I’m not a polymer scientist, but since olefin metathesis pops up so much in polymers, I’ve picked up a few things here and there. Most polymer applications of olefin metathesis are polymerizations or functional modifications, but the recent paper by Guan and Leibler has stretched the boundaries of what’s possible – by metathesis or any other chemistry.
Let’s start with the basics. Thermoplastic polymers are comprised of linear chains of atoms. At low temperature, the motion of the segments along the chain is limited resulting in a glassy or semi-crystalline solid. But when heated above the glass transition or melt temperature, the chains have enough energy to overcome the intermolecular interactions and the polymer flows like a liquid. This makes thermoplastics easy to melt-process into all sorts of shapes, and also to recycle – PET water bottles can be ground, melted, and re-molded. Thermoplastics will dissolve in organic solvents also (Remember the first time you tried to rinse that plastic weighing dish with dichloromethane?). But if you crosslink the linear chains of a polymer, the polymer chains are no longer free to move past each other and you’ll end up making a very different material. Cross-linked polymers tend to be better suited for cyclic loading (like in a tire or a wind blade) than thermoplastics, but they can’t be re-used by melting or dissolving. Your vulcanized tire is incredibly durable but you have to break carbon-carbon bonds before it loses its tire shape – good for its application, but problematic for recycling. If you designed a polymer with reactive cross-links that could exchange with one another, you might end up with a material that has some of the properties of both cross-linked and uncross-linked systems. Guan and Leibler did just that. They’re not the first people to have this idea, but there are few chemistries that can pull it off if you want to maintain the number of cross-links but rearrange them so that the properties are maintained as the material is molded.
The group lightly cross-linked a cis-1,4-polybutadiene by using radical chemistry, and after purification, they swelled the sample with dichloromethane and loaded it with 0.005 – 0.01 mol% of the 2nd Generation Grubbs Catalyst (if I did my math right that’s 0.08 – 0.16 wt%). Some of the chains were broken during this process as the catalyst exchanged its benzylidene group for a polymer backbone olefin, but when the solvent was removed, what was left was a rubber loaded with catalyst. I won’t go into the all the details on mechanical tests or control experiments they performed (and explained well enough so that even I could understand them.) However, Guan and Leibler convincingly supported their assertion that the Grubbs-loaded rubber was undergoing dynamic crosslinking under strain – changing form but remaining a cross-linked rubber.
The authors left me with the mental image of a rubber band that stays stretched out if you hold it for a while, and also with the provocative closing thought that catalyst poisoning after shape-shifting would get you back to a traditional elastomeric material.