On the stoichiometry of the semiconductor pyrite (FeS2)

The compound semiconductor pyrite (FeS2) has recently attracted considerable attention as a possible absorber material for thin-film solar cells. In the literature, a controversy exists on the degree of its non-stoichiometry. In the present paper an analysis of the literature concerning this question is given. In spite of many reports on significant deviations from the ideal stoichiometry, no reliable proof exists that pyrite has an existence range which exceeds compositional variations of the order of 1 at %. To support this result, a theoretical estimation of the vacancy formation energy in pyrite has been performed. Based on the macroscopic cavity model of van Vechten and Phillips the vacancy formation energies E(V)(S) = 1.66 eV of a sulphur vacancy and E(V)(Fe) = 2.18 eV of an iron vacancy were calculated. The formation energy of a complete Schottky defect E(V)(Fe + 2S) = 5.50 eV is in good agreement with a second value E(V)(Fe + 2S) = 5.80 eV, which was derived from Kroger's rule: E(V) approximate to 0.5H(sub), where H-sub is the sublimation energy of pyrite. These fairly high values support the conclusion of this paper that pyrite has only slight deviations from the nominal sulphur-to-iron ratio of 2.0. However, in order to confirm this conclusion, new reliable experiments are necessary. Based on the present results, pyrite can be regarded as comparable with compound semiconductors such as CdS or CdTe, the stoichiometry of which can be controlled by appropriate methods.