One of the objectives of the PhD work, is to develop an analysis tool for MRGS spectra.
A first step for such analysis tool is a characterization of the detection system. Measurements with calibrated gamma ray sources in a well-defined geometry are foreseen with available CZT and LaBr detectors. CZT detector spectra exhibit a low energy tail which needs to be appropriately described in view of a quantitative analysis of the peaks.
The usual fitting techniques applied for Germanium detectors are not suitable for the analysis of these spectra. In addition, the commercially available tools are often as black boxes where little information is given about the fitting procedure and parameters and their associated uncertainty budget. When applied to CZT detector spectra analysis poor results were obtained. Other gamma-ray codes developed for CZT detectors such as sIGAle are not commercially available.
In a first stage, the aim is to implement an analysis method based on peak fitting. This will allow to have an analysis tool to determine net peak areas and treat spectra of medium complexity with a possibility to deconvolute overlapping peaks as is the case in gamma-ray spectra of uranium.
In a second stage, a method based on a response method will be studied. This method relies on the predetermined detector response to gamma-rays of given nuclides to compose the gamma ray spectrum that best fits to the experimental data. Monte Carlo methods based on the code MCNP-CP will be used to investigate and account for the correlated photon-electron transport and effects related to the measurements geometry. A tool based on the response method has the potential to treat more complex spectra such as the one arising from Pu bearing samples.
The analysis tool will be benchmarked with measured data acquired with available CZT and with LaBr in collaboration with the Institute for Reference Materials and Measurements (IRMM). The data are measured on U and Pu standards.
In addition, a benchmark using synthetic spectra can also be done. The synthetic spectra can be generated with Monte Carlo methods by including the obtained detector characteristics.
The goal of the PhD is to give an answer to the applicability of MRGS on Pu and U bearing materials:
- what is the accuracy to be expected (in comparison to HRGS),
- what are isotopic compositions of Pu, U where the method works or fails;
- what is the impact of the spectrum quality (e.g. what are acceptable measurement times)
- how is uncertainty dealt with.