Innovative optimization studies to advance the current state of isotope production target design at the high flux...

A growing demand for various stable and radioactive isotopes is limiting the ability to incorporate new and innovative target designs for isotope production (IP) at the US Department of Energy's High Flux Isotope Reactor (HFIR) located at Oak Ridge National Laboratory. This demand has increased reliance on historical and conservative approaches, which has forced scientists to operate in a "just-in-time" manner, producing routine yield calculations with few advancements. The goal of this work is to demonstrate the effectiveness and applicability of a recently proposed HFIR-specific optimization framework; as a test case, its efficacy is demonstrated for a routinely produced isotope, ¹⁸⁸W. The developed framework relies on a Python-based metaheuristic (MH) optimization algorithm coupled with stochastic radiation transport and deterministic activation/depletion/decay codes. The algorithm automatically designs, simulates, and evaluates thousands of unique target design candidates based on user-defined variable ranges, mock input files, and an objective function to maximize ¹⁸⁸W production. By using the current target design as the baseline, various new and innovative target designs were found to return higher specific activity (SA), a key radioisotope production metric. The variables explored included the target placement in HFIR, the number of W rings in the target, the option to separate each ring with various spacer materials, and the vertical thicknesses of the rings and spacers. One advantage of this optimization framework is the flexibility and robustness of a newly developed and computationally efficient target design simulation model. Six optimization studies are presented in this article and included nearly 20,000 simulations. These simulations showed that Al-based spacers integrated into a four-ring design increased SA by up to 20% compared with the current eight-ring target design without spacers. Moreover, the MH framework predicted that minimizing the W ring thickness and maximizing spacer thickness would contribute to an even higher SA increase of 31%. Exotic fissile flux-boosting and coproducing spacer materials were also explored, and although they did not outperform the most optimal Al-based spacers, they did contribute to higher SA than the current target design and demonstrated the efficacy of a dual-target system. These results demonstrate the usefulness of the proposed general-use optimization methods for ¹⁸⁸W IP target design and indicate much wider applicability for other IP campaigns at HFIR and similar state-of-the-art facilities.