The Manhattan Project – The US Enters the War

Following the Pearl Harbor attack on December 7, 1941, the United States entered WW II (Germany and Italy declared war on the U.S. three days later).

In January 1942, President Franklin D. Roosevelt gave secret, tentative approval for the development of an atomic bomb. The Army Corps of Engineers was assigned the task and set up the Manhattan Engineer District to manage the project. A bomb research and design laboratory was to be built at Los Alamos, New Mexico. Due to uncertainties regarding production effectiveness, two possible fuels for the reactors were to be produced with uranium enrichment facilities at Oak Ridge, Tennessee, and plutonium production facilities at Hanford, Washington.

Just before the attack on Pearl Harbor Compton’s committee had concluded that a critical mass of between 2 and 100 kilograms of uranium-235 would produce a powerful fission bomb and that for $50-$100 million isotope separation in sufficient quantities could be accomplished. Bush forwarded the findings to Roosevelt under a cover letter on November 27, 1941. Roosevelt did not respond until January 19, 1942 when he did, it was as commander in chief of a nation at war. The President’s hand-written note read,

“V.B. OK returned
I think you had best keep this in your own safe
FDR”.

In 1942 it had been estimated that military and industrial targets in Germany could be devastated with 500,000 tons of TNT bombs, which would be the equivalent to 1 to 10 tons of uranium-235. It was said that Bush had instructed everyone not to discuss the expense of producing uranium-235 to avoid arousing government fears of excessive costs.

As already mentioned, with the written agreement of Roosevelt, Bush immediately put Eger Vaughan Murphree, a chemical engineer with the Standard Oil Company, in charge of engineering studies and supervising pilot plant construction and any laboratory-scale investigations. He confirmed Urey, Lawrence, and Compton as program chiefs. Urey headed up work including diffusion and centrifuge methods and heavy-water studies. Lawrence took electro-magnetic and plutonium responsibilities, and Compton ran chain reaction and weapon theory programs.

Following this Compton decided to concentrate the work for which he was responsible at the University of Chicago. The Columbia group under Fermi, and its accumulated material and equipment, and the Princeton group, which had been studying resonance absorption, were moved to Chicago in the course of spring 1942.

From the meeting of December 18, 1941 Compton decided that they needed to demonstrate a chain reaction within 6 months. For the exponential piles and measurements of physical constants he wanted 80 men and $340,000 for Columbia and Princeton, and for Chicago he wanted 58 men and $278,000. He also wanted $500,000 for pile materials. Before Pearl Harbor these figures would have been fantastic, after, they became reasonable. In a meeting of his new group on January 3, 1942 he found that his three teams could not agree on anything. Finally they did agree on the experimental data for the ‘pile’ but not on specifications. Oppenheimer‘s name again appeared as the one to coordinate paper studies on fast-neutrons. Another meeting on the January 18, 1942 made everyone realise that each had made some mistakes (including Fermi underestimating the poisoning effect of boron). A plan was outlined in which a bomb would be ready in early 1945. They also decided on the purchase of 30 tons of uranium metal. All the teams had argued for research to be concentrated on one site. Compton finally decided to bring Columbia and Princeton to Chicago. For space he appropriated part of the cyclotron room (in the Service Building), most of Eckhart Hall, and the area under the north and west stands of Stagg Field (American football fields).

Under Lawrence the investigation of large-scale electro-magnetic separation was accelerated at the University of California at Berkeley and a related separation project was started at Princeton.

Before the meeting of December 18, 1941 Lawrence had already started to convert the 37-inch cyclotron to a mass spectrometer.

Research and development on the diffusion process and on the production of heavy-water continued at Columbia under Urey, under the general supervision of Murphree. The centrifuge work continued at the University of Virginia under Beams while the Columbia centrifuge work was transferred to the laboratories of the Standard Oil Development Co. at Bayway, New Jersey.

With the entry of the U.S. into war the situation changed from having too little money and no deadlines, to a situation with a clear goal, plenty of money, and too little time.

On March 9, 1942, Bush sent a report to the President reflecting general optimism but placing proper emphasis on the tentative nature of conclusions. Compton had promised a multiplication factor of 1.14, but they had overestimated the expected results with uranium metal and the projected first pile experiment suddenly grew by 15%. Fortunately this discouraging information came too late to influence the decision of Roosevelt. The report of Bush contemplated completion of the project in 1944. In addition, the report contained the suggestion that the Army be brought in during the summer of 1942 for construction of full-scale plants.

The entire heavy-water program was under review in March and April 1942. The reviews followed a visit to the United States in February and March 1942 by Francis Simon, Hans Heinrich von Halban, and Wallace Alan Akers from England. In a memorandum of April 1, 1942 addressed to Bush, Conant reviewed the situation and reported on conferences with Compton and Briggs. His report pointed out that extremely large quantities of heavy-water would be required for a plutonium production plant employing heavy-water instead of graphite as a moderator. For this reason, he reported adversely on the suggestion that Halban be invited to bring to this country the 165 litres of heavy-water which he then had in England.

In a memorandum written to Bush on May 14, 1942, Conant estimated that there were five separation or production methods which were about equally likely to succeed: the centrifuge, diffusion, and electro-magnetic methods of separating uranium-235, and the uranium-graphite pile and the uranium-heavy-water pile methods for producing plutonium. All the production methods were considered about ready for pilot plant construction and perhaps even for preliminary design of production plants. If the methods were to be pushed to the production stage, a commitment of $500 million would be entailed. Although it was too early to estimate the relative merits of the different methods accurately, it was presumed that some methods would prove to be more rapid and efficient than others. It was feared, however, that elimination of any one method might result in a serious delay.

It was also thought that the Germans might be some distance ahead of the United States in a similar program.

Research on uranium required uranium ore, and obtaining sufficient supplies was the responsibility of Murphree and his group. Fortunately, enough ore was on hand to meet the projected need of 150 tons through mid-1944. Twelve hundred tons of high-grade ore were stored on Staten Island, and Murphree made arrangements to obtain additional supplies from Canada and the Colorado Plateau, the only American source. Uranium in the form of hexafluoride was also needed as feed material for the centrifuge and the gaseous and thermal diffusion processes. Philip Abelson, who had moved from the Carnegie Institution to the U.S. Naval Research Laboratory, was producing small quantities, and Murphree made arrangements with E. I. du Pont de Nemours and Company (their TNX Division for the atomic energy program) and the Harshaw Chemical Company of Cleveland to produce hexafluoride on a scale sufficient to keep the vital isotope separation research going.

In fact in addition to the 1,200 tons on Staten Island the Belgian owners also had 100 tons of oxide at a refinery on the shores of Lake Ontario. Also the Eldorado Mine near the Artic Circle could produce 300 tons per year. And there was at least 500 tons of oxide in waste sludge from vanadium refineries on the Colorado Plateau.
Obtaining uranium metal was problematic. As a powder it is highly pyrophoric and very difficult to melt and cast. In addition there was a supplementary requirement of less than 2 parts per million boron. During 1942 only a few kilograms could be produced at any time.
Finally the centrifuges, gaseous-diffusion and thermal-diffusion processes needed uranium hexafluoride, which until 1940 could only be produced in laboratory conditions. In 1942 the Planning Board contracted for a pilot plant able to produce 3-5 kg/day uranium hexafluoride.
There was also a request for one to two tons of heavy-water, and finally in 1942 a technique was found allowing them to build a pilot plant. Whilst heavy-water would have been the better moderator for a ‘pile’ they simply did not have enough. With 5 tons heavy-water they could have built a small pile and produce 100’s of grams of element-94 (I’m giving up using ‘….’ around pile, and pile, reactor and core all mean the same thing). With 30 tons they could have imagined producing a kilogram of element-94 per week. But with the best will in the world they might have had 5 tons in mid-1943, and 30 tons only in mid-1945. And it must be admitted that at the time the results with graphite were encouraging, the first experiments in Chicago gave a ƙ of 0.94 +/- 0.02. With a similar configuration Fermi also obtained a ƙ of 0.9 in Columbia. The feeling was that by refining the experiment and better quality materials they could obtain a ƙ of more than 1.0. In fact by July 1942 Fermi had obtained a ƙ of 0.995 just by improving the experiment and refining the calculation.

Uranium metal was also required. Uranium ore is usually refined as uranium oxide, often called black oxide, or uranium salt. In early 1942 only small quantities of uranium metal were available. In fact uranium ore had been stockpiled from a mine in northwest Canada, but the mine had to be re-opened when a large new order for uranium ore was made in 1941. The National Bureau of Standards developed a new process for removing all impurities by a single extraction method, and the Mallinckrodt Chemical Works developed a large-scale production unit for uranium oxide. This was then converted into uranium tetrafluoride, or green salt, which is the feed material for most uranium metal-making processes.

Graphite was available in the U.S., but what was needed was a very pure graphite, in particular with no boron. The National Bureau of Standards traced the boron in commercial graphite to the coke used for its production. Coke was substituted with petroleum, which solved the problem.

Heavy-water was another problem. Arrangements had been made post June 1943, but there was an immediate need, and the world’s entire stock of heavy-water (aprox. 400 pounds) was in the hands of the British. In fact Joliot-Curie had sent his collaborators, Hans von Halban and Lew Kowarski, with 165 litres of heavy-water, first to Britain, then to Canada.

There was a major requirement for several thousand tons of copper for the magnet coils used by Lawrence. But copper was a critical war material, so silver was used. Initially when they asked for 15,000 tons of silver from the Under Secretary of the U.S. Treasury, the answer was “we usually talk of silver in ounces”. The first transfer was for 175 million fine troy ounces, i.e. 6,000 tons, but the final figure turned out to 14,700 tons.