Date: 14 février 2024
Heure: Dès 11h00
Endroit: A-2553 - Campus MIL
Titre: Nature, Theory, and Chemoinformatics: My Contributions To Three Important Threads Within Organic Chemistry
Résumé:
Two significant events in organic chemistry occurred prior to our understanding of chemical structure: the destruction of vitalism with the preparation of urea from inorganic materials by Wohler (1828) and the discovery of benzene from oil gas by Faraday (1825). These began two independent and concurrent threads within organic chemistry that are relevant to this presentation: the synthesis of natural products and materials of theoretical interest. Eventually, computational advances and algorithms by the third quarter of the 20th century enabled organic chemists to develop computer programs based on heuristics and pattern recognition, prompting a new thread: computer assisted organic chemistry. In this presentation I will describe research whose original intent was to elaborate on and continue these three important themes.
First, I will describe our contributions to the field of natural products with the synthesis of hyacinthacine A1, a polyhydroxylated pyrrolizidine that functions as an efficient glycosidase inhibitor.1 The synthesis exploited two key steps: a chemo- and diastereoselective cross-benzoin reaction and a furan photooxygenation-amine cyclization. The latter step was incorporated into a three-step one-pot process to deliver the final polyhydroxylated pyrrolizidine framework.
The second topic will be related to the synthesis of theoretically interesting planar [10]annulenes, which are the next higher neutral Hückel species. Unfortunately, the parent structure is non-planar and thus unsuitable for aromaticity studies. The origin of the non-planarity of the parent framework is due to the large deviation from ideal sp2 bond angles (120°) that are required to generate a D10h symmetric decagon (144°). Fortunately, previous computational explorations suggested that cyclopropanation may relax the bond angle, allowing a planar minimum to be obtained.2 Herein I will describe the successful implementation of this strategy towards the preparation of two planar [10]annulenes.34
The third topic will describe our efforts to ameliorate enantioselective catalysis with chemoinformatics and machine learning. Enantioselective catalysis is highly valuable for both industrial and academically inclined chemists but remains a costly endeavour. Typical approaches rely heavily on intuition derived from minimal experimental data. Unfortunately, our intuition is often fallible and subject to bias. In cases where only negative data (poor %ee’s) are obtained, the problem becomes twofold, and our intuition becomes even less valuable. Herein I will describe a strategy that enables us to determine if a phosphinoxazoline (PHOX) ligand framework is suitable in a transformation with a minimal number of experiments and without resorting to intuition, a process that depends on the initial featurization and clustering.
references
(1) Parmar, K.; Haghshenas, P.; Gravel, M. Total Synthesis of (+)-Hyacinthacine A1 Using a Chemoselective Cross-Benzoin Reaction and a Furan Photooxygenation–Amine Cyclization Strategy. Organic Letters 2021, 23 (4), 1416-1421. DOI: 10.1021/acs.orglett.1c00090.
(2) Sulzbach, H. M.; Schleyer, P. v. R.; Jiao, H.; Xie, Y.; Schaefer, H. F. A [10]Annulene Isomer May Be Aromatic, After All! Journal of the American Chemical Society 1995, 117 (4), 1369-1373. DOI: 10.1021/ja00109a021.
(3) Parmar, K.; Blaquiere, C. S.; Lukan, B. E.; Gengler, S. N.; Gravel, M. Synthesis of a highly aromatic and planar dehydro [10]annulene derivative. Nature Synthesis 2022, 1 (9), 696-700. DOI: 10.1038/s44160-022-00135-z.
(4) Parmar, K.; Blaquiere, C.; Lukan, B.; Gengler, S.; Gravel, M. Is Schleyer’s Predicted All-Cis [10]Annulene an Unstable Framework?ChemRxiv 2022. This content is a preprint and has not been peer-reviewed. DOI: 10.26434/chemrxiv-2022-zhvtf
