[Pw_forum] U

[Gabriele Sclauzero]

Il giorno 06/lug/09, alle ore 10:06, ali kazempour ha scritto:
>
> Dear all
> I want to study the effect of U(habburd parameter on band gap).  
> According to method described in coccocini phd thesis , we are able  
> to obtain U by setting alpha to different values and   the  
> difference in the slope of ns Vs. alpha and ns0 Vs. alpha0  is U  
> parameter.
> My question is : Do I Optimize the lattice constant After obtaining  
> U from above procedure or not?
> Does above method works for any bulk material that have d orbital?

I think that in principle the correct procedure should be to recompute  
U for every lattice parameter value you will use to get the total  
energy and then fit to the equation of state. This is because of the  
internally consistent definition of U, which depends on the local  
environment surrounding the Hubbard atoms. At each value of the  
lattice parameter this environment changes, and the hybridization of  
the Hubbard orbitals is influenced by it. Anyway I guess that if you  
computed the U at a value of the lattice parameter not distant from  
the optimal one (theoretical), maybe you can use that value of U to  
get the total energy in the neighborhood of the optimal value. On the  
contrary, the estimation of the bulk modulus may be much more  
sensitive of the value of U which you employ.


GS


Gabriele Sclauzero, PhD Student
   SISSA & CNR-INFM Democritos,
   via Beirut 2-4, 34014 Trieste (Italy)
   email: sclauzer at sissa.it
   phone: +39 040 3787 511
   skype: gurlonotturno




-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://www.democritos.it/pipermail/pw_forum/attachments/20090706/87e89e08/attachment.htm