Rina Ibragimova, Mikhail S. Kuklin, Tigany Zarrouk, Miguel A. Caro
{"title":"Unifying the description of hydrocarbons and hydrogenated carbon materials with a chemically reactive machine learning interatomic potential","authors":"Rina Ibragimova, Mikhail S. Kuklin, Tigany Zarrouk, Miguel A. Caro","doi":"arxiv-2409.08194","DOIUrl":null,"url":null,"abstract":"We present a general-purpose machine learning (ML) interatomic potential for\ncarbon and hydrogen which is capable of simulating various materials and\nmolecules composed of these elements. This ML interatomic potential is trained\nusing the Gaussian approximation potential (GAP) framework and an extensive\ndataset of C-H configurations obtained from density functional theory. The\ndataset is constructed through iterative training and structure-search\ntechniques that generate a broad range of configurations to comprehensively\nsample the potential energy surface. Furthermore, the dataset is supplemented\nwith relevant bulk, molecular, and high-pressure structures. Finally,\nlong-range van der Waals interactions are added as a locally parametrized\nmodel. The accuracy and generality of the potential are validated through the\nanalysis of different simulations under a wide range of conditions, including\nweak interactions, high temperature, and high pressure. We show that our CH GAP\nmodel describes different problems such as the formation of simple and complex\nalkanes, aromatic hydrocarbons, hydrogenated amorphous carbon (a-C:H), and CH\nsystems at extreme conditions, while retaining good accuracy for pure carbon\nmaterials. We use this model to generate hydrocarbons of different sizes and\ncomplexity without prior knowledge of organic chemistry rules, and to highlight\nintrinsic limitations to the simultaneous description on intra and\nintermolecular interactions within a single computational framework. Our\ngeneral-purpose ML interatomic potential has the capability to significantly\nadvance research in the field of H-containing carbon materials and compounds,\nparticularly in the areas where longer dynamics, reactivity and large-scale\neffects may be important.","PeriodicalId":501304,"journal":{"name":"arXiv - PHYS - Chemical Physics","volume":null,"pages":null},"PeriodicalIF":0.0000,"publicationDate":"2024-09-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - PHYS - Chemical Physics","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2409.08194","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
We present a general-purpose machine learning (ML) interatomic potential for
carbon and hydrogen which is capable of simulating various materials and
molecules composed of these elements. This ML interatomic potential is trained
using the Gaussian approximation potential (GAP) framework and an extensive
dataset of C-H configurations obtained from density functional theory. The
dataset is constructed through iterative training and structure-search
techniques that generate a broad range of configurations to comprehensively
sample the potential energy surface. Furthermore, the dataset is supplemented
with relevant bulk, molecular, and high-pressure structures. Finally,
long-range van der Waals interactions are added as a locally parametrized
model. The accuracy and generality of the potential are validated through the
analysis of different simulations under a wide range of conditions, including
weak interactions, high temperature, and high pressure. We show that our CH GAP
model describes different problems such as the formation of simple and complex
alkanes, aromatic hydrocarbons, hydrogenated amorphous carbon (a-C:H), and CH
systems at extreme conditions, while retaining good accuracy for pure carbon
materials. We use this model to generate hydrocarbons of different sizes and
complexity without prior knowledge of organic chemistry rules, and to highlight
intrinsic limitations to the simultaneous description on intra and
intermolecular interactions within a single computational framework. Our
general-purpose ML interatomic potential has the capability to significantly
advance research in the field of H-containing carbon materials and compounds,
particularly in the areas where longer dynamics, reactivity and large-scale
effects may be important.
我们提出了一种通用的机器学习(ML)碳氢原子间势,它能够模拟由这些元素组成的各种材料和分子。这种 ML 原子间位势是利用高斯近似位势(GAP)框架和从密度泛函理论获得的 C-H 构型扩展数据集进行训练的。该数据集是通过迭代训练和结构搜索技术构建的,可生成广泛的构型,从而对势能面进行全面采样。此外,数据集还补充了相关的块体、分子和高压结构。最后,长程范德瓦耳斯相互作用被添加为局部参数化模型。通过分析各种条件下的模拟(包括弱相互作用、高温和高压),我们验证了该势垒的准确性和通用性。我们的研究表明,我们的 CH GAP 模型可以描述不同的问题,例如在极端条件下简单和复杂烷烃、芳香烃、氢化无定形碳(a-C:H)和 CH 系统的形成,同时对纯碳材料保持良好的准确性。我们利用该模型生成了不同大小和复杂程度的碳氢化合物,而无需事先了解有机化学规则,并强调了在单一计算框架内同时描述分子内和分子间相互作用的内在局限性。我们的通用 ML 原子间势能极大地推动了含氢碳材料和化合物领域的研究,尤其是在长动力学、反应性和大尺度效应可能很重要的领域。