In 1848 a man sorted crystals into left and right by eye. In 2003 a sealed flask, given one extra molecule in two million, made almost nothing but one hand. The committee said that, other than life itself, no one had done that before.
Henri Kagan, 95, and Kenso Soai, 76, won this year’s chemistry prize for showing how a slight preference for one mirror image can grow. Life did something like it to its amino acids and sugars. Where its first preference came from is still open, and so is whether it was a preference at all.
Buy a sprig of spearmint and a jar of caraway seed, and smell one and then the other. The two scents come from one molecule and its mirror image: the same atoms, joined in the same order. Your nose can tell them apart.
Hold a corkscrew up to a mirror and the one in the glass spirals the other way. Either would open a bottle, because a cork has no preference. A nut threaded for one would refuse the other. Your nose is threaded, and so is the rest of you. Life builds its proteins from one mirror form of amino acid and its genes from one form of sugar. No one knows why.
Louis Pasteur, the chemist who first pulled a substance apart into its two mirror forms, thought in 1860 that this might be the one line that could be drawn between the chemistry of the living and the dead. Whatever a chemist made from simple ingredients came out as an even mixture of the two forms. Only living things, and what came from them, made one form alone.
On the morning of October 7, in Stockholm, the Royal Swedish Academy of Sciences announced that this year’s chemistry prize would go to two men who had spent their working lives near that line. Henri Kagan, in France, had shown in 1986 that a flawed catalyst could do better than anyone expected. Kenso Soai, in Japan, had built a flask in which one form makes more of itself until the other has all but gone. Of what Soai’s flask had done, the Academy wrote: “Other than life itself, no one had previously achieved this feat.”
It is a generous sentence, and a strange one. It sets a flask of glass beside the whole of biology, at the same task. Follow how the flask got there, and the compliment begins to run the other way.
I
In the Morning
Pull the cork from a bottle of white wine that has spent the night in the refrigerator, and you may find a crust of small glassy crystals on its underside. Winemakers call them wine diamonds. They are harmless, a salt of tartaric acid, the sharp acid of grapes. Every one of them is built from the same mirror form of that acid, because the grape makes only one.
In the early nineteenth century tartaric acid was an industry. At the Kestner works in Thann, in Alsace, it was boiled out of the crust that wine leaves inside its casks and sold to the region’s cloth printers. In the early 1820s, mostly in winter, stray crystals turned up in the vats. They had the same makeup as tartaric acid, atom for atom, but they did not behave like it. The new substance was named for its source, acide racémique, from the Latin for a bunch of grapes. The works never made it again. The name stayed: a chemist still calls any even mixture of two mirror forms racemic, which means, if you trace it back, “of grapes.”
The difference showed in light. Pass light through certain crystals and it comes out vibrating in a single plane, like a rope shaken only up and down. Send that light through a solution of the grape’s tartaric acid and the plane turns a little to the right. Jean-Baptiste Biot, a physicist in Paris, had spent much of his life measuring such turns. Send it through the acid from Thann and it came out unturned. The same atoms, and no turn. Nobody could say why.
In 1848 Louis Pasteur was a young chemist at the École Normale in Paris, more interested in the shapes of crystals than in wine. A leading German crystal chemist had reported that a salt of the grape’s acid and the same salt of the Thann acid were alike in every way he could find, except for what they did to light. Pasteur made the Thann salt himself and looked at its crystals more closely than anyone had. On each one were small extra faces. On some crystals they leaned to the right, on others to the left, so that each crystal was the mirror image of another. He sorted them into two piles by eye. Dissolved, one pile turned light to the right, exactly as the grape’s acid did. The other turned it to the left, by the same amount. Mixed in equal weights, the two turned it not at all. The acid from Thann was the grape’s acid and its mirror twin, in equal shares.
Temperature mattered. This salt separates into left and right crystals only when it is cool; a few degrees warmer, it crystallizes as a single mixed kind, with nothing to sort. When Pasteur wrote out the method in 1850 for anyone who wanted to repeat it, it read like a recipe. Let the solution stand for a few days, until some of the crystals are several centimeters long. Take them out in the morning, because the day’s rising warmth dissolves them a little and erases the small faces. Pour off the liquid, wipe each crystal with crumpled blotting paper, and sort them one by one by the lean of their faces.
Biot did not take the result on trust. He had Pasteur make the salt again from Biot’s own acid, soda and ammonia, and leave it to crystallize in Biot’s own rooms at the Collège de France. Pasteur sorted the crop in front of him and said which pile would turn the light which way. Biot made the solutions himself, and looked first at the one that should turn it to the left: a form of the acid that, until Pasteur’s sorting, no one had ever held in a bottle. He saw it turn. Telling the story in 1860, Pasteur remembered the old man taking him by the arm: “My dear child, I have loved science so much throughout my life that this makes my heart throb.”
In the years that followed, Pasteur found two more ways to pull the forms apart. He mixed the Thann acid with an alkaloid from the bark of the cinchona tree, the tree that gives quinine. The tree’s molecule has a form of its own, so it got on differently with the two acids, and one came out of solution before the other. Then he let a common mold loose in a solution of the even mixture. The mold ate one form and left the other alone, and the liquid, which had done nothing to light when he began, slowly began to turn it. Each way of sorting needed something that could already tell left from right: a man’s eye, a tree’s alkaloid, a mold’s appetite.
He also tried to make one form from nothing. He grew crystals beside powerful magnets, and set up mirrors to make the sun appear to cross the sky backwards over growing plants. The magnets failed, and so did the mirrors.
By 1860, lecturing to the chemists of Paris, he set out what it all came to. That living things made one form alone established, in his words, “perhaps the only well-marked line of demarcation that can at present be drawn between the chemistry of dead matter and the chemistry of living matter.”
II
Borrowed Hands
Even inside a living body, a form does not always hold. Leave a molecule long enough, and warm enough, and it drifts toward an even mixture of its two forms. One amino acid in the dentin of your teeth drifts this way, year after year, so steadily that a tooth can be used to estimate its owner’s age. Thalidomide, the sedative that in the late 1950s and early 1960s harmed children before they were born, was sold as an even mixture of its two forms, and in the blood either form turns into the other within hours. One form has to be held against that drift. Getting it in the first place is harder.
For most of a century after Pasteur, a chemist could make one form only by starting from something that already had one. In 1904 a German chemist managed a slight excess of one form with brucine, an alkaloid from the seeds of the strychnine tree. Then, from the 1970s, came pure one-handed catalysts: molecules that help a reaction along and come out unchanged, ready to help again, so that each can stamp its form on a great many molecules of product. That work was honored with the chemistry prize in 2001. One of those catalysts, built on the grape’s tartrate, is the one that turns up in Kagan’s flask.
In 1986 every careful chemist wanted a purer catalyst. At Orsay, a university town south of Paris, Kagan made his worse on purpose, to test an idea he had first considered, he later wrote, in the late 1960s. He mixed the grape’s tartrate with a measured share of its mirror twin until it was only partly one-handed, and ran the reaction to see what came out. The rule chemists used was a straight line, drawn and seldom tested: a catalyst only partly one form would make a product only partly one form, in proportion, because each mirror-twin catalyst molecule would make mirror-twin product. By that line a flawed catalyst could only make a product flawed in the same proportion. It came out purer than the line allowed.
Kagan’s explanation is a picture anyone can hold. His catalyst works in pairs. Two of the same form make a pair that works. A right paired with a left, which Kagan called a “marriage,” makes a pair that barely works, and it sits out. The minority form, always the scarcer partner, ends up mostly married and idle, so the pairs still at work are mostly of the majority’s form. The worse catalyst behaves like a better one.
Pasteur had shown a room full of chemists the same kind of attraction. At the end of his first lecture in 1860 he poured together strong solutions of the grape’s acid and its mirror twin. The mixture grew warm, the liquid set solid, and the room applauded. In that crystal, each form takes more readily to its twin than to its own kind.
Drawn as a graph, with the catalyst’s purity along the bottom and the product’s up the side, Kagan’s results bowed above the old straight line, and the bend became an instrument. The shape of the curve shows whether a catalyst’s molecules work alone or in company, which no microscope can show, and chemists still draw it to look inside reactions they cannot watch. Later, a drug company’s chemists put the bend to work. They made a reagent from pinene, out of pine resin, that was far from pure, and got back a product almost entirely of one form.
Every hand in this part of the story came from something alive: the seeds of one tree, the resin of another, the grape. The product can beat the line, but the maker’s hand still came from the grape.
III
The Flask That Chose
In 1953, two flasks stood for the beginning of life.
One was real. In a laboratory in Chicago, a graduate student sealed water and the gases then thought to make up the early Earth’s air inside a loop of glass, and sent sparks through it for a week to stand in for lightning. By the end of the first day the water was pink; by the end of the week it was a deep, murky red. In it were amino acids, the building blocks of proteins, made from nothing but gas, water and electricity. They came in both forms, evenly. Whatever gave life one form, lightning alone had not.
The other existed only on paper. In Bristol that year F. C. Frank, a physicist who spent most of his career on how crystals grow and where they go wrong, published a short note in a biochemistry journal. He imagined a molecule that does three jobs: it makes more of itself; it makes more in its own form; and it gets in its mirror twin’s way. Start such a mixture perfectly even, let chance tip it the slightest bit, and it would run on by itself until only one form was left. He closed with one careful sentence: “A laboratory demonstration may not be impossible.” His flask stayed on paper until 1995.
Crystals got there first. Late in the 1980s, chemists in North Carolina were growing crystals of sodium chlorate, a salt whose molecules have no mirror form at all, though its crystals do: each one packs its atoms with a twist one way or the other. Left alone, a flask of it grows left and right crystals in roughly equal numbers. Stirred, it grew crystals nearly all of one kind, and which kind changed from flask to flask. By one account the stirring began on a whim, after one of them watched a crystal sink to the bottom of the flask. When they sent the result to a journal, one of its referees repeated it. The usual explanation is that the stirring broke the first crystal to form into fragments, and each fragment seeded new crystals in its own form. It was a world of one hand, made by nothing but motion and chance. But the hand lived in the stack, not the molecule. Dissolve the crystals and it is gone. That is why the Academy’s phrase still stands. It is about molecules.
Kenso Soai has spent most of his working life at the Tokyo University of Science. His chemistry joins a zinc compound to a small ring-shaped molecule to make an alcohol, and the zinc compound catches fire in air, so the work runs dry and sealed. In 1990 Soai found an alcohol that made more of itself in its own form: a little of one form in the flask helped turn out more of the same. But each batch came out less one-handed than the catalyst that made it. Frank had said what was missing. A molecule that only makes more of itself can at best keep the ratio it starts with, and in practice it slips; it must also get in its mirror twin’s way.
Five years later Soai found a molecule that did all three of Frank’s jobs. Take a little of what one flask made and use it to start the next, and each new batch came out more one-handed than the catalyst that made it, round after round. What Soai had built was an amplifier for left and right.
How his molecule got in its mirror twin’s way took longer to see. Chemists who measured the heat the reaction gives off as it runs found it behaving as though its catalyst worked in pairs. Frank’s last condition, the one Soai’s first molecule lacked, turned out to be Kagan’s marriage. In Soai’s flask too, a right paired with a left sits out. Not every account of the reaction agrees, but more than one does, and in those the two halves of this year’s prize meet in a single flask.
In 2003 his group gave the amplifier its hardest test. They mixed a trace of the one-handed alcohol into an even mixture until the excess was one molecule in two million. By the third run the flask was making almost nothing but one form.
Then they gave it nothing one-handed at all. Thirty-seven flasks were started with none of the alcohol in either form, sealed and run. Every one came out leaning, some a little and some heavily: eighteen one way, nineteen the other. A group in Texas had found the same the year before.
Something tipped each flask. It could be a stray trace of anything one-handed, a speck of dust from something once alive. It could, it seems, be nothing more than which few molecules happened to form first.
None of them chose.
IV
A Cause or an Accident
Kagan’s bend is useful. Soai’s reaction is good for nothing but itself: it makes one obscure alcohol nobody needs. A method is worth what it makes. A proof is worth what it rules out, and Soai’s flasks ruled out the oldest suspicion in this story, that only life could give molecules one hand from none. In the seeded flask, a hand diluted almost to nothing grew back. In the unseeded ones, a hand appeared where none had been put in.
It is not, the committee is careful to add, how life did it. Life’s chemistry runs in water, and water ruins this one.
In 1957 physicists found, to the disbelief of some of the best of them, that one force in nature can tell left from right: the weak force, which governs some kinds of radioactive decay. By some calculations it favors the form life’s amino acids take, by a margin far smaller than the chance excess of a few molecules that can tip a flask on its own.
Soai’s flask turned out to be finer than any instrument that measures the turning of light. His group tipped it with quartz, whose crystals come in left and right. They tipped it with crystals of the salt that sorts itself when stirred, and with something lit by light twisted one way. They tipped it with a molecule whose two sides differed by one extra neutron in a single carbon atom, and the alcohol followed the neutron. That turned “could this tip anything?” into an experiment. Which thing tipped life, it cannot say.
An instrument that fine cannot be shielded. Had something in the room been leaning on all thirty-seven flasks, they would have leaned together. They hadn’t. Whatever tipped them favored neither side. An amplifier makes a cause indistinguishable from an accident. Soai ran thirty-seven flasks and could count. Life, as far as anyone knows, ran once.
V
Other Than Life Itself
The news of the prize found Soai out shopping. That evening, on the telephone, he spoke first of Pasteur, who had tried before him, and called his own part a small contribution to the problem.
The problem runs back through everything alive. In a living thing, every hand comes from a hand before it. The grape makes its acid with enzymes built from one-handed amino acids, which were made in turn by enzymes before them. Followed back, the chain has to start somewhere. Where did the grape get its own?
On a Sunday morning, September 28, 1969, a fireball crossed the sky over Victoria, in southeastern Australia, and broke apart above the small town of Murchison. Pieces fell across the paddocks, and one came through a roof and landed in hay. People who picked them up said they smelled of methylated spirits.
The stone is a fragment of an asteroid as old as the solar system. Two chemists in Arizona found in it amino acids that life doesn’t use, leaning toward the form life does use. That amino acids fall from space is familiar by now; these arrived leaning. Because life doesn’t use them, the lean can’t be contamination from the people who picked up the stones. These amino acids barely flip, unlike the one in your teeth, so the lean has lasted almost as long as the solar system has existed.
Murchison’s parent asteroid was not always dry. Early on, it was warm enough inside for ice to melt and seep through its rock. Later, chemists at NASA compared stones. Where water had once soaked through them, these amino acids leaned, Murchison’s by nearly a fifth. In stones water had barely touched, they didn’t lean at all.
So the stone is a witness from outside every laboratory. It carries a seed nobody can name, and an amplifier that seems to have worked in water, where Soai’s flask cannot. How a lean in these amino acids could pass to the ones life uses is still an open question.
The Academy set a flask beside the whole of life, as if the flask had risen to life’s level. It may be the other way round. What life did to its molecules may need nothing a flask cannot do: a tilt no one meant, made larger by molecules that make more of themselves, somewhere in water, once. Or it may need something not yet imagined. The stone does not say. It is what it was on the morning it fell: a stone that arrived leaning, for reasons nobody can yet explain.
⁂
This essay answers the Royal Swedish Academy of Sciences’ award of the Nobel Prize in Chemistry 2026 to Henri B. Kagan and Kenso Soai (October 7, 2026). It draws on the Academy’s press release, popular information and scientific background. The committee’s phrase is “previously achieved” in the press release and “ever achieved” in the popular information; the press release is followed. Kagan’s finding is from C. Puchot, O. Samuel, E. Duñach, S. Zhao, C. Agami and H. B. Kagan, Journal of the American Chemical Society 108 (1986). Soai’s first self-making alcohol is from K. Soai, S. Niwa and H. Hori, Journal of the Chemical Society, Chemical Communications (1990), and his amplifying one from K. Soai, T. Shibata, H. Morioka and K. Choji, Nature 378 (1995). The press release describes Soai’s 2003 success in one sentence that draws on two papers, the seeded experiment (Sato, Urabe, Ishiguro, Shibata and Soai, Angewandte Chemie, 2003) and the unseeded one (Soai and colleagues, Tetrahedron: Asymmetry, 2003); the figures here are from the scientific background and the papers.
Pasteur is quoted from his 1860 lectures to the Société chimique de Paris in the Alembic Club translation, including his own account of Biot’s verification. Gerald Geison and Howard Flack have tested his 1860 account of what led him to the discovery against his notebooks, and Joseph Gal has defended it; the essay relies on it only for Biot’s verification. His method for the separation is from his 1850 memoir, “Recherches sur les propriétés spécifiques des deux acides qui composent l’acide racémique,” Annales de chimie et de physique, 3rd series, 28. The blotting paper and the morning harvest are given as Hervé This quotes them in the Notes académiques de l’Académie d’agriculture de France (2021), from whom the account of the Kestner works is also drawn. The familiar picture of Pasteur sorting crystals under a microscope with tweezers comes from later biography, not his own accounts, and repeating his separation has not always succeeded.
Kagan’s word “marriage” is from his review in Oil & Gas Science and Technology (2007); his first consideration of the idea in the late 1960s is from his personal account in Synlett (2001), as cited by the committee. F. C. Frank is quoted from Biochimica et Biophysica Acta 11 (1953). The 1953 spark experiment follows Stanley Miller’s paper in Science (May 15, 1953). The stirred-salt result is from D. K. Kondepudi, R. J. Kaufman and N. Singh in Science (1990); the whim, the sinking crystal and the referee are from Charles Day’s account in Physics Today (April 2005); and the usual explanation is disputed (C. Viedma, Physical Review Letters, 2005). The disbelief of 1957 is Wolfgang Pauli’s among others, in his letter to Victor Weisskopf of January 17, 1957. The meteorite findings are from J. R. Cronin and S. Pizzarello, Science (1997), and D. P. Glavin and J. P. Dworkin, Proceedings of the National Academy of Sciences (2009). The idea that one form of thalidomide was harmless rests on work in mice; an earlier study found both forms harmful in rabbits, and the two convert into each other in the body. Soai is paraphrased from his telephone interview with Nobel Prize Outreach on October 7, 2026.
Chemists described but not named: Eilhard Mitscherlich, the German crystal chemist; Willy Marckwald, 1904; K. Barry Sharpless, the 2001 work; Stanley Miller, Chicago, 1953; D. K. Kondepudi, R. J. Kaufman and N. Singh, North Carolina, 1990; Daniel Singleton and Lam Vo, Texas, 2002; the Merck process chemists; Donna Blackmond, John M. Brown and colleagues, the heat measurements, 2001; John Cronin and Sandra Pizzarello, Arizona, 1997; Daniel Glavin and Jason Dworkin, NASA, 2009.
Kenso Soai was born in Hiroshima in 1950; Henri Kagan in Boulogne-Billancourt in 1930. Drafted with Claude Opus 5.5; literary editing by Claude Opus 5.5; outside review by Claude Fable 5.1; copy editing by Gemini Argon (the configured models; the serving model on any turn may differ). Images generated with Gemini Argon from prompts drafted with Claude Opus 5.5.