[Wien] Ghost band come from large charge fluctuation

liyh lyhua at fudan.edu.cn
Mon Feb 28 13:59:11 CET 2005


Dear wien user,

        I am calculating the large superlattices.
Now I meet a problem with ghost bands which caused by large charge fulctuation.
The superlattices are build in 1x1 in x and y direction, but the scale in z direction are 
very large, up to 60 atom layers (about 94 angstroms), the total atoms in the superlattice are 
about 80.
For si there are no porblem, when we use pratt mixing scheme in the first iterations, and set the 
mixing parameter to 0.01.
But for some polar materials this method failed. Because the semiconductor is polar and there are 
an interface in this superlattice, so the charge fulaction are so large, and get a ghost bands in the 
second iteration. We still can not avoid this problem even we set the mixing parameter to 0.00001.
I know the initial energy prameter and charge density are come from atoms, so there are no problem for 
the first iteration. But the charge fulaction between the first and second iteration is too large if we use 
a polar interface, then the ghost bands appeared.
these are the energy parameters and the calculated eigenvalue in the second iteration.
 ______________________________________________________________
               .................        
          ATOMIC SPHERE DEPENDENT PARAMETERS FOR ATOM  Ga
          OVERALL ENERGY PARAMETER IS   -0.0007
          OVERALL BASIS SET ON ATOM IS LAPW
          E( 0)=   -0.1030
             APW+lo
          E( 0)=   -0.5440
             LOCAL ORBITAL
          E( 1)=    0.1200
             APW+lo
          E( 1)=   -0.5300
             LOCAL ORBITAL
          E( 2)=    0.1200
             APW+lo
          E( 2)=   -0.7130
             LOCAL ORBITAL 
..................................
          ATOMIC SPHERE DEPENDENT PARAMETERS FOR ATOM  As
          OVERALL ENERGY PARAMETER IS   -0.0007
          OVERALL BASIS SET ON ATOM IS LAPW
          E( 0)=   -0.0420
             APW+lo
          E( 0)=   -0.5320
             LOCAL ORBITAL
          E( 1)=    0.0940
             APW+lo
          E( 2)=    0.0470
             APW+lo
          E( 2)=   -2.2980
             LOCAL ORBITAL
............................................
       K=   0.00000   0.00000   0.00000            1
:RKM  : MATRIX SIZE 4429LOs: 792  RKM= 7.00  WEIGHT= 1.00  PGR:
       EIGENVALUES ARE:
        -2.7433474   -2.7431768   -2.7431768   -2.7420130   -2.7420130
        -2.7326936   -2.7324469   -2.7324469   -2.7312467   -2.7312467
        -2.7300525   -2.7300085   -2.7300085   -2.7287968   -2.7287967
        -2.6952517   -2.6950644   -2.6950644   -2.6939024   -2.6939024
        -2.6915573   -2.6914734   -2.6914734   -2.6902141   -2.6902141
        -2.6380089   -2.6378341   -2.6378341   -2.6366109   -2.6366109  <-----As 3d(-2.7--- -1.7)
        -2.6307358   -2.6306405   -2.6306405   -2.6294444   -2.6294444        El=-2.29
        -2.5701874   -2.5699363   -2.5699363   -2.5687394   -2.5687394

        -2.5448468   -2.5447474   -2.5447474   -2.5435526   -2.5435526
        -2.4752486   -2.4752021   -2.4752021   -2.4739306   -2.4739306
        -2.4330375   -2.4327828   -2.4327828   -2.4316496   -2.4316496
        -2.3610463   -2.3610384   -2.3610384   -2.3597561   -2.3597561
        -2.3016877   -2.3016785   -2.3016785   -2.3004509   -2.3004509
        -2.2449789   -2.2448014   -2.2448014   -2.2435894   -2.2435894
        -2.1724769   -2.1724326   -2.1724326   -2.1712203   -2.1712203
        -2.1199731   -2.1199391   -2.1199391   -2.1186647   -2.1186647

        -2.0367174   -2.0365472   -2.0365472   -2.0353862   -2.0353861
        -1.9885085   -1.9885085   -1.9883856   -1.9871598   -1.9871598
        -1.9134396   -1.9134396   -1.9133645   -1.9121696   -1.9121696
        -1.8832318   -1.8831874   -1.8831874   -1.8819215   -1.8819215
        -1.8256205   -1.8256126   -1.8256126   -1.8243857   -1.8243857
        -1.8090128   -1.8089888   -1.8089888   -1.8077115   -1.8077115
        -1.7640895   -1.7640810   -1.7640810   -1.7628624   -1.7628624
        -1.7597053   -1.7597053   -1.7595772   -1.7583481   -1.7583481

        -1.1803019   -1.1745726   -1.1728348   -1.1728348   -1.1689268
        -1.1689268   -1.1661294   -1.1661294   -1.1619556   -1.1619556
        -1.1603716   -1.1545904   -1.1545904   -1.1481634   -1.1481634
        -1.1405540   -1.1319026   -1.1319026   -1.1253180   -1.1253180<------------Ga 3d(-1.17---- -0.53)
        -1.1153949   -1.1097725   -1.1097724   -1.1032379   -1.1032379                El=-0.71
        -1.0873363   -1.0774549   -1.0774549   -1.0746084   -1.0707801
        -1.0707801   -1.0436705   -1.0409026   -1.0357421   -1.0357421
        -1.0291190   -1.0291190   -1.0228871   -1.0123914   -1.0123914

        -1.0112487   -1.0058416   -1.0058416   -0.9813694   -0.9660938
        -0.9467778   -0.9412180   -0.9412180   -0.9364906   -0.9346377
        -0.9346377   -0.9227757   -0.9227757   -0.9193048   -0.9163658
        -0.9163658   -0.9101156   -0.8543591   -0.8276364   -0.8205872
................................................
I think in this situation it is not good to use the atoms charge density as 
initail charge density, but I dont know how to do with it. Are there 
any body have some suggest with it?
 
Yonghua

Fudan university of China





More information about the Wien mailing list