In this study, we use DFT+U to derive spin model Hamiltonians consisting of Heisenberg exchange interactions up to the fourth nearest neighbors and bi-quadratic interactions. We map DFT+U results of several magnetic configurations to the Heisenberg spin model Hamiltonian to estimate Heisenberg exchanges. We demonstrate that the number of magnetic configurations should be at least twice the number of exchange parameters to estimate exchange parameters correctly. To calculate biquadratic interaction, we propose specific non-collinear magnetic configurations that don’t change the energy of the Heisenberg spin model. We use classical Monte Carlo simulations

No abstract

In this study, we use DFT+U to derive spin model Hamiltonians consisting of Heisenberg exchange interactions up to the fourth nearest neighbors and bi-quadratic interactions. We map DFT+U results of several magnetic configurations to the Heisenberg spin model Hamiltonian to estimate Heisenberg exchanges. We demonstrate that the number of magnetic configurations should be at least twice the number of exchange parameters to estimate exchange parameters correctly. To calculate biquadratic interaction, we propose specific non-collinear magnetic configurations that don't change the energy of the Heisenberg spin model. We use classical Monte Carlo simulations to evaluate DFT+U results. We obtain the temperature dependence of magnetic susceptibility and specific heat to determine the Curie-Weiss and Néel temperatures. The MC simulations reveal that although the biquadratic interaction can not change the Néel temperature, it modifies the order parameter. We indicate that for a fair comparison between classical MC simulations and experiments, we need to consider the quantum effect by applying (S + 1)/S correction in classical Monte Carlo simulations.