Title: Oxygen Vacancy Formation Energy in Metal Oxides: High Throughput Computational Studies and Machine Learning Predictions Show Chinese title

Title: 金属氧化物中的氧空位形成能：高通量计算研究和机器学习预测

Abstract: The oxygen vacancy formation energy ($\Delta E_{vf}$) governs defect dynamics and is a useful metric to perform materials selection for a variety of applications. However, density functional theory (DFT) calculations of $\Delta E_{vf}$ come at a greater computational cost than the typical bulk calculations available in materials databases due to the involvement of multiple vacancy-containing supercells. As a result, available repositories of direct calculations of $\Delta E_{vf}$ remain relatively scarce, and the development of machine learning models capable of delivering accurate predictions is of interest. In the present, work we address both such points. We first report the results of new high-throughput DFT calculations of oxygen vacancy formation energies of the different unique oxygen sites in over 1000 different oxide materials, which together form the largest dataset of directly computed oxygen vacancy formation energies to date, to our knowledge. We then utilize the resulting dataset of $\sim$2500 $\Delta E_{vf}$ values to train random forest models with different sets of features, examining both novel features introduced in this work and ones previously employed in the literature. We demonstrate the benefits of including features that contain information specific to the vacancy site and account for both cation identity and oxidation state, and achieve a mean absolute error upon prediction of $\sim$0.3 eV/O, which is comparable to the accuracy observed upon comparison of DFT computations of oxygen vacancy formation energy and experimental results. Finally, we demonstrate the predictive power of the developed models in the search for new compounds for solar-thermochemical water-splitting applications, finding over 250 new AA$^{\prime}$BB$^{\prime}$O$_6$ double perovskite candidates. Show Chinese abstract

Abstract: 氧空位形成能（$\Delta E_{vf}$）决定了缺陷动力学，并且是用于多种应用材料选择的有用指标。 然而，由于涉及多个含空位的超细胞，密度泛函理论（DFT）计算的$\Delta E_{vf}$的计算成本比材料数据库中通常可用的体材料计算要高。 因此，直接计算$\Delta E_{vf}$的可用存储库仍然相对稀少，开发能够提供准确预测的机器学习模型具有重要意义。 在本工作中，我们解决了这两个问题。 我们首先报告了对超过1000种不同氧化物材料中不同独特氧位点的氧空位形成能的新高通量DFT计算结果，这些结果共同构成了目前已知的最大直接计算氧空位形成能的数据集。 然后，我们利用得到的$\sim$2500$\Delta E_{vf}$个值来训练具有不同特征集的随机森林模型，研究了本文引入的新特征以及文献中之前使用的特征。 我们展示了包含特定于空位位点的信息的特征的优势，并考虑了阳离子身份和氧化态，预测的平均绝对误差为$\sim$0.3 eV/O，这与氧空位形成能的DFT计算与实验结果比较中观察到的准确性相当。 最后，我们展示了所开发模型在寻找用于太阳能热化学水分解应用的新化合物方面的预测能力，发现了超过250种新的AA$^{\prime}$BB$^{\prime}$O$_6$双钙钛矿候选材料。