Construction of sulfur stereocentres by asymmetric geminate recasting

Abstract

Radical pairs generated by light- or heat-induced bond cleavage play a central role in biochemical transformations and the synthesis of pharmaceuticals, polymers, and industrial chemicals. When such cleavage occurs at a stereocentre in chiral molecules, recombination of the produced radicals can lead to either enantiomer, typically resulting in racemization. Achieving selective conversion of racemic mixtures into single enantiomers is highly desirable yet challenging, due to the uncontrolled behaviour of free radicals. Here we show that stereocontrol over these reactions can be achieved through asymmetric geminate recasting—a process in which homolysis and recombination occur within a solvent cage under the influence of a chiral photocatalyst. This strategy enabled the selective construction of chiral sulfur stereocentres via deracemization of sulfinamides, providing access to valuable sulfur-containing building blocks. The approach opens unexplored possibilities for controlling stereochemistry in radical reactions and may inspire broader applications in asymmetric synthesis, medicinal chemistry, and materials development.

Homolytic cleavage of a single bond in the most general form produces two radicals that are no longer bound to each other and that, having attained translational freedom, can react independently as separate chemical entities. In solution, the dissociation process is not instantaneous, and the radicals can remain in proximity of each other, that is, as a geminate radical pair, because surrounding solvent molecules (the solvent cage) prevent their immediate separation. These fundamental reactions underlie many modern synthetic methods, industrial production of polymers and feedstock chemicals and an array of biological processes ^1,2,3,4,6 Of the many reactions that the two radicals can undergo, recombination of the geminate radical pair stands out because it reconstitutes the progenitor molecule by reforming the previously cleaved covalent bond (Fig. 1a). As such, geminate recombination is a common side reaction in synthetic radical chemistry and polymerization that may lead to diminished reaction yields and radical initiation efficiencies^1,7.

Fig. 1. Asymmetric geminate recasting and photocatalytic deracemization of sulfinamides.

a, Formation of independently reacting radicals after homolytic cleavage of a covalent bond and regeneration of the progenitor molecule by recombination of the radical pair. b, Asymmetric geminate recasting promoted by a chiral photocatalyst as a platform for deracemization to circumvent the kinetic limitations of out-of-cage radical pair reactions. k_r and k_s are rate constants for the out-of-cage recombination and other side reactions. c, Scarcity and roles of sulfur chirality in life and materials sciences, synthesis and catalysis. d, Construction of synthetically and medicinally valuable sulfur(IV) and sulfur(VI) stereocentres by photocatalytic deracemization of sulfinamides based on asymmetric geminate recasting. e, Reaction development and survey of structural effects on the photocatalytic activity in deracemization of sulfinamides. PC, photocatalyst.

Geminate recasting, which comprises a sequence of homolysis and geminate recombination, is also a well-documented and undesirable pathway for racemization of chiral compounds. The loss of optical purity is caused by the formation of planar or configurationally labile radicals that can react at either of the prochiral faces^1,8,9,10 While the radical process requires heat or light as an external stimulus for bond homolysis, racemization is intrinsically thermodynamically favourable because the entropy of a racemic mixture is higher than that of pure enantiomers.

Given the ubiquity of geminate recasting across a variety of covalent compounds^1,5, imparting enantiocontrol to this fundamental process and thus countering the thermodynamic preference for racemization could unlock a broadly generalizable strategy for producing single enantiomers from racemic mixtures. To be successful, the approach would necessitate a chiral catalytic system that could divert the reaction to one of the enantiomers through involvement in both phases of geminate recasting and before the two radicals diffuse away from each other (Fig.1.b). Furthermore, geminate recombination is substantially more efficient than an out-of-cage encounter of escaped reactive radicals that are present at very low concentrations and that can undergo a variety of deleterious reactions independently of each other^4,5. In this process, the solvent cage restricts the separation of the nascent radical pair, thus favouring the catalyst-controlled enantioselective recombination over the inefficient and unselective reaction of free radical pairs. A reaction design based on catalytic geminate recasting would thus circumvent the limitations of conventional catalytic approaches that rely on random encounters of reacting partners in the bulk of a solution.

Such catalytic reaction design could also facilitate homolysis by harnessing an appropriate external stimulus. To this end, light-driven reactions may be able to provide the energy for homolytic cleavage at ambient and even cryogenic temperatures, which generally increases the efficiency of geminate recombination⁴. Additionally, the involvement of photochemically accessible excited states in the deracemization process can circumvent the limitations imposed by microscopic reversibility if the bond cleavage and radical recombination steps occur in different electron and spin states¹¹. Collectively, these considerations indicate that a suitable chiral photocatalytic system capable of promoting the homolysis and recombination phases could enable efficient deracemization. This premise has been supported by recent pioneering studies on efficient deracemization reactions mediated by chiral photocatalytic systems^11,12. The mechanistic manifolds that underlie the asymmetric photocatalysis in these systems rely on several key processes that drive deracemization. For example, photosensitized isomerization of double bonds in triplet excited states enabled deracemization reactions of allenes and alkenes that were pioneered by Bach^13,14,15,16, Dang and Yu¹⁷ and Luo^18,19. Photocatalytic systems based on electron, proton, and hydrogen atom transfer processes were also reported by Miller and Knowles²⁰, Meggers²¹, Bach^22,23,24, and Ye²⁵. Additionally, photocatalytic approaches based on ring-opening reactions were developed by Bach and Gilmour ^26,27,28,29.

Importantly, current photocatalytic deracemization strategies leverage transformations that preserve the structural unity of the substrate throughout the deracemization process, as exemplified by triplet excited states in the double bond isomerization of allenes^{13,14,15,16,17,18} and biradical species derived from cyclopropanes^28,29. The core structures of the substrates also remain intact during proton or hydrogen transfer^20,21. In contrast, attainment of translational freedom by two unbound radicals upon substrate homolysis is a notable challenge in the development of an efficient deracemization strategy because of the complexity and high rates of the subsequent independent reactions that the two radicals can undergo.

We posited that the asymmetric geminate recasting concept could be validated by addressing the synthetically important problem of catalytic construction of sulfur stereocentres. To this end, we were inspired by insightful mechanistic investigations by Cram and Booms that indicated that homolysis of the S–N bond drives the thermal racemization of sulfinamides³⁰ Notably, they reported that a secondary aromatic amino group on the S(IV) stereocentre predisposed the sulfinamide to racemization via an in-cage dissociation–recombination pathway. This observation is consistent with the relative facility of the S–N bond cleavage (bond dissociation free energy of 21.7 kcal/mol for 1a), which makes this pathway more energetically favourable than the high-barrier inversion of the sulfur stereocentre (ΔG^≠ = 44.5 kcal/mol). Since thermal homolysis leads to a singlet radical pair, in-cage radical recombination is kinetically facile, as it does not require a spin transition. This type of in-cage recombination of thermally generated radical pairs was recently successfully employed by Ha-Cheong and Smith to control the selectivity of [1,2]-rearrangements of allylic ammonium ylides³¹. On the other hand, photocatalytic homolysis would likely produce a triplet radical pair that would require triplet–singlet interconversion (spin transition) and could render geminate recombination less efficient. However, previous studies have indicated that large spin–orbital coupling in radical pairs containing a sulfinyl radical can efficiently promote triplet–singlet interconversion³². Furthermore, a suitably designed photocatalytic system could also leverage the effect of heavy atoms that accelerate spin transition and can increase the efficiency of recombination³³.

The selection of sulfinamides as substrates is also advantageous from a synthesis perspective. Sulfur stereocentres display diverse substitution patterns and oxidation states and hence provide opportunities for fine-tuning molecular properties and expanding the chemical space of drug candidates, catalysts, and reagents (Fig 1.c)^{34,35,36,37,38,39,40} Despite the medicinal and synthetic potential, broader exploration of sulfur chirality is severely limited by the dearth of generalizable asymmetric strategies for accessing chiral sulfur functional groups^41,42. The S(IV) stereocentre in sulfinamides can serve as a synthetically versatile precursor to a wide range of other chiral S(IV) and S(VI) functionalities. However, enantioselective synthesis of chiral sulfinamides remains underdeveloped and continues to rely on chiral stoichiometric reagents or multistep derivatization of other sulfur stereocentres^37,41. In light of these constraints, catalytic approaches to chiral sulfinamides have emerged at the forefront of synthetic methodology, underscoring both the urgency and challenges of establishing enantioselective synthetic avenues to this unexplored class of chiral functional groups^{43,44,45,46,48,49}. Significantly, current catalytic strategies for accessing chiral sulfinamides predominantly rely on ground-state reactions of closed-shell intermediates.

Homolytic cleavage produces pairs of radicals that can undergo independent deleterious reactions and can prevent efficient recombination, presenting a challenge for catalytic deracemization. Here, we show how asymmetric geminate recasting can be used to circumvent this challenge and address the synthetically important problem of producing sulfur chirality. The approach leverages the reactivity of photocatalytically generated sulfur- and nitrogen-centred open shell species for deracemization of sulfinamides (Fig 1.d) and provides a blueprint for further exploration of the framework in asymmetric synthesis.

Results

Our initial survey of various chiral Lewis acids identified indium(III) bromide complex In-1 as a singularly efficient photocatalyst for the light-driven deracemization of sulfinamide 1a (Fig. 1.e and Supplementary Tables S1 and S2). The enantioselectivity was improved at lower temperatures, and optimum performance was achieved at −55 °C. Under optimal conditions, irradiation of the reaction with violet LED (light-emitting diode) light afforded the product in 80% yield and 94% e.e. The catalytic activity of In-1 is unique among the many surveyed metal complexes; all three structural elements—indium, bromine, and the salen ligand—proved to be essential for achieving high enantioselectivity. For example, neither salen complexes of an array of other main group and transition metals nor complexes with other chiral ligands demonstrated any catalytic activity. Replacement of the bromine atom with the lighter chorine was likewise detrimental to the enantioselectivity of the reaction. The chiral diaminocyclohexane fragment and the proximal ortho tert-butyl groups in the salen ligand were also indispensable, whereas the distal para tert-butyl groups had a smaller effect on the enantioselectivity.

We next explored the substrate scope of this method (Table 1). A variety of sulfinamides bearing substituents on the aromatic ring of the amino group proved to be suitable substrates. Fluoro-substituents, chloro-substituents and bromo-substituents in the ortho, meta and para positions were well tolerated, and the corresponding sulfinamides 1b, 1c, 1f–1h, 1o, 1p, 1s and 1t were produced in 72–90% yields and 91–98% e.e. The sterically more demanding ortho-biphenyl group (1e) and the medicinally valuable trifluoromethyl and trifluoromethoxy groups (1d, 1i, 1j) were also readily accommodated, while also demonstrating the feasibility of a gram scale synthesis ((S)-1i). The cyano, ester, keto, alkyne, methoxy and tertiary amino groups were likewise compatible with the reaction conditions, affording products 1k–1n, 1q and 1r with enantioselectivities in the range of 90–96% e.e. In addition, the heteroaromatic pyrazine ring was equally well tolerated (1u). Furthermore, substrates bearing other functionalities on the nitrogen atom, including alkene (1v, 1w) and ester (1x, 1y) groups, as well as cyclic sulfinamide 1z, underwent efficient deracemization with high enantioselectivities.

Table 1. Scope of substituted sulfinamides.

All yields are of isolated products. The experimental details are provided in the supplementary materials. Reactions were carried out with ent-In-1 for 1r, 2i, 2s, 3b, 3c and the gram scale for 1i. Reaction with 1r was conducted with 12 mol% ent-In-1. The absolute configurations of 1t, 1i, and 2o were determined by X-ray crystallographic analysis. The absolute configurations for all other products were assigned by analogy.

Substituents in the aromatic ring on the sulfur atom were also equally well tolerated, including halogens (2a–2d, 2j, 2k, 2m), trifluoromethoxy and difluoromethoxy groups (2e, 2f) and trifluoromethyl (2g), phenyl (2h), cyclohexyl (2i) and methoxy (2l) groups. In addition, benzothiophene-derived sulfinamide 2n was produced with high enantioselectivity. Finally, sulfinamides 2o–2u, which feature substituted aromatic and heteroaromatic groups on both the sulfur and the nitrogen, were obtained with 92–98% e.e. To further demonstrate the utility of the reaction, we applied it to a series of biomedically relevant substrates of greater structural complexity (Table 1). For example, a sulfinamide derivative of the antidiabetic drug empagliflozin 3a was produced in a good yield and 90% diastereomeric access (d.e.). Modified versions of the nonsteroidal anti-inflammatory drugs flurbiprofen (3b) and deracoxib (3c) proved to be equally suitable substrates, furnishing products with 93% d.e. and 94% e.e., respectively.

We next aimed to examine if deracemization products could be converted to other valuable chiral sulfur functionalities by stereospecific substitution and atom transfer reactions (Fig. 2.a). The N-alkylaniline group in the sulfinamides was efficiently displaced with nucleophilic Grignard reagents, affording sulfoxide derivatives of empagliflozin (4) and celecoxib (5) with excellent enantiospecificity. Furthermore, the secondary amino group in the antihistamine drug desloratadine was amenable to enantiospecific reaction with sulfinamide (S)-2b under basic conditions, affording corresponding derivative 6 in a good yield. Sulfonimidamides featuring a S(VI) stereocentre were likewise readily generated from sulfinamide products (Fig. 2.b). For example, iron-catalysed nitrene transfer⁵⁰ afforded sulfonimidamide 7 with excellent enantiospecificity. On the other hand, a sulfonimidamide analogue of the antidepressant drug sertraline 8 with an unprotected NH group was readily accessed by substitution of the aniline residue in sulfinamide (S)-2a and subsequent imino group transfer⁵¹ to the sulfur stereocentre. Other chiral S(VI) functionalities could also be introduced using sulfinamide products as sulfinyl group donors (Fig. 2.b). For example, sulfinamide (S)-1i was converted to sulfonimidate derivative of the antiarthritis drug oxaprozin 9 by a two-step sequence that entailed swapping the amide group in a reaction with n-heptylamine and subsequent coupling with alcohol 10 mediated by tert-butyl hypochlorite⁴⁸ under basic conditions. To illustrate the utility of sulfinamides as precursors to chiral sulfoximines, sulfinamide 2s was converted to cyclin-dependent kinase 9 (CDK9) inhibitor 11, which has previously been synthesized only as a racemate⁵². The concise enantiospecific synthesis entailed nucleophilic substitution of the aniline group in sulfinamide 2s with a Grignard reagent and a Rh-catalysed conversion of the intermediate sulfoxide to sulfoximine⁵³. The subsequent Pd-catalysed arylation of aniline 12 afforded sulfoximine 11 with >99.9% e.e. and a 51% overall yield after single recrystallization.

Fig. 2. Synthetic applications of sulfinamide products.

a, Conversion to sulfoxides and other sulfinamides by nucleophilic substitution of the aniline moiety. b, Construction of medically important chiral sulfur(VI) functionalities, including sulfonimidamide (7 and 8), sulfonimidate (9) and sulfoximine (11) groups. The structures of 7 and 8 were confirmed by X-ray crystallographic analysis. c, Examples of archetypical chiral ligands and catalysts featuring sulfur stereocentres and their synthesis from sulfinamide (S)-1i. The experimental details are provided in Supplementary Information. e.s., enantiospecificity; Ac, acetyl; ⁿBu, n-butyl; ^tBu, tert-butyl; Ph, phenyl; Ts, p-toluenesulfonyl; KHMDS, potassium bis(trimethylsilyl)amide; Xantphos, 4,5-bis(diphenylphosphino)-9,9-dimethylxanthene.

Given the growing importance of chiral sulfur compounds in asymmetric catalysis⁵⁴, we sought to expand the utility of the reaction to the synthesis of representative chiral ligands and catalysts (Fig. 2.c). To this end, a sulfur stereocentre could be successfully installed in ferrocene 13 by directed ortho metalation and subsequent sulfinyl group transfer with sulfinamide (S)-1i. The dimethylamino group in the side chain was then converted to a phosphine. The synthetic sequence furnished a prototype of hitherto unexplored ferrocene-based chiral sulfoxide phosphine ligands (14) that are structurally related to the widely used Josiphos diphosphines⁵⁵ and feature a sulfur stereocentre in addition to the ferrocene chiral plane and the stereogenic side chain. Moreover, a sequence of a base-promoted amino group exchange with diamine 15 and a reaction with isothiocyanate 16 provided simple entry to thiourea 17. This derivative is representative of hydrogen bond donor catalysts with a pendant chiral sulfinamide group that is essential for enabling highly enantioselective transformations⁵⁶.

Taken together, these results demonstrate that the substitution of the aniline group in the deracemization-derived sulfinamide products proceeds with high stereospecificity and can facilitate access to a variety of chiral sulfur(IV) and sulfur(VI) functional groups.

Prompted by the observation of high enantioselectivity imparted by indium catalyst In-1, we sought to elucidate the mechanistic basis of its photocatalytic activity in the light-driven deracemization of sulfinamides. UV–Vis absorption profiles of sulfinamide 1a and catalyst In-1 at reaction ratios indicated that In-1 is the primary light-absorbing species in the violet LED light range centred at 400 nm (Fig. 3.a). Consistent with the postulated role of In-1 as a photocatalyst, deracemization was observed only when the reactants were irradiated with light (Fig. 3.b). Taken together, these results indicate that the deracemization reaction proceeds via a photocatalyst In-1-mediated pathway. Accordingly, we posited that photoexcitation of a complex of catalyst In-1 and sulfinamide could provide sufficient energy for the cleavage of the weak sulfur–nitrogen bond and produce a geminate sulfinyl–aminyl radical pair.

Fig. 3. Mechanistic and computational studies of the photocatalytic deracemization of sulfinamides.

a, UV–vis absorption profiles of sulfinamide rac-1a, catalyst In-1 and the reaction mixture in dichloromethane. b, Kinetic profiles of the deracemization of sulfinamide 1a in the dark (green circles) and with violet LED light (blue triangles). c, Spin adducts of the sulfinyl (19) and aminyl (20) radicals with DMPO detected by high-resolution mass spectrometry and X-band EPR spectroscopy. The simulated EPR spectrum was rendered with a 61:39 ratio of 19 and 20. d, Crossover experiments with racemic sulfinamides 1a and 3d. e, Relationship between the e.e. values of catalyst In-1 and product 1a. f, Sulfinyl group–catalyst In-1 binding mode determined from the X-ray crystal structure of catalyst–sulfinamide adduct 21. Hydrogen atoms and solvent molecules are omitted for clarity. Only the adduct with (R)-1g is shown. g, Density functional theory-derived Gibbs free energy profile of the catalytic deracemization of sulfinamides in units of kcal mol⁻¹. h, Transition state structures TS-S and TS-R.

Following homolytic bond cleavage, some of the radicals may escape the solvent cage and can be detected experimentally⁵. To investigate whether sulfinyl and aminyl radicals are produced from sulfinamides under catalytic deracemization conditions, the reaction was carried out in the presence of the nitrone spin trap DMPO (18). The spin adducts of sulfinyl and aminyl radicals 19 and 20 were readily detected by electron paramagnetic resonance (EPR) spectroscopy and high-resolution mass spectrometry (HRMS) (Fig. 3.c) only when the reaction mixture was irradiated with LED light. These results corroborate the involvement of light-driven homolytic S–N bond fission in the deracemization reaction. In line with this conclusion, no sulfinate ester was observed when the reaction was carried out in the presence of methanol as the sulfinyl cation scavenger (see page 10 in the Supplementary Information for details), ruling out heterolytic S–N bond cleavage.

We next sought to elucidate the roles of in-cage and out-of-cage radical processes in the deracemization reaction by performing crossover experiments with racemic sulfinamides 1a and 3d under photocatalytic conditions (Fig. 3.d). Radicals that escape from the solvent cage can then randomly combine with other radicals they encounter in the solution. If the deracemization reaction predominantly proceeds by an out-of-cage mechanism, these processes would lead to a random exchange between structurally similar sulfinyl and amino groups and an accumulation of crossover products at rates that are comparable with the rates of formation of noncrossover products. Conversely, reactions that predominantly proceed by in-cage recombination would produce crossover products in low yields. The observation of low yields would be consistent with the contribution of a minor fraction of radicals that escaped the solvent cage. Experimental studies revealed that crossover products 1t and 3e were formed in single digit yields that were incommensurate with the observed high levels of enantiomeric enrichment (See Supplementary Figures S5 and S6 and the accompanying discussion for details of the additional crossover and radical scavenger experiments). Taken together, these results are consistent with a photocatalytic deracemization reaction that predominantly occurs via asymmetric geminate recasting.

To probe the structural identity of the catalytically active species, the enantiomeric excess of sulfinamide 1a was monitored as a function of the enantiomeric purity of catalyst In-1. The observed linear relationship between the enantiomeric excess of catalyst In-1 and that of sulfinamide product 1a (Fig. 3.e) indicates that the catalytically active species contains only one molecule of indium complex In-1.

The nature of the interactions between the catalyst and the sulfinamide substrate in the reaction mixture was next investigated by means of diffusion-ordered nuclear magnetic resonance spectroscopy (DOSY NMR). This spectroscopic technique enables the measurement of the diffusion coefficients of dissolved species, which can be used to calculate their molecular weights, thereby providing insight into complexation processes in solution⁵⁷. The experimentally determined molecular weight of the In(salen)-containing species in the reaction mixture of catalyst In-1 and racemic sulfinamide 1a (MW = 1025.1, ±5.6%) was consistent with a 1 : 1 1a–In-1 complex (MW = 970.9).

Informed by this insight, we next sought to elucidate the binding mode of the sulfinyl group in the putative complex with catalyst In-1 by X-ray crystallographic analysis. Crystals suitable for X-ray analysis were obtained with racemic sulfinamide 1g. Crystallographic studies revealed that photocatalyst–substrate complex 21 was composed of In-1 and sulfinamide 1g at a 1 : 1 stoichiometry, in line with the conclusions of the DOSY NMR experiments (Fig. 3.f). The sulfinyl group is bound to the indium atom through an oxygen–indium bond (2.34 Å). Additionally, a stabilizing interaction between the sulfur atom of the sulfinyl group and the proximal oxygen of the salen ligand was evident from the short S···O_salen distance (2.89 Å).

Density functional theory (DFT) analysis was performed next to elucidate the mechanistic details and the origin of the enantioselectivity of the asymmetric geminate recasting. (Fig. 3.g). Photoexcitation of diastereomeric photocatalyst–sulfinamide complexes 22-R and 22-S and subsequent intersystem crossing produced triplet states ³22-R and ³22-S. The energy accumulated in the photochemical process rendered the barrierless stereoablative homolytic scission of the weak S–N bond thermodynamically favourable and resulted in the formation of triplet geminate pair ³[23+24] composed of photocatalyst-derived radical intermediate 23 and aminyl radical 24 from both diastereomeric precursors ³22-R and ³22-S. Analysis of the spin density distribution in intermediate 23 revealed some delocalization of the spin density to the photocatalyst fragment, although the sulfinyl fragment largely retained its radical character. Subsequent spin transition produced singlet geminate pair ¹[23+24], setting the stage for the asymmetric radical recombination step. The recombination of radicals 23 and 24 preferably proceeds through transition state structure TS-R (ΔΔG^≠ = 2.6 kcal/mol). The stereochemical outcome of this most favourable pathway leads to the R-configured sulfinamide, which is consistent with the experimentally observed absolute configuration of the deracemization products.

Analysis of noncovalent interactions (NCIs) indicated that the favoured transition state structure benefited from increased C–H---π and π-stacking interactions of the aminyl radical with the sulfinyl fragment and the salen ligand of the catalyst, including the proximal tert-butyl groups (Fig. 3.h). In accordance with the experimentally observed structural features of sulfinyl-bound adduct 21, both transition state structures exhibited stabilizing interactions between the sulfur atom of the sulfinyl group and the heteroatoms of the salen ligand.

Furthermore, computational evaluation of alternative reaction pathways that proceed by electron transfer, inversion of the sulfur stereocentre, or formation of ionic intermediates indicated that these reactions involve strongly endergonic processes and high reaction barriers (See Supplementary Figures S8–S11 and the accompanying discussion for further details).

Discussion

Collectively, the experimental and computational studies are consistent with a photocatalytic asymmetric geminate recasting mechanism that enables efficient light-driven deracemization of sulfinamides. Harnessing this concept, we developed a stereoselective strategy for accessing a variety of medicinally and synthetically valuable chiral sulfur functionalities. Given the fundamental nature of geminate recasting, we anticipate that the demonstrated photocatalytic asymmetric approach will be applicable in a broad range of structural settings.

Methods:

General procedure for deracemization of sulfinamides

A 10 mL test-tube equipped with a magnetic stirbar was charged with photocatalyst In-1 (7.4 mg, 0.01 mmol, 10 mol%), a racemic sulfinamide (0.1 mmol), and degassed anhydrous dichloromethane (2 mL). The reaction tube was capped tightly, and the cap was sealed with parafilm. The reaction mixture was stirred at −55 °C in an acetone bath for 10 minutes. LED lights (λ_max = 400 nm) were then turned on, and the reaction mixture was stirred (500–550 rpm) for 30–50 h. The reaction mixture was then concentrated, and the product was purified by preparative thin layer chromatography to give the desired enantioenriched sulfinamide.