The Looking-Glass Molecule: Henri Kagan, Kenso Soai and the 2026 Nobel Prize in Chemistry

Looking-glass milk

Near the start of Through the Looking-Glass, Alice holds up her kitten to the mirror and wonders aloud what life is like on the other side. "Perhaps Looking-glass milk isn't good to drink," she says. Lewis Carroll meant it as a whimsy, but he had touched, perhaps without knowing it, on one of the deepest strangenesses in chemistry. The molecules of milk, like the molecules of almost everything alive, have a handedness. Their mirror images would be chemically identical in every way we normally measure, with the same weight, the same boiling point and the same formula. Yet a body built from one hand would find the other hand foreign, and sometimes poisonous.

This is the puzzle behind the 2026 Nobel Prize in Chemistry, awarded on 7 October to Henri B. Kagan, now in his mid-nineties and a professor emeritus at Université Paris-Sud, and Kenso Soai, professor emeritus at Tokyo University of Science, "for the discovery of non-linear effects and autocatalysis in asymmetric organic synthesis." It is a forbidding phrase, the sort that makes people turn the page. But what it describes seems to me one of the most beautiful things chemistry has shown us: how a world that should be perfectly balanced between left and right can tip, and keep tipping, until one hand has won.

Pasteur's tweezers

The story is usually begun with Louis Pasteur, and rightly so. In 1848, as a young chemist of twenty-five, Pasteur was examining crystals of a salt of tartaric acid, a by-product of winemaking. Other chemists had reported a puzzle: one form of the acid rotated polarized light, while a second form, chemically indistinguishable, did not. Under his microscope Pasteur saw that the crystals of the inactive form were of two kinds, tiny mirror images of each other, like left and right gloves. With tweezers and great patience he sorted them into two piles and dissolved each. One solution turned polarized light to the right and the other to the left; mixed together, they cancelled.

Pasteur later said that chance favours only the prepared mind, and there was certainly chance in his discovery. But there was something more. He had grasped that molecules themselves must have a shape in space, a handedness, long before anyone could picture what a molecule looked like. A decade later he made a second, stranger observation. When microorganisms fed on tartaric acid, they consumed the right-handed form and left the left-handed form untouched. Life, it seemed, could tell its left from its right.

Chemists call such molecules chiral, a word Lord Kelvin coined in the 1890s from the Greek for "hand." The two mirror forms are enantiomers. And life, as it turned out, is almost perfectly one-handed, or homochiral. The amino acids in our proteins are all of the L form; the sugars in DNA and RNA are all D. A protein assembled from mixed hands would not fold; a double helix of mixed sugars would not wind.

This poses a question that has troubled chemists for well over a century. When we make a chiral molecule in the laboratory from ordinary, non-chiral ingredients, we always get both hands in exactly equal amounts, a racemic mixture. Nature, so far as anyone could tell, had no reason to prefer one hand over the other. How, then, did life end up with only one?

A physicist's thought experiment

In 1953 the Bristol physicist Frederick Charles Frank wrote a short paper about it. He had been told, apparently with some surprise on his part, that chemists still had no explanation for life's asymmetry. So he offered a mathematical one. Imagine, he wrote, "a chemical substance which is a catalyst for its own production and an anti-catalyst for the production of its optical antimer." A molecule, in other words, that makes more of itself and suppresses its mirror image.

Frank's model needs three things: a product that catalyzes its own formation (autocatalysis); a catalyst that favours its own hand (asymmetric catalysis); and some way for each hand to hold back the other (mutual inhibition). Given these, any small imbalance, however slight, would grow until one hand dominated completely. He ended his paper, with characteristic British understatement, by remarking that "a laboratory demonstration may not be impossible."

For decades it remained a sort of chemical riddle, the kind a lecturer poses to a seminar and nobody can answer. Asymmetric catalysis was achieved, slowly, and its pioneers were honoured with Nobel Prizes in 2001 and 2021. But mutual inhibition, the strangest of Frank's conditions, had never been seen in a catalytic reaction, and no one had combined all three.

Kagan's doubt

Henri Kagan, working in Orsay in the early 1980s, was one of many chemists trying to coax reactions into making one enantiomer rather than both. The usual approach was to use a catalyst built from a metal atom and a chiral organic molecule, a ligand, that lends its handedness to the reaction. Everyone assumed that the catalyst's purity passed straight through to the product. A ligand that was half pure would give a product that was half pure; the relation was a straight line. It was one of those assumptions so obvious that no one thought to test it.

Kagan's gift, it seems to me, was a kind of disciplined doubt. He asked what the catalyst actually looked like. A metal atom, he reasoned, often holds not one ligand but two. If the ligand were a mixture of left and right, the metal would gather them in three combinations: two left, two right, or one of each. The left–left and right–right catalysts are mirror images and should behave as mirror images. But the mixed, left–right catalyst is a different creature altogether, a diastereomer, with its own stability and its own speed. If it were sluggish, it would quietly sequester the minority hand, and the product would come out purer than the ligand that made it.

In 1986 Kagan and his colleagues published three reactions in the Journal of the American Chemical Society that broke the straight line. In two the curve sagged below it, and in one, the Sharpless epoxidation of geraniol, it bowed upward: a positive non-linear effect. The chiral information had not merely been transmitted; it had been amplified. The heterochiral catalyst, like a pair of mismatched dancers, had stepped off the floor and taken the minority with it.

What began as a curiosity became a tool. When Ryoji Noyori used a zinc catalyst with an amino-alcohol ligand of only 15% enantiomeric excess, he obtained a product of 98% ee. Process chemists at Merck reduced a ketone with a reagent made from pinene of 70% optical purity and got a product of 95% ee. And chemists everywhere learned to read the bend of the curve as a clue, a kind of clinical sign that tells them how many chiral partners are at work in the hidden machinery of a catalytic cycle.

Soai's reaction

Kenso Soai, born in Hiroshima in 1950, had been studying a particular class of reactions: the addition of zinc reagents to aldehydes, guided by chiral alcohols. These reactions were known for strong non-linear effects. Looking at the structures of his catalysts and his products, Soai noticed that they resembled each other, and he asked a question of startling simplicity. Could a product be made to catalyze its own formation?

In 1990 he showed that it could. With pyridine-3-carbaldehyde and diisopropylzinc, a product alcohol of 86% ee catalyzed the formation of more of itself. But each generation was less pure than the last (the new batch came out at 35% ee), like a photocopy of a photocopy. Autocatalysis alone, it turned out, was not enough; Frank had been right that mutual inhibition was also needed.

The decisive experiment came five years later, in Nature. Soai switched to a pyrimidine-5-carbaldehyde, and this time the copies grew sharper. A seed of 5% ee gave 55% after one run and 87% after the next, levelling off near 90%. For the first time, all three of Frank's conditions had been met in a flask.

The chemistry itself is not exotic. It is a carbonyl addition of the most familiar kind, the same arrows a student draws for a Grignard reaction, with zinc in place of magnesium. The electrons of the carbon–zinc bond reach across to the aldehyde carbon, and the carbon–oxygen π bond swings over to bind the zinc.

What the arrows cannot show, and what makes the reaction extraordinary, is which face of the flat aldehyde the isopropyl group attacks. That choice is made by the product itself. The zinc alkoxide that forms gathers, it now appears, into clusters of four, and these clusters guide the next addition toward their own hand.

In 2003 Soai's group pushed the experiment to an almost absurd limit. They seeded the reaction with alcohol of 0.00005% ee, a balance so fine that among two million molecules there is roughly one extra of one hand. After a single run the product was 57% ee; after the second, 99%; after the third, more than 99.5%. The Nobel Committee calculates the amplification at a factor of about 630,000.

Something from nothing

There remained one further, almost metaphysical, question. What would happen if one gave the reaction no chiral seed at all?

No real sample, however carefully prepared, is ever exactly balanced. Molecules are discrete, and chance ensures that one hand or the other will always be slightly ahead, if only by a handful among trillions. The chemist Kurt Mislow called this hidden imbalance cryptochirality. For any ordinary reaction it is irrelevant, lost in the noise. But the Soai reaction is no ordinary reaction.

In 2003 Soai's group ran it again and again from purely achiral starting materials, with nothing chiral added. The products were never racemic. In thirty-seven experiments, eighteen came out predominantly in one hand and nineteen in the other, with enantiomeric excesses ranging from 15% to 91%. Daniel Singleton and L. K. Vo at Texas A&M had independently observed the same thing a year before. The outcome was a coin toss, but every toss came up decisively. Chance proposed a winner, and the chemistry amplified it until there was no doubt. (The 🎲 button in the model above plays the same game.)

I find this one of the most moving results in modern chemistry. It is not, as the Nobel Committee is careful to say, the way life actually became one-handed; the Soai reaction depends on zinc compounds that could never survive in water, and the true origin of biological homochirality remains a mystery that may never be settled. But it shows, in a single flask, that such a thing can happen, that a world of perfect symmetry can break its own balance and choose. The reaction is so sensitive, indeed, that it has been used as a kind of chemical seismograph, registering the faint chiral influences of circularly polarized light, of mineral crystals, even of molecules whose only asymmetry lies in the placement of a heavier isotope.

A thirty-year puzzle

How the Soai reaction works has proved remarkably difficult to discover, and the story of that search has its own fascination. Donna Blackmond and John Brown measured the heat the reaction gives off and argued for a catalyst built of two zinc alkoxides; later kinetic modelling suggested four. Calculations proposed a strange "square-macrocycle-square," and in 2015 Soai's colleagues crystallized the zinc alkoxides and found just such tetramers: two zinc–oxygen squares linked through the pyrimidine nitrogens into a twelve-membered ring. In 2020 Scott Denmark and his collaborators proposed that the aldehyde nestles "floor-to-floor" against two zincs of a homochiral tetramer, a fit the mismatched tetramer cannot offer. At nearly the same time Oliver Trapp proposed a different picture altogether, built on fleeting zinc hemiacetals. Both may be right, each for its own substrate. A molecule that can choose a hand from nothing has kept its own secrets for thirty years.

Coda

When the prize was announced, Soai was reached by telephone. "There are many, many, many excellent researchers in this field," he said. "So I'm very especially glad to receive this prize." Erick Carreira, a chemist at ETH Zürich, put it more simply: the implications for the origins of chirality on earth are "amazing. It just blows your mind."

There is a pleasing symmetry, if one may use the word, in a prize that honours two discoveries about broken symmetry. Kagan's was a discovery of doubt: that an obvious straight line was not straight at all. Soai's was a discovery of persistence: that a chemical copy could grow sharper, not fainter, with each generation, until a whisper became a certainty. Between them they have shown that the asymmetry of life need not be a miracle, only chemistry patient enough to remember which hand came first.

Alice, I suspect, would have been delighted. Looking-glass milk may not be good to drink. But we now understand a little better why the milk on our side of the mirror is the way it is.

Sources

1. The Royal Swedish Academy of Sciences. Press release: The Nobel Prize in Chemistry 2026. 7 October 2026. nobelprize.org
2. The Royal Swedish Academy of Sciences. Popular science background: They solved chemistry's asymmetric mystery. 2026. nobelprize.org
3. The Nobel Committee for Chemistry. Scientific background: Non-linear effects and autocatalysis in asymmetric organic synthesis. 7 October 2026. nobelprize.org
4. Puchot, C.; Samuel, O.; Duñach, E.; Zhao, S.; Agami, C.; Kagan, H. B. "Nonlinear effects in asymmetric synthesis. Examples in asymmetric oxidations and aldolization reactions." J. Am. Chem. Soc. 1986, 108, 2353–2357. doi.org
5. Soai, K.; Shibata, T.; Morioka, H.; Choji, K. "Asymmetric autocatalysis and amplification of enantiomeric excess of a chiral molecule." Nature 1995, 378, 767–768.
6. Sato, I.; Urabe, H.; Ishiguro, S.; Shibata, T.; Soai, K. Angew. Chem. Int. Ed. 2003, 42, 315–317 (amplification from 0.00005% ee).
7. Soai, K. et al. Tetrahedron: Asymmetry 2003, 14, 185–188 (absolute asymmetric synthesis; 37 experiments).
8. Singleton, D. A.; Vo, L. K. J. Am. Chem. Soc. 2002, 124, 10010–10011; Org. Lett. 2003, 5, 4337–4339.
9. Frank, F. C. "On spontaneous asymmetric synthesis." Biochimica et Biophysica Acta 1953, 11, 459–463.
10. Athavale, S. V.; Simon, A.; Houk, K. N.; Denmark, S. E. Nature Chemistry 2020, 12, 412–423.
11. Barbu, B. "2026 Chemistry Nobel awarded for chiral chemistry advances." Chemical & Engineering News, 7 October 2026. cen.acs.org
12. Carroll, L. Through the Looking-Glass, and What Alice Found There. Macmillan, 1871; chapter 1.
13. Pasteur, L. "Sur les relations qui peuvent exister entre la forme cristalline, la composition chimique et le sens de la polarisation rotatoire." Annales de Chimie et de Physique 1848, 24, 442–459.
14. Kelvin, Lord (W. Thomson). Baltimore Lectures on Molecular Dynamics and the Wave Theory of Light, 1904, where he defines "chiral" (a term he first used in the 1890s).