Abstract
We introduce machine-learned potentials for Ag-Pd to describe the energy of alloy configurations over a wide range of compositions. We compare two different approaches. Moment tensor potentials (MTPs) are polynomial-like functions of interatomic distances and angles. The Gaussian approximation potential (GAP) framework uses kernel regression, and we use the smooth overlap of atomic position (SOAP) representation of atomic neighborhoods that consist of a complete set of rotational and permutational invariants provided by the power spectrum of the spherical Fourier transform of the neighbor density. Both types of potentials give excellent accuracy for a wide range of compositions, competitive with the accuracy of cluster expansion, a benchmark for this system. While both models are able to describe small deformations away from the lattice positions, SOAP-GAP excels at transferability as shown by sensible transformation paths between configurations, and MTP allows, due to its lower computational cost, the calculation of compositional phase diagrams. Given the fact that both methods perform nearly as well as cluster expansion but yield off-lattice models, we expect them to open new avenues in computational materials modeling for alloys.
Highlights
The technology frontier relies on the exceptional performance of next-generation materials
Conventional interatomic potentials (IPs), such as Lennard–Jones, embedded atom method (EAM), modified EAM, Tersoff, Stillinger–Weber, and so on, typically provide six to eight orders of magnitude speed-up compared to density functional theory (DFT) calculations, and due to their simple, physically motivated forms, they are somewhat robust in the sense that their predictions for low energy structures are plausible
Due to the efficiency of its polynomial basis of interatomic distances and angles, Moment tensor potentials (MTPs) is significantly faster than Gaussian approximation potential (GAP) and has already been shown to be capable of reaching equivalent accuracy for modeling chemical reactions[34], single-element systems[35,36], single-phase binary systems[37], or ground states of multicomponent systems[25]. We demonstrate that both GAP and MTP are capable of fitting the potential energy function of a binary metallic system, the Ag-Pd alloy system, with DFT accuracy across the full space of configuration and composition for solid and liquid systems
Summary
The technology frontier relies on the exceptional performance of next-generation materials. Conventional interatomic potentials (IPs), such as Lennard–Jones, embedded atom method (EAM), modified EAM, Tersoff, Stillinger–Weber, and so on, typically provide six to eight orders of magnitude speed-up compared to DFT calculations, and due to their simple, physically motivated forms, they are somewhat robust in the sense that their predictions for low energy structures are plausible. Their quantitative accuracy is typically quite poor compared to DFT, especially in reproducing macroscopic properties. This transferability problem requires researchers to take care of constructing, applying, and validating IPs, and in particular makes it a rather tenuous proposition to use them to discover and predict new structures and novel properties
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a callDisclaimer: All third-party content on this website/platform is and will remain the property of their respective owners and is provided on "as is" basis without any warranties, express or implied. Use of third-party content does not indicate any affiliation, sponsorship with or endorsement by them. Any references to third-party content is to identify the corresponding services and shall be considered fair use under The CopyrightLaw.
