Medical Isotope Production Without Highly Enriched Uranium (2009)

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

7 Conversion to LEU-Based Production of Molybdenum-99: Technical Considerations T he objective of this chapter is to describe and discuss the important technical considerations for conversion of molybdenum-99 (Mo-99) production from highly enriched uranium (HEU) to low enriched uranium (LEU). This chapter is intended to support the discussion of con- version feasibility that appears in Chapter 10. The focus of this chapter is on conversion of the HEU targets that are currently being used to produce Mo-99 for medical use (Chapter 2). With two exceptions, all of the reactors that are currently being used for large- scale production of Mo-99 (Chapter 3) have already been converted to LEU fuel. The exceptions are the Safari-1 Reactor in South Africa and the Belgian Reactor II (BR2) in Belgium. Safari-1 is in the process of converting (see Chapter 3), and BR2 will convert when a suitable LEU fuel becomes avail- able. A general discussion of research reactor fuel conversion is provided in Chapter 11. TARGET DESIGN AND PROCESSING As noted in Chapter 1, almost all of the Mo-99 produced for medical use in the world today is made using HEU targets. These targets consist of an HEU “meat,” usually a uranium oxide or uranium metal alloy, contained within a metal or metal alloy cladding (Chapter 2). Three basic a ­ pproaches exist for converting these targets to LEU: 90 

TECHNICAL CONSIDERATIONS 91 • Direct replacement of the HEU in the target with LEU (with an increase in the number of targets that are irradiated). • Increase the mass of U-235 in the LEU target by increasing target size. • Increase the mass of U-235 in the target by changing the composi- tion of the target meat. These approaches are described in the following sections. Direct Replacement of the HEU in the Target with LEU HEU and LEU have essentially the same physical and chemical proper- ties, so the direct replacement of HEU by LEU in the target meat would pose no particular target design, fabrication, or testing challenges. The LEU target would have the same geometry, heat transfer, and chemical process- ing properties as the equivalent HEU target and could be irradiated and processed in essentially the same manner. Assuming the same target design and uranium density, the yield of Mo-99 from the LEU target would be only about 20 percent of the HEU target it replaces owing to its reduced uranium-235 (U-235) mass and increased neutron capture. Consequently, approximately five times as many LEU targets would have to be irradiated and processed to produce the same amount of Mo-99 as a single HEU target, and up to five times as much volume of waste from target process- ing might be produced as a result. Some producers have suggested that their facilities might not be able to accommodate these higher throughput requirements without substantial modification. Facility modifications might not be necessary, however, if certain pro- cess changes are made. Current target dissolution processes (see Chap- ter 2) operate well below solubility limits using containers that are small relative to the hot cells in which they sit. Higher throughputs could be accommodated by increasing container sizes and/or increasing material concentrations in the solvent. Liquid waste can also be converted to ­solids by precipitation, evaporation, or calcination to substantially reduce its volume. Moreover, given its lower U-235 enrichment, this solid waste can be more closely packed together in storage containers and facilities with- out increasing criticality risks. Because the LEU targets produce the same amount of fission heat and heat-producing fission products as HEU targets   Most HEU targets have 93 percent U-235 enrichments; LEU targets would have 19.75 percent enrichments. Neutron capture in LEU targets (primarily by uranium-238 [U-238]) would be about 15 percent higher than in equivalent-sized HEU targets.   Increasing material concentrations in the solvent could lead to criticality problems if HEU targets are used but would not be a problem for LEU targets. 

92 MEDICAL ISOTOPE PRODUCTION WITHOUT HIGHLY ENRICHED URANIUM with the same Mo-99 yield, heat management requirements would also be the same. Another consideration for direct replacement is reactor irradiation capacity. Most of the world’s supply of Mo-99 is produced by irradiat- ing HEU targets in multipurpose, multiuser facilities (Table 3.2). Reactor operators’ ability to accommodate larger numbers of LEU targets could be limited because of other user demands on reactor resources. Increase the Mass of U-235 in the LEU Target by Increasing Target Size Additional U-235 could be incorporated into an LEU target by increas- ing the volume of the target material (i.e., the target meat). This approach would reduce target throughput requirements in the reactor but would not substantially change the other material throughput requirements described previously. Also, space limitations in the reactor target irradiation positions might preclude the use of substantially larger targets. Increase the Mass of U-235 in the Target by Changing the Composition of the Target Meat The HEU targets used for most current Mo-99 production are u ­ ranium-aluminum alloys (Table 2.2) having uranium densities approach- ing 1.6 g/cm3. To obtain an equivalent mass of U-235 in an LEU target of the same size, a uranium density of about 8 g/cm3 would be required. Higher-density LEU targets could be made of several materials: • Uranium metal targets. Argonne National Laboratory has led the development of a uranium metal target (Figure 7.1) in cooperation with several organizations. Recent progress is described by Vandegrift et al. (2007), Bakel et al. (2008), and Wiencek et al. (2008)., The target consists of a thin (typically 100- to 150- micron) LEU metal foil wrapped in an   Although the focus of this discussion is on targets, the same considerations apply for the conversion of reactor fuel from HEU to LEU as will be discussed in Chapter 11. Targets and fuels have the same basic sandwich design and differ primarily in size and configuration.   The primary participants are Comisión Nactional de Energía Atómica (CNEA, Argentina), MURR (United States), and Indonesian National Atomic Energy Agency (BATAN). CNEA is providing advice on target design and has carried out tests on irradiated foils. BATAN and the Australian Nuclear Science and Technology Organisation (ANSTO) have also test-irradiated these foils. MURR is evaluating target fabrication approaches and modeling target thermal properties. In early November 2008, MURR also began irradiating and processing small (5 g) targets.   Compagnie pour l’ Etude et la Réalisation de Combustibles Atomiques (CERCA, France) is also investigating LEU foil targets in cooperation with the Missouri University Research Reactor (Allen et al., 2007). 

TECHNICAL CONSIDERATIONS 93 A Weld end closed U foil located between tubes B 7-1a new FIGURE 7.1  (a) LEU metal foil targets developed by Argonne National Laboratory. (b) Views into the hot cell at the Missouri University Research Reactor (MURR) showing the irradiated LEU foil target being removed from the target cladding in preparation for dissolution. SOURCES: Courtesy of George Vandergrift, Argonne National Laboratory, and the University of Missouri, respectively. 

94 MEDICAL ISOTOPE PRODUCTION WITHOUT HIGHLY ENRICHED URANIUM aluminum or nickel foil barrier and encapsulated in a cylindrical alumi- num cladding. The aluminum or nickel foil serves as a recoil barrier and prevents the uranium foil from bonding with the aluminum cladding. The cylindrical design was selected to improve target structural integrity and heat transfer and facilitate physical target disassembly after irradiation. However, these targets could also be fabricated as plates (see Allen et al., 2007). The primary advantage of this target material is its high uranium density (~19 g/cm3): Gram for gram, uranium foil targets can produce as much or more Mo-99 than currently used HEU targets under the same irradia­tion conditions. The uranium foil is potentially compatible with the alkaline and acidic dissolution processes that are currently employed by large-scale producers. The foils used for this development work have been produced by Argonne National Laboratory by hot and cold rolling and by the Korea Atomic Energy Research Institute (KAERI, South Korea) using a casting method (Kim et al., 2004).10 The KAERI foils are economical to produce but contain pinholes and surface irregularities and are of uneven thick- nesses. These irregularities would not necessarily preclude the use of these foils for Mo-99 production, but could make it more difficult to qualify the targets for use and more expensive to produce targets on a produc- tion ­basis. Cold rolling these foils would eliminate these irregularities but is ­ labor intensive. KAERI is working on improving the consistency of its foils,11 and Argonne is investigating other potential sources for obtaining large quantities of these foils.12 Work is underway at the University of Missouri to develop foil target designs that can be used for high volume production of Mo-99 (Solbrekken et al., 2008).   After the target is irradiated the aluminum cladding and foil sandwich are mechanically separated and the uranium foil is chemically processed. The separation of the foil from the cladding prior to processing reduces the mass of material that must be chemically dissolved, which is another advantage of this target design.   For example, the HEU targets now being used by the Institut National des Radioéléments (IRE) to produce Mo-99 contain 3.7 g of U-235. If the HEU meat in these targets were replaced with an LEU metal foil of the same thickness, it would contain 16.6 g of U-235 (Wiencek et al., 2008).   The alkaline process requires the use of hydrogen peroxide to oxidize the uranium metal.   Argonne National Laboratory has also developed a modified Cintichem process to dissolve these targets and recover Mo-99. This process has Mo-99 recovery efficiencies of over 90 per- cent, which is similar to the recovery efficiencies for the alkaline and acidic processes that are currently being used by large-scale producers. See Bakel et al. (2008) for recent progress. 10 Hot rolling refers to heating metal above its recrystalization temperature before rolling to form sheets. Cold rolling is conducted at room temperature to maintain a metal’s original crystalline structure. Casting involves melting the metal and pouring it into a mold. 11 George Vandegrift, Argonne National Laboratory, written communication, July 14, 2008. 12 The Y-12 site in Tennessee has the equipment and materials to produce these foils, for example. 

TECHNICAL CONSIDERATIONS 95 • Uranium-aluminum dispersion targets. CNEA has developed and is using high-density LEU-aluminum dispersion targets to produce Mo-99 for its domestic market (Kohut et al., 2000; Cestau et al., 2008). The target meat has a uranium density of about 2.9 g/cm3 (IAEA, 2003, Annex 1), which is obtained by increasing the ratio of uranium aluminide to alumi- num in the target meat. The aluminum serves as a binder in the target meat. The mass of U-235 in the target meat is about twice that of conventional uranium-aluminum alloy targets. These targets are compatible with the alkaline dissolution processes that are currently used by most large-scale Mo-99 producers (see Chapter 2). However, these targets would still not have enough U-235 mass to serve as direct replacements for the HEU t ­ argets used by current producers. • Uranium silicide targets. Uranium silicide (U3Si2) was initially devel­ oped as an LEU replacement fuel for research reactors (see Chapter 11). Its use as a target material would represent a natural extension of that application. The primary advantage of this material is its higher uranium density13 (4.8 g/cm3) relative to uranium-aluminum dispersions and its ease of ­ fabrication into targets (see Kolar and Wolterbeek, 2004). However, uranium silicide is difficult to dissolve14 and cannot be processed using conventional alkaline or acidic dissolution processes (Chapter 2). • Uranium-molybdenum targets. As described in Chapter 11, work is currently under way to develop high-density LEU fuels using uranium- m ­ olybdenum alloys. The goal of this work is to develop fuels that have ura- nium densities in the range of 7–9 g/cm3, which are within the range needed for direct replacement of HEU in targets. However, uranium-­molybdenum ­alloys are unsuitable for use for Mo-99 production because of their high Mo-98 content. The Mo-98 in the target would dilute the Mo-99 produced during irradiation, reducing its specific activity sufficiently to make it unusable.15 13  Uranium silicide fuel having a uranium density of 4.8 g/cm3 has been qualified for use in research and test reactors (USNRC, 1988). Argonne has fabricated fuel plates with uranium densities of up to 6.1 g/cm3, and CERCA has fabricated fuel plates with uranium densities of up to 6.0 g/cm3, but these have not been qualified for use as reactor fuel. See http://www.rertr. anl.gov/QualFuel.html and Durand et al. (1992). 14  The Argonne National Laboratory study (Buchhold and Vandegrift, 1995) on processing uranium silicide concluded that: (1) Neither of the alkaline solvents that are typically used to dissolve uranium-aluminum alloy targets (NaOH or NaOH/NaNO3) dissolves uranium silicide. (2) Uranium silicide can be dissolved in NaOH if hydrogen peroxide is added. This solvent dissolves uranium silicide at an acceptable rate, but agglomerates of the material form during dissolution and must be broken up to obtain rapid dissolution. (3) Uranium silicide will dissolve in nitric acid, but gelatinous silicic acid forms unless the Si concentration is main- tained at less than 0.1 molar. (4) The fluoride ion dissolves uranium silicide but complicates waste treatment and disposal because of its corrosive nature. See also Cols et al. (2000). 15  Researchers at Delft University are investigating methods to separate Mo-99 from Mo-98; see Appendix D. 

96 MEDICAL ISOTOPE PRODUCTION WITHOUT HIGHLY ENRICHED URANIUM The design of the LEU target has important implications for target dissolution, Mo-99 recovery, and waste management (see Chapter 2 for a discussion of these issues). The two principal processes that are currently used for HEU target dissolution and Mo-99 recovery are broadly similar (Chapter 2): Irradiated HEU targets are dissolved in acidic or alkaline solutions, and Mo-99 is recovered through a series of chemical process- ing steps followed by sorption onto an alumina column or other media. Over the years, Mo-99 producers have added proprietary improvements to their processes to reduce processing time, reduce product impurities, and improve Mo-99 recovery. These processes are technically mature and have allowed producers to achieve good consistency in the quantity and quality of their Mo-99 product. Producers would probably prefer to convert to LEU-based produc- tion without having to make major changes to their target dissolution and Mo-99 recovery processes. The ideal approach for conversion is exemplified by the path taken by CNEA when it developed high-density LEU-aluminum dispersion targets. These targets were described previously. The CNEA- developed target is the same size and approximately the same U-235 mass as the HEU target it replaced. CNEA was able to produce this target by increasing the uranium density and thickness of the target meat and reduc- ing the cladding thickness. As a result, CNEA had to make relatively minor adjustments to its target dissolution and Mo-99 recovery processes during conversion. Moreover, CNEA was able to convert to LEU-based produc- tion while maintaining HEU-based production—and to carry out both of these activities in a single set of hot cells. Cestau et al. (2007) reported that the efficiency and stability of CNEA’s LEU-based process is similar to the HEU-based process it replaced. Mallinckrodt, IRE, and MDS Nordion will probably not be able to f ­ ollow this conversion path, primarily because there are no LEU-aluminum or LEU-oxide materials with sufficiently high uranium densities (~ 9 g /cm3) to serve as direct replacement for HEU targets, nor are such materials on the horizon. These producers could certainly develop LEU-aluminum t ­ argets following the CNEA approach, but they would still have to irradi- ate, process, and manage up to two or three times more targets and wastes. Alternatively, these producers could use LEU metal targets but would likely have to modify their target dissolution processes. However, because LEU metal targets have such high uranium densities, material throughputs (both for target irradiation and processing) would likely be smaller than for cur- rently used processes. Nuclear Technology Products (NTP) Radioisotopes uses 45 percent HEU to produce Mo-99 instead of the 93 percent HEU that is used by the other three large-scale producers. Consequently, it must process ap- proximately twice as much HEU material to produce the same amount of 

TECHNICAL CONSIDERATIONS 97 Mo-99 as other large-scale producers. NTP could probably convert to LEU targets using a CNEA-type target design without markedly increasing target throughput or processing requirements. RESEARCH AND DEVELOPMENT (R&D) TO SUPPORT TARGET CONVERSION As the foregoing discussion illustrates, there are a number of techni- cal options available to producers for converting to LEU-based targets for production of Mo-99. However, there is no “best” or “one size fits all” approach. Each producer must choose a conversion path based on its own assessment of cost, time, and technical practicability. R&D will be essential for making wise selections. Some of the necessary R&D is already in prog- ress (e.g., LEU metal foil targets) or has been completed (e.g., high density LEU-aluminum dispersion targets). This work provides a good starting point for understanding the range of available conversion options. Producers will need to focus their R&D on specific target design and fabrication, process development, and waste management operations. A ­ lthough these operations are discrete and can in principle be investigated separately, R&D will be more effective in terms of cost, time, and outcomes if these operations are treated as a system to be optimized for LEU-based production. The work being carried out by Argonne National Laboratory and its collaborators, which was described previously in this chapter, is a good example of a systems-focused approach. In its discussions with large- scale producers, however, the committee did not see any clear evidence of such a systems-level focus.16 This is likely due (at least in part) to the fact that most producers do not have all of the R&D capabilities in-house that are needed to address such systems-optimization problems. Also, because producers consider their processes to be proprietary they may be reluctant to seek outside assistance. The primary objective of systems-focused R&D is to develop LEU tar- gets that are well suited to downstream process operations, beginning with their irradiation in reactors and subsequent chemical processing through waste management (including consideration of recovery of uranium from process wastes). Systems-focused R&D might proceed as follows: • Target fabrication. Development work on LEU targets would be initially aimed at producing designs that mimic (to the extent feasible given the forgoing discussion on target design options) the characteristics of currently utilized HEU targets. This would serve to reduce the number of changes in downstream target irradiation and processing operations. 16  Producers did not share all of the details of their R&D work with the committee. 

98 MEDICAL ISOTOPE PRODUCTION WITHOUT HIGHLY ENRICHED URANIUM Ideally, the targets would be roughly equivalent in U-235 mass to the HEU targets they would replace so that they produce roughly equivalent amounts of Mo-99. • Process development. To the extent that process changes are r ­ equired to accommodate LEU targets, R&D would focus on target dis- solution efficiency, Mo-99 recovery efficiency and purity, and minimization of process waste streams. • Waste management. The vast majority of the uranium in the target eventually becomes waste no matter what process is used; only a very small fraction (typically about 3 percent) of the U-235 in the target undergoes fission. However, the volume and form of the waste can have a substantial impact on the difficulty and costs for its management, storage, and ultimate disposal. The committee was told by some producers that there are numerous process uncertainties that must be performed at both the front and back ends of the Mo-99 production process before conversion to LEU-based production will be feasible. At the front end, the important uncertainties include the target dissolution rates, chemistry, and process liquid properties. At the back end, important uncertainties include product yields, product quality and consistency, and waste volumes. Some producers have cited the lack of available hot cell space as an impediment to addressing these uncertainties. In the committee’s judgment, much if not most of the necessary pro- cess development work can be resolved at relatively low cost using well- e ­ stablished process development and testing procedures. Current producers have decades of experience in handling and processing HEU targets to r ­ ecover Mo-99. Conversion to LEU-based production is not likely to ­require substantial changes to current processing equipment or processing flow sheets. Consequently, access to hot cells would not be required for most of the needed R&D work. Target processing equipment is small—it would easily fit on a large laboratory bench (see Chapter 2). This makes it possible to carry out testing at full scale and at relatively low cost. Most of the front-end process uncertainties can be resolved through “cold” (i.e., nonradioactive) testing. Such testing allows LEU target m ­ aterials to be evaluated using conventional wet-laboratory facilities. B ­ ecause target irradiation times for Mo-99 production are short and U-235 burn-ups in the targets are low, unirradiated LEU targets have essentially the same material and chemical properties as irradiated LEU targets. Con- sequently, issues such as the following can be evaluated with cold testing on ­ unirradiated targets: efficiencies and sizes of separations equipment, reagent volumes and concentrations, dissolution rates, dissolution chemis- try, treatment of process gases, and process liquid throughputs. In the case 

TECHNICAL CONSIDERATIONS 99 of alkaline dissolution, uranium precipitation rates and filtration rates can also be determined. Many of the important back-end uncertainties can be assessed through tracer testing (“slightly hot” testing) using only very small amounts of radio­active material. These tests could use stable Mo-98 that is spiked with tracer amounts of Mo-99 and radioactive impurities of concern for target processing to determine Mo-99 recovery efficiencies and purity. Because only tracer amounts are used, these tests could probably be carried out in hoods. Of course, full-scale testing with irradiated LEU targets would be required to demonstrate that the process works as designed. Additionally, some process uncertainties that can only be resolved through full-scale test- ing with irradiated LEU targets, for example, product purity testing. These tests could probably be carried out in a single hot cell over a period of a few weeks to months. To the committee’s knowledge, none of the major producers are doing much actual development work17 on LEU targets and process, including the use of cold or slightly hot testing as part of their conversion strategies. The committee views this as a missed opportunity. FINDINGS Several important technical considerations for converting Mo-99 pro- duction from HEU targets to LEU targets were described and discussed in this chapter. Based on this information, the committee finds that: • There are three basic approaches for converting HEU targets to LEU: Direct replacement of HEU in the target with LEU; increasing the mass in U-235 in the target by increasing target size; or increasing the mass of U-235 in the target by changing target composition. Each approach has advantages and disadvantages. • There are no technical barriers to conversion of Mo-99 produc- tion from HEU targets to LEU targets. Production using LEU targets is technically feasible and in fact is being carried out by CNEA in Argentina and shortly will be applied by the Australian National Nuclear Science and Technology Organisation (ANSTO) using CNEA technology (see Chap- ter 3). The committee sees no technical barriers to scaling up to large-scale 17  HEU-based producers did not provide details of any development work aimed at conver- sion, either in presentations to the committee or in discussions during site visits. On this basis the committee assumes that development work such as is described above has not been done. The committee is aware of a conversion feasibility study that was carried out by Atomic E ­ nergy of Canada Limited (AECL), but that organization was unwilling to share the results of that study with the committee. 

100 MEDICAL ISOTOPE PRODUCTION WITHOUT HIGHLY ENRICHED URANIUM production. However, such scale-up could require additional investments in facilities and personnel. • There is no single “best” approach for conversion: Each producer must choose a conversion path based on its own assessment of cost, time, and technical practicability. There are opportunities for modifying current target or process designs that would allow producers to convert within their existing facilities. • R&D will be essential for making wise selections about conversion approaches. Most of the needed R&D can be carried out using cold test- ing and radioactive tracer testing at full scale and at relatively low cost in conventional laboratory facilities. Except for some specific testing needs, access to expensive hot cell facilities would not be required. • Based on the information presented to it by producers, the commit- tee did not see any evidence that such R&D was being carried out. Additional information about the prospects for conversion for existing HEU-based Mo-99 producers is provided in Chapter 10.