Introduction

Since 1887 and the first documented allene synthesis by Burton and Pechmann1, this peculiar class of cumulene has fascinated the organic practitioners. Indeed, the allene motif is found in a plethora of natural products2, which stimulated the development of synthetic approaches. In addition, over the years allenes found applications in medicinal chemistry programs and in material chemistry3,4.

Besides, allenes proved to be excellent synthetic building blocks to forge more complex structures5,6,7,8,9, upon further synthetic manipulations10,11,12,13,14. Among them, allenyl silanes are an important class of versatile synthetic intermediates15, which could be for instance used as propargylic anion surrogate, as pioneered by Danheiser16,17. These compounds are linchpin in organic synthesis and were used in strategic transformations in natural product synthesis, for instance18. Therefore, it is not surprising that significant efforts were devoted to elaborate practical reaction manifold to forge these compounds. The synthetic approaches mostly rely on the metal catalyzed silylation of propargylic derivatives, using Cu, Ni, Rh or Pd catalysts (Fig. 1, Eq. (1))19,20,21,22,23,24,25,26,27,28,29,30,31,32,33,34. Complementary, the electrophilc silylation of metaled allenes, generated from either carbometallation or direct metalation, offers a straightforward access to this scaffold (Fig. 1, Eq. (2))35,36,37. As another approach, the Wittig reaction on silylated ketene was also depicted (Fig. 1, Eq. (3))38. Finally, the formation of the allene moiety was reported starting from silylated substrates (Fig. 1, Eq. (4))39,40,41,42. Importantly, most of these methods require either the use of transition metal catalysts (Cu, Ni, Rh, Pd, Ir), stoichiometric amount of strong bases or harsh reaction conditions. Hence, the development of mild and more sustainable reaction conditions to forge this class of compounds is highly sought after.

Fig. 1: State of the art and present work.
Fig. 1: State of the art and present work.
Full size image

Eq. (1). Transition metals catalyzed synthesis of allenyl silanes. Eq. (2). Metallated allenes as precursors of allenyl silanes. Eq. (3). Silylated ketene as precursor of allenyl silane (Wittig approach). Eq. (4). Silylated alkynes of propargyl derivatives as precursor of allenyl silanes.

Over the last decades, organic electrosynthesis has known a significant resurgence43,44,45,46,47,48,49,50,51,52,53,54,55. This renewal is mainly due to the impact of organic electrosynthesis on the development of sustainable process. Indeed, synthetic organic electrochemistry contributes to the reduction of waste generation, since it can replace hazardous oxidizing or reducing reagents by simple electrons. On the other hand, electrosynthesis usually provides high chemoselectivity with mild reaction conditions and could allow access to new reactivity56,57,58. Moreover, if electricity arises from renewable source, electrosynthesis fulfills the criteria of sustainability. Hence, organic electrosynthesis is a pivotal research area to offer alternative sustainable reaction manifold by decreasing the environmental footprint, a contemporary concern of the community for the public welfare.

Quite surprisingly, few electrochemical approaches to synthesize allenes were reported to date59,60. In addition, to our knowledge, the electrochemical synthesis of allenyl silanes remained scarce and mostly relied on the use of transition metal catalysts. In 2023, Li reported an electrochemical approach using a Cobalt catalyst and a sacrificial anode to forge allenyl silanes from propargylic derivatives61. Concomitantly, Yue and Rueping described an electroreductive coupling to forge allenyl silanes from 1,3-enynes using a Nickel catalyst62.

Thus, as part of our ongoing research program dedicated to organic electrosynthesis63,64,65,66,67,68,69, we sought to elaborate a versatile electrochemical method to synthesize allenyl silanes. Indeed, based on our experience on the oxidative electrogeneration of silyl radical70,71, and other reports72,73,74,75,76,77,78,79,80,81, we surmised that a plausible addition of the latter on a propargylic derivative would lead to the formation of the desired allenyl silanes82. This approach, using a borosilane, avoids the use of electromediators, usually required for the formation of silyl radical from silanes, or harsh reductive conditions71. Indeed, the formation of the silyl radical using silane usually needs the addition of a HAT catalyst, while the use of chlorosilane required strong reductive conditions72,73,74,75,76,77,78,79,80,81. Our approach focuses on the oxidation of a boronate intermediate, which has a low oxidation potential, proving a smooth formation of the silyl radical70.

In this work, we develop the electrochemical synthesis of allenyl silanes under oxidative conditions, using the Suginome reagent. This approach is also extended to the synthesis of allenyl boronated using a similar strategy.

Results and discussion

Optimization of the allenyl silane synthesis

At the outset of our investigations, we studied the reaction of the propargyl acetate derivative I with the Suginome reagent (PhMe2Si–BPin). After a meticulous set of optimizations83, we were pleased to delineate the optimal reaction conditions. Using 2 equivalents of the Suginome reagent, nBu4NCl as the electrolyte (0.1 M), stainless steel electrodes at both the cathode and the anode in a 9:1 mixture of CH3CN and CH3OH with a constant current of 30 mA and a charge of 3 F.mol-1, the allenyl silane 1 was obtained in a 74% NMR yield and a pleasant 68% isolated yield (Table 1, entry 1). From this optimization, we found that the 9:1 mixture of CH3CN and CH3OH was crucial for the outcome of the reaction (entries 2 and 3). Then, the replacement of the SST anode led to no reaction (entry 4), while the use of a platinum cathode furnished a lower yield (entry 5). A current of 30 mA was the optimum (entry 6). Our attempt to decrease the amount of electrolyte was deleterious for the reaction outcome (entry 7). Interestingly, the reaction conditions were compatible with methyl and tert-butyl carbonate propargyl derivatives, in place of the acetate, albeit with slightly lower yields (entries 8 and 9).

Table 1 Optimization of the reaction conditions

Scope and limitations of the reaction

Then, with the settled optimized reaction conditions we decided to explore the scope and limitations of the method (Fig. 2).

Fig. 2: Electrochemical synthesis of allenyl silanes – scope and limitations.
Fig. 2: Electrochemical synthesis of allenyl silanes – scope and limitations.
Full size image

Reaction conditions: propargyl acetate (0.3 mmol), PhMe2Si–BPin (0.6 mmol), nBu4NCl (0.4 mmol), CH3CN/CH3OH (9:1, 4 mL), 30 mA, r.t., air.a diasteroisomeric ratio determined by 1H NMR.

First, we introduced substituents on the aromatic ring of the secondary propargyl acetate. The presence of halogens was well tolerated, giving access to the corresponding di-substituted allenes 24 in very good yields. Then, the methyl substituent, whatever its position on the phenyl ring did not alter the reaction outcome and allenes 57 were obtained in very good yields. The allenyl silane 8 having a methoxy group on the aromatic ring was obtained in a moderate 50% yield, while the benzodioxolane derivative 9 was obtained in a low 21% yield. The heteroaromatic ring (i.e. benzothiophene) was tolerated and the allene 10 was isolated in 45% yield. Then, the reactions with alkyl derivatives were performed. Pleasingly, allenyl silanes 1113 were isolated in moderate to very good yields. The presence of an olefin was also tolerated affording the allene 14 in 45% yield, showcasing a faster addition on the alkyne than on the alkene. Cycloalkyl substituents did not hamper the reaction and allenyl silanes 15 and 16 were obtained in 56% and 53%. In addition, the reaction was tolerant to the presence of a thioether moiety, as observed with allene 17, which was isolated in 74% yield. Then, we tested our reaction conditions on tertiary propargyl acetates to access trisubstituted allenyl silanes. Pleasingly, the method allowed the formation of the allenes 18 and 19, bearing aromatic or alkyl substituent in very good yields (82% and 78%, respectively). The presence of a chlorine atom, a TBS-protected alcohol or an ester group was tolerated, as showcased with allenes 20, 21 and 22. The cyclic derivatives 23, 24 and 25 were obtained in moderate to good yields. This straightforward protocol was then used to synthesize complex allenyl silanes. The Mestranol derived allenyl silane 26 was isolated in an excellent 95% yield. The procedure was then applied to propargyl acetates derived from Dehydroepiandrosterone and Pregnenolone. The corresponding allenyl silanes 27 and 28 were formed in very good yields, 69% and 79% respectively, the allene 28 obtained as a 9:1 mixture of diasteroisomer83. Interestingly, the reaction can be scale up. The allenyl silanes 11 and 19 were obtained in 46% and 63% isolated yield, on a 1.5 mmol and 1.2 mmol scale respectively. Overall, the reaction proceeded well and highlighted a good functional group tolerance, although some substrates remained reluctant in our hand (Fig. 2). Indeed, the presence of nitrile motif was not tolerated, resulting in the decomposition of the starting propargylic acetate. The reaction was sensitive to the steric hindrance, as shown with the fluorenone derivatives. Unfortunately, the substitution of the alkyne was not tolerated, hampering the access to tetrasubstituted allenyl silanes. Finally, primary propargyl acetates were also reluctant under our reaction conditions.

Then, we sought to extend our strategy to the synthesis of the synthetically useful allenyl boranes84,85,86,87,88,89,90,91, using B2Pin2 as the boron source. Indeed, inspired by our initial foray in the electrooxidative generation of boryl radical66,67,69, we hypothesized that the boryl radical92,93,94,95,96,97,98,99,100,101,102,103,104,105,106,107,108,109,110,111 would have a similar reactivity and would lead to the formation of the allenyl boronate from an appropriate propargylic derivative. This approach would complement the existing state of the art and would offer a metal free and sustainable approach toward this versatile class of compound. Indeed, the seminal synthetic access to these compounds relied on the electrophilic borylation of allenyl metals (MgX, Li…). The metal catalyzed borylation reaction on propargylic derivatives (Cu, Pd, Au, Fe, Ag)112,113,114,115,116,117,118,119,120, along with the hydrometallation/borylation sequence on enynes121,122,123,124,125 remain the most popular approaches. Complementary, Studer and Friese recently reported the unique radical approach to these compounds, taking advantage of the redox properties of oxalic ester derived from tertiary propargylic alcohols126. Hence, we tested our electrochemical approach with the reaction of B2Pin2 with the propargyl carbonate II.

Optimization of the allenyl boronated synthesis

After an extensive set of optimizations83, we devised the optimal reaction conditions (Table 2, entry 1). Using stainless steel electrodes (cathode and anode), Et4NBF4 (0.05 M) as the electrolyte, NaOCH3 as the additive (1 equiv) in CH3OH, a constant current of 25 mA and rapid alternating polarity (rAP, 150 ms) at room temperature, the allenyl boronic ester 29 was obtained in a decent 56% NMR yield and a 46% isolated yield. The use of rapid alternating polarity was crucial to avoid the electrodes passivation and degradation of the product.

Table 2 Optimization of the reaction conditions for the synthesis of 29

During the optimization study, we found that NaOCH3 was the best additive, since the other bases (NaOH or Na2CO3) did not improve the yield (entry 2). The use of stainless-steel cathode and anode was the optimal setup, as other electrodes gave either poor yield or no reaction (entries 3 and 4). The screening of solvent highlights methanol as the best solvent, since other combinations led to lower yields (entries 6 and 7). Finally, the pulse duration of rapid alternating polarity was investigated and an optimal rAP of 150 ms was found (entries 8 and 9). Then, with these optimized reaction conditions in hand we explored the scope and the limitations of this catalyst-free electrochemical synthesis of allenyl boronic esters (Fig. 3).

Fig. 3: Electrochemical synthesis of allenyl boranes.
Fig. 3: Electrochemical synthesis of allenyl boranes.
Full size image

Reaction conditions: propargyl carbonate (0.2 mmol), B2Pin2 (0.4 mmol), nBu4NBF4 (0.02 mmol), CH3ONa (0.2 mmol), CH3OH (4 mL), 25 mA, rAP (150 ms) r.t., air.a determined by 1H NMR using CH3NO2 as the internal standard.

Scope of the reaction with B2Pin2

First, other diborane were tested. Pleasingly, the PinB–Baam was tested and the corresponding allenyl borane 30 was isolated in a modest 20%. The PinB–Bdan was also tested, however the reaction led a complex reaction mixture without trace of the desired allene. Then, the reaction conditions were then applied to other propargylic derivatives with longer alkyl chain on the alkyne. The allenyl boranes 31 and 32 were isolated in moderate yields (28% and 32%), despite a good NMR yield in the case of 32, mainly due to tedious purification of the products due to their moderate stability over silica gel. The cyclohexyl derivative 33 was obtained in a modest 48% isolated yield and a 69% NMR yield. Interestingly, the allenyl boranes 34 was isolated in a pleasant 52% yield. Likewise, the boronate 35 was obtained in a 50% isolated yield. The reaction with dimethylated propargyl derivatives led to the formation of allenes 36, 37, 38 and 39 in low to modest yields (12% to 40%). Finally, other disubstituted allenes were synthesized with very low yields (40 and 41, 10% and 5% isolated yields). The observed low yields are mainly explained by the stability of the allenyl boranes in the reaction media. Indeed, these substrates can readily be subjected to a protodeboration event. Then, we sought to get insight into the reaction mechanism (Fig. 4).

Fig. 4: Mechanistic investigations and plausible reaction mechanism.
Fig. 4: Mechanistic investigations and plausible reaction mechanism.
Full size image

A Study of the stereoretention of the process. B Radical trapping experiment. C Cyclic voltammetry studies. D Plausible reaction mechanism—for the sake of clarity ligand on the boryl radical was omitted.

Study of the reaction mechanism

First, the stereoretentive nature of the process was studied. Thus, we conducted the reaction with the enantioenriched propargylic derivatives (S)-III and (R)-III (Fig. 4A). Under standard reaction conditions, the corresponding allene 11 was obtained in similar yields with a complete enantiospecificity in the case of (R)-III and a slight alteration of the e.r., when the (S) isomer was used (e.r. = 93:7, e.s. = 96%). Likewise, (S)-IV was used as a model substrate and the enantioenriched product 12 was obtained in a similar e.r. (e.s. = 98%) These results demonstrate the enantiospecificity of the reaction and a possible access to enantioenriched allenes from enantioenriched chiral propargylic acetates. To confirm the presence of the silyl radical in the reaction mixture, the reaction of the model substrate in the presence of 1,1-diphenylethene was conducted under standard conditions (Fig. 4B). The HRMS analysis of the crude reaction mixture allowed the detection of the silyl radical addition product on the 1,1-diphenylethene, confirming the involvement of the silyl radical83.

Then, we conducted cyclic voltammetry (CV) measurements to support a plausible reaction mechanism (Fig. 4C). First, the CV measurements of the model proparylic acetate I83, and carbonate II did not showcase an oxidation nor a reduction event under our conditions. These results preclude an initial reduction of the propagylic derivatives under our conditions (Cheng, Li and co-workers (reference 59) described a possible initial reduction of the propargylic acetate (E°1/2 = –2.24 V to –2.50 V vs Ag/Ag+) in DMF, which has a larger electroactivity field). The measurement of a mixture of B2Pin2 and NaOMe, to mimic the cathodic formation of methoxide from methanol, revealed a single oxidation event (blue curve, Ep = +1.26 V vs SCE). Thus, a possible oxidation of an in situ formed borate species from B2Pin2 and methoxide, which eventually would lead to the formation of a boryl radical, since this oxidation appears irreversible. Likewise, and in agreement with previous reports70,71,127, a plausible oxidation of a borate formed from the Suginome reagent could lead to the formation of the corresponding silyl radical (despite all our effort, we have not been able to collect the CV measurement of the Suginome reagent (Me2PhSi–BPin), using NaOMe, mainly due to a fast decomposition of the borosilane). Finally, the CV measurement in the presence of all reaction partners highlights a similar profile as the one observed with B2Pin2 and NaOMe (Fig. 2C, green curve).

Hence, with all these data in hand and previous reports69,70,111, we suggested the following mechanism (Fig. 4D). First, B2Pin2 reacted with methoxide, generated from the reduction of the solvent (i.e MeOH) at the cathode to form the corresponding borate A. The latter is then oxidized to generate B, which collapse into the boryl radical C (the boryl radical is probably stabilized by the solvent or a methoxide anion to form a stable 7 electrons boryl radical. For clarity, the solvent or the Lewis base was omitted from Fig. 4C). This radical was reacted with the propargylic derivative to form a transient vinyl radical D. This radical might then react through two distinct pathways to forge the desired allene. First, the β-scission of the C–O bond could form the allene and the corresponding methoxycarbonyloxy radical (MeOCOO•) (path a). This radical, which highlights a very slow decarboxylation rate (3.8 × 103 M s-1)128,129, can be reduced at the cathode to form the corresponding anion. This anion might then quickly extrude CO2 to form methoxide130. Even though this class of C–O fragmentation has been scarcely reported, some examples might support this pathway131,132,133,134,135. Alternatively, a radical ionic pathway, similar to the well-known fragmentation of β-(phosphatoxy)alkyl radical136, might lead to the formation of a radical cation E along with methoxide and CO2 (generated from MeOCO2-) (path b). This cation E could be reduced at the cathode to deliver the targeted allene. Although, the mechanistic elucidation of this final event is yet impossible in our hand, we presume that the β-scission of the C–O bond might be the favored pathway. Indeed, the formation of the methoxycarbonyloxy radical MeOCOO• radical, resulting from the fragmentation, is prompt to be reduced and then undergo a fast decarboxylation136,137. Moreover, the stereoretentive nature of the process (see Fig. 4A), supports a concerted mechanism, without the involvement of a radical cation prompt to racemize. Note that an anionic pathway, resulting from the direct borylation of the substrates has been ruled out138,139. Indeed, control experiments without current in the presence of B2Pin2 and NaOMe, led to no reaction83. In the case of PhMe2Si–BPin a complete decomposition of the reagent was observed without trace of the allenyl silane.

Finally, to showcase the synthetic utility of the products some transformations were carried out (Fig. 5).

Fig. 5: Synthetic utility of the products.
Fig. 5: Synthetic utility of the products.
Full size image

aZ:E ratio determined by 1H NMR. b Diastereoisomeric ratio determined by 1H NMR.

Synthetic utility of the products

First, the regioselective hydroboration catalyzed by copper was attempted on allenyl silane 11. Pleasingly, the vinyl boronate 42 was obtained in 40% yield with a 2:1 Z:E selectivity. The allenyl silane 19 was then used to access the alkenyl silane 43 with a propargylic quaternary carbon center. Then, the protodesilylation of 10 and 18 was achieved in moderate yields, 40% and 49%, respectively, mainly due to the volatility of the final allenes. The synthetic utility of allenyl boranes was also demonstrated. First, the Suzuki cross coupling of 33 with iodotoluene provided the tetrasubstituted allene 46 in a 45% yield. Then, the hydroxymethylation of 33 gave the primary alcohol 47 in an excellent 97% yield. Finally, the oxidation of 33 allows the formation of the trisubstituted alkene 48 in a 93% yield and a 1:1.24 Z:E ratio.

In summary, we reported herein the mild and catalyst-free electrochemical synthesis of allenyl silanes and allenyl boronates. The formation of allenyl silanes proceeded smoothly with a good functional group tolerance. The reaction conditions were applied to the synthesis of a large panel of allenyl silanes, which were isolated in good yields (28 examples, 45–82%). This practical method allows the formation of di- and trisubstituted allenyl silanes and avoids the use of sensitive reagents or expensive transition metal catalyst. Under modified reaction conditions, the electrochemical synthesis of allenyl boronic esters was developed. The applicability of the method was illustrated through the synthesis of 13 examples, which were isolated in low to moderate yields (5–57%). Finally, a mechanistic analysis of the reaction supported a plausible addition of a silyl or boryl radical, generated from an oxidation of a transient borate, on the propargylic precursor. These original approaches toward the synthesis of allenyl silanes and boranes, broaden the current arsenal of the organic chemist. We believe that these novel manifolds will inspire further developments toward the sustainable synthesis of silicon- and boron-containing building blocks.

Methods

General procedure for the synthesis of allenyl silanes

PhMe2Si–BPin (0.6 mmol, 157.3 mg, 2.0 equiv.) and nBu4NCl (0.4 mmol, 111.2 mg, 0.1 M) were weighted under Argon in a glovebox and added in an ElectraSyn vial (5 mL) with a stir bar. The propargylic acetate (0.3 mmol, 1.0 equiv.) solubilized in anhydrous CH3CN (3.6 mL) and CH3OH (0.4 mL) was added. The ElectraSyn vial cap equipped with anode (Stainless steel, SST, 8 mm × 52 mm × 2 mm, immersion length is 25 mm) and cathode (SST) was inserted into the mixture. After pre-stirring for 5 minutes, the reaction mixture was electrolyzed under air and a constant current of 30 mA. The reaction is monitored by TLC and stopped until total conversion of the starting material substrate was observed. The reaction mixture was then concentrated under reduced pressure. The residue was diluted with ethyl acetate, filtered through a pad of silica and celite and concentrated. The crude mixture was purified by flash column chromatography on silica gel or preparative thin layer chromatography (PTLC).

General procedure for the synthesis of allenyl boronates

The ElectraSyn 5 mL vial was wrapped with parafilm to improve sealing. Then the ElectraSyn vial (5 mL) with a stir bar was charged with propargylic carbonate (0.2 mmol, 1.0 equiv.), B2Pin2 (0.4 mmol, 101.6 mg, 2.0 equiv.), tetraethylammonium tetrafluoroborate (Et4NBF4, 0.2 mmol, 43.4 mg, 0.05 M) and CH3ONa (0.2 mmol, 10.8 mg, 1.0 equiv.) was added anhydrous CH3OH (4 mL). After the resulting mixture was purged with an argon balloon for ca 10 s. An ElectraSyn vial cap equipped with anode (SST) and cathode (SST) was inserted into the mixture. The mixture was electrolyzed at room temperature under a constant current of 25 mA with rapid alternating polarity (150 ms, 3.33 Hz) until the completion of the reaction monitored by TLC. The reaction mixture was concentrated under reduced pressure. The residue was diluted with ethyl acetate, filtered through a pad of silica and celite and concentrated. CH3NO2 (1.0 equiv.) was added as an internal standard to the mixture, the NMR yield of the product was determined by 1H NMR. Then the resulting mixture was purified by flash column chromatography on silica gel (Santai Tech. Inc., 25 g), flow rate: 25 mL/min (pentane/ethyl acetate, gradient: 100:0 to 95:5).