Solute segregation at the interface between the aluminum (Al) matrix and the () phase decreases the interfacial energy, impedes the coarsening of precipitates, and enhances the thermal stability of such precipitates. In this study, we employ density functional theory to systematically calculate solute segregation energies of 42 solute elements at the coherent and semi-coherent interfaces between the two phases, as well as mixing energies of these elements within the Al and Cu sublattices of the phase. Using correlation analysis and machine learning methods, we establish the relationship between the solute segregation energy and 20 selected atomic descriptors. Metalloid and late transition metal elements are predicted as potential candidates for enhancing the thermal stability of Al–Cu alloys. We observe that the solute segregation energy at the interfacial site of the semi-coherent interface correlates with the atomic size of solute atoms and their solubilities within the phase. The developed machine learning models exhibit the potential to predict solute segregation energies at various sites of the coherent and semi-coherent interfaces. Overall, our study provides valuable insights into the stabilizing potential of individual elements at the /Al interface in Al–Cu alloys.

中文翻译：

---

研究 Al-Cu 合金中[公式省略]/Al 的界面偏析：利用密度泛函理论和机器学习的综合研究


铝（Al）基体和（）相之间界面处的溶质偏析降低了界面能，阻碍了析出物的粗化，并增强了此类析出物的热稳定性。在本研究中，我们采用密度泛函理论系统地计算了 42 种溶质元素在两相之间的相干和半相干界面处的溶质偏析能，以及这些元素在相的 Al 和 Cu 亚晶格内的混合能。利用相关分析和机器学习方法，我们建立了溶质偏析能和 20 个选定原子描述符之间的关系。类金属和后过渡金属元素被预测为增强 Al-Cu 合金热稳定性的潜在候选元素。我们观察到，半共格界面的界面位置处的溶质偏析能与溶质原子的原子尺寸及其在相内的溶解度相关。开发的机器学习模型表现出预测相干和半相干界面各个位置的溶质偏析能量的潜力。总的来说，我们的研究为 Al-Cu 合金中 /Al 界面上各个元素的稳定潜力提供了有价值的见解。