Our paper “Complementary site-selectivity in arene functionalization enabled by overcoming the ortho constraint in palladium/norbornene catalysis” published in Nature Chemistry can be read from

https://www.nature.com/articles/s41557-018-0074-z

Poly-substituted arenes are ubiquitous structural motifs found in small molecule-based pharmaceuticals and agrochemicals. One of the routine way to functionalize an arene is through electrophilic aromatic substitution (EAS) followed by cross-coupling reactions. Due to the nature of EAS reaction, the substitution will occur at more electron-rich and more accessible sites. But, what if we want substitutions at the positions disfavored for EAS reaction? Despite the breakthroughs in directing group strategies and steric-sensitive C−H borylation/silylation methods, solutions to such a quest for “complementary site-selectivity” remain highly sought after.

In the Dong lab, we are interested in developing the palladium/norbornene chemistry, which is known as the Catellani-type reaction. In such a reaction, an electrophile (an oxidant) reacts at the ortho position, whereas a nucleophile is coupled with the ipso carbon of the aryl halides. But, one limitation of this chemistry that really bothers us is the “ortho constraint”, that is, there must be an ortho substituent in the aryl halide. There are several challenges to overcome such an “ortho constraint”, for example the second unwanted C-H metalation is too facile. More discouraging for us is the early finding by Catellani that β-carbon elimination become feasible only when both ortho positions are substituted. Nevertheless, we are stimulated by the opportunity not only to greatly expand the scope for Catellani-type reactions, but also to provide a unique tool to achieve complementary arene functionalization.

Our original idea to solve this problem is simple: when we eliminate the ortho substituent from the substrate, we can put substituent in the catalyst! After carefully examining the transition states, we decided to put substituent on the bridgehead position of the norbornene, which could potentially solve both two challenges. For me personally, it is a quite exciting project because we are rationally modifying and designing a catalyst in order to realize a specific goal.

After a long period of negative results, our first hit came when I tried to screen different conditions for the methyl ester substituted norbornene. I got very excited for a while, but quickly calmed down because it was only 2% yield; and even worse, the unwanted di-product is in a significant amount. Nevertheless, such result encouraged me to screen more extensively on different norbornenes. Luckily, the norbornene we found in the first hit is almost the worst choice: when we changed to alkyl substituted norbornene, the reaction worked much better.

When the optimization of the reaction is done, I was very happy to get a special opportunity to learn DFT calculation at University of Pittsburg. Prof. Liu gave me the warmest welcome in that cold winter and I was quite excited to reproduce the experimentally observed selectivity in silico.

After that, we continued to explore its scope. We demonstrated that the complementary selectivity could be achieved via a two-step sequence as proposed. One interesting aspect of this chemistry is the late-stage functionalization of natural products.  Taking Strychnine as an example, we can put functional group at the site disfavored for EAS reaction in two steps, while the prior approach took 7 steps. We believe that this chemistry could be a new starting point for Catellani-type reaction and we are trying to extend such proof-of-concept to more practical chemistry so that it could become a useful tool for synthetic and medicinal chemist.