发表机构
Fudan University; University of Delaware（复旦大学; 特拉华大学）
机构由 AI 辅助整理，请以论文原文为准。AI 中文总结
提出一种基于小超胞和少量DFT数据训练点缺陷机器学习原子间势的高效方案，仅需四次弛豫即可准确预测大超胞缺陷形成能，并验证了多尺寸数据与纯净超胞对提升外推能力的作用。
AI 中文摘要
机器学习原子间势（MLIP）能够以接近第一性原理的精度处理大规模材料体系，并已被广泛用于加速点缺陷模拟。然而，MLIP的训练通常依赖于大量的DFT数据。这一问题对于带电缺陷尤为突出，因为带电缺陷需要大超胞的DFT计算以避免长程库仑相互作用和有限尺寸效应，使得数据集的构建在计算上非常昂贵。在本工作中，我们提出了一种基于小超胞（少于100个原子）和有限数量DFT计算的高效MLIP训练方案，用于中性和低电荷态点缺陷。该方案仅需四次DFT结构弛豫即可构建训练数据集，且训练得到的专用于该缺陷的MLIP能够以较小的误差（大多小于0.3 eV）预测更大超胞（超过200个原子）中的缺陷形成能。以GaN、SiO$_2$和$\rm Cu_2ZnSnS_4$中的缺陷为代表案例，我们评估了该方案在预测不同超胞尺寸下缺陷总能和结构弛豫方面的外推能力。结果表明，仅使用单个小缺陷超胞训练的MLIP在处理大超胞时会产生严重误差。纳入多个超胞尺寸的缺陷数据可提高MLIP的预测性能，而添加无缺陷的纯净体相超胞则进一步提升了准确性。这些结果为以较低计算成本训练缺陷体系的MLIP模型提供了实用指导。
英文摘要
Machine learning interatomic potentials (MLIPs) can treat large-scale material systems with near first-principles accuracy and have been widely used to accelerate point-defect simulations. However, the training of MLIPs usually relies on large amounts of DFT data. This issue is particularly pronounced for charged defects, for which DFT calculations of large supercells are required to avoid long-range Coulomb interactions and finite-size effects, making the construction of datasets computationally expensive. In this work, we propose an efficient MLIP training scheme for neutral and lowly charged point defects based on small supercells (less than 100 atoms) and limited number of DFT calculations. The scheme requires only four DFT structural relaxations to construct the training dataset and the trained MLIPs dedicated for the defect can predict the defect formation energies in larger supercells (over 200 atoms) with small errors (mostly smaller than 0.3 eV). Using defects in GaN, SiO$_2$, and $\mathrm{Cu}_2\mathrm{ZnSnS}_4$ as representative examples, we evaluate the extrapolation capability of this scheme for predicting defect total energies and structural relaxations across different supercell sizes. The results show that an MLIP trained only on single small defect supercell produces severe errors when treating large supercells. Incorporating defect data from multiple supercell sizes improves the predictive performance of the MLIP, while adding pristine defect-free bulk supercells further enhances the accuracy. These results provide practical guidance for training MLIP models of defect systems with low computational costs.
