Results from one of the molecular dynamics simulations in the paper, showing a 400-fs time trace of the atomic nuclei. Motion of the liquid is shown in blue, and DF in white. The interlaced starlike shapes arise from significant vibrational excitation in the newly formed DF

Today’s issue of Science contains a report that I wrote along with colleagues at Rutherford Appleton Laboratory (UK) and Heriot-Watt University (UK), providing atomically-resolved detail describing how a chemical reaction happens in a liquid.

This is an area that I’ve been working on for awhile now – specifically trying to figure out how atomic and molecular energy transfer works. I first got interested in this stuff a few years back, when I showed that the product yields of the well-known “hydroboration” reaction in organic chemistry are very sensitive to how quickly energy in the reacting molecule dissipates into the surrounding liquid. This result was a surprise for me: all chemical reactions involve energy being deposited into molecules, but it’s generally neglected in descriptions of how chemical reactions in liquids work. The conventional thinking is that molecules can’t hold on to their energy long enough for it to make much of a difference. This study showed that the molecules involved in the reaction were holding onto their energy deposits for long enough to have a profound effect on the identity of the reaction products.

People have known for some time that if you deposit energy into a molecule using a laser light pulse, it can hold onto its excitation for quite awhile, even in a liquid.  But most of the liquid chemistry occurring in nature happens under so-called “thermal” conditions, where there’s no laser pulses involved. Under thermal conditions, the general assumption is that energy deposited in a molecule by way of a chemical reaction is dissipated so fast that it’s not really worth worrying about, and couldn’t possibly affect the reaction outcome. Today’s Science paper combines some very fast time-resolved spectroscopy and state-of-the-art molecular dynamics simulations (that was my role!) to examine a simple chemical reaction in a liquid that produces a DF molecule. The results provide compelling evidence – yet again – that even under thermal conditions, the reaction product holds onto its energy for a rather long time. What’s particularly surprising about this result is that we intentionally chose a so-called ‘strongly coupled’ system, where we expected this not to be the case. But alas, even under these conditions, DF is produced with lots of excitation (nearly three quanta!), and the excitation persists. The simulation framework (run on a massively parallel supercomputing architecture described further in an arXiv paper and implemented in CHARMM and TINKER) showed that the newly formed DF holds onto its energy while it searches out hydrogen bond partners with its neighboring liquid molecules. The simulations also showed that DF’s time-resolved experimental spectra is the result of two effects – loss of energy, and formation of hydrogen bonds – each of which wants to move the spectra in opposite directions, and therefore mostly cancel each other out.

I’m excited about this stuff, especially because I’ve started thinking about whether it transfers over to larger molecules in organic and biochemistry. Results generated using the same simulation framework recently showed that these sorts of effects can actually lead to solvent-induced enantioselectivity for much larger molecules, which is a fascinating result that we’re looking at in further detail.