About this Author
DBL%20Hendrix%20small.png College chemistry, 1983

Derek Lowe The 2002 Model

Dbl%20new%20portrait%20B%26W.png After 10 years of blogging. . .

Derek Lowe, an Arkansan by birth, got his BA from Hendrix College and his PhD in organic chemistry from Duke before spending time in Germany on a Humboldt Fellowship on his post-doc. He's worked for several major pharmaceutical companies since 1989 on drug discovery projects against schizophrenia, Alzheimer's, diabetes, osteoporosis and other diseases. To contact Derek email him directly: Twitter: Dereklowe

Chemistry and Drug Data: Drugbank
Chempedia Lab
Synthetic Pages
Organic Chemistry Portal
Not Voodoo

Chemistry and Pharma Blogs:
Org Prep Daily
The Haystack
A New Merck, Reviewed
Liberal Arts Chemistry
Electron Pusher
All Things Metathesis
C&E News Blogs
Chemiotics II
Chemical Space
Noel O'Blog
In Vivo Blog
Terra Sigilatta
BBSRC/Douglas Kell
Realizations in Biostatistics
ChemSpider Blog
Organic Chem - Education & Industry
Pharma Strategy Blog
No Name No Slogan
Practical Fragments
The Curious Wavefunction
Natural Product Man
Fragment Literature
Chemistry World Blog
Synthetic Nature
Chemistry Blog
Synthesizing Ideas
Eye on FDA
Chemical Forums
Symyx Blog
Sceptical Chymist
Lamentations on Chemistry
Computational Organic Chemistry
Mining Drugs
Henry Rzepa

Science Blogs and News:
Bad Science
The Loom
Uncertain Principles
Fierce Biotech
Blogs for Industry
Omics! Omics!
Young Female Scientist
Notional Slurry
Nobel Intent
SciTech Daily
Science Blog
Gene Expression (I)
Gene Expression (II)
Adventures in Ethics and Science
Transterrestrial Musings
Slashdot Science
Cosmic Variance
Biology News Net

Medical Blogs
DB's Medical Rants
Science-Based Medicine
Respectful Insolence
Diabetes Mine

Economics and Business
Marginal Revolution
The Volokh Conspiracy
Knowledge Problem

Politics / Current Events
Virginia Postrel
Belmont Club
Mickey Kaus

Belles Lettres
Uncouth Reflections
Arts and Letters Daily
In the Pipeline: Don't miss Derek Lowe's excellent commentary on drug discovery and the pharma industry in general at In the Pipeline

In the Pipeline

« CETP, Alzheimer's, Monty Hall, and Roulette. And Goats. | Main | The Chemistry Jobs Market »

January 25, 2013

Down With the Ullmann

Email This Entry

Posted by Derek

Have I mentioned recently what a pain the rear the Ullmann reaction is? Copper, in general? Consider it done, then. I'm trying to make biaryl ethers, not something I'd usually do, and these reactions are the traditional answer. One of my laws of the lab, though, is that when there are fifty ways of doing some reaction in the literature, it means that there's no good way to do it, and the Ullmann is the big, hairy, sweaty example of just that phenomenon. Even when it works, there are worries. But you have to get it to work first. . .

Comments (22) + TrackBacks (0) | Category: Life in the Drug Labs


1. MTK on January 25, 2013 11:50 AM writes...

Obviously not suitable for many substituent patterns, particularly electron rich ones, but I really like using nucleophilic aromatic substitution to form aryl ethers.

If you can find the right aryl fluoride, it can work like a champ.

Permalink to Comment

2. petros on January 25, 2013 12:01 PM writes...

When I made some biaryls I found zinc couplings worked well, and were reasonable to work up

Permalink to Comment

3. InfMP on January 25, 2013 12:01 PM writes...

Use the Chan-Lam-Evans instead. I've done Ullmans at 30% go to Chan-Lams at >99% on 50g scale.

My friends swear by N,N-dimethylglycine as Ullman ligand

Permalink to Comment

4. Derek Lowe on January 25, 2013 12:15 PM writes...

I'm a big fan of SnAR reactions, too, but my electronics are wrong in this case. N,N-dimethylglycine, proline, et al. are also failing me here (although I have had them work pretty well in the past, for N-couplings - problem with using proline is that sometimes it comes in and couples, too).

The Chan-Lam-Evans is an excellent idea; I'd completely forgotten about that reaction. I'll go set one of them up, as soon as I finished my luxurious lunch here at my desk!

Permalink to Comment

5. processchemist on January 25, 2013 1:08 PM writes...

Did you evaluate Ar-OH + Ar-X with KF in DMSO? I don't remember if this reaction has a name or what...

Permalink to Comment

6. Luysii on January 25, 2013 1:19 PM writes...

"when there are fifty ways of doing some reaction in the literature, it means that there's no good way to do it"

True in spades for the treatment of any disease.

Permalink to Comment

7. Erik on January 25, 2013 5:40 PM writes...

Definitly a painful reaction. A few years back I struggled with trying to identify a pseudo general procedure for pmc work. I found Copper (ii) chloride (May have been Cu(I)Cl...was awhile ago and I don't have my files on the train) with 2,2,6,6 tetramethylheptane-3,5-Dione as ligand at 130c in NMP worked as well as you could hope. Its not flawless but it was the winner more often than not the few times Ive attempted an ulman since.

Permalink to Comment

8. partial agonist on January 25, 2013 5:42 PM writes...

It seems to me to be a very, very substrate-sensitive reaction-- one procedure might work great for a series of compounds, but apply it to another scaffold and you just crash and burn.

That's what continues to amaze me about Suzuki and a few other Pd reactions. You try the same procedure on 100 different substrates and guess what, most of them actually work (80%, 90%?).

Permalink to Comment

9. L J on January 25, 2013 6:03 PM writes...

We did almost all of our difficult Ullmann couplings in the microwave with 1,10-phenanthroline, 10 mins at 80 to 100 deg. Watch that you get all the copper off the sides of the glass and down into the liquid though. We had several "go off" before we realized the metal made hotspots.

Permalink to Comment

10. dvizard on January 25, 2013 7:54 PM writes...

I might have to dig out my Master thesis, I remember doing something like that and losing a lot of time with it...

Permalink to Comment

11. Beef Supreme on January 25, 2013 8:19 PM writes...

In my experience, the copper source and ligand need to be tweaked for each substrate class. I did a small set with CuI/1,10-phenanthroline that worked OK in toluene at 100 with Cs2CO3 (c. 50% isolated). Then install a single ortho methyl group and 0% yield for every reaction. Switched to N,N'-ethylenediamine and CuOAc, and those worked ok.

Seems like the halide matters too, which makes sense if you think about the differences in disproportionation of CuCl CuBr and CuI

Permalink to Comment

12. Secondaire on January 25, 2013 8:34 PM writes...

"In my experience, the copper source and ligand need to be tweaked for each substrate class."


"when there are fifty ways of doing some reaction in the literature, it means that there's no good way to do it"

Oh man. This reminds me of an old Amos Smith joke:

Q. Why are there thirty ways to cleave a dithiane?
A. Because none of them work.

Permalink to Comment

13. Curt F. on January 26, 2013 10:09 AM writes...

The reaction might not be so great, but the encyclopedia is pretty good. Long Live Ullmann's!

Permalink to Comment

14. simpl on January 28, 2013 8:30 AM writes...

Judging by the number of times I have set up and inactivated CuI in our production system, if it makes it through development, it doesn't scale up well, either.

Permalink to Comment

15. nitrosonium on January 28, 2013 9:23 AM writes...

agree with #3. chan-lam is the way to go if you can get away with it. easy to run and work up. and really great yields. i spent weeks trying to make the necessary biaryl-ether linkage through Ullmann/Buchwald rxns. all i got was black tar. tried the Chan-Lam on 15g and got 95% with no chromatography

Permalink to Comment

16. processchemist on January 28, 2013 9:29 AM writes...

In my experience, yes, every substrate needs tuning, but plain Ulmann (or Rose,mund), is the best when you're in the 10g-1000 g scale. DMF (NEVER DMSO), CuI, 160 °C and most of the times the job is done, when the nucleophile is N. For really electron poor rings, DMSO/base 100 °C worked great, but DMF stinked... so there's definitely a solvent effect to make the landscape "more interesting".....

Permalink to Comment

17. Kent G. Budge on January 28, 2013 10:32 AM writes...

"One of my laws of the lab, though, is that when there are fifty ways of doing some reaction in the literature, it means that there's no good way to do it,"

Kind of how I feel about diabetic diets. The fact that the recommendations are all over the chart, from Bernstein's "As few carbs as possible" to the old ADA recommendation of 70% calories from carbs, suggests that nothing really works better than anything else.

Which doesn't keep me from sticking fanatically to what seems to work for me...

Permalink to Comment

18. another process chemist on January 28, 2013 12:37 PM writes...

We just did a few on 20 kg scale each. Our preferred conditions (tailored for our substrate, as mentioned above) for a pretty non-nucleophilic phenol:

1 eq. Phenol
1.5 eq. Phenol
2.1 eq. Cesium Carbonate
0.75 eq. Cu(I)Cl (99% better than 99.999%)
0.375 eq. TMHD
20 mL Dioxane/g of Phenol
100 °C, 10-15 hours

75-77% yield on 20 kg scale.

It can be done, but has been the bane of our existence for a long time. We've just found we can make some substrate changes and use an SNAr, which is a great help.

Permalink to Comment

19. another process chemist on January 28, 2013 12:39 PM writes...

Sorry about the mistake in the previous post. Should be:

1.0 eq. Phenol
1.05 eq. Aryl Bromide

Permalink to Comment

20. Anonymous on February 27, 2013 7:05 AM writes...

How did your Chan-Lam reaction go? Trying it out myself and the yields are beyond pathetic.

Permalink to Comment

21. Derek Lowe on February 27, 2013 7:55 AM writes...

#20, I hate to confirm your results, but I had very poor luck with the Chan-Lam conditions myself. With some combinations, the reaction seemed to go to about 10% and stop in its tracks, and with others, it never ever got that far. I tried six or eight variations, but with no success. Colleagues here report that they've had it work on another system, with decent yields, but with glacial slowness. I kept things running for a long time, but mine never went anywhere. So it's another very substrate-dependent reaction, apparently - like the Ullmann itself.

Permalink to Comment

22. Galen on March 1, 2013 8:48 PM writes...

Tried the heat and beat approach with copper metal?

1 eq. phenol
1.2 eq NaH

stir 1h, then
2 eq aryl halide
.5 eq copper powder
more DMF

heat 160C for 60h

only works well if product is distill-able as black goo is the result.

Permalink to Comment


Remember Me?


Email this entry to:

Your email address:

Message (optional):

The Last Post
The GSK Layoffs Continue, By Proxy
The Move is Nigh
Another Alzheimer's IPO
Cutbacks at C&E News
Sanofi Pays to Get Back Into Oncology
An Irresponsible Statement About Curing Cancer
Oliver Sacks on Turning Back to Chemistry