60th Annual Report on Research 2015

Research during the first year of funding has advanced many of the goals delineated in our grant application. First, we completed our initial study if intramolecular C–H insertion that largely focused on the synthesis of benzodihydrofurans, with initial results on related ring systems. This work was published at the end of 2014 in J. Am. Chem. Soc. Second, we have invested significant effort exploring intermolecular X–H insertion reactions of alcohols, amines, and thiols. Third, we have begun exploring catalyst diversity for the synthesis of alkyl-substituted benzodihydrofurans and indanes, which were our toughest substrates in the initial study. Finally, we have designed a new series of substrates that explore the synthesis of complex, polycyclic products that will open up new vistas for asymmetric catalysis.

Our first paper explored the asymmetric synthesis of benzodihydrofurans using a one-pot process that generates diazo compounds and their respective carbenoids in situ. This process proved to be highly effective for a high number of substrates and high yields and/or stereoselectivities were observed for a total of twenty substrates, including one example in which the rate of insertion had to exceed that of elimination. In addition, we showed that the reaction worked for the synthesis of an indane and an indoline, i.e. the carbon and nitrogen analogs of the parent system. Finally, the method was applied to the first asymmetric synthesis of an oligoreseveratrol natural product, namely E-delta-viniferin. Although this work has been published for less than one year, it has already been highly cited.

Based on the success of the intramolecular C–H insertion, we chose to explore the intermolecular reactions of diarylcarbenoids toward X–H bonds wherein X is either nitrogen, oxygen or sulfur. We screened a total of 75 catalytic conditions with 7 different substrates and 17 catalysts without ever observing significant formation of the desired product. In many cases, the principle component of the reaction mixture was the azine, i.e. the product of the nitrogen of the diazo compound attacking the carbon center of the metal carbenoid to produce a symmetric product. Collectively, these results demonstrated that the X–H insertion substrates were either too bulky or otherwise not nucleophilic enough to approach the carbenoid. We did observe that in the absence of catalysts, our diazo-generating reaction formed the basis of a mild esterification reaction that was applicable to variety of carboxylic acid substrates. In collaboration with the Hein group (UC Merced), we were able to use infrared reaction monitoring in situ (react-IR) to demonstrate that there is no buildup of the diazo intermediate. These results have been prepared for publication and will be submitted shortly.

The C–H bonds of alkyl chains are generally regarded as the most challenging insertion substrates. Although our initial benzodihydrofuran-forming condition worked in reasonably high yield and selectivity, there was significant room for improvement when compared to the benzylic or allylic counterparts. In addition, we wanted to more generally explore indane formation. Our lead result with the Rh₂(PTAD)₄ catalyst proceeded in good yield, but with modest diastereomer ratio (dr, 86:14) and low enantiomer ration (er, 70:30). We have elected to explore unsymmetrical catalysts developed by Prof. Joseph Fox (University of Delaware). He has graciously provided 28 different catalysts in which one chiral ligand has been displaced with an achiral carboxylic acid. These catalysts showed better selectivity in some of the Rh-catalyzed reactions developed in his group. We are currently screening each catalysts for two reactions, i.e. benzodihydrofuran and indane formation, in two different solvents using GCMS and chiral HPLC to get conversion, dr, and er for each reaction run. So far none of the catalysts is superior to our lead, but we are not yet done with the first half of the screen.

In our most recent efforts, we are exploring substrates that might normally be prone to elimination in order to access complex, polycyclic products. Specifically, we are making silyl-hydrazones, which can be de-protected in situ with CsF and we have documented that the fluoride does not interfere with the rhodium catalyst. In addition, the rhodium might be able to serve the dual catalytic role of enabling the hydrazone synthesis, given that a Lewis acid is normally required. The net process will involve a one-pot transformation that proceeds from ketone directly to C–H insertion product.