Alexey S. Kotykhov, Konstantin Gubaev, Vadim Sotskov, Christian Tantardini, Max Hodapp, Alexander V. Shapeev, Ivan S. Novikov
{"title":"与磁力拟合提高了磁矩张势的可靠性","authors":"Alexey S. Kotykhov, Konstantin Gubaev, Vadim Sotskov, Christian Tantardini, Max Hodapp, Alexander V. Shapeev, Ivan S. Novikov","doi":"arxiv-2405.07069","DOIUrl":null,"url":null,"abstract":"We propose a novel method for fitting machine-learning interatomic potentials\nwith magnetic degrees of freedom, namely, magnetic Moment Tensor Potentials\n(mMTP). The main feature of the methodology consists in fitting mMTP to\nmagnetic forces (negative derivatives of energies with respect to magnetic\nmoments) derived from spin-polarized density functional theory calculations. We\ntest our method on the bcc Fe-Al system with different composition.\nSpecifically, we calculate formation energies, equilibrium lattice parameter,\nand total cell magnetization. Our findings demonstrate a precise match between\nvalues calculated with mMTP and those obtained by DFT at zero temperature.\nAdditionally, using molecular dynamics, we estimate the finite temperature\nlattice parameter and capture the cell expansion as was previously revealed in\nexperiment. We demonstrate that mMTPs fitted to magnetic forces, increase the\nrelaxation reliability, which is the percent of successfully relaxed structures\n(i.e. with almost zero forces, stresses, and magnetic moments after the\noptimization of geometry). Eventually, we show that the proposed methodology\ncan provide an accurate and reliable mMTP with reduced number of\ncomputationally complex spin-polarized density functional theory calculations.","PeriodicalId":501211,"journal":{"name":"arXiv - PHYS - Other Condensed Matter","volume":"18 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-05-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Fitting to magnetic forces improves the reliability of magnetic Moment Tensor Potentials\",\"authors\":\"Alexey S. Kotykhov, Konstantin Gubaev, Vadim Sotskov, Christian Tantardini, Max Hodapp, Alexander V. Shapeev, Ivan S. Novikov\",\"doi\":\"arxiv-2405.07069\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"We propose a novel method for fitting machine-learning interatomic potentials\\nwith magnetic degrees of freedom, namely, magnetic Moment Tensor Potentials\\n(mMTP). The main feature of the methodology consists in fitting mMTP to\\nmagnetic forces (negative derivatives of energies with respect to magnetic\\nmoments) derived from spin-polarized density functional theory calculations. We\\ntest our method on the bcc Fe-Al system with different composition.\\nSpecifically, we calculate formation energies, equilibrium lattice parameter,\\nand total cell magnetization. Our findings demonstrate a precise match between\\nvalues calculated with mMTP and those obtained by DFT at zero temperature.\\nAdditionally, using molecular dynamics, we estimate the finite temperature\\nlattice parameter and capture the cell expansion as was previously revealed in\\nexperiment. We demonstrate that mMTPs fitted to magnetic forces, increase the\\nrelaxation reliability, which is the percent of successfully relaxed structures\\n(i.e. with almost zero forces, stresses, and magnetic moments after the\\noptimization of geometry). Eventually, we show that the proposed methodology\\ncan provide an accurate and reliable mMTP with reduced number of\\ncomputationally complex spin-polarized density functional theory calculations.\",\"PeriodicalId\":501211,\"journal\":{\"name\":\"arXiv - PHYS - Other Condensed Matter\",\"volume\":\"18 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-05-11\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - PHYS - Other Condensed Matter\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2405.07069\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - PHYS - Other Condensed Matter","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2405.07069","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Fitting to magnetic forces improves the reliability of magnetic Moment Tensor Potentials
We propose a novel method for fitting machine-learning interatomic potentials
with magnetic degrees of freedom, namely, magnetic Moment Tensor Potentials
(mMTP). The main feature of the methodology consists in fitting mMTP to
magnetic forces (negative derivatives of energies with respect to magnetic
moments) derived from spin-polarized density functional theory calculations. We
test our method on the bcc Fe-Al system with different composition.
Specifically, we calculate formation energies, equilibrium lattice parameter,
and total cell magnetization. Our findings demonstrate a precise match between
values calculated with mMTP and those obtained by DFT at zero temperature.
Additionally, using molecular dynamics, we estimate the finite temperature
lattice parameter and capture the cell expansion as was previously revealed in
experiment. We demonstrate that mMTPs fitted to magnetic forces, increase the
relaxation reliability, which is the percent of successfully relaxed structures
(i.e. with almost zero forces, stresses, and magnetic moments after the
optimization of geometry). Eventually, we show that the proposed methodology
can provide an accurate and reliable mMTP with reduced number of
computationally complex spin-polarized density functional theory calculations.