.
Nakayama Y., Pauzauskie P. J., Radenovic A., Onorato R. M., Saykally R. J., Liphardt J., and
Yang P. D. (2007). Tunable nanowire nonlinear optical probe. Nature 447(7148),
1098–1101
Oudar J.L., Chemla D.S. (1977). Hyperpolarizabilities of the nitroanilines and their relations
to the excited state dipole moment. J. Chem. Phys 66, 2664-2670
Pu Y., Centurion M., Psaltis D. (2008). Harmonic Holography–a new holographic principle.
Appl. Opt. 47 A103-A110
Rekaï, E. D., Baudin, J. B., Jullien, L., Ledoux, I., Zyss, J., Blanchard-Desce, M. Chem.Eur. J.
2001, 7, 4395
Revillod G., Duboisset J., Russier-Antoine I., Benichou E., Bachelier G., Jonin Ch., Brevet P.F.
(2008). Multipolar Contributions to the Second Harmonic Response from Mixed
DiA-SDS Molecular Aggregates. J. Phys. Chem. C 112, 2716
Richards B., and Wolf E. (1959). Electromagnetic diffraction in optical systems. 2. Structure
of the image field in an aplanatic system. Proceedings of the Royal Society of London
Series a-Mathematical and Physical Sciences 253, 358–379
Rodríguez E. M., Speghini A., Piccinelli F., Nodari L., Bettinelli M., Jaque D., and Sole J. G.
(2009). Multicolour second harmonic generation by strontium barium niobate
nanoparticles. J. Phys. D Appl. Phys. 42(10), 4
Rodriguez E. V., de Araujo C. B., Brito-Silva A. M., Ivanenko V. I., and Lipovskii A. A.
(2009), Hyper-Rayleigh scattering from BaTiO3 and PbTiO3 nanocrystals. Chem.
Phys. Lett. 467(4-6), 335–338
Sandeau N., Le X.L. , Chauvat D., Zhou C., Roch J. F., and Brasselet S. (2007). Defocused
imaging of second harmonic generation from a single nanocrystal. Opt. Express
15(24), 16051–16060
Sanz N., Baldeck P. L., Nicoud J. F., LeFur Y., and Ibanez A. (2001), Solid State Sci. 3, 867.
Shen Y., Swiatkiewicz J., Winiarz J., Markowicz P., Prasad P. (2000), Second-harmonic and
sum-frequency imaging of organic nanocrystals with photon scanning tunneling
microscope. Appl.Phys.Lett. 77, 2946 (2000)
Schön P., Munhoz F., Gasecka A., Brustlein S., and Brasselet S. (2008). Polarization distortion
effects in polarimetric two-photon microscopy. Opt. Express 16(25), 20891–20901
Schön P., Behrndt M., Ait-Belkacem D., Rigneault H., Brasselet S. (2010). Polarization and
Phase Pulse Shaping applied to Structural Contrast in Nonlinear Microscopy
Imaging. Phys. Rev. A 81, 013809
Terhune R.W., Maker P.D., SavageC.M. (1965). Measurements of Nonlinear Light Scattering.
Phys. Rev. Lett. 14, 681-684.
Toury T., Brasselet S., Zyss J. (2006) Electro-optical microscopy: mapping nonlinear polymer
films with micrometric resolution Opt. Lett. 31(10), 1468-1470
Trebino R., DeLong K.W., Fittinghoff D.N., Sweetser J.N., Krumbügel M.A., Richman B.A.,
and Kane D.J. (1997). Measuring ultrashort laser pulses in the time-frequency
domain using frequency-resolved optical gating. Rev. Sci. Instrum. 68, 3277
270 Nanocrystals

Treussart F., Botzung-Appert E., Ha-Duong N.-T., Ibanez , Roch J.-F., Pansu R. (2003).
Second Harmonic Generation and Fluorescence of CMONS Dye Nanocrystals
Grown in a Sol-Gel Thin Film. ChemPhysChem 4, 757–760
Wang X., Zhang Y., Wang G., Zhang C., Fu D., Shen Y., Lu Z., Cui Y., Ochiai S., Uchida Y.,
Kojima K. and Ohashi A. Second-order nonlinear optical properties of
centrosymmetric nanoparticles studied by hyper-Rayleigh scattering (HRS)
technique. (2005). Chinese Optics Letters, Vol. 3, Issue S1, S125-S127
Wnuk P., Le L. X., Slablab A., Tard C., Perruchas S., Gacoin T., Roch J. F., Chauvat D., and
Radzewicz C. (2009). Coherent nonlinear emission from a single KTP nanoparticle
with broadband femtosecond pulses. Opt. Express 17(6), 4652–4658
Yew E. Y. S., and Sheppard C. J. R. (2006). Effects of axial field components on second
harmonic generation microscopy. Opt. Express 14(3), 1167–1174
Yi T., Tancrez N., Clément R., Ledoux-Rak I., Zyss J. (2004). Organic-MPS3 nanocomposites
with large second-order nonlinear optical response. J. Lumin, 110, 389-395
Yi T., Clément R., Haut C., Catala L., Gacoin T., Tancrez N., Ledoux I., Zyss J. (2005). J-
Aggregated Dye-MnPS3 Hybrid Nanoparticles with Giant Quadratic Optical
Nonlinearity. Adv. Mater. 17, 335-338
Yokoyama, S., Nakahama, T., Otomo, A., Mashiko, S (2000). Enhancement of the Molecular
Hyperpolarizability in Multi-choromophoric Dipolar Dendrons. J. Am. Chem. Soc.,
122, 3174-3181
Zielinski M., Oron D., Chauvat D., and Zyss J. (2009). Second-harmonic generation from a
single core/shell quantum dot. Small 5(24), 2835–2840
Zyss J. and Oudar J.L. (1982). Relations between microscopic and macroscopic lowest order
optical nonlinearities of molecular crystals in one and two-dimensional units. Phys
Rev A 26 (4), 2028-2048
Zyss J., Nicoud J.F., Coquillay M. (1984). Chirality and hydrogen-bonding in molecular
crystals for phase-matched second-harmonic generation:
N-(4-nitrophenyl)-(L)-prolinol (NPP). J. Chem. Phys. 81, 4160
Zyss J. and Ledoux I. (1994) Nonlinear Optics in Multipolar Media: Theory and
Experiments. Chem.Rev.94,77-105
Hybrid colloidal nanocrystal-organics based light emitting diodes 271

12
X

Hybrid colloidal nanocrystal-organics based

light emitting diodes
Aurora Rizzo, Marco Mazzeo and Giuseppe Gigli
National Nanotechnology Laboratory (NNL), CNR-Nanoscienze, Università del Salento,
Italian Institute of Technology (IIT) via Morego, 30 I-16163 Genova, Italy
Via per Arnesano Km 5 I73100 Lecce, Italy

1. Introduction
In the last few years a new and promising field of research defined “nanoscience” has
attracted an increasing interest. This new field involves the capability to manipulate,
fabricate and characterize the material structure at a nanometer scale. Research dedicated to
this field has an interdisciplinary character and covers physics, engineering, chemistry,
material science and, molecular biology. The recent possibility to manipulate the matter
with nanoscopic resolution revealed novel physical and chemical properties of low
dimensional structures in between bulk materials and atomic sizes. Among all the
nanostructured materials, semiconductor quantum dots (QDs) or nanocrystals have been
used in a wide array of fields such as electronic and optoelectronic devices. The most recent
techniques in the preparation of nanocrystals are the chemical methods and in particular the
colloidal nanocrystals synthesis. These techniques allow for the growth of semiconductor
nanocrystals in solution and consequently their integration in the organic electronic
technology. The most effective technique is the synthesis by thermal decomposition of
precursors in hot coordinating solvents. This allows the growth of semiconductor (II-VI, III-
V), (Murray et al., 1993; Murray et al., 2001) metal (Murray et al., 2001; Sun et al., 1999) and
metal oxide nanocrystals with high quality, (Rockenberger et al., 1999) almost free from
defects and narrow size distribution (less than 5%).
In particular the mostly used material in optoelectronic application synthesized by this
technique is the luminescent CdSe QDs. Usually the CdSe cores are surrounded by a wider
band gap semiconductor (such as ZnS) that passivates the surface states of nanocrystals with
a consequent increase in the photoluminescence. In order to ensure the solubility, the QDs
are covered by an organic capping group.
These structures are smaller in size than the diameter of a Bohr exciton in a bulk crystal of
the same material. By reducing the size of the nanocrystals core, the quantum confinement
of the electron, hole and exciton is increased, with a consequent increase in the exciton
energy. For instance in CdSe nanocrystals, the quantum confinement increases the exciton
energy from a bulk bandgap of 1.7 eV (λ=730nm red edge of visible spectrum) to any value
up to 2.7 (λ=450nm blue luminescence). The spectral tunability can access almost all the
visible range with a single material set. Moreover the saturated color emission (linewidths of
<35nm Full Width at Half Maximum), the high luminous efficiency even in the blue region,
272 Nanocrystals

the possibility to tailor the external chemistry without affecting the emitting core as well as
the inorganic nature make these structures ideal candidates in the fabrication of hybrid
organic/inorganic light emitting diodes (LEDs) potentially more stable and with a longer
lifetime with respect to the fully organic counterpart.
In recent years, indeed the development of organic light emitting diode (OLED) showed
several basic physical problems that are very difficult to overcome. One is the relatively
short lifetime if compared with other electroluminescent devices. Lifetime have been
extended using new chemistry to prevent crystallization and sophisticated packaging
schemes to avoid the degradation of the organic emitting molecules by water and oxygen.
The chemistry optimization must be repeated for each emitter and differences in the aging
time of different molecules emitting in the red, green and blue region poses problems in the
OLED color stability. Inorganic QDs are robust and high luminescence lumophores but they
present poor charge conduction properties. Recently many efforts have been devoted in the
successful integration of these nanocrystalline materials in OLED technology. However the
studies on photocurrent and injection in thin disordered films of nanocrystals between
metal electrodes revealed that transport and injection in these films cannot be described in
the framework of the band model of semiconductor (Ginger et al., 2000; Leatherdale et al.,
2000). Nanocrystals are particles separated one to each other by organic surfactant. The high
degree of disorder in these films suggests that hopping and tunneling between localized
states are the principal mechanism for the charge transport (Ginger et al., 2000). The
hopping model is similar to the one described for the organic semiconductors, but in the
case of nanocrystals the energetic disorder arises from the size distribution of the particles
and the geometric disorder comes from their spatial separation, also induced by the
surfactant. Moreover colloidal nanocrystals have poor charge transport properties especially
in the case of thick multilayer films. To avoid the problem of poor charge conduction in LED
fabrication, colloidal nanocrystals have to be merged with organic semiconductor hole and
electron transporting/injection materials. In the last few years several hybrid device
structures based on colloidal QDs and organic materials have been reported.
In this chapter, the recent development in the fabrication of Hybrid organic/QD-LEDs will
be reviewed; In particular the diverse approaches to fabricate multilayer devices and white
light emitting devices will be discussed in details.

2. Hybrid Light Emitting Diodes based on Polymer/QDs

The fist work on colloidal semiconductors nanocrystal LEDs demonstrated a device
structure in which five monolayer of CdSe quantum dots were deposited on the top of spin
coated and thermal converted semiconducting p-paraphenylene vinylene (PPV) layer.
Nanocrystal layers were bound on the polymer with exane dithiol functionalization (Colvin
et al. 1994). The light emission arises from the recombination of holes injected into the PPV
polymer layer with electrons injected into the multilayer film of CdSe nanocrystals. The low
external quantum efficiency of the devices (between 0.001 and 0.1%) can be attributed either
to the low photoluminescence (PL) efficiency of the CdSe core QDs, or to the poor electron
conduction through the five layer thick QD film (Morgan et al., 2002). Moreover, the use of
the hexane dithiol for the device fabrication adds impurities that can trap the charges and
quenching the QD excitons.
Hybrid colloidal nanocrystal-organics based light emitting diodes 273

Subsequently electroluminescence has been obtained by incorporating CdSe QDs into a thin
film (100 nm) of polyvinylcarbazole (PVK), a photostable hole conducting polymer, and an
oxadiazole derivative (t-Bu-PBD), an electron transport species (Dabbousi, 1995). Three
devices with different size of QDs (~32Å, ~40Å, ~60Å diameter) have been fabricated by
spin coating nanocrystals with the mixture of PVK and t-Bu-PBD as an emitting layer
(Figure 1). The maximum external quantum efficiency in air and at room temperature of
these devices is quite low (~0.0005%) and it decreases with decreasing of the QD size.

Fig. 1. Absorption (dashed lines), photoluminescence (exc. 457.9 nm) (dotted Lines), and
electroluminescence (solid lines) spectra for 32 Å (a),(b), 40Å (c),(d), and 60 Å (e),(f) diameter
CdSe nanocrystallite/PVK/PBD devices at room temperature (a), (c), (e) and at 77 K (b), (d),
(f). (Reprinted with permission, Dabbousi et al. , 1995. Copyright 1995, American Institute of
Physics).

A significant improvement over the previous devices has been obtained by using CdSe(CdS)
core/shell QDs (Schlamp et al., 1997). These core/shell nanocrystals consist in a CdSe core
surrounded by a shell of epitaxially grown CdS. The core and shell energy levels alignment
confines the holes in the core while electrons are delocalized in throughout the structure.
This results in a higher PL quantum yield (> 50% in solution at room temperature) and
photo-oxidative stability of the CdSe(CdS) compared with the CdSe core samples. The active
nanocrystal multilayer has been deposited on a thermal converted PPV polymer (Figure 2).
274 Nanocrystals

Fig. 2. PL (solid lines) and EL (dashed) of samples of (25, 5) CdSe(CdS) (a) and (34, 7)
CdSe(CdS) (b) nanocrystals. The PL was collected in solution at room temperature. Inset:
device configuration. (Reprinted with permission, Schlamp et al., 1997. Copyright 1997,
American Institute of Physics).

The devices emit from red to green with external quantum efficiency of 0.22% at a
brightness of 600 cd/m2 and a current density of 1 A/cm2. Despite the potential advantages
of using QDs as emitters, all these QD-LED structures have efficiencies that are far below of
the all organic LED technology ones.
The most limiting factor in the QD devices performances is the poor conductivity of the
colloidal nanocrystals compared to semiconductor materials (Leatherdale et al., 2000) and in
all the structures reported above the QDs act both as emitters and electron transport species.
A multilayer device configuration can allow an independent optimization of materials for
charge injection, transport and, emission.

3. High efficiency hybrid LED insulating QD layer function

3.1 QD-LED by phase segregated nanocrystal monolayer
In 2002 an innovative QD-LED structure incorporating a monolayer of CdSe/ZnS core/shell
QDs sandwiched between hole and electron transporting organic layers was proposed (Coe
et al., 2002). In this device structure colloidal QDs act only as a lumophores and they do not
participate to the charge conduction process. The organic layers transport charge carriers to
the vicinity of the QD monolayer from which the luminescence originates. In this way the
problem of QD poor charge conductivity due to the insulating layer of surfactant, that coat
their surface, was overcome. For the device fabrication a solution of QDs and organic small
molecule material is spin coated on an Indium Tin Oxide anode. During the spin-coating
process the QDs, covered by an aliphatic capping layer, phase separate from the aromatic
small molecules, such as N,N’-diphenyl-N,N’-bis(3-methylphenyl)-(1,1’-biphenyl)-4,4’-
diamine (TPD), and form a layer on the organic surface.
The thickness and the coverage of the nanocrystals layer depends on the QD concentration
in the solution. For the device fabrication the QD concentration is optimized in order to
obtain a single close packed monolayer. By exploiting the phase segregation technique they
fabricated high luminance devices with emission in the red-green region achieving a
maximum external quantum efficiency >2%.(Coe et al., 2002, Coe-Sullivan et al., 2003; Coe-
Sullivan et al., 2005) (Fig. 3).
Hybrid colloidal nanocrystal-organics based light emitting diodes 275

Fig. 3. Performance metrics of the QD-LED, with external quantum efficiency of in excess of
2% and a maximum luminescence over 7,000cd/m2. Inset: EL spectrum with a saturated
color red emission peaking at 615nm , and full width at half maximum of 27nm (Coe-
Sullivan et al., 2005. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with
permission.).

3.2 Multilayer devices by cross-linked hole transporting layer

Recently an alternative strategy for QD-LED fabrication by spin-coating CdSe/ZnS QDs
onto a thermally cross-linked hole transporting layer, polystyrene (PS)- N,N’-diphenyl-
N,N’-bis(4-n-butylphenyl)-(1,1’-biphenyl)-4,4’-diamine(TPD)-perfluorocyclobutane (PFCB)
was reported (Zhao, 2006). Following the deposition of the QD layer, the electron
transporting layer, 1,3,5-tris(N- phenylbenzimidazol-2yl)benzene (TPBI), was deposited by
thermal evaporation. EL spectrum and device structure are reported in fig. 4.

Fig. 4. Absorption (dashed line), photoluminescence (dotted line), and electroluminescence

spectra (solid line) for CdSe/CdS QD-LEDs with a structure ITO/PS-TPD-PFCB (30
nm)/CdSe QDs/ TPBI (40 nm)/Ca (30 nm)/Ag (120 nm) at a voltage of 6.0 V. Structures of
the multilayered CdSe/CdS QD-LEDs and cross-linked PS-TPD-PFCB are shown in the
inset of the figure. (Reprinted with permission, Zhao et al., 2006.Copyright 2006, American
Chemical Society).
276 Nanocrystals

The maximum external quantum efficiency of these devices was 0.8% at 100 cd/m2 and a
maximum brightness in excess of 1000 cd/m2.
The same group improved the performance of these devices structure by using thermal
annealing of the quantum dot layer reaching an external quantum efficiency of 1.6% at a
brightness of 100 cd/m2 (Niu et al., 2007).
Although QD-LEDs with QDs monolayer have the advantages of higher efficiency and
lower turn-on voltages, their output power, maximum luminance and color purity are
limited owing to the low chromophore quantity and the poor confinement of excitons in the
active QD region.
Recently high-performance red, orange, yellow and green QD-LEDs based on QDs with a
CdSe core and a ZnS or CdS/ZnS shell have been reported (Sun et al., 2007). Their
maximum luminance reached 9,064, 3,200, 4,470 and 3,700cd/m2 (Fig. 5).

(a) (b)
Fig. 5. (a) Current/voltage and luminance/voltage characteristics of red-, orange-, yellow-
and green-emitting QD-LEDs. The insets show schematic device configurations of the
corresponding QD-LEDs. (b) EL spectra of the red-, orange-, yellow- and green- emitting
QD-LEDs operating at different luminances voltages. The insets show images of the devices
under operation. (Reprinted with permission, Sun et al., 2007. Copyright 2007, Nature
Publishing Group).
Hybrid colloidal nanocrystal-organics based light emitting diodes 277

The superior performances of the QD-LEDs arise from the careful preparation of highly
purified, uniform and monodispersed colloidal core–shell QDs, and optimization of the
thicknesses of the polymer HTL, the QD layer and the ETL. Indeed, for use in QD-LEDs, all
the QDs were subjected to a multistep purification process to remove the organic ligands.
After purification, the red, orange and yellow QDs were in solid powder form with a
quantum yield (QY) higher than 30% in toluene, whereas the green QDs were still in liquid
solution, with a QY of 10% in toluene, owing to their smaller size. During the purification
process, the removal of organic ligands from the QDs produces surface defects that may
trap charges.
A trilayer structure of indium tin-oxide(ITO)/poly(ethylenedioxythiophene):polystyrene
sulphonate (PEDOT:PSS) (25 nm)/HTL (45 nm)/QDs (15–20 nm)/ETL (35 nm)/Ca (15
nm)/Al (150 nm) was used in the fabrication of the QD-LEDs, with ITO/PEDOT:PSS as the
anode, poly(N,N0-bis(4-butylphenyl)-N,N0-bis(phenyl) benzidine) (poly-TPD) spin-coated
from its chlorobenzene solution as HTL, QD layers spin-coated from their toluene solutions
as emissive layers, tris(8-(hydroxyl-quinoline) aluminium (Alq3) as ETL, and Ca/Al as the
cathode. In the QD-LEDs, the HTL and ETL thicknesses were both optimized to confine the
injected electrons and holes to recombine predominantly within the QD layer and provide
optimal hole and electron transportation.

3.3 QD-light emitting diodes with metal oxides charge transport layers
For practical device application it is fundamental to improve the shelf-life robustness.
Recently the use of sputtered amorphous inorganic semiconductors as robust charge
transport layers was reported (Caruge, 2008). In particular 20nm thick film of NiO is
deposited by radiofrequency (RF)-sputtering on the ITO electrode; the active layer consists
of ZnCdSe alloyed QDs, deposited by spin-coating, with red emission peak at 638nm. The
electron transporting layer is an optically transparent film of alloyed ZnO and SnO2. NiO
and ZnO:SnO2 are p-type and n-type semiconductors respectively (Fig. 6).

Fig. 6. (a) Device configuration: the Ni is used as hole transporting layer, the QD as active
luminescent layer, the ZnO:SnO2 as electron transporting layer. (b) Band diagram
determined from UV photoemission spectroscopy and optical absorption measurements.
(Reprinted with permission, Caruge et al., 2008. Copyright 2008, Nature Publishing Group).

A maximum EQE of nearly 0.1% was demonstrated, the peak brightness was measured to be
1950 cd m-2 at 19.5 V and 3.73 Acm-2. Comparable brightness and J-V characteristics is
claimed when the devices were tested after being stored for four days, in contrast to
unpackaged organic QD-LEDs, which cannot withstand prolonged atmospheric exposure.
278 Nanocrystals

Recently high efficiency QD-LEDs by using a sol gel TiO2 electron transporting layer was
reported (Cho et al., 2009); the TiO2 layer is deposited by spin-coating, enabling the QD-LED
to be fabricated by means of an all-solution process (Fig. 7). The emissive layer is made of a
crosslinked QD layer directly spin-coated on a poly[(9,9-dioctylfluorenyl-2,7-diyl)-co-(4,40-
(N-(4-sec-butylphenyl)) diphenylamine)] (TFB) hole transporting layer. The linker molecule,
1,7-diaminoheptane, is attached to the QD through exchange with the pre-existing
surfactants or by binding to empty sites on the QD surface. The crosslinking is obtained by a
post-deposition thermal treatment. The crosslinking of the QD layer greatly improves the
luminance and luminous efficiency.

Fig. 7. (a) Device structure (left) and cross-sectional TEM image (right) of the QD-LED. TFB,
poly[(9,9-dioctylfluorenyl-2,7-diyl)-co-(4,40-(N-(4-sec-butylphenyl))diphenylamine)]. Scale
bar, 100 nm. (b) Energy band diagram. (Reprinted with permission, Cho et al., 2009.
Copyright 2009, Nature Publishing Group).

It is demonstrated that the energy band offset between the QD and the HTL was reduced
from 1.5 to 0.9 eV that leads to an increase by over a factor of 10 in the maximum luminous
efficiency because of more efficient hole injection and enhanced charge balance. Moreover
the metal oxide TiO2 layer shows an improved electron injection superior if compared with
the standard organic Alq3; this is a consequence of the lower band offset (0.4 eV) for Al/TiO2
compared to that for Al/Alq3 (1.2 eV). And higher electron mobility of the sol-gel TiO2 (1.7
× 10-4 cm2 V-1 s-1) compared to that for Alq3 (~1.0 × 10-4 cm2 V-1 s-1). The turn-on voltage in
the TiO2-based QD-LED also significantly decreased to 1.9 V, which is smaller than the QD
bandgap of 2.1 eV and much lower than that observed for the Alq3-based device (~4.0 V).
The overall device performances are improved, in particular the luminance reach the value
of 12380 cd m-2 and a maximum power efficiency of 2.41 lm W-1.

4. White Hybrid Light Emitting Devices

To obtain white light, all the three primary colors (red, green and blue) have to be produced
simultaneously. Since it is difficult to obtain all primary emissions from a single molecule,
excitation of more than one organic species is often necessary, thus introducing color
stability problems. Due to the different degradation rates of the employed organic
compounds, the emission color of the device can, in fact, changes with time. The CdSe
semiconductor QDs exhibit a size-dependent color variation due to quantum confinement
effect, which covers almost whole the visible range (Suzuki et al., 1996; Ginger et al., 2000;
Coe-Sullivan et al., 2005). Additionally, the fluorescence efficiency and, in particular the
stability of the nanocrystals, can be greatly improved by modifying the particle surface.
These characteristics can be merged with peculiar properties of organic materials, such as
Hybrid colloidal nanocrystal-organics based light emitting diodes 279

flexibility and ease of processing, to give rise to a novel class of low cost hybrid white-LEDs
with improved lifetime and color stability. We proposed two different white emitting device
structures based on colloidal semiconductor QDs. In the first structure, red emitting
CdSe/ZnS QDs are blended in a blue emitting polyfluorene (Li et al., 2005). The matrix
polymer provides a simple device preparation process due to its high processability and a
blue emitting component in the device emission. The function of green component in the
white emission spectrum is given to a green emitting electron transporting material
evaporated on top of the polymer/QDs film. In the second structure (Li et al., 2006), three
different size of CdSe/ZnS QD samples emitting at wavelength of 490nm, 540nm, 620nm are
dispersed in a high mobility hole organic matrix. White bright electroluminescence is
obtained from the ternary nanocrystal composites.

4.1 Organic-Inorganic Hybrid White LEDs Based on Polymer and QDs

Host-guest systems are typically employed to obtain white light emission by exploiting two
mechanisms, namely, Förster energy transfer (Förster, 1959; Lakowicz, 1983) and charge
transfer (Utsugi et al., 1992; Suzuki et al., 1996). All these processes have to be accurately
controlled in order to obtain white EL. We demonstrated that a balanced white emission is
obtained in hybrid system poly[(9,9-dihexylfluoren- 2,7-diyl)-alt-co- (2-methoxy-5- {2-
ethylhexyloxy} phenylen-1,4-diyl)](PFH-MEH):QDs/ Alq3 when Förster energy transfer in
the guest-host system is accomplished by charge transfer from PFH-MEH and Alq3 to QDs
during the electrical excitation (Li et al., 2005). The absorption and PL spectra of all
components and possible energy/charge-transfer pathways were shown in fig. 8.

(a) (b)

(c)

Fig. 8. (a) Absorption and (b) Photoluminescence (PL) spectra of device components. (c)
Possible pathways leading to emissive states in device ITO//PEDOT-PSS// PFH-MEH
(PF):CdSe/ZnS//Alq3//Ca/Al. (Reprinted with permission, Li et al., 2005. Copyright 2005,
American Institute of Physics).
280 Nanocrystals

Electroluminescence measurements have been carried out on ITO//PEDOT-PSS//PFH-

MEH:CdSe/ZnS//Alq3//Ca/Al structures. In fig. 9 the EL spectra for different PFH-
MEH:CdSe/ZnS concentration ratios are reported.

Fig. 9. EL spectra for the device ITO//PEDOT-PSS//PFH-MEH:CdSe/ZnS//Alq3// Ca/Al

with different ratio (PFH-MEH:CdSe/ZnS, c%) (Reprinted with permission, Li et al., 2005.
Copyright 2005, American Institute of Physics).

In order to achieve white EL emission, the different color components have to be accurately
balanced by controlling both the Förster energy transfer and charge transfer mechanisms. To
this aim we fabricated devices with different PFH-MEH:QDs concentration ratios, namely
200:1 and 300:1. At low QDs concentration (300:1), the possible pathways are process I and
II. Process I involves the transfer of a hole from a PFH-MEH cation radical (PF+•) to an Alq3
anion radical (Alq3-•). Process II involves the transfer of an electron from an Alq3 anion
radical (Alq3-•) to a PFH-MEH cation radical (PF+•). Both the mechanisms result in excited
Alq3 and PFH-MEH molecules which can decay radiatively, originating blue and green EL
emission. Negligible red emission is instead originated from the low amount of QDs. By
increasing the concentration of CdSe/ZnS QDs (concentration ratio 200:1), the possible
pathway are process I, II and III. A relevant additional role is assumed by the following
processes: Förster energy transfer to QDs from excited PFH-MEH and Alq3 molecules,
sequential charge transfer of a hole from PFH-MEH followed by transfer of an electron from
Alq3 or charge transfer of an electron from Alq3 followed by transfer of a hole charge from
PFH-MEH (processes III). Efficient electroluminescence at the three primary colors is thus
obtained from PFH-MEH, Alq3 and QDs with a balanced white spectrum with CIE
(0.30,0.33). Maximum External Quantum Efficiency of 0.24% is measured at 1mAcm-2 and
11V. In these QDs based white-LEDs, holes are considered to be injected from the ITO
electrode through PEDOT:PSS layer into the polymer hole conductor and are eventually
transported to the QDs. Similarly, the electrons are considered to be injected from Ca/Al
cathode into the Alq3 and are eventually transported to the QDs. Since the high electron
affinity of QDs the electrons are better confined within the surface PFH-MEH:QDs/Alq3, thus
enhancing the balance between opposite carriers in the region where more efficient radiative
Hybrid colloidal nanocrystal-organics based light emitting diodes 281

exciton recombination can occur. In particular, charge transfer processes to CdSe /ZnS core-
shell quantum dots are found to be the key element for well balanced white emission.

4.2 Bright White LEDs from Ternary Nanocrystals Composites

To obtain a high performance hybrid device in which QDs act as lumophores is the
occurrence of efficient exciton recombination in the inorganic nanocrystals. This is usually
inhibited by the poor electron conduction of the inorganic species which limits exciton
formation (Leatherdale et al., 2000). Therefore low EL efficiency is observed in
QDs/polymers blend based devices. (Colvin et al., 1994; Dabbousi et al., 1995; Schlamp et
al., 1997). The phase-segregation technique proposed by Coe et al. allows for the fabrication
of high efficiency hybrid monochromatic emission devices, but it involves a narrow QDs
size distribution to form high-coverage monolayers, which is compatible only with
monochromatic emitting QD-LEDs (Coe et al., 2002; Coe-Sullivan et al., 2003; Coe-Sullivan
et al., 2005). To date, efficient QD-LEDs that emit white light only from nanocrystals are still
a challenge due to the lack of proper fabrication techniques.
In order to obtain efficient simultaneous emission from different sizes of QDs composites,
we proposed a novel device structures in which exciton formation in the inorganic QDs is
not exclusively obtained by direct charge injection but by the accurate control of the
energy/charge transfer mechanisms from the organic host (Li et al., 2006). We demonstrate
the first efficient hybrid light-emitting device, with near white emission from chemically and
optically stable ternary nanocrystal composites dispersed in an organic matrix. White bright
emission is obtained from homogenous blends, without phase segregation between the
active ternary QDs composites and the organic matrix 4,4,N,N-diphenylcarbazole (CBP),
exploiting the energy transfer and charge-trapping properties of the different species. The
proposed approach is a new general method for the fabrication of potential long operating
lifetime, high efficiency white light-emitting devices. The emission mechanisms and the
structure of the devices and are shown in the fig. 10a, 10b. In fig. 10c the Photoluminescence
(PL) spectra corresponding to isolated lake blue, green and red quantum dots in solid state
are shown. The EL spectrum at 13V is shown in fig. 11a. Spectral peaks at wavelengths of
490nm, 540nm and 618nm are attributed to the emission of lake blue, green and red
quantum dots, respectively, in agreement with the PL spectra in the solid state (see fig. 10c).
A complete quenching of the CBP band is observed in the EL spectrum, resulting in efficient
emission at the three primary colors from the QDs, giving rise to a balanced near white
overall emission [CIE (0.32,0.45)] at 13V. Furthermore, the color coordinates do not
remarkably change with increasing the operating voltage in the range 10V-28V. Possible
pathways leading to emissive states are shown in fig. 10a. In our device, holes are injected
from the ITO contact through the PEDOT:PSS layer into the CBP host matrix, and are
transported towards the QDs. Similarly, electrons are injected from the Ca/Al cathode into
the Alq3 and are transported to the QDs. Then exciton generation on QDs occurs via two
parallel processes: (I) charge-trapping and (II) Förster energy transfer from CBP and Alq3. In
the former process, electrons may be trapped in the QDs owing to the relative energy
alignment of the LUMO levels of CBP, Alq3 and QDs, thus, efficient exciton formation on the
QDs can occur after recombination with high mobility holes. In the latter process, excitons
form on organic molecules CBP (IIa) and Alq3 (IIb), then undergo Förster energy transfer to
the lower-energy QDs sites, where they can recombine radiatively.
282 Nanocrystals

Fig. 10. (a) Proposed simplified energy level diagram of devices and possible exciton
creation mechanisms on the QDs: (I) charge-trapping and (II) Förster energy transfer. (b)
The structure of the device. (c) PL spectra corresponding to isolated lake blue, green and red
quantum dots, measured in solid state. (Li et al. , 2006. Copyright Wiley-VCH Verlag GmbH
& Co. KGaA. Reproduced with permission.).

(a) (b)
Fig. 11. Electroluminescence (EL) spectra and characteristics of the ternary QDs device. (a)
EL spectrum for the device: ITO//PEDOT:PSS//CBP:QDs (B,G,R:18,2,1) //Alq3//Ca/Al.
(b) V-I (dot), V-L (circle) characteristics of the device. Inset: A photo taken from the working
device. (Li et al. , 2006. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with
permission.).
Hybrid colloidal nanocrystal-organics based light emitting diodes 283

Possible pathways leading to emissive states are shown in fig. 10a. In our device, holes are
injected from the ITO contact through the PEDOT:PSS layer into the CBP host matrix, and
are transported towards the QDs. Similarly, electrons are injected from the Ca/Al cathode
into the Alq3 and are transported to the QDs. Then exciton generation on QDs occurs via
two parallel processes: (I) charge-trapping and (II) Förster energy transfer from CBP and
Alq3. In the former process, electrons may be trapped in the QDs owing to the relative
energy alignment of the LUMO levels of CBP, Alq3 and QDs, thus, efficient exciton
formation on the QDs can occur after recombination with high mobility holes. In the latter
process, excitons form on organic molecules CBP (IIa) and Alq3 (IIb), then undergo Förster
energy transfer to the lower-energy QDs sites, where they can recombine radiatively. The
maximum brightness of the device is 1050 cd m-2 at 58 mA cm-2, which corresponds to a
current efficiency of 1.8 cd A-1, and a turn-on voltage of 6V are measured in air atmosphere
fig. 11b. To our knowledge, this is the highest efficiency hybrid device with white emission
only from ternary QDs composites, whose luminance satisfies lighting application
requirements (i.e. 1000 cd m-2).

5. Hybrid Light Emitting Diodes by Micro-Contact Printing

In previous sections different approaches for hybrid LEDs fabrication have been reported.
All these device structures are limited by the chemical properties of the employed materials
and by the deposition techniques. The main consequence is a restriction of the organic
materials choice, especially for the holes transporting and injection layers, which is crucial
for the optimization of the device emission. In fact, unlike organic small molecules, QDs
cannot be deposited by thermal evaporation because of their high molecular weight and
only wet deposition techniques, such as spin-coating and drop-casting are available.
Consequently, the separation of the QD active layer from the organic transport layers and
the fabrication of multilayer organic/inorganic structures are still challenging. The lack of
an appropriate QD deposition technique strongly limits the implementation of Hybrid LEDs
in the heterojunction devices technology. This technology could result in an improvement of
the devices efficiency and lifetime.
In this frame we developed two totally dry, simple, and inexpensive deposition techniques
to transfer colloidal semiconductor QDs onto organic substrates (Rizzo et al., 2008a; Rizzo et
al., 2008b). These innovative techniques are modification of the standard microcontact
printing (μCP). In the standard μCP, a poly(dimethylsiloxane) (PDMS) stamp is inked by the
material which is then transferred onto the solid substrate by conformal contact of the
elastomeric stamp and the substrate. However this technique cannot be used for the
deposition of colloidal semiconductor nanocrystals. PDMS stamp, indeed, can be readily
swelled by a number of non polar solvents such as toluene and chloroform, in which
nanocrystals for optoelectronic applications are usually dispersed in.
The first technique we developed is named μCP-Double Transfer because it consists in a two
step process. First the nanocrystals are deposited from a toluene solution on a float glass
substrate, when the solvent is dried, the PDMS stamp pad is brought in conformal contact
with the nanocrystal film, then the PDMS stamp is peeled away and the nanocrystals are
transferred to the surface of the stamp. Finally, QDs are deposited by single step
microcontact printing from the PDMS stamp to the surface of previously evaporated organic
thin films. (Rizzo et al., 2008a).
284 Nanocrystals

The second technique consists in protecting the PDMS stamp pad with a SU-8 photoresist
layer deposited by spin coating and cross linked in order to make it insoluble to toluene.
(Rizzo et al., 2008b). The two techniques that we propose avoided the swelling of the PDMS
stamp and combined the ease of wet deposition processes with the possibility to grow a
heterojunction QDs-LED using thermal evaporation. White and red electroluminescence
from colloidal QDs has been obtained demonstrating that these deposition techniques are
fully compatible with OLED technology.

5.1 Hybrid Light-Emitting Diodes from μ-CP Double-Transfer

of Colloidal Semiconductor CdSe/ZnS Quantum Dots onto Organic Layers
The sketch of the Double Transfer technique is reported in fig. 12. Drop-casting is the
simplest technique for the deposition of a QDs solution on the glass substrate. The degree of
order in the resulting assembly is mainly determined by the rate of solvent evaporation:
slowing down the evaporation of solvent, in our case toluene, and preventing it from a rapid
dewetting is crucial for a successful double transfer. A slow evaporation allows nanocrystals
to diffuse and self assemble into large periodical (long range order) structures with fewer
defects. In order to slow down the evaporation after the drop deposition, the substrates
were placed in an over-saturated environment of toluene vapour.

Fig. 12. A schematic illustration of the procedure used to transfer colloidal nanocrystals on
organic thin films. 1. Colloidal QDs are deposited from toluene solution on a cleaned float
glass substrate; 2. when the solvent is dried, a PDMS stamp pad is brought in conformal
contact with the surface of the nanocrystals film; 3. The PDMS stamp is then peeled away and
the nanocrystals are transferred to the surface of the stamp; 4. QDs are deposited by single step
microcontact printing from the PDMS stamp to the surface of previously evaporated organic
thin films; 5. Nanocrystals are transferred on the organic substrate (Rizzo et al., 2008a.
Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.).

The quality of the drop-casted films was verified by Transmission Electron Microscope
(TEM) measurements. In fig. 13a and 13b we compare two samples realized depositing 20μL
of the same colloidal solution on a TEM grid in over-saturated environment (fig. 13a) and
Hybrid colloidal nanocrystal-organics based light emitting diodes 285

under ambient condition (fig. 13b). After a slow dewetting of the solvent, (about 30 minutes)
the nanocrystals form a compact layer on a scale of several microns. For a fast solvent
evaporation (less than 5 min) the film presents the formation of multilayer domains and
microscopic voids. Deposition of a uniform QDs layer on the top of the organic layers has a
paramount importance for the correct functioning of the hybrid LED devices. The layer
quality of QDs formed on the surface of PDMS reflects the morphology of the QD film on
the glass substrate, i.e. it depends on the concentration of the nanocrystals solution and on
their assembling on the glass surface.

(a) (b)

(c) (d)
Fig. 13. a) TEM images of nanocrystals deposited by drop-casting on TEM grid after slow
evaporation of the solvent and b) fast evaporation of the solvent. c) Confocal microscopy
image of QDs transferred by double transfer μCP onto CBP organic substrates from a glass
substrate after slow evaporation of the solvent; d) QDs on CBP transferred from a glass
substrate after fast evaporation of the solvent. (Rizzo et al., 2008a. Copyright Wiley-VCH
Verlag GmbH & Co. KGaA. Reproduced with permission.).

In fig. 13c and 13d we show the confocal microscope images of QDs transferred on CBP
layer by μCP Double Transfer from glass substrates after slow (fig. 13c) and fast evaporation
(fig. 13d) of the solvent. The morphology of 50×150μm QDs stripes results uniform for
samples made after slow evaporation of the solvent and very dishomogeneous for the one
made after a fast evaporation of the solvent. This is also a further proof of how effective
slower solvent evaporation is in assuring highly homogeneous film of colloidal QDs at the
286 Nanocrystals

micrometer scale. Increasing 15 times the concentration of QDs in toluene solution, the
thickness of the transferred layer can vary from few layers to more than 100 nm.
The transferring mechanism of QDs from the glass substrate to the stamp can be explained
by considering the strong interaction between the PDMS stamp surface and the colloidal
nanocrystal hydrophobic surfactant. This can be attributed to the existence of highly mobile
hydrophobic oligomers on the PDMS stamp surface. Moreover PDMS has a very low surface
energy of 21.6dyn/cm thus allowing the transfer of the QDs on a higher surface energy
substrate, such as the CBP. In this contest the organic material acts as glue for the
nanocrystals.
In order to realize a hybrid LED that emits light from nanocrystals alone, a high coverage
and compact QDs layer has to be deposited on the top of the organic material. For this
reason in the device fabrication we decided to use a flat PDMS pad for the QDs transfer. The
device structure (fig. 14a, inset) consists of a thermally evaporated N,N'-Bis(naphthalen-1-
yl)-N,N'-bis(phenyl)benzidine (α-NPD) hole injection layer (HIL) doped with 2,3,5,6-
tetrafluoro-7,7,8,8-tetracyano-quinodimethane (F4-TCNQ) and CBP hole transporting layer
(HTL). A QDs active layer of 80 nm is transferred on the organic substrate by the Double
Transfer technique reported above. The 2,9-dimethyl-4,7-diphenyl-1,10-phenanthroline
(BCP) hole blocking layer (HBL), the Alq3 electron transporting layer (ETL), and the
electrodes are evaporated after the QDs deposition.

(a) (b)
Fig. 14. Structure and performance of a hybrid LED fabricated by μCP double transfer. a)
Current-density–voltage and luminescence–voltage characteristics of the hybrid LED; inset:
a schematic illustration of the device structure. b) Electroluminescence (solid line) and PL
(dashed line) spectra; the inset shows a picture of a functional 1.5mmx1.5mm area device
(Rizzo et al., 2008a. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with
permission.).

In the proposed device the holes and electrons are delivered to the nanocrystals layer
through the injection and transport layers. The electroluminescence (EL) spectrum and the
picture of the working device in fig. 14b demonstrate that the emission is dominated by QDs
band edge emission. Photoluminescence (PL) spectrum of the nanocrystals is reported in
dashed line in the same figure.
In fig. 14a a current density versus voltage and luminescence versus voltage characteristics
are reported. The maximum brightness of the device is 90 cd m-2 at 20 mA cm-2, which
corresponds to an external quantum efficiency of 0.15%.
Hybrid colloidal nanocrystal-organics based light emitting diodes 287

5.2 White Electroluminescence from a µ-CP

Deposited CdSe/ZnS Colloidal Quantum Dot Monolayer
In Figure 15 a schematic sketch of the second approach for QD deposition is displayed
(Rizzo et al., 2008b).

1 2 3 4
Fig. 15. A schematic sketch of the modified μCP technique used to deposit colloidal
nanocystals on organic thin films. 1. The SU-8 is spin-coated on the top of a flat PDMS stamp
pad, then exposed to UV radiation and subsequently thermally cross-linked in order to
make the film insoluble to toluene; 2. a mixed blue, green and red CdSe/ZnS QD solution is
spin-coated at on the SU-8 protected PDMS; 3. the inked stamp is then inverted on the
surface of a previously evaporated organic thin film and peeled away after 30 second; 4.
nanocrystals are transferred on the organic substrate. (Rizzo et al., 2008b. Copyright Wiley-
VCH Verlag GmbH & Co. KGaA. Reproduced with permission.

In fig. 16 we report atomic force microscopy (AFM) images of the same film. Fig. 16a was
taken at the interface between the CBP and QDs layer at the edge of a pinhole defect. By the
profile (fig. 16b) we can estimate QD film thickness of about 10nm. In fig. 16c, 16d, 16e, and
16f AFM images of the QD film at different resolutions are reported. Pictures in fig. 16a, 16c,
16d, point out the uniform distribution of the QDs on the organic material with an average
roughness of about 5nm. Moreover in the higher resolution images in fig. 16e and 16f is
possible to discriminate the two phases: QDs from 4-10 nm in diameter on the top and
organic CBP as underlayer. The film such prepared was used for the fabrication of white
emitting hybrid LED. In Fig. 17 the electroluminescence spectra at different applied voltages
are reported, showing an increase of the green and blue components relative to the red one
when voltage is increased from 5V to 10V. This is consistent with the larger barrier for holes
injection in the smaller QDs. Two dominant QD excitation mechanisms are present in the
hybrid device: the Förster energy transfer and the charge trapping. In the charge trapping
process the electrons may be trapped in the QD owing to the absence of any energy barrier
between the organic Alq3 and BCP layers and CdSe/ZnS one. For these charged QDs the
barrier to holes injection and transfer from the α-NPD/CBP bilayer is reduced. Thus, exciton
formation and consequent radiative recombination occur on QD sites upon an acceptance of
high mobility holes from CBP. The presence of a α-NPD (420nm) small band emission is due
to the presence of pinhole defects, grain boundaries, interstitial spaces in the QD thin layer
that allow the formation of exciton on organic sites. The HOMO level alignment of α-
NPD/CBP suggests an accumulation of holes in the α-NPD at the HIL/HTL interface.
288 Nanocrystals

Fig. 16. Atomic force microscopy (AFM) characterization of the transferred nanocrystal
layer. a) AFM image of the layer at the edge of a pinhole defect and b) profile corresponding
to a); c), d), e), f), AFM images of the QD film at different scanning sizes. (Rizzo et al., 2008b.
Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with permission.).

(a) (b)
Fig. 17. a) EL spectra of the device at different applied voltages normalized at the red peak.
b) Current-Voltage I-V characteristic and current efficiency (CE) for the hybrid device. .
(Rizzo et al., 2008b. Copyright Wiley-VCH Verlag GmbH & Co. KGaA. Reproduced with
permission.).

By increasing the voltage and, consequently the current, the fraction of emission coming
from the organic layer increases. The development of these two techniques enables us to
separate the deposition of the organic materials from the QDs deposition and the fabrication
of multilayer complex device structures without any restriction in the organic material
choice. By taking advantage from the independent processing of QD we fabricated red and
white emitting multilayer LEDs with a doped holes injection layer.
Hybrid colloidal nanocrystal-organics based light emitting diodes 289

6. Conclusion
Several approaches have been followed for the incorporation of inorganic colloidal
semiconductor nanocrystals in organic light emitting diodes technology. As a results we first
reported two methods for the fabrication of white emitting hybrid organic/inorganic LEDs
wherein colloidal CdSe/ZnS QDs are blended in a blue emitting polymer or in a high
mobility hole transporting organic small molecule. The major breakthrough has been
reached with the fabrication of white hybrid LEDs from ternary nanocrystal composites.
CdSe/ZnS QDs with different sizes simultaneously emit light at different wavelength (blue,
red and green region) to achieve white emission. By accurately controlling the Förster
energy transfer and the charge trapping on QDs sites we were able to fabricate the highest
efficiency hybrid device with emission only from QD composites. The maximum satisfies
the lighting application requirements. Moreover the development of two innovative micro-
contact printing techniques allow the fabrication of multilayer hybrid LEDs structure
without restrictions in the organic under-layers. These deposition methods permit the
integration of colloidal QDs in the promising p-doping organic technology. The recent
results demonstrate that the successful integration of colloidal nanocrystals in OLED
technology could enable the creation of a new set of promising devices with a possible
improvement in the color stability and the device lifetime.

7. References
Murray, C. B.; Norris, D. J.; Bawendi, M. G. Journal of the American Chemical Society 1993, 115,
(19), 8706-8715.
Murray, C. B.; Sun, S. H.; Gaschler, W.; Doyle, H.; Betley, T. A.; Kagan, C. R. Ibm Journal of
Research and Development 2001, 45, (1), 47-56.
Sun, S. H.; Murray, C. B. Journal of Applied Physics 1999, 85, (8), 4325-4330.
Rockenberger, J.; Scher, E. C.; Alivisatos, A. P. Journal of the American Chemical Society 1999,
121, (49), 11595-11596.
Ginger, D. S.; Greenham, N. C. Journal of Applied Physics 2000, 87, (3), 1361-1368.
C. A. Leatherdale, C. R. Kagan, N. Y. Morgan, S. A. Empedocles, M. A. Kastner, and M. G.
Bawendi, “Photoconductivity in CdSe quantum dot solids”, Phys. Rev. B, 62, 2669-
2680, 2000.
V. L Colvin, M. C. Schlamp, and A. P. Alivisatos, “Light-emitting diodes made from
cadmium selenide nanocrystals and a semiconducting polymer”, Nature, 370, 354-
357, 1994.
Morgan, N. Y.; Leatherdale, C. A.; Drndic´, M.; Jarosz, M. V.; Kastner, M. A. & Bawendi, M.
(2002). Electronic transport in films of colloidal CdSe nanocrystals, Phys. Rev. B, 66,
075339
Dabbousi, B. O.; Bawendi, M. G.; Onitsuka; O. & Rubner, M. F. (1995). Electroluminescence
from CdSe quantum-dot/polymer composites, App. Phys. Lett., 66, 1316-1318
Schlamp, M. C.; Peng, X.; & Alivisatos, A. P. (1997). Improved efficiencies in light emitting
diodes made with CdSe(CdS) core/shell type nanocrystals and a semiconducting
polymer”, J. App. Phys., 82, 5837-5842
Coe, S.; Woo, W. K.; Bawendi, M.; & Bulovic, V. (2002). Electroluminescence from single
monolayers of nanocrystals in molecular organic devices Nature, 420, 800-803
290 Nanocrystals

Coe-Sullivan, S.; Woo, W. K.; Steckel,J. S.; Bawendi, M. G.; & Bulovic, V. (2003). Tuning the
performance of hybrid organic/inorganic quantum dot light-emitting devices, Org.
Electron. 4, 123-130
S. Coe-Sullivan, J. S. Steckel, W. K. Woo, M. G. Bawendi, & V. Bulovic, (2005) Large-Area
Ordered Quantum-Dot Monolayer via Phase Separation During Spin-Casting, Adv.
Func. Mater. 15, 1117-1124
Zhao, J.; Bardecker, J. A.; Munro, A. M.; Liu, M. S.; Niu, Y. H.; Ding, I.-K.; Luo, J.; Chen, B;
A. K.-Y., Jen; & Ginger, D. S. (2006). Efficient CdSe/CdS Quantum Dot Light-
Emitting Diodes Using a Thermally Polymerized Hole Transporting Layer, Nano
Lett. 6, 463-467
Niu, Y. H.; Munro, A. M.; Cheng, Y.-J.; Tian, Y. Q.; Liu, M. S.; Zhao, J.; Bardecker, J. A.;
Plante, I. J.-L.; Ginger, D. S.; & Jen, A. K.-Y. (2007). Improved Performance from
Multilayer Quantum Dot Light-Emitting Diodes via Thermal Annealing of the
Quantum Dot Layer, Adv. Mater. 19, 3371-3376, 2007
Sun, Q.; Wang, Y. A.; Li, L. S.; Wang, D. Y.; Zhu, T.; Xu, J.; Yang, C. H.; & Li, Y. F. (2007).
Bright, multicoloured light-emitting diodes based on quantum dots Nature
Photonics, 1, 717-722.
Caruge, J. M. Halpert, J. E. Wood, V. Bulovic´, V. & Bawendi, M. G. (2008). Colloidal
quantum-dot light-emitting diodes with metal-oxide charge transport layers Nature
Photonics 2, 247-250
Cho, K.-S.; Lee, E. K.; Joo, W.-J.; Jang, E.; Kim, T.-H.; Lee, S. J.; Kwon, S.-J.; Han, J. Y.; Kim,
B.-K.; Choi, B. L.; & Kim, J. M. (2009). High-Performance Crosslinked Colloidal
Quantum-Dot Light-Emitting Diodes Nat. Photonics, 3, 341– 345
Suzuki, H.; & Hoshino, S. (1996). Effects of doping dyes on the electroluminescent
characteristics of multilayer organic light-emitting diodes, Journal of Applied Physics
79, (11), 8816-8822
Li, Y. Q.; Rizzo, A.; Mazzeo, M.; Carbone, L.; Manna, L.; Cingolani, R.; & Gigli, G. (2005)
White organic light-emitting devices with CdSe/ZnS quantum dots as a red emitter
Journal of Applied Physics, 97, 113501
Li, Y. Q.; Rizzo, A.; Cingolani, R.; & Gigli, G. (2006). Bright white-light-emitting device from
ternary nanocrystal composites Advanced Materials, 18, 2545-2548.
Förster, T. (1959). Transfer mechanisms of electronic excitation, Discuss. Faraday Soc, 27, 7-17
Lakowicz, J. R. Principles of Fluorescence Spectroscopy. Plenum (New York): 1983.
Utsugi, K.; & Takano, S. (1992). Luminescent Properties of Doped Organic EL Diodes Using
Naphatalimide Derivative, Journal of the Electrochemical Society, 139, 3610-3615.
Rizzo, A.; Mazzeo, M.; Palumbo, M; Lerario, G.; D’Amone, S.; Cingolani, R.; & Gigli, G.
(2008). Hybrid Light-Emitting Diodes from Microcontact-Printing Double-Transfer
of Colloidal Semiconductor CdSe/ZnS Quantum Dots onto Organic Layers, Adv.
Mater. 20, 1886-1891
Rizzo, A.; Mazzeo, M.; Biasiucci, M.; Cingolani, R.; & Gigli, G. (2008) White EL from
Microcontact Printing Deposited Colloidal Quantum Dot Monolayer, Small 4, 2143-
2147
Seasoning ferroelectric liquid crystal with
colloidal nanoparticles for enhancing application flavors 291

0
13

Seasoning ferroelectric liquid crystal with colloidal

nanoparticles for enhancing application flavors
Jung Y. Huang
Department of Photonics, Chiao Tung University
The Republic of China at Taiwan

1. Introduction
The properties of nanostructured materials can be significantly different from those of bulk
materials. This prospect has attracted immense academic and industrial interests to develop
new synthetic schemes of nano building blocks and self-assembling methodologies for
efficiently manipulating a variety of building blocks into functional materials. Liquid crystal
(LC) materials possess both spatial order and mobility at the molecular level, thus may serve
suitable candidates for self-assembly of nanoscale materials. The LC-based nanoscience (Heg-
mann et al. (2007)) has recently made significant progress to cover the topics of the synthesis
of nanomaterials using LCs as templates, the design of LC nanomaterials, self-assembly of
nanomaterials in LC phases, and defect formation in LC-nanoparticle suspensions.

For the synthesis of nanomaterials and self-assembly of nano objects in LC, researchers had
demonstrated the shape-selective synthesis of gold nanoparticles in a liquid crystal medium
and assembled these nanoparticles to spontaneously form ribbonlike patterns (Mallia et
al (2007)). For the design of LC nanomaterials, suspensions of ferroelectric nanoparticles
in a nematic liquid crystal (NLC) host had been shown to possess an enhanced dielectric
anisotropy and are sensitive to the sign of an applied electric field (Cheon et al. (2005)). Large
colloidal particles in a LC-nanoparticle suspension can cause strong director deformations
and that usually lead to defect formation in LC matrices. By controlling the particles’ shapes
researchers can disturb the uniform alignment of the surrounding nematic host to generate
highly directional pair interactions (Lapointe et al (2009)). Therefore, researchers in the
LC-based nanoscience are now able to tailor colloidal interactions to realize high-precision
self-assembly in anisotropic nematic fluids.

Small particles dispersed in LC at high concentration create almost a rigid LC suspension.

From simulation (Hung & Bale (2009)), a nanocube in a NLC medium with perpendicular
anchoring at the nanocube’s faces tends to align in such a way that none of its faces is parallel
or perpendicular to far-field LC director n(r). A triangular nanoprism with homeotropic
anchoring of the nematic at its surfaces tends to align with both its long axis and one of its
rectangular faces perpendicular to n(r). For systems of nanoprisms in NLC, inverted parallel
arrays (the long axes of the nanoprisms are parallel, and one of the prisms is inverted with
respect to the other one) are thermodynamically more stable than linear arrays (the long axes
of the particles are collinear and the particles have the same orientation), which in turn are
292 Nanocrystals

more stable than parallel arrays (the long axes of the particles are parallel and the particles
have the same orientation). The minima observed in the potentials of mean force curves
for the inverted parallel and linear arrays are significantly deeper than that observed for
the parallel array. These NLC-mediated, anisotropic interparticle interactions can make the
particles bind together at specific locations, and thus could be exploited to assemble the
particles into ordered structures with different morphologies.

At low concentrations, LC nanocolloids appear similar to a pure LC with unique properties

(Ouskova et al. (2003)). The diluted suspensions are stable, because the small concentration
of nanoparticles does not significantly perturb the director field in the LC, and interaction
between the particles is weak. Most importantly, the nanoparticles can share their intrinsic
properties with the LC matrix due to the anchoring with the LC. For engineering appli-
cations (Kobayashi (2009)), nanoparticle-embedded LCDs using nanoparticles of metals,
semiconductors, inorganic oxide dielectrics and polymers had been shown to reduce both
the threshold voltage and response time, to improve frequency modulation electro-optical
response, and even to yield a new display mode.

Unlike NLC, which possesses only orientational order, ferroelectric liquid crystals (FLC)
exhibit smectic layer structures in which the director in the smectic layers tilts with respect to
the layer normal (Lagerwall (1999)). Clark and Lagerwall (Clark & Lagerwall (1980)) invented
the concept of surface-stabilized ferroelectric liquid crystal (SSFLC) to realize the attractive
properties of FLC materials, which include microsecond response time, bistability and wide
viewing angle. However, to meet the industrial application criteria, further improvements
on the properties of FLC are required. Modifying an existing FLC material by doping with
appropriate nanoparticles could produce new FLC by blending instead of synthesizing new
mesogenic molecules and therefore had attracted significant interest (Kaur et al. (2007);
Li et al (2006); Reznikov et al. (2003); Shiraishi et al. (2002)). However, the behaviors of a
dilute suspension of colloidal nanometer-scale particles in smectic medium remain elusive.
Further study to yield insight into the interaction between nanoparticle and FLC species
(Huang et al. (2008)) is crucial not only for designing better blending schemes but also for the
understanding of this new material system.

This chapter aims to provide such an insight. By using FLC doped by ZnO nanocrystals
(nc-ZnO) as a model system, we illustrate several important concepts such as what material
properties can be modified by the doping and what physical processes likely be involved.
We first applied the capacitance-voltage (CV) measurement technique to reveal the roles of
multiple dipolar species in a typical SSFLC film (Li & Huang (2009)). We found that nc-ZnO
doping into the SSFLC not only enhances the dynamic polarization of the beneficial dipolar
species but also minimizes the polarization response of the species with counter effects. We
then exploited two-dimensional infrared (2D IR) absorption correlation technique to verify
that ZnO nanocrystals are dispersed uniformly into the FLC host and the uniform doping
leads to an improved aligning structure of FLC molecules (Huang et al. (2008)). We proposed
that the ZnO nanodots tie up surrounding FLC dipolar species and yield a molecular binding
effect. Finally, dynamic light scattering (DLS) was employed to show that nc-ZnO doping
increases the elastic constant of FLC and renders the mechanical response of the FLC film into
under-damping.
Seasoning ferroelectric liquid crystal with
colloidal nanoparticles for enhancing application flavors 293

2. Preparation and Characterization of Materials

2.1 Material Preparation
Zinc oxide nanocrystals were chosen for this study in view of the low-cost synthetic process
with non-toxic chemicals. We synthesized colloidal ZnO nanoparticles with the procedure
reported in the literature (Goyal et al. (1993); Meulenkamp (1998)). To obtain a stabilized
nc-ZnO colloid, we coated the nc-ZnO with 3-(trimethoxysilyl) propyl methacrylate (TPM)
and purified the colloid as reported in the literature. The TEM image reveals the nc-ZnO
colloid to have a size distribution of a mean diameter of 3.2 nm and a full width-at-half
maximum (FWHM) of 1 nm (Huang et al. (2008)).

FELIX 017/100 FLC (from Clariant) with spontaneous polarization of Ps =47 nC/cm2 at 25o C
−28◦ C 73◦ C 77◦ C 85◦ C
exhibits a thermotropic transition sequence of Cr ←→ SmC*←→ SmA*←→ N*←→ Iso. The
material properties of the FLC related to this study are summarized in Table 1. To prepare
doped FLC material, an appropriate amount of ZnO nano powder was added into the FLC
and then homogenized with ultrasonic at 85o C for 40 minutes. The substrates of each SSFLC
cell are indium tin oxide (ITO)-coated glass plates of 12 mmx15 mm in dimension, which
were coated with SE7492 polyimide (from Nissan Chemical) and rubbed unidirectionally
to produce uniform FLC alignment. We assembled SSFLC cells with two anti-parallelly
rubbed substrates and maintained the cell gap with 2-µm diameter silica balls dispersed in a
UV-curable gel NOA65. The nc-ZnO doped FLC in the isotropic phase was filled into a test
cell and then cooled slowly to 35o C. All SSFLC cells prepared for this study exhibit a stable
single SmC* domain under an optical conoscope.

2.2 Material Characterization

The characterization of nanoparticles requires highly sophisticated analytical tools. Such
tools must be suited for the studies where the nanoparticles are dilutely doped into a complex
medium. Generally speaking a combination of analytical techniques results in the best
characterization. In this study, a variety of characterization techniques were used to establish
(i) whether the nanoparticles were distributed uniformly in the ferroelectric liquid crystal
medium. (ii) Was there any influence of nc-ZnO doping on the structural ordering and
dynamic response of the SSFLC? (iii) What is the physical picture related to the interaction be-
tween the nc-ZnO and ferroelectric liquid crystal molecules? Combined capacitance-voltage
measurement, two-dimensional infrared absorption correlation technique, and dynamic light
scattering seem to be useful for characterizing the nc-ZnO doped ferroelectric liquid crystal
in the smectic C* phase.

The C-V measurements of the SSFLC cells (Li & Huang (2009)) were carried out at 25o C with
a computer-controlled HP4284A LCR meter, which provides a bias voltage range of 35 V
and a sinusoidal wave of 1.0 V with a frequency range of 20 Hz?1 MHz. We first used C-V
measurement technique to ensure each empty cell to have a voltage-independent capacitance
about 717±35 pF at 1 kHz.

Figure (1) presents the schematic of a SSFLC cell in an infrared absorption spectrometer
(Huang et al. (2008)). Fourier-transform infrared (FTIR) spectra from 900 to 3500 cm−1
with 4-cm−1 resolution were recorded with a liquid nitrogen-cooled HgCdTe detector and
a home-made data acquisition electronics. For time-resolved FTIR measurement, bipolar
294 Nanocrystals

square-wave pulses were used to drive the test cell. The waveform is comprised of a
+10V pulse from 0 to 140 µsec, followed by a field-free period extending from 140 to 500
µsec. Then a -10V pulse extends from 500 to 640 µsec, followed by a field-free duration
from 640 to 1000 µsec. For each polarization direction of the incident infrared beam, a
total of 32 time-resolved interferograms were acquired. The 2D correlation analysis with
varying infrared polarization angles were calculated based upon an algorithm developed by
Noda (Noda (1993)), and was implemented in a software 2Dshige by S. Morita (Morita (2004)).

Fig. 1. Configuration of SSFLC in a FTIR apparatus. Z denotes the rubbing direction, which is
also the layer normal of the smectic layers. E is the direction of the applied electric field. M
on the right diagram denotes the IR dipole, which tilts from the molecular long axis ξ with an
angle β and can rotate about the ξ-axis by γ.

Figure (2) presents the schematic diagram of the DLS apparatus used to probe thermal fluc-
tuations of a SSFLC cell. A He-Ne laser with a wavelength of 632.8 nm was used as the
excitation light source. The test cell was inserted in a setup with crossed polarizer and ana-
lyzer. We adjusted the incident light polarization to be along to the rubbing direction of the
test cell. We adopted a scattering geometry with the scattering vector lying on the cell’s sub-
strates and being perpendicular to the rubbing direction. This experimental geometry ensures
that the incident photons can only be scattered by the fluctuation of the azimuthal angle ϕ(t)
of FLC directors. Furthermore, a forward scattering geometry with a small scattering angle
(∼ 4o ) was used to ensure only the fundamental mode of thermal excitation of the cell to be
observed. The scattered photons were detected with a silicon avalanche photodiode (SPCM-
AQR-15, PerkinElmer) and then fed into the Flex02-01D digital correlator (Correlator.com, NJ)
to calculate the intensity autocorrelation function in real time. The SSFLC test cells in all mea-
surements were kept at a specific temperature with ±0.1o C precision.
Referring to the coordinate system described in Fig. 1, the intensity of the transmitted light
through the setup can be expressed as (Huang et al. (2008))
Γ
I = I0 [sin2 2θsin2 ]cos2 ϕ, (1)
2

where Γ = k∆n d 2 denotes the phase retardation caused by the SSFLC film and d is the film
thickness.
Seasoning ferroelectric liquid crystal with
colloidal nanoparticles for enhancing application flavors 295

Fig. 2. Experimental configuration of DLS. M: mirror; P1 and P2: polarizers; L: Lens; D:

diaphragm. Here the polarization of the incoming beam is set to be parallel to the rubbing
direction of the SSFLC cell, but perpendicular to both the scattering plane and the polarization
analyzing direction for the output scattering beam.

3. Experimental Results
Note that ZnO possesses a wurtzite structure that renders ZnO nanoparticle with a di-
ameter of 2R0 =3.2 nm to have a permanent dipole moment of µ ZnO =50 Debye (Nann &
Schneider (2004); Shim & Guyot-Sionnest (1999)). It is inspiring to first estimate the en-
ergy density of the dipolar interaction resulting from nc-ZnO doping. Being nano-size ob-
jects, the nc-ZnO particles shall behave like a molecular dopant in SSFLC. The dipole-dipole
interaction energy between nc-ZnO
 and one of the surrounding dipolar species µCO reads
V = µ ZnO µCO (1 − 3 cos2 θ ) (4πε 0 r3 ). Because the dipole-dipole interaction depends on
the relative orientations θ, the molecules can exert forces on one another and in fact can not
rotate freely. The orientation with lower energy is favoured so there is a non-zero dipolar
interaction energy. At a uniform doping concentration of one weight percent, we estimated
one ZnO nanoparticle to be surrounded by about n FLC =15000 molecules within a sphere of
Rmax =14 nm centered at the nc-ZnO. The thermal average of the dipole-dipole interaction
energy density between nc-ZnO and surrounding FLC molecules becomes (Keesom (1921);
Magnasco et al. (2006))

2 µ2CO µ2ZnO n FLC NZnO

UaZnO− FLC = − ; (2)
3 (4πε 0 )2 k B T R30 R3max

We estimated UaZnO− FLC ≈ 1000J · m−3 with NZnO = 1.4 × 1023 m−3 , µCO =1.5 Debye,
and k B T = 4 × 10−21 J. This interaction energy has an alignment effect similar to that FLC
molecules will experience within a distance of 100 nm from an alignment surface of an
anchoring strength 1 × 10−4 J · m−2 and shall produce an observable effect in our SSFLC cells.

Indeed, by inserting a doped SSFLC cell into a setup of crossed polarizer and analyzer and
rotating the cell about the substrate normal, we found the resulting azimuthal pattern of the
optical transmittance has a reduced light leakage in the dark state by 9 times while the optical
throughput in the bright state is increased by 1.2 times, leading to an increased contrast by a
factor of 10 (Li & Huang (2009)).
296 Nanocrystals

3.1 The Influence of nc-ZnO Doping on the CV Characteristics of SSFLC

The origin of the observed improved alignment of FLC molecules by nc-ZnO doping can be
revealed in more detail with the C-V characterization technique. Figs. 3(a) and 3(b) present
two typical CV characteristic curves of SSFLC obtained with a negative-to-positive voltage
sweep. We can use Preisach model to derive the CV characteristic curve of a SSFLC film with
multiple dipolar species (Li & Huang (2009))
n
Psi · δi
Ccell (Vex ) = Cconst + ∑ cosh2 [δ ∓
· (Vex ± VCi )]
A. (3)
i =1 i
where Cconst denotes the voltage-independent part of the capacitance of FLC; A the cell area;
P the polarization, and Vex the driving voltage. Psi in Eq. (3) represents the spontaneous
polarization as all dipoles of the ith species are fully aligned to an external field. VCi is the
coercive voltage at which the electric polarization of the ith species vanishes during switching
from one orientation to the other orientation (Blinov & Chigrinov (1993); Demus et al (1998)).
δi = log{[1 + ( Pr /Ps )i ] [1 − ( Pr /Ps )i ]}/VCi is a constant related to species with Pr and Ps
denoting the remnant and spontaneous polarizations, respectively. The (+/−) sign refers
to an increasing/decreasing Vex . The CV curves of Fig. 3 can fit to Eq. (3) with four dipolar
species being involved in the SSFLC films. From the data fitting, we found that by doping
with nc-ZnO, the peak capacitance contributed by species 1 is increased while capacitances
from species 2, 3 and 4 are reduced. ZnO nanocrystal doping also decreases Vc of species 1 by
1.2V, while increases Vc of species 2 and 4 by 1.5V and 0.5V, respectively.

(a) (b)

Fig. 3. CV characterization of SSFLC. The CV characteristic curves (open squares) at 1 kHz

and the corresponding fitting results (solid lines) from the four dipolar species in a SSFLC cell
(a) without and (b) with nc-ZnO doping.

One of the major advantages of SSFLC is microsecond response time for true video rate
applications. Therefore, it is also interesting to investigate how Ps of the four dipolar species
behaves at high driving frequencies. Figure 4(a) shows that nc-ZnO doping results in an
increase in Ps of species 1 while that from the remaining three species are diminished; the
doping-induced spontaneous polarization change (∆Ps ) of the four species decays with
frequency, reflecting more and more FLC molecules can not catch up the driving field at high
frequency. These findings strongly support the notion that doping with nc-ZnO enhances
the dynamic polarization of the beneficial dipolar species while it minimizes the polarization
response of the species with counter effects.
Seasoning ferroelectric liquid crystal with
colloidal nanoparticles for enhancing application flavors 297

(a) (b)

Fig. 4. Polarization response from dipolar species.(a) The doping-induced spontaneous po-
larization changes (∆Ps ) of the four dipolar species in a SSFLC cell as a function of driving
frequency. (b) Schematic drawing illustrating the dipolar interaction of species 1 in an un-
doped SSFLC cell (left) and the dipolar interaction of species 1 with nc-ZnO in a doped SSFLC
cell (right). The inset figure shows the smectic layer structure of SSFLC and Pspe1 denotes the
spontaneous polarization of species 1.

Figure 4(b) presents a schematic drawing illustrating our concept about the dipolar interac-
tion of species 1 in an undoped and a doped SSFLC cell. Dipoles associated with species 1
could adjust their spatial distribution and likely aggregate on those regions near the north
and south poles of the nc-ZnO, and accordingly align in the same direction with the nc-ZnO
dipole. The number of the dipolar species 1 lying on the equatorial plane of the nc-ZnO
becomes smaller. By this way, Ps from species 1 can therefore be increased by nc-ZnO doping.

3.2 2D IR Study on the Effect of nc-ZnO Doping in SSFLC

Two-dimensional IR correlation is a technique where the spectral intensity is plotted as
a function of two independent spectral variables. By spreading spectral peaks along the
second dimension, one can gain an advantage of sorting out complex or overlapped spectral
features that normally cannot be resolved in a one-dimensional spectrum. Generalized 2D
IR spectroscopy is an effective mathematical tool to elucidate spectral details of a dynamic
system (Ozaki & Noda (2000)). The data yielded from the technique are usually presented
with synchronous and asynchronous plots. In our case, the synchronous plot reveals the
information about the in-plane order and the similarity in the IR azimuthal angular patterns,
while the asynchronous plot offers the information about the dissimilarity in the azimuthal
angular patterns. Therefore, the information about the alignment of submolecular species and
their field-induced switching dynamics can be effectively probed with the 2D IR absorption
correlation technique.

In Fig. 5, the field-free 2D synchronous plots in the 2820-3000 cm−1 region were shown on
the left column and the asynchronous on the right-hand side for the undoped (first row) and
the nc-ZnO doped (second row) SSFLC cells. The major auto-peak appearing at 2926cm−1
is assigned to the anti-symmetric stretching of CH2 group (a-CH2) and reflects an in-plane
alignment order of a-CH2. The appearance of the cross-peaks A: 2926 (a-CH2) vs. 2840 cm−1
(s-CH2), B: 2926 vs. 2876 cm−1 (s-CH3), C: 2926 (a-CH2) vs. 2900 cm−1 and D: 2975 (a-CH3)
298 Nanocrystals

(a) (b) (c)

(d)

Fig. 5. 2D IR plots of CH stretching modes. Synchronous (a and c) and asynchronous (b and d)

2D IR correlation plots from the alkyl-chain modes of SSFLC cells without (a and b)and with
nc-ZnO doping (c and d).

vs. 2926 cm−1 indicates the IR-active dipoles of CH2 and CH3 to be angularly correlated.
Notice that the asynchronous cross-peak can reflect relative orientation difference of IR-active
dipoles. Two cross peaks B and B in Figure 5(b) can therefore be attributed to be s-CH3
from different species. The cross peaks A, B and C were found to be positive while B and
D to be negative. By defining Φ0 as the apparent angle of the maximum IR absorbance,
the asynchronous cross-peak can be expressed as Ψ a (ν1 , ν2 ) = 0.11 sin 2[(Φ0 (ν1 ) − Φ0 (ν2 )]
(Huang & Shih (2006)), indicating that IR-active modes associated with the cross peaks A, B,
and C relative to that of the a-CH2 mode at 2926-cm−1 have smaller Φ0 while peaks B and D
have larger Φ0 .

In the synchronous plot, the major auto-peak of the SSFLC doped with nc-ZnO appearing
at 2926 cm−1 is more intense than the undoped sample, indicating an improved alignment
order with nc-ZnO doping. In the asynchronous plot, the cross peaks C and D of the doped
sample become in phase, suggesting that CH2 and CH3 in the doped SSFLC are packed into
a more ordered structure. This shall happen only when the ZnO nanodots are dispersed
uniformly in the FLC medium.

In Fig. 6, the synchronous 2D IR plots from the core groups, the ring C=C stretchings of the
FLC were presented. The cross-peaks of C(1514 vs. 1608 cm−1 ), B(1584 vs. 1608 cm−1 ) are
more distinctive with nc-ZnO doping, indicating a more ordered structure of the core groups
in the doped SSFLC cell.
Seasoning ferroelectric liquid crystal with
colloidal nanoparticles for enhancing application flavors 299

(a) (b)

Fig. 6. 2D IR plots of core groups. Synchronous 2D IR correlation plots from the core groups
of SSFLC cells (a) without and (b) with nc-ZnO doping.

By using time-resolved 2D IR correlation technique, the field-induced switching of SSFLC

can be finely resolved into the reorientation of submolecular species. We had successfully
exploited a global 2D phase defined as tan−1 [Ψ a (ν1 , t ; ν1 , 0) Ψs (ν1 , t ; ν1 , 0)] to reflect the
orientational variations of IR-active modes (Huang et al. (2008)). The resulting 2D IR phase
angles of the IR-active modes associated with the FLC cores and alkyl chains are presented
in Fig. 7. The phase angles change sign with the polarity of the driving field, indicating
that during the field-induced reorientation the IR-active molecular dipoles projected onto the
substrate surface point to two opposite sides of their corresponding field-free directions. The
CH2/CH3 stretching modes associated with the alkyl chains are loosely connected, resulting
in larger angular spreads than that the core groups. By doping nc-ZnO into SSFLC, the
orientational variations of the IR-active dipoles reveal smaller angular spreads, suggesting
doping with nc-ZnO yields more concerted reorientation processes.

3.3 Dynamic Light Scattering from pure SSFLC and nc-ZnO Doped SSFLC
The experimental results described above clearly indicate that a dilute suspension of colloidal
nanometer-scale particles in SSFLC film can significantly improve both the field-free align-
ment and the field-induced dynamic response of the film at the submolecular level. However,
due to the unavoidable random impacts from thermal excitation, the SSFLC molecules behave
stochastically. The random force constantly pushes FLC molecules away from the orientations
of minimal elastic energy. As the FLC molecules return to their equilibrium orientations, they
release the elastic energy into the rotational kinetic energy and through which, the kinetic
energy is dissipated via the viscous dragging force. To advance our knowledge of the new
ferroelectric liquid crystal material, we further conducted a series of dynamic light scattering
measurements.

Dynamic light scattering (DLS) is an efficient technique to probe the stochastic behavior
of a material. The methodology can yield an autocorrelation function G (t) =  Is (t) Is (0)
(Berne & Pecora (2000)) of the optical scattering intensity Is (t) resulting from the SSFLC film
under study. Figure 8(a) presents the autocorrelation curves of undoped SSFLC acquired at
varying temperatures. We can find several characteristic features from the measured DLS
curves. First, autocorrelation at short delay time G (t → 0) increases with temperature of the
300 Nanocrystals

Fig. 7. Dynamic responses of functional groups. Time courses of 2D IR correlation phase

angles of IR-active molecular normal modes of SSFLC with (open symbols) or without (filled
symbols) nc-ZnO doping. The dashed line denotes the bipolar square wave form of applied
electric field used.

SSFLC film. Secondly, at higher temperature the autocorrelation at long delay time G (t → ∞)
relaxes to a higher level than that at lower temperature. Note that the value of G (0) can be
related to the 
total number of fluctuating domains lying in the optically illuminated region
by N = G (0) ( G (0) − 1). Although the SSFLC cells used for this study were prepared to
have single domain, the cells do not fluctuate in unison. Instead, they are divided into about
500 coarse grains. By using G (0) of Fig. 8(a) and a light spot of 20 µm, we estimated each
thermally fluctuating domains to have a diameter of about 1 µm.

Fig. 8(b) presents the DLS curve of doped SSFLC with 1% nc-ZnO measured at 45o C. An
interesting oscillating feature can be seen in the curve, which most likely originates from
an increase in elastic constant by the nc-ZnO doping. The increased elastic constant could
convert the mechanical response of the SSFLC from over-damping to under-damping. We
also noted that doping with 1% nc-ZnO significantly increases DLS signal, indicating that
the doping may reduce the total number of fluctuating domains in the optically illuminated
region.

Researchers had noted that stochastic resonances (Gammaitoni et al. (1998)) can occur in a
system with a proper combination of nonlinearity and stochastic excitation. The stochastic
resonances had been exploited to minimize the ill effects of noise on the system. SSFLC is
essentially a nonlinear system. FLC molecules in a SSFLC device typically encounter two
driving forces: one from a deterministic control field and the other a stochastic thermal
excitation force. One may be curious about how the FLC molecules behave under the
influences of these two forces. To answer the question, we investigated the DLS of SSFLC
by using a sinusoidal bipolar waveform of varying frequencies with Vpp =1 V. The measured
DLS curves are presented in Fig. 9. Note that without the interruption of the thermal
excitation the correlated DLS response of FLC molecules to the periodic waveform shall
persist forever. Instead, we found that the envelopes of the DLS curves decrease as delay time
Seasoning ferroelectric liquid crystal with
colloidal nanoparticles for enhancing application flavors 301

(a) (b)

Fig. 8. DLS data of SSFLC. (a) The measured autocorrelation curves of DLS from pure SSFLC
at various temperatures (the curves with black, blue, red, and purple colors correspond to
DLS at 65o C, 55o C, 45o C, and 35o C, respectively. (b) The DLS autocorrelation curve of SSFLC
at a doping level of 1% meaured at 45o C.

increases, revealing clearly the randomizing effect of the fluctuating thermal force. Based on
the rotational viscosity and spontaneous polarization of the FLC material, we expect that the
FLC molecules shall not respond to an applied field with a frequency of 10 kHz or above.
At such high frequencies, the FLC molecules remain essentially at the original orientation
with a small random excursion caused by the thermal excitation force. Indeed, the DLS curve
driven at 10 kHz (see the green-colored line in Fig. 9(a)) exhibits a flat response. However, it
is surprising to find that after doping with nc-ZnO (see Fig. 9(b)), the FLC molecules become
responsive even at 10 kHz.

3.4 The Model of Dynamic Light Scattering from SSFLC

To retrieve the underlying physics, we developed a stochastic model to describe the dynamic
light scattering behavior of SSFLC. The model is comprised of two parts: the first part de-
scribes the dynamic light scattering process from a medium with a fluctuating index of refrac-
tion, and the second part focuses on the description of the fluctuating process of the medium.
In this section, we will focus our attention on the second part of the model by reducing SSFLC
to an effective medium with a moment of inertia per unit volume ρ and a rotational viscosity
coefficient η. Furthermore, we assumed the SSFLC film to be well below the transition tem-
perature of smectic C* phase. Thus, we can focus on the fluctuating behavior of the azimuthal
angle ϕ of the SSFLC director. Thus, the equation of motion of SSFLC with thermal force FR (t)
becomes (Chandrasekhar (1992))

∂2 ϕ ( t ) ∂ϕ(t) δE f ree
ρ 2
+η = + FR (t). (4)
∂t ∂t δϕ
We further simplified the description of elastic deformation energy of SSFLC with single elas-
tic constant K. The total free energy per unit area can then be expressed as

1 ∂ϕ
Efree = d [ Ksin2 θ ( )2 − Ps E cos ϕ]. (5)
2 ∂x
302 Nanocrystals

(a) (b)

Fig. 9. Forced DLS of SSFLC. (a) The measured autocorrelation curves of DLS from pure SS-
FLC at various driving frequencies. (b) The DLS autocorrelation curves of SSFLC at a doping
level of 0.5% meaured at 45o C.

Here d denotes the thickness of the SSFLC layer. The equation of motion in absence of external
electric field finally reads

∂2 ϕ ( t ) ∂ϕ(t) ∂2 ϕ ( t )
ρ +η + Ksin2 θ = FR (t). (6)
∂t2 ∂t ∂x2
We implemented a strong anchoring boundary condition on ϕ by using ϕ(0, t) = 0, ϕ(d, t) =
0. Assuming the rubbing direction of the SSFLC to be along the z-axis and by applying the
technique of the separation of variables, we can expand ϕ as a series of sinusoidal functions
∞ 
as ϕ( x, t) = ∑ ψ(t) sin(nπx d). We exploited a forward scattering geometry with small
n =1
scattering angle to ensure only the fundamental mode of thermal excitation to be detected.
This can greatly simplify Eq. (6) to yield

∂2 ψ (t) ∂ψ(t)
ρ +η + K̃ψ(t) = FR (t). (7)
∂t2 ∂t

with K̃ = Ksin2 θ (π d)2 . This is essentially a damped harmonic oscillator driven by a stochas-
tic thermal force FR (t). The thermal force is supposed to vary extremely rapidly over the
time of any observation, causing the excitation effect to be summarized by its first and second
moments

 FR (t) = 0,  FR (t) FR (0) = 2 η k B T δ(t). (8)

Here k B is the Boltzmann constant and T is the temperature. The bracket notation ·· denotes
an average with respect to the distribution of the realizations of the stochastic variable FR (t).
The important result of the second moment is well known as the Fluctuation-Dissipation
Theorem (FDT) (Kubo (1966)), which relates the strength of the fluctuating force to the
magnitude of the dissipation. It expresses the balance between dissipation which tends to
drive any system to a completely dead state and fluctuation force which tends to keep the
system alive. This balance is required to have a thermal equilibrium state at long times.

By applying Laplace transform on Eq. (7), we can derive an analytical solution of Eq. (7) as
Seasoning ferroelectric liquid crystal with
colloidal nanoparticles for enhancing application flavors 303

s1 t
  st st
s 2 e s2 t
ψ ( t ) = ψ (0) s1 e s1 − e 1 −e 2
η
−s2 + ψt (0) + ρ ψ (0) s1 − s2 +
t . (9)
1 s ( t − ) s ( t − )
ρ ( s1 − s2 ) 0
dτFR (τ ) e 1 τ −e 2 τ

  
∂ψ  K̃ 2 η 2 2 η 2
and s1 = − 2ρ + i ( ρ ) − ( 2ρ ) , s2 = − 2ρ − i ( K̃ρ ) − ( 2ρ ) .
η η
where ψt (0) = ∂t 
t =0
By using Eq. (8), we can further calculate the autocorrelation of dynamic light scattering in-
tensity as

G (t) − 1 =  I (t) I (0)  I (0) I (0) − 1
 s1 t
2
s 2 e s2 t s t
e s2 t
∝ N1 { ψ(0) s1 e s1 − + [ψt (0) + ρ ψ(0)] e s11 −
η
− s − s + . (10)
 2s t 2 2 
2 η kB T e 1 −1 e2s2 t −1 2 ( s + s ) t
2 2 2s1 + 2s2 + s1 + s2 ( 1 − e 1 2 ) }
ρ ( s1 − s2 )
Eq. (9) indicates
 that although the typical mechanical
 response
 of SSFLC is over damping
with K̃ ρ< η (2ρ), it can become under damping (K̃ ρ> η (2ρ)) if the elastic constant can
be increased to conquer the viscous effect.

For an over-damped SSFLC, the acceleration  is quickly damped by dragging force, causing
the rotational kinetic energy term ρ∂2 ψ(t) ∂t2 to be negligible in Eq. (4). In this case, the
equation of motion becomes
∂ψ(t)
η + K̃ψ(t) = FR (t). (11)
∂t
The autocorrelation of dynamic light scattering intensity is then simplified to be

ψ2 (0) 2  k T 
G (t) − 1 ∝ {ψ (0)e−2 tK̃ η − B (e−2 tK̃ η − 1)}. (12)
N K̃
The autocorrelation at the short delay time limit can be found to approach a value of

1 1 k T
|ψ(0)|4 = ( B )2 .
G (0) − 1 ∝ (13)
N N K̃
In a liquid crystal medium, the effective elastic constant is usually proportional to the square
of order parameter S as K̃ = K̃0 S2 ( T ), which decreases as temperature is increased. This leads
to a higher short-time-limit of the intensity autocorrelation as temperature is increased. At
  2
long delay time, the intensity autocorrelation decays to G (∞) − 1 = N1 k B T K̃ , implying a
higher background level at higher temperature. These asymptotic behaviors explain well the
experimental observation shown in Fig. 8(a).
Both the elastic constant and rotational viscosity of the FLC decrease as temperature is
increased. Eq. (12) predicts the decay time of the intensity autocorrelation function to be 0.3
sec at 35 o C, 0.2 sec at 45 o C, 0.15 sec at 55 o C, and 0.1 sec at 65 o C, respectively. By inserting
the material parameters into Eq. (11), we can solve the stochastic differential equation
numerically at the four temperatures. We can use the solution to calculate the intensity
autocorrelation function. The calculated DLS curves are presented in Fig. 10(a), which fit to
the experimental data (shown as symbols) quite well. By using a larger elastic constant and
solving Eq. (7) numerically with a thermal force corresponding to a temperature of 45o C, we
can calculate an intensity autocorrelation curve. The result is presented in Fig. 10(b), which
successfully reproduces the oscillating feature of the measured DLS curve shown in Fig. 8(b).
304 Nanocrystals

(a) (b)

Fig. 10. Simulated DLS curves of SSFLC. (a) The autocorrelation functions at a series temper-
ature. The symbols represent the experimental data and the lines are the simulation results.
The related parameters used in simulation are listed in table 1. (b) A simulation result for
underdamped dynamics. Note that the oscillation behavior begins at delay time 0.02 seconds,
which equals to the experiment result of SSFLC doped with 1% ZnO at 45o C.

In response to an external electric field applied along the x-direction, the equation of motion
becomes

∂ϕ(t) ∂2 ϕ ( t )
η = −Ksin2 θ + Ps E sin ϕ + FR (t). (14)
∂t ∂x2
Assuming the SSFLC to have a bookshelf geometry with uniform profile along the x-axis, the
first term in the right hand side of the equation can be neglected, which reduces the equation
to

dϕ
η = Ps E sin ϕ + FR (t). (15)
dt

Here Ps η plays a major role in determining the response time of FLC driven by an applied
electric field E. In this case, we encounter two driving forces with Ps E sin ϕ being determinis-
tic, and the other FR (t) having a stochastic nature. It is interesting to investigate the dynamics
with simulation and compare it with the experimental results. The parameters used in the
simulation include the spontaneous polarization Ps = 47 nC/cm2 , cell thickness d  = 2 µm,
and rotational viscosity η = 105 mPa · s. The  magnitude of random force strength 2η k B T
was adjusted to meet the condition of 2k B T η  0.5, which corresponds to a temperature of
about 25 o C. As shown in Fig. 11(a) for a undoped SSFLC, the simulation curve (the solid
line) calculated with Eq. (15) can successfully reproduce the measured DLS curve (the open
circles) measured at 100 Hz. By increasing the driving frequency to 10 kHz, the simulation
predicts a flat DLS response, which agrees well with the experimental results (open circles)
presented in Fig. 11(b)). The surprising result of Fig. 9(b), showing that the FLC molecules
become responsive
 at 10 kHz after doping with nc-ZnO, can be attributed to a doping-induced
increase in Ps η. Currently we have sampled only very limited parameter space. We believe
many interesting phenomena remain to be explored in SSFLC with or without nc-ZnO doping.
Seasoning ferroelectric liquid crystal with
colloidal nanoparticles for enhancing application flavors 305

(a) (b)

Fig. 11. Simulated curves of forced DLS from SSFLC. The comparison of experimental data
(symbols) and simulated curve (solid line) of DLS at 25o C by applying a sinusoidal waveform
with Vpp =1V and (a) f =100 Hz, (b) f =10 kHz on a SSFLC cell.

4. Importance of the Fundamental Mode of Molecular Orientation Fluctuation in

Liquid Crystal
Why LC molecules under thermal excitation fluctuate collectively in orientation and how
important to investigate the forward DLS from the resulting orientational fluctuation with
small scattering angle? The answers to these questions may lie in the fundamental concept of
symmetry and broken symmetry. In Fig. 12, we present a schematic diagram to summarize
the essential concepts to be discussed.

Typically, a crystal belongs to one of the 230 space groups, which have both the positional
and translational order of constituent atoms or ions. The most important behavior of a
condensed-phase material is a collective excitation of its constituent elements. One type of
the collective excitations in a crystal is phonon, in which each constituent nuclei oscillates
coherently with its neighbors at the same frequency. The oscillatory excitations in a crystal
are under-damped, resulting from the positionally ordered configurations that reduce the
chance of lattice nuclei to bump into each other and enable the lattice to oscillate with a price
of relatively low energy.

For materials with lower symmetry, they can be either positionally or orientationally
disordered. Liquid crystals are positionally disordered, but have an orientational order.
The orientational order reflects the degree of the molecular tendency to align to a specific
direction. In LCs, the analogy to phonon in a crystal is orientational fluctuations of LC
molecules. The collective excitations in LCs are typically over-damped, originating from the
positional disorder of LC constituents. This leads to that inertial forces are much smaller than
the viscous force. FDT describes the relation between the molecular fluctuations and their
energy dissipations in equilibrium.

The dispersion relations for phonons in a crystal and orientational fluctuations in LC offer
further insight into the underlying physics. A zero-frequency mode can exist in both material
systems. The long-wavelength limit of acoustic phonon modes in a crystal always renders
306 Nanocrystals

Fig. 12. Schematic showing the comparison of positionalal order in a crystal and orientational
order in a nematic liquid crystal. The two figures at the bottom are the dispersion relations

ω (q), which is the oscillating frequency of the phonons with q denoting the wavevector of
 
the collective excitation. n0 is the average orientation of the LC director, and δ n denotes a
fluctuation of the director with a relaxation time of τ.

into the zero-frequency mode, whereas the zero-frequency mode of LC originates from the
long-wavelength limit of orientational fluctuation. However, the q-dependent behaviors are
quite different. The acoustic phonon of a crystal exhibits a linear dispersion of ω (q) ∝ q. The
dispersion relation in LC is parabolic τ −1 ∝ q2 , a characteristic property of an over-damping
system.

The existence of the zero-frequency mode in fact originates from the spontaneous breaking
of the symmetry in a system. In the context of Goldstone’s theorem (Strocchi (2008)), which
states if the ground state of a many-body system has a broken symmetry, a gapless branch
of collective excitations must exist in the system to restore the lost symmetry. It is this
zero-frequency mode that makes all the neighboring LC molecules to change the orientations
coherently when external perturbations such as thermal forces are exerted on them. Forward
Seasoning ferroelectric liquid crystal with
colloidal nanoparticles for enhancing application flavors 307

DLS with small scattering angle offers us a unique opportunity to probe deeply into the
zero-frequency mode of the orientational fluctuation.

5. Conclusion
In this chapter, we reviewed our recent advances on the ordered structure and dynamic
responses of ferroelectric liquid crystal to an external deterministic force and stochastic ther-
mal excitation. We studied how the ordered structure and dynamic responses were tailored
by doping the FLC with ZnO nanoparticles. Combined capacitance-voltage measurement,
two-dimensional infrared absorption correlation technique, and dynamic light scattering
were shown to be valuable to yield clear picture for the complex system. Specifically,
the CV characterization method revealed that doping FLC with nc-ZnO can improve the
electro-optical response by increasing the spontaneous polarization and reducing the coercive
voltage of the beneficial dipolar species while minimizing the polarization of the species with
counter effects. Our 2D IR correlation data showed that at the submolecular level the ZnO
nanocrystals were dispersed uniformly into the SSFLC medium. This uniform dispersion
yields stronger correlations among the IR-active modes. The resulting stronger correlations
lead to a more concerted reorientation dynamics. Dipolar interaction of a ZnO nanodot with
surrounding C=O groups of FLC molecules was proposed to produce a doping-induced
binding effect with an energy stability of about 1000 J/m3. The doping-induced stability
also revealed in the dynamic light scattering study. Fluctuations in the optical response of a
medium have traditionally been regarded as noise. We found that fluctuations in DLS can be
used as a source of information for the complex system. By comparing the theoretical and
experimental results of DLS, we found that an increase in elastic constant and spontaneous
polarization by nc-ZnO doping can convert the mechanical response of the SSFLC from
over-damping to under-damping. We offered a general discussion on the collective excitation
of molecular orientation in SSFLC based on the Goldstone’s theorem and argued that the
zero-frequency mode may exist in the system to restore the lost symmetry. Based on the
theoretical and experimental studies, we concluded that nc-ZnO doping into FLC can become
a simple yet effective way to tailor the field-induced switching properties of an existing FLC.

This research was supported by the National Science Council of the Republic of China under
grant NSC 97-2112-M-009-006-MY3. The author also likes to thank his graduate students
Liu S. Li, Meng M. Cheng, and Bin K. Chen for their contribution and devotion to this research.

oC

η(mPas.s) θ (o ) Ps (nC/cm2) K̃ (N/m2) τ = (η 2K̃ ) (s)
35 105 14 47 0.17 0.31
45 60 13 39 0.15 0.20
55 35 12 32 0.12 0.15
65 20 10 25 0.10 0.10
Table 1. The Related Parameters of FELIX 017/100 FLC used for the simulation
308 Nanocrystals

6. References
Berne, B. J. & Pecora, R. (2000). Dynamic Light Scattering with Application to Chemistry, Biology
and Physics, Dover Publications, Mineola and New York.
Blinov, L. M. & Chigrinov, V. G. (1993). Electrooptic Effects in Liquid Crystal Materials, Springer-
Verlag, ISBN-10: 0387947086, Berlin.
Chandrasekhar, S. (1992). Liquid Crystal, Cambridge University Press, Cambridge,United
Kingdom.
Cheon, C. I.; Li, L.; Glushchenko, A.; West, J. L.; Reznikov, Y.; Kim, J. S. & Kim, D. H. (2005).
Electro-Optics of Liquid Crystals Doped with Ferroelectric Nano-Powder, Society for
Information Display Digest of Technical Papers, Vol. 36, pp. 1471-1473
Clark, N. A. & Lagerwall, S. T. (1980). Submicrosecond bistable electro-optic switching in liq-
uid crystals. Appl. Phys. Lett., Vol. 36, No. , page numbers (899-901).
Demus, D.; Goodby, J.; Gray, G. W.; Spiess, H.-W.; Vill, V. & Lagerwall, S. T. (1998). Chiral
Smectic Liquid Crystals: Ferroelectric Liquid Crystals, In: Handbook of Liquid Crystals
Vol 2B, Demus, D.; Goodby, J.; Gray, G. W.; Spiess, H. W. & Vill, V. (Eds.), page
numbers (515-664), WILEY-VCH Verlag GmbH, New York.
Gammaitoni, L.; Hanggi, P.; Jung, P. & Marchesoni, F. (1998). Stochastic Resonance. Reviews of
Modern Physics, Vol. 70, No. 1, page numbers (223-287).
Goyal, D. J.; Agashe, C.; Takwale, M. G.; Bhide, V. G.; Mahamuni, S. & Kulkarni, S. K. (1993).
Stochastic Resonance. J. Mater. Res. , Vol. 8, No. , page numbers (1052-1054).
Hegmann, T; Qi, H. & Marx, V. M.(2007). Nanoparticles in Liquid Crystals: Synthesis, Self-
Assembly, Defect Formation and Potential Applications. Journal of Inorganic and
Organometallic Polymers and Materials, Vol. 17, No. 3, page numbers (483-508).
Huang, J. Y. & Shih, W. T. (2006). Probing field-induced submolecular motions in a ferroelectric
liquid crystal mixture with time-resolved two-dimensional infrared spectroscopy. J.
Phys. C , Vol. 18, No. 32, page numbers (7593-7603).
Huang, J. Y.; Li, L. S. & Chen, M. C. (2008). Probing Molecular Binding Effect from Zinc Ox-
ide Nanocrystal Doping in Surface-Stabilized Ferroelectric Liquid Crystal with Two-
Dimensional Infrared Correlation Technique. J. Phys. Chem. C, Vol. 112, No. 14, page
numbers (5410-5415).
Hung, F. R. & Bale, S. (2009). Faceted nanoparticles in a nematic liquid crystal: defect struc-
tures and potentials of mean force. Molecular Simulation, Vol. 35, No. 10-11, page num-
bers (822aV834).
˛
Kaur, S.; Singh, S. P.; Biradar, A. M.; Choudhary, A. & Sreenivas, K. (2007). Enhanced electro-
optical properties in gold nanoparticles doped ferroelectric liquid crystals. Appl. Phys.
Lett., Vol. 91, No. 2, page numbers (023120).
Keesom, W. H. (1921). Phys. Z., Vol. 22, No. , page numbers (129).
Kobayashi, S. (2009). Nanoparticle-Embedded Liquid Crystal Display Devices. EKISHO, Vol.
13, No. , page numbers (241-249).
Kubo, R. (1966). The fluctuation-dissipation theorem. Rep. Prog. Phys., Vol. 29, No. , page num-
bers (255-).
Lagerwall, S. T. (1999). Ferroelectric and Antiferroelectric Liquid Crystals, Wiley-VCH, Weinheim.
Lapointe, C. P.; Mason, T. G. & Smalyukh, I. I. (2009). Shape-Controlled Colloidal Interactions
in Nematic Liquid Crystals. Science, Vol. 326, No. , page numbers (1083-1086).
Li, F.; Buchnev, O.; Cheon, C. I.; Glushchenko, A.; Reshetnyak, V.; Reznikov, Y.; Sluckin, T. J.
& West, J. L. (2006). Orientational Coupling Amplification in Ferroelectric Nematic
Colloids. Phys. Rev. Lett., Vol. 97, No. 14, page numbers (147801-4).
Seasoning ferroelectric liquid crystal with
colloidal nanoparticles for enhancing application flavors 309

Li, L. S. & Huang, J. Y. (2009). Tailoring switching properties of dipolar species in ferroelec-
tric liquid crystal with ZnO nanoparticles. J. Phys. D: Appl. Phys., Vol. 42, No. , page
numbers (125413-125417).
Magnasco, V.; Battezzati, M.; Rapallo, A. & Costa, C. (2006). Keesom coefficients in gases.
Chem. Phys. Lett. , Vol. 428, No.4-6, page numbers (231-235).
Mallia, A.; Vemula, P. K.; John, G.; Kumar, A.; Pulickel, M. & Ajayan, P. (2007). In Situ Synthesis
and Assembly of Gold Nanoparticles Embedded in Glass-Forming Liquid Crystals.
Angewandte Chemie International Edition, Vol. 46, No. , page numbers (3269-3274).
Meulenkamp, E. A. (1998). Synthesis and Growth of ZnO Nanoparticles. J. Phys. Chem. B, Vol.
102, No. , page numbers (5566-5572).
Morita, S. (2004). 2Dshige. Kwansei-Gakuin University, http://scitech.ksc.kwansei.
ac.jp/_ozaki/.
Nann, T. & Schneider, J. (2004). . Chem. Phys. Lett., Vol. 384, No. , page numbers (150-).
Noda, I. (1993). Two-Dimensional Correlation Method Applicable to Infrared, Raman and
Other Types of Spectroscopy. Appl. Spectrosc., Vol. 47, No. , page numbers (1329-1336).
Ouskova, E.; Buchnev, O.; Reshetnyak, V.; Reznikov, Y. & Kresse, H. (2003). Dielectric re-
laxation spectroscopy of a nematic liquid crystal doped with ferroelectric Sn2P2S6
nanoparticles. Liquid Crystals, Vol. 30, No. 10, page numbers (1235-1239).
Ozaki, Y. & Noda, I. (2000). Two-Dimensional Correlation Spectroscopy, AIP Conf. Proc. Vol 508,
AIP, Melville, New York.
Reznikov, Y.; Buchnev, O.; Tereshchenko, O.; Reshetnyak, V.; Glushchenko, A. & West, J.
(2003). Appl. Phys. Lett., Vol. 82, No. , page numbers (1917-).
Shim, M. & Guyot-Sionnest, P. (1999). Permanent Dipole Moment and Charges in Colloidal
Quantum Dots. J. Chem. Phys., Vol. 111, No. , page numbers (6955-6964).
Shiraishi, Y., Toshima, N., Maeda, K., Yoshikawa, H., Xu, J. & Kobayashi, S. (2002). . Appl. Phys.
Lett., Vol. 81, No. , page numbers (2845-2847).
Strocchi, F. (2008). Breaking of Continuous Symmetries. Goldstone’s Theorem, Symmetry Breaking,
Lecture Notes in Physics Vol. 643, Springer, New York.
310 Nanocrystals
Organic nanocrystals for nanomedicine and biophotonics 311

14
X

Organic nanocrystals for

nanomedicine and biophotonics
Koichi Baba1*, Hitoshi Kasai2,3, Kohji Nishida1 and Hachiro Nakanishi2
1. Osaka University, 2. Tohoku University, 3. PRESTO, Japan Science and Technology
Japan

1. Introduction
In this chapter, we will describe the advanced methodology for opening the new gate of
nanomedicine and bioimaging using organic nanocrystals formulation, which is free from
organic solvent and nano carrier. The organic nanocrystals aqueous dispersion is prepared
by our originally developed technique, the reprecipitation method.
The importance of the contribution of nanotechnology to biomedical application has been
greatly increased in few decades. Especially the current scientific endeavours on the
researches of nanotechnology based nanomedicine and bioimaging have been giving us the
novel and important information of biological events in vitro and in vivo. When focusing
bioimaging and nanomedicine, we definitely treat the fluorescent dyes and drugs. However,
currently used numerous organic compounds including drugs and fluorescent dye probes
are hydrophobic in nature. Therefore, for their using in drug administration to human and
animals or in dye staining imaging in living cells in vitro and in vivo, organic solvents are
widely used for their increasing water solubility. However, organic solvent has not only the
problems of causing systemic toxicity and cytotoxic activity, but also disturb obtaining
accurately the biological experimental data, which may be modified by the influence of
organic solvent. Many researchers have been making the scientific endeavour to clear these
problems. Using nano carrier such as liposome, nanocapsule, nanosphere, dendrimer,
emulsion, and surfactant have concluded successfully in some part of increasing water
solubility of drugs and dyes. However, these carries still suffer from systemic toxicity and
cytotoxic activity caused by their own. Additionally, the nano carriers have disadvantage of
low amount loading of drugs and dyes in carrier and of their leakage.
To overcome these problems, recently the organic solvent and carrier free method for
photodynamic cancer therapy and for bioimaging of living cells in vitro using organic
nanocrystals are reported including our group (Baba et al., 2007 and 2009). Organic
nanocrystals were prepared by the reprecipitation method (details are described in the
flowing section 2). Employing these anticancer drug and fluorescent dye nanocrystals are
superior in anticancer function and excellent fluorescent cell imaging, which are at least
comparable with conventional manner. The crystals are most densely packed structure of
molecules and nanocrystals are too. Therefore, the amounts of molecules in nanocrystals,
which consist of 100% of molecules, are higher than that of the same size of nano carriers,
312 Nanocrystals

and nanocrystals are free from leakage problems. The details of using organic nanocrystals
for nanomedicine and bioimaging are described in the section 3. Future perspective of
organic nanocrystals in nanomedicine and bioimaging are described in the section 4. Until
now, to our knowledge, the scientific researches dealing with pure organic nanocrystals for
nanomedicine and bioimaging are quite less. In this chapter, we will introduce recent
researches of organic nanocrystals in biomedical applications.

2. The reprecipitation method, which is the preparation

method of organic nanocrystals
The reprecipitation method is bottom-up type of preparing technique of organic
nanocrystals (Kasai et al., 1992), which is the solvent exchange method from good solvent to
poor solvent. For example of the preparation, first the targeting hydrophobic compound for
nanocrystallization is dissolved in good solvent acetone (in mM order), then the acetone
solution (200 μl) is speedy injected into vigorously stirred poor solvent water (10 ml) using
microsyringe. Then the sudden exchanges of solubility from good to poor cause the
precipitation, and when the concentration, temperature, and kinds of solvent are adequately
selected, nanocrystals are stably dispersed in water, with usually giving the negative
charged -potential on the surface of nanocrystals. The -potential helps nanocrystals for
preventing their aggregation because of the created balanced stability by negative charged
repulsion force of them. The schematic images of the reprecipitation method are shown in
Figure 1. The reprecipitation method is excellent for precise controls of their size and
morphology compared with that of the other conventional top-down procedure such as
milling method. The typical size control range is from several tens nanometres to several
micrometers (Fig. 2). The imaging pictures of water dispersion of nanocrystals are shown in
Figure 3. The transparent and solution-like appearances are because of the quite low light

Fig. 1. The scheme of the reprecipitation method.

Organic nanocrystals for nanomedicine and biophotonics 313

scattering originated in nano-sized crystals. These dispersions have both liquid and crystal
characteristics. Therefore, we can directly measure the physiochemical properties of crystals
as the liquid dispersions, or we can fabricate the nanocrystals-layered structure by collecting
these nanocrystals from dispersion (Masuhara et al, 2001). Until now, our scientific
endeavour revealed the size-controlled and size-dependent optoelectrical properties of
functional organic nanocrystals such as nonlinear optical materials (Nakanishi & Katagi,
1998), π-conjugate functional dyes (Kasai et al., 1996), organic pigment for colour filter of
liquid crystal display (Miyashita et al., 2008), and fullerene (Masuhara et al., 2009). Recently,
numerous researches focusing on the optoelectrical property of organic nanocrystals have
been reported by several groups (Zhao et al., 2008, An et al., 2004). However, the
applications of organic nanocrystals on biomedical researchers are quite less (Sin et al., 2006,
Bwambok et al., 2009), instead of their radical increasing importance. Therefore, recent our
interest and challenging topics are applying organic nanocrystals technology toward the
applications of drugs for medical treatments and fluorescent probe for bioimaging.

Fig. 2. Size control of organic nanocrystals in same compound.

Fig. 3. Imaging pictures of organic nanocrystals water dispersion.

314 Nanocrystals

3. The recent achievements of organic nanocrystals

in nanomedicine and biophotonics
3.1 Organic nanocrystals used for nanomedicine
Recently, as with developing the nanoparticles preparation technology, several reports are
mentioning about the fabrication technique of drug nanocrystals. However, almost all the
reports are concerning about the drug formulation in industrial interest (Kipp, 2004). As our
knowledge, quite few reports mentioning about the scientific aspects of using pure organic
nanocrystals forms of drugs. Prof. Prasad groups firstly demonstrated the using of
photosensitizing anticancer drugs in photodynamic cancer therapy (Baba et al., 2007). The
concept of this study based on how to increase the water solubility of hydrophobic drugs
without using solubilising agents such as surfactants and nanocarrier, which cause
cytotoxicity, and also to exhibit the high efficacy of the drug.
Not only the photosensitizing drugs, but currently used many of pharmaceutical agents
including various drugs are hydrophobic in nature (about 60% of drug products). Therefore,
special formulations are required to make their aqueous dispersion for delivery of these
drugs. Usually, surfactants or nanocarrier based delivery vehicles are used. In the cancer
therapy treatment, once systemic administration of drugs are carried out, these nanocarriers
including drugs are gradually taken up by tumour tissues according to the concept of the
“enhanced permeability and retention effect” (Maeda et al., 2006), which rely on the
characteristics of tumour tissues in trapping and retaining circulating nanoparticles because
of their leaky stricture. The nanocarriers are member of such as liposomes, polymeric
micelles, oil dispersions micelles, polymeric nanoparticles, drug encapsulated polymer
complexes. However, such carriers tend to increase the systemic toxicity caused by own.
Therefore, there is increasing interest in the developing the novel type of drug formulation
and delivery approaches without auditioning any solubilising agents such as surfactants
and/or nanocarriers. One novel method proposed for this was using the reprecipitation
method. Though the organic nanocrystals formations of hydrophobic compounds have been
well studied in optoelectrical materials, there was no report of using this method for drug
delivery.
To demonstrate the concept, the nanocrystal formulation of the hydrophobic drug of
photodynamic therapy was selected, and its efficacy was compared with the conventional
surfactant-supported formulation. Photodynamic therapy is a promising approach for
curing several types of cancers as well as some dermatological and ophthalmic diseases such
as age meditated macular degeneration. The main advantage of photodynamic therapy
compared with the conventional cancer chemotherapy is in the localized treatment using
selective light exposure to the tumour tissue. Usually, photodynamic therapy is carried out
by systemic administration of photosensitizing drugs, followed by exposing the light to the
tumour tissue. Basically, visible or near-infrared light are selected, and after the light is
exposed, the photosensitizing molecules transfers their excited state energy to oxygen
molecular in the surroundings, and as the result, reactive oxygen are formed, so-called
singlet oxygen. The singlet oxygen has ability to distract the structure of cells, and cell
deaths are induced in tumour tissue. Though the destruction process requires the
combination of photosensitizing, light, and oxygen, there is enough advantage for
photodynamic therapy in selective distraction of tumour tissues. However, one of the major
problems of photodynamic therapy is in the poor water solubility of photosensitizing drugs,
thus making their stable formulation for systemic administration are highly required. One of
Organic nanocrystals for nanomedicine and biophotonics 315

the overcoming these problems are using solubilising agent such as nanocarrier. However,
these nanocarriers often have the problem of drug leakage from carriers, resulting in fewer
efficacies. Furthermore, rejection reactions caused by nanocarriers are one of the anxiety
problems. Therefore, the ideal formulation for safe and for good efficacy photodynamic
therapy requires the solubilising agent free method. Thus, in this time, as new method for
the delivery of water insoluble drugs, which is free from using additional solubilising agents
was reported. This new drug formulation was fabricated by the reprecipitation method. The
resulting nanocrystals were monodispersed and taking stably dispersion, with diameter
about 100 nm.
In this research, there was an interesting finding that the fluorescence and photodynamic
activity of the drug nanocrystals were initially quenched in aqueous media because of the
peculiarity of crystal structure; however both recovered under in vitro and in vivo conditions
with the treatment of serum, resulting in creating molecular form. The recovery of
fluorescence and photodynamic activities were verified by confocal fluorescent microscopy
observation of cells and cellular phototoxicity assay. Efficacy of the nanocrystal formulation
in vitro as well as in vivo was found to be comparable with that of the same drug formulated
in the conventional surfactant delivery vehicle. This study was first example of the use of
organic nanocrystals prepared by the reprecipitation method.

3.2 Organic nanocrystals used for bioimaging

Fluorescence microscopy observation is one of the most broadly utilized imaging techniques
in biomedical researches, which allows the noninvasive imaging of cells and tissues with
molecular specificity. These imaging requires fluorescent dyes to enter cells and tissues
before visualization. Currently used dyes include classes of the coumarin, rhodamine,
fluorescein, and carbocyanine, as well as their derivatives. 3,3´-Dioctadecyloxacarbocyanine
perchlorate [DiO: DiOC18(3)] is the long-chain dialkylcarbocyanine dye, and commonly
used for visualizing the anterograde and retrograde neuronal tracers in living cells. This
lipophilic carbocyanine is also employed for many other applications, including the tracking
of cell migration and lipid diffusion in membranes through fluorescence recovery after
photobleaching (Gordon et al., 1995), the labeling of lipoproteins (Lohne et al., 1995), and
cytotoxicity assays (Johann et al., 1995). Nevertheless, the hydrophobic nature of these dyes
require, in many cases, organic solvents [e.g. dimethylsulfoxide (DMSO) and
dimethylformamide] and surfactants for successful cell imaging with increased water
solubility of dyes. However, unfortunately, organic solvents and surfactants themselves
tend to increase cytotoxicity in vitro and in vivo. On the other hand, the direct applications of
micronized crystals have also been investigated, although such crystals are probably not
small enough to allow the diffusion of dyes into cells, thus imaging efficency is poor.
Furthremore, not only DiO dyes, but also many of other fluorescent dyes are hydrophobic in
nature, including perylene, which is widely studied hydrophobic material as functioning
with high quantum yield in organic electroluminescence devices. Although using perylene
is potentially useful for bioimaging, the preparation of aqueous dispersions of perylene dyes
requires special formulation techniques, similar to those of DiO. One of the very useful
technique for dispersing hydrophobic compounds in water is the “reprecipitation method”,
which has been used for the nanocrystallization of organic optoelectronic materials
exhibiting size-dependent optical properties. Recently, by using this reprecpitation mehod,
we demonstrated an organic solvent free bioimaging method employing nanocrystals of
316 Nanocrystals

hydrophobic fluorescent dye, and their applications to in vitro were evaluated by confocal
fluorescent laser microscopy observation (Baba et al., 2009). The fluorescent dyes that we
used were DiO and perylene.
The aqueous dispersions of DiO and perylene prepared using the reprecipitation method
have been dialyzed for 6 h to remove acetone, thereby resulting in organic solvent–free
dispersions were obtained (Fig. 4). The resulting particle sizes and morphologies were less
than 150 nm for the DiO nanocrystals and 155 nm for the perylene nanocrystals, which was
revealed by scanning electron microscopy observation, and these values correspond
reasonably well with those determined through dynamic light scattering measurements
(Fig. 5). The average  -potentials of the DiO and perylene nanocrystals were ca. +36 and –10
mV, respectively. The relatively high positive surface charge of the DiO nanocrystals,
presumably derived from the positively charged molecular structure, imparted this water
dispersion with excellent stability. In the case of the perylene nanocrystals, the factors
causing their negative surface potential remains unclear, although negative charging of
organic nanocrystals in water is well known empirically to inhibit particle aggregation. The
aqueous dispersion of perylene nanocrystals was stable for more than 3 months. The
crystallinity of their prepared DiO and perylene nanocrystals were confirmed by powder X-
ray diffraction analysis.
The obtained UV–Vis absorption and fluorescence emission spectra of these nanocrystals are
significantly different from that of their solution form, namely we observed suppression and
broadening in the absorption spectra and decreases in emission intensities. The decreases in
fluorescence intensities are well-known aggregation effects for fluorescent molecules. In our
case, this phenomenon was due to the low solubility of the hydrophobic dyes, with
corresponding aggregation effects in aqueous system. A decreased fluorescent yield will
induce a decreased imaging efficiency.

Fig. 4. Absorption spectra of water dispersions of (a) DiO nanocrystals and (b) perylene
nanocrystals. (i) before and (ii) after dialysis. The disappearences of the absorbance of
acetone (at λ max = 265 nm) were confirmed.
Organic nanocrystals for nanomedicine and biophotonics 317

Fig. 5. Size distributions of (a) DiO and (b) perylene nanocrystals dispersed in water. The
peak sizes of the DiO particles were 22 nm (intensity: 58%) and 146 nm (intensity: 42%); the
peak size of the perylene particles was 155 nm (intensity: 100%; i.e., monodispersity).

However, there was interesting finding that the fluorescence of the DiO and perylene
nanocrystals were recovered by elapsed time under in vitro conditions in the presence of
10% fetal bovine serum. It seemed that this recovery in fluorescence in the presence of fetal
bovine serum induced the DiO and perylene nanocrystals to reasonably good soluble.
Although the exact mechanism for the improved solubility is not cleared yet, but we
consider that it may rely on the affinity interactions between the hydrophobic moieties of
the proteins and the hydrophobic surfaces of the nanocrystals, which cause the dissolution
of nanocrystals, resulting in converting nanocrystals into the molecular forms of the dyes,
thereby, leading to fluorescence recovery.
318 Nanocrystals

Fig. 6. In vitro fluorescent confocal images of cells recorded 1 h and 7 hrs after incubation
with (i), (ii) nanocrystals of (a) DiO and (b) perylene, and (iii), (iv) dyes solution in DMSO of
(a) DiO and (b) perylene.

To evalute if the organic nanocrystals were useful for the bioimaging of living cells in vitro,
we performed confocal fluorescence imaging of tumor cells using the DiO and perylene
Organic nanocrystals for nanomedicine and biophotonics 319

nanocrystal formulations (Fig. 6). Although the fluorescence signals in the nanocrystals were
initially weak, it increased after cellular uptake (1 h) as similar to the behavior of
nanocrystals in the fetal bovine serum-containing medium in bothe case of perylene
nanocrystals and DiO nanocrytals. Compared with the using of solution form of dyes
disolved in DMSO, where exhibiting higher fluorescence signals from at the initial stage of
dye dosing in culture medium, nanocrystals revealed comparable levels of cellular uptake
by elapsed time. Therefore, the long-term cellular uptake of the DiO and perylene dyes in
their nanocrystal formulations were similar to that of the dyes in their solution formulations.
The DiO dyes specifically stained the cell membrane; the perylene dyes stained the
cytoplasm.
On the other hand, we investigated whether nanocrystal formation was really necessary for
efficient cell imaging, especially remarking on the crystal size. Aqueous dispersions of
microcrystals, with the average sizes were in the micrometer range, were prepared through
sonication of the bulk crystals in water. The concentrations of these dispersions were
adjusted to be the same as those of the nanocrystals. Figure 7 displays scanning electron
microscopy images of the DiO and perylene microcrystals used for cell imaging. In both
cases, their sizes were on the order of several to a few tens of micrometers. In microcrystals,
no significant recovery in fluorescence occurred in the cell culture medium containing 10%
fetal bovin serum during 6 hrs after the addition of the DiO microcrystals, and whereas a
slight recovery of fluorescence in the case of the perylene microcrystals was observed.
Presumably, one of the main reasons for the significantly lower fluorescence recoveries of
the microcrystals was in their size. When the size of a crystal increases, its surface area per
unit volume decreases, as a result, its interactions with the serum components become weak,
and resulting in decreased solubility. Figure 8a presents the confocal microscopy images we
obtained when applying the DiO microcrystals. Interestingly, only the surrounding areas
where the microcrystals were attached to the cells

Fig. 7. Scanning electron microscopy images of (a) DiO and (b) perylene microcrystals.

displayed fluorescence (e.g., yellow circle in Figure 8a). The direct application of DiO
microcrystals to fluorescence imaging of cells is a commonly used technique. Nevertheless,
we observed no fluorescence from those cells not presenting any attached DiO microcrystals,
320 Nanocrystals

meaning that this imaging technique is poorly efficient. Figure 8b reveals that, in the case of
the added perylene microcrystals, the fluorescence imaging was less efficient than that
obtained using the corresponding nanocrystal formulation (Fig. 6b, ii) Clearly, the low
solubility of the perylene microcrystals in the serum led to poor imaging. Additionally, we
also imaged cells lacking added dyes as a control, and we observed no fluorescence signals
under the corresponding measurement conditions (Fig. 8c). Thus, the fluorescence images
we obtained using nanocrystals, dyes in DMSO treatments, and microcrystals were not the
result of autofluorescence of the cells.

Fig. 8. Fluorescent confocal images of (a) DiO microcrystals, (b) perylene microcrystals, and
(c) control: the pictures of fluorescent images (left), tranceperant images (middle), and
overlaped images of the left and middle (right). The yellow circle in (a: middle) represents
the microcrystals. These fluorescent images were obtained after 7 hrs incubation.
Organic nanocrystals for nanomedicine and biophotonics 321

Fig. 9. Double staining images of cells using DiO and perylene nanocrystals. The green and
blue fluorescent colours come from DiO and perylene, respectively.

When we investigated double staining imaging of both the DiO and perylene nanocrystals
in cells in culture medium we initially detected each fluorescence signal individually by
confocal fluorescent laser microscopy observation, and Figure 9 presents the over laped
pictures of them. DiO resulted in significant membrane staining (Fig. 9: green color),
whereas perylene stained in mostly cytoplasmic (Fig. 9: blue color). One of the reason for
this behavior might be due to differences in -potentials of nanocrystals/dyes, namely the
DiO particles had a positive -potential (+36 mV), which may have aided in their trapping
on the slightly negatively charged cell membrane. We also confirmed that the fluorescence
obtained were not the result of autofluorescence of the cells. Although the detail
mechanisms of endocytosis of the nanocrystals/dyes remained unclear, we were successful
in recording a double-staining cell images using the formulation of organic nanocrystals.

3.3 Application of the reprecipitation method toward dye doped polymeric nanoparticles
We will introuduce the other approach for fabricating bioimaging tools using the
reprecipitation mehtod, where the fluorescent dye doped biodegradable polymer
nanoparticles were demonstrated (Baba et al., 2005). The polymer used was poly (D,L-
lactide-co-glycolide), and the dyes used were infrared emitting dyes and two photon
excitable fluorescent dyes. Current technologies in fluorescent imaging have advanced
significantly in developing for multiprobe applications in the study of biological events.
Especially, for further advanced fluorescent imaging capability, organic fluorophores
absorbing and emitting in near infrared region have been developed. Such near infrared
322 Nanocrystals

dyes have the advantage for providing the better signals avoiding the autofluorescence
caused by ultra violet or visible light. Secondly, tissue penetration of near infrared lighg is
great due to low absorption of excitation light in tissue. However, many of these dyes have
less solubility and emissions in aqueous systems, making them undiserable for biological
applications. The infrared emitting dyes used as demonstration shows fluorescence around
1.1 to 1.35 μm in organic solvents, but was significantly quenched in aqueous media.
Whereas, two photon excitable fluorescent dye has a polar D-π-A structure (Lin et al., 2004),
in which the π-system is end-capped by an electron donor (D) and an electron acceptor (A),
having one of the most effective molecular models for both second- and third-order
nonlinear optical materials, thus such dyes have great potential for two photon imaging.
However, as with these dyes lacks the solubility properties for the application of biological
systems. The nanoparticle technology had the ability to encapsulate such hydrophobic dyes
into some matrix for producing the aqueous dispersion, enabling functionality in biological
systems. For example, the fabrication of dye-doped polymeric nanoparticle water
dispersions were quite effective method for bioimaging of such hydrophobic fluorescent
dyes. The biodegradable poly (D,L-lactide-co-glycolide) were useful for preparing particles,
which have been well investigated as a drug and gene delivery vehicle (Jain, 2000). The
applying of the reprecipitation method, which is simple and surfactant-free fabrication
method, for the incorporation of hydrophobic, near infrared emitting dyes and two photon
excitable emitting dyes into poly (D,L-lactide-co-glycolide) nanoparticles resulted in
creating an aqueous dispersion of nanoparticle, with their sizes of less than 100 nm. Then,
the optical properties of dye solution, dye nanocrystals water dispersion, and dye doped
polymeric nanoparticle water dispersion for these two dyes were investigated, respectively.
For infrared emitting dyes, the fluorescence from dye solution was in the spectral range of
~1.1 to 1.35 μm. Whereas, in the case of dye nanocrystal water dispersion, even at the same
concentration of dye solution, the emission was not noticeable. This was because of the
fluorescence quenching, estimated to be including collisional quenching, static quenching,
excited state reactions, electron transfer and energy transfer. The main reason for dye
fluorescence quenching in water dispersion might comes from the crystallization of dye, as
it is well known for several kinds of organic molecules. Similar changes in absorption were
usually associated with the appearance of the molecular H-aggregates, which were
nonradiative. However, when fabricating the dye doped poly (D,L-lactide-co-glycolide)
nanoparticle water dispersion, it exhibitted fluorescence in the aqueous environment. A
series of systematic studys of the relationship between fluorescence and micelle
microemulsion in water environment has been reported. Our results also indicated that dye
was successfully encapsulated in the polymer matrix. Thus, crystallization of dyes were
prevented, and also dyes were protected from contact with the water environment, thus
fluorescence quenching was avoided. In the case of two photon excitable emitting dyes,
optical properties were investigated especially forcusing on dye nanocrystals and dye
doped polymeric nanoparticles. The size of dye nanocrystals was 35 ± 9 nm, and the sizes of
these polymeric nanoparticles were controlled as ca. 36 ± 8 nm, 50 ± 11 nm and 96 ± 19 nm,
respectively, using the reprecipitation mehod. According to the procedure of the
reprecipitation mehod, when the initial concentration of polymer solution increased, the size
of polymeric particles also increased. There have been several reports on fabrication of poly
(D,L-lactide-co-glycolide) particles with sizes from several hundreds nanometer to few tens
micrometer. However, polymeric nanoparticles less than 100 nm in size were succesuffly
Organic nanocrystals for nanomedicine and biophotonics 323

prepared. Even though, two photon excitable fluorescent dye nanocrystals showed
fluorescence in aqueous media, the fluorescence intensity of dye-doped polymeric
nanoparticles was higher than that of the dye nanocrystals. Furthermore, we observed the
increasing fluorescence intensity of dye doped polymeric nanoparticles, with an increasing
in their particle size. The role of the polymer for influencing the fluorescence intensity
should be similar to the case of infrared emitting dyes. Namely, when two photon excitable
fluorescent dyes were surrounded by the polymer matrix, the crystallization of dyes and
contact of dyes with the water environment were prevented, resulting in increasing
fluorescent intensity. The reason for increasing the fluorescence intensity with an increasing
the polymeric particle size was that when the ratio of polymer to dye increased,
correspondingly, number of dyes individually embedded in the polymer matrix increased
without interaction between dyes. The uptake of two photon excitable fluorescent dye
nanocrystals and their dyes-doped polymeric nanoparticles by cancer cells in culture media
were investigated. Confocal laser microscopy of two-photon imaging revealed a difference
in the fluorescence intensity and staining pattern of fluorescence in cells incubated with dye
nanocrystals (34 ± 11 nm in size) and the dye-doped polymeric nanoparticle (32 ± 10 nm in
size), respectively. The fluorescence intensity obtained from cells incubated with dye-doped
polymeric nanoparticles were about ten times higher than that of dye nanocrystals. These
results suggested that the coating of fluorescence dye by adequate polymer was an
important methodology for efficient cell imaging. We can say that not only nanocrystals
formation, but alos dye-doped polymeric nanoparticles prepared by the reprecipitation
mehod are effective way for efficient bioimaging.

4. Future perspective of organic nanocrystals in nanomedicine

and biophotonics
4.1 Organic nanocrystals for biophotonics
Herein, we will report the new type anti-photo breaching fluorescent organic nanocrystals
used in bioimaging, which have the potential to overcome the problems of quantum dots
and molecular probe. Currently, using quantum dot and molecular probe for bioimaging
and nanomedicine are numerously reported. Quantum dots have been applied for detection
and imaging in several areas in the life sciences, ranging from microarray technologies to
fluorescence in situ hybridization to in vitro imaging (Medintz et al., 2005). Despite many
superior optical properties, such as size-tuneable absorption and emission, extremely broad
and intense absorption enabling a unique flexibility in excitation, high fluorescence
quantum yields even in the NIR wavelengths and large two-photon action cross-section as
compared to organic dyes, the solutions for using quantum dots have so far been individual
ones. Mainly, quantum dots need surface modification by complicated procedure to increase
the water solubility and to avoid the cytotoxic caused by leaking of heavy metal ions.
Additionally, we need patient to find a solution to the challenges of their particular
experimental system against the benefits of the advanced spectroscopic features of quantum
dots. On the other hand, the optical property of the molecular probe depend on the
electronic transitions involved and can be fine-tuned by elaborate design strategies if the
structure-property relationship is known for the given class of dye (Mason, 1999). The
emission of organic dyes typically originates either from an optical transition delocalized
over the whole chromospheres or form intramolecular charge transfer transitions. The
324 Nanocrystals

majority of common fluorophores, such as fluorescein, rhodamines, and most cyanines, are
resonant dyes that are characterized by slightly structured, comparatively narrow
absorption and emission bands that often mirror each other, a small solvent polarity-
insensitive Stokes shift, high molar absorption coefficients, and niderate-to-high
fluorescence quantum yields. The major problems of these dyes are both less water soluble
in nature, thus need organic synthesis making hydrophilic salts form or using solubilising
agents, and the weak resistance against photo breaching. We recently developed the new
type of fluorescent organic nanocrystals for bioimaging. These organic nanocrystals have
advantage in high water solubility, cytotoxicities-free caused by heavy metal ions, and quite
strong against photo breaching. The organic nanocrystals were class of organic pigments,
quinacridone deliverities. The organic pigments are very rigid compound, which have
strong light resistance and weather fastness, thus have been used for several applications
such as colour filter and coating materials. Therefore, making water dispersion of organic
pigments is quite difficult because of their poor water solubility. However, we found that if
applying the reprecipitation method, fine dispersion of quinacridone nanocrystals were
obtained (Ujiiye-Ishii et al., 2006). Under the confocal laser fluorescent microscopy
observation after dosing quinacridone nanocrytals in cell culture medium, we have been
successfully obtained fine and tough fluorescent imaging during nearly 15 min in laser
irradiation; despite the output power of laser source was nearly 100%. One can see that
almost no fluorescent breaching was observed in the imaging pictures at the beginning and
the ending of the observation (Fig. 10). Despite the merit of strong light resistance originated
from crystal structure, the fluorescent intensity of quanicridone nanocrystals were not
strong like such as fluorescein because of the quenching effect caused by crystallization.
Thus, we are now further investigating to find the new approach that satisfies both high
fluorescent intensity and light resistance. We believe that these kinds of right resistance
fluorescent organic nanocrystals are candidate for new type of bioimaging tools in near
future.

Fig. 10. Fluorescent confocal images of cells stained by quinacridone nanocrystals. The
fluorescent images observed at the (a) beginning stage and (b) ending stage.
Organic nanocrystals for nanomedicine and biophotonics 325

4.2 The size effect of nanocrystals for nanomedicine and biophotonics

The size is the important factor in biological events especially for cellular uptake, systemic
circulation in the case of systemic administration, and solubility in living organism, which
affect the efficacy of drugs and efficient fluorescent cell imaging. Actually, we found that
hydrophobic nanocrystals that had poor solubility in water had good solubility in serum
component, and which is peculiar to nano-sized crystals and not to more than micro-sized
crystals. These findings give us the very important information for considering the strategy
in nanomedicine and bioimaging for the selection of efficient and appropriate approaches.
For the advanced applications, one of the strategies we can select for using organic
nanocrystals in bioimaging and nanomedicine is taking the step of designing molecular
structure at first, which have the hydrophobic characteristics and the sight specific targeting
moiety. If the fluorescent dyes and drugs are hydrophobic, their stable organic nanocrystals
aqueous dispersions are well prepared by the reprecipitation method. Upon, taking
administration of these nanocrystals in vivo and in vitro, the nanocrystals will start
dissolving in living organism, with gradually generating the drugs/dyes activity by taking
molecular formulation. If the drugs and dyes that have site specific moiety, the drugs and
dyes can bind to the desired part such as affected site and cells with minimally dosing with
specifically. The dissolving kinetics, disposition, and body distribution of nanocrystals will
be basically aligned by controlling the size of particles. These favourable peculiarities will be
achieved by without applying any external solubilising agents such as organic solvents,
surfactants, and nanocarriers. This nanocrystal technology in the fields of biology should be
applied for not only limited scientific researches, but also for medical treatment in near
future. This simplest nanocrystal approach, which we have demonstrated, will be mostly
close way toward the clinical approach, which holds down the risk caused by carries and
the cost caused by manufacturing. We can say that the nanocrystals aqueous dispersion or
their dried powder will be useful for drugs/diagnostic fluorescent agents in such as
administrations of oral, intravascular, inhalational, transdermal, nasotracheal, and ocular.

5. Conclusion
In this chapter we described three topics; first, we explained how to prepare the organic
nanocrystals in aqueous dispersion system using the reprecipitation method. Second, we
referred the recent our achievements of organic nanocrystals in nanomedicine and
biophotonics. Third, we remarked the future direction of organic nanocrystals in
nanomedicine and biophotonics. We believe that our organic nanocrystals technology,
recent results, and ideas will be helpful especially for biochemist, biophysicist, nanoscientist,
and medical scientist for tremendous advances in their specialized fields and in their
multidisciplinary.

6. References
An, B. K. (2004). Strongly fluorescent organogel system comprising fibrillar self-assembly of a
trifluoromethyl-based cyanostilbene derivative. J. Am. Chem. Soc., 126, 10232-10233
Baba, K. (2005). Infrared emitting dye and/or two photon excitable fluorescent dye
encapsulated in biodegradable polymer nanoparticles for bioimaging. Materials
Research Society Symposium Proceedings, 845, 209-214
326 Nanocrystals

Baba, K. (2007). New Method for Delivering a Hydrophobic Drug for Photodynamic
Therapy Using Pure Nanocrystal Form of the Drug. Mol. Pharm., 4, 289-297
Baba, K. (2009). Organic solvent-free fluorescence confocal imaging of living cells using pure
nanocrystal forms of fluorescent dyes. Jpn. J. Appl. Phys., 48, 117002/1-117002/4
Bwambok, D. K. (2009). Near-infrared fluorescent nanoGUMBOS for biomedical imaging.
ACS Nano 3, 3854-3860
Gordon, G. W. (1995). Analysis of simulated and experimental fluorescence recovery after
photobleaching. Data for two diffusing components. Biophys. J., 68, 766-778
Jain, R. A. (2000). The manufacturing techniques of various drug loaded biodegradable
poly(lactide-co-glycolide) (PLGA) devices. Biomaterials, 21, 2475-2490
Johann, S. (1995). A versatile flow cytometry-based assay for the determination of short- and
long-term natural killer cell activity. J. Immunol. Methods, 185, 209-216
Kasai, H. (1992). A novel preparation method of organic microcrystals. Jpn. J. Appl. Phys., 31,
L1132-L1134
Kasai, H. (1996). Size-dependent colors and luminescences of organic microcrystals. Jpn. J.
Appl. Phys., 35, L221-L223
Kipp, J. E. (2004). The role of solid nanoparticle technology in the parenteral delivery of
poorly water-soluble drugs. Int. J. Pharm., 284, 109-122
Lin, T.-C. (2004). Degenerate nonlinear absorption and optical power limiting properties of
asymmetrically substituted stilbenoid chromophores. J. Mater. Chem., 14, 982-991
Lohne, K. (1995). Standardization of a flow cytometric method for measurement of low-
density lipoprotein receptor activity on blood mononuclear cells. Cytometry., 20,
290-295
Maeda, H. (2006). The EPR effect and polymeric drugs: A paradigm shift for cancer
chemotherapy in the 21st century. Adv. Polym. Sci., 193, 103−121
Mason, W. T. (1999). Fluorescent and luminescent probes for biological activity 2nd edn. Academic
press, London
Masuhara, A. (2001). Hetero-multilayered thin films made up of polydiacetylene
microcrystals and metal fine particles. J. Macromol. Sci. Pure Appl. Chem., A38, 1371-
1382
Masuhara, A., (2009). Fullerene fine crystals with unique shapes and controlled size. Jpn. J.
Appl. Phys., 48, 050206/1-050206/3
Medintz, I. L. (2005). Quantum dot bioconjugates for imaging, labelling and sensing. Nat.
Mater., 4, 435-446
Miyashita, Y. (2008). A new production process of organic pigment nanocrystals. Mol. Cryst.
Liq. Cryst., 492, 268-274
Nakanishi, H. & Katagi, H. (1998). Microcrystals of polydiacetylene derivatives and their
linear and nonlinear optical properties. Supramol. Sci., 5, 289-295
Sin, K. K. (2006). A highly sensitive fluorescent immunoassay based on avidin-labeled
nanocrystals. Anal. Bioanal. Chem., 384, 638-644
Ujiiye-Ishii, K. (2006). Mass-Production of Pigment Nanocrystals by the Reprecipitation
Method and their Encapsulation. Mol. Cryst. Liq. Cryst. 445, 177-183
Zhao, S. Y., (2008). Low-Dimensional Nanomaterials Based on Small Organic Molecules:
Preparation and Optoelectronic Properties. Adv. Mater., 20, 2859-2876