Ankerite (Ank) grains frequently contain cores of almost pure dolomite (Dol) in the garnet zone and at lower grades of three areally extensive stratigraphic units from both Buchan and Barrovian terrains in northern New England, USA. Dol cores and Ank rims have Fe/(Fe+Mg) in the range 0.006-0.011 and 0.16-0.23, respectively. Ank grains with dolomite cores have many features closely analogous to K-feldspar rims around albite cores produced experimentally by incomplete hydrothermal reaction between albite and KCl solution [1]: (a) Ank-Dol contacts are irregular but sharp at the micron scale, (b) The orientations of the Ank and Dol crystal lattices are identical, (c) Fe/(Fe+Mg) of Ank is often slightly greater at the Ank-Dol contact than further away, (d) The Ank-Dol contact is often decorated with open pores. Based on these similarities, Ank appears to have replaced detrital Dol by a dissolution-reprecipitation mechanism during diagnesis or very low grade regional metamorphism.

This chapter contains sections titled: Introduction Prebiotic mineral evolution I – evidence from meteorites Prebiotic mineral evolution II – crust and mantle reworking The anoxic Archean biosphere The Great Oxidation Event A billion years of stasis The snowball Earth The rise of skeletal mineralization Summary Acknowledgements References

The X-CO2 recorded by mineral-fluid equilibria in contact metamorphosed siliceous carbonates commonly defines two groups of rocks in the same aureole. One group records relatively low X-CO2 that results from infiltration of chemically reactive H20-rich fluid. The other records relatively high X-CO2, up to 0.99, that results from decarbonation reactions with little or no infiltration. A complementary dichotomy in apatite compositions exists in five contact aureoles in Italy, Scotland, and the U.S.A. Apatite in the low-X-CO2 group is close to an F-OH solid solution. Apatite in the high-X-CO2 group is a relatively Cl-rich Cl-F-OH solution. The halogen content of fluid coexisting with analyzed apatite was characterized in two aureoles to determine the origin and significance of the dichotomy in apatite composition. Calculated a(HF)/a(H2O), a(HF)/a(HCl), a(HF), and mF(T) (the total F molality of fluid) are systematically higher in fluid coexisting with the low-X-CO2 group. In contrast, a(HCl)/a(H2O) in the high-X-CO2, group may be higher than or overlap with a(HCl)/a(H2O) in the low-X-CO2 group. Calculated a(HCl) and M-ClT in the high-X-CO2 group are lower than or overlap with a(HCl) and M-ClT in the low-X-CO2 group. The Cl-rich apatites in the high-X-CO2 group are explained by crystallization at relatively low a(H2O), a(HF) and mFT rather than at high a(HCl) or M-ClT. In comparison, the F-OH apatites in the low-X-CO2 group formed by infiltration of rock by and equilibration with relatively H2O-rich, high mf(T)/MClT fluid, reflecting the same metasomatic process responsible for F-rich humite-group, minerals and skams in many contact aureoles. Calculated halogen contents indicate that the non-CO2 fraction of fluid in equilibrium with both groups had modest, seawater-like salinity, and that the reactive H2O- and F-rich fluid that infiltrated the low-X-CO2 group had a plutonic source.

Progress ( j) of the infiltration-driven reaction, 4olivineþ 5CO2 þ H2O 1⁄4 talc þ 5magnesite, that occurred during Barrovian regional metamorphism, varies at the cm-scale by a factor of 3 5 within an 3 m volume of rock. Mineral and stable isotope compositions record that XCO2, d Ofluid, and d Cfluid were uniform within error of measurement in the same rock volume. The conventional interpretation of small-scale variations in j in terms of channelized fluid flow cannot explain the uniformity in fluid composition. Small-scale variations in j resulted instead because (a) reactant olivine was a solid solution, (b) initially there were small-scale variations in the amount and composition of olivine, and (c) fluid composition was completely homogenized over the same scale by diffusion–dispersion during infiltration and subsequent reaction. Assuming isochemical reaction, spatial variations in j image variations in the (Mg þ Fe)/ Si of the parent rock rather than the geometry of metamorphic fluid flow. If infiltration-driven reactions involve minerals fixed in composition, on the other hand, spatial variations in j do directly image fluid flow paths. The geometry of fluid flow can never be determined from geochemical tracers over a distance smaller than the one over which fluid composition is completely homogenized by diffusion–dispersion.

Inverse calculations reveal the three-dimensional geometry of time-integrated fluid flux over a 120 km(2) area during peak Barrovian regional metamorphism in southeastern Vermont. Prograde changes in whole-rock CO2, O-18, and C-13 and calculated fluid compositions at the peak of metamorphism were inverted assuming tracer mass balance to obtain the time-integrated fluid flux in three dimensions. Peak metamorphic fluid How was spatially nonuniform with flux magnitudes ranging from similar to0 to 3.10(5) Mol fluid/cm(2) rock and flux directions ranging from vertical (upward and downward) to horizontal. Averaged over the entire study area, the magnitude of the time-integrated metamorphic fluid flux vector is similar to3.4.10(4) moll fluid/cm(2) rock. The average flux vector trends 45degrees to the southwest and points upward at 36degrees from the present horizontal, parallel to formation boundaries on a regional scale. Fluids in the terrain carried similar to3.10(3) moll CO2/cm(2) rock toward Earth's surface during the peak of metamorphism. Results suggest that local cross-layer transport processes are secondary to terrain-scale metamorphic fluid flow in driving prograde decarbonation reactions. Regional structure exerts a first-order control on the gross geometry of peak metamorphic fluid flow.

The P-T conditions of both equilibria were determined precisely by reversal experiments in a piston-cylinder apparatus. On the basis of 8 experiments, brackets for the rutile-magnesite-geikielite equilibrium are 7.0-7.1 kbar at 800 degreesC, 8.6-8.7 kbar at 850 degreesC, and 10.5-10.7 kbar at 900 degreesC. On the basis of 9 experiments, brackets for the zircon-magnesite-baddeleyite-forsterite equilibrium are 7. 1-7.7 kbar at 800 degreesC, 9.2-9.4 kbar at 850 degreesC, and 10.7-10.9 kbar at 900 degreesC. Considering experimental uncertainties in P (+/-300 bars) and T (+/- 3 degreesC), equilibrium curves calculated from both the Berman and the Holland and Powell databases pass through all brackets. Molar Gibbs free energy of formation from the elements at 1 bar and 298 K for geikielite and zircon, derived front the experiments and consistent with the Berman database, are -1481.94 +/- 0.67 kJ and - 1917.54 +/- 1.25 kJ respectively. Corresponding values consistent with the Holland and Powell database are -1479.30 +/- 0.74 kJ and -1918.47 +/- 1.49 kJ. Application of the two equilibria indicate that: (1) the mole fraction of CO2 in fluid was 0.54-1.00 when geikielite and baddeleyite formed during contact metamorphism of siliceous dolomites in the Ballachulish aureole, Scotland;. (2) the activity of CO2 could have been as low as 2.10(-5) during ultra-high pressure metamorphism of magnesite-bearing ecologites and (3) the activity of CO2 was <0.18 during one instance of mantle metasomatism.

Abstract Forsterite in metamorphosed siliceous dolomites commonly contains calcite inclusions that are significantly more magnesian than even the most-magnesian calcite grains in the matrix of the same rock. Calcite inclusions in forsterite from dolomite-bearing rocks therefore have an unexploited potential for providing more-accurate estimates of the peak T of metamorphism than do matrix calcite grains. Some consequences for estimating fluid pressure and for the development of models of heat and fluid transport during contact metamorphism are reviewed. Measured compositions of calcite inclusions in forsterite from five contact aureoles indicate that significant growth of forsterite grains continues over a 50-200 °C range in T after the initial nucleation and growth of the mineral during heterogeneous mineral-fluid reaction. The continued growth may occur by an annealing mechanism driven by reduction in interfacial free energy.

Quartz-calcite sandstones experienced the reaction calcite + quartz = wollastonite + CO2 during prograde contact metamorphism at P = 1500 bars and T = 560 degreesC. Rocks were in equilibrium during reaction with a CO2-H2O fluid with XCO2 = 0.14. The transition from calcite-bearing, wollastonite-free to wollastonite-bearing, calcite-free rocks across the wollastonite isograd is only several millimeters wide. The wollastonite-forming reaction was driven by infiltration of quartz-calcite sandstone by chemically reactive H2O-rich fluids, and the distribution of wollastonite directly images the flow paths of reactive fluids during metamorphism. The mapped distribution of wollastonite and modeling of an O-isotope profile across a lithologic contact indicate that the principal direction of flow was layer-parallel, directed upward, with any cross-layer component of flow <0.1% of the layer-parallel component. Fluid flow was channeled at a scale of 1-100 m by pre-metamorphic dikes, thrust and strike-slip faults, fold hinges, bedding, and stratigraphic contacts. Limits on the amount of fluid, based on minimum and maximum estimates for the displacement of the wollastonite reaction front from the fluid source, are (0.7-1.9) x 10(5) cm(3) fluid/cm(2) rock. The sharpness of the wollastonite isograd, the consistency of mineral thermobarometry, the uniform measured O-18-O-16 fractionations between quartz and calcite, and model calculations all argue for a close approach to local mineral-fluid equilibrium during the wollastonite-forming reaction.

Patterns in the occurrence of minerals in metamorphic rocks suggest additional opportunities for investigating chemical and physical processes during metamorphism. Three such patterns are reviewed. First, trace minerals in metamorphic rocks commonly occur with regular distributions indicating their participation in prograde reactions that can be mapped as isograds. Examples include the distribution of allanite and monazite in pelitic rocks and of zircon and baddeleyite in siliceous dolomites. Recognition of these isograds points to the potential for developing a chronology of specific chemical reactions during metamorphism and for defining the P-T conditions of those reactions. Second, the mineralogical products of retrograde metamorphism in many cases occur in distinctive associations that are consistent with partial mineral-fluid equilibrium. Examples include the distribution of retrograde calcite, quartz, and tremolite in siliceous limestones and of retrograde tremolite, dolomite, brucite, and serpentine in siliceous dolomites from contact aureoles. Among other things, application of partial equilibrium to retrograde metamorphic rocks leads to constraints on the amount and direction of fluid flow in contact aureoles as they cool. Third, pseudomorphs are typically absent from prograde metamorphic rocks but are common in retrograde metamorphic rocks. The distribution may be explained by the effect of "force of crystallization." The pattern of occurrence of pseudomorphs thus suggests novel phenomena during metamorphism that develop from an interplay between chemical and mechanical processes.

▪ Abstract Stable isotopic, mineralogical, and chemical alteration in metamorphic terranes is evidence for reactive fluid flow during metamorphism. In many cases, the amount and spatial distribution of the alteration can be quantitatively interpreted using transport theory in terms of fundamental properties of metamorphic flow systems such as time-integrated flux, flow direction, and Peclet number. Many estimates of time-integrated flux in the upper and middle crust are surprisingly large, 105–106 cm3 fluid/cm2 rock; estimates for the lower crust are much smaller. Rather than pervasive and uniform, reactive fluid flow in all metamorphic environments is channelized on scales of <1–104 m. Channelization results from heterogeneous permeability structures controlled by features such as lithologic layering, contacts, folds, fractures, and faults. Consequently flow may be in the direction of either decreasing or increasing temperature or isothermal. Site-specific thermal-hydrologic models of metamorphic terranes that explicitly consider chemical reactions and dynamic permeability structures will help resolve outstanding questions with regard to the driving forces and duration of flow, metamorphic permeability distributions, and how deformation controls fluid flow.

White mica from the Liassic black shales and slates in Central Switzerland was analysed by transmission electron microscopy (TEM) and electron microprobe to determine its textural and compositional evolution during very low‐grade prograde metamorphism. Samples were studied from the diagenetic zone, anchizone and epizone (T ≈100°–450 °C). Phyllosilicate minerals analysed include illite/smectite (I/S), phengite, muscovite, brammallite, paragonite, margarite and glauconite. Textural evolution primarily is towards larger, more defect‐free grains with compositions that approach those of their respective end‐members. The smectite‐to‐illite transformation reduced the amounts of the exchange components SiK−1Al−1, MgSiAl−2, and Fe3+Al−1. These trends continue to a lesser degree in the anchizone and epizone. Correlations between the proportion of smectite in I/S and the composition of I/S indicate that smectite layers may contain a high layer charge. Illite in I/S bears a compositional resemblance to macrocrystalline phengite in some samples, but is different in others. Paragonite first appears in the upper diagenetic zone or lower anchizone as an interlayer‐deficient brammallite, and it may be mixed with muscovite on the nanometre scale. Owing to the small calculated structure factor for paragonite‐muscovite superstructures, conventional X‐ray powder diffraction cannot distinguish between mixed‐layer structures and a homogeneous compositionally intermediate solid solutions. However, indirect TEM evidence shows that irregularly shaped domains of Na‐ and K‐rich mica exist below 10 nm. Subsequent coarsening of domains at higher grades produced discrete paragonite grains at the margins of muscovite crystals or in laths parallel to the basal plane of the host muscovite. Margarite appears in the epizone and follows a textural evolution similar to paragonite in that mixtures of margarite, paragonite, and muscovite may initially occur on the nanometre scale. However, no evidence of interlayer‐poor margarite has been found.

Oxygen, carbon and strontium isotopic profiles across the margin of the Waterville limestone member are used to investigate advective and diffusive transport during metamorphism of the Waterville Formation in south-central Maine, USA. Rb-Sr isotopic systematics were homogenized on the similar to 10cm hand-specimen scale at ages that are within error of the 376 +/- 6 Ma RbSr whole-rock age of the syn-metamorphic Hallowell pluton. This ir consistent with a plutonic heat source for this low-pressure andalusite- and sillimanite-grade Acadian metamorphic terrane. Advective dis placements of all three isotope profiles at the garnet-grade Blue Rock Quarry indicate fluid flow to the east into the limestone, and the oxygen-isotope profile implies a time-integrated fluid flux of 3.2 +/- 1.4 m(3)/m(2) (2 sigma error). This moss layer flux is insufficient to cause the observed reaction prepress of the muscovite + ankerite + quartz to biotite + anorthite (in plagioclase) + calcite reaction in the similar to 100m thick Waterville limestone member and much of the fluid flow responsible may have been layer parallel. The Isotope profiles indicate advective-diffusive homogenization over distances of 1.5m (delta(13)C) to 6m (delta(18)O) and such homogenization distances are difficult to reconcile with observations of order of magnitude variations in reaction progress on the centimetre scale or less. It is Possible that infiltration occurred during events short lived compared with diffusion, that the reactions started at different temperatures dependent on bulk composition or that diffusion of water from layers with less reactants to layers with more reactants war important in driving the biotite-producing reaction. However variations of fluid composition inferred from the mineral assemblages are apparently inconsistent with diffusion driving reaction progress, and models of precursor arsemblages do not indicate significant compositional control of the temperature of the first appearance of biotite in the rocks. Irrespective of the details of flow and diffusive exchange on the centimetre scale, the average reaction progress in the Waterville limestone member requires significant layer-parallel fluid fluxes.

Isograds were mapped in siliceous dolomites from the Ballachulish aureole on the basis of the formation of geikielite (MgTiO3), baddeleyite (ZrO2), and qandilite (Mg2TiO4) by the following model reactions: rutile + dolomite = geikielite + calcite + CO2, zircon + 2 dolomite = baddeleyite + forsterite + 2 calcite + 2 CO2, and geikielite + periclase = qandilite. The (T, X(CO2)) conditions of reaction inferred from (1) mineral equilibria in pelitic rocks, (2) calcite + dolomite thermometry, and (3) the diopside + dolomite + forsterite + calcite + tremolite and dolomite + periclase + calcite equilibria are: geikielite isograd (640-655 degrees C, 0.76-0.80); baddeleyite isograd (660-710 degrees C, 0.76-0.95); qandilite isograd (725-755 degrees C, <0.08). These T-X(CO2) conditions for the geikielite and baddeleyite isograds are the same within error of those independently estimated from rutile + dolomite + geikielite + calcite and zircon + dolomite + baddeleyite + forsterite + calcite equilibria using Berman's thermodynamic data base. Rutile, zircon, geikielite, and baddeleyite are common in dolomites from at least two other contact aureoles in Scotland and Montana. Although the minerals occur in concentrations <0.01%, they appear to have been in local equilibrium during metamorphism both with each other and with coexisting carbonates and silicates and are potentially useful but currently unexploited records of the physical conditions of metamorphism.

Field, petrolopic and geochemical data were used to characterize fluid infiltration and partial melting during metamorphism of pelitic rocks in the contact aureole of the Onawa pluton, central Maine, USA. Mineral assemblages delineate five metamorphic Zones within the contact aureole: chlorite zone, andalusite-cordierite (a-c) zone, alkali feldspar zone, sillimanite zone and leucocratic-vein (l-v) zone, The sequence of observed mineral assemblages and mineral-fluid reactions calculated by mass balance is similar to those observed in other contact aureoles. Pressure of contact metamorphism is similar to 3 kbar, an the basis of optimum geothermobarometry calculations. Metamorphic temperatures vary from similar to 500 degrees C in the andalusite-cordierite zone to similar to 650 degrees C in the leucocratic-vein zone. Data front field observations, mineral textures, observed reaction stoichiometry, geothermometry and major-element geochemistry suggest that the leucocratic veins of the l-v zone represent crystallized, partial melts. Two overall calculated mineral reactions are responsible for vein formation:qtz + alk fsp + plag + biot + Al-silicate --> cord + ilm + hydrous leuc vein (R9)qtz + plag + biot + Al-silicate --> cord + alk fsp + ilm + hydrous leuc vein (R10)which can be modeled as combinations of two NKFMTASH melting reactions:qtz + alk fsp + plag + biot + sill + H2O --> ilm + H2O-saturated melt (M1)Progress of (M1) and (M2) was measured in eight samples, and reaction (M1) is the dominant melt-forming reaction in all samples. Partial melting (and vein formation) was therefore driven by infiltration of the l-v zone by H2O-rich fluids. Calculated time-integrated fluid fluxes for l-v zone samples range from 0 . 9 x 10(4) to 3 . 1 x 10(4) mol/cm(2), and flow was in the direction of increasing temperature.

Contact-metamorphic assemblages in ophicarbonate from the Bergell aureole correspond either to model isobaric invariant T-X(CO2) points [Atg-Cal-Di-Tr-Fo (6 samples) and Atg-Cal-Tr-Fo-Dol (2)] or to isobaric univariant T-X(CO2) curves [Tr-Cal-Di-Atg (18), Tr-Dol-Atg-Cal (I), Atg-Cal-Fo-Di (1), and Atg-Cal-Tr-Fo (1)]. Calcite-dolomite thermometry and mineral-fluid equilibria in the invariant assemblages record T = 440-540 degrees C at P = 3.5 kbar. Epuilibrium metamorphic fluids were very H2O rich with X(CO2) = 0.001-0.027. In the invariant assemblages Tr + Fo were produced by prograde decarbonatian-dehydration reactions. In contrast, measured modes and reaction textures in samples with univariant assemblages indicate that Tr was produced by carbonation reactions. The apparent paradox of simultaneous decarbonation reactions in the model isobaric invariant assemblages and carbonation reactions in univariant assemblages is resolved by local mineral-fluid epuilibrium and fluid flow through ophicarbonates in the direction of decreasing temperature as the aureole heated. Time-integrated flux (q) was computed from measured reaction progress in 28 samples for models of bath horizontal and vertical down-temperature flow. Results are similar with q decreasing rapidly from (0.2-5.1) x 10(5) cm(3) fluid/cm(2) rock 1.3-1.7 km from the intrusion to 0-0.6 x 10(5) cm(3)/cm(2) at 1.8-4.0 km. The decrease in q is more consistent with vertical than horizontal flow. Variations in time-integrated flux of more than an order of magnitude are recorded by samples from the same outcrop. The absence of carbonate in adjacent metaperidotite indicates that flow was confined to the ophicarbonate. Channelized spatially heterogeneous, vertical flow can be explained by the brecciation and strong vertical foliation of the ophicarbonate relative to surrounding massive metaperidotite. Generation of metamorphic fluids by decarbonation-dehydration reactions within the ophicarbonates explains larger average flux 1-2 km from the intrusion compared with more distal points.

Modal, mineral chemical, and whole-rock chemical data for 266 samples of pelitic schist, micaceous sandstone, and micaceous limestone were used to determine the amount and direction of fluid flow during Buchan-type regional metamorphism in south-central Maine and during Barrovian-type regional metamorphism in east-central Vermont. Metamorphism occurred in Maine at pressure almost-equal-to 3.5 kb (10-15 km depth) and temperature increasing from almost-equal-to 400-degrees-C in the biotite zone to almost-equal-to 550-degrees-C in the sillimanite zone. Metamorphism in Vermont occurred at pressure almost-equal-to 7 to 8 kb (25-30 km depth) and temperature increasing from almost-equal-to 475-degrees-C in the biotite zone to almost-equal-to 550-degrees-C in the kyanite zone. Field, petrologic, and isotopic data suggest that pervasive fluid flow associated with prograde mineral reactions in maine and Vermont was mostly parallel to lithologic layering, subhorizontal, and in the direction of increasing temperature. Calculated amounts of fluid flow, recorded by progress of prograde devolatilization reactions in the different lithologies and measured as a volumetric time-integrated flux, are similar in both areas and lie in the following ranges: pelitic schists, (1-26) . 10(5) cm3 fluid/cm2 rock; micaceous sandstones, (1-42) . 10(5) cm3/cm2; micaceous limestones, (0.02-24) . 10(5) cm3/cm2. Evaluation of errors introduced from a variety of sources suggests that calculated values are probably uncertain by no more than an order of magnitude. Substantial reactive fluid flow appears to be a fundamental aspect of regional metamorphism to depths of at least 30 km in northern New England at grades ranging from those of the biotite zone to those of the sillimanite zone. Although flow was not excluded from any particular rock type, geometry was greatly controlled at the outcrop scale by enhanced flow in more permeable layers (such as pelitic schists) and restricted flow in other layers (such as relatively pure carbonate rocks). In situ metamorphic rock permeability, estimated from values of time-integrated flux and Darcy's law, is in the range 6 . 10(-16) to 1 . 10(-12) cm2 (0.06-100 mud) with an uncertainty of +/-2 orders of magnitude. Metamorphic fluid flow was an essential driving force of prograde mineral reactions but probably had negligible influence on the thermal history of either terrane.