Do Computed Grain Boundary Energies Depend on Interatomic Potentials? Universal Trends Employing EAM-X

It is well established that grain boundaries (GBs) greatly influence the observable properties of a wide range of engineering and functional materials. Classical atomistic simulations employing the Embedded Atom Method (EAM) have emerged as a powerful technique to simulate GB phenomena. However, folded into such simulations is the dependency of GB structure and properties on the particular choice of EAM parameters. To address this question, we follow a direction of investigation that has not been generally explored, namely we simplify the EAM function space to a small but efficient set of parameters, called EAM-X [1]. Then, we study a set of GBs with various geometries and calculate their energies in the complete EAM-X parameter space. We find that variations in GB energy with EAM parameters can be larger than variations due to GB geometry; an effect that has not been quantified before. The atomistic data are used to determine a fit of the GB energy in EAM parameter space, which can be used to obtain boundary energies in real FCC elements by selecting corresponding points in this parameter space. We find generally at best a moderate correlation between GB energy and shear modulii, and we discuss the relationship to prior work along these lines. Our work highlights the need to consider sensitivity to details of empirical potentials when performing quantitative studies of GB physics. <i>SNL is managed and operated by NTESS under DOE NNSA contract DE-NA0003525 (SAND2022-1056 A).</i><br/><br/>[1] M. S. Daw, and M. Chandross. "Simple parameterization of embedded atom method potentials for FCC metals." Acta Materialia 248 (2023): 118771.