Carbon–Carbon Bond Formation at a Neutral Terminal Carbido Ligand: Generation of Cyclopropenylidene and Vinylidene Complexes

Abstract

Olefin metathesis is an important tool for organic and polymer synthesis. However, some key functional groups are not tolerated even by Ru-based catalysts. We recently showed that vinyl esters can deactivate [Ru(CHPh)(PCy3)2Cl2] (1) [3] by quantitative formation of [Ru(C)(PCy3)2Cl2] (2) [4,5] A rare neutral terminal carbido complex, 2 is surprisingly stable and has few reported reactions. However, protonation of 2 by strong acid yields catalysts that rapidly initiate olefin metathesis. Thus, 2 is both a precursor to and a decomposition product of olefin metathesis catalysts. We see 2 as a potential source of a C1 fragment. Accordingly, we describe herein the first C C bond-forming reaction of this unusual compound. The terminal carbido ligand in 2 is a poor nucleophile, as shown by its failure to react with MeI, MeCOCl, and PhCH2Br. Although 2 does not react with a variety of alkenes and alkynes (see the Supporting Information), it reacts cleanly with MeO2CC CCO2Me (dimethyl acetylenedicarboxylate, DMAD) over 4 h in C6H6. A new blue-purple complex, 3, is formed as the carbido signal for 2 (C NMR: d= 471.8 ppm) is replaced by a new signal at d= 195.7 ppm. The H NMR spectrum evinces formation of a 1:1 adduct of 2 with DMAD. Formation of the cyclopropenylidene complex [Ru{=CC2(CO2Me)2}(PCy3)2Cl2] (Scheme 1) accounts for these observations. Several cyclopropenylidene complexes exist. Unlike 3, however, the cyclopropenylidene units in these complexes are substituted by phenyl or electrondonating groups. [Ru(C)(H2IMes)(PCy3)Cl2] (4 ; H2IMes= 4,5-dihydro-1,3-bis(mesityl)imidazol-2-ylidene) reacts similarly with DMAD, but the reaction is not clean since the product reacts further with DMAD before all of 4 has been consumed. However, 4 reacts more cleanly with HC CCO2Me (see the Supporting Information). Single-crystal X-ray diffraction confirmed the structure of 3. Figure 1 depicts a thermal ellipsoid plot of one of the two chemically equivalent but crystallographically independent molecules of 3 in the crystal. The data establish the expected connectivity in 3, but the large uncertainty associated with the Ru=C bond length of 1.846(10) A precludes comparison with those in related alkylidene complexes. The cyclopropenylidene ring lies in the Cl-Ru-Cl plane. The structure shows significant bond localization in the cyclopropenylidene fragment. These distances closely resemble those observed in free Scheme 1. Formation of 3 and ring-opening reactions. HBpin=pinacolborane, Ar=3,5-Me2C6H3.