Modular Access to Quaternary α‑Cyano Carbonyl Compounds via NiH Catalysis

Abstract

Quaternary α-cyano carbonyl compounds are an important class of molecules in organic synthesis due to their versatility as precursors to essential building blocks and their potential applications in pharmaceutical synthesis. Despite the importance of these sterically congested and highly functionalized structures, the development of a general and practical synthetic platform remains a persistent challenge. Herein, we report a nickel-hydride-catalyzed hydrofunctionalization of α,β-unsaturated nitriles with acyl fluorides as well as their carbamoyl and formyl derivatives. This synthetic protocol enables the modular synthesis of a diverse range of quaternary α-cyano carbonyl compounds, including ketones, amides, and esters, under mild conditions and without the need for an external ligand. Combined DFT and experimental mechanistic studies reveal that the present NiH catalysis proceeds via regioselective hydrometalation, followed by the formation of a nickel-keteneiminate intermediate through rearrangement and subsequent nucleophilic substitution reaction between the nickel-keteneiminate and the carbonyl fluorides.

Introduction

The nitrile functional group is commonly found in biologically active compounds and serves as a versatile synthetic handle for downstream modifications. Consequently, significant efforts have been devoted to developing efficient strategies for synthesizing nitrile-containing compounds. Among these, methods for the synthesis of α-cyano carbonyl compounds, particularly those bearing quaternary carbon centers at the α-position, have garnered considerable attention in recent decades. − This interest is primarily driven by the inherent synthetic challenges posed by quaternary carbon centers and their broad synthetic utility. These compounds are notable for their ability to be transformed into other valuable building blocks, such as β-amino acids, β-amino amides, and β-hydroxy nitriles, including sterically congested structures (Figure A, left). , Additionally, they can serve as key synthetic intermediates in drug synthesis, as exemplified by the synthesis of verapamil (Figure A, right). Quaternary α-cyano carbonyl compounds can be accessed through several strategies, including electrophilic and nucleophilic cyanation of carbonyls and their α-halo derivatives, ,− arylation of malononitriles, and carbonylative coupling of aryl halides with secondary alkyl nitriles. Alternatively, these compounds can be synthesized via the reaction of in situ-generated metal keteneiminates or pre-generated silyl ketene imines with carbonyl electrophiles (Figure B). , While this approach provides a distinct disconnection strategy and holds promise for broader access to quaternary α-cyano carbonyls, it faces several significant challenges. Specifically, the generation of metal keteneiminates requires a strong base, and silyl ketene imines are highly sensitive to moisture, complicating their handling. , Moreover, the scope of carbonyl electrophiles is largely limited to acyl electrophiles, with only rare examples involving carbamoyl and formyl electrophiles. ,,, Therefore, the development of a more efficient and practical synthetic strategy for this C–C bond disconnection, one that enables the modular synthesis of quaternary α-cyano carbonyl compounds, is highly desirable.

1.

Synthesis of quaternary α-cyano carbonyl compounds: Prior art and this work.

The aforementioned strategy could be realized by reacting α,β-unsaturated nitriles with carbonyl electrophiles under transition metal-hydride catalysis (Figure C). The key to this reaction design lies in identifying an appropriate catalytic system capable of facilitating regioselective hydrometalation, in situ generation of metal keteneiminate, and subsequent reaction with a broad range of carbonyl electrophiles. However, while this approach appears to be both feasible and conceptually straightforward, to the best of our knowledge, only two examples have been reported to date. A pioneering study by Szymczak demonstrated the reductive coupling of α,β-unsaturated nitriles with symmetric anhydrides under RuH-catalysis. More recently, Xia developed a Ru-catalyzed hydroacylation of acrylonitrile employing aldehydes or alcohols via a hydrogen transfer strategy. Although these studies represent important advances, both methods rely on limited substrates, specifically, four examples of symmetric anhydrides in the former and acrylonitrile as the sole Michael acceptor to afford the quaternary α-cyano carbonyl compounds in the latter, thereby limiting its broader synthetic applicability.

Building on our ongoing interest in developing nickel-hydride-catalyzed hydrofunctionalization reactions , and leveraging acyl fluorides, carbamoyl fluorides, and fluoroformates in catalysis, , we envisioned a general and efficient method for synthesizing quaternary α-cyano carbonyl compounds through NiH-catalyzed hydrofunctionalization of α,β-unsaturated nitriles with these carbonyl electrophiles. Herein, we report the successful realization of this goal (Figure D). The reaction operates under mild conditions and does not require an exogenous ligand. A wide range of acyl fluorides, along with their carbamoyl and formyl derivatives, can be employed as carbonyl electrophiles, enabling the modular synthesis of α-cyano ketones, amides, and esters bearing quaternary carbon centers. Mechanistic studies, including DFT calculations, reveal that the reaction proceeds via regioselective hydrometalation, followed by the formation of a nickel-keteneiminate intermediate.

Results and Discussion

We began our optimization efforts by examining a range of reaction conditions using (Z)-2,3-diphenylacrylonitrile (1a) and benzoyl fluoride (2a) as the model substrate and acyl electrophile, respectively (Table ).

1. NiH-Catalyzed Hydroacylation of 1a with 2aOptimization of Reaction Conditions .

entry	catalyst (mol %)	base (equiv)	hydrosilane	additive (equiv)	yield, 3a (%)
1	NiCl2·DME (15)	CsF (2.0)	PMHS		49
2	CoBr2·DME (15)	CsF (2.0)	PMHS		n.d.
3	Cu(OAc)2 (15)	CsF (2.0)	PMHS		<5
4	NiCl2·DME (15)	CsF (2.0)	Ph2SiH2		28
5	NiCl2·DME (15)	CsF (2.0)	(EtO)3SiH		51
6	NiCl2·DME (15)	KF (2.0)	(EtO)3SiH		43
7	NiCl2·DME (15)	K3PO4 (2.0)	(EtO)3SiH		67
8	NiCl2·DME (15)	K3PO4 (2.0)	(EtO)3SiH	2,6-lutidine (1.0)	84
9	NiCl2·DME (15)	K3PO4 (2.0)	(EtO)3SiH	2,6-lutidine (0.3)	70
10	NiCl2·DME (15)	K3PO4 (2.0)	(EtO)3SiH	2,6-lutidine (2.0)	>99
11	NiCl2·DME (10)	K3PO4 (0.5)	(EtO)3SiH	2,6-lutidine (2.0)	>99
12		K3PO4 (0.5)	(EtO)3SiH	2,6-lutidine (2.0)	n.d.
13	NiCl2·DME (10)		(EtO)3SiH	2,6-lutidine (2.0)	n.d.
14	NiCl2·DME (10)	K3PO4 (0.5)		2,6-lutidine (2.0)	n.d.
15	NiCl2·DME (10)	K3PO4 (0.5)	(EtO)3SiH	2,6-lutidine (2.0)	n.d.
16	NiCl2·DME (10)	K3PO4 (0.5)	(EtO)3SiH	2,6-lutidine (2.0)	11

^aReaction conditions: 1a (2.0 equiv), 2a (0.1 mmol, 1.0 equiv), catalyst, base, hydrosilane (2.0 equiv), additive, and DMF (1.0 mL) for 12 h at room temperature. ¹H NMR yields are given.

^b 2b was used in lieu of 2a.

^c 2c was used in lieu of 2a.

We observed that employing NiCl₂·DME (15 mol %) as the catalyst, CsF (2.0 equiv) as the base, and PMHS (2.0 equiv) as the hydride source yielded the desired quaternary α-cyano ketone product 3a in a 49% yield (entry 1). Substituting NiCl₂·DME with alternative first-row transition metal catalysts, such as CoBr₂·DME or Cu­(OAc)₂, resulted in either no reaction or only trace product formation (entries 2 and 3). Further hydrosilane screening identified (EtO)₃SiH as a more effective agent for this hydroacylation, whereas Ph₂SiH₂ produced the product in a lower yield (entries 4 and 5). Testing various bases revealed that the use of K₃PO₄ improved the yield, while KF resulted in a reduced reaction efficiency (entries 6 and 7). To further improve the reaction efficiency, we evaluated different additives. Of the additives tested, the use of 1.0 equiv of 2,6-lutidine improved the yield of the desired product to 84%, while a catalytic amount (30 mol %) of 2,6-lutidine produced a comparable yield (entries 8 and 9). Notably, employing 2.0 equiv of 2,6-lutidine significantly enhanced the yield, resulting in nearly quantitative product formation (entry 10).

We also observed that the yield of 3a remained quantitative even when the catalyst and base loading were reduced to 10 and 50 mol %, respectively (entry 11). Control experiments confirmed that the nickel catalyst, base, and hydrosilane are each essential for successful hydroacylation (entries 12–14). Finally, testing alternative acylating reagents revealed that other acyl electrophiles, such as benzoyl chloride (2b) and benzoic anhydride (2c), either yielded no product or significantly decreased the reaction efficiency (entries 15 and 16). Throughout the optimization process, we observed that only the α-acylated product was obtained, with no detection of the β-acylated product.

The successful reaction optimization prompted us to investigate the mechanism of the present nickel-hydride-catalyzed hydroacylation. To this end, we divided the catalytic cycle into two major steps: the regioselective hydronickelation of α,β-unsaturated nitrile and subsequent acylation.

The mechanism of each step was then examined in detail. First, a series of mechanistic experiments were conducted to elucidate the hydronickelation step (Scheme ). To examine the regioselectivity of the hydronickelation step, density functional theory (DFT) calculations were performed (Scheme A). Consistent with the experimental results, the computational findings indicated that hydronickelation from Int-A, which forms the α-nickelated intermediate Int-B via TS-1 (15.6 kcal/mol), would proceed much more readily than hydronickelation from Int-A′, leading to β-metalated species Int-B′ via TS-2 (30.3 kcal/mol). A deuterium labeling experiment was then carried out using nickel-catalyzed hydroacylation of 1a with 2a as the model reaction in the presence of deuterated hydrosilane (Ph₂SiD₂, Scheme B). The result showed the exclusive incorporation of deuterium at the β-position of both the acylation product (3a- d ) and the alkyl nitrile side product (4- d ), with no deuterium detected in the recovered starting material (1a). This outcome suggests that hydrosilane serves as the hydride source and that the regioselective hydronickelation is irreversible. Additionally, we conducted the reaction of alkyl nitrile 4 with 2a under the standard reaction conditions to rule out the possibility that the reaction proceeds through the reduction of α,β-unsaturated nitrile followed by base-assisted acylation of alkyl nitrile. As shown in Scheme C, desired product 3a was not observed, while 4 was nearly fully recovered, allowing us to exclude this pathway.

1. Mechanistic Studies on Hydronickelation Step .

^a Computational level: Gaussian 09 SMD­(N,N-dimethylformamide)-UB3LYP/6–311+G**|SDD­(Ni)//UB3LYP/6–31G**|LANL2DZ (Ni). See the Supporting Information (SI) for the detailed experimental procedures.

We next sought to explore the detailed mechanism underlying the acylation step (Scheme ). Starting from the α-nickelated intermediate Int-B, we considered two possible mechanistic pathways: formation of a nickel-keteneiminate intermediate followed by nucleophilic acyl substitution with acyl fluoride and acylation via the oxidative addition of acyl fluoride followed by reductive elimination. To assess which pathway is more plausible, DFT calculations were performed (Scheme A). The C-metalated nitrile intermediate Int-B could be converted to the N-metalated keteneiminate Int-C by passing through transition state TS-3, which is 9.6 kcal/mol higher in energy. The calculated structure of nickel­(I)-keteneiminate Int-C features an elongated N–C1 bond of 1.195 Å, a shortened C1–C2 bond of 1.363 Å, and a bent Ni–N–C1 angle of 124.6°, closely resembling the structure of the ruthenium-keteneiminate complex isolated and reported by Szymczak and co-workers. Int-C then reacts with benzoyl fluoride (2a) to form the tetrahedral intermediate Int-D, with a barrier of 10.3 kcal/mol. Int-D is subsequently converted to Int-E, which undergoes C–F bond cleavage via TS-5, located only 1.0 kcal/mol higher in energy, to yield the desired product and (DMF)₂NiF. The (DMF)₂NiF species then reacts with hydrosilane to regenerate the catalytically active nickel­(I)-hydride. On the other hand, the oxidative addition of benzoyl fluoride (2a) to Int-B may occur, leading to the formation of the nickel­(III)-acyl species Int-F with a kinetic barrier of 5.3 kcal/mol. Int-F is subsequently converted to Int-G, which then undergoes reductive elimination via TS-7 with a barrier of 13.3 kcal/mol. Starting from Int-B, the overall energy barrier for the pathway involving the formation of a nickel-keteneiminate followed by nucleophilic acyl substitution is 1.8 kcal/mol lower than that of the oxidative addition–reductive elimination pathway. While this small energy difference does not completely rule out the oxidative addition–reductive elimination mechanism, the keteneiminate pathway appears to be energetically more favorable in this hydroacylation reaction.

2. Mechanistic Studies on the Acylation Step .

^a Computational level: Gaussian 09 SMD­(N,N-dimethylformamide)-UB3LYP/6–311+G**|SDD­(Ni)//UB3LYP/6–31G**|LANL2DZ (Ni). See the SI for the detailed experimental procedures.

To obtain further experimental evidence for the reaction between nickel-keteneiminate and acyl fluoride, we treated silyl ketene imine 5, an analog of in situ-generated nickel-keteneiminate in the present acylation, with 2a (Scheme B). The reaction produced the desired product 3a in nearly quantitative yield, providing additional, albeit indirect, evidence that nickel-keteneiminate can indeed react with acyl fluoride via a nucleophilic acyl substitution mechanism. However, note that the formation of the silyl ketene imine (Int-C′) and its subsequent reaction with acyl fluoride are less favorable in the current catalytic system. Calculations indicate that the transformation of Int-C to Int-C′ is endergonic by 11.4 kcal/mol, rendering the formation of Int-C′ unlikely under the reaction conditions. Additionally, although Int-C′ is proposed to form following the turnover-limiting hydronickelation step, it was neither detected nor accumulated under the standard reaction conditions, even in the absence of the acyl electrophile, as determined by HRMS (Scheme C). Taken together, these computational and experimental results suggest that the present hydroacylation indeed proceeds via nucleophilic acyl substitution involving the nickel-keteneiminate (Int-C), rather than the silyl ketene imine (Int-C′).

Finally, we sought to investigate the role of the 2,6-lutidine additive, which is known to react with acyl electrophiles to form N-acyllutidinium salts, to determine whether it similarly interacts with acyl fluoride to generate a cationic acylium intermediate during the reaction. To this end, we monitored the reaction between 2a and 2,6-lutidine in DMF-d ₇ using ¹H and ¹⁹F NMR (Scheme D). Throughout the analysis, no new signals appeared, suggesting that acyl fluoride itself rather than an in situ-generated N-acyllutidinium intermediate reacts with the nickel-keteneiminate intermediate to form the desired hydroacylation product. While further mechanistic studies are needed, 2,6-lutidine might serve as an additional base in the current NiH catalysis, thereby improving the overall reaction efficiency.

A plausible catalytic cycle for the present hydroacylation reaction is proposed based on our mechanistic studies and prior literature reports (Scheme ). , The cycle is initiated by the in situ generation of a catalytically active nickel­(I)-hydride species (I) via the reaction of the Ni­(II) precatalyst with hydrosilane in the presence of base.

3. Plausible Catalytic Cycle.

This Ni–H species undergoes irreversible, regioselective hydronickelation of the α,β-unsaturated nitrile substrate to form the corresponding α-nickelated intermediate (II), rather than the β-isomeric species (II′), as supported by both experimental and computational studies. Intermediate II is proposed to be in equilibrium with a rearranged nickel-keteneiminate species (III), which undergoes nucleophilic acyl substitution with an acyl fluoride to furnish the desired hydroacylation product and a nickel fluoride species (IV). Subsequent reaction of IV with hydrosilane regenerates the active nickel-hydride species (I), thereby completing the catalytic cycle. While the mechanistic pathway involving the nickel-keteneiminate intermediate is more consistent with our experimental and computational findings, we cannot fully exclude an alternative pathway wherein oxidative addition of the acyl fluoride to intermediate II generates a Ni­(III)–acyl species (V), followed by reductive elimination.

We next explored the scope of acyl fluorides and α,β-unsaturated nitriles in the present hydroacylation reaction under the optimized reaction conditions (Table ). Aroyl fluorides bearing diverse substituents at its aryl ring, including phenyl (3c), tert-butyl (3d), alkoxy (3e, 3f and 3i), dimethylamino (3g), thiomethyl (3h), and chloro (3j), participated in the hydroacylation reaction to afford the desired α-cyano ketones with a quaternary carbon center in moderate to excellent efficiency. Moreover, aroyl fluorides substituted with electron-withdrawing groups, such as ester (3k), acetoxy (3l), trifluoromethyl (3m), trifluoromethoxy (3n), and cyano (3o), were efficiently converted to the corresponding hydroacylation products. 2-Naphthoyl fluoride (3p) and heteroaroyl fluorides such as thiophenyl (3q), indolyl (3r), and benzofuranyl (3s) derivatives were also compatible. Additionally, not only aromatic but also aliphatic acyl fluorides (3t, 3u, 3v, and 3w) could be accommodated in this transformation. Under the reaction conditions, α-aryl cinnamonitrile derivatives containing alkoxy, bromo, chloro, cyano, fluoro, ester, nitro, and trifluoromethyl substituents were all compatible (3b–3s). Furthermore, α,β-unsaturated nitriles substituted with a heteroaryl group, such as thiophene (3t), dibenzofuran (3u), and pyridine (3v and 3w), also underwent the hydroacylation, affording the corresponding products in moderate to good yields.

2. Scope of Hydroacylation of α,β-Unsaturated Nitriles with Acyl Fluorides .

^aReaction conditions: α,β-unsaturated nitrile (2.0 equiv), acyl fluoride (0.2 mmol, 1.0 equiv), NiCl₂·DME (10 mol %), K₃PO₄ (2.0 equiv), (EtO)₃SiH (2.0 equiv), 2,6-lutidine (2.0 equiv), DMF (1.0 mL) for 12 h at room temperature. Isolated yields are given.

^bK₃PO₄ (50 mol %) was used.

^cThe reaction was performed at 0.4 M in DMF solvent.

^dThe reaction was performed at 0.2 M in DMF solvent.

^e15 mol % of NiCl₂·DME was used.

^fK₃PO₄ (3.0 equiv) was used.

^gCsF was used instead of K₃PO₄.

^hPMHS was used instead of (EtO)₃SiH.

ⁱRun for 24 h.

^jK₂CO₃ (1.0 equiv) and K₃PO₄ (1.0 equiv) were used instead of K₃PO₄ (2.0 equiv).

^kTMDSO was used instead of (EtO)₃SiH.

^l(E)-cinnamonitrile was used.

^m E/Z mixture of crotononitrile was used.

^{n 1}H NMR yields are given.

^o2-benzylacrylonitrile was used.

^p E/Z mixture of 2-methyl-3-phenylacrylonitrile was used.

In addition to α-aryl cinnamonitrile derivatives, α-aryl-β-alkyl (3x) and cyclic α,β-unsaturated nitrile (3y) could participate in this transformation. Notably, a cyanoacrylate derivative could be employed as a substrate, affording the desired product bearing a quaternary carbon center substituted with two distinct carbonyl groups (3z). It was also found that not only internal but also terminal α,β-unsaturated nitriles could be transformed into the corresponding hydroacylation products with good efficiency (3aa and 3ab). In the present hydroacylation, α-cyano ketones bearing a tertiary carbon center were also obtainable by employing cinnamonitrile or crotononitrile as substrates, albeit in low yields (3ac and 3ad). We note that hydroacylation of α,β-unsaturated nitriles bearing strong electron-donating substituents at the β-position was challenging (3ae and 3af). In addition, the reaction was unsuccessful when applied to α-alkyl-substituted α,β-unsaturated nitriles (3ag–3ai).

Encouraged by the successful development of hydroacylation of α,β-unsaturated nitriles with acyl fluorides, we next investigated whether carbamoyl fluorides and fluoroformates could participate in the reaction with α,β-unsaturated nitriles under the established nickel-hydride catalysis to yield the desired hydrocarbamoylation and hydroesterification products, respectively. Upon screening of reaction conditions, we found that model carbamoyl fluoride 6a and fluoroformate 7a could be efficiently coupled with 1a to produce desired quaternary α-cyano amide 8a and ester 9a, respectively, in excellent yields. This was achieved under conditions slightly modified from the hydroacylation setup, with CsF replacing K₃PO₄ as the base (Table , entry 1). Control experiments revealed that the hydrocarbamoylation and hydroesterification did not proceed in the absence of nickel catalyst, base, or hydrosilane (entries 2–4).

3. NiH-Catalyzed Hydrocarbamoylation and Hydroesterification of 1a with 6a or 7aControl Experiments .

entry	deviation from standard conditions	yield, 8a (%)	yield, 9a (%)
1	none	98	88
2	w/o NiCl2·DME	n.d.	n.d.
3	w/o CsF	n.d.	n.d.
4	w/o (EtO)3SiH	n.d.	n.d.
5	w/o 2,6-lutidine	72	71
6	using other electrophiles (6b or 7b) instead of 6a or 7a	36	n.d.,

^aReaction conditions: 1a (2.0 equiv), 6a or 7a (0.1 mmol, 1.0 equiv), NiCl₂·DME (10 or 15 mol %), CsF (2.0 equiv), (EtO)₃SiH (2.0 equiv), 2,6-lutidine (2.0 equiv) and DMF (1.0 mL) for 12 h at room temperature. ¹H NMR yields are given.

^bNiCl₂·DME (10 mol %) was used.

^cNiCl₂·DME (15 mol %) was used.

^d 6b was used instead of 6a.

^e 7b was used instead of 7a.

^fYield of 9b.

The reaction efficiency of both hydrocarbamoylation and hydroesterification diminished when 2,6-lutidine additive was not employed (entry 5). We note that, similar to hydroacylation, a dramatic decrease in reaction efficiency was observed when carbamoyl chloride 6b and chloroformate 7b were employed instead of 6a and 7a, respectively (entry 6).

In terms of the substrate scope for hydrocarbamoylation, it was found that various carbamoyl fluorides could be coupled to a series of α,β-unsaturated nitriles to afford the desired products (Table A). Specifically, N,N-dialkyl carbamoyl fluorides, including those bearing morpholine (8c), piperazine (8d), and piperidine (8e) groups, were successfully employed in this carbamoylation, yielding the desired quaternary α-cyano amide products in moderate to good yields. Additionally, N,N-aryl,alkyl carbamoyl fluorides containing indoline (8f), 2-pyridyl (8g), 4-chlorophenyl (8h), 2-furanylmethyl (8i), and 2-thiophenylmethyl (8j) groups were also compatible in this transformation. The scope of hydroesterification was also examined using 1-adamantyl fluoroformate as the esterification reagent (Table B). A wide range of functional groups on α,β-unsaturated nitriles, including methoxy (9b), cyano (9c), thiomethyl (9e), and trifluoromethyl (9f), was well tolerated. Additionally, substrates containing heteroaromatic rings, such as thiophene (9d) and pyridine (9g), successfully participated in this transformation, yielding the corresponding quaternary α-cyano esters. Not only α-aryl cinnamonitrile derivatives but also α-aryl-β-alkyl unsaturated nitriles could be coupled with the fluoroformate (9h and 9i). Finally, it was found that 2-benzylidenemalononitrile could undergo the desired hydroesterification, affording the corresponding α,α′-dicyano ester in good yield (9j). While the scope of carbamoylation and esterification was found to be quite broad, carbamoyl fluorides bearing nitrile or allyl groups were incompatible with the current hydrocarbamoylation conditions (8k and 8l). Moreover, substrates including cyclic α,β-unsaturated nitrile, cinnamonitrile, and crotononitrile remained unreactive under the present hydroesterification reaction (9k–9m). ,

4. Scope of Hydrocarbamoylation and Hydroesterification of α,β-Unsaturated Nitriles with Carbamoyl Fluorides or Fluoroformate .

^aReaction conditions: α,β-unsaturated nitrile (2.0 equiv), carbamoyl fluoride or 1-adamantyl fluoroformate (0.2 mmol, 1.0 equiv), NiCl₂·DME (10 or 15 mol %), CsF (2.0 equiv), (EtO)₃SiH (2.0 equiv), 2,6-lutidine (2.0 equiv) and DMF (2.0 mL) for 12 h at room temperature. Isolated yields are given.

^bNiCl₂·DME (15 mol %) was used.

^cThe reaction was performed at 0.2 M in DMF solvent.

^dRun for 24 h.

^eNiCl₂·DME (15 mol %) was used.

^{f 1}H NMR yields are given.

^g(E)-cinnamonitrile was used.

^h E/Z mixture of crotononitrile was used.

We then aimed to showcase the synthetic utility of the present nickel-catalyzed hydrofunctionalization protocol (Scheme ). First, the modification of natural products and drug derivatives was performed (Scheme A). Acyl fluorides derived from probenecid, febuxostat, oxaprozin, and naproxen successfully participated in the reaction, affording the desired hydroacylation products (10a–10d) in synthetically acceptable yield. In addition, the reaction with an estrone-derived carbamoyl fluoride also gave the corresponding α-cyano amide product (10e). The α-tocopherol derived α,β-unsaturated nitrile substrate also underwent a hydroesterification reaction with adamantyl fluoroformate, affording the desired product (10f).

4. Synthetic Application,,

^a NiCl₂·DME (10 mol %) was used.

^b NiCl₂·DME (15 mol %) was used.

^c Reagents and conditions: (i) CeCl₃ (1.0 equiv), NaBH₄ (1.2 equiv), EtOH, room temperature, 2 h; (ii) DAST (30.0 equiv), CH₂Cl₂, room temperature, 12 h; (iii) K₂CO₃ (2.0 equiv), H₂O₂ (5.0 equiv), DMSO, room temperature, 20 h; (iv) NiCl₂·6H₂O (3.0 equiv), NaBH₄ (20.0 equiv), (Boc)₂O (3.0 equiv), MeOH, 0 °C to room temperature, 12 h; (v) conc. HCl (0.7 M),1,4-dioxane, reflux, 12 h. Isolated yields are given. See the SI for the detailed experimental procedures.

The robustness and practicality of the present reaction system were further validated through the gram-scale syntheses of α-cyano carbonyl compounds (Scheme B). It was observed that the reaction of 1a with carbonyl fluorides 2a, 6a, or 7a on a 5.0 mmol scale proceeded smoothly, affording the corresponding products in synthetically useful yields. We note that although the yields were somewhat reduced, PMHS was utilized as a less toxic and more cost-effective hydride source in place of (EtO)₃SiH for these larger-scale syntheses.

Given that the obtained quaternary α-cyano ketones, amides, and esters are versatile building blocks in organic synthesis, several derivatizations of these compounds were investigated (Scheme C). The α-cyano ketone 3a was converted to the corresponding alcohol (11) in excellent yield and with high diastereoselectivity under the conventional Luche reduction conditions. Reaction of 3a with DAST afforded the corresponding difluoromethylene compound (12) in moderate yield. Hydrolysis of α-cyano amide 8a resulted in the formation of the corresponding diamide (13). Compound 8a could also be transformed to the β-amino amide (14) via nickel-mediated reduction of the nitrile followed by Boc-protection. Finally, reduction of nitrile in α-cyano ester 9a, followed by hydrolysis under acidic conditions, was accomplished to yield the corresponding β-amino acid in its HCl salt form (15).

Conclusions

In summary, we have developed an efficient and general synthetic method for quaternary α-cyano carbonyl compounds via nickel-hydride-catalyzed hydrofunctionalization of α,β-unsaturated nitriles with acyl fluorides, carbamoyl fluorides, and fluoroformate. This protocol operates under mild conditions without the need for an external ligand. The synthetic practicality and versatility of this method are further demonstrated through the late-stage modification of pharmaceutical derivatives, gram-scale syntheses, and product derivatizations into high-value molecules. Computational and experimental mechanistic studies reveal that the present nickel-catalyzed hydrofunctionalization proceeds through regioselective hydrometalation, followed by the formation of a nickel-keteneiminate intermediate and subsequent nucleophilic substitution. Ongoing efforts are aimed at further elucidating the mechanistic details of the hydrocarbamoylation and hydroesterification as well as developing an asymmetric variant of this transformation.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/jacsau.5c00835.

• Detailed experimental procedures, full characterization data, copies of ¹H, ¹³C, and ¹⁹F NMR spectra of the new compounds (PDF)

The authors declare no competing financial interest.