Electroreductive Nickel-Catalyzed Intramolecular Hydrosilylation of Alkenes

Abstract

A highly efficient and scalable method for synthesizing novel silylated heterocycles with potential applications in medicinal chemistry, fragrances, and materials is reported. The process involves NiBr₂bipy-catalyzed electrochemical intramolecular hydrosilylation of alkenes, converting a variety of arylsilanes with high regioselectivity. The Si–H group exclusively adds to the carbon–carbon double bond via a 5-exo-trig cyclization. The reaction employs a sacrificial anode process, requiring only a “catalytic” amount of electricity and can therefore be readily scaled up for gram-scale synthesis. Mechanistic investigations, such as cyclic voltammetry, corroborate the proposed mechanism.

Introduction

Hydrosilylation of π-bonds is one of the most industrially valuable transition metal-catalyzed reactions, offering a highly atom-economic organic transformation. It provides the raw materials for the production of silicon-based materials. , Silicon-containing molecules, particularly silylated heterocycles, are gaining increasing attention in medicinal chemistry , and fragrances. As functionalized organosilanes are important building blocks, the development of new C–Si bond-forming reactions has been extensively researched. Since Sommer’s pioneering work in 1947, which involved trichlorosilane with 1-octene in the presence of a peroxide radical initiator, significant efforts have led to the development of efficient transition metal-catalyzed hydrosilylation reactions. These reactions utilize metals such as platinum, palladium, rhodium, and ruthenium. More recently, abundant non-noble metals like iron, cobalt, and nickel have also been employed as effective catalysts. Additionally, recent advances have shown that Lewis acids, particularly electron-deficient boranes, can catalyze this metal-free transformation. Inspired by classic radical hydrosilylation of alkenes, visible-light-induced catalytic approaches have also been developed. One critical challenge in these hydrosilylation reactions is regioselectivity, which can result in either Markovnikov or anti-Markovnikov additions. This selectivity is highly substrate-dependent and can be controlled by reaction conditions or functional groups. While hydrosilylations are commonly used in their intermolecular version, intramolecular reactions remain less explored.

In recent years, organic electrosynthesis has experienced a remarkable renaissance, marked by significant advances in homogeneous electrocatalysis. This revival has triggered considerable interest in the electrochemical synthesis of organic silanes. In addition to numerous reports on the electroreductive activation of chlorosilanes, the hydrosilylation of acrylonitrile or cyclohexene has been achieved through an anodic oxidation step involving hydrosilane. Following this pioneering study, He and coworkers recently reported an electrochemical radical silylation of activated olefins. Since the direct anodic oxidation of hydrosilane is challenging due to its high redox potential, the silyl radical was generated through Hydrogen Atom Transfer (HAT) using N-hydroxyphthalimide as a redox mediator.

Despite the significance of hydrosilylation reactions, the synthesis of silacycles through the intramolecular hydrosilylation of alkenes or alkynes remains relatively unexplored. Although some examples have been reported using Speier’s catalyst (Scheme a), rhodium , or ruthenium for the activation of silylated ethers (Scheme b), and arylsilanes (Scheme c), , the field remains largely underdeveloped. Our group has developed various synthetic methodologies to access new families of silylated and germylated binuclear heterocycles. More recently, we have focused on activating the Si–H or Ge–H bond to form these heterocycles. In particular, we reported a nickel-catalyzed electroreductive intramolecular hydrosilylation of internal alkynes with high regio- and stereoselectivities. Building on these initial results, we turned our attention to the intramolecular hydrosilylation of alkenes. We report here the outcomes of this electroreductive hydrosilylation catalyzed by a nickel-2,2’-bipyridine complex, using the sacrificial anode process. This method yields silylated heterocycles with high regioselectivity control (Scheme d).

1. Intramolecular Silylation of Alkenes.

Results and Discussion

Based on our initial results with alkynes using electroreductive nickel catalysis, we first employed the NiBr₂bipy complex as a catalyst, in DMF at 80 °C, for our model reaction, combined with the sacrificial anode process. The optimization was conducted using aryl silane 2a, which is easily obtained in two steps from commercially available 2-iodobenzyl bromide. First, a nucleophilic substitution between 2-iodobenzyl bromide and an in situ-generated cuprate yields 1-allyl-2-iodobenzene 1a in 79% yield. The cuprate is prepared by adding a solution of vinyl magnesium bromide in the presence of CuI complexed with 2,2’-bipyridine. A subsequent halogen–metal exchange reaction, followed by substitution with diphenylchlorosilane, yielded precursor 2a in 71% yield (Scheme ).

2. Synthesis of Arylsilanes.

Initial tests using a nickel plate as the anode and nickel foam as the cathode yielded promising results, showing partial consumption of the starting aryl silane 2a (Table , entry 1). Decreasing the temperature to 50 °C allowed for complete conversion, and benzosilole 3a was isolated in 30% yield (Table , entry 2). At room temperature, 3a was isolated with a very good yield of 95% (Table , entry 3). Finally, as demonstrated with alkynes, if the catalytic cycle is based on Ni(0), a small amount of electrons can be used. We ran the electrolysis until 2 F/mol of Ni­(II) was passed (13 min at 25 mA, 20 C), and the reaction was stirred 2 h until complete consumption of the aryl silane. Under these conditions, we achieved total conversion, and benzosilole 3a was isolated in 98% yield (Table , entry 4).

1. Optimization of the Intramolecular Electrochemical Hydrosilylation of 2a .

entry	charge (F/mol)	temperature (°C)	conversion (isolated yield)
1	2	80	67%
2	2	50	100% (30%)
3	2	rt	100% (95%)
4	0.2	rt	100% (98%)

^aReaction conditions, unless otherwise specified: 2a (1 mmol), NiBr₂bipy (0.10 mmol), NBu₄BF₄ (0.10 mmol), DMF (5 mL). The electrolysis was run under argon at a constant current intensity of 25 mA.

^bIsolated yields are based on the initial aryl silane 2a.

^cThe reaction was stirred for 2 h after the end of the electrolysis (13 min).

In all cases, complete regioselectivity was achieved in favor of the 5-exo-trig cyclized product 3a. Given this excellent result, we did not attempt further optimization of the electrochemical conditions. The final optimized conditions are as follows: a nickel plate as the anode and nickel foam as the cathode, in DMF at room temperature. The electrolysis is stopped only after 13 min at a constant current of 25 mA (20 C), and the reaction mixture is then stirred for approximately 2 h before quenching.

With the optimized conditions established, we decided to explore the scope of the reaction. We first evaluated the influence of substituents on the double bond. Various alkenes 2b–e were prepared following the protocol in Scheme , using different vinyl cuprates and then applied to electrolysis. However, for all the aryl silanes 2b–e used, we generally did not obtain the cyclized product, even when the electrolysis was run at higher temperatures or for 2 F/mol of silane (200 C) (Table , entries 1–6). The major product remained as the starting material. It appears that substitution on the double bond inhibits cyclization, likely due to steric hindrance. In the case of aryl silane 2c, only isomerization of the double bond was observed, with the Z/E ratio changing from 1:1 to 7:3 after electrolysis (Table , entry 4). Starting from the crotyl derivative 2e, we obtained 20% of the desired benzosilole 3e, along with a significant amount of starting material 2e (Table , entry 7). Extending the electrolysis time to 2 h did not increase cyclization but resulted in substantial degradation.

2. Influence of the Substitution of the Double Bond in the Intramolecular Electrochemical Hydrosilylation .

^aReaction conditions, unless otherwise specified: 2a (1 mmol), NiBr₂bipy (0.10 mmol), NBu₄BF₄ (0.10 mmol), DMF (5 mL). The electrolysis was run at room temperature, under argon, at a constant current intensity of 25 mA for 13 min (20 C).

^bIsolated yields, based on initial aryl silane 2.

^cThe starting silane 2 was recovered at the end of the reaction.

^dThe electrolysis was run at a constant current intensity of 25 mA for 2 h 15 min (200 C).

Given these poor results, it seems clear that the substitution of the double bond prevents cyclization from occurring. We therefore decided to modulate the link between the double bond and the aryl fragment and thus prepared vinyl ether 1f in order to introduce the silyl moiety using the Turbo-Grignard reagent, which yielded silane 2f in 72% yield (Scheme ).

3. Synthesis and Reactivity of Arylsilane 2f .

Compound 2f was then subjected to the optimized conditions, and this time the cyclization was complete, ultimately affording dihydrobenzooxasilole 3f with an excellent yield of 99%. In order to evaluate the scope of the process, we prepared various substituted arylsilanes 2g–q bearing an allyl or vinyl ether group. The aryl allyl iodides 1g–k were prepared in two steps from benzylic alcohols, which were converted to benzylic bromides with PBr₃, before reacting with vinyl cuprate under the conditions shown in Scheme . As for the vinyl ether derivatives, they were obtained in three steps from substituted phenols after electrophilic iodination with NIS. Once all the compounds were obtained, we evaluated the scope of the intramolecular nickel-catalyzed hydrosilylation reaction (Table ).

3. Influence of the Substitution of the Alkyl Moiety in the Intramolecular Electrochemical Hydrosilylation .

^aReaction conditions, unless otherwise specified: 2 (1 mmol), NiBr₂bipy (0.10 mmol), NBu₄BF₄ (0.10 mmol), DMF (5 mL). The electrolysis was run at room temperature, under argon, at a constant current intensity of 25 mA for 13 min (20 C).

^b92% yield based on starting material recovery.

^c96% yield based on starting material recovery.

^dThe electrolysis was run under argon at a constant current intensity of 25 mA for 2 h and 15 min (200 C).

In all cases, the reaction process follows a 5-exo-trig cyclization step, and a single regioisomer is obtained with excellent chemical yields. The reaction proved effective with electron-donating substituents such as methyl (3g), methoxy (3i) or t-butyl (3o), with nearly quantitative yields in each case. The reaction also tolerates electron-withdrawing groups, with a quantitative yield of product 3k bearing a trifluoromethyl moiety. Halogens are also supported, with a 96% yield for product 3l bearing a fluorine atom, and 93% yields for products 3j and 3m, both bearing a chlorine atom. Phenyl (3n) and naphthyl (3h) substituents also gave quantitative yields. We finally decided to modulate the substituent carried by the silicon. When a phenyl group was replaced by a methyl group, we observed a slight decrease in chemical yields due to incomplete conversion. However, compounds 3p and 3q were obtained at 92 and 94%, respectively, relative to the recovered starting material. Compound 3p was isolated at 84% yield in the form of two diastereoisomers in a 1:1 ratio. 3q, meanwhile, was isolated with a yield of 69%, and a moderate diastereoselectivity of 2:1. The NOESY experiment confirms that the main diastereoisomer has both methyl groups on the same side of the ring. In the case of the dimethylsilyl group 2r, we obtained only a fairly low yield of 31% of the cyclized product 3r under optimized conditions, with a large amount of starting material. However, we found that the conversion (and yield) could be increased by applying a charge of 2 F/mol instead of 0.2 F/mol, performing electrolysis throughout the cyclization process. 3r was isolated with a moderate yield of 52%, which can be explained by its high volatility, leading to losses during the purification step. To complete this work, we decided to prepare arylsilane 2s in order to evaluate the possibility of obtaining a 6-membered ring. 2s is obtained in two steps from 2-iodo-N-methylaniline (see Supporting Information). Again, a charge of 2 F/mol was necessary to completely consume the silane, and compound 3s was isolated at 60% as a single isomer after a 6-exo-trig cyclization step.

To better understand the mechanism involved in this nickel-catalyzed intramolecular hydrosilylation reaction, electrochemical studies were conducted by using cyclic voltammetry. It is particularly important to consider the small number of electrons, which suggests a Ni(0)-based catalytic cycle. The black curve in Figure corresponds to the electrochemical behavior of NiBr₂bipy alone. Reduction occurs at −1.2 V/SCE, and during the reverse scan, the reoxidation wave of Ni(0) can be observed. The addition of arylsilane (red curve, Figure ) does not alter the reduction curve of Ni­(II), but during the reverse scan, the reoxidation of Ni(0) disappears, which is characteristic of the oxidative addition of electrogenerated Ni(0) in the Si–H bond.

1.

Cyclic voltammograms obtained in DMF + 0.1 M NBu₄BF₄ at a gold disc microelectrode (0.25 mm diameter), with a scan rate of v = 0.2 V·s^–1 at r.t. for (a) 10^–2 M NiBr₂bipy (black curve) and (b) 10^–2 M NiBr₂bipy with 5 molar equiv. of ArSiH 2i (red curve).

We conducted control experiments. When the reaction was carried out without NiBr₂bipy, with a charge of 0.2 F/mol, no cyclized product was detected (Scheme a). Furthermore, when the reaction was run in the presence of NiBr₂bipy but without electrolysis, the starting material was completely recovered without any trace of the cyclized product (Scheme b). These results are consistent with the need to generate Ni(0) for the cyclization to occur. We then performed the reaction in DMF at room temperature, using 10 mol % of Ni­(COD)₂ instead of NiBr₂bipy, without electricity or a reducing agent. Within 1 h, complete conversion was achieved, yielding 3f with a 99% yield (Scheme c). These results showed that the Ni(0) generated in situ was the active species in the catalytic cycle and that it was regenerated without the need for another reducing agent. In order to verify that the reaction does not proceed via a radical pathway, we studied the formation of the cyclized product by adding 1 equiv of TEMPO after 15 min (immediately after the end of the Ni­(II) salt electrolysis). After 2 h of reaction, a 98% yield of the cyclic product 3f was obtained, thus demonstrating that no radical intermediate was involved (Scheme d).

4. Control Experiments.

We then extended our study to complementary experiments by preparing a deuterated Si–D precursor. 2f-D was obtained from the iodinated precursor 1f by metal–halogen exchange with the Turbo-Grignard reagent, followed by substitution with Ph₂Si­(OMe)₂. The second methoxy group was then reduced with LiAlD₄. The 2f-D was finally obtained in an overall yield of 56% (Scheme ). The cyclization of 2f-D proceeds efficiently, even though a lower yield of 68% is obtained for 3f-D compared to 99% for its hydrogenated version. Deuterium incorporation was observed only in the methyl group, with an incorporation rate of 100%.

5. Electrochemical Cyclization of Deuterated Arylsilane 2f-D .

Finally, intermolecular deuterium/hydrogen exchange is observed when electrolysis is carried out in a cell containing an equimolar mixture of products 2f-D and 2a, leading to complete conversion in both cyclized products with 50% deuterium incorporation in each (Scheme ). This can be attributed to an H/D exchange occurring after the oxidative addition, just prior to the cyclization step.

6. Intermolecular Exchange of Deuterium during the Electrochemical Cyclization of 2f-D and 2a .

As previously discussed, the mechanism proceeds through a Ni(0)-based catalytic cycle, as supported by the requirement for only a small number of electrons. Based on these experiments, and considering the “catalytic” amount of electrons, we therefore propose the mechanism depicted in Scheme . In line with results reported for alkyne hydrosilylation, we suggest that the process involves a silylnickelacycle intermediate, consistent with a classical Chalk–Harrod mechanism.

7. Proposed Mechanism for the Electrochemical Nickel-Catalyzed Hydrosilylation of 2 .

The catalytic cycle begins with the two-electron reduction of NiBr₂bipy at the cathode, generating the active Ni(0)­bipy catalyst. Subsequently, the oxidative addition of Ni(0) into the Si–H bond affords intermediate A. A migratory insertion then produces silanickelacycle B. Finally, reductive elimination releases the cyclic product 3 while regenerating Ni(0), thereby accounting for the substoichiometric amount of electrons required for the reaction.

Conclusions

In summary, we have developed an efficient new approach to access silylated heterocycles via intramolecular hydrosilylation of alkenes using an electrochemically activated Ni­(II) complex. This methodology enables the formation of five-membered rings with high regioselectivity, consistently affording the 5-exo-trig product in isolated yields ranging from 52 to 100%. Extension of the strategy to six-membered rings was also achieved, albeit with lower chemical yields. Overall, this procedure is attractive due to its high efficiency, operational simplicity, and absence of noble metals or stoichiometric reducing agents.

Experimental Section

General Procedure for the Nickel-Catalyzed Intramolecular Hydrosilylation of Aryl Silanes

Electrolysis was performed in a one-compartment cell equipped with a nickel foam cathode (approximately 2 cm²) and a nickel rod anode. NBu₄BF₄ (40 mg, 0.1 mmol, 0.1 equiv) and silane 2 (1.0 mmol, 1 equiv) were dissolved in DMF (5 mL) under an argon atmosphere, followed by the addition of NiBr₂bipy (37 mg, 0.1 mmol, 0.1 equiv). The electrolysis was conducted at room temperature under argon with a constant current of 25 mA at 2 F/mol with respect to the nickel precatalyst (20 C, 13 min). After electrolysis, the reaction mixture was stirred for an additional 2 h at room temperature until complete consumption of the silylated aromatic compound 2. The mixture was then hydrolyzed with H₂O and extracted with EtOAc. The combined organic layers were washed with brine, dried over MgSO₄, concentrated under reduced pressure, and purified by column chromatography using an appropriate solvent.

The Supporting Information is available free of charge at https://pubs.acs.org/doi/10.1021/acsomega.5c13201.

• Synthetic procedures, characterization data, spectroscopic data, and copies of NMR spectra (PDF)

The manuscript was written through the contributions from all authors. All authors have given approval to the final version of the manuscript.

The authors declare no competing financial interest.