About polarization direction

[chinaye]

Dear developers,
in yambo.in generated by using the command yambo -b -o b -y h,
there are two options determining the direction of the polarization. If the BLongDir option is
0.000000 | 0.000000 | 1.000000 |, must LongDrXs be 0.000000 | 0.000000 | 1.000000 |? Thank you very much.

[andrea marini]

Dear developers,
in yambo.in generated by using the command yambo -b -o b -y h,
there are two options determining the direction of the polarization. If the BLongDir option is
0.000000 | 0.000000 | 1.000000 |, must LongDrXs be 0.000000 | 0.000000 | 1.000000 |? Thank you very much.

Dear Jianfei Ye,

the short answer to your question is: not necessarily. Whenever you calculate a response function you need to specify the direction of the electric field that defines the polarization for small transferred momenta. You can find the relevant equations here. In your specific case you are calculating the static response function needed to integrate the Bethe Salpeter kernel (LongDrXs) and the Bethe Salpeter absorption itself (BLongDir).

As explained here the static dielectric function is integrated in the BS kernel and all its components are summed up. So the direction of the long wavelength limit only represents the line along which integration of the Brillouin Zone is performed. For this reason the choice of LongDrXs is not so relevant unless you are studying a very anisotropic material.

The choice of BLongDir, instead, is important as it defines the polarization of the external perturbing field.