Degree Name

Doctor of Philosophy


School of Physics


Methods of determining the biological effects of radiation based on the concepts of linear energy transfer as a physical descriptor of the radiation field and absorbed dose are flawed due to limitations in these quantities. For example, radiation with the same linear energy transfer and absorbed dose may have very different biological effects due to the potential for very different track structures and energy deposition spectra. Methods of estimating biological effects of radiation which are independent of linear energy transfer and absorbed dose, such as track structure theory and microdosimetry are described.

Track structure theory was applied to heavy charged particle relative thermoluminescence efficiencies in LiF:Mg,Ti and LiF:Mg,Cu,P. The Unified Interaction Model of dose response was applied to the experimental dose response of 100 keV synchrotron x-rays. The Unified Interaction Model has had success in explaining the thermoluminescence mechanisms responsible for the dose response of the various peaks in the glow curve, particularly the supralinearity at dose levels above ~1 Gy. Track structure theory requires matching of the dose response function with the spectra of secondary particles liberated by heavy charged particles slowing down in material. The energies of such particles are typically of the order of a few keV. Measurements at these energies have proven difficult and so the dose response at ultra-low electron energies has been estimated from the Unified Interaction Model by extrapolation of the maximum dose response values from available data. Calculations of relative proton and α-particle thermoluminescence efficiencies from track structure theory in LiF:Mg,Ti and LiF:Mg,Cu,P differ significantly (by factors up to ~30 for 4.95 MeV α-particles and ~10 for 1.43 MeV protons) from experimentally measured values. However, uncertainties in the experimental measurements, uncertainties arising from the estimation of the dose response function and possible uncertainties in previous calculations of radial dose distributions are significant. More sophisticated calculations and experimental measurements at ultra-low photon/electron energies are indicated for future studies.

Microdosimetry was applied in heavy ion fields relevant to heavy ion therapy and space radiation fields using two generations of silicon-on-insulator microdosimeter arrays. Radiation protection for these applications requires the ability to measure the rapidly changing lineal energy of the particle with high precision. The high spatial resolution of both microdosimeters due to their 10 μm thickness was demonstrated by measurements of 4He and/or 12C ion beams at the HIMAC (Japan) and HIT (Germany) heavy ion therapy facilities. Contributions from secondary particles, particularly neutrons were observed, demonstrating the ability of microdosimetry to measure the lineal energy of components of the unknown spectrum of secondary radiation. Differences observed in the lineal energy spectra measured by first and second generation microdosimeters were attributed to slight differences in the mean chord lengths and chord length distributions due to the different sensitive volume geometries. Microdosimeters were also able to accurately reproduce intricate dose plans. Out-of-field measurements with the microdosimeters positioned lateral to the heavy ion field showed that the majority of secondary radiation originates inside the treatment volume upstream of the beam.

A third generation microdosimeter, developed using n-type silicon-on-insulator (nSOI) and epitaxial technologies, was developed with the aim of increasing the sensitive surface area and yield of sensitive volumes. Charge collection studies indicated a 100% yield which is a significant improvement over the previous generations. However, both nSOI and epitaxial devices were found to suffer from charge sharing between sensitive volume and guard ring structures, as well as an enhanced energy response to heavy ions. In addition, the epitaxial microdosimeters were found to be highly susceptible to radiation damage. A coincidence analysis of ion beam induced charge collection designed to investigate the anomalous response of nSOI microdosimeters confirmed the occurrence of charge sharing between the sensitive volume and guard ring. The guard ring was applied as a veto electrode to discriminate shared charge, which was shown to improve the charge collection geometry and the measured energy deposition spectrum. The effective sensitive surface area of a single cell was reduced from ~20 μm to ~8 μm, which is closer to the 10 μm diameter of the nominal sensitive volume. However, as the geometry of the sensitive volume is still not properly understood this technique is useful for characterisation but not for experimental microdosimetry due to the requirement of a well-defined sensitive volume. The anomalous energy response was investigated using ion beam induced charge collection and spectroscopy with 12C, 4He and H ions of various energies and linear energy transfers. No correlation between particle LET and the energy over-response was found. The enhanced energy response was hypothesised to be a result of a displacement current induced in the active SOI layer by charge carriers induced in the substrate due to the parasitic capacitance of the SiO2. This hypothesis was investigated using the response of the device to 148Gd α-particles, whose range is less than the thickness of the active SOI layer. The enhanced energy response was not observed, indicating, although not confirming, that the enhanced energy response is a result of a displacement current. A second hypothesis for the cause of the enhanced energy response was that the thickness of the active SOI layer is greater than the value of 10 _m provided by the device manufacturer. A scanning electron microscopy study coupled with energy dispersive x-ray spectroscopy on an nSOI microdosimeter provided no evidence of the SiO2 insulating layer which limits the thickness of the SOI active layer to 10 μm. To confirm the viability of this technique for observing the SiO2 layer, the same investigation was performed on a second generation SOI microdosimeter. The SiO2 layer was clearly observed at a depth of 9.6±0.2 μm with a thickness of 1.9±0.2 μm, in agreement with the device specifications. This finding explains the enhanced energy observed, however, the question as to why full energy deposition is not observed remains unanswered.