Image scanned from a negative held in the Dunitz Papers, 1.012.9
Portrait of Dorothy Hodgkin by Bryan Organ, 1982.
Although J. Monteath Robertson had invented the method of isomorphous replacement, we never used it during my years at Glasgow. For the sort of problem I was working on, there was no opportunity to apply the heavy atom method — there were no heavy atoms in the compounds studied — nor to make use of the Patterson function. What we did learn was how to look at an X-ray diffraction diagram and see the main features of the molecular transform, an ability that proved to be very useful, both then and later. One day during my Glasgow work, I read Dorothy Crowfoot’s paper on the structure of cholesteryl iodide. (5) This was one of the first three-dimensional analyses of what was then regarded as a complex molecule — the penicillin work was not yet published. I wrote to Dorothy Hodgkin (neé Crowfoot) to ask if she would accept me as a post-doctoral research worker, and she agreed to do so. I had no money, but the Carnegie Trust for the Universities of Scotland, which had provided financial support to me as a research student in Glasgow, generously decided to continue and even to augment my stipend.
After my arrival in Oxford in late 1946, my first project was to try to determine the detailed crystal structure of a calciferol derivative. From earlier comparisons of unit cell dimensions of sterols and related compounds, it had been tentatively concluded that the calciferol molecule, in spite of its broken B ring, had more or less the same shape as a normal sterol, such as cholesterol. (6) I was sure this must be incorrect. As Dorothy herself was doubtful about the matter, we decided to try to obtain a suitable heavy atom derivative. Fortunately, we were able to persuade John (Kappa) Cornforth and his wife Rita to help us, but it took several months before they could come up with a suitable crystalline derivative, the 4-iodo-5-nítrobenzoate of calciferol. From a two-dimensional projection down the short unit cell axis, (7) it was clear that the molecule was indeed s-anti about the central bond of the broken B ring and not s-syn, as had been assumed in the early work. With more than 40 atoms (not counting hydrogens) in the molecule, this was at that time the most complex structure to have been determined at atomic resolution — a record that did not endure very long. I was able to do the calculations required for the two-dimensional projection by hand, with the help of Beevers-Lipson strips (8) — does anyone still remember what they were? — but the completion of the full three-dimensional analysis had to wait many years until appropriate computing facilities became available. (9)
During the waiting time for suitable crystals, we had a visit from a member of the Medical Research Council staff. It had been found that the drug stilbamidine, a 4,4'-diamídostilbene derivative, undergoes photochemical change to a more toxic substance when its solutions are exposed to light. The toxic product could be converted to a hydrocarbon, C28H24, melting temperature 163°C, identical to the product obtained many years earlier by irradiation of stilbene itself (10) and identified as 1,2,3,4-tetraphenylcyclobutane, together with a small amount of an isomer, melting temperature 149°C. The question of interest to the Medical Research Council was: Could we tell which stereoisomer was which? From the unit-cell dimensions and space group information, the higher melting compound was soon identified as the centrosymmetric isomer. As I had nothing better to do at the time, I practiced my skill in trial-and-error analysis by determining the crystal structure in projection down the short (5.77 Å) monoclinic axis (11) and then also tested my endurance by calculating lines and sections of the three-dimensional electron density distribution, based on visual estimates of all photographically recorded reflections within the CuKα sphere of reciprocal space. (12) From today's perspective, when the measurements could be made in a few hours and the calculations in a few seconds, it is hard to imagine how much drudgery was involved in such an exploit in those days, working with paper, pencil and Beevers-Lipson strips. Why on earth did I take it on, and why did I persist? No one was pushing me. Perhaps I merely wanted to show that I could do it.
Fame is the spur that the clear spirit doth raise
(That last infirmity of noble mind)
To scorn delights and live laborious days.
My tetraphenylcyclobutane work did not bring me fame but it did bring me to Caltech. Indirectly. From my results, the bond distances in the cyclobutane ring appeared to be longer than the standard carbon-carbon single-bond distance of 1.54 Å, while those in the phenyl groups were normal for benzene rings. However, according to the recently developed "bent bond" model, (13) bonds in small carbocyclic rings were expected to be slightly shorter than1.54 Å, as had been found, indeed, for cyclopropane and spiropentane from gas-phase electron diffraction. (14) Were the long bonds found in tetraphenylcyclobutane an intrinsic property of the cyclobutane ring? Or were they in some way connected with the presence of the four phenyl substituents? Or were they merely attributable to experimental error? When I discussed this problem with Verner Schomaker during his visit to Oxford in the early summer of 1948, we decided that the problem called for a gas-phase electron diffraction study of cyclobutane itself. I was interested in learning this technique, and, through Schomaker's intervention, Pauling offered me a research fellowship to come to Caltech.
At Oxford, I became friends with Gerhard Schmidt, who, although a few years older than I, was still working for his D. Phil. Schmidt, a Jewish emigré from Germany, had lost valuable years, as he been arrested in 1940 as an "enemy alien," sent to the internment camp in the Isle of Man and then deported to Australia. Eventually, as a result of intervention by Sir Robert Robinson, Professor of Organic Chemistry at Oxford University, he was able to resume his studies. Schmidt had learned his organic chemistry from Robinson and did his best to improve and extend my more limited and old-fashioned views on the subject. We often worked into the night. Gerhard went on to found that marvelous school of structural chemistry at the Weizmann Institute. He told me the following story, which deserves to be secured for posterity. In one of his Oxford lectures, Robinson had drawn on the blackboard an organic formula with a pentavalent carbon atom. After some hesitation as to whether he should interrupt his famous and reputedly irascible teacher, Gerhard tentatively raised his hand. Robinson stopped. "Well," he asked, with a disapproving stare. "Excuse me, Professor, but in the alkaloid structure on the left of the board you have drawn a pentavalent carbon atom". Robinson glared at Gerhard, turned to the board, erased the superfluous bond, glared again at Gerhard, and continued with the lecture. Twenty years later, during a visit to the Weizmann Institute, Sir Robert again wrote an incorrect formula in the course of a lecture. By this time, Gerhard was himself a well known scientist and had enough self-confidence to immediately draw the lecturer's attention to his error. Robinson glared at Gerhard, turned to the board, glared back at Gerhard, and exclaimed: "You again, Schmidt!"