Structural, morphological and optical properties of rf ...

6 downloads 0 Views 2MB Size Report
Available online 30 March 2016. Cadmium sulfide (CdS) nanocrystalline semiconductor thin films were deposited using the technique of magne- tron sputtering ...
Materials and Design 100 (2016) 198–203

Contents lists available at ScienceDirect

Materials and Design journal homepage: www.elsevier.com/locate/matdes

Structural, morphological and optical properties of rf – Sputtered CdS thin films Ovidiu Toma a, Lucian Ion a, Sorina Iftimie a, Adrian Radu a, Stefan Antohe a,b,⁎ a b

University of Bucharest, Faculty of Physics, 405 Atomistilor Street, PO, Box MG-11, 077125, Magurele-Ilfov, Romania Academy of Romanian Scientists, 54 Splaiul Independentei, Bucharest, Romania

a r t i c l e

i n f o

Article history: Received 25 November 2015 Received in revised form 21 March 2016 Accepted 22 March 2016 Available online 30 March 2016 Keywords: Cadmium sulfide Spectroscopic ellipsometry Window layers Solar cells RF-sputtering

a b s t r a c t Cadmium sulfide (CdS) nanocrystalline semiconductor thin films were deposited using the technique of magnetron sputtering in radio – frequency plasma. Structural, morphological and optical characterizations of the prepared CdS thin films were carried out. The films were polycrystalline and well textured, with (002) crystallographic planes oriented parallel to the surface (crystallite sizes were between 30.6 nm up to 48.2 nm). RMS surface roughness was in the nanometer range and decreases with increasing film thickness. Optical constants (refractive indices and extinction coefficients) as well as film thicknesses (between 56.4 nm and 174.6 nm) and surface rugosities were computed by spectroscopic ellipsometry. This optical method was combined with optical spectrophotometry (absorption coefficients, optical transmittances, etc.) in UV – VIS – NIR for a better verification of the results. The range for the obtained values of optical band gaps is between 2.3 eV and 2.36 eV. © 2016 Elsevier Ltd. All rights reserved.

1. Introduction Among AII - BVI compounds cadmium sulfide (CdS) is one of the most used materials in electronic and optoelectronic applications, such as thin film transistors (TFTs) [1], light emitting diodes (LEDs) [2], photodetectors [3], gas sensors [4], solar cells [5–7], etc. There is a rich literature on theoretical and experimental investigations on CdS, regarding for example the influence of the various deposition conditions [8–10], the influence of the post–deposition thermal treatments (annealings) [11–12], defects control [13], influence of the ionizing radiations on electrical properties [14,16], influence of the substrate temperatures [17], etc. The physical properties of CdS such as the wide optical band gap of 2.35 eV in its thermodynamically stable würtzite phase, its good chemical and mechanical stabilities, recommend it as an almost ideal semiconductor material for applications in electronics and optoelectronics. Using various growth techniques, it can be cast in different nanostructures like nanowires [18,19], nanorods [20], nanotubes [21], nanobelts [22], etc. Radio – frequency magnetron sputtering is a reliable technique used in CdS thin films deposition, allowing a good control of film uniformity and thickness over large area substrates [23, 24]. Here we report our results on CdS thin films depositon by rfmagnetron sputtering and their characterization. Our films are intended

⁎ Corresponding author at: University of Bucharest, Faculty of Physics, 405 Atomistilor Street, PO, Box MG-11, 077125, Magurele-Ilfov, Romania. E-mail address: santohe@solid.fizica.unibuc.ro (S. Antohe).

http://dx.doi.org/10.1016/j.matdes.2016.03.117 0264-1275/© 2016 Elsevier Ltd. All rights reserved.

for use as n – type semiconductor window layers for thin film based second generation of solar cells, in combination with different absorber layers, such as CdTe, CuInSe2 or ZnTe.

2. Experimental details High purity (99.99%) solid CdS targets of 2 in. diameters and 3 mm thickness provided by Neyco Company were used. The sputtering deposition system from Tectra is capable working both in DC and RF modes. The films were deposited in RF regime using an RF generator working at a frequency of 13.56 MHz. Optical BK 7 glasses were used as substrates. The thicknesses of CdS targets are enough to provide an uniform sputtering onto the optical glass substrates. Before starting the depositions the glass substrates were ultrasonically cleaned in acetone and isopropyl alcohol for 15 min. All the depositions were performed at ambient temperature, using a target – to – substrate distance of 80 mm. Spectral argon was used as working gas, while the deposition pressure inside the sputtering reactor was fixed at 4.6 × 10−3 mbar. The sputtering power was fixed at 50 W during all the depositions [15]. Films with four selected thickness values have been deposited (see Table 1), in order to evaluate the influence of this deposition variable on their structural, morphological and optical properties. The obtained samples were denoted as S1 (56.4 nm thickness), S2 (111.9 nm), S3 (153.5 nm) and S4 (174.6 nm). Quartz crystal monitoring (QCM) technique was employed to measure CdS films thickness during the depositions.

O. Toma et al. / Materials and Design 100 (2016) 198–203

Table 1. Deposition parameters of CdS thin films. Sample number

Sputtering time (min)

S1 S2 S3 S4

5 10 14 15

Thickness as computed by spectroscopic ellipsometry (nm) 56.4 111.9 153.5 174.6

The structural properties of CdS thin films deposited by rf – magnetron sputtering were investigated by grazing incidence X – ray diffraction (GIXRD) using a Bruker D 8 Discover diffractometer. The monochromatised CuKα1 radiation (λ = 1.5406 Ǻ) was used. The

199

scattered intensity was scanned in the 2θ range between 10° - 70° with a step size of 0.08°. The morphological investigations of CdS thin films were made by atomic force microscopy (AFM) using a microscope from TopoMetrix at a resonance frequency of 320 kHz. The silicon cantilever was used in the contact mode, working in air, with a force of 10 pN (the lateral resolution was around 20 nm). The surface morphology was also investigated by cross – section scanning electron microscopy (Cross – Section SEM) using a Tescan Vega XMU – II microscope. The optical properties of the films were investigated by spectroscopic ellipsometry (SE) technique using a variable angle spectroscopic ellipsometer Woollam WVASE 32 coupled with a scanning monochromator HS – 190 with the spectral range between 250 nm up to 1.7 μm. The ellipsometric data were coupled with additional optical data obtained by UV – VIS – NIR optical spectrophotometry using a double beam Perkin Elmer Lambda 35 spectrophotometer, with the spectral range between 190 nm and 1.1 μm. 3. Results and discussion

Fig. 1. GIXRD patterns of CdS films S1-S4 (from bottom to top).

Diffraction patterns recorded in grazing incidence geometry (GIXRD), at an incidence angle of 2°, are shown in Fig. 1. This geometry is commonly used for phase analysis of thinner films, to maximize the signal from the CdS layers by reducing the intensity of the radiation reflected by the substrate. The films are polycrystalline, consist of würtzite phase CdS and are preferentially oriented with (002) crystallographic planes parallel to the surface. This direction has the highest growth rate for the magnetron sputtered CdS thin films [5], but also (110) and (103) peaks are observed in the case of thicker films. Microcrystalline properties of the films were investigated by performing a profile analysis of (002) reflection recorded in Bragg-Brentano theta-theta geometry. Fig. 2 shows the experimental results and the line profile as obtained by fit. A double-Voigt method [25] was used: the peaks were fit analytically using Voigt profiles and the Lorentzian and Gaussian breadths were determined and corrected for instrumental broadening. The red line is the result of the fit with Voigt function. The residuals of the fit procedure are also shown for each case. For determining the crystalline coherence length (crystallite size) and strain, it was assumed that the first is given by the Lorentzian integral breadth, while the micro-strain distribution is Gaussian. The dependence of the crystallite size Def and mean-square strain bε2 N 1/2 on the thickness of the films is shown in Fig. 3. A trend can be observed, with both the crystallite size and the microstrain being reduced with increasing the film thickness. In the growth

Fig. 2. Line profile analysis for (002) peak, measured in Bragg-Brentano theta-theta geometry, for films S1 (a), S2 (b), S3 (c) and S4 (d).

200

O. Toma et al. / Materials and Design 100 (2016) 198–203 Table 3 Surface roughnesses, Skewness parameter (Ssk) and Kurtosis parameter (Sku) obtained for rf-magnetron sputtered CdS thin films. Sample S1 S2 S3 S4

Fig. 3. Dependence of the crystallite size and mean-square strain corresponding to (002) reflection on the film thickness.

Table 2 Structural parameters of CdS films. Sample S1 S2 S3 S4

a (Å)

c (Å)

D(002) ef (nm)

bε2 N 1/2

4.118 4.162 4.127 4.188

6.728 6.742 6.749 6.750

48.2 54.3 35.7 30.6

1.01 × 10−2 6.72 × 10−3 4.90 × 10−3 6.22 × 10−3

conditions considered here, due to the low temperature of the substrate during growth, the migration of the atoms adsorbed at the growing film surface is limited. The influx of (adsorbed) atoms is relatively high, which is typical for plasma based deposition techniques. Consequently, adsorbed atoms don't have enough time to reach their equilibrium positions in the crystalline structure, structural defects are forming and

RMS roughness (nm)

Ssk

Sku

3.67 2.84 2.36 1.67

0.44 0.28 0.09 0.06

3.29 2.84 2.76 2.69

their accumulation results in mechanical stress building up in (002) oriented columnar crystallites during film growth. This can result in altering the normal packing of (002) crystalline planes. It is the relaxation of this stress (at newly developed grain boundaries) which can result in the observed decrease of the crystalline coherence length. All structural parameters are collected in Table 2. For comparison, ideal CdS lattice constants are aid = 4.141 Å and cid = 6.7198 Å (PDF2 41–1049). Variation of calculated lattice constants with respect to ideal values is due to a macro-strain effect; here a tensile macro-strain can be observed along the direction perpendicular to (002) crystalline planes. The morphological characterizations of the surfaces corresponding to CdS thin films deposited by radio – frequency assisted magnetron sputtering were made using atomic force microscopy (AFM) and cross – section scanning electron microscopy (Cross SEM). Bi – dimensional images for CdS thin films deposited with different sputtering times are shown in Fig. 4. As can be seen from these images the CdS surfaces have compact and very uniform aspects, with grains height at the order of 1–2 nm. Due to the fact that magnetron sputtering works at a molecular level, very small particles of CdS are ejected onto the optical glass substrates leading to very tight and uniform patterns of the films. Quantitative data regarding important surface parameters (such as surface roughness, statistical parameters corresponding to the peaks distributions over the film surfaces, etc.) for CdS thin films were extracted from the AFM images showed in Fig. 4, by using a specialized image processing software (WSxM 4.0). All these data are collected in Table 3. One can observe a slight decrease of CdS films roughness with increasing the sputtering time (the film thickness) during depositions. The two statistical parameters collected in Table 3 are: Skewness (Ssk) parameter

Fig. 4. AFM bidimensional images for sputtered CdS thin films. Two different sampling areas were used for AFM images corresponding to each CdS films, respectively, 1 × 1 [μm x μm] for upper row images and 5 × 5 [μm x μm] for lower row images.

O. Toma et al. / Materials and Design 100 (2016) 198–203

201

Fig. 5. Surface histograms for sputtered CdS thin films: sample S1 (up left), S2 (up right), S3 (down left) and S4 (down right).

which is a 3rd order statistical parameter used to describe how symmetric a statistical distribution is (zero value is for the perfect Gaussian distribution), respectively, Kurtosis (Sku) parameter which is a 4th order

statistical parameter used to describe how sharp or how broad a statistical distribution is (values above 3 indicate sharp distributions, while values below 3 indicate broader distributions).

Fig. 6. Cross section SEM images with film thicknesses for CdS thin films.

Fig. 7. Refractive indices vs. wavelength for rf-magnetron sputtered CdS thin films with different sputtering times (for samples from S2 to S4).

202

O. Toma et al. / Materials and Design 100 (2016) 198–203

Table 4 Spectroscopic ellipsometry results for CdS thin films deposited by rf-magnetron sputtering. Sample Fitting thickness (nm) S2 S3 S4

110 140 170

Mean-square Computed error thickness (MSE) (nm)

Refractive index (at λD

2.965 11.73 8.154

2.501 2.527 2.546

111.9 153.5 174.6

= 590 nm)

Statistical distributions of peaks over the CdS film surfaces were analysed and their corresponding histograms are plotted in Fig. 5. In each histogram the number of statistical events (on vertical axis) is plotted against the surface topography (on horizontal axis and expressed in nm). From the statistical analysis of peaks distributions is observed that thicker CdS films (samples S3 and S4) have more symmetric distributions of peaks and smaller rugosity. Also the surface peaks distribution histograms given for each sample confirm the obtained values for Kurtosis parameter, which indicates that S3 and S4 samples have the broadest peaks distributions while the thinner sample S1 has the sharpest peaks distribution. AFM technique was combined with cross – section SEM morphological investigations showed in Fig. 6. From Fig. 6 it can be seen that all rf-magnetron sputtered CdS thin films with different sputtering times have smooth surfaces and due to the optimal distance of 8 cm chosen between target and substrate during deposition surfaces have no visible damage and there are no drops formed during sputtering. The measured thicknesses of each CdS sample as showed on each SEM image are between 56.16 nm and 171.91 nm. The obtained values are close to the values computed by spectroscopic ellipsometry. The optical properties of CdS thin films deposited by rf-magnetron sputtering with different sputtering times were investigated by spectroscopic ellipsometry (SE). The ellipsometric spectra [26,27] were recorded for three different incident angles of light on the samples (65, 70 and 75°). The selection of these incident light angles was made using a simulation process adapted to the material characteristics, in order to generate more data for the parameters of interest (Ψ and Δ angles). The recorded ellipsometric spectra were fitted by building an appropriate optical model consisting in three layers: a bottom isotropic layer corresponding to the glass substrates, an anisotropic layer for CdS and a top isotropic layer to consider rugosities of CdS films [28]. The dispersion model chosen for fitting (Ψ, Δ) spectra was Adachi – New Forouhi (ANF) model [29]. In Fig. 7 the dependencies of the refractive index on the incident photon wavelengths are plotted for three films. After choosing the optical model the fitting of the ellipsometric spectra was made until a minimum value was obtained for the mean squared error (MSE). Normal dispersion behaviour is observed for all CdS thin films, the refractive index decreasing while increasing the

incident photons wavelengths in the transparent region. The ellipsometric data are given in Table 4. Although the sputtering power was maintained constant at 50 W during all depositions, the variation of the sputtering time slightly influenced the optical constants of CdS films, both real and imaginary parts of the complex refractive index increasing with the increase of the sputtering time (the values of the refractive index given for classic sodium wavelength at 590 nm are between 2.501 and 2.546 when the sputtering time is increased from 10 min. up to 15 min.). The variations of the computed extinction coefficient (the imaginary part of the complex refractive index) with photon wavelengths are shown in Fg. 8. It is observed from the Fig. 8 that the extinction coefficient as computed by spectroscopic ellipsometry for CdS thin films deposited with different sputtering times, decrease rapidly with the increase of the wavelength. The change of graphs shape around 460 nm corresponds to the inter-band transition energy value at 2.7 eV (absorption edge) for all hexagonal CdS samples.The spectroscopic ellipsometry (SE) results were combined with classic spectrophotometry analysis. The transmission spectra for rf-magnetron sputtered CdS thin films are shown in Fig. 9. All CdS thin films have high transmissions (60% to 80%) at long wavelengths, in the transparent region proving that these films are perfectly suitable for manufacturing of high quality semiconductor window layers in second generation PV cells. In Fig. 10, the optical band gap values are computed by using the dispersion relation near the fundamental absorption edge corresponding to direct band gap semiconductors [30]. The values of optical band gaps for CdS thin films of different thickness were obtained by extrapolating the linear portion of the (αhυ)2 vs. hυ graph on hυ axis at α = 0. A short decrease in the optical band gap for rf-magnetron sputtered CdS thin films by increasing the sputtering time can be observed. This is usually attributed both to the grain size growth with the increase of films thickness and also with the reduction in strain when the thickness increased [31]. Since the effect of grain size on the optical band gap is negligible for large grains (like in the case of our samples) the thickness dependence of band gap is due to the observed (as GIXRD studies have confirmed) decrease of lattice micro-strains with the increase of rf-sputtering time. 4. Conclusions CdS thin films with various thicknesses were deposited onto optical glass substrates by rf-magnetron sputtering at ambient temperature. The films were polycrystalline and showed a pronounced texture, with (002) planes parallel to the surface. The crystallite size, as determined from X-ray diffraction data, decreases slightly with increasing film thickness. The surface of the films is relatively smooth and free of drops. All analysed films show an optical transmittance in the range of 60%–90% in the transparent region. The thin films of CdS prepared by rf-magnetron sputtering technique are very well textured, they have superior properties regarding the uniformity in thickness, smooth control

Fig. 8. Extinction coefficients vs. wavelength for rf-magnetron sputtered CdS thin films with different sputtering times (from S2 to S4).

O. Toma et al. / Materials and Design 100 (2016) 198–203

Fig. 9. Transmission spectra for rf-magnetron sputtered CdS thin films.

of thickness, reduced roughness, as compared with those prepared by thermal vacuum evaporation technique. All these properties playing an important role in the reducing of the number of defect states present at the interface of CdS with another A2B6 compound used for electronic and optoelectronic devices, including photovoltaic cells. More that, some preliminary studies not published yet, suggest that this technique applied to CdS thin film lead to more stable film against the ionizing radiation, then an advantage in the case of photovoltaic cells for space application. That is why, the CdS thin films prepared by rf-magnetron sputtering are perfectly suitable candidates for producing of relatively low cost, efficient and stable photovoltaic cells used both in terrestrial and space applications. Acknowledgements Suport from Executive Unit for Financing Education Higher, Research Development and Innovation (UEFISCDI) through grants no. PN II 64/2013 and PN-II 288/2014 is acknowledged. References [1] I. Mejia, A.L. Salas-Villasenor, A. Avendano-Bolivar, J. Horvath, H. Stiegler, B.E. Gnade, M.A. Quevedo-Lopez, Low temperature hybrid CMOS circuits based on chalcogenides and organic TFTs, IEEE Electr. Device L 32 (2011) 1086–1088. [2] R. Grover, R. Srivastava, O. Rana, A.K. Srivastava, K.K. Maurya, K.N. Sood, D.S. Mehta, M.N. Kamalasanan, Electroluminescence from hybrid organic-inorganic LEDs based on thermally evaporated CdS thin films, J. Lumin. 132 (2012) 330–336.

Fig. 10. Tauc plots for calculation of optical band gaps for CdS thin films of different thickness.

203

[3] B. Pradhan, A.K. Sharma, A.K. Ray, A simple hybrid inorganic-polymer photodiode, J. Phys. D. Appl. Phys. 42 (2009) 165308 (4pp). [4] A. Giberti, D. Casotti, G. Cruciani, B. Fabbri, A. Gaiardo, V. Guidi, C. Malagu, G. Zonta, S. Gherardi, Electrical conductivity of CdS films for gas sensing: Selectivity properties to alcoholic chains, Sensor Actuat. B – Chem. 207 (2015) 504–510. [5] O. Toma, S. Iftimie, C. Besleaga, T.L. Mitran, V. Ghenescu, O. Porumb, A. Toderas, M. Radu, L. Ion, S. Antohe, New investigations applied on cadmium sulphide thin films for photovoltaic applications, Chalcogenide Lett. 8 (2011) 747–756. [6] S. Antohe, S. Iftimie, V. Ghenescu, R. Constantineanu, M. Gugiu, M. Ion, I. Stan, A. Radu, L. Ion, Effect of protons irradiation on the performances of CdS/CdTe photovoltaic cells for space applications, Rom. Rep. Phys. 64 (2012) 1153–1162. [7] F. Lisco, P.M. Kaminski, A. Abbas, J.W. Bowers, G. Claudio, M. Losurdo, J.M. Walls, High rate deposition of thin film cadmium sulphide by pulsed direct current magnetron sputtering, Thin Solid Films 574 (2015) 43–51. [8] M. Kim, B. K. Min, C. D. Kim, S. H. Lee, H. T. Kim, S. K. Jung, S. H. Sohn, Study of the physical property of the cadmium sulphide thin film depending on the process condition, Curr. Appl. Phys. 10 (2010) S 455-S 458. [9] L.V. Garcia, M.I. Mendivil, G.G. Guillen, J.A. Aguilar-Martinez, B. Krishnan, D. Avellaneda, G.A. Castillo, T.K. Das-Roy, S. Shaji, CdS thin films prepared by laser assisted chemical bath deposition, Appl. Surf. Sci. 336 (2015) 329–334. [10] J.N. Alexander, S. Higashiya, D. Caskey, H. Efstathiadis, P. Haldar, Deposition and characterization of cadmium sulphide (CdS) by chemical bath deposition using an alternative chemistry cadmium precursor, Sol. Energ. Mat. Sol. C 125 (2014) 47–53. [11] H.E. Maliki, J.C. Bernede, S. Marsillac, J. Pinel, X. Castel, J. Pouzet, Study of the influence of annealing on the properties of CBD-CdS thin films, Appl. Surf. Sci. 205 (2003) 65–79. [12] B. Ghosh, K. Kumar, B.K. Singh, P. Banerjee, S. Das, Growth of CdS thin films on indium coated glass substrates via chemical bath deposition and subsequent air annealing, Appl. Surf. Sci. 320 (2014) 309–314. [13] N. Hernandez-Como, V. Martinez-Landeros, I. Mejia, F.S. Aguirre-Tostado, C.D. Nascimento, G.M. Azevedo, C. Krug, M.A. Quevedo-Lopez, Defect control in room temperature deposited cadmium sulfide thin films by pulsed laser deposition, Thin Solid Films 550 (2014) 665–668. [14] V. Ruxandra, S. Antohe, The effect of the electron irradiation on the electrical properties of thin polycrystalline CdS layers, J. Appl. Phys. 84 (1998) 727–733. [15] C. Besleaga, L. Ion, S. Antohe, AZO thin films synthesized by rf-magnetron sputtering: The role of deposition power, Rom. Rep. Phys. 66 (2014) 993–1001. [16] S. Antohe, L. Ion, V.A. Antohe, M. Ghenescu, H. Alexandru, Defects induced by ionizing radiations in A II – B VI polycrystalline thin films used as solar cell materials, J. Opt. Adv. M. 9 (2007) 1382–1394. [17] J. Schaffner, E. Feldmeier, A. Swirschuk, H.J. Schimper, A. Klein, W. Jaegermann, Influence of substrate temperature, growth rate and TCO substrate on the properties of CSS deposited CdS thin films, Thin Solid Films 519 (2011) 7556–7559. [18] H. Dang, V. Singh, S. Rajaputra, S. Guduru, J. Chen, B. Nadimpally, Cadmium sulfide nanowire arrays for window layer applications in solar cells, Sol. Energ. Mat. Sol. C 126 (2014) 184–191. [19] M. Ghenescu, L. Ion, I. Enculescu, C. Tazlaoanu, V.A. Antohe, M. Sima, M. Enculescu, E. Matei, R. Neumann, O. Ghenescu, V. Covlea, S. Antohe, Electrical properties of electrodeposited CdS nanowires, Phys. E. 40 (2008) 2485–2488. [20] A.B. Wong, S. Brittman, Y. Yu, N.P. Dasgupta, P. Yang, Core-shell CdS-Cu2S nanorod array solar cells, Nano Lett. 15 (2015) 4096–4101. [21] X. Kuang, Y. Ma, C. Zhang, H. So, J. Zhang, B. Tang, A new synthesis strategy for chiral CdS nanotubes based on a homochiral MOF template, Chem. Commun. 51 (2015) 5955–5958. [22] X. Wang, J. Li, Q. Li, B. Chen, G. Song, W. Zhang, L. Shi, B. Zou, R. Liu, Yellow-light generation and engineering in zinc-doped cadmium sulfide nanobelts with lowthreshold two-photon excitation, Nanotechnology 25 (2014) 325702 (8pp). [23] J.H. Lee, D.J. Lee, Effects of CdCl2 treatment on the properties of CdS films prepared by r.F. Magnetron sputtering, Thin Solid Films 515 (2007) 6055–6059. [24] B.S. Moon, J.H. Lee, H. Jung, Comparative studies of the properties of CdS films deposited on different substrates by R.F. Sputtering, Thin Solid Films 511-512 (2006) 299–303. [25] J.I. Langford, R. Delhez, T.H. de Keijser, E.J. Mittemeijer, Profile analysis for microcrystalline properties by the Fourier and other methods, Aust. J. Phys. 41 (1988) 173–187. [26] R.M. Azzam, N.M. Bashara, Ellipsometry and Polarized Light Ed North – Holland, Amsterdam, 1977. [27] H. Fujiwara, Spectroscopic Ellipsometry – Principles and Applications Ed John Wiley & Sons, London, 2007. [28] O. Toma, L. Ion, M. Girtan, S. Antohe, Optical, morphological and electrical studies of thermally vacuum evaporated CdTe thin films for photovoltaic applications, Sol. Energy 108 (2014) 51–60. [29] H. Yoshikawa, S. Adachi, Optical constants of ZnO, Jpn. J. Appl. Phys. 36 (1997) 62376234. [30] J. Tauc, Amorphous and Liquid Semiconductors, Plenum Press, New York, 1974. [31] J.P. Enriquez, X. Mathew, Influence of the thickness on structural, optical and electrical properties of chemical bath deposited CdS thin films, Sol. Energ. Mat. Sol. C 76 (2003) 313–322.