npj Computational Materials最新文献

{"title":"Transferable machine learning model for multi-target nanoscale simulations in hydrogen-carbon system from crystal to amorphous","authors":"Weiqi Chen, Zhiyue Xu, Kang Wang, Lei Gao, Aisheng Song, Tianbao Ma","doi":"10.1038/s41524-025-01629-y","DOIUrl":"https://doi.org/10.1038/s41524-025-01629-y","url":null,"abstract":"<p>Carbon materials especially with hydrogenation have attracted wide attention for their novel physical and chemical properties and broad application prospects. A systematic theoretical simulation method accurately describing atomic interactions for hydrogen-carbon systems is crucial for the design of carbon-based materials and their industrial applications. Multiphases of hydrogenated carbon materials, from crystal to amorphous, with covalent network and diverse chemical reactions bring huge difficulties to construct a general interatomic potential under various conditions. Here, we demonstrate a transferable active machine learning scheme with separated training of sub-feature spaces and target-oriented finetuning, and construct a general-purpose pre-trained machine learning potential (MLP) for hydrogen-carbon systems. The pre-trained MLP is further efficiently transferred to three target spaces of deposition, friction and fracture with scale reliability. This work provides a robust tool for the theoretical research of hydrogen-carbon systems and a general scheme for developing transferable MLPs in multiphase systems across compositional and conditional complexity.</p>","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"93 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-05-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143903122","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}

{"title":"High throughput computational screening and interpretable machine learning for iodine capture of metal-organic frameworks","authors":"Haoyi Tan, Yukun Teng, Guangcun Shan","doi":"10.1038/s41524-025-01617-2","DOIUrl":"https://doi.org/10.1038/s41524-025-01617-2","url":null,"abstract":"<p>The removal of leaked radioactive iodine isotopes in humid air environments holds significant importance in nuclear waste management and nuclear accident mitigation. In this study, high-throughput computational screening and machine learning were combined to reveal the iodine capture performance of 1816 metal-organic framework (MOF) materials under humid air conditions. Initially, the relationship between the structural characteristics of MOF materials (including density, surface area and pore features) and their adsorption properties was explored, with the aim of identifying the optimal structural parameters for iodine capture. Subsequently, two machine learning regression algorithms—Random Forest and CatBoost, were employed to predict the iodine adsorption capabilities of MOF materials. In addition to 6 structural features, 25 molecular features (encompassing the types of metal and ligand atoms as well as bonding modes) and 8 chemical features (including heat of adsorption and Henry’s coefficient) were incorporated to enhance the prediction accuracy of the machine learning algorithms. Feature importance was assessed to determine the relative influence of various features on iodine adsorption performance, in which the Henry’s coefficient and heat of adsorption to iodine were found the two most crucial chemical factors. Furthermore, four types of molecular fingerprints were introduced for providing comprehensive and detailed structural information of MOF materials. The 20 most significant Molecular ACCess Systems (MACCS) bits were picked out, revealing that the presence of six-membered ring structures and nitrogen atoms in the MOF framework were the key structural factors that enhanced iodine adsorption, followed by the presence of oxygen atoms. This work combined high-throughput computation, machine learning, and molecular fingerprints to comprehensively and systematically elucidate the multifaceted factors governing the iodine adsorption performance of MOFs in humid environments, establishing a robust and profound guideline framework for accelerating the screening and targeted design of high-performance MOF materials.</p>","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"15 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-05-02","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143901472","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}

{"title":"Electron-mediated anharmonicity and its role in the Raman spectrum of graphene","authors":"Nina Girotto Erhardt, Aloïs Castellano, J. P. Alvarinhas Batista, Raffaello Bianco, Ivor Lončarić, Matthieu J. Verstraete, Dino Novko","doi":"10.1038/s41524-025-01610-9","DOIUrl":"https://doi.org/10.1038/s41524-025-01610-9","url":null,"abstract":"<p>The Raman active G mode in graphene exhibits a strong coupling to electrons, yet the comprehensive treatment of this interaction in the calculation of its temperature-dependent Raman spectrum remains incomplete. In this study, we calculate the temperature dependence of the G-mode frequency and linewidth, and successfully explain the experimental trend by accounting for the contributions arising from the first-order electron-phonon coupling, electron-mediated phonon-phonon coupling, and standard lattice anharmonicity. The generality of our approach enables its broad applicability to study phonon dynamics in materials where both electron-phonon coupling and anharmonicity are important.</p>","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"43 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-04-29","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143889482","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}

{"title":"Strain and ligand effects in the 1-D limit: reactivity of steps","authors":"Onyinyechukwu Goodness Njoku, Paige Fronczak, Kara Smeltz, Ian T. McCrum","doi":"10.1038/s41524-025-01616-3","DOIUrl":"https://doi.org/10.1038/s41524-025-01616-3","url":null,"abstract":"<p>The predictive design of alloy (electro)catalysts is necessary to identify catalysts more active, selective, stable, and low-cost than the pure metals. Our fundamental understanding of the catalytic behavior of alloys is limited however as it is typically derived from that of flat, “pristine” surfaces, not the industrially-relevant, defect-rich surfaces found on nanoparticles. We use density functional theory (DFT) modeling to probe strain, ligand, and ensemble effects on transition metal surfaces with step-defects. We find the response of the step to strain and ligand effects is much smaller in magnitude and sometimes opposite in direction to that of a flat surface, due to the breaking of two-dimensional symmetry at the step. Insight gained from flat surfaces alone is therefore not sufficient to understand (alloy) nanoparticles; defect sites must be explicitly considered. We additionally find that the one-dimensional, bimetallic ensemble created by the selective decoration of step defects can break adsorbate scaling, yielding surface alloys with potentially enhanced catalytic performance.</p>","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"76 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143876016","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}

{"title":"Photoinduced ferroelectric phase transition triggering photocatalytic water splitting","authors":"Jun Wen, Zhi-rui Luo, Lin-can Fang, Wen-xian Chen, Gui-lin Zhuang","doi":"10.1038/s41524-025-01601-w","DOIUrl":"https://doi.org/10.1038/s41524-025-01601-w","url":null,"abstract":"<p>Utilizing two-dimensional (2D) ferroelectric semiconductors for photocatalytic water splitting (PWS) to produce clean hydrogen fuel shows promise but faces performance regulation challenges. This study employs real-time time-dependent density functional theory (rt-TDDFT) and first-principle calculations to propose a “one stone, two birds” strategy: light induces ferroelectric phase transitions and triggers PWS on monolayer Hf₂Ge₂S₆. Electronically, monolayer Hf₂Ge₂S₆ exhibits excellent stability, mechanical properties, an appropriate band gap, optimal band edge positions, and broad light absorption. Its ferroelectric (FE) phase promotes oxygen evolution reaction(OER), while the paraelectric (PE) phase enhances hydrogen evolution reaction(HER). Specifically, applying 10% compressive strain effectively suppresses OER on the FE phase, while a mere 2% tensile strain can induce complete spontaneity in HER on the PE phase. Finally, rt-TDDFT simulation results demonstrate that laser pulses can drive effective ion displacements of Ge atoms in monolayer Hf₂Ge₂S₆ and thereby generate the transition from FE to PE, which is attributed to the maintenance of charge distribution asymmetry through internal atomic electron transfers. More importantly, this recyclable ferroelectric photocatalyst, activated by light and electric fields, effectively prevents performance drawbacks from pure electric fields, demonstrating that a photoelectric alternating field can regulate PWS performance. These findings demonstrate that a photoelectric alternating field is an effective strategy to regulate photocatalytic performance for PWS.</p>","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"44 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-04-25","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143876009","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}

{"title":"A machine learning model with minimize feature parameters for multi-type hydrogen evolution catalyst prediction","authors":"Chao Wang, Bing Wang, Changhao Wang, Aojian Li, Zhipeng Chang, Ruzhi Wang","doi":"10.1038/s41524-025-01607-4","DOIUrl":"https://doi.org/10.1038/s41524-025-01607-4","url":null,"abstract":"<p>The vast chemical compositional space presents challenges in catalyst development using traditional methods. Machine learning (ML) offers new opportunities, but current ML models are typically limited to screening a single catalyst type. In this work, we developed an efficient ML model to predict hydrogen evolution reaction (HER) activity across diverse catalysts. By minimizing features, we introduced a key energy-related feature <i>φ</i> = <span>({{rm{Nd}}0}^{2}/{rm{psi }}0)</span>, which correlates with HER free energy. Using just ten features, the Extremely Randomized Trees model achieved <i>R</i>² = 0.922. We predicted 132 new catalysts from the Material Project database, among which several exhibited promising HER performance. The time consumed by the ML model for predictions is one 200,000th of that required by traditional density functional theory (DFT) methods. The model provides an efficient approach for discovering high-performance HER catalysts using a small number of key features and offers insights for the development of other catalysts.</p>","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"42 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143872932","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}

{"title":"Uncertainty quantification for neural network potential foundation models","authors":"Jenna A. Bilbrey, Jesun S. Firoz, Mal-Soon Lee, Sutanay Choudhury","doi":"10.1038/s41524-025-01572-y","DOIUrl":"https://doi.org/10.1038/s41524-025-01572-y","url":null,"abstract":"<p>For neural network potentials (NNPs) to gain widespread use, researchers must be able to trust model outputs. However, the blackbox nature of neural networks and their inherent stochasticity are often deterrents, especially for foundation models trained over broad swaths of chemical space. Uncertainty information provided at the time of prediction can help reduce aversion to NNPs. In this work, we detail two uncertainty quantification (UQ) methods. Readout ensembling, by finetuning the readout layers of an ensemble of foundation models, provides information about model uncertainty, while quantile regression, by replacing point predictions with distributional predictions, provides information about uncertainty within the underlying training data. We demonstrate our approach with the MACE-MP-0 model, applying UQ to the foundation model and a series of finetuned models. The uncertainties produced by the readout ensemble and quantile methods are demonstrated to be distinct measures by which the quality of the NNP output can be judged.</p>","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"69 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143866932","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}

{"title":"ELEQTRONeX: A GPU-accelerated exascale framework for non-equilibrium quantum transport in nanomaterials","authors":"Saurabh S. Sawant, François Léonard, Zhi Yao, Andrew Nonaka","doi":"10.1038/s41524-025-01604-7","DOIUrl":"https://doi.org/10.1038/s41524-025-01604-7","url":null,"abstract":"<p>Non-equilibrium electronic quantum transport is crucial for existing and envisioned electronic, optoelectronic, and spintronic devices. Encompassing atomistic to mesoscopic length scales in the same nonequilibrium device simulations has been challenging due to the computational cost of high-fidelity coupled multiphysics and multiscale requirements. In this work, we present ELEQTRONeX (<b>ELE</b>ctrostatic <b>Q</b>uantum <b>TR</b>ansport modeling <b>O</b>f <b>N</b>anomaterials at <b>eX</b>ascale), a massively parallel GPU-accelerated framework for self-consistently solving the nonequilibrium Green’s function formalism and electrostatics in complex device geometries. By customizing algorithms for GPU multithreading, we achieve significant improvement in computational time, and excellent scaling on up to 512 GPUs and billions of spatial grid cells. We validate our code by computing band structures, current-voltage characteristics, conductance, and drain-induced barrier lowering for various 3D configurations of carbon nanotube field-effect transistors, and demonstrate its suitability for complex device/material geometries where periodic approaches are not feasible, such as arrays of misaligned carbon nanotubes requiring fully 3D simulations.</p>","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"34 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143872933","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}

{"title":"Advancing organic photovoltaic materials by machine learning-driven design with polymer-unit fingerprints","authors":"Xiumin Liu, Xinyue Zhang, Ye Sheng, Zihe Zhang, Pan Xiong, Xuehai Ju, Junwu Zhu, Caichao Ye","doi":"10.1038/s41524-025-01608-3","DOIUrl":"https://doi.org/10.1038/s41524-025-01608-3","url":null,"abstract":"<p>To enhance the power conversion efficiency (PCE) of organic photovoltaic (OPV) cells, the identification of high-performance polymer/macromolecule materials and understanding their relationship with photovoltaic performance before synthesis are critical objectives. In this study, we developed five algorithms using a dataset of 1343 experimentally validated OPV NFA acceptor materials. The random forest (RF) algorithm exhibited the best predictive performance for material design and screening. Additionally, we explored a newly developed polymer/macromolecule structure expression, polymer-unit fingerprint (<i>PUFp</i>), which outperformed the molecular access system (MACCS) across diverse machine learning (ML) algorithms. <i>PUFp</i> facilitated the interpretability of structure-property relationships, enabling PCE predictions of conjugated polymers/macromolecules formed by the combination of donor (D) and acceptor (A) units. Our <i>PUFp</i>-ML model efficiently pre-evaluated and classified numerous acceptor materials, identifying and screening the two most promising NFA candidates. The proposed framework demonstrates the ability to design novel materials based on <i>PUFp</i>-ML-established feature/substructure-property relationships, providing rational design guidelines for developing high-performance OPV acceptors. These methodologies are transferable to donor materials, thereby supporting accelerated material discovery and offering insights for designing innovative OPV materials.</p>","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"13 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143866933","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}

{"title":"Insights into oxygen diffusion in rare earth disilicate environmental barrier coatings","authors":"Shiqiang Hao, Richard P. Oleksak, Ömer N. Doğan, Michael C. Gao","doi":"10.1038/s41524-024-01508-y","DOIUrl":"https://doi.org/10.1038/s41524-024-01508-y","url":null,"abstract":"<p>Environmental barrier coatings (EBC) are crucial for the use of SiC-based ceramic matrix composites in high-temperature combustion environments, yet knowledge of oxygen diffusion in these coatings is limited. This study investigates oxygen diffusion dynamics in the β-RE₂Si₂O₇ system to minimize oxygen penetration in rare earth disilicates. We analyze defect formation energy under varying oxygen conditions, identifying key diffusion mechanisms. In oxygen-rich environments, the most favorable neutral interstitial oxygen diffuses along the [110] direction. In oxygen-poor conditions, neutral oxygen vacancies rotate around Y and Si atoms, exhibiting a diffusivity of 6.59×10⁻²² m²/s at 1500 K for β-Yb₂Si₂O₇. Under intermediate oxygen levels, charged interstitial oxygen diffuses via concerted interstitialcy along the [001] direction with a diffusivity of 6.21×10⁻¹⁷ m²/s. Additionally, alloying rare earth Y with Er and Yb increases diffusion barriers, contributing to improved EBC performance in extreme environments. The insights gained provides valuable guidance for designing robust coatings tailored to withstand extreme operational environments.</p>","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"14 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-04-24","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"143866931","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}