4-Sulfanyl-1,2,3-triazole as a Powerful Ligand in CuAAC to Synthesize 1,4-Substituted 1,2,3-Triazoles Under Solvent-Free and Low Catalyst Loading

Abstract

4-Sulfanyl-1,2,3-triazole (L1) accelerated the solvent-free CuAAC efficiently with low catalyst loading (0.1 mol% for common azides and 1 mol% for sulfonyl azides). L1 exhibited higher catalytic activity compared to 1,4-substituted 1,2,3-triazole without sulfanyl group (5a) and sulfide, demonstrating that coordination of both sulfanyl and 1,2,3-triazole moieties with copper was critical to enhance the activity of L1. The Cu(OAc)₂/L1 catalytic system displayed high selectivity in synthesis of alkynyl- or azido-involved 1,2,3-triazoles. The di-copper system Cu(OAc)₂/CuBr/L1 promoted the reaction of electron-deficient and less reactive sulfonyl azides well, generating N-sulfonyl-1,2,3-triazoles in good yields, and L1 showed better performance than 1,3-di-o-tolylthiourea (L′). Other features of this protocol included recyclable ligand, 1:1 substance ratio, high yields, wide substance scope, and easily scaled up and facile purification of most products.

1. Introduction

1,4-Substituted 1,2,3-triazoles are potent moieties in bioactive compounds like agrochemicals and pharmaceuticals because they serve as facile-to-install pharmacophores and display various biological activities [1,2]. Cu-catalyzed azide–alkyne cycloaddition (CuAAC) [3,4], a valuable protocol to construct 1,4-substituted 1,2,3-triazoles, has been widely applied in synthetic chemistry [1,5], glycoscience [6], chemosensors [7], material science [8], protein science [9], biochemistry [10], drugs [11] and agrochemicals [12]. Traditional CuAAC is usually carried out in organic solvents with copper salt or complex as a catalyst and an excess amount of amine as a base [13]. To promote CuAAC efficiently in the absence of excess amine, several ligands have been developed in recent years (Scheme 1a); for example, tris-(benzyltriazolylmethyl)amine (Hm-TBTA) promoted CuAAC efficiently with 1 mol% catalyst loading [14]. Bisoxazoline (BOX) derivatives (catalyst loading, [Cu]: 10 mol%, ligand: 12%) [15,16,17,18], 1,8-naphthyridines (catalyst loading, [Cu]: 10 mol%, ligand: 13%) [19] and tris((1-cyclopentyl-1H-1,2,3-triazol-4-yl)methyl)amine (TCPTA) (catalyst loading, 5 mol%) [20] were fabricated as well, but most of these reactions proceeded in volatile organic solvents with relatively high catalyst loading. Although N-heterocyclic carbene (NHC)-Cu(I) (catalyst loading, 5 mol%) [21] and triazolyl-prolinamide (Pro-1) (catalyst loading, [Cu]: 5 mol%, ligand: 10%) [22,23] could promote CuAAC in H₂O, the catalyst loading also remained to improve. Tris(triazolylmethyl)amine (BTTAA) enabled CuAAC to be performed inside live cells at very low concentrations, but Na ascorbate was required [24]. Moreover, most of the above ligands were somewhat complex and difficult to synthesize. Thus, developing other simple, readily manufactured and powerful ligands was still required for the importance of CuAAC. On the other hand, establishing solvent-free CuAAC with very low catalyst loading at rt not only reduced cost and energy consumption, but also was necessary for conducting biorthogonal CuAAC in living organisms due to the toxicity of copper catalyst [25,26]. Recently, solvent-free CuAAC protocols were developed using 1,3-bis(2,6-dimethylphenyl)thiourea-based copper [27], Cu-MOF [28] and ZrO₂ ball mill/Cu(OAc)₂ [29] catalytic systems (Scheme 1b). But these reactions suffered some drawbacks like relatively high catalyst loading and heating. Establishing solvent-free CuAAC with wide substance scope under rt and low catalyst loading was still challenging. 4-Sulfanyl-1,2,3-triazoles were firstly and systematically synthesized by Peng and Orita from 1-phosphinyl-2-sulfanylethynes which were obtained from ethynyl(diphenyl) phosphine oxide by bromination and coupling with thiols [30], but the one-pot synthesis of 4-sulfanyl-1,2,3-triazoles from commercially available (bromoethynyl)triisopropylsilane and their application as ligands were not disclosed. Herein, 4-sulfanyl-1,2,3-triazole (L1), which was accessed facilely by a one-pot desilylation/CuAAC protocol, was developed as a powerful ligand for CuAAC (Scheme 1c). To our delight, L1 enabled the solvent-free CuAAC to be carried out smoothly with a 1:1 substance ratio under rt and low catalyst loading, producing various 1,4-substituted 1,2,3-triazoles including bioactive compounds, azido- or alkynyl-functionalized 1,2,3-triazoles and N-sulfonyl-1,2,3-triazoles in high yields.

2. Results and Discussion

Initially, we synthesized 4-sulfanyl-1,2,3-triazoles using a one-pot method (Scheme 2). The Cu(MeCN)₄PF₆-catalyzed cross-coupling between thiols (1) and (bromoethynyl)triisopropylsilane afforded sulfanylethynes (2) [31], and the following one-pot desilylation/CuAAC between 2 and benzyl azide (3a) formed the desired L1 and L2 in 81% and 62% yields, respectively. When 1,3-bis(azidomethyl)benzene (0.4 equiv.) was applied, bis(4-sulfanyl-1,2,3-triazoles) (L5) was obtained in 76% yield. Adding N,N′-dimethyl-1,2-ethanediamine (DMEDA) realized the formation of L3 and L4 in reasonable yields. This reaction was facile for a gram-scale synthesis, with 0.6021 g (5 mmol) of 1a producing 1.0974 g (3.9 mmol) of L1 in 78% yield.

The solvent-free CuAAC of ethynes (4) and azides (3) using L1–L5 as ligands was then performed (Scheme 3). When CuSO₄/L1 was used, the desired 5a was formed in 16% yield from phenylacetylene (4a) and (azidomethyl)benzene (3a) (entry 1). Among L1–L5, L1 gave the highest yield (entry 1–5). Notably, L1 displayed better performance than 5a, Ph₂S, PhSMe, Bu₂S and difurfurylsulfide (entry 6 vs. 8–12, 16 vs. 18), demonstrating that the coordination of both sulfanyl and 1,2,3-triazole moieties with copper was critical to enhance the catalytic performance of L1. The di-copper system (Cu(OAc)₂/CuBr/L1) enabled the reaction to complete in 0.5 h and showed higher catalytic activity than the mono-copper catalytic system (Cu(OAc)₂/L1) (entry 16 vs. 6). Without ligand, 5a was formed in 48% yield in the mono-copper system within 5.5 h and 28% yield in the di-copper system within 0.5 h (entries 7 and 17). In the reported CuAAC in CH₂Cl₂/CH₃OH [12], L1 also exhibited higher activity than Hm-TBTA (entry 15 vs. 14). After examining the reaction carefully (seen in Supplementary Materials), the optimized conditions were obtained as follows: a mixture of 4a (2.0 mmol), 3a (2.0 mmol), Cu(OAc)₂ (0.002 mmol) and L1 (0.002 mmol), at rt under air for 5.5 h, giving 5a in 99% yield (0.4659 g) (entry 6). In a gram-scale synthesis, 1.0213 g (10.0 mmol) of 4a formed 2.3294 g (9.9 mmol) of 5a in 99% yield (entry 13).

The substance scope was then extended (Scheme 4). The CuAAC of benzyl azides was firstly investigated. Phenyl ethynes with electron-donating (OMe, Et, Pr) and electron-withdrawing (MeOC(O), NO₂, Cl, Br, F, CF₃) groups gave 5b–5k in 90–99% yields. 2-Ethynyl-substituted thiophene, pyridine and pyrazine gave 5l–5n in 89–92% yields. Alkyl ethynes formed 5o–5q in excellent yields. Alkenyl and ester groups were tolerated, forming 5r and 5s in 88% and 86% yields, respectively. When benzyl azides with Me, CF₃, Br or Cl groups were used, 5t–5z were obtained in 87–99% yields. Naphthyl and quinolyl azides gave 5aa and 5ab in 89% and 86% yields, respectively. Diphenylmethyl azide survived to produce 5ac in 84% yield. Moreover, the less reactive alkyl and phenyl azides [32] were applicable as well. Primary and secondary alkyl azides provided 5ad–5ah in 86–94% yields. 2,3-Dihydrobenzofuranyl-, tetra-hydrofuranyl- and phenylthiol-involved alkyl azides afforded 5ai–5ak in excellent yields. The reaction of 1-azido-4-methylbenzene and ethynes formed 5al–5aq in 80–93% yields. Trimethylsilylacetylene was also tolerated to form 5ar in 72% yield. Bioactive compounds were also obtained: 5as (97% yield) with heat shock protein 90-inhibitory activities [33], 5at (87%) with phytotoxic activity [34], 5au (85%) with fungicide activity [11], 5av (83%) with anti-plasmodial activity [35], 5aw (90%) inhibiting the Nrf2-Keap1 protein–protein interaction [36], 5ax (89%) with antidepressant-like effect [37], and 5ay (84%) reducing the mycelial growth of colletotrichum gloeosporioides [38]. All products were purified by washing, which avoided the tedious column chromatography.

Generally, alkynyl- or azido-involved 1,2,3-triazoles, versatile intermediates in organic synthesis, were difficult to synthesize via asymmetric CuAAC in organic solvents for the further conversions of dissolved products [15,16,17,18,19,20,39,40,41]. In this regard, our solvent-free protocol exhibited high selectivity due to the tiny reactions of precipitated products. As disclosed in Scheme 5, the reaction of ethynes and diazides formed 1,2,3-triazoles (6) bearing an unreacted azido in good yields. Meta-, ortho- and para-substituted benzyl diazides were applicable, giving 6a–6c in 72–85% yields. Ethynes with electron-donating (OMe, Me, Me₂N) and electron-withdrawing (F, Cl, CF₃, MeO(O)C, NO₂) groups afforded 6d–6k in 81–90% yields. 2-Ethynylpyridine and 2-ethynylthiophene were tolerated, forming 6l and 6m in good yields. Alkyl diazide was suitable as well (6n, 69% yield). Subjection of diyne to this reaction produced 1,2,3-triazole (7) with an unaltered alkynyl in high yields. Benzyl and alkyl azides afforded 7a–7f in 67–91% yields. The reaction of diazides and 1,3-diethynylbenzene proceeded well to afford 7g and 7h in moderate yields. In most cases, the azido in 6 and alkynyl in 7 were unconverted for the deposition of products. Treatment of 1,3-diethynylbenzene with benzyl azide and 1-azidohexane afforded bis(1,2,3-triazoles) (8a, 8b) in 93% and 90% yields, respectively. When 1,3-bis(azidomethyl)benzene was employed, 8c was formed in 95% yield. Similarly, 8d and 8e were also provided in high yields.

N-Sulfonyl-1,2,3-triazoles were not only key synthons in denitrogenative transformation [42] but also displayed antitumor activity [43]. Sulfonyl azides displayed lower reactivities in CuAAC than common azides due to electron deficiency. In the reaction between 4a and benzenesulfonyl azide (3aa), N-sulfonyl-1,2,3-triazole (9a) was formed in less than 7% yield in the presence of Cu(OAc)₂/L1 even when the catalyst loading was elevated to 2 mol% (Scheme 6, entry 1–3). CuBr/L1 also provided 9a in less than 10% yield (entry 4). Combination of Cu(OAc)₂, CuBr and L1 with catalyst loading of 0.25 mol% enhanced the yield of 9a to 32% (entry 5). Increasing the catalyst loading to 1.0 mol% gave the best yield (entry 7 vs. 6 and 8). However, ligand 1,3-di-o-tolylthiourea (L′) afforded 9a in 76% yield (entry 9). The best conditions were provided as follows: a mixture of 4a (1.0 mmol), 3aa (1.0 mmol), Cu(OAc)₂ (0.01 mmol), CuBr (0.01 mmol) and L1 (0.01 mmol), at rt under air for 20 h, producing 9a in 99% yield (0.2825 g) (entry 7).

With L1 as a ligand, phenyl ethynes with either electron-withdrawing (F, Br) or electron-donating (Me, Et, OMe) group were applicable under the above best reaction conditions, producing the corresponding 9c–9h in 88–94% yields, while 9c–9h were generated in 66–82% yields when 1,3-di-o-tolylthiourea (L′) was employed. These above results demonstrated that L1 displayed better performance than L′ in the solvent-free CuAAC of sulfonyl azides (Scheme 7).

A plausible mechanism for the above di-copper system was proposed. Firstly, activation of Cu(OAc)₂ by ligand (L1) generated L1-Cu(OAc)₂ (0). Coordination of 0 with 4a formed the intermediate A, which then underwent H atom abstraction from ethyne by OAc and produced B. Di-copper intermediate C was formed by incorporation of CuBr and 3aa with B with elimination of HOAc. Intramolecular cyclization of C gave the six-membered metallacycle D by C1–N1 bond formation. Formation of C2–N2 in D gave the five-membered cycle E with the liberation of CuBr. Coordination of HOAc with E yielded F, which underwent H transformation to afford 9a and regenerate the catalyst (0) (Scheme 8). The coordination of both sulfanyl and 1,2,3-triazole moieties in L1 with copper was critical to enhance the activity of L1 possibly due to the formation of more-stable six-membered metallacycle D, which then lowered the activation energy of this CuAAC reaction.

The ligand L1 could be recycled using the reaction of 4a and 3a to synthesize 5a as a model (10 mmol scale). After workup with NH₄Cl aq. and extracting with ethyl acetate, the recycled ligand L1 was pure enough and could be reused without further purification. Subjecting the recycled L1 to the second cycle afforded 5a in 98% yield. L1 could be reused for up to ten successive cycles. In this cycle, we found that it took 8 h to achieve 95% yield up to five cycles. After five catalytic cycles, 9 h was required to achieve excellent yields (Figure 1).

3. Materials and Methods

All reagents were commercially available and used without further purification. The dehydrated solvents were purchased and redistilled before use. Glassware was dried in an oven and heated under reduced pressure before use. Column chromatography was undertaken on silica gel (300–400 mesh) using a proper eluent system. Analytical thin-layer chromatography (TLC) was performed on Haiyang TLC silica gel GF254 (Qingdao, China, 0.25 mm) plates. Proton, carbon and fluorine nuclear magnetic resonance spectra (¹H, ¹³C and ¹⁹F NMR) were recorded on a Bruker-400 (Billerica, MA, USA, 400 MHz for ¹H NMR, 101 MHz for ¹³C NMR and 376 MHz for ¹⁹F NMR spectroscopy) spectrometer with solvent resonance as the internal standard (¹H NMR, CDCl₃ at 7.26 ppm, DMSO-d₆ at 2.50 ppm; ¹³C NMR, CDCl₃ at 77.16 ppm, DMSO-d₆ at 39.52 ppm). Chemical shifts are reported in ppm (δ) relative to internal tetramethylsilane (TMS). Data are reported as follows: chemical shift, multiplicity (s = singlet, d = doublet, t = triplet, q = quartet, m = multiplet); coupling constants (J) in hertz. Melting points were measured using a melting point meter RY-1G. Crystal data were acquired at 296 K on a Rigaku Oxford Diffraction Supernova Dual Source (Akishima, Japan), Cu at zero, equipped with an Atlas S2 CCD using Cu Kα radiation. HRMS data were acquired using the Waters G2-Xs qtof mass spectrometer (Milford, MA, USA) under Electron Spray Ionization conditions. Ethynes 4y [34] and 4z [38] were prepared according to the reported methods; other ethynes 4a–4x were purchased from Energy, SigmaAldrich (St. Louis, MO, USA) and Leyan (Shanghai, China).

Procedure for Synthesis of L1–L5. Cu(MeCN)₄PF₆ (0.1 mmol) was dissolved in degassed acetonitrile (9 mL) completely under nitrogen. A Schlenk tube of 100 mL equipped with a magnetic stir bar was charged with (bromoethynyl)triisopropylsilane (1.0 or 2.0 mmol), thiol 1 (1.1 or 2.2 mmol), dtbbpy (0.2 mmol), 2,6-lutidine (2.0 mmol) and degassed acetonitrile (20 mL) under nitrogen. Then the above acetonitrile solution of Cu(MeCN)₄PF₆ was added to the Schlenk tube. The mixture was stirred at rt for 15–30 min, and then was concentrated under vacuum to provide a crude reaction mixture. The residue was extracted with EtOAc and saturated NH₄Cl aqueous solution. The organic layer was washed with brine and dried over Na₂SO₄. The solvents were evaporated after filtration. The crude residue was subjected to short-column chromatography on silica gel (hexanes) to extract alkynyl sulfides 2 [24], which were used for the next step without further purification. The crude product 2 was dissolved in THF (10 mL) and TBAF (1.0 M in THF, 1.0 mmol) was added at 0 °C. The mixture was stirred at rt for 10 h. Then CuI (0.1 mmol), azide 3 (1.0 mmol) and nPrOH (20 mL) were added, and the mixture was stirred at 100–110 °C for 15 h. The product was extracted with EtOAc after the mixture was quenched with saturated NH₄Cl aqueous solution. The combined organic layer was washed with brine and dried over Na₂SO₄. The solvents were evaporated after filtration. The crude residue was subjected to column chromatography on silica gel (petroleum ether/EtOAc, 4:1) to extract L1–L5 in pure forms.

General Procedure for the Synthesis of 5, 6, 7 and 8. A mixture of Cu(OAc)₂, L1 (0.0006 g, 0.002 mmol), 4 (2.0 mmol) and 3 (2.0 mmol) was stirred under air at rt for 5.5–15 h. Then the reaction mixture was washed with water twice followed by cold petroleum ether (or petroleum ether/ethyl acetate (10/1)) three times or the reaction mixture was subjected to column chromatography on silica gel to produce the desired 5a–5ay (purified by washing with water and cold petroleum ether), 6a–6n (purified by column chromatography), 7a–7h (purified by column chromatography), and 8a–8e (purified by washing with water and petroleum ether/ethyl acetate (10/1)) in pure forms.

General Procedure for the Synthesis of 9. A mixture of CuBr, Cu(OAc)₂, L1, 4 and 3 was stirred under air at rt for 20 h or 32 h. Then the reaction mixture was washed with water twice followed by petroleum ether three times to produce the desired 9 in pure form.

4. Conclusions

In summary, 4-sulfanyl-1,2,3-triazole (L1) with excellent catalytic performance in CuAAC was obtained facilely in gram scale using a one-pot method. Using Cu(OAc)₂/L1 as a catalytic system, the solvent-free CuAAC between common azides and ethynes with a substance ratio of 1:1 proceeded smoothly at rt under air and low catalyst loading conditions, which not only avoided volatile organic solvents and reduced cost and energy consumption, but also exhibited high atomic economy. The di-copper system Cu(OAc)₂/CuBr/L1 accelerated the reaction of sulfonyl azides well, generating N-sulfonyl-1,2,3-triazoles in high yields.