Skirting the Transition State

Now nearly three-quarters of a century old, transition-state theory dominates how chemists think about the way chemical reactions proceed. The theory suggests that reactants are separated from products by a critical chemical configuration--the transition state--that defines the reaction's rate. But in certain cases, "a chemical reaction can proceed via a path that completely bypasses that reaction's transition state," says Joel M. Bowman, a theoretical chemist at Emory University, Atlanta.

Bowman bases that suggestion on recent work he's published with physical chemist Arthur G. Suits of Wayne State University, Detroit. Using high-powered computational work and detailed experimental studies, their team has demonstrated that photoexcited formaldehyde (H₂CO) can decompose to H₂ and carbon monoxide via a path that skirts that reaction's well-established transition state entirely [Science, published online Oct. 21, http://dx.doi.org/10.1126/science.1104386]. Because "there's nothing special about formaldehyde," Suits suggests that such transition-state-skirting pathways may not be at all unusual.

"Do these observations turn transition-state theory on its ear? I don't think so. But they do point out a key limitation of what is an old but very robust theory," comments physical chemist C. Bradley Moore, who is Northwestern University's vice president for research.

Formaldehyde decomposition has long been a model system for those studying transition-state theory because the reaction is simple enough to treat with high-level theoretical models, and the products are easily detectable. During the 1980s and early 1990s, Moore and his coworkers characterized the dynamics of formaldehyde decomposition in great depth. In particular, his lab showed that the reaction proceeds through a skewed, high-energy transition state in which both hydrogen atoms are on the same side of the CO. The neighboring hydrogen atoms then react to give H₂, which bounces away, leaving CO behind.

TO GLEAN an even more detailed picture of this model reaction, Suits used a high-resolution ion-imaging technique to determine the rotational and vibrational energies of the H₂ and CO resulting from the decomposition of photoexcited formaldehyde. His detailed measurements revealed that decomposition proceeds through two distinct pathways, each of which gives rise to products with different rotational and vibrational signatures: The dominant pathway goes through the reaction's well-defined transition state and yields rotationally excited CO and H₂ with little vibrational energy. Another pathway--the existence of which Moore had suggested but never was able to prove--produces CO with little rotational energy and highly vibrationally excited H₂.

To probe the nature of this second, unusual pathway, Suits turned to theoreticians. Using a brand-new global potential-energy surface for H₂CO developed in collaboration with Lawrence B. Harding of Argonne National Laboratory, Bowman and coworkers performed high-level calculations to create a movie of this second pathway. The movie reveals that one of H₂CO's hydrogen atoms breaks off and roams around before bumping into the second hydrogen atom and forming H₂. At no point in this second pathway does the reaction go through its transition state, Bowman notes. At sufficiently high energy, as much as a third of the decomposing H₂CO molecules go through this pathway, he estimates.

It remains to be seen how common it is for reactions to skirt their transition state. "It may not be that unusual," Suits suggests. He points out that because hydrogen is so lightweight, it's the atom most likely to do the roaming required to sidestep the transition state. Such a roaming hydrogen may well be able to react with an atom other than hydrogen, but reactions that eliminate H₂ may be the most common examples, he admits.

Although these new observations don't undermine traditional transition-state theory, Bowman and Suits argue that they are part of a growing body of work suggesting that the theory can not thoroughly describe chemical reactivity. Recent work from the groups of Stephen J. Klippenstein at Sandia National Laboratory and William L. Hase at Texas Tech University has hinted that transition-state theory may not be able to thoroughly explain other simple chemical reactions. "A lot of things can happen during chemical reactions that no one ever expected from transition-state theory," Hase notes. He hopes that this growing body of work will challenge theoreticians to search for ways to incorporate these observations into our understanding of chemical reactivity.