I am trying to understand what it means that k sampling is not implemented for hybrid functionals. Does it mean that the structure is still treated periodically with gamma point or it means that it is assumed to be a closed system without k points taking into the account. I am confused because when I print out the Kohn Sham Hamiltonian, there is only one block and it seems like there is no list of neighboring cells.
hybrid functionals can only be used in periodic calculations using
the Gamma point only. In this case all operator matrices are
over periodic Gaussian functions.
I am planning to do transport simulations using the KS Hamiltonian with hybrid functionals. but the Hamiltonian that is printed (using &AO_MATRICES%%KOHN_SHAM_MATRIX .TRUE.) doesn't have different blocks for neighboring cells. It is probably attributed to the fact that K point keys aren't used with hybrids, even though gamma point is considered in the simulations. If I do hybrid simulations with k point scheme of "GAMMA" I get an error. Is there any way to get the different blocks of Hamitonian (corresponding to neighboring cells) similar to what I can get with PBEs?
HFX calculations use a special scheme for the Gamma point. The
integrals are summed up early and no operator matrices between
cells is constructed.
The k-point code (even if you only use the Gamma point) has no
HFX part. So you cannot get these matrices from CP2K, it is not
implemented.
> Thank you very much, It helped me clear up a lot of confusion.
> One last question, so It would be pointless for me to go through the source code and try to print raw Hamiltonians myself?