High Specific Activity of Radium Isotopes in Baryte from the Czech Part of the Upper Silesian Basin—An Example of Spontaneous Mine Water Treatment

Radium-bearing barytes (radiobarytes) have been known since the beginning of the 20th century. They are mainly found as precipitates of low-temperature hydrothermal solutions. In anthropogenic environments, they frequently occur as crusts on oil industry equipment used for borehole extraction, in leachates from uranium mill tailings, and as a by-product of phosphoric acid manufacturing. Recently, we recognized Ra-rich baryte as a precipitate in the water drainage system of a bituminous coal mine in the Czech part of the Upper Silesian Basin. The precipitate is a relatively pure baryte, with the empirical formula (Ba0.934Sr0.058Ca0.051Mg0.003)Σ1.046S0.985O4.000. The mean specific activity of 226Ra was investigated by the two-sample method and it equals 39.62(22) Bq/g, a level that exceeds known natural occurrences. The values for 228Ra and 224Ra are 23.39(26) Bq/g and 11.03(25) Bq/g. The radium content in the baryte is 1.071 ng/g. It is clear that the Ra-rich baryte results from the mixing of two different mine waters—brines rich in Ba, Sr, and isotopes 226Ra and 228Ra and waters that are affected by sulfide weathering in mine works. When this mixing occurs in surface watercourses, it could present a serious problem due to the half-life of 226Ra, which is 1600 years. If such mixing spontaneously happens in a mine, then the environmental risks will be much lower and will be, to a great, extent eliminated after the closure of the mine.


Introduction
The radium isotope 226 Ra is a daughter product of 230 Th alpha decay in the 238 U decay series, which is also known as the uranium or radium series. The radium isotopes 228 Ra and 224 Ra are the daughter products of 228 Ac, resp. 228 Th decay in the 232 Th decay series, also known as the thorium series [1]. Radium is enriched in some naturally occurring radioactive materials (NORMs), but even higher concentrations can be found in some technologically enhanced naturally occurring radioactive materials (TENORMs).
In the anthropogenic environment, the most common concentrator of radium, by far, is radium-rich baryte, followed by radium-rich gypsum [17]. These materials frequently occur as crusts on oil industry equipment used for borehole extraction [18][19][20][21][22], in leachates from uranium mill tailings [23][24][25][26], and as by-products of phosphoric acid manufacturing [27][28][29][30]. The radium content of these anthropogenic materials is much higher than that of natural phases and it can reach 1 × 10 3 Bq/g in the samples from the first source mentioned.
In both NORMs and TENORMs, the formation of Ra-rich baryte causes the removal of radium dissolved in water [31,32]. Experimental investigations of the formation of solid solutions between radium and baryte have proved that this process happens at fast kinetic rates [33,34] and, therefore, available radium can be fixed if enough Ba 2+ and SO 4 2are available in solution.
The crusts of unknown minerals were found during a review of the water drainage system at thě CSA Mine, Doubrava Shaft ( Figure 1), which extracts bituminous coal from the Karviná Formation of the Upper Silesian Basin, Czech Republic. The crusts were highly radioactive, emitting particles by alpha decay. Baryte was proven to be a major component of this incrustation, with radium content being responsible for the radioactivity. The aim of this paper is to characterize the genesis of this material, its composition, and the specific activity of the radium isotopes. The results will help to elucidate the behaviour and possible hazards of this potentially hazardous TENORM. Among natural occurrences, the most common are trace amounts of radium in aquifers [2,3]. Increased concentrations have been reported in mineral and thermal waters [4][5][6][7][8][9][10] and in brines [11][12][13][14]. Among natural mineral phases, radium-rich baryte (also known as radian baryte or radiobaryte) is baryte that contains trace amounts of radium. The natural occurrences of radiobaryte were first noticed at the beginning of the 20th century. Such occurrences are connected with low-temperature hydrothermal fluorite-baryte mineralization [15,16], and the radioactivity can reach 8 Bq/g.
In the anthropogenic environment, the most common concentrator of radium, by far, is radiumrich baryte, followed by radium-rich gypsum [17]. These materials frequently occur as crusts on oil industry equipment used for borehole extraction [18][19][20][21][22], in leachates from uranium mill tailings [23][24][25][26], and as by-products of phosphoric acid manufacturing [27][28][29][30]. The radium content of these anthropogenic materials is much higher than that of natural phases and it can reach 1 × 10 3 Bq/g in the samples from the first source mentioned.
In both NORMs and TENORMs, the formation of Ra-rich baryte causes the removal of radium dissolved in water [31,32]. Experimental investigations of the formation of solid solutions between radium and baryte have proved that this process happens at fast kinetic rates [33,34] and, therefore, available radium can be fixed if enough Ba 2+ and SO4 2-are available in solution.
The crusts of unknown minerals were found during a review of the water drainage system at the ČSA Mine, Doubrava Shaft ( Figure 1), which extracts bituminous coal from the Karviná Formation of the Upper Silesian Basin, Czech Republic. The crusts were highly radioactive, emitting particles by alpha decay. Baryte was proven to be a major component of this incrustation, with radium content being responsible for the radioactivity. The aim of this paper is to characterize the genesis of this material, its composition, and the specific activity of the radium isotopes. The results will help to elucidate the behaviour and possible hazards of this potentially hazardous TENORM.

The Upper Silesian Basin
The Upper Silesian Basin belongs to the eastern domain of the Central European Variscides. It is a typical foreland basin [35]. The post-erosional boundary of the basin has a roughly triangular shape

The Upper Silesian Basin
The Upper Silesian Basin belongs to the eastern domain of the Central European Variscides. It is a typical foreland basin [35]. The post-erosional boundary of the basin has a roughly triangular shape that extends from Poland southward into the Czech Republic. The area of this important European bituminous coal basin exceeds 7400 km 2 [36].
The sedimentary sequence that fills the Czech part of the basin has been discussed by, e.g., [37,38]. Cyclic alternation of clastic sediments with coalbeds is typical for the whole sedimentary record [39]. Intercalations of volcanoclastic material are frequent [40]. The older type of sedimentation from the Serpukhovian stage contains marine horizons [39], while younger formations that range from the Bashkirian to early Kasimovian stage originate purely in terrestrial settings [41].
The basement of the basin is formed by the sedimentary cover of the Brunovistulicum and it is specifically formed by sediments of the Cambrian, Ordovician, Devonian, and Mississippian [42,43]. The basin fill is overlain primarily by Triassic, Neogene, and Quaternary deposits and Permian and Jurassic sedimentary sequences in the Polish part. The southern (Czech) part of the basin is overlain by Neogene deposits of the Carpathian Foredeep and it is overlain by nappes of the Outer Carpathians (Jurassic to Paleogene) further to the south [37].

Geochemical and Hydrogeochemical Background
While considering that uranium is a primary source of radium, it is necessary to mention that uranium minerals are not known in the Upper Silesian Basin. Local coal contains up to 44 ppm U and 24 ppm Th in coal ashes [44], which is not higher dramatically than the Clarke value for world bituminous coals-15 ppm for U and 23 ppm for Th [45]. Whether the uranium is bound to the organic or inorganic material in the coal is not yet known with certainty. The U and Th contents of sediments [46,47] and tuffs [47] have also been published. In such samples, both of the elements are present as the isomorphic admixtures, mostly in zircon and minerals of the monazite and apatite groups, U is probably also sorbed in clay minerals and/or organic matter.
The hydrogeological and hydrogeochemical conditions in the Czech part of the Upper Silesian Basin are known to be complex. The basin contains nine water-bearing systems at different stratigraphic levels [48] that differ in the contents of dissolved gases and elements.
The geochemistry and origin of the water have been studied, such as in a study by [49]. Four of the water-bearing systems that they noted are of interest for the purposes of this study. First, the systems are connate waters of the Miocene sea formation. Their total mineralization varies, with values up to 150 g/L. The systems are virtually sulfate and uranium free, and the Ba 2+ content does not exceed 40 mg/L. Only one value for the 226 Ra content is available-33 pg/L. SO 4 -free brines of a pre-Tortonian hot climate recharge are also characterized by high mineralization, variable Ba and Sr content, and 226 Ra contents of up to 3100 pg/L. The oldest paleoinfiltration brines, assumed to be of Permian age and it could have a salinity of up to 230 g/L. They do not contain U and SO 4 2-, but they are extremely rich in Ba 2+ (up to 1800 mg/L), Sr 2+ (up to 450 mg/L), and 226 Ra (up to 6000 pg/L). Finally, mixed waters are waters affected by the mixing of any other types, mainly due to coal mining activities [49].

Materials and Methods
The piece of old water pipe in question came from the 10th floor of the Doubrava Shaft. It was used for pumping waste water to the main pumping station in the by-pass of the 10th floor. The investigated material forms a beige precipitate crust ("scale") that is several centimetres thick. It is porous and shows indications of a botryoidal surface ( Figure 2). X-ray powder diffraction measurements were conducted while using a Bruker-AXS D8 Advance instrument (Institute of Geological Engineering, VŠB-Technical University of Ostrava, operator D. Matýsek) with 2θ/θ geometry and measured using a LynxEye position sensitive detector under the following conditions: radiation: CoKα/Fe filter, voltage: 40 kV, current: 40 mA, step-by-step mode of 0.014 • 2θ with an interval of 0.25 s per step, and the summation of at least five successive measurements based on the complexity of the recording. The samples were prepared by pulverisation in an agate mortar and transfer onto a low-background silicon holder. The data were digitally processed while using Bruker Diffrac Suite software (Bruker Corporation, Billerica, MA, USA). The Rietveld method in Bruker Topas version 4.2 was used to verify the accuracy of the qualitative evaluation of the measurements.
Wavelength-dispersive X-ray spectroscopic analyses (WDS) were performed at the Laboratory of Electron Microscopy and Microanalysis at the Faculty of Science at Masaryk University in Brno on an electron microprobe (model CAMECA SX100 and analyst R.Čopjaková). The standards used were leucite (Rb Lα), SrSO 4   Two samples of the Ra-rich baryte powder were prepared and then analysed for their 226 Ra content. The first sample, containing m = 4.47(2) g of the Ra-rich baryte powder, was mixed with 1.90(1) mL of 226 Ra water solution of known activity, A0 = 2712(24) Bq, provided by Eurostandard, Czech Republic. The second sample, which contained the same amount (4.47 g) of the Ba powder, was mixed with 1.90(1) mL of distilled water. Both of the samples were sealed in 5 mL polyethylene containers while using a special wax to prevent 222 Rn leakage and left for 28 days to establish secular equilibrium between 226 Ra and its decay products 222 Rn, 214 Pb, and 214 Bi. The gamma spectra of the samples were then measured using a 30% HPGe spectrometer (GC3018, Canberra) equipped with 10 cm lead + 1 mm cadmium + 1 mm copper shielding to reduce the background. The spectra were analysed while using Genie2000 software (Mirion Technologies, Canberra, Australia).
The two-sample method enables the determination of 226 Ra activity without the need to calibrate the gamma spectrometer. There is also no need to make corrections for gamma self-absorption in the samples. We only assume that the 226 where R1 and R2 are the detected gamma rates for the first and the second samples, respectively, RB is the background gamma rate, and is the detection efficiency. The unknown 226 Ra activity can then be calculated from: Two samples of the Ra-rich baryte powder were prepared and then analysed for their 226 Ra content. The first sample, containing m = 4.47(2) g of the Ra-rich baryte powder, was mixed with 1.90(1) mL of 226 Ra water solution of known activity, A 0 = 2712(24) Bq, provided by Eurostandard, Czech Republic. The second sample, which contained the same amount (4.47 g) of the Ba powder, was mixed with 1.90(1) mL of distilled water. Both of the samples were sealed in 5 mL polyethylene containers while using a special wax to prevent 222 Rn leakage and left for 28 days to establish secular equilibrium between 226 Ra and its decay products 222 Rn, 214 Pb, and 214 Bi. The gamma spectra of the samples were then measured using a 30% HPGe spectrometer (GC3018, Canberra) equipped with 10 cm lead + 1 mm cadmium + 1 mm copper shielding to reduce the background. The spectra were analysed while using Genie2000 software (Mirion Technologies, Canberra, Australia).
The two-sample method enables the determination of 226 Ra activity without the need to calibrate the gamma spectrometer. There is also no need to make corrections for gamma self-absorption in the samples. We only assume that the 226 where R 1 and R 2 are the detected gamma rates for the first and the second samples, respectively, R B is the background gamma rate, and is the detection efficiency. The unknown 226 Ra activity can then be calculated from: The specific activity A m is then obtained from A m = A/m. The standard uncertainties of the measured quantities are shown in parentheses, and the standard uncertainties of the calculated quantities are obtained from the Gaussian law of uncertainty propagation.
The analyses of water were performed in the accredited labs of Labtech Company. The acid-base reaction was measured with an Orion pH meter (Model 710A), and the total mineralization of the individual samples of water was determined by the gravimetric method. The contents of major anions (Cland SO 4 2-) were established by ion chromatography while using an IC 90 chromatograph that was manufactured by Dionex Co (Sunnyvale, CA, USA). Iodides and bromides were determined by redox titration, and the concentrations of cations and heavy metals (Ca, Mg, Na, K, Li, Al, Fe, Ba, and Sr) were determined by the ICP-OES method while using an Jobin Activa spectrometer. The contents of the CO 2 fixation forms were calculated from their neutralizing capacities while using an acid-base titration method.

Results
Powder X-ray diffraction revealed that baryte is the prevailing constituent of the precipitate (scale). The precipitate also contains traces of kaolinite and quartz and also possibly an amorphous phase containing iron. The cell parameters calculated using the Rietveld method correspond with data given in the literature for substituted baryte ( Table 1). The difference in the "a" axis can be ascribed to the substitution of Sr into the Ba positions of the crystal lattice of baryte, which causes the shrinking of the "a" distance of the unit cell [51], and possibly also to substitution of Ca into the same position [52].  Table 2 gives the chemical composition of studied precipitate, since the pure baryte spots could not be measured to their thickness ( Figure 3) and porosity of the aggregate. This factor also affected the totals of the measurement. We ascribed content of K, Fe, P, Si, Al, and Cl to detected minor amounts of kaolinite, quartz, and amorphous Fe-bearing phase. Empirical formula (Ba 0.934 Sr 0.058 Ca 0.051 Mg 0.003 ) Σ1.046 S 0.985 O 4.000 given as the mean of seven determinations (Table 2) correspond well with the ideal baryte. The major isomorphic admixtures are Sr and Ca (both up to 0.06 apfu), and there is also minor content of Mg (up to 0.01 apfu).   The two-sample method described in Section 3 enabled us to determine the 226 Ra activity and specific activity for the five most intensive gamma lines from the 226 Ra decay chain ( Table 3). The weighted arithmetic mean of the specific activity of 226 Ra calculated from all five detected gamma lines equals 39.62(22) Bq/g (Table 3). If we consider an older non-SI unit of radioactivity, the curie (Ci), which corresponds to 3.7 × 10 10 decays per second of 1 g 226 Ra [55], then the content of radium 226 Ra in the baryte is 1.071 ng/g. Table 3. Detected gamma rates R1, R2, and RB (first and second sample and background, respectively) for the five most intensive gamma lines and the calculated 226 Ra activities A and specific activities Am. Combined standard uncertainties are shown in parentheses. The known activity of 226 Ra in the two samples was used for the efficiency calibration of the HPGe spectrometer that enabled us to determine the activity of the other two radium isotopes from the 232 Th series that were identified in the samples, namely 228 Ra and 224 Ra ( Table 4). The weighted arithmetic means of the specific activity of 228 Ra calculated from the two detected gamma lines (from The two-sample method described in Section 3 enabled us to determine the 226 Ra activity and specific activity for the five most intensive gamma lines from the 226 Ra decay chain ( Table 3). The weighted arithmetic mean of the specific activity of 226 Ra calculated from all five detected gamma lines equals 39.62(22) Bq/g (Table 3). If we consider an older non-SI unit of radioactivity, the curie (Ci), which corresponds to 3.7 × 10 10 decays per second of 1 g 226 Ra [55], then the content of radium 226 Ra in the baryte is 1.071 ng/g. The known activity of 226 Ra in the two samples was used for the efficiency calibration of the HPGe spectrometer that enabled us to determine the activity of the other two radium isotopes from the 232 Th series that were identified in the samples, namely 228 Ra and 224 Ra ( Table 4). The weighted arithmetic means of the specific activity of 228 Ra calculated from the two detected gamma lines (from the 228 Ac decay) equals 23.39(26) Bq/g. For 224 Ra (from 212 Pb and 212 Bi decays), we get 11.03(25) Bq/g. The ratios of the specific activities of 226 Ra and 228 Ra, and 224 Ra and 228 Ra then equal 1.694 (21) and 0.472 (12), respectively. It should be noted that the latter ratio is identical to the ratio of specific activities of 228 Th and 228 Ra. The analysed mine water samples (P2307 and P41 in Table 5) indicate Miocene brines. Typically, the SO 4 2− concentrations are below the detection limit. However, Br/I is much higher than the range that was published by [49], so they may, in fact, belong to the group of SO 4 -free brines of a pre-Tortonian hot climate recharge. Samples P2111, P1804, and P4505 in Table 4 represent mixed brines from mine works with sulfate efflorescences. An increased SO 4 2content is typical for them. Depletion in Ba and

Nuclide
Sr is an effect of the precipitation of baryte.

Discussion
The examples of the mine water chemical composition (Table 5) reveal the story behind the radium-rich baryte. Original Miocene brines are rich in Ba, Sr, and Ra, but are virtually SO 4 free.
Radium, especially 226 Ra and 228 Ra, is generated by the alpha decay of 238 U, respectively 232 Th from accessory minerals (zircon, monazite, and apatite). Alpha recoil is the process that concentrates 226 Ra and 228 Ra in the brines [56]. Rock massif is oxygen deficient, and water can also contain free CH 4 from natural coal degassing, so it does not cause sulfide decomposition. Once the water reaches the mine works, which have different geochemical conditions, it combines with bacterial activity to participate in sulfide decomposition in both the coal and surrounding rocks [57]. Typically, the SO 4 2content is increased. Ba reacts with sulfate anions and precipitates as baryte once such waters of divergent composition are mixed. Strontium and radium are hence incorporated into its crystal lattice as isomorphic admixtures due to their similar ionic radius and coordination number.
Similar findings have also been reported from the Polish part of the Upper Silesian Basin. Radium-rich mine waters are present, according to [49], and they pose a possible hazard to the environment, since ca. one-half of the radioactive nuclides (mostly radium) flowing into the mines are discharged to surface waters [58]. There, the radioactive nuclides become part of the sediment, fixed in baryte, witherite (BaCO 3 ), or metal oxyhydroxides [59]. However, we were able to document and describe what happens to radionuclides that do not leave the underground mines. Radium is at least partly fixed in baryte, and, with some exceptions, such as the possible recycling of the steel parts of water drainage and pumping systems, the baryte does not enter surface ecosystems, due to spontaneous water mixing. Baryte is considered a relatively stable mineral [60,61] and it should not release radium after the closure of mines and their flooding. Therefore, the spontaneous precipitation of Ra-rich baryte is the natural analogue of the treatment method tested by [62] for radium removal from mine waters. However, radium that is fixed in baryte could be released by activities of sulfate-reducing bacteria [63], reported from both laboratory [64][65][66] and field [66] environments.

Conclusions
Radium is a significant environmental pollutant. The isotope 226 Ra is the most significant, while considering its long half-life (1600 years), radioactivity, toxicity, and ability to be incorporated into animal and human bones due to its geochemical similarity with calcium [67]. Bituminous coal mining in the Upper Silesian Basin presents a substantial source of Ra. This element could enter terrestrial aquatic and atmospheric environments due to the pumping of radium-rich brines. The spontaneous fixation of 226 Ra in Ra-rich baryte due to the mixing of waters of different chemical nature is beneficial if it occurs underground, because the radium is fixed far from possible entry points to biogenic cycles or contact with living organisms. Experiments have shown that similar processes, i.e., blending Ra-bearing fluids with acid mine drainage (AMD) waters, could be used to remove radium and barium from hydraulic fracturing fluids [68] or mine waters [62]. However, the evaluation of radiation hazards for miners would require a separate study, since there are many factors that affect the precise place of baryte precipitation, presence of miners at those locations, and handling with the material containing precipitate.