TY - JOUR
AU1 - Kennedy, Konnor J
AU2 - LeBlanc, Alexandre
AU3 - Pirkkanen, Jake
AU4 - Thome, Christopher
AU5 - Tai, T C
AU6 - LeClair, Robert
AU7 - Boreham, Douglas R
AB - Abstract Living systems have evolved in the presence of naturally occurring ionising radiation. REPAIR is a research project investigating the biological effects of sub-natural background radiation exposure in SNOLAB, a deep-underground laboratory. Biological systems are being cultured within a sub-background environment as well as two control locations (underground and surface). A comprehensive dosimetric analysis was performed. GEANT4 simulation was used to characterise the contribution from gamma, muons and neutrons. Additionally, dose rates from radon, 40K and 14C were calculated based on measured activity concentrations. The total absorbed dose rate in the sub-background environment was 27 times lower than the surface control, at 2.48 ± 0.20 nGy hr−1, including a >400-fold reduction in the high linear energy transfer components. This modelling quantitatively confirms that the environment within SNOLAB provides a substantially reduced background radiation dose rate, thereby setting the stage for future sub-background biological studies using a variety of model organisms. INTRODUCTION Ionising natural background radiation (NBR) has been present at the surface of the planet throughout the evolution of biological life. There is growing evidence that this NBR may play an essential role for life on earth(1–4). This NBR originates from both cosmic and terrestrial sources, and it contributes to our background dose rate through external exposures as well as inhalation and ingestion(1). Although the biological effects of ionising radiation have been robustly studied at doses above NBR levels, there are limited studies that have investigated the biological effects of the absence of cosmic and terrestrial NBR(4–6). A limiting factor behind this type of research is the number of facilities where sub-NBR studies can be conducted due to the radiation shielding and infrastructure requirements. The major challenge to sub-NBR studies is how to reduce the cosmic radiation component of NBR(1, 7). A large portion of the cosmic radiation component is muons, produced through interactions of galactic cosmic radiation with Earth’s atmosphere(8, 9). A solution to this is conducting experiments in deep underground research facilities, where the inherent rock overburden acts as shielding against cosmic radiation. One such study is Researching the Presence and Absence of Ionizing Radiation (REPAIR), which is conducted at SNOLAB, a research laboratory located 2 km deep underground within an active mine in Sudbury, Ontario(1). The SNOLAB muon flux is reduced by a factor of 50 million from the surface(10). REPAIR is studying the hypothesis that sub-NBR levels of radiation are detrimental to biological life by experimentally studying these systems both above ground and below within SNOLAB. An additional challenge presented with an underground research environment is to efficiently reduce the inherently elevated levels of radon gas and gamma radiation that emanate naturally from surrounding rock(1, 2, 10–12). To this end, REPAIR has designed, engineered and commissioned a specialised tissue culture incubator (STCI) for biological experimentation underground within SNOLAB(7). To reduce radon, the air fed into the STCI is aged as to allow significant decay of 222Rn given its 3.8-d half-life. The STCI is completely sealed from the higher radon laboratory environment and is equipped with an airlock and purge mechanism for bringing samples into its interior(7). Within the STCI there is a ‘lead castle’ constructed with lead bricks. The interior of the lead castle is maintained at appropriate environmental conditions for cell culture growth (37°C, 5% CO2). To examine the biological effects of sub-NBR, REPAIR is culturing cells within the STCI in SNOLAB, which acts as the ‘sub-background’ experimental environment. Cells are also simultaneously being cultured at the Northern Ontario School of Medicine (NOSM) in a standard biological incubator, which acts as the normal NBR ‘surface control’. One environmental variable that cannot be controlled for in the experimental setup is air pressure. Air pressure levels within SNOLAB can be up to 1.2 times higher than on surface(10). To account for potential influences of pressure, a second set of control cells are being cultured within SNOLAB in a standard biological incubator that acts as the ‘underground control’. Oxygen, nitrogen and carbon dioxide are tightly controlled at the same concentrations across all three incubators. These underground cells have a reduced cosmic ray background but are exposed to higher levels of radon. The underground control also accounts for any influences that transportation of cells to and from SNOLAB might have. In order to study the effects of a sub-NBR environment, an accurate characterisation of the background radiation dose rates in all three experimental setups is required. Radiation exposure in each of the three environments occurs due to a combination of natural sources of radon, gamma, neutron and muons. However, the underground and surface environments will differ in the relative contribution of each radiation type(1). For example, underground environments tend to have much higher concentrations of radon and as a result have larger exposures to high linear energy transfer (LET) alpha particles, whereas the above ground dose is primarily due to low LET muons(1, 11, 13). Due to the extremely low dose rate from several NBR components, particularly within the STCI, direct measurements of radiation levels using dosemeters is not practical. The length of time that it would take to achieve a measurable dose would be large. Therefore, the dose rate must be estimated through simulations. To accomplish this large-scale characterisation, the geometry and tracking (GEANT4) simulation toolkit was used to construct a particle transport simulation using a varied number of sources, all based on real-world designs and gamma, neutron and muon flux measurements(14). GEANT4 utilises Monte Carlo methods to simulate the generation of various particle types and energies in a randomised distribution. An additional source of background radiation is from isotopes present in the liquid cell culture nutrient growth media, specifically the beta radiation component of 40K and 14C(12). Dose rates from radon, 40K and 14C were calculated separately based on their measured activity concentrations. METHODS Environments REPAIR utilises three different radiation environments to perform radiobiological experiments. The radiation environments are as follows: Surface control: This is a standard CO2 cell culture incubator that is located at the NOSM. Cells in this incubator are exposed to surface NBR. Underground control: This is a standard CO2 incubator that is located within the Life Sciences Laboratory in SNOLAB. The underground incubator controls for the higher-pressure levels within SNOLAB as well as acting as a transportation control. Cells in this incubator are shielded from cosmic radiation but are exposed to higher levels of radon and terrestrial gamma radiation. Sub-background: This is the STCI that was constructed in SNOLAB to achieve a sub-background environment by reducing cosmic radiation, radon gas and gamma rays. The background radiation dose rate in each of the three environments was estimated based on six different sources of radiation: gamma rays, neutrons, muons, radon, 40K and 14C. The dose rate from gammas, neutrons and muons were determined by GEANT4 simulation. Flux measurements for each of these three radiation types was previously taken in the underground laboratory and at the surface laboratory. Simulations were performed to determine the actual dose rate within the incubation environments (sub-background, surface control, underground control) due to attenuation through the materials they are constructed from. The dose rates from radon, 40K and 14C were calculated based on activity concentration measurements that were taken directly within the experimental environments or within the cell culture media, which is a liquid supplemented with organic compounds that the cells feed off for proliferation. GEANT4 model To determine the dose rates within the three environments from gamma rays, neutrons and muons, a detector geometry was first constructed within GEANT4. For the sub-background location, the simulation environment was constructed using dimensional blueprints of the STCI(7). The STCI exterior was constructed using stainless steel panels with an acrylic window. The dimensions of the STCI within GEANT4 were 152.4 cm across (x-direction), 87.31 cm high (y-direction) and 86.17 cm deep (z-direction). The thickness of the stainless steel and acrylic walls was 0.5 cm. The interior of the STCI contains a lead ‘castle’, inside which the biological samples are contained. The dimensions of the internal volume of the lead castle are 60.96 cm across, 30.48 cm high and 20.32 cm deep. The thickness of the lead bricks is 10 cm. The geometries of the surface and underground control environments were approximated to be the same dimensions as the external steel geometry of the STCI, without the 10-cm lead castle. Additionally, the acrylic window was removed in the simulation, and replaced with stainless steel, given that there is no acrylic window in standard incubators. The scoring volume in the sub-background simulation was the entire interior of the lead castle. An identical-sized scoring volume was used in the surface and underground control environments. Water was used as the material in the scoring volume, given that human cells are predominantly composed of water. Around the periphery of the scoring volume, a 1.2-mm-thick polystyrene layer was applied, which was to simulate the material composition of the wall of the culture flasks in which biological samples are grown. In order to make each wall specific to the thickness in question, rectangular prisms were constructed and subsequently rotated vectorially around the scoring volume (which was placed effectively at the origin). The resultant HepRep visualisation is shown in Figure 1. Figure 1 Open in new tabDownload slide HepRep visualisation of the sub-background and control environments. (A) The sub-background environmental geometry and the lead bricks that encase the scoring volume. (B) 2D visualisation of the lead bricks encasing the flasks and scoring volume. (C) Geometry without the lead bricks used for the underground control and surface control simulations. (D) 2D visualisation without the lead bricks used for the underground and surface control simulations Figure 1 Open in new tabDownload slide HepRep visualisation of the sub-background and control environments. (A) The sub-background environmental geometry and the lead bricks that encase the scoring volume. (B) 2D visualisation of the lead bricks encasing the flasks and scoring volume. (C) Geometry without the lead bricks used for the underground control and surface control simulations. (D) 2D visualisation without the lead bricks used for the underground and surface control simulations GEANT4 sources Monte-Carlo simulations using GEANT4 were performed on the three environmental scenarios to determine the background radiation dose rate for each of the three particle types. GEANT4 10.5 was used in this study. The QBBC physics list(15) was used due to its versatility in accurately simulating the various particle types that are present in each laboratory environment. The global particle source (GPS) source geometry consists of a sphere that encompasses the entire scoring volume geometry in x, y and z, with a radius equivalent to the largest length of the detector geometry. To ensure that particles are generated as close to our geometry as possible, a rectangular prism of air was constructed to encompass the entire detector geometry and was designated as the mother volume of every single subsequent logical wall volume. The GPS source was then set to confine particle generation on the surface of this air prism in a cosine angular distribution. This source methodology is very similar to those employed by Lampe et al.(16). This angular distribution ensures that the source sphere is Lambertian in intensity, which means that the intensity at any given observed point is going to be uniform and truly isotropic(17). A normalisation factor was applied since there is a difference in surface area between the scintillator that took the measurements as described previously by Peracchi et al. and Lund and Jevremovic(18, 19). The normalisation is as follows: $$\begin{equation} \dot{N_r}=\dot{N_s}\times \frac{\phi_r}{\phi_s} \end{equation}$$(1) With |$\dot{N_r}$| representing the actual normalised particle rate, |$\dot{N_s}$| representing the simulated particle rate within GEANT4, and |${\phi}_r$| and |${\phi}_s$| representing the real and simulated fluences of gamma photons, respectively. The units of fluence being in m−2 in this case. Since the particle number generated is assumed to be equal in the underground control environmental scenario (in terms of N particles), and therefore cancel out of the equation, the normalisation formula then becomes: $$\begin{equation} \dot{N_r}=\dot{N_s}\times \frac{SA_s}{SA_r} \end{equation}$$(2) With |${SA}_{\mathrm{s}}$| indicating the simulated surface area (the rectangular air prism), and |${SA}_r$| the surface area of the scintillator used to take the measurements. The formula for the surface area of the rectangular prism is as follows: $$\begin{equation} SA=2\left( xz+ yx+ yz\right) \end{equation}$$(3) With x representing the length of either the scintillator crystal or the air prism, y the height and z the width all of which are in m. The resulting units being m2. Gamma Flux measurements for gamma rays were previously taken in SNOLAB and at the surface control laboratory (Table 1)(7). The gammas originate from isotope decay within rock, as well as higher energy gammas from the atmosphere (at surface). The energies ranged from 0.7 to 26.5 MeV. The measurements were taken using a Sodium Iodide (NaI) scintillation detector (Saint-Gobain, Crystal Model# 4X4H16-NP) as previously described by Pirkkanen et al.(7). The measurements were taken over the period of 6.76 and 6.93 d for the underground and surface control environments, respectively, with more than 75 million gamma particles measured. The isotope decay peaks were then fit using Interspec (Sandia National Laboratory v1.0.5) and Cambio (Sandia National Laboratory v120822) software and channel to energy conversion coefficients formed, based on the 40K and 208TI peaks. From here, a probability histogram was constructed for both the above and below ground spectra by simply taking the ratio of counts per energy bin divided by total counts in all bins to generate probability of emission at specific energies. The resultant histogram was then input into the GPS macro file as a user defined histogram. For every run, 150 million gamma particles were simulated. Given the differing particle numbers and live times as collected at SNOLAB and the surface laboratory reported by Pirkkanen et al.(7), the resultant GEANT4 absorbed dose was normalised to the time it took each respective spectra to generate these particles. The resultant dose rates were simulated five times and their respective error was calculated based on the standard deviation between each iterative simulation. Table 1 Gamma, neutron and muon flux measurements used in GEANT4 simulations. Particle type . Surface (m−2 min−1) . Underground (m−2 min−1) . Gamma(7) 4.1 × 104 8.2 × 104 Fast neutron(10) 6.0 × 103 2.8 × 100 Thermal neutron(10) — 2.9 × 100 Muon(10) 1.0 × 104 1.9 × 10−4 Particle type . Surface (m−2 min−1) . Underground (m−2 min−1) . Gamma(7) 4.1 × 104 8.2 × 104 Fast neutron(10) 6.0 × 103 2.8 × 100 Thermal neutron(10) — 2.9 × 100 Muon(10) 1.0 × 104 1.9 × 10−4 Open in new tab Table 1 Gamma, neutron and muon flux measurements used in GEANT4 simulations. Particle type . Surface (m−2 min−1) . Underground (m−2 min−1) . Gamma(7) 4.1 × 104 8.2 × 104 Fast neutron(10) 6.0 × 103 2.8 × 100 Thermal neutron(10) — 2.9 × 100 Muon(10) 1.0 × 104 1.9 × 10−4 Particle type . Surface (m−2 min−1) . Underground (m−2 min−1) . Gamma(7) 4.1 × 104 8.2 × 104 Fast neutron(10) 6.0 × 103 2.8 × 100 Thermal neutron(10) — 2.9 × 100 Muon(10) 1.0 × 104 1.9 × 10−4 Open in new tab Muons and neutrons The flux values for neutrons and muons were previously measured within SNOLAB (Table 1)(10). The muons and neutrons were simulated in an isotropic fashion, from a circular surface source of radius 73 cm placed 44 cm above the sub-background environmental geometry, similar to the methods used by Lampe et al.(16, 20). This modification to the source geometry was done so that particles arriving at the source volume geometry would as closely model cosmic rays as achievable. Additionally, 54% μ+ and 46% μ− were simulated separately and then summed at the end of the simulation(20). The average muon energy at sea level is 4 GeV(21), which is what was simulated in the model. Fast neutrons were simulated to have an energy of 10.5 MeV (assuming a range of 1–20 MeV) and thermal neutrons were simulated to have an energy of 0.025 eV(22). In contrast to the simulated gamma photons, 1.5 million particles were simulated to ease computation time. The dose rate in between runs showed a relative error under 1% and simulating further particles would not lead to an appreciable amount of greater precision. The same time correction as was applied to the gammas was also applied to the neutrons and muons. The resultant dose rates were simulated five times and their respective error was calculated based on the standard deviation between each iterative simulation. Radon Radon concentrations were measured in each of the three environments in order to calculate the radon dose rate. Radon was measured using a RAD7 real-time radon monitor (DURRIDGE Company Inc., USA) as previously described(7). Radon was measured directly within the STCI, as well as in the surface and underground laboratory air. Since the access door to the above ground and underground incubators are routinely opened, it was assumed that the radon concentrations within the incubators matched the laboratory air concentration. The radon concentrations were 123.45 ± 15.53 Bq m−3 for the underground control, 3.67 ± 2.14 Bq m−3 for surface control and 0.79 ± 0.93 Bq m−3 for sub-background. The assumption was made that only radon gas (and not radon decay products present in the laboratory air) would diffuse into the cell culture media (liquid the cells are grown in), and that the radon concentration in the media would be the same as the in-air measurement. It was then assumed that once radon had diffused into the media, the alpha or beta emitted from 222Rn and each successive daughter product (up to 210Pb that has a half-life of 22 y) would deposit all of its energy in the scoring volume. The formula used to calculate the radon dose rate was: $$\begin{equation} \dot{D_R}={\sum}_{i=0}^n{A}_i{E}_i \end{equation}$$[4] where |$\dot{D_R}$| represents the absorbed dose rate from radon, |${A}_i$| the activity concentration of the ith isotope (in Bq kg−1) and Ei the energy of the particle released from the ith isotope. The activity concentration of each isotope (⁠|${A}_i$|⁠) was the same as the measured in-air concentration of radon gas. The particle energies (⁠|${E}_i$|⁠) were as follows: a 5.59 MeV alpha particle for 222Rn, a 6.11 MeV alpha particle for 218Po, a 0.29 MeV beta particle for 214Pb, a 0.66 MeV beta particle for 214Bi and a 7.83 MeV β particle for 210Pb(26). 40K and 14C The cell culture media in which cells are grown contributes dose primarily from the beta decay of 40K and 14C. The activity concentration of each isotope was measured in a 48.8 g sample of culture media. The 40K activity was measured using a high-purity germanium detector (Canberra 2011, Canberra Coaxial Detector) at SNOLAB as previously described(23). The sample was left in the detector for 6.819 d, and the count rate was established at the end of the counting cycle. The methods used for determining the sample activity have been previously described(23). Briefly, the measurement window was centred where the specific energy was expected to occur, and the signal was the total number of counts within this window. A background was determined by taking windows containing the same number of channels above and below the energy peak. A net peak/count rate was then determined by subtracting out the counts of these background windows. The 14C measurements were performed at the André E. Lalonde Accelerator Mass Spectrometry Laboratory at the University of Ottawa as previously described(24). Briefly, the sample was combusted using a Thermo Flash 1112 elemental analyser, and the CO2 trapped within a Pyrex seal. The combusted liquid sample was then analysed by a 3 MV tandem accelerator mass spectrometer as outlined in Kieser et al.(25). Count rates were then established from the sample. The concentration of 40K measured in the cell culture media was 8005.1 ± 628.9 mBq kg−1 (mean ± standard deviation between counts). The concentration of 14C measured in the culture media was 612.2 ± 4.0 mBq kg−1. To simplify the calculations, it was assumed that all of the energy from each beta particle was deposited within the scoring volume. Additionally, the gamma component of the beta decay was not considered in the model since ~ 90% of the decays emit only a beta particle. Equation 5 was used to calculate the 40K and 14C dose rate in the exact same fashion as radon, with the activity concentration in (Bq kg−1) as well as using the average energy of the beta particles from ICRP 107 (40K = 522 keV and 14C = 49.5 keV)(26). Since the cells in all three environments are grown in the same culture media, |$\dot{D_I}$|was the same. The formula used to calculate the 40K and 14C dose rate was: $$\begin{equation} \dot{D_I}=A\overline{E} \end{equation}$$[5] with |$\dot{D_I}$| representing the dose rate of both the 40K and 14C isotopes, A representing the activity concentration (in Bq kg−1) and |$\overline{E}$| the average energy of the beta emission from ICRP 107(26). Direct dosimetry To verify the accuracy of the model, dose rates were directly measured in the surface control laboratory using a pressurised ionisation chamber survey meter (451P, Fluke Biomedical). Measurements were taken in three different locations: within the cell culture incubator with the door closed, within the laboratory outside of the incubator and outside on the roof of the building. Two separate measurements were taken at each location over 24 h. Given that the ionisation chamber measures gammas and X-rays (>25 keV) as well as charged particles (>1 MeV), the measured dose was considered to capture the gamma and muon component of NBR. RESULTS The calculated dose rates for each particle type as well as the total dose rates are shown in Table 2. The dose rate from gamma rays includes primary gamma photons that make their way into the scoring volume, as well as the large cascade of secondary electrons that are produced from interactions in stainless steel layer, the lead castle (for the STCI) and the polystyrene layer. The total gamma dose rate was the highest in the underground control environment, where it was 1.33 times greater than on surface. The dose rate was substantially lower in the sub-background environment by several orders of magnitude. The decrease in neutron dose rate was quite visible in the two underground locations, compared with the surface control, given that there is a full three order of magnitude reduction in neutron flux below ground in SNOLAB. Within the sub-background environment, there was a 2.66-fold reduction in neutron dose rate compared with the underground control and a 2675-fold reduction compared with the surface control. The underground and sub-background muon flux was on the order of fGy hr−1 (10−15 Gy hr−1). As a result, for both the sub-background and underground control environments the muon dose rate was considered negligible. The muon dose rate for the surface control was 55.27 ± 0.40 nGy hr−1. Table 2 Absorbed dose rate calculations for the three radiation environments. Dose rates were calculated in a water scoring volume. Particle type . Surface control (nGy hr−1) . Underground control (nGy hr−1) . Sub-background (nGy hr−1) . Gamma 5.78 ± 0.03 7.67 ± 0.01 0.0427 ± 0.0013 Neutron 4.52 ± 0.04 0.0045 ± 0.0002 0.00169 ± 0.00002 Muon 55.27 ± 0.40 Negligible Negligible 222Rn 0.044 ± 0.014 1.45 ± 0.17 0.009 ± 0.011 40K 2.41 ± 0.19 2.41 ± 0.19 2.41 ± 0.19 14C 0.0175 ± 0.0001 0.0175 ± 0.0001 0.0175 ± 0.0001 Low LETa 63.48 ± 0.62 10.10 ± 0.20 2.47 ± 0.19 High LETb 4.56 ± 0.05 1.45 ± 0.17 0.01 ± 0.01 Total 68.04 ± 0.67 11.55 ± 0.37 2.48 ± 0.20 Particle type . Surface control (nGy hr−1) . Underground control (nGy hr−1) . Sub-background (nGy hr−1) . Gamma 5.78 ± 0.03 7.67 ± 0.01 0.0427 ± 0.0013 Neutron 4.52 ± 0.04 0.0045 ± 0.0002 0.00169 ± 0.00002 Muon 55.27 ± 0.40 Negligible Negligible 222Rn 0.044 ± 0.014 1.45 ± 0.17 0.009 ± 0.011 40K 2.41 ± 0.19 2.41 ± 0.19 2.41 ± 0.19 14C 0.0175 ± 0.0001 0.0175 ± 0.0001 0.0175 ± 0.0001 Low LETa 63.48 ± 0.62 10.10 ± 0.20 2.47 ± 0.19 High LETb 4.56 ± 0.05 1.45 ± 0.17 0.01 ± 0.01 Total 68.04 ± 0.67 11.55 ± 0.37 2.48 ± 0.20 aLow LET = Gamma, Muon, 40K, 14C. bHigh LET = Neutron, 222Rn. Open in new tab Table 2 Absorbed dose rate calculations for the three radiation environments. Dose rates were calculated in a water scoring volume. Particle type . Surface control (nGy hr−1) . Underground control (nGy hr−1) . Sub-background (nGy hr−1) . Gamma 5.78 ± 0.03 7.67 ± 0.01 0.0427 ± 0.0013 Neutron 4.52 ± 0.04 0.0045 ± 0.0002 0.00169 ± 0.00002 Muon 55.27 ± 0.40 Negligible Negligible 222Rn 0.044 ± 0.014 1.45 ± 0.17 0.009 ± 0.011 40K 2.41 ± 0.19 2.41 ± 0.19 2.41 ± 0.19 14C 0.0175 ± 0.0001 0.0175 ± 0.0001 0.0175 ± 0.0001 Low LETa 63.48 ± 0.62 10.10 ± 0.20 2.47 ± 0.19 High LETb 4.56 ± 0.05 1.45 ± 0.17 0.01 ± 0.01 Total 68.04 ± 0.67 11.55 ± 0.37 2.48 ± 0.20 Particle type . Surface control (nGy hr−1) . Underground control (nGy hr−1) . Sub-background (nGy hr−1) . Gamma 5.78 ± 0.03 7.67 ± 0.01 0.0427 ± 0.0013 Neutron 4.52 ± 0.04 0.0045 ± 0.0002 0.00169 ± 0.00002 Muon 55.27 ± 0.40 Negligible Negligible 222Rn 0.044 ± 0.014 1.45 ± 0.17 0.009 ± 0.011 40K 2.41 ± 0.19 2.41 ± 0.19 2.41 ± 0.19 14C 0.0175 ± 0.0001 0.0175 ± 0.0001 0.0175 ± 0.0001 Low LETa 63.48 ± 0.62 10.10 ± 0.20 2.47 ± 0.19 High LETb 4.56 ± 0.05 1.45 ± 0.17 0.01 ± 0.01 Total 68.04 ± 0.67 11.55 ± 0.37 2.48 ± 0.20 aLow LET = Gamma, Muon, 40K, 14C. bHigh LET = Neutron, 222Rn. Open in new tab The absorbed dose rate from radon gas was the highest in the underground control and markedly lower in the sub-background environment by a factor of 161-fold. The dose rate from beta decay in the cell culture media is constant across all three environments, given that cells are cultured in the same media in each location with the same concentration of 40K and 14C. The dose rates were 2.41 ± 0.19 nGy hr−1 for 40K and 0.0175 ± 0.0001 nGy hr−1 for 14C. Overall, the highest total absorbed dose rate was the surface control environment, which is primarily due to the large muon flux. The total absorbed dose in the underground control was 6 times lower than the surface control. In the underground control, the total absorbed dose rate was predominantly a mixture of gamma, alpha and beta. The absorbed dose rate in the sub-background environment was the lowest at 2.48 ± 0.20 nGy hr−1, 27-fold lower than the surface control. Almost all of the dose in the sub-background environment was from beta decay in the culture media (~97%). In all three environments, the majority of absorbed dose resulted from low LET radiation. However, the high LET components had a much greater reduction in the sub-background environment compared with both control locations, where it was over 100-fold lower. The measured dose rates were highest outside on the roof of the surface laboratory at 95.83 ± 5.89 nGy hr−1. The roof of the building and the walls of the cell culture incubator provided some shielding, resulting in slightly lower dose rates in the other two locations. The dose rate inside the laboratory and inside the cell culture incubator were 83.33 ± 0.01 nGy hr−1 and 80.83 ± 0.59 nGy hr−1 respectively. The measured dose rate within the cell culture incubator matched closely to the GEANT4 modelled dose rate from gamma and muons in the surface control, which was 61.05 ± 0.43 nGy hr−1 (Table 2). DISCUSSION Dosimetry simulations were performed to quantify the background radiation dose rate within the REPAIR project’s newly established sub-NBR environment in SNOLAB, as well as in the two control locations. The GEANT4 particle physics simulation was used to quantify the dose rate from gamma rays, muons and neutrons. The radon, 40K and 14C dose rates were calculated using activity concentration measurements. The results of these calculations and simulations confirmed that the sub-background environment in SNOLAB within the STCI has a dose rate below what is found in the surface or underground control environments. Previous biological experiments conducted by REPAIR showed altered growth and hatch weights of lake whitefish embryos reared within the SNOLAB underground environment compared with surface(27). However, the higher levels of radon gas and gamma contaminants found underground within SNOLAB were unable to be controlled for at the time of these experiments. These factors were hypothesised to have influenced these effects, as we have previously shown that low doses of ionising radiation can increase whitefish morphology and growth(28). The need to control these increased NBR components (gamma and radon) was a significant impetus for the construction of the sub-background STCI. Gamma The total gamma dose rate within the sub-background environment showed a drastic decrease compared with the surface and underground control environments, with a fold reduction of 135.36 and 179.63, respectively. This was expected since the lead castle geometry in the STCI was designed to reduce > 99% of the gamma flux within SNOLAB(7). The sub-background environment gamma dose rate of 0.0427 ± 0.0013 nGy hr−1 is the lowest gamma dose rate reported in the sub-NBR literature. Many deep underground labs such as the Gran Sasso National Laboratory (LNGS) and the Waste Isolation Pilot Project (WIPP) report shielded gamma dose rates in the range of 3.6–15.6 nGy hr−1, respectively(3, 29). WIPP also interestingly used KCl salt (and the secondaries/gammas produced by 40K) as the primary means of simulating a natural background environment within their underground facility, where a full cell culture environment is housed within a pre-WWII era shipping container. Lampe et al.(16) (based out of Modane, France) reported an absorbed gamma dose rate of <1 nGy hr−1, and the radiation research centre out of Osaka Prefecture University reported a shielded gamma dose rate of 2.1 nGy hr−1(30). The Centre National de la Recherche found that their gamma dose rates were of the range of 0.1–0.27 mGy y−1 (or 11.4–30.8 nGy hr−1), which also used 10-cm lead bricks surrounding their incubation environments, similar to what is used by REPAIR at SNOLAB(31–33). The Subsurface Experiment of Life in Low Radiation (SELLR) group measured 0.5 nGy hr−1 within their experimental setup, which used a similar lead castle apparatus as REPAIR(34). The geographic locations, rock compositions as well as detection methods all varied within these studies, with the majority electing to utilise Thermoluminescent dosimeter (TLD)-based methods(3, 4, 16, 30). TLDs, as well as many other types of dosemeters, do not typically have the sensitivity to accurately quantify sub-NBR levels of radiation. Given that sub-NBR dose rates are low, it necessitated using simulation-based means to characterise the environment. Neutron The neutron dose rate in the sub-background environment was slightly lower compared with the underground control environment; however, the fold reduction was not nearly to the same degree as was seen with gamma rays. This was expected since lead is not an effective shielding material against neutron radiation. The Modane lab simulations yielded a surface control dose rate similar in magnitude to the one reported here, namely 4.4 vs 4.52 nGy hr−1, respectively(16, 20). Their shielded neutron dose rate was reported as much <1 nGy hr−1, similar to the results generated here (0.00169 nGy hr−1) for the sub-background environment. Other groups have constructed shielded cell incubators that were coated in paraffin, which reduced the dose to 2 from 8 nGy hr−1(30). As the neutron flux was already 1000-fold lower underground within SNOLAB, a low Z shielding material such as water or paraffin wax was not considered necessary in the design of the STCI. Muon Muons were simulated in a similar geometric fashion to the Modane group(20). Our calculated surface control dose rate of 55.27 nGy hr−1 is comparable to their 45 nGy hr−1. The attributable difference is likely due to the different spectral inputs for the muons. Like many of the other studies that operate deep underground laboratories, the muon flux underground was found to be negligible in the underground and sub-background environments given that the dose was on the order of 1 fGy hr−1(3, 4). Many labs have elected to use pre-established altitude measurements of cosmic muon flux to establish absorbed dose rates, such as those reported by the United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR), which are slightly lower than our calculated dose rate(4, 35, 36). Radon The high radon concentrations in SNOLAB substantially increased the high LET dose rate in the underground control compared with the sub-background environment. The ambient radon concentration within SNOLAB is almost 34-fold higher compared with the surface and accounts for one of the largest proportions of deposited dose rate within the underground control environment. The ambient radon concentration at SNOLAB is very similar to the tunnel of the LNGS (50–120 Bq m−3)(6). The dose rate attributed to the LNGS is similar to the dose rate reported in this paper (0.17 and 1.7 nGy hr−1 at LNGS)(4, 5). However, the LNGS radon dose rate is based on a model by Jostes et al.(37), which is a hypothetical measure of how damaging radon is to a single cell. WIPP had lower ambient radon concentration than SNOLAB (15.6 Bq m−3). WIPP’s dose rate, based on a Monte-Carlo N Particle simulation, ranged from 1 to 4 pGy hr−1, an order of magnitude lower than the dose rate reported here. WIPP also had a 4.25 times lower radon concentration in their surface control laboratory compared with our surface control(38). Other experiments have elected not to include radon in their dosimetry considerations(16, 20). This was due to their particular model system being tightly closed in a test tube. In many studies, including REPAIR, the experimental organisms are cultured within vessels that have active air exchange with the surrounding environment. As a result, a contributing dose from the increased levels of radon in the air is unavoidable in the underground control environment. However, the STCI is successful at mitigating ambient radon concentration within the cell growth environment. 40K and 14C The largest contributor to the sub-background environment dose rate is from 40K in the cell culture media. Our calculated dose rate of 2.41 ± 0.19 nGy hr−1 is similar in magnitude to that estimated by the SELLR experiment in the UK(34). The SELLR group were using bacterial models, and their specific activity measurements included bacterial samples within the measured culture media. The 40K measurements in this paper were performed with liquid cell culture nutrient media, which included calf serum as well penicillin/streptomycin. Many previous sub-NBR studies had not considered the dose rate effects of the growth media and primarily focused on reducing the environmental radiation dose rates(3, 5). The WIPP group examined 40K in the growth media, with their dose rate being slightly lower to the REPAIR and SELLR groups (0.75 nGy hr−1 in Castillo and Smith(39)). Other literature has reported that 40K contributed 99% of the dose to their particular sub-background model system(35), which is very similar to what is seen here, with 40K accounting for ~ 97% of the sub-background dose rate. The Modane lab found that their 40K dose rate was 26 nGy hr−1(24,); however the constituents of their culture media contained much higher activity per kilogram than the culture media used by REPAIR. LNGS found that since the 40K did not exhibit a substantial signal above background, they equated the 40K activity in cells to that of within a human, producing a large value (19 nGy hr−1)(4). It is important to note the various different deep underground laboratories use many different biological models with different amounts of potassium within the culture media. There is also some absorbed dose given by 14C, which is present in all organic life on earth. Many of the previously mentioned deep underground research group do not consider the absorbed dose rate from 14C, which is understandable given that it does not contribute large amounts to the overall dose rate. The Modane lab considered 14C and found a dose rate of much less than 1 nGy hr−1(16), similar to what is reported here. The NBR environment at the surface was directly measured using an ionisation chamber. Given the agreement between the measured and simulated dose rates (80.83 vs 61.05 nGy hr−1), this lends a degree of confidence to the simulated values. Our modelled dose rates are also in agreement with the dose rates reported by UNSCEAR, who estimate a cosmic dose rate at sea level of 31 nGy hr−1 from charged particles and photons(36). Our surface laboratory is at an elevation of ~350 m above sea level. According to UNSCEAR, at that elevation, ~75% of cosmic ray dose is from muons, 15% from photons/electrons and 10% from neutrons(36). In our modelled data, muons make up 84%, gammas make up 8.8% and neutrons 6.9%. These numbers are within acceptable ranges, considering the assumptions and approximations that were included in our model, as well as uncertainties in the UNSCEAR estimates based on the specific geographic location of our surface control laboratory. When solely considering the amount of terrestrial and cosmic radiation (ignoring the contribution from isotopes in the cell culture media), the sub-background environment has a much more substantial reduction in background dose rate compared with the underground and surface control. Removing 40K and 14C from the dose calculation in Table 2 produces a sub-background absorbed dose rate that is over 1000-fold lower than the surface control. There exists the potential to conduct experiments with 40K and 14C eliminated from culture growth media. This has been performed in the past, and there is a pronounced negative effect on the growth rate of cells cultured in 39K media compared with 40K(40), similar to what is shown when other components of NBR are removed. When either higher concentrations of 40K were reintroduced to the growth media or cells were exposed to a radiation source, they recovered normal growth rates(40). The 39K avenue is significantly more expensive than standard cell culture media; however, it is something that will be explored in future experiments conducted by REPAIR. CONCLUSION The STCI was designed and engineered to create a sub-background environment appropriate for biological experimentation in the absence of naturally occurring ionising radiation. The GEANT4 simulation and measurement-based calculations support that there is a substantial decrease in dose rate in the sub-background environment compared with surface and underground control environments. The lead castle reduces gamma radiation within the underground lab environment, and the radon mitigation strategies within the sub-background environment reduce radon dose rates. In conclusion, the results of this dosimetric modelling lend confidence that future experiments conducted by REPAIR are occurring within what is truly a sub-NBR and novel environment. FUNDING The REPAIR project is supported through a Natural Sciences and Engineering Research Council of Canada (NSERC) Collaborative Research and Development (CRD) grant and MITACS Accelerate grants in partnership with Bruce Power Inc. and the Nuclear Innovation Institute. ACKNOWLEDGEMENTS The authors would like to thank SNOLAB and its staff for support through underground space, logistical and technical services. Special thanks to Dr Ian Lawson and Lina Anselmo for their assistance in acquiring the gamma photon measurements and the interpretation of the results. Thank you to Dr Simon Lees for insights as principal investigator on the CRD grant as well as manuscript review. Thanks to Dr Xiao-Lei Zhao and Carley Crann of the A. E Lalonde AMS laboratory for their assistance in measuring the 14C content. Additional thanks to Remington Hill for his involvement in constructing the GPS, overall help with GEANT4 and the visualisation drivers and Forest Mahoney for his help in trouble shooting and refining the GPS as well as general structure of the geometries. Thank you to Dr Mike Oliver and Michel Lapointe who assisted with conducting the background measurements at NOSM and logistical support. SNOLAB operations are supported by the Canada Foundation for Innovation and the Province of Ontario Ministry of Research and Innovation, with underground access provided by Vale at the Creighton mine site. References 1. Thome , C. , Tharmalingam , S., Pirkkanen , J., Zarnke , A., Laframboise , T. and Boreham , D. R. The REPAIR Project: examining the biological impacts of sub-background radiation exposure within SNOLAB, a deep underground laboratory . Radiat. Res. 188 ( 4.2 ), 470 – 474 ( 2017 ). Google Scholar Crossref Search ADS PubMed WorldCat 2. Zagórski , Z. P. Possible role of radon in prebiotic chemistry and in early evolution of life on Earth . Nukleonika. 55 ( 4 ), 555 – 558 ( 2010 ). Google Scholar OpenURL Placeholder Text WorldCat 3. Carbone , M. C. et al. The Cosmic Silence experiment: on the putative adaptive role of environmental ionizing radiation . Radiat. Environ. Biophys. 48 ( 2 ), 189 – 196 ( 2009 ). Google Scholar Crossref Search ADS PubMed WorldCat 4. Fratini , E. , Carbone , C., Capece , D., Esposito , G., Simone , G., Tabocchini , M. A., Tomasi, M., Belli, M., and Satta, L. Low-radiation environment affects the development of protection mechanisms in V79 cells . Radiat. Environ. Biophys. 54 (2), 183–194 ( 2015 ). Google Scholar OpenURL Placeholder Text WorldCat 5. Carbone , M. C. , Pinto , M., Antonelli , F., & Balata , M. Effects of deprivation of background environmental radiation on cultured human cells . Nuovo Cimento. B , 125 ( 4 ), 469–477 ( 2010 ). Google Scholar OpenURL Placeholder Text WorldCat 6. Morciano , P. , Cipressa , F., Porrazzo , A., Esposito , G., Tabocchini , M. A. and Cenci , G. Fruit flies provide new insights in low-radiation background biology at the INFN underground Gran Sasso National Laboratory (LNGS) . Radiat. Res. 190 ( 3 ), 217 ( 2018 ). Google Scholar Crossref Search ADS PubMed WorldCat 7. Pirkkanen , J. et al. A novel specialized tissue culture incubator designed and engineered for radiobiology experiments in a sub-natural background radiation research environment . J. Environ. Radioact. 228 ( 2021 ), 106512 ( 2021 ). Google Scholar OpenURL Placeholder Text WorldCat 8. Bacioiu , I. Equivalent dose rate by muons to the human body . Radiation protection dosimetry , 147 ( 3 ), 380 – 385 ( 2011 ). 9. Wissmann , F. Variations observed in environmental radiation at ground level . Radiat. Prot. Dosimetry 118 ( 1 ), 3 – 10 ( 2006 ). Google Scholar Crossref Search ADS PubMed WorldCat 10. SNOLAB. SNOLAB Users Handbook . Retrieved from http://snolab2008.snolab.ca/snolab_users_handbook_rev02.pdf. 11. Ford , R. SNOLAB: review of the facility and experiments. In: AIP Conference Proceedings Sep 28 (Vol. 1441 , No. 1 , pp. 521 – 524 ). American Institute of Physics. 12. Shahbazi-Gahrouei , D. , Setayandeh , S. and Gholami , M. A review on natural background radiation . Adv. Biomed. Res. 2 ( 1 ), 65 ( 2013 ). Google Scholar Crossref Search ADS PubMed WorldCat 13. Furukawa , S. , Nagamatsu , A., Nenoi , M., Fujimori , A., Kakinuma , S., Katsube , T. et al. Space radiation biology for “living in space.” . Biomed. Res. Int. 2020 , 4703286 ( 2020 ). Google Scholar OpenURL Placeholder Text WorldCat 14. Agostinelli , S. et al. GEANT4—a simulation toolkit . Nuclear instruments and methods in physics research section A: Accelerators, Spectrometers, Detectors and Associated Equipment, 506 (3), pp. 250–303. OpenURL Placeholder Text WorldCat 15. Ribeiro , R. M. and Souza-Santos , D. Comparison of the neutron ambient dose equivalent and ambient absorbed dose calculations with different GEANT4 physics lists . Radiat. Phys. Chem. [Internet]. 139 ( April ), 179 – 183 ( 2017 ). http://dx.doi.org/10.1016/j.radphyschem.2017.05.021. Google Scholar OpenURL Placeholder Text WorldCat 16. Lampe , N. , Marin , P., Castor , J., Warot , G., Incerti , S., Maigne , L., Sarramia, D., Breton, V. Background study of absorbed dose in biological experiments at the Modane Underground Laboratory . In EPJ Web of Conferences 2016 (Vol. 124 , p. 00006). EDP Sciences. OpenURL Placeholder Text WorldCat 17. Ikeuchi , K. Lambertian reflectance. In: Encyclopedia of Computer Vision . ( Boston, MA : Springer ) pp. 441 – 443 ( 2014 ). Google Scholar Crossref Search ADS Google Preview WorldCat COPAC 18. Peracchi , S. , Vohradsky , J., Guatelli , S., Bolst , D., Tran , L. T., Prokopovich , D. A., Rosenfeld, A. B. Modelling of the silicon-on-insulator microdosimeter response within the International Space Station for astronauts’ radiation protection . Radiat. Meas. [Internet] 128 ( January ), 106182 ( 2019 ). https://doi.org/10.1016/j.radmeas.2019.106182. Google Scholar OpenURL Placeholder Text WorldCat 19. Lund , M. and Jevremovic , T. Enhanced GEANT4 Monte Carlo simulations of the space radiation effects on the International Space Station and Apollo missions using high-performance computing environment . Acta. Astronaut [Internet] 165 ( January ), 219 – 228 ( 2019 ). https://doi.org/10.1016/j.actaastro.2019.09.018. Google Scholar OpenURL Placeholder Text WorldCat 20. Lampe , N. , Biron , D. G., Brown , J. M. C., Incerti , S., Marin , P., Maigne , L., Sarramia, D., Seznec, H. and Breton, V. Simulating the impact of the natural radiation background on bacterial systems: implications for very low radiation biological experiments . PLoS One 11 ( 11 ), 1 – 19 ( 2016 ). Google Scholar Crossref Search ADS WorldCat 21. Cecchini , S. and Spurio , M. Atmospheric muons: experimental aspects . Geoscientific Instrumentation, Methods and Data Systems . 1 ( 2 ), 185 – 96 (2012). 22. Preskitt , C. A. Neutron physics. In: Proceedings of the symposium held at Rensselaer Polytechnic Institute, May 1961 . Yeater , M. L., Ed. ( New York : Academic Press ) ( 1962 ) xiii + 303 pp. Illus. $12. Science (80-). 1963 . Google Scholar Google Preview OpenURL Placeholder Text WorldCat COPAC 23. Cleveland , B. T. and Lawson , I. T. Measurement of the γ activity of a NdCl3 sample with the SNOLAB Ge detector . Energy (keV) 195 (200), 205 ( 2008 ). 24. Crann , C. A. , Murseli , S., St-Jean , G., Zhao , X., Clark , I. D. and Kieser , W. E. First status report on radiocarbon sample preparation techniques at the A.E. Lalonde AMS Laboratory (Ottawa, Canada) . Radiocarbon. 59 ( 3 ), 695 – 704 ( 2017 ). Google Scholar Crossref Search ADS WorldCat 25. Kieser , W. E. , Zhao , X. L., Clark , I. D., Cornett , R. J., Litherland , A. E., Klein , M., Mous, D. J. W., and Alary, J-F. The André E. Lalonde AMS Laboratory—The new accelerator mass spectrometry facility at the University of Ottawa . Nucl. Instruments Methods Phys. Res. Sect. B Beam Interact with Mater Atoms [Internet] 361 , 110 – 114 ( 2015 ). http://dx.doi.org/10.1016/j.nimb.2015.03.014. Google Scholar Crossref Search ADS WorldCat 26. International Commission on Radiological Protection . Nuclear decay data for dosimetric calculations. ICRP Publ 107 . Ann. ICRP 38 ( 3 ) ( 2008 ). OpenURL Placeholder Text WorldCat 27. Pirkkanen , J. , Zarnke , A. M., Laframboise , T., Lees , S. J., Tai , T. C., Boreham , D. R., and Thome, C. A research environment 2 km deep-underground impacts embryonic development in Lake Whitefish (Coregonus clupeaformis) . Front. Earth Sci. 8 ( July ), 1 – 9 ( 2020 ). Google Scholar OpenURL Placeholder Text WorldCat 28. Thome , C. , Mitz , C., Hulley , E. N., Somers , C. M., Manzon , R. G., Wilson , J. Y., and Boreham, D.R. Initial characterization of the growth stimulation and heat-shock-induced adaptive response in developing Lake Whitefish embryos after ionizing radiation exposure . Radiat. Res. 188 ( 4.2 ), 475 – 485 ( 2017 ). Google Scholar Crossref Search ADS PubMed WorldCat 29. Van Voorhies , W. A. , Castillo , H. A., Thawng , C. N. and Smith , G. B. The phenotypic and transcriptomic response of the caenorhabditis elegans nematode to background and below-background radiation levels . Front. Public Health 8 ( October ), 1 – 12 ( 2020 ). Google Scholar PubMed OpenURL Placeholder Text WorldCat 30. Kawanishi , M. , Okuyama , K., Shiraishi , K., Matsuda , Y., Taniguchi , R., Shiomi , N., Yonezawa M. and Yagi T. Growth retardation of paramecium and mouse cells by shielding them from background Radiation . J. Radiat. Res. 53 ( 3 ), 404 – 10 ( 2012 ). Google Scholar Crossref Search ADS PubMed WorldCat 31. Planel , H. , Soleilhavoup , J. P., Tixador , R., Richoilley , G., Conter , A., Croute , F., Caratero, C., and Gaubin, Y. Influence on cell proliferation of background radiation or exposure to very low, chronic γ radiation . Health Phys. 52 ( 5 ), 571 – 8 ( 1987 ). Google Scholar Crossref Search ADS PubMed WorldCat 32. Planel , H. , Soleilhavoup , J. P., Tixador , R., Croute , F. and Richoilley , G. Demonstration of a stimulating effect of natural ionizing radiation and of very low radiation doses on cell multiplication. In: Biological and Environmental Effects of Low Level Radiation, Vol. 1, pp. 127-133. International Atomic Agency, Vienna, 1976. See also Stimul. Newslet. 1, 29–34 (1970). 33. Conter , A. , Dupouy , D. and Planel , H. Demonstration of a biological effect of natural ionizing radiations . Int. J. Radiat. Biol. 43 ( 4 ), 421 – 432 ( 1983 ). Google Scholar OpenURL Placeholder Text WorldCat 34. Wadsworth , J. , Cockell , C. S., Murphy , A. S., Nilima , A., Paling , S., Meehan , E., Toth, C., Scovell, P. and Cascorbi, L. There’s plenty of room at the bottom: low radiation as a biological extreme . Front. Astron. Sp Sci. 7 ( August ), 1 – 14 ( 2020 ). Google Scholar OpenURL Placeholder Text WorldCat 35. Lampe , N. et al. Reducing the ionizing radiation background does not significantly affect the evolution of Escherichia coli populations over 500 generations . Sci. Rep. 9 ( 1 ), 1 – 6 ( 2019 ). Google Scholar Crossref Search ADS PubMed WorldCat 36. United Nations Scientific Committee on the Effects of Atomic Radiation . Sources and Effects of Ionizing Radiation, United Nations Scientific Committee on the Effects of Atomic Radiation UNSCEAR 2000 Report to the General Assembly, with Scientific Annexes . UNSCEAR 2000 Report ( 2000 ). 37. Jostes , R. F. et al. In vitro exposure of mammalian cells to radon: dosimetric considerations . Radiat. Res. 127 ( 2 ), 211 ( 1991 ). Google Scholar Crossref Search ADS PubMed WorldCat 38. Castillo , H. , Li , X., Schilkey , F. and Smith , G. B. Transcriptome analysis reveals a stress response of Shewanella oneidensis deprived of background levels of ionizing radiation . PLoS One 13 ( 5 ), 1 – 22 ( 2018 ). Google Scholar Crossref Search ADS WorldCat 39. Castillo , H. and Smith , G. B. Below-background ionizing radiation as an environmental cue for bacteria . Front. Microbiol. 8 ( February ), 1 – 7 ( 2017 ). Google Scholar PubMed OpenURL Placeholder Text WorldCat 40. Luckey , T. D. Ionizing radiation promotes protozoan reproduction . Radiat. Res. 108 ( 2 ), 215 – 221 ( 1986 ). Google Scholar Crossref Search ADS PubMed WorldCat © The Author(s) 2021. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/open_access/funder_policies/chorus/standard_publication_model)
TI - DOSIMETRIC CHARACTERISATION OF A SUB-NATURAL BACKGROUND RADIATION ENVIRONMENT FOR RADIOBIOLOGY INVESTIGATIONS
JF - Radiation Protection Dosimetry
DO - 10.1093/rpd/ncab120
DA - 2021-09-08
UR - https://www.deepdyve.com/lp/oxford-university-press/dosimetric-characterisation-of-a-sub-natural-background-radiation-Lv0caPHeP2
SP - 114
EP - 123
VL - 195
IS - 2
DP - DeepDyve
ER -