Filling transitions on rough surfaces: inadequacy of Gaussian surface models

TL;DR

This study numerically investigates wetting on various rough surfaces, revealing that Gaussian models are inadequate as they produce qualitatively different phase diagrams compared to non-Gaussian topographies, emphasizing careful use of Gaussian assumptions.

Contribution

The paper demonstrates that Gaussian surface models do not accurately represent wetting phenomena on rough surfaces, highlighting the importance of non-Gaussian models for realistic predictions.

Findings

Good agreement with analytical interface-displacement theory

Critical points at positive contact angles in all studied cases

Gaussian surfaces exhibit qualitatively different phase diagrams

Abstract

We present numerical studies of wetting on various topographic substrates, including random topographies. We find good agreement with recent predictions based on an analytical interface-displacement-type theory \cite{Herminghaus2012, Herminghaus2012a}. The phase diagrams are qualitatively as predicted, but differently in this study the critical points are found to lie within the physical parameter range (i.e., at positive contact angle) in all cases studied. Notably, it is corroborated that Gaussian random surfaces behave qualitatively different from all non-Gaussian topographies investigated, exhibiting a qualitatively different phase diagram. This shows that Gaussian random surfaces must be used with great care in the context of wetting phenomena.