Hi Mahesh,

Hybrid DFT functionals are a still evolving feature in PWscf.

Only a few functionalities are implemented.

The following hybrid functionals are implemented: Hartree-Fock, PBE0,B3LYP, HSE (see Modules/functionals.f90 for updated info and more details).

Usually in PWscf the functional to be used is read from pseudopotential files but we do not have so far a pseudopotential generator for hybrid functionals so one needs to use pseudopotentials generated with some other functionals (eg. LDA, PBE, BLYP) and force the usage of a hybrid functional using input variable "input_dft" in &SYSTEM namelist.

Also you have to insert this additional flags:

nqx1 = #,

nqx2 = #,

nqx3 = #,

x_gamma_extrapolation = .TRUE.

exxdiv_treatment = gygi-baldereschi

Usually these calculations are very expensive, where basically the computational cost depends on the number of q-points chosen to represent the BZ integration, and depends in general on the localization of the Wannier functions of the system.

To choose the number of q-points in the &SYSTEM namelist there are three variables nqx1,nqx2,nqx3 that define the regular q-grid in the BZ in a way similar to the automatic k-points generation. Their value must be compatible with the k-points used.

Good Luck!

Cheers,

A.

Thanks Mr. A. Pidatella

hi Mahesh,

Did you success in calculating the bandgap of TiO2 by using HSE ???

Because I am a new one in WE. And I am trying to calculate the bandgap of ZnO by HSE.

But the scf out send me the message " Error in routine read_namlist (17) ..... reading namelist system"

And I do not know why.

If you succeed in your calculation please give me an advice about this subject.

Hello this error is related to a fault in your input file in block of system ,prbably you forget a comma or you wrote wrong word

Hi Mahesh,

I am inspired by your questions. To the best of my knowledge, no hybrid pps are available now. However, our group is working on this, and several works on this field are on the way. I think these works will be of great importance to the development of DFT in condensed matter physics.

 Dear All,

Now, I also want to use QE to calculate the band gap with HSE functional, but it is found that the results is not good, much smaller than that of previous reports.

I just use the PBE pseudo-potential in HSE calculation.

Dear All,

You have to manually set it in your input in the &SYSTEM block, simply add a line for your desired functional as: input_dft = 'hse' But beware that this will significantly increase your computational costs compared to the GGA functional read from your pseudopotentials.