Abstract

The use of the classical Heisenberg model which incorporates only transverse spin degrees of freedom has only limited success for description of the metallic magnetism at finite temperature, since temperature and magnetic disorder induced longitudinal variations of the atomic spin moments might become large in the itinerant electron systems away from the limit of localized moments. In order to incorporate the longitudinal spin fluctuations in finite temperature simulation schemes a simple extended version of the Heisenberg model which allows for an on-site spin magnitude variation controlled by the one-site energy terms is widely used during the recent decade for ab-initio mapping and statistical simulations. Here, we apply and discuss such ab-initio based scheme for the canonical itinerant ferromagnetic metals (Fe, Co, Ni) and recently discovered high temperature antiferromagnet - V3Al, in conjunction with standard spherical integration metrics in classical spin state and the recently proposed linear one. We also examine the dependence of the results on the choice of the exchange and correlation potential in ab-initio total energy calculations. We compare the respective uncertainties in the calculated values of the magnetic ordering temperature and temperature dependent spin moment magnitude to the difference in the results which relate to the choice of the metrics.

Highlights

  • Finite temperature magnetism of band electrons in d-transition metals poses a long standing problem in solid state physics[1] due to the fact that its proper description requires to consider the effects of electron correlation in the intermediate coupling regime

  • Correlation effects within spinpolarized d-bands are not so strong to lead to a metal-insulator transition and/or to a strong localization of the electronic states like in most of the Rare-Earth metals and intermetallic compounds

  • The correlation effects are important to consider and the essential progress has been done in the past three decades in the field of the itinerant magnetism by developing finite temperature correlated theories of the transition metals based on the Hubbard model and their hybrids with Local Spin Density Approximation (LSDA).[3]

Read more

Summary

Introduction

Finite temperature magnetism of band electrons in d-transition metals poses a long standing problem in solid state physics[1] due to the fact that its proper description requires to consider the effects of electron correlation in the intermediate coupling regime. Despite that one-electron theories, like Hartree-Fock or the Local Spin Density Approximation (LSDA), provide an explanation of the non-integer value of the spin moment in metals and often a reasonable description of the magnetically ordered ground state even on a first-principles basis, the one-electron picture (Stoner-like) theory completely fails to describe finite temperature properties of the itinerant metallic magnets.[1,2] the correlation effects are important to consider and the essential progress has been done in the past three decades in the field of the itinerant magnetism by developing finite temperature correlated theories of the transition metals based on the Hubbard model and their hybrids with LSDA..

Methods
Results
Conclusion
Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call