AuBr3-catalyzed azidation of per-O-acetylated and per-O-benzoylated sugars

Herein we report, for the first time, the successful anomeric azidation of per-O-acetylated and per-O-benzoylated sugars by catalytic amounts of oxophilic AuBr3 in good to excellent yields. The method is applicable to a wide range of easily accessible per-O-acetylated and per-O-benzoylated sugars. While reaction with per-O-acetylated and per-O-benzoylated monosaccharides was complete within 1–3 h at room temperature, the per-O-benzoylated disaccharides needed 2–3 h of heating at 55 °C.

Of the gold-catalyzed N-glycosylation reactions, Yu et al. demonstrated the effective purine and pyrimidine nucleoside synthesis using per-O-acyl/per-O-benzoyl furanosyl and pyranosyl o-hexynylbenzoates [23]. Subsequently, Hotha and co-workers utilized propargyl 1,2-orthoesters and alkynyl glycosyl carbonate donors for the synthesis of pyrimidine nucleosides [24,25]. In addition, N-glycosides are also accessible by AuCl 3 /phenylacetylene-promoted Ferrier rearrangement of glycals [17], thus, demonstrating the efficient catalysis by alkynophilic and carbophilic Au complexes. Although the alkynophilicity and carbophilicity of Au complexes are well explored, very little is known about the role played by the oxophilicity of gold [26] towards the glycosylation reactions.
Generally, easily accessible per-O-acetylated and per-Obenzoylated sugars are not regarded as effective glycosyl donors in glycosylation reactions since they require harsh reaction conditions due to the deactivating effect of the ester groups. In a recently reported gold(III)-mediated reaction Vankar and co-workers disclosed that a AuCl 3 -phenylacetylene complex promotes the O-glycosylation of armed 1-O-acetyl pyranosides and furanosides [17,27]. The authors also observed that 5 mol % AuCl 3 alone promoted the O-glycosylation albeit in low yields, thus indicating the possible utility of the oxophilic character of Au(III) towards the acetylated sugars.
Next we tested the slightly more Lewis acidic AuCl 3 in this reaction, and found that 10 mol % of AuCl 3 were essential for complete consumption of the starting material. As AuBr 3 is less hygroscopic than AuCl 3 and thus easier to handle, all further experiments were conducted with AuBr 3 only. Interestingly, we noticed that, when stirring the reaction mixture with 4 Å molecular sieves powder to remove moisture prior to the addition of the catalyst no product was formed. This observation suggested that in addition to the coordination of AuBr 3 to the lone pairs of the anomeric acetate carbonyl oxygen, probably the Brønsted acid, HBr, generated from AuBr 3 and water present in the reaction medium is also participating in the catalytic cycle.
Also no reaction was observed when peracetylated galactose and 3 equiv of trimethylsilyl azide were stirred at room temperature in the absence of AuBr 3 as the catalyst. Additionally, the treatment of peracetylated galactose with NaN 3 instead of trimethylsilyl azide at room temperature for 6 h also yielded no product. In view of the above observations, a plausible catalytic cycle is proposed in Supporting Information File 1.
As anticipated, the anomeric azidation of peracetylated ribofuranose (Table 1, entry 7) proceeded well even at 0 °C within 2 h to give the product in 93% yield. However, the azidation of peracetylated 2-deoxy-D-glucosamine was slow and required one equivalent of AuBr 3 and heating at 55 °C for 48 h to reach completion. In this case the desired product β-azido 2,3,4,6acetyl-D-glucosamine (9) could be obtained in 74% yield. The need of using higher amounts of catalyst in this reaction could be attributed to the possible coordination of AuBr 3 with the amide. Having successfully accomplished the gold(III)-catalyzed azido glycosidation of per-O-acetates, we next turned our attention to per-O-benzoylated sugars. Gratifyingly, using 12 mol % AuBr 3 , the easily accessible per-O-benzoylated mannopyranose and glucopyranose ( Table 2, entries 1 and 2) were readily converted into the corresponding 2,3,4,6-tetra-O-benzoyl-α-Dmannopyranosyl azide (10) and 2,3,4,6-tetra-O-benzoyl-β-Dglucopyranosyl azide (11) in excellent yields within 3 h reaction at room temperature. It is noteworthy that the present method can be successfully applied to perbenzoylated sugars with a slightly higher catalyst loading given the fact that they these sugars are more deactivated than the corresponding acetates. Conversely, the reaction of 1,2,3,4-tetra-O-benzoyl-L-rhamnopyranoside (Table 2, entry 3) proceeded within 1 h at room temperature giving 2,3,4-tri-O-benzoyl-α-L-rhamnopyranosyl azide (12) in 71% yield. Furthermore, C5-O-TBDPS-protected perbenzoylated arabinofuranose ( Table 2, entry 4) afforded the desired azide 13 in 70% yield along with some amounts of desilylated product. Furthermore, azidation of perbenzoylated maltose and lactose ( Table 2, entries 5 and 6) did not proceed at room temperature and required heating at 55 °C for 2 h to provide the desired products β-D-maltopyranosyl azide and β-D-lactopyranosyl azide 14 and 15 in 91% and 84% yields, respectively. We found these results very intriguing as the rate of N-glycosylation of benzoylated glycosyl donors which is usually considered low, could be achieved using a cat-alytic amount of the mildly Lewis acidic AuBr 3 and an excellent azide source, trimethylsilyl azide.
Further, we checked the possibility of O-glycosylation and C-glycosylation of peracetylated sugars with 10 mol % AuBr 3 , but the starting materials remained unaffected. Finally, the potential of the gold(III)-catalyzed azidation for large scale applications was demonstrated by performing a gram-scale synthesis on glucose peracetate giving product 2 in 90% yield.

Conclusion
In summary, a facile methodology demonstrating the ability of Au(III) in catalyzing the azidation of deactivated sugars was shown. The reaction proceeds in the absence of molecular sieves without forming lactols as byproducts. This operationally simple protocol enables the synthesis of various N-glycoconjugates offering a wide range of applications and further demonstrates the value of gold catalysis in carbohydrate chemistry.

Experimental
General experimental methods: Chemicals and materials were obtained from commercial sources and used without further purification unless otherwise noted. 1 H and 13 C NMR spectra were recorded on a 400 MHz and 100 MHz spectrometer, respectively using CDCl 3 as the solvent. Chemical shifts (δ) are given in ppm. For perbenzoate compounds 10-15, tetramethyl silane was used as internal standard. Electrospray ionization (ESI) was used for high resolution mass spectrometry (HRMS). An FTIR spectrometer was used for recording IR spectra and only major peaks are reported in cm −1 . Optical rotations were measured on a polarimeter using sodium light (D line at 589 nm). Column chromatography was performed on silica gel (120-200 mesh) using mixtures of ethyl acetate and hexane as the eluents.
General procedure for the anomeric azidation: To a solution of peracetylated or perbenzoylated sugars (300 mg) in 4 mL of dry DCM at room temperature, TMSN 3 (3 equiv) was added followed by the addition of AuBr 3 (amounts of the catalyst are given in Table 1 and Table 2). The reaction mixture was stirred either at room temperature or heated to 55-60 °C as mentioned in the Table 1 and Table 2. Then, the reaction was quenched by adding triethylamine (20 μL). The mixture was concentrated in vacuo and the crude product was purified by column chromatography. Alternatively, the reaction can be quenched by adding sodium bicarbonate solution followed by extraction with DCM (2 × 20 mL). The combined organic layers were washed with water, brine and dried over Na 2 SO 4 and concentrated to dryness. The residue was purified by column chromatography on silica gel using petroleum ether (bp 60-70 °C) and EtOAc.

Supporting Information
Supporting Information File 1 Plausible catalytic cycle, experimental data and copies of 1 H and 13 C NMR spectra of glycosyl azides 1-15 were provided.