No binding of Pt2

[Joris Kuipers]

Hello everyone,

I hope someone can help me or if someone has experienced this problem too.

No matter what kind of method you use (CCSD, Multi with CC, MRCI, MRPT2,
etc.) no more than 1 milihartree of binding is found for Pt2 opposed to
the GAMESS program package which gives a binding of ~ 2 eV.

I compared the way molpro describes the orbitals and what kind of
references (configurations) should be taken into account. I'm using the
ECP60MHF effective core potential (to deal with the large Pt) and using
the corresponding s, p and d basis.

I'm suspecting that these core potentials are not fully functional for
larger elements so that they do not deal properly with the relativistic
effects.

Does anyone recknognize this problem or does anybody know what else could
be causing this discrepancy? As many of you know Pt2 has more binding than
the vanderwaals forces! I hope someone can give me some feedback.

Taken from my input:

 geometry={angstrom;
           Pt1
           Pt2,1,R1(k)}


 gprint,basis,orbital,cs,cp
 basis = {
 ecp, Pt,ECP60MHF
 spd,Pt,ECP60MHF;
 }


Kind regards,

Joris Kuipers



Leiden Institute of Chemistry, Gorlaeus Laboratories.
Leiden University;Einsteinweg 55 / P.O. Box 9502 2300RA, Leiden The Netherlands.
Phone:        +31 71 5274400    TeleFax:      +31 71 5274537
Department:                     Phone:        +31 71 527