fcc Fe, antiferromagnetic lattice definition, tspin_2.in  [SOLVED]

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

Moderator: bguster

Locked
snow
Posts: 7
Joined: Fri Aug 07, 2015 6:53 pm

fcc Fe, antiferromagnetic lattice definition, tspin_2.in

Post by snow » Wed Dec 06, 2017 4:03 am

The tspin_2.in and tspin_3.in tutorials for AF fcc Fe both have this:

rprim
0.5 -0.5 0.0
0.5 0.5 0.0
0.0 0.0 1.0
xred
0.0 0.0 0.0
0.5 0.0 0.5
spinat
0.0 0.0 4.0
0.0 0.0 -4.0

I might be wrong but it seems like for atom 2 xred should be 0.5 0.5 0.5. Or were these intended to be xcart values? I'm thinking about this in a picture where the xyz coordinates line up with the conventional cell. Then all of the positions of the fcc lattice are generated. But that symmetry suggests FM nearest neighbors, which must be wrong. Maybe that's accounted for with the Shubnikov symmetries, however in tspin_3.in, the Shubnikov symmetries aren't used, and the cell parameters have not changed.

Plus when I look at what is given by the tutorial, I can't seem to generate all of the fcc lattice points, regardless of whether they're FM or AF.

Anyway, I hope the above is correct, because I don't see an easy way to create an AF fcc lattice. I suppose this is related to the Shubnikov. If so, could I get a reference to something good about this?

Thanks if you can clear this up. Take care. -Ryan

snow
Posts: 7
Joined: Fri Aug 07, 2015 6:53 pm

Re: fcc Fe, antiferromagnetic lattice definition, tspin_2.in  [SOLVED]

Post by snow » Tue Dec 12, 2017 7:16 pm

I think I have now realized that the lattice involved is not a simple AF system, but as the note within the tutorial input files mention, it is a spin-spiral.

Then it makes sense. The first two lattice vectors are to conventional face-points, the third to a conventional vertex. One face-point is internal to the volume and is anti-aligned to the origin. The two face-points of the lattice vectors are then collinear/aligned with the origin, I assume in agreement with the spin-spiral structure.

As for a simple fcc AF structure, it doesn't appear to exist. E.g. if you start to consider the origin up and all nearest neighbors down, then check and you'll see that some of the nearest neighbors have each other as nearest neighbors. So apparently there is some frustration, or at least the simple fcc AF structure doesn't exist.

Locked