[Wien] Ghost Band and Spin Orbit Coupling Band Calculation for Bi

[Peter Blaha]

A qhick check of the data you sent reveals several possible problems:

> run_lapw -so
> x lapw1 -band
> x lapwso
> x spaghetti.

If you want SO bands, you need   x spaghetti -so

>  QTL-B VALUE .EQ.   10.47804   in Band of energy   - 0.41098   ATOM=    
> 1   L=  0

Yes, you have ghostbands.

>           ATOMIC SPHERE DEPENDENT PARAMETERS FOR ATOM  Bi1      
>           OVERALL ENERGY PARAMETER IS    0.3000
>           OVERALL BASIS SET ON ATOM IS LAPW
>           E( 2)=   -0.1700   E(BOTTOM)=   - 0.360   E(TOP)=    0.020
>              APW+lo
>           E( 2)=    0.3000
>              LOCAL ORBITAL
>           E( 0)=    0.1200   E(BOTTOM)=   -0.660   E(TOP)=    0.900
>              APW+lo
>           E( 0)=    0.0005
>              LOCAL ORBITAL
>           E( 1)=    0.3000
>              APW+lo
> 
>        K=   0.05000   0.05000   0.05000            1
> :RKM  : MATRIX SIZE  146LOs:  30  RKM= 6.85  WEIGHT= 2.00  PGR:   
>        EIGENVALUES ARE:
>         -0.4106008   -0.2248095   -0.2248095   -0.1990911   -0.1990911
>         -0.1972398   -0.1132902   -0.1053402   -0.1053402   -0.0790426
>         -0.0790426    0.2221233    0.3902385    0.4176217    0.4176217
>          0.6946178    0.7236864    0.7249391     0.7249391    0.8138934
>          0.8557083    0.8557083    0.9123181    1.0734300    1.0734300
>          1.6118069    1.6118069    1.6974122    1.8054132    1.8605886
>          1.8605886    1.9831581    1.9831581    2.0456323    2.1117763
>          2.1117763    2.2090826    2.2090826    2.2516513    2.2673514
> 
>          2.3173559     2.3952378    2.3952378    2.4165511    2.4623928
>          2.4770960    2.4770960    2.5534038    2.6654239    2.6654239
>          2.6758456    2.7499461    2.7794123    2.7879459    2.7879459
>          2.8609063    2.8609063    2.9093091     2.9367524    2.9367524
>          2.9973414    3.0675987    3.0675987    3.1607611    3.1607611
>          3.1626438    3.2016070    3.2016070    3.3103052
>        ********************************************************
> 
>        NUMBER OF K-POINTS:         110
>    0.0   0.0 angle (M,z), angle (M,x) deg
>  RELATIVISTIC LOs:
>  on atom            1 e(           1 )=  -5.82749999999998    
>  
>  
>         SPIN-ORBIT EIGENVALUES:
>      K=   0.05000   0.05000   0.05000            1
>       MATRIX SIZE=  150   WEIGHT= 2.00
>      EIGENVALUES ARE:
>           -5.8257559   -5.8253744   -5.8245165   - 5.8241350   -3.2827295
>           -3.2825076   -3.2819354   -3.2817976   -3.2815757   -3.2811225
>           -3.2803820   -3.2795692   -0.4105841   -0.4105841   -0.2629474
>           -0.2629474   -0.2604524   -0.2604524   -0.2526100   -0.2526100
>           -0.2503525   -0.2503525   -0.1507602   -0.1507602   -0.1087296
>           - 0.1087295   -0.0995931   -0.0995931   -0.0554644   -0.0554644
>           -0.0529689   -0.0529689   -0.0488590   -0.0488590    0.2221748
>            0.2221751    0.3502867    0.3502903    0.4424264    0.4424298
> 
>            0.4641082    0.4641102    0.6621771    0.6621804    0.7122195
>            0.7122199    0.7509970    0.7509991    0.7635048    0.7635063
>            0.8133581    0.8133582    0.8491332    0.8491333    0.8691713
>            0.8691713    0.9148027    0.9148030     1.0743300    1.0743301
>            1.0751390    1.0751390
>        ********************************************************

When you compare no-so and so-eigenvalues, you see big changes and 
eigenvalues appear at -5 to -3 eV. This cannot come from the effect of 
SO !!!

>  I have increased energy parameter of the L=0 level in atom 1 to 5, but 
> that doesn't seem to solve the problem.
> My .in1 file is given by:
> WFFIL        (WFPRI, SUPWF)
>   7.00       10    4 (R-MT*K-MAX; MAX L IN WF, V-NMT
>   0.30    5  0      (GLOBAL E-PARAMETER WITH n OTHER CHOICES, global 
> APW/LAPW)
>  2   -1.54      0.010 CONT 1
>  2    0.30      0.000 CONT 1
>  0   -0.73      0.010 CONT 1
>  0    5    0.000 CONT 1

This file is fixedformat, so the line above is not valid!!!

> WFFIL
>  4  1  0                      llmax,ipr,kpot
>  -10.0000   1.50000           emin,emax (output energy window)
>    1.  1.  1.                 direction of magnetization (lattice vectors)
>  1                           number of atoms for which RLO is added
>  1   -4.97      0.005       atom number,e-lo,de (case.in1), repeat NX times

You add a p-RLO at - 4.97 Ry !    Why ??
Do you have a p "semicore state, i.e. in case.in1 a p (l=1) LO ?
NO.
So remove this and set NLO-number to zero.

Restart from non-SO results (provided they are ok).