- Email: [email protected]
Uranium distribution and radon exhalation from Brazilian dimension stones
Applied Radiation and Isotopes 70 (2012) 808–817
Contents lists available at SciVerse ScienceDirect
Applied Radiation and Isotopes journal homepage: www.elsevier.com/locate/apradiso
Uranium distribution and radon exhalation from Brazilian dimension stones P.G.Q. Amaral, T.M.B. Galembeck, D.M. Bonotto n, A.C. Artur Departamento de Petrologia e Metalogenia, Instituto de Geociˆencias e Ciˆencias Exatas, Universidade Estadual Paulista (UNESP), Av. 24-A No. 1515, C.P. 178, CEP 13506-900, Rio Claro, Sa~ o Paulo, Brazil
a r t i c l e i n f o
a b s t r a c t
Article history: Received 24 August 2011 Received in revised form 23 October 2011 Accepted 24 October 2011 Available online 9 November 2011
This paper provides evaluations of the radiometric behavior and exhalation patterns of radon gas in decorative and dimension stones explored in the Brazilian states of Minas Gerais and Espı´rito Santo, given the importance of determining radon gas concentrations in human-inhabited environments. A total of 10 silicate rock types were studied, featuring different petrographic/petrophysical characteristics given by seven magmatic rocks (three of which are granitic pegmatites) and three metamorphic rocks. The study, comprising radiometric data of U and monitoring of 222Rn gas exhalation, shows a strong correlation between petrographic parameters and the physical properties of rocks. U levels ranged between 2.9 and 37 ppm, revealing a good coherence between the presence and the absence of radioactive element-bearing accessory minerals for each rock type. The rate of radon exhalation from the stones is related to the petrographic/petrophysical features of each material. By comparing the 222 Rn level generated by a rock to the amount effectively emanated by it, the rate of emanated gas proves to be insignificant; also, a rock that produces more Rn will not always emanate more. Simulations performed to estimate the radon levels inside residences or any given indoor environment showed that nine samples attained values below the 4 pCi/L EPA limit, whereas one was above that limit. & 2011 Elsevier Ltd. All rights reserved.
Keywords: Dimension stones Radon exhalation Indoor environment
1. Introduction The distribution of uranium in the Earth’s crust is related to magmatic activities during the Earth formation processes. The mean abundance of uranium in the continental crust is 1.3 ppm, with higher concentrations in the upper continental crust (2.7 ppm) and inferior in the lower crust (0.2 ppm) (Ulbrich and Gomes, 1981). Uranium occurrence is more pronounced in granitic rocks, granitic pegmatites and syenites, appearing in a wide variety of minerals and found in most rock-forming minerals, secondary and accessory minerals, or forming their own minerals. In this case, it is concentrated in few less-abundant species, such as uraninite (UO2). The distribution of uranium, present in trace amounts in igneous rock-forming minerals such as quartz and feldspar, may be linked to processes of isomorphous replacement, concentration and adsorption in imperfect minerals, inclusion in the crystal lattice and in microcrystals, as suggested by Pertlik et al. (1974). Biotite retains 19% to 22% of total uranium, and heavy minerals such as zircon, monazite, apatite, magnetite, ilmenite and riebeckite contain 61% to 65% of the uranium present in a given rock type (Moreira-Nordemann, 1977).
n
Corresponding author. Tel.: þ55 19 35269244; fax: þ55 19 35249644. E-mail address: [email protected] (D.M. Bonotto).
0969-8043/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.apradiso.2011.10.010
Radon gas (222Rn), a product of uranium (238U) decay series, represents the second leading cause of lung cancer in the United States after cigarette smoking, resulting in approximately 20 thousand deaths a year, according to studies by the US Environmental Protection Agency (EPA, 2003). The exhalation rate of radon gas from a given stone is related to U levels, reflecting the origin of the rock as well as the influence of textural and structural features of the material. Whenever 226Ra decays, a radon atom and alpha particle are formed and simultaneously ejected in opposite directions. This mechanism can result in expulsion of the radon atom from the crystal or molecular lattice in which the radon atom was formed. The petrophysical characteristics of rocks (microfissure condition, grain size, arrangement, alteration degree and contact surfaces between constituents) have direct influence on permeability, a parameter that can contribute to radon gas exhalation by facilitating the internal transport and flow to the outside environment of rocks. In 2008, US news outlets reported that increased radon levels found inside homes could be linked to countertops made from dimension stones (ABIROCHAS, 2008; Murphy, 2008). The same article mentioned that similar claims had been made periodically over the last decade by manufacturers and distributors of materials competing with dimension stones (Murphy, 2008). Because Brazil has taken an important role in the decorative and dimension stone business due to the vast lithological variety
P.G.Q. Amaral et al. / Applied Radiation and Isotopes 70 (2012) 808–817
found in its territory, with the United States as its main client (ABIROCHAS, 2011), there is a clear need to study this theme, focusing not only on the commercial aspects, but also on human health. This work features assessments of 222Rn emanates from silicatebased dimension stones sourced from different geological environments in the states of Minas Gerais and Espı´rito Santo, southeastern Brazil.
2. Sampling and physical characterization The analyzed samples include magmatic rocks of different lithologic types used for decorative purposes that are commercially mined in Brazil and distributed overseas (diorite, syenite, charnockite, monzogranite and three granitic pegmatites), as well as three metamorphic rocks. The locations of all rocks can be found in the geological maps in Figs. 1 and 2, while Table 1 shows the petrographic/petrophysical characteristics of the analyzed samples.
809
A total of 10 test specimens were built for each sample, in a roughly cubic shape, each side measuring about 5 cm and weighing between 300 and 400 g. The specimens were rinsed in running water and oven-dried for 24 h at 110 1C. The specimens were then removed from the oven, cooled at room temperature and weighed individually to obtain the dry weight (mass C). Next, the test specimens were immersed and saturated in a desiccator for 24 h, using a vacuum pump for the first three hours. After 24 h, the specimens were removed from the water, superficially dried and air-weighed, thereby obtaining their saturated weight (mass B). They were then weighed again on a hydrostatic scale, suspended in a recipient containing water, to obtain the submerged weight (mass A). The values obtained in this trial result from the following equations: dry apparent specific mass (kg/m3)¼C/(BA); saturated apparent specific mass (kg/m3)¼ B/(B A); apparent porosity (%)¼100 (B C)/(BA); apparent water absorption (%)¼100 (B C)/C. Lastly, the arithmetic mean of the physical indices was calculated according to NBR norm 12766 (ABNT, 1992) for each lithologic type. Table 2 contains the obtained results.
Fig. 1. Simplified geological map with the location of the rock specimens belonging to Minas Gerais State (CPRM, 2003). 1—Rock I (GC); 2—Rock F (MS); 3—Rock E (TI).
810
P.G.Q. Amaral et al. / Applied Radiation and Isotopes 70 (2012) 808–817
Fig. 2. Simplified geological map with the location of the rock specimens belonging to Espı´rito Santo State (CPRM, 2003). 4—Rock H (IL); 5—Rock C (VL); 6—Rock A (PSG); 7—Rock J (CB); 8—Rock B (OI); 9—Rock D (CC); 10—Rock G (GI).
3. Radiometric measurements 3.1. Gamma readings Each sample was submitted to gamma spectrometry. Preparation consisted of: selection, breaking rocks with the aid of a hammer, quartering, trituration in a jaw crusher, crushing in an oscillating mill in agate pot until reaching 200 meshes, and placement in cylindrical aluminum capsules containing approximately 70 g of powder. A 30-day period was then observed in order to reach secular radioactive equilibrium in the uranium series. To quantify uranium, a gamma-ray spectrometer was used, formed basically by a gamma sensor and electronic circuitry splitting the radiation striking the crystal into two or more energy components, where an interaction occurs between the gamma radiation emitted by a radioactive source and a NaI(Tl) scintillator crystal, producing a low-amplitude pulse in the anode of a photomultiplier, in which each pulse is pre-amplified and sent to an amplifier (Bonotto, 1996). The pre-amplifier is connected to the amplifier, which sends the signals to a card in the computer,
converting them into data using software Maestro II, by EG&G ORTEC. In order to calibrate the system to obtain uranium, the standards employed were acquired from the New Brunswick Laboratory of the US Department of Energy, located in Argonne, IL, USA, for pitchblende (codes NBL-101-A, NBL-102-A, NBL-103A, NBL-104-A, NBL-105-A) and monazite sand (codes NBL-106-A and NBL-107-A). The count rate was then determined in the 214Bi photopeak, obtaining the effective intensity of uranium (Iu) (count rate by mass) and later the uranium concentration (Cu) in ppm. The calibration curve of CU as a function of IU was drawn, which is linear and expressed by the equation log CU ¼ 1.057 (log IU)þ 2.578 (Duarte, 2002). 3.2. Radon readings The AlphaGUARD device, manufactured by Genitron, was used for radon quantification. It is an alpha spectrometer consisting of an optimized ion pulse chamber, in which the gas is diffused into a cylindrical chamber containing a fiberglass filter, retaining all products of radon decay.
P.G.Q. Amaral et al. / Applied Radiation and Isotopes 70 (2012) 808–817
811
Table 1 Petrographic/petrophysical characterization of the samples. Mineralogy (%)
Rock A (PSG)
Rock B (OI)
Rock C (VL)
Rock D (CC)
Rock E (TI)
Quartz Potassium feldspar Plagioclase Biotite Hornblende Clinopyroxene Orthopyroxene Opaques Apatite Zircon Titanite Allanite Rutile Garnet Tourmaline Topaz Muscovite Sericite Chlorite Epidote Bastite Goethite Iron hydroxides Carbonates Heavy minerals (%) Grain size (mm) Range Prevalence/matrix Color Structure Texture Microcracks/mm2 Micro cracks (%) Filled intra-grain Unfilled intra-grain Filled inter-grain Unfilled inter-grain Contact (%) Concave/convex Straight Serrated Mineral alteration Potassium feldspar Plagioclase Biotite Hypersthene Classification
5.0 4.0 58.0 9.0 4.0 4.0 12.0 3.5 0.5 Trace Trace – – – – – – Trace Trace Trace Trace – – Trace 33.0
4.0 58.0 22.0 4.5 7.0 – – 2.0 Trace Trace 2.0 Trace Trace – – – – Trace Trace Trace – – – 0.5 16.0
25.0 40.0 19.0 6.0 0.5 1.0 3.0 0.5 0.4 0.1 0.5 – Trace 4.0 – – – Trace Trace Trace Trace – – Trace 1.,0
31.6 31.0 25.0 7.5 – – – 0.4 1.0 0.4 1.0 0.1 Trace – – – – 2.0 Trace Trace – – – Trace 12.4
31.0 40.0 15.5 0.5 – – – 0.5 Trace – – – – – Trace 0.5 7.5 1.5 – Trace – 1.5 1.5 – 13.5
0.5–10 2–6 Dark gray Isotropic Equigranular 1.73
0.5–25 2–6 Brown Isotropic Porphyritic 2.90
0.3–30 3–7 Dark green Isotropic Porphyritic 6.39
0.3–30 2.5–5 Clear gray Isotropic Inequigranular 2.21
2–80 10–30 Green Isotropic Pegmatite 4.20
47.2 23.0 18.7 11.1
24.4 62.2 11.0 2.4
41.2 48.8 5.4 4.6
33.6 62.4 2.7 1.3
58.2 36.3 3.8 1.7
49.8 47.8 2.4
62.8 31.6 5.6
58.7 36.9 4.4
61.8 32.4 5.8
40.5 44.2 15.3
Incipient Incipient Incipient Incipient Hypersthene quartz diorite
Incipient Incipient Incipient – Hornblende quartz syenite
Incipient Incipient Weak Incipient Garnet charnockite
Incipient Moderate Incipient – Monzogranite
Weak Weak Muscovitized – Hydrothermalized granitic pegmatite
Mineralogy (%)
Rock F (MS)
Rock G (GI)
Rock H (IL)
Rock I (GC)
Rock J (CB)
Quartz Potassium feldspar Plagioclase Biotite Opaques Apatite Zircon Titanite Garnet Allanite Sillimanite Topaz Muscovite Spinel Sericite Epidote Goethite Iron hydroxides Carbonates Heavy minerals (%) Grain size (mm) Range Prevalence/matrix Color
29.0 42.0 14.5 Trace 0.5 Trace Trace – – – – 0.3 8.7 – 2.0 Trace 1.5 1.5 – 14.5
25.0 41.0 24.0 5.0 Trace Trace – – 1.0 – 0.5 – 3.5 – Trace Trace – Trace – 10.0
35.0 39.0 17.0 3.5 Trace Trace – – 4.0 – 1.5 – – Trace Trace Trace – – Trace 9.0
28.0 39.0 21.0 8.5 Trace Trace Trace – – – 2.5 – Trace – Trace Trace – 1.0 – 12.0
26.0 28.0 34.0 5.5 Trace Trace Trace Trace – Trace Trace – 3.0 – 2.0 1.5 – Trace – 12.0
2–50 10–20 Yellow
2–200 20–80 Clear yellow
0.2–25 2–5 Yellow
0.5–10 2–4 Yellow brown
0.3–25 3–8 Reddish pink
812
P.G.Q. Amaral et al. / Applied Radiation and Isotopes 70 (2012) 808–817
Table 1 (continued ) Mineralogy (%)
Rock A (PSG)
Rock B (OI)
Rock C (VL)
Rock D (CC)
Rock E (TI)
Structure Texture
Isotropic Pegmatite
Isotropic Pegmatite
Migmatitic Inequigranular
Gneiss Inequigranular
Microcraks/mm2 Micro cracks (%) Filled intra-grain Unfilled intra-grain Filled inter-grain Unfilled inter-grain Contact (%) Concave/convex Straight Serrated Mineral alteration Potassium feldspar Plagioclase
3.4
3.2
Gneiss Granoblastic/ Inequigranular 2.8
4.9
3.8
52.0 46.6 0.0 1.4
35.4 58.2 1.4 5.0
19.6 60.1 17.3 3.0
24.6 68.8 3.3 3.3
57.0 40.2 1.9 0.9
53.6 34.8 11.6
64.0 25.1 10.9
66.5 29.5 4.0
59.1 34.4 6.5
53.6 34.8 11.6
Incipient Incipient to moderate Incipient Granitic pegmatite
Incipient Incipient
Incipient Incipient
Incipient Incipient
Weak to Intense Granitic pegmatite
Weak Quartz-feldspathic granulite with sillimanite
Weak Biotite gneiss migmatized
Incipient Moderate to Intense Incipient Migmatite gneiss
Biotite Classification
Table 2 Results of the analysis of some physical parameters.
Table 3 General features of the samples studied in this investigation.
Sample (code)
Dry bulk density (kg/m3)
Saturated bulk density (kg/m3)
Apparent porosity (%)
Water absorption (%)
Rock Rock Rock Rock Rock Rock Rock Rock Rock Rock
2997 2719 2657 2683 2617 2645 2589 2653 2592 2613
3003 2728 2665 2691 2626 2657 2603 2662 2606 2621
0.56 0.92 0.81 0.78 0.89 1.27 1.45 0.91 1.41 0.79
0.19 0.34 0.30 0.29 0.34 0.48 0.56 0.34 0.54 0.30
A (PSG) B (OI) C (VL) D (CC) E (TI) F (MS) G (GI) H (IL) I (GC) J (CB)
The signals or pulses are quantified and digitized when exiting the chamber. The electrode center reads ionization rates as pulses, which are processed by a very sensitive pre-amplifier unit, in which three independent channels in a digital processing network make it possible to efficiently differentiate between real alpha events and various interference events, measure radon concentrations, and evaluate their high concentrations (Balca´zar et al., 2002). The device was configured to take measurements in real time, every minute. The samples were plate cut (Table 3), cleaned in an alcoholsoaked cloth and placed in an oven at 70 1C for approximately 24 h, to remove the moisture caused by cleaning. After removal from the oven, they were air-cooled and placed inside glass storage bottles. The samples were kept in Pyrex glass bottles (19 L), sealed with cork attached to a system of valves, allowing air to enter and escape. Air was sucked in a vacuum up to 450 mmg/Hg (safety limit, set not to surpass the resistance limit of the bottles and below the manufacturer’s limit of 650 mmg/Hg, to avoid implosion), not allowing air to enter or escape. Accumulation periods of 5, 10, 15, 20 and 25 day were preset for the samples. The exhaled radon amount was measured at the end of each accumulation period for all stored rocks by
Sample
Code
Surface treatment
Number of samples
Size (cm3)
Rock Rock Rock Rock Rock Rock Rock Rock Rock Rock
PSG OI VL CC TI MS GI IL GC CB
Polished Polished Polished Polished Polished, resined and screened Polished, resined and screened Polished, resined and screened Polished Polished Gross
5 4 4 4 4 4 4 4 4 4
30 4.5 1.5 30 4.5 2 30 4.5 2 30 4.5 2 30 4.5 2 30 4.5 2 30 4.5 2 30 4.5 2 30 4.5 2 30 4.5 2
A B C D E F G H I J
connecting the AlphaGUARD device to the valve system (Fig. 3). It is important to enforce that this experimental protocol did not involve the samples crushing to 200 meshes like in the gamma readings (Section 3.1), i.e. the procedure for evaluating the Rn emanation was applied to intact dimension stone slabs. Thus, the rocks’ pore system and respective attributes (porosity and permeability) have not been modified due to crushing, that avoided changing the emanation efficiencies.
4. Results and discussion 4.1. Physical indices All studied stones showed dry bulk density above the mean value of 2550 kg/m3 for Brazilian rocks (Fraza~ o and Farjallat, 1995). Values regarding the specific mass of the rocks reflect their mineralogy, in which rocks with higher percentages of heavy minerals tend to feature higher specific mass. With regard to porosity and water absorption, most of the analyzed rocks showed results lower than average values for Brazilian rocks presented by Fraza~ o and Farjallat (1995); that is, lower than 1.0% for porosity and 0.4% for water absorption. This group includes PSG hypersthene diorite, OI syenite, VL charnockite, CC monzogranite, TI granitic pegmatite, IL granulite and CB gneissified
P.G.Q. Amaral et al. / Applied Radiation and Isotopes 70 (2012) 808–817
Table 4 Uranium content and accumulation time. Code
PSG OI VL CC TI MS GI IL GC CB
Fig. 3. System for measuring the Rn gas exhaled from the studied rocks: (1) alpha spectrometer, (2) air pump, (3) Pyrex bottle, (4) valves system and (5) direction of the air flow.
migmatite. The remaining rocks (MS/GI granitic pegmatites and GC migmatized gneiss) had levels above the mentioned values, with rock 7 (GI) as the material with the highest indices (Table 2). The values for these indices are related to petrographic/petrophysical characteristics inherent to each rock type, such as textural homogeneity/heterogeneity, inter-mineral contact relations, microfissure behavior and presence of specific minerals of certain rocks. The intergranular vs. intragranular porosity must control both liberation of Rn and permeability to the surface of the finished product, also occurring in case of selective Rn sorption on different minerals. 4.2. Uranium and radon readings The rocks showed values between 2.9 and 37.0 ppm of uranium (Table 4). The lowest values (up to 10 ppm) were obtained in OI syenite (2.9 ppm), followed by MS granitic pegmatite (5.2 ppm), GI granitic pegmatite (6.4 ppm), VL charnockite and IL granulite (7.7 ppm). This wide range of U levels basically reflects the different genetic nature, degree of evolution and mineralogy composition of the studied rocks. It should be mentioned that U levels obtained in this work showed adequate coherence with the variation intervals for this element as presented by Salas et al. (2003) for non-alkaline, ornamental quartz-feldspathic granitic and gneissic rocks from Minas Gerais State. The Rn exhaled by each rock sample was repeatedly measured after finishing the accumulation period, i.e. 5, 10, 15, 20 and 25 day. Some rocks yielded extremely low values, inclusive in the longest accumulation period (25 day) like PSG hypersthene diorite and VL charnockite. Two materials showed relatively high values (above 1000 Bq/m3) in repeated measurements, i.e. GI and TI granitic pegmatites (Table 4). The radon emanation rate should be constant under the radioactive equilibrium condition between Ra and Rn. However, the Rn exhaled in the rock did not reach this situation due to the factors pointed out in Section 4.1. Consequently, emanation was not constant as the Rn evolution curves in the monitored samples did not reach equilibrium after the 25-days accumulation period. The curves were obtained using a second order polynomial trendline from point zero scaled according to the emanation rate
813
222
Rn exhaled from the rock samples according to the
U
222
Rn (Bq/m3)
(ppm)
5 day
10 day
15 day
20 day
25 day
11.05 2.91 7.67 37.03 22.53 5.20 6.41 7.70 18.47 20.44
1.33 10.33 4.33 33.33 678.00 251.67 856.67 16.33 66.33 330.00
8.33 8.67 1.00 51.33 822.33 404.67 981.33 18.33 90.67 600.67
14.33 20.00 10.33 62.67 1029.00 376.00 1550.00 14.67 104.33 651.00
3.67 8.00 1.00 74.67 1089.00 472.67 1953.33 22.67 118.00 642.00
12.00 21.67 10.00 69.33 1135.67 471.67 2483.33 23.00 98.33 752.00
in order to understand their evolution in relation to the Rn emanation (Fig. 4). Because Rn gas can get trapped in the rock, the U level alone is not able to indicate which rock will emanate the highest Rn gas amount. This may be evidenced by comparing rocks with homogenous and heterogeneous textural aspects. Rock D (CC monzogranite) exhibits the highest U level and one of the lowest Rn emanation rate, while GI granitic pegmatite features the highest Rn emanation rate and a very low U level (Fig. 5). In addition to the amount of Rn generated, exhalation depends on the crystal structure of minerals and of the rock fabric. The fracture density of the rock matrix tends to facilitate the escape of Rn gas, as well the porous media matrix flow that is linked to the degree to which pores within the material are interconnected. Given that 222Rn from 238U has a longer half-life compared to other Rn isotopes, it has a greater probability of escaping the rock’s structure and reaching the outside environment. Certain types of microfissures may contain filling material with radioactive elements. The closer they are to the surface, the easier it becomes for radon gas to escape from the rock. Therefore, the greater the number of inter-grain microfissures, combined with trans-grain fissures, which can be over 5 to 6 cm long in the case of TI, MS and GI pegmatites, the greater will be the Rn exhalation rate (Fig. 6). The presence of minerals containing U in their chemical makeup, such zircon and allanite, leads to partial destruction of the crystalline structures of adjacent minerals, usually accompanied by micropore irradiation, which can theoretically facilitate the Rn migration. Fig. 7 shows that the porosity due to microgaps and the U amount are factors affecting the Rn exhalation in the rocks studied. 4.3. Radon emanation coefficient Each sample showed a different behavior for radon exhalation, expressed by emanation efficiency as defined by Wanty et al. (1992) through the following equation: E ¼ ð222 RnÞ air 222
ð
RnÞ air þ ð222 RnÞ rock
Wanty et al. (1992) considered that E is almost always lower than 0.5 (usually lower than 0.3). The U content (ppm) makes it possible to estimate the 226Ra activity (Bq/kg) by employing conversion factor 12.34. This value corresponds to 222Rn generated in the rock, under radioactive equilibrium conditions. The radon exhaled in the air, [Rn] (in Bq/m3), is converted to Bq/kg using the equation [Rn] V/m, in which V¼19 L and m is the sample weight (kg).
814
P.G.Q. Amaral et al. / Applied Radiation and Isotopes 70 (2012) 808–817
Fig. 4. Radon exhaled according to the accumulation period. Samples described in Table 1.
Table 5 contains the values obtained for E, showing that although some rocks exhibited low U levels, they exhaled more Rn gas than others possessing high levels of this radioelement.
This is likely related to the Rn arrangement in the crystalline minerals structure in the rock, with microgaps behaving as a favorable aspect for exhalation.
P.G.Q. Amaral et al. / Applied Radiation and Isotopes 70 (2012) 808–817
Fig. 5. Relationship between the U content and Rn exhaled after the accumulation period of 25 day.
815
The GI granitic pegmatite showed the highest exhalation efficiency (E¼0.169), among the studied rocks. It also featured the highest value of apparent porosity among all analyzed rocks (Table 2). The CC monzogranite contains the highest U concentration among the rocks analyzed, producing the most Rn. Nevertheless, its E value is comparable to those of PSG hypersthene diorite, VL charnockite and IL quartz-feldspar granulite that are rocks featuring much lower U content. Thus, the apparent porosity, low microfissuring degree/permeability and strong mineral engagement shown by CC monzogranite are important parameters for justifying its E value. Compared to other rocks, GC migmatized gneiss biotite and CB gneissified migmatite showed similar U levels, producing nearsimilar 222Rn values (in Bq/kg of rock). Rock J (CB) achieved higher emanation efficiency than rock I (GC), possibly due to its
Fig. 6. Relationships among the microcracks types and Rn exhaled after the accumulation period of 25 day.
Fig. 7. Relationships among the U content, apparent porosity and Rn exhaled after the accumulation period of 25 day.
816
P.G.Q. Amaral et al. / Applied Radiation and Isotopes 70 (2012) 808–817
Table 5 Radon emanation efficiency from radiometric data. Code
PSG OI VL CC TI MS GI IL GC CB
Table 6 Simulation of the Rn exhaled to a hypothetical indoor circulation environment.
U
226 Ra and 222Rn calculated from U content
222
Rn exhaled (after 25 day)
Rn emanation efficiency
(ppm)
226 Ra (Bq/kg)
222 Rn (Bq/kg)
222
Rn (Bq/m3)
222 Rn (Bq/kg)
E
136.36 35.91 94.65 456.95 278.02 64.17 79.10 95.02 227.92 252.23
136.36 35.91 94.65 456.95 278.02 64.17 79.10 95.02 227.92 252.23
12.00 21.67 10.00 69.33 1135.67 471.67 2483.33 23.00 98.33 752.00
0.07 0.13 0.07 0.44 8.01 3.14 16.10 0.14 0.69 4.66
0.001 0.004 0.001 0.001 0.028 0.047 0.169 0.001 0.003 0.018
11.05 2.91 7.67 37.03 22.53 5.20 6.41 7.70 18.47 20.44
Code
PSG OI VL CC TI MS GI IL GC CB
Simulated rock volume utilized (m3)
1.82 1.82 1.82 1.82 1.82 1.82 1.82 1.82 1.82 1.82
Simulated environment area (m2)
90.86 90.86 90.86 90.86 90.86 90.86 90.86 90.86 90.86 90.86
Exhaled Rn measured after 25 day (Bq/m3)
12.0 21.7 10.0 69.3 1135.7 471.7 2483.3 23.0 98.3 752.0
Simulated Rn exhaled to the indoor environmenta (Bq/m3)
(pCi/L)
1.43 2.54 1.27 8.25 145.40 57.67 293.06 2.62 12.47 83.65
0.04 0.07 0.03 0.22 3.93 1.56 7.92 0.07 0.34 2.26
a 148 Bq/m3 and 4 pCi/L are the threshold Rn concentration values recommended for homes (EPA, 2003).
petrographic/petrophysical characteristic. Even though it also features lower apparent porosity, it has a larger number of filled microfissures; as such, the location of U in the stone becomes one important factor for the efficiency. When comparing CB gneissified migmatite to TI hydrothermal granitic pegmatite, the E factor is higher for the pegmatite, which, in addition to featuring higher 222Rn level (in Bq/kg) also has greater apparent porosity, thus, contributing to more pronounced Rn exhalation from these rocks. Both MS and TI granitic pegmatites have similar petrographic characteristics determined by optical microscopy but different U content. The latter experienced significant U-enrichment as an effect of hydrothermal action, leading to different 222Rn levels, even though MS pegmatite has a significantly higher apparent porosity than TI granitic pegmatite. 4.4. Use as decorative stones in residences Ornamental stones have several uses in buildings, including flooring, baseboards, doorsills, doorjambs and wall siding, as well as in table tops, countertops and washbasins. They are often used as a single material or in different stone combinations to achieve a distinct effect. Based on the quantification of radon obtained from the samples and knowing the contribution to the exhalation rate from stones used for decorative purposes in buildings, the Chyi (2008) method was adapted to simulate how the use of the rocks would impact the radon levels inside residences or any given indoor environment. To that end, a 90.86 m2 area and 2.80 m high circulation environment was considered, where stone was applied as flooring. Factors such as indoor-outdoor air exchange were disregarded because one of the greatest risks of radon is its accumulation in a given environment, usually indoors. The formula Rnce ¼[Rne] F2/F1 was adopted, in which [Rne] is the amount of radon exhaled by the rock (in Bq/m3) that is provided by the Rn readings. Factor 1 (air dilution factor), F1, represents the recurrent air volume from within the circulation environment (Vce) in relation to the air volume inside the bottle (Vb), i.e. F1 ¼Vce/Vb. Factor 2 (material amplification factor), F2, is the ratio between the area of the material applied in the circulation environment (Ace) and the area of the sample used to monitor Rn (As), i.e. F2¼ Ace/As. The results for Rnce are given in Bq/m3 and later converted to pCi/L, as expressed in Table 6. In the case of the rock that surpassed the 4 pCi/L limit suggested by EPA (2003), the
evaluation was made only considering the flooring of bathing areas and a 3 m 0.60 m 0.02 m countertop in the kitchen, thereby allowing a comparison of the air in the house as a whole and in those environments. Six stones under consideration in the simulation showed values much below the 4 pCi/L limit suggested by EPA (2003): PSG hypersthene diorite, OI syenite, VL charnockite, CC monzogranite, GC migmatized gneiss and IL granulite. Two stones achieved intermediate values, one belonging to the pegmatite group (MS, 1.56 pCi/L) and another belonging to the metamorphic rocks group (CB migmatite, 2.26 pCi/L). These stones can influence the total radon levels when added to other existing materials in a given environment. The GI and TI granitic pegmatites showed high values; the former (7.92 pCi/L) stood above the 4 pCi/L EPA (2003) limit and the latter (3.93 pCi/L) was very close to the limit, requiring caution when used in enclosed or poorly ventilated environments. When GI granitic pegmatite is applied only as flooring in bathroom areas of the hypothetical environment, a reduction is observed in the added radon level for the air of the entire residence, standing at 0.77 pCi/L. Nevertheless, the 7.92 pCi/L concentration remains in the bathroom areas. When a countertop is added in the kitchen, radon levels reach 1.46 pCi/L, with overall residence added levels increasing to 0.89 pCi/L. It is concluded that, although the residence now has radon levels below that suggested by EPA (2003), bathroom areas remain above the 4 pCi/L limit and the kitchen area begins to exert a certain influence in case more radon-releasing materials are added. This simulation shows certain precautions with regard to the specification of rocky materials to be used in buildings, especially in spacious and enclosed environments, with radon gas exhalation rates representing an important factor to be considered. The increase in radon levels in the case of dimension stones is related to the amount of material used in a given environment, i.e. the greater the volume of stone used, the higher the added radon levels will be.
5. Conclusion Radiometric data on uranium and the results of the monitoring of 222Rn gas exhalation closely match the genetic nature of petrographic types and feature a good correlation with their respective petrographic parameters and physical indices. By quantifying Rn exhalation in the studied rocks, it was concluded that it is not only the quantity of U in a stone that
P.G.Q. Amaral et al. / Applied Radiation and Isotopes 70 (2012) 808–817
should be analyzed as a parameter of radon gas exhalation, but also petrographic/petrophysical aspects such as their microporous network. The comparison between the quantity of 222Rn produced by a given stone and the amount emanated from the rock, shown as factor E (emanation efficiency), demonstrates that not all the radon gas produced in the stone is actually emanated, and that the rock that produces the most Rn will not always emanate more. Thus, it should be emphasized that radon exhalation, as well as being related to uranium amounts, is also linked to the ease with which radon gas is able to escape from the crevices of a rock. It is therefore concluded that radon gas exhalation in the studied dimension stones results from the petrographic/petrophysical aspects of those rocks. Among the rocks under consideration, six showed values much below the 4 pCi/L EPA (2003) limit when applied as flooring: PSG hypersthene diorite, OI syenite, VL charnockite, CC monzogranite, IL granulite and GC migmatized gneiss. Two stones achieved intermediate values, between 1.5 and 2.5 pCi/L: MS pegmatite and CB migmatite. The TI and GI granitic pegmatites showed high values: the former was very close to the 4 pCi/L limit, whereas the latter stood above the limit, both requiring caution when used in closed or poorly ventilated environments. The increase in radon levels in a given environment depends on the quantity of stone applied, its exhalation rate, as well as the volume of air ambient considering ventilation (indoor/outdoor air exchange), so that if one of the conditioning factors is increased, the other is reduced/increased.
Acknowledgments To Brazil’s CNPq (National Council for Scientific and Technological Development) for the financial support. One anonymous reviewer is thanked by helpful comments that improved the readability of the manuscript.
817
References ABIROCHAS (Associac- a~ o Brasileira da Indu´stria de Rochas Ornamentais), 2008. Radioatividade natural em materiais rochosos, Informe 18/2008, /http:// www.abirrochas.com.brS. ABIROCHAS (Associac- a~ o Brasileira da Indu´stria de Rochas Ornamentais), 2011. ~ Balanc- o das exportac- oes brasileiras de rochas ornamentais para os EUA em 2010, Informe 06/2011, /http://www.abirrochas.com.brS. ABNT (Associac- a~ o Brasileira de Normas Te´cnicas), 1992. Rochas para revestimento - determinac- a~ o da massa especı´fica aparente, porosidade aparente e absorc- a~ o d’a´gua aparente. NBR 12766. ABNT, Sa~ o Paulo. Balca´zar, M., Lo´pez, A., Streil, T., Cha´vez, A., 2002. Radon leakage assessment in a controlled radon laboratory. Geofı´sica Int. 41 (3), 281–284. Bonotto, D.M., 1996. Comportamento hidrogeoquı´mico do 222Rn e iso´topos de 238U ~ controladas em laborato´rio e em sistemas naturais. Post e 234U sob condic- oes Ph.D. Thesis, UNESP-Universidade Estadual Paulista, Rio Claro, SP, 223 pp. Chyi, L.L., 2008. Radon Testing of Various Countertop Materials. Final Report. The University of Akron: Akron, 15 pp. CPRM (Servic-o Geolo´gico do Brasil), 2003. Geologia, Tectoˆ nica e Recursos Minerais do Brasil. CPRM, Brası´lia, DF. Duarte, C.R., 2002. Radioelementos naturais na a´rea do projeto Rio Preto (GO). Ph.D. Thesis, UNESP-Universidade Estadual Paulista, Rio Claro, SP, 169 pp. EPA (US Environmental Protection Agency), 2003. Assessment of risks from radon in homes. EPA, Washington, DC 99 pp.. Fraza~ o, E.B., Farjallat, J.E.S., 1995. Caracterı´sticas tecnolo´gicas das principais rochas silica´ticas brasileiras usadas como pedras de revestimento. In: Atas do 11 Congresso Internacional da Pedra Natura. Lisboa, Portugal, pp. 47–58. Moreira-Nordemann, L.M., 1977. Etude de la vitesse d’alteration des roches au moyer de l’uranium utilize comme traceur natural. Application a deux bassins du nordest du Bresil. Tese (Doctorat d’ E´tat), Universite´ Pierre et Marie Curie, Paris, 162 pp. Murphy, K., 2008. What’s lurking in your countertop?. The New York Times /http:// www.nytimes.com/2008/07/24/garden/24granite. html?pagewanted=1&_r=1&sq=radon&st=cse&scp=1S. Pertlik, F., Roger, J.J.N., Adams, J.A.S., 1974. Uranium. In: Wedepohl, K.H. (Ed.), Handbook of Geochemistry, Springer-Verlag, Berlin, p. 92.0192.08. Salas, H.T., Nalini Jr., H.A., Mendes, J.C., 2003. Radioatividade em rochas ornamentais do Brasil. In: Vidal, F.W.H., Lima, M.A.A.B. (Eds.), Anais do IV Simpo´sio de Rochas Ornamentais do Nordeste. Fortaleza, CE, pp. 68–73. Ulbrich, H.H.G.J., Gomes, C.B., 1981. Alkaline rocks from continental Brazil. Earth Sci. Rev. 17, 135–154. Wanty, R.B., Lawrence, E.P., Gundersen, L.C.S., 1992. A theoretical model for the flux of radon from rock to groundwater. In: Gattes, A.E., Gundersen, L.C.S. (Eds.), Geologic Controls of Radon, Geological Society of America, Boulder, CO, USA.
Contents lists available at SciVerse ScienceDirect
Applied Radiation and Isotopes journal homepage: www.elsevier.com/locate/apradiso
Uranium distribution and radon exhalation from Brazilian dimension stones P.G.Q. Amaral, T.M.B. Galembeck, D.M. Bonotto n, A.C. Artur Departamento de Petrologia e Metalogenia, Instituto de Geociˆencias e Ciˆencias Exatas, Universidade Estadual Paulista (UNESP), Av. 24-A No. 1515, C.P. 178, CEP 13506-900, Rio Claro, Sa~ o Paulo, Brazil
a r t i c l e i n f o
a b s t r a c t
Article history: Received 24 August 2011 Received in revised form 23 October 2011 Accepted 24 October 2011 Available online 9 November 2011
This paper provides evaluations of the radiometric behavior and exhalation patterns of radon gas in decorative and dimension stones explored in the Brazilian states of Minas Gerais and Espı´rito Santo, given the importance of determining radon gas concentrations in human-inhabited environments. A total of 10 silicate rock types were studied, featuring different petrographic/petrophysical characteristics given by seven magmatic rocks (three of which are granitic pegmatites) and three metamorphic rocks. The study, comprising radiometric data of U and monitoring of 222Rn gas exhalation, shows a strong correlation between petrographic parameters and the physical properties of rocks. U levels ranged between 2.9 and 37 ppm, revealing a good coherence between the presence and the absence of radioactive element-bearing accessory minerals for each rock type. The rate of radon exhalation from the stones is related to the petrographic/petrophysical features of each material. By comparing the 222 Rn level generated by a rock to the amount effectively emanated by it, the rate of emanated gas proves to be insignificant; also, a rock that produces more Rn will not always emanate more. Simulations performed to estimate the radon levels inside residences or any given indoor environment showed that nine samples attained values below the 4 pCi/L EPA limit, whereas one was above that limit. & 2011 Elsevier Ltd. All rights reserved.
Keywords: Dimension stones Radon exhalation Indoor environment
1. Introduction The distribution of uranium in the Earth’s crust is related to magmatic activities during the Earth formation processes. The mean abundance of uranium in the continental crust is 1.3 ppm, with higher concentrations in the upper continental crust (2.7 ppm) and inferior in the lower crust (0.2 ppm) (Ulbrich and Gomes, 1981). Uranium occurrence is more pronounced in granitic rocks, granitic pegmatites and syenites, appearing in a wide variety of minerals and found in most rock-forming minerals, secondary and accessory minerals, or forming their own minerals. In this case, it is concentrated in few less-abundant species, such as uraninite (UO2). The distribution of uranium, present in trace amounts in igneous rock-forming minerals such as quartz and feldspar, may be linked to processes of isomorphous replacement, concentration and adsorption in imperfect minerals, inclusion in the crystal lattice and in microcrystals, as suggested by Pertlik et al. (1974). Biotite retains 19% to 22% of total uranium, and heavy minerals such as zircon, monazite, apatite, magnetite, ilmenite and riebeckite contain 61% to 65% of the uranium present in a given rock type (Moreira-Nordemann, 1977).
n
Corresponding author. Tel.: þ55 19 35269244; fax: þ55 19 35249644. E-mail address: [email protected] (D.M. Bonotto).
0969-8043/$ - see front matter & 2011 Elsevier Ltd. All rights reserved. doi:10.1016/j.apradiso.2011.10.010
Radon gas (222Rn), a product of uranium (238U) decay series, represents the second leading cause of lung cancer in the United States after cigarette smoking, resulting in approximately 20 thousand deaths a year, according to studies by the US Environmental Protection Agency (EPA, 2003). The exhalation rate of radon gas from a given stone is related to U levels, reflecting the origin of the rock as well as the influence of textural and structural features of the material. Whenever 226Ra decays, a radon atom and alpha particle are formed and simultaneously ejected in opposite directions. This mechanism can result in expulsion of the radon atom from the crystal or molecular lattice in which the radon atom was formed. The petrophysical characteristics of rocks (microfissure condition, grain size, arrangement, alteration degree and contact surfaces between constituents) have direct influence on permeability, a parameter that can contribute to radon gas exhalation by facilitating the internal transport and flow to the outside environment of rocks. In 2008, US news outlets reported that increased radon levels found inside homes could be linked to countertops made from dimension stones (ABIROCHAS, 2008; Murphy, 2008). The same article mentioned that similar claims had been made periodically over the last decade by manufacturers and distributors of materials competing with dimension stones (Murphy, 2008). Because Brazil has taken an important role in the decorative and dimension stone business due to the vast lithological variety
P.G.Q. Amaral et al. / Applied Radiation and Isotopes 70 (2012) 808–817
found in its territory, with the United States as its main client (ABIROCHAS, 2011), there is a clear need to study this theme, focusing not only on the commercial aspects, but also on human health. This work features assessments of 222Rn emanates from silicatebased dimension stones sourced from different geological environments in the states of Minas Gerais and Espı´rito Santo, southeastern Brazil.
2. Sampling and physical characterization The analyzed samples include magmatic rocks of different lithologic types used for decorative purposes that are commercially mined in Brazil and distributed overseas (diorite, syenite, charnockite, monzogranite and three granitic pegmatites), as well as three metamorphic rocks. The locations of all rocks can be found in the geological maps in Figs. 1 and 2, while Table 1 shows the petrographic/petrophysical characteristics of the analyzed samples.
809
A total of 10 test specimens were built for each sample, in a roughly cubic shape, each side measuring about 5 cm and weighing between 300 and 400 g. The specimens were rinsed in running water and oven-dried for 24 h at 110 1C. The specimens were then removed from the oven, cooled at room temperature and weighed individually to obtain the dry weight (mass C). Next, the test specimens were immersed and saturated in a desiccator for 24 h, using a vacuum pump for the first three hours. After 24 h, the specimens were removed from the water, superficially dried and air-weighed, thereby obtaining their saturated weight (mass B). They were then weighed again on a hydrostatic scale, suspended in a recipient containing water, to obtain the submerged weight (mass A). The values obtained in this trial result from the following equations: dry apparent specific mass (kg/m3)¼C/(BA); saturated apparent specific mass (kg/m3)¼ B/(B A); apparent porosity (%)¼100 (B C)/(BA); apparent water absorption (%)¼100 (B C)/C. Lastly, the arithmetic mean of the physical indices was calculated according to NBR norm 12766 (ABNT, 1992) for each lithologic type. Table 2 contains the obtained results.
Fig. 1. Simplified geological map with the location of the rock specimens belonging to Minas Gerais State (CPRM, 2003). 1—Rock I (GC); 2—Rock F (MS); 3—Rock E (TI).
810
P.G.Q. Amaral et al. / Applied Radiation and Isotopes 70 (2012) 808–817
Fig. 2. Simplified geological map with the location of the rock specimens belonging to Espı´rito Santo State (CPRM, 2003). 4—Rock H (IL); 5—Rock C (VL); 6—Rock A (PSG); 7—Rock J (CB); 8—Rock B (OI); 9—Rock D (CC); 10—Rock G (GI).
3. Radiometric measurements 3.1. Gamma readings Each sample was submitted to gamma spectrometry. Preparation consisted of: selection, breaking rocks with the aid of a hammer, quartering, trituration in a jaw crusher, crushing in an oscillating mill in agate pot until reaching 200 meshes, and placement in cylindrical aluminum capsules containing approximately 70 g of powder. A 30-day period was then observed in order to reach secular radioactive equilibrium in the uranium series. To quantify uranium, a gamma-ray spectrometer was used, formed basically by a gamma sensor and electronic circuitry splitting the radiation striking the crystal into two or more energy components, where an interaction occurs between the gamma radiation emitted by a radioactive source and a NaI(Tl) scintillator crystal, producing a low-amplitude pulse in the anode of a photomultiplier, in which each pulse is pre-amplified and sent to an amplifier (Bonotto, 1996). The pre-amplifier is connected to the amplifier, which sends the signals to a card in the computer,
converting them into data using software Maestro II, by EG&G ORTEC. In order to calibrate the system to obtain uranium, the standards employed were acquired from the New Brunswick Laboratory of the US Department of Energy, located in Argonne, IL, USA, for pitchblende (codes NBL-101-A, NBL-102-A, NBL-103A, NBL-104-A, NBL-105-A) and monazite sand (codes NBL-106-A and NBL-107-A). The count rate was then determined in the 214Bi photopeak, obtaining the effective intensity of uranium (Iu) (count rate by mass) and later the uranium concentration (Cu) in ppm. The calibration curve of CU as a function of IU was drawn, which is linear and expressed by the equation log CU ¼ 1.057 (log IU)þ 2.578 (Duarte, 2002). 3.2. Radon readings The AlphaGUARD device, manufactured by Genitron, was used for radon quantification. It is an alpha spectrometer consisting of an optimized ion pulse chamber, in which the gas is diffused into a cylindrical chamber containing a fiberglass filter, retaining all products of radon decay.
P.G.Q. Amaral et al. / Applied Radiation and Isotopes 70 (2012) 808–817
811
Table 1 Petrographic/petrophysical characterization of the samples. Mineralogy (%)
Rock A (PSG)
Rock B (OI)
Rock C (VL)
Rock D (CC)
Rock E (TI)
Quartz Potassium feldspar Plagioclase Biotite Hornblende Clinopyroxene Orthopyroxene Opaques Apatite Zircon Titanite Allanite Rutile Garnet Tourmaline Topaz Muscovite Sericite Chlorite Epidote Bastite Goethite Iron hydroxides Carbonates Heavy minerals (%) Grain size (mm) Range Prevalence/matrix Color Structure Texture Microcracks/mm2 Micro cracks (%) Filled intra-grain Unfilled intra-grain Filled inter-grain Unfilled inter-grain Contact (%) Concave/convex Straight Serrated Mineral alteration Potassium feldspar Plagioclase Biotite Hypersthene Classification
5.0 4.0 58.0 9.0 4.0 4.0 12.0 3.5 0.5 Trace Trace – – – – – – Trace Trace Trace Trace – – Trace 33.0
4.0 58.0 22.0 4.5 7.0 – – 2.0 Trace Trace 2.0 Trace Trace – – – – Trace Trace Trace – – – 0.5 16.0
25.0 40.0 19.0 6.0 0.5 1.0 3.0 0.5 0.4 0.1 0.5 – Trace 4.0 – – – Trace Trace Trace Trace – – Trace 1.,0
31.6 31.0 25.0 7.5 – – – 0.4 1.0 0.4 1.0 0.1 Trace – – – – 2.0 Trace Trace – – – Trace 12.4
31.0 40.0 15.5 0.5 – – – 0.5 Trace – – – – – Trace 0.5 7.5 1.5 – Trace – 1.5 1.5 – 13.5
0.5–10 2–6 Dark gray Isotropic Equigranular 1.73
0.5–25 2–6 Brown Isotropic Porphyritic 2.90
0.3–30 3–7 Dark green Isotropic Porphyritic 6.39
0.3–30 2.5–5 Clear gray Isotropic Inequigranular 2.21
2–80 10–30 Green Isotropic Pegmatite 4.20
47.2 23.0 18.7 11.1
24.4 62.2 11.0 2.4
41.2 48.8 5.4 4.6
33.6 62.4 2.7 1.3
58.2 36.3 3.8 1.7
49.8 47.8 2.4
62.8 31.6 5.6
58.7 36.9 4.4
61.8 32.4 5.8
40.5 44.2 15.3
Incipient Incipient Incipient Incipient Hypersthene quartz diorite
Incipient Incipient Incipient – Hornblende quartz syenite
Incipient Incipient Weak Incipient Garnet charnockite
Incipient Moderate Incipient – Monzogranite
Weak Weak Muscovitized – Hydrothermalized granitic pegmatite
Mineralogy (%)
Rock F (MS)
Rock G (GI)
Rock H (IL)
Rock I (GC)
Rock J (CB)
Quartz Potassium feldspar Plagioclase Biotite Opaques Apatite Zircon Titanite Garnet Allanite Sillimanite Topaz Muscovite Spinel Sericite Epidote Goethite Iron hydroxides Carbonates Heavy minerals (%) Grain size (mm) Range Prevalence/matrix Color
29.0 42.0 14.5 Trace 0.5 Trace Trace – – – – 0.3 8.7 – 2.0 Trace 1.5 1.5 – 14.5
25.0 41.0 24.0 5.0 Trace Trace – – 1.0 – 0.5 – 3.5 – Trace Trace – Trace – 10.0
35.0 39.0 17.0 3.5 Trace Trace – – 4.0 – 1.5 – – Trace Trace Trace – – Trace 9.0
28.0 39.0 21.0 8.5 Trace Trace Trace – – – 2.5 – Trace – Trace Trace – 1.0 – 12.0
26.0 28.0 34.0 5.5 Trace Trace Trace Trace – Trace Trace – 3.0 – 2.0 1.5 – Trace – 12.0
2–50 10–20 Yellow
2–200 20–80 Clear yellow
0.2–25 2–5 Yellow
0.5–10 2–4 Yellow brown
0.3–25 3–8 Reddish pink
812
P.G.Q. Amaral et al. / Applied Radiation and Isotopes 70 (2012) 808–817
Table 1 (continued ) Mineralogy (%)
Rock A (PSG)
Rock B (OI)
Rock C (VL)
Rock D (CC)
Rock E (TI)
Structure Texture
Isotropic Pegmatite
Isotropic Pegmatite
Migmatitic Inequigranular
Gneiss Inequigranular
Microcraks/mm2 Micro cracks (%) Filled intra-grain Unfilled intra-grain Filled inter-grain Unfilled inter-grain Contact (%) Concave/convex Straight Serrated Mineral alteration Potassium feldspar Plagioclase
3.4
3.2
Gneiss Granoblastic/ Inequigranular 2.8
4.9
3.8
52.0 46.6 0.0 1.4
35.4 58.2 1.4 5.0
19.6 60.1 17.3 3.0
24.6 68.8 3.3 3.3
57.0 40.2 1.9 0.9
53.6 34.8 11.6
64.0 25.1 10.9
66.5 29.5 4.0
59.1 34.4 6.5
53.6 34.8 11.6
Incipient Incipient to moderate Incipient Granitic pegmatite
Incipient Incipient
Incipient Incipient
Incipient Incipient
Weak to Intense Granitic pegmatite
Weak Quartz-feldspathic granulite with sillimanite
Weak Biotite gneiss migmatized
Incipient Moderate to Intense Incipient Migmatite gneiss
Biotite Classification
Table 2 Results of the analysis of some physical parameters.
Table 3 General features of the samples studied in this investigation.
Sample (code)
Dry bulk density (kg/m3)
Saturated bulk density (kg/m3)
Apparent porosity (%)
Water absorption (%)
Rock Rock Rock Rock Rock Rock Rock Rock Rock Rock
2997 2719 2657 2683 2617 2645 2589 2653 2592 2613
3003 2728 2665 2691 2626 2657 2603 2662 2606 2621
0.56 0.92 0.81 0.78 0.89 1.27 1.45 0.91 1.41 0.79
0.19 0.34 0.30 0.29 0.34 0.48 0.56 0.34 0.54 0.30
A (PSG) B (OI) C (VL) D (CC) E (TI) F (MS) G (GI) H (IL) I (GC) J (CB)
The signals or pulses are quantified and digitized when exiting the chamber. The electrode center reads ionization rates as pulses, which are processed by a very sensitive pre-amplifier unit, in which three independent channels in a digital processing network make it possible to efficiently differentiate between real alpha events and various interference events, measure radon concentrations, and evaluate their high concentrations (Balca´zar et al., 2002). The device was configured to take measurements in real time, every minute. The samples were plate cut (Table 3), cleaned in an alcoholsoaked cloth and placed in an oven at 70 1C for approximately 24 h, to remove the moisture caused by cleaning. After removal from the oven, they were air-cooled and placed inside glass storage bottles. The samples were kept in Pyrex glass bottles (19 L), sealed with cork attached to a system of valves, allowing air to enter and escape. Air was sucked in a vacuum up to 450 mmg/Hg (safety limit, set not to surpass the resistance limit of the bottles and below the manufacturer’s limit of 650 mmg/Hg, to avoid implosion), not allowing air to enter or escape. Accumulation periods of 5, 10, 15, 20 and 25 day were preset for the samples. The exhaled radon amount was measured at the end of each accumulation period for all stored rocks by
Sample
Code
Surface treatment
Number of samples
Size (cm3)
Rock Rock Rock Rock Rock Rock Rock Rock Rock Rock
PSG OI VL CC TI MS GI IL GC CB
Polished Polished Polished Polished Polished, resined and screened Polished, resined and screened Polished, resined and screened Polished Polished Gross
5 4 4 4 4 4 4 4 4 4
30 4.5 1.5 30 4.5 2 30 4.5 2 30 4.5 2 30 4.5 2 30 4.5 2 30 4.5 2 30 4.5 2 30 4.5 2 30 4.5 2
A B C D E F G H I J
connecting the AlphaGUARD device to the valve system (Fig. 3). It is important to enforce that this experimental protocol did not involve the samples crushing to 200 meshes like in the gamma readings (Section 3.1), i.e. the procedure for evaluating the Rn emanation was applied to intact dimension stone slabs. Thus, the rocks’ pore system and respective attributes (porosity and permeability) have not been modified due to crushing, that avoided changing the emanation efficiencies.
4. Results and discussion 4.1. Physical indices All studied stones showed dry bulk density above the mean value of 2550 kg/m3 for Brazilian rocks (Fraza~ o and Farjallat, 1995). Values regarding the specific mass of the rocks reflect their mineralogy, in which rocks with higher percentages of heavy minerals tend to feature higher specific mass. With regard to porosity and water absorption, most of the analyzed rocks showed results lower than average values for Brazilian rocks presented by Fraza~ o and Farjallat (1995); that is, lower than 1.0% for porosity and 0.4% for water absorption. This group includes PSG hypersthene diorite, OI syenite, VL charnockite, CC monzogranite, TI granitic pegmatite, IL granulite and CB gneissified
P.G.Q. Amaral et al. / Applied Radiation and Isotopes 70 (2012) 808–817
Table 4 Uranium content and accumulation time. Code
PSG OI VL CC TI MS GI IL GC CB
Fig. 3. System for measuring the Rn gas exhaled from the studied rocks: (1) alpha spectrometer, (2) air pump, (3) Pyrex bottle, (4) valves system and (5) direction of the air flow.
migmatite. The remaining rocks (MS/GI granitic pegmatites and GC migmatized gneiss) had levels above the mentioned values, with rock 7 (GI) as the material with the highest indices (Table 2). The values for these indices are related to petrographic/petrophysical characteristics inherent to each rock type, such as textural homogeneity/heterogeneity, inter-mineral contact relations, microfissure behavior and presence of specific minerals of certain rocks. The intergranular vs. intragranular porosity must control both liberation of Rn and permeability to the surface of the finished product, also occurring in case of selective Rn sorption on different minerals. 4.2. Uranium and radon readings The rocks showed values between 2.9 and 37.0 ppm of uranium (Table 4). The lowest values (up to 10 ppm) were obtained in OI syenite (2.9 ppm), followed by MS granitic pegmatite (5.2 ppm), GI granitic pegmatite (6.4 ppm), VL charnockite and IL granulite (7.7 ppm). This wide range of U levels basically reflects the different genetic nature, degree of evolution and mineralogy composition of the studied rocks. It should be mentioned that U levels obtained in this work showed adequate coherence with the variation intervals for this element as presented by Salas et al. (2003) for non-alkaline, ornamental quartz-feldspathic granitic and gneissic rocks from Minas Gerais State. The Rn exhaled by each rock sample was repeatedly measured after finishing the accumulation period, i.e. 5, 10, 15, 20 and 25 day. Some rocks yielded extremely low values, inclusive in the longest accumulation period (25 day) like PSG hypersthene diorite and VL charnockite. Two materials showed relatively high values (above 1000 Bq/m3) in repeated measurements, i.e. GI and TI granitic pegmatites (Table 4). The radon emanation rate should be constant under the radioactive equilibrium condition between Ra and Rn. However, the Rn exhaled in the rock did not reach this situation due to the factors pointed out in Section 4.1. Consequently, emanation was not constant as the Rn evolution curves in the monitored samples did not reach equilibrium after the 25-days accumulation period. The curves were obtained using a second order polynomial trendline from point zero scaled according to the emanation rate
813
222
Rn exhaled from the rock samples according to the
U
222
Rn (Bq/m3)
(ppm)
5 day
10 day
15 day
20 day
25 day
11.05 2.91 7.67 37.03 22.53 5.20 6.41 7.70 18.47 20.44
1.33 10.33 4.33 33.33 678.00 251.67 856.67 16.33 66.33 330.00
8.33 8.67 1.00 51.33 822.33 404.67 981.33 18.33 90.67 600.67
14.33 20.00 10.33 62.67 1029.00 376.00 1550.00 14.67 104.33 651.00
3.67 8.00 1.00 74.67 1089.00 472.67 1953.33 22.67 118.00 642.00
12.00 21.67 10.00 69.33 1135.67 471.67 2483.33 23.00 98.33 752.00
in order to understand their evolution in relation to the Rn emanation (Fig. 4). Because Rn gas can get trapped in the rock, the U level alone is not able to indicate which rock will emanate the highest Rn gas amount. This may be evidenced by comparing rocks with homogenous and heterogeneous textural aspects. Rock D (CC monzogranite) exhibits the highest U level and one of the lowest Rn emanation rate, while GI granitic pegmatite features the highest Rn emanation rate and a very low U level (Fig. 5). In addition to the amount of Rn generated, exhalation depends on the crystal structure of minerals and of the rock fabric. The fracture density of the rock matrix tends to facilitate the escape of Rn gas, as well the porous media matrix flow that is linked to the degree to which pores within the material are interconnected. Given that 222Rn from 238U has a longer half-life compared to other Rn isotopes, it has a greater probability of escaping the rock’s structure and reaching the outside environment. Certain types of microfissures may contain filling material with radioactive elements. The closer they are to the surface, the easier it becomes for radon gas to escape from the rock. Therefore, the greater the number of inter-grain microfissures, combined with trans-grain fissures, which can be over 5 to 6 cm long in the case of TI, MS and GI pegmatites, the greater will be the Rn exhalation rate (Fig. 6). The presence of minerals containing U in their chemical makeup, such zircon and allanite, leads to partial destruction of the crystalline structures of adjacent minerals, usually accompanied by micropore irradiation, which can theoretically facilitate the Rn migration. Fig. 7 shows that the porosity due to microgaps and the U amount are factors affecting the Rn exhalation in the rocks studied. 4.3. Radon emanation coefficient Each sample showed a different behavior for radon exhalation, expressed by emanation efficiency as defined by Wanty et al. (1992) through the following equation: E ¼ ð222 RnÞ air 222
ð
RnÞ air þ ð222 RnÞ rock
Wanty et al. (1992) considered that E is almost always lower than 0.5 (usually lower than 0.3). The U content (ppm) makes it possible to estimate the 226Ra activity (Bq/kg) by employing conversion factor 12.34. This value corresponds to 222Rn generated in the rock, under radioactive equilibrium conditions. The radon exhaled in the air, [Rn] (in Bq/m3), is converted to Bq/kg using the equation [Rn] V/m, in which V¼19 L and m is the sample weight (kg).
814
P.G.Q. Amaral et al. / Applied Radiation and Isotopes 70 (2012) 808–817
Fig. 4. Radon exhaled according to the accumulation period. Samples described in Table 1.
Table 5 contains the values obtained for E, showing that although some rocks exhibited low U levels, they exhaled more Rn gas than others possessing high levels of this radioelement.
This is likely related to the Rn arrangement in the crystalline minerals structure in the rock, with microgaps behaving as a favorable aspect for exhalation.
P.G.Q. Amaral et al. / Applied Radiation and Isotopes 70 (2012) 808–817
Fig. 5. Relationship between the U content and Rn exhaled after the accumulation period of 25 day.
815
The GI granitic pegmatite showed the highest exhalation efficiency (E¼0.169), among the studied rocks. It also featured the highest value of apparent porosity among all analyzed rocks (Table 2). The CC monzogranite contains the highest U concentration among the rocks analyzed, producing the most Rn. Nevertheless, its E value is comparable to those of PSG hypersthene diorite, VL charnockite and IL quartz-feldspar granulite that are rocks featuring much lower U content. Thus, the apparent porosity, low microfissuring degree/permeability and strong mineral engagement shown by CC monzogranite are important parameters for justifying its E value. Compared to other rocks, GC migmatized gneiss biotite and CB gneissified migmatite showed similar U levels, producing nearsimilar 222Rn values (in Bq/kg of rock). Rock J (CB) achieved higher emanation efficiency than rock I (GC), possibly due to its
Fig. 6. Relationships among the microcracks types and Rn exhaled after the accumulation period of 25 day.
Fig. 7. Relationships among the U content, apparent porosity and Rn exhaled after the accumulation period of 25 day.
816
P.G.Q. Amaral et al. / Applied Radiation and Isotopes 70 (2012) 808–817
Table 5 Radon emanation efficiency from radiometric data. Code
PSG OI VL CC TI MS GI IL GC CB
Table 6 Simulation of the Rn exhaled to a hypothetical indoor circulation environment.
U
226 Ra and 222Rn calculated from U content
222
Rn exhaled (after 25 day)
Rn emanation efficiency
(ppm)
226 Ra (Bq/kg)
222 Rn (Bq/kg)
222
Rn (Bq/m3)
222 Rn (Bq/kg)
E
136.36 35.91 94.65 456.95 278.02 64.17 79.10 95.02 227.92 252.23
136.36 35.91 94.65 456.95 278.02 64.17 79.10 95.02 227.92 252.23
12.00 21.67 10.00 69.33 1135.67 471.67 2483.33 23.00 98.33 752.00
0.07 0.13 0.07 0.44 8.01 3.14 16.10 0.14 0.69 4.66
0.001 0.004 0.001 0.001 0.028 0.047 0.169 0.001 0.003 0.018
11.05 2.91 7.67 37.03 22.53 5.20 6.41 7.70 18.47 20.44
Code
PSG OI VL CC TI MS GI IL GC CB
Simulated rock volume utilized (m3)
1.82 1.82 1.82 1.82 1.82 1.82 1.82 1.82 1.82 1.82
Simulated environment area (m2)
90.86 90.86 90.86 90.86 90.86 90.86 90.86 90.86 90.86 90.86
Exhaled Rn measured after 25 day (Bq/m3)
12.0 21.7 10.0 69.3 1135.7 471.7 2483.3 23.0 98.3 752.0
Simulated Rn exhaled to the indoor environmenta (Bq/m3)
(pCi/L)
1.43 2.54 1.27 8.25 145.40 57.67 293.06 2.62 12.47 83.65
0.04 0.07 0.03 0.22 3.93 1.56 7.92 0.07 0.34 2.26
a 148 Bq/m3 and 4 pCi/L are the threshold Rn concentration values recommended for homes (EPA, 2003).
petrographic/petrophysical characteristic. Even though it also features lower apparent porosity, it has a larger number of filled microfissures; as such, the location of U in the stone becomes one important factor for the efficiency. When comparing CB gneissified migmatite to TI hydrothermal granitic pegmatite, the E factor is higher for the pegmatite, which, in addition to featuring higher 222Rn level (in Bq/kg) also has greater apparent porosity, thus, contributing to more pronounced Rn exhalation from these rocks. Both MS and TI granitic pegmatites have similar petrographic characteristics determined by optical microscopy but different U content. The latter experienced significant U-enrichment as an effect of hydrothermal action, leading to different 222Rn levels, even though MS pegmatite has a significantly higher apparent porosity than TI granitic pegmatite. 4.4. Use as decorative stones in residences Ornamental stones have several uses in buildings, including flooring, baseboards, doorsills, doorjambs and wall siding, as well as in table tops, countertops and washbasins. They are often used as a single material or in different stone combinations to achieve a distinct effect. Based on the quantification of radon obtained from the samples and knowing the contribution to the exhalation rate from stones used for decorative purposes in buildings, the Chyi (2008) method was adapted to simulate how the use of the rocks would impact the radon levels inside residences or any given indoor environment. To that end, a 90.86 m2 area and 2.80 m high circulation environment was considered, where stone was applied as flooring. Factors such as indoor-outdoor air exchange were disregarded because one of the greatest risks of radon is its accumulation in a given environment, usually indoors. The formula Rnce ¼[Rne] F2/F1 was adopted, in which [Rne] is the amount of radon exhaled by the rock (in Bq/m3) that is provided by the Rn readings. Factor 1 (air dilution factor), F1, represents the recurrent air volume from within the circulation environment (Vce) in relation to the air volume inside the bottle (Vb), i.e. F1 ¼Vce/Vb. Factor 2 (material amplification factor), F2, is the ratio between the area of the material applied in the circulation environment (Ace) and the area of the sample used to monitor Rn (As), i.e. F2¼ Ace/As. The results for Rnce are given in Bq/m3 and later converted to pCi/L, as expressed in Table 6. In the case of the rock that surpassed the 4 pCi/L limit suggested by EPA (2003), the
evaluation was made only considering the flooring of bathing areas and a 3 m 0.60 m 0.02 m countertop in the kitchen, thereby allowing a comparison of the air in the house as a whole and in those environments. Six stones under consideration in the simulation showed values much below the 4 pCi/L limit suggested by EPA (2003): PSG hypersthene diorite, OI syenite, VL charnockite, CC monzogranite, GC migmatized gneiss and IL granulite. Two stones achieved intermediate values, one belonging to the pegmatite group (MS, 1.56 pCi/L) and another belonging to the metamorphic rocks group (CB migmatite, 2.26 pCi/L). These stones can influence the total radon levels when added to other existing materials in a given environment. The GI and TI granitic pegmatites showed high values; the former (7.92 pCi/L) stood above the 4 pCi/L EPA (2003) limit and the latter (3.93 pCi/L) was very close to the limit, requiring caution when used in enclosed or poorly ventilated environments. When GI granitic pegmatite is applied only as flooring in bathroom areas of the hypothetical environment, a reduction is observed in the added radon level for the air of the entire residence, standing at 0.77 pCi/L. Nevertheless, the 7.92 pCi/L concentration remains in the bathroom areas. When a countertop is added in the kitchen, radon levels reach 1.46 pCi/L, with overall residence added levels increasing to 0.89 pCi/L. It is concluded that, although the residence now has radon levels below that suggested by EPA (2003), bathroom areas remain above the 4 pCi/L limit and the kitchen area begins to exert a certain influence in case more radon-releasing materials are added. This simulation shows certain precautions with regard to the specification of rocky materials to be used in buildings, especially in spacious and enclosed environments, with radon gas exhalation rates representing an important factor to be considered. The increase in radon levels in the case of dimension stones is related to the amount of material used in a given environment, i.e. the greater the volume of stone used, the higher the added radon levels will be.
5. Conclusion Radiometric data on uranium and the results of the monitoring of 222Rn gas exhalation closely match the genetic nature of petrographic types and feature a good correlation with their respective petrographic parameters and physical indices. By quantifying Rn exhalation in the studied rocks, it was concluded that it is not only the quantity of U in a stone that
P.G.Q. Amaral et al. / Applied Radiation and Isotopes 70 (2012) 808–817
should be analyzed as a parameter of radon gas exhalation, but also petrographic/petrophysical aspects such as their microporous network. The comparison between the quantity of 222Rn produced by a given stone and the amount emanated from the rock, shown as factor E (emanation efficiency), demonstrates that not all the radon gas produced in the stone is actually emanated, and that the rock that produces the most Rn will not always emanate more. Thus, it should be emphasized that radon exhalation, as well as being related to uranium amounts, is also linked to the ease with which radon gas is able to escape from the crevices of a rock. It is therefore concluded that radon gas exhalation in the studied dimension stones results from the petrographic/petrophysical aspects of those rocks. Among the rocks under consideration, six showed values much below the 4 pCi/L EPA (2003) limit when applied as flooring: PSG hypersthene diorite, OI syenite, VL charnockite, CC monzogranite, IL granulite and GC migmatized gneiss. Two stones achieved intermediate values, between 1.5 and 2.5 pCi/L: MS pegmatite and CB migmatite. The TI and GI granitic pegmatites showed high values: the former was very close to the 4 pCi/L limit, whereas the latter stood above the limit, both requiring caution when used in closed or poorly ventilated environments. The increase in radon levels in a given environment depends on the quantity of stone applied, its exhalation rate, as well as the volume of air ambient considering ventilation (indoor/outdoor air exchange), so that if one of the conditioning factors is increased, the other is reduced/increased.
Acknowledgments To Brazil’s CNPq (National Council for Scientific and Technological Development) for the financial support. One anonymous reviewer is thanked by helpful comments that improved the readability of the manuscript.
817
References ABIROCHAS (Associac- a~ o Brasileira da Indu´stria de Rochas Ornamentais), 2008. Radioatividade natural em materiais rochosos, Informe 18/2008, /http:// www.abirrochas.com.brS. ABIROCHAS (Associac- a~ o Brasileira da Indu´stria de Rochas Ornamentais), 2011. ~ Balanc- o das exportac- oes brasileiras de rochas ornamentais para os EUA em 2010, Informe 06/2011, /http://www.abirrochas.com.brS. ABNT (Associac- a~ o Brasileira de Normas Te´cnicas), 1992. Rochas para revestimento - determinac- a~ o da massa especı´fica aparente, porosidade aparente e absorc- a~ o d’a´gua aparente. NBR 12766. ABNT, Sa~ o Paulo. Balca´zar, M., Lo´pez, A., Streil, T., Cha´vez, A., 2002. Radon leakage assessment in a controlled radon laboratory. Geofı´sica Int. 41 (3), 281–284. Bonotto, D.M., 1996. Comportamento hidrogeoquı´mico do 222Rn e iso´topos de 238U ~ controladas em laborato´rio e em sistemas naturais. Post e 234U sob condic- oes Ph.D. Thesis, UNESP-Universidade Estadual Paulista, Rio Claro, SP, 223 pp. Chyi, L.L., 2008. Radon Testing of Various Countertop Materials. Final Report. The University of Akron: Akron, 15 pp. CPRM (Servic-o Geolo´gico do Brasil), 2003. Geologia, Tectoˆ nica e Recursos Minerais do Brasil. CPRM, Brası´lia, DF. Duarte, C.R., 2002. Radioelementos naturais na a´rea do projeto Rio Preto (GO). Ph.D. Thesis, UNESP-Universidade Estadual Paulista, Rio Claro, SP, 169 pp. EPA (US Environmental Protection Agency), 2003. Assessment of risks from radon in homes. EPA, Washington, DC 99 pp.. Fraza~ o, E.B., Farjallat, J.E.S., 1995. Caracterı´sticas tecnolo´gicas das principais rochas silica´ticas brasileiras usadas como pedras de revestimento. In: Atas do 11 Congresso Internacional da Pedra Natura. Lisboa, Portugal, pp. 47–58. Moreira-Nordemann, L.M., 1977. Etude de la vitesse d’alteration des roches au moyer de l’uranium utilize comme traceur natural. Application a deux bassins du nordest du Bresil. Tese (Doctorat d’ E´tat), Universite´ Pierre et Marie Curie, Paris, 162 pp. Murphy, K., 2008. What’s lurking in your countertop?. The New York Times /http:// www.nytimes.com/2008/07/24/garden/24granite. html?pagewanted=1&_r=1&sq=radon&st=cse&scp=1S. Pertlik, F., Roger, J.J.N., Adams, J.A.S., 1974. Uranium. In: Wedepohl, K.H. (Ed.), Handbook of Geochemistry, Springer-Verlag, Berlin, p. 92.0192.08. Salas, H.T., Nalini Jr., H.A., Mendes, J.C., 2003. Radioatividade em rochas ornamentais do Brasil. In: Vidal, F.W.H., Lima, M.A.A.B. (Eds.), Anais do IV Simpo´sio de Rochas Ornamentais do Nordeste. Fortaleza, CE, pp. 68–73. Ulbrich, H.H.G.J., Gomes, C.B., 1981. Alkaline rocks from continental Brazil. Earth Sci. Rev. 17, 135–154. Wanty, R.B., Lawrence, E.P., Gundersen, L.C.S., 1992. A theoretical model for the flux of radon from rock to groundwater. In: Gattes, A.E., Gundersen, L.C.S. (Eds.), Geologic Controls of Radon, Geological Society of America, Boulder, CO, USA.