Dear ADF-users,

Does anyone have experience with the combination of COSMO and the
FRAGMENTS approach (fragments larger than one atom)? Lately, I was trying
to understand what ADF is doing in such cases, but I've got some very very
puzzling results, for example:

I did a geometry optimization of the complex, using the
solvation-keyword. For the fragments, I tried different things:
(a) single point calculation, using the solvation-keyword, geometry taken
from the complex.
(b) single point calculation, without the solvation-keyword, geometry
taken from the complex.

Then, using the 'fragments approach', I did two single point calculations
of the complex, using the solvation keyword, with TAPE21-files of the
fragments
- from calculation (a) -> (1)
- from calculation (b) -> (2)

To find out what happened, I also calculated the bond energy by
subtracting the total enery of the fragments from that of the complex. So,
I took the total energy of the complex from the geometry optimization and
the total energy of the fragments
- from calculation (a) -> (3)
- from calculation (b) -> (4)

Now, what I would expect, is that (1) and (3) would be very similar, as
well as (2) and (4). (The difference should be only the BSSE, what should
be small using TZ2P.) Also, as for one of the fragments the energy with
solvation is very different from that without solvation, I expect (1) and
(2) to be very different. I thought that (1) would be the correct way to
find to bonding energy in solution.

But this is what I get: (2) and (4) are (almost) the same, but also very
close to (1). And (3) is totally different from the other three.

Is there someone who can explain me this, or who knows which of the
results is actually physically meaningful?? All comments are welcome!!!

Kind regards,
Els
Received on 2003-07-25 10:08:36