CCL: Level of Theory for zwitterionic species
- From: Justin Finnerty <justin.finnerty *
uni-oldenburg.de>
- Subject: CCL: Level of Theory for zwitterionic species
- Date: Fri, 06 Nov 2009 13:17:52 +0100
Sent to CCL by: Justin Finnerty [justin.finnerty(~)uni-oldenburg.de]
On Thu, 2009-11-05 at 10:11 -0500, Sarah Ann Kebbell
sarah.kebbell2_-_mail.dcu.ie wrote:
> Sent to CCL by: "Sarah Ann Kebbell"
[sarah.kebbell2^-^mail.dcu.ie]
> Dear CCLers,
>
> just wondering if anybody has an previous experience in locating
intermediates that are zwitterionic in nature and if so, is B3LYP/6-31G(d) a
sufficient level of theory to locate zwitterions at or do I need to change the
theory/basis set to do so?
>
> Many thanks,
> Sarah
Dear Sarah,
myself, Rainer Koch and a number of others have worked with the
well-known experimentalist Prof Curt Wentrup in the field of reactive
intermediates. We have successfully used B3LYP/6-31G(d) to predict the
base geometry of a range of formally zwitterionic intermediates. This
was always combined with improved energy predictions using single-point
calculations with large basis-sets and/or better theories. As Tanja
said, solvation can be a major factor and we have had to use both the
implicit solvation model (eg SCRF) and more rarely explicit solvation.
However, the improvement in computing resources now means we use the
6-311++G** basis set for both geometry optimistation and
thermochemistry. We are finding this to give much improved results for
determining reaction pathways that had formerly eluded us. We are also
investigating alternative functionals, such as MPW1K, as possible
alternatives to the venerable B3LYP.
One point that I see is that the zwitterionic character is not readily
apparent in the molecular orbital picture, particularly for gas-phase
calculations. It usually requires full NBO style analysis to get a more
"chemist'S intuitive" picture with formal charges on atoms. However,
I,
myself, prefer the MO picture where you consider "zwitterions" as
having
volumes with increased or decreased electron density rather than the
notion of formal charges.
Cheers
Justin
--
Dr Justin Finnerty
Rm W3-1-165 Ph 49 (441) 798 3726
Carl von Ossietzky Universität Oldenburg