Geochemical fixation of rare earth elements into secondary minerals in sandstones beneath a natural fission reactor at Bangombé, Gabon

To study geochemical processes for migration and fixation of fissiogenic rare earth elements (REE) in association with uranium dissolution, in situ isotopic analyses using an ion microprobe were performed on U- and REE-bearing secondary minerals, such as coffinite, françoisite, uraniferous goethite, and uraninite found in a sandstone layer 30 to 110 cm beneath a natural fission reactor at Bangombé, Gabon. Phosphate minerals such as phosphatian coffinite and françoisite with depleted ²³⁵U (²³⁵U/²³⁸U = 0.00609 to 0.00638) contained large amount of fissiogenic light REE, while micro-sized uraninite grains in a solid bitumen aggregate have normal U isotopic values (²³⁵U/²³⁸U = 0.00725) and small amount of fissiogenic REE components. The proportions of fissiogenic and non-fissiogenic REE components in four samples from the core of BAX03 vary in depth ranging from 30 cm to 130 cm beneath the reactor, which suggests mixing between fissiogenic isotopes from the reactor and non-fissiogenic isotopes from original minerals in the sandstone. Significant chemical fractionation was observed between Ce and the other REE in the secondary minerals, which shows evidence of an oxidizing atmosphere during their formation. Pb-isotopic analyses of individual minerals do not directly provide chronological information because of the disturbance of U-Pb decay system due to recent geologic alteration. However, systematic Pb-isotopic results from all of the minerals reveal the mobilization of fissiogenic isotopes, Pb and U from the reactor in association with dolerite dyke intrusion ∼0.798 Ga ago and the formation of the secondary minerals by mixing event between 2.05 Ga-old original minerals and reactor materials due to recent alteration.     

Isotopic evidence for trapped fissiogenic REE and nucleogenic Pu in apatite and Pb evolution at the Oklo natural reactor

A part of the boundary layer of reactor zone 10 at the Oklo natural reactor shows a unique petrologic texture, which contains high-grade uraninite and massive apatite concretions. In order to study distribution behavior of fission products around the boundary between the reactor zone and the wall rock and to clarify the relation of migration mechanisms of fission products with geochemical factors, in-situ isotopic analyses of Nd, Sm, Gd, Pb and U in uraninite and apatite from the sample were performed by Sensitive High Resolution Ion Microprobe (SHRIMP). Sm and Gd isotopic ratios of uraninite and apatite show evidence of neutron irradiation with fluence between 4.4-6.8×10¹⁹ n/cm². Judging from the isotopic anomalies of Nd and U, the apatite coexisting with the uraninite plays an important role in trapping fissiogenic LREE and nucleogenic ²³⁹Pu into the structure. Systematic Pb isotopic data from apatite, uraninite, galena and minium suggest the following chronological interpretations.

1.

The apatite formed 1.92±0.01 Ga ago and trapped fissiogenic light REE and nucleogenic ²³⁹Pu that migrated from the reactor during the criticality.

2.

The uraninite around the boundary between reactor and sandstone dissolved once 1.1∼1.2 Ga ago.

3.

Galena grains were formed by U-Pb mobilization in association with the intrusion of dolerite dyke 0.45∼0.83 Ga ago.

4.

Minium was derived from recent dissolution of galena under locally oxidizing conditions.

     

Formation and geochemical significance of micrometallic aggregates including fissiogenic platinum group elements in the Oklo natural reactor, Gabon

Metallic aggregates with a size of a few tens μm and consisting mainly of Ru, Rh, Pd, Te, Pb, As, Sb, S and Bi were found in the acid residue of SD37-S2/CD uraninite taken from Oklo natural reactor zone (RZ) 13. Quantitative analyses of major elements using an electron probe microanalyzer and in situ isotopic analyses of Zr, Mo, Ru, Pb and U using a sensitive high-resolution ion microprobe were performed on the metallic aggregates to determine the geochemical behaviors of fission products and actinides and to ascertain the processes of formation of the aggregates in the RZs. The chemical compositions of the aggregates investigated in this study are significantly different from those reported previously, showing lower Pb content and no correlation between the contents of Pb and S in the individual grains. The ²³⁵U/²³⁸U ratios in metallic aggregates vary significantly from 0.00478 to 0.01466, indicating chemical fractionation between U and Pu during the formation of the aggregates. The Pb isotopic data indicate that most of the Pb in the aggregates decayed from 2.05 Ga-old uraninite that existed in the RZ originally and that there was chemical fractionation between U and Pb in some aggregates. The Zr and Mo isotopic ratios, ⁹⁰Zr/⁹¹Zr and ⁹⁵Mo/⁹⁷Mo, for most of the aggregates had small variations, which can be simply explained by constant separate mixing of fissiogenic and nonfissiogenic components. On the other hand, a large variation in the ⁹⁹Ru/¹⁰¹Ru ratio (0.324-1.73) cannot be explained only by a two component mixing theory; thus, chemical fractionation between Tc and Ru during the reactor criticality is suggested. The large variations in the ²³⁵U/²³⁸U and ⁹⁹Ru/¹⁰¹Ru isotopic ratios suggest that the aggregates formed under various redox conditions owing to the radiolysis of water.     

SHRIMP in situ isotopic analyses of REE,Pb and U in micro-minerals bearing fission products in the Oklo and Bangombé natural reactors: A review of a natural analogue study for the migration of fission products

In the Precambrian, parts of the Oklo, Okélobondo and Bangombé uranium deposits of the Republic of Gabon, central Africa, functioned as natural fission reactors. Many elements in the Oklo and Bangombé uranium deposits show variations in isotopic composition caused by a combination of nuclear fission, neutron capture and radioactive decay. Isotopic studies provide useful information to understand the behavior of radionuclides in geological media. In our recent work, in situ REE, Pb and U isotopic analyses of individual tiny minerals in and around reactor zones have been performed using a SHRIMP (Sensitive High Resolution Ion Microprobe). The isotopic results of the SHRIMP analyses on micro-minerals found in and around the Oklo and Bangombé natural reactors are reviewed in this paper. The data suggest the selective uptake behavior of (1) Ra into illite, and (2) Pu into apatite, (3) the formation process of secondary minerals bearing fissiogenic REE and depleted U, (4) evidence of nuggets (?-particles) bearing fissiogenic platinum group elements (PGE), and (5) from the U–Pb systematics of highly altered zircons, the redistribution of U and Pb.     

Orogenesis vs. diagenesis: Can we use organic-rich shales to interpret the tectonic evolution of a depositional basin?

Black shales from the southern Appalachian Basin and the southwest Welsh Basin have anomalous U–Pb and Nd model ages suggesting syn- and post-depositional resetting of the Sm–Nd and U–Pb isotopic systems. This alteration to the primary detrital signature of these two shale sequences is indicative of black shale diagenetic/depositional processes that obscure paleo-environmental and provenance information recorded prior to and during deposition. The trace element and isotopic signatures of these two shale sequences reveal a syn-/post-depositional history that is de-coupled from the coeval orogenic history of the region making it difficult to reconstruct the tectonic and oceanographic conditions present at the time of deposition.Both the Ordovivian Welsh Basin and the Devonian Appalachian Basin sequences host REE- and U-bearing diagenetic phosphate minerals that play a critical role in the whole rock REE and U budgets. In the Welsh Basin shales, early diagenetic apatite and a later monazite phase dominate the REE budget and cause the redistribution of REE early in the basin's history (ca. 460 Ma). This redistribution is recorded by the Sm–Nd system (450 ± 90 Ma) and the Nd model ages that are anomalously old by as much as 20% (T_DM > 2.0 Ga). This early history is complicated by a Permo-Triassic fluid event affecting the whole rock U-budget and resetting the U–Pb isotopic system at 193 ± 45 Ma. The Appalachian Basin sequence appears to have a much less complicated history yet still records a significant disturbance in both the Sm–Nd isotopic system (392 ± 76 Ma) and the Pb isotopic system (340 ± 50 Ma) at about the time of deposition (ca. 365 Ma).These two sequences suggest a pattern of diagenetic disturbance common to black shales. These processes are unique to black shales and must be considered when interpreting provenance and paleo-environmental information from the black shale sequences. Although these rocks are susceptible to alteration, the alteration may provide extensive information on the post-depositional history of the basin while still retaining some primary depositional information. If black shale processes are considered during the interpretation of isotopic and trace element signatures from organic-rich shales, it may be possible to recover an extensive basin history.     

Mobilization and mechanisms of retardation in the Oklo natural reactor zone 2 (Gabon)—inferences from U, REE, Zr, Mo and Se isotopes

Mineralogical and isotopic studies were carried out on the natural nuclear reaction zone 2 from the Oklo deposit to evaluate the mobility of several nuclear reaction products in response to the alteration of the reaction zone and to identify the mechanisms which could retard the transport of released radionuclides. To address these issues, in situ isotopic analyses by SHRIMP and a selective extraction procedure were performed to constrain the structural location of nuclear reaction products (exchangeable and non exchangeable) and their association with mineral phases. The distribution patterns of U, REE, Zr and Mo isotopes reveal that substantial amounts were released from the core and migrated through the hydrothermal alteration halo over metric distances, owing to uraninite dissolution and advective transport by hydrothermal solutions during and soon after criticality. The results emphasize the mobility of Zr at Oklo, this element being often considered as “immobile” during water–rock interactions. The main output is the demonstration of the net effects of sorption and coprecipitation processes. Chlorite and to a lesser extent illite were found to have adsorbed significant amounts of U, REE, Zr (and probably Th) and less sorbing elements such as Mo. Coprecipitation of secondary UO₂ and P-rich coffinite within the alteration halo is also an important means of retardation. The concentration of radionuclides released from the reactor were probably high and they display solubility limited transport behaviour. No retention effect was found for Se in the immediate vicinity of the reactor and this element may have moved farther from its source of production. These results have interesting implications for the evaluation of long-term containment of radionuclides. They provide a simple illustration of the performance of a clay barrier in the uptake of radionuclides by sorption onto clays and reincorporation in secondary U-minerals. This study also demonstrates the robustness of these retention processes over extremely long periods of time.     

Characterization of a La-Ce-Sr-Ca aluminous hydroxy phosphate in nuclear zone 13 in the Oklo uranium deposit (Gabon)

A La-Ce-Sr-Ca aluminous hydroxy phosphate of the goyazite-florencite-crandallite series was observed in the nuclear reactor zone 13 in the Oklo uranium deposit (Gabon). It has a trigonal elementary cell with a?=?6.98 ± 0.03 Å and c?=?16.35?±?0.04?Å (space group P3m). It occurs in a U-rich sample depleted in ²³⁵U (²³⁵U/²³⁸U?=?0.0047). Its chemical composition, measured by ion microprobe analyses, shows that it contains fission products (Zr, LREE) whereas the isotopic composition of Ba and Sr is natural. Ion microprobe imaging indicates that the nearby uraninite crystals are depleted in fissiogenic LREE. This study shows that the migration of some fission products, probably related to a dolerite dike intrusion, is limited by the formation of aluminous hydroxy phosphates.     

Rare earth elements, yttrium and H, O, C, Sr, Nd and Pb isotope studies in mineral waters and corresponding rocks from NW-Bohemia, Czech Republic

The sparkling waters from the area of Kyselka near Karlovy Vary at the western slope of the Doupovske hory, Bohemia (Czech Republic), and CO₂-poor waters from two underground boreholes at Jachymov, Krusne hory, Bohemia, have been studied with the aim of characterizing the distribution of rare earth elements, yttrium, and H, O, C, Sr, Nd, Pb isotopes during the low-temperature alteration processes of the host rocks. Additionally, leaching experiments were performed at pH 3 on the granitic and basaltic host rocks from Kyselka and the granite of Jachymov. All REE patterns of the granite- and the basalt-derived waters from the Kyselka area are different from those of their source rocks and the leachates of the latter. This elucidates the inhomogeneous distribution of REE and Y among the solid phases in the altered magmatic rocks. The Eu and Ce anomalies in granite-derived waters are inherited, the Y anomaly is achieved by fluid migration. Yttrium is always preferentially leached by mineral waters, whereas Y/Ho ratios of rocks and their leachates are very similar. The REE abundances in waters from the wells in Jachymov are derived from rocks intensely leached and depleted in easily soluble REE-bearing minerals, whereas the granites and basalts from Kyselka still contain soluble, REE-bearing minerals. A comparison of REE/Ca patterns of the experimental leachates with those of the mineral waters elucidate the high retention of REE in rocks during water–rock interaction. In strongly altered rocks Sr isotope ratios of mineral waters and rocks differ widely, whereas the corresponding Nd isotope ratios are very similar. ²⁰⁷Pb/²⁰⁸Pb, ²⁰⁶Pb/²⁰⁸Pb and ²⁰⁶Pb/²⁰⁷Pb ratios in mineral waters are independent from U/Th ratios in the rocks. ²⁰⁶Pb/²⁰⁸Pb and ²⁰⁶Pb/²⁰⁷Pb are lower in mineral waters than in their source rocks and their leachates, which indicates that Pb is primarily derived from solid phases that do not contain significant contents of leachable U and Th. Thus, mineral waters, although CO₂ rich, only interact with surface films on minerals and not with the bulk of the minerals as in the leaching experiments.Calculation of mixing ratios of waters from the granitic and basaltic sources of the waters from the Kyselka area yield about 40% of water from the underlying granite in water recovered from the basalt, whereas the granite-derived water is mixed with only about 5% of the water from the basalt.     

Pb-Pb dating of black shales from the Lower Cambrian and Neoproterozoic strata, South China

Black shales occur widely in the Lower Cambrian and Neoproterozoic strata on the Yangtze Platform, South China. In this study, Lower Cambrian black shales from Xiuning section and Late Neoproterozoic black shales from Weng’an section were studied and Pb isotopic compositions were analyzed following a stepwise acid-leaching technique. The ²⁰⁶Pb/²⁰⁴Pb ratios in both sections show large variations, from 18.906 to 43.737 in the Weng’an section and from 24.811 to 38.110 in the Xiuning section. In contrast, the ranges for ²⁰⁷Pb/²⁰⁴Pb and ²⁰⁸Pb/²⁰⁴Pb values in both sections are relatively smaller from 15.649 to 17.126 and 37.744-38.199 in the Weng’an section, and from 16.034 to 16.783 and 38.602-39.391 in the Xiuning section, respectively. These data yielded two Pb isotope isochron ages of 536±39 and 572±36 Ma, respectively. These ages well accord with other published data and we suggest that they represent the depositional ages for the Lower Cambrian Hetang Formation and the upper Neoproterozoic Doushantuo Formation in South China.     

Weathering process at the natural fission reactor of Bangombé

 The uranium deposits in the basin of Franceville (Gabon) host the only natural fission reactors known in the world. Unique geological conditions favoured a natural fission reaction 2 Ga ago. This was detected by anomalous isotopic compositions of uranium and rare earth elements (REE), which are produced by the fission reaction. In total, 16 reactor zones were found. Most of them are mined out. The reactor zone of Bangombé, is only 10–11 m below the surface. This site has been influenced by surface weathering processes. Six drill cores have been sampled at the site of the reactor zone of Bangombé during the course of the study and only one drill core (BAX 08) hit the core of the reactor. From these data and previous drilling campaigns, the reactor size is estimated to be 10 cm thick, 2–3 m wide and 4–6 m long. The migration of fission products can be traced by the anomalous isotope ratios of REE because of the fission process. The ¹⁴⁹Sm/¹⁴⁷Sm ratio close to the reactor zone is only 0.28 (normal: 0.92) because of the intense neutron capture of ¹⁴⁹Sm and subsequent transmutation, whereas ¹⁴⁷Sm is enriched by the fission reaction. Similar changes in isotopic patterns are detectable on other REE. The isotope ratios of Sm and Nd of whole rock and fracture samples surrounding the reactor indicate that fission-genic REE migrated only a few decimetres above and mainly below the reactor zone. Organic matter (bitumen) seems to act as a trap for fission-genic REE. Additional REE-patterns show less intense weathering with increasing depth in the log profile and support a simple weathering model. Received: 26 November 1999 · Accepted: 2 May 2000     

Pb/Pb geochronology, petrography and chemistry of Zr-rich accessory minerals (zirconolite, tranquillityite and baddeleyite) in mare basalt 10047

In situ U-Pb isotopic measurements were carried out by ion microprobe on the Zr-rich accessory minerals zirconolite [CaZrTi₂O₇], tranquillityite [Fe₈²⁺(ZrY)₂Ti₃Si₃O₂₄] and baddeleyite [ZrO₂] in low-K, high-Ti mare basalt 10047 collected during the Apollo 11 mission. The analysed minerals are concentrated in pockets of late-stage mesostasis that comprises an intergrowth of silica, barian K-feldspar and Si-Al-K glass, from a phaneritic, subophitic, basalt comprising mainly pyroxene, plagioclase, ilmenite, cristobalite and troilite. Most Zr-rich minerals are unaltered, however, some tranquillityite is replaced by a complex intergrowth of zirconolite, baddeleyite, ilmenite and fayalite, suggesting that the mineral became unstable during crystallization. Several baddeleyite crystals have also undergone alteration to secondary zircon. Zirconolite was analysed in thin section 10047,11 and tranquillityite and baddeleyite in 10047,227, using a ∼6 μm primary ion beam. Both zirconolite and tranquillityite have significant U and low initial Pb contents, and are highly suitable for Pb/Pb dating. Fifteen analyses of zirconolite give a ²⁰⁷Pb/²⁰⁶Pb age of 3708 ± 7 Ma (²⁰⁷Pb/²⁰⁶Pb:²⁰⁴Pb/²⁰⁶Pb isochron; 95% confidence, including renormalisation of ratios) and twenty-five analyses of tranquillityite give 3710 ± 6 Ma. The ²⁰⁷Pb/²⁰⁶Pb dates are consistent with each other and refine results from an earlier study. Baddeleyite data were less precise, mainly due to lower secondary ionisation efficiency. Our results show that zirconolite and tranquillityite can provide precise isotopic dates and, given their presence in other samples, they represent important U-Pb chronometers for refining lunar geology.     

Uraninite recrystallization and Pb loss in the Oklo and Bangombé natural fission reactors, Gabon

The Oklo and Bangombé natural fossil fission reactors formed ca. 2 Ga ago in the Franceville basin, Gabon. The response of uraninite in the natural reactors to different geological conditions has implications for the disposal of the UO₂ in spent nuclear fuel. Uraninite and galena from two reactor zones, RZ16 at Oklo and RZB at Bangombé, were studied to clarify the chronology and effect of alteration events on the reactor zones. In addition, ion microprobe U-Pb analysis of zircons from a dolerite dyke in the Oklo deposit were completed to better constrain the age of the dyke, and thereby testing the link between the dyke and an important alteration event in the reactor zones.The analyzed uraninite from RZ16 and RZB contains ca. 6 wt% PbO, indicating a substantial loss of radiogenic Pb. Transmission electron microscopy showed that microscopic uraninite grains in the reactor zones consist of mainly defect-free nanocrystalline to microcrystalline aggregates. However, the nanocrystalline regions have elevated Si contents and lower Pb contents than coarser uraninite crystallites. Single stage model ages of large, millimeter-sized galena grains at both RZ16 and RZB correlate well with the age of the Oklo dolerite dyke, 860 ± 39 Ma (2σ). Thus, the first major Pb loss from uraninite occurred at both Oklo and Bangombé during regional extension and the intrusion of a dyke swarm in the Franceville basin, ∼860-890 Ma ago. Uraninite Pb isotopes from RZ16 and RZB give lower ages of ca. 500 Ma. These ages agree with the “chemical” ages of the uraninite, and show that an ancient Pb loss occurred after the intrusion of the dolerite dykes. The presence of nanocrystallites in the reactor uraninite indicates internal recrystallization, which may have occurred around 500 Ma, resulting in the 6wt% PbO uraninite. It is suggested that leaching by fluid interaction triggered by the Pan-African orogeny was important during this second Pb-loss event. Thus, there are indications that uraninite at both the Oklo and Bangombé natural reactors has experienced at least two ancient episodes of Pb loss associated with internal recrystallization. These recrystallization events have occurred without significantly depleting the 2 Ga fission products compatible with the uraninite structure.     

Provenance ages and alteration histories of shales from the Middle Archean Buhwa greenstone belt, Zimbabwe: Nd and Pb isotopic evidence

Shales of the ca. 3.0 Ga Buhwa Greenstone Belt, Zimbabwe, were derived from a compositionally diverse provenance whose ages, determined by ion probe analyses of detrital zircons in interbedded sandstones, range from 3.8 to 3.1 Ga. Geochemical data for the shales were previously interpreted to indicate that sediments had been derived from an intensely weathered source. REE concentrations in the shales were interpreted to suggest that the provenance was compositionally mixed, with components of felsic (tonalite and alkalic granitoid) and mafic rocks. Sm/Nd and Nd isotopic compositions of these rocks can be used to model initial Nd isotopic ratios at the time of sedimentation (ε_Ndsed), as well as model crustal formation ages (T_DM). The former, at the age of sedimentation, range from +0.6 to −10.8, consistent with a range of provenance ages. The latter range from 4.46 Ga to 2.99 Ga. The oldest crustal formation ages, up to 0.7 Ga older than known detrital components, are interpreted here to indicate that the Sm-Nd system of the sediments experienced open system behavior. The implied alteration would have included an increase in Sm/Nd by about 20-25 percent, probably in the form of preferential loss of Nd with respect to Sm. The Pb isotopic compositions of whole rock samples are quite radiogenic, with a range of ²⁰⁶Pb/²⁰⁴Pb from 25.5 to 154. An array of ten samples lies scattered about a line with a ²⁰⁷Pb/²⁰⁴Pb -²⁰⁶Pb/²⁰⁴Pb slope age of about 2.73 Ga. Five individual samples were sequentially leached to further test the timing and characteristics of this U-Th-Pb alteration event. These arrays of a whole rock, three leach steps, and a residue also form linear Pb-Pb arrays (one is more scattered) with ages ranging from 2260 ± 360 Ma to 2824 ± 170 Ma, suggesting that all samples experienced a latest Archean to earliest Proterozoic enrichment in U/Pb. This age range also may be the approximate age of Sm/Nd enrichment for the shales. All samples, both whole rocks and leached samples, lie grouped on a ²⁰⁸Pb/²⁰⁴Pb - ²⁰⁶Pb/²⁰⁴Pb diagram around a line with ²³²Th/²³⁸U = 3.5 (2.9 to 3.9). Because of the lack of large differences in the Th/U of the samples through large ranges of U/Pb, we interpret this consistency in Th/U to mean that the shales of the Buhwa belt experienced Pb loss, rather than U and Th gain. Circumstances that may be responsible for Pb loss in a sedimentary basin include loss of saline fluids during basin dewatering. Such an event would likely have been related to folding associated with the thrusting and magmatic intrusion of the adjacent Limpopo Belt, suggesting that uplift, dewatering, and geochemical and isotopic alteration can be genetically related.     

Mineralogic investigation into occurrence of high uranium well waters in upstate South Carolina, USA

High levels of U (up to 5570 μg/L) have been discovered in well waters near Simpsonville, South Carolina, USA. In order to characterize the mineralogical source of the U and possible structural controls on its presence, a deep (214 m) well was cored adjacent to one of the enriched wells. The highest gamma-ray emissions in the recovered core occur in coarse biotite granite at a depth just below 52 m. A slickenlined fault plane at 48.6 m and narrow pegmatite layers at depths of 113, 203 and 207 m also yield high gamma-ray counts. Thin sections were made from the above materials and along several subvertical healed fractures. Uraninite and coffinite are the principal U-rich minerals in the core. Other U-bearing minerals include thorite and thorogummite, monazite, zircon and allanite. Primary uraninite occurs in the biotite granite and in pegmatite layers. Secondary coffinite is present as tiny (<5 μm) crystals dispersed along fractures in the granite and pegmatites. Coffinite also occurs along the slickenlined fault plane, where it is associated with calcite and calcic zeolite and also replaces allanite. Coffinite lacks radiogenic Pb, hence is considerably younger than the uraninite.Dissolution of partially oxidized Ca-rich uraninite occurring in the surficial biotite granite (or secondary coffinite in fracture zones) is likely the main source for the current high levels of U in nearby area wells. The high-U well waters have a carbonate signature, consistent with pervasive calcite vein mineralization in the core. Aqueous speciation calculations suggest U transport as an uranyl (U⁶⁺) hydroxyl-carbonate complex. Later reduction resulted in secondary precipitation along fractures as a U⁴⁺ mineral (i.e., coffinite).     

Geochemical characteristics of the Xuanwei Formation in West Guizhou: Significance of sedimentary environment and mineralization

The Upper Permian Xuanwei Formation widely occurs in western Guizhou, unconformably overlying the Emeishan basalts, and mainly consists of black shales. It is ∼170 m thick at Cuyudong Village, Weining County, West Guizhou, China, where the samples of black shale and sandy shale were collected and analyzed. The shales mainly contain SiO₂, 18.9%–44.1%, Al₂O₃, 14.8%–52.8%, Fe₂O₃, 1.0%–41.2%, LOI, 3.2%–21.1%, TiO₂, 1.0%–6.7%, and MgO, 0.2%–2.5%. The contents of all other major elements are lower than 1.0%. It is shown that the black shales have higher contents of Fe₂O₃ and LOI than normal shales. The siderites occurred in the black shales with higher contents of Fe₂O₃, which may be attributed to hydrothermal activities on seafloor. All analyzed shale samples have extremely high Ga, 47.8×10⁻⁶–109.9×10⁻⁶ (70.5×10⁻⁶ on average), higher than the industrial mining standard of Ga Resource Industry Standard. The total contents of rare-earth elements (REE) of 9 black shale samples vary from 213×10⁻⁶ to 1460×10⁻⁶, suggesting that these black shales are enriched in REE. The shale-normalized REE patterns display both positive and negative Ce anomalies (Ce/Ce* from 0.5 to 1.7), revealing that the Xuanwei shales were precipitated under oxic and anoxic conditions. The Rb-Sr chronological diagram of 6 shale samples in the Xuanwei Formation shows an age of 255±12 Ma. Strontium isotopic ratios (⁸⁷Sr/⁸⁶Sr)_t0 range from 0.70635 to 0.70711, suggesting that these Xuanwei black shales might be derived from chemical weathering of the Emeishan basalts.     

Natural fractionation of U/U

The isotopic composition of U in nature is generally assumed to be invariant. Here, we report variations of the ²³⁸U/²³⁵U isotope ratio in natural samples (basalts, granites, seawater, corals, black shales, suboxic sediments, ferromanganese crusts/nodules and BIFs) of ∼1.3‰, exceeding by far the analytical precision of our method (≈0.06‰, 2SD). U isotopes were analyzed with MC-ICP-MS using a mixed ²³⁶U-²³³U isotopic tracer (double spike) to correct for isotope fractionation during sample purification and instrumental mass bias. The largest isotope variations found in our survey are between oxidized and reduced depositional environments, with seawater and suboxic sediments falling in between. Light U isotope compositions (relative to SRM-950a) were observed for manganese crusts from the Atlantic and Pacific oceans, which display δ²³⁸U of −0.54‰ to −0.62‰ and for three of four analyzed Banded Iron Formations, which have δ²³⁸U of −0.89‰, −0.72‰ and −0.70‰, respectively. High δ²³⁸U values are observed for black shales from the Black Sea (unit-I and unit-II) and three Kupferschiefer samples (Germany), which display δ²³⁸U of −0.06‰ to +0.43‰. Also, suboxic sediments have slightly elevated δ²³⁸U (−0.41‰ to −0.16‰) compared to seawater, which has δ²³⁸U of −0.41 ± 0.03‰. Granites define a range of δ²³⁸U between −0.20‰ and −0.46‰, but all analyzed basalts are identical within uncertainties and slightly lighter than seawater (δ²³⁸U = −0.29‰).Our findings imply that U isotope fractionation occurs in both oxic (manganese crusts) and suboxic to euxinic environments with opposite directions. In the first case, we hypothesize that this fractionation results from adsorption of U to ferromanganese oxides, as is the case for Mo and possibly Tl isotopes. In the second case, reduction of soluble U^VI to insoluble U^IV probably results in fractionation toward heavy U isotope compositions relative to seawater. These findings imply that variable ocean redox conditions through geological time should result in variations of the seawater U isotope compositions, which may be recorded in sediments or fossils. Thus, U isotopes might be a promising novel geochemical tracer for paleo-redox conditions and the redox evolution on Earth. The discovery that ²³⁸U/²³⁵U varies in nature also has implications for the precision and accuracy of U-Pb dating. The total observed range in U isotope compositions would produce variations in ²⁰⁷Pb/²⁰⁶Pb ages of young U-bearing minerals of up to 3 Ma, and up to 2 Ma for minerals that are 3 billion years old.     

U-Pb systematics of an Upper Carboniferous black shale from South Yorkshire, UK

Open-system behaviour of uraniferous shales, which has been known for many years, has discouraged attempts to use U-Pb geochronology to date sedimentary systems. Techniques now available can facilitate better understanding of their geochemical evolution and their possible use in geochronometry. For the U-rich Alton (G. listeri) Marine Band, a combined fission track mapping, electron optical and sequential chemical extraction study confirms that uranium is incorporated into francolite, an early diagenetic precipitate. U-Pb analyses of uranium-rich (>1000 ppm) francolite nodules are discordant and imply ages ∼50-150 Ma younger than the date of sedimentation. Pb isotopic analysis suggests that uranium daughters continually leaked from the francolite, ²³⁸U daughters being released more efficiently than those of ²³⁵U. Extrapolation of the U-Pb data to concordia produces an age consistent with the time of sedimentation. These features are also displayed by other uranium-rich shales such as the Swedish Kolm Measures, despite uranium being incorporated into different phases. Preferential loss of ²³⁸U daughters from fine-grained particles due to alpha recoil could explain the unusual U-Pb isotopic composition, in both examples. Further work would be justified to investigate the application of U-Pb isotopic analysis of such material to date sedimentary sequences.     

Fate of trace elements during alteration of uraninite in a hydrothermal vein-type U-deposit from Marshall Pass, Colorado, USA

Alteration of uraninite from a hydrothermal vein-type U-deposit in Marshall Pass, Colorado, has been examined by electron microprobe analysis in order to investigate the release and migration of trace elements W, As, Mo, Zr, Pb, Ba, Ce, Y, Ca, Ti, P, Th, Fe, Si, Al, during alteration, under both reducing and oxidizing conditions. The release of trace elements from uraninite is used to establish constraints on the release of fission product elements from the UO₂ in spent nuclear fuels. Uraninite occurs with two different textures: (1) colloform uraninite and (2) fine-grained uraninite. The colloform uraninite contains 1.04-1.75 wt% of WO₃, 0.16-1.70 wt% of As₂O₃, 0.06-0.88 wt% of MoO₃; whereas, the fine-grained uraninite retains 2.25-4.93 wt% of WO₃, up to 5.76 wt% of MoO₃, and 0.26-0.60 wt% of As₂O₃. The near constant concentration of incompatible W in the colloform uraninite suggests W-incorporation into the uraninite structure or homogeneous distribution of W-rich nano-domains. Incorporation of W and Mo into the uraninite and subsequent precipitation of uranyl phases bearing these elements are critically important to understanding the release and migration of Cs during the corrosion of spent nuclear fuel, as there is a strong affinity of Cs with W and Mo. Zoning in the colloform texture is attributed to variation in the amount of impurities in uraninite. For unaltered zones, the calculated amount of oxygen ranges from 2.08 to 2.32 [apfu, (atom per formula unit)] and defines the stoichiometry as UO_2+x and U₄O₉; whereas, for the altered zones of the colloform texture, the oxygen content is 2.37-2.48 [apfu], which is probably due to the inclusion of secondary uranyl phases, mainly schoepite. The supergene alteration resulted in precipitation of secondary uranyl minerals at the expense of uraninite. Four stages of colloform uraninite alteration are proposed: (i) formation of an oxidized layer at the rim, (ii) corrosion of the oxidized layer, (iii) precipitation of U⁶⁺-phases with well-defined cleavage, and (iv) fracture of the uraninite surface along the cleavage planes of the U⁶⁺-phases.     

U–Th–Pb–REE systematics of organic-rich shales from the ca. 2.15 Ga Sengoma Argillite Formation, Botswana: Evidence for oxidative continental weathering during the Great Oxidation Event

The redox-sensitive geochemical behavior of uranium permits the use of Th/U ratios as a geochemical proxy for the oxidation state of the atmosphere during deposition. Due to the effects of post-depositional uranium mobility on Th/U ratios during events involving oxygenated fluids, direct measurements of Th/U ratios are often misleading even for drill core samples. Because both of these elements radioactively decay and produce lead isotopes, the Pb isotope composition may reflect the depositional Th/U ratio, although the Th/U ratios induced by changes shortly after deposition may not be distinguished from the true depositional Th/U ratios. In order to effectively evaluate the time-integrated Th/U ratio (κ_a), values for the initial depositional Pb isotope composition must be determined or accepted from the models for the whole Earth.While the timing for the rise of atmospheric oxygen is reasonably well constrained now, its effect on continental weathering and ocean redox state remains poorly constrained and debated. The ca. 2.15 Ga Sengoma Argillite Formation of Botswana contains organic-rich shales deposited during the Great Oxidation Event. The slope of the ²⁰⁷Pb/²⁰⁴Pb–²⁰⁶Pb/²⁰⁴Pb array of shales from the Sengoma Argillite Formation corresponds to a Pb–Pb age that is within analytical error of the depositional age and is, therefore, inferred to be the time by which the time-integrated thorogenic and uranogenic lead growth started. The time-integrated lead growth corresponds to an average κ_a of 2.63 (± 0.62, 1σ) for the organic-rich shales of the Sengoma Argillite Formation. This is lower than Th/U ratios measured in Archean shale suites or estimated for the Archean–Proterozoic average upper continental crust [Taylor, S.R. and McLennan, S.M., 1985. The Continental Crust: Its Composition and Evolution. Blackwell, Oxford, 312 pp.], which indicates that these samples were enriched in uranium with respect to thorium (and perhaps lead) at the time of deposition. In the modern ocean, sediments are enriched in uranium under reducing conditions by reduction of the water-soluble uranyl ion, which is delivered to the ocean by oxidative weathering of continental crust. To evaluate the potential mobility of Th, U, and Pb during post-depositional processes, the concentrations of the rare earth elements (REE) were also determined. Interelement ratios of the largely immobile REE (in this study, La/Nd and Gd/Er) can be used as a proxy for the Th/U ratio, as the geochemical behavior of the lanthanide and actinide elements is similar under a variety of conditions. Furthermore, close similarity in the chondrite-normalized REE patterns and small range in La/Nd and Gd/Er ratios in studied samples indicate that variations in κ_a values are not likely to have been controlled by mixing of one or more REE-, Th-, and U-rich heavy minerals from the multiple detrital sources. Our study of shales from the ca. 2.15 Ga Sengoma Argillite Formation indicates that decoupling of U from Th, most likely related to the oxidative continental weathering, began by 2.15 Ga, at the latest.     

Early Permian peak metamorphism recorded in U–Pb system of black slates from the Ogcheon metamorphic belt,South Korea,and its tectonic implication

We report Pb–Pb whole rock and uraninite CHemical Th–U–total Pb Isochron MEthod (CHIME) ages of carbonaceous black slates from the Ogcheon metamorphic belt, South Korea. The Pb isotopic data of whole rock samples yield ²⁰⁷Pb/²⁰⁶Pb ages of 283±33 and 291±13 Ma for two outcrops from the northeastern part of the belt. The uraninite CHIME age is estimated at 283±26 and 281±27 Ma for the northeastern and the middle part of the belt, respectively. All of the above ages are identical within error ranges, and represent the timing of peak metamorphism after the late Precambrian intraplate rifting. On the other hand, the ²⁰⁷Pb/²⁰⁶Pb whole rock age for the southwestern part of the belt is estimated at 194±27 Ma, probably representing the timing of contact thermal metamorphism associated with the intrusion of Jurassic granitic plutons. Rb–Sr isotopic data of the black slates do not define any meaningful isochron. The early Permian metamorphic age of this study does not support any tectonic scheme in favor of major tectonometamorphism at either the Silurian–Devonian or the Triassic time. Instead, it corroborates the probability that the two zones in the Ogcheon belt, the Ogcheon metamorphic belt and the Taebaegsan zone, were separated from each other before the development of major structural framework in the former. Our data do not support an idea that the Ogcheon belt corresponds to the continuation of the Triassic collision belt in east central China.