Total: 1
This work applies Bayesian optimization to a loop composed of PIC simulations of laser electron acceleration and Monte Carlo (MC) simulations of bremsstrahlung-induced nuclear reactions, to maximize the production of molybdenum-99, the precursor of the most used radiopharmaceutical in nuclear medicine, metastable technetium-99. PIC and MC simulations are computationally intensive, and besides reducing the time spent, the Bayesian optimization coupling both simulations resulted in an improvement of an order of magnitude in the $^\text{99}$Mo yield over a previous work, in which the output of an optimization loop based solely on PIC simulations was used a posteriori to estimate $^{99}\mathrm{Mo}$ production through a MC simulation.