The Evisa complex is predominantly composed of rare-metal granite of peralkaline affinity. A distinctive attribute of this granite is its notable abundance of rare-earth elements (REE) and, specifically, an exceptional fractionation of the heavy REE (HREE), which took place mostly during magmatic stages. Most of the granite at Evisa is equigranular, miaskitic to agpaitic in composition. The granite contains late-magmatic pockets showing pegmatitic textures. Both the granite and these pegmatites are intersected by granitic to mafic dykes. The granite exhibits pervasive alteration, manifesting as the partial replacement of magmatic minerals such as amphibole, elpidite, zircon, and pyrochlore by hydrothermal phases. Concomitant with this alteration process, there was emplacement of quartz veins and formation of hydraulic breccia. Compared to the granite, pegmatites show similar light REE (LREE) contents, but are enriched in the middle (M)REE and HREE. Felsic dykes are slightly enriched in LREE and more so in MREE-HREE. In the breccia, all REE are two to four times higher than in the granite. Globally, in granite and pegmatite, LREE are mostly hosted in pyrochlore and monazite-(Ce), MREE in pyrochlore and zircon, HREE in zircon, pyrochlore and amphibole. Pyrochlore alteration is archetypal of the REE mobility in the granite, as evidenced by a remarkable exchange of LREE and HREE when igneous LREE-rich pyrochlore was replaced by HREE-rich hydrothermal pyrochlore. In this study, we emphasize how, in peralkaline granites, hydrothermal alteration that followed igneous crystallization led to a chain reaction whereby several generations of REE-bearing minerals precipitated sequentially, replacing the previous one.

The chemistry of detrital apatite has increasingly been employed for provenance studies and to constrain the composition of the parent magma, yet this tool has to date rarely been applied to the topic of Archean crustal evolution. In this study, we present new halogen and trace element compositions analyzed via EPMA and LA-ICP-MS in detrital apatites collected from Paleoarchean metasediments in the Barberton Greenstone Belt (BGB), South Africa, and in magmatic apatites of the surrounding tonalite-trondhjemite-granodiorite (TTG) granitoid-gneisses. Detrital apatites from both the Hooggenoeg (deposited at ∼ 3.43 Ga) and Schapenburg (deposited at ∼ 3.24 Ga) (meta-)sedimentary rocks retained trace element chemistry consistent with a primary igneous origin. They show strong fractionation of their REE spectra (LaN/LuN = 29.4 and 5.43 respectively) and low Eu/Eu* (∼0.5 and 0.8 respectively). On the other hand, the Hooggenoeg apatites have high Sr (600–1200 ppm) and moderate Mn content (400–800 ppm), are F-rich (∼2.2 wt%) and have Cl content ∼ 0.4 wt%; whereas the Schapenburg apatites have uniformly low Sr (∼200 ppm), higher Mn (800–1800 ppm) together with higher F (∼3.2 wt%) and lower Cl (∼0.01 wt%) contents. Strikingly, the characteristics of neither of these populations match with those of apatites from the Barberton TTGs but rather overlap well with apatites crystallized in Neoarchean granitoids, notably sanukitoids, either low-SiO2 ones in the case of the Hooggenoeg apatites, or high-SiO2 for the Schapenburg ones. Therefore, rocks akin to the sanukitoid suite were already present and exposed to surface alteration in the Paleoarchean felsic crust at the time the sediments were deposited. This implies that the felsic crust around the BGB at 3.43–3.24 Ga was more compositionally diverse than the exposed rock record (dominated by TTGs) suggests. In turn, we propose that the onset of observed granitoid diversity in the Neoarchean is an artifact related to a preservation bias, i.e. the higher (upper crustal) emplacement level of K-rich granitoids compared with mid-crustal TTGs made them more prone to erosion before ca. 3 Ga, while the onset of collisional processes at 3.0–2.5 Ga allowed to “shield” these high-level intrusions in the newly amalgamated continental masses and preserve them to our time.

Abstract The Varuträsk deposit (Skellefte District, northern Sweden) is a well-zoned and extremely differentiated granitic pegmatite of the petalite subtype and is of historic significance since it was investigated to determine the first phosphate transformation sequences. New petrographic and geochemical data indicate that minerals of the triphylite–lithiophilite solid solution, as well as wolfeite and montebrasite, are the main primary phosphates crystallizing during the magmatic stage. Iron-manganese phosphates mainly originate from the wall zone of the pegmatite, and their Fetot/(Fetot + Mntot) ratios decrease from 0.627 to 0.186 when approaching the inner intermediate zone. Fractionation trends of triphylite–lithiophilite are consistent with the chemical evolution of closely associated tourmaline-group minerals, confirming that phosphates evolve in equilibrium with the pegmatitic melt and with the coexisting mafic silicates, rather than in a closed system. The occurrence of Al phosphates in the inner pegmatite zones, associated with Li- and Cs-bearing minerals, clearly indicates that the fractionation degree required for the formation of these phosphates is higher than that needed for the crystallization of iron-manganese phosphates. Moreover, montebrasite shows a limited incorporation of fluorine (F < 3.76 wt.%), indicating an unusually low F concentration in the melt (≈1 wt.% F). During the magmatic-hydrothermal stage, a high-temperature aqueous fluid exsolves from the saturated residual melt, initiating the replacement of primary phosphates by high-temperature metasomatic secondary species. The F content of the fluid is lower than that of the melt, as shown by the crystallization of F-poor secondary montebrasite in veins crosscutting the primary montebrasite. Triphylite–lithiophilite remains unaltered or undergoes a partial to complete replacement by the more oxidized species of the ferrisicklerite–sicklerite and heterosite–purpurite series or by alkali-rich varulite during the Na-metasomatic phase. Several substitution mechanisms have been observed in alluaudite-type phosphates, and electron-microprobe analyses indicate the presence of varulite NaNaMn(MnFe3+)(PO4)3, hagendorfite NaNaMn(Fe2+Fe3+)(PO4)3, alluaudite □NaMnFe3+2(PO4)3, and of a new possible oxidized species □□Mn3+Fe3+2(PO4)3. Finally, at lower temperatures, OH and/or H2O-rich phosphates crystallize, and triphylite is replaced by veins of vivianite and fairfieldite under poorly oxidizing conditions. A second stage of low-temperature hydrothermal transformations is characterized by an increasing Ca activity in the aqueous fluid, combined with more oxidizing conditions, thus allowing the crystallization of Ca- and Fe3+-bearing secondary species such as jahnsite s.l., a replacement product of vivianite and fairfieldite. Apatite-group minerals also crystallized during the Ca-metasomatic phase, and their compositions show significant increases of the Cl, Mn, and Sr contents, as well as a decrease of the F content, thus highlighting changes in the fluid composition during the pegmatite evolution.

It is well established that "lepidolites" are mixtures between a dioctahedral pole, i.e., muscovite (Mus) K(Al2)[Si3 AlO10](OH)2, and two trioctahedral poles, i.e., trilithionite (Trl) K(Li1.5Al1.5)[Si3AlO10](F,OH)2 and polylithionite (Pol) K(Li2Al)[Si4O10](F)2. Due to variations in xF = F/(F + OH) in the Trl component, four components must be considered: Mus, F-Trl, OH-Trl, and F-Pol. Based on 22 analyses from the Tanco pegmatite, the mole fractions (xi) of each of the components were calculated by solving a linear matrix using the EPMA Si-Al-F and ICP-MS Li values. The values of x(Mus) range from 0.1 to 0.4 and are perfectly correlated with the values of octahedral vacant sites. The values of x(OH-Trl) reach as high as 0.7 with a large proportion in the 0.35-0.42 range. These results demonstrate that all the Li components are trioctahedral and that the name "Li-muscovite" is not appropriate. This allows for the calculation of fluoride concentrations in aqueous solutions in equilibrium with "lepidolite" using the HCh software by Yuri Shvarov (Lomonosov Moscow State University). The thermodynamic properties for the F-Trl and F-Pol components are known from experimental work, but those for the OH-Trl component remain unknown. For a preliminary approach, these properties were estimated using those of OH-and F-phlogopite as models. The results of the calculations are presented for 1 kbar and T = 400 and 500 degrees C. The solutions in equilibrium with "lepidolite" exhibit a high ionic strength (I), leading to log(mF-total) values ranging from-.-0. 7 to-.-0. 4 at 400 degrees C, and from-.-1. 2 to-.-0. 9 at 500 degrees C. These high F concentrations are consistent with calculations based on the experimental calibration of the geofluorimeter.

The debate about early Earth differentiation focuses on the processes responsible for the formation of protocrust(s) and continental crust of felsic (SiO2 ≥ 55 weight %) composition. One aspect of this debate is how Hadean zircons fit into an ultramafic environment. On the basis of experiments, thermodynamic modeling, and elemental partitioning, we show that felsic melts could have been generated by shallow interaction between primordial serpentinized peridotite and basaltic magmas on Earth and Mars. On the basis of the hafnium isotopic evolution of Hadean detrital zircons worldwide, we infer that these interactions allowed for the formation of extensive Hadean felsic crust (4.4 to 4.5 billion years ago), which, in turn, would account for up to 50% of the present continental crustal mass. A similar process may have occurred on Mars. The serpentinized protocrust had a dual role in the primitive planetary environment: to provide ingredients for the continental crust and to enable life to emerge on water-bearing terrestrial planets.

The main constituent of the planetary lithosphere is the dominant silicate mineral, olivine α-(Mg,Fe)2SiO4, which, along with associated minerals and the olivine-hosted inclusions, records the physical–chemical conditions during the crystal growth and transport to the planetary surface. However, there is a lack of physical–chemical information regarding the kinetic factors that regulate crystal growth during melt–rock, fluid–rock, and magma–rock interactions. Here, we conducted an experimental reaction between hydrated peridotite rock and basaltic melt and coupled this with a structural and elemental analysis of the quenched products by high-resolution transmission electron microscopy. The quenched products revealed crystallographically oriented oxide nanocrystals of ilmenite (Fe,Mg)(Ti,Si)O3 that grew over the newly formed olivine in the boundary layer melt of the reaction zone. We established that the growth mechanism is epitaxial and is common to both experimental and natural systems. The kinetic model developed for shallow (<1 GPa) crystal growth requires open system conditions and the presence of melt or fluid. It implies that the current geodynamic models that consider natural ilmenite–olivine assemblage as a proxy for deep to ultra-deep (>>1 GPa) conditions should be revised. The resulting kinetic model has a wide range of geological implications—from disequilibrium mineral growth and olivine-hosted inclusion production to mantle metasomatism—and helps to clarify how geological reactions proceed at depth.

Understanding the behavior of trace elements such as Zr-Hf, REE, and Th-U during magmatic and hydrothermal processes is crucial for unraveling the evolution of rare-metal peralkaline granites. These elements play key roles in geochemical cycles, mineral formation, and economic mineral deposits. The Evisa intrusion, a rare-metal peralkaline granite, is an ideal setting to study such processes due to its diverse zircon population, which records a continuous transition from early igneous to late hydrothermal precipitation and alteration. This evolution spans temperatures from more than 600 degrees C to approximately 150 degrees C, with zircon demonstrating exceptional diversity in textures and compositions. In-situ U-Pb dating of both igneous and hydrothermal zircon suggests they precipitated coevally at similar to 270 Ma, indicating the synchronous formation and alteration of zircon during the crystallization and cooling of the Evisa granite. Hf isotopic compositions vary from 0 to + 10 epsilon Hf(t) for igneous zircon and from -5 to + 5 epsilon Hf(t) for hydrothermal zircon, reflecting a shift from a mantle-derived to a more crustal signature. Both U-Pb and Hf isotopic data support a rift-related context for the Corsica terrane during the late Permian. Zircon in this granite is highly enriched in REE (median value > 2 wt% for Y + REE) and variably enriched in U and Th (up to several wt.%), with igneous zircon exhibiting higher Ce anomalies and HREE/LREE ratios than hydrothermal zircon. Alteration of zircon results in the formation of pores (ranging from nm to mu m scale) and the redistribution of trace metals between newly-formed mineral inclusions, newly-crystallized zircon, and the aqueous fluid. This redistribution, combined with the breakdown of other REE-minerals and availability of suitable ligands in the fluid, controlled the cyclic enrichment of REE in hydrothermal zircon and influenced their fate in the Evisa granite. Hydrothermal remobilization of Hf at low temperature led to significant Hf isotopic variations (up to > 20 epsilon Hf(t) unit) in newly crystallized zircon, likely due to mass-dependent kinetic fractionation. Discrepancies in the Zr/Hf ratio (from similar to 0.5 to 2.5) in successive growth zones of hydrothermal zircon were interpreted as resulting from fractionation due to mass-independent effects on bond strength during transport, probably in the form of fluoride complexes. This study not only provides new insights into the behavior of Zr-Hf, Th-U and REE in rare-metal peralkaline granites but also underscores the need for caution when interpreting zircon compositions, particularly when using commonly employed geochemical tools, such as Ti-based thermometers.

Plusieurs filons d'albitite à corindon, riche en minéraux accessoires, recoupent les lherzolites de trois massifs pyrénéens. Ces roches magmatiques seraient le produit d’une fusion partielle de faible degré d'une source harzburgitique préalablement enrichie par métasomatose carbonatitique. Des cristaux de corindon, vivement colorés en bleu, ont suscité le lancement d'un atelier de lapidaire par un amateur motivé qui livre son expérience.

Nodular chromite ore deposits are found in ophiolites and crop out in structures interpreted as former dykes. Most of them are transposed parallel to the plastic foliation of the host peridotite. Many studies have been conducted in order to decipher the origin and evolution of nodular chromite but the outstanding lack of consensus paves the way for an integrated field, geochemical and microstructural approach to be carried out. We sampled the well-characterized Maqsad chromitite dyke that crops out at the top of the mantle-crust dunitic transition zone in the Oman ophiolite and that was not affected by transposition. The spectacular variations in nodule size and texture and their distribution within the dyke have been perfectly preserved which is a rather unique situation. We selected about 40 nodules representative of the shapes and size variability of this ore deposit. Nodules have been classified in three categories: Type-1 nodules are large skeletal chromite grains associated with amphibole and olivine filling their former porosity; Type-3 nodules have a central nucleus of chromite/silicate surrounded by a mantle of close-packed chromite grains. Their shape is best described as almond-like, and they may reach 3 cm in length; Type-2 includes all of the intermediate nodules shapes and sizes between type-1 and type-3 varieties. Electron Probe MicroAnalyser (EPMA) transects and maps show that mineral chemical variations in type-1 nodules and in the nuclei of type-2 and type-3 nodules record out of equilibrium crystal growth. They exhibit high XCr and relatively low-TiO2 and likely resulted from transient interactions between hydrothermal fluids and basaltic melts. Contrary to type-1 and nuclei, the mantles of type-2 and -3 nodules have lower Cr2O3 contents, decreasing toward the nodule edges. They are richer in TiO2 than type-1 nodules but concentration patterns of this element along edge to edge transects are quite variable. The chromite mantles of type-2 and -3 are best understood as the fractional crystallization products of a parent melt of type-1 nodules in different pFluids and/or redox conditions. The analysis of the distribution of the misorientation axis between the skeletal grains and the adjacent chromite grains in the mantle (N = 222 in 28 nodules) revealed a clustering around a [111] axis across the whole range of misorientation angles. Accordingly, we suggest that the growth of the Maqsad nodules is achieved by accretion of finer euhedral chromite grains onto a skeletal chromite grain by juxtaposition of their flat crystal facets. Combined EDS-EBSD mapping together with EPMA transects revealed that a similar to 1-mm-thick rim of high XCr, displaying a homothetic shape to its corresponding nuclei, is a common feature in the mantle when the nodule sizes exceed about 1 cm. We interpret this observation as the result of the accretion of grains with higher XCr compositions during the nodule construction. Overall, our new study of the Maqsad nodular chromitite highlights the peculiar and transient conditions needed to give rise to the textural and chemical complexity that is preserved in these enigmatic rocks.

Le gisement de talc-chlorites de Trimouns résulte de l’altération métasomatique de métasédiments paléozoïques, par des fluides aqueux et salés. Ce processus s’effectue entre 250 et 300 ºC à une profondeur inférieure à 5 km. Daté entre 120 et 90 millions d’années, il est lié à l’amincissement de la croûte continentale lors de l'hyper-extension provoquée par l’ouverture du Golfe de Gascogne.

How did felsic crust form and segregate from the primordial olivine-rich mantle?Based on experiments on serpentinite-basalt interaction, thermodynamic modeling, and trace element partitioning, we hypothesize that felsic material that subsequently segregated as the Hadean crust could have been generated due to shallow (< 0.5 GPa) interaction of the primordial ultramafic hydrated proto-crust with basaltic melts [1, 2].We performed a new series of experiments on variable ratios of serpentinite-to-basalt at 0.1 GPa, 1100°C and 5 days duration.An optimal serpentinite percentage favorable for production of felsic (62 -69 wt% SiO 2 ) melts varies between 50 and 80 wt% (Fig. 1).The melts coexist with high-Mg olivine-rich harzburgitic, dunitic or wehrlitic residue, depending on the starting serpentinite-to-basalt ratio.Subsequent segregation of the melts from the ultramafic residue and primordial protocrust is able to produce 20 -40 % of felsic crust at the beginning of the Hadean eon.Values of Lu/Hf = 0.16 and, thus, 176 Lu/ 177 Hf = 0.022 for the experimental felsic glasses plotted in the ɛHf(T) vs. age (Ma) of Figure 2 are consistent with data from Hadean detrital zircons worldwide [3][4][5][6][7].Extensive melting could be possible at the beginning of the Hadean, where the basaltic melts from the magma ocean infilled fractures in the quenched ultramafic protocrust.Renewed melting could still occur by a similar process due to low-velocity impacts during the whole Hadean.Such a mechanism of felsic crust formation that could start at 4.5 Ga required prior alteration of the early-solidified magma ocean by liquid water and may have been also effective on Mars.

Active felsic magmatism has been rarely probed in situ by drilling but one recent exception is quenched rhyolite sampled during the 2009 Iceland Deep Drilling Project (IDDP). We report finding of rare zircons of up to ∼100 µm in size in rhyolite glasses from the IDDP-1 well products and the host 1724 AD Viti granophyres. The applied SHRIMP U-Th dating for both the IDDP and the Viti granophyre zircons gives zero-age (±2 kyr), and therefore suggests that the IDDP-1 zircons have crystallized from an active magma intrusion rather than due to the 20–80 ka post-caldera magmatic episodes recorded by nearby domes and ridges. Ti-in-zircon geothermometer for Viti granophyre reveals zircon crystallization temperatures ∼800°C–900°C, whereas IDDP-1 rhyolite zircon cores show Ti content higher than 100 ppm, corresponding to temperatures up to ∼1,100°C according to the Ti-in-zircon thermometer. According to our thermochemical model at such elevated temperatures as 1,100°C, rhyolitic magma cannot be saturated with zircon and zircon crystallization is not possible. We explain this controversy by either kinetic effects or non-ideal Ti incorporation into growing zircons at low pressures that start to grow from nucleus at temperatures ∼930°C. High temperatures recorded by IDDP-1 zircon together with an occurrence of baddeleyite require that the rhyolite magma formed by partial melting of the host granophyre due to basaltic magma intrusion. Zr concentration profiles in glass around zircons are flat, suggesting residence in rhyolitic melt for >4 years. In our thermochemical modeling, three scenarios are considered. The host felsite rocks are intruded by: 1) a basaltic sill, 2) rhyolite magma 3) rhyolite sill connected to a deeper magmatic system. Based on the solution of the heat conduction equation accounting for the release of latent heat and effective thermal conductivity, these data confirm that the rhyolite magma could be produced by felsic crust melting as a result of injection of a basaltic or rhyolite sill during the Krafla Fires eruption (1975 AD).

Understanding radionuclides mass transfer mechanisms in monazite (LREEPO4) and the resulting features, from the micro- to the nanoscale, is critical to its use as a robust U–Th–Pb geochronometer. A detailed multiscale characterisation of discordant monazite grains from a granulite which records a polymetamorphic history explores the mechanisms of Th and Pb mobility in crystals. Some monazite grains display Th-rich linear features (0.1–1 µm thick) forming a regular network throughout the grain. They are interpreted as resulting from fluid ingress along crystallographically controlled pathways. Nanoscale features termed ‘clusters’ (Ø < 10 nm) are composed of radiogenic Pb (Pb*) ± Si ± Ca and are localised within monazite lattice defects. Their formation results from the competition, over millions of years, of both radiation damage production allowing element mobility (by diffusion) and accumulation in defects and α-healing inducing their trapping. Nanophases (Ø = 0.02–1 µm) containing Pb* are present in all grains and correspond to galena (PbS) or sesquioxide of Pb (Pb2O3). They are associated with a chemically varied suite of amorphous silicate (± Al, Mg, Fe) phases or sulphur (e.g. FeS). They are interpreted as precipitates within monazite crystals. They are formed during replacement mechanism of monazite through fluid interactions. Two generations of Pb*-bearing nanophases exist supported by previous geochronological data. The shielding effect of garnet and rutilated quartz (host minerals), limiting fluid access, induces plentiful Pb*-bearing nanophases precipitation (fluid saturation enhanced) and limits Pb*-loss at the grain scale. This multiscale study provides new insights for interpretations of meaningless geochronological information, thanks to nanoscale investigations.

The Perseverance rover is exploring Jezero crater on Mars since February 2021. SuperCam instrument is part of the scientific payload, combining five different techniques in order to characterize the targets to sample: LIBS (Laser-Induced Breakdown Spectroscopy), Raman and Visible and Infrared (VISIR) spectroscopies, Sound Recording and a colored Imager. Thirty-six calibration targets have been developed for this instrument. Twenty-three of these calibration targets are dedicated to the LIBS technique, corresponding to rocks and/or minerals. Raman and VISIR spectroscopy have their own calibration targets but may also use the LIBS mineral targets as reference on Mars to test mineral detections. The choice and the fabrication of these calibration targets have been made in order to fulfill the science intents of the mission, as well as the technical and science intents of each of the SuperCam technique. These calibration targets have been spectrally characterized and they have shown to be chemically homogeneous at the SuperCam LIBS scale. Their elementary compositions are given, using two different quantitative methods. The composition of the calibration targets will be used as reference for future assessment of the quantitative capability of the SuperCam LIBS technique.

In the Echassieres district of the French Massif Central, occur several outstanding magmatic/hydrothermal systems enriched in strategic metals, such as the Beauvoir rare-metal granite. In this contribution we propose a systematic approach, based on mica trace chemistry, to decipher the different events leading to mineralization. Twelve groups of micas were defined by their specific petrographic features and/or location in the district. Their trace element composition, obtained by LA-ICP-MS, varies widely from one mica group to another, although homogeneous signatures within groups could be distinguished. Some of the trace elements are remarkably enriched, such as W in igneous lepidolite and Sn in greisen muscovite, both of which occur in the Beauvoir granite. A statistical approach based on a set of multivariate analyses highlights that the trace chemistry of micas is inherited from their source, whether hydrothermal or igneous, thus providing a signature for their origin. This approach also shows that differences in major element composition (i.e., different mica species) impact only slightly the trace-element signature. For instance, muscovite and zinnwaldite from one granite have a coherent signature, but they contrast with same mica species in another granite or hydrothermal veins. It is thus possible to genetically link two different mica species from remote locations, or inversely, to recognize different origins for a same mica species in the same sample (e.g., superposed alterations). A second, important implication is that trace-element signatures of micas provide a record of metal remobilization and transport. In the Echassieres district, greisen alteration of the Beauvoir granite caused dissolution of W-rich (ca. 290 ppm in average) lepidolite and cassiterite (SnO2). Newly-formed greisen muscovite incorporated most remobilized Sn (ca. 1000 ppm in average), while W precipitated in distal quartz veins as wolframite. As a consequence, Sn is concentrated in the granite, while W occurs outside of it. This is also underlined by the gradual Sn decrease and W increase recorded in micas from distal veins. Finally, wolframite-bearing veins do not contain cassiterite, validating mica trace-element chemistry as a powerful tool to decipher Sn-W ore forming hydrothermal processes.

The ophiolite of Sivas (Turkey) was studied in order to define the chronology of different alteration events related to a series of serpentinization and carbonation episodes. Six samples were investigated, representative of different types of ophicalcite (partially carbonated serpentinite). X-ray diffraction (XRD) and Mössbauer spectroscopy were used to determine the bulk mineralogy and the bulk Fe3+/Fetot ratio, respectively. Electron microprobe and secondary ion mass spectrometer (SIMS) analyses were also conducted to identify the chemical composition of different mineral phases in addition to the carbon and oxygen isotopic compositions of calcite. An initial, i.e. pre-obduction, phase of olivine and pyroxene serpentinization was followed by a brecciation event associated with precipitation of massive serpentine. This first alteration event occurred during exhumation of the peridotites to the seafloor, followed by a carbonation event at temperatures in the range 35‒100 °C. A low-temperature (∼35 °C) carbonation event occurred between 90 and 65 Ma. Finally, a reheating of the system likely occurred after the obduction at 55‒40 Ma, resulting in a carbonation episode followed by late serpentinization. Our study presents the first direct evidence of serpentinization after obduction. In that geological context, the hydrogen produced during the proposed multiphase serpentinization may have been trapped by the salt deposits overlying the ophiolite but subsurface data will be necessary to define potential traps and reservoirs; further studies are also needed to determine whether the serpentinization process is still ongoing.

The Nkob talc deposit located in the central Anti-Atlas Pan-African belt is hosted by magnesian and siliceous metacarbonates in the contact aureole of the Ediacaran Amassine granite. It consists of a stratified succession of green serpentine-rich marbles, black dolomitic marbles, talcitites and phlogopitites (variously retrogressed into chloritites) enclosed within andalusite metapelite hornfels and quartzites. Two main metamorphic stages have been defined for the formation of the marbles and talcitites: (1) high-temperature recrystallization occurred under amphibolite facies conditions ( 500 degrees C). Temperature increase during granite emplacement led to the formation of forsterite, tremolite, phlogopite (+/- diopside) bearing assemblages in siliceous marbles and to transformation of dolomitic (calcite-free) marbles from pure dolostones. This high temperature stage is also characterized by the growth of andalusite porphyroblasts in the metapelites surrounding the deposit. (2) A low temperature retrograde hydrothermal stage characterized by the reactive infiltration of aqueous silica-rich fluid (presumably derived from the granitic body), in greenschist facies conditions (<520 degrees C). Hydrous silicate minerals, mostly serpentine and talc (+calcite), were formed during this stage within the rocks (after fluid infiltration along grain boundaries) or within veins (marking channeled fluid circulation). Talc in ores formed by two processes at Nkob: acicular talc near the granite formed after the breakdown of tremolite in a temperature range of 350 to 500 degrees C, while tabular talc in the external aureole formed via reaction between dolomite and siliceous-aqueous fluids in temperatures below 350 degrees C. The different types of marbles derive from pure to slightly siliceous dolomite but green and green-black marbles were strongly affected by interactions with siliceousaqueous fluids forming serpentine-phlogopite veins with calcite reaction zones. The talcitite layers were also affected by similar hydrothermal H2O-Si enrichment but their dolomitic precursor probably contained a more important detrital silicate fraction.