[Pw_forum] Kubo-Greenwood with Quantum Espresso

[Taylor, Decarlos E CIV USARMY RDECOM ARL (US)]

Hi,

I'm trying (first attempt) to compute some electrical conductivities of a
few materials and have found a few codes that work with Quantum Espresso
(BoltzTrap for example).  I would also like to implement the Kubo-Greenwood
expression, using the bands and momentum matrix elements provided by Quantum
Espresso.  However, when I use bands.x (and the keyword "lp=.true." to
compute the momentum elements) only the matrix elements between the
conduction and valence bands are provided as output (e.g. <valence | p |
conduction >).  The <valence|p|valence> elements are not computed.  Are
these terms not required for the KG formula for electrical conductivity or
is there a way to get Quantum Espresso to compute them?  I've looked at the
write_p_avg.f90 subroutine in the /PP/src directory, where these elements
are computed, and "could" change the loop indices to include all orbitals
instead of just the valence/conduction bands as currently written.  But I'm
sure that would require adjustments to other parts of the code so I'm
hesitant to try that.  


Thanks for any input,
Carlos 


DeCarlos E. Taylor, PhD
Research Chemist - Multiscale Modeling
Aberdeen Proving Ground, MD  21005-5069
Phone: 410-306-0853
Fax: 410-306-1909
decarlos.e.taylor.civ at mail.mil




-------------- next part --------------
A non-text attachment was scrubbed...
Name: smime.p7s
Type: application/pkcs7-signature
Size: 5635 bytes
Desc: not available
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20161101/122e0ced/attachment.p7s>