The Effect of Doping Ratios on Structure, Composition, and Electrical Properties of Absorber Formed by Thermal Sintering
Chalcopyrite compounds of copper indium gallium diselenide (CIGS) absorber were fabricated by using binary-particle (, , and ) precursors with thermal sintering method. The binary-particle ink was firstly prepared by milling technology and then printed onto a soda lime glass substrate, which was baked at a low temperature to remove solvents and form a dry precursor. Following milling, the average particle size of agglomerated CIGS powder is smaller than 1.1 μm. Crystallographic, stoichiometric, and electrical properties of precursor CIGS films with various doping amounts of had been widely investigated by using thermal sintering in a nonvacuum environment without selenization. Analytical results reveal that the CIGS absorption layer prepared with a doping ratio of 3 has a chalcopyrite structure and favorable composition. The mole ratio of Cu : In : Ga : Se of this sample was 1.03 : 0.49 : 0.54 : 1.94, and related ratios of Ga/(In + Ga) and Cu/(In + Ga) were 0.52 and 0.99, respectively. The resistivity and carrier concentration were 3.77 ohm-cm and 1.15 E + .
Chalcopyrite compounds of copper indium gallium diselenide (CIGS) and related alloys are among the most promising materials for photovoltaic applications . CIGS-based solar cells have potential portable applications because of their high absorption coefficient, large absorption spectrum, and high stability. Recent investigations have revealed efficiencies as high as 20.3% for thin-film CIGS-based solar cells .
A set of low-cost approaches with the realistic potential to reduce direct manufacturing costs has been developed in the last five years. Industrial production of CIGS solar cell is mostly based on vacuum processes, which depends on a high initial investment in manufacturing machines. These low-cost techniques typically involve the use of simple and fast nonvacuum deposition methods and the molecular prefixing of the film composition in a precursor layer, which is chemically and thermally treated to form a high-quality CIGS film.
In recent years, CIGS absorbing layers had been fabricated using nanoparticles [3–5] and other coating methods. For nanoparticle-based process, there are two approaches to be used to form the CIGS absorption layer. One is the preparation of a precursor film by synthesizing CIGS powder  and the other one is a mechanochemical process (MCP)  or a rotary ball milling (RBM) method implemented before nonvacuum coating . For MCP, four elemental metal powders of Cu, In, Ga, and Se are mixed and milled by the planetary ball milling technology, yielding a particle size of 1 μm after milling . For RBM, the precursor has quaternary compound composition ratios of Cu/(In + Ga) = 0.95, Ga/(In + Ga) = 0.39, and Se/(Cu + In + Ga) = 0.75. After milling, the particle size of the agglomerated CIGS powder is less than 100 nm . Thus, thermal sintering can cause precursor films forming the dense absorption layer for use in solar cells . All steps performed in a nonvacuum environment are relatively simple and inexpensive, such as ease of scalup  and excellent control of the ratios of the metal concentrations. To fabricate low-cost and printable CIGS thin-film solar cells, Lee et al.  developed a precursor solution by using a nanoparticle-based method in 2011. In that process, a nearly carbon-free CIGS film can be obtained by applying a three-step heat treatment process: the first step is the elimination of carbon residues by air annealing, the second step is the formation of CIGS alloys by sulfurization, and the third step is the grain growth and the densification of CIGS films by selenization.
This work proposes another process to fabricate a CIGS absorption layer by using binary-particle precursors, such as Cu2Se, In2Se3, and Ga2Se3, through thermal sintering. Crystallographic, stoichiometric, and electrical properties of CuIn0.5Ga0.5Se2 absorber prepared with binary-particle precursors affected by the doping amount of Ga2Se3 were examined and analyzed through various measurements, such as scanning electron microscopy (SEM), X-ray diffractometry (XRD), energy dispersive spectrometry (EDS), Hall measurement, and thermogravimetry analyzer (TGA). Consequently, an optimized doping amount of the Ga2Se3 for forming CuIn0.5Ga0.5Se2.0 absorption layers could be obtained.
Three binary-alloy powders were mixed with appropriate solvents, whose initial stoichiometric ratios are presented in Table 1. Those powders were then ground into binary-alloy particles by using a mill technology to form the ink that was required for preparing the film precursors. Three commercially available binary-alloy powders of In2Se3, Ga2Se3, and Cu2Se with grain sizes of approximately 74 μm (Williams Advanced Materials Technologies Taiwan Co., Ltd., 99.99%) were used as starting materials for forming CIGS absorption layers in a nonvacuum environment.
To obtain the ink for forming CIS and CIGS precursor films, three starting materials were mixed with solvent and zirconium beads; each zirconium bead (diameter: 0.5 mm and mass: 20 g) was ground continuously in a mill for eight hours to produce the required ink. The SEM photograph shows that the average grain size of agglomerated CIGS powders is less than 1.1 μm, as shown in Figure 1. After the ink printed onto a soda lime glass substrate, the solvent was removed by placing the sample into an oven at a temperature of 120°C and maintaining for 5 min. Then, the temperature was increased from 120 to 200°C at the same rate and maintained at 200°C for 10 min. Once the oven turned off, the expected dry precursor film could be obtained when the sample was cooled down to the room temperature. Subsequently, the CuIn1-x GaxSe2.0 absorption layer for use in solar cells can be formed by thermal sintering.
The weight loss of Cu2Se, In2Se3, and Ga2Se3 measured by TGA is as shown in Figure 2. The data shows that the loss of Cu2Se is 2% at 520°C, whereas the losses of In2Se3 and Ga2Se3 are 32.5% and 25%, respectively, at 420°C. Based on these data, the sintering temperature was set at 420°C and maintained for 30 min; then the temperature was increasing from 420 to 520°C and maintained at 520°C for one hour. Subsequently, prepared precursor samples were heated in a thermal annealing furnace filled with nitrogen. Thus, CuIn1-x GaxSe2.0 films with highly uniform crystalline structure could be formed. After one hour, the temperature was dropped below 50°C. Since the amount of Ga2Se3 can affect the film structure of CuIn1-x GaxSe2.0 sample, the properties of the sample can be analyzed according to this point.
3. Results and Discussion
3.1. Structural Properties
Figure 3 presents the XRD patterns of samples that were sintered with various amounts of Ga2Se3. The XRD measurement was executed by scanning the diffraction angle from 20 to 60° and using a grazing incidence angle of 1°. In Figure 3, XRD patterns display that sintered CIGS samples have many diffraction peaks at (112), (220)/(204), and (312)/(116), corresponding to different crystalline structures, respectively.
From sample CIS to sample CIGS-4, the XRD peak at (112) shifts from = 26.575° to = 27.175°. According to the results reported by Balboul et al. , the compositions were determined by calculating the peak shift with the ratio of In : Ga. As Ga content increases, the doublet of peaks, that is, (220)/(204) peak and (116)/(312) peak, is splitting into two individual parts. This splitting indicates that the deviation of tetragonality, that is, , was resulted by Ga substitution. These results are ascribed to the small ionic size of Ga, 0.62 Å, compared to that of In, 0.81 Å. As Ga2Se3 ratio increases, all signals shift towards larger diffraction angles. Figure 3 shows this effect on the signal of (112). This noticeable shift is due to the decrease in lattice constants “” and “” as reported by Olejníček et al. . The chalcopyrite structure of CuIn0.5Ga0.5Se2.0 was modeled using the space group I42d (122) with standard lattice constants of = 5.673 Å and = 11.322 Å (JCPDS-40-1488) [11, 12]. Table 1 and Figure 3 reveal that the data of the sintered sample with a Ga2Se3 doping ratio of 3 (CIGS-3) are very close to those on the JCPDS card (40-1488) for Cu(Ga0.5In0.5)Se2.0 at the angles of 27.075°, 44.975°, and 53.325°, respectively.
The stoichiometric compositions of CIGS samples having various Ga2Se3 doping ratios were analyzed by using EDS data, as shown in Table 1. An attempt was also made to optimize the Cu, In, Ga, and Se contents in the CIGS absorption layer of solar cells. The Ga/(In + Ga) and Cu/(In + Ga) mole ratios presented in Table 1 were obtained by using various Ga2Se3 doping ratios. For samples sintered with various Ga2Se3 doping ratios, their ratios of Ga/(In + Ga) and Cu/(In + Ga) are shown in Figure 4. Since the mole ratios of Ga/(In + Ga) and Cu/(In + Ga), respectively, are 0.52 and 0.99, the doping ratio of Ga2Se3 for CIGS-3 sample is the optimal. For this sample, its doping ratio of Ga2Se3 is 3. This reveals that the Cu : In : Ga : Se mole ratio of CIGS-3 is 1.03 : 0.49 : 0.54 : 1.94, which is approximately 1 : 0.5 : 0.5 : 2.0 reported elsewhere . For fixed Cu content, the Ga/(In + Ga) ratio and the at (112)-peak of each sintered sample corresponding to various Ga2Se3 doping ratios are shown in Figure 5. Furthermore, the main XRD peaks (112) shown in Figure 3 display a noticeable shift that the higher the angle, the larger the doping ratio of Ga2Se3, which has also been reported elsewhere [14–17]. This is attributed to relatively small Ga atoms substituting for larger In atoms in the chalcopyrite structure , resulting in a decrease in the lattice parameter.
3.3. Electrical Properties
Parameters related to electrical, structural, and compositional properties of each CIGS sample are listed in Table 1. The carrier concentration and the electrical resistivity of all sintered CIGS samples were determined by Hall measurement , as shown in Figure 6. Measuring data shows that all samples are p-type. At room temperature, each sample has an electrical resistivity between 0.408 and 44.8 ohm-cm and an associated carrier concentration from 2.44 E + 15 to 6.25 E+ 18 cm−3. The CIGS-1 sample has the lowest resistivity of 0.408 ohm-cm and the highest carrier concentration of 6.25 E + 18 cm−3. The CIS sample has the highest resistivity of 44.8 ohm-cm and the lowest carrier concentration of 2.44 E + 15 cm−3. For specified Ga content, Figures 7 and 8, respectively, show the ratios of Cu/(In + Ga) and Ga/(In + Ga) of each sintered sample and their related resistivity and carrier concentration. These two figures reveal that CIGS-1 had the highest ratio of Cu/(In + Ga), the lowest resistivity of 0.408 ohm-cm, and the highest carrier concentration of 6.25 E + 18 cm−3. Sample CIGS-4 has a lower ratio of Cu/(In + Ga) and a higher ratio of Ga/(In + Ga), as well as the highest resistivity of 41.8 ohm-cm and a carrier concentration of 1.66 E + 17 cm−3. For sample CIGS-3, however, its resistivity and carrier concentration are 3.77 ohm-cm and 1.15 E + 18 cm−3, respectively. Moreover, according to the pieces of literature reported by Turcu and Rau and Zhang et al. [19, 20], the Cu-rich films phase has lower electrical resistivity and higher carrier concentration, whereas the Cu-poor films phase has higher electrical resistivity and lower carrier concentration.
Based on the data shown in Section 3.2, the ratio of Cu/(In + Ga) of the sample CIGS-3 is nearly equal to one, so that it is very suitable for the CuIn1-x GaxSe2.0 absorbing layer having the required resistivity and carrier concentration [21–26]. For this sample, the ratios of Cu/(In + Ga) = 0.99 and Ga/(In + Ga) = 0.52 were optimal because the mole ratio of Cu : In : Ga : Se is very close to the expected value of 1 : 0.5 : 0.5 : 2.0 . At the same time, the data of CIGS-3 is also close to that on the JCPDS card (40-1488) for Cu(Ga0.5In0.5)Se2.0. Additionally, the data of at the peak of (112) increases with the doping amount of Ga2Se3. This is consistent with the result reported by other several works [14–17].
In this paper, a CIGS absorber prepared by using binary-particle (Cu2Se, In2Se3, Ga2Se3) precursors via thermal sintering method had widely been investigated. The required binary-particle ink for fabricating the absorbing layer was prepared by using the milling technology. The SEM photograph reveals that the average particle size of the agglomerated CIGS powder is smaller than 1.1 μm after milling. Analytical results obtained from measurements of SEM, XRD, EDS, Hall, and TGA reveal that the CIGS-3 sample has a chalcopyrite structure and a favorable composition of CuIn0.5Ga0.5Se2.0. The Cu : In : Ga : Se mole ratio of CIGS-3 is 1.03 : 0.49 : 0.54 : 1.94, and its resistivity and carrier concentration are 3.77 ohm-cm and 1.15 E + 18 cm−3, respectively. Among those sintered samples prepared in our laboratory, the data of CIGS-3 is very close to the data on the JCPDS card (40-1488) for Cu(Ga0.5In0.5)Se2.0. In summary, our investigations mainly focused on crystallographic, stoichiometric, and electrical properties of CuIn0.5Ga0.5Se2.0 thin film and obtained the optimized doping amount of Ga2Se3 in a CIGS absorbing layer.
Conflict of Interests
All authors declare that there is no conflict of interests regarding the publication of this paper.
The authors would like to thank the National Science Council, Taiwan, for financially supporting this research under Contracts nos. NSC 99-2221-E-155-051 and NSC 100-2221-E-155-045.
C. P. Liu and C. L. Chuang, “Fabrication of copper–indium–gallium–diselenide absorber layer by quaternary-alloy nanoparticles for solar cell applications,” Solar Energy, vol. 86, no. 9, pp. 2795–2801, 2012.View at: Google Scholar
B. Vidhya, S. Velumani, J. A. Arenas-Alatorre, A. Morales-Acevedo, R. Asomoza, and J. A. Chavez-Carvayar, “Structural studies of mechano-chemically synthesized CuIn1−xGaxSe2 nanoparticles,” Materials Science and Engineering B, vol. 174, no. 1–3, pp. 216–221, 2010.View at: Publisher Site | Google Scholar
D. Suri, K. Nagpal, and G. Chadha, “X-ray study of CuGaxIn1−xSe2 solid solutions,” Journal of Applied Crystallography, vol. 22, pp. 578–583, 1989.View at: Google Scholar
W. N. Shafarman, R. Klenk, and B. E. McCandless, “Device and material characterization of Cu(InGa)Se2 solar cells with increasing band gap,” Journal of Applied Physics, vol. 79, no. 9, pp. 7324–7328, 1996.View at: Google Scholar
L. J. van der Pauw, “A method of measuring specific resistivity and Hall effect of discs of arbitrary shape,” Philips Research Reports, vol. 13, pp. 1–9, 1958.View at: Google Scholar
R. D. L. Kristensen, S. N. Sahu, and D. Haneman, “Flash evaporation of CuInSe2 films,” Solar Energy Materials, vol. 17, no. 5, pp. 329–345, 1988.View at: Google Scholar
A. N. Tiwari, D. K. Pandya, and K. L. Chopra, “Electrical and optical properties of single-phase CuInS2 films prepared using spray pyrolysis,” Thin Solid Films, vol. 130, no. 3-4, pp. 217–230, 1985.View at: Google Scholar