Close space sublimation of CdTe for solar cells and the effect of underlying layers

Abstract

This work has focused on the design, construction and testing of a close space sublimation system for CdTe deposition. In addition, it also focused on variations to the treatment and fabrication procedures of the transparent conducting oxide and CdS layers prior to the CdTe deposition, in order to influence the structure and electrical properties of the CdTe/CdS interface. CdTe was deposited by the physical vapour process, close space sublimation. The equipment used was custom built for this work and is therefore described in detail. Some of the deposition parameters of the equipment were varied in order to allow a comparison between the equipment in this work and those reported in the literature. Bilayers of CdS have been created by depositing two individual CdS layers, on top of one another, whilst also varying the annealing treatment of the first layer. These bilayers were included in devices and which subsequently underwent a series of etches to allow analysis of the materials. The bilayers were shown to effect the preferred orientation of the CdTe layer deposited on top, but the results suggested the bilayers had a negative effect on the amount of current recombination in the depletion region. The roughness of the transparent conducting oxide coated glass substrates used in this work, was proposed as a way of influencing the roughness of the CdS layer deposited on top and therefore the roughness of the layer the CdTe is deposited on. It was proposed that the CdTe material deposited on a roughened substrate may have an increased grain size; no evidence for this was found for samples including CdS layers. A second phase of CdTe was observed and assigned to the presence of a CdSxTe1−x alloy. The sulphur content of the alloy was found to increase as the roughness of the substrate increased. CdS was deposited by the wet chemistry process chemical bath deposition. In situ monitoring of the deposition was carried out by studying the re ection of light from the interface at the end of an optical fibre placed within the solution. The results indicated that the in situ monitoring was sensitive to the initial deposition but the sensitivity reduced as the film thickness approached 150 nm. The in situ monitoring technique and ellipsometry measurements of the CdS showed good agreement with the theory of two deposition mechanisms occurring during the deposition. The results showed that CdS grown by chemical bath deposition has a two layer structure consistent with; a low density CdS grown by homogeneous deposition, on top of dense CdS grown by ion by ion deposition. An ellipsometry model for the CdS and CdTe layers is presented, along with its optical properties over the wavelength range 245 to 1690 nm.