Ring-closing metathesis has become an essential tool for C-C bond formation as demonstrated by the profound impact on total synthesis in recent years.1 The first examples date back to 1980 and involved the use of tungsten-based catalysts for the preparation of macrocycles from dialkenyl keto-esters and dialkenyl ketones (Villemin and Tsuji, respectively).2 In 1993, Grubbs and coworkers reported the first example of a carbocyclization using a functional-group tolerant molybdenum-based carbene catalyst (silyl ethers, esters, alcohols, and benzyl ethers were all tolerated).3 The Grubbs group subsequently synthesized a new air-stable ruthenium-carbene complex which was able to catalyze the cyclization of a variety of dienes in good yields (Scheme 1).4 Numerous research groups have since used RCM to synthesize highly complex molecules. What follows is a brief overview highlighting a few representative examples of RCM.
In 2004, Wood and coworkers reported the synthesis of ingenol.5 This natural product has fascinated synthetic chemists for the past 20 years due to its promising biological activity and structure, which features four distinctive ring systems. Wood and coworkers devised an ingenious approach involving a ring-opening/cross metathesis and a ring-closing metathesis step (Scheme 2). With only 2 mol % of the Grubbs catalyst (1st Generation), the norbornene derivative afforded the desired dienes in nearly quantitative yield. It is important to note that the polymerization of norbornene was suppressed by using high dilution conditions and excess ethylene gas. Further functionalization of the diene set the stage for the ring-closing metathesis step. The desired product was obtained in 76% yield in the presence of 25 mol % of the Grubbs-Hoveyda catalyst (2nd Generation).
Traditional Grubbs catalyst systems are five-coordinate ruthenium complexes containing a neutral ligand, which is typically a phosphine, or in the case of the Hoveyda-Grubbs catalysts, a styrenyl ether. Ligand dissociation is required to provide the active catalyst, but is slow at low temperatures. Therefore, traditional Grubbs catalysts are not reactive at low temperatures. The Piers group developed preformed 4-coordinate cationic complexes that do not require ligand dissociation prior to reaction. The reaction progress was examined for the cyclization of diethyldiallylmalonate to provide the corresponding cyclopentene derivative using both the Piers catalyst and the Grubbs catalyst at 0 °C (Scheme 3).6 At this temperature, the Grubbs catalysts (2nd Generation) was found to be a weak initiator and after 4 h, the reaction had progressed to 25% completion, while the cationic complex had progressed to > 90% completion after 2 h. The initiation rate of the cationic catalyst at 0 °C was found to be comparable to the initiation rate of the Grubbs catalyst (2nd Generation) at 35 °C.
In a subsequent study, Piers and coworkers examined the intermediates in an olefin metathesis reaction by NMR at –50 °C (Scheme 4). This was the first direct observation of a ruthenacyclobutane intermediate and provided evidence for a symmetrical intermediate.7 It also illustrated the stabilizing effect of the N-heterocyclic carbene ligand on the Ru(IV) species.
More recently, Stoltz and Grubbs jointly reported the use of ringclosing metathesis in the key step of the total synthesis of (+)–elatol, a spiro[5.5]undecane-containing natural product which has drawn significant interest due to its antibiofueling, antibacterial, and antifungal activity as well as its cytotoxicity. The researchers developed an asymmetric synthesis of this molecule utilizing RCM (with Hoveyda- Grubbs 2nd Generation catalyst) in order to provide a key intermediate en route to the natural product. (Scheme 5).8
In their preparation of sominone (R = OH, 22ß-O) and related analogs, Matsuya and coworkers used RCM to generate tetrasubstituted olefins. The RCM reaction was utilized in the preparation of the lactone side chain (Scheme 6). While ring closure affording the tetrasubstituted olefin with the Hoveyda-Grubbs catalyst (2nd Generation) proceeded in low yields (15–24%), it allowed a rapid and modular entry into the desired motif.9
Donohoe and co-workers have developed several synthetic approaches to the synthesis of furans which incorporate RCM as a key step.10 In 2007, the researchers synthesized a series of di- and tri-substituted furans using both the Hoveyda-Grubbs catalyst (2nd Generation) and the Grubbs catalyst (2nd Generation), followed by addition of PPTS or TFA, respectively (Scheme 7).
In 2008, Donohoe and coworkers devised a synthesis of (–)‑ (Z)-deoxypukalide, a member in a class of marine natural products which have exhibited a range of biological activities, including neurotoxicity and anti-inflammatory effects. In the total synthesis of (–)-( Z)-deoxypukalide, mixed acetal formation with acrolein diethyl acetal served to install the requisite allyl group and following RCM, acidic hydrolysis furnished the desired furan in good overall yield. The final step of the synthesis utilized a second RCM with the Grubbs catalyst (2nd Generation) to form the butenolide unit in 72% yield (Scheme 8).1
Alkyne metathesis has been a useful tool for C-C bond formation since the discovery of structurally well-defined metal alkylidynes by Schrock and coworkers.12 These complexes have found use in the synthesis of complex natural products and in material science.13 The limitations of these catalysts include air- and moisture-sensitivity as well as incompatibilities with substrates that contain donor sites. Fürstner and co-workers have recently developed an air-stable molybdenum catalyst for alkyne metathesis, which proved versatile and compatible with substrates containing donor sites.14 The authors also examined the use of this catalyst in the ring-closing cross metathesis (RCAM) of a variety of alkynes in generating various macrocycles in good yields (Scheme 9).