Electrophotocatalytic Si–H Activation Governed by Polarity-Matching Effects

Abstract

Open AccessCCS ChemistryRESEARCH ARTICLE1 May 2022Electrophotocatalytic Si–H Activation Governed by Polarity-Matching Effects Yangye Jiang, Kun Xu and Chengchu Zeng Yangye Jiang Faculty of Environment and Life, Beijing University of Technology, Beijing 100124 Google Scholar More articles by this author , Kun Xu *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] Faculty of Environment and Life, Beijing University of Technology, Beijing 100124 Google Scholar More articles by this author and Chengchu Zeng *Corresponding authors: E-mail Address: [email protected] E-mail Address: [email protected] Faculty of Environment and Life, Beijing University of Technology, Beijing 100124 Google Scholar More articles by this author https://doi.org/10.31635/ccschem.021.202101010 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Trialkylsilanes are important building blocks in organic synthesis; however, their widespread use in redox chemistry is limited by their high oxidation potentials and comparably high bond dissociation energies (BDEs) of Si–H and α–Si–C–H bonds (>92 kcal mol−1). Herein, we report a new strategy for Si–H bond homolysis enabled by the synergistic combination of electrooxidation, photoinduced ligand-to-metal charge transfer (LMCT), and radical-mediated hydrogen atom transfer (HAT). Governed by the polarity-matching effect, the HAT to electrophilic MeO· or [Cl-OHCH3]· from the more hydridic Si–H instead of a C–H bond allows the selective generation of silyl radicals. This electrophotocatalytic protocol provides rapid access to Si-functionalized benzimidazo-fused isoquinolinones with broad functional-group compatibility. Mechanistic studies have shown that n-Bu4NCl is essential to the electrooxidation of CeCl3 to form the Ce(IV) species. Download figure Download PowerPoint Introduction Organosilanes are of great interest in the fields of medicinal chemistry,1,2 organic synthesis,3,4 and organic electronics.5,6 Currently, the most attractive approach for Si-incorporation involves the interception of silyl radicals with alkenes or heterocycles. In this regard, the homolytic cleavage of liable Si–X (X = Si,7,8 B,9 COOH,10 etc.11–13) bonds has been identified as a powerful tool to obtain silyl radicals. A more practical and atom economic alternative is the homolysis of Si–H bonds in silicon hydrides. The classical activation of Si–H bonds to form silyl radicals relies on the combination of stoichiometric peroxide with an initiator (Scheme 1a).14–18 However, limitations that include harsh conditions and functional group incompatibility have driven the identification of new strategies for Si–H bond activation. Recently, photoredox catalysis19–28 has emerged as an appealing alternative for silyl radical formation from silicon hydrides. However, the substrates were invariably limited to phenylsilanes or (TMS)3SiH with labile Si–H bonds (Scheme 1b).11,29–35 As pioneered by Fagnoni et al.,36 decatungstate was found to be an effective hydrogen atom transfer (HAT) photocatalyst for aromatic tertiary silanes activation (Scheme 1c). Nevertheless, only poor selectivities were observed for trialkylsilanes owing to the comparably high bond dissociation energies (BDEs) of Si–H and α–Si–C–H bonds (>92 kcal mol−1).37 Wu and colleagues38,39 recently developed a mild and generally applicable platform for silyl radical formation using quinuclidin-3-yl acetate or triisopropylsilanethiol as the highly selective HAT reagent (Scheme 1d). Considering the valuable synthetic utilities of silyl radicals, the establishment of a practical and mechanistically alternative approach for the selective formation of silyl radicals from readily available trialkylsilanes remains highly desirable but synthetically challenging. Scheme 1 | (a–e) Representative strategies for Si–H activation to produce silyl radicals. Download figure Download PowerPoint Organic electrosynthesis utilizes the electron as a redox reagent to obtain reactive intermediates in a sustainable manner. Its application to carbon- and heteroatom-centered radicals generation has therefore been flourishing over the past decade.40–55 However, in comparison with carbon radicals, the electrochemical generation of silyl radicals remains largely unexplored. Recently, Lin and coworkers56 reported an elegant electroreductive approach to obtain silyl radicals from Si–Cl-containing compounds using Mg as a sacrificial anode. Aligned with our ongoing interest in organic electrosynthesis,57–59 we were intrigued by the possibility of electrochemically generating silyl radicals starting with trialkylsilanes. However, the direct electrooxidation of silicon hydrides to produce silyl radicals results in limited functional group compatibility and poor selectivity, mainly due to the high oxidation potentials of silicon hydrides.60 Since the HAT strategy allows the direct activation of substrates without the limitation of redox potentials,61,62 the synergy of electrooxidation and HAT provides new opportunities for the formation of silyl radicals at much lower operating potentials than those of analogous direct Si–H electrooxidations. Among many HAT reagents, we envision that MeO· should be ideal for the following reasons: (1) Hydrogen atom abstraction from an Si–H bond (BDE up to 96 kcal mol−1)63 by MeO· (BDE of O–H: 105 kcal mol−1)64 is thermodynamically favorable; (2) As hydrogen is more electronegative than silicon (electronegativity 2.2 vs 1.9 on the Pauling scale), the HAT triggered by electrophilic MeO· would preferably occur at more hydridic Si–H instead of C–H bonds according to the polarity-matching effect65; (3) The byproduct MeOH generated from the HAT process is easily removed from the reaction mixture. However, the success of this proposal hinges on the facile generation of energetically challenging MeO· from MeOH at a high oxidation potential to subsequently trigger HAT with trialkylsilanes to afford the desired silyl radicals. Motivated by the seminal works from Zuo66–71 who proved the direct generation of RO· from alcohols via cerium-catalyzed ligand-to-metal charge transfer (LMCT), we herein report a synergy that combines electrooxidation, photoinduced LMCT, and methoxyl-radical-mediated HAT for the selective generation of silyl radicals from silicon hydrides (Scheme 1e). Si–H bonds with high BDEs were activated through a polarized transition state. The synthetic utility of this electrophotocatalytic72–79 strategy is demonstrated by the rapid access to Si-functionalized benzimidazo-fused isoquinolinones with broad functional-group compatibility, which are prevalent in many biologically active molecules and advanced materials.80,81 Our working hypothesis is inspired by recent reports demonstrating the generation of alkoxy radicals via photoinduced LMCT excitation of Ce(IV)-OR complexes.66–70,82–89 As shown in Scheme 2, anodically generated Ce(IV) could coordinate with MeOH to afford transient Ce(IV)-OMe, which can undergo homolysis via photoinduced LMCT to give transient MeO·. Governed by the polarity-matching effect, HAT to electrophilic MeO · from the more hydridic Si–H in preference to a C–H bond allows the selective generation of silyl radicals. However, Si–H activation by in situ generated chlorine radical cannot be excluded at this moment.90 Subsequently, the silyl radical is intercepted by compound A to give radical B, which triggers a radical cyclization followed by Ce(IV) oxidation and proton release to deliver Si-functionalized benzimidazo-fused isoquinolinone E as the product. However, two main challenges must be solved to realize the mechanistic hypothesis. First, the MeO· should be generated at a relatively low concentration. Otherwise, the competition between Si–H and C–H cleavage during the HAT process would lead to poor selectivity. Second, the in situ generated Ce(IV)-OMe complex should undergo LMCT quickly to avoid its cathodic reduction. Scheme 2 | Mechanistic hypothesis. Download figure Download PowerPoint Experimental Methods General procedure for electrophotochemical synthesis: An undivided cell was equipped with a carbon felt anode (1.0 × 1.0 cm2) and a foamed nickel cathode (1.0 × 1.0 cm2) and connected to a direct current (DC) regulated power supply. To the cell was added 1 (0.3 mmol), 2 (1.0 mL), CeCl3·7H2O (0.06 mmol, 20 mol %), n-Bu4NCl (0.3 mmol), and CH3CN:cholorobenzene (5 mL, V/V = 1:1). The reaction cell was placed 3 cm away from the light-emitting diodes (LEDs) (390 nm, 20 W). The mixture was electrolyzed using constant current conditions (2 mA) at 50 °C under magnetic stirring. When thin-layer chromatography (TLC) analysis indicated that the electrolysis was complete (witnessed by the disappearance of the 1), the solvent was removed by distillation. The product was then extracted with dichloromethane (DCM) (3 × 20 mL), dried over Na2SO4, and concentrated in vacuo. The residue was purified by column chromatography on silica gel to afford the desired pure product. Further details including experimental procedures and additional data can be found in the Supporting Information. Results and Discussion Following the working hypothesis, the reaction parameters were established using 1a and triethylsilane ( 2a) as model substrates (Table 1). When the reaction mixture was electrolyzed in an undivided cell equipped with a carbon felt anode and a Ni foam cathode with CeCl3·7H2O as the catalyst, and n-Bu4NCl in MeOH/CH3CN/chlorobenzene as the electrolytic solution, while being irradiated using 20 W 390 nm LEDs at 50 °C for 9 h, the desired polycyclic product 3aa was obtained in 66% yield (entry 1). Replacing the anode with a graphite plate or changing the cathode to a Ni plate led to decreased yields (entries 2 and 3). The optimization of current density showed that 2 mA/cm2 was optimal (entries 1 vs 4 and 5). Reducing the amount of silane and Ce catalyst led to lower yields (entries 6 and 7). Anhydrous CeCl3 showed a similar reaction efficiency compared with CeCl3·7H2O (entry 8). The choice of MeOH as the precursor of HAT reagent had a significant influence on the chemical yield. Decreasing or increasing the amount of MeOH afforded lower yields (entries 9 and 10). Replacing chlorobenzene with other halogenated benzenes including PhCF3 and PhBr failed to yield the desired product (entries 11 and 12). Irradiation of the reaction mixture by 440 nm LEDs or white light reduced the yields to 49% and 31%, respectively (entries 13 and 14). A series of control experiments showed that cerium catalyst, MeOH, light irradiation, and current were all essential to the success of the transformation (entries 15–18), showcasing the synergistic cooperation between electrooxidation and photocatalysis. Table 1 | Optimization of Reaction Conditionsa Entry Deviation from Standard Conditions Yield (%)b 1 None 66 2 Graphite plate as anode 43 3 Ni plate as cathode 44 4 I = 1 mA 20 5 I = 3 mA 42 6 Et3SiH (0.7 mL) 37 7 CeCl3·7H2O (0.03 mmol) 31 8 Anhydrous CeCl3 as the catalyst 67 9 MeOH (0.05 mL) 39 10 MeOH (0.2 mL) 56 11 PhCF3 or PhBr instead of PhCl Trace 12 1,3-Dichlorobenzene instead of PhCl 0 13 440 nm LEDs 49 14 White light 31 15 No CeCl3 0 16 No MeOH 9 17 No light 7 18 No electricity Trace aReaction conditions: 1a (0.3 mmol), 2a (1 mL), CeCl3·7H2O (0.06 mmol), n-Bu4NCl (0.3 mmol), MeOH (0.1 mL) in CH3CN:chlorobenzene (1:1, 5 mL), graphite felt anode and foamed Ni cathode (working area: 1 cm2), undivided cell, I = 2 mA, 390 nm LEDs (20 W), 50 °C, 9 h. bIsolated yield. With optimal reaction conditions established, we began to examine the substrate scope of the silicon hydrides. As shown in Scheme 3, trialkylsilanes showed excellent selectivities affording the corresponding polycyclic products 3a– 3f up to 81% yield, while the competitive reaction at the C–H adjacent to silicon was not observed. Aromatic tertiary silanes tolerated optimal conditions well, affording Si-products 3g– 3i with up to 73% yield. When 1,4-bis(dimethylsilyl)benzene was employed as the substrate, only one of the Si–H bonds underwent the tandem radical reaction to furnish the corresponding product 3j in 39% yield. (TMS)3SiH, with its lower BDE, was shown to be a suitable substrate to generate polycyclic product 3k in 72% yield. However, trimethoxysilane failed to give product 3l probably due to the instability of the corresponding silyl radical. As compound 1a was unstable even without electrolysis, the decomposition of 1a was found to be a major side reaction for these transformations. Scheme 3 | The scope of hydrosilanes. Reaction conditions: 1a (0.3 mmol), 2a–2l (1 mL for liquid silicon hydrides, and 6 mmol for solid silicon hydrides), CeCl3·7H2O (0.06 mmol), n-Bu4NCl (0.3 mmol), MeOH (0.1 mL) in CH3CN: chlorobenzene (V/V 1:1, 5 mL), graphite felt anode and Ni foam cathode (working area: 1 cm2), undivided cell, I = 2 mA, 390 nm LEDs (20 W), 50 °C, 9 h; isolated yield. Download figure Download PowerPoint Having identified the silicon hydride substrate scope, we next focused on determining the scope of N-substituted benzimidazole derivatives. As shown in Scheme 4, a variety of N-substituted benzimidazoles underwent the tandem radical cyclization reactions smoothly to afford the Si-incorporated products 4– 20 in good to moderate yields. For the Ar1 moiety, the presence of electron-donating groups decreased the reaction efficiency compared with systems containing electron-withdrawing substituents ( 4 vs 5 and 6). For the Ar2 moiety, both electron-donating (OMe) and -withdrawing (CN) groups were well-tolerated ( 10 and 16). A series of functional groups including alkene ( 11), borate ( 13), ester ( 14), sulfonate ( 15), and nitrile ( 16) were all compatible with the optimal conditions, giving the corresponding products up to 67% yield. Replacing the R group from methyl to phenyl led to a lower yield ( 17). Notably, substrates derived from l-menthol, (–)-nopol, and diacetone-d-glucose were also well-tolerated to give products 18– 20 with up to 64% yield, thus showcasing the preparative potential of this electrophotocatalytic transformation. Scheme 4 | The scope of N-substituted benzimidazoles. Reaction conditions: 1 (0.3 mmol), 2a or 2c (1 mL), CeCl3·7H2O (0.06 mmol), n-Bu4NCl (0.3 mmol), MeOH (0.1 mL) in CH3CN:chlorobenzene (V/V 1:1, 5 mL), graphite felt anode and foamed Ni cathode (working area: 1 cm2), undivided cell, I = 2 mA, 390 nm LEDs (20 W), 50 °C, 9 h; isolated yield. Download figure Download PowerPoint In addition to N-substituted benzimidazoles, N-substituted indole was also a suitable substrate under optimal conditions, giving the corresponding polycyclic product 22 in 71% yield (Scheme 5). Scheme 5 | The construction of indole-based polycyclic product. Download figure Download PowerPoint To gain insight into the reaction mechanism, cyclic voltammetry (CV) experiments were conducted (Figure 1). First, substrates 1a and 2a showed similar oxidation potentials (curves b and c), which suggests that the direct oxidation of both compounds might proceed simultaneously. Second, CV of CeCl3·7H2O did not display obvious oxidation peak (curve e), while the mixture of CeCl3·7H2O with n-Bu4NCl showed an obvious new oxidation peak at 0.68 V (curve g). However, the mixture of CeCl3·7H2O with other tetrabutylammonium salts including n-Bu4NI, n-Bu4NBr, and n-Bu4NPF6 did not afford a new oxidation peak (see Supporting Information Figure S3–S5 for details). These results reveal that exogenous chloride was essential to the electrooxidation of Ce(III) to form Ce(IV) species that trigger the subsequent LMCT process needed for the generation of MeO· as the HAT reagent. This information coupled with the much lower oxidation potential of Ce(III) compared with that of Si–H containing reagents and the alkene moiety in 1a, leads us to the proposed electrophotocatalytic cycle shown in Scheme 2 starting with the oxidation of Ce(III) to Ce(IV). Figure 1 | CV of related compounds in 0.1 M LiClO4/CH3CN/PhCl/MeOH with Pt disk as the working electrode, Pt wire as the counter electrode, and Ag/AgNO3 (0.1 M in CH3CN) as the reference electrode: (1) Background, (2) 2a (5 mM), (3) 1a (5 mM), (4) background, (5) CeCl3·7H2O (3 mM), (6) n-Bu4NCl (3 mM), (7) n-Bu4NCl (9 mM) + CeCl3·7H2O (3 mM). Download figure Download PowerPoint To further demonstrate the generation of Ce(IV) species under electrooxidative conditions assisted by exogenous chloride, UV–vis experiments were carried out. As shown in Figure 2, the mixture of CeCl3·7H2O and n-Bu4NCl in CH3CN/MeOH showed an absorption band with λmax = 250 nm. After the electrolysis of this mixture for 20 min, a new absorption band appeared with λmax = 395 nm. These results are consistent with our experimental evidence that the highest reaction efficiency was observed with irradiation at 390 nm. However, when n-Bu4NCl was replace by n-Bu4NPF6, no absorption shift was observed after electrolysis (see Supporting Information Figure S7 for details). Figure 2 | UV–vis absorption spectra of a solution of the 1:1 mixture of CeCl3·7H2O and n-Bu4NCl in CH3CN/MeOH before (black line) and after electrolysis (red line). Download figure Download PowerPoint Conclusions We have developed an electrophotocatalytic platform by synergistically combining electrooxidation, cerium-catalyzed LMCT, and radical-mediated HAT for the selective generation of silyl radicals from Si–H bonds with high BDEs. This synergistic scenario makes use of electrons as clean oxidants and an earth-abundant cerium salt as the photocatalyst to induce the formation of MeO· or [Cl-OHCH3]· from readily available MeOH via LMCT, which in turn acts as an electrophilic HAT reagent to activate more hydridic Si–H bonds instead of C–H bonds to selectively produce silyl radicals according to the polarity-matching effect. This strategy provides a mild and generally applicable entry to diversified Si-functionalized benzimidazo-fused isoquinolinones. The much lower operating potential of this methodology compared to that of direct Si–H oxidation provides the basis for broad functional group compatibility. Mechanistic studies have showed that n-Bu4NCl is essential to the electrooxidation of CeCl3 to form the Ce(IV) species. The application of this strategy for the functionalization of other inert substrates is currently under investigation in our laboratory. Supporting Information Supporting Information is available and includes general experimental considerations, detailed procedures, spectral characterization, cyclic voltammograms, and additional data. Conflict of Interest There is no conflict of interest to report. Funding Information This work was supported by grants from the National Key Technology R&D Program (no. 2017YFB0307502), the National Natural Science Foundation of China (no. 21871019), and Beijing Municipal Education Committee Project (nos. KZ202110005003 and KM202110005006).