Improved photoelectrochemical performance of Cu2ZnSnS4(CZTS) thin films prepared using modified successive ionic layer adsorption and reaction (SILAR) sequence

Electrochimica Acta 150 (2014) 136–145

Contents lists available at ScienceDirect

Electrochimica Acta

journal homepage: www.elsev ier .com/ locate /e lectacta

Improved photoelectrochemical performance of Cu2ZnSnS4 (CZTS) thinfilms prepared using modified successive ionic layer adsorption andreaction (SILAR) sequence

M.P. Suryawanshi a,b, S.W. Shin c, U.V. Ghorpade a,d, K.V. Gurav a, C.W. Hong a,G.L. Agawane a, S.A. Vanalakar a, J.H. Moon a, Jae Ho Yun e, P.S. Patil b, Jin Hyeok Kim a,*,A.V. Moholkar b,*aOptoelectronics Convergence Research Center, Department of Materials Science and Engineering, Chonnam National University, 300, Yongbong-Dong,Buk-Gu, Gwangju 500-757, South Koreab Thin Film Nanomaterials Laboratory, Department of Physics, Shivaji University, Kolhapur 416-004, (MH) IndiacCenter for Nanomaterials and Chemical Reactions, Institute for Basic Science, Daejeon 305-701, South KoreadAnalytical Chemistry and Material Science Research Laboratory, Department of Chemistry, Shivaji University, Kolhapur 416-004, MH, Indiae Photovoltaic Research Group, Korea Institute of Energy Research, 71-2 Jang-Dong, Yuseong-Gu, Daejeon 305-343, South Korea

A R T I C L E I N F O

Article history:

* Corresponding authors. Tel.: +82 62 53E-mail addresses: [email protected]

(A.V. Moholkar).

http://dx.doi.org/10.1016/j.electacta.2014.100013-4686/ã 2014 Elsevier Ltd. All rights

0 1709, fax.kr (J.H. Ki

.124reserved.

A B S T R A C T

A promising successive ionic layer adsorption and reaction (SILAR)methodwith amodified sequence hasbeen designed and developed to deposit CZTS thin films with high absorbing characteristics. Theinfluence of copper concentration in the precursor solution on the structural, compositional,morphological and optical properties and on the photoelectrochemical performance of the CZTS thinfilms has been investigated. The copper concentration in the precursor solution was varied from 0.02 to0.008M in 0.004M intervals with constant Zn/Sn ratio. The films deposited using the optimized copperconcentration of 0.012M exhibit a prominent CZTS phase with a Cu-poor and Zn-rich composition, adense microstructure and optical band gap energy of 1.43 eV. The device based on a Cu-poor (0.012M)CZTS absorber layer exhibits the highest current density of 15.23mA/cm2 with a power conversionefficiency of 3.81%.

ã 2014 Elsevier Ltd. All rights reserved.

1. Introduction

Cu2ZnSnS4 (CZTS) is an emerging absorber material used inthin-film solar cells (TFSCs), sharing many properties with Cu(InxGa1-x)Se2 (CIGS). Thismaterial is structurally analogous to CIGSand has an absorption coefficient exceeding 104 cm�1 and a directband gap energy of 1.4–1.5 eV [1,2], which is suitable forphotovoltaic applications. Additionally, it contains non-toxic,low-cost, earth-abundant elements, as In and Ga are replaced byZn and Sn, making it attractive for large-scale applications [1,2].

Several attempts have been utilized to fabricate efficient TFSCsbased on CZTS absorber layers [3,4]. The highest efficiency of 12.6%has been reported for solution-processed CZTS thin films, whilevacuum-based sputtering methods have been achieved anefficiency of 10.8% in 14 cm2 cells. Katagiri et al. [5–9] reported

: +82 62 530 1699.m), [email protected]

efficiencies of 1.08%, 3.93% and 6.77% for chemical compositions ofCu/(Zn+Sn)=0.99, 0.73 and 0.87, respectively, for CZTS TFSCsprepared by vacuum evaporation. S. Ahmed et al. obtained anefficiency of 7.3% for CZTS TFSCs prepared by the sulfurization ofelectroplated precursors with Cu/(Zn+Sn) = 0.78 and a Zn/Sn ratioof 1.35 [10]. Thus, the highest efficiencies were obtained underCu-poor and Zn-rich growth conditions, as Cu-poor conditionsenhance the formation of Cu vacancies (VCu), which gives rise toshallow acceptors in CZTS, while Zn-rich conditions suppress thesubstitution of Cu at Zn sites, which gives rise to relatively deepacceptors [11]. Therefore, control of the chemical composition iscrucial for fabricating high-efficiency CZTS TFSCs.

Successive ionic layer adsorption and reaction (SILAR) is a well-established method that is low-temperature, scalable, andinexpensive compared to other solution-based non-vacuummethods. It is also particularly ecofriendly because it avoids theusage of toxic or dangerous reagents and organic solvents. Twodifferent strategies for the preparation of CZTS thin films by SILAR,namely, the traditional single-step approach [12,13], and sequen-tial stacking of sulfide layers [14], have been investigated so far.

M.P. Suryawanshi et al. / Electrochimica Acta 150 (2014) 136–145 137

A few general conclusions can be drawn from the comparison ofthese two approaches. The sequential stacking of sulfide layers bySILAR allows for more precise stoichiometric tuning compared tothe traditional single-step SILAR method but is potentially moretime-consuming and cumbersome and therefore less interestingfor industrial applications. Indeed, the use of a single-step SILARmethod for the deposition of quaternary compounds such as CZTSis quite challenging due to the different adsorptivities of thecations (Cu+, Sn4+, Zn2+) in a single cationic bath [15]. Althougheffort has been devoted to the deposition of CZTS thin films usingthe traditional single-step SILAR approach, the deposited CZTSfilms feature a porous morphology, markedly heterogeneouscoverage and poor crystallinity along with stoichiometry devia-tions [12,13].

In our previous report, we investigated the synergistic effect ofanionic bath immersion time on the properties and photoelec-trochemical performance of CZTS thin films prepared by a modifiedSILAR sequence [15]. It was found that the anionic bath immersiontime has a profound effect on the surface morphology of CZTS thinfilms, while the modification of the SILAR cycle sequence byseparating the cationic precursors into two different cationic bathseasily allows the chemical composition of the precursor films to becontrolled. The chemical compositionplays a vital role in fabricatinghigh-efficiency CZTS TFSCs. It is well known that the composition ofprecursor films using the SILAR method can be easily controlled bysimply varying the concentrationof precursor elements [16]. Hence,after a literature survey, a successful attempt has been made todepositCu-poorthinfilmsusingthemodifiedSILARsequencesimplyby varying the copper concentration in the precursor solution.Furthermore, the influence of chemical composition on theproperties and photoelectrochemical performance of CZTS thinfilms has been investigated.

2. Experimental

2.1. Chemicals

Copper (II) sulfate (CuSO4), zinc sulfate (ZnSO4), tin (II) sulfate(SnSO4) and sodium sulfide (Na2S) were purchased from Aldrich.Europium (III) nitrate (Eu(NO3)3) was also purchased from Aldrich.All reagents were used as received. Aqueous solutions wereprepared with water purified by a Millipore Milli-Q system.

2.2. Synthesis of CZTS precursor films

CZTS thin films with high absorbing characteristics weredeposited using the SILAR method with modified sequencing asdescribed in our previous report [15]. The deposition of CZTSprecursor films was achieved using 0.02-0.008M CuSO4, 0.08MSnSO4, 0.05M ZnSO4 and 0.16M Na2S. The copper concentrationwas systematically varied from0.02 to 0.008M in 0.004M intervalsto deposit Cu-poor CZTS precursor films. An immersion time of30 sec was used for each cationic precursor solution, whereas animmersion time of 5 sec was used for the anionic bath precursorsolution to deposit precursor films with S-poor compositions,which after sulfurization resulted into bigger grains. The depositedCZTS precursor films with different copper concentrations werethen sulfurized under a H2S (5%) +N2 (95%) atmosphere for 1h at580 �C. The corresponding CZTS samples were designated asCZTS0.02, CZTS0.016, CZTS0.012 and CZTS0.008 depending on the Cuconcentration.

2.3. Fabrication of the photoelectrochemical (PEC) device

Photoelectrochemical (PEC) solar cells with an active area of0.25 cm2 were fabricated using a standard two-electrode

configuration. The sulfurized CZTS with different copper concen-trations was used as the photoelectrode and indium doped tinoxide (ITO) as the counter electrode, which was sealed with theworking electrode using a spacer (� 1mm) of polyacrylamide. Thedistance between the CZTS photoelectrode and ITO counterelectrode was 0.4 cm. The redox electrolyte was 0.1M europiumnitrate [Eu(III)(NO3)3], which acts as an electron scavenger. The PECcharacteristics were measured by a Sol2A Oriel instrument(Newport Corporation, USA) with Keithley-4200 source meterunder 1.5 AM.

2.4. Characterization

The structural properties of sulfurized CZTS samples werestudied by X-ray diffractometry using an X-ray diffractometer(XRD, X’pert PRO, Philips, Eindhoven, Netherlands) operated at40kV and 30mA with CuKa radiation (l =1.5406Å) and Ramanspectroscopy using a Raman microscope (LabRam HR800 UV,Horiba Jobin-Yvon, France) with an excitation wavelength of514nm. The chemical composition was determined by X-rayfluorescence spectroscopy (Rigaku, EDXL-300). The surface mor-phology was observed using field emission scanning electronmicroscopy (FE-SEM) (Hitachi S4800, Japan). The chemical bindingenergies of the sulfurized CZTS samples were examined by high-resolution X-ray photoelectron spectroscopy (HR-XPS, VGMultilab2000, Thermo VG Scientific and UK) at room temperature. Thebinding energies in the spectrometer were calibrated using thecarbon 1s line at 285.0 eV. UV-visible absorbance spectra of thesamples were obtained using a UV-visible spectrophotometer(Cary 100, Varian, Mulgrave, Australia).

3. Results and Discussion

Fig. 1(a) shows the XRD patterns of CZTS thin films withdifferent copper concentrations (CZTS0.02, CZTS0.016, CZTS0.012,CZTS0.008) sulfurized at 580 �C for 1h. All of the samples arepolycrystalline and consist of a CZTS kesterite phase, as indicatedby a strong reflection along the (112) plane, irrespective of copperconcentration [17]. Other planes corresponding to (2 0 0), (2 2 0)and (312) also appeared with weak intensities. However, sampleCZTS0.02 shows an extra diffraction peak from the Cu2-XS secondaryphase marked by # along with a kesterite CZTS. The molybdenum(Mo) peaks are noted by * in the spectrum. No characteristic peaksfor other impurities are observed in any sample except for theCZTS0.02 sample, which confirms the formation of a prominentCZTS phase. The mean values of a = 0.543nm and c =1.085nm arein good accord with the reported values (a = 0.5427nm andc= 1.0848nm) [18,19]. The measured c/a ratio of 1.99 agreed wellwith the value of 1.998 for an ideally tetragonal I4

n ostructure

obtained from the standard JCPDS data 26-0575. It is also observedthat as the copper concentration increases, the intensity of the(112) diffraction peak increases, which indicates that more grainsare aligned along the (112) preferred orientation. In addition, thefull width at half maximum (FWHM) of a diffraction peak from the(112) plane narrows with increasing copper concentration. Theaverage crystallite sizes of these samples are estimated usingScherrer’s equation based on the (112) peak. The grain sizes of theCZTS0.02, CZTS0.016, CZTS0.012 and CZTS0.008 samples are 71.96nm,63.26nm, 53.74nm and 47.60nm, respectively. This reduction ingrain size with decreasing copper concentration shows theimprovement of the CZTS crystallites under Cu-rich conditions[20].

Secondary phases are frequently formed along with CZTSduring both synthesis and sulfurization. Previous studies havesuggested that this compound is formed by reactions among thecorresponding binary sulfides, such as Cu2-XS, ZnS and SnS2, or

[(Fig._1)TD$FIG]

Fig. 1. XRD patterns (a) and Raman spectra (b) of sulfurized CZTS0.02, CZTS0.016,CZTS0.012 and CZTS0.008 samples.

Table 1Compositional ratio and elemental composition of the as-deposited precursor filmand sulfurized CZTS0.02, CZTS0.016, CZTS0.012 and CZTS0.008 samples.

Sample As-deposited Sulfurized thin films

Cu/(Zn + Sn) Cu(at %)

Zn(at %)

Sn(at %)

S(at %)

Cu/(Sn + Sn) Zn/Sn

CZTS0.02 0.95 25.8 12.6 12.5 50.9 1.03 1.01CZTS0.016 0.88 24.9 12.6 13.9 51.1 0.94 0.91CZTS0.012 0.83 23.8 13.2 12.9 50.2 0.91 1.02CZTS0.008 0.75 22.1 13.7 13.1 50.6 0.82 1.05

138 M.P. Suryawanshi et al. / Electrochimica Acta 150 (2014) 136–145

ternary sulfides, such as Cu2SnS3. The most frequently foundsecondary phases in CZTS compounds are binary compounds, suchas Cu2-XS and ZnS, and ternary phases, such as Cu2SnS3 [21,22].However, XRD results alone often cannot confirm the presence ofthese secondary phases [22,23]. In particular, the diffractionpattern of CZTS is indistinguishable from that of Cu2SnS3 with acubic structure and ZnS with a zinc blende structure. Thus, weemployed Raman spectroscopy to confirm the phase purity of CZTSthin films. Fig. 1(b) shows the Raman spectra of CZTS samples withdifferent copper concentrations. All samples exhibit a strong peakat 334 cm�1with additional shoulder peaks at 2 92 and 366 cm�1,which confirms the formation of kesterite CZTS [24,25]. However,additional peaks at 311 and 469 cm�1 are observed for thesamples CZTS0.016 and CZTS0.02, respectively, which may beattributed to the formation of SnS2 and Cu2-XS, respectively. Theformation of SnS2 in the CZTS0.016 sample may be attributed to theslightly higher Sn content during sulfurization, while the Cu-richCZTS0.02 sample similarly forms a Cu2-XS secondary phase underthese conditions.

Table 1 shows the chemical composition of sulfurized CZTS0.02,CZTS0.016, CZTS0.012, and CZTS0.008 samples with different copperconcentrations in the as-deposited samples. It is clear that thesesamples have slightly higher Cu/(Zn+Sn) ratios relative to the as-deposited samples. However, the slightly higher copper content inthe CZTS0.02 sample formed the Cu2-XS secondary phase during

sulfurization, which is consistent with the XRD and Raman results.The Zn/Sn ratios of the sulfurized samples are greater than 1 exceptthat of the CZTS0.016 sample, just as in the precursor films. Thisdecrease in the Zn/Sn ratio in sample CZTS0.016 may be due to theslight increase in the Sn content during sulfurization, which resultsin the formation of a SnS2 secondary phase, corroborating theRaman results for sample CZTS0.016. Nevertheless, a range of Cu-poor and Zn-rich thin films may be deposited using the SILARmethod with modified sequencing, and the chemical compositionof the sulfurized thin films can be controlled fairly well by simplyvarying the chemical composition of the precursor SILAR solutions.

Fig. 2 (a-h) shows the FE-SEM images of as-deposited CZTSprecursor samples for different copper concentrations. Thesignificant effect of copper concentration on the microstructureand thickness of the precursor films can be clearly observed. TheCZTS0.02 sample, which has the highest copper concentration,exhibits a nanocrystalline morphology with agglomerated grains,which form building blocks over the surface. The cross-sectionalimage shows the presence of a continuous, compact layer of� 0.84mm thickness. For the CZTS0.016 sample, unreactedagglomerated particles with a porous morphology can be clearlyobserved in Fig. 2(b). The cross-sectional image shows thepresence of a compact and continuous layer of a reduced thicknessof � 0.82mm. A further decrease in the copper concentration forCZTS0.012 and CZTS0.008 results in the formation of agglomeratedparticles, which gradually form a particle assembly covering thewhole surface. Reduced thicknesses of � 0.69 and � 0.65mm canbe clearly observed in the cross-sectional images ( Fig. 2g & h) forthe CZTS0.012 and CZTS0.008 samples, respectively. The decrease inthe thickness of the as-deposited sample is attributed to thereduction in the copper concentration of the cationic precursorsolution, which provides fewer cations for surface adsorptionduring immersion. All of the as-deposited samples were thensulfurized at 580 �C for 1h in H2S (5%) +N2 (95%) atmosphere toincorporate S into their lattices, which are sulfur-deficient becauseof the dramatically reduced anionic bath immersion time (5 sec).Fig. 3 shows surface (a-d) and cross-sectional (e-g) FE-SEM imagesof the CZTS0.02, CZTS0.016, CZTS0.012 and CZTS0.008 samples. Allsulfurized samples are compact, uniform and without voids andcracks, except for the CZTS0.016 sample, which exhibited somevoids. However, a decrease in the grain sizewith decreasing copperconcentration was observed. A similar trend was observed in thecase of CIS and CIGS films [17,26]. The larger grain size in the Cu-rich CZTS0.02 sample may be attributed to the presence of Cu2-XS,which enhances the grain growth. It is well known that theformation of Cu2-XS phase is highly sulfur pressure dependent [26][26a]. The Cu-S phase diagram suggests that all Cu-S phases can beformed during temperature ramping [26][26b]. The Cu2-XS ishighly cation conducting phase that provides Zn and Sn exchangeleading to the formation of CZTS [26][26c]. However, at a givenhigh temperature and with a sufficient S activity, excess Cu in theCZTS film insteadmight partially exist as a liquid Cu2-XS. This liquidcan aid mass transport to form larger grain size in the CZTS film.The cross-sectional image shows the presence of a compact and

[(Fig._2)TD$FIG]

Fig. 2. FE-SEM images of as-deposited CZTS0.02, CZTS0.016, CZTS0.012 and CZTS0.008 samples.

M.P. Suryawanshi et al. / Electrochimica Acta 150 (2014) 136–145 139

uniform layer of CZTS of � 1mm thickness over the Mo (Fig. 3e).The as-deposited porous CZTS0.016 sample with unreactedagglomerated particles offers a large amount of room for S toincorporate fully during sulfurization, resulting in the formation of

larger grains with an increased thickness of� 0.97mm. In addition,the formation of a MoS2 layer between the interfaces of CZTS/Mowith increased thickness was also observed. However, the as-deposited CZTS0.012 and CZTS0.008 samples with a completely

[(Fig._3)TD$FIG]

Fig. 3. FE-SEM images of CZTS0.02, CZTS0.016, CZTS0.012 and CZTS0.008 samples sulfurized at 580 �C for 1 h in H2S (5%) + N2(95%) atmosphere.

140 M.P. Suryawanshi et al. / Electrochimica Acta 150 (2014) 136–145

grown assembly of particles resulted in a compact and uniformsurface with thicknesses of � 0.89 and 0.72mm, respectively, aftersulfurization treatment, as shown in Fig. 3(g & h). These resultssuggest that the morphological properties are predominantly

reliant on the composition of the precursor solution used todeposit the CZTS precursor films.

X-ray photoelectron spectroscopy was used to confirm thepresence of four constituent elements in all sulfurized CZTS0.02,

[(Fig._4)TD$FIG]

Fig. 4. High-resolution XPS spectra of Cu 2p (a), Zn 2p (b), Sn 3d(c) and S 2p (d) for sulfurized CZTS0.02, CZTS0.016, CZTS0.012 and CZTS0.008 samples andmagnified XPS spectra ofCu 2p (e) and Sn 3d (f) for sulfurized CZTS0.02 and CZTS0.016, respectively.

[(Fig._5)TD$FIG]

Fig. 5. The variation of (ahn)2 vs. (hn) for the estimation of the optical band gap

M.P. Suryawanshi et al. / Electrochimica Acta 150 (2014) 136–145 141

CZTS0.016, CZTS0.012 and CZTS0.008 samples. Fig. 4(a) shows the Cu 2pXPS spectra of all sulfurized samples. There are two peaks located at952.17and932.95 eV,with19.84-eVpeak separations, indicating theformationofCu+ forall sulfurizedsamples [27,28].Ontheotherhand,the peak observed at 932.25 eV is attributed to the Cu2-XS phase forthe CZTS0.02 sample (Fig. 4(e)) [29], which is consistent with theRaman results for this sample. The two Zn 2p peaks at 1045.7 and1022.5 eV in Fig. 4(b) suggest the presence of Zn2+ for all sulfurizedsamples [27,28]. The Sn3dcorepeaksat494.65and486.35 eVcanbeattributed to Sn4+ [27,28] (Fig. 4(c)). Sample CZTS0.016 shows anextrapeakat 487eV fromSn in theSnS2phase(Fig. 4(f)),which is in accordwith the Raman results [30]. Fig. 4(d) features a slight doublet of S2p1/2 and S 2p3/2, with peaks at 162.95 and 161.8 eV and a 1.15-eVpeak separation, which are consistent with the expected values(160–164eV) of S in sulfide phases [27,28,31].

The absorption data were analyzed and the band gap estimatedusing classical relationship between the absorption coefficient (a)and the photon energy (hn):

ahvð Þ2 ¼ hv� Eg

142 M.P. Suryawanshi et al. / Electrochimica Acta 150 (2014) 136–145

The optical band gap energy values can be obtained byextrapolating the linear portion of the plot of (ahn)2 against hntoa=0, as shown in Fig. 5. The observed band gap energy values are1.31, 1.40, 1.43 and 1.51 eV for CZTS0.02, CZTS0.016, CZTS0.012 andCZTS0.008, respectively. The band gap energies increase as the Cu/(Zn + Sn) ratio decreases. Cu-poor films are found to have largeroptical band gaps than Cu-rich films. A similar trend has beenobserved in the case of CZTS thin films [32,33]. The changes in theextent of p-d hybridization between the Cu d-levels and S p-levelsled to the formation of a Cu2-XS secondary phase given the highercopper content of the CZTS0.02 sample, shifting the opticalabsorption edge towards lower energy values [34]. However, theoptical band gap energy of 1.44 eV for the CZTS0.012 samplecorrespondswellwith that reported and is near the optimumvaluefor photovoltaic applications.

We investigated the photoelectrochemical (PEC) performanceof PEC devices based on CZTS0.02, CZTS0.016, CZTS0.012 and CZTS0.008samples by measuring the photocurrent density with a two-electrode configuration of the photocathode, which was immersedin 0.1M europium nitrate [Eu(III)(NO3)3] as an electron scavengerunder UV (100mW) illumination. Indium-doped tin oxide (ITO)was used as the counter electrode.

PEC characterization was preferred over the fabrication of thesolid-state device because it allows the rapid, non-destructiveevolution of CZTS absorber layers and minimizes the electricalshorting of the metal back contact of the device with the frontcontact. In addition, the conformal contact of the electrolyte withthe grains in CZTS thin filmsminimizes the distance overwhich theminority charge carrier (electrons) can diffuse to reduce the redoxspecies before they recombinewith the photogenerated holes. Themeasurement of PEC devices allows a variety of samples to betested to optimize parameters of interest, as its fabrication processis simple, cost-effective and not very time-consuming. Addition-ally, the fabrication only requires the use of a counter electrode,electrolyte and working electrode [35].

The selection of redox species, whose redox potential does notlead to the degradation of the semiconductor photoelectrode isessential for studying the photoelectrochemical performance. AEu3+/Eu2+ redox couple was chosen for the measurement of thephotoelectrochemical performance of PEC devices based onCZTS0.02, CZTS0.016, CZTS0.012 and CZTS0.008 electrodes. The stan-dard electrode potential for Eu3+/Eu2+ is � -0.35eV (versus SHE),and it fulfills most of these requirements [36,37]. Fig. 6 shows theband position diagram for ZnO, CdS and CZTS in electrolyte withEu3+/Eu2+ redox species, which reveals that the unoccupied level ofthe Eu3+/Eu2+ couple overlaps with the conduction band (CB) ofCZTS [38]. Thus, easy transfer of electrons from the CB to the redox

[(Fig._6)TD$FIG]

Fig. 6. Band position diagram of ZnO, CdS and CZTS semiconductors in electrolytewith Eu3+/Eu2+ redox species.

system can be expected. When the CZTS electrode is immersed inan electrolyte and exposed to light, the absorption of photonsgenerates electron-hole pairs. The minority carrier density isgenerally higher than the majority carrier density duringillumination [39]. The generated minority carriers are driven bythe electric field within the space charge region. These carriersmove toward the electrode/electrolyte interface and are trans-ferred across the interface to reduce one of the redox species. Thisphenomenon accounts for photocurrent generation. It should benoted that the photoelectrochemical transfer of electrons at the p-type semiconductor solution interface involves transition throughinterfacial barriers to accepter states in solution [40,41], which isgraphically illustrated in Fig. 7.

Fig. 8(a) & (b) show the current density versus voltage (J-V)characteristics of PEC devices based on CZTS0.02, CZTS0.016,CZTS0.012 and CZTS0.008 electrodes without and with illumination,respectively. Fig. 8(a) shows the ideal diode behavior with arectifying ratio originating from the formation of a junctionbetween the CZTS photoelectrode and electrolyte. The gradualincrease in the cathodic photocurrent with increasingly negativepotential indicates that the thin films were p-type [38]. Table 2shows the various solar cell parameters for the devices, where Vocis the open-circuit voltage, Jsc is the short-circuit current density, FFis the fill factor, h is the power conversion efficiency, Rs is the seriesresistance, Rsh is the shunt resistance, nd is the ideal factor and Jo isthe reverse saturation current density. As shown in Table 2, theCZTS0.012-based PEC device has the highest power conversionefficiency (PCE) of 3.81%, which is attributed to the fact that itpossesses the highest Jsc of 15.23mA/cm2 and Voc of 0.51V. Thisdevice features the highest Jsc due to the large grains and lack ofvoids of the CZTS0.012 sample (Fig. 3(c)). The conversion efficiencyof the polycrystalline TFSCs is strongly related to the grain size, as alarge grain size in the absorber layer maximizes both the minoritycarrier diffusion length and the built-in potential in TFSCs [42,43].The dependency of the solar cell efficiency on the copperconcentration shown in Table 2 clearly reveals the enhancementof the efficiency of the devices with decreasing copper concentra-tion. This may be explained by the fact that the removal of copper-based secondary phases is more highly favored as the Cu/(Zn + Sn)composition ratio decreases. Chen et al. reported that Cu vacancies(Vcu) and the substitution of Cu in Zn sites (CuZn) leads to thecreation of acceptors in CZTS. The acceptor transition energy levelsfor Vcu and CuZn are 0.02 and 0.10 eV, respectively, which are bothabove the VBM. Because CuZn is a relatively deep acceptor, theshallow acceptor of Vcu improves the efficiency of CZTS TFSCs. TheCu-poor composition suppresses CuZn formation and enhances Vcu

formation [11]. Consequently, the Jsc of the PEC cell based on Cu-poor CZTS is higher than that of slightly Cu-poor and stoichiomet-ric CZTS solar cells. However, an overly low Cu content could leadto a failure of the suppression of CuZn formation and thereby theformation Vcu, whichwould further degrade the Jsc and thereby the

[(Fig._7)TD$FIG]

Fig. 7. Graphical illustration of the photoelectrochemical transfer of electrons at thep-type semiconductor solution interface and the corresponding energy leveldiagram.

[(Fig._8)TD$FIG]

Fig. 8. J-V curves without (a) and with illumination (b) and semi-logarithmic plot (c) for PEC cells based on CZTS0.02, CZTS0.016, CZTS0.012 and CZTS0.008 samples.

M.P. Suryawanshi et al. / Electrochimica Acta 150 (2014) 136–145 143

device performance. Although controlling the Cu/(Zn+Sn) ratio viathe copper concentration in the precursor films significantlyenhanced the power conversion efficiencies of CZTS devices, thevalues are lower than those measured in polycrystalline CZTSTFSCs. This may be attributed to the relatively low shuntresistances (Rsh) and slightly high series resistances (Rs) of theCZTS devices. The relatively low Rsh values were due to shorts orleaks, whereas the slightly high Rs values were due to theformation of a thick MoS2 layer at the Mo/CZTS interface duringsulfurization.

The ideality factor nd and the reverse saturation current densityJo can shed light on the recombination mechanism of the devices.Therefore, we further analyzed the measured J-V characteristicsusing the lumped circuit model for solar cells and obtained thediode parameters following the approach described elsewhere[44,45]. The current equation of a solar cell device underillumination can be expressed as [46]

J ¼ JoexpeVndkT

� 1� �

(1)

where J and V are the diode current density and voltage,respectively; Jo is the reverse saturation current density; e is theelectronic charge; nd is the ideality factor; k is the Boltzmannconstant; and T is temperature.

Table 2Various solar cell parameters obtained for PEC cells based on CZTS0.02, CZTS0.016, CZTS0

Sample Voc (V) Jsc (mA/cm2) F. F. Efficienc

CZTS0.02 0.42 12.88 0.43 2.33CZTS0.016 0.46 13.04 0.48 2.87CZTS0.012 0.51 15.23 0.49 3.81CZTS0.008 0.39 8.81 0.38 1.31

The saturation current density (Jo) is obtained by extrapolatingthe linear portion of ln Jo versus V (Fig. 8 (c)) to zero bias, and theideal factor is calculated from the slope of the linear portion of theplot as follows:

nd ¼ ekT

� � @V@lnJo

� �(2)

The ideal factor (nd) for an ideal diode is 1, but it may vary from1 to 2 depending on the relationship between the diffusion currentand recombination current. When the diffusion current exceedsthe recombination current, then the ideality factor is closer to 1. Inthe reverse case, the ideality factor is closer to 2 [46]. The values ofnd, which were calculated from Fig. 8(c) and are listed in Table 2,are 2.23 and 2.67 for the devices based on samples CZTS0.02 andCZTS0.016, respectively. As the obtained values are greater than thenormal range, it can be inferred that defects exist in the quasi-neutral region as well as the junction and are responsible for therecombination of the carriers at the junction [47]. In contrast, theCZTS0.012 and CZTS0.008 devices have values of 1.63 and 1.89,respectively, which suggest that the recombination is less than thatof the CZTS0.02 and CZTS0.016 devices. The values of Jo for CZTS0.02,CZTS0.016, CZTS0.012 and CZTS0.008 devices are 3.56�10�4,1.23�10�5, 4.01�10�5 and 1.33�10�4, respectively. Note thathigh values of nd and Jo will lower the Voc and FF values of the solar

.012 and CZTS0.008 absorber layers.

y h (%) Rs (V) Rsh (V) nd Jo (mA/cm2)

63 500 2.23 3.56 � 10�4

33 600 2.67 1.23 � 10�5

52 500 1.63 7.28 � 10�5

90 500 1.89 2.32 � 10�4

144 M.P. Suryawanshi et al. / Electrochimica Acta 150 (2014) 136–145

cell. This overall decrease in Jo is due to the decreases in the defectdensity and carrier recombination. The CZTS0.012 device has thelowest nd and Jo values, indicating that it features the lowest carrierrecombination. Thus, it is reasonable to state that device CZTS0.012exhibits the best performance, which agrees well with the resultsfrom h. It can be concluded from the results obtained herein that itis important to precisely control the copper composition to achieveCu-poor CZTS for fabricating highly efficient TFSCs.

Although the achieved PCE for the PEC device based on SILAR-deposited CZTS thin films is the highest achieved to date, the devicestill suffers from lower efficiency, which is due to (i) its lower fillfactor and (ii) its higher series resistance. Hence, an efficiency closeto the highest reported values can likely be achieved by enhancingthe thin film quality to improve the fill factor and decrease theseries resistance. Further efforts are underway in our laboratory toimprove the photon conversion efficiency.

4. Conclusions

We designed and developed a novel, simple and cost-effectiveapproach using the SILAR method with modified sequence tosynthesize CZTS thin filmswith outstanding PV characteristics. Thecomposition-dependent properties and photoelectrochemicalperformance of the CZTS thin films have been investigated. Thechemical composition of CZTS thin films can be controlled fairlywell by simply varying the concentration of the precursor solution.The XRD results confirm that CZTS thin films are polycrystalline innature, having a kesterite structure. Larger grains have beenobserved under Cu-rich grain growth conditions. The optical bandgap of CZTS thin films shifted to higher energies as the Cu/(Zn + Sn)ratio of the CZTS thin films decreased. The PEC device preparedusing the Cu-poor CZTS0.012 exhibited the highest Jsc of 15.23mA/cm2 and a power conversion efficiency of 3.81%

Acknowledgments

This work was supported by the Human Resources Develop-ment Program (No. 20124010203180) of the Korea Institute ofEnergy Technology Evaluation and Planning (KETEP) grant fundedby the Korean Ministry of Trade, Industry and Energy andsupported partially by the University Grant Commission (UGC),NewDelhi, throughmajor research project F. No. 41-945/2012 (SR).

References

[1] K. Jimbo, R. Kimura, T. Kamimura, S. Yamada, W.S. Maw, H. Araki, K. Oishi, H.Katagiri, Cu2ZnSnS4-type thin film solar cells using abundant materials, ThinSolid Films 515 (2007) 5997.

[2] K. Moriya, K. Tanaka, H. Uchiki, Fabrication of Cu2ZnSnS4 thin-film solar cellprepared by pulsed laser deposition, Jpn. J. Appl. Phys. 46 (2007) 5780.

[3] M.P. Suryawanshi, G.L. Agawane, S.M. Bhosale, S.W. Shin, P.S. Patil, J.H. Kim, A.V.Moholkar, CZTS based thim film solar cells: a status review, Mater. Technol. 28(2013) 98.

[4] U. Ghorpade, M. Suryawanshi, S.W. Shin, K. Gurav, P. Patil, S. Pawar, C.W. Hong,J.H. Kim, S. Kolekar, Towards environmentally benign approaches for thesynthesis of CZTSSe nanocrystals by a hot injection method: a status review,Chem. Commun. 50 (2014) 11258.

[5] H. Katagiri, N. Sasaguchi, S. Hando, S. Hoshino, J. Ohashi, T. Yokota, Preparationand evaluation of Cu2ZnSnS4 thin films by sulfurization of E-B evaporatedprecursors, Sol. Energy Mater. Sol. Cells 49 (1997) 407.

[6] T. Kobayashi, K. Jimbo, K. Tsuchida, S. Shinoda, T. Oyanagi, H. Katagiri,Investigation of Cu2ZnSnS4-based thin film solar cells using abundantmaterials, Jpn. J. Appl. Phys. 44 (2005) 783.

[7] H. Katagiri, Cu2ZnSnS4 thin film solar cell, Thin Solid Films 480-481 (2005)426.

[8] H. Katagiri, K. Jimbo, S. Yamada, T. Kamimura, W.S. Maw, T. Fukano, T. Ito, T.Motohiro, Enhanced conversion efficiencies of Cu2ZnSnS4-based thin filmsolar cells by using preferential etching technique, Appl. Phys. Exp. 1 (2008)041201.

[9] H. Katagiri, K. Jimbo, W.S. Maw, K. Oishi, M. Yamazaki, H. Araki, A. Takeuchi,Development of CZTS-based thin film solar cells, Thin Solid Films 517 (2009)2455.

[10] S. Ahmed, K.B. Reuter, O. Gunawan, L. Guo, L.T. Romankiw, H. Deligianni, A highefficiencyelectrodepositedCu2ZnSnS4solarcell,Adv.EnergyMater.2 (2012)253.

[11] S. Chen, J.H. Yang, X.G. Gong, A. Walsh, S.H. Wei, Intrinsic point defects andcomplexes in the quaternary kesterite semiconductor Cu2ZnSnS4, Phy. Rev. B81 (2010) 245204.

[12] S.S. Mali, B.M. Patil, C.A. Betty, P.N. Bhosale, Y.W. Oh, S.R. Jadkar, R.S. Devan, Y.R.Ma, P.S. Patil, Novel synthesis of kesterite Cu2ZnSnS4 nanoflakes by successiveionic layer adsorption and reaction technique: characterization andapplication, Electrochim. Acta 66 (2012) 216.

[13] N.M. Shinde, D.P. Dubal, D.S. Dhawale, C.D. Lokhande, J.H. Kim, J.H. Moon,Room temperature novel chemical synthesis of Cu2ZnSnS4 (CZTS) absorbinglayer for photovoltaic application, Mater. Res. Bull. 47 (2012) 302.

[14] M.P. Suryawanshi, S.W. Shin, U.V. Ghorpade, K.V. Gurav, G.L. Agawane, C.W.Hong, J.H. Yun, P.S. Patil, J.H. Kim, A.V. Moholkar, A chemical approach forsynthesis of photoelectrochemically active Cu2ZnSnS4 (CZTS) thin films, Sol.Energy 110 (2014) 221.

[15] M.P. Suryawanshi, P.S. Patil, S.W. Shin, K.V. Gurav, G.L. Agawane, M.G. Kang, J.H.Kim, A.V. Moholkar, The synergistic effect anionic bath immersion time on thephotoelectrochemical performance of CZTS thin films prepared by a modifiedSILAR sequence, RSC Adv. 4 (2014) 18537.

[16] H.M. Pathan, C.D. Lokhande, Deposition of metal chalcogenide thin films bysuccessive ionic adsorption and reaction (SILAR) method, Bull. Mater. Sci. 27(2004) 85.

[17] A.V. Moholkar, S.S. Shinde, G.L. Agawane, S.H. Jo, K.Y. Rajpure, P.S. Patil, C.H.Bhosale, J.H. Kim, Studies of compositional dependent CZTS thin films solarcells by pulsed laser deposition technique: An attempt to improve theefficiency, J. Alloys Comp. 544 (2012) 145.

[18] W. Schäfer, R. Nitsche, Tetrahydral quaternary chalcogenides of the type Cu2-II-IV-S4(Se4), Mater. Res. Bull. 9 (1974) 645.

[19] B. Ananthoju, F.J. Sonia, A. Kushwala, D. Bahadur, N.V. Medhekar, M. Aslam,Improved structural and optical properties of Cu2ZnSnS4 thin films viaoptimized potential in single bath deposition, Electrochim. Acta 137 (2014)154.

[20] K.V. Gurav, S.M. Pawar, SeungWook Shin, M.P. Suryawanshi, G.L. Agawane, P.S.Patil, Jong-Ha Moon, J.H. Yun, Jin Hyeok Kim, Electrosynthesis of CZTS thinfilms by sulfurization of CZT precursor: Effect of soft annealing treatment,Appl. Surf. Sci. 283 (2013) 74.

[21] J.J. Scragg, T. Ericson, T. Kubart, M. Edoff, C. Platzer-Björkman, Chemicalinsights into the stability of Cu2ZnSnS4 films during annealing, Chem. Mater.23 (2011) 4625.

[22] P.A. Fernandes, P.M.P. Salomé, A.F. da Cunha, Growth and Raman scatteringcharacterization of Cu2ZnSnS4 thin films, Thin Solid Films 517 (2009) 2519.

[23] P.A. Fernandes, P.M.P. Salomé, A.F. da Cunha, Study of polycrystallineCu2ZnSnS4 films by Raman scattering, J. Alloys Comp. 509 (2011) 7600.

[24] K.V. Gurav, J.H. Yun, S.M. Pawar, S.W. Shin, M.P. Suryawanshi, Y.K. Kim, G.L.Agawane, P.S. Patil, J.H. Kim, Pulsed electrodeposited CZTS thin films: Effect ofduty cycle, Mater. Lett. 108 (2013) 316.

[25] M. Valdes, M. Modibedi, M. Mathe, T. Hillie, M. Vazquez, ElectrodepositedCu2ZnSnS4 thin films, Electrochim. Acta 128 (2014) 393.

[26] (a) J.A. Thorton, T.C. Lommasson, H. Talieh, B.H. Tseng, Reactive sputteredCuInSe2, Sol. Cells 24 (1998) 1;(b) H. Rodriguez-Alvarez, I.M. Kotschau, H.W. Schock, Pressure dependent real-time investigations on the rapid thermal sulfurization of Cu–In thin films, J.Crystal Growth 310 (2008) 3638;(c) D.J. Chakrabarti, D.E. Laughlin, The Cu-S (copper-sulfur) system, Bull. AlloyPhase Diagrams 4 (3) (1983) 254;(d) S. Wibmann, K.D. Becker, Tracer diffusion of indium in CuInS2, Solid StateIon 101-103 (1997) 539.

[27] S.C. Riha, B.A. Parkinson, A.L. Prieto, Solution-based synthesis andcharacterization of Cu2ZnSnS4 nanocrystals, J. Am. Chem. Soc. 131 (2009)12054.

[28] P.C. Dai, X.N. Shen, Z.J. Lin, Z.Y. Feng, H. Xu, J.H. Zhan, Band-gap tunable(Cu2Sn)x/3Zn1-xS nanoparticles for solar cells, Chem. Commun. 46 (2010) 5749.

[29] P. Velásquez, D. Leinen, J. Pascual, J.R. Ramos-Barrado, R. Cordova, H. Gómez, R.Schrebler, XPS SEM, EDX and EIS study of an electrochemically modifiedelectrode surface of natural chalcocite (Cu2S), J. Electroanal. Chem. 510 (2001)20.

[30] SeungWook Shin, S.M. Pawar, C.Y. Park, J.H. Yun, J.H.Moon, J.Y. Kim, Lee Studieson Cu2ZnSnS4 (CZTS) absorber layer using different stacking orders inprecursor thin films, Sol. Energy Mater. Sol. Cells 95 (2011) 3202.

[31] C.L. Azanza Ricardo, M.S. Suait, M. Muller, P. Scardi, Production of Cu2(Zn,Fe)SnS4 powders for thin films solar cell by high energy ball milling, J. PowerSources 230 (2013) 70.

[32] K. Tanaka, Y. Fukui, N. Moritake, H. Uchiki, Chemical composition dependenceof morphological and optical properties of Cu2ZnSnS4 thin films deposited bysol-gel sulfurization and Cu2ZnSnS4 thin film solar cell efficiency, Sol. EnergyMater. Sol. Cells 95 (2011) 838.

[33] A.I. Inamdar, S. Lee, K.Y. Jeon, C.H. Lee, S.M. Pawar, R.S. Kalubarme, C.J. Park, H.Im,W. Jung, H. Kim, Optimized fabrication of sputter deposited Cu2ZnSnS4 thinfilms, Sol. Energy 91 (2013) 196.

[34] M. Suryawanshi, S.W. Shin,W.R. Bae, K. Gurav,M.G. Kang, G. Agawane, P. Patil, J.H. Yun, J.Y. Lee, A. Moholkar, J.H. Kim, Kesterite CZTS nanocrystals: pH-dependent synthesis, Phys. Status Solidi (a) 211 (2014) 1531.

[35] S.C. Riha, S.J. Fredrick, J.B. Sambur, Y. Liu, A.L. Prieto, B.A. Parkinson,Photoelectrochemical characterization of nanocrystalline thin-film Cu2ZnSnS4photocathodes, ACS Appl. Mater. Inter. 3 (2011) 58.

M.P. Suryawanshi et al. / Electrochimica Acta 150 (2014) 136–145 145

[36] P.K. Sarswat, M.L. Free, An evaluation of depletion layer photoactivity inCu2ZnSnS4 thin film, Thin Solid Films 520 (2012) 4422.

[37] P.K. Sarswat, M.L. Free, Demonstration of a sol-gel synthesized bifacialCZTS photoelectrochemical cell, Phys. Status Solidi (a) 208 (2011)2861.

[38] J.J. Scragg, P.J. Dale, L.M. Peter, G. Zoppi, I. Forbes, New routes to sustainablephotovoltaics: evaluation of Cu2ZnSnS4 as an alternative absorber layer, Phys.Status Solidi (b) 245 (2008) 1772.

[39] R. Memming, The role of energy levels in semiconductor-electrolyte solar cells,J. Electrochem. Soc. 125 (1978) 117.

[40] Shahed U.M. Khan, John O’M. Bockris, A model for electron transfer at theilluminated p-type semiconductor-solution interface, J. Phys. Chem. 88 (1984)2504.

[41] C. Guillen, J. Herrero, D. Lincot, Photovoltaic activity of electrodeposited p-CuInSe2/electrolyte junction, J. Appl. Phys. 76 (1994) 359.

[42] T. Todorov, O. Gunawan, T. Gokmen, D.B.Mitzi, Solution-processed Cu(In,Ga)(S,Se)2 absorber yielding a 15.2% efficient solar cell, Prog. Photovolt. Res. Appl. 21(2013) 82.

[43] K. Tanaka, N. Moritake, H. Uchiki, Prepartion of Cu2ZnSnS4 thin films bysulfurizing sol-gel deposited precursors, Sol. EnergyMater. Sol. Cells 91 (2007)1199.

[44] S.S. Hegedus,W.N. Shafarman, Thin-film solar cells: device measurements andanalysis, Prog. Photovolt. Res. Appl. 12 (2004) 155.

[45] M.A. Contreras, K. Ramanathan, J. AbuShama, F. Hasoon, D.L. Young, B. Egaas, R.Noufi, Diode characteristics in state-of-the-art ZnO/CdS/Cu(In1-xGax)Se2 solarcells, Prog. Photovolt. Res. Appl. 13 (2005) 209.

[46] S.M. Sze, Physics of semiconductor devices, Science 258 (1992) 1474.[47] M. Patel, I. Mukhopadhyay, A. Ray, Study of the junction and carrier lifetime

properties of a spray-deposited CZTS thin-film solar cell, Semicond. Sci.Technol. 28 (2013) 055001.