input files for dft+u with paw for molecule

Total energy, geometry optimization, DFT+U, spin....

Moderator: bguster

Locked
mina
Posts: 110
Joined: Mon Sep 27, 2010 8:44 am

input files for dft+u with paw for molecule

Post by mina » Thu Dec 29, 2016 9:17 pm

Dear User
I am new with abinit and i want to do some dft+u calculation for molecule.i see the tbase1_1.in for H2 molecule. would you please guide me
in molecules calculation we have to just define big acell? is it correct to define xangst for cartesian cordinates in angstrom length?
this is my input file :
natom 13
typat 1 1 1 1 1 1 2 2 2 2 2 2 3
znucl 6 1 27
ntypat 3
acell 20 20 20 angstrom
angdeg 90 90 90
xangst -0.6989050000 -1.2105300000 0
0.6988994957 -1.2105287487 0
1.3978133190 0 0
0.6988994957 1.2105287487 0
-0.6988994957 1.2105287487 0
-1.3978133190 0 0
-1.2453809199 -2.1571025050 0
1.2453809199 -2.1571025050 0
2.4908032154 0 0
1.2453809199 2.1571025050 0
-1.2453809100 2.1571025050 0
-2.4908032154 0 0
0 0 1.50
Best regards
Mina

lucamontana
Posts: 14
Joined: Tue Mar 15, 2016 3:09 am

Re: input files for dft+u with paw for molecule

Post by lucamontana » Sat Dec 31, 2016 1:47 am

Dear mina,

your input is correct, but i think that you have to further increase acell up to 60 angstrom in each direction in order to eliminate
artificial screening from periodic images to a large extent, in particular as you have a relatively heavy element such as Co.

I suggest you to resort to Quantum Espresso, as calculation becomes very time consuming upon system size increase,
and QE is much faster than Abinit.

If you need further helps, let me know.

Best wishes
LUCA

Locked