Dear Y.P. Wang,<br><br>The Wannier code as it stands right now is capable<br>of computing "bulk" Quantum Conductance automatically.<br>"bulk" means that your system must be periodic (like a pristine<br>
carbon nanotube or a pristine nanowire for example).<br>As for a "Lead Conductor Lead" calculation (denoted 'lcr' in<br>the Wannier input file) you need to give 5 matrices to the code<br>in order to compute the Quantum Conductance. The five matrices<br>
are :<br><br> H_L that you should call <seedname>_htL.dat, where seedname<br> is the same as your master <seedname>.win file.<br> This matrix should correspond to the left semi-infinte lead.<br>
<br> H_LC you should call it <seedname>_htLC.dat.<br> This matrix should correspond to the Left Lead - Conductor<br> coupling matrix.<br><br> H_C you should call it <seedname>_htC.dat.<br>
This is the matrix corresponding to the Conductor region or<br> "central part" of your system (which probably contains the<br> defect)<br><br> H_CR you should call it <seedname>_htCR.dat.<br>
This matrix corresponds to the Conductor - Right Lead coupling<br> matrix<br><br> H_R you should call it <seedname>_htR.dat.<br> This corresponds to the Right Lead matrix.<br><br>
You should look at the Wannier user guide to find the format of those files.<br>Those are basically Hamiltonian matrices expressed in the Wannier basis of<br>the corresponding piece of your system (Left Lead, Conductor, Right Lead).<br>
<br>Also you should know that the next release of the Wannier90 code should<br>have this 'lcr' type of calculations <b>completely automated</b> such that the <br>code will directly compute the Quantum Conductance automatically. So you<br>
will not have to do anything, just like in the 'bulk' case!<br><br>Cheers,<br><br>Nicolas Poilvert<br><br><br>-- <br>POILVERT Nicolas<br>PhD candidate,<br>Dpt of Materials Science and Engineering<br>Massachusetts Institute of Technology<br>
77, massachusetts avenue<br>Cambridge, MA 02139<br>USA<br>work: (617) 452-4212<br><a href="mailto:nicolas.poilvert@gmail.com">nicolas.poilvert@gmail.com</a><br>