Monday, August 8, 2022

Well, *that's* an interesting letter

Tucked in the letters to the editor in C&EN, an unusual comment about the death by exposure to dimethylmercury of Dartmouth chemistry professor Karen Wetterhahn: 

I read the article on Karen Wetterhahn with a profound sense of loss and sorrow that has not fully abated after 25 years. I was one of Karen’s graduate students, and I have come to recognize that she trained her other students and me very well.

I’d like to share an anecdote about Karen’s teaching style. Karen cultivated an air of omniscience, which certainly drove her students to prepare well for discussions with her about their research. One of her favorite questions was, “Don’t you know?,” implying that the student had not done their homework. Karen had wide-ranging knowledge, but she also had human limitations. Once, I called Karen’s bluff and confessed that I didn’t know the answer to a question, so I asked her what the answer was. We both chuckled when she admitted that she didn’t know either. Nevertheless, all her students learned the importance of asking insightful questions.

This brings me to a deeply troubling point raised in the article. I don’t agree with the conclusion about how Karen was poisoned. Karen taught me that if you disagree, you better have data on your side, so here goes. The New England Journal of Medicine article estimated that Karen likely absorbed about 1,344 mg of mercury, meaning she likely absorbed 0.44 mL of dimethylmercury. To do so meant she had to have been splashed with more than that—probably closer to 1 mL since some of the compound would be lost to evaporation or remain in the glove. This is a lot more than a drop or two.

When I was in Karen’s lab, I did some experiments using coaxial nuclear magnetic resonance tubes, which allowed a small volume of an external standard between the tubes. I don’t know what Karen was using for an NMR tube, but in currently available technology, where the reference goes into the center of a larger sample tube, typical volumes for the inner reference standard for a 5 mm tube are 60 ┬ÁL, while the outer sample volume is 10×. If Karen was using less than 0.1 mL of dimethylmercury, how could she have absorbed 10× what she was transferring? (Her lab notebooks might provide insight.) My supposition is that either she was splashed with more dimethylmercury than what was released from the pipette through her glove, or there was another method of ingestion, conceivably involving the deliberate actions of another individual.

Samuel Brauer
Shelton, Connecticut

Editor’s note: An investigation into Karen Wetterhahn’s death concluded, “The rapid, monophasic, first-order increase in the mercury content of hair is consistent with either one or several episodes of exposure to dimethylmercury beginning on or about August 14, 1996, and is consistent with the evidence (reports from coworkers and information from labeled vials and laboratory notebooks) that a single accidental exposure to dimethylmercury occurred on August 14. . . . Our patient’s accidental exposure may have resulted from both transdermal absorption of the liquid (given the lack of protection provided by disposable latex gloves) and inhalation of vapors (even though the work was conducted under a fume hood)” (N. Engl. J. Med. 1998, DOI: 10.1056/NEJM199806043382305). Wetterhahn did not record in her lab notebook the quantities she used or planned to use, according to John Winn, a Dartmouth professor emeritus of chemistry, who was chair of the department when Wetterhahn died.

I'm not an analytical chemist, so I can't pretend to have an educated opinion about who is right or who is wrong, but it seems that there are more possible explanations other than deliberate poisoning...

(Read all the letters for lots of articles about dimethylmercury in the good old days...)


  1. Coaxials for 10 mm NMR tubes do exist, but external referencing is still a viable method. Not an explanation, but is worth pointing out.

    I'm not sure why foul play is being suggested now...

    On a slightly different note (and I'm not sure if people have already mentioned this before), the best NMR standard for 199/201Hg NMR today is still neat Me2Hg. Sure we could use something less toxic like mercury chloride (like those calomel electrodes), but it's just not the same. It's not just purely for chemical shift referencing - it's also for the initial power and pulse length setup. (See also:


  2. I remember being told about this incident while I was in undergrad and that made me far more aware of PPE (especially compatible gloves) while attending a school that was pretty lax with PPE and my grad school was even worse. Especially when people would get DCM or chloroform on their gloved hands and wonder why their hands were burning 30 seconds later...

  3. Here is a plausible scenario: she could have accidentally knocked over an open small bottle of neat Me2Hg in the glovebox - those thick glovebox gloves are so clumsy - and immediately realized that an entire 5g bottle spilled in the glovebox is a serious screwup, then worked diligently to clean up the mess, changed the gloves, moved all the towels soaked with Me2Hg out of the glovebox, into some fume hood, and she did not report that mishap initially because the safety departments tend to be so unhelpful and everything leads to long investigation and risk re-assessment, which could lead to restrictions for her project. During the cleanup she got exposed some more, but fell sick only after considerable delay, only then she reported it.

    Also the glovebox gloves could have had a small needle puncture - since changing glovebox gloves is a chore, and students accidentally puncture the fingers of gloves quite often (and certain brands of gloves also crumble in creases between fingers) the common practice is just to use electrical vinyl tape and tape that puncture from inside - it holds for a bit longer and it is "good enough" not to spoil oxygen levels as there is a mild overpressure inside the glovebox


looks like Blogger doesn't work with anonymous comments from Chrome browsers at the moment - works in Microsoft Edge, or from Chrome with a Blogger account - sorry! CJ 3/21/20