Dear Prof. Blaha, dear All,
I did not mention that the only reason for swapping equivalent atoms is
to look at a partial charge of a particular equivalent atom using the
QTL code. "x lapw2" sums all the equivalent atoms and therefore averages
out things like the "hidden spin polarization" (of Zunger&Freeman).
I made some tests with bulk WSe2. I made a bulk calculation with spin
and SOC. WSe2 bulk case.struct file has only 2 non-equivalent atoms: 4
equivalent Se and 2 equivalent W.
1). I looked at fat bands of one of the W atoms using "x qtl".
2). I swapped the equivalent positions of W atoms in case.struct (I
edited case.struct file leaving everything else intact). WSe2 is 2H so
this means that now the first equivalent W position is in the adjacent
layer.
3). I again looked at fat bands using "x qtl".
This seems to work, the spin-up fat bands swapped between the upper and
lower bands at K, as expected when swapping to the adjacent layer in
WSe2.
So this would mean one can control which equivalent atom "x qtl" is
looking at by manually editing case.struct. I assume "x qtl" is always
looking at the first equivalent atom in case.struct.
This would mean no new convergence is needed (which would be really
great).
Could you please confirm?
Best,
Lukasz
P.S.: In my previous email, wherever I mentioned 'non-equivalent,' I
actually meant 'equivalent.' I apologize for any confusion this may have
caused.
On 2024-12-26 10:14, pluto via Wien wrote:
Dear Prof. Blaha, dear All,
I converged large slab that contains some Mn atoms with spin, SOC, and
GGA+U. I saved the calculation with save_lapw. Here is one of the Mn
atoms in the case.struct:
ATOM -5: X=0.00000000 Y=0.50000000 Z=0.22529133
MULT= 4 ISPLIT= 8
-5: X=0.00000000 Y=0.50000000 Z=0.77470867
-5: X=0.50000000 Y=0.50000000 Z=0.22529133
-5: X=0.50000000 Y=0.50000000 Z=0.77470867
Mn2 NPT= 781 R0=.000050000 RMT= 2.50000 Z: 25.00000
I would like to reconverge this calculation, but swap the
non-equivalent positions into:
ATOM -5: X=0.50000000 Y=0.50000000 Z=0.22529133
MULT= 4 ISPLIT= 8
-5: X=0.50000000 Y=0.50000000 Z=0.77470867
-5: X=0.00000000 Y=0.50000000 Z=0.22529133
-5: X=0.00000000 Y=0.50000000 Z=0.77470867
Mn2 NPT= 781 R0=.000050000 RMT= 2.50000 Z: 25.00000
Can I do the following:
1). restore_lapw (in a new folder)
2). edit case.struct as explained above, that is swap non-equivalent Mn
atoms
3). Run the SCF calculation with -NI
Or perhaps I need to do init_lapw and init_so_lapw again?
Best,
Lukasz
_______________________________________________
Wien mailing list
[email protected]
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:
http://www.mail-archive.com/[email protected]/index.html
_______________________________________________
Wien mailing list
[email protected]
http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
SEARCH the MAILING-LIST at:
http://www.mail-archive.com/[email protected]/index.html
