Overlayer-induced valence states, and evidence for charge transfer in Na/GaP(110) and Na/GaAs(110): A comparative photoemission study

In the study of the formation of the interfacial electrical properties of the metal-semiconductor contact, the precise relationship between the Schottky barrier height and properties of the semiconductor (electron affinity, surface states, defect states), metal (work function, electronegativity), and interface (induced gap states, impurity states) remains unclear. It is now apparent that such junctions have very different chemical and electrical properties depending on factors such as the reactivity of the metal-semiconductor system, the growth temperature, and the surface preparation prior to metal contact deposition. Even comparison with the detailed calculations available for certain systems is difficult due to the disruption of the surface layers as a result of chemical reaction and interdiffusion observed under normal experimental conditions. Alkali metal contacts to III-V semiconductor surfaces have received much recent attention by a wide range of experimental techniques which utilize ultrahigh vacuum conditions and well-characterized semiconductor surfaces for uncontaminated metal contact growth. Cs overlayers on III-V surfaces have been particularly widely studied and are found to be unreactive. However, at room temperatures, the sticking coefficient of this metal limits the coverage to around a single monolayer, although thicker layers have been grown at reduced temperatures of – 100 K. However, the Na-GaAs( 110) interface has been shown to be reactive at room temperature and inert at 85 K. Detailed calculations are limited to Na and K contacts to GaAs(1lO), and these have provided a detailed description of the geometric and electronic structure of these interfaces. Total energy calculations suggest a transfer of charge between the outer valence level of the alkali metal and the empty surface state associated with the Ga atoms, although the extent of charge transfer remains a question of debate. Experimental studies have not been able, as yet, to reveal directly such charge transfer and the corresponding influence on the developing interfacial potential barrier. The technique of surface-sensitive core and valence level photoelectron emission spectroscopy (PES) is well suited to the parallel monitoring of the chemical and electronic properties of a metal-semiconductor interface as the metal layer is deposited on the clean (110) surface. This study compares the development of the related Na/GaP( 110) and Na/GaAs( 110) interfaces in the light of the available theoretical predictions for such contacts. In both cases, shifts in substrate and overlayer core levels are interpreted as due to charge transfer between the Na3s state and the empty surface state of the semiconductor surface, which gives rise to a new feature in the region of the semiconductor band gap.