CuGaInSe2 Thin Film Grown by a New Method: Oscillating Temperature Selenization of Metallic Elemental Layers in a Se Vapour

The CuInSe2, with its direct band gap of 1.0 eV and high absorption coefficient, is one of the most promising semiconductors for thin film solar cells. Devices based on this material [1] (with a little amount of Ga) have reached 16.9% conversion efficiency which is the highest efficiency so far reported for any kind of thin film solar cell. However, Stolt et al. [1] use a multisource evaporation system for the preparation of CuGaInSe2 films, a technique which is not suitable for the fabrication of large area devices. Another technique, which is easily scalable, is the selenization of metallic elemental layers in an atmosphere containing H2Se. This technique has given quite good results. For example Basol et al. [2] have reported an efficiency of about 12-13% for solar cells obtained by preparing CulnSe2 film by selenization in H2Se atmosphere. This technique has a drawback, which consists in the use of H2Se a very poisoning gas. In our laboratory we have developed a new simple method to grow thin films of this material. We have grown the CuGaInSe2 thin films by selenizing the Cu, Ga, In elemental layers in a Se vapour. The main characteristic of the process consists in the fact that the first part of the film is grown with the substrate temperature oscillating between 200-230°C and 400- 500°C. These films have been used to fabricate CuGaInSe2/CdS solar cells with a conversion efficiency close to 12% |4J. Most of the tested cells exhibit an open-circuit voltage of 420-430 mV, a short circuit current close to 40 mA/cm2 and fill factor of 0.65-0.68 with efficiency, which in the best cases was close to 12%. Probably, if the junction is made with a thin layer of CdS obtained by the chemical bath deposition [3|, a larger efficiency can he achieved. From the spectral response of these cells it appears that the absorber layer has a gap dose to 1 eV, which is typical of CuInSe2, and not of a mixed CuGaInSe2 compound. On the other hand both open-circuit voltage and short-circuit currents observed in these cells are typical of CuInSe2. This method of growing the CuGaInSe2 films probably separates CuInSe2 from CuGaSe2 and it seems that CuInSe2 grows on the surface of the film. In conclusion, a new promising process to prepare CuInSe2 films containing a certain amount of Ga has been successfully experimented. The process is scalable and does not imply the use of poisonous gas. [1] L. Stolt, Photovoltaic Insiders Report, Vol . XII n. 6, June 1993 [2] B.M. Basol, V. Kapur and A. Halany, Proc. 22th IEEE Photov. Spec. Conf., IEEE, New York (1991), 893 [3] Lincot et al., Proc. llth E.C. Photov. Solar Energy Conf., October 1992, Harwood Academic Publ., Switzerland, p. 870 [4] N. Romeo, A. Bosio and V. Canevari, Proc. 12th E.C. Photovoltaic Solar Energy Conf. Amsterdam, The Netherlands, 11-15 April 1994, in press