Cycloaddition reactions between alkenes and dienes are textbook organic chemistry and are often a headache for undergraduate students learning chemistry. Most chemists think of Diels-Alder and thermally allowed formations of six-membered rings when they hear of cycloaddition reactions. Our group discovered some time ago that iron catalysts with pyridine(diimine) ligands promote rare examples of [2+2] cycloadditions of unactivated alkenes or dienes to form four- rather than six-membered rings under thermal conditions.1-3 One particularly exciting example is the iron-catalyzed reaction between butadiene and ethylene. While often cited as the prototypical Diels-Alder reaction (that really doesn’t happen under typical laboratory conditions),4 an iron catalyst promotes the selective formation of vinylcyclobutane, highlighting the preference for squares over hexagons (Figure 1).

Figure 1. Intermolecular [4+2] and [2+2] cyclization of butadiene and ethylene. 

Vinylcyclobutane can be considered to be ethylene substituted with a cyclobutane group on one carbon. Rose Kennedy, a postdoc in the lab at the time and now an Assistant Professor at the University of Rochester realized this, and reasoned that butadiene and vinylcyclobutane could also engage in [2+2] cycloaddition. Each propagating cycloaddition would leave another vinyl group eligible for further [2+2] cycloadditions. In practice, by reducing the amount of ethylene relative to butadiene, Rose obtained a higher boiling, more viscous material that she recognized had enchained multiple butadienes, generating organic molecules with multiple cyclobutanes.

These discoveries inspired a new postdoc and lead author on the current paper, Megan Mohadjer Beromi, to explore the possibility that butadiene alone could undergo a cycloaddition polymerization to give a unique macromolecular architecture, as butadiene can act as both a mono-olefin and a di-olefin. This turned out to be true as exposure of neat diene to a pyridine(diimine) iron bis(dinitrogen) precatalyst generated an oligomer with cyclobutane repeat units terminated on each end by a vinyl group. These end groups define what is known as a telechelic material and offer the opportunity to add functional groups to the end of the chain through well-established alkene chemistry. More interestingly, however, was the observation that re-exposure of the cyclo(oligomer) to the iron catalyst under vacuum resulted in retro-[2+2] cycloaddition – meaning that the oligomerization can be reversed and the new material is chemically recyclable! This is the first time that chemical recyclability back to monomer has been observed from a polyolefin – a material prepared from one of the major feedstock hydrocarbons used on enormous scale in the chemical industry. While in principle revolutionary, there were (and still are) practical challenges we faced in the laboratory. First and foremost, the iron catalyst coordinates butadiene, so driving the equilibrium to the side of free diene was a challenge. We found that sequestering the butadiene in 5 Å molecular sieves favored deoligomerization. Another challenge that persists is the morphology of the higher molecular weight material. The good news is that we made a hard plastic; the bad is that it is hard to do chemistry on a sparingly soluble material!

We knew when we observed the cycloaddition oligomerization of butadiene that we had done something special. After all, the original polymerization of butadiene, discovered by Sergei Lebdev in 1910, produces acyclic material that make up our car tires and shoe soles, among other applications (Figure 2).5 What good, if any, is the material we made? There were questions even more fundamental than that – what really did we make? What is the regio- and stereochemistry of the enchainment? What are the bulk materials properties?  To do this, we needed expertise beyond what was available in our research group at Princeton. We reached out to scientists at the ExxonMobil Chemical Company in Baytown, Texas – a team that routinely studies catalysts that make commercial polyolefins and that has expertise ranging from organometallic chemistry and catalysis to theory and materials characterization. We are fortunate to work with such a diverse and talented team as the material offered quite a few surprises.
 Figure 2. Polymerization of butadiene leading to acyclic polymers (top) and polymers of squares (bottom). 

Alex Carpenter was the lead and he fortunately was able to scale the oligomerization reaction and conduct it safely in neat diene at higher temperatures. Equally as important, Alex assembled a team of scientists to tackle the characterization of this new and interesting material. In a heroic effort, Sarah Mattler collected 1H and 13C NMR data and conducted additional two-dimensional experiments and spent many hours fully assigning the spectrum and ultimately established the cyclobutyl repeat unit and the 1,3-regiochemistry of the enchainment. It is not often one gets to assign the structure of new polyolefin but it is indeed challenging work. Joseph Throckmorton studied the X-ray scattering of the material and establishing its liquid crystalline behavior – a rarity for a polyolefin. Jarod Younker was able to accurately simulate the superstructure and explain this behavior. With this team, we were able to fully characterize the polymer of squares as a material with truly distinct, and exciting, physical properties. 

1. Bouwkamp, M. W.; Bowman, A. C.; Lobkovsky, E.; Chirik, P. J. J. Am. Chem. Soc. 2006, 128, 13340-13341.
2. Russell, S. K.; Lobkovsky, E.; Chirik, P. J. J. Am. Chem. Soc. 2011, 133, 8858-8861.
3. Hoyt, J. M., Schmidt, V. A.; Tondreau, A. M.; Chirik, P. J. Science 2015, 349, 960-964. 
4. Joshel, L. M.; Butz, L. W. J. Am. Chem. Soc. 1941, 63, 3350-3352.
5. Feldman, D. Polymer history. Des. Monomers Polym. 2008, 11, 1-15.