Kink dynamics in a one-dimensional growing surface

TL;DR

This paper investigates the dynamics of kinks on a one-dimensional growing surface, revealing how symmetry-breaking terms influence kink profiles, velocities, and coarsening behavior, with implications for surface morphology evolution.

Contribution

It introduces a detailed analytical study of kink classes and their dynamics under symmetry-breaking conditions, extending previous models to more realistic scenarios.

Findings

Two classes of kinks ($A$ and $B$) are analytically characterized.

Symmetry-breaking causes $A$ kinks to narrow and $B$ kinks to widen, maintaining a constant product of widths.

Kink separation scales as $ ext{ln} t$ without noise and as $t^{1/3}$ with noise.

Abstract

A high-symmetry crystal surface may undergo a kinetic instability during the growth, such that its late stage evolution resembles a phase separation process. This parallel is rigorous in one dimension, if the conserved surface current is derivable from a free energy. We study the problem in presence of a physically relevant term breaking the up-down symmetry of the surface and which can not be derived from a free energy. Following the treatment introduced by Kawasaki and Ohta [Physica 116A, 573 (1982)] for the symmetric case, we are able to translate the problem of the surface evolution into a problem of nonlinear dynamics of kinks (domain walls). Because of the break of symmetry, two different classes ($A$ and $B$) of kinks appear and their analytical form is derived. The effect of the adding term is to shrink a kink $A$ and to widen the neighbouring kink $B$, in such a way that the product of their widths keeps constant. Concerning the dynamics, this implies that kinks $A$ move much faster than kinks $B$. Since the kink profiles approach exponentially the asymptotical values, the time dependence of the average distance $L(t)$ between kinks does not change: $L(t)\sim\ln t$ in absence of noise, and $L(t)\sim t^{1/3}$ in presence of (shot) noise. However, the cross-over time between the first and the second regime may increase even of some orders of magnitude. Finally, our results show that kinks $A$ may be so narrow that their width is comparable to the lattice constant: in this case, they indeed represent a discontinuity of the surface slope, that is an angular point, and a different approach to coarsening should be used.