Skip to Content

Chemical News

Dissolving One Way And Another

All right, fellow chemists, you’ve got this hydrophobic/hydrophilic thing down, right? I’m glossing over the fact that our intuition about those things can be wrong, as can much of the software used to estimate it – we at least know about these concepts and have a physical picture of compounds that like to dissolve in water versus the ones that don’t.

But wait! There’s more! (As the old cable-TV ads used to say when they were selling you no-sharpen knives and the like). There are finer gradations of compound behavior in water, and the terms kosmotrophic and chaotropic are worth adding to your vocabulary if you don’t have them at hand. (Plus, you’ll stand out at parties even more than you do now). This new review goes into good detail on them, but the difference is easy to state: kosmotrophic solutes are those that tend to produce or stabilize ordered water structure, and chaotrophic ones break water structure up.

If you think about what happens when some species gets out into bulk water, there’s always a cavity formed in what used to be the hydrogen-bonded water network. It has to stretch out to accommodate the new molecule/ion/whatever, and it’s the energetic cost of doing so that determines how soluble a given species is. Kosmotrophic species tend to be very small, with high charge density (think fluoride and sulfate). The water shell around these things is quite oriented, which means that compared to the normal structure of bulk water, things are actually more organized than they were to start with. (You may now add an unfavorable entropic term to your energy calculations!) But this also happens around true hydrophobic solutes, since the waters around them can form no hydrogen bonds with the solute molecules. Their orientations become more organized as well (solvation is entropically unfavorable). Giving such a hydrophobic species a chance to desolvate, to get out of the bulk water and into, say, a protein’s binding site instead is entropically favorable from the standpoint of the water molecules involved.

But consider, as the authors of the paper have, what happens when you have a big borate cluster anion (dodecaborate, with a -2 charge). That is not a hydrophobic species – but it still likes to desolvate (and associate with, for example, the cavity of a cyclodextrin). That’s the same sort of thing you’d expect from a big round hydrophobic molecule, but it turns out that the thermodynamics are completely different. The borate cluster, when it dives into the cyclodextrin, has a big favorable enthalpic change, whereas the entropic change is actually unfavorable. That’s because the water molecules around it were actually quite disorganized (it’s a chaotrope), and the bulk water that forms up behind it as it heads into the cyclodextrin cavity is actually more orderly than before. You get the same sort of overall free energy change for a chaotrope and a hydrophobe on desolvating into a nonpolar cavity, but you get there through different means.

Ang. Chemie/Wiley Publishing

The authors propose that the traditional list of chaotropic ions (things like fluoroborate, perchlorate, and iodide) be augmented with “superchaotropes” that have noticeably larger effects. Those are your decaborates and polytungstates (and other polyoxometalates): large charged cluster ions that show very large entropic effects on structured water indeed. The chart shows the relationships between these things; as you can see, there are several factors at work simultaneously to give a range of behaviors. As the example above shows, some of these things look the same, but work in different ways.

It’s a neat thought experiment to extend solvation all the way out to the right-hand side of the scale: a cavity. You get those forming in ultrasound baths, and they’re notoriously weird (temperatures of over 15000K, pressures of up to 1000 bar, for a very short period of time in a very small space). I’d never thought of a cavity, a speck of actual vacuum, as a solute, but perhaps a vacuum is the ultimate hydrophobic species, since there’s absolutely nothing – as in literally nothing – for the water molecules around it to interact with.

As medicinal chemists, we’re probably not going to be working with many polyoxometalate drug candidates. But these concepts are definitely worth thinking about, even if we’re not heading for the superchaotrope part of the scale. (The interaction of such species with proteins and membrane, though, is of considerable theoretical interest – the review linked above has some leading references). The whole story of ligands binding to proteins is crucially influenced by the solvation/desolvation behavior of those ligands, of course, and this gives you a more detailed mental picture of it. It’s also useful to keep these concepts in mind as you think about what conditions are really like in the cytosol and inside the nucleus (all the recent excitement about phase-change condensates in cells comes in as well). I do wonder, as drug candidate molecules push into new, higher-molecular-weight higher-complexity space, if we might see some unusual solvation behavior start to show up that isn’t so important with smaller species (I have no particular evidence in mind, just sheer speculation).

Water! We think of it in everyday life as the default solvent, the standard by which we measure everything. But it’s very strange stuff indeed, and it just gets stranger and more complicated the longer you look at it. . .



12 comments on “Dissolving One Way And Another”

  1. Chad Irby says:


    Could it be possible that a kosmotrophic additive might actually lower the volume of water by making the molecules line up?

    We’ve seen odder things.

    1. AR says:

      Wouldn’t that increase the volume like ice structures do?

    2. Quintin says:

      I think we’re all familiar with the fact that mixtures of ethanol and water have a smaller volume than their pure constituents. You’d think ethanol would be a chaotropic solute, being basically water with a little hydrophobic tail. So yeah.

  2. Peter Kenny says:

    I’m guessing that looking at the electrostatic potential minima (Vmin) associated with the hydrogen bond acceptor atoms of these anions will shed some light on interactions with water molecules. I’ve linked an article (see Figure 4) in which we looked at Vmin values for nitrate anion. The calculations suggest that each oxygen atom is more likely than nitrogen to accept a hydrogen bond.

  3. Old Timer says:

    I think it’s kosmotropic–you know, so you sound credible at those parties 🙂

    1. Thomas Lumley says:

      -tropic, moving towards vs -trophic, eating.

      I’d quite like a chaotrophic reagent

  4. Uncle Al says:

    This conceptualization begs a breathless press release containing no intellectual content whatsoever (superintellophobic cavity, e.g., Social Justice) demanding female and diversity population of medicinal chemistry quietly exceeding 100% – followed by a long line of thermodynamics with a d’Alembert operator lest anybody attempt to make sense of it. Exemplar:
    … Luce Irigaray sources physics’ defects.

    One can imagine any number of predatory journals bidding for the rights.

    1. zero says:

      Go back to your cave. Breathless press releases have nothing to do with politics or social change and everything to do with ad revenue (or in some cases attracting investment).

      As someone with no meaningful use for the information Derek provides in this blog, I still appreciate the change in perspective. This post provides an unusually accessible insight into the behavior of small molecules in water that I found enjoyable.

  5. cynical1 says:

    And did they reconcile these ideas with the Rule of 5? I don’t see anything about multiples of 5 up there. (Forgive my sarcasm…………)

  6. Colloid crystals in water: particles spaced and organized on a scale of microns.

    That, and lots of other fascinating stuff on this page:

    Some of it sounds implausible, but they are talking about easily observable phenomena, e.g. the centimeters-long bridge of water you can create between two beakers with a voltage applied between them.

    Over a decade ago, I wandered into a lab where they were working on long-range structures in water (I’m sure they’ve published by now) – IIRC they had solvated species with lots of charges per particle/molecule with interesting numbers, e.g. 5+ on one particle and 7- on the complementary particle (I may even remember that they got up to over 10 charges per particle). They got very complicated long-range structures. From something they said, I replied, “Huh, that reminds me of Penrose tiles” and they looked at each other significantly.

  7. My post didn’t appear – let’s try without the URL (it’s linked on my name).

    The page looks like pseudoscience, but they’re reporting on directly observable experiments that seem directly relevant to this article. For example, water next to a hydrophilic membrane can be ordered enough to exclude particles and molecules for hundreds of microns.

    Colloid crystals – individual particles spaced and latticed on a scale of microns. (That’s how I found the page; I remembered reading about those years ago.)

    The centimeters-long bridge of water that can be stretched between two beakers with a voltage applied.

    Over a decade ago (I’m sure they’ve published by now) I wandered into a lab that was dissolving nanoparticles(?) with large and interesting numbers of charges apiece, e.g. 7+ on one particle and 15- on the other. They made some very interesting partially ordered colloid crystals. I said, “That reminds me of Penrose tiles” and they looked at each other significantly.

  8. Project Osprey says:

    Different take on the same issue. Also from this year.


Leave a Reply

Your email address will not be published. Required fields are marked *

Time limit is exhausted. Please reload CAPTCHA.