npj Computational Materials最新文献_第6页

SLM-MATRIX: a multi-agent trajectory reasoning and verification framework for enhancing language models in materials data extraction SLM-MATRIX：一个多智能体轨迹推理和验证框架，用于增强材料数据提取中的语言模型

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-07-19 DOI: 10.1038/s41524-025-01719-x

Xin Li, Zhixuan Huang, Shu Quan, Cheng Peng, Xiaoming Ma

{"title":"SLM-MATRIX: a multi-agent trajectory reasoning and verification framework for enhancing language models in materials data extraction","authors":"Xin Li, Zhixuan Huang, Shu Quan, Cheng Peng, Xiaoming Ma","doi":"10.1038/s41524-025-01719-x","DOIUrl":"https://doi.org/10.1038/s41524-025-01719-x","url":null,"abstract":"Small Language Models offer an efficient alternative for structured information extraction. We present SLM-MATRIX, a multi-path collaborative reasoning and verification framework based on SLMs, designed to extract material names, numerical values, and physical units from materials science literature. The framework integrates three complementary reasoning paths: a multi-agent collaborative path, a generator–discriminator path, and a dual cross-verification path. SLM-MATRIX achieves an accuracy of 92.85% on the BulkModulus dataset and reaches 77.68% accuracy on the MatSynTriplet dataset, both outperforming conventional methods and single-path models. Moreover, experiments on general reasoning benchmarks such as GSM8K and SVAMP validate the framework’s strong generalization capability. Ablation studies evaluate the effects of agent number, Mixture-of-Agents (MoA) depth, and discriminator design on overall performance. Overall, SLM-MATRIX presents an effective approach for high-quality material information extraction in resource-constrained and offers new insights into structured scientific text understanding tasks.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"10 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-07-19","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144664621","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Fine-tuning foundation models of materials interatomic potentials with frozen transfer learning 基于冻结迁移学习的材料原子间势的微调基础模型

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-07-18 DOI: 10.1038/s41524-025-01727-x

Mariia Radova, Wojciech G. Stark, Connor S. Allen, Reinhard J. Maurer, Albert P. Bartók

{"title":"Fine-tuning foundation models of materials interatomic potentials with frozen transfer learning","authors":"Mariia Radova, Wojciech G. Stark, Connor S. Allen, Reinhard J. Maurer, Albert P. Bartók","doi":"10.1038/s41524-025-01727-x","DOIUrl":"https://doi.org/10.1038/s41524-025-01727-x","url":null,"abstract":"Machine-learned interatomic potentials are revolutionising atomistic materials simulations by providing accurate and scalable predictions within the scope covered by the training data. However, generation of an accurate and robust training data set remains a challenge, often requiring thousands of first-principles calculations to achieve high accuracy. Foundation models have started to emerge with the ambition to create universally applicable potentials across a wide range of materials. While foundation models can be robust and transferable, they do not yet achieve the accuracy required to predict reaction barriers, phase transitions, and material stability. This work demonstrates that foundation model potentials can reach chemical accuracy when fine-tuned using transfer learning with partially frozen weights and biases. For two challenging datasets on reactive chemistry at surfaces and stability and elastic properties of tertiary alloys, we show that frozen transfer learning with 10–20% of the data (hundreds of datapoints) achieves similar accuracies to models trained from scratch (on thousands of datapoints). Moreover, we show that an equally accurate, but significantly more efficient surrogate model can be built using the transfer learned potential as the ground truth. In combination, we present a simulation workflow for machine learning potentials that improves data efficiency and computational efficiency.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"677 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-07-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144652617","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

5d orbital induced room temperature quantum anomalous Hall effect in TbCl TbCl中5d轨道诱导的室温量子反常霍尔效应

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-07-17 DOI: 10.1038/s41524-025-01732-0

Jianqi Zhong, Jianzhou Zhao, Jinyu Zou, Gang Xu

引用次数: 0

Cross-modality material embedding loss for transferring knowledge between heterogeneous material descriptors 跨模态材料嵌入损失在异质材料描述符之间传递知识

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-07-16 DOI: 10.1038/s41524-025-01723-1

Gyoung S. Na

{"title":"Cross-modality material embedding loss for transferring knowledge between heterogeneous material descriptors","authors":"Gyoung S. Na","doi":"10.1038/s41524-025-01723-1","DOIUrl":"https://doi.org/10.1038/s41524-025-01723-1","url":null,"abstract":"Despite the remarkable successes of transfer learning in materials science, the practicality of existing transfer learning methods are still limited in real-world applications of materials science because they essentially assume the same material descriptors on source and target materials datasets. In other words, existing transfer learning methods cannot utilize the knowledge extracted from calculated crystal structures when analyzing experimental observations of real-world chemical experiments. We propose a transfer learning criterion, called cross-modality material embedding loss (CroMEL), to build a source feature extractor that can transfer knowledge extracted from calculated crystal structures to prediction models in target applications where only simple chemical compositions are accessible. The prediction models based on transfer learning with CroMEL showed state-of-the-art prediction accuracy on 14 experimental materials datasets in various chemical applications. In particular, the prediction models with CroMEL achieved R2-scores greater than 0.95 in predicting the experimentally measured formation enthalpies and band gaps of the experimentally synthesized materials.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"7 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144645644","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Refining coarse-grained molecular topologies: a Bayesian optimization approach 细化粗粒度分子拓扑：贝叶斯优化方法

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-07-16 DOI: 10.1038/s41524-025-01729-9

Pranoy Ray, Adam P. Generale, Nikhith Vankireddy, Yuichiro Asoma, Masataka Nakauchi, Haein Lee, Katsuhisa Yoshida, Yoshishige Okuno, Surya R. Kalidindi

{"title":"Refining coarse-grained molecular topologies: a Bayesian optimization approach","authors":"Pranoy Ray, Adam P. Generale, Nikhith Vankireddy, Yuichiro Asoma, Masataka Nakauchi, Haein Lee, Katsuhisa Yoshida, Yoshishige Okuno, Surya R. Kalidindi","doi":"10.1038/s41524-025-01729-9","DOIUrl":"https://doi.org/10.1038/s41524-025-01729-9","url":null,"abstract":"Molecular Dynamics (MD) simulations are vital for predicting the physical and chemical properties of molecular systems across various ensembles. While All-Atom (AA) MD provides high accuracy, its computational cost has spurred the development of Coarse-Grained MD (CGMD), which simplifies molecular structures into representative beads to reduce expense but sacrifice precision. CGMD methods like Martini3, calibrated against experimental data, generalize well across molecular classes but often fail to meet the accuracy demands of domain-specific applications. This work introduces a Bayesian Optimization-based approach to refine Martini3 topologies—specifically the bonded interaction parameters within a given coarse-grained mapping—for specialized applications, ensuring accuracy and efficiency. The resulting optimized CG potential accommodates any degree of polymerization, offering accuracy comparable to AA simulations while retaining the computational speed of CGMD. By bridging the gap between efficiency and accuracy, this method advances multiscale molecular simulations, enabling cost-effective molecular discovery for diverse scientific and technological fields.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"10 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-07-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144645646","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Multiscale computational framework linking alloy composition to microstructure evolution via machine learning and nanoscale analysis 通过机器学习和纳米级分析将合金成分与微观结构演变联系起来的多尺度计算框架

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-07-15 DOI: 10.1038/s41524-025-01730-2

Jaemin Wang, Hyeonseok Kwon, Sang-Ho Oh, Jae Heung Lee, Dae Won Yun, Hyungsoo Lee, Seong-Moon Seo, Young-Soo Yoo, Hi Won Jeong, Hyoung Seop Kim, Byeong-Joo Lee

{"title":"Multiscale computational framework linking alloy composition to microstructure evolution via machine learning and nanoscale analysis","authors":"Jaemin Wang, Hyeonseok Kwon, Sang-Ho Oh, Jae Heung Lee, Dae Won Yun, Hyungsoo Lee, Seong-Moon Seo, Young-Soo Yoo, Hi Won Jeong, Hyoung Seop Kim, Byeong-Joo Lee","doi":"10.1038/s41524-025-01730-2","DOIUrl":"https://doi.org/10.1038/s41524-025-01730-2","url":null,"abstract":"Achieving targeted microstructures through composition design is a core challenge in developing structural materials for high-performance applications. This study introduces a multiscale Integrated Computational Materials Engineering (ICME) framework that combines CALPHAD-based thermodynamic modeling, machine learning, molecular dynamics, and diffusion kinetics to link alloy chemistry to microstructural evolution. Machine learning models trained on 750,000 CALPHAD-derived datapoints enabled rapid screening of two billion compositions based on thermodynamic criteria. An advanced screening step incorporated nanoscale physical descriptors that capture mechanisms governing precipitate coarsening and dynamic recrystallization. Applied to wrought Ni-based superalloys, the framework identified twelve compositions predicted to form fine intragranular γ′ precipitates within coarse γ grains; one was experimentally validated, with microscopy confirming the predicted microstructure. While demonstrated for Ni-based systems, the methodology is broadly generalizable. This work highlights the power of integrating high-throughput composition screening with atomistic-scale evaluation to accelerate microstructure-driven materials design beyond equilibrium thermodynamics.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"1 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144629912","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

A neural master equation framework for multiscale modeling of molecular processes: application to atomic-scale plasma processes 分子过程多尺度建模的神经主方程框架：在原子尺度等离子体过程中的应用

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-07-15 DOI: 10.1038/s41524-025-01677-4

Shoubhanik Nath, Joseph R. Vella, David B. Graves, Ali Mesbah

{"title":"A neural master equation framework for multiscale modeling of molecular processes: application to atomic-scale plasma processes","authors":"Shoubhanik Nath, Joseph R. Vella, David B. Graves, Ali Mesbah","doi":"10.1038/s41524-025-01677-4","DOIUrl":"https://doi.org/10.1038/s41524-025-01677-4","url":null,"abstract":"Plasma-surface interactions (PSI) play a crucial role in microelectronics fabrication; however, their multiscale nature and array of complex, often unknown interactions make computational modeling of PSIs extremely difficult. To this end, we propose a general neural master equation (NME) framework that uses master equations to describe the dynamics of a molecular process, wherein neural networks learned from atomistic simulations represent unknown transitions between different system states. By leveraging the physics-based structure of master equations and data-driven state transitions, the NME framework promotes generalizability and physics interpretability, and can bridge disparate length and time scales. The framework is demonstrated for multiscale modeling of Si atomic layer etching and reactive ion etching, where the learned NME-based surface kinetic models exhibit good predictive and extrapolative capabilities for predicting experimentally relevant observables as a function of process parameters. The NME-based surface kinetic models obey physical constraints, which are violated in models based on neural ordinary differential equations. The proposed NME framework for multiscale modeling of molecular processes can pave the way for the discovery of new chemistries and materials in atomic-scale plasma processes.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"670 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144630138","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Technical review: Time-dependent density functional theory for attosecond physics ranging from gas-phase to solids 技术评论：从气相到固体的阿秒物理的随时间密度泛函理论

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-07-15 DOI: 10.1038/s41524-025-01715-1

Shunsuke A. Sato, Hannes Hübener, Umberto De Giovannini, Angel Rubio

{"title":"Technical review: Time-dependent density functional theory for attosecond physics ranging from gas-phase to solids","authors":"Shunsuke A. Sato, Hannes Hübener, Umberto De Giovannini, Angel Rubio","doi":"10.1038/s41524-025-01715-1","DOIUrl":"https://doi.org/10.1038/s41524-025-01715-1","url":null,"abstract":"First-principles electron dynamics calculations can be applied in the investigation of a wide range of ultrafast phenomena in attosecond physics. They offer unique microscopic insight into light-induced ultrafast phenomena in both gas and condensed phases of matter, and thus, they are a powerful tool to develop our understanding of the physics of attosecond phenomena. We specifically review techniques employing time-dependent density functional theory (TDDFT) for investigating attosecond and strong-field phenomena. First, we describe this theoretical framework that enables the modeling of perturbative and non-perturbative electron dynamics in materials, including atoms, molecules, and solids. We then discuss its application to attosecond experiments, focusing on the reconstruction of attosecond beating by interference of two-photon transitions (RABBIT) measurements. We also briefly review first-principles calculations of optical properties of solids with TDDFT in the linear response regime and their extension to calculations of transient optical properties of solids in non-equilibrium phases, by simulating experimental pump-probe setups. We further demonstrate the application of TDDFT simulation to high-order harmonic generation in solids. First-principles calculations have predictive power, and hence they can be utilized to design future experiments to explore non-equilibrium and nonlinear ultrafast phenomena in matter and characterize and control metastable light-induced quantum states.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"22 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144640410","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Modeling crystal defects using defect informed neural networks 基于缺陷信息神经网络的晶体缺陷建模

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-07-15 DOI: 10.1038/s41524-025-01728-w

Ziduo Yang, Xiaoqing Liu, Xiuying Zhang, Pengru Huang, Kostya S. Novoselov, Lei Shen

{"title":"Modeling crystal defects using defect informed neural networks","authors":"Ziduo Yang, Xiaoqing Liu, Xiuying Zhang, Pengru Huang, Kostya S. Novoselov, Lei Shen","doi":"10.1038/s41524-025-01728-w","DOIUrl":"https://doi.org/10.1038/s41524-025-01728-w","url":null,"abstract":"Most AI-for-Materials research to date has focused on ideal crystals, whereas real-world materials inevitably contain defects that play a critical role in modern functional technologies. The defects break geometric symmetry and increase interaction complexity, posing particular challenges for traditional ML models. Here, we introduce Defect-Informed Equivariant Graph Neural Network (DefiNet), a model specifically designed to accurately capture defect-related interactions and geometric configurations in point-defect structures. DefiNet achieves near-DFT-level structural predictions in milliseconds using a single GPU. To validate its accuracy, we perform DFT relaxations using DefiNet-predicted structures as initial configurations and measure the residual ionic steps. For most defect structures, regardless of defect complexity or system size, only 3 ionic steps are required to reach the DFT-level ground state. Finally, comparisons with scanning transmission electron microscopy (STEM) images confirm DefiNet’s scalability and extrapolation beyond point defects, positioning it as a valuable tool for defect-focused materials research.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"2 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144630276","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Stochastic 3D reconstruction of cracked polycrystalline NMC particles using 2D SEM data 利用二维扫描电镜数据对裂纹多晶NMC颗粒进行随机三维重建

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-07-15 DOI: 10.1038/s41524-025-01695-2

Philipp Rieder, Orkun Furat, Francois L. E. Usseglio-Viretta, Jeffery Allen, Peter J. Weddle, Donal P. Finegan, Kandler Smith, Volker Schmidt

{"title":"Stochastic 3D reconstruction of cracked polycrystalline NMC particles using 2D SEM data","authors":"Philipp Rieder, Orkun Furat, Francois L. E. Usseglio-Viretta, Jeffery Allen, Peter J. Weddle, Donal P. Finegan, Kandler Smith, Volker Schmidt","doi":"10.1038/s41524-025-01695-2","DOIUrl":"https://doi.org/10.1038/s41524-025-01695-2","url":null,"abstract":"Li-ion battery performance is strongly influenced by the 3D microstructure of its cathode particles. Cracks within these particles develop during calendaring and cycling, reducing connectivity but increasing reactive surface, making their impact on battery performance complex. Understanding these contradictory effects requires a quantitative link between particle morphology and battery performance. However, informative 3D imaging techniques are time-consuming, costly and rarely available, such that analyses often have to rely on 2D image data. This paper presents a novel stereological approach for generating virtual 3D cathode particles exhibiting crack networks that are statistically equivalent to those observed in 2D sections of experimentally measured particles. Consequently, 2D image data suffices for deriving a full 3D characterization of cracked cathodes particles. Such virtually generated 3D particles could serve as geometry input for spatially resolved electro-chemo-mechanical simulations to enhance our understanding of structure-property relationships of cathodes in Li-ion batteries.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"23 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-07-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144629995","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":1,"RegionCategory":"材料科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0