The central approximation made in classical molecular dynamics simulation of materials is the interatomic potential used to calculate the forces on the atoms. Great effort and ingenuity is required to construct viable functional forms and find accurate parameterizations for potentials using traditional approaches. Machinelearning has emerged as an effective alternative approach to develop accurate and robust interatomic potentials. Starting with a very general model form, the potential is learned directly from a database of electronic structure calculations and therefore can be viewed as a multiscale link between quantum and classical atomistic simulations. Risk of inaccurate extrapolation exists outside the narrow range of time-and length-scales where the two methods can be directly compared. In this work, we use the Spectral Neighbor Analysis Potential (SNAP) and show how a fit can be produced with minimal interpolation errors which is also robust in extrapolating beyond training. To demonstrate the method, we have developed a new tungsten-beryllium potential suitable for the full range of binary compositions. Subsequently, large-scale molecular dynamics simulations were performed of high energy Be atom implantation onto the (001) surface of solid tungsten. The new machine learned W-Be potential generates a population of implantation structures consistent with quantum calculations of defect formation energies. A very shallow (< 2nm) average Be implantation depth is predicted which may explain ITER diverter degradation in the presence of beryllium.

A natural extension of the descriptors used in the Spectral Neighbor Analysis Potential (SNAP) method is derived to treat atomic interactions in chemically complex systems. Atomic environment descriptors within SNAP are obtained from a basis function expansion of the weighted density of neighboring atoms. This new formulation instead partitions the neighbor density into partial densities for each chemical element, thus leading to explicit multielement descriptors. For N elem chemical elements, the number of descriptors increases as N ( ) elem 3 , while the computational cost of the force calculation as implemented in LAMMPS is limited to N ( ) elem 2 and the favorable linear scaling in the number of atoms is retained. We demonstrate these chemically aware descriptors by producing an interatomic potential for indium phosphide capable of capturing high-energy defects that result from radiation damage cascades. This new explicit multielement SNAP method reproduces the relaxed defect formation energies with substantially greater accuracy than weighted-density SNAP, while retaining accurate representation of the bulk indium phosphide properties.

No abstract

Article history:Available online xxxx a b s t r a c tWe compare the hydrogen and helium clustering characteristics of three interatomic potential energy models intended for simulation of plasma-facing materials for fusion applications. Our simulations compare a Finnis-Sinclair potential and two different Tersoff-style bond order potentials created by Juslin et al. (2005) and Li et al. (2011), respectively, with respect to both helium and hydrogen clustering behavior in tungsten. We find significant differences between the Juslin and Li potentials in terms of both hydrogen and helium clustering behavior as well as the spatial distribution of hydrogen below the surface. These simulations are an important test on the road to more accurate models of gas clustering and surface evolution of tungsten divertors in ITER and other plasma devices.