Abstract
Gamma ray spectroscopy is routinely used to measure radiation in a number of situations. These include security applications, nuclear forensics studies, characterisation of radioactive sources, and environmental monitoring. For routine studies of environmental materials, the amount of radioactivity present is often very low, requiring spectroscopy systems which have to monitor the source for up to 7 days to achieve the required sensitivity. Recent developments in detector technology and data processing techniques have opened up the possibility of developing a highly efficient Compton Suppressed system, that was previously the preserve of large experimental collaborations. The accessibility of Monte-Carlo toolkits such as GEANT4 also provide the opportunity to optimise these systems using computer simulations, greatly reducing the need for expensive (and inefficient) testing in the laboratory. This thesis details the development of such a Compton Suppressed, planar HPGe detector system. Using the GEANT4 toolkit in combination with the experimental facilities at AWE, Aldermaston (which include HPGe detection systems, scintillator based detector systems, advanced shielding materials and gamma-gamma coincidence systems), simulations were built and validated to reproduce the detector response seen in the 'real-life' systems. This resulted in several improvements to the current system; for the shielding materials used, terrestrial and cosmic radiation were minimised, while reducing the X-ray fluorescence seen in the primary HPGe detector by an order of magnitude. With respect to the HPGe detector itself, an optimum thickness was identified for low energy (<300 keV) radiation, which maximised the efficiency for the energy range of interest while minimising the interaction probability for higher energy radionuclides (which are the primary cause of the Compton continuum that obscures lower energy decays). A combination of secondary detectors were then optimised to design a Compton Suppression system for the primary detector, which could improve the performance of the current Compton Suppression system by an order of magnitude. This equates to a reduction of the continuum by up to a factor of 240 for a nuclide such as Co-60, which is crucial for the detection of low-energy, low-activity emitters typically swamped by such a continuum. Finally, thoroughly optimised acquisition and analysis software has also been written to process data created by future high sensitivity gamma coincidence systems. This includes modules for the creation of histograms, coincidence matrices, and an ASCII to binary converter (for historical data) that has resulted in an analysis speed increase of up to ~20000 times when compared to the software originally used for the extraction of coincidence information. Modules for low-energy time-walk correction and the removal of accidental coincidences are also included, which represent a capability that was not previously available.