Substrate-Controlled Direct α-Stereoselective Synthesis of Deoxyglycosides from Glycals Using B(C6F5)3 as Catalyst

B(C6F5)3 enables the metal-free unprecedented substrate-controlled direct α-stereoselective synthesis of deoxyglycosides from glycals. 2,3-Unsaturated α-O-glycoside products are obtained with deactivated glycals at 75 °C in the presence of the catalyst, while 2-deoxyglycosides are formed using activated glycals that bear no leaving group at C-3 at lower temperatures. The reaction proceeds in good to excellent yields via concomitant borane activation of glycal donor and nucleophile acceptor. The method is exemplified with the synthesis of a series of rare and biologically relevant glycoside analogues.


■ INTRODUCTION
Deoxyglycosides are an important class of carbohydrates commonly found in nature as part of biologically active glycoconjugates. 1 These compounds are characterized by the lack of substitution at one or several positions around the carbohydrate ring, often at C-2, which makes them more challenging to synthesized than their fully oxygenated counterparts. The absence of directing groups adjacent to the anomeric center to bias the nucleophile approach during the glycosylation reaction, often leads to mixtures of anomers and/ or  Owing to their relevance in drug discovery, research efforts have focused in achieving their stereoselective synthesis. 3 Coupling reactions involving glycals are one of the most atom-efficient routes to access deoxyglycosides. 4 Traditional methods to yield the corresponding 2-deoxyglycoside rely on acid catalyzed direct addition of an alcohol to the glycal, which often leads to either moderate to low yields and/or variable selectivities which are dependent on the nature of the OH nucleophile (e.g., primary vs secondary, axial vs equatorial) with 2,3-unsaturated glycosides and hydrolyzed starting material as common side products. 2a,3a,q However, most acidcatalyzed processes to access 2,3-unsaturated glycosides directly, which are also versatile synthons in organic chemistry, 5 tend to use harsh promoters, required specific protected building blocks, and often lead to moderate overall yields and diastereocontrol, which has limited their utility. 3a Given the range of glycoside donor and OH nucleophile reactivity profiles, there is currently no universal Lewis acid glycosylation promoter that can be used for the activation of glycals. 3 Thus, there is a need to find improved and more general catalysts to access these high value glycosides.
Our group has been interested over the past few years in the development of practical, selective and catalytic methods for the direct activation of glycals using thiourea-based organo-catalysts, 4b−d as well as palladium and gold catalytic activators. 9 Encouraged by our previous work, we undertook synthetic studies toward the development of a metal-free and improved organocatalytic method for the activation of glycals.
Trivalent boron reagents are often employed as Lewis acids because of their ubiquitous electrophilic nature and ability to reversibly form bonds with oxygen and thus are attractive catalysts in glycosylation chemistry including examples in regioselective glycosylations. 3a,p,6 Among the boron-based Lewis acids available, B(C 6 F 5 ) 3 (BCF) has demonstrated extensive versatility in a wide variety of reactions including borylation, hydrogenation, hydrosilylation, frustrated Lewis pair (FLP) chemistry, and Lewis acid catalysis. 7 In the context of glycosylation chemistry, the utility of BCF has only been shown in the activation of fully substituted trichloroacetimidate glycosyl donors in good to excellent yields and moderate to good diastereoselectivity, 8 with no examples reported in the synthesis of deoxyglycosides.

■ RESULTS AND DISCUSSION
Initial studies began by screening BCF for its ability to promote the stereoselective glycosylation of peracetylated galactal 1a with glucoside acceptor 2a in the presence of different catalyst loadings, solvents, and temperatures. It was found that 5 mol % BCF in toluene at 75°C was the optimum condition to yield the corresponding 2,3-unsaturated glycoside 3a after 2 h (88%, α:β 30:1, entry 1 in Table 1). Reactions were less efficient at lower catalyst loadings or at lower temperatures; changing the solvent to CH 2 Cl 2 or CH 3 CN was also detrimental to the reaction (see Table S1 in the Supporting Information (SI) for solvent and temperature screen details). Having established the optimal reaction conditions, our attention then turned to exploring the substrate scope of the reaction between 1a and a range of OH nucleophiles 2b−2k (Table 1). In all cases, reactions proceeded smoothly within 1.5−4 h and in good to excellent yields and a clear preference for the α-products, demonstrating the reaction is tolerant of primary, secondary, and phenolic OH nucleophiles, as well as common alcohol protecting groups (e.g., acetals, ethers, and esters).
Next, we explored whether "armed" glycosides lacking a leaving group at C-3 could undergo BCF-activation and give substitution products selectively. To probe this, reactions between perbenzylated galactal 9a, acceptor 2a and BCF were screened at different catalyst loadings, solvents and temperatures as before. Best results were found when 5 mol % B(C 6 F 5 ) 3 was used in toluene at 50°C to give 2deoxyglycoside 10a after 2 h (88%, α/β > 30:1, entry 1, Table 2). As before, reactions were less efficient at lower catalyst loadings or temperatures below 50°C and changing Scheme 1. BCF-Catalysed Synthesis of 2,3-Unsaturated Glycosides from "Disarmed" Glycals (Pathway A) and 2-α-Deoxyglycosides from "armed" Glycals (Pathway B) Table 1. Glycosylation Reactions with Galactal 1a c a Isolated yield. b Determined by 1 H-NMR. c Reaction did not proceed in the absence of catalyst. d Activation with BF 3 .OEt 2 afforded 3a in 19% as a 6:1 mixture of anomers.

The Journal of Organic Chemistry
Article the solvent to CH 2 Cl 2 , CH 3 CN or CF 3 Ph was also detrimental (SI Table S2 for full details).
To explore the substrate scope of the glycosylation, galactals 9b−d and glucals 11a−c were reacted with a range of primary and secondary OH nucleophiles 2a, 2e, or 2g under the optimized reaction conditions. In all cases, reactions proceeded smoothly in yields of 62−94% and high α-selectivity (20:1− 30:1 α:β), with secondary OHs requiring longer reaction times (entries 4 and 5 vs 1−3). Subsequently, a series of differentially protected galactals 9b−d and glucals 11a−c bearing benzyl, methoxymethyl acetal, silyl ethers and acetals, and siloxane protecting groups were prepared and subjected to the reaction conditions to investigate the effect of glycal donor on the reaction. Pleasingly, reactions involving all galactals were complete within 1−7 h, in good yields (71−82%) and high αselectivities (20:1 to 30:1 α/β) (entries 6−8). The reaction was also amenable to glycosylations with glucal substrates, albeit required longer reaction times (17 h) and afforded the glycoside products in moderate to good yields (54−86%) with similarly high α-stereocontrol. Siloxane protected donors 11b and 11c gave better yields that perbenzylated glucal 11a (entries 9−13) as expected. 4c These results further highlight that the catalytic system works well across a range of reactivity profiles in both the glycal moiety and nucleophile acceptor.

Article
To probe the mechanism of this versatile reaction, deuterated perbenzylated galactal 15 was reacted with 2a to yield α-glycoside 16a and 16b (90% yield) as a 2:1 mixture of cis/trans products, with a preference for syn addition of both H and O-nucleophile across the double bond (Scheme 4A).
Addition of K 2 CO 3 to the reaction between either 1a or 9a with 2a inhibited the reaction, which supports an acid catalyzed process. Monitoring the reaction between 1a or 9a and hexafluoroisopropanol 17 (Scheme 4B) by 1 H NMR over 60 min at 45°C or 90 min at RT, respectively, only showed anomeric signals corresponding to the starting material and product, without any observable changes in the anomeric ratio of the product throughout the time scales of the reaction (SI Figures S1 and S3). 13 Moreover, subjecting a 4:1 α/βanomeric mixture of 10a to the reaction conditions in the presence of acceptor 2a gave no change in the anomeric ratio (see SI for details). These results suggest the reaction proceeds via short-lived intermediates and that the high α-selectivity is not likely the result of anomerization. 19 F-NMR of the reactions (SI Figures S2 and S4) showed the appearance of fluorinated signals assigned to products 18 and 19, respectively, and also shifts associated with the formation of other BCF species, suggesting the presence of BCF-adducts. Moreover, 1 H NMR spectroscopy studies in Toluene-d 8 of a 1:1 mixture of BCF with galactal donor 1a or 9a, also showed H-shifts associated with the enol ether alkene protons, in each case (SI Figures S5 and S7). 19 F-NMR of the same mixtures showed additional signals associated with several distinct BCFspecies (SI Figures S6 and S8), suggesting activation of the glycal enol ethers by BCF can take place and formation of adducts. Interestingly, 1 H NMR equimolar mixtures of B(C 6 F 5 ) 3 and OH nucleophile 2a at room temperature showed proton shifts associated with 2a ( Figure S9). This effect was more evident in the 19 F-NMR spectra of the same mixtures (SI Figure S10) which showed the shift of the fluorine signals from the catalysts and appearance of different fluorinated species, further supporting the formation of an adduct between the catalyst and the OH nucleophile. This is in agreement to previous reports of glycosyl acceptor activation with boron-based catalysts such as BCF and PhBF 2 in the acid−base activation of trichloroacetimidate glycosyl donors. 8b, 12 As our preliminary findings suggest, BCF could act as a Lewis acid to promote the effective allylic rearrangement 13 (A) of deactivated glycals such as 1a to form transient oxocarbenium ion (B) that can undergo nucleophilic substitution by the BCF-activated nucleophile adduct (H BCFOR) in a stereoselective manner to give 2,3-unsaturated glycosides. In the presence of more reactive glycals, which lack a leaving group at C-3 (e.g., 9a), enol ether direct activation to form oxacarbenium ion (D) might take place, which after nucleophilic substitution by the BCF-activated nucleophile and concomitant protonolysis leads to deoxyglycoside products. In both instances, there is a clear preference for an α-face nucleophilic approach, likely due to sterics and a favorable anomeric effect 14 (Scheme 5, top). However, in the presence of Lewis basic oxygen atoms, BCF coordination to the pyran oxygen in the donor is also possible. Therefore, an acid−base catalyzed mechanism whereby the boron ate adduct promotes both oxocarbenium ion formation and nucleophile activation can not be discarded and it is likely to occur in parallel (Scheme 5, bottom). Further investigations are ongoing to better understand the mechanism of this reaction.

■ CONCLUSIONS
In summary, we have described the unprecedented BCFcatalyzed substrate-controlled stereoselective synthesis of αdeoxyglycosides directly from glycals. We show that 2,3unsaturated α-O-glycoside products are obtained with deactivated glycals at 75°C, while 2-α-deoxyglycosides are formed with activated glycals lacking a leaving group at C-3 at slightly lower temperatures. This metal-free and versatile reaction is applicable to a range of glycal donors and nucleophile acceptors, and is tolerant of most common protecting groups. The reaction proceeds with good to excellent yields and high selectivity for the α-anomer. We exemplify the robustness and utility of the approach in the stereoselective synthesis of a series of oligosaccharides, glycosyl-amino acids, and other glyco-conjugates including rare glycosides analogues of α-L-Rhodinose α-L-cinerulose and α-L-aculose. Work from our lab is currently underway to exploit this chemistry for the stereoselective synthesis of other important glycosides.
2h, 2i, 2j, and 2k were obtained from Sigma-Aldrich. Galactal donors 9b and 9c and glycosyl acceptors 2a, 2c, and 2d were prepared following literature procedures, 4d while glucal 11b and 11c and glycosyl acceptor 2g were synthesized by Balmond et al. reported methods. 4c Dry solvents were obtained by distillation using standard procedures or by passage through a column of anhydrous alumina using equipment from Anhydrous Engineering (University of Bristol) based on the Grubbs' design. Reactions requiring anhydrous conditions were performed under nitrogen; glassware and needles were either flame-dried immediately prior to use or placed in an oven (150°C) for at least 2 h and allowed to cool either in a desiccator or under reduced pressure; liquid reagents, solutions, or solvents were added via syringe through rubber septa; solid reagents were added via Schlenk type adapters. Teflon rings were used between the joints of the condensers and round-bottom flasks. Reactions were monitored by TLC on Kieselgel 60 F254 (Merck). Detection was by examination under UV light (254 nm) and by charring with 10% sulfuric acid in ethanol. Flash column chromatography was performed using silica gel [Merck, 230−400 mesh (40−63 μm)]. Extracts were concentrated in vacuo using both a Buchi rotary evaporator (bath temperatures up to 40°C) at a pressure of either 15 mmHg (diaphragm pump) or 0.1 mmHg (oil pump), as appropriate, and a high vacuum line at room temperature. 1 H NMR and 13 C NMR spectra were measured in the solvent stated at 400 or 500 MHz. Chemical shifts are quoted in parts per million from residual solvent peak (CDCl 3 : 1 H−7.26 ppm and 13C−77.16 ppm) and coupling constants (J) given in Hertz. Multiplicities are abbreviated as b (broad), s (singlet), d (doublet), t (triplet), q (quartet), m (multiplet), or combinations thereof. Mass spectra were determined by the University of Bristol mass spectrometry service by electrospray ionization (SI) modes. The units of the specific rotation, (deg·mL)/(g·dm), are implicit and are not included with the reported value. Concentration c is given in g/ 100 mL.
General Glycosylation Procedure. Glycal donor (1.0 equiv), OH nucleophile acceptor (0.75 equiv), and B(C 6 F 5 ) 3 (5 mol %) were weighed into an oven-dried microwave vial, sealed and placed under vacuum for 1 h. Then the vial was filled with N 2 followed by the addition of ∼1.0 mL of anhydrous toluene. The solutions were stirred and heated at 75°C for Ferrier glycosylation and 50°C for 2-deoxy glycosylation until the reaction was determined to be complete by either TLC or NMR analysis of the crude material (times are given in Tables S1 and S2 and Tables 1 and 2 of the main manuscript). The reaction mixture was concentrated in vacuo and the dried residue was purified by silica gel column chromatography.