Michel Cuney

Crustal generation of the Himalayan leucogranites

Abstract Detailed studies of the Himalayan two-mica leucogranites, such as the Manaslu pluton, indicate that they have very uniform mineralogical, petrological and structural characteristics. One can relate their occurrence to the thickest zones of the underlying Tibetan Slab. In these zones, migmatization attains its greatest development and vertical extension. The granite is emplaced at first along the main disharmonic plane above the Main Central Thrust (MCT), at the top of the Tibetan Slab (infrastructure). Ductile deformation of the granite is variable; the granite being syn-to late-kinematic with regard to the functioning of the MCT. Major elements are very homogeneous (except for Na and K) implying that P-T conditions of melting were relatively uniform. The melted material was of a similar composition over a vast volume, and the percentage of melting was small (10–15%). Trace elements are highly variable. Some are characteristic for very evolved material (Ta, Rb, Cs, U) or show the link with the Tibetan Slab (Ba-Sr), whilst others are problematic (Th, REE). REE and Th abundances, being much less in the granite than in the Tibetan Slab, imply that they have been extracted during one of the main stages of formation, possibly by monazite. Radiogenic (Pb, Sr, Nd) and stable (O) isotopes are consistent with the origin of the granite from the Tibetan Slab. However, the heterogeneous Sr isotopic ratios make age dating difficult and imply poor mixing or little fluid interaction during its evolution. H-isotope data indicate that magmatic compositions of the main body of the Manaslu granite have been preserved. Late or post-magmatic alterations are extremely local in the main pluton. To the north, another belt of two-mica granites occurs whose characteristics are very similar to the High Himalaya belt. They were probably generated in a similar way during this recent intracontinental evolution.

Submagmatic microfractures in granites

Microstructures in Hercynian granite plutons of the Pyrenees indicate fracturing of plagioclase crystals in the presence of melt. This is represented mainly by quartz and/or feldspar that fills fractures. Electron-microprobe analyses demonstrate that, in compositional agreement with the rims of the zoned plagioclase, the feldspar infillings are more sodic with decreasing melt fraction. Fracturing resulted from concentration of stress at contacts between grains while the granite was submagmatic, i.e., when the melt was below the critical fraction for magmatic flow. Material compositionally far from eutectic is petrologically the best candictate to develop submagmatic fractures, because during cooling the volume fraction of melt remains low for a large temperature range. A strong tendency for microfractures to be oblique to flow foliation may constitute a new criterion for shear sense in magmas.

Meta-igneous origin of Hercynian peraluminous granites in N.W. French Massif Central: implications for crustal history reconstructions

Seventy samples of Hercynian peraluminous granites (Guéret, Millevaches and Saint Sylvestre massifs) and metamorphic units of the Limousin area were analysed for Rb−Sr and Sm−Nd. The source rocks of the peraluminous granites can be found in the metamorphic rocks of Limousin, among them meta-igneous rocks were largely predominent over meta-sedimentary rocks in the source of the three granites. Millevaches and Guéret granites were generated by the partial melting of rocks comprising meta-volcanics and meta-sediments, whereas the Saint Sylvestre granite was produced exclusively by the melting of late Precambrian granites. This leads to confusing TDMNdvalues, the confusion being amplified by the segregation of monazite during the petrogenetic evolution of the peraluminous granites, which leads to dramatic fractionation in Sm/Nd ratios. The data of the present study tend to demonstrate that peraluminous granites do not give a good representation of isotopic mean crustal estimates. Late Precambrian time seems, however, to have been a period of extensive crustal generation in Western Europe.

Alteration of monazite and zircon and lead migration as geochemical tracers of fluid paleocirculations around the Oklo–Okélobondo and Bangombé natural nuclear reaction zones (Franceville basin, Gabon)

Abstract Large-scale light rare earth element (LREE), uranium, lead and phosphorus migration has been evidenced in the FA Lower Proterozoic sandstones of the Franceville basin (Gabon) hosting Oklo natural nuclear reaction zones (RZ) in relation with extensive accessory mineral alteration by highly saline diagenetic brines (28.7 wt.% NaCl eq. to 30 wt.% CaCl2 eq.) at about 140°C and 1 kbar. Monazite is the most severely altered accessory mineral in the coarse-grained sandstones of the basal FA formation. Detrital monazite crystals are altered to Th–OH silicate microcrystalline phase with very low concentrations of U and LREE. The Th/La ratio increase from non-altered (Th/La∼0.27) to altered sandstones (Th/La∼1.14) shows that about 76% of the LREE was leached. This corresponds to the leaching of 2.01×109 metric tons at the scale of the FA formation in the Franceville basin. Similarly, the Th/U increase from monazite (Th/U=18.6) to the Th-silicate phase (Th/U=88.7) is interpreted as a result of an alteration by oxidizing brines with leaching of U together with LREE and P. It corresponds to the leaching of 9.06×106 metric tons of uranium. This amount of uranium largely exceeds the known uranium reserves from the Franceville basin. In zircon crystals, the cores are generally homogeneous, weakly fractured and well preserved as attested by the Archean ages (2867±24 and 2865±51 Ma) obtained by ionic microprobe analysis on zircon of the FA Formation, respectively, from the marginal and central parts of the basin. Their composition corresponds to the pure end-member (Zr,Hf)(SiO4), poor in Th and U (Th/U∼1). At the contrary, their rims, which present several growth zones with cracks fillings, are enriched in REE, P, Th and U with higher Th/U ratios (5–10). Both altered monazite and altered zircon contain galena as numerous inclusions in the outer growth zones and as crack fillings. For example, in zircon, the Pb of galena crystals (3–23 wt.%) largely exceeds the amount of Pb (maximum 0.1 wt.%) that would have been produced in situ by radioactive decay in this mineral. Nearly all the lead were introduced into altered zones of accessories. Dissolution of accessory minerals occurred at 2000 Ma, producing a porous and distorted crystal structure which has allowed a later incorporation of Pb. Galena inclusions in altered zircons located in the vicinity of reactor zones have radiogenic lead compositions. Altered zircon rims and galena inclusions in altered zircon located far from reactor zones have non-radiogenic Pb isotopic compositions, confirming the external origin of lead. Pb isotopic evolution models indicate a crystallization age sometime after 1000 Ma, both for galena located close to and far from U mineralizations and reactor zones, which may be synchronous with a regional extension event contemporaneous with intrusion of dolerite dyke swarms, between 1000 and 750 Ma, at the scale of the Franceville basin. The present study also illustrates the different retention capacities of accessory mineral for elements representing analogs of the radiotoxic nuclides in the relatively extreme natural conditions created by the circulation of moderately hot and chloride-rich fluids during the diagenesis of a sedimentary basin.

Uranium in granitic magmas: Part 2. Experimental determination of uranium solubility and fluid-melt partition coefficients in the uranium oxide-haplogranite-H2O-NaX (X = Cl, F) system at 770°C, 2 kbar

The solubility of uranium oxide was investigated in both aqueous halide (Cl, F) fluid and granitic melt in equilibrium in the system uranium oxide-haplogranite-H2O-NaCl (0.1–5.0 molal), NaF (0.1–0.5 molal) at 770°C, 2 kbar, and fO2 conditions controlled by Ni-NiO, Fe3O4-Fe2O3, and Cu2O- CuO buffers. Three distinct uranium oxides UO(2+x) with x = 0.01 ± 0.01; 0.12 ± 0.02; and 0.28 ± 0.02, respec- tively, were obtained in both chloride and fluoride systems, under the three fO2 conditions cited above. Changes in the composition of aqueous solutions and silicate melt were observed after the runs. These changes were more pronounced for the fluoride-bearing experiments. Quench pH decreased from 5.9 to 2.1 with increasing chloride molality from 0.085–4.38 molal. For fluoride solutions, the decrease of pH from 5.4 to 3.4 corresponded to the increase of fluoride molality from 0.02–0.23 molal. The U solubility in chloride solutions was in the range 10–967 ppm. For the same molality, fluoride solutions appeared to dissolve up to twenty times more uranium than chloride solutions. The increase of halide molality and oxidation led to increase the U solubility. The U solubility in silicate glasses was in the range 10-1.8 × 104 ppm and increased with increasing oxidation and halide concentration. In addition, increasing agpaicity also increased U solubility in the chloride system. This effect was not observed in the fluoride system. The chloride concentration in the silicate melt increased from 100-790 ppm with increasing initial aqueous chloride concentration from 0.1–5.0 m. The fluoride concentration in the silicate melt increased from 2.8 × 103 to 1.1 × 104 ppm with increasing initial fluoride concentra- tion from 0.1–0.5 m. In the chloride system, the partition coefficient of U (log D)(U)fluid/melt) increased from -1.2–0 with increasing agpaicity from 0.92–1.36, for increasing chloride concentration from 0.085-4.38 molal and for increasing fO2 from 10−15 to 10−4 bar. In the fluoride system, a linear correlation was established between the partition coefficient of U and the log fO2. In F-rich system, D(U)fluid/melt values was in the range 2.4 × 10−2-4.2 × 10−2 for increasing fluoride concentration from 0.02–0.22 molal and for the same increasing of fO2. In the chloride system, the partition coefficients of Na (D (Na)fluid/melt) and K (D) (K)fluid/melt) are in good agreement up to 1.0 m NaCl with the two linear equations established by Holland (1972) : D (Na)fluid/melt = 0.46 × (Cl)(m) (1) and D(Na)fluid/melt = 0.34 × (Cl)(m) (2). However, in initial 5.0 m NaCl, slopes of Eqns. 1 and 2 decreased to 0.41 and 0.16, respectively. Data obtained in the present study provide useful information for the understanding of the behaviour of U in the fractionation processes of halide rich magmas. Fluid/melt partition coefficients higher than one, favorable for the genesis of magmatic U mineralization, can be reached for peraluminous leucogran- ites in equilibrium with chloride-rich solutions.

U-Pb, Rb-Sr and Sm-Nd chronology of granitic basement, hydrothermal albitites and uranium mineralization (Lagoa Real, South-Bahia, Brazil)

The granites orthogneisses, hydrothermal albitities and rocks which have suffered uranium mineralization from the Lagoa Real District (South State of Bahia, Brazil) have been investigated by U-Pb, Rb-Sr and Sm-Nd techniques. U-Pb values on zircons from the granitic protolith give an age of 1725 Ma; U-Pb on U-mineralization dates the primary mineralization at 1395 Ma and indicates a reworking at 480 Ma, which may represent the age of the thrusting of the Lagoa Real complex over Espinhaço metasediments during the Brazilian orogeny. These two dates are given by Rb-Sr on albitites, but from sparse information and are not supported by unequivocal arguments. The uranium deposition and sodium metasomatism, however, cannot be linked either with the thermal history of the granite or with Brazilian thrusting. Sm-Nd gives scattered results which are suggestive of autochtonous reworkings of REEs. These results lead to the following suggestions (1) magmatic activity of subalkaline affinity existed within the Sao Francisco craton at about 1.7 Ga (2) such subalkaline plutons are likely sources for U-mineralization (3) at about 1.4 Ga an unknown event caused hydrothermal activity leading to U-deposition at Lagoa Real (4) Brazilian overthrusting at about 480 Ma did not play any genetic role in the hydrothermal activity.

A detailed fluid inclusion study in silicified breccias from the Kombolgie sandstones (Northern Territory, Australia): inferences for the genesis of middle-Proterozoic unconformity-type uranium deposits

The relative chronology and detailed chemistry of paleofluids circulating at the base of the Kombolgie Sub-basin were investigated in the East Alligator River district (Northern Territory, Australia), where world-class unconformity-type uranium deposits are located. The chemistry of fluid inclusions was determined using in-situ analysis (Raman microprobe and laser-induced breakdown spectroscopy [LIBS]) and by observing the melting sequences by microthermometry. This study revealed the occurrence of three distinct fluids: (i) a sodium-rich brine that corresponds to a diagenetic fluid percolating at the bottom of the Kombolgie sandstones at a temperature close to 150±15 °C; (ii) a calcium-rich brine, probably corresponding to a residual brine in evaporitic environment that has evolved by fluid–rock interactions with the basement lithologies; and (iii) a low salinity fluid, heated in the basement, injected into the base of the sandstone cover. H2 and O2 and/or traces of CH4 were detected in the vapor phase of some fluid inclusions, especially in the low salinity ones in quartz breccia samples taken above mineralized areas. Hydraulic brecciation of the sandstone was associated with a pressure decrease favoring fluid mixing and the subsequent cementation of breccias. According to the fluid inclusion study and other geologic constrains, the minimum thickness of the Sub-Kombolgie Basin is estimated at 4 km. Drusy quartz breccias with evidence of fluid mixing are quite common at the base of the Kombolgie Basin, but not necessarily linked to U-mineralization. However, it is proposed that the presence of gases such as H2 and O2 in fluid inclusions, which results from water radiolysis, constitutes an indicator of gas linked to significant U concentrations deeper in the basement rocks.

Significance of aluminum phosphate-sulfate minerals associated with U unconformity-type deposits: The Athabasca basin, Canada

Abstract Aluminum phosphate-sulfate (APS) minerals formed around the Athabasca unconformity-type deposits and those from their Australian counterparts are chemically very similar showing the same continuum between the diagenetic Sr-rich APS minerals of the barren sandstones and the LREE-rich composition of the APS minerals in the hydrothermally altered sandstone. The P- and LREE-rich compositions were controlled by the transport and the redistribution of P and LREE elements released from the dissolution of phosphate minerals (principally monazite) in the basement rocks and in the basin during the syn-ore alteration processes. The S/Sr ratio measured in the APS minerals from unaltered sandstone away from the unconformity and any mineralization is preserved during the syn-ore alteration processes suggesting that the fluids involved in both the deep burial diagenetic processes and the syn-ore alteration system were derived from a similar diagenetic reservoir in both the Athabasca and Kombolgie regions. The trioctahedral chlorite host-rock alteration around the Australian basement-hosted U deposits, as compared to the illite and sudoite associated with the Athabasca basement-hosted, along with the more LREE-rich APS compositions in the Australian deposits, suggests that the pH and oxygen fugacity (fO₂) of the syn-ore fluids differed in the alteration systems of the two regions at the time of the U deposition.

Experimental study of Th-bearing LaPO 4 (780 degrees C, 200 MPa); implications for monazite and actinide orthophosphate stability

Abstract A complete solid solution has been hydrothermally synthesized between the two end- members LaPO4 and (Ca05Th0.5)PO4 at 780 ℃ and 200 MPa, indicating that there is no limitation in temperature and pressure conditions corresponding to those of granitic magmas for Th insertion in natural monazites. The composition limits of the (A3+1-2xB2+xC4+x)PO4 compounds crystallized in the monazite structure-type are determined by both raverage = (1 - 2x) [9]rA³+ + x [9]rB²+ + x [9]rC⁴+ and rratio = (1 - x)[9]rA³+ + [9]rB²+/(1-x)[9]rA³+ + x [9]rC⁴+ parameters (where [9]rA is the ionic radius of the A element in ninefold coordination). The upper and lower values of these parameters are 1.216 Å ≥ raverage ≥ 1.107 Å and 1.238 ≥ rratio ≥ 1. The incorporation of large amounts of trans-uranium elements in the monazite structure is deduced from this model. The limitations and geochronological inferences of this model are discussed.

Metallogenesis in the French part of the Variscan orogen. Part I: U preconcentrations in pre-Variscan and Variscan formations — a comparison with Sn, W and Au

Abstract In the French part of the Variscan orogen, the pre-Variscan metamorphic rocks and the Variscan granites resulting from pure crustal recycling are enriched in uranium and tin. Data on tungsten and gold are too scarce for a reasonable estimation of the average concentrations of these elements to be obtained. The spatial distribution of uranium preconcentrations is transverse to the main Variscan structural domains, and the distribution of the Sn-W mineralizations prompts the same conclusions. U, Sn, W and Au metal enrichment in crustal domains in Europe seems to result from late Proterozoic to early Palaeozoic basic and acid magmatism. For uranium, at least two main stages of partial melting, early Palaeozoic and Late Carboniferous, lead to enrichment of the metal in peraluminous leucogranites; these leucogranites represent the main source of late Variscan uranium and tin. Within the peraluminous granitic complexes, the main metal source is related to the late intrusion of small granitic cupolas strongly enriched in metals. In the case of uranium the efficiency of the hydrothermal remobilization depends on the proportion of metal located in uraninite, which is easily leachable by hydrothermal solutions. The emplacement of these cupolas is structurally controlled in well-defined areas by large shear zones. These shear zones are continuously active during the successive emplacement of the granitic magmas. Most of the hydrothermal uranium mineralization is related to the Permian reactivation of the shear zones in a brittle stage. These zones then channel the hydrothermal fluids involved in the last metal concentration step, leading to the formation of the orebodies. The efficiency of metal remobilization is dependent on a close spatial relationship between magmatic structures enriched in uranium and brittle structures channeling the hydrothermal fluids.