I am continuously impressed by the publications that have appeared since Prof. Keith Fagnou’s shocking passing a little over a year ago. The chemical community still mourns; it is clear from these post-mortem publications that Fagnou’s – and his clearly dedicated and talented graduate students and post-docs – brilliance lives on. (Note – this is the same article that appears on Chemical Crystallinity.)
The chemistry that Fagnou has truly spearheaded, direct C-H functionalization, is a method of forming C-C, C-N, C-B, etc bonds without having to prepare one of the coupling partners, as in traditional transition-metal catalyzed cross-coupling reactions. Palladium, rhodium and ruthenium are commonly used catalysts in direct C-H functionalization reactions. Fagnou has published a great deal on arylation reactions of a wide variety of substrates and even a bit on direct benzylation reactions. Some fairly recent reviews are linked in a previous post at my own blog.
A recent publication in Journal of Organic Chemistry (doi: 10.1021/jo102081a), “Predictable and Site-Selective Functionalization of Poly(hetero)arene Compounds by Palladium Catalysis,” published by David Lapointe and coworkers, explores the development of two approaches to selectively functionalizing multi-ring systems – 1) using site-selective reaction conditions, and 2) a pathway with a particular order of reactivity according to a concerted metalation-deprotonation (CMD) mechanism. It is well-known in the field that a great many (hetero)arenes can be functionalized with (painfully) rigorous fine-tuning of the catalyst, ligand, additives, and other reaction conditions. Some substrates have been more difficult to functionalize than others, and selectivity of particular positions on these rings is always an issue – this publication tackles both issues.
To explore site-selective functionalization, the group used compounds with more than one available C-H bond for direct functionalization, and using multiple protocols specific for specific C-H bonds (Larossa’s conditions for C2 arylation of indoles, Gaunt’s Cu-catalyzed C3 arylation of indoles which is actually selective for meta to amido groups, and their own protocols for arylation of perfluorobenzenes and aromatic N-oxides) were able to successfully and selectively functionalize targeted C-H bonds in moderate yields. Here is an example with some decent yields, with reaction times ranging from 16 – 24 hours:
The alternative approach relies upon the CMD pathway as the operative mechanism, which favors electron-deficient substrates. Several years ago, Echavarren published support of this mechanism by finding a preference for the most acidic C-H bond and requirement for a carbonate base, and Fagnou established the use of a pivalate additive, which was speculated to play a crucial role via CMD. A recent mechanistic paper with aromatic N-oxides as the substrates strongly supports this mechanism. The metal first inserts into the aryl-X bond, as expected, and in the key transition state, the pivalate coordinated to the metal deprotonates the C-H bond while the palladium forms a bond to the same C. Reductive elimination (not shown) releases the arylated product.
In the current paper DFT calculations were found to agree quite well compared to competition reaction results of a series of heterocycles to elucidate the order of reactivity of the substrates. Those presented in the paper are as follows, in order of reactivity – this is extremely convenient for the synthetic chemist who would like to utilize this chemistry. And it’s just plain neat – the kind of thing that will hopefully end up in a textbook someday. (Note: the last two substrates are either switched in the text or switched in the image – they don’t agree in the paper and I haven’t looked at the supporting information closely.)
|Reaction conditions: 0.5 eq. of each of two heteroarenes in the competition experiment, 0.125 eq. 4-bromotrifluorobenzene, Pd(OAc)2 5 mol%, PCy3.HBF4 (10 mol%), PivOH (30 mol%), K2CO3 (1.5 eq.), DMA (0.3M), 100ºC.|
And finally, for an example of the method in action – note that the difference between using this method and the previously described is that here, there aren’t necessarily general optimized conditions available for each of the substrate classes here. Examples of a few of these are peppered throughout the arylation literature but they aren’t like indoles, pyridines, N-oxides, perfluorobenzenes, imidazoles, and pyrazoles and don’t have their own special set of conditions (that I’m aware of at the moment). Yields of included substrates range from 65-80%. Instead of optimizing conditions for each, the site of reactivity can be predicted with good specificity – here the indolizine C-H bond over the more electron-rich thiophene’s:
Instead of an aryl bromide, benzyl chloride can be used as the coupling partner as well, with published yields from 55-84%.
- Lapointe, D., Markiewicz, T., Whipp, C. J., Toderian, A., Fagnou, K. (2011). Predictable and Site-Selective Functionalization of Poly(hetero)arene Compounds by Palladium Catalysis Journal of Organic Chemistry : 10.1021/jo102081a