Low Temperature Serpentinite Replacement by Carbonates during Seawater Influx in the Newfoundland Margin

: Serpentinite replacement by carbonates in the seafloor is one of the main carbonation processes in nature providing insights into the mechanisms of CO 2 sequestration; however, the onset of this process and the conditions for the reaction to occur are not yet fully understood. Preserved serpentine rim with pseudomorphs of carbonate after serpentine and lobate-shaped carbonate grains are key structural features for replacement of serpentinite by carbonates. Cathodoluminescence microscopy reveals that Ca-rich carbonate precipitation in serpentinite is associated with a sequential assimilation of Mn. Homogeneous δ 18 O values at the µ m-scale within grains and host sample indicate low formation temperature ( < 20 ◦ C) from carbonation initiation, with a high fluid to rock ratio. δ 13 C (1–3 ± 1% (cid:24) ) sit within the measured values for hydrothermal systems ( − 3–3% (cid:24) ), with no systematic correlation with the Mn content. δ 13 C values reflect the inorganic carbon dominance and the seawater source of CO 2 for carbonate. Thermodynamic modeling of fluid / rock interaction during seawater transport in serpentine predicts Ca-rich carbonate production, at the expense of serpentine, only at temperatures below 50 ◦ C during seawater influx. Mg-rich carbonates can also be produced when using a model of fluid discharge, but at significantly higher temperatures (150 ◦ C). This has major implications for the setting of carbonation in present-day and in fossil margins. This pressure corresponds to the lithostatic pressure expected at drill site 1277, where drilling was performed at 4600 m water-depth up to 180 m below the seaﬂoor. As expected, tests of the respective impacts of temperature and pressure on the thermodynamic equilibrium show that temperature is the primary control of the stable assemblage whereas pressure variations are less important in the pressure and temperature ranges relevant to MORs (4 to 350 ◦ C and 20 to 50 MPa).


Geological Setting
In the Newfoundland hyper-extended margin and conjugate Iberian margin, subcontinental lithospheric mantle was exhumed on the seafloor by detachment faulting. During the mantle exposure on the seafloor, a superposition of near-seafloor processes i.e., tectonic, hydrothermal and sedimentary brecciation occurred, resulting in tectono-sedimentary breccias [31][32][33]. Ophicarbonates are observed from the proximal to the distal margin of both the Newfoundland and Iberian margins. Carbonation processes overlap the brecciation, leading to calcite veins, replacement of serpentine and carbonate cement. The tectono-sedimentary breccias rework exhumed serpentinized footwall and thus preserved every stage of carbonation from its onset. Previous work from the Iberian margin has estimated that the temperature of carbonation in calcite veins and cement was < 45 • C based on δ 18 O VSMOW values of 25% to 35% [1,3,[34][35][36]. The carbonation was thus interpreted as a near-surface cold process [35].
Here, we study samples from the International Ocean Discovery Program (IODP) site 1277 ( Figure 1; [37]). The drill core contains serpentinized peridotites and locally magmatic intrusions (e.g., alkaline sills of 124-112 Ma, followed by Mid-Ocean Ridge Basalt-type magmatism <112 Ma at the onset of oceanic spreading, [38,39] and references therein). The drill core is also composed of brecciated serpentinized peridotites over 20 m (Figure 1c); we focus on these serpentinite samples that show evidence of replacement by carbonates.
Here, we study samples from the International Ocean Discovery Program (IODP) site 1277 ( Figure 1; [37]). The drill core contains serpentinized peridotites and locally magmatic intrusions (e.g., alkaline sills of 124-112 Ma, followed by Mid-Ocean Ridge Basalt-type magmatism < 112 Ma at the onset of oceanic spreading, [38,39] and references therein). The drill core is also composed of brecciated serpentinized peridotites over 20 m (Figure 1c); we focus on these serpentinite samples that show evidence of replacement by carbonates.

Microscopy
Scanning Electron Microscopy images were acquired using a Tescan Mira LMU FE-SEM (ISTe, University of Lausanne (UniL), Lausanne, Switzerland) operated at an acceleration voltage of 20 kV and a probe current of 20 nA. We used cold cathodoluminescence (CL) microscopy (model 8200 MkII (OPEA TM ), ISTe, UniL, Lausanne, Switzerland) to qualitatively estimate the variation in Mn content [41]. Light orange corresponds to a high Mn concentration whereas dark red is associated with low Mn concentrations. Electron microprobe analyses were conducted on a JEOL JKA-8530F (UniL, Lausanne, Switzerland), under beam of 15 nA and current of 10 nA as proposed by Lane and Dalton [42].

Secondary Ion Mass Spectrometer (SIMS) Analyses of Calcite
We measured carbonate grains in the core of the serpentine mesh textures using a Secondary Ion Mass Spectrometer Cameca 1280HR. We used a 10 kV Cs + primary beam, a ~1.5 nA current, resulting in a ~15 µ m beam size. The electron flood gun, with normal incidence, was used to compensate charges.
For oxygen isotope measurements, 16 O and 18 O were analyzed with a mass resolving power of 2400 and collected in 2 faraday cups (10 10 and 10 11 Ω) in multi-collection mode. Each analysis takes ~4 min, including pre-sputtering (30 s) and automated centering of secondary electrons. This setting allowed an average reproducibility of ±0.28‰ (2 standard deviations (2 SD)) on an in house calcite reference material (UNIL_C1; δ 18 OVSMOW = 18.29 ± 0.12‰ (2 standard error (2 SE))) at the beginning of each session. A set of 4 analyses of the calcite standard were also measured every 12 analyses for monitoring the instrument stability. The variation of the calcite standard over the 24 h session was

Microscopy
Scanning Electron Microscopy images were acquired using a Tescan Mira LMU FE-SEM (ISTe, University of Lausanne (UniL), Lausanne, Switzerland) operated at an acceleration voltage of 20 kV and a probe current of 20 nA. We used cold cathodoluminescence (CL) microscopy (model 8200 MkII (OPEA TM ), ISTe, UniL, Lausanne, Switzerland) to qualitatively estimate the variation in Mn content [41]. Light orange corresponds to a high Mn concentration whereas dark red is associated with low Mn concentrations. Electron microprobe analyses were conducted on a JEOL JKA-8530F (UniL, Lausanne, Switzerland), under beam of 15 nA and current of 10 nA as proposed by Lane and Dalton [42].

Secondary Ion Mass Spectrometer (SIMS) Analyses of Calcite
We measured carbonate grains in the core of the serpentine mesh textures using a Secondary Ion Mass Spectrometer Cameca 1280HR. We used a 10 kV Cs + primary beam, a~1.5 nA current, resulting in a~15 µm beam size. The electron flood gun, with normal incidence, was used to compensate charges.
For oxygen isotope measurements, 16 O and 18 O were analyzed with a mass resolving power of 2400 and collected in 2 faraday cups (10 10 and 10 11 Ω) in multi-collection mode. Each analysis takes 4 min, including pre-sputtering (30 s) and automated centering of secondary electrons. This setting allowed an average reproducibility of ±0.28% (2 standard deviations (2 SD)) on an in house calcite reference material (UNIL_C1; δ 18 O VSMOW = 18.29 ± 0.12% (2 standard error (2 SE))) at the beginning of each session. A set of 4 analyses of the calcite standard were also measured every 12 analyses for monitoring the instrument stability. The variation of the calcite standard over the 24 h session was ±0.52% (2 SD), suggesting a magnetic drift, for which the data were corrected. The error reported for samples reflected the 2 SD bracket between the surrounding set of standard analyses.
Minerals 2020, 10, 184 4 of 24 12 C and 13 C, measured in a second SIMS session, were separated using a mass resolving power of 4600, and collected in 2 faraday cups (both 10 11 Ω). Each analysis takes~4-6 min, including pre-sputtering (45 s) and automated centering of secondary electrons. This setting allowed an average reproducibility of ±0.6% (2 SD) on a calcite in house standard (UNIL_C1; δ 13 C VPDB = 0.64 ± 0.08% (2 SE)) at the beginning of the session. A set of 4 analyses of a calcite standard was also measured every 10 unknowns; the variation of the calcite standard over the entire session (~3 h) was ±0.8% (2 SD), suggesting a slight magnetic drift. The error reported for samples reflected the 2 SD bracket between the surrounding set of standard analyses.
During both sessions, the faraday cups were calibrated at the beginning of the sessions using a calibration routine. Mass calibrations were performed at the beginning of the sessions and again after 12 h for the 18 We converted the measured δ 18 O VSMOW values from calcite into temperatures using the standard procedure of Kim and O'Neil [43] and assumed that calcite crystallized in equilibrium with seawater with a δ 18 O VSMOW = 0% at present-day, and a δ 18 O VSMOW = −1.2% for an ice-free world [44].

Thermodynamic Modeling
Fluid/rock interaction was modelled at the equilibrium with the EQ3/6 software [45]. The evolution of the mineralogy and the fluid chemistry were predicted for oceanic serpentinized peridotite during hydrothermal fluid recharge or discharge. We model hydrothermal circulation in one dimension with a medium discretized in a sequence of 50 to 60 boxes. The temperature was fixed in each box and either linearly increases (recharge) or decreases (discharge) from one box to the other along the flow path ( Figure 2). The simulations consisted of 3000 to 7000 iterations subdivided into two steps. First, the composition of the fluid phase in equilibrium with the solid was calculated independently in each box. This induces mass transfer between the solid and the fluid. The aim of this model was to determine the impact on mineralogical composition of fluid-mediated mass transfer from low to high temperatures (recharge) or high to low temperatures (discharge). Therefore, during the second step, the fluid of modified composition was transferred from one box to the next one in flow direction, while seawater with fixed composition was introduced into the first box ( Figure 2). This latter step conserves mass. Time is not a parameter of the model since time dependent processes such as fluid flow, diffusion or reaction kinetics were not considered. To determine the extent of fluid-rock interaction, we used the cumulative mass of fluid introduced in the first box of the model ((W/R) d ; scaled to 1 kg of solid initially present in this box). This parameter could seem similar to the water to rock ratio commonly used in thermodynamic calculations (e.g., [46]), but it does not correspond to a single calculation at the equilibrium since over the number of iterations it integrates the amount of fluid added in the first box (it is thus a "dynamic" water to rock ratio). This water to rock ratio should ideally be of 2 × 10 −2 to correspond to the porosity of~5% measured in serpentinized peridotites [47]. However, with such a low amount of fluid, the computing time is too long to obtain significant mass transfer leading to carbonates formation. Therefore, at each iteration, we added into the model 38 kg (discharge) or 380 kg (recharge) of fluid (for 1 kg of solid) in the first box. Thermodynamic calculations at the equilibrium were performed with the database of Johnson et al. [48] with additional data for ferrous and ferric serpentine and ferrous brucite from Klein et al. [49] (see [50] for details). No organic components were considered. In the calculations, we suppressed the reduction of carbon by H 2 to account for the slow kinetics of methanogenesis reactions [51]. The composition of the reacting serpentine was initially fixed at Mg 2.77 Fe 2+ 0.13 Fe 3+ 0.13 Si 1.94 O 5 (OH) 4 based on microprobe analyses. Serpentine composition can then evolve as a result of fluid-rock interactions. Iron oxidation state in serpentine was not determined, therefore, half of the iron was assumed to be trivalent, based on the compilation of serpentine composition by Evans [52], in agreement with the thermodynamic calculations by Klein et al. [48] and the µ-X-ray Absorption Near Edge Structure serpentine analyses on mesh textures by Andreani et al. [53]. The composition of seawater introduced in the first box of the model is given in Table S3. Simulations were run at a constant pressure of 50 MPa for simplicity. This pressure corresponds to the lithostatic pressure expected at drill site 1277, where drilling was performed at 4600 m water-depth up to 180 m below the seafloor. As expected, tests of the respective impacts of temperature and pressure on the thermodynamic equilibrium show that temperature is the primary control of the stable assemblage whereas pressure variations are less important in the pressure and temperature ranges relevant to MORs (4 to 350 • C and 20 to 50 MPa).
Minerals 2020, 10, x FOR PEER REVIEW 5 of 24 MPa for simplicity. This pressure corresponds to the lithostatic pressure expected at drill site 1277, where drilling was performed at 4600 m water-depth up to 180 m below the seafloor. As expected, tests of the respective impacts of temperature and pressure on the thermodynamic equilibrium show that temperature is the primary control of the stable assemblage whereas pressure variations are less important in the pressure and temperature ranges relevant to MORs (4 to 350 °C and 20 to 50 MPa). We applied equilibrium thermodynamics at low temperature in far from the equilibrium conditions. Reaction kinetics may prevent the achievement of equilibrium in these conditions. Taking into account reaction kinetics is fraught with uncertainty due to the lack of data on reaction rates in hydrothermal conditions and the need for modeling other time-dependent processes such as fluid flow. Several studies investigated the coupling between reaction kinetics and fluid flow in hydrothermal systems [54][55][56], but they simplified the system to a single chemical reaction and are therefore not able to model the changes in chemical composition investigated here. We therefore chose to calculate the system composition at the equilibrium towards which the system tends [57]. Such calculations were successfully applied to processes occurring at MORs (e.g., [47,48]). The use of equilibrium thermodynamics may be challenged by the production of metastable phases during reaction. The main minerals formed here are serpentine and carbonates. Metastable serpentine minerals such as chrysotile have similar composition and thermodynamic properties to stable phases We applied equilibrium thermodynamics at low temperature in far from the equilibrium conditions. Reaction kinetics may prevent the achievement of equilibrium in these conditions. Taking into account reaction kinetics is fraught with uncertainty due to the lack of data on reaction rates in hydrothermal conditions and the need for modeling other time-dependent processes such as fluid flow. Several studies investigated the coupling between reaction kinetics and fluid flow in hydrothermal systems [54][55][56], but they simplified the system to a single chemical reaction and are therefore not able to model the changes in chemical composition investigated here. We therefore chose to calculate the system composition at the equilibrium towards which the system tends [57]. Such calculations were successfully applied to processes occurring at MORs (e.g., [47,48]). The use of equilibrium thermodynamics may be challenged by the production of metastable phases during reaction. The main minerals formed here are serpentine and carbonates. Metastable serpentine minerals such as chrysotile have similar composition and thermodynamic properties to stable phases (lizardite; [58]). They will thus have a limited effect on the model outputs. However, Mg-carbonates (i.e., magnesite and dolomite) have precipitation kinetics at least 6 orders of magnitude slower than calcite precipitation at ambient conditions [59]. The high activation energy of Mg-carbonate precipitation makes Mg-carbonates formation possible at high temperature. Because of the slow Mg-carbonate kinetic reaction, we run both simulations including and excluding Mg-carbonate precipitation. Finally, we assumed thermodynamic equilibrium in each box of the model. This requires that transport processes (i.e., diffusion and advection) are fast enough compared to the reaction to achieve equilibrium at a scale of several meters. carbonate kinetic reaction, we run both simulations including and excluding Mg-carbonate precipitation. Finally, we assumed thermodynamic equilibrium in each box of the model. This requires that transport processes (i.e., diffusion and advection) are fast enough compared to the reaction to achieve equilibrium at a scale of several meters.

Brecciation and Carbonate Crystallization
The 20 m-thick layer of tectono-sedimentary breccias drilled at Site 1277 ( Figure 1; [37]) contains several calcite generations recognized by their crosscutting relationship and angular to rounded clasts of serpentinite ( Figure 3). The calcite occurs in three contexts: it precipitates in veins crosscutting the clasts and formed during tectonic brecciation (type 1; Figure 4a       In cathodoluminescence (CL) images, the calcite veins (type 1) display a color banding with an alternation of black and orange (Figure 4b). The banding is either parallel to fracturing, showing the opening direction of the vein (Figure 4b), or concentric, leading to scalenohedral crystals (Figure 5b). New grains appear to have nucleated in the vein at fracture edges, leading to single crystals with irregular shapes and low Mn-content (Figure 4d; dark in CL image). In addition, calcite grain growth occurs in parallel to the fracture opening direction (Figure 4b). It is common to observe in CL that the In cathodoluminescence (CL) images, the calcite veins (type 1) display a color banding with an alternation of black and orange (Figure 4b). The banding is either parallel to fracturing, showing the opening direction of the vein (Figure 4b), or concentric, leading to scalenohedral crystals (Figure 5b). New grains appear to have nucleated in the vein at fracture edges, leading to single crystals with irregular shapes and low Mn-content (Figure 4d; dark in CL image). In addition, calcite grain growth occurs in parallel to the fracture opening direction (Figure 4b). It is common to observe in CL that the cores of the scalenohedron-shaped calcites are black (Mn-poor) whereas the last band is the brightest (Mn-rich; up to 1 wt. %, Figure 4d). The scalenohedral sparitic calcite grains found in veins vary in size from 10 to 100 µm. The banding in the veins display sharp contacts with the calcite in the cement due to brecciation (Figure 4c). The cement (type 2) is characterized by small, grain-sized calcite (<50 µm), with no CL-banding; CL colours are dark orange, with neither variation at sample or core scale (Figures 4b and 5b).
The serpentinite clasts are made of a network of mesh-like, regularly spaced fractures with two perpendicular orientations. Within the core of this regular mesh texture, 50 to 200 µm wide calcite grains are found (type 3) (Figure 5a,b). The mesh texture is preserved during replacement by carbonates (i.e., pseudomorphic replacement). In the core of the mesh, calcite grains grow concentrically from a nucleus, until reaching the rim of the mesh (Figure 5b). Depending on the nucleus location, the bands may not be continuous in the calcite grain. The CL colour of the bands is dark near the nucleus and again, bright orange in the outermost one ( Figure 5b).
The banding observed in calcites of types 1 and 3 is similar (Figures 4b and 5b). At the contact between calcite and serpentine, the calcite grains are sharp and angular above the µm-scale (Figure 5c,d), and below the µm-scale the grains can display lobate shapes and are concave towards the serpentine (Figure 5d). The reaction front separates the well-crystallized calcite and fibrous serpentine and spans several micrometers. It is composed of <0.5 µm-thick calcite needles pervasively distributed in a disordered serpentine matrix (Figure 5d).
The µm-scale banding does not allow us to perform bulk isotopic analysis, as we would have obtained a mean value for all the bands. Therefore, we measured in-situ δ 18 O in the different bands with the SIMS. We acquired profiles through calcite grains in veins (type 1) and in the serpentine mesh core (type 3). Only CL bands larger than 20 µm could be analyzed (white crosses on Figures 4b and 5b; Table S2). Measured δ 18 O VSMOW values cluster in a relative restricted range for pseudomorphous grains (30.8 ± 0.4% to 32.6 ± 0.4% , n = 14), whereas calcite in veins are homogeneous (28.4 ± 0.4% , n = 49) ( Figure 6; Table S2). In all the calcites measured, no variation of δ 18 O VSMOW can be observed from core to rim. δ 13 C VPDB measurements vary from 1.09 ± 0.63% to 3.07 ± 0.85% (Table S2) but do not show systematic variation with the Mn-banding. Our dataset sits within the range of measurements available for the Iberia margin ( Figure 6).  Table S1.

Thermodynamic Modeling of Carbonation
Opposite flow directions in the recharge and the discharge models are responsible for the formation of different mineralogical assemblages (Figures 7 and 8) and aqueous fluid compositions (Figures 9 and 10). This is due to the variations in mineral solubility with temperature. If mineral solubility increases with temperature (silicates), fluid transfer from low to high temperature (recharge) will lead to dissolution, whereas fluid transfer from high to low temperature (discharge) will lead to precipitation. Therefore, mass transfer will differ for models considering recharge or discharge. An opposite trend is expected for minerals with solubility decreasing with temperature (carbonates). Rock composition is observed to change along sharp boundaries (reaction fronts) for  Table S1.

Thermodynamic Modeling of Carbonation
Opposite flow directions in the recharge and the discharge models are responsible for the formation of different mineralogical assemblages (Figures 7 and 8) and aqueous fluid compositions (Figures 9  and 10). This is due to the variations in mineral solubility with temperature. If mineral solubility increases with temperature (silicates), fluid transfer from low to high temperature (recharge) will lead to dissolution, whereas fluid transfer from high to low temperature (discharge) will lead to precipitation. Therefore, mass transfer will differ for models considering recharge or discharge. An opposite trend is expected for minerals with solubility decreasing with temperature (carbonates). Rock composition is observed to change along sharp boundaries (reaction fronts) for both recharge (Figure 7) and discharge models (Figure 8). In both models, the initial serpentine, exclusively composing the rock, is progressively replaced by talc as the fluid circulates through the rock. The talc is ultimately replaced by quartz (or clays at temperatures below ca. 40 • C) for the largest amount of fluid having circulated through the rock. The mineralogical assemblage can thus become quartz + carbonate and the modelled rock is a listvenite. Talc appears for lower (W/R) d when the precipitation of Mg-carbonates is allowed in the model, than when Mg-carbonates precipitation is excluded (Figures 7 and 8). These mineralogical evolutions require that the Mg and Si initially contained in serpentine are transported outside the rock as aqueous species in the fluid (Figures 9 and 10).
Carbonates are only formed at temperatures below 150 • C for the discharge model, and 100 • C for the recharge model (Figures 7 and 8e-h). In the discharge model, Ca concentration in the fluid slightly increases from 1.1 × 10 −4 to 4.5 × 10 −4 mol/kg as temperature decreases. The limitation of the carbonate stability field is thus mainly associated with pH decrease from 8.5 to 5 as temperature increases from 4 to 250 • C ( Figure 10). In the recharge model, the limitation of the carbonate stability field is associated with anhydrite (CaSO 4 ) formation at temperature above 100 • C as shown, for example in Figure 9a The distribution and composition of carbonates are different, even though the reactions of serpentine alteration were similar for discharge and recharge. The carbonates are first formed at the lowest temperature (4 • C) in the recharge model (Figure 7b,f). As fluid/rock interactions increase, the precipitation front progressively migrates towards higher temperatures. This front forming a serpentine + carbonate assemblage in equilibrium with a fluid of low Si concentration (Figure 9) is followed by a second front of talc + carbonate precipitation leading to increase in carbonate modal contents up to 60 mol. % (Figure 7h) and in Si concentration up to one order of magnitude (Figure 9h). Carbonates first replace serpentine at high temperatures (~150 • C) in the discharge model compared to the recharge model where this replacement starts at the lowest temperature (4 • C). The modal amounts of carbonate in the boxes are found to progressively decrease along the fluid flow path by one order of magnitude from 150 • C to 50 • C. Serpentine replacement by carbonates requires larger amount of fluids (or larger porosities for the same number of time steps in our model) to be observed in the case of recharge than in the case of discharge. This is due to the higher serpentine solubility where serpentine dissolution occurs in the discharge model at 150 • C than in the recharge at 20 • C. When Mg-carbonate precipitation is included in the model, carbonates are dolomite in the whole investigated temperature range in the recharge model and at temperatures below 20 • C in the discharge model. At temperatures above 20 • C, carbonates precipitate as magnesite in the discharge model, suggesting that Ca-rich carbonates are preferentially formed at low temperature. When Mg-carbonate precipitation is excluded, Ca-carbonates are only found to precipitate in the recharge model at temperatures below 110 • C (Figure 7). Minerals 2020, 10, x FOR PEER REVIEW 11 of 24

Serpentine Replacement by Carbonate During Seawater Influx in the NF Margin
We identify three mechanisms of carbonation in the Newfoundland serpentinites: veining, cementation and pseudomorphic replacement. Tectonic deformation during mantle exhumation along passive rifted margins leads to cataclasis in the footwall of detachment faults [32] and to fluid circulation. Both mechanisms result in calcite vein formation (e.g., [5,60]). They follow the pre-existing fracture network of the serpentinized peridotite. Near the seafloor, sedimentation is active, leading to calcite cementation into serpentinite breccia. All along this tectono-sedimentary sequence, serpentine is partially replaced by carbonate in the core of the mesh texture. To determine when replacive carbonation occurs during the mantle exhumation, we measured the temperature record of calcite in the inferred oldest carbonate.
During mantle exhumation, the anisotropic thermal contraction of peridotite together with tectonic stresses generates a primary microfracturing responsible for the onset of serpentinization [61]. This fracture network is then re-used for reaction induced fracturing [56,62,63]. The serpentine growing along the microfractures is organized in lizardite pseudocolumns constituting the rim of the mesh texture [61,63,64]. For higher reaction degree, preserved olivine grains can be altered into isotropic serpentine or proto-serpentine [65,66]. The core of meshes can be composed of either preserved olivine or proto-serpentine, depending on the extent of the serpentinization reaction. Residual olivine grains are extremely rare in the Iberia and Newfoundland exhumed serpentine basement, where the degree of serpentinization is >90% complete [67]. Therefore, the different habits of serpentine crystals and their coherence is probably responsible for the preferential carbonate growth in the mesh core of the studied sample. The lobate shape of calcite grains suggests serpentine replacement. Our interpretation is reinforced by the fact that calcite needles at the reaction front start to grow from isotropic serpentine. In the veins, calcite grains grow with a similar CL banding, but the absence of lobate shape suggests that these grains are not replacive.
Hydrothermal systems are an important source of manganese as described in several present-day active hydrothermal vents, e.g., East Pacific Rise, [68] or in the acidic hydrothermal systems (pH < 5) from the MAR where Mn concentration varies from 59 to 2250 µM [18]. Ca 2+ from calcite can be substituted by Mn 2+ to form a solid solution of MnCO 3 and CaCO 3 [69]. CL of calcite is activated by Mn 2+ , whose emission colour is orange [70]. Temperature is not strongly influencing the Mn uptake in calcite grains [71]. Therefore, the Eh-dependent solubility of Mn into the fluid allows using CL colour variation as a proxy for oxygen fugacity variations in the fluid [72].
The CL-banding sequence is comparable between calcite grains growing in veins and in the serpentine mesh core, indicating a synchronous growth. The low content of Mn in calcite grains core reflects oxidizing conditions in both cases. In contrast, the last band in calcite is enriched in Mn testifying for more reduced fluids [72]. The fluid responsible for carbonation is oxidizing from the beginning of the reaction, even though fluids formed during serpentinization are reduced due to the coupled oxidation of the ferrous iron from olivine and pyroxenes to form magnetite [73]. This indicates that carbonation, from its onset, occurs after serpentinization near the seafloor, through interaction with oxidizing seawater.
The measured δ 18 O are homogeneous in calcite precipitating in the veins (δ 18 O VSMOW ≈ 31% ) and vary by less than 2% in the calcite formed in the cores of the serpentine meshes. There is no systematic variation in the δ 18 O with respect to the Mn content variation or the location in the grain. The narrow ranges of the in-situ isotopic measurements at the tens of micrometer scale indicate limited variations of the composition of infiltrating fluids, both spatially, as suggested by the bulk data at MORs, and temporally, as indicated by the absence of zoning. Assuming seawater isotopic composition, the δ 18 O measurements give carbonation temperatures between T min −1 to 11 • C and T max 4 to 18 • C (Table S2). This low temperature of carbonate formation is consistent with experiments of inorganic carbonate precipitation at 25 • C [74][75][76]. It also suggests that serpentinization and carbonation can be two temporally decoupled processes. Indeed, the highest serpentinization rates measured in experiments are found in the range of 250-300 • C [77,78].
The temporal decoupling between serpentinization and carbonation can also be shown by subtracting δ 18 O calcite from δ 18 O serpentine [79]. For this calculation, we used a mean value calculated from a compilation of δ 18 O serpentine from the Iberia margin (Leg 103, [34]; Leg 149, [3]) and from MAR [80]. The mean value for δ 18 O serpentine used here is 10.65% . We also used a mean value for δ 18 O calcite from the NF margin i.e., 31% . If calcite was formed in equilibrium with serpentine, δ 18 O calcite -δ 18 O serpentine ≈ 10% . Here, the difference exceeds 20% (Figure 6 and Table S2), therefore, calcite and serpentine are not at equilibrium. Fast advective fluid transport (e.g., in a fracture network) during fluid recharge or discharge would lead to limited interaction between the fluid and the serpentine, and to the preservation of the fluid isotopic signature. However, pervasive fluid flow in a low permeable rock favors the fluid-serpentine interactions and can thus modify the initial isotopic signature of the fluid and of the precipitating carbonate [81]. In addition to the fluid-serpentine disequilibrium, the extensive carbonation as well as the absence of δ 18 O zoning in calcite indicate that carbonation is fluid-buffered. As the fluid has a seawater signature, carbonation probably occurs near the surface during the onset of recharge into the serpentinized peridotites. The reproducibility of data in the MAR and Iberia-Newfoundland margins suggests that the recharge is a common process. δ 13 C VPDB measured in NF margin are in agreement with previous data from the conjugated Iberia margin (i.e., between −3% and 3% ) and are typical for marine hydrothermal systems (e.g., [1]). This suggests that C is dominantly inorganic derived from seawater as suggested for the Iberian margin [1]. Thermodynamic modeling indicates that such a precipitation of seawater-derived inorganic carbon is possible and requires the circulation of more than 10 3 kg of seawater per kg of rock (Figures 7  and 8). The need for high dynamic water to rock ratios is associated with the low concentration of inorganic carbon in seawater (~2 mM). There are no systematic correlations between Mn-banding and δ 13 C variation.

Carbonation in Passive Margins and Slow-Spreading MORs
We compiled calcite δ 18 O and δ 13 C bulk measurements from the Iberia and Newfoundland margins and from the Central MAR ophicarbonates ( Figure 6). Based on their δ 18 O record and calculated temperatures of formation, we infer that the ophicarbonates can be separated into two distinct groups, (1) calcite formed during seawater influx in an evolving carbonation front; or (2) during hydrothermal fluid discharge, in the case that during recharge, fluid-peridotite interaction was small.
As calcites from the Iberia margin and MAR, formed by veining, cementation and pseudomorphic replacement, record high δ 18 O, we infer this calcite to be formed during seawater influx. This has already been proposed in calcite veins from Iberia with O isotopic compositions varying from 27.6% to 31.2% [1] and in sedimentary carbonates (e.g., botryoidal calcite) at the Iberia margin (δ 18 O from 30% to 32% ; [2,35]) and in the MAR (δ 18 O from 33% to 35% ; [23]).
Calcite formed during hydrothermal fluid discharge is created by veining or hydrothermal deposits (e.g., [21,23,24]). In the MAR, carbonates found in veins and deposits are calcite and aragonite [21]. Mg content in calcite can be as high as 28 wt. % [24]. These high temperature calcite veins are found at the vicinity of hydrothermal vents, and record δ 18 O between 26% and 8% (Mid-Atlantic Ridge, Kane Fracture Zone [26]; ODP Leg 209, [27]; Lost City, [21,23]). However, those samples represent a minority of the collected samples (see references in Table S1). Such low δ 18 O calcite indicates carbonation temperatures higher than 100 • C [23]. No pseudomorphic replacement is observed in environments in close relationship with an active hydrothermal vent. Based on O and C isotopic signature of replacive calcites, we interpret carbonate replacement in the samples from Newfoundland, the Iberian margin and the MAR as related to near-surface alteration during seawater influx.
We demonstrate that carbonation on passive margins is a low temperature, near surface process. This is true, even if magmatism is present as demonstrated at Site 1277, where syn-and post-rifting magmatism was identified (e.g., [39]). Indeed, calcite recorded temperatures below 20 • C. In the case of the Iberia-Newfoundland margins, magmatism has also no effect on the isotopic record, and hence does not influence carbonate precipitation.

Insights from Numerical Modeling
Carbonates are formed at two distinct temperature ranges at MORs, either below 10 • C or locally at the Lost City above 100 • C in veins feeding hydrothermal vents. Interestingly, this difference in carbonation temperature is also the main difference between the two simulations performed here with recharge and discharge models. Our simulations reproduced carbonate formation during recharge at near-surface conditions. They also predict carbonate formation at approximately 150 • C during discharge. Moreover, carbonate production required a higher fluid amount ((W/R) d ) for recharge than for discharge models, suggesting more open system conditions. Talc formation is decoupled from carbonate precipitation in the recharge models only, leading to rocks exclusively composed of serpentine and carbonates as observed in the natural samples. In all the simulations, Ca-rich carbonates were only formed at low temperature, whereas carbonation at temperature above 150 • C only produced Mg-rich carbonates. Based on these results, we interpret carbonate replacement in the samples from Newfoundland, the Iberian margin and the MAR as related to near-surface alteration during recharge. Carbonation at high temperature (>100 • C; [23]) is also expected to occur at MORs during discharge and, in particular, near hydrothermal vents such as the Lost City. The simulations predict the observation of calcite in the natural samples studied here, but they do also predict Ca-rich carbonates formation at temperatures above 100 • C, as is observed at the Lost City Hydrothermal Field. Due to missing compatible thermodynamic data, we do not include phases such as layered double hydroxides (coalingite-pyroaugite, LDHs), and hydrous Mg-carbonates (hydromagnesite and nesquehonite) (e.g., [82]).

Implications for Ophicarbonates Formation Preserved in Ophiolites
Ophicarbonates have been described first in remnants of the central Jurassic Tethys, in the Klosters, Totalp and Arosa ophiolites, in the Central-Eastern Alps [4] and in the Chenaillet ophiolite in the Western Alps [9]. Over the last 30 years, multiple studies have been conducted with a renewed interest, since they contain a key to the understanding of mantle exhumation to the seafloor, and because of their implication as an analogue for present-day passive margins. In addition, carbonation of serpentinite and other mantle rocks is a potential engineering solution to CO 2 sequestration [83]. Alpine ophicarbonates are interpreted to be the result of oceanic alteration of serpentinized peridotites during mantle exhumation associated with the Jurassic hyperextension phase [13], already proposed by Weissert and Bernoulli [4] and discussed by Früh-Green et al. [84].
It is important to note that carbonates are mostly calcites, and minor aragonite [2], never dolomite. Mineralogy of carbonates encountered in the Alpine Tethys ophicarbonates is mostly calcite, and minor dolomite (e.g., in Val Ventina, Alps, [82]; and in the Chenaillet, [13] and references therein). We compiled δ 18 O and δ 13 C measured in ophicarbonates from the literature in several locations in the Alpine Tethys. The data set contains bulk measurements on carbonate veins [84], disseminated carbonates [13], as well as sedimentary ophicarbonates [4,85,86].
Regional metamorphism experienced by ophicarbonates in the Platta nappes did not exceed 200 to 250 • C ( [87] and references therein). The δ 18 O values of carbonates range from 21% to 14% [85] in ophicarbonates towards the thrust plane, which separates them from serpentinites in the Platta nappe. Metamorphism and fluid flow during metamorphism seems to play in important role in the δ 18 O record in ophicarbonates, and hence a careful re-evaluation of these values should be attempted to ascertain the initial condition of formation of these carbonates prior to metamorphism. This explanation is supported by the low δ 18 O values also recorded in sedimentary ophicarbonates (from 8.6% to 24.7% in the Western and Central Alps [4,13,86,88]). These low δ 18 O values correspond to temperatures from 40 to 70 • C (Table S1). As in the Iberian margin, near-surface sedimentary carbonation precipitation also occurs in a range of temperatures below 20 • C, we infer that these temperatures are too high for true sedimentary deposition.
A way to optically verify if carbonate re-equilibration by metamorphism has started is to observe the Mn-banding in carbonates by cathodoluminescence. Diffusion of Mn is approximately 10 orders of magnitude faster (e.g., 1 mm at 300 • C in 25 Ma) than diffusion of O in calcite (30 µm to 10 −6 µm for the same setting [89][90][91]. Many carbonates precipitating in shallow hydrothermal settings will obtain rhythmic layers of CL-active elements, as is shown in this study. Since diffusion of these elements (like Mn) is fast, they will disappear upon heating, and are obliterated by recrystallization. This means that if a carbonate preserves its banding structure acquired during carbonation on the seafloor, it should also preserve its δ 18 O signature. Therefore, a test for recognizing re-equilibration in ophicarbonates from ophiolites would be the presence of thin CL-banding. This criterion can only be applied if diffusion is the only process acting during re-equilibration since calcite dissolution-reprecipitation may also modify the composition of the carbonates at low temperature. Nevertheless, this will also eliminate potentially the isotopic signature.
To follow up, we estimate the δ 18 O expected in calcite that is supposed to be equilibrated with the surrounding serpentine during metamorphism. δ 18 O serpentine in Alpine ophiolites ranges from 4.3% to 13.4% [84]. An alternative to the onset of differential re-equilibration during Alpine metamorphism could be that instead of precipitating during seawater influx, the ophicarbonates were formed during hydrothermal fluid discharge. We want to emphasize that this explanation would explain only the high temperatures recorded by the δ 18 O in carbonates, as the textures and brecciation processes are similar to the present-day passive margin samples and are easily explained by the seawater influx model.
In an ophiolite where hydrothermal vents are clearly observed (e.g., Chenaillet), it should be easier to establish a relationship between carbonation and the distance between the hydrothermal fluid pathways. Magmatism may also play an important role in the high temperature recorded by ophicarbonates in distal parts of the Ocean Continent Transitions such as the Chenaillet, as late syn-rift to early post-rift magmatism with intrusions crosscutting the detachment faults are observed [28].
In the case of mostly discharge-driven carbonation, as the Alpine Tethys ophicarbonates are a good analogue for the present-day Iberia passive margin, near-surface ophicarbonates with high δ 18 O (25-35% ) should also be reported in the remnant ophiolites, which is still not the case.

Implications for CO 2 Mineral Sequestration
The storage of CO 2 through carbonate precipitation appears to be one of the solutions to compensate for the anthropogenic greenhouse gas emissions [92]. Among the Earth's crust rocks, ultramafic rocks have one of the highest potentials for carbon sequestration due to their high magnesium content [6,[93][94][95][96][97][98][99]. The results of this study provide constraints for the process of CO 2 sequestration in ultramafic rocks. We show here that low-temperature carbonation on the seafloor (T < 50 • C) produces Ca-carbonates. The calcium source is the seawater, and high water to rock ratios are needed to induce carbonate precipitation. The high MgO content of the ultramafic rock is therefore not an asset for CO 2 storage in the conditions prevailing on the seafloor. However, thermodynamic modeling indicates that high temperature carbonation (T > 100 • C) requires smaller water to rock ratios since magnesium can be incorporated into the carbonates. Mimicking the high temperature process thus appears to be more relevant to develop an efficient CO 2 storage solution. Such a solution would also need to circumvent the potential issues associated with the modification of the porous network by carbonation. Serpentine reaction occurring at high temperature to form talc and then quartz indeed requires significant magnesium and silica transport.

Conclusions
We provide microtextural observations of pervasive serpentine replacement by calcite and synchronous calcite growth in veins in exhumed mantle from the Newfoundland passive margin. We interpret those calcite grain cores as the first carbonate to grow in exhumed mantle. This is shown by the crosscutting relationship between brecciation and calcite grains textures revealed by CL images. Replacive calcite maintains the serpentine mesh texture and grows as scalenohedral crystals with a characteristic Mn-compositional banding. Preciseness of SIMS allowed us to measure O and C isotopic composition of each band in a single calcite grain. Measured O isotopes highlight no systematic variation, C isotope measurements display seawater range values and O isotope thermometry reveals that carbonation is cold (<20 • C) since the onset of the reaction.
Our thermodynamic modeling predicts Ca-rich carbonate crystallization near-surface during seawater influx and hydrothermal fluid discharge. Si-rich phases appear in the system with carbonation front evolution through time and space (e.g., talc).
In the discharge model, this stability field limitation is associated with carbonate dissociation into HCO 3 − , aq as the pH decreases with temperature. In the recharge model, the limitation of the carbonate stability field is associated with anhydrite (CaSO 4 ) formation at temperatures above 100 • C. Anhydrite formation only occurs in the first box of the model at high temperature in the discharge model (250 • C). We summarize the petrological, geochemical and numerical modeling results of this study in Figure 11, where the differences between recharge and discharge regarding carbonate precipitation are highlighted. Replacive calcite maintains the serpentine mesh texture and grows as scalenohedral crystals with a characteristic Mn-compositional banding. Preciseness of SIMS allowed us to measure O and C isotopic composition of each band in a single calcite grain. Measured O isotopes highlight no systematic variation, C isotope measurements display seawater range values and O isotope thermometry reveals that carbonation is cold (<20 °C) since the onset of the reaction. Our thermodynamic modeling predicts Ca-rich carbonate crystallization near-surface during seawater influx and hydrothermal fluid discharge. Si-rich phases appear in the system with carbonation front evolution through time and space (e.g., talc).
In the discharge model, this stability field limitation is associated with carbonate dissociation into HCO3 − ,aq as the pH decreases with temperature. In the recharge model, the limitation of the carbonate stability field is associated with anhydrite (CaSO4) formation at temperatures above 100 °C. Anhydrite formation only occurs in the first box of the model at high temperature in the discharge model (250 °C). We summarize the petrological, geochemical and numerical modeling results of this study in Figure 11, where the differences between recharge and discharge regarding carbonate precipitation are highlighted.
According to our petrological and modeling results, we wonder why there are no cold ophicarbonates sampled yet in the Alpine ophiolites, and we propose that, in the case of absence of post-rift melting, O and C isotopic composition might have been re-equilibrated during Alpine deformation.
We show that carbonation of ultramafic rocks is more efficient (lower water to rock ratio) at high temperature (>100 °C) since Mg-bearing carbonates can be formed. This result may guide the development of future engineering solutions for CO2 sequestration. Figure 11. Conceptual model for the relationship between carbonation, hydrothermal circulation and magmatism. Two different models are proposed without (a) and with (b) magmatic extrusions. They both include recharge and discharge zones. Low-temperature carbonation (<50 °C) occurs in the recharge zones leading to Ca(±Mg)-carbonate formation (c). High-temperature carbonation starts at 150 °C in the discharge zones where Mg-carbonates only precipitate (c). Carbonation induces more magnesium and silica transport in the discharge zones leading to serpentine replacement by talc and then quartz as the water to rock ratio increases. These elements may precipitate at the surface in hydrothermal chimney as it is observed at the Lost City Hydrothermal Field [23]. The discharge model requires large scale fluid transport in permeable zones whereas carbonation in the recharge model is widespread and occurs through pervasive fluid transport. Figure 11. Conceptual model for the relationship between carbonation, hydrothermal circulation and magmatism. Two different models are proposed without (a) and with (b) magmatic extrusions. They both include recharge and discharge zones. Low-temperature carbonation (<50 • C) occurs in the recharge zones leading to Ca(±Mg)-carbonate formation (c). High-temperature carbonation starts at 150 • C in the discharge zones where Mg-carbonates only precipitate (c). Carbonation induces more magnesium and silica transport in the discharge zones leading to serpentine replacement by talc and then quartz as the water to rock ratio increases. These elements may precipitate at the surface in hydrothermal chimney as it is observed at the Lost City Hydrothermal Field [23]. The discharge model requires large scale fluid transport in permeable zones whereas carbonation in the recharge model is widespread and occurs through pervasive fluid transport. According to our petrological and modeling results, we wonder why there are no cold ophicarbonates sampled yet in the Alpine ophiolites, and we propose that, in the case of absence of post-rift melting, O and C isotopic composition might have been re-equilibrated during Alpine deformation.
We show that carbonation of ultramafic rocks is more efficient (lower water to rock ratio) at high temperature (>100 • C) since Mg-bearing carbonates can be formed. This result may guide the development of future engineering solutions for CO 2 sequestration.
Supplementary Materials: The following are available online at http://www.mdpi.com/2075-163X/10/2/184/s1; Table S1: Compilation of carbon and oxygen isotope composition on calcite in serpentinized peridotites collected on the seafloor and in metamorphic environments. Table S2: Carbon and oxygen isotope composition on calcite from IODP Site 1277. Table S3: Electron microprobe analyses of serpentine and carbonate, and seawater composition used in thermodynamic modeling.