When we do spin polarized calculation then we get magnetic moment after each iteration if system is magnetic. In that case how does confirm type of magnetism ferro or antiferro in the system. Is there any limit of minimum magnetic moment for any type of magnetism.
How does report bandgap for spin polarized calculation ? In spin polarized calculation we get two bandgaps one for spin up and other for spin-down. Both have different bandgap values and experimentally we have only one bandgap value. What should be the exact value of bandgap for spin polarized calculations. For example we want to see trend in bandgap increasing or decreasing by dopant in case of spin polarized calculation. Any basic literature regarding the same. Thank You All
Hello Shilendra,
There is a specific flag in VASP to ask it to write down at the end of the OUTCAR the magnetic moment of each atom (LORBIT with setting =11, if my memories are corrected, thus check)
I'm not an expert of magnetism, but you can play a bit with the RWIGS as the wavefunctions will be projected on the spheres centered on each nucleus with a radius of RWIGS.
Knowing the magnetic moment of each nucleus you can see if it's FM or AFM type I or type II.
There is also other flag, NONCOLLINEAR if i remember well, but i can't remember how to use it, for this check the handsession of Martijn
MARSMAN, the IV or V maybe (the one about magnetism), i guess it is for the magnetic structure with a rotating magnetic moment along a direction, but I will not affirm it.
For the problem of 2 bandgap numerically / 1 bandgap experimentally, I never simulate semicond thus I let this question to the experts, but keep in mind that the experiments are not at 0 K whereas we are with VASP.
Also, it just a thought but it might be just the smallest bandgap which counts, thus you would have only jump of electron with a given spin between the valence to the conduction band . But it just a thought
You can check the free energy for ferromagnetic system and anti-ferromagnetic system. From DOS calculation you have to again run the system for all spin in + state (FM) and all spin in different combination to get AFM. Compare the free energy of the system and the lower value of ground state corresponds stable state. You will get many papers regarding this. If you need, I can send you some papers. give me your e-mail.
Hope this will help you.
As an example of Homnath Luitel's comment : comparison of free energy for the different magnetic phase,
you can take a look to the article (particularly the fig 2) of the article :
Article Ab initio full-potential study of the structural and magneti...
Dear @Homnath, thank you for your answer !
Here free energy means Gibbs free energy or ground state energy of the system. I use VASP code and I know spin polarized PDOS calculation. Just I want to know proper tags in VASP for making system ferromagnetic and antiferromagnetic. Generally we give LORBIT=11 for PDOS and MAGMOM value for ferro and antiferro system.
So for checking system is giving ferro-magnetism or antiferromagnetism we have to check both combination with MAGMOM value and then according to ground state energy which one is lower we will say this system is either ferro or antiferro. Am I right ? Any other way checking without MAGMOM value ferromagnetism and antiferromagnetism.
How does report final magnetic moment. Since in each iteration we get magnetic moment and final magnetic moment in OUTCAR file. In OUTCAR files it includes all s p d f orbitals. I am getting 2.0 magnetic moment after convergence but in OUTCAR total I am getting 1.65. This means 1.65 magnetic moment for whole system in which out of 64 atoms major contribution of magnetic moment coming from two oxygen atoms 28 and 44 (why magnetic moment only from two oxygen atoms).
You can send me some basic good paper on the topic my email id is [email protected] (upload limit is 16 MB). You can also send me INCAR file of ferro and antiferro setup.
Shilendra
Dear Shilendar
Be carreful, the last column in the outcar is the total local moment of/around the atom, it is not magntism of f state electron.
The magnetism report in, the outar is the difference between electronic density spin up and spin down inside the sphere of radius rwigs centred on each atom.
In the oszicar the magnetism is the total magnetic moment of all the supercell.
The differe.ce come from the fact that spheres are not filling the all volume of the system, there are interstitials regions not counting in het spheres and having a magnetic moment.
Dear Gargee Bhattacharyya,
I did not play a lot with magnestism aspect so I am not confident in the answer i would gave concerning the RWIGS. I can only say and advice that you should look at the local magnetic moment of the atoms under consideration in function of RWIGS, taking care of not use RWIGS values leading to a too big overlapping between the associated spheres of 2 adjacent atoms (see the VASP RWIGS webpage :https://cms.mpi.univie.ac.at/vasp/vasp/RWIGS.html)
I hope someone else will answer to your question because I am curious about it.
Concerning your second question, the magnetism given by the OSZICAR could be useful in case of ferromagnetic system.
But, you could probably feel at this stage that if the magnetic moment in OSZICAR is zero, it does not imply that your system is paramagnetic, it could be anti-ferromagnetic of type 1 or 2. Thus it would be the magnetic moment of the atoms that would allow you to know that. (I used a lot of example based on metallic systems because it is on these systems I am working on, I dont have any other example about it).
To fisnish, I would say that if you have a sufficiently good RWIGS value, the total magnetic moment coming from the sum of all the atomic magnetic moment in the OUTCAR should not be too different from the value in the OSZICAR.
Once again, I have a little expericence about magnetism thus, I will let a more skilled person confirm or contradict what I said.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Coding