The bonding of alkali metals to semiconductor surfaces: a direct and inverse photoemission study

The bonding character (ionic versus covalent) of alkali metals on semiconductor surfaces is controversial. This concerns the amount of charge transfer and the question where a possible metallization takes place (alkali overlayer or semiconductor substrate). To give an answer we have employed direct and inverse photoemission spectroscopy to Na, K, and Cs on Si(111)2×1 and 7 × 7 and on GaAs(110), and K on Si(100)2 × 1. We find that there is no iso-electronic behavior, e.g., monolayer coverages may result in a metallic surface {Si(111)2 × 1K} or leave the surface semiconducting {Si(111)1 × 1Na, Si(111)√3 × √3R30°Cs} with specific surface atom re-arrangements. The reason may lie in the increasing competition between the alkali-semiconductor and alkali-alkali interaction as a function of coverage, both being dependent on the atomic radii of the alkali-metal atoms. Our results will be compared to theoretical calculations as well as to the findings of other experimental techniques. © 1992.