npj Computational Materials最新文献_第3页

Machine learning and data-driven methods in computational surface and interface science 计算表面和界面科学中的机器学习和数据驱动方法

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-07-01 DOI: 10.1038/s41524-025-01691-6

Lukas Hörmann, Wojciech G. Stark, Reinhard J. Maurer

引用次数: 0

Optimization of Coulomb energies in gigantic configurational spaces of multi-element ionic crystals 多元素离子晶体巨大构型空间中库仑能的优化

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-07-01 DOI: 10.1038/s41524-025-01690-7

Konstantin Köster, Tobias Binninger, Payam Kaghazchi

引用次数: 0

Data-driven low-rank approximation for the electron-hole kernel and acceleration of time-dependent GW calculations 数据驱动的电子空穴核的低秩近似和时变GW计算的加速

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-07-01 DOI: 10.1038/s41524-025-01680-9

Bowen Hou, Jinyuan Wu, Victor Chang Lee, Jiaxuan Guo, Luna Y. Liu, Diana Y. Qiu

{"title":"Data-driven low-rank approximation for the electron-hole kernel and acceleration of time-dependent GW calculations","authors":"Bowen Hou, Jinyuan Wu, Victor Chang Lee, Jiaxuan Guo, Luna Y. Liu, Diana Y. Qiu","doi":"10.1038/s41524-025-01680-9","DOIUrl":"https://doi.org/10.1038/s41524-025-01680-9","url":null,"abstract":"Many-body interactions are essential for understanding non-linear optics and ultrafast spectroscopy of materials. Recent first principles approaches based on nonequilibrium Green’s function formalisms, such as the time-dependent adiabatic GW (TD-aGW) approach, can predict nonequilibrium dynamics of excited states including electron-hole interactions. However, the high-dimensionality of the electron-hole kernel poses significant computational challenges. Here, we develop a data-driven low-rank approximation for the electron-hole kernel, leveraging localized excitonic effects in the Hilbert space of crystalline systems to achieve significant data compression through singular value decomposition (SVD). We show that the subspace of non-zero singular values remains small even as the k-grid grows, ensuring computational tractability with extremely dense k-grids. This low-rank property enables at least 95% data compression and an order-of-magnitude speedup of TD-aGW calculations. Our approach avoids intensive training processes and eliminates time-accumulated errors, seen in previous approaches, providing a general framework for high-throughput, nonequilibrium simulation of light-driven dynamics in materials.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"27 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144533948","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

Pretrained E(3)-equivariant message-passing neural networks with multi-level representations for organic molecule spectra prediction 用于有机分子光谱预测的预训练E(3)等变消息传递神经网络

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-07-01 DOI: 10.1038/s41524-025-01698-z

Yuzhi Xu, Daqian Bian, Cheng-Wei Ju, Fanyu Zhao, Pujun Xie, Yuanqing Wang, Wei Hu, Zhenrong Sun, John Z. H. Zhang, Tong Zhu

{"title":"Pretrained E(3)-equivariant message-passing neural networks with multi-level representations for organic molecule spectra prediction","authors":"Yuzhi Xu, Daqian Bian, Cheng-Wei Ju, Fanyu Zhao, Pujun Xie, Yuanqing Wang, Wei Hu, Zhenrong Sun, John Z. H. Zhang, Tong Zhu","doi":"10.1038/s41524-025-01698-z","DOIUrl":"https://doi.org/10.1038/s41524-025-01698-z","url":null,"abstract":"Fast and accurate spectral prediction plays a crucial role in molecular design within fields such as pharmaceutical and materials science. Nevertheless, predicting molecular spectra typically requires quantum chemistry calculations, posing significant challenges for fast predictions and high-throughput screening. In this paper, we propose an equivariant, fast, and robust model, named EnviroDetaNet, which integrates molecular environment information. EnviroDetaNet employs an E(3)-equivariant message-passing neural network combining intrinsic atomic properties, spatial features, and environmental information, allowing it to comprehensively capture both local and global molecular information. Compared to state-of-the-art machine learning models, EnviroDetaNet excels in various predictive tasks and maintains high accuracy even with a 50% reduction in training data, demonstrating strong generalization capabilities. Ablation studies confirm that molecular environment information is crucial for improving model stability and accuracy. EnviroDetaNet also shows outstanding performance in spectral predictions for complex molecular systems, making it a powerful tool for accelerating molecular discovery.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"27 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144533950","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

Learning atomic forces from uncertainty-calibrated adversarial attacks 从不确定性校准的对抗性攻击中学习原子力

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-07-01 DOI: 10.1038/s41524-025-01703-5

Henrique Musseli Cezar, Tilmann Bodenstein, Henrik Andersen Sveinsson, Morten Ledum, Simen Reine, Sigbjørn Løland Bore

{"title":"Learning atomic forces from uncertainty-calibrated adversarial attacks","authors":"Henrique Musseli Cezar, Tilmann Bodenstein, Henrik Andersen Sveinsson, Morten Ledum, Simen Reine, Sigbjørn Løland Bore","doi":"10.1038/s41524-025-01703-5","DOIUrl":"https://doi.org/10.1038/s41524-025-01703-5","url":null,"abstract":"Adversarial approaches, which intentionally challenge machine learning models by generating difficult examples, are increasingly being adopted to improve machine learning interatomic potentials (MLIPs). While already providing great practical value, little is known about the actual prediction errors of MLIPs on adversarial structures and whether these errors can be controlled. We propose the Calibrated Adversarial Geometry Optimization (CAGO) algorithm to discover adversarial structures with user-assigned errors. Through uncertainty calibration, the estimated uncertainty of MLIPs is unified with real errors. By performing geometry optimization for calibrated uncertainty, we reach adversarial structures with the user-assigned target MLIP prediction error. Integrating with active learning pipelines, we benchmark CAGO, demonstrating stable MLIPs that systematically converge structural, dynamical, and thermodynamical properties for liquid water and water adsorption in a metal-organic framework within only hundreds of training structures, where previously many thousands were typically required.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"646 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144533955","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

Transition state structure detection with machine learningś 基于机器学习的过渡状态结构检测

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-07-01 DOI: 10.1038/s41524-025-01693-4

Yitao Si, Yiding Ma, Tao Yu, Yifan Wu, Yingzhe Liu, Weipeng Lai, Zhixiang Zhang, Jinwen Shi, Liejin Guo, Oleg V. Prezhdo, Maochang Liu

{"title":"Transition state structure detection with machine learningś","authors":"Yitao Si, Yiding Ma, Tao Yu, Yifan Wu, Yingzhe Liu, Weipeng Lai, Zhixiang Zhang, Jinwen Shi, Liejin Guo, Oleg V. Prezhdo, Maochang Liu","doi":"10.1038/s41524-025-01693-4","DOIUrl":"https://doi.org/10.1038/s41524-025-01693-4","url":null,"abstract":"Transition structure calculations via quantum chemistry methods have become a staple in modern chemical reaction research. Yet, success rates in optimizing transition structures rely heavily on rational initial guesses and expert supervision. We develop a machine learning approach that utilizes a bitmap representation of chemical structures to generate high-quality initial guesses for modeling transition states of chemical reactions. The core of the approach comprises a convolutional neural network methodology with a genetic algorithm. An extensive dataset derived from quantum chemistry computations is built, providing sufficient data on which the model can be trained, validated and tested. By applying the method to typical bi-molecular hydrogen abstraction reactions involving hydrofluorocarbons, hydrofluoroethers, and hydroxyl radicals—reactions critical in atmospheric fluoride degradation and global warming potential evaluation, yet extremely challenging to model, we achieve transition state optimizations with an impressive, verified success rate of 81.8% for hydrofluorocarbons and 80.9% for hydrofluoroethers. The reported work demonstrates the effectiveness of employing visual representation in chemical space exploration tasks and opens new avenues for the transition structure modeling.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"10 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144533956","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 machine learning approach to designing and understanding tough, degradable polyamides 一种设计和理解坚韧、可降解聚酰胺的机器学习方法

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-07-01 DOI: 10.1038/s41524-025-01696-1

Yoshifumi Amamoto, Chie Koganemaru, Ken Kojio, Atsushi Takahara, Sayoko Yamamoto, Kazuki Okazawa, Yuta Tsuji, Toshimitsu Aritake, Kei Terayama

{"title":"A machine learning approach to designing and understanding tough, degradable polyamides","authors":"Yoshifumi Amamoto, Chie Koganemaru, Ken Kojio, Atsushi Takahara, Sayoko Yamamoto, Kazuki Okazawa, Yuta Tsuji, Toshimitsu Aritake, Kei Terayama","doi":"10.1038/s41524-025-01696-1","DOIUrl":"https://doi.org/10.1038/s41524-025-01696-1","url":null,"abstract":"The development of environmentally friendly plastics has received renewed attention for a sustainable society. Although the trade-off between toughness and degradability is a common challenge in biodegradable polymers, the design of biodegradable polymers to overcome these issues is often difficult. In this study, we demonstrated that machine learning techniques can contribute to the development of multiblock polyamides composed of Nylon6 and α-amino acid segments that are mechanically tough and degradable. Multi-objective optimization based on Gaussian process regression for the degradation rate, strain at break, and Young’s modulus (the last two parameters correspond to toughness) suggested appropriate α-amino acid sequences for polyamides endowed with both properties. Ridge regression revealed that the physical factors associated with the sequences, as well as the higher-order multiblock-derived structures (such as the crystal lattice structure, melting points, and hydrogen bonding), were essential for endowing these polymers with satisfactory properties among the multimodal measurement/calculation data. Our method provides a useful approach for designing and understanding environment-friendly plastics and other materials with multiple properties based on machine learning techniques.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"19 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144533951","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

Addressing data scarcity in nanomaterial segmentation networks with differentiable rendering and generative modeling 用可微分渲染和生成建模解决纳米材料分割网络中的数据稀缺性

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-07-01 DOI: 10.1038/s41524-025-01702-6

Dennis Possart, Leonid Mill, Florian Vollnhals, Tor Hildebrand, Peter Suter, Mathis Hoffmann, Jonas Utz, Daniel Augsburger, Mareike Thies, Mingxuan Gu, Fabian Wagner, George Sarau, Silke Christiansen, Katharina Breininger

{"title":"Addressing data scarcity in nanomaterial segmentation networks with differentiable rendering and generative modeling","authors":"Dennis Possart, Leonid Mill, Florian Vollnhals, Tor Hildebrand, Peter Suter, Mathis Hoffmann, Jonas Utz, Daniel Augsburger, Mareike Thies, Mingxuan Gu, Fabian Wagner, George Sarau, Silke Christiansen, Katharina Breininger","doi":"10.1038/s41524-025-01702-6","DOIUrl":"https://doi.org/10.1038/s41524-025-01702-6","url":null,"abstract":"Nanomaterials’ properties, influenced by size, shape, and surface characteristics, are crucial for their technological, biological, and environmental applications. Accurate quantification of these materials is essential for advancing research. Deep learning segmentation networks offer precise, automated analysis, but their effectiveness depends on representative annotated datasets, which are difficult to obtain due to the high cost and manual effort required for imaging and annotation. To address this, we present DiffRenderGAN, a generative model that produces annotated synthetic data by integrating a differentiable renderer into a Generative Adversarial Network (GAN) framework. DiffRenderGAN optimizes rendering parameters to produce realistic, annotated images from non-annotated real microscopy images, reducing manual effort and improving segmentation performance compared to existing methods. Tested on ion and electron microscopy datasets, including titanium dioxide (TiO2), silicon dioxide (SiO2), and silver nanowires (AgNW), DiffRenderGAN bridges the gap between synthetic and real data, advancing the quantification and understanding of complex nanomaterial systems.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"47 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-07-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144534060","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

Agent-based multimodal information extraction for nanomaterials 基于agent的纳米材料多模态信息提取

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-06-23 DOI: 10.1038/s41524-025-01674-7

R. Odobesku, K. Romanova, S. Mirzaeva, O. Zagorulko, R. Sim, R. Khakimullin, J. Razlivina, A. Dmitrenko, V. Vinogradov

{"title":"Agent-based multimodal information extraction for nanomaterials","authors":"R. Odobesku, K. Romanova, S. Mirzaeva, O. Zagorulko, R. Sim, R. Khakimullin, J. Razlivina, A. Dmitrenko, V. Vinogradov","doi":"10.1038/s41524-025-01674-7","DOIUrl":"https://doi.org/10.1038/s41524-025-01674-7","url":null,"abstract":"Automating structured data extraction from scientific literature is a critical challenge with broad implications across domains. We introduce nanoMINER, a multi-agent system combining large language models and multimodal analysis to extract essential information from scientific research articles on nanomaterials. This system processes documents end-to-end, utilizing tools such as YOLO for visual data extraction and GPT-4o for linking textual and visual information. At its core, the ReAct agent orchestrates specialized agents to ensure comprehensive data extraction. We demonstrate the efficacy of the system by automating the assembly of nanomaterial and nanozyme datasets previously manually curated by domain experts. NanoMINER achieves high precision in extracting nanomaterial properties like chemical formulas, crystal systems, and surface characteristics. For nanozymes, we obtain near-perfect precision (0.98) for kinetic parameters and essential features such as Cmin and Cmax. To benchmark the system performance, we also compare nanoMINER to several baseline LLMs, including the most recent multimodal GPT-4.1, and show consistently higher extraction precision and recall. Our approach is extensible to other domains of materials science and fields like biomedicine, advancing data-driven research methodologies and automated knowledge extraction.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"18 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144341315","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

Machine learning revealed giant thermal conductivity reduction by strong phonon localization in two-angle disordered twisted multilayer graphene 机器学习揭示了双角无序扭曲多层石墨烯中强声子局域化的巨大热导率降低

IF 9.7 1区材料科学

npj Computational Materials Pub Date : 2025-06-23 DOI: 10.1038/s41524-025-01678-3

Jingwen Wang, Zheng Zhu, Tianran Jiang, Ke Chen

{"title":"Machine learning revealed giant thermal conductivity reduction by strong phonon localization in two-angle disordered twisted multilayer graphene","authors":"Jingwen Wang, Zheng Zhu, Tianran Jiang, Ke Chen","doi":"10.1038/s41524-025-01678-3","DOIUrl":"https://doi.org/10.1038/s41524-025-01678-3","url":null,"abstract":"In two-dimensional (2D) layer-stacked materials, the twist angle between layers provides extensive freedom to explore novel physics and engineer remarkable thermal transport properties. We discovered that the cross-plane thermal conductivity of multilayer graphene can be effectively controlled by arranging the layers with two specific twist angles in a defined sequence. Disorderly aperiodic twisted graphene layers lead to the localization of phonons, substantially reducing the cross-plane thermal transport via the interference of coherent phonons. We employed non-equilibrium molecular dynamics simulations combined with machine learning approach, to study heat transport in the two-angle disordered multilayer stacks, and identified within the constrained structural space the optimal stacking sequence that can minimize the cross-plane thermal conductivity. Compared to pristine graphite, the optimized structure can reduce thermal conductivity by up to 80%. Through analysis of phonon transport properties across different structures, we revealed the underlying physical mechanism of phonon localization.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"269 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-06-23","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"144341317","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