In this paper, Ruffman et al. evaluate the effect of dynamics in metal-based catalysts, particularly Ga-In and Ga-Pt surfaces. This exciting and timely paper focuses on capturing dynamic effects in catalysis concerning adsorption energies, which are critical descriptors to understanding catalysis.

The paper addresses this topic using ab initio molecular dynamics simulations with the NVT ensemble. It suggests a framework to capture the dynamic behavior of a few metal-based catalytic surfaces and their impact on the adsorption energies and, therefore, in catalysis. The novelty in the paper lies in addressing this concept, not that much in the methods used, which, in the end, are AIMD simulations and single-point energy corrections. Nevertheless, I find the paper deserves publication in Chemical Science for the importance of the proposed concept, provided the following aspects are revised.

The “high”/” low” energy term in the text can be a bit misleading. I understood what the authors meant, but where possible, in the text, I would clarify that high energy means less stable and low energy means more stable (at least, this is how I understood it).

The main text does not mention the DFT functional used (PBEsol) and the code (VASP). This can be found in the ESI, but I think this information is important enough to move it to the main text—at least the most critical aspects of the computational settings.

While the paper focuses on liquid metal, dynamic effects can also play a role on nanoparticles/even relatively large metal surfaces of solid catalysts, especially under the conditions required most of the time in heterogeneous catalysis. This needs to be clarified, besides the simulations being in the gas phase, not the liquid phase. Some papers have addressed this issue in the context of catalytic reactions. Therefore, I think this could be explained in the text, citing the relevant works in the field (such as those using metadynamics) and increasing the paper’s scope to a more general audience.

A critical missing aspect is the entropic contribution. Could the authors provide this for a selected of the evaluated systems to unravel the impact of that? That would bring the paper to a new level. I don’t ask for assessing the reactivity, i.e., transition states, as I think it is out of scope, but that would be nice, too.

The authors have appropriately considered my comments. Thus, I accept the paper.