[Pw_forum] HSE output for yambo TDDFT calculation

[Michael Friedrich]

Dear QE users,

I hope there is anybody who has experience with the output generation for yambo in case of hybrid functionals, especially for TDDFT.
Usually one does a scf and a subsequent nscf calculation in pw.x. The nscf calculation, however, is not possible in the case of hse. So I’m using the output from the scf calculation with the input posted below.
In principle, the developers from yambo don’t see a problem doing a TDDFT calculation on top of hybrid functional calculation but have either not yet really tried it.
My problem is that the output does apparently not provide all input for yambo that is needed as information on the reciprocal lattice vectors is missing (RL vectors        (WF):0)
 <--->  :: Electrons             : 68.00000
 <--->  :: Temperature       [ev]: 0.000000
 <--->  :: Lattice factors [a.u.]: 8.757014  8.757014  8.757014
 <--->  :: K-points              :  38
 <--->  :: Bands                 :  60
 <--->  :: Spinor components     : 1
 <--->  :: Spin polarizations    : 1
 <--->  :: Spin orbit coupling   : no
 <--->  :: Symmetries   [spatial]: 3
 <--->  ::                [T-rev]: yes
 <--->  :: Max WF components     :  8469
 <--->  :: RL vectors        (WF):0
 <--->  :: RL vectors    (CHARGE):  67681
 <--->  :: XC potential          : HSE06(XC)
 <--->  :: Atomic species        : 3
 <--->  :: Max atoms/species     :  6

which is probably the reason for the tddft run to stop, though p2y works without complaining.
Here’s the input I used in pw.x:

 &control
    calculation = 'scf'
    restart_mode='from_scratch',
    verbosity='high'
    prefix='LN',
    pseudo_dir = '/scratch/michaelf/work/input/qe/',
    outdir='.'
    wf_collect= .true. 
 /
 &system    
    ibrav=  0, 
    nat=  10, 
    ntyp= 3,
    ecutwfc =78.0,
    nbnd = 60,
    input_dft='hse',
    nqx1 = 1,
    nqx2 = 1,
    nqx3 = 1, 
    force_symmorphic=.true.
 /
 &electrons
!    mixing_beta = 0.7 ,
    conv_thr =  1.0d-10,
    electron_maxstep = 500
 /
ATOMIC_SPECIES
Li   2.0    03_Li_ONCV.UPF
Nb   2.0    41_Nb_ONCV.UPF
O    2.0    08_O_ONCV.UPF
ATOMIC_POSITIONS crystal
Li       0.283300360   0.283300360   0.283300360
Nb       0.000000000   0.000000000   0.000000000
O        0.855421834   0.616149204   0.219876866
O        0.219876866   0.855421834   0.616149204
O        0.616149204   0.219876866   0.855421834
Li       0.783300360   0.783300360   0.783300360
Nb       0.500000000   0.500000000   0.500000000
O        0.116149204   0.355421834   0.719876866
O        0.719876866   0.116149204   0.355421834
O        0.355421834   0.719876866   0.116149204
K_POINTS automatic
6 6 6  1 1 1
CELL_PARAMETERS bohr
   5.647324820  -0.000000000   8.757014575
  -2.823662411   4.890726758   8.757014575
  -2.823662411  -4.890726758   8.757014575

Maybe there is someone who knows more about this? Is there a possibility to provide the same output to yambo from a hse run as from a scf—>nscf DFT run? As far as I know there shouldn’t be much of a difference in the format of the WF and in principle it should be possible?!

Thank you
Michael Friedrich
Universität Paderborn
Germany
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20151209/3b086b4b/attachment.html>