Mechanism of phosphine borane deprotection with amines: the effects of phosphine, solvent and amine on rate and efficiency.

The kinetics of borane transfer from simple tertiary phosphine borane adducts to a wide range of amines have been determined. All data obtained, including second-order kinetics, lack of cross-over, and negative entropies of activation for reaction of triphenylphosphine borane with quinuclidine and triethylamine, are consistent with a direct (SN 2-like) transfer process, rather than a dissociative (SN 1-like) process. The identities of the amine, phosphine, and solvent all impact substantially on the rate (k) and equilibrium (K) of the transfer, which in some cases vary by many orders of magnitude. P-to-N transfer is more efficient with cyclic amines in apolar solvents due to reduced entropic costs and ground-state destabilisation. Taken as a whole, the data allow informed optimisation of the deprotection step from the stand-point of rate, or synthetic convenience. In all cases, both reactants should be present at high initial concentration to gain kinetic benefit from the bimolecularity of the process. Ultimately, the choice of amine is dictated by the identity of the phosphine borane complex. Aryl-rich phosphine boranes are sufficiently reactive to allow use of diethylamine or pyrrolidine as a volatile low polarity solvent and reactant, whereas more alkyl-rich phosphines benefit from the use of more reactive amines, such as 1,4-diaza[2.2.2]bicyclooctane (DABCO), in apolar solvents at higher temperatures.