[QE-users] About electron-phonon coupling matrix elements

[Han Yang]

Dear all,

I have been using Phonon package ( version 6.1) to compute electron-phonon coupling matrix elements. Following example 3 in Phonon package, I computed the elements of Al(bcc) and the elements are written in _ph0/al.phsave/elph.*.*.xml. According to phonon user’s guide and developer’s manual, the elements in these files are coupling matrix in the basis of normal mode, ie g_{q nu} and it has been scaled to phonon frequencies and masses. However, if I manually change masses in Phonon input file, the elements do not change. So, are you sure the elements are in normal mode basis and are scaled to masses? Also, what is the unit for the elements written in those files? I assume it is Rydberg?

Another question is about the elements at k=gamma and q=gamma. Because the el-ph coupling matrix elements are computed with < psi_mk+q | dvscf | psi_nk >. At k=gamma and q=gamma, I believe these quantities should be real numbers, right? However, I get complex numbers for Al at k=gamma, q=gamma. Here are first several elements:

         <K_POINT.1>
         <COORDINATES_XK type="real" size="3" columns="3">
    0.000000000000000E+000  0.000000000000000E+000  0.000000000000000E+000
         </COORDINATES_XK>
         <PARTIAL_ELPH type="complex" size="100">
    1.173453794238444E-007, 5.551115123125783E-017
   -2.279566019787527E-002, 3.586239646544477E-001
   -4.862786020311256E-001, 2.450200129592524E-001
    3.623894552425257E+000,-5.173393946467373E-001
   -6.097101185043030E-009, 1.009496251869191E-008
    2.988349845356453E-010, 9.034043582656359E-010
    2.767334139619493E-010, 1.470082250376592E-010

Do I make any mistake in computing or interpreting the results? 

Here’s my input file for phonon calculation:
 &inputph
      tr2_ph=1.0d-10,
      prefix='al',
      fildvscf='aldv',
      amass(1)=26.98,
      fildyn='./_ph0/al.dyn',
      outdir='./',
      electron_phonon='yambo',
      el_ph_sigma=0.005,
     el_ph_nsigma=10,
     asr = .true.
     zue = .false.
     epsil = .false.
     trans=.true.,
     ldisp=.true.
     nq1=5, nq2=5, nq3=5
  /

Thank you very much!

Han.
University of Chicago
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20180824/8c789874/attachment.html>