The ABINIT Users Mailing List ( CLOSED )

[Fabien Bruneval <fabien.bruneval@polytechnique.fr>]

Hi!

You are perfectly right: the small imaginary part in the denominator for 
chi0 was added to avoid divergences when calculating chi0 for real 
frequencies.
Since the poles of chi0 are located along the real axis, this is true 
that you can set the hardcoded variable etadelta to 0, when you 
calculate chi0 for imaginary frequencies (as it is the case in the 
plasmon-pole models).

This small imaginary part is necessary for the calculation without 
plasmon-pole model. But furthermore, I thought this was completely 
harmless for the calculations with plasmon-pole model, as the built-in 
tests of abinit were all passed successfully by the new code 5.1.x.

Please tell me if you have other questions regarding the new implementation.


Fabien




dylu@ucdavis.edu wrote:
> Hello, I did some tests on the dielectric matrix calculation of bulk silicon recently, and found that the results from Abinit 5.1.2 and Abinit 4.6.5 are 
> not consistent. Taking the results from lesson 1 of the GW tutorial for example,
> chi0 and epsilon^-1 computed from Abinit 5.1.2 are not Hermitian, while those 
> from Abinit 4.6.5 are (input file attached below).
>
> The disagreement comes from i*eta introduced in the denominator of chi0 as a 
> broadening factor in routines cchi0.F90 and cchi0q0.F90 of abinit 5.1.2. In 
> the code, etadelta=0.1 eV contributing an imaginary part of -0.002 for 
> diagonal matrix elements of epsilon^-1(G,G',q=0,omega=0). Typical changes in 
> other elements are
> in the order of 0.001.
>
> I assume "etadelta" may relate to the new implemented plasmon-pole models or 
> non-plasmon-pole model in the higher version. Is there any other considerations?
> Is it all right to simply not using "etadelta" at all in calculating 
> epsilon^-1 with abinit 5.1.2? I could actually reproduce the results in 4.6.5 by seting 
> etadelta=0.
>
> Thanks
> Deyu Lu
>
> Below is my input:
> ndtset      2
>
> # Definition of parameters for the calculation of the KSS file
> nbandkss1   -1         # Number of bands in KSS file (-1 means the maximum 
> possible)
> nband1      9         # Number of (occ and empty) bands to be computed
> istwfk1     10*1
>
> # Calculation of the screening (epsilon^-1 matrix)
> optdriver2  3        # Screening calculation
> getkss2     -1       # Obtain KSS file from previous dataset
> nband2      17       # Bands to be used in the screening calculation
> ecutwfn2    2.1      # Cut-off energy of the planewave set to represent the 
> wavefunctions
> ecuteps2    3.6      # Cut-off energy of the planewave set to represent the 
> dielectric matrix
> ppmfrq2    16.7 eV  # Imaginary frequency where to calculate the screening
>
> # Data common to the three different datasets
>
> # Definition of the unit cell: fcc
> acell  3*10.217        # This is equivalent to   10.217 10.217 10.217
> rprim  0.0  0.5  0.5   # FCC primitive vectors (to be scaled by acell)
>        0.5  0.0  0.5
>        0.5  0.5  0.0
>
> # Definition of the atom types
> ntypat  1         # There is only one type of atom
> znucl 14          # The keyword "znucl" refers to the atomic number of the
>                   # possible type(s) of atom. The pseudopotential(s)
>                   # mentioned in the "files" file must correspond
>                   # to the type(s) of atom. Here, the only type is Silicon.
>
> # Definition of the atoms
> natom 2           # There are two atoms
> typat  1 1        # They both are of type 1, that is, Silicon.
> xred              # Reduced coordinate of atoms
>       0.0  0.0  0.0
>       0.25 0.25 0.25
>
>
> # Definition of the k-point grid
> kptopt  1            # Option for the automatic generation of k points,
> nkpt    10
> ngkpt   4 4 4
> nshiftk 4
> shiftk  0.5 0.5 0.5  # These shifts will be the same for all grids
>         0.5 0.0 0.0
>         0.0 0.5 0.0
>         0.0 0.0 0.5
>
> # Use only symmorphic operations
> symmorphi 0
>
> # Definition of the planewave basis set (at convergence 16 Rydberg 8 Hartree)
> ecut 8.0          # Maximal kinetic energy cut-off, in Hartree
>
> # Definition of the SCF procedure
> nstep   10        # Maximal number of SCF cycles
> toldfe  1.0d-6    # Will stop when this tolerance is achieved on total energy
> diemac  12.0      # Although this is not mandatory, it is worth to
>                   # precondition the SCF cycle. The model dielectric
>                   # function used as the standard preconditioner
>                   # is described in the "dielng" input variable section.
>                   # Here, we follow the prescription for bulk silicon.