[Pw_forum] electron-phonon interation: matrix element

[Vi Vo]

Dear All,

I had question before, but my question is probably is not clear enough, so I 
rephrase it here.

 I would like to use phonon-electron code in QE to calculate the matrix element 
only, not the el-phonon coupling constant.  From example07, first a scf at very 
dense grid is calculated and followed by a scf at coarser k-grid (suitable for 
phonon calculation).  Since in the example, k-grids have to be provided in the 
input for all scf calculations and phonon, and the kpt is sampled by 
Monkhorst-Pack (MP), I guess the matrix elements involve only the states at the 
special k-points and q-points specified by MP.  Now if I want to calculate the 
matrix elements involving those states at other k-point, such as the one 
corresponding to minimum of conduction band, is there a fast way for me to do to 
obtain this?   


I have run few tests on example07, but I failed (I just wanted to see how the 
code run).  In the 2nd scf, instead of entering a grid of kpt 8x8x8, I enter 
directly the kpt-grid that I obtained from running kpoints.x for 8x8x8, the run 
crashed and stop at the 1st eletron-phonon calculation.  Clearly, I input the 
same kpt-grid, but the code is unable to run.  The reason why I did this is that 
I like to see whether I provide a kpt list for the scf run the code can run. 
Could you explain to me why this doing this way, it does not work.

Vi Vo
Department of Chemical Engineering,
University of Houston


      
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20100815/c9c84a6e/attachment.html>