Materials Science in Semiconductor Processing ] (]]]]) ]]]–]]]
Contents lists available at ScienceDirect
Materials Science in Semiconductor Processing journal homepage: www.elsevier.com/locate/mssp
Effect of processing parameter on structural, optical and electrical properties of photovoltaic chalcogenide nanostructured RF magnetron sputtered thin absorbing films P.K. Mishra a,n, V. Dave b, R. Chandra b, J.N. Prasad a, A.K. Choudhary b a b
University Department of Physics, Ranchi University, Ranchi 834008, India Institute Instrumentation Centre, Indian Institute of Technology, Roorkee 247667, India
a r t i c l e i n f o
Keywords: CIGS thin film Sputtering parameter XRD EDAX Band gap
abstract The aim of this work was to develop high quality of CuIn1 xGaxSe2 thin absorbing films with x (Ga/In þGa) o0.3 by sputtering without selenization process. CuIn0.8Ga0.2Se2 (CIGS) thin absorbing films were deposited on soda lime glass substrate by RF magnetron sputtering using single quaternary chalcogenide (CIGS) target. The effect of substrate temperature, sputtering power & working pressure on structural, morphological, optical and electrical properties of deposited films were studied. CIGS thin films were characterised by X-ray diffraction (XRD), Field emission scanning electron microscope (FESEM), Energy dispersive X-ray spectroscopy (EDAX), Atomic force microscopy (AFM), UV– vis–NIR spectroscopy and four probe methods. It was observed that microstructure, surface morphology, elemental composition, transmittance as well as conductivity of thin films were strongly dependent on deposition parameters. The optimum parameters for CIGS thin films were obtained at a power 100 W, pressure 5 mT and substrate temperature 500 1C. XRD revealed that thin film deposited at above said parameters was polycrystalline in nature with larger crystallite size (32 nm) and low dislocation density (0.97 1015 lines m 2). The deposited film also showed preferred orientation along (112) plane. The morphology of the film depicted by FE-SEM was compact and uniform without any micro cracks and pits. The deposited film exhibited good stoichiometry (Ga/Inþ Ga ¼ 0.19 and In/In þ Ga ¼ 0.8) with desired Cu/In þ Ga ratio (0.92), which is essential for high efficiency solar cells. Transmittance of deposited film was found to be very low (1.09%). The absorption coefficient of film was 105 cm 1 for high energy photon. The band gap of CIGS thin film evaluated from transmission data was found to be 1.13 eV which is optimum for solar cell application. The electrical conductivity (7.87 Ω 1 cm 1) of deposited CIGS thin film at optimum parameters was also high enough for practical purpose. & 2014 Elsevier Ltd. All rights reserved.
1. Introduction CuIn1 xGaxSe2 (CIGS) has emerged as one of the most important polycrystalline semiconducting material in solar
n
Corresponding author. Tel.: þ91 8171070991, þ918102389917. E-mail addresses:
[email protected],
[email protected] (P.K. Mishra).
cell application because of their stability against photo degradation. It belongs to the group I–III–VI2 semiconducting material and possesses chalcopyrite structure. It exhibits high optical absorption coefficient ( 105 cm 1) for high energy photons (1.5 eV). Also, its high conversion efficiency and excellent anti irradiation performance make it possible to reduce the thickness of absorbing layer in few micrometre [1]. Its tuneable band gap energy (1.0–1.7 eV) allows us tailoring the optical band gap for
1369-8001/$ - see front matter & 2014 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.mssp.2014.01.021
Please cite this article as: P.K. Mishra, et al., Materials Science in Semiconductor Processing (2014), http://dx.doi.org/ 10.1016/j.mssp.2014.01.021i
P.K. Mishra et al. / Materials Science in Semiconductor Processing ] (]]]]) ]]]–]]]
2
this, there is no systematic research that deals with optimisation of processing parameter of RF magnetron sputtering to achieve high quality of CuIn0.8Ga0.2Se2 thin films. In present work, CuIn0.8Ga0.2Se2 (CIGS) thin films were fabricated on soda lime glass by RF magnetron sputtering using single quaternary target without any selenization. The desired phase as well as microstructure, good stoichiometry, low band gap and high conductivity has been optimised by varying the processing parameter including substrate temperature, sputtering power and working pressure.
high efficiency solar cell [2]. The efficiency of photovoltaic (PV) cell strongly depends on many factors including stoichiometry, microstructure, presence of impurity phases and surface roughness for a given area of deposited film. The basic requirement for high efficiency solar cell based on CIGS is high mobility, chalcopyrite structure with good stoichiometry and microstructure with columnar grains [3]. Apart from this, Cu deficient absorbing layer is the favourable condition for high performance of solar cell because majority carrier (hole) concentration increases by increasing the density of Cu vacancies [4]. Solar cells based on CIGS have shown long term stability. The PV cell using CIGS as absorbing layer have reached the record efficiency and recently it has been grown up to 19.9% and 20.3% reported by NREL [5] and Zentrum fur wasserstoff [6] respectively. Various researchers have used different techniques to deposit CIGS thin films for high efficiency solar cell. Among these techniques, co-evaporation has emerged as main deposition technique [7]. Three stage coevaporation processes for CIGS thin films deposition involve the evaporation of Cu, In, Ga and Se using Curich as well as Cu- poor growth condition. However, this method requires precise control of individual source elements during deposition. Also, this technique involves some disadvantages like costly equipment and slow deposition rate hence; it is difficult to apply this technique for large scale production. The module efficiency based on co-evaporation process has reached only 12–13%. Another method for CIGS thin films fabrication is post selenization of precursor layer [8]. The module efficiency using post selenization process has reached up to 16%. The benefit of this technique is the maintenance of uniformity of film over large area with high deposition rate. However, this process involves toxic gases (Se or H2Se). Hence, this technique is harmful concerning the environment as well as health. Various scientists have also used different deposition techniques for CIGS thin films deposition like chemical bath deposition[9], electrodeposition [10], electron beam evaporation [11], inkjet printing [12] and laser assisted deposition [13]. Several groups have also used magnetron sputtering of ternary and quaternary alloy followed by post selenization [8]. Post selenization was performed to improve the crystallinity as well as to get the high quality chalcopyrite product. This is a simplified fabrication method because of reducing the number of targets. Wei et al. reported that the best solar cell efficiency can be achieved with x (Ga/InþGa) r0.3 [14]. It is also reported that when x40.3, the efficiency is decreased because open circuit voltage is not proportionally increased with band gap. Hence CuIn1 xGaxSe2 thin film works effectively only for a limited range of x. There is a limited study on CIGS thin films with x ¼0.2. Apart from
2. Experimental CuIn0.8Ga0.2Se2 (CIGS) thin films were deposited on ultrasonically cleaned soda lime glass (SLG) substrate by RF magnetron sputtering. A 2 in. diameter, single quaternary chalcogenide CIGS target with purity 99.999% was used as sputtering target. The elemental composition (atomic %) of CIGS target was Cu¼25%, In ¼20%, Ga¼5% and Se¼50%. CIGS target and SLG substrate was placed inside the vacuum chamber. The distance between target and substrate was kept 6 cm. Prior to the film deposition; chamber was first evacuated to high vacuum of 3 10 6 Torr with the help of turbo molecular pump backed by rotary pump. Ar gas was used as sputtering gas with a flow of 20 sccm. During deposition substrate temperature, power and pressure was varied keeping deposition time 40 min. Before each thin film deposition, pre sputtering was performed for 15 minutes. The deposition parameters for CIGS thin films are shown in Table 1.
3. Characterisation Structural analysis of deposited CIGS thin films was carried out by Bruker D8 Advance X-ray diffractometer (XRD) using CuKα radiation (λ¼ 0.154016 nm) at a tube angle of 21. Surface morphology as well as cross-sectional view of CIGS thin films were analysed by field emission scanning electron microscopy (FE-SEM: FEI Quanta 200 F). Elemental composition of CIGS thin films was investigated by energy dispersive X-ray analysis (EDAX) attached with FE-SEM. Surface roughness of deposited films was evaluated using Atomic force microscopy (AFM: NTEGRA) operated in semi contact mode. Optical transmittance and absorbance of CIGS thin films were measured by UV–vis– NIR spectrophotometer (Varian Cary 5000) in 400– 1500 nm wavelength range. Electrical conductivity of deposited CIGS thin films was measured by four probe method equipped with Kithley electrical measurement unit (2182 A nanovoltmeter, 6221 current source).
Table 1 Deposition parameters for CIGS thin films. Target
Substrate Base pressure (Torr)
CuIn0.8Ga0.2Se2 SLG
3.0 10 6
Working pressure range (mTorr)
Power range (Watt)
Substrate temperature range (1C)
Ar gas flow (sccm)
Deposition time (min)
3.5–9
40–100
RT-500
20
40
Please cite this article as: P.K. Mishra, et al., Materials Science in Semiconductor Processing (2014), http://dx.doi.org/ 10.1016/j.mssp.2014.01.021i
P.K. Mishra et al. / Materials Science in Semiconductor Processing ] (]]]]) ]]]–]]]
4. Result and discussion CIGS thin films were deposited at different substrate temperature, sputtering power and working pressure. The structural, surface morphological, elemental, optical and electrical properties of deposited CIGS thin films were comparatively studied at different processing parameter.
4.1. Structural analysis 4.1.1. Effect of substrate temperature Substrate temperature (Tsub) is the important processing parameter for growth of CIGS thin films. It helps in enhancement of crystallinity of deposited CIGS thin films. Various groups have reported that device performance can be improved by increasing the Tsub [15,16]. The surface morphology and elemental composition of CIGS thin films can be altered by changing the substrate temperature [8,17]. In the present study, CIGS thin films were deposited on SLG substrate keeping sputtering power and working pressure fixed at 40 W and 5 mTorr respectively. The substrate temperature during deposition was varied from room temperature to 500 1C. Fig. 1a shows the XRD spectra of CIGS thin films deposited at various substrate temperatures. It was observed that the film deposited at 100 1C was amorphous in nature while at substrate temperature higher than 100 1C, film shows polycrystalline behaviour.
3
It is due to the fact that at substrate temperature higher than 100 1C, Cu, In, Ga and Se atoms reacted with each other which causes the appearance of CIGS phases. All deposited CIGS thin films at higher substrate temperature (4100 1C) exhibited preferred orientation along (112) plane at about 26.71 [11]. Other phase reflection along (220/204) and (312) direction were also observed. By increasing the substrate temperature, narrowing of (112) plane was observed. The peak intensity of other phase planes (220/204) and (312) also increases with the enhancement of the substrate temperature. Thus, higher substrate temperature is beneficial for the growth of polycrystalline film without forming any complex phase. Average crystallite size (D) of deposited CIGS thin films were calculated by the Scherrer's formula [18] D¼
0:9λ β cos θ
ð1Þ
where λ¼wavelength of CuKα radiation (0.154056 nm), β¼full width at half maxima in radian, θ¼diffraction angle in degree. The value of average crystallite size was listed in Table 2. It was observed that with increase in temperature, crystallite size of CIGS thin films increases. The increment in grain size with rise in substrate temperature is due to the fact that as substrate temperature rises, atoms deposited on the glass substrate gets much thermal energy from substrate to diffuse and react with each other, aggregating to form good CIGS thin films with large crystallites.
Fig. 1. XRD pattern of CIGS thin films on SLG substrate at (a) different substrate temperature (b) sputtering power and (c) working pressure.
Please cite this article as: P.K. Mishra, et al., Materials Science in Semiconductor Processing (2014), http://dx.doi.org/ 10.1016/j.mssp.2014.01.021i
P.K. Mishra et al. / Materials Science in Semiconductor Processing ] (]]]]) ]]]–]]]
4
The value of dislocation density (L) and strain (ε) in the deposited films were calculated using the formula [18] L¼
1
ð2Þ
D2
and ε¼
β cos θ 4
ð3Þ
The calculated values of L and ε are shown in Table 2. It was observed that with increase in temperature lattice imperfection present in deposited films was decreased. The probable reason may be that as the temperature rises from 200 1C to 500 1C, there is an enhancement of crystallite size, consequently there is a simultaneous movement of atoms from grain boundary to crystallites which causes a decrease of dislocation density as well as strain in CIGS thin films [18]. Lattice parameters (a and c) of developed films were determined using the miller indices. The value of lattice parameters of CIGS thin films is given in Table 2. It was observed that at high substrate temperature (500 1C), film was consistent with lattice parameter of CIGS thin films [19]. It is clear from the above discussion that film deposited at higher temperature (500 1C) possesses larger crystallite size, less strain and minimum dislocation density. Substrate temperature was not increased beyond 500 1C since it was near to the softening temperature of glass substrate. Hence, other deposition parameters were varied at fixed substrate temperature for further investigation of desired properties of thin absorbing film. 4.1.2. Effect of sputtering power XRD spectra of CIGS thin films deposited at different sputtering power(40 W–100 W) and at fixed substrate
temperature (500 1C) and working pressure (5 mTorr) is shown in Fig. 1(b). All deposited films were polycrystalline in nature with chalcopyrite structure indexed by preferred orientation along (112) plane. Other phases corresponding to (220/204) and (312) planes were also observed. XRD spectra of all deposited films under different sputtering power do not show any complex peaks reveal that deposited films form phase pure chalcopyrite structure [11,19]. The diffraction peak intensity corresponding to (112) orientation was found to increase with increasing sputtering power, which reveals that crystallinity of deposited CIGS thin films increases with sputtering power. The growth of (220/204) plane was also enhanced with RF power. Since the phase plane corresponding to (112) orientation always favours lowest surface energy hence, grain growth along (112) plane remains dominant for all sputtering power [20]. The crystallite size, dislocation density and strain of deposited CIGS thin films under different sputtering power were calculated using the Eqs. (1), (2) and (3) and their values are shown in Table 3. The value of lattice parameters was also calculated for the film deposited at different sputtering power. It was observed that film deposited at high power as well as at high substrate temperature possesses larger crystallite size (32 nm) than the film grown at low power (22 nm). The probable reason may be that as the sputtering power is increased from 40 W to 100 W, the kinetic energy of the sputtered particles increases which causes crystallite density to increase and hence, crystallite size increases [21]. Dislocation density as well as strain in the developed films was decreased with sputtering power due to motion of interstitial atoms from its grain boundary to the crystallites. The above mentioned analysis reveals that high substrate temperature and high sputtering power are amicable for desired absorbing layer properties. The RF
Table 2 Variation of crystallite size, dislocation density, strain, lattice parameter and thickness of CIGS thin films with substrate temperature. Substrate temperature (1C)
200 300 400 500
Crystallite size (nm)
10 15 18 22
Dislocation density ( 1015 lines m 2)
10.0 4.44 3.08 2.06
Strain ( 10 3)
3.47 2.31 1.93 1.58
Lattice parameters (Å)
Thickness measurement by FE-SEM cross section
a
c
(nm)
5.75 5.76 5.76 5.76
11.90 195 11.65 200 11.65 210 11.62 240
Table 3 Variation of crystallite size, dislocation density, strain, lattice parameter and thickness of CIGS thin films with sputtering power. Sputtering power (W)
40 60 80 100
Crystallite size (nm)
22 24 26 32
Dislocation density ( 1015 lines m 2)
2.06 1.73 1.47 0.97
Strain ( 10 3)
1.58 1.45 1.34 1.08
Lattice parameter (Å)
Thickness measurement by FE-SEM cross section
a
c
(nm)
5.76 5.75 5.75 5.74
11.62 11.62 11.61 11.60
240 407 544 868
Please cite this article as: P.K. Mishra, et al., Materials Science in Semiconductor Processing (2014), http://dx.doi.org/ 10.1016/j.mssp.2014.01.021i
P.K. Mishra et al. / Materials Science in Semiconductor Processing ] (]]]]) ]]]–]]]
power was not increased beyond 100 W for more improvement of desired properties of absorbing layer because of the fact that during deposition substrate is also heated by plasma. Further investigation on CIGS thin films was carried out by varying working pressure at fixed power and substrate temperature.
4.1.3. Effect of sputtering pressure The CIGS thin films were deposited at different pressure varying from 3.5 mTorr to 9 mTorr. During sputtering, power and substrate temperature was fixed at 100 W and 500 1C respectively. The XRD pattern of CIGS thin films deposited under different working pressure is shown in Fig. 1(c). All deposited films were characterised by distinct peaks corresponding to (112) plane. Other phases corresponding to (220/204) and (312) planes were also observed at different working pressure. All the CIGS films were polycrystalline in nature with tetragonal phase [19]. By increasing the working pressure, the peak intensity corresponding to preferred orientation along (112) plane was found to decrease. Peaks corresponding to other planes (220/204) and (312) almost disappears with rise in sputtering pressure. The crystallite size, dislocation density, and strain of deposited film at various working pressure were calculated using the Eqs. (1), (2) and (3). The value of lattice parameters (a and c) of deposited thin films at different working pressures were also determined. It was observed that by increasing the pressure from 3.5 mTorr to 9 mTorr, average crystallite size was decreased from 35 nm to 19 nm (Table 4). It is due to the fact that the increase in sputtering pressure leads to increase of collision between sputtered CIGS particles and Ar atoms. As a result, mean free path for CIGS atoms become shorter and energy of sputtered particles arriving at the substrate surface falls. So less energy gains for surface diffusion, this further causes low surface mobility of sputtered particles. Since grain growth is always associated with high surface mobility of condensed particles, hence crystallinity is decreased with increase in sputtering pressure [22]. The decrease in crystallinity also leads to the increase of dislocation density and strain in the deposited film. The texture coefficient of CIGS thin films on SLG substrate is calculated from XRD peak using the following formula [8,23]. TC ¼
IðhklÞ Ið112Þ þIð220=204Þ þ Ið312Þ
ð4Þ
where I is relative intensity obtained from XRD data and (hkl) represents (112), (220) and (312) orientation of CIGS thin films. Fig. 2 shows the variation of texture coefficient of CIGS thin films with sputtering pressure. The variation of texture coefficient with sputtering pressure is due to the competition between surface energy and strain energy of crystallites. Since grain growth along (112) orientation exhibits lower surface energy than the (220/204) orientation [20]. Hence, film exhibits strong texture for (112) orientation for all working pressure. It was observed that film deposited at higher working pressure (9 mTorr) possesses slightly higher texture coefficient of value 0.81 for (112) orientation than the film deposited at low pressure (3.5 mTorr) which has a texture coefficient of 0.79. It is clear from above experimental data that film synthesised at 500 1C, 100 W and 3.5 mTorr exhibited larger crystallite size with high peak intensity along (112) plane. However, film deposited at higher working pressure possess slightly higher texture coefficient but with reduced crystallite size. Hence, higher substrate temperature 500 1C, higher sputtering power 100 W and low working pressure 3.5 mTorr is beneficial for polycrystalline feature of CIGS thin film. 4.2. Surface morphology 4.2.1. Effect of substrate temperature Fig. 3a shows the surface morphology of CIGS thin films deposited at different substrate temperature while inset
Fig. 2. Variation of texture coefficient of CIGS thin films with working pressure.
Table 4 Variation of crystallite size, dislocation density, strain, lattice parameter and thickness of CIGS thin films with working pressure. Sputtering pressure (mTorr)
3.5 5 7 9
Crystallite size (nm)
35 32 24 19
Dislocation density ( 1015 lines m 2)
0.81 0.97 1.73 2.77
5
Strain ( 10 3)
0.99 1.08 1.38 1.83
Lattice parameter (Å)
Thickness measurement by FE-SEM cross section
a
c
(nm)
5.74 5.74 5.77 5.77
11.61 11.60 11.77 11.76
893 868 526 450
Please cite this article as: P.K. Mishra, et al., Materials Science in Semiconductor Processing (2014), http://dx.doi.org/ 10.1016/j.mssp.2014.01.021i
6
P.K. Mishra et al. / Materials Science in Semiconductor Processing ] (]]]]) ]]]–]]]
Fig. 3. FE-SEM cross-sectional view and surface morphology of CIGS thin films deposited at different (a) substrate temperature (b) sputtering power and (c) working pressure.
views are the corresponding FE-SEM cross-sectional observation of CIGS thin films. From these micrographs, it was observed that substrate temperature strongly influence the microstructure of CIGS thin films. Particle size of thin films was increasing with rise in temperature as indicated by FE-SEM image which confirms the correctness of XRD data. Particle size of the films deposited at Tsub ¼ 500 1C was seems to be larger in comparison to other films particles size. All deposited films surface were free from pits and micro cracks. It was also observed that films compactness enhances with rise in temperature which is essential for good absorbing layer [24]. The thickness of films was measured by FE-SEM cross-section and data are listed in Table 2. 4.2.2. Effect of sputtering power Surface morphology of CIGS thin films deposited at different sputtering power and at fixed substrate temperature
and working pressure is shown in Fig. 3b while inset figures are corresponding cross-sectional view of CIGS films. It is clear from the figure that higher thickness (868 nm) of film was achieved at 100 W. It is because of increment in deposition rate of CIGS thin films with rise in sputtering power. It was also observed that film deposited at 100 W possess compact and had uniform morphology. The deposited film was also free from pin holes and microcracks. The thickness of deposited CIGS thin films at different sputtering power are listed in Table 3. 4.2.3. Effect of sputtering pressure Fig. 3c highlights the surface morphology of CIGS thin films at different working pressure. The cross-sectional images of corresponding thin films are also shown in the figure. It was observed from cross-sectional SEM image that with increase in pressure, thickness of films decreases.
Please cite this article as: P.K. Mishra, et al., Materials Science in Semiconductor Processing (2014), http://dx.doi.org/ 10.1016/j.mssp.2014.01.021i
P.K. Mishra et al. / Materials Science in Semiconductor Processing ] (]]]]) ]]]–]]]
The probable reason may be that higher working pressure was associated with large scattering of sputtered CIGS particles and Ar ions inside the chamber. As a result of this, less number of particles arrives at the substrate surface. Film deposited at 3.5 mTorr also possesses uneven surface which may be due to the presence of microrods in the film. These microrods are not beneficial for solar cell fabrication because these rods causes short circuit in the cell [24]. Hence, it must be avoided. Film deposited at higher pressure was free from microrods. It was also observed that the surface of the films deposited at pressureZ5 mTorr was even, compact, uniform and free from microcracks, pits and holes. Thus, it was observed that film deposited at power 100 W, substrate temperature 500 1C and pressure 5 mTorr is free from holes, microcracks and pits. Film also possesses compact and uniform morphology with desired crystallinity. However, film deposited at very low working pressure (3 mTorr) covered an uneven surface with some microrods which is not supposed to be positive for cell performance. 4.3. Elemental analysis 4.3.1. Effect of substrate temperature It was observed that by varying substrate temperature, atomic% ratio of In/InþGa and Ga/(InþGa) was also changing and it was found to be 0.79 and 0.21 respectively at 500 1C (as shown in Fig. 4a) which was near to the stoichiometry of target material (Ga/InþGa ¼0.2 and In/InþGa¼ 0.8). But films contain large number of Cu atoms at higher temperature (Fig. 4a). The atomic % ratio of Cu/InþGa possessed by CIGS thin films at higher temperature (400 1C and 500 1C) was found to be41 which is not suppose to be positive for cell performance. So for desired content of Cu atoms, other parameters should vary so that Cu/InþGa ratio lie in specific range (0.75–0.98) for high efficiency solar cell. 4.3.2. Effect of sputtering power Variation of atomic % of Cu, In, Ga and Se of CIGS thin films with sputtering power is shown in Fig. 4b. It was observed that by increasing the power from 40 W to 100 W at fixed substrate temperature 500 1C and pressure 5 mTorr, the stoichiometry of deposited films approximately moves towards the composition of bulk i.e. target material (CuIn0.8Ga0.2Se2). The atomic % ratio of Ga/InþGa and In/InþGa possessed by CIGS thin film at 100 W was 0.19 and 0.80 respectively (Fig. 4b) which was near to the target stoichiometric% ratio. Also, film was found to be slightly Cu deficient (Cu/InþGa ¼0.92) (as shown in Fig. 4b) which is essential for good photovoltaic effect and high efficiency solar cell [25]. 4.3.3. Effect of sputtering pressure Fig. 4(c) shows the variation of atomic% of constituent elements with working pressure. It was observed that films deposited at lower working pressure (3.5 mTorr, 5 mTorr) possess stoichiometry that approximately matches with the target material while at higher pressure (7 mTorr, 9 mTorr) the film became off stoichiometric. The
7
atomic% ratio of Ga/Inþ Ga and In/InþGa was found to be 0.19 and 0.80 respectively at 5 mTorr which was near to the target atomic % ratio (Fig. 4c). In addition to this, atomic% ratio of Cu/InþGa (CIG) was found to be in the range of 0.75 rCIG r0.98 at low working pressure (as shown in Fig. 4c) hence, these films show photovoltaic effect. Film deposited at 9 mTorr exhibits 0.62 CIG and thus, do not lie in the desired CIG range. Hence, it will not show good photovoltaic effect. It was also observed that by decreasing the working pressure up to 5 mTorr, atomic % ratio of CIG increases which indicate that film deposited at low working pressure (5 mTorr) possesses higher efficiency. The above discussion on elemental analysis clearly reveals that higher substrate temperature, higher working power and lower sputtering pressure is favourable condition for stoichiometric thin film growth. 4.4. AFM analysis 4.4.1. Effect of substrate temperature Fig. 5a shows the 3D AFM image of CIGS thin films at different substrate temperature. Using AFM data analysing software the rms surface roughness of deposited CIGS thin films was studied. The rms surface roughness of a material is given by [26]. Srms ¼
1 ½N∑j ðZ i Z a Þ1=2
f or j ¼ i 1 to N
ð5Þ
where Zi ¼height of ith point and Za ¼ arithmetic average height within measurement area. It was observed that with rise in substrate temperature (300 1C to 500 1C) surface roughness was also increased (2 nm to 7 nm). It was also observed that film is sufficiently smooth to use as an absorbing layer at higher temperature [27]. The enhancement in surface roughness of CIGS thin films with increase in substrate temperature is due to the growth of grains with temperature. Also, the growth of grains with preferred orientation was dictated by grain boundary diffusivity, adatoms mobility, film thickness and induced thermal stress [23]. Hence, these factors are also responsible for increment in surface roughness. 4.4.2. Effect of sputtering power The rms value of surface roughness of CIGS thin film deposited at different working power (40 W–100 W) and at fixed substrate temperature (500 1C) and pressure (5 mTorr) was also analysed. Three dimensional AFM images of CIGS thin films at different working power are shown in Fig. 5b). It is clear from the figure that film deposited at 100 W is rougher than the 40 W deposited film. The rms value of surface roughness was found to increase from 7 nm to 12 nm with increase in sputtering power from 40 W to 100 W. The surface roughness was found to be increasing with sputtering power but the range of the roughness (7–12 nm) was so small that the films can be considered as smooth to use as an absorbing layer. The probable reason in increment of surface roughness was the enhancement of deposition rate with working power due to which adatoms mobility increases which
Please cite this article as: P.K. Mishra, et al., Materials Science in Semiconductor Processing (2014), http://dx.doi.org/ 10.1016/j.mssp.2014.01.021i
8
P.K. Mishra et al. / Materials Science in Semiconductor Processing ] (]]]]) ]]]–]]]
Fig. 4. Atomic% and mutual atomic % ratio of different elements in CIGS thin films as a function of (a) substrate temperature (b) sputtering power and (c) working pressure.
causes the growth of crystallite size. The higher crystallite size in turn was responsible for higher surface roughness of deposited films [23]. 4.4.3. Effect of sputtering pressure The 3D AFM images of CIGS thin films deposited at different working pressure is shown in Fig. 5 (c). It was observed that by increasing the working pressure from 3.5 mTorr to 9 mTorr keeping substrate temperature fixed at 500 1C and power at 100 W, surface roughness was decreased from 26 nm to 6 nm. It is because of the fact that by increasing the working pressure, crystallite size
decreases and thus, surface roughness of thin films also decreases. The rms value of surface roughness for 5 mTorr deposited film was found to be 12 nm. The film possessing this much value of roughness can be treated as sufficiently smooth film and hence, can be widely used as an absorbing layer. It is clear from the above discussion that the surface roughness can be altered by varying the processing parameter. The film deposited at higher substrate temperature (500 1C), higher sputtering power (100 W) and lower working pressure (5 mTorr) is somewhat rougher ( 12 nm) but can be treated as smooth absorbing layer.
Please cite this article as: P.K. Mishra, et al., Materials Science in Semiconductor Processing (2014), http://dx.doi.org/ 10.1016/j.mssp.2014.01.021i
P.K. Mishra et al. / Materials Science in Semiconductor Processing ] (]]]]) ]]]–]]]
Fig. 5. 3D AFM image of CIGS thin films at different (a) substrate temperature (b) sputtering power and (c) working pressure.
Please cite this article as: P.K. Mishra, et al., Materials Science in Semiconductor Processing (2014), http://dx.doi.org/ 10.1016/j.mssp.2014.01.021i
9
10
P.K. Mishra et al. / Materials Science in Semiconductor Processing ] (]]]]) ]]]–]]]
4.5. Optical analysis 4.5.1. Effect of substrate temperature The transmittance and absorbance spectra of CIGS thin films deposited at different substrate temperature are shown in Fig. 6a). It is clear from the transmittance spectra that CIGS thin films exhibit 22% transmittance at 500 1C
while 28% at 300 1C. The lower value of transmittance at 500 1C is because of higher value of films thickness (240 nm) possessed by CIGS thin films. CIGS thin film also possesses high surface roughness of value 7 nm at 500 1C. As a result, scattering loss at rougher surface increases which in turn reduces transmittance in CIGS thin film [18]. The transmittance of CIGS thin films do not
Fig. 6. Optical transmittance and absorbance spectra of CIGS thin films deposited at different (a) substrate temperature (b) sputtering power and (c) working pressure.
Please cite this article as: P.K. Mishra, et al., Materials Science in Semiconductor Processing (2014), http://dx.doi.org/ 10.1016/j.mssp.2014.01.021i
P.K. Mishra et al. / Materials Science in Semiconductor Processing ] (]]]]) ]]]–]]]
possess interference pattern in long wavelength region (41100 nm) indicating that film deposited at low power and high substrate temperature do not exhibit thickness uniformity [8].
4.5.2. Effect of sputtering power Fig. 6b shows the optical transmittance and absorbance spectra of CIGS thin films deposited at different sputtering power and at fixed substrate temperature and working pressure. It is clear from the transmittance spectra that with increase in working power, transmittance of deposited films decreases. The transmittance of deposited film at 100 W was found to be 1.09% at 800 nm wavelength which is very small in comparison to the transmittance of film deposited at 40 W ( 22%). It is due to the fact that by increasing the sputtering power, thickness of film on glass substrate increases this further causes the opaqueness in the film. Transmittance spectra of CIGS thin films deposited at 80 W and 100 W possesses interference pattern in infrared region which indicates that films exhibit well thickness on glass substrate [8]. It was also observed from absorbance spectra that absorbance of CIGS thin films increases with power. Fig. 6b also shows that films possess high absorbance in visible region while near infrared region it decreases sharply. This result is consistent with other result obtained by Han et al. [20].
4.5.3. Effect of sputtering pressure The variation of transmittance and absorbance spectra of CIGS thin films with working pressure is highlighted in Fig. 6c. It was found that by increasing the pressure from 3.5 mTorr to 9 mTorr at fixed temperature 500 1C and power 100 W, transmittance value rises from 0.2% to 24% and absorbance value decreases near NIR range (800 nm). It may be due to thickness of the film. The thickness of the films decreases from 893 nm to 450 nm (as evidenced by FE-SEM cross-section) as pressure was increased from 3.5 mTorr to 9 mTorr. As a result, opaqueness of the films decreases. Also, increment in surface roughness (as evidenced by 3D AFM image) with decrease in pressure further causes more scattering loss and thus reduces transmittance. Transmittance spectra at lower working pressure also exhibited good interference pattern in infrared region. The appearance of interference pattern revealed that films deposited at low pressure exhibits uniform thickness at substrate. Absorption coefficient of films deposited at different working pressure was calculated using the relation [28] T¼
½ð1 RÞ2 e αt ½1 R2 e 2αt
11
Further Eq. (7) is reduced to simpler form as [9] α¼
2:303 logð1=TÞ t
ð8Þ
The value of α was calculated using the relation (8) and it was found that the value of α was 105 cm 1 for high energy photons as shown in Fig. 7. The band gap of CIGS thin films was calculated using the Tauc relation [8] αhv ¼ Aðhv Eg Þn
ð9Þ
where h¼Planck's constant, ν ¼frequency of radiation, Eg ¼band gap, n ¼1/2 to 2 Since CIGS is a direct band gap semiconducting material so value of n was chosen 1/2 [8]. Fig. 8 shows the plot of (αhν)2 vs. hν. By extrapolating the straight line part of the (αhν)2 curves band gap was evaluated. The band gap is showing increasing trend with increase in working pressure and it increases from 1.11 eV to 1.36 eV as the pressure varies from 3.5 mTorr to 9 mTorr. It may be probably due to change in grain size (as evidenced by XRD).It is well known fact that larger grain size have lesser number of grain boundaries which in turn reduces the
Fig. 7. Variation of absorption coefficient of CIGS thin films with working pressure.
ð6Þ
where R and T are reflectance and transmittance of film, α is absorption co efficient of film and t is the thickness of deposited film. For higher absorbance (αt41), the interference effects due to internal reflection and reflectance at normal incidence are negligible. In this condition the Eq. (6) becomes T ¼ e αt
ð7Þ
Fig. 8. Variation of band gap of CIGS thin films with working pressure.
Please cite this article as: P.K. Mishra, et al., Materials Science in Semiconductor Processing (2014), http://dx.doi.org/ 10.1016/j.mssp.2014.01.021i
P.K. Mishra et al. / Materials Science in Semiconductor Processing ] (]]]]) ]]]–]]]
12
Fig. 9. Variation of film thickness and conductivity of CIGS thin films with (a) substrate temperature (b) sputtering power and (c) working pressure.
scattering phenomenon and consequently, reduction in band gap. It has been also reported that the band gap of CIGS thin films varies with x [29] (x ¼Ga/InþGa) according to the relation Eg ¼ 1:02 þ0:67 x þ b xðx 1Þ
ð10Þ
where b is called optical bowing coefficient and its value is 0.21 eV [30]. The value of b is independent of x [31]. Using the value of ‘b’ in Eq. (10) Eg is found to be 1.12 eV for x¼0.2, which is found to be in good agreement with band gap value of films deposited at 5 m Torr (1.13 eV). 4.6. Electrical properties 4.6.1. Effect of substrate temperature Fig. 9a shows the variation of conductivity of CIGS thin films with substrate temperature. The conductivity of deposited films was calculated using the formula [25]. s¼
1 Rs d
ð11Þ
where Rs ¼sheet resistance of deposited film at different substrate temperature, d ¼thickness of film [25]. It was observed that as the substrate temperature increases conductivity of CIGS thin films also increases linearly. It is due to the fact that with the increase in temperature,
grain size (XRD data) of CIGS thin films increases. As a result, potential barrier and height for the charge carriers were reduced to a very low value. So charge carriers have to cross minimum number of grain boundaries during the electrical transport which in turn enhances the conductivity of CIGS thin films. Such result was also observed by Shah et al. where highly conducting thin film was obtained at high substrate temperature [32]. It was also observed that films deposited at higher substrate temperature (400–500 1C) exhibit very high conductivity. This can be explained by EDAX results. EDAX data (Fig. 4a shows that thin films deposited at higher temperature contain excess of Cu contents which causes the existence of CuxSe which exhibits a very high conductivity and metallic electrical behaviour [25]. 4.6.2. Effect of sputtering power The variation of conductivity of CIGS thin films with sputtering power is shown in Fig. 9b. It was observed that conductivity of deposited films decreases exponentially with sputtering power. It was also found that film deposited at 100 W exhibits low electrical conductivity (7.87 Ω 1 cm 1) than the film deposited at 40 W which possesses very high conductivity (200 Ω 1 cm 1). The probable reason may be the non stoichiometric behaviour of CIGS thin films at low power. The EDAX data
Please cite this article as: P.K. Mishra, et al., Materials Science in Semiconductor Processing (2014), http://dx.doi.org/ 10.1016/j.mssp.2014.01.021i
P.K. Mishra et al. / Materials Science in Semiconductor Processing ] (]]]]) ]]]–]]]
reveals that atomic% ratio of Cu/InþGa for 100 W deposited sample is 0.92 while for 40 W deposited sample it is about 1.28 which signifies that film deposited at low power contains excess of Cu (at %) contents and hence, there is existence of CuxSe, which is responsible for high conductivity of deposited film [25]. 4.6.3. Effect of sputtering pressure Variation of conductivity as a function of working pressure is shown in Fig. 9c. It was observed that conductivity of CIGS thin films increases with decreasing pressure. Such result was also observed by Scofield et al. where high conducting material was obtained at low working pressure [33]. The increment in conductivity is because of the fact that thickness of films increases as the pressure decreases. Consequently, the grain size increases which cause the reduction of intercrystalline barrier. Hence, charge carriers have to cross minimum number of potential barrier height during the electrical charge flow. It was observed that low power and high substrate temperature is not beneficial for electrical properties of CIGS thin film due to non stoichiometric behaviour (high atomic% of Cu) of deposited film. The film deposited at high sputtering power (100 W), low working pressure (5 mTorr) and high substrate temperature exhibited good electrical conductivity which signifies the good stoichiometric behaviour and higher crystallinity of deposited film. The film deposited at high working pressure possesses low conductivity due to reduction in crystallinity as well as stoichiometry of thin film. 5. Conclusion CIGS thin absorbing films for solar cell were deposited by RF magnetron sputtering using single quaternary alloy target. The effect of processing parameters on phase structure, morphology, optical and electrical properties of CIGS thin films was investigated. All the deposited CIGS thin films were indexed with chalcopyrite phase structure without any precipitation of secondary phase. The film microstructure and phase structure were found out to be better at higher substrate temperature, higher working power and low working pressure. The CIGS thin absorbing layer sputtered at 500 1C, 5 mTorr and 100 W showed well crystallised microstructure with compact and uniform morphology strongly (112) plane oriented, which is supposed to be positive for solar cell performance. The absorption coefficient of deposited film was found to be of the order of 105 cm 1 in the region of high photon energy. Also, film at the optimum parameters exhibits low optical band gap (1.13 eV). The overall chemical composition of thin absorbing film was close to that of target material. The Cu/InþGa atomic% ratio of deposited CIGS thin film was found to be 0.92 which is desirable for high efficiency solar cell. The surface roughness of deposited film at optimised parameter was found to be 12 nm which showed that film is sufficiently smooth to use as an absorbing layer. Film was also found to be good conducting at optimum parameters. Apart from this, the present work represented the optimised parameter of high quality CIGS absorbing thin film fabricated using sputtering technique
13
excluding selenization process. In future, these optimised parameters can be used to fabricate absorbing layer below the window layer.
Acknowledgement One of the authors P.K. Mishra is thankful to DST New Delhi, for financial support towards this research work. References [1] Y.H. Jo, B.C. Mohanty, Y.S. Cho, Appl. Surf. Sci. 256 (2010) 6819–6823. [2] S. Schleussner, U. Zimmermann, T. Watjen, K. Leifer, M. Edoff, Sol. Energy Mater. Sol. Cells 95 (2011) 721–726. [3] S. Niki, M. Contreras, I. Repins, M. Powalla, K. Kushiya, S. Ishizuka, K. Matsubara, Prog. Photovolt. Res. Appl. 18 (2010) 453–466. [4] C. Xu, H. Zhang, J. Parry, S. Perera, G. Long, H. Zheng, Sol. Energy Mater. Sol. Cells 117 (2013) 357–362. [5] I. Repins, M.A. Contreras, B. Egaas, C.D. Hart, J. Scarf, C.L. Perkins, B. To, R. Noufi, Prog. Photovolt. Res. Appl. 16 (2008) 235–239. [6] P. Jackson, D. Hariskos, E. Lotter, S. Paeterl, R. Wuerz, R. Menner, W. Wischmann, M. Powalla, Prog. Photovolt. Res. Appl. 19 (2011) 894–897. [7] S. Jung, S.J. Ahn, J.H. Yun, J. Gwak, D. Kim, K. Yoon, Curr. Appl. Phys. 10 (2010) 990–996. [8] Z. Yu, Y. Yan, S. Li, Y. Zhang, C. Yan, L. Liu, Y. Zhang, Y. Zhao, Appl. Surf. Sci. 264 (2013) 197–201. [9] M. Chandramohan, S. Velumani, T. Venkatachalam, Mater. Sci. Eng. B 174 (2010) 205–208. [10] V.S. Saji, I.H. Choi, C.W. Lee, Sol. Energy 85 (2011) 2666–2678. [11] M. Venkatachalam, M.D. Kannan, S. Jayakumar, R. Balasundaraprabhu, A.K. Nandakumara, N. Muthukumarasamy, Sol. Energy Mater. Sol. Cells 92 (2008) 571–575. [12] W. Wang, Y.W. Su, C. Chang, Sol. Energy Mater. Sol. Cells 95 (2011) 2616–2620. [13] T. Nakada, S. Shirakata, Sol. Energy Mater. Sol. Cells 95 (2011) 1463–1470. [14] S.H. Wei, S.B. Zhang, A. Zunger, Appl. Phys. Lett. 72 (1998) 3199–3201. [15] W.N. Shafarmann, J. Zhu, Thin Solid films 361–362 (2001) 473–477. [16] J. Kessler, C. Chityuttakan, J. Scoldstrom, L. Stolt, Thin Solid films 431–432 (2003) 1–5. [17] M. Mohammidikish, F. Davar, M.R.L. Estarki, J. Cluster Sci. 24 (2013) 217–231. [18] P. Matheswaran, B. Gokul, K.M. Abhirami, R. Sathyamoorthy, Mater. Sci. Semicond. Process. 15 (2012) 486–491. [19] M. Venkatachalam, M.D. Kannan, S. Jayakumar, R. Balasundaraprabhu, N. Muthukumarasamy, Thin Solid Films 516 (2008) 6848–6852. [20] A. Han, Y. Zhang, W. Song, B. Li, W. Liu, Y. Sun, Semicond. Sci. Technol. 27 (2012) 8pp. (035022). [21] D.J. Kwak, M.W. Park, Y.M. Sung, Vacuum 83 (2008) 113–118. [22] P. Gao, L.J. Meng, M.P. dos Santos, V. Teixeira, M. Andritschky, Appl. Surf. Sci. 173 (2001) 84–90. [23] V. Chawla, R. Jayaganthan, A.K. Chawla, R. Chandra, J. Mater. Process. Technol. 209 (2009) 3444–3451. [24] A.J. Zhou, D. Mexi, X.G. Kong, X.H. Xu, L.D. Feng, X.Y. Dai, T. Gao, J. Z. Li, Thin solid Films 520 (2012) 6068–6074. [25] Z. Yu, C. Yan, T. Huang, W. Huang, Y. Yan, Y. Zhang, L. Liu, Y. Zhao, Appl. Surf. Sci. 258 (2012) 5222–5229. [26] V. Dave, P. Dubey, H.O. Gupta, R. Chandra, Thin Solid Films 549 (2013) 2–7. [27] L. Zhang, Q. He, W. Jiang, F.F. Liu, C.J. Li, Y. Sun, Sol. Energy Mater. Sol. Cells 93 (2009) 114–118. [28] Z.S.E. Mandouh, M.S. Selim, Thin Solid Films 371 (2000) 259–263. [29] O. Lundberg, M. Edoff, L. Stolt, Thin Solid Films 480–481 (2005) 520–525. [30] S.H. Wei, A. Zunger, J. Appl. Phys. 78 (1995) 3846–3856. [31] S.H. Han, F.S. Hasoon, J.W. Pankow, A.M. Hermann, D.H. Levi, Appl. Phys. Lett. 87 (2005) 151904-(1)–151904-(3). [32] N.M. Shah, J.R. Ray, V.A. Kheraj, M.S. Desai, C.J. Panchal, B. Rehani, J. Mater. Sci. 44 (2009) 316–322. [33] J.H. Scofield, A. Duda, D. Albin, B.L. Ballared, P.K. Predecki, Thin Solid Films 260 (1995) 26–31.
Please cite this article as: P.K. Mishra, et al., Materials Science in Semiconductor Processing (2014), http://dx.doi.org/ 10.1016/j.mssp.2014.01.021i