2. Initiation of Metathesis Precatalysts
2.1. How Do Metathesis Precatalysts Initiate?
Typical alkene metathesis precatalysts take the form displayed in
Figure 2.3, consisting of a ruthenium(II) center, a carbene with substituent R, two anionic ligands X (typically chloride), a nondissociating ligand L (typically a trialkylphosphine or NHC), and a dissociating ligand L', which is most often either a phosphine or a chelating alkoxyarene. While the nature of X, L, L', and R all influence the initiation rate and mechanism, it is the nature of L and X that determine the catalytic activity of the active species itself; complexes
G2,
M2, and
GH2 all produce the
same active species, albeit via different mechanisms and at different rates.
A thorough understanding of the initiation mechanism and rate is important, as this determines the rate at which active catalyst is released into the reaction. Faster initiators do not always perform best in metathesis reactions,
24 presumably due to the much lower stability of intermediate species compared to the precatalyst complexes. Polymer chemists often make use of “latent” precatalysts, which are designed to initiate very slowly at room temperature, or only upon specific stimuli such as acid or light of a specific wavelength
25; these allow for the convenient handling of precatalyst/monomer mixtures before the polymerization reaction is begun. Straightforward reactions to form five- and six-membered rings without difficult substitution patterns are often conducted in a straightforward manner with rapid initiators, while more thermally stable slower-initiating precatalysts have been found to perform best for the preparation of challenging tetrasubstituted alkenes (
vide infra).
22 Figure 2.3 Components of a typical ruthenium-based metathesis precatalyst.
Early work had suggested mechanisms for metathesis reactions catalyzed by
G1 in which the alkene might coordinate first, followed by phosphine dissociation, or where both phosphine ligands remain coordinated throughout the catalytic cycle. However, it was later established that intermediates in the catalytic cycle were monophosphine ruthenium complexes, and not bisphosphine species; Lloyd-Jones has previously discussed and summarized some of these early studies.
26One can envision three potential mechanisms for initiation reactions of these species: associative, dissociative, and interchange (
Scheme 2.3); these can even be considered as a spectrum between two extremes (associative and dissociative). In the associative mechanism, the alkene substrate coordinates to the metal center to form an 18e
- intermediate, followed by dissociation of L' and rearrangement to the 16e
- ?2 complex. In the dissociative mechanism, the order of the steps is reversed; L' first dissociates to yield a 14e
- intermediate which then binds alkene. In the interchange mechanism, binding of the substrate and release of the L' group occurs simultaneously, via a single transition state.
Scheme 2.3 Potential mechanisms for metathesis precatalyst initiation: (a) associative, (b) dissociative, and (c) interchange.
Differences in experimental behavior would be expected between the three mechanisms. For the dissociative mechanisms, the initiation rate should be essentially independent of the concentration and nature of the alkene substrate, once the association of the substrate proceeded at a rate higher than that of the dissociation step (i.e., saturation behavior). It has been shown that association of the alkene to a 14e
- ruthenium carbene complex is essentially barrierless.
27 A positive enthalpy of activation would be expected, as the Ru–L' bond is cleaved, along with a positive entropy of activation, the magnitude of which would depend on whether (and how) R and L' are tethered, with larger values for untethered L'. The rate of initiation would be expected to be reduced in the presence of alternative ligands such as excess L'. The associative and interchange mechanisms would proceed at rates dependent on the concentration and nature of the alkene substrate, being favored in the presence of high concentrations of small, electron-rich alkenes, the best ligands requiring the least energetically expensive rearrangement of ligands around the metal center. Both would be expected to have a smaller, yet still positive, enthalpy of activation as the R–L' bond is weakened and the ligands must move closer together to accommodate the incoming substrate. The entropy of activation would be expected to be large and negative, accounting for most of the rotational and translational entropy of the substrate molecule being lost in the associative mechanism, and a significant proportion of it in the interchange mechanism.
The mechanisms in operation for some important classes of alkene metathesis catalyst have been studied; density functional theory (DFT) studies have also provided useful insights into reaction mechanisms. In particular, the initiation mechanisms of “Grubbs-type” and “Hoveyda-type” complexes have been explored. In each case, insights into the effects of structure on initiation rate have been achieved.
2.2. Well-characterized Systems
2.2.1. Complexes Bearing a Dissociating Phosphorus-Based Ligand
The initiation mechanisms of four classes of complex are considered. These classes are “Grubbs-type,” where the alkylidene is a benzylidene and L' is a phosphine; indenylidene, where the alkylidene is 3-phenylindenylidene and L' is a phosphine; “Cazin-type,” where L' is a phosphite; and methylidene complexes (
Figure 2.4). Grubbs-type and indenylidene complexes are known in both bis(phosphine) (such as
G1, where L=L'=PCy3) and NHC/phosphine forms (such as
G2, where L=
SIMes and L'=PCy3), while only heteroleptic NHC/phosphite Cazin-type complexes have been disclosed in the peer-reviewed literature. Notably, methylidene complexes are key intermediates in
all metathesis reactions involving terminal alkenes and so it is important to understand how these might reenter the catalytic cycle.
Figure 2.4 The classes of complex considered in this section.
2.2.1.1. Grubbs-Type Complexes
Grubbs-type benzylidene complexes
15,
16 are one of the most common categories of metathesis precatalyst that are employed in target synthesis, predominantly due to their wide availability and their long-term stability if relatively straightforward precautions are taken.
A watershed moment in the understanding about the initiation mechanism of these complexes was the study conducted by Grubbs and coworkers in 2001.
28,
29 By this point, it was known that second-generation systems were typically far more active in metathesis transformations than their first-generation counterparts.
16 In the initial communication, the authors set out to distinguish between the associative and dissociative mechanisms. The degenerate exchange between free and bound phosphine was explored using 31P magnetization transfer experiments; the resonance for the free phosphine was selectively inverted and the peak heights of free and bound phosphine were measured after different mixing times (
Scheme 2.4). This data then allowed the
rate of degenerate phosphine exchange to be measured, which surprisingly revealed that the more active
G2 complex underwent initiation approximately 100-fold
slower than
G1 (
kobs=0.13±0.01s
-1 versus 9.6±0.2s
-1) at 80 °C in toluene. This rate constant was independent of the concentration of excess PCy3 present in solution, as would be expected for a dissociative process. In addition, the activation parameters determined from variable temperature experiments for
G1 (?H‡=23.6±0.5 kcal mol
-1; ?S‡=12±2 cal K
-1mol
-1) and
G2 (?H‡=27±2 kcal mol
-1; ?S‡=13±6 cal K
-1mol
-1) are indicative of a dissociative mechanism due to the quite large and positive enthalpy and entropy values.
Scheme 2.4 Magnetization transfer experiments to explore phosphine exchange rates in
G1 and
G2.
29 Subsequent experiments probed the reaction of these species with vinyl ether compounds, which irreversibly yield...
