Mass transfer and fluid evolution in late-metamorphic veins, Rhenish Massif (Germany): insight from alteration geochemistry and fluid-mineral equilibria modeling

Abstract

Element mobility and fluid-rock interaction related to the formation of late-metamorphic quartz veins have been studied by combination of mineral chemistry, whole-rock geochemistry, mass balance analysis and fluid-mineral equilibria modeling. The quartz veins are hosted by very low-grade metasedimentary rocks of the fold-and-thrust belt of the Rhenish Massif (Germany). The veins record two stages of evolution, a massive vein filling assemblage with elongate-blocky quartz, chlorite, apatite and albite, and a later open space filling assemblage with euhedral crystals of quartz, ankerite-dolomite and minor calcite and sulfides. Detailed mass balance analysis of an alteration profile adjacent to a representative quartz vein demonstrates that element mobility is restricted to the proximal zone. The most important element changes are gain of Ca, Fe, Mg, Mn, P and CO₂, and loss of Si, K and Na. The data demonstrate that wall-rock carbonation is one of the main alteration features, whereas mobility of Si, K and Na are related to dissolution of quartz and destruction of detrital feldspar and muscovite. The whole-rock geochemical data, in conjunction with fluid composition data and pressure-temperature estimates, were used as input for fluid-mineral equilibria modeling in the system Si-Al-Fe-Mg-Ca-Na-K-C-S-O-H-B-F-Cl. Modeling involved calculation of rock-buffered fluid compositions over the temperature interval 100–500 °C, and reaction-path simulations where a rock-buffered high-temperature fluid reacts with fresh host-rocks at temperatures of 400, 300 and 200 °C. Calculated rock-buffered fluid compositions demonstrate that retrograde silica solubility is a strong driving force for quartz leaching in the temperature-pressure window of 380–450 °C and 0.5 kbar. These conditions overlap with the estimated temperatures for the initial stage of vein formation. Reaction-path models show that high-temperature alteration can produce the observed silica leaching, suggesting that fast advection of external hot fluids from deeper crustal levels was essential for the early stage of vein formation. Fluid advection must have occurred as multiple pulses, which allowed for periods of influx of fluids that leached quartz, alternating with periods of cooling and quartz precipitation in the veins. Reaction-path models at high temperatures (300–400 °C) do not produce carbonate alteration, whereas fluid-rock reaction at 200 °C produces carbonate alteration, consistent with the temperature estimates for the late-stage vein carbonate assemblage. Comparison between modeling results and geochemical data suggests that the observed alteration features are the product of fluid-rock reaction under conditions where the external fluid gradually cooled down and evolved with time. The results of this study highlight the importance of late-orogenic fluid migration for the formation of quartz vein arrays in fold-and-thrust belts.

Introduction

Metamorphic veins in orogenic belts are hosted by a wide range of rock types ranging from metasediments to metavolcanics and they are one of the most important indicators of mass transfer during deformation and metamorphism (Ague 1994a, b, 2011). Quartz vein arrays provide clear evidence for small- to large-scale crustal fluid flow, and yield key information about the nature and composition of fluids that migrate through the upper and middle crust (Bickle and Mckenzie 1987; Ague 2003; Cox 2007; Penniston-Dorland and Ferry 2008; Van Noten et al. 2011; Jacques et al. 2013). Fluid flow and mass transfer play a significant role in the mineralogical, chemical, and structural evolution of the continental crust (Etheridge et al. 1983; Mullis et al. 1994; Nesbitt and Muehlenbachs 1995; Oliver 1996). Continental collision transports entire crustal terrains to deeper levels where rocks are subjected to higher temperature-pressure conditions, where hydrous minerals, carbonates and sulfides become increasingly unstable and release large volumes of metamorphic fluids (Cartwright and Buick 2000). Development of major shear zones and faults during deformation facilitates drainage of fluids from large rock volumes and successive migration from deeper to shallower levels of the crust (Yardley 1983; Dipple and Ferry 1992; Yardley 1997; Connolly 1997; Jamtveit and Yardley 1997; Bons 2001). Focused fluid flow results in redistribution of heat, isotopes, and chemical species (Yardley 1983, 1997; Bickle and Mckenzie 1987; Chamberlain and Rumble 1988; Connolly and Thompson 1989; Connolly 1997; Ague 2003, 2011), and the presence of fluids impacts on mineral reactions and rock deformation behavior (Etheridge et al. 1983, 1984; Cox et al. 1987; Baumgartner and Ferry 1991; Ferry and Dipple 1991; Oliver 1996; Ague 1997; Cox 2007). Interaction of deep crustal fluids with shallow-crustal rocks typically results in strong chemical potential gradients, which drives the development of alteration haloes and selvages around veins and fault zones. If time-averaged fluid fluxes are sufficiently high, the alteration reactions can lead to major changes in the bulk composition of the wall-rocks (Oliver and Bons 2001).

Detailed studies of alteration haloes using whole-rock geochemistry and stable and radiogenic isotopes allows to quantitatively reconstruct fluid-rock interaction and mass transfer (Cox 2007; Penniston-Dorland and Ferry 2008; Wagner et al. 2010). Small-scale geochemical analysis of cross sections through alteration zones makes it possible to constrain the scale of mass transfer, to calculate the amount of mass that has been exchanged between rock and fluid, and to estimate the volume of fluid involved (Ague 1994a, b, 2011; Mountain and Williams-Jones 1996; Halter et al. 1996). Mass balance analysis allows to discriminate among elements that have behaved rather immobile during fluid-rock interaction and those that were highly mobile, and to quantify gains and losses of mobile elements (Gresens 1967; Grant 1986, 2005). This information constrains whether metamorphic veins have formed in response to crustal-scale fluid advection along temperature-pressure gradients or whether local derivation of material driven by chemical potential gradients was dominant (Jamtveit and Yardley 1997; Oliver and Bons 2001; Masters and Ague 2005). In particular, the presence of large volumes of silica that fill metamorphic quartz veins leads to the question whether they represent fossilized large-scale fluid flow pathways in metamorphic terrains, the product of local mobilization of material from the surrounding host-rocks, or a combination of both (Yardley and Bottrell 1992; Penniston-Dorland and Ferry 2008; Wagner et al. 2010). Resolving external and local contributions to vein-wall-rock systems in metamorphic belts requires to combine geochemical analysis of alteration zones, mass balance calculations, chemical microanalysis of fluid inclusions (Miron et al. 2013; Marsala et al. 2013) and geochemical modeling of fluid-rock interaction (Cleverley and Oliver 2005; Mernagh and Bierlein 2008).

This study reconstructs the fluid-rock interaction processes and fluid evolution of late-metamorphic veins in the low-grade metasedimentary rocks of the fold-and-thrust belt of the Rhenish Massif (Germany). Previous work has addressed the field relations, mineral assemblages, mineral chemistry, vein textures (Wagner and Cook 2000), stable isotope geochemistry (Wagner et al. 2010), pressure-temperature evolution and fluid inclusion compositions (Marsala et al. 2013). The data demonstrate that the veins formed from advecting external fluids that were substantially hotter than the metasedimentary rocks hosting the veins, and that fluid-rock disequilibrium caused wall-rock alteration and mass transfer. Building on this work, the present study combines small-scale whole-rock geochemical analysis of alteration envelopes with fluid-mineral equilibria modeling. The spatially-resolved geochemical dataset is used for quantitative mass balance analysis and evaluation of the scale of element mobility. The whole-rock geochemical data are combined with fluid inclusion compositions to model progressive fluid-rock interaction using a reaction-path Gibbs energy minimization code.

Geological setting

The investigated quartz veins are hosted in very low-grade metasedimetary rocks (metapelites, metasiltstones and metasandstones) of the Hunsrück nappe complex, a major tectono-stratigraphic unit of the fold-and-thrust belt of the Rhenish Massif, SW Germany (Fig. 1). The Rhenish Massif belongs to the external Rhenohercynian zone of the Central European Variscan orogenic belt (Oncken et al. 1999), and is mainly composed of metamorphosed sedimentary and volcano-sedimentary rock units. The rocks formed between Early Devonian and Late Carboniferous, and were deposited during basin development in a passive continental margin setting (Matte 1991; Oncken et al. 1999). The Lower Devonian rocks are sandstones, siltstones and pelites with few intercalations of tuffitic horizons (Franke et al. 1978). Middle-Upper Devonian to Lower Carboniferous units are lithologically more heterogeneous and comprise shallow-marine platform carbonates, volcano-sedimentary rocks, submarine volcanics, pelites and black shales (Franke et al. 1978).

Fig. 1

Geological map of the fold-and-thrust belt of the Rhenish Massif (Germany), modified and redrawn after Oncken et al. (1999). The black star indicates the sampling location (Kaub). The insert shows the large-scale Variscan geodynamic framework of central Europe and the location of the Alpine front. Abbreviations: S.T. Siegen Thrust, B.T. Boppard Thrust, T.T. Taunus Thrust, RH Rhenohercynian zone, ST Saxothuringian zone, MO Moldanubian zone

Progressive NW-SE oriented compression affected the Rhenish Massif during the Late Carboniferous, resulting in a SW-NE strike of the first-order structural features such as the major cleavage, fold axes and imbricated nappes, which are separated by major thrust zones (Fig. 1). The main tectono-stratigraphic units are separated by the Siegen, Boppard and Taunus thrust systems, with the Boppard thrust system and the Taunus thrust system (Fig. 1) bounding the Hunsrück nappe complex to the north and south (Oncken et al. 1999). Progressive deformation resulted in development of the first pervasive cleavage (S₁) related to the major folding and thermal peak, which was partially overprinted by two subsequent generations of incipient crenulation cleavage (S₂ and S₃) (Oncken 1984, 1988). The deformation ages are constrained by K-Ar and Ar-Ar dating of syn-deformational phyllosilicates, showing a systematic decrease from 325 to 300 Ma across the belt (Ahrendt et al. 1983; Plesch and Oncken 1999).

The peak metamorphic conditions were estimated from deformation microtextures, mineral assemblages, coal rank and illite crystallinity data, and generally do not exceed sub-greenschist facies conditions (Oncken 1991; Oncken et al. 1995). The southernmost tectono-stratigraphic unit is the only one that recorded slightly higher metamorphic conditions, up to lower greenschist facies (Oncken 1991; Dittmar et al. 1994), with temperatures and pressures estimated at 320–350 °C and 4–6 kbar (Massone 1995; Oncken et al. 1995). The metamorphic peak conditions in the Hunsrück nappe complex have been estimated from mineral chemistry, vitrinite reflectance and illite crystallinity data, and are in the range of 270–300 °C and 2–4 kbar (Wolf 1978; Oncken 1991). Following essentially orthogonal compressional deformation, the Late-Paleozoic strike-slip regime in Central Europe (Arthaud and Matte 1977; Henk 1997) resulted in rotation of the regional stress field from NW-SE towards E-W (Oncken et al. 1999). As a consequence, the pervasive S₁ cleavage was partially overprinted, resulting in back-rotation of structures and re-activation of fault zones (Oncken 1991). The formation of late-metamorphic quartz vein systems in the Hunsrück nappe complex is related to late-orogenic brittle overprinting structures. The temperature-pressure conditions during the initial stage of vein formation were estimated as 370–420 °C and 0.2–0.7 kbar from chlorite thermometry, oxygen isotope geothermometry and fluid inclusion data (Wagner and Cook 2000; Wagner et al. 2010). Detailed fluid inclusion studies show a systematic decrease in temperature down to 130–240 °C (Wagner and Cook 2000; Marsala et al. 2013). Stable isotope data from veins and related wall- and host-rocks demonstrate both internal and external contributions, and that veins formed in a continuum between rock- and fluid-buffered conditions (Wagner et al. 2010).

Characteristics of quartz veins and their wall- and host-rocks

Field documentation and sampling of quartz veins and their wall- and host-rocks was done in one of the large underground roofing slate mines in the Kaub area (Fig. 1). This allowed to map the distribution of vein systems on a km-scale and to obtain well-documented and fresh sample material of veins in their structural context. Detailed sampling focused on representative large extension veins (Fig. 2a), and on detailed cross sections of their associated alteration envelopes that extended into fresh host-rocks within the same lithological unit.

Fig. 2

Field photographs of the late-metamorphic quartz veins and their host-rocks. a Lens-shaped extension vein, with sharp contacts with the host-rock. b Exposure of host-rock slates with clearly visible S₁ foliation

The host-rocks are mainly composed of a thick sequence of dark grey-brown monotonous slates (Fig. 2b), which are mainly metasiltstones with local intercalations of rather pure metapelites and few horizons of metasandstones. A pervasive slaty cleavage (S₁) is always present in the host-rocks, striking SW-NE with almost subvertical dip. Rarely, a later incipient crenulation cleavage (S₂) is visible, predominantly in the vicinity of fault zones. The slaty cleavage is narrow-spaced and rather continuous in fine-grained rock types (Fig. 3a), but is more widely-spaced and discontinuous in coarser-grained lithologies (Fig. 3b). The foliation in fine-grained rocks is defined by a dense layering of sheet silicate minerals, mainly chlorite and muscovite, in a very fine-grained quartz matrix with porphyroclasts of quartz (Fig. 3c and e), chlorite-mica stacks (Fig. 3c), pyrite aggregates (Fig. 3d), feldspar (Fig. 3e) and patchy carbonates (Fig. 3f). By contrast, the foliation in more coarse-grained rocks shows an alternation of anastomosing thin layers of chlorite + muscovite surrounding lithons composed of quartz and carbonate, with some irregularly distributed porphyroclasts of chlorite, muscovite and quartz. The wall-rocks adjacent to the quartz veins have typically a slightly darker color than the fresh host-rocks, but there are no obvious mineralogical differences between the altered slates and their unaltered precursors.

Fig. 3

Microphotographs in transmitted light illustrating key textures of host-rock slates. a Continuous foliation marked by dense layers of subparallel-oriented phyllosilicates (chlorite and white mica). b Coarsely-spaced foliation with alternation of phyllosilicate layers and quartz-carbonate lithons. c Porphyroclast composed of chlorite (Chl) and muscovite (Mus), surrounded by fine-grained quartz (Qtz) and phyllosilicates that are arranged along the main foliation. d Pyrite porphyroclast composed of recrystallized framboids. e Albite (Ab) porphyroclast with sigmoidal shape. Crossed polarizers. f Partly replaced carbonate (Carb) porphyroclast. Crossed polarizers

The quartz veins occur isolated or in small groups (two to five) and they are always surrounded by a network of smaller veinlets and fractures with similar structure and mineral assemblage. The shape is tabular or lenticular with thicknesses between 15 and 80 cm, and the contacts with the wall-rocks are sharp (Fig. 2a). The veins strike NW-SE with subvertical dip, and they crosscut the main foliation of the host-rocks (S₁) at high angles. Two different successive stages in the development of the veins were recognized, which are the massive vein filling assemblage (elongate-blocky quartz, chlorite, apatite and albite) and the late-stage open vein filling assemblage (euhedral quartz, ankerite-dolomite, calcite, chalcopyrite, and pyrite).

The massive vein filling assemblage represents 90–95 % of the vein volume and is mainly composed of fibrous to elongate-blocky quartz (Bons and Jessell 1997). The quartz typically grows perpendicular to vein-wall-rock contacts and rims slate fragments. These fragments are frequently chloritized and partly replaced by vein minerals (Fig. 4a and b). Crack-seal textures (Ramsay 1980) parallel to the vein-wall-rock contacts are sometimes visible in the elongate-blocky quartz. They are restricted to the early stages of quartz growth and were not found in the central part of the veins (Fig. 4c). Chlorite is always present as inclusions within elongate-blocky quartz, and is the first mineral crystallizing from the fluid or replacing wall-rock fragments. The chlorite shows a characteristic vermicular shape (Fig. 4e), composed of numerous minute crystals (50–150 μm). Euhedral albite crystals are rare and restricted to contacts between vein quartz and wall-rock fragments (Fig. 4d).

Fig. 4

Microphotographs in transmitted light illustrating quartz vein textures. a Elongate-blocky quartz (Qtz) enclosing wall-rock fragments that are altered and partly replaced by quartz and chlorite. b The same image with crossed polarizers. Note the considerable grain-size variation of the elongate-blocky quartz crystals. c Crack-seal texture inside coarse-grained elongate-blocky quartz. The crack-seal texture is defined by multiple bands of solid inclusions. Crossed polarizers. d Albite (Ab) crystals growing at the contact to a wall-rock fragment. Crossed polarizers. e Vermicular chlorite (Chl) intergrown with elongate-blocky quartz. f Euhedral crystals of ankerite-dolomite (Ank) and quartz present as late-stage open-space filling of an open cavity. Crossed polarizers

In the late open-space filling stage, cavities and vugs in the center of the veins were filled by euhedral crystal assemblages. Quartz is the first and most abundant mineral, showing prismatic shapes with rhombohedral terminations. Euhedral to subhedral ankerite-dolomite is intergrown with quartz crystals or has grown on top of them (Fig. 4f). Rare calcite, chalcopyrite and pyrite crystals have grown on the crystal faces of euhedral quartz and ankerite-dolomite. The transition between the massive vein filling and the subsequent open space filling assemblage is rather continuous, as demonstrated by elongate-blocky quartz crystals that continue to grow euhedrally into open fissures and cavities.

Pressure-temperature conditions of vein formation

The formation conditions and pressure-temperature evolution of the quartz veins are well-constrained from fluid inclusion data and other geothermobarometric information. Estimated pressure-temperature conditions for the early vein stage (elongate-blocky quartz, chlorite, albite and apatite) are in the range of 370–420 °C and 0.2–0.7 kbar, based on chlorite thermometry, quartz-apatite oxygen isotope thermometry and fluid inclusion isochores (Wagner and Cook 2000; Wagner et al. 2010). The fluid inclusion data demonstrate fluid cooling down to temperatures of 130–240 °C (pressure corrected homogenization temperatures) for the late-stage open-space filling assemblage (euhedral quartz cystals and ankerite-dolomite in open cavities). Detailed fluid inclusion studies of euhedral crystals from open cavities show that three successive fluid inclusion generations record systematically lower temperatures of 210–225, 148–164 and 124–139 °C. The temperature decrease is paralleled by a decrease in fluid salinity from 4.7–5.3 to 3.1–3.9 and then to 1.4–1.7 wt.% equivalent NaCl (Marsala et al. 2013). Microanalysis of fluid inclusions using laser-ablation inductively-coupled plasma-mass spectrometry (LA-ICPMS) shows that the changes in temperature and bulk salinity are correlated with a decrease in the concentration of many elements that include the alkalis, but also fluid-volatile elements such as B, As and Sb (Marsala et al. 2013). The fluid evolution is interpreted in terms of fluid mixing between a hot metamorphic fluid of moderate salinity and a cooler dilute fluid that was of meteoric origin. The fluid evolution is consistent with vein formation in a shallow crustal environment and records ingression of meteoric water during late-orogenic uplift (Marsala et al. 2013).

Methodology

Mineral chemistry

The major and minor element composition of mineral phases (chlorite, albite, carbonates) in the veins and in the wall- and host-rocks was determined by electron probe micro-analysis (EPMA) using a JEOL-8200 instrument with 5 wavelength-dispersive spectrometers. The operating conditions were 15 kV acceleration voltages and 20 nA beam current for silicates and 2 nA beam current for carbonates. Standards used for silicates were: wollastonite (Si-K_α and Ca-K_α) clinopyroxene (Na-K_α), K-feldspar (K-K_α), hematite (Fe-K_α), corundum (Al-K_α), periclase (Mg-K_α), rutile (Ti-K_α) and pyrolusite (Mn-K_α). Standards used for carbonate analysis were: wollastonite (Ca-K_α), magnesite (Mg-K_α), fluoro-apatite (P-K_α and F-K_α), barite (Ba-L_α), siderite (Fe-K_α), scapolite (Cl-K_α), strontianite (Sr-L_α) and rhodochrosite (Mn-K_α). The trace element composition of chlorite and carbonate phases was determined by LA-ICPMS microanalysis. Analysis used an excimer ArF laser-ablation system (ETH GeoLas prototype), operating at 193 nm wavelength, coupled to a Perkin Elmer Elan 6100 DRC quadrupole ICP mass spectrometer (Günther et al. 1997). Data processing was done with the SILLS software package (Guillong et al. 2008), using NIST-610 as external standard. The CaO concentration obtained from EPMA was used as internal standard for carbonates, whereas the SiO₂ concentration was used for chlorites.

Whole-rock geochemistry

Selected major and trace element concentrations in wall- and host-rock samples were determined with wavelength-dispersive X-ray fluorescence (XRF) analysis using a PANalytical Axios instrument. Major and selected trace elements (Rb, Sr, Ba, Nb, Zr, Hf, Ga, Zn, Cu, Ni, Co, Cr, V, Sc, Pb, Th, U) were measured on fused glass beads obtained from standard lithium-tetraborate assay. Pressed rock powder pills were used to analyze the concentration of S, F, W, Sn, Sb and Cs. In addition, the FeO content was independently determined at the University of Lausanne using wet chemistry and colorimetric methods (Wilson 1960; Johnson and Maxwell 1981), which permitted to calculate the Fe₂O₃/FeO ratio using total Fe obtained from XRF analysis. The carbon content was measured with a UIC Inc. CM 5012 carbon dioxide analyzer using coulometric titration. Total carbon content is analyzed by furnace combustion (CM 5200 Autosampler) and subsequent coulometric determination, whereas the inorganic carbon content is determined by acid digestion (CM 5130). Organic carbon is then calculated as the difference between total carbon and inorganic carbon content. The water content of the samples was not independently measured, but calculated by mass balance from the loss on ignition, the volatile concentrations (CO₂, C_org, S) and the gain on ignition obtained from the FeO and Fe₂O₃ content. LA-ICPMS analysis on the fused glass beads were performed to determine As and rare earth elements concentration, and the same type of analysis was done on the pressed powder pills to obtain Li and B concentration. Instrumentation, analytical procedures and data processing techniques were the same as applied for trace element analysis in minerals.

Illite and chlorite crystallinity

Textured, air dried, < 2 μm fractions of host-rock samples were analyzed to determine the illite and chlorite crystallinity, i.e., the Kübler index (KI) and the Árkai index (AI). The KI represents the full width at half maximum (FWHM) in Δ°2θ for (001) basal reflection of illite/muscovite and the AI the FWHM in Δ°2θ for (001) and (002) basal reflections of chlorite. Sample preparation and instrument settings were the same as described previously (Kisch 1991; Árkai et al. 2000). X-ray diffraction (XRD) patterns were recorded with an automated Bruker (AXS D8 Advance) X-ray powder diffractometer equipped with a Lynxeye superspeed detector. The program EVA 3.0 (Bruker AXS) was used for identifying the mineral phases in the XRD patterns. Data quantification was done with a Rietveld method using the AutoQuan software package.

Geochemical modeling

Fluid-mineral equilibria modeling was performed in the model system Si-Al-Fe-Mg-Ca-Na-K-C-S-O-H-B-F-Cl. The calculations were done with the GEM-Selektor Gibbs free energy minimization package (Wagner et al. 2012; Kulik et al. 2013), which allows to simulate complex multicomponent-multiphase equilibria that simultaneously account for aqueous speciation, mineral solubility, mineral stability and mineral solid-solution behavior. The GEM-Selektor code makes it possible to calculate simple static isothermal-isobaric equilibria, but has also a provision for reaction path simulations such as titration and flow-through models and simple 1D reactive-transport models (Reed 1998).

The thermodynamic dataset covers all main aqueous species, rock-forming minerals and their solid-solutions in the model system. It is based on the model used by Dolejs and Wagner (2008), with additions for aqueous carbon, sulfur, boron and fluorine species and the respective minerals (carbonates, graphite, sulfides, anhydrite, fluorite). Thermodynamic properties of aqueous species were calculated with the Helgeson-Kirkham-Flowers (HKF) equation of state (Helgeson and Kirkham 1976; Helgeson et al. 1981; Tanger and Helgeson 1988). Data for charged and neutral aqueous species (Shock and Helgeson 1988; Shock et al. 1989, 1997; Johnson et al. 1992; Sverjensky et al. 1997) were augmented by critically evaluated data for aluminum (Tagirov and Schott 2001) and boron (Akinfiev et al. 2006) species. The end-member data for rock-forming minerals were taken from the Holland and Powell (1998) database, and the data for sulfur minerals from Evans et al. (2010). Data for fluorite came from Robie and Hemingway (1995).

Activity-composition relationships for mineral solid-solutions were described by ideal, non-ideal symmetric and asymmetric models (Powell and Holland 1993, 1999; Holland and Powell 1996, 1998, 2003; Holland et al. 1998; Coggon and Holland 2002; Wei and Powell 2003; Dale et al. 2005; Warren and Waters 2006; Evans et al. 2010). Aqueous activity coefficients were calculated with the extended Debye-Hückel equation, and the activity coefficient of water solvent from the osmotic coefficient (Helgeson et al. 1981; Walther 1997, 2001). The extended Debye-Hückel model is not strictly valid for saline brines, but thermodynamic analysis of experimental data in moderately concentrated NaCl and KCl solutions (to 4.5 mol kg⁻¹) at temperatures and pressures up to 600 °C and 3 kbar has shown that it represents the data within experimental uncertainty (Sharygin et al. 2002).

Mineral chemistry data

Combined EPMA and LA-ICPMS analysis was performed on representative samples of carbonates (Table 1) and chlorite (Table 2) from the quartz veins. In addition, the major element composition of carbonates, chlorite, white mica and albite (Table 3) from host- and wall-rocks was determined by EPMA, but trace elements could not be analyzed due to the small grain size.

Table 1 Representative major and trace element compositions (EPMA and LA-ICPMS data) of vein carbonate minerals (CO₂ calculated by stoichiometry)

Table 2 Representative major and trace element compositions (EPMA and LA-ICPMS data) of vein chlorites (H₂O calculated by stoichiometry)

Table 3 Representative major element compositions (EPMA data) of minerals from altered and unaltered host-rock slate (CO₂ and H₂O calculated by stoichiometry)

All analyzed chlorite compositions including vein and wall-rock chlorites are very similar. The Fe/(Fe + Mg) ratios of vein chlorites lie in a very narrow range of 0.62–0.64, comparable to wall- and host-rock chlorites that have Fe/(Fe + Mg) ratios of 0.60–0.66. Chlorites from veins have low but detectable concentrations of MnO in the range of 0.15–0.22 wt.%, very similar to those of chlorites from the wall- and host-rock samples which are in the range 0.21–0.28 wt.%. Vein-hosted chlorites contain detectable concentrations of Li, As, P, Ti, V, Cr, Co, Ni, and Zn (Table 2). The vein hosted carbonates are essentially ankerite-dolomite, with the Fe/(Fe + Mg) ratio in the range of 0.37–0.53 and low to moderate MnO concentrations of 1.1–2.5 wt.% (Table 1). Carbonates from the host-rocks have similar major element compositions (Table 3), with Fe/(Fe + Mg) ratios of 0.40–0.46 and MnO concentrations of 1.5–2.1 wt.% (Table 2). The vein ankerite-dolomite contains detectable concentrations of Na, Al, Sr, B, Cr and Zn (Table 1). The white mica in the wall-rocks conforms to almost pure muscovite and the feldspars are essentially pure albite.

Alteration geochemistry and mass balance analysis

Analysis of alteration geochemistry has been performed on a representative cross section that extends from the vein-wall-rock contact into the fresh host-rock slate. The cross section has a length of about 30 cm and is associated with a 10–20 cm thick vein. The sample has been cut into 18 slabs in intervals of 1 cm for the first 10 cm, 2 cm for the next 10 cm, and 3 cm for the last 10 cm (Fig. 5). The spatial resolution is finer in the most proximal part of the alteration envelope, because we expected to see more chemical changes in this part of the cross section. Each slab has been analyzed for the full suite of geochemical data, including standard major and trace elements, Fe²⁺/Fe³⁺ ratio, loss on ignition (LOI), CO₂, C_org, S and the REE (Table 4).

Fig. 5

Pictures illustrating the geometry of the studied alteration zone and the sampling pattern. a Image of the wall-rock sample indicating the location of the 18 slabs cut out for whole-rock analysis. b Sketch of the vein-wall-rock geometry illustrating the location of the samples in relation to the quartz vein. S₁ is the main foliation of the host-rocks

Table 4 Composition of each slab sample analyzed from the vein wall-rock cross section KAU-35 (Fig. 5), sorted in order of increasing distance from the vein contact

Plots of the element and oxide concentrations as a function of the distance from the vein contact (Fig. 6) give a first-order indication of element mobility and the spatial extent of the alteration zone. Especially the profiles of SiO₂, (FeO + Fe₂O₃), MgO, CaO, P₂O₅, Li, Rb, Ni, and Zn show concentration changes, and these appear to be restricted to the first 12–14 cm next to the vein-wall-rock contact (Fig. 6). Beyond this point, no significant gains or losses of elements are visible and the concentrations of most elements remain constant within analytical error. For a vein of 10–20 cm thickness (average about 15 cm), the scale of major mass changes is therefore restricted to approximately 15 cm distance from the vein contact. An interesting feature of the alteration profile is the appearance of a sharp peak in the concentrations of a group of elements between 7 and 10 cm distance from the vein contact (Fig. 6). In this interval, the concentrations of Ca, Fe, Mg, Mn, CO₂, P, Li, Sr and Zn show a sharp increase above the curve of the alteration profile.

Fig. 6

Geochemical profiles displaying the concentrations of SiO₂, FeO + Fe₂O₃, MgO, MnO, CaO, Na₂O, K₂O, P₂O₅, CO₂, Al₂O₃, TiO₂, LOI, Li, Rb, Sr, B, Ni, Zn, As, Sb, Ba as a function of the distance from the vein-wall-rock contact

Mass balance calculations were performed along the alteration profile using an immobile element reference frame, following the methods of Grant (1986) and Mountain and Williams-Jones (1996) (Table 5). We selected Zr as the most immobile element, because it shows no systematic changes along the alteration profile, whereas Al that is commonly considered as a suitable immobile element reference frame, clearly shows changes in the cross section (Fig. 6). Calculations using the isocon method (Grant 1986) made it possible to identify additional immobile elements and to obtain the relative mass changes for all the mobile elements. When the average concentrations of elements (or oxides) in the altered wall-rocks (1–15 cm distance from vein) are plotted as function of concentrations in the least altered precursor host-rock (16–29 cm distance from vein), clear trends are visible (Fig. 7). Isocon analysis indicates considerable gain of Fe, Mg, Mn, Ca, P and CO₂, and loss of Si, Na and K. The trace elements gained are Li, Sr, Co, Ni, Zn, Sn, Cu, F, Sb and the REE, while B, Rb, As, Ba, Cr, V, Nb and Sc have been lost.

Table 5 Mass changes for major, trace and rare earth elements along the alteration profile

Table 6 Summary of fluid inclusion compositions, based on data from Marsala et al. (2013), recalculated into molality units (mol kg⁻¹ water solvent)

Table 7 Recalculated rock and fluid compositions used as input in fluid-mineral equilibria modeling

Fig. 7

Isocon diagram comparing the average composition of proximal samples from the alteration envelope (1–15 cm distance from the vein) with the average composition of the least altered slates (16–29 cm distance). Major element oxides are plotted as wt.%. and trace elements as ppm. The isocon line based on the assumption of immobile Zr is given for reference

REE geochemistry

The REE concentrations were normalized to Post-Archean Australian Sediments (PAAS) (McLennan 1989), in order to relate them to a typical crustal sedimentary source. PAAS-normalized REE patterns are essentially flat for the fresh and least altered slate samples (14–29 cm distance from the vein), indicating no REE fractionation in the distal part of the alteration profile (Fig. 8). By contrast, the PAAS-normalized REE patterns of the proximal samples (first 12 cm) show a pronounced bell-shaped enrichment in the middle REE (MREE). These patterns demonstrate significant REE mobility in the proximal part of the alteration profile, and the scale of REE mobility conforms to the scale where mobility of other major and trace elements is observed (Fig. 6). The REE patterns of the proximally altered wall-rocks can be compared to REE patterns of the vein carbonates (Fig. 9) and apatite (Wagner et al. 2010), with vein carbonates (ankerite-dolomite) and apatite showing a similar bell-shaped enrichment in the MREE. The absolute concentrations of all REE are higher in proximally altered wall-rocks compared to the least-altered host-rocks, which can be explained by passive REE enrichment due to silica leaching from altered wall-rocks. The combination of wall-rock carbonation, apatite enrichment and passive REE enrichment is able to explain the strong MREE enrichment and the higher REE concentrations in altered wall-rocks. This interpretation is consistent with the increase in Ca, Fe, Mg, Mn and CO₂ in samples close to the vein-wall-rock contact.

Fig. 8

Shale-normalized (PAAS) rare earth element (REE) patterns of whole-rock samples arranged in sequence from the vein-wall-rock contact (top) to the most distal and least altered host-rock (bottom). The samples in the proximal alteration zone display a bell-shaped enrichment in the middle REE

Fig. 9

Shale-normalized (PAAS) rare earth element (REE) patterns of vein carbonates from the metamorphic quartz veins. a Sample KAU-32. b Sample KAU-19

Illite and chlorite crystallinity

The average value of illite crystallinity (KI) for all samples from the investigated cross section is 0.23 ± 0.01, whereas the average chlorite crystallinity (AI) is 0.17 ± 0.01. When plotted as function of the distance from the vein contact, both the KI and the AI show no systematic change that could be related to effects of fluid-rock interaction (Fig. 10). The average value of the illite crystallinity value of 0.23 would indicate metamorphic conditions of approximately 300 °C, corresponding to the epizone (Kübler and Jaboyedoff 2000).

Fig. 10

Diagram illustrating the variation of illite (KI) and chlorite crystallinity (AI) as a function of distance from the vein-wall-rock contact

Geochemical modeling of fluid-mineral equilibria

The quantitative analysis of alteration geochemistry provides important insight into element mobility and mass transfer during fluid-rock interaction. Combination of the alteration geochemistry data with vein mineralogy, textural characterization, pressure-temperature evolution (Wagner and Cook 2000; Wagner et al. 2010; Marsala et al. 2013) and chemical composition of fluid inclusions (Marsala et al. 2013) results in an unique set of constraints to test geochemical-thermodynamic models for fluid-rock reaction. In turn, thermodynamic modeling of fluid-mineral equilibria makes it possible to link the observed fluid compositions and mineral assemblages with important system parameters such as pressure-temperature conditions, oxidation state and pH, and the fluid/rock ratio. We used simulations of rock-buffered equilibria (at constant low fluid/rock ratio) to evaluate how the fluid composition changes as function of temperature and pressure. We then used the results from the rock-buffered simulations to test if the fluid inclusions were trapped under conditions of local fluid-rock equilibrium or if they would represent an externally derived fluid that had equilibrated with similar host-rocks at higher pressure-temperature conditions. Subsequently, we performed reaction-path simulations at variable fluid/rock ratios (Heinrich et al. 1996) to investigate which alteration features are produced via reaction of the fresh host-rocks with a fluid that has equilibrated with similar rocks at higher pressure-temperature conditions.

Model design and input data

The migration path of aqueous metamorphic fluids in low-grade metasedimentary rocks, from fluid production via mineral dehydration at deeper levels to vein formation in shallower parts of the metamorphic nappe pile, extends over a considerable range in temperatures and pressures. We have therefore modeled the fluid evolution by performing simulations of rock-buffered fluid-mineral equilibria (at a constant mass-based fluid/rock value of 0.1) at several isobars (2.5, 1.0 and 0.5 kbar) and at temperatures of 200–500 °C. The chosen temperature-pressure conditions cover the metamorphic peak conditions estimated for the Hunsrück nappe complex (270–300 °C and 2–4 kbar; Wolf 1978; Oncken 1991) and tectono-stratigraphic units of the Rhenish Massif metamorphosed at higher metamorphic grades (Oncken 1991; Dittmar et al. 1994; Massone 1995; Oncken et al. 1995), as well as the conditions estimated for the late-metamorphic quartz veins (130–400 °C at 0.4–0.6 kbar; Wagner and Cook 2000; Marsala et al. 2013). Reaction-path simulations were performed at variable fluid/rock ratios (ranging from 0.01 to 1000) for temperatures of 200, 300 and 400 °C and a fixed pressure of 0.5 kbar. The reaction-path simulations were run at several temperatures to evaluate the effect of fluid-rock reaction under conditions of a progressively cooling fluid flow system, consistent with the fluid inclusion record.

Rock-buffered and reaction-path simulations were run for contrasting end-member scenarios of reducing and oxidizing fluid conditions, because the oxidation state of the fluid largely depends on the amount of reactive organic matter and pyrite in the host-rocks which is difficult to evaluate. Reducing fluid conditions were produced by including the measured concentration of organic carbon (about 0.3 wt.%) into the bulk rock composition, whereas oxidizing fluids were produced by decreasing the amount of organic carbon by an order of magnitude. These end-member compositions reflect the range in reducing organic carbon measured in the lithologically variable metasedimentary rocks (comprising metapelites, metasiltstones and metasandstones) of the Hunsrück nappe complex. Because of the unusually elevated Al concentrations in the fluid, demonstrated by fluid inclusion studies using LA-ICPMS microanalysis (Marsala et al. 2013), we have also investigated the effect of adding fluorine to the chemical system. Experimental studies have shown that Al solubility is greatly enhanced at temperatures of 400–450 °C in the presence of fluorine, due to the formation of very stable Al and Na-Al hydroxyfluoride complexes (Tagirov et al. 2002a, b). Reflecting the variations in bulk rock input data (reducing or oxidizing conditions) and the chemical system (presence or absence of fluorine), four calculation series of rock-buffered and four calculation series of reaction-path simulations were performed; a summary of the parameters and conditions for each series is given in Tables 6, 7, and 8.

Table 8 Summary of parameters and conditions used for fluid-mineral equilibria calculations

The bulk rock composition used in the rock-buffered simulations was taken from the whole-rock geochemical data of a fresh host-rock sample that is representative for average metasediments of the Hunsrück nappe (Wagner and Cook 2000; Wagner and Jochum 2002). The bulk rock composition used in the reaction-path simulations was taken from the most distant rock sample of the alteration profile. The analyzed bulk rock compositions were converted into the SiO₂-Al₂O₃-Fe₂O₃-FeO-MgO-CaO-Na₂O-K₂O-CO₂-C-H₂O-FeS₂(−CaF₂) composition space and normalized to 100 wt.% (Table 7). This was done by first subtracting TiO₂, MnO and stoichiometric amounts of CaO (to account for modal apatite) from the bulk-rock composition. Subsequently, the analyzed S concentration was converted into a stoichiometric amount of FeS₂, and an equivalent amount of FeO was subtracted. For simulations that considered F, the measured F was converted into CaF₂ and a stoichiometrically equivalent amount of CaO was subtracted. The model fluid input composition for rock-buffered simulations was constructed from the measured average salinity, sulfur and boron concentrations, which are the essential fluid-volatile elements. The fluid composition was recalculated from the microthermometric and LA-ICPMS data for fluid inclusions of the first generation, and expressed as molalities of NaCl, H₂S and H₃BO₃ (Table 6). The input fluid composition for reaction-path simulations was taken from the results of rock-buffered simulations, and represents a multicomponent equilibrium fluid. We selected the fluid composition obtained at 450 °C and 0.5 kbar as starting fluid for the reaction-path simulations, because these conditions reflect the uppermost limit for formation of the quartz veins. Reaction-path simulations were run over a range in initial mass-based fluid/rock ratio from 0.01 to 1000. Due to mineral dissolution and precipitation, the equilibrium fluid/rock ratio is different from the initial fluid/rock ratio, and all diagrams that illustrate the results of reaction-path calculations use the equilibrium fluid/rock ratio.

Results of rock-buffered simulations

Figures 11, 12 and 13 summarize the modeling results obtained at 2.5, 1.0 and 0.5 kbar for the fluorine-free system, including bulk fluid composition, selected aqueous species abundance, pH, oxidation state and mineral stability. The total element solubility curves and the other fluid parameters have essentially similar temperature trends at all three pressures, indicating that pressure effects are minor compared to temperature effects. The only notable exception is shown by Si, which shows a retrograde solubility behavior above 400 °C at 0.5 kbar (Fig. 14), whereas solubility is prograde at all other pressure-temperature conditions investigated. Elements that are affected by variations in the oxidation state are mainly C, Ca and S. The higher C solubility above 340–360 °C in reducing fluids is essentially caused by higher concentrations of CH₄(aq) compared to carbonate species (compare Fig. 11c and d), which impacts on the stability of carbonate minerals in the host-rock (compare Fig. 11k and l). Higher Ca solubility below 300 °C in the oxidizing fluid is due to CaSO₄ being present in higher concentrations at lower temperatures (Fig. 11c), which results in anhydrite saturation (Fig. 11k). As expected, sulfur shows different solubility curves for reducing and oxidizing fluids. Under reducing fluid conditions, the S solubility curve increases smoothly with temperature (Fig. 11b). By contrast, the solubility curve for the oxidizing fluid shows a concave shape with a minimum between 360 and 320 °C and higher values at lower and higher temperatures (Fig. 11a). The high S concentration below 300 °C is caused by high concentrations of sulfate species, mainly CaSO₄ and SO₄ ²⁻ (Fig. 11c and e).

Fig. 11

Calculated rock-buffered fluid composition and whole-rock mineralogy as function of temperature at 2.5 kbar. Left column illustrates results for oxidizing conditions (0.03 wt.% C_org in the rock) and right column those for reducing conditions (0.3 wt.% C_org in the rock). a and b Total molalities of elements in the fluid. c and d Distribution of the main Ca and C species. e and f Distribution of the main S species. g and h Calculated pH compared to the neutral pH value. i and j Oxidation state (logarithm of oxygen fugacity) compared to the fayalite-quartz-magnetite (FQM) and hematite-magnetite (HM) mineral buffers. k and l Stable mineral phases

Fig. 12

Calculated rock-buffered fluid composition and whole-rock mineralogy as function of temperature at 1.0 kbar. Left column illustrates results for oxidizing conditions (0.03 wt.% C_org in the rock) and right column those for reducing conditions (0.3 wt.% C_org in the rock). a and b Total molalities of elements in the fluid. c and d Distribution of the main Ca and C species. e and f Distribution of the main S species. g and h Calculated pH compared to the neutral pH value. i and j Oxidation state (logarithm of oxygen fugacity) compared to the FQM and HM mineral buffers. k and l Stable mineral phases

Fig. 13

Calculated rock-buffered fluid composition and whole-rock mineralogy as function of temperature at 0.5 kbar. Left column illustrates results for oxidizing conditions (0.03 wt.% C_org in the rock) and right column those for reducing conditions (0.3 wt.% C_org in the rock). The curves terminate at 450 °C, because at higher temperatures the water density would be below 0.33 g cm⁻³, outside the permissible range of the HKF equation of state for aqueous electrolytes (Johnson et al. 1992). a and b Total molalities of elements in the fluid. c and d Distribution of the main Ca and C species. e and f Distribution of the main S species. g and h Calculated pH compared to the neutral pH value. i and j Oxidation state (logarithm of oxygen fugacity) compared to the FQM and HM mineral buffers. k and l Stable mineral phases

Fig. 14

Comparison of the total calculated silica solubility in rock-buffered fluids for 2.5, 1.0 and 0.5 kbar. The curve at 0.5 kbar intersects the window of retrograde silica solubility at temperatures between 400 and 450 °C

The pH is not affected by oxidation state and pressure variations, but essentially controlled by buffering silicate and carbonate mineral assemblages in the host-rock (i.e., chlorite, muscovite, albite, ankerite-dolomite, calcite). The pH values are close to neutral at high temperatures and increase to more alkaline with decreasing temperature, ranging between 5 and 7 (Fig. 11g and h). The fluid oxidation state is not significantly affected by pressure, but as expected the oxygen fugacity values are several orders of magnitude higher in the oxidizing fluid (Fig. 11i and j). At high temperatures, the oxidation state for both simulations is more reduced relative to the FQM (Fayalite-Quartz-Magnetite) and HM (Hematite-Magnetite) mineral buffers, and increasingly shifts to more oxidizing conditions with decreasing temperature.

In Fig. 15, selected calculated element ratios are directly compared to the measured values from LA-ICPMS analysis of fluid inclusions. The element ratios and pressure-corrected homogenization temperatures of the fluid inclusions are plotted as average values for each main fluid inclusion generation along with error bars that bracket the minimum and maximum values. The average fluid inclusion temperatures are around 240 °C for fluid generation 1 and decrease to 130 °C for fluid generation 3. As noticed earlier, the Ca solubility behaves differently under oxidizing and reducing conditions. Within the temperature range that is defined by the pressure-corrected homogenization temperatures, the calculated Ca/Na ratios for the oxidizing fluid are slightly higher than the measured ones (Fig. 15a), whereas the Ca/Na ratios for the reducing fluid are close to the measured ones (Fig. 15b). The calculated K/Na and Mg/Na ratios are not affected by the fluid oxidation state. The K/Na ratios are generally in good agreement with the measured fluid inclusion data (Fig. 15c and d). By contrast, the analyzed Mg/Na ratios and fluid inclusion temperatures are not in agreement with the calculated values (Fig. 15e and f). The calculated curves would match with the measured Mg/Na ratios at temperatures of 350–400 °C.

Fig. 15

Calculated element ratios in rock-buffered fluids as function of temperature, compared to average measured fluid inclusion (FI) data (Marsala et al. 2013). Error bars represent maximum and minimum values for all individual fluid inclusions from FI generations 1 to 3. Left column illustrates results for oxidizing conditions (0.03 wt.% C_org in the rock) and right column those for reducing conditions (0.3 wt.% C_org in the rock). a and b Log(Ca/Na). c and d Log(K/Na). e and f Log(Mg/Na)

Adding fluorine to the system results in a substantial increase in Al solubility by one to three orders of magnitude (Fig. 16), whereas the oxidation state of the fluid has no significant impact on Al solubility. The effect of F becomes more important with increasing temperature, and particularly above 350–400 °C the Al concentrations increase dramatically to values that are more than 3 orders of magnitude higher than in the F-free system (Fig. 16a and b). The measured Al concentrations in the fluid inclusions match with the calculated solubilities in the F-bearing system at temperatures of 390–440 °C (Fig. 16a and b). The F solubility curves between 400 and 450 °C show a similar trend as does Al (Fig. 16c and d), suggesting a chemical coupling between the two elements. Inspection of the speciation results shows that Al and Na-Al hydroxyfluoride species, i.e., Al(OH)F₂, NaAl(OH)₃F and NaAl(OH)₂F₂, become increasingly stable as temperature increases, and they dominate Al speciation above about 350 °C (Fig. 16e and f), in agreement with experimental results in the Na-Al-F-O-H system (Tagirov et al. 2002a, b).

Fig. 16

Effect of fluorine addition on calculated rock-buffered fluid compositions, plotted as function of temperature at 0.5 kbar. Left column illustrates results for oxidizing conditions (0.03 wt.% C_org in the rock) and right column those for reducing conditions (0.3 wt.% C_org in the rock). a and b Al concentration in fluorine-bearing system (black curves) compared to the F-free system (grey curves). c and d Total Ca and F concentrations in the fluid. e and f Distribution of the main Al species. Note that solubility enhancement in the F-bearing system is mainly due to the very stable NaAl(OH)₂F₂ complex. g and h Distribution of the main Ca and F species. i and j Stable mineral phases

Results of reaction-path simulations

The main results of the reaction-path simulations are summarized in Figs. 17, 18 and 19. The plots illustrate mineral abundance, relative mass changes, total element solubilities in the fluid, fluid pH and oxidation state as function of increasing fluid/rock ratio for simulations at 400, 300 and 200 °C. The initial mineral assemblage at 400 °C at low fluid/rock ratios comprises quartz, albite, chlorite and muscovite, for both oxidizing and reducing conditions (Fig. 17a and b). In the calculations at 300 and 200 °C, ankerite-dolomite is stable instead of albite (Figs. 18a, b and 19a, b). The main difference between calculations with the reducing fluid compared to those with the oxidizing fluid is the precipitation of large amounts of graphite from the fluid at high fluid/rock ratios. The amount of precipitated graphite increases at lower temperatures, resulting in graphite making up most of the rock mineralogy (Figs. 18b and 19b). In addition, increasing the fluid/rock ratio results in formation of iron sulfides, either pyrrhotite + pyrite at 400 °C (Fig. 17a) or pyrite alone at lower temperatures (Figs. 18a, b and 19a, b). Albite, or ankerite-dolomite at lower temperatures, always reacts out around fluid/rock ratios of about 10, and chlorite disappears after albite or ankerite-dolomite at higher fluid/rock ratios at 300 and 400 °C. In the runs at 200 °C, chlorite disappears at very low fluid/rock ratios, which is related to increasing carbonation (formation of siderite-magnesite) at intermediate fluid/rock ratios (Fig. 19a, b). The formation of iron sulfides usually results in gain of FeO and S to the rock at high fluid/rock ratios, especially under oxidizing conditions that suppress the precipitation of graphite (Figs. 17c, 18c and 19c). The precipitation of graphite in the simulations with the reducing starting fluid adds considerable amounts of C to the rock (Figs. 17d, 18d and 19d). The mass gain of C becomes dominant at the highest fluid/rock ratios, which are probably not attained in the metasediments hosting the vein-wall-rock system because this style of alteration is not observed.

Fig. 17

Calculated fluid composition, mineral assemblage and mass changes as function of fluid/rock ratio at 400 °C and 0.5 kbar. The fluid-rock interaction model assumes that rock-buffered fluid (450 °C) reacts with the fresh host-rock. Left column illustrates results for oxidizing conditions (0.03 wt.% C_org in the rock) and right column those for reducing conditions (0.3 wt.% C_org in the rock). a and b Mineral abundance (wt.%). c and d Percentage mass changes relative to the whole-rock composition of the fresh host-rock. e and f Total molalities of elements in the fluid. g and h Calculated pH compared to the neutral pH value. i and j Oxidation state (logarithm of oxygen fugacity) compared to the FQM and HM mineral buffers

Fig. 18

Calculated fluid composition, mineral assemblage and mass changes as function of fluid/rock ratio at 300 °C and 0.5 kbar. Left column illustrates results for oxidizing conditions (0.03 wt.% C_org in the rock) and right column those for reducing conditions (0.3 wt.% C_org in the rock). a and b Mineral abundance (wt.%). c and d Percentage mass changes relative to the whole-rock composition of the fresh host-rock. e and f Total molalities of elements in the fluid. g and h Calculated pH compared to the neutral pH value. i and j Oxidation state (logarithm of oxygen fugacity) compared to the FQM and HM mineral buffers

Fig. 19

Calculated fluid composition, mineral assemblage and mass changes as function of fluid/rock ratio at 200 °C and 0.5 kbar. Left column illustrates results for oxidizing conditions (0.03 wt.% C_org in the rock) and right column those for reducing conditions (0.3 wt.% C_org in the rock). a and b Mineral abundance (wt.%). c and d Percentage mass changes relative to the whole-rock composition of the fresh host-rock. e and f Total molalities of elements in the fluid. g and h Calculated pH compared to the neutral pH value. i and j Oxidation state (logarithm of oxygen fugacity) compared to the FQM and HM mineral buffers

Mineral transformations produce sharp changes in the solubility curves of some elements. They are more evident in the low-temperature calculations due to the increasing temperature difference between the reacting fluid and the rock (Figs. 18e, f and 19e, f). The concentration of all elements in the fluid is usually constant up to a fluid/rock ratio of 10, when albite or ankerite-dolomite have completely reacted out. Beyond this point, the S solubility always increases and Ca solubility decreases (Fig. 18e and f). The solubility of Mg increases as well, but the trend reverses when chlorite (or siderite-magnesite at 200 °C) disappears (Fig. 18e and f). The pH values remain close to neutral as long as the initial rock-buffer mineral assemblage is stable. Successive mineral transformations shift the pH to more acidic values (Fig. 18g and h). The oxidation state does not appear to be strongly affected by variations in fluid/rock ratio, and the values lie always between the FQM and HM mineral buffers (Fig. 18i and j). Compared to the simulations in the F-free system, the calculation series with F show only small differences, with the most notable one being the precipitation of fluorite at higher fluid/rock ratios.

Discussion and conclusions

The results of our combined alteration geochemistry and modeling study complement the existing datasets on vein mineralogy, pressure-temperature evolution, stable isotope geochemistry and fluid composition, and provide important constraints on the driving forces for fluid-rock reaction and vein formation. The insight gained from the approach adopted in our work results in a multi-faceted picture of the vein-wall-rock relations, and provides useful insight into vein formation processes in very low-grade metamorphic rocks in other fold-and-thrust belts.

Controls on alteration geochemistry

Although there is no macroscopically visible alteration envelope surrounding the veins and no major change in the wall-rock mineralogy, the detailed geochemical mass balance analysis demonstrates element mobility and defines the scale of mass transfer as the first 12–14 cm adjacent to the vein-wall-rock contact (Fig. 6). The essential mass changes are loss of Si, K and Na and gain of Fe, Ca, Mg, CO₂ and P. The element suite that is gained in the alteration halo conforms to the elements that make up the Fe-Mg vein carbonates (ankerite-dolomite) and apatite. This is supported by the REE patterns of vein carbonates and apatite, compared to those of wall-rocks from the proximal part of the alteration envelope which show a similar enrichment in the middle REE. Therefore, carbonation is one of the main alteration reactions in the proximal wall-rock zone. The reaction-path modeling using an oxidizing input fluid produces carbonate alteration at moderate fluid/rock ratios at 200 °C, whereas the simulations at higher temperatures do not produce such carbonate alteration. The temperature window for the carbonation reaction around 200 °C is in good agreement with the temperatures of 120–230 °C estimated for deposition of vein carbonates inside open cavities along with euhedral quartz crystals (Wagner and Cook 2000; Marsala et al. 2013).

Silica leaching is mainly related to quartz dissolution, and decrease in K and Na is likely related to reactions involving detrital feldspars and muscovite-illite. The average amount of silica that was leached from the wall-rocks is on the order of about 5 g per 100 g of precursor host-rock (Table 5). If this amount is compared with the vein thickness of 10–20 cm, leaching from the wall-rock alteration zone is not sufficient to explain the formation of the quartz veins, and an additional external fluid source for silica is clearly required. This finding is in agreement with earlier conclusions based on oxygen isotope systematics suggesting that the vein quartz is partly derived from leaching of the altered wall-rocks and partly from advecting external fluids (Wagner et al. 2010).

The retrograde silica solubility in the temperature window 400–450 °C at 0.5 kbar (Fig. 14) appears to be the essential driving force for silica leaching in the alteration halo, in good agreement with the pressure-temperature estimates for the early stage of vein formation and incipient alteration (Wagner and Cook 2000; Wagner et al. 2010). Taken together, the data are best interpreted in such a way that the early stage of vein formation during late-orogenic uplift is related to advection of external fluids that were supplied from a deeper and substantially hotter source. The external fluids were produced by metamorphic devolatilization at conditions that clearly exceed the peak metamorphic conditions inferred for the rock units of the Hunsrück nappe complex currently exposed at the surface. Considering previous estimates for the initial stage of vein formation (i.e., early elongate-blocky quartz and chlorite), which are around 400 °C and 0.5 kbar (Wagner and Cook 2000), fluid advection from the hot zone of fluid production must have been rather fast and close to adiabatic, preventing considerable heat loss to the surrounding host-rocks. Interaction between the hotter fluid and the colder wall-rocks resulted in quartz leaching, because the pressure-temperature path would intersect the window of retrograde quartz solubility. Because the fluids also deposited large amounts of quartz inside the veins, fluid advection must have occurred as multiple pulses, which allowed for periods of influx of hot external fluids that leached quartz alternating with periods of fluid cooling and quartz precipitation.

Comparison between calculated element ratios and the measured fluid inclusion data gives somewhat inconclusive results. While the analyzed Ca/Na and K/Na ratios would suggest that the fluids are compositionally close to rock-buffered fluids that equilibrated with typical metasedimentary rocks of the Hunsrück nappe complex around 200 °C, the Mg/Na ratios and the Al concentrations are in better agreement with fluid-rock equilibrium having been attained at temperatures of at least 400 °C. These results explain why application of empirical solute geothermometers to element ratios obtained by LA-ICPMS microanalysis of fluid inclusions was not successful (Marsala et al. 2013). The mixed messages obtained from comparison between modeled rock-buffered fluid compositions and analyzed fluid inclusion data might indicate that the fluids have been produced by devolatilization of different source rocks at high temperatures, or that fluid production occurred during a transient stage that did not allow for attainment of full fluid-rock equilibrium. Discrete and transient pulses of fluid production that could result in disequilibrium fluid compositions could relate to prograde metamorphic reactions that involve successive breakdown and dehydration of clay and hydrated mica phases. Experimental and thermodynamic modeling studies of the dehydration behavior of clays and hydrated micas have shown that water loss occurs discontinuously via a series of reactions that result in breakdown of discrete water layers (Vidal and Dubacq 2009; Dubacq et al. 2010). The discontinuous water release results in a sequence of discrete pulses of aqueous fluid production as clay-rich rocks are subjected to a prograde metamorphic gradient.

Controls on Al solubility

The unusually high Al concentrations that were analyzed by LA-ICPMS in the fluid inclusions from quartz veins in the Kaub area, but also in fluid inclusions hosted by late-stage euhedral quartz crystals from other similar veins in the Hunsrück nappe complex (Marsala et al. 2013), can be explained by the strong Al solubility enhancement that is caused by fluorine complexation. The host-rocks contain sufficient F (500–700 ppm) and the low Ca concentrations (below 0.5–1 wt.% CaO) that control fluorine concentrations in the fluid (coupled through the fluorite solubility product) allow for building up sufficiently high fluorine levels in rock-buffered fluids. The substantial increase in Al solubility above 350–400 °C is caused by the high stability of Al and Na-Al hydroxyfluoride complexes, in particular NaAl(OH)₃F and NaAl(OH)₂F₂ (Fig. 16e). The thermodynamic data for these complexes are well-constrained by carefully conducted solubility experiments at 400 and 450 °C and 0.5 and 1.0 kbar (Tagirov and Schott 2001; Tagirov et al. 2002a, b), which clearly demonstrated the combined effect of Na and F on increased Al solubility. The experiments cover exactly the pressure-temperature range that is relevant for of the late-metamorphic quartz veins of our study (Wagner and Cook 2000; Marsala et al. 2013).

By comparison, measured Al concentrations in fluid inclusions from low-grade metasedimentary rocks in the accretionary-wedge of the Swiss Alps (the so-called Bündnerschiefer) are considerably lower (Miron et al. 2013) than those analyzed in fluid inclusions of the quartz veins of the Rhenish Massif (Marsala et al. 2013). The metasedimentary rocks of the Bündnerschiefer are lithologically different from the rocks of the Hunsrück nappe complex and contain large volumes of calc-schists along with metapelites and metasiltstones. Therefore, Ca concentrations in the rocks are much higher (typically 2–4 wt.% CaO), which will result in depressing F concentrations in rock-buffered fluids. This, in turn, will result in lower concentrations of Na-Al hydroxyfluoride species that control the Al solubility budget in fluorine-bearing systems at temperatures above 350–400 °C and low pressures. At higher pressures above 4–5 kbar, Al solubility might be enhanced by Al-Si complexing (Manning 2007; Beitter et al. 2008; Newton and Manning 2008) even in fluorine-poor systems.

Evolution of the vein-wall-rock system

The main features of the alteration envelope are silica leaching and addition of elements related to Fe-Mg carbonate deposition in the wall-rocks. Silica leaching can be essentially explained by the pressure-temperature path of the advecting external fluids intersecting the window of retrograde quartz solubility between 400 and 450 °C (and extending down to 350–360 °C if the fluids were quartz-saturated at 450 °C), and is clearly a feature of high-temperature alteration. In contrast to the Si alteration, the carbonate alteration cannot be explained by high-temperature fluid-rock interaction, regardless whether the fluid is oxidizing or reducing. The reaction-path simulations at 400 and 300 °C do not produce wall-rock carbonation, but rather result in progressive dissolution of ankerite-dolomite that is stable at low fluid/rock ratios. Only low-temperature fluid-rock reaction around 200 °C is able to produce carbonate alteration. This is consistent with the field observations and fluid inclusion data that demonstrate that vein hosted ankerite-dolomite (and younger calcite) precipitated during the late-stage in open vugs and cavities at temperatures of 120–230 °C (Wagner and Cook 2000; Marsala et al. 2013).

Taken together, the field observations, fluid inclusion data and modeling results suggest that alteration and vein formation cannot be explained by a single-stage high-temperature fluid-rock interaction process, but require at least two sequential stages of fluid-rock reaction, with the first stage causing silica leaching at high temperatures and the second stage causing wall-rock carbonation at much lower temperatures. This suggests that quartz vein formation and the associated hydrothermal alteration are the product of multiple fluid pulses reacting with the host-rocks under conditions where the fluid system gradually cooled down with time. The late-metamorphic quartz veins in the fold-and-thrust belt of the Rhenish Massif therefore record a continuum of evolving fluid composition that reflects decreasing metamorphic fluid production, temperature and pressure decrease in response to orogenic uplift, and the additional cooling effect of ingression of meteoric water at shallower depths (Marsala et al. 2013). The key role of an externally derived fluid for formation of the late-metamorphic quartz vein arrays highlights the importance of late-orogenic fluid migration for the evolution of fold-and-thrust-belts.