Friday, September 20, 2013

Hydrogen bonding highlights

I have really enjoyed the Hydrogen Bonding conference this week. There are about 120 people which is a good size.
The diversity of the topics covered is a testimony to just how ubiquitous, important, and challenging hydrogen bonds are. The participants ranged from old timers who have probably been to all twenty conferences to many newcomers like me who tthis was their first time. People were quite friendly and any contested discussion was quite cordial. I also enjoyed the lack of hype or sef-promotion. I got a lot of positive feedback for my simple model, showing that the field is quite open to new people and new approaches, particularly from physicists.


Here is a somewhat random commentary. If you want more details, ask.

Double proton transfer in porphycenes. This proceeds via a concerted mechanism and quantum tunneling is clearly present. Jacek Waluk's group has some beautiful results.

Water. [The basic questions never go away!]
Ali Hassanalli gave a nice talk about Proton transport through the water gossamer. Car-Parrinnello simulations show the preponderence of directed rings of water molecules.
What is the local environment around an hydroxide anion (OH-)?
There was animated discussion about whether it is H-bonded to four or five of the surrounding water molecules.

High resolution infra-red (IR) spectroscopy is a powerful probe that was widely featured. One problem is identifying lines with specific vibrations. This often requires comparison with an electronic structure calculation (usually DFT-based or MP2). I think this usually means all the frequencies have to be scaled by some mysterious factor (0.94-0.99). For strong hydrogen bonds identifying the lines is difficult.

Computational chemistry featured heavily as a tool for understanding specific experiments and to gain insight into the nature of bonding in specific complexes. Perhaps MP2 [Hartree-Fock + second order perturbation theory] featured more than DFT [with some dispersion corrected functionals]. CCSD was mentioned but not higher level methods (e.g. CAS-SCF). They would certainly be possible for some of the small molecules studied. Atoms In Molecules Theory featured significantly. Energy decomposition analysis featured some. As an ignorant outsider I am somewhat skeptical about how robust the decomposition is and so how much reliable insight it can provide.

Halogen bonding. Some of the motivation for resurgence of interest is that 20 per cent of drugs actually involve halogen bonds. The 1969 Nobel Prize in Chemistry, was partly for halogen bonding.
[Aside: they are also present in superconducting organic charge transfer salts, along with practically every other type of bond].
It seems to have some similarities to hydrogen bonding. Theoretically this can be seen within Weinhold's Natural bond orbital donor-acceptor picture.  It was claimed it can be viewed as a 4 electron, 3 orbital bond. However, no real evidence for this was actually presented. I think this could be nicely shown with a CAS-SCF treatment looking at different active spaces.
Interestingly, they have a stronger tendency to symmetric bonds than H-bonds.

New types of bonding: Halogen bonds, Charge Inverted Hydrogen bonds, Carbon bonds, Agostic, .... Various speakers claimed to have discovered new types of bonds. Some people think this is just nomenclature. Some suggested the "new" bonds" were really just "old bonds." Others think that there are fundamental issues here. It is interesting that Charles Coulson [my hero] actually claimed that bonds, like molecular orbitals, are a figment of our imagination and cannot be rigorously defined from a quantum point of view. I think the advent of natural orbitals and Atoms in Molecules Theory show that he was wrong on both counts.

Biomolecules featured some, but not as much as might have been expected, given the crucial role, H-bonds play in biomolecular function. Several participants who work on biomolecules told me they really enjoyed being at a meeting with a more fundamental focus.

Vibrational circular dichroism and Raman optical activity featured significantly. They is sensitive to the chirality of molecules and so can detect alpha-helices and beta-sheets in biomolecules. This is particularly important for studying carbohydrates [sugars and starch] which make up a large fraction of the planet's biomass. Unlike proteins they do not crystallise and so there is very little structural information about them.
Interesting to me, is that there are fundamental questions about the spatial extent of coherence [quantum or classical?] of vibrational excitations along alpha-helices and beta-sheets.

Microsolvation. How many water molecules does it take to make an acid? Four! i.e., if you add four water molecules to HCl the latter will dis-associate.

No comments:

Post a Comment