by Thomas Hager
But there was also a more fundamental problem: the way x-ray crystallographers approached their work. In order to be certain of their results, researchers—nearly always physicists in the 1920s—usually started from the supposition that any arrangement of atoms that matched the compound's chemical formula should be considered as a structural possibility. Finding the correct one was a process of eliminating those that did not strictly fit the x-ray and density data. This rigorous method gave a firm answer—you could be certain you had the right structure because you had eliminated, one by one, all other alternatives—but it was also something like trying to find a needle in a haystack by picking up each stalk of hay, holding it next to a picture of a needle, and tossing it aside if it did not match. The more complex the crystal, the bigger the haystack.
As the targets of x-ray crystallographers became larger, this work became harder, more time-consuming and repetitious, requiring the employment of teams of "human computers" to do the needed mathematical work. It was not to Pauling's taste, and he began looking for shortcuts. He had always read widely and critically; now he began to synthesize what he knew about chemistry and x-ray crystallography into a structural worldview, a sense of how atoms preferred to fit together. If he was reading an article on a crystal structure and the conclusion seemed reasonable, he added it to his mental library. But if something sounded wrong—bond angles awry or atoms misplaced—Pauling would fine-comb the work, sometimes refiguring structures from published x-ray photos, sometimes shooting and analyzing new ones. He often found that whatever struck him as wrong was, in fact, a mistake. But occasionally, and more excitingly, the thing that struck him as wrong might actually prove to be correct. Then it was his worldview that would have to change to incorporate the new information. This was how he learned. Through this process of analysis, incorporation of new data, and refinement of worldview, Pauling’s molecular structural sense over the years would become uncanny; he would be able to come up with plausible structures and discard the chaff almost instantly. Observers would later call this Pauling's "chemical intuition." The phrase is not quite right, though; it smacks too much of the emotional, the irrational. Pauling's ability was entirely rational, based on thousands of hours of careful reading, mentally pruning, sorting, and filing tens of thousands of chemical facts.
This deep understanding of chemical structure allowed Pauling to break away from the old ways of doing x-ray crystallography. "My attitude was, why shouldn't I use the understanding that I have developed of the nature of crystals in inorganic substances to proceed to predict their structures?" he asked. The physicists doing x-ray crystallography might have to consider each structure with the right number of atoms as a possibility because they did not know any better. Pauling, however, knew that most hypothetical structures would be unlikely for one reason or another on chemical grounds. The atoms would prefer to fit in a more limited number of ways. Before starting, he could eliminate most of the unreasonable ideas and get down to a small number of most likely candidates.
Prediction was the way to break out of the crystallographer's conundrum. But if accurate prediction was to be useful to other researchers, it required rules, sets of principles that could be applied to a number of cases to demonstrate why certain structures were likely or unlikely. By the late 1920s, Pauling knew that the same basic structural patterns were often repeated in different crystals. The repeated patterns happened for a reason; they had to be governed by rules. If he could find the rules, he could predict the structures of unknown crystals.
- - -
In his laboratory in Manchester, England, the British physicist William Lawrence "Willie" Bragg was thinking along the same lines.
Bragg was a complex fellow. Shy and old-fashioned, gentlemanly in the Victorian manner, fond of birds and gardening, Bragg seemed in many ways more like a country squire than a prominent scientist. Even "his outward appearance and behavior were rather like those of a prosperous farmer, dressed not quite in the latest fashion," a colleague noted.
He was also one of the brightest physicists in England and, beneath the outward reserve, gifted with a tough competitive streak. He was also under a great deal of pressure. He had achieved international fame early—perhaps too early—when, with his father, also William, he codeveloped the science of x-ray crystallography after Laue's initial discovery. The Bragg equation for the diffraction of x-rays by a crystal was the foundation upon which the science was built; the Braggs' textbook was the bible of the field. The work with his father had earned Willie Bragg a Nobel Prize at age twenty-five, making him the youngest man ever to win one. A rush of international fame and directorship of the prestigious Manchester laboratories followed.
He did not much like administration, but he did enjoy research. Bragg was gifted with an ability to cut away extraneous details and get to the heart of a scientific problem, and in the late 1920s he was working on a big one. Bragg's interest centered on a large and important family of minerals called the silicates, complex ionic crystals formed of silicon, oxygen, and various metal atoms. Silicates were among the most common minerals on earth—the family included everything from talc to topaz—and also the most complex. Solving their various structures would be a coup for any crystallographer. Bragg, like Pauling, was working on shortcuts and was very proud of an approach he developed in which he viewed silicate structures in terms of packing together spheres of different-sized ions, rather like packing together marbles in a jar. Bragg's work showed that ionic crystals in general kept their atoms as close together as possible, "close packing," he called it. In his scheme, the way in which larger ions like oxygen packed next to each other would determine the crystal's basic structure, and the smaller ones would tuck into the spaces in between. Using this approach, he was, he believed, on the way to making sense of all silicate structures.
As Pauling put it, "Bragg thought that that was his field." But he was in for a rude awakening. Pauling, too, was interested in silicates and intended to give Bragg some competition. Although the English researcher was only eleven years older, Pauling considered Bragg "a member of the older generation of scientists, who had blazed a trail that I was attempting to follow." And while Bragg's close-packing scheme was a good start, there were other approaches Pauling thought could bring one closer to a solution.
Like Bragg, Pauling thought of the ions in these types of crystals as spheres of distinct sizes, and he knew the sizes inside and out—after all, it had been Pauling who first developed a quantum-mechanics approach to determining ionic radii. But Bragg and Pauling differed in how they thought these spheres behaved. Whereas Bragg worked from the idea that ionic crystals had no molecular-type structure within them—he saw them instead as extended patterns of separate ions, of the sort he and his father had found in table salt—Pauling viewed silicates in terms of basic units. An atom of silicon, for instance, could behave much like carbon; it, too, formed a tetrahedron when binding to four oxygens. Bragg had shown that oxygens for their part often packed together in ionic crystals six at a time to form an octahedron. Tetrahedra and octahedra: Pauling started with these basic building-block shapes.
In puzzling this over, Pauling made a breakthrough. He ingeniously distilled what he knew about quantum mechanics, ionic sizes, published crystal structures, and the dictates of chemistry into a set of simple rules indicating which joining patterns were most likely. The most important of these, called the electrostatic valence rule, used what was known about one element's binding capacity for others, its valency, as the basis for a formula for determining how many blocks would likely meet at a given corner. A key was the realization that in these minerals a central element's valency could be distributed in fractions to ions of opposite sign. Other rules dealt with the sharing of faces and edges. Taking all of these rules together, and considering as well Bragg's ideas about close-packing, Pauling was able to outline a relatively simple step-by-step procedure for eliminating scores of unlikely crystal structures and predicting the most like
ly ones.
He first published his rules in late 1928 as a contribution to a set of papers written in honor of Sommerfeld's sixtieth birthday, a fitting tribute to the man who had taught him to use whatever was needed to get to a good solution. The next year, he put them forward in more detail in the JACS. They were soon known widely among crystallographers as "Pauling's rules." And they worked. In his first papers, Pauling used his rules to solve the structure of two complex silicate crystals, brookite and topaz. His rules helped extend the reach of x-ray crystallography to tougher structures than had previously been solvable.
But more than his rules were involved. In his silicate work, Pauling developed an entire approach to solving difficult x-ray crystallography patterns, a general method that he would use time and again over the next decades. It started by using known chemical principles to set up a set of structural rules. If the rules were stringent enough, many theoretical structures could be eliminated on the basis of chemical considerations alone, leaving only a few of the most plausible remaining. Pauling would then make models of those few to decide which was best. The models allowed him to view the structures in three dimensions, to see what worked and what did not, to fiddle and twist until things lined up right. If the models showed that certain structures packed their atoms too tightly or too loosely to seem reasonable, they, too, could be eliminated. Finally, just one would emerge as most likely. The properties of this hypothetical structure, including the x-ray diffraction pattern it would likely produce, could then be compared to those of the actual substance. If they matched, the structure was assumed to be correct.
Mustering everything he knew about chemistry and physics, and adding to it his new interest in model building, Pauling was able to leap to a solution where others were left mired in a swamp of confusing x-ray data. A few years later, Pauling was describing his approach to an acquaintance, Karl Darrow, who told him that it already had a name: the stochastic method. Darrow referred Pauling to a 1909 chemistry text in which the author talked about the long-disused Greek term that could be translated as "to divine the truth by conjecture." In a simple way, the stochastic approach could be seen as nothing more than an educated guess, a hypothesis like any other scientific hypothesis. Anybody could guess at a molecular structure, and while a comparison of the molecule's properties to those calculated for the hypothetical structure might eliminate the guessed-at structure as wrong, it was difficult to say that the hypothetical structure was rigorously correct because the comparison between the guess and reality would be based almost invariably on limited experimental data. But the way Pauling used it, the stochastic method was not a simple guessing game. You had to know enough about chemistry and crystallography to pare away all but a single structure. As he put it, "Agreement on a limited number of points cannot be accepted as verification of the hypothesis. In order for the stochastic method to be significant the principles used in formulating the hypothesis must be restrictive enough to make the hypothesis itself essentially unique; in other words, an investigator who makes use of this method should be allowed one guess."
Pauling had broken through a very complex problem using his stochastic method, and he would continue to employ it against even tougher puzzles during the next three decades. Sometimes his one guess would be wrong; far more often he would be right. This ability to "divine the truth by conjecture" would allow him to vault over his competition in solving the thorniest chemical problems. Eventually, it would bring him his greatest triumphs and earn him the reputation of a person who could almost magically dream up solutions where others had failed. But it was all the result of very hard work, deep chemical knowledge, and a willingness, a daring, to make that one guess.
- - -
Before the publication of his rules for solving complex ionic crystals, Pauling had been known as a promising young crystallographer. Afterward, he was propelled into the first rank. Lawrence Bragg, for one, was stunned when he read Pauling's work. This young American had appeared out of nowhere and beaten him at his own game. Some of the ideas Pauling put forward were similar to those Bragg had been using in his own attack on silicates, but much was new. The electrostatic valence rule especially was clear and very useful; years later Bragg would generously call it "the cardinal principle of mineral chemistry."
Pauling employed his rules with great success. In 1929 and 1930 he worked out the structure of mica, a silicate whose tendency to split into thin, flexible, transparent sheets Pauling discovered was due to a layered crystal structure with strong bonds in two directions and weak bonds in the third. He then compared mica to silicates that, while similar in chemical makeup, differed greatly in form. Talc, he found, also had a layered structure, but one that was held together so weakly in two directions that it crumbled instead of split. Another group of silicates called zeolites interested researchers because of their ability to absorb some gases, including water vapor, but not others. Pauling discovered that zeolites were honeycombed with passages so tiny that they formed molecular sieves, letting in only molecules small enough to squeeze through and keeping out others.
A Much-Wanted Man
Pauling's rules were important, but the young man was hungry for more. He read everything in physics and chemistry, tried to attend every Caltech seminar given in those fields and biology, was interested in everything and alive to any problems he might solve. In 1930 he turned back toward physics, clearing up a puzzle that had been posed by researchers at Berkeley. In theory, heat capacity—the amount of energy needed to raise the temperature of an object—should decline to nothing for substances at absolute zero. But the Berkeley group was surprised to find that their experimental work with molecular hydrogen pointed to a significantly higher value. Pauling explained it: All you have to do, he said in a paper in the Physical Review, was assume that hydrogen molecules could rotate while bound into crystals. His quantum-mechanical demonstration of the point was an advance not only in the study of the heat capacity of solids but also helped to explain things like the transition from one crystalline form to another. Scientific American spotlighted the work as one of two "standout" discoveries in fundamental chemistry for 1930. The other for that year was Pauling's and Bragg's general solution of the structure of silicates.
Pauling's mastery of quantum mechanics and its application to chemistry was now becoming well known in research circles, and job offers began coming in. During the late 1920s the best American universities were snapping up young researchers who knew the new physics, preferably one of the few U.S. students who had learned it on foreign fellowships, like Pauling, John Slater, Ed Condon, John Van Vleck, or Robert Mulliken, or, failing that, immigrant Europeans such as Samuel Goudsmit. Pauling, along with a dozen or so of his quantum-literate peers, suddenly found himself in a seller’s market.
G. N. Lewis told Pauling that he had visited Caltech in 1928—a special occasion, for he rarely traveled—intending to offer Pauling a job at Berkeley. He was waylaid by Noyes, however, who "forbade him to do so." Noyes did grant Pauling one term's leave in the spring of 1929 to serve as a visiting professor at Berkeley. While there, Pauling received an even greater honor, a call to Harvard. The great school's chemistry department was in the midst of being restructured after the death of its longtime leader, the first American to win the Nobel Prize in chemistry, Theodore Richards. The Richards chair had been offered to both Lewis and Richard Tolman, but both had decided against going east. By the time the offer reached Pauling, the job had been downsized to an associate professorship without mention of the chair. Still, it was enticing. Harvard was the oldest, most prestigious, best-endowed university in America; its physics and chemistry departments were renowned. "The prospect of becoming associated with the Chemistry Division of Harvard University appeals to me very much," Pauling quickly wrote back. However, his teaching duties prevented his coming east for a visit until at least May, he wrote, asking for a delay.
Pauling had seen enough of academia to know how the promotion game worked. Tolman,
for instance, leveraged his Harvard offer into thousands of dollars in research moneys, support for a faculty member he wanted to hire, and a promise that he would not have to take part in publicity activities. Pauling wrote Noyes and asked him frankly what he could expect in the way of a Caltech counteroffer. Noyes did not want to lose Pauling, either. At Noyes's and Millikan's urging, the Caltech executive council in March awarded Pauling a promotion the next fall to associate professor (after only two years as an assistant professor), substantial pay raises over the coming two years, support for a laboratory assistant, two more graduate students, and travel money for a European trip.
But Harvard remained in the hunt. There was talk of building new courses in crystal structure and chemical physics around Pauling, even creating a new department devoted to the young researcher's brand of what was now being called "quantum chemistry." Pauling remembered (although no corroborating documents seem to exist) that the Harvard offer was eventually raised to full professorship, successor to the Richards chair.
So Pauling visited, arriving in Cambridge for a week in early May. He was treated royally, staying in the home of organic chemist James Bryant Conant (who was soon to become president of Harvard), touring the new chemical laboratories, presenting seminars, and attending receptions. He was twenty-eight years old and flattered by the attention, but he also found things—some big, some little—he did not like. Whereas Caltech was becoming famous for allowing researchers a free hand to develop their own unique approaches to science, at Harvard, Pauling found, subdisciplines such as organic chemistry and physical chemistry had ossified into separate fiefdoms. There was a sense of backbiting and politicking and a hoarding of talent he did not like. A product of the egalitarian American West, Pauling also received his first taste of eastern class snobbery. "Here was a society where there were a lot of important people who were important just because of birth. They had money and stature not based on their own abilities," he remembered. "I thought I would be a sort of second-class citizen at Harvard."