The ABINIT Users Mailing List ( CLOSED )

[Andrey Khavryuchenko <akhavr@gmail.com>]

Pierre-Matthieu,

"AP" == Anglade Pierre-Matthieu wrote:

 AP> Abinit, up to v5.1 is only doing fully periodic calculations. The most
 AP> proper way for your problem is to use big to very big cells. And
 AP> evaluate the convergence of results with respect to cell size.
 AP> To include extra charges you must add a compensation charge in the
 AP> background. see keyword "charge".

Thanks.

Nuno,

"NAGB" == Nuno A G Bandeira wrote:

 NAGB> Andrey Khavryuchenko wrote:
 >> Hi!
 >> What is the proper way to calculate isolated, non-periodic system?  How 
 >> can
 >> I define, then, it's charge and multiplicity?

 NAGB> You must first define your model with 1 k point.

 NAGB> kptopt 0
 NAGB> nkpt 1

Already done.

 NAGB> Consult the pdf manual (v. 4.33) and search for the strings
 NAGB> occupation' and 'charge'. Something is bound to come up.

Ok.

 NAGB> Also you may want to consider the appropriateness of using planewaves
 NAGB> as a suitable basis set. It may take a lot longer than using "normal"
 NAGB> basis sets for molecular models for the same level of accuracy. 

I've turned to abinit because at future step I'll need to perform a
modelling of 2d-periodic system (slab) and what results be comparable and
done with the same tool.

But it looks like I have to use different software for that steps.

[...]

 NAGB> Gaussian is an old dynosaur that has the advantage of being
 NAGB> ecclectic.

:)

-- 
Andrey V Khavryuchenko            
Software Development Company http://www.kds.com.ua/