On Friday, December 23, 2005, at 07:37 AM, Alvin Chang wrote:

> As I know, BP in ADF denot BPV86 in GAUSSIAN ,not BP86.The main
> difference is the usage of different correlation of LDA.
> in BP, it use VWN
> in BP86,it use Perdew81

Historically, within ADF you would specify Becke-Perdew which includes
then
VWN local correlation; this option has in private versions and in
several
papers been referred to as BP86. Therefore, it has been added as
shortcut
within ADF.

In private communication with Perdew and co-workers, while I was
implementing
the TPSS functional in ADF, it was found out that the ADF BP86 gives a
much
better performance for the G2-set than the Gaussian BP86 (which uses
PZ81
local correlation); therefore the VP86 shortcut was introduced into G03.

In summary, BP86 within ADF is equal to Becke-Perdew, and uses VWN local
correlation.

末末末末末末末末末末末末末末末末末末末末末末
dr. Marcel Swart

Theoretische Chemie (kamer R152)
Vrije Universiteit Amsterdam
Faculteit der Exacte Wetenschappen

De Boelelaan 1083
1081 HV Amsterdam
The Netherlands

Tel +31-20-5987619
Fax +31-20-5987629
E-mail m.swart_at_few.vu.nl
Web http://www.few.vu.nl/~swart
末末末末末末末末末末末末末末末末末末末末末末
Received on 2005-12-23 12:04:47