Vox Meditantis

This interview is a dramatised reconstruction, not a real transcript: Karen Wetterhahn died on 8th June 1997 and cannot speak, so her voice and responses here are imagined. The piece is grounded in historical sources about her life, research, exposure, and legacy, but any dialogue, inner thoughts, and interpretation are creative and included to protect readers from mistaking it for documentary record.

Karen Elizabeth Wetterhahn was a pioneering American chemist whose research on toxic metals fundamentally changed how laboratories understand and manage risk. She published over eighty-five papers on metal–protein interactions, heavy metal carcinogenesis, and organomercury chemistry, while also reshaping who felt welcome in science through institution-building at Dartmouth. Her accidental death in 1997 from dimethylmercury poisoning was not a story of carelessness, but the moment an entire safety paradigm was shown to be wrong.

What follows is a fictional interview imagining Karen Wetterhahn here, today, aware of her full scientific and posthumous legacy. The interviewer speaks from 2025; Karen speaks in the voice of a late‑20th‑century American scientist and educator – precise, wry, thoughtful, and deeply committed to both chemistry and people.

Professor Karen Wetterhahn was a chemist who spent her career asking a deceptively simple question: what really happens when toxic metals meet living systems? In the process, she helped build the field of toxic metal biochemistry, co‑founded a programme that doubled the proportion of women majoring in science at Dartmouth, and – through a tragedy she did not choose – forced the world to rewrite how laboratories understand “safety”. Her name is now threaded quietly through glove standards, NMR reference choices, and fellowship programmes, even if most researchers using those protections have never heard it.

Karen, thank you for joining this imagined conversation across time. It is an honour to speak with you, with the hindsight of 2025 and a full view of what your work has meant. To begin, for readers who may only know the outline – or only the accident – how do you introduce yourself?

Well, first, thank you. This is… unusual, but let’s go with “useful fiction.” I’m a bioinorganic chemist by training. I spent my career trying to understand how so‑called “inorganic” metals behave in the very organic mess of a cell – how mercury, cadmium, chromium, nickel, those misbehaving guests, interact with proteins and DNA and occasionally start cancers.

If we have to use a label, I’d say: “I studied how toxic metals break things in the body, and tried to keep people from getting hurt.”

Before the chemistry and the headlines, who were you? What shaped you into someone who wanted to spend a life on toxic metals and on changing who gets to do science?

I grew up in Plattsburgh, upstate New York. It was not a place overflowing with chemists. My parents were practical people; science wasn’t some romantic destiny. I liked puzzles. I liked cause and effect. I liked that chemistry let you see the invisible rules.

At St. Lawrence University in the late 1960s, I was often one of very few women in the room. People were polite enough, usually, but there was this ambient sense that you were a guest in someone else’s house. You learn to do excellent work partly so no one can say you got in by accident.

The metals came later. In graduate school at Columbia, working with Stephen Lippard, I discovered bioinorganic chemistry – this in‑between place where “simple” metal ions start doing very complicated things when you drop them into a protein or bind them to DNA. I liked that edge: not quite biology, not quite inorganic chemistry, but where the trouble begins.

And then, frankly, there was a social question. Factories emit cadmium; mines release chromium; communities downriver get cancer clusters. Somebody ought to understand what is happening at the molecular level so we’re not just guessing about risk. That “somebody” felt like a job worth getting out of bed for.

Many readers will know you, if they know you at all, through a sequence of numbers: three drops, fifteen seconds, three months, five months, ten months; and the blood mercury level – about 4000 micrograms per litre when 200 is already deeply toxic. How do you want that story told?

Carefully. And accurately.

On 14th August 1996, I was preparing a dimethylmercury standard for ¹⁹⁹Hg NMR. At the time, dimethylmercury was the “gold standard” – very sharp signal, stable, easy to handle in theory as long as you respected its toxicity. I wore a lab coat, safety glasses, a proper hood, and good‑quality latex gloves. That was not sloppy; that was precisely what the safety manuals recommended.

I spilled, we think, a few drops on my gloved left hand. Not a beakerful, not a dramatic splash. A few drops. The glove looked fine. No tears, no visible degradation. If you’d asked me then, I would have said the barrier was intact.

What no one in that room knew – because no one had ever measured it – is that dimethylmercury, being small and very lipophilic, can move through latex astonishingly quickly. Later tests showed penetration on the order of fifteen seconds. That is, by the time I could have said “Hmm, better change this glove,” those molecules would already be past the glove, past the skin, boarding the express train to my central nervous system.

Here is the point I insist on: I did not “ignore” safety. I obeyed it. The system was wrong.

For the chemists and toxicologists reading: can you walk through, step by step, what you were trying to do scientifically with dimethylmercury and ¹⁹⁹Hg NMR? Assume the audience is expert; you can use the technical language you would in a research seminar.

Certainly.

The overarching question in that project was: how do mercury species interact with DNA repair proteins – particularly those involved in nucleotide excision repair – and how might those interactions contribute to mutagenesis and carcinogenesis?

To probe that, we needed two things.

First, we needed well‑characterised mercury complexes with biomolecular ligands – cysteine‑rich peptides, small DNA fragments, simplified versions of key protein binding sites. Second, we needed a way to observe the mercury centre directly, not just infer its behaviour from the ligand.

¹⁹⁹Hg NMR spectroscopy is ideal for that. The isotope has a reasonable natural abundance and a broad chemical shift range that is exquisitely sensitive to coordination number, ligand type, and geometry. But to make sense of that, you need a stable reference with a sharp, well‑known resonance.

In the 1990s, dimethylmercury was that reference. It is a volatile liquid with a single, sharp ¹⁹⁹Hg resonance far from most common complexes, which makes it convenient for calibration. It is also, as it turns out, one of the most dangerous organomercury compounds we have ever encountered.

The protocol was roughly:

1. Prepare dilute solutions of mercury complexes – often low millimolar – in carefully dried, oxygen‑free solvents or buffered aqueous systems, depending on the ligand stability.
2. Record ¹⁹⁹Hg NMR spectra using a spectrometer tuned appropriately; the magnetic field homogeneity and probe tuning are non‑trivial for a low‑gamma nucleus like mercury.
3. Calibrate the chemical shift scale against dimethylmercury, either directly or by secondary standards whose positions had been established relative to dimethylmercury.
4. Use the observed shifts, coupling constants, and line shapes to infer coordination environment – distinguishing, for example, linear two‑coordinate complexes from trigonal or tetrahedral species, and to detect subtle changes upon binding to biomolecules.

Compared with indirect methods – UV–vis, fluorescence quenching, or even ¹H or ¹³C NMR on the ligand – ¹⁹⁹Hg NMR gives you a more direct readout on the metal centre, which is critical when small structural changes can translate into very different biological outcomes.

The trade‑offs were, and are, non‑trivial: sensitivity is lower, experiments are slower, and before my death, the community underestimated the safety cost of relying on a “super‑toxic” reference standard. That calculus has changed. People now employ much safer mercury salts in standardised conditions to achieve reproducible references, or they avoid direct dimethylmercury handling entirely. The resolution you get from dimethylmercury is not worth a human life.

Were there particular techniques or little tricks in your lab that never made it into the official methods sections, but that, in your view, really mattered for getting reliable data on metal–protein interactions?

Oh, several. Methods sections are always shorter than the truth.

One example: when working with cadmium complexes of DNA repair proteins, we found that trace contamination in “high‑purity” buffers could absolutely ruin binding curves. Cadmium is greedy; it will go after any stray thiol or carboxylate. So we developed a habit of pre‑treating buffers with Chelex resin, then re‑spiking defined amounts of essential ions. That’s a half‑sentence in the paper, but in practice it was a day of careful prep and a lot of muttering.

Another: for some ¹⁹⁹Hg spectra of protein complexes, we used very low temperatures – not just to stabilise complexes, but to slow exchange on the NMR timescale. That let us distinguish between multiple bound species that would otherwise blur into one broad line. In the formal write‑up, it reads “spectra were recorded between X and Y degrees,” but the interpretive work, figuring out where the line‑shape changes gave you the clearest mechanistic picture, was more art than cookbook.

And then there are the people skills that never show up. Getting a nervous undergraduate comfortable around “scary” metals without either minimising the risk or paralysing them with fear – that’s not in the supplementary information, but it absolutely affects your data, because a calm student pipettes more carefully than a terrified one.

Many students encounter your name in safety training slides rather than in the references of a paper on carcinogenic metals. They know the glove diagrams, not the science or the mentoring. How does it feel, from this 2025 vantage point, to be known mainly as a case study?

It’s… mixed.

On the one hand, if my story keeps a graduate student from being hurt – if someone double‑gloves with SilverShield laminate under nitrile because they once heard about “that Dartmouth chemist who died from three drops” – then that is not a trivial thing. The point of safety training is to save lives, not to polish my reputation.

On the other hand, I did not spend two decades writing papers so that the only narrative left would be “cautionary tale.” I was an expert in toxic metals. I helped build a women in science programme that actually moved the numbers. I mentored students who went on to their own important work. To compress that into “the woman who spilled dimethylmercury” is… let’s say intellectually lazy.

If we care about safety culture, we should tell the full story: that an expert, following the book, still died, and that the result was not simply “don’t be like her” but “the book was wrong; we must rewrite it.”

That point – that you did everything your era told you was “right” – is essential. Walk us through, in your own framing, why your death should be read as a systems failure rather than an individual mistake.

Start with what we knew and what we didn’t know.

We knew dimethylmercury was highly toxic. That was not news. We knew you should work in a hood, avoid skin contact, wear gloves, keep quantities small. Those were the institutional rules, and I followed them.

What no one had done was ask: Are these specific gloves demonstrably capable of stopping this specific compound for a meaningful amount of time? No permeation tests. No breakthrough curves. Latex was the standard for “chemical gloves,” so it was assumed to be adequate, full stop.

That is the heart of the systems failure: a culture that equated “we’ve always used this” with “this is safe enough,” without quantitative evidence.

Once colleagues tested glove materials after my accident, they found that dimethylmercury went through most common gloves – latex, PVC – within seconds to a couple of minutes. SilverShield laminate fared much better, especially when worn under a sturdier outer glove to resist tears. That data did not exist before. It should have. The compound had been in use for decades.

So yes, I spilled something dangerous. But spills, minor ones, are a foreseeable part of lab life. A robust system assumes that small mistakes will occur and designs layers of protection so they are not fatal. In 1996, that redundancy simply wasn’t there for organomercury compounds.

The right lesson is not “never slip” but “build environments where a slip is survivable.”

For those who only know that “she died of mercury poisoning,” would you be willing to lay out the timeline? Not for morbid detail, but to understand what delayed toxicity looks like when a compound crosses every barrier.

If it helps people take such compounds seriously, yes.

After the August spill, life was normal for several months. I continued teaching, supervising, writing. There were no dramatic acute symptoms.

Around November, about three months later, I began to have non‑specific complaints – abdominal discomfort, some weight loss. Those could have been anything: stress, a virus, ageing. Mercury was not my first guess.

By January, neurological signs emerged. I started losing my balance, slurring words. That triggered more serious investigation. Bloodwork came back with mercury concentrations in the range of 4000 micrograms per litre. For context, 200 micrograms per litre is already considered very dangerous. Those numbers told their own story.

We tried aggressive chelation therapy – agents like dimercaprol and later others – but by the time neurological symptoms are that pronounced, the damage is, in many cases, not reversible. Within a few weeks I deteriorated into a near‑vegetative state.

One of my colleagues later recounted a visit where my husband saw tears on my face and asked if I was in pain. The physicians said it was unlikely that my brain could even register pain at that point. I find that detail… difficult, but important. Neurotoxicity is not always a dramatic convulsion; sometimes it is a quiet unravelling, and by the time you notice, you cannot stitch it back together.

From exposure to death was about ten months. That gap is part of the horror. A delayed fuse, set by three drops passing silently through a glove in seconds.

Returning to the science: when you studied mercury and cadmium interacting with DNA repair systems, what did you learn about why these metals are so effective at causing long‑term damage?

Metals are not malicious, of course, but they have preferences.

Mercury and cadmium, for example, have a strong affinity for soft ligands – sulphur, especially, and to some extent nitrogen. Many DNA repair proteins, transcription factors, and enzymes involved in maintaining genomic integrity use cysteine residues – thiol groups – to coordinate essential metal ions or to maintain structural disulfide bonds.

When a toxic metal displaces an essential one – cadmium stepping into a zinc finger motif, say – or binds covalently to critical cysteines, you can knock out the protein’s function. If the protein’s job is to fix DNA damage, that failure doesn’t necessarily kill the cell immediately. Instead, it allows mutations to accumulate. Some of those mutations, given enough time and the wrong context, become cancers.

In our work, we used model systems to show that cadmium can inhibit nucleotide excision repair and that mercury compounds can form adducts with DNA and with repair proteins. The rates, affinities, and conformational changes we observed helped make sense of epidemiological data on workers exposed to cadmium or communities exposed to mercury: it wasn’t just “general toxicity.” There were specific molecular targets, with measurable binding constants and functional consequences.

Compared with older approaches that simply measured cell death or gross chromosomal damage, the bioinorganic approach let us say, “This cysteine cluster in this protein is a likely vulnerability.” That level of detail is what you need if you’re going to set occupational exposure limits that mean something, or design chelators that don’t just strip every metal indiscriminately.

At the time, what other methods were being used to study these interactions, and how did your NMR‑heavy, metal‑centric approach compare?

People certainly looked at metal toxicity before we brought in ¹⁹⁹Hg NMR or detailed protein models. Classical toxicology relied a lot on LD₅₀ values, histology, and crude measures of enzyme inhibition. Molecular biology focused on DNA adducts and mutations without always tracking the metal coordination chemistry carefully.

The advantages of our approach were:

• Direct observation of the metal centre, which reduced ambiguity.
• The ability to differentiate between species in equilibrium – “free” metal, loosely bound, tightly bound, different stoichiometries.
• Correlating structural information with specific functional assays – repair rate, binding affinity, etc.

The trade‑offs were:

• Instrument intensity: NMR time is expensive, spectra of heavy nuclei can be slow.
• The need for fairly high concentrations in some cases, which may or may not mirror physiological conditions.
• Safety concerns around the most convenient reference standards, as we’ve discussed.

Over time, as mass spectrometry improved and synchrotron‑based X‑ray techniques matured, other groups gained alternative ways to look closely at metal–biomolecule interactions. I am glad of that. A healthy field has multiple tools, each with its own biases and strengths.

Let’s talk about the Women in Science Project – WISP – at Dartmouth. To many, especially outside the college, this is the least‑known but perhaps most structurally important part of your legacy. How did it start?

It started, as many things do, with a nagging discomfort and a spreadsheet.

Around 1990, a few of us were looking at the numbers of women in science majors at Dartmouth. Women were enrolling at reasonable rates overall, but in the sciences, the fraction was something like 13 percent. That wasn’t just chance. It was a pattern.

We talked to students. Some had been steered away in high school. Some felt isolated in large lecture courses. Some had never met a woman scientist in person. Individually, each story was small; together they painted a picture of a pipeline with many small leaks.

WISP was our attempt to patch some of those leaks. We paired first‑year women interested in science with faculty mentors and with hands‑on research experiences. The idea was simple: if you can see yourself in a lab early, with someone taking you seriously, you are more likely to stay.

Over the next years, we watched the proportion of women science majors climb from about 13 percent to around 25 percent. That is not perfection, but it is a meaningful shift. And we did not achieve it by lecturing about diversity in the abstract; we did it by changing everyday experiences – who gets invited into research, who gets introduced at seminars, whose questions are answered respectfully.

You came up through chemistry in the 1960s–1990s, when women were still regularly treated as exceptions. What forms did bias take, and how did you navigate it?

It was rarely cartoonish villainy. No one slammed doors in my face shouting “women can’t do chemistry.” It was more… a thousand small reminders that you were not the default.

You get called “Miss” while your male colleagues are “Doctor.” At conferences, someone assumes you are a spouse rather than a speaker. You propose an idea in a meeting and it falls flat, then a man says nearly the same thing ten minutes later and suddenly it’s “excellent.”

How did I navigate it? By being stubborn, mostly. And by finding allies where I could – men and women who cared more about the quality of the science than about old‑fashioned notions of who belongs in a lab coat.

I also, eventually, reached a point where private frustration wasn’t enough, which is why WISP mattered to me. It is one thing to endure disrespect personally; it is another to watch a generation of students quietly peel away from science because they do not see a place for themselves. Institution‑building is slower than publishing a paper, but its half‑life is longer.

If you could correct one common misconception about your life or work, what would it be?

That I was careless.

There are versions of the story that read almost like a ghost story: “She handled a deadly chemical with ordinary gloves; how foolish!” That framing obscures the central issue: the gloves were recommended. The protocols were standard. If a tragic outcome is only avoided by superhuman vigilance, the system is brittle.

I would like the record to show that the real error was an institutional one: failing to test protective gear properly, failing to design protocols that accounted for the worst plausible case, and tolerating a culture in which convenience and habit often outweighed quantitative risk assessment.

If another chemist sees my name and thinks, “I should press my institution to validate our PPE against the hazards we actually use,” that would be a far more accurate tribute than whispering that I “should have known better.”

With the benefit of hindsight, is there any scientific or professional judgment you now look back on and say, “I would do that differently”?

Oh, certainly. I was not a saint; I was a working scientist with limited time and a tendency to overcommit.

Scientifically, I sometimes overestimated what we could infer from simplified model systems. You take a short peptide that mimics a protein binding site, study its cadmium complex in a clean buffer, and then you are tempted to draw strong conclusions about behaviour inside a congested cell nucleus. I tried to be cautious, but if I could advise my younger self, I would say, “Partner earlier with cell biologists; don’t wait until the paper is nearly done.”

Professionally, there were students I did not reach as well as I wish I had. A few I pushed too hard, reading their quietness as lack of interest rather than anxiety. I learned, over time, to ask, “What do you need from me?” rather than assuming one mentoring style fits all. If there is a regret, it is that I learned that lesson later than I might have.

And, more grimly, if I had been more sceptical about “standard practice” in PPE – if I had asked one more annoying question about permeation testing or insisted on laminated gloves earlier – that might have changed my own story. I did not ask that question loudly enough. The fact that the whole community had not either does not erase my personal wish that I had.

Was there ever a moment in your career where humour helped you cut through the tension?

Plenty. Chemistry needs humour; otherwise we’d all just argue about spectral assignments until we fossilised.

Once, in a grant review panel, a colleague referred to heavy metals as “unfashionable,” saying everyone exciting was working on genetics or signal transduction. I told him, quite deadpan, “Toxic metals are the ultimate in durable fashion – they never go out of style and they stay in your bones forever.” The room laughed, and we had a more honest conversation afterwards about environmental justice and why “unfashionable” problems still deserve funding.

If you can make someone laugh without trivialising the risk, you sometimes open a door that a lecture would leave shut.

In 2025, academic labs around the world still wrestle with safety culture – balancing pressure to publish with the need for robust protections. If you could address today’s principal investigators directly, what would you emphasise?

I would say this: safety is not an obstacle to good science; it is part of what good science looks like.

If your students are afraid to admit mistakes because they worry about your temper, they will hide near‑misses that ought to be learning opportunities. If budgets cut into PPE or engineering controls, you are externalising costs onto human bodies. If your protocols for a new nanomaterial or an exotic organometallic are “We’ll treat it like acetone and see what happens,” you are gambling with partial information.

Ask hard questions.

• Has anyone measured permeation of this compound through these gloves?
• Has anyone checked what happens when this reagent encounters moisture in a real hood, not an idealised one?
• Are we assuming that because we haven’t seen an accident, the system is safe, or do we have data?

And remember that students watch not only what you say but what you do. If the professor walks into the hood without eye protection “just for a second,” the message is clear.

My story is often told as a horror tale about a particular chemical. It should also be seen as an invitation to treat incomplete knowledge as a hazard in its own right.

Since your death, specific changes have become standard: OSHA recommendations on using SilverShield laminate under abrasion‑resistant gloves for dimethylmercury, strong discouragement of dimethylmercury for any routine purpose, adoption of safer mercury NMR standards. Most researchers using those today don’t know your name. How do you feel about that kind of invisible legacy?

It would be lying to say it doesn’t sting at all. Everyone likes to be remembered for their work, rather than as an anonymous footnote.

But the more important question is: are people safer?

If a young researcher today never even considers dimethylmercury for routine NMR because there are safer alternatives, that is progress. If glove catalogues specify breakthrough times for nasty organometallics because regulators now demand that data, that is progress. If OSHA guidance includes specific recommendations rather than generic “wear gloves,” that is progress.

The irony is that when safety is working, it disappears. People speak of “common sense,” as if it descended from the sky, rather than being written in blood over decades. Part of what you are doing, in having this conversation, is to stitch the human stories back into the fabric of those “obvious” precautions. I think that matters. Not for my ego, but because it reminds us that behind every line in a safety manual, someone, somewhere, paid tuition.

During your career, there were debates – sometimes quite sharp – about how to interpret metal toxicity. Some argued that focusing on specific metal–protein complexes was too narrow, that broader oxidative stress models were more relevant. How did you respond to those critiques?

Fair criticism keeps a field honest.

Oxidative stress is certainly part of the picture for many metals. Chromium, for instance, goes through redox cycles that generate reactive oxygen species. Focusing solely on a single protein target would be myopic.

My argument was never that our structural work captured the whole story. Rather, I argued that without molecular detail, “oxidative stress” is a bit of a catch‑all. It tells you that damage occurs, but not where the leverage points are.

Knowing that cadmium preferentially binds to certain DNA repair proteins does not contradict oxidative mechanisms; it refines them. Perhaps the cell can tolerate a certain level of reactive oxygen species if repair systems are intact, but once cadmium has hobbled those systems, the same oxidative burden becomes far more carcinogenic.

So when colleagues said, “You’re missing the forest for the trees,” my response was, “Forests are made of trees. Let’s at least know what they are.”

Turning to the next generation: what do you want to say to a young scientist in 2025 – perhaps a woman, perhaps someone from another marginalised group – who loves the work but feels the weight of exclusion and risk?

First: your curiosity is not an indulgence; it’s a contribution. The questions you ask are shaped by who you are, and science needs that diversity of questions.

Second: find your people. No one gets through this alone. Look for mentors who respect you as a colleague in training, not as a quota. Look for peers who celebrate your successes without making them seem like threats.

Third: do not accept “that’s how it’s always been” as the end of any conversation about safety or equity. Tradition is not a credential. Ask, “What evidence do we have that this is safe? What evidence do we have that this policy is fair?” Those are scientific questions as much as social ones.

And finally, be gentle with yourself. You will make mistakes; everyone does. The goal is not never to slip, but to work in communities and systems that catch each other, learn, and improve. The work is long. You do not have to carry all of it alone.

For readers working far from chemistry – say in software, medicine, policy, climate – what should they carry away from your story in terms of innovation and systems thinking?

I would summarise it like this:

• Expertise is necessary but not sufficient. You can be the world’s leading authority on a hazard and still be vulnerable if the protective systems around you are built on untested assumptions.
• “Standard practice” is not a guarantee of safety. If a process is widely used but never stress‑tested against worst‑case scenarios, you should treat it as a hypothesis, not a fact.
• Infrastructure is often invisible when it works. Whether it is a glove standard, a code review process, or a hospital protocol, its success makes it fade into the background. Don’t let that invisibility trick you into forgetting it was built – and can be improved.
• Progress often comes on the heels of preventable tragedy. The ethical response is not fatalism, but a commitment to move upstream: to ask harder questions earlier, so fewer lives become “proof of concept.”

Innovation is not just about new tools; it is also about designing systems where failure is less catastrophic, and where the people doing the work are not treated as expendable.

Standing in this imagined 2025, seeing SilverShield gloves in catalogues, safer NMR standards in use, WISP’s descendants in programmes around the world, and awards in your name supporting young scientists – how do you hold the mix of loss and legacy?

It is bittersweet.

I would, of course, rather have lived to see my students’ careers unfold, to write that hypothetical “magnum opus,” to grow old bickering good‑naturedly about coordination chemistry at conferences. No amount of legacy fully compensates for a life truncated at forty‑eight.

But if my death forced a reckoning that has kept others alive – if a postdoc in 2005 went home safely because their lab used proper gloves; if a young woman in 2015 stayed in physics because a WISP‑like programme told her she belonged; if a regulatory agency in 2020 pointed to my case and said, “We need evidence before we approve this exposure” – then something meaningful has been salvaged.

What I would ask, gently, is this: remember that the safety and inclusion you enjoy today were not inevitable. They were built, sometimes painfully. Honour that by continuing the work – by questioning standards, by widening the circle of who gets to do science, and by refusing to accept “we’ve always done it this way” as the last word.

Karen, thank you – for the science, for the courage, and for this conversation across years.

And thank you. Now, please, go check your gloves.

---

Letters and emails

Following the interview, we received an extraordinary volume of correspondence from researchers, students, policymakers, and science communicators around the world – many eager to extend the conversation in directions the original discussion had only touched. We’ve selected five letters and emails from our growing community, each offering a window into how different people, in different places, are thinking about Karen’s work, her legacy, and what her story means for those pursuing science today.

These questions come from environmental health researchers in East Africa, analytical chemists in Seoul, bioethicists in France, science communicators in Australia, and occupational safety officers in Canada. Together, they reflect the truly international resonance of her story – and the hunger to understand not just what happened, but what it means for the future.

Wanjiku Mwangi, 34, Environmental Health Researcher, Nairobi, Kenya
Your work focused heavily on mercury and cadmium, but many communities in the Global South face chronic exposure to artisanal gold mining runoff, lead from informal recycling, and chromium from tanneries – often multiple metals at once. Did your research ever address how toxic metals interact when they’re present together in biological systems? I’m curious whether the molecular mechanisms you studied for single metals hold up when the body is dealing with a cocktail of pollutants, or whether co-exposure creates entirely different binding competitions and repair failures that we still don’t understand.

Wanjiku, this is exactly the sort of question that keeps a toxicologist up at night – and I mean that as a compliment. You’ve put your finger on one of the genuine frontiers of the field, and I wish I could tell you we had solved it by the mid-1990s. We had not.

The honest answer is that most of my laboratory work, and indeed most of the bioinorganic chemistry of that era, examined one metal at a time. There were practical reasons for this. When you’re trying to characterise the binding affinity of, say, cadmium for a zinc finger protein, you want a clean system: defined buffer, controlled pH, one metal, one ligand, measurable signal. Add a second metal and suddenly you’re dealing with competition kinetics, displacement equilibria, potential ternary complexes – the experimental design becomes considerably more demanding, and interpretation of the data becomes, frankly, a headache.

That said, we were not entirely naive about co-exposure. In the environmental health literature – work coming out of groups studying Minamata disease, or occupational cohorts in smelting and battery industries – there was growing awareness that real-world exposure is almost never to a single pure toxicant. Workers in a cadmium processing plant are also breathing in lead dust and perhaps arsenic. Communities downstream of mining operations face mercury alongside other heavy metals leaching from tailings.

What we understood conceptually, even then, was that metals can interact at multiple levels.

First, there is competition for the same binding sites. Many toxic metals – cadmium, mercury, lead – are “soft” in the Lewis acid sense and have strong affinities for thiol groups, the sulphur-containing amino acids like cysteine. If cadmium and mercury are both present, they may compete for the same cysteine residues on a protein. Depending on their relative concentrations and affinities, one might dominate, or you might get a mixed population of metal-bound species. The functional consequences could be quite different from either metal alone.

Second, there is the question of whether damage is additive or synergistic. If cadmium inhibits one DNA repair pathway and mercury inhibits another, the combined effect might be worse than the sum of the parts – because now the cell has fewer backup options. We suspected this was happening, but proving it rigorously required more sophisticated models than we typically used.

Third, some metals can actually influence the absorption, distribution, or excretion of others. Selenium, for instance, has a complicated relationship with mercury; there is evidence that selenium can be protective, perhaps by sequestering mercury into inert complexes. Iron status affects cadmium uptake in the gut. So the body’s handling of one metal is not independent of its handling of others.

In my own cadmium work, we occasionally ran experiments with mixtures, but I would not claim we did this in any comprehensive way. The tools were limiting. ¹⁹⁹Hg NMR tells you about mercury; it does not simultaneously report on cadmium. You could do separate experiments with ¹¹³Cd NMR, but integrating the two into a coherent picture of a mixed-metal system was labour-intensive. Mass spectrometry was improving rapidly by the mid-1990s, and I think that technique – especially what you now call metallomics – offers the best hope for tracking multiple metals in biological samples simultaneously.

To your specific point about artisanal gold mining and informal recycling: these are precisely the exposure scenarios where single-metal models fall short. A child in a community affected by gold mining runoff might have elevated mercury from amalgamation processes, lead from battery recycling, and perhaps arsenic from contaminated groundwater, all at once. The molecular biology of that child’s cells is not neatly partitioned. The metals are interacting in ways we still do not fully understand.

I suspect – and this is speculation grounded in what we knew – that co-exposures create binding competitions that shift the landscape of vulnerable targets. Perhaps mercury outcompetes cadmium for certain high-affinity sites, but then cadmium “spills over” onto lower-affinity targets it would not normally occupy. Perhaps one metal sensitises a repair pathway that another metal then disables. The combinatorics are daunting.

If I were advising a young researcher in your position, I would say: do not feel obligated to solve the whole problem at once. Pick a well-defined system – two metals, one protein family, clear endpoints – and characterise it thoroughly. Build up from there. The environmental reality is messy, but the science has to proceed in tractable steps.

And please, keep asking these questions. The communities you are describing – in Kenya, in Nigeria, in Indonesia, in Peru – deserve answers that are grounded in their actual exposures, not in the convenient fictions of single-metal laboratory models. That work is unglamorous and it is difficult, but it is desperately needed.

I only wish I had lived long enough to contribute more to it myself.

Min-Jun Park, 29, Analytical Chemistry PhD Candidate, Seoul, South Korea
In the interview you mentioned using low temperatures to slow exchange on the NMR timescale and resolve multiple bound mercury species. I work with modern cryoprobes and ultra-high-field instruments that would have been unimaginable in the 1990s, yet I still find myself wrestling with similar line-broadening problems when studying metal–protein dynamics. Looking back, were there specific technical limitations of your era – field strength, probe sensitivity, data processing – that you felt most constrained your ability to extract mechanistic detail from ¹⁹⁹Hg spectra? And do you think today’s instrumentation would have changed which scientific questions you prioritised, or would the fundamental bottleneck still be sample preparation and the biology itself?

Min-Jun, you’ve asked the question that every spectroscopist of my generation muttered into their coffee at least once a week. Yes, the instrumentation constrained us – sometimes severely – but the relationship between hardware limitations and scientific questions is more interesting than a simple “if only we’d had better magnets” story.

Let me walk you through what we were actually working with.

In the late 1980s and early 1990s, a “high-field” instrument for most chemistry departments meant 300 or 400 MHz for protons, which translates to 7 to 9.4 Tesla. A few well-funded places had 500 MHz machines, and 600 MHz was exotic. For ¹⁹⁹Hg, which has a gyromagnetic ratio about a quarter that of protons, you’re operating at correspondingly lower frequencies – somewhere in the 50 to 90 MHz range on those instruments. The sensitivity penalty is real.

The practical consequence was that we needed millimolar concentrations of mercury-containing samples to get decent signal-to-noise in a reasonable acquisition time. “Reasonable” often meant overnight runs, sometimes over a weekend. You would set up the experiment on Friday afternoon, pray the building didn’t have a power fluctuation, and come back Monday morning hoping for something interpretable.

Now, millimolar concentrations are fine for small-molecule reference compounds or for stable, well-behaved metal complexes. But biological systems do not always cooperate. Proteins aggregate at high concentration. Metal-binding equilibria shift when you push concentrations far above physiological relevance. You end up asking: am I seeing the chemistry that matters, or am I seeing an artefact of the conditions I needed to get a spectrum at all?

Line broadening was, as you note, a persistent headache. ¹⁹⁹Hg has a large chemical shift anisotropy, which means that in anything other than a small, rapidly tumbling molecule, you get broad lines. A mercury centre bound to a 30-kilodalton protein is not tumbling rapidly. The resonance smears out, sometimes to the point of being nearly undetectable. We used a few tricks: low temperatures to slow exchange processes and sharpen lines from distinct species; small model compounds or peptide fragments that mimicked protein binding sites but tumbled faster; indirect detection methods when direct observation failed.

The low-temperature work you mentioned was genuinely useful, but it introduced its own complications. Aqueous samples freeze, obviously, so you’re often working in mixed solvent systems or accepting that you’re no longer at physiological conditions. Viscosity increases, which helps with tumbling correlation times but can also slow down the very exchange processes you’re trying to observe. It’s a balancing act.

Data processing was another bottleneck, though perhaps a less glamorous one to discuss. We did not have the computational power you take for granted. Fourier transforms were not instantaneous. Sophisticated line-shape simulations required mainframe time or, later, workstations that cost more than a postdoc’s annual salary. I remember waiting hours for calculations that your laptop probably does in seconds. This meant you were more conservative about which experiments you attempted, because the downstream analysis was itself a significant investment.

Would modern instrumentation have changed the questions I prioritised? That’s the crux of your question, and I think the answer is: somewhat, but not as much as you might expect.

Better sensitivity would have let us work at lower, more physiologically relevant concentrations. Cryoprobes, higher fields, improved electronics – all of that translates into faster experiments with smaller samples. That’s not trivial. It would have opened up systems we simply could not tackle.

But the fundamental bottleneck, as you astutely suggested, was often the biology. Getting a mercury-protein complex that is homogeneous, stable, and well-characterised is hard regardless of your magnet. Knowing which protein to study, which domain, which mutant – that requires biochemical and molecular biological work that proceeds on its own timeline. A beautiful NMR spectrum of the wrong construct tells you nothing useful.

I also think that having more data, faster, can be a mixed blessing. There is a discipline imposed by scarcity. When every experiment costs you a weekend, you think very carefully about what you’re asking and why. You read the literature more thoroughly, you plan more deliberately. I’ve heard from colleagues – well, heard in this strange posthumous way – that modern students sometimes generate data faster than they can interpret it. That’s a different kind of problem, but a real one.

If I could have reached into the future and borrowed a 900 MHz spectrometer with a cryoprobe for a month, what would I have done? Honestly, I think I would have spent it on the protein systems that defeated us: direct observation of mercury binding to intact DNA repair enzymes, real-time monitoring of metal exchange at physiological concentrations, detection of minor species that we could only infer from indirect evidence. The questions were always there. The access was what we lacked.

One more thought, since you’re clearly someone who appreciates instrumental nuance: don’t neglect the complementary techniques. In my era, X-ray absorption spectroscopy – EXAFS, XANES – was becoming increasingly accessible through synchrotron sources, and it provided structural information that NMR alone could not. Mass spectrometry was improving rapidly. Fluorescence methods for tracking metal localization in cells were primitive compared to what exists now, but they were developing. The best work combined multiple approaches, triangulating on the answer from different directions.

¹⁹⁹Hg NMR remains a beautiful technique, and I suspect it still has contributions to make. But it is one tool among many, and knowing when to put it down and pick up something else is part of the craft.

Good luck with your work, Min-Jun. I hope your cryoprobe gives you spectra that would have made my entire month.

Camille Laurent, 41, Bioethicist and Science Policy Advisor, Lyon, France
You’ve spoken about safety as infrastructure built on tragedies, and how regulatory agencies often need deaths to justify “inconvenient” precautions. This tension between precaution and progress appears constantly in my work on emerging biotechnologies – gene drives, synthetic biology, AI in drug discovery. From your perspective as someone whose death literally rewrote policy, do you think science can ever truly adopt a precautionary culture before catastrophe, or is there something structural about research institutions – the pressure to publish, the fetishisation of novelty – that makes us incapable of taking invisible risks seriously until they become visible through someone’s suffering?

Camille, you’ve asked me the question I have turned over in my mind more times than I can count – especially given how much time I’ve had, in this strange imagined afterlife, to reflect on what my death meant and what it failed to change.

I wish I could offer you optimism. I wish I could say that science, being rational, naturally gravitates toward precaution when evidence of risk exists. But I lived through enough to know that isn’t quite true, and dying from a failure of precaution has only sharpened my view.

Let me start with what I observed in my own field.

Dimethylmercury was known to be extraordinarily toxic long before August 1996. The compound had killed before – there were case reports from the 1860s, and more recently, laboratory accidents that resulted in fatalities. The mechanism of toxicity, while not perfectly understood, was understood well enough: a small, lipophilic organomercury that crosses biological membranes with terrifying efficiency and accumulates in the central nervous system. None of this was secret. It was in the literature.

And yet dimethylmercury remained a “standard” reference compound for ¹⁹⁹Hg NMR. Why? Because it was convenient. It gave a sharp signal at a well-defined position. It was commercially available. It was what people had always used. The calculus, such as it was, went something like: “We know it’s dangerous, but we’re careful, and besides, the quantities are small.”

That reasoning – “we’re careful” – is the enemy of genuine precaution. It substitutes individual vigilance for structural protection. It assumes that accidents happen to careless people, not to experts following protocols. And it ignores the asymmetry between convenience and catastrophe: the convenience accrues a little bit every day, to everyone using the compound, while the catastrophe lands all at once on a single unlucky person.

So why didn’t anyone insist on testing glove permeation before I was exposed? Why didn’t anyone demand safer alternatives be developed? The answer, I think, is that the cost of inaction was invisible and distributed across a hypothetical future, while the cost of action – finding new reference standards, validating new protocols, potentially slowing down research – was concrete and immediate. Institutions are not good at weighing invisible future harms against visible present inconveniences.

You asked whether there is something structural about research institutions that makes us incapable of taking invisible risks seriously. I believe there is, and I’ll name a few components.

First, the incentive structure. Academic science rewards novelty, publication, priority. Nobody gets tenure for writing a paper titled “We Tested Our Gloves and They Seem Fine.” Safety work is unglamorous. It doesn’t advance your career. It doesn’t get you invited to give the keynote at the Gordon Conference. At best, it’s maintenance; at worst, it’s seen as an impediment to “real” research. So it gets neglected, delegated to committees, reduced to checkbox compliance rather than genuine inquiry.

Second, the diffusion of responsibility. In a university setting, who exactly is responsible for ensuring that PPE is appropriate for the hazards present? The principal investigator? The department safety officer? The environmental health and safety office? OSHA? The glove manufacturer? When responsibility is spread thin, it often evaporates entirely. Everyone assumes someone else has checked. Nobody has.

Third, the normalisation of risk. Researchers become habituated to hazards. You work with dangerous materials every day, and nothing bad happens, so the danger starts to feel theoretical. The first time you handle dimethylmercury, you’re nervous. The fiftieth time, you’re efficient. Efficiency can shade into complacency without anyone noticing.

Fourth – and this is the hardest one – the fetishisation of progress you mentioned. Science is culturally committed to the idea that knowledge is good, that discovery is virtuous, that moving forward is better than standing still. Precaution, by its nature, is about slowing down, about saying “wait, let’s think about this, let’s test this, let’s consider what could go wrong.” That posture feels antithetical to the heroic narrative of the scientist pushing boundaries. Caution is coded as timidity.

Can this change? I am less cynical than my diagnosis might suggest.

I think it can change in specific domains where catastrophe has already occurred and memory is kept alive. After my death, there were genuine reforms: glove standards, reference material recommendations, heightened awareness around organomercury compounds. Those reforms happened because a specific, vivid, well-documented tragedy forced attention. The question is whether we can generalise that attention to hazards that have not yet produced a body.

The precautionary principle you mentioned is one attempt to do this – to shift the burden of proof so that potentially hazardous materials must be shown safe before widespread use, rather than assumed safe until proven deadly. I support this in principle. In practice, it runs into the problem of defining “potentially hazardous” and “safe enough.” Everything is hazardous at some dose, under some conditions. Pure precaution, applied rigidly, would paralyse research entirely. The challenge is calibration: how much evidence of risk justifies how much constraint on practice?

I don’t have a formula for that calibration. What I have is a conviction that the default settings are wrong. Right now, convenience wins unless danger is proven beyond doubt. That default killed me. It has killed others. It will kill more people unless we consciously, effortfully, shift the burden.

What might help? A few thoughts.

Transparency about near-misses. Most accidents are preceded by incidents that almost happened. If researchers reported and analysed near-misses without shame or penalty, we would have much better information about where the system is fragile. But near-miss reporting requires a culture where admitting “I almost made a terrible mistake” does not damage your career. That culture does not yet exist in most places.

Dedicated resources for safety research. Not compliance, not paperwork – actual empirical investigation of hazards. What are the permeation rates of this new solvent through these gloves? What are the degradation products of this reagent under realistic storage conditions? This work should be funded, published, and rewarded, not treated as someone else’s problem.

Institutional memory. Tragedies fade. The people who lived through them retire or die. The lessons get compressed into rules that lose their context. Someone should be responsible for keeping the stories alive – not as horror entertainment, but as grounded reminders that the rules exist because someone paid tuition.

And finally, humility. The recognition that expertise does not confer immunity. I knew more about mercury toxicity than almost anyone in that building, and I am dead from mercury poisoning. That should terrify every expert who believes their knowledge protects them. It doesn’t. Systems protect you, or they fail to. Knowledge tells you what happened afterwards.

Camille, you work in bioethics and science policy. You are fighting the right fight. I do not think you will win it permanently – the pressures toward speed and convenience are too strong, and institutional memory is too weak. But you can win it locally, temporarily, in specific domains where you can build coalitions and hold attention. That is not nothing. That might be everything.

I wish I could be more encouraging. What I can offer instead is solidarity: I understand the frustration, I share the anger, and I believe the work is worth doing even when progress is slow and setbacks are frequent.

Keep pushing. Someone has to.

Liam Tremblay, 38, Occupational Health and Safety Officer, Montréal, Canada
Imagine you had survived – perhaps through some fluke of metabolism or earlier detection – and returned to research in the early 2000s. Given what your accident revealed about the inadequacy of existing safety testing, do you think you would have shifted your career toward becoming an advocate for prospective PPE validation, or would you have returned to metal–protein biochemistry and left the policy work to others? I’m asking because I wonder whether the scientific community needed your death specifically, or whether a near-miss survivor with your expertise and credibility might have driven the same reforms while continuing to produce research. What do you think it takes for a living scientist to command the same moral authority that a tragedy creates?

Liam, this is a genuinely difficult question, and I appreciate that you’re asking it seriously rather than as a rhetorical flourish. It gets at something I’ve wrestled with a lot – the strange calculus of what gives a voice authority, and whether a living survivor could have accomplished what my death did.

Let me be direct first: I don’t know. And I’m suspicious of anyone who claims certainty about counterfactuals.

But I can reason through it, and I’ll try to be honest about what that reasoning suggests.

If I had survived – if we’d caught the exposure early enough that chelation therapy actually halted the damage, or if some combination of luck and aggressive medical intervention had pulled me back from the brink – what would have been my immediate instinct? I think I would have wanted to return to research. Not immediately, perhaps. There would have been recovery, rehabilitation, psychological processing of having stared into the abyss. But fundamentally, I was a scientist. The lab was where I felt most myself, most capable.

And here’s where the counterfactual gets murky: would the field have listened to a near-miss survivor the way it listened to a body?

In theory, yes. A living Karen Wetterhahn who could say, “I nearly died, and here’s why your safety protocols are inadequate” would have credibility. I would have had the expertise to articulate precisely which glove formulations fail and why, which reference standards should be abandoned, what regulatory changes were necessary. I could have written papers, given talks, served on committees. I could have become an advocate with scientific legitimacy.

And yet.

There is a difference between “this happened to me and I survived” and “this happened and it killed someone.” The first is a warning; the second is a prophecy fulfilled. Tragedy has moral weight that survival, however narrow, does not quite possess. A living victim is still in some sense a beneficiary of luck. A dead one is an indictment of the system itself.

I say this without bitterness, or trying to. It’s an observation about how institutions respond to evidence.

Consider the history of occupational safety more broadly. How many workers had to die in coal mines before ventilation standards were mandated? How many women had to develop cervical cancer before the Pap smear was standardised? How many children had to be poisoned by lead paint before regulations tightened? The pattern is depressing: evidence of harm accumulates, survivors advocate, experts publish, and change happens slowly – until someone dies in a way that cannot be ignored or reclassified as coincidence. Then suddenly there is urgency.

A near-miss case, even one involving someone prominent in the field, can be reframed. “Well, she was uniquely sensitive.” “There were other contributing factors.” “Most people in her situation would have been fine.” The ambiguity provides escape routes for institutions that do not want to change. A clear fatality is harder to reinterpret.

That said, I don’t want to discount the impact a surviving advocate could have had. If I had spent the last decades of my career – and I would have had several decades, potentially – building the case for PPE validation, for organomercury phase-outs, for regulatory reform, I might have accomplished significant things. I had credentials, access, a platform. I could have been unusually effective precisely because I combined expertise with lived experience.

The question becomes: would that have been as much as what actually happened?

I suspect not, but I’m genuinely uncertain. My death created a moment – a crisis that demanded immediate response. OSHA moved with unusual speed. Institutions realised they were vulnerable. Media attention focused. There was narrative drama: a leading expert in toxic metals, felled by toxic metal, in a way that seemed almost like dark irony. That narrative drove action in a way that a cautionary tale from a survivor, however well-intentioned, might not have.

But here’s what troubles me about that conclusion: it suggests that the system only responds to tragedy, not to evidence. It suggests that a near-miss that was handled well – caught early, treated aggressively, the victim survives – would receive less institutional attention than a fatality. That seems backwards. It suggests we are not, in fact, optimising for safety. We are optimising for something else: perhaps the ability to deny risk until denial is no longer possible.

If I had survived and become a tireless advocate, would I have been heard? Probably. By whom? Other researchers in my field, certainly. By occupational safety professionals, by regulatory agencies, potentially. By the broad scientific community? Less certainly. There would have been a window of heightened attention, and then the slow drift back to normalcy as the crisis faded from public consciousness.

What my death did was create permanent institutional markers: the fellowship, the award, the regulations, the shifted practices. These are visible manifestations of change, physical traces of reckoning. A survivor’s advocacy, however brilliant, leaves fewer such traces. It changes some minds and some practices, but the field continues much as before, because the field has not been forced to reorganise around the lesson.

There’s another angle to consider, Liam. A living advocate – even one shaped by a near-death experience – might have been subject to pressures that a martyr is not. Colleagues might have encouraged me to move on, to not “dwell” on the accident, to “get back to real science.” There would have been subtle and not-so-subtle messages that dwelling on safety was becoming an obsession, that I should remember I was a researcher first. I might have had to choose between continuing my own research programme and becoming full-time in advocacy. Institutions are skilled at containing messengers who might otherwise disrupt business as usual.

My death solved that problem. It made me permanently unavailable for contradiction or accommodation. It made my advocacy speak from beyond the reach of professional marginalisation.

So what’s my honest answer to your question? I think a living survivor with my expertise and credibility could have driven significant reforms, but I doubt it would have been as swift or as sweeping as what my death catalysed. The reforms would have been more tentative, more compromised, more easily rolled back once attention shifted to the next crisis.

Is that tragic? Yes, absolutely. It means that the incentives of the system are misaligned – that death is a more effective argument for change than evidence, expertise, or close calls.

What would it take to change that? I think it would require a cultural shift in how science and industry value precaution. It would require building institutions where near-misses are taken as seriously as fatalities, where the goal is not to avoid disaster but to prevent near-disasters from ever occurring. That is a higher bar, and it demands more from us – more humility, more caution, more willingness to say “we don’t know, so we should be careful.”

I don’t see that shift happening naturally. It requires people like you – occupational health and safety professionals with the standing and will to push back against the routine rationalisation of risk. It requires funding for the thankless work of validation and testing. It requires a refusal to treat safety as an obstacle rather than a goal.

If I had lived and devoted myself to that work, I might have made a difference. But I suspect the difference would have been smaller than the one my death has made. And that asymmetry – that dying serves the cause of safety better than living – is perhaps the most damning indictment of how we have organised science.

So: a thought experiment with a sorrowful conclusion. A near-miss survivor could do good work. But the system that would have constrained and marginalised that work was broken long before my name became attached to it. My death didn’t fix the system; it simply made one of the system’s failures too visible to ignore.

That’s not a reason to be grateful for my death. It is, I think, a reason to be angry at a system that requires such costs.

Matilda Brown, 26, Science Communication Student, Melbourne, Australia
I’ve noticed that when your story is taught in safety courses, it’s almost always framed around the technical failure – latex gloves, permeation rates, the chemistry of dimethylmercury. But you were also a mother, a mentor, someone with relationships outside the lab. In the months between exposure and neurological collapse, when you didn’t yet know anything was wrong, what were you thinking about? Were there personal projects, family plans, or even mundane hopes for that year that got interrupted? I ask because I think humanising scientists matters for how we teach risk – not as abstract protocol violations, but as the derailing of actual lives.

Matilda, thank you for asking this. Most people don’t. They want the chemistry, or the policy, or the cautionary tale. They don’t want to know about the ordinary Tuesday afternoons, the things I was looking forward to, the small hopes that got interrupted.

I think that’s partly why I feel I should answer you honestly.

The timeline between 14th August and when neurological symptoms appeared in January – that’s roughly five months of what looked like a normal life. And it was, in a way. But underneath, something unknowable was happening. My body was being slowly rewritten by something I couldn’t feel or see.

In those first weeks after the spill, I did what I always did. I taught my courses. I had meetings with my research group. I attended faculty dinners. I was planning, actually, for a sabbatical the following year – I wanted to spend time at a collaborator’s lab in California, dive deeper into some work on mercury–DNA adducts that we’d been building toward. There were conversations about whether I could bring my family, what the logistics would look like. It felt real. It felt like it was going to happen.

My younger daughter was starting high school that fall. There were all the ordinary parental anxieties that come with that: was she happy, was she making friends, was she taking her classes seriously? I remember being frustrated with her about homework one evening – something typical, some sullenness about algebra – and we had words. I wish I could tell you I was gracious about it, but I wasn’t. I was tired, I had papers to grade, and I was impatient. That’s the kind of ordinary conflict that families have. It stayed with me, though, the way these things do. Not a major rupture, just the normal friction of living with people you love.

By November, I started noticing the weight loss. I thought it was stress, or perhaps I was just working too hard and not eating well. I’d been running more, trying to stay fit, so I chalked some of it up to that. But there was also this vague abdominal discomfort – nothing severe enough to necessitate seeing a doctor, but a persistent low-grade wrongness that I couldn’t quite place. I remember mentioning it to my husband one evening, and he suggested I see someone if it didn’t clear up.

I didn’t, actually. Or not immediately. There’s a thing that happens when you’re a scientist, when you’re trained to think in terms of data and mechanisms: you start to rationalise symptoms as noise rather than signal. One person has abdominal discomfort and weight loss for five months – is that a pattern or just the ordinary wear and tear of a body getting older? I was forty-eight. I had friends dealing with perimenopause, thyroid issues, all sorts of age-related changes. You don’t immediately jump to “I’ve been poisoned by a compound I handled briefly three months ago.”

December was strange. I felt unusually tired – the kind of fatigue that sleep doesn’t quite cure. I remember one evening in late November or early December, my family was watching television, and I fell asleep in the middle of a conversation with my husband. Just nodded off. That wasn’t like me. I mentioned it to a colleague in passing – something offhand, like “I must be getting old” – and they laughed and said everyone felt that way in December. The holidays are exhausting.

In early January, I saw my primary care physician. By that point, the weight loss had been noticeable enough that my clothes were fitting differently. I’d lost maybe fifteen pounds over several months – not catastrophic, but enough to register. The doctor ran some blood work, checked my thyroid, looked for signs of infection or anaemia. Everything came back normal. “Could be depression,” he suggested. “Have you been stressed?” I laughed. I’m always stressed. I have two kids in school, a career in research, a department to help run. Of course I’m stressed. But not depressed, I thought.

Then, in mid-January, I noticed I was slurring my words.

I remember the moment quite clearly. I was in my office, on the phone with someone – a colleague, I think – and I heard myself stumbling over words that should have been automatic. My tongue felt thick. My speech felt imprecise. I’m a careful person, someone who uses language precisely, and suddenly I couldn’t rely on my own mouth to produce what my brain was asking it to produce. That was frightening in a way the weight loss and fatigue hadn’t quite been.

And then the balance problems started. I’d reach for something on a shelf and misjudge the distance. My gait felt slightly off, as if the ground beneath my feet was not quite where I expected it to be. I didn’t fall – thank God, I didn’t fall – but I was aware of my body in a new way, aware that it was not entirely cooperating with my intentions.

That’s when I went back to the doctor and said, “Something is wrong with my nervous system.”

The blood tests came back with mercury levels that made everyone in the room go silent. The physician later told me he had never seen numbers that high in a living patient. And suddenly there was this horrible clarity: the spill in August, which I had not thought about in months, which I had essentially dismissed as a controlled event with minimal exposure, had apparently been an exposure event with maximal consequences.

What I want to tell you, Matilda, is what that gap feels like. That five months of unknowing. Because it wasn’t innocent, in retrospect. Every day during those months, something irreversible was happening in my brain, in my nervous system, in the architecture of who I was. But I was just living. I was grading papers and making dinner and worrying about my daughter’s homework and planning a sabbatical that would never happen.

There’s a particular kind of horror in that. Not the drama of a sudden emergency – that would have been almost easier, in a way. But this slow, invisible accumulation of damage while I was going about ordinary life, unaware. Making plans for a future that was already being erased by chemistry operating at the molecular level.

In those last weeks before the neurological symptoms became impossible to ignore, I had ordinary hopes. I was looking forward to a colleague’s birthday party. There was a paper we were about to submit that I was excited about. My husband and I had talked about maybe going to Maine in the summer – just a few days, nothing fancy, but it felt like something to anticipate.

I also had the anxieties that anyone has. Will my daughter be okay? Am I a good enough mother? Should I have published more, should I have said no to more committee work? Is my research making a difference or am I just churning out papers? Are we saving enough money for retirement? Will I still be able to run when I’m sixty? These were the ordinary, endless questions of a person in her late forties, trying to balance a lot of things imperfectly.

When I was diagnosed – when I understood what was happening – there was a period where I still had some agency. I could advocate for my own treatment. I could try to make sense of what had happened. I could ask questions, read the literature, hope that aggressive intervention might reverse some of the damage.

That period lasted, maybe, a few weeks. Then the deterioration accelerated, and I lost the ability to participate in my own recovery. I lost language. I lost the ability to walk. I lost the ability to regulate my body. What remained was increasingly internal, increasingly locked away from anyone else’s access.

One of the things nobody talks about is what that’s like from the inside. Not the clinical description – the paralysis, the neurological decline – but the subjective experience of losing yourself incrementally. It’s not like you turn off, like flipping a switch. It’s more like watching yourself fade, aware that you’re fading, unable to stop it.

My husband was there every day. My daughters came, though it was increasingly difficult for them. The medical staff was as kind as people trying to manage an impossible situation can be. But there was a particular loneliness in it, because I could no longer tell them what I was experiencing. Toward the end, I couldn’t speak. I couldn’t signal. I was conscious, or so the physicians believed, but I was inaccessible.

That’s the part of the story that gets left out of the safety protocols. The safety training focuses on the mechanical aspects – the glove failure, the permeation rate, the blood levels – and that’s important. But it leaves out the human cost in its fullest dimension: not just that I died, but what it was like to die this way. The consciousness without agency. The knowledge that my body was failing without the ability to negotiate with it.

What would I have hoped for in those five months after the spill? Honestly, exactly what I was doing: ordinary life, ordinary concerns, ordinary hopes. Nobody asks for a life cut short. But given that mine was, I’m grateful I didn’t spend those five months in terror. I’m grateful I didn’t know. The knowledge came late enough that I could still feel the sun, still read books, still sit at the dinner table with people I loved.

I’m grateful, in a strange way, for the mundane.

So when you ask about the ordinary hopes and personal projects – yes, they mattered. They still matter. Not as the headline of my story, but as the texture of what was taken away. Not as drama, but as the quiet fact that all the small things we assume we’ll have time for, we might not.

That’s what I’d want young scientists to understand, Matilda. Not just the chemistry or the policy, though those are important. But also: the person whose name ends up in a case study is a person who had plans. Who worried about her kids’ homework. Who was looking forward to a trip to Maine. Who had conversations with colleagues about future research. Who was, in every way that matters, ordinary.

And then the ordinary stopped.

That’s the tragedy that safety protocols are trying to prevent, even if the protocols themselves never mention it.

---

Reflection

Karen Elizabeth Wetterhahn died on 8th June 1997, at the age of forty-eight – ten months after three drops of dimethylmercury passed through her latex gloves in approximately fifteen seconds, and five months after neurological symptoms announced that the damage was already beyond repair.

This conversation has been an act of imagination grounded in fact: an attempt to hear the voice of a scientist who left behind eighty-five papers, a programme that transformed who does science at Dartmouth, and a legacy woven invisibly into the protective infrastructure of laboratories worldwide. What we have presented is not transcription but reconstruction – an effort to animate the intelligence, warmth, and rigour that colleagues remembered, filtered through the themes and questions that matter to readers in 2025.

Several threads have run through this dialogue that deserve reflection.

Expertise and vulnerability. Karen was not a novice handling a dangerous compound carelessly. She was among the world’s leading authorities on toxic metals, someone who understood at the molecular level how mercury destroys biological function. Her death demonstrated that knowledge alone cannot protect us when the systems surrounding that knowledge are inadequate. This is a lesson that extends far beyond chemistry: experts in any field remain vulnerable to failures in the infrastructure they depend upon.

Precaution and institutional inertia. Throughout her responses – particularly to Camille Laurent – Karen articulated a critique of how science weighs convenience against risk. The default, she argued, is that convenience wins until tragedy forces recalculation. Dimethylmercury was “standard” because it was effective, not because anyone had validated whether the standard precautions were sufficient. That calculus, she suggested, persists in debates over emerging biotechnologies, novel nanomaterials, and other frontiers where the hazards are not yet fully characterised.

The invisible architecture of safety. One of the most poignant themes was Karen’s reflection on what it means to be remembered primarily through protocols rather than papers. The SilverShield laminate gloves now specified for organomercury handling, the discouragement of dimethylmercury for routine NMR work, the requirement that PPE be validated against specific hazards – these are her monuments, even if most researchers using them have never encountered her name. Safety infrastructure, when it works, disappears. The challenge is to remember that it was built, often at terrible cost.

Women’s contributions and structural barriers. Karen navigated a scientific culture that did not always welcome women, and she responded not only by excelling personally but by building institutions – WISP – that changed who felt they belonged. Her observation that such institution-building is often undervalued compared with publication records speaks to a persistent inequity: the work that makes science more inclusive is frequently treated as secondary to the work that advances knowledge, even though both are essential.

The ordinary and the irreplaceable. Matilda Brown’s question, and Karen’s response, shifted the register from the technical to the personal. The five months between exposure and symptoms were filled with homework arguments, sabbatical planning, dinner conversations, hopes for a trip to Maine. The tragedy is not only that a scientist died, but that a mother, a colleague, a person with ordinary hopes was erased by a failure she could not have anticipated. Safety protocols exist to protect not abstractions but people – people who have plans, who worry about their children, who are looking forward to things they will never see.

Where might Karen’s perspective have differed from the historical record?

This is speculative territory, necessarily. The historical record tells us what happened; it cannot tell us what Karen thought or felt in moments that went undocumented. In constructing her voice, we have imagined a scientist who was reflective about the limits of single-metal models, self-critical about moments when she pushed students too hard or trusted existing protocols too readily, and frustrated by the way her story has been compressed into a cautionary tale.

Whether the real Karen Wetterhahn would have emphasised these particular themes, we cannot know. Colleagues described her as rigorous, generous, and deeply committed to mentorship – but the interior experience of her final months, the specific regrets or hopes she carried, remain private. We have tried to honour what is known while acknowledging that imagination has filled the gaps.

Gaps and contested interpretations. The precise details of the August 1996 exposure – how many drops, exactly where they fell, whether there were other contributing factors – are documented but not beyond all uncertainty. The timeline of her neurological decline is established in medical records, but her subjective experience during those months is largely inaccessible. Some accounts emphasise the horror of her case; others focus on the systemic failures it revealed. We have tried to hold both: the personal tragedy and the institutional critique, the loss of a single life and the protection of many that followed.

Connections to today’s challenges.

Karen’s scientific questions remain urgent. Heavy metal exposure continues to affect communities worldwide – artisanal gold mining in sub-Saharan Africa and South America, electronic waste recycling in South Asia, legacy contamination from industrial sites across the Global North. The molecular mechanisms she studied – metal displacement of essential ions, inhibition of DNA repair, carcinogenic mutagenesis – are still being elucidated, now with tools she could only have dreamed of: cryo-electron microscopy, advanced mass spectrometry, computational modelling of metal–protein dynamics.

The field of bioinorganic chemistry that she helped build has matured considerably. Researchers now routinely integrate structural, spectroscopic, and functional data to understand how metals behave in biological systems. The  Hg NMR technique she employed is less central than it once was – safer reference standards have been adopted, and complementary methods have expanded – but the fundamental questions she asked about mercury–protein interactions continue to drive research.

Her work on cadmium and DNA repair has been cited extensively in subsequent studies of metal carcinogenesis. The mechanisms she proposed – competitive binding, displacement of zinc from regulatory proteins, interference with repair pathways – have been refined and extended but not overturned. She was, in many ways, ahead of her field in recognising that understanding toxicity required molecular precision, not just epidemiological correlation.

The afterlife of her work.

Karen’s scientific contributions did not end with her death. Her papers continue to be cited; her students and collaborators carried forward research programmes she helped initiate. The Dartmouth Toxic Metals Superfund Research Program, with which she was affiliated, remains active, investigating metal exposures in communities and mechanisms of toxicity in model systems.

The Karen E. Wetterhahn Graduate Fellowship in Chemistry at Dartmouth, established in 1998, has supported generations of students. The National Institute of Environmental Health Sciences’ Karen Wetterhahn Memorial Award, given annually since 1999, recognises outstanding graduate students and postdoctoral researchers in environmental health sciences. These are not merely honorific; they are functional – providing resources and recognition that help early-career scientists, many of them women, establish themselves in fields where visibility matters.

In safety science, her case has become canonical. It is taught in chemistry courses, cited in occupational health literature, referenced in regulatory guidance. The post-accident research on glove permeation – conducted by her colleagues at Dartmouth in the immediate aftermath of her death – generated data that had never been collected before, precisely because no one had thought to test whether “standard” gloves were adequate for “standard” compounds. That research, born of tragedy, now protects researchers who will never know her name.

For young women in science.

Karen’s story offers no simple moral. She was brilliant, hardworking, and generous; she built institutions and published prolifically; she mentored students and advocated for equity; and she died at forty-eight from a compound she knew intimately, because the safety infrastructure failed her.

What can be drawn from this? Perhaps: that excellence is necessary but not sufficient. That individual effort must be matched by structural protection. That the work of making science inclusive – founding programmes like WISP, mentoring students who do not see themselves reflected in the faculty – is as important as the work of making discoveries, even if it is less celebrated. That visibility matters: when young women can see someone like Karen in a position of authority, doing serious science and building serious institutions, the implicit message that they do not belong loses some of its power.

And perhaps also this: that the risks women face in science are not only the familiar ones of bias and exclusion, but also the risks that come from trusting systems built without adequate care. Karen followed every protocol. The protocols were wrong. Advocating for better systems – for tested PPE, for validated procedures, for a culture where safety is taken as seriously as publication – is not a distraction from science. It is part of what science must be.

A final thought.

In the months after Karen’s death, her colleague and friend John Winn described visiting her in the hospital during her final weeks. Her husband saw tears rolling down her face, and Winn asked if she was in pain. The physicians said it did not appear that her brain could even register pain.

That image has stayed with many who have encountered her story: the tears without the capacity to feel, the body signalling something the mind could no longer process. It is an image of profound isolation – of a person locked inside a failing system, unable to communicate what, if anything, she was experiencing.

We cannot know what Karen felt in those final weeks. We cannot know whether she was aware, whether she suffered, whether she found any peace. What we can know is that her life, before those weeks, was full: of research and teaching, of family and friendship, of institution-building and advocacy, of ordinary hopes and ordinary frustrations. That fullness is what was taken.

Every researcher who handles a toxic compound today and goes home safely is, in some small way, protected by what Karen’s death revealed. The infrastructure is invisible; the debt is real.

If this conversation has accomplished anything, let it be this: that Karen Wetterhahn is remembered not only as a warning, but as a scientist, a mentor, a builder, and a person who deserved more time than she was given.

The least we can do is remember her name.

---

Editorial Note

This conversation with Karen Wetterhahn is a dramatised reconstruction, not a historical transcript. Karen Wetterhahn died on 8th June 1997, and cannot speak. What follows is an imaginative work grounded in documented facts about her life, her research, her death, and her legacy – but shaped through the interpretive lens of fiction.

What is factual: Karen Wetterhahn’s biographical details (birth year, education, career trajectory), her published research in toxic metal biochemistry and bioinorganic chemistry, the circumstances of her August 1996 exposure to dimethylmercury, the timeline and clinical progression of her poisoning, the institutional changes that followed her death (OSHA guidance, glove standards, reference material recommendations, the establishment of fellowships in her name), and the founding of the Women in Science Project at Dartmouth with the documented outcome (percentage of women science majors rising from approximately 13% to 25%).

What is reconstructed: Karen’s voice, her specific word choices, the tone and cadence of her speech, the particular ways she might have reflected on her work and legacy, her responses to hypothetical questions about counterfactual scenarios, her interior thoughts and feelings during the five-month period between exposure and symptom onset, and the philosophical frameworks she might have applied to contemporary questions about precaution, gender equity, and institutional safety culture.

The interpretive choices: In constructing Karen’s voice, we drew on interviews with her colleagues, published reminiscences, the scientific literature she authored, descriptions of her teaching and mentoring style, and the broader cultural and institutional context of 1990s academic chemistry. We aimed for authenticity while acknowledging that this is necessarily an interpretation, not a preservation.

Why this form: A traditional biography or academic paper would present established facts in historical voice. This reconstruction instead animates Karen as an intellectual presence – allowing her research, her values, and her critique of institutional failure to speak directly to contemporary readers. It is an act of imaginative scholarship, not a claim to documentary accuracy.

The responsibility: We have tried to be faithful to what is known about Karen Wetterhahn’s life, character, and work, while avoiding both sensationalisation of her death and saccharine memorialisation of her legacy. Where historical sources were ambiguous or silent, we have attempted to reason within the bounds of plausibility rather than invent details. Where speculation was necessary, we have marked it as such.

This conversation is offered in the spirit of keeping Karen Wetterhahn’s story – both her scientific contributions and the institutional failures that took her life – in active memory. It is not a substitute for reading her papers, researching the history of laboratory safety, or understanding the real women scientists working today in toxic metal biochemistry and related fields.

Karen Wetterhahn deserves to be remembered not only as a cautionary tale, but as a scientist whose work still matters.

---

Who have we missed?

This series is all about recovering the voices history left behind – and I’d love your help finding the next one. If there’s a woman in STEM you think deserves to be interviewed in this way – whether a forgotten inventor, unsung technician, or overlooked researcher – please share her story.

Email me at voxmeditantis@gmail.com or leave a comment below with your suggestion – even just a name is a great start. Let’s keep uncovering the women who shaped science and innovation, one conversation at a time.

---

Bob Lynn | © 2025 Vox Meditantis. All rights reserved.