Sei sulla pagina 1di 5

www.advenergymat.

de
www.MaterialsViews.com

Sulfur-Doped Nickel Oxide Thin Film as an Alternative


Communication

to Pt for Dye-Sensitized Solar Cell Counter Electrodes


Guan Hong Guai, Ming Yian Leiw, Chee Mang Ng, and Chang Ming Li*

Dye-sensitized solar cells (DSSCs) have attracted intense and redox couples,[19,20] although its low conductivity and poor
research attention because of their low cost and high efficiency charge carrier mobility limit its actual charge transfer process. To
in converting solar energy into electricity.[1,2] Typically, a DSSC improve the charge transfer impedance of NiO, a sulfur-doped
constitutes a dye-loaded mesoporous TiO2 photoelectrode, an nickel oxide (S–NiO) thin film is fabricated for a DSSC counter
electrolyte containing iodide/tri-iodide (I−/I3−) redox couples, electrode and its electrocatalytic activity towards the I−/I3−
and a catalytic counter electrode which reduces I3− ions.[3,4] redox process is investigated. By using an electrophoretic depo-
Since both the photo and counter electrodes are exposed to the sition method, a porous S–NiO thin film with a nanorod-like
electrolyte, a high DSSC performance is greatly dependent on a network is fabricated on a FTO substrate to greatly enhance its
fast reduction reaction at the counter electrode to suppress the effective reaction surface area. The surface composition and
possible reduction reaction by a charge recombination process at morphology of the S–NiO thin film are characterized using
the photoelectrode.[5,6] To achieve such selectivity, noble platinum scanning electron microscopy (SEM), X-ray photoemission
(Pt) is often used as a catalyst on fluorine-doped tin oxide (FTO) spectroscopy (XPS), and energy dispersive spectroscopy (EDS).
conducting glass to enhance the reduction rate at the counter The results obtained from cyclic voltammetry (CV) and alter-
electrode.[7] However, Pt is expensive and non-sustainable nating current (A.C.) impedance measurements indicate that
for long-term applications. Carbon-based materials such as the S–NiO\FTO (SNO) electrode possesses high electrocatalytic
polymers, carbon black, and carbon nanotubes have also been activity towards the I−/I3− redox process as compared to both
reported to replace Pt catalysts.[8,9] However, these carbon mate- plain NiO\FTO (NO) and FTO electrodes. When a low loading
rials have much weaker catalytic activities than Pt and thus of S–NiO (∼40 to 50 μg cm−2) is applied as a DSSC counter elec-
require large carbon loadings (loading ∼300 to >1000 μg cm−2) trode, a performance close to that of Pt and much better than
to compensate for the lowered electrocatalytic activity, resulting plain FTO and NiO is achieved.
in poor device energy density.[9–11] Therefore, a low cost and The surface morphology of the S–NiO thin film electro-
robust alternative counter electrode with high electrocatalytic deposited on FTO glass in Figure 1a depicts a porous network
activity towards I3− reduction is in high demand. consisting of intertwined nanorods with diameter ca. 60 nm
Transition metal based materials have been demonstrated to and length ∼200 nm. Based on the cross-sectional FESEM
possess outstanding catalytic performance in redox reactions image of the film, the thickness of the film is approximately
because of their multiple oxidation states.[12,13] However, very 60–140 nm depending on the deposition duration. Because
few of such materials have been studied and employed as a of its thin porous nature, the SNO electrode possesses a low
DSSC counter electrode.[14,15] In this study, nickel oxide (NiO) optical absorbance similar to bare FTO as shown in Figure 1b.
is selected as a potential electrode material because of its good Compared to the opaque or non-transparent electrode formed
charge capacity,[16] high chemical stability,[17,18] and a valence by Pt nanoparticles or carbon based materials, the high optical
band energy (∼–4.96 eV) comparable to the redox potential of transmittance of the SNO electrode (∼80%) appreciably allows
I−/I3− for favorable charge transfer between the metal oxide possible application to other DSSC layouts, such as a tandem
structure or metal-foil-supported DSSCs.
The corresponding XPS survey spectrum against binding
G. H. Guai, Prof. C. M. Li
energy (BE) in Figure 1c shows that five elements, Ni, O, S,
Center for Advanced Bionanosystems and School of Chemical
and Biomedical Engineering Sn, and C, are detected and the various XPS peaks are labeled
Nanyang Technological University accordingly. The presence of a C 1s peak (BE 284.2 eV) is likely
70 Nanyang Drive, Singapore 637457, Singapore attributed to the adventitious carbon from the equipment. Since
E-mail: Ecmli@ntu.edu.sg the S–NiO film is porous, a Sn 3d peak (BE 485.8 eV) from the
G. H. Guai, M. Y. Leiw, Dr. C. M. Ng underlying FTO substrate is also slightly detected. As the XPS
GlobalFoundries Singapore Pte. Ltd.,
60 Woodlands Industrial Park D, Street 2, Singapore 738406, Singapore peak position is highly sensitive to the chemical environment
M. Y. Leiw of the elements, high resolution XPS spectra are obtained to
School of Electrical and Electronic Engineering further characterize the oxide film. As shown in Figure 1d, the
Nanyang Technological University measured Ni 2p peak position of the NiO film is located at BE
50 Nanyang Avenue, Singapore 639798, Singapore 854.4 eV with increased energy as compared to that of pure Ni
Prof. C. M. Li metal (BE 853.1 eV). The oxide-associated O 1s peak position
Institute for Clean Energy & Advanced Materials
is determined at 530.4 eV (see inset (i) of Figure 1d). Similarly,
Southwest University
Chongqing 400715, P.R. China the measured S 2p peak is also shifted from BE 164 to 168.2 eV,
and is contributed by a sulphate (SO4) group (see inset (ii) of
DOI: 10.1002/aenm.201100582 Figure 1d). Thus, the S–NiO thin film could be confirmed to

334 wileyonlinelibrary.com © 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Energy Mater. 2012, 2, 334–338
www.advenergymat.de
www.MaterialsViews.com

Communication
Figure 1.  a) Surface morphology of a S–NiO\FTO (SNO) electrode. The inset shows its cross-sectional FESEM image. b) UV–vis optical absorbance
and transmittance (inset) spectra of a S–NiO thin film and bare FTO substrate. c) XPS survey spectrum of S–NiO\FTO (SNO) electrode. d) XPS high
resolution Ni 2p, O 1s, and S 2p spectra of a S–NiO thin film.

be a composite consisting of NiO and NiSO4 with the former I− − ∗


3 + 2 ↔ I + I2 (1)
as the main component. The higher BE peak positions of both
measured Ni 2p and S 2p peaks are a result of the attraction I∗2 + 2e−↔ 2I− + 2 (2)
of valence electrons in the Ni-O and Si-O bonds towards the
more electronegative O atoms. Consequently, both Ni and S In addition, the linear relationship of the reduction peak cur-
possess slightly positive net charges which increase the binding rent (Ipc) versus ν illustrated in inset (i) of Figure 2a indicates
energies with the core electrons. The EDS spectrum obtained that the reaction process of the I−/I3− redox couple is a surface
further confirmed the presence of Ni, S, and O in the SNO elec- adsorption-limited electrocatalytic process. Thus, the Ipc should
trode (see Figure S1, Supporting Information). be proportional to the adsorption quantity of the reacting ion
To understand the catalytic mechanism of the S–NiO thin (Γ) by the relation:[21]
film, the electrochemical behaviors of a SNO electrode toward
 
the I−/I3− redox reaction were studied by CV and impedance Ipc = (n · F )2 · ν ·  / ( R · T ) (3)
measurements. Figure 2a displays the CV curves of a SNO
electrode with 48.1 μg cm−2 S–NiO loading in an iodide-based where n is the number of electrons appearing in the half-
electrolyte at different scan rates (ν). When the potential is reaction for the redox couple (i.e., n = 2), F is the Faraday constant,
decreased from +1.0 V to –0.6 V, two reaction peaks occur. R is the universal gas constant, and T is the absolute tempera-
Since the cyclic voltammograms of the S–NiO film is alike that ture. Based on Equation 3, the adsorption quantity of I3− for the
of reported works for Pt catalysts, the I3− reduction pathway S–NiO electrode is determined to be 6.89 × 10−10 mol s−1.
on the oxide film can be assumed to be similar.[11,21] Peak A Figure 2b shows the linear voltammograms of different SNO
(at ∼ +0.6 V) denotes the adsorption process while peak R repre- electrodes with increasing oxide loading. Compared to bare
sents the I−/I3− reduction reaction (at ∼–0.3 V). As shown below FTO, the presence of a S–NiO film significantly increases the
in Equation 1 and 2, I3− is first adsorbed onto the unoccupied Ipc from 0.16 to 1.06 mA (SNO4), indicating that S–NiO has
S–NiO catalytic surface (Θ) and reduced to form intermediates, a much stronger electrocatalytic activity towards reduction of
followed by reduction of the adsorbed molecules (I2*) to I− and I3− than FTO. This is because a higher loading of S–NiO tends
desorption.[5] It is noted that the peak at +0.3 V is not a redox to increase the effective catalytic surface area, thus a larger Ipc
peak as the peak current remains zero and does not vary with occurs and the onset of the redox peak R is also shifted more
increasing ν. Instead, it may be a result of the overpotential positively from –0.41 to –0.27 V, signifying a better intrinsic
compensation of the SNO electrode. catalytic activity of the SNO electrode than FTO. In addition, the

Adv. Energy Mater. 2012, 2, 334–338 © 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim wileyonlinelibrary.com 335
www.advenergymat.de
www.MaterialsViews.com

the SNO electrode. It is known that in a voltammogram, the


Communication

peak current is proportional to the electrode surface area while


a more positive onset potential indicates a faster charge transfer
rate in a reduction process. Thus, S dopants in the S–NiO film
improve both charge transfer and reaction surface area for
better electrocatalytic performance.
Figure 2c illustrates the A.C. impedance Nyquist plots of
various SNO electrodes, in which the diameter of the semi-
circle represents the Faraday resistance of the electrochemical
reaction. In an electrochemical reaction, a lower Faraday resist-
ance indicates a faster electron transfer rate.[22] Since NiO is
semiconductive and has a lower charge carrier mobility as com-
pared to Pt, the poor conductivity of a thicker oxide film tends
to limit electron transport to the FTO current collector, which
in turn results in a slower electron transfer from the S–NiO\
FTO interface to the S–NiO\electrolyte catalytic surfaces. Thus,
the Faraday resistance is increased from ∼450 to ∼3000 Ω when
the S–NiO loading increases from 27.2 (SNO2) to 63.6 μg cm−2
(SNO6). The charge transfer comparisons by the A.C. imped-
ance measurements agree well with the CV results.
In a DSSC, the electrochemical reduction of the I−/I3− redox
couple can occur on both photo and counter electrodes, and a
slow reaction rate at the counter electrode can cause a serious
charge recombination in the photoelectrode to reduce the power
conversion efficiency (PCE). Thus, optimization of the S–NiO
loading for a high charge transfer rate is required to achieve
good catalytic performance. In this work, a S–NiO loading
of 48.1 μg cm−2 is found to have a relatively low resistance
(∼1000 Ω) while achieving a high Ipc (1.06 mA).
The current–voltage (I–V) characteristics of DSSCs made
from SNO counter electrodes with different S–NiO loadings is
depicted in Figure 3 (and Table S1, Supporting Information).
A conventional platinized electrode, with a Pt loading of 5.7 μg
cm−2, achieved a PCE of 6.08% and is used for comparison. It
is shown clearly that when an S–NiO thin film is introduced
onto a plain FTO substrate, the device performance is greatly
improved from 0.08% (of FTO electrode) to 3.23% (of SNO1
electrode), representing a significant forty times increment.

Figure 2.  a) Cyclic voltammograms of a S–NiO\FTO (SNO) electrode


at increasing scan rate from 0.05–0.20 V s−1. A Pt foil and Ag/AgCl elec-
trode were used as the counter and reference electrodes, respectively.
Labels A and R signify the adsorption current peak and the I−/I3− redox
current peak, respectively. The inset illustrates the relation between the
reduction peak current (Ipc) versus scan rate (ν) for the SNO electrode.
b) Reduction linear voltammograms for bare FTO, plain NiO\FTO (NO),
and various SNO electrodes with increasing S–NiO loading at a scan rate
of 0.10 V s−1. c) A.C. impedance spectra of FTO, NO, and different SNO
electrodes with increasing S–NiO loading. Figure 3.  Photocurrent–voltage characteristics of DSSCs with various
counter electrodes, namely S–NiO\FTO (SNO), platinized FTO (Pt),
plain NiO\FTO (NO), and bare FTO substrates. The inset illustrates the
NO electrode is also found to have a lower Ipc (0.57 mA) and a relationship between the power conversion efficiency (PCE) versus S–NiO
more negative onset of redox peak R (–0.47 V) as compared to loading for DSSCs with SNO counter electrodes.

336 wileyonlinelibrary.com © 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Energy Mater. 2012, 2, 334–338
www.advenergymat.de
www.MaterialsViews.com

Although the FTO surface could be exposed to electrolyte foil as the counter-electrode. By controlling the deposition time, different

Communication
through the porous structure of the S–NiO layer, a bare FTO loadings of S–NiO can be precisely controlled and deposited on different
electrode has a much poorer electrocatalytic activity than a FTO surfaces for DSSC performance evaluation. After the S–NiO formed
a thin film, the oxide coated electrode was dried at 70  °C for 1 h. A
S–NiO\FTO electrode. Thus the performance of a SNO based plain NiO thin film on a FTO substrate was prepared using the above-
device should be mainly dependent on S–NiO loading as proven described method without CH4N2S, followed by thermal annealing at
by the optimization shown in Figure 3. A SNO based DSSC is 350 °C for 30 min. A conventional platinized electrode was also prepared
also found to have a much higher PCE than a device based on a by sputtering Pt onto a FTO substrate using a JFC-1600 fine coater (JOEL
NO electrode (0.31%). The significantly enhanced performance Ltd., Japan). All chemicals used were purchased from Sigma–Aldrich.
of a S–NiO film over plain NiO and FTO films is apparently For clearer elaboration, the S–NiO\FTO and plain NiO\FTO electrodes
are referred to as SNO and NO electrodes respectively.
contributed by the presence of S dopants, which significantly
Assembly of DSSC: A DSSC photoelectrode was prepared by screen-
improve both charge transfer and reaction surface area for high printing a layer of TiO2 paste (DSL 18NR-T, Dyesol Co.) on a clean FTO
electrocatalytic activity at the counter electrode\electrolyte inter- substrate, followed by thermal annealing at 450  °C for 30 min. The
face as supported by the electrochemical results (see Figure 2). thickness of the resultant mesoporic TiO2 film was approximately 10 μm.
With greater S–NiO loading, the short circuit current density A 4 μm scattering over-layer consisting of 400 nm TiO2 nanoparticles
(Jsc) and open-circuit voltage (Voc) for SNO based devices are was then deposited and heated at 500  °C for 30 min, followed by an
immersion in N719 dye solution (Solaronix SA, 0.3 × 10−3 m) composed
gradually increased to ∼13.5 mA cm−2 and ∼770 mV, respec-
of a 50:50 mixture of acetonitrile (Sigma–Aldrich) and tert-butanol
tively, to finally achieve a PCE of 5.04% (see Figure 3 inset). It is (Sigma–Aldrich) for 20 h at room temperature for dye adsorption. A
also found that SNO based DSSCs often have a lower fill factor sandwich-type DSSC device was assembled by placing the dye-loaded
(FF) than Pt based DSSCs, which is likely caused by the higher photoelectrode on each prepared counter-electrode. The two electrodes
electron transfer resistance than in the latter device. Further- were clipped firmly together and a drop of liquid electrolyte (HL-HPE,
more, compared to other non-Pt based counter electrodes such Dyesol Co.) was injected into the cell. The active area of the cell was
as carbon black and carbon nanotubes, only a low loading of defined by a mask with an aperture of 0.25 cm2.
Characterization: Cyclic voltammetry (CV) measurement was carried
S–NiO (48.1 μg cm−2) is required to achieve a DSSC perform- out at scan rates between 0.05 and 0.20 V s−1 in a three-electrode
ance close to that of a Pt electrode. Hence the SNO electrode cell setup using a CHI 760B electrochemical workstation. The cell
has a prominent advantage of using much less materials for setup comprised a S–NiO coated, plain NiO coated, or non-coated
higher energy and power density to replace a Pt counter elec- substrate as the working electrode, an Ag/AgCl reference electrode,
trode. Investigations to further improve the electrocatalytic and a Pt foil as the counter-electrode. The electrolyte was LiI/I2 (10 ×
activity of the SNO electrode are currently underway in our lab. 10−3 m, molar ratio 9:1) in acetonitrile, with LiClO4 (0.1 m) added as the
supporting electrolyte. Electrochemical A.C. impedance spectroscopy
In summary, a transparent S-doped NiO\FTO (SNO) electrode
(EIS) was performed for the fabricated electrodes by using the same
with low oxide loading is successfully prepared as a counter elec- three-electrode cell setup with a CHI 760B workstation. The EIS were
trode for DSSCs. The electrode material has been characterized recorded over a frequency range of 0.1 to 105 Hz at an applied voltage
by FESEM, XPS, and EDS. Electrochemical behaviors of the SNO of 5 mV. To study the surface morphology of the fabricated electrodes,
electrode in a I−/I3− redox system are systematically studied, indi- field-emission scanning electron microscopy (FESEM) was performed
cating that the I3− reduction pathway by the oxide film electrode using a Jeol JSM-6701F microscope. The element stoichiometry of the
S–NiO thin film was determined by X-ray photoelectron spectroscopy
follows a two-step process and is surface adsorption limited. The
(XPS) using a Kratos Analytical AXIS Ultra model with a monochromatic
enhanced electrocatalytic activity of the SNO electrode is very likely Al Kα source. Energy dispersive spectroscopy (EDS) was performed
contributed by the S dopants in the NiO film which significantly using an EDS model 6587 from Oxford Instrument coupled with a SEM
improve both charge transfer and electrochemical reaction surface JEOL JSM5600LV and Oxford Instrument Link ISIS system software. A
area. However, a very high loading of S–NiO tends to cause high Shimadzu UV-2450 spectrophotometer was used to record the optical
interfacial impedance despite providing a large effective surface. absorbance and transmittance of the prepared electrode between the
Hence, optimization of the SNO electrode is required to achieve wavelengths of 300 and 800 nm. The current–voltage characteristics of
the DSSCs were measured with a Keithley 2420 multimeter under one
a high electrocatalytic effect between the reaction surface and sun condition using a solar light simulator (Abet Technologies S2000
charge transfer rate towards the I−/I3− redox process. This work with a 550-W Xe lamp and an AM 1.5 filter, 100 mW cm−2).
offers a promising Pt-free counter electrode material with high
energy density to achieve good DSSC performance close to the
conventional platinized electrode. The high optical transmittance
of the SNO electrode also provides potential applications in metal- Supporting Information
foil-supported DSSCs or power-producing windows. Supporting Information is available from the Wiley Online Library or
from the author.

Experimental Section
Acknowledgements
Preparation of Counter Electrodes: To fabricate the S–NiO\FTO (SNO)
electrode, a precursor mixture containing NiCl2 (5 × 10−3 m) and CH4N2S This work was financially supported by Center for Advanced Nanosystems,
(150 × 10−3 m) in pH 10 KOH solution was first prepared and aged for Nanyang Technological University (Singapore), and GlobalFoundries
3 h at room temperature. Electrophoretic deposition of S–NiO on the Singapore Pte. Ltd.
cleansed FTO glass substrate (Nippon Glass Sheet, ∼12 Ω −1) was
then accomplished using a CHI 760B electrochemical workstation (CH Received: September 28, 2011
Instrument Inc., Shanghai) at a constant voltage of –0.8 V in a three- Revised: November 24, 2011
electrode cell setup with a Ag/AgCl reference electrode and a planar Pt Published online: February 6, 2012

Adv. Energy Mater. 2012, 2, 334–338 © 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim wileyonlinelibrary.com 337
www.advenergymat.de
www.MaterialsViews.com

[1] I. K. Ding, J. Zhu, W. Cai, S.-J. Moon, N. Cai, P. Wang, [12] Y. Wang, C. X. Guo, X. Wang, C. Guan, H. Yang, K. Wang, C. M. Li,
Communication

S. M. Zakeeruddin, M. Grätzel, M. L. Brongersma, Y. Cui, Energy Environ. Sci. 2011, 4, 195.


M. D. McGehee, Adv. Energy Mater. 2011, 1, 52. [13] Y. Wang, X. Wang, C. M. Li, Appl. Catal. B 2010, 99, 229.
[2] Y. Luo, D. Li, Q. Meng, Adv. Mater. 2009, 21, 4647. [14] M. Wang, A. M. Anghel, B. T. Marsan, N.-L. Cevey Ha,
[3] F. Huang, D. Chen, X. L. Zhang, R. A. Caruso, Y.-B. Cheng, Adv. N. Pootrakulchote, S. M. Zakeeruddin, M. Grätzel, J. Am. Chem.
Funct. Mater. 2010, 20, 1. Soc. 2009, 131, 15976.
[4] J. Halme, P. Vahermaa, K. Miettunen, P. Lund, Adv. Mater. 2010, 22, [15] Q. W. Jiang, G. R. Li, S. Liu, X. P. Gao, J. Phys. Chem. C 2010, 114,
E210. 13397.
[5] L. Bay, K. West, B. Winther-Jensen, T. Jacobsen, Sol. Energy Mater. [16] W. Lv, F. Sun, D.-M. Tang, H.-T. Fang, C. Liu, Q.-H. Yang,
Sol. Cells 2006, 90, 341. H.-M. Cheng, J. Mater. Chem. 2011, 21, 9014.
[6] N. Papageorgiou, Coord. Chem. Rev. 2004, 248, 1421. [17] J. Y. Lee, K. Liang, K. H. An, Y. H. Lee, Synth. Met. 2005, 150, 153.
[7] L. Chen, W. Tan, J. Zhang, X. Zhou, X. Zhang, Y. Lin, Electrochim. [18] H. Kumagai, M. Matsumoto, K. Toyoda, M. Obara, J. Mater. Sci.
Acta 2010, 55, 3721. Lett. 1996, 15, 1081.
[8] T. N. Murakami, M Grätzel, Inorg. Chim. Acta 2008, 361, 572. [19] K. Nakaoka, J. Ueyama, K. Ogura, J. Electroanal. Chem. 2004, 571, 93.
[9] G. Wang, W. Xing, S. Zhuo, J. Power Sources 2009, 194, 568. [20] G. Boschloo, A. Hagfeldt, Acc. Chem. Res. 2009, 42, 1819.
[10] H. Choi, H. Kim, S. Hwang, W. Choi, M. Jeon, Sol. Energy Mater. Sol. [21] Y. Tang, X. Pan, C. Zhang, S. Dai, F. Kong, L. Hu, Y. Sui, J. Phys.
Cells 2011, 95, 323. Chem. C 2010, 114, 4160.
[11] K. Imoto, K. Takahashi, T. Yamaguchi, T. Komura, J. Nakamura, [22] C. M. Li, C. Q. Sun, S. Song, V. E. Choong, G. Maracas, X. J. Zhang,
K. Murata, Sol. Energy Mater. Sol. Cells 2003, 79, 459. Front Biosci. 2005, 10, 180.

338 wileyonlinelibrary.com © 2012 WILEY-VCH Verlag GmbH & Co. KGaA, Weinheim Adv. Energy Mater. 2012, 2, 334–338

Potrebbero piacerti anche