For my inaugural post on Colorblind Chemist, I’d like to talk about allylic fluorination.
Fluorine has captured the interest of medicinal chemists for a while now; the introduction of the wily halogen can modulate all sorts of pharmacokinetic properties, most importantly being the lipophilicity, which generally tracks fairly well with oral bioavailability. Even better, the newest hotness in the field of fluorine chemistry for the past decade or so has been Positron Emission Tomography (PET), which uses radioactive fluorine-18 atoms to track molecules of interest in biological systems with good signal-to-noise. These two desires have shot fluorination to the forefront of a number of methodology groups’ to-do list, including that of my former advisor, Tobias Ritter. In fact, just recently the Ritter group put out a beautiful paper on the facile fluorination of aromatic substrates, which lends itself directly into PET chemistry (there are a number of constraints on what makes a reaction PET-compatible, including fluorine source, reaction time, and byproduct profile, which are discussed in the linked paper for those who are interested).
So then, where does the state of the art lie? If we ignore PET-compatible constraints, there are actually a number of catalytic methods to generate selective fluorinations, where substrates range from aromatic to aliphatic with varying degrees of success. Today I’d like to focus on allylic fluorination, specifically because this strange reaction system was published in Organometallics recently, and I honestly can’t tell if I love it or hate it.
So what’s the big deal here? Well, a priori, we have a palladium-based system that stoichiometrically generates allylic fluorides from allylic bromides. Or does it perform C-F Activation/Arylation? Technically it does both.
So let’s take a step back and think about what we’re working with. This reaction would have been exciting as hell 17-20 years ago when the only reliable manner for generating allylic fluorides was to use molten caesium fluoride as the fluoride source, but in recent years we’ve had all sorts of break-throughs from the Gouverneur labs at Oxford to create linear allylic fluorides from a-b unsaturated esters, and several iterations of a system from the Doyle labs that enantioselectively generates branched allylic fluorides from allylic chlorides. We don’t even need to consider Nguyen’s trichloroacetamide system to realize that the bar for allylic fluorination is fairly high.
So what about C-F activation? Well, it turns out that is pretty well precedented as well; in fact, it’s the entire focus of the Hughes group at Dartmouth. So really, what has this paper achieved? This is where I draw the line about whether I love or hate this paper; the practical answer is nothing. The products of this method are precedented, even through palladium catalysis, and so in terms of adding access to new molecules, there really isn’t anything novel. Sure, the method can alkylate a perfluoropyridine AND simultaneously fluorinate an allylic bromide, but honestly, if the day comes where I just happen to need those two things at the same general time, I’m not going to pull the obscure phosphine ligand needed for this off the shelf (or rather, synthesize it, as it isn’t commercially available) to do this.
So, what in the hell am I getting at? To be succinct, this paper does exactly one thing well: it accomplishes fluorination with a very non-traditional fluorine source. Who cares? Well, if we think about the current fluorine sources, which are admittedly rather abundant, they tend to have their subtle drawbacks. For instance, use CsF that’s been exposed to air too long, and it’ll hydrate to the point of being practically useless. Expose AgF to air or light for too long, and it will decompose, and don’t even get me started on the cationic fluorine sources. Further, nearly all of these have some issues with organic solubility, which may not be an issue in PET scenarios, but for med chem, it can be a serious issue if your fluorine source doesn’t agree with your choice of solvent. In this paper, we use a card-carrying organic fluorine source, which is by all approximations air stable, and pluck a fluorine off and put it on our substrate. If you’re really enthusiastic, you could even recycle the spent fluoropyridine with the Ritter method and call yourself square.
Is perfluoropyridine the ideal “organic” fluoride source? Of course not. However, for a relatively simple molecule that’s reasonably priced, it’s certainly not horrible, but the big thing is that it’s something different. I believe that fluorination reactions will make up a very important tool kit of reactions for both medicinal chemists and diagnosticians alike, and thinking up alternatives to CsF and DAST are likely what will enable the next generation of catalysts. So, hats off to Braun et al, it’s probably not what you were going for, but you gave me some food for thought.