Sticking in the quantum regime:H2andD2on Cu(100)

Abstract

Several questions of importance have yet to be clarified concerning the sticking of a light particle on a surface. Among these are the following: the relation between permanent "sticking" and temporary "trapping," the "scaling" of the sticking coefficient with incident angle, and the role of quantum-mechanical resonance phenomena in promoting sticking. We consider these and other questions in connection with data taken for the sticking of ${\mathrm{H}}_2$ and ${\mathrm{D}}_2$ molecular beams in the energy range 8-45 meV and for angles of incidence between 0° and 60° on a cold (∼10-K) Cu(100) surface. The sticking coefficient at zero coverage, $S_0$, measured with use of partial monolayer desorption, displays a background that falls off with increasing energy on which are superposed well-defined peaks at characteristic energies that coincided with features in the specular reflectivity and with the condition for "selective adsorption." The peaks are due to the formation of quasibound states that are entered via the surface corrugation, the rotational anisotropy, or both acting together, and decay via sticking and backscattering channels. These states have been studied extensively in elastic scattering measurements. We demonstrate their decay via sticking channels and show that these channels give important contributions to the resonance widths. By comparing the dependence of the observed background sticking on energy and angle with calculations, we conclude that positive-energy trapping on initial collision is prevalent at wide incident angles. The trapping probability is governed primarily by the normal rather than total energy. However, many of the trapped particles revert to the gas phase and do not contribute to the sticking coefficient. As a result, the background sticking coefficient of ${\mathrm{H}}_2$ and ${\mathrm{D}}_2$ on Cu(100) does not display a simple scaling behavior as the angle of incidence changes.