[Introductory remarks by Crellin Pauling]

David Shoemaker: I went to Caltech in 1942 on graduation from Reed College, where my senior professor, Arthur Scott, introduced me to Linus Pauling through his 1940 edition of The Nature of the Chemical, Bond out of which he had taught a course that I took. I had written to Dr. Pauling asking for a job for the summer of 1942 since I was vulnerable to the draft, so he put me to work with Professor Carl Niemann on an attempt to create antibodies to histamine. Pauling had postulated that, if a protein molecule such as gamma globulin were partially denatured by mild heating in the presence of a potential hapten such as histamine, the partially unfolded protein molecule would re-fold in the presence of the hapten, assuming a surface structure complementary to that of the hapten, and thus be a potential antibody. In my novice hands, at least, it did not work, and I never did find out how close it was to working, but it was my first exposure to the bold inventiveness of Linus Pauling.

One year of conventional graduate course work was followed by three years of war work, the first of which was under Reuben Wood on the Pauling Oxygen Meter, another example of Pauling's inventiveness, in which the paramagnetic susceptibility of an atmospheric sample was measured by the deflective force on a lightly suspended test body in an inhomogeneous magnetic field as a measure of the content of paramagnetic oxygen in the sample. Then I joined Pauling's rocket propellant group as editor of technical reports. My job was not only to correct spelling and grammar but to make the reports look like they were written by Dr. Pauling himself a frightening challenge to a novice like me. This got me into trouble with other workers on the project who felt, perhaps with some justice, that their reports did not need to be corrected by the likes of me. I suspect that at least one of this afternoon's speakers was in that category. I will say that if any one person should be credited with teaching me to spell, it was Linus Pauling.

Bill Lipscomb and Ken Hedberg were contemporary with me as graduate students. Bill taught me a lot about growing crystals, and launched me on a small crystal structure investigation which I never finished. Ken was my office mate in 58 Crellin. He never tires to this day in reminding me of my incessant cigarette smoking, using as an ashtray a shell casing strapped to my desk, acting as a sort of chimney for cigarette butts that continued to smolder. I happened to be present in that office when Linus Pauling picked up Ken's key chain ornament with a photograph of a beautiful naked girl standing on a rock in a mountain stream and identified the rock as basalt.

After the war I decided to do crystal structure work on amino acids under Dr. Robert Corey. Dr. Verner Schomaker took a strong interest in my work and together he and I computed the first ever three-dimensional Patterson Function on an IBM punched-card tabulating machine, and with it determined the structure of the amino acid threonine. Meanwhile, Dr. Pauling had been working on the development of valence theory as applied to metals and alloys, and I developed an interest in that subject. So when Dr. Pauling asked me if I would like to spend a year abroad on a Guggenheim Fellowship, I chose metal structure theory as my research topic. In due course the fellowship was awarded, and I received a letter from Niels Bohr inviting me to spend my fellowship year at the Institute for Theoretical Physics in Copenhagen. During that year I had little occasion to be homesick, for my good friends Verner and Judy Schomaker were also there, and there was never any difficulty finding a fourth for bridge. Moreover, I spent a sizeable fraction of the year at Oxford University with guess who Linus Pauling, who was there to give some lectures, plus a couple of weeks in Paris with the Schomakers and the Paulings to attend a couple of conferences on valence theory.

I arrived in Oxford bearing a dozen fresh eggs and a large wheel of ripe Danish brie cheese, having heard that things were tough in England after the war. I learned that Linus Pauling had difficulty finding enough nutritious food, so I gave him my eggs. I was invited to the Paulings for dinner one day, so I took the cheese along. After dinner we sat by the fire, popped chestnuts, and consumed the cheese in its entirety. A few days later I visited Pauling in the apartment in Balliol College that he was using for his scholarly work. He had an electric space heater turned on its side, and on it was a pot of boiling water. In the pot was one of the eggs I had given him. In that room I saw history being made. With a pair of scissors he was cutting cardboard to make models of planar amide groups and taping them together to form a helix. Thus, in the early spring of 1948, was born the α-helix.¹

In Paris I learned that the Paulings were -- among other things -- broad minded; they did not object when I invited their sixteen-year-old daughter, Linda, and their youngest son Crellin, to go with me to that most famous burlesque show, the Folies Bergère. I also enjoyed the fireworks when Dr. Pauling was debating metal valence theory with band theory advocates such as John Slater and Neville Mott.

During those days most physicists working on metals and alloys tended to prefer the band theory approach, while Pauling continued to prefer the valence bond approach, modified to account for the special properties of metals. For example, where physicists tended to assign valences of one or two to copper, silver, and gold, and zero (!) to the transition metals, Pauling assigned 5.44 to copper, and so forth, and 5.78 or 6 to the transition metals, to take account of the stronger bonding evidenced by the densities which were much larger than those of sodium and calcium. This did not go down at all well with the physicists, and I must admit that I had to swallow rather hard at times. As paper after paper came from Pauling, elaborating his theory with modification after modification, I began to feel that his theory was losing some of its predictive power. Moreover, in Copenhagen I was surrounded by physicists, including many visitors from Britain and Europe. As a result, my one published paper from Copenhagen was a band theory study. That paper may have had something to do with my getting a job in 1951 as Assistant Professor at MIT, for among my interviewers for that job was Professor John Slater of the MIT Physics Department, who seemed to be impressed by that paper. However, I did not develop any great confidence in myself as a theoretician, and I was happy to accept Pauling's offer of a position as Senior Research Fellow at Caltech, a post that I held for three years, doing X-ray structure work on alloys. During some of that time my Research Assistant was Linda Pauling.

During those years it seemed as though almost every day there was a memo in my mailbox from Linus Pauling suggesting a structure for some alloy or other. Many of those were for the sigma phase, a phase first found in the iron-chromium system where it was a nuisance in stainless steels. Since the grain size was very small, the diffraction work was done by the powder method. I made several ball-and-stick models, trying unsuccessfully to verify Pauling's proposals. Finally Gunnar Bergman, my first graduate student, and I found in a powder sample a very few grains that we thought might be single crystals large enough (100 microns) to examine by single crystal methods. We were successful in determining the structure. Pauling was in Chicago, so we sent the results to him by telegram, and received a congratulatory message from him in return. Much of my subsequent career and that of my wife Clara stems from that structure; together we determined the structures of about a couple of dozen alloys belonging to the family which we called "sigma phase related" at first; later we called the alloys of the family "tetrahedrally close-packed" because all of the interstices were tetrahedral. Over the years we were frequently in contact with Pauling, who continued to have an interest in our work.

One day in 1960 the telephone rang in our house in Lexington, Massachusetts, and it was Linus Pauling. He had a problem. He had accepted an invitation to participate in a debate with John Slater and Neville Mott on bonding in metals during the 1960 crystallography congress in Cambridge, England, but the U.S. Passport Office, at the behest of Pauling's critics in Congress (who held him in suspicion of disloyal conduct during the McCarthy years), had withdrawn his passport. Would I take his place in the debate? As my bowels were turning to water, I heard myself utter a feeble assent. I already said that I had some difficulty with his valence bond theory as applied to metals and alloys so I had little confidence in how well I would perform. The debate took place, and I escaped with my life, in part because the physicists were easy on me and in part because, unbeknownst to me, Pauling had asked Bob Rundle to back me up. This he did, and he gave a very good talk, showing that he understood and appreciated the subject matter better than I did.

The last decade of our relationship with Linus Pauling was the era of quasicrystals, those strange materials which seemed to violate well-established crystallographic principles by possessing five-fold symmetry in single crystal electron diffraction patterns. Indeed, the materials seemed to possess in most cases the full symmetry of the icosahedron. One day Pauling showed up at our house in Corvallis, and as we were drinking tea on our patio he astounded us by expressing his strong disbelief in the quasicrystal hypothesis. The so-called quasicrystals, he said, were really multiple twins of a probably cubic crystal, with a twinning angle of 72 degrees, the supplement of which, 108 degrees, was close to the tetrahedral angle of 109.5 degrees. Shortly afterwards he voiced this hypothesis at an American Crystallographic Association meeting, and it was accepted overwhelmingly. What stimulated his thinking was a statement by Shechtman that the X-ray powder pattern obtained from this material "could not be indexed by any Bravais lattice," "a statement I knew to be wrong." Most of the materials under discussion had approximate formulae MnAl₄-MnAl₆. For the face-centered cubic structure that Pauling proposed to be twinned, he borrowed from the gas hydrate structures investigated by Claussen, and Sten Samson's structure of NaCd₂. He arrived at a unit cell length a₀ of 26.73 Å. The predicted powder pattern seemed to fit Shechtman's data except for a scale factor which Pauling thought was 15 percent off, but Shechtman and Bendersky succeeded in persuading Pauling that the error was in his unit cell, so Pauling reluctantly reduced a₀ to 23.36, corresponding to 820 atoms per unit cell.

Meanwhile I programmed my personal computer to calculate a powder diagrams for various values of the unit cell length, and to plot the deviation of observed values of the reciprocal spacing against the unit cell length over a wide range of values. I was pleased to inform Dr. Pauling that I got a very close fit to his value of 26.73 Å. Soon afterwards Pauling had to change his value to 23.36 Å, and I was unable to get any support for that value from my computer. Pauling attempted to find support for his twinned hypothesis in the electron diffraction patterns, but I am obliged to say that our own examination of these patterns left us skeptical of his interpretations.

More recently Clara and I have come to believe that the quasicrystals are very closely related to our tetrahedrally close-packed structures, or to a closely related family of structures containing aluminum, possessing in common with them the same kinds of coordination polyhedra, especially icosahedra. Clara determined the structure of a so-called mu phase MnAl₄ with hexagonal structure and a very large unit cell. This she examined closely in various orientations in comparison with features of a postulated quasicrystal with decagonal symmetry, and found striking similarities. These results were sent to Linus Pauling, who expressed his interest in his reply.

Meanwhile, new quasicrystal phases continue to be found in many systems, including some that appear to be thermodynamically stable. Work on icosahedral and decagonal phases has reached the point of actually assigning atomic positions in those phases. The reality of these phases can no longer be seriously doubted. However, the structures derived or postulated for these materials were not without the mark of Linus Pauling, especially as regards the linkage and clustering of icosahedra, in particular the 104-atom complex found in Mg₃₂(Al,Zn)₄₉, the structure of which was derived by Pauling by the stochastic method.

In sum, Linus Pauling is the most remarkable person I have ever known, for his intellectual brilliance and for his bold willingness to risk being wrong. His strategy of generating lots of ideas and throwing out the bad ones was overall extremely successful, and the bad ones that did get through have not significantly detracted from his genius. In addition, he had a wonderful sense of humor, and a warm regard for his students and colleagues. We will all miss him.