[DFTB-Plus-User] bcc Fe in DFTB+

[Peter Klaver]

Hello people,

I wanted to see how well DFTB+ can reproduce DFT results for bcc Fe with some self-interstitial atoms in the lattice (radiation damage). I started with defect-free Fe. For bulk bcc Fe I find a magnetic moment of 2.13 Bohr magneton, which is quite good, but a lattice parameter of 2.942 Å, where experimental data extrapolated to 0 K gives somewhere halfway between 2.84-2.85 Å. So it’s 3.4% off. For an earlier use of DFTB+, on graphene and graphane, I found bond lengths were usually less than 1% different from DFT and still less than 2% for the worst cases.

The README with the pbc Slater-Koster parameterisation files says “The iron parameters have been applied to structural and magnetic properties of iron clusters [Iron], but should also give reasonable results for bcc bulk iron.” I wonder: is the 3.4% deviation in lattice parameter reasonable, or is it an indication that I might be doing something wrong?

I sought for DFTB+ papers on bcc Fe, but couldn’t find much. Has anyone here used DFTB+ for bcc Fe, and if so, what were your findings as to how well the pbc Slater-Koster files reproduce bcc Fe properties, in particular the lattice spacing?

One point where, as a relatively inexperienced DFTB+ user, I wasn’t too sure if I did things right was in the spin constants. I put in

SpinConstants = {
ShellResolvedSpin = Yes
Fe = {
-0.013 -0.009 -0.003 -0.009 -0.011 -0.001 -0.003 -0.001 -0.015
}
}

If anyone is working on bcc Fe, did you put in a similar line for that?

Kind regards,
Peter
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://mailman.zfn.uni-bremen.de/pipermail/dftb-plus-user/attachments/20260304/aab06c67/attachment.htm>