Accurate binding energy database based on first-principles calculations for Monte Carlo simulations of aluminum-based alloy

Abstract The evaluation of binding energy in alloys is crucial for assessing structural stability and predicting new materials in design. Thus, calculating binding energy has become a central focus. Empirical potential functions provide computational efficiency and the ability to explore large-scale material systems without the need for high-performance computing resources. However, limitations such as reduced precision and restricted applicability to novel materials remain a challenge. To address this issue, a new method based on first-principles calculations is introduced to accurately determine the binding energy in vacancy-containing Al alloys. The study focuses on four binary alloys: Al-Si, Al-Zr, Al-Mg, and Al-Ti. The root mean square error is employed to evaluate the quality of this novel binding energy database, demonstrating that it achieves precision comparable to first-principles calculations within Monte Carlo (MC) simulations. The results confirm that this database successfully replicates the aging process in Al alloys while offering superior computational speed compared to first-principles calculations without compromising accuracy. This advancement significantly enhances the precision of atomic modeling, improving both predictive accuracy and computational efficiency for novel materials. Furthermore, this potential opens the door for broader applications in alloy design, paving the way for the discovery of new materials with optimized properties.