Bayesian Optimization of Photochemically Induced Acousto-Optics in Ozone Gas for Enhanced Diffraction Efficiency

Traditional acousto-optics involves modulating the refractive index of a crystal using acoustic waves, but this limits the damage threshold under high-energy laser exposure. To address this, recent developments have extended the technique to gaseous media, where acoustic waves are generated by localized UV light absorption, resulting in spatially modulated gas heating. However, this setup introduces stringent requirement on experimental setup, including pump and probe beam alignment as well as the delay time between laser beams, making manual optimization tedious and time-consuming. In this work, we focus on optimizing the diffraction efficiency of photochemically induced acousto-optics in an ozone-oxygen gas mixture using Bayesian optimization (BO). A Gaussian process surrogate model is employed to predict diffraction efficiency, balancing exploration and exploitation to minimize the number of experimental trials. Our findings demonstrate that BO can significantly reduce the experimental workload while achieving high diffraction efficiency, making it a powerful tool for optimizing acousto-optic interactions in gaseous media. This approach offers a promising framework for optimizing complex experimental setups.