Automated discovery of a robust interatomic potential for aluminum
Los Alamos National Laboratory（洛斯阿拉莫斯国家实验室）
Justin S. Smith, Benjamin Nebgen, Nithin Mathew, Jie Chen, Nicholas Lubbers, Leonid Burakovsky, Sergei Tretiak, Hai Ah Nam, Timothy Germann, Saryu Fensin, Kipton Barros
Accuracy of molecular dynamics simulations depends crucially on the interatomic potential used to generate forces. The gold standard would be first-principles quantum mechanics (QM) calculations, but these become prohibitively expensive at large simulation scales. Machine learning (ML) based potentials aim for faithful emulation of QM at drastically reduced computational cost. The accuracy and robustness of an ML potential is primarily limited by the quality and diversity of the training dataset. Using the principles of active learning (AL), we present a highly automated approach to dataset construction. The strategy is to use the ML potential under development to sample new atomic configurations and, whenever a configuration is reached for which the ML uncertainty is sufficiently large, collect new QM data. Here, we seek to push the limits of automation, removing as much expert knowledge from the AL process as possible. All sampling is performed using MD simulations starting from an initially disordered configuration, and undergoing non-equilibrium dynamics as driven by time-varying applied temperatures. We demonstrate this approach by building an ML potential for aluminum (ANI-Al). After many AL iterations, ANI-Al teaches itself to predict properties like the radial distribution function in melt, liquid-solid coexistence curve, and crystal properties such as defect energies and barriers. To demonstrate transferability, we perform a 1.3M atom shock simulation, and show that ANI-Al predictions agree very well with DFT calculations on local atomic environments sampled from the nonequilibrium dynamics. Interestingly, the configurations appearing in shock appear to have been well sampled in the AL training dataset, in a way that we illustrate visually.
Abstract Interatomic potentials constitute the key component of large-scale atomistic simulations of materials. The recently proposed physically-informed neural network (PINN) method combines a high-dimensional regression implemented by an artificial neural network with a physics-based bond-order interatomic potential applicable to both metals and nonmetals. In this paper, we present a modified version of the PINN method that accelerates the potential training process and further improves the transferability of PINN potentials to unknown atomic environments. As an application, a modified PINN …
Choice of appropriate force field is one of the main concerns of any atomistic simulation that needs to be seriously considered in order to yield reliable results. Since, investigations on mechanical behavior of materials at micro/nanoscale has been becoming much more widespread, it is necessary to determine an adequate potential which accurately models the interaction of the atoms for desired applications. In this framework, reliability of multiple embedded atom method based interatomic potentials for predicting the elastic properties was investigated. Assessments were carried out for differe…
Automated Machine Learning Pipeline: Large Language Models-Assisted Automated Dataset Generation for Training Machine-Learned Interatomic Potentials
NYU, Department of Chemistry（纽约大学化学系）；Initiative for Computational Catalysis, Flatiron Institute（计算催化计划，Flatiron研究所）；NYU, Department of Physics（纽约大学物理系）；Courant Institute of Mathematical Sciences, NYU（数学科学学院，纽约大学）；NYU-ECNU Center for Computational Chemistry（纽约大学-复旦大学计算化学中心）；Simons Center for Computational Physical Chemistry, NYU（Simons计算物理化学中心，纽约大学）
Many rotational invariants for crystal structure representations have been used to describe the structure-property relationship by machine learning. The machine learning interatomic potential (MLIP) is one of the applications of rotational invariants, which provides the relationship between the energy and the crystal structure. Since the MLIP requires the highest accuracy among machine learning estimations of the structure-property relationship, the enumeration of rotational invariants is useful for constructing MLIPs with the desired accuracy. In this study, we introduce high-order linearly i…
dpti: An Automated Thermodynamic Integration Workflow for Phase Diagram Calculations with Machine Learning Interatomic Potentials
University of Alabama at Birmingham（阿拉巴马大学伯明翰分校）；Princeton University（普林斯顿大学）；University of Science and Technology of China（中国科学技术大学）；Suzhou Institute for Advanced Research, University of Science and Technology of China（中国科学技术大学苏州高等研究院）；Suzhou Big Data & AI Research and Engineering Center（苏州大数据与人工智能研究与工程中心）；AI for Science Institute（科学智能研究所）；DP Technology（深势科技）；Institute of Applied Physics and Computational Mathematics（应用物理与计算数学研究所）；College of Engineering, Peking University（北京大学工学院）
In this letter we propose a new methodology for crystal structure prediction, which is based on the evolutionary algorithm USPEX and the machine-learning interatomic potentials actively learning on-the-fly. Our methodology allows for an automated construction of an interatomic interaction model from scratch replacing the expensive DFT with a speedup of several orders of magnitude. Predicted low-energy structures are then tested on DFT, ensuring that our machine-learning model does not introduce any prediction error. We tested our methodology on a problem of prediction of carbon allotropes, den…
Defect-limited thermal transport in AlN using pretrained machine-learning interatomic potentials
Seoul National University（首尔大学）；Research Institute of Advanced Materials, Seoul National University（首尔大学先进材料研究所）；Center for AI and Natural Sciences, Korea Institute for Advanced Study（韩国高等研究院人工智能与自然科学中心）；Incheon National University（仁川国立大学）
Large scale atomistic simulations with suitable interatomic potentials are widely employed by scientists or engineers of different areas. Quick generation of high-quality interatomic potentials is of urgent need under present circumstances, which largely relies on the developments of potential construction methods and algorithms in this area. Quantities of interatomic potential models have been proposed and parameterized with various methods, such as analytic method, force-matching approach and multi-object optimization method, in order to make the potentials more transferable. Without apparen…
PySlice: Routine Vibrational Electron Energy Loss Spectroscopy Prediction with Universal Interatomic Potentials
Vanderbilt University（范德堡大学）；Oak Ridge National Laboratory（橡树岭国家实验室）；University of Virginia（弗吉尼亚大学）；University of Washington（华盛顿大学）；Uppsala University（乌普萨拉大学）
We devise automated workflows for the calculation of Helmholtz and Gibbs free energies and their temperature and pressure dependence and provide the corresponding computational tools. We employ non-equilibrium thermodynamics for evaluating the free energy of solid and liquid phases at a given temperature and reversible scaling for computing free energies over a wide range of temperatures, including the direct integration of $P$-$T$ coexistence lines. By changing the chemistry and the interatomic potential, alchemical and upscaling free energy calculations are possible. Several examples illustr…
Modeling non-empirical and highly flexible interatomic potential energy surfaces (PES) using machine learning (ML) approaches is becoming popular in molecular and materials research. Training an ML-PES is typically performed in two stages: feature extraction and structure-property relationship modeling. The feature extraction stage transforms atomic positions into a symmetry-invariant mathematical representation. This representation can be fine-tuned by adjusting on a set of so-called "hyper-parameters" (HPs). Subsequently, an ML algorithm such as neural networks or Gaussian process regression…
We present a systematic methodology, built within the Open Knowledgebase of Interatomic Models (OpenKIM) framework (https://openkim.org), for quantifying properties of grain boundaries (GBs) for arbitrary interatomic potentials (IPs), GB character, and lattice structure and species. The framework currently generates results for symmetric tilt GBs in cubic materials, but can be readily extended to other types of boundaries. In this paper, GB energy data are presented that were generated automatically for Al, Ni, Cu, Fe, and Mo with 225 IPs; the system is installed on openkim.org and will contin…
Cutting Through the Noise: On-the-fly Outlier Detection for Robust Training of Machine Learning Interatomic Potentials
Cavendish Laboratory, Department of Physics, University of Cambridge（卡文迪许实验室，物理系，剑桥大学）；Lennard-Jones Centre, University of Cambridge（兰纳德-琼斯中心，剑桥大学）；Yusuf Hamied Department of Chemistry, University of Cambridge（尤苏夫·哈米德化学系，剑桥大学）
It has been a challenge to accurately simulate Li-ion diffusion processes in battery materials at room temperature using {\it ab initio} molecular dynamics (AIMD) due to its high computational cost. This situation has changed drastically in recent years due to the advances in machine learning-based interatomic potentials. Here we implement the Deep Potential Generator scheme to \textit{automatically} generate interatomic potentials for LiGePS-type solid-state electrolyte materials. This increases our ability to simulate such materials by several orders of magnitude without sacrificing {\it ab …
We present a stochastic modeling framework for atomistic propagation of a Mode I surface crack, with atoms interacting according to the Lennard-Jones interatomic potential at zero temperature. Specifically, we invoke the Cauchy-Born rule and the maximum entropy principle to infer probability distributions for the parameters of the interatomic potential. We then study how uncertainties in the parameters propagate to the quantities of interest relevant to crack propagation, namely, the critical stress intensity factor and the lattice trapping range. For our numerical investigation, we rely on an…
VibroML: an automated toolkit for high-throughput vibrational analysis and dynamic instability remediation of crystalline materials using machine-learned potentials
The allotropes of boron continue to challenge structural elucidation and solid-state theory. Here we use machine learning combined with random structure searching (RSS) algorithms to systematically construct an interatomic potential for boron. Starting from ensembles of randomized atomic configurations, we use alternating single-point quantum-mechanical energy and force computations, Gaussian approximation potential (GAP) fitting, and GAP-driven RSS to iteratively generate a representation of the element's potential-energy surface. Beyond the total energies of the very different boron allotrop…
The atomic cluster expansion (ACE) was proposed recently as a new class of data-driven interatomic potentials with a formally complete basis set. Since the development of any interatomic potential requires a careful selection of training data and thorough validation, an automation of the construction of the training dataset as well as an indication of a model's uncertainty are highly desirable. In this work, we compare the performance of two approaches for uncertainty indication of ACE models based on the D-optimality criterion and ensemble learning. While both approaches show comparable predi…
Self-Optimizing Machine Learning Potential Assisted Automated Workflow for Highly Efficient Complex Systems Material Design
Key Laboratory of Material Simulation Methods \& Software of Ministry of Education, College of Physics, Jilin University, Changchun 130012, China；Research Center for Crystal Materials, State Key Laboratory of Functional Materials；Devices for Special Environmental Conditions, Xinjiang Key Laboratory of Functional Crystal Materials, Xinjiang Technical Institute of Physics；Chemistry, Chinese Academy of Sciences, 40-1 South Beijing Road, Urumqi 830011, China；State Key Laboratory of Superhard Materials, College of Physics, Jilin University, Changchun 130012, China；Department of Chemistry, University College London, Gordon Street, London, WC1H 0AJ, UK；International Center of Future Science, Jilin University, Changchun 130012, China；Key Laboratory of Physics；Technology for Advanced Batteries of Ministry of Education, College of Physics, Jilin University, Changchun, 130012, China
John A. Paulson School of Engineering and Applied Sciences, Harvard University（哈佛大学约翰·A·保尔森工程与应用科学学院）；Robert Bosch LLC, Research and Technology Center（罗伯特·博世有限责任公司研究与技术中心）
We present a way to dramatically accelerate Gaussian process models for interatomic force fields based on many-body kernels by mapping both forces and uncertainties onto functions of low-dimensional features. This allows for automated active learning of models combining near-quantum accuracy, built-in uncertainty, and constant cost of evaluation that is comparable to classical analytical models, capable of simulating millions of atoms. Using this approach, we perform large scale molecular dynamics simulations of the stability of the stanene monolayer. We discover an unusual phase transformatio…
Interatomic potential models based on machine learning (ML) are rapidly developing as tools for materials simulations. However, because of their flexibility, they require large fitting databases that are normally created with substantial manual selection and tuning of reference configurations. Here, we show that ML potentials can be built in a largely automated fashion, exploring and fitting potential-energy surfaces from the beginning (de novo) within one and the same protocol. The key enabling step is the use of a configuration-averaged kernel metric that allows one to select the few most re…
NepTrain and NepTrainKit: Automated Active Learning and Visualization Toolkit for Neuroevolution Potentials
College of physics and electronic information engineering, Guilin University of Technology, Guilin, 541008, P. R. China（桂林理工大学物理与电子信息工程学院）；School of Interdisciplinary Science, Beijing Institute of Technology, Beijing 100081, P. R. China（北京理工大学 interdisciplinary science 学院）；School of Mechanical and Electrical Engineering, Xinyu University, Xinyu, 338004, P. R. China（新余大学机械与电子工程学院）；School of Science, Harbin Institute of Technology (Shenzhen), Shenzhen 518055, China（哈尔滨工业大学（深圳）科学学院）；College of Physical Science and Technology, Bohai University, Jinzhou 121013, P. R. China（渤海大学物理科学与技术学院）
Hydrogen uptake and hydride formation in Al$_x$CoCrFeNi high-entropy alloys: First-principles, universal-potential, and experimental study
Ruhr-Universität Bochum（波鸿鲁尔大学）；Max-Planck-Institut for Sustainable Materials（马克斯·普朗克可持续材料研究所）；University of Stuttgart（斯图加特大学）；Deutsches Elektronen-Synchrotron DESY（德国电子同步辐射中心）；Goethe University Frankfurt（法兰克福歌德大学）；Federal Institute for Materials Research and Testing (BAM)（联邦材料研究与测试研究所）
Based on ab initio molecular dynamics simulations, we determined the melting curve of magnesium (Mg) up to ~460 GPa using the solid-liquid coexistence method. Between ~30 and 100 GPa, our melting curve is noticeably lower than those from static experiments, but is in good agreement with recent shock experiments. Up to ~450 GPa, our melting curve is generally consistent with the melting points from first-principles calculations using the small-cell coexistence method. We found that, at high pressures of a few hundred GPa, due to the strong softening of interatomic interactions in the liquid pha…
Rapid and Automated Alloy Design with Graph Neural Network-Powered LLM-Driven Multi-Agent Systems
Massachusetts Institute of Technology（麻省理工学院）；Schwarzman College of Computing（计算学院）；Center for Computational Science and Engineering（计算科学与工程中心）；Laboratory for Atomistic and Molecular Mechanics (LAMM)（原子与分子力学实验室）
