«William Geoffrey West A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy (Nuclear Engineering ...»
ABSTRACT Carbon-doped, anion-defective aluminum oxide has become a widely-used and effective medium for personnel dosimetry applications using optically stimulated luminescence (OSL) techniques. Though the commercial products currently using this material employ light-tight packaging to prevent light-induced effects on the OSL signal, the material could be employed in environments where package integrity cannot be assured. This paper reports on the results of an experiment performed to quantify the effects of sunlight exposure on α-Al2O3:C. Samples of commercially available Luxel® material were exposed to carefully-recorded levels of sunlight both before and after irradiations to determine the nature and magnitude of both activation and fading phenomena in this material. The results confirm that both fading and activation processes are seen in this material and indicate that the material reaches an equilibrium dose level in response to prolonged sunlight exposure equivalent to a dose of ~15 mGy under the experimental test conditions.
Optically stimulated luminescence (OSL) has become a commercially successful dosimetry method during the past 10 y, as evidenced by the success of the Luxel® dosimeter manufactured and distributed by Landauer, Inc. In addition, the OSL phenomenon is widely used in the dating of geologic samples and has been proposed for a variety of other applications.
The use of OSL as a dosimetry method is attractive for a number of technical reasons, some of which are: 1) the current material of choice for this application, αAl2O3:C, is highly sensitive in that it emits a large amount of stimulated luminescence per unit of radiation dose absorbed, 2) the optical read-out method is fast and relatively simple, and 3) the OSL technique lends itself to repeated read-out of samples.
Unfortunately, the large photo-ionization cross-section that makes alumina such a sensitive OSL dosimeter also introduces some complexities to its behavior which must be
studies have performed limited treatments of the effects of low-level ‘daylight’ exposure on this material(4-5). In addition, a seminal investigation of the wavelength dependence of light-induced fading of the material using monochromatic excitation methods was performed by Walker, et al(6). However, a quantitative and comprehensive study of the reaction of this material to direct sunlight exposure under controlled and repeatable conditions has not been performed to date. It is important for researchers and users of this material in real-world environments to know both the magnitude and nature of the
related to environmental or personnel dosimetry. Whether the material undergoes dosimetrically significant changes in 1 µs or 1 h is extremely important when a dosimetrist or physicist is attempting to determine whether a light exposure may or may not have occurred, analyzing a potential new application for the material that may involve natural light, or designing handling protocols for investigating or using this material in an outdoor setting. To date, no other study has provided this quantitative information in a controlled and fully-documented manner.
This experiment was designed to measure the amount of OSL signal fading due to direct sunlight exposure over several orders of magnitude immediately after a radiation exposure. In addition, this experiment also measured the amount of OSL signal induction caused by direct sunlight exposure on samples which had been previously bleached of their dosimetric OSL signal and were not subsequently irradiated. The results of this experiment may be used to better understand the effects of handling this material for even brief time periods in sunlit environments. In addition, this experiment also yields additional and potentially useful data for researchers interested in the phenomological characteristics of aluminum oxide’s OSL behavior.
MATERIALS Aluminum oxide (α-Al2O3:C) provided by Landauer, Inc. in the form of laminated
Luxel® material was used for this experiment. Twenty individual samples of α-Al2O3:C
were formed using a standard-size single hole punch to produce 6.35 mm (1/4”) diameter
alcohol to remove any potentially fluorescent contaminants before use.
METHODS This experiment was performed at the Los Alamos National Laboratory Luminescence Geochronology Lab. The experimental setup consisted of a Risø Model DA-15 Automated OSL/Thermoluminescence (TL) Reader System. This reader allows continuous blue light stimulation centered at 470 nm. The reader also had an integral beta irradiator employing a pneumatically activated 1.48 GBq 90Sr/90Y source, capable of delivering ~15 mGy s-1 to a sample. The detection subsystem of this OSL/TL reader consists of a low-background bi-alkali photomultiplier tube with a quartz window, operated in photon counting mode. U-340 filters are employed to filter the emission spectrum and these filters transmit a light bandwidth centered at 340 nm, with a fullwidth half-maximum value of ~80 nm. For additional technical information regarding the specifications of this commercially available system, the reader is referred to the technical thesis regarding its development and construction(7).
Though the alumina samples had been previously exposed to both room light and low-level background radiation for several months prior to the testing, all samples were first ‘cleared’ using the OSL reader by stimulation with blue light until no remaining dosimetric OSL signal above background was detectable. Due to the laminated nature of the Luxel® material, a total thermal annealing of the material is not possible. As such, it would be expected that the tested material could have a small residual deep trap
Appropriate treatment and consideration of this deep-trap population is addressed in the Discussion section of this paper. It is noted that the OSL signal output from the clearing sequences verified that the accumulated dose on the material since receipt from the OSL material manufacturer was less than 2 mGy in all cases. After clearing, ten of the samples were irradiated to 15 mGy using the OSL reader’s integral beta source. The other ten samples were not irradiated. All of the samples were then placed in identical, individually-labeled light-tight containers for transport from the laboratory to a suitably level and unobstructed outdoor location for the sunlight exposure. The sunlight exposures occurred on the grounds of Los Alamos National Laboratory during midsummer at an elevation of ~2200 m (7277 ft) above sea level. Meteorological conditions during the sunlight exposures were sunny with little or no clouds in the sky and no cloud cover of the sun itself. The light levels for the sunlight exposure were measured using a Quantum Instruments Photometer LX possessing a current calibration certificate. The light level during the exposures was 100,000 lux (± 5000 lux). The outdoor ambient temperature during the exposures was 40°C. When exposed, the samples were placed flat on a solid white piece of paper. Samples (one each of the irradiated and non-irradiated batches) were exposed to 0, 1, 5, 15, 30, 60, 300, 900, and 1800 s of unfiltered direct sunlight, respectively. Anticipated 3600 s samples were not used because of approaching cloud cover near the end of the sunlight exposure period.
After sunlight exposure, the samples were returned to their light-tight containers for transport back to the laboratory darkroom and immediately read out using both the infrared and blue excitation sources. All of the read-outs were continuous-illumination
points per read-out. The results of the read-out phases were collected by the OSL reader software and those data were analyzed and plotted.
RESULTS Figure 2.1 shows a composite plot of the blue-light read-outs of the samples which were illuminated by sunlight without prior irradiation. This graph clearly shows that as sunlight exposures increase, the magnitude of the resulting OSL signals increase, up to a point. Furthermore, in overlaying a plot of a sample irradiated to 15 mGy with no sunlight exposure, it becomes evident that the OSL signal’s readout ‘half-life’ for the sunlight-activated samples is considerably different from that of an irradiated sample of similar peak height. This discrepancy will be discussed further in the next section.
Figure 2.1: OSL response of a-Al2O3:C samples exposed to varying durations of direct sunlight.
Results shown for blank (opened squares), 1 s (open circles), 5 s (open hexagons), 30 s (open triangles), 5 min (open diamonds), 30 min (open inverse triangles) and a comparison sample irradiated to 15 mGy but with no sunlight exposure (open stars).
samples that were not irradiated) were dropped during darkroom handling after sunlight exposure and, based on their measurements, it is presumed that they were flipped over before reading. The opaque nature of the Luxel® material resulted in readings which were ~30% lower than would be expected based on the general trend. Because these samples were considered to be compromised, their corresponding data points are not shown in the figures. Figure 2.2 shows the change in the peak OSL signal (an average of the first three OSL channel readings) versus the amount of sunlight exposure in units of equivalent exposure versus lux-hours.
Note that because of the different shape of the radiation-induced OSL curve versus the sunlight-induced OSL curve, the choice of what measurement (peak height, area, or some other value) to use for equating a dose to an OSL signal is somewhat Figure 2.2: Peak OSL signal of a-Al2O3:C in units of equivalent dose versus total direct sunlight exposure.
duration peaks cannot be measured, the peak height is used as the reference value. Using this method, the data show a potentially significant induction of OSL signal in aluminum oxide roughly equivalent to 12 mGy at maximum.
The next portion of our experiment was the study of the impact of direct sunlight exposure on samples which had been irradiated just prior to their exposure. Figure 2.3 shows a composite plot of the blue-light read-outs of these samples. The time axis of this plot was truncated to 60 s for readability. The graph clearly shows that as sunlight exposure increases, a more complex change in the magnitude and shape of the OSL signal results. A plot of the OSL signal peak height versus sunlight exposure time of both the previously-irradiated and non-irradiated samples is shown in Figure 2.4. These figures will be discussed in the next section.
Figure 2.3: OSL response of a-Al2O3:C samples exposed to varying durations of direct sunlight after receiving a 15 mGy beta dose.
Results shown for blank (open squares), 1 s (open circles), 5 s (open hexagons), 15 s (open triangles), 30 s (open diamonds), 1 min (open inverse triangles), 5 min (open stars), 15 min (closed circles) and 30 min (closed squares).
DISCUSSION Based on the results obtained, it appears that direct exposure to sunlight produces both immediate and significant light-induced fading and activation of this material in time periods on the order of seconds and at magnitudes on the order of mGy.
These two processes appear to compete and eventually reach a saturation condition independent of whether the material had been dosed prior to sunlight exposure or not.
This saturation condition may be a steady-state equilibrium condition though our sunlight exposures did not extend to time periods long enough to verify this possibility. The saturation condition under the experimental conditions was roughly equivalent to an absorbed dose in the material of 12 mGy, which could certainly prove to be a confounding effect on environmental or personnel dosimetry measurements. A previous study briefly treated the daylight-induced activation of α-Al2O3:C, but this study was
daylight was direct or indirect and whether it was filtered or unfiltered(5). In addition, the material in that study had been exposed to a 2 Gy gamma dose and annealed to 600K immediately before the sunlight activation, resulting in a different initial condition for the material. This initial condition would not normally be expected to occur with the commercially available alumina-based dosimeter since the relatively low melting point of its polyester substrate precludes high-temperature thermal annealing. This previous study found an equivalent TL response of roughly 90 µGy at a daylight exposure of 100 luxhours, which is roughly a factor of 45 lower than our data. Another study(4) examined the fading of α-Al2O3:C when exposed to daylight filtered by normal window glass for 8 h.
This study found a residual OSL signal dose of 0.4 µGy. However, the ultra-violet (UV) filtering effect of the window glass as well as the possibility that the sunlight exposure was indirect and thus composed of a different spectral composition than direct sunlight could explain the difference in our signal induction findings. With respect to the well-documented importance of a material’s dose ‘pre-history’ to its subsequent OSL behavior, our experiment was designed to more closely approximate the conditions expected from use of the as-found commercial product for a period on the order of months. It is noted that the low level of background radiation experienced by the samples prior to the experimentation was small with respect to both the dose used during the irradiation portion of the experiment, as well as the equivalent doses realized by prolonged exposure to sunlight. As such, the effect of this particular element of prehistory on the material’s deep trap population, when compared with the effect from the
material approached an equilibrium condition.