Nano Energy (2013) 2, 174–189

Available online at www.sciencedirect.com

journal homepage: www.elsevier.com/locate/nanoenergy

REVIEW

Fabrication of flexible dye sensitized solar

cells on plastic substrates
Hasitha C. Weerasinghea,b,1, Fuzhi Huanga,1, Yi-Bing Chenga,n

a
Department of Materials Engineering, Monash University, Wellington Road, Clayton, Victoria 3800, Australia
b
Bio21 Institute, University of Melbourne, 30 Flemington Road, Parkville, Victoria 3010, Australia

Received 16 August 2012; received in revised form 14 October 2012; accepted 14 October 2012
Available online 23 October 2012

KEYWORDS Abstract
Photovoltaic; This article introduces the latest progress in the research on fabrication of flexible dye-
Dye sensitized solar sensitized solar cells (DSSCs), in particular the choice of flexible plastic substrates replacing the
cell; conventionally used rigid glass substrates. A major challenge for making DSSCs on plastic
Flexible substrates; substrates is the temperature limitation of the substrate in producing the TiO2 working
Plastic substrates;
electrode. Several low-temperature fabrication methods for nano-porous TiO2 films, such as
Roll to roll;
ball-milling, acid/water treatments, chemical vapor deposition and electrophoretic deposition,
Low temperature
processing as well as recently studied chemical and mechanical film processing methods such as chemical
sintering, hydrothermal treatment, microwave irradiation, and different compression techni-
ques are extensively discussed here. It also presents studies on new fabrication methods of
flexible counter electrodes and recently developed new materials particularly useful for
flexible DSSCs. At last, the developments and prospects of fabricating large scale flexible DSSC
modules and their durability are discussed.
& 2012 Elsevier Ltd. All rights reserved.

Introduction cells dominate the commercial market at the present time,

but their manufacturing involves energy intensive pro-
Photovoltaic (PV) technology has been realized as a suitable cesses and high vacuum conditions, which lead to high
renewable power source for the fulfilment of increasing world manufacturing costs. In this regard, dye sensitized solar
energy consumption with least impact on the environment. cells (DSSCs) [1,2] and organic solar cells (OSCs) [3–5] have
In the last two decades there has been a growing interest for demonstrated their potentials as relatively low-cost alter-
fabrication of low-cost solar cells based on molecular materi- natives to silicon solar cells. An attractive advantage of
als and semiconductor nano-structures. Silicon based solar OSCs is the potential for roll-to-roll printing on polymer
substrates. The absence of a liquid electrolyte also makes
n
Corresponding author. Tel.: +61 3 9905 4930; processing and sealing of OSC devices comparatively
fax: +61 3 9905 4940. simpler. However, electrochemical solar cells using TiO2
E-mail address: [email protected] (Y.-B. Cheng). electrodes, in combination with Ru-based dyes and iodine-
1
Contributed equally to this paper. based liquid electrolyte are on the verge of entering the

2211-2855/$ - see front matter & 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.nanoen.2012.10.004
Fabrication of flexible dye sensitized solar cells 175

photovoltaic market due to its superior power conversion

efficiency (PCE) over the organic solar cells and relatively
low production cost. Efficiencies over 12.5% have already
been achieved for the lab size devices [6] and over 8% TCO coated substrate
efficiency has been recently reported for devices using a p-
type compound (CsSnI3 xFx) as a solid state electrolyte [7]. Working
Lately, there has been an increased demand for light- Electrode
weight, flexible electronic devices as they can be produced
in large scales using printing techniques at room tempera-
ture with a much reduced production cost. The concept of - - (Electrolyte)
I/ I3
flexible polymer based solar cells has thus attracted much
interest amongst researchers in the field of DSSCs. However Counter
there are many unanswered issues related to the fabrication Platinum coated Counter electrode Electrode
of flexible DSSCs. In this article, the technological develop-
TiO2 Dye
ment of DSSCs on flexible substrates is reviewed, with a
Particles Molecules
focus on polymer substrates and the factors related to the
slurry preparation, deposition of films and processing of
electrodes to improve the mechanical and photovoltaic
properties of the device.

Structure and working principle of DSSCs

The structure of a dye sensitized solar cell mainly consists of

three major components, namely: the working electrode
which is usually a dye molecule coated nanocrystalline
porous TiO2 film deposited on a transparent conductive
oxide (TCO) coated substrate, the counter electrode which
is often a platinum-coated TCO deposited substrate and
an electrolyte containing usually an I /I3 redox couple.
Figure 1(A) illustrates the components in a typical DSSC. Figure 1 (A) Structure and (B) principle of a dye sensitized
In the 10 mm thick mesoporous working electrode, TiO2 solar cell.
particles are bonded to each other and to the TCO glass
substrate by sintering at 450–550 1C. Additionally, the
mesoporous structure of the TiO2 film provides a large
surface area to accommodate a high amount of dye the TiO2 surface. Figure 1(B) shows the operation principle
molecules which enhances the amount of light absorp- of the DSSC. Dye molecules absorb the incident photons and
tion [8]. A variety of nano crystalline semiconductor parti- get excited from a low-energy state (HOMO-highest occu-
cles (mainly TiO2) can be prepared by the sol–gel technique. pied molecular orbital) to a high-energy state (LUMO-lowest
The TiO2 film is usually deposited on a TCO coated glass unoccupied molecular orbital). The excited state of the dye
substrate using TiO2 slurry containing various solvents and molecule is capable of injecting electrons into the CB of the
organic binders. The binders in the TiO2 slurry reduce the TiO2 particle. The oxidized dye is then regenerated and
stresses of the film during drying, and assist in the becomes ready for the next excitation by obtaining elec-
formation of well-adhered titania layers on the TCO glass trons from the redox electrolyte converting iodide (I ) in to
substrates. These electrodes are then sintered at 450– tri-iodide (I3 ). The electrons injected to the CB of TiO2
550 1C to remove the binders in the film, as well as to form percolate through the TiO2 particles in the film until they
chemical bonding between TiO2 particles and the TiO2 are collected at the TCO contact of the photoelectrode, and
particles with the TCO layer on the substrate via a solid are fed in to the external circuit. The tri-iodide species (I3 )
state diffusion mechanism. Sintered electrodes are soaked are converted back to iodide (I ) via the platinum catalyst
in a dye solution to allow for the dye adsorption, and then at the counter electrode by the electrons traveled through
sandwiched with a platinum-coated counter electrode. the external circuit.
The Pt coating is usually prepared either by thermal
decomposition or by sputtering techniques. The voids
between TiO2 particles and the space between the TiO2 Flexible DSSCs
layer and the counter electrode are filled with a liquid
electrolyte containing I and I3 redox couple. Generally, Advantageous of flexible DSSCs
these types of solar cells are exposed to the solar radiation
from the working electrode side of the TCO glass, known as For any photovoltaic technology, flexible solar panels offer
front illumination (Figure 1(A)). some advantages over classic rigid solar panels. When it
The working principle of the DSSC is based on the comes to the installation, plastic substrate based PVs can be
absorption of photons and excitation of the dye, followed installed quickly on buildings with flat or curved surfaces
by fast electron injection into the conduction band (CB) of because of the light weight and flexible nature of these
176 H.C. Weerasinghe et al.

types of PVs. Therefore, although DSSCs fabricated on TCO Preparation of TiO2 pastes for flexible DSSCs
coated glass have achieved efficiencies above 12%, there is
a great deal of interests in cheaper, flexible solar cells using Preparation of suitable TiO2 pastes with appropriate rheol-
transparent indium tin oxide (ITO) coated plastic electrodes ogy to produce high quality mechanically stable films with-
(usually polyethylene naphthalate (PEN) and polyethylene out organic binders is the first major challenge facing when
terephthalate (PET)) in place of the glass. Flexibility and fabricating DSSCs on plastic substrates. In order to prevent
light weight are two important advantages of using plastic film cracks due to the large agglomerates, the TiO2 nano-
substrates. Flexible solar cells would be adaptable to particles are required to be well dispersed in the paste. Our
different shapes of surfaces and light weight would enable work showed the importance of the ball-milling technique
them as a mobile power source for portable electronic for the preparation of well dispersed TiO2 slurries using
devices. relatively cheap Degussa P25 TiO2 powder as the raw
Flexible DSSCs are important not only for user conve- material [11]. De-agglomeration by such a milling led to
nience, but also for realizing roll-to-roll mass production. an increase of the film strength and the surface area,
The cost of solar cells would be significantly reduced by this improving dye adsorption in titania films. A maximum power
fast, continuous manufacturing process. TCO coated glass conversion efficiency of 4.2% was achieved for the plastic
substrates account for a significant portion of the cost of the based solar cells prepared using the ball milling technique.
device. A replacement of polymer substrates for glass is It is very important to evaluate the inter-particle bonding of
expected to reduce the materials cost as well. the mesoporous nanocrystalline metal oxide films. However,
currently there is no such method developed to measure the
inter-particle bonding between metal oxide nano particles.
Here, a method of nanoindentation was firstly developed for
Challenges in flexible DSSCs this purpose. A nano-scratch test was performed with the
nanoindentation instrument to measure the force required
Deposition of TiO2 nano crystalline films on plastic for making a scratch, which has shown its potential as a
substrates for working electrodes has become a great more quantitative alternative to the conventional testing
challenge, as most plastic substrates limit the thermal methods such as bending deformation [12,13], pencil
treatment to around 150 1C. This precludes using any scratch test [14–16] and measuring the resistance of the
organic binders during the film deposition, as they would film for the ultra-sonication [17]. In a typical nanoscratch
normally remain in the TiO2 film processed below 400 1C. test, a stylus or a diamond tip is moved over the film
Absence of binders in TiO2 pastes would result in the surface. The test is performed by the combination of
formation of cracks in the film during drying and thus two types of tip movement with respect to the film surface
increase electrical resistivity. Decreased adhesion of the in normal (vertical) and lateral (horizontal) motions.
TiO2 film with the substrate and poor electrical contact Forces exerted on the tip are monitored in normal and
between TiO2 particles are the other difficulties when lateral directions while forming a scratch with known
plastic substrates are used due to low temperature proces- depth and length, and they are used to compare the film
sing. Poor stability of the devices due to the ingress of strength [11]. Using this technique, the mechanical strength
moisture and oxygen through the highly permeable plastic of the mesoporous TiO2 films was found to increase with
substrates is a major problem faced by both DSSC and OSC longer milling time of the slurries, corresponding to some
technologies based on plastic substrates. Thus a rigorous improvements in conversion efficiency of plastic based DSSC
encapsulation technology that has no significant effect on devices [11].
the production cost would be essential when commercializ- For printing processes, such as screen printing and gravure
ing these modules. printing, pastes with high viscosity are required. However,
The working electrode of a DSSC consists of a porous TiO2 commonly used water or alcohol based binder free titania
film of 10 mm thickness. When electrons are injected from slurries usually exhibit very low viscosity. Therefore investiga-
the dye molecules into TiO2 particles, they must travel tion of the factors associated with the paste rheology and
through a large number of titania nanoparticles (106 and their influence on the inter-particle connectivity has been a
107) in order to reach the TCO layer [9]. In glass-based key interest for fabricating flexible DSSCs. To date, only a few
DSSCs, the inter-particle chemical connection (so called highly viscous pastes suitable for a single step coating process
necking) and strong connectivity of the titania layer with on ITO coated plastic substrates have been developed [18,19].
the substrate are achieved by the high temperature sinter- The rheological behavior of ethanol based titania pastes was
ing process, which are essential for achieving high efficiency investigated and it was found that with the addition of
DSSCs. Furthermore, due to lack of binders, the TiO2 slurries ammonia, hydrochloric acid and water, the viscosity of TiO2
used for casting films often have difficulties to wet hydro- pastes could vigorously be changed (Figure 2) [20,21]. The
phobic polymer substrates, resulting in poor adhesion with change in the viscosity has been correlated with the measured
the substrates [10]. The restriction of high temperature zeta potential of the colloidal titania pastes. Improved
heat treatments is a major factor responsible for the inter-particle connectivity and hence better solar cell
relatively low conversion efficiency of plastic based DSSCs. performance was achieved for the ethanol based pastes
Thus, some modifications are required in the preparation of containing acid or water. Maximum light to electrical energy
TiO2 pastes and alternative techniques have to be employed conversion efficiencies of 4.9% and 5.0% were obtained for
to improve the bonding between TiO2 particles and the TiO2 the plastic based dye-sensitized solar cells fabricated using
film with the substrate for the fabrication of DSSCs on water and acid-added slurries, respectively. A similar study
plastics with improved efficiencies. was reported by Park et al. [22] to prepare highly viscous
Fabrication of flexible dye sensitized solar cells 177

Figure 2 (A) Comparative images of change of viscosity for an ethanol based TiO2 slurry (a) before and (b) after adding HCl
(0.025 M). (B) Change in viscosity of an ethanol based titania paste (30 wt% solid loading) prepared with different water
percentages. ((a) =0%, (b)= 5%, (c) = 10%, (d)= 20%, (e) =30%, (f) = 40%, (g)= 60%, (h) =80% and (i)= 100%) [20,21].

TiO2 colloidal pastes by adding concentrated ammonia into solar cell. The low efficiency was attributed to the
acidic colloid solutions. increased resistance of the ITO layer due to the hydrother-
Another method to increase the viscosity and the con- mal treatment. The other limitation of the process is that
nectivity of TiO2 particles is to mix of small amounts of it is incompatible with continuous printing for large-scale
Ti-monomers into TiO2 pastes [17,23,24]. Normally, a hydro- manufacturing.
thermal treatment is carried on after the mixing of the A recent result with a high efficiency (5.8%) for plastic DSSCs
Ti-monomers. During the drying process and the hydro- fabricated using a highly viscous TiO2 paste (PECC-K01, Peccell
thermal treatment, Ti-monomers could transfer into crys- Co. Ltd) was reported by Yujiro et al. at Toin University of
talline TiO2, forming firm chemical bonding between TiO2 Yokohama, Japan [25]. This paste was prepared by mixing
nanoparticles in the film. A study found hydrothermal various sizes of TiO2 particles (10–150 nm) with a small amount
treatments at (100 1C) achieved fairly high efficiencies of larger TiO2 particles (250 nm) to enhance light scattering. In
(4.2%) for TiO2 films prepared on FTO coated glass sub- addition, an aqueous colloidal sol of 10–30 nm TiO2 particles
strates [24]. But different results were observed when the was used as an inter-particle connection agent. The effect of
low temperature hydrothermal treatment was used to make particle size on the cell efficiency was also studied. It was found
plastic based DSSCs. Zhang et al. [24] reported the use of that higher dye adsorption and photocurrent were obtained for
three different pastes of P25 TiO2 powder mixed with TiCl4, an average particle size of 60 nm. It was assumed that the
TiOSO4 and Ti-tetraisopropoxide for low temperature TiO2 interparticular connection among TiO2 particles was produced
electrode preparation. Only the Ti-tetraisopropoxide con- by the dehydration of hydrogen bonded network of TiO2
taining paste could be tested on ITO-PET substrates due to nanoparticles, as shown in Equation (1). This process is also
the high acidity of the other two pastes, damaging the ITO referred to as ‘‘chemical sintering’’ [26]. However, it is still
coating. The coated TiO2 films were subjected to a hydro- unclear about the thickness and the crystalline nature of the
thermal treatment in a Teflon-lined autoclave at 100 1C for inter-particle layer formed by such colloidal sols, which would
12 h and the process only achieved a 2.5% efficient flexible be critical for the further improvement of the cell efficiency.
178 H.C. Weerasinghe et al.

(1)

Preparation and processing of TiO2 working Electrophoretic deposition is a material processing technique
electrodes for flexible DSSCs in which charged particles in a suspension are moved toward an
oppositely charged electrode and are deposited onto a sub-
Film preparation strate under externally applied DC electric field (Figure 3(A)).
It has advantages over other coating processes due to less
complexity, controllability of the film thickness and ability
Different low temperature coating and deposition techniques
to deposit films with preferred patterns irrelevant to the
have been attempted and reported for the preparation of TiO2
substrate shape. Power conversion efficiencies exceeding 6%
electrodes on flexible substrates. Generally the preparation of
have recently been reported by Grinis et al. [33] for the
TiO2 electrodes on plastic substrates has been carried out by
electrophoretically deposited mesoporous TiO2 nanoparticles
casting a paste containing TiO2 nanoparticles dispersed in
on plastic substrates. Spray deposition can be considered as
solvents of a low boiling point using doctor blading technique.
another low cost and simple process to directly deposit thin
However, a number of other direct deposition methods, such as
nanoporous films with precise thickness using colloidal solu-
electrophortic deposition [27–33], chemical vapor deposition
tions. This technique can also be integrated into a continuous
(CVD) [34,35] spray deposition [36,37] and pulsed laser deposi-
reel-to-reel deposition system. However some post-treatments
tion (PLD) [38] techniques have been successfully employed for
are required to improve the inter-particle connectivity and the
the preparation of TiO2 electrodes on flexible substrates.
film/substrate adhesion to achieve comparatively high power
conversation efficiencies for the spray deposited electrodes on
flexible substrates. Cha et al. has attempted utilizing a dye
spray technique to prepare pre-sintered TiO2 micro-balls and
achieved efficiencies just over 2% (Figure 3(B)), [37].
For the preparation of TiO2 photoelectrode films, doctor
blading has been widely used for the small scale preparation
of TiO2 films [11,20,21,39]. This method has not yet been
proved as a suitable technique that can be applied in a large-
scale roll-to-roll fabrication. Thus, screen printing, gravure/
reversed gravure and spraying techniques have been con-
sidered as alternate methods for the continuous preparation
of TiO2 electrodes. Fabrication of DSSCs on flexible substrates
totally relies on the development of large scale printability of
mechanically stable thick TiO2 films (5–15 mm) on the sub-
strates. So far only few works has been reported on the large
scale preparation of TiO2 films for flexible DSSCs.

Film processing techniques

The printed TiO2 films require an additional processing step
to improve the necessary inter-particle contact for their
effective performance as an electrode material, as well as
to improve the mechanical stability, namely good film-
substrate adhesion. Organic binders were used for making
flexible DSSCs, but absence of high temperature sintering of
the TiO2 film on polymer substrates resulted in incomplete
necking of the particles due to the presence of residual
organics in the film. UV/ozone treatment of the deposited
TiO2 films was used by several groups as a method for
eliminating adsorbed organic impurities on the surface of
TiO2 films (Figure 4) [40], but it only achieved a modest
improvement in efficiency of the devices [41–43]. Moreover,
the effect of UV radiation on improving the inter-particle
connection was also investigated as another low tempera-
ture sintering method for flexible DSSCs [40,44].
Figure 3 A schematic illustrations of (A) electrophoretic Some alternative sintering techniques have been attempted
deposition and (B) spray-drying techniques [37]. to improve the necking of TiO2 nanoparticles without damaging
Fabrication of flexible dye sensitized solar cells 179

and ceramic industries to make strong green bodies of

complex shapes. In the CIP process, the same pressure is
applied through a liquid medium in all directions uniformly,
independent of the size and the shape of the pressed
material. High pressure can be applied through the CIP to
very large size materials, being bulky pieces or thin films.
Because the pressure is applied through a liquid medium,
the technique has unique advantages for pressing thin films
and non-flat solar cells. As observed in Figure 5, due to the
deaggregation by CIP compression, the CIP-pressed films
became more transparent and showed increased light
transmittance with higher pressure [39]. Such increase in
transparency of the electrodes can also be associated with
the improved film uniformity and reduction of the film
thickness due to the compression. Both normal and lateral
forces measured by the nanoindentation scratch test on the
CIPed TiO2 films showed a significant increase with increas-
ing the CIP pressure. Electrodes processed by the CIP
method have yielded high conversion efficiencies of 6.3%
and 7.4% for simulated sun lights of 100 and 15 mW cm 2
respectively, a marked improvement compared to that
Figure 4 ATR–FTIR spectra of the TiO2 film after different
without the CIP pressing [46]. Although CIP is a batch
times of irradiations (From bottom to top: 0, 1, 2, and 6 h)
process, it is a mature manufacturing technology. Large
indicating that most of the organic substances are eliminated
CIP facilities are commercially available for pressing a very
after the UV treatment [40].
large number of solar cell modules in one batch, thus can be
compatible with commercial scale manufacturing.
the plastic substrate. Microwave heating can be a selective When the low temperature prepared TiO2 nanocrystalline
heating process, depending on materials responses to the film is compressed, the connection between the TiO2
microwave radiation. For example, many polymers have low nanocrystals is not a chemical bonding but merely a physical
loss tangent and thus are transparent to microwave, while contact. Thus, there are still barriers for the electron
many oxides are good microwave absorbers. Uchida et al. transport especially at high electron density. Normally the
heat-treated TiO2 films on ITO/PET substrates in a microwave PCE of DSSCs is lower at high light intensity compared to
oven at 2.45 GHz and 28 GHz frequencies and achieved 0.74% that at low light intensity. In order to improve the PCE of
and 2.2% conversion efficiency respectively [45]. The effi- the plastic DSSC, Dürr et al. developed a new method,
ciency improvement by higher frequency heating was mainly permitting the use of slurries containing organic binders and
resulted from a significant increase in the photocurrent of high temperature sintering for flexible TiO2 photoelectrodes
the DSSC device. However, more work is needed to verify if [50]. This ‘‘lift off’’ technique is based on the formation of a
TiO2 particle interconnection was improved in the microwave sintered titania film on a glass substrate and then transfer-
heating. ring it to a plastic substrate. Efficiency close to 6% was
Mechanical compression is one of the widely used alter- reported for the transferred electrode systems on plastic
native techniques for making DSSCs on plastics and the substrates. Initially nano crystalline TiO2 films were screen
technique can be compatible to a roll-to-roll production line printed on gold layer deposited glass substrates. Inter-
[46–48]. In this technique, an external pressure is applied on particle connection of TiO2 films was achieved by high
the dried TiO2 films deposited on polymer substrates in temperature sintering. The sintered TiO2 film was separated
order to enhance the particle-particle and film-substrate from the glass substrate by chemical dissolution of the gold
connectivity. So far the highest efficiency for the polymer layer and then transferred on to an ITO-PEN substrate as
based DSSC is reported by the Arakawa group in Japan, using shown in Figure 6 [50]. Despite the high efficiency achieved,
a uniaxial compression technique and modified TiO2 pastes this is not a commercially viable technique due to high
[47]. In their work, a compressive pressure was applied on expenses and the complexity of transferring steps required
the doctor bladed TiO2 films using a uniaxial Mini Test Press during the process.
(Toyo Seiki) and the maximum pressure applied on the films The mechanical compression methods, especially the CIP
was limited to 100 MPa. Maximum efficiency and maximum technique, can easily achieve mechanically stable TiO2 films
photocurrent of 8.1% and 14.5 mA/cm2 were reported in on plastic substrates. But the high temperature sintering
this work. However, it remains a challenge to achieve high does show the advantage on the electron transport com-
uniformity in a 10 mm thick TiO2 film by such a uniaxial pared with the compression processes. To combine the two
pressing, especially when the TiO2 film area is large, as it techniques together, a new concept of ‘‘building blocks’’
would require the two pressing plates to be perfectly has recently been developed by Cheng’s group to construct
parallel and to have mirror flat surfaces. On the other high efficiency working electrodes on plastic substrates with
hand, a cold isostatic pressing (CIP) technique has been a novel and facile approach [51]. The building blocks are the
applied for the first time by Cheng’s group to press TiO2 agglomeration units in which nanocrystalline TiO2 particles
films on plastic substrates for flexible DSSCs [39,49]. CIP is a are constructed in a mesoporous structure. By pre-treating
powder compaction technology widely used in metallurgy the build blocks at high temperatures, it can induce strong
180 H.C. Weerasinghe et al.

Figure 5 (A) Appearance of the TiO2 electrodes (a) before and (b) after processed by CIP. (B) Variation of the transmittance of the
TiO2 films on ITO-PEN substrates processed at different CIP pressures. All the films had the same thickness (12 mm) before CIP
compression. [39].

chemical connections between the nanocrystals. These two types of electrodes. The building blocks could be pre-
sintered building blocks can then be used to build the sintered at 500–650 1C, thus there would be comparable
working electrode films on plastic substrates by a compres- inter-particle necking for electron transport within the
sion process. Submicrometer-sized mesoporous TiO2 beads beads as that in the sintered nanoparticles. The main
shown in Figure 7(a) [52] were employed as an example of barrier comes from the relatively poor contact between
building blocks to demonstrate the superiority of using such the beads. The plastic DSSCs made from building blocks
a processing strategy. The beads were synthesized via a showed much higher PCE (6.6%) than that (4.9%) made from
combination of a sol-gel self-assembly process and a sol- powdered nanoparticles (18 nm) [51]. Electrochemical
vothermal process [53,54]. Such mesoporous titania beads Impedance Spectroscopy (EIS) studies clearly showed that
have the following features: (1) high surface areas for the building block electrode had much faster electron
effective adsorption of sensitizers, as a result of the diffusion rate (near 1 order of magnitude higher) than that
presence of abundant and well-interconnected mesopores in the powered nanoparticle electrode. Ke et al. also
within the beads [54]; (2) strong light scattering capability employed the similar TiO2 beads structure to build the
due to the comparable beads size to the wavelength of the electrode, and found fast electron transport in the elec-
visible light [52] and (3) fast electron transportation via the trode [56]. Furthermore, the pre-sintered beads were
densely-packed and chemically-bonded anatase network treated with a 40 mM TiCl4 solution, followed by a high
inside the beads [55] as clearly shown in Figure 7(b), [51]. temperature sintering to form core-shell structured TiO2
The beads building blocks were compared with the pow- beads, which was an essential step to increase the photo-
dered nanoparticles when constructed into working electro- current involved in the glass substrate based DSSCs pre-
des for flexible DSSCs by the CIP compression technique. viously [57]. The treated building blocks were used to
Figure 7(c and d) [51] shows the schematic graphs of the make films on plastic substrates via the CIP process. The
Fabrication of flexible dye sensitized solar cells 181

Figure 6 Individual steps of lift-off and transfer process [50].

Figure 7 (a) SEM [52] and (b) TEM images of titania mesoporous beads, and schematic representation of titania films constructed
from (c) powdered nanoparticles and (d) building blocks [51].

final PCE was greatly improved up to 7.5%, an efficiency of flexible DSSCs, it is necessary to minimize the difference
rarely reported in the literature for plastic DSSCs. The between processing of electrodes on flexible and glass
improvement is attributed to the enhancement of the photo- substrates. The 1D structures provide a possible resolution.
current due to the TiCl4 treatment, from 13.18 mA/cm2 to The 1D structures, such as ZnO nanowires and TiO2 nano-
15.41 mA/cm2 [51]. tubes can be directly grown on a substrate via a relatively
Although the efficiency of flexible DSSCs has been gradu- low temperature processing condition [58,59]. Such 1D
ally improved, it is still lower than that of glass substrate- structures not only hold fast electron diffusion channels,
based DSSCs. In order to further improve the performance but also provide the free space between nanorods/wires/
182 H.C. Weerasinghe et al.

Figure 8 Picture of TiO2 nanotubes grown over 10  10 mm2 areas on Kapton HN substrate and scheme of nanotube array formation [59].

tubes allowing for bending. Vomiero et al. anodized a ZnO nanowires on the optical fiber to prepare a 3D fiber-
titanium layer deposited on a PET/Kapton substrate by based DSSC device, as shown in Figure 10. The device gave a
sputter coating to grow TiO2 nanotubes for flexible devices, higher power output when light illuminated from the core of
as shown in Figure 8 [59]. the optical fiber than from the side. By the 3D design, the
The Kapton substrate used in Figure 8 has a higher heat- light interaction surface area is increased and then the
resistant than PET or PEN polymer substrates, thus would be device performance was greatly improved. The efficiency of
more feasible to produce photoanodes at a higher tempera- the fiber based 3D DSSC device was 120% higher than that of
ture. But the processing temperature would still need to be ZnO nanowire devices made on the flat substrate.
maintained lower than that employed for glass substrate- Various different film processing techniques have been
based devices (usually 500 1C). Alternatively, metal sub- successfully demonstrated by different research groups
strates such as stainless steel [60–62] and Ti foil [63] have around the world. Table 1 is a summary for a number of
attracted a lot of interests as they are flexible and can film deposition and processing techniques used for the
sustain high temperature sintering. However, as the metal preparation of DSSCs on plastic substrates and the corre-
substrate is not transparent, the device required to be sponding power conversion efficiency of the devices.
illuminated from the back side, in which case some photons
would be lost due to the absorption of counter electrode Low temperature fabrication of counter
materials (mainly Pt) and the electrolyte. To resolve this
issue, semi-transparent metal meshes have been employed
electrodes on plastic substrates
as substrates to allow front illumination [64].
Zou et al. has made an interesting wire based flexible As one of the main components in DSSCs, counter electrode
device where a metal wire was coated with TiO2 and plays a very important function in the process of photo-to-
adsorbed dyes as a working electrode, and another wire electricity conversion. Similar to the working electrode, a
was coated with Pt as a counter electrode [65]. The metal number of issues arise when counter electrodes are pre-
wire offers high flexibility and good conductivity. The two pared on plastic substrates. As the high temperature
wires were twisted together and filled with electrolyte to process is not allowed, the deposition of counter electrode
form a device as shown in Figure 9 (a, b). Based on this materials on plastic substrates requires novel low tempera-
novel idea, they developed a fiber-shaped weave-able DSSC, ture methods different to conventional high temperature
in which a thinner counter electrode wire wristed on the methods used in the glass-based counter electrodes. Sputter
working electrode wire, as shown in Figure 9 (c, d) [66] and coating of Pt on flexible substrates is a widely used
a bifacial wired flexible DSSC, in which the counter elec- technique reported in the literature; however this is a
trode wire and working electrode wire were paralleled relatively expensive approach. In order to reduce the cost,
alternately, as shown in Figure 9 (e, f) [67]. counter electrode materials other than Pt, such as carbon
As the metal wire is non-transparent for the light, [71–74], graphene [75,76], carbon nanotubes [77], poly(3,4-
alternatively, optical fibers have been employed as the ethylenedioxythiophene) (PEDOT) [78–80], and CoS [80,81],
substrate [68]. Firstly, a ZnO:Al film of 130 nm was depos- and other deposition techniques have been employed for
ited on the fiber by atomic layer deposition to introduce a making counter electrodes.
conductive coating. Then a mesoporous TiO2 film was
coated and sensitized with dyes. After adding the gelati- Low temperature deposition
nized iodine electrolyte by dip-coating, the carbon-based
counter electrode was coated to complete the device. By In general, counter electrode materials/precursors are
this configuration, the light can travel along the optical prepared in a slurry/solution, and then coated onto plastic
fiber and be scattered to the sensitized electrode to substrates by different coating techniques, followed by a
generate photocurrent. Weintraub et al. [69] have coated low temperature annealing process. H2PtCl6 is a commonly
Fabrication of flexible dye sensitized solar cells 183

Figure 9 (a) Optical photo of a twisted WSF-DSSC (Uncut; radius = 0.2 mm), (b) SEM photo of a WSF-DSSC [65], (c) SEM of the
fiber-shaped solid state photoanode with (right) and without (left) the counter electrode, (d) optical image of the fiber-shaped solar
cell with length of 30 cm and remarkable flexibility [66], (e) cross section and sketch of a flexible solar cell [67] and (f) photograph
of a solar-cell module with an area of 1.50 cm2 (0.50 cm  3.0 cm) [67].

used precursor used to produce a Pt coating on substrates thermal decomposition, Pt counter electrodes can also be
for counter electrodes in DSSCs. To prepare Pt counter made by chemically reducing H2PtCl6. By dip-coating the
electrode on a conductive plastic substrate, H2PtCl6 can be substrate in Sn2 + and H2PtCl6 solutions alternately, a Pt
coated on the substrate followed by thermal decomposition layer has been coated on a conductive plastic substrate
or chemical reduction. For example, in a vacuum environ- [83]. Sun et al. [84] employed ethylene glycol as a solvent
ment, H2PtCl6 can be decomposed to form Pt by heating at and a reduction agent for H2PtCl6 and made a Pt counter
120 1C for 2 h [82]. Following this method, a platinum/ electrode at 180 1C. In another process, H2PtCl6/terpineol
single-wall carbon nanotube (Pt/SWCNT) film has been solutions were screen printed on conductive plastic sub-
sprayed onto the conductive plastic substrate to strates and then reduced by treating it in a NaBH4 solution
form an effective counter electrode [82]. Except for the followed by a hydrothermal treatment at 100 1C [85].
184 H.C. Weerasinghe et al.

Carbon based materials have demonstrated excellent the counter electrode was reported, which would compro-
catalytic properties and thus are attractive less-expensive mise the fill factor of the prepared devices [86].
alternatives to Pt for DSSC applications. They can be
made into counter electrodes on plastic substrates in low
temperature annealing processes. A polypyrrole/f-MWCNT Electrochemical deposition
(multiwalled carbon nanotubes) aqueous slurry has been
drop-casted on an ITO-PEN film, and dried at 60 1C under As a room temperature processing technique, the electro-
vacuum to form a counter electrode [86]. The resulted films chemical deposition method has attracted a lot of interests
have exhibited improved adhesion with the plastic sub- [87–90]. Counter electrodes deposited by the electrochemi-
strate. A flexible DSSC device made from such counter cal method can possess high conductivity due to improved
electrodes showed a PCE of 4.0% (Figure 11 [86]). Although adhesion between the counter electrode film and the
the efficiency is lower than that of Pt counter electrode, substrate. Compared to many other deposition techniques,
this method showed a facile process to prepare counter electrochemical deposition is limited to produce relatively
electrodes and a further increase in PCE was expected by thin films. However, this would not be a major obstacle for
optimizing the thickness of the counter electrode film. DSSC applications, in which the counter electrode is nor-
Carbon-nanofiber was also made as a counter electrode by mally less than 10 mm thick.
a doctor-blading technique, followed by annealing at 100 1C The counter electrodes have also been prepared by
for 5 h in a vacuum oven [71]. Such counter electrodes means of direct current (DC) and pulse current (PC)
showed a quite comparable catalytic property as a Pt electrodepositing methods. Normally, the PC electrode-
counter electrode on glass. However, due to the low positing method was preferred as it showed some advan-
temperature annealing process, higher sheet resistance in tages of controllable particle size, stronger adhesion,
reduced internal stress and uniform film morphology
compared to the DC electrodepositing. By applying
a pulse current in an H 2PtCl 6 solution, uniform Pt
nanoclusters of 40 nm composed of 3 nm particles were
coated on plastic substrates [87]. Similarly, a CoS coun-
ter electrode on a plastic substrate was produced by
placing the substrate in an alkaline (pH = 10) aqueous
solution containing 5 mM cobalt(II) chloride and 150 mM
thiourea, with a potential of 0.8 V vs. Ag/AgCl [81].
The resulted counter electrode was tested for flexible
DSSC devices and showed a PCE of 6.5% in conjunction
with a Z907 sensitizer and a eutectic melt electrolyte.
The performance of the CoS counter electrode is com-
parable to that of the device made from Pt counter
Figure 10 J–V curves of the fiber-based DSSC under one full- electrode [81].
sun illumination (AM 1.5 illumination, 100 mW cm 2) [69]. The Electrophoretic deposition described in the previous
illumination is (1) normal to the fiber axis (NA; 2D case) and (2) sections is another electrochemical deposition method
parallel to the fiber axis (PA; 3D case). successfully utilized for the preparation of counter

Table 1 A brief summary of different film deposition and processing techniques used for preparation of DSSCs on plastic
substrates.

Paste Pre-treatment Deposition technique Post-treatment PCE (%) Ref.

TiO2, water None Doctor blading Mechanical pressing 8.1 [47]

TiO2 beads TiCl4 treatment Doctor blading CIP compression 7.5 [51]
TiO2 P25, ethanol None Doctor blading CIP compression 6.3 [39]
TiO2 P25, ethanol, acetylacetone, None Electrophoretic Compression 6.2 [33]
iodine, acetone, and water additives deposition
TiO2 P25, ethanol Acid, base Doctor blading None 5 [21]
treatment
TiO2 P25, ethanol, water None Doctor blading None 4.9 [20]
TiO2 P25, ethanol Ball milling Doctor blading None 4.2 [11]
TiO2 (F-5, G-2 from Showa Denko) None Electrophoretic UV treatment 4.1 [70]
deposition
TiO2 100 nm and 12 nm particles, None Doctor blading Chemical sintering 3.1 [26]
TiO2 P25, butanol
TiO2 (TAYCA Corp., TKS-201) Hydrothermal Doctor blading Microwave heating 2.2 [45]
TiO2 P25, ethanol, hydroxybenzoic acid Pre-sintering Spray-dried None 2.1 [37]
Fabrication of flexible dye sensitized solar cells 185

to the plastic substrate under a pressure applied by a

heavy weight. Similarly, Lee et al. employed a ‘‘growth-
detachment-transfer’’ process to coat vertically aligned N-
doped carbon nanotube arrays on the conductive plastic
substrate [93].
In order to reduce the cost and to improve the perfor-
mance of the flexible DSSC, researchers have also tried
to replace the expensive ITO coated substrate with other
conductive materials at the counter electrode side [94–97].
Wang et al. employed titanium carbide, tungsten oxide, and
vanadium nitride as catalysts and at the same time to
provide conductivity for electrons [95]. Alternatively, Huang
et al. used aligned carbon nanotube/polymer composite
films that are transparent, flexible, and highly conductive as
the counter electrode [96].

Conductive substrates and blocking layers

Figure 11 Photocurrent density–voltage curves (J–V) of DSSCs
with Ppy/f-MWCNT and Pt films on PEN substrates as the
Conductive substrates
counter electrodes under AM1.5 illumination of 100 mW cm 2,
respectively. The inset shows photos of Pt (left) and Ppy/f- In flexible DSSCs, a TiO2 porous film is coated on a
MWCNT (right) films on PEN substrates [84]. conductive substrate, usually ITO coated PET or PEN. ITO
is a quite expensive material due to large demands and
electrodes on plastic substrates. For example, an efficient relatively low availability. Thus other alternative conductive
counter electrode has been prepared on a plastic ITO-PEN materials have been studied for the replacement of the ITO
substrate, by immersing in a platinum colloid solution with layer. Such alternative materials should be transparent,
an applied d.c field of 1.6 V cm 1 followed by a heat- highly conductive, bendable, and non-reactive with the
treatment at 60 1C for 30 min to remove water and ethanol. electrolyte. These conductive materials must exhibit low
Flexible DSSCs prepared using the counter electrodes absorption and no scattering for the incident light. Further-
deposited by this method have shown higher PCE (5.8%) more, high conductivity is required to reduce the sheet
compared to the counter electrodes prepared by the resistance of the device, which is one of major factors
sputter-coating method (5.3%) [89]. affecting the PCE of the device. In addition, the conductive
Similar to the electrophoretic deposition method, counter layer should demonstrate high endurance on bending defor-
electrodes on plastics substrates can also be fabricated by a mation, which would confine the thickness of the layer.
simple electrospinning technology. In a typical synthesis, an Ito et al. demonstrated a TiO2–Ag–TiO2 film as a replace-
electrospinning solution, for example, a mixture of Polyaniline ment of the ITO layer, where TiO2 acts as a dielectric
emeraldine (PANI), 10-camphorsulfonic acid (CSA) and poly- material while Ag providing the required high conductivity
lactic acid (PLA) in chloroform was loaded into a plastic [98]. Both materials exhibit low absorption in the visible
syringe, and then electrospun at a rate of 1.0 mL h 1 and an light region of the solar spectrum. The sputtered layers
electric field of 25 kV. The conductive plastic substrates were (20 nm TiO2/10 nm Ag/100 nm TiO2) had sheet resistance of
placed onto an aluminum collector to collect films and the 8 O/&, average transmittance of 82% and reflectance (400–
thickness of the film is controlled by the electrospinning time 700 nm) of 6.3%, and showed higher PCE compared to the
[90]. This method can provide a fast and facile fabrication of DSSC devices fabricated on FTO and ITO substrates [98].
low-cost counter electrodes on plastic substrates and is Magnetron sputtered aluminum-doped zinc oxide (AZO)
suitable for potential roll-to-roll production. films have also been investigated as the conductive layer
on plastic substrates. a PCE of 2.1% was achieved in DSSCs
with the AZO substrate. This was an inexpensive alternative
Other deposition methods approach for producing mechanically and chemically stable
conductive films on flexible substrates [99].
A pulsed laser deposition method has been employed to
deposit graphite onto conductive plastic substrates as a low- Blocking layer
cost alternative to platinum for the counter electrode [91].
The properties of the pulsed laser deposited films can be In order to inhibit the recombination between electrons in the
adjusted by tuning of experimental variables such as laser conductive substrate and holes in the electrolyte/hole con-
pulse width, wavelength, ablation time, substrate-target ductor, normally a pinhole free, extremely thin dense layer is
distance, laser fluence, and inert gas pressure [91]. Wan deposited on the top of the conductive layer as a blocking
et al. adopted a transferring technique to coat graphene layer [100,101]. On the glass substrate, this is normally
counter electrode on the conductive plastic substrate [92]. achieved by spray pyrolysis of titanium di-isopropoxide bis-
Graphene dispersion has been vacuum filtrated using a (acetylacetonate) solution [102] or by immersing the sub-
mixed cellulose ester membrane and the resulted mem- strates in chemical bath of TiCl4 aqueous solution [55]. Both
brane with graphene has been transferred and adhered on these processes require high temperature heat treatment
186 H.C. Weerasinghe et al.

(4400 1C) for converting amorphous TiO2 to crystals, and obtained in the commonly used acetonitrile based devices
thus are not suitable for plastic substrates. However several [18]. Lee et al. [107] have investigated the long-term
techniques have been attempted for the preparation of a stability of the flexible DSSC using an MPN based electrolyte
blocking layer at low processing temperatures although they and found that the performance of the devices maintained
are still not fully desirable to be used in flexible DSSCs. Dip- 96.9% after more than 1000 h prolonged aging under one sun
coating method followed by the microwave treatment has light irradiation at 60 1C. Furthermore, solvent-free ionic
been attempted by Hart et al. [103] to prepare a dense TiO2 liquid based electrolytes have also been studied in flexible
layer at low temperatures. A TiO2 barrier layer has also been DSSCs [81,108,109]. Flexible devices made from ionic liquid
prepared by the sputtering and e-beam evaporation methods electrolytes have retained 85% of its initial PCE after 1000 h
[104]. Furthermore, Chen et al. [105] have prepared an of light soaking at 60 1C [81].
ultrathin layer of TiO2 on the conductive plastic substrate by To prevent potential leakage of solvent based electrolytes,
sputtering a TiO2 target, and found that the addition of the quasi-solid-state or solid-state flexible DSSCs have been
compact layer increased the PCE by more than 25%. More developed by introducing polymer gels into the electrolytes
efforts in this area are required to develop an effective or by replacing the liquid electrolyte with a solid-state
blocking layer on plastic substrates in order to adapt newly polymer hole conductor [110–120]. Lee et al. [113] used a
developed electrolytes, such as cobalt based systems and poly(ethylene oxide)-based gel electrolyte for solidifying
other solid-state hole-transport materials efficiently used on flexible DSSCs and incorporated quaternized ammonium
glass-based DSSCs. iodide-containing conjugated polymer electrolytes (CPEs), in
which the ion-supporting side chains supplied channels for ion
transport and the polymer backbones provided pathways for
Stability and up-scaling hole transporting back to the contact electrode. After the
device was operated for 100 h, the PCE dropped merely by
Sealing and stability 15%. By using the gel electrolyte, a new approach could be
developed to laminate flexible DSSCs, in which a gel electro-
Poor stability is one of the major issues that render lyte sheet could be laminated between a TiO2 electrode and
commercialization of DSSCs on flexible substrates. This is the counter-electrode to form a device [110].
mainly caused by the leakage and deterioration of the
electrolyte due to imperfections in the sealing material Modules and up-scaling
and process [106]. Sealing of the electrolyte between the
working and counter electrodes has become a more con- A single flexible DSSC device can only produce 0.6–0.8 V and
cerned issue in plastic substrate-based DSSCs due to the 10 mA/cm2. The amount of power generated by such a single
poor adhesion of sealing materials with the polymer sub- device has limited usage in real applications. Therefore,
strate surface. Thus, when using Surlyn or Bynel materials different types of modules are required to produce sufficient
to seal the two plastic substrates in a device, the films tend power for applications. A module in parallel connection can
to delaminate more easily compared to the glass-based provide a high current for powering small electronics, and that
DSSCs. The delamination creates defects in the sealed edges in series connection can offer a high voltage for charging
allowing the electrolyte to leak. Thus, alternative sealing batteries. By combining the different types of modules to form
materials, such as silicone elastomers and epoxy resins a large area device, it can provide both high voltage and
are currently being examined. However, finding a suitable current. When the device becomes large with high current, it
sealing material that can be applied at low temperature requires a low internal resistance, to minimize the energy lost
and has high resistance to the corrosive components in the through the thermal effect associated with high resistance,
electrolyte still remained as a challenge. which in turn would damage the device. Normally, a metal grid
Except for screening out of new sealing materials, more that has high endurance to the corrosive electrolyte corrosion is
attentions have been focussed on the modification of coated between the working electrode strips to collect and
electrolytes. Ikegami et al. [106] found that electrolyte transport photogenerated current. Miyasaka [121] have made
compositions had a significant effect on the devices’ long- DSSC modules on plastic substrate, which produced 7.2 V as
term stability. LiI, which is a commonly used compound open-circuit photovoltage and 250–300 mA as short-circuit
in electrolyte solutions, was found to react with the ITO photocurrent. His group has also fabricated large modules
coating on plastic substrates. Devices with electrolyte producing a stable DC voltage more than 110 V under indoor
containing LiI showed degradation after outdoor exposure illumination [122].
of one month. TiO2 film found to peel-off partially due to DSSCs on plastic substrates have advantages of light
the dissolution of the ITO on the substrate. However, the weight and thin body compared with glass substrate-based
devices made from a LiI-free electrolyte have shown DSSCs. However, a more significant advantage is the flex-
improved performance over 220 h at 55 1C and 95% humidity ibility, which not only opens more possible applications, but
conditions under dark [102]. On the other hand, solvents also makes the mass production much easier. Attempts have
usually used in electrolyte solutions are highly volatile and been made in producing flexible modules by printing
can easily leak through the imperfections in the sealing. techniques (Figure 12). The roll-to-roll process, commonly
Therefore, 3-methoxypropionitrile (MPN) as a non-volatile employed in the printing industry, can be used to fabricate
solvent with high boiling point and low vapor pressure has the flexible DSSCs in a fast production line. For such a
been used as a solvent for electrolytes used in plastic based process, all layer components consisting of the devices
devices. Flexible devices containing MPN as the solvent have would be printable. This offers a very attractive prospect
achieved PCE above 5%, and are comparable to the PCE for mass production of low cost photovoltaic devices.
Fabrication of flexible dye sensitized solar cells 187

improve the stability and lifetime of flexible DSSCs by pre-

venting electrolyte evaporation and moisture ingress on to the
devices. Another alternative approach to improve the lifetime
is to substitute the volatile liquid electrolyte with ionic liquid
electrolytes or solid-state hole conductors. Although there has
been a little work reported so far on the fully solid-state plastic
DSSC due to the difficulty in fabricating the blocking layer,
different quasi solid-state hole conductor materials and poly-
mer gel electrolytes have shown promising results in the plastic
substrate based DSSCs with much improved stability and
considerably high efficiencies. When the fully solid-state hole
conductors were applied in the flexible device, it is expected
that the stability would be improved further. Some novel
designs of the flexible DSSCs like fiber-shape and 3D structures
would improve the performance of the devices and broaden
the real applications of the flexible devices. Flexible DSSCs
printed on plastic substrates are envisaged to be a promising,
low cost technology for renewable energy in the future.

References

Figure 12 A flexible DSSC module produced by screen print- [1] P. Wang, S.M. Zakeeruddin, P. Comte, R. Charvet,
ing. (Courtesy to the Victoria Organic Solar Cell Consortium R. Humphry-Baker, M. Grätzel, Journal of Physical Chemistry
(VICOSC), Australia). B 107 (2003) 14336–14341.
[2] S. Noda, K. Nagano, E. Inoue, T. Egi, T. Nakashima, N. Imawaka,
M. Kanayama, S. Iwata, K. Toshima, K. Nakada, K. Yoshino,
Synthetic Metals 159 (2009) 2355–2357.
Conclusions and outlook [3] X. Xi, Q. Meng, F. Li, Y. Ding, J. Ji, Z. Shi, G. Li, Solar Energy
Materials and Solar Cells 94 (2010) 623–628.
Plastic substrate-based DSSCs are showing increased poten- [4] M. Tuomikoski, P. Kopola, H. Jin, M. Ylikunnari, J. Hiitola-
tial for commercialization owing to its unique features, such Keinanen, M. Valimaki, M. Aikio, J. Hast, Roll-to-Roll Manufac-
as light weight, flexibility, and the feasibility for roll-to-roll turing of Organic Photovoltaic Modules, in: European Microelec-
speedy mass production. However, further improvements of tronics and Packaging Conference, IEEE (2009) 217–220.
power conversion efficiency and stability of DSSCs on plastic [5] B. Muhsin, J. Renz, K.H. Drue, G. Gobsch, H. Hoppe,
substrates would be required before commercial production Synthetic Metals 159 (2009) 2358–2361.
and applications. The relatively low photovoltaic perfor- [6] A. Yella, H.-W. Lee, H.N. Tsao, C. Yi, A.K. Chandiran,
M.K. Nazeeruddin, E.W.-G. Diau, C.-Y. Yeh, S.M. Zakeeruddin,
mance in DSSCs on plastic substrates is mainly attributed to
M. Grätzel, Science 334 (2011) 629–634.
the low temperature processing limited by the thermal [7] I. Chung, B. Lee, J. He, R.P.H. Chang, M.G. Kanatzidis,
stability of the plastic substrate. Various low temperature Nature 485 (2012) 486–489.
processing methods have been developed for producing [8] A. Zaban, A. Meier, A.J. Nozik, B.A. Gregg, Proceedings of
efficient working electrodes and new materials and methods the Symposium on Electrode Materials and Processes for
have been studied to develop efficient counter electrodes Energy Conversion and Storage IV 97–13 (1997) 306–316.
for plastic DSSCs. With great efforts devoted and more [9] K.D. Benkstein, N. Kopidakis, J. van de Lagemaat, A.J. Frank,
novel methods developed, the PCE of plastic based DSSCs The Journal of Physical Chemistry B 107 (2003) 7759–7767.
has been continuously improved from 4–5% to 6–7% over [10] E.S. Carreras, F. Chabert, D.E. Dunstan, G.V. Franks, Journal
the last decade and lately the highest PCE of 8.1% has been of Colloid and Interface Science 313 (2007) 160–168.
[11] H.C. Weerasinghe, P.M. Sirimanne, G.P. Simon, Y.B. Cheng,
reported. Furthermore, many different novel structures
Journal of Photochemistry and Photobiology a-Chemistry 206
have been designed for the flexible photoanodes. The 1D (2009) 64–70.
structure semiconductors can be directly grown on the [12] C.Y. Jiang, X.W. Sun, K.W. Tan, G.Q. Lo, A.K.K. Kyaw,
flexible substrate to minimize the difference in process- D.L. Kwong, Applied Physics Letters 92 (2008) 143101.
ing techniques between glass substrate-based device and [13] Y. Li, K. Yoo, D.-K. Lee, J.H. Kim, N.-G. Park, K. Kim, M.J. Ko,
flexible substrate-based device, providing a very attractive Current Applied Physics 10 (2010) E171–E175.
approach to achieve high performance flexible devices. [14] A.D. Pasquier, M. Stewart, T. Spitler, M. Coleman, Solar
Promising results published by many research groups have Energy Materials and Solar Cells 93 (2009) 528–535.
shown that the low temperature prepared counter electro- [15] H.-W. Chen, C.-Y. Hsu, J.-G. Chen, K.-M. Lee, C.-C. Wang,
des have competitive performance as the high temperature K.-C. Huang, K.-C. Ho, Journal of Power Sources 195 (2010)
6225–6231.
treated ones. It is highly expected that in the near future
[16] T. Miyasaka, N. Ikeda, T.N. Murakami, K. Teshima, Chemistry
the PCE of plastic based DSSCs would be further improved Letters 36 (2007) 480–487.
to a higher level, comparative to the glass-based DSSCs. [17] D. Zhang, J.A. Downing, F.J. Knorr, J.L. McHale, The Journal
Despite the fact that the poor stability being a major obstacle of Physical Chemistry B 110 (2006) 21890–21898.
for the plastic based DSSCs, robust sealing and encapsulation [18] Y. Kijitori, M. Ikegami, T. Miyasaka, Chemistry Letters 36
techniques are being investigated by many research groups to (2007) 190–191.
188 H.C. Weerasinghe et al.

[19] N.G. Park, K.M. Kim, M.G. Kang, K.S. Ryu, S.H. Chang, [48] H. Santa-Nokki, J. Kallioinen, T. Kololuoma, V. Tuboltsev,
Y.J. Shin, Advanced Materials 17 (2005) 2349–2353. J. Korppi-Tommola, Journal of Photochemistry and Photo-
[20] H.C. Weerasinghe, G.V. Franks, J.D. Plessis, G.P. Simon, Y.- biology A: Chemistry 182 (2006) 187–191.
B. Cheng, Journal of Materials Chemistry 20 (2010) 9954–9961. [49] Y. Peng, J.Z. Liu, K. Wang, Y.-B. Cheng, International Journal
[21] H.C. Weerasinghe, P.M. Sirimanne, G.V. Franks, G.P. Simon, of Photoenergy 2011 (2011) 410352.
Y.B. Cheng, Journal of Photochemistry and Photobiology A: [50] M. Dürr, A. Schmid, M. Obermaier, S. Rosselli, A. Yasuda,
Chemistry 213 (2010) 30–36. G. Nelles, Nature Materials 4 (2005) 607–611.
[22] N.-G. Park, M.G. Kang, K.S. Ryu, S.H. Chang, Y.-J. Shin, [51] F. Huang, D. Chen, Q. Li, R.A. Caruso, Y.-B. Cheng, Applied
Advanced Materials 17 (2005) 2349–2353. Physics Letters 100 (2012) 123102.
[23] D. Zhang, T. Yoshida, H. Minoura, Chemistry Letters (2002) [52] F.Z. Huang, D.H. Chen, X.L. Zhang, R.A. Caruso, Y.B. Cheng,
874–875. Advanced Functional Materials 20 (2010) 1301–1305.
[24] D. Zhang, T. Yoshida, K. Furuta, H. Minoura, Journal of [53] D.H. Chen, L. Cao, F.Z. Huang, P. Imperia, Y.B. Cheng,
Photochemistry and Photobiology A: Chemistry 164 (2004) R.A. Caruso, Journal of the American Chemical Society 132
159–166. (2010) 4438–4444.
[25] T. Miyasaka, M. Ikegami, Y. Kijitori, Journal of The Electro- [54] D.H. Chen, F.Z. Huang, Y.B. Cheng, R.A. Caruso, Advanced
chemical Society 154 (2007) A455–A461. Materials 21 (2009) 2206–2210.
[26] X. Li, H. Lin, J.B. Li, N. Wang, C.F. Lin, L.Z. Zhang, Journal of [55] F. Sauvage, D.H. Chen, P. Comte, F.Z. Huang, L.-P. Heiniger,
Photochemistry and Photobiology A: Chemistry 195 (2008) Y.B. Cheng, R.A. Caruso, M. Grätzel, ACS Nano 4 (2010)
247–253. 4420–4425.
[27] H.-W. Chen, K.-C. Huang, C.-Y. Hsu, C.-Y. Lin, J.-G. Chen, [56] C.R. Ke, J.M. Ting, Journal of Power Sources 208 (2012)
C.-P. Lee, L.-Y. Lin, R. Vittal, K.-C. Ho, Electrochimica Acta 316–321.
56 (2011) 7991–7998. [57] B.C. O’Regan, J.R. Durrant, P.M. Sommeling, N.J. Bakker,
[28] H.-W. Chen, C.-P. Liang, H.-S. Huang, J.-G. Chen, R. Vittal, Journal of Physical Chemistry C 111 (2007) 14001–14010.
C.-Y. Lin, K.C.W. Wu, K.-C. Ho, Chemical Communications 47 [58] M. Law, L.E. Greene, J.C. Johnson, R. Saykally, P.D. Yang,
(2011) 8346–8348. Nature Materials 4 (2005) 455–459.
[29] W.-H. Chiu, K.-M. Lee, W.-F. Hsieh, Journal of Power Sources [59] A. Vomiero, V. Galstyan, A. Braga, I. Concina, M. Brisotto,
196 (2011) 3683–3687. E. Bontempi, G. Sberveglieri, Energy & Environmental
[30] T. Miyasaka, Y. Kijitori, T.N. Murakami, N. Kawashima, Science 4 (2011) 3408–3413.
Proceedings of the Society of Photo-Optical Instrumentation [60] Y. Jun, J. Kim, M.G. Kang, Solar Energy Materials and Solar
Engineers 5215 (2004) 219–225. Cells 91 (2007) 779–784.
[31] T. Miyasaka, Y. Kijitori, T.N. Murakami, M. Kimura, S. Uegusa, [61] J.H. Park, Y. Jun, H.-G. Yun, S.-Y. Lee, M.G. Kang, Journal of
Chemistry Letters (2002) 1250–1251. the Electrochemical Society 155 (2008) F145–F149.
[32] J.H. Yum, S.S. Kim, D.Y. Kim, Y.E. Sung, Journal of Photo- [62] X. Huang, P. Shen, B. Zhao, X. Feng, S. Jiang, H. Chen, H. Li,
chemistry and Photobiology A: Chemistry 173 (2005) 1–6. S. Tan, Solar Energy Materials and Solar Cells 94 (2010) 1005–1010.
[33] L. Grinis, S. Kotlyar, S. Ruhle, J. Grinblat, A. Zaban, [63] M.G. Kang, N.G. Park, K.S. Ryu, S.H. Chang, K.J. Kim,
Advanced Functional Materials 20 (2010) 282–288. Chemistry Letters 34 (2005) 804–805.
[34] T. Miyasaka, T.N. Murakami, Applied Physics Letters 85 (2004) [64] X. Fan, F. Wang, Z. Chu, L. Chen, C. Zhang, D. Zou, Applied
3932–3934. Physics Letters 90 (2007) 073501.
[35] T.N. Murakami, Y. Kijitori, N. Kawashima, T. Miyasaka, Chem- [65] X. Fan, Z. Chu, F. Wang, C. Zhang, L. Chen, Y. Tang, D. Zou,
istry Letters 32 (2003) 1076–1077. Advanced Materials 20 (2008) 592–595.
[36] J. Halme, J. Saarinen, P. Lund, Solar Energy Materials and [66] D. Wang, S. Hou, H. Wu, C. Zhang, Z. Chu, D. Zou, Journal of
Solar Cells 90 (2006) 887–899. Materials Chemistry 21 (2011) 6383–6388.
[37] S.I. Cha, B.K. Koo, K.H. Hwang, S.H. Seo, D.Y. Lee, Journal of [67] Y. Fu, Z. Lv, S. Hou, H. Wu, D. Wang, C. Zhang, D. Zou,
Materials Chemistry 21 (2011) 6300–6304. Advanced Energy Materials 2 (2012) 37–41.
[38] H. Pan, S.H. Ko, N. Misra, C.P. Grigoropoulos, Applied Physics [68] M. Toivola, M. Ferenets, P. Lund, A. Harlin, Thin Solid Films
Letters 94 (2009) 071117. 517 (2009) 2799–2802.
[39] H.C. Weerasinghe, P.M. Sirimanne, G.P. Simon, Y.-B. Cheng, [69] B. Weintraub, Y.G. Wei, Z.L. Wang, Angewandte Chemie-
Progress in Photovoltaics: Research and Applications 20 International Edition 48 (2009) 8981–8985.
(2012) 321–332. [70] T. Miyasaka, Y. Kijitori, Journal of the Electrochemical
[40] D. Gutie rrez-Tauste, I. Zumeta, E. Vigil, M.A. Herna ndez- Society 151 (2004) A1767–A1773.

Fenollosa, X. Domenech, J.A. Ayllón, Journal of Photochem- [71] G. Veerappan, W. Kwon, S.-W. Rhee, Journal of Power
istry and Photobiology A: Chemistry 175 (2005) 165–171. Sources 196 (2011) 10798–10805.
[41] T.N. Murakami, Y. Kijitori, N. Kawashima, T. Miyasaka, Journal [72] J. Chen, K. Li, Y. Luo, X. Guo, D. Li, M. Deng, S. Huang,
of Photochemistry and Photobiology A: Chemistry 164 (2004) Q. Meng, Carbon 47 (2009) 2704–2708.
187–191. [73] B. Zhao, H. Huang, P. Jiang, H. Zhao, X. Huang, P. Shen,
[42] T. Oekermann, D. Zhang, T. Yoshida, H. Minoura, Journal of D. Wu, R. Fu, S. Tan, Journal of Physical Chemistry C 115
Physical Chemistry B 108 (2004) 2227–2235. (2011) 22615–22621.
[43] T. Clark, J.D. Ruiz, H. Fan, C.J. Brinker, B.I. Swanson, [74] J. Chen, B. Li, J. Zheng, J. Zhao, H. Jing, Z. Zhu, Electro-
A.N. Parikh, Chemistry of Materials 12 (2000) 3879–3884. chimica Acta 56 (2011) 4624–4630.
[44] G. Mincuzzi, L. Vesce, A. Reale, A.D. Carlo, T.M. Brown, [75] K.S. Lee, Y. Lee, J.Y. Lee, J.-H. Ahn, J.H. Park, Chem-
Applied Physics Letters 95 (2009) 103312. SusChem 5 (2012) 379–382.
[45] S. Uchida, M. Timiha, H. Takizawa, M. Kawaraya, Journal of [76] J.D. Roy-Mayhew, D.J. Bozym, C. Punckt, I.A. Aksay, ACS
Photochemistry and Photobiology A: Chemistry 164 (2004) 93–96. Nano 4 (2010) 6203–6211.
[46] H. Lindström, A. Holmberg, E. Magnusson, L. Malmqvist, [77] R.A. Sayer, S.L. Hodson, T.S. Fisher, Journal of Solar
A. Hagfeldt, Journal of Photochemistry and Photobiology A: Energy Engineering-Transactions of the Asme 132 (2010) 021007.
Chemistry 145 (2001) 107–112. [78] A. Kanciurzewska, E. Dobruchowska, A. Baranzahi, E. Carlegrm,
[47] T. Yamaguchi, N. Tobe, D. Matsumoto, T. Nagai, H. Arakawa, A. Fahlman, M. Fahlman, M.A. Girtu, Proceedings of the Society
Solar Energy Materials and Solar Cells 94 (2010) 812–816. of Photo-Optical Instrumentation Engineers 6656 (2007) 665611.
Fabrication of flexible dye sensitized solar cells 189

[79] A. Kanciurzewska, E. Dobruchowska, A. Baranzahi, [111] N. Ikeda, T. Miyasaka, Chemistry Letters 36 (2007) 466–467.
E. Carlegrim, M. Fahlman, M.A. Girtu, Journal of Optoelec- [112] Q. Li, J. Wu, Z. Tang, Y. Xiao, M. Huang, J. Lin, Electro-
tronics and Advanced Materials 9 (2007) 1052–1059. chimica Acta 55 (2010) 2777–2781.
[80] P. Sudhagar, S. Nagarajan, Y.-G. Lee, D. Song, T. Son, W. Cho, [113] R.-H. Lee, T.-F. Cheng, J.-W. Chang, J.-H. Ho, Colloid and
M. Heo, K. Lee, J. Won, Y.S. Kang, Acs Applied Materials & Polymer Science 289 (2011) 817–829.
Interfaces 3 (2011) 1838–1843. [114] L.-C. Chen, J.-M. Ting, Y.-L. Lee, M.-H. Hon, Journal of
[81] M. Wang, A.M. Anghel, B. Marsan, N.-L.C. Ha, Materials Chemistry 22 (2012) 5596–5601.
N. Pootrakulchote, S.M. Zakeeruddin, M. Graetzel, Journal [115] M.A. De Paoli, A.F. Nogueira, D.A. Machado, C. Longo,
of the American Chemical Society 131 (2009) 15976–15977. Electrochimica Acta 46 (2001) 4243–4249.
[82] Y. Xiao, J. Wu, G. Yue, J. Lin, M. Huang, Z. Lan, Electro- [116] C. Longo, A.F. Nogueira, M.A. De Paoli, H. Cachet, Journal of
chimica Acta 56 (2011) 8545–8550. Physical Chemistry B 106 (2002) 5925–5930.
[83] X.-L. He, M. Liu, G.-J. Yang, S.-Q. Fan, C.-J. Li, Applied [117] S.A. Haque, E. Palomares, H.M. Upadhyaya, L. Otley,
Surface Science 258 (2011) 1377–1384. R.J. Potter, A.B. Holmes, J.R. Durrant, Chemical Commu-
[84] K. Sun, B. Fan, J. Ouyang, Journal of Physical Chemistry C nications (2003) 3008–3009.
114 (2010) 4237–4244. [118] C. Longo, J. Freitas, M.A. De Paoli, Journal of Photochemistry
[85] L. Chen, W. Tan, J. Zhang, X. Zhou, X. Zhang, Y. Lin, and Photobiology A: Chemistry 159 (2003) 33–39.
Electrochimica Acta 55 (2010) 3721–3726. [119] D. Wei, H.E. Unalan, D. Han, Q. Zhang, L. Niu, G. Amaratunga,
[86] S. Peng, Y. Wu, P. Zhu, V. Thavasi, S.G. Mhaisalkar, T. Ryhanen, Nanotechnology 19 (2008) 424006.
S. Ramakrishna, Journal of Photochemistry and Photobiology [120] K.C. Mandal, A. Smimov, D. Peramunage, R.D. Rauh, Materials
a-Chemistry 223 (2011) 97–102. Research Society Symposium Proceedings 737 (2003) 739–744.
[87] S.S. Kim, Y.C. Nah, Y.Y. Noh, J. Jo, D.Y. Kim, Electrochimica [121] T. Miyasaka, Key Engineering Materials 451 (2011) 1–19.
Acta 51 (2006) 3814–3819. [122] T. Miyasaka, M. Ikegami, Journal of Photopolymer Science and
[88] G. Zhu, L. Pan, T. Lu, X. Liu, T. Lv, T. Xu, Z. Sun, Electro- Technology 23 (2010) 269–277.
chimica Acta 56 (2011) 10288–10291.
[89] X. Yin, Z. Xue, B. Liu, Journal of Power Sources 196 (2011) Dr. Hasitha C. Weerasinghe received his MSc
2422–2426. degree in Physics from the University of
[90] S. Peng, P. Zhu, Y. Wu, S.G. Mhaisalkar, S. Ramakrishna, Rsc South Florida, USA and PhD degree from
Advances 2 (2012) 652–657. the Department of Materials Engineering of
[91] K.P. Acharya, H. Khatri, S. Marsillac, B. Ullrich, P. Anzenbacher, Monash University. Hasitha is currently
M. Zamkov, Applied Physics Letters 97 (2010) 201108. working as a postdoctoral researcher in
[92] L. Wan, S. Wang, X. Wang, B. Dong, Z. Xu, X. Zhang, B. Yang, Professor Andrew Holmes’ laboratory at
S. Peng, J. Wang, C. Xu, Solid State Sciences 13 (2011) 468–475. the Bio21 Institute of University of Mel-
[93] K.S. Lee, W.J. Lee, N.-G. Park, S.O. Kim, J.H. Park, Chemical bourne and he is involved in printing,
Communications 47 (2011) 4264–4266. encapsulation and durability testing of
[94] S. Hou, X. Cai, Y. Fu, Z. Lv, D. Wang, H. Wu, C. Zhang, Z. Chu, Organic Photovoltaic (OPV) and Dye-Sensitized solar cell (DSSC)
D. Zou, Journal of Materials Chemistry 21 (2011) 13776–13779. devices on plastic substrates. His research interests include fabrica-
[95] Y. Wang, M. Wu, X. Lin, Z. Shi, A. Hagfeldt, T. Ma, Journal of tion and characterization of DSC and OPV devices, barrier encapsu-
Materials Chemistry 22 (2012) 4009–4014. lation, new materials on improving stability and conducting life-
[96] S. Huang, L. Li, Z. Yang, L. Zhang, H. Saiyin, T. Chen, time testing devices.
H. Peng, Advanced Materials 23 (2011) 4707–4710.
[97] H. Sun, Y. Luo, Y. Zhang, D. Li, Z. Yu, K. Li, Q. Meng, Journal
of Physical Chemistry C 114 (2010) 11673–11679. Dr. Fuzhi Huang received his Ph.D. in
[98] S. Ito, T. Takeuchi, T. Katayama, M. Sugiyama, M. Matsuda, chemistry (2009) from The University of
T. Kitamura, Y. Wada, S. Yanagida, Chemistry of Materials 15 Melbourne, Australia. He is currently work-
(2003) 2824–2828. ing as a research fellow in the Department
[99] D.-J. Kwak, J.-H. Kim, B.-W. Park, Y.-M. Sung, M.-W. Park, of Materials Engineering, Monash University.
Y.-B. Choo, Current Applied Physics 10 (2010) S282–S285. His research interest is developing new
[100] A. Burke, S. Ito, H. Snaith, U. Bach, J. Kwiatkowski, materials and techniques for high efficiency
M. Grätzel, Nano Letters 8 (2008) 977–981. flexible dye-sensitized solar cells.
[101] P.J. Cameron, L.M. Peter, Journal of Physical Chemistry B 107
(2003) 14394–14400.
[102] L. Kavan, M. Grätzel, Electrochimica Acta 40 (1995) 643–652.
[103] J.N. Hart, D. Menzies, Y.-B. Cheng, G.P. Simon, L. Spiccia,
Journal of Sol–Gel Science and Technology 40 (2006) 45–54. Yi-Bing Cheng is a professor in materials
[104] T.B. Meyer, A.F. Meyer, D. Ginestoux, Proceedings of the Society engineering at Monash University, Australia.
of Photo-Optical Instrumentation Engineers 4465 (2002) 13–20. He completed his undergraduate and Master
[105] P. Chen, Y.-W. Lo, T.-L. Chou, J.-M. Ting, Journal of the studies at Wuhan University of Technology,
Electrochemical Society 158 (2011) H1252–H1257. China, and received a Ph.D. degree from
[106] M. Ikegami, J. Suzuki, K. Teshima, M. Kawaraya, T. Miyasaka, University of Newcastle-upon-Tyne, U.K. in
Solar Energy Materials and Solar Cells 93 (2009) 836–839. 1989. He joined Monash in 1991 after work-
[107] K.-M. Lee, W.-H. Chiu, M.-D. Lu, W.-F. Hsieh, Journal of ing for three years as a Research Associate
Power Sources 196 (2011) 8897–8903. in Newcastle, U.K. He specialises in ceramic
[108] D. Kuang, J. Brillet, P. Chen, M. Takata, S. Uchida, H. Miura, materials and composites. His research
K. Sumioka, S.M. Zakeeruddin, M. Graetzel, ACS Nano 2 interests cover the areas of ceramic processing, microstructure
(2008) 1113–1116. development, phase characterisation, and materials for energy
[109] J. Chen, H. Lin, X. Li, X. Zhao, F. Hao, S. Dong, Electro- related applications. His current interest in dye-sensitized solar
chimica Acta 56 (2011) 6026–6032. cells is to develop nanoporous working and counter electrodes for
[110] A. Du Pasquier, Electrochimica Acta 52 (2007) 7469–7474. monolithic devices and flexible cells on polymer substrates.