Laser-Induced Breakdown Spectroscopy for Molten Salts

Abstract

Many current processes in the nuclear industry rely on sample extraction and ex-situ services for analysis. To reduce the risks of exposure of personnel to ionising radiation, enhance operational efficiency of future processes, and ensure nuclear material safeguarding, in-situ process monitoring solutions suitable for the hostile environments of nuclear processes must be developed. This thesis addresses the lack of available instrumentation in the context of molten salt fuel reprocessing – a process that is currently being developed in the UK. One proposed solution is a submersible laser-induced breakdown spectroscopy (LIBS) probe that would provide compositional melt analysis in real-time. This thesis describes a multidisciplinary approach to the instrument development and addresses the key questions and practical considerations for such a device. The optical design of the probe is based on fibre optic laser delivery and signal collection, to enable subsurface measurements of the melt without exposing sensitive equipment to the melt environment. To address the challenges associated with pulsed laser to fibre coupling, a custom vacuum coupling chamber was designed. The design incorporates several novel features, namely an adaptable optical design and custom optical fibre feedthroughs. Radiation testing of optical components was carried out to evaluate the influence of radiation induced attenuation on quantitative LIBS analysis, and guide optical component selection for the probe head. High OH and low OH fibres were compared, and while transmittance in both was found to reduce when subjected to ionising radiation, the high OH fibres exhibited greater tolerance. UV-fused silica lenses were found not to be affected by radiation at doses typical for the application environment. Quantitative analysis of LIBS data by univariate and multivariate techniques were evaluated in the context of process monitoring and radiation. Univariate calibration curves were found to be cumbersome and therefore unsuitable for industrial applications. Unsupervised classification using principal component analysis was found to be an effective analysis technique, and most efficient when a reduced spectral window was used as opposed to the full spectrum. Partial least squares regression was found to be most appropriate for process monitoring, as it enabled simultaneous consideration of numerous emission peaks while taking concentration into account for regression. The performance of all techniques tested was adversely affected by radiation, but not significantly enough to compromise operation. Discrepancies are expected between experiments and a real-world scenario however, as LIBS analysis was performed with annealed fibres and not inside an active environment. A microchip laser was evaluated as an alternative excitation source via the quantitative analysis of gadolinium in simulated nuclear debris samples. The low pulse energies and greater shot-to-shot variation resulted in inferior performance for quantitative analysis when compared to an actively Q-switched laser. The effects of radiation and elevated temperatures, evaluated from published literature, were also found to be too significant for the molten salt process. Although of significant interest for other LIBS applications, the microchip laser was found to be unsuitable for a molten salt monitoring application.

Date of Award	1 Aug 2023
Original language	English
Awarding Institution	The University of Manchester
Supervisor	Paul Wright (Supervisor) & Barry Lennox (Supervisor)