Borane-catalyzed selective dihydrosilylation of terminal alkynes: reaction development and mechanistic insight

Abstract

Here, we describe simple B(C₆F₅)₃-catalyzed mono- and dihydrosilylation reactions of terminal alkynes by using a silane-tuned chemoselectivity strategy, affording vinylsilanes and unsymmetrical geminal bis(silanes). This strategy is applicable to the dihydrosilylation of both aliphatic and aryl terminal alkynes with different silane combinations. Gram-scale synthesis and conducting the reaction without the exclusion of air and moisture demonstrate the practicality of this methodology. The synthetic utility of the resulting products was further highlighted by the structural diversification of geminal bis(silanes) through transforming the secondary silane into other silyl groups. Comprehensive theoretical calculations combined with kinetical isotope labeling studies have shown that a prominent kinetic differentiation between the hydrosilylation of alkynes and vinylsilane is responsible for the chemoselective construction of unsymmetrical 1,1-bis(silanes).

---

Introduction

Organosilanes have a multitude of applications in a variety of fields, ranging from organic synthesis to functional materials and pharmaceutical sciences.¹ Geminal bis(silanes) are particularly attractive due to their versatile synthetic utilities, and valuable structural and conformational properties.^2,3 However, the scarcity of practical protocols for their synthesis limits their structural diversity and their further synthetic utilities. Previous strategies for the synthesis of geminal bis(silanes) are mainly based on prefunctionalized precursors, such as retro-[1,4]-Brook rearrangement of 3-silyl allyloxy-silanes,^3a,b,d insertion of benzylic carbenes into disilanes,⁴ and double C–Si coupling of geminal dibromides.⁵

Synthesis of geminal bis(silanes) through dihydrosilylation of alkynes is an ideal way due to the high step- and atom-economy. Recently, several research groups have reported dihydrosilylation reaction of alkynes through transition-metal catalysis. For example, Lu's group and Zhu's group reported the catalytic 1,1-dihydrosilylation of terminal aliphatic alkynes with Co- and Fe-catalysts, affording symmetrical 1,1-bis(silanes), respectively (Scheme 1a).^6a,7 Besides, Cui's group also reported the dihydrosilylation of internal alkynes with rare-earth ate complexes, providing access to 1,1-di- or 1,1,1-trisilylated products.⁸ More importantly, Lu and co-workers elegantly developed sequential catalysis with a combination of two Co-catalysts (CoBr₂, Xantphos and CoBr₂, OIP (OIP = oxazoline-iminopyridine)) to access unsymmetrical 1,1-bis(silanes) with aliphatic terminal alkynes (Scheme 1b).^6b Although significant advances have been achieved, currently available synthetic protocols are mainly applicable to terminal aliphatic alkynes or internal alkynes. The synthesis of 1,1-bis(silanes) with arylacetylenes is less reported. Scope investigation on Fe-catalyzed dihydrosilylation of terminal alkynes has shown that the reaction of 1-arylacetylene with hydrosilane only affords the monohydrosilylated product due to the steric effect.^7a,9

Scheme 1 Catalytic dihydrosilylation of alkynes. R¹ = alkyl; R² = aryl, alkyl or silyl group; R = alkyl or aryl.

Recently, the combination of commercially available tris(pentafluoro) borane (B(C₆F₅)₃)^10,11 and hydrosilanes (Si–H) has emerged as a highly versatile catalytic system for the hydrosilylation, silylation and selective defunctionalization of organic molecules to valuable intermediates,^12–15 providing a powerful alternative to the classical transition-metal catalysis. It should be noted that Gevorgyan's group has previously disclosed B(C₆F₃)₃-catalyzed hydrosilylation of alkenes about twenty years ago,^13c but the catalytic hydro- or dihydrosilylation of terminal alkynes with the versatile B(C₆F₅)₃/Si–H system remains unknown. We herein report the development of a metal-free strategy for the selective dihydrosilylation of aliphatic or aryl-terminal alkynes with B(C₆F₅)₃ as the catalyst (Scheme 1c). This method allows the direct construction of unsymmetrical geminal bis(silanes) with a quaternary and a secondary silyl group, in which the Si–H bonds can be readily converted to other silyl groups. Mechanistic studies reveal that tertiary silanes (like EtMe₂SiH) favor the hydrosilylation of alkynes (1^st hydrosilylation) while primary hydrosilanes (such as PhSiH₃ and BuSiH₃) favor the hydrosilylation of vinylsilane (2^nd hydrosilylation).

Results and discussion

Reaction development

Although the realization of this strategy could provide a conceptually different approach for the synthesis of 1,1-bis(silanes), several possible problems should be overcome: first, hydrosilylation of alkynes generally yields alkenyl silanes, which are difficult for further hydrosilylation owing to the steric effect.^7a,9 Second, the control of chemoselectivity in the presence of two different hydrosilanes is also problematic (Scheme 2a). Third, the rapid and irreversible 1,1-carboboration reaction between a terminal alkyne and B(C₆F₅)₃ may lead to catalyst deterioration (Scheme 2b).^16,17 As a result, despite the broad application of B(C₆F₅)₃ in hydrosilylation reactions, its capability in the dihydrosilylation of alkynes has, to our knowledge, not been demonstrated.

Scheme 2 Challenges in catalytic 1,1-dihydrosilylation of terminal alkynes with B(C₆F₅)₃.

We first examined the reaction conditions for catalytic monohydrosilylation of a terminal alkyne (Scheme 3a). At room temperature, the hydrosilylation of phenylacetylene 1a using B(C₆F₅)₃/EtMe₂SiH combination only affords Z-vinylsilane 3aa in 31% yield.¹⁸ This may probably be attributed to the occurrence of 1,1-carboboration reaction between phenylacetylene and B(C₆F₅)₃.^16,17 From a mechanistic point of view, the hydrosilylation pathway, involving a concerted three-molecular Si-centered displacement of a boron-coordinated hydride by the alkyne transition state (S_N²@Si),^13,19 requires more entropy-cost than the two-molecular transition state toward the carboboration pathway. Therefore, we hypothesized that low temperatures might facilitate the entropy-disfavored hydrosilylation pathway. We were delighted to find that the yield was significantly improved from 58% to 84% when lowering the temperature from 5 to −40 °C. Then, the effect of hydrosilylation reagent was examined in the presence of 5 mol% of B(C₆F₅)₃ at −40 °C. Other hydrosilanes, including tertiary, secondary, and primary silanes, were not as effective as EtMe₂SiH for monohydrosilylation of phenylacetylene (see Table S1† in the ESI for optimization details). It is worth noting that with secondary hydrosilane Et₂SiH₂, Z-vinyl silane was formed in 55% yield but accompanied by 25% yield of dihydrosilylated product. These results indicate that low temperature and suitable hydrosilanes are necessary to synthesize vinylsilane from terminal alkynes using B(C₆F₅)₃ as the catalyst. In addition to phenylacetylene, the reactivity of aliphatic terminal alkynes and internal alkynes was also investigated for the B(C₆F₅)₃-catalyzed monohydrosilylation reaction. As shown in Scheme 3b, this method is applicable to both terminal and internal alkynes. It should be noted that the hydrosilylation of 1-phenylpropyne exhibits excellent regioselectivity and efficiency, the corresponding Z-vinylsilane could be prepared in 89% yield (9 mmol scale) with only 0.25 mol% of B(C₆F₅)₃.

Scheme 3 Initial investigation on the monohydrosilylation of terminal alkynes. ^aPerformed at −40 °C. ^bWith 5 mol% B(C₆F₅)₃. ^cWith 10 mol% B(C₆F₅)₃. ^dIn 9 mmol scale, performed at room temperature with 0.25 mol% B(C₆F₅)₃. ND = not detected.

After the monohydrosilylation reaction conditions were established, we then explored the dihydrosilylation of alkynes (Table 1). Using EtMe₂SiH as the sole hydrosilane, only traces of dihydrosilylated product 5aaa were observed, and most of the monohydrosilylated product 3aa (76% NMR yield) was left unreacted at −20 °C in the presence of 5 mol% of B(C₆F₅)₃ (entry 1). Further increasing the catalyst loading from 5 mol% to 10 mol% did not significantly improve the yield of 5aaa (entry 2). Replacing EtMe₂SiH by an equimolar mixture of EtMe₂SiH and Et₂SiH₂, unsymmetrical geminal bis(silane) 4aab was formed in 66% yield (entry 3), but accompanied by 34% yield of symmetrical 1,1-bis(silane) 5abb (dihydrosilylated by Et₂SiH₂). When a combination of EtMe₂SiH and PhSiH₃ was used, phenylacetylene underwent two sequential hydrosilylation reactions to give the corresponding unsymmetrical 1,1-bis(silane) 4aac in 92% isolated yield (entry 4).²⁰ Reaction performed overnight at a lower temperature (−40 °C) did not decrease the yield of the desired product (entry 5).

Table 1 Optimization of the dihydrosilylation reaction of alkynes^a

Entry	Si–H combinations	3	4	5
a Reaction conditions: 1a (0.4 mmol), EtMe2SiH (2a, 0.60 mmol), Si–H (2a–2c, 0.60 mmol), B(C6F5)3 (0.04 mmol), in dichloromethane (1.0 mL). Yields were determined by NMR analysis; isolated yield in parentheses. b 5 mol% B(C6F5)3. c With 3.0 equiv. of 2a. d At −40 °C.
1^b^,^c	2a	76%	—	<5%
2^c	2a	68%	—	29%
3	2a/2b = 1 : 1	ND	66%	34%
4	2a/2c = 1 : 1	ND	>95% (92%)	ND
5^d	2a/2c = 1 : 1	ND	>95%	ND

---

Examination of substrate scope

We then evaluated a series of terminal alkynes to probe the substrate scope (Scheme 4). Substrates with electron-withdrawing (1b–1i) or electron-donating (1j–1n) substituents underwent dihydrosilylation with good efficiency, transforming the terminal alkyne into the corresponding 1,1-bis(silanes) in reaction with an equimolar mixture of EtMe₂SiH and PhSiH₃. Functional groups, including halides (F, Cl, Br, 1b–1h), trifluoromethyl (CF₃, 1i), and trimethylsilyl (TMS, 1o) were tolerated. 3-Triisopropylsilyl-O-substituted phenylacetylene was selectively dihydrosilylated to yield unsymmetrical 1,1-bis(silane) 4pac. The reaction of 1-ethynyl-4-[[tris(1-methylethyl)silyl]oxy]benzene with EtMe₂SiH and PhSiH₃ under standard conditions only affords the symmetrical geminal silane 4qcc in 27% (dihydrosilylated by PhSiH₃). This is probably due to the fact that the kinetics or reaction mechanism of the substrate with strong electron-donating substituents is different from phenylacetylenes with weak electron-donating substituents. Similarly, alkyne 1r with a cyclic thioether also did not afford the corresponding unsymmetrical bis(silane) 4rac according to the general procedure, but afforded 4rcc in 15% yield (dihydrosilylated by PhSiH₃). The unsymmetrical 1,1-bis(silane) 4rac could be obtained in 42% yield by slightly changing the reaction conditions. Aside from phenylacetylene derivatives, a β-naphthyl and a thiophene-3-yl as in 1s and 1t were also compatible with the reaction conditions, leading to products 4sac and 4tac, respectively, with moderate and slightly lower yields. However, functional groups including ether, ester, aldehyde, ketone, carboxylic acid, and amide were not tolerated because of the high reduction or defunctionalization ability of B(C₆F₅)₃/hydrosilane combinations. For example, when 1-(4-ethynylphenyl)ethanone was subjected to the mixture of B(C₆F₅)₃ and hydrosilane at −20 °C, we could only isolate the decarbonylated 1,1-bis(silane) 4uac. In addition to phenylacetylenes, terminal alkynes with aliphatic side chains were also viable (Scheme 4, bottom). Oct-1-yne 1v and 6-chlorohex-1-yne 1w could be dihydrosilylated to give the related 1,1-bis(silanes) 4vac and 4wac by simply stirring the reaction mixture at −40 °C for 5 hours and then at room temperature for an additional 24 hours. The reactivity difference between aryl and aliphatic terminal alkynes can be rationalized by the stability of different types of carbon cations involved. However, the reaction of the internal alkyne, 1-phenyl-1-propyne, only affords the monohydrosilylated compound as the major product. Different from Fe-catalyzed dihydrosilylation of terminal alkynes, in which only aliphatic terminal alkynes can be transformed into the corresponding symmetrical geminal bis(silanes) with primary silanes,^7a our approach is applicable to both aryl and alkyl-substituted terminal alkynes.

Scheme 4 Substrate scope. Isolated yields after flash chromatography. ^a20 mol% B(C₆F₅)₃ was used. ^bPhSiH₃ was added in 5 hours delay after the initial EtMe₂SiH addition. ^cDihydrosilylated by PhSiH₃. ^dStirred at −40 °C for 5 h, and then 24 h at room temperature.

Then, the reactivity observed in Table 1 intrigued us to construct geminal bis(silanes) with other hydrosilane combinations. As shown in Scheme 5, using Et₂SiH₂ as the sole silane, a symmetrical geminal bis(silane) 5abb with two tertiary silane moieties was obtained in excellent yield. Moreover, the reaction of Et₂SiH₂/PhSiH₃ and EtMe₂SiH₂/BuSiH₃ combination with phenylacetylene also worked effectively and selectively to give the unsymmetrical 1,1-bis(silane) 4abc and 4aad in 62% and 81% yield, respectively. To further identify the silane effect in the second hydrosilylation step, we subjected 3aa to the standard procedures with an equimolar mixture of Et₂SiH₂ and PhSiH₃, and detected the exclusive formation of 4aac (Scheme 5, middle; selectively hydrosilylated by PhSiH₃). These results confirmed that the primary silane is more reactive than secondary and tertiary hydrosilanes in the second hydrosilylation step (vinylsilane hydrosilylation). This trend is consistent with the size of the hydrosilanes. Although the use of EtMe₂SiH and Et₂SiH₂ combination gives a mixture of two different dihydrosilylation products (cf.Table 1, entry 3), the unsymmetrical 1,1-bis(silane) 4aab can be obtained in 81% yield by simply adding Et₂SiH₂ in 5 hours delay after the initial EtMe₂SiH addition (Scheme 5, bottom).

Scheme 5 Reactivity of other hydrosilane combinations. Isolated yields after flash chromatography. ^aA slight amount of symmetric geminal bis(silane) was detected (unsymmetric : symmetric > 10 : 1). ^bEt₂SiH₂ was added in 5 hours delay after the initial EtMe₂SiH addition.

To demonstrate the practicality of this protocol in preparative organic synthesis, a gram-scale reaction was conducted. As shown in Scheme 6a, in the presence of 10 mol% B(C₆F₅)₃, 2.7 grams of 4aac could be prepared with phenylacetylene in 90% yield. Moreover, the reaction could be conducted with all reagents handled in the open atmosphere without exclusion of air and moisture; the yield was 90% in 1 mmol scale (Scheme 6b).²¹ The two Si–H bonds in the secondary silane of 4aac can be readily converted to Si–O (6 and 7) and Si–F bonds (8) (Scheme 6c), enriching the structural diversity of the geminal bis(silanes).

Scheme 6 Reaction practicality and product transformation.

Mechanism of borane-catalyzed dihydrosilylation of terminal alkynes

Comparison of the influence of hydrosilanes on alkyne hydrosilylation and vinylsilane hydrosilylation. The experimental results collected in Scheme 7 show that for the hydrosilylation of phenylacetylene, replacement of the ethyl group in EtMe₂SiH by phenyl (PhMe₂SiH) leads to a reduced yield of vinylsilane (84% versus 46%). Hydrosilylation of 1a with Et₂SiH₂ exhibits higher reactivity than PhMeSiH₂, albeit accompanied by the formation of symmetric geminal bis(silane); both Ph₂SiH₂ and PhSiH₃ are inactive for the hydrosilylation of phenylacetylene. These results suggest that hydrosilylation of phenylacetylene is controlled by the electronic effect, and the observed reactivity order is consistent with the electron-donating ability (Me > Ph > H) of hydrosilanes as measured by Mayr et al.²² However, the reactivity of vinylsilane hydrosilylation is significantly different from alkyne hydrosilylation. Tertiary silane is less reactive than secondary and primary hydrosilane. Replacement of the two ethyl groups in Et₂SiH₂ by phenyl groups leads to little reactivity. Besides, we also performed the hydrosilylation reaction of 3aa in the presence of two secondary hydrosilanes Et₂SiH₂ (2b) and PhMeSiH₂ (2e). The reaction rate of 3aa with PhMeSiH₂ is faster than that of Et₂SiH₂ (4aab/4aae = 1 : 3.5, see Scheme S2 and Fig. S3† for details). This result confirms the importance of the steric effect in the hydrosilylation of vinylsilane. However, when performing the hydrosilylation reaction of 3aa in the presence of two primary silanes PhSiH₃ and BuSiH₃, alkylsilane BuSiH₃ shows higher reactivity than PhSiH₃; and the corresponding geminal bis(silanes) were obtained in 2.4 : 1 ratio (see Scheme S2 and Fig. S4† for details). To conclude from these control experiments, we can get the reactivity order for the hydrosilylation of vinylation: EtMe₂SiH < Et₂SiH₂ < PhMeSiH₂ < PhSiH₃ < BuSiH₃. These observations suggest that hydrosilylation of vinylsilane mainly depends on the steric effect, but electronic effects also play a role in the hydrosilylation of vinylsilane with primary hydrosilanes.

Scheme 7 Comparison of the influence of hydrosilanes on alkyne hydrosilylation and vinylsilane hydrosilylation. ^aHydrosilylation of phenylacetylene 1a: performed at −40 °C with 10 mol% B(C₆F₅)₃ in 0.4 mmol scale. ^bHydrosilylation of vinylsilane 3aa: performed at −20 °C with 5 mol% B(C₆F₅)₃ in 0.2 mmol scale. ^cDihydrosilylated product formed with Et₂SiH₂, see Table S1† for details.

DFT calculations on the full catalytic cycle. Computational studies with the M06-2X functional²³ were undertaken to gain more mechanistic insight. The solvent effect was treated with the PCM²⁴ solvation model and with dichloromethane as the solvent in structural optimization and single-point energy calculations (see ESI† for computational details).²⁵ The dihydrosilylation of phenylacetylene 1a with Me₃SiH/PhSiH₃ combination was selected as the model reaction. The complete free energy profile is displayed in Fig. 1 (see Fig. S6† for optimized structures). The sequential dihydrosilylation of alkyne proceeds through a classical Piers–Oestreich-type mechanism,^13,19 wherein substrate 1a (alkyne hydrosilylation, in brown line) or in situ formed vinylsilane 3aa′ (vinylsilane hydrosilylation, in blue line) attacks the Si center in the B(C₆F₅)₃-hydrosilane η1-adduct I or IV (viaTS_I-II or TS_3aa′-V) to form the ion-pair species II or V. The barrier for Si–H bond cleavage in adduct I with alkyne (viaTS_I-II) is computed to be 18.7 kcal mol⁻¹, and the Si–H bond cleavage in adduct IV with vinylsilane as a nucleophile (TS_3aa′-V) requires a slightly higher activation barrier of 20.3 kcal mol⁻¹. Both ion-pair species II and IV are readily reorganized to their isomers III and VI,¹⁹ in which the anion B–H is oriented towards the carbon cation center (vinyl cation or β-silylcarbenium ion). Then, hydride transfer from the borohydride to the electrophilic carbon atom in the ion-pair III (viaTS_III-3aa′, ΔG^‡ = 18.2 kcal mol⁻¹) or VI (viaTS_VI-4aa′c, ΔG^‡ = 18.2 kcal mol⁻¹) generates the trans-hydrosilylation product 3aa′ or dihydrosilylated product 4aa′c, respectively. In addition, the cis-hydrosilylation pathway involving hydride transfer from borohydride to the vinyl cation at the same side of the silyl group (leading to E-3aa′) was also considered. This process is less kinetically favorable than the trans-hydrosilylation pathway (ΔG^‡ = 24.1 versus 18.2 kcal mol⁻¹, see Fig. S7†). Along the whole dihydrosilylation pathway, nucleophilic attack of vinylsilane 3aa′ at the Si center of the B(C₆F₅)₃-PhSiH₃ adduct IV (TS_3aa′-V) with an activation barrier of 20.3 kcal mol⁻¹ is the rate-limiting step, which is consistent with the mild reaction conditions.

Fig. 1 Gibbs free energy profile (298.15 K) for the B(C₆F₅)₃-catalyzed dihydrosilylation of alkyne (in kcal mol⁻¹). Performed at the M06-2X/cc-pVTZ/(PCM, CH₂Cl₂)//M06-2X/6-311G(d,p)/(PCM, CH₂Cl₂) level of theory. r.d.s = rate-determining step.

Rationalization of the reactivity difference of hydrosilanes

Based on the reactivity difference observed in dihydrosilylation of terminal alkynes that depend on hydrosilanes, we then focused on gaining insights into the origin of chemoselectivity in this catalytic system.

Hydrosilylation of terminal alkynes (1^st hydrosilylation). To compare the reactivity of silanes on the hydrosilylation of alkynes, we computed the energy barriers of the 1^st hydrosilylation step (I → II) with representative hydrosilanes (Me₃SiH, Me₂SiH₂, and PhSiH₃). As shown in Fig. 2, the activation barrier for the secondary hydrosilane Me₂SiH₂ here is slightly higher than that with Me₃SiH (19.4 versus 18.7 kcal mol⁻¹); but both of which are lower than the barrier of the carboboration reaction between B(C₆F₅)₃ and phenylacetylene 1a (21.6 kcal mol⁻¹, see Fig. S8† in the ESI). With PhSiH₃, however, the related S_N²@Si transition state could not be located after extensive endeavors (Fig. S10†). Our further calculations show that the resulting ion-pair intermediate doesn't exist as a minimum structure on the potential energy surface (Fig. 2, red dashed line). These results suggest that both Me₃SiH and Me₂SiH₂ are suitable for the hydrosilylation of terminal alkynes, but the primary silane PhSiH₃ is not effective. NPA charge calculations reveal that the benzylic carbon of the ion-pair intermediate II′ for Me₂SiH₂ (Fig. 2b, II′: 0.358e) is more positively charged than that in II (for Me₃SiH, II: 0.302e). These computational results also support that the hydrosilylation of the terminal alkyne is probably controlled by the electronic effect, being consistent with these experimental results.

Fig. 2 (a) Comparison of the carboboration pathway and the B(C₆F₅)₃-catalyzed hydrosilylation of alkyne. Performed at the M06-2X/cc-pVTZ/(PCM, CH₂Cl₂)//M06-2X/6-311G(d,p)/(PCM, CH₂Cl₂) level of theory (energies are in kcal mol⁻¹); (b) optimized structures and NPA charges for ion-pair species II (with Me₃SiH) and II′ (with Me₂SiH₂).

To understand the origin of reactivity difference, we undertook NMR analysis on the hydrosilylation reactions of 1a with EtMe₂SiH or Et₂SiH and PhSiH₃, respectively. The crude ¹H NMR spectra clearly show the formation of the hydrosilylation or dihydrosilylation product with EtMe₂SiH or Et₂SiH₂ in the presence of 20 mol% B(C₆F₅)₃ (Fig. S1 in the ESI†). We were able to detect the characteristic peaks of B(C₆F₅)₃ (i.e., 128.1, 144.0, 160.9 ppm) from the ¹⁹F NMR spectra (Fig. S2 in the ESI†). However, only a trace amount of dihydrosilylated product was detected with PhSiH₃ as the silane source. The ¹⁹F NMR spectrum of the reaction mixture of PhSiH₃, 1a and 20 mol% of B(C₆F₅)₃ confirms the deterioration of B(C₆F₅)₃ in the case of PhSiH₃. We attribute this to the occurrence of the carboboration reaction of 1a with B(C₆F₅)₃ or other unknown oligomerization or polymerization reactions (see Scheme S1† for Computational analysis). These results are in line with our theoretical calculations that both EtMe₂SiH and Et₂SiH₂ are suitable silanes for the first hydrosilylation step.

Hydrosilylation of vinylation (2^nd hydrosilylation). The choice of silane combinations has a significant influence on the dihydrosilylation of alkynes: (1) the use of EtMe₂SiH as the sole silane mainly gave the monohydrosilylation product; (2) the use of EtMe₂SiH/PhSiH₃, or Et₂SiH₂/PhSiH₃, or EtMe₂SiH/Et₂SiH₂ combination could produce the corresponding unsymmetrical geminal bis(silanes), as shown in Table 1 and Scheme 5. Our calculations reveal that the experimentally observed reactivity is in good accordance with the computed activation barriers of the 2^nd S_N²@Si transition states (vinylsilane hydrosilylation, Fig. 3). As the size of hydrosilane increases, the Gibbs free energy barrier increases from 20.3 kcal mol⁻¹ with the primary silane PhSiH₃ to 22.5 kcal mol⁻¹ with the secondary silane Me₂SiH₂, and to 26.5 kcal mol⁻¹ with the tertiary silane Me₃SiH. In TS_3aa′a′ (vinylsilane hydrosilylation with Me₃SiH), the trimethylsilyl (TMS) group in Me₃SiH introduces large steric congestion around B(C₆F₅)₃ to destabilize the transition state.²⁶ This leads to increased distances between the reacting atoms in TS_3aa′a′ (d_Si–H = 2.12 Å and d_Si–C = 2.55 Å) compared to those in TS_3ab′b′ (d_Si–H = 2.05 Å and d_Si–C = 2.28 Å) and those in TS_3aa′-V (d_Si–H = 1.88 Å and d_Si–C = 2.26 Å). In addition, with Me₂SiH₂ as the substrate, we located another transition state (TS_{3ab′b′-iso}) (an isomer of TS_3ab′b′) with the silyl group rotated by 120°. This transition state has a higher activation barrier than TS_3ab′b′ (24.3 versus 22.5 kcal mol⁻¹), indicating that the steric effect is a key factor in the 2^nd hydrosilylation step.

Fig. 3 Optimized geometries of Si–H cleavage transition states with vinylsilane as the nucleophile. Reaction barriers are given in kcal mol⁻¹. Color code: H, white; C, gray; B, pink; F, green; Si, brown.

Kinetic isotope effect (KIE) determination for the hydrosilylation of the vinylsilane step. To further gain insight into the rate-determining step, intermolecular KIE was determined from the initial reaction rates of two parallel reactions of vinylsilane 3aa with PhSiH₃ and PhSiD₃ (Scheme 8 and Fig. S5†). An inverse secondary deuterium KIE (k_H/k_D = 0.81) was observed in the hydrosilylation reaction of vinylsilane in the presence of 2.25 mmol% of B(C₆F₅)₃ (Scheme 8a). KIE values of 1.4–1.9 have been reported for the hydrosilylation of carbonyl compounds and heterodehydrocoupling of phosphines and hydrosilanes.^13b,27 These values were interpreted in terms of hydride transfer.²⁸ To identify the origin of the observed kinetic isotope effect, KIE calculations using the theory of the Bigeleisen–Mayer equation²⁹ and the rigid-rotor harmonic oscillator approach³⁰ were conducted on the related transition states involved in the hydrosilylation of vinylsilane (see ESI† for Computational details).³¹ As shown in Scheme 8b, our calculations indicate that the measured inverse secondary KIE is related to the S_N²@Si transition state involving the displacement of hydride at the silicon atom with vinylsilane as the nucleophile (TS_3aa′-V, KIE_comp = 0.75, Table S4†), rather than the hydride transfer step (KIE_comp = 1.39). The inverse secondary KIE may come from the hybridization change of the silicon center from tetravalent in PhSiH₃ to pentavalent in the transition state TS_3aa′-V. With these computational and experimental results, we can conclude that the Si–H bond cleavage in adduct IV (viaTS_3aa′-V) featuring a pentavalent silicon species is involved in the rate-determining step.

Scheme 8 Kinetic isotope effect determination for the borane-catalyzed hydrosilylation reaction of vinylsilane 3aa.

Based on the experimental and computational studies, we summarized the mechanistic details of the borane-catalyzed dihydrosilylation of alkynes in Scheme 9. The selective formation of unsymmetrical geminal bis(silanes) is a kinetically controlled process, in which the alkyne hydrosilylation can readily occur with EtMe₂SiH. In turn, the hydrosilylation of vinylsilane with primary silane is much more efficient than that with EtMe₂SiH. The high selectivity toward the unsymmetrical geminal bis(silane) rather than symmetrical products could be rationalized by a combination of steric and electronic effects.

Scheme 9 Mechanistic proposal for borane-catalyzed dihydrosilylation of terminal alkynes giving rise to unsymmetrical geminal bis(silanes). BCF = B(C₆F₅)₃.

Conclusions

We have developed a borane-catalyzed selective dihydrosilylation reaction of terminal alkynes to obtain synthetic valuable unsymmetrical geminal bis(silanes), using a combination of a tertiary silane (or secondary silane) and primary silane. The method exhibits broad substrate scope and can be run in the presence of air and moisture. The resulting silane reagents can be readily functionalized through Si–H bond transformation, increasing the structural diversification of geminal bis(silanes). Combined theoretical calculations and experimental studies shed light on the mechanism, which can account for the high chemoselectivity toward unsymmetrical geminal bis(silanes). We expect our present work to stimulate future studies of boron/silane-related transformations as well as the synthetic utilities of unsymmetrical geminal bis(silane) products obtained through this protocol.

Data availability

The datasets supporting this article have been uploaded as part of the ESI.†

Author contributions

G. W., X. S., and L. G. performed the experiments. G. W. carried out the theoretical calculations. G. L. computed the KIE values. X. S. and X. L. characterized the compounds. G. W. and S. L. designed the project, analyzed the results and wrote the manuscript.

Conflicts of interest

There are no conflicts to declare.

Acknowledgements

G. W. and S. L. thank the National Natural Science Foundation of China (Grant Nos. 22073043, 21903043, 21833002, and 21673110), and the Fundamental Research Funds for the Central Universities (020514380243) for financial support. We thank Prof. Jing Ma and Prof. Xu Cheng for their helpful suggestion on the work. All theoretical calculations were performed on the High-Performance Computing Center (HPCC) of Nanjing University.

Notes and references

1. The Chemistry of Organic Silicon Compounds, ed. Z. Rappoport and Y. Apeloig, Wiley, New York, 1998, vol. 2, p. 1687.
2. For a recent review on geminal bis(silanes), see: L. Gao, Y. Zhang and Z. Song, Synlett, 2013, 24, 139–144.
3. For selected examples, see: (a) D. R. Williams, A. I. Morales-Ramos and C. M. Williams, Org. Lett., 2006, 8, 4393–4396; (b) Z. Yin, Z. Liu, Z. Huang, Y. Chu, Z. Chu, J. Hu, L. Gao and Z. Song, Org. Lett., 2015, 17, 1553–1556; (c) Z. Liu, X. Lin, N. Yang, Z. Su, C. Hu, P. Xiao, Y. He and Z. Song, J. Am. Chem. Soc., 2016, 138, 1877–1883; (d) Z. Chu, K. Wang, L. Gao and Z. Song, Chem. Commun., 2017, 53, 3078–3081; (e) Y. Zhang, Q. Guo, X. Sun, J. Lu, Y. Cao, Q. Pu, Z. Chu, L. Gao and Z. Song, Angew. Chem., Int. Ed., 2018, 57, 942–946.
4. Z. Liu, H. Tan, T. Fu, Y. Xia, D. Qiu, Y. Zhang and J. Wang, J. Am. Chem. Soc., 2015, 137, 12800–12803.
5. H. Hazrati and M. Oestreich, Org. Lett., 2018, 20, 5367–5369.
6. (a) Z. Cheng, S. Xing, J. Guo, B. Cheng, L.-F. Hu, X.-H. Zhang and Z. Lu, Chin. J. Chem., 2019, 37, 457–461; (b) J. Guo, H. Wang, S. Xing, X. Hong and Z. Lu, Chem, 2019, 5, 881–895.
7. (a) M.-Y. Hu, J. Lian, W. Sun, T.-Z. Qiao and S.-F. Zhu, J. Am. Chem. Soc., 2019, 141, 4579–4583; (b) M.-Y. Hu, P. He, T.-Z. Qiao, W. Sun, W.-T. Li, J. Lian, J.-H. Li and S.-F. Zhu, J. Am. Chem. Soc., 2020, 142, 16894–16902.
8. W. Chen, H. Song, J. Li and C. Cui, Angew. Chem., Int. Ed., 2020, 59, 2365–2369.
9. (a) R. J. P. Corriu, M. Granier and G. F. Lanneau, J. Organomet. Chem., 1998, 562, 79–88; (b) P.-F. Fu, J. Mol. Catal. A: Chem., 2006, 243, 253–257.
10. For reviews on the chemistry of electron-deficient boron Lewis acids: (a) W. E. Piers and T. Chivers, Chem. Soc. Rev., 1997, 26, 345–354; (b) M.-A. Légaré, C. Pranckevicius and H. Braunschweig, Chem. Rev., 2019, 119, 8231–8261.
11. Application of B(C₆F₅)₃ in Frustrated Lewis Pair (FLP) chemistry, see: (a) G. C. Welch, R. R. San Juan, J. D. Masuda and D. W. Stephan, Science, 2006, 314, 1124–1126; (b) P. A. Chase, G. C. Welch, T. Jurca and D. W. Stephan, Angew. Chem., Int. Ed., 2007, 46, 8050–8053; (c) G. C. Welch and D. W. Stephan, J. Am. Chem. Soc., 2007, 129, 1880–1881; (d) P. Spies, S. Schwendemann, S. Lange, G. Kehr, R. Fröhlich and G. Erker, Angew. Chem., Int. Ed., 2008, 47, 7543–7546.
12. For recent reviews, see: (a) D. Weber and M. R. Gagné, in Organosilicon Chemistry: Novel Approaches and Reactions, ed. T. Hiyama and M. Oestreich, Wiley-VCH, Weinheim, 2019, pp. 33–85; (b) M. Oestreich, J. Hermeke and J. Mohr, Chem. Soc. Rev., 2015, 44, 2202–2220; (c) W. Meng, X. Feng and H. Du, Acc. Chem. Res., 2018, 51, 191–201; (d) T. Hackel and N. A. McGrath, Molecules, 2019, 24, 432; (e) H. Fang and M. Oestreich, Chem. Sci., 2020, 11, 12604–12615.
13. (a) D. J. Parks and W. E. Piers, J. Am. Chem. Soc., 1996, 118, 9440–9941; (b) D. J. Parks, J. M. Blackwell and W. E. Piers, J. Org. Chem., 2000, 65, 3090–3098; (c) M. Rubin, T. Schwier and V. Gevorgyan, J. Org. Chem., 2002, 67, 1936–1940; (d) S. Rendler and M. Oestreich, Angew. Chem., Int. Ed., 2008, 47, 5997–6000; (e) A. Y. Houghton, J. Hurmalainen, A. Mansikkamäki, W. E. Piers and H. M. Tuononen, Nat. Chem., 2014, 6, 983–988.
14. Selected examples of defunctionalization reactions with borane/silane combinations: (a) V. Gevorgyan, M. Rubin, J.-X. Liu and Y. Yamamoto, J. Org. Chem., 2001, 66, 1672–1675; (b) L. L. Adduci, T. A. Bender, J. A. Dabrowski and M. R. Gagné, Nat. Chem., 2015, 7, 576–581; (c) T. Mahdi and D. W. Stephan, Angew. Chem., Int. Ed., 2015, 54, 8511–8514; (d) J. Lowe, B. Bowers, Y. Seo and M. R. Gagné, Angew. Chem., Int. Ed., 2020, 59, 17297–17300.
15. For some recent transformations based on the borane/silane catalytic system, see: (a) N. Gandhamsetty, S. Joung, S.-W. Park, S. Park and S. Chang, J. Am. Chem. Soc., 2014, 136, 16780–16783; (b) C. K. Hazra, N. Gandhamsetty, S. Park and S. Chang, Nat. Commun., 2016, 7, 13431; (c) Y. Ma, B. Wang, L. Zhang and Z. Hou, J. Am. Chem. Soc., 2016, 138, 3663–3666; (d) Y. Ma, L. Zhang, Y. Luo, M. Nishiura and Z. Hou, J. Am. Chem. Soc., 2017, 139, 12434–12437; (e) Z.-Y. Zhang, Z.-Y. Liu, R.-T. Guo, Y.-Q. Zhao, X. Li and X.-C. Wang, Angew. Chem., Int. Ed., 2017, 56, 4028–4032; (f) C. K. Hazra, J. Jeong, H. Kim, M. H. Baik, S. Park and S. Chang, Angew. Chem., Int. Ed., 2018, 57, 2692–2696Z.-Y. Liu, M. Zhang and X.-C. Wang, J. Am. Chem. Soc., 2020, 142, 581–588.
16. (a) C. Chen, F. Eweiner, B. Wibbeling, R. Fröhlich, S. Senda, Y. Ohki, K. Tatsumi, S. Grimme, G. Kehr and G. Erker, Chem.–Asian J., 2010, 5, 2199–2208; (b) C. Chen, G. Kehr, R. Fröhlich and G. Erker, J. Am. Chem. Soc., 2010, 132, 13594–13595; (c) C. Chen, T. Voss, G. Kehr, R. Fröhlich and G. Erker, Org. Lett., 2011, 13, 62–65; (d) G. Kehr and G. Erker, Chem. Commun., 2012, 48, 1839–1850; (e) G. Kehr and G. Erker, Chem. Sci., 2016, 7, 56–65.
17. A. Bismuto, G. S. Nichol, F. Duarte, M. J. Cowley and S. P. Thomas, Angew. Chem., Int. Ed., 2020, 59, 12731–12735.
18. Although borane-catalyzed hydrosilylation of internal alkynes is known, its application in the hydrosilylation of terminal alkynes has not been disclosed. For the hydrosilylation of internal alkynes with the borane/silane system, see: (a) L. D. Curless and M. J. Ingleson, Organometallics, 2014, 33, 7241–7246; (b) Y. Kim, B. Ramesh, R. B. Dateer and S. Chang, Org. Lett., 2017, 19, 190–193; (c) X. Zhao, D. Yang, Y. Zhang, B. Wang and J. Qu, Org. Lett., 2020, 22, 970–975.
19. For computational analysis, see: K. Sakata and H. Fujimoto, J. Org. Chem., 2013, 78, 12505–12512.
20. Lewis acids, such as AlCl₃ and EtAlCl₂, which were used in the hydrosilylation of alkynes to produce vinylsilanes, see: T. Sudo, N. Asao, V. Gevorgyan and Y. Yamamoto, J. Org. Chem., 1999, 64, 2494–2499 . However, AlCl₃ did not produce 1,1-bis(silane) 4aac using a combination of EtMe₂SiH and PhSiH₃..
21. For air- or moisture-tolerant borane Lewis acid-catalyzed reactions, see: (a) K. Ishihara, N. Hanaki, M. Funahashi, M. Miyata and H. Yamamoto, Bull. Chem. Soc. Jpn., 1995, 68, 1721–1730; (b) V. Fasano, J. E. Radcliffe, A. Gyömöre, M. Bakos, T. Földes, I. Pápai, A. Domján and T. Soós, ACS Catal., 2015, 5, 5366–5372; (c) D. J. Scott, T. R. Simmons, E. J. Lawrence, G. G. Wildgoose, M. J. Fuchter and A. E. Ashley, ACS Catal., 2015, 5, 5540–5544; (d) V. Fasano, J. E. Radcliffe and M. J. Ingleson, ACS Catal., 2016, 6, 1793–1798.
22. (a) H. Mayr and G. Hagen, J. Chem. Soc., Chem. Commun., 1989, 91–92; (b) M. Horn, L. H. Schappele, G. Lang-Wittkowski, H. Mayr and A. R. Ofial, Chem. -Eur. J., 2013, 19, 249–263.
23. (a) Y. Zhao and D. G. Truhlar, Theor. Chem. Acc., 2008, 120, 215–241; (b) Y. Zhao and D. G. Truhlar, Acc. Chem. Res., 2008, 41, 157–167.
24. J. Tomasi and M. Persico, Chem. Rev., 1994, 94, 2027–2094.
25. C. Y. Legault, CYLview, 1.0b, Universitéde Sherbrooke, 2009, http://www.cylview.org.
26. Morandi et al elegantly proved that the type of silane is also critical to the regioselective deoxygenation of 1,2-Diols, see: (a) N. Drosos and B. Morandi, Angew. Chem., Int. Ed., 2015, 54, 8814–8818; (b) G.-J. Cheng, N. Drosos, B. Morandi and W. Thiel, ACS Catal., 2018, 8, 1697–1702.
27. L. Wu, S. S. Chitnis, H. Jiao, V. T. Annibale and I. Manners, J. Am. Chem. Soc., 2017, 139, 16780–16790.
28. H. Mayr, N. Basso and G. Hagen, J. Am. Chem. Soc., 1992, 114, 3060–3066.
29. J. Bigeleisen and M. G. Mayer, J. Chem. Phys., 1947, 15, 261–267.
30. (a) D. J. O'Leary, P. R. Rablen and M. P. Meyer, Angew. Chem., Int. Ed., 2011, 50, 2564–2567; (b) A. Fong, M. P. Meyer and D. J. O'Leary, Molecules, 2013, 18, 2281–2296.
31. Quantum mechanical tunneling effects were also considered for both methods using the one-dimensional parabolic approximation. (a) R. P. Bell, Chem. Soc. Rev., 1974, 3, 513–544; (b) R. P. Bell, The Tunnel Effect in Chemistry; Chapman and Hall: New York, 1980; (c) D. B. Northrop, J. Am. Chem. Soc., 1999, 121, 3521–3524.

---

Footnotes

† Electronic supplementary information (ESI) available: Experimental details, characterization, and computational data. See DOI: 10.1039/d1sc02769g

‡ These authors contributed equally to this work.

---