MAGMOM tag

Message

zlch5 · #1 Post by **zlch5** » Wed Feb 13, 2008 4:49 am

Hello,

I am calculating CH free radical in a cubic cell. I found for different initial settings of MAGMOM, the total potential converges to different energies and the final mag. are also different. I use PBE potential in my calculations.

Here is my INCAR:

SYSTEM= CH
PREC = Accurate
ISMEAR= 0
SIGMA= 0.01
ISPIN = 2
NSW= 20
IDIPOL= 4
DIPOL= 0.5 0.5 0.5
LDIPOL= .TRUE.
IBRION = 2
POTIM = 0.2

The result is: 1 F= -.57293829E+01 E0= -.57265619E+01 d E =-.572938E+01 mag= 1.0000

But if I set MAGMOM= 2 1
The result is: 1 F= -.53967144E+01 E0= -.53967144E+01 d E =-.539671E+01 mag= 3.0000

Both energies are converged. I wonder if the different setting of MAGMOM affect the final results.

Thank you

#2 Post by **admin** » Wed Feb 13, 2008 12:21 pm

you can enforce a certain spin state by setting MAGMOM if the system has a stable high-spin and low-spin configuration.. However, if you compare the total energies you see that the low-spin configuration with mag=1 is more stable by 0.333 eV in your case.

boris · #3 Post by **boris** » Thu Feb 14, 2008 7:46 am

[quote="admin"]you can enforce a certain spin state by setting MAGMOM if the system has a stable high-spin and low-spin configuration...[/quote]

Does it mean that setting different MAGMOMs affect the initial filling of orbitals that can lead to a final ground state with either a low or high spin configuration ?

#4 Post by **admin** » Thu Feb 14, 2008 10:59 am

yes, in principle that can happen if there are (meta) stable high-and low spin configurations of the system which are separated by a barrier in the energy (potential) hypersurface.

boris · #5 Post by **boris** » Thu Feb 14, 2008 12:01 pm

Thus, is it better to set MAGMOMs according to what we expect them to be in the ground state? Or rather setting them according to the initial atomic configuration?

I'm encountering this problem with UO2. In their atomic configuration, the magnetic moments for U and O are respectively 4 and 2.

But in UO2, it is well known that U and O are ioniv (U4+ and O2-), that makes the magnetic moments be respectively 2 and 0.

My problem is :
When setting MAGMOM = 2 0 , the final total energy for the ground state is : E=-929.067261 eV.

When setting MAGMOM = 4 2 , the final total energy for the ground state is : E=-929.886684 eV.

For both cases, the total magnetic moment is 0 MuB (antiferromagnetic) and magnetic moments on U and O slightly differ depending on the initial MAGMOM setting.

I can't manage to explain this difference in energy.