From owner-chemistry-: at :-ccl.net Mon Apr 17 09:18:01 2006
From: "Eduard Matito ematito_-_gmail.com" <owner-chemistry[#]server.ccl.net>
To: CCL
Subject: CCL: AIM charges - HF or DFT?
Message-Id: <-31531-060417082333-32057-LZYz/IFsS+shWG2ClZPKIg[#]server.ccl.net>
X-Original-From: Eduard Matito <ematito * gmail.com>
Content-Transfer-Encoding: 7bit
Content-Type: text/plain; charset=ISO-8859-1; format=flowed
Date: Mon, 17 Apr 2006 14:21:37 +0200
MIME-Version: 1.0


Sent to CCL by: Eduard Matito [ematito===gmail.com]
Dear David,

> First of all, is a HF or a DFT wave function better?  What about the 
> hybrid=
> B3LYP?=20=20
>  
>
B3LYP geometries are far better than HF ones, but HF is actually the 
best monodeterminantal wavefunction
by definition, so I would rather consider using HF wavefunction. In 
addition to QTAIM charges a useful tool
for the analysis of the electron distributions are the so-called 
localization/delocalization indexes (LI and DI).
See for instance the following papers:
X. Fradera, M. A. Austen, and R. F. W. Bader, J. Phys. Chem. A, 1999, 
103, 304.
J. Poater, M. Sola, M. Duran, and X. Fradera, Theo. Chem. Acc., 2002, 
107, 362.

> Second, is a large basis set still considered to be important?  What 
> about =
> diffuse functions?=20=20
>  
>
Most of these 3-D molecular space decomposition schemes (as QTAIM or 
fuzzy-atom) are less sensible to
basis set dependence than the Hilbert-space counterparts. Unlike Hibert 
space analysis, which needs of atom
centered functions, 3D schemes do not have problems to use diffuse or 
bond functions.

> Third, I am considering purchasing AIM2000.  Are AIM2000 users 
> satisfied?  =
> Does it have limitations I should be aware of?  Is there competing 
> software=
> I should consider?  For non-computational-chemistry reasons, I have 
> only W=
> indows XP operating system.
>  
>
The group of Bader gives the PROAIM program for free. It contains all 
applications needed for the calculation
of QTAIM properties. The good points of AIM2000 are its nice graphical 
interface and that the calculation
of LI and DI are done automatically (however you can easily write a 
50lines code to calculate DIs and LIs
> from *.int files generated with PROAIM for a monodeterminantal 
wavefunction).

Hope it helps,

Eduard Matito
Institute of Computational Chemistry
University of Girona (Spain)
e-mail: eduard(at)iqc.udg.es
http://iqc.udg.es/~eduard