C(α)-Proton Transfer from 2-(1-Hydroxybenzyl)oxythiamin: The Unit Br⊘nsted Slope Overestimates the Amount of Bond Formation to the Base Catalyst in the Transition State

Rate constants for C (α)-hydron transfer from racemic 2-(1-hydroxybenzyl)oxythiamin (HBOT) in oxygen-containing (cacodylate, phosphate, or alcohol) and primary amine buffers are reported. Thermodynamically unfavorable C(α)-H transfer from HBOT (pK_a = 15 ± 1) shows general-base catalysis with a Br⊘nsted β value of ≧0.95, which suggests rate-limiting diffusional separation of the conjugate buffer acid from the C(α)-carbanion/enamine. The calculated rate constant for the reverse protonation of the C(α)-carbanion/enamine by buffer acids, k_bh = 10^4±1 M^-1 s^-1, is independent of pK_a^BH with α ≦ 0.05, but is far below the diffusion-controlled limit. The primary kinetic isotope effects for cacodylate catalysis, k_H/k_T ⁼ 1-8 ± 0.1 and k_H/k_D = 1.5 ± 0.1 in H₂O, obey the Swain-Schaad relation and require incomplete proton transfer in the rate-limiting transition state. These results are consistent with the suggestion that a value of ad α≈ -0.2 for desolvation of the buffer acid offsets α = 0.2 for protonation to give α_obsd = 0 for some carbanions. General-base catalysis is detectable because there is a 10²⁹-fold negative deviation from the Br⊘nsted correlation for hydroxide ion.