Carbon 105 (2016) 33e41
lable at ScienceDirectContents lists avaiCarbon
journal homepage: www.elsevier .com/locate/carbonInkjet-printed graphene electrodes for dye-sensitized solar cells
David Dodoo-Arhin a, b, Richard C.T. Howe a, Guohua Hu a, Yinghe Zhang c,
Pritesh Hiralal d, Abdulhakeem Bello e, Gehan Amaratunga d, Tawfique Hasan a, *
a Cambridge Graphene Centre, University of Cambridge, 9 JJ Thomson Avenue, Cambridge, CB3 0FA, UK
b Department of Material Science and Engineering, University of Ghana, P.O. Box Lg 77, Accra-Legon, Ghana
c Faculty of Science and Engineering, Waseda University, Tokyo, Japan
d Department of Engineering, University of Cambridge, 9 JJ Thomson Avenue, Cambridge, CB3 0FA, UK
e Department of Physics, University of Pretoria, Pretoria, 0028, South Africaa r t i c l e i n f o
Article history:
Received 20 January 2016
Received in revised form
21 March 2016
Accepted 5 April 2016
Available online 6 April 2016* Corresponding author.
E-mail address: th270@cam.ac.uk (T. Hasan).
http://dx.doi.org/10.1016/j.carbon.2016.04.012
0008-6223/© 2016 Elsevier Ltd. All rights reserved.a b s t r a c t
We present a stable inkjet printable graphene ink, formulated in isopropyl alcohol via liquid phase
exfoliation of chemically pristine graphite with a polymer stabilizer. The rheology and low deposition
temperature of the ink allow uniform printing. We use the graphene ink to fabricate counter electrodes
(CE) for natural and ruthenium-based dye-sensitized solar cells (DSSCs). The repeatability of the printing
process for the CEs is demonstrated through an array of inkjet-printed graphene electrodes, with ~5%
standard deviation in the sheet resistance. As photosensitizers, we investigate natural tropical dye ex-
tracts from Pennisetum glaucum, Hibiscus sabdariffa and Caesalpinia pulcherrima. Among the three natural
dyes, we find extracts from C. pulcherrima exhibit the best performance, with ~0.9% conversion efficiency
using a printed graphene CE and a comparable ~1.1% efficiency using a platinum (Pt) CE. When used with
N719 dye, the inkjet-printed graphene CE shows a ~3.0% conversion efficiency, compared to ~4.4% ob-
tained using Pt CEs. Our results show that inkjet printable graphene inks, without any chemical func-
tionalization, offers a flexible and scalable fabrication route, with a material cost of only ~2.7% of the
equivalent solution processed Pt-based electrodes.
© 2016 Elsevier Ltd. All rights reserved.1. Introduction
Graphene has attracted significant interest in a variety of elec-
tronic and optoelectronic applications, including in photovoltaics.
This is because it can serve multiple purposes: as a transparent
conducting electrode [1,2], as a channel for charge transport [3] and
as a counter electrode material [4]. It has been reported that the
edge plane sites of graphene exhibit faster electron-transfer ki-
netics than the basal plane sites [5]. Thus, exfoliated graphene
flakes/platelets may offer good electrocatalytic properties. Coupled
with this, electrical conductivity and chemical stability makes
graphene an attractive counter electrode material.
Graphene can be prepared via various top-down and bottom-up
approaches; the most widely exploited ones being mechanical
cleavage [6], chemical vapour deposition (CVD) [7], and solution
based methods such as chemical exfoliation [8] and ultrasonic-
assisted liquid phase exfoliation (UALPE) [9e12]. Thoughmicromechanically cleaved graphene is ideal for fundamental
studies due to the high quality of the exfoliated material, the low
yield of the process renders it unsuitable for large-scale applica-
tions [6]. In recent years, CVD has been scaled up to produce large-
area, high quality graphene [2,7] accompanied by a significantly
improved understanding of the growth mechanisms and graphe-
neecatalyst interaction [13]. However, the high temperature pro-
duction process and subsequent transfer to target substrate is not
always ideal for photovoltaic devices. In addition, CVD grown
mono- or few-layer graphene has limited exposed edges, a key
requirement for enhanced electrocatalytic properties [5,14]. Solu-
tion processing, meanwhile, allows scalable production of disper-
sions consisting of single- and few-layer graphene flakes under
ambient conditions [9e12]. These can be exploited as inks using
existing functional printing and coating techniques, enabling gra-
phene to be deposited onto substrates such as silicon and glass as
well as flexible materials [15e18]. Inkjet printing is of particular
interest, allowing additive patterning, direct writing without the
use of masks or stencils and low cost [15,17,19].
Amongst the different photovoltaic devices, dye-sensitized solar
34 D. Dodoo-Arhin et al. / Carbon 105 (2016) 33e41cells (DSSCs) offer certain advantages, including low cost materials
[20,21], economic fabrication, good low-light conversion effi-
ciencies (>12%) and many colour design possibilities [22,23]. The
DSSC structure consists of a ~10 mm thick mesoporous network of
nanocrystalline semiconductor oxide (e.g. titanium dioxide, TiO2)
deposited onto a conducting electrode or transparent conducting
electrode (e.g. FTO, ITO etc), and an electrolyte. The porous network
of nanocrystals (10e30 nm) provides the large surface area
necessary for adsorption of a thin layer of dye molecules, typically a
ruthenium(II) bipyridyl dye [24] or an organic dye [21,25,26], to
allow for optimum light harvesting. Absorption of light by a dye
molecule creates an excited molecular electronic state. The dye
rapidly returns to its original oxidation state via electron transfer
from iodide ions in the I
/I
3 redox electrolyte. The I
3 ions formed
by oxidation of I
 diffuse a short distance (<50 mm) through the
electrolyte to the cathode, which is coated with a thin layer of
platinum catalyst, where the regenerative cycle is completed by
electron transfer to reduce I
3 to I
 [27].
Two of the most important factors which influence the perfor-
mance of a DSSC are the dye used as a sensitiser, and the counter-
electrode material. The key properties of the dye are the absorption
across the solar spectrum and the adsorption/adhesion of the dye
molecules to the surface of the semiconductor oxide [28]. The most
widely used sensitizers are those based on heavy transition metal
co-ordination compounds (e.g. ruthenium (Ru) polypyridyl com-
plexes) [23,29] due to their efficientmetal-to ligand charge transfer,
intense charge-transfer absorption across the visible range and
long excited lifetime [23]. However, Ru-polypyridyl based com-
plexes are expensive (~1200 $/g) [30], and contain a heavy rare
earth metal, which is environmentally undesirable [31]. Thus,
investigation into environmentally-friendly, economic and readily
available dyes, including those extracted from plants, remains a
strong interest [21,25,26,32,33].
Dyes produced from plant extracts are advantageous due to
their wide availability, simple extraction process, usability without
further purification, environmental sustainability, and low cost
[21,25,26,32,33]. Anthocyanin molecules (water-soluble vacuolar
pigments [34]) are found in tissues of higher plants (i.e. land plants
that have lignified tissues or xylems for transporting water and
minerals throughout the plant), and are responsible for their
redeblue range of pigments, depending on pH [21,25,26,32e36].
Anthocyanin molecules contain carbonyl and hydroxyl groups,
which can adsorb on to the surface of porous TiO2 films, leading to
photoelectron transfer from the anthocyanin molecule to the con-
duction band of TiO2 [21].
The counter-electrode, meanwhile, should be catalytically active
and electrically conducting. It should also exhibit a low over-
potential for rapid reduction of the redox couple to carry the
generated photocurrent across the width of the solar cell. Platinum
(Pt) is the most commonly used counter electrode material in
DSSCs [37]. However, while the required platinum loading for op-
timumperformance of the solar cell is small (~3.2 g/m2) [38,39], the
dissolution of the platinum film in the corrosive I
3 /I
 electrolyte
and high-temperature heat treatment (~300e400 C) [37,40]
required for good Pt-substrate adhesion limits their use on flex-
ible substrates and in low-cost applications. This has necessitated
the search for chemically stable and cost-effective counter-elec-
trode materials for DSSCs such as various carbon nanomaterials
[14,41e46] conductive organic polymers [47] and inorganic semi-
conductors [48]. In this context, graphene has emerged as a
promising low-cost electrode material candidate, which provides
both the conductive pathways and catalytic properties
[41e43,49e54].
The performance of DSSCs with a graphene counter electrode is
dependent on the structure of graphene. Although CVD grapheneproduces continuous layers with high electrical conductivity and
comparable charge transfer resistance (Rct) to platinum [54], it has
a very limited number of active sites for I
/I
3 electrocatalysis [55].
On the other hand, single and few-layer graphene nano-flakes
obtained via solution processing have high density of active edge
sites, offering high catalytic activity, and modest electrical con-
ductivity [15,17,18] at a fraction of the cost. In addition, solution
processed graphene can be used in combination with a variety of
deposition technologies (such as inkjet [15e18,56], screen printing
[42], spray coating [17,44] or flexographic printing [43]) forming a
versatile platform not only for DSSCs but also for any devices
requiring such electrode materials for large scale fabrication. We
note that there have been reports in the literature of graphene inks
being used for DSSCs [14,41e43]. However, these typically use
functionalised forms of graphene such as graphene oxide [41e43].
While the presence of functional groups on graphene sheets can aid
their exfoliation and dispersion into solvents, the electrical prop-
erties of the material are compromised, and cannot be fully
restored to those of pristine graphene [57].
Here, we present formulation of an inkjet printable pristine
graphene ink for the fabrication of graphene-based CEs as a Pt
alternative. Through the use of a polymer stabiliser, we tune the
rheology of the ink, allowing stable and repeatable CE printing. We
also investigate three natural tropical dye extracts as photosensi-
tizers as a low-cost alternative to standard Ru-based dyes. We
demonstrate that with only ~2.7% of the materials cost of platinum
our printed graphene counter electrode, without any structural or
chemical optimization, shows comparable performance to Pt CEs
with both natural and Ru-based dyes as sensitizers.
2. Preparation and characterisation of natural dye sensitizer
extracts
2.1. Extraction of natural dyes
Clean, fresh specimens of Hibiscus sabdariffa, Caesalpinia pul-
cherrima and Pennisetum glaucum are oven dried at 60 C and
crushed into fractionlets. To extract the natural dyes, 5 g of each of
these samples are put in 60 mL of ethanol for 5 days at room
temperature without exposure to light. The dye solutions are then
filtered to remove the solid residues. The concentrations of the dyes
in the final solutions are ~0.4 g/L.
2.2. Characterisation of natural dyes
Anthocyanins and their derivatives, which belong to a group of
natural phenolic compounds, show a broad absorption band in the
visible spectral range [21,25]. This absorption band is due to charge
transfer transitions from the highest occupied molecular orbital
(HOMO) to the lowest unoccupied molecular orbital (LUMO) [58].
As such, the response of these dyes to light absorption is solvent
and pH sensitive [34]. Indeed, anthocyanin molecules tend to show
a red flavylium form in acidic medium and purple deprotonated
quinonodial form as the solution pH increases [21,34,59,60].
Fig. 1 shows representative vis-NIR absorption spectra of the
natural dyes in ethanol. The absorbance peaks (C. pulcherrima:
~422, ~446, ~474 nm, P. glaucum: ~492 nm, H. sabdariffa: ~548,
666 nm) are attributed to the presence of anthocyanin, carotene,
flavone and xanthophyll molecules [61,62]. We also carry out
Fourier transform infrared spectroscopy (FTIR) to confirm the
presence of various functional groups in the dye extracts, see
Supplementary Information for details. For comparison, the inor-
ganic ruthenizer 535-bisTBA (N719) dye, which is an outstanding
solar light absorber and charge-transfer sensitizer, shows a broad
absorption peak at ~518 nm, attributed to its metal to ligand charge
D. Dodoo-Arhin et al. / Carbon 105 (2016) 33e41 35
Table 1
Electrochemical band levels for the organic dyes.
Dyes Eox (V) Ered (V) Ip (HOMO) (eV) Ea (LUMO) (eV)
C. pulcherrima 0.2706 
0.375 
4.671 
4.025
H. sabdariffa 0.324 
0.248 
4.724 
4.152
P. glaucum 0.302 
0.279 
4.702 
4.121
Fig. 1. Optical absorption spectra of the photosensitizers normalized to the highest
absorbance peak. (A colour version of this figure can be viewed online.)transfer transition [23].
The dyes are then characterized by Cyclic Voltammetry (CV) to
investigate corresponding HOMO and LUMO energy levels and
their oxidation (removal of electron from HOMO energy level) and
reduction (addition of electron to LUMO energy level) potentials.
The onset potentials of oxidation and reduction of the dyes can be
correlated to their electron affinity and ionization potential ac-
cording to the empirical relationship proposed in Refs. [63,64] on
the basis of comparisons between theoretical and experimental
electrochemical measurements. CV measurements of the natural
dyes are taken using a three-electrode electrochemical cell at
100 mV
1 constant scan rate under nitrogen atmosphere. A 1.6 mm
diameter glassy carbon electrode is used as the working electrode,
and a platinum wire as the auxiliary electrode. All potentials are
measured and quoted against an Ag/AgCl (3 M KCl) electrode in a
double bridge filled with 3 M KCl solution (upper part) and sup-
porting electrolyte solution (lower part). CV measurements show
distinct oxidation and reduction processes for all the three dyes as
shown in Fig. 2. The curves on the forward and reverse scans
indicate that they are diffusion controlled, and hence the electrode
process is rapid and reversible [63,64]. Thus, the data from the
forward and reverse processes can be used for the calculation of the
thermodynamic parameters [63,64]. The estimated HOMO and
LUMO energy levels of the dyes, determined from the oxidation and
reduction peaks from the CV measurements are presented in
Table 1. For comparison, the Ruthenizer 535-bisTBA (N719) dye,
with typical absorption peak (lmax) of ~535 nm [ε/104 M
1cm
1],Fig. 2. Cyclic voltammograms of the natural dyes. (A colour version of this figure can
be viewed online.)has HOMO and LUMO levels of 
5.34 eV and
3.43 eV, respectively
[30].3. Preparation and characterisation of graphene inks
3.1. Graphene ink preparation
The graphene ink is formulated from graphene prepared via
ultrasonic-assisted liquid phase exfoliation (UALPE). In this process,
bulk graphite flakes are mixed into a solvent, and the mixture is
ultrasonicated. The ultrasound causes high-frequency pressure
variations in the solvent, forming localised microcavities. These
cavities are inherently unstable and rapidly collapse, producing
high shear forces sufficient to overcome theweak interlayer van der
Waals forces and exfoliate graphene flakes from the bulk crystal
[9,10]. The graphene produced by this route is chemically pristine,
since the UALPE process involves no chemical pre- or post-
treatment. The exfoliation process, as well as the stability of the
resultant dispersion, is dependent on the intermolecular in-
teractions between the solvent and the graphene flakes. On a basic
level, the exfoliation process can be explained in terms of surface
energy [11,12], although a more refined understanding can be
gained from the Hansen solubility parameters (HSPs) [65,66]. It was
determined that the best suited solvents have dispersive (dD), polar
(dP) and hydrogen bonding (dH) parameters that match those
experimentally derived for graphene (d ½D,Graphene ¼ 18 MPa ,
dp,Graphene ¼ 9.3 MPa½, d ½H,Graphene ¼ 7.7 MPa ) [65]. However, the
best suited solvents (e.g. N-methylpyrollidone e NMP, cyclohexa-
none) typically present processing challenges due to their high
boiling points (NMP ~ 202 C, cyclohexanone~ 155 C), as well as
poor wettability due to their high surface tension (NMP~
44.6 mNm
1, cyclohexanone~ 34.4 mNm
1) [67].
For ease of processing, it would therefore be advantageous to
develop graphene inks from low boiling-point solvents. Alcohols, in
particular, are attractive as they additionally offer low surface
tension (~20e25 mN m
1), allowing good wetting of the substrate
[56,68]. Indeed, they are commonly used as the primary or sec-
ondary solvent in the majority of graphics and functional inks [69].
Among the common alcohols, isopropyl alcohol (IPA) can exfoliate
graphene with meta-stable dispersion [70] due to the mismatch in
HSPs (dD ¼ 15.8 MPa1/2, dP ¼ 6.1 MPa1/2, d ¼ 15.4 MPa1/2H ) [65]
compared to graphene. The poor long-term stability of a pure IPA
based graphene dispersion means that it cannot be used for inkjet
printing, as it will aggregate and block the printing nozzles. UALPE
with ionic surfactants and nonionic polymers are commonly used
strategies to stabilize graphene [9e11,56,71,72]. In addition to sta-
bilization, the addition of polymers offers modification of viscosity,
allowing optimisation of the ink formulation for inkjet printing
[19]. We use polyvinylpyrrolidone (PVP), a polymer analogy to NMP
(PVP has N-substituted pyrrolidone rings similar to those of NMP)
which has been used before to stabilize carbon nanostructures,
such as nanotubes and graphene in different solvents [56,72,73].
To prepare the ink, 100 mg of graphite flakes (SigmaeAldrich,
100 mesh) with 1.5 mg PVP (Sigma Aldrich, average molecular
weight 10 kDa) is ultrasonicated in 10 mL IPA for 12 h at ~15 C
using a low-power bath sonicator. This starting PVP concentration
36 D. Dodoo-Arhin et al. / Carbon 105 (2016) 33e41enables stabilization in addition to the desired ink rheological
properties specific to our inkjet printing nozzle [17,19,56].
Following sonication, the dispersion is centrifuged at ~4000 rpm
(~1500 g) for 1 h to remove the unexfoliated flakes. Finally, the ink
is filtered via vacuum filtration (glass fiber filter, 1 mm pore),
removing any large particles that may otherwise clog the inkjet
nozzles during the printing process [19]. The final ink (a photo-
graph of which is shown as an inset to Fig. 3) is stable, and forms no
visible aggregates even over several months.3.2. Graphene ink characterisation
The concentration of dispersed graphene may be estimated by
considering the optical absorption of the ink, and using the Beer-
eLambert law:
Al ¼ allc (1)
where c is the graphene concentration (gL
1), l is the distance the
light passes through the dispersion (m), and Al and al are the
absorbance (a.u.) and material dependent optical absorption coef-
ficient (Lg
1m
1) at wavelength l (nm), respectively. Fig. 3 shows
the optical absorbance spectrum of the dispersion diluted to 10 vol
% to reduce scattering losses during measurement, and with the
background absorbance from the solvent and PVP subtracted [74].
The spectrum is mostly featureless as expected, due to the linear
dispersion of Dirac electrons [75]. The peak in the UV region is a
signature of the van Hove singularity in the graphene density of
states [75]. Using a e2460 Lg
1m
1660nm [11], we estimate the
concentration of graphene in the undiluted dispersion as 0.42 gL
1.
The dimensions of the exfoliated flakes are measured via atomic
forcemicroscopy (AFM). The as prepared ink is diluted to 5 vol% and
drop-casted onto a Si/SiO2 wafer. The sample is then annealed at
400 C for 30 min, sufficient to decompose any residual PVP [76]
without affecting the graphene flakes [18,77]. This gives isolated
and polymer free flakes on the substrate, allowing their accurate
measurement. This is achieved using a Bruker Dimension Icon AFM
in ScanAsyst™ mode. The distributions of measured flake thick-
nesses and lateral dimensions are shown in Fig. 4, along with the
images of typical flakes. The average flake thickness is (5.9 ± 0.2)
nm, with ~56% of flakes 4 nm thick (equivalent to <10 layers,
assuming ~0.7 nm measured thickness for a monolayer flake and
~0.35 nm increase for each subsequent layer) [78]. The average
lateral dimension is (196 ± 6) nm. This gives a relatively high edge
to surface area ratio of ~1:17. The disordered sites at the edges of
flakes act as active catalytic sites for the I
/I
3 redox couple reactionFig. 3. Optical absorption spectrum of the graphene in dispersion diluted to 10 vol% to
avoid scattering effects. Inset e photograph of a cuvette containing the as-prepared
graphene ink. (A colour version of this figure can be viewed online.)to carry photocurrent across the width of the DSSC, making the as-
produced graphene ink a potential counter electrode material for
DSSCs.
When formulating an ink for inkjet printing, various parameters
of the ink govern the successful formation of droplets. First, the
particle size (i.e. graphene flake dimensions) in the ink must be
considered, as large particles in the ink can clog nozzles, or disrupt
the ink flow [19]. The nozzle size of the DMP-2831 used in this work
is 22 mm, significantly larger than the average flake dimension
measured via AFM, confirming that nozzle clogging should not
present any issues. Stable drop generation (single droplet genera-
tion for each electrical impulse, without the formation of satellite
droplets) and jetting of ink (avoiding deviation of droplet trajec-
tory) is of primary importance for high quality inkjet printing.
Unstable jetting may lead to uncontrolled amount of ink deposition
on to undesired locations. In inkjet printing, a figure of merit, Z, is
commonly used to consider the printability of inks and is defined
as: Z ¼ (gra)½/mwhere, g is surface tension of the ink (mNm
1), r is
the density of ink (gcm
3), m is the viscosity of the ink (mPas) and a
is the nozzle diameter (mm). As a rule of thumb, it is commonly
accepted that 1 < Z < 14 is required for stable drop-on-demand
inkjet printing [19]. Inks with Z < 1 are too viscous for droplet
ejection, while those with Z > 14 will produce satellite droplets in
addition to the primary droplet, reducing printing reproducibility
[19]. We stress that the range of Z values should be considered as a
guide only. Inks with Z values outside this range may also be
printable, in particular, by controlling the shape and amplitude of
the electrical pulses for droplet generation.
To determine the Z value of the graphene-PVP in IPA ink, we
measure g via a pendant droplet method, and m using a parallel
plate rheometer. Both the measurements are conducted at room
temperature. We measure g ~28.0 mNm
1 and m ~2.3 mPas. The
measured density of the ink is r ~0.8 gcm
3. With a ¼ 22 mm, we
calculate Z ~9.6, falling well into the recommended range for stable
jetting. This is confirmed by experimental results, where we
observe consistent single drop ejection for our ink; Fig. 5a. From the
position of the droplet at different times after jetting, we calculate
an average jetting speed of (3.44 ± 0.03) ms
1.
We next confirm the uniformity of our print process by pre-
paring an array of electrodes on a glass substrate. 9 squares of
6 mm  6 mm are printed at a platen temperature of 60 C, as
shown in Fig. 5b. The printed squares are highly uniform, showing
no coffee-ring effect. After printing, the samples are annealed at
400 C for 30 min to remove residual PVP. The volume of ink
printed is sufficient to achieve a conductive network of graphene
flakes through the printed film (i.e. above the percolation
threshold) [56]. The sheet resistance of the electrodes is measured
for their consistency, as shown in Fig. 5c. The printed electrodes
show good electrical uniformity, with <5% standard deviation in
sheet resistance between them.
4. DSSC fabrication and photovoltaic characteristics
4.1. DSSC fabrication
Fabrication of a DSSC involves several steps. First, two FTO
(fluorine-doped SnO2) conductive glass substrates (~2 cm  2 cm,
8 U/,, Solaronix) are cleaned in a detergent solution using an ul-
trasonic bath for 10 min, rinsed with deionised water and ethanol,
followed by oven drying at 90 C. One of the FTO substrates is then
immersed in a 40 mM aqueous solution of TiCl4 (Sigma Aldrich) at
~70 C for 30 min (for good mechanical adhesion of the TiO2) and
washed with deionised water and ethanol before drying.
The photoelectrode is prepared by depositing a ~9 mm thick TiO2
film of TiO2 paste (Solaronix) on to the TiCl4 treated FTO conductive
D. Dodoo-Arhin et al. / Carbon 105 (2016) 33e41 37
Fig. 4. AFM characterisation of the dispersed graphene flakes. (a) image and (b) section of a typical graphene flake. Distributions of (c) flake thickness and (d) flake lateral di-
mensions. (A colour version of this figure can be viewed online.)
Fig. 5. Inkjet printing of graphene ink. (a) Jetting process of the ink, showing stable single droplet ejection. (b, c) Printed graphene squares for printing consistency characterisation,
showing (b) photograph and (c) normalised sheet resistance for the electrodes. (A colour version of this figure can be viewed online.)glass substrate by doctor-blading. The active area is
~0.25e0.36 cm2. The electrode is then gradually preheated over a
period of ~60 min, and then sintered at 500 C for 30 min in air.
After cooling to ~80 C, the TiO2 electrode is immersed in the dye-
ethanol solution for 12e24 h at 50 C under gentle magnetic stir-
ring. Next, excess dye is washed away with anhydrous ethanol.
A standard, platinum based counter electrode is prepared by
drop casting ~60 ml of 10 mM H2PtCl6eIsopropyl alcohol (SigmaAldrich) solution onto the other FTO-glass substrate with pre-
drilled holes at room temperature and then heating gradually to
~450 C for 30 min. The graphene CE is prepared by inkjet printing
the graphene ink on the pre-drilled FTO glass substrate at 60 C
platen temperature. The active graphene deposited area is
2.0 cm  1.5 cm, leaving the edge for electrical contact. The sub-
strate is gradually heated above 150 C for ~30 min to improve
graphene adhesion, and to remove the polymer stabilizer. The
38 D. Dodoo-Arhin et al. / Carbon 105 (2016) 33e41
Fig. 6. CurrenteVoltage characteristic curves for (a) graphene ink CE based cells, (b)
platinum CE based cells sensitized with (i) P. glaucum (square), (ii) H. sabdariffa (circle),
(iii) C. pulcherrima (triangle), and (iv) Ruthenizer N719 (diamond) and (c) Comparison
of the efficiencies for platinum and graphene ink based cells. (A colour version of this
figure can be viewed online.)relative costs of these two alternative electrodes (as calculated by
considering the relative costs of the raw materials e see Supple-
mentary Information for calculation) are - £14/m2 (solution pro-
cessed graphene) and £520/m2 (solution processed platinum),
meaning that the printed graphene electrode represents ~97.3%
reduction in materials cost.
The dye-adsorbed TiO2 electrode and the platinum or graphene
counter electrode are assembled into a sandwich-type cell with a
60 mm Surlyn thermoplastic film (Solaronix) to separate the pho-
toanode and the counter electrode. An Iodolyte Z-50 (Solaronix)
electrolyte solution (I
/I
3 ) is injected into the cell via the back-
drilled hole in the counter electrode. The hole is then sealed us-
ing a Surlyn film to complete the device assembly.
4.2. DSSC photoelectrochemical characterization
An Oriel solar simulator with an AM 1.5 G filter is used as the for
white-light source (100 mW/cm2) for currentevoltage character-
ization of the DSSCs. The light intensity is measured using a digital
light meter. The short-circuit current density (Jsc) is calculated by
dividing the measured photocurrent by the actual active area
(~0.36 cm2) of the solar cell.
The photoelectrical parameters of the graphene ink and plat-
inum counter electrode-based DSSCs sensitized by the natural and
ruthenium dyes, under AM 1.5 (100 mW/cm2) simulated solar
illumination are shown in Fig. 6 and Table 2. The results are aver-
aged from 3 devices, with <5% standard deviation in values be-
tween them. The efficiency (h) and fill factor (FF) are calculated as
reported elsewhere [21]. Preliminary tests on the cells show that
over a period of 24 h of measurement, the devices remain stable
with <10% change in the conversion efficiency.
Cells fabricated with graphene inks and sensitized with natural
dyes reached an efficiency (h %) of 0.9, >80% of that of platinum
based natural dye cells (h ~1.1%). Ruthenium based cells with gra-
phene ink electrodes reached (h ~3.0%) ~70% that of platinum based
cells (h ~4.4%).
The open circuit voltage of the graphene natural dye cells yield
Voc ¼ 580 mV, representing ~83% of the platinum based cells. For
the ruthenium dye based cells, Voc ¼ 640 mVwas obtained, ~98% of
that for the platinum based cells (650 mV). A positive shift in the
iodide/tri-iodide redox energy level will increase the open circuit
voltage (Voc) and vice versa, thereby influencing the energy con-
version efficiency of the cell [46,79,80].
From the JV curves in Fig. 6, it can be deduced that the graphene
ink electrode cells sensitized with ruthenium dyes, exhibit a lower
shunt resistance (RSH) and a higher series resistance (RSE) than the
equivalent Pt-based cells. RSH and RSE can be calculated as [81]:
dV
RSH ¼ (2)dJ V¼0
dV
RSE ¼ (3)dJ J¼0
This shows that RSH;Gr ~43.3U, and RSE;Gr ~14.6 kU, while RSH;Pt
~26.1U, and RSE;Pt ~23.9 kU. Although this effect negligibly in-
fluences the Voc, it influences the Jsc. The lower RSH for the graphene
CEs could be attributed to factors such as current leakage paths on
the graphene based electrodes, while the higher RSE could arise
from the lower conductivity (i.e. higher sheet resistance) of the
graphene electrode (600 U/,) compared to that of the Pt based
electrode (15 U/,), or to slower electro-kinetics in graphene than
in Pt [82,83].
From the obtained results, it can also be observed that the
natural-dye-based DSSCs produce low efficiencies compared to theinorganic dye (N719) based cells. For a dye molecule to be an
excellent photosensitizer in a DSSC, it must possess several
carbonyl (C]O) or hydroxyl (eOH) groups capable of chelating to
the Ti(IV) sites on the porous nanostructured TiO2 surface. The
chemical adsorption of these dye molecules to the TiO2 nano-
particles in DSSCs occurs generally due to the condensation of the
alcoholic-bound protons with the hydroxyl groups on the surface of
the nanostructured TiO2 as well as the contribution from the
chelating effect of the two nearest hydroxyl group towards Ti(IV)
D. Dodoo-Arhin et al. / Carbon 105 (2016) 33e41 39
Table 2
Photoelectrical parameters of DSSCs fabricated with graphene and platinum counter electrodes.
Photosensitizer dye Active area (cm2) Voc Jsc FF h n-ideal factor
(mV) % of Pt (mA/cm2) % of Pt (%) % of Pt (%) % of Pt
Graphene/FTO counter electrode
P. glaucum ~0.36 520 92.85 1.38 87.34 69.5 94.47 0.50 83.33 1.99
H. sabdariffa ~0.36 500 92.59 1.16 52.25 62.6 82.83 0.4 44.44 2.68
C. pulcherrima ~0.36 580 82.85 2.15 96.85 70.9 103.91 0.9 81.81 2.07
Ruthenizer N719 ~0.36 640 98.46 7.44 76.46 62.0 88.77 3.0 68.18 3.52
Platinum counter electrode
P. glaucum ~0.36 560 1.58 73.57 0.6 1.721
H. sabdariffa ~0.36 540 2.22 75.57 0.9 1.48
C.pulcherrima ~0.36 700 2.28 68.23 1.1 2.86
Ruthenizer N719 ~0.36 650 9.73 69.84 4.4 2.45sites on the semiconductor nanocrystalline layer [28,60,84,85]. It is
also very important that the dye molecules do not aggregate (pep
stacking) on the photoelectrode (TiO2) surface to avoid non-
radiative decay of the excited state to the ground state, which often
occurs with thicker films [86]. However, natural or organic dyes
with good intermolecular pep* interactions help to achieve a
closer packing of the dye molecules on to the TiO2 surface. This
better protects the surface from contact with the electrolyte,
thereby resulting in a high open-circuit voltage and hence high
efficiencies [87]. The C. pulcherrima dye comprises carotenoids
(Carotene and Xanthophyll) and anthocyanins while those from H.
sabdariffa and P. glaucum have anthocyanins as their major com-
ponents. The relatively high efficiency of C. pulcherrima dye based
cells (Gr and Pt CEs) could be attributed to a better intermolecular
pep* interactions in the dye arising from these carotenoids. We
note that further investigation is required to fully understand the
interactions and how they may be improved. In particular, it is
likely that the efficiency could be improved further by the addition
of different stabilizing functional groups or binders to the dye
molecules. This could be achieved through acid treatments
(CH3COOH, HNO3 or HCl) [88], or the addition of sugar molecules
[89] or a co-adsorbent (e.g. chenodeoxycholic acid) [90].
Changes to the electrolyte and to the exfoliated graphene may
also aid the DSSC efficiency. Cobalt polypyridine complex electro-
lytes in place of the iodide/triiodide may improve efficiencies
especially in graphene based cells [20,41]. Meanwhile, since the
catalytic activity of the graphene is key to the device performance,
increasing the density of the catalytic sites (e.g. by chemical func-
tionalisation of the graphene) may enhance the efficiency. How-
ever, a balance must be maintained to avoid excessively lowering
the conductivity of the CE.5. Conclusions
We have demonstrated low-cost, environmentally friendly al-
ternatives for two key elements of DSSCs. Plant-extract dyes have
been used as photosensitisers, in place of ruthenium-based dyes,
where their simple extraction procedure, wide availability, and
environmentally friendly nature make them promising alternative
sources of sensitizers for DSSCs. Meanwhile, inkjet-printed gra-
phene has been used as a counter-electrode in the place of plat-
inum. A polymer-stabilised isopropanol-based graphene ink has
been developed via UALPE of chemically pristine graphite. The ink
is formulated for inkjet printing, where it shows consistent single-
droplet jetting. This, combined with the low boiling point and low
surface tension of the ink allows repeatable printing onto sub-
strates including FTO/glass used for DSSC fabrication. The inkjet-
printed graphene CEs exhibit promising electrocatalytic activity
toward I /I
3 redox couple comparable to the Pt-based CEs. Furtherstudies are now underway into the kinetics of the dye molecules,
device structure and life-time of the fabricated solar cells. Overall,
the low materials cost for the graphene electrodes (~2.7% of the
equivalent solution processed Pt cost) makes them highly attractive
for the development of low-cost DSSCs. These inkjet-printed gra-
phene electrodes may additionally have wider applications in fully
solution processed and printable perovskite solar cells, allowing
their economic fabrication.
Acknowledgements
Authors acknowledge support from CAPREX, Cambridge Africa
Alborada Fund, Carnegie-University of Ghana Next Generation of
Africa Academics programme and the Royal Academy of Engi-
neering (RAEng) through a research fellowship (Graphlex).
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.carbon.2016.04.012.
References
[1] X. Wang, L. Zhi, N. Tsao, Z
. Tomovic, J. Li, K. Müllen, Transparent carbon films
as electrodes in organic solar cells, Angew. Chem. 47 (2008) 2990e2992.
[2] S. Bae, H. Kim, Y. Lee, X. Xu, J.-S. Park, Y. Zheng, et al., Roll-to-roll production of
30-inch graphene films for transparent electrodes, Nat. Nanotechnol. 5 (2010)
5748.
[3] N. Yang, J. Zhai, D. Wang, Y. Chen, L. Jiang, Two-dimensional graphene bridges
enhanced photoinduced charge transport in dye-sensitized solar cells, ACS
Nano 4 (2010) 887e894.
[4] L. Wan, S. Wang, X. Wang, B. Dong, Z. Xu, X. Zhang, et al., Room-temperature
fabrication of graphene films on variable substrates and its use as counter
electrodes for dye-sensitized solar cells, Solid State Sci. 13 (2011) 468e475.
[5] W. Yuan, Y. Zhou, Y. Li, C. Li, H. Peng, Zhang, et al., The edge- and basal-plane-
specific electrochemistry of a single-layer graphene sheet, Sci. Rep. 3 (2013)
2248.
[6] K.S. Novoselov, A.K. Geim, S.V. Morozov, D. Jiang, Y. Zhang, S.V. Dubonos, et al.,
Electric field effect in atomically thin carbon films, Science 306 (2004)
666e669.
[7] X. Li, W. Cai, J. An, S. Kim, J. Nah, D. Yang, et al., Large-area synthesis of high-
quality and uniform graphene films on copper foils, Science 324 (2009)
1312e1314.
[8] S. Stankovich, D.A. Dikin, R.D. Piner, K.A. Kohlhaas, A. Kleinhammes, Y. Jia, et
al., Synthesis of graphene-based nanosheets via chemical reduction of exfo-
liated graphite oxide, Carbon 45 (2007) 1558e1565.
[9] J.N. Coleman, Liquid-phase exfoliation of nanotubes and graphene, Adv. Funct.
Mater 19 (2009) 3680e3695.
[10] J.N. Coleman, Liquid exfoliation of defect-free graphene, Acc. Chem. Res. 46
(2013) 14e22.
[11] T. Hasan, F. Torrisi, Z. Sun, D. Popa, V. Nicolosi, G. Privitera, et al., Solution-
phase exfoliation of graphite for ultrafast photonics, Phys. Status Solidi 247
(2010) 2953e2957.
[12] Y. Hernandez, V. Nicolosi, M. Lotya, F.M. Blighe, Z. Sun, S. De, et al., High-yield
production of graphene by liquid-phase exfoliation of graphite, Nat. Nano-
technol. 3 (2008) 563e568.
[13] C. Mattevi, H. Kim, M. Chhowalla, A review of chemical vapour deposition of
graphene on copper, J. Mater. Chem. 21 (2011) 3324e3334.
40 D. Dodoo-Arhin et al. / Carbon 105 (2016) 33e41[14] L. Kavan, J.H. Yum, M. Gra€tzel, Optically transparent cathode for dye-
sensitized solar cells based on graphene nanoplatelets, ACS Nano 5 (2011)
165e172.
[15] F. Torrisi, T. Hasan, W. Wu, Z. Sun, A. Lombardo, T.S. Kulmala, et al., Inkjet-
printed graphene electronics, ACS Nano 6 (2012) 2992e3006.
[16] F. Withers, H. Yang, L. Britnell, A.P. Rooney, E. Lewis, A. Felten, et al., Heter-
ostructures produced from nanosheet-based inks, Nano Lett. 14 (2014)
3987e3992.
[17] R.C.T. Howe, G. Hu, Z. Yang, T. Hasan, Functional inks of graphene, metal
dichalcogenides and black phosphorus for photonics and (opto)electronics,
Proc. SPIE 9553 (2015) 95530R.
[18] E.B. Secor, P.L. Prabhumirashi, K. Puntambekar, M.L. Geier, M.C. Hersam, Inkjet
printing of high conductivity, flexible graphene patterns, J. Phys. Chem. Lett. 4
(2013) 1347e1351.
[19] I.M. Hutchings, G.D. Martin, Inkjet Technology for Digital Fabrication, John
Wiley and Sons, Chichester U.K., 2012.
[20] A. Yella, H.-W. Lee, H.N. Tsao, C. Yi, A.K. Chandiran, M.K. Nazeeruddin, et al.,
Porphyrin-Sensitized Solar Cells with Cobalt (II/III)-Based Redox Electrolyte
Exceed 12 Percent Efficiency, Science 334 (2011) 629e634.
[21] G. Calogero, A. Bartolotta, G. Di Marco, A. Di Carlo, F. Bonaccorso, Vegetable-
based dye-sensitized solar cells, Chem. Soc. Rev. 44 (2015) 3244e3294.
[22] B. O'Regan, M. Gr€atzel, A low-cost, high-efficiency solar cell based on dye-
sensitized colloidal TiO2 films, Nature 353 (1991) 737e740.
[23] M. Gra€tzel, Photoelectrochemical cells, Nature 414 (2001) 338e344.
[24] J. Bisquert, D. Cahen, G. Hodes, S. Rühle, A. Zaban, Physical chemical principles
of photovoltaic conversion with nanoparticulate, mesoporous dye-sensitized
solar cells, J. Phys. Chem. B 108 (2004) 8106e8118.
[25] G. Calogero, J.-H. Yum, A. Sinopoli, G. Di Marco, M. Gr€atzel, M.K. Nazeeruddin,
Anthocyanins and betalains as light-harvesting pigments for dye-sensitized
solar cells, Sol. Energy 86 (2012) 1563e1575.
[26] S. Hao, J. Wu, Y. Huang, J. Lin, Natural dyes as photosensitizers for dye-
sensitized solar cell, Sol. Energy 80 (2006) 209e214.
[27] D. Matthews, P. Infelta, M. Gra€tzel, Calculation of the photocurrent-potential
characteristic for regenerative, sensitized semiconductor electrodes, Sol. En-
ergy Mater. Sol. Cells 44 (1996) 119e155.
[28] K. Tennakone, G.R.R.A. Kumara, A.R. Kumarasinghe, P.M. Sirimanne,
K.G.U. Wijayantha, Efficient photosensitization of nanocrystalline TiO2 films
by tannins and related phenolic substances, J. Photochem. Photobiol. A Chem.
94 (1996) 217e220.
[29] N.Y.M. Iha, Supramolecular photochemistry and solar cells, An. Acad. Bras.
Cienc 72 (2000) 67e74.
[30] Solaronix Ruthenizer 535-bisTBA, http://shop.solaronix.com/ruthenizer-535-
bistba.html.
[31] Y. Amao, T. Komori, Bio-photovoltaic conversion device using chlorine-e6
derived from chlorophyll from Spirulina adsorbed on a nanocrystalline TiO2
film electrode, Biosens. Bioelectron. 19 (2004) 843e847.
[32] A. Polo, N.Y.M. Iha, Blue sensitizers for solar cells: natural dyes from calafate
and jaboticaba, Sol. Energy Mater. Sol. Cells 90 (2006) 1936e1944.
[33] C.G. Garcia, A.S. Polo, N.Y. Murakami Iha, Fruit extracts and ruthenium poly-
pyridinic dyes for sensitization of TiO2 in photoelectrochemical solar cells,
J. Photochem. Photobiol. A Chem. 160 (2003) 87e91.
[34] Ø.M. Andersen, M. Jordheim, Anthocyanins. Encycl. Life Sci, John Wiley and
Sons, Chichester U.K., 2010.
[35] G.R.A. Kumara, S. Kaneko, M. Okuya, B. Onwona-Agyeman, A. Konno,
K. Tennakone, Shiso leaf pigments for dye-sensitized solid-state solar cell, Sol.
Energy Mater. Sol. Cells 90 (2006) 1220e1226.
[36] N.J. Cherepy, G.P. Smestad, M. Gr€atzel, J.Z. Zhang, Ultrafast electron injection:
implications for a photoelectrochemical cell utilizing an anthocyanin dye-
sensitized TiO2 nanocrystalline electrode, J. Phys. Chem. B 101 (1997)
9342e9351.
[37] N. Papageorgiou, Counter-electrode function in nanocrystalline photo-
electrochemical cell configurations, Coord. Chem. Rev. 248 (2004)
1421e1446.
[38] G. Smestad, C. Bignozzi, R. Argazzi, Testing of dye sensitized TiO2 solar cells I:
Experimental photocurrent output and conversion efficiencies, Sol. Energy
Mater. Sol. Cells 32 (1994) 259e272.
[39] X. Fang, T. Ma, G. Guan, M. Akiyama, T. Kida, E. Abe, Effect of the thickness of
the Pt film coated on a counter electrode on the performance of a dye-
sensitized solar cell, J. Electroanal. Chem. 570 (2004) 257e263.
[40] E. Olsen, G. Hagen, S. Eric Lindquist, Dissolution of platinum in methoxy
propionitrile containing LiI/I2, Sol. Energy Mater. Sol. Cells 63 (2000) 267e273.
[41] J.D. Roy-Mayhew, D.J. Bozym, C. Punckt, I.A. Aksay, Functionalized graphene as
a catalytic counter electrode in dye-sensitized solar cells, ACS Nano 4 (2010)
6203e6211.
[42] D.W. Zhang, X.D. Li, S. Chen, H.B. Li, Z. Sun, X.J. Yin, et al., Graphene Nanosheet
Counter-electrodes for Dye-sensitized Solar Cells. 3rd International Nano-
electronics Conference (INEC), IEEE, Hong Kong, China, 2010, pp. 610e611.
[43] J. Baker, D. Deganello, D.T. Gethin, T.M. Watson, Flexographic printing of
graphene nanoplatelet ink to replace platinum as counter electrode catalyst in
flexible dye sensitised solar cell, Mater. Res. Innov. 18 (2014) 86e90.
[44] Y. Peng, J. Zhong, K. Wang, B. Xue, Y.-B. Cheng, A printable graphene enhanced
composite counter electrode for flexible dye-sensitized solar cells, Nano En-
ergy 2 (2013) 235e240.
[45] A. Kay, M. Gra€tzel, Low cost photovoltaic modules based on dye sensitized
nanocrystalline titanium dioxide and carbon powder, Sol. Energy Mater. Sol.Cells 44 (1996) 99e117.
[46] E. Ramasamy, W.J. Lee, D.Y. Lee, J.S. Song, Nanocarbon counterelectrode for
dye sensitized solar cells, Appl. Phys. Lett. 90 (2007) 173103.
[47] K. Saranya, M. Rameez, A. Subramania, Developments in conducting polymer
based counter electrodes for dye-sensitized solar cells e an overview, Eur.
Polym. J. 66 (2015) 207e227.
[48] M. Wu, X. Lin, A. Hagfeldt, T. Ma, A novel catalyst of WO2 nanorod for the
counter electrode of dye-sensitized solar cells, Chem. Commun. 47 (2011)
4535e4537.
[49] A. Kaniyoor, S. Ramaprabhu, Thermally exfoliated graphene based counter
electrode for low cost dye sensitized solar cells, J. Appl. Phys. 109 (2011)
124308.
[50] H. Wang, S.L. Leonard, Y.H. Hu, Promoting effect of graphene on dye-
sensitized solar cells, Ind. Eng. Chem. Res. 51 (2012) 10613e10620.
[51] R. Cruz, J.P. Araújo, L. Andrade, A. Mendes, Transparent graphene-based
counter-electrodes for iodide/triiodide mediated dye-sensitized solar cells,
J. Mater. Chem. A 2 (2014) 2028e2032.
[52] W. Kwon, J.-M. Kim, S.-W. Rhee, Electrocatalytic carbonaceous materials for
counter electrodes in dye-sensitized solar cells, J. Mater. Chem. A 1 (2013)
3202e3215.
[53] M.-S. Wu, Y.-J. Zheng, Electrophoresis of randomly and vertically embedded
graphene-nanosheets in activated carbon film as a counter electrode for dye-
sensitized solar cells, Phys. Chem. Chem. Phys. 15 (2013) 1782e1787.
[54] J.D. Roy-Mayhew, I.A. Aksay, Graphene materials and their use in dye-
sensitized solar cells, Chem. Rev. 114 (2014) 6323e6348.
[55] T.N. Murakami, M. Gra€tzel, Counter electrodes for DSC: application of func-
tional materials as catalysts, Inorg. Chim. Acta 361 (2008) 572e580.
[56] S. Santra, G. Hu, R.C.T. Howe, A. De Luca, S.Z. Ali, F. Udrea, J.W. Gardner, et al.,
CMOS integration of inkjet-printed graphene for humidity sensing, Sci. Rep. 5
(2015) 17374.
[57] S. Park, R.S. Ruoff, Chemical methods for the production of graphenes, Nat.
Nanotechnol. 4 (2009) 217e224.
[58] J.M. Rehm, G.L. McLendon, Y. Nagasawa, K. Yoshihara, J. Moser, M. Gr€atzel,
Femtosecond electron-transfer dynamics at a sensitizing dye
semiconductor
(TiO2), Interface. J. Phys. Chem. 100 (1996) 9577e9578.
[59] K. Wongcharee, V. Meeyoo, S. Chavadej, Dye-sensitized solar cell using natural
dyes extracted from rosella and blue pea flowers, Sol. Energy Mater. Sol. Cells
91 (2007) 566e571.
[60] Q. Dai, J. Rabani, Photosensitization of nanocrystalline TiO2 films by pome-
granate pigments with unusually high efficiency in aqueous medium, Chem.
Commun. (2001) 2142e2143.
[61] F. Delgado-Vargas, A.R. Jimenez, O. Paredes-Lopez, Natural pigments: carot-
enoids, anthocyanins, and betalains d characteristics, biosynthesis, process-
ing, and stability, Crit. Rev. Food Sci. Nutr. 40 (2000) 173e289.
[62] S. Roy, C. Llewellyn, E.S. Egeland, G. Johnsen, Phytoplankton Pigments, Cam-
bridge University Press, Cambridge, U.K, 2011.
[63] D.M. de Leeuw, M.M.J. Simenon, A.R. Brown, R.E.F. Einerhand, Stability of n-
type doped conducting polymers and consequences for polymeric micro-
electronic devices, Synth. Met. 87 (1997) 53e59.
[64] J.L. Bredas, R. Silbey, D.S. Boudreaux, R.R. Chance, Chain-length dependence of
electronic and electrochemical properties of conjugated systems: poly-
acetylene, polyphenylene, polythiophene, and polypyrrole, J. Am. Chem. Soc.
105 (1983) 6555e6559.
[65] C.M. Hansen, Hansen Solubility Parameters: a User's Handbook, Boca Raton
FL: CRC Press, Inc., 2007.
[66] Y. Hernandez, M. Lotya, D. Rickard, S.D. Bergin, J.N. Coleman, Measurement of
multicomponent solubility parameters for graphene facilitates solvent dis-
covery, Langmuir 26 (2010) 3208e3213.
[67] C.L. Laws, Thermophysical Properties of Chemicals and Hydrocarbons, William
Andrew Inc, Norwich, NY, 2008.
[68] F.M. Fowkes, Contact Angle, Wettability, and Adhesion, American Chemical
Society, Washington, D.C., 1964.
[69] H. Kipphan, Handbook of Print Media: Technologies and Production Methods,
Springer, Berlin, Germany, 2001.
[70] A. O'Neill, U. Khan, P.N. Nirmalraj, J. Boland, J.N. Coleman, Graphene disper-
sion and exfoliation in low boiling point solvents, J. Phys. Chem. C 115 (2011)
5422e5428.
[71] T. Hasan, P.H. Tan, F. Bonaccorso, A.G. Rozhin, V. Scardaci, W.I. Milne, et al.,
Polymer-assisted isolation of single wall carbon nanotubes in organic solvents
for optical-quality nanotube-polymer composites, J. Phys. Chem. C 112 (2008)
20227e20232.
[72] T. Hasan, V. Scardaci, P. Tan, A.G. Rozhin, W.I. Milne, A.C. Ferrari, Stabilization
and “debundling” of single-wall carbon nanotube dispersions in N-methyl-2-
pyrrolidone (NMP) by polyvinylpyrrolidone (PVP), J. Phys. Chem. C 111 (2007)
12594e12602.
[73] M.J. O'Connell, P. Boul, L.M. Ericson, C. Huffman, Y. Wang, E. Haroz, et al.,
Reversible water-solubilization of single-walled carbon nanotubes by poly-
mer wrapping, Chem. Phys. Lett. 342 (2001) 265e271.
[74] C.F. Bohren, D.R. Huffman, Absorption and Scattering of Light by Small Par-
ticles, Wiley, Weinheim, Germany, 1998.
[75] V.G. Kravets, A.N. Grigorenko, R.R. Nair, P. Blake, S. Anissimova,
K.S. Novoselov, et al., Spectroscopic ellipsometry of graphene and an exciton-
shifted van Hove peak in absorption, Phys. Rev. B 81 (2010) 155413.
[76] K.-S. Chou, C.-C. Chen, Fabrication and characterization of silver core and
porous silica shell nanocomposite particles, Microporous Mesoporous Mater
D. Dodoo-Arhin et al. / Carbon 105 (2016) 33e41 4198 (2007) 208e213.
[77] J. Li, F. Ye, S. Vaziri, M. Muhammed, M.C. Lemme, M. O€stling, Efficient inkjet
printing of graphene, Adv. Mater. 25 (2013) 3985e3992.
[78] A. Gupta, G. Chen, P. Joshi, S. Tadigadapa, P.C. Eklund, Raman scattering from
high-frequency phonons in supported n e graphene layer films, Nano Lett. 6
(2006) 2667e2673.
[79] C.-H. Zhou, Y. Yang, J. Zhang, S. Xu, S.-J. Wu, H. Hu, et al., Enhanced electro-
chemical performance of the counterelectrode of dye sensitized solar cells by
sandblasting, Electrochim. Acta 54 (2009) 5320e5325.
[80] M. Dürr, A. Yasuda, G. Nelles, On the origin of increased open circuit voltage of
dye-sensitized solar cells using 4-tert-butyl pyridine as additive to the elec-
trolyte, Appl. Phys. Lett. 89 (2006) 061110.
[81] C.Y. Jiang, X.W. Sun, K.W. Tan, G.Q. Lo, A.K.K. Kyaw, D.L. Kwong, High-bend-
ability flexible dye-sensitized solar cell with a nanoparticle-modified ZnO-
nanowire electrode, Appl. Phys. Lett. 92 (2008) 143101.
[82] K. Miettunen, M. Toivola, G. Hashmi, J. Salpakari, I. Asghar, P. Lund, A carbon
gel catalyst layer for the roll-to-roll production of dye solar cells, Carbon 49
(2011) 528e532.
[83] J. Baker, J.D. McGettrick, D.T. Gethin, T.M. Watson, Impedance characteristics
of transparent GNP-Pt ink catalysts for flexible dye sensitized solar cells,J. Electrochem. Soc. 162 (2015) H564eH569.
[84] G.P. Smestad, Education and solar conversion: demonstrating electron
transfer, Sol. Energy Mater. Sol. Cells 55 (1998) 157.
[85] S. Meng, J. Ren, E. Kaxiras, Natural dyes adsorbed on TiO2 nanowire for
photovoltaic applications: enhanced light absorption and ultrafast electron
injection, Nano Lett. 8 (2008) 3266e3272.
[86] J.-H. Yum, P. Chen, M. Graetzel, M.K. Nazeeruddin, Recent developments in
solid-state dye-sensitized solar cells, Chem. Sus Chem. 1 (2008) 699.
[87] M. Amaresh, K. Markus, R. Fischer, P. Bauerle, Metal-free organic dyes for dye-
sensitized solar cells: from structure: property relationships to design rules,
Angew. Chem. Int. Ed. 48 (2009) 2474e2499.
[88] K. Hee-Je, K. Dong- Jo, S.N. Karthick, K.V. Hemalatha, C. Justin Raj, C. Youngson,
Curcumin dye extracted from curcuma longa L, used as sensitizers for efficient
dye-sensitized solar cells, Int. J. Electrochem. Sci. 8 (2013) 8320e8328.
[89] F. Queiroz, C. Oliveira, O. Pinho, I.M. Ferreira, Degradation of anthocyanins and
anthocyanidins in blueberry jams/stuffed fish, J. Agric. Food Chem. 57 (22)
(2009) 10712e10717.
[90] W. Yinni, W. Zhisheng, A. Zhongwei, C. Xinbing, C. Pei, L. Qianfeng, Highly
efficient dye-sensitized solar cells by co-sensitization of organic dyes and co-
adsorbent chenodeoxycholic acid, Chin. J. Chem. 32 (2014) 474e478.