Chapter 14: The Feed Materials Program
The Manhattan District’s acquisition of uranium- and thorium-bearing ores was only the initial step in providing the essential materials for the large-scale electromagnetic, diffusion, and pile processes.1 The District also had to bring under contract and to monitor the operation of a complex network of processing plants for refining and converting the ore, first into pure concentrates of uranium oxide (black oxide) or sodium uranate (soda salt) and then into the chemical feed forms of uranium dioxide and trioxide, uranium tetrafluoride and hexafluoride, and uranium metal. Thus for the Army, development and management of the feed materials program, begun by the Office of Scientific Research and Development (OSRD) and Stone and Webster in 1941–42, proved to be one of its most challenging and difficult tasks in administering the atomic bomb project.2
Program Organization and Support Activities
With ore acquisition activities proceeding apace, District Engineer Marshall in October 1942 formed a Materials Section to monitor the shipment of uranium-bearing ores and other materials from mines, tailing piles, storage depots, and processing plants, as well as their treatment through various stages of refinement and conversion into feed materials. He selected Lt. Col. Thomas T. Crenshaw as section head and assigned several District officers already familiar with some aspect of materials procurement to assist him – including Capt. Phillip L. Merritt, a geologist by training, and Capt. John R. Ruhoff, a chemical engineer who, when serving as the St. Louis area engineer, had overall responsibility for the District’s uranium metal production. Also, because of Colonel Nichols’s deep involvement in the earlier OSRD acquisition program, Marshall had his deputy continue to give his special attention and expertise to the District’s feed materials program.3
The relocation of District headquarters from New York to Oak Ridge in mid-August 1943 occasioned a complete reorganization of the Materials Section. (See Chart 2.) Colonel Nichols, now the district engineer, decided to leave the materials group in New York City, close by the ports of entry and storage points for ores coming from overseas and also convenient to the headquarters of many of the firms under contract to supply feed materials. He redesignated the section as the Madison Square Area Engineers Office and, with transfer of Colonel Crenshaw to Oak Ridge as the officer in charge of all Clinton operations, assigned Ruhoff, recently promoted to lieutenant colonel, to be the Madison Square area engineer. Colonel Ruhoff took over administration of a burgeoning materials group, numbering nearly four hundred by early 1944, and an indication of its key role during the period of the project’s greatest activity, from late 1943 to the fall of 1944, was Nichols’s practice of coming to New York for weekly meetings with Ruhoff and his staff.4
The Madison Square staff, three-quarters of which worked in the New York City area and the rest at various points in the field, oversaw a program comprised essentially of four operations: a search for additional raw materials; their procurement in whatever form might be available; their refinement; and their conversion into feed materials. To ensure a steady flow of raw and semi-refined materials to the project’s processing plants, staff members closely monitored the scheduling of ore shipments from Africa to the port of New York; made or expedited arrangements for their storage; approved procurement of partially processed uranium-containing materials; and assisted in contracting with crude ore refining firms (African Metals Corporation, Eldorado Mining and Refining Company, and Vitro Manufacturing Company) to obtain uranium oxide, uranium sludge, radium and radioactive lead, and similar products. They also oversaw various research programs (Princeton and Yale Universities, Massachusetts Institute of Technology, and National Bureau of Standards), supplying them with an ever-increasing variety of other chemicals and special materials.
Staff members in the field provided liaison between the Madison Square office and seven area offices reporting to Ruhoff (Chart 4). Of these, two – the Murray Hill Area Engineers Office in New York and the Colorado Area Engineers Office in Grand Junction – monitored materials procurement, while five – Iowa (in Ames), St. Louis, Wilmington, Beverly (near Boston), and Tonawanda (near Buffalo) – oversaw feed materials processing operations.
Source: MDH. Bk. 7, Vol. 1, App. B4, DASA.
A single area engineer administered the Iowa and St. Louis offices, traveling between the headquarters located at Iowa State College and the Mallinckrodt Chemical Works. In Wilmington, the busy area engineer at Du Pont kept a check on that firm’s production of feed materials in addition to overseeing its plutonium program. In Beverly, the area engineer supervised the District’s contract for uranium metal production with Metal Hydrides, Inc. And in Tonawanda, the area engineer had responsibility for contracts with the Linde Air Products Company, a subsidiary of the Union Carbide and Carbon Corporation, for production and chemical processing of uranium oxide into its dioxide and salt forms and with the Electro Metallurgical Company for production of uranium metal; later he supervised contracts with the Harshaw Chemical Company of Cleveland, which made uranium tetrafluoride and uranium hexafluoride, and with the Hooker Electrochemical Company of Niagara Falls, New York, which reclaimed uranium from slag produced in the mining of carnotite and other ores.5
Feed Materials Procurement Raw Materials
From 1943 to the end of the war the Manhattan Project steadily increased its supplies of uranium ore, to ensure sufficient stores for conversion into the black oxide needed for the feed materials processing plants. Ore procurement activities, which reached a high point in 1944 and then leveled off somewhat in early 1945, were concentrated in three major areas: Africa, Canada, and the United States. Project leaders were aware in 1943 that the wartime needs of the bomb program were likely to exhaust both the immediately available domestic and Canadian deposits, and the security implications of this situation ultimately led to a District policy of using, to the greatest extent possible, ore from foreign sources.6
The most significant foreign source of natural uranium was the Belgian Congo, where the Belgian mining firm, Union Miniere du Haut Katanga, controlled all mineral rights. Following negotiations, the District procured the African ore through Union Miniere’s subsidiary, the African Metals Corporation. For the period October 1942 to December 1944, cost of 30,000 tons of Congo ore containing 3,800 tons of black oxide totaled more than $9 million, based on the price of oxide averaging about $1.12 per pound. The District’s Washington Liaison Office arranged for purchase of additional Congo ore, containing more than 3,100 tons of oxide and costing more than $10 million.7
All Canadian ore, procured through the Eldorado Mining and Refining Company (formerly, until June 1943,
Eldorado Gold Mines), came from the Great Bear Lake area. In May 1943, with completion of Stone and Webster’s initial purchase order (15 July 1942), the District negotiated another contract with a representative of Eldorado’s sales agency in the United States. But procurement officials soon experienced serious difficulties in implementing this contract and decided to terminate it. Colonel Ruhoff, acting in his capacity as chief of the newly constituted Madison Square Area Engineers Office, agreed in September to the terms of a new contract with Eldorado; he approved a second agreement in December 1944. For the period July 1942 to December 1944, cost of 4,200 tons of Canadian ore containing 1,137 tons of black oxide was slightly over $6.6 million, based on the price of oxide varying from about $1.95 to over $4.00 per pound.8
Domestic sources of natural uranium were in the Colorado Plateau region of the states of Colorado, Utah, and New Mexico. The uranium in this region occurred in carnotite ores, which also contained vanadium – an element urgently needed in the war effort because of its use as a hardening agent in the manufacture of steel. District procurement officials, learning in late 1942 that those firms actively mining carnotite ores and refining vanadium did not extract the relatively small amount of uranium in the refuse materials, began negotiations in early 1943 to acquire these tailings. Because these tailings were in the form of sand, and thus too heavy for economical shipment, they arranged contracts with several vanadium operators – the government-owned and -financed Metals Reserve Corporation,9 the privately owned and operated Vanadium Corporation of America, and the United States Vanadium Corporation, a Union Carbide subsidiary – and proposed they convert the tailings into concentrates (sludges). The advantage of the concentrates was that they would yield a higher percentage of uranium for conversion into black oxide and that, in this form, shipment to the Buffalo-area processing firms would be a less costly operation. For the period November 1942 to February 1945, cost of 380,000 tons of carnotite sands containing 1,350 tons of black oxide was more than $2.1 million, based on the price of oxide averaging about $0.80 per pound.10
Uranium ore from North America yielded considerably less black oxide than that from Africa, primarily because
of the much greater oxide content of the latter. The African ore from the Belgian Congo contained an estimated average of over 2-percent black oxide, whereas Canadian ore from the Great Bear Lake area assayed at somewhat more than 0.5 percent and domestic ore from the Colorado Plateau region at 0.25 percent. For this reason, the combined quantity of estimated black oxide in uranium ore purchased from the North American sources accounted for only one-third of the total contracted for the entire project.11
So tremendously important to the success of the atomic project was securing and processing raw ores that this operation tended to obscure another significant activity of the District’s feed materials program: procurement of special materials. A number of these materials were difficult to obtain in the quantities needed or completely unavailable from commercial sources. Hence, their procurement was often not simply a matter of District officials approving a purchase order or letting a contract, but required planning and implementing means for the radical expansion of such limited sources as existed or for even approving construction of entirely new plants. Two separate sections in the Madison Square Area Engineers Office had responsibility for special procurement – the Special Materials Branch and the Special Projects Branch.12
Demands for special chemicals and other materials of the project’s research and production facilities increased rapidly in 1943 and 1944. For testing and operating atomic piles there was need for radium and radioactive lead as a neutron source, graphite and beryllium as neutron moderators, and helium as a coolant; for the heavy water project at Trail, nickel chromium for a catalyst; for the gaseous diffusion project, elemental fluorine and a variety of fluorinated chemicals, including those suitable for cleaning, cooling, lubricating, and sealing; for the manufacture of uranium metal, magnesium and calcium; and for the design and test of the bomb at Los Alamos, a seemingly endless list of materials – bismuth, tungsten, boron, beryllium, and many others.
The quantity and variety of special materials needed by the project presented the Madison Square staff with a whole spectrum of challenging problems. Some proved to be relatively simple. For example, radium and radioactive lead, which were byproducts of uranium processing, could be obtained from the same firms that refined the ore. In the early period, project officials purchased most of the radium required through the New York firm, Canadian Radium and Uranium Corporation, which procured most of its supply from Eldorado Mining and Refining. In 1943, however, difficulties in reaching agreement on contractual terms and prices caused them to turn to Joseph A. Kelly, who acted as agent for the Radium Chemical Company of New York. After 1943, Kelly supplied most
of the radium required by the project. As for radioactive lead, the District obtained most of its requirement for this material from Eldorado Mining’s ore-refining operations at Port Hope, Ontario.13
Acquisition of a suitable pile moderator was one of the most difficult procurement problems in the early months of the project’s plutonium program. Pile designers finally decided to employ graphite rather than heavy water or beryllium, because it was the only one of these neutron-absorbing substances available in quantity from commercial sources and because Metallurgical Laboratory scientists and researchers at the National Carbon and Speer Carbon Companies recently had devised a process that would produce an adequate supply of high-grade graphite for the program. This success with graphite did not, however, end interest in obtaining beryllium and heavy water for experimental purposes. Project scientists, particularly those at Los Alamos, showed an increasing interest in beryllium metal in the later years of the war. Only a single American firm, Brush Beryllium Company of Lorain, Ohio, produced beryllium commercially for the fabrication of certain alloys. From 1943 to 1946, the Madison Square staff concentrated its efforts on increasing the production capacity of this firm, assisting it in obtaining priorities on new equipment and other materials from the War Production Board and also in expanding its plant. By 1945, these measures had led to a substantial increase in production of beryllium metal.14
Generally speaking, District procurement officials had to cope with no more than the usual stringencies of the tight wartime economy in obtaining moderate quantities of such elements as magnesium, calcium, bismuth, tungsten, boron, and helium. Because early decisions for helium-cooled production piles appeared to forecast a future need for very large amounts of the gaseous element, they arranged with the Bureau of Mines, which controlled helium distribution, for large-scale procurement, including transfer directly of funds from the War Department to Interior to pay the costs. In 1943, they also assisted in negotiation of a contract with the General American Transportation Company of Chicago for purchase of special tank cars to ship the helium to Hanford. But the decision by pile designers later that year to use water as the primary coolant greatly reduced the need for helium, and the District materials group sharply cut back the earliest procurement schedules for the element.15
Another material that presented special procurement problems was elemental fluorine, to include its chemical derivatives. This highly corrosive, and therefore hazardous-to-handle, element was the choice of the project designers for combining with uranium to make the gaseous feed material (uranium hexafluoride) for operating several of the main production plants.
Because of the huge requirements of just the gaseous diffusion plant, as well as the problems of shipment, the designers decided to build a fluorine gas production plant right at the diffusion plant site. The District’s materials group also played a significant role in letting contracts and overseeing the activities of a number of private research institutions (Johns Hopkins, MIT, Purdue) and chemical firms (American Cynamid, Du Pont, General Chemical, Harshaw Chemical, Hooker Electrochemical, Kinetic Chemicals, Penn Salt) in the development and supply of the numerous fluorinated hydrocarbon chemical compounds – in the form of coolants, sealants, and lubricants – needed to operate the plants safely and efficiently with the highly corrosive feed material.16
Feed Materials Production
The initial phase of the feed materials production network was conversion of the uranium-bearing crude ore into pure concentrates of black oxide and soda salt by various industrial firms under contract to the District. In each case the refining treatment was quite similar and involved subjecting the crude ore to the successive processes of pulverization into a sandlike material, acid immersion, precipitation to eliminate impurities, and roasting (drying).
Eldorado Mining at its Port Hope refinery processed all Canadian ore and some Congo ore into black oxide, whereas the Vitro Manufacturing Company at its Cannonsburg (Pennsylvania) refinery processed only Congo ore into soda salt. Designed only for treating the higher-grade Congo and Canadian ores, neither the Eldorado nor Vitro plants could properly process the carnotite concentrates from the Colorado Plateau region. Aware that the Linde Air Products Company had produced for the OSRD a satisfactory grade of black oxide from carnotite concentrates, the District’s Materials Section at the end of 1942 made arrangements with Linde to refine new stocks of concentrates at its plant in Tonawanda, New York, as well as to produce other feed materials for the project. With assistance of the Tonawanda area engineer, Linde expanded its black oxide production facilities, but, by late 1943, was phasing out domestic ores and using its facilities to refine higher-yielding African ores.17
Figures compiled by the Madison Square Area Engineers Office, beginning in September 1943, show that the amount of uranium from all sources available for refinement in the United States and Canada, and the quantity of black oxide and soda salt extracted from this ore, grew dramatically from 1943 to 1945. Thus, at the end of September 1943, the Manhattan District had available 2,920 tons of uranium ore and produced 1,660 tons of black oxide and soda salt. A year later, the quantities rose
to 5,640 tons available and 3,500 tons of black oxide and soda salt produced. And at the close of September 1945, the figures stood respectively at 6,600 tons of ore and 5,150 tons of black oxide and soda salt.18
The final phase in the feed materials production network was the conversion of black oxide and soda salt, through a series of chemical treatments, into one of the several chemical feeds suitable for processing in the electromagnetic, diffusion, and pile plants. The first step changed black oxide or soda salt into brown oxide (uranium dioxide) or orange oxide (uranium trioxide), the latter an important feed material for the electromagnetic process in its early stages of development. The second step transformed brown oxide into green salt (uranium tetrafluoride). The third, and final, step converted green salt into one of a number of uranium compounds – for example, gaseous uranium hexafluoride for the gaseous and liquid diffusion processes and the electromagnetic process in its last stage of development – or into uranium metal, the prime feed material for the pile process.19
Because the OSRD had made considerable progress in arranging contracts with industrial firms to provide for each of the different chemical treatments required to produce feed materials, the principal task remaining for Manhattan leaders was that of shaping the project’s feed materials processors into a production network capable of supplying most of the feeds for the Clinton and Hanford production plants, regardless of the adverse effects of sabotage, technical failures, or other inhibiting factors. By early 1943, having extended OSRD contracts and negotiated new agreements, they organized and expanded this network so that, in effect, it comprised three parallel chemical-processing chains, the first link in each chain consisting of processors of both brown and orange oxide; the second, those of green salt; and the third, those of uranium metal.20
Mallinckrodt, Du Pont, and Linde comprised the brown and orange oxide links. Mallinckrodt, which had pioneered in development of the highly efficient ether process for refining uranium under the leadership of Ruhoff, provided the most important link. During the course of the wartime project, it produced nearly 4,200 tons of brown and orange oxide, nearly twice the output of the other two firms, and including almost all of the oxide used by the electromagnetic project. In cooperation with Yale University, it continued research that culminated in design and construction of a plant for continuous extraction of brown oxide from raw ore (pitchblende), not completed until 1946. The Du Pont plant, built adjacent to the company’s big Chambers Chemical and Dye Works across the Delaware River from Wilmington in Deep Water, New Jersey, processed
mainly scrap and by-products material to produce almost 2,000 tons of brown oxide. Linde, operating the third plant, processed black oxide from its own refinery to produce a total of about 300 tons of brown oxide.21
Four chemical firms comprised the green salt links. Three were the same firms that produced brown oxide and the fourth was the Harshaw Chemical Company of Cleveland, which the OSRD had originally brought under contract to produce green salt in the summer of 1942. District procurement officials drew up new contracts for a substantially enlarged output in the fall of 1942 – with Harshaw in September and the other three companies in November. These contracts, except for that with Harshaw, remained in effect for the duration of the war and resulted in production of more than 7,200 tons of green salt: 2,926 by Mallinckrodt, 2,060 by Linde, 1,640 by Harshaw, and 608 by Du Pont. When more uranium hexafluoride was needed for the diffusion plants, the Madison Square Area Engineers Office renegotiated the contract with Harshaw, providing in a new agreement that the Cleveland firm convert black oxide into green salt and then into uranium hexafluoride. At the same time, the Madison Square office also arranged to have Harshaw raise its output of uranium tetrachloride, which it had been producing in small quantities since early 1943, to meet a sudden increase in demand for the electromagnetic production plan.22
Four commercial firms and a college formed the uranium metal links. Mallinckrodt, Du Pont, Electro Metallurgical, Metal Hydrides and Iowa State, at one time or another, were involved in metal production for the wartime atomic project, although only the first three firms constituted the permanent links in the parallel feed materials chains. Uranium metal procurement dated back to the earliest days of the atomic energy program, because the material was required for laboratory research and experimentation. Both the National Bureau of Standards and the OSRD had let contracts to university research laboratories and commercial chemical firms to develop a process for mass production of uranium metal of a high degree of purity. The processes devised by Metal Hydrides proved to have serious drawbacks. Iowa State, however, had developed a method for reducing green salt with calcium (later, magnesium proved more effective) at high temperatures inside a steel bomb and recasting the end product into metal in an induction-heated furnace. So successful was this method that Iowa State itself employed it to manufacture a considerable amount of metal for the project. Subsequently, the Army let contracts to Mallinckrodt, Du Pont, and Electro Metallurgical to produce metal using the steel bomb method.23
When the Army took over direction of materials procurement, it continued the metal-production contracts
with Metal Hydrides and Iowa State and negotiated new contracts with Electro Metallurgical and Du Pont. In several instances, District officials had to monitor construction of additional plant buildings, at government expense, to expedite the production of uranium metal under these contracts. Metal Hydrides and Du Pont had serious operating problems that limited their output of metal, although Metal Hydrides subsequently developed a highly successful metal-recasting operation. Nevertheless, by the time District officials shut down most production of new metal in late 1943 – Iowa State continued its output until late 1944 – the various contractors had manufactured several thousand tons. By late August 1944, the Madison Square area engineer reported delivery of nearly 3,500 tons of metal to Hanford and Clinton, comprised of 1,000 tons from Electro Metallurgical, 900 from Iowa State, 650 from Mallinckrodt, 610 from Metal Hydrides, and lesser amounts from other processors. These deliveries included both new metal and metal recast into ingots from turnings and other scraps from machining and fabricating operations.24
Quality Control Program
One factor that made materials procurement difficult was the almost universal requirement for previously unheard of standards of quality. In the feed materials program, for example, procurement schedules required that uranium metal contain no more than 0.1 of 1 percent of impurities that would affect its efficiency in the pile-operating process. Similarly stringent standards were established for graphite, fluorinated chemicals, and other materials. Because most of the commercial contractors who furnished these materials were unprepared to carry out the physical and analytical tests necessary to maintain these high standards, the materials group had to build up its own quality control organization.25
In February 1943, Colonel Crenshaw’s staff began negotiations with Princeton, MIT, the chemical section of the Metallurgical Laboratory, and the National Bureau of Standards, with the objective of forming these research institutions into a central quality control laboratory group. The plan was to have the scientists at each institution analyze and test samples from the uranium metal production plants, as well as to devise more effective methods of metal analysis, to furnish personnel and facilities when needed to supplement those of the manufacturing plants, to investigate other materials, and to provide general guidance for the control program. In addition, the Metallurgical Laboratory was to carry out physical tests of brown oxide and finished metal for the pile process. Because all of these institutions already were engaged in some aspect of analysis and testing of uranium, the Materials Section simply supplemented or revised existing contracts with them to provide the necessary
organization of the central quality control laboratory group.26
By spring, the Materials Section had completed satisfactory arrangements with Princeton, MIT, and the Bureau of Standards. Colonel Crenshaw reported in May that these three institutions were “doing an excellent job, and have attacked the problem as a job of commercial analysis, which is the case.”27 The Metallurgical Laboratory expressed a preference for carrying out its part of the analytical work under its existing overall research contract, but Crenshaw opposed this, because he knew it would prevent the Materials Section from exercising direct control over the laboratory’s part in the analytical program. The reasons why the laboratory did not want such a contract soon became apparent: The scientists did not relish performing routine analysis and testing of metal samples because it took time and used facilities they would rather devote to more original and challenging research and development activities.28
Colonel Crenshaw arranged a meeting with Richard L. Doan, associate director of the University of Chicago’s Clinton Laboratories in Tennessee, and George E. Boyd, chief of the analytical chemistry group at the Metallurgical Laboratory. The two scientists agreed that the Metallurgical Laboratory would continue to perform routine chemical analysis and testing of brown oxide and uranium metal until the workload in this area declined. This would occur shortly, they knew, when Iowa State completed facilities for quality testing its own metal output. By fall of 1943, the other institutions had taken over most of the routine chemical analytical work that the Metallurgical Laboratory had been doing. The Madison Square area engineer attested to the effectiveness of the quality control program when, at the end of November, he reported to Colonel Nichols that the feed materials program was making metal of a higher degree of purity than any previously produced by the atomic energy project.29
Development of the feed materials program ahead of the fissionable materials production and weapon programs was a matter of necessity, for the latter were completely dependent upon an adequate supply of the feed and other materials essential to their operation. In less than two years, the Manhattan District’s materials organization was able to expand the already existing OSRD program, solving serious technical problems and securing the requisite priorities to meet on schedule the requirements for the research and development, testing, and start-up in operations of the major production plants for the manufacture of fissionable materials. By late 1944 and in 1945, the District could begin to phase out, or reduce, some aspects of the program and to give some attention to the postwar requirements of the atomic energy program.