Quasi-Chemical Theory with Cluster Sampling from Ab Initio Molecular Dynamics: Fluoride (F$^-$) Anion Hydration

TL;DR

This paper improves the prediction of fluoride ion hydration free energies by using cluster sampling from ab initio molecular dynamics, addressing challenges posed by complex hydrogen bonding and energy landscapes.

Contribution

It introduces a method combining quasi-chemical theory with cluster sampling from ab initio molecular dynamics to enhance hydration free energy predictions for anions.

Findings

Cluster sampling from ab initio MD improves hydration free energy accuracy.

Implicit solvation models can introduce errors for anion clusters.

Careful application of quasi-chemical theory reduces anharmonic effects.

Abstract

Accurate predictions of the hydration free energy for anions typically have been more challenging than for cations. Hydrogen bond donation to the anion in hydrated clusters such as $\mathrm{F(H_2O)}_n{}^-$ can lead to delicate structures. Consequently, the energy landscape contains many local minima, even for small clusters, and these minima present a challenge for computational optimization. Utilization of cluster experimental results for the free energies of gas-phase clusters shows that, even though anharmonic effects are interesting, they need not be troublesome magnitudes for careful applications of quasi-chemical theory to ion hydration. Energy-optimized cluster structures for anions can leave the central ion highly exposed and application of implicit solvation models to these structures can incur more serious errors than for metal cations. Utilizing cluster structures sampled from ab initio molecular dynamics simulations substantially fixes those issues.