# How to model an anti-ferromagnetic crystal in quantum ESPRESSO properly?

I'm fairly new to DFT. I would like to plot the DOS and band structure of Cobalt Oxide (CoO). But since CoO is anti ferro magnetic, how would I approach the problem? would I create an input file wherein I should assign starting magentization to spins as follows:

by using names like Co1 and Co2 in the Atomic species name card. Or would the act of putting nspin = 2, along with a starting magnetisation for all cobalt atoms suffice? Maybe someone who is experience could shed some light on how to properly model an anti-ferro magentic system.

Thank you.

• Do you want to model the antiferromagnetic interactions themselves? Or just to enforce antiferromagnetic ordering on the simulation? – taciteloquence Nov 4 '20 at 17:18
• +1. Excellent first question and welcome to Stack Exchange! Thank you for contributing your question here, and we hope to see much more of you! Also nice diagram for the spins in a magnet, you may be interested to see that I used the same one in this question: mattermodeling.stackexchange.com/q/136/5 (which I closed voluntarily because the question was too hard for anyone to answer, but it can be re-opened if you do think you have an answer to it). – Nike Dattani Nov 9 '20 at 17:16

• The key point is the setting of starting_magnetization. You can find an explanation on the pw.x input description page.

• Here I give you an example to set this tag. The structure picked here is the monolayer CrI$$_3$$. To achieve an AFM configuration, a square lattice with four Cr atoms is built as follows:

• Antiferromagnetic configuration setting: Cr1 ($$\uparrow$$) / Cr2 ($$\downarrow$$) / Cr3=Cr1 ($$\uparrow$$) / Cr4=Cr2 ($$\downarrow$$)

• All I atoms are non-magnetic.

• QE scf input card:

 &CONTROL
calculation='scf'
prefix='CrI3'
outdir='./tmp/'
pseudo_dir='./pseudo/'
etot_conv_thr=1.0d-05
/
&SYSTEM
ibrav=0
nat=16
ntyp=3
ecutwfc=60
ecutrho=720
occupations='smearing'
smearing='mp'
degauss=0.02
nspin=2
starting_magnetization(1)= 3  !note that ntype=3
starting_magnetization(2)=-3
starting_magnetization(3)= 0
/
&ELECTRONS
conv_thr=1.0d-8
mixing_beta=0.7
electron_maxstep=200
/
ATOMIC_SPECIES
Cr1  51.996 cr_pbe_v1.5.uspp.F.UPF
Cr2  51.996 cr_pbe_v1.5.uspp.F.UPF
I    126.9045 I.pbe-n-kjpaw_psl.0.2.UPF
ATOMIC_POSITIONS crystal
Cr1  0.0000000000  0.0000000000   0.3304567040
Cr2  0.0000000000  0.3333333430   0.3297766150
Cr1  0.5000000000  0.5000000000   0.3304567040
Cr2  0.5000000000  0.8333333730   0.3297766150
I    0.1495049890  0.1672649980   0.4083212910
I    0.6495050190  0.6672649980   0.4083212910
I    0.1743501420  0.4911199810   0.4083212910
I    0.6743501420  0.9911199810   0.4083212910
I    0.6761449580  0.3416150210   0.4083212910
I    0.1761449580  0.8416150210   0.4083212910
I    0.8504950400  0.1660683450   0.2519120570
I    0.3504950400  0.6660683160   0.2519120570
I    0.3256500360  0.3422133330   0.2519120570
I    0.8256500360  0.8422133330   0.2519120570
I    0.8238549830  0.4917183820   0.2519120570
I    0.3238549830  0.9917184110   0.2519120570
CELL_PARAMETERS angstrom
6.8670001030       0.0000000000       0.0000000000
0.0000000000      11.8939924240       0.0000000000
0.0000000000       0.0000000000      20.0000000000
K_POINTS automatic
4 2 2 0 0 0

• Total magnetization in each electronic step:

• Once you know how to do the self-consistent calculation, all left calculations are easily performed.
• Fantastic, super detailed answer! – taciteloquence Nov 10 '20 at 17:53