<html><body><div>Dear prof. Blaha, <br></div><div><br></div><div>wow, thanks a lot for such a detailed answer. This was very helpful indeed.</div><div><br></div><div>Best regards</div><div>Pavel<br></div><br><blockquote data-email="peter.blaha@tuwien.ac.at">Hi,
<br>Here are my comments. Most of them similar to what Laurence said.
<br>
<br>> I'm trying to calculate a band structure of Tb3Ga5O12 magneto-optical
<br>> crystal (cubic Ia-3d, 80 atoms). While I consider myself quite
<br>>
<br>> Luckily I'm not shooting completely blind as I have some high-quality
<br>> optical data where we can see some (very weak but also quite sharp and
<br>> hence noticeable f-f transitions in the band gap so I have some idea
<br>> how the Tb f states at least should look like). Significant optical
<br>> absorption start around 4eV but below that I see some very weak
<br>> electronic transitions in the 0.2-0.9eV range, around 2.5 and 3.5eV
<br>> (reportedly between f states located in the band gap). So I expect at
<br>> least three bunches of f states in band gap one occupied and the others
<br>> unoccupied.
<br>
<br>Unfortunately, I don't believe that these optical f-f transitions can be
<br>described by DFT. These are crystal-field splitted multiplet
<br>excitations, which are usually not accessible by DFT.
<br>
<br>PS: Optical transitions create an electron-hole pair and excitonic
<br>(correlation) effects can be very large.
<br>XPS creates a free electron and a hole and although this is also not a
<br>ground state, it is usually better described by groundstate DFT.
<br>
<br> From your chemical formulae one expects Tb3+, i.e. a fully occupied
<br>spin-up 4f band and a single 4f electron occupied in spin-dn.
<br>Of course, PBE gives a metal and the 4f-dn states are pinned at EF.
<br>An orbital potential can split these states and single out a single 4f
<br>electron/atom. However, with orbital potentials in many cases one can
<br>obtain several different orbitals occupied, which depends on the
<br>starting density matrix. In other words, your solution may not be the
<br>ground state, but a metastable state.
<br>Therefor I'd do first GGA+SO, and "hope" that this gives me a bit larger
<br>occupancy of the "correct" 4f orbital. When you then calculate the
<br>density matrix from this solution, you may run in the lowest energy
<br>orbitally-ordered state. Eventually, you could also start from different
<br>density matrices and see to which solutions you converge and compare
<br>total energies (these manipulations are simpler in DFT+U than in EECE).
<br>
<br>RMTs: Since we cannot use HDLOs for orbital potentials, too large
<br>spheres are not good. However, (in particular for 3d systems) small
<br>spheres mean that only 80-90% of the d-charge is inside the sphere and
<br>thus gets shifted by the orbital potential. Thus one needs a larger U
<br>(or alpha) to get similar results with smaller RMTs.
<br>For later 4f atoms, however, the 4f are very localized (in Tb with
<br>RMT=2.0 97% of the 4f charge is inside spheres (see case.outputst). My
<br>personal choice would be RMT = 2.1 to 2.2).
<br>
<br>Relaxation: Yes, you can safely relax the O atoms when SO is switched
<br>off for them and the heavy atoms are fixed in case.inM.
<br>If this is just a powder X-ray structure, the O-positions could be quite
<br>wrong.
<br>
<br>Most 4f systems would be anti-ferromagnets, but with very low Neel
<br>temperature, which means that the energy difference between an AFM and
<br>FM ordering is very small. These are local moments and they do not care
<br>too much how their neighbors are polarized.
<br>>
<br>> Regarding the f electron correction I opted for onsite hybrid and
<br>> initialized it with init_orb_lapw -eece.
<br>> UG says that its better to use LDA for the exchange potential so I
<br> > copied case.in0 to case.in0eece_lapw where I replaced "XC_PBE" on the
<br> > first line with "EX_PBE VX_LDA EC_PBE VC_PBE".
<br>
<br>This is a misunderstanding. I'd use PBE in case.in0 since the Ga/O
<br>states should be much better described by PBE. However, for the double
<br>counting correction, LDA is numerically preferred and the UG says:
<br> "This is possible by copying case.in0 to case.in0eece_lda and specify
<br>VX_LDA". Note: it is case.in0eece_lda, not case.in0eece_lapw
<br>
<br>EECE vs DFT+U is a matter of taste. EECE has one adjustable parameter,
<br>DFT+U 1-2 (U and J). For 4f systems the "effective U" (J=0) is often not
<br>justified since the intraatomic J may be important. It may have quite
<br>some influence on the orbital magnetic moment.
<br>Anyway, both are approximations and for a proper gap you may need mBJ+U
<br>(or mBJ+EECE) with a smaller U (alpha).
<br>
<br>> The onsite hybrid calculation converged fine, I get a nice splitting of
<br>> the f states (albeit a bit too much maybe).
<br>> The other options would be +U obviously, I went for the hybrid because
<br>> it felt more rigorous, but I would also appreciate comments if someone
<br>> has maybe better experience with +U?
<br>>
<br>> Next step was to initialize spin-orbit interaction with init_so_lapw. I
<br>> started with the default 001 but I want to also try other directions
<br>> later and compare. I opted for no relativistic LOs (no support in
<br>> optics) and enabled it only for Tb and Ga. symetso created a new
<br>> structure (most notable I have more Tb inequivalent positions) and than
<br>> I manually fixed case.inso case.indm and case.inorb as the init_so
<br>> script warned me. I also guessed I should fix case.ineece (that seemed
<br>> straightforward) but than I thought I should also fix case.in2eece.
<br>> Reading UG gives the impression that case.in2eece is normal case.in2
<br>> with extra EECE on the first line and than the optional 3a and 3b
<br>> lines. In the case.in2eece created automatically with init_orb_lapw -
<br>> eece the 3a and 3b lines looked like:
<br>> 1
<br>> 1 1 3
<br>> However reading UG this actually seems wrong? Because UG says (Section
<br>> 7.9 page 166) the format for optional 3b is just two values:
<br>> jatom rho, l rho
<br>> so I wonder if the UG is wrong or if I'm actually applying the hybrid
<br>> correction to p instead of f?
<br>
<br>This is an error in the UG. You would see it quickly, if you correct p
<br>states instead of f.
<br>
<br>PS: The only important input file is case.ineece (and maybe
<br>case.in0eece_lda). The runeece script (and not init_orb) generates the
<br>other *in*eece files on the fly in every iteration.
<br>
<br>>
<br>> Also, is there anything else I should fix manually after intializing
<br>> the so on top of eece? Or should I do it the other way around (first so
<br>> and then eece)? The reasoning for doing first eece was that I get a
<br>> metal with plain PBE and an insulator with the onsite hybrid, so I
<br>> thought it might be easier to converge if I start so from insulator
<br>> (but I still use TEMP smearing just to be sure I don't end with
<br>> convergence problems if I get a metal during the convergence as the
<br>> expected unoccupied occupied f-f distance is very small.)
<br>>
<br>> I was also considering mBJ later, just to get some feeling how the
<br>> conduction bad would shift but I'm not sure if this would work or not
<br>> on top of eece and so?
<br>>
<br>> One last question is regarding the forces. From reading the UG I
<br>> understood that it should be OK to relax the oxygen positions with
<br>> onsite hybrid and so (as long as I don't have so or eece enabled for O
<br>> atoms). Is this correct? So will just switching to MSR1a and running
<br>> normal runsp -so -eece work or are some other fixes needed?
<br>>
<br>> Best regards
<br>> Pavel
<br>> _______________________________________________
<br>> Wien mailing list
<br>> Wien@zeus.theochem.tuwien.ac.at
<br>> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
<br>> SEARCH the MAILING-LIST at: http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
<br>
<br>--
<br>--------------------------------------------------------------------------
<br>Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
<br>Phone: +43-1-58801-165300
<br>Email: peter.blaha@tuwien.ac.at WIEN2k: http://www.wien2k.at
<br>WWW: http://www.imc.tuwien.ac.at
<br>-------------------------------------------------------------------------
<br>_______________________________________________
<br>Wien mailing list
<br>Wien@zeus.theochem.tuwien.ac.at
<br>http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
<br>SEARCH the MAILING-LIST at: http://www.mail-archive.com/wien@zeus.theochem.tuwien.ac.at/index.html
<br></blockquote></body></html>