Clustering of Metals on Semiconductors Surfaces: Relation to Metallicity, Surface Diffusion, Growth Modes and Schottky Barriers
Photoelectron spectroscopy (PES) has been one of the primary techniques with which Schottky barrier formation on III-V semiconductor substrates has been measured and correlated with overlayer morphology and chemistry. One important class of overlayers is the unreactive materials such as In, Ga and Ag, which cluster on III-V substrates kept at room temperature. Partially reactive materials such as Al and Au have also been observed to cluster on these substrates. Clustered systems have a nonuniform distribution of pinning sites on the surface, and as a result, nonuniformities in the surface potential are also anticipated.1 Such nonuniform surface potentials complicate the interpretation of band bending measurements from these systems. It is the purpose of this paper to describe a model treating the clustering problem.
KeywordsSurface Potential Metal Cluster Depletion Region Cluster Diameter Surface Fermi Level
Unable to display preview. Download preview PDF.
- 9.K. Miyano, D. M. King, C. J. Spindt, T. Kendelewicz, R. Cao, Z. Yu, I. Lindau and W. E. Spicer, Phys Rev. B (to be published)Google Scholar
- 11.PISCES and STRIDE are semiconductor device modelling programs that can simultaneously solve the Poisson equation and the electron and hole continuity equations over regions of semiconductor, metal, and insulator. In this paper the equilibrium case is of interest, and only the Poisson equation needs to be solved. See M. R. Pinto, C. S. Rafferty, and R. W. Dutton, PISCES-II-Poisson and Continuity Equation Solver, Stanford University Technical Report, Stanford University, 1984.Google Scholar
- 13.PISCES and STRIDE assume that the ionized donors and acceptors provide a constant background of charge. For example the charge in the depletion region is taken to be uniformly distributed. Even for the highest substrate dopings considered in this study, the ionized donors and acceptors are actually separated by tens of Å. As pointed out in Ref. 1, a more accurate calculation would take the discrete distribution of this charge into account.Google Scholar
- 14.PISCES is not necessary for a one-dimensional analysis. 1-d as established by PISCES is always in excellent agreement with the distance of 90% potential variation that is computed from the simple depletion approximation: (1 – 0.1) x1-d, where x1-d is expressed in formula (1).Google Scholar