[Pw_forum] 3 questions about elph file of Calculation of electron-phonon interaction coefficients

Eyvaz Isaev eyvaz_isaev at yahoo.com
Sat Apr 16 22:16:35 CEST 2005


Hi,
 
> 
> 1st question, which Gaussian Broadening value should
> be right?

It depends on convergence of \lambda over the
"degauss" parameter. You should try to locate some
interval for degauss where  changes for \lambda are
not considerable.

> 
> 2nd question, what gamma values stand for?
> 

That is so-called phonon linewidth, and caused by
coupling of two electrons at the Fermi surface.
It is used to obtain the spectral function 
\alpha^2F(\omega) which in turn used to calculate
\lambda.

I suggest before doing such a kind of calculations you
should read some texbook. At least, see Savrasovs'
paper in PRB. 

> 3rd question, why gamma value corresponding to the
> biggest two lambda(
> 7)&(8) got gamma=********** GHz?
> 

It is due to formatted record. It means that you have 
got some \gamma >999.999999GHz.
Besides, some \gamma you have got (sorry!) are
negative. 
It should be positive by definition, you can try check
convergence of \gamma in dependent on cutoff energy
and k-points sampling.
 
Bests,
Eyvaz.

> 
> Thanks a lot and best regards,
> 
> 
> 
> LIU, LEI
> 
>  
> 
>  
> 
> The below is part of the output elph file at gamma
> point I got: 
> 
>  
> 
>      Atomic displacements:
> 
>      There are   6 irreducible representations
> 
>  
> 
>      Representation     1      2 modes - To be done
> 
>  
> 
>      Representation     2      1 modes - To be done
> 
>  
> 
>      Representation     3      1 modes - To be done
> 
>  
> 
>      Representation     4      1 modes - To be done
> 
>  
> 
>      Representation     5      2 modes - To be done
> 
>  
> 
>      Representation     6      2 modes - To be done
> 
>      PHONON       :  1m30.44s CPU time
> 
>  
> 
>      electron-phonon interaction  ...
> 
>  
> 
>      Gaussian Broadening:   0.010 Ry, ngauss=   1
> 
>      DOS =  5.198430 states/spin/Ry/Unit Cell at Ef=
>  8.662562 eV
> 
>      double delta at Ef =693.041579
> 
>      lambda( 1)= -0.000133   gamma= -0.003329 GHz
> 
>      lambda( 2)= -0.000133   gamma= -0.003329 GHz
> 
>      lambda( 3)=  0.000000   gamma=  0.000000 GHz
> 
>      lambda( 4)= -0.000739   gamma= -0.370550 GHz
> 
>      lambda( 5)= -0.000739   gamma= -0.370550 GHz
> 
>      lambda( 6)=  0.000000   gamma=  0.000000 GHz
> 
>      lambda( 7)=  5.420169   gamma=********** GHz
> 
>      lambda( 8)=  5.420169   gamma=********** GHz
> 
>      lambda( 9)=  0.000811   gamma=  1.719525 GHz
> 
>      Gaussian Broadening:   0.020 Ry, ngauss=   1
> 
>      DOS =  4.897174 states/spin/Ry/Unit Cell at Ef=
>  8.654971 eV
> 
>      double delta at Ef =364.930422
> 
>      lambda( 1)=  0.000579   gamma=  0.013666 GHz
> 
>      lambda( 2)=  0.000579   gamma=  0.013666 GHz
> 
>      lambda( 3)=  0.000000   gamma=  0.000000 GHz
> 
>      lambda( 4)=  0.004037   gamma=  1.906981 GHz
> 
>      lambda( 5)=  0.004037   gamma=  1.906981 GHz
> 
>      lambda( 6)=  0.000000   gamma=  0.000000 GHz
> 
>      lambda( 7)=  3.742847   gamma=********** GHz
> 
>      lambda( 8)=  3.742847   gamma=********** GHz
> 
>      lambda( 9)=  0.000427   gamma=  0.851899 GHz
> 
>                                     ......
> 
>      Gaussian Broadening:   0.100 Ry, ngauss=   1
> 
>      DOS =  4.261566 states/spin/Ry/Unit Cell at Ef=
>  8.675412 eV
> 
>      double delta at Ef = 75.872648
> 
>      lambda( 1)= -0.012885   gamma= -0.264534 GHz
> 
>      lambda( 2)= -0.012885   gamma= -0.264534 GHz
> 
>      lambda( 3)=  0.002440   gamma=  0.121909 GHz
> 
>      lambda( 4)=  0.003925   gamma=  1.613456 GHz
> 
>      lambda( 5)=  0.003925   gamma=  1.613456 GHz
> 
>      lambda( 6)= -0.000183   gamma= -0.112090 GHz
> 
>      lambda( 7)=  1.048348   gamma=********** GHz
> 
>      lambda( 8)=  1.048348   gamma=********** GHz
> 
>      lambda( 9)=  0.000068   gamma=  0.117765 GHz
> 
>  
> 
>  
> 
> 

__________________________________________________
Do You Yahoo!?
Tired of spam?  Yahoo! Mail has the best spam protection around 
http://mail.yahoo.com 



More information about the Pw_forum mailing list