Introduction The embrace of PET by medical clinicians has been reluctant (ΔT ≈ 20 yr) primarily due to the scale of the infrastructure that is needed. The capital cost of a cyclotron (≈ 106 USD) is now dwarfed by the demand for compliance to recent regulatory standards. This is a recurring expense, not only imposing an order-of-magnitude increase in staffing and operating costs, but damping the enthusiasm of researchers recalling the brisk pace of research in earlier days. Now an academic site, with little interest or opportunity to scale up production for wider distribution, is burdened by the new regulatory terrain of good manufacturing practice (GMP), mandated for translational studies that will reach only a few subjects. With our production resources held within a basic science department, the Medical Physics cyclotron facility at the University of Wisconsin has sought a sustainable pathway. We now anchor the operating budget by providing high-value, long-lived radionuclides to off-site users, to buffer the fluctuations of local demand for conventional PET synthons. Material and Methods: The tools of the trade The radioisotopes discussed here belong to the 3-d and 4-d sub shell, but are now moving into the rare-earths, with applications ranging from - targeted molecular imaging agents, - internal radionuclide therapy using to Auger electron-emitters, - to basic physics experiments using 163Ho (t1/2 ≈ 4500 yr) to determine the mass of the neutrino. Rather than focusing on the dozens of radionuclides produced, a number of tools deserve mention, as they support a variety of targets, reactions and products. These will be listed in order (A-G) from cyclotron to extraction to analysis. A. Two cyclotrons are used, a legacy RDS 112 (#1; 1985) and a GE PETtrace (2009). Neutron and gamma detectors are monitored during the long irradia-tions, signaling any subtle changes in the running conditions. (1). The PET-trace is fitted with a quick-change variable degrader target (2), as well as a beam-line with a 5-port (0 o, ±15 o, ±30 o) vertical switching magnet (3). The downward directed beam ports provide support for solid targets (e.g. Ga, S, Se, Te) that melt at low temperature. The irradiation of gas targets employs a generalized manifold to handle inert gases such as 36Ar for the production of 34mCl, as well as natural Kr and Xe for making Rb and Cs isotopes to act as fission product surrogates. These products are captured on a stainless steel target chamber liner, and rinsed off with warm water. The alkali metals are convenient tracers to study the ion exchange trapping process, pivotal in future 99Mo production from solution reactors (4). B. The preparation of malleable solid targets employs a 10-ton hydraulic bench press, and a jeweler’s mill to roll out foils from pellets, pressed between Nb foils to avoid contamination. C. Binary alloys are smelted in a programmable 1600o tube furnace under argon flow (eg. NiGa4). Alternatively, an induction furnace now permits highly localized heating of the binary metal charge, while allowing mechanical agitation during the smelting process. D. Electroplating onto gold discs is used for various enriched target material or the alloys above where quantitative recovery is essential, or where heat transfer from high beam current is demanding. E. The separation chemistry, prior to che-lation to targeted molecular imaging agents, is performed in LabView-driven, home-built “black boxes” resident in mini-cells (Radiation Shielding Inc.). F. Analysis of the targets after irradiation makes use of HPGe spectroscopy for gammas and characteristic X-rays of decay (e.g. rare earths). The elemental constitution of target alloys is deter-mined prior to irradiation by X-ray fluorescence analysis, excited by 109Cd and 241Am sources. G. Finally, broad-band elemental analysis at the ppb level now makes use of a microwave plasma atomic emission spectrometer (Agilent 4200), to be de-scribed elsewhere in this meeting. Results and Conclusions The tools above (A-G) are employed in the pro-duction of the expanded list of radionuclides offered by our cyclotron group to both local and off-site colleagues. The list below is ordered in terms of decreasing use, from regular production for national distribution (64Cu, 89Zr), to weekly inhouse use (44Sc, 66,68Ga, 68,69,71Ge, 72As, 61Cu, 86Y), to infrequent production for multi-site collaborations (163Ho, 95mTc, 206Bi): Radionuclide Target Employs 64Cu 64Ni/Au A, D, G 89Zr natY A, E, G 44Sc natCa A, B, E, F, G 66, 68Ga Zn/Ag A, B, D, E, F, G 68, 69, 71Ge Ga, GaO2 A, B, C, E,F 72As GeO2 A, B, E, F 52Mn natCr A, E, F, G 76, 81mBr SeO A, E, F 34mCl, Rb, Cs noble gas A, E, F 95mTc,163Ho Mo, Dy A, E, F TABLE 1. Target materials and processes. The production of long-lived radionuclides lends itself to crowd-sourcing, with distributed irradia-tion at virtually any site with a suitable accelera-tor and a relaxed beam schedule. A number of unique challenges do arise that don’t appear in the usual production of conventional cyclotron products such as 11C or 18F. Contamination by stable metals, inadvertently introduced by target pressing or beam-induced sputtering from degraders, can cause serious interference downstream limiting effective specific activity. Long-lived manganese isotopes are ubiquitous. And some very high value products are simply not within the reach of small cyclotrons, such as 52Fe and 67Cu, being too far off the line of beta stability. In conclusion, the research leading to a doctoral degree necessarily must focus on the physics and chemistry of novel radionuclides and tracers. On the other hand, clinical and translational research needs established imaging agents, with little room for innovation within the regulatory constraints. Our experience at Wisconsin has led us to a balancing act, with our routine production of clinical doses countered with our research program to provide high-value radionu-clides for our collaborative work with our basic science colleagues.

Introduction The embrace of PET by medical clinicians has been reluctant (ΔT ≈ 20 yr) primarily due to the scale of the infrastructure that is needed. The capital cost of a cyclotron (≈ 106 USD) is now dwarfed by the demand for compliance to recent regulatory standards. This is a recurring expense, not only imposing an order-of-magnitude increase in staffing and operating costs, but damping the enthusiasm of researchers recalling the brisk pace of research in earlier days. Now an academic site, with little interest or opportunity to scale up production for wider distribution, is burdened by the new regulatory terrain of good manufacturing practice (GMP), mandated for translational studies that will reach only a few subjects. With our production resources held within a basic science department, the Medical Physics cyclotron facility at the University of Wisconsin has sought a sustainable pathway. We now anchor the operating budget by providing high-value, long-lived radionuclides to off-site users, to buffer the fluctuations of local demand for conventional PET synthons. Material and Methods: The tools of the trade The radioisotopes discussed here belong to the 3-d and 4-d sub shell, but are now moving into the rare-earths, with applications ranging from - targeted molecular imaging agents, - internal radionuclide therapy using to Auger electron-emitters, - to basic physics experiments using 163Ho (t1/2 ≈ 4500 yr) to determine the mass of the neutrino. Rather than focusing on the dozens of radionuclides produced, a number of tools deserve mention, as they support a variety of targets, reactions and products. These will be listed in order (A-G) from cyclotron to extraction to analysis. A. Two cyclotrons are used, a legacy RDS 112 (#1; 1985) and a GE PETtrace (2009). Neutron and gamma detectors are monitored during the long irradia-tions, signaling any subtle changes in the running conditions. (1). The PET-trace is fitted with a quick-change variable degrader target (2), as well as a beam-line with a 5-port (0 o, ±15 o, ±30 o) vertical switching magnet (3). The downward directed beam ports provide support for solid targets (e.g. Ga, S, Se, Te) that melt at low temperature. The irradiation of gas targets employs a generalized manifold to handle inert gases such as 36Ar for the production of 34mCl, as well as natural Kr and Xe for making Rb and Cs isotopes to act as fission product surrogates. These products are captured on a stainless steel target chamber liner, and rinsed off with warm water. The alkali metals are convenient tracers to study the ion exchange trapping process, pivotal in future 99Mo production from solution reactors (4). B. The preparation of malleable solid targets employs a 10-ton hydraulic bench press, and a jeweler’s mill to roll out foils from pellets, pressed between Nb foils to avoid contamination. C. Binary alloys are smelted in a programmable 1600o tube furnace under argon flow (eg. NiGa4). Alternatively, an induction furnace now permits highly localized heating of the binary metal charge, while allowing mechanical agitation during the smelting process. D. Electroplating onto gold discs is used for various enriched target material or the alloys above where quantitative recovery is essential, or where heat transfer from high beam current is demanding. E. The separation chemistry, prior to che-lation to targeted molecular imaging agents, is performed in LabView-driven, home-built “black boxes” resident in mini-cells (Radiation Shielding Inc.). F. Analysis of the targets after irradiation makes use of HPGe spectroscopy for gammas and characteristic X-rays of decay (e.g. rare earths). The elemental constitution of target alloys is deter-mined prior to irradiation by X-ray fluorescence analysis, excited by 109Cd and 241Am sources. G. Finally, broad-band elemental analysis at the ppb level now makes use of a microwave plasma atomic emission spectrometer (Agilent 4200), to be de-scribed elsewhere in this meeting. Results and Conclusions The tools above (A-G) are employed in the pro-duction of the expanded list of radionuclides offered by our cyclotron group to both local and off-site colleagues. The list below is ordered in terms of decreasing use, from regular production for national distribution (64Cu, 89Zr), to weekly inhouse use (44Sc, 66,68Ga, 68,69,71Ge, 72As, 61Cu, 86Y), to infrequent production for multi-site collaborations (163Ho, 95mTc, 206Bi): Radionuclide Target Employs 64Cu 64Ni/Au A, D, G 89Zr natY A, E, G 44Sc natCa A, B, E, F, G 66, 68Ga Zn/Ag A, B, D, E, F, G 68, 69, 71Ge Ga, GaO2 A, B, C, E,F 72As GeO2 A, B, E, F 52Mn natCr A, E, F, G 76, 81mBr SeO A, E, F 34mCl, Rb, Cs noble gas A, E, F 95mTc,163Ho Mo, Dy A, E, F TABLE 1. Target materials and processes. The production of long-lived radionuclides lends itself to crowd-sourcing, with distributed irradia-tion at virtually any site with a suitable accelera-tor and a relaxed beam schedule. A number of unique challenges do arise that don’t appear in the usual production of conventional cyclotron products such as 11C or 18F. Contamination by stable metals, inadvertently introduced by target pressing or beam-induced sputtering from degraders, can cause serious interference downstream limiting effective specific activity. Long-lived manganese isotopes are ubiquitous. And some very high value products are simply not within the reach of small cyclotrons, such as 52Fe and 67Cu, being too far off the line of beta stability. In conclusion, the research leading to a doctoral degree necessarily must focus on the physics and chemistry of novel radionuclides and tracers. On the other hand, clinical and translational research needs established imaging agents, with little room for innovation within the regulatory constraints. Our experience at Wisconsin has led us to a balancing act, with our routine production of clinical doses countered with our research program to provide high-value radionu-clides for our collaborative work with our basic science colleagues.