[gpaw-users] get_eels_spectrum() - does this give the full "interacting" response function?

[Trevor Hardcastle]

Thanks Kirsten. I've noticed that when I use get_eels_spectrum(xc='ALDA'), my graphene spectrum goes to zero everywhere with sensible vacuum thicknesses (> 10 Angstroms). I'm assuming that this is due to the divergence of the ALDA kernal at n(r)=0 mentioned by Wachsmuth et al. PRB 88, 075433 (2013). (http://journals.aps.org/prb/abstract/10.1103/PhysRevB.88.075433) The authors of this study used a real-space grid in the vacuum direction to somehow circumvent this issue. Does GPAW! have any similar tricks?
 
Thanks,

Trevor


-----Original Message-----
From: gpaw-users-bounces at listserv.fysik.dtu.dk [mailto:gpaw-users-bounces at listserv.fysik.dtu.dk] On Behalf Of Kirsten Andersen
Sent: 19 November 2014 09:55
To: gpaw-users at listserv.fysik.dtu.dk
Subject: Re: [gpaw-users] get_eels_spectrum() - does this give the full "interacting" response function?

Hi Trevor,

The eels spectrum is calculated from macroscopic dielectric function, which is obtained from the inverse dielectric matrix - this corresponds to using the interacting density response function where local field effects are included. See for example the tutorial at the gpaw homepage:
https://wiki.fysik.dtu.dk/gpaw/tutorials/dielectric_response/dielectric_response.html?highlight=df#gpaw.response.df.DielectricFunction

If I remember correctly the output returns a file with the loss spectrum as two columns without and with local field effects. 

If you're only using RPA (and not ALDA) the Dyson equation is actually not solved in the code. Instead the dielectric matrix is found from
chi0: 
epsilon = I - V  chi0
And then epsilon is inverted to account for local field effects. This is equivalent to obtaining chi first and the using epsilon^-1 = I + V chi as written in the method paper. 

Best regards
Kirsten



On Tue, 2014-11-18 at 19:14 +0000, Trevor Hardcastle wrote:
> Hello all,
> 
>  
> 
> I’d like to check that I properly understand what equations are 
> actually being solved, and what steps are being taken analytically, 
> when I command GPAW! to do the following:
> 
>  
> 
> from gpaw.response.df import DielectricFunction
> 
> df = DielectricFunction(calc='gs-diagonalised.gpw')    
> 
> df.get_eels_spectrum(xc=’RPA’)
> 
>  
> 
> Does  get_eels_spectrum() use the Dyson series expression to go beyond 
> the Adler-Wiser formula for “chi0” and calculate the full 
> “interacting” density response function “chi” using TDDFT? That is, 
> does it follow the procedure summarised in equations (4) – (12) of Jan 
> et al. PRB 83, 245122 (2011)?
>  http://journals.aps.org/prb/abstract/10.1103/PhysRevB.83.245122
> 
>  
> 
> Many thanks,
> 
> Trevor
> 
>  
> 
> ************************
> 
> Dr. Trevor P. Hardcastle
> 
> EPSRC Doctoral Prize Fellow
> 
> Room B20.C
> 
> Engineering Building
> 
> Institute for Materials Research
> 
> University of Leeds
> 
> LS2 9JT, United Kingdom
> 
> t.p.hardcastle at leeds.ac.uk
> 
> ************************
> 
>  
> 
> 
> _______________________________________________
> gpaw-users mailing list
> gpaw-users at listserv.fysik.dtu.dk
> https://listserv.fysik.dtu.dk/mailman/listinfo/gpaw-users


_______________________________________________
gpaw-users mailing list
gpaw-users at listserv.fysik.dtu.dk
https://listserv.fysik.dtu.dk/mailman/listinfo/gpaw-users