The Rožná uranium deposit (Bohemian Massif, Czech Republic): shear zone-hosted, late Variscan and post-Variscan hydrothermal mineralization
- First Online:
- Cite this article as:
- Kříbek, B., Žák, K., Dobeš, P. et al. Miner Deposita (2009) 44: 99. doi:10.1007/s00126-008-0188-0
- 325 Views
Three major mineralization events are recorded at the Rožná uranium deposit (total mine production of 23,000 t U, average grade of 0.24% U): (1) pre-uranium quartz-sulfide and carbonate-sulfide mineralization, (2) uranium, and (3) post-uranium quartz-carbonate-sulfide mineralization. (1) K–Ar ages for white mica from wall rock alteration of the pre-uranium mineralization style range from 304.5 ± 5.8 to 307.6 ± 6.0 Ma coinciding with the post-orogenic exhumation of the Moldanubian orogenic root and retrograde-metamorphic equilibration of the high-grade metamorphic host rocks. The fluid inclusion record consists of low-salinity aqueous inclusions, together with H2O-CO2-CH4, CO2-CH4, and pure CH4 inclusions. The fluid inclusion, paragenetic, and isotope data suggest that the pre-uranium mineralization formed from a reduced low-salinity aqueous fluid at temperatures close to 300°C. (2) The uraniferous hydrothermal event is subdivided into the pre-ore, ore, and post-ore substages. K–Ar ages of pre-ore authigenic K-feldspar range from 296.3 ± 7.5 to 281.0 ± 5.4 Ma and coincide with the transcurrent reorganization of crustal blocks of the Bohemian Massif and with Late Stephanian to Early Permian rifting. Massive hematitization, albitization, and desilicification of the pre-ore altered rocks indicate an influx of oxidized basinal fluids to the crystalline rocks of the Moldanubian domain. The wide range of salinities of fluid inclusions is interpreted as a result of the large-scale mixing of basinal brines with meteoric water. The cationic composition of these fluids indicates extensive interaction with crystalline rocks. Chlorite thermometry yielded temperatures of 260°C to 310°C. During this substage, uranium was probably leached from the Moldanubian crystalline rocks. The hydrothermal alteration of the ore substage followed, or partly overlapped in time, the pre-ore substage alteration. K–Ar ages of illite from ore substage alteration range from 277.2 ± 5.5 to 264.0 ± 4.3 Ma and roughly correspond with the results of chemical U–Pb dating of authigenic monazite (268 ± 50 Ma). The uranium ore deposition was accompanied by large-scale decomposition of biotite and pre-ore chlorite to Fe-rich illite and iron hydrooxides. Therefore, it is proposed that the deposition of uranium ore was mostly in response to the reduction of the ore-bearing fluid by interaction with ferrous iron-bearing silicates (biotite and pre-ore chlorite). The Th data on primary, mostly aqueous, inclusions trapped in carbonates of the ore substage range between 152°C and 174°C and total salinity ranges over a relatively wide interval of 3.1 to 23.1 wt% NaCl eq. Gradual reduction of the fluid system during the post-ore substage is manifested by the appearance of a new generation of authigenic chlorite and pyrite. Chlorite thermometry yielded temperatures of 150°C to 170°C. Solid bitumens that post-date uranium mineralization indicate radiolytic polymerization of gaseous and liquid hydrocarbons and their derivatives. The origin of the organic compounds can be related to the diagenetic and catagenetic transformation of organic matter in Upper Stephanian and Permian sediments. (3) K–Ar ages on illite from post-uranium quartz-carbonate-sulfide mineralization range from 233.7 ± 4.7 to 227.5 ± 4.6 Ma and are consistent with the early Tethys-Central Atlantic rifting and tectonic reactivation of the Variscan structures of the Bohemian Massif. A minor part of the late Variscan uranium mineralization was remobilized during this hydrothermal event.
KeywordsUraniumBohemian MassifRožnáCzech Republic
Numerous unpublished studies on the geology, mineralogy, and geochemistry of Rožná were summarized by Arapov et al. (1984) and Vilhelm et al. (1984). The uranium mineralization was dated at 280–260 Ma (U–Pb, uraninite; Anderson et al. 1988). Fluid inclusions in host rocks and ores were studied by Vosteen and Weinoldt (1997), Dobeš et al. (2001), and Hein et al. (2002). The isotope composition of carbonates and sulfides was determined by Žák et al. (2001). Selenide mineralization that accompanies uranium mineralization at Rožná was studied by Johan and colleagues (1971, 1976, 1978), and the rare earth element, yttrium and zircon mobility associated with uranium mineralization is discussed in René (2008). During the last 15 years, a detailed structural, petrological, mineralogical, and geochemical investigation of the deposit together with stable isotope, fluid inclusion, cathodoluminescence, and organic-geochemical research were performed to characterize structural, geochemical, and time controls on ore deposition. The results of this project, including primary data, were summarized by Kříbek and Hájek (2005), and a short version is presented in this paper. The aim of this comprehensive study is to characterize the sources of the uranium-bearing fluids and the p–T conditions of the successive fluid events during the post-orogenic evolution of the Bohemian Massif. The data allow to propose a descriptive model of the shear zone-hosted uranium mineralization and to relate its origin to the late Variscan and post-Variscan geological history of the Bohemian Massif.
Geological outline of the eastern part of the Bohemian Massif
Geology of the Rožná uranium deposit
Longitudinal fault structures hosting uranium mineralization are crosscut and segmented by steep, ductile to brittle NW–SE- and SW–NE-striking fault zones that host post-uranium carbonate-quartz-sulfide mineralization.
Ore bodies and macroscopic types of uranium mineralization
- (Ad 1)
The main ore-bearing structures of Rožná are represented by fault zones R1 and R4 (Figs. 3 and 4). The zones comprise massive cataclasite 4–15 m thick or several subparallel cataclasite bodies 1–3 m thick, traceable over the horizontal distance of 15 km. The zones strike 340–355° and dip west at an angle of 45–70°. The ore is composed of uraninite, coffinite, (U,Zr)-silicates, and minor montroseite disseminated in chloritized, hematitized, limonitized, argillized, and pyritized coherent and incoherent cataclasites and fault breccias (Fig. 5a). Ore lenses are up to 90,000 m2 large and are 3.5 m thick on average. The grade is around 0.5% U, up to 10% U locally.
- (Ad 2)
Ore zones R2 and R3 and carbonate veins represent pinnate structures genetically related to the master faults of the Rožná shear zone. These structures are traceable over a distance of 4–5 km and dip west at an angle of 55–90°. Compared to the master faults R1 and R4, they host a large number of carbonate veins up to 2 m thick with predominantly uraninite mineralization. Ore bodies reach 27,000 m2 in size and have 2 m in average thickness. The average grade is 0.6% U.
- (Ad 3)
Ore bodies in desilicified, albitized, and hematitized, commonly highly porous rocks occur adjacent to the individual brittle structures (Fig. 5b), and their shape usually follows the original metamorphic foliation of the host rocks. Disseminated coffinite is accompanied by less common uraninite and U-Zr-silicate. The ore bodies reach a maximum size of 30,000 m2 with a grade of 0.10–0.15% U, exceptionally 0.3% U.
- (Ad 4)
Disseminated uranium mineralization bound to oblique fault zones is usually hosted by quartz-carbonate-sulfide breccias at intersections with diagonal and longitudinal structures. Compared to other types of mineralization, the ore bodies are small (1,000–5,000 m2) but contain relatively high-grade ore (average grade 0.8% U, maximum 20% U in some ore shoots).
Materials and methods
Ore samples were collected from different levels and stopes of the Rožná I and Rožná II mines, mainly from levels 18–24 at depths of 900–1,200 m below surface or from the rock and mineral collection of both mines.
Thin sections and polished sections of rocks and ores were studied by polarized-light microscopy. Cathodoluminescence microscopy was performed on an optical microscope equipped with hot cathode HC2/LM (Simon-Neuser, Bochum, Germany).
Chemical compositions of rock-forming minerals were determined using the CamScan 3200 LINK ISIS microprobe (Czech Geological Survey, Prague), JEOL JXA-50 microprobe (Institute of Geology, Academy of Sciences of the Czech Republic), CAMECA SX-50 microprobe (Institute of Mineralogy, University of Hannover), and MICROSPEC 2A electron microscope (Department of Mineralogy and Petrology, Masaryk University, Brno). Uranium mineralization was studied using the CamScan 4DV-LINK AN 10000 microprobe at the State Glass Institute, Hradec Králové.
XRD characteristics of clay minerals and their relative proportions in clay- and silt-size fractions were determined using a Philips PW 7310 diffractometer with CuKα radiation and Ni filter standard set. Samples for XRD were prepared by hand crushing and rock disintegration in demineralized water using a sonic dismembrator for approximately 10–15 min. Clay (<2 μm) and fine silt (2–5 μm) fractions were then separated by sedimentation and centrifugation.
The 57Fe Mössbauer spectra of the powderized chlorite and biotite samples were measured using 57Co in Rh source in transmission geometry at room temperature. The velocity scale was calibrated against the spectrum of pure α-Fe foil. The analyses were carried out at the Institute of Material Physics, Brno.
Fluid inclusions were measured in siderite, calcite, quartz, and sphalerite of various paragenetic stages using a Chaixmeca heating–freezing stage (Poty et al. 1976) and a LINKAM THM combined heating- and freezing-stage, both under standard conditions. Salinity was determined after Bodnar and Vityk (1995), composition of the aqueous phase was evaluated after Borisenko (1977). P–V–T–X characteristics of H2O–CO2–CH4 inclusions were evaluated after Bakker and Diamond (2000) and Van den Kerkhof and Thiery (2001).
Minerals for stable isotope determinations were separated by hand under the binocular microscope. SO2 gas for sulfur isotope measurement was prepared from sulfides after Grinenko (1962). CO2 gas from carbonates was prepared according to McCrea (1950). Quartz and silicate for oxygen isotope analyses were processed following the method of Clayton and Mayeda (1963). Measurement of carbon, oxygen, and sulfur isotope composition was done on a Finnigan MAT 251 IRMS at the Czech Geological Survey. The general reproducibility of δ13C and δ18O values of carbonates and δ34S values of sulfides was ±0.1‰, and that of δ18O values of quartz and silicates was ±0.2‰.
Samples for deuterium isotope (δD) determination were preheated in a ceramic oven to remove surface water and then heated in vacuo at 750°C. The released water was reduced over hot uranium metal and the resulting H2 gas was collected on charcoal at −196°C. Isotopic analyses were performed on a Finnigan MAT 251 gas isotope ratio mass spectrometer. Data were corrected for 3H+ production in the ion source and reported in the standard delta notation as per mil deviation from V-SMOW. External reproducibility is ±2.3‰ (1σ) based on repeat analyses of the Cambrian Formation Water Standard (CFW). The determinations were carried out by Actlabs Laboratories, Canada.
The extractable organic matter (EOM) of solid bitumen was extracted using a Soxhlet HT2 instrument (Tecator Company) and analyzed by gas chromatography–mass spectrometry. The extract was analyzed using a HP 5MS column (length 30 m, interior diameter 0.25 μm) in a Hewlett-Packard HP 5890 gas chromatogram with a flame ionization detector. The mass spectrometer (Hewlett-Packard 5985) source was at 220°C, using electron impact (70 eV) ionization, at 0.2 s per decade.
K–Ar age data were obtained from the <1-, <3-, and <5-μm separates of illite/white mica-rich samples. K–Ar dating was performed by Actlabs Laboratories, Canada. Data were stored and processed by a Hewlett-Packard MX-21-E data system. Details of the analytical procedures are given in Kapusta et al. (1997).
Chemical U–Pb dating on monazite was done using a JEOL JX 8600 microprobe equipped with four WDX spectrometers (Institute of Mineralogy, Salzburg University) and a CAMECA SX 100 microprobe (Dionýz Štúr Institute of Geology, Bratislava) using the method of Montel et al. (1996) and data processing after Ludwig (1980). The analytical error ranged between 0.04–0.05 wt% for Th, 0.025–0.030 wt% for U, and 0.010 wt% for Pb. Monazite samples with known concordant U–Pb ages (TIMS) were used as standards.
Stages and types of vein-style and disseminated mineralization
Pre-uranium stage mineralization
The disseminated or veinlet-type quartz-sulfide mineralization of the pre-uranium stage is characterized by polygonal quartz, pyrrhotite, and xenomorphic pyrite with inclusions of chalcopyrite and arsenopyrite. Quartz is frequently replaced by siderite of the younger, vein-type carbonate-sulfide mineralization, which comprises siderite, ankerite-dolomite, galena, dark brown-colored sphalerite, and minor chalcopyrite, pyrite, tetrahedrite, boulangerite, and bournonite. Chalcopyrite is present in the form of xenomorphic aggregates or inclusions in sphalerite.
Uranium stage mineralization
The uranium stage can be schematically subdivided into the pre-ore, ore, and post-ore substages.
The pre-ore substage is characterized by widespread albitization, hematitization, chloritization, and carbonatization of metasedimentary rocks and migmatites. Veins and veinlets of K1 and K2 calcite commonly contain minute needle-shaped hematite and euhedral quartz grains.
Post-uranium stage mineralization
Quartz-carbonate-sulfide mineralization of the post-uranium stage forms either veins or veinlets that cut older uranium-bearing mineral assemblages or cements uraniferous carbonate breccias. The paragenetically oldest part of the mineralization is represented by quartz and yellow–brown-colored sphalerite followed by galena, and to a lesser extent by chalcocite, bournonite, tetrahedrite, freibergite, chalcopyrite, cubanite, arsenopyrite, dyscrasite, and native antimony enclosed in carbonate matrix. Tabular pyrrhotite and isometric pyrite are common. Gangue minerals are mainly quartz, carbonate (K7 calcite), barite, and minor chlorite and fluorite. In cathodoluminescence images, carbonate shows bright orange luminescence and distinct zonal structure (Fig. 8c). The paragenetically youngest part of the mineralization is represented by zeolites and vermiculite.
Petrology of retrograde-metamorphic and hydrothermal alteration styles
Large-scale host rock alteration styles at Rožná can be distinguished as (1) related to retrograde-metamorphic reactions in sheared host rocks and to the pre-uranium mineralization, (2) related to the pre-ore, ore, and post-ore substages of the uranium mineralization, and (3) alteration that accompanies quartz-carbonate-sulfide mineralization of the post-uranium stage (Fig. 6).
Retrograde-metamorphic and pre-uranium stage alteration
Representative composition of chloritized biotite, inherited chlorite, and authigenic chlorite at the Rožná deposit
Structural formulae based on 28 oxygen atoms
Uranium stage alteration
Temperatures of formation of authigenic chlorite calculated on the basis of its chemical composition after different authors
Cathelineau and Neiva (1985)
Zang and Fyfe (1995)
Pre-ore substage chlorite (n = 12), °C
Post-ore substage chlorite (n = 4), °C
Post-uranium stage chlorite (n = 6), °C
Representative chemical analyses of muscovite and illite from the Rožná deposit
Structural formulae, occupancy and charge, based on 20 oxygen atoms (after Stevens 1946)
The hydrothermal alteration of the ore substage overlaps with the pre-ore substage alteration in time and space, but is generally confined to the vicinity of the uranium ore bodies. In the peripheral part of the ore substage alteration halo, which extends up to 80 m from the ore bodies, biotite, previously altered biotite, and inherited chlorite are transformed to Fe-rich [Fe/(Fe+Mg) = 0.6–0.8] chlorite (Fig. 9, Table 1, analysis 5). As revealed by Mössbauer spectroscopy, the Fe3+/Fe2+ratio in inherited chlorite is high (0.087), which points to the oxidative character of the ore fluids. In the inner part of the halo, adjacent to and within high-grade uranium ore bodies, Fe-rich chlorite is almost entirely decomposed into a mixture of XRD-amorphous hydrooxides of iron and Fe-rich illite (Fig. 10e). Illite from the ore substage alteration halo differs from the pre-ore substage muscovite in its low cation content in interlayer positions (1.1–1.2) and in higher viRII/(viRII+viRIII) and viAl/(viAl+viFeIII) ratios (0.04–0.06 and 0.35–0.8, respectively), calculated on the basis of 20 O (Fig. 11, Table 3, analyses 3 and 4). The Fe2O3 content ranges between 4.5 and 10.5 wt%. Anhydrite commonly occurs in association with Fe-rich illite and iron hydrooxides. Minor authigenic chlorite that overgrows uraninite and coffinite in carbonate veins is characterized by higher Fe/(Fe+Mg) ratios (0.5–0.6) compared to those of pre-ore substage authigenic Mg-rich chlorite. Authigenic apatite and monazite form accessories in ore stage hydrothermally altered rocks.
Compared with the voluminous pre-ore and syn-ore substage alteration, post-ore substage alteration is restricted to small authigenic chlorite-carbonate veins and veinlets (Fig. 10f) and disseminations of chlorite and pyrite in carbonatized host rocks. The chemical composition of authigenic chlorite in carbonate veins is comparable with the chemistry of ore substage authigenic chlorite and its chemistry indicates temperatures ranging from 150°C to 170°C, depending on the thermometer used (Table 2).
Post-uranium stage alteration
Alteration styles that accompany post-uranium quartz-carbonate-sulfide mineralization comprise carbonatization, silicification, and argillization of the host rock. The clay mineral assemblage is dominated by illite with minor kaolinite and chlorite. No mixed layers and smectite group minerals were detected in the <2-μm fraction. Compared with illite that accompanies uranium mineralization, illite from post-uranium alteration has lower Fe3+ content [viAl/(viAl+viFeIII) = 0.35–0.80] and more variable K content in interlayer positions (1.11–1.65, Fig. 11, Table 3, analyses 5 and 6). In contrast to relatively Na-rich uranium-related muscovite and illite, Na contents in post-uranium illite are low. Rare authigenic post-uranium stage chlorite (Fig. 9, Table 1, analysis 8) forms fan-like, light green aggregates, locally replacing fluorite. The temperature of its formation was estimated at 170°C to 190°C (Table 2).
Retrograde-metamorphic fluids and fluids from pre-uranium mineralization
In early barren quartz of quartz-sulfide veins, only secondary H2O–CO2–CH4 and CO2–CH4 inclusions were found. These inclusions are characterized by Tm CO2 of −59.1°C to −59.3°C, which indicates XCH4 in gaseous phase from 0.14 to 0.15. The Th CO2(L) range between 17.6°C and 26.0°C, which corresponds to a density of the gaseous phase between 0.697 and 0.802 g/cm3. Tm clat CO2 range between 7.6°C and 9.2°C, which corresponds to salinities of 1.5–4.5 wt% NaCl eq. (Fig. 12a). Several types of primary inclusions were found in late quartz in quartz-sulfide veins, which host pyrrhotite, pyrite, sphalerite, and chalcopyrite mineralization. Primary H2O–CO2–CH4 and CO2–CH4 inclusions have variable liquid to vapor ratios (LVR). Tm CO2 range from −58.1°C to −59.4°C, and XCH4 in the gaseous phase is 0.08–0.16. Th CO2(L) range between −1.8°C and 22.4°C, which indicates CO2 densities of 0.749–0.940 g/cm3. In several H2O–CO2–CH4 inclusions, CO2 homogenized to the gaseous phase at temperatures of 8.9–9.9°C (DCO2 = 0.130–0.134 g/cm3). In these inclusions, Tm CO2 range from −59.2°C to −59.4°C, thus XCH4 in the gaseous phase is approximately 0.22. Tm clat CO2, measured in only one inclusion, is 9.8°C, indicating a low salinity of about 1 wt% NaCl eq. The variable LVRs in H2O–CO2–CH4 inclusions indicate that they were trapped in a heterogeneous environment under conditions of partial immiscibility of a H2O-rich and a CO2±CH4-rich phase. Low salinities of the aqueous solution (approximately 3 wt%) imply complete miscibility of the aqueous and gaseous phases at 300°C (Hendel and Hollister 1981). The irregular LVRs of the studied inclusions suggest that these were trapped at temperatures close to or lower than 300°C. Only few primary CH4 inclusions were found in some quartz grains. Their Th CH4(L) are between −95.7°C and −98.8°C, which corresponds to DCH4 of 0.28–0.30 g/cm3. In rare secondary CH4 inclusions, CH4 homogenized to liquid at temperatures between −119.8°C and −129°C, indicating DCH4 of 0.35–0.37 g/cm3. Temperatures of melting of ice of primary H2O-only fluid inclusions range from −0.4°C to −1.2°C, which corresponds to low fluid salinities (0.7–2.1 wt% NaCl eq).
Fluids from uranium mineralization
Dominantly aqueous inclusions and, to a lesser extent, CH4, CO2–CH4, and H2O–CO2–CH4 inclusions were observed in quartz of the pre-ore substage. Homogenization temperatures (Th) of the primary H2O inclusions in quartz and K1 to K2 carbonates (Fig. 6) range between 119°C and 206°C, and salinities between 7.3 and 21.8 wt% NaCl eq (temperatures of final ice melting Tm = −4.6°C to −19.1°C; Fig. 12a). Temperatures of first ice melting (Tfm) range between −52°C and −54°C, thus indicating a high content of CaCl2 and probably also MgCl2 or FeCl2 in the solution (Fig. 12b). Th were measured in only one sample with inclusions with uniform LVRs and ranged between 181°C and 206°C. For other inclusions, Th estimation was prevented by variable LVRs. Primary H2O–CO2–CH4 inclusions with variable LVRs were found in some quartz crystals. These inclusions display Tm CO2 of −59.3°C to −59.5°C, thus XCH4 in the gaseous phase is 0.33–0.35. Th CO2 to vapor vary from −5.5°C to −7.5°C (DCO2 = 0.076–0.080 g/cm3). The salinity of the aqueous solution is approximately 7.5 wt% NaCl eq (Tm clat CO2 = 5.4–5.6°C).
Primary H2O inclusions with variable LVRs occur in K3 and K4 ore substage carbonate (Fig. 6). Homogenization temperatures for inclusions with a consistent LVR (0.9) are 152–174°C. The salinities are highly variable (Fig. 12a) and range from 3.1 to 23.1 wt% NaCl eq (Tm = −1.8°C to −21.0°C). Tfm range from −52°C to −54.2°C and indicate high CaCl2 concentrations in the solution (Fig. 12b).
Primary aqueous inclusions in post-ore substage K5 and K6 carbonate (Fig. 6) and in associated quartz show variable LVRs. Few inclusions with LVRs of approximately 0.9 display Th of 84–128°C. Salinities of aqueous solutions in calcite and quartz range from 0.5 to 12.4 wt% NaCl eq (Tm = −0.3°C to −8.6°C), and Tfm between −35°C and −41°C (Fig. 12a,b). The given range of Tfm values cannot be interpreted unambiguously. The probable components of the aqueous solution are NaCl, MgCl2, FeCl2, and KCl (Borisenko 1977). Moreover, coexisting primary H2O, CH4, or H2O–CH4 inclusions with variable LVRs are exceptionally found in quartz that forms a part of the chlorite-pyrite-calcite-bitumen post-ore substage mineral association. The inclusions are believed to be trapped under conditions of H2O and CH4 immiscibility. Th of H2O-only inclusions are 118–131°C and salinities are 8–9 wt% NaCl eq. Th of CH4 to the gaseous phase (−85.2°C to −872°C) indicate its low density (0.081–0.094 g/cm3).
Fluids from post-uranium mineralization
The fluid inclusion characteristics of this stage are highly variable. Early quartz in association with sphalerite contains primary aqueous inclusions with LVRs of approximately 0.9 with Th of 165–178°C and high salinities (25.0–25.5 wt% NaCl eq, Fig. 12a). In paragenetically younger minerals, the salinities of aqueous solutions range between 3.2–16.3 wt% NaCl eq in quartz, 3.9–11.8 wt% NaCl eq in calcite, and 4.5–8.7 wt% NaCl eq in barite. The Tfm value of −56.2°C in barite-hosted inclusions (Fig. 12b) indicates CaCl2–NaCl–H2O composition of the solution. In drusy quartz, monophase liquid-only inclusions are found, indicating low temperatures (<100°C) of the latest stage of fluid activity.
Geochemistry of stable isotopes
Generally positive δ34S values of metamorphosed (pre-Variscan) stratabound sulfides reflect the isotope composition of sulfate sulfur of marine origin (Kříbek et al. 1996, 2002). The isotope composition of pre-uranium stage quartz-sulfide mineralization (δ34S = −5 to +28‰, CDT) thus reflects a host rock source of sulfur. The pre-uranium stage carbonate-sulfide mineralization (Fig. 6), however, has a narrow compositional range between −3‰ and +4‰ (CDT), and the calculated δ34S value of the fluid phase close to 0‰ points to a homogeneous, lower crustal source of sulfur (Fig. 13).
Carbon and oxygen
The δ13C values of pre-ore substage K1 and K2 carbonates and ore substage K3 and K4 carbonates of the uranium mineralization are very homogeneous (δ13C = −3‰ to −7‰), and their composition corresponds to a deep-seated origin of carbon. A recalculation to fluid δ13C values is complicated by the fact that the H2CO3/HCO3− ratio is not precisely known and probably varied during the low-temperature stages. A δ13C range of −4‰ to −8‰ can be estimated for the temperature of the main phase of the uranium deposition (approximately l50°C, based on fluid inclusion studies).
Compared with ore carbonates, calcites of the post-ore substage of the uranium mineralization stage (calcite K5 and K6) and carbonates of quartz-carbonate-sulfide mineralization of the post-uranium stage (calcite K7) are characterized by more variable δ13C values (Fig. 15), thus indicating multiple sources of carbon including host rock carbonates and percolating meteoric waters. In contrast to the relatively homogeneous oxygen isotope composition of early generations of K1 to K4 carbonate (δ18O = +13‰ to +18‰ SMOW), the δ18O values of the late generations K5 to K7 are more variable (+4‰ to +17‰), thus indicating multiple sources of solutions, including meteoric waters.
Oxygen and hydrogen
Isotope composition of oxygen and hydrogen of silicates in the host rock and altered host rock at the Rožná deposit
δ18O (‰ SMOW)
δD (‰ SMOW)
HOR6/metagranite (garnet-biotite granitic gneiss)
HOR5/retrograde-metamorphic quartz-muscovite vein
Pre-uranium stage alteration
SP-8/hanging wall of sulfide-siderite vein
Uranium stage alteration, pre-ore substage
ME-10/albitized and hematitized gneiss
ME-8/albitized and hematitized pegmatite
ME-1/albitized, chloritized and carbonatized biotite gneiss
ZA-14A/albitized and chloritized biotite gneiss
ZA-2/albitized, hematitized and carbonatized gneiss
ZA-10/Chloritized and carbonatized gneiss
Uranium stage alteration, ore substage
ZA-14A/Altered gneiss, hangingwall of uraninite-calcite vein
Inherited, Fe-rich chloritea
ZA-14B/Altered gneiss, footwall of uraninite-calcite vein
Inherited Fe-rich chloritea
During the pre-ore substage of the uranium mineralization, however, the δD values substantially increase (−39‰ to −58‰; n = 6). Because of difficulties with the separation of pre-ore and ore substage chlorites, only two samples of ore substage chlorite that overgrows uraninite accumulations were analyzed. The δD data of these chlorite samples are −87‰ and −89‰.
Timing of hydrothermal events
K–Ar age of illite, mica, and K-feldspar
K–Ar ages of muscovite, K-feldspar, and illite from the Rožná deposit
Grain size (μm)
40Ar air (%)
Coarse-grained retrograde muscovite, quartz-muscovite veinlet
307.6 ± 8.7
V1-XVIII1 VP RGS-R6 (20. vein)
Fine-grained white mica, alteration halo of the pre-uranium carbonate (siderite)-sulfide vein
307.5 ± 6.0
312.1 ± 7.7
Coarse-grained K-feldspar, pre-ore substage of the uranium mineralization
312.0 ± 6.2
Coarse-grained K-feldspar, pre-ore substage, albitized and hematitized gneiss
308.0 ± 7
Authigenic adularia, pre-ore substage
281.6 ± 5.4
Authigenic, fine-grained adularia, pre-ore substage
296.3 ± 7.5
Illite, argillitized breccia with uranium mineralization (U mineralization slightly younger)
277.2 ± 5.5
276.1 ± 5.1
274.9 ± 4.7
Illite, argillitized breccia with uranium mineralization (U mineralization slightly older)
275.6 ± 5.9
273.9 ± 5.2
264.1 ± 4.3
Illite, argillitized hangingwall of the post-uranium, quartz-carbonate-sulfide mineralization
233.7 ± 4.7
231.5 ± 4.5
227.5 ± 4.6
Chemistry and chemical U–Pb age of monazite
Concentrations of U, Th, and Pb, and chemical U–Pb age of monazite in the host rock and altered host rock at the Rožná deposit (concentrations in wt%)
Monazite in quartz, biotite gneiss
468 ± 56
Monazite in feldspar, biotite gneiss
434 ± 52
Monazite in feldspar, biotite gneiss
480 ± 57
Monazite in feldspar, biotite gneiss
410 ± 49
Monazite in feldspar, biotite gneiss
324 ± 39
Monazite in feldspar, biotite gneiss
390 ± 47
Monazite in feldspar, biotite gneiss
336 ± 40
Monazite in biotite, biotite gneiss
449 ± 54
Monazite in feldspar, biotite gneiss
350 ± 42
Hydrothermal monazite, pre-ore substage alteration, gneiss
286 ± 66
Hydrothermal monazite, ore substage alteration, gneiss
264 ± 48
Hydrothermal monazite, ore substage alteration, gneiss
260 ± 57
Hydrothermal monazite, ore substage alteration, gneiss
263 ± 31
Graphite and organic geochemistry of bitumen
The host biotite gneiss at Rožná contains up to 0.4 wt% graphite (Cgraph) in the form of lamellae up to 0.2 mm large. Reflectivity of graphite (Rmax = 10% to 13%, Rmin = 1.5% to 2.0%) and distinct optical anisotropy of graphite lamellae reveal a high structural ordering. The uraniferous mylonites and cataclasites of the Rožná shear zone have elevated Cgraph contents (1.0 to 3.5 wt%). Graphite lamellae in slightly sheared rock are commonly bent and distorted. With an increase of plastic deformation, graphite flakes are pulverized to form black, submicroscopic particles that are concentrated, together with sericite and chlorite, in planes of maximum strain. There is no evidence of recrystallization or neoformation of graphite. During the superimposed brittle deformation, graphite pigment and fine-grained phyllosilicate particles were mechanically reconcentrated into “graphite seams” up to several millimeters thick.
Mineral parageneses, sources of metals, and P–T–X characteristics of fluids
The formation of the pre-uranium stage quartz-sulfide and carbonate-sulfide mineralizations at Rožná and the widespread pyritization, chloritization, sericitization, and carbonatization of the adjacent host rock can be related to a generally reduced retrograde-metamorphic fluid phase. The δ18O of quartz and δ18O and δD of mica in the alteration halos of both mineralization types are consistent with this interpretation. Variable liquid to vapor ratios and Th CO2 for inclusions in quartz of the quartz-sulfide and carbonate-sulfide mineralization and in quartz from retrograde-metamorphic quartz-muscovite veins indicate the trapping of inclusions under conditions of immiscibility of H2O and CO2 phases (below the solvus of the low-salinity H2O–CO2 system), at temperatures probably below 300°C. Using a different set of ore samples, Vosteen and Weinoldt (1997) and Hein et al. (2002) reported a similar temperature range (280°C to 300°C) for pre-uranium stage carbonate-sulfide mineralization. The results of the fluid inclusion studies correspond with the temperature estimate based on sulfide thermometry (approximately 300°C). Quartz-sulfide and carbonate-sulfide mineralizations, however, differ substantially in the isotope compositions of their sulfides. A wide range of δ34S values of sulfides from the older quartz-sulfide mineralization (δ34S = +7‰ to +25‰) is attributed to local remobilization of sulfur from host rocks that contain sulfides and sulfates with mostly positive δ34S values (Kříbek et al. 1996, 2002). In contrast, the narrow spread of the δ34S data (−3‰ to +4‰) of the younger carbonate-sulfide mineralization points to a deep-seated, homogeneous sulfur reservoir and to focussed fluid flow along structural channels, i.e., ductile to brittle shear zones, during fracture propagation to the deeper crust. This scenario seems to be supported by the high δ18O and low δ13C values of carbonate (siderite and ankerite) indicative of a high-temperature equilibration of fluids with host rock, including hydrolytic decomposition of graphite.
As evidenced by prominent hematitization, especially in deeper parts of Rožná, the barren, pre-ore substage of the uraniferous fluid flow event indicates deep infiltration of oxidized, surface-derived fluids to the crystalline basement. Deep circulation of fluids of the pre-uranium substage gave rise to desilicified, hematitized, and albitized rocks. The petrological and geochemical character of the pre-ore altered rocks is very similar to that of episyenites described from numerous uranium deposits of the Variscan belt of Europe (Leroy 1984; Cathelineau 1986; Cuney et al. 1989). In contrast to episyenites, which are usually products of late magmatic alteration of granites, the hydrothermally altered rocks at Rožná are of metamorphic origin. The fluids responsible for the alteration of rocks of the pre-ore substage differ from the earlier low-salinity metamorphic fluids in their generally higher but highly variable salinities (7.3 to 21.8 wt% NaCl eq). Important differences in the salinity of the pre-uranium fluids probably reflect the large-scale mixing of chemically heterogeneous basinal brines with meteoric water. Compared with metamorphic fluids that produced the pre-uranium stage mineralization (mostly NaCl fluids), pre-ore substage fluids contain high amounts of CaCl2 and possibly MgCl2 and FeCl2, which corresponds to the widespread albitization and hematitization of the host rock and formation of authigenic Mg-chlorites. The oxidized character of the fluids is evidenced by hematitization and the higher Fe3+/Fe2+ ratio in chloritized biotite and protochlorite in hematitized rocks compared with the same ratio in chloritized biotite from pre-uranium alteration. Based on the Th of the aqueous inclusions, the temperature of the hydrothermal fluids of this substage can be estimated at 200°C. Hein et al. (2002) reported similar temperatures (Th = 179°C to 236°C) and lower salinities (1.1 to 8.1 wt% NaCl eq), based on fluid inclusion data on pre-ore substage carbonates. Chlorite thermometry of pre-ore, Mg-rich authigenic chlorites yielded temperatures in the range of 260°C to 310°C. These roughly correspond to the temperature calculated using the oxygen thermometer in the quartz–albite pair (Žák et al. 2001). Using the volumetric properties of fluid inclusions and pressure correction (Potter 1979), based on chlorite and isotope thermometers, the pre-ore mineralization (K1 and K2 calcite generation) is estimated to have formed at 0.6 to 1.3 kbar.
The isotope composition of the fluids of the pre-ore substage can be estimated at +6‰ to +10‰ δ18O and 0‰ to −30‰ δD, respectively. The high δ18O and δD values of pre-ore fluids can be best explained as due to the infiltration of basinal fluids from now eroded Late Stephanian and Early Permian cover rocks into the crystalline basement. Subtropical to tropical and, at least in some periods, arid climatic conditions are evidenced by the occurrence of anhydrite in Permian sedimentary rocks of the Bohemian Massif (Skoček et al. 1977). Elevated evaporation probably produced lake waters with relatively high δ18O and δD characteristics. For the Permo-Triassic waters of Europe, Wilkinson et al. (1995) reported δD values between −14‰ and −3‰. Turpin et al. (1981) calculated δD of −5‰ for basinal fluids giving rise to the late Variscan vein mineralization of Portugal. The δ18O and δD characteristics of basinal brines can be shifted to even higher values on fluid–evaporite interaction (Sheppard 1986). The reaction of basinal waters with hematite and anhydrite will lead to an oxidation state of the fluid well above the magnetite-hematite buffer (Sverjensky 1987). Oxidizing basinal brines can be very effective in leaching uranium. Compared with the average contents of uranium and thorium in Moldanubian gneisses (3.6 ppm U, 11.1 ppm Th, Th/U = 3.22, n = 68; René 2002), unaltered host gneiss and migmatites of the Rožná-Olší uranium ore field are generally slightly enriched in uranium (5.2 ppm U, 8.4 ppm Th, Th/U = 3.27, n = 14). Uranium in host gneisses and migmatites is essentially hosted by monazite, which contains 5.4 wt% Th and 0.9 wt% U on average, and to a lesser extent, by thorogummite, cheralite, and zircon (Leichmann et al. 2002). In barren, pre-ore substage hydrothermally altered rocks, monazite, thorogummite, and cheralite are missing and zircon is slightly corroded. Therefore, the source of uranium may be in the hydrothermal decomposition of uranium-bearing accessories, as proposed for the unconformity-type uranium deposits in Canada by Hecht and Cuney (2000). A similar scenario, i.e., the extraction of uranium from crystalline rocks by Permian formation waters was proposed by Turpin et al. (1990) to explain the origin of uranium in hydrothermal uranium deposits in Hercynian granites of the Massif Central, France.
Uranium mineralization followed, or partly overlapped in time, the pre-ore substage alteration. Uranium-enriched fluids were probably driven in the direction of thermal and pressure gradients toward the permeable zones of pyritized and graphite-enriched cataclastic zones that frequently cut and displace pre-ore altered rocks. Several generations of ore and gangue fault breccia cemented by carbonate and, to a lesser extent, by quartz document multiple opening and healing of the ore-controlling brittle structures and the possibility of tectonically driven flow and seismic pumping of ore-bearing fluids. In the same time period, the porous, albitized, and hematitized rocks that originated during the pre-ore alteration substage served as a “trap” for the uranium mineralization. The same scenario was proposed for uranium-bearing episyenites at uranium deposits of the Massif Central in France (Leroy 1984). Uranium mineralization is represented by uraninite and at least two generations of Y-rich coffinite (commonly zirconium-enriched coffinite); brannerite being far less abundant. In addition to zirconium-enriched coffinite, (U,Zr)-silicate phase was detected in the ore. U-silicate phase with 64 at.% Zr was recorded in bitumens at the Oklo-Okélobondo uranium deposit in Gabon by Jensen and Ewing (2001). At Rožná, uranium is bound by microheterogeneous substances formed by a mixture of U, Ti, Si, Ca, Fe, and Zr. High concentrations of these generally immobile elements, i.e., Ti, Y, and Zr, in hydrothermal fluids are usually attributed to the formation of fluoride or phosphate complexes (Gieré 1990; Rasmussen 2005). However, apatite and monazite occur only as accessories in hydrothermally altered rocks associated with uranium mineralization, and fluorite forms a part of the younger post-uranium quartz-carbonate-sulfide mineralization. In F- and P-poor hydrothermal systems, the mobility of Ti, Y, and Zr may be promoted by sulfate complexing (Rubin et al. 1993). Oxidation of pyrite and formation of anhydrite during the emplacement of the uranium mineralization at Rožná are consistent with this scenario. Selenide mineralization that accompanies uraninite and coffinite is slightly younger than uranium mineralization. As relatively high concentrations of selenium were reported from pre-uranium sulfides by Vencelides (1991), it is highly probable that the oxidation of rock sulfides during the introduction of oxidative, uranium-bearing fluids resulted in the oxidation of both sulfur and selenium. Contrary to SO42−, SeO42− can be easily reduced to form selenides at relatively high oxidation state. This scenario is consistent with the occurrence of both anhydrite and selenides in ore substage hydrothermally altered host rocks. P–T–X conditions of the ore substage of the uranium mineralization can be estimated using fluid inclusion, stable isotope, and chlorite thermometry data. The Th values of primary inclusions range between 152°C and 174°C, and their total salinities span a relatively wide interval of 3.1 to 23.1 wt% NaCl eq. Hein et al. (2002) obtained a similar temperature (Th = 130–166°C) for primary inclusions of the ore stage K3 and K4 carbonates, but lower salinity (3.1 and 7.7 wt% NaCl eq). Isotope compositions of oxygen and hydrogen in Fe-rich ore substage chlorite differ from those of pre-ore chlorite. Percival et al. (1993) reported a similar difference between pre-ore and Fe-rich ore chlorites for the Cigar Lake uranium deposit in Saskatchewan, Canada. Using the data of Suzoki and Epstein (1976), they interpreted the differences in the δD isotope composition as a result of a higher concentration of D in Mg- and Al-rich hydrous minerals relative to Fe-bearing minerals. However, taking into account the significant differences between the δD values of pre-ore and ore chlorites at Rožná, it cannot be ruled out that the δD values for ore chlorite reflect an oxidation of the D-depleted organic molecules in the hydrothermal solution (organic waters, Sheppard 1986).
Uranium precipitation can be controlled by many factors, including a decrease in temperature and pressure, mixing, boiling, and degassing of fluids, decrease in pH, or a decrease in oxygen fugacity (Evert et al. 1997). Moreover, sorption and subsequent reduction of uranium species on the surface of silicates or hydrooxides of iron may play a substantial role in uranium immobilization. Fluid inclusion studies at Rožná provided no evidence of boiling or degassing of uraniferous fluids. Multiple dissolution and reprecipitation of several generations of carbonates as evidenced by CL studies indicate that the pH of the uranium-bearing solutions varied slightly along the boundary of the carbonate stability field. As kaolinite is entirely missing in the hydrothermally altered rocks, the pH of the solutions was probably buffered by the equilibrium between illite and K-feldspar in the range of 7 to 8. Typical alterations that accompany uranium mineralization include the formation of Fe3+-enriched chlorite and the decomposition of previously formed chloritized biotite, protochlorite, and matrix chlorite to Fe-rich illite and hydrooxides of iron (goethite and amorphous Fe hydrooxides). The large extent of altered rocks and the abundant presence of Fe3+-bearing minerals associated with the uranium ore suggest that the deposition of uranium minerals was a response to the reduction of the ore-bearing fluid by direct interaction with ferrous iron-bearing silicate (biotite and chlorite). A similar scenario was proposed for Australian deposits by Wilde et al. (1989). In addition to Fe2+ phyllosilicates, pre-uranium pyritization significantly contributed to the reducing capacity of the host rock. In contrast, the role of graphite in the ore-bearing solution is probably negligible due to the low reactivity of graphite at temperatures below 300°C (French 1966; Frost 1979).
Homogenization temperatures of post-ore substage carbonate inclusions range between 84°C and 131°C. Compared with the ore substage inclusions, the salinities are lower (0.5 and 12.4 wt% NaCl eq), which may indicate increasing mixing of formation water with surface-derived rainwater. Hein et al. (2002) reported almost the same homogenization temperatures (80°C to 130°C) for inclusions from post-ore substage carbonates but higher salinities (10.5 to 18.6 wt% NaCl eq). The pressures of 0.2 and 0.7 kbar were estimated for the post-ore mineralization (K5–K6 calcite generation) after pressure correction (Potter 1979) using chlorite and stable isotope thermometers. Exceptional low-density CH4 inclusions of this substage can represent either secondary overprinted, reequilibrated older inclusions or products of radiation-induced polymerization of organic molecules as was proposed for the Příbram uranium deposit by Kříbek et al. (1999) and for Canadian uranium deposits by Landais and Dereppe (1985) and Wilson et al. (2003). Because of the occurrence of solid uraniferous bitumens at Rožná, we prefer the second possibility. The origin of organic molecules during diagenetic and catagenetic transformation of organic matter of the Stephanian and Lower Permian sedimentary rocks was documented by the presence of biogenic tracers (phytane and pristane, i.e., products of the diagenetic decomposition of the chlorophyll molecule) in extractable parts of bitumens at Rožná.
The chalcocite-pyrrhotite-pyrite-chalcopyrite-galena mineral assemblage of the post-uranium stage and the disequilibrium assemblage of pyrrhotite and barite indicate variable sulfidation state. This variation is also expressed in the distinct zoning of carbonate and quartz crystals as revealed by CL images (Kříbek and Hájek 2005; Vosteen and Weinoldt 1997). Fluid inclusions in quartz that forms part of the early mineral paragenesis have Th of 165°C to 178°C and very high salinities around 25 wt% NaCl eq (Hein et al. 2002 report 14.0 to 18.6 wt% NaCl eq). Temperatures obtained for the early stage of mineralization range from 100°C to 200°C using the galena-sphalerite sulfur isotope thermometer and from 170°C to 190°C using the chlorite thermometer. The pressure of trapping of quartz-calcite-sulfide mineralization (K7 calcite generation in association with early quartz and sphalerite) was estimated at 0.2 to 0.4 kbar. Low-salinity inclusions in paragenetically younger K7 carbonates, barite, and associated quartz (3.2–16.3 wt% NaCl eq) probably document an increasing mixing of residual brines in crystalline rocks (“shield brines”; Negrel and Casanova 2005; Leybourne et al. 2006) derived from external sources (brines of Permian basins) or generated in situ by low-temperature crystalline rock–water interactions with meteoric water. In drusy quartz, monophase liquid-only inclusions are present indicating low temperatures (<100°C) of the latest stage of mineralization. The wide spread of sulfide sulfur isotope data indicates either a remobilization of sulfur from older stratabound mineralization that occurs in the host rocks at Rožná or, more likely, a different degree of isotope fractionation between oxidized (sulfate) and reduced (sulfide) forms of sulfur species in the fluid.
Timing of the mineralization and relation to the late Variscan and post-Variscan geological evolution of the Bohemian Massif
- (Ad 1)
All mineralization styles at Rožná are bound to faults and conjugate structures of the shear zone that originated in the upper part of the passively transported Moldanubian nappes during the deep-level core wedging of the Brunovistulian in the east beneath the Moldanubian orogenic root between 340 and 337 Ma (Schulmann et al. 2005). The structures were reactivated during NE–SW dextral transpressive shearing parallel to the Brunovistulian foreland at 330–325 Ma (Schulmann et al. 2005) and later, during the early stage of post-orogenic extension in the Moldanubian domain (325 to 300 Ma; Handler et al. 1991). The K–Ar radiometric ages of white mica from the Rožná mylonites (307.6 ± 8.7 Ma) and from sericite in the wall rock of the siderite-sulfide veins (312 ± 6.0 to 307 ± 7.7 Ma) are consistent with the time span of the post-orogenic extension. It can be, therefore, hypothesized that the origin of fluids giving rise to the quartz-sulfide and carbonate-sulfide mineralizations at Rožná was related to the isothermal decompression, uplift, and hydration of the upper crust following overthrusting and large-scale melting of the lower crust of the Moldanubian domain. This model was suggested for many late Variscan siderite-sulfide vein-type deposits of Variscan Europe (Hein 1993; Radvanec et al. 2004).
- (Ad 2)
Uranium mineralization is significantly separated in time from the previous stage. Radiometric K–Ar dates of authigenic K-feldspar (adularia) in albitized and hematitized rock of the pre-ore substage (296.3 ± 7.5 and 281 ± 7.0 Ma) correspond to the opening of the Late Stephanian to Lower Autunian transpressional or transtensional grabens in the Bohemian Massif (Pešek et al. 2001). The regional distribution of the Late Stephanian and Permian sedimentation in the Bohemian Massif was probably much more extensive compared to the now preserved remnants. Franců et al. (1998), for example, using reflectance data of organic matter in Permian basins, documented that at least 2,800 m of Autunian and younger sediments were eroded in the Boskovice Graben (Fig. 2) before the Early Cretaceous. The initial stage of graben formation (300 to 290 Ma) was characterized by differential block movement in a generally brittle deformation regime. The uranium-enriched basinal fluids were transported along the direction of pressure and temperature gradients, i.e., toward the cataclastic, chloritized and pyritized zones in the time span of 280 to 260 Ma (K–Ar illite and chemical U–Pb monazite dating). These ages correspond with previously published U–Pb ages on uraninite of 280 to 260 Ma (Anderson et al. 1988). The circulation of hydrothermal fluids probably ceased because of the tectonic inversion of the Lower Permian basins during the Late Autunian and Saxonian.
- (Ad 3)
Radiometric K–Ar ages of illite in the host rock of the post-uranium stage (quartz-carbonate-sulfide mineralization) of 233.7 ± 4.7 to 227.5 ± 4.6 Ma reveal a reactivation of uranium-bearing cataclastic shear zones during the Middle to Late Triassic. Tectonic and thermal reactivation of ore-bearing structures and the formation of new, crosscutting faults can be probably attributed to the post-Variscan, early Alpine transcurrent movements all over central Europe (Brandmayr et al. 1995). The reactivation of Variscan and pre-Variscan structures in this period probably reflects the early tensional phase in the Paleotethys-Central Atlantic region (Ziegler 1996).
Our data allow to establish a genetic model of the shear zone-hosted uranium deposit at Rozna and to integrate the mineralization history into the late Variscan and post-Variscan geological evolution of the Bohemian Massif. The origin of the pre-uranium, quartz-sulfide, and carbonate-(siderite-ankerite)-sulfide mineralizations is related to post-orogenic extension of the Moldanubian crust that was accompanied by retrograde-metamorphic alteration of high-grade mylonitized rocks and cataclasites during the Westphalian. Within this period, rocks of the shear zone were pyritized and chloritized, thus forming favorable redox traps.
The origin of the uranium mineralization is associated with infiltration of oxidized basinal brines of the Upper Stephanian and Lower Permian basins into the crystalline basement along deep brittle structures that opened during the late Variscan transcurrent crustal block reorganization. Relicts of Upper Carboniferous and Lower Permian clastic sedimentary rocks in the Moldanubian domain indicate that the areal extent of the Permian sedimentation was much larger than today. It is, therefore, highly probable that Permian sedimentary rocks unconformably covered a substantial part of the periphery of the Moldanubian domain. Infiltration of basinal brines effectively leached uranium from U-bearing accessories and silicates in the crystalline rocks. Quick uplift and tectonic ablation of the basement enabled seismic pumping of ore-bearing fluids toward the chloritized and pyritized zones of cataclasites where uranium was gradually precipitated due to the interaction of ore fluids with reducing host rock lithologies (ferrous silicates and pyrite). The principal role of formation waters of the Upper Stephanian and Permian basins in the origin of uranium mineralization at Rožná thus parallels the model for unconformity-type uranium deposits in Canada and Australia. Within this scenario, the Rožná deposit may represent the basement rock-hosted part of an unconformity-type deposit, the overlying Permian sedimentary cover being almost entirely eroded during the pre-Cretaceous period. Compared with the spatially restricted but high-grade unconformity-type deposits, the relatively large extent of the uranium mineralization and the low ore grade at Rožná points to a relatively defocused fluid system.
In the Middle to Late Triassic, zones of uraniferous cataclasites were reactivated and crosscut by newly formed fault systems. During this hydrothermal event, uranium mineralization was locally remobilized and quartz-carbonate-sulfide mineralization formed. The wide range of salinities of fluids probably documents mixing of residual brines in crystalline rocks (“shield brines”) with meteoric water.
In the context of hydrothermal uranium mineralization of the Variscan belts of western and central Europe, the Rožná uranium deposit and some other deposits of the Moldanubicum of the Bohemian Massif (such as the Mähring, Olší, Dyleň, Zadní Chodov, Okrouhlá Radouň deposits) differ from those of the Massif Central and Armorican Massif in France and from those in the Erzgebirge (Krušné hory Mountains) in Germany/Czech Republic. “Disseminated-type” uranium deposits of the Massif Central and Armorican Massif are generally linked with plutons of fractionated muscovite granites, which are considered the main source of uranium. Rožná and other deposits of the Moldanubicum of the Bohemian Massif show no spatial relationship between mineralization and granitic plutons. Apart from the generally vein-type deposits in the Erzgebirge (Krušné hory Mountains) and the Příbram uranium deposit, which are characterized by the proximity of granitic plutons and by very weak hydrothermal alteration, the Rožná deposit shows extensive albitization, chloritization, and hematitization of the crystalline country rocks and a mostly disseminated style of mineralization. Reactions of hydrothermal fluids with the host rocks can also be held responsible for the much wider variability in chemical composition of the fluids compared to the vein-type deposits. On the other hand, a feature common to all uranium deposits of Variscan Europe is their approximately same age (280–260 Ma) and their spatial association with tectonic structures formed or reactivated in the latest Stephanian or Early Permian. Hydrothermal fluid circulation along structures controlling the mineralization was probably accelerated by elevated heat flow. This is evidenced by the high coalification degree of the Stephanian coal (up to the anthracite stage) in the pull-apart basins which developed on these structures.
This study is based on results of tens of unpublished, and formerly held secret, reports available exclusively from the archives of the Diamo State Enterprise. For many years, the deposit was studied by Russian geologists, i.e., I.M. Bayushkin, V. Ye. Boytsov, J.M. Dymkov, P.A. Ivanov, B.P. Yurgenson, V.S. Katargin and, at different stages of exploitation, by Czech geologists, i.e., F. Fediuk, J. Komínek, M. Kvaček, J. Malec, F. Novák, P. Pauliš, V. Rozhoň, Z. Uhlík, S. Vilhelm, J. Vokoun, V. Zrůstek, and many others. The present paper could have never originated without their detailed geological, mineralogical, and structural studies.
The manuscript benefited from reviews by M. Pagel and an anonymous reviewer. Last but not the least, we wish to thank P. Sulovský and J. Zimák for their interpretation of chlorite chemistry, I. Vavřín, Z. Korbelová, and V. Šrein for microprobe analyses, J. Schneeweis for Mössbauer spectra of phyllosilicates, B. Humer for the determination of chemical U–Pb ages of monazite, and J. Adamovič for his English corrections.