Nickel-Catalyzed Heteroaromatic C-H Alkylation via Ligand-to- Ligand Hydrogen Transfer

Abstract: Regioand enantioselective functionalization of heteroarene C–H bonds in the absence of directing groups is a long-standing challenge in the field of C–H activation. Herein, we present an approach involving nickel-catalyzed intermolecular enantioselective C–H alkylation of heteroarenes. The process can be carried out under mild conditions using nickel(0) catalysts with N-heterocyclic carbene (NHC) ligands in the absence of Lewis acid co-catalysts. A series of NHC nickel complexes stabilized with 1,5-hexadiene were synthesized via an operationally simple approach, resulting in improved functional group tolerance and a wide substrate scope. Mechanistic investigations indicate that the transformation involves a ligand-to-ligand hydrogen transfer (LLHT) pathway where the C–H bond activation precedes a rate-determining reductive elimination step. Heteroaromatic rings are common motifs in natural products and FDA approved pharmaceuticals.[1] Direct functionalization of heteroaromatic C–H bonds has the potential to streamline the synthesis of complex molecules by avoiding the need for pre-functionalization steps.[2] Despite their prevalence, aromatic C-H bonds are relatively inert and often similar in reactivity, rendering their selective functionalization challenging, especially in intermolecular, non-directed processes.[3] Over the past several decades, enormous advances have been made in the field of selective C–H functionalization through the exploration of transition metal catalysis. These newly developed methods offer orthogonal opportunities to existing strategies for the synthesis of natural products, functional materials as well as exploration of structure-activity relationships by avoiding the pre-installation of functionalities required for traditional crosscouplings.[4] Selective transformation of heteroaromatic C-H bonds, however, typically rely on the use of directing groups and secondor third-row transition metal catalysts.[5] To this end, functionalizing heteroarenes that lack directing groups using earth abundant transition metals remains underdeveloped.[6] Our interest in nickel-catalyzed C–H functionalization stems from nickel’s relatively high earth abundance and complementary reactivity compared with second and third-row transition metals.[7] Recently, a nickel catalyzed C–H alkenylation reaction that first reported by Nakao and Hiyama[8] was extensively explored computationally by Perutz and co-workers.[9] The authors proposed that C-H activation proceeds through a novel ligand-to-ligand hydrogen transfer (LLHT) pathway rather than the formation of a discrete nickel-hydride via oxidative addition. Under this framework, a variety of aromatic C–H bonds have been functionalized using the combination of nickel(0) pre-catalysts and NHC or phosphorus-based ligands.[10] Enantioselective LLHT processes are more rare, with intramolecular asymmetric cyclization being described by Ye[11], Cramer,[12] Ackermann,[13] and Shi[14] groups, enabled by the design of novel ligand scaffolds (Scheme 1). In 2018, the Ye group reported a Ni-Al bimetallic enantioselective C–H exo-selective cyclization of imidazoles with alkenes promoted by secondary phosphine oxide (SPO) ligands.[11a] Shortly after, the Cramer group developed a modular synthesis of IPhEt ligand family, which was first reported by Galway and co-workers.[15] These catalytic systems were successfully applied to the C–H functionalization of pyridones[12a, 12c], pyrroles and indoles.[12d] Independently, the Ackermann group discovered that the combination of nickel and JoSPOphos allowed for the asymmetric C– H functionalization of imidazoles without the need of external Lewis acids, and they further illustrated this novel catalysis with a well-defined nickel(II)-JoSPOphos complex.[13a] Scheme 1. Nickel catalyzed asymmetric heteroaromatic C–H alkylation via ligand-to-ligand hydrogen transfer (LLHT). While important advances have been made in enantioselective intramolecular cyclization processes, the corresponding enantioselective intermolecular transformations have not been developed and represent an important gap in the field. To address this challenge, we report herein a method for the intermolecular enantioselective C–H alkylation of heteroarenes under mild conditions using a hindered NHC ligand with backbone chirality. An NHC ligand motif not previously utilized in enantioselective LLHT processes is paired with a strategy for the utilization of discrete 1,5-hexadiene-supported nickel(0) complexes to improve catalytic activity, increase functional group tolerance, and widen substrate scope of the transformation. We began our efforts by exploring the alkylation of benzoxazole with norbornene (nbe) using nickel catalysts. After screening several common chiral ligands for enantioselective nickel catalysis, we found that NHCs derived from commercially available chiral diamines provided appreciable enantioinduction while maintaining good reactivity. Upon optimization of the reaction conditions through variation of ligand structure, solvent, temperature, and additives (see supporting information), we set out to demonstrate the utility of this method by exploring the scope of heteroarenes (Scheme 2). The optimized conditions involve in-situ preparation of an active chiral catalyst via the use of 5 mol % L5•HBF4, KHMDS and Ni(COD)2. The process is tolerant of electron-neutral and electron-rich benzoxazoles substituted at various positions (1a-1f) to give the alkylation products with good yields and enantioselectivities. Selective functionalization of benzofurans were also observed for C2 C–H bonds leaving functional groups such as boronic esters intact (1g-1i). LLHT products were not observed when using 5-bromobenzofuran as substrate (1j), which could be attributed to the facile activation of C(sp2)-Br bond in the presence of low-valent nickel catalyst. Benzimidazoles previously studied by Ackermann[13a] and Ye[11a] for enantioselective cyclizations were also reactive in our intermolecular system. We showed that nbe reacted with N-Me and N-Ph benzimidazoles (1k, 1l) to yield the C–H functionalization products with a moderate control for the enantioselectivity at 60 °C (vide infra). Interestingly, a C5-alkylated 1,2,4-triazole was exclusively obtained regioand enantioselectively despite the relatively more hindered environment (1m). Other nitrogen containing heteroaromatic rings including caffeine (1n) and 3-methylquinazolin-4(3H)-one (1o) also underwent efficient couplings to give the desired product in good yields and 90:10 and 87:13 e.r. respectively. All of alkylation products were obtained as a single, exocyclic diastereomer. Scheme 2. Enantioselective heteroaromatic C–H functionalization scope. Reactions were performed at 0.10 mmol scale and yields were reported for isolated products. a Reaction was carried out at 60 °C. When examining benzoxazoles bearing electron-withdrawing substituents, however, the desired alkylation products were not obtained using the standard protocol described above. We postulated that the increased acidity of benzoxazole’s C-2 C-H bond could lead to deleterious side reactions with 1,5-cyclooctadiene (COD) from the nickel pre-catalyst, resulting in unproductive off-cycle intermediates. The diminished reactivity of benzoxazoles bearing electron-withdrawing groups has similarly been observed by Ong and co-workers in their study of hydroarylation of cyclic dienes.[16] The inhibitory effect of COD in C–H activations was previously described by our laboratory in the development of alkyne hydroarylation reactions involving LLHT pathways.[7a] Such phenomenon originates from the participation of COD in a competing LLHT C–H activation that results in a stable off-cycle p-allyl complex that minimizes the concentration of the active catalyst for productive catalysis. This side reaction was prevented by employing 1,5-hexadiene-supported nickel pre-catalysts[17] leading to a more efficient alkyne hydroarylation at reduced temperatures. Motivated by this finding, we set out to develop the synthesis for NHC nickel complexes stabilized with 1,5-hexadiene from easily accessible nickel precursors under operationally simple condition. Inspired by work done by Wilke and co-workers on the synthesis of ‘bis(olefin)nickel-ligand complex’[18] as well as Belderrain and Nicasio’s nickel bis-styrene complex,[19] we anticipated that ligand exchange from COD to 1,5-hexadiene followed by trapping with free NHC ligands would allow facile access to a variety of sterically demanding NHC nickel complexes. Using this approach (see supporting information), we were able to synthesize 1,5-hexadiene supported nickel complexes bearing IMes, IPrMe, IPr*OMe as well as the optimal ligand for the enantioselective transformation described above (L5 in Scheme 2), whose structure was unambiguously confirmed using x-ray crystallography (Scheme 3). Scheme 3. Synthesis of NHC nickel complexes stabilized with 1,5-hexadiene. ORTEP diagrams of L5-Ni(C6H10) with thermal ellipsoids at 50% probability. Hydrogens have been omitted for clarity. With these novel catalysts in hand, we first tested their reactivity using benzoxazole (1a), and comparable results were obtained. Moreover, the synthesis of enantioenriched alkylation products could be carried out using as low as 2.5 mol % L5-Ni(C6H10) at 1.0 mmol scale without diminished enantioselectivity. We employed this catalyst in previously problematic substrates including those that feature electron-withdrawing substituents. While the in-situ-generated chiral catalyst at 10 mol % loading gave no reaction, 1p and 1q underwent alkylation smoothly to give desired products using the discrete nickel catalyst L5-Ni(C6H10), where excellent enantioselectivities and clean reaction profiles were observed. Most surprisingly, products arising from chemoselective C–H functionalization were observed as the sole product with 93:7 e.r. in the pres