3—
Business

The discovery of Na²⁴ brought Lawrence into a new relationship with patents. In the case of the cyclotron, he had been persuaded to seek protection in order to block a possible monopoly by Raytheon or other firms. In the case of artificial radioelements, he participated in an effort to secure a monopoly position in a future radio-pharmaceutical industry. Here he and the Research Corporation were stymied by the difficulties in their case and by the practices of the U.S. Patent Office. A few European physicists did manage to secure patents on nuclear processes. The competitive environment influenced the direction of the work undertaken at the Laboratory, but did not restrict its openness to visitors or lessen its generosity to colleagues.

A Corner on the Market

Early in January 1932 the Research Corporation's lawyer, A.P. Knight, completed a draft application for a patent on the cyclotron, an instrument to produce high-speed ions by successive impulses "in a compact or relatively small apparatus." The ions, according to Knight, might be "utilized in any suitable manner, for example, for application to the disintegration or synthesis of atoms, or for general investigations of atomic structure, or for therapeutic investigations or applications." (Here the lawyer foresaw applications that Lawrence, still mired in machine design, probably had not; lawyers might be useful adjuncts to research teams.) Knight claimed injecting, accelerating, focusing, deflecting, and extracting the ions as patentable. The patent examiner rejected them all, as was the custom, enforced greater precision in language, and allowed the claims in September 1933.^[113] The patent was granted in February 1934.

Shortly after the Research Corporation received Lawrence's assignment of rights in the cyclotron, its leaders visited Berkeley

to see the cyclotron in action and to encourage its inventor to stay alert to patentable material developed in the Laboratory. Lawrence became almost an agent for the Research Corporation. In September 1933 he proposed patenting the water-cooled anode at the tip of the coil in Sloan's x-ray machine; in 1934 he wanted to patent a cheap cloud chamber for lecture use; in 1938 he brought Beams' ultracentrifuge to the Research Corporation; in 1939/40 it was the turn of a new oscillator developed by Sloan and Lauriston Marshall, and in 1941 of an application to industrial radiography by a Berkeley engineer. Lawrence wrote the engineer, for whom he interceded with the Research Corporation, without a hint of the old physicist's ethos: "If you are going after any patents along this line, you have my blessing."^[114]

The apparently fallow years in the preceding recital were, in fact, Lawrence's busiest and most vexatious time with patent affairs. A week after he had identified Na²⁴ , he recommended his discovery to Knight and Poillon as "almost ideal for biological work," a novelty that "might ultimately supersede radium in usefulness." He added that by running fast enough, Knight could get an application to the Patent Office before news of Lawrence's type of radiosodium appeared in print. Poillon was ecstatic: the discovery fulfilled half the prophecy he had made six months earlier to the president of the Chemical Foundation, that Lawrence's results qualified him for a quick Nobel prize and that his further work would help bring about "the production of synthetic radium." Poillon ordered Knight to run.^[115] Knight had the application in hand a week later; it emphasized that in Lawrence's process the radioactive material was chemically identical with the target and, moreover, that the chemical sodium does not harm the human body. The emphasis on the chemical identity of target and

product aimed to elude Fermi's patent on Na²⁴ made by (n,a ) on aluminum. This was perfectly proper, since patent law made a clear distinction between process and product: controlling a product does not mean controlling all ways to make it, and vice versa.^[116]

The patent examiner observed that the novelty claimed was production via (d,p) and that Crane and Lauritsen had priority of publication of results obtained in that manner. "There is no invention in utilizing the well known effects of deuteron bombardment upon any particular light metal," he held, and rejected all Lawrence's claims.^[117] Although Knight insisted that the work of Crane and Lauritsen had nothing to do with Na²⁴ , and although Lawrence declared that excitation by (d,p) was first observed at Berkeley, the examiner held firm. Meanwhile the Laboratory had found that (d,p) could make another important radiosubstance, P³² , which Lawrence urged Knight to try to patent together with the (d,p) process.^[118] To this last proposal the examiner returned a crippling objection: the (d,p) process as described by Lawrence, Lewis, and Livingston in 1933, not the application of the process to sodium in 1934, should be the precedent; and, if Lawrence could not show that at that time he had the idea of activating the substances he bombarded, the game was up.^[119] (A patent application could not then be filed on the strength of the paper of 1933, since filing had to occur within two years of publication.) This objection had at least this merit: it revealed how far the various parties were prepared to go to secure control of Lawrence's artificial radium.

The examiner held that since Na²⁴ and P³² were produced by (d,p) in the experiments of Lawrence, Lewis, and Livington, they had been discovered then even though no one knew it. That construction of the legal mind nonplussed Lawrence. He proposed to

Knight that they drop the attempt to patent radioelements. After all, he said, we have the cyclotron, "which I am sure will always be the apparatus that produces the radioactive substances." Lawrence thus reversed the estimate he had made during the patenting of the cyclotron. Then, while Poillon's Knight duelled with the examiner, Lawrence had doubted that the cyclotron had any commerical possibilities in the near term; and, when the duel concluded happily, he had questioned whether the Research Corporation should go to the additional expense of securing a Canadian patent. "The patent application on the cyclotron is very much of a gamble. It may never be of great worth," he had written Knight. "And yet," he added, "developments may come which would make it of tremendous value."^[120] The developments that revalued the cyclotron were, of course, the discoveries of artificial radioactivity and neutron excitation.

Poillon declined to hide behind the cyclotron. He appealed to Lawrence's patriotism: if the generous Research Corporation were to withdraw, grasping Caltech would rush into the vacuum. "I know how repugnant it is for any right-thinking scientist to become embroiled in a discussion concerning priority of discovery. . . . However, California Technology is quite a 'powerful Katinka' and is out for both intellectual recognition and financial return whenever proper and possible. . . . Under these conditions might it not be possible to straighten up a little bit in your claims for patent priority?" Lawrence could scarcely deny this appeal from his benefactor for help against his rival. "It is entirely proper," he replied, "for us to look out for the commercial aspects of our work, if this can be done in a dignified and proper way." He had not counselled withdrawal from distaste for battling Caltech, but from conviction that the patented cyclotron was protection enough.^[121]

Knight's strategy was to obtain an affidavit from Lawrence that he had had the idea of "irradiation" by deuterons before June 1933. That might allow patenting of (d,p); it might also require

Lawrence to claim discovery of artificial radioactivity. Knight's Washington correspondents arranged a meeting between Lawrence and the examiner, which they judged to be encouraging. Lawrence abandoned his application for radiosodium, substituted a claim for the (d,p) process, and swore that he had ordered the experiments of 1933.^[122] The examiner rejected all claims based on the experiments and the affidavit and found a new objection: Cockcroft and Walton must have had some deuterons among their hydrogen ions; they therefore must have made something radioactive by (d,p); and consequently, by patent logic, they discovered without knowing it what Lawrence claimed as his own. The only way out seemed to be an affidavit from Lawrence's colleagues that he suggested the original bombardments with deuterons in order to search for radioactivity. This Lawrence was reluctant to seek—his colleagues might not like his running away with a patent on their joint work—but he would cooperate if necessary. He supplied a second affidavit, which the examiner again rejected as insufficient, since it did not declare (what would have been perjury) that Lawrence examined the product of the irradiation for radioactivity. Knight's correspondents judged that a sufficiently strong affidavit would win the day, but had the sense to doubt "whether upon the actual facts of the situation a fully satisfactory affidavit can be furnished."^[123]

Lawrence had gone at least as far as he could. He wrote Knight: "The more I think about the matter the less enthusiasm I have for further endeavors to patent the process for producing the artificial radioactive substances." He retreated to his old position, much stronger in 1939 than it had been in 1935: "I feel that the cyclotron affords the only means of producing the radioactive materials in appreciable quantities; therefore with the cyclotron protected we have essential control of the matter." This time Poillon concurred, observing that the Research Corporation controlled

not only the cyclotron, but essential features of the Van de Graaff generator too.^[124]

Poillon had hoped to create and control a new industry of radio pharmaceuticals. Hence in the mid 1930s the Research Corporation invested redundantly in cyclotrons on the same principle that had built its cartel in the precipitation business: by expecting or requiring its grantees to assign improvements in the art to the Corporation. The theory is clear from the justification of a grant to Columbia to "enlarge the cyclotron . . . so that its field of application may be extended, and the equipment thus be made more effective for the preparation of artificial radioactive elements." In service of the same program, the Research Corporation supported work on the separation and application of biologically interesting stable isotopes, like C¹³ and O¹⁷ , at Columbia, and the ultracentrifuges of Jesse Beams at Virginia and of J.W. McBain at Stanford.^[125] The purpose was highminded. As Poillon put it, "we do not in any way want to prevent scientists from having the free use of any discoveries that are made but if we can assess industry a reasonable sum, we will have just that much more to give to scientific research."^[126] The methods, however, were those of the entrepreneur and the patent lawyer.

In 1940 it appeared likely that the Research Corporation would enjoy large royalties from its cyclotron patent. Several corporations, notably Westinghouse and American Cyanimid, deliberated building production cyclotrons for profit.^[127] Radiophosphorus had shown promise in the treatment of certain sorts of leukemia. The annual cost of treating all Americans so afflicted was reckoned at between $200,000 and $500,000. Since the university cyclotrons in operation in 1940 could not supply the demand, let alone the requirements for other therapies and applications, commercial

production of radioisotopes under patents held by the Research Corporation seemed imminent and humanitarian. American Cyanimid hesitated only over the worry that universities might undersell commercial radioelements if they started to charge for the products of their cyclotrons. Both Poillon and Lawrence reassured Cyanimid that the demand would be so large, and the output of university cyclotrons available for therapy so small, that the commercial market would not be affected by the policies of university laboratories. In preparation for a windfall, the Research Corporation pushed Lawrence to patent certain cyclotron improvements.^[128] The war ended these initiatives in two ways: by providing other lines of work for the interested parties and by creating, in the atomic pile, a much more efficient engine for the production of radioisotopes than the cyclotron. After the war, the Research Corporation wrote to all cyclotron laboratories to grant royalty-free licenses "for educational, scientific, experimental and research purposes." That amazed many. As the director of the Biochemical Research Foundation (Bartol) wrote in acknowledgement of this largesse: "I never knew there was such a patent."^[129]

The Research Corporation had not cared to exercise its rights when the primary consumers of artificial radioelements were research teams in universities and hospitals. And, as Lewis had done with heavy water, Lawrence distributed the fruits of the cyclotron gratis throughout the world. To be sure, the product spread the fame of the machine that produced it; but the Laboratory made its gift in a true spirit of scientific cooperation. Lawrence had several reasons for not charging even the cost of production: he thereby retained the right to support only projects he thought worthwhile; he had to avoid giving the men in the Laboratory the impression that they were cogs in a business; and he wanted to repay in some measure the support he had received

from charitable foundations and public bodies.^[130] The industry Poillon had envisaged did develop, after his patent expired. The first commercial cyclotrons for radioisotope production were made by the Collins Radio Company in the 1950s. In 1957 the former chief engineer at the Laboratory, William Brobeck, marketed cyclotrons for neutron therapy. By 1970 the annual sales of cyclotron-produced radioisotopes exceeded $3 million. It was but a small part of a big business—some $50 million a year—in radioisotopes for research, diagnosis, and therapy.^[131] Poillon had the right idea but the wrong machine.

Other Players

Early in 1935 the chief Italian journal of physics, Nuovo Cimento , pointed out that Lawrence was behind Fermi's group in the discovery of Na²⁴ . The Research Corporation likewise came late in the effort to patent it. On October 26, 1934, Fermi's group obtained an Italian patent covering activation by the absorption of fast or slowed neutrons and the products of the process, including radiosodium, as well. This violation of the physicist's ethos originated not with Fermi but with his patron O.M. Corbino, who had close ties to what high-tech industry then existed in Italy. "Age gave him wisdom," Mrs. Fermi writes, "[and] the boys were used to following his advice."^[132]

But they, too, had been anticipated by "the inventor of all things."^[133] In March 1934, a month or so after learning about artificial radioactivity and before Fermi's group had demonstrated the efficacy of neutrons, Szilard applied for a British patent on the "transmutation of chemical elements." His "invention," which he never reduced to art, had three parts: generation of neutrons to provoke reactions; separation of radioisotopes produced by the

(n,g ) process; and utilization of the heat liberated in the transmutation. Szilard's eccentric genius is displayed to full advantage in his method of obtaining neutrons. He planned to use deuterons accelerated by a high-tension device to create neutrons in collisions with light nuclei like beryllium or deuterium (he had made good use of the indications then accumulating of the d-d reaction that had misled Lawrence); he also proposed getting his neutrons from light, via (g ,n), and sketched an apparatus for making and absorbing photoneutrons.

The productivity of the transmutation evidently depends upon the number of neutrons at work. Szilard observed that if there exists a nucleus that when struck by a neutron liberates another without capturing the first, a very rapid buildup of a free neutronic population might occur. Mixing these hypothetical neutron multipliers with the material to be transmuted would increase the efficiency of transmutation; and a large enough sample of the material capable of sustaining the chain reaction (n,2n) would make a fine explosive. For the rest, Szilard proposed to separate isotopes made by (n,g ), which are chemically identical to their parents, by exploiting a process he did take the trouble to test.^[134] His scheme as of June 1934, including provision for extracting power, appears in figure 4.2.^[135]

Szilard considered assigning his first British patent on isotope and energy production to the Research Corporation in return for a grant for three years to continue research on the subject.^[136] Instead, he licensed it to a relative of Brasch's, a Havana importer named Isbert Adam, in return for $15,000 in research support. (Szilard thereby made more money from radioactivity than the Research Corporation, even after subtracting the $7,000 he subsequently repaid Adam to reacquire the patents in 1943.)^[137] In March 1936, when Szilard obtained a second patent on chain

[Full Size]

Fig. 4.2
Szilard's cornucopia of radioelements. Deuterons from the source 1 make
fast neutrons from the beryllium target 28, which spread through a sphere
3 composed of all elements that might multiply neutrons and that might
develop energy on absorbing neutrons. The tubes 107, 110, 111 contain a
coolant that delivers the heat of the reaction to an engine not shown.
Szilard, CW, 1 , 650.

reactions, he had reason to believe that multiplication of neutrons was possible. In his applications of 1934, he had imagined three sorts of interactions: a neutron multiplication (n,2n), a neutron conversion (n,2n ), and a neutron reduction (2n ,n), where 2n is a hypothetical heavy neutron with twice the mass of an ordinary one. In his definitive specification of June 1934, the basis of his patent of 1936, Szilard made the success of the chain reaction depend upon the existence of heavy neutrons, and offered indium, which in his experiments suffered an (n,4n ) reaction, as an element of the conversion type.^[138]

He mobilized fellow Hungarian refugees Eugene Wigner and Michael Polanyi to procure the material needed to initiate a chain reaction. He acquired a cylinder of beryllium, and access to a big radium bomb in a London hospital, to make photoneutrons. He convinced himself that indium could give out at least one double neutron via (n,2n) or (n,2n ).^[139] A trip to the United States provided leisure to weigh the whale he had hooked; and in March 1935 Szilard filed for an American patent on a large-scale transmutation process, similar to the earlier schemes, but with uranium and water (to slow the neutrons) as the neutron multiplier. All this was before the discovery of fission.^[140] The responsibility for the chain reaction grew too heavy for Szilard to carry and in order to keep it secret he offered to assign the patent detailing it (his second British patent on transmutation) to the War Office. The official who examined the gift could see no value in it. Szilard had better luck with the Admiralty, which accepted his assignment in March 1936.^[141]

As Szilard explained his actions to Fermi and to his British colleagues, he had never considered the patents to be his private property. He proposed to Fermi that they share responsibility for

controlling a fund secured by the promise of their patents. "It must be awkward for any scientist to have a personal interest from such patents," Szilard wrote, "while other scientists, who also could have taken out such patents, refrain from doing so." As for the research the fund might support, Szilard did not see the wisdom of the course of the Research Corporation and its Berkeley client; "I personally do not think very much of producing radioactive elements for medical purposes and I should not like to be responsible for inducing manufacturers to embark upon such an enterprise at present."^[142] (In that he was not entirely free from duplicity, since he wrote by the same mail to his patron Adam that the first priority was a systematic search for long-lived elements suitable for medical purposes). Nor did Szilard think much of the cyclotron. As he wrote to encourage Adam: "The artificial production of radioactive isotopes in California that you mention depends on a principle different [from mine], which I think is not susceptible of development and will have scarcely any commercial importance."^[143]

For development of his more promising scheme, Szilard thought that he could do with perhaps 1,000 pounds sterling a year and, if Fermi came in, 5,000 pounds for three years, less than a sixth of Lawrence's rate of consumption.^[144] Still the sum was not easy to raise. The agent of the Italian group, G.M. Giannini, agreed that the combined patents would make a nice set and professed an interest in cooperation; Segrè liked the idea of capitalizing the patents for a research fund, and for the researchers, "which would also indirectly advance science." But nothing came of it. Szilard continued to consider himself a disinterested broker.^[145] The Italians preferred to make money and hoped for a

time to set up an industrial concern; they turned down an option on the patents offered by Metropolitan-Vickers on Allibone's recommendation and ended by selling their European rights for $3,000 to Philips of Eindhoven, which had set up a nuclear section partly as a result of a visit from Segrè.^[146] They failed to interest any American corporation in their radioactive technique. Still, they obtained more than they dreamt of in the 1930s when, after much haggling, the U.S. government, which had exploited their technique during the war, paid them $400,000 as "just compensation."^[147] As for Szilard, he felt obliged to explain his altruistic policies to the major British physicists, and, when Fermi's group went its own way, to license Adam.^[148]

There is another round to the story. In 1938 Szilard removed his headquarters from England to New York City, where he rightly expected to find greater scope for his schemes. He now concentrated on improving the neutron source. From data on neutron yields provided in papers from Berkeley and Rome, Szilard calculated that the energy that could be stored in transmuted radioactive atoms might be one hundred times the energy of the bombardment required to make the neutrons to make the transformations. Economics had given its blessing; only a little ingenuity was required to make a nuclear-powered airplane. "Perhaps we ought to think of new methods for producing really strong neutron beams." Szilard had proposed to Brasch to scale up a hightension machine to slam electrons into metal walls at 10 MeV. The resulting x rays would sire neutrons in profusion from a beryllium target.^[149]

For money Szilard appealed to Lewis Strauss, a Wall Street financier with an interest in radiation therapy for cancer, who thus entered on his controversial career in atomic energy. As a light inducement to cooperation, Szilard offered to give to any taxexempt nonprofit corporation Strauss might wish to set up for producing radioelements a nonexclusive license to exploit whatever of his rights he had not sold to Adam. Strauss tried to enlist Westinghouse, General Motors, and General Electric; but all that came of it was an introduction to another supplicant, "whose friendship for the following twenty years was one of the finest experiences of my [Strauss's] life." That was Lawrence, who this time got nothing from Strauss. The future friend had decided to support Brasch and Szilard, if only a suitable place for Brasch's experiments could be found. One was. Toward the end of 1938 the ever-acquisitive Millikan offered space at Caltech, on the understanding that it would cost him nothing.^[150]

Brasch intended to reach for 15 MV, which Millikan thought "exceedingly interesting and thrilling," and also expensive, over $100,000. Strauss doled out his money in droplets; Brasch complained that he could not exist on "homeopathic doses" of dollars, and raised the estimate to $200,000; by 1940 the adventure had come to an end nowhere near its goal.^[151] Meanwhile fission had been discovered, and the royal road to atomic energy. Szilard kept Strauss apprised of progress by telegram, although they were almost neighbors, and asked his benefactor to find him another. Szilard had in mind Alfred Loomis, a retired investment banker and a first-rate amateur physicist, who soon became one of Lawrence's main advisors and supporters.^[152] Szilard wished Loomis to help underwrite the cost of experiments he planned to try at Columbia University, where he had a guest appointment, to determine how fission might be exploited for atomic energy. To complete the circle, the new émigré professor of physics at Columbia, Enrico Fermi, was then engaged in the same line of work. We shall return to their unequal competition.