With the recognition of a major impact of olefin metathesis in modern organic synthesis,¹ recent years have witnessed a surge of productive research efforts in Ru catalyst design,^2-10 driven by the fascinating challenge to improve the already-impressive performances of Grubbs catalyst, (PCy₃)₂Cl₂Ru=CHPh,² and its upgraded second 1a² and third 1b³ generations. Parallel significant progress was made with the advent of Grubbs-Hoveyda catalyst 1c based on the "release/return" concept.⁴ Enhanced versions of the latter^5-8 exhibiting a broader application profile^5,11 were obtained by modifying the chelating arm upon introduction of a bulky phenyl group (Blechert's catalyst 1d)⁵ or an electron-withdrawing substituent (our own "nitro" catalyst 1e)⁶ in such a way to destabilize the oxygen/metal interaction, thereby favoring a faster access to the key "propagating" 14 electron species.¹²

Also quite spectacular was the report by Piers⁹ of an elegant direct access to a 14 electron catalyst based on a phosphonium alkylidene, obtained without liberation of any auxiliary ligand susceptible to be recaptured by the active species.

As recently mentioned by Fogg,¹⁰ however, higher activity may come at the price of catalyst lifetime. So, in the context of green chemistry, there is still an obvious need to find the best compromise between desirable (but antinomic) properties such as the stability of the precatalyst, a high initiation rate, a high selectivity for challenging substrates, and the possibility of catalyst recycling.

In the present attempt to enhance the leaving group properties of the styrenyl ether, we were led to functionalize its aliphatic end group by attachment of an ester function, originally selected in account of its electron withdrawing properties.

The required ligand 3f was prepared according to Scheme 1 and reacted with 1a in the presence of CuCl,^5,13 to produce the air stable green complex 1f. The molecular structure of 1f, established by X-ray diffraction (Figure 1), revealed the occurrence of an additional coordination of the carbonyl oxygen of the ester function to the metal (2.54 Å).

	Figure 1 Molecular structure of the catalyst 1f (see ESI for details).
	Scheme 1. Two-Step Preparation of Catalysts 1f and 1g

The new complex 1f was subject to a series of comparative catalytic ring-closing metathesis (RCM) reactions (at 0 C) and cross-metathesis (CM) with methyl acrylate (at 25-26 C) with a set of benchmark catalysts 1c-e (Table 1). Interestingly, in the more challenging CM reaction, the 1f catalyst's performance was matching that recorded with Blechert's catalyst (Table 1, entry 2). The ester catalyst 1f appeared to be particularly insensitive to impurities, as established by a selected CM reaction deliberately carried out in an opened tube and in commercial grade dichloromethane (entry 2). Significantly, 1f catalyses the cross metathesis of a model olefin, 4c,^6a,b with acrylonitrile (95% yield after 5 h at 25 C with 3 mol % of catalyst loading), or methacrylonitrile (56% yield after 16 h at 40 C with 5 mol % of catalyst loading).¹⁴

In a logical extension of this approach, we also prepared the closely related complex 1g, incorporating both the electron-withdrawing NO₂ group^6,13 and the ester. An interesting point is that the combined independent effects of the coordinating ester and the electron-withdrawing groups are additive. This is particularly obvious in the RCM of 4a at 0 C, where 99% is achieved within 1 h with only 1 mol % of catalyst 1g (Table 1, entry 1), and in the RCM of diethyldiallylmalonate 4b (Figure 2).

Figure 2 Relative conversion rates for RCM of 4b, by 1a (Grubbs), 1c (Hoveyda), 1d (Blechert), 1f, and 1g (this work) using 1 mol % catalyst in CH2Cl2 at 0 C. See ref 15 and ESI for more details.

Particularly appealing are the high turnover numbers recorded with 1g in the RCM of dienes 4f (TON = 1367), 4g (TON = 3200), and in enyne 4h cycloisomerization (TON = 560). Selectivity, a further advantage of the catalyst, can be appraised in the synthesis of 5i, a pheromone of the Leopard moth (Zeuzera pyrina), obtained in good yield without homo-dimerization product, and requiring only a very low catalyst loading (0.3 mol % of 1g), just ten times less than required with already very active⁶ 1e to obtain the same result. This highlights the fact that apparently "minor" changes in the structure of Hoveyda-Grubbs type catalysts can dramatically improve their performances and also broaden the scope of their applications, which can be taken as a rewarding justification of relevant research efforts in this direction.

Acknowledgment

Research supported by the Polish Academy of Sciences (President of Polish Academy of Sciences Fellowship to M.B.), by the CNRS, and through a European Community Marie Curie Action (contract no. HPMT-CT-2001-00398). A gift of chemicals from Chemetall GmbH (Frankfurt am Main, Germany) is gratefully acknowledged.