Characterization and application of 3D silicon microdosimeters
MetadataVis full innførsel
The effect of ionizing radiation on biological matter differs significantly between the various types of radiation. For the same amount of absorbed energy, some forms of radiation are much more effective in inducing biological response than others, having a higher radiation quality. Not only does the radiation quality differ between the particle species, but it also depends on the particles’ energy. Microdosimetry is an experimental and theoretical scientific field where the energy deposition in micrometric volumes is used to quantify the radiation quality. The strength of microdosimetry is that although the underlying physics is complex, the radiation quality is defined in principally simple terms which are quantifiable and measurable and can provide input to radiobiological models. At the heart of the microdosimetry is the detector, or microdosimeter, which is used to measure energy depositions. For 75 years the tissue equivalent proportional counter (TEPC) has been the gold standard for microdosimetry, but over the last two decades silicon detectors have been developed as an alternative. The main objective of this work has been to characterize and test a new generation of silicon microdosimeters with five slightly different designs. Electrical characteristics were measured and the microdosimeters have been tested with several soft photon sources and an Am-241 alpha source. The charge collection efficiency (CCE) was determined by comparing the results to that of a commercial PIN diode for spectroscopy. One of the microdosimeters was investigated in a microbeam with the ion beam induced charge collection (IBICC) technique with C-12 ions, revealing the sensitivity of the different parts of the microdosimeter and produced radiation damage effects. A microdosimeter was also used to measure the energy deposition at all depth of an absorber in a 15 MeV proton beamline used for radiobiological experiments. The results were compared to both a MC simulation and the dose measurements from a commercial ionization chamber (IC). The measurements in the proton beam were conducted to further characterize the microdosimeter and was used as a microdosimetric characterization of the beamline. Since the silicon microdosimeters are not tissue equivalent (TE) the measurements from the 15 MeV beamline were corrected with a novel tissue correction function presented here and compared to a previously used method from literature. The measurements showed that the silicon microdosimeters are fully depleted at 5 V with a dark current of approximately 0.1 nA and capacitance below 80 pF. Photon sources between 8 and 60 keV showed 100% CCE for all microdosimeters. The alpha particles produced spectra with a peak at 1445 keV, which were in line with MC simulation. The spectra also had a very large fraction of events below 100 keV and a low amplitude constant band of events between 100 and 1200 keV not visible in the simulations. The IBICC experiment showed homogeneous charge collection at the centre of the SVs but they had a clear sensitivity gradient at the edges giving rise to lower energy events from the monoenergetic beam. The high LET C-12 microbeam produced surface damage, where charge in the oxide layer made the volume between the SVs sensitive. The effects from the surface damage were reduced effectively by increasing the bias voltage from 5 to 15 V. In the 15 MeV proton beamline, the energy deposition spectra at all depths of the polyamide absorber matched well with the MC simulations apart from a slight shift towards higher energy depositions at the entrance. MC simulations of the proton beam showed that the tissue correction function had a maximum error of 1.1% while previously used methods gave up to 15% error. The comparison with the IC indicated that the tissue corrected microdosimeter reproduced the relative depth dose profile well, although the comparison suffered from slightly different measurement positions with respect to the absorbers. The measured tissue corrected dose-mean lineal energy was between 8 and 35 keV/µm and matched well with simulations of a tissue composed microdosimeter except for a 12% difference at the entrance. An alternative type of microdosimeter is also presented and discussed, where a stack of high granularity pixel sensors can be used to track all the particles entering and generated within the microdosimeter. The specifications from the ALPIDE detector with a 5 µm resolution along the two dimensions of the sensor plane are used in the discussion. 12 µm resolution can be achieved in the depth direction by stacking the sensors densely but would be reduced by inserting tissue equivalent material between the sensors to make the detector more biological relevant. The ALPIDE can coarsely measure the energy deposition in each layer by allowing clusters of pixels to fire when struck by a particle. A design with the current ALPIDE detector should be able track primary particles entering the detector well but would have issues with tracking most of the secondary electrons as they would need at least 50 keV to be separable from the primary particle. Further studies of such a microdosimeter should be conducted through MC simulations to determine the necessary specifications for such a tracking microdosimeter. In summary, the measurements with the microdosimeters agrees well with simulations and can be an alternative to TEPCs. The microdosimeters small size makes them excellent for measurements at various depths in therapeutic beamlines such that the relative biological effectiveness (RBE) can be assessed. The microdosimeters are inexpensive to mass produce and they are easy to operate, this makes them readily available for use in conjunction with research, radiation therapy and radiation protection. The work presented here can support other users of the microdosimeter when planning, measuring and analysing results. This work also aids in the development of new and better microdosimeters.