SELFE computes magnetic ordering temperatures and heat capacity from DFT-derived exchange parameters combined with an efficient self-consistent single-site sampling scheme, matching established methods on BCC-Fe and FCC-Fe with up to 142 times fewer averaging steps.
Ab initio calculation of the magnetic Gibbs free energy of materials using magnetically constrained supercells
1 Pith paper cite this work. Polarity classification is still indexing.
abstract
We present a first-principles approach for the computation of the magnetic Gibbs free energy of materials using magnetically constrained supercell calculations. Our approach is based on an adiabatic approximation of slowly varying local moment orientations, the so-called finite-temperature disordered local moment picture. It describes magnetic phase transitions and how electronic and/or magnetostructural mechanisms generate a discontinuous (first-order) character. We demonstrate that the statistical mechanics of the local moment orientations can be described by an affordable number of supercell calculations containing noncollinear magnetic configurations. The applicability of our approach is illustrated by firstly studying the ferromagnetic state in bcc Fe. We then investigate the temperature-dependent properties of a triangular antiferromagnetic state stabilizing in two antiperovskite systems Mn$_3$AN (A = Ga, Ni). Our calculations provide the negative volume expansion of these materials as well as the ab initio origin of the discontinuous character of the phase transitions, electronic and/or magnetostructural, in good agreement with experiment.
fields
cond-mat.mtrl-sci 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Magnetothermal Properties with Sampled Effective Local Field Estimation
SELFE computes magnetic ordering temperatures and heat capacity from DFT-derived exchange parameters combined with an efficient self-consistent single-site sampling scheme, matching established methods on BCC-Fe and FCC-Fe with up to 142 times fewer averaging steps.