Viscous flow model of ion-induced pattern formation: consistency between theory and experiment

TL;DR

This study demonstrates that a viscous flow model can consistently explain critical angles, wavelengths, and stress data in ion-induced surface pattern formation on silicon and germanium, supporting its role in a unifying theory.

Contribution

The paper shows that a viscous flow model explains experimental data on ion-induced pattern formation without needing erosion or redistribution effects.

Findings

Viscous flow model matches critical angle data.

Model explains wavelength and stress measurements.

Supports viscous flow as key mechanism in pattern formation.

Abstract

It is known that ion-irradiation can lead surfaces to spontaneously develop patterns with characteristic length scales on the order of only a few nanometers. Pattern formation typically occurs only for irradiation angles beyond a critical angle, which varies with projectile, target, and irradiation energy. To date, there is no completely unifying physical theory. However, since predictions of the critical angle can be extracted from linear stability analysis of a given model, the ability to explain critical angle selection is a simple but important test of model validity. In this paper, we survey all existing critical angle, wavelength, and in-plane stress data for noble gas broad beam ion-irradiation of silicon and germanium by argon, krypton and xenon at energies from 250eV to 2keV. While neglecting the effects of erosion and redistribution, which are widely regarded as key contributors to ion-induced pattern formation, we find that a viscous flow model is capable of explaining these three types of experimental data simultaneously using only parameters within a physically-reasonable range, and in a manner consistent with the defect kinetics of the amorphous layer. This bolsters the case for viscous flow and stress-driven instabilities as important components of an eventual, unifying model of nanoscale pattern formation under ion bombardment.