Exact meaning of Strength from sort_and_report

[xtbu]

Dear all,
The dipoles extracted from ndb.dip_R with the unit of length (Bohr) or chargr*length?
And since the unit of residuals is energy, then the |BS_R|^2 should be energy*length. But why BS_R/q0_def_norm are in length unit as Palummo posted in the previous reply?

Thanks!

Xiangtian

[Daniele Varsano]

Dear Xiangtan,

Here we use Hartree atomic units. The dipoles have charge*lenght units (e=1).
BS_R is defined as <c|e^iq.r|v> which for q=0 is then treated as q0.<c|r|v> where q0 is a small momentum (10^-5).
Thus BS_R/qo has length unit.

The dielectric permittivity in atomic units has e^2/(ao*Eh) units, or equivalently 4*pi*eps_0.

Best,
Daniele

[xtbu]

Thanks, Daniele. I got the point!

Xiangtian

[gsethi]

Dear Robert,

Could you tell me how were you able to print |BS_R|^2 in ypp?

Thanks a lot.

Regards,
Sethi

I used GaAs as a test system, where I get |BS_R|^2 about 1e-7 (a_0^2 ?) (I modified the code to print it directly in ypp) , which yields about 4e-28 (m^2 ?), which is far too small (should be about 1e-20 m^2) for the transition dipole matrix element of the that specific exciton.

[rbw]

Good morning Sethi,
no problem.
In ypp/excitons/excitions_sort_and_report.F is a line
Residuals(:) = real(BS_R(:)*conjg(BS_R(:)),SP)
I think currently it's line 126.
There, Residuals(:) = q^2 |BS_R|^2, with q=1e-5 bohr

Personally I added
+ ! matrix element times 1/q^2 * bohr2meter^2
+ Residuals_mod(:) = Residuals(:) * 1e10 * (5.29177e-11)**2

Residuals_mod now hold the matrix elements in units of m^2, which is what I needed.
I modded the output a bit, but easiest would probably to change this line (currently 167)
rv(:3)=(/real(BS_E(j2),SP)*HA2EV,Residuals(j2),real(j2,SP)/)
to use the changed residuals variable.
(The rv array is printed out by ypp as part of its standard output).

I hope this helps.
Cheers,
Robert

[haseebphysics1]

Dear Yamboers,

Can you suggest how one can print the optical transition dipole elements against the K-points of the BZ? as shown in the following Fig (c)



Thanks,
Haseeb

[Daniele Varsano]

Dear Haseeb,
once you have extracted the transition dipoles (this can be done using e.g. python scripts as here: viewtopic.php?p=7783#p7783), you have the dipoles for c,v bands and the k points and you can represent it as you prefer and build a figure as in the example you post.
Best,
Daniele

[Fadil]

Dear all,

I want to calculate the oscillator strengths of a system by using dipole matrix elements and exciton weights, which for a specific excitation it should be related with |\Sum_{k,k',n,n'} D_{k,k',n,n'} * C_{k,k',n,n'}|^{2}.
D is dipole matrix elements, I am extracting them from 'ndb.dip_iR_and_P' with a python script in full_BZ.
And for C_{k,k',n,n'}^2, I am using 'o-BSE.exc_weights_at_x' files for a specific excitonic weight. (in order to obtain C values in the full_BZ correctly, I turned off all symmetry considerations in the system)

I was expecting the ratio between the oscillator strengths of the excitons I obtained from the 'O-BSE.exc_E_sorted' file and calculated with the above method to be at least similar. But they are different.

Do you have any idea what I am missing?

[Daniele Varsano]

Dear Fadil,
please sign your post with your name and affiliation, this is a rule of the forum and you can do once for all by filling the signature in your user profile.
Beside prefactor you need to check accurately here I can see a problem due to the fact that the strength reported in 'O-BSE.exc_E_sorted', are normalised to the maximum value:

you can check in:
ypp/excitons/excitons_sort_and_report.F

Code: Select all

130    if (trim(R_normalize)=='yes') then
131      Rmax = maxval(Residuals)
132      Residuals=Residuals/Rmax
133      write(dump_ch,'(a,1e10.5)') " Maximum Residual Value = ",Rmax
134    endif

the normalisation factor is written in the output as " Maximum Residual Value " and you need to rescale the strength to that value.

Best,
Daniele

[Fadil]

Dear Daniele,

Thank you, I found it under 'o-BSE.exc_qpt1_E_sorted'.
However, it does not solve my problem as I need to rescale each excitation with the same normalization factor.
The ratio of oscillator strength between the different excitons that I obtained by summing the dipole matrix elements
and weights is still not compatible with the result of the yambo.
Is there a way to write the residuals for a particular exciton before summing over the k-points and bands?

Best