by Thomas Hager
A. A. Noyes was sure that more insights would come. X-ray crystallography was invented by European physicists, but Noyes would bring it to America and make it a chemists' tool. Within three years of the Braggs' first published findings, Noyes was calling the field of x-ray crystallography "the most important one in physical chemistry today," and in 1916 he advised one of his MIT graduate students studying in Germany, C. Lalor Burdick, to stop by the Braggs' laboratory in England on his way back home to learn about their x-ray techniques. Burdick returned with enough knowledge to build America's first x-ray spectrometer at MIT, and Noyes had him build a second, improved version—"the best of its day," Burdick remembered—in Pasadena in 1917. Results began flowing quickly, and by the time Pauling arrived, x-ray crystallography was the single most important chemical research tool at Caltech, the source of fifteen of the chemistry division's first twenty published papers.
Noyes had a lot riding on the technique. Chemistry was the study of the behavior of molecules. Noyes was becoming increasingly convinced that behavior depended on structure. And now it was possible to "see" the structure of molecules. By assigning Pauling to Dickinson's laboratory, Noyes was pointing the promising student toward what he believed would be the future of chemistry.
- - -
Pauling threw himself into his laboratory work but soon found himself stymied. Noyes had suggested that Pauling try to solve the structure of lithium hydride, but after spending three weeks in October working on the problem, Pauling discovered that a team in Holland had beaten him to the structure. During the next weeks he tried a number of other compounds, making crystals by melting chemicals in an electric furnace and cooling them slowly, then cutting and mounting the good ones and taking them through an initial analysis to see if they were simple enough to solve. None of them were. (One he tried, sodium dicadmide, turned out to be one of the most complex inorganic molecules known, involving more than one thousand atoms. It would be thirty-five years before one of Pauling's colleagues solved its structure.) Pauling's frustration grew with each failure.
After two months of unsuccessful efforts with fifteen different substances, he was rescued by Dickinson. His mentor led him to the chemistry stockroom, grabbed a nodule of molybdenite—a shiny black mineral composed of molybdenum and sulfur—from the shelf, showed Pauling an innovative way of preparing thin slices mounted on glass slides, and walked him through the preparation of the Laue photographs. Why Dickinson picked molybdenite is a mystery. Perhaps he thought its relatively simple chemical formula, MoS. would translate into a simple crystal structure. Perhaps he was just lucky. In any case, the mounted slices were in good shape, the unit cell was small, and within a month Dickinson and Pauling had determined the structure—an interesting one, the first to be described that had six atoms of a nonmetal, sulfur, forming an equal-sided prism around a single atom of a metal, molybdenum.
Pauling was elated. "The achievement made a great impression on me. The process of structure determination involved a series of logical arguments, which were presented to me by Dickinson in a meticulous way, with emphasis on rigor," he later wrote. "I was pleased to learn that questions about the nature of the world could be answered by carefully planned and executed experiments." His pleasure was coupled with a deep sense of satisfaction—the feeling that comes from using skill and brains to bring to light one of nature's hidden things as fully and surely as the human mind is capable. He had made a discovery.
He was now a scientist.
Dickinson, of course, had analyzed crystals before, and satisfied that his student now knew the ropes, he went back to other work. Pauling assumed that the next step would be publishing his first scientific paper. But "I waited for a month and nothing happened," he remembered. So, on his own, he wrote up the results of the molybdenite work for publication and presented the paper to Dickinson.
Shortly afterward, Noyes called Pauling to his office. He sat the young man down and delicately turned the conversation to the subject of scientific attribution. Noyes noted that the molybdenite paper carried only Pauling's name; he was afraid, he said, that Pauling may have forgotten that Professor Dickinson had also been involved in the work. "This was, of course, a shock to me," Pauling said. "I realized that I had just ignored completely the efforts that he [Dickinson] had made and his activities in instructing me." The paper, "The Crystal Structure of Molybdenite," was revised and received for publication by the Journal of the American Chemical Society in April 1923. The authors were, in order, Roscoe G. Dickinson and Linus Pauling. "I think it was a good experience," said Pauling, "in that it pointed out to me how easy it is to underestimate the contributions that someone else has made."
After his shaky start, Pauling became a proficient crystallographer. Dickinson soon had so much faith in his student that when he went to Europe for a year in 1924 on a fellowship, he put Pauling in charge of the x-ray lab, an experience that strengthened Pauling's independence as a researcher. Pauling also assumed the role of teacher, introducing other students to the technique and overseeing their work. After his return, Dickinson gradually became interested in other fields of research, and Pauling inherited the mantle of Caltech's resident x-ray expert. Before receiving his doctorate, Pauling would publish, on his own and with others, an impressive total of six more crystal structures.
In addition to establishing his reputation as one of the nation's more promising young crystallographers, Pauling's early work provided him with a new way of looking at the world. He spent so much time analyzing the depth, height, and width of crystal units, learning everything he could about the sizes of atoms and the lengths of the bonds between them, that from then on he would see everything chemical in terms of structure. Molecules, he began to understand viscerally, were built out of atoms, just as buildings were built out of bricks and beams. There was nothing random about their structure. They were connected at certain angles to make certain shapes; this was architecture at the scale of hundred-millionths of a centimeter.
There was a purely aesthetic pleasure in discovering and describing these forms, and there was more. The ways in which molecules were built said a great deal about how they behaved. In molybdenite, for instance, Dickinson and Pauling found that the sulfur atoms were farther apart than the Braggs had measured in other minerals. In his first paper Pauling related those stretched sulfur-sulfur bonds to molybdenite's tendency to cleave easily. He also began reviewing all known crystal structures to try to find out why the Braggs' sulfur bonds were shorter than he and Dickinson had found and developed a sense that the length of chemical bonds varied, depending on type: Bonds in which the two atoms shared electrons equally—for which Langmuir in 1919 coined the term "covalent"—were generally shorter (and stronger) than ionic bonds, in which one atom held the electrons more closely than the other. Lewis, in his papers on the cubical atom, had theorized that a range of bond types were possible, depending on how equally the electrons were shared between two atoms. And now, as he reviewed the literature and solved more structures himself, Pauling found that, as Lewis had hypothesized, it wasn't an either-or case. Some bonds appeared to be intermediate between the two general types.
Physical chemists had, for the past forty years, tended to ignore chemical structure in their concentration on reactions; after all, before crystallography there had been no way to describe structure precisely. But that was changing. Now there was a growing realization that the properties of a substance were based on its structure.
- - -
Noyes was phasing out his teaching, concentrating on administration instead. During his first term at Caltech, Pauling took the last class Noyes ever taught: chemical thermodynamics, a swan song for the nineteenth-century approach to physical chemistry based on classical Newtonian physics.
It was also the only chemistry course Pauling would take as a graduate student. He had had plenty of chemistry at OAC but was starved for physics and math. The rest of his graduate career was filled with course after course in the two field
s.
Math took its place as a useful tool, a technique whose mastery was required for learning physics. And here his guide was yet another— and perhaps the most brilliant—of Noyes's MIT students, Richard Chace Tolman. After earning his doctorate in 1910, he had bounced around, from the University of Michigan to the University of Cincinnati, then four years with G. N. Lewis at Berkeley, two at the University of Illinois, and three in government service, searching for a position compatible with his wide interests. Tolman's interests were boundless: thermodynamics, statistical mechanics, kinetics, theoretical physical chemistry, even cosmology. His major interest outside of chemistry, however, was theoretical physics. He wrote the first book in English discussing Einstein's theory of special relativity and was one of a handful of American chemists—or physicists, for that matter—who understood the importance of the exciting new European advances in quantum physics. Noyes lured Tolman to Caltech the year before Pauling arrived and gave him an unusual dual title that reflected his interests: professor of physical chemistry and mathematical physics. Noyes called him "a great prize ... a very unusual combination of experimental and theoretical knowledge. . . . Indeed, with the possible exception of G. N. Lewis of Berkeley, there is no physical chemist in the country who ranks with him." Tolman was the cornerstone in Noyes's plan to create a chemistry program fully integrated with modern physics, and during Pauling's student years he carried the brunt of graduate teaching. While Noyes became increasingly concerned with the undergraduate curriculum at Caltech in the 1920s, Tolman epitomized the spirit of the graduate program: clear, critical, and focused on the cutting edge of research.
Students saw him as an impressive intellectual, erudite in any number of fields, and he looked the part, with his high forehead and neatly trimmed mustache. He was, like Noyes, a New Englander, born into a well-to-do Massachusetts family and educated at MIT. And if it wasn't for Noyes and a few other men he felt comfortable with, members of Pasadena's private clubs, he would never have stayed at Caltech. His leftist political sympathies—Pauling remembered students reciting a ditty about "Richard the Red and his brother Ed" (Edward Tolman was a professor at Berkeley)—were at odds with Southern California's conservatism, and his sense of culture was affronted by what he saw as the region's crass commercialism. He was a Brahmin who would always feel apart from Caltech. "I feel that there is not much pressure in New England to advertise and grab, and that decisions are not made on the basis of publicity," he wrote in the late 1920s while considering an offer from Harvard. "In coming to Harvard I should feel that I was coming back to my own people, whose ways and traditions I understand." Only Caltech's unmatchable salaries kept him in place.
Whatever they paid was worth it: Tolman was the most brilliant professor at Caltech. His lectures were masterpieces of organization; he would fill blackboards with figures before class and logically, compellingly, steer his students through the complexities of the new physics. He would engage his students directly, pausing in his lecture to pick out a fellow and quiz him about a point being discussed. It kept his students awake and forced them to think quickly. Pauling followed his first Tolman class with everything else the man taught, including relativity theory and an excellent course in statistical mechanics. He impressed Tolman, as he had nearly all of his teachers, and he was soon helping Tolman prepare manuscripts for publication.
But Tolman had his greatest effect on Pauling by introducing the young chemist to quantum theory.
Bohr's Atom
Before he came to Caltech, Pauling's only college work in physics had been a three-term introductory course at OAC specially oriented to the needs of chemical engineers. It barely touched on the newest ideas from Europe, where a small group of physicists led by a young Dane named Niels Bohr were questioning how the world is made.
Bohr wanted to understand the atom, and he came to England before the war to study with Ernest Rutherford, who had just put forward his dynamic solar-system atomic model in which electrons orbited the nucleus. Bohr's first great achievement was to link Rutherford's atom to other new findings in physics, most importantly the puzzling ability of elements to give off or take in energy in specific, discrete amounts—the small packets of energy that the German physicist Max Planck in 1901 had named quanta.
Planck's heresy was to propose that energy, like light and heat, was not continuous and smooth, as Newton had thought, but instead existed in discrete bits. Planck's ideas worked to explain some paradoxical phenomena like black-body radiation but flew in the face of other cherished physical ideas. Despite mounting evidence in favor of quanta—some of the most important marshaled by a young theorist named Albert Einstein—there was still no agreement among physicists that quanta were anything more than a convenient fiction.
Bohr, however, was ready to assert that quanta were not only real but essential to understanding the atom. In 1913 he published a paper proposing a model of the atom in which electrons circled the nucleus in flat, circular orbits, as Rutherford had visualized. But Bohr said that only certain-size orbits were possible and that these were restricted by quantum rules. Add a quantum of energy and the Bohr electron "jumped" from one orbit to a more energetic one; falling back to a more stable orbit, it released energy, sometimes as visible light.
Using an instrument called the spectroscope to closely study the light given off by luminous bodies, physicists had for decades been working on a mystery. When burned, different elements emitted light not evenly across the spectrum but only at specific wavelengths. The pattern of wavelengths given off by each element was unique, a sort of glowing fingerprint. And the patterns themselves were fascinating, tantalizing, regular enough to promise some underlying order, yet complex enough to defy explanation. When seen through a spectroscope, it seemed as though each element were displaying, in notes of light, the fingering positions for a unique chord. No one, however, had figured out why the atoms were playing those particular tunes.
Until Bohr. This was the most impressive achievement of the Bohr model of the atom: He correlated the fall of an electron between quantum orbits to the frequency of the bright lines seen with a spectroscope when looking at incandescent gases.
At least for hydrogen. Hydrogen is the simplest element, with only one electron, and Bohr's atomic model could be used to calculate the most obvious of hydrogen's spectral lines. But despite years of work he couldn't get his theory to do a good job of accounting for a number of the very fine lines in the spectrum of hydrogen, nor could he get it to work for any element more complex than hydrogen—and, of course, every other element was more complex than hydrogen.
But his success with the hydrogen spectrum led a number of other physicists—chief among them Laue's old boss, the German theorist Arnold Sommerfeld, head of the Institute of Theoretical Physics in Munich—to work on refining his model. During World War I, Sommerfeld, with the help of two assistants detained as enemy aliens (one of them the Russian-born future Caltech professor Paul Epstein), extended Bohr's atom to include, in addition to simple circular orbits, a complex of elliptical crossing and penetrating orbits in which the movement of electrons were corrected by applying Einstein's theory of relativity. The result was a more complex model of the atom, one that could be made to fit with most of the fine structure of the hydrogen spectrum and could plausibly be extended to many-electron atoms.
What became known as the Bohr-Sommerfeld model of the atom took center stage after the war, and this was what Pauling learned from Tolman. It was a dynamic model with fast-moving electrons, as opposed to the static, cubical atom of Lewis and Langmuir that Pauling had found so appealing at OAC. By the time he was at Caltech, the chemists' static atom was becoming an object of ridicule to physicists, its "loafer" electrons, as Caltech's head Robert Millikan jeered in a 1924 speech, "sitting around on dry goods boxes at every corner, ready to shake hands with, or hold on to, similar loafer electrons in other atoms." Physicists knew that electrons had to move somehow to keep from crashing into the nucleus.
At the same time, the physicists' dynamic atom was becoming more palatable to chemists. Sommerfeld's elongated, elliptical orbits gave the Bohr atom some of the three-dimensional specificity demanded by chemists: If one end of an ellipse was closer to the nucleus, the electron's orbit would extend from the nucleus like a fat arm, offering a visualizable way of making that handshake, in a particular way, with other atoms. In the early 1920s Bohr himself remade his flat orbits into three-dimensional shells, much more like Lewis's cubes. And there was movement toward compromise from the chemist's side as well. Lewis ventured the idea that his stationary electrons might instead represent the average positions of more mobile particles. By 1923, only seven years after proposing his cubical atom, Lewis was ready to accept the Bohr-Sommerfeld model—at least in the case of the hydrogen atom— even though there was still no explanation of how atoms could form bonds with one another.
At its most refined stage in the early 1920s, the Bohr-Sommerfeld model of the atom was a captivating work of the imagination. Drawings done to Bohr's specifications showed atoms like gorgeous geometric flowers petaled with intricate layers of interpenetrating electron orbits. For the few years Pauling was a graduate student at Caltech, those complex atomic constructions, with their pulsing, wheeling, harmonious electron orbits and the cordlike sets of spectral lines, appeared to represent, as Sommerfeld said, "the true music of the spheres."
But the melody was all wrong. How could electrons disappear from one orbit and reappear in another without existing anywhere in between—a "quantum leap" deemed impossible by classical physics? No one knew. How could negatively charged electrons circle the positively charged nucleus without losing energy, as Newton demands of moving charged bodies? As great a physicist as Millikan was reduced to answering, "God did not make electrons that way." Even with Sommerfeld's corrections, there were still spectral phenomena, especially in more complex atoms, that the model couldn't explain. There was a breakdown of theory. Classical physics simply didn't seem to work at the level of the atom, but Bohr's quantum theory didn't seem to work, either. As the physicist George Gamow wrote, "It looked for a while as though either the physicists or physics itself had become completely insane."
