ARTICLE IN PRESS
Applied Radiation and Isotopes 60 (2004) 583–587
Radon emanation measurements using silicon photodiode detectors ! J.L. Gutie! rrez*, M. Garc!ıa-Talavera, V. Pen˜a, J.C. Nalda, M. Voytchev, R. Lopez Laboratorio LIBRA, Edificio I+D, Campus Miguel Delibes, Universidad de Valladolid, Valladolid 47011, Spain
Abstract Driven by the global concern about radon hazards, a wide variety of methods to measure radon and its decay products have been developed. Pin silicon photodiodes are increasingly applied in this field, their main advantages being high detection efficiency for alpha particles and low cost. In this paper, we present a system to determine the emanation factor for 222Rn from porous material based on a pin photodiode. This equipment is valid both for field and laboratory measurements, allowing to monitor the external emanation conditions by means of temperature, humidity and pressure sensors. To illustrate the capabilities of the system, we present two case studies of samples with high and low 226Ra content. The activity of this radionuclide in the samples had been previously determined by g-ray spectrometry. r 2003 Elsevier Ltd. All rights reserved. Keywords: Radon; Emanation factor; Silicon photodiodes; Accumulation chamber
1. Introduction Radiation dose due to 222Rn represents more than 50% of the annual dose equivalent received by the general population (UNSCEAR, 1993). Thus, great effort has been dedicated to develop instruments for the measurement of radon and its decay products, most of which are based on the detection of the alpha particles emitted by 222 Rn (5.5 MeV), 218Po (6.0 MeV) and 214Po (7.7 MeV). In particular pin silicon photodiodes, which were traditionally used in light generation and detection, are increasingly used in this field (Voytchev et al., 2001; Choi et al., 2001). Their main advantages regarding radon determinations are high detection efficiency for alpha particles and low cost. Under normal conditions, the main sources of radon gas in indoor air are the soil or bedrock beneath a dwelling and, to a less extent, building materials (Nazaroff and Nero, 1988). Therefore, the measurement of 222Rn emanation from different materials provides an essential information regarding prevention and mitigation strategies. To quantify the amount of radon available in a given medium for transport to the surface *Corresponding author. Fax: +34-98-34-23-013. E-mail address:
[email protected] (J.L. Guti!errez).
through the pores or cracks, the emanation factor is used. This parameter depends not only on properties of the material, as the radium distribution, but also on temperature, pressure and moisture content (Morawska and Philips, 1992). For this reason, it is important to carry out the emanation experiments under controlled conditions, which cannot often be achieved in the field. In this paper, we propose a method to measure the emanation of porous materials using a pin silicon photodiode. We have devised a system equipped with temperature, humidity and pressure sensors, suitable for both laboratory and in situ determinations. It presents a good sensitivity, with minimum measurable exhalation rates below those typical from natural materials. We describe the setup and calibration of the equipment for laboratory measurements. As an application, we determine the radon emanation factor for samples of fly ashes and uraniferous soils, whose 226Ra content has been previously determined by g-ray spectrometry.
2. Experimental We have built a radon measuring system based on a pin silicon photodiode (Hamamatsu S3584-09) with a
0969-8043/$ - see front matter r 2003 Elsevier Ltd. All rights reserved. doi:10.1016/j.apradiso.2003.11.080
ARTICLE IN PRESS J.L. Guti!errez et al. / Applied Radiation and Isotopes 60 (2004) 583–587 Air outlet
Air inlet
Photodiode's box
Aluminum chamber
Methacrylate column Sample
Radon emanation
Fig. 1. Schematic view of the set-up used for the radon emanation measurements in the laboratory.
surface area of 9 cm2. It is connected to a Hamamatsu H4083 preamplifier and this, in turn, to a Canberra ASA-100 multichannel analyser and a PC using GeniePC software (Canberra) for spectrum analysis. Alternatively, for field applications, the photodiode and preamplifier can be connected to a microcomputer with TMCA software. The detector is installed within a cylindrical aluminium chamber of 19 cm height and 22 cm diameter, which is connected to the ground to prevent attachment of radon daughters to the walls. An air inlet and an air outlet on top of the chamber (see Fig. 1) allow to establish a circulating system aided by a pump from Sarad Environmental Instruments. Such a system makes it possible to use the photodiode for continuous measurements of radon in air, by forcing ambient air to circulate through the chamber. Besides, in emanation experiments, the circulating system is used to purge the radon out by nitrogen gas before counting. To perform emanation measurements, the bottom of the chamber has to be removed. For in situ determinations, the open chamber is directly placed over the soil. For measurements in the laboratory, it is coupled to a methacrylate column of 49 cm height. Because of the light sensitivity of photodiodes, the column has to be covered with aluminum paper or any other opaque material during the measurement. To monitor the emanation conditions, the chamber is equipped with temperature (HD 9216 Delta Ohmn), humidity (HD 9216 Delta Ohmn) and pressure (Digima junior-Special Instruments) sensors.
3. Study of the photodiode response
will be counted. The critical detection angle (i.e., the minimal angle between the diode plane and the direction of the incident alpha particle at which the particle could be detected) is 10 for this type of photodiodes (Klein et al., 1993). Although silicon pin detectors are also sensitive to beta particles and photons, the contribution of these type of radiations due to radon and its daughters has been proved to be non-detectable (Voytchev, 2000). In Fig. 2 we show a spectrum acquired with our system for a high radon concentration in the chamber. The continuum distribution on the left part of the spectrum corresponds to 222Rn and to progeny distributed in the volume around the photodiode. The two peaks at 6.0 and 7.7 MeV correspond to the 218Po and 214 Po atoms that have been deposited on the photodiode surface. Besides, a peak caused by electronic noise appears in the lower channels. A linear dependence of the diode response as a function of the activity concentration has been observed in each of these three regions, namely the 222Rn continuum and the 218Po and 214Po peaks. This linearity allows the use of this system for radon measurements. Two different procedures were followed to determine the counting rates in each region: *
*
Integral areas were computed in three predefined spectral windows. The two polonium peaks were fitted to Gaussian functions convoluted with exponential tails and the continuum was reproduced using a polynomial function in order to compute the net areas in every region.
The two above-mentioned procedures preserve linearity in the response under laboratory conditions. Nevertheless, variations in the detection efficiency with the temperature and the humidity have been
1024 218 Po
768 214Po
Counts
584
512
256
222Rn+daughters
0 0
The silicon photodiode detects alpha particles emitted by 222Rn, 218Po and 214Po atoms within the chamber. Due to the low range of alpha particles in air, only those emitted within a close volume around the photodiode
5927
11854
Energy (keV) Fig. 2. Spectrum acquired with our system for a radon concentration in the chamber of 23 kBq/m3. Acquisition time was set to 2 min.
ARTICLE IN PRESS J.L. Guti!errez et al. / Applied Radiation and Isotopes 60 (2004) 583–587
reported for these measuring devices. Variation in counting rates can be neglected within the usual range of relative humidity (37–92%). But changes with temperature are above 2% per C for any of the three spectral regions (Voytchev et al., 2001). Therefore, the dependence of the detection efficiency with temperature must be calibrated to be able to use the system for field measurements. In the laboratory, the temperature, humidity and pressure conditions are kept stable during the measurement. We have calibrated the detection efficiency of our system at 22 C temperature, 60% relative humidity and 0.936 atm pressure using a Pylon RNC Calibrated Radon Gas Source. The inner gas was ventilated to purge the radon in the chamber. The RNC was then connected to the chamber and a known volume was pumped to it using an alpha-pump (Genitron Instruments GmbH). A radon activity, A0, of 228157662 Bq/m3 was transferred to the chamber. Once the RNC was disconnected, we started to take measurements after waiting for 12 min in order to attain secular equilibrium for 222Rn and 218Po. The evolution of the counting rate in the 218Po peak can be described by the following function: N ¼ N0 eðlþlf Þt ;
ð1Þ
N(s1) being the counting rate in the 218Po peak, l (s1), the 222Rn decay constant and t(s) elapsed time since the RNC was disconnected. The counting rate at t=0 (N0) and the parameter lf ; which indicates losses of radon from the chamber by leakage, can be determined by means of a least-squares fit. Since the activity of 222Rn in the chamber in the initial time (A0) is known, we can calculate the activity at any time t, using the value fitted for lf ; namely (5.7970.16)106 s1. Thus, the detection efficiency for the 218Po peak can be determined by a linear fit of the measured counting rates and the calculated 222Rn activities. Specifically, the detection efficiency of the system using the 218Po peak is (9.7570.34) 106 cps/Bq m3.
4. Theoretical approach Radon emanation measurements are usually performed by enclosing or covering the sample to be measured. We can distinguish two processes within the enclosure. First, the generation of radon in the porous material due to 226Ra decay, part of which is released to the pore system. Second, the transport of radon to the surface through the pores or cracks by diffusion (advection processes are not considered). Given the geometry of our accumulation system and assuming that the walls are non-absorbent, we can apply a one-dimensional theory to describe the
585
radon diffusion: @cðx; tÞ D @2 cðx; tÞ ¼ lcðx; tÞ þ f; @t p @x2
ð2Þ
c being the concentration of radon within the pores (Bq m3), D the diffusion coefficient (m2 s1), p the porosity of the material, l the radon decay constant and f (Bq m3 s1) the production rate of radon in the pore air, which can be expressed as f¼
f lCRa r ; p
ð3Þ
where f is the emanation factor, r the bulk density (kg m3) and CRa the radium concentration (Bq kg1) of the material. On the other hand, radon flux at the surface is given by Fick’s law. The concentration of radon inside the chamber is given by the following equation: @cðx; tÞ dca ðtÞ S ðl þ lf Þca ðtÞ; ¼ D ð4Þ dt Va Vs @x s
where ca is the radon concentration in the accumulation chamber, S the exhalation area, Va the accumulation volume, and Vs the volume of the sample. This equation satisfies the boundary condition cðx; tÞjs ¼ ca ðtÞ . Nevertheless, one should take into account that the fact of enclosing the sample might change the emanation conditions. In general, when enclosing a sample exhaling into free space, the free exhalation (E0) will change with time towards the bound exhalation (Eb). The ratio between free and bound exhalation rate is given by (Cosma et al., 2001) E0 pS Ll d ¼1þ tanh ; ð5Þ Eb ðVa Vs Þ ðl þ lf Þ L pffiffiffiffiffiffiffiffiffiffiffi where L ¼ D=pl and d is the thickness of the sample. Assuming E0 EEb ; the growth of radon concentration inside the chamber can be explained by the following function: E0 S ½1 eðlþlf Þt : ð6Þ ca ðtÞ ¼ ðVa Vs Þðl þ lf Þ Parameters E0 and lf can be determined by means of a non-linear squares fit. We have performed numerical simulations in order to ensure that the condition E0 EEb will be fulfilled for any of the samples to be studied, specifically soils, mine tailings and fly ashes. The aim of such simulations is to determine the maximum thickness that can be allowed for the sample in the methacrylate column: increasing the volume of the sample to measure improves detection sensitivity, but also increases the second term in Eq. (5). We have assumed that porosity of the samples, p, ranges from 20% to 40% and that diffusion coefficient, D, vary from 6 107 to 7 106 m2 s1. We generated pair of values (p,D) by randomly sampling in the intervals given above, following a uniform probability distribution.
ARTICLE IN PRESS J.L. Guti!errez et al. / Applied Radiation and Isotopes 60 (2004) 583–587
586
These pairs of values were then used to evaluate Eq. (5), where lf was set to the value determined in the efficiency calibration. According to these results, we selected to filled the 49 cm column up to a height of 16 cm.
0.6 0.5 N(s–1)
0.4 0.3 0.2
5. Applications
0.1
The radon emanation factor, f, can be obtained using the following expresion: E0 : lCRa rd
ð7Þ
We measured the 226Ra activity for the sample under study (CRa) by g-ray spectrometry, using the 351.9-keV photopeak from 214Pb. A petri dish is filled with the sample, sealed and, after storing for 1 month to attain secular equilibrium, counted in an HPGe detector. The radon tighness of the sealed petri dishes was verified for various uranium-rich samples. We compared the activity values derived from the 214Pb 351.9-keV photopeak and the 226Ra 186.0-keV photopeak, after substracting the 235U contribution. This was obtained from its emissions at 163.3 and 205.1 keV.1 No systematic deviations were observed between the activities calculated in both ways showing that the dishes do not actually show radon leakage. Next, we present the application of our system to determine the radon emanation factor for samples of uraniferous soils and fly ashes. In Fig. 3, we show the evolution of the radon concentration in the chamber after enclosing the soil sample. Measurement intervals were set to 120 min.The experimental data points were fitted to Eq. (5) using the Levenberg–Marquardt algorithm in order to determine parameters E0 and lf ; as well as their associated uncertainties. The value of lf agrees with the one determined in the efficiency calibration within the statistical uncertainties. On the other hand, the 226Ra activity (2120730 Bq/kg) of the sample was determined using the 351.9-keV photon from 214Pb. The emanation factor was calculated to be 0.2570.03, according to Eq. (7). Analogously, we determined the emanation factor for a sample of fly ashes. Due to their lower 226Ra activity (A=8070.4 Bq/kg), measurement intervals had to be set to 12 h. Fig. 4 shows one of the spectra corresponding to this sample. The emanation factor in this case is 0.3570.05.
0.4x106 Time (seconds)
0.7x106
Fig. 3. Evolution of the counting rate (N) for the 218Po peak during an exhalation measurement for a uraniferous soil sample. Experimental measurements (points) were fitted by a weighted least-squares fit to the theoretical equation (solid line). Error bars are smaller than the symbol size.
32
222Rn
+daughters 218Po
24
Counts
f ¼
0.1x106
214Po
16
8
0 6027
12054
Energy (keV) Fig. 4. Spectrum adquired during the exhalation measurement of a fly ashes sample. Acquisition time was set to 12 h.
can be applied to laboratory and field determinations, and, particularly, to measure accurately radon emanation rates from porous materials. The possibility to monitor the external conditions, as temperature, humidity and pressure, will allow us in the future to investigate the influence of radium speciation in the emanation factor.
6. Conclusions
Acknowledgements
We have developed a low-cost continuous radon measurement system using a silicon pin photodiode. It
We are thankful to Dr. L.S. Quindos and the staff of his laboratory for assistance in the experiment calibration and to M. Bordonada (ENUSA S.A.) for providing the uraniferous samples. This work has been supported by Junta de Castilla y Leon and Iberdrola.
1 235 U photopeaks require applying coincidence–summing corrections factors (see Garcia-Talavera et al., 2001).
ARTICLE IN PRESS J.L. Guti!errez et al. / Applied Radiation and Isotopes 60 (2004) 583–587
References Choi, E., Komori, M., Takahisa, K., Kudomi, N., Kume, K., Hayashi, K., Yoshida, S., Ohsumi, H., Ejiri, H., Kishimoto, T., Matsuoka, K., Tasaka, S., 2001. Highly sensitive radon monitor and radon emanation rates for detector components. Nucl. Instrum. Methods A 459, 177–181. Cosma, C., Dancea, F., Jurcut, T., Ristou, D., 2001. Determination of 222Rn emanation fraction and diffusion coefficient in concrete using accumulation chambers and the influence of humidity and radium distribution. Appl. Radiat. Isot. 54, 467–473. Garcia-Talavera, M., Laedermann, P., D!ecombaz, M., Daza, M.J., Quintana, B., 2001. Coincidence summing corrections for the natural decay series in g-ray spectrometry. Appl. Radiat. Isot. 54, 769–776. Klein, D., Devillard, C., Chambaudet, A., Barillon, R., 1993. Development of measuring techniques using a silicon diode
587
pin detector for continuous radon monitoring. Nucl. Tracks Radiat. Meas. 22 (1–4), 369–372. Morawska, L., Phillips, C.R., 1992. Dependence of the radon coefficient on radium distribution and internal structure of the material. Geochim. Cosmochim. Acta 57, 1783–1797. Nazaroff, W.W., Nero, A.V., 1988. Radon and Its Decay Products In Indoor Air. Wiley, New York. Voytchev, M., 2000. Etude et developement d’une metrologie des particules alpha utilisant la technologie des photodiodes—application a la mesure du radon et de ses descendants. Ph.D. Tesis, Universite de Franche. Voytchev, M., Klein, D., Chambaudet, A., Georgiev, G., 2001. The use of silicon photodiodes for radon and progeny measurements. Health Phys. 80 (6), 590–596. UNSCEAR, 1993. Sources and effects of ionizing radiation, UNSCEAR 1993 report.
