Dear Dr. German Samolyuk, Thank you for the detailed reply. I checked your example but I do not understand how do you calculate the zed and config values according to the vacancies? For mixture of Cu with 1/29 i was expecting the zed value to be 28.
Best regards, Novendra Novendra On Tue, Apr 4, 2017 at 3:29 PM German Samolyuk <[email protected]> wrote: Dear Novendra Novendra, The only way how to do it within pseudopotential method I found, was to make pseudopotential with both zed and config values equal to fractional number. For example, for Cu: zed=28.9, config='[Ar] 3d9.5 4s1.4 4p0.0' corresponds to mixture of Cu with 1/29 vacancies. It's not exactly the same as mixing with "empty" pspot, but more like rigid band approximation for particular component. If you need do do something more accurate, probably, one of the CPA KKR packages will be more appropriate. One more comment regarding modeling vacancies with VCA. You probably should remember that even in closed packed structures like most of the metals relaxation of surrounding vacancy atoms positions is significant. This effect is not included within VCA approach. Best, German Samolyuk ORNL On Mon, Apr 3, 2017 at 2:34 PM, Novendra Novendra <[email protected]> wrote: Dear QE developers, I am trying to prepare input file for a system with a positional disorder. I have seen some example where VCA can be applied for system with substitional disorder, where in that case a virtual PP can be prepared by mixing PP of different elements. In the case of positional disorder, is it possible to prepare a virtual PP by mixing with an "empty" PP? Thank you, Novendra Novendra Ph.D Student, Department of Chemical Engineering & Materials Science University of California, Davis _______________________________________________ Pw_forum mailing list [email protected] http://pwscf.org/mailman/listinfo/pw_forum _______________________________________________ Pw_forum mailing list [email protected] http://pwscf.org/mailman/listinfo/pw_forum
_______________________________________________ Pw_forum mailing list [email protected] http://pwscf.org/mailman/listinfo/pw_forum
