You want to elaborate on this? Is it because of the STOs? As to GTO's I use TZVP/P basis sets, which are not too bad usually.
Yes, the STO's are always much better. But not only that. It is a very clever program, making the calculations very efficient when looking at CPU-time, parallellization, those kind of things.
> Furthermore, did you also include relativistic effects ? yes, scalar relativistic via the ECP.
Well, that's not really what I meant. You assume that the ECP's mimic the relativistic effects you are after, but it is not the same as including the relativistic effects themselves !!!
> I guess when you look at Tungsten/Molybdenum, these can be quite significant.Again this is a YES to ADF, since in this program these effects are> easily taken into account.
Also, I know from work from my Professor (yet unpublished I think), who is an expert in Relativistic Quantum Chemistry (J.G. Snijders), that these relativistic effects are needed for heavy atom dimers. If you don't include them (properly), the calculations are not performing very well.
Marcel.