Atomistic spin dynamics with quantum colored noise

Abstract: The accurate prediction of temperature-dependent magnetization dynamics is a fundamental challenge in computational magnetism. While Atomistic Spin Dynamics (ASD) simulations have emerged as a powerful tool for studying magnetic phenomena, their classical nature leads to significant deviations from experimental observations, particularly at low temperatures. Here we present a comprehensive implementation of quantum-corrected ASD into the Vampire software package, based on the open-system Landau-Lifshitz-Gilbert equation with a quantum thermostat. Our implementation incorporates memory effects along with colored noise derived from quantum-mechanical considerations that improve the description of the equilibrium magnetization. We demonstrate excellent quantitative agreement with experimental magnetization curves for nickel and gadolinium across the full temperature range. Our results establish that incorporating quantum environmental effects and colored noise substantially enhances the predictive capabilities of ASD simulations, providing a robust framework for modeling temperature-dependent magnetic phenomena in localized moment magnetic systems.