Experimental investigation of the effect of indium content on the CuIn5S8 electrodes using electrochemical impedance spectroscopy

Materials Research Bulletin 61 (2014) 519–527

Contents lists available at ScienceDirect

Materials Research Bulletin journal homepage: www.elsevier.com/locate/matresbu

Experimental investigation of the effect of indium content on the CuIn5S8 electrodes using electrochemical impedance spectroscopy M. Gannouni *, I. Ben Assaker, R. Chtourou Laboratoire de Photovoltaïque, Centre de Recherches et des Technologies de l'Energie, Technopole Borj Cedria, Bp 95, Hammamlif 2050, Tunisia

A R T I C L E I N F O

A B S T R A C T

Article history: Received 26 May 2014 Received in revised form 27 October 2014 Accepted 29 October 2014 Available online 4 November 2014

This paper reports on the use of electrochemical impedance spectroscopy to investigate the electrochemical behavior of spinel CuIn5S8/electrolyte interface. The CuIn5S8 spinel films have been potentiostatically deposited onto indium tin oxide (ITO)-coated glass substrate. CuCl2 and InCl3 mixed solutions with different [Cu]/[In] ratios were used as cation precursor and Na2S2O3 as the anion precursor in acidic solution and at room temperature. The effect of the [Cu]/[In] ratio in the precursor solution on the structural, chemical stoichiometry, and morphological properties of prepared samples, as well as the electrochemical behavior of the CuIn5S8/electrolyte interface was investigated. The electrochemical impedance spectroscopy data have been modeled using an equivalent circuit approach. Several parameters such as, flat-band potential and free carrier concentration were determined by the change in the Mott–Schottky plots. ã 2014 Published by Elsevier Ltd.

Keywords: A. Interfaces A. Semiconductors A. Thin films C. Impedance spectroscopy

1. Introduction I–III–VI ternary semiconductors with the general formulae of I– III–VI2 and I–III5–VI8 have potential as photo absorbers in solar cells, optoelectronics devices, and photoelectrochemical cells. These ternary semiconductor materials can be prepared to have either n- or p-type conductivity, depending on the synthesis methods and the composition of elements in samples [4,5]. For applications of solar energy conversion, CuIn5S8 semiconductor has a suitable band gap to be a photo-absorber in solar or photoelectrochemical cells [1–3]. Furthermore, this material does not contain toxic elements like Se and Ga, which is an advantage in comparison with the other absorbers like CuInSe2 and CuGa(In)Se2. This ternary compound belongs to the CuIn2n + 1S3n + 2 family with n = 2, and is formed in the in-rich side of the pseudo-binary Cu2S–In2S3 system [9]. At the extreme limits of structural tolerance to off-stochiometry of chalcopyrite phases, this compound stabilizes due to the ordering of the neutral defect pairs (2VCu
1 + InCu2+) and (2CuIn
2 + InCu2+) in the Cu–In–VI2 phase [46]. CuIn5S8 is a cubic spinel with indium found principally on octahedral sites rather than on tetrahedral sites as found in most chalcopyrite structures [10]. However, n-CuInS2 chalcopyrite is known to be difficult to grow [47]. Therefore, CuIn5S8 spinel phase being typically n-type and having a similar band gap of 1.5 eV,

* Corresponding author. Tel.: +216 27 192 196; fax: +216 79 325 934. E-mail address: [email protected] (M. Gannouni). http://dx.doi.org/10.1016/j.materresbull.2014.10.070 0025-5408/ ã 2014 Published by Elsevier Ltd.

matches well with the solar spectrum for energy conversion [6]. CuIn5S8 is visible-light-active materials with high-absorption coefficients, suitable band gaps, good radiation stability, and easy conversion between n- and p-type carrier types which permits a variety of potentially low-cost homo and hetero junction [5,7]. An efficiency of about 9.1% was associated with the p-CuInS2/CuI/nCuIn5S8 tandem structure solar cell [8]. A variety of physical and chemical techniques have been employed to fabricate this material [11–13]. Among these methods, electrodeposition as a cheap and facile method is widely used for the co-deposition of copper indium sulfide thin films under particular conditions [14–17]. It is easy to scale up to produce films with good quality and large area [18]. But a potential problem in the development of this technique is the control of the sample composition which is directly related to the electrodeposition conditions and concentrations. Studies concerning the CuIn5S8 semiconductor/electrolyte interface (namely polysulfide electrolyte) have been investigated in some detail by many authors [19–22]. Even though these studies demonstrated the importance of these semiconductors as electrodes in photoelectrochemical cells, previous efforts lacked a systematic study necessary to understand the phenomena related to the CuIn5S8 semiconductor/electrolyte interface by means of electrochemical impedance spectroscopy technique (EIS). This latter technique constitutes a powerful tool for studying semiconductor/electrodes barrier, which measures the response of a sample under an AC stimulus in which the frequency is varied over a wide range. EIS has wide applicability, and has been used for the study of ionic conductors, dielectric materials, semiconductors,

520

M. Gannouni et al. / Materials Research Bulletin 61 (2014) 519–527

Table 1 Deposition parameters of the samples and atomic composition determined by XRF analysis. [Cu]/[In] ratio

Bath composition (mM) 2+

0.5 0.4 0.33 0.28 0.25 0.22

3+

Atomic ratio obtained from XRF analysis (at%)

Cu

In

S2O3

1 1 1 1 1 1

2 2.50 3 3.50 4 4.50

10 10 10 10 10 10

2


KCl

Cu (at%)

In (at%)

S (at%)

Cu:In:S

250 250 250 250 250 250

8.5 8.1 7.9 7.2 6.7 6.6

27.2 32.5 34.9 37.8 39.9 40.8

64.2 59.3 57.1 54.8 53.3 52.5

1:3.18:7.51 1:3.97:7.24 1:4.40:7.21 1:5.21:7.54 1:5.91:7.90 1:6.17:7.96

solar cells, fuel cells, batteries and corrosion [23,24]. This technique has the potential to provide information about physical and electrical properties of semiconductor materials such as, type of semiconductor, charge carriers density (N), and flat band potential (Vfb) [25]. The objective in electrochemical impedance spectroscopy is to correlate features of impedance spectra with their underlying micro-structural origins by means of appropriate and reasonable equivalent circuit. All these considerations explain our interest for studying the electrochemistry of CuIn5S8 films by EIS. The aim of this paper is to investigate the effect of [Cu]/[In] ratio in the electroplating solution on the structural, morphological and stability of the onestep potentiostatically deposited CuIn5S8 thin film onto conductive and transparent indium tin oxide (ITO)-coated glass substrates from an acidic solution. These fundamental investigations are aiming at the clarification of electrochemical behavior and physical phenomena taking place at the CuIn5S8 electrode/Na2SO4 electrolyte solution interface utilizing electrochemical impedance spectroscopy. 2. Experimental

2.2. Characterizations The crystal phase of the films was analyzed using an X-ray diffractometer (automated Bruker D8 advance) with CuKa (l = 1.541 Å´) radiations in the 2u range of 10–70 . The obtained structural data were examined with X'Pert HighScore Plus program. The morphology of samples was analyzed using a scanning electron microscope (SEM, JEOL JSM-6700), with an accelerating voltage of 15 kV. The impact of the precursor’s molar ratio onto the composition of samples was estimated by X-ray fluorescence (XRF) using a spectrometer MagiX PW2403. 2.3. Electrochemical measurements The measurement of electrochemical impedance investigations and Mott–Schottky plots were carried out using a computercontrolled potentiostat (PGSTAT 30) equipped with a frequency response analyzer and connected to a three-electrode cell (as described above). The surface area of the working electrode is about 0.5 cm2. All electrochemical measurements were released in an electrolyte solution of 0.5 M Na2SO4 (pH 6.5) [26]. EIS measurements were carried out at the open-circuit potential. A

2.1. Materials preparation !

30

40

* CuInS2

311

. In

50

60

70

ITO

ITO

ITO

ITO

20

ITO

10

.

.

Cu/In=0.22

. Cu/In=0.25

Cu/In=0.28

! (a.u) Intensity

*

Cu/In=0.33

*

*

Cu/In=0.40

*

Cu/In=0.50

10

20

30 40 50 Bragg angle 2 (degree)

533 622 444

620

440 531

422 511

!

331

400

222

220

311

*

111

The electrodeposition has been carried out potentiostatically using an Autolab potentiostat/galvanostat PGSTAT 30 (Eco Chemie BV) connected to a three-electrode cell (K0269A Faraday Cage, Par). The working electrode was (ITO)-coated glass substrate (average area = 1 cm2,r  5.0  10
5 V cm), the reference electrode was an Ag/AgCl (3 M NaCl) and a platinum plate was used as counter electrode. Before using, all (ITO)-coated glass substrates were ultrasonically cleaned during 15 min with acetone and isopropanol, respectively and rinsed with deionized water and finally dried in air at room temperature. The copper to indium molar ratio [Cu]/[In] in precursor solution was varied during the deposition of samples. The solution baths, which were well stirred, contained 1 mM CuCl2, (Sigma–Aldrich, 97%) for copper, 2–4.5 mM InCl3, (Sigma–Aldrich, 98%) for the indium and 10 mM Na2S2O3 5H2O, (Fluka, 99%) (check Table 1). All the precursors were dissolved in deionized water with 0.25 M KCl as the supporting electrolyte. The pH of the solution was adjusted to 3.0 by adding drops of concentrated 1.0 M HCl in order to decrease the formation of metal complexes such as In(OH)3. The uniform and well adherent CuIn5S8 spinel thin films were deposited at optimized deposition potential of
1.0 V (vs. Ag/AgCl). Details of the deposition process are reported in a recent past work [17]. The deposition time (td) was kept at 15 min with magnetically stirring [4]. The obtained samples have been rinsed with deionized water and then dried in air at a room temperature. Due to the amorphous nature of as-deposited films and in order to improve their crystallinity, all as-deposited films have been annealed in N2 atmosphere at 350  C for 60 min. The detailed analysis of thermal treatments on the crystal structure of CuIn5S8 thin films has been investigated in past work [5].

60

CuIn5S8 JCPDS N.24-361

70

Fig. 1. XRD patterns of electrodeposited CuIn5S8 thin films onto ITO-(glass) substrate prepared from mixing precursor with [Cu]/[In] ratios between 0.22 and 0.5.

M. Gannouni et al. / Materials Research Bulletin 61 (2014) 519–527

sinusoidal AC perturbation of 10 mV was applied to the electrodes over the frequency range of 0.01–105 Hz. Mott–Schottky plots were carried out in the range of
1 to 0 V (vs. Ag/AgCl) reference electrode under frequency of 1 kHz. 3. Results and discussion 3.1. Structural analysis Fig. 1 provides a comparison of typical XRD patterns of CuIn5S8/ ITO thin films obtained from different [Cu]/[In] ratios in the solution bath. As can be seen in this figure, all samples have shown CuIn5S8 peaks with preferential orientation along the (3 11) plane at around 2u = 27.84 (JCPDS card no 24-361) [27]. The crystal phase of the CuIn5S8 films is affected by varying the copper to indium molar ratio in precursor solution. Otherwise, the crystal phase of samples changed from in-poor- to in-rich samples with a decrease in the [Cu]/[In] molar ratio in the solution bath. In-poor samples obtained from precursor solution (0.33  [Cu]/[In] 0 .5) show a solid mixtures of spinel (CuIn5S8)/chalcopyrite (CuInS2). The peaks corresponding to the secondary phase CuInS2 are located at 32.41 and 46.33 (JCPDS card no. 65-2732) [28]. With

521

decreasing [Cu]/[In] ratio up to 0.28, the signal from this secondary phase decreases and finally disappears, which indicates that the chalcopyrite phases may play an important promotion role in the formation of CuIn5S8 via electrodeposition technique. This result is almost in agreement with that observed by Hodes et al. [29], which have indicated that CuIn5S8 spinel phase starts to be formed and mixed with some chalcopyrite when the Cu/In ratio is made less than 0.5. For sample deposited from molar ratio of ([Cu]/ [In] = 0.28), XRD pattern shows only the peaks associated to the CuIn5S8 cubic spinel phase, indicating that the crystalline nature of the CuIn5S8 improves at this corresponding molar ratio. Indeed according to Binsma et al. [30], the increase of relative amounts of indium causes the progressive substitution of copper into tetrahedral sites. This substitution allows the transition from a mixed CuIn5S8/CuInS2 structure, to a single CuIn5S8 structure, in which copper and indium are statistically disordered on the tetrahedral sites. With a [Cu]/[In] molar ratio lesser than 0.28, two peaks at about 2u = 32.92 and 36.44 are observed, and identified to the elemental In (JCPDS card no. 085-1409) [31] in these in-rich samples. From Fig. 2(a), we note also, a small shift in the (3 11) preferential orientation to higher angles with the decrease in [Cu]/ [In] ratios, indicating that In3+ ions were incorporated into the

Fig. 2. (a) Enlarged 3 11 main peak vs. Bragg angle 2u at various [Cu]/[In] ratios, (b) 3 11 intensity peak and FWHM vs. [Cu]/[In] ratio.

522

M. Gannouni et al. / Materials Research Bulletin 61 (2014) 519–527

Fig. 3. SEM micrographs of CuIn5S8 thin films electrodeposited onto ITO-(glass) substrate at different [Cu]/[In] ratios between 0.22 and 0.5.

lattice of samples by increasing the Indium content. As can be seen in Fig. 2(b), this preferential orientation intensity do not increase monotonously with [Cu]/[In] ratios. The strongest, well-defined (3 11) reflection was observed for sample obtained from [Cu]/ [In] = 0.28, indicating high crystalline quality. The low intensity and broadness of (3 11) peaks for in-poor and in-rich films suggest their worse crystallinity, which may be ascribed to the large compositional deviation from the stoichiometry and/or the coexistence of multi-phase due to inhomogenous Indium incorporation. Both, the change in the (3 11) main peak intensity and the

successive shift may be due to the increasingly irregular arrangement of the lattice structure and to the occurrence of intrinsic defects in the crystal structure following the increase of the amount of indium. 3.2. Morphology and stoichiometry analyses Fig. 3 shows the scanning electron micrographs of CuIn5S8 thin films electrodeposited onto ITO-glass substrate with different [Cu]/ [In] ratios. As can be seen from this figure, the effect of [Cu]/[In]

Fig. 4. SEM cross-sectional image of CuIn5S8 thin film grown onto ITO-(glass) substrate with [Cu]/[In] = 0.28.

M. Gannouni et al. / Materials Research Bulletin 61 (2014) 519–527

ratios is visible on the aspect of the surface film. All microstructures consist of dense layers with bigger grains and/or large agglomerates. As illustrated in Fig. 3((a)–(c)), films prepared with molar ratio between (033  [Cu]/[In]  0.5) show rough organization with some holes on the surface. The micrographs of these films reveal non-uniform grain sizes and non-homogeneous surface with mixture of smaller and larger clusters. This effect indicates that possible formation of the CuInS2 segregated phase at the surface was caused by the lower In content in these film. For the micrograph of film obtained from ([Cu]/[In] = 0.28) (Fig. 3(d)), we notice a reduction of clusters size giving a densely packed and a nearly homogeneous surface consists of large agglomerates. This reduction reinforces the densification of the CuIn5S8 films and reduces the leakage current due to grain boundaries. When [Cu]/ [In] ratio departures from 0.28, the films return to a poor surface morphology displaying grains/particles with uneven sizes. This is clearly seen from Fig. 3((e) and (f)), in which isolated white particle are formed which might be results to the in-rich phase. On the other hand, Fig. 4 provides a typical cross-sectional SEM image of the CuIn5S8/ITO film derived from ([Cu]/[In] = 0.28), from which, it can be seen that the average agglomerates size and film thickness are likely to be in the micrometer range for all films. These features are in agreement with those observed by AFM analysis in our previous work [4]. Table 1 lists the composition analysis results for CuIn5S8 film with various [Cu]/[In] ratios. As expected from this study, the indium atomic ratio in our samples increases with indium concentration in the solution bath. As can be seen in Table 1, when the [Cu]/[In] ratio in solution bath varies from 0.5 to 0.33, XRF analysis shows a composition ratio of [Cu]:[In]:[S] varying from 1:3.18:7.51 to 1:4.40:7.21, indicating that an indium-poor CuIn5S8 is deposited on the ITO substrate for high [Cu]/[In] ratios. When the precursor copper-to-indium ratio in the plating bath is maintained at 0.28, the chemical composition tends to improve in the film and shows a nearly stoichiometric composition ratio of [Cu]:[In]:[S] 1:5.21:7.54. While, in case of samples deposited at low precursor copper-to-indium ratio in the plating bath (0.22  [Cu]/ [In]  0.25), the chemical composition shows again a deviation from the stoichiometric with a composition ratio of [Cu]:[In]:[S] 1:6.17:7.96, indicating that an indium-rich CuIn5S8 is deposited on the ITO substrate. As a result from this study, XRF analysis clearly indicate the impact of the precursor’s molar ratio onto the composition of electrodeposited CuIn5S8 films, and are also in good agreement with XRD analyses mentioned above.

523

presence of two time constants and are likely to have the same magnitude of frequency as that observed in phase angle peaks. Fig. 6 illustrates typical Nyquist plots (imaginary impedance (
Z00 ) vs. real impedance (Z0 )) of the CuIn5S8 film in neutral 0.5 M Na2SO4 solution for different [Cu]/[In] ratios. Clearly, the Nyquist diagrams exhibit two semicircles, in high and low frequencies with their diameters changing upon varying the [Cu]/[In] ratio (as evident in the phase angle plot, Fig. 5(a)). This indicates that the CuIn5S8 semiconductor/electrolyte interface is composed of at least two different electrochemical processes. If we do not take into account the role played by surface states and the impedance of counter electrode (small area 0.5 cm2), the first hemisphere at high frequencies could be probably associated with a space charge, and the second one at low frequencies to the double layer. This enables the determination of two capacitive contributions involved in the measured impedance. This assumption could be interpreted as follows: after the contact of the semiconductor surface with the electrolyte, the thermodynamic equilibrium is established on both sides of the interface. This equilibration happens through electron transfer across the interface, which results in the formation of the space charge layer inside the semiconductors characterized by CSC,

3.3. Impedance studies of CuIn5S8 film in neutral 0.5 M Na2SO4 solution Fig. 5, shows representative Bode (modulus and phase angle vs. frequency) diagrams of CuIn5S8 in neutral 0.5 M Na2SO4. One can observe two phase angle peaks in the Bode plot (Fig. 5(a)) that correspond to two constant phases. The first one on the left hand side is related to low frequencies, the phase angle magnitude changes between (
48 and
66 ), which is less negative than that of a perfect capacitor (
90 ). The second one on the right hand side is determined at high frequencies, the correspondent phase angle magnitude remains nearly constant in the vicinity of
77 and
85 (close to
90 ), with a shift toward lower frequencies by decreasing the [Cu]/[In] ratios. The change of the phase angle magnitudes and the peak shift may originate from frequency dispersion and the easy transfer of charge in presence of higher contents of In, respectively. This behavior will be detailed in the next part. The modulus-impedance vs. frequency plots (Fig. 5(b)), also exhibits tow maximums. These two maximums clearly indicate the

Fig. 5. Bode plots of CuIn5S8 thin films electrodeposited onto ITO-(glass) substrate at different [Cu]/[In] ratios between 0.22 and 0.5.

524

M. Gannouni et al. / Materials Research Bulletin 61 (2014) 519–527

Fig. 6. Nyquist diagrams for CuIn5S8 film in 0.5 M Na2SO4 at different [Cu]/[In] ratios in the range of 0.22– 0.5.

and the electric Helmholtz double layer inside the electrolyte solutions described by CH [41]. Because the number of available states per unit energy in the electrolyte solutions (highly concentrated) far exceeds the number present in a semiconductor, it can be assumed that the width of the space charge layer is much

larger than the width of the Helmholtz layer, which yields CSC at high frequencies much lesser than CH at low frequencies. To make out the electrical property changes of different layers, it is necessary to establish an equivalent circuit model to fit the measured impedance spectra. This is based on the fact that the

M. Gannouni et al. / Materials Research Bulletin 61 (2014) 519–527

525

Fig. 7. Equivalent circuit compatible with the experimental impedance data in Figs. 5 and 6 for CuIn5S8 electrodes in 0.5 M Na2SO4.

parallel combination of a resistor and a capacitor produces a semicircle in the Nyquist plot [32]. Furthermore, to obtain a satisfactory impedance simulation of CuIn5S8 in Na2SO4 solution, constant phase element (CPE) is more suitable to describe the behavior of a non-ideal capacitor where capacitance depends on frequency. In fact, as described earlier, CuIn5S8 electrodes show rough surface with some inhomogeneities, which may cause frequency dispersion due to heterogeneous current density distribution. Considerable research on this subject has been reported previously [33–35]. The impedance of the CPE element is given by the formula [36]: Fig. 8. Mott–Schottky (M–S) plots (1 /C2 vs. V) of CuIn5S8 film in 0.5 M Na2SO4.

1 ¼ Q 0 ðjwÞn Z

(1)

where Q0 is constant parameters and the parameter n can assume values between 0 and 1. When n = 1, the CPE is a pure capacitance. The CPE behavior can be quantified by plotting the imaginary part of the impedance as a function of frequency in logarithmic coordinates (which is not shown here). In this study the slopes (which means the exponent n) are esteemed in the range of (0.93–0.99) and (0.90–0.98) in high and low frequencies, respectively (check Table 2). According to Brug et al. [37], the corresponding capacity can be extracted from the CPE parameters n, and Q0 by the following relation: 1

C¼

ðRQ 0 Þn R

(2)

Fig. 7, presents the equivalent circuit compatible with the experimental impedance measurements. In this electrical equivalent circuit, Rs represents the intercept value of the impedance spectrum at high frequency side with the real axis, which corresponds to the electrolyte resistance. Q1 and R1 at high frequency represents a constant phase element corresponding to space charge capacitance and the space charge resistance, respectively. Q2 and R2 at low frequency are the electrical elements related to double layer capacitance and charge transfer resistance. Table 2 illustrates the equivalent circuit parameters for the impedance spectra of CuIn5S8 in different [Cu]/[In] ratios. As can be inferred from these data, the electrolyte resistance, Rs, remains constant over the entire [Cu]/[In] ratio range with the values between 81 V and 89 V. A decrease of the [Cu]/[In] ratio resulted in a significant reduction of the space charge resistance R1 (from 1398 V to 625 V), and in an elevation of the space charge capacity

Q1 in the range 0.12 mF–4.96 mF. Such effect could be explained by the rapid transfer charge in the CuIn5S8 semiconductor electrode and by the high accumulation of electrons in presence of higher contents of indium. Another noticeable result is the huge value of R2 and Q2 corresponding to the double layer. Both R2 and Q2 values are almost in the same order of magnitude over the entire [Cu]/[In] ratios range, and varying in the range of 3332–4255 V and 14.56–56.20 mF, respectively. The Mott–Schottky theory is commonly used to evaluate the doping concentration and flat-band potential based on the equation [38]:  
kT 1 2 ¼ V
V (3)
 fb e C 2SC ee0 S2 N where, e, e0, e, k and T have their usual meaning dielectric constant of the CuIn5S8 (e = 10) [39], permittivity of a vacuum, elementary electric charge, Boltzmann constant and absolute temperature, respectively, whereas S denotes the surface area of the electrode (S = 0.5 cm2), N stands for the density of donor (ND) or acceptor (NA) in the semiconductor, V is the externally applied potential with respect to the reference electrode and Vfb is the flat band potential, i.e. the potential corresponding to the situation in which there is no accumulation of charge in the semiconductor and the energy bands show no bending. The shape of the Mott–Schottky plot is an indicator of the conductivity type of the semiconductor, where, negative sign is for p-type and positive for n-type semiconductor [40]. The value of the free charge carrier concentration can be calculated from the slope of 1/CSC2 vs. V plot and Vfb can be obtained from the intercept on the potential axis (1/CSC2 = 0). These values are important in determining the efficiency of a semiconductor as an electrode in photoelectrochemical cells. There are two capacitances to be considered, that of the space

Table 2 Equivalent circuit parameters of CuIn5S8 film with different [Cu]/[In] ratios in Na2SO4 Electrolyte, with relative errors of 4–5%. [Cu]/[In]

Rs (V)

R1 (V)

Q1 (mF)

n1

R2 (V)

Q2 (mF)

n2

0.5 0.4 0.33 0.28 0.25 0.22

89 84 81 87 88 81

1398 1371 898 738 697 625

0.12 0.28 0.81 1.67 2.23 4.96

0.99 0.98 0.93 0.97 0.95 0.96

3332 4062 3964 3979 3908 4255

20.77 50.12 27.14 56.20 21.05 14.56

0.90 0.93 0.94 0.98 0.95 0.91

526

M. Gannouni et al. / Materials Research Bulletin 61 (2014) 519–527

Table 3 Mott–Schottky parameters for CuIn5S8 film with different [Cu]/[In] ratios in Na2SO4 electrolyte [Cu]/[In]

Vfb(V)

ND (x1016 cm-3)

LD(x10
8 m)

WD(x10
7 m)

Type

0.5 0.4 0.33 0.28 0.25 0.22


0.44
0.46
0.48
0.50
0.51
0.51

1.23 2.38 2.84 34.22 41.1 74.3

3.40 2.44 2.24 0.64 0.58 0.43

2.98 2.15 1.97 0.56 0.51 0.38

n n n n n n

charge region and that of the double layer. Since these capacitances are in series, the total capacitance is the sum of their reciprocals. As the space charge capacitance is much smaller than double layer capacitance when the measured frequency is high enough (>1 kHz), the contribution of the double layer capacitance to the total capacitance is negligible. Therefore, the capacitance value calculated from this model is assumed to be the value of the space charge capacitance. Thus, the validity of the Mott–Schottky analysis is based on the assumption that the capacitance of the semiconductor/electrolyte interface is that of the capacitance of the space charge semiconductor. Moreover, as described above the exponent (n) of CPE pertaining to space charge capacitor (Q1 in Table 2) was found to be in the range 0.93–0.99, hence the CPE can still be considered as a capacitor for the purpose of generating Mott–Schottky plot [41]. As shown in Fig. 8, the Mott–Schottky plots for the CuIn5S8 in 0.5 M Na2SO4 electrolytes at various [Cu]/[In] ratios display a quasilinear behavior (due to the heterogeneous in composition) at potentials more superior than
0.5 V vs. Ag/AgCl with a positive slope, indicating that we are dealing with an n-type CuIn5S8 semiconductor. This behavior is associated with the most probable defects VS, InCu which can introduce two donor levels. As can be seen from Table 3, both values of Vfb and ND (the density of donor) are altered by varying the amount of indium. The estimated flat band potentials (Vfb) of the samples lie in the range of
0.44 V to–0.51 V and became more negative with the decrease in the [Cu]/[In] ratio. The carrier densities of the samples obtained from the Mott–Schottky plots are in the range of 1.23  1016 cm
3 to 7.43  1017 cm
3 and tend to increase with decreasing [Cu]/[In] ratio. This change in the values of Vfb and ND is likely to depend on the synthesis methods, surface states modifications and the composition of elements in samples. The experimentally determined carrier density, flat band potentials (Vfb), and type of our samples should be compared to those obtained in previous investigations. Using Hall measurement, Endo et al. [42] have found that Culn5S8 single crystal is n-type semiconductor and has a carrier concentration in the range of 6.3  1013
1.2  1019 cm
3. On the other hand, Makhova et al. [43] have prepared CuIn5S8 spinel thin films by sulphurization of Cu–In alloy films in sulfur vapor, and by using Van der Pauw technique, they have shown an n-type conductivity, and a carrier concentration of 20  1017 cm
3. In another study, Cahen et al. [44] have studied the photochemical response of CuIn5S8 semiconductor/polysulfide interface, and by means of Mott–Schottky plots, their results have shown an n-type conductivity of this semiconductor, an abnormally high carrier concentration in the range of 1021–1022 cm
3 and a flat-band potential in the order of
0.4,
0.45 V. Within the Mott–Schottky approximation, the thickness of the space charge layer WD and Debye length LD are given by [45]:  W¼

2ee0 e ND




kT V
V fb
e

1=2 (4)

LD ¼

ee0 kT e2 N D

(5)

Both LD and WD, corresponding to V
Vfb = 1 V, are collected in Table 3. From this table, with higher carrier concentration and increased capacitance, it is expected to have a decreased space charge width and Debye length. 4. Conclusions In this study, CuIn5S8 ternary compound was deposited on ITOcoated glass substrates using one-step electrodeposition. The influence of the [Cu]/[In] molar ratio in the solution bath on the structural, morphological, and electrochemical behavior in inert Na2SO4 electrolyte of samples was investigated. These results indicate that the ratio of [Cu]/[In] in the precursor solution bath plays an important role in controlling the microstructure of films produced using the electrodeposition route. XRD results show that the samples consist of the mixed phase of chalcopyrite and spinel CuIn5S8 for [Cu]/[In] ratios of less than 0.28 in solution bath. The single CuIn5S8 spinel phase and nearly homogeneous surface as well as nearly stoichiometric composition were obtained from the starting molar ratios of [Cu]/[In] = 0.28 in the precursor solution. Electrochemical impedance spectroscopy was employed to analyze the capacitive nature of CuIn5S8/Na2SO4 interface. The interface was modeled by an equivalent circuit consisting of two parallel resistor–– capacitor blocks in series that signify semiconductor and electrolyte sides of the interface. Following this electrochemical study, we show that the [Cu]/[In] ratio has a significant influence on the equivalent circuit parameters. From the Mott–Schottky plot, it was concluded that all samples show an n-type conduction with a flat potential in the range of
0.44 V to –0.51 V, which can be shifted negatively by varying the starting molar ratios of [Cu]/[In]. The density of majority carrier was found to be in the range of 1.23  1016 cm
3 to 4.1 1017 cm
3 and can be increased by decreasing the [Cu]/[In] ratio. These changes can be correlated with the synthesis methods, solid XRD and XRF composition. Future investigations focus on the effect of [Cu]/[In] ratio on the photoelectrochemical responses and solar cell properties of these CuIn5S8. The work is presently under way and the results will be reported subsequently. References [1] K.W. Cheng, C.M. Huang, G.T. Pan, W.S. Chang, T.C. Lee, T.C.K. Yang, Chem. Eng. Sci. 65 (2010) 74–79. [2] K.W. Cheng, C.M. Huang, Y.L. Huang, H.J. Chuang, Y.C. Wu, Thin Solid Films 520 (2011) 469–474. [3] J. Flahaut, L. Domange, M. Guittard, M. Ourmitchi, J. Kamsu, Bull. Soy. Chim. Fr. 28 (1961) 2382. [4] M. Gannouni, I. Ben Assaker, R. Chtourou, Superlattice. Microstruct. 61 (2013) 22–32. [5] M. Gannouni, M. Kanzari, J. Alloys Compd. 509 (2011) 6004–6008. [6] A. Usujima, S. Takeuchi, S. Endo, T. Irie, Jpn. J. Appl. Phys. 20 (1981) L505–L507. [7] M. Gannouni, M. Kanzari, Appl. Surf. Sci. 257 (2011) 10338–10341. [8] L. Makhova, R. Szargan, I. Konovalov, Thin Solid Films 472 (2005) 157–163.

M. Gannouni et al. / Materials Research Bulletin 61 (2014) 519–527 [9] F. Py, J.O. Fourcade, J.C. Jumas, J. Solid State Chem. 99 (1992) 319–328. [10] N.M. Gasanly, S.A. El-Hakim, L.G. Gasanova, A.Z. Magomedov, Phys. Status Solidi (b) 158 (1990) k1–k4. [11] L.V. Makhova, I. Konovalov, R. Szargan, Phys. Status Solidi (a) 201 (2004) 308–311. [12] A.F. Qasrawi, N.M. Gasanly, Cryst. Res. Technol. 38 (2003) 1063–1070. [13] R. Nomura, Y. Seki, H. Matsuda, Thin Solid Films 209 (1992) 145–147. [14] R.N. Bhattacharya, D. Cahen, G. Hodes, Sol. Energ. Mater. 10 (1984) 4145. [15] S. Nakamura, A. Yamamoto, Sol. Energ. Mater. Sol. Cells 75 (2003) 81–86. [16] A.M. Martinez, A.M. Fernandez, L.G. Arriaga, U. Cano, Mater. Chem. Phys. 95 (2006) 270–274. [17] M. Gannouni, I. Ben Assaker, R. Chtourou, J. Electrochem. Soc. 160 (8) (2013) H446–H451. [18] I.M. Dharmadasa, J. Haigh, J. Electrochem. Soc. 153 (2006) G47–G52. [19] G. Dagan, S. Endo, G. Hodes, G. Sawatzky, D. Cahen, Sol. Energ. Mater. 11 (1984) 57–74. [20] G. Hodes, T. Engelhard, D. Cahen, Thin Solid Films 128 (1985) 93–106. [21] G. Hodes, T. Engelhard, J.A. Turner, D. Cahen, Sol. Energ. Mater. 12 (1985) 211–219. [22] R.S. Becker, T. Zheng, J. Elton, J. Electrochem. Soc. 132 (1985) 1824. [23] S.M.R. Niya, M. Hoorfar, J. Power Sources 240 (2013) 281–293. [24] L. Zhang, F. Liu, K. Brinkman, K.L. Reifsnider, A.V. Virkar, J. Power Sources 247 (2014) 947–960. [25] A. Vijayakumara, T. Du, K.B. Sundaram, Appl. Surf. Sci. 242 (2005) 168–176. [26] A. Aghassi, M. Jafarian, I. Danaee, F. Gobal, M.G. Mahjani, J. Electroanal. Chem. 661 (2011) 265–269. [27] M. Robbins, M.A. Miksovsky, Mater. Res. Bull. 6 (1971) 359–364. [28] Y. Zou, D. Li, D. Yang, Mater. Chem. Phys. 132 (2012) 865–869.

527

[29] G. Hodes, T. Engelhard, C.R. Herrington, L.L. Kazmerski, D. Cahen, Progr. Cryst. Growth Char. 10 (1984) 345–351. [30] J.J.M. Binsma, L.J. Giling, J. Bloem, J. Cryst. Growth 50 (1980) 429. [31] H.E. Swanson, R.K. Fuyat, Natl. Bur. Stand. (U. S.), Circ. 539 (1954) 3. [32] A.J. Bard, L.R. Faulkner, Electrochemical Methods, Fundamentals and Applications, Second ed., John Wiley & Sons, 2001. [33] G. Oskam, D. Vanmaekelbergh, J.J. Kelly, J. Electroanal. Chem. 315 (1991) 65–85. [34] Y.B. Wang, R.K. Yuan, M. Wilander, Appl. Phys. A: Mater. Sci. Process 63 (1996) 481–486. [35] Z. Hens, W.P. Gomes, Phys. Chem. Chem. Phys. 1 (1999) 3617–3625. [36] O. Dhibi, A. Ltaief, S. Zghal, A. Bouazizi, Superlattice. Microstruct. 60 (2013) 548–560. [37] G.J. Brug, A.L.G. van den Eeden, M.S. Rehbach, J.H. Sluyters, J. Electroanal. Chem. 176 (1984) 275. [38] K.W. Cheng, Y.C. Wu, Y.T. Hu, Mater. Res. Bull. 48 (2013) 2457–2468. [39] S. Kitamura, S. Endo, T. Irie, J. Phy. Chem Solids 46 (1985) 881–885. [40] G.T. Pan, M.H. Lai, R. ChengJuang, T.W. Chung, T.C.K. Yang, Sol. Energ. Mater. Sol. Cells 94 (2010) 1790–1796. [41] W.P. Gomes, D. Vanmaekelbergh, Electrochim. Acta 41 (1996) 967–973. [42] S. Endo, S. Kitamura, T. Irie, S. Takeuchi, Il NuovoCimento D 2 (1983) 1875–1879. [43] L. Makhova, R. Szargan, I. Konovalov, Thin Solid Films 472 (2005) 157–163. [44] D. Cahen, G. Dagen, G. Hodes, Y. Mirovsky, Y.W. Chen, J.C.W. Folmer, P.J. Ireland, R. Noufi, J.A. Turner, K.J. Bachmann, S. Endo, C. Rincon, G.A. Sawatzky, M. Tomkiewicz, Progr. Cryst. Growth Char. 10 (1984) 263–270. [45] H. Wang, L. Pilon, Electrochim. Acta 63 (2012) 55–63. [46] N. Khemiri, M. Kanzari, Solid State Commun. 160 (2013) 32–36. [47] L. Makhova, I. Konovalov, Thin Solid Films 515 (2005) 5938–5942.