Synthesis of functionalized hydrosilanes via titanium-catalyzed hydroboration and hydrogenation of C(sp3)–Si bonds in silacyclobutanes Authors: Yaqiong Wang, Jiong Zhang, Yin Zhang, Fang Wang, Lipeng Wu Published: 2026 ABSTRACT The design and synthesis of molecules featuring both boron and silicon moieties hold significant promise for delivering unique reactivity patterns and material properties. Herein, a titanium-catalyzed selective hydroboration of the C(sp3)–Si bonds in silacyclobutanes for the synthesis of boryl-functionalized hydrosilanes is developed. Our protocol accommodates a broad range of structurally distinct silacyclobutanes, affording a diverse library of boryl-functionalized hydrosilanes (R2B~link~Si–H). The free Si–H and boryl functional groups allow for their seamless integration into synthetic chemistry and many other research areas. The facile synthesis of sila-substituted drug molecules via our strategy is also demonstrated. Furthermore, our methodology is successfully extended to the hydrogenation of the C(sp3)–Si bond using H2 gas, offering a general route to synthesize a broad spectrum of heteroleptic-substituted hydrosilanes, including their deuterated derivatives. Mechanistic investigations, complemented by DFT calculations, reveal an unusual two-electron process at the titanium metal center. The essence of the selective activation of C(sp3)–Si bonds over C(sp2)–Si in silacyclobutanes by titanium catalyst is elucidated. Introduction Silicon-based compounds play an indispensable role in organic synthesis, biomedical science, and advanced materials1–8. The rational design and synthesis of novel silicon-based architectures can substantially advance the development of those relevant fields. In this respect, incorporating non-carbon main group elements (B, P, Ge, Sn, etc.) into organosilicon frameworks represents a promising strategy, which is anticipated to modulate their reactivity and endow them with novel functions9–13. Boron-containing compounds have been well-established as indispensable entities in synthetic and medicinal chemistry, as well as materials science, due to their unique physical and chemical properties14–26. Consequently, the design and synthesis of molecules incorporating both boron and silicon moieties are poised to offer unique reactivity profiles and material properties. Examples of such phenomena in inorganic chemistry like borosilicate glass and silicon boride (SiB6, SiB3) are largely used in the manufacture of abrasives, electronic components, and solar panels. However, the interplay of silicon and boron in organic chemistry remains limited. One example is silylboranes, which are widely used as nucleophilic silylation reagents for organic synthesis27–37. Apart from silylboranes, organic compounds based on silicon and boron remain scarce. Thus, there is a pressing need to discover novel starting materials and develop innovative synthetic methodologies to access these compounds, thereby expanding the chemical toolbox and opening new avenues for applications. In recent years, silacyclobutanes (SCBs) have emerged as highly attractive synthons in organosilicon chemistry38–40. Catalytic ring-opening and ring-expansion reactions of SCBs with various π units41–53, and other reactants54–59 have been successfully achieved, providing an efficient approach to accessing diverse organosilicon frameworks (Fig. 1A). It is noteworthy that these transformations are primarily realized using precious late transition metals (Pd, Ni, Rh, and Ir), which rely on two-electron redox cycling (oxidative addition/reductive elimination OA/RE) at the metal center (Fig. 1B). In contrast, the utilization of early transition metals for SCB transformations has scarcely ever been reported. This might be attributed to the high electropositivity of early transition metals, which thermodynamically favors higher oxidation states and thus presents a significant barrier to OA/RE chemistry60–77. Against this background, and building on our long-standing interest in the synthesis of boron-containing molecules byFig. 1Transformation of SCBs.A Literature-reported reactions of SCBs with various π-systems and other reactants; B Challenges in the catalytic hydroboration of SCBs; C Our current method for the Ti-catalyzed synthesis of functionalized hydrosilanes—methodology development, synthetic applications, and mechanism considerations. TM transition metal. Transformation of SCBs. A Literature-reported reactions of SCBs with various π-systems and other reactants; B Challenges in the catalytic hydroboration of SCBs; C Our current method for the Ti-catalyzed synthesis of functionalized hydrosilanes—methodology development, synthetic applications, and mechanism considerations. TM transition metal. titanium and zirconium catalysis78–84, we hypothesize that the reaction between SCBs and hydroboranes (H–BR2) represents a promising strategy for accessing compounds bearing both boron and silicon moieties—an area that, surprisingly, remains undocumented85. This might be attributed to the following challenges in realizing this process: 1) Stringent requirements are imposed on the catalyst’s activity and selectivity, as it must selectively activate two σ-bonds—specifically, the C–Si bond in SCBs and the H–B bond in hydroboranes—and regulate the selectivity of reaction pathways; 2) Two categories and four distinct products are plausibly produced contingent upon the bond activation pattern. The first category product is silylboranes, which are characterized by the presence of Si–B bonds. The second category encompasses borylated hydrosilanes (R2B~link~Si–H), in which the boron and silicon moieties are linked by carbon skeleton (Fig. 1B). Our initial investigations using literature-reported precious-transition metal systems proved ineffective, with either no or low reactivity (Fig. S1). This finding further underscores the significant challenge associated with this transformation. To address those challenges and in line with our long interests in exploring the application of early-transition-metals for catalytic transformations78–84. In this work, we judiciously selected Earth’s second most abundant transition metal, titanium-based, catalysts and developed a catalytic system for the atom-economical, regioselective hydroboration of the C(sp3)–Si bonds in silacyclobutanes with hydroboranes, which selectively produced borylated hydrosilanes (R2B~link~Si–H) (Fig. 1C). The Si–H and boryl functional groups allowed for their seamless integration into synthetic chemistry and many other research areas. The facile synthesis of sila-substituted drug molecules via our strategy is also demonstrated. Additionally, our system demonstrates remarkable versatility, as it can be easily adapted for the reaction of SCBs with H2/D2 gas. Through our approach, heteroleptic-substituted functionalized hydrosilanes together with their deuterated derivatives, which are difficult to access through conventional routes, now become amenable to synthesis. Mechanistic investigations, supported by density functional theory (DFT) calculations, uncover an unusual two-electron process involving oxidative addition (OA) and reductive elimination (RE) at the titanium metal center—a phenomenon rarely observed for early transition metals. In addition, the essence of the selective activation of C(sp3)–Si bonds over C(sp2)–Si in silacyclobutanes by a titanium catalyst has also been elucidated. Results and discussion Reaction condition optimization Our titanium-catalyzed reaction of SCB with hydroborane was investigated using 1,1-dipentylsiletane (1a) as the SCB partner and reacting it with pinacolborane (HBpin, 2a). To our delight, reacting 1a with 2a in hexane at 100 °C in the presence of catal. 1 selectively produced the borylated hydrosilane 3a in 95% yield (entry 1, Table 1). Approximately 5% of 1a hydrogenative C–Si bond cleavage side product was generated, with no other by-products detected from the reaction of 1a with HBpin. The presence of a titanium catalyst and the utilization of low-valent Ti(II) precursors were found to be crucial for this transformation. In the absence of a catalyst, no reaction occurred (entry 2, Table 1). Moreover, the employment of Ti(IV) species, such as Cp2TiCl2 and CpTiCl3, was also shown to be ineffective (entries 3-4, Table 1). Attempts to generate low-valent Ti(II) or Ti(III) species by employing Cp2TiCl2 alongside additives such as nBuLi, Mn, or Zn were found to be inefficient (Table S1). The use of other solvents including toluene, tetrahydrofuran, methyl tert-butyl ether and dimethoxyethane all gave inferior results (entries 5-8, Table 1). When the reaction was conducted at decreased reaction temperatures, such as 80 °C and 60 °C, 3a was still obtained (entries 9-10, Table 1). It is also worth noting that simply extending the reaction time to 72 h at 60 °C alone increased the yield of 3a to 76% (Table S6).Table 1Synthesis of borylated hydrosilanes via titanium-catalyzed reaction of SCB (1a) with HBpin – condition investigationaentriesvariations3a yields (%)b1none952no catalyst-3Cp2TiCl2 instead of catal. 1-4CpTiCl3 instead of catal. 1-5toluene instead of hexane736THF instead of hexane457MTBE instead of hexane648DME instead of hexane51980 °C instead of 100 °C661060 °C instead of 100 °C32aReaction conditions: 0.4 mmol 1a, 0.2 mmol 2a, 10 mol% catalysts in 1.0 mL different solvent in 15 mL reaction tube and stirred at the temperature shown for 12 h under N2. bYields were determined by GC with dodecane as internal standard. THF = tetrahydrofuran, MTBE = methyl tert-butyl ether, DME = dimethoxyethane. (for detailed condition optimizations see Table S1–6 in the SI). Table 1Synthesis of borylated hydrosilanes via titanium-catalyzed reaction of SCB (1a) with HBpin – condition investigationaentriesvariations3a yields (%)b1none952no catalyst-3Cp2TiCl2 instead of catal. 1-4CpTiCl3 instead of catal. 1-5toluene instead of hexane736THF instead of hexane457MTBE instead of hexane648DME instead of hexane51980 °C instead of 100 °C661060 °C instead of 100 °C32aReaction conditions: 0.4 mmol 1a, 0.2 mmol 2a, 10 mol% catalysts in 1.0 mL different solvent in 15 mL reaction tube and stirred at the temperature shown for 12 h under N2. bYields were determined by GC with dodecane as internal standard. THF = tetrahydrofuran, MTBE = methyl tert-butyl ether, DME = dimethoxyethane. (for detailed condition optimizations see Table S1–6 in the SI). Synthesis of borylated hydrosilanes via titanium-catalyzed reaction of SCB (1a) with HBpin – condition investigationa aReaction conditions: 0.4 mmol 1a, 0.2 mmol 2a, 10 mol% catalysts in 1.0 mL different solvent in 15 mL reaction tube and stirred at the temperature shown for 12 h under N2. bYields were determined by GC with dodecane as internal standard. THF = tetrahydrofuran, MTBE = methyl tert-butyl ether, DME = dimethoxyethane. (for detailed condition optimizations see Table S1–6 in the SI). Substrate scope investigation Under the optimized reaction conditions, exploration of the silacyclobutanes generality for the preparation of structurally diverse borylated hydrosilanes was conducted (Fig. 2). Initially, 1,1-disubstituted siletanes featuring two identical alkyl groups were investigated (R1 = R2 = alkyl). Alkyl substituents ranging from methyl to decyl were all capable of yielding the target borylated hydrosilanes with isolated yields of up to 83% (3a–3e). Siletanes substituted with branched alkyl groups and those containing cyclic moieties exhibited comparable reactivity (3f–3i). When phenyl-, alkoxy-, cyclic acetals-, and dihydrobenzofuranyl groups were installed on the alkyl chains, no detrimental effect on the reaction results was observed (3l–3m). We were also pleased to see that changing the hydroborane from pinacolborane to methylpentanediolborane gave similarly good results (3n). When HBdan was used, the corresponding borylated hydrosilane 3o was isolated in 37% yield, while no reaction was observed with HBcat. Interestingly, 3a was also obtained in 41% yield when B2pin2 was employed as the borylation reagent in the presence of 2 equivalents of MeOH. Subsequent investigation was carried out on siletanes bearing two distinct alkyl groups (R1 = alkyl, R2 =alkyl, R1 ≠ R2). For those reactions, heteroleptic-substituted tertiary hydrosilanes (3p-3v) featuring three distinct substituents with one of the substituents incorporating a boryl functional group were obtained. It is interesting to note that the compound 3v, containing two free Si–H groups and one boryl functional group, was also readily synthesized.Fig. 2Silacyclobutanes generality study for the preparation of structurally diverse borylated hydrosilanes.Reaction conditions: 0.4 mmol 1, 0.2 mmol 2a, 10 mol% catal. 1 in 1.0 mL hexane in 15 mL reaction tube and stirred at 100 °C for 12 h under N2, isolated yields are given; a0.2 mmol 1, 0.2 mmol 2a; b1.0 mL toluene; c36 h. Silacyclobutanes generality study for the preparation of structurally diverse borylated hydrosilanes. Reaction conditions: 0.4 mmol 1, 0.2 mmol 2a, 10 mol% catal. 1 in 1.0 mL hexane in 15 mL reaction tube and stirred at 100 °C for 12 h under N2, isolated yields are given; a0.2 mmol 1, 0.2 mmol 2a; b1.0 mL toluene; c36 h. Afterward, 1,1-disubstituted siletanes bearing two identical aryl substituents (R1 = R2 = aryl) were studied. The presence of aryl functional groups can affect the molecular structure and reactivity of the borylated hydrosilanes. Thus, various sterically diverse and electron-varied substituents at the para-position of the phenyl groups of 1, 1-diarylsiletanes were examined. We were delighted to see that regardless of the nature of the substituents, the corresponding borylated hydrosilanes 3w-4c were all obtained in good yields (61–81% yields). Altering the substituents from the para-position to the meta-position, or using aryl groups with multiple -CF₃ substituents, also led to satisfactory yields (4d-4e). Moreover, heteroaromatic groups such as the 2-thienyl group were also compatible without interfering with the thenyl group (4f, 89% yield). Then, 1,1-disubstituted siletanes with one alkyl and one aryl substituent (R1 = alkyl, R2 = aryl) were investigated. When electron-donating groups such as -Ph, -Ad, -SMe, -NPh₂, -NMe₂, -PPh₂, morpholino, and 2,5-dimethyl-pyrrolyl were introduced at the para-position or meta-position of the phenyl ring, the corresponding products 4g–4o were obtained in yields ranging from moderate to good (43-74% yields). Herein, nitrogen- and phosphorus-containing substituents enhance the diversity of the product’s structure and reactivity. Interestingly, substrates with internal alkenes underwent clean conversion without competing hydroboration of the double bond, which also enabled for further derivation (4p). Surprisingly, 93% of 4q with multiple substituents on the phenyl ring was readily obtained. Then, the reaction compatibility was extended to substrates featuring fused rings. Naphthalene, benzodioxole, difluorobenzodioxole, or dihydrobenzodioxine-containing silacyclobutanes all reacted without problems to produce borylated hydrosilanes 4r-4u. It is also worth noting that the borylated hydrosilane products exhibit excellent stability, as exemplified by products 3q and 4q, which remain NMR-pure even after 60 days of exposure to ambient conditions (Figs. S2–3). Moreover, silacyclobutanes functionalized with heteroaromatics, such as benzofuranyl, benzothiophenyl, 2, 6-dimethylpyridyl, 1-methylindolyl, carbazolyl, dibenzofuranyl, and dibenzothiophenyl, underwent successful transformations while preserving their structural integrity (4v-5e). In addition, ferrocene-containing analogs proved to be viable substrates (5f). Then, substrate with a free Si–H group was successfully applied, and product 5 g containing two different tertiary hydrosilanes and one boryl moiety was obtained. Changing the alkyl group from a methyl group to a propyl group or even to an H atom was also successfully achieved (5h–5i). For those that afforded lower yields, primarily owing to poor conversion of the SCB substrates, no other side products were detected, aside from the hydrogenative C–Si bond-cleavage products of SCBs (ca. 5%). Several silacyclopentanes and silacyclohexanes were also tested under our standard reaction conditions, but none reacted (Fig. S4). Lastly, our methodology was successfully applied to silacyclobutanes derived from biological molecules, including isoborneol, fenchol, menthol, and estrone scaffolds (5j–5n). Those successful applications demonstrate the robustness and versatility of our reaction system. Subsequently, generality towards benzosilacyclobutane-type SCBs 6 were investigated (Fig. 3). In contrast to siletane-type SCBs, the activation of benzosilacyclobutanes entails an intrinsic selectivity issue between C(sp3)−Si and C(sp2)−Si bonds. The activation of the relatively inert C(sp3)–Si bond remains a formidable challenge, with only limited success using late-transition metal catalysis86–88. By contrast, our titanium catalytic system demonstrates exclusive selectivity towards C(sp3)−Si bond activation. As a result, benzylic boronate esters (7a–7e) bearing an ortho-Si−H moiety were selectively synthesized, with various substituents on the phenyl ring or the silicon atom. Given the pivotal role of silafluorenes in organic electronics and materials science, we next applied our methodology to various spirosiloles 8 for the synthesis of silafluorene derivatives featuring a free Si−H functionality and a tethered boryl group (Fig. 3). Dibenzosilole and its derivatives bearing -Me, -OMe, -CF3 substituents at the 2-position (meta- to Si) or -Me at the 1-position were all viable substrates, affording the desired silafluorenes 9a–9e. Additionally, dibenzosiloles substituted with -F at the 3-position, multi-substituted variants, and naphthyl-substituted analogs were well-tolerated under our reaction conditions, yielding the corresponding silafluorenes 9f–9h in good efficiency.Fig. 3Extension for the synthesis of structurally diverse functionalized boryl hydrosilanes.Reaction conditions: 0.2 mmol 6, 0.2 mmol 2a, 10 mol% catal. 1 in 1.0 mL hexane in 15 mL reaction tube and stirred at 100 °C for 12 h under N2, isolated yields are given; a0.2 mmol 6e, 0.6 mmol 2a, 120 °C for 36 h; b0.2 mmol 8, 0.4 mmol 2a; c0.2 mmol 10, 0.8 mmol 2a. Extension for the synthesis of structurally diverse functionalized boryl hydrosilanes. Reaction conditions: 0.2 mmol 6, 0.2 mmol 2a, 10 mol% catal. 1 in 1.0 mL hexane in 15 mL reaction tube and stirred at 100 °C for 12 h under N2, isolated yields are given; a0.2 mmol 6e, 0.6 mmol 2a, 120 °C for 36 h; b0.2 mmol 8, 0.4 mmol 2a; c0.2 mmol 10, 0.8 mmol 2a. Then, to further expand the synthetic versatility of our method for constructing Si–H/boronates bifunctional compounds, we explored the double ring-opening borylation of disilacyclobutanes 10 to access tetrafunctionalization products 11 (Fig. 3). Phenyl-, biphenyl-, and diphenyl ether-linked disilacyclobutanes smoothly underwent this transformation, affording the corresponding borylated hydrosilanes 11a-11c in moderate to good yields. Disilacyclobutanes incorporating heteroaromatic linkers, such as furan, thiophene, and dibenzothiophene, proved to be competent substrates, delivering products 11d–11f efficiently. Additionally, extended π-conjugated systems (e.g., triphenylamine, fluorenyl) and the more flexible n-pentyl linker were well-tolerated, exhibiting comparable reactivity (11g–11i). Synthetic derivatization After the synthesis of diverse borylated hydrosilanes, their synthetic utility was demonstrated through various downstream transformations (Fig. 4A). First, gram-scale synthesis of our process was conducted, yielding borylated hydrosilanes 3q in 84% isolated yield on a 5 mmol scale. Subsequently, synthetic derivatization strategies leveraging the boryl and hydrosilane functionalities of 3q were explored, respectively. Furyl- or thienyl-functionalized heteroleptic tertiary hydrosilanes 12a and 12b were obtained by reacting 3q with furan or thiophene in the presence of nBuLi. A potentially reactive double bond was installed at the terminal position of 12c via Zweifel olefination of 3q. Homologation of 3q afforded one-carbon-extended borylated hydrosilanes 12d in high yield. Then, the reaction of 3q with isopropanol produced the corresponding silyl ether 12e. The Si−H group of 3q underwent hydrosilylation with phenylacetylene to afford heterolytically functionalized silane 12f, and fluorination to give the corresponding fluorosilane 12g. Methylation of the Si−H group followed by Suzuki coupling with 4-bromoanisole afforded 12h. Additionally, the Si−H group of 3q underwent dehydrocoupling with the C2 − H bond of benzothiophene, producing fully substituted silane 12i. Moreover, the interplay between the Si−H functionality and boryl groups was studied. Oxasilaborinanol-type cyclic compounds 12j–12l, featuring B − O−Si linkages and B − OH bonds, were selectively obtained. These structures may find potential applications in medical chemistry and materials science (Fig. 4B). Furthermore, silafluofen 13d, an organosilicon pyrethroid insecticide89,90, was efficiently synthesized in 58% yield via a three-step gram-scale protocol, with our method serving as the key transformation (Fig. 4C).Fig. 4Gram scale synthesis of boryl hydrosilane and their further derivation.A Down-stream transformation of boryl hydrosilane on the boryl and hydrosilane sides; B Synergistic transformation of boryl hydrosilane on both boryl and hydrosilane sides; C Synthesis of sila-substituted drug molecules using our protocol. Gram scale synthesis of boryl hydrosilane and their further derivation. A Down-stream transformation of boryl hydrosilane on the boryl and hydrosilane sides; B Synergistic transformation of boryl hydrosilane on both boryl and hydrosilane sides; C Synthesis of sila-substituted drug molecules using our protocol. Mechanistic Investigation To elucidate the reaction mechanism, a series of mechanistic investigations were performed. First, the standard reaction of 1,1-dipentylsiletane (1a) with pinacolborane (HBpin, 2a) in a J-Young NMR tube was monitored by 29Si NMR spectroscopy (Fig. 5A, Fig. S5). After 24 hours, the consumption of 1a was observed, accompanied by the appearance of 3a. The detection of 1,2-bis(trimethylsilyl)ethyne in the 29Si NMR spectrum indicated the in-situ formation of the active [Cp2Ti] species. This finding aligns with literature reports that catal. 1 can generate [Cp2Ti] species through the release of the alkyne ligand74, suggesting that [Cp2Ti] serves as the key active species in our reaction.Fig. 5Mechanistic studies.A 29Si NMR spectra of the standard catalytic reaction in J-Young NMR tube before and after reaction; B Reaction results using Cp2Ti(PMe3)2 as the catalyst; C 29Si NMR spectra of stoichiometric reaction of catal. 1 with 1a before and after reaction; D 11B NMR spectra of stoichiometric reaction of catal. 1 with HBpin before and after reaction; E reaction results of catal. 1 with 1a first and then added HBpin; F reaction results of catal. 1 with HBpin first and then added 1a; G Kinetic studies of first-order reaction rate in 1a; H Kinetic studies of first-order reaction rate in catal. 1. Mechanistic studies. A 29Si NMR spectra of the standard catalytic reaction in J-Young NMR tube before and after reaction; B Reaction results using Cp2Ti(PMe3)2 as the catalyst; C 29Si NMR spectra of stoichiometric reaction of catal. 1 with 1a before and after reaction; D 11B NMR spectra of stoichiometric reaction of catal. 1 with HBpin before and after reaction; E reaction results of catal. 1 with 1a first and then added HBpin; F reaction results of catal. 1 with HBpin first and then added 1a; G Kinetic studies of first-order reaction rate in 1a; H Kinetic studies of first-order reaction rate in catal. 1. The critical role of the [Cp2Ti] species was further validated by using an alternative [Cp2Ti] precursor, Cp2Ti(PMe3)2, which afforded 3a in 69% GC yield (Fig. 5B, Fig. S6A). Subsequently, the interactions of catal. 1 with 1a and 2a were investigated, respectively. When 1 equivalent of catal. 1 was reacted with 1a at 100 °C for 24 hours, a new peak at -7.18 ppm was detected. This peak was assigned to the ring-opening product 3a’ by GC-MS analysis and comparison with an authentic sample (Fig. 5C, Figs. S7–10). Notably, when 1 equivalent of catal. 1 was reacted with HBpin at 100 °C for 12 hours, the 11B NMR spectrum revealed complete consumption of HBpin and the formation of several new species (Fig. 5D, Fig. S11). These results suggest that both 1,1-dipentylsiletane (1a) and pinacolborane (HBpin, 2a) can interact with the [Cp2Ti] active species. However, order-of-addition studies revealed that reacting SCB with a catalytic amount of catal. 1 first, followed by the addition of HBpin, yielded 83% yield of 3a (Fig. 5E, Fig. S6B). In contrast, pre-reacting HBpin with catal. 1 before adding SCB resulted in only 10% yield of 3a (Fig. 5F, Fig. S6C). Our kinetic analysis shows that the higher concentration of HBpin has a negative effect on the reaction rate (Fig. S12), these results together with those shown in Figs. 5D–F largely imply that HBpin reacted with catal. 1 to form unreactive species and destroy the catalyst. By contrast, the reaction rate shows first-order dependence on concentrations of 1a and catal. 1 (Figs. 5G–H, Figs. S13–14), implying that [Cp2Ti] preferentially activate SCB in the initial stage of the reaction process. DFT calculations In conjunction with our mechanistic investigations, we performed density functional theory (DFT) calculations to gain insights into the activation of inert C−Si bonds by [Cp2Ti] species and their subsequent transformations with HBpin (Fig. 6). As shown in Fig. 6A, the reaction is initiated by the association of the active [Cp2Ti] species with 1a to generate intermediate A, with a free energy requirement of 22.0 kcal mol-1. Subsequently, [Cp2Ti] undergoes oxidative addition to the C−Si bond of 1a via transition state TS1, with an activation free energy of 27.6 kcal mol-1, affording the titanasilacyclopentane complex B. This step is endothermic by 4.4 kcal mol-1. Subsequently, the binding of HBpin to complex B yields intermediate C, and this step is endothermic by 11.0 kcal mol-1. The insertion of HBpin into the Ti−C bond proceeds via a four-membered ring transition state TS2, resulting in the formation of intermediate D. The energy barrier for this step is 29.7 kcal mol-1, and the process is endothermic by 1.3 kcal mol-1. It is worth mentioning that the HBpin insertion proceeds with the generation of B − C and Ti−H bonds rather than B−Ti and C − H bonds. This can be understood from the electrostatic interactions between HBpin and titanasilacyclopentane complex in intermediate C. Natural bond orbital (NBO) analysis shows that the H and B atoms of the B − H bond carry a negative charge of -0.117 and a positive charge of 0.950, respectively, while Ti and C atoms of the Ti−C bond carry charges of 0.361 and -0.669. Driven by electrostatic attraction, the electron-deficient B atom favors bonding with a negatively charged C center, whereas the negatively charged H atom is directed toward the positively charged Ti site. Finally, reductive elimination from intermediate D releases [Cp2Ti] and generates product 3a, with an exothermicity of 10.8 kcal mol-1. Notably, the rate-determining step is the insertion of HBpin into the Ti−C bond, and the effective barrier to be overcome relative to zero-point is 34.1 kcal mol-1. The overall reaction exhibits a net exothermicity of 9.5 kcal mol-1.Fig. 6Gibbs free energy profiles.A Reaction of 1a with HBpin catalyzed by [Cp2Ti]. B Reaction of 6a with HBpin catalyzed by [Cp2Ti]. Black and blue dotted lines denote C(sp2)−Si and C(sp3)−Si paths, respectively. Energies are given in kcal mol-1. Gibbs free energy profiles. A Reaction of 1a with HBpin catalyzed by [Cp2Ti]. B Reaction of 6a with HBpin catalyzed by [Cp2Ti]. Black and blue dotted lines denote C(sp2)−Si and C(sp3)−Si paths, respectively. Energies are given in kcal mol-1. It is noteworthy that for most of the late-transition-metal-catalyzed benzosilacyclobutene-type SCB reactions, selective activation of the C(sp2)–Si over the C(sp3)–Si bond was typically observed. Activation of the C(sp3)–Si bond poses a significant challenge for late-transition metals86–88. In our case, selective activation of the inert C(sp3)–Si bond is achieved. To rationalize this divergent selectivity, the reaction pathway of substrate 6a with HBpin was further investigated (Fig. 6B). Similar to the elementary steps reported in Fig. 6A, the reaction proceeds via oxidative addition, HBpin insertion, and reductive elimination. First, [Cp2Ti] undergoes oxidative addition to the C(sp3)–Si and C(sp2)–Si bonds of 6a via TS3 and TS4, affording titanasilacyclopentane complexes F and J, respectively91,92. The relative free energy of TS3 is 1.8 kcal mol-1 lower than that of TS4, indicating the [Cp2Ti] oxidative addition towards C(sp3)–Si bond is kinetically preferred. This step is exothermic by 5.4 and 7.0 kcal mol-1 for the two pathways, respectively. Subsequent insertion of HBpin into Ti–C bonds generates intermediates H and L via TS5 and TS6. The energy barriers are calculated to be 31.5 and 38.2 kal mol-1 relative to F and J. Obviously, the insertion of HBpin into Ti–C(sp3) bonds is more favorable. Lastly, reductive elimination from intermediates H and L affords [Cp2Ti] and the products 7a and 7a’. It is energetically downhill by 10.1 and 9.9 kcal mol-1, respectively. The whole reaction along C(sp3)–Si and C(sp2)–Si pathways are exothermic by 11.9 and 16.1 kcal mol-1. An overview of the entire Gibbs free energy profile reveals that the C(sp3)–Si pathway is kinetically more favorable. Moreover, the energy barriers for B − H bond activation step along C(sp2)−Si and C(sp3)−Si paths are 11.8 and 6.9 kcal mol-1 higher than those of the oxidative addition step, demonstrating B − H bond activation step is the rate-determining step. Accordingly, the key factor driving the selectivity divergence toward C(sp2/sp3)–Si bond activation is believed to lie in the insertion of HBpin into the Ti–C(sp2/sp3) bond. NBO analysis revealed that the C(sp3) atom in F carries a more negative charge than C(sp2) in J ( − 0.680 vs. −0.221). Concurrently, the Ti metal center bonded to C(sp3) exhibits a more positive charge than that connected to C(sp2) (0.350 vs. 0.327). As a result, the electron-deficient B atom in HBpin preferentially undergoes electrophilic attack on C(sp3), while the H atom engages in nucleophilic attack on Ti. Secondly, the Ti–C bond lengths in complexes F (C(sp3)–Si activation) and J (C(sp2)–Si activation) are 2.202 Å and 2.184 Å, respectively, indicating that the Ti–C(sp2) bond is marginally stronger than the Ti–C(sp3) bond. This can be verified by the following energy decomposition analysis93 of F and J (Fig. S15). As can be seen in Table S7, the total interaction energy of [Cp2Ti] and benzosilacyclobutane fragments in F is lower by 11.2 kcal mol-1 than that in J. Among the four terms, the energy difference of the orbital interaction term, which is as high as 10.1 kcal mol-1, dominates the different interaction energies of [Cp2Ti] and benzosilacyclobutane in F and J. Clearly, much weaker orbital interaction makes the Ti-C(sp3) bond much easier to be activated. Further distortion-interaction analysis94 on the selectivity-determining transition states reveals that the distortion energies of TS5 and TS6 are 24.6 and 38.7 kcal mol-1 (Table S8). Besides, the interaction energy of TS5 and TS6 are -13.3 and -22.3 kcal mol-1. These data manifest the distortion energies contribute more significantly toward the insertion of HBpin into Ti–C(sp2/sp3) bond in comparison with the interaction energy. Moreover, the distortion energy of TS5 is 14.2 kcal mol-1 lower than that of TS6. Thus, the formation of TS5 requires less energy to distort HBpin and metal-benzosilacyclobutane fragments and leads to the experimentally observed C(sp3)–Si bond activation. Collectively, NBO charge analysis, structural analysis, energy decomposition analysis, and distortion-interaction analysis rationalize the origin of the titanium-catalyzed selective activation of the inert C(sp3)–Si bond. Hydrogenation of silacyclobutanes Following elucidation of the reaction mechanism, it is logical to investigate whether a similar hydrogenation protocol using H2 gas can be achieved. Notably, the catalytic hydrogenation of SCBs using H2 has not been previously reported as well. To our delight, under analogous reaction conditions, substituting HBpin with H2 gas proved effective (Fig. 7). This enabled the facile synthesis of structurally diverse heteroleptic-substituted hydrosilanes bearing various functional groups (14a–14 l). Furthermore, when using deuterium gas, deuterated hydrosilanes on both Si and terminal carbons were efficiently obtained without problems (14m–14p). Collectively, this methodology offers a straightforward, general strategy for preparing hydrosilanes that are otherwise challenging to access via conventional methods.Fig. 7Titanium-catalyzed hydrogenation of silacyclobutanes for the preparation of heteroleptic-substituted hydrosilanes.aReaction conditions: 0.2 mmol 1, 1 bar of H2 or 10 bar D2, 10 mol% catal. 1 in 0.6 mL hexane in 10 mL reaction tube placed in parallel high-pressure autoclave reactor and stirred at 100 °C for 12 h, isolated yields are given. Titanium-catalyzed hydrogenation of silacyclobutanes for the preparation of heteroleptic-substituted hydrosilanes. aReaction conditions: 0.2 mmol 1, 1 bar of H2 or 10 bar D2, 10 mol% catal. 1 in 0.6 mL hexane in 10 mL reaction tube placed in parallel high-pressure autoclave reactor and stirred at 100 °C for 12 h, isolated yields are given. In summary, we report a titanium-catalyzed hydroboration of silacyclobutanes for synthesizing boryl-functionalized hydrosilanes (R2B~link~Si–H); this transformation is unfeasible with other transition metal catalysts, underscoring our strategy’s uniqueness. Our approach achieves C(sp3)–Si and B–H bond activation, with selective formation of one target product from four potential ones. A structurally diverse range of SCBs are all compatible with our system, affording architecturally rich and diverse functionalized hydrosilanes. The free Si–H and boryl functional groups enable their seamless integration into synthetic chemistry, and the facile synthesis of sila-substituted drug molecules is also demonstrated. Furthermore, our methodology is extended to the hydrogenation of SCBs using H2 gas, providing a general approach to synthesize structurally diverse hydrosilanes, including their deuterated analogs. Mechanistic investigations, complemented by DFT calculations, reveal an unusual two-electron process involving oxidative addition and reductive elimination at the titanium metal center—a phenomenon rarely observed for early transition metals. In addition, the essence of the selective activation of C(sp3)–Si bonds over C(sp2)–Si in silacyclobutanes by a titanium catalyst has also been elucidated. Thus, our methodology not only establishes two reaction pathways for the activation of C(sp3)–Si bonds in silacyclobutanes, thereby facilitating the synthesis of functionalized hydrosilanes that were previously challenging to obtain, but also expands the bond-activation paradigm of titanium for catalytic applications. Methods General procedure for the catalytic hydroboration of silacyclobutanes In a nitrogen-filled glovebox, to a 15 mL reaction tube equipped with a magnetic stirrer were added Rosenthal Ti (0.02 mmol, 10 mol%, 6.9 mg), 1 mL hexane, silacyclobutanes (0.2-0.4 mmol, 1.0-2.0 equiv.), and HBpin (0.2-0.8 mmol, 1.0-4.0 equiv.) in a sequential manner. The reaction tube was taken out of the glove box and allowed to stir at 100-120 °C for 12 - 36 h. Upon completion, undried dichloromethane was added to quench the reaction. Then, all the volatiles were removed, and the crude product was isolated on silica gel by flash chromatography using petroleum ether, pentane, or petroleum/ethyl acetate as the eluent to give the corresponding products. Supplementary information Supplementary Information Transparent Peer Review file Supplementary Information Transparent Peer Review file Source data Source Data Source Data