Grazing‐incidence small‐angle X‐ray scattering (GISAXS) is introduced as a method of studying discontinuous thin films. In this method, the incident beam is totally externally reflected from the substrate followed by small‐angle scattering of the refracted beam by the thin film. The experiment described establishes the ability of GISAXS to provide size information for islands formed in the initial stages of thin film growth. The data presented are for gold films of 7 and 15 Å average thicknesses on Corning 7059 glass substrates. The advantages of this technique are that it is non‐destructive, can be done in situ, provides excellent sampling statistics, does not necessarily require a synchrotron source, and is not limited to thin or conducting substrates.

No abstract

No abstract

The most general analysis of diffuse X-ray scattering to obtain information about local atomic arrangements in disordered binary solid solutions has been presented by Borie and Sparks. It involves an expansion of the diffuse intensity into various components, each describing separately the local order or moments of the local displacement distribution. Each of these components has a different symmetry and is periodic in reciprocal space under the assumption that the ratio of atomic scattering factors does not vary drastically with sin θ/λ. This assumption does not hold in many systems of practical interest. Furthermore, even if valid, the information about local displacements is of limited value since the displacements of the two atomic species cannot be determined separately. Presented here is an analysis of a modification of the Borie and Sparks method, first suggested by Tibballs, which minimizes these problems. The displacement intensity components are broken down further into terms each involving a different atomic species. These new terms are strictly periodic in reciprocal space and the variation of the atomic scattering factor ratio is actually used to separate them. Diffuse scattering for a number of binary alloys was simulated by computer and subsequently analyzed by this method, including some commonly occuring errors (counting statistics, background errors, etc.). The errors propagating to the final SRO coefficients were never worse than those involved in a Borie and Sparks analysis, and the displacement coefficients for each atomic species could be determined within reasonable errors. An on-line computer program for making measurements with this technique is described

No abstract

No abstract

No abstract

No abstract