Simulated liquids point to new solutions.

Simulated liquids point to new solutions

Detergents lifting dirt from clothes, paint disappearing into a turpentine-soaked rag and cellular proteins folding into their biologically active shapes all depend on the way chemicals dissolve in various liquids.

Two theoretical chemists have run new computer simulations of water and non-water liquids that help clarify why some liquids dissolve specific solutes better than others. They say their work could give molecular biologists a better understanding of how proteins fold and function, while helping materials scientists predict how various ingredients will behave in new mixtures.

Liquids fall into two broad categories--polar (such as water) and nonpolar (such as turpentine) -- which differ in the abilities of their molecules to form hydrogen bonds among themselves and with other molecules. Although the intermolecular attractive power of a hydrogen bond reaches only about one-tenth that of the covalent bonds that link the atoms within a molecule, the weaker bonds help determine such important properties as a liquid's boiling point, viscosity and ability to dissolve specific solutes.

Any chemistry student can explain why table salt dissolves well in water but not so well in hexane, a nonpolar organic solvent. Electrically polar water molecules, with their segregated regions of positive and negative charge, cluster around the salt's charge, sodium and chloride ions. Hydrogen bonding then helps the ion-centered clusters to merge into the surrounding water. Because the nonpolar hexane molecules do not form such clusters, they don't pull salt into solution as readily.

But even professional chemists have difficulty accounting for why nonpolar liquids outdo water in dissolving electrically neutral gas molecules such as methane. Lawrence R. Pratt of Los Alamos (N.M.) National Laboratory and Andrew Pohorille of the NASA Ames Research Center in Mountain View, Calif., set out to answer this basic question.

Using computer simulations of water and five nonpolar solvents, they calculated "the likelihood of finding, at an arbitrary point in the solvent, an atomic-sized cavity that could accommodate the solute." Though water has more overall cavity space than the nonpolar solvents, its spaces are distributed in smaller, less flexible packets, the chemists report in the June 20 JOURNAL OF THE AMERICAN CHEMICAL SOCIETY. Thus it appears that water, compared with a nonpolar solvent such as carbon tetrachloride, would have more difficulty rearranging its molecules to make room for small solute molecules such as methane.

Most previous models of solubility based on solvent cavities have portrayed molecules as hard spheres, a simplification that works well for nonpolar solvents such as hexane, notes theorist Frank H. Stillinger of AT&T Bell Laboratories in Murray Hill, N.J. But for probing subtle differences between the dissolving powers of water and nonpolar solvents, models that include the effects of hydrogen bonds on cavity size should prove more fruitful, Stillinger and Pratt assert.