Synthesis of a novel resorcin[4]arene–glucose conjugate and its catalysis of the CuAAC reaction for the synthesis of 1,4-disubstituted 1,2,3-triazoles in water

The Cu(i)-catalyzed azide–alkyne cycloaddition (CuAAC) in aqueous media using resorcin[4]arene glycoconjugate (RG) is reported. The eight β-d-glucopyranoside moieties constructed on the resorcin[4]arene upper rim provide a pseudo-saccharide cavity that offers a suitable host environment for water-insoluble hydrophobic azido and/or alkyne substrates in water. The utility of RG was established as an efficient inverse phase transfer catalyst for the CuAAC in water as a green approach for the synthesis of 1,4-disubstituted 1,2,3-triazole species. The catalytic utility of RG (1 mol%) was demonstrated in a multicomponent one-pot CuAAC for various azido/alkyne substrates. The RG acts as a molecular host and a micro-reactor resulting in the 1,4-disubstituted 1,2,3-triazoles in excellent yield.


Introduction
The classical Huisgen 1 cycloaddition reaction for the synthesis of 1,2,3-triazole involves thermal 1,3-dipolar cycloaddition of organic azides with alkynes, though in low yield and mixed regioselectivity. Sharpless' 2 and Meldal's 3 research groups later independently developed the improved procedure involving the copper(I)-catalyzed Huisgen 1,3-dipolar cycloaddition reaction, which is the widely studied 'click' reaction (Scheme 1). The impact of the copper catalyzed azide-alkyne click reaction in various branches of science is increasing exponentially as evidenced from numerous recent reviews available in the literature since 2010. 4 The three most common facile protocols for CuAAC include (i) use of copper(I) salts (mostly in organic solvents), (ii) the reduction of a copper(II to I), and (iii) oxidation of Cu(0 to I). Of the three protocols described above, the method employing in situ reduction of copper(II) salts is known to be more practical and can be carried out in aqueous conditions. From review of literature, it is easy to conclude that water is an appropriate choice as a solvent for the CuSO 4 /sodium ascorbate catalyzed click protocol, which results in the formation of the triazole in high yields and with excellent regioselectivity. However, despite the efficiency of the CuAAC reaction, there are limitations to using the procedure especially when the substrates are not water-soluble. The protocol in essence requires deoxygenated conditions in the presence of mixed aprotic organic solvents such as THF, CH 3 CN, CH 2 Cl 2 , toluene, etc. and due to the oxidative tendency of the copper(I), a higher catalyst concentration throughout the reaction is needed. To stabilize the catalyst, several phosphine-based complexes and amine-based (bound with different heterocyclic donors) ligands have been used for rate acceleration. 5 Additionally, a number of heterogeneous Cu catalysts, 6 including amberlyst resin-supported, 7 polymer-supported, 8 and zeolite-supported 9 have been explored to catalyze the triazole formation. To speed up the azide-alkyne reaction, use of surfactants and phase transfer catalyst, 10 the microwave 11 and ultrasound irradiations 12 have also been reported. 13 Recently, our research group reported resorcin [4]arene cavitand glycoconjugates (RCGs) 14 as inverse phase transfer catalysts with abilities to catalyze organic reactions in aqueous media. We also reported on the RCGs ability to catalyze the formation of 1,4-disubstituted 1,2,3-triazoles in water without the addition of any co-organic solvents. 14 It is noteworthy to mention that we were the rst to establish the concept of the spatial directionality of b-D-glycopyranoside units on the resorcin [4]arene rigid structure "cavitand". The RCGs possesses a unique molecular host system "pseudo-saccharide bucket" which can encapsulate organic substrates and catalyze chemical reactions in water. 14 In this manuscript, we describe the synthesis of a resorcin [4] arene glycoconjugate (RG) (Fig. 1) and its application as a microreactor for the synthesis of 1,4-disubstituted 1,2,3-triazole species in aqueous media via the CuAAC reaction. RG structure consists of eight b-D-glucopyranoside moieties constructed on the phenolic parts on the resorcin [4]arene upper rim via multiple 1,4disubstituted 1,2,3-triazole linkages. The eight arm resorcin [4] arene glycoconjugate offers an enlarged exible pseudosaccharide cavity that can act as a molecular vessel for waterinsoluble azido and/or alkyne substrates in aqueous environment.

Results and discussion
Synthesis of eight arm resorcin [4]arene glycoconjugate (RG) For the synthesis of the novel RG, resorcin [4]arene 1 (ref. 14 and 15) was rst synthesized upon the acid-catalyzed cyclocondensation reaction of methyl resorcinol with heptanal. Compound 1 was then treated with propargyl bromide in the presence of potassium carbonate in reuxing acetone to achieve the octa-propargyl resorcin [4]arene intermediate 2 (Scheme 2).
Resorcin [4]arene 2 was characterized conclusively from its NMR and spectral data. In its 1 H-NMR spectrum, the benzylic protons (-CH 3 , H a ) were found as a singlet at 2.25 ppm. The alkyne protons (-C^CH, H f ) were observed as a triplet at 2.50 ppm (J ¼ 2.4 Hz) and the propargyl methylene protons (-OCH 2 C^, H c,c 0 ) showed as two set of double doublets with J ¼ 15.3 and 2.4 Hz at 4.20 and 4.36 ppm. Its 13 C-NMR spectrum had resonances for the alkyne carbons C e and C f at 74.8 and 79.5 ppm, respectively, and the propargyl methylene carbons (-OCH 2 C^, C c ) were at 60.3 ppm (Fig. 2). The molecular formula (C 80 H 96 O 8 ) of compound 2 was conrmed from its m/z measurement using ESI-Q-TOF HRMS: observed 1207.7034 (M + Na) + , calculated 1207.7003 (M + Na) + .
The structures of the RG and its octaacetoxy precursor 4 were established from the respective 1 H-and 13 C-NMR data. Namely, the absence of the alkyne protons (H f ), the emergence of the triazole protons (1H, at 7.69 ppm) and the shi of the propargyl protons (H c , from 4.11 ppm to 4.50 ppm) conrmed the structure of 4 (Fig. 3). In the 1 H-NMR spectra of RG (Fig. 3b), the disappearance of the acetate protons (-OCOCH 3

Optimizing CuAAC in water catalyzed by RG
The CuAAC reactions was investigated to examine the catalytic activity in aqueous environment. For optimization, the CuAAC reaction of benzyl azide (1.0 mmol) with phenyl acetylene (1.05 equiv.) with/without the addition of a catalyst (1 mol%) was carried out (Fig. 4). The reactions were performed in the presence of copper sulfate (1 mol%) and sodium L-ascorbate (3 mol%) in 10 mL of distilled deionized water at 80 C (Fig. 4). The chemical structures and synthetic procedure for RCG catalysts evaluated in Fig. 4 have been previously reported by our research group. 14 The coupling reaction was much slower in absence of added RCG catalyst and in presence of compound 14 ( Fig. 1), 14 which lacks the spatial directionality of the RCGs. While the RCGs catalyzed the CuAAC reaction between benzyl azide and phenyl acetylene, there were signicant differences in substrate conversion to suggest a dependence of catalytic activity on chemical structure (Fig. 4). Remarkably, the CuAAC reaction was almost completed with more than 93% conversion only in 10 minutes when RG (1 mol%) was added but only 26% conversion was observed in its absence (no catalyst). Obviously, the fast CuAAC in the presence of RG indicates that it provides a unique molecular environment that is capable of catalyzing the CuAAC reaction efficiently.

Scoping the CuAAC in water using RG
To scope the CuAAC reactions catalyzed by RG in aqueous media, we have investigated coupling of substituted benzyl azides 1a-4a with aromatic and aliphatic alkynes 1b-4b ( Table 1).
As recorded in Table 1, the RG catalyzed CuAAC reactions of a variety of substituted benzyl azides in water and all reactions led to the desired 1,4-disubstituted 1,2,3-triazole products (2c-16c) in high yields (>90% isolated). Not surprisingly, it was determined that the reaction took longer to reach completion as the hydrophobicity of the substrate pair increased, as evidenced by their clog P values (Table 1). For example, the coupling between the 4-bromo benzyl azide (4a, clog P ¼ 4.13) and 1hexyne (4b, clog P ¼ 2.51) took nearly 65 minutes to completion while the reactions between propargyl alcohol (1b, clog P ¼ À0.34) and 3-methoxy benzyl azide (1a, clog P ¼ 3.18) was completed in 15 minutes.
To further evaluate the effectiveness of the RG as a catalyst in CuAAC of hydrophobic substrates, we carried out the coupling of phenyl acetylene (3b) with alkylated ortho-azido phenols (5a-9a) of increasing steric bulk and clog P values from 2.55-4.95. A   Table 1 CuAAC of substituted benzyl azides 1a-4a and alkyne substrates 1b-4b in the presence of RG a comparative study of the CuAAC reaction in absence and presence of the RG is shown in Table 2.
As expected, the coupling of o-azido phenols (5a, clog P ¼ 2.44) with phenyl acetylene 1b catalyzed by RG resulted in triazole 17c in 96% yield and even without added catalyst the yield was 64% due to its higher hydrophilicity (entry 1). However, replacing the hydroxyl functionality with alkoxy groups, i.e. -OMe, -OEt, -OBu, -OPh of progressively higher bulk, hydrophobicity, and clog P value led to much slower reactions in absence of the RG. The reactions catalyzed by the RG resulted in near quantitative conversion in about 30 minutes irrespective of the bulk or hydrophobicity of the azide substrates, attesting to its effectiveness as a catalyst in the CuAAC reactions in water even for the much bulkier and hydrophobic substrates.

Di-CuAAC reactions in water using RG
Simultaneous multiple CuAAC reactions have found interest in synthesis of the polymers and dendrimers and the need for efficient catalyst that can catalyze reactions in water has never been greater. 17 We have therefore investigated RG for catalyzing the di-CuAAC reactions in water of di-propargyl benzene derivatives 5b-7b with substituted benzyl azides 1a-4a (Table 3). Interestingly, the di-CuAAC reaction were completed within 45 min resulting in the desired bis-1,2,3-triazole products 22c-33c in gram quantities.

Multicomponent one-pot CuAAC reactions in water
Organic azides are not always considered safe for handling because of their toxic and shock sensitive nature and there have been alternative methods employed for their in situ synthesis. The azides can be prepared from their corresponding halides upon the addition of sodium azide. Hence, a multicomponent one-pot CuAAC reaction between in situ generated azide from its corresponding precursor and alkyne is highly desirable. 18 We have investigated the multicomponent one-pot CuAAC reactions of phenyl acetylene 3b with the aryl bromides and sodium azide catalyzed by the RG. Complete conversion were accomplished within 25-55 minutes even for a bulkier aryl system (naphthyl) to achieve the 1,2,3-triazoles in excellent isolated yield (90-96%) ( Table 4, entries [1][2][3][4][5]. In addition, reaction with a-bromo acetophenones (entries 6 and 7), a-bromoesters (entry 8) and ally bromide (entry 9) led to quantitative conversion within 25-30 min to the desired triazoles 35c-37c in 91-95% isolated yield. However, the multi component CuAAC failed when saturated aliphatic bromides were used as substrates, i.e., butyl, heptyl and dodecyl bromides (entries [10][11][12]. We suspect the p-p interactions play an important role in the encapsulation, in water, of the substrates molecule in the hydrophobic cavity of the RG and the lack of these interactions in the aliphatic alkyl bromide does not allow their encapsulation in the RG cavity and hence no observed reaction.

Inclusion complex of RG with benzyl azide (2a) and phenyl acetylene (3b) in D 2 O
We set out to probe the encapsulation of the substrates by the RG. Specically, we have studied the encapsulation of benzyl azide and phenyl acetylene by RG using NMR. The 1 H-NMR spectra of the guest (2a & 3b) were recorded in the presence of the host (RG) in a 1 : 1 molar ratio for 2 mM concentrated solution in D 2 O at 25 C (Fig. 5). In the 1 H-NMR, the aromatic protons (H 2 , H 3 , H 4 ) in 2a were shied up eld from 7.35 ppm to 6.90 ppm upon addition of the RG (Fig. 5a). Similarly, the aromatic protons (H 6 , H 7 , H 8 ) in 3b were shied upeld from 7.35 and 7.45 ppm to 6.98  Table 3 Di-CuAAC of di-alkynes 5b-7b with benzyl azide derivatives 1a-4a in water in the presence of RG a,b a Reaction condition: di-alkyne derivative (1 mmol) and azides (2.1 mmol), Cu(II) (1 mol%), Na-L-ascorbate (3 mol%), RG (1 mol%), water (10 mL), 45 min, 80 C. b Isolated yields. and 7.15 ppm, respectively upon addition of RG (Fig. 5b). The shielding of the guest 1 H NMR resonances upon addition of RG indicated their encapsulation in the RG cavity. The encapsulation of the substrates in the RG cavity may explain the catalytic activity of the RG. In addition, the observation that the CuAAC reactions of the aromatic and p bond containing azides were accelerated in the presence of RG (Table 4), the encapsulation of the guest most likely involves p-p host-guest interactions.
Proposed mechanism for the CuAAC using RG Based upon the reaction catalysed and the encapsulation observed, the proposed mechanism for the CuAAC reaction catalyzed by RG may proceed via step shown in Fig. 6. Starting with inclusion of the alkyne and azide substrates in the pseudo-b-Dglucopyranoside cavity of RG. The closed proximity of the alkyne and azide in the presence of Cu(I) catalyst accelerate the cycloaddition process resulting in the copper-triazole complex followed by protonation and dissociation of the desired 1,4-di substituted 1,2,3triazole (Fig. 6). 19 The binding of the Cu(I) to the multiple triazoles in the RG may also facilitate the coupling reaction inside the RG cavity.

Conclusion
In conclusion, a new resorcinarene sugar conjugate (RG) is reported. It was determined that RG is an efficient catalyst for the CuAAC reaction in water at only 1 mol%. Based upon the reaction catalysis and the encapsulation observed, RG catalytic mechanism, we believe, begins with inclusion of the alkyne and azide substrates resulting in the copper-triazole complex followed by protonation and dissociation to the desired 1,4-di substituted 1,2,3-triazole. RG was found to catalyze the coupling of the alkyne/azide pair and also of the multicomponent alkyl bromide/ sodium azide/alkyne to the 1,4-disubstituted 1,2,3-triazole products in excellent yield within short period of time in water.

Fig. 6
Proposed mechanism for the CuAAC in aqueous media catalyzed by RG.
Typical procedure for three component one-pot CuAAC catalyzed by RG Phenyl acetylene 1b (1 mmol) and aryl/alkyl bromides (1.05 equiv.) with sodium azide (1.1 equiv) were added to a solution of copper(II) sulfate pentahydrates, Na-ascorbate and RG (1 : 3 : 1 mol%) in 10 mL de-ionized distilled water. The reaction mixture was then heated to 80 C for 20-55 min. Aer completion, the solid reaction product was ltered off, dried, and weighed to calculate the isolated yield. When the conversion was low (entries 10-12, Table 4), the reaction mixture was extracted with DCM (2 Â 5 mL). The combined organic layer was collected and dried over MgSO 4 . The solvent was then removed using a rotary evaporator and 1 H-NMR was taken in CDCl 3 .