The ABINIT Users Mailing List ( CLOSED )

[Jean-Christophe Charlier <charlier@pcpm.ucl.ac.be>]

Dear Sir,

We have recently performed an ab initio calculation (including structural

optimization) of a (4,2) nanotube containing 56 atoms in its unit cell.

It seems that this calculation is close to the limit of what ABINIT is capable

to do on conventional computing architectures.

For your specific simulation, I would suggest you to use a software based on 

localized orbitals (such as SIESTA, GAUSSIAN, ...) rather than on

plane waves (like ABINIT, PWSCF, ...), because you'll pay a lot for the vacuum

space around the nanotube in the unit cell.

In addition, the software SIESTA is free for universities !

You could find more informations in the following link :

 

http://www.uam.es/departamentos/ciencias/fismateriac/siesta/

With best regards,

JC Charlier

**********************************************************

Charlier Jean-Christophe

Chercheur Qualifié FNRS

Université Catholique de Louvain

Unité P.C.P.M., Place Croix du Sud, 1 (Boltzmann)

1348 Louvain-la-Neuve, BELGIUM

Tel:   (32) 10 47 35 64

Fax:  (32) 10 47 34 52

**********************************************************

On Nov 25, 2005, at 7:23 AM, SNU wrote:

Hi,

I'm new in this calculation field.

Last week I visited a group, which purchased software for an

incredible large amount of money, to do abinitio calculations

on Carbon Nano Tubes (CNTs); a system typically contained

about 200 to 250 atoms.

On a cluster of 6 PCs, each with single 3 GHz Intel CPU,

a typical calculation takes about one week.

The people there were not familiar with ABINIT, even never

had heard of ABINIT!

I wonder if ABINIT is capable of doing such calculations too:

- a CNT of about 200 atoms

- calculate the binding energy between a functional group

  (approx. 10 atoms) with the wall of a CNT.

Would it be worth to use ABINIT for such systems, or is ABINIT

the wrong software for this topic?

Thanks,

Rob.