el-ph matrix elements in polar materials

[zahra]

Dear all,

In the following paper it is stated that g (el-ph matrix elements) are separated into short- and long-range contributions where then the short-range contribution is dealt with standard wannier interpolation and both contributions are added up after interpolation at arbitrary k and q points (fine grid?)
https://journals.aps.org/prl/pdf/10.110 ... 115.176401
My question is that how is then the long-range contribution is treated? Is this part calculated within QE?

Also, for polar 2D material it is then believed that the Frohlich interaction should be treated differently when we are in long wave length limit (q<1/|d| where d is the interlayer distance):
https://journals.aps.org/prb/pdf/10.110 ... .94.085415
In fact it is then understood that the diverging behavior of g matrix elements are the result of non-zero interaction between the images of the 2D system, resembling the behavior of a bulk system. Therefore this group have suggested the modification of Coulomb potential within the QE code (they have used the QE code for g matrix elements calculation).
My question is that: is this what happens in EPW too? And is there a way to prevent this feature and implement the same procedure as they have done but in EPW? or, one also has to consider the QE modification?

I have done calculation on monolayer MoS2 and I observed diverging behavior in there near \Gamma point for phonons.

Many thanks,
Zahra

[sponce]

Hello,

The calculation of the splitting into short and long-range for polar materials needs to be done only when you do Wannier interpolation.
Such splitting is done in the public EPW version.

A similar approach to EPW has been developed by the group of M. Calandra and is directly inside QE. They also have coded the short/long range
splitting but I do not know if this is inside the public version of QE.

For the Coulomb truncation in the case of 2D materials, their developement is, as far as I'm aware, not publicly available in QE.

There is no Coulomb truncation in EPW at the moment. It is something I intend to look at eventually.

Best,

Samuel

[zahra]

Hi,

Thank you so much.
But isn't it right that one should have Coulomb truncation in QE at first place to have it then in EPW? As far as I know the most important part of the el-ph interaction is calculated within QE (i.e. dvscf files)

Best Regards,
Zahra

[sponce]

Hello,

This is correct.

Sam