Hi, Probably you will want to apply a shift to obtain a correct bandgap. If you want to do it with the option shit=..., first read PRB 48, 11789 (1993). epsilon.x has implemented what is called "naive scissors correction". The correction is easy to implement, and I have a working one for version 4.3. Sooner or later it will get to the distribution, (if I am not wrong).
Use LDA, as in the PRB article, and I think also semilocal GGA functionals are fine. Do not use a hybrid functional. -- Eduardo Menendez Proupin Departamento de Qu?mica Fisica Aplicada Facultad de Ciencias Universidad Aut?noma de Madrid 28049 Madrid, Spain Phone: +34 91 497 6706 ---------- Mensaje reenviado ---------- From: Layla Martin-Samos <lmartinsa...@gmail.com> To: Iwan Darmadi <iwan_darmadi at rocketmail.com>, PWSCF Forum < pw_forum at pwscf.org> Cc: Date: Thu, 29 Nov 2012 09:31:24 +0100 Subject: Re: [Pw_forum] Frequency Dependent Dielectric Function Dear Iwan, take a look at epsilon.x inside PP directory. Be careful, epsilon.x only implements RPA (Random Phase Approximation) dielectric function with no local fields effects and without the nonlocal part of the pseudopotential. cheers Layla 2012/11/29 Iwan Darmadi <iwan_darmadi at rocketmail.com> Dear All, I am Iwan Darmadi, undergraduate physics student, Universitas Indonesia. Currently I am running TM doped ZnO simulation using Quantum Espresso. At the moment, I plan to calculate dielectric function (frequency dependent) of my model. I have read the manual and other sources but I hardly find how to calculate them. My question is, does QE have feature to calculate frequency dependent dielectric function ? If yes, could anyone tell me the main steps to calculate them ? -------------- next part -------------- An HTML attachment was scrubbed... URL: http://pwscf.org/pipermail/pw_forum/attachments/20121129/346b3ed9/attachment.html