[Pw_forum] Potential alignment in charged defect supercell

[毛飞]

Dear all,
 
I am studying the charged defects in semiconductor with the help of Q.E. codes. It is known that, when study the formation energy of the charged defects with supercell method, the finite-size effects should be carefully considered.
 
When I calculate the total energy of the charged defect cell (for example, tot_charge=1) in the periodic boundary condition calculations, a compensating jellium background is inserted to remove energy divergences. By setting assume_isolated= makov-payne which corrects the image charge interactions, as said in the variable instructions, an estimate of the vacuum level is then calculated so that eigenvalues can be properly aligned. Can I think the makov-payne method only partially align the electrostatic potential of the charged cell? After I have defined the two variables in the input file, do I need continue to align the electrostatic potential far from the defect with that of the corresponding bulk cell (defect free) after the calculation, in order to exclude the interaction between the background density and the real, phycial charges in the supercell.
 
A question more technically, how to obtain the the atomic-site (local) electrostatic potential as a function of the distance in the supercell with Q.E. codes.
 
Any suggestions or comments are appreciated.
Evan
USC, China





-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.quantum-espresso.org/pipermail/users/attachments/20160901/c4b500b0/attachment.html>