THE EXPERIMENTAL MANIFESTATIONS OF CORNER-CUTTING TUNNELING

A family of potential energy surfaces has recently been developed, based on a linear, 3-body model, which permits the calculation of rate constants that accurately mimic a large body of data for hydride transfer reactions in solution. Two of the parameters of the symmetrical surface have now been systematically varied so as to alter the importance of hydrogen tunneling, and the effects on measurable parameters have been examined. On all reasonable variants of the surface, at temperatures approximately 300 K, most hydride transfer events occur by tunneling, and tunneling generally accounts for a large fraction of deuteride and tritide transfer events as well. At this temperature most tunneling occurs at donor-acceptor distances slightly larger than those of the transition states (corner-cutting). Tunneling increases the rate of hydride transfer by about a factor of 10. The isotopic difference in Arrhenius activation energies, E(a)(D) - E(a)(H), is increased by about 1 kcal mol-1 and the ratio of preexponential factors, A(H)/A(D), is reduced by tunneling. However, it appears that completely unambiguous experimental proof that tunneling occurs would be impossible to obtain at approximately 300 K, although tunneling becomes clearly evident at much lower temperatures. The commonly observed steric magnification of the kinetic isotope effect requires no special modification of the potential energy surface. It can be reproduced by increasing the equilibrium donor-acceptor distance. It is partly due to tunneling and partly due to isotopic differences in the variational transition states. An increase in kinetic isotope effect when the hydride donor is modified so as to increase the equilibrium constant is a symptom of tunneling. The most reliable experimental indications of tunneling obtainable at approximately 300 K are probably a combination of these effects, without significant exceptions. In systems with only one isotopically substituted hydrogen, the Swain-Schaad relation appears to hold as well in the presence of tunneling as in its absence. It appears to be a reliable way of estimating a tritium isotope effect from the corresponding deuterium isotope effect, but it is not useful for identifying the origin of the effects.