Computational framework for efficient navigation of atomic configurational space using physically guided actions and transformations

Computational methods like Monte Carlo (MC) or molecular dynamics simulations paired with empirical or machine learning potentials have been shown to be good at sampling the atomic configurational space of complex materials. However, it has also been shown that it takes on the order of ~10⁷ steps to estimate thermodynamic observables using MC simulations [1]. We present a framework that allows us to determine what changes of atomic configurations have the most effect on materials properties to help us identify compositions that result in desired properties, in a computationally efficient manner. The workflow is as follows: (i) transform atomic configurations by performing actions (atomic swaps) based on Shannon entropy and local electronegative derived from n-grams, (ii) keep track of short-range chemical and structural order [2] and global entropy [3] after every transformation for physical consistency, (iii) predict formation energy of transformed configurations using a Kernel Ridge Regression model trained on n-grams histograms. Using our framework, we find a minimum formation energy configuration for the MoNbTaW high-entropy alloy that exhibits spinodal decomposition consistent with experimental results. This configuration is reached in three orders of magnitude fewer steps and with less computational cost compared to other methods. Our framework provides an alternative, more directed method to sample configuration space and expedite the design and discovery of new materials.

 

References 

1. Liu, X., Zhang, J., Yin, J., Bi, S., Eisenbach, M., & Wang, Y. (2021). Monte Carlo simulation of order-disorder transition in refractory high entropy alloys: A data-driven approach. Computational Materials Science, 187, 110135. 

1. Cowley, J. M. (1950). An approximate theory of order in alloys. Physical Review, 77(5), 669. 

1. Rao, Y., & Curtin, W. A. (2022). Analytical models of short-range order in FCC and BCC alloys. Acta Materialia, 226, 117621.