[Pw_forum] Fermi level determination

Gabriele Sclauzero sclauzer at sissa.it
Mon Jan 11 10:34:35 CET 2010



If you are treating your system as an insulator (i.e. no smearing), there will be no Fermi 
energy, of course.
If you need to find the band gap you should look at the DOS and its integral: the energy 
at which the integral equals the number of valence electrons corresponds to the highest 
occupied state (top of the valence band).
In order to see the conduction band in your DOS you may need to increase the value of nbnd 
from its default (nelec/2 for insulators).

HTH

GS

Vit wrote:
> As I've mentioned above there is no Fermi energies information in  
> *.scf.out and *.dos.out (nscf calculation).
> Are there any other way? Or any other suggestions?
> Here is the end of scf.out file to be sure.
> 
> 	k =-0.0821-0.1560-0.0775 ( 2356 PWs) bands (ev):
> 
> 	 -3.0847 -3.0847 -2.6903 -2.6903 -2.3638 -2.3638 3.4750 3.4750
> 	 3.9668 3.9668 4.8754 4.8754 5.6593 5.6593 6.0363 6.0363
> 	 6.6648 6.6648 6.8761 6.8761 7.0072 7.0072 7.4343 7.4343
> 	 8.1220 8.1220 9.0760 9.0760 11.3965 11.3965
> 
> 	! total energy = -240.51977828 Ry
> 	 Harris-Foulkes estimate = -240.51977829 Ry
> 	 estimated scf accuracy < 9.1E-09 Ry
> 
> 	 The total energy is the sum of the following terms:
> 
> 	 one-electron contribution = 51.57190027 Ry
> 	 hartree contribution = 17.06985894 Ry
> 	 xc contribution = -102.98259683 Ry
> 	 ewald contribution = -206.17894067 Ry
> 
> 	 convergence has been achieved in 30 iterations
> 
> 	 Writing output data file Mo2S3.save
> 
> 	 PWSCF : 6h45m CPU time, 7h21m wall time
> 
> With best wishes,
> Koroteev Victor
> PhD student, NIIC
> 
> 
> 
>> Vit wrote:
>>> Dear all,
>>> I'm running calculation of Mo2S3 layers and need to determine Fermi
>>> level. I'm running DOS calculation and there is no info about Fermi
>>> level in dos.out file. What should I do?
>> Use the Fermi energies reported at the end of the scf or nscf PWscf  
>> calculation you
>> performed before running dos.x. If your calculations are converged  
>> w.r.t. to k-point
>> sampling, scf and nscf values of E_F should reasonably agree with each  
>> other.
>>
>> GS
>>
>>> In file is attached.
>>> With best regards,
>>> Koroteev Victor
>>> PhD student NIIC
>>>
>>>
>>> ------------------------------------------------------------------------
>>>
>>> _______________________________________________
>>> Pw_forum mailing list
>>> Pw_forum at pwscf.org
>>> http://www.democritos.it/mailman/listinfo/pw_forum
> 
> 

-- 


o ------------------------------------------------ o
| Gabriele Sclauzero, PhD Student                  |
| c/o:   SISSA & CNR-INFM Democritos,              |
|        via Beirut 2-4, 34014 Trieste (Italy)     |
| email: sclauzer at sissa.it                         |
| phone: +39 040 3787 511                          |
| skype: gurlonotturno                             |
o ------------------------------------------------ o



More information about the users mailing list