AVS 45th International Symposium
    Surface Science Division Wednesday Sessions
       Session SS1-WeM

Paper SS1-WeM5
Direct Determination of the Interplay between Gas-Surface Reactivity, Thin Film Strain and Alloy Surface Segregation: Growth of Strained Heteroepitaxial Si@sub 1-x@Ge@sub x@ Thin Films on Si(100)

Wednesday, November 4, 1998, 9:40 am, Room 308

Session: Physics of Semiconductors
Presenter: A.M. Lam, Cornell University
Authors: A.M. Lam, Cornell University
Y.J. Zheng, Cornell University
J.R. Engstrom, Cornell University
Correspondent: Click to Email

Silicon-Germanium alloys have attracted attention for more than a decade. The addition of Ge to Si epitaxial films leads to a variety of effects and phenomena, some well understood, some not. Almost untouched are phenomena related to the dissociative adsorption of the relevant thin film precursors on surfaces of the growing alloy. Knowledge of the corresponding sticking coefficients is key to developing relationships between gas phase, surface and sub-surface/bulk compositions. Concerning alloy (Ge) segregation, virtually all work conducted to date has relied on a single in situ experimental probe. We shall present here a comprehensive study of the heteroepitaxial growth of Si@sub 1-x@Ge@sub x@ alloys on Si(100) employing no less than 3 complementary in situ probes: supersonic molecular beam scattering, x-ray photoelectron spectroscopy (XPS), and low-energy ion scattering spectrometry (LEISS). Our results for the first time demonstrate the importance of strain in determining gas-surface reactivity in these heteroepitaxial alloys-- the reactivity of the strained Si@sub 1-x@Ge@sub x@ overlayers is not simply an appropriate combination of pure component (bulk, relaxed) reactivities. On the other hand, use of the complementary in situ surface analyses of XPS and LEISS has allowed us to determine both the extent and nature of Ge surface segregation, and its dependence on film growth conditions. In all cases the surface Ge concentration exceeds that found in the bulk (often by an order of magnitude). A two-site model (involving surface and bulk) cannot account for both the XPS and LEISS results, rather a model invoking Ge enrichment in the subsurface layers is required to explain the data.