26th Australian Institute of Physics Congress and the Australian and New Zealand Conference on Optics and Photonics

R-matrix analysis of proton-boron fusion for energy production and dose-enhancement in proton-beam cancer therapy

Not scheduled

20m

Belinda Hutchinson Building (The University of Sydney )

Poster AIP | Nuclear and Particle Physics (NUPP)

Description

Proton-boron fusion is a resonant reaction with a large positive Q-value that decays via multiple channels into three $\alpha$ particles. These properties make it a leading candidate for aneutronic fusion energy production and show significant promise for using boron as a dose enhancement agent in proton therapy in cancer treatment. Modelling these applications requires a precise reaction cross-section evaluation over a broad energy range, however, this remains poorly constrained.

In this contribution, I will discuss recent work to evaluate the proton-boron fusion cross-section using phenomenological R-matrix analysis, a robust framework for interpreting resonant reaction data, enabling cross-sections to be interpolated and extrapolated to unmeasured angles and energies consistently with the underlying reaction dynamics.

Previous evaluations of the reaction have been hampered by inconsistencies between experimental data sets and a limited assessment of the uncertainties in the resultant R-matrix description. Dataset uncertainties arise primarily from differences in energy calibration and systematic uncertainties in target thickness, detector deadtime, beam current and detector efficiency. A rigorous survey of all available datasets has been conducted to assess their quality. New experimental data has also been measured at the ANU Ion Implantation Lab. While this is primarily intended as a proof-of-principle for making cross-section measurements in this facility, careful energy normalisation and uncertainty quantification will ideally provide a well-calibrated dataset for the R-matrix analysis.

To improve the uncertainty assessment of the final fits, a Markov Chain Monte Carlo (MCMC) routine has been integrated into a new phenomenological R-matrix code. Unlike conventional $\chi^2$ minimisation, which is prone to becoming trapped in local minima, MCMC explores the full parameter space by sampling probability distributions, offering greater robustness for high-dimensional problems and natively providing well-quantified uncertainties on the final cross-section. This work will ultimately lay the foundation for an open-source fitting framework extendable to other low-energy scattering reactions.