1

Dye-Sensitized Solar Cells: History, Components, Configuration,

and Working Principle
S.N. Karthick 1 , K.V. Hemalatha 2 , Suresh Kannan Balasingam 3 , F. Manik Clinton 4 , S. Akshaya 5 ,
and Hee-Je Kim 6
1
Electrochemical Materials and Devices Lab, Department of Chemistry, Bharathiar University, Coimbatore, Tamil Nadu, India
2 Department of Chemistry, Coimbatore Institute of Technology (CIT), Coimbatore, Tamil Nadu, India
3
Department of Materials Science and Engineering, Faculty of Natural Sciences, Norwegian University of Science and Technology
(NTNU), Trondheim, Norway
4
Electrochemical Materials and Devices Lab, Department of Chemistry, Bharathiar University, Coimbatore, Tamil Nadu, India
5
Electrochemical Materials and Devices Lab, Department of Chemistry, Bharathiar University, Coimbatore, Tamil Nadu, India
6
School of Electrical and Computer Science Engineering, Pusan National University (PNU), Gumjeong-Ku, Jangjeong-Dong,
Busan, Republic of Korea

1.1 Introduction
The ever-growing human population requires the consumption of energy in various forms,
and therefore researchers in energy field focus on energy harvesting from various sources. The
nonrenewable energy sources such as fossil fuels are running out, which cannot be replen-
ished in our life time. The nonrenewable energy sources are carbon-based fossil fuels such as
coal, petroleum, and natural gas that emits greenhouse gases (for example carbon dioxide) that
cause global warming, a serious threat to the world and mankind. At present, worldwide around
three-fourth of the electricity is obtained from the nonrenewable sources that cannot be reused
or recycled [1]. Many countries such as Japan, China, France, Ukraine, and India depend on
nuclear power stations for the production of electricity and also they are facing several harmful
issues from these power plants that lead to environmental pollution [2]. Therefore, the focus
of scientists mainly rely on the renewable energy-based energy conversion devices. Solar, wind,
hydroelectric, biomass, and geothermal are some of the examples of renewable energy resources
available in our earth. Of these, solar energy is an important source of renewable energy, which
is available throughout a day all over the year, basically inexhaustible in nature. In case of solar
energy, radiation obtained from the sunlight is capable of producing heat and light, causes pho-
tochemical reactions, and generates electricity. As the electricity becomes a first and foremost
basic need for the mankind, this impressive energy source can be utilized for the conversion of
solar to electrical energy using solar cell technology. The strength of solar energy is magnani-
mous as it provides us about 10 000 times more energy that is higher than the world’s daily need
of energy consumption [1]. The earth receives such a huge amount of energy every day, we are
fortunate to harness it using suitable solar cell technologies. Regrettably, though solar energy
is free of cost, the highly expensive technologies required for its conversion and storage which
limit the technology to reach the wider community.
The concept of solar energy harvesting has been evolving since eighteenth century. Edmond
Becquerel, a French scientist, has first discovered the photovoltaic (PV) effect in 1839 [3].
This effect has become a starting point for the solar energy harvesting applications. The
Interfacial Engineering in Functional Materials for Dye-Sensitized Solar Cells, First Edition.
Edited by Alagarsamy Pandikumar, Kandasamy Jothivenkatachalam and Karuppanapillai B. Bhojanaa.
© 2020 John Wiley & Sons, Inc. Published 2020 by John Wiley & Sons, Inc.
2 1 Dye-Sensitized Solar Cells: History, Components, Configuration, and Working Principle

Types of solar
cell

Silicon based Semiconductor Emerging

materials materials materials

Crystalline Amorphous Chalcogenides Group iii–v DSSC

Quantum
a-Si:H dots
CdTe GaInP
Single Perovskite
crystalline
CZnTS GaAs
Organic
Multi
crystalline CIGS Other

Figure 1.1 Classification of solar cells technologies. Source: Ibn-Mohammed et al. 2017 [7]. Reprinted with
permission of Elsevier.

light energy to electrical energy conversion is being done by the special photoactive devices
called photovoltaic cells. When the semiconductor material absorbs light, it produces elec-
trical voltage and this effect is called photovoltaic effect. The first-generation solar cells use
crystals of silicon to attain this effect. In 1887, German scientist Hertz has first examined
the photoelectric effect and found that photons present in the light are capable of ejecting
free electrons from a solid surface (usually conductor) to create power. However, based on
his preliminary results, he also found that the same process produced more power when the
conductor is incident with UV light rather than more intense visible light [4, 5]. Based on this
phenomenon, the modern solar cells rely on the photoelectric effect to convert sunlight into
electricity. Generation of solar cells are broadly classified into three different categories based
on the materials properties and the period of time they evolved. For example, silicon-based
materials (crystal and amorphous) belong to the first-generation solar cells, semiconductor
materials (III–IV group chalcogenides/phosphide materials) belong to the second-generation
solar cells, and emerging materials (Figure 1.1) belong to the third-generation solar cells. From
1953 to 1956, physicists at Bell Laboratory fabricated silicon solar cells with 6% efficiency,
which is more efficient than selenium. This discovery paved a way for identifying the capability
of the solar cells to power up the electrical equipment; again, the experimentation continued
to improve the performance and attempted to make new devices for commercialization.
Approximately, after 10 years, the second-generation thin-film solar cells were developed. The
second-generation solar cells are often described as emerging thin-film solar cells that converts
30% of the solar radiation into electrical energy [6]. The semiconductor materials used in this
generation is copper indium gallium selenide (CIGS), cadmium telluride (CdTe), and gallium
arsenide (GaAs). The devices made out of these materials are commercially available and used
in space crafts. After 1990s, the third-generation solar cell technologies have been emerged.
This new generation photovoltaic technologies include dye-sensitized solar cells (DSSCs),
organic/polymer solar cells, quantum dot solar cells, perovskite solar cells, etc. The power
conversion efficiency of these third-generation solar cells are lower than silicon-based solar
cells and thin-film solar cells, but it has its own advantages such as low processing costs and
less environmental impact that induce the intensive research and development in this area.
 1.2 History of Dye-sensitized Solar Cells 3

The production cost of first- and second-generation solar cells are reported to be more than
1 US$/W. The third-generation PV technologies aim to produce the large-scale electricity with
a low cost of less than 0.5 US$/W. This can be achieved with the module cost at the rate of
70 US$/m2 revealing 14% efficiency [1] for third-generation solar cells.

1.2 History of Dye-sensitized Solar Cells

During 1839, Becquerel [3] found that a voltage was produced when two platinum electrodes
were immersed in the electrolyte containing a metal halide salt when it is illuminated with light.
After that, the illumination of light on the organic dye is capable of producing electricity at the
semiconductor electrodes in the electrochemical cell was discovered during 1960s [8]. Later, the
phenomenon of photoexcitation was studied primarily at University of California during 1970s
in order to simulate the mechanism of photosynthesis by extracting the chlorophyll pigment
from spinach and using ZnO as semiconductor electrode material in electrochemical cells [9].
The mechanism was recognized when the photoexcitation of dye molecules injected electrons
into the conduction band of the n-type semiconductor scaffold material and found that the dye
molecules adsorbed on the semiconductor monolayer material was responsible for maximum
yield. This study forms the basis of the bionic or biomimetic approach of sensitizing semicon-
ductor materials for electrons excitation. Again, in 1972, the generation of electricity through
dye sensitization was demonstrated by Tributsch [10]. In the study conducted by Michio Mat-
sumura et al., in the year of 1980 [11], they inferred that the efficiency of the dye molecules
could be improved by fine tuning the porosity of the semiconductor oxide material, but the
stability of the dye is a mere challenge in the dye-sensitized photocell. Then the progress in
this field was initiated in 1991 by Prof. Michael Grätzel who performed the sensitized electro-
chemical PV device made of dye sensitization on semiconductor TiO2 material and named it
as “dye-sensitized solar cell” [12]. They architectured the device toward the new conceptual of
PV energy generation. The new conceptual approach is based on the three billion years old idea
of nature’s photosynthetic activity. They inferred that PV device is modeled based on the light
absorption and electron transfer mechanisms in plant. When the sunlight falls on the chloro-
phyll molecule, the electron generated from chlorophyll is transferred from one molecule to
the other till it reaches the chlorophyll reaction center. From the reaction center, the electron
is transferred to the energy storage molecule. The chlorophyll lacks with one electron is being
grabbed from the surrounding water molecule. This cyclical process is being imitated for har-
vesting energy using sun’s radiation through synthetic dye material. In the first attempt, they
received <2% efficiency from the DSSCs. It has its own drawback with the chlorophyll dye such
as adsorption range and stability of dye molecules. To increase the performance of the device,
they used the ruthenium-based dyes and polymer gel electrolyte with good thermal stability in
the year of 2003 and reached around 6% maximum efficiency [13]. In 2005, researchers con-
centrated on improving the quantum efficiency of DSSCs, by modifying the morphology of the
photoanode scaffold material and improving the electron transfer in the semiconductor, design-
ing the particle with a small size having high surface area for more dye adsorption, etc. Then
they replaced the ruthenium-based dye with low-cost dyes and achieved around 5% efficiency
in 2005 [14]. In 2008 [15], 8% efficiency was achieved due to the iodine-based redox liquid elec-
trolyte. Then the researchers attempted to solve the corrosive nature of the electrolyte, which
helped to improve the photovoltage, efficiency, and the stability of devices. In this period, the
efficiency of the devices increased up to 11.5% [16]. Finally, the efficiency was further improved
more than 13% using porphyrin ring as dye molecules and cobalt(II/III) redox shuttle as the
electrolyte [17, 18]. Later in 2009, DSSCs were first commercialized by G24 Power Limited,
South Lake Drive, Imperial Park, Newport, United Kingdom.
4 1 Dye-Sensitized Solar Cells: History, Components, Configuration, and Working Principle

1.3 Components of DSSCs

The basic components of DSSCs primarily consist of transparent conducting oxide (TCO)
film-coated glass substrates, dye, photoanode, electrolytes, and counter electrode (CE).

1.3.1 Conductive Glass Substrate

The transparent conductive oxide-coated glass was used as substrates for both photoanode
and counter electrode. The coating of this layer is required for the collection of electrons
ejected from the photoanode (dye-coated TiO2 ) and pass it to the counter electrode through
outer circuit. Therefore, it is called current collector. Generally, TCO films are made by both
fluorine-doped tin oxide (FTO) and indium-doped tin oxide (ITO). This film has a very low elec-
trical resistance of <20 Ω/sq at room temperature. The conductive substrate should be made
of highly abundant material, low cost with a maximum transparency of solar radiation [19].

1.3.2 Photoanode
A typical photoanode material consists of dye molecules-adsorbed semiconductor oxide mate-
rial coated on TCO substrate, which acts as a working electrode in DSSCs. The most widely
used semiconductor material is an anatase phase of titanium dioxide (TiO2 ), having a band gap
of 3.2 eV. It is an inexpensive, nontoxic, and abundant material present in the earth’s crust. TiO2
paste could be made using some binders such as terpineol, ethyl cellulose, polyethylene glycol,
ethylene glycol, and so on, for a strong adhesion on the FTO glass plate. Screen printing or
doctor blade techniques are being used for the coating of viscous TiO2 paste onto the substrate.
Removal of binder is usually performed by annealing the metal oxide paste-coated FTO glass
plate at 450–500 ∘ C for 30 minutes. Due to annealing, the metal oxide has good adhesion with
TCO glass substrate, interparticle contact, and eventually forms the nanostructured porous
electrode. The thickness of the formed porous electrode film is around 10–15 μm and the inter-
connected nanoparticle size ranges from 15 to 30 nm. Titanium dioxide has three different
crystalline phases such as anatase, rutile, and brookite. The images of different TiO2 crystal
structures is shown in Figure 1.2. Apart from the titanium dioxides, the other semiconductor
metal oxides such as ZnO, SiO2 , Zn2 SnO4 , CeO2 , WO3 , SrTiO3, and Nb2 O5 [20–26] were also
used as the photoanode scaffold material in DSSCs. The performance of the device was poor
when these materials are used in the photoanode. When ZnO is used in the photoanode doped
with magnesium, it showed the efficiency of 4.11% [27]. The efficiency was 6.8% with SiO2 ,
3.8% with Zn2 SnO4 , CeO2 does not show any energy conversion efficiency, WO3 showed the
efficiency of 1.46%, SrTiO3 , and Nb2 O5 showed 4.8% and 2%, respectively. TiO2 is concluded
to be the best semiconductor photoanode material because of its good efficiency and stability
compared to the other materials. Among the above metal oxides, ZnO showed better efficiency
because it has a band gap close to that of TiO2 and found to contain the higher carrier mobility.
Hence, it is found to be the efficient photoanode that can replace TiO2 . However, the recombina-
tion of the photoexcited charge carriers and the instability of the ZnO surface toward the acidic
dyes were the major setback for the ZnO to provide high conversion efficiency. Researchers
tried to improve the efficiency by modifying the structure, surface morphology, particle size,
thickness optimization of photoanode, and so on.

1.3.3 Counter Electrode

The three important functions of counter electrode as a potential catalyst is (i) It is responsi-
ble for the completion of process, that is the oxidized redox couple is reduced at the surface
 1.3 Components of DSSCs 5

Anatase Rutile

a
b
c
b a
c
Brookite

O
Ti

a
c

Figure 1.2 Different crystal structures of titanium dioxide (anatase, rutile, brookite). Source: Haggerty et al.
2017 [28]. Licensed under CC BY 4.0. https://www.nature.com/articles/s41598-017-15364-y#rightslink.

of counter electrode by accepting electrons. In solid-state DSSCs, the oxidized redox couple
gets reduced by collecting the electrons from the counter electrode through ionic transport
material. (ii) It collects electrons from the outer circuit and transfers them into inner cell elec-
trolyte solution. So the ultimate role of the counter electrode is to collect electrons from the load
and transfer them to the cell for circulation. (iii) The counter electrode acts as the mirror and
reflects the unabsorbed light to the photoanode for the enhanced utilization of the sunlight [29].
To satisfy these basic functions, the counter electrode should possess high catalytic activity,
highly conductivity and reflectivity, high surface area with porous nature, chemical corrosion
resistance, high chemical and mechanical stability, energy level of counter electrode should
match with the potential of the redox electrolyte, good adhesiveness toward TCO substrate. It
is impossible to achieve all these parameters simultaneously with a single material. For example
carbon-based counter electrode was found to achieve the high-surface area with porous nature,
chemical corrosion resistance but it does not reflect the light. The counter electrode has much
influence on photovoltaic parameters. The maximum photovoltage determined at zero current
is due to the energy difference between the redox mediator and the semiconductor metal oxide.
Under load, the output voltage seems less than the open-circuit voltage. This loss in voltage is
due to the overpotential of the counter electrode that derives from the delivery of the current
through electrolyte and electrolyte/CE interface. Platinum paste is used as a counter electrode
material in standard cells for the completion of redox reaction. The reduction of tri-iodide
was carried out by the counter electrode. The rate of reduction of tri-iodide to iodide at the
counter electrode is important since after the redox reaction, formation of iodide is important
to regenerate dye molecules at the anode side of the cell. For a workable cell, a slow reaction is
carried out at the anode side and the fast reaction takes place at the cathode side. Apart from
the platinum material, carbon, graphite, and conductive polymers are also used as counter elec-
trode materials, platinum is still the best choice of catalysts due to its facile iodide/tri-iodide
redox kinetics. Counter electrodes are being prepared by various preparation techniques such
as sputtering, chemical vapor deposition, hydrothermal, electrodeposition, etc.. The platinum
prepared by electrodeposition technique by Zhong et al., showed higher electrocatalytic activity
6 1 Dye-Sensitized Solar Cells: History, Components, Configuration, and Working Principle

than the pure Platinum [30]. Also, the light reflection value of platinum is higher than the car-
bon materials, which causes more reflection of light into the cell. Bimodal mesoporous carbon
embedded with nickel prepared by thermal pyrolysis under nitrogen environment by Wu et al.
showed the highest efficiency of 8.6% which was higher than that of Pt as counter electrode [31].
CoS films as counter electrodes prepared by Huo et al., by electrophoretic deposition and ion
exchange deposition technique showed the efficiency of 7.72% [32]. Carbon nanotube (CNT)
arrays were prepared by Nam et al. by the chemical vapor deposition technique. He inferred that
the highly ordered structure of CNT would rapidly diffuse the electrolyte into the electrode and
also helps in fast electron transport and achieved efficiency was 10.04% [33]. The conductive
polymers such as polyaniline, polypyrrole, polyethylene dioxythiophene, and polythiophene
derivatives were also used as counter electrodes. Anothumakkool et al. prepared polyethy-
lene dioxythiophene-impregnated cellulose paper as a flexible counter electrode through in situ
polymerization technique that achieved the efficiency of 6.1% which is on par with the efficiency
achieved with Pt/FTO as counter electrode [34]. Since platinum is a very expensive material,
recently researchers focus on many different metals and chalcogenide-based alternative counter
electrodes for DSSCs.

1.3.4 Electrolytes
The electrolyte is yet another important compound playing an essential role in DSSCs. The
presence of redox mediator in the electrolyte acts as a charge or ion transporter between the
counter electrode and photoanode. The redox couple should not observe the light at the visible
range. It should be chemically stable and reversible. It should have low viscosity for the rapid
diffusion of charge carriers. It should not desorb the dye or disperse the semiconductor mate-
rial adhered to the FTO plate. The most common, stable, and highly efficient redox mediator
is iodide/tri-iodide redox couple. This redox mediator is promising due to its high penetration
power into semiconductor nanoporous film, slow recombination loss, and the fast regenera-
tion of dye molecules. This redox shuttle is used at present as an electrolyte for any standard
cells. It is found to be an ideal electrolyte having a low viscosity, negligible vapour pressure, opti-
mum boiling point, and dielectric properties. It also has important properties such as chemical
inertness, environmental sustainability, and easy processing. It is believed that the redox cou-
ple encourages the charge transport between photoanode and counter electrodes. When the
dye molecules inject electrons into the semiconductor metal oxide, the tri-iodide in the elec-
trolyte helps in reducing the oxidized dye as quickly as possible to the ground state [35]. The
iodide, an electron acceptor moves to the counter electrode and receives its lost electrons and
regenerate as tri-iodide. The circuit is completed by the electron migration through the external
load. For dye and electrolyte regeneration, the redox potential of the iodide/tri-iodide should
be considered.

1.3.4.1 Types of Solvents Used in Electrolytes

The solvent used for electrolytes preparation in DSSCs are categorized into aqueous, organic
and ionic liquid-based solvents.

Aqueous Electrolyte The first investigation with aqueous electrolyte (various pH ranging from
2 to 7) was done by Grätzel between pH ranging from 2 to 7 was done by Grätzel and O’Regan
group by photoelectrochemical dye sensitization of porous TiO2 electrodes. The research on
water-based electrolytes was also fully concentrated for its nonflammability, nonvolatility,
cost-effectiveness, and environmental compatibility. The aqueous electrolyte that is acidified
for the better adsorption of dye molecules toward the TiO2 is possible only when the pH is
below the iso-electric point of TiO2 which is 5. Next, the shift in the position of conduction
 1.3 Components of DSSCs 7

band edge of TiO2 takes place toward the negative direction of the energy scale while pH is
decreasing [36].
Lindquist et al. demonstrated an experiment by adding small amount of water in nonaqueous
electrolyte. The addition of water leads to elevation in photovoltage and decrease in
photocurrent density [37]. It is inferred that addition of water weakens the bonding between
dye molecules and the metal oxide and undergoes desorption of dye molecules. So a special
dye variety should be considered when pure or mixed aqueous electrolyte is used. According
to O’Regan and coworkers addition of 20% of water into nonaqueous electrolyte enhances the
power energy conversion efficiency from 5.5% to 5.7% [38].

Organic Solvents-based Electrolytes Grätzel and coworkers developed the alcohol-based solvents
for the preparation of lithium iodide electrolyte. It gives an efficiency of 11%, however while
using aprotic solvents as electrolytes the bond between semiconducting material and dye
become weaker and show instability toward aqueous solvents due to the hydrophilicity in
nature [39]. Then nitrile-based solvents such as methyl cyanide, glutaronitrile, and valeronitrile
were used as solvents for the electrolyte, but these are toxic in nature and expensive. So the
derivative of nitrile like 3-methoxypropionitrile (MPN) and methoxyacetonitrile (MAN) were
used. MPN showed comparatively good stability and exhibited 7.6% efficiency [35].

Ionic Liquids-Based Electrolytes Ionic liquids were recently found to be the most promising sol-
vents used in electrolyte mixture because of their stability, conductivity, and low vapor pres-
sure. The long-term stable ionic liquid electrolyte was first employed in 1996 by Papageorgiou
et al. using methyl-hexyl-imidazolium iodide as solvent in DSSCs rather than other liquid elec-
trolytes [40]. Xi and his team performed the experiments with nonimidazolium ionic liquids,
i.e. S-ethyltetrahydrothiophenium tricyanomethide and S-ethyltetrahydrothiophenium iodide,
showed better efficiency of 7.2% and 6.9%, respectively [41].

1.3.4.2 Alternative Redox Mediators

Though iodide/tri-iodide redox shuttle has many advantages, it also has its dark-side that
they will easily corrode the substrates like silver and copper metal joints when they are used
as current collectors and the photovoltage was also suppressed by the redox potential of
the iodide/tri-iodide mediator. It also absorbs the visible light and some amount of pho-
tons when the dye is irradiated. So the research moves toward some other alternatives to
iodide/tri-iodide mediator [42]. The bromide/tri-bromide redox couple is found to be one of
the alternatives for iodide/tri-iodide electrolyte because of its high positive redox potential.
This bromide/tri-bromide redox couple is suitable if eosin Y dye is used as a sensitizer
[43]. Pseudohalogen redox couples such as SeCN− /(SeCN)3 − and SCN− /(SCN)3 − were also
used as the redox mediators having more positive equilibrium potential of 0.19 and 0.43 V,
respectively, than I− /I3 − [44]. A halogen-free sulfur-based mediator like disulfide/thiolate
(T− /T2 ) is also used as the redox couple with ruthenium-based dye as a sensitizer and achieved
the efficiency of 6.44% [45]. Here, T− is 5-mercapto-1-methyltetrazole anion and T2 is its
dimer. Cobalt(II/III) redox shuttle was also used as the electrolyte, which increases the overall
efficiency to 13% [17]. SpiroMeOTAD, a solid hole conductor is capable of transporting charge,
and dye regeneration could be used as charge mediators [42]. Polymer electrolytes are also
under extensive investigation. Ionic liquids are another promising alternative electrolytes that
are known for their advantages such as mutable, excellent chemical, and thermal stability,
high-ionic conductivity. So finding the superior redox couple is one of the main challenges for
researchers working in DSSCs field.
8 1 Dye-Sensitized Solar Cells: History, Components, Configuration, and Working Principle

1.3.5 Dyes
The dye plays the centralized role in DSSCs by ejecting the electrons on irradiation and
initiating the mechanism. The following are the good qualities of dyes. It should be adsorbed
on photoanode and capture the sunlight at all wavelengths (𝜆) below 920 nm. It should have
the strong anchoring groups such as –COOH, –SO3 H, –H2 PO3 for the adsorption toward
photoanode. The excited state of the dye should have higher energy than the conduction band
of the photoanode material for the easy transport of electrons from the dye to the photoanode.
For the p-type DSSCs, the HOMO level of dye should have a more positive potential than
the valence band of p-type semiconductors. For the regeneration of dyes, it should have
more positive oxidized state level than the redox potential of electrolyte. Optimization of
the molecular structure of dye prevents the dye aggregation on the semiconductor electrode
surface. The dye should have the high stability toward light and heat. Many photosensitizers
such as metal complexes, porphyrins, phthalocyanines, and metal-free organic dyes were
used for the past decades. Normally, the metal complexes have a central metal ion with
ancillary ligand having atleast one anchoring group. Metal ligand charge transfer process is
responsible for the visible light absorption in the solar spectrum. The metal ligand complex
(cis-bis(4,4′ -dicarboxy-2,2′ -bipyridine)diisothiocyanato-ruthenium(II)) named N719 dye was
taken as the reference dye in DSSCs due to its improved power conversion efficiency. A
thiocyanato derivative cis-(SCN)2bis(2,2′ -bipyridyl-4,4′ -dicarboxylate)ruthenium(II) coded
as N3 dye is an another sensitizer taken as the reference for the standard cell showed the
efficiency of 10% having the longer excited state lifetime of ≈20 seconds. Again, to improve
the efficiency of the sensitizers to near-infrared (NIR) region, ligands of the Ru complex have
been modified by Grätzel and his workers. They developed N749 dye also called as black dye
where Ru is bonded with three thiocyanato and one terpyridine ligands substituted with three
carboxyl groups. The power conversion efficiency was 10.4% using black dye. The structure of
N719, N3, black dyes are shown in Figure 1.2. A series of Ru complexes with phenanthroline
ligand Ru(dcphen)2 (NCS)2 –(TBA)2 as an anchor was studied as the photosensitizer showed
the efficiency of 6.6%. The other metal complexes were prepared using Os, Re, Fe, Pt, and Cu
as a central metal atom [46]. Some of the other sensitizers such as synthetic dyes such as N749,
Z907, N712, D35, and natural dyes extracted from flowers, fruits, and vegetables were also
investigated. The rockstar dyes of DSSCs are N719, N3, and black dyes structures which are
shown in Figure 1.3.
Choi et al. synthesized a novel synthetic organic dye using nonplanar bis-dimethyl-
fluorenylamino moiety and a terthiophene unit into the organic framework and cyanoacrylic
acid moiety as an acceptor and anchoring group. With the novel synthetic dye, the achieved
efficiency was 8.60%, which is on par to the efficiency of a standard cell [47]. Wang et al.
used coumarin dyes based on the concept of donor-pi–conjugated bridge-acceptor, with –CN
groups as electron acceptor and showed the power conversion efficiency of 7.6% [48]. Hanaya
and coworkers during 2015 achieved an efficiency of 11.2% using alkoxysilyl and carboxy
anchor dye as cosensitizers [18]. Hemalatha et al. extracted the natural sensitizers carotenoid
and anthocyanin dye from Rosa chinensis and Kerria japonica flowers, respectively, and for the
dye stability, sugar molecules were added to the dye. The power conversion was very low and
does not exceed 0.29% [49].

1.4 Configuration of DSSCs

1.4.1 Metal Substrates for Photoanode and Glass/TCO for Counter Electrode
The basic configuration of DSSCs consists of TCO as a glass as substrate. Dye coated with
TiO2 as photoanode. In this configuration, TCO film transmits photons and conducts electrons
 1.4 Configuration of DSSCs 9

O N(CH2CH2CH2CH3)3

O C

COOH
HOOC
N
N
N
Ru
N O
N C O
N
C
S C N(CH2CH2CH2CH3)3
S
(a)

COOH

HOOC

N COOH

N
N
Ru
N
C N
N
S
C COOH

S
(b)

HOOC

O O

(CH3CH2CH2CH2)3N O C O N(CH2CH2CH2CH3)3
C
N
N N
Ru
N
N
C N
C
S
C S

S
(c)

Figure 1.3 Structures of (a) N719, (b) N3, and (c) black dyes used in DSSCs.

and glass provides the back support. Platinum thin-film coated TCO/glass is used as a counter
electrode, N719 dye as a sensitizer and iodide/tri-iodide as an electrolyte. Compactness with
flexible DSSCs can be achieved by modifying the configuration of DSSCs. The modification may
be done in TCO substrate, electrode, dye, electrolyte, or redox mediator. The configuration of
DSSCs is attempted using thin metal foil as a substrate replacing both TCO and rigid glass sub-
strate. It can withstand high sintering temperature and when TiO2 is coated on the metal foil,
the interconnectivity of the particle increases and internal resistance was also reduced. Initially,
Grätzel’s group deposited TiO2 on the titanium foil [50]. The first solid-state flexible DSSCs
were developed by Meyer et al. using titanium and stainless steel foil as a substrate and formed
the TiO2 material over the substrate by rapid sintering method. TiO2 was adsorbed with the dye
followed by the insertion of spiro-OMeTAD a hole-transporting material into the active matrix
10 1 Dye-Sensitized Solar Cells: History, Components, Configuration, and Working Principle

of TiO2 . This configuration showed very low efficiency of 0.8% under 1 sun illumination [51].
These metal substrates were then replaced by several other metals such as Pt, Co, Ti, Al, and
W [52] as flexible substrates for photoanode, but at the interface, they form a corresponding
insulating oxides that affect the conductivity. This could be retarded by making the ITO layer as
an interface in between metal substrate and TiO2 because it prevents the direct contact of TiO2
with the metal that avoids the oxidation of metals. In these configurations, TCO–glass sub-
strates were used as the counter electrode that allows the photons during the back illumination
process [50].

1.4.2 Metal Substrates for Counter Electrode and Glass/TCO for Photoanode
Recently, researchers focused on enhancing the efficiency of DSSCs using metal substrates as
counter parts of the cells such as platinized nickel, stainless steel, titanium, aluminum, cop-
per. But Al and Cu used as counter electrodes showed the weak stability over iodide/tri-iodide
electrolyte. So they remain unfit for making the cell, whereas, Pt–nickel and Pt–stainless steel
counter electrodes gave high-electrical conductivity and low-internal resistance. The platinum
that is coated over the nickel using the sputtering, electroplating technique gives an efficiency
of 5.4%, which is comparatively higher than that of the Pt–TCO glass substrate [53]. Kim and
Rhee investigated stainless steel-based counter electrodes (stainless steel–Pt) on the perfor-
mance of DSSCs. The platinum thin-film sputtered on stainless steel substrate counter electrode
showed the highest efficiency of 7.7% when coupled with glass–FTO/titanium oxide nanopar-
ticles (TNPs) as a photoanode [54].

1.4.3 Metal Substrate for Photoanode and Polymer Substrate for Counter Electrode
The flexible DSSCs with TNPs on titanium foil as working electrode and poly(ethylene
naphthalate)-indium-doped tin oxide/platinum-thinfilm, PEN-ITO/Pt as a counter electrode
were fabricated by Grätzel and coworkers that exhibits an efficiency of 7.2% with back
illumination process. Though the efficiency was highest, it is lower than that of the efficiency
achieved by glass substrates. In another configuration, stainless steel substrate coated with
ITO and SiOx followed by TiO2 paste as working electrode and Pt coated polymer was used as
a counter electrode showing the efficiency of 4.2% at back illumination [55] which is still lower
than DSSCs configured by glass substrate. Grätzel and coworkers build the DSSCs with TiO2
nanotube coated on the titanium foil as the working electrode and glass–ITO/Pt substrate
as the counter electrode. On back illumination at counter electrode, the efficiency was about
3.3%. Same configuration is built using different polymer substrates such as PEDOT, PET, PEN
instead of glass (polymer–ITO/Pt) as the counter electrode that exhibited the efficiency of
3.6% at back illumination [56].

1.4.4 Polymer Substrates for Flexible DSSCs

The rigid glass substrate was replaced with flexible transparent polymer flims such as
poly(ethylene terephthalate) (PET), poly(ethylene naphthalate) (PEN), and poly(ether sulfone)
(PES) coated with TiO2 as a photoanode. The TiO2 material works well only at the calcination
temperature of 450 ∘ C, which is not the favorable for the transparent conducting low melting
point polymer films. So the calcination temperature was limited to 200 ∘ C. Though this
temperature is not optimized for the fabrication of TiO2 , various low temperature method-
ologies such as microwave heating, hydrothermal crystallization, mechanical compression,
electrophoretic deposition, chemical sintering were adopted for the fabrication of flexible solar
cells irrespective of fabrication methods. The polymer substrates-based flexible solar cells was
 1.5 Working Principle of DSSCs 11

found to give a very low efficiency compared to glass substrates because of the poor adhesion
of the TiO2 material on polymer substrate and the lack of interparticle contact between TiO2
particles due to the low calcination temperature [50]. But Arakawa and coworkers obtained
the highest efficiency of 8.1% by fabricating flexible DSSCs using plastic substrate, coating
TiO2 paste on the substrate by mechanical pressing method [57]. This is the highest efficiency
of DSSCs using plastic as a substrate.

1.4.5 Glass/TCO-Free Metal Substrates for Flexible DSSCs

The conductive mesh-based flexible DSSCs without TCO substrates were fabricated by Fan
et al. using stainless steel mesh coated with porous TiO2 as photoanode and 0.5 mm thick Pt foil
as counter electrode. The efficiency was 1.49% with 10 μm thick TiO2 layer [58]. Yoshida et al.
configured DSSCs with TCO-free substrate. They used stainless steel mesh-TiO2 as a work-
ing electrode and Pt-coated titanium foil as counter electrode that showed the efficiency of
5.56% from front illumination [59]. Quasi solid-state DSSCs using a gel electrolyte and stain-
less steel mesh as a substrate were fabricated by Huang and coworkers. MgO deposited TiO2
colloidal solution was spray coated on the stainless steel mesh and sintered at high tempera-
ture followed by deposition of 1D TiO2 nanofibers by electrospinning technique. Stainless steel
substrate coated with polypyrrole or the platinum foil was used as the counter electrode. The
efficiency was 2.8% with platinum foil as counter electrode, whereas the polypyrrole-coated
stainless steel mesh achieved the efficiency of 2.3% [60]. A p–n junction-based flexible DSSCs
using copper mesh as a substrate were fabricated by Heng et al. A p-type cuprous iodide was
first deposited on copper mesh followed by the deposition of nanostructured TiO2 by electro-
spun technique. It was then sensitized with Ru-based dye sensitizer. Pt foil acts as the counter
electrode. The device was irradiated at Cu-mesh/CuI layer, the direct contact of Cu layer with
TiO2 forms the p–n junction that enabled the separation of electron and hole and transports
the charge effectively which enhanced the J SC value and efficiency (4.73%) of the cell [61].

1.4.6 Glass/TCO-Free Metal Wire Substrates for Flexible DSSCs

In 2008, Fan et al. designed twisted wire-shaped DSSCs using stainless steel-wire coated with
TNPs as working electrodes and poly(vinylidene fluoride) (PVDF) polymer-coated Pt-wire as
counter electrodes [62]. Both the electrodes were twisted together and configured to form the
wire-shaped DSSCs. Cylinder-shaped DSSCs were fabricated by Liu and Misra by inserting
Ti-wires/TiO2 -nanotube as working electrodes into the capillary tube along with Pt wires. The
efficiency was found to be 2.78% for 55 μm long titanium dioxide nanotubes (TNTs) [63]. Liu
et al. constructed TiO2 -coated titanium wires as working electrodes and platinum-coated tita-
nium foil as counter electrodes that exhibited the efficiency of 4.1% by front-side illumination
[64].
The following is the list of works done by various scientists all over the world using different
metallic and polymeric substrates as the working and counter electrodes for flexible DSSCs (see
Table 1.1).

1.5 Working Principle of DSSCs

The working principle of DSSCs involves the following processes:
1) Light absorption
2) Charge separation
3) Charge collection
12 1 Dye-Sensitized Solar Cells: History, Components, Configuration, and Working Principle

Table 1.1 Various configuration evolved in dye-sensitized solar cells.

Substrate Working electrode Counter electrode Done by

Metal substrate and Ti/TiO2 FTO/Pt Kim et al.

TCO – glass Ti/ZnO FTO/Pt Fan et al.
substrate
SS/TiO2 FTO/Pt Vijayakumar et al.
Zn/ZnO FTO/Pt Gao et al.
Metal substrate and Ti and SS/TiO2 ITO–PET/Pt Chang et al.
TCO – polymer Ti/ZnO ITO–PEN/Pt Lin et al.
substrate
Ti/TiO2 ITO–PEN/Pt-SWCNT Xiao et al.
TNT/SrO–TiO2 ITO–PET Chen et al.
Both as metal Ti/TiO2 Ti/Pt Xiao et al.
substrate SS/TiO2 SS/Pt Bonilha et al.
SS/TiO2 Glass paper/Pt Cha et al.
Ti/TiO2 Ti/Pt/PEDOT Wu et al.

Source: Balasingam et al. 2013 [50]. Reproduced with permission of Royal Society of Chemistry.

The mechanism resembles and imitates the natural processes of plants that convert sunlight
into chemical energy called photosynthesis. In DSSCs, solar to electrical energy conversion
occurs by ruthenium-based dye-sensitized nanocrystalline TiO2 photoanode. In DSSCs, charge
separation is carried out by a kinetic approach in the same manner as photosynthesis process
that leads to photochemical action. The well-configured DSSCs have its unique and specific
mechanism [65].
Prior to the discussion about the working principle of DSSCs, the factors influencing the
solar to electrical conversion efficiency of the cell should be considered. The efficiency of DSSCs
depends on the three criterial aspects: (i) the frontier molecular orbitals of dye (i.e. highest occu-
pied molecular orbital (HOMO) and the lowest unoccupied molecular orbital, LUMO). (ii) On
the Fermi-level of scaffold material and (iii) the redox-potential of the electrolyte shuttle. The
energy gap between the HOMO and LUMO of the dye molecules should be as short as possible.
So that electrons can be easily excited from ground state to excited state by the absorption of
photons. Then the conduction band of titanium dioxide must be lower to the LUMO level of
dye molecules to facilitate the excited electrons flow from dye to TiO2 material and then circu-
late to external circuit via TCO substrate. Working principle of DSSCs is quite interesting due
to the flow of electrons from the photoanode to counter electrode, which generates electricity
(Figure 1.4). The working mechanism of the DSSCs are as follows.
The heart of the system is a mesoscopic semiconductor oxide film, which is placed in contact
with a redox electrolyte or an organic hole conductor. Attached to the surface of the nanocrys-
talline film is a monolayer of the dye sensitizer. Under sunlight illumination, the dye sensitizer
will absorb photons (light) and the energy of the photon is sufficient to excite the electron from
HOMO to LUMO level of the dye molecule and become photoexcited. The charge separation
is attained across dye and TiO2 interface, where an electron is located in the TiO2 and a hole
is located in the oxidized dye molecule. The adsorbed dye molecules inject electrons from the
LUMO level of dye into the conduction band of nanocrystalline anatase phase TiO2 which is
lower in energy than the excited state of the dye (Figure 1.5). The electrons transferred through
the porous network of TiO2 eventually reach the back contact of the working electrode where
charge collection occurs. The extracted charge can subsequently perform electrical work in the
external circuit and eventually return to the counter electrode where platinum is coated on the
 1.5 Working Principle of DSSCs 13

FTO glass Counter electrode

TiO2 dye Electrolyte FTO glass

Injection
–0.5 Dye•
E versus NHE (V)

Maximum
0 hv voltage

Red Ox
Mediator
0.5 Diffusion
Interception

1.0
Dye°/Dye–

Figure 1.4 Schematic diagram of the working principle of DSSCs.

hγ

Ti O
e–

N719 Dye

Anatase (TiO2)

Figure 1.5 Interaction of dye and TiO2 . Reprinted with permission from ref. [66]
14 1 Dye-Sensitized Solar Cells: History, Components, Configuration, and Working Principle

FTO glass substrate that acts as a counter part of the cell. This platinum (2+) which already
gets oxidized by the electrolyte receives the electron from the external circuit and is reduced
to the platinum (0). Now again, the platinum will get oxidized by donating the electron to the
electrolyte. The electron enters into the iodine electrolyte that has a potential to carry out the
redox reaction and regenerate the dye molecule. In the electrolyte, there is an iodine molecule
(I2 ) and iodide ion (I– ) that normally is combined as I3 – (triiodide ion). This ion accepts the
electron from the counter electrode and gets reduced and forms three iodide ions (3I– ). Again,
these iodide ions get oxidized to triiodide ions by donating the electrons to dye. So the liquid
redox electrolyte will complete the circuit by reducing the oxidized dye. The voltage generated
under illumination corresponds to the difference between the Fermi level of the electron in the
solid and the redox potential of the electrolyte [67]. Overall, the device generates electric power
from light without suffering any permanent chemical transformation.

1.5.1 Electron Transfer Mechanism in DSSCs

TiO2 ∕S + Photon (h𝜐) → S∗ (Excitation process)

TiO2 ∕S∗ + TiO2 → TiO2 ∕S+ + e−(CB) (Injection process)

TiO2 ∕2S+ + 3I− → TiO2 ∕2S + I−3 e−(CE) (Regeneration)

I−3 + 2e− → 3I− (Reduction)

I−3 + 2e− → 3I− (Re − caption in dark reaction)

TiO2 ∕S+ + e−(CB) → TiO2 ∕S (Recombination in dark reaction)

1.5.2 Photoelectric Performance

Under the radiation of solar spectrum with photon flux, the photon-energy-to-electricity con-
version efficiency is defined as follows:
JSC × VOC × FF
𝜂=
J0
where J SC , short circuit current; V OC , open circuit voltage; FF, fill factor; and I 0 , photon flux.
When the cell is exposed to illumination from the direct sun, the photocurrent and voltage
drop can be interpreted through I–V characteristics curve. The short circuit current I SC varies
with dye concentration. The magnitude of the photocurrent depends upon the nature and con-
centration of the dye used. When the concentration of the dye is high, the performance of the
cell is high. On the other hand, the lower concentration of dyes reduces the efficiency of the
cell and the magnitude of the photocurrent [68]. The incident photon-to-current conversion
efficiency (IPCE) depends upon the magnitude of the short circuit current, the photoresponse,
and the light intensity as given below:
1240 (eV nm) × Jsc (μA∕cm2 )
IPCE (𝜆) =
𝜆 (nm) × I (μW∕cm−2 )
where 𝜆 is the wavelength of the adsorbed photon; I, light intensity at wavelength 𝜆.
 References 15

Acknowledgments
The corresponding author, Dr. S.N. Karthick would like to thank the funding support of Uni-
versity Grant Commission - Basic Scientific Research (UGC-BSR) Start-Up-Grant (2018-2020)
from the University Grant Commission, New Delhi, India, the National Research Foundation of
Korea (Project No. 2015R1D1A4A01019537) and BK 21 PLUS, creative human resource devel-
opment programme for IT convergence, Republic of Korea.

References
1 Kalyaanasundram, K. (2010). Fundamental Sciences in Dye Sensitized Solar Cells, vol. 1, 543.
EPFL Press, Published by CRC Press.
2 Horvath, A. and Rachlew, E. (2016). Ambio 45: 38–49.
3 Williams, R. (1960). J. Chem. Phys. 32: 1505.
4 Mc Quarrie, D.A. (2016). Viva Student Edition. Sausalito, CA: University Science Books.
5 Wong, D., Lee, P., Shenghan, G. et al. (2011). Eur. J. Phys. 32: 1059–1064.
6 Chopra, K.L., Paulson, P.D., and Dutta, V. (2004). Prog. Photovoltaics Res. Appl. 12: 69–92.
7 Ibn-Mohammed, T., Koh, S.C.L., Reaney, I.M. et al. (2017). Renew. Sustain. Energy Rev. 80:
1321–1344.
8 Gerischer, H., Michel-Beyerle, M.E., Rebentrost, F., and Tribijxch, H. (1968). Electrochim.
Acta 13: 1509–1515.
9 Tributsch, H. and Calvin, M. (1971). Photochem. Photobiol. 14: 95–112.
10 Tributsch, H. (1972). Photochem. Photobiol. 16: 261–269.
11 Matsumura, M., Matsudaira, S., and Tsubomura, H. (1980). Ind. Eng. Chem. Prod. Res. Dev.
19: 415–421.
12 O’Regan, B. and Grätzel, M. (1991). Nature 353: 737–740.
13 Wang, P., Zakeeruddin, S.M., Moser, J.E. et al. (2003). Nat. Mater. 2: 402–407.
14 Wang, Q., Campbell, W.M., Bonfantani, E.E. et al. (2005). J. Phys. Chem. B 109:
15397–15409.
15 Bai, Y., Cao, Y., Zhang, J. et al. (2008). Nat. Mater. 7: 626–630.
16 Chen, C.-Y., Wang, M., Li, J.-Y. et al. (2009). ACS Nano 3: 3103–3109.
17 Mathew, S., Yella, A., Gao, P. et al. (2014). Nat. Chem. 6: 242–247.
18 Kakiage, K., Aoyama, Y., Yano, T. et al. (2015). Chem. Commun. 51: 15894–15897.
19 Yang, Z., Gao, S., Li, T. et al. (2012). ACS Appl. Mater. Interfaces 4: 4419–4427.
20 Lin, C.-Y., Lai, Y.-H., Chen, H.-W. et al. (2011). Energy Environ. Sci. 4: 3448–3455.
21 Choi, J.-W., Kang, H., Lee, M. et al. (2014). RSC Adv. 4: 19851–19855.
22 Tan, B., Toman, E., Li, Y., and Wu, Y. (2007). J. Am. Chem. Soc. 129: 4162–4163.
23 Sayyed, S.A.A.R., Beedri, N.I., Kadam, V., and Pathan, H.M. (2016). Mater. Sci. 39:
1381–1387.
24 Zheng, H., Tachibana, Y., and Kalantar-zadeh, K. (2010). ACS Langmuir 26: 19148–19152.
25 Kim, C.W., Suh, S.P., Choi, M.J. et al. (2013). J. Mater. Chem. A 1: 11820–11827.
26 Sayama, K., Sugihara, H., and Arakawa, H. (1998). Chem. Mater. 10: 3825–3832.
27 Justin Raj, C., Prabakar, K., Karthick, S.N. et al. (2013). J. Phys. Chem. C 117: 2600–2607.
28 Haggerty, J.E.S., Schelhas, L.T., Kitchaev, D.A. et al. (2017). Sci. Rep. 7: 15232.
29 Wu, J., Lan, Z., Lin, J. et al. (2017). Chem. Soc. Rev. 46: 5975–6023.
30 Zhong, C., Hu, W.B., and Cheng, Y.F. (2011). J. Power Sources 196: 8064–8072.
31 Wu, M., Chen, C., Chen, Y., and Shih, H. (2016). Electrochim. Acta 215: 50–56.
32 Huo, J., Zheng, M., Tu, Y. et al. (2015). Electrochim. Acta 159: 166–173.
16 1 Dye-Sensitized Solar Cells: History, Components, Configuration, and Working Principle

33 Nam, J., Park, Y., Kim, B., and Lee, J. (2010). Scr. Mater. 62: 148–150.
34 Anothumakkool, B., Agrawal, I., Bhange, S.N. et al. (2016). ACS Appl. Mater. Interfaces 8:
553–562.
35 Wu, J., Lan, Z., Lin, J. et al. (2015). Chem. Rev. 115: 2136–2173.
36 Kalyanasundaram, K., Vlachopoulos, N., Krishnan, V. et al. (1987). J. Phys. Chem. 91:
2342–2347.
37 Liu, Y., Hagfeldt, A., Xiao, X.R., and Lindquist, S.E. (1998). Sol. Energy Mater. Sol. Cells 55:
267–281.
38 Law, C.H., Pathirana, S.C., Li, X.O. et al. (2010). Adv. Mater. 22: 4505–4509.
39 Nazeeruddin, M.K., Liska, P., Moser, J. et al. (1990). Helv. Chim. Acta 73: 1788–1803.
40 Papageorgiou, N., Athanassov, Y., Armand, M. et al. (1996). J. Electrochem. Soc. 143: 3099.
41 Xi, C., Cao, Y., Cheng, Y. et al. (2008). J. Phys. Chem. C 112: 11063.
42 Burschka, J., Dualeh, A., Kessler, F. et al. (2011). J. Am. Chem. Soc. 133: 18042.
43 Boschloo, G. and Hagfeldt, A. (2009). Acc. Chem. Res. 42: 1819–1826.
44 Oskam, G., Bergeron, B.V., Meyer, G.J., and Searson, P.C. (2001). J. Phys. Chem. B 105:
6867–6873.
45 Wang, M., Chamberland, N., Breau, L. et al. (2010). Nat. Chem. 2: 385–389.
46 Shalini, S., Balasundaraprabhu, R., Satish Kumar, T. et al. (2016). Int. J. Energy Res. 40:
1303–1320.
47 Choi, H., Baik, C., Kang, S.O. et al. (2008). Angew. Chem. Int. Ed. 47: 327–330.
48 Wang, Z.-S., Cui, Y., Dan-oh, Y. et al. (2008). J. Phys. Chem. C 112: 17011–17017.
49 Hemalatha, K.V., Karthick, S.N., Justin Raj, C. et al. (2012). Spectrochim. Acta, Part A 96:
305–309.
50 Balasingam, S.K., Kang, M.G., and Jun, Y. (2013). Chem. Commun. 49: 1471–1487.
51 Meyer, T.B., Meyer, A.F., and Ginestoux, D. (2002). Proc. SPIE Int. Soc. Opt. Eng. 13: 4465.
52 Man, G.K., Park, N.G., Kwang, S.R. et al. (2005). Chem. Lett. 804.
53 Ma, T., Fang, X., Akiyama, M. et al. (2004). J. Electroanal. Chem. 574: 77–83.
54 Kim, J.M. and Rhee, S.W. (2012). J. Electrochem. Soc. 159: B6–B11.
55 Ito, S., Ha, N.L.C., Rothenberger, G. et al. (2006). Chem. Commun.: 4004.
56 Kuang, D., Brillet, J., Chen, P. et al. (2008). ACS Nano 2: 1113–1116.
57 Yamaguchi, T., Tobe, N., Matsumoto, D. et al. (2010). Sol. Energy Mater. Sol. Cells 94:
812–816.
58 Fan, K., Chen, J., Yang, F., and Peng, T. (2012). J. Mater. Chem. 22: 4681–4686.
59 Yoshida, Y., Pandey, S.S., Uzaki, K. et al. (2009). Appl. Phys. Lett. 94: 093301.
60 Huang, X., Shen, P., Zhao, B., Feng, X., et al. (2010). Sol. Energy. Mater. Sol. Cells 94:
1005–1010.
61 Heng, L., Wang, X., Yang, N. et al. (2010). Adv. Funct. Mater. 20: 266–271.
62 Fan, X., Chu, Z., Wang, F. et al. (2008). Adv. Mater. 20: 592–595.
63 Liu, Z. and Misra, M. (2010). ACS Nano 4: 2196–2200.
64 Liu, Y., Li, M., Wang, H. et al. (2010). J. Phys. D: Appl. Phys. 43: 205103.
65 Grätzel, M. (2003). J. Photochem. Photobiol., C 4: 145–153.
66 Grätzel, M. (2001). Nature 414: 338–344.
67 Zhao, Z. and Liu, Q. (2007). J. Phys. D: Appl. Phys.41: 025105.
68 Yanagida, S., Yu, Y., and Manseki, K. (2009). Acc. Chem. Res. 42: 1827–1838.