The ability to engage cationic substrates such as iminium and oxocarbenium ions in transition-metal-catalyzed C–C bond formation is a challenging task that has recently seen successful application.[1] In this context, we reported that a low-valent Ni catalyst facilitates unprecedented Suzuki–Miyaura cross-coupling reactions with allylic N,O- and O,O-acetal substrates.[2,3] Mechanistic studies revealed that boronic acids mediate allylic C–O activation, with oxidative addition occurring between the resulting iminium or oxocarbenium ion intermediate and the Ni catalyst (Scheme 1a).[4] The demonstration that a transition-metal catalyst can oxidatively insert into these prochiral intermediates offers a number of exciting possibilities for reaction design and enantioselective synthesis. Herein, we demonstrate that this activation mode enables the enantioselective synthesis of α-substituted 2,3-dihydro-4-pyridones by Negishi cross-coupling with N-acyl pyridinium ions (Scheme 1b). Scheme 1 Nickel-catalyzed cross-coupling with iminium ions. α-Substituted piperidines are among the most prevalent scaffolds in biologically active small molecules, and also serve as building blocks for natural product and pharmaceutical synthesis.[5] The stereoselective addition of carbon-centered nucleophiles to activated pyridines is a particularly attractive route to these core structures.[6] Seminal work from the groups of Comins and Charette established the possibility of using stoichiometric chiral acylating agents to control the stereo-chemical outcome of C–C bond formation.[7,8] However, only two examples have been described for the catalytic enantio-selective addition of organometallic reagents to prochiral N-acyl/alkyl pyridinium ions.[9] Both of these reactions likely proceed through the addition of a chiral [M]-R species into a prochiral pyridinium salt (M =Cu or Rh). As such, the strategy is limited to reactions with either highly nucleophilic R groups or highly electrophilic pyridinium ions. As the oxidative addition/reductive elimination mechanism that we elucidated for cross-coupling with N-acyl quinolinium salts should not be subject to these same limitations, we envisioned that its application to enantioselective cross-coupling with pyridinium ions could provide a complementary approach to these methods. Furthermore, successful development of such a reaction would demonstrate for the first time that iminium ion activation by Ni0 is subject to highly enantioselective C–C bond formation. One challenge apparent at the outset of our endeavors was potential catalyst poisoning in the presence of free pyridine. To address this issue, we chose to use 4-methoxy-pyridine as a substrate, because it shows substantial formation of a pyridinium salt with chloroformates at −78°C.[10] To enable facile transmetalation at low temperature, a Negishi reaction platform was selected.[11] Notably, when we initiated the reaction at −78°C with warming to RT, the combination of [{(methallyl)NiCl}2] (7.5 mol%) and (R)-Monophos (L1; 18 mol%) was found to promote arylation of 4-methoxypyridine in the presence of phenyl chloroformate and 4-FC6H4ZnBr with low but measurable ee (Table 1, entry 1).[12] Commercial phosphoramidite ligands L2–L4, which bear substituents at the 3 and 3′ positions of the binaphthyl backbone, induced more promising levels of asymmetric induction (up to 91% ee, entries 2–4).[13] Accordingly, an extensive library of 3,3′-substituted ligands was prepared and evaluated, revealing that ligand L7 was optimal (entry 7). With this ligand, the 2,3-dihydro-4-pyridone product 2a was obtained in 95% ee, albeit with moderate reaction efficiency. Table 1 Ligand optimization.[a] Upon selection of an optimal ligand (L7), we pursued further optimization of the reaction parameters. Among the Ni sources examined, NiBr2·diglyme was most promising, as its use increased the yield of the reaction while maintaining high levels of enantioselectivity (Table 2, entry 2). This result is particularly attractive because NiBr2·diglyme is air stable, thus permitting the reactions to be set up on the benchtop. In the absence of a Ni source, the reaction performs poorly, demonstrating that Ni induces substantial rate acceleration (entry 5). Additional changes to the reaction conditions, such as decreasing the amount of nucleophile or changing the identity of the acylating agent, had a negative effect on reactivity (entries 6–8). On the other hand, decreasing the Ni loading from 15 mol% to 10 mol% provided nearly identical levels of enantioselectivity and reaction efficiency (entry 9). A gram-scale reaction set up on the benchtop using a catalyst loading of 5 mol% delivered 2a in 73% yield and 93% ee (entry 10). Table 2 Optimization of reaction conditions.[a] Having identified conditions to prepare 2a in high yield and enantiomeric excess, a survey of aryl zinc nucleophiles was conducted to evaluate the scope of the reaction (Scheme 2).[14] Phenyl zinc bromide afforded 2,3-dihydro-4-pyridone 2b in 92% ee, which could be recrystallized to >99% ee in 78% yield. Generally, ortho-, meta-, and para-substituted zinc nucleophiles are well tolerated. Reactions with electron-withdrawing zinc reagents fared exceptionally well, delivering products bearing 3,5-difluoro (2c), 4-Cl (2d), 4-CF3 (2e), 4-cyano (2 f), and 4-sulfonamide (2g) groups in 96–99% ee. Notably, zinc reagents substituted with methyl ester and pivaloate groups afforded products (2j, 2k) that are otherwise inaccessible by standard Grignard methods. Moreover, a reaction with a heteroaromatic nucleophile furnished 2l in excellent ee and yield, highlighting the potential of this strategy to deliver medicinally relevant products. Whereas electron-neutral aryl zinc nucleophiles such as 4-vinyl (2m), 4-Me (2n) and 2-naphthyl (2o) performed modestly in terms of enantioselectivity, electron-rich nucleophiles such as 4-OMe (2p) underwent reaction with no stereoinduction owing to a competitive racemic background reaction (77% yield of isolated product without nickel).[15] An additional limitation is that other electrophile partners, including pyridine, 2-methoxy-pyridine, and 4-dimethylaminopyridine (DMAP), are not competent under the optimized conditions. Scheme 2 Scope of aryl zinc nucleophiles in the reaction. Yields and ee values are the average of two runs on a 0.5 mmol scale. [a] Yield and ee after recrystallization from ether. [b] Reaction warmed to RT. [c] Reaction warmed to −20°C. [d] Derived ... To understand the mechanism of this transformation, we studied the stoichiometric reaction of Ni0 with the pyridinium salt derived from 4-methoxypyridine and phenyl chloroformate, which was generated in situ. For simplicity, we chose to conduct our studies with PPh3, as it is a competent ligand for the reaction of interest, providing (±)-2a in 87% yield under otherwise standard reaction conditions.[13] In the event, an air-sensitive allyl–NiII adduct 3 was produced in 87% yield (Scheme 3), the structure of which was confirmed by single crystal X-ray diffraction (Figure 1). When subjected to 4-FC6H4ZnBr, allyl adduct 3 underwent C–C bond formation, providing 2a in 25% yield [Eq. (1)]. Furthermore, 3 is catalytically competent, providing racemic 2a in 68% yield at 10 mol% catalyst loading [Eq. (2)]. Taken together, these data provide compelling evidence for a redox mechanism distinct from that typically considered for transition-metal-catalyzed addition reactions to pyridinium ions. Figure 1 Solid-state structure of allyl NiII complex 3. Ellipsoids set at 30% probability. Hydrogen atoms omitted for clarity. Scheme 3 Stoichiometric reaction between Ni0 and a pyridinium ion. (1) (2) With these data in mind, we propose the reaction mechanism shown in Scheme 4. Oxidative addition of Ni0 into the C–N π bond of the pyridinium salt provides a NiII allyl intermediate analogous to complex 3.[16] Subsequent transmetalation with ArZnBr gives diorganonickel intermediate 4, which can undergo reductive elimination to regenerate the Ni0 catalyst and complete the catalytic cycle.[17] The presence of the methoxy substituent at C4 presumably serves as a blocking group, favoring the observed regioisomeric dihydropyridine 5.[18] Acid hydrolysis of 5 then affords the 2,3-dihydro-4-pyridone product 2. Scheme 4 Proposed catalytic cycle. In conclusion, we have described a novel nickel-catalyzed enantioselective Negishi cross-coupling reaction of 4-methoxypyridinium salts. A broad range of synthetically valuable enantioenriched 2,3-dihydro-4-pyridones can be obtained from commercially available or readily prepared starting materials and an air-stable, inexpensive NiII precatalyst. Preliminary mechanistic data support the intermediacy of a NiII π-allyl intermediate generated upon ionic oxidative addition of Ni0 to the pyridinium electrophile. The study provides an exciting indication of the generality of this mode of activation and its amenability to asymmetric catalysis.