<p>Arsenic (As) is an element that is well known for its toxic capabilities. It is odorless and colorless and is known to contaminate the drinking water of populations in several parts of the world. Routine monitoring of arsenic exposure is usually performed with urine, hair or nail, where samples are collected for laboratory analysis. Arsenic’s strong affinity to keratin rich tissues make skin another possible measurement site, in addition to the latter two tissues mentioned above. In some cases, skin samples are extracted for analysis. This is painful and invasive and is not ideal for <em>in vivo</em> monitoring of arsenic. The ability to quantify elemental concentration non-destructively is the major calling card of x-ray fluorescence (XRF). To that end, work was started on development of XRF detection systems for arsenic. The technique has shown promise for other elements and dramatic improvements in As detection capabilities were previously found when going from a radioisotope-based x-ray source to an x-ray tube based approach.</p> <p>This thesis documents the comparison of three x-ray tube based detection systems intended for the measurement of arsenic in skin. Two benchtop systems were used, with a) extended development of the previously assembled system and b) the first use of a separate detection system. Two handheld x-ray analyzers (portable detection systems) were also investigated in stand mode, where they were attached to a purpose-built mounting stand, provided by the manufacturer, during all analysis. Polyester resin phantoms were used to model arsenic in skin and a nylon backing was used to represent as bulk tissue behind skin. During the course of the work, modifications were made to the laboratory setup associated with the benchtop approaches.</p> <p>A benchtop polychromatic Mo anode x-ray tube based x-ray fluorescence (XRF) detection system was the first system used in this work. Through modifications to the existing design of the system, the lowest minimum detection limit (MDL) achievable was found to be (0.611±0.001) ppm normalized to gross scatter, where ppm is ug of arsenic per gram of dry weight (resin). The measurement time was ~1800 seconds real time. The equivalent (skin) and whole body effective doses delivered were (19±3) uSv and (163±47) uSv respectively. The corresponding direct (un-normalized) MDL was (0.499±0.002) ppm, in agreement with that found previously. Modifications to the system allowed a reduction in the localized effective dose delivered, to achieve this MDL, from (0.64±0.03) uSv previously to (0.14±0.04) uSv here.</p> <p>Next, the current work investigated two handheld x-ray analyzers provided by InnovX. A PiN diode detector based Alpha 4000S model unit (W anode x-ray tube) and a Silicon Drift Detector (SDD) based Delta model (Au anode x-ray tube). Both units were operated in benchtop mode: they were mounted in a stand and a phantom was placed on a kapton exit window. The lowest gross-scatter normalized and direct detection limit with the Alpha 4000S unit was (1.649±0.002) ppm and (1.651±0.002) ppm respectively. The equivalent and whole body effective doses delivered were found to be (9.4±2.2) mSv and (94±22) uSv respectively. The localized effective dose was (6.4±1.5) X 10<sup>-3</sup> uSv. By comparison, the Delta unit produced a gross-scatter and direct normalized detection limit of (0.570±0.002) ppm and (0.558±0.002) ppm respectively. The equivalent dose delivered was found to be (19.0±9.0) mSv. The corresponding localized and whole body effective doses delivered were (9.7±4.6) X 10<sup>-3</sup> uSv and (190±90) uSv respectively.</p> <p>The last system used in the current research was a monochromatic Ag anode x-ray tube based XRF setup. A doubly curved crystal (DCC) was used to select the Ag K-alpha line and focused the beam to a spot size of mm<sup>2</sup> at the focal length. The phantoms were placed at a farther distance where the beam had expanded to a larger area. The lowest Compton scatter normalized detection limit with the Si(Li) detector was found to be (0.696±0.002) ppm. After characterizing its performance in a range of energies, a silicon drift detector was also used on this system. It had the benefit of higher throughput capabilities and superior resolution. The housing of the detector was sufficiently small that it could be placed closer to the phantom surface than the Si(Li) detector. The lowest Compton-scatter normalized detection limit with the SDD was (0.441±0.003) ppm in 1800 seconds real time. The equivalent dose was found to be (11±2) mSv and the localized and whole body effective doses were found to be (3.92±0.87) X 10<sup>-3</sup> uSv and (110±23) uSv respectively. A significantly lower system dead time was observed with the SDD. Finally, Monte Carlo simulations of the system were performed to evaluate the performance of three ratios when their phantom measurement values were compared against simulations of skin. Results were found to be in agreement to withinin vivo concentration of arsenic in skin (ICRP).</p> <p>Finally, EDXRF measurements were performed on bulk cores of skin, <em>ex vivo</em>. While it was not possible to detect arsenic in the samples, due to the samples being collected from members of the public as opposed to an exposed population, a depth profile of numerous skin samples, starting from the surface and running straight down, was obtained for calcium, iron and copper.</p> / Doctor of Philosophy (PhD)
| Identifer | oai:union.ndltd.org:mcmaster.ca/oai:macsphere.mcmaster.ca:11375/14113 |
| Date | January 2014 |
| Creators | Desouza, Elstan |
| Contributors | McNeill, Fiona, Fleming, David, Chettle, David, Medical Physics |
| Source Sets | McMaster University |
| Detected Language | English |
| Type | thesis |
Page generated in 0.0024 seconds