Thermal sputtering as a gas-dynamic process

Abstract

Thermal sputtering, which has traditionally been interpreted either as true vaporization or as vaporization from an elastic collision spike, was suggested by Urbassek and Michl to be more correctly a gas-dynamic process. We here develop a gas-dynamic analysis based on one-dimensional unsteady adiabatic expansion (UAE). We assume UAE from a system with finite extent, with an initially uniform temperature which exceeds the thermodynamic critical temperature (Ttc), and with an initially uniform density like that of the target. We take the UAE as ceasing abruptly and becoming collision-free molecular flight when the gas density falls to a critical value. We conclude that the experimental claim that the energy flux appropriate to thermal sputtering has Maxwellian form is largely acceptable given the experimental uncertainties, that the reported temperatures are also largely acceptable, and that the yields will greatly exceed those predicted for collisional sputtering. Perhaps the most important conclusion, however, is that there is nothing unphysical about the inequality T > Ttc, a problem that has long beset models of thermal sputtering based on true vaporization. © 1990.