Posted on Feb 8, 2012
For print, a PDF version is available here.
Last edited August 2012: polishing, per requests in comments.
The idea is interesting but I would much rather see the worked-out equations than cartoons+hand waving. The heart of the discussion, which is that random temperature fluctuations produce randomized rotations, effectively canceling out electrostatic attractions, is a simple exercise in statistical mechanics. It’d be an important problem to tackle and the cartoons are not a compelling approach.
Furthermore your arguments are nearly good enough to make such a strong conclusion. “In real liquids, permanent dipole-dipole interactions actually play less important role than induced dipoles”? You cannot pick and choose your experiments just to emphasize this fact while ignoring the VAST amount of conflicting literature. Your statement is only feasibly valid in molecules large in 1 or 2 dimensions, (polymers, beta-sheets). If you do a good enough job working out the details I think your ideas may be quite important, but this webcomic is, IMO, bad science.
“Furthermore your arguments are nearly good enough to make such a strong conclusion”
Thanks for the feedback. Mea culpa: I should have clarified that none of this is “my idea” (and certainly none of it is new science). The p-chem with derivations can be found in most phys chem textbook (e.g. Atkins). For this reason, I’m curious about the conflicting literature – can you point me to some examples in isotropic liquids where contributions from permanent dipole is more important than LDF? (Water is the only one that I remember being an exception (even after discounting for h-bonding), with its high polarity and low polarizibility.)
I’m also confused about the statement regarding polymer/beta-sheets. Those are precisely where I’d expect the contribution of permanent dipoles being non-negligible, since there is restriction of molecular motions. Could you clarify?
Interesting points, I’m currently embroiled in a discussion of whether carbon dioxide can serve as an electrophile to induce dipoles in a phenolic ring. And if it can, what is the strength and the longevity of the interaction. I guess we need to try to dissolve phenol in liquid CO2 and see what happens.
Off top of my head, I’m going to guess that it *would* form the phenolic carbonate, in a reaction analogous to amine scrubbing. Back of the envelop estimate say that this reaction is overall exothermic, and the equilibrium would be driven to the carbonate by virtue of CO2 as solvent.