This study introduces a computationally efficient approach for determining the alloy structures expected under working conditions. The method combines a compact lattice Hamiltonian with Monte Carlo sampling and parameters fitted to density functional theory (DFT) data for archetypal nanoparticle arrangements. Applied to Pd-Pt-Cu, Ni-Pd-Cu, and Co-Rh-Cu nanoparticles, the model reproduced distinct ordering patterns and showed that temperature-driven disorder can strongly alter surface composition and reactivity. The approach also enabled direct estimation of mixing free energies, offering a practical route to more realistic simulations and design of multimetallic nanocatalysts.
- Arravind Subramanian
- Mikhail V. Polynski
- Sergey M. Kozlov
