Small Nuclear Quantum Effects in Scattering of H and D from Graphene

Abstract

We study nuclear quantum effects in H/D sticking to graphene, comparing scattering experiments at near-zero coverage with classical, quantized, and transition-state calculations. The experiment shows H/D sticking probabilities that are indistinguishable from one another and markedly smaller than those expected from a consideration of zero-point energy shifts of the chemisorption transition state. Inclusion of dynamical effects and vibrational anharmonicity via ring-polymer molecular dynamics (RPMD) yields results that are in good agreement with the experimental results. RPMD also reveals that nuclear quantum effects, while modest, arise primarily from carbon and not from H/D motion, confirming the importance of a C atom rehybridization mechanism associated with H/D sticking on graphene.

An experimentally validated neural-network potential energy surface for H-atom on free-standing graphene in full dimensionality

“…By computing several hundred thousand classical trajectories we demonstrated the utility of the PES by simulating angle-and energy-resolved H-and D-atom scattering experiments similar to those recently published. 4,17 The theoretical distributions are remarkably close to those seen in experiment. They accurately capture the branching between a quasi-elastic channel that samples only the physisorption well and a high-energy-loss channel that results from trajectories that traverse the chemisorption well.…”

“…16 In this paper, we present the first HDNN-PES for H atoms interacting with free standing graphene, which we validate against data obtained from H and D scattering experiments using graphene grown on Pt(111), experiments that are similar to those recently reported elsewhere. 17 Compared to the REBO-EMFT PES, 4 we achieve substantially reduced fitting errors without sacrificing computational performance. Using MD, we show that experimentally obtained H/D-atom energy loss and angular distributions are faithfully reproduced.…”

OrbNet: Deep learning for quantum chemistry using symmetry-adapted atomic-orbital features

“…60 The GFN family of methods 60,65,66 have proven to be extremely useful for the simulation of large molecular system (1000s of atoms or more) with time-tosolution for energies and forces on the order of seconds. However, this applicability can be limited by the accuracy of the semi-empirical method, 48,67 thus creating a natural opportunity for "delta-learning" the difference between the GFN1 and DFT energies on the basis of the GFN1 features. Specifically, we consider regression labels associated with the difference between high-level DFT and the GFN1-xTB total atomization energies,…”

Path-integral approximations to quantum dynamics

“…All that is required is that the initial condition be describable in terms of smooth imaginary-time loops. This has been exploited recently with the development of nonequilibrium RPMD [133] (following earlier explorations of non-equilibrium CMD [134,135]), which has had interesting recent applications to hydrogen-graphene scattering [136,137], gas-surface reactions [138], and to excited-state dynamics [139]. Closely related to this are the recent developments of microcanonical RPMD [140][141][142].…”