The Impossible Rock


The Impossible Rock


The Impossible Rock


The Impossible Rock

A strange mineral in an Italian museum, secret diaries, gem smugglers, a meteorite, and an expedition to a remote part of the world... The extraordinary global hunt for the origins of a rock that was not supposed to exist.

Paul Steinhardt (BS ’74) once imagined a kind of crystal that was not supposed to exist in nature. Decades later, with the help of an Italian mineralogist, he found one. Tracing its origins would take Steinhardt, his son Will (BS ’11), and an international team of researchers to one of the most remote parts of the world, and open new mysteries about the formation of the solar system.


In the fall of 2008, Luca Bindi sat in his lab at the Museum of Natural History at the University of Florence in Italy, poring over data analysis. Then the head of mineralogy, Bindi was conducting a series of X-ray analyses on a selection of micromounts, tiny rock samples displayed in showcases in the museum’s main exhibition hall, specks only a bit larger than grains of sand. One of the samples, catalog number 46407/G, made him sit up straight. If he was reading the data correctly—and he was pretty sure he was—46407/G exhibited properties strongly indicative of an incredibly rare type of matter: a quasicrystal.

If so, it would be the first natural quasicrystal ever discovered.

Quasicrystals were not supposed to exist naturally. In fact, up until three decades ago, they weren’t supposed to exist at all. The laws that govern the structure of crystals dictate that they exhibit certain properties, including specific symmetries. Quasicrystals defy such boundaries, with compositions unlike any of their brethren. Their initial discovery in 1984 stirred deep controversy and overturned existing preconceptions about matter. Still, they are exceedingly rare, and had only been made in laboratories under tightly controlled conditions. So it then became the common wisdom that nature must not have elected to make them. If this sample was actually a quasicrystal, and if it was naturally made, it would contradict what had for nearly a century been considered a fundamental law.

Sample 46407/G might change the way we look at the world.

Bindi put his findings into an email a few days later. “The results are incredibly promising,” he wrote—the science-speak equivalent of “Eureka!”


Across the Atlantic, Bindi’s email pinged the inbox of Paul Steinhardt (BS ’74), a theoretical physicist and cosmologist at Princeton University. Steinhardt is a warm but quiet man, pensive and careful with his words. As the director of the Princeton Center for Theoretical Science, he is also regarded as one of the nation’s leading cosmologists, widely known for developing the first working model of cosmic inflation (and then decades later recanting it in favor of a radically different theory).

Steinhardt also happens to be one of the world’s foremost authorities on quasicrystals; three decades ago, he practically coined the term.

So what exactly is a quasicrystal?

When most people think of a crystal, they conjure up images of jewelry cases filled with multihued minerals with sharp, angular edges. It turns out that these shapes are the result of their atomic structures, which arrange into neat, regular, and repeating geometrical patterns. Now think of tiles spread across the surface of a floor: Depending on the shape of the tiles, the pattern produces certain symmetries. While there are thousands of crystals, their shapes can be grouped into only four specific symmetries—either two-, three-, four-, or six-fold.

The first hints of this were discovered more than a century ago in 1912, when the German physicist Max von Laue sent beams of X-rays through crystals, capturing beautiful portraits of the resulting diffraction patterns. This quickly opened the door to further discoveries about the makeup of matter. X-ray crystallography, as it became known, offered a new method that allowed researchers to peer into the dazzling and complex atomic structures of minerals, viruses, proteins, and even DNA. Yet compared with other forms of matter, crystals proved to be nature’s minimalists, and the understanding of their rigid architecture and symmetry was considered a closed case.

“It was so fundamental, you learned it in elementary school,” said Steinhardt. “Two, three, four, or six folds—period. That was it.”

There are, of course, other symmetrical compositions in the universe. The tiling on a soccer ball, composed of pentagons and hexagons, includes five-fold symmetries. Mosques throughout the Middle East display Girih tiling with two-, five-, and 10-fold symmetries that produce ornate patterns dizzying in complexity and beauty. In 1976, mathematical physicist Roger Penrose proposed a set of just two tiles that could cover a plane infinitely in nonperiodic fashion, full of five-fold symmetries.

But to have such exotic formations occur within crystals…that was forbidden by the laws of material science as we knew them. Such a crystal would be impossible.

Turns out, Steinhardt has a problem with the word “impossible.”

“I think that when someone says ‘impossible,’ I always want to know: Do you mean it violates the laws of physics?” Steinhardt asks. “Or do you mean that it would be very, very interesting?”

Steinhardt decided to put the crystal law to the test. In the early 1980s, while a faculty member at the University of Pennsylvania, he and fellow researcher Dov Levine developed a theoretical framework arguing that solids with forbidden symmetries could occur, at least in principle. They termed this new phase of matter “quasiperiodic crystals,” or “quasicrystals.” By sheer coincidence, around the same time another materials scientist, Dan Shechtman, who was at Johns Hopkins University, stumbled upon just such a formation lurking within an aluminum-manganese alloy and published his findings in 1984, just a few months before Steinhardt and Levine published their own paper.

“Shechtman had a sample without a theory,” Steinhardt says. “We had the theory without a sample.”

The name stuck, but that was all. The reaction from the scientific community was swift and withering. The very idea of quasicrystals, most experts argued, was so heretical to the laws of crystallography that Shechtman, Steinhardt, and their colleagues had to be in error. “There are no such thing as quasicrystals,” two-time Nobel laureate and Caltech professor Linus Pauling (PhD ’25) famously declared, “only quasi-scientists.” Pauling labored to explain the strange patterns with another theory, but the evidence mounted as more examples of quasicrystals continued to be found. Within a decade, the tide of scientific thinking began to turn. “I think Pauling thought that such an extraordinary claim required extraordinary evidence,” Steinhardt said. “In the end, I believe he felt that he served by pushing us to eliminate all other options and strengthen our case.”

In 1998, Steinhardt moved from Pennsylvania to the physics department at Princeton, and his primary attention drifted to other research interests. He also focused on his role as a father to his four young children.

Still, one issue continued to trouble him: Hundreds of quasicrystals had been discovered by researchers around the world, but all had been created in a lab. Why had none been found to occur naturally?

“These things could be made, but only under very controlled processes,” Steinhardt said. “But what if nature had already figured out another way to make them?” If so, might one already be hiding within some existing collection? He just had to figure out how to find the needle in a very, very large haystack.

The Internet offered a solution. For a number of years, mineralogists had been cataloging their samples in an international database. This included information from routinely performed powder-diffraction tests, a kind of quick and dirty version of the type of X-ray diffraction test used to detect quasicrystals. Powder diffractions weren’t enough to make a positive identification, but they could be used to rule out samples. In 2001, Steinhardt and fellow physicist Peter Lu published an open invitation, along with a methodology, for researchers to help them search their collections. No one responded.

Years passed. Steinhardt’s children grew, went on to high school, and then began to think about college. Then, in 2007, one mineralogist offered to help: Luca Bindi in Florence.




Bindi is, in many ways, a study in contrast to Steinhardt. Where the latter can be collected and circumspect, Bindi is energetic, a fast and reactive speaker with an infectious enthusiasm for his work. In 2007, he was newly appointed as the head of the division of mineralogy at the Museum of Natural History at the University of Florence, the curator of more than 50,000 mineral samples collected from around the world. Bindi thought that if a quasicrystal was hiding in someone’s collection, it might as well be his, and so he agreed to enlist in Steinhardt’s search.

“Steinhardt had already identified several candidates, but none worked out,” Bindi said. “I decided to look at samples with aluminum alloys, which had compositions similar to known quasicrystals.” Over the course of the next year, Bindi would slice razor-thin portions of already tiny grains, then mount them on a glass fiber for examination. It was slow, tedious work that required a great deal of precision. It also meant the destruction of large portions of the material.

Bindi’s eureka moment came on October 2008, when he prepared 46407/G. The rock grain contained a unique blend of copper, zinc, and aluminum interwoven with other minerals—and a few grains of something Bindi couldn’t identify. When he conducted the X-ray diffraction test, it was filled with the spikes and patterns that Steinhardt and Lu predicted. Bindi was stunned. “I absolutely felt, ‘My God, I think I’ve found it!’” —but his excitement was met with an unexpected splash of cold water.

Screen Shot 2015-11-24 at 11.22.46 AM.png

“Extraordinary claims require extraordinary evidence,” Steinhardt reminded. He then asked Bindi to send him the sample directly.

When the package arrived at Princeton that December, Steinhardt frowned. The sample was incredibly small, and the work Bindi had already performed meant that there was little material left over. “I had to really squint just to see a couple of the grains,” Steinhardt said. “I was doubtful that we would have enough for all of the experiments we wanted to perform.”

On the morning of New Year’s Eve 2008, Steinhardt trudged across Princeton’s frozen and dead-quiet campus to the university’s PRISM Imaging Center. With no one else in the building, Steinhardt and the center’s director loaded the sample into a powerful electron microscope and switched it on. The screen lit up with a diffraction pattern of dazzling white dots arranged in a starburst pattern, a bit like looking at a streetlight through a kaleidoscope. Steinhardt leaned in; he recognized the pattern immediately.

Without question, it was a quasicrystal.

“The first spot that we hit was just spectacular,” Steinhardt said. “I guess I had expected to see something with many defects, but this diffraction pattern was as nice as any I had ever seen.”

Steinhardt dashed an email back to Bindi, “Happy Quasi–New Year.”


Now that Steinhardt was in possession of what he knew to be a quasicrystal, two questions confronted him: How exactly did this rock come to be? And how did it wind up in Bindi’s collection? The first was a scientific problem, the second more of one for a detective. Both proved to be vexing. Steinhardt started by taking the sample to Lincoln Hollister, a professor of geosciences at Princeton and a leading petrologist—someone who investigates the origins of rocks. Hollister took a look, then offered some bad news. “What you have here is impossible,” he said.

There it was, that word again, “impossible.”

Hollister elaborated that the rock couldn’t be natural. First, the sample contained metallic aluminum, which is present on Earth only in man-made form. Aluminum has a strong affinity for oxygen, and binds to it to make aluminum oxide. More puzzling, the substance contained copper, which doesn’t mix with aluminum except in rare manufactured alloys. Steinhardt pressed, “Could it be a meteorite?” They took the sample to specialists in meteorites, who initially ruled out the possibility that it was extraterrestrial.

Hollister’s conclusion: The sample was artificial, likely a piece of slag, or waste from a factory. It could not have formed naturally.

“I have to admit, it was pretty depressing,” Steinhardt said. After the initial excitement, he was now unsure just what he had. Was he really looking at a piece of slag? Could runoff from some mining facility truly produce something as exotic as a quasicrystal? If the rock was not natural, then its significance was severely diminished.

He needed more information.



Back in Italy, Bindi had also hit a wall. “It was the fall of 2009. I remember sitting at a dinner party one night, utterly frustrated,” Bindi said. He regaled fellow guests with his own efforts to find the source of 46407/G. The sample was labeled khatyrkite, a name that referred to where it was purported to have been found: Khatyrka, located in the Koryak Mountains, a remote mountain range in far-northeastern Siberia. There were some other samples of khatyrkite in museums, but testing revealed all to be fakes, which is less rare than one might imagine; mineralogy is a collector’s game, and prone to counterfeiting.

This meant that there were only two confirmed rocks: theirs and the holotype—the original specimen—which was housed in the St. Petersburg Museum in Russia. There was a good deal of information on the holotype in the 1985 paper naming it. If the samples were related, then they would know much more about their rock’s origins. But other than sharing a name, there was no direct evidence linking the two, and the museum would not allow direct testing. They reached out to the paper’s author, a retired platinum specialist named Leonid Razin, but were greeted with suspicion. While Razin officially named the mineral khatyrkite, their interactions left Steinhardt uncertain how much the former Soviet scientist actually knew, or whether he had even been the one to retrieve it.

With the holotype a dead end, it added pressure to trace the exact steps of 46407/G. Bindi found that his museum had acquired the specimen in a collection bought from a Dutch gem dealer named Nicholas Koekoek.

“But who is this Koekoek?” Bindi lamented to his friends at the dinner. “He’s nowhere to be found on the Internet, and the last name is a common one. It’s like looking for John Smith.”

“I know an old woman named Koekoek,” one of the fellow dinner guests, who also lived in Amsterdam, said. “I can ask her. Maybe she’ll know somebody.” Bindi shrugged, at that point resigned that any lead, however thin, was worth a shot. Then, days later, he got a call with surprising news: The woman was in fact the gem dealer’s widow. Bindi booked the soonest possible flight to Amsterdam.

Meeting in her dim apartment, Koekoek’s widow shared with Bindi a “secret diary” that had belonged to her husband. Within its pages was an entry detailing the khatyrkite: It came from a Romanian smuggler identified only as Tim. Bindi tried for months to track Tim down, but could find no trace of him. “This ‘Tim’ was even more mysterious than the first gem dealer,” Bindi said. “There was no evidence he had traded with anyone else.” So Bindi went back to the widow six months later, hoping for a little more information.

Then, to Bindi’s immense surprise, the widow went to her bookshelves and retrieved a second diary—a “secret, secret diary.” Koekoek apparently kept two versions of his notes, one to protect his sources and presumably himself, and one that acted as his definitive record. This second secret diary revealed a different name as the supplier of the khatyrkite, and a familiar one: Leonid Razin.

“So there Razin was again,” Steinhardt said. “He was the source of both the sample in St. Petersburg and ours
in Florence.”

Steinhardt remained wary of Razin, and began to feel somewhat adrift in his search. The Princeton professor’s world was in the cosmos. He traded in theories and experimentation, not gemstones, smugglers, and secret diaries. Still, both scientists and detectives required one thing in common: hard evidence.

Hoping for another clue, Steinhardt began to pore back over the original 1985 paper when he stopped on a name: Valery Kryachko. Most had assumed Kryachko was a local miner who perhaps had helped the science team, and that was all. But Steinhardt suddenly recalled having seen the same name in other scientific papers. Kryachko might be more involved than previously thought.

Steinhardt and Bindi found Kryachko, now in his 60s living part-time in Moscow, and reached out via email. To their delight—they learned that he was a mineralogist, and, in fact, he had also heard of them. Kryachko had read with interest the articles published in academic journals about their quasicrystal search, though he had no idea he might somehow be connected.

Kryachko confirmed that, indeed, he had been the one to find the khatyrkite. As a young geologist, he had been sent in 1979 to the Siberian peninsula by Razin to search for
platinum and gold deposits. He didn’t find any gold, but Kryachko did come back with some small, interesting nuggets he couldn’t identify and gave them to Razin.

“We couldn’t believe it,” Steinhardt said. “We called just hoping for a lead on our sample, and we found the very man who may have retrieved it.”

Even better, Kryachko said that he remembered the exact spot—and would be willing to take them there.


“I have heard the stories about quasicrystals my entire life,” said Paul’s son, Will Steinhardt. Tall and slender, with what friends describe as an adventurous demeanor and quick wit, Will Steinhardt grew from a boy to a young man over the course of his father’s search. “A lot of people treated my father as this renowned scientist. But to me he was just my dad—and one of his interests was looking for this weird kind of crystal.” Still, something must have made an impression; Will Steinhardt eventually chose to follow in his father’s footsteps and began attending Caltech in 2007, where he majored in geophysics (BS ’11).

Will was 20 and a sophomore when 46407/G was found, and as the elder Steinhardt undertook his investigation, the father and son would spend long hours discussing the rock together by phone and over holiday visits. “I think my dad would readily admit that he’s not a geologist,” said Will Steinhardt. “Here I was being educated at what I believe is the best geosciences program in the world.”

The two agreed that Caltech might be able to help more directly, and Paul Steinhardt brought the sample to Pasadena for testing, where he was referred to John Eiler, the Robert P. Sharp Professor of Geology and Professor of Geochemistry. Testing the samples at the Caltech Microanalysis Center, Eiler found silicates and oxides that bore a distinctive oxygen-isotope fingerprint dating back to the pre-solar system. The sample was extraterrestrial.

“No mineral on Earth bears such a signature,” Eiler said. “There was your answer, and it was the least likely one: This rock fell on our heads from space.”

Now Paul Steinhardt had two important pieces of evidence: He knew that the rock was a fragment of a meteorite, and he also knew where it was found. There was one thing left to do. He called his son and asked if he wanted to join an expedition to Russia.

“There aren’t too many times you get to go to a remote part of the planet in search of a one-of-a-kind meteorite,” laughed Will Steinhardt. “You know…just a typical camping trip with your dad.”


In July 2011, a team of scientists, including Paul and Will Steinhardt and Bindi, gathered outside Anadyr, a town in the northeast of Siberia. There they met Kryachko and a small contingent of Russian mineralogists who would be their guides. Never an outdoorsman, Steinhardt was now venturing into one of the most remote regions in Russia. Weather ruled out traveling by air, so they would have to drive in snowcat trucks, which were essentially large metal boxes with tank treads meant to carve their own roads.

Crossing the tundra was a grueling four-day journey. The ground, packed under snow and ice for most of the year, had thawed to a slushy mud that made walking impossible. Bears were known to be close by, migrating to the streams for the wild salmon, which were so abundant you could reach in and catch them with your hands. Then there were the swarms of mosquitoes, which immediately attacked any exposed skin. “You would unzip your hood to eat, and they would fly inside your mouth,” Will Steinhardt said.

Kryachko led them to the small streambed where, 32 years before, he had first sifted the soil. They had finally arrived at the source.

Paul Steinhardt surveyed the area. He tried to remain realistic about their chances of finding another quasicrystal sample, which he estimated to be one in 1,000. The more feasible objective, he reminded everyone, was to learn as much as they could about the surrounding environment. Still—he allowed himself a bit of optimism. Over the next 10 days, the team worked to dredge mud from the river. The thick blue clay broke their shovels, so Will Steinhardt resorted to using his hands to fill 40-pound buckets, which they would boil down. Kryachko then panned the remnants in the riverbed and boiled it again, exactly as he had before, until all that remained was a thimbleful of dust. Will Steinhardt marveled at Kryachko’s speed and dexterity, “You could see the decades of experience in his hands.”

Bindi examined some of the samples through a microscope in his tent. The more he looked, the more enthusiastic he became. “There were a number of samples that looked very similar to ours in Florence,” Bindi said. There was, of course, no way to know whether or not they had a quasicrystal until they returned—but Bindi’s optimism was infectious, and the team felt buoyed by the sense that they were on to something.

Paul Steinhardt found that he could now relax a bit, which allowed him to appreciate another unique part of the experience—working with his son as a scientist. “One of the great privileges of being a father is the ability to form a relationship with your child as an adult,” Steinhardt said. “I got to see how excellent a field scientist Will is.”

Will Steinhardt echoed the sentiment, “I came not just because I know geoscience, obviously, but also because I know my father. And after this trip…I felt I knew him better.”

Their last night, Kryachko offered a heartfelt toast, remarking that being associated with the search was a highlight of his career. Bindi felt the same, “This project has really become a part of me. I owe a great deal to my collaboration with Paul, and am proud to count him as a terrific friend.”

Everyone then placed bets on whether or not they had found even one sample containing a quasicrystal. “By the time we left, I thought our odds had improved to one in 100,” Paul Steinhardt said. Bindi wagered one in 20.

EVERYONE PLACED BETS on whether or not they had found even one sample containing a quasicrystal. “By the time we left, I thought our odds had improved to one in 100,” Paul Steinhardt said.

EVERYONE PLACED BETS on whether or not they had found even one sample containing a quasicrystal. “By the time we left, I thought our odds had improved to one in 100,” Paul Steinhardt said.



Bindi was right to be optimistic. Back in Princeton, tests conducted on one of the larger samples revealed the same composition of aluminum alloy and—lurking within—a grain of quasicrystal, the very same variety that had been found in 46407/G.

“All of that crazy, crazy stuff we went through—it just paid off,” Paul Steinhardt said. “To see that exact same X-ray diffraction pattern was immensely satisfying.”

While Steinhardt and Bindi were conducting their research, Dan Shechtman was awarded the Nobel Prize in Chemistry for spotting the first quasicrystal nearly three decades earlier. The quasicrystal had gone from a theoretical object, to a contested curiosity, to an accepted man-made substance, to—when Steinhardt and Bindi published their full findings in the spring of 2012—a part of nature.

There are, it turns out, a number of mysteries packed into the tiny specks of Paul Steinhardt and Bindi’s extraterrestrial rock: Not only the presence of never-before-seen combinations of elements like metallic aluminum, but how they formed and what that tells us about the evolution of our solar system.

“We thought that a certain type of matter could not exist—and it can. Then we thought nature couldn’t make it—and it does,” Paul Steinhart said. “The existence of this one little rock tells us that we missed something, which raises the question: What else might we be missing?” Like any good mystery, the story continues, but the investigators have assembled a compelling case for the journey of 46407/G:

Four-and-a-half billion years ago, objects in the pre-solar system smashed together, resulting in an object with strange and unusual properties, including particles of quasicrystal. The asteroid lingered in space until 15,000 years ago, when it fell to Earth, crashing into the mountains of what would eventually be northeastern Russia. With time and erosion, the shattered pieces slowly migrated, coming to rest in a small streambed, where they were sifted in 1979 by a young Russian mineralogist panning for gold. One piece was sent to a museum in St. Petersburg, the other passed through several hands before coming to the University of Florence, where it was labeled and placed into storage. Decades later, a researcher spotted signs of a dazzling crystalline structure lurking inside, and sent it to Paul Steinhardt, the man who was among the first to imagine that such a crystal could even exist, and who had been searching the world for just such a rock.

What would be the chances? Incredibly low. But not, it turns out, impossible.

No items found.

More posts like this

Quantum Hoop Dreams

Quantum Hoop Dreams

How Dean Oliver (BS ’90) became known as the father of basketball analytics.

Looking for E.T.

Looking for E.T.

After five decades hunting for intelligent life in the universe, we have yet to find any evidence. Does it mean there’s nothing out there? Or do we need to rethink what we’re searching for?