Abstract
Copper indium gallium selenide (CIGS) films were deposited for the first time by the brush electrodeposition technique. X-ray diffraction studies indicated the formation of single phase chalcopyrite CIGS. Lattice parameters, dislocation density, and strain were calculated. Band gap of the films increased from 1.12 eV to 1.63 eV as the gallium concentration increased. Room temperature transport parameters of the films, namely, resistivity increased from 0.10 ohm cm to 12 ohm cm, mobility decreased from 125 cm2V−1s−1 to 20.9 cm2V−1s−1, and carrier concentration decreased from 4.99 × 1017 cm−3 to 2.49 × 1016 cm−3 as the gallium concentration increased. Photosensitivity of the films increased linearly with intensity of illumination and with increase of applied voltage.
1. Introduction
In recent years, quaternary chalcopyrite compound (CIGS) has been one of the most promising absorber materials for high efficiency thin film solar cells [1]. Thin film solar cells based on coevaporated CIGS absorbers ( close to 0.3) have reached up to 20.3% conversion efficiencies at the laboratory scale by using a process requiring high vacuum [2]. From “lab to large-scale production,” one of the main challenges is to find an alternative deposition method using nonvacuum equipment that yields economically viable solar cells and is easily scalable. From this point of view, electrochemical deposition is a simple and nonvacuum technique and has a natural advantage of large-area deposition [3]. Therefore, over the last two decades, there has been considerable work done on the growth of CIGS thin films using electrodeposition technique [4–6]. So far all electrodeposited (ED) CIGS films need a selenization step under a Se-containing atmosphere to recrystallize the films, as in most cases electrodeposition is employed at low temperature. In the present work we report the properties of CIGS films deposited by the brush plating technique for the first time.
2. Experimental Methods
Brush plating was carried out using Selectron Power Pack MODEL 150A-40 V. Layers were brush plated on tin oxide coated conducting substrates of about 50 cm2, which is the negative electrode. The stylus, consisting of a carbon rod wrapped in cotton wool, served as the anode. The cotton wool was held in position by a porous sleeve. A sketch of the system is shown in Figure 1. Prior to plating, the stylus was wired to the power supply and the cotton wool was soaked in the electrolyte. The stylus was then brought into contact with the substrate and moved at uniform speed. An electrical current was found passing whenever the stylus was in contact with the substrate. This is associated with the acceleration of the ions in the electrolyte trapped within the cotton wool, which were subsequently reduced at the substrate to form the CIGS film. CIGS films were deposited on tin oxide coated conducting glass substrates at a temperature of 80°C. The precursors used were 20 mM SeO2, 30 mM CuCl2, and the concentration of the indium chloride and gallium chloride precursors varied, as shown in Table 1, to obtain films of different composition. The deposition current density was 1.0 mA cm−2. The pH was maintained at 1.5 by HCl. Thickness of the films estimated by Mitutoyo surface profilometer varied in the range of 0.5–1.1 μm with an increase in substrate temperature. The films were characterized by Xpertpanalytical X-ray diffraction unit with Cuk radiation. Optical measurements were recorded using a Hitachi UV-Vis-IR spectrophotometer. Composition of the films was estimated by EDS attachment to JOEL SEM. X-ray photoelectron spectroscopic (XPS) studies were made using VGMKII system with Mgk radiation. Surface morphology of the films was studied by molecular imaging atomic force microscope. Electrical measurements were made by Hall Van der Pauw Method. The dark and photoconductivity measurements were carried out at room temperature. The measuring system consists of a regulated dc power supply (Aplab) in series with the sample and Keithley electrometer (model 610 C). A tungsten lamp of 200 W was used for illumination. The intensity of the light source was measured with the help of a power meter. The spectral distribution of photocurrent was measured with the help of photophysics monochromator.
| (a) = 0.1 |
| (b) = 0.4 |
| (c) = 0.9 |
3. Results and Discussion
Microstructural parameters were estimated by studying the X-ray diffractograms of films of different composition. Figure 2 shows the X-ray diffraction patterns of films of different composition. All the figures indicate the prominent peaks corresponding to (112), (220)/(204), (312)/(116). These are characteristic of the chalcopyrite phase. No other phases were observed in the X-ray diffractograms, indicating the formation of single phase material. The peaks shifted from CuInSe2 side to CuGaSe2 side as the concentration of Ga increased in the films. The lattice parameters were calculated using the following relation [7]: where “” and “” are the lattice parameters and “” is the lattice spacing. Figure 3 shows the variation of “” and “” with increase of gallium concentration. This behaviour is similar to an earlier report [8].
(a)
(b)
(c)
(d)
The grain size of the films has been calculated using Scherrer’s formula: where is the wavelength of X-ray used, the full width half-maximum (FWHM), and the Bragg angle. The grain size varied from 30 nm to 70 nm as the gallium concentration increased. The dislocation density , defined as the length of dislocation lines per unit volume of the crystal, has been evaluated using the formula [9]
The microstructural parameters are presented in Table 2. From the table it is observed that the dislocation density decreases with the increase of grain size. Information on the particle size and strain for the films was obtained from the full width at half-maximum of the diffraction peaks. The full width at half-maximum can be expressed as a linear combination of the contributions from the particle size and strain through the relation [10] The plot of versus allows us to determine both strain and particles size from slope and intercept of the graph. The estimated values for films deposited at different duty cycles are listed in Table 2. The deviation in the lattice parameter values from the bulk value observed in the present case clearly suggests that the grains in the films are under stress. Such a behaviour can be attributed to the change of nature, deposition conditions, and the concentration of the native imperfections developed in thin films. This results in either elongation or compression of the lattice and the structural parameters. The density of the film is therefore found to change considerably in accordance with the variations observed with the lattice constant values. The stress developed at higher Ga concentrations is likely to be due to the formation of native defects developed from the lattice misfit or dislocations. The defects have a probability to migrate parallel to the substrate surface so that the films will have a tendency to expand and develop an internal tensile stress. This type of change in internal stress is always predominant by the observed recrystallization process in polycrystalline films. The stress relaxation is mainly considered as due to dislocation glides formed in the films. The decrease of internal stress may be attributed to a decrease in dislocation density. The reduction in the strain and dislocation density with decrease of Ga concentration may be due to the reduction in concentration of lattice imperfections at lower Ga concentrations. Similar behaviour was reported earlier [11].
Composition of the films was estimated by recording the EDS spectrum of the films deposited from different composition (Table 3). It is observed that films with lower concentrations of gallium were copper rich. As the gallium concentration increased, the films became nearly stoichiometric. This is due to the fact that as the concentration of gallium chloride increases, more flux of gallium ions are available for deposition compared to the flux of indium ions, which results in higher concentration of gallium thus decreasing the Cu/(Ga + In) ratio. Based on the defect chemistry model of ternary compounds [12], compositional deviations of the films can be expressed by nonstoichiometry parameter . The parameter is related to the electronic defects. For , the film has a -type conductivity and it has an -type conductivity for < 0. In this study the value of is greater than zero, and the films exhibit -type conductivity.
Surface morphology of the films (Figure 3) studied in an area of 3 μm × 3 μm indicated that the grain size increased from 40 nm to 100 nm as the In concentration increased. The surface roughness also increased from 0.25 nm to 3.2 nm with increase of indium concentration. The surface roughness increases due to the increase of grain size.
XPS analysis was performed to identify the chemical binding states of the constituents of the films deposited at 50% duty cycle. Figure 4 represents the XPS spectra of films of different composition. Cu core level spectrum. The observed peak located at 931.9 eV coincides with the binding energy for Cu electrons emitted from CuGaSe2 compound, and the peak at 951.2 eV corresponds to the binding energy for Cu electrons emitted from Cu element. The Ga core level spectrum was observed at 20 eV. One peak for the Se core level spectrum was observed. The peak at 54 eV corresponds to the electronic state of Se . The peaks shifted as follows: Cu from 932.38 eV to 932.216 eV, In from 441.15 eV to 440.75 eV, In 3 from 449.15 eV to 448.75 eV, Se from 53.7 eV to 53.8 eV, and Ga from 19.60 eV to 19.20 eV, as In concentration decreases from 0.9 to 0.1.
(a)
(b)
(c)
(d)
Figure 5 shows the transmission spectra of the films deposited at 80°C. The spectra exhibit interference fringes, and the value of the refractive index was estimated by the envelope method [13] as follows: where is the refractive index of the substrate and and are the maximum and minimum transmittances at the same wavelength in the fitted envelope curve on a transmittance spectrum. The value of the refractive index calculated from the above equations was 2.60. The refractive index decreases with wavelength (Figure 6). The value of the absorption coefficient () was calculated using the relation where “” is the thickness of the film and the other parameters have the usual meaning as given for (6). The band gap of the films increased from 1.12 eV to 1.63 eV as the gallium concentration increased (from versus plot) (Figure 7). The values of the band gap agree well with the earlier report [14].
The room temperature transport parameters were measured by Hall Van der Pauw technique by providing gold ohmic contact. The influence of composition on the resistivity of the films is shown in Table 4. The magnitude of the room temperature resistivity increased from 0.1 ohm cm to 12.00 ohm cm as the gallium concentration increased. The resistivity values are comparable with an earlier report [15]. The variation in resistivity with gallium concentration can be explained in terms of the Cu/(Ga + In) ratio obtained from EDAX measurements. The Cu/(Ga + In) ratio is greater than unity for all compositions. The films exhibit -type conductivity, as the concentration of gallium increases, the resistivity increases. The variation of room temperature mobility and carrier density with increase of gallium concentration is also shown in Table 4. The increase of resistivity can also be explained in terms of the decrease of carrier density with increase of gallium concentration.
Various crystalline imperfections in the film, such as vacancies, dislocations, and grain boundaries, act as trapping or recombination centers of the carriers and play an important role in photoconduction. These traps act as localized positive potential centers for electrons and negative potential centers for holes. Therefore, some localized discrete energy levels are formed in the band gap, in the vicinity of the conduction and valence bands, respectively. Figure 8 shows the variation of photocurrent with light intensity of films of different composition. The photocurrent is found to increase with an increase of indium concentration due to increase in film thickness.
As the thickness of the film increases, the crystalline nature increases (Table 2), and this helps in the improvement of photocurrent. The increase in photocurrent is attributed to an increase in the majority carrier concentration and/or an increase in impurity centers acting as traps for minority carriers. The variation of photocurrent with applied voltage in films of different composition is shown in Figure 9. The photocurrent increases with an increase in voltage.
Photocurrent spectra of films of different composition are shown in Figure 10. The photocurrent spectra show a peak near the absorption edge, which was also observed by Rose [16]. The band gap of the films determined from the spectral response coincided with the band gap determined from optical absorption measurements for that composition. The low photocurrent in the short wavelength range may be due to the high absorption coefficient, and only surface region where defect states give a shorter life time is excited. In the high wavelength region the radiation is only partially absorbed, giving rise to less photocurrent than the peak value. Tails of the spectra extending to long wavelength are attributed to direct excitation of the carriers from the defect levels. Similar results have been reported by several workers for CuInSe2 films [17, 18].
| (a) = 0.9 |
| (b) = 0.6 |
| (c) = 0.3 |
| (d) = 0.1 |
Photosensitivity is the ratio of the increase in conductivity of the material in the presence of light to the conductivity in darkness and is given by the relation where and represent the current under illumination and in the dark, respectively. It seems that some transitions that create additional free carriers effectively increase the free life time increasing the photosensitivity of the material. Figure 11 shows a plot of photosensitivity versus light intensity of thin films of different compositions. Thinner films exhibit moderate photosensitivity, whereas thicker films are found to exhibit higher photosensitivity. Crystallographical imperfections acting as trapping centers will enhance the photosensitivity, whereas the recombination centers decrease the photosensitivity.
4. Conclusions
The brush plating technique has been successfully employed for the deposition of films of different compositions. Films with crystallite size in the range of 30–70 nm were deposited. The band gap of the films could be varied in the range of 1.12 eV to 1.63 eV as the gallium concentration increased. The resistivity of the films was in the range of 0.1 ohm cm to 12 ohm cm. Linear photocurrent-voltage characteristics were observed in all cases. The photocurrent increased linearly with intensity of illumination.