The Desorption Rate at Liquid-Solid Interface

TL;DR

This study models atom desorption at a liquid-solid interface using simulations, revealing that the desorption rate follows Arrhenius behavior but mean-field theories fail due to barrier recrossing and fluctuations.

Contribution

It introduces a combined Monte Carlo and molecular dynamics approach to analyze desorption rates and tests the limitations of mean-field theories in this context.

Findings

Desorption rate coefficient follows Arrhenius relation.

Mean-field theories do not accurately predict the desorption rate.

Barrier recrossing and fluctuations cause deviations from theoretical predictions.

Abstract

We use a simple generic model to study the desorption of atoms from a solid surface in contact with a liquid, by using a combination of Monte Carlo and molecular dynamics simulations. The behavior of the system depends on two parameters: the strength $\epsilon_{LS}$ of the solid-liquid interaction energy and the strength $\epsilon_{LL}$ of the liquid-liquid interaction energy. The contact with the solid surface modifies the structure of the adjacent liquid. Depending on the magnitude of the parameters $\epsilon_{LS}$ and $\epsilon_{LL}$ the density of the liquid oscillates with the distance from the surface for as far as five atomic layers. For other values of the parameters the density of the liquid near the surface is much lower than that of the bulk liquid, a process sometimes called dewetting. To describe the desorption rate we have determined the number $N(t)$ of atoms that were adsorbed initially and are still adsorbed at time $t$. The average of this quantity decays exponentially and this allows us to define a desorption rate coefficient $k_{d}$. We found that $k_{d}$ satisfies the Arrhenius relation. The logarithm of the pre-exponential is a linear function of the activation energy (the so-called compensation effect). We have also examined two approximate mean-field like theories of the rate constant: one calculates the activation free energy for desorption; the other uses transition state theory applied to the potential of mean force. Both methods fail to reproduce the exact desorption rate constant $k_{d}$. We show that this failure is due to the presence of substantial recrossing of the barrier (which introduces errors in transition state theory) and the presence of large fluctuations in the desorption rate of individual molecules whose effect is not captured by mean-field theories.