Distinguishing Between Pathways for Transmetalation in Suzuki-Miyaura Reactions
Brad P. Carrow and John F. Hartwig*
J. Am. Chem. Soc. 2011, 133, 2116.
The Suzuki-Miyaura cross-coupling reaction (Fig. 1) is definitely one of the most important reactions in synthetic chemistry, especially in industrial and medical settings. So one would think that the mechanism of this classic reaction should be fully elucidated by now, right? (The reaction was discovered back in the early 70s!) Although everybody knows that the reaction proceeds by the “classic” cross-coupling mechanism – oxidative addition of the aryl/vinyl/alkyl halide to Pd to form LPdRX (X = halide), transmetallation of the aryl/vinyl/alkyl boron species onto Pd to form RPdR’L and BX, and then reductive elimination from three-coordinate Pd to form R-R’ and PdL – the exact mechanism of transmetallation, the purported rate-limiting step, remained unclear. That is, until the work of Hartwig and Carrow in this paper.
Suzuki couplings are often run in water/THF with base (such as K2CO3) present, which we know from freshman chemistry leads to the generation of OH- in situ; the presence of OH- accelerates the rate of the reaction, so it must be involved somehow in transmetallation. There are two possible mechanisms for transmetallation that could explain this effect: the attack of hydroxide on the boron species (in this case, a boronic acid, forming a trihydroxyborate), which is more reactive towards transmetallation onto Pd, or attack of hydroxide on Pd to form a PdOH complex, which is more reactive than the parent PdX complex (Figure 2).
The Hartwig group approached this problem simply and elegantly: why not make and isolate an aryl trihydroxyborate and react it with a ArPdX complex, and do the same with an ArPdOH complex and an aryl boronic acid? Assessing the relative rates of the two reactions would then give insight into which process is faster. This would be a test of the kinetic competence of the two intermediates: do either react fast enough to be relevant to the reaction mechanism? However, before testing their kinetic competence the authors needed to test the chemical competence of the trihydroxyborate and PdOH complex.
Using 31P NMR to observe the generation of free phosphine ligand (L) upon reaction of LPdArZ (Z = Cl, OH) complexes, the authors found that trihydroxyborates react only very slowly with LPdArCl complexes (<10% conversion after 11 hours, estimated rate constant of 1.7 x 10-7); in contrast, the reaction of a PdOH complex with p-tolylboronic acid proceeded with a rate constant of 10-3 s-1. The conclusion that transmetallation occurs via a PdOH intermediate, however, also demands that PdOH complexes form under the normal Suzuki conditions. Again using 31P NMR, the authors were able to establish and watch the equilibrium between PdOH dimer and monomer complexes with the corresponding PdX complexes using tetraalkylammonium halide salts in THF/water. While the equilibrium tends to favor the PdX complex, the Keq is within one order of magnitude of 1 in all reported cases. Thus, PdOH complexes were established to be both kinetically and thermodynamically competent intermediates in the Suzuki-Miyaura cross coupling reaction with boronic acids. Lastly, the authors demonstrate that the same is most likely true for boronic esters, the second most common boron-containing species used in Suzuki couplings.