[Wien] SP+SO calculation

Tue Jul 20 19:53:09 CEST 2004

Dear Prof. Peter Blaha,

Thanks for your reply.  In my calculation, the Fe-3s and Fe-3p are treaded
as semi-core. About the case.in2(c) file, I always use the default values,
the four values in second line are: " -9.0   100.0  0.50 0.05".

I have tried to change all the possible parameters to do the SP+SO
calculation, but no help. The program cannot find the semicore states of Fe.
In the following I would like to show you the strange EPL001 and EPH001
values of Fe printed in case.scf and case.output2up files of SP+SP
calculation after the convergent SP calculation.

Could you please give me suggestions, or maybe in my case, Fe only has
valence states?

By the way, for the cases of HCP-Fe and hexagonal Fe2C (SG_ 164_P-3m1), they
are fine, no problem for the SP+SO calculation. For the cases of Fe6C1
(SG_162_P-31m), Fe6C2 (SG_182_P6322) and Fe6C3 (SG_162_P-31m), they have the
same problems: can not find the semicore states in SP+SO calculations.

Best regards,
Shunli Shang

*************************
Only SP calculation for Fe6C1 (the spin-up results for the last cycle, the
correct results):
Energy to separate semicore and valencestates:  0.23488
:FER  : F E R M I - ENERGY(TETRAH.M.)=   0.83700
:PCS001: PARTIAL CHARGES SPHERE =  1
S,P,D,F,PX,PY,PZ,D-Z2,D-X2Y2,D-XY,D-XZ,D-YZ
:QTL001: 1.0647 3.0116 3.7450 0.0032 1.0032 1.0062 1.0021 0.7337 0.7465
0.7141 0.7909 0.7597
        Q-s-low E-s-low   Q-p-low E-p-low   Q-d-low E-d-low   Q-f-low
E-f-low
:EPL001:  1.0005 -5.5980    2.9508 -3.1402    0.0108 -0.0751
   0.0004 -0.1835
        Q-s-hi  E-s-hi    Q-p-hi  E-p-hi    Q-d-hi  E-d-hi    Q-f-hi  E-f-hi
:EPH001:  0.0641  0.4686    0.0608  0.5810    3.7342  0.6751    0.0029
0.6113

*************************
SP+SO calculation (the spin-up results for the first cycle of SP+SO
calculation after the convergent SP calculation)
Energy to separate semicore and valencestates:-999.00000
:FER  : F E R M I - ENERGY(TETRAH.M.)=   0.82510
:PCS001: PARTIAL CHARGES SPHERE =  1
S,P,D,F,PX,PY,PZ,D-Z2,D-X2Y2,D-XY,D-XZ,D-YZ
:QTL001: 1.0641 3.0471 3.7042 0.0037 1.0056 0.8872 0.8226 0.7236 0.7379
0.7061 0.7854 0.7510
        Q-s-low E-s-low   Q-p-low E-p-low   Q-d-low E-d-low   Q-f-low
E-f-low
:EPL001:  0.0000 10.0000    0.0000 10.0000    0.0000 10.0000    0.0000
10.0000
        Q-s-hi  E-s-hi    Q-p-hi  E-p-hi    Q-d-hi  E-d-hi    Q-f-hi  E-f-hi
:EPH001:  1.0641 -5.2356    3.0471 -3.0663    3.7042  0.6695
   0.0037 -0.0744

*************************
SP+SO calculation (some strange results shown in case.output2up file for the
first cycle of SP+SO calculation after the convergent SP calculation).
number of occupied bands:         103
   highest energy:  0.873929976078531
  Bandranges (emin - emax):
band           1  -9.98505085982580       -5.64126602440761
band           2  -8.59789957521298       -5.64059028470795

band          49  -2.92032502323060       -5.96874600200437D-002
band          50  -2.91947964393872       -5.59045829250948D-002
band          51  -9.73140285226681D-002  0.362743687682807
band          52  -9.28189398976315D-002  0.380337011265983

*************************

----- Original Message ----- 
From: "Peter Blaha" <pblaha at zeus.theochem.tuwien.ac.at>
To: <wien at zeus.theochem.tuwien.ac.at>
Sent: Tuesday, July 20, 2004 12:37 PM
Subject: Re: [Wien] SP+SO calculation

> From your description I can see, that with  runsp -so  everything is ok
> (QTL-B values below 5 are not perfect, but not a real problem).
>
> Do you treat Fe 3p states in the bandstructure ? In that case you
definitely
> should be able to find a seperation energy.
>
> You must examin case.scf and case.output2up/dn  (bandranges). Together
> with a meaningfull case.in2 (did you change anything there in the 2nd line
?)
> it should work.
>
> Regards
>
> > (4) perform the "runsp_lapw -so" calculation.
> >
> > In each SCF cycle there are messages:
> >
> > QTL-B VALUE .EQ.    (<5)
> >
> > Energy to separate semicore and valencestates:-999.00000
> >
> > After convergence, the obtained properties seem OK (?), e.g. the HFF and
> > Total energy.
> >
> >
> >
> > (5) perform the "runsp_lapw -so -in1new 2" calculation
> >
> > In each SCF cycle there is message:
> >
> > Energy to separate semicore and valencestates:-999.00000
> >
> > The obtained HFF is unreasonable. (after 20 times SCF cycles, still no
> > convergent evidence).
> >
> >
> >
> > Could you please tell me why the energy to separate semicore and
> > valencestates is so large (-999.00)?  and could you please give me some
> > suggestions?  Please let me know if you need more information.
>
>
>
>                                       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/
> --------------------------------------------------------------------------
>
> _______________________________________________
> Wien mailing list
> Wien at zeus.theochem.tuwien.ac.at
> http://zeus.theochem.tuwien.ac.at/mailman/listinfo/wien
>
>