[Wien] lapw2 error in cycle 20 : 'l2main' - QTL-B.GT.15., ghostbands, check scf files

[Donghui Guo]

Dear all,

Now I am calculating a (1x1x3) supercell with large vaccum. After many iterations, I got a error as mentioned in subject.

I know it is a well disscussed problem but I still not sure how to deal with it because the situation is kind of different.

First of all, I knew it might happen in case of "spheres are 'extremely' small or you have heavy elements".  It is also the problem puzzled me. In case of Ni2P bulk, the RMT for Ni is 2.16 bohr and that for phosphorus is 1.91 bohr. It works well. But after I made slab by "x supercell", the RMT values I can set decreased to 1.74 bohr for Ni and 1.90 for P, separately. I am sure that the vacuum is big enough and I still cannot increase the RMTs due to NN calculation .

************************
The error message shows in lapw2.error is:
'l2main'  - QTL-B.GT.15., ghostbands, check scf files

***************************
case.scf2 shows:

   QTL-B VALUE .EQ.  134.20549   in Band of energy    0.39310   ATOM=   11   L=  1
    Check for ghostbands or EIGENVALUES BELOW XX messages
    Adjust your Energy-parameters or use -in1new switch, check RMTs  !!!


:WARN : QTL-B value eq. 134.21  in Band of energy   0.39310   ATOM=   11   L=  1
:WARN : You should change the E-parameter in case.in1 or use -in1new switch

************************************
I found it in FAQ that the most proper way should be rerun the run_lapw with the -in1new switch. Is it means I should set the -in1new switch to about 0.2 ev below fermi energy and rerun run_lapw? If so, the problem is that the fermi energy is located around 0.48, I have no idea how to increase the energy parameter from 0.3?  

Any suggestion is appreciated.

Thanks in advance.

****************************************
Code version: WIen2k_08.3
Platform: openSUSE 10.3
MKL: 10.0.010
*******************************************

Best regards, 
Yours, 
Donghui Guo
2008-12-11