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

« Novartis Gets Out of RNAi | Main | The Latest Protein-Protein Compounds »

April 16, 2014

Professor Fukuyama's Solvent Peaks

Email This Entry

Posted by Derek

See Arr Oh expresses some doubts about all the NMR spectral corrections we've been seeing lately. He's specifically referring to Bethany Halford's interview piece, and he has this to say after reading it:

If your group focuses on "clean up your spectra" more than "purify your compounds better," that's a communications issue. If a professor with a large group sees nothing but perfect spectra all day, two thoughts should crop up:

1. "I must have the smartest, most efficient students in the world," or...
2. "Something's fishy here."

Perfect-looking data should always be a cause for concern in any experiment. My guess is that Prof. Fukuyama was closer to Option One, though, possibly in the variant of "My group has such high standards!" But high standards or not, a series of perfect, flat, NMR spectra with no solvent and no impurities is rather hard to achieve in total synthesis, considering the quantities that are being used. Load up the tube with 50mg of material and you can make a lot of stuff look good, but you don't have fifty mgs at step thirty-four, do you? I remember putting everything I had into one NMR tube (or worse, one polarimeter tube) in my own total synthesis days, and I carried the thing down to the machine like it was a bag of gold.

But there's no doubt that in a big group, there will be people who try to slip things past the boss. I've seen it myself; I'm sure that a lot of you have. And if you're giving the boss exactly what the boss wants to see - perfection - then it's going to be a lot easier. These spectral problems look like a collaborative effort to me - expectations from above, willingness from below. And there are a lot of other groups that have done (and, I feel sure, still do) the same thing. Zapping the solvent peaks in the NMR is the least of it, in some cases.

Update: added a direct link to the Fukuyama/Yokoshima interview.

Comments (19) + TrackBacks (0) | Category: The Dark Side | The Scientific Literature


1. Hap on April 16, 2014 8:24 AM writes...

Isn't part of a professor's job to teach students how not to fool themselves? It's hard to do that if you're playing "What I can't see increases the yield" with your students' NMR spectra. As noted by Quintus via Chemjobber here, someone in authority had to be in on the analysis of the spectra, so "I didn't see it" doesn't hold water, either. Given the state of most compounds after N steps of a total synthesis, a perfectly clean spectrum should have been cause for concern.

On top, unless peak manipulation and other forms of scientific dishonesty are pretty rampant, how did the professor end up with all of the cheats in his group? Is his group just a really bad judge of character? (Given the "It's someone else's fault" line, that is a possibility.) It seems more plausible that the group or the professor either implicitly or explicitly said "These spectra have to look nice, or you're not getting out of here, and I don't care how you do it.", and the students listened. If you set up perverse incentives, and back them up (you probably wouldn't have to do so explicitly), then voila! Clean spectra and high yields.

Permalink to Comment

2. Gyges on April 16, 2014 8:24 AM writes...

I wonder what interesting discoveries they've lost under the blizzard of snow-paper. The tiny impurities that suggested an interesting side reaction that could've been really important if they'd but tried to characterise the products.

Worse still, others will not bother working in the same area; others who don't use tippex and would find these rogue peaks.

Discoveries not lost for ever but not discovered as soon as they should have been.

Why is this methodology funded and supported?

Permalink to Comment

3. NMH on April 16, 2014 8:32 AM writes...

My post-doc advisor got extremely angry if anyone admitted to a mistake. Now, this guy was a little on the crazy side, but I dont think he is unusual. Not really good for the intellectual honesty of lab if you have an advisor with that attitude.

Permalink to Comment

4. Teddy Z on April 16, 2014 8:57 AM writes...

When I was in grad school, I knew a grad student who was the expert for the group of adding in the reference peak for all IR spectra that the PI insisted must be taken.

Permalink to Comment

5. Anon on April 16, 2014 9:31 AM writes...

I knew a girl back in grad school who had a PI (without saying names he was THE prominent figure in metastasis) continually suggest he would let her go if she "couldn't get the experiment to work." She ended up quitting and entering medical writing. When you have someone over your shoulder threatening to hinder your career (which would essentially end it, in the current funding environment) not everyone is willing to give up what they've work towards.

Permalink to Comment

6. John Wayne on April 16, 2014 11:26 AM writes...

You get the sort of behavior that you reward.

Permalink to Comment

7. JC on April 16, 2014 12:59 PM writes...

Rotovap down from carbon tetrachloride. Problem solved.

Permalink to Comment

8. Hap on April 16, 2014 1:24 PM writes...

@7: That only works for solvent peaks - it doesn't help for impurities, which would be the harder things to remove (especially on small scale in a total synthesis, where you don't have much to lose and purifications that clean the material do so at cost of product mass).

On the other hand, perhaps it solves the product mass issue (it has been alleged that the CCl4 soln was a standard technique for at least one big synthetic group, for precisely this reason - advisor wanted yields so...) - "Look, I got 120% yield! Guess I'd better leave it on the high vac for a moment until it gets below 110%."

Permalink to Comment

9. Anonymous on April 16, 2014 2:01 PM writes...

I'm sure this suggestion has been brought up before in similar discussions, but what reasons exist for not requiring, or simply voluntarily including, the raw FID files as part of the SI?

Permalink to Comment

10. a. nonymaus on April 16, 2014 2:28 PM writes...

Re: 9
Raw FID files can be manufactured. Just decide what are the peaks that you want with the linewidth that you want, perform an inverse Fourier transform, add some noise, and save in the desired format. Without automated validation similar to CIF-check, requiring FIDs won't help much.

I do know well the feeling of loading the world's supply of a compound into an NMR tube and blocking off all weekend to acquire scans. The spectra had merely O.K. snr and showed the presence of an isomer. But, we analysed the whole spectrum, rationalised the presence of the two species, and didn't reach for the white-out.

Permalink to Comment

11. Anonymous on April 16, 2014 11:24 PM writes...

Many times I've reached the desired purity or above it by throwing the material in hexanes, spinning on the rotavap without vacuum for 1 hour in a warm water bath, and then decanting off the hexanes. It's amazing how well it works when even your best repeated chromatography efforts fail. Of course you can't do this though on really small scales

Permalink to Comment

12. sepisp on April 17, 2014 3:52 AM writes...

Can someone explain where this utter idiocy of requiring 100% purity in each step in a 50-step synthesis comes from? Is it still the good old days of determining identity with refractive index and melting point? Some sort of a misguided academic or intellectual "purity"?

Solvent peaks might be due to the student being lazy and not high-vacuuming for long enough, but does it really matter if we introduce a 1-2% error in an academic total synthesis yield? Identity check is anyway with NMR.

Determination of selectivities, side products and impurities should be the focus of these efforts, not something to avoid. This sort of "99% or die" attitude might somewhat explain why the relevancy and even value of total synthesis, and indeed the whole of synthetic and medicinal organic chemistry is questioned.

And what's bad is that it's these same PIs that the current system rewards. That, combined with Asian students trained to not think and do exactly what the boss says produces nothing but practically useless papers.

Permalink to Comment

13. eugene on April 17, 2014 11:11 AM writes...

"Many times I've reached the desired purity or above it by throwing the material in hexanes, spinning on the rotavap without vacuum for 1 hour in a warm water bath, and then decanting off the hexanes."

That is a really good idea! I'll try it next time I have an oil or solid that doesn't dissolve in hexanes. The rotovap is perfect for this lazy solution that bypasses recrystallization.

Permalink to Comment

14. Anonymous on April 17, 2014 3:14 PM writes...

If you're going to fudge a spectrum, why not just make up the entire spectrum *instead* of doing any practical work?

Permalink to Comment

15. Anonymous on April 17, 2014 3:15 PM writes...

^ Then you'd need to add fake impurity peaks to make it look more real.

Permalink to Comment

16. Concerned Citizen on April 18, 2014 10:33 AM writes...

This really hits close to home. I witnesses this happen multiple times during grad school. One can blame the student certainly but In my experience the blame rest squarely on the PI.

In one case a student presented a yield of ~20% on a reaction to his PI. The PI said no do it again I want better. Again the the student repeated being very careful at each point during the reaction and once again got ~20%. The PI again declared that unacceptable and to do it again. Repeat this cycle many times with the PI getting more and more angry and mean each time. Then one day after a particular chewing out session from the boss this student got a 90% yield, the reaction was wonderful now. And the PI stopped being mean to him. Can anybody guess what lesson the student learned from the PI?

The all to common tactic of demanding certain results from students and being cruel or mean when they can not get them inevitably leads to the students giving the PI what ever they want by any means necessary.

Permalink to Comment

17. cliffintokyo on April 18, 2014 9:00 PM writes...

Thanks! I just realized where all the 'shoot the messenger' bosses in pharma industry come from.

Permalink to Comment

18. Anonymous on April 22, 2014 5:20 AM writes...

This story does amuse me. Working in a Japanese company at a non-Japan site, I see this type of 'data airbrushing' regularly in med chem meetings. A 'validated target' where the molecules have the same effect in the KO animals as they do in wild-type, the insoluble candidate molecule, 'oral drug' candidate that needed to be injected to see in-vivo name it. In fact, one of the authors of a retracted paper is a colleague leading one of the above programs..SNAP!!

Permalink to Comment

19. Artem on April 29, 2014 8:28 AM writes...

Hi folks, there are different situations and different compounds for NMR.
1. I dealt myself with people, which hands are growing from the right hands, so they were able to synthesize such compounds, that nobody could reproduce their synthesis afterwards. I am sure, that they were not liers.
2. Once I did a nice reaction and I have derived my target compound, then I have spent 6 months - and I was desperate to reproduce that experiment myself....
3. I have perfect NMR data, but I did not lie.... I have no reason to lie.... I am just proud, that I have nice experience in synthetic chemistry.

3. I have perfect NMR data, but I did not lie.... I have no reason to lie.... I am just proud, that I have nice experience in synthetic chemistry.

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