I don't know how GW handles net charge. I remember GW and the energy of the poles of green function are the energy of adding and removing one electron. The GW gap is essentially the electron affinity (adding electron). The extra energy cost of having non-zero net change is a little confusing to me. Does this cost scale down with system size? Given a supercell with 20 atoms, would the extra charge energy cost (1e smeared over 20 atoms) be less than an unit cell with with 2 atoms (1e smeared over 2 atoms)? The number of k-mesh sampling also in some sense corresponds to "supercell for electron wavefunction". Would this affect the energy cost?
I want to ask the Question that I want purches my PC and I want to khow that which system is best for VASP software for work If I Purches the the i7 and 6th gen. will I be able to work on this system?
I was frustrated too much when I saw hedin's last equation. But I can understand it now because Feynman's diagram kept on my head :)
Please ask questions regarding VASP in the VASP forum: www.vasp.at/forum/
I don't know how GW handles net charge. I remember GW and the energy of the poles of green function are the energy of adding and removing one electron. The GW gap is essentially the electron affinity (adding electron). The extra energy cost of having non-zero net change is a little confusing to me. Does this cost scale down with system size? Given a supercell with 20 atoms, would the extra charge energy cost (1e smeared over 20 atoms) be less than an unit cell with with 2 atoms (1e smeared over 2 atoms)?
The number of k-mesh sampling also in some sense corresponds to "supercell for electron wavefunction". Would this affect the energy cost?
I want to ask the Question that I want purches my PC and I want to khow that which system is best for VASP software for work If I Purches the the i7 and 6th gen. will I be able to work on this system?
How to perform NEB calculations