[Pw_forum] nonzero dos at Fermi level

[Nguyen Doan Sau]

Hi there, 
I have a problem related to DOS at fermi level. Let me start with my scf file first
-------------
 &system
    ibrav=0, celldm(1)=0 , nat=22, ntyp=4,
      ecutwfc =37.0, occupations='smearing', smearing='gauss', degauss=0.01, nelec=156
    nspin= 1,
 /
 &electrons
  electron_maxstep= 300
    mixing_mode = 'plain'
    mixing_beta = 0.2
    conv_thr =  1.0d-6
    diagonalization= 'david'
 /
--------------
My pseudopotentials are norm conversing, which are generated by  ld1.x and their transferability is OK.
here is the output
----------
the Fermi energy is     3.4391 ev

!    total energy              =   -1213.21835062 Ry
     Harris-Foulkes estimate   =   -1213.21835085 Ry
     estimated scf accuracy    <       0.00000034 Ry

     The total energy is the sum of the following terms:

     one-electron contribution =    -711.80653324 Ry
     hartree contribution      =     352.57845405 Ry
     xc contribution           =    -222.78424774 Ry
     ewald contribution        =    -631.18916691 Ry
     smearing contrib. (-TS)   =      -0.01685678 Ry

     convergence has been achieved in  32 iterations
----------------------
and Dos
3.000  0.2551E+02  0.1178E+03
  3.100  0.2931E+02  0.1207E+03
  3.200  0.3391E+02  0.1241E+03
  3.300  0.3375E+02  0.1275E+03
  3.400  0.2724E+02  0.1302E+03
  3.500  0.2299E+02  0.1325E+03
  3.600  0.2335E+02  0.1348E+03
  3.700  0.2043E+02  0.1369E+03
  3.800  0.1297E+02  0.1381E+03
  3.900  0.9369E+01  0.1391E+03
  4.000  0.8969E+01  0.1400E+03

--------------
My question are concerned to:
+ DOS is still not zero at fermi level E=3.4391
+ Integrated DOS is not equal number of electron (156) , even I try DeltaE=0.001
-------------
I have thought about the question
+First, My compound is insulator, an oxide with optical bandgap of 4 eV. But in scf input, I use occupations='smearing'. Shoud I change it to occupations='fixed' and increase nbnd>156/2 and delete smearing variable.
------------
Is there anyone had this problem before and please help me to fix that.

Sau Nguyen
U of Houston



-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20100827/5bf49c56/attachment.html>