Multimetallic catalysed cross-coupling of aryl bromides with aryl triflates

Abstract

The advent of transition-metal catalysed strategies for forming new carbon-carbon bonds has revolutionized the field of organic chemistry, enabling the efficient synthesis of ligands, materials, and biologically active molecules. In cases where a single metal fails to promote a selective or efficient transformation, the synergistic cooperation of two distinct catalysts--multimetallic catalysis--can be used instead. Many important reactions rely on multimetallic catalysis, such as the Wacker oxidation of olefins and the Sonogashira coupling of alkynes with aryl halides, but this approach has largely been limited to the use of metals with distinct reactivities, with only one metal catalyst undergoing oxidative addition. Here, we demonstrate that cooperativity between two group 10 metal catalysts--(bipyridine)nickel and (1,3-bis(diphenylphosphino)propane)palladium--enables a general cross-Ullmann reaction (the cross-coupling of two different aryl electrophiles). Our method couples aryl bromides with aryl triflates directly, eliminating the use of arylmetal reagents and avoiding the challenge of differentiating between multiple carbon-hydrogen bonds that is required for direct arylation methods. Selectivity can be achieved without an excess of either substrate and originates from the orthogonal reactivity of the two catalysts and the relative stability of the two arylmetal intermediates. While (1,3-bis(diphenylphosphino)propane)palladium reacts preferentially with aryl triflates to afford a persistent intermediate, (bipyridine)nickel reacts preferentially with aryl bromides to form a transient, reactive intermediate. Although each catalyst forms less than 5 per cent cross-coupled product in isolation, together they are able to achieve a yield of up to 94 per cent. Our results reveal a new method for the synthesis of biaryls, heteroaryls, and dienes, as well as a general mechanism for the selective transfer of ligands between two metal catalysts. We anticipate that this reaction will simplify the synthesis of pharmaceuticals, many of which are currently made with pre-formed organometallic reagents, and lead to the discovery of new multimetallic reactions.

Extended Data Fig. 1. Varying Ligands on the Nickel and Palladium Catalysts

Several different phosphine ligands for palladium (a) and amine ligands for nickel (b) were investigated. While selectivity and yield of cross-product were sensitive to the identity of the phosphine ligand (a), a variety of different amine ligands were nearly equally effective (b). Reactions conducted with L15, L17, and L18 used 5 mol% catalyst loadings, were heated to 40 °C, and were monitored for 64 hours.

Extended Data Fig. 2. The Effect of Potassium and Fluoride Salts on the Selectivity of the Multimetallic Catalyzed Cross Ullman Reaction

The presence of potassium fluoride in cross couplings has been demonstrated to improve the product yield and rate of reactions while minimizing the formation of dimeric products. This “KF effect” can be attributed to a variety of different factors, including the formation of “ate” complexes, the removal of a halide from a metal complex, or the formation of a fluoride bridged complex, which could facilitate an oxidative addition or transmetalation step^,. To investigate whether KF could be beneficial for the nickel and palladium multimetallic system, various amounts of the additive were included under standard reaction conditions. The resulting selectivity and rate data are found in (a) and (b). Once the advantageous role of KF was confirmed, other potassium and fluoride salts were subsequently tested. The resulting selectivity and rate data are found in (c) and (d). The presence of potassium resulted in faster reaction rates, while the presence of fluoride reduced the amount of bianisole and biphenyl byproducts, improving the yield of cross product. Reactions without additives took 32 hours to complete.

Fig. 1

A general cross-Ullman reaction catalyzed by a combination of nickel and palladium. Although a nickel(0/II) and palladium (0/II) cycle is depicted, alternative mechanisms are also possible, such as a nickel(I/III) cycle.

Fig. 2

Nickel and Palladium Catalyst Selectivities. (a) (dppp)PdCl₂ and (bpy)NiBr₂ (10 mol% each); (b) (bpy)NiBr₂ (10 mol%); and (c) (dppp)PdCl₂ (10 mol%) were used along with approximately 0.5 mmol of each starting material. Concentrations determined by GC analysis, corrected with an internal standard. The low mass balance in (b) is due to the formation of benzene (0.13 mmol) from hydrodehalogenation of 2.

Fig. 3. Substrate Scope. (a) Aryl and vinyl bromides with aryl and vinyl triflates. (b) Other aryl electrophiles. (c) Relative reactivity of the palladium and nickel catalysts

^a No KF was used. ^b Three equivalents of 3-bromothiophene were used. ^c Two equivalents of 2-bromocyclohexenone were used.