Electrochemical Glycosylation via Halogen-Atom-Transfer for C-Glycoside Assembly

Glycosyl donor activation emerged as an enabling technology for anomeric functionalization, but aimed primarily at O-glycosylation. In contrast, we herein disclose mechanistically distinct electrochemical glycosyl bromide donor activations via halogen-atom transfer and anomeric C-glycosylation. The anomeric radical addition to alkenes led to C-alkyl glycoside synthesis under precious metal-free reaction conditions from readily available glycosyl bromides. The robustness of our e-XAT strategy was further mirrored by C-aryl and C-acyl glycosides assembly through nickela-electrocatalysis. Our approach provides an orthogonal strategy for glycosyl donor activation with expedient scope, hence representing a general method for direct C-glycosides assembly.


INTRODUCTION
C-glycosides represent a privileged carbohydrate motif in natural products and pharmaceutical compounds. 1The diverse biological functions led to a shift in focus from O-glycosides to more stable C-glycosides over the last decades and have thus revolutionized antiviral treatments, exemplified by the widespread use of C-ribosyl nucleoside prodrug, Remdesivir, in the early stage of COVID-19 infection. 2Additionally, the replacement of uridine with C-glycoside pseudo uridine led to the development of effective mRNA vaccines against COVID-19, culminating in the 2023 Nobel Prize for Physiology and Medicine for groundbreaking nucleoside base modification. 3Thus, selective glycosidic C−C bond construction methods are of significant value for enriching the viable C-glycoside motifs. 4Recent advances in metal-catalyzed cross-couplings, 5 photochemistry, 6 and C−H glycosylation 7 provide transformative platforms for C-glycosylation.Despite indisputable advances, limitations, such as the tedious synthesis of glycosyl donors, expensive photocatalysts, and the use of stochiometric amounts of oxidants or reductants, remain to be considered for a sustainable transformation.
In the recent decade, halogen-atom transfer (XAT) has gathered considerable attention, 8 since it proved to be an efficient strategy to activate alkyl halides by exploiting, inter alia, nucleophilic α-amino alkyl radicals as XAT reagent, which can be formed through the single-electron (SET) oxidation of simple tertiary amines by excited photocatalysts (Scheme 1a). 9Notably, while the electrochemical anodic oxidation of tertiary amines is a known reaction, it is typically regarded as a sacrificial oxidation event aimed at preserving anodic electrodes in reductive electrochemical transformations (Scheme 1a). 10 The effort to further explore the ability of α-aminoalkyl radicals as HAT reagents in electrochemistry remains rare. 11lectrochemical glycosylation, pioneered by Noyori and Kurimoto 12 and Yoshida et al., 13 was established as a reliable method for glycosyl donor activation using thio-/seleno-/ telluro-glycoside donors, particularly in synthesizing O-glycosides via oxocarbenium cation species (Scheme 1b). 14However, the electrochemical C-glycosylation has, to the best of our knowledge, thus far not been described.Within our program on C-glycosylation, 15 and electrocatalysis, 16 we questioned whether the electrochemically generated α-amino alkyl radical could serve as an in situ generated XAT reagent to enable glycosyl halide activation, affording glycosyl radical intermediate for subsequent glycosylations.This indeed may induce unprecedented reactivity profiles not achievable with oxocarbenium ion intermediates.Herein, we have devised an e-XAT-mediated anomeric addition for C-alkyl glycoside synthesis with the dual-functional Hunig base diisopropylethyl-amine (DIPEA) as the XAT reagent and anodic sacrificial reagent at exceptionally mild reaction conditions.Remarkably, the e-XAT process was likewise paired with nickel-electrocatalysis, enabling the assembly of C-aryl glycosides.Moreover, both anomers of C-aryl glycosides and C-acyl glycosides could be stereoselectively obtained through the judicious choice of the supporting ligand control 17 (Scheme 1c).

RESULTS AND DISCUSSION
Glycosyl halides represent user-friendly glycosyl donors 18 and have been explored for anomeric Giese-addition with organotin hydrides via tin-radical initiated XAT process. 19In the pursuit of sustainable C-glycosylation, tin-free variants have been developed utilizing either photoredox catalysis 20 or metal catalysis. 21In this context, we initially commenced our studies for the envisioned electrochemical anomeric addition of glycosyl donors with 4-vinyl-1,1′-biphenyl (2aa) to access C-alkyl glycoside (Table 1).Thus, glycosyl sulfones 1c−1f were probed with a zinc plate as the anode and graphite felt (GF) as the cathode material, affording product 3a either in a trace amount of product or low yields.Glycosyl bromide 1a proved efficient to generate the desired product 3a in 87% yield (Table 1, entry 1).By contrast, glycosyl chloride 1b was an unsuitable glycosyl donor for electrochemical C-glycosylation.Notably, the byproducts, such as 3b, were observed due to the electrochemical 2e − reduction 22 and the subsequent β-OAcelimination.Next, we explored replacing the Zn anode with GF and screened different sacrificial amines, such as Et 3 N, DIPEA, and piperidine.DIPEA was identified as the best choice and thus avoided the use of the Zn plate as a sacrificial anode in undivided cell (Table 1, entries 2−4).Replacing GF anode with Pt delivered product 3a in 63% yield (Table 1, entry 6).A slight modification greatly improved the yield to 89% with glycosyl bromide 1a as the limiting reagent which helps to Scheme 1. Halogen-Atom Transfer (XAT) and Electrochemical C-Glycosylation reduce the formation of byproduct 3b (Table 1, entries 7−9).The control experiments verified the essential roles of the amine and the electricity (Table 1, entries 10 and 11).
To demonstrate the applicability of the electrochemical anomeric addition reaction (Scheme 3), structurally complex alkenes 2aq−2as were examined for direct late-stage modification, giving products 25−27.The reaction with acrylate 2at bearing a galactose moiety enabled the Cdisaccharide assembly (28).When hybrid glycosyl bromide donors bearing natural products and drug derivatives, such as ibuprofen 1m, ciprofibrate 1n, and oxaprozin 1o, were employed as the substrates, highly functionalized glycoconjugates (29−31) were selectively and efficiently obtained.Furthermore, disaccharides, including lactose 1p, cellobiose 1q, isomaltose 1r, and maltose 1s, proved suitable for versatile C-glycosides assembly in a single reaction (32−35).The robustness of our e-XAT protocol was mirrored by the Scheme 2. Scope of the Alkenes and Glycosyl Bromides maltotriosyl radical addition, yielding product 36 in 65% yield.
Additionally, the glycosyl radical with the dehydroalanine moiety gave glycosylated-amino acid (37−41) Encouraged by the success of our e-XAT for C-alkyl glycoside synthesis, we wondered whether the merger of the anodic e-XAT process with cathodic reductive nickel catalysis would indeed set the stage for the assembly of synthetically relevant C-aryl glycoside.This paired nickel-catalyzed reductive electrocatalysis would thereby offer a sustainable solution to address the current limitations of metal-catalyzed reductive C-glycosylation, such as the use of external stochiometric metal reductants, expensive photocatalysts in metalla-photoredox catalysis, sacrificial anode materials, or difficult to access glycosyl donors.Moreover, this paired electrocatalysis represents a conceptually distinct strategy for halogen-atom transfer in molecular synthesis.
To this end, we probed the envisioned e-XAT reductive cross-electrophile coupling 23 using galactosyl bromide 1a and methyl 4-iodobenzoate 2ba with the aid of nickelaelectrocatalysis (Scheme 4).
The desired product 42a was obtained in 19% yield, albeit with a low anomeric selectivity (α: β = 3:1) with Ni(acac) 2 as Scheme 3. Scope of the Electrochemical Glycoconjugation catalyst, bipyridine L1 as ligand, and DIPEA serving both as XAT-agent and reductant in an undivided cell.Next, a set of bidentate bipyridine ligands L2−L4 were tested and revealed that the substituents at the para-position can considerably alter the yield (L3 and L4).With a substituent at the ortho-position (L2) the α-anomeric arylation product was selectively formed.A similar substituent effect was observed when employing phenanthroline ligands.Also, here the ligand-bearing C2substituents outperformed ligands with groups at the C3-or C4-position (L5-L10) in terms of reactivity and selectivity.Notably, the ligand L6 gave the desired product in 51% yield with an excellent α:β ratio of 20:1.A slight increase in the amount of DIPEA resulted in a further improved yield of 72%, notably without loss of anomeric selectivity.Interestingly, this selectivity could be switched to anomeric β-arylation product 42ba, when exploiting the tridentate ligands L11 and L12.
Here, L12 afforded the desired product 42ba with a moderate yield and excellent anomeric selectivity.The yield could be further improved with increased amounts of the Hunig base.Remarkably, this e-XAT-mediated nickel-electrocatalysis could be transferred to glycosyl acylation using commercially available benzoyl chloride as the acylating reagent.The anomeric selectivity could again be controlled by the judicious choice of the ligand.Hence, L2 afforded the C-acyl glycosides 43a with a high level of α-selectivity, while the β-acylation product 43ba was obtained with tridentate ligand L11.Thus, our strategy provides a stereodivergent, and sustainable assembly of C-acyl-glycosides, 24 thereby avoiding rather difficult-to-access glycosyl donors. 25heme 5. Scope of Nickelaelectro-Catalyzed Glycosyl α-Arylation Then, the robustness of our nickelaelectro-catalyzed reductive glycosyl arylation was probed with a diverse array of substrates featuring functional groups, such as ester (42a and 49), and trifluoromethyl (45 and 46).Likewise, cyano (47) and ketone (48) groups were well tolerated in the nickelaelectro-catalyzed cross-electrophile coupling.Furthermore, structurally complex glycosyl bromides, such as isomaltose, lactose, cellobiose, and maltotriose, were arylated with excellent α-selectivities (50-53) (Scheme 5).
Of particular note, our nickelaelectrocatalyzed C-glycosylation offers stereodivergent access to both αand β-isomers of C-aryl glycosides and C-acyl glycosides by nickel catalyst with slightly different tridentate terpyridine ligands (Scheme 6).The terpyridine/Nickel catalytic system outcompetes the inherent stereoelectronic effects of α-isomer due to the steric hindrance of C2 substituents with axial ligated nickel complex, leading to exclusive β-isomers.The scope of nickelaelectrocatalytic β-C-aryl glycosylation was explored, and excellent anomeric selectivities and functional group tolerance were observed (42ba−42be).Notably, the β-acylation, which has not been reported, likewise demonstrated feasible albeit with moderate yields with galactosyl bromide (43ba−43be).
To gain insights into the working mode of the e-XATmediated C-glycosylation, the formal electroreductive C-alkyl glycoside synthesis was initially probed using stoichiometric zinc powder as a reductant.The failure to form the Gieseaddition product 21a suggested that single-electron reduction of glycosyl bromide 1g was not operative (Scheme 7A). 26urthermore, a deuterium labeling experiment yielded the product [D]-3 with 70% deuterium incorporation at the benzylic position, being suggestive of a cathodic reduction of the benzylic radical to the corresponding anion to be involved(Scheme 7B).To elucidate the e-XAT process, a paired electrolysis was attempted to trap the α-amino alkyl radical.With dicyanobenzene 2bh as a cathodic reducing reagent product 54 was isolated in 58% yield, highlighting a formed α-amino alkyl radical through the anodic oxidation of the Hunig base (E ox = 0.86 V) (Scheme 7C). 27A control experiment without the Hunig base resulted only in trace amounts of the product, further supporting the key role of the α-amino alkyl radical in the e-XAT process (Scheme 7D).Scheme 6. Scope of Nickelaelectro-Catalyzed Glycosyl β-Arylation and β-Acylation Then, the Ni−I complex was prepared from the facile oxidative addition. 28A catalytic amount of Ni−I complex was employed in the electroreductive cross-electrophile coupling reaction, and a comparable yield of 56% was observed (Scheme 7E).These findings support that a nickel(0/II) manifold could be catalytically relevant.
To further assess the versatility of our e-XAT strategy, we investigated the reductive arylation of tert-butyl 3-iodoazetidine-1-carboxylate (1u), resulting in the formation of product 55 in 59% yield.Likewise, methyl 4-iodobenzoate (2ba) was identified as a viable substrate for the e-XAT cross electrophile coupling, indicating the enabling potential of our strategy beyond glycosyl bromide electrophiles (Scheme 7F).
Based on the above-mentioned investigations, a plausible mechanism for our electrochemical C-glycosylation was suggested (Scheme 8).Initially, anodic oxidation generates an α-amino alkyl radical of the Hunig base.Subsequently, XAT occurs via homolytic C−Br cleavage, generating glycosyl radical I.The C-alkyl glycoside product is then formed by anomeric radical addition, cathodic reduction, and final protonation.
In contrast, when merging this manifold with nickel catalysis, a plausible catalytic cycle for nickelaelectro-catalyzed reductive glycosyl arylation proceeds via a nickel(0/II/III/I) pathway, with initial oxidative addition of the aryl iodide onto nickel(0) IV to nickel(II) V.In the meantime, the e-XAT generates the persistent glycosyl radical I, which in turn delivers nickel(III) species VI.Finally, reductive elimination from nickel(III) VI provides the desired C-aryl glycosides, while the thus-obtained nickel(I) species VII undergoes cathodic reduction.

CONCLUSIONS
In conclusion, we have reported on an efficient and selective electrochemical C-glycosylation, providing a platform for the assembly of diverse C-alkyl glycosides and structurally complex C-glycoconjugates.The photocatalyst-free e-XAT strategy was not limited to radical conjugate additions.Indeed, paired electrocatalysis by nickelaelectro-catalysis enabled glycosyl Scheme 7. Mechanistic Studies anomeric arylation and acylation.This merger of e-XAT with reductive nickel catalysis reflects the outstanding versatility of our e-XAT approach for C-aryl and C-acyl glycoside synthesis.It is noteworthy, that the judicious choice of the supporting ligand allowed here for full selectivity control at the anomeric center.
■ ASSOCIATED CONTENT