The work performed during the project can be divided in two main parts:
(a) Fluorination of aryl boronic acid esters enabled by bismuth redox catalysis (published in Science in 2020, 10.1126/science.aaz2258).
To develop a catalytic reaction based on high-valent bismuth catalysis that could overcome limitations presented by transition metals, we design a ligand scaffold based on a sulfoximine motif bearing an electron-withdrawing CF3 group. This scaffold allowed us to study three independent organometallic steps individually. First, we focused on the reductive elimination step. To do so, we oxidized the corresponding Bi(III)-aryl complexes with XeF2, obtaining Bi(V)F2-aryl compounds. These compounds were submitted to thermal decomposition, delivering high yields of fluorobenzene. After demonstrating that reductive elimination and oxidation are possible, transmetallation of a variety of phenylboronic acids was surveyed, obtaining high yields of the corresponding aryl bismines. Then, oxidation was performed with N-fluoro-2,6-dichloropyridinium fluoride, a mild oxidant compared to xenon difluoride that after oxidation provides a weakly coordinating pyridine and therefore a cationic Bi(V) intermediate. The protocol proved general with a variety of para-substituted aryl-fluorides including vinyl functionalities, and it also boded well with naphthalene and ortho- and meta- substituents.
At this point we had already demonstrated the individual fundamental organometallic steps worked independently. We then turned our attention to merging all these steps into a catalytic cycle. After optimization of several reaction conditions, a variety of aryl boronic esters were smoothly converted to the corresponding aryl fluorides using 10 mol% of tetrafluoro bismine as the catalyst. Substitution in para-position was also tolerated as exemplified by the presence of trimethylsilyl or alkynyl groups, while meta-substituted aryl boronic esters presented more difficulties, affording moderate yields. Polyaromatic systems and sterically crowded compounds were also amenable for fluorination. Importantly, the reaction did not proceed in absence of Bi, which highlights its key role in this transformation. The development of this catalytic fluroination allowed us to overcome some limitations presented by transition metals with the development of a catalytic cycle based on the Bi(III)/Bi(V) redox couple, a feat that remained elusive to date. Indeed, this mode of reactivity represents a step forward not only in mimicking transition metal-like behavior,
but also in developing alternative reactivity which is beyond their scope.
(b) Bismuth-catalyzed oxidative coupling of aryl boronic acids with triflate and nonaflate salts (published in the Journal of the American Chemical Society in 2020, 10.1021/jacs.0c05343).
At this point, we imagined that it could be possible to expand this reactivity to other coupling partners. Indeed, we found inspiration in a work reported by Mukaiyama and co-workers, which describes the stoichiometric oxidative coupling phenyl bismines and triflic acid through the postulated Bi(V) intermediate depicted here, releasing phenyl triflate in 29% yield. This pioneering example is extremely important, as the construction of carbon-triflate bonds is inaccessible to transition metals, mainly due to the weak nucleophilic character of the triflate anion, which results in its weakly coordinating properties as ligand.
After a brief optimization of the reaction conditions, and with the knowledge gathered from the project (a), the scope of this transformation was investigated using a variety of aryl boronic acid derivatives. The methodology boded well with a variety of para- and ortho-substituted aryl boronic acids, including electron-donating and electron-withdrawing groups. Remarkably, when the steric encumbrance at the ortho-position was further increased, excellent yields of the corresponding triflate were obtained. The methodology accommodates meta-substituents, as well as unsaturated moieties such as alkynyl and vinyl groups. The reaction is proposed to follow a Bi(III)/Bi(V) catalytic cycle based on organometallic steps such as transmetallation, oxidative addition, and reductive elimination of C-OTf and C-ONf bonds. Overall, an unprecedented oxidative coupling of arylboronic acids with triflate and nonaflate salts has been developed exploiting the reactivity of the Bi(III)/Bi(V) redox couple, a transformation that it is yet to be reported utilizing transition metals as catalysts.
