Computational and Experimental Studies in Selective Organocatalysis

Abstract

Chemical synthesis has transformed the ability of scientists and engineers to interact with the molecular world. Yet despite almost two centuries of considerable effort, small-molecule synthesis remains a challenging task. Hundreds of new reactions are discovered every year, but few possess the requisite selectivity and generality needed to be useful for routine synthesis, and elucidation of their mechanism and underlying catalytic principles is rarely conducted. In this work, we describe a variety of efforts at the interface of organic, computational, and analytical chemistry which seek to address the linked problems of discovering selective organocatalysts and understanding the mechanism by which they operate. In Chapter 1, we report the development of a new analytical method that combines chiral stationary phase supercritical fluid chromatography with mass spectrometry-based detection to enable enantiodetermination of pooled crude reaction mixtures, greatly increasing analytical throughput. This advance allows us to perform multi-substrate screening to discover catalysts possessing good substrate scope, which we demonstrate in the optimization of a Brønsted acid catalyst for the enantioselective Pictet–Spengler reaction. In Chapter 2, we disclose the results of a mechanistic study aimed at understanding a hydrogen chloride/hydrogen-bond donor co-catalyzed Prins cyclization of alkenyl aldehydes which exhibited dramatic rate acceleration compared to the background reaction. Our studies reveal that the catalyst reacts with hydrogen chloride to form a new chiral acid in situ with a higher pKa than hydrogen chloride, which nevertheless reacts faster owing to favorable catalyst-controlled positioning of the chloride anion to electrostatically stabilize the major transition state. In Chapter 3, we report a computational study of our group’s regio- and stereoselective glycosylation of minimally protected glycosyl acceptors. The computational model described— the first of hydrogen-bond-donor-catalyzed glycosylation of glycosyl phosphate donors—contains features of the transition state previously hypothesized on the basis of experimental results, and lends support to the proposed “4H” binding mechanism. In Chapter 4, we describe the development of an enantioselective protio-semipinacol reaction of unactivated vinylic cyclopropanols. Motivated by the question of how high enantioselectivity can be achieved in a low-barrier 1,2-rearrangement, we conduct an experimental and computational mechanistic investigation and come to the surprising conclusion that protonation to form a formally achiral carbocation in fact exerts stereocontrol over the subsequent rearrangement step: the rearrangement is so rapid that the carbocation is locked in a given chiral conformation, rendering the rearrangement effectively stereospecific. Finally, in chapter 5 we detail a spectroscopic and computational study of solutions of hydrogen chloride in diethyl ether, aimed at assigning the solution structure of hydrogen chloride. In situ IR spectroscopy, combined with density-functional theory and molecular dynamics, provides evidence for the existence of oxonium ions formed from complete proton transfer to diethyl ether. This observation explains the often-inhibitory effect of diethyl ether on hydrogen chloride-catalyzed reactions and has intriguing implications for catalyst design.