Journal of Radioanalytical and Nuclear Chemistry (2021) 328:911–923 https://doi.org/10.1007/s10967-021-07709-9 Assessment of natural radioactivity and radon exhalation rate associated with rock properties used for construction in greater Accra region, Ghana Francis Otoo1,2,3 · Emmanuel Ofori Darko1,2 · Massimo Garavaglia3 · Oscar Kwaku Adukpo1,2 · Joseph Kwabena Amoako1,2 · Joseph Bremang Tandoh2,4 · Stephen Inkoom1,2 · Samuel Nunoo5 · Simon Adu2,6 Received: 26 November 2020 / Accepted: 1 April 2021 / Published online: 27 April 2021 © Akadémiai Kiadó, Budapest, Hungary 2021 Abstract Natural radioactivity, radon exhalation and rock properties have been studied using gamma spectrometry, CR-39 detectors and weighing methods. Statistical tool was used to study the relationship between the activity concentration and properties associated with the rocks. The purpose is to determine the rocks with good properties and lower radioactivity levels in order to reduce radiological effect when used as building materials. Gneiss recorded good properties while quartzite was found to contain poor and lower value of radiological parameters as compared to granite and sandstone. Radon-222 was found to correlate well with the properties than radionuclides of radium-226, thorium-232 and potassium-40. Keywords CR-39 · HPGe · Rocks · NORM/radon · Statistical analysis Introduction and shaping without secondary flaws, and durability. The various ways in which rock is used for construction depends Rock aggregate is the most abundant natural resource and largely upon the geological properties, quality to resistance widely used as a construction material. Rocks normally used to weathering, water absorption capacity degradation, pre- in the country are granite, quartzite, granite, marble, gneiss, vailing environmental conditions and availability. However, and sandstones. It is commonly used in the building indus- studies conducted on rock samples indicated that both rock tries in the form of concrete because of their properties such properties and naturally occurring radionuclides are asso- as resistance to weathering, textural and colour patterns, ciated with composition of the rocks. Naturally occurring crushing strength, abrasive strength, amenability to cutting radioactive materials (NORM) are dominated by the activity concentrations of the uranium and thorium radionuclides and potassium-40. The most important factors affecting the * Francis Otoo radon exhalation rate are the distribution of the parent radio- kwaotoo@yahoo.com nuclide of radium and the internal structure of the rocks [1]. 1 Radon which is the direct decay daughter of radium is the Radiation Protection Institute, Ghana Atomic Energy Commission, P.O. Box LG 80, Legon, Accra, Ghana largest source of the radiation exposure to the populace due 2 to its greater mobility as compared to radium, thorium and School of Nuclear and Allied Sciences, University of Ghana, P.O. Box AE1, Atomic Campus, Legon, Accra, Ghana potassium which are permanently present in solid matter. 3 Radon gas can easily leave rocks by escaping into fractures Centre for Radiation Protection, Agency for Environmental Protection, Friuli Venezia Giulia, Via 42 Colugna, and openings into enclosed area. 33100 Udine, Italy Studies conducted on rocks have shown that radon gas 4 National Nuclear Research Institute, Ghana Atomic Energy decreases during degradation and weathering as a result of Commission, P.O. Box LG 80, Legon, Accra, Ghana higher levels of porosity which occur during these processes 5 Department of Earth Science, University of Ghana, P.O. Box [2–5]. Water absorption was also found to affect the NORM LG 58, Legon, Accra, Ghana concentration and transport parameters for radon gas [6–8]. 6 Nuclear Regulatory Authority, Atomic Energy, P. O. Box AE The properties such as durability, water absorption content, 50, Neutron Avenue House 1&2, Accra, Ghana apparent and bulk densities are linked to the quality of the Vol.:(012 3456789) 912 Journal of Radioanalytical and Nuclear Chemistry (2021) 328:911–923 rocks used as a construction material [9–13]. Studies done obtained from Building and Road Research Institute (BRRI) by some other researchers indicated that water absorption of the Council for Scientific and Industrial Research (CSIR), contents, porosity, pore sizes and other geological proper- Kumasi-Ghana. These data were used to determine the types ties present in the rocks have adverse influence on the con- of samples to be collected and the sampling points. centration of radionuclides and the movement of the radon Total of sixty-two (62) rock samples from nineteen (19) gas [2–5, 14–16]. In Ghana, most of the studies done only locations from quarries, pits and building sites were sam- focussed on natural radioactivity in environmental samples pled across the two regions as shown in Fig. 1. The rocks from mines and some building materials [17–21]. Some were collected randomly based on availability and acces- studies have also focussed on the radon exhalation rate in sibility. Three (3) or four (4) samples were taken from dif- soils and rocks [20, 22]. ferent areas of the same study location. The samples were Natural radionuclides and radon exhalation rate levels in then placed into different labelled black polythene bags and rocks are of particular interest to this research because it is sent to the laboratory for preparation and analysis. These the most common and naturally abundance in rocks. Natural rocks are mostly used in construction of houses, bridges, and radioactivity and their properties are part of the composition roads in the study area. Naming and description of the rocks of the rocks throughout the earth crust. Studies conducted in were done at the Department of Earth Science, University some environmental samples including rocks showed some of Ghana, whilst their properties were also determined at level of correlation of radioactivity with geological proper- Central Materials Laboratory of Ghana Highway Authority, ties [2, 23, 24]. However, there has not been known studies Accra, Ghana. The activity concentration and radon analysis that determine their linkage or relationship with rock proper- were done at Centre for Radiation Protection of Agency for ties in Ghana. As a result, there is paucity of knowledge and Environmental Protection of Friuli Venezia Giulia, (ARPA, data on how to select good rocks with low level of radionu- FVG), Udine, Italy. For these analyses, the samples were clides for construction industries. pulverized, homogenized, air-dried and sieved to a uniform This study used Pearson’s correlation coefficient, cluster mixture with a particle size of about 2 mm, and sealed in and principal component analysis to measure the relationship 500 mL beakers. The sealed samples were weighed stored between activity concentrations of radon-222, radium-226, at room temperature for a period of 3–4 weeks to allow 238U thorium-232, potassium-40 and durability, water absorption and 232Th decay series to reach radioactive equilibrium with content, bulk and apparent densities as rock properties. This the short-lived progenies [21, 27]. The study rocks are clas- is with the view to determine the low-NORM rock with qual- sified as basin granite (BG), mica granite (MG), fresh foli- ity properties suitable for construction. This would also help ated quartzite (FFQ), and micaceous quartzite (MQ), highly to prevent rocks with high level of radiological content being weathered quartzite (HWQ), gritty pebbly sandstone (GPS), used as building materials in Ghana. fine grained sandstone (FGS), garnet hornblende gneiss (GHG), garnet pyroxene gneiss GGP and hornblende gneiss (HG). Materials and methodology Geology of the study area Determination of activity concentration, radon exhalation rate and rock properties The study area covers the Greater Accra and some parts of Central Region of Ghana. The geology is made up of dif- Calibration of high‐purity germanium detector (HPGe) ferent types of rocks and soils. The main rock types of the study area are Precambrian Dahomeyan schists, granodior- Full energy peak efficiency calibration was done with a ites, granitic gneiss, and amphibolites to late Precambrian multigamma-ray liquid source (from 46.5 to 1836.1 keV) Togo series comprising mainly of quartzite, phyllites, phy- with density 1 ρ/cm3 in the same geometry in as the samples litones, and breccias [25, 26]. Many of these deposits are to be measured [28]. The detector is surrounding by lead being exploited in an uncontrolled manner for constructional shield of 100 mm lined with copper, cadmium and plexi- purposes. glass sheets. The specific activities of the standard source used for the calibration are; 210Pb (6.63 × 1 05 Bq), 241Am Sampling and sample preparation (6.60 × 1 04 Bq), 109Cd (4.39 × 1 05 Bq), 57Co (1.23 × 1 04 Bq), 123mTe (7.65 × 103  Bq), 113Sn (2.72 × 104  Bq), 85Sr Rock samples most commonly used for construction in the (1.05 × 1 04 Bq), 137Cs (9.81 × 1 04 Bq), 88Y (4.63 × 1 04 Bq) study area across the two regions along the southern zone 60Co (1.16 × 105 Bq). For the measurement of samples with of Ghana are marbles, sandstone, quartzite, gneiss and gran- higher density, corrections for self-absorption were carried ites. Geological and processed building material data were out with the GESPECOR software [29]. 1 3 Journal of Radioanalytical and Nuclear Chemistry (2021) 328:911–923 913 Fig. 1 Map of the study area Activity concentration beaker under identical measurement conditions as the sam- ples. Counting time was 72,000 s. The data acquisition, The measurement of the activity concentrations of 226Ra display and on-line spectrum analysis were carried out and 232Th was done by using the energy lines of the daugh- using the Genie 2000 V3.3 (1) spectroscopy software from ter products. For a nuclide having more than one peak Canberra [20, 21, 27]. The activity concentration (AC) in in the spectrum, the activity concentration was obtained (Bq/kg) of each radionuclide in any given sample was cal- as the weighted average activity at each peak. The emis- culated from the spectrum using the following analytical sions of 214Bi (609.31, 1120.29, 1764.49 keV) and 214Pb expression. (295.22, 351.93 keV) were assumed to represent the activ- ( ) ity of 238U. The lines of 212Bi (727.33 keV), 228Ac (209.25, N226 232 40 samAc Ra, Th, K = (1) 409.46, 463.0, 794.95, 911.20, 964.77, 968.97 keV) and P(E) ∗ n(E) ∗ T ∗ Mc sam 212Pb (238.63, 300.09 keV) were used to evaluate the where M (kg) is the mass of sample, N (cps) is the activity of 232Th. 40K was determined using its only γ-ray sam samnet peak area for the sample in the peak range, P(E) is the line of peak 1460.82 keV. Prior to sample measurement, gamma emission probability, T (s) is the counting time in the background was determined with an empty Marinelli c  1 3 9 14 Journal of Radioanalytical and Nuclear Chemistry (2021) 328:911–923 seconds, and η(E) is the photo peak efficiency which had been V2 cRn obtained from the standard solution. Radon Exhalation rate ERn (Bq/m h) = (2)SaTc Radon exhalation rate measurement where Vc volume of diffusion chamber (m3), Sa is the surface area of the sample ( m2), Rn is the decay constant of radon A tight closed vessel technique with cylindrical jar was (1 ∕s )  is the calibration factor of detector (track/cm2d/(Bq/ used for radon exhalation rate measurements from rocks. m3),  is the measured surface density of tracks (tracks/cm2) A known weight of each rock sample with 78.5 c m2 sur- and Tc is the effective exposure (s) in the diffusion chamber. face area was used for radon exhalation rate. The detection system were done by installing detectors at the bottom of a Determination of rock properties chamber that covers at a distance of 22 cm from the surface of the sample in order to count for only radon (222Rn) and The rock properties were used to determine the quality of the prevent thoron from evading the surface of the detectors rock samples use for buildings. The rock properties consid- [30]. The radon exhalation measurements were performed ered for this study are durability, water absorption contents by placing the rock samples at the bottom of glass containers (WAC), bulk (BUK) and apparent density (APP). The rock of 10 cm diameter and 25 cm height. The ratio of volumes of properties were determined using the standard procedures the containers and samples were more than ten (10) to pre- proposed by International Organizations [9–11] and adopted vent or reduce the probability of back diffusion [31]. Before by the Materials Division of Ghana Highway Authority, CR-39 detector was installed in the container, the samples Accra, Ghana. were completely sealed and stored for one (1) month so that 226Ra will attain equilibrium with its progeny. This step was Durability test necessary to ensure that the radon gas and its decay products are confined within the sample. The detectors were care- The durability (DBT) of the rock was determined using Los fully installed and the containers were hermetically sealed Angeles Abrasion Test (LAAT) [9–11]. It is an abrasion test for three (3) months. in which loss in mass is caused by the impact of steel balls dropping on the geological aggregate samples in a rotating Preparation and analysis of detectors drum. The overall durability of rock sample was estimated using the following mathematical expressed as follows. The analysis of the detectors were done by preparing 4000 ml of distilled water in a thermostatic bath and every minute AWAG − ARAG one spoon of 1000 g pellet sodium hydroxide was added LAAT% = × 100 (3)AWAG until the container was completely empty and the solution reaches temperature of 90 °C. RadoSlide with chips was where AWAG is the total weight of aggregate and ARAG is then mounted on an etching carousel and chemically etched the total weight retained after rotation for 500 revolutions in the thermostatic bath with a solution for a time period of [9–11]. 4 h and 15 min. It was then covered and the stirrer in the thermostatic bath stirred the water vigorously throughout the Water absorption period to ensure uniformity of the solution. A new solution made of 36 ml of 96% diluted acetic acid and 4000 ml of The water absorption contents (WAC) is defined as the distilled water was prepared and poured into the thermostatic increase in mass due to water present in the pores of the rocks. bath for fifteen (15) minutes to wash the dosimeters from the The WAC was done to know the water holding capacity and to chemicals. Finally, the detectors were washed in 4000 ml measure the strength or quality of the rocks for building and of distilled water for fifteen (15) minutes and later drained construction. It was determined by weighing rock aggregate into a wash basin in the laboratory. The cover door of ther- and then placed in pycnometer with tap water at a temperature mostatic bath was opened and the etched carousel removed of 28 °C with a cover of water above the sample. Immediately and placed in tray for rinsing and drying for four (4) days after the immersion of the sample aggregate into the water, the after which the detector holder with the radon test plastic entrapped air was again removed from the sample by shaking chips were taken for evaluation with Radometer. The latent the pycnometer. The samples were then weighed in surface tracks formed on the detectors were scanned and counted in dry until no further moisture could be expel and allowed to dry 144 fields using an optical microscope of 40× magnification for 15 min to ensure completely surface dry. The surface dried objective lens. The tracks density left on track films were aggregate is then weighed and kept in an oven at a temperature then used to evaluate the radon exhalation rate using the of 110 °C for 24 h. It was then weighed again after removing following equation [15]: from oven and allowed to cool to a room temperature. This 1 3 Journal of Radioanalytical and Nuclear Chemistry (2021) 328:911–923 915 process was repeated for other rocks using international stand- MPw is the mass of pycnometer and water (g) MPDAGW is ard procedures [9–11]. the mass of pycnometer, water and dry rock samples (g) and The water absorption content (WAC) was calculated and MPDAG is the mass of pycnometer and dry rock samples (g). expressed in percentage using the expression below; MAG W × 100 (4) Result and discussionAG= MDA where MAG is the mass of rock samples (g) and M is the Activity concentration and rock properties DA mass dry rock samples (g). Bulk density is define as the total mass of the soil or rocks The measured activity concentrations of the naturally occur-226 232 40 including s solids, water and air per unit volume, while appar- ring radionuclides of Ra, Th and K and rock proper- ent density is define as mass of rock solid per unit volume of ties were determined and presented in Tables 1 and 2. The the rock. natural radioactivity and rock properties were found not The Bulk Density of rock samples (BUK) is given as to be uniformly distributed throughout the locations and rock types. This may be due to geological properties, pores BUK = WT×S.G (5) sizes, differences in densities, mineral composition, moisture AG content, pores and the structural size of the rocks [2, 16]. where S.GAG is the calculated specific gravity of the rock The obtained results showed the mean activity concentra- and ρWT is the density of the water at test temperature. tions of the 226Ra, 232Th and 40K in rocks from the different Apparent density of the rock samples (APP) was calculated location ranging from 226Ra (5–54) 29 ± 16 Bq/kg, 232Th using the equation below (18–65) 42 ± 14 Bq/kg and 40K (65-1222) 551 ± 378 Bq/  kg as presented in Table 1. The corresponding values for WT×M DAG A =D ( ) (6) rock properties also varied from durability (45–77) 60 ± 11, ( M −M − (M −M )) PW P PDAGW PDAG water absorption content (0.4–2) 1 ± 0.5, bulk (2–3) 3 ± 0.2 where ρ is the density of water at test temperature, M and apparent densities (2–3) 3 ± 0.2 presented in Table 2 WT DAG is the mass of dry rock samples (g). respectively. The highly weathered quartzites (HWQ) from Ofankor and Pokuase obtained lowest values for 226Ra, 232Th Table 1 Radon exhalation rate Study location Sample code Sample no. ERn (mBq/m2) Radionuclides activity concentra- and activity concentration of tion (Bq/kg) radionuclides 226Ra 232Th 40K Mean Mean Mean Mean Achimota FGS 3 107 ± 10 34 ±3 41 ± 4 315 ±32 Afienya HG 3 129 ± 24 52 ±5 55 ± 6 878 ±80 Anyarkode GHG 3 118 ± 10 42 ±4 51 ± 5 1200 ±106 Asutuare GGP 3 121 ± 13 43 ±4 57 ± 6 1222 ±109 Atomic HWQ 3 36 ± 3 6 ±0.6 20 ± 2 65 ±6 Atomic Hills MQ 3 97 ± 9 34 ±3 40 ± 3.9 331 ±33 Bukor Allkope GHG 3 118 ± 11 42 ±4 50 ± 4 1200 ±104 Dodowa FFQ 4 65 ± 6 26 ±2 36 ± 2 454 ±40 Dominase MG 4 97 ± 9 32 ±3 60 ± 6 739 ±65 Haatso MQ 3 97 ± 9 31 ±4 39 ± 3 339 ±32 Kasoa BG 3 88 ± 8 29 ±2 65 ± 7 739 ±64 Mccarthy MQ 4 102 ± 10 41 ±4 43 ± 5 340 ±30 Ofankor HWQ 3 33 ± 3 5 ±0.3 19 ± 2 70 ±8 Oyarifa FFQ 3 55 ± 4 16 ±2 39 ± 3 445 ±32 Oyibi FFQ 3 49 ± 4 14 ±1 37 ± 2 447 ±32 Pokuase HWQ 3 39 ± 3 6 ±0.4 18 ± 2 67 ±8 Shai Hills HG 4 131 ±13 54 ± 6 57 ±6 880 ±77 Tesano GPS 3 40 ± 4 6 ±0.6 35 ± 2 405 ±44 Weija MQ 4 97 ± 9 34 ±6 40 ± 3 331 ±29 1 3 916 Journal of Radioanalytical and Nuclear Chemistry (2021) 328:911–923 Table 2 Rock properties Study location Sample code Sample no. Rock properties DBT WAC BUK APP (%) (%) (g/cm3) (g/cm3) Mean Mean Mean Mean Achimota FGS 3 68 ± 16 2 ± 0.2 2 ± 0.2 2 ± 0 2 Afienya HG 3 47 ± 10 0.4 ± 0.02 3 ± 03 3 ± 0.3 Anyarkode GHG 3 50 ± 12 0.5 ± 0.04 3 ± 0.2 3 ± 0.3 Asutuare GGP 3 47 ± 10 0.5 ± 0.1 3 ± 0.3 3 ± 0.3 Atomic HWQ 3 77 ± 16 2 ± 0.02 2 ± 0.2 2 ± 0.1 Atomic Hills MQ 3 65 ± 16 1 ± 0.01 3 ± 0.2 3 ± 0.3 Bukor Allkope GHG 3 49 ± 11 0.5 ± 0.03 3 ± 0.2 3 ± 0.3 Dodowa FFQ 4 60 ± 15 1 ± 0.01 3 ± 0.2 3 ± 0.2 Dominase MG 4 49 ± 12 0.8 ± 0.1 3 ± 0.2 3 ± 0.2 Haatso MQ 3 68 ± 13 1 ± 0.01 3 ± 0.2 3 ± 0.3 Kasoa BG 3 48 ± 10 0.8 ± 0.1 3 ± 0.3 3 ± 0.2 Mccarthy MQ 4 66 ± 15 1 ± 0.01 3 ± 0.2 3 ± 0.2 Ofankor HWQ 3 78 ± 16 2 ± 0.02 2 ± 0.2 2 ± 0.2 Oyarifa FFQ 3 62 ± 13 1 ± 0.01 3 ± 0.2 3 ± 0.2 Oyibi FFQ 3 63 ± 16 1 ± 0.01 3 ± 0.2 3 ± 0.2 Pokuase HWQ 3 73 ± 15 2 ± 0.01 2 ± 0.1 2 ± 0.1 Shai Hills HG 4 45 ± 9 0.4 ± 0.02 3 ± 0.2 3 ± 0.3 Tesano GPS 3 65 ± 16 1 ± 0.01 3 ± 0.3 3 ± 0.2 Weija MQ 4 67 ± 16 1 ± 0.01 3 ± 0.2 3 ± 0.2 and 40K as shown in the Table 1 respectively. The maximum potassium isotopes permanently present in the solid matter values of 226Ra, 232Th and 40K were obtained in hornblende of the rocks [2, 4, 14, 22, 33]. gneiss (HG), basin granite (BG), garnet hornblende gneiss The radon exhalation rate was found to be in the range of (GHG) from Shai Hills, Kasoa and Asutuare respectively. 33 to 131 mBq/m2h with the mean value of 85 ± 34 mBq/ The activity concentrations of 232Th were much higher than m2h. The highest radon exhalation rate was found in the 226Ra. 40K was found to contain the highest radionuclide hornblende from Shail Hills with lower water absorption and content from the rocks in all the studied locations. This result durability values. It can be attributed to pores size, mineral was in an agreement with some studies done by [18, 21, 29, composition as well as the porosity of the rocks, radon gas 32]. It could be attributed to potassium isotopes being most makes very difficult to escape its origin. The lowest radon abundant radionuclide in the environment. Higher potassium exhalation rate obtained in a highly weathered quartzite associated with the rocks does not really cause a major radi- with higher levels of water absorption and durability but ological effect as compared to radium and thorium. This is with lower densities was also found in Ofankor area. It may because 40K is a beta minus emitter during decay of which attribute to the more pore space as a result of weaker and most of these rays escape from the body representing very fragile crystalline texture associated with highly weathered small fraction of the background dose rate from all-natural quartzite. It was observed that where 226Ra concentration sources. The highly weathered quartzite was found to con- was higher, the 222Rn exhalation rate is higher in the studied tain lower levels of radionuclides and densities while water areas as presented and described in Table 1. The seventeen absorption and durability values were found to be high. percent (17%) of the radon exhalation values were found The gneiss rocks were found to contain lower levels of to be greater than the world average of 125 mBq/m2h [34]. water absorption and durability values but with higher den- sities and radionuclides especially hornblende and garnet hornblende from Shai Hills and Anyarkode as presented in Pearson’s correlation coefficient analysis Tables 1 and 2. Lower values of water absorption and dura- bility and higher bulk and apparent densities may due to Statistical correlation analysis was carried out using Pear- structural compactness in the gneiss. The highest concen- son’s statistical tool. Correlation coefficient was performed tration for 226Ra, 232Th and 40K in the gneiss, may be due to in order to measure direction of a linear relationship and higher amounts of uranium and thorium series as well as the determine the strength of the linkage between the pairs of the 1 3 Journal of Radioanalytical and Nuclear Chemistry (2021) 328:911–923 917 radionuclides and rock properties associated with the stud- Table 4 Cluster membership Cases 3 Clusters ied rock. The correlation pairs for this study: 222Rn/226Ra, 222Rn/232Th, 222Rn/40K, 222Rn/DBT, 222Rn/APP, 222Rn/BUK, DBT 1 222Rn/WAC, 226Ra/232Th, 226Ra/40K, 226Ra/DBT, 226Ra/ APP 3 APP, 226Ra/BUK, 226Ra/WAC, 232Th/40K, 232Th/DBT, 232Th/ BUK 2 APP, 232Th/BUK, 232/WAC, 40K/DBT, 40K/APP, 40K/BUK, WAC 1 40K/WAC, DBT/APP, DBT/BUK, DBT/WAC, APP/BUK, 222Rn 2 APP/WAC and BUK/WAC are presented and described in 226Ra 2 Table 3. The correlation was significant at the 0.01 level 232Th 2 (2-tailed). The positive and negative coefficients were found 40K 2 to exist between the natural radionuclides and rock prop- erties. The highest positive correlation coefficient of 0.83 was observed in radionuclides of 222Rn and 226Ra. This indi- This may be due to the fact that subjecting the rocks to the cated that 222Rn exhalation rate was mainly dependent on high levels of durability will result in the degradation of the decay of the 226Ra radionuclide present in the rock. It the interface structure or decrease in compactness of the can be also explained that the higher the radium level, the rock. This may lead to higher porosity which results in lower greater the chances of the rocks having high levels of radon level of radon gas passing through the studied rocks [6]. The gas. This result is in an agreement with the other research same reason may be attributed to the good negative correla- work [21, 35, 36]. The correlation of 222Rn with 232Th and tion coefficient for the pairs: 226Ra/DBT, 233Th/DBT, 226Ra/ 40K, 226Ra with 232Th and 40K, 232Th with 40K showed high WAC, 232Th/WAC, 40K/WAC and 40K/DBT. levels degree of positive correlation coefficient as presented The highest positive correlation coefficient of 0.71 was in Table 4. It may be due to the radionuclides level being found between a pair of variable 222Rn/APP as shown in influenced by the same sources of geological properties from Table 3. The highest positive coefficient recorded in 222Rn/ their origin [4, 14, 36]. The radon exhalation rate was found APP may be due to high mobility of radon gas present in the to correlate well with rock properties than the natural radio- rock samples without water content. This may also account nuclides of 232Th and 40K as shown in Table 4. This may be for the highest positive coefficient of apparent density with attributed to the high level of mobility of the radon gas as radon gas than that of the bulk density. The positive correla- compared to the activity concentration of the radionuclides tion observed is an indication that increase in the densities of 226Ra, 232Th and 40K which is permanently present in the will lead to decrease in porosity of the rocks, which will solid matter. result in an increase in the radon exhalation rate. The posi- The highest negative correlation coefficient of 0.79 was tive correlation associated with radionuclides and densities obtained in pairs of 222Rn and WAC. The negative correla- can be linked to the same reason with that of radon for the tion indicated that, radon gas exhalation is inversely pro- pairs: 226Ra/APP, 226Ra/BUK, 232Th/APP, 232Th/BUK, 40K/ portional to the water absorption content. In other words, APP and 40K/BUK. Mostly, activity concentration of radio- saturated rocks will have less radon gas exhaled as compared nuclides increases inversely with the grain size [6, 8, 24, 37] to the dry rocks [15, 33, 35]. The negative correlation can and, in proportion with the density of the rock aggregates also be linked to decrease in mass of the water content in [8]. the pores which lead to an increase in the radon gas exha- This may be the reason for the positive and negative lation from the rocks. The good negative relationship was Pearson’s correlation coefficient observed for natural radio- also observed between 222Rn and DBT as shown in Table 3. nuclides and rock properties as presented in the Table 3. Table 3 Correlation Coefficient 222Rn 226Ra 232Th 40K DBT APP BUK between radionuclides and rock properties of rock samples 222Rn 226Ra 0.83** 232Th 0.75** 0.79** 40K 0.71** 0.74** 0.69** DBT − 0.77** − 0.70** − 0.59** − 0.69** APP 0.71** 0.64** 0.62** 0.60** − 0.54** BUK 0.67** 0.60** 0.61** 0.41** − 0.51** 0.59** WAC − 0.79** − 0.78** − 0.69** − 0.67** 0.69** − 0.71** − 0.61** **significant at the 0.01 level (2-tailed) 1 3 918 Journal of Radioanalytical and Nuclear Chemistry (2021) 328:911–923 The result showed that radionuclides content in the studied observed in DBT and WAC, this can be linked to similarities rocks are been influenced by behaviour of different geologi- of geological structural patterns or porosity associated with cal properties from different locations. the rocks. These similarities maybe attributed to the negative correlation coefficient between the DBT and WAC with all the natural radionuclides as shown in Table 3. Cluster analysis The cluster analysis was done by using the average link- age method and correlation coefficient distance as shown in Cluster analysis was performed to identify similar charac- dendrogram Fig. 2; Table 3. The result for the dendrogram teristics or features among natural radionuclides and rock and cluster membership showed that, all eight parameters properties. Clusters 1 and 2 were formed based on geological were grouped into three statistically significant clusters. properties of the rock parameters. The clustering of all natu- All the natural radionuclides of 222Rn, 226Ra, 232Th and 40K ral radionuclides of 222Rn, 226Ra, 232Th and 40K infer cluster formed a single cluster (cluster 3) which was distinct from 1 as shown in the Table 4; Fig. 2. This may attribute to the the other two clusters (1 and 2) and comprised of (APP, fact that they are all naturally occurring radioactive materials BUK) and DBT, WAC). and their decay series occurred in the same natural origin. The second cluster, cluster 2 is made up of APP and BUK; these parameters possess similarities in their densities Principal component analysis which is dependent on the composition of the rock structure. These similarities may be the reason for the positive correla- The principal component analysis (PCA) was employed to tion coefficient of APP and BUK with all the natural radio- study the variables of different groups of natural radionu- nuclides as shown in Table 3. The third clusters were also clides and rock properties to determine their correlation, Fig. 2 Dendrogram obtained by the site clustering analysis with both radionuclides and rock properties data for rock samples 1 3 Journal of Radioanalytical and Nuclear Chemistry (2021) 328:911–923 919 similar behaviour and common sources of origin [38, 39]. The Varimax rotation with Kaiser Normalization method was used to determine differences between radionuclides and rock properties shown in Table 5. In Table 5, correla- tion matrix, the eigen values and eigen vectors are extracted to explain the number of significant factors and the percent of variance. Factor analysis performed on the data revealed three main components as confirmed by clustering member- ship and dendrogram analysis. These three factors contrib- uted 93.9% of the cumulative variance measured on the data. The component plot showed clusters in 3 dimensions. The Fig. 3 as depicted below shows a factor analysis plot of all the study parameters. 226Ra, 232Th and 222Rn can be observed clustering while 40K was separated from the other radionu- clides as depicted in Fig. 3. The closeness between 226Ra and 232Th can be linked to their decay series that occurs together in nature while 40K may be emanated from the source of origin which is different from the 226Ra and 232Th [36]. The factor analysis also showed that the closet between the pair of BUK/APP and DBT/WAC were found to be in different separate quadrant of the box plot as depicted in Fig. 3. Fig. 3 Box plot for both radionuclides and geological properties data This can be attributed to the different structure forma- tion, location, types and properties of the rocks. This may be the reason for Pearson’s negative and positive correlation 226Ra and 232Th concentrations observed may be due to coefficient between the pair of BUK/APP and DBT/WAC radium and thorium series which occur together in nature respectively. The factor plot confirmed the earlier results but they are chemically dissimilar and behave differently obtained from the cluster and dendrogram. during separation of parent and daughter decay processes. Factor analysis with only radiological data was done in order to determine further variance [38, 39]. It was revealed in Fig.  4 that, 222Rn and 226Ra were closely bounded, separated from 232Th and 40K. The closeness of the radon and radium can be attributed to 222Rn as the parent radionuclide of 226Ra and existence depends on an amount of activity concentration of 226Ra present in the rocks. It may be also due to the 222Rn which is the daugh- ter radionuclide produced from the decay of 226Ra within the 238U decay series. This is in agreement with Pearson’s correlation coefficient result. The further distance between Table 5 Rotated component matrixa Component 1 2 3 DBT − 0.85 − 0.34 − 0.35 APP 0.43 0.44 0.77 BUK 0.24 0.86 0.34 WAC − 0.65 − 0.67 − 0.27 222Rn 0.71 0.62 0.28 226Ra 0.74 0.61 0.23 232Th 0.73 0.56 0.20 40K 0.88 0.20 0.35 a Rotation converged in 8 itera- tions Fig. 4 Box plot for only radionuclides 1 3 920 Journal of Radioanalytical and Nuclear Chemistry (2021) 328:911–923 Radiological hazards indices Gamma activity concentration index Radium equivalent activity The gamma activity concentration index used to determine investigate level for practical monitoring of the building The radium equivalent activity R aeq is a weighted sum of materials was estimated using Eq. (8) proposed by European activity concentration of the 226Ra, 232Th and 40K radio- Commission [41]. nuclides for non-uniform distribution in the construction materials. It is based on the assumption that 370 Bq/kg of AcRa AcTh AcKI = + + 226 x (8) Ra, 259 Bq/kg of 232Th and 4,810 Bq/kg of 40K produce 300 200 3000 the same gamma ray dose rate [40]. It was estimated using where  A cRa, AcTh, AcK are the activity concentration the Eq. 7. (Bq/kg) of radium, thorium and potassium in the rocks Ra = Ac + 1.43Ac + 0.077Ac respectively.eq Ra Th K (7) The gamma activity concentration index (Ix) was found where ARa, ATh and AK are the activity concentration of to be ranged of 0.1–0.8 with mean value of 0.4 ± 0.3. The 226Ra, 232Th and 40K in Bq/kg respectively. The rocks for GGP and HWQ from Asutuare and Atomic recorded highest building must be rejected if the radium equivalent activity and lowest indexes respectively. The results obtained were exceeds 370 Bq/kg. This is to reduce radiation hazards asso- less than the proposed value of annual effective dose of 1 ciated with rocks. It can be seen from Table 6, that radium mSv [42]. equivalent levels estimated from the activity concentration of radionuclides were found to be varied 18–200 Bq/kg with mean of 108 ± 64 Bq/kg. The GGP and HWQ from Asutuare Internal hazards index and Atomic recorded highest and lowest values showed in Table 6. The result was found to be less than the proposed The internal hazard index (Hin) from rock was also calculated acceptable level of 370 Bq/kg in all the rocks [40]. from the activity concentration of radionuclides using the equation proposed by [43]. Table 6 Radiological hazards of the studied rocks Study location Sample code Sample no. Radiological hazards Raeq (Bq/kg) Gamma Index (Ix) Internal ( Hin) DR (nGy/h) Mean Mean Mean Mean Achimota FGS 3 108 ± 11 0.4 ± 0.04 0.4 ± 0.04 50 ± 5 Afienya HG 3 193 ± 12 0.7 ± 0.1 0.7 ± 0.1 92 ± 9 Anyarkode GHG 3 194 ± 13 0.8 ± 0.1 0.6 ± 0.1 95 ± 9 Asutuare GGP 3 200 ± 14 0.8 ± 0.1 0.7 ± 0.07 97 ± 9 Atomic HWQ 3 18 ± 1 0.1 ± 0.01 0.1 ± 0.01 8 ± 7 Atomic Hills MQ 3 108 ± 11 0.4 ± 0.04 0.4 ± 0.03 50 ± 5 Bukor Allkope GHG 3 196 ± 12 0.8 ± 0.1 0.6 ± 0.04 95 ± 9 Dodowa FFQ 4 98 ± 9 0.4 ± 0.02 0.3 ± 0.02 47 ± 5 Dominase MG 4 134 ± 12 0.5 ± 0.04 0.5 ± 0.03 64 ± 6 Haatso MQ 3 100 ± 9 0.4 ± 0.03 0.4 ± 0.03 47 ± 4 Kasoa BG 3 126 ± 11 0.5 ± 0.03 0.4 ± 0.03 61 ± 6 Mccarthy MQ 4 126 ± 15 0.5 ± 0.04 0.5 ± 0.03 58 ± 6 Ofankor HWQ 3 19 ± 2 0.1 ± 0.01 0.1 ± 0.01 9 ± 1 Oyarifa FFQ 3 73 ± 6 0.3 ± 0.02 0.2 ± 0.01 36 ± 3 Oyibi FFQ 3 69 ± 6 0.3 ± 0.02 0.2 ± 0.01 34 ± 3 Pokuase HWQ 3 20 ± 2 0.1 ± 0.01 0.1 ± 0.01 9 ± 1 Shai Hills HG 4 199 ± 15 0.7 ± 0.1 0.7 ± 0.1 94 ± 9 Tesano GPS 3 46 ± 6 0.2 ± 0.01 0.1 ± 0.01 23 ± 1 Weija MQ 4 108 ± 12 0.4 ± 0.03 0.2 ± 0.03 50 ± 5 1 3 Journal of Radioanalytical and Nuclear Chemistry (2021) 328:911–923 921 AcRa AcTh AcK Conclusions Hin = + + (9)185 259 4810 The measurement and statistical analysis of natural radionu- where  AcRa, AcTh, AcK are the activity concentration (Bq/ clides and rock properties in construction materials within kg) of radium, thorium and potassium in the rocks respec- Central and Greater Accra regions of Ghana has been stud- tively. The values of Hin must be less than unity to in order ied. The quartzites were found to contain lower levels of the to determine that rocks have negligible hazardous effects of radionuclides with highly weathered quartzite from Atomic radon and its short-lived progeny to the respiratory organs. having smallest activity concentrations for all the radionu- From the Table 6, the calculated values ranged from 0.1 to clides. High values of water absorption, durability and lower 0.7 with mean value of 0.4 ± 0.2. The GGP and HWQ from densities were also recorded in highly weathered quartzite. Asutuare and Atomic recorded highest and lowest indexes. The gneiss rocks from all the studied locations were to have The result was found to be less than annual effective dose good rock properties. This could be attributed to its min- of 1 mSv [44]. eralogical composition with interlocked grain boundaries. Hornblende gneiss and highly quartzite rocks recorded high- est and lowest radon exhalation rate levels. Highest positive Absorbed dose rate correlation coefficient between the radionuclides was found in 222Rn and 226Ra. The highest positive and negative corre- The absorbed dose rates (DR) due to gamma radiations in air lations coefficients between the radionuclides and rock prop- at 1 m above the ground surface for the uniform distribution erties were found in radon exhalation with apparent density of the radionuclides of 226Ra, 232Th and 40K in construc- and water absorption content. Radon was found to corre- tion materials was determined. It was calculated based on late well with the rock properties than 226Ra, 232Th and 40K. known activity concentration of 226Ra, 232Th and 40K and Three clusters and principal components were obtained in conversion factors of 0.462 nGy/h per Bq/kg for 226Ra, 0.604 rock properties and radionuclides. A single cluster and much nGy/h per Bq/kg for 232Th and 0.0417 nGy/h per Bq/kg for closely bonded was found between the 222Rn and 226Ra in the 40K [44]. It was computed using the equation proposed by studied rocks. The radon gas was found to be correlated very UNSCEAR, 2000. well with the rock properties than the natural radionuclides. 17% of the radon exhalation values found to be greater than D = 0.462Ac + 0.604Ac + 0.0417Ac R Ra Th K (10) the world average of 125.0 mBq/m2h. The calculated mean where  AcRa, AcTh, AcK are the activity concentration (Bq/ absorbed dose rate was less than the world average terrestrial kg) of radium, thorium and potassium in the rocks respec- tively. The absorbed dose rate was found to be varied in the range 8–97 nGy/h with mean value of 54 ± 31 nGy/h as presented in Table 6. The highest absorbed dose rate recorded in GGP from Asutuare was 1.8 times greater than the world average terrestrial value [44]. The lowest value in HWQ from Atomic was found to be 6.9 less than the world average value. The 24% of the studied rocks from all the locations were found to be greater than the world average of 55 nGy/h [44]. Comparing the calculated radiological hazards and rock properties The gneiss rock recorded higher values for all the radiologi- cal hazards indices, densities and lower levels for durability and water absorption contents as depicted in Fig. 5. Sand- stone on other hand obtained higher levels for durability and water absorption. The rock recorded radiological hazards indices greater than quartzite but less than the granite rocks as described in Fig. 5; Table 6. Quartzite was found to con- Fig. 5 Comparison of radiological hazards and rock properties in the tain lower values of radiological hazards than all the rocks. studied rocks 1 3 9 22 Journal of Radioanalytical and Nuclear Chemistry (2021) 328:911–923 value of 55 nGy/h. Radiological hazards indices associated In: ASTM Designation C-131-01. American Society for Testing with the concentrations of radium, thorium and potassium Materials, Philadelphia Philadelphia, USA. were all found to be less than world average values proposed 10. EC (2013) Test for Rock and Physical Properties of Aggregates-Part 6: Determination of Particle Density and Water Absorption, by ICRP, UNSCEAR and NEA-OECD. European Committee for Standardization, EN 1097-6, Brussels Generally, gneiss rocks were found to be contained good 11. EC (1999b) Tests for General Properties of Aggregates-Part 2: rock properties with relatively higher levels of natural radio- Methods for Reducing Laboratory Samples, European Committee activity than other rocks especially quartzites which possess for Standardization. EN 932-2 12. Marinos P, Hoek E, Marinos V (2006) Variability of the rock poor characteristics and lower levels of NORM. However, it properties of rock masses quantified by the geological strength is recommended that further work which may include expan- index: the case of ophiolites with special emphasis on tunnelling. sion of the study to cover other areas in order to have accu- Bull Eng Geol Environ 65/2:129–142 rate representative value for risk estimation for the general 1 3. Ndukauba E, Akaha CT (2012) Engineering-geological evaluation of rock materials from Bansara, Bamenda Massif Southeastern population in the country. Studying both rock properties and Nigeria, as aggregates for pavement construction. Geoscience NORM especially radon in rocks for constructions may be 2(5):107–111. https:// doi.o rg/ 10. 5923/j. geo. 20120 205. 01 better than radionuclides alone. Results of this study will 14. Sakoda A, Hanamoto K, Ishimori Y, Nagamatsu T, Yamaoka K be useful in selecting the quality rocks with lower activity (2008) Radioactivity and radon emanation fraction of the granites sampled at Misasa and Badgastein. Appl Radiat Isot 66:648–652 concentration of NORM for construction industries within 15. Schery SD, Gaeddert DH, Wilkening MH (1982) Transport of the study area in Ghana. Radon from fractured rock. J Geophys Res 87(B4):2969–2976 1 6. Strong KP, Levins DM (1982) Effect of moisture content on radon Acknowledgements The authors are very grateful to the staff of Cen- emanation from uranium ore and tailing. Health Phys 42:27–32 tre for Radiation Protection, Agency for Environmental Protection, 17. Darko EO, Faanu A, Awudu AR, Emi-Reynolds G, Yeboah J, Friuli Venezia Giulia, Udine, Italy and Central Materials Laboratory, Oppon OC, Ahaho HK (2009) Public exposure to hazards associ- Ghana Highway Authority, Accra, Ghana for their assistance and the ated with natural radioactivity in open-pit mining in Ghana. Rad use of their facilities for this study. This research was supported by the Prot Dosim 138(1):45–51 International Atomic Energy Agency (IAEA) and Radiation Protection 1 8. Faanu A, Adukpo OK, Tettey Larbi L, Lawluvi H, Kpeglo DO, Institute, Ghana Atomic Energy Commission in the form of fellowship Darko EO, Emi Reynolds G, Awudu RA, Kansaana C, Amoah P training. et al (2016) Natural radioactivity levels in soils, rocks and water at a mining concession of Perseus gold mine and surrounding towns in Central Region of Ghana. Springer Pls 5:98 19. Faanu A, Ephraim J, Darko EO (2010) Assessment of public expo- sure to naturally occurring radioactive materials from mining and References mineral processing activities of Tarkwa Goldmine in Ghana. Envi- ron Monit Assess 180(1–4):15–29 20. Otoo F, Darko EO, Garavaglia M, Giovani C, Pividore S, Andam 1. Morawska L, Phillips C (1993) Dependence of the radon emana- AB, Amoako JK, Adukpo OK, Inkoom S, Adu S (2018) Public tion coefficient on radium distribution and internal structure of the exposure to natural radioactivity and radon exhalation rate in con- material. Geochim Cosmochim Acta 57:1783–1797. https:// doi. struction materials used within Greater Accra Region of Ghana. org/ 10.1 016/ 0016-7 037(93) 90113-B Sci Afr 1:e00009 2. Brai M, Belli S, Hauser S, Puccio P, Rizzo S, Basile S, Marrale 2 1. Otoo F, Darko EO, Emi-Reynolds G, Adukpo OK, Andam AB M (2006) Correlation of radioactivity measurements, air kerma (2012) Preliminary Studies on 226Ra, 232Th and 40K concentra- rates and geological features of sicily. Radiat Meas 41:461–470 tions from quartzite rocks used as building materials. Radiat Prot 3. Mujahid SA, Hussain S, Dogar AH, Karim S (2005) Determina- Environ 34:119–128 tion of porosity of different materials by radon diffusion. Radiat 2 2. Otoo F, Arhin I, Darko EO (2020) Studies of radon levels, radium Meas 40:106–109 concentration and estimated effective dose in dwellings and soils 4. Sakoda A, Nishiyama Y, Hanamoto K, Ishimori Y, Yamamoto Y, in gold mining towns in Abirem of Eastern region of Ghana. Kataoka T, Kawabe A, Yamaoka K (2010) Differences of natu- Radiat Protect Dosim. https:// doi. org/ 10.1 093/r pd/ ncaa1 29 ral radioactivity and radon emanation fraction among constituent 2 3. Adam AMA, Eltayeb MAH (2012) Multivariate statistical analy- minerals of rock or soil. Appl Radiat Isot 68:1180–1184. https:// sis of radioactive variables in two phosphate ores from Sudan. J doi.o rg/ 10. 1016/j. aprad iso.2 009. 12. 036 Environ Radioact 107:23–43 5. Stranden E, Kolstad A, Lind B (1984) The influence of moisture 24. Szarlowicz K, Marcin Stobinski M, Hamerlik L, Peter Bitusik P and temperature on radon exhalation. Radiat Prot Dosim. https:// (2019) Origin and behavior of radionuclides in sediment core: a doi.o rg/1 0.1 093/r pd/7.1 -4.5 5 case study of the sediments collected from man-made reservoirs 6. Misdaq M, Ktata A, Bakhchi A (2000) A new method for study- located in the past mining region in Central Slovakia. Environ Sci ing the transport of radon and thoron in various building materi- Pollut Res 26(71 22 26):7115. 7115–7122 https://d oi.o rg/1 0.1 007/ als using CR–39 and LR–115 solid state nuclear track detectors. s11356- 019-0 4136-y Radiat Meas 32:35–42 25. Dickson KB, Benneh G (2004) A new geography of Ghana. Long- 7. Nazaroff WW (1992) Radon transport from soil to air. Rev Geo- mans Group Limited, London phys 30:137–160 26. Kesse GO (1985) The mineral and rock resources of Ghana. A. A 8. Schuiling RD, De Meijer RJ, Riezebos HJ, Scholten MJ (1985) Balkema, Rotterdam Grain size distribution ofdifferent mineral in a sediment as a func- 27. Ngachin M, Garavaglia M, Giovani C, Kwato Njock MG, tion of their specific density. Geol Mijnbouw 64:199–203 Nourreddine A (2007) Assessment of natural radioactivity and 9. ASTM (2006) Resistance to degradation of small-size coarse associated radiation hazards in some Cameroonian building mate- aggregate by abrasion and impact in the Los Angeles Machine. rials. Radiat Meas 42:61–67 1 3 Journal of Radioanalytical and Nuclear Chemistry (2021) 328:911–923 923 28. Eckert & Ziegler (2011) Nuclide Mixtures NG1 to NG6, Braun- 3 8. Jackson JE (1991) A user’s guide to principal components. Wiley, schweig Germany New York 29. Sima O, Arnold D, Dovlete CGESPECOR (2001) A versatile tool 3 9. Murtagh F, Heck A, (1987) Multivariate data analysis, Kluwer in gamma-ray spectrometry. J Radio Nucl Chem 248:359–364 Murtagh and Heck, 1987. Multivariate data analysis, Kluwer, 30. Durrani SA, Ilic R (1997) Radon measurements by etched track Dordrecht. http://s trule.c s.q ub.a c.u k/~ fmurta gh/b ooks/h oldal l.p df. detectors. World Scientific Publishing Co, Singapore 40. Beretka I, Mathew PI (1985) Natural radioactivity of Australian 3 1. Nazaroff WW, Nero AV (1988) Radon and its decay products in building materials, waste and byproducts. Health Phys 48:87–95 indoor air. Willey, New York 4 1. EC (1999a) Radiological protection principles concerning the 3 2. Joshua EO, Ademola JA, Akpanowo MA, Oyebanjo OA, Olorode natural radioactivity of building materials. Directorate-General DO (2009) Natural radionuclides and hazards of rock samples Environment. Nuclear Safety and Civil Protection 112 collected from Southeastern Nigeria. Radiat Meas 44:401–404 4 2. ICRP. The 2007 Recommendations of the International Commis- 3 3. Mehra R, Singh S, Singh K (2006) A study of uranium, sion on Radiological Protection (2007) Annals on Recommenda- radium, radon exhalation rate and indoor radon in the environs tions of the ICRP 37:2–4. https://w ww.i crp.o rg/d ocs/I CRP_P ubli of some areas of Malwa region, Punjab. Indoor Built Environ cation_1 03 15(5):499–505 4 3. Krieger R (1981) Radioactivity of construction materials. Beton- 34. UNSCEAR (1988) United Nations Scientific Committee on the werk Fertigteil-Technik 47:468–473 effects of atomic radiation sources and effects of ionizing radia- 4 4. UNSCEAR (2000) United Nations Scientific Committee on the tion. Sources and effect of ionizing radiation. UNSCEAR, New effects of atomic radiation sources and effects of ionizing radia- York tion. Sources and effect of ionizing radiation. UNSCEAR, New 35. Lee SC, Kim CK, Lee CK, Kang HD (2001) Natural radionuclides York contents and radon exhalation rates in building materials used in South Korea. Radiat Prot Dosim 94:269–274 Publisher’s note Springer Nature remains neutral with regard to 3 6. Kathren RL (1998) NORM Sources and their origins. Appl Radiat jurisdictional claims in published maps and institutional affiliations. Isot (Incorporating Nuclear Geophysics) 49(3):149–168 37. Abumurad KM, Al-Tamimi MH (2001) Emanation power of radon and its concentration in soil and rocks. Radiat Meas 34:423–442 1 3