Photoelectrochemical properties of AgInS2 thin films prepared using electrodeposition

Photoelectrochemical properties of AgInS2 thin films prepared using electrodeposition

Solar Energy Materials & Solar Cells 95 (2011) 453–461 Contents lists available at ScienceDirect Solar Energy Materials & Solar Cells journal homepa...

2MB Sizes 0 Downloads 44 Views

Solar Energy Materials & Solar Cells 95 (2011) 453–461

Contents lists available at ScienceDirect

Solar Energy Materials & Solar Cells journal homepage: www.elsevier.com/locate/solmat

Photoelectrochemical properties of AgInS2 thin films prepared using electrodeposition Chih-Hao Wang a, Kong-Wei Cheng b, Chung-Jen Tseng a,n a b

Department of Mechanical Engineering, National Central University, No. 300, Jhongda Road, Chungli 32001, Taiwan Department of Chemical and Materials Engineering, Chang Gung University, Taoyuan 333, Taiwan

a r t i c l e in fo

abstract

Article history: Received 15 April 2010 Accepted 31 August 2010

Ternary silver–indium–sulfide samples were deposited on fluorine-doped tin oxide (FTO) coated glass substrates using a one-step electrodeposition method. A new procedure for the deposition of AgInS2 samples is reported. The effect of the [Ag]/[In] molar ratio in solution bath on the structural, morphological, and photoelectrochemical properties of samples was examined. X-ray diffraction patterns of samples show that the films are the AgInS2 phase. The thickness, direct band gap, and indirect band gap of the films were in the ranges 209–1021 nm, 1.82–1.85 eV, and 1.44–1.51 eV, respectively. The carrier densities and flat-band potentials of films obtained from Mott-Schottky and open-circuit potential measurements were in the ranges of 4.2  1019–9.5  1019 cm  3 and  0.736 to  0.946 V vs. the normal hydrogen electrode (NHE), respectively. It was found that the samples with molar ratio [Ag]/[In] ¼ 0.8 in solution bath had a maximum photocurrent density of 9.28 mA/cm2 with an applied bias of + 1.0 V vs. an Ag/AgCl electrode in contact with electrolyte containing 0.25 M K2SO3 and 0.35 M Na2S. The results show that high-quality AgInS2 films can be deposited on FTO-coated glass substrates for photoelectrochemical (PEC) applications. & 2010 Elsevier B.V. All rights reserved.

Keywords: Thin film Photocurrent Electrodeposition AgInS2

1. Introduction After Fujishima and Honda reported that water can be split into H2 and O2 using photoelectrochemical (PEC) cells [1], many studies have reported the use semiconductor materials for hydrogen production under irradiation [2–6]. Silver indium disulfide (AgInS2) is an I–III–VI (I ¼Ag, Cu; III ¼Al, In, Ga; VI¼S, Se, Te) ternary semiconductor with two crystal phases: chalcopyrite and orthorhombic [7]. The chalcopyrite phase of AgInS2 is stable when annealed at temperatures lower than 620 1C and the orthorhombic phase is stable when annealed at temperatures higher than 620 1C [8]. AgInS2 is an attractive semiconductor material for PEC, optoelectronic, and photovoltaic applications [8–10] because of its relatively large absorption coefficient and suitable energy band gap. It was reported that chalcopyrite AgInS2 can be a photoabsorber in an efficient CdS/AgInS2/CuInSe2 tandem solar cell because the lattice structure of AgInS2 is compatible with CuInSe2 and CdS semiconductors [11]. Wu et al. [12] reported that a visible-light-active ZnS/AgInS2/AgIn5S8 film electrode has a good stability in aqueous solutions containing Na2S and K2SO3 ions. A maximum photocurrent density of 6.28 mA/cm2 was obtained for an AgInS2 sample with an applied

n

Corresponding author. Tel.: + 886 3 4267348; fax: + 886 3 4254501. E-mail addresses: [email protected], [email protected] (C.-J. Tseng).

0927-0248/$ - see front matter & 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.solmat.2010.08.030

voltage maintained at + 1 V vs. a saturated calomel electrode (SCE) reference electrode in a solution bath containing Na2S and K2SO3 under a 300 W Xe lamp illumination with the light intensity of 100 mW/cm2 [12]. Several techniques have been reported for the preparation of AgInS2, such as hot wall epitaxy (HWE) [13], hot pressing (HP) [14], vertical gradient freezing (VGF) [15], melting of Ag–In–S elements [16], and spray pyrolysis methods (SP) [17–19]. Among these methods, electrodeposition is suitable for the preparation of ternary semiconductors such as CuInS2 [20], CuInSe2 [21], and CdIn2S4 [22]. Electrodeposition, which is suitable for large-scale production, is a low-cost, low-temperature, and eco-friendly method for the deposition of these ternary semiconductor film electrodes [23]. Few studies have reported about the deposition of AgInS2 using the electrodeposition method. In the present study, an electrodeposition method for the growth of AgInS2 samples on FTO-coated glass substrates with various [Ag]/[In] molar ratios in the precursor solution is reported. The structural, optical, and PEC properties of the films are investigated.

2. Experimental details Samples were made in a conventional three-electrode system using a potentiostat (model CHI 6081C, CH Instrument). The working, counter, and reference electrodes were FTO-coated glass

454

C.-H. Wang et al. / Solar Energy Materials & Solar Cells 95 (2011) 453–461

Table 1 Deposition parameters of the samples in this study. Sample

(a) (b) (c) (d) (e) (f)

[Ag]/[In] in solution

0.6 0.7 0.8 0.9 1.0 1.2

Concentration (M) AgNO3

In(NO3)3  5H2O

KNO3

TEA

Na2S2O3  5H2O

0.015 0.015 0.015 0.015 0.015 0.015

0.025 0.021 0.019 0.017 0.015 0.013

0.015 0.015 0.015 0.015 0.015 0.015

0.005 0.005 0.005 0.005 0.005 0.005

0.3 0.3 0.3 0.3 0.3 0.3

substrates, a platinum plate, and an Ag/AgCl (sat. KCl) electrode, respectively. The samples were grown on FTO-coated glass substrates (Uni-Onward Co., Taiwan) using co-electrodeposition. The substrates (5 cm  5 cm) were cut into several pieces (  1 cm  1.5 cm) and ultrasonically cleaned in ethanol, deionized water, and acetone for 30 min, respectively. Then, the substrates were carefully washed using deionized water and blown dry with ultra-pure nitrogen gas before experiments. The silver to indium molar ratio [Ag]/[In] in precursor solution was varied during the deposition of samples. The solution baths, which were well stirred, contained 15 mM silver nitrate (AgNO3, Riedel-de Haen, 99.99%), 13–25 mM indium nitrate (In(NO3)3  5H2O, Aldrich, 99.99%), 15 mM potassium nitrate (KNO3, Aldrich, 99.99%), 5 mM triethanolamine (TEA) ((C2H5O)3N, Merck, 99.0%), and 300 mM sodium thiosulfate (Na2S2O3  5H2O, J.T. Baker; 99.8%). The pH value of the bath was kept at 1 using concentrated H2SO4 (BASF Co., Taiwan, 96%) in order to decrease the formation of metal complexes such as In(OH)3 [24]. Na2S2O3 was used as the sulfur source because it can be used as a S2  ion source for the growth of I–III–VI2 semiconductor thin films using one-step electrodeposition [25]. Detailed deposition parameters of the samples used in this study are given in Table 1. The magnetic stirring was set at 100 rpm during deposition in order to increase the uniformity of films. The temperature of the solution bath was kept at 257 1 1C using a recirculation system (model nRC-10L, Cheng Seng Scientific Co., Taiwan). After electrodeposition, the films were rinsed in deionized water. These samples were annealed in a vacuum quartz tube at a temperature of 400 1C for 1 h. The crystal phase of the films was analyzed using an X-ray ˚ diffractometer (Siemens D5005) with Cu Ka (l ¼1.5405 A) radiation in the 2y range of 20–701. The scan rate was set at 31/min. The morphology and composition of samples were examined using a field-emission scanning electron microscope (FE-SEM, JEOL JSM-7401F) with an energy dispersive analysis of X-ray spectroscope (EDAX, Oxford INCA Energy 350). The surface roughness of films was investigated using an atomic force microscope (AFM, Park systems X70) in a non-contact mode. The thickness of samples was estimated using a surface profiler (Surfcorder ET3000). The transmittance and reflectance spectra of samples were obtained using an UV–vis spectrophotometer (JASCO Model V-670) with an integrating sphere in a wavelength range of 300–1200 nm at room temperature. Mott-Schottky measurements for the evaluation of the flat-band potentials of samples were performed using a potentiostat (model CHI 6081C) with a frequency response analyzer. The applied potentials were set in the range of  1.4 to+0.2 V (versus Ag/AgCl reference electrode). The frequency of the Mott-Schottky plots was set at 1 kHz in this study. The flat-band potentials of samples were also investigated using the open-circuit potential (OCP) method [26]. The PEC performances of films were measured in an electrochemical cell with K2SO3 (0.25 M) and Na2S (0.35 M) as the

electrolyte (pH¼13). The electrolyte was prepared using double deionized water and degassed by purging with nitrogen gas before each experiment. All measurement were carried out in a nitrogen environment at temperature of 25 1C. A conventional three-electrode system was employed in this study, where the sample, a Pt plate (  1  1 cm2), and an Ag/AgCl (sat. KCl) were the working, counter and reference electrodes, respectively. A copper wire was attached to the conducting layer of the FTO-coated glass substrates using silver paste. Epoxy resin was used to seal the conduction layer and corners of the samples. The current density of samples, as a function of applied voltage, was varied from  1.2 to+1.0 V (vs. Ag/AgCl reference electrode) using a 300 W xenon (Xe) lamp as the light source with the light intensity kept at 100 mW/cm2. The light intensity from the Xe lamp was measured using a power meter (Newport 1918-C). The current density–applied voltage plots of samples were recorded by a potentiostat (CHI 6081C) device in darkness and under illumination. The incident-photon-to-current efficiency (IPCE) of samples in the monochromatic light wavelength of 300–800 nm was measured using a 300 W xenon lamp combined with a monochromator (Oriel Cornerstone 130). The light intensity was measured using an optical power meter (Newport, 1918-C).

3. Results and discussion 3.1. Ag, In, and S electrodeposition analysis using cyclic voltammetry Cyclic voltammetry (CV) is an effective electroanalytical tool for finding redox couples and confirming the electrode potential during the deposition process [27]. Fig. 1 shows the cyclic voltammetry for single element of Ag + , In3 + , and S2 O2 ions 3 and a mixture of Ag, In, and S in solution bath using an applied potential range of + 0 to  1.6 V vs. the Ag/AgCl electrode. Three reduction peaks at 0.2 to  0.38 V (A1),  0.8 to  1.4 V (A2), and  0.8 to  1.1 V (A3) (vs. Ag/AgCl) can be observed in Fig. 1(a), corresponding to the reductions of silver, indium, and sulfur (S2 O2 3 ) ions, respectively. The reduction peaks of silver, indium, and sulfur are more negative than the values reported from Standard Electrode Potentials [28] (differences between our experimental results and standard electrode potential, E0 values are: Ag 0.96 V, In 0.62 V, and S  1.08 V). The possible reason was the complex agent in solution bath [29]. The concentration of complex agent (TEA) was high enough to keep most of Ag + and In3 + form [Ag(TEA) + ] and [In(TEA)3 + ] complexes. Free Ag + or In3 + ions in electrical double layer decrease, which result in the reduction peaks of silver, indium ions become more negative than those reported using Standard Electrode Potential. Such phenomenon can be explained using the Nernst equation [30]. Fig. 1(b) shows the cyclic voltammetry results for a mixture of silver, indium, and S2 O2 (molar ratio of Ag:In:S2 O2 3 3 ¼1:1:20) in precursor solution bath. Three reduction peaks can be observed

C.-H. Wang et al. / Solar Energy Materials & Solar Cells 95 (2011) 453–461

455

2 2S2 O2 3 2S4 O6

ð4Þ

 þ S2 O2 3 þ H 2S þ HSO3

ð5Þ

The reduction of indium in the mixtures containing Ag + , In3 + , and S2O23  ions started at  1.0 V (vs. Ag/AgCl) and reached a maximum reaction rate of deposition at  1.4 V (vs. Ag/AgCl) in our experiment. However, in order to avoid In-rich samples, such as the AgIn5S8 phase on the substrates, the co-deposition potential of samples was set to 1.0 V vs. the Ag/AgCl electrode in the present study. The total reaction for the deposition of the silver–indium–sulfide system can be written as xAg þyIn þ zS-AgX InY SZ

ð6Þ

3.2. Crystal phase characterization

Fig. 1. Cyclic voltammetry patterns of samples prepared in AgNO3 (0.015 M), In(NO3)3  5H2O (0.015 M), Na2S2O3  5H2O (0.3 M) with TEA (0.005 M) aqueous solution, (a) unitary Ag + , In3 + , and S2 O2 3 system and (b) ternary Ag–In–S system.

XRD patterns of annealed Ag–In–S film deposited on FTO-glass substrates are shown in Fig. 2. All samples were polycrystalline in nature and a preferential (1 1 2) orientation was observed at 2y ¼27.41 (JCPDS number 46–1088) for all deposited samples. With [Ag]/[In] molar ratio in deposition bath from 0.6 to 0.9 (sample (a)–(d)), the crystal phases of samples are chalcopyrite AgInS2 (JCPDS number 25–1330) and orthorhombic AgInS2 (JCPDS number 25–1328) mixed phases. For [Ag]/[In] molar ratio in deposition bath greater 1.0 (samples (e) and (f)), the crystal phase of samples is chalcopyrite AgInS2 phase. Similar results were also reported for the spray pyrolysis method using Ag-rich aqueous solutions [11]. The intensity of the (1 1 2) orientation peak increase with increase in [Ag]/[In] molar ratio from 0.6 to 0.8 in solution bath. When the [Ag]/[In] molar ratio is greater than 0.9, the intensity of (1 1 2) orientation peak of samples decrease. This may be due to the increasingly irregular arrangement of the lattice structure for samples with an increase in [Ag]/[In] molar ratio in solution [11]. Table 2 shows the atomic ratios of Ag, In, and S for the samples obtained using an X-ray energy dispersive analyzer (EDAX). It was observed that the samples with a nearly stoichiometric composition of AgInS2 were obtained from the starting molar ratios of [Ag]/[In]¼0.6 1.0 in the precursor solution. This indicates that AgInS2 thin films can be produced with various precursor concentration ratios using the electrodeposition route. Fig. 3 shows the element atomic ratios of [Ag]/[In] and 2[S]/([Ag]+3[In]) in the films using

at potential positions of  0.38 to 0.7 V (B1:Ag + ), 0.8 to  1.0 3+ (B2:S2O2) vs. the Ag/AgCl 3 ), and  1.0 to  1.4 V (B3:In electrode, respectively. The reduction potentials of silver, indium, and S2 O2 3 ions in the mixture were a little more negative than that of the single Ag + , In3 + , or S2 O2 3 ions in the deposition bath. It could be due to an overpotential caused by diffusion limitations of the ions in the solution used in the electrodeposition process [31]. The chemical reactions that occur during the electrodeposition used in this study may be as follows: (a) the ions diffuse to the electrode surface in the electrolyte; (b) adsorbed ions are reduced at the cathode; (c) silver, indium, and sulfur atoms migrate to the reduction reaction zone and produce ternary compound semiconductors [32]. The reactions and the Standard Electrode Potentials, E0 for the growth of AgInS2 thin film during electrodeposition are Ag þ þ e 2Ag In3 þ þ 3e 2In

E0 ¼ 0:2Vðvs: Ag=AgClÞ E0 ¼ 0:8 V ðvs:Ag=AgClÞ

 þ HSO 3 þ5 H þ 4e 2S þ 3H2 O

E0 ¼ 0:8V ðvs: Ag=AgClÞ

ð1Þ ð2Þ ð3Þ

Fig. 2. X-ray diffraction patterns of samples fabricated using the electrodeposition route with various [Ag]/[In] molar ratios in precursor solution and deposition potential kept at + 1.0 V (vs. Ag/AgCl electrode).

456

C.-H. Wang et al. / Solar Energy Materials & Solar Cells 95 (2011) 453–461

Table 2 Physical properties of the samples prepared using the electrodeposition method. Sample

(a) (b) (c) (d) (e) (f) a b

[Ag]/[In] in solution

0.6 0.7 0.8 0.9 1.0 1.2

Thickness (nm)

1021 799 633 543 470 209

Atomic ratio in samples

Ag

In

S

1 1 1 1 1 1

1.23 1.14 1.12 0.99 0.96 0.47

2.34 2.17 2.05 1.96 1.91 1.90

Roughness of samples (nm)

Direct Eg (eV)

Indirect Eg (eV)

a ENHE fb (V)

b ENHE fb (V)

Carrier density (cm  3)

Conduction type

674.84 511.95 309.13 301.37 202.47 200.21

1.85 1.82 1.83 1.84 1.85 1.85

1.51 1.46 1.44 1.46 1.50 1.48

 0.746  0.766  0.816  0.786  0.742  0.736

 0.916  0.926  0.946  0.926  0.916  0.826

6.07  1019 8.25  1019 9.53  1019 7.38  1019 5.45  1019 4.21  1019

n n n n n n

Obtained from Mott-Shottky plots. Obtained from open-circuit potential.

[Ag]/[In] molar ratio in solution. A possible reason is decrease in thickness of samples with increase in [Ag]/[In] molar ratio [35]. These results indicate that the ratio of [Ag]/[In] in the precursor solution bath plays an important role in controlling the microstructure of films produced using the electrodeposition route. Fig. 5 shows the AFM images of samples deposited with various ratios of [Ag]/[In] in the precursor solution. The root mean square (RMS) roughness (Rq) was defined using the following formula [36]: vffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi u n u1 X Rq ¼ t ð7Þ ðy yÞ2 ni¼1 i where n is the number of data points, yi the vertical coordinate of the image, and y the mean value of the yi. It was found that the RMS roughness of all film surfaces is in the range 200–675 nm. Detailed results are shown in Table 2. 3.4. Optical properties Fig. 3. The molar ratios of [Ag]/[In] and 2[S]/([Ag] + 3[In]) in the films as a function of the [Ag]/[In] molar ratio in the deposition bath.

EDAX analysis as a function of the [Ag]/[In] molar ratio in the precursor solution. The [Ag]/[In] atomic ratio of films increases with an increase in ratio of [Ag]/[In] in the precursor solution bath. The same results were reported for the co-deposition of ternary semiconductor thin films [33]. With molar ratios of less than 1.0 (samples (a)–(e)) in precursor solution, the 2[S]/([Ag]+3[In]) molar ratios in films obtained using EDAX analysis approached to 1.0, which indicates that there is no sulfur deficit in samples (a)–(e). With [Ag]/[In] molar ratios in solution bath41.0, the formation of excess sulfur was observed. Because the applied voltage was set at  1.0 V vs. the Ag/AgCl electrode in our experiments, the deposition rate of indium in sample (f) was low, which resulted in a relative excess of sulfur and silver in the sample. The thickness of the films was measured using surface profile measurements. The results are shown in Table 2. The thicknesses of films are in the range 209–1021 nm and decrease with increase in [Ag]/[In] atomic ratio in the solution bath. A possible reason is the decrease in the indium growth rate in our experiments with increase in [Ag]/[In] molar ratio in solution bath. This result is consistent with the EDAX analysis of samples.

3.3. Microstructure and morphology The FE-SEM images of samples with various deposition parameters are shown in Fig. 4. The microstructures of the surface of samples consist of small agglomerate plate-like particles. These microstructures become more uniform on the surface of the substrates with increase in [Ag]/[In] molar ratio in the solution (samples (a)–(f)). Similar results were observed for Ag–In–S thin films [34]. The grain size of samples decreases with increase in

The transmittance and reflectance spectra of samples with various [Ag]/[In] molar ratios in solution bath are shown in Fig. 6(a) and (b). It can be observed that the absorption edge approached 620 nm with an [Ag]/[In] molar ratio in the range 0.6–0.9 in solution bath. The absorption edge of samples approached to that of bulk AgInS2 [37]. The results agree well with the EDAX analysis and XRD patterns of samples (see Table 2 and Fig. 1). For [Ag]/[In] molar ratios greater than 0.9 in solution bath, the thickness of the samples decreases, which results in a noticeable increase in sample transmittance and reflectance [38,39]. Transmittance spectra (T), reflectance spectra (R), and thickness (t) can be used to estimate the absorption coefficient (a) using the Manifacier model [24,40,41] with the following formula:   1 1R ð8Þ a ¼ ln t T The absorption coefficient (a) is related to the energy band gap (Eg) of samples by:

ahn ¼ AðhnEg Þn

ð9Þ

where hn is the energy of the incident photon, A a constant; Eg represents the energy band gap of samples, and n is a constant. The band gap of samples can be obtained by extrapolating the linear portions of the respective curve to (ahn)1/n ¼0, where n is 2 for an indirect band gap transition and 1/2 for a direct band gap transition. To determine the possible transitions of samples, (ahn)1/n vs. hn was plotted for various values of n. Figs. 7 and 8 show the direct and indirect energy band gaps of samples obtained with various [Ag]/[In] molar ratios in solution. Direct and indirect energy band gaps of samples obtained from

C.-H. Wang et al. / Solar Energy Materials & Solar Cells 95 (2011) 453–461

457

Fig. 4. SEM images of samples deposited on the FTO-coated glass substrates with different [Ag]/[In] molar ratios in deposition bath. (a) [Ag]/[In] ¼ 0.6, (b) [Ag]/[In]¼ 0.7, (c) [Ag]/[In] ¼0.8, (d) [Ag]/[In]¼ 0.9, (e) [Ag]/[In] ¼1.0 and (f) [Ag]/[In]¼1.2.

Figs. 7 and 8 are shown in Table 2. The direct and indirect energy band gaps of samples varied from 1.82 to 1.85 eV and from 1.44 to 1.51 eV, respectively. These values of the direct energy band gap of AgInS2 are similar to those published in the literature (Eg varies from 1.83 to 2.07 eV) [10,18]. This study found that the energy band gap for samples (a)–(c) decreases with increase in [Ag]/[In] molar ratio in precursor solution. An increase in crystallinity and a decrease in the sulfur concentration in samples could cause such a shift [42]. In addition, with [Ag]/[In] molar ratios Z0.9 in solution bath, the energy band gap increased in our experiments because of the decrease in the crystallinity of the samples. 3.5. Photoelectrochemical characterization The flat-band potential of samples can be estimated using the Mott-Schottky (M-S) equation [43]:    1 2 kT ¼  EE fb eee0 ND e Cs2

ð10Þ

The Mott-Schottky (M-S) plot shows the potential dependence of Cs2 of a semiconductor electrode under depletion conditions, where e is the electron charge ( +e for electrons,  e for holes, e¼1.602  10  19 C), e the dielectric constant of the AgInS2 films, e0 is the permittivity of vacuum (8.85  10  14 F/cm), ND the donor carrier density, Efb the flat-band potential of the semiconductor, E the applied potential, k the Boltzmann constant (1.38  10  23 J/K), and T the absolute temperature (K). The carrier density of samples can be obtained from the slope of the 1/Cs2 vs. E plots. Efb can be obtained from the extrapolation for 1/Cs2 ¼0. Fig. 9 shows Mott-Schottky plots for films in K2SO3 (0.25 M) and Na2S (0.35 M) aqueous solution. The flat-band potential and the carrier density of films are shown in Table 2. It can be observed that the slope of the M-S plot is positive, which indicates that the samples are n-type semiconductors. The flat-band potential of samples can also be measured using open-circuit potential (OCP) under illumination [24]. The flat-band potentials and carrier density of samples are listed in Table 2. In this study, the flat-band potentials of the films are in the range of  0.736 to 0.946 V vs. the normal hydrogen electrode (NHE) from M-S plots and OCP methods.

458

C.-H. Wang et al. / Solar Energy Materials & Solar Cells 95 (2011) 453–461

Fig. 5. The AFM images of samples deposited on the FTO-coated glass substrates with different [Ag]/[In] molar ratios in deposition bath. (a) [Ag]/[In] ¼ 0.6, (b) [Ag]/[In]¼0.7, (c) [Ag]/[In]¼ 0.8, (d) [Ag]/[In] ¼0.9, (e) [Ag]/[In]¼1.0, (f) [Ag]/[In]¼ 1.2.

A negative shift in the flat-band potential was observed for compositions with [Ag]/[In] molar ratios of less than 0.8 in solution bath; sample (c) ([Ag]/[In]¼0.8) had the most negative shift of flat-band potential. For [Ag]/[In] molar ratios higher than 0.8, a shift to more positive values in the flat-band potential of the film was observed with increase in [Ag]/[In] molar ratio in solution bath. This indicates that the flat-band potential shifts to the negative direction with increase in concentration of sulfur deposited on the film. This effect was discussed in a previous study [44]. This result is consistent with the EDAX analysis of samples. The photoresponse of samples with various [Ag]/[In] precursor ratios in electrolyte was used to study the application of solar to hydrogen conversion. Fig. 10 shows the current density–applied voltage plots for samples (a), (c), and (f) using the chopping method for evaluating the PEC response of samples. The pH value of the electrolyte was 13. All films show a photoenhancement effect in the positive potential direction under illumination. These results indicate that the films are n-type semiconductors, which agrees well with the experimental results of M-S plots. It was

observed that the photocurrent density of samples increased from 3.9 to 9.28 mA/cm2 with an external bias of + 1.0 V vs. Ag/AgCl when the [Ag]/[In] molar ratio in solution bath was increased from 0.6 to 0.8 (samples (a)–(c)). In addition, the photocurrent densities of samples decreased (from 9.28 to 1.77 mA/cm2) with increase in [Ag]/[In] molar ratio in solution bath (samples (c)–(f)). A possible reason is the effect of film thickness; although larger film thickness tends to increase the amount of light absorption, it also causes higher electron transport resistance and increases the recombination of electron–hole pairs in the samples, resulting in smaller photocurrent density. Such observations were reported in the other studies [45–47]. Another possible reason is the carrier density of samples. Efficient separation of the photoexcited carriers was observed in the samples with larger carrier density because the steep potential field built up across a thinner space charge layer suppresses recombination [48]. As shown in Fig. 11, the maximum photocurrent density value was 9.28 mA/ cm2 at an applied bias of +1.0 V (vs. Ag/AgCl). Sample (c) shows a good PEC response due to its optimal thickness and large carrier density.

C.-H. Wang et al. / Solar Energy Materials & Solar Cells 95 (2011) 453–461

459

Fig. 8. The plot of (ahn)1/2 vs. hn for films.

Fig. 6. Optical spectra of samples in this study. (a) Transmittance spectra and (b) reflectance spectra.

Fig. 9. Mott-Schottky plots of samples in K2SO3 (0.25 M) and Na2S (0.35 M) solution.

Ag/AgCl reference electrode. The pH value of the electrolyte was 13. The IPCE of samples was determined using the following relation [49]: IPCE ð%Þ ¼

1240  isc  100 l  Iinc

ð11Þ

where isc (A/cm2) is the photocurrent of the system, l (nm) the incident photon wavelength, and Iinc (W/cm2) the intensity of the corresponding incident monochromatic light power. The IPCE of samples (a), (c), and (f) approaches to zero at the light wavelength of 640, 660, and 640 nm, respectively, corresponding to band gaps of 1.94, 1.87, and 1.94 eV for samples (a), (c), and (f), respectively. These results agree well with optical band gap of samples obtained from transmission and reflectance data. The IPCE of samples shows maximum values of 32%, 63%, and 9% at 560 nm for sample (a), (c), and (f), respectively, which indicated that the photocurrent of samples is due to the band gap excitation under illumination. Fig. 7. The plot of (ahn)2 vs. hn for films.

4. Conclusions Fig. 12 shows the incident-photon-to-current efficiency (IPCE) spectra of samples under monochromatic light illumination in the light wavelength of 300–800 nm at applied voltage of + 1.0 V vs.

The photoelectrochemical properties of electrodeposited AgInS2 thin films on FTO-coated glass substrates were investi-

460

C.-H. Wang et al. / Solar Energy Materials & Solar Cells 95 (2011) 453–461

Fig. 11. The photoenhancement current density plot of samples in K2SO3 (0.25 M) and Na2S (0.35 M) solution (pH value of electrolyte ¼13).

Fig. 12. Incident photon conversion efficiency (IPCE) of samples in K2SO3 (0.25 M) and Na2S (0.35 M) at applied voltage of +1.0 V vs. Ag/AgCl with various [Ag]/[In] molar ratios in deposition bath. (a) [Ag]/[In] ¼ 0.6 (c) [Ag]/[In]¼ 0.8 (f) [Ag]/[In]¼ 1.2 (pH value of electrolyte ¼13).

Fig. 10. The photocurrent density–applied voltages plots of samples in K2SO3 (0.25 M) and Na2S (0.35 M) solution (I) sample (a), (II) sample (c), (III) sample (f) (pH value of electrolyte ¼ 13).

gated. XRD results show that the samples consist of the mixed phase of chalcopyrite and orthorhombic AgInS2 for [Ag]/[In] ratios of less than 1.0 in solution bath. The single chalcopyrite phase was

observed with a precursor ratio [Ag]/[In]Z1.0 in solution. Nearly stoichiometric ratio AgInS2 was obtained from the starting molar ratios of [Ag]/[In]¼0.6 1.0 in the precursor solution, as determined from EDAX analyses. The direct and indirect band gap energies of samples were found to be in the ranges of 1.82–1.85 and 1.44–1.51 eV, respectively. The flat-band potentials of the AgInS2 films were in the range of  0.736 to  0.946 V vs. the normal hydrogen electrode (NHE) as obtained using the Mott-Schottky and OCP methods. All samples show n-type conductivity using the M-S plot and photocurrent density–applied voltage measurements. The maximum photocurrent density of films was found to be 9.28 mA/cm2 with a precursor ratio of [Ag]/[In]¼0.8 in solution with an applied bias of +1.0 V vs. the Ag/AgCl electrode under illumination with a 300 W xenon-arc lamp at a light intensity of 100 mW/cm2. The samples show maximum incident photon-to-current conversion efficiency of 63% at 560 nm with applied voltage maintained at +1.0 V vs. Ag/AgCl. Our results show that a low-cost electrochemical deposition method can be used to prepare AgInS2 ternary thin films for photoelectrochemical applications.

C.-H. Wang et al. / Solar Energy Materials & Solar Cells 95 (2011) 453–461

Acknowledgements The authors would like to thank Chang Gung University and the department of applied chemistry of National Chiao Tung University for XRD, AFM, SEM-EDAX analysis, and FE-SEM measurements.

References [1] A. Fujishima, K. Hoda, Electrochemical photolysis of water at a semiconductor electrode, Nature 238 (1972) 37–38. [2] A.J. Bard, M.A. Fox, Artificial photosynthesis: solar splitting of water to hydrogen and oxygen, Acc. Chem. Res 28 (1995) 141–145. [3] O. Khaselev, J. Turner, A monolithic photovoltaic–photoelectrochemical device for hydrogen production via water splitting, Science 285 (1998) 425–427. [4] X. Qian, X. Zhang, Y. Bai, T. Li, X. Tang, E. Wang, S. Dong, Photoelectrochemical characteristics of a-Fe2O3 nanocrystalline semiconductor thin film, J. Nanopart. Res 2 (2000) 191–198. [5] S.U.M. Khan, J. Akikusa, Photoelectrolysis of water to hydrogen in p-SiC/Pt and p-SiC/ n-TiO2 cells, Int. J. Hydrogen Energy 27 (2002) 863–870. [6] J. Turner, Sustainable hydrogen production, Science 305 (2004) 972–974. [7] J.L. Shay, J.H. Wernick, Ternary Chalcopyrite Semiconductors: Growth, Electronic Properties and Applications, Pergamon Press, New York, 1975, pp. 1–75. [8] L.H. Lin, C.C. Wu, C.H. Lai, T.C. Lee, Controlled deposition of silver indium sulfide ternary semiconductor thin films by chemical bath deposition, Chem. Mater. 20 (2008) 4475–4483. [9] R.W. Bir kmire, Recent progress and critical issues in thin film polycrystalline solar cells and modules, in: Proceedings of the 26th Photovoltaic Specialists Conference, IEEE, Anaheim, CA, 1997, pp. 295. [10] Z. Aissaa, M. Amlouk, T.B. Nasrallah, J.C. Bernede, S. Belgacema, Effect of S/In concentration ratio on the physical properties of AgInS2-sprayed thin films, Sol. Energy Mater. Sol. Cells 91 (2007) 489–494. [11] M.L. Albor Aguilera, M. Ortega-Lopez, V.M. Sanchez Resendiz, J. Aguilar Hernandez, M.A. Gonzalez Trujillo, Some physical properties of chalcopyrite and orthorhombic AgInS2 thin films prepared by spray pyrolysis, Mater. Sci. Eng. B 102 (2003) 380–384. [12] C.C. Wu, K.W. Cheng, W.S. Chang, T.C. Lee, Preparation and characterizations of visible light-responsive (Ag–In–Zn)S thin-film electrode by chemical bath deposition, J. Taiwan Inst. Chem. Eng 40 (2009) 180–187. [13] S.H. Youa, K.J. Hong, B.J. Lee, T.S. Jeong, C.J. Youn, J.S. Park, S.N. Baek, Temperature dependence of band gap and photocurrent properties for the AgInS2 epilayers grown by hot wall epitaxy, J. Cryst. Growth 245 (2002) 261–266. [14] K. Yoshino, A. Kinoshita, K. Nomoto, T. Kakeno, S. Seto, Y. Akaki, T. Ikari, Pressure dependence of AgInS2 crystals grown by hot-press method, Phys. Status Solidi C 3 (2006) 2648–2651. [15] K. Yoshino, N. Mitani, M. Sugiyama, S.F. Chichibu, H. Komaki, T. Ikari, Optical and electrical properties of AgIn(SSe)2 crystals, Physica B 302-303 (2001) 349–356. [16] G. Delgado, A.J. Mora, C. Pineda, T. Tinoco, Simultaneous Rietveld refinement of three phases in the Ag–In–S semiconducting system from X-ray powder diffraction, Mater. Res. Bull. 36 (2001) 2507–2517. [17] M.O. Lopez, A.M. Acevedo, O.S. Feria, Physical properties of AgInS2 films prepared by chemical spray pyrolysis, Thin Solid Films 385 (2001) 120–125. [18] Z. Aissa, T. Ben Nasrallah, M. Amlouk, J.C. Bernede, S. Belgacem, Some physical investigations on AgInS2 sprayed thin film, Sol. Energy Mater. Sol. Cells 90 (2006) 1136–1146. [19] M.L. Albor Aguilera, J.R. Aguilar Herna´ndez, M.A. Gonza´lez Trujillo, M. Ortega Lo´pez, G. Contreras Puente, Photoluminescence studies of chalcopyrite and orthorhombic AgInS2 thin films deposited by spray pyrolysis technique, Thin Solid Films 515 (2007) 6272–6275. [20] A.M. Martinez, L.G. Arriaga, A.M. Fernandez, U. Cano, Band edges determination of CuInS2 thin films prepared by electrodeposition, Mater. Chem. Phys. 88 (2004) 417–420. [21] E. Tzvetkova, N. Stratieva, M. Ganchev, I. Tomov, K. Ivanova, K. Kochev, Preparation and structure of annealed CuInSe2 electrodeposited films, Thin Solid Films 311 (1997) 101–106. [22] A.V. Kokate, M.R. Asabe, S.D. Delekar, L.V. Gavali, I.S. Mull, P.P. Hankare, B.K. Chougule, Photoelectrochemical properties of electrochemically deposited CdIn2S4 thin films, J. Phys. Chem. Solids 67 (2006) 2331–2336.

461

[23] D. Lincot, Electrodeposition of semiconductors, Thin Solid Films 487 (2005) 40–48. [24] K.W. Cheng, S.C. Wang, Influence of chelating agents on the growth and photoelectrochemical responses of chemical bath-synthesized AgIn5S8 film electrodes, Sol. Energy Mater. Sol. Cells 93 (2009) 307–314. [25] S. Nakamura, A. Yamamoto, Preparation of CuInS2 films with sufficient sulfur content and excellent morphology by one-step electrodeposition, Sol. Energy Mater. Sol. Cells 49 (1997) 415–421. [26] S.U. Khan, M.J.O’m. Bockris, The open circuit potential in solar cells, J. Appl. Phys. 52 (1981) 7270–7274. [27] Riyanto, M.R. Othman, J. Salimon, Analysis of ethanol using copper and nickel sheet electrodes by cyclic voltammetry, Malays. J. Anal. Sci. 11 (2007) 379–387. [28] D.R. Lide, Handbook of Chemistry and Physics, CRC Press, Boca Raton, Florida, 1992, pp. 8–17. [29] C.C. Chang, C.J. Liang, K.W. Cheng, Physical properties and photoresponse of Cu–Ag–In–S semiconductor electrodes created using chemical bath deposition, Sol. Energy Mater. Sol. Cells 93 (2009) 1427–1434. [30] F.F. Scholz, A.M. Bond, Electroanalytical Methods: Guide to Experiments and Applications, Springer, New York, 2002, pp. 10–48. [31] W. Plieth, Electrochemistry for Materials Science, Elsevier, Boston, 2008, pp. 143–167. [32] Y. Lai, F. Liu, Z. Zhang, J. Liu, Y. Li, S. Kuang, J. Li, Y. Liu, Cyclic voltammetry study of electrodeposition of Cu(In,Ga)Se2 thin films, Electrochim. Acta 54 (2009) 3004–3010. [33] T.S. Liang, Z. Zainal, T.W. Tee, I. Hamadneh, Potentiostatic deposition of copper indium disulfide thin films: effect of cathodic potentials on the optical and photoelectrochemical properties, Malays. J. Anal. Sci. 12 (2008) 600–608. [34] K.W. Cheng, C.M. Huang, G.T. Pan, W.S. Chang, T.C. Lee, C.K. Yang, Effect of Sb on the growth and photoelectrochemical response of AgIn5S8 film electrodes created by solution growth technique, Chem. Eng. Sci. 65 (2010) 74–79. [35] V. Kapaklis, P. Poulopoulos, V. Karoutsos, Th. Manouras, C. Politis, Growth of thin Ag films produced by radio frequency magnetron sputtering, Thin Solid Films 510 (2006) 138–142. [36] L.Y. Beaulieu, A.D. Rutenberg, J.R. Dahn, Measuring thickness changes in thin films due to chemical reaction by monitoring the surface roughness with in situ atomic force microscopy, Microsc. Microanal. 8 (2002) 422–428. [37] K. Okamoto, K. Kinoshita, Electrical and optical properties of AgInS2, Solid State Electron. 19 (1976) 31–32. [38] H.F. Zhang, R.J. Liu, H.F. Liu, C.X. Lei, D.T. Feng, C.K. Yuan, Mn-doped ZnO transparent conducting films deposited by DC magnetron sputtering, Mater. Lett. 64 (2010) 605–607. [39] M. Fahoume, O. Maghfoul, M. Aggour, B. Hartiti, F. Chrabi, A. Ennaoui, Growth and characterization of ZnO thin films prepared by electrodeposition technique, Sol. Energy Mater. Sol. Cells 90 (2006) 1437–1444. [40] M. Rusu, I. Salaoru, M.E. Popa, G.I. Rusu, On the optical characteristics of CdS thin films deposited by quasi-closed volume technique, Int. J. Mod. Phys. B 18 (2004) 1287–1297. [41] M.H. Habibi, N. Talebian, The effect of annealing on structural, optical and electrical properties of nanostructured tin doped indium oxide thin films, Acta Chim. Slov 52 (2005) 53–59. [42] M. Mathew, R. Jayakrishnan, P.M. Ratheesh Kumar, C.S. Kartha, K.P. Vijayakumar, Anomalous behavior of silver doped indium sulfide thin films, J. Appl. Phys. 100 (2006) 033504-1–033504-8. [43] C.M. Rangel, M. da Cunha Belo, Semiconductor electrochemistry and localised corrosion, Port. Electrochim. Acta 22 (2004) 295–304. [44] D.W. Chen, Y.Q. Liu, X.Y. Yi, G.Z. Xu, Influence of excess S2  ion on the interfacial photocatalysis in superfine CdS system and band-potential matching, Acta Phys. Chim. Sin 17 (2001) 781–787. [45] D. Chauhan, V.R. Satsangi, S. Dass, R. Shrivastav, Preparation and characterization of nanostructured CuO thin films for photoelectrochemical splitting of water, Bull. Mater. Sci. 29 (2006) 709–716. [46] Y. Sun, C.J. Murphy, K.R. Reyes, E.A. Reyes, J.M. Thornton, N.A. Morris, D. Raftery, Photoelectrochemical and structural characterization of carbondoped WO3 films prepared via spray pyrolysis, Int. J. Hydrogen Energy 34 (2009) 8476–8484. [47] V.B. Patil, G.S. Shahane, L.P. Deshmukh, Studies on photoelectrochemical (PEC) cells: a correlation between electrochemical and material properties, Mater. Chem. Phys. 80 (2003) 625–631. [48] E. Dorjpalam, M. Takahashia, Y. Tokudaa, T. Yoko, Controlling carrier density and its effect on I–V characteristics of the anatase-TiO2 thin films prepared by a sputter deposition method, Thin Solid Films 483 (2005) 147–151. [49] L. Su, L. ingang Zhang, J. Fang, M. Xu, Z. Lu, Electrochromic and photoelectrochemical behavior of electrodeposited tungsten trioxide films, Sol. Energy Mater. Sol. Cells 58 (1999) 133–140.