Page 1 of 1
Use relativistic pseudopotentials while performing MAE calculations in VASP?
Posted: Sat May 18, 2024 1:24 pm
by farah_shahzadi
I am currently working on a project that involves calculating the magnetocrystalline anisotropy energy (MAE) of a magnetic material using VASP. I have encountered some discussions suggesting that relativistic effects can significantly impact the accuracy of MAE calculations.Do I need to use relativistic pseudopotentials while performing MAE calculations in VASP?
Re: Use relativistic pseudopotentials while performing MAE calculations in VASP?
Posted: Tue May 21, 2024 7:27 am
by jonathan_lahnsteiner2
Dear farah,
This question can not be answered in general because it will strongly depend on the material if relativistic effects are important or not. To switch on relativistic corrections in vasp you should use
spin orbit coupling.
The pseudopotentials itself include sclalar relativistic effects. Please check this
forum post for further information.
All the best Jonathan
Re: Use relativistic pseudopotentials while performing MAE calculations in VASP?
Posted: Thu May 23, 2024 5:43 am
by farah_shahzadi
Dear Jonathan,
Thank you for your response. I would appreciate it if you could further clarify the following: What specific factors in the material determine the significance of relativistic effects when calculating the magnetocrystalline anisotropy energy?
Best Regard,
Farah Shehzadi
Re: Use relativistic pseudopotentials while performing MAE calculations in VASP?
Posted: Thu May 23, 2024 11:12 am
by marie-therese.huebsch
Hi,
could you explain what you mean by "relativistic effects can significantly impact the accuracy of MAE "?
The MAE itself is a relativistic effect as it is caused by SOC.
There are no special "relativistic pseudopotentials". The usual pseudopotentials are already scalar relativistic. As a side remark: When choosing the pseudopotential mind that the states that cause MAE, i.e., that have a finite magnetization, should be in the valence. You can check that by analyzing the atomic configuration.
Of course it is always possible that you need a full 4 component Dirac Hamiltonian to describe your system, but that cannot be done with VASP. Basically it is a matter of comparing your results with the experiment. But even if you have disagreement with the experiment, I think it is more likely that you need a different/higher level method to include exchange correlation effects before going to Dirac theory.
Anyways, please let me know what you have read about the accuracy of MAE. I would be interested.
Marie-Therese