New Insights into Decomposition of Metathesis-Active Methylidenes

Title: New Insights into Decomposition of Metathesis-Active Methylidenes
Authors: Rufh, Stephanie
Date: 2017
Abstract: Olefin metathesis is one of the most powerful tools in current use for the assembly of carbon-carbon bonds. The advent of easily-handled ruthenium catalysts, along with new awareness of ring-closing metathesis (RCM), led to widespread adoption of metathesis methodologies in synthetic organic chemistry. Relative robustness to air and water, and tolerance for an impressive range of functional groups, is attested to by the recent appearance of RCM in pharmaceutical manufacturing processes. However, challenges remain, particularly with respect to the ease with which these catalysts undergo loss of the critical Ru=CHR functionality that enables metathesis. This limits not only metathesis productivity, but also selectivity, because the Ru products can promote unwanted C=C migration reactions. Many decomposition pathways have been found to culminate in activation of the N-heterocyclic carbene (NHC; most typically an H2IMes) ligand. In consequence, NHC truncation is a widely accepted design solution. This thesis work was aimed at assessing the validity of this proposition from several perspectives. Hundreds of metathesis catalysts are now known, but the second-generation Grubbs catalyst (GII) is still most widely used. This work began with examining the role of Lewis bases in promoting decomposition of the second-generation Grubbs catalyst, RuCl2(H2IMes)(PCy3)(=CHPh) GII, under conditions of metathesis. Small Lewis donors, ranging from pyridine and DMSO to THF and water, were shown to greatly accelerate loss of the methylidene ([Ru]=CH2) ligand. Associative donor binding was shown to promote dissociation of PCy3, a powerful nucleophile which then attacks at the methylidene carbon. Attempts to probe the reversibility of this step revealed no reformation of the [Ru]=CH2 unit, suggesting that nucleophilic attack to form a [Ru]–CH2PCy3+ moiety is the key decomposition event. Ensuing C-H activation of a ligand (e.g. the o-methyl of H2Mes or IMes), with abstraction of a chloride ligand, results in elimination of [MePCy3]Cl. Two potential preventative measures were examined. These involved, first, the use of phosphine scavengers to intercept the liberated PCy3 ligand during metathesis, and second, truncation of an N-mesityl group to N-methyl to inhibit C-H activation. While phosphine scavenging proved somewhat successful in increasing metathesis yields, lower activity was observed for the catalyst with the modified NHC ligand. The negative impact of a smaller NHC on catalyst activity is of particular interest, given that curtailed NHC bulk has been proposed as the key to improved catalyst robustness, as noted above. ix To clarify this issue, truncation of the nitrogen substituents was taken to the extreme by replacing both N-mesityl groups with N-methyl. Accordingly, the reaction chemistry of the first-generation Grubbs catalyst, RuCl2(PCy3)2(=CHPh) GI with 1,3,4,5-tetramethylimidazol-2-ylidene (IMe4) was explored. The small size of this NHC ligand precludes access to a stable mono-IMe4 catalyst corresponding to GII: rather, an initial cis-RuCl2(IMe4)2(=CHPh) species is thought to rapidly convert to a highly stable face-bridged dimer, the observed product. The latter complex exhibits minimal catalytic activity, even in the trivial RCM of diethyl diallylmalonate (DDM). Synthesis of the corresponding edge-bridged dimer was accomplished, but activity remained poor. To simulate an active catalyst, IMe4 was added to GIm. Complete loss of the methylidene ligand was observed within minutes at RT. Importantly, however, decomposition was traced to rapid bimolecular coupling, with loss of the methylidene ligand as ethylene, rather than nucleophilic abstraction. Coupling is inhibited for the benzylidene system by the increased steric bulk of [Ru]=CHPh relative to [Ru]=CH2, a feature well established in more conventional catalyst systems. We conclude that moving from H2IMes to a smaller NHC to inhibit C-H activation is not a practical solution for improving catalyst robustness. Rather, catalyst redesign should focus on protecting the [Ru]=CH2 moiety to prevent loss of the methylidene ligand via either nucleophilic abstraction or bimolecular coupling.
CollectionThèses - Embargo // Theses - Embargo