Sustainable Thioetherification via Electron Donor-Acceptor Photoactivation using Thianthrenium Salts

Abstract

The synthesis of sulfides has been widely studied because this functional subunit is prevalent in biomolecules and pharmaceuticals, as well as being a useful synthetic platform for further elaboration. Thus, various methods to build C–S bonds have been developed, but typically they require the use of precious metals or harsh conditions. Electron donor-acceptor (EDa) complex photoactivation strategies have emerged as versatile and sustainable ways to achieve C–S bond formation, avoiding challenges associated with previous methods. This work describes an open-to-air, photoinduced, site-selective C–H thioetherification from readily available reagents via EDA complex formation, tolerates a wide range of different functional groups. Moreover, C(sp²)–halogen bonds remain intact using this protocol, allowing late-stage installation of the sulfide motif in various bioactive scaffolds, while allowing yet further modification through more traditional C-X bond cleavage protocols. Additionally, various mechanistic investigations support the envisioned EDA complex scenario.

Graphical Abstract

A versatile electron donor-acceptor (EDa) complex thioetherification was developed using simple organic compounds (thianthrenium salts and thiols) and visible-light irradiation under open-to-air conditions. This protocol allows the retention of the C-X bond, as well as the late-stage thioetherification of biomolecules.

Introduction

Given its intrinsic advantages, electron donor-acceptor (EDa) complex photoactivation has recently become a powerful synthetic tool to assemble new bonds in a sustainable manner.¹ This approach enables radical formation that is induced by visible-light activation under mild conditions via charge transfer interaction between, typically, two colorless organic compounds.^1,2 The scientific community has found this strategy attractive because of its ability to generate radicals in an operationally simple manner without exogenous photocatalysts, reporting several visible-light assisted transformations in an effort to access new chemical space in a simple way.² Notably, some of these methods were focused on the construction of C–S bonds, achieving the synthesis of sulfides under mild conditions and inert atmosphere, avoiding the use of precious metals (Figure 1).³

Figure 1.

Strategies to synthesize thioethers via EDA complex photoactivation.

Sulfides or thioethers are important scaffolds in the materials science,⁴ agrochemical,⁵ and pharmaceutical⁶ industries, as well as being valuable synthetic precursors to access higher oxidation state sulfur-containing compounds.^7,8b Thus, many synthetic routes to these materials have been developed in recent decades.^8-13 The Stadler-Ziegler reaction⁹ and transition metal-catalyzed cross-couplings¹⁰ represent the traditional approach to C–S bond construction, and in the last decade various photoredox,¹¹ metallaphotoredox,¹² and radical-nucleophilic aromatic substitution¹³ reactions have emerged as complementary and improved protocols. However, most of these reported methods require strong bases, high temperatures, air-sensitive ligands, or precious metals. These issues were addressed by the recent appearance of photoinduced EDA complex thioetherifications (Figure 1). Thus, Miyake and co-workers described the formation of C(sp²)–S bond formation via charge transfer interaction between aryl halides (acceptors) and thiolate anions (donors).^3a Additionally, Akiyama et al. and Song et al. reported different visible-light-assisted EDA complex methods to forge C(sp³)–S bonds readily. Akiyama and co-workers designed a combined EDA complex and hydrogen atom transfer (HAT) system,^3b while Song and coworkers used various alkyl N-hydroxyphthalimide esters as electron-acceptors.^3c However, these reactions are air sensitive and/or proceed through C(sp²)–X reduction.

Recently, Procter and co-workers reported an efficient Giese-type aryl radical addition and an insertion of the cyanide group via EDA complex photoactivation using aryl dibenzothiophenium or -phenoxathinium salts as electron-acceptors, respectively.¹⁴ Sulfonium salts are well known as versatile reagents for a highly site-selective C–H activation from commodity reagents.¹⁵ In particular, Ritter et al. described a novel arene thianthrenation reaction that allows late-stage functionalization.¹⁶ Given the special properties of such thianthrenium salts and their susceptibility to undergo single electron reduction, several applications were developed in cross-coupling¹⁷ and metallaphotoredox catalysis.¹⁸ Notably, one of those application described the synthesis of aryl thioethers by photoredox/Cu dual catalysis.^17b

In conjunction with recent investigations performed by our research group in the field of photoinduced EDA complexes¹⁹ and inspired by the work reported by Procter et al.,¹⁵ we envisioned the formation of visible-light absorbing EDA complexes between thianthrenium salts (acceptors) and thiolate anions (donors), as a way to forge a new C(sp²)–S bonds (Figure 1). Thianthrenium salts were anticipated to function as radical precursors, releasing a radical after a single electron reduction event with subsequent loss of recyclable thianthrene, allowing the insertion of various thiols onto the skeleton of simple arene compounds or more complex bioactive motifs from readily available reagents and visible light irradiation.

Results and Discussion

To evaluate the feasibility of the conceived visible light-mediated thioetherification, thianthrenium (TT⁺) salt 1a and 4-chlorothiophenol (2a) were utilized as model substrates as indicated in Scheme 1. An initial solvent and base screening showed that the combination of dimethyl sulfoxide (DMSO) and potassium carbonate generated the best results (entry 1-8). The use of polar aprotic solvents with high dielectric constants (κ) and the capacity to solubilize all of the components of the reaction were best suited to allow the formation and stabilization of the envisaged EDA complex. DMSO satisfies both specifications: it possesses a high κ value (47) and perfectly dissolves all the reagents. On the other hand, an examination of various inorganic and organic bases revealed that carbonates and potassium phosphate tribasic prove to be suitable bases for this thioetherification reaction. The less expensive K₂CO₃ ($0.08/gram) gave better results than Cs₂CO₃ ($3.78/gram), which had been used in other reported EDA complex thioetherifications.^3a-b Notably, an open-to-air experiment afforded identical results to those found under inert atmosphere, indicating that oxygen did not affect the transformation (entry 9). Additionally, different light sources were evaluated (entry 10). Irradiation at 427 nm (Kessil PR160L) instead 390 nm (Kessil PR160L) had no dramatic effect on the reactivity, while the use of 455 nm (Kessil A160We), 525 nm (Kessil PR160L) or blue LED strips (light-emitting diode, λ_max = 456 nm) resulted in a decreased formation of sulfide 3a, and dimerization of 2a to 1,2-bis(4-chlorophenyl)disulfane was detected (see Supporting Information). More detailed optimization parameters can be found in the Supporting Information. Control experiments with no base and no light confirm the necessity of these conditions to enable the C(sp²)–S bond formation and rule out a nucleophilic aromatic substitution mechanism.

Scheme 1.

Optimization of Reaction Conditions: Optimization of reactions were performed using 1a (0.10 mmol), 2a (0.20 mmol), and base (0.20 mmol), in dry degassed solvent (1.0 mL, c = 0.1 M) under purple Kessil® irradiation (λ_max = 390 nm) for 2 h at rt. ^aCalculated using 1,3,5-trimethoxybenzene as internal standard (IS) from the crude mixture. ^bOpen-to air. ^cIsolated yield of 3a on 0.3 mmol scale. nr: no reaction.

With suitable conditions in hand, the scope of the EDA complex photoactivation method was investigated. A first evaluation of various (hetero)aryl thiols was performed using TT⁺ 1a or TT⁺ 1b salts as model substrates (Scheme 2). Initially, para-chloro- and para-bromothiophenol afforded the corresponding sulfides 3a and 3b in excellent yields, while the use of the brominated regioisomers meta- (3g) and ortho-bromothiophenol (3j) proved to be less competent. In contrast, 2-fluorothiophenol yielded sulfide 3i in 91% yield, presumably owing to the smaller size of the fluorine substituent. Both, electron-rich methoxy- and electron-deficient trifluoromethyl-substituted thiophenols displayed similar reactivity, efficiently generating the desired thioethers 3c and 3e, respectively. Notably, free aniline (3d), carboxylic acid (3f, 3o), phenol (3h, 3t), and boronic acid (3s) functional groups were tolerated under the reaction conditions, avoiding the use of protecting groups. The corresponding sulfides were isolated in low to moderate yields along with methyl 2-methoxybenzoate byproduct, derived from the in situ formed aryl radical via hydrogen atom transfer (HAT). A sterically hindered 2,6-dimethylbenzenethiol was also found to be a suitable coupling partner in this photoinduced transformation (3k). Pentafluorothiophenol was incorporated as a dimer through an initial nucleophilic substitution under the basic reaction conditions (3l). Additionally, electron-poor and electron–rich heteroaryl thiols were amenable, affording the desired pyridyl- (3m, 3o), pyrimidyl- (3n, 3p), benzo thiazolyl- (3q), and benzo imidazolyl (3r) sulfides in excellent yields. Unfortunately, C(sp²)–S coupling with alkyl thiols was unproductive, with disulfides resulting from dimerization of the thiols being detected, along with unreacted TT⁺-1a (see Supporting Information, Section 4d).

Scheme 2.

Aryl Thiol Scope. Reaction conditions: 1a (0.30 mmol), 2a (0.45 mmol) and K₂CO₃ (0.45 mmol), in dry DMSO (3.0 mL, c = 0.1 M) under open-to-air blue Kessil® irradiation (λ_max = 427 nm) at rt. ^aK₂CO₃ (0.60 mmol).

4-Chlorothiophenol, pyridine-2-thiol and 2,6-dimethylthiophenol were selected as model substrates to test the reactivity of different TT⁺ 1 salts. As shown in Scheme 3, aldehyde (3u-v), amide (3w, 3x, 3aa), strained cyclopropyl (3x), ester (3x, 3y, 3aa), and ether (3u-v, 3aa, 3ad-ae) groups were efficiently incorporated. Although C–halogen bonds are known to be reduced in various EDA complex transformations to generate aryl radicals,^3a-b,13a halogen substituents survived under the developed reaction conditions to afford brominated sulfides 3ab-ae and iodinated sulfide 3af in good yields. Retention of the halide handles in this protocol complements other reported thioether syntheses that proceed via reduction of the C–halogen bond,^3a-b and allows further diversification of the sulfide structure by known procedures, such as metallaphotoredox catalysis.²⁰

Scheme 3.

Aryl Thianthrenium Salt Scope. Reaction conditions: 1a (0.30 mmol), 2a (0.45 mmol) and K₂CO₃ (0.45 mmol), in dry DMSO (3.0 mL, c = 0.1 M) under open-to-air blue Kessil® irradiation (λ_max = 427 nm) at rt. ^a2a (0.33 mmol) and K₂CO₃ (0.33 mmol). ^bK₂CO₃ (0.60 mmol). ^cGram scale reaction

The amenability of this EDA complex photoactivation strategy on C–H late-stage functionalization^16c was demonstrated by the structural modification of several highly functionalized pharmaceuticals and biomolecules. Thus, the sulfide moiety was successfully introduced in the skeleton of anti-inflammatory compounds salicin (3ah), flurbiprofen (3al-am) and fenbufen (3an-ao), lipid-lowering agents gemfibrozil (3aj) and clofibrate (3ak), anti-fungal agent bifonazole (3ai), as well as insecticide and photocatalyst xanthone (3ag). Additionally, sulfide 3ao from fenbufen was synthesized on gram scale, further showcasing the synthetic utility of this photoinduced thioetherification protocol.

To elucidate the reaction mechanism of this photoinduced thioetherification process, mechanistic studies were performed as shown in Figure 2 and in the Supporting In-formation. The analysis of the UV/Vis absorption spectra of each of the reaction components and mixtures in DMSO evidenced the formation of an intermediate EDA complex between TT⁺ salt 1a and the thiolate anion of 2a (Figure 2b). TT⁺ 1a (black line) and 4-chlorothiophenol (2a) present absorption bands in the near ultraviolet range, and the mixture of 2a with K₂CO₃ (pink line) displays a small absorption in the visible-light region. The TT⁺ salt 1a and 2a mixture (dark green line) shows a small absorption band displacement with an insignificant tailing and visible-light absorption, while the mixture of reaction components (olive green band) exhibits a significant bathochromic shift with a visible-light absorption tailing to the 425-560 nm range (see Supporting Information). Both bathochromic shifts derived from the charge-transfer absorption of a new molecular aggregate, and the set of observations of the individual reaction components are in agreement with the visual appearance of each solution (Figure 2a). The solution of TT⁺ 1a with 2a stayed colorless, while the solution of TT⁺ 1a with 2a in the presence of K₂CO₃ changed to an intense yellow color. Additionally, these studies revealed the important role of K₂CO₃ in the establishment of a powerful EDA complex via in situ formation of the corresponding thiolate anion of 2a. The formation of this intermediate EDA complex was also confirmed by the Job Plot²¹ and Benesi-Hildebrand²² (Figure 2d) experiments, where a 1:1 stoichiometric relation between TT⁺ salt 1a and the thiolate anion of 2a was confirmed, with an association constant of 7.8 M-1 in DMSO (see Supporting Information).

Figure 2.

Mechanistic Studies: A. Visual appearance of individual reaction components and mixtures thereof. B. UV/Vis absorption spectra measured in DMSO (0.1 M) at 0.3 mmol scale. C. Job plot for a mixture of the thiolate anion of 2a and thianthrenium salt 1a in DMSO (0.1 M). D. Radical trapping experiment with TEMPO. E. Proposed mechanism to synthesize thioethers 3 via EDA complex photoactivation.

The addition of TEMPO (2,2,6,6-tetramethyl-1-piperidinyloxy) under standard conditions resulted in the inhibition of the thioetherification reaction, observing the dimerization of 2h to 3,3'-disulfanediyldiphenol and isolation of the corresponding TEMPO adduct 4. The structure of 4 was confirmed by NMR and HRMS analysis, as evidence of aryl radical generation from TT⁺ salt 1b (Figure 2d). Furthermore, the photochemical quantum yield Φ for this transformation is 89, thus indicating an efficient chain radical mechanism (see Supporting Information).²³

Given previous reports,^3a-b especially the thioetherification described by Bunnett et al.,^13a and the mechanistic findings described herein, a plausible mechanism for this C(sp²)–S coupling is proposed in Figure 2E. The proposed mechanism involves the formation of an intermediate EDA complex that induces a unimolecular nucleophilic substitution (S_RN1) radical chain. Under basic conditions, the thiolate anion A is formed in situ from thiol 2. Association of the electron-rich anion A with the electron-poor TT⁺ salt 1 forms a new molecular aggregate B in the ground state. Irradiation and excitation of this new EDA complex B at 427 nm triggers an intra-complex electron transfer event from anion A to TT⁺ salt 1, generating disulfide and thianthrene byproducts, along with the corresponding transient radical C. Based on Miyake and co-workers’ density functional theory (DFT) calculations^3a and the lack of reactivity observed with alkyl thiols, it is feasible to propose that the association of anion A with TT⁺ salt 1 proceeds via π–π interaction. The resulting aryl radical C escapes from the solvent cage and couples with another anion A, affording the intermediate diaryl sulfide radical anion D. A subsequent single-electron reduction event of another TT⁺ salt 1 (E_red ≈ −1.5 V vs. SCE^18a) by the resulting organic reductant D (E_red ≈ −2.7 V vs. SCE²⁴) promotes the formation of the desired sulfide 3, regenerating aryl radical C and liberating thianthrene byproduct. The regeneration of C maintains the radical chain, and thianthrene byproduct can be recovered from the large-scale reaction in 89% yield. Because the thianthrene byproduct is the precursor to the synthesis of TT⁺ salt 1,^16a recycling of this byproduct can be effected as shown Figure 2E, providing a sustainable visible light-assisted process.

Conclusion

In summary, an operationally simple thioetherification protocol was developed from rapidly accessed or commercially available reagents [thianthrenium salts, (hetero)aryl thiols, and potassium carbonate]. Given its mild conditions, this EDA complex–S_RN1 method exhibits a broad range of functional group compatibility (for example, free boronic or carboxylic acids were amenable under the reaction conditions) and allows the retention of C–halogen bonds as an important complement to other reported methods. Moreover, the sulfide motif was incorporated in highly functionalized biomolecules and pharmaceuticals, demonstrating the synthetic utility and applicability of this process in late-stage C–H functionalization. Although a nucleophilic aromatic substitution scenario could be envisaged, mechanistic investigations and control experiments dismissed this pathway in favor of a charge transfer interaction mechanism.