Sulfoximines, isosteres of sulfones, are valuable motifs in many scenarios [1-3]. Besides the traditional utilization as catalysts and ligands in synthetic chemistry [4-9], sulfoximines are also intensively investigated in medicinal chemistry as key pharmacophores (Scheme 1a, left) [10-14]. In addition, studies on sulfoximines have also branched out into materials science. For example, Takata and Bolm introduced sulfoximine units into polymers (Scheme 1a, middle) [15,16]; Yang et al. designed a sulfoximine unit as a new acceptor for thermally activated delayed fluorescence (TADF) emitters featuring tunable electron-accepting ability (Scheme 1a, right) [17,18]. What goes beyond sulfones is that sulfoximines can conveniently regulate physicochemical properties through N-H functionalization. Synthetic protocols for N-functionalization of sulfoximines have been widely explored, such as N-phosphorylation, acylation, halogenation, alkynylation, alkylation, arylation [19-22]. Despite vast established methods for N-H functionalization, N-silylation of sulfoximines has been less explored. Traditionally, chlorosilanes or silylamine reagents are used to conduct silylation of sulfoximines, which usually have problems due to the hygroscopically sensitive starting materials [23-27]. In 2015, Cheng and co-workers reported a direct oxidative cross-coupling reaction towards N-silylated sulfoximines via a copper-catalyzed N-silylation of sulfoximines with various 3° silanes in the presence of stoichiometric DTBP (Scheme 1b) [28].

To the best of our knowledge, the enantioselective N-silylation of sulfoximine is still unknown to date. In fact, there is currently a lack of an efficient catalytic asymmetric synthesis to access Si-stereogenic silylamine compounds. Few attempts could provide Si-stereogenic silylamine derivatives with enantiomeric excess of around 30% after stabilization via a one-pot deprotonation/boronation procedure [29,30]. The challenges may lie in the following aspects: (a) Si-N bond is usually unstable due to the high moisture sensitivity of silylamines [31-34] (b) a longer Si-N bond (1.74 Å) compared with C-N bond (1.47 Å) would prevent the formation of compact transition states, making it difficult to achieve high enantiocontrol [35-38]. Given the unique properties and versatile utilities of both sulfoximines and Si-stereogenic silanes, the development of efficient enantioselective N-silylation of sulfoximine is highly demanded.

Owing to the electrostatic attractions between sulfur and nitrogen, as well as the severe lone-pair lone-pair repulsions between oxygen and nitrogen, NH in sulfoximine is acidic and the silylated Si-N bond is relatively stable for separation and identification [39]. Meanwhile, during the past few years, a handful of elegant methods have been developed for constructing Si-stereogenic centers [40-48], especially the direct desymmetrization of dihydrosilanes via catalytic asymmetric dehydrogenative coupling towards Si-stereogenic compounds (Si-CADC) [49-51]. Herein, we report a rhodium/diphosphine-catalyzed enantioselective N-silylation of sulfoximines. This protocol gives access to a variety of Si-stereogenic N-silylated sulfoximines in decent yield (up to 99%) with excellent stereoselectivity (up to 99%), featuring high atom economy, and a cleaner manner with H₂ as the sole byproduct (Scheme 1c).

At the outset of our studies, we examined this cross-coupling reaction between tert-butylphenylsilane 1a and diphenylsulfoximine 2a using copper or rhodium as the catalyst with chiral phosphine ligands. Under copper catalysis, we could obtain good yields of the targeted cross-coupling product 3aa, albeit with poor stereocontrol (Table 1, entries 1 and 2). When we moved on to rhodium catalyst, an obvious increase of enantioselectivity was observed using [Rh(cod)Cl]₂ combined with BINAP L2 as the catalyst and chiral ligand respectively (Table 1, entry 4). To our delight, when Josiphos L5 was applied with [Rh(cod)Cl]₂ in this reaction, both the yield and enantioselectivity could be obtained from moderate to good (Table 1, entry 7). Further switching the substituted groups on the phosphorus atoms of the Josiphos ligand, almost quantitative yield and enantiomeric excess (ee) were obtained (Table 1, entry 8). Under the rhodium-catalyzed conditions, other diphosphine ligands such as Ph-BPE L1, Segphos L3, and QuinoxP* L4 gave barely stereocontrol with low to moderate yields (Table 1, entries 3, 5 and 6). More polar solvents such as THF, and DCE did not show a positive impact on the reaction (Table 1, entries 9 and 10). It is worth mentioning that, the reaction could proceed smoothly without the use of any ligand under [Rh(cod)Cl]₂-catalyzed conditions with a high yield (Table 1, entry 11). In summary, the best reaction conditions we identified is that starting material dihydrosilane 1a and sulfoximine 2a could react with each other to give desired coupling product 3aa with both quantitative yield and ee in the presence of [Rh(cod)Cl]₂ catalyst with Josiphos L6 as chiral ligand at room temperature in the solvent of toluene.

With the optimized conditions in hand, next we investigated the substrate scope to establish a general methodology for the construction of enantioenriched N-silylated sulfoximines (Scheme 2). First, functional group tolerance on the aromatic ring of dihydrosilane substrates was tested. Both electron-donating methyl (1b), methoxy (1c), trimethylsilyl (1d), and electron-withdrawing fluorine (1e), chlorine (1f) groups all proceeded robustly with diphenylsulfoximine 2a to afford the corresponding silicon-stereogenic sulfoximine products 3ba-3fa in good yields (77%-92%) with excellent enantioselectivities (95%-97% ee). Extended aromatic rings such as 1-naphthyl (1g), and 2-naphthyl (1h) groups were also well accommodated in this transformation. This catalytic system is also suitable for heterocycles such as benzothiophene (1i), indole (1j), and benzodioxole (1k), which provided the corresponding chiral N-silylated sulfoximine products with excellent yields (89%-94%) and stereoselectivities (95%-96% ee). Morpholine (1l) and triphenylamine (1m) units which frequently appeared in material molecules as donors also proceeded smoothly, and afforded the target products 3la and 3ma with almost quantitative yields and ee, indicating a promising application potential of this methodology in material area. Nonaromatic tert-butyl dihydrosilane substrate 1n was submitted to the reaction, however, only 54% ee of the product 3na was obtained. Ortho-substituted phenyl cyclohexyl dihydrosilanes maintained to give products 3oa-3qa with good ee around 83%, while phenyl cyclohexyl dihydrosilane could only yield desired product 3ra with 66% ee. Replacing the tert-butyl group with isopropyl, cyclohexyl, and 2,5-dimethylphenyl groups reduced the stereoselectivity comparing product 3aa with 3ra-3ta. Meanwhile, a range of diaryl sulfoximines 2b-2i bearing electron-donating and withdrawing groups were tested in the reaction. The corresponding enantioenriched N-silylated sulfoximines 3ab-3ai could be generated in decent yields with excellent ee. In addition, cyclic sulfoximine 2j also went through the reaction smoothly without a significant drop in efficiency. Non-symmetric sulfoximines 2k and 2l functioned well under the catalytic conditions, but without the control of the sulfur-stereogenic center. Changing the phenyl substituent to benzyl group on sulfoximine (2m) reduced both the reactivity and enantioselectivity.

To illustrate the utility of this methodology, we designed a sulfoximine substrate 2n featuring a 9,9-dimethyl-9,10-dihydroacridine unit as a typical donor (D) and a sulfoximine unit itself as an elegant acceptor (A). Under the standard reaction conditions, the corresponding enantioenriched N-silylated sulfoximine product 3an was obtained in 91% yield with 97% ee. Molecular modeling of 3an by DFT revealed an almost perpendicular D-A conformation with a dihedral angle of 85.7° in the ground state (Fig. 1). Efficient spatial separation of HOMO and LUMO distribution in the ground state, together with a small singlet-triplet energy gap (ΔE_ST) of 0.02 eV were disclosed by theoretical calculation. These properties showed a potential possibility for the entitled product to find utility in TADF materials [52-56]. Moreover, a microsecond level of fluorescent lifetime τ_F = 2.08 µs for 3an in CH₂Cl₂ (10⁻⁵ mol/L) was detected at room temperature (Section 6 in Supporting information). The photophysical properties of 3an elucidate that the UV-vis absorption maxima and emission maxima of 3an are around 284 and 504 nm, respectively, thus featuring a large Stokes shift of 220 nm. Meanwhile, Φ_PL of 3an in CH₂Cl₂ (10⁻⁵ mol/L) at room temperature was measured to be 15% by integrating sphere.

To extend the scope of this enantioselective N-silylation of sulfoximines, we evaluated the reaction between bis-dihydrosilanes 1u, 1v, and sulfoximine 2a. Under the standard conditions, the enantioenriched bis-silicon-stereogenic sulfoximine 3ua and 3va could be successfully obtained by the double dehydrogenative coupling reaction. Notably, the bis-silicon-stereogenic sulfoximine featuring Si-H bonds can readily be utilized as the chiral monomer for the polymerization reaction. Given the important application of sulfoximine unit, the direct transformation of bis-silicon-stereogenic sulfoximine 3va into its corresponding polymer 4va was conducted (Scheme 3). When 3va and 1,4-bis(dimethyl(vinyl)silyl) benzene were charged with Karstedt catalyst in THF at 120 ℃, the hydrosilylation polymerization proceeded efficiently providing the chiral polymer 4va in 75% yield with number-average molecular weights M_n = 7.3 kg/mol, and small polymer dispersity index (PDI = 1.1). This chiral organosilicon polymer featuring sulfoximine pendants could be highly attractive to organic material chemists [57-61].

In summary, we have developed the first enantioselective N-silylation of sulfoximines via a rhodium/diphosphine-catalyzed asymmetric dehydrogenative cross-coupling, which gave access to a variety of chiral N-silylated sulfoximines featuring versatile Si-H bonds in good yields with excellent enantioselectivities. Sulfoximine featuring a D-A structure which frequently appeared as functional material fragments proceeded in this reaction smoothly. Moreover, chiral sulfoximine-containing polymer could be accessed conveniently from the enantiopure bis-Si-stereogenic N-silylated sulfoximine. We believe that this operational benign and effective methodology for constructing silicon-stereogenic N-silylated sulfoximines will embrace potential applications in different scenarios.

The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.

We are grateful for financial support from the National Natural Science Foundation of China (Nos. 22122102, 22101120, 22271134), Guangdong Provincial Key Laboratory of Catalysis (No. 2020B121201002), Guangdong Pearl River Talent Program (No. 2019QN01Y628), Shenzhen Science and Technology Innovation Committee (No. RCJC20221008092723013), and Key Laboratory of Organosilicon Chemistry and Material Technology of Ministry of Education (No. KFJJ2022012). The theoretical calculation part is supported by center for computational science and engineering at Southern University of Science and Technology. The photophysical data were obtained using equipment maintained by Southern University of Science and Technology Core Research Facilities.

Supplementary material associated with this article can be found, in the online version, at doi:10.1016/j.cclet.2023.108783.