Re: CCL:TM force fields

[Per-Ola Norrby <peo ( ( at ) ) compchem.dfh.dk>]

 	Hi all,
	I'd like to take up this thread again.  Also, I have a
 question at the end:
 Isaac B. Bersuker wrote:
> The open (or quasi-open) d shell is the main reason of
>  peculiar
>  properties of  TM systems (TMS)
>  with regard to modeling:
> (1) 3D delocalization and hence strong mutual influence of  the M-L
>  bonds on each other (e.g.,
>  cis- and trans-effect),
> (2) "back donation" and hence local excitation of the
>  ligands
>  (chemical activation by
>  coordination),
> (3) degeneracy and pseudodegeneracy and hence the strong Jahn-Teller
>  and pseudo JT effects.
> All this together makes imposible to get reasonable parametrization
>  of a force field (with
>  transferable parameters) that could be used in modeling similar to
>  that of organic and/or
>  non-TMS, in general.
> I am sorry that these important features of TMS remain unknown to
>  people engaged in their
>  investigation.
	I completely agree that these effects can make modelling of
 transition metals challenging.  However, I disagree strongly with the
 word "impossible".  We are still far from a force field that
 can do
 this for all conceivable complexes without parameterization, but wide
 transferability is not a requirement for force field modelling.
 Personally, I design force fields for specific problems (you can
 check my web page for examples, see below).  Setting up the force
 field requires about the same time as doing e few QM/MM calculations,
 then I have access to a tool that allows rapid conformational
 searching.  In the types of problems I'm working with (asymmetric
 catalysis), a good sampling of the conformational space is essential.
 The force fields are still 3-4 orders of magnitude faster than QM/MM
 calculations in this case, and seemingly almost as accurate.  An
 attractive combination is to use the force field as a search tool,
 followed by refinement of a few selected structures by QM/MM.
	For the concerns listed above, we've only done simple things
 like introduction of specific cross terms to reproduce trans
 induction, but others have done a lot more.  I'm aware of work by
 Landis, Comba, Deeth, and Carlsson, solving different aspects of
 these problems.  I would like to see more.  If you could send me
 references on how to treat the above problems, I'll summarize to the
 list.
 	Best regards,
 	Per-Ola
 --
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  *  Per-Ola Norrby, Associate Professor
  *  The Royal Danish School of Pharmacy, Dept. of Med. Chem.
  *  Universitetsparken 2, DK 2100 Copenhagen, Denmark
  *  Tel.   +45-35306506, fax +45-35306040
  *  Email: peo ( ( at ) ) dfh.dk (preferred), peo ( ( at ) ) compchem.dfh.dk
  *  WWW:   http://compchem.dfh.dk/PeO/