Computational mapping of full interfacial segregation spectrum for designing core-shell θ' precipitates in Al-Cu-based alloys

Abstract

With rising demand for lightweight high-temperature structural materials in aerospace and automotive industries, developing heat-resistant aluminum alloys has become a critical challenge. This work employs high-throughput density functional theory and machine learning to systematically map the comprehensive segregation profile of 26 alloying elements across the entire plane of the semi-coherent {100} θ′ |{100} Al interface in Al Cu alloys. By evaluating segregation energies in diverse atomic environments, we establish a full interfacial segregation spectrum that categorizes elements into distinct groups: (1) Mo, W, Cr, Mn, Fe, Co, and Ni preferentially occupying interfacial Cu sites; (2) Be, Ge, Si, and Ga displaying lower Al-site affinity than Cu; (3) Cd, Na, Li, Mg, Y, and Sc exhibiting stronger Al-site affinity than Cu; (4) Ti, V, Zr, Nb, Hf, and Ta showing negligible segregation; and (5) Au, Ag revealing standalone encapsulation capability. Random forest modeling and SHAP analysis reveal that size effects dominate segregation behavior, while chemical bonding plays a secondary role. These insights guide the design of heat-resistant Al Cu alloys with core-shell θ' precipitates, proposing a targeted micro-alloying strategy combining Cu-site segregators ( e.g. , Mn, Cr, Fe, Co), L1 2 -phase formers (Sc, Zr, Hf, Ti), and encapsulation elements (Ag) to synergistically enhance interfacial stability and suppress coarsening.