Ruthenium catalysts bearing cyclic (alkyl)(amino)carbene
(CAAC)
ligands can attain very high productivities in olefin metathesis,
owing to their resistance to unimolecular decomposition. Because the
propagating methylidene species RuCl2(CAAC)(CH2) is extremely susceptible to bimolecular decomposition, however,
turnover numbers in the metathesis of terminal olefins are highly
sensitive to catalyst concentration, and hence loadings. Understanding
how, why, and how rapidly the CAAC complexes partition between the
precatalyst and the active species is thus critical. Examined in a
dual experimental–computational study are the rates and basis
of initiation for phosphine-free catalysts containing the leading
CAAC ligand C1
Ph
, in which a
CMePh group α to the carbene carbon helps retard degradation.
The Hoveyda-class complex HC1
Ph
(RuCl2(L)(CHAr), where L = C1
Ph
, Ar = C6H3-2-O
i
Pr-5-R; R = H) is compared with its nitro-Grela analogue
(nG-C1
Ph
; R = NO2)
and the classic Hoveyda catalyst HII (L = H2IMes; R = H). t-Butyl vinyl ether (tBuVE) was employed as substrate, to probe the reactivity of these
catalysts toward olefins of realistic bulk. Initiation is ca. 100×
slower for HC1
Ph
than HII in C6D6, or 44× slower in CDCl3. The rate-limiting step for the CAAC catalyst is cycloaddition;
for HII, it is tBuVE binding. Initiation
is 10–13× faster for nG-C1
Ph
than HC1
Ph
in either
solvent. DFT analysis reveals that this rate acceleration originates
in an overlooked role of the nitro group. Rather than weakening the
Ru–ether bond, as widely presumed, the NO2 group
accelerates the ensuing, rate-limiting cycloaddition step. Faster
reaction is caused by long-range mesomeric effects that modulate key
bond orders and Ru-ligand distances, and thereby reduce the trans
effect between the carbene and the trans-bound alkene in the transition
state for cycloaddition. Mesomeric acceleration may plausibly be introduced
via any of the ligands present, and hence offers a powerful, tunable
control element for catalyst design.