In some articles, i often read the authors mention that "spin-polarization effect" is implemented because the system is 'magnetic'. But i wish to know how can we 'preempt' whether the system is magnetic or not. Do we need some preliminary calculations to detect this?
It is based on the nature of atoms you have in the system. For examle we know iron or cobalt are magnetic. Thus if you have them in your system, you need to consider the spin-polarization.
I beleive including spin-polarization increase the calculation time. You can include it for the materials that are not magnetic and you should get the same results as the outcomes without using it. Please confirm this through literature or by running enough simulation.
Be mindful of your system temperature. You should exclude spin polarization for atoms above their curie temperature.
Every compound is by itself ‘magnetic’. The presence of unpaired electrons (spin multiplicity 2S+1 > 1) will determine a paramagnetic behaviour, while if all electrons are paired (2S+1 = 1) a molecule will be diamagnetic. In the more complex cases of spin crossover (SCO: Article Structural Diversity on a New Series of Halogenated Quinolyl...
) compounds unpaired spins can be magnetically or antiferromagnetically coupled. This can happen also in noninnoncent complexes, such as neutral bis(1,2-dithiolene) complexes of nd8 metal ions (see for exampleArticle Diradical Character of Neutral Heteroleptic Bis(1,2-dithiole...
). Many reviews are available on this topic.
I guess you are doing a first-principle calculation in solid materials. For that, you could see Materials Project, an online database for the computational properties of a vast variety of materials. You can see whether it is NM (non-magnetic), FM(ferromagnetic) or AFM (anti-ferromagnetic). According to this, you will be able to correctly set the initial magnetic moments in a first principle calculation.
- Badges
- Science topic
- Similar topics
- Computer Science
- Program