High-Dimensional Atomistic Neural Network Potential to Study the Alignment-Resolved O₂ Scattering from Highly Oriented Pyrolytic Graphite

Abstract

A high dimensional and accurate atomistic neural network potential energy surface (ANN-PES) that describes the interaction between one O2 molecule and a highly oriented pyrolytic graphite (HOPG) surface has been constructed using the open-source package (aenet). The validation of the PES is performed by paying attention to static characteristics as well as by testing its performance in reproducing previous ab initio molecular dynamics simulation results. Subsequently, the ANN-PES is used to perform quasi-classical molecular dynamics calculations of the alignment-dependent scattering of O2 from HOPG. The results are obtained for 200 meV O2 molecules with different initial alignments impinging with a polar incidence angle with respect to the surface normal of 22.5° on a thermalized (110 and 300 K) graphite surface. The choice of these initial conditions in our simulations is made to perform comparisons to recent experimental results on this system. Our results show that the scattering of O2 from the HOPG surface is a rather direct process, that the angular distributions are alignment dependent, and that the final translational energy of end-on molecules is around 20% lower than that of side-on molecules. Upon collision with the surface, the molecules that are initially aligned perpendicular to the surface become highly rotationally excited, whereas a very small change in the rotational state of the scattered molecules is observed for the initial parallel alignments. The latter confirms the energy transfer dependence on the stereodynamics for the present system. The results of our simulations are in overall agreement with the experimental observations regarding the shape of the angular distributions and the alignment dependence of the in-plane reflected molecules.

Nitric Oxide Scattering off Graphene Using Surface-Velocity Map Imaging

“…A thermal desorption process would yield n = 1, indicating that despite the significant loss in kinetic energy, a trapping-desorption mechanism does not seem to play a significant role here. The simulated polar angle distribution follows a cos distribution, i.e., a little wider than the experimentally measured distribution, but still very narrow and much narrower than the angular distributions measured for NO scattered off graphite, , albeit these were measured at non-zero incidence angles. This demonstrates that both in the experiment and the simulations, a direct scattering mechanism closely along the surface normal is by far the dominant process.…”

“…There is some qualitative agreement in that the scattered NO molecules lose more than 60% of their mean kinetic energy (from 0.4 to 0.15 eV), but this energy loss is less pronounced compared to the experiments. This discrepancy is likely due to the limitations of the employed force field and could be improved using density functional theory (DFT)-based force fields, − but the overall trend observed is reproduced qualitatively. Future scattering experiments with aligned NO molecular beams could serve as a benchmark for improving the force fields.…”

“…Interatomic forces were truncated after 7 Å. It should be noted that more sophisticated neural network potential energy surfaces have recently been calculated even for surface scattering processes related to the one presented here; − however, not all of these studies take the metal substrate into account, and the literature force fields mentioned abovewhile not verified for the NO-graphene interactionare likely to be sufficient to at least qualitatively reproduce the trends observed experimentally.…”

Photoinduced Desorption Dynamics of CO from Pd(111): A Neural Network Approach

“…In the last few years, the use of neural network (NN)-generated multidimensional PESs has become an accurate alternative to ab initio molecular dynamics (AIMD) to describe the dynamics of diverse gas–surface processes 28 − 39 and also the dynamics at solid–liquid water interfaces. 40 − 43 In particular, for these studies, the development of the atomistic neural network (AtNN) approach has constituted a major advancement.…”

“…Moreover, since the NN-PES can also be made a function of the surface atom coordinates, the treatment of both the independent surface atom movement and the surface temperature effects has also been performed, which has allowed for accounting for energy exchange between the molecule and the surface along the dynamics. 31 − 33 , 35 − 39 …”