Materials Science in Semiconductor Processing 74 (2018) 267–276

Contents lists available at ScienceDirect

Materials Science in Semiconductor Processing

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

An investigation of the dye-sensitized solar cell performance using MARK

graphene-titania (TrGO) photoanode with conventional dye and natural
green chlorophyll dye
⁎
Siti Zubaidah Siddick, Chin Wei Lai , Joon Ching Juan
Nanotechnology & Catalysis Research Centre (NANOCAT), Institute of Postgraduate Studies (IPS), University of Malaya, 3rd Floor, Block A, Kuala Lumpur, Malaysia

A R T I C L E I N F O A B S T R A C T

Keywords: Development of innovative low-cost and efficient DSSCs (dye-sensitized solar cell) device for solar energy
Natural dye conversion is essential. In this present study, the performance of DSSCs based on graphene-titania (TrGO)
Reduced graphene oxide (RGO) photoanode using conventional N719 (synthetic dye) and natural green chlorophyll dye (organic dye) were
Titanium dioxide (TiO2) extensively studied, analysed and discussed. TrGO was successfully synthesized via one step sol precipitation
Dye sensitized solar cell (DSSCs)
peptization technique using titanium (IV) isopropoxide (TTIP) and reduced graphene oxide (rGO) as starting
TiO2-RGO
materials. It was found that TrGO-2 exhibited the highest photocurrent generation among samples regardless of
Nanocomposite photoanode
dye type. DSSCs device fabricated using N719 dye showed 9.15 mA cm−2 of photocurrent with 3.95% efficiency
whereas DSSCs device using natural chlorophyll green dye exhibited photocurrent density of 3.43 mA cm−2 with
0.67% efficiency. From the Tauc plot, band gap was narrowed from 3.17 eV of TiO2 to 2.78 eV of TrGO con-
sidering the graphene extended the absorption range of the film to the visible light region. The main reason was
associated with the better surface contact of the TiO2 dopants with the rGO film resulting in a lower charge
transfer resistance.

1. Introduction sensitizers in DSSCs. This is attributed to the fact that the electronic
structures of the plant pigments have the ability to interact with sun-
Historically, M. Gratzel and O’Regan successfully demonstrated the light and turn it into an electric current by altering the wavelength
first 3rd-generation solar cells known as DSSCs which converted the [10]. In the present study, green chlorophyll extracted from the pandan
light energy into electricity via the electron-transfer event between the leaves (Pandanus amarylliforus) was extracted as the natural green
dye (sensitizer) and the metal oxide in 1991 [1–6]. In 2005, Nazeer- chlorophyll dye pigment. Specifically, chlorophyll could be classified as
uddin et al. [4] reported that the champion DSSCs sensitized with N719 a unique pigment ascribed to its ability to conduct photosynthesis
dye (ruthenium complexes) illustrated approximately 10% of power converting the light energy to transduction energy in plant. Ad-
conversion efficiency under AM 1.5. However, DSSCs using ruthenium ditionally, chlorophyll (a mixture of 2 pigment complexes, namely
complexes suffered from shortcomings in terms of their high cost (due chlorophyll a and b) is an attractive candidate as the sensitizer in DSSCs
to the limited production of ruthenium) and hazardous nature of ru- attributed to its tendency to absorb blue and red lights. Also, many
thenium (non-environmental friendliness of the heavy metal) [7]. research works have been focussed on the preparation of porphyrin-
In recent years, dyes derived from natural sources as the photo- type organic dye from chlorophyll due to its low lost, ease of pre-
sensitizer for DSSCs have been extensively studied as the possible in- paration and eco-friendly. Although the commercialized N719 (ruthe-
expensive and eco-friendly alternatives to the conventional material nium transition metal polypyridyl complexes) is the most widely used
with high light-harvesting efficiency, [5,8,9]. In order to bring the re- synthetic dye with photoelectric conversion efficiency (PCE) of > 10%
newable solar energy to the point of commercial readiness, substantial [11], DSSCs equipped with natural dyes as the sensitizers have several
research efforts towards the development of hybrid semiconductor/ advantages over the synthetic N719 dye considering the latter contains
photoelectrode for high-efficient DSSCs have been developed. heavy metal and is not cost effective.
Accordingly, natural dyes especially cyanine, chlorophyll, antho- In other respect, it has been reported that DSSCs photoanode com-
cyanin, carotene and flavonoid have been extensively studied as posed by the thick mesoporous TiO2 nanoparticles (~ 10 to 15 µm)

⁎
Corresponding author.
E-mail addresses: sitizubaidah48@gmail.com (S.Z. Siddick), cwlai@um.edu.my (C.W. Lai).

http://dx.doi.org/10.1016/j.mssp.2017.10.046
Received 31 July 2017; Received in revised form 23 October 2017; Accepted 27 October 2017
1369-8001/ © 2017 Elsevier Ltd. All rights reserved.
S.Z. Siddick et al. Materials Science in Semiconductor Processing 74 (2018) 267–276

network would serve as a large-specific-surface-area anchoring site for 2.2. Preparation of graphene oxide (GO) and reduced graphene oxide
the dye molecules to absorb in the photoanode. In an open literature,
titanium dioxide (TiO2) itself has recorded a remarkable high photo- Huang’s method for the preparation of graphene oxide (GO) [21]
conversion efficiency (PCE) in DSSCs. However, TiO2 tends to suffer was adopted in the present study [18]. Accordingly, the one-pot che-
from shortcomings in terms of its relatively large band gap, which re- mical oxidation of graphite was carried out at room temperature for 3
sults in the fast recombination rate. One possible explanation is that the days to allow the oxidation of graphite for the preparation of GO. The
injected electron tends to travel in the random colloidal particle matrix solution was then subjected to centrifugation at 10,000 rpm to form the
and grain boundaries of TiO2, creating a random transfer route followed GO gel. The dried GO powder was then subjected to reduction process
by the trap-limited diffusion process [12]. In other words, the charge to prepare the reduced graphene oxide. Specifically, GO solution was
carrier recombination rate increases tremendously, which leads to the prepared by adding the GO powder to a flask containing 3 mg/ml of
reduction of the PCE in DSSCs. Taking these facts into serious con- purified water. The hydrazine solution was immediately added to the
sideration, it is of great important to introduce a highly conductive GO solution, followed by refluxing in an oil bath at 80 °C overnight. The
material in metal oxide semiconductor to enhance the PCE. black precipitate was then filtered and dried to obtain the rGO powders.
Correspondingly, reduced graphene oxide (rGO) which is in-
expensive and abundance has been attracted tremendous research in-
terest in the field of photovoltaic DSSCs [13–18]. Due to its remarkable 2.3. Synthesis of TrGO (graphene-titania) nanocomposite
properties such as good flexibility, zero bandgap, high mobility carrier
coupled with large surface area, graphene appeared as one of the pro- In the present work, the precursor solution [solution A] was pre-
mising carbonaceous materials that has been widely used as photo- pared by dissolving 11 ml TTIP in 50 ml propanol. In the meanwhile,
anodes in DSSCs device [19–23]. It is noteworthy to mention that the 0.01 g rGO powders were dispersed in 50 ml DI water, followed by
rGO has a matching conduction band with TiO2 and thereby, a rapid sonication for 2 h. 1.5 ml nitric acid serving as an oxidizing agent was
charge transfer could be triggered between the rGO and TiO2 surface. then added to the as-prepared rGO solution to prepare solution B.
Also, the photo-induced electrons can move through rGO bridge rather Solution A was added dropwise into the solution B with a constant
than TiO2 – TiO2 grain boundary as a result of the diminished charge stirring and then allowed to stand for aging for 6 h at 80 °C. Next, the
recombination [24]. In view of these facts, the combination of TiO2 and collected sol was dried at 60 °C to obtain the TrGO nanocomposite
rGO to produce the TiO2/rGO composites is an alternative route to powder. Finally, the nano-powder was annealed at 450 °C for 6 h, and
further improve the conduction pathways and PCE of DSSCs from the then allowed to cool until it has reached room temperature. The same
point of photo-induced electrons at photoanode to the charge collector procedures were repeated to prepare TrGO nano-composites containing
electrode. different weight percentages of rGO; TrGO-1 0.01 g, TrGO-2 0.03 g and
In the present research study, a novel method (one step sol pre- TrGO-3 0.05 g, respectively.
cipitation peptization technique) to encapsulate TiO2 into the rGO
photo-anode film was developed with the aim to enhance the photo-
current of DSSCs by minimizing the electron transfer resistance in the 2.4. Preparation of TiO2 paste and TrGO paste
device and intensifying the electron transport rate. Innovative ap-
proaches by using one step sol precipitation peptization technique to Fluorine-doped tin oxide covered glass (FTO) was used as the sub-
synthesis hybrid of TiO2/rGO composites as photoanode are crucial for strate in the present work to fabricate the photoanode. Accordingly, the
developing efficient DSSCs system. Notably, one-step one step sol pre- FTO glass was first soaked in acetone and then subsequently cleaned in
cipitation peptization technique contributed well TiO2 particles in well an ultrasonic bath for about 15 min to remove all contaminants present
crystalline formed without agglomeration on rGO sheets. Furthermore, on the surface of the substrate, followed by rinsing with DI water and
the performance of the photoanode sensitized with natural dye has also isopropanol. To prepare TrGO paste, 2.0 g TrGO powder was initially
been studied to evaluate the photocurrent performances and absorption dispersed into 10 ml absolute ethanol and left to stir at room tem-
compatibility between the natural dye and TiO2/rGO photoanode. It is perature for 2 h, followed by sonication for approximately 30 min until
also of great important to tailor the morphology of the photoanodes by a homogeneous suspension was formed. Finally, a drop of trixton-100
optimizing the ratio of TiO2 and rGO in the composites to provide more was added to the solution. Adopting the doctor blade technique, the
conductive pathways for photo-induced electron transfer. To the best of TrGO paste was spread on the FTO substrate, followed by drying at
our knowledge, there have been very limited reports of one-step room temperature for about 30 min and heating at 100 °C for 10 min to
synthesis of TrGO composite through sol precipitation peptization prepare the photoanode electrode. The semiconductors were then
technique that resulted in high PCE in DSSCs. Therefore, this study aims subjected to a final thermal sintering process at 450 °C for 30 min. For
to optimize the TiO2/rGO ratio for improving the PCE in the natural the sake of comparison, a pure TiO2 paste and photoanode were fab-
green chlorophyll- and N719-sensitized DSSCs. ricated using the same procedure as of those TrGO.

2. Materials and methods 2.5. Dye preparation

2.1. Material Two different types of dyes (i.e., conventional N719 dye and natural
extracted green chlorophyll dye) were prepared in the present study.
Natural graphite powder (< 20 µm), analytical grade sulphuric acid Correspondingly, N719 (ruthenizer 535-bis TBA) was diluted with
(H2SO4, 95.0–98.0%), phosphoric acid (H3PO,85.0%), potassium per- ethanol to prepare the 0.5 mM N-719 dye ethanol solution. In the
manganate (KMnO4,99.9%), hydrogen peroxide (H2O2, 30 wt%), hy- meanwhile, Soxhlet extraction method was adopted to extract the
drochloric acid (HCl, 37.0%), hydrazine solution 37%, titanium (IV) natural green chlorophyll dye from pandan leaves. Specifically, 10 g of
isopropoxide (TTIP, 97.0%), ethanol (EtOH, 99.0%), nitric acid (HNO3, pandan leaves (source of chlorophyll) was initially cut into smaller
68.0%), fluorine tin oxide (FTO) conducting glass slides (7 Ω/sq) and pieces and filled in the net cloth. 250 ml ethanol was used as the solvent
triton X-100 laboratory scale were purchased from Sigma Aldrich. N719 for the extraction process. Following this, extraction of dye pigment
(ruthenizer 535-bis TBA) and electrolyte EL-HPE were supplied by was carried out in a Soxhlet extractor overnight. Next, the chlorophyll
Dyesol. pigment solution was placed in rotary evaporator to prepare the
chlorophyll pigment in gel form.(Fig. 1)

268
S.Z. Siddick et al. Materials Science in Semiconductor Processing 74 (2018) 267–276

Fig. 1. N719 commercialize dye and natural green chlorophyll dye.

Fig. 3. Fourier Transform Infrared (FTIR) spectra of pure TiO2, TrGO-1, TrGO-2 and
2.6. The cell assembling of DSSCs and characterization TrGO-3.

Fig. 2 illustrates the schematic diagram for assembling process of 3. Results and discussions
DSSCs. Accordingly, the pure TiO2- and TrGO-based photoanodes were
sensitized by soaking the aforementioned electrodes in the dye solution 3.1. Characterization of TrGO photoanode
containing 0.5 mM N719 and natural green dye overnight. Then, pho-
toanodes were rinsed with ethanol to remove unabsorbed molecules FTIR spectroscopy was used in the present study to evaluate the
from the surface and dried. Next, the DSSCs were assembled by sand- functional groups on TrGO and pure TiO2. As illustrated in Fig. 3, both
wiching the sensitized photoanode and carbon counter electrode by TiO2 and TrGO showed a broad band below 1000 cm−1, implying Ti-O-
introducing 6.0 µL EL-HPE electrolyte between the electrodes. The Ti and Ti-O-C stretching vibrations within the molecules, respectively
current density–voltage (J-V) characteristics of DSSCs were then mea- [25–27]. It was also noted that the stretching peak of C-O at about
sured under AM 1.5 stimulated solar illumination with a power density 1084 cm−1, which appeared in pure TiO2, disappeared in the TrGO
of 100 mW/cm2. The electrochemical impedance spectroscopy (EIS) nanocomposites. From Fig. 3, it was also observed that TrGO nano-
analysis of DSSCs was conducted using a Methohm Autolab with fre- composite exhibited some characteristic peaks, which could be assigned
quency range from 0.1 Hz to 100 kHz at the open circuit conditions to the C˭C stretching (1541 cm−1), C˭O stretching (1695 cm−1) and
under 100 mW/cm2 of irradiation to identify the photoconversion ef- –OH stretching (3440 cm−1) [25]. There was also a strong peak located
ficiency of DSSCs coupled with the electron movement and re- at 2357 cm−1 assigned to the CO2 contamination [28]. Also, as dis-
combination across the device. The morphologies of the electrodes were played in Fig. 3, there were peaks in the region between 3600 and
studied using a high resolution-transmission electron microscope (HR- 3900 cm−1 corresponding to the formation of oxygen vacancies in-
TEM, JEOL) and field emission scanning electron microscope (FESEM, duced by the formation of Ti3+ species [29], while the peaks at 3724
Joel). On top of that, raman spectroscopy, photoluminescence (PL) and 3845 cm−1 were attributed to the Ti3+-OH formation. Moreover,
spectra, fourier transform infrared (FTIR), UV–vis spectroscopy were the peak at 3674 cm−1 corresponded to the Ti4+ in the TiO2.
employed in the present work to characterize the chemical and struc- Raman spectroscopy was employed in the present study to char-
tural properties of the photoelectrode. acterize the structural properties of the graphene materials. Fig. 4
shows the Raman spectra of pure TiO2 and TrGO. It is clear that both
the TiO2 and TrGO possessed the anatase phase consisting of five major
vibration peaks at 144 cm−1 (Eg), 197 cm −1 (Eg), 395 cm−1 (B1b),
513 cm−1 (A1b), and 635 cm−1 (Eg). This is in line with the finding
reported by Otakar et al. [30]. Also, it is noteworthy that TrGO showed

Fig. 2. Schematic diagram for DSSCs assembling method.

269
S.Z. Siddick et al. Materials Science in Semiconductor Processing 74 (2018) 267–276

Fig. 5. Photoluminescence (PL) spectra of pure TiO2; TrGO-1; TrGO-2 and TrGO-3.

pair of electrons [33]. Furthermore, as shown in Fig. 5, pure TiO2

photoanode displayed higher PL intensity compared to that of TrGO
samples. The reduced intensity of PL spectra was in concomitant to the
decrease of the weight percent of TiO2 in the nanocomposite since the
PL intensity was mainly influenced by the recombination rate of elec-
tron-hole pairs. It has also been reported in literatures that a lower
Fig. 4. Raman spectroscopy of TiO2 and TrGO (TrGO-1, TrGO-2 and TrGO-3). intensity of PL spectra implied an increase in charge separation and a
decrease in the rate of electron-hole pair recombination [34–36]. In-
terestingly, TrGO-1 obtained the highest intensity PL peak among
two broad vibration bands within 1200–1700 cm−1, in which the D
samples and assigned as high recombination rate of electron-hole pairs
band (attributed to the breathing mode of A1g symmetry involving
and slower the electrolyte redox reaction. The reason might be attrib-
phonons near the K zone boundary) appeared at 1309 cm−1, while the
uted to the insufficient rGO dopants to decorate the TiO2 surface and
G band (originates from sp2 in-plane vibration mode of carbon atoms)
resulted more point or lattice defects within the bulk of TrGO-1 sample,
at 1583 cm−1. As reported by Ahmad et al. [31], the intensity ratio of D
which increased the number of recombination centers significantly.
and G bands(ID/IG) could represent the structure, defects and size of the
These recombination centers obviously decreases the mobility of photo-
graphene. Correspondingly, ID/IG was found to be 0.38 suggesting a low
induced electrons due to the increased series of resistance caused by
level of defect, high crystallinity and small size of rGO in TrGO. Ad-
these trap states.
ditionally, as displayed in Fig. 4, Raman 2D band at about 2600 cm−1
Fig. 6 displays the absorption spectra of pure TiO2 and TrGO-based
was detected in TrGO-1 implying the presence of aggregates of some
nanocomposites. Correspondingly, these spectra give the information
rGO layers in the nanocomposites which might be associated with the
regarding to the absorption ability of the photoanode and the bandgap
restacking of rGO layers during the synthesis process [32]. The Raman
of the anatase semiconductor oxide film. In general, these typical ab-
band shifts and broadening in Eg mode (144 cm−1) was also observed in
sorptions with an intense transition in the UV–visible region were used
Fig. 4. One possible explanation [29], reported that the peak shifts and
to measure the intrinsic band gap absorption of samples where the
broadening might originate from the surface defects such as the pre-
electron transited from the valence band (VB) into conduction band
sence of oxygen vacancy and peak broadening which roughly show
(CB) (O2p – Ti3d). It is noteworthy that the optical absorbance
pools of oxygen vacancies.
The PL spectrum was used to identify the behavior of photo-gen-
erated electron and holes in the semiconductor layer of DSSCs. Fig. 5
shows the PL spectra of TiO2 and TrGO nano-composites. The presence
of PL peaks in the wavelength range of 450–750 nm indicated that TiO2
and TrGO nanocomposites are the indirect band gap semiconductors. It
was also determined that the presence of a peak near 330 nm was
mainly attributed to the ligand to metal charge transfer in TiO2 (O2- to
Ti4+), while the peaks at 380 nm, 450 nm and 660 nm signified the
surface oxygen vacancy related defects. Furthermore, one may find that
all peaks were blue shifted in the Raman peak. This implies the trans-
formation of un-coordinated Ti3+ ions to the surface of oxygen va-
cancy. The green band with peak position at 540 nm was also an in-
dication of the surface oxygen vacancy of anatase (101) phase.
The red shift of the PL spectra between 600 and 620 nm is attributed
to the oxygen-vacancy-related defects as a result of the oxygen atom
removal by annealing in an inert nitrogen atmosphere where the re-
moval of the oxygen atom will create electron traps under the con-
duction band [19]. Specifically, the electron traps usually locate at
0.7–1.6 eV below the conduction band edge energy. Ti4+ will be re-
duced to Ti3+considering the removal of oxygen atom can create lone Fig. 6. Absorption plot of pure TiO2; TrGO-1; TrGO-2 and TrGO-3.

270
S.Z. Siddick et al. Materials Science in Semiconductor Processing 74 (2018) 267–276

3.2. Dye absorption spectra

Fig. 10 indicates the UV–vis absorption spectra for both natural

green chlorophyll dye (extracted from the pandan Leaves) and con-
ventional N719 dye. It has been reported in the previous studies that
the absorption range for conventional N719 dye was around
400–600 nm, while the absorption range for green chrorophyll dye was
in 400–500 nm with some broad peaks around 520–620 nm and
620–700 nm. Taking these facts into consideration, it could be con-
cluded that the absorbance of natural green chlorophyll dye was more
efficient than that of the N719 dye in the absorption range of
400–500 nm. However, N719 would have a better absorbance perfor-
mance than the natural green chlorophyll dye at higher absorbance
range (such as 500–600 nm).

3.3. Photovoltaic performances of DSSCs

Fig. 7. Tauc Plot of pure TiO2; TrGO-1; TrGO-2 and TrGO-3. Figs. 11 and 12 show the photocurrent density-voltage (J-V) char-
acteristic of the N719- and green chlorophyll dyes-sensitized DSSCs
under 100 mW cm−2 stimulated solar irradiation, while Tables 2 and 3
spectrum of all the films was in the range of ~ 300 to 430 nm. The
summarized and tabulated their photoelectrochemical characteristics.
photon absorption ability in TrGO was also determined to be higher
Fig. 13 illustrates the comparison of PCE of DSSCs based on the com-
than that of those pure TiO2 (3.17 eV) due to the peak shift towards the
mercialize N719 and natural green chlorophyll dyes. It is noteworthy
blue region. Based on the results obtained, TrGO-3 possessed the
that the DSSC fabricated with the rGO photoanodes showed an overall
highest absorbance in the visible-light range at approximately 430 nm,
better performance than that of those with pure TiO2. This could be
corresponding to band gap energy of about 2.78 eV. In addition, the
related to the reduction in resistances of each layer, the charge trans-
blue shifting of the absorbance spectra also gives us a hint that the
port and recombination properties. Accordingly, the photocurrent
crystalline size of TrGO was smaller than that of TiO2.
density-voltage curves fill factor (FF) (Eq. (1)) and the overall conver-
Fig. 7 shows the Tauc plot, while Table 1 tabulates the energy band
sion efficiency (%) (Eq. (2)) were calculated using these relations:
gap (from the UV spectra) of pure TiO2 and TrGO films. It was found
that, based on Table 1, the optical energy band gap decreased pro- FF = Pmax / Isc Voc (1)
gressively with increasing rGO weight percentage in TrGO composite.
ŋ = Voc Isc FF / Pin. (2)
The smallest band gap was detected to be approximately 2.78 eV
(TrGO-3). The reduction of band gap was mainly attributed to the en- where Pmax (maximum power output) is obtained from Imax (maximum
hancement of electron movement to the conduction band induced by photocurrent density) and Vmax (maximum photovoltage). Isc and Voc
rGO. represent short circuit current and open circuit voltage, respectively,
FE-SEM and HR-TEM were used to determine the morphology of the while Pin is the power of incident light of the overall conversion effi-
nanocomposites. Based on the surface morphology of TiO2/rGO (Figs. 8 ciency.
and 9), two main conclusions could be drawn: (i) TiO2 loaded between From the J-V curves, it is interesting to note that TrGO-2 nano-
rGO sheets with little agglomeration; and (ii) the crumpling together of composite showed the highest photocurrent generation regardless of
rGO layer sheets with TiO2 grown on it. On top of that, FE-SEM images the dye type used, and the N719 dye-sensitized TrGO-2 illustrated a
(as shown in Fig. 9) suggest that the TiO2 microspheres were en- higher photocurrent efficiency (3.95%) compared to that of the natural
capsulated well in rGO sheets with the increase in rGO content in the green chlorophyll dye-sensitized TrGO-2 (0.67%). It was also found
TrGO nanocomposite. From HR-TEM images (Fig. 8), it was found that that, in Fig. 13, the overall photocurrent efficiency (ŋ) of DSSCs im-
the TiO2 particles with the particle size of ~ 17–28 nm were in-situ proved by the addition of rGO into the photoanode up to 0.03 g (i.e.,
grown on the surface of rGO during the synthesis of TrGO particle, TrGO-2). This is associated with the fact that the addition of rGO en-
while the fringe spacing of TiO2 was estimated to be 0.35 nm ensure the hanced the transfer path for the ultrafast ejected electrons to travel
presents of (101) anatase. Based on these observations, it is presumably between TiO2 colloids [37]. Additionally, rGO which acts as the pro-
that the TrGO composite possesses enhanced charge transport con- tection layer for the photoanode also prevented the liquid electrolyte
sidering the photo-induced electrons could move through the rGO from reaching the FTO substrate. As a result, these phenomena con-
bridge rather than solely the Ti-Ti grain as a result of the intimate tributed to an increase in the current density while minimizing/redu-
contact between the TiO2 and rGO. Also, the photoanodes absorption of cing the recombination losses of charge carrier generation. On top of
dye was relatively improved by the addition of rGO ascribed to the photocurrent efficiency and current density, TrGO-1 and TrGO-2 also
larger surface area for dye loading. Considering these facts, it is thereby showed a massive improvement in Voc. Correspondingly, the Voc im-
expected that the photoelectric properties of the DSSCs could be sig- provement led to the reduction in the Fermi level in the photoanode
nificantly enhanced. and the Nerst potential of redox couple in the electrolyte when it was
under illumination [24]. However, DSSCs containing higher amount of
rGO (i.e., TrGO-3) lost its performance as a consequence of the drop in
Table 1 the conversion efficiency even though TrGO-3 possesses the lowest
Energy band gap from UV spectra.
energy band gap. This is mainly due to the antagonistic effect between
Sample Band gap (eV) the rGO and dye molecules to compete for the visible light. In other
words, the photon of visible light absorbed by the dye molecules would
Pure TiO2 3.17 become less if the rGO content is high considering rGO only absorbed
TrGO-1 3.02
approximately 4% of light [40,41]. This would give rise to less excited
TrGO-2 2.85
TrGO-3 2.78 electron traveling in the DSSCs device and eventually lead to a lower
current density and PCE of DSSCs.

271
S.Z. Siddick et al. Materials Science in Semiconductor Processing 74 (2018) 267–276

Fig. 8. HR-TEM images of TrGO-2 (a) 50k magnification (b) 100k magnification.

Apart from that, as shown in Fig. 13, it was interesting to point out structures of the natural dye.
that there was a slight decrease in the photocurrent density-voltage Considering the aforementioned facts, it is of great important to
curves fill factor (FF) of the TrGO-based DSSCs compared to that of optimize the loading of TiO2 dopants attached to the rGO sheets to
those TiO2-based DSSCs sensitized with N719. However, it is note- maximize the photovoltaic performance. It has been widely accepted
worthy that the efficiency of the TrGO-based DSSCs was much higher that the photon absorption could be significantly improved with a good
than those based on pure TiO2, regardless of the dye type. Certainly, combination between TiO2 and rGO. Nevertheless, excessive TiO2 na-
both dyes have a positive impact on the photoanodes after the addition noparticle dopants attached to rGO photoanode might cause electron
of rGO in the electrolyte, especially the natural green chlorophyll dye. It leakage in the DSSCs device because of the poor photons absorption
is also undeniable that conventional N719 dye showed a better per- [38]. Also, the aim of the present study was to enhance the absorption
formance than the natural dye, but it is strongly believed that a stable of the dye via optimizing the ratio of TiO2 and rGO dopants within
and high performance yet environmentally friendly dye could be pre- photoanode to maximize the photovoltaic performance of DSSCs. Ac-
pared by proper altering the R-chains and stabilizing the porphyrin cordingly, the electron transport path can be improved by incorporating

Fig. 9. FE-SEM images of TiO2 grow on graphene film on 50k magnification (a) TiO2; (b) TrGO-1; (c) TrGO-2 (d) TrGO-3.

272
S.Z. Siddick et al. Materials Science in Semiconductor Processing 74 (2018) 267–276

Table 2
Photocurrent density-voltage efficiency (PCE) characteristic DSSCs using conventional
N719 dye.

Sample Voc(mV) Jsc(mA cm−2) FF(%) ŋ(%)

Pure TiO2 604 6.55 0.73 2.89

TrGO-1 619 8.53 0.67 3.53
TrGO-2 644 9.15 0.67 3.95
TrGO-3 613 7.96 0.67 3.26

Table 3
Photocurrent density-voltage efficiency (PCE) characteristic DSSCs using natural green
dye.

Sample Voc(mV) Jsc(mA cm−2) FF(%) ŋ(%)

Pure TiO2 305 2.96 0.39 0.35

TrGO-1 313 3.22 0.45 0.45
TrGO-2 435 3.43 0.45 0.67
TrGO-3 322 3.30 0.47 0.50
Fig. 10. UV–vis absorbance plot for natural green chlorophyll dye and conventional
N719.

of rGO. From the previous study, it was known that the resistance be-
tween the grain and path could be minimized by incorporating rGO into
TiO2 as rGO tends to build the junction bridge between TiO2
[24,39–46]. This condition might help in enhancing the charge se-
paration, electron collector and transportation of photo-induced elec-
trons as well as reducing the probability of recombination losses. This is
also beneficial for the photo-induced electrons to transfer without suf-
fering from massive resistance [37,41,47,48].

3.4. EIS analysis of DSSCs

Electrochemical impedance spectroscopy (EIS) was employed to

determine the electronic properties behavior and the internal resistance
in the DSSCs device. Accordingly, EIS test was conducted under the
illumination condition of 100 mW/cm2 over the frequency range of
100 kHz to 0.001 Hz at the amplitude of 20 mV. The EIS was then
analysed using the stimulated model of internal circuit of DSSCs to
Fig. 11. Photocurrent density-voltage characteristic of pure TiO2; TrGO-1; TrGO-2; TrGO- identify the resistance of electron charge between each surface mate-
3 and TrGO-4 using N719 dye. rials.
Fig. 14 shows the schematic diagram of the structure of DSSCs.
Accordingly, the series resistance (Rs) represents the sheet resistance
between FTO and photoanode; first semicircle (RCE) represents the
charge transfer resistance between the counter electrode and electrolyte
at high frequency region; second semicircle (RCT) is related to the
charge transfer between the interface of photoanode/dye/electrolyte;
third semicircle represents the response in low frequency region cor-
responding to the resistance and the interface hindrance characteristics
redox I−3/I- electrolyte [37,46]. Tables 4 and 5 tabulate the stimulated
fitted values of Nyquist curve from autolab and the electron lifetime
were obtained from the body phase.
Fig. 15 illustrates the Nyquist plot. It is clear that there were 3
semicircles. The first semicircle was barely visible due to the over-
lapping of the large second semicircle. Based on the results reported in
Tables 4 and 5, the Nyquist plot showed almost similar values of Rs and
RCE. This gave us a hint that the materials for counter electrode and
electrolyte were the same for all devices. For N719 dye, the RCT values
of TiO2, TrGO-1, TrGO-2 and TrGO-3 were reported to be 126.0 Ω,
Fig. 12. Photocurrent density-voltage characteristic of pure TiO2; TrGO-1; TrGO-2; TrGO- 92.3 Ω, 50.2, and 52.3 Ω respectively. Meanwhile, natural green dye
3 and TrGO-4 using Natural Green Chlorophyll dye. recorded 351 Ω, 284 Ω, 130 Ω, 132 Ω for pure TiO2, TrGO-1, TrGO-2
and TrGO-3, respectively. The semicircle arc in Fig. 15 shows that the
rGO into TiO2 photoanode as rGO could provide path to increase the rGO contained DSSCs were decreased except for TrGO-3 compared to
dye molecule absorption on the photoanode surface [37,39]. Also, TiO2 reflecting faster electron transfer between the interface
when the surface of photoanode was illuminated, the ultrafast electron [24,39,43,45,46,49]. These values represent the electrocatalytic ac-
transport could be enhanced with the support of excellent conductivity tivity in the DSSCs, in which the lowest value of the RCT represents the
highest electrocatalytic activity in DSSCs. Taking this fact into account,

273
S.Z. Siddick et al. Materials Science in Semiconductor Processing 74 (2018) 267–276

Fig. 13. DSSCs analysis of all samples; a) current density; b) fill factor; c) voltage short circuit; d) efficiency.

it can be concluded that TrGO-2 nanocomposite recorded the highest

electrocatalytic activity, but the lowest recombination resistance and
electron transmission resistance in the photoanode. Correspondingly, a
new path would be created in the structure of photoanode when rGO
was incorporated into the photoanode. This condition allows the photo-
induced electrons to travel quickly and fill the anode surface up. Ad-
ditionally, rGO helps in interfacial electron transportation and reduces
the electron-hole recombination losses [37,39,41–46]. With optimized
TiO2 dopants in rGO film, the minimal defect in the photoanode con-
Fig. 14. Schematic diagram of DSSCs structure that fitting into resistance, capacitance
and Nyquist.
tributes towards (i) a good surface contact between the semiconductor
and dye, (ii) performance enhancement of DSSCs, (iii) supressed the
recombination losses across the semiconductor, and (iv) charge transfer
Table 4
enhancement between the interfaces [24,37,39,41–46].
EIS analysis of N719 dye DSSCs.
In order to support the EIS analysis, bode plots was used to obtain
Sample Rs(Ω) RCT(Ω) RCE(Ω) Bode peak (Hz) Electron lifetime τn (ms) the information of the lifetime of the electronics transfer and re-
combination of the electron holes in the DSSCs photoanode-electrolyte
Pure TiO2 20.9 126.0 38.6 63.096 2.52 interface. From the bode plot, the electron lifetime (τn ) could be ob-
TrGO-1 21.5 92.3 37.1 19.953 8.38
TrGO-2 20.0 50.2 37.6 10.000 15.92
tained from the peak angular frequency using Eq. (3):
TrGO-3 20.9 52.3 38.2 19.953 8.38
1
τn =
2πfmax (3)
Table 5
EIS analysis of natural green chlorophyll dye DSSCs. where fmax is the peak frequency of bode phase. As shown in Tables 3
and 4, the recorded electron lifetimes for N719 dye were found to be
Sample Rs(Ω) RCT(Ω) RCE(Ω) Bode peak (Hz) Electron lifetime τn (ms) 2.52 ms, 8.38 ms, 16.32 ms and 8.38 ms, while the electron lifetime for
natural green chlorophyll dye were 1.59 ms, 3.18 ms, 6.34 ms and
Pure TiO2 22.0 351 39.4 100.00 1.59
TrGO-1 21.4 284 39.7 50.119 3.18
4.00 ms. From the bode plot, shifting of the characteristic peak into
TrGO-2 20.3 130 37.7 25.119 6.34 lower frequency demonstrate with increase in electron transfer. The
TrGO-3 21.3 132 38.1 39.811 4.00 recombination in the photoanode/dye/electrolyte interface reduced
when there is an increase of electron lifetime [37]. From the tables, the

274
S.Z. Siddick et al. Materials Science in Semiconductor Processing 74 (2018) 267–276

Fig. 15. EIS analysis (A & B) Nyquist plot for commercialize N719 dye and & natural green chlorophyll dye; (B & D) Bode plot for commercialize N719 dye and & natural green chlorophyll
dye.

increase in electron lifetime was TrGO-2 regardless type of dye which 2016), University Malaya Prototype Grant (RU005G-2016), Global
suggests the recombination rate of injected electron have improved and Collaborative Programme- SATU Joint Research Scheme (ST007-2017),
leads to faster diffusion rate, thus increasing Jsc and PCE values PPP Grant University Malaya (PG045-2015B), and Grand Challenge
[37,39,47,48]. Grant (GC002A-15SBS).

4. Conclusion References

In this work, pure TiO2 and TrGO-based photoanode films in DSSCs [1] M. Grätzel, Conversion of sunlight to electric power by nanocrystalline dye-sensi-
tized solar cells, J. Photochem. Photobiol. 164 (2004) 3.
application were explored to improve the DSSCs performances. Based
[2] E. Singh, H.S. Nalwa, Graphene-based dye-sensitized solar cells: a review, Sci. Adv.
on the FTIR and Raman spectroscopy, it was confirmed that the pure Mater. 7 (2015) 1863.
TiO2 and rGO nanocomposite could be successfully synthesized using [3] E. Singh, H.S. Nalwa, Graphene-based bulk-heterojunction solar cells: a review, J.
one step sol precipitation peptization technique. HRTEM and FESEM Nanosci. Nanotechnol. 15 (2015) 6237.
[4] M.K. Nazeeruddin, E. Baranoff, M. Grätzel, Dye-sensitized solar cells: a brief over-
images also revealed the growth of the TiO2 on the RGO surface giving view, Sol. Energy 85 (6) (2011) 1172.
rise to an increase in the light absorption ability of DSSCs. The ab- [5] S. Ito, T.N. Murakami, P. Comte, P. Liska, C. Grätzel, M.K. Nazeeruddin, M. Grätzel,
sorption spectra and tauc plot of the film showed that the band gap in Fabrication of thin film dye sensitized solar cells with solar to electric power con-
version efficiency over 10%, Thin Solid Films 516 (14) (2008) 4613.
TrGO nanocomposites was lower than that of those pure TiO2 film [6] M. Grätzel, Photovoltaic performance and long-term stability of dye-sensitized
considering rGO acts as a bridge for the electron transportation. meosocopic solar cells, C. R. Chim. 9 (5) (2006) 578.
Furthermore, it was determined that the overall efficiency of DSSC was [7] E. Singh, K.S. Kim, G.Y. Yeom, H.S. Nalwa, Two-dimensional transition metal di-
chalcogenide-based counter electrodes for dye-sensitized solar cells, RSC Adv. 7
increased with increasing the rGO content. However, relatively high (2017) 28234.
rGO content in the TiO2/rGO photoanode was undesirable (TrGO-3). [8] S. Ito, T. Saitou, H. Imahori, H. Uehara, N. Hasegawa, Fabrication of dye-sensitized
EIS analysis indicated that TrGO-2 had a smaller charge transfer re- solar cells using natural dye for food pigment: monascus yellow, Energy Environ.
Sci. 3 (2010) 905.
sistance due to the increase in the electron transport and lower re-
[9] I.C. Maurya, P. Srivastava, L. Bahadur, Dye-sensitized solar cell using extract from
combination. Optimizing TrGO film is a must to obtain an efficient petals of male flowers Luffa cylindrica L. as a natural sensitizer, Opt. Mater. 52
photoanode to elevate the overall DSSCs characteristic and to prevent (2016) 150.
[10] S. Shalini, R. Balasundara Prabhu, S. Prasanna, T.K. Mallick, S. Senthilarasu,
any electron leakage due to surface facets induced by the rGO in the
Review on natural dye sensitized solar cells: operation, materials and methods,
TiO2 matrix. Renew. Sustain. Energy Rev. 51 (2015) 1306.
[11] H. Chang, H.M. Wu, T.L. Chen, K.D. Huang, C.S. Jwo, Y.J. Lo, Dye-sensitized solar
cell using natural dyes extracted from spinach and ipomoea, J. Alloy. Compd. 495
Acknowledgments
(2010) 606.
[12] Y.B. Tang, C.S. Lee, J. Xu, Z.T. Liu, Z.H. Chen, Z.B. He, Y.L. Cao, G.D. Yuan,
The authors thankfully acknowledge the financial support by H.S. Song, L.M. Chen, L.B. Luo, H.M. Cheng, W.J. Zhang, I. Bello, S.T. Lee,
University Malaya Research Fund Assistance (BKP) Grant (BK096- Incorporation of graphenes in nanostructured TiO2 films via molecular grafting for

275
S.Z. Siddick et al. Materials Science in Semiconductor Processing 74 (2018) 267–276

dye-sensitized solar cell application, ACS Nano 4 (2010) 3482. oxide using a natural reducing agent, Ceram. Int. 41 (2015) 9505.
[13] K. Novoselov, Nobel lecture: graphene: materials in the flatland, Rev. Mod. Phys. 83 [33] C. Jin, B. Liu, Z. Lei, J. Sun, Structure and photoluminescence of the TiO2 films
(2011) 837. grown by atomic layer deposition using tetrakis-dimethylamino titanium and
[14] Y.W. Zhu, S. Murali, W. Cai, X. Li, J.W. Suk, J.R. Potts, R.S. Ruoff, Graphene and ozone, Nanoscale Res. Lett. 10 (2015) 1.
graphene oxide: synthesis, properties, and applications, Adv. Mater. 22 (2010) [34] C.P.P. Wong, C.W. Lai, K.M. Lee, S.B. Abd. Hamid, Advanced chemical reduction of
3906. reduced graphene oxide and its photocatalytic activity in degrading reactive black
[15] H.J. Choi, S.M. Jung, J.M. Seo, D.W. Chang, L.M. Dai, J.B. Baek, Graphene for 5, Materials 8 (2015) 7118.
energy conversion and storage in fuel cells and supercapacitors, Nano Energy 1 [35] N. Raghavan, S. Thangavel, G. Venugopal, Enhanced photocatalytic degradation of
(2012) 534. methylene blue by reduced graphene-oxide/titanium dioxide/zinc oxide ternary
[16] W.W. Liu, X.B. Yan, Q.J. Xue, Multilayer hybrid films consisting of alternating nanocomposites, Mater. Sci. Semicond. Process. 30 (2015) 321.
graphene and titanium dioxide for high-performance supercapacitors, J. Mater. [36] S.K. Gupta, R. Desai, P.K. Jha, S. Sahoo, D. Kirin, Titanium dioxide synthesized
Chem. C 1 (2013) 1413. using titanium chloride: size effect study using Raman spectroscopy and photo-
[17] C.C. Xiang, M. Li, M.J. Zhi, A. Manivannan, N. Wu, Reduced graphene oxide/tita- luminescence, J. Raman Spectrosc. 41 (2010) 350.
nium dioxide composites for supercapacitor electrodes: shape and coupling effects, [37] G.D. Sharma, D. Daphnomili, K.S.V. Gupta, T. Gayathri, S.P. Singh, P.A. Angaridis,
J. Mater. Chem. 22 (2012) 19161. T.N. Kitsopoulos, D. Tasis, A.G. Coutsolelo, Enhancement of power conversion ef-
[18] N.M. Huang, H.N. Lim, C.H. Chia, M.A. Yarmo, M.R. Muhamad, Simple room- ficiency of dye-sensitized solar cells by co-sensitization of zinc-porphyrin and
temperature preparation of high-yield large-area graphene oxide, Int. J. Nanomed. thiocyanate-free ruthenium(II)-terpyridine dyes and graphene modified TiO2 pho-
6 (2011) 3443. toanode, RSC Adv. 3 (2013) 22412.
[19] P. Afzali, Y. Abdi, E. Arzi, Gated graphene/titanium dioxide-based photodetector. , [38] F.W. Low, C.W. Lai, S.B. Abd Hamid, Surface modification of reduced graphene
J. Nanopart. Res. 16 (2014) 2695. oxide film by Ti ion implantation technique for high dye-sensitized solar cells
[20] L. Li, H. Duan, X. Wang, C. Luo, Adsorption property of Cr(VI) on magnetic me- performance, Ceram. Int. 43 (2017) 625.
soporous titanium dioxide-graphene oxide core-shell microspheres, New J. Chem. [39] A. Sacco, S. Porro, A. Lamberti, M. Gerosa, M. Castellino, A. Chiodoni, S. Bianco,
38 (2014) 6008. Investigation of transport and recombination properties in graphene/titanium di-
[21] C. Huang, C. Li, G. Shi, Graphene based catalysts, Energy Environ. Sci. 5 (2012) oxide nanocomposite for dye-sensitized solar cell photoanodes, Electrochim. Acta
8848. 131 (2014) 154.
[22] A.K. Geim, Graphene: status and prospects, Science 324 (2009) 1530. [40] F.A. Jumeri, H.N. Lim, Z. Zainal, N.M. Huang, A. Pandikumar, S.P. Lim, Dual
[23] Q. Xiang, J. Yu, M. Jaroniec, Graphene-based semiconductor photocatalysts, Chem. functional reduced graphene oxide as photoanode and counter electrode in dye-
Soc. Rev. 41 (2012) 782. sensitized solar cells and its exceptional efficiency enhancement, J. Power Sources
[24] Y. Zhou, L. Chen, W. Tu, Z. Li, C. Bao, H. Dai, T. Yu, J.G. Liu, Z. Zou, Enhanced 293 (2015) 712.
photovoltaic performance of a dye-sensitized solar cell using graphene-TiO2 pho- [41] J. Zhi, H. Cui, A. Chen, Y. Xie, F. Huang, Efficient highly flexible dye sensitized solar
toanode prepared by a novel in situ simultaneous reduction-hydrolysis technique, cells of three dimensional graphene decorated titanium dioxide nanoparticles on
Nanoscale 5 (2013) 3481. plastic substrate, J. Power Sources 281 (2015) 404.
[25] L. Karimi, M.E. Yazdanshenas, R. Khajavi, A. Rashidi, M. Mirjalili, Using graphene/ [42] J.D. Roy-Mayhew, I.A. Aksay, Graphene materials and their use in dye-sensitized
TiO2 nanocomposite as a new route for preparation of electroconductive, self- solar cells, Chem. Rev. 114 (2014) 6323–6348.
cleaning, antibacterial and antifungal cotton fabric without toxicity, Cellulose 21 [43] F.W. Low, C.W. Lai, S.B. Abd Hamid, Facile synthesis of high quality graphene oxide
(2014) 3813. from graphite flakes using improved hummer s technique, J. Nanosci. Nanotechnol.
[26] J.C. Yu, L. Zhang, Z. Zheng, J. Zhao, Synthesis and characterization of phosphated 15 (2015) 6769.
mesoporous titanium dioxide with high photocatalytic activity, Chem. Mater. 15 [44] J. Fan, S. Liu, J. Yu, Enhanced photovoltaic performance of dye-sensitized solar
(2003) 2280. cells based on TiO 2 nanosheets/graphene composite films, J. Mater. Chem. 22
[27] C.W. Lai, F.W. Low, S.W. Chong, C.P.P. Wong, S.Z. Mohamed Siddick, J.C. Juan, (2012) 17027.
S.B. Abdul Hamid, An overview: recent development of titanium dioxide loaded [45] X. Fang, M. Li, K. Guo, Y. Zhu, Z. Hu, X. Liu, B. Chen, X. Zhao, Improved properties
graphene nanocomposite film for solar application, Curr. Org. Chem. 19 (2015) of dye-sensitized solar cells by incorporation of graphene into the photoelectrodes,
1882. Electrochim. Acta 65 (2012) 174.
[28] L. Zhang, S.M. Yuan, X.J. Lu, Amperometric nonenzymatic glucose sensor based on [46] M. Zhu, X. Li, W. Liu, Y. Cui, An investigation on the photoelectrochemical prop-
a glassy carbon electrode modified with a nanocomposite made from nickel(II) erties of dye-sensitized solar cells based on graphene–TiO2 composite photoanodes,
hydroxide nanoplates and carbon nanofibers, Microchim. Acta 181 (2014) 365. J. Power Sources 262 (2014) 349–355.
[29] F.W. Low, C.W. Lai, S.B. Abd Hamid, Study of reduced graphene oxide film in- [47] J. Zhao, J. Wu, F. Yu, X. Zhang, Z. Lan, J. Lin, Improving the photovoltaic per-
corporated of TiO2 species for efficient visible light driven dye-sensitized solar cell, formance of cadmium sulfide quantum dots-sensitized solar cell by graphene/ti-
J. Mater. Sci. Mater. Electron. 28 (2017) 3819. tania photoanode, Electrochim. Acta 96 (2013) 110.
[30] O. Frank, M. Zukalova, B. Laskova, J. Kürti, J. Koltai, L. Kavan, Raman spectra of [48] S.W. Chong, C.W. Lai, S.B. Abd Hamid, Controllable electrochemical synthesis of
titanium dioxide (anatase, rutile) with identified oxygen isotopes (16, 17, 18), Phys. reduced graphene oxide thin-film constructed as efficient photoanode in dye-sen-
Chem. Chem. Phys. 14 (2012) 14567. sitized solar cells, Materials 9 (2) (2016) 69.
[31] I. Ahmad, J.E. McCarthy, A. Baranov, Y. Gun'ko, Development of graphene nano- [49] T.T. Wu, J.M. Ting, Bridging TiO2 nanoparticles using graphene for use in dye-
platelet based counter electrodes for solar cells, Materials 8 (2015) 5953. sensitized solar cells, Int. J. Energy Res. 38 (2014) 1438.
[32] S.W. Chong, C.W. Lai, S.B. Abd. Hamid, Green preparation of reduced graphene

276