A fellow researcher, working over at The Competition, sent along a couple of good questions. He's a biologist, and was reading the posts here earlier this week about compound repositories. He writes:

"I would like to pose to you a question I have tried for years to get chemists to address: why has no one yet come up with a better "universal" solvent than DMSO for dissolving compounds for routine screening in assays? As you point out, DMSO has numerous liabilities for this purpose (perhaps more on the biological side), and yet we all continue to use it routinely, because there doesn't seem to be a better alternative. In the various screening labs in which I have been involved for the past dozen or so years, we have tested a number of possible alternative solvents on an ad hoc basis, none of which seemed any better. Surely modern organic chemistry can do better?"

All I can offer my colleague is this: modern organic chemistry may not be quite as powerful as you've been led to believe. Perhaps the problem is that you've been listening to too many of us modern organic chemists. We do tend to go on.

He's right that DMSO has its down side, and there are some that I didn't even mention. For one, DMSO and air make for a decent oxidizing system, enough to cause trouble in electron-rich molecules. Things will start to change color on you in a DMSO solution that's been left open. And it does have its biological problems. Too much DMSO in an assay system will cause the proteins involved to change their conformations, probably inactivating them. At the very least, your data start to go haywire.

Is there anything we mighty chemists can do about this? Well. . .actually. . .no, probably not a whole lot. The problem is, anything that has "universal solvent" properties is going to have "universal denaturant" properties when it comes to large biomolecules. Proteins, carbohydrates, and nucleic acids are made to hang around with water. Anything that isn't water is going to cause trouble, sooner or later.

Coming up with a solvent that acts just like water, but isn't, may well be impossible. Water's just too weird. Its boiling point and viscosity are way off the estimates you'd get from looking at related things like ammonia and hydrogen sulfide, due to its extraordinary hydrogen-bonding powers. And it's those bonds that do the trick with biomolecules. It's sui generis - there's no other molecule that small that can do hydrogen bonds that strongly, at those angles, in two directions at once.

DMSO gets by because it's also small (although not as small as water.) It's the smallest sulfoxide possible, so it has the most character. The key is that the sulfur and oxygen atoms in the sulfoxide bond have a lot of charge on them - the oxygen's nearly a minus charge; the sulfur's nearly a plus. That dipole lets it really line up with any polar groups a molecule might have, and its two methyl groups give it a chance to dissolve some hydrocarbons that water won't accept. (Larger sulfoxide analogs just add greasiness, and are less powerful solvents. That's the wrong direction, and you can't go any further the other way.)

And all the other attempts at DMSO substitutes tend to follow that same path, things that are polar because of their high dipole moments. Some of the also-rans are N-methylpyrrolidone (NMP), DMPU, and the toxic HMPA. None of them are as good as DMSO, and they all suffer from its disadvantages. There may well be some funky structures out there that haven't been given a fair hearing, but I sure can't think of many right now. I'm afraid that we're just going to have to live with DMSO, and respect water's magical powers for what they are.