Solvent isotope effects as a probe of general catalysis and solvation in phosphoryl transfer

Abstract

Phosphoryl transfer to methanol from tris(p-nitrophenyl) phosphate (PNNN), methyl bis(p-nitrophenyl) phosphate (PMNN), and dimethyl p-nitrophenyl phosphate (PMMN) exhibits general base catalysis by acetate ion but no detectable catalysis by acetic acid. For PNNN, acetate catalysis produces normal solvent isotope effects kROH/kROD of 1.68 ± 0.01 at high ionic strength (0.475) and 1.77 ± 0.04 at low ionic strength (0.048). A linear proton inventory indicates most simply that the isotope effect arises from a one-proton catalytic bridge in the transition state, although this model cannot strongly be distinguished from a generalized solvation effect. Reactions of methoxide ions produce slight inverse isotope effects kROD/kROH of 1.1–1.2, far smaller than the inverseeffect of about 2.5 expected for complete and uncompensated desolvation of the reactant-state methoxide ion. The transition state is thus stabilized by substantial interaction with the solvent. The proton inventory for the least reactive substrate PMMN (relative rate constant 1) is suggestive of transition-state stabilization by a combination of one-proton catalytic bridge(s) and distributed sites, while the proton inventory for the most reactive substrate PNNN (relative rate constant 1388) suggests only generalized transition-state solvation (many distributed sites); the proton inventory for PMNN, a substrate of intermediate reactivity (relative rate constant 60), suggests an intermediate situation. The data are consistent with a model in which transition states with exterior concentrations of charge favor stabilization of the charge by isotope-fractionating one-proton bridges, while transition states with distributed charge favor stabilization of the charge by many distributed sites. Key words: phosphoryl transfer, proton inventories, solvent isotope effects.