Optimal experimental design (OED) is necessary for maximizing utility of expensive timeconsuming complex experiments, particularly when dealing with high-dimensional design spaces. In this talk, we discuss the application of Bayesian OED for optimizing a photoionization mass spectrometry experiment targeting chemical rate measurements for low temperature oxidation of propane in a high-pressure photolysis reactor [1]. The Bayesian framework provides well-grounded means for handling uncertainty in experimental conditions as well as parameters of the physical system and instrument models, and for the probabilistic modeling of information gain. It also allows handling both experimental measurement noise and model error in a unified landscape. We discuss first the construction and Bayesian calibration of a model for the photolysis reactor and the time-of-flight mass spectrometer instrument [2]. We outline our utilization of a Gaussian process model error construction, and highlight necessary localization and subsampling strategies for computational feasibility. We then use this surrogate in a Bayesian OED construction, where we target maximization of an expected utility function that measures the information gain provided by the experiment for a given design, for model parameters of interest. We discuss requisite approximations and dimensionality reduction necessary for tractability, given large data size and high dimensional model inputs/outputs. We examine the performance of the construction, and illustrate the effective optimization of experimental conditions, and the impact on Bayesian estimation of parameters of interest.