Ni/NHC-catalyzed C5-H alkylation and alkenylation of challenging furan(thiophene)-2-carboxaldehydes enabled by recyclable imine protecting group

Abstract

Transition-metal-catalyzed C-H alkylation of heteroaromatics with alkenes represents an atom-economical and cost-effective strategy for accessing industrially and pharmaceutically relevant compounds. However, the selective C5-H alkylation of biomass-derived furfural and its isosteric analog, thiophene-2-carboxaldehyde, highly challenging yet industrially vital substrates, has remained elusive. Herein, we disclose a Ni/NHC-catalyzed strategy for the C5-H alkylation of furan- and thiophene-2-carboxaldehydes with styrenes and norbornene, enabled by a readily installable and recyclable N-PMP (p-methoxyphenyl) imine protecting group. This method also achieves selective C5-H alkenylation with internal alkynes. Mechanistic studies suggest that C-H alkylation proceeds via a ligand-to-ligand hydrogen transfer (LLHT) pathway. The N-PMP imine group effectively suppresses undesirable benzoin condensation of these reactive aldehydes and prevents NHC trapping in Breslow intermediates, a major catalyst deactivation pathway. The protecting group is efficiently cleaved under acid hydrolysis, yielding C5-functionalized aldehydes, while the liberated anisidine can be recycled for imine substrate preparation. This work also highlights the largely unexplored potential of the N-aryl imine group as the protecting group for distal C(sp²)-H functionalization of heteroaromatic aldehydes under Ni catalysis.

A Ni/NHC-catalyzed strategy for C5-H alkylation and alkenylation of furan- and thiophene-2-carboxaldehydes with alkenes and alkynes has been developed, leveraging an easily installable and recyclable N-PMP imine protecting group. This group suppresses unwanted benzoin condensation of the reactive aldehydes and avoids NHC trapping in Breslow intermediates—a key catalyst deactivation pathway.

Introduction

Atom- and step-economical C–H alkylation and alkenylation of heteroaromatic substrates with alkenes and alkynes, the so-called hydroheteroarylation reactions, represent highly promising, cost-effective, and environmentally benign strategy for the synthesis of alkylated and alkenylated heteroaromatic scaffolds demanded in biomedical research, medicine, agrochemistry, polymers, and materials chemistry^1–6. Nickel catalysts have drawn considerable attention in hydroheteroarylation reactions owing to their high efficiency and relative low cost of nickel compared to noble metals (Fig. 1a)^7–20.

Fig. 1. 2-Functionalyzed 5-alkyl and 5-alkenyl furans and thiophenes.

a Nickel-catalyzed alkylation and alkenylation of furan(thiophene) derivatives with alkenes and alkynes. b Representative examples of practically valued compounds. c This work: C5-H alkylation and alkenylation of furan(thiophene)-2-carboxaldehydes enabled by a recyclable imine protecting group.

Heterocyclic substrates containing various functional groups are particularly attractive for C-H functionalization because they allow the synthesis of new, selectively functionalized molecules capable of further structural modifications^21–23. However, the presence of functional groups in addition to active C-H bonds raises the problem of selectivity of reactions of such substrates, especially under conditions of metal catalysis. Formyl group represents one of the most widely used functional group; however, C-H alkylation and alkenylation of heterocyclic rings in heteroaromatic aldehydes can be problematic due to the possibility of C-H activation of the formyl group and competitive hydroacylation reactions^24–26. Although examples of selective C-H alkylation and alkenylation of some heteroaromatic substrates with a formyl group under Ni catalysis are reported in the literature^15,27, heteroaromatic aldehydes are still remain challenging substrates.

Among the heteroaromatic aldehydes, furfural (furan-2-carboxaldehyde) is considered a promising renewable biobased feedstock for the production of chemicals and fuels^28–38. The global production of furfural from agricultural residues and forestry wastes is ~370 kt/a (according to 2016 data)³¹ and has the potential to increase to 1 Gt/a in the future, exceeding the current demand for chemicals and aviation fuels of 0.4 and 0.2 Gt/a, respectively³⁴. Among furfural derivatives, 5-alkyl- and 5-alkenyl-furan-2-carboxaldehydes have received application as reagents in the preparation of medicinally oriented compounds^39,40, plant effectors⁴¹, and photoactive materials (Fig. 1b)^42–45. Conventional synthesis of 5-alkylated and 5-alkenylated furfurals via Friedel-Crafts reactions faces challenges due to furfural’s instability, while alternative routes using C5-prefunctionalized furfural derivatives (lithiated^46–48, halogenated^49–77, or borylated^43,78–80) generate substantial waste despite good selectivity. Recent advances in C–H functionalization of furfural⁸¹, such as C5-H alkylation with alkyl halides under Pd^82–84, Fe⁸⁵ or photoinduced Pd^86,87 catalysis, or with carboxylic acid peroxides under Fe catalysis⁸⁸, as well as Pd-catalyzed oxidative alkenylation with alkenes^89–92 offer more sustainable approaches, but these methods remain constrained by limited substrate compatibility and stringent reaction conditions.

C-H alkylation and alkenylation of furfural with alkenes and alkynes seems to be highly promising and environmentally benign strategy for the synthesis of furfural derivatives, though it still remains immature. Approaches to the selective C5-H alkenylation of furfural with alkynes under Rh and Pd catalysis have been developed^93,94. However, to the best of our knowledge, similar reactions between furfural and alkenes, yielding 5-alkylfurfurals, remain unexplored. The only example of furfural acetal alkylation with decene-1 under catalysis with a Ni(0)/MHC complex with a hindered IPr*^OMe ligand has been reported by Hartwig and coworkers¹³. However, to our knowledge, the synthesis of 5-functionalized furan- and isosteric thiophene-2-carboxaldehydes by metal-catalyzed hydroheteroarylation reactions of appropriate aldehydes or their protected derivatives has not yet been reported.

Herein, we report an efficient approach for the selective C5-H alkylation of furfural, thiophene-2-carbaldehyde and some of their derivatives with styrenes and norbornene, as well as the selective C5-H alkenylation with internal alkynes, under Ni/NHC catalysis (Fig. 1c). Key to this approach is the use of an N-PMP (p-methoxyphenyl) imine protecting group, which is readily installed, removable in situ, and recyclable. Mechanistic studies show that the N-PMP imine group plays a dual role: (1) suppressing deleterious benzoin condensation and (2) preventing deactivation of the Ni/NHC catalyst by highly reactive aldehydes. Notably, despite the potential of the N-PMP imine group to act as an N-donor directing group—which could favor C3-H functionalization—the reaction exhibits exceptional selectivity for C5-H alkylation/alkenylation. An important novelty of this study lies in demonstrating the utility of imine groups as protecting groups for remote C–H functionalization of carbonyl-containing heteroaromatics under Ni catalysis.

Results and discussion

Attempted alkylation and alkenylation of unprotected furfural

We started our study by investigating various Pd- and Ni-based catalytic systems for the reaction of furfural (1a) with styrene (2a) in 1,4-dioxane (Fig. 2, Supplementary Table S1). The catalytic systems were generated in situ from Pd(dba)₂, Pd(OAc)₂, or NiCp₂ (nickelocene) in combination with phosphine ligands 5a-e or NHC proligands 6a-d (Fig. 2a). Additionally, well-defined Ni/NHC complexes 7a-f were evaluated as precatalysts (Fig. 2a).

Fig. 2. Attempted alkylation of furfural (1a) with styrene (2a).

a Ligands, Pd and Ni salts and complexes used in the study. b Attempts to react 1a with 2a under Pd and Ni catalysis. Reaction conditions: furfural 1a (1 mmol), styrene 2a (0.25 mmol), [Pd] or [Ni] (5-10 mol%), (pro)ligand (5-10 mol%, if required, Table S1), HCOONa (100 mol%, or another additive, if required, Table S1), dioxane (1 mL), 16 h, 110 °C, argon atmosphere. c The main route of catalyst decomposition under the action of furfural.

Unfortunately, none of the tested Pd or Ni catalytic systems facilitated the formation of the desired alkylation product 3a (Fig. 2b, Supplementary Table S1; extended discussion in the ESI, Section S3.1). Attempted alkenylation of furfural with diphenylacetylene under Pd/NHC and Ni/NHC catalysis was also unsuccessful (Supplementary Table S2). The conversion of styrene and diphenyl acetylene did not exceed 10%. However, compound 4 was identified as the major product in both Pd/NHC- and Ni/NHC-catalyzed reactions (Fig. 2b, Supplementary Tables S1 and S2). Notably, furfural conversion was nearly complete under Ni/NHC catalysis, producing 4 in 54–76% (attempted alkylation) and 63–80% (attempted alkenylation) yields. Unidentified tar-like furfural decomposition byproducts were also observed.

Compound 4 arises from the NHC-catalyzed benzoin condensation of furfural^95–100, mediated by NHCs generated via deprotonation of azolium salt-NHC-proligands 6a-d or decomposition of Ni/NHC complexes 7a–f^101,102. The significantly lower yields of 4 under Pd/NHC catalysis, as opposed to Ni/NHC, can be explained by the greater stability of Pd/NHC complexes under the reaction conditions.

Further studies revealed that furfural rapidly decomposes Ni/NHC complexes 7a-f in the presence of HCOONa (Supplementary Fig. S2; Supplementary Information, Section S3.1) via formation of the Breslow intermediate (BI, Fig. 2c). The high reactivity of furfural promotes Ni–C(NHC) bond dissociation by trapping the NHC as the BI, a key intermediate in the benzoin condensation cycle^98,99. However, the BI is thermally unstable and gradually degrades (Supplementary Fig. S3), leading to NHC ligand depletion and eventual breakdown of the Ni/NHC catalytic system. While the benzoin condensation of aldehydes catalyzed by NHCs in the presence of Ni complexes is well-documented¹⁰³, the aldehyde-induced decomposition of M/NHC complexes via NHC trapping in the Breslow intermediate (Fig. 2c) represents, to our knowledge, a previously unreported phenomenon.

Thus, the failure of furfural alkylation and alkenylation under Ni/NHC catalysis, despite previous successes with C5-alkylation of other furan derivatives^14,16, can be attributed to furfural-induced degradation of Ni/NHC systems and its high propensity for NHC-catalyzed benzoin condensation.

Protecting group selection and reaction optimization

To suppress Ni/NHC catalyst deactivation and competing benzoin condensation, we investigated readily available furfural derivatives containing protected formyl groups specifically, acetal 8 (previously employed by Hartwig’s group)¹³ and N-arylimines 9a–d as substrates for the reaction with styrene (Fig. 3). It should be noted that in metal-catalyzed C-H functionalization reactions, imine groups have been predominantly used as N-donor directing^104–109 or functional^{24,26,110–112} groups. Surprisingly, the potential of the imine group as a protecting group in transition metal-catalyzed distal C-H functionalization reactions remains understudied.

Fig. 3. Alkylation of furfural derivatives with a protected formyl group.

Reaction conditions: compound 8,9a-d (0.25 mmol), styrene 2a (0.3 mmol), 7d (10 mol%), HCO₂Na (0.25 mmol), 1,4-dioxane, 110 °C, 16 h.

The stable Ni/NHC precatalyst 7d and sodium formate (as a precatalyst activator) were used under conditions analogous to those recently reported for the alkylation and alkenylation of heterocyclic compounds (Fig. 3)¹⁴. Alkylation of substrate 8 proved unsuccessful, with only trace amounts of the proposed product 10 detected by GC-MS. In contrast, imines 9a–d yielded the target alkylated products 11a–d in 41–95% yield (GC-MS). Notably, no benzoin condensation product (4) was observed by GC-MS, underscoring the stability of the imine protecting group under the reaction conditions. The observed reactivity difference between compounds 8 and 9 is probably due to the different electronic properties of their substituents at position 2 of the furan ring. The electron-donating 1,3-dioxolane group in 8 increases the electron density of the furan ring, thereby decreasing the C-H acidity and consequently increasing the reactivity under the C-H alkylation conditions studied. In contrast, the imine-protected carbonyl group in 9 retains its electron withdrawing character through negative resonance effect, thus maintaining sufficient C–H acidity and reactivity of the furan nucleus. The identity of the N-aryl group in the imine-protected substrates (9a–d) had a significant impact on the yields of alkylated products (11a–d). The significantly lower yield of product 11c (41%) compared to products 11a, 11b, and 11d (Fig. 3) can be explained by proposed partial catalyst inhibition due to nickel coordination with the N,N-dimethylamino group in substrate 9c. Substrates 9b and 9d produced comparable yields of the alkylation products 11b (94%) and 11d (95%), respectively Fig. 3. However, the N-PMP imine group seems the best choice for protecting the carbonyl group in alkylating substrates due to the significantly higher yield of compound 9d (91%) compared to 9b (24%) during preparation from furfural and the corresponding aromatic amines (Supplementary Information, Section S2.2), as well as the better crystallizability of the N-PMP imines derived from numerous aldehydes.

Based on the experimental results (Fig. 3), compound 9d, synthesized through the condensation of furfural with p-methoxyaniline, emerged as the most effective substrate for achieving selective C5-H alkylation of the furan ring. Consequently, the N-PMP group was identified as the optimal protecting group for nickel-catalyzed C5-H functionalizations involving furfural and related aldehyde substrates.

The reaction conditions were optimized (Table 1 and Supplementary Table S3). Nickel systems with phosphine ligands were found to be completely inactive (Table S3, entries 1-6). Among the well-defined Ni/NHC complexes 7a–f (Table 1, entries 1-6), complex 7d (entry 4) was found to be the most efficient precatalyst in the presence of HCOONa as an activator. The inefficiency of complexes 7a (entry 1) and 7b (entry 2) may be related to the detrimental effect of the sigma-coordinated second NHC (in 7a) or Cp ligand (in 7b). The low efficiency of complexes 7c (entry 3) and 7e (entry 5), which have the same structural type as 7d, may be explained by the less suitable steric parameters provided by the N-aryl substituents in the NHC ligands. These results agree with a previous report identifying mesityl-substituted NHCs as optimal ligands for the (NHC)Ni(Cp)Cl-catalyzed C–H alkylation of heteroaromatics with styrene¹⁴. We also tested complex 7a as a self-activated precatalyst in strongly alkaline media using tBuONa as an additive (the second NHC can serve as a sacrificial ligand, from Ni(II) to Ni(0) reductant, in alkaline medium)^113,114, but the yield of 11d was moderate (entry 7) and decreased with prolonged reaction time (entry 8), apparently due to the instability of the reaction product 11d under harsh reaction conditions. Complex 7b, which was reported as a thermally activated precatalyst due to the possibility of Ni(II) reduction to Ni(0) via reductive elimination of two Cp ligands at elevated temperatures¹⁸, was also found to be an underperforming precatalyst, as only a moderate yield of 11d was obtained (entry 9). Considering the results of the experiments described in entries 1–9 (Table 1), the catalytic system consisting of complex 7d and sodium formate was found to be the most effective. Different solvents were also tested, and 1,4-dioxane was found to be the best (compare entry 4 with entries 10–13). Increasing the reaction temperature from 110 °C to 120 °C (and higher) did not increase the yield, whereas decreasing the temperature to 100 °C (and lower) significantly decreased the yield (Table 1, comparing entries 4, 14, and 15; Supplementary Table S3, entries 22–25). Therefore, 110 °C was selected as the optimal reaction temperature. Systematic variation of complex 7d and HCOONa loading failed to improve the yield of 11d (Supplementary Table S3, entries 26–28). The reaction time proved critical: shorter durations substantially reduced the yield (Supplementary Table S3, entries 29–30), while extending to 24 h afforded 11d in 98% GC-MS yield (91% isolated yield; Table 1, entry 16). Control experiments confirmed the essential role of both 7d and HCOONa, as no product formation occurred in the absence of either (Supplementary Table S3, entries 32-33). This requirement for sodium formate is consistent with its proposed role in reducing Ni(II) to active Ni(0) species via intermediate Ni-hydride complexes (e.g., 7f)¹⁴. Supporting this mechanism, preformed 7f catalyzed the reaction without HCOONa (Table 1, entry 17). An in situ generated Ni/NHC system from NiCp₂ and IMes·HCl proved less effective than well-defined precatalyst 7d (Table 1, entries 16 vs 18). These results established the conditions of entry 16 (Table 1) as optimal for the alkylation.

Table 1.

Optimization of the reaction conditions

Entry	Ni/NHC	Additive	Solvent	Time, h	T, oC	Yield of 11d, (%)
1	7a	HCO2Na	dioxane	16	110	trace
2	7b	HCO2Na	dioxane	16	110	trace
3	7с	HCO2Na	dioxane	16	110	trace
4	7d	HCO2Na	dioxane	16	110	95
5	7e	HCO2Na	dioxane	16	110	37
6	7f	HCO2Na	dioxane	16	110	91
7	7a	tBuONa	o-xylene	1	150	51
8	7a	tBuONa	o-xylene	20	150	trace
9	7b	–	o-xylene	16	150	61
10	7d	HCO2Na	o-xylene	16	110	81
11	7d	HCO2Na	toluene	16	110	86
12	7d	HCO2Na	DMF	16	110	0
13	7d	HCO2Na	NMP	16	110	0
14	7d	HCO2Na	dioxane	16	120	94
15	7d	HCO2Na	dioxane	16	100	64
16	7d	HCO2Na	dioxane	24	110	98 (91)[a]
17	7f	–	dioxane	16	110	89
18	NiCp2[b] + IMes·HCl[b]	HCO2Na	dioxane	24	110	85

^[a] Isolated yield.

^[b] 10 mol%.

Synthetic scope

With the optimized conditions in hand, we explored the scope and limitations of the reaction (Fig. 4). Furan N-arylimines 9d–e and thiophene N-arylimines 9g–i were investigated in the reactions with different alkenes (Supplementary Fig. S1). Arylation products such as 11d and similar suffered from partial hydrolysis of the imine protecting group during isolation and purification (especially during column chromatography). Therefore, the primary reaction products were subjected to hydrolysis of the imine group with a 1 M aqueous HCl solution without isolation to give the target aldehydes 3, which were isolated and purified. Substrates 9d, 9e, 9g-i were successfully alkylated with various styrenes and norbornene to give products 3a–o in 31-90% isolated yields (Fig. 4a). The presence of an aryl group at position 3 of the furan or thiophene ring did not significantly affect the reaction yields (products 3f, 3g, 3m and 3n), whereas all attempts to alkylate the N-PMP derivative of 4-phenylfurfural (9j, Supplementary Fig. S1) as well as the N-PMP derivative of N-methylpyrrole (9k, Supplementary Fig. S1) were unsuccessful. Apparently, the reaction is highly sensitive to steric and, probably, electronic factors of the heteroaromatic substrate.

Fig. 4. Scope of the Ni/NHC-catalyzed alkylation and alkenylation of furan- and thiophene-2-carboxaldehyde N-PMP-imines.

a Ni/NHC catalyzed alkylation. b Ni/NHC catalyzed alkenylation. ^a Reaction conditions: 7d (10 mol %), HCO₂Na (0.25 mmol), 9d–i (0.25 mmol), 2a–f (0.3 mmol) or 12a–h (0.5 mmol), dioxane (1 mL), 110 °C, argon atmosphere, 24 h, then hydrolysis with 1 M aq. HCl at 25 °C for 16 h; isolated yields are shown. ^b Isolated yields of a mixture of stereoisomers. ^c At 140 °C in o-xylene.

Styrenes 2a-e (Supplementary Fig. S1) reacted selectively to give branched regioisomers 3a–g,i–n, which is consistent with previously reported regioselectivity for ligand-controlled Ni/NHC-catalyzed hydroarylation reactions (NHC = IMes or similar NHCs with moderate steric bulkiness) in the absence of special Lewis acid additives that can affect regioselectivity^{14,17,115–117}. Aliphatic linear alkenes such as 1-hexene, acrylates, and stilbenes were nearly inactive and provided only trace yields of alkylation products (Supplementary Fig. S1). The same alkenes were found to be inactive in the alkylation of many other heteroaromatic substrates under similar conditions¹⁴. In contrast, norbornene, which has a more reactive double bond due to ring tension, gave product 3h in 76% isolated yield. However, in the reaction with the thiophene derivative 9g, norbornene gave only a moderate yield of the alkylation product 3o (31%), presumably due to the lower reactivity of 9g compared to 9d (~50% of the starting 9g remained intact after the reaction) and the greater losses of norbornene due to side polymerization. Remarkably, the reaction involved only the C5-H bond of the furan (thiophene) ring for all alkenes. Products that might be formed via side activation of the imine C-H functional group, such as hydroalkenylation^118,119, or reductive alkylation^120,121, were not detected.

The scalability of the developed procedure was demonstrated by the gram-scale syntheses of compounds 3a and 3i in 79% and 72% yields, respectively (see SI, sections S2.6 and S2.7). Remarkably, p-methoxyaniline formed after hydrolysis of the alkylated imine intermediate can be recovered and reused. This was demonstrated in the gram-scale syntheses of compounds 3a and 3i. Aqueous HCl solutions formed after the isolation of products 3a and 3i were alkalinized with aqueous NH₃, and p-methoxyaniline was extracted with ether and then reacted with furfural or thiophene-2-carboxaldehyde to give starting compounds 9d and 9g, which were isolated (after crystallization from hexane) in 70% and 60% yields, respectively, and successfully used as substrates in the alkylation reactions without further purification.

In addition, substrates 9d–i were selectively alkenylated with internal alkynes 12a-h to give products 13a–q in 54-86% isolated yield (Fig. 4b). A twofold molar ratio of alkyne to heterocyclic substrate was used in the alkenylation due to alkyne losses by side trimerization under the reaction conditions. Due to the reduced electrophilicity of alkynes containing TMS groups, the reaction with these alkynes was carried out at 140 °C in o-xylene to obtain better yields of compounds 13c–e,k,l–n (Fig. 4b). Alkenylation, in most cases, afforded the E-stereoisomer (with a cis-orientation of the aryl or alkyl groups relative to the double bond in the alkenyl moiety) as the only isolated product. Only products 13b (Z-isomer content ~15%) and 13j (Z-isomer content ~20%) were isolated as mixtures of stereoisomers. This result is consistent with the commonly observed stereoselectivity of Ni-catalyzed hydroheteroarylations of alkynes, which typically afford E-stereoisomers as the dominant products^{14,27,122–124}. Reactions with unsymmetrical alkynes containing a TMS group on one of the acetylenic carbons selectively afforded regioisomers with a TMS group on the ethylenic carbon distal to the furan(thiophene) moiety, in agreement with previously observed regioselectivity of Ni-catalyzed hydroheteroarylations of similar alkynes^{14,27,122,123}.

The structures of compounds 3a–o and 13a–q were unambiguously confirmed by NMR spectroscopy, including NOESY, ¹H-¹³C HSQC, and ¹H-¹³C HMBC experiments (see SI, section S5).

Mechanistic considerations

A plausible reaction mechanism (Fig. 5a) involves the initial in situ formation of active Ni(0)/NHC complexes by reduction of complex 7d with sodium formate via the formation of the intermediate (NHC)Ni(Cp)H complex 7f, which undergoes reductive elimination of cyclopentadiene to give active Ni(0)/NHC complexes such as 14 (L = cyclopentadiene or solvent, etc.), which enter the catalytic cycle¹⁴. Based on literature data, two alternative mechanisms of catalysis involving complex 14, namely substrate C-H oxidative addition^125,126 and ligand-to-ligand hydrogen transfer (LLHT)^126,127, can be postulated.

Fig. 5. Mechanistic studies.

a Proposed mechanism of Ni/NHC-catalyzed C-H alkylation. b Kinetic isotope effect (KIE) measurement. c Attempted competitive C3-H alkylation. d Observation of C3-D for C3-H exchange. e The suggested roots for catalyst deactivation.

The first mechanism, observed experimentally for Ni/NHC catalyzed reactions of some heteroaromatic substrates^{14,17,117,128–130}, involves coordination of the substrate to the complex 14 to form the intermediate complex A. Of course, complexes with alternative coordination modes can also participate in the equilibrium with the intermediate A as resting states (as an example, Ni-N coordination in the apical position to the metal is shown in B, and similar interactions can occur in other structures, which are omitted for simplicity). Complex A undergoes oxidative addition (OA) of the substrate C-H bond to form a heteroleptic nickel hydride complex 15, which reversibly coordinates alkene (or alkyne) to form complex 16. Subsequent insertion of alkene (or alkyne) into a Ni–H bond yields the organonickel intermediate 17, which reductively eliminates the alkylation (or alkenylation) product 11d. In this type of mechanism, the oxidative C-H addition is usually considered to be the rate-determining step.

The alternative LLHT mechanism, accepted for Ni/NHC catalyzed reactions of many heteroaromatic substrates^{13,127,131–135}, involves sequential coordination of the substrate 9d and alkene (or alkyne) with 14 to give the intermediate 18, which undergoes concerted oxidative addition-migratory insertion to give the intermediate 17 in one step, without the formation of the hydride complexes 15 and 16. In the LLHT mechanism, reductive elimination of the alkylation (or alkenylation) product^127,132,134, or LLHT^131,133,135, is typically rate-determining step.

To gain more insight into the reaction mechanism, we performed electrospray ionization mass spectrometry (ESI-MS) and NMR spectroscopy studies of the reaction mixtures. We also conducted kinetic experiments with the deuterated substrate. Contrary to previous observations^14,117,129, attempts to detect the hydride intermediate (15) by ¹H NMR were unsuccessful. In parallel experiments on the alkylation of 9d and deuterated 9d-D(5) with styrene (Fig. 5b, Supplementary Fig. S4), we determined the kinetic isotope effect (KIE). The low value of the determined KIE = 1.08 suggests that the C-H activation is not involved in the rate-limiting step, making the LLHT mechanism more likely. However, only ~50% of the deuterium incorporation was observed in the coupling product 11d-D (Scheme S1). The deuterium loss could be due to reversible oxidative addition of the substrate C − D bond to Ni with subsequent deuterium for protium exchange or, alternatively, ligand-to-ligand hydrogen transfer (LLHT) followed by β-H elimination. Considering numerous literature data on Ni catalyzed hydroheteroarylation¹³⁶, the widely accepted LLHT mechanism with the rate-limiting reductive elimination step seems to be the most likely reaction route for the alkylation of furfural imines. However, alternative oxidative C − H addition mechanism could not be completely ruled out.

Another interesting question concerns the origin of the high selectivity for C5-H alkylation/alkynylation and the absence of C3-H functionalization byproducts. In many metal-catalyzed C-H functionalizations, the imino group serves as a directing group^104–109, suggesting that C3-H alkylation of the furan (or thiophene) core might be expected^137,138. However, when substrate 9l (where the C5 position of the furan is blocked by a methyl group) was subjected to alkylation, no C-H functionalization products were observed (Fig. 5c and discussion in Supplementary Information, Section S4.2). Notably, alkylation of deuterium-labeled substrate 9d-D(3) (with deuterium at the furan C3 position) yielded exclusively the C5-H alkylation product, yet over half of the deuterium was exchanged for protium. This implies reversible nickelation of the C3-D(H) bond (Fig. 5d and Section S4.3). While C3-H metallation thus occurs, C3-alkylation products are conspicuously absent. We attribute this to the pronounced sensitivity of nickel-catalyzed C-H alkylation/alkenylation to steric constraints. Specifically, steric hindrance from the N-aryl imino group likely disfavors the formation of the requisite transition state involving the alkene (or alkyne) and C3-nickelated intermediates, thereby suppressing C3-H alkylation/alkenylation.

Plausible routes for catalyst deactivation include the formation of the H-NHC (6b) and R-NHC (19 and 20) coupling products¹⁰¹, as detected by the ESI-MS spectra of the reaction mixtures (Fig. 5e, Supplementary Fig. S6). These deleterious NHC reductive elimination routes have been previously observed in heteroarylation reactions of alkenes and alkynes with various substrates¹⁴.

We also investigated the possible role of Ni nanoparticles (Ni NPs), which are expected to be formed by the decomposition of Ni/NHC complexes, in the catalysis of the investigated alkylation and alkenylation reactions. The sampling of colloidal metal particles from reaction mixtures after the reaction between compounds 9d and 2a catalyzed by complex 7d in the presence of HCOONa was carried out using a “nanofishing” technique. The “nanofishing” procedure is implemented by dipping a fresh thin carbon-coated grid directly into a reaction mixture to capture dynamic metal particles from the solution¹³⁹. Transmission electron microscopy (TEM) study revealed the presence of metal-containing nanoparticles in the reaction mixture (Supplementary Fig. S7). The reaction solutions were centrifuged, black precipitates were collected, washed, and analyzed by energy dispersive X-ray analysis (EDX). The presence of Ni in the collected precipitate was clearly demonstrated (Supplementary Fig. S8). It should be noted that the formation of Ni NPs is not surprising, since such precedents are known for reactions involving Ni(0) species in catalytic process^140,141. Participation of NHC-ligated Ni NPs in catalysis as alternative active centers^140,142,143, or Ni NPs as metal reservoirs¹⁴⁴, should not be completely excluded. Therefore, a hot-filtration test was performed to clarify the potential role of Ni NPs in the studied catalytic systems using the example of the reaction between compounds 9d and 2a (Supplementary Fig. S9). The reaction was carried out at 110 °C in an argon atmosphere in two parallel experiments. In the control experiment (without hot filtration of the reaction mixture), GC-MS analysis revealed a 47% yield of 11d after 8 h and a 97% yield after 24 h from the start. In the parallel experiment, the yield of 11d also reached about 50% initially. The reaction mixture was then subjected to hot filtration under an argon atmosphere to remove the catalyst. The filtrate was then transferred to a new reactor and subjected to further heating and stirring for another 16 h under identical conditions, resulting in a 79% yield of 11d at the end of the experiment (Supplementary Fig. S9). The data suggest that the majority of the active nickel species are predominantly present in the solution. However, the decrease in 11d yield after hot filtration does not exclude the possibility of a cocktail-type mechanism involving both homogeneous and nanoparticulate Ni/NHC active species (e.g., NHC-ligated Ni NPs). Another proposed reason for the decrease in 11d yield after hot filtration could be the partial decomposition of active Ni/NHC complexes by the filter material (celite) used in the hot filtration (see discussion in SI, section S4.6). Overall, the data obtained suggest that catalysis with homogeneous Ni/NHC complexes dominates.

Conclusions

The present study introduces an efficient approach to overcome the limitations of the Ni/NHC catalytic systems in reactions of highly active aldehyde substrates, leading to catalyst deactivation. The approach utilizes N-aryl imine protection of the formyl group to enable distal C(sp²)-H functionalization of the heteroaromatic core under Ni catalysis. While imine groups have been widely used as N-donor directing or functional groups in C-H functionalization chemistry, their utility as protective groups in metal-catalyzed C-H transformations has remained largely unexplored.

An efficient method for the C5-H alkylation of furan- and thiophene-2-carboxaldehydes with styrenes and norbornene under Ni/NHC catalysis has been developed. The key to the process is the use of a recyclable N-PMP imine protecting group. Thorough studies of the reaction have shown that the protecting group allows suppression of deleterious benzoin condensation, which leads to rapid decomposition of aldehyde substrates. In addition, N-PMP protection prevents deactivation of the catalytic system by highly reactive aldehydes that decompose Ni/NHC complexes by trapping NHC ligands in the Breslow intermediates. The developed approach is also suitable for the synthesis of C5-alkenylated derivatives of furfural and thiophene-2-carboxaldehyde via reactions with internal alkynes. Other practical advantages of the N-PMP imine protecting group include the simplicity of its installation (by condensation of the corresponding aldehyde with p-methoxyaniline), the readiness of its removal (by acid-catalyzed hydrolysis without isolation of intermediate alkylated and alkenylated imines), and the possibility of recovering and reusing anisidine after deprotection of the formyl group.

The Ni/NHC catalytic system is easily generated from bench-stable, air-tolerant (IMes)Ni(Cp)Cl precatalyst and sodium formate, the Ni(II) to Ni(0) reductant, without the use of unstable Ni(0) sources or hazardous Ni(II) reductants.

To the best of our knowledge, the developed approach represents the first example of the synthesis of C5-alkylated furfurals and thiophene-2-carboxaldehydes via metal-catalyzed hydroheteroarylation with alkenes.

Methods

Synthesis of compounds 3a–o (general procedure)

An oven-dried vial equipped with a magnetic stirring bar and a septum was charged with 7d (12 mg, 0.025 mmol, 10 mol %), compound 9d,e,g–i (0.25 mmol), alkene 2a–f (0.3 mmol) and dioxane (1 mL). The resulting mixture was purged with argon through a syringe via a septum and then heated at 110 °C with vigorous stirring within 24 h (see Fig. 4a). After cooling to room temperature, the mixture was diluted with dioxane (3 mL) and filtered through a short pad of Celite. The solvent was then rotary evaporated under vacuum, and the residue was dissolved in diethyl ether (4 mL). 1 M aqueous HCl (4 mL) was added to the ether solution, and the mixture was stirred overnight at 25 °C. The ether layer was separated by decanting. The aqueous layer was extracted with diethyl ether (3 × 5 mL). The ether solution separated from the reaction mixture, and the extracts were combined, dried over MgSO₄ and rotary evaporated to dryness. The oily residue obtained was chromatographed on SiO₂ (elution with hexane-EtOAc 10:1 mixture) to obtain desired products 3a–o.

Synthesis of compounds 13a–q (general procedure)

An oven-dried vial equipped with a magnetic stirring bar and a septum was charged with 7d (12 mg, 0.025 mmol, 10 mol %), compound 9d–i (0.25 mmol), alkyne 12a–i (0.5 mmol) and dioxane (1 mL, for alkyne 12a–c, h, i) or o-xylene (1 mL, for alkyne 12d–g). The resulting mixture was purged with argon through a syringe via a septum and then heated at 110 °C (for dioxane) or 140 °C (for o-xylene) with vigorous stirring within 24 h (see Fig. 4b). The resulting mixture was then treated as described for the synthesis of products 13a–q.

Author contributions

O.V. Khazipova, O.V. Khazipov, and K.E. Shepelenko designed and performed synthetic experiments, analyzed experimental data. A.S. Kashin carried out the TEM and EDX analysis. O.V. Khazipov, Yu Zhang, V.M. Chernyshev, and V.P. Ananikov drafted the manuscript with contributions from all authors. All authors participated in the discussions.

Peer review

Peer review information

Communications Chemistry thanks Nupur Goswami and the other anonymous reviewers for their contribution to the peer review of this work. Peer reviewer reports are available.

Data availability

The data supporting the findings of this study are included in the paper or the Supplementary Information and are also available upon request from the corresponding author. The Supplementary Information contains full details of the experimental procedure, optimization of conditions, kinetic plots, characterization of compounds, along and supplementary references, copies of ¹H NMR and ¹³C NMR spectra, TEM images, EDX spectra. All data are available from the corresponding author upon request.

Competing interests

The authors declare no competing interests.

Supplementary information

The online version contains supplementary material available at 10.1038/s42004-025-01653-5.