X-ray photoelectron spectroscopy studies of the role of etching in improving CdSe films for photoelectrochemical solar energy conversion

Abstract

Polycrystalline thin film CdSe photoelectrochemical (PEC) cells have shown a great promise for low cost terrestrial power generation in the past several years 1-5. Many details of the CdSe films, particularly the atomic levels are not understood. Controlled studies of the atomic composition of the surface and structure of films have been reported 6. It is known 7 that chemical etching has an important effect on the solar conversion efficiency of such films but the reasons are not fully understood. We report here the results of X-ray photoelectron spectroscopy (XPS) and scanning electron microscopy (SEM) measurements on the effects of etching of CdSe photoanodes. The CdSe samples used in this study were prepared in a manner similar to previous work 6, by electrolysis of HC1 solution containing CdC12 and SeO 2. A carbon plate was used as the anode and was spaced 2 cm from the titanium substrate. The potential of the substrate was maintained constant with respect to a saturated calomel electrode by a potentiostatic technique and by varying the SeO 2 concentration. The current was passed for a period of 2 h with continuous stirring. The samples were then annealed in air for 2 h above 400 °C. For each experiment, two samples were prepared together, one in the form of a strip and the other as a cylinder suitable for use with the XPS equipment. The strip was used for currentvoltage measurements to check the performance of the surfaces. The counterelectrode and the electrolyte have been described earlier 4. XPS measurements were made with a Varian lEE-15 spectrometer. Annealed unetched films showed the presence of cadmium, selenium and chlorine in the ratios 1:1:0.1 as well as the usual carbon and oxygen. Titanium was present on some samples but not on others. Cadmium and selenium peak positions were identical with those of polycrystalline CdSe (Cleveland Crystals Inc.). The etching solution (aqua regia) was a 12~o aqueous solution of 4:1 HC1-HNO 3 (by volume). The heights of various peaks were measured after each of a series of etching steps. The source of chlorine is probably the HC1 used for the deposition of the CdSe films. Even after the first 2 s of etching, a strong decrease in the chlorine peak was observed. A consistent explanation is that chlorine diffuses to the surface during annealing and is subsequently removed by etching. There was no evidence for the presence of chlorine in the surface of unannealed films. Etching increased the intensity of the titanium peak slightly, if titanium was present before etching. This may be due to the