Decomposition of magnetization in mixalch calculation

Running Abinit Ground state
1

Dear abinit users,

Can someone help me to find out how to extract magnetization of dopant element using mixalch calculation? Is there any command?

For example I have the following mixalch block:
############################
npsp 3
znucl 50 27 26
ntypat 2
ntypalch 1
mixalch 2.5/3 0.5/3
##########################
where I substitute Co by Fe and I give the following spin block
#################
nsppol2 2
nspden2 2
spinat2 0.0 0.0 3.0
#################

It gives me the total magnetization in the cell. I want to calculate magnetic moment for both Co and Fe.

Let me know if there is possibility to calculate that.

Best,
Rajesh

2

Hi Rajesh,

in a word, you can’t. Mixalch implements the VCA, which considers all atoms to be the same average of the 2 species. It’s relatively accurate as you are mixing the pseudopotentials, and not just using a single nucleus with an average charge, but still, you can not distinguish the 2 species. With prtdensph you can get more local information around the atom, to complement the global data you have.

To distinguish you should do supercells with a finite fraction of dopants, and look locally.

M.