Unknown magnetic ground state
Posted: Wed Nov 05, 2014 8:04 pm
Dear VASP users,
I want to indicate the magnetic ground state of a bulk crystal which is not known experimentally, I know I have to set up different magnetic ordering which might be many Antiferromagnetic orderings(AF1, AF2, AF3, ...) and one Ferro ..., and in the end just comparing their SCF energy to find the lowest one.
My first question is:
Should the orders in a specific ordering for example AF2, must be fixed? if not, do not you think that all different orderings converge finally to the same state? as these orders we indicate in the beginning are just a initial ordering and they are probably allowed to change as SCF approaches to the convergence.
The second question is:
I was wondering if different orders (AF1, AF2, ... , Ferro ...), must be optimized and minimized first? and then take the optimized version of every order and run the usual SCF for every of them. Because as I know the magnetic state is a effective parameter on the final optimized structure. If the per-optimized is not that necessary, then which structure must be considered as platform to setup different magnetic ordering on it, maybe optimized one none-magnetically.
I hope I could clearly asked my question and any comment would be highly appreciated.
Best regards,
Salman
I want to indicate the magnetic ground state of a bulk crystal which is not known experimentally, I know I have to set up different magnetic ordering which might be many Antiferromagnetic orderings(AF1, AF2, AF3, ...) and one Ferro ..., and in the end just comparing their SCF energy to find the lowest one.
My first question is:
Should the orders in a specific ordering for example AF2, must be fixed? if not, do not you think that all different orderings converge finally to the same state? as these orders we indicate in the beginning are just a initial ordering and they are probably allowed to change as SCF approaches to the convergence.
The second question is:
I was wondering if different orders (AF1, AF2, ... , Ferro ...), must be optimized and minimized first? and then take the optimized version of every order and run the usual SCF for every of them. Because as I know the magnetic state is a effective parameter on the final optimized structure. If the per-optimized is not that necessary, then which structure must be considered as platform to setup different magnetic ordering on it, maybe optimized one none-magnetically.
I hope I could clearly asked my question and any comment would be highly appreciated.
Best regards,
Salman