As someone who used a lot of carbohydrates as chiral pool starting materials in grad school, I regard this paper as the next thing to witchcraft. Even folks without carbohydrate experience appreciate readily that there are sugars that you hear about all the time (such as glucose, mannose, and galactose) and some that you hardly ever hear about at all (such as allose, gulose, and talose). That’s not even getting into the deoxys and the aminos and all the rest of the derivatives; those guys are straight-up members of the sugar club, but they’re just rare in nature (and correspondingly expensive in commerce).
Emil Fischer famously synthesized his way through the carbohydrates in the early days of the field and confirmed the stereochemical relationships between them, a tour de force that still ranks as one of the truly impressive feats of organic chemistry. So it’s not like you can’t synthesize all sorts of rare carbohydrates, but it can be quite painful and expensive to do so. Carbohydrate chemistry has long been The Land of Protecting Groups, and for good reason, since there’s a hydroxyl group coming off of nearly every carbon that can host one. Now it’s true that each of those hydroxyls on each different sugar backbone has its own personality (hostile, in many cases), and there are whole lists of odd little reactions that will open up or protect one particular one (or two adjacent ones). Carbohydrate chemistry is like steroid chemistry in that regard – these crowded ring systems have all sorts of reactions that you wouldn’t think you could do but you can, and a similar list of things that look perfectly plausible on paper but just don’t work.
So the paper mentioned above, from the Wendlandt group at MIT, is particularly interesting because it’s working on completely unprotected sugar systems. It features a hydrogen abstraction/donation mechanism (through photoredox catalysis) that basically just reaches in and epimerizes free hydroxy groups. There have been several other papers in recent years that target reactions such as C-H alkylation on such systems, with the shared goal of getting out of the aforementioned Protecting Group Land. This paper shares some similarities to those earlier ones – for example, seeing reactions at the C-3 hydroxy more often than not.
The authors freely admit that reasons for the selectivity of the reaction are not clear. From what I can see, the alpha-methyl glycosides tend to epimerize the C3 OH, and the beta-methyl ones tend to react at C2. Free sugars (OH at the anomeric position) react as well, albeit in lower yields. 2-deoxy sugars can be epimerized at their 3 position, which is something that basically you just haven’t been able to do. Even disaccharides like sucrose or raffinose have single OH groups isomerized, and the reaction can also be done on nucleosides and C-glycosides as well. Yields vary, but as I said earlier, I’m just amazed that this can be done at all. Figuring out the reasons for the selectivities, which clearly seem to be under kinetic control and not thermodynamic, and seeing if other OH groups can be dragged into participating will be something to watch for. Rare sugars might end up not being quite as rare. . .