Thursday, February 21, 2019


Our work on electrochemical Birch reduction was published today. In case you don't want to read the manuscript, here's the Cliff notes:

  • Pfizer approached us, saying traditional Birch is bad news for med chemists (alkali metals are spooky; condensing ammonia is gross; the whole thing is a lot of effort). They asked us to make a better Birch.
  • We developed some conditions to effect the Birch transformation with electrochem. Basically, the conditions are mix substrate, LiBr, electricity, DMU, a key phosphoramide additive (TPPA), and THF. No ammonia, no alkali metals, no cryogenics.
  • Intuitively, one might naively assume that what's going on here is that the LiBr is reduced to Li(0), and from there the reaction goes through a standard chemical Birch mechanism. We assumed that, naively (at first).
  • Turns out, that mechanism is not at all the case, as demonstrated by much experimental and computational mechanistic study. Instead, the substrate is reduced directly on the electrode. Further, Li(0) is never generated at all. Safe stuff. Cool beans.
  • We went back to Pfizer with this. They seemed pretty happy.
  • Oh, also the reaction works in flow, on >100 g scale.
  • Oh, also, the key phosphoramide additive, that makes everything work, came from the Li-ion battery literature.
A huge shout-out has to go to our collaborators: Shelley Minteer and her group (led on the student side by postdoc David) for electroanalytical work, Matt Neurock and his group (led on the student side by grad student Sagar) for computational work, and our longtime buddies at Asymchem (led by Longrui) for scaleup and flow work. We should also note that this whole collaboration is driven by the CCI electrochemistry consortium, which has been a tremendous way to put people from various subfields of electrochemistry in touch with one another.

Anyway, normally, this is the part of the blog where we'd tell some funny anecdotes about the project. Unfortunately, the whole team is way too excited to get back in the hood for subsequent projects, so we didn't write anything. In lieu of anecdotes, please accept a few vaguely-electrochemical-Birch-related poems:


We set out to Birch some arenes;
to make a scary reaction more clean.
It turned out lack of lithium
gave us complete freedom
to safely Birch on our Electrosyeeeens


Birch in Winter frost
with Electrochemistry
No lithium, good.


Poor Arthur Birch
did much research.
He was always sad,
for his reaction was bad.
Now we made it better
but alas, he is deader.

Oh, also we made a video, objectively comparing the time it takes to run a classical, chemical Birch, to the new electrochemical protocol. We hope you enjoy it:

- Electrochemical Birch team


  1. These poems are hilarious! Good job guys!

  2. Excellent work by a group of brilliant people! Bravissimo!

    However, the seemingly questionable yields reported in Fig.4 using Na dispersion/15-Crown-5 have inevitably undermined the credit of this wonderful ammonia-free Birch surrogate...

    1. Thanks for the feedback! We too were somewhat surprised to see how low many of the yields were with the Na-dispersion method. It's possible that we were running the reaction wrong, somehow, but we certainly used our best air- and water-free technique, and also used brand new reagents from Sigma. The yields reported are simply what we observed.

      Regardless, we believe it's good to have multiple synthetic tools in your quiver, and hope that this electrochemical method provides another orthogonal approach to Birch beyond the ones already published.

      ~ Birch team

    2. Thank you for your comments. Actually, I did try to reproduce Birch reductions on some substrates in Figure 4 using Na-dispersion/15-crown-5 method. Here are some of my observations:

      1) When 15-crown-5 was added dropwise into Na-dispersion in THF at 0 degC, blue color was only observed upon the first few drops of 15-C-5. The final suspension had the same color as if no 15-crown-5 was ever added.
      2) When the Na-dispersion/15-C-5 suspension in THF was cooled down to -50 degC or below, the dark blue color was observed and persisted. However, when warmed to 0 degC, the blue color disappeared.

      Therefore I was assuming that the electride was generated efficiently only at low temperature (e.g. -60 degC), at which point a solution of substrate and iPrOH in THF was added slowly. The blue color disappeared upon the addition. The temperature was maintained between -60 to -40 degC, In most of the cases, the blue color reappeared after 15-30 min's stirring, generally indicating the completion of the reactions. The progress could alternatively be monitored by TLC, which did not affect the reactions.

      Below are some results of Birch reductions using Na-dispersion/15-crown-5 method; the yields were determined by NMR using 1,3,5-trimethoxybenzene as an internal reference:

      1) Reduction of naphthalene: product 6 (in Figure 4) was obtained in 86% yield. (vs. 6% in this report)
      2) Reduction of 4-methylanisole: product 11 (in Figure 4) was obtained in 96% yield. (vs. 15% in this report)
      3) Reduction of 3-phenylpropionic acid: product 22 (in Figure 4) was obtained in 75% yield, together with 8% S. M. and 16% over-reduced product (vs. 14% in this report)
      4) Reduction of 1,3,5-tri-tert-butylbenzene: no reduction was observed, which was consistent with this report
      5) Reduction of 6-methoxyquinoline: the corresponding 1,2,3,4-tetrahydroquinoline was the only product observed (in 60% yield, vs. NR in this report) The color of the final reaction mixture was yellow (vs. dark blue in most cases)

      To this end, I believe Na-dispersion/15-crown-5 method is indeed a much more efficient and reliable Birch surrogate than what was described in this report. It would be of great help to the massive readers if the experiments involving Na-dispersion/15-crown-5 (mainly in Fig.4) could be re-performed and the corresponding yields could be revised.

      Nevertheless, the aforementioned discrepancies in some yields using Na-dispersion/15-crown-5 method do not obscure the value of this work, which would certainly open a door in the ere of electrochemistry.

    3. Thanks for your experiments!

      We ran our Na-dispersion/15-C-5 comparisons at 0 degC, which are the conditions reported in the literature. As with your notes, at 0 degC, we didn't observe any persistent blue color of the reaction. The temperature difference could explain the discrepancies between your results and ours.

      We should also note that our comparisons were independently run by two coauthors on the paper, with very similar results. So, at least across our lab, the comparison controls seem reproducible. We understand these kinds of head-to-head comparisons are serious; so we take extra diligence to exactly follow the methods reported in the lit, and reproduce them across several pairs of hands.

      At the end of the day, practitioners of the Birch should be aware of all of these methods (i.e., traditional Birch; Na-dispersion; electrochemical Birch) and decide for themselves what is most practical, cost-effective, and reproducible. It's always good to have more tools in the toolkit!

      ~ Birch team

  3. This comment has been removed by the author.