PHD Thesis

COUPLED PLASMONIC STRUCTURES
FOR SENSING, ENERGY AND

SPECTROSCOPY APPLICATIONS
A DISSERTATION SUBMITTED TO
THE GRADUATE SCHOOL OF ENGINEERING AND SCIENCE
OF BILKENT UNIVERSITY
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR
THE DEGREE OF
DOCTOR OF PHILOSOPHY
IN
MATERIALS SCIENCE AND NANOTECHNOLOGY PROGRAM
By
Sencer Ayas
August, 2015
I certify that I have read this thesis and that in my opinion it is fully adequate, in
scope and in quality, as a thesis of the degree of Doctor of Philosophy.
………………………………….
Assist. Prof. Dr. Aykutlu Dana (Advisor)
………………………………….
Assist. Prof. Dr. Necmi Biyikli
………………………………….
Assoc. Prof. Dr. Ozgur Oktel

………………………………….
Assoc. Prof. Dr. Hamza Kurt
………………………………….
Assist. Prof. Dr. Mehmet Solmaz
Approved for the Graduate School of Engineering and Science:
………………………………….
Prof. Dr. Levent Onural

Director of the Graduate School
ABSTRACT
COUPLED PLASMONIC STRUCTURES FOR SENSING, ENERGY

AND SPECTROSCOPY APPLICATIONS
Sencer Ayas
Ph.D. in Material Science and Nanotechnology Program
Supervisor: Assist. Prof. Dr. Aykutlu Dana
August, 2015
Recent advances in nanofabrication and characterization methods have enabled
the study of novel optical phenomena, thus boosting the research in nanophotonics
and plasmonics. Metal nanostructures offer a route for the excitation of surface
plasmons by confining the light in sub-wavelength dimensions, yielding extremely
high electromagnetic field intensities. Moreover, coupling different plasmon modes
offers a rich optical dispersion which cannot be obtained inherently by using single
plasmonic resonator. In this thesis, we first present a detailed study of simple coupled
plasmonic structures based on metal-insulator-metal structure. Then, we use similar
structures to devise novel optical platforms in various applications such as surface
enhanced Raman spectroscopy (SERS), surface enhanced infrared absorption
spectroscopy (SEIRA) and plasmon enhanced hot-electron devices.
The first part of this thesis concentrates on coupled plasmonic structures and their
spectroscopy and photodetector applications. Firstly, we study these structures
numerically and analytically and show surface enhanced Raman spectroscopy
(SERS) as a possible application with uniform signal intensities over large areas.
Then, fabricating these plasmonic surfaces with sub-10nm gaps over large areas lead
to development of single molecule Raman spectroscopy platforms. As an energy
related application, a contact free characterization method is developed to probe hot
electrons where similar coupled plasmonic surfaces are employed as hot electron
devices. Finally, using aluminum and its native aluminum oxide hierarchical
plasmonic surfaces are fabricated and its spectroscopy applications are demonstrated.
In the second part of, we develop interference-coating-based sensing platforms in
the visible and infrared wavelengths. Despite large field enhancements, plasmonic
iv
structures suffer from low signal intensities due to low mode volumes. To overcome
this issue we propose another strategy, namely using interference coatings with small
and uniform electric field enhancements over large mode volumes. These surfaces
outperform the conventional plasmonic surfaces when they are used as infrared
absorption spectroscopy platforms. Finally, similar surfaces are employed as
colorimetric sensor platforms to sense monolayer and bilayer proteins simply by
change in the surface color.
Keywords: Surface plasmons, coupled plasmons, spectroscopy, hot-electrons,

interference coatings, optical sensors
v
ÖZET
Eşlenmiş Plazmonik Yapıların Sensör, Enerji ve Spektroskopi

Uygulamaları
Sencer Ayas
Malzeme Bilimi ve Nanoteknoloji Programı, Doktora
Tez Yöneticisi: Assist. Prof. Dr. Aykutlu Dana
Ağustos, 2015
Nanofabrikasyon ve karakterizasyon metotlarındaki yeni gelişmeler yenilikçi

optik çalışmaları hızlandırmış, bu da nanofotonik ve plazmonik çalışmalarında ciddi
bir artışa sebep olmuştur. Metal nano-yapılar ışığı dalga boyundan çok küçük
boyutlara hapsederek yüzey plazmonlarının uyarılması sağlayarak elektromanyetik
alan şiddetini arttırmaktadır. Diğer yandan, farklı plazmon modlarının eşlenmesi tek
plazmonik rezonatör yapısıyla elde edilemeyecek çok zengin bir optik bant yapısı
sunmaktadır. Bu tezde, ilk olarak metal-dielektrik-metal rezonatör yapısı baz alınarak
eşlenik plazmon yapıları detaylı olarak çalışılmıştır. Sonra, benzer yapılar
kullanılarak yenilikçi optik platformlar üretilerek yüzey artırımlı Raman
spektroskopisi (SERS), yüzey artırımlı kızılötesi emilim spektroskopisi (SEIRA) ve
plazmon artırımlı sıcak elektron aygıt uygulamaları gösterilmiştir.
Tezin ilk kısmı plazmonik yapılar ve bu yapıların spektroskopi ve fotodetektör
uygulamalarına yoğunlaşmıştır. İlk olarak, bu yapıları analitik ve numerik olarak
çalışarak geniş alanlarda sabit artırımlı SERS uygulaması gösterilmiştir. Sonra,
benzer yapıların 10nm'den küçük aralıklı olarak üretilmesiyle tek molekül Raman
spektroskopisi uygulaması gösterilmiştir. Enerji uygulaması olarak da benzer eşlenik
plazmonik yapıları kullanılarak sıcak elektronların ölçüldüğü XPS tabanlı bir yöntem
geliştirilmiştir. Son olarak da aluminyum ve doğal aluminyum oksit tabakası
kullanılarak hiyerarşik plazmonik yapıları üretilerek spektroskopi uygulamaları
gösterilmiştir.
Tezin ikinci kısmında görünür ve kızılötesi dalga boylarında girişim kaplamasına
dayanan sensör platformları geliştirilmiştir. Plazmonik yapılar elektromanyetik alan
şiddetini çok fazla arttırmalarına rağmen, moleküllerin sinyal şiddetleri mod
vi
hacminin küçüklüğünden dolayı azdır. Bu problemi aşmak için yeni girişim
kaplamalara dayalı yeni bir strateji önerilmektedir. Bu yüzeyler kızılötesinde
kullanıldıklarında plazmonik yüzeylerden daha iyi performans göstermektedirler.
Son olarak da benzer yüzeyler görünür bölgede tek ve çift tabakalı proteinleri renk
değişimine bağlı olarak tespit etmek için önerilmektedir.
Anahtar kelimeler: Yüzey plazmonları, eşlenik plazmonlar, spektroskopi, sıcak

elektronlar, girişim kaplamaları, optik sensörler
vii
Acknowledgement
I would like to thank my PhD advisor Aykutlu Dâna for accepting me as a PhD
student in his group. This thesis would not be successful without his efforts, guidance
and support. I have learnt a lot from him; mathematical modeling, coding and tips
about clean room processes are few of them. I wouldn't ask for a better advisor.
I would like to thank Prof. Salim Çıracı for accepting me to PhD program.
I would like to thank my thesis committee; Assist. Prof. Necmi Bıyıklı, Assoc.
Prof. Mehmet Özgür Oktel, Assist. Prof. Mehmet Ertuğrul Solmaz and Assoc. Prof.
Hamza Kurt.
I would like to thank some of my present and past group members; Dr. Gökhan
Bakan, Dr. Kemal Çelebi, Hasan Güner, Andi Çupallari, Mustafa Ürel, Ahmet Emin
Topal and Serkan Karayalçın. I would especially thank Gökhan and Kemal for being
good friends and editing this thesis' drafts. I want to thank Andi for being good friend
and companion. He is one of the handy people I have known who made things
happen easily; especially clean room processes. I want to thank Mustafa for AFM
roughness measurements. I would like to thank Mehmet Kanık for FTIR
measurements, graphical illustrations and for being a good friend.
I want to thank Mehmet Ertuğrul Solmaz for partially supporting me for the last 8
months of my PhD.
I would like to thank clean room engineers and technicians; particularly Semih
Yaşar, Murat Dere and Mustafa Güler.
I greatly acknowledge TUBITAK through grant 111M344.
Finally, I thank my family for encouraging and supporting me to complete my
PhD work.
viii
Contents
ABSTRACT .......................................................................................................... iv
ÖZET ..................................................................................................................... vi
Acknowledgement .............................................................................................. viii
Contents ................................................................................................................. ix
List of Figures ...................................................................................................... xii
List of Tables ...................................................................................................... xxv
Chapter 1 Introduction ...................................................................................... 1
Chapter 2 Well-Defined Coupled Plasmonic Structures ................................ 7
2.1. Surface Plasmons at Metal-dielectric Interface ........................................... 8
2.2. Surface Plasmons in Metal-Insulator-Metal Resonators ........................... 11
2.3. Raman Enhancement on a Broadband Meta-surface................................. 13
2.3.1. Introduction ...................................................................................... 13
2.3.2. Description of the Plasmonic Meta-Surface ..................................... 14
2.3.3. Coupled Meta-Atoms and Meta-Molecules ..................................... 17
2.3.4. Experimental Results ........................................................................ 21
2.3.5. Discussion of Spatial Uniformity of Enhancement at the Nanoscale

24
2.4. Edge Rounding Induced Broadband Plasmonic Metamaterial Absorber

Structures ............................................................................................................... 27
2.4.1. Simulation Geometry and Method ................................................... 29
2.4.2. Results .............................................................................................. 30
ix
2.5. Conclusion .................................................................................................. 34
Chapter 3 Nano-particle Based Large Area Broadband Plasmonic

Surfaces and Their Applications ............................................................................ 35
3.1. Large Area Plasmonic Metasurfaces ......................................................... 36
3.2. Single Molecule Raman Events Using Large Area Plasmonic Surfaces .... 40
3.2.1. Observation of Single Molecule Raman Events Using Smart Phone's

Camera 42
3.3. Probing Hot-electron Effects in Wide Area Plasmonic Surfaces Using XPS
................................................................................................................................ 48
3.3.1. Introduction ...................................................................................... 48
3.3.2. Results .............................................................................................. 49
3.4. Conclusion .................................................................................................. 55
Chapter 4 Exploiting Native Aluminum Oxide for Multispectral Aluminum

Plasmonics 56
4.1. Introduction ................................................................................................ 57
4.2. Native Oxide Based Plasmonic Meta-surfaces........................................... 58
4.3. Nanoparticle Based Large Area Surfaces .................................................. 64
4.4. Surface Enhanced Infrared Absorption (SEIRA) Spectroscopy and Surface

Enhanced Raman Spectroscopy (SERS) Using Hierarchical Plasmonic Surfaces 67
4.5. All Aluminum Hierarchical Surfaces in the Infrared ................................. 75
4.6. Conclusion .................................................................................................. 82
Chapter 5 .............................................................................................................. 83
Interference Coating Based Sensing Platforms ................................................ 83
5.1. Thermally Tunable Ultrasensitive Vibrational Spectroscopy Platforms

Based on Thin Phase Change Films ...................................................................... 83
5.1.1. Introduction ...................................................................................... 84
5.1.2. Results .............................................................................................. 86
5.1.3. Thermally Tunable IR Spectroscopy Platforms Using Ge2Sb2Te5... 93
x
5.2. Strong Interference Based Colorimetric Sensors ....................................... 99
5.2.1. Introduction ...................................................................................... 99
5.2.2. Results ............................................................................................ 101
5.2.3. Protein Binding Experiment ........................................................... 106
5.3. Conclusion ................................................................................................ 107
Chapter 6 ............................................................................................................ 109
Conclusion .......................................................................................................... 109
Bibliography....................................................................................................... 111
Appendix A ........................................................................................................ 135
A.1. Effective index simulation in MIM geometry ..................................... 135
A.2. Field Profile Simulation in Multilayer Coatings ............................... 136
A.3. Transfer Matrix Code......................................................................... 137
xi
List of Figures
Figure 2-1 Surface plasmon excitation scheme at metal-dielectric interface. ............. 8

Figure 2-2. Dispersion relation of surface plasmons at metal-air interface (Solid blue
curve). Dispersion of photons in air and glass are shown with solid and dashed red
curves. ........................................................................................................................ 10
Figure 2-3. Surface plasmon excitation scheme for Kretchmann geometry (a) and
angular reflectance spectrum for λ=600nm (b). ......................................................... 10
Figure 2-4. Geometry of metal-insulator-metal resonator geometry. ........................ 11
Figure 2-5. (a) Schematic cross section of metal–insulator–metal (MIM) meta-
material structures. The metal layers are evaporated Ag, and the dielectric layer (0–
40 nm thick) is Al2O3 deposited. (b) Scanning electron micrograph of a
representative structure is shown; scale bar 1 μm. (c) Effective index (neff) and
impedance (ZMIM) of the MIM fundamental TM mode as a function of wavelength
for various dielectric layer thicknesses (3, 13, 23, 33, and 43 nm) are given for
convenience. (d) Depending on the frequency of excitation, different resonant modes
can be excited within the MIM section. In the one-dimensional case, the MIM can be
viewed as a transmission line, whose propagation constant (or effective index)
depends on the gap between the top and bottom metal layers. Truncation at both ends
results in a Fabry–Perot (FP)-type resonator. Electric field intensity squared is
plotted for a 200 nm long MIM structure with 10 nm dielectric gap excited at 555
nm, corresponding to a resonance of the FP resonator. (e) When the same structure is
excited at 660 nm, another resonance is excited displaying a different field
distribution. (f) When the gap is reduced to zero, one resonant mode persists in the
vertical transmission line cavity formed with side walls of the top metal layer as
conducting planes. (g) Cartoons show the field amplitudes of modes shown in (d, e,
and f). ......................................................................................................................... 15
Figure 2-6. (a) Cavity resonance (LFP) can be thought as corresponding to a light
atom, with a single localized state, and (b) MIM can be thought as corresponding to a
xii
heavy atom, with a multiple localized states. (c) In a one-dimensional periodic
arrangement of the light atoms, coupling through the surface plasmon mode can
introduce a coupling of the LFPs and propagating bands can be formed. (d)
Similarly, in a one-dimensional periodic arrangement of the MIM structures with a
relatively thick top metal and a thin gap dielectric, the MIM modes and the LFP
mode formed between consecutive MIMs are hybridized to form a diatomic
molecule. The molecules are coupled to form the band structure of the diatomic
crystal. (e) Effect of mode coupling is observed in the reflection coefficient, plotted
as a function of wavelength and dielectric thickness, h, for a one-dimensional crystal.
Numerical computation results are shown for a 250 nm period structure with 180 nm
MIM width, for normal incidence. Resonance wavelengths are calculated using the
Fabry–Perot resonator model for the MIM structures (black lines; see text for details)
that are superimposed. Antisymmetric modes (even mode number m) are not coupled
to the free-space modes at normal incidence and, therefore, contribute no absorption
(arrow (i)). The LFP mode wavelength is estimated using the LC model (vertical line
around 520 nm, denoted by arrow (ii)). (f) Reflection coefficient is plotted as a
function of wavelength and dielectric thickness h at a 10° angle of incidence. Due to
the broken symmetry, even modes also contribute to absorption. (g) Lumped circuit
model for the diatomic unit cell consists of a transmission line capacitively coupled
to a localized LC resonator representing the LFP mode. The model is used to
calculate the resonance frequencies of the modes, plotted as dots, on top of the
computational results in (f). The model captures the essential features of numerical
calculations such as the resonance frequencies and anticrossing behavior due to
coupling of modes. ..................................................................................................... 19
Figure 2-7. (a) Band structure of a monatomic meta-surface is calculated using the
circuit model in Figure 2g and the ABCD matrix approach and superimposed on the
reflection obtained through numerical calculations. When the dielectric thickness is
zero, only the bands formed are through coupling of the LFP (arrow I) modes via the
surface plasmon mode (arrow II) are observed. (b) Increasing the gap to 3 nm results
in the appearance of a large number of MIM modes and formation of a diatomic
meta-surface. (c) Increasing the gap to 13 nm results in a reduction of the number of
MIM modes, while increasing the absorption in the coupled bands due to improved
impedance matching of the surface with free-space propagating modes. Further band
xiii
structures are shown in (d) and (e) for 180 and 160 nm MIM widths, with 250 nm
period. Color bar shows corresponding reflectance values. ....................................... 20
Figure 2-8. Theoretical and experimental reflectance plotted as a function of
wavelength at normal incidence for (a) one-dimensional, 250 nm period structures
with nanowire widths of 180 nm (I), 170 nm (II), 160 nm (III), 150 nm (IV), 140 nm
(V) and (b) one-dimensional, 300 nm period structures with nanowire widths of 200
nm (I), 190 nm (II), 180 nm (III), 170 nm (IV), 160 nm (V). (c) TM mode reflectance
for a 10 nm band around 550 nm observed with different numerical apertures (20×,
50×, and 100× objectives) on one-dimensional 250 nm period 150 nm width
structures. Insensitivity to numerical aperture demonstrates the quasi-omni-
directional absorption of the surfaces. ....................................................................... 21
Figure 2-9. (a) Reflectance map (540–600 nm band) acquired using a 20× objective
(NA 0.4) on a 250 nm period structure with 200 nm MIM width. (b) Raman spectra
map (intensity of 591 cm–1 band, collected with 100 μW excitation power at 532
nm, 20× objective, 100 ms dwell time per pixel) of Cresyl Violet monolayer on the
same structure. Scale bar is 10 μm. (c) Reflectance as a function of wavelength for
several locations and (d) corresponding Raman spectra. Reflectance is plotted at two
different locations (I) and (II) as referenced to location (III). Due to non-uniformity
of the fabrication process, a gradient of the resonance wavelength is observed from
top-right position to bottom-left position. When the absorption overlaps with
excitation and emission wavelengths, improved Raman scattering is observed. Inset
shows chemical structure of Cresyl Violet. (e) Raman signal is collected using a
longer integration time (22 s, 100 μW excitation power) on a planar silver surface,
unpatterned MIM, and 250 nm period MIM regions. Although Cresyl Violet exhibits
no Raman signal on the plane metal surface, some enhancement is seen on
unpatterned MIM regions, possibly due to the surface roughness of the top layer. .. 23
Figure 2-10. (a) Superimposed Raman spectra collected from 1600 individual spots
over an area of 10 μm × 10 μm (100 μW excitation power, 100× objective, and 40
ms dwell time per pixel). (b) Histograms of intensity of two spectral locations shown
by arrows (I) and (II) demonstrate uniform signal intensity within ca. 10% of average
value for the fluorescence (arrow II) and ca. 20% for the Raman signal (arrow I). (c)
Raman map formed using the 591 cm–1 Raman band (scale bar 2 μm) where the
contrast is enhanced to show several dead-spots with submicrometer dimensions,
demonstrating the high-resolution imaging capability with such substrates. ............ 24
xiv
Figure 2-11. (a) Enhancement factors (EF) averaged over various regions (see inset
in c) for a MIM structure with 50 nm periodicity, 10 nm top metal thickness, 20 nm
dielectric gap, and 30 nm top metal width. (b) Maximum value of EF for different
regions for the geometry in panel a. (c) Average EF values for a MIM structure with
100 nm periodicity, 20 nm top metal thickness, 20 nm dielectric gap, and 80 nm top
metal width. (d) Maximum value of EF for different regions for the geometry in
panel c. (e) Average EF values for a MIM structure with 250 nm periodicity, 50 nm
top metal thickness, 20 nm dielectric gap, and 230 nm top metal width. (f) Maximum
value of EF for different regions for the geometry in panel e. It is seen that shrinking
the MIM size results in fewer resonances and improved average enhancement over
the unit cell, especially on the top surface (10 nm thick slab over the top metal). The
maximum values of the EFs are much higher than the average, showing the inherent
spatial non-uniformity of enhancement. Comparing panels a–c, it is seen that tuning
of the resonances through choice of geometry greatly improves EF for the
wavelength range of interest (532 to 650 nm). .......................................................... 26
Figure 2-12. Sem images of polarization independent plasmonic metamaterial
surface. The fabricated structures are dis-shaped despite they are aimed perfect
squares. ....................................................................................................................... 28
Figure 2-13. (a) Schematic description of the proposed MMA structures. Top (b) and
side (c) view of simulated structures. Periods in x and y directions are 250nm. W is
width of the patches and R is the curvature of the corners. ....................................... 29
Figure 2-14. (a) Absorption spectra of the MMA structure with perfect nano-square
patches (R=0nm) for tox= 0 (blue) and tox= 30nm (green) (b) Absorption spectra for
various nano-square patch widths (tox= 30 nm). Magnetic (c) and Electric (d) field
profiles at the resonances shown in (a). ..................................................................... 31
Figure 2-15. Effect of the curvature on the absorption spectra for W=200nm (a,b) and
W=210nm (c,d) for tox=30nm. Effect of the oxide thickness on the absorption spectra
for W=200nm and R=100nm (e-f). The magnetic plasmon mode splits into two
modes at smaller radius as the width of the resonator increases. ............................... 32
Figure 2-16. Origin of resonances for W=200nm, R=100nm and tox=30nm.
Simulated z component of electric field profiles for λ=550nm (a) and λ=600nm (b).
Calculated electrical field profiles for modes m=1, n=2 (c) and m=3, n=1 (d). ........ 33
xv
Figure 2-17. Effect of side angle on the absorption spectra. (a) Cross section of the
simulated structure. (b-c) Absorption spectra with respect to the side angle. As the
slope of the cone increases absorption of magnetic resonance mode decreases. ....... 34
Figure 3-1. (a) Scanning Electron Micrographs (SEM) of plasmonic surfaces with 1,
3 and 5 nm mass thickness Ag overlayer shows coarsening and percolation of Ag
nanoislands. Scale bars 250 nm. Plasmonic field enhancement is greater as the
nanoislands approach each other, reducing the inter-particle gap. 3 nm mass
thickness sample exhibits greatest hot spot density. (b) Schematic view of the
substrate showing layer structure. (c) Reflectance of the surfaces near normal
incidence for 1, 3 and 5 nm mass thickness Ag over-layer and 30 nm HfO2. Gray
band shows the wavelength region of interest for Raman scattering excited by 532
nm light. Ag mass thickness of 3 nm results in a wide band meta-surface. (d) The
reflectance is plotted for HfO2 thicknesses of 5, 10, 20 and 30 nm. (e) Dependence of
reflectance on angle of incidence is plotted for 20, 30, 40, 50, 60, 70 and 80 degrees.
The 30 nm HfO2/ 3nm Ag surface is quasi-omni-directional, maintaining high
absorption over a wide wavelength range at angles up to 60 degrees. ...................... 37
Figure 3-2. . (a) Maximum |E|2 field enhancement factor as a function of wavelength,
plotted for 10, 20 and 30 nm dielectric thickness for a periodic arrangement of Ag
nanoislands (40 nm period, 10 nm thickness, 35 nm island size, 75 degree sidewall
angle), shows increase in the enhancement around 550 nm as thickness increases to
30 nm. (b) Calculated reflectance of the periodic arrangement, plotted as a function
of wavelength. (c) Maximum |E|2 field enhancement factor as a function of
wavelength, for a quasi-random surface derived from SEM data as a function of
dielectric thickness. (d) Calculated reflectance of the quasi-random arrangement,
plotted as a function of wavelength and dielectric thickness. A dielectric thickness of
40 nm is also included in the calculations, as it better fits the experimental
reflectance for 30 nm HfO2 shown in Figure 3-3. ...................................................... 38
Figure 3-3. SEM derived surface model for field enhancement and reflectance
calculations. a) The SEM data is used to extract silver island shapes. b) Field
enhancement factor (|E|2 for 550 nm excitation) at the top metal-dielectric interface.
c) Field enhancement in the middle point of upper boundary of the top metal and
dielectric interface. d) Field enhancement at the upper boundary of the metal islands
.................................................................................................................................... 39
xvi
Figure 3-4. (a) Cross sectional magnetic field profile for a periodic arrangement of
metal-insulator-metal resonators (40 nm period, 35 nm width, 20 nm thickness, 75
degree sidewall angle Ag, on 20 nm HfO2, on Ag) at 430 nm excitation wavelength
and (b) at 700 nm excitation wavelength. (c) For a top metal thickness of 10 nm,
fields have enhancement at the top surface (excited at 550 nm) as compared to (d) a
top metal thickness of 20 nm (all scale bars are 20 nm wide). .................................. 39
Figure 3-5. (a) Time series of the integrated intensity within 0 to 3500 cm -1, as
recorded by the cooled CCD spectrometer on a particularly bright hot-spot, shows
blinking events as fluctuations in the intensity (500 µW of excitation at 532 nm). (b)
The time series plotted to feature the full spectrum exhibits Raman lines that
fluctuate in intensity and frequency, characteristic of single molecule SERS. (c)
Occasionally, a fluorescent molecule is captured in the hot-spot (at time 35 s in (a)).
(d) Typical Raman spectra during a SERS blink (at time 76 s in (a)) exhibits a broad
fluorescence superimposed with distinct Raman bands. ............................................ 42
Figure 3-6. a) Schematic of the measurement set-up for using the camera of a smart-
phone for imaging with a confocal Raman system. b) Photograph of the measurement
set-up showing various components described in (a). ............................................... 43
Figure 3-7. (a) Frames from video captures using the smart phone camera on
plasmonic surfaces with 1, 3 and 5 nm mass thickness Ag over-layer and 5, 10, 20,
30 nm HfO2 dielectric layer thickness. Excitation laser is defocused to illuminate an
area 50 µm in diameter. Arrows denote blink events on a fluorescence background of
the Ag nano-island layer. Scale bar is 20 µm wide. (b) Video frames are analysed to
extract a histogram of blink event intensity for the 3 nm Ag samples at varying HfO 2
thickness. A dielectric thickness of 30 nm produces brightest blink events. As the
bottom Ag layer is removed (dielectric thickness infinite), blink events can still be
observed, however at a decreased rate and intensity. Inset shows the potential source
of blinking, i.e. surface diffusion of physisorbed volatile organic compounds into and
out of hot-spots........................................................................................................... 43
Figure 3-8. a) Schematic of the spectrometer configuration. b) Photograph of the
measurement set-up showing various components described in (a). ......................... 44
Figure 3-9. (a) Configuration for using the smart phone camera as a low resolution
spectrometer. The collection fiber from the Raman set-up is collimated and dispersed
with a transmission grating (300 lines per mm) before entering the camera. (b)
Raman spectra of silicon and ethanol on silicon collected with 13 mW of 532 nm
xvii
excitation using the cooled CCD spectrometer are shown. (c) Smart phone camera
recordings of the two orders of the dispersed input light in the snap-shot mode. Solid
blue lines show integrated pixel intensity and dotted red line shows superimposed
Raman spectra convolved with a lineshape function that represents the point spread
function (PSF) of the optical configuration. Insets show actual camera excerpts. (d)
Close-up of the second order diffraction region of the camera output (solid lines)
superimposed with Raman spectra shown in (b) convolved with the PSF of the
configuration. ............................................................................................................. 46
Figure 3-10. (a) False color coded excerpts from series of smart phone camera
snapshot captures during blink events on the plasmonic substrate, in the spectrometer
configuration (1 mW excitation at 532 nm). The spectral region is cropped, rotated
and stitched for each frame. Inset shows actual camera color coding of the same data.
Integration time per frame is ~ 1 sec. (b) False color coded excerpts from a video
sequence recorded at 30 frames per second. Although video recording is at lower
resolution, distinct spectral features during blink events can be observed (also see
Supporting Video). The SERS spectra of plasmonic surfaces (30 nm HfO 2 thickness)
were also recorded using the cooled CCD spectrometer, treated with 10 nM
Methylene Blue solution in (c) and 1 µM Cresyl Violet solution in (d) using 100 µW
excitation. Corresponding spectra are also captured using the smart phone camera as
shown in (e) and (f). Insets show excerpts of the region of interest from actual
camera captures in snapshot mode. ............................................................................ 47
Figure 3-11. (a) Cross-section of the MIM surface during an XPS measurement with
laser illumination. Jx represents the X-ray induced photoemission current into the
vacuum. (b) The band diagram of the MIM junction under X-ray and laser
illumination. Hot-electron energy distributions are shown with rectangles above the
Fermi level, due to absorption of light in the top and bottom metal layers. (c) A
simplified circuit model with various currents acting on the top metal layer which
acquires a steady state voltage of Vs. Current JT through the dielectric is modeled
within a first order approximation by the resistor RT................................................. 49
Figure 3-12. (a) Schematic description of the MIM grating structures used in
polarization dependent measurements. (b) Calculated field profiles of the structures
for TE and TM polarizations. Field enhancement is minimal for TE polarization,
while plasmonic enhancement is present in the gap for TM polarization. (Scale bar
50 nm). (c) XPS spectra of the Ag 3d peaks acquired on gratings labeled as (I) and
xviii
(II) for two polarizer orientations. As the polarizer is rotated from 0° to 90°, apparent
binding energy shifts of the Ag 3d3/2 and 3d5/2 are observed for the gratings with
different orientations. Each polarization excites only one of the gratings, for which
the plasmonic modes are excited................................................................................ 50
Figure 3-13. (a) Scanning electron micrograph of MIM surface fabricated by atomic
layer deposition of HfO2 on Ag and self-organized formation of Ag nanoislands on
HfO2 upon 3 nm thick Ag evaporation (Scale bar 250 nm). (b) Reflectance of the
surfaces for incidence angles ranging from 20° to 80° shows a broad plasmonic
absorption peak around 580 nm. (c) Photo-response of the MIM surface when
illuminated by 532 nm excitation, measured by XPS. (d) Consecutive measurements
of the XPS spectrum under dark and illuminated conditions exhibit repeatable
differential shifts of the Ag 3d lines. (e)–(h) Same as in (a)–(d) except the top Ag
mass thickness is 5 nm and a semicontinuous Ag film is formed instead of MIM
nano-islands and no surface photovoltage is observed. ............................................. 52
Figure 3-14. a) XPS spectra of the Ag 3d peaks measured on the MIM surface for
dark and illuminated conditions, using lasers of 650, 532, and 445 nm wavelength, 20
mW power. The binding energies shift to negative due to hot-electron tunneling from
the bottom metal to the top Ag island. Greater shifts are observed for increasing
photon energy. (b) As the illumination intensity (445 nm) is increased from 5 to 20
mW, greater negative shift of the binding energy is observed which saturates at high
intensities. .................................................................................................................. 53
Figure 4-1. (i) Thermal evaporation of Al on silicon substrates. 3nm germanium is
evaporated as an adhesion layer prior to Al deposition. (ii) Formation of a thin
aluminum oxide (Al2O3) layer upon exposing the Al films to air for e-beam
lithography. (iii) ~100nm PMMA is spin coated and annealed at 180˚C for 120
seconds. (iv) Patterns are formed using e-beam lithography. (v) 50nm of Al or Ag is
evaporated. (vi) Lift-off process to form the final structures. .................................... 60
Figure 4-2. Verification of NO-MIM structures through observations of resonances
in NIR. (a) Schematic of formation of NO-MIM surfaces during fabrication process
(Top patterned layer height is 50 nm). (b) Simulated reflectance spectrum of NO-
MIM structures for various Al2O3 thicknesses. While no plasmonic resonance is
observed in the absence of native oxide, plasmonic resonances in NIR are observed
due to the native oxide layer. Resonances blueshift with increasing oxide thickness.
(c) Experimental reflection spectra of the fabricated nanodisc NO-MIM structure.
xix
Diameter and period of the discs are 250nm and 400nm, respectively. (Inset) SEM
image of the nanodisc array. ...................................................................................... 60
Figure 4-3. Characterization of the native oxide thickness over a time span of 24 days
using XPS. (a) XPS spectrum of an aluminum film. High energy shoulder
corresponds to the aluminum oxide peak. (b) Evolution of the calculated native oxide
thicknesses. ................................................................................................................ 61
Figure 4-4. Resonance tuning in the NIR and MIR range using anisotropic Al bar
arrays. (a) Schematic of asymmetric NO-MIM structures with nonidentical periods
and widths along x and y directions to excite multiple modes in IR (Top patterned
layer height is 50 nm). (b) Typical reflection spectra showing multiple resonances in
the NIR and MIR range due to the asymmetry of the structures with P x=500nm,
Py=2000nm and Wx=310nm and Wy=1000nm (black curve), 1500nm (blue curve)
(top) and Wx=250nm (red curve), 310nm (blue curve) for W y=1000nm (bottom).
(Inset) SEM images of corresponding structures. Red dashed lines on the cartoon
illustrations indicate the axis on which the structures’ width is modified. (c)
Experimentally observed resonance wavelengths as functions of the Al bar width
either along the short or long axis (blue dots). Calculated resonances by treating the
NO-MIM structures as Fabry-Perot (FP) resonators consisting of truncated
waveguides (green and red curves). A better fit of the FP resonator model to the
experiments can be achieved if the dielectric function of Al2O3 is assumed to be
wavelength dependent (green curve), as opposed to a constant (red curve, n=1.6).
Vertical lines on the curves are errors bars corresponding to 0.5nm uncertainty in the
oxide thickness. (d) Effective refractive index of the waveguide forming the FP
resonator as a function of Al2O3 thickness, using the wavelength dependent dielectric
function. ..................................................................................................................... 63
Figure 4-5. Simultaneous resonances in visible and NIR regime using 1D NO-MIM.
(a) Schematic of 1D NO-MIM structures. tgr=50nm and P=250nm. Experimental
results showing tuning of resonance wavelengths as W changes between 110 and
150nm, (b) in the visible and (c) in the NIR regime. Simulation results for the
experimented structures shown in (d-e) being in a good agreement with the
experimental results. Each spectrum curve is shifted by 1 along y-axis for clarity. (f)
First order (m=1) and third order (m=3) mode profiles corresponding to NIR (ii) and
visible (i) resonances as indicated with arrows in (d) and (e). The even mode (m=2)
xx
is not observed for normal incidence. In the simulations 5nm of native oxide layer
thickness is assumed. ................................................................................................. 64
Figure 4-6. SEM images of Ag nanoparticles on silicon and their size distribution. 65
Figure 4-7. Nano-particle based large area NO-MIM structures. (a) Reflection
spectra for Ag nano-particle based NO-MIM structures. Resonance blue-shifts if the
time period between two deposition processes increases (dashed and solid red lines).
No resonance is observed if the deposition of Al and Ag is done without breaking the
vacuum (blue solid line). (Inset) SEM images of 3nm Ag on Si(i) and Al(ii). (c)
Simulation results and simulated structure (inset) for tox=3nm and 5nm. Simulated
structure is a periodic array of truncated cones with D=35nm, P=40nm, tgr=20nm and
θ=75 degrees .............................................................................................................. 66
Figure 4-8. SEIRA on NO-MIM structures with patterned Ag top layer fabricated by
e-beam lithography. Reflection spectra of these structures before (blue) and after
(green) DDT monolayer formation for (a) Wx,y=300nm and (b) Wx,y=350nm.
Periods along x and y directions are the same (Px=Py=50nm). (c) and (d) First
derivative of the reflection spectra in (a) and (b), respectively. Molecular signatures
of PMMA and DDT are more visible in the first derivative curves in (c) and (d)
compared to the reflection spectra in (a) and (b). (e) An SEM image of the fabricated
structures. (Scale bar: 1 μm ). .................................................................................... 68
Figure 4-9. (a) Infrared reflection spectrum of the monolayer DDT on bare Ag mirror
measured under Grazing Angle Illumination after background subtraction. (b)
Infrared reflection of the monolayer DDT on the NO-MIM structures with Ag top
layer after background subtraction. ............................................................................ 69
Figure 4-10. SEIRA detection of molecular monolayers on hierarchical NO-MIM
structures. (a) Schematic of hierarchical NO-MIM structures before and after
nanoparticle deposition with top patterned layer height of 50 nm. Periods and widths
along x and y directions are the same, Px=Py=500nm and Wx=Wy=350nm. (b) SEM
image of NO-MIM structures after 3nm Ag deposition (Scale bar: 500nm). (c) FTIR
reflection spectrum of thick DDT solution on a bare Al film. Region of interest is
highlighted in grey. (Inset) Molecular sketch of DDT molecule. (d) FTIR reflection
spectra of NO-MIM structures for: (i) just after fabrication of all Al NO-MIM
structures, (ii) after plasma cleaning of PMMA residues, (iii) after the formation of
hierarchical NO-MIM structures by 3nm Ag deposition, (iv) after DDT molecular
monolayer formation on hierarchical NO-MIM structures. (e) FTIR reflectance
xxi
between 2500 and 3500 cm-1, and its derivative (f) for better visualization of
stretching of monolayer DDT molecules. Arrows mark the DDT and PMMA
signatures in (e). Molecular signatures (2930 cm-1 and 2955 cm-1) of PMMA are
observable before oxygen plasma cleaning. No distinctive bands are observed after
oxygen plasma (ii) and Ag deposition (iii). ............................................................... 70
evaporated as an adhesion layer prior to Al deposition. (ii) Formation of a thin
aluminum oxide (Al2O3) layer upon exposing the Al films to air for e-beam
lithography. (iii) ~100nm PMMA is spin coated and annealed at 180˚C for 120
seconds. (iv) Patterns are formed using e-beam lithography. (v) 50nm of Al is
evaporated. (vi) Lift-off process to form all Al MIM structures. (vii) O2 plasma
cleaning for PMMA residue removal and formation of a thin oxide film on the top Al
structures. (viii) Formation of Ag nanoparticles through evaporation of 3nm Ag film.
.................................................................................................................................... 71
Figure 4-12. (a) Reflectance spectra for Ag nanoparticles-based NO-MIM (blue), all
Al NO-MIM (red) and hierarchical NO-MIM (green) surfaces in the visible regime.
(b) Reflectance spectra for all Al NO-MIM and hierarchical NO-MIM (red) surfaces
in the MIR regime. ..................................................................................................... 72
Figure 4-13. Characterization of NO-MIM structures fabricated using Focused Ion
Beam (FIB) milling. (a) Fabrication steps for NOMIM structures: (i) Thermal
evaporation of 100nm of Al on silicon substrates, (ii) formation of a thin aluminum
oxide (Al2O3) layer upon exposing the Al films to air, (iii) thermal evaporation of
50nm of Ag, (iv) FIB patterning of the top Ag layer. (b) Measured (100μmx 100μm)
area with respect to the fabricated area (50μmx 50μm). (c) Top-view and cross-
section (inset) SEM images of FIB fabricated surfaces. (d) Reflectance spectra for
the FIB fabricated surfaces with shown widths and periods. ..................................... 73
Figure 4-14. SERS detection of molecular monolayers on hierarchical NO-MIM
structures. (a) Raman Spectra of a thick DDT solution on 80nm Al coated silicon. (b)
SERS spectra of monolayer DDT on hierarchical NOMIM surfaces and an Al coated
Si substrate. Integration time and laser power are 100ms and 0.5mW, respectively,
for hierarchical NOMIM surfaces; where they are 10s and 13mW, for Al films on Si.
(c) Simulated field profiles for all Al NO-MIM (MIM1), Ag nano-particle based NO-
MIM (MIM2) and hierarchical NO-MIM (MIMIM) structures. (d) Optical
micrograph of hierarchical NO-MIM surfaces. (e) SERS mapping of DDT on
xxii
hierarchical NO-MIM surfaces. Brighter and darker regions emphasize hierarchical
NO-MIM and nano-particle based NO-MIM structures, respectively. (f) Wide-field
CMOS camera image of blink events on NO-MIM structures. Blink events are only
observed on hierarchical NO-MIM structures. .......................................................... 75
Figure 4-15. (a) 3D and (b) side view of the hierarchical aluminum surfaces. Top Al
layer thicknesses (tgr1 and tgr2) are 50nm. The thicknesses of the bottom and top oxide
layers are denoted as tox1 and tox2, respectively. (c) Top-down SEM image of the
fabricated structures. (d) Fabrication of hierarchical Al structures. (i) Deposition of
80 nm thick Al. (ii) Formation of native Al2O3 after exposure of the Al films to air.
(iii) PMMA coating. (iv) Pattering PMMA with e-beam lithography. (v) Deposition
of the 1st 50 nm thick Al. (vi) Formation of the 2nd native Al2O3 layer after vacuum
break. (vii) Deposition of the 2nd 50 nm thick Al layer. (viii) Lift-off process. ........ 77
Figure 4-16. Depth dependent XPS characterization of Al surfaces with thin native
Al2O3 on top. (a) Just after the deposition of Al films. (b) After the fabrication of
hierarchical Al structures. .......................................................................................... 78
Figure 4-17. (a) Measured and (b) simulated reflectance spectra for W=290nm,
300nm, 320nm and 340nm (Period=500nm). (c) Magnetic field intensities for the
resonances shown in (b). ............................................................................................ 79
Figure 4-18. Simulated reflection spectra for the hierarchical MIM structure with two
oxide layers assuming different extinction coefficients for the native oxide layer.... 80
Figure 4-19. Simulated reflection spectra for (a) varying t ox-1 and tox-2, (b) varying tox-
2 with tox-1=5nm. W=300 nm, P=500 nm for all cases .............................................. 81
Figure 4-20. (a) 3D schematics of higher order hierarchical aluminum surfaces with
multiple MIM resonators. Simulated reflectance spectra for structures with (b) 3, (c)
4 oxide layers. (d) Corresponding magnetic field profiles for the resonances shown in
(c). The thickness of the top Al layers is 50nm, W=300nm and P=500nm for both
geometries. ................................................................................................................. 82
Figure 5-1. Sensing performance of various optical platforms. 5 nm PMMA (a)
suspended in air, (b) on semi-continuous CaF2 substrate, (c) on 300 nm a-Si/Al
surface, (d) on 400 nm a-Si/Al surface. (i) Simulated electric field enhancement
profiles on cartoon illustrations at λ: 5780 nm (1/λ: 1732 cm-1). (ii) Simulated electric
field enhancement factors on the surface of each structure, (iii) far-field signal
(transmittance/reflectance) spectra, (iv) signal from the major PMMA band. The
signal is the difference between the far-field signals w/ and w/o PMMA layer.
xxiii
Measured optical properties of PECVD deposited a-Si are used for the simulations
(n: 3.3 in the IR, see Figure S11). Optical properties of PMMA are fitted to a
Lorentzian oscillator (Figure S12) centered around 5780 nm. .................................. 87
xxiv
List of Tables
Table 1. RGB values of sensing platforms with different SiO2 thicknesses. ........... 104
xxv
Chapter 1
Introduction
Surface plasmons are known as propagating electromagnetic waves at metal-

dielectric interfaces. The field of plasmonics deals with interaction of
electromagnetic waves with metal surfaces and nanostructures. Since, the surface
plasmon field is tightly confined to the interface; electromagnetic field intensity is
enhanced at the metal-dielectric interface. Various optical sensors[1]–[3],
interconnects[4]–[7], photodetectors[8]–[11], optical elements[12] and photovoltaic
devices[13], [14] utilize such unique characteristic of surface plasmons. There are
different ways of exciting surface plasmons at metal-dielectric interfaces such as
electrons and light[15]–[17]. Among the excitation sources, light is the most
common and easiest way of exciting surface plasmons[18].
Metallic nanoparticles, on the other hand, provide an easier route for plasmonics
research due to ease of fabrication and characterization methods[15], [19], [20].
Various fabrication methods are utilized plasmonic metallic nanoparticles. E-beam
lithography[21], [22], focused ion-beam lithography[23], nano-imprint
lithography[24], [25], interference lithography[26] and nano-sphere lithography[27],
[28] are among the top-down fabrication methods where nanoparticles with uniform
size distributions can be fabricated. Despite the uniformity and resolution of
fabricated devices, top-down fabrication methods like e-beam lithography and
focused-ion-beam lithography suffer from high fabrication costs and small
fabrication areas. Periodic nanopatterns can be fabricated over large areas using
interference lithography and nanosphere lithography, even at wafer-scale[1], [26].
Physical vapor deposition (PVD) of nanoparticles on the other hand emerges as one
of the easiest way of fabricating nanoparticles over large areas without using
1
chemical processes[29], [30]. Thermal evaporation and e-beam evaporation are
among the most used PVD methods to fabricate metal nanoparticles[29], [31].
However, the control of physical vapor deposition of nanoparticles lacks order and
uniformity due to stochastic nature of nanoparticle formation process[32], [33].
Dewetting of metal films results in nanoparticles during PVD process due to the
surface energy difference between the substrate and metal film. Depending on the
substrate, the thickness of the metal film, the size of nanoparticles and the
interparticle distance change. The size of nanoparticles also differs depending on the
deposition rate and the thickness of the metal[30]. Moreover, the post-fabrication
processing of metal nanoparticles such as thermal and laser annealing modify the
final shape and the interparticle distance of nanoparticles[34], [35]. On the other
hand, fabrication of nanoparticles by chemical means gives flexibility to fabricate
nanoparticles in low-costs with large amounts with fair size distributions[36].
Nanoparticles can be fabricated in various shapes which further provide different
functionalities for specific applications[20].
When nanoparticles are brought together, they start interacting optically[37], [38].
The interaction between nanoparticles can be intuitively understood as the coupling
of individual resonators which results a shifted energy spectrum compared to the
individual resonators' energies[39]. Due to coupling, plasmonic coupled systems
offer a very rich energy spectrum[40], [41]. Since, energy spectrums of coupled
structures are broad or multispectral they are desired for various nanophotonics and
plasmonic applications. Coupled plasmonic structures are fabricated by either top-
down or chemical methods[19], [42]. The aggregating nanoparticles in solutions
results another plasmon peak in the red wavelengths which is due coupling of
nanoparticles[43]. Scattering spectra of nanoparticles changes drastically when they
get close to metallic planes due to hybridization of surface plasmon and localized
plasmon modes and resulting gap plasmon modes in between[44]. Gap plasmon
modes are also known as metal-insulator-metal (MIM) modes when they are excited
between two metal planes. MIM geometry offers rich physics as well as plasmon
enhanced applications rooted in classical and quantum electromagnetics[45], [46].
Negative index metamaterials, superlensing, metamaterial absorbers, color printing,
photodetectors, plasmonic sensors and surface enhanced spectroscopies are few of
the novel applications of MIM based plasmonic structures[22], [47]–[54].
2
One of the applications of plasmonic surfaces studied here is label-free sensing.
Labeled sensing platforms are tagged with fluorescent molecules or quantum dots
which are prone to bleaching or deterioration[55], [56]. Label-free sensing strategies
are developed to overcome the difficulties that labeled sensing platforms face.
Surface plasmon sensors (SPR) are the known as the golden standard for the label-
free sensing platforms. Bulk prism based SPR systems are among the widely used
plasmonic sensing platforms to probe bio-molecular binding events. Various sensing
platforms have been developed on nanoparticles, gratings and plasmonic
metamaterials utilizing top-down and bottom-up fabrication methods. Spectroscopic
methods combined with plasmonic structures offer another route for label-free
sensing methods. Surface enhanced Raman spectroscopy (SERS) provides
vibrational information and molecular signatures of molecular moieties[57]–[59].
Typically, Raman spectroscopy requires large amount of molecules to probe
molecules due to low probability of release of a Raman scattered photon. When, a
monochromatic light (typically laser light) interacts with molecules' vibrational
modes, scattered photons have either higher energy (Anti-Stokes) or lower energy
(Stokes) than the excitation source. Anti-stokes signals are generally filtered out
from the Raman spectra due to their low intensities. Since Raman signal scales with
𝐸 = |𝐸0 |4 , enhancing electromagnetic field intensity near the probe molecules
increases the Raman signal enormously. In SERS enhancing electromagnetic field
intensity by utilizing nanostructured metal surfaces enhances the Raman signal 4-10
orders of magnitude, thus enabling the possibility of detecting single molecules. On
the other hand, debate on the origin of single molecule Raman events is still
unresolved. The charge transfer between the metal and molecules might lead to the
single molecule Raman events. This mechanism is also known as chemical
enhancement factor. SERS platforms should provide electromagnetic field
enhancement at the excitation and Stoke's wavelengths. Double resonant
nanoantennas, metamaterials and plasmonic gratings are good examples for good
SERS platforms where each resonance is accompanied with field enhancement[25],
[60], [61]. Broadband plasmonic surfaces are desired as SERS substrates whose
resonances span all the vibrational bands of a molecule.
Infrared absorption spectroscopy is a well established label-free tool used in
biological and chemical sciences to identify molecular moieties by utilizing specific
molecular vibrational bands in the infrared wavelengths[62]. Since the selection rules
3
in IR and Raman spectroscopy are different, the vibrational information of these
techniques is complementary[63]. However, the absorption efficiencies of molecules
for specific bands are low in the IR which limits the use of IR spectroscopy for small
amount of molecules, molecular or protein monolayers. Different techniques are
developed in the IR to identify molecules in small amounts. Attenuated total
reflectance (ATR), infrared reflection absorption spectroscopy (IR-RAS) are two
widely used tools to identify molecules[64], [65]. In ATR, electromagnetic field
decays exponentially when light is sent through a high index prism to a low index
medium. The enhanced field at the prism medium interface is used to increase
absorption of molecules. Similar to ATR, electromagnetic field is enhanced when a
p-polarized light is sent to a substrate at oblique incidences on which molecules are
deposited. Metal nanoparticles and deterministically designed plasmonic metal
nanostructures emerge as an alternative tool to detect minute amount of
molecules[58], [66], [67]. The electromagnetic field is highly enhanced and
concentrated around nanoparticles which results in the enhancement of the
absorption of the molecules in close proximity. The absorption of the molecules is
sensed in the far field signal (reflection/transmission spectrum). This powerful
technique is also known as surface enhanced infrared absorption spectroscopy
(SEIRA). However, there are some limitations to SEIRA. Vibrational bands of
molecules are embedded in plasmonic resonance background in the farfield
reflection spectra and data post processing is required to extract positions of
vibrational bands. All of the vibrational bands cannot be detected because the
electromagnetic field cannot be enhanced over very large bandwidths (3-20μm band)
using single resonant plasmonic structures. Moreover, the positions and shapes of
vibrational modes are modified due to strong coupling between plasmonic
resonances and vibrational modes. Although electromagnetic field enhancement is on
the order of 103-105 signal intensities suffer from low vibrational intensities (~1-5%).
Therefore, field enhancements with high mode volumes are preferred over highly
localized field enhancements.
Energy harvesting is another application area that plasmonics studies target.
Plasmon enhanced photovoltaics is emerged as alternative to current solar cell
technologies by either incorporating metallic nanophotonics structures to thin films
or silicon solar cells or providing novel solar cell designs[13], [49], [68]–[71]. Most
efforts to utilize plasmons in solar cells are based on enhancing electromagnetic field
4
and increasing the absorption in semiconductor films to decrease the costs and
thicknesses. Despite the huge amount of literature, the success of plasmons in
photovoltaics area is debatable[72]. Still, the plasmonics promise a bright future for
silicon photonics industry for the IR regime[8]. Structuring Schottky contacts with
plasmonic structures enabled to use of silicon in photodetectors in the IR
wavelengths with lower energy than silicon bandgap. Generation and capture of hot-
electrons are the main mechanism for photodetection in these devices. In a Schottky
diode, due to band alignment between metal and semiconductor layer, there is a band
offset that is smaller than the band gap of the semiconductor. When light is absorbed
by the metal contact, electrons are generated, which are either thermalizes or
contribute to current through tunneling or hopping above the barrier. The electrons
that have energy higher than the band offset are called hot-electrons. However, due
to reflective nature of metals the most of the incident light is reflected, thus the
generation rate of hot electrons is small. By patterning Schottky contacts, hot-
electron generation can be increased due to plasmon enhanced electromagnetic
field[8], [14]. MIM resonators emerge as one of the novel hot electron
photodetection devices[52], [73]. Similar to Schottky based hot-electron devices, by
patterning the metal contacts plasmons are employed to increase hot-electron
generation rate. Since these devices utilize high bandgap insulators between two
metal contacts, the photoresponsivities are small compared to Schottky based hot
electron devices.
Perfect absorption of light is important for aforementioned and can be obtained
using patterned metallic surfaces through exciting plasmons. On the other hand,
strong interference coatings offer an easier way for perfect light absorption by using
ultrathin absorbing semiconducting films[74], [75]. When a thin dielectric film is
deposited on a metal surface, resonant light absorption occurs at thicknesses; 𝑡 ≅
λ/4n, where n is the dielectric constant and λ is the resonance wavelength. If the
dielectric constant increases the resonance thickness becomes 𝑡 < λ/4n due to
increased nontrivial phase at metal dielectric interface. If the dielectric film has a
non-zero extinction coefficient (𝑛 = 𝑛̃ + 𝑖𝑘̃), the thickness decreases to 𝑡 ≪ λ/4n
and depending on the refractive index function of metal and dielectric resonance
thickness changes. Non-zero extinction coefficient results small phase accumulation
at oblique angles and the absorption becomes wide-angle. Since these types of
coatings exhibit distinct colors depending on the resonance wavelength, most of the
5
applications concentrate on reflective type color filters[74]–[78]. Light absorption on
these coatings can be exploited as novel photodetectors.
In this thesis, we first develop a circuit model to understand the coupling
mechanism of a coupled system based on MIM geometry. Our findings are supported
with experimental results by fabricating and characterizing the proposed structures.
These structures are utilized as SERS substrates with unity enhancement over large
areas. Fabrication imperfections are exploited to achieve triple band plasmonic
absorber geometry due to excitation of three different plasmon modes. Then,
following theoretical and analytical study, MIM structures with nanoparticle top
layer are fabricated. By optimizing the conditions for nanoparticle formation, large
area, wide angle and broadband plasmonic surface are achieved. Single molecule
Raman blinking events are observed with these structures using a confocal Raman
microscope equipped with cooled CCD spectrometer. Single molecule Raman events
are also detected when a mobile phone's camera used instead of cooled CCD
spectrometer. To the best of our knowledge this is the first observation of a single
molecule Raman blinking event using a cmos camera. Same surfaces are also used as
a photodetector to probe hot electrons using a x-ray photoemission spectroscopy
(XPS) system. The final part of plasmonic studies includes aluminum based
plasmonic surfaces. By exploiting native oxide layer on Al forming during the
fabrication processes, we are able to realize plasmonic surfaces in the visible and IR
wavelengths. These surfaces are used as SERS and SEIRA platforms simultaneously.
To address the issues of plasmonic surfaces such as low signal intensities narrow
bandwidth, small surface area and costs, thin film interference coatings are proposed
as novel sensing platforms in the infrared and visible wavelengths. These coatings
perform better than conventional plasmonic surfaces when they are used as infrared
sensing platforms. Similar surfaces are used as colorimetric sensor platforms to sense
protein monolayers and bilayers.
6
Chapter 2
Well-Defined Coupled Plasmonic

Structures
Plasmonic metamaterials allow confinement of light to deep subwavelength

dimensions, while allowing for the tailoring of dispersion and electromagnetic mode
density to enhance specific photonic properties. Optical resonances of plasmonic
molecules have been extensively investigated; however, benefits of strong coupling
of dimers have been overlooked. In this part, a novel plasmonic geometry and its
surface enhanced Raman spectroscopy application are demonstrated. Then,
fabrication related issues are exploited for a multispectral metamaterial absorber.
Before going into details of this chapter, surface plasmon at different metal dielectric
interfaces are studied.
Parts of this chapter were published as;
1. Sencer Ayas, Hasan Guner, Burak Turker, Oner Ekiz, Faruk Dirisaglik ,Ali
Kemal Okyay and Aykutlu Dana, "Raman Enhancement on a Broadband Meta-
Surface", ACS Nano, 2012, 6 (8), pp 6852–6861
2. Sencer Ayas, Gokhan Bakan and Aykutlu Dana, "Rounding corners of nano-
square patches for multispectral plasmonic metamaterials absorbers" Optics
Express, Vol. 23, No. 9, 2015
7
2.1. Surface Plasmons at Metal-dielectric Interface
To understand the coupling conditions, a very simple geometry (metal-dielectric
interface) is used to give the basics of surface plasmons [79]. We start with the
assumption of evanescently decaying and propagating surface plasmons in the z and
x-directions respectively as shown in Figure 2-1.
Figure 2-1 Surface plasmon excitation scheme at metal-dielectric interface.
For a TM polarized wave, electric and magnetic fields are given as

𝐻𝑦 (𝑧) = 𝐴𝑒 𝑖𝛽𝑥 𝑒 −𝑘𝑑𝑧 (2.1)
𝐴𝑘𝑑 𝑖𝛽𝑥 −𝑘 𝑧
𝐸𝑥 (𝑧) = 𝑖 𝑒 𝑒 𝑑 (2.2)
𝜔𝜀𝑜 𝜀𝑑
for z>0 (dielectric region)
𝐻𝑦 (𝑧) = 𝐵𝑒 𝑖𝛽𝑥 𝑒 𝑘𝑚𝑧 (2.3)
𝐵𝑘𝑚 𝑖𝛽𝑥 𝑘 𝑧
𝐸𝑥 (𝑧) = −𝑖 𝑒 𝑒 𝑚 (2.4)
𝜔𝜀𝑜 𝜀𝑚
for z<0 (metal region).

By equating Hy and Ex at z=0 we obtain A=B and Akd/εd=-Bkm/ εm. Hence:
𝑘𝑚 𝑘
= − 𝜀𝑑 (2.5)
𝜀𝑚 𝑑
Since propagation constants and 𝜀𝑑 are real, surface plasmons are supported for
materials at the metal-dielectric interface which have negative dielectric constants.
Typically, metals have negative dielectric functions above plasma frequency.
Although, plasma frequencies of metals are in the deep-UV regime, some metals
have real dielectric functions in the visible portion of the spectrum due to interband
transitions. For example, gold (Au) has an interband transition around 500nm which
8
limits the use of Au in the UV wavelengths. Using βx2=ko2εd,m-(kd,m)2 and Equation
𝜀 𝜀
2.5 together results surface plasmon dispersion relation 𝛽𝑥 = 𝑘𝑜 √𝜀 𝑑+𝜀𝑚 . In a
𝑑 𝑚
similar manner, we can derive the surface plasmon coupling condition for TE
polarized waves.
For a TE polarized wave electric and magnetic fields are given as
𝐸𝑦 (𝑧) = 𝐴𝑒 𝑖𝛽𝑥 𝑒 −𝑘𝑑𝑧 (2.6)
𝐴𝑘𝑑 𝑖𝛽𝑥 −𝑘 𝑧
𝐻𝑥 (𝑧) = 𝑖 𝑒 𝑒 𝑑 (2.7)
𝜔𝜇
for z>0 (dielectric region)
𝐸𝑦 (𝑧) = 𝐵𝑒 𝑖𝛽𝑥 𝑒 𝑘𝑚𝑧 (2.8)
𝐻𝑥 (𝑧) = −𝑖 𝑒 𝑒 𝑚 (2.9)
𝜔𝜇
for z<0 (metal region).

The continuity equation for Ey and Hx at the metal-dielectric interface (z=0)
results in A=B and Akd+ Bkm=0 where we obtain the relation A(kd+ km)=0. Since
kd,m>0, A=0 is the only condition for TE mode. Thus, Ey=0, Hx=0 and the surface
plasmons are not supported for TE polarized waves.
So far, a propagating wave at a metal-dielectric interface is assumed to derive
surface plasmon dispersion relation. However, excitation of surface plasmons at
metal-air interface requires special coupling geometries such as Kretschmann
geometry, since the dispersion relation of surface plasmons and photonic dispersion
relation do not cross as shown in Figure 2-2. In Kretshcmann geometry, light is
coupled through a prism rather than air. Such excitation scheme results on a
resonance when the incidence angle is scanned for a certain wavelength as shown in
Figure 2-3. Surface plasmon resonance (SPR) sensors are developed by utilizing
such resonant characteristic. Moreover, the reflectance spectrum depends on the
dielectric environment that surface plasmons are coupled. Biomolecular interactions
and binding events can be detected using SPR sensors by probing the reflected light
intensity due to the change of dielectric environment close to the metal-dielectric
interface during binding events. Small molecules, viruses, protein interactions and
bacteria can also be detected using SPR sensors.
9
Figure 2-2. Dispersion relation of surface plasmons at metal-air interface (Solid blue curve).
Dispersion of photons in air and glass are shown with solid and dashed red curves.
Figure 2-3. Surface plasmon excitation scheme for Kretchmann geometry (a) and angular
reflectance spectrum for λ=600nm (b).
10
2.2. Surface Plasmons in Metal-Insulator-Metal
Resonators
Different surface plasmon schemes are proposed for specific applications. Metal-
insulator-metal (MIM) resonators are among the widely used geometry in plasmonics
and metamaterial applications due to their ease of fabrication and tunability, while
controlling surface plasmon propagation by tuning effective refractive index. Since
the coupled plasmonic geometry in this study is based on MIM resonators, the
general dispersion relations and effective dielectric constants are studied here. The
geometry of an MIM resonator is shown in Figure 2-4. The thickness of dielectric
layer is d. The metal thicknesses are assumed semi-infinite for top and metal layers.
Figure 2-4. Geometry of metal-insulator-metal resonator structure.
The electric and magnetic field profiles in top metal, dielectric and bottom
metal regions are given as
𝐻𝑦 (𝑧) = 𝐴𝑒 𝑖𝛽𝑥 𝑒 −𝑘𝑚𝑧 (2.10)
𝐴𝑘𝑚 𝑖𝛽𝑥 −𝑘 𝑧
𝐸𝑥 (𝑧) = 𝑖 𝑒 𝑒 𝑚 (2.11)
for z>d/2 (top metal region).
𝐻𝑦 (𝑧) = 𝐵𝑒 𝑖𝛽𝑥 𝑒 𝑘𝑚𝑧 (2.12)
𝐸𝑥 (𝑧) = −𝑖 𝑒 𝑒 𝑚 (2.13)
for z<-d/2 (bottom metal region).
11
𝐻𝑦 (𝑧) = 𝐶𝑒 𝑖𝛽𝑥 𝑒 −𝑘𝑑 𝑧 + 𝐷𝑒 𝑖𝛽𝑥 𝑒 𝑘𝑑 𝑧 (2.14)
𝐶𝑘𝑑 𝑖𝛽𝑥 −𝑘 𝑧 𝐷𝑘𝑑 𝑖𝛽𝑥 𝑘 𝑧
𝐸𝑥 (𝑧) = 𝑖 𝑒 𝑒 𝑑 −𝑖 𝑒 𝑒 𝑑 (2.15)
𝜔𝜀𝑜 𝜀𝑑 𝜔𝜀𝑜 𝜀𝑑
for d/2<z<-d/2 (dielectric region).
The continuity of Hy and Ex at z=-d/2 and z=d/2 results in
−𝑘𝑑 𝑑 𝑘𝑑 𝑑
𝐴𝑒 −𝑘𝑚𝑑/2 = 𝐶𝑒 2 + 𝐷𝑒 2 (2.16)
𝑘𝑚 −𝑘 𝑑/2 𝑘𝑑 −𝑘𝑑 𝑑 𝑘𝑑 𝑘𝑑 𝑑
𝐴 𝑒 𝑚 = 𝐶𝑒 2 − 𝐷𝑒 2 (2.17)
𝜀𝑚 𝜀𝑑 𝜀𝑑
𝑘𝑑 𝑑 −𝑘𝑑 𝑑
𝐵𝑒 −𝑘𝑚𝑑/2 = 𝐶𝑒 2 + 𝐷𝑒 2 (2.18)
𝑘𝑚 −𝑘 𝑑/2 𝑘𝑑 𝑘𝑑 𝑑 𝑘𝑑 −𝑘𝑑 𝑑
−𝐵 𝑒 𝑚 = 𝐶𝑒 2 − 𝐷𝑒 2 (2.19)
𝜀𝑚 𝜀𝑑 𝜀𝑑
After eliminating A and B using Equations 2.16-17 and 2.18-19 respectively, we
obtain
𝜀𝑚 𝑘𝑑 −𝑘𝑑 𝑑 𝜀𝑚 𝑘𝑑 𝑘𝑑 𝑑
(1 − ) 𝐶𝑒 2 = (1 + )𝐷𝑒 2 (2.20)
𝑘𝑚 𝜀𝑑 𝑘𝑚 𝜀𝑑
𝜀𝑚 𝑘𝑑 𝑘𝑑 𝑑 𝜀𝑚 𝑘𝑑 −𝑘𝑑 𝑑
(1 + ) 𝐶𝑒 2 = (1 − )𝐷𝑒 2 (2.21)
If we multiply both sides of Equations 2.20 and 2.21, 𝐶 = ±𝐷 is achieved. Then
Equation 2.20 separates into two equations as follows
(1 − ) 𝑒 2 − (1 + )𝑒 2 = 0 (2.22)
(1 − ) 𝑒 2 + (1 + )𝑒 2 = 0 (2.23)
Equations 2.22 and 2.23 result in two distinct solutions which are
𝑘𝑑 𝑑 𝜀𝑚 𝑘𝑑
tanh( )=− (2.24)
2 𝑘𝑚 𝜀𝑑
𝑘𝑑 𝑑 𝜀𝑑 𝑘𝑚
tanh( )=− (2.25)
2 𝑘𝑑 𝜀𝑚
Different parameters can be extracted from Equations 2.24 and 2.25 such as
effective refractive index and propagation length which are important parameters for
plasmonic waveguide applications. Effective refractive index parameter can be
extracted using Equations 2.24-25 and βx2=ko2εd,m-(kd,m)2 where βx=koneff.
12
2.3. Raman Enhancement on a Broadband Meta-surface
2.3.1. Introduction
Plasmonic excitations of metallic nanostructures have attracted a great deal of

attention in the past decades, due to the rich variety of geometric configurations, the
associated optical properties and phenomena, and the wide range of present and
potential future applications[18], [80]. Propagating and localized plasmons have been
utilized in the design of photonic structures to efficiently couple free-space
propagating light onto highly confined surface modes, resulting in the enhancement
of electromagnetic field intensities[81], [82]. Nonlinear optical effects benefit from
plasmonic field enhancement[81], and plasmonics have the potential to be an
enabling technology for quantum optics and all-optical information processing[83],
[84]. It has been shown that plasmonic field enhancement allows the observation of
Raman scattering from single molecules with low excitation powers down to
microwatts[85], [86]. The lack of reliability resulting from the spatially non-uniform
nature of plasmonic field enhancement can be a problem for applications requiring
repeatability. In the case of surface-enhanced Raman scattering (SERS), regions with
high enhancement (so-called hot spots) are typically major contributors to the
observed signal. Raman intensity enhancement is estimated
through ISERS ≅I0|E(ωexc)E(ωdet)/E0(ωexc)E0(ωdet)|2, where ωexc and ωdet are the
excitation and detection frequencies, and E and E0 are the electric field intensities
with and without the presence of plasmonic structures. Defining an enhancement
factor, EF(ω) = |E(ω)/E0(ω)|2, overall Raman enhancement factor can be written as
the product of excitation and detection factors, EFSERS =EF(ωexc)EF(ωdet). Spatial
non-uniformity of the electric field directly translates into a spatial non-uniformity
of EFSERS and can be an important disadvantage for repeatability. Hot spots are
typically formed when two metal regions come close (within a few nanometers) to
each other, and even periodic structures may display a wide distribution of
enhancement factors[87]. In order to achieve high and spatially uniform field
enhancement, engineered surfaces that exhibit plasmon modes at both the excitation
and scattering wavelengths are needed[53], [60], [88], [89]. Previously, metal
nanoparticle clusters (bottom-up approach) and sparse structures or biharmonic
gratings with double resonances (top-down approach) were used for this purpose[90].
13
Bottom-up approach substrates offer the advantage of simplicity and offer
uniformity over large scales; on the other hand, the statistical nature of production
processes, in principle, prevents completely hot-spot-free and uniform enhancement
on a microscopic scale. Sparse arrays of multiply sized nanoantennas, or
multiperiodic structures, may be designed to enhance fields at both excitation and
scattering frequencies, however, exhibit spatially non-uniform spectral properties
over the extent of a wavelength. In contrast, plasmonic meta-materials possess
subwavelength periodicity. When fabricated using a top-down approach, they have
the potential to offer spatially uniform, wide band coupling required for uniform
Raman enhancement. In this part of the thesis, we discuss the application of
plasmonic meta-surfaces to SERS. We describe the resonances and field
enhancements of a closely packed metal–insulator–metal configuration, which we
refer to as multiply coupled plasmonic meta-material (MCPM). Strong coupling of
the modes is shown to be effective for a subset of geometric parameters. Despite the
seemingly common and simple geometry, we show that, under strong coupling of
various types of modes, a rich band structure can be engineered within a wide
spectral range, allowing highly controllable field enhancement at excitation and
scattering frequencies, thereby allowing highly repeatable SERS.
2.3.2. Description of the Plasmonic Meta-Surface
A cross-sectional schematic of the one-dimensional periodic MIM structure is

shown in Figure 2-5a (a representative SEM micrograph is given in Figure 2-5b).
MIM elements with relatively thick top metal layers (d≥50 nm), with widths of 150–
200 nm, are repeated with periods of 200 to 300 nm, forming an inter-MIM spacing
of 10–80 nm. MIM structures can be viewed as nanoscale planar waveguides[91],
and individual MIM elements have modes confined inside the dielectric region
(referred to as waveguide mode). Localized resonances result from the Fabry–Perot
(FP)-type resonator formed within the MIM waveguide, terminated at both ends. The
propagation wavevector βSPP and effective index neff= βSPP/k0 of the fundamental TM
mode of a MIM waveguide can be calculated by solving for βSPP in the following
𝑘𝑑 𝑤
equation 𝑘𝑚 𝜀𝑑 + 𝑘𝑑 𝜀𝑚 tanh( ) = 0 here kd = (βSPP2 – εdk02)1/2, km = (βSPP2 –
2
εmk02)1/2, and k0 = 2π/λ is the propagation wave vector in free space for wavelength λ
which is also given by Equation 2.25. Typically, βSPP is a complex number, and the
14
real part of βSPP can be used to infer the effective wavelength. Assuming a unity
relative magnetic permeability for propagating TM modes inside the MIM
waveguide, the effective index (neff) can also be used to estimate the impedance of
the transmission line, ZMIM = 120π/neff (see Figure 2-5c). The propagation constant
can be used to estimate the resonances of a FP cavity. The roundtrip phase is written
as φ(λ) = 2βSPPw + 2Δφ(λ), where 2βSPPw is the phase shift due to propagation, and
2Δφ(λ) is the total phase shift acquired upon reflection at the ends of the resonator. A
FP resonator exhibits resonances when |1 – R0 exp iφ(λ)| is a minimum, where R0 is
the magnitude of the reflection coefficient. Using the above formulation, absorption
resonances of uncoupled MIM waveguides have been calculated. In general, Δφ(λ)
may be calculated through the reflection coefficient at the end of the waveguide,
taking into account the impedance mismatch between the waveguide mode and free-
space modes[92].
Figure 2-5. (a) Schematic cross section of metal–insulator–metal (MIM) meta-material

structures. The metal layers are evaporated Ag, and the dielectric layer (0–40 nm thick) is
Al2O3 deposited. (b) Scanning electron micrograph of a representative structure is shown;
scale bar 1 μm. (c) Effective index (neff) and impedance (ZMIM) of the MIM fundamental TM
mode as a function of wavelength for various dielectric layer thicknesses (3, 13, 23, 33, and
43 nm) are given for convenience. (d) Depending on the frequency of excitation, different
resonant modes can be excited within the MIM section. In the one-dimensional case, the
MIM can be viewed as a transmission line, whose propagation constant (or effective index)
depends on the gap between the top and bottom metal layers. Truncation at both ends results
in a Fabry–Perot (FP)-type resonator. Electric field intensity squared is plotted for a 200 nm
long MIM structure with 10 nm dielectric gap excited at 555 nm, corresponding to a
resonance of the FP resonator. (e) When the same structure is excited at 660 nm, another
resonance is excited displaying a different field distribution. (f) When the gap is reduced to
15
zero, one resonant mode persists in the vertical transmission line cavity formed with side
walls of the top metal layer as conducting planes. (g) Cartoons show the field amplitudes of
modes shown in (d, e, and f).
Optical properties of arbitrary metallic nanostructures have been commonly

analyzed using optical first-principles calculations. Optical properties of simple
geometric configurations, on the other hand, can be explained in terms of lumped
circuit elements[93]–[97]. The waveguide perspective is practically adequate, as the
experimentally observed optical properties of isolated structures can be well
explained using this approach[98]. Figure 2-5d,e shows the electric field profiles
when such structures are excited at two different wavelengths for a dielectric
thickness of 12 nm. Standing wave patterns within the MIM are observed when the
surface is excited near its resonances, validating the FP resonator approach (cartoons
in Figure 2-5g show E-field along the structure for various modes in Figure 2-5d–f).
In the one-dimensional case, when the dielectric thickness is reduced to zero, the
MIM waveguide modes disappear; however, the remaining structure still exhibits a
localized mode (Figure 2-5f). This remaining vertical cavity mode (referred to as
LFP, short for localized Fabry–Perot) is formed due to FP-type resonance in the
vertical metallic cavity between the top metal layers of consecutive MIM
regions[94]. The resonance frequency of the LFP can be approximately calculated
using lumped models or using the waveguide approach used above. The fundamental
resonance of closely spaced MIM structures also have been modeled previously
using lumped LC elements[94], [99]. Using such a model, the LFP mode frequency
can be approximated by ωLFP = ((Ce + Cm + (Cm2 + Ce2)1/2)/(LCmCe))1/2, where L ≅
0.5 μhw + 3w/(2εOhwP2) is the inductance per unit width (μ is the magnetic
permeability, h is the dielectric thickness, and w is the width of the
MIM), Ce ≅ cεOh/g is the capacitance per unit width due to coupling between MIM
top metals (ε is the dielectric constant, h is the dielectric thickness, g is the gap
between MIMs, and c is a correction parameter), Cm ≅ 0.25εOεh/g is the capacitance
per unit width due to coupling of top metal layer of MIMs to the metallic ground
plane. The model ignores resistive losses.When the areal density of such structures is
increased to improve surface coverage, strong coupling of individual structures
becomes inevitable, modifying the spectral properties, as well as the band structure.
16
2.3.3. Coupled Meta-Atoms and Meta-Molecules
The LFP mode can be considered as a light meta-atom, possessing a single energy
level (Figure 2-6a). Similarly, the MIM structure constitutes a heavy meta-atom, with
more than one energy level (Figure 2-6b). When the atoms are brought into contact
by a coupling mechanism, bands emerge similar to the coupled resonator optical
waveguide (CROW)[100] A periodic arrangement of grooves (i.e., LFPs) results in a
monatomic crystal (Figure 2-6c). In the monatomic case, coupling is primarily
through the surface plasmon mode (SPP). Collective excitations in nanoparticle
arrays have been previously reported, and for large interparticle spacing, weak
coupling occurs through radiative routes[101]–[103]. On the other hand, when MIMs
are arranged in a closely packed periodic fashion, MIM modes and the LFP modes
strongly couple due to overlapping modes, rather than due to radiative coupling,
resulting in a diatomic crystal (Figure 2-6d). The coupled MIM waveguides could
also be viewed to be similar to a resonant guided wave network
(RGWN)[104]; however, coupling of waveguides through an intermediate localized
resonance in the vertical cavity distinguishes the MCMP from the RGWN. The
coupling of the MIM and LFP modes results in an anticrossing behavior, as shown
in Figure 2-6e, f. In numerical calculations, it is seen that anti-crossing-over behavior
due to coupling of MIM modes with the LFP (around h ∼ 17 nm in Figure 2-6e, f)
leads to an improvement of the absorption bandwidth.
The angular dependence of coupling to the resonances is best visualized in the
band structure of the plasmonic meta-surface. In order to intuitively understand the
dependence of the band structure on design parameters, we construct a circuit model,
as shown in Figure 2-6g. The MIM section is modeled using a transmission line, and
the LFP is modeled using an LC resonator. A capacitor is used to model the
coupling. The one-dimensional band structure is calculated by constructing an
ABCD matrix for the lumped element model of the unit cell. In order to obtain the
band structure, the ratio of the input and output parameters of the ABCD matrix are
assumed to be exp(ikΛ) for a given propagation vector k and period Λ. In order to
find the resonance frequencies, one needs to solve for ω through
𝐴 𝐵 𝑖𝑘Λ
|𝐼 − [ ]𝑒 | = 0 (2.26)
𝐶 𝐷
where I is the 2 × 2 identity matrix.
17
The ABCD matrix of the unit cell is constructed by multiplying ABCD matrices
for the transmission line representing the MIM section with ABCD matrices for the
lumped elements for coupling and LFP sections of the model. Resistive or radiative
losses are ignored in the model. The ABCD matrix for the transmission line section
is given by
cos(𝛽𝑙) 𝑗𝑧𝑜 sin(𝛽𝑙)
𝐴𝐵𝐶𝐷TL = [ ] (2.27)
𝑗𝑧𝑜 sin(𝛽𝑙) cos(𝛽𝑙)
The coupling capacitor has an impedance of Xc = 1/(jωCcoupling). The ABCD

matrix for the coupling capacitor is given by
1 𝑋𝑐
𝐴𝐵𝐶𝐷coupling = [ ] (2.28)
0 1
The parallel LC resonator representing the localized Fabry–Perot (LFP) mode has
an impedance XLFP given by XLFP = 1/(1/(jωLFP) + jωCFP). The ABCD matrix for the
LC resonator representing the mode is given as
1 0
𝐴𝐵𝐶𝐷LFP = [ ] (2.29)
1/𝑋𝐿𝐹𝑃 1
The ABCD matrix for the unit cell, ABCDuc, is calculated by multiplying the
matrices in correct order, ABCDuc = ABCDTL × ABCDcoupling × ABCDLFP × ABCDcoupling.
The resonance frequencies calculated by the matrix method are shown on top of
numerical results in Figure 2-6e,f. The model correctly captures the essential features
of the numerical results, such as mode frequencies and anticrossing behavior.
18
Figure 2-6. (a) Cavity resonance (LFP) can be thought as corresponding to a light atom, with
a single localized state, and (b) MIM can be thought as corresponding to a heavy atom, with
a multiple localized states. (c) In a one-dimensional periodic arrangement of the light atoms,
coupling through the surface plasmon mode can introduce a coupling of the LFPs and
propagating bands can be formed. (d) Similarly, in a one-dimensional periodic arrangement
of the MIM structures with a relatively thick top metal and a thin gap dielectric, the MIM
modes and the LFP mode formed between consecutive MIMs are hybridized to form a
diatomic molecule. The molecules are coupled to form the band structure of the diatomic
crystal. (e) Effect of mode coupling is observed in the reflection coefficient, plotted as a
function of wavelength and dielectric thickness, h, for a one-dimensional crystal. Numerical
computation results are shown for a 250 nm period structure with 180 nm MIM width, for
normal incidence. Resonance wavelengths are calculated using the Fabry–Perot resonator
model for the MIM structures (black lines; see text for details) that are superimposed.
Antisymmetric modes (even mode number m) are not coupled to the free-space modes at
normal incidence and, therefore, contribute no absorption (arrow (i)). The LFP mode
wavelength is estimated using the LC model (vertical line around 520 nm, denoted by arrow
(ii)). (f) Reflection coefficient is plotted as a function of wavelength and dielectric thickness
h at a 10° angle of incidence. Due to the broken symmetry, even modes also contribute to
absorption. (g) Lumped circuit model for the diatomic unit cell consists of a transmission
line capacitively coupled to a localized LC resonator representing the LFP mode. The model
is used to calculate the resonance frequencies of the modes, plotted as dots, on top of the
computational results in (f). The model captures the essential features of numerical
calculations such as the resonance frequencies and anticrossing behavior due to coupling of
modes.
19
In order to illustrate the effect of geometric parameters on the optical resonances
and show the quasi-omni-directional nature of the MCPM, we calculate the bands of
several structures, as shown in Figure 2-7a–e. The monatomic crystal (no dielectric)
is essentially a lamellar grating (cavity depth 50 nm, width 50 nm, period 250 nm),
and the LFP mode couples through the SPP to form a CROW-like band (Figure
2-7a). As the dielectric gap is introduced, MIM modes emerge in the reflection
spectrum. For a thin dielectric of h = 3 nm, a large number of modes that span the
visible spectrum emerge (Figure 2-7b). Consecutive bands with odd and even MIM
modes show improved absorption at low and high angles due to their symmetry (a
similar observation was reported for MIM resonator arrays in the terahertz frequency
range[92]. The band structure, which is calculated by solving Equation 2.26, displays
bands that coincide with reflection minima seen in numerical calculations. As the gap
is increased to 12 nm (Figure 2-7c–e), the number of bands is reduced. The effect of
the surface plasmon mode becomes visible at high frequencies and is included in
band structure calculations as a perturbation. The coupling of the SPP to the
MIM/LFP structure has the effect of pushing and concentrating the bands to around
500 nm at high angles (kΛ/π ∼ 1) in Figure 2-7a–e.
Figure 2-7. (a) Band structure of a monatomic meta-surface is calculated using the circuit
model in Figure 2g and the ABCD matrix approach and superimposed on the reflection
obtained through numerical calculations. When the dielectric thickness is zero, only the
bands formed are through coupling of the LFP (arrow I) modes via the surface plasmon
mode (arrow II) are observed. (b) Increasing the gap to 3 nm results in the appearance of a
large number of MIM modes and formation of a diatomic meta-surface. (c) Increasing the
gap to 13 nm results in a reduction of the number of MIM modes, while increasing the
absorption in the coupled bands due to improved impedance matching of the surface with
free-space propagating modes. Further band structures are shown in (d) and (e) for 180 and
160 nm MIM widths, with 250 nm period. Color bar shows corresponding reflectance
values.
20
2.3.4. Experimental Results
The absorption spectra of the surfaces have been recorded at normal incidence
using a low numerical aperture (NA = 0.05) objective for both illumination and light
collection. The reflection spectra are seen to agree with simulation results for various
MIM widths and periods (Figure 2-8a, b). Evidence for the quasi-omni-directional
character of the one-dimensional structures can be seen in the high spatial resolution
reflection maps shown in Figure 2-8c (average reflectance for a 10 nm spectral band
centered at 550 nm is shown). When the structures are imaged by different numerical
aperture lenses (objectives with 20×, 50×, and 100× magnification and respective
numerical apertures of 0.4, 0.7, and 0.95), the absorption is seen to be quite
independent of the numerical aperture. This is in accordance with the expected quasi-
omni-directional character of the band structures around 550 nm.
Figure 2-8. Theoretical and experimental reflectance plotted as a function of wavelength at

normal incidence for (a) one-dimensional, 250 nm period structures with nanowire widths of
180 nm (I), 170 nm (II), 160 nm (III), 150 nm (IV), 140 nm (V) and (b) one-dimensional,
300 nm period structures with nanowire widths of 200 nm (I), 190 nm (II), 180 nm (III), 170
nm (IV), 160 nm (V). (c) TM mode reflectance for a 10 nm band around 550 nm observed
with different numerical apertures (20×, 50×, and 100× objectives) on one-dimensional 250
nm period 150 nm width structures. Insensitivity to numerical aperture demonstrates the
quasi-omni-directional absorption of the surfaces.
For most applications in sensing and spectroscopy, the locations with high field
enhancements must be exposed to allow for adsorption of molecules. Such regions
are present in the MCPM (LFP region and top surface), where the enhancement can
be utilized for sensing or Raman spectroscopy. It should be noted that absorption and
field enhancement are not necessarily proportional. However, numerical calculations
show that if the resistive losses can be ignored, there is a strong correlation between
the absorption and field enhancement. Typically, this is the case when the excitation
21
frequency is reasonably below the plasma frequency of the metal and absorption is
enhanced primarily due to plasmonic resonances. In our wavelength range of interest
(400 nm to NIR), dielectric constant of silver allows this approximation. The MCPM
can feature multiple absorption bands (and correlated field enhancement) that cover
the entire excitation and scattering wavelengths (for example 250 period, 50 nm top
metal, 12 nm dielectric thickness, and 200 nm top metal width). The advantage of the
MCPM structures in plasmon-enhanced Raman spectroscopy is demonstrated using
Cresyl Violet dye as the example molecule. Using low powers of about 100–300 μW
(measured by a placing a photodiode at the sample location in a separate
measurement), the Raman spectra are recorded at various locations on the
unpatterned MIM substrate (Figure 2-9a, arrow III) and on the MCMP (Figure 2-9a,
arrows I and II). The unpatterned MIM reference produces little observable Raman
or fluorescence signal, whereas the MCPM produces a pronounced enhancement,
which is theoretically estimated to be on the order of 5 × 105 to 106. The TM
polarized reflectance (10 nm band around 550 nm, Figure 2-8a) has a slightly non-
uniform spectral response over a distance of about 50 μm, attributed to the exposure
nonuniformity during the e-beam lithography. Raman signal map (591 cm–1 peak)
collected at the same location is shown in Figure 2-9b. A spatial non-uniformity is
also seen in the Raman signal (Figure 2-9b), correlated with the reflectance non-
uniformity (Figure 2-9a). As expected, enhancement of the Raman signal is
proportional to the absorption within the Raman band wavelength range (Figure
2-9c,d). When a longer averaging time of 22 s is used, still the Raman signal is
absent on a flat Ag reference surface for the low power level used (Figure 2-9e). The
fabrication related non-uniformity is insignificant over smaller length scales, and we
superimpose the Raman spectra collected from 1600 locations within a 10 μm × 10
μm square region in Figure 2-10a. The fluorescence and Raman enhancements are
uniformly enhanced (about 10% for fluorescence and about 20% for Raman) over the
region, as is seen in the histograms (Figure 2-10b) shown for two wavelengths. When
the Raman signal is represented as an image map, several defects with
submicrometer diameters are clearly resolved as dark spots in the image (Figure
2-10c), indicating true high-resolution imaging capability using the MCPM (image
contrast is enhanced to clarify the dark spots).
22
Figure 2-9. (a) Reflectance map (540–600 nm band) acquired using a 20× objective (NA 0.4)
on a 250 nm period structure with 200 nm MIM width. (b) Raman spectra map (intensity of
591 cm–1 band, collected with 100 μW excitation power at 532 nm, 20× objective, 100 ms
dwell time per pixel) of Cresyl Violet monolayer on the same structure. Scale bar is 10 μm.
(c) Reflectance as a function of wavelength for several locations and (d) corresponding
Raman spectra. Reflectance is plotted at two different locations (I) and (II) as referenced to
location (III). Due to non-uniformity of the fabrication process, a gradient of the resonance
wavelength is observed from top-right position to bottom-left position. When the absorption
overlaps with excitation and emission wavelengths, improved Raman scattering is observed.
Inset shows chemical structure of Cresyl Violet. (e) Raman signal is collected using a longer
integration time (22 s, 100 μW excitation power) on a planar silver surface, unpatterned
MIM, and 250 nm period MIM regions. Although Cresyl Violet exhibits no Raman signal on
the plane metal surface, some enhancement is seen on unpatterned MIM regions, possibly
due to the surface roughness of the top layer.
23
Figure 2-10. (a) Superimposed Raman spectra collected from 1600 individual spots over an
area of 10 μm × 10 μm (100 μW excitation power, 100× objective, and 40 ms dwell time per
pixel). (b) Histograms of intensity of two spectral locations shown by arrows (I) and (II)
demonstrate uniform signal intensity within ca. 10% of average value for the fluorescence
(arrow II) and ca. 20% for the Raman signal (arrow I). (c) Raman map formed using the 591
cm–1 Raman band (scale bar 2 μm) where the contrast is enhanced to show several dead-
spots with submicrometer dimensions, demonstrating the high-resolution imaging capability
with such substrates.
2.3.5. Discussion of Spatial Uniformity of Enhancement at the Nanoscale
Due to the periodicity of the meta-surface, Raman enhancement appears to be

uniform in the far field when the diffraction-limited spot size is larger than the meta-
material period. However, it is important to understand the variations of local
enhancement factor within a unit cell. We plot the average and maximum
enhancement factors for three different geometries, as shown in Figure 2-11, for
MIM periods of 50, 100, and 250 nm. The averages are taken over rectangular
sections over the top surface (10 nm thick slab is chosen), within the LFP region (air
gap region between the top metals) and inside the MIM gap, within the dielectric
(see Figure 2-11c inset). The dielectric gap of the MIM structure does not contribute
to the Raman signal since molecules cannot be placed there after fabrication.
However, enhancement in this region may be important for other applications;
therefore, we plot enhancement factors for this region for convenience. The
correlation of EF values for the MIM and LFP regions also demonstrates the
coupling of LFP and MIM resonances. When comparing average and maximum EF
of different regions and geometries, it is seen that for smaller structures (50 nm
24
period, 30 nm width, Figure 2-11a), a single resonance is present in the wavelength
range of interest (532 to 640 nm), and average field enhancements at the top surface
are greatly improved compared to 100 nm structures (Figure 2-11c) and 250 nm
structures (Figure 2-11e). It is also seen that the LFP region has a stronger average
enhancement factor as compared to the top surface and MIM gap. The maximum
enhancement factors can be relatively large (EFexc ∼ 1000, EFSERS ∼ 106) for both
small and larger periods (Figure 2-11b, d, and f). From the data, it can be concluded
that, despite the apparent uniform enhancement in diffraction-limited far field
measurements, enhancements are mostly localized to the LFP region for larger
structures. For smaller structures (∼50 nm period), although there is local non-
uniformity and the LFP region still contributes the greater portion to the enhanced
signal, the average EF values for the top surface and the LFP region have improved
ratio, indicating improved uniformity.
The MCPM features propagation of multiple plasmon modes in the visible range
of the spectrum, through direct capacitive coupling or coupling via a localized
resonant mode. The simplicity of the MCPM allows fabrication of an ultrathin meta-
surface with tailorable bands at wavelengths toward the near-UV. Bandwidths are
comparable to thicker wideband absorbers[105]. Due to the quasi-omni-directional
coupling of light to the plasmon modes, high spatial resolution SERS imaging is
possible. The diffraction-limited spot size is greater than the period of the meta-
material, and plasmonic enhancement of the Raman signal essentially occurs with a
unity surface fill factor. High-resolution imaging capability can potentially improve
biomolecular sensing[106], [107]. Raman excitation wavelength and scattering
wavelengths coincide with different bands. Advancing the analogy of the meta-
material with a semiconducting crystal, the Raman transition can be regarded as a
photonic transition between two bands, where the transition is accompanied with the
release of a phonon. Raman scattering constitutes an example for nonlinear optical
phenomena, and in general, the MCPM can find application in cases where enhanced
nonlinearity at low optical powers is desirable, such as enhancement of photonic
interactions in all-optical classical or quantum information processing.
25
Figure 2-11. (a) Enhancement factors (EF) averaged over various regions (see inset in c) for
a MIM structure with 50 nm periodicity, 10 nm top metal thickness, 20 nm dielectric gap,
and 30 nm top metal width. (b) Maximum value of EF for different regions for the geometry
in panel a. (c) Average EF values for a MIM structure with 100 nm periodicity, 20 nm top
metal thickness, 20 nm dielectric gap, and 80 nm top metal width. (d) Maximum value of EF
for different regions for the geometry in panel c. (e) Average EF values for a MIM structure
with 250 nm periodicity, 50 nm top metal thickness, 20 nm dielectric gap, and 230 nm top
metal width. (f) Maximum value of EF for different regions for the geometry in panel e. It is
seen that shrinking the MIM size results in fewer resonances and improved average
enhancement over the unit cell, especially on the top surface (10 nm thick slab over the top
metal). The maximum values of the EFs are much higher than the average, showing the
inherent spatial non-uniformity of enhancement. Comparing panels a–c, it is seen that tuning
of the resonances through choice of geometry greatly improves EF for the wavelength range
of interest (532 to 650 nm).
26
2.4. Edge Rounding Induced Broadband Plasmonic
Metamaterial Absorber Structures
Perfect absorption of light can be achieved by exciting surface plasmons on
textured metallic surfaces[108]. Metamaterials (MMs) are known as nanostructured
metal surfaces to tailor the optical properties. So far, MMs are used to design
negative index materials[47], optical sensor platforms[109], surface enhanced Raman
spectroscopy substrate[110], solar cells[14], and optical bolometer imaging
systems[111], [112]. Metamaterial absorbers (MMAs) are arrays of patterned metal
insulator metal structures that have almost perfect absorption for certain wavelengths
where impedance of MMAs are perfectly matched to the air’s impedance near the
resonant wavelengths. Nearly perfect MMAs are demonstrated in the
microwave[113]–[115], terahertz[116], [117], infrared (IR)[21], [35], [97], [118]–
[123] and visible frequencies[124]–[130]. Metamaterial absorbers in the visible
spectrum can also be realized by using nanoparticles separated from a thick metallic
ground plane with a dielectric spacer[131]–[133]. Typically, MMAs have narrow
bands or single resonances attributed to their identical resonators in every unit
cell[127], [128], [134], [135]. On the other hand, multispectral characteristic of a
metamaterial plasmonic surface has great potential for various applications such as
infrared absorption spectroscopy and surface enhanced Raman spectroscopy where
the presence of resonances are generally accompanied by high field
enhancements[63]. Recently, using more complex geometries broadband or
multispectral MMAs are demonstrated in the visible and IR wavelengths.
Multiplexing different resonator structures in a unit cell is another way to obtain
multispectral absorption which requires precise control of dimensions of multiple
nanostructures[114], [136]–[142]. However, multispectral MMAs based on the
simultaneous excitation of electrical and magnetic plasmon modes is the least
exploited approach[126], [143], [144]. MMAs are usually fabricated by using
electron beam (e-beam) and optical lithography. The fabrications of nanostructures,
however, are typically accompanied by process imperfections. Disc structures
become more like truncated cones due to finite sidewall angle and square structures
have rounded corners due to lithographic nonidealities. These fabrication related
features are exploited in our MMA structure which has already multispectral
absorbance bands due to the excitation of both electrical and magnetic plasmon
27
modes. Particularly, when corners of the top nano-square patches are rounded, a new
resonance is observed to emerge. The proposed structure takes into account the
imperfections of common fabrication techniques and is therefore more realistic.
Unlike the multiplexed metamaterial structures having multiple resonators with
different geometries in a unit cell, the multiband plasmonic structure studied here
only requires the control of width of the structures to obtain multispectral and
broadband response.
Figure 2-12. SEM images of polarization independent plasmonic metamaterial surface. The
fabricated structures are disc-shaped despite they are aimed perfect squares.
28
2.4.1. Simulation Geometry and Method
The schematic description of the proposed structure is shown in Figure 2-13. The
structure is based on metal/insulator/nano-square array geometry. Silver is used for
both bottom metallic plane and top nano-square arrays. Alumina (Al2O3) is used as
the dielectric spacer layer. The thickness of the bottom (ground) plane is 100nm for
all the simulations where transmission is almost zero in the visible range. The
thickness of the top layer is chosen 50nm, allowing the coupling of localized surface
plasmon modes due to the interaction of adjacent square patches. Periods of square
patches along x, y directions are 250nm. The simulations are based on the Finite
Difference Time Domain (FDTD) method (Lumerical FDTD). The material
properties (complex and real parts of dielectric function) for silver and alumina are
obtained from the program's material database and fitted to a 12 order function (Palik
data). The mesh size used in FDTD analysis is 2nm for both x and y directions and
1nm for z direction. A broadband electromagnetic plane wave for normal incidence
illumination is used with 400nm-750nm wavelength range. Bloch boundary
conditions are chosen for x and y directions and perfectly matched layer (PML) is
used along the z direction. Simulation durations are further decreased by using
asymmetric and symmetric boundary conditions along x and y directions.
Figure 2-13. (a) Schematic description of the proposed MMA structures. Top (b) and side
(c) view of simulated structures. Periods in x and y directions are 250nm. W is width of the
patches and R is the curvature of the corners.
29
2.4.2. Results
In Figure 2-14a dual band absorber structure is demonstrated when the curvatures
are omitted in the simulations and perfect square patches are assumed (R=0nm).
When there is no oxide in the structures, single resonance at λ=530nm is observed
due to the cavity resonance between the patches. When, a tox= 30nm of oxide spacer
is included, the resonance splits into 2 modes at λ1=480nm and λ2=605nm as shown
in Figure 2-14a. To understand the origin of these resonances electric and magnetic
field distributions are plotted as seen in Figure 2-14c. Without the oxide layer, layer
magnetic field is localized in the grooves and electric field is higher at the top of the
adjacent ridges. This resonance resembles an electric plasmon mode. With the oxide
layer, magnetic field is also confined in the grooves at λ1=480nm. For λ2=605nm,
magnetic field is confined in the spacer layer. Hence, resonances at λ1=480nm and
λ2=605nm are electrical and magnetic plasmon modes respectively. Then, we
investigated effect of the nano-square patch width as shown in Figure 2-14b. The
resonance at shorter wavelength is unaffected with increasing W which also suggests
that the resonance is electric plasmon mode[126], [145]. However, as the width
increases the resonance at higher wavelength shifts to longer wavelengths which
verifies a magnetic plasmon mode. This resonance can be explained in terms of
standing waves by 𝜆𝑟 = 2𝑛𝑒𝑓𝑓 W + 𝜃 where neff is the effective index of MIM mode,
θ is the phase term due to reflections from the edges. Resonance wavelength linearly
depends on W[146].
30
Figure 2-14. (a) Absorption spectra of the MMA structure with perfect nano-square patches
(R=0nm) for tox= 0 (blue) and tox= 30nm (green) (b) Absorption spectra for various nano-
square patch widths (tox= 30 nm). Magnetic (c) and Electric (d) field profiles at the
resonances shown in (a).
When the structures are fabricated using electron beam lithography, due to the
finite point-spread function width of the exposure process, the corners of the square
patches get rounded. In order to demonstrate the effect of round corners in the
absorption spectra, curvature is added to the corners of nano-square patches in the
simulation (See Figure 2-13). Figure 2-15 shows the emergence of a third mode
whose resonance frequency is affected by the radius of curvature of the corners. For
W=200nm, R=0 nm up to R=40nm there are two distinct resonances as shown in
Figure 2-15a, c. When the curvature is further increased, the magnetic plasmon mode
splits into two. For R=100nm where the nano-square patches become perfect discs,
three resonances are observed with almost equal wavelength spacing. The absorption
of each resonance is higher than 85%. For a small change of the width of the patches,
to W=210nm, the magnetic plasmon mode splits into two at a smaller curvature
value (R~30nm). The wavelength spacing between the resonances is higher for the
maximum curvature value (R=105nm) and, near perfect absorption is seen for the
middle resonance around 550-600 nm. We further investigated the effect of dielectric
thickness on the absorption of triple band plasmonic structures as shown in Figure
31
2-15 e, f. The electric plasmon mode at λ~500nm is almost independent of the spacer
thickness as expected. However, the magnetic plasmon modes strongly depend on
the spacer thickness which is due to the change in effective refractive index of the
MIM waveguide formed by the top patch and bottom Ag layer, with the increasing
spacer thickness.
Figure 2-15. Effect of the curvature on the absorption spectra for W=200nm (a,b) and
W=210nm (c,d) for tox=30nm. Effect of the oxide thickness on the absorption spectra for
W=200nm and R=100nm (e-f). The magnetic plasmon mode splits into two modes at smaller
radius as the width of the resonator increases.
In order to understand the origins of the resonances for the perfect nano-disc top
layer (R=100nm, W=200nm), we have plotted the electric field profiles at λ=550nm
and λ=600nm. The analytical approaches for circular nano patch antennas have been
recently demonstrated[147], [148]. Following the steps shown in previous literature,
the Helmholtz equation for metal-insulator-metal (MIM) geometry is solved in the
cylindrical coordinates which is
𝜕 𝜕𝐸𝑧 1 𝜕 2 𝐸𝑧 𝜕 2 𝐸𝑧
(𝜌 ) + 2 (𝜌 2
)+ 2
+ 𝑘 2 𝐸𝑧 = 0 (2.30)
𝜕𝜌 𝜕𝜌 𝜌 𝜕𝜑 𝜕𝑧
The solution to Equation 2.30 is given by 𝐸𝑧 = 𝐽𝑚 (𝑘𝑠𝑝 𝜌) cos(𝑚𝜑 + 𝜗) (𝐴𝑒 𝑘𝑧 𝑧 +

𝐵𝑒 −𝑘𝑧 𝑧 ) with Neumann boundary condition 𝜕𝐸𝑧 /𝜕𝜌=0 where 𝑘𝑠𝑝 is the wave-vector
for the MIM cavity. The solution to electric field distribution is given in terms of
even and odd modes as E (O)mn where n is the order of the roots for 𝜕 (𝐽𝑚 (𝑘𝑠𝑝 𝜌)) /
32
𝜕𝜌 =0. For odd and even values of m odd and even order modes are excited,
respectively. The simulated and calculated mode profiles are shown in Figure 2-16.
For λ=550nm which is due to the corner curvature effect, O12 mode is excited as
shown in Figure 2-16a, c while the magnetic plasmon mode at λ=600nm resembles
O31 mode after the corner curvature effect (Figure 2-16 b, d).
Figure 2-16. Origin of resonances for W=200nm, R=100nm and tox=30nm. Simulated z
component of electric field profiles for λ=550nm (a) and λ=600nm (b). Calculated electrical
field profiles for modes m=1, n=2 (c) and m=3, n=1 (d).
Fabrication of MMAs requires lithography and liftoff processes which results in

slanted sidewalls of patterned surfaces. Figure 2-17a shows the cross section of a
typical MMA structure. To demonstrate the effect of the side angle on the absorption
spectra of the triple band absorber structure, we simulated a truncated cone shaped
top metallic layer rather than a perfect disc (corresponding to sidewall angle θ=90).
Absorption spectrum of such a structure is shown in Figure 2-17b, c. As the side
angle decreases down to θ=75 degrees we can observe a resonance at λ=600nm, but
for smaller side angles this resonance diminishes. A broad absorption band covering
400 to 550 nm can be observed for smaller sidewall angles. For θ=85 degrees or
33
higher, we still observe triple bands. Such a structure with high side angles can be
fabricated using inductively coupled plasma or reactive ion etching , after e-beam
lithography on MIM films .
Figure 2-17. Effect of side angle on the absorption spectra. (a) Cross section of the simulated
structure. (b-c) Absorption spectra with respect to the side angle. As the slope of the cone
increases absorption of magnetic resonance mode decreases.
2.5. Conclusion
In this chapter 1D and 2D periodic coupled plasmon meta-surfaces are presented.
For the 1D case we have developed a circuit model to understand the coupling
mechanisms of different modes. Due to wide-band width of these surfaces, Raman
signals can be amplified for both excitation and stokes wavelengths. The proposed
geometry is also subwavelength (Period=250nm) which results in unity Raman
enhancements over large areas. Moreover by utilizing the analogy of heavy and light
molecule analogy, Raman transition can be seen as a photonic transition between two
bands. For the 2D case, we demonstrated a triple band plasmonic metamaterial
surface due to edge rounding effects occurred during the fabrication process. The
emerging third resonance is attributed to the excitation of circular cavity modes.
34
Chapter 3
Nano-particle Based Large Area
Broadband Plasmonic Surfaces and
Their Applications
In this chapter we present a large area plasmonic surface based on metal-insulator-

nano-particle geometry which supports localized surface plasmon modes over large
bandwidths. The broad bandwidth is due to excitation of the electric and magnetic
plasmon modes. The electric plasmon mode is due to inter-particle coupling. Field is
confined to the insulator layer for the magnetic plasmon mode. We used these
surfaces for single molecule Raman spectroscopy and as hot-electron detectors. For
the first time, we have observed single molecule Raman events using a mobile
phone's camera. Moreover, we have also developed a contact free characterization
technique using x-ray photoemission spectroscopy (XPS) by probing hot-electrons
emitted from individual metal-insulator-nano-particle photodetector.
Parts of this chapter were published as;
1. Sencer Ayas, Andi Cupallari,Yasin Kaya, O.O. Ekiz, Aykutlu Dana,"Counting
Single Molecules With a Mobile Phone Camera Using Plasmonic Enhancement",
ACS Photonics,1,1,2014
2. Sencer Ayas(*), Andi Cupallari(*) and Aykutlu Dâna,"Probing Hot-Electron
Effects in Wide Area Plasmonic Surfaces using X-Ray Photoelectron Spectroscopy",
Appl. Phys. Lett.,105, 221608 (2014) (* equal contribution).
35
3.1. Large Area Plasmonic Metasurfaces
We use the same geometry we have developed in the chapter 2 for the
applications presented here. However, fabrication of such surfaces requires top-down
fabrication techniques such as electron-beam lithography and focused ion beam
(FIB). We have developed a large area plasmonic surface by dewetting of thin Ag
films over insulator/metal surfaces. Substrates fabricated by dewetting of silver films
were studied before for their plasmonic properties and SERS enhancement[149],
[150]. Metals near percolation threshold have localized surface plasmon resonance
due to their nanoparticle nature. By changing fabrication conditions, resonances of
the surface can be tuned. Placing metal nanoparticles over a metal surface, with an
dielectic separator in between, results in strong changes in overall plasmonic
properties due to the interaction of metal nanoparticles with the metal film[44],
[151]–[153]. It is seen that Ag nanoislands form at a mass thickness of 1 nm,
coarsening takes place at 3 nm and percolation begins at 5 nm mass thickness (Figure
3-1a). For 3 nm top metal mass thickness, the gap between individual nanoparticles
is about 5-10 nm and nanoisland thickness is 10 nm on average. HfO2 thickness of 30
nm and Ag nanoisland layer mass thickness of 3 nm produce a wide band perfect
absorber as measured by spectroscopic ellipsometry (Figure 3-1c, d). The
metasurfaces are quasi-omnidirectional and absorption is above 90% for the 30 nm
thick HfO2 sample over a wavelength range of 450 to 800 nm, for angle of incidence
of up to 60 degrees (Figure 3-1e).
36
Figure 3-1. (a) SEM of plasmonic surfaces with 1, 3 and 5 nm mass thickness Ag overlayer
shows coarsening and percolation of Ag nanoislands. Scale bars 250 nm. Plasmonic field
enhancement is greater as the nanoislands approach each other, reducing the inter-particle
gap. 3 nm mass thickness sample exhibits greatest hot spot density. (b) Schematic view of
the substrate showing layer structure. (c) Reflectance of the surfaces near normal incidence
for 1, 3 and 5 nm mass thickness Ag over-layer and 30 nm HfO2. Gray band shows the
wavelength region of interest for Raman scattering excited by 532 nm light. Ag mass
thickness of 3 nm results in a wide band meta-surface. (d) The reflectance is plotted for HfO2
thicknesses of 5, 10, 20 and 30 nm. (e) Dependence of reflectance on angle of incidence is
plotted for 20, 30, 40, 50, 60, 70 and 80 degrees. The 30 nm HfO2/ 3nm Ag surface is quasi-
omni-directional, maintaining high absorption over a wide wavelength range at angles up to
60 degrees.
Electromagnetic properties are calculated for periodic nano-islands with 75 degree

sidewall angles (Figure 3-2a, b), as well as for SEM data based model surfaces
(Figure 3-2c, d and Figure 3-3) using computational tools. Plasmonic metamaterial
surfaces have been demonstrated to have broadband absorption and superior field
enhancements that are fabricated using electron beam lithography[127]. In contrast,
the surfaces presented here require no lithography, hence are easy to fabricate over
large areas. The resonance wavelengths can be tuned by changing the thickness of
spacer layer and by control of island diameter and thickness. Optimized surfaces with
30 nm HfO2 dielectric thickness and 10 nm island thickness have broadband plasmon
resonances over the whole visible spectrum with almost 90% average absorption.
37
Figure 3-2. . (a) Maximum |E|2 field enhancement factor as a function of wavelength,
plotted for 10, 20 and 30 nm dielectric thickness for a periodic arrangement of Ag
nanoislands (40 nm period, 10 nm thickness, 35 nm island size, 75 degree sidewall
angle), shows increase in the enhancement around 550 nm as thickness increases to
30 nm. (b) Calculated reflectance of the periodic arrangement, plotted as a function
of wavelength. (c) Maximum |E|2 field enhancement factor as a function of
wavelength, for a quasi-random surface derived from SEM data as a function of
dielectric thickness. (d) Calculated reflectance of the quasi-random arrangement,
plotted as a function of wavelength and dielectric thickness. A dielectric thickness of
40 nm is also included in the calculations, as it better fits the experimental
reflectance for 30 nm HfO2 shown in Figure 3-3.
38
Figure 3-3. SEM derived surface model for field enhancement and reflectance calculations.
a) The SEM data is used to extract silver island shapes. b) Field enhancement factor (|E|2 for
550 nm excitation) at the top metal-dielectric interface. c) Field enhancement in the middle
point of upper boundary of the top metal and dielectric interface. d) Field enhancement at the
upper boundary of the metal islands.
Figure 3-4. (a) Cross sectional magnetic field profile for a periodic arrangement of metal-
insulator-metal resonators (40 nm period, 35 nm width, 20 nm thickness, 75 degree sidewall
angle Ag, on 20 nm HfO2, on Ag) at 430 nm excitation wavelength and (b) at 700 nm
excitation wavelength. (c) For a top metal thickness of 10 nm, fields have enhancement at
the top surface (excited at 550 nm) as compared to (d) a top metal thickness of 20 nm (all
scale bars are 20 nm wide).
39
The broadband absorption is due to both the inter-particle and particle-metal film
couplings. Mode confinements are shown for a periodic nano-island arrangement as
shown in Figure 3-4. Inter-particle coupling is attributed to a dipolar coupling
between individual nano-particles, which is known as electrical resonance where
magnetic field is confined between the nano-particles (Figure 3-4a). The particle-
metal film coupling, typically referred to as the magnetic resonance has the magnetic
field confined to the spacer layer (Figure 3-4b) [94]. A coupling of the two types of
resonances is also present and enhances the absorption in the intermediate
wavelength region between the two resonances. The thicknesses of the Ag islands are
important for having an optimal enhancement (Figure 3-4c, d). For single molecule
SERS enhancement, the surfaces need to be optimized, and optimal SERS
enhancement is achieved when plasmonic resonances cover the range of excitation
and scattering frequencies[154]. We optimized the SERS enhancement at excitation
and scattering wavelengths (532 to 650 nm wavelength range, corresponding to
Raman range of 0 to 3500 cm-1) by sweeping the fabrication parameters. The
dielectric layer is very important in achieving the enhancement levels presented here,
and provides about an order of magnitude improvement. This effect is studied
recently by independent groups, using SiO2 as the spacer[30].
3.2. Single Molecule Raman Events Using Large Area

Plasmonic Surfaces
Plasmonic structures can enhance the Raman signal significantly, thereby
increasing the signal from single or few molecules to levels that can be detected
using a cooled CCD spectrometer or a photomultiplier. In order to characterize the
spectral content of the single molecule blink events, we record the time dependent
Raman spectra on a bright hot spot, using a cooled CCD spectrometer, as shown in
Figure 3-5. The integrated intensity between 0-3500 cm-1 is plotted as a function of
time in Figure 3-5a. The integration time of the spectrometer is 100 ms. Sudden
changes of the spectrum are observed as peaks in the time series. The spectra show
distinct Raman bands which fluctuate in both intensity and frequency, which is
commonly interpreted as a positive indication for single molecule level sensitivity
(Figure 3-5b). The debate about the interpretation of fluctuating Raman bands as
40
evidence for single molecule SERS is ongoing in the literature[155]–[159]. It is
claimed that, thermally activated diffusion and Brownian motion of molecules are
possibly responsible for the observed SERS signals[160]–[162]. Fluctuations of
intensity of optical emission from silver nanoparticles were claimed to be
independent of the intentional presence of probe molecules but an inherent feature
related to Ag nanoparticles[163]. However, we note that in our experiments, when
untreated surfaces are subjected to airborne molecules, for example by spraying a
mixture of carbon compounds from a fragrance bottle, a sudden increase of blinking
is observed. Diffusion of large molecules on metal surfaces has been directly
observed and characterized using scanning tunneling microscopy[164]–[166]. Also,
spurious lines are routinely present in blinking SERS spectra[167]. Based on such
previous observations, we attribute the unidentified spectra to the presence of
adsorbed volatile organic compounds present in ambient air[168], [169]. It is
observed that the plasmonic substrates also enhance the fluorescence signal, and
once in a while a broad fluorescence spectrum is captured as shown in Figure 3-5c.
Such enhancements of the fluorescence were previously studied using plasmonic
antennae[170]. The optical spectrum during most blink events exhibit Raman bands,
shown in Figure 3-5d, typically superimposed on a broad fluorescence background
from the Ag nano-island layer. In comparison, no signal can be observed on a flat Ag
surface.
41
Figure 3-5. (a) Time series of the integrated intensity within 0 to 3500 cm-1, as recorded by
the cooled CCD spectrometer on a particularly bright hot-spot, shows blinking events as
fluctuations in the intensity (500 µW of excitation at 532 nm). (b) The time series plotted to
feature the full spectrum exhibits Raman lines that fluctuate in intensity and frequency,
characteristic of single molecule SERS. (c) Occasionally, a fluorescent molecule is captured
in the hot-spot (at time 35 s in (a)). (d) Typical Raman spectra during a SERS blink (at time
76 s in (a)) exhibits a broad fluorescence superimposed with distinct Raman bands.
3.2.1. Observation of Single Molecule Raman Events Using Smart Phone's

Camera
We have used mobile phone's camera in two configurations. First, we captured the
blink events directly placing the smart phone to the eyepiece of Raman setup as
shown in Figure 3-6.
The blinks are captured using the smart phone camera at 30 fps, and are shown in
Figure 3-7a. The relative enhancement of the substrates as a function of dielectric
thickness and Ag over-layer thickness can be seen in the increase of fluorescence
signal from the Ag nanoisland layer. A dielectric thickness of 30 nm and an Ag over-
layer mass thickness of 3 nm produces the highest blink rate with high blink
intensities, an observation which correlates well with the reflection spectra of the
surfaces. The blinks are counted using QuickPalm and histograms are generated as
42
shown in Figure 3-7b[171]. The pixel noise superimposed on the fluorescence
background causes the QuickPalm algorithm to detect erroneous blink events with
low peak intensity, causing an increased blink count at low intensities.
Figure 3-6. a) Schematic of the measurement set-up for using the camera of a smart-phone
for imaging with a confocal Raman system. b) Photograph of the measurement set-up
showing various components described in (a).
Figure 3-7. (a) Frames from video captures using the smart phone camera on
plasmonic surfaces with 1, 3 and 5 nm mass thickness Ag over-layer and 5, 10, 20,
30 nm HfO2 dielectric layer thickness. Excitation laser is defocused to illuminate an
area 50 µm in diameter. Arrows denote blink events on a fluorescence background of
the Ag nano-island layer. Scale bar is 20 µm wide. (b) Video frames are analyzed to
extract a histogram of blink event intensity for the 3 nm Ag samples at varying HfO 2
thickness. A dielectric thickness of 30 nm produces brightest blink events. As the
bottom Ag layer is removed (dielectric thickness infinite), blink events can still be
observed, however at a decreased rate and intensity. Inset shows the potential source
of blinking, i.e. surface diffusion of physisorbed volatile organic compounds into and
out of hot-spots.
43
Figure 3-8. a) Schematic of the spectrometer configuration. b) Photograph of the
measurement set-up showing various components described in (a).
Second configuration for probing single molecule events is shown in Figure 3-8.
This configuration is simply a spectrometer configuration. The collection fiber output
is collimated using a lens and a transmission grating is positioned before the focusing
lens of the camera. In this configuration, wavelength separation of the light can be
achieved with reasonable linearity. A similar configuration has been recently used to
record the spectra of a white light source and perform label-free biomolecular
detection[172]. The wavelength dependent response is not uniform due to the color
filters on the camera, however we do not attempt spectral equalization and use the
data as it is extracted from the snapshots and video captures. We record the
unenhanced Raman spectrum of silicon and ethanol on silicon using the cooled CCD
spectrometer as shown in Figure 3-9b and compare the results with the smartphone-
44
spectrometer data shown in Figure 3-9c. A 600 lines per mm (lpmm) transmission
grating allows observation on 0th, 1st and 2nd order diffraction orders, as shown in the
insets. A close-up of the 2nd order region is shown in Figure 3-9d, superimposed with
the original Raman spectra shown in Figure 3-9b, convolved with a point spread
function of the smartphone-spectrometer. The resolution (> 180 cm-1 for 532 nm
excitation) is limited by the diffractive power of the grating as well as the diameter of
the fiber (25 µm diameter) and focal length of the collimator (10.99 mm, F220-
SMA-B Thorlabs). The fluctuating Raman spectra can be recorded using the SERS
substrate as shown in Figure 3-10a, where multiple snapshots are recorded
consecutively, using a 600 lpmm grating, with an integration time on the order of 1
sec. Distinct spectral features can be observed during blink events. The spectral
features can still be observed at 30 fps, as shown in Figure 3-10b. In order to
demonstrate the superior signal intensity of SERS, we record the Raman spectra on
plasmonic substrates treated with 10 nM Methylene Blue solution and 1 µM Cresyl
Violet solution in using 100 µW excitation ( Figure 3-10c, d). Corresponding spectra
are also captured using the smart phone camera as shown in Figure 3-10e, f.
Although the fluctuating spectra contain signatures that can be attributed to Cresyl
Violet and Methylene Blue, spurious signals are also present that show presence of
unidentified molecules occasionally producing Raman bands. As SERS spectra are
typically modified as compared to bulk spectra, we do not attempt to identify the
spurious molecules.
Although the experiments were performed on a non-portable microscope system,
the setup does not differ from a fluorescence measurement setup and can be
potentially miniaturized. Use of low excitation powers (0.1 to 10 mW) also favors
the potential for miniaturization. It must be noted that, although electromagnetic
calculations estimate an overall enhancement factor of about 107, the large intensity
of the blinking signals suggest that the chemical or first layer contribution to SERS is
not negligible. Such chemical enhancement can be significant, up to two orders of
magnitude chemical enhancement due to complex mechanisms leading to charge
transfer between the metal and the probe molecule has been predicted[173].
Therefore, the SERS spectra differ significantly from unenhanced Raman spectra,
and should be understood beyond simpler electromagnetic models. On the other
hand, SERS spectra are repeatable in themselves, and the plasmonic substrates that
enable the observations can be fabricated on large area substrates without the need
45
for top-down patterning, therefore can be produced at low cost. The width of the
spectra acquired using a 600 lpmm grating is about 1/8th of the overall field of view.
In principle, by using a grating with higher groove density, the dispersion can be
increased and resolution of the spectrometer configuration can be improved to better
allow discrimination of spectra. However, for using the full field of view of the
camera, additional optics may be required to improve the resolution to about 20 cm -1.
The remarkable sensitivity of the smart phone camera combined with high plasmonic
enhancement of the optical signal, may pave the way for low cost hand-held systems
which can be used in the analytical study of samples at a single molecule level.
Figure 3-9. (a) Configuration for using the smart phone camera as a low resolution
spectrometer. The collection fiber from the Raman set-up is collimated and dispersed
with a transmission grating (300 lines per mm) before entering the camera. (b)
Raman spectra of silicon and ethanol on silicon collected with 13 mW of 532 nm
excitation using the cooled CCD spectrometer are shown. (c) Smart phone camera
recordings of the two orders of the dispersed input light in the snap-shot mode. Solid
blue lines show integrated pixel intensity and dotted red line shows superimposed
Raman spectra convolved with a lineshape function that represents the point spread
function (PSF) of the optical configuration. Insets show actual camera excerpts. (d)
Close-up of the second order diffraction region of the camera output (solid lines)
superimposed with Raman spectra shown in (b) convolved with the PSF of the
configuration.
46
Figure 3-10. (a) False color coded excerpts from series of smart phone camera snapshot
captures during blink events on the plasmonic substrate, in the spectrometer configuration (1
mW excitation at 532 nm). The spectral region is cropped, rotated and stitched for each
frame. Inset shows actual camera color coding of the same data. Integration time per frame is
~ 1 sec. (b) False color coded excerpts from a video sequence recorded at 30 frames per
second. Although video recording is at lower resolution, distinct spectral features during
blink events can be observed (also see Supporting Video). The SERS spectra of plasmonic
surfaces (30 nm HfO2 thickness) were also recorded using the cooled CCD spectrometer,
treated with 10 nM Methylene Blue solution in (c) and 1 µM Cresyl Violet solution in (d)
using 100 µW excitation. Corresponding spectra are also captured using the smart phone
camera as shown in (e) and (f). Insets show excerpts of the region of interest from actual
camera captures in snapshot mode.
47
3.3. Probing Hot-electron Effects in Wide Area
Plasmonic Surfaces Using XPS
3.3.1. Introduction
Plasmonic structures have attracted increasing attention in the recent years,

enabling engineering of optical properties of surfaces[174]–[176]. Hot-electron
effects (HEE) in plasmonic structures have been investigated as alternative
mechanisms for solar energy harvesting and photo-detection[177]–[180]. HEE rely
on transfer of hot carriers from one contact to another through a thin dielectric barrier
and provide a semiconductor free route for the photoconversion. A relevant thin film
structure is the metal-insulator-metal (MIM) configuration, where optical properties
of the surface can be tuned through plasmonic resonances, while separately allowing
engineering of carrier transfer and collection. Hot-electron transfer competes with
thermalization in the conversion of the absorbed light into a DC or voltage[181],
[182]. The transient electron energy distributions of metallic surfaces excited by light
pulses are thermalized in short time scales ranging from 100 fs to 1 ps. Photoelectron
emission with pulsed X-ray, UV, and visible light sources was used in investigation
of non-equilibrium energy distributions of carriers on surfaces[183]–[185]. Such
pulsed or pump-probe measurements provide direct information about carrier
dynamics, however require advanced light sources. X-Ray Photoelectron
Spectroscopy (XPS) was used to investigate surface photovoltaic and
photoconductivity effects in nanocomposite surfaces without using femtosecond or
picoseconds light or X-ray pulses[186]–[190]. Here, we extend such use of XPS to
investigate HEE in plasmonic MIM structures. Laser excitation is used to illuminate
the plasmonic surfaces (Figure 3-11a). The top metal of the MIM structures acts as the
plasmonic antenna (metal nanodiscs or gratings/stripes) that provides enhanced
optical absorption. Plasmonic enhancement leads to more efficient hot-electron
generation. HEE result in surface photovoltages and are observed by comparing
shifts of binding energy (BE) of the top metal islands for dark and illuminated
conditions.
48
3.3.2. Results
A schematic description of the band diagram of a MIM surface under X-ray and
visible light illumination is shown in Figure 3-11b. A top metal layer, Ag, is separated
from a bottom metal layer by a thin dielectric barrier. Under X-ray illumination, the
photoelectrons that escape from the top metal islands result in a positive current (Jx),
charging the surface positively (Figure 3-11a). As a consequence, band bending
occurs as shown in Figure 3-11b. In the absence of a compensating electron gun, the
surface potential reaches steady-state due to the presence of direct (or trap assisted)
tunneling current from the bottom metal to the surface, denoted by J t in Figure 3-11b.
When the surface is illuminated, absorption takes place at the top and bottom metals
and the hot-electrons acquire an energy distribution, as shown by rectangles in Figure
3-11b on both sides of the junction. The electron mean free path is on the order of
20–30 nm, which is longer than the dielectric thickness. Therefore, hot-electrons with
energies greater than the barrier height can be partially transmitted towards the top
metal island where they are thermalized. This causes a negative shift of the observed
BE of Ag. In contrast, hot-electrons generated in the top metal layer are not able to
tunnel from the positively charged island.
Figure 3-11. (a) Cross-section of the MIM surface during an XPS measurement with laser
illumination. Jx represents the X-ray induced photoemission current into the vacuum. (b) The
band diagram of the MIM junction under X-ray and laser illumination. Hot-electron energy
distributions are shown with rectangles above the Fermi level, due to absorption of light in
the top and bottom metal layers. (c) A simplified circuit model with various currents acting
on the top metal layer which acquires a steady state voltage of Vs. Current JT through the
dielectric is modeled within a first order approximation by the resistor RT.
49
Figure 3-12. (a) Schematic description of the MIM grating structures used in polarization
dependent measurements. (b) Calculated field profiles of the structures for TE and TM
polarizations. Field enhancement is minimal for TE polarization, while plasmonic
enhancement is present in the gap for TM polarization. (Scale bar 50 nm). (c) XPS spectra of
the Ag 3d peaks acquired on gratings labeled as (I) and (II) for two polarizer orientations. As
the polarizer is rotated from 0° to 90°, apparent binding energy shifts of the Ag 3d3/2 and
3d5/2 are observed for the gratings with different orientations. Each polarization excites only
one of the gratings, for which the plasmonic modes are excited.
Recently, a nanostripe antenna was used to demonstrate plasmonic hot-electron
current enhancement in MIM devices[52]. Polarization dependence of the
enhancement was measured to show that optical absorption in the metal layers was
dominated by plasmonic effects. Here, we demonstrate that HEE related surface
photovoltages can be measured in similar MIM structures using XPS. Periodic MIM
structures are fabricated using standard techniques. A 70 nm thick Ag layer is
covered with 5 nm HfO2, deposited using Atomic Layer Deposition (ALD) at 100 °C.
Metal stripes are patterned by electron beam lithography and lift-off, resulting in
50 nm thick Ag stripes. The width, period, and length of the stripes are 150 nm,
250 nm, and 50 μm, respectively. The area of the patterns (250 μm2) matches the
XPS data collection spot size. In order to elucidate the plasmonic nature of
50
absorption in the metal layers, two grating structures perpendicular to each other are
fabricated (Figure 3-12a). The MIM gratings feature wide angle and localized
resonances in their optical response[191]. Field localization is calculated using
Finite-Difference Time-Domain (FDTD) simulations for the TE and TM
polarizations (Figure 3-12b). The plasmonic absorption Pa in the metal layers is
resistive and location dependent and given by,
1
𝑃𝑎 (𝑟) = 〈𝑗(𝑟⃗). 𝐸(𝑟⃗)〉 = 2 𝜀𝑜 𝜀𝑖 |𝐸|2 (3.1)
where 𝑗(𝑟⃗) is the current density, 𝜀𝑜 is the vacuum electric permittivity, 𝜀𝑖 is the
imaginary part of the relative electric permittivity of the metal, ω is the angular
frequency, and E(r) is the local electric field. Total absorption can be calculated by
integrating the absorbed power over the material boundaries. It is seen in Figure
3-12b that TE polarization does not excite the plasmonic mode and little or no field
penetrates the metal within the gap region. In this case, hot-electron generation is
minimal. In contrast, TM polarization excites the plasmonic resonance and enhances
hot-electron generation in the top and bottom metal layers, within the regions facing
the dielectric gap. The hot-electrons tunnel across the dielectric and disturb the
charge equilibrium of the surface Ag layers and cause shifts in the apparent BE of Ag
under 532 nm illumination at 40 mW/mm2 (Figure 3-12c). The work-function of Ag is
assumed to be 4.7 eV. The HfO2/Ag conduction band barrier is 2.05 eV and this
facilitates partial tunneling of hot-electrons for the used laser wavelengths 445 nm
(2.77 eV), 532 nm (2.34 eV), and 650 nm (1.92 eV)[192]. The laser is incident at 45°
and the polarization can be chosen to be parallel or at an angle to the surface plane by
a polarizer. Under dark conditions, the Ag 3d peaks whose native binding energies
are 374.35 eV (Ag 3d3/2) and 368.24 eV (Ag 3d5/2) are shifted towards positive
energies (to 370.2 eV for the Ag 3d5/2 peak) due to X-Ray induced charging. When
the polarization has a vertical component, a negative BE shift is observed for the
grating that is perpendicular to the light propagation vector (from 370.2 eV to
369.1 eV for the Ag 3d5/2 peak, II in Figure 3-12c), while the gratings parallel to the
excitation (I in Figure 3-12c) do not exhibit a significant shift. When the polarization
is rotated, the responses of the periodic stripe antennas are reversed, highlighting the
plasmonic origin of the observed shifts.
51
Figure 3-13. (a) Scanning electron micrograph of MIM surface fabricated by atomic
layer deposition of HfO2 on Ag and self-organized formation of Ag nanoislands on
HfO2 upon 3 nm thick Ag evaporation (Scale bar 250 nm). (b) Reflectance of the surfaces
for incidence angles ranging from 20° to 80° shows a broad plasmonic absorption peak
around 580 nm. (c) Photo-response of the MIM surface when illuminated by 532 nm
excitation, measured by XPS. (d) Consecutive measurements of the XPS spectrum under
dark and illuminated conditions exhibit repeatable differential shifts of the Ag 3d lines. (e)–
(h) Same as in (a)–(d) except the top Ag mass thickness is 5 nm and a semicontinuous Ag
film is formed instead of MIM nano-islands and no surface photovoltage is observed.
In order to demonstrate that HEE can be observed in wide area plasmonic
surfaces, we use spontaneously organized MIM surfaces (Figure 3-13). The surfaces
are fabricated by depositing 5 nm thick ALD deposited HfO2 (at 100 °C) on an 80 nm
thick Ag layer on silicon. A thin Ag layer of 3 nm mass thickness is evaporated on
the HfO2, which forms Ag nano-islands of average diameter of 30 nm and thickness
of 10 nm[29], [191]. The surface exhibits a broad plasmonic absorption band around
580 nm (Figure 3-13b) and a repeatable photoresponse as measured by XPS, as
shown in Figure 3-13c, d. When a semicontinuous film is formed on top of HfO2 by
depositing 5 nm thick Ag, no photoresponse can be observed (Figure 3-13e-h). The
lack of surface photovoltage in the 5 nm Ag case is attributed to reduced plasmonic
absorption and lower density of openings in the film where the excitation can couple
to into modes between the metal layers.
52
Figure 3-14. a) XPS spectra of the Ag 3d peaks measured on the MIM surface for dark and
illuminated conditions, using lasers of 650, 532, and 445 nm wavelength, 20 mW power. The
binding energies shift to negative due to hot-electron tunneling from the bottom metal to the
top Ag island. Greater shifts are observed for increasing photon energy. (b) As the
illumination intensity (445 nm) is increased from 5 to 20 mW, greater negative shift of the
binding energy is observed which saturates at high intensities.
When the surface is illuminated by lasers of wavelength 445, 532, and 650 nm (22
mW/mm2), the BE shifts on the 3 nm Ag surface are observed to be dependent on
photon energy (Figure 3-14a). The observed binding energies are shifted from
370.2 eV for the Ag 3d5/2 under dark conditions to 369.01 ± 0.03 eV for 445 nm,
369.33 ± 0.03 eV for 532 nm, and 369.81 ± 0.03 eV for 650 nm excitation. The
apparent BE is dependent on excitation power (Figure 3-14b) for 445 nm excitation at
5, 10, and 20 mW/mm2. The shifts, with respect to observed BE under dark
conditions (at 370.2 eV), are towards the native (grounded, bulk) BE of Ag 3d5/2 at
368.4 eV and increase with increasing excitation power density, saturating at high
intensities.
The observations can be understood in terms of steady state currents, including
hot-electron, tunneling, and photoemission currents. The IV characteristics of MIM
junctions under illumination have been previously studied in the context of energy
harvesting and photo-detection. If no other internal current sources are present, the
open circuit voltage is independent of excitation power and is given by
ℎ𝑣 − 𝜑𝑒 𝐽𝐻𝑏𝑜𝑡𝑡𝑜𝑚
𝑉𝑜𝑐 = − [1 − 𝑡𝑜𝑝 ] (3.2)
𝑒 𝐽𝐻
where JH's and Voc are hot-electron currents and open-circuit voltage, hν is the
photon energy, φe is the barrier height, and e is the electronic charge. In the case of
the MIM junction illuminated by light and X-rays, the photoelectron current Jx must
also be included in calculation of the equilibrium condition. For VS > 0, i.e., a net
53
positive charging of the top islands, the hot-electron currents can be written
𝐻
as 𝐽𝑏𝑡𝑚 = −𝛾𝑃𝑎𝑏𝑡𝑚 (ℎ𝜈 − 𝜑𝑒)/ℎ𝜈 and 𝐽𝑡𝑜𝑝
𝐻
= 𝛾𝑃𝑎𝑡𝑜𝑝 𝑚𝑎𝑥(ℎ𝜈 − 𝜑𝑒 − 𝑒𝑉𝑆, 0)/ℎ𝜈,
where 𝑃𝑎𝑏𝑡𝑚 and 𝑃𝑎𝑡𝑜𝑝 are the absorbed power at the bottom and top metal. Here, γ is
a proportionality constant, which is included to approximately take into account the
hot-electron density and velocity inside the metals. Assuming the surface is
positively charged, i.e., VS > 0, the circuit diagram in Figure 3-11c can be used to
infer the steady state surface voltage VS. For VS larger than (Eph−φe)/e≈0.72 V for
445 nm illumination, all hot-electrons from the top metal are assumed to be reflected
and 𝐽𝑐𝑡𝑜𝑝 =0 . It was demonstrated that the dielectric behaves as a resistive layer under
X-ray exposure, due to continuous ionization within the layer[193], [194]. Resistance
of a 4 nm thick SiO2 layer was determined to be 8.0 ± 1.0 MΩ in an XPS
measurement specifically designed to extract electronic properties. We assume that
the HfO2 layer behaves similarly, like both a tunnel barrier for hot-electrons and a
Poole-Frenkel conductor, i.e.,𝐽𝑡 ≅ 𝑉𝑆 /𝑅𝑡 . A more accurate quantitative analysis
requires direct measurement of the photoelectron current or effective resistance of
the HfO2 layer (as done in Ref. [193] ). The circuit model is still relevant for
providing a qualitative understanding of the observed shifts. The steady state
condition for VS > 0 is
𝑚𝑎𝑥(ℎ𝜈 − 𝜑𝑒 − 𝑒𝑉𝑆, 0) (ℎ𝜈 − 𝜑𝑒) 𝑉𝑆
𝐽𝑥 + 𝛾𝑃𝑎𝑡𝑜𝑝 − 𝛾𝑃𝑎𝑏𝑡𝑚 = (3.3)
ℎ𝜈 ℎ𝜈 𝑅𝑇
which predicts a surface voltage of VS=RTJx for dark conditions (𝑃𝑎𝑏𝑡𝑚 =
𝑃𝑎𝑡𝑜𝑝 =0 ). For the case of high power illumination (𝑃𝑎𝑏𝑡𝑚 , 𝑃𝑎𝑡𝑜𝑝 =0→∞), Equation
(ℎ𝜈−𝜑𝑒 )
3.3 predicts a surface voltage 𝑉𝑆 = [1 − 𝑃𝑎𝑏𝑡𝑚 /𝑃𝑎𝑡𝑜𝑝 ], which is equivalent to
𝑒
the open circuit voltage given in Equation 3.3. Using Jx≈ 72 nA/mm2 measured for a
similar XPS system, we estimate the resistance of the 5 nm thick HfO2 layer as 25
MΩ over 1 mm2 surface area. According to the data shown in Figure 3-14b for
445 nm excitation, 5 mW incident power causes a surface voltage shift of 256 mV
over a 300 μm diameter illumination area, corresponding to a hot-electron current of
2.5 nA. The current is used to estimate the external quantum efficiency (EQE) as
11× 10−6 and responsivity as 500 nA/W, for the 300 μm diameter measurement area.
The calculated responsivity compares favorably with the 75 nA/W measured for the
same wavelength in stripe nanoantenna devices[52]. At higher excitation densities at
54
445 nm (Figure 3-14b), the surface voltage approaches 0.54 V, which would require
a 𝐽𝐻𝑏𝑜𝑡𝑡𝑜𝑚 ⁄𝐽𝐻𝑡𝑜𝑝 ratio of 0.25 according to Equation 3.2.
3.4. Conclusion
In this chapter, we have developed a nano-particle based braodband and wide
angle large area plasmonic meta-surface. The broad bandwidth is due to the
excitation of electrical and magnetic plasmon modes. Then, we have used these
surfaces for single molecule Raman spectroscopy and photo-detector applications.
For the first time, we have imaged single molecule Raman event using a mobile
phone's camera. We have also developed a contact free technique to probe hot-
electrons using XPS.
55
Chapter 4
Exploiting Native Aluminum Oxide

for Multispectral Aluminum
Plasmonics
Aluminum, despite its abundance and low cost, is usually avoided for plasmonic
applications due to losses in visible/infrared (IR) regimes and its inter-band
absorption at 800nm. Yet, it is compatible with silicon complementary metal oxide
semiconductor (CMOS) processes, making it a promising alternative for integrated
plasmonic applications. It is also well known that a thin layer of native Al2O3 layer is
formed on aluminum when exposed to air which must be taken into account properly
while designing plasmonic structures. For the first time we report exploitation of the
native Al2O3 layer for fabrication of periodic metal-insulator-metal (MIM) plasmonic
structures that exhibit resonances spanning a wide spectral range, from the near
ultraviolet (UV) to mid-infrared (MID-IR) region of the spectrum. Through
fabrication of silver nano-islands on aluminum surfaces and MIM plasmonic surfaces
with a thin native Al2O3 layer, hierarchical plasmonic structures are formed and used
in surface enhanced infrared spectroscopy (SEIRA) and surface enhanced Raman
spectroscopy (SERS) for detection of self-assembled monolayers of dodecanethiol
(DDT).
Part of this chapter is published as;
Sencer Ayas, Ahmet Emin Topal, Andi Cupallari, Hasan Güner, Gokhan Bakan
and Aykutlu Dâna,"Exploiting Native Al2O3 for Multispectral Plasmonic Structures",
ACS Photonics,1,12,2014
56
4.1. Introduction
Recent advancement in plasmonics enabled the development of better performing
plasmonic materials for the ultra-violet (UV)[195]–[199] and infrared (IR)[200]–
[207] portion of the light spectrum. Typically gold (Au) and silver (Ag) are most
common materials used to fabricate nanostructures to study novel plasmon-enhanced
and enabled optical phenomena such as negative refraction,[208], [209]
transformation optics,[12] surface plasmon sensors,[2], [59] surface enhanced Raman
spectroscopy (SERS),[210], [211] surface enhanced infrared absorption spectroscopy
(SEIRA),[64], [67] plasmon enhanced solar cells and detectors[14], [72]. Au has an
internal band transition around 500nm which limits utilization of gold towards the
UV portion of the visible spectrum[34]. Due to its chemically inert properties,
stability and its tailorable binding to biomolecules, Au is widely used for surface
plasmon resonance sensor applications working at visible wavelengths closer to NIR
regime. Ag is considered as the optimal material for plasmonic applications in the
visible spectrum due to its low loss compared to other metals[34]. However, Ag
suffers from atmospheric sulfur contamination and oxidation[212]. Aluminum arises
as a promising material for UV[213], [214] and deep UV plasmonic applications[16],
[197], [198], [215], [216] owing to its high plasma frequency. Al has high losses
from visible to IR range as well as an interband absorption around 800 nm which
makes it less favorable as a NIR plasmonic material[34], [195], [217]. Still,
localized plasmon resonances in Al have been demonstrated in several geometries,
including nanoparticles,[213], [216], [218] triangles,[197], [214], [219] discs,[198],
[220], [221] rods and nanoantennas[16], [222], [223]. Relative abundance of Al can
be advantageous for the design of plasmonic absorbers in solar energy conversion, or
57
for inexpensive, disposable SERS and fluorescence enhancing substrates. Due to
CMOS compatibility, Al plasmonic nanostructures would be optimal for silicon
integrated optoelectronic applications, such as integrated biomolecular sensing,[3],
[224] on-chip plasmonic nanoantennas, waveguides and interconnects[7], [208],
[225]–[227]. Furthermore, the surface oxide of Al self-terminates at a thickness of
3−5 nm, forming a durable protective layer and preserving the metal, making this
material a convenient choice for plasmonic applications that require durability[223],
[228]. In this part of this thesis, we demonstrate exploitation of native oxide (NO)
layer to form metal-insulator-metal (MIM) structures with resonances spanning a
wide spectral range by enabling extreme mode confinement limit where multiple
modes can be excited on the same structure due to ultrathin native oxide spacer layer.
Nano-particle based MIM structures are demonstrated over large areas using a thin
dewetting Ag top layer near the percolation threshold on naturally oxidized
aluminum. The optical properties of such MIM structures is tuned by simply
allowing a thicker native oxide layer to form by longer exposure to ambient
conditions. The native oxide also allows simple fabrication of hierarchical plasmonic
surfaces that exhibit simultaneous resonances in the visible and infrared regimes.
Such hierarchical multispectral MIM structures feature simultaneously surface
enhanced Raman spectroscopy (SERS) and surface enhanced infrared absorption
spectroscopy (SEIRA).
4.2. Native Oxide Based Plasmonic Meta-surfaces

Periodic MIM structures on aluminum bottom layers are studied where the
insulator is the native oxide layer which is naturally formed during the fabrication
process that eliminates the need for depositing an insulator layer. Typically,
fabrication process of native oxide based MIM (NO-MIM) structures requires
multiple metal deposition processes (See Figure 4-1). After depositing the first Al
58
layer, the vacuum is broken for PMMA coating and e-beam lithography, which is
accompanied by formation of a thin layer of Al2O3 on the surface of Al. The first set
of plasmonic structures are formed through fabrication of nanodisc arrays on a
silicon substrate coated with a thick layer of aluminum to demonstrate NIR
resonance due to presence of native oxide. Figure 4-2a illustrates the formation of
NO-MIM plasmonic surfaces with top nanodisc array. The fabricated NO-MIM
structures show a resonance at NIR regime (Figure 4-2c) as characterized by a Fourier
Transform Infrared Reflection (FTIR) microscope system equipped with a 15x
Cassegrain objective. A knife edge aperture is used to limit measurement area to
regions containing lithographically defined structures (100μm x 100μm). FDTD
calculations show that a resonance emerges with the presence of a thin Al2O3 layer
and blueshifts with increasing oxide thickness (Figure 4-2b). Evolution of the native
oxide thickness is monitored by x-ray photoelectron spectroscopy (XPS) as shown in
Figure 4-3. Uniformity of the oxide thickness is verified by XPS measurements at
various spots on a continuous Al film. The thickness of the native oxide layer is also
characterized by spectroscopic ellipsometer measurements. However, it is not
possible to determine the exact value of the native aluminum oxide thickness due to
the surface roughness of aluminum. Neglecting the surface roughness of Al which
has 1.5nm of RMS value, the thickness of the native oxide layer is found to be 5.2nm
after 24-hour exposure to air (20˚C, 35% relative humidity).
59
evaporated as an adhesion layer prior to Al deposition. (ii) Formation of a thin aluminum
oxide (Al2O3) layer upon exposing the Al films to air for e-beam lithography. (iii) ~100nm
PMMA is spin coated and annealed at 180˚C for 120 seconds. (iv) Patterns are formed using
e-beam lithography. (v) 50nm of Al or Ag is evaporated. (vi) Lift-off process to form the
final structures.
Figure 4-2. Verification of NO-MIM structures through observations of resonances in NIR.

(a) Schematic of formation of NO-MIM surfaces during fabrication process (Top patterned
layer height is 50 nm). (b) Simulated reflectance spectrum of NO-MIM structures for various
Al2O3 thicknesses. While no plasmonic resonance is observed in the absence of native oxide,
plasmonic resonances in NIR are observed due to the native oxide layer. Resonances blue-
shift with increasing oxide thickness. (c) Experimental reflection spectra of the fabricated
nanodisc NO-MIM structure. Diameter and period of the discs are 250nm and 400nm,
respectively. (Inset) SEM image of the nanodisc array.
60
Figure 4-3. Characterization of the native oxide thickness over a time span of 24 days using
XPS. (a) XPS spectrum of an aluminum film. High energy shoulder corresponds to the
aluminum oxide peak. (b) Evolution of the calculated native oxide thicknesses.
In order to demonstrate the versatility and the tunability of NO-MIM structures,
anisotropic structures that support resonances in the NIR and MIR regimes are
investigated with long and short axes periods of 2μm by 500nm, respectively.
Schematic of these structures is shown in Figure 4-4a. SEM images of the fabricated
structures are shown in Figure 4-4b. Tunability of reflection spectra by increasing
either long or short axis width is shown in Figure 4-4b. The reflection depths get
weaker at these wavelengths due to weaker coupling of long wavelengths to the thin
native oxide layer. In an MIM cavity the effective propagation constant and effective
refractive index depend on the thickness of the dielectric film. The plasmonic
resonances are strongly dependent on the period and the width of the surfaces.
Within a first order approximation, the MIM mode of a patterned surface can be
modeled as a standing wave that reflects back and forth from the edges of the top
patterned layer, where individual resonators are assumed to be non-interacting. For
closely spaced MIM resonators, coupling becomes important. Period of MIM
structures affects the coupling of these individual nanocavities and typically for a
constant MIM width, the interaction increases for smaller periods due to stronger
intercavity coupling. Otherwise, period of an array of MIM resonators is assumed to
be not dominant in determining the individual resonator response. In order to find the
resonance wavelengths, one has to solve the characteristic equation given by,
2𝜋
𝑊𝑆,𝐿 𝑛𝑒𝑓𝑓 = 𝜋 − 𝜑𝑆,𝐿 (4.1)
𝜆
61
where WS,L are the widths along the short and long axes, neff is the effective
refractive index of the propagating mode and 𝜑𝑆,𝐿 is the phase term due to the
reflection from the terminations. The effective refractive index, neff, is calculated by
solving the Maxwell equations analytically for continuous MIM waveguide as shown
in Figure 4-4d (See also Section 2.2). Figure 4-4c shows the calculated resonance
frequencies of NO-MIM modes with experimentally measured resonance frequencies
for varying widths along either short or long axes. An Al2O3 thickness of 5±0.5 nm is
deduced from the model which agrees well with the ellipsometer measurements. It is
observed that a dispersive dielectric function explains the observed resonance
frequencies better than a constant dielectric function for Al2O3. Moreover, the cavity
size of 1000nm x 310nm asymmetric resonators is about λ3/500,000 assuming 5nm
thick native oxide layer.
A simple MIM structure with 1D Al bar array offers tunable simultaneous
resonances in both visible and NIR regime due to accommodation of the first and
higher order modes when period and width of 1D structures are chosen properly
(P=250nm). Only odd order modes are excited for normal illumination, due to
symmetry. Even order modes can be excited at oblique angles due to the broken
symmetry condition. Schematic and an SEM image of the fabricated 1D NO-MIM
structures are shown in Figure 4-5a. Figure 4-5b shows the measured reflection spectra
for the visible regime with increasing 1D structure width. The reflection spectra are
measured with fiber coupled white light, polarized along the grating axis, using
reflection from an aluminum mirror as the reference. Resonance wavelength red-
shifts with increasing width (Figure 4-5). NIR reflectance spectra of the same
structures are measured with the FTIR microscope using un-polarized NIR light
source (Figure 4-5c). The resonance wavelength starts from 1200nm and shifts to
1400nm with increasing width. Measured visible and NIR reflectance spectra agree
well with the simulations shown in Figure 4-5c,d. Simulations performed with
neglecting NO show no distinct resonances which also indicates that the resonances
are due to the presence of native oxide layer. Although almost zero reflection is
observed in the simulations, the reflection depths for NIR measurements are smaller
compared to the simulated values due to the use of un-polarized light. Surface
roughness is neglected in the simulations which might account for some of the
discrepancies such as resonance wavelength mismatch between experiments and
simulations. Magnetic field distributions are simulated as shown in Figure 4-5f to
62
further understand the origin of the observed visible and NIR resonances. Magnetic
field is confined to the oxide layer for both resonances. For the NIR resonance, first
order or fundamental mode is excited. For visible, third order resonance is excited
where there are three magnetic field peaks in the oxide layer (λ1=
̃ 3λ3).
Figure 4-4. Resonance tuning in the NIR and MIR range using anisotropic Al bar arrays. (a)
Schematic of asymmetric NO-MIM structures with nonidentical periods and widths along x
and y directions to excite multiple modes in IR (Top patterned layer height is 50 nm). (b)
Typical reflection spectra showing multiple resonances in the NIR and MIR range due to the
asymmetry of the structures with Px=500nm, Py=2000nm and Wx=310nm and Wy=1000nm
(black curve), 1500nm (blue curve) (top) and Wx=250nm (red curve), 310nm (blue curve)
for Wy=1000nm (bottom). (Inset) SEM images of corresponding structures. Red dashed lines
on the cartoon illustrations indicate the axis on which the structures’ width is modified. (c)
Experimentally observed resonance wavelengths as functions of the Al bar width either
along the short or long axis (blue dots). Calculated resonances by treating the NO-MIM
structures as Fabry-Perot (FP) resonators consisting of truncated waveguides (green and red
curves). A better fit of the FP resonator model to the experiments can be achieved if the
dielectric function of Al2O3 is assumed to be wavelength dependent (green curve), as
opposed to a constant (red curve, n=1.6). Vertical lines on the curves are errors bars
corresponding to 0.5nm uncertainty in the oxide thickness. (d) Effective refractive index of
the waveguide forming the FP resonator as a function of Al2O3 thickness, using the
wavelength dependent dielectric function.
63
Figure 4-5. Simultaneous resonances in visible and NIR regime using 1D NO-MIM. (a)
Schematic of 1D NO-MIM structures. tgr=50nm and P=250nm. Experimental results
showing tuning of resonance wavelengths as W changes between 110 and 150nm, (b) in the
visible and (c) in the NIR regime. Simulation results for the experimented structures shown
in (d-e) being in a good agreement with the experimental results. Each spectrum curve is
shifted by 1 along y-axis for clarity. (f) First order (m=1) and third order (m=3) mode
profiles corresponding to NIR (ii) and visible (i) resonances as indicated with arrows in (d)
and (e). The even mode (m=2) is not observed for normal incidence. In the simulations 5nm
of native oxide layer thickness is assumed.
4.3. Nanoparticle Based Large Area Surfaces

The studies presented in this part so far includes a lithography step to pattern the
top layer which always yields a native oxide layer. To study the existence of native
oxide layer, NO-MIM structures based on self-forming Ag nanoparticles on Al
bottom layer is fabricated with and without exposing Al layer to air by breaking the
vacuum. Such approach also offers larger area plasmonic surfaces compared to top
down fabrication techniques such as e-beam lithography and focused ion beam
64
milling. For this study, Ag nanoparticles with an average size of 25 nm are obtained
by evaporating 3nm Ag film (See Figure 4-6). Exposing Al surfaces to air for longer
duration, hence forming a thicker oxide causes a blue-shift in the reflection spectrum
(Figure 4-7a). Sequential deposition of Ag on Al films without breaking the vacuum
does not show a resonance due to lack of an oxide layer. The effect of native oxide
layer on Ag nanoparticle based MIM structures is simulated by modeling the
nanoparticle film as an array of truncated cones as seen in Figure 4-7b. Increasing
exposure time is captured through increasing NO thickness in the simulations which
shows a blue-shift.
Figure 4-6. SEM images of Ag nanoparticles on silicon and their size distribution.
65
Figure 4-7. Nano-particle based large area NO-MIM structures. (a) Reflection spectra for Ag
nano-particle based NO-MIM structures. Resonance blue-shifts if the time period between
two deposition processes increases (dashed and solid red lines). No resonance is observed if
the deposition of Al and Ag is done without breaking the vacuum (blue solid line). (Inset)
SEM images of 3nm Ag on Si(i) and Al(ii). (c) Simulation results and simulated structure
(inset) for tox=3nm and 5nm. Simulated structure is a periodic array of truncated cones with
D=35nm, P=40nm, tgr=20nm and θ=75 degrees
66
4.4. Surface Enhanced Infrared Absorption (SEIRA)
Spectroscopy and Surface Enhanced Raman
Spectroscopy (SERS) Using Hierarchical Plasmonic
Surfaces
The studied plasmonic structures in this part of the thesis are used by sensing
organic materials under Raman and infrared absorption spectroscopy which give
information about the molecular structures and vibration spectra of molecules. The
symmetry requirements in these two phenomena are different and they give
complementary information about a molecule. Typically, high concentration of
molecules is required in these techniques. On the other hand, SERS and SEIRA can
magnify the signal levels up to 108 folds where resonant light structures are used to
enhance field intensities enabling observation of the molecular vibrational signatures
from few molecules. By taking advantage of multispectral characteristic of NO-MIM
surfaces we demonstrate SERS and SEIRA on the same structures. In the
experiments dodecanethiol (DDT) is used as the probe molecule. DDT cannot form a
monolayer on aluminum. However, DDT molecular formation on Ag has been long
studied[229]. Therefore, we first tried to demonstrate SEIRA on NO-MIM structures
with a Ag top layer (Ag/Al2O3/Al). The resonance frequency of the structures is
around 3500 cm-1 which is close to the C-H stretching of both DDT and PMMA (
2930 and 2955 cm-1 ). However, during fabrication resist residues, i.e. polymethyl
methacrylate (PMMA) for e-beam lithography, are left on or under the patterned
metals, resulting vibrational signatures coincide with those of DDT. Despite the
presence of PMMA residues, the molecular signature of DDT after monolayer
formation can be observed as shown in Figure 4-8. The number of molecules
contributing to the SEIRA signal is estimated to be 4.6 femtomole on such periodic
67
structures. Less than 1 attomole sensitivity can be achieved by taking measurements
from a single NO-MIM resonator, if a light source with very high SNR is used such
as synchrotron light source.[230] The SEIRA enhancement factor is calculated as
2200 for these structures using a bare Ag film as the reference (Figure 4-9).
Figure 4-8. SEIRA on NO-MIM structures with patterned Ag top layer fabricated by e-beam
lithography. Reflection spectra of these structures before (blue) and after (green) DDT
monolayer formation for (a) Wx,y=300nm and (b) Wx,y=350nm. Periods along x and y
directions are the same (Px=Py=500nm). (c) and (d) First derivative of the reflection spectra
in (a) and (b), respectively. Molecular signatures of PMMA and DDT are more visible in the
first derivative curves in (c) and (d) compared to the reflection spectra in (a) and (b). (e) An
SEM image of the fabricated structures. (Scale bar: 1 μm ).
68
Figure 4-9. (a) Infrared reflection spectrum of the monolayer DDT on bare Ag mirror
measured under Grazing Angle Illumination after background subtraction. (b) Infrared
reflection of the monolayer DDT on the NO-MIM structures with Ag top layer after
background subtraction.
In order to eliminate PMMA signatures from residues, the NO-MIM
structures are treated in oxygen plasma. Since, oxygen plasma deteriorates top silver
layer, a different type of resonator structure with multiple hierarchical MIM cavities
is developed. Figure 4-10a illustrates the hierarchical NO-MIM (metal-insulator-
metal-insulator-metal, MIMIM) structures developed for SEIRA measurements to
eliminate the PMMA residues with oxygen plasma. Oxygen plasma does not harm
Al, but increase oxide thickness on Al slightly. The fabrication steps of such
structures are shown in Figure 4-11. First, an all-aluminum NO-MIM structure is
fabricated by e-beam lithography with 500nm period and 350nm width that exhibits
resonance in the IR (around 3500 cm-1). Then the structures are cleaned in oxygen
plasma, for PMMA residue removal. By evaporating 3nm of Ag on cleaned
Al/Al2O3/Al resonators, final MIMIM structures are obtained. Simulations show that
depositing Ag nanoparticles on the Al NO-MIM structures change these structures’
reflection spectra significantly in the visible regime, whereas it only causes a slight
blue-shift in the IR regime (Figure 4-12). Figure 4-10b shows SEM image of the
fabricated structure. Ag nanoparticles are formed both on and between the
rectangular Al patterns. We first measured the FTIR reflection spectra of DDT
69
solution on bare Al surface to determine the molecular bands of DDT close to
resonance of the fabricated structures as shown in Figure 4-10c. Then, FTIR
reflection spectra are recorded in each step of the fabrication process as shown in
Figure 4-10d, e. The resonance wavelengths do not change significantly after oxygen
plasma and Ag deposition. The molecular signature of PMMA residues after liftoff
can be determined before oxygen plasma. After oxygen plasma, PMMA residues are
cleaned and no molecular bands for PMMA are observed. Following 3 nm Ag
evaporation, the structures are treated with 1 mM DDT in ethanol solution for
monolayer formation. The surfaces are washed with ethanol to remove unattached
DDT molecules, and reflection spectra are recorded. Molecular signatures of DDT
molecules can be identified without any post processing in the reflection spectra. The
first derivative of each reflection spectrum is shown in Figure 4-10f, which better
differentiates the DDT and PMMA signatures[231].
Figure 4-10. SEIRA detection of molecular monolayers on hierarchical NO-MIM structures.

(a) Schematic of hierarchical NO-MIM structures before and after nanoparticle deposition
with top patterned layer height of 50 nm. Periods and widths along x and y directions are the
same, Px=Py=500nm and Wx=Wy=350nm. (b) SEM image of NO-MIM structures after 3nm
Ag deposition (Scale bar: 500nm). (c) FTIR reflection spectrum of thick DDT solution on a
70
bare Al film. Region of interest is highlighted in grey. (Inset) Molecular sketch of DDT
molecule. (d) FTIR reflection spectra of NO-MIM structures for: (i) just after fabrication of
all Al NO-MIM structures, (ii) after plasma cleaning of PMMA residues, (iii) after the
formation of hierarchical NO-MIM structures by 3nm Ag deposition, (iv) after DDT
molecular monolayer formation on hierarchical NO-MIM structures. (e) FTIR reflectance
between 2500 and 3500 cm-1, and its derivative (f) for better visualization of stretching of
monolayer DDT molecules. Arrows mark the DDT and PMMA signatures in (e). Molecular
signatures (2930 cm-1 and 2955 cm-1) of PMMA are observable before oxygen plasma
cleaning. No distinctive bands are observed after oxygen plasma (ii) and Ag deposition (iii).

evaporated as an adhesion layer prior to Al deposition. (ii) Formation of a thin aluminum
oxide (Al2O3) layer upon exposing the Al films to air for e-beam lithography. (iii) ~100nm
PMMA is spin coated and annealed at 180˚C for 120 seconds. (iv) Patterns are formed using
e-beam lithography. (v) 50nm of Al is evaporated. (vi) Lift-off process to form all Al MIM
structures. (vii) O2 plasma cleaning for PMMA residue removal and formation of a thin
oxide film on the top Al structures. (viii) Formation of Ag nanoparticles through evaporation
of 3nm Ag film.
71
Figure 4-12. (a) Reflectance spectra for Ag nanoparticles-based NO-MIM (blue), all Al NO-
MIM (red) and hierarchical NO-MIM (green) surfaces in the visible regime. (b) Reflectance
spectra for all Al NO-MIM and hierarchical NO-MIM (red) surfaces in the MIR regime.
We have also used another method to fabricate PMMA-free NO-MIM
surfaces with top Ag layer using focused ion beam (FIB) milling (Figure 4-13). Since
the area that can be fabricated with FIB is smaller compared to e-beam lithography,
the resonance contrast in the reflection spectrum from these surfaces is weak (Figure
4-13d). Other nanofabrication techniques can be used to eliminate the molecular
signatures coming from the residues such as nano-stencil-lithography[231].
72
Figure 4-13. Characterization of NO-MIM structures fabricated using Focused Ion Beam
(FIB) milling. (a) Fabrication steps for NOMIM structures: (i) Thermal evaporation of
100nm of Al on silicon substrates, (ii) formation of a thin aluminum oxide (Al2O3) layer
upon exposing the Al films to air, (iii) thermal evaporation of 50nm of Ag, (iv) FIB
patterning of the top Ag layer. (b) Measured (100μmx 100μm) area with respect to the
fabricated area (50μmx 50μm). (c) Top-view and cross-section (inset) SEM images of FIB
fabricated surfaces. (d) Reflectance spectra for the FIB fabricated surfaces with shown
widths and periods.
The presence of closely spaced Ag nanoparticles on top of the hierarchical
NO-MIM resonators with resonances in the visible and infrared presents an
opportunity to perform SERS and SEIRA simultaneously. The Raman spectrum of a
thick DDT solution on an Al covered Si substrate, recorded in 10 s using 13 mW
excitation at 532 nm is shown in Figure 4-14a. In comparison, SERS data of DDT
monolayers formed on hierarchical NO-MIM surfaces is shown in Figure 4-14b,
collected using 0.5 mW excitation and 100 ms integration time. Despite shifts and
changes in intensity of several lines, main features of the spectra are comparable to
73
Raman data shown in Figure 4-14a. No distinct spectral features of DDT are observed
on bare Al films on Si (Figure 4-14b), since DDT does not stick on Al. It should be
noted that, by coinciding the visible resonances of the NO-MIM structures with the
resonances of Ag nanoparticle layer, an enhancement of the field intensities is
predicted by computational study of MIMIM structures (Figure 4-14c, showing |E|
distribution). While a small field enhancement is obtained for all Al NO-MIM
structures around 532nm, Ag nanoparticle based NO-MIM structures show improved
field enhancement due to better impedance matching of the resonances to free space
modes, as well as due to reduced interparticle spacing. In contrast, hierarchical NO-
MIM structures show further improvement in field enhancement, due to overlapping
of the visible resonances of the Al and Ag island layers. Visible resonance is
relatively weak, and accounts for the diminished reflectance over the hierarchical
NO-MIM structures as shown in Figure 4-14d. The hierarchical NO-MIM structure
however shows improved SERS performance as shown in the Raman map in Figure
4-14e. When the surface is illuminated by a defocused 532 nm excitation and
inspected with a monochrome CMOS camera, blink events[232], typically associated
with single molecule SERS are observed on the hierarchical portions of the sample
(Figure 4-14f). Although FDTD calculations suggest overall SERS enhancement of
1.6x105, we attribute the blinking spots to the presence of hot spots with greater field
enhancement, due to the random nature of the top Ag nanoisland layer. We have
estimated the enhancement factors as 2.2x103 for SEIRA and 103 for SERS on our
multispectral plasmonic structures. These values are comparable to the ones reported
in the literature[63], [206], [207] and one order of magnitude smaller than the
reported values in Ref. [233] .
74
Figure 4-14. SERS detection of molecular monolayers on hierarchical NO-MIM structures.
(a) Raman Spectra of a thick DDT solution on 80nm Al coated silicon. (b) SERS spectra of
monolayer DDT on hierarchical NOMIM surfaces and an Al coated Si substrate. Integration
time and laser power are 100ms and 0.5mW, respectively, for hierarchical NOMIM surfaces;
where they are 10s and 13mW, for Al films on Si. (c) Simulated field profiles for all Al NO-
MIM (MIM1), Ag nano-particle based NO-MIM (MIM2) and hierarchical NO-MIM
(MIMIM) structures. (d) Optical micrograph of hierarchical NO-MIM surfaces. (e) SERS
mapping of DDT on hierarchical NO-MIM surfaces. Brighter and darker regions emphasize
hierarchical NO-MIM and nano-particle based NO-MIM structures, respectively. (f) Wide-
field CMOS camera image of blink events on NO-MIM structures. Blink events are only
observed on hierarchical NO-MIM structures.
4.5. All Aluminum Hierarchical Surfaces in the Infrared

So far we presented hierarchical surfaces utilizing Ag nano-particles on top of
NO-MIM surfaces. Here, we present all-aluminum broadband hierarchical plasmonic
surfaces using multiple Al and native Al2O3 layers. The hierarchical plasmonic
surfaces are simply fabricated by sequential deposition and oxidation of Al films
through breaking the vacuum of the physical vapor deposition system to expose the
Al films to air. The fabricated hierarchical structures have multiple resonances in the
MIR wavelengths as characterized by FTIR measurements and confirmed by
simulations. Simulations of higher order hierarchical plasmonic surfaces show
broader spectrum and reveal the importance of the native Al2O3 layers thicknesses.
These structures have potential for applications that require durability, multispectral
75
characteristics such as surface enhanced spectroscopies, optical waveguides, color
printing and thermal emitters.
The 3D schematics and top-down scanning electron microscopy (SEM) images of
the fabricated hierarchical structures are shown in Figure 4-15. The top layer of these
structures is designed to have arrays of squares with widths around 300nm and a
period of 500nm along both x- and y-axis. The thickness of each top Al layer is
chosen as 50nm. Typically the fabrication of such structures requires deposition of
multiple dielectric and metal layers. Using Al eliminates the deposition of dielectric
layers through the formation the native Al2O3. The thickness of the oxide layer
depends on the exposure time of the Al surface to air. Since the native Al2O3 is
ultrathin, even small changes in its thickness result in large shifts in the resonance
wavelengths of the fabricated structures. The hierarchical Al surfaces are fabricated
using e-beam lithography and multiple Al deposition steps. The fabrication steps are
shown in Figure 4-15c and summarized here. First, silicon substrates are coated with
80nm thick Al using thermal evaporation following deposition of a thin layer of
germanium (5nm) adhesion layer. The Ge layer also helps achieving smoother Al
films as confirmed by AFM measurements. The first Al layer grows a thin Al 2O3
layer when exposed to air. Then the samples are spin-coated with 250nm PMMA for
e-beam lithography. Once the PMMA is patterned, 50nm thick Al is deposited. This
first Al layer is exposed to air to form the second native oxide layer by breaking
vacuum of the deposition chamber. Then the final Al layer is deposited and the final
structures are formed using lift-off process. The area of the fabricated surfaces is
50μm x 50μm. The thicknesses of the top and bottom oxide layers are expected to be
different. A Thermo K-Alpha monochromated high-performance X-ray
photoelectron spectroscopy (XPS) system is used to characterize the thickness of the
native Al2O3 layers.
76
Figure 4-15. (a) 3D and (b) side view of the hierarchical aluminum surfaces. Top Al layer
thicknesses (tgr1 and tgr2) are 50nm. The thicknesses of the bottom and top oxide layers are
denoted as tox1 and tox2, respectively. (c) Top-down SEM image of the fabricated structures.
(d) Fabrication of hierarchical Al structures. (i) Deposition of 80 nm thick Al. (ii) Formation
of native Al2O3 after exposure of the Al films to air. (iii) PMMA coating. (iv) Pattering
PMMA with e-beam lithography. (v) Deposition of the 1st 50 nm thick Al. (vi) Formation of
the 2nd native Al2O3 layer after vacuum break. (vii) Deposition of the 2nd 50 nm thick Al
layer. (viii) Lift-off process.
Depth dependent x-ray photoelectron energy spectra of the fabricated Al surfaces
are shown in Figure 4-16. The depth dependent binding energy spectrum is obtained
by etching the surfaces with ion gun and measuring the energy of the emitted
photoelectrons from the etched surface. To characterize the thickness of the top oxide
layer, the depth dependent binding energy spectrum for a thick Al film (80nm) on a
silicon substrate is measured just in an hour after the deposition of the film (Figure
4-16a). The high energy peak corresponds to the aluminum-oxygen binding which
indicates the formation of the native Al2O3 film. After 2nm of etch depth, the high
energy peak reduces to 20% of its original intensity value. For 3nm of etch depth, the
high energy peak almost disappears. Hence, the thickness of the native Al 2O3 on a
freshly deposited Al film is deduced to be 2-3 nm. Figure 4-16b shows the depth
77
dependent XPS energy spectra for the Al surfaces outside the patterned area after the
fabrication of hierarchical Al structures. The high energy peak is stronger compared
to the freshly deposited Al films. After 5nm of etching, the high energy peak reduces
more than 80% and it only slightly changes for further etching. Hence, it is
concluded that the thickness of the native Al2O3 on the Al films experienced the
fabrication steps is about 5nm or more. The thicker native oxide on the bottom Al
films is expected to be due to longer exposure to air and the chemical processes
during the fabrication. The use of acetone, isopropanol alcohol and water might also
affect the final thickness of the films.
Figure 4-16. Depth dependent XPS characterization of Al surfaces with thin native Al 2O3 on
top. (a) Just after the deposition of Al films. (b) After the fabrication of hierarchical Al
structures.
The reflection spectra of the hierarchical Al structures are characterized using
Fourier transform infrared spectrometer (FTIR) equipped with a 15x Cassegrain
objective in the MIR regime. Since the area of the fabricated hierarchical Al surfaces
is 50 μm x 50 μm a knife edge aperture is used to collect light reflected from the
patterned area. As the reference sample, an 80 nm thick Al coated silicon substrate is
used. Owing to two different oxide thicknesses, two distinct MIR resonances are
observed (Figure 4-17a). These resonances red-shift with increasing widths as
expected. These structures are simulated using 2.5nm and 5nm thicknesses for the
top and bottom oxide layers respectively (Figure 4-17b). The simulated spectral
responses are similar to the experimental results, however the depths of the
experimental resonances are weaker. This discrepancy might be due to surface
roughness of the Al films which is not accounted for in the simulations. Moreover,
the native oxide layer, which might be lossy by its nature, is assumed to be lossless
in the MIR simulations (See Figure 4-18). The simulated electric and magnetic field
78
distributions are shown in Figure 4-17c corresponding to the resonances denoted with
(i) and (ii) in Figure 4-17b. Since magnetic field is confined to the Al2O3 layers, the
resonances are magnetic plasmon modes by its nature. Magnetic fields are confined
to the bottom (5nm) and top (2.5nm) native Al2O3 layers for short and long
wavelength resonances, respectively. The confinement of longer wavelengths to the
thinner oxide layer can be intuitively explained by increased effective refractive
index with decreasing oxide thickness, accompanied by a red-shift in the resonance
wavelength[234](λr=2neffW).
Figure 4-17. (a) Measured and (b) simulated reflectance spectra for W=290nm, 300nm,
320nm and 340nm (Period=500nm). (c) Magnetic field intensities for the resonances shown
in (b).
79
Figure 4-18. Simulated reflection spectra for the hierarchical MIM structure with two oxide
layers assuming different extinction coefficients for the native oxide layer.
The effect of the top and bottom oxide thicknesses on the reflection spectra are
further studied by varying these thicknesses in simulations (Figure 4-19). The
resonance wavelengths shift drastically with sub-nm changes in the ultrathin oxide
layers (2-5 nm) as shown in Figure 4-19 for W=300nm. When both oxide layers are
neglected in the simulations, no resonances in the MIR are observed as expected
(Figure 4-19a). If the thinner top oxide is neglected, only the short wavelength
resonance is excited where it is confined to the bottom oxide layer. If the thicker
bottom oxide is neglected, only the longer wavelength resonance is excited where the
magnetic field is confined to top oxide layer. The effect of the native oxide layer on
the longer wavelength resonance is investigated by varying the thickness of the top
oxide layer in between 0.5nm and 3.5 nm for 5nm thick bottom oxide layer (Figure
4-19b). As the top oxide thickness decreases, the longer wavelength resonance red-
shifts. As the thickness approaches the bottom oxide’s value (5nm) the resonances
start overlapping. As a result, the top and bottom oxide thicknesses have to be
different to observe a multispectral response. If the oxide thicknesses are the same,
only one mode is excited with stronger confinement of light into both oxide layers.
80
Figure 4-19. Simulated reflection spectra for (a) varying tox-1 and tox-2, (b) varying tox-2 with
tox-1=5nm. W=300 nm, P=500 nm for all cases
The use of the native oxide as the spacer layer offers an easy fabrication route to
realize higher order hierarchical resonator structures (Figure 4-20a). The optical
response and field distribution of such structures with 3 and 4 oxide layers are
studied using simulations. Since the resonance wavelength of an MIM resonator
strongly depends on the thickness of the oxide layer, it is possible to excite as many
modes as the number of oxide layers. Hence 3 distinct resonances are observed when
3 oxide layers with different thicknesses are used (Figure 4-20b). Similar to the
previous simulation results with two oxide layers, when top two oxide thicknesses
are chosen to be the same, only two resonances are observed. The study is repeated
for 4 oxide layers and 3 different thickness combinations: 1) all thicknesses different,
2) two of the thicknesses are the same, 3) three of the thicknesses are the same
resulting in 4, 3 and 2 distinct resonances, respectively (Figure 4-20c, d). As the
number of layers with the same thickness increases, the bandwidth decreases, yet
light couples into the MIM mode determined by the common oxide thickness more
strongly.
81
Figure 4-20. (a) 3D schematics of higher order hierarchical aluminum surfaces with multiple
MIM resonators. Simulated reflectance spectra for structures with (b) 3, (c) 4 oxide layers.
(d) Corresponding magnetic field profiles for the resonances shown in (c). The thickness of
the top Al layers is 50nm, W=300nm and P=500nm for both geometries.
4.6. Conclusion
In conclusion, we have demonstrated the use of Al and its native Al 2O3 to
fabricate MIM resonators spanning the visible and IR wavelengths using top-down
fabrication techniques. Large area plasmonic surfaces are fabricated by depositing
thin Ag films near percolation threshold and the resonance wavelength is tuned by
longer exposure of Al layer to air. We fabricated hierarchical plasmonic surfaces
using multiple native oxide layers and the percolating Ag films and used these
surfaces as SERS and SEIRA substrates for the detection of monolayer DDT. Then
we have then improved the bandwidth of native oxide based plasmonic surfaces by
sequential deposition and oxidation of Al films through breaking the vacuum of the
physical vapor deposition system by exposing the Al films to air. Different
resonances are confined to different native oxide layers with different thicknesses.
82
Chapter 5
Interference Coating Based Sensing

Platforms
In the previous chapters we have designed and fabricated plasmonic structures to

study novel optical phenomena such as SERS, SEIRA and plasmonic hot-electron
effects. In this chapter, we shift our focus to interference coatings and their
applications. We present two novel applications of these types of coatings. For the
first application, we have developed a novel sensing platform in the IR. By utilizing
phase change property of Ge2Sb2Te5 (GST), tunable surfaces are fabricated and used
for sensing ultrathin (~5nm) polymer films, self-assembled and protein monolayers.
The performance of such surfaces is comparable or better than the plasmonic
counterparts. As a second application, we propose and fabricate a sensing platform
based on strong interference coatings. Unlike interference coatings, the thicknesses
of such surfaces are about t=λ/8n. The sensing mechanism is based on the color
change due to accumulation of bio-molecules on the sensing surface. Protein
monolayer and bilayers are detected using these surfaces.
5.1. Thermally Tunable Ultrasensitive Vibrational

Spectroscopy Platforms Based on Thin Phase Change
Films
Infrared spectroscopy enables identification of materials through characterization
of the vibrational modes of molecular moieties which are typically observed as small
dips in the reflection/transmission spectrum. The vibrational mode signals increase
with enhancement of electric field in the vicinity of the probe molecules. Plasmonic
83
surfaces with very high but localized field enhancements are typically used for
sensing of small amounts of materials. Here, we propose using thin, unpatterned
phase changing GST films deposited on wafer-scale metal ground planes with
modest field enhancement factors (|E|2/|Eo|2~4) on the entire GST surface for
ultrasensitive vibrational spectroscopy. Optical properties of the top GST layer are
tuned by annealing the surfaces leading to tuning of the wavelengths at which
electric field is enhanced. The ease of fabrication combined with ultra-sensitivity of
these surfaces offer convenient optical platforms for the infrared vibrational
spectroscopy of materials ranging from monolayers to µm thickness in size.
5.1.1. Introduction
Infrared vibrational spectroscopy is one of the widely used tools in

chemistry[235] and biology[62], [236] for characterization of materials. The
molecular signatures of materials can be sensed in the far-field signals (reflection or
transmission spectrum). Current infrared measurement techniques such as Fourier
Transform Infrared Spectroscopy (FTIR) require large amounts of molecules to
collect significant information. Attenuated total internal reflection[237] and grazing
angle objective[65] methods are commonly used to enhance the signals due to
vibrational modes of small amount of materials to be sensed. The intensity of far-
field molecular vibrational bands scale with square of local electric field intensity
(|E|2) as stated by the Beer's Law[238]. The wavelength dependent percentage of
absorbed power can be estimated as
𝑑𝑉 4𝜋𝑛(λ)𝑘(λ)
𝐴(λ) = ∫0 |𝐸(𝑟, λ)|2 𝑑𝑉 (5.1)
λ
Thus, increasing near-field intensities in the vicinity of the probe molecules

results in increased signal intensities of the vibrational modes. Recently, rough metal
surfaces[58], patterned metal structures[239] and nanoantennas[240] have been
demonstrated to enhance the vibrational modes of molecular monolayers in the
infrared by increasing field intensities through exiting plasmons. This technique is
widely known as Surface Enhanced Infrared Absorption Spectroscopy (SEIRA). In
SEIRA, patterned plasmonic structures that have resonances close to the vibrational
bands of a molecule are used to enhance the far-field signal intensities. Depending on
the vibrational bands to be enhanced, the geometry of the plasmonic structure has to
be tuned[67]. Smart designs with multispectral[233] responses have been
demonstrated to enhance different infrared bands of a molecule.
84
Plasmonic surfaces used in SEIRA are usually fabricated using e-beam and
optical lithography and multi-step processes, limiting the sensing area and increasing
the cost. Although there are easier methods to fabricate plasmonic surfaces such as
interference[26] and nanosphere lithography[241], they typically do not allow
wavelength tunability over large bandwidths. Field enhancement factors (|E|2/|Eo|2) in
plasmonic structures are very high but localized. The field enhancement calculations
are based on the assumption that most of the contributing vibrational band signal
comes from the high field enhancement regions. Although SEIRA enhancement of
plasmonic structures are typically estimated to be in the range of 102-105, the far-
field signal intensity changes due to vibrational bands of monolayers are typically
small (<10%). Designs like pedestal shaped plasmonic antennas[242] have been
proposed to increase the far-field signal through increasing the enhanced field area
accessible to the probe molecules[243], [244]. Here we propose using entire surface
of an optical platform with modest field enhancements (~4) similar to what is
implemented in recent reports of electrically tunable graphene Salisbury
screens[245]–[247]. The optical platforms are formed by depositing thin dielectric
layers (Si and GST) on metal films which show optical resonances accompanied by
field enhancements on the dielectric surfaces in the mid-infrared (MIR) regime.
Thickness (t) of the lossless dielectric films (amorphous Si and GST) is determined
as λV/4n (λV: vibrational mode wavelength, n: refractive index). Crystalline GST
exhibits a non-zero extinction coefficient (k) in the IR resulting in strong interference
effects[74] (t< λV/4n). Very thin (5 nm) films of poly(methyl methacrylate) (PMMA)
and monolayers of bovine serum albumin (BSA) and octadecanthiol (ODT)
molecules are used to test the performance of the surfaces. The studied surfaces are
proposed as highly-sensitive, angle-independent infrared spectroscopy platforms
which can be fabricated with low cost, very fast owing to the simple structure with
two continuous layers, without patterning, on large-area rigid or flexible substrates,
e.g., Al foil, and can be tuned by varying the thickness or phase of the dielectric
layer. Unlike plasmonic surfaces used in SEIRA, the proposed surfaces do not
exhibit Fano type vibrational bands due to the strong coupling of plasmonic and
vibrational modes. Hence, complicated post-processing methods required for
plasmonic structures are eliminated using these surfaces.
85
5.1.2. Results
A very-thin (~nm) suspended layer of a material enables the most basic infrared
spectroscopy method as the vibrational modes can be observed as small dips in the
transmission spectrum with an electric field enhancement factor of ~1 (Figure
5-1a(i),(ii)). For a ultrathin absorbing layer, the Equation 5.1 becomes 𝐴(λ) =
4𝜋𝑛(λ)𝑘(λ)/λ|𝐸(λ)|2 𝑡. The far-field signals of the vibrational modes are expected
to be detectable but small (Figure 5-1a (iii), (iv)). For 1732 cm-1 k~0.5, thus for
t=5nm intensity change in the far-field transmission is about 1%. This signal
intensity change is in the order of signal intensities (1-5%) obtained with plasmonic
surfaces. Achieving a suspended layer of any material, e.g., 5 nm PMMA, over a
rather large area (~mm), however, is not possible. A thin (~50 nm) suspended
membrane, such as Si3N4, can be used to address this problem while introducing
other issues such as limited area, fragility of the substrate, and observation of strong
signals from vibrational bands of the membrane material. The IR transmission
measurements of a 5 nm PMMA coated 50 nm thick Si3N4 membrane show
detectable but small signal for the major PMMA band at 1732 cm-1 and very strong
signals (~40%) for the Si-N vibrational bands (Figure 5-3).
86
Figure 5-1. Sensing performance of various optical platforms. 5 nm PMMA (a) suspended in
air, (b) on semi-continuous CaF2 substrate, (c) on 300 nm a-Si/Al surface, (d) on 400 nm a-
Si/Al surface. (i) Simulated electric field enhancement profiles on cartoon illustrations at λ:
5780 nm (1/λ: 1732 cm-1). (ii) Simulated electric field enhancement factors on the surface of
each structure, (iii) far-field signal (transmittance/reflectance) spectra, (iv) signal from the
major PMMA band. The signal is the difference between the far-field signals w/ and w/o
PMMA layer. Measured optical properties of PECVD deposited a-Si are used for the
simulations (n: 3.3 in the IR). Optical properties of PMMA are fitted to a Lorentzian
oscillator (Figure 5-2) centered around 5780 nm.
Figure 5-2. Measured refractive index and extinction coefficient of PMMA. The optical
parameters are extracted using spectroscopic IR ellipsometer measurements of 130 nm
PMMA film on a Si wafer using the Lorentzian oscillator model for each PMMA band. Only
the vibrational band at 5780 nm is used for the simulations.
87
Figure 5-3. Transmission spectra of 50 nm thick suspended SiNx membranes with and
without 5 nm PMMA layer. The vibrational band of SiNx is observed as a significant dip in
the transmission spectra. The signal from the PMMA band at 1732 cm-1, however, is small.
A thick IR transparent substrate, such as CaF2 or MgF2, can also be used to detect
the vibrational bands on the transmission spectrum (Figure 5-1b) with an
enhancement factor of ~0.7 on the surface. Although the vibrational mode signals are
expected to be smaller compared to the suspended layer configuration, such IR
substrates have the advantages of strength and large area. Alternatively vibrational
mode signals can be detected on the reflection spectrum using a reflective substrate
such as metal-coated Si wafer. Electric field enhancement factor increases from
almost zero on the metal surface to 4 at a distance quarter wavelength (λ/4) away
from the surface[245], [246]. The vibrational band signal is expected to be
maximized when the probe material is λV/4 away from the metal surface where the
enhancement factor is ~4. While this can be achieved by forming a thin membrane
over a metal surface, the most trivial way is using a dielectric coated reflective layer.
The enhancement factor at the air-dielectric interface in this case is given as
2
|𝐸|2 𝑟12 + 𝑟23 𝑒 2𝑖𝛽
=( + 1) (5.2)
|𝐸𝑜 |2 1 + 𝑟12 𝑟23 𝑒 2𝑖𝛽
where rxy=(ñx-ñy)/( ñx+ñy), ñ is the complex refractive index, β=(2π/λ)ñ2t, and layers
1, 2, and 3 are air, dielectric, and metal, respectively. The total reflection coefficient
(rt) is 1 when r12 is zero (n2=1) and decreases with increasing n2 (-1 < r12 < 0). Using
a relatively high refractive index material such a-Si requires smaller thickness but
slightly reduces the enhancement factor to ~3.7. If the dielectric thickness is not
88
exactly λV/4n2, the vibrational mode signal can still be sensed owing to non-zero
field enhancement (Figure 5-1c). The signal intensity is maximized for t= λV/4n2
(Figure 5-1d). The enhancement of electric field extends up to λ/4 above the dielectric
surface, enabling sensing of thicker probe materials, e.g., cells, which is typically not
possible for plasmonic surfaces due to low spatial extent (<100 nm) of field
enhancements[248]. For a certain dielectric material of λV/4n2 thickness, the
enhancement factor reaches its theoretical limit of 4 for r23=-1 which is achieved for
a perfect electrical conductor (PEC) layer (Figure 5-4). Hence the more closely a
metal behaves like a PEC in the MIR, the higher the enhancement factor is.
Simulation results for various metals, including Al, Au and Ag, show that Al and Au
can achieve the highest enhancement factors (~ 3.7) in the MIR for a-Si dielectric
layer. Hence, Al is chosen as the ground plane material owing to its low cost,
abundance and CMOS compatibility for the fabrication of the surfaces. Using Al
ground planes enables comparing IR absorption and sensing performances of these
surfaces to those of based on thin Al foils which are cheap, bendable and large area
substrates.
89
Figure 5-4. Simulated electric field enhancements on the surfaces of a-Si layers with various
thicknesses on metals and a perfect electrical conductor (PEC) layer. The enhancement
factor increases with increasing wavelength for metals as their optical properties approach
those of a perfect electrical conductor.
The initial surfaces are fabricated using amorphous silicon (a-Si) as the dielectric
layer on Al coated Si substrates following the simulations in Figure 5-1. The surfaces
are coated with 5 nm PMMA layer, used as the probe material to test the sensing
performance of the surfaces. The a-Si thickness is chosen to tune the surface’s first
order resonance to be close to the PMMA’s strongest vibrational band at 1732 cm -1
(λ~5.77 µm). The center to edge variation in the a-Si thickness (400 – 460 nm)
results in resonances in between 5.3 and 6.1 µm, revealing the major PMMA band
with signal intensities of 6-7 % after background subtraction (Figure 5-5). The
sensing performance of the surface is still good at extreme incident angles (75̊) for
both TE and TM polarizations (Figure 5-6), enabling relaxed design options for
sensing systems based on these surfaces. These surfaces are successful at identifying
PMMA, however also exhibit very strong Si-H vibrational bands[238], [249] at
90
~2010 cm-1. The presence of Si-H bands is attributed to SiH4 gas used for the
deposition of the films in a plasma enhanced chemical vapor deposition (PECVD)
system. The strong signal due to the Si-H vibrational bands would prevent sensing
materials with vibrational bands in this regime, limiting the use of the surfaces
formed by PECVD deposited a-Si layers. The Si-H and PMMA vibrational bands in
the reflection spectrum are not observed for PECVD deposited, thinner a-Si layers
whose primary resonances are at lower wavelengths (Figure 5-7).
Figure 5-5. Experimental demonstration of infrared sensing using a-Si layers with different
thicknesses (400-460 nm). (a) Reflectance spectra of the as fabricated (black) and 5 nm
PMMA coated surfaces (red). Arrows mark the Si-H band at 2010 cm-1 and PMMA and 1732
cm-1. (b) The signal intensities after background subtraction. The background signal is
calculated by smoothing the reflectance curves using the moving average method. The signal
intensity decreases as the resonance shifts away from the major PMMA band.
91
Figure 5-6. Angle dependent infrared ellipsometer measurement results for 5 nm PMMA
coated 400 nm PECVD Si on Al surface (The normal incidence results are shown in Figure 2
of the main document). The incident angle is varied between 25̊ to 75̊ with 5̊ steps. The
signal intensity due to the PMMA vibrational band at 1732 cm-1 reduces to %1.7 at 75̊ from
% 5.6 at 25̊. For TM polarization, the resonances and the signals from the PMMA and Si-H
bands are strong up to high incident angles.
Figure 5-7. Reflection spectra of 5 nm PMMA coated PECVD deposited a-Si surfaces with
170 and 460 nm thicknesses on Al. No significant Si-H or PMMA bands are observed for
170 nm a-Si surface.
92
5.1.3. Thermally Tunable IR Spectroscopy Platforms Using Ge2Sb2Te5
Ge2Sb2Te5 (GST) in amorphous phase has a similar refractive index function to

that of a-Si in the MIR and does not have a-Si layer’s inherent surface problems.
Chalcogenides, in general, are preferred for IR applications owing to their high IR
transparency, low temperature-coefficient of refractive index and low
dispersion[250]. GST, in particular, is a phase-change material, exhibiting
amorphous to face-centered cubic structure (c-GST) transition at ~150 C which is
accompanied by a large change in its electrical and optical properties. Phase-change
material based electrical data storage (phase-change memory, PCM) devices, which
target the non-volatile memory market, take advantage of the contrast between the
electrical resistivity of the material in amorphous and crystalline phases[251]. The
contrast in the optical properties of phase-change materials like GST has been used
in the optical data storage media for decades[252]. Although optical properties of the
phase-change materials have long been studied, number of reports on optical
devices/surfaces using these materials have recently increased due to their potential
use in plasmonic applications in the IR[253], [254] and strong-interference-effect
based applications in the visible regime[255], [256]. GST thicknesses used in these
devices/surfaces are typically small (7 - 50 nm). A rather thick GST (>80 nm) on a
metal layer exhibits optical resonances in the IR[257]. The optical responses of these
surfaces are tuned over a larger wavelength range owing to large refractive index
contrast between amorphous and crystalline phases (Figure 5-8).
93
7
c-GST
6
5
a-GST
4
n
1
0 1000 2000 3000 4000 5000
4 c-GST
3
k
1 a-GST
0
0 1000 2000 3000 4000 5000
λ (nm)
Figure 5-8. Refractive index and extinction coefficient of a-GST and c-GST films. The
optical parameters are extracted by spectroscopic ellipsometer measurements of GST films
on Si wafers at various angles in the 300 – 1700 nm range. The optical parameters are
extracted within and beyond this wavelength range using the Tauc-Lorentz oscillator model
for a-GST films and the Tauc-Lorentz + Drude oscillator models for c- GST films. The
refractive index of c-GST predicts higher resonance wavelengths than what is observed in
measurements. This is attributed to diffusion of Al into GST and reducing the effective
refractive index of the GST layer during crystallization of the GST films. Diffusion of Al
into GST can be avoided by introducing a thin barrier layer, such as TiN, in between Al and
GST layers.
The measured reflection spectra of GST films on Al show great potential for the
sensing purposes as the near-IR (NIR) and MIR spectrum are covered by varying the
GST thickness and can be further tuned by crystallizing the films (Figure 5-9).
Crystallization of a-GST layer consistently red-shifts the resonance owing to the
larger refractive index of c-GST. When crystalline to amorphous volume ratio is
changed gradually, instead of complete crystallization, a mixed optical response is
94
achieved. Partial crystallization is performed by laser annealing of different portions
of a-GST films on Al surfaces (Figure 5-9).
Figure 5-9. NIR and MIR results for GST layers on Al. (a) Reflection spectra of a-GST
(black) and c-GST (blue) covered Al surfaces for indicated GST thicknesses. The
crystallized films are expected to be 5 % thinner than the as-fabricated a-GST films[257]. (b)
Calculated electric field (|E|2/|Eo|2) enhancements across cross-sections of 350 nm a-GST and
c-GST films on Al and corresponding reflection spectra.
Alternatively, the gradual crystallization can be achieved by controlling the

annealing duration of a-GST films at a high temperature (<150 C), which requires
simultaneous monitoring of the optical response of the surfaces. For sensing
experiments, only complete amorphous or complete crystalline films are used, since
the areas of the laser scanned surfaces are limited to 100 x 100 µm2. The calculated
maximum field enhancement factors on the surface of 350 nm films is ~3.5 for a-
GST and ~1.4 for c-GST (Figure 5-9). The lower enhancement factor for c-GST, at
the resonance is attributed to the high refractive index (n) and non-zero extinction
coefficient (k) of c-GST, which makes c-GST a lossy dielectric in the MIR. As a
result, crystallizing a-GST films enables tuning the resonance to a higher wavelength
95
while reducing the enhancement factor. The higher order resonances for both phases
are observed to be sharp with field confinements within the films (Figure 5-9).
Figure 5-10. Optical microscope image and reflection spectra of 100 µm x 100 µm GST
surfaces (t: 350 nm) on Al. Crystalline to amorphous ratio increases from (i) to (iv), (i) being
completely amorphous and (iv) being completely crystalline. Lighter color lines are
crystalline regions. The surface is laser-scanned using a green laser with 1 µm spot size
equipped on a Raman system.
Sensing performance of the GST surfaces are tested using very thin (5 nm)
PMMA layer on GST films with 200 and 350 nm thicknesses (Figure 5-11) which
target the PMMA vibrational bands around 3000 and 1500 cm-1, respectively. 200
nm a-GST surface reveals the bands at 2997, 2952 and 1732 cm-1. 200 nm c-GST
surface shows the PMMA bands at 2997, 2952 cm-1 with lower signal intensities, and
the 1732 cm-1 band with a higher signal intensity owing to its higher wavelength
resonance (Figure 5-11). The major PMMA band at 1732 cm-1 is observed as ~7 %
drop in the reflection spectrum of 350 nm a-GST surface. 350 nm a-GST is
particularly good at sensing all the vibrational bands larger than 1000 cm-1. 350 nm
c-GST surface exhibits a smaller signal at 1732 cm-1 due to its smaller enhancement
factor despite its deeper resonance and performs better for higher wavelength PMMA
bands such as 1151 cm-1 compared to 350 nm a-GST surface. Monolayer sensing
96
performance of 350 nm a-GST surface is tested by binding BSA on the surface. The
BSA vibrational bands at 1652 and 1531 cm-1 are observed on the reflection
spectrum of 350 nm a-GST surface with a signal intensity of 6 % at 1652 cm-1. This
signal intensity is larger than 3-4 % which are reported for plasmonic surfaces in the
literature[64], [240]. The GST surfaces are further tested with ODT molecules which
forms a monolayer on Au. 250 nm a-GST surface is functionalized by depositing 1.5
nm Au on top. The top Au layer is chosen to be very thin to form Au nano-islands
instead of a continuous film on the GST surface in order not to modify the optical
response significantly (Figure 5-12). Both ODT bands at 2849 and 2917 cm-1 are
sensed by 250 nm a-GST and c-GST surfaces. The signal due to the ODT band at
2917 cm-1 is 3 % for 250 nm a-GST surface, despite a slight mismatch between the
surface’s resonance and the targeted ODT bands (Figure 5-13), being similar to the
enhanced performance of pedestal gold structuress from the literature[242].
Figure 5-11. Sensing performances of GST surfaces. (a) Reflection spectra of 200 and 350
nm GST surfaces covered with 5 nm thick PMMA or monolayer BSA. (b) The signal
intensities after background subtraction. The background signal is calculated by smoothing
the measurement curves using the moving average method. The sensed vibrational bands of
PMMA and BSA are marked with dashed lines and labeled.
97
Figure 5-12. Scanning electron microscopy (SEM) images of 350 nm a-GST films on Al (a)
with and (b) without 1.5 nm Au top layers. (c) SEM image of 1.5 nm Au on Si.
98
Figure 5-13. (a-b) Reflection spectra of ODT covered 1.5 nm Au/250 nm GST/Al surfaces.
Arrows mark the ODT bands at 2849 and 2917 cm-1. (c) Signal intensities due to the ODT
bands after background subtraction. Signal intensities at 2917 cm-1 are % 3 for a-GST and %
1 for c-GST surface.
5.2. Strong Interference Based Colorimetric Sensors

5.2.1. Introduction
Label free detection of biomolecular interaction has attracted great attention due
to its importance for pharmacological and biomedical applications in both health and
research areas[1], [32], [258]–[262]. Different types of label-free sensing platforms
99
have been developed to overcome the difficulties that the label-based sensing
platforms face[33], [55], [263]. Surface plasmon based sensors are among the most
studied label-free sensors[33], [56], [224], [264]–[266]. Different types of platforms
utilizing surface plasmons have been developed. Surface enhanced Raman
spectroscopy (SERS)[210], [267]–[270] and surface enhanced infrared absorption
spectroscopy (SEIRA)[63], [64], [67] use the vibrational bands of molecules that
interact with localized or surface plasmons. Although, in SERS, the scattered Raman
signal is too complicated to determine specific biomolecular interactions, bio-
detection at the single molecule sensitivity is possible[198], [211], [270]. On the
other hand, in SEIRA the vibrational signatures of different molecules exhibit similar
bands limiting the probing of biomolecular interactions. Other than vibrational
spectroscopies, surface plasmon sensors (SPR) emerges as one of the prominent
label-free sensor technologies by taking advantage of local refractive index change
during biomolecular binding events[1], [33], [224], [265], [271]–[273]. SPR
technology is further improved for probing multiple biomolecular events by
incorporating imaging equipment to the system[56]. Generally, these types of SPR
systems rely on the reflected light intensity that is captured by a camera and is a type
of refractometric sensor. Among the refractometric sensors, interference based
refractometric sensors draw attention due to their ease of use and lab-on-chip
compatibility[258], [274], [275]. These types of sensors rely on the local refractive
index change due to the change in the thicknesses of bio-molecules. These sensors
are typically fabricated on silicon chips and composed of thick layer of oxide which
exhibits interference fringes in the reflection spectra. The biomolecular interactions
are investigated by illuminating the chip with either a broadband or narrow light
source and collecting the reflected light with either a spectrometer or a camera.
Colorimetric sensing mechanism is in the core of the current label-based sensing
mechanisms such as ELISA[56], [261], [276]. Recently, a new type of sensing
mechanism using plasmonic nano-particles has emerged[272], [277]. These type of
sensors are based on the chemical-reaction driven color change occurred due to the
presence of target molecules due to either change in size of particles or the distance
between them. Metal ions and bio-molecules have been detected using these type of
sensors[278].
A reflective surface, such as a thin metal film, covered with a thin layer of lossy
dielectric show optical resonances (strong interferences) and can show perfect
100
absorption at certain wavelengths depending on the complex refractive indices (n+ik)
of the metal and dielectric layers. The strong interference resonances are due to
destructive interference of the light reflected form the top and bottom of the lossy
dielectric layer with phases different from 0 or π. The analysis of the strong
interference effects and the n and k conditions which provide resonances are
elucidated by Kats et al[74]. Unlike the thin film interference coatings (t≈ λ/4n) these
surfaces exhibit resonances for thinner dielectric coatings (t≈ λ/8n). Thinner
dielectric coatings also enable a wide angular spectral response due to small phase
accumulation. Since then, strong interference effects have been investigated for
coloring surfaces using gold and germanium[74], tunable, dynamic infrared filters
using sapphire and VO2[279], light sensing using Ag and amorphous Si[77] and
advanced optical data storage using Au and Ge2Sb2Te5 (GST)[280]. So far, most of
the studies based on these types of coatings are utilized for color printing
applications. In this part of the thesis a novel colorimetric sensor based on
interference coatings using highly absorptive layer is presented. Unlike the
refractometric interference sensors, a lossy dielectric layer is deposited on top of a
mirror or metal surface. The proposed sensor surfaces can be fabricated on large
areas and have wide angle optical response. Ultra-thin dielectric coatings and protein
monolayer are detected using these sensors.
5.2.2. Results
The proposed sensing surface is shown in Figure 5-14. The structure consists of a
thin amorphous silicon (~20nm) layer on top of a thick (~80 nm) aluminum film.
Similar surfaces using different absorptive dielectric films have been demonstrated to
have resonances in the visible spectrum. Such surfaces are utilized mostly for color
printing applications. It has been shown that even small variations in the thickness of
absorptive dielectric films with high refractive indices results in color change. The
proposed sensing principle is based the color change due to molecular binding events
as shown in Figure 5-14a. The optical microscope image of a fabricated surface is
shown in the inset of Figure 5-14b. The image is taken under halogen light
illumination with 5x magnification. Since the surface color strongly depends on the
spectrum of the illumination light, the surface shows different colors under different
light sources. For this work, we used a halogen bulb integrated to optical microscope
for all measurements. The reflection spectrum shows a resonance with almost perfect
101
absorption at ~510nm as shown in Figure 5-14b. The resonance, hence the color of
the surface, is wide angle for both polarizations (Figure 5-14 c, d). The resonance
wavelength of the surface can be tuned by increasing the thickness of a-Si layer.
Thicker a-Si thickness also results similar reflectance spectra due to excitation of
higher order interference modes (See Figure 5-15). However, wavelength shift due to
dielectric layer for thicker a-Si layer negligible due to higher electric confinement to
the a-Si layer and smaller electric field intensity at air/a-Si interface.
Figure 5-14. Optical characterization and the working principle of the proposed sensing
platforms. (a) Cross section illustration of the sensing platform. For broadband excitation
(white light), surfaces reflects distinct color. The color of reflected light changes upon
deposition of a very thin dielectric layer (~3-5nm) or protein monolayer. (b) Reflection
spectra of surfaces. (Inset) Microscope image of the surface illuminated with halogen light.
Measured reflectance spectra map for (c) s and (d) p polarizations.
102
Figure 5-15. Comparison of sensing performances of surfaces with different a-Si thicknesses.
(a) Simulated reflected spectra for 20nm and 80nm a-Si on Al. Simulated reflectance spectra
without(solid curve) and with(dashed curve) 5nm dielectric layer(n=1.5) for 20nm(b) and
80nm a-Si thicknesses.
In order to characterize the colorimetric sensing performance, dielectric films with

varying thicknesses deposited on top the surfaces. Similar experiments have been
performed by depositing Al2O3 on germanium coated gold films recently to enhance
the color contrast by e-beam evaporation. Here, SiO2 films are deposited by e-beam
evaporation with different thicknesses (5, 10, 15 and 20nm) on our sensing surfaces.
The thickness and refractive index of deposited films are characterized by variable
angle spectroscopic ellipsometer (n~1.5). The reflectance spectra shows a red-shift in
the resonance wavelength when the thickness of the SiO2 increases (Figure 2 a, b).
The color of the sensing surfaces is shown in Figure 2c. The intensity of the blue
channel changes significantly due to increase in SiO2 thickness as shown in Table 1.
However, the green and red channels do not show a significant trend. The intensity
change of blue channel is attributed to the blue-shift in resonance wavelength where
blue portion (~400-450nm) of the spectrum is reflected more and the red portion of
the spectrum(~650-700nm) is reflected less. We have also deposited Al2O3 films by
atomic layer deposition (ALD) at 80oC with three different thicknesses (50, 100 and
150 cycles). The thicknesses of Al2O3 films are measured to be 5, 10 and 15nm using
ellipsometer. The refractive index of Al2O3 is close to that of SiO2 and proteins
(~1.5) which is different than the refractive index values reported in the literature.
Al2O3 deposited surfaces also show wavelength shift with increasing oxide thickness
similar to SiO2 deposited surfaces (See Figure 5-17). We have also demonstrated
color change by depositing ultrathin PMMA layer on the sensor surface. 5nm PMMA
film is spin-coated on the surfaces shows also color change as shown in Figure S6.
The square regions in Figure S6b are patterned by e-beam lithography and PMMA-
free region have different color than the PMMA coated regions.
103
Figure 5-16. Optical characterization of sensing platforms with different SiO2 thicknesses.
(a) Reflection spectra with 0, 5, 10, 15 and 20nm SiO2 thicknesses. (b) Reflectance spectra
around the resonance wavelength. (c) Optical images of the surfaces with (i) 0nm, (ii) 5nm,
(iii) 10nm, (iv) 15nm and (v) 20nm SiO2 thickness. (vi) SiO2 spots with different thicknesses
are deposited on the sample surfaces with a shadow mask
Table 1. RGB values of sensing platforms with different SiO2 thicknesses.
0nm SiO2 5nm SiO2 10nm SiO2 15nm SiO2 20nm SiO2
R 221 222 225 223 218
G 94 85 84 89 98
B 58 69 97 110 117
104
Figure 5-17. Optical characterization of sensing platforms with different Al2O3 thicknesses.
(a) Reflection spectra with 0, 5, 10 and 15nm Al2O3 thicknesses. (b) Reflectance spectra
around the resonance wavelength. (c) Optical images of the surfaces with (i) 0nm, (ii) 5nm,
(iii) 10nm and (iv) 15nm Al2O3 thickness. (d) Intensity values for RGB channels for different
SiO2 thicknesses.
Figure 5-18. Optical characterization of sensing platforms with PMMA coating. (a)
Reflection spectra with and without 5nm PMMA coating. (b) Optical image of the PMMA
coated surface. The square patterns are PMMA-free regions.
105
5.2.3. Protein Binding Experiment
In order the test the biomolecular interaction of our sensor platform, we

performed binding event using mono and bi-layer bovine serum albumin (BSA). For
monolayer and bi-layer BSA coatings, we adopted a well known procedure[281].
The schematic of protein binding experiment is shown in Figure 5-19a. The sensor
surfaces are immersed in BSA/citrate buffer solution for 2 hours for monolayer
formation. Then the surfaces are washed with citrate buffer to remove un-bound BSA
proteins. The surfaces are immersed in dextrane sulfate/citrate buffer solution for 1
hour to negatively charge the positively charged BSA surface for BSA layer coating.
The reflectance spectra of BSA coated surfaces are shown in Figure 5-19b. The
resonance wavelength red-shifts for monolayer BSA coating (∆λ~8nm). After the
second BSA layer the resonance wavelength shift is about (∆λ~18nm). Higher
resonance wavelength shift for the second BSA layer is attributed to the different
optical properties and thicknesses of positively and negatively charged BSA layers.
There is a significant difference between the surfaces as shown in Figure 5-19c. For
bi-layer BSA coated surface, the color is close to purple due to the increased
reflected light intensity at shorter wavelengths (~400-450nm). The surfaces can also
be differentiated from the photographs of the same surfaces captured with a mobile
phone's camera as shown in Figure 5-19d. The color change due to protein binding is
also be detected with naked eye.
106
Figure 5-19. Colorimetric sensing of protein binding events. (a) Schematic of protein binding
experiments. (i) BSA monolayer is coated on the sensor surface by physobsorption.(ii)
Binding dextrane layer to BSA protein to negatively charge the BSA protein surface. (iii)
Second BSA layer is covalently bound to dextran layer. (b) Reflectance spectra for bare
surface, monolayer and bi-layer BSA coated surfaces. (c) Optical microscope images of bare
(i), monolayer (ii) and bi-layer (iii) BSA coated surfaces. (d) Photograph of the same surface
taken with a cell phone camera.
5.3. Conclusion
In the first part of this chapter, we propose using thin dielectric layers, such as
aGST and a-Si, coated metal ground planes as optical platforms for ultrasensitive IR
vibrational spectroscopy measurements. Switching to a-GST overcomes the a-Si
layer’s drawbacks and enables further tuning of the optical resonance by
transitioning to c-GST at ~150 C. Crystallization of GST films on Al layers red-
shifts the resonance owing to the higher refractive index and non-zero extinction
coefficient of c-GST while the vibrational band signals at higher wavelengths are
enhanced. The GST surfaces are successful at sensing monolayers of BSA and ODT,
proving themselves as highly sensitive IR sensing platforms. The proposed surfaces
are fabricated very fast with one deposition step, without any patterning process and
on large substrates such as Al coated Si substrates or thin Al foils. High sensitivity,
angle- independent response and large area of the surfaces reduce the strict IR
sensing measurement requirements such as the need for magnification or specific
107
measurement angle which are typically used for lithographically defined small-area
surfaces. While the cutting edge plasmonic based SEIRA studies aim to reach single
molecule sensitivity[67] by extremely enhancing localized fields, the proposed
surfaces perform better at detecting of ultra-thin and monolayers of materials with
further advantages of fast, inexpensive, easy fabrication, wafer-scale dimensions, and
thermal tunability.
In the second part, we have demonstrated a novel colorimetric sensor platform
based on strong interference effects exhibited by highly absorbing dielectric films on
metal surfaces. First, the sensing performance is tested using different dielectric
coatings with different thicknesses. The resonance wavelength shifts of the surfaces
are also prominent which can be used as sensing mechanism. The color change can
easily be detected after very thin dielectric film deposition. Then, protein binding
experiments are performed through monolayer and bi-layer BSA coatings on the
sensing surfaces. Although our experiments are performed in air with monolayer and
bi-layer BSA protein coatings, real time binding experiments can be performed by
functionalizing the surfaces with protein antibodies. Since, there are no nano-
fabrication process, e-beam or optical lithography steps, these sensor platforms can
be fabricated on laboratory glass slides over large areas in large scales. These sensor
surfaces can be integrated to mobile phone based o portable point of care devices
which can be advantageous in low-resource areas for the detection of disease
biomarkers.
108
Chapter 6
Conclusion
In this thesis, we first demonstrate simple coupled plasmonic geometry in the

visible spectrum. A simple electrical circuit model is developed to understand optical
properties. Surface enhanced Raman spectroscopy (SERS) is one of the important
applications of these of plasmonic surfaces since multiple resonances exhibited by
these surfaces can be exploited for either different laser excitation wavelength or
Raman wavelengths. Confocal Raman experiments have shown that these surfaces
exhibit unity Raman enhancement over large areas due to subwavelength
periodicities. Although, these coupled plasmonic surfaces display broadband and
multispectral spectral response, the use of e-beam lithography limits the fabrication
of these surfaces over large areas. Similar plasmonic geometry with nanoparticle top
layers can be fabricated over large areas showing similar optical properties using
simple physical deposition methods. Fabricated surfaces perfectly absorb light for
almost the whole visible spectrum up to 60 degrees angle of incidence. Single Raman
events are detected using nanoparticle based coupled plasmonic surfaces using
confocal Raman system. Single molecule Raman events can also be detected using a
mobile phone's camera either with camera or spectrometer mode. Similar surfaces
with thinner insulator layers are used to study plasmon enhanced hot electron effects.
Hot-electron effects are probed using an XPS system with external laser excitation
without using electrical contacts. The energy shifts of photoelectrons caused by laser
excitation are used to extract photoresponsivities. The photoresponsivities display
fair values, when compared to hot-electron currents of similar geometries. Aluminum
based metal-insulator-metal plasmonic surfaces are fabricated by exploiting the
native oxide layer to show resonances in the visible and IR wavelengths. The
resonance wavelengths are simply tuned by exposing the surfaces to air for longer
durations. The thicknesses of native oxide layer are characterized by XPS.
109
Nanoparticle based hierarchical aluminum surfaces are used SERS and SEIRA
applications. All aluminum hierarchical surfaces are demonstrated with multiple
resonances in the IR using native oxide layer as the spacer layer. In the last part, two
novel applications of regular and strong interference coatings are demonstrated.
Tunable surfaces are fabricated using phase change materials in the IR wavelengths.
By simply heating the surfaces up-to phase change temperatures, the resonance
wavelength is tuned. We have demonstrated SEIRA using these surfaces. By tuning
resonance wavelength, different IR bands are enhanced. Amorphous silicon based
strong interference coatings are fabricated and used as colorimetric sensor platforms.
Monolayer and bilayer proteins can be detected using these surfaces as the change in
surface color.
110
Bibliography
[1] H. Im, H. Shao, Y. Il Park, V. M. Peterson, C. M. Castro, R. Weissleder, and

H. Lee, “Label-free detection and molecular profiling of exosomes with a
nano-plasmonic sensor.,” Nat. Biotechnol., vol. 32, no. 5, pp. 490–5, May
2014.
[2] J. Homola, S. S. Yee, and G. Gauglitz, “Surface plasmon resonance sensors:

review,” Sensors Actuators B Chem., vol. 54, no. 1–2, pp. 3–15, Jan. 1999.
[3] J. Homola, “Present and future of surface plasmon resonance biosensors.,”

Anal. Bioanal. Chem., vol. 377, no. 3, pp. 528–39, Oct. 2003.
[4] J. A. Dionne, E. Verhagen, A. Polman, and H. A. Atwater, “Are negative

index materials achievable with surface plasmon waveguides? A case study of
three plasmonic geometries,” Opt. Express, vol. 16, no. 23, p. 19001, Nov.
2008.
[5] E. Feigenbaum and M. Orenstein, “Perfect 4-way splitting in nano plasmonic

X-junctions,” Opt. Express, vol. 15, no. 26, p. 17948, Dec. 2007.
[6] M. C. Gather, K. Meerholz, N. Danz, and K. Leosson, “Net optical gain in a

plasmonic waveguide embedded in a fluorescent polymer,” Nat. Photonics,
vol. 4, no. July, pp. 457–461, 2010.
[7] R. F. Oulton, V. J. Sorger, D. a. Genov, D. F. P. Pile, and X. Zhang, “A hybrid

plasmonic waveguide for subwavelength confinement and long-range
propagation,” Nat. Photonics, vol. 2, no. 8, pp. 496–500, Jul. 2008.
[8] A. Sobhani, M. W. Knight, Y. Wang, B. Zheng, N. S. King, L. V Brown, Z.

Fang, P. Nordlander, and N. J. Halas, “Narrowband photodetection in the
near-infrared with a plasmon-induced hot electron device.,” Nat. Commun.,
vol. 4, p. 1643, Jan. 2013.
[9] A. Akbari, R. N. Tait, and P. Berini, “Surface plasmon waveguide Schottky

detector,” Opt. Express, vol. 18, no. 8, p. 8505, Apr. 2010.
[10] A. Akbari and P. Berini, “Schottky contact surface-plasmon detector

integrated with an asymmetric metal stripe waveguide,” Appl. Phys. Lett., vol.
95, no. 2, p. 021104, Jul. 2009.
[11] V. E. Babicheva, N. Kinsey, G. V. Naik, M. Ferrera, A. V. Lavrinenko, V. M.

Shalaev, and A. Boltasseva, “Towards CMOS-compatible nanophotonics:
Ultra-compact modulators using alternative plasmonic materials,” Opt.
Express, vol. 21, no. 22, p. 27326, Nov. 2013.
111
[12] T. Zentgraf, Y. Liu, M. H. Mikkelsen, J. Valentine, and X. Zhang, “Plasmonic
Luneburg and Eaton lenses.,” Nat. Nanotechnol., vol. 6, no. 3, pp. 151–5, Mar.
2011.
[13] J. Qi, X. Dang, P. T. Hammond, and A. M. Belcher, “Highly efficient

plasmon-enhanced dye-sensitized solar cells through metal@oxide core-shell
nanostructure.,” ACS nano, vol. 5, no. 9. pp. 7108–16, 27-Sep-2011.
[14] M. W. Knight, H. Sobhani, P. Nordlander, and N. J. Halas, “Photodetection

with active optical antennas.,” Science, vol. 332, no. 6030, pp. 702–4, May
2011.
[15] J. B. Lassiter, H. Sobhani, M. W. Knight, W. S. Mielczarek, P. Nordlander,

and N. J. Halas, “Designing and deconstructing the fano lineshape in
plasmonic nanoclusters.,” Nano Lett., vol. 12, no. 2, pp. 1058–62, Feb. 2012.
[16] M. W. Knight, L. Liu, Y. Wang, L. Brown, S. Mukherjee, N. S. King, H. O.

Everitt, P. Nordlander, and N. J. Halas, “Aluminum plasmonic
nanoantennas.,” Nano Lett., vol. 12, no. 11, pp. 6000–4, Nov. 2012.
[17] N. Yamamoto, S. Ohtani, and F. J. García de Abajo, “Gap and Mie plasmons
in individual silver nanospheres near a silver surface.,” Nano Lett., vol. 11, no.
1, pp. 91–5, Jan. 2011.
[18] H. Raether, Surface Plasmons on Smooth and Rough Surfaces and on

Gratings, vol. 111. Springer-Verlag: Berlin, 1988.
[19] A. J. Pasquale, B. M. Reinhard, and L. Dal Negro, “Engineering photonic-

plasmonic coupling in metal nanoparticle necklaces.,” ACS Nano, vol. 5, no.
8, pp. 6578–85, Aug. 2011.
[20] R. Scattering, F. Le, D. W. Brandl, Y. A. Urzhumov, H. Wang, J. Kundu, N. J.

Halas, J. Aizpurua, and P. Nordlander, “Metallic Nanoparticle Arrays : A
Common,” vol. 2, no. 4, pp. 707–718, 2008.
[21] N. Liu, M. Mesch, T. Weiss, M. Hentschel, and H. Giessen, “Infrared perfect

absorber and its application as plasmonic sensor.,” Nano Lett., vol. 10, no. 7,
pp. 2342–8, Jul. 2010.
[22] X. Liu, T. Tyler, T. Starr, A. Starr, N. Jokerst, and W. Padilla, “Taming the
Blackbody with Infrared Metamaterials as Selective Thermal Emitters,” Phys.
Rev. Lett., vol. 107, no. 4, Jul. 2011.
[23] E. J. Smythe, M. D. Dickey, J. Bao, G. M. Whitesides, and F. Capasso,

“Optical Antenna Arrays on a Fiber Facet for in Situ Surface-Enhanced
Raman Scattering Detection 2009,” Nano Lett., vol. 9, no. 3, pp. 1132–1138,
2009.
112
[24] S. Rodriguez, a. Abass, B. Maes, O. Janssen, G. Vecchi, and J. Gómez Rivas,
“Coupling Bright and Dark Plasmonic Lattice Resonances,” Phys. Rev. X, vol.
1, no. 2, pp. 1–7, Dec. 2011.
[25] A. Kocabas, G. Ertas, S. S. Senlik, and A. Aydinli, “Plasmonic band gap

structures for surface-enhanced Raman scattering,” Opt. Express, vol. 16, no.
17, p. 12469, Aug. 2008.
[26] A. A. Yanik, A. E. Cetin, M. Huang, A. Artar, S. H. Mousavi, A. Khanikaev,

J. H. Connor, G. Shvets, and H. Altug, “Seeing protein monolayers with naked
eye through plasmonic Fano resonances.,” Proc. Natl. Acad. Sci. U. S. A., pp.
1–6, Jun. 2011.
[27] G. H. Chan, J. Zhao, E. M. Hicks, G. C. Schatz, and R. P. Van Duyne,

“Plasmonic Properties of Copper Nanoparticles Fabricated by Nanosphere
Lithography,” Nano Lett., vol. 7, no. 7, pp. 1947–1952, Jul. 2007.
[28] B. Kasemo, C. Langhammer, and Z. Michael, “Nanodisk Plasmons : Material

Damping Mechanisms,” ACS Nano, vol. 5, no. 4, pp. 2535–2546, 2011.
[29] S. Ayas, G. Cinar, A. D. Ozkan, Z. Soran, O. Ekiz, D. Kocaay, A. Tomak, P.

Toren, Y. Kaya, I. Tunc, H. Zareie, T. Tekinay, A. B. Tekinay, M. O. Guler,
and A. Dana, “Label-free nanometer-resolution imaging of biological
architectures through surface enhanced Raman scattering.,” Sci. Rep., vol. 3,
p. 2624, Jan. 2013.
[30] D. Wang, W. Zhu, M. D. Best, J. P. Camden, and K. B. Crozier, “Wafer-scale

metasurface for total power absorption, local field enhancement and single
molecule Raman spectroscopy.,” Sci. Rep., vol. 3, p. 2867, Jan. 2013.
[31] A. I. Maaroof and D. S. Sutherland, “Optimum plasmon hybridization at

percolation threshold of silver films near metallic surfaces,” J. Phys. D. Appl.
Phys., vol. 43, no. 40, p. 405301, Oct. 2010.
[32] B. Zhang, R. B. Kumar, H. Dai, and B. J. Feldman, “A plasmonic chip for

biomarker discovery and diagnosis of type 1 diabetes,” Nat. Med., vol. 20, no.
8, pp. 948–953, 2014.
[33] S. M. Tabakman, L. Lau, J. T. Robinson, J. Price, S. P. Sherlock, H. Wang, B.

Zhang, Z. Chen, S. Tangsombatvisit, J. a Jarrell, P. J. Utz, and H. Dai,
“Plasmonic substrates for multiplexed protein microarrays with femtomolar
sensitivity and broad dynamic range.,” Nat. Commun., vol. 2, p. 466, 2011.
[34] G. V Naik, V. M. Shalaev, and A. Boltasseva, “Alternative plasmonic

materials: beyond gold and silver.,” Adv. Mater., vol. 25, no. 24, pp. 3264–94,
Jun. 2013.
[35] J. Wang, Y. Chen, X. Chen, J. Hao, M. Yan, and M. Qiu, “Photothermal

reshaping of gold nanoparticles in a plasmonic absorber.,” Opt. Express, vol.
19, no. 15, pp. 14726–34, Jul. 2011.
113
[36] Y. C. Cao, R. Jin, and C. a Mirkin, “Nanoparticles with Raman spectroscopic
fingerprints for DNA and RNA detection.,” Science, vol. 297, no. 5586, pp.
1536–40, Aug. 2002.
[37] H. Liu, Y. M. Liu, T. Li, S. M. Wang, S. N. Zhu, and X. Zhang, “Coupled

magnetic plasmons in metamaterials,” Phys. status solidi, vol. 246, no. 7, pp.
1397–1406, Jul. 2009.
[38] R. Kekatpure, E. Barnard, W. Cai, and M. Brongersma, “Phase-Coupled

Plasmon-Induced Transparency,” Phys. Rev. Lett., vol. 104, no. 24, pp. 1–4,
Jun. 2010.
[39] P. Nordlander, C. Oubre, E. Prodan, K. Li, and M. I. Stockman, “Plasmon

Hybridization in Nanoparticle Dimers,” Nano Lett., vol. 4, no. 5, pp. 899–903,
May 2004.
[40] P. Nordlander and E. Prodan, “Plasmon Hybridization in Nanoparticles near

Metallic Surfaces,” Nano Lett., vol. 4, no. 11, pp. 2209–2213, Nov. 2004.
[41] J. B. Lassiter, J. Aizpurua, L. I. Hernandez, D. W. Brandl, I. Romero, S. Lal, J.

H. Hafner, P. Nordlander, and N. J. Halas, “Close Encounters between Two
Nanoshells 2008,” Nano Lett., vol. 8, no. 4, pp. 1212–1218, 2008.
[42] S. Y. Lee, J. J. Amsden, S. V Boriskina, A. Gopinath, A. Mitropolous, D. L.

Kaplan, F. G. Omenetto, and L. Dal Negro, “Spatial and spectral detection of
protein monolayers with deterministic aperiodic arrays of metal
nanoparticles.,” Proc. Natl. Acad. Sci. U. S. A., vol. 107, no. 27, pp. 12086–
12090, 2010.
[43] N. J. Halas, S. Lal, W. S. Chang, S. Link, and P. Nordlander, “Plasmons in

strongly coupled metallic nanostructures,” Chem. Rev., vol. 111, no. 6, pp.
3913–3961, 2011.
[44] A. Moreau, C. Ciracì, J. J. Mock, R. T. Hill, Q. Wang, B. J. Wiley, A.

Chilkoti, and D. R. Smith, “Controlled-reflectance surfaces with film-coupled
colloidal nanoantennas.,” Nature, vol. 492, no. 7427, pp. 86–9, Dec. 2012.
[45] C. Ciracì, X. Chen, J. J. Mock, F. McGuire, X. Liu, S.-H. Oh, and D. R.

Smith, “Film-coupled nanoparticles by atomic layer deposition: Comparison
with organic spacing layers,” Appl. Phys. Lett., vol. 104, no. 2, p. 023109, Jan.
2014.
[46] C. Ciracì, R. T. Hill, J. J. Mock, Y. Urzhumov, a I. Fernández-Domínguez, S.

a Maier, J. B. Pendry, a Chilkoti, and D. R. Smith, “Probing the ultimate limits
of plasmonic enhancement.,” Science, vol. 337, no. 6098, pp. 1072–4, Aug.
2012.
[47] J. Valentine, S. Zhang, T. Zentgraf, E. Ulin-Avila, D. A. Genov, G. Bartal,

and X. Zhang, “Three-dimensional optical metamaterial with a negative
refractive index.,” Nature, vol. 455, no. 7211, pp. 376–9, Sep. 2008.
114
[48] S. Xiao, V. P. Drachev, A. V Kildishev, X. Ni, U. K. Chettiar, H.-K. Yuan,
and V. M. Shalaev, “Loss-free and active optical negative-index
metamaterials.,” Nature, vol. 466, no. 7307, pp. 735–8, Aug. 2010.
[49] Y. Wang, T. Sun, T. Paudel, Y. Zhang, Z. Ren, and K. Kempa, “Metamaterial-

plasmonic absorber structure for high efficiency amorphous silicon solar
cells.,” Nano Lett., vol. 12, no. 1, pp. 440–5, Jan. 2012.
[50] A. Tittl, P. Mai, R. Taubert, D. Dregely, N. Liu, and H. Giessen, “Palladium-

based plasmonic perfect absorber in the visible wavelength range and its
application to hydrogen sensing.,” Nano Lett., vol. 11, no. 10, pp. 4366–9,
Oct. 2011.
[51] M. Khorasaninejad, S. Mohsen Raeis-Zadeh, H. Amarloo, N. Abedzadeh, S.

Safavi-Naeini, and S. S. Saini, “Colorimetric sensors using nano-patch surface
plasmon resonators.,” Nanotechnology, vol. 24, no. 35, p. 355501, 2013.
[52] H. Chalabi, D. Schoen, and M. L. Brongersma, “Hot-electron photodetection

with a plasmonic nanostripe antenna.,” Nano Lett., vol. 14, no. 3, pp. 1374–80,
Mar. 2014.
[53] Y. Chu, M. G. Banaee, and K. B. Crozier, “Double-Resonance Plasmon

Substrates for Surface-Enhanced Raman Scattering Stokes Frequencies,” ACS
Nano, vol. 4, no. 5, pp. 2804–2810, 2010.
[54] K. Chen, R. Adato, and H. Altug, “Dual-band perfect absorber for

multispectral plasmon-enhanced infrared spectroscopy.,” ACS Nano, vol. 6,
no. 9, pp. 7998–8006, Sep. 2012.
[55] A. J. Qavi, A. L. Washburn, J. Y. Byeon, and R. C. Bailey, “Label-free

technologies for quantitative multiparameter biological analysis,” Anal.
Bioanal. Chem., vol. 394, no. 1, pp. 121–135, 2009.
[56] O. Tokel, F. Inci, and U. Demirci, “Advances in plasmonic technologies for

point of care applications,” Chem. Rev., vol. 114, no. 11, pp. 5728–5752,
2014.
[57] D. L. Jeanmaire and R. P. Van Duyne, “Surface raman

spectroelectrochemistry,” J. Electroanal. Chem. Interfacial Electrochem., vol.
84, no. 1, pp. 1–20, Nov. 1977.
[58] T. R. Jensen, R. P. Van Duyne, S. A. Johnson, and V. A. Maroni, “Surface-

Enhanced Infrared Spectroscopy: A Comparison of Metal Island Films with
Discrete and Nondiscrete Surface Plasmons,” Appl. Spectrosc., vol. 54, no. 3,
pp. 371–377, Mar. 2000.
[59] A. D. McFarland and R. P. Van Duyne, “Single Silver Nanoparticles as Real-

Time Optical Sensors with Zeptomole Sensitivity,” Nano Lett., vol. 3, no. 8,
pp. 1057–1062, Aug. 2003.
115
[60] M. G. Banaee and K. B. Crozier, “Mixed dimer double-resonance substrates
for surface-enhanced Raman spectroscopy.,” ACS Nano, vol. 5, no. 1, pp.
307–14, Jan. 2011.
[61] E. Kim, F. Wang, W. Wu, Z. Yu, and Y. Shen, “Nonlinear optical

spectroscopy of photonic metamaterials,” Phys. Rev. B, vol. 78, no. 11, Sep.
2008.
[62] C. Kendall, M. Isabelle, F. Bazant-Hegemark, J. Hutchings, L. Orr, J. Babrah,

R. Baker, and N. Stone, “Vibrational spectroscopy: a clinical tool for cancer
diagnostics.,” Analyst, vol. 134, no. 6, pp. 1029–45, Jun. 2009.
[63] C. D. Andrea, A. Toma, C. Huck, F. Neubrech, E. Messina, E. Di Fabrizio, M.

Lamy, D. La Chapelle, P. G. Gucciardi, B. Fazio, and O. M. Marago, “Optical
Nanoantennas for Multiband Surface-Enhanced Infrared and Raman
Spectroscopy,” ACS Nano, vol. 7, no. 4, pp. 3522–3531, 2013.
[64] R. Adato, A. a Yanik, J. J. Amsden, D. L. Kaplan, F. G. Omenetto, M. K.

Hong, S. Erramilli, and H. Altug, “Ultra-sensitive vibrational spectroscopy of
protein monolayers with plasmonic nanoantenna arrays.,” Proc. Natl. Acad.
Sci. U. S. A., vol. 106, no. 46, pp. 19227–32, Nov. 2009.
[65] Y. Ishino and H. Ishida, “Grazing Angle Metal-Overlayer Infrared ATR

Spectroscopy,” Appl. Spectrosc., vol. 42, no. 7, pp. 1296–1302, Sep. 1988.
[66] J. Kundu, F. Le, P. Nordlander, and N. J. Halas, “Surface enhanced infrared

absorption (SEIRA) spectroscopy on nanoshell aggregate substrates,” Chem.
Phys. Lett., vol. 452, no. 1–3, pp. 115–119, Feb. 2008.
[67] L. V Brown, K. Zhao, N. King, H. Sobhani, P. Nordlander, and N. J. Halas,

“Surface-enhanced infrared absorption using individual cross antennas tailored
to chemical moieties.,” J. Am. Chem. Soc., vol. 135, no. 9, pp. 3688–95, Mar.
2013.
[68] F. P. Garcia de Arquer, A. Mihi, D. Kufer, and G. Konstantatos,

“Photoelectric Energy Conversion of Plasmon-Generated Hot Carriers in,”
ACS Nano, vol. 7, no. 4, pp. 3581–3588, 2013.
[69] A. K. Pradhan, T. Holloway, R. Mundle, H. Dondapati, and M. Bahoura,

“Energy harvesting in semiconductor-insulator-semiconductor junctions
through excitation of surface plasmon polaritons,” Appl. Phys. Lett., vol. 100,
no. 6, p. 061127, Feb. 2012.
[70] S. Mubeen, J. Lee, N. Singh, S. Krämer, G. D. Stucky, and M. Moskovits, “An

autonomous photosynthetic device in which all charge carriers derive from
surface plasmons.,” Nat. Nanotechnol., vol. 8, no. 4, pp. 247–51, Apr. 2013.
[71] F. Wang and N. A. Melosh, “Plasmonic energy collection through hot carrier
extraction.,” Nano Lett., vol. 11, no. 12, pp. 5426–30, Dec. 2011.
116
[72] H. A. Atwater and A. Polman, “Plasmonics for improved photovoltaic
devices.,” Nat. Mater., vol. 9, no. 3, pp. 205–13, Mar. 2010.
[73] F. Wang and N. A. Melosh, “Power-independent wavelength determination by

hot carrier collection in metal-insulator-metal devices.,” Nat. Commun., vol. 4,
p. 1711, Jan. 2013.
[74] M. A. Kats, R. Blanchard, P. Genevet, and F. Capasso, “Nanometre optical

coatings based on strong interference effects in highly absorbing media.,” Nat.
Mater., vol. 12, no. 1, pp. 20–4, Jan. 2013.
[75] M. A. Kats, R. Blanchard, S. Ramanathan, and F. Capasso, “Thin-Film

Interference in Lossy, Ultra-Thin Layers,” Opt. Photonics News, vol. 25, p.
40, 2014.
[76] F. F. Schlich and R. Spolenak, “Strong interference in ultrathin

semiconducting layers on a wide variety of substrate materials,” Appl. Phys.
Lett., vol. 103, no. 21, p. 213112, Nov. 2013.
[77] H. Song, L. Guo, Z. Liu, K. Liu, X. Zeng, D. Ji, N. Zhang, H. Hu, S. Jiang,
and Q. Gan, “Nanocavity enhancement for ultra-thin film optical absorber,”
Adv. Mater., vol. 26, no. 17, pp. 2737–2743, 2014.
[78] M. A. Kats, S. J. Byrnes, R. Blanchard, M. Kolle, P. Genevet, J. Aizenberg,

and F. Capasso, “Enhancement of absorption and color contrast in ultra-thin
highly absorbing optical coatings,” Appl. Phys. Lett., vol. 103, no. 10, p.
101104, 2013.
[79] S. A. Maier, “Plasmonics: Fundamentals and Applications,” Springer, 2007.
[80] W. L. Barnes, A. Dereux, and T. W. Ebbesen, “Surface plasmon

subwavelength optics.,” Nature, vol. 424, no. 6950, pp. 824–30, Aug. 2003.
[81] A. K. Popov, S. A. Myslivets, T. F. George, and V. M. Shalaev, “Four-wave

mixing, quantum control, and compensating losses in doped negative-index
photonic metamaterials,” Opt. Lett., vol. 32, no. 20, p. 3044, Oct. 2007.
[82] S. Palomba, S. Zhang, Y. Park, G. Bartal, X. Yin, and X. Zhang, “Optical

negative refraction by four-wave mixing in thin metallic nanostructures.,” Nat.
Mater., vol. 11, no. 1, pp. 34–8, Jan. 2012.
[83] J. T. Choy, B. J. M. Hausmann, T. M. Babinec, I. Bulu, M. Khan, P.

Maletinsky, A. Yacoby, and M. Lončar, “Enhanced single-photon emission
from a diamond–silver aperture,” Nat. Photonics, vol. 5, no. 12, pp. 738–743,
Oct. 2011.
[84] I. L. Chuang and Y. Yamamoto, “Simple quantum computer,” Phys. Rev. A,

vol. 52, no. 5, pp. 3489–3496, Nov. 1995.
117
[85] K. Kneipp, Y. Wang, H. Kneipp, L. Perelman, I. Itzkan, R. Dasari, and M.
Feld, “Single Molecule Detection Using Surface-Enhanced Raman Scattering
(SERS),” Phys. Rev. Lett., vol. 78, no. 9, pp. 1667–1670, Mar. 1997.
[86] S. Nie, “Probing Single Molecules and Single Nanoparticles by Surface-

Enhanced Raman Scattering,” Science (80-. )., vol. 275, no. 5303, pp. 1102–
1106, Feb. 1997.
[87] Y. Fang, N.-H. Seong, and D. D. Dlott, “Measurement of the Distribution of

Site Enhancements in Surface-Enhanced Raman Scattering,” Science (80-. ).,
vol. 321, no. 5887, pp. 388–392, Jul. 2008.
[88] J. J. Baumberg, T. A. Kelf, Y. Sugawara, S. Cintra, M. E. Abdelsalam, P. N.

Bartlett, and A. E. Russell, “Angle-resolved surface-enhanced Raman
scattering on metallic nanostructured plasmonic crystals.,” Nano Lett., vol. 5,
no. 11, pp. 2262–7, Nov. 2005.
[89] A. Kocabas, G. Ertas, S. S. Senlik, and A. Aydinli, “Plasmonic band gap

structures for surface-enhanced Raman scattering,” Opt. Express, vol. 16, no.
17, p. 12469, Aug. 2008.
[90] W. J. Cho, Y. Kim, and J. K. Kim, “Ultrahigh Density Array of Silver

Nanoclusters for SERS Substrate with High Sensitivity and Excellent
Reproducibility.,” ACS Nano, vol. 6, no. 1, pp. 249–255, Nov. 2011.
[91] I. P. Kaminow, W. L. Mammel, and H. P. Weber, “Metal-clad optical

waveguides: analytical and experimental study.,” Appl. Opt., vol. 13, no. 2, pp.
396–405, Feb. 1974.
[92] Y. Todorov, L. Tosetto, J. Teissier, A. M. Andrews, P. Klang, R. Colombelli,

I. Sagnes, G. Strasser, and C. Sirtori, “Optical properties of metal-dielectric-
metal microcavities in the THz frequency range,” Opt. Express, vol. 18, no.
13, p. 13886, Jun. 2010.
[93] N. Engheta, A. Salandrino, and A. Alù, “Circuit Elements at Optical

Frequencies: Nanoinductors, Nanocapacitors, and Nanoresistors,” Phys. Rev.
Lett., vol. 95, no. 9, Aug. 2005.
[94] J. Zhou, E. N. Economon, T. Koschny, and C. M. Soukoulis, “Unifying

approach to left-handed material design,” Opt. Lett., vol. 31, no. 24, p. 3620,
Dec. 2006.
[95] L. Fu, H. Schweizer, H. Guo, N. Liu, and H. Giessen, “Synthesis of

transmission line models for metamaterial slabs at optical frequencies,” Phys.
Rev. B, vol. 78, no. 11, Sep. 2008.
[96] Y. Chu, D. Wang, W. Zhu, and K. B. Crozier, “Double resonance surface

enhanced Raman scattering substrates: an intuitive coupled oscillator model,”
Opt. Express, vol. 19, no. 16, p. 14919, Jul. 2011.
118
[97] M. Pu, C. Hu, M. Wang, C. Huang, Z. Zhao, C. Wang, Q. Feng, and X. Luo,
“Design principles for infrared wide-angle perfect absorber based on
plasmonic structure,” Opt. Express, vol. 19, no. 18, p. 17413, Aug. 2011.
[98] H. T. Miyazaki and Y. Kurokawa, “Squeezing Visible Light Waves into a 3-

nm-Thick and 55-nm-Long Plasmon Cavity,” Phys. Rev. Lett., vol. 96, no. 9,
p. 097401, Mar. 2006.
[99] H. Liu, D. Genov, D. Wu, Y. Liu, J. Steele, C. Sun, S. Zhu, and X. Zhang,
“Magnetic Plasmon Propagation Along a Chain of Connected Subwavelength
Resonators at Infrared Frequencies,” Phys. Rev. Lett., vol. 97, no. 24, Dec.
2006.
[100] A. Yariv, Y. Xu, R. K. Lee, and A. Scherer, “Coupled-resonator optical

waveguide: a proposal and analysis,” Opt. Lett., vol. 24, no. 11, p. 711, Jun.
1999.
[101] A. N. Grigorenko, A. K. Geim, H. F. Gleeson, Y. Zhang, A. A. Firsov, I. Y.

Khrushchev, and J. Petrovic, “Nanofabricated media with negative
permeability at visible frequencies.,” Nature, vol. 438, no. 7066, pp. 335–8,
Nov. 2005.
[102] V. Giannini, G. Vecchi, and J. Gómez Rivas, “Lighting Up Multipolar Surface

Plasmon Polaritons by Collective Resonances in Arrays of Nanoantennas,”
Phys. Rev. Lett., vol. 105, no. 26, Dec. 2010.
[103] B. Auguié and W. Barnes, “Collective Resonances in Gold Nanoparticle

Arrays,” Phys. Rev. Lett., vol. 101, no. 14, Sep. 2008.
[104] E. Feigenbaum and H. A. Atwater, “Resonant Guided Wave Networks,” Phys.

Rev. Lett., vol. 104, no. 14, p. 147402, Apr. 2010.
[105] K. Aydin, V. E. Ferry, R. M. Briggs, and H. A. Atwater, “Broadband

polarization-independent resonant light absorption using ultrathin plasmonic
super absorbers,” Nat. Commun., vol. 2, p. 517, Nov. 2011.
[106] S. Wang, X. Shan, U. Patel, X. Huang, J. Lu, J. Li, and N. Tao, “Label-free
imaging, detection, and mass measurement of single viruses by surface
plasmon resonance.,” Proc. Natl. Acad. Sci. U. S. A., vol. 107, no. 37, pp.
16028–32, Sep. 2010.
[107] N. Liu, M. L. Tang, M. Hentschel, H. Giessen, and A. P. Alivisatos,

“Nanoantenna-enhanced gas sensing in a single tailored nanofocus.,” Nat.
Mater., vol. 10, no. 8, pp. 631–636, May 2011.
[108] T. V. Teperik, F. J. García de Abajo, a. G. Borisov, M. Abdelsalam, P. N.

Bartlett, Y. Sugawara, and J. J. Baumberg, “Omnidirectional absorption in
nanostructured metal surfaces,” Nat. Photonics, vol. 2, no. 5, pp. 299–301,
Apr. 2008.
119
[109] E. Cubukcu, S. Zhang, Y.-S. Park, G. Bartal, and X. Zhang, “Split ring
resonator sensors for infrared detection of single molecular monolayers,”
Appl. Phys. Lett., vol. 95, no. 4, p. 043113, Jul. 2009.
[110] C. Cao, J. Zhang, X. Wen, S. L. Dodson, N. T. Dao, L. M. Wong, S. Wang, S.

Li, T. A. Phan, and Q. Xiong, “Metamaterials-BasedLabel-Free Nanosensors
for Conformation and Affinity Biosensing,” ACS Nano, vol. 7, no. 9, pp.
7583–7591, 2013.
[111] F. B. P. Niesler, J. K. Gansel, S. Fischbach, and M. Wegener, “Metamaterial

metal-based bolometers,” Appl. Phys. Lett., vol. 100, no. 20, p. 203508, 2012.
[112] V. Savinov, V. a Fedotov, P. a J. de Groot, and N. I. Zheludev, “Radiation-

harvesting resonant superconducting sub-THz metamaterial bolometer,”
Supercond. Sci. Technol., vol. 26, no. 8, p. 084001, Aug. 2013.
[113] N. Landy, S. Sajuyigbe, J. Mock, D. Smith, and W. Padilla, “Perfect

Metamaterial Absorber,” Phys. Rev. Lett., vol. 100, no. 20, p. 207402, May
2008.
[114] J. W. Park, P. Van Tuong, J. Y. Rhee, K. W. Kim, W. H. Jang, E. H. Choi, L.

Y. Chen, and Y. Lee, “Multi-band metamaterial absorber based on the
arrangement of donut-type resonators,” Opt. Express, vol. 21, no. 8, pp. 9691–
9702, 2013.
[115] H. Wakatsuchi, S. Greedy, C. Christopoulos, and J. Paul, “Customised

broadband metamaterial absorbers for arbitrary polarisation.,” Opt. Express,
vol. 18, no. 21, pp. 22187–98, Oct. 2010.
[116] M. Diem, T. Koschny, and C. Soukoulis, “Wide-angle perfect

absorber/thermal emitter in the terahertz regime,” Phys. Rev. B, vol. 79, no. 3,
p. 033101, Jan. 2009.
[117] H. Tao, N. I. Landy, C. M. Bingham, X. Zhang, R. D. Averitt, and W. J.

Padilla, “A metamaterial absorber for the terahertz regime : Design ,
fabrication and characterization Abstract :,” vol. 16, no. 10, pp. 1494–1496,
2008.
[118] J. Hao, J. Wang, X. Liu, W. J. Padilla, L. Zhou, and M. Qiu, “High

performance optical absorber based on a plasmonic metamaterial,” Appl. Phys.
Lett., vol. 96, no. 25, p. 251104, Jun. 2010.
[119] X. Liu, T. Starr, A. F. Starr, and W. J. Padilla, “Infrared Spatial and Frequency
Selective Metamaterial with Near-Unity Absorbance,” Phys. Rev. Lett., vol.
104, no. 20, p. 207403, May 2010.
[120] J. Hao, L. Zhou, and M. Qiu, “Nearly total absorption of light and heat
generation by plasmonic metamaterials,” Phys. Rev. B, vol. 83, no. 16, p.
165107, Apr. 2011.
120
[121] X. Liu, T. Tyler, T. Starr, A. F. Starr, N. M. Jokerst, and W. J. Padilla,
“Taming the Blackbody with Infrared Metamaterials as Selective Thermal
Emitters,” Phys. Rev. Lett., vol. 107, no. 4, p. 045901, Jul. 2011.
[122] R. Alaee, C. Menzel, C. Rockstuhl, and F. Lederer, “Perfect absorbers on

curved surfaces and their potential applications,” Opt. Express, vol. 20, no. 16,
p. 18370, Jul. 2012.
[123] L. P. Wang and Z. M. Zhang, “Wavelength-selective and diffuse emitter

enhanced by magnetic polaritons for thermophotovoltaics,” Appl. Phys. Lett.,
vol. 100, no. 6, p. 063902, Feb. 2012.
[124] P. Zhu and L. Jay Guo, “High performance broadband absorber in the visible
band by engineered dispersion and geometry of a metal-dielectric-metal
stack,” Appl. Phys. Lett., vol. 101, no. 24, p. 241116, 2012.
[125] M. G. Nielsen, A. Pors, O. Albrektsen, and S. I. Bozhevolnyi, “Efficient

absorption of visible radiation by gap plasmon resonators.,” Opt. Express, vol.
20, no. 12, pp. 13311–9, Jun. 2012.
[126] J. Wang, C. Fan, P. Ding, J. He, Y. Cheng, W. Hu, G. Cai, E. Liang, and Q.
Xue, “Tunable broad-band perfect absorber by exciting of multiple plasmon
resonances at optical frequency,” Opt. Express, vol. 20, no. 14, p. 14871, Jun.
2012.
[127] K. Aydin, V. E. Ferry, R. M. Briggs, and H. a Atwater, “Broadband

polarization-independent resonant light absorption using ultrathin plasmonic
super absorbers.,” Nat. Commun., vol. 2, p. 517, Jan. 2011.
[128] J. A. Bossard, L. Lin, S. Yun, L. Liu, D. H. Werner, and T. S. Mayer, “Near-

Ideal Optical Metamaterial Bandwidth,” ACS Nano, vol. 8, no. 2, pp. 1517–
1524, 2014.
[129] R. Alaee, C. Menzel, U. Huebner, E. Pshenay-Severin, S. Bin Hasan, T.

Pertsch, C. Rockstuhl, and F. Lederer, “Deep-subwavelength plasmonic
nanoresonators exploiting extreme coupling.,” Nano Lett., vol. 13, no. 8, pp.
3482–6, Aug. 2013.
[130] X. Wang, C. Luo, G. Hong, and X. Zhao, “Metamaterial optical refractive

index sensor detected by the naked eye,” Appl. Phys. Lett., vol. 102, no. 9, p.
091902, 2013.
[131] M. K. Hedayati, M. Javaherirahim, B. Mozooni, R. Abdelaziz, A.

Tavassolizadeh, V. S. K. Chakravadhanula, V. Zaporojtchenko, T. Strunkus,
F. Faupel, and M. Elbahri, “Design of a perfect black absorber at visible
frequencies using plasmonic metamaterials.,” Adv. Mater., vol. 23, no. 45, pp.
5410–4, Dec. 2011.
[132] M. Elbahri, M. K. Hedayati, V. S. Kiran Chakravadhanula, M. Jamali, T.

Strunkus, V. Zaporojtchenko, and F. Faupel, “An omnidirectional transparent
121
conducting-metal-based plasmonic nanocomposite.,” Adv. Mater., vol. 23, no.
17, pp. 1993–7, May 2011.
[133] A. Moreau, C. Ciracì, J. J. Mock, R. T. Hill, Q. Wang, B. J. Wiley, A.

Chilkoti, and D. R. Smith, “Controlled-reflectance surfaces with film-coupled
colloidal nanoantennas.,” Nature, vol. 492, no. 7427, pp. 86–9, Dec. 2012.
[134] Z. H. Jiang, S. Yun, F. Toor, D. H. Werner, and T. S. Mayer, “Conformal

Dual-Band Near-Perfectly Coating,” ACS Nano, vol. 5, no. 6, pp. 4641–4647,
2011.
[135] Y. Cui, K. H. Fung, J. Xu, H. Ma, Y. Jin, S. He, and N. Fang, “Ultra-
broadband Light Absorption by a Sawtooth Anisotropic Metamaterial Slab.,”
Nano Lett., Feb. 2012.
[136] B. Zhang, Y. Zhao, Q. Hao, B. Kiraly, I.-C. Khoo, S. Chen, and T. J. Huang,
“Polarization-independent dual-band infrared perfect absorber based on a
metal-dielectric-metal elliptical nanodisk array,” Opt. Express, vol. 19, no. 16,
p. 15221, Jul. 2011.
[137] C.-W. Cheng, M. N. Abbas, C.-W. Chiu, K.-T. Lai, M.-H. Shih, and Y.-C.
Chang, “Wide-angle polarization independent infrared broadband absorbers
based on metallic multi-sized disk arrays,” Opt. Express, vol. 20, no. 9, p.
10376, Apr. 2012.
[138] J. Hendrickson, J. Guo, B. Zhang, W. Buchwald, and R. Soref, “Wideband

perfect light absorber at midwave infrared using multiplexed metal
structures.,” Opt. Lett., vol. 37, no. 3, pp. 371–3, Feb. 2012.
[139] E. Lansey, I. R. Hooper, J. N. Gollub, A. P. Hibbins, and D. T. Crouse, “Light

localization, photon sorting, and enhanced absorption in subwavelength cavity
arrays.,” Opt. Express, vol. 20, no. 22, pp. 24226–36, Oct. 2012.
[140] J. Le Perchec, Y. Desieres, N. Rochat, and R. Espiau de Lamaestre,

“Subwavelength optical absorber with an integrated photon sorter,” Appl.
Phys. Lett., vol. 100, no. 11, p. 113305, 2012.
[141] W. Ma, Y. Wen, and X. Yu, “Broadband metamaterial absorber at mid-

infrared using multiplexed cross resonators,” Opt. Express, vol. 21, no. 25, pp.
154–156, 2013.
[142] H. Ko, D.-H. Ko, Y. Cho, and I. K. Han, “Broadband light absorption using a
multilayered gap surface plasmon resonator,” Appl. Phys. A, no. Mmdm, May
2014.
[143] P. Pitchappa, C. P. Ho, P. Kropelnicki, N. Singh, D.-L. Kwong, and C. Lee,

“Dual band complementary metamaterial absorber in near infrared region,” J.
Appl. Phys., vol. 115, no. 19, p. 193109, May 2014.
122
[144] S. Ayas, H. Guner, B. Turker, O. O. Ekiz, F. Dirisaglik, A. K. Okyay, and A.
Dana, “Raman Enhancement on a Broadband,” ACS Nano, vol. 6, no. 8, pp.
6852–6861, 2012.
[145] D. Li, L. Qin, X. Xiong, R.-W. Peng, Q. Hu, G.-B. Ma, H.-S. Zhou, and M.
Wang, “Exchange of electric and magnetic resonances in multilayered
metal/dielectric nanoplates,” Opt. Express, vol. 19, no. 23, p. 22942, Oct.
2011.
[146] C. Koechlin, P. Bouchon, F. Pardo, J. Pelouard, and R. Ha, “Analytical

description of subwavelength plasmonic MIM resonators and of their
combination Abstract :,” vol. 21, no. 6, pp. 4641–4647, 2013.
[147] R. Filter, C. Rockstuhl, and F. Lederer, “Circular Optical Nanoantennas - An

Analytical Theory,” Phys. Rev. B, vol. 85, no. 125429, pp. 1–16, 1887.
[148] F. Minkowski, F. Wang, A. Chakrabarty, and Q.-H. Wei, “Resonant cavity

modes of circular plasmonic patch nanoantennas,” Appl. Phys. Lett., vol. 104,
no. 2, p. 021111, Jan. 2014.
[149] J. G. Bergman, D. S. Chemla, P. F. Liao, A. M. Glass, A. Pinczuk, R. M. Hart,

and D. H. Olson, “Relationship between surface-enhanced Raman scattering
and the dielectric properties of aggregated silver films,” Opt. Lett., vol. 6, no.
1, p. 33, Jan. 1981.
[150] A. I. Maaroof and D. S. Sutherland, “Optimum plasmon hybridization at

percolation threshold of silver films near metallic surfaces,” J. Phys. D. Appl.
Phys., vol. 43, no. 40, p. 405301, Oct. 2010.
[151] J. J. Mock, R. T. Hill, A. Degiron, S. Zauscher, A. Chilkoti, and D. R. Smith,

“Distance-dependent plasmon resonant coupling between a gold nanoparticle
and gold film.,” Nano Lett., vol. 8, no. 8, pp. 2245–52, Aug. 2008.
[152] S. Mubeen, S. Zhang, N. Kim, S. Lee, S. Krämer, H. Xu, and M. Moskovits,

“Plasmonic properties of gold nanoparticles separated from a gold mirror by
an ultrathin oxide.,” Nano Lett., vol. 12, no. 4, pp. 2088–94, Apr. 2012.
[153] D. Y. Lei, A. I. Fernandez-Dominguez, Y. Sonnefraud, K. Appavoo, R. F.

Haglund, J. B. Pendry, and S. A. Maier, “Revealing Plasmonic Gap Modes in
Particle-on-Film Systems Using Dark-Field Spectroscopy.,” ACS Nano, Jan.
2012.
[154] N. M. B. Perney, J. J. Baumberg, M. E. Zoorob, M. D. B. Charlton, S.

Mahnkopf, and C. M. Netti, “Tuning localized plasmons in nanostructured
substrates for surface-enhanced Raman scattering,” Opt. Express, vol. 14, no.
2, p. 847, Jan. 2006.
[155] J. A. Dieringer, R. B. Lettan, K. A. Scheidt, and R. P. Van Duyne, “A

frequency domain existence proof of single-molecule surface-enhanced
123
Raman spectroscopy.,” J. Am. Chem. Soc., vol. 129, no. 51, pp. 16249–56,
Dec. 2007.
[156] S. R. Emory, R. A. Jensen, T. Wenda, M. Han, and S. Nie, “Re-examining the

origins of spectral blinking in single-molecule and single-nanoparticleSERS,”
Faraday Discuss., vol. 132, pp. 249–259, Apr. 2006.
[157] Z. Wang and L. J. Rothberg, “Origins of blinking in single-molecule Raman

spectroscopy.,” J. Phys. Chem. B, vol. 109, no. 8, pp. 3387–91, Mar. 2005.
[158] E. C. Le Ru, M. Meyer, and P. G. Etchegoin, “Proof of single-molecule

sensitivity in surface enhanced Raman scattering (SERS) by means of a two-
analyte technique.,” J. Phys. Chem. B, vol. 110, no. 4, pp. 1944–8, Feb. 2006.
[159] K. A. Bosnick and L. E. Brus, “Fluctuations and Local Symmetry in Single-

Molecule Rhodamine 6G Raman Scattering on Silver Nanocrystal Aggregates
†,” J. Phys. Chem. B, vol. 106, no. 33, pp. 8096–8099, Aug. 2002.
[160] S. Khatua and M. Orrit, “Toward single-molecule microscopy on a smart

phone.,” ACS Nano, vol. 7, no. 10, pp. 8340–3, Oct. 2013.
[161] A. Otto, “Theory of First Layer and Single Molecule Surface Enhanced
Raman Scattering (SERS),” Phys. status solidi, vol. 188, no. 4, pp. 1455–
1470, Dec. 2001.
[162] Y. Maruyama, M. Ishikawa, and M. Futamata, “Thermal Activation of

Blinking in SERS Signal,” J. Phys. Chem. B, vol. 108, no. 2, pp. 673–678,
Jan. 2004.
[163] P. C. Andersen, M. L. Jacobson, and K. L. Rowlen, “Flashy Silver

Nanoparticles,” J. Phys. Chem. B, vol. 108, no. 7, pp. 2148–2153, Feb. 2004.
[164] Y. Shirai, A. J. Osgood, Y. Zhao, K. F. Kelly, and J. M. Tour, “Directional

control in thermally driven single-molecule nanocars.,” Nano Lett., vol. 5, no.
11, pp. 2330–4, Nov. 2005.
[165] M. Schunack, T. R. Linderoth, F. Rosei, E. Lægsgaard, I. Stensgaard, and F.

Besenbacher, “Long Jumps in the Surface Diffusion of Large Molecules,”
Phys. Rev. Lett., vol. 88, no. 15, p. 156102, Mar. 2002.
[166] J. Weckesser, J. V. Barth, and K. Kern, “Direct observation of surface

diffusion of large organic molecules at metal surfaces: PVBA on Pd(110),” J.
Chem. Phys., vol. 110, no. 11, p. 5351, Mar. 1999.
[167] A. Otto, “What is observed in single molecule SERS, and why?,” J. Raman
Spectrosc., vol. 33, no. 8, pp. 593–598, Aug. 2002.
[168] S. K. Brown, M. R. Sim, M. J. Abramson, and C. N. Gray, “Concentrations of

Volatile Organic Compounds in Indoor Air - A Review,” Indoor Air, vol. 4,
no. 2, pp. 123–134, Jun. 1994.
124
[169] M. Phillips, J. Herrera, S. Krishnan, M. Zain, J. Greenberg, and R. N. Cataneo,
“Variation in volatile organic compounds in the breath of normal humans,” J.
Chromatogr. B Biomed. Sci. Appl., vol. 729, no. 1–2, pp. 75–88, Jun. 1999.
[170] A. Kinkhabwala, Z. Yu, S. Fan, Y. Avlasevich, K. Müllen, and W. E.

Moerner, “Large single-molecule fluorescence enhancements produced by a
bowtie nanoantenna,” Nat. Photonics, vol. 3, no. 11, pp. 654–657, Oct. 2009.
[171] R. Henriques, M. Lelek, E. F. Fornasiero, F. Valtorta, C. Zimmer, and M. M.

Mhlanga, “QuickPALM: 3D real-time photoactivation nanoscopy image
processing in ImageJ.,” Nat. Methods, vol. 7, no. 5, pp. 339–40, May 2010.
[172] D. Gallegos, K. D. Long, H. Yu, P. P. Clark, Y. Lin, S. George, P. Nath, and

B. T. Cunningham, “Label-free biodetection using a smartphone.,” Lab Chip,
vol. 13, no. 11, pp. 2124–32, Jun. 2013.
[173] A. Otto, “The ‘chemical’ (electronic) contribution to surface-enhanced Raman

scattering,” J. Raman Spectrosc., vol. 36, no. 6–7, pp. 497–509, Jun. 2005.
[174] S. A. Maier and H. A. Atwater, “Plasmonics: Localization and guiding of

electromagnetic energy in metal/dielectric structures,” J. Appl. Phys., vol. 98,
no. 1, p. 011101, Jul. 2005.
[175] J. A. Schuller, E. S. Barnard, W. Cai, Y. C. Jun, J. S. White, and M. L.

Brongersma, “Plasmonics for extreme light concentration and manipulation.,”
Nat. Mater., vol. 9, no. 3, pp. 193–204, Mar. 2010.
[176] H. A. Atwater and A. Polman, “Plasmonics for improved photovoltaic

devices.,” Nat. Mater., vol. 9, no. 3, pp. 205–13, Mar. 2010.
[177] F. Wang and N. A. Melosh, “Power-independent wavelength determination by

hot carrier collection in metal-insulator-metal devices.,” Nat. Commun., vol. 4,
p. 1711, 2013.
[178] K. Kempa, M. J. Naughton, Z. F. Ren, A. Herczynski, T. Kirkpatrick, J.

Rybczynski, and Y. Gao, “Hot electron effect in nanoscopically thin
photovoltaic junctions,” Appl. Phys. Lett., vol. 95, no. 23, p. 233121, Dec.
2009.
[179] F. Wang and N. A. Melosh, “Plasmonic energy collection through hot carrier
extraction.,” Nano Lett., vol. 11, no. 12, pp. 5426–30, Dec. 2011.
[180] M. A. Green, “Third generation photovoltaics: Ultra-high conversion

efficiency at low cost,” Prog. Photovoltaics Res. Appl., vol. 9, no. 2, pp. 123–
135, Mar. 2001.
[181] F. B. Atar, E. Battal, L. E. Aygun, B. Daglar, M. Bayindir, and A. K. Okyay,

“Plasmonically enhanced hot electron based photovoltaic device,” Opt.
Express, vol. 21, no. 6, p. 7196, Mar. 2013.
125
[182] T. P. White and K. R. Catchpole, “Plasmon-enhanced internal photoemission
for photovoltaics: Theoretical efficiency limits,” Appl. Phys. Lett., vol. 101,
no. 7, p. 073905, Aug. 2012.
[183] Y. H. Wang, H. Steinberg, P. Jarillo-Herrero, and N. Gedik, “Observation of

Floquet-Bloch states on the surface of a topological insulator.,” Science, vol.
342, no. 6157, pp. 453–7, Oct. 2013.
[184] J. Cao, Y. Gao, R. J. D. Miller, H. E. Elsayed-Ali, and D. A. Mantell,

“Femtosecond photoemission study of ultrafast electron dynamics on
Cu(100),” Phys. Rev. B, vol. 56, no. 3, pp. 1099–1102, Jul. 1997.
[185] M. Bauer, C. Lei, K. Read, R. Tobey, J. Gland, M. M. Murnane, and H. C.

Kapteyn, “Direct Observation of Surface Chemistry Using Ultrafast Soft-X-
Ray Pulses,” Phys. Rev. Lett., vol. 87, no. 2, p. 025501, Jun. 2001.
[186] O. O. Ekiz, K. Mizrak, and A. Dâna, “Chemically specific dynamic

characterization of photovoltaic and photoconductivity effects of surface
nanostructures.,” ACS Nano, vol. 4, no. 4, pp. 1851–60, Apr. 2010.
[187] H. Sezen, A. A. Rockett, and S. Suzer, “XPS Investigation of a CdS-Based

Photoresistor under Working Conditions: Operando–XPS,” Anal. Chem., vol.
84, no. 6, pp. 2990–2994, Mar. 2012.
[188] H. Sezen and S. Suzer, “Charging/discharging dynamics of CdS and CdSe

films under photoillumination using dynamic x-ray photoelectron
spectroscopy,” J. Vac. Sci. Technol. A Vacuum, Surfaces, Film., vol. 28, no. 4,
p. 639, Jun. 2010.
[189] S. Suzer, H. Sezen, and A. Dâna, “Two-dimensional X-ray photoelectron

spectroscopy for composite surface analysis.,” Anal. Chem., vol. 80, no. 10,
pp. 3931–6, May 2008.
[190] S. Suzer and A. Dâna, “X-ray photoemission for probing charging/discharging

dynamics.,” J. Phys. Chem. B, vol. 110, no. 39, pp. 19112–5, Oct. 2006.
[191] S. Ayas, H. Güner, B. Türker, O. Ö. Ekiz, F. Dirisaglik, A. K. Okyay, and A.

Dâna, “Raman Enhancement on a Broadband Meta-Surface,” ACS Nano, vol.
6, no. 8, pp. 6852–6861, Aug. 2012.
[192] M. L. Huang, Y. C. Chang, Y. H. Chang, T. D. Lin, J. Kwo, and M. Hong,

“Energy-band parameters of atomic layer deposited Al2O3 and HfO2 on
InxGa1-xAs,” Appl. Phys. Lett., vol. 94, no. 5, p. 052106, Feb. 2009.
[193] G. Ertas, U. K. Demirok, A. Atalar, and S. Suzer, “X-ray photoelectron

spectroscopy for resistance-capacitance measurements of surface structures,”
Appl. Phys. Lett., vol. 86, no. 18, p. 183110, Apr. 2005.
126
[194] H. Cohen, “Chemically resolved electrical measurements using x-ray
photoelectron spectroscopy,” Appl. Phys. Lett., vol. 85, no. 7, p. 1271, Aug.
2004.
[195] P. R. West, S. Ishii, G. V. Naik, N. K. Emani, V. M. Shalaev, and a.

Boltasseva, “Searching for better plasmonic materials,” Laser Photon. Rev.,
vol. 4, no. 6, pp. 795–808, Nov. 2010.
[196] Y. Jeyaram, S. K. Jha, M. Agio, J. F. Löffler, and Y. Ekinci, “Magnetic

metamaterials in the blue range using aluminum nanostructures,” Opt. Lett.,
vol. 35, no. 10, p. 1656, May 2010.
[197] A. Taguchi, Y. Saito, K. Watanabe, S. Yijian, and S. Kawata, “Tailoring

plasmon resonances in the deep-ultraviolet by size-tunable fabrication of
aluminum nanostructures,” Appl. Phys. Lett., vol. 101, no. 8, p. 081110, 2012.
[198] S. K. Jha, Z. Ahmed, M. Agio, Y. Ekinci, and J. F. Löffler, “Deep-Ultraviolet

Surface-Enhanced Resonance Raman Scattering of Adenine on Aluminum
Nanoparticle Arrays.,” J. Am. Chem. Soc., vol. 134, no. 4, pp. 1966–1969, Jan.
2012.
[199] Y. Yang, J. M. Callahan, T.-H. Kim, A. S. Brown, and H. O. Everitt,

“Ultraviolet nanoplasmonics: a demonstration of surface-enhanced Raman
spectroscopy, fluorescence, and photodegradation using gallium
nanoparticles.,” Nano Lett., vol. 13, no. 6, pp. 2837–41, Jun. 2013.
[200] M. Kanehara, H. Koike, T. Yoshinaga, and T. Teranishi, “Indium tin oxide

nanoparticles with compositionally tunable surface plasmon resonance
frequencies in the near-IR region.,” J. Am. Chem. Soc., vol. 131, no. 49, pp.
17736–7, Dec. 2009.
[201] G. Garcia, R. Buonsanti, E. L. Runnerstrom, R. J. Mendelsberg, A. Llordes, A.

Anders, T. J. Richardson, and D. J. Milliron, “Dynamically modulating the
surface plasmon resonance of doped semiconductor nanocrystals.,” Nano
Lett., vol. 11, no. 10, pp. 4415–20, Oct. 2011.
[202] J. M. Luther, P. K. Jain, T. Ewers, and a P. Alivisatos, “Localized surface

plasmon resonances arising from free carriers in doped quantum dots.,” Nat.
Mater., vol. 10, no. 5, pp. 361–6, May 2011.
[203] S. Q. Li, P. Guo, L. Zhang, W. Zhou, T. W. Odom, T. Seideman, J. B.

Ketterson, and R. P. H. Chang, “Infrared Plasmonics with Indium–Tin-Oxide
Nanorod Arrays,” ACS Nano, vol. 5, no. 11, pp. 9161–9170, 2011.
[204] G. V Naik, J. Liu, A. V Kildishev, V. M. Shalaev, and A. Boltasseva,

“Demonstration of Al:ZnO as a plasmonic component for near-infrared
metamaterials.,” Proc. Natl. Acad. Sci. U. S. A., vol. 109, no. 23, pp. 8834–8,
Jun. 2012.
127
[205] V. W. Brar, M. S. Jang, M. Sherrott, J. J. Lopez, and H. A. Atwater, “Highly
confined tunable mid-infrared plasmonics in graphene nanoresonators.,” Nano
Lett., vol. 13, no. 6, pp. 2541–7, Jun. 2013.
[206] M. Abb, Y. Wang, N. Papasimakis, C. H. de Groot, and O. L. Muskens,

“Surface-enhanced infrared spectroscopy using metal oxide plasmonic antenna
arrays.,” Nano Lett., vol. 14, no. 1, pp. 346–52, Jan. 2014.
[207] L. Yu, A. Rosenberg, and D. Wasserman, “All-semiconductor plasmonic

nanoantennas for infrared sensing.,” Nano Lett., vol. 13, no. 9, pp. 4569–74,
Sep. 2013.
[208] H. J. Lezec, J. A. Dionne, and H. A. Atwater, “Negative refraction at visible

frequencies.,” Science, vol. 316, no. 5823, pp. 430–2, Apr. 2007.
[209] T. Xu, A. Agrawal, M. Abashin, K. J. Chau, and H. J. Lezec, “All-angle

negative refraction and active flat lensing of ultraviolet light.,” Nature, vol.
497, no. 7450, pp. 470–4, May 2013.
[210] K. E. Shafer-Peltier, C. L. Haynes, M. R. Glucksberg, and R. P. Van Duyne,

“Toward a glucose biosensor based on surface-enhanced Raman scattering.,”
J. Am. Chem. Soc., vol. 125, no. 2, pp. 588–93, Jan. 2003.
[211] S. Nie, “Probing Single Molecules and Single Nanoparticles by Surface-

Enhanced Raman Scattering,” Science (80-. )., vol. 275, no. 5303, pp. 1102–
1106, Feb. 1997.
[212] J. C. Reed, H. Zhu, A. Y. Zhu, C. Li, and E. Cubukcu, “Graphene-enabled

silver nanoantenna sensors.,” Nano Lett., vol. 12, no. 8, pp. 4090–4, Aug.
2012.
[213] M. H. Chowdhury, K. Ray, S. K. Gray, J. Pond, and J. R. Lakowicz,

“Aluminum nanoparticles as substrates for metal-enhanced fluorescence in the
ultraviolet for the label-free detection of biomolecules.,” Anal. Chem., vol. 81,
no. 4, pp. 1397–403, Feb. 2009.
[214] G. H. Chan, J. Zhao, G. C. Schatz, and R. P. Van Duyne, “Localized Surface

Plasmon Resonance Spectroscopy of Triangular Aluminum Nanoparticles,” J.
Phys. Chem. C, vol. 112, no. 36, pp. 13958–13963, Sep. 2008.
[215] D. O. Sigle, E. Perkins, J. J. Baumberg, and S. Mahajan, “Reproducible Deep-

UV SERRS on Aluminum Nanovoids,” J. Phys. Chem. Lett., vol. 4, no. 9, pp.
1449–1452, May 2013.
[216] G. Maidecchi, G. Gonella, R. Proietti Zaccaria, R. Moroni, L. Anghinolfi, A.

Giglia, S. Nannarone, L. Mattera, H.-L. Dai, M. Canepa, and F. Bisio, “Deep
ultraviolet plasmon resonance in aluminum nanoparticle arrays.,” ACS Nano,
vol. 7, no. 7, pp. 5834–41, Jul. 2013.
128
[217] M. a Ordal, L. L. Long, R. J. Bell, S. E. Bell, R. R. Bell, R. W. Alexander, and
C. a Ward, “Optical properties of the metals Al, Co, Cu, Au, Fe, Pb, Ni, Pd,
Pt, Ag, Ti, and W in the infrared and far infrared.,” Appl. Opt., vol. 22, no. 7,
pp. 1099–20, Apr. 1983.
[218] J. Martin, J. Proust, D. Gérard, and J. Plain, “Localized surface plasmon

resonances in the ultraviolet from large scale nanostructured aluminum films,”
Opt. Mater. Express, vol. 3, no. 7, p. 954, Jun. 2013.
[219] G. Maidecchi, G. Gonella, R. Proietti Zaccaria, R. Moroni, L. Anghinolfi, A.

Giglia, S. Nannarone, L. Mattera, H.-L. Dai, M. Canepa, and F. Bisio, “Deep
ultraviolet plasmon resonance in aluminum nanoparticle arrays.,” ACS Nano,
vol. 7, no. 7, pp. 5834–41, Jul. 2013.
[220] B. Kasemo, C. Langhammer, and Z. Michael, “Gold, Platinum, and

Aluminum Nanodisk Plasmons: Material Independence, Subradiance, and
Damping Mechanisms,” ACS Nano, vol. 5, no. 4, pp. 2535–2546, 2011.
[221] S. J. Tan, L. Zhang, D. Zhu, X. M. Goh, Y. M. Wang, K. Kumar, C.-W. Qiu,

and J. K. W. Yang, “Plasmonic color palettes for photorealistic printing with
aluminum nanostructures.,” Nano Lett., vol. 14, no. 7, pp. 4023–9, Jul. 2014.
[222] Y. Ekinci, H. H. Solak, and J. F. Löffler, “Plasmon resonances of aluminum

nanoparticles and nanorods,” J. Appl. Phys., vol. 104, no. 8, p. 083107, 2008.
[223] M. W. Knight, N. S. King, L. Liu, H. O. Everitt, P. Nordlander, and N. J.

Halas, “Aluminum for plasmonics.,” ACS Nano, vol. 8, no. 1, pp. 834–40, Jan.
2014.
[224] F. Mazzotta, G. Wang, C. Hägglund, F. Höök, and M. P. Jonsson,

“Nanoplasmonic biosensing with on-chip electrical detection.,” Biosens.
Bioelectron., vol. 26, no. 4, pp. 1131–6, Dec. 2010.
[225] E. Ozbay, “Plasmonics: merging photonics and electronics at nanoscale

dimensions.,” Science, vol. 311, no. 5758, pp. 189–93, Jan. 2006.
[226] A. Hryciw, Y. C. Jun, and M. L. Brongersma, “Plasmonics: Electrifying

plasmonics on silicon.,” Nat. Mater., vol. 9, no. 1, pp. 3–4, Jan. 2010.
[227] J. a. Dionne, L. a. Sweatlock, M. T. Sheldon, A. P. Alivisatos, and H. A.

Atwater, “Silicon-Based Plasmonics for On-Chip Photonics,” IEEE J. Sel.
Top. Quantum Electron., vol. 16, no. 1, pp. 295–306, 2010.
[228] C. Langhammer, M. Schwind, B. Kasemo, and I. Zoric, “Localized Surface

Plasmon Resonances in Aluminum Nanodisks,” Nano Lett., 8 (5), pp 1461–
1471, vol. 8, no. 5, p. pp 1461–1471, 2008.
[229] M. Surfaces, P. E. Laibinis, G. M. Whitesides, and D. L. Allara, “Comparison

of the Structures and Wetting Properties of Self-Assembled Monolayers of,”
pp. 7152–7167, 1991.
129
[230] F. Neubrech, A. Pucci, T. Cornelius, S. Karim, A. García-Etxarri, and J.
Aizpurua, “Resonant Plasmonic and Vibrational Coupling in a Tailored
Nanoantenna for Infrared Detection,” Phys. Rev. Lett., vol. 101, no. 15, p.
157403, Oct. 2008.
[231] S. Aksu, M. Huang, A. Artar, A. a Yanik, S. Selvarasah, M. R. Dokmeci, and

H. Altug, “Flexible plasmonics on unconventional and nonplanar substrates.,”
Adv. Mater., vol. 23, no. 38, pp. 4422–30, Oct. 2011.
[232] S. Ayas, A. Cupallari, O. O. Ekiz, Y. Kaya, and A. Dana, “Counting

Molecules with a Mobile Phone Camera Using Plasmonic Enhancement,”
ACS Photonics, vol. 1, no. 1, pp. 17–26, Jan. 2014.
[233] H. Aouani, M. Rahmani, H. Šípová, V. Torres, K. Hegnerová, M. Beruete, J.

Homola, M. Hong, M. Navarro-Cía, and S. a. Maier, “Plasmonic
Nanoantennas for Multispectral Surface-Enhanced Spectroscopies,” J. Phys.
Chem. C, vol. 117, no. 36, pp. 18620–18626, Sep. 2013.
[234] P. Jouy, Y. Todorov, a. Vasanelli, R. Colombelli, I. Sagnes, and C. Sirtori,

“Coupling of a surface plasmon with localized subwavelength microcavity
modes,” Appl. Phys. Lett., vol. 98, no. 2, p. 021105, 2011.
[235] A. Hartstein, J. Kirtley, and J. Tsang, “Enhancement of the Infrared

Absorption from Molecular Monolayers with Thin Metal Overlayers,” Phys.
Rev. Lett., vol. 45, no. 3, pp. 201–204, Jul. 1980.
[236] R. Adato and H. Altug, “In-situ ultra-sensitive infrared absorption

spectroscopy of biomolecule interactions in real time with plasmonic
nanoantennas.,” Nat. Commun., vol. 4, p. 2154, Jan. 2013.
[237] P. Bassan, A. Sachdeva, J. Lee, and P. Gardner, “Substrate contributions in

micro-ATR of thin samples: implications for analysis of cells, tissue and
biological fluids.,” Analyst, vol. 138, no. 14, pp. 4139–46, Jul. 2013.
[238] D. Dregely, F. Neubrech, H. Duan, R. Vogelgesang, and H. Giessen,

“Vibrational near-field mapping of planar and buried three-dimensional
plasmonic nanostructures,” Nat. Commun., vol. 4, p. 2237, Jan. 2013.
[239] A. E. Cetin, M. Turkmen, S. Aksu, D. Etezadi, and H. Altug, “Multi-resonant

compact nanoaperture with accessible large nearfields,” Appl. Phys. B, vol.
118, no. 1, pp. 29–38, Oct. 2014.
[240] A. E. Cetin, D. Etezadi, and H. Altug, “Accessible Nearfields by

Nanoantennas on Nanopedestals for Ultrasensitive Vibrational Spectroscopy,”
Adv. Opt. Mater., vol. 2, no. 9, pp. 866–872, Sep. 2014.
[241] S. Cataldo, J. Zhao, F. Neubrech, B. Frank, C. Zhang, P. V Braun, and H.

Giessen, “Hole-mask colloidal nanolithography for large-area low-cost
metamaterials and antenna-assisted surface-enhanced infrared absorption
substrates.,” ACS Nano, vol. 6, no. 1, pp. 979–85, Jan. 2012.
130
[242] C. Huck, A. Toma, F. Neubrech, M. Chirumamilla, J. Vogt, F. De Angelis,
and A. Pucci, “Gold Nanoantennas on a Pedestal for Plasmonic Enhancement
in the Infrared,” ACS Photonics, vol. 2, no. 4, pp. 497–505, Apr. 2015.
[243] Y. Shen, J. Zhou, T. Liu, Y. Tao, R. Jiang, M. Liu, G. Xiao, J. Zhu, Z.-K.
Zhou, X. Wang, C. Jin, and J. Wang, “Plasmonic gold mushroom arrays with
refractive index sensing figures of merit approaching the theoretical limit.,”
Nat. Commun., vol. 4, p. 2381, Jan. 2013.
[244] Y. Kaya, S. Ayas, A. E. Topal, H. Guner, and A. Dana, “Sensitivity

comparison of localized plasmon resonance structures and prism coupler,”
Sensors Actuators B Chem., vol. 191, pp. 516–521, Feb. 2014.
[245] M. S. Jang, V. W. Brar, M. C. Sherrott, J. J. Lopez, L. Kim, S. Kim, M. Choi,

and H. A. Atwater, “Tunable large resonant absorption in a midinfrared
graphene Salisbury screen,” Phys. Rev. B, vol. 90, no. 16, p. 165409, Oct.
2014.
[246] V. Thareja, J.-H. Kang, H. Yuan, K. M. Milaninia, H. Y. Hwang, Y. Cui, P. G.

Kik, and M. L. Brongersma, “Electrically tunable coherent optical absorption
in graphene with ion gel.,” Nano Lett., vol. 15, no. 3, pp. 1570–6, Mar. 2015.
[247] V. W. Brar, M. C. Sherrott, M. S. Jang, S. Kim, L. Kim, M. Choi, L. A.

Sweatlock, and H. A. Atwater, “Electronic modulation of infrared radiation in
graphene plasmonic resonators,” Nat. Commun., vol. 6, p. 7032, May 2015.
[248] F. Neubrech, S. Beck, T. Glaser, M. Hentschel, H. Giessen, and A. Pucci,

“Spatial extent of plasmonic enhancement of vibrational signals in the
infrared.,” ACS Nano, vol. 8, no. 6, pp. 6250–8, Jun. 2014.
[249] D. J. Michalak, S. R. Amy, D. Aureau, M. Dai, A. Estève, and Y. J. Chabal,

“Nanopatterning Si(111) surfaces as a selective surface-chemistry route.,” Nat.
Mater., vol. 9, no. 3, pp. 266–71, Mar. 2010.
[250] B. J. Eggleton, B. Luther-Davies, and K. Richardson, “Chalcogenide

photonics,” Nat. Photonics, vol. 5, no. 3, pp. 141–148, Dec. 2011.
[251] S. Raoux, F. Xiong, M. Wuttig, and E. Pop, “Phase change materials and
phase change memory,” MRS Bull., vol. 39, no. 08, pp. 703–710, Aug. 2014.
[252] M. Wuttig and N. Yamada, “Phase-change materials for rewriteable data

storage.,” Nat. Mater., vol. 6, no. 11, pp. 824–32, Nov. 2007.
[253] A.-K. U. Michel, D. N. Chigrin, T. W. W. Maß, K. Schönauer, M. Salinga, M.

Wuttig, and T. Taubner, “Using low-loss phase-change materials for mid-
infrared antenna resonance tuning.,” Nano Lett., vol. 13, no. 8, pp. 3470–5,
Aug. 2013.
[254] A.-K. U. Michel, P. Zalden, D. N. Chigrin, M. Wuttig, A. M. Lindenberg, and

T. Taubner, “Reversible optical switching of infrared antenna resonances with
131
ultrathin phase-change layers using femtosecond laser pulses,” ACS
Photonics, vol. 1, no. 9, pp. 833–839, Aug. 2014.
[255] P. Hosseini, C. D. Wright, and H. Bhaskaran, “An optoelectronic framework

enabled by low-dimensional phase-change films,” Nature, vol. 511, no. 7508,
pp. 206–211, Jul. 2014.
[256] F. F. Schlich, P. Zalden, A. M. Lindenberg, and R. Spolenak, “Color

Switching with Enhanced Optical Contrast in Ultrathin Phase-Change
Materials and Semiconductors Induced by Femtosecond Laser Pulses,” ACS
Photonics, vol. 2, no. 2, pp. 178–182, Feb. 2015.
[257] K. Shportko, S. Kremers, M. Woda, D. Lencer, J. Robertson, and M. Wuttig,

“Resonant bonding in crystalline phase-change materials.,” Nat. Mater., vol.
7, no. 8, pp. 653–8, Aug. 2008.
[258] M. Cretich, M. R. Monroe, A. Reddington, X. Zhang, G. G. Daaboul, F.

Damin, L. Sola, M. S. Unlu, and M. Chiari, “Interferometric silicon biochips
for label and label-free DNA and protein microarrays,” Proteomics, vol. 12,
no. 19–20, pp. 2963–2977, 2012.
[259] M. Cretich, A. Reddington, M. Monroe, M. Bagnati, F. Damin, L. Sola, M. S.

Unlu, and M. Chiari, “Silicon biochips for dual label-free and fluorescence
detection: Application to protein microarray development,” Biosens.
Bioelectron., vol. 26, no. 9, pp. 3938–3943, 2011.
[260] R. Jenison, S. Yang, a Haeberli, and B. Polisky, “Interference-based detection

of nucleic acid targets on optically coated silicon.,” Nat. Biotechnol., vol. 19,
no. 1, pp. 62–65, 2001.
[261] P. Mitchell, “F EATURE A perspective on protein microarrays Proteins , not

genes , are the true targets of medicines , but their analysis by array
technology still poses significant challenges for drug developers .,” Nat.
Biotechnol., vol. 20, pp. 225–229, 2002.
[262] P. F. Predki, “Functional protein microarrays: Ripe for discovery,” Curr.

Opin. Chem. Biol., vol. 8, no. 1, pp. 8–13, 2004.
[263] M. Zhao, X. Wang, and D. D. Nolte, “Molecular interferometric imaging.,”

Opt. Express, vol. 16, no. 10, pp. 7102–7118, 2008.
[264] A. F. Coskun, A. E. Cetin, B. C. Galarreta, D. A. Alvarez, H. Altug, and A.

Ozcan, “Lensfree optofluidic plasmonic sensor for real-time and label-free
monitoring of molecular binding events over a wide field-of-view,” Sci. Rep.,
vol. 4, p. 6789, 2014.
[265] A. E. Cetin, A. F. Coskun, B. C. Galarreta, M. Huang, D. Herman, A. Ozcan,

and H. Altug, “Handheld high-throughput plasmonic biosensor using
computational on-chip imaging,” Light Sci. Appl., vol. 3, no. 1, p. e122, 2014.
132
[266] M. Perino, E. Pasqualotto, a De Toni, D. Garoli, M. Scaramuzza, P. Zilio, T.
Ongarello, a Paccagnella, and F. Romanato, “Development of a complete
plasmonic grating based sensor and its application to Self Assembled
Monolayer detection,” vol. 53, no. 26, pp. 24–26, 2014.
[267] U. S. Dinish, F. C. Yaw, A. Agarwal, and M. Olivo, “Development of highly

reproducible nanogap SERS substrates: Comparative performance analysis
and its application for glucose sensing,” Biosens. Bioelectron., vol. 26, no. 5,
pp. 1992–1987, 2011.
[268] A. Gopinath, S. V Boriskina, W. R. Premasiri, L. Ziegler, B. M. Reinhard, and

L. Dal Negro, “Plasmonic nanogalaxies: multiscale aperiodic arrays for
surface-enhanced Raman sensing.,” Nano Lett., vol. 9, no. 11, pp. 3922–9,
Nov. 2009.
[269] A. Barhoumi and N. J. Halas, “Label-free detection of DNA hybridization

using surface enhanced Raman spectroscopy.,” J. Am. Chem. Soc., vol. 132,
no. 37, pp. 12792–3, Sep. 2010.
[270] L. Rodríguez-Lorenzo, R. a Alvarez-Puebla, I. Pastoriza-Santos, S. Mazzucco,

O. Stéphan, M. Kociak, L. M. Liz-Marzán, and F. J. García de Abajo,
“Zeptomol detection through controlled ultrasensitive surface-enhanced
Raman scattering.,” J. Am. Chem. Soc., vol. 131, no. 13, pp. 4616–8, Apr.
2009.
[271] M. R. Gartia, A. Hsiao, A. Pokhriyal, S. Seo, G. Kulsharova, B. T.

Cunningham, T. C. Bond, and G. L. Liu, “Colorimetrics: Colorimetric
Plasmon Resonance Imaging Using Nano Lycurgus Cup Arrays (Advanced
Optical Materials 1/2013),” Adv. Opt. Mater., vol. 1, no. 1, pp. 1–1, 2013.
[272] R. de la Rica and M. M. Stevens, “Plasmonic ELISA for the ultrasensitive

detection of disease biomarkers with the naked eye,” Nat. Nanotechnol., no.
October, pp. 2–5, 2012.
[273] R. de la Rica and M. M. Stevens, “Plasmonic ELISA for the detection of

analytes at ultralow concentrations with the naked eye.,” Nat. Protoc., vol. 8,
no. 9, pp. 1759–64, 2013.
[274] E. Ozkumur, J. W. Needham, D. a Bergstein, R. Gonzalez, M. Cabodi, J. M.

Gershoni, B. B. Goldberg, and M. S. Unlü, “Label-free and dynamic detection
of biomolecular interactions for high-throughput microarray applications.,”
Proc. Natl. Acad. Sci. U. S. A., vol. 105, no. 23, pp. 7988–7992, 2008.
[275] G. G. Daaboul, R. S. Vedula, S. Ahn, C. a. Lopez, a. Reddington, E. Ozkumur,

and M. S. Ünlü, “LED-based Interferometric Reflectance Imaging Sensor for
quantitative dynamic monitoring of biomolecular interactions,” Biosens.
Bioelectron., vol. 26, no. 5, pp. 2221–2227, 2011.
[276] R. L. Stears, T. Martinsky, and M. Schena, “Trends in microarray analysis,”

Nat. Med., vol. 9, no. 1, 2003.
133
[277] S. Kumar Vashist, “Nanoparticles-Based Naked-Eye Colorimetric
Immunoassays for In Vitro Diagnostics,” J. Nanomed. Nanotechnol., vol. 05,
no. 01, pp. 1–3, 2014.
[278] J. M. Slocik, J. S. Zabinski, D. M. Phillips, and R. R. Naik, “Colorimetric

response of peptide-functionalized gold nanoparticles to metal ions,” Small,
vol. 4, no. 5, pp. 548–551, 2008.
[279] M. A. Kats, D. Sharma, J. Lin, P. Genevet, R. Blanchard, Z. Yang, M. M.

Qazilbash, D. N. Basov, S. Ramanathan, and F. Capasso, “Ultra-thin perfect
absorber employing a tunable phase change material,” Appl. Phys. Lett., vol.
101, no. 22, p. 221101, Nov. 2012.
[280] F. F. Schlich, P. Zalden, A. M. Lindenberg, and R. Spolenak, “Color

Switching with Enhanced Optical Contrast in Ultrathin Phase-Change
Materials and Semiconductors Induced by Femtosecond Laser Pulses,” ACS
Photonics, vol. 2, no. 2, pp. 178–182, Feb. 2015.
[281] M. Piliarik, H. Sípová, P. Kvasnička, N. Galler, J. R. Krenn, and J. Homola,

“High-resolution biosensor based on localized surface plasmons.,” Opt.
Express, vol. 20, no. 1, pp. 672–80, 2012.
134
Appendix A
A.1. Effective index simulation in MIM geometry
clc;clear all;
LM=1000:10:10000;
AL_rakic;
nal=interp1(nk(:,1),nk(:,2),LM);
kal=interp1(nk(:,1),nk(:,3),LM);
emetal=(nal-sqrt(-1)*1e0*kal).^2;
dgap=[1e-9:1e-9:6e-9];
for pp=1:length(dgap)
for lll=1:length(LM)
d=dgap(pp);
ed=1.6*1.6;
lambda=LM(lll)*1e-9;
w=2*pi*3e8/lambda;
em=emetal(lll);
k0=2*pi/lambda;
kd=k0*sqrt(ed);
km=k0*sqrt(em);
b=0:k0/100:50*k0;
kd=sqrt(b.^2-ed*k0*k0);
km=sqrt(b.^2-em*k0*k0);
xx=ed*km+em*kd.*tanh(kd*d/2);
x2=abs(real(xx));
qx=abs(imag(xx));
bbb=b(find(x2==min(x2)));
neff(pp,lll)=bbb/k0;
end
end
135
A.2. Field Profile Simulation in Multilayer Coatings
clear all
close all
clc
m=1;
w=300:1:1000;
Al_rakic; %Refractive index of Al from Rakic
nag=interp1(nk(:,1),nk(:,2),w,'cubic','extrap');
kag=interp1(nk(:,1),nk(:,3),w,'cubic','extrap');
aSiPVD_nk;%Refractive index of aSi

nd=interp1(nk(:,1),nk(:,2),w,'cubic','extrap');
kd=interp1(nk(:,1),nk(:,3),w,'cubic','extrap')+0.0;
rr=zeros(1,length(w));
ii=1;
theta=THETA;
for tt=[80]
for WL=w
n=[1.0 1.5 nd(ii)-sqrt(-1)*kd(ii) nag(ii)-sqrt(-1)*kag(ii) 1.5];
t=[1200 tt 20 100 200 pol='tm';
[EE position2 r t]=fresnelmultilayer_field(WL,n,t,pol,1);
E(ii,:)=EE;
rr2(ii)=r;
ii=ii+1;
end
ii=1;
rrr(m,:)=rr2;
m=m+1;
end
plot(w,(abs(rrr).^2))
figure
imagesc((abs(rrr).^2))
figure
plot(position2,abs(E3([201],:)).^2);
136
A.3. Transfer Matrix Code
function [EE position r tt]=fresnelmultilayer_field(lamda,n,t,pol,res)

nn=length(n);
tt=length(t);
if (pol=='te')
Ao=[1 1;n(1) -n(1)];
end
if (pol=='tm')
Ao=[1 1;n(1) -n(1)];
end
M=eye(2);
for i=2:nn-1
k=(2*pi/lamda)*n(i);
Tm=[exp(sqrt(-1)*k*t(i)) 0;0 exp(-sqrt(-1)*k*t(i))];
Am=[1 1;n(i) -n(i)];
Amm=Am^-1;
M=M*Am*Tm*Amm;
end
M=M;
kt=2*pi/lamda*n(nn);
dm=sum(t(2:(nn-1)));
X=Ao^-1*M*[exp(-sqrt(-1)*kt*dm);exp(-sqrt(-1)*kt*dm)*n(nn)];
%%Calculate Transmission and reflection coefficients

tt=1/X(1,1);
r=X(2,1)/X(1,1);
Eb=r*1;
Et=tt*1;
E=zeros(1,sum(t)/res);
EE=[];tf=0;
El=[1;Eb];
Al=Ao;
zz=[];
for i=2:nn-1
z=[tf:res:tf+t(i)-res];
zz=[zz z];
tf=t(i)+tf;
k=(2*pi/lamda)*n(i);
Tm=[exp(sqrt(-1)*k*t(i)) 0;0 exp(-sqrt(-1)*k*t(i))];
Am=[1 1;n(i) -n(i)];
Er=Tm^-1*Am^-1*Al*El;
137
Ei=Er;
EE=[EE Ei(1,1)*exp(-sqrt(-1)*k*(z-tf))+Ei(2,1)*exp(sqrt(-1)*k*(z-tf))];
El=Er;
Al=Am;
end
k=(2*pi/lamda)*n(1);
kt=(2*pi/lamda)*n(nn);
z1=[-t(1):res:-res];z2=sum(t(2:nn-1)+res):res:sum(t(2:nn-1))+t(nn);
EE=[1*exp(-sqrt(-1)*k*(z1))+Eb*exp(sqrt(-1)*k*(z1)) EE Et*exp(-sqrt(-
1)*kt*(z2))];
position=[z1 zz z2];
138

PHD Thesis

Uploaded by

Copyright:

Available Formats

PHD Thesis

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

PHD Thesis

Uploaded by

Copyright:

Available Formats

COUPLED PLASMONIC STRUCTURES

FOR SENSING, ENERGY AND

THE GRADUATE SCHOOL OF ENGINEERING AND SCIENCE

IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR

MATERIALS SCIENCE AND NANOTECHNOLOGY PROGRAM

Assist. Prof. Dr. Aykutlu Dana (Advisor)

Assist. Prof. Dr. Necmi Biyikli

Assoc. Prof. Dr. Ozgur Oktel

Assoc. Prof. Dr. Hamza Kurt

Assist. Prof. Dr. Mehmet Solmaz

Approved for the Graduate School of Engineering and Science:

Prof. Dr. Levent Onural

COUPLED PLASMONIC STRUCTURES FOR SENSING, ENERGY

Keywords: Surface plasmons, coupled plasmons, spectroscopy, hot-electrons,

Eşlenmiş Plazmonik Yapıların Sensör, Enerji ve Spektroskopi

Nanofabrikasyon ve karakterizasyon metotlarındaki yeni gelişmeler yenilikçi

Anahtar kelimeler: Yüzey plazmonları, eşlenik plazmonlar, spektroskopi, sıcak

Acknowledgement .............................................................................................. viii

List of Figures ...................................................................................................... xii

List of Tables ...................................................................................................... xxv

Chapter 1 Introduction ...................................................................................... 1

Chapter 2 Well-Defined Coupled Plasmonic Structures ................................ 7

2.1. Surface Plasmons at Metal-dielectric Interface ........................................... 8

2.2. Surface Plasmons in Metal-Insulator-Metal Resonators ........................... 11

2.3. Raman Enhancement on a Broadband Meta-surface................................. 13

2.3.1. Introduction ...................................................................................... 13

2.3.2. Description of the Plasmonic Meta-Surface ..................................... 14

2.3.3. Coupled Meta-Atoms and Meta-Molecules ..................................... 17

2.3.4. Experimental Results ........................................................................ 21

2.3.5. Discussion of Spatial Uniformity of Enhancement at the Nanoscale

2.4. Edge Rounding Induced Broadband Plasmonic Metamaterial Absorber

2.4.1. Simulation Geometry and Method ................................................... 29

2.4.2. Results .............................................................................................. 30

Chapter 3 Nano-particle Based Large Area Broadband Plasmonic

3.1. Large Area Plasmonic Metasurfaces ......................................................... 36

3.2.1. Observation of Single Molecule Raman Events Using Smart Phone's

3.3.1. Introduction ...................................................................................... 48

3.3.2. Results .............................................................................................. 49

3.4. Conclusion .................................................................................................. 55

Chapter 4 Exploiting Native Aluminum Oxide for Multispectral Aluminum

4.1. Introduction ................................................................................................ 57

4.2. Native Oxide Based Plasmonic Meta-surfaces........................................... 58

4.3. Nanoparticle Based Large Area Surfaces .................................................. 64

4.4. Surface Enhanced Infrared Absorption (SEIRA) Spectroscopy and Surface

4.5. All Aluminum Hierarchical Surfaces in the Infrared ................................. 75

4.6. Conclusion .................................................................................................. 82

Interference Coating Based Sensing Platforms ................................................ 83

5.1. Thermally Tunable Ultrasensitive Vibrational Spectroscopy Platforms

5.1.1. Introduction ...................................................................................... 84

5.1.2. Results .............................................................................................. 86

5.1.3. Thermally Tunable IR Spectroscopy Platforms Using Ge2Sb2Te5... 93

5.2.1. Introduction ...................................................................................... 99

5.2.2. Results ............................................................................................ 101

5.2.3. Protein Binding Experiment ........................................................... 106

5.3. Conclusion ................................................................................................ 107

Chapter 6 ............................................................................................................ 109

Conclusion .......................................................................................................... 109

Appendix A ........................................................................................................ 135