On Thu, 13 Dec 2001, kiran boggavarapu wrote:
> As anticipated shifting from Gaussian to ADF create
> some intial hiccups esp. with ADF's FRagment energy
> and lack of total energy computation.

Actually, this is a common misconception. It -is- possible
to compute total energy in ADF, using the same reference
convention as, say Gaussian (ie bare nuclei and infinitely
separated electrons). To do this, one would have to
calculate ADF bonding energies of bare nuclei (by using
the CHARGE keyword appropriately), and subtract these
(per-atom) energies from the bonding energy of the system.
Because bonding energies of the bare nuclei tend to be
-very- large, it is advisable to perform such calculations
with an ACCINT setting higher than default. It is also
possible to calculate Gaussian-style total energy within
ADF directly, by specifying suitably charged atomic
fragments.

Apart from a few special cases (usually involving direct
comparison with published total energies, using bare-nuclei
energy reference) there is no benefit to the Gaussian-style
total energies.

(Depending on the version of ADF you are using, small
 source code changes may be required to allow bare nucleus
 calculations to succeed).

Serguei

---
Dr. Serguei Patchkovskii
Tel: +1-(613)-991-2719
Fax: +1-(613)-947-2838
E-mail: Serguei.Patchkovskii_at_nrc.ca
Research Council Officer
Theory and Computation Group
Steacie Institute for Molecular Sciences
National Research Council Canada
Room 2158, 100 Sussex Drive
Ottawa, Ontario
K1A 0R6 Canada