Ring closing metathesis (RCM) has evolved into one of those coveted “predictable” reactions in organic synthesis. Sure, there are things that can go wrong, but for forming unstrained rings, it’s a great way to go.
As with any time using a Grubbs catalyst system, you want to avoid certain functional groups. Coordinating groups (basic amines, sulfides, phosphines, etc.) can inhibit catalyst turnover. They’re especially problematic if situated such that they can form a chelate to a ruthenium alkylidene intermediate (think about the structure of the Hoveyda-Grubbs catalyst). If they can form a five-membered chelate, ureas, amides, and even ethers can cause problems.
Also keep in mind that acyclic compounds will adopt different conformations based on their substitution pattern (and stereochemistry). It’s most often seen in medium-sized rings, but changing protecting groups or oxidation states can sometimes be an easy fix to a troubling cyclization (for a few examples, see ref. 1-3).
Ring size and concentration
It’s a common misconception that RCM reactions need to be run dilute. If it’s your first time through on a small scale and you don’t care how much solvent and catalyst you use, then go ahead, dilute away. But once you start to scale up even a little bit, you’ll start to see the benefits of pushing to higher concentrations.
- Small rings (5 & 6): These reactions can often be run quite concentrated (>0.5 M), and can often be run neat.
- Medium rings (7+): These reactions should be run more dilute (<0.5 M).
- Large rings (14+): These should be run as dilute as is practical. There are a few examples of successful macrocyclizations that are run more concentrated, but generally you should plan on ~10 mM.
Kinetics vs. thermodynamics
When you’re making a large ring that has the potential for E/Z isomers, you can sometimes have different kinetic and thermodynamic products. It’s worth trying to select for the kinetic product by using a 1st Generation Grubbs catalyst (which is slow to react with 1,2-disubstituted olefins), or to select for the thermodynamic product by using a 2nd Generation catalyst (which will react with both substrate and products to give a thermodynamic mixture). It’s tough to predict which is which, but with a little luck, you’ll be able to select for either one.4
Another aspect of this idea comes into play when making relatively strained rings. In situations where there’s a driving force for ring re-opening, you can use this trick to your advantage. In an example that’s near to my heart, chemists in John Wood’s group5 showed that when forming the strained 7-membered ring shown below, a trisubstituted olefin (IIb) can be made in good yield under conditions that give only trace amounts of corresponding disubstituted olefin (IIa).
It seems counterintuitive, since trisubstituted olefins are supposed to be harder to form than disubstituted ones. What’s going on? Re-subjection of the products of these reactions to the reaction conditions explains it. Disubstituted olefin IIa decomposes (presumably via ring-opening), whereas trisubstituted olefin IIb is stable. So it seems the RCM reaction is under kinetic control and the metastable product IIb is formed effectively only because it doesn’t re-enter the catalytic cycle.
1. Creighton, C. J.; Reitz, A. B. Org. Lett. 2001, 3, 893.
2. Shu, C.; Zeng, X.; Hao, M.-H.; Wei, X.; Yee, N. K.; Busacca, C. A.; Han, Z.; Farina, V.; Senanayake, C. H. Org. Lett. 2008, 10, 1303.
3. Bajwa, N.; Jennings, M. P. Tetrahedron Lett. 2008, 49, 390.
4. Look here for a discussion on kinetic and thermodynamic product distributions in macrocyclizations.
5. Haifeng Tang, Ph.D. thesis, Yale 2002, p. 47