Oil and Water Do Mix

Sometimes, all that's needed to overturn generations of conventional chemical wisdom is a vigorous stir.

With the firmly entrenched notion that oil and water don't mix, chemists frequently approach organic reactions from a like-needs-like perspective. Water-insoluble reactants require at least some amount of an organic solvent, such as benzene, in which to dissolve before the reaction can proceed at a decent rate.

In recent decades, though, water has shed its reputation for being incompatible with organic chemistry. It's already received attention for its attractiveness as an environmentally friendly solvent. Even more provocative is the ever-accumulating pile of cases where plain old, unadulterated H₂O actually steps up the pace of some organic reactions previously thought dependent on an organic medium.

Scripps Research Institute chemistry professor and Nobel Laureate K. Barry Sharpless and his colleagues have thrust this unexpected phenomenon into the spotlight. Stirring insoluble reactants in water into a vinaigrette-like suspension, they find, makes a number of well-loved but sluggish organic reactions leap into furious action.

Though an abundance of literature exists showing that many organic reactions thrive in aqueous environments (Mannich condensations, for example), the solutions are generally very dilute and essentially homogeneous--that is, the reactants are actually very slightly soluble in water, and small quantities are completely dissolved. But these aren't synthetically practical conditions because the quantities of product generated would be too small.

A number of researchers say part of what's new about the Sharpless work is its focus on the effect at synthetically useful concentrations, in a heterogeneous--undissolved--environment. "He's really forming immiscible phases," says William L. Jorgensen, a chemistry professor at Yale University. "He's making observations that are quite profound."

A case in point: A cycloaddition reaction between quadricyclane and dimethyl azodicarboxylate (DMAD) normally takes several hours to complete in toluene, and two days with no solvent. But in plain water, the reaction runs its course in 10 minutes.

The Scripps team, including chemistry professors M. G. Finn and Valery V. Fokin and postdoc Sridhar Narayan, detail the behavior in a collection of important organic reactions--ene reactions, nucleophilic substitutions, Claisen rearrangements, and Diels-Alder reactions. Because the insoluble reactants float on top of water before they're stirred in, Sharpless' group has given the method the name "on water" (Angew. Chem. Int. Ed. 2005, 44, 3275). The effect isn't limited to liquid reactants--a ground-up solid can react with a liquid in the same speeded-up manner. 

WATER-BASED organic chemistry is the traditional purview of green chemists, whose goal is to replace syntheses using toxic apolar solvents with those using water. For years, they've been observing that some organic reactions do well in water, says Chao-Jun Li, a chemistry professor at McGill University. Li details this in an upcoming issue of Chemical Reviews.

"Most people working in organic chemistry in water have noticed the phenomenon," notes Li. "But although we all knew this, we never really did a detailed study."

Why has it taken this long for the effect to generate a buzz outside the green chemistry community?

"We're trained to put stuff in solution," Finn speculates. It's counterintuitive, he says, to expect two reactants to react faster in a medium in which they don't dissolve than in one without any solvent at all.

"I think it is a real breakthrough in synthetic organic chemistry," says Jan B. F. N. Engberts, a chemistry professor at the University of Groningen, the Netherlands, who has both studied and followed the phenomenon (Nature 2005, 435, 746). "Apart from the excellent synthetic results and the often easy workup, it goes without saying that water really is a 'green solvent' and very cheap."

Organic chemists cite Columbia University chemistry professor Ronald Breslow's hallmark 1980 study of the Diels-Alder reaction--the ubiquitous synthetic workhorse used in pharmaceutical and industrial chemistry--as the genesis of interest in organic chemistry in water. Breslow's group was attempting to carry out a Diels-Alder reaction inside a cyclodextrin cavity, but the researchers found to their surprise that their experimental control--doing the reaction in plain water--went even faster.

In 1990, Jorgensen and his then-graduate student James F. Blake simulated Breslow's Diels-Alder reaction on a computer. They showed that the acceleration in water arises not only from the hydrophobic effect (the tendency of organics to turn away from polar water molecules), but also from a strengthening of hydrogen bonds from water to the dienophile's carbonyl oxygen.

In the ensuing years, other researchers, including Joseph J. Gajewski, a chemistry professor at Indiana University, and Engberts, studied the Diels-Alder reaction in water, coming to the same conclusion.

Chemists familiar with the history of aqueous organic reactions also often cite Paul A. Grieco, a chemistry professor at Montana State University, in Bozeman, Mont., who performed one of the first on-water-like experiments, with a Diels-Alder reaction of heterogeneous mixtures in plain water. Grieco proposed that the hydrophobic ends of reactant molecules cluster together to form micelles. Some experimental evidence supports that. Recently, researchers found that alcohol forms micelles when it's stirred in water (Nature 2002, 416, 829).

Sharpless' team began to suspect water's effect on organic reactions over several years during their research on "click chemistry," their system of ideal covalent and irreversible reactions (C&EN, Feb. 16, 2004, page 63). "It just became more and more obvious that we should be leaving out every solvent other than water, in contrast to the natural inclination of adding a cosolvent," Sharpless says.

In a series of experiments, keeping their reaction volumes constant and using the same magnetic stirrer, they meticulously documented the on-water effect.

Other researchers are now taking note. B. Mikael Bergdahl, associate organic and bioorganic chemistry professor at San Diego State University, and his colleagues report that the popular carbon-carbon double-bond-forming Wittig reaction does well in water (Tetrahedron Lett. 2005, 46, 4473).

The effect, however, is by no means general. Researchers observe that many reactions don't respond to water at all. Which raises the pressing question: What's behind the on-water effect?

RESEARCHERS hypothesize that a combination of effects are involved.

The dilute, homogeneous solution phenomenon that gained attention with the early Diels-Alder work--in which organic reactant molecules orient themselves away from polar water--likely contributes. As Sharpless says, "When you force things that normally stay away from each other to be next to each other, the situation gets pretty intense."

But that's not the whole story. Molecular interactions at the interface of reactant suspension droplets and water undoubtedly play a major role, scientists believe. Stirring creates an oil-on-water emulsion of nanometer-scale droplets, says Christian Reichardt, a chemistry professor retired from the University of Marburg, Germany, and author of what some chemists refer to as the "bible" on solvation, "Solvent and Solvent Effects in Organic Chemistry."

"Why the reactions taking place in such nanodroplets are much faster is not quite clear," Reichardt says.

The answer could lie with water's uniquely high "cohesive energy density"--a measure of how much the liquid wants to stay together. At the boundary between water and an insoluble organic droplet, water molecules are forced to interact with molecules other than themselves, which they don't like to do and which generates species with high-energy ground states. This may speed up the reaction between the organics in an effect analogous to heating.

The relationship between organic reactants and water is apparently unique. If the on-water effect was merely due to apolar reactants' dislike of polar solvents, then other polar solvents should in general produce the same effect as water. But polar solvents like dimethyl sulfoxide do little to boost reaction speed. And Sharpless' test of the reaction of quadricyclane and DMAD in perfluorohexane, a solvent that resists dissolving organic molecules, did not perform much better than the slow, no-solvent reaction.

In another mechanistic puzzle, reactions performed by Sharpless' group using heavy water, D₂O, were slower than with water by a factor of four. "This far exceeds the effect expected for changes in hydrogen-bond donation and may even be outside that expected for the cohesive energy-density effect," Gajewski says. A possible explanation is D₂O's viscosity, which is much greater than that of water.

A new wave of research, from the Scripps group and others, aims to answer those questions. They're already at work determining the importance of the size of emulsive droplets on reaction rate and on precisely what constitutes "vigorous" stirring. "The rate of the reaction does depend on how well you stir it," Finn notes.

Jorgensen is modeling the quadricyclane-DMAD reaction on water. And more measurements of reaction kinetics are refining the picture.

The Scripps group wants other organic chemists to jump in the game. They can send their experiences with on-water types of reactions to onwater@scripps.edu, for which the group will give the contributors public credit.

Water's unique power could transform the way people think about organic chemistry, Sharpless predicts. "The future of chemistry is really bright if we keep it simple."