[Wien] [SPAM?] LSDA + U calculation

[Peter Blaha]

a) What result do you get for GGA ? The first thing is to try to reproduce the published
calculations. You should have large moments at Cr (2 mB) and smaller negative ones (0.7 uB) on Os.

b) Are you sure you entered the structure correctly ? This is a R-spacegroup and WIEN2k requires
    to enter the positions in R coordinates.

c) For which atoms did you put U ??? For sure you should put it to Cr (larger U ?) and eventually to
    Os (smaller U, like the 0.15 Ry in your input).

d) I hope, you are using a new WIEN2k version, because in some older version there was a bug for
    LDA+U with complex vorb components.

e) Otherwise, I can only advise you to be systematic like:

spin-pol GGA calc.
save
initso
runsp -so
save    (and check your results)
add indm(c), inorb
runsp -so -orb

eventually, you can also try once to skip the -so step and do everything without SO (but Os is quite heavy).

Without doing the calculations I can hardly say anything, except that I would not understand how Cr
can loose its moment in LDA+U.

> To clarify in more detail, I reproduce Sr2CrOsO6 calculated using FPLO program.
>
> In experiment, this material is ferrimagnet(FI) insulating state which is revised from the previous question.
>
> In my calculation for LSDA, I got a FI metallic state. And then to got a FI insulating state, I calculated +U and SOC. Unfortunately, I got the results that are non-magnetic
> metallic state.
>
> This is the method that I did.
>
> For +U, I copied the files, inorb and indm.
>
> For inorb file
>
> -----------------------------------------------------
>
> 120nmod, natorb, ipr
> PRATT1.0BROYD/PRATT, mixing
> 2 1 2iatom nlorb, lorb
> 5 1 2iatom nlorb, lorb
> 1nsic 0..AFM, 1..SIC, 2..HFM
> 0.1471 0.0 U J (Ry)Note: we recommend to use U_eff = U-J and J=0
> 0.1471 0.0U J (Ry)Note: we recommend to use U_eff = U-J and J=0
> -------------------------------------------------------------------
>
> For indm file
>
> --------------------------------------------------------------------
>
> -9.Emin cutoff energy
> 2number of atoms for which density matrix is calculated
> 212index of 1st atom, number of L's, L1
> 512index of 1st atom, number of L's, L1
> 0 0r-index, (l,s)index
> --------------------------------------------------------------------
>
> $ runsp_lapw –orb –ec 0.00001 –cc 0.00001
>
> For SOC,
>
> First, I did initso.
>
> 1.Spin direction : Z-direction
>
> 2.Which atom : Every atoms
>
> 3.Emax : 5Ry
>
> 4.RLO : No
>
> 5.Spin polarized case : Yes
>
> 6.SOC : Yes
>
> 7.K-point : Default
>
> 8.Kgen : No
>
> Finally, $runsp_lapw –so –ec 0.00001 –cc 0.00001.
>
> In my calculation, that’s all.
>
> So, my questions are
>
> 1.Is my calculation process right?
>
> 2.In SOC case, Should I consider the RLO option?
>
> 3.In +U case, is there another considering option for 5d-system?
>
> Except for my questions, can you give another advice to me?
>
> Thank you for your help, in advance.
>
>
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien

-- 

                                       P.Blaha
--------------------------------------------------------------------------
Peter BLAHA, Inst.f. Materials Chemistry, TU Vienna, A-1060 Vienna
Phone: +43-1-58801-15671             FAX: +43-1-58801-15698
Email: blaha at theochem.tuwien.ac.at    WWW: http://info.tuwien.ac.at/theochem/
--------------------------------------------------------------------------