Isotope effects in proton-transfer equilibria

Large hydrogen isotope effects were found experimentally very soon after the discovery of deuterium in 1932, and there is now an extensive literature on the subject, recently supplemented by work with tritium. Much of this relates to the kinetics of reactions involving proton transfer, but there is also a large amount of information on acid-base equilibria. A recent review estimated that about 300 papers on isotope effects are now published each year, and a considerable proportion of these relate to the isotopes of hydrogen. The present chapter will be devoted to the effect of isotope substitution on acid-base equilibrium constants, while Chapter 12 will deal with kinetic isotope effects in proton transfer reactions. 

It was recognized by La Mer and Noonan (1939 Noonan and La Mer, 1939) that the comparison of chemical processes in isotopically different solvents involves a general medium or transfer effect, in addition to any effect due to the occurrence of exchange equilibria. The point was further developed by Kingerley and La Mer (1941) in relation to acid-base equilibria, but these authors also showed that in many cases the transfer contribution is of minor importance. Because of the relative size of the two contributions, most workers have paid little or no attention to the transfer contribution. Precisely the opposite view, namely that the entire or, at least, a large part of the solvent isotope effect on proton transfer processes was due to the transfer effect, was propounded by Long and his group (Halevi et al., 1961) but has been abandoned in their more recent work. 

Measurements of the deuterium isotope effect for unsymmetrical di-Schiff bases fully confirmed the interrelation between proton transfer equilibria in both intramolecular hydrogen bonds.46... 

Kinetic Acidities in the Condensed Phase. For very weak acids, it is not always possible to establish proton-transfer equilibria in solution because the carbanions are too basic to be stable in the solvent system or the rate of establishing the equilibrium is too slow. In these cases, workers have turned to kinetic methods that rely on the assumption of a Brpnsted correlation between the rate of proton transfer and the acidity of the hydrocarbon. In other words, log k for isotope exchange is linearly related to the pK of the hydrocarbon (Eq. 13). The a value takes into account the fact that factors that stabilize a carbanion generally are only partially realized at the transition state for proton transfer (there is only partial charge development at that point) so the rate is less sensitive to structural effects than the pAT. As a result, a values are expected to be between zero and one. Once the correlation in Eq. 13 is established for species of known pK, the relationship can be used with kinetic data to extrapolate to values for species of unknown pAT. 

Strictly speaking, the inverse isotope effects do not prove the mechanism of Eq. (37) because rate effects per se can tell us only about differences between reactants and transition states, not about equilibria preceding the rate-determining step (Melan-der and Myhre, 1959). We can confidently conclude that proton transfer was essentially complete before the transition state was reached, since if it were still occurring at the transition state a normal primary isotope effect on hydrogen transfer would result (kTi o/kiy Q >1). Thus the transition state is effectively that for decomposition of SH+ whether or not SH+ was formed in a discrete equilibrium step. 

The subject of isotope effects in H2O-D2O mixtures has been treated in some detail because kinetic studies in these solvents have been recently used extensively to obtain detailed information about the nature of transition states in proton-transfer reactions. The problems involved are essentially the same as arise for equilibria, in particular with respect to transfer effects, and there is the added difficulty that fractionation factors and transfer activity coefficients for the transition state must either be guessed at by analogy with stable species, or derived from the kinetic measurements themselves. On the other hand, the numerical value of /c /k is often more favourable for distinguishing between different possibilities than are the values of K /K commonly met with in equilibrium systems, and there is also another piece of experimental information, the so-called product isotope effect, which is sometimes helpful. These kinetic problems will be discussed briefly in the next chapter. 

The only isotope effects which are usually of significance in electroorganic mechanism considerations are those involving H and D in (a) primary kinetic isotope effects, (b) secondary solvent isotope effects where reactions are compared in pure H2O and D2O or pure h- and rf-alcohols, and (c) in prior protonation equilibria, e.g., with ketone reduction. Primary kinetic isotope effects having a magnitude of > 2.5 may be expected in reactions that involve a rearrangement with proton participation, e.g., keto-enol tautomerism prior to an electron transfer step. Most other H/D isotope effects arise from protonation equilibria prior to the rate-controlling electron transfer step (e.g., in ketone reduction RR CO + RR COH ) and the isotope effect is... 

---