Lincoln’s pronouncement sounded definitive. At this point, most geologists, respecting his reputation and hearing his stern tone of voice and all it conveyed, would have thanked him for his advice and ended their investigation right there and then.
Sitting before him, though, was a stubborn theoretical physicist who was a neophyte when it came to geology but all too familiar with impossible challenges. So I persisted and asked Lincoln the same question I always ask myself whenever I hear the word “impossible.”
“When you say ‘impossible’ . . . Do you mean impossible like 1 + 1 = 3? Or do you mean very, very unlikely? And if true, very, very interesting?”
Thankfully, Lincoln did not seem to think my question was impertinent because he did not immediately throw me out of his office. Instead, he paused a few moments to consider the question. When he finally spoke, his voice had returned to a more normal volume.
“I suppose,” he said thoughtfully, “that if I were forced to come up with a natural explanation, I would need to find conditions where aluminum could easily separate from oxygen. It would require ultra-high pressures, which can be found three thousand kilometers below the surface of the Earth near the core-mantle boundary.
“But then,” he continued to speculate, “if you could manage to make metallic aluminum and form your quasicrystal, you would need to find a mechanism to get it to the surface of the Earth rapidly without the mineral decomposing and the aluminum reacting with oxygen on the way up.”
For a moment, I was worried that he might consider this an impassable roadblock. Happily, I was wrong.
“There is a conceivable way this can happen,” he offered. “You may know of Jason Morgan, the Princeton geoscientist who helped to establish the modern theory of plate tectonics.
“Jason retired a few years ago. He had a theory that there might be superplumes, tube-like upwellings of material from the core-mantle boundary to the surface. The superplumes, if they exist, would be giant versions of the well-known plumes that created the Hawaiian Islands.
“The superplume idea has never been proven,” Lincoln continued. “But if your sample was made at the core-mantle boundary and carried to the surface in this way, it would be the first direct evidence of his idea.”
When Lincoln finished, my eyes must have been the size of saucers. Our sample is not impossible after all, I thought. And, if it turns out to be of natural origin, it is going to be extremely important.
After a brief silence, I timidly offered my pet idea. “If the problem is keeping aluminum away from oxygen, is it possible that the sample was made in space, perhaps inside a meteorite?”
The idea that meteorites might be a source of quasicrystals had occurred to me before. I had been thinking about it for several years and had even mentioned the possibility to Luca, although we never pursued it. I was not aware at the time, though, that my question was terribly naive. I thought there was little or no oxygen in space when, in fact, meteoroids and asteroids are full of oxygen bonded to other elements.
Fortunately, Lincoln did not point out the error in my thinking. “I do not know much about meteorites,” he said, “but I know someone who does.”
Lincoln was referring to Glenn MacPherson, the head of the Division of Meteorites at the Smithsonian National Museum of Natural History. Glenn received his PhD from Princeton in 1981. Lincoln had known him for decades and had even recommended him for his current position.
Lincoln offered to arrange a visit for me to meet Glenn at his office in Washington, D.C. He also offered to accompany me on the trip, and I eagerly accepted. I interpreted the offer to be a positive sign that the legendary geologist was somewhat excited by our finding.
Once I returned to my office, I emailed Luca in Italy to tell him about my meeting with Lincoln. Luca was aware of Lincoln’s professional reputation and held him in high esteem. I tried to be as upbeat as possible and did not dwell on the fact that Lincoln’s first impression was that the sample was just an ordinary piece of scrap metal.
Luca, like me, was not aware that metallic aluminum had never been found before in nature. That bit of news meant we had even more reason to worry about the reaction to the Science article we had submitted. Not only were we reporting an impossible new form of matter, a quasicrystal, we were also announcing that we had found natural metallic aluminum, making our discovery doubly impossible.
Luca was impressed with Lincoln’s superplume idea and at how well the conversation seemed to have gone overall. But in truth, I was becoming concerned. Both of the proposed explanations, superplumes and meteorites, sounded like long shots.
A week later, Luca and I received good news from the editors of Science magazine. The manuscript describing our discovery of the first natural quasicrystal had passed the first round of review. This was promising. The fact that the editors had not rejected our paper outright meant they did not consider the evidence to be ridiculous, even though the quasicrystal included metallic aluminum. The real test, though, would be the next round of scientists reviewing the article. They would all be experts in the field who, much like Lincoln, would probably consider the reported discovery of natural metallic aluminum to be preposterous.
* * *
WASHINGTON, D.C., JANUARY 24, 2009: “Impossible!”
As Lincoln and I climbed the steps to the entrance of the Smithsonian, Glenn MacPherson stood at the top of the stairs. He was holding open the giant door to the Natural History Museum while blurting out his opinion of the Florence sample loud enough for everyone to hear.
I did not know that Lincoln had prepped Glenn about the subject of our meeting and that, by the time we arrived, he had already given our sample careful consideration. So I was completely startled by the way Glenn introduced himself.
Glenn was taller than Lincoln and me. He was slender with dark hair graying at the temples, had a dark mustache and, unlike Lincoln, the look of someone who spent all of his time in the laboratory.
Glenn ushered Lincoln and me inside and helped us register for the special badges we would need to enter the inner sanctum of the Smithsonian. He then proceeded to lead us on a long, labyrinthine path to his office that included countless corridors, elevators, a slew of security doors, and then, yet again, even more corridors. The entire time we followed him through all of those endless passageways, Glenn continued to pummel me with all of the reasons our sample could not possibly be natural.
The presence of metallic aluminum, the same concern that Lincoln had raised, was only the first of the problems, according to Glenn. Once we finally arrived at a meeting room near his office, he asked us sit down at a big table where he proceeded to lay out a series of key papers and data demonstrating how impossible it is to form metallic aluminum naturally on Earth.
“And as for meteorites . . .” Glenn said ominously, as he began to take aim at my pet theory. In all of his experience with meteorites, and Glenn had seen all types, he assured us, he had never seen a single example that contained any metallic aluminum or aluminum alloys.
Glenn was convinced, absolutely convinced, that our sample was . . . and then he uttered the dreaded four-letter word . . . S-L-A-G. Slag is a catch-all name for a useless by-product of an industrial process. Slag meant unnatural. Slag meant we had failed to find what I thought we had found. Slag was the ugly word I did not want to hear.
Glenn was not finished with his case, however. Another big problem, he explained, was our claim that the metallic aluminum was supposedly mixed with metallic copper in three different minerals in our sample: khatyrkite, cupalite, and our quasicrystal. That is also impossible, he asserted. Just as aluminum has an affinity for oxygen, copper has an affinity for sulfur.
The two metals are found in different classes of minerals because of their different chemical bonding patterns. It would be inconceivable, according to Glenn, that they could naturally form a metallic alloy like khatyrkite, cupalite, or our quasicrystal through any natural geochemical process.
A third iss
ue was the absence of any corrosion. How could a sample containing aluminum metal survive on the surface of the Earth without any sign of rust?
Glenn then continued to rattle off a seemingly endless list of additional reasons why the sample could not possibly be natural.
As I listened and took notes, I could tell that Glenn had put considerable thought into the issue and presumed that he was trying to impress Lincoln, his former mentor. At first, Lincoln tried to defend the case for a natural quasicrystal by presenting his novel idea about superplumes. But he eventually wilted under Glenn’s continued barrage. By the time we left the Smithsonian a few hours later, Lincoln seemed completely convinced by Glenn’s conclusion: Our sample had to be an artificial by-product of an aluminum smelter or laboratory.
The Smithsonian meeting could have marked the end of the investigation. I was sure that Lincoln and Glenn thought they would never hear from either me or Luca again.
But I was not swayed by Glenn’s arguments. They all relied on a variety of reasonable but, truth be told, unproven scientific assumptions. And while Glenn had presented abundant evidence supporting his case, all of his evidence was, by definition, a reflection of what had been observed in the past. None of it proved that it was impossible to find something new in the future.
I preferred to view the situation another way. If Glenn was wrong and the Florence sample was not slag, then it represented something even more spectacular than we first imagined. It would not only serve to prove the existence of natural quasicrystals, it would also overturn widely accepted assumptions about the kinds of minerals that can form in nature.
Once Lincoln and I returned to Princeton, I wrote an email to Luca containing a full and brutally honest summary of the visit. I hit the send button, and wondered if the disappointing news would make him decide to abandon the project. I did not have to wait very long for an answer. Within minutes, a reply appeared in my mailbox.
Luca had no intention of giving up. He was confident that our sample was natural. Not only that, he was every bit as committed to the investigation as I was and equally determined to work with me to prove the case scientifically. Luca and I acknowledged to each other that we were charting a dangerous course. It would be a very public battle that could, despite our best efforts, prove to be very embarrassing.
In order to proceed, we would need to formulate a new strategy. And we would need our two harshest critics, Lincoln and Glenn, to play a key role.
ELEVEN
* * *
BLUE TEAM vs. RED TEAM
PRINCETON AND FLORENCE, JANUARY 25, 2009: Luca and I were under enormous pressure. We had written and submitted the scientific paper announcing our discovery and the review process was well under way. But now we were getting major pushback from Lincoln and Glenn, neither of whom agreed with our conclusions.
The sample was slag, they argued. We had been hoodwinked. Natural khatyrkite, our natural quasicrystal, and metallic aluminum were most certainly impossible.
Their opposition put us in a terrible bind. If the Science paper was published and later turned out to be wrong, as Lincoln and Glenn believed, the damage to our reputations and impact on future research projects would be enormous. On the other hand, if we pulled back and withdrew the paper, the reversal would draw attention and raise suspicions. The search for natural quasicrystals would lose credibility in the scientific community and perhaps come to an end altogether.
The only way out of the predicament was to do our best to work quickly to resolve the central scientific issue before publication. Is the quasicrystal natural or slag? In our view, the evidence strongly tilted toward the sample being natural. But we needed more than that. We needed substantial evidence, enough to sway our harshest critics.
It was essential to keep Lincoln and Glenn involved in the investigation. For one thing, the four of us made a good team. Their scientific expertise was complementary to ours. The fact that they were both extremely skeptical was an advantage, I thought. No matter how hard we might try, I did not believe Luca and I could trust ourselves to be completely objective.
The first principle is that you must not fool yourself and you are the easiest person to fool.
—Richard Feynman, “Cargo Cult Science,” 1974
My early mentor, Richard Feynman, delivered an elegant speech at my Caltech commencement ceremony about the dangers of a phenomenon known as confirmation bias. It is a well-known human frailty that has been studied for decades. People in every walk of life tend to ignore evidence that runs contrary to their preexisting beliefs and eagerly accept evidence that appears to support them. Feynman’s message was that the more you believe in something, the more vulnerable you are to making a mistake.
I have always strongly embraced this philosophy. Over time, I developed a tried-and-true solution to the problem—I always seek out people for my research teams whose role, in my mind, is to be the fiercest critic imaginable. My in-house critics must be harsher than anyone else who might challenge the work if it were ever published. I assign critics to the “red team” and advocates to the “blue team.” The goal is for the two teams to duke it out in a merciless, but friendly, competition until the scientific truth is revealed.
Lincoln and Glenn were both so negative at this point that they were perfect candidates for the red team. That was to be their implicit role, even though we never discussed it directly. Luca and I would represent the blue team and be primarily responsible for gathering the probative evidence.
The blue team advocates, Luca and I, immediately began holding daily meetings over the Internet to discuss our research efforts, which quickly became something of a roller-coaster ride. A thrill at one moment and a fright the next. Before too long, we found ourselves addicted to the adrenaline rush.
Luca suggested we engage in typewritten chats instead of verbal conversations, which was remarkably prescient. The written records would turn out to be a valuable resource during the tortuous course of our investigation. We often went back to them to check facts and refresh our memories.
Our intense daily chats inevitably turned into a rivalry: Which one of us could discover the most interesting item? We competed to find the best new scientific paper, best new Internet contact, best new hint about the origin of the Florence sample, and best new laboratory measurements of the remaining specks. On most days, Luca was the clear winner. But I enjoyed an upset victory every now and then.
Our first priority was to figure out how and when the sample labeled “Khatyrkite” made its way to Luca’s mineral museum.
Luca combed the museum archives and dug up correspondence dating back more than two decades. The letters revealed that his museum acquired the khatyrkite in 1990 as part of a larger purchase of 3,500 specimens. Curzio Cipriani, Luca’s predecessor, paid roughly $30,000 for the entire lot. It was amusing to learn that our now-precious khatyrkite once had a whopping street value of less than ten dollars.
According to the records, Cipriani had purchased the specimens from a private mineral collector in Amsterdam named Nico Koekkoek. It was promising information, but woefully incomplete. None of the old paperwork included any contact information.
* * *
AMSTERDAM, HOLLAND, FEBRUARY 2009: Luca and I jumped online and started plowing through Dutch telephone directories. We spotted numerous Koekkoeks, but no one named Nico. We found a number of other mineral dealers and bombarded them with emails, both in English and Dutch, pleading for help. Despite a month of concerted effort, we failed to unearth a single clue.
Without Nico Koekkoek, I wondered, how could we ever hope to establish the origin of the Florence sample?
It was a disappointing dead end, but Luca and I were already deeply engrossed in other aspects of the investigation. Time was so short that we had no choice but to pursue many different ideas simultaneously.
One of our biggest headaches was that Lincoln and Glenn kept insisting that the aluminum-containing metal alloys in the Florence sample were nothing but sla
g. Granted, khatyrkite and cupalite were listed in the International Mineralogical Association catalog of recognized minerals. But neither Lincoln nor Glenn trusted the analysis connected to those entries. Metallic aluminum without oxygen was the sticking point. Impossible, they both said dismissively.
Luca and I thought we could persuade them that the metal alloys were natural by finding another sample of khatyrkite in a different collection. We knew the source would have to be unimpeachable in order to convince them.
We began by checking prestigious museums with gigantic mineral collections, like the Smithsonian in Washington, D.C., and the American Museum of Natural History in New York City. No luck in either place, which was, frankly, unexpected and a bit concerning. Then we turned to museums with more modest collections, some of which had mineral catalogs we could review online. Once again, no luck, which was now becoming a bit more worrisome. We then began checking lesser collections in smaller museums, academic institutions, and individual collections around the world.
We reached out to international mineral dealers. Did any of them have khatyrkite, or had they ever sold khatyrkite to anyone? We checked Mindat.org, an excellent public database of minerals used by amateur and professional mineralogists. Did anyone on that website have any khatyrkite?
* * *
NORTHFIELD, MINNESOTA, MARCH 2009: The upshot of our exhaustive worldwide search was a grand total of four potential sources of khatyrkite. Three of the samples were in North America and Western Europe. A fourth, perhaps the most promising, was in St. Petersburg, Russia.
I was especially excited when Luca discovered that one of the samples was held in the mineral collection at Carleton College in Northfield, Minnesota. A sample kept at an academic institution is certain to be authentic, I thought. I was even more confident when I learned that Carleton’s leading geology professor, Cameron Davidson, was a Princeton graduate and one of Lincoln’s former students.
The Second Kind of Impossible Page 13