Characterization of OSL dosimeters used at the ELI-beamlines laser-driven accelerator facility

The Extreme Light Infrastructure (ELI) Beamlines [1] is a state-of-the-art laser-driven accelerator facility located on the outskirts of the city of Prague, Czech Republic. It will operate the most intense non-military laser system with ultra-high power up to 10 PW, concentrated plasma intensities up to 10²⁴ W cm⁻², and ultra-short laser pulses of the order of few femtoseconds. This will open new frontiers for basic research as well as for areas such as medical imaging and diagnostics, radiotherapy, new materials, and x-ray optics. In-house experiments are already taking place since the first half of 2018. While first user calls were issued in 2019 and first user experiments were successfully performed later that same year.

A laser facility as unique as ELI Beamlines poses unique challenges from a radiation protection (RP) standpoint [2]. In fact, lasers of a few hundred TWs are already capable of producing a large amount of ionizing radiation when interacting with suitable targets [3, 4]. Therefore, to ensure the safety of personnel, users, general public, and the environment, a number of protective and preventive measures have been implemented. These include extensive Monte Carlo simulations, a comprehensive interlock system, personal dosimeters, and a robust and redundant radiation monitoring system. This radiation monitoring system is designed to monitor prompt radiation levels, activation, and contamination in the work place and surrounding areas. It includes several detector technologies, both active and passive.

Passive luminescence detectors are widely used in various areas [5] dealing with radiation detection and dosimetry, including personal and area monitoring, medical dosimetry, nuclear forensics, geochronology, and space dosimetry. Beryllium oxide optically stimulated luminescence (BeO-OSL) detectors are amongst the most used passive detectors exploiting luminescence phenomena. BeO-OSLs are an integral part of the radiation monitoring system at ELI Beamlines. They are placed throughout the experimental building and around the facility premises as part of a long term environmental monitoring effort. Moreover, additional ones are placed in the proximity of experimental station depending on the needs and requests of the RP group.

BeO-OSLs were chosen based on their favourable dosimetric characteristics [5] which include a high sensitivity to photon ionizing radiation, a linear dose response over six orders of magnitude (∼5 µGy to ∼5 Gy), a low effective atomic number (Zeff = 7.2) which makes it a near tissue-equivalent material, relatively low cost, and they do not need to be protected against electromagnetic pulses (EMPs). EMPs constitute, in fact, a major hazard at ELI Beamlines. Moreover, these detectors have been successfully tested in pulsed radiation fields [6].

The BeO-OSLs used at ELI Beamlines were supplied by Dosimetrics GmbH and Materion Ceramics. They are square chips of 4.7 × 4.7 × 0.5 mm³ and have a nominal density of 2.85 g cm⁻³. Annealing is achieved in air atmosphere by heating the detectors at 700 °C for 15 min in a laboratory chamber furnace by LAC s.r.o. This temperature is sufficient to empty deep traps in the materials and avoid photo-transfer [7]. Subsequently, but before irradiation, the detectors are covered with a 10 µm of aluminium foil to shield them from visible light. In fact, it has been proven that a residual signal is observed from the BeO-OSLs if they are exposed to sunlight or artificial light [8].

After irradiation, the BeO-OSLs are readout in a Lexsyg Smart reader from Freiberg Instruments GmbH. The signal estimation procedure reflects the one outlined in [9]. Each chip is stimulated for 120 s using blue light (wavelength 458 nm). In these conditions, >99% of the signal is read at the current stimulation power of 85 mW cm⁻². The readout signal is corrected for background and signal sensitivity to account for different production batches.

ELI Beamlines relies on the National Radiation Protection Institute (Czech acronym: SÚRO) to calibrate its BeO-OSL detectors. SÚRO is, in fact, an accredited calibration laboratory (ISO/IEC 17025:2017 [10]). The latest calibration campaign was carried out in May 2020. Five sets of 16 BeO-OSL chips were irradiated up to five different air kerma values. An extra set was used for the environmental background measurement. Due to operational reasons, two different Cs-137 sources were used. The source activity and all other relevant parameters are listed in table 1 for each irradiation point. In all cases, a 3 mm build-up polymethyl methacrylate layer was placed in front of the detectors. Each chip was readout separately and background and sensitivity corrections were applied. Subsequently, the mean signal of the set was taken and the standard deviation on the mean was estimated. The latter never exceeds 4%. The dosimeters were readout less than three days after irradiation. This is sufficiently adequate since it is known that the signal for these type of detectors fades by ∼5% during the first hour, but remains relatively stable after that [11]. Results are shown in figure 1. The BeO-OSLs show a good linear response over the entire air kerma range investigated. The response, R, of the OSL dosimeters to the delivered air kerma, Ka, is given by equation (1):

$$\begin{equation}R = \left[ {\left( {1.05 \pm 0.02} \right) \times {{10}^3}\mu {\text{G}}{{\text{y}}^{ - 1}}} \right] \cdot Ka.{ }\end{equation} \tag{ 1 }$$

Zoom In Zoom Out Reset image size

In addition, FLUKA Monte Carlo [12, 13] simulations were also performed to estimate the air kerma measured by the detectors. The CERN FLUKA release was used [14]. Unfortunately, it is not possible to obtain air kerma directly in FLUKA. Therefore, ambient dose equivalent was instead scored and converted to air kerma following the official recommendations in ISO 4037–3:2019 (Calibration of area and personal dosemeters and the measurement of their response as a function of energy and angle of incidence). Figure 2 shows an excellent agreement data vs. Monte Carlo.

Zoom In Zoom Out Reset image size

The energy dependence response of the BeO-OSL was as well tested. Six sets of 16 BeO-OSL chips were irradiated up to 1 mGy of air kerma using an Isovolt Titan x-ray tube at the SÚRO laboratories. An extra set was used for the environmental background measurement. The dosimeters were placed 184.2 cm from x-ray tube exit window. Six different photon energies were investigated using two different voltage settings (160 kV and 320 kV). The dosimeters were readout out following the same process described in the previous section. The results are shown in figure 3, where the average BeO-OSL signal is plotted as a function of the photon energy and normalised to the expected signal for 1 mGy of air kerma from Cs-137, as calculated by equation (1). The results show an underestimation of the accrued radiation for energies lower than ∼250 keV. A sigmoid function is fitted to the data. Thus, if the radiation energy is known, the underestimation can be corrected by calculations. The seen underestimation is consistent with absorbed dose energy dependence [16].

Zoom In Zoom Out Reset image size

BeO-OSL dosimeters have been used for many years and in various fields of applications thanks to their favourable dosimetric characteristics. These passive dosimeters are extensively used at ELI Beamlines where an exhaustive and thorough characterization process is in place. This allows for reliable and coherent radiation measurements. During the latest calibration campaign, a subset of the dosimeters was irradiated using Cs-137 sources. Their response was studied as a function of the delivered air kerma. After background subtraction and sensitivity correction, it was seen that the BeO-OSLs have a linear response over the entire investigated air kerma range. A calibration curve was estimated using a linear regression. FLUKA Monte Carlo simulation were performed and compared to data. An excellent agreement data vs Monte Carlo was seen. This proves a good understanding of the experimental setup and the ability of correctly describing the dosimeters in simulations. The latter is an asset for any future radiation protection study.

The energy response of the BeO-OSLs was also studied using an x-ray tube. The delivered air kerma was fixed to 1 mGy and photon energy was varied. The data were normalised to Cs-137 and fitted with a sigmoid function (see figure 3). An underestimation of the accrued radiation is seen for energies lower than ∼250 keV.