The David Milstein group aims to develop challenging, unprecedented, sustainable, green catalytic reactions. As such, many successful projects are developed upon untold initial unsuccessful experiments that only former lab members know. 

I started to work with David in July 2012 as a postdoc. The first thing that David asks every new postdoc is to talk to lab members and find out what they want to tackle. I picked up one of the most difficult projects that have been investigated since the late 1980's in the group. For this catalytic reaction, reductive elimination was believed to be the crux of the cycle. This is why I designed the bulky PN ligand 1, which was assumed to promote reductive elimination like Buckwald ligands do while maintaining a metal-ligand cooperativity function developed by the Milstein group (Figure 1). 

Figure 1. A design principle of ligand 1.

Unfortunately, after a few months of study, we concluded that the aimed reaction could not be achieved using 1. This brought us to repurpose 1 for another challenging reaction, Fe-catalyzed olefin metathesis. Initially, we tried to prepare Fe-carbene complexes using carbene transfer methods used to prepare Ru-carbenes (Figure 2).

Figure 2. Envisioned carbene transfer method. 

Soon, we found that we can not make the carbene complex using a diazo compound, as it oxidizes the phosphine ligands, as observed by Grubbs, who pioneered this captivating challenge of Fe-catalyzed olefin metathesis decads ago. After a year of failure, we managed to isolate an iron-carbene complex using a different ligand set. However, as is the case of other isolable Fe-carbene complexes (examples), it did not show olefin metathesis reactivity. After another year of failure, I gave up on this method and started to examine an alpha-H elimination method using Fe(III)-alkyl complexes  (Figure 3).

Figure 3. Envisioned alpha-H elimination method. 

This method turned out to be promising. Initially, we obtained a ring-opening metathesis polymerization (ROMP) polymer of norbornene in a low yield, and soon we improved the yield to > 95% using a ligand with PⁱPr₂ group in place of the P^tBu₂ group. However, this result came with a big problem, irreproducibility. The activity of the catalyst was uncontrollably different from one batch to the other. Thus, a frustrating search for the key ingredient was started. After two years of detective work and four years of postdoc, we decided to finish up the project without ultimate success. With large disappointment, I started collecting ¹H NMR data of polymers. 

During the detective work, there was a polymer that often forms in high yield when I use Fe(II) precursors, and I could not obtain ¹H NMR of it due to insolubility. Considering the precedence of Fe(II)-alkyl catalyzed addition polymerization, I thought that it is the undesired addition polymerization polynorbornene that has low solubility, unlike ROMP polynorbornene (Figure 4).

Figure 4.  Possible Fe(II)-alkyl catalyzed addition polymerization.

This polymer always came out as a dish around a stirring bar, and after washing and drying, it becomes a mass with a marshmallow-like texture that is very different from Ru-catalyzed ROMP polynorbornene (Figure 5).

Figure 5. Polynorbornene prepared by Fe or Ru-catalyst. 

On June 16, 2016, a month before the end of my postdoc contract, I took the polymer for NMR analysis. I noticed that the polymer became brittle while standing on the bench for two months. The brittle polymer was soluble in CDCl₃, and to my surprise, and explosion of excitement, the ¹H NMR showed spectra of ROMP polymer !!! (Figure 6) I rushed into David's office.

Figure 6. Our first ¹H NMR spectra of Fe-catalyzed ROMP polynorbornene.

It was Thursday, the end of a weekday in Israel. I drank the tastiest Goldstar beer in my life with David and group members. After additional six years of journey, here we finally report our discovery. You can also read it here for free. Now, after reading this story, you may understand how supporting and optimistic David is. I sincerely thank for it.

In my own lab, I hope I can make a moment as exciting as I had in Israel.

Is anyone interested in working with me? 

Please feel free to leave a comment below.