Activation of Thioglycosides with Copper(II) Bromide

Reported herein is a new protocol for glycosidation of alkyl and aryl thioglycosides in the presence of copper(II) bromide. While the activation with CuBr2 alone was proven suitable for reactive glycosyl donors, the activation of less reactive donors was more efficient in the presence of triflic acid as an additive. A variety of thioglycoside donors in reactions with different glycosyl acceptors were investigated to determine the initial scope of this reaction.


Introduction
A myriad of approaches have been developed for the chemical synthesis of glycosidic linkages [1][2][3]. Glycosyl halides [4,5], glycosyl imidates [6], and thioglycosides [7] have become the most prominent glycosyl donors utilized in chemical glycosylation and oligosaccharide synthesis. First synthesized by Fischer in 1909 [8], thioglycosides are stable towards a majority of protecting group manipulations albeit can be readily activated in the presence of mild thiophilic reagents [7,[9][10][11][12]. Among a plethora of known thiophilic promoters, transition metals have been used for decades. In fact, the first known method for thioglycoside activation makes use of mercury(II) salts [13]. Nevertheless, only recently the activation of conventional thioglycosides through the direct coordination of a metal salt with the anomeric sulfur has been investigated in detail. First reported by Pohl et al., a sub-stoichiometric amount of Ph 3 Bi(OTf) 2 was able to activate propylthio glycosides [14,15]. Subsequently, Sureshan et al. [16] performed activation of thioglycosides using a sub-stoichiometric amount of AuCl 3 . Zhu et al. [17] also showed that propargylthio glycosyl donors are activated through the direct coordination of Au(III) to the sulfur atom rather than the remote pathway via the alkyne functionality. As a part of our efforts toward the development of novel methods for glycosylation, previously we reported that alkyl/aryl thioglycosides can be activated with palladium(II) bromide (PdBr 2 ) [18]. It has been found that the activation can be performed in the presence of PdBr 2 only, but an additive (propargyl bromide) accelerates the activation process. A preliminary mechanistic analysis relying on 1 H NMR spectroscopy revealed that propargyl bromide could form a more reactive reaction intermediate and possibly acts as the leaving group scavenger. For example, when thiogalactoside 1 was activated for glycosidation with glycosyl acceptor 2 in the presence of PdBr 2 only, the reaction was slow and disaccharide 3 was obtained in 76% yield in 48 h. However, in the presence of propargyl bromide, disaccharide 3 was obtained in 96% yield within 24 h (Scheme 1A). It was postulated that the reaction proceeds via the intermediacy of complex A. In an effort to identify greener and more accessible transition metal salts, reported herein is our investigation of copper(II)-promoted activation of thioglycosides.
By means of a personal communication with Professor Marra, the authors hav that the choice of the reaction solvent is essential. Marra and co-workers observed activation of thioglycosides promoted by Cu(OTf)2 proceeded much faster in ace (or MeCN/DCM), a solvent in which Cu(OTf)2 is highly soluble. The solvent w pected to act as a ligand judged by the color change of the reaction mixture. Ma co-workers have also suspected that reactions in MeCN proceeded through a sin tron transfer, whereas for reactions in DCM the copper salt acted as a thiophilic was also noted that these reaction conditions were better suited for the activ "armed" glycosyl donors such as per-O-benzylated thioglycoside 4. These obse were made in 2005, and to the best of our knowledge no follow-up investigation emerged.
To elaborate on this discovery, for our own study we chose copper(II) brom pric bromide) as a significantly cheaper and less moisture sensitive altern Cu(OTf)2. For the initial experimentation, we selected a highly reactive galactosyl [23] equipped with the superarming protecting group pattern [24], and a conv primary glycosyl acceptor 2 [25]. When the reaction of glycosyl donor 1 with ac was promoted with 0.8 equiv of CuBr2 in DCM at rt, disaccharide 3 [26] was obt 45% in 24 h (Table 1, entry 1). Complete 1,2-trans selectivity achieved in this re due to the participation of the neighboring 2-O-benzoyl ester group, a well-know in carbohydrate chemistry [27]. Encouraged by this promising result, we perform vations in the presence of 1.5, 2.0, and 2.5 equiv of CuBr2. As a result, disaccharid

Results and Discussion
Our interest to copper(II) was sparked by a report by Dondoni, Marra, and Massi wherein the synthesis of disaccharides was achieved by the activation of ethylthio glycosides promoted by copper(II) triflate [19]. We were fascinated by these results because Cu(OTf) 2 is a well-known activator for reactive glycosyl donors, such as halides 5 or thioimidates [20][21][22], but not conventional alkyl/aryl thioglycosides. In accordance with the reported protocol, thioglycoside 4 was reacted with excess of acceptor 2 to afford the corresponding disaccharide 5 in a high yield albeit with poor stereoselectivity (Scheme 1B) [19].
By means of a personal communication with Professor Marra, the authors have learnt that the choice of the reaction solvent is essential. Marra and co-workers observed that the activation of thioglycosides promoted by Cu(OTf) 2 proceeded much faster in acetonitrile (or MeCN/DCM), a solvent in which Cu(OTf) 2 is highly soluble. The solvent was suspected to act as a ligand judged by the color change of the reaction mixture. Marra and co-workers have also suspected that reactions in MeCN proceeded through a single-electron transfer, whereas for reactions in DCM the copper salt acted as a thiophilic metal. It was also noted that these reaction conditions were better suited for the activation of "armed" glycosyl donors such as per-O-benzylated thioglycoside 4. These observations were made in 2005, and to the best of our knowledge no follow-up investigation has yet emerged.
To elaborate on this discovery, for our own study we chose copper(II) bromide (cupric bromide) as a significantly cheaper and less moisture sensitive alternative to Cu(OTf) 2 . For the initial experimentation, we selected a highly reactive galactosyl donor 1 [23] equipped with the superarming protecting group pattern [24], and a conventional primary glycosyl acceptor 2 [25]. When the reaction of glycosyl donor 1 with acceptor 2 was promoted with 0.8 equiv of CuBr 2 in DCM at rt, disaccharide 3 [26] was obtained in 45% in 24 h (Table 1, entry 1). Complete 1,2-trans selectivity achieved in this reaction is due to the participation of the neighboring 2-O-benzoyl ester group, a well-known effect in carbohydrate chemistry [27]. Encouraged by this promising result, we performed activations in the presence of 1.5, 2.0, and 2.5 equiv of CuBr 2 . As a result, disaccharide 3 was obtained in 70, 75, and 96% yield, respectively (entries 2-4). Therefore, we chose to conduct all subsequent experiments in the presence of 2.5 equiv of the promoter. solvent did not result in the anticipated gains. Although reactions in MeCN/1,2-DCE or neat MeCN were significantly faster at rt, 6 h or 1.5 h, the yields of disaccharide 3 were also reduced to 70 or 80%, respectively (entries 8 and 9). As a result of this optimization study, we identified reaction conditions listed in entry 6, CuBr2 (2.5 equiv) in 1,2-DCE at rt, as the most suitable conditions for the activation of donor 2. It should be noted that other copper(II) salts such as Cu(OTf)2 or CuCl2 gave inferior results under these reaction conditions. Thus, disaccharide 3 was obtained in 24 or 16% yield, respectively, in 24 h (entries 10 and 11). An attempt to perform the activation in the presence of copper(I) bromide resulted in no reaction after 24 h (entry 12). When the developed reaction conditions comprising 2.5 equiv CuBr2 in 1,2-DCE at rt were applied to the activation of per-O-benzoylated glycosyl donor 6 [28,29] for reaction with acceptor 2, the corresponding disaccharide 7 [30] was obtained in a disappointing yield of 36% after 24 h (entry 13). This result was not totally unexpected due to a These reactions were not swift and required 24 h to complete, and to enhance the reaction rates, we investigated the effect of the reaction temperature. When the reaction was performed at 40 • C, it was completed within 4 h. However, the shorter reaction time has also translated into a reduced yield of 84% for disaccharide 3 (entry 5). We then investigated the effect of the reaction solvent. When the reaction was carried out in 1,2-DCE at rt, disaccharide 3 was produced in an excellent yield of 98% and the reaction time was slightly reduced to 22 h (entry 6). Like in the case of reactions in DCM, increasing the reaction temperature (to 80 • C in this case) led to a decreased reaction time to 3 h, but also a decreased yield of 85% for disaccharide 3 (entry 7). Further investigation of the reaction solvent did not result in the anticipated gains. Although reactions in MeCN/1,2-DCE or neat MeCN were significantly faster at rt, 6 h or 1.5 h, the yields of disaccharide 3 were also reduced to 70 or 80%, respectively (entries 8 and 9). As a result of this optimization study, we identified reaction conditions listed in entry 6, CuBr 2 (2.5 equiv) in 1,2-DCE at rt, as the most suitable conditions for the activation of donor 2. It should be noted that other copper(II) salts such as Cu(OTf) 2 or CuCl 2 gave inferior results under these reaction conditions. Thus, disaccharide 3 was obtained in 24 or 16% yield, respectively, in 24 h (entries 10 and 11). An attempt to perform the activation in the presence of copper(I) bromide resulted in no reaction after 24 h (entry 12).
When the developed reaction conditions comprising 2.5 equiv CuBr 2 in 1,2-DCE at rt were applied to the activation of per-O-benzoylated glycosyl donor 6 [28,29] for reaction with acceptor 2, the corresponding disaccharide 7 [30] was obtained in a disappointing yield of 36% after 24 h (entry 13). This result was not totally unexpected due to a significantly lower reactivity of donor 6 due to its disarming protecting group pattern [31] (all esters) in comparison with the superarmed donor 1. Hence, we conducted further optimization of reaction conditions to enhance glycosylations with donor 6. This included varying all major factors, solvent, temperature, and additives, while keeping the amount of CuBr 2 constant. Pilot investigation of the reaction solvent and temperature led to only marginal improvements. Thus, reaction in MeCN produced disaccharide 7 in 50% yield in 6 h (entry 14). Reaction in 1,2-DCE performed at 80 o C led to the formation of 7 in 70% in 4 h (entry 15). The desired solution was found by implementing TfOH as an additive for this reaction, and the optimal amount was found to be 0.5 equiv. Thus, glycosidation of donor 6 with acceptor 2 in the presence of CuBr 2 (2.5 equiv) and TfOH (0.5 equiv) in DCE at rt led to the formation of disaccharide 7 in 96% yield in 24 h (entry 16)! A control experiment performed with the reactive glycosyl donor 1 in the presence of CuBr 2 (2.5 equiv) and TfOH (0.5 equiv) in DCE at rt led to disaccharide 3 in a reduced yield of 73% (entry 17). This result was rationalized by the enhanced rate of hydrolysis of the glycosyl donor that is taking place in the presence of TfOH (see the SI for NMR monitoring experiments). Also observed was that in the absence of glycosyl acceptor, the glycosyl bromide will be slowly produced in the presence of CuBr 2 alone. Traces were present between 4 and 20 h. In the presence of CuBr 2 and TfOH, hemiacetal, the product of donor hydrolysis is produced first (2-4 h), and then slow production of the glycosyl bromide begins (20 h).

Experimental
General. All chemicals used were reagent grade and used as supplied. The ACS grade solvents used for reactions were purified and dried in accordance with standard procedures. Column chromatography was performed on silica gel 60 (70-230 mesh), reactions were monitored by TLC on Kieselgel 60 F254. The compounds were detected by examination under UV light and by charring with 10% sulfuric acid in methanol. Solvents were removed under reduced pressure at <40 °C. CH2Cl2 and ClCH2CH2Cl were distilled from CaH2 directly prior to the application. Molecular sieves (3 Å), used for reactions, were crushed and activated in vacuo at 390 °C during 8 h in the first instance and then for 2-3 h at 390 °C directly prior to application. 1 H NMR spectra were recorded at 400 MHz, 13 C NMR spectra were recorded at 100 MHz. The 1 H NMR chemical shifts are referenced to tetramethylsilane (δH = 0 ppm) or CHCl3 (δH = 7.26 ppm) for 1 H NMR spectra for solutions in CDCl3. Anomeric ratios (if applicable) were determined by comparison of the integral intensities of relevant signals in 1 H NMR spectra (see the Supplementary Material).

Entry
Acceptor Product, Yield

Experimental
General. All chemicals used were reagent grade and used as supplied. The ACS grade solvents used for reactions were purified and dried in accordance with standard procedures. Column chromatography was performed on silica gel 60 (70-230 mesh), reactions were monitored by TLC on Kieselgel 60 F254. The compounds were detected by examination under UV light and by charring with 10% sulfuric acid in methanol. Solvents were removed under reduced pressure at <40 °C. CH2Cl2 and ClCH2CH2Cl were distilled from CaH2 directly prior to the application. Molecular sieves (3 Å), used for reactions, were crushed and activated in vacuo at 390 °C during 8 h in the first instance and then for 2-3 h at 390 °C directly prior to application. 1 H NMR spectra were recorded at 400 MHz, 13 C NMR spectra were recorded at 100 MHz. The 1 H NMR chemical shifts are referenced to tetramethylsilane (δH = 0 ppm) or CHCl3 (δH = 7.26 ppm) for 1 H NMR spectra for solutions in CDCl3. Anomeric ratios (if applicable) were determined by comparison of the integral intensities of relevant signals in 1 H NMR spectra (see the Supplementary Material).

Experimental
General. All chemicals used were reagent grade and used as supplied. The ACS grade solvents used for reactions were purified and dried in accordance with standard procedures. Column chromatography was performed on silica gel 60 (70-230 mesh), reactions were monitored by TLC on Kieselgel 60 F254. The compounds were detected by examination under UV light and by charring with 10% sulfuric acid in methanol. Solvents were removed under reduced pressure at <40 °C. CH2Cl2 and ClCH2CH2Cl were distilled from CaH2 directly prior to the application. Molecular sieves (3 Å), used for reactions, were crushed and activated in vacuo at 390 °C during 8 h in the first instance and then for 2-3 h at 390 °C directly prior to application. 1 H NMR spectra were recorded at 400 MHz, 13 C NMR spectra were recorded at 100 MHz. The 1 H NMR chemical shifts are referenced to tetramethylsilane (δH = 0 ppm) or CHCl3 (δH = 7.26 ppm) for 1 H NMR spectra for solutions in CDCl3. Anomeric ratios (if applicable) were determined by comparison of the integral intensities of relevant signals in 1 H NMR spectra (see the Supplementary Material).

Experimental
General. All chemicals used were reagent grade and used as supplied. The ACS grade solvents used for reactions were purified and dried in accordance with standard procedures. Column chromatography was performed on silica gel 60 (70-230 mesh), reactions were monitored by TLC on Kieselgel 60 F254. The compounds were detected by examination under UV light and by charring with 10% sulfuric acid in methanol. Solvents were removed under reduced pressure at <40 °C. CH2Cl2 and ClCH2CH2Cl were distilled from CaH2 directly prior to the application. Molecular sieves (3 Å), used for reactions, were crushed and activated in vacuo at 390 °C during 8 h in the first instance and then for 2-3 h at 390 °C directly prior to application. 1 H NMR spectra were recorded at 400 MHz, 13 C NMR spectra were recorded at 100 MHz. The 1 H NMR chemical shifts are referenced to tetramethylsilane (δH = 0 ppm) or CHCl3 (δH = 7.26 ppm) for 1 H NMR spectra for solutions in CDCl3. Anomeric ratios (if applicable) were determined by comparison of the integral intensities of relevant signals in 1 H NMR spectra (see the Supplementary Material).

Experimental
General. All chemicals used were reagent grade and used as supplied. The ACS grade solvents used for reactions were purified and dried in accordance with standard procedures. Column chromatography was performed on silica gel 60 (70-230 mesh), reactions were monitored by TLC on Kieselgel 60 F254. The compounds were detected by examination under UV light and by charring with 10% sulfuric acid in methanol. Solvents were removed under reduced pressure at <40 °C. CH2Cl2 and ClCH2CH2Cl were distilled from CaH2 directly prior to the application. Molecular sieves (3 Å), used for reactions, were crushed and activated in vacuo at 390 °C during 8 h in the first instance and then for 2-3 h at 390 °C directly prior to application. 1 H NMR spectra were recorded at 400 MHz, 13 C NMR spectra were recorded at 100 MHz. The 1 H NMR chemical shifts are referenced to tetramethylsilane (δH = 0 ppm) or CHCl3 (δH = 7.26 ppm) for 1 H NMR spectra for solutions in CDCl3. Anomeric ratios (if applicable) were determined by comparison of the integral intensities of relevant signals in 1 H NMR spectra (see the Supplementary Material).

Experimental
General. All chemicals used were reagent grade and used as supplied. The ACS grade solvents used for reactions were purified and dried in accordance with standard procedures. Column chromatography was performed on silica gel 60 (70-230 mesh), reactions were monitored by TLC on Kieselgel 60 F254. The compounds were detected by examination under UV light and by charring with 10% sulfuric acid in methanol. Solvents were removed under reduced pressure at <40 °C. CH2Cl2 and ClCH2CH2Cl were distilled from CaH2 directly prior to the application. Molecular sieves (3 Å), used for reactions, were crushed and activated in vacuo at 390 °C during 8 h in the first instance and then for 2-3 h at 390 °C directly prior to application. 1 H NMR spectra were recorded at 400 MHz, 13 C NMR spectra were recorded at 100 MHz. The 1 H NMR chemical shifts are referenced to tetramethylsilane (δH = 0 ppm) or CHCl3 (δH = 7.26 ppm) for 1 H NMR spectra for solutions in CDCl3. Anomeric ratios (if applicable) were determined by comparison of the integral intensities of relevant signals in 1 H NMR spectra (see the Supplementary Material).

Experimental
General. All chemicals used were reagent grade and used as supplied. The ACS grade solvents used for reactions were purified and dried in accordance with standard procedures. Column chromatography was performed on silica gel 60 (70-230 mesh), reactions were monitored by TLC on Kieselgel 60 F254. The compounds were detected by examination under UV light and by charring with 10% sulfuric acid in methanol. Solvents were removed under reduced pressure at <40 °C. CH2Cl2 and ClCH2CH2Cl were distilled from CaH2 directly prior to the application. Molecular sieves (3 Å), used for reactions, were crushed and activated in vacuo at 390 °C during 8 h in the first instance and then for 2-3 h at 390 °C directly prior to application. 1 H NMR spectra were recorded at 400 MHz, 13 C NMR spectra were recorded at 100 MHz. The 1 H NMR chemical shifts are referenced to tetramethylsilane (δH = 0 ppm) or CHCl3 (δH = 7.26 ppm) for 1 H NMR spectra for solutions in CDCl3. Anomeric ratios (if applicable) were determined by comparison of the integral intensities of relevant signals in 1 H NMR spectra (see the Supplementary Material).

Experimental
General. All chemicals used were reagent grade and used as supplied. The ACS grade solvents used for reactions were purified and dried in accordance with standard procedures. Column chromatography was performed on silica gel 60 (70-230 mesh), reactions were monitored by TLC on Kieselgel 60 F254. The compounds were detected by examination under UV light and by charring with 10% sulfuric acid in methanol. Solvents were removed under reduced pressure at <40 °C. CH2Cl2 and ClCH2CH2Cl were distilled from CaH2 directly prior to the application. Molecular sieves (3 Å), used for reactions, were crushed and activated in vacuo at 390 °C during 8 h in the first instance and then for 2-3 h at 390 °C directly prior to application. 1 H NMR spectra were recorded at 400 MHz, 13 C NMR spectra were recorded at 100 MHz. The 1 H NMR chemical shifts are referenced to tetramethylsilane (δH = 0 ppm) or CHCl3 (δH = 7.26 ppm) for 1 H NMR spectra for solutions in CDCl3. Anomeric ratios (if applicable) were determined by comparison of the integral intensities of relevant signals in 1 H NMR spectra (see the Supplementary Material).

Experimental
General. All chemicals used were reagent grade and used as supplied. The ACS grade solvents used for reactions were purified and dried in accordance with standard procedures. Column chromatography was performed on silica gel 60 (70-230 mesh), reactions were monitored by TLC on Kieselgel 60 F254. The compounds were detected by examination under UV light and by charring with 10% sulfuric acid in methanol. Solvents were removed under reduced pressure at <40 °C. CH2Cl2 and ClCH2CH2Cl were distilled from CaH2 directly prior to the application. Molecular sieves (3 Å), used for reactions, were crushed and activated in vacuo at 390 °C during 8 h in the first instance and then for 2-3 h at 390 °C directly prior to application. 1 H NMR spectra were recorded at 400 MHz, 13 C NMR spectra were recorded at 100 MHz. The 1 H NMR chemical shifts are referenced to tetramethylsilane (δH = 0 ppm) or CHCl3 (δH = 7.26 ppm) for 1 H NMR spectra for solutions in CDCl3. Anomeric ratios (if applicable) were determined by comparison of the integral intensities of relevant signals in 1 H NMR spectra (see the Supplementary Material).

Experimental
General. All chemicals used were reagent grade and used as supplied. The ACS grade solvents used for reactions were purified and dried in accordance with standard procedures. Column chromatography was performed on silica gel 60 (70-230 mesh), reactions were monitored by TLC on Kieselgel 60 F 254 . The compounds were detected by examination under UV light and by charring with 10% sulfuric acid in methanol. Solvents were removed under reduced pressure at <40 • C. CH 2 Cl 2 and ClCH 2 CH 2 Cl were distilled from CaH 2 directly prior to the application. Molecular sieves (3 Å), used for reactions, were crushed and activated in vacuo at 390 • C during 8 h in the first instance and then for 2-3 h at 390 • C directly prior to application. 1 H NMR spectra were recorded at 400 MHz, 13 C NMR spectra were recorded at 100 MHz. The 1 H NMR chemical shifts are referenced to tetramethylsilane (δ H = 0 ppm) or CHCl 3 (δ H = 7.26 ppm) for 1 H NMR spectra for solutions in CDCl 3 . Anomeric ratios (if applicable) were determined by comparison of the integral intensities of relevant signals in 1 H NMR spectra (see the Supplementary Material).

Method A-General Procedure for Glycosidation of Thioglycosides in the Presence of CuBr 2
A mixture containing thioglycoside donor (30 mg, 0.040-0.050 mmol), glycosyl acceptor (0.030-0.040 mmol), and freshly activated molecular sieves (3 Å, 90 mg) in dry 1,2-dichloroethane (C 2 H 4 Cl 2 , DCE, 1.0 mL) was stirred under argon for 1 h at rt. After that, copper bromide (CuBr 2 , 2.5 equiv to donor, 0.100-0.125 mmol) was added and the reaction mixture was stirred for 24 h at rt. The solids were filtered off through a pad of Celite and rinsed successively with DCM. The combined filtrate (~20 mL) was washed with H 2 O (2 × 5 mL). The organic phase was separated, dried with MgSO 4 , and concentrated under reduced pressure. The residue was purified by column chromatography on silica gel (ethyl acetate-hexanes gradient elution) to afford a disaccharide derivative in yields listed in tables and below.

Method B-General Procedure for Glycosidation of Thioglycosides in the Presence of CuBr 2 and TfOH
A mixture containing thioglycoside donor (30 mg, 0.040-0.050 mmol), glycosyl acceptor (0.030-0.040 mmol), and freshly activated molecular sieves (3 Å, 90 mg) in dry 1,2-dichloroethane (C 2 H 4 Cl 2 , DCE, 1.0 mL) was stirred under argon for 1 h at rt. After that, copper bromide (CuBr 2 , 2.5 equiv to donor, 0.100-0.125 mmol) and TfOH (0.50 equiv to donor, 0.020-0.025) was added and the reaction mixture was stirred for 24 h at rt. The solids were filtered off through a pad of Celite and rinsed successively with DCM. The combined filtrate (~20 mL) was washed with H 2 O (2 × 5 mL). The organic phase was separated, dried with MgSO 4 , and concentrated under reduced pressure. The residue was purified by column chromatography on silica gel (ethyl acetate-hexanes gradient elution) to afford a disaccharide derivative in yields listed in tables and below.

Conclusions
A new method for the activation of thioglycosides has been developed. The activation with CuBr 2 can be sluggish with unreactive thioglycoside donors. In these cases, the outcome can be improved in the presence of triflic acid additive. Upon standardizing the basic reaction conditions for both reactive and unreactive glycosyl donors, further examination of various thioglycosides has been performed. In most cases, our activation system was effective and predictable, but the product yields were largely dependent on the reactivity of the glycosyl donor. We believe that this study will fundamentally contribute to other developments of copper-catalyzed reactions with carbohydrates [66][67][68][69][70][71][72][73][74][75]. Further optimization of the reaction conditions, investigation of the reaction mechanism, and its application in automated synthesis of glycans are currently underway in our laboratory.