I am working on a Heusler alloy computationally in the base of DFT. Now i need to specify that what kind of magnetism (itinerant or localized) is dominant.
As a preliminary step, I would consider spin charge density maps (charge density spin up minus spin down) and make a conclusion in what a manner spin density is distributed, being either localized on particular centers, or distributed more homogeneously in the unit cell. ,
Localized is a safer, but also a more conservative starting point. Note that, the answer may depend on the actual chemistry of your alloy (there are plenty of indications that different Heusler alloys may have sufficiently different magnetic properties that we can't rely on one-model-for-all approach). The previous good advice (VV Banikov) will get you to a certain point, but you will then need to develop your own intuition about your material and its phase diagram "relatives". As a matter of proper process, one would check what kind of experimental data are available, before you go too deep into parameters' choices and modeling.
In the periodic table the magnetism exists only for transition metals (3 d orbitals, Fe, Co, Ni compounds) with itinerant magnetism or with rare earths (4 f orbitals) with local magnetism. Let me to say just the main physical characteristics for both phases:
1. In a metallic state the itinerant magnetism is characterized for having spin polarization of the bands, i.e. the spin density states of spin up ↑ is different of the ones of spin down ↓.The shift between ↑ and ↓ Fermi bands becomes energetically possible only if electron polarization lowers the Coulomb energy as the Stoner criterion says (for instance, the gain in exchange energy exceds the loss in kinetic one for itinerant ferromagnetism). Thus the first step to characterize the itinerance of ferromagnetic state is to analyze the density of states of the material.
2. The second step is to see what important is the crystal field associated to the lattice discrete symmetries (which separates in energy the eg and t2g states of the d orbitals for the transition metals). If you have a spherical symmetry surrounding the magnetic ion then you can apply the useful Hund's rules (high and low spin transitions). In the octahedral symmetry the eg states have higher energy than the t2g while in tetrahedral happens the opposite.
3. In the d metals the crystal field energy is much higher than the spin-orbit coupling in opposition of what happens with the rare earhs. This makes that the f orbital magnetism is mainly localized. Assuming that he f levels are filled according to Hund‘s rule and Russel-Saunders coupling for the LS spins is a good approximation
I hope this helps.
- Badges
- Science topic
- Similar topics
- Solid State Physics
- Devices