Pub Date : 2025-12-17DOI: 10.1038/s41524-025-01905-x
Kai Liu, Zixiong Wei, Wei Gao, Poulumi Dey, Marcel H. F. Sluiter, Fei Shuang
Universal machine-learning interatomic potentials (uMLIPs) are emerging as foundation models for atomistic simulation, offering near-ab initio accuracy at far lower cost. Their safe, broad deployment is limited by the absence of reliable, general uncertainty estimates. We present a unified, scalable uncertainty metric, U, built from a heterogeneous ensemble that reuses existing pretrained MLIPs. Across diverse chemistries and structures, U strongly tracks true prediction errors and robustly ranks configuration-level risk. Using U, we perform uncertainty-aware distillation to train system-specific potentials with far fewer labels: for tungsten, we match full density-functional-theory (DFT) training using 4% of the DFT data; for MoNbTaW, a dataset distilled by U supports high-accuracy potential training. By filtering numerical label noise, the distilled models can in some cases exceed the accuracy of the MLIPs trained on DFT data. This framework provides a practical reliability monitor and guides data selection and fine-tuning, enabling cost-efficient, accurate, and safer deployment of foundation models.
{"title":"Heterogeneous ensemble enables a universal uncertainty metric for atomistic foundation models","authors":"Kai Liu, Zixiong Wei, Wei Gao, Poulumi Dey, Marcel H. F. Sluiter, Fei Shuang","doi":"10.1038/s41524-025-01905-x","DOIUrl":"https://doi.org/10.1038/s41524-025-01905-x","url":null,"abstract":"Universal machine-learning interatomic potentials (uMLIPs) are emerging as foundation models for atomistic simulation, offering near-ab initio accuracy at far lower cost. Their safe, broad deployment is limited by the absence of reliable, general uncertainty estimates. We present a unified, scalable uncertainty metric, U, built from a heterogeneous ensemble that reuses existing pretrained MLIPs. Across diverse chemistries and structures, U strongly tracks true prediction errors and robustly ranks configuration-level risk. Using U, we perform uncertainty-aware distillation to train system-specific potentials with far fewer labels: for tungsten, we match full density-functional-theory (DFT) training using 4% of the DFT data; for MoNbTaW, a dataset distilled by U supports high-accuracy potential training. By filtering numerical label noise, the distilled models can in some cases exceed the accuracy of the MLIPs trained on DFT data. This framework provides a practical reliability monitor and guides data selection and fine-tuning, enabling cost-efficient, accurate, and safer deployment of foundation models.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"14 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145765588","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}
Pub Date : 2025-12-17DOI: 10.1038/s41524-025-01895-w
Jisu Kim, Jiho Lee, Sangmin Oh, Yutack Park, Seungwoo Hwang, Seungwu Han, Sungwoo Kang, Youngho Kang
Pretrained universal machine-learning interatomic potentials (MLIPs) have revolutionized computational materials science by enabling rapid atomistic simulations as efficient alternatives to ab initio methods. Fine-tuning pretrained MLIPs offers a practical approach to improving accuracy for materials and properties where predictive performance is insufficient. However, this approach often induces catastrophic forgetting, undermining the generalizability that is a key advantage of pretrained MLIPs. Herein, we propose reEWC, an advanced fine-tuning strategy that integrates Experience Replay and Elastic Weight Consolidation (EWC) to effectively balance forgetting prevention with fine-tuning efficiency. Using Li6PS5Cl (LPSC), a sulfide-based Li solid-state electrolyte, as a fine-tuning target, we show that reEWC significantly improves the accuracy of a pretrained MLIP, resolving well-known issues of potential energy surface softening and overestimated Li diffusivities. Moreover, reEWC preserves the generalizability of the pretrained MLIP and enables knowledge transfer to chemically distinct systems, including other sulfide, oxide, nitride, and halide electrolytes. Compared to Experience Replay and EWC used individually, reEWC delivers clear synergistic benefits, mitigating their respective limitations while maintaining computational efficiency. These results establish reEWC as a robust and effective solution for continual learning in MLIPs, enabling universal models that can advance materials research through large-scale, high-throughput simulations across diverse chemistries.
{"title":"An efficient forgetting-aware fine-tuning framework for pretrained universal machine-learning interatomic potentials","authors":"Jisu Kim, Jiho Lee, Sangmin Oh, Yutack Park, Seungwoo Hwang, Seungwu Han, Sungwoo Kang, Youngho Kang","doi":"10.1038/s41524-025-01895-w","DOIUrl":"https://doi.org/10.1038/s41524-025-01895-w","url":null,"abstract":"Pretrained universal machine-learning interatomic potentials (MLIPs) have revolutionized computational materials science by enabling rapid atomistic simulations as efficient alternatives to ab initio methods. Fine-tuning pretrained MLIPs offers a practical approach to improving accuracy for materials and properties where predictive performance is insufficient. However, this approach often induces catastrophic forgetting, undermining the generalizability that is a key advantage of pretrained MLIPs. Herein, we propose reEWC, an advanced fine-tuning strategy that integrates Experience Replay and Elastic Weight Consolidation (EWC) to effectively balance forgetting prevention with fine-tuning efficiency. Using Li6PS5Cl (LPSC), a sulfide-based Li solid-state electrolyte, as a fine-tuning target, we show that reEWC significantly improves the accuracy of a pretrained MLIP, resolving well-known issues of potential energy surface softening and overestimated Li diffusivities. Moreover, reEWC preserves the generalizability of the pretrained MLIP and enables knowledge transfer to chemically distinct systems, including other sulfide, oxide, nitride, and halide electrolytes. Compared to Experience Replay and EWC used individually, reEWC delivers clear synergistic benefits, mitigating their respective limitations while maintaining computational efficiency. These results establish reEWC as a robust and effective solution for continual learning in MLIPs, enabling universal models that can advance materials research through large-scale, high-throughput simulations across diverse chemistries.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"20 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-12-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145765592","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}
Pub Date : 2025-12-16DOI: 10.1038/s41524-025-01860-7
Aileen Luo, Tao Zhou, Ming Du, Martin V. Holt, Andrej Singer, Mathew J. Cherukara
Coherent X-ray scattering techniques are critical for investigating the fundamental structural properties of materials at the nanoscale. While advancements have made these experiments more accessible, real-time analysis remains a significant bottleneck, often hindered by artifacts and computational demands. In scanning X-ray nanodiffraction microscopy, which is widely used to spatially resolve structural heterogeneities, this challenge is compounded by the convolution of the divergent beam with the sample’s local structure. To address this, we introduce DONUT (Diffraction with Optics for Nanobeam by Unsupervised Training), a physics-aware neural network designed for the rapid and automated analysis of nanobeam diffraction data. By incorporating a differentiable geometric diffraction model directly into its architecture, DONUT learns to predict crystal lattice strain and orientation in real-time. Crucially, this is achieved without reliance on labeled datasets or pre-training, overcoming a fundamental limitation for supervised machine learning in X-ray science. We demonstrate experimentally that DONUT accurately extracts all features within the data over 200 times more efficiently than conventional fitting methods.
相干x射线散射技术对于研究纳米尺度材料的基本结构特性至关重要。虽然进步使这些实验更容易获得，但实时分析仍然是一个重要的瓶颈，经常受到工件和计算需求的阻碍。在扫描x射线纳米衍射显微镜中，这一挑战由于发散光束与样品局部结构的卷积而变得更加复杂。扫描x射线纳米衍射显微镜被广泛用于空间解析结构异质性。为了解决这个问题，我们引入了DONUT (Unsupervised Training with Optics for Nanobeam Diffraction)，这是一个物理感知的神经网络，旨在快速、自动地分析纳米束衍射数据。通过将可微的几何衍射模型直接集成到其结构中，DONUT可以实时学习预测晶格应变和取向。至关重要的是，这是在不依赖标记数据集或预训练的情况下实现的，克服了x射线科学中监督机器学习的基本限制。实验证明，与传统拟合方法相比，DONUT准确提取数据中的所有特征的效率提高了200倍以上。
{"title":"DONUT: physics-aware machine learning for real-time X-ray nanodiffraction analysis","authors":"Aileen Luo, Tao Zhou, Ming Du, Martin V. Holt, Andrej Singer, Mathew J. Cherukara","doi":"10.1038/s41524-025-01860-7","DOIUrl":"https://doi.org/10.1038/s41524-025-01860-7","url":null,"abstract":"Coherent X-ray scattering techniques are critical for investigating the fundamental structural properties of materials at the nanoscale. While advancements have made these experiments more accessible, real-time analysis remains a significant bottleneck, often hindered by artifacts and computational demands. In scanning X-ray nanodiffraction microscopy, which is widely used to spatially resolve structural heterogeneities, this challenge is compounded by the convolution of the divergent beam with the sample’s local structure. To address this, we introduce DONUT (Diffraction with Optics for Nanobeam by Unsupervised Training), a physics-aware neural network designed for the rapid and automated analysis of nanobeam diffraction data. By incorporating a differentiable geometric diffraction model directly into its architecture, DONUT learns to predict crystal lattice strain and orientation in real-time. Crucially, this is achieved without reliance on labeled datasets or pre-training, overcoming a fundamental limitation for supervised machine learning in X-ray science. We demonstrate experimentally that DONUT accurately extracts all features within the data over 200 times more efficiently than conventional fitting methods.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"6 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-12-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145771080","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}
Pub Date : 2025-12-15DOI: 10.1038/s41524-025-01900-2
Shengli Jiang, Michael A. Webb
Modifying solution viscosity is a key functional application of polymers, yet the interplay of molecular chemistry, polymer architecture, and intermolecular interactions makes tailoring precise rheological responses challenging. We introduce a computational framework coupling topology-aware generative machine learning, Gaussian process modeling, and multiparticle collision dynamics to design polymers yielding prescribed shear-rate-dependent viscosity profiles. Targeting thirty rheological profiles of varying difficulty, Bayesian optimization identifies polymers that satisfy all low- and most medium-difficulty targets by modifying topology and solvophobicity, with other variables fixed. In these regimes, we find and explain design degeneracy, where distinct polymers produce near-identical rheological profiles. However, satisfying high-difficulty targets requires extrapolation beyond the initial constrained design space; this is rationally guided by physical scaling theories. This integrated framework establishes a data-driven yet mechanistic route to rational polymer design.
{"title":"Generative active learning across polymer architectures and solvophobicities for targeted rheological behavior","authors":"Shengli Jiang, Michael A. Webb","doi":"10.1038/s41524-025-01900-2","DOIUrl":"https://doi.org/10.1038/s41524-025-01900-2","url":null,"abstract":"Modifying solution viscosity is a key functional application of polymers, yet the interplay of molecular chemistry, polymer architecture, and intermolecular interactions makes tailoring precise rheological responses challenging. We introduce a computational framework coupling topology-aware generative machine learning, Gaussian process modeling, and multiparticle collision dynamics to design polymers yielding prescribed shear-rate-dependent viscosity profiles. Targeting thirty rheological profiles of varying difficulty, Bayesian optimization identifies polymers that satisfy all low- and most medium-difficulty targets by modifying topology and solvophobicity, with other variables fixed. In these regimes, we find and explain design degeneracy, where distinct polymers produce near-identical rheological profiles. However, satisfying high-difficulty targets requires extrapolation beyond the initial constrained design space; this is rationally guided by physical scaling theories. This integrated framework establishes a data-driven yet mechanistic route to rational polymer design.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"16 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-12-15","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145759489","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}
Pub Date : 2025-12-14DOI: 10.1038/s41524-025-01909-7
Wenqi Xiong, Yu Jia
Activating ferroelectricity in two-dimensional (2D) bilayer materials is essential for enabling non-volatile memory and logic functionalities. While interlayer sliding driven by van der Waals interactions can break inversion symmetry, achieving out-of-plane (OOP) polarization through this mechanism remains challenging in highly symmetric 2D materials—particularly those that are centrosymmetric, such as graphene, hexagonal boron nitride (hBN), and transition metal dichalcogenides (TMDs). Here, we propose a general strategy to activate OOP ferroelectricity by intercalating inert atoms into the interlayer space. Using bilayer graphene, hBN, and MoS2 as model systems, we reveal that such intercalation lowers the symmetry from nonpolar D3d to polar C3v, enabling reversible polarization switching via lateral displacement of the intercalants. This resulting semi-sliding ferroelectricity preserves the structural and electronic integrity of the host materials, and features ultralow switching barriers along with atomic-scale dipole control—where each intercalated atom acts as an independent, reversible memory bit. Importantly, the polarization magnitude scales linearly with the electrostatic potential difference across the bilayer, providing a quantitative and tunable design rule. Our findings establish a universal and material-agnostic framework for realizing low-power, ultrahigh-density 2D ferroelectric devices on otherwise nonpolar platforms.
{"title":"General strategy for activating ferroelectricity in bilayer 2D materials with intercalating inert atoms","authors":"Wenqi Xiong, Yu Jia","doi":"10.1038/s41524-025-01909-7","DOIUrl":"https://doi.org/10.1038/s41524-025-01909-7","url":null,"abstract":"Activating ferroelectricity in two-dimensional (2D) bilayer materials is essential for enabling non-volatile memory and logic functionalities. While interlayer sliding driven by van der Waals interactions can break inversion symmetry, achieving out-of-plane (OOP) polarization through this mechanism remains challenging in highly symmetric 2D materials—particularly those that are centrosymmetric, such as graphene, hexagonal boron nitride (hBN), and transition metal dichalcogenides (TMDs). Here, we propose a general strategy to activate OOP ferroelectricity by intercalating inert atoms into the interlayer space. Using bilayer graphene, hBN, and MoS2 as model systems, we reveal that such intercalation lowers the symmetry from nonpolar D3d to polar C3v, enabling reversible polarization switching via lateral displacement of the intercalants. This resulting semi-sliding ferroelectricity preserves the structural and electronic integrity of the host materials, and features ultralow switching barriers along with atomic-scale dipole control—where each intercalated atom acts as an independent, reversible memory bit. Importantly, the polarization magnitude scales linearly with the electrostatic potential difference across the bilayer, providing a quantitative and tunable design rule. Our findings establish a universal and material-agnostic framework for realizing low-power, ultrahigh-density 2D ferroelectric devices on otherwise nonpolar platforms.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"35 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-12-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145746831","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}
Pub Date : 2025-12-14DOI: 10.1038/s41524-025-01849-2
Mingqian Li, Lifeng Wang, Zhuoqun Zheng
A coarse-grained neuroevolution potential (CGNEP) for multilayered graphene based on an ab initio accuracy dataset is developed for mesoscale molecular dynamics simulations. The information loss in coarsening process is discussed and divided into intralayer part and interlayer part. The CGNEP describes the interlayer shear introduced by van der Waals interactions well by modifying the descriptor of NEP. The mechanical properties and vibration frequencies of structures of different sizes are well predicted via CGNEP. Compared with the traditional empirical CG potential, the CGNEP possesses interlayer properties of the structure of graphene and maintains the ability for higher mapping ratio coarsening. The frequencies of a 12-layer graphene membrane with a length and width of 1 μm are directly calculated via the CGNEP with a 64:1 mapping ratio and compared with the experimental results. The proposed CGNEP may be further used for other multilayered CG 2D materials.
{"title":"Coarse-grained machine learning potential for mesoscale multilayered graphene","authors":"Mingqian Li, Lifeng Wang, Zhuoqun Zheng","doi":"10.1038/s41524-025-01849-2","DOIUrl":"https://doi.org/10.1038/s41524-025-01849-2","url":null,"abstract":"A coarse-grained neuroevolution potential (CGNEP) for multilayered graphene based on an ab initio accuracy dataset is developed for mesoscale molecular dynamics simulations. The information loss in coarsening process is discussed and divided into intralayer part and interlayer part. The CGNEP describes the interlayer shear introduced by van der Waals interactions well by modifying the descriptor of NEP. The mechanical properties and vibration frequencies of structures of different sizes are well predicted via CGNEP. Compared with the traditional empirical CG potential, the CGNEP possesses interlayer properties of the structure of graphene and maintains the ability for higher mapping ratio coarsening. The frequencies of a 12-layer graphene membrane with a length and width of 1 μm are directly calculated via the CGNEP with a 64:1 mapping ratio and compared with the experimental results. The proposed CGNEP may be further used for other multilayered CG 2D materials.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"70 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-12-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145746797","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}
Pub Date : 2025-12-14DOI: 10.1038/s41524-025-01858-1
Longlong Wu, David Yang, Wei Wang, Shinjae Yoo, Ross J. Harder, Wonsuk Cha, Aiguo Li, Ian K. Robinson
Visualization of internal deformation fields in crystalline materials helps bridge the gap between theoretical models and practical applications. Applying Bragg coherent diffraction imaging under X-ray dynamical diffraction conditions provides a promising approach to the longstanding challenge of investigating the deformation fields in micron-sized crystals. Here, we present an automatic differentiation-based reconstruction method that integrates dynamical scattering theory to accurately reconstruct deformation fields in large crystals. Using this forward model, our simulated and experimental results demonstrate that three-dimensional local strain information inside a large crystal can be accurately reconstructed under coherent X-ray dynamical diffraction conditions with Bragg coherent X-ray diffraction imaging. These findings open an avenue for extending the investigation of local deformation fields to microscale crystals while maintaining nanoscale resolution, leveraging the enhanced coherence and brightness of advanced X-ray sources.
{"title":"Unveiling nano-scale crystal deformation using coherent X-ray dynamical diffraction","authors":"Longlong Wu, David Yang, Wei Wang, Shinjae Yoo, Ross J. Harder, Wonsuk Cha, Aiguo Li, Ian K. Robinson","doi":"10.1038/s41524-025-01858-1","DOIUrl":"https://doi.org/10.1038/s41524-025-01858-1","url":null,"abstract":"Visualization of internal deformation fields in crystalline materials helps bridge the gap between theoretical models and practical applications. Applying Bragg coherent diffraction imaging under X-ray dynamical diffraction conditions provides a promising approach to the longstanding challenge of investigating the deformation fields in micron-sized crystals. Here, we present an automatic differentiation-based reconstruction method that integrates dynamical scattering theory to accurately reconstruct deformation fields in large crystals. Using this forward model, our simulated and experimental results demonstrate that three-dimensional local strain information inside a large crystal can be accurately reconstructed under coherent X-ray dynamical diffraction conditions with Bragg coherent X-ray diffraction imaging. These findings open an avenue for extending the investigation of local deformation fields to microscale crystals while maintaining nanoscale resolution, leveraging the enhanced coherence and brightness of advanced X-ray sources.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"20 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-12-14","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145746796","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}
Pub Date : 2025-12-13DOI: 10.1038/s41524-025-01862-5
Kailai Lin, Matthew J. Coley-O’Rourke, Eran Rabani
The semi-empirical pseudopotential method (SEPM) has been widely applied to provide computational insights into the electronic structure, photophysics, and charge carrier dynamics of nanoscale materials. We present “DeepPseudopot”, a machine-learned atomistic pseudopotential model that extends the SEPM framework by combining a flexible neural network representation of the local pseudopotential with parameterized non-local and spin-orbit coupling terms. Trained on bulk quasiparticle band structures and deformation potentials from GW calculations, the model captures many-body and relativistic effects with very high accuracy across diverse semiconducting materials, as illustrated for silicon and group III-V semiconductors. DeepPseudopot’s accuracy, efficiency, and transferability make it well-suited for data-driven in silico design and discovery of novel optoelectronic nanomaterials.
{"title":"Deep-learning atomistic semi-empirical pseudopotential model for nanomaterials","authors":"Kailai Lin, Matthew J. Coley-O’Rourke, Eran Rabani","doi":"10.1038/s41524-025-01862-5","DOIUrl":"https://doi.org/10.1038/s41524-025-01862-5","url":null,"abstract":"The semi-empirical pseudopotential method (SEPM) has been widely applied to provide computational insights into the electronic structure, photophysics, and charge carrier dynamics of nanoscale materials. We present “DeepPseudopot”, a machine-learned atomistic pseudopotential model that extends the SEPM framework by combining a flexible neural network representation of the local pseudopotential with parameterized non-local and spin-orbit coupling terms. Trained on bulk quasiparticle band structures and deformation potentials from GW calculations, the model captures many-body and relativistic effects with very high accuracy across diverse semiconducting materials, as illustrated for silicon and group III-V semiconductors. DeepPseudopot’s accuracy, efficiency, and transferability make it well-suited for data-driven in silico design and discovery of novel optoelectronic nanomaterials.","PeriodicalId":19342,"journal":{"name":"npj Computational Materials","volume":"1 1","pages":""},"PeriodicalIF":9.7,"publicationDate":"2025-12-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"145746833","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}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：