[Wien] Problems ... when calculating the effective U

[Karel Vyborny]

Dear Yu,
just've noticed the message below... Are you still working with DFT+U for 
Fe3Al? I don't think you should take the remark of Laurence Marks (about 
using +U only for insulators) as a statement of "absolute validity",  my 
own experience is that metals may need it too:

https://journals.aps.org/prb/abstract/10.1103/PhysRevB.97.125109

Even if CuMnAs is a system in some sense "not too far" from an insulator,
I'd be curious to know what conclusion can be drawn for a good metal such 
as Fe3Al.

Kind regards,

Karel


PS: Most likely, if you want to do better than DFT for Fe3Al there would 
be other ways than DFT+U...

--- x ---
dr. Karel Vyborny
Fyzikalni ustav AV CR, v.v.i.
Cukrovarnicka 10
Praha 6, CZ-16253
tel: +420220318459
web: https://spintronics.fzu.cz
      http://unix12.fzu.cz/~vybornyk


On Sun, 5 Feb 2023, 晨晨 via Wien wrote:

> 
> Dear WIEN2k experts and users,
> 
> I am a beginner of WIEN2k. The version is WIEN2k_21.1. Now, I am learning to
> calculate the effective U of Fe in Fe3Al, following the note ??Calculating
> the effective U in APW methods. NiO?? by Georg K.H. Madsen and Pavel Novak.
> When runsp the constrained calculation, the following messages and error
> appeared:
> 
> LAPW0 END
> 
> LAPW1 END
> 
> LAPW1 END
> 
> LAPW2 END
> 
> LAPW2 END
> 
> CORE  END
> 
> CORE  END
> 
> ERROR: NEC01 charge leakage too large
> 
> Could you please help me to check which part of my modification went wrong. 
> Thank you for your reply in advance.
> 
> The specific procedures and related files are as follows:
> 
> 1. The supercell of Fe3Al has been modified the first Fe. Then run x sgroup
> and copy case.struct_sgroup to case.struct.
> 
> 2. Modify inst so that the part of the first Fe looks like:
> 
>   Fe
> 
> Ar 3
> 
> 3, 2,2.0  P
> 
> 3, 2,2.0  P
> 
> 3,-3,2.5  P
> 
> 3,-3,0.0  P
> 
> 4,-1,1.0  N
> 
> 4,-1,0.5  N
> 
> Then run the following commands:
> 
>   x lstart
> 
>   x dstart
> 
>   x dstart ?Cup
> 
>   x dstart ?Cdn
> 
> runsp
> 
> It's working fine up here.
> 
> 3. Set two calculations. One with 3.5 (up), 3 (dn). The other with 3.5 (up),
> 2 (dn). Make the bellowing 4 .inc files. And modified case.in1 and case.in2.
> 
> case.incup_+.50:
> 
>   7 1.00  0  NUMBER OF ORBITALS (EXCLUDING SPIN), SHIFT, IPRINT
> 
> 1,-1,2               ( N,KAPPA,OCCUP)
> 
> 2,-1,2               ( N,KAPPA,OCCUP)
> 
> 2, 1,2               ( N,KAPPA,OCCUP)
> 
> 2,-2,4               ( N,KAPPA,OCCUP)
> 
> 3,-1,2               ( N,KAPPA,OCCUP)
> 
> 3, 2,4               ( N,KAPPA,OCCUP)
> 
> 3,-3,3               ( N,KAPPA,OCCUP)
> 
>  
> 
> case.incdn_+.50:
> 
> 7 1.00  0  NUMBER OF ORBITALS (EXCLUDING SPIN), SHIFT, IPRINT
> 
> (The first 5 lines are the same as the file above)
> 
> 3, 2,4               ( N,KAPPA,OCCUP)
> 
> 3,-3,2               ( N,KAPPA,OCCUP)
> 
>  
> 
> case.incup_+.5-1:
> 
> 7 1.00  0  NUMBER OF ORBITALS (EXCLUDING SPIN), SHIFT, IPRINT
> 
> (The first 5 lines are the same as the file above)
> 
> 3, 2,4               ( N,KAPPA,OCCUP)
> 
> 3,-3,3               ( N,KAPPA,OCCUP)
> 
>  
> 
> case.incdn_+.5-1:
> 
> 6 1.00  0  NUMBER OF ORBITALS (EXCLUDING SPIN), SHIFT, IPRINT
> 
> (The first 5 lines are the same as the file above)
> 
> 3, 2,4               ( N,KAPPA,OCCUP)
> 
>  
> 
> Corresponding part of case.in1:
> 
>   0.30    4  0      (GLOBAL E-PARAMETER WITH n OTHER CHOICES, global
> APW/LAPW)
> 
>  1    0.30     0.0000 CONT 1
> 
>  1   -3.97     0.0001 STOP 1
> 
>  2   20.30     0.0010 CONT 1
> 
>  0    0.30     0.0000 CONT 1
> 
>  
> 
> The Original Fe3Al supercell had 408 valence electrons. I put 6.5 and 5.5
> electrons into the core.
> 
> Corresponding part of case.in2_+.50:
> 
> TOT             (TOT,FOR,QTL,EFG,FERMI)
> 
>    -12.0   401.5   0.50 0.05  1   EMIN, NE, ESEPERMIN, ESEPER0, iqtlsave
> 
>  
> 
> Corresponding part of case.in2_+.5-1:
> 
> TOT             (TOT,FOR,QTL,EFG,FERMI)
> 
>    -12.0   402.5   0.50 0.05  1   EMIN, NE, ESEPERMIN, ESEPER0, iqtlsave
> 
>  
> 
> Then Run the one constrained calulations.
> 
>  cp case.in2_+.50 case.in2
> 
>  cp case.incup_+.50 case.incup
> 
> cp case.incdn_+.50 case.incdn
> 
> runsp
> 
> Then the following messages and error appeared here:
> 
> LAPW0 END
> 
> LAPW1 END
> 
> LAPW1 END
> 
> LAPW2 END
> 
> LAPW2 END
> 
> CORE  END
> 
> CORE  END
> 
> ERROR: NEC01 charge leakage too large
> 
>  
> 
> Run the other constrained calulations.
> 
>  cp case.in2_+.5-1 case.in2
> 
>  cp case.incup_+.5-1 case.incup
> 
> cp case.incdn_+.5-1 case.incdn
> 
> runsp
> 
> For the _+.5-1, the same thing happens.
> 
>  
> 
> I am looking forward your reply. Thank you in advance.
> 
> Best regards,
> 
> Yu
> 
> 
>