Thursday, 5 December 2013

Pancake bonding

π-system radicals may interact with each other in a special way. An actual recombination of the radicals is not possible due to the delocalization in the π-system. But the binding between two such radicals is still a lot stronger than would be expected from pure van-der-Waals interactions. This phenomenon is called "pancake bonding".

There is a new paper out by us, studying this: "Study of the Diradicaloid Character in a Prototypical Pancake-Bonded Dimer: The Stacked Tetracyanoethylene (TCNE) Anion Dimer and the Neutral K2TCNE2 Complex" in ChemPhysChem.

Pancake bonding is quite demanding from a computational point of view, since you need to describe the static correlation of the radicals as well as dynamic correlation, which determines the dispersion interaction. We used the multireference averaged quadratic coupled cluster (MR-AQCC) method, which should be able to deal with these effects. MR-AQCC is very similar to multireference configuration interaction only that there is an additional correction for size-extensivity.


Anonymous said...

Which code implements MR-AQCC? Gaussian09? How probable to find such pancakes at the surface while chemisorption? PS Sorry for asking, cannot access the article.

Felix said...

Hi, we were using Columbus:
it is a free program with an efficient and highly parallel implementation of MR-CI and MR-AQCC.

These TCNEs form crystals. I don't know what would happen at the surface, though

if you send me an email, I can see if I can send you an author's copy of the article. if you are interested in that