Received: 4 November 2018 Revised: 5 June 2019 Accepted: 12 June 2019

DOI: 10.1002/er.4710

RESEARCH ARTICLE

Synthesis of TiO2 nanoparticles at low hydrothermal

temperature and its performance for DSSC sensitized using
natural dye extracted from Melastoma malabathricum L.
seeds

Nofrijon Sofyan1 | Aga Ridhova2 | Akhmad H. Yuwono1 | Arief Udhiarto3 |

Jeffrey W. Fergus4

1
Department of Metallurgical and
Summary
Materials Engineering, Universitas
Indonesia, Depok 16424, Indonesia Efficiency of a dye‐sensitized solar cell (DSSC) device depends on its semicon-
2
Research Center for Metallurgy and ductor layer and the sensitizing dye to absorb the light. This work seeks to
Materials, Indonesian Institute of obtain the best solvent for the natural dye extraction from Melastoma
Sciences, Tangerang Selatan, Banten
15314, Indonesia malabathricum L. seeds. The extracted dye is used as sensitizer on TiO2 nano-
3
Department of Electrical Engineering, particles produced via hydrothermal but optimized at relatively low tempera-
Universitas Indonesia, Depok 16424, ture. Infrared characterization of the extracted dyes showed differences in
Indonesia
4
functional groups using different solvents, whereas ultraviolet visible examina-
Materials Research and Education
Center, Auburn University, Auburn,
tion of the dyes showed differences in intensity along the spectrum ranges of
Alabama 36849, USA 600 to 400 nm with maximum absorption around 550 to 500 nm. Thermal anal-
ysis revealed that the natural dye should be stable around room temperature.
Correspondence
Nofrijon Sofyan, Faculty of Engineering, Analysis on the synthesized TiO2 nanoparticles showed that the average crys-
Department of Metallurgical and tallite size reported in the previous work is consistent with crystallite sizes
Materials Engineering, Universitas
observed in the transmission electron microscope images. Photoactivity exam-
Indonesia, Depok 16424, Indonesia.
Email: nofrijon.sofyan@ui.ac.id ination showed that the DSSC sensitized using natural dye extracted with eth-
anol containing 20% distilled water on TiO2 synthesized at 150°C has an
Funding information
Ministry of Research, Technology, and
efficiency of 5.7%, whereas the one on commercial TiO2 P25 Degussa has an
Higher Education, Republic of Indonesia, efficiency of 3.0%. The DSSC device sensitized using commercial dye on TiO2
Grant/Award Number: WCP No. 123.49/ synthesized at 150°C has an efficiency of 4.4%, whereas the one on TiO2 P25
D2.3/KP/2018 123.49/D2.3/KP/2018;
World Class Professor Program 2018
Degussa has an efficiency of 4.0%. This result is promising for further develop-
ment of the DSSC device using TiO2 nanoparticles synthesized at low hydro-
thermal temperature and sensitized with the natural dye.

KEYWORDS
dye‐sensitized solar cell, Melastoma malabathricum L., natural dye, power conversion efficiency,
solvent extraction

1 | INTRODUCTION photocatalyst in environmental purification,2,3 and elec-

trochemical devices.4 The discovery that TiO2 can be used
The use of titanium dioxide (TiO2) in wide range of as a semiconductor in the development of dye‐sensitized
applications has been well‐known including sensors,1 solar cell (DSSC)5 has driven even more intensive research
Int J Energy Res. 2019;43:5959–5968. wileyonlinelibrary.com/journal/er © 2019 John Wiley & Sons, Ltd. 5959
5960
and led investigators to find a rapid and cheap way to syn-
thesize TiO2 nanoparticles. One of the methods to synthe-
size TiO2 nanoparticles is through hydrothermal method6;
however, the process temperature is relatively high.7 To
further reduce the cost of the synthesis process, prepara-
tion of TiO2 nanoparticles at lower temperature would be
preferable.
One of the areas being developed for renewable energy
source is DSSC. The working principle of this type of solar
cell resembles the process of photosynthesis through the
photo‐electrochemical principle.5 The main problem of
this DSSC is due to its low efficiency as compared with
the established technology such as silicon‐based solar cell.
As of early 2019, the recorded maximum efficiency for the
DSSC cell is 11.9 % as compared with 26.7% in Si crystalline
cell or 29.1 % in GaAs thin film.8 There are many factors
that affect the efficiency of a DSSC device. These factors
include material conductivity, crystallinity, thickness, sur-
face area of the cell, dye sensitivity, counter electrode, and
electrolyte.9 In terms of the sensitizing dye, the most com-
mon dyes used as sensitizer are usually synthetic dyes
based on complexes of ruthenium dyes.10,11 The price of
these synthetic dyes are relatively expensive, so alternative
dyes are needed to further reduce the production cost. The
most promising option is to use natural dyes, which are
commonly available in nature by harnessing pigments,
such as natural anthocyanin.10 Several investigators have
sought to use natural dye as sensitizer in the DSSC device.
For example, betanin and betalain have been extracted
from beet root with an efficiency of 2.7% and 0.4%, respec-
tively.12,13 Other investigators have used natural dye from
monascus yellow with an efficiency of 2.3%.14
One of the natural dye sources is Melastoma
malabathricum L., which is a purple‐bluish fruit flowering
plant that can easily grow as shrubs in the wild and or
weeds in plantation. This plant has fruit with a berry‐like
capsule containing numerous seeds coated with red sweet
astringent pulp that can be easily extracted to produce a
natural dye via solvent extraction. Because it can easily
FIGURE 1 Melastoma malabathricum L. plant (A) and evolution of M. malabathricum L. flower from bud to fruit (B) [Colour figure can
be viewed at wileyonlinelibrary.com]
SOFYAN ET AL.
grow and become weeds, this plant is usually unwanted
in any plantation. However, instead of removing it as a
weed, it can be potentially grown as a farm for its natural
dye. Some investigators have explored this natural dye for
several purposes. For example, Aziz et al have extracted
the dye from this plant using methanol for use as an anti-
corrosion coating, and they have found that the extracted
dye has high resistance to oxidation.15 Singh et al have also
extracted the dye from this plant using distilled water and
pure ethanol and used it as a sensitizer in a DSSC device.
The efficiency of the device, however, is still low, 1.37%
and 0.72 % for dyes extracted using distilled water and eth-
anol, respectively.16
There are two objectives of this work. The first is to
examine the characteristics of visible light absorbed by
the natural dye extracted from M. malabathricum L. seeds
using various solvents and performance of the resulting
dye as a sensitizer in the DSSC device. The second is to
examine the properties of TiO2 nanoparticles produced
via hydrothermal method but optimized at relatively
low temperature synthesis17 as a semiconductor layer in
the DSSC device.
2 | R ESEAR CH M ETHODOLOGY
2.1 | Dye extraction
M. malabathricum L. ripe fruits were taken from the wild
near Mount Bungsu, West Sumatra and were separated
from the stalks. Figure 1A shows the picture of M.
malabathricum L. plant, and Figure 1B shows its evolu-
tion from the bud to the fruit. Different from the work
of Singh et al,16 in this work, the seeds were taken out
of the fruits and dried under indirect sunlight for 1 week
to remove the excess moisture. The dried fruits were
crushed into pieces in a mortar and moved to a glass bea-
ker. Mixtures of 10 g of dried crushed seeds and 100 mL
of four different solvents, ie, ethanol absolute (Merck),
SOFYAN ET AL.
ethanol containing 20% distilled water, ethanol contain-
ing 1% acetic acid (99.7%, Sigma‐Aldrich), and ethanol
containing 1% HCl (37 %, Merck), were placed in four
separate beakers and agitated using a magnetic stirrer at
an ambient temperature for 12 hours. The dye solution
was then filtered from the seeds residue and stored in a
dark glass bottle for the next use. The steps in this natural
dye extraction process are shown in Figure 2.
The product of extraction was characterized using an
attenuated total reflection Fourier transform infrared
spectroscopy (ATR FTIR, PerkinElmer Spectrum 2),
ultraviolet‐visible spectroscopy (UV‐vis, PerkinElmer
Lambda 25), and differential scanning calorimetry (DSC,
Perkin Elmer STA 6000). As a comparison, a commercial
organic dye (RK1, Solaronix) was similarly characterized.
The commercial dye RK1 is an organic pigment with
chemical formula C46H40N4O2S2. The chemical structure
of this dye is given in Figure 3.18
2.2 | Synthesis of TiO2
The synthesis of TiO2 followed the procedure described in
literature.6 Slight modification to the procedure was car-
ried out to lower the hydrothermal temperature process
as described in our previous work17 and is explained as
following. A volume of 40.5‐mL titanium tetra
isopropoxide (TTIP) was mixed with 112.5 mL of distilled
water under a magnetic stirrer at 15°C for 1 hour. The
precipitates were filtered and washed with 50 mL of dis-
tilled water before being transferred to tetra methyl
ammonium hydroxide (TMAH) solution. The mixture
was refluxed at 100°C for 4 hours. The resulting colloids
were hydrothermally treated at three temperature varia-
tions 100°C, 125°C, and 150°C for 12 hours in a Teflon
autoclave. The suspension was centrifuged at 5000 rpm
to obtain TiO2 nanoparticles. The obtained TiO2 nano-
particles were calcined at 450°C for 2 hours before being
characterized using X‐ray diffraction (XRD, ARL OPTX‐
2050) for the crystal structure and field emission scanning
electron microscope (FE SEM, FEI Inspect F50) for sur-
face morphology. The results have been reported in our
previous work.17 In this current article, additional
FIGURE 2 Solvent extraction from Melastoma malabathricum L [Colour figure can be viewed at wileyonlinelibrary.com]
5961
FIGURE 3 The chemical structure of RK1 organic dye
information on the high‐resolution imaging using trans-
mission electron microscope (TEM, FEI Tecnai G2 S‐
Twin) is included.
2.3 | Device fabrication
The DSSC device was fabricated as described in our previ-
ous work19 and is explained as following. Semiconductor
layer was prepared by adding an amount of methanol into
TiO2 nanoparticles along with two drops of TRITON X‐100
(Sigma Aldrich). The blend was mixed to obtain a homoge-
neous phase. Two conductive electrodes made of fluorine‐
doped tin oxide (FTO) glasses were prepared and ultrason-
ically cleaned. One was used as the anode for deposition of
semiconductor material. The other FTO glass was used for
deposition of counter electrode using platinum paste
(Platisol T/SP, Solaronix). After the deposition, the anode
was soaked in the dye solution for 12 hours and then rinsed
with ethanol. The two FTO electrodes were then put
together and filled with the electrolyte, which is composed
of iodide/tri‐iodide (I−/I3−) solution (Iodolyte, Solaronix).
The DSSC device activities were recorded and character-
ized using electrochemical impedance spectroscopy (Hioki
3522‐50) under room light condition and semiconductor
parameter analyzer at a standard illumination of 100
mW/cm2 (HP Agilent 4145B).
3 | R ESULTS A ND DISCUSSIONS
The as‐synthesized TiO2 nanoparticles characterized using
XRD confirmed to be anatase and has been reported in our
previous work.17 Slightly different from our previous
report, however, there is an indication of small brookite
5962
peak around 31.6°. The brookite phase has also been
observed by Zhao et al, who reported that brookite could
be produced using hydrothermal method.20 According to
Xu et al, the formation of brookite phase is influenced by
chemical composition, pressure, and temperature.21 In
this work, it seems that during the hydrothermal synthesis,
the reaction has not totally produced the anatase phase;
however, the dominant factors for this brookite formation
have not been determined. The average crystallite size
determined using XRD peak widths reported in our previ-
ous work17 are consistent with crystallite sizes observed in
the TEM images given in Figure 4.
The dye obtained from the extraction and the interac-
tion between the dye and TiO2 after sensitizing were con-
firmed for its functional groups using infrared
spectroscopy. This was performed to confirm the struc-
ture of the anthocyanin. Further, the absorption within
the visible light range, the range where the absorption
could take place, was also confirmed using UV‐vis spec-
troscopy. The results for the functional groups are given
in Figure 5, the schematic illustration of the basic chem-
ical structure of anthocyanin and its interaction mecha-
nism possibility with TiO2 nanoparticles are given in
Figure 6, whereas the UV‐vis absorption spectra are pre-
sented in Figure 7.
Figure 5A shows the infrared spectra of commercial
sensitizer RK1 and the natural dye extracted from M.
malabathricum L. using various solvents. The spectra
are similar with only little differences in the C═O stretch
vibration group around 1650 cm−1. Details of the identi-
fied functional groups that correspond to various absorp-
tion bands are summarized in Table 1. The wavenumbers
of the functional groups indicate that the extracted dyes
are anthocyanin. Figure 5B shows infrared spectra of
the sensitized P25 TiO2 nanoparticle with the extracted
dyes, whereas Figure 5C shows infrared spectra of the
synthesized TiO2 sensitized with the dye extracted using
the mixture of ethanol and distilled water and pure P25
TiO2 nanoparticle with no sensitization as a comparison.
FIGURE 4 Transmission electron microscope (TEM) images of TiO2 nanoparticles synthesized at (A) 100°C, (B) 125°C, and (C) at 150°C.
Inset is selected area diffraction of the image showing single crystalline of TiO2
SOFYAN ET AL.
This was performed to examine the interaction between
the natural dye functional groups and TiO2 after sensitiz-
ing. As can be seen in Figure 5B, the main pattern of P25
TiO2 nanoparticles sensitized with the natural dye extract
is similar to the graph shown in Figure 5A but with low
intensity and the presence of Ti–O bond around 400 to
600 cm−1.16,22 An anomaly that is yet to be further con-
firmed occurs for the P25 TiO2 sensitized using the dye
extracted with ethanol in which its spectrum has higher
intensity compared with the other spectra. The carboxyl
and hydroxyl group in natural dye are expected to bond
with TiO2 nanoparticles as the anode in DSSC.23 The
interaction between TiO2 nanoparticles and the func-
tional group of the dye, when exposed to sunlight, will
trigger the electron transfer from the dye molecules into
the conduction band of TiO2.24
Comparison of the absorption spectra obtained from
UV‐vis spectroscopy for commercial sensitizer and natural
dye extracted using various solvents is presented in
Figure 7A. Absorption spectra for the extracted dyes sensi-
tized synthesized TiO2 and the extracted dyes sensitized
TiO2 P25 Degussa are given in Figure 7B,C, respectively.
As can be seen in Figure 7A, the maximum absorption
spectrum for commercial sensitizer RK1 lies in between
500 and 450 nm with an intensity of 1.3%, while all of
the natural dye extracts have almost the same intensity
along the spectrum range of 600 to 400 nm with maximum
absorption spectrum around 550 to 500 nm and intensity
around 0.5% to 1.0%. The absorption spectral ranges of
the extracted dyes are wider than that of RK1. This wide
range of absorption band will increase the ability of natu-
ral dye extracts to absorb more photon energy over a wide
spectrum, which will increase the production of photo-
electron.16 An increase in the number of photoelectrons
produced by the extracted dye in the visible light range will
increase the efficiency of DSSC device. The lower absor-
bance in the near UV region would be beneficial for
high‐light intensity but not for the low‐light condition.25
In Figure 7B, the absorption spectrum characteristic of
SOFYAN ET AL. 5963

FIGURE 5 Superimposed of infrared spectra: (A) commercial sensitizer RK1 and the extracted dyes; (B) natural dye sensitized TiO2 P25;
and (C) synthesized TiO2 sensitized using the natural dye extracted with ethanol containing 20% distilled water and TiO2 P25 with no
sensitization [Colour figure can be viewed at wileyonlinelibrary.com]

maxima of the semiconductor TiO2 P25 Degussa after sen-

FIGURE 6 (A) The basic chemical structure of anthocyanin and
(B) the possibility of its interaction mechanism with TiO2
nanoparticles
the extracted dye sensitized synthesized TiO2 is revealed.
It can be seen that the absorption range of sensitized syn-
thesized TiO2 follows the absorption range of pure
extracted dye in Figure 7A. The absorption of pure synthe-
sized TiO2 and commercial dye sensitized synthesized
TiO2, however, is much lower as compared with the
extracted dye sensitized synthesized TiO2. The same is
true for the natural dye sensitized commercial TiO2 P25
Degussa in Figure 7C. It seems that the commercial dye
and natural extracted dyes cannot increase the absorption
sitization, except the one sensitized using the natural dye
extracted using ethanol. The reason for this phenomenon
could be because of the phases present in the structure,
ie, anatase and rutile.17 Another reason could be because
of dye stability in the mixed phase; however, further work
would be needed to confirm these reasons.
Thermal stability of an extracted dye is important since
the unstable dye will degrade easily. Differential thermal
calorimetry results for the extracted dyes are given in
Figure 8, which show that the endothermic peak occurred
at a temperature around 120°C with dye extracted using
pure ethanol and ethanol with distilled water and at 100
to 120°C with dye extracted using ethanol with addition
of HCl and acetic acid. This endothermic activity indicates
that the evaporation of residual water from the sensitizer
occurred at that particular temperature. The differences
in the magnitudes of the endothermic peaks are due to
the difference of water content in solvent with the addi-
tion of acid. The solvent without acid has higher water
5964 SOFYAN ET AL.

FIGURE 7 Ultraviolet‐visible (UV‐vis) spectra of (A) commercial sensitizer RK1 and the extracted dyes; (B) extracted dyes sensitized
synthesized TiO2; (C) extracted dyes sensitized TiO2 P25 Degussa [Colour figure can be viewed at wileyonlinelibrary.com]

TABLE 1 Identified functional groups of the extracted dyes using

various solvents, sensitized P25 TiO2, sensitized synthesized TiO2
using the ethanol‐H2O extracted dye, and pure P25 TiO2

Wavenumbers, Functional
Specimen cm−1 Groups

Extracted dyes 3150‐3500 Alcohol OH stretch

2950‐3000 ─CH stretch
1650 C═O stretch vibration
1500‐1250 Aromatic
1085 C─OH stretch
1050 C─O─C stretch
vibration
875 Aldehyde
Sensitized TiO2 P25 3247 ─OH
1607 C═O stretch
FIGURE 8 Thermal characteristic of the extracted dyes using
vibrations
various solvents [Colour figure can be viewed at
1035 C─O─C stretch
wileyonlinelibrary.com]
vibration
400‐600 Ti─O
Sensitized synthesized 3352 ─OH content as compared with solvent containing acid. Since
TiO2 1635 C═O stretch
all the changes occur above 100°C, the dye should be sta-
vibrations
ble when used at around room temperature.
400‐600 Ti─O
Figure 9 shows the graphical conversion for absorption
Pure P25 400‐600 Ti─O
fitting spectrum (AFS) method of UV‐vis spectroscopy
SOFYAN ET AL. 5965

results from TiO2 P25 Degussa and synthesized TiO2 at TABLE 2 Band gap energy value of P25 from AFS method with
hydrothermal temperature variations sensitized using solvent variations and synthesized TiO2 at hydrothermal tempera-
the extracted dyes. Band gap energy values for each of ture variations
the material calculated from graphical extrapolation are Sample 1/λ (1/nm) Eg, eV
shown in Table 2, which shows that the band gap energy
P25 + Ethanol 0.00270 3.347
values for each sample sensitized with the extracted dyes
are higher than that of TiO2 nanoparticles from P25 P25 + Ethanol‐H2O 0.00265 3.285
Degussa. The addition of dyes to TiO2 nanoparticles P25 + Ethanol‐HCl 0.00260 3.223
seems to change the band gap energy slightly, which P25 + Ethanol‐acetic acid 0.00280 3.471
could be due to the interaction between TiO2 and the dif- P25 + RK1 0.00245 3.037
ferent functional groups of the dyes.26 The exact reason o
150 C + RK1 0.00258 3.198
for this change, however, is yet to be further confirmed.
o
Electrochemical study was performed to characterize 125 C + RK1 0.00270 3.347
o
the electrochemical characteristic of the DSSC device. 100 C + RK1 0.00285 3.534
The study was performed on the DSSC sensitized with Abbreviation: AFS, absorption fitting spectrum.
the dye extracted using the solvent of ethanol containing
20 % distilled water only on the TiO2 hydrothermally syn- sensitized using the natural dye, the commercial dye on
thesized at 150°C and the commercial TiO2 P25 Degussa. the TiO2 P25 Degussa, and the natural dye on the com-
For comparison, the results from the one sensitized using mercial TiO2 P25 Degussa. High value of a semicircle
the commercial dye RK1 are also given. The results from radius indicates large value of the charge transfer resis-
this electrochemical examination are given in Figure 10. tance.27 The same trends can be observed in the Bode
The electrochemical characteristics of the DSSC device phase diagram in Figure 10B and the Bode magnitude
are represented in the forms of impedance spectra of diagram in Figure 10C. An anomaly, however, is found
Nyquist, Bode phase, and Bode magnitude diagrams in the frequency above 100 Hz in which the tail in the
given in Figure 10A‐C, respectively. The filled circles are Bode magnitude for both RK1 and natural dye sensitized
the experimental data, whereas the solid lines are the TiO2 P25 Degussa have higher |Z| value as compared with
fitted results. As indicated by the Nyquist plot in those RK1 and natural dye sensitized synthesized TiO2.
Figure 10A, the electrochemical characteristics of all In this instance, at the frequency above 100 Hz, the mod-
DSSC devices show a similar semicircle representing the ulus of TiO2 P25 Degussa is higher than the synthesized
interface resistance of electrons but with a different value TiO2. There is a possibility that the difference in the crys-
indicated by the differences in the semicircle radius. In tal structure affects the modulus characteristic of TiO2
the descending order, the DSSC device sensitized using P25 Degussa as compared with the synthesized TiO2. It
commercial dye RK1 on TiO2 synthesized at 150oC has has been shown that TiO2 P25 Degussa has mixed ana-
the largest semicircle radius followed by the one tase and rutile phases,17 whereas the synthesized TiO2

FIGURE 9 Absorption fitting spectrum (AFS) method for TiO2 P25 Degussa nanoparticle sensitized using the extracted dyes (A) and
synthesized TiO2 at hydrothermal temperature variations (B) [Colour figure can be viewed at wileyonlinelibrary.com]
5966 SOFYAN ET AL.

FIGURE 10 Impedance spectra of the dye‐sensitized solar cell (DSSC) device measured under room light condition: (A) Nyquist
representation; (B) Bode phase; and (C) Bode magnitude [Colour figure can be viewed at wileyonlinelibrary.com]

contains almost pure anatase. This phenomenon is yet to

be further confirmed, however, as it would apparent in
the photoactivity characterization explained in the fol-
lowing, it has an impact on the power conversion effi-
ciency (PCE) of the DSSC device.
As for the electrochemical characterization, the
photoactivity characteristic was examined only for the
DSSC sensitized with the dye extracted using the solvent
of ethanol containing 20% distilled water on the TiO2
hydrothermally synthesized at 150°C and the commercial
TiO2 P25 Degussa and the one sensitized using the com-
mercial dye RK1 for comparison. The results in the forms
of J‐V curves are given in Figure 11. The resulting param- FIGURE 11 J‐V curve characteristics of the dye‐sensitized solar
eter for PCE from each of the dyes was calculated and is cell (DSSC) device fabricated from synthesized TiO2 and TiO2 P25
shown in Table 3. The PCE obtained from the use of nat- Degussa sensitized using commercial dye and the one extracted
ural dye extracted using ethanol containing 20% distilled from Melastoma malabahricum L. with ethanol containing 20% of
water on TiO2 hydrothermally synthesized at tempera- distilled water [Colour figure can be viewed at wileyonlinelibrary.
ture 150°C (5.7%) is higher than the PCE from the use com]
of natural dye extracted using ethanol containing 20% dis-
TABLE 3 Performance characteristics of the DSSC device
tilled water on commercial TiO2 P25 Degussa (3.0%) as
well as the PCE obtained from the use of commercial Sample VOC, V JSC, mA/cm2 FF PCE, %
dye on TiO2 hydrothermally synthesized at temperature o
150 C + Ethanol‐H2O 0.50 8.18 0.14 5.69
150°C (4.4 %) and the PCE from the use of commercial
P25 + Ethanol‐H2O 0.27 10.02 0.11 3.00
dye on commercial TiO2 P25 Degussa (4.0 %).
o
The achieved PCE can be related to the parameters 150 C + RK1 0.45 9.97 0.10 4.40
a
used in this work, ie, different solvents for the dye extrac- P25 + RK1 0.48 10.28 0.08 4.02
tion and temperature for the TiO2 synthesis. Based on the Abbreviation: DSSC, dye‐sensitized solar cell.
UV‐vis characterization results, commercial dye has the a
Sofyan et al.19
highest absorbance value, whereas the extracted dye with
various solvent has wide range of absorption but with low ethanol, and ethanol with acetic acid from the highest to
absorbance value as compared with that of RK1 commer- the lowest, but the band gap alone does not determine
cial dye. The bandgap energy increases with the use of sol- the PCE because the highest PCE was achieved by the
vent ethanol and HCl, ethanol and distilled water, pure mixture of ethanol and water. In this case, even though
SOFYAN ET AL.
the addition of acid into the solvent used for the extrac-
tion may increase the anthocyanin recovery,28 it may also
promote the disintegration of the semiconductor used in
the DSSC and thus reduce the efficiency.
In the case of synthesis temperature, the best PCE was
obtained from the temperature of 150°C. This would be
understandable since the higher the temperature, the bet-
ter the chance of the atoms to rearrange themselves and
form the crystalline phase as characterized through the
crystallite differences shown in the TEM images previ-
ously. DSSC fabricated using commercial P25 Degussa
has lower efficiency when compared with that of TiO2
hydrothermally synthesized at 150°C. In terms of electro-
chemical characteristic, even though charge transfer resis-
tance of commercial RK1 and natural dyes sensitized
synthesized TiO2 is higher than that of TiO2 P25 Degussa,
however, above 100 Hz, modulus resistance of TiO2 P25
Degussa is higher than that of synthesized TiO2. This dif-
ference is expected to be due to differences in the crystal
structure of both nanoparticles in which the synthesized
TiO2 contains almost single phase of anatase crystal struc-
ture, but commercial P25 Degussa contains a mixed phase
of rutile and anatase crystal structures.17 The anatase crys-
tal structure has high energy‐band‐gap and electron affin-
ity as compared with the rutile crystal structure.29
In terms of the sensitizing dye, the finding agrees with
other results19 in which the PCE of the DSSC device sen-
sitized using the dye extracted using ethanol containing
20% distilled water is higher than that of pure ethanol.
The current PCE is also higher than the PCE obtained
by Singh et al.16 Although the difference is yet to be fur-
ther confirmed, it is expected that the difference in the
process during the extraction could result in the differ-
ence in the obtained dye. This difference could further
affect the dye sensitizing activity toward TiO2 nanoparti-
cles and thus the performance of the DSSC device. Other
investigators have also obtained quite different results
from the use of natural dyes. For example, the dye
extracted from beet root in the forms of betanin and
betailain resulted in two different efficiency, 2.7% and
0.4%, respectively.12,13 Other investigators have obtained
an efficiency as high as 2.3% from the natural dye
monascus yellow.14 It is undeniable that further efforts
are needed to reduce this variability since different natu-
ral resources could result in different efficiency.
4 | CONCLUSION
The use TiO2 nanoparticles synthesized at relatively low
hydrothermal temperature and sensitized with natural
dye extracted from M. malabathricum L. in a DSSC device
has been successfully realized. The characterization
5967
results from FTIR, UV‐vis, and DSC of the dye extracted
using different solvents showed that there are differences
amongst the various solvents. The use of ethanol as solvent
as well as the addition of various acid and distilled water
showed differences in transmittance and absorbance
values of the extracted dye. The absorbance peak from
each extracted dye showed wavelength of 500 to 550 nm.
The highest PCE of the DSSC device (5.69 %) was obtained
from TiO2 prepared through hydrothermal synthesis at
150°C and sensitized using the dye extracted with ethanol
containing 20% distilled water. This result is promising for
further development of the DSSC device using natural sen-
sitizer. For the future research work, since the current
result only characterized the extracted dye qualitatively,
other tools to quantitatively characterize the results could
be useful. Purification of the extracted dye could be also a
main challenge. It would be also interesting to examine
the stability of the dye during the lifetime of the DSSC
device. Another challenging issue could be on enlarging
the scale, both for harnessing the natural dye and optimiz-
ing the low temperature synthesis process for TiO2, espe-
cially if these could be developed in industrial scale.
ACKNOWLEDGEMENTS
The authors acknowledge the funding from the World
Class Professor Program 2018, Ministry of Research,
Technology, and Higher Education, Republic of Indonesia
No. 123.49/D2.3/KP/2018. The authors would like also to
thank Mr. Achmad Subhan from Physics Research Cen-
ter, Indonesian Institute of Sciences, for the EIS work.
ORCID
Nofrijon Sofyan https://orcid.org/0000-0001-7814-9022
Aga Ridhova https://orcid.org/0000-0001-5619-4340
Akhmad H. Yuwono https://orcid.org/0000-0001-9887-
6446
Arief Udhiarto https://orcid.org/0000-0001-8861-1129
Jeffrey W. Fergus https://orcid.org/0000-0002-8067-1992
RE FER EN CES
1. Maziarz W, Kusior A, Trenczek‐Zajac A. Nanostructured TiO2‐
based gas sensors with enhanced sensitivity to reducing gases.
Beilstein J Nanotechnol. 2016;7:1718‐1726. https://doi.org/
10.3762/bjnano.7.164
2. Binas V, Venieri D, Kotzias D, Kiriakidis G. Modified TiO2 based
photocatalysts for improved air and health quality. J Materiomics.
2017;3(1):3‐16. https://doi.org/10.1016/j.jmat.2016.11.002
3. Ma D, Liu A, Li S, Lu C, Chen C. TiO2 photocatalysis for C–C
bond formation. Cat Sci Technol. 2018;8(8):2030‐2045. https://
doi.org/10.1039/C7CY01458A
5968
4. Zu G, Li H, Liu S, Li D, Wang J, Zhao J. Highly efficient mass
determination of TiO2 nanotube arrays and its application in
lithium‐ion batteries. Sustain Mater Technol. 2018;17:e00079.
https://doi.org/10.1016/j.susmat.2018.e00079
5. Grätzel M. Photoelectrochemical Cells. Nature. 2001;414(6861):
338‐344. https://doi.org/10.1038/35104607
6. Rehan M, Lai X, Kaleb GK. Hydrothermal synthesis of titanium
dioxide nanoparticles studied employing in situ energy disper-
sive X‐ray diffraction. CrstEngComm. 2011;13(11):3725‐3732.
https://doi.org/10.1039/C0CE00781A
7. Mehnane HF, Wang C, Kondamareddy KK, et al. Hydrothermal
synthesis of TiO2 nanoparticles doped with trace amounts of
strontium, and their application as working electrodes for dye
sensitized solar cells: tunable electrical properties & enhanced
photo‐conversion performance. RSC Adv. 2017;7(4):2358‐2364.
https://doi.org/10.1039/C6RA26012H
8. Green MA, Hishikawa Y, Dunlop ED, et al. Solar cell efficiency
tables (Version 53). Prog Photovolt Res Appl. 2019;27(1):3‐12.
https://doi.org/10.1002/pip.3102
9. Desilvestro H. What Physical Factors Affect Current‐Voltage.
Swiss: Dyesol Ltd; 2008.
10. Calogero G, Yum J‐H, Sinopoli A, Marco GD, Gratzel M,
Nazeeruddin MK. Anthocyanins and betalains as light‐harvesting
pigments for dye‐sensitized solar cells. Solar Energy. 2012;86(5):
1563‐1575. https://doi.org/10.1016/j.solener.2012.02.018
11. Chen W‐C, Kong F‐T, Li ZQ, et al. Superior Light‐Harvesting
Heteroleptic Ruthenium (II) Complexes with electron‐donating
antennas for high performance dye‐sensitized solar cells. ACS
Appl Mater Interfaces. 2016;8:19410‐19417. https://doi.org/
10.1021/acsami.6b04411
12. Sandquist C, McHale JL. Improved efficiency of betanin‐based
dye‐sensitized solar cells. J Photochem Photobiol A: Chem. 2011;
221(1):90‐97. https://doi.org/10.1016/j.jphotochem.2011.04.030
13. Oprea CI, Dumbrav A, Enache I, Georgescu A, Gîrt MA. A com-
bined experimental and theoretical study of natural betalain
pigments used in dye‐sensitized solar cells. J Photochem
Photobiol A Chem. 2012;240:5‐13. https://doi.org/10.1016/j.
jphotochem.2012.05.003
14. Ito S, Saitou T, Imahori H, Uehara H, Hasegawa N. Fabrication
of dye‐sensitized solar cells using natural dye for food pigment:
Monascus yellow. Energ Environ Sci. 2010;3(7):905‐909. https://
doi.org/10.1039/C000869A
15. Aziz N, Nor N, Mohd‐Adnan A, Taha R, Arof A. Study of antho-
cyanin stability derived from the fruit pulp of Melastoma
malabathricum in a coating system. Pigm Resin Technol.
2012;41(4):223‐229. https://doi.org/10.1108/03699421211242455
16. Singh LK, Karlo T, Pandey A. Performance of Fruit Extract of
Melastoma malabathricum L. as Sensitizer in DSSCs. Spectrochim
Acta Part A: Mol Biomol Spectros. 2014;118:938‐943. https://doi.
org/10.1016/j.saa.2013.09.075
17. Sofyan N, Ridhova A, Yuwono AH, Udhiarto A. Preparation of
anatase TiO2 nanoparticles using low hydrothermal tempera-
ture for dye‐sensitized solar cell. IOP Conf Series: Mater Sci
Eng. 2018;316:0120551‐0120556. https://doi.org/10.1088/1757‐
899X/316/1/012055
18. Joly D, Pellejà L, Narbey S, et al. A robust organic dye for dye
sensitized solar cells based on iodine/iodide electrolytes
SOFYAN ET AL.
combining high efficiency and outstanding stability. Sci Rep.
2014;4(4033):1‐7. https://doi.org/10.1038/srep04033
19. Sofyan N, Ridhova A, Yuwono AH, Udhiarto A. Fabrication of
solar cells with TiO2 nanoparticles sensitized using natural dye
extracted from mangosteen pericarps. Int J Technol.
2017;8(7):1229‐1238. https://doi.org/10.14716/ijtech.v8i7.692
20. Zhao H, Liu L, Andinob JM, Li Y. Bicrystalline TiO2 with con-
trollable anatase–brookite phase content for enhanced CO2
photoreduction to fuels. J Mater Chem A. 2013;1(28):8209‐8216.
https://doi.org/10.1039/C3TA11226H
21. Xu J, Li K, Wu S, Shi W, Peng T. Preparation of brookite titania
quasi nanocubes and their application in dye‐sensitized solar
cells. J Mater Chem A. 2015;3(14):7453‐7462. https://doi.org/
10.1039/C4TA06746K
22. Hema M, Arasi AY, Tamilselvi P, Anbarasan R. Titania nano-
particles synthesized by sol‐gel technique. Chem Sci Trans.
2013;2(1):239‐245. https://doi.org/10.7598/cst2013.344
23. Al‐Alwani MA, Mohamad AB, Kadhum AA, Ludin NA. Effects
of solvent on the extraction of natural pigments and adsorption
onto TiO2 for dye‐sensitized solar cell application. Spectrochim
Acta Part A: Mol Biomol Spectros. 2015;138:130‐137. https://
doi.org/10.1016/j.saa.2014.11.018
24. Narayan MR. Review: Dye sensitized solar cells based on natural
photosensitizers. Renew Sustain Energy Rev. 2012;16(1):208‐215.
https://doi.org/10.1016/j.rser.2011.07.148
25. Safie NE, Ludin NA, Hamid NH. Electron Transport Studies of
Dye‐Sensitized Solar Cells Based on Natural Sensitizer Extracted
from Rengas (Glutta spp.) and Mengulang (Heritiera elata)
Wood. BioResource. 2017;12(4):9227‐9243. https://doi.org/
10.15376/biores.12.4.9227‐9243
26. Chang H, Kao M‐J, Chen T‐L, Chen C‐H, Cho K‐C, Lai X‐R.
Characterization of natural dye extracted from wormwood and
purple cabbage for dye‐sensitized solar cells. Int J Photoenergy.
2013;2013:159502:1‐8. https://doi.org/10.1155/2013/159502
27. Mehmood U, Aslam Z, Al‐Sulaiman FA, et al. Electrochemical
impedance spectroscopy and photovoltaic analyses of dye‐
sensitized solar cells based on carbon/TiO2 composite counter
electrode. J Electrochem Soc. 2016;163(5):H339‐H342. https://
doi.org/10.1149/2.0111606jes
28. Chirinos R, Rogez H, Campos D, Pedreschi R, Larondelle Y.
Optimization of extraction conditions of antioxidant phenolic
compounds from mashua (Tropaeolum tuberosum Ruíz &
Pavón) tubers. Sep Purif Technol. 2007;55(2):217‐225. https://
doi.org/10.1016/j.seppur.2006.12.005
29. Park NG, van de Lagemaat J, Frank AJ. Comparison of Dye‐
Sensitized Rutile‐ and Anatase‐Based TiO2 Solar Cells. J Phys Chem
B. 2000;104(38):8989‐8994. https://doi.org/10.1021/jp994365l
How to cite this article: Sofyan N, Ridhova A,
Yuwono AH, Udhiarto A, Fergus JW. Synthesis of
TiO2 nanoparticles at low hydrothermal
temperature and its performance for DSSC
sensitized using natural dye extracted from
Melastoma malabathricum L. seeds. Int J Energy
Res. 2019;43:5959–5968. https://doi.org/10.1002/
er.4710