Magmatic influence on reaction paths and element transport during serpentinization

Chemical Geology 274 (2010) 196–211

Contents lists available at ScienceDirect

Chemical Geology j o u r n a l h o m e p a g e : w w w. e l s ev i e r. c o m / l o c a t e / c h e m g e o

Magmatic influence on reaction paths and element transport during serpentinization Niels Jöns ⁎, Wolfgang Bach, Frieder Klein 1 Department of Geosciences, Klagenfurter Straβe, University of Bremen, 28359 Bremen, Germany

a r t i c l e

i n f o

Article history: Received 25 January 2010 Received in revised form 12 April 2010 Accepted 14 April 2010 Editor: R.L. Rudnick Keywords: Abyssal peridotite Alteration Metasomatism Plagiogranite Serpentinization

a b s t r a c t Small-scale shear zones are present in drillcore samples of abyssal peridotites from the Mid-Atlantic ridge at 15°20′N (Ocean Drilling Program Leg 209). The shear zones act as pathways for both evolved melts and hydrothermal fluids. We examined serpentinites directly adjacent to such zones to evaluate chemical changes resulting from melt–rock and fluid–rock interaction and their influence on the mineralogy. Compared to fresh harzburgite and melt-unaffected serpentinites, serpentinites adjacent to melt-bearing veins show a marked enrichment in rare earth elements (REE), strontium and high field strength elements (HFSE) zirconium and niobium. From comparison with published chemical data of variably serpentinized and melt-unaffected harzburgites, one possible interpretation is that interaction with the adjacent melt veins caused the enrichment in HFSE, whereas the REE contents might also be enriched due to hydrothermal processes. Enrichment in alumina during serpentinization is corroborated by reaction path models for interaction of seawater with harzburgite–plagiogranite mixtures. These models explain both increased amounts of alumina in the serpentinizing fluid for increasing amounts of plagiogranitic material mixed with harzburgite, and the absence of brucite from the secondary mineralogy due to elevated silica activity. By destabilizing brucite, nearby melt veins might fundamentally influence the low-temperature alteration behaviour of serpentinites. Although observations and model results are in general agreement, due to absence of any unaltered protolith a quantification of element transport during serpentinization is not straightforward. © 2010 Elsevier B.V. All rights reserved.

1. Introduction In recent years, there has been an increasing interest in ultramafichosted hydrothermal systems at the seafloor, where peridotites are exposed and interact with seawater (e.g., Gràcia et al., 2000; Kelley et al., 2001; Charlou et al., 2002; Douville et al., 2002; Paulick et al., 2006; Schmidt et al., 2007; Melchert et al., 2008; Klein and Bach, 2009). The resulting serpentinization affects the mineralogy, rheology, seismic and magnetic structure of the oceanic crust (e.g., Miyashiro et al., 1969; Escartín et al., 2001; Fujiwara et al., 2003; O'Driscoll and Petronis, 2009). Additionally, hydration reactions in the abyssal peridotites lead to oxidation of ferrous iron in primary minerals to ferric iron in secondary assemblages (e.g., McCollom and Bach, 2009), and the reducing conditions involved are expressed by formation of hydrogen and methane, which provide fuel for microbial communities (e.g., O'Brien et al., 1998; Kelley et al., 2005; Alt et al., 2007; Delacour et al., 2008). Fluid flow through the oceanic crust causes redistribution of elements. High-temperature fluids have the potential to leach certain ⁎ Corresponding author. Tel.: +49 421 218 65404. E-mail addresses: [email protected] (N. Jöns), [email protected] (W. Bach), [email protected] (F. Klein). 1 Present address: Geology and Geophysics Department, Woods Hole Oceanographic Institution, MA 02543, USA. 0009-2541/$ – see front matter © 2010 Elsevier B.V. All rights reserved. doi:10.1016/j.chemgeo.2010.04.009

elements from crustal rocks, transport them and lead to precipitation of minerals elsewhere. A profound understanding of the underlying processes and pathways of fluid flow is of interest not only for economic geologists, but also for explanation of the fluid chemistry at hydrothermal vents and for estimating the bulk composition of the oceanic crust, the input into subduction zones. Thus, the detailed study of the character and pressure–temperature framework of serpentinization reactions is relevant for understanding the complex interaction among lithosphere, hydrosphere and biosphere, contributing to comprehensive insights into global geochemical cycles. The composition and general anatomy of the oceanic crust is believed to be well understood from seismology as well as petrological and geochemical studies on ophiolite complexes and drillcore samples (e.g., Anonymous, 1972; de Wit and Stern, 1976). However, the more detailed the studies get, the more they add knowledge about inhomogeneities to the model of a laterally continuous oceanic crust. Such features, which are particularly common in slow-spreading crust, include chemical heterogeneities, e.g., ultramafic oceanic core complexes (Boschi et al., 2006; Smith et al., 2006; Ildefonse et al., 2007) and intrusions of evolved melts (Coleman and Peterman, 1975; Aldiss, 1981; Koepke et al., 2007; Jöns et al., 2009), but also structural heterogeneities such as detachment faults and fracture zones. Concerning fluid flow and element transport, both sources of heterogeneities are important to consider. Fracturing promotes and focuses fluid flow and thus, to some extent, countervails the volume expansion that accompanies mineral

N. Jöns et al. / Chemical Geology 274 (2010) 196–211

hydration reactions. Fracturing therefore allows for fluid-dominated alteration regimes (high water/rock mass ratios), whereas hydration reactions under static conditions promote rock-dominated alteration (lower water/rock mass ratios). Additionally, intrusions of melt generally follow pre-existing fractures (e.g., Kaczmarek and Müntener, 2008; Jöns et al., 2009), resulting in co-occurring chemical and structural heterogeneities. Assuming fluid flow along melt-influenced fractures, the chemical characteristics of the fluid are likely dominated by the interaction with a gabbroic or more evolved melt, even though the host rocks are of distinct chemical composition (e.g., abyssal peridotites). In the present study, we examine the mineral and bulk-rock chemistry of serpentinized abyssal harzburgites. These rocks host shear zones with abundant magmatic material and are thus an ideal object for studying the combined mineralogical and geochemical effects of serpentinization, magmatism and fluid flow, which might eventually influence vent fluid chemistry of associated hydrothermal systems. 2. Geological setting Samples for this study were drilled during ODP Leg 209 (Kelemen et al., 2004), which is located at the slow-spreading Mid-Atlantic ridge (MAR) near the prominent 15°20′N fracture zone (Fig. 1). Leg 209 drilled mainly abyssal peridotites and gabbros exposed in the footwalls of major detachment faults on the flanks of the MAR. Fluids issuing from the active Logatchev hydrothermal field, located between Sites 1268 and 1270, are characterized by high concentrations of dissolved dihydrogen and methane, indicative of peridotite–seawater interactions. The widespread fluid–peridotite interaction is also evidenced by H2 and CH4 anomalies in the water column (Charlou et al., 1991; Bogdanov et al., 1997). Based on structural and lithological observations, Schroeder et al., (2007) interpret the MAR in the 15°N area as a nonvolcanic ridge with abundant localized faulting. At Site 1270 of Leg 209, four holes were drilled near the top of an exposed long-lived normal

197

fault on the side of the rift valley (Kelemen et al., 1998; Schroeder et al., 2007). Although this part of the MAR is volcanically starved, gabbroic intrusions and melt veins are abundant (Kelemen et al., 2004). Veins containing zircon and apatite have been found in small-scale shear zones, and they are interpreted as plagiogranitic melts derived from hydrous partial melting of gabbroic rocks (Jöns et al., 2009). In the present study, we use serpentinite samples adjacent (max. 2 cm distance) to such evolved melt veins (the corresponding host rocks to samples used by Jöns et al., (2009), sample locations given in Table 1). 3. Methods 3.1. Phase identification Mineral phases were identified by polarizing microscopy. In addition, whole-rock samples were pulverized and examined with a Panalytical X'Pert PRO X-ray diffractometer (XRD) at the University of Bremen with Cu Kα radiation. Each powder pellet was scanned from 2Θ = 4° to 75° with a voltage of 45 kV and a current of 40 mA. Data evaluation was done using R. Petschick's MacDiff software (version 4.2.5; available for download from http://servermac.geologie. uni-frankfurt.de/Staff/Homepages/Petschick/RainerE. html). XRD patterns of selected samples can be found in Appendix A. 3.2. Electron microprobe analyses Major element mineral chemical compositions have been determined using a JEOL JXA-8900 electron microprobe at the University of Kiel (Germany). For most silicate minerals an accelerating voltage of 15 kV was used, with exception of olivine and spinel (20 kV). Measurement spot sizes were typically 1–5 μm in diameter. Standards were either natural minerals or synthetic materials. The CITZAF method of Armstrong (1995) was used for correction of the raw counts.

Fig. 1. Bathymetric map (prepared using the Generic Mapping Tools of Wessel and Smith, 1991, 1998) of the central Atlantic Ocean showing the location of ODP Leg 209 Site 1270 (equidistant conic projection).

198

N. Jöns et al. / Chemical Geology 274 (2010) 196–211

Table 1 List of samples of serpentinized harzburgites used for whole-rock geochemical analyses. All samples are taken in a maximum distance of 2 cm from chlorite-amphibole-bearing veins on small-scale shear zones. Sample no.

1

2

3

4

5

6

Latitude Longitude Hole Core Section Length of sample (cm) Depth (mbsf)

14°43.284′N 44°53.091′W 1270C 1R 1 10 0.22

14°43.284′N 44°53.091′W 1270C 1R 1 5 0.64

14°43.270′N 44°53.084′W 1270D 3R 1 8 19.67

14°43.270′N 44°53.084′W 1270D 3R 2 3 21.90

14°43.270′N 44°53.084′W 1270D 3R 1 4 19.63

14°43.270′N 44°53.084′W 1270D 4R 1 5 24.04

3.3. Whole-rock geochemistry and density Major elements were measured with a Philips PW1480 XRF spectrometer at the University of Kiel on glass beads. These were prepared from 3.6 g lithium-tetraborate (Li2B4O7) and 0.6 g sample powder on an Oxiflux 5-stage burner. The USGS rock powder standard BHVO1 (Flanagan, 1976) was used for control over the standardization. Accuracy and precision are better than 1% (1σ) for the major oxides (see Appendix A of van der Straaten et al., 2008). The loss on ignition (LOI) was determined after ignition of the samples at 1000 °C for 12 h. Trace element concentrations were measured by ICP-MS (Agilent 7500cs) at the University of Kiel. 100 mg of dried and pulverized sample powder were digested using HF–HNO3 in acid digestion bombs at 180 °C for four days and subsequently boiled and fumed off with HClO4. Before measurement, the final 1% HNO3 solution was 20fold diluted and spiked for internal standardisation (2.5 ng ml−1 Be, In and Re). For analytical quality control, the international reference standard UB-N (Govindaraju and Roelandts, 1989), BHVO-2 (Wilson, 1997), as well as procedural blanks were prepared and analyzed along with the sample series (see Appendix B). From standard and duplicate analyses the analytical precision is estimated to be b5%. Further details on the analytical procedure can be found in Garbe-Schönberg (1993) and John et al. (2008). Density measurements were performed using 25 ml Gay-Lussac pycnometers. Each sample was measured three times, resulting in a precision b5% (1σ). 3.4. Laser ablation ICP-MS Trace element compositions of minerals have been analyzed at the University of Bremen (Germany) using a NewWave UP193 solid state laser (output wavelength: 193 nm), which is coupled to a ThermoScientific Element2 high-resolution ICP-MS. The laserbeam diameter was typically between 50 and 100 μm. Helium was used as carrier gas (0.4 l min−1) in a standard sample chamber with subsequent addition of argon (ca. 0.8 l min−1). After 20 s measurement of the blank signal, each sample was ablated for 50 s on a fixed spot; the laser fired with a repetition rate of 5 Hz and the irradiance was ca. 1 GW cm−2. Depending on the mineral composition, either silicon or calcium or magnesium was used as an internal standard element, which had previously been determined by electron microprobe analysis. Calibration was done using glass standards NIST610 and NIST612, with preferred averages from (Pearce et al., 1997). Analytical quality was controlled by repeated analyses of the international glass standards BCR2-G (Jochum et al., 2006) and GOR132-G (Jochum et al., 2005, see Appendix C). 3.5. Geochemical reaction path modeling Reaction path models have been set up to predict the mineralogical evolution during hydrous alteration of harzburgite and the influence of a nearby felsic melt vein. Results of the model calculations were compared with the petrographic observations. Models are calculated using the EQ3/6 software package (Wolery, 2002) with a customized database of equilibrium constants (log K values; calcu-

lated using Supcrt92; Johnson et al., 1992) for minerals and aqueous species. The calculations are performed along an assumed geothermal gradient with the following temperature [°C]/pressure [bar] pairs: 2/ 250, 50/270, 100/330, 150/400, 200/500, 250/600, 300/725, 350/860, 400/970, 450/1100, 500/1250, 550/1400, 600/1600, 650/1800, 700/ 2000, 750/2250, 800/2500, 850/2800, 900/3150. Ideal solid solutions for olivine (forsterite–fayalite), serpentine (chrysotile–greenalite), brucite (brucite–Fe-brucite; with V0 for Fe-brucite from Wolery and Jove-Colon, 2004), orthopyroxene (enstatite–ferrosilite), clinopyroxene (diopside–hedenbergite), chlorite (clinochlore–daphnite), garnet (andradite–grossular), talc (talc–minnesotaite), amphibole (tremolite–ferroactinolite–pargasite–ferropargasite) and plagioclase (albite–anorthite) have been taken into account. We have chosen chrysotile as the Mg end-member for the serpentine solid solution, which is justified by only minimal differences in the thermodynamic properties of chrysotile and lizardite (e.g., Evans, 2004; Frost and Beard, 2007), the typical serpentine minerals occurring in abyssal peridotites. To account for metastable mineral equilibria, we suppressed antigorite, calcite, aragonite, dolomite, magnesite, huntite, monticellite, corundum and nepheline. The calculated model involved speciation of seawater at 25 °C and heating to 100 °C. During further heating from 100 to 800 °C, the fluid was allowed to equilibrate with mixtures of harzburgite (using the composition of the JP-1 standard; Imai et al., 1995) and plagiogranite (composition kindly provided by J. Koepke and given in Jöns et al. (2009)) in variable proportions. The initial fluid/rock (w/r) mass ratio was 1 at 100 °C, but was variable at higher temperatures due to different proportions of hydrous minerals. Although higher w/r ratios likely have existed on the contiguous meltbearing shear zone, such a low w/r estimate for the serpentinites is justified by similarity of mineral assemblages to other localities (cp. Klein et al. (2009). In an additional model, which is based on the calculations of Jöns et al. (2009), we evaluated the effect of varying w/ r ratio on the peridotite mineral assemblage. A JP-1 harzburgite is titrated into a fluid that had equilibrated with a 70:30 mixture of harzburgite and plagiogranite. Such a composition has been deduced by Jöns et al. (2009) based on bulk major element chemistry of the shear zone rocks. A seawater-derived fluid is assumed to equilibrate at an initial w/r ratio of 1 at 100 °C, which varies upon heating due to changing of e.g., mineral assemblages. Although higher w/r ratios might be present on shear zones, we have chosen this conservative estimate, as it perfectly reproduces the observed chlorite-amphibole fault schist mineral assemblages (cp. Jöns et al., 2009). 4. Petrography and mineral chemistry Samples investigated in this study are strongly serpentinized harzburgites. The focus lay on rocks that occur adjacent (i.e. in b2 cm distance) to small-scale shear zones, which show evidence for influence of evolved melts. The former magmatic minerals are completely altered to chlorite-amphibole veins (Fig. 2a) with accessory zircon and apatite. They have been studied in detail by Jöns et al. (2009) and interpreted as former plagiogranites derived from hydrous partial melting of gabbroic intrusions. Due to the restricted amount of sample material in the drill core (diameter of the

N. Jöns et al. / Chemical Geology 274 (2010) 196–211

199

Fig. 2. Thin section photomicrographs of serpentinite samples from Site 1270. (a) Overview of a serpentinized harzburgite crosscut by a melt-bearing shear zone. In general, serpentinites show typical mesh textures (after olivine) with local occurrence of bastites (replacement of orthopyroxene; see upper left and lower right of the photograph). In shear zones, which are crosscut by late-stage picrolite veins, chlorite and amphibole form elongate bands. Plane polarized light. (b) Serpentinite meshes rimmed by magnetite, which are crosscut by late-stage serpentine veins. A bastite pseudomorph after orthopyroxene is shown in the upper left part of the picture. Crossed nicols. (c) A thick and chemically zoned serpentine vein crosscutting the serpentinite. Crossed nicols. (d) Late-stage carbonate vein crosscutting a serpentinite. The vein consists of idiomorphic dolomite crystals in a calcite matrix. Cathodoluminescence image.

core and recovery rate), it was not possible to sample the serpentinites in variable distances from the altered melt veins. 4.1. Altered melt veins The millimeter-sized veins crosscutting the serpentinites contain crystals of euhedral brownish magnesiohornblende (XMg = Mg/ (Mg + Fe2+) = 0.86–0.96) in a matrix of colorless to light green fibrous chlorite (XMg = 0.82–0.95). The hornblende crystals form δtype porphyroclasts (e.g., Passchier and Simpson, 1986) with needles of actinolite or tremolite (XMg = 0.85–0.93) growing in pressure shadows. Common accessory zircon is elongate with slightly rounded corners. Titanium-in-zircon thermometry (Ferry and Watson, 2007) indicates that zircon crystallized at temperatures around 820 °C, which is a realistic estimate for growth from an evolved melt (Jöns et al., 2009). Apatite is less common and it is characterized as a solid solution between hydroxylapatite (XOH = [OH]/([OH]+ Cl+ F) = 0.47– 0.62), chlorapatite (XCl = 0.25–0.45) and minor fluorapatite component (XF = 0.05–0.13). Partially serpentinized olivine within the veins is obviously sheared in from the host peridotite; however, olivine in the shear zone is generally less altered than olivine in the surrounding serpentinites. Whole-rock oxygen isotope data for mechanically separated melt veins range from + 3.0 to + 4.2‰ (relative to VSMOW), which is interpreted to reflect hydrous alteration due to focused fluid flow on the shear zone (Jöns et al., 2009). 4.2. Serpentinized harzburgites The adjacent serpentinites are generally fine-grained. Olivine, spinel, and occasionally orthopyroxene, are preserved relict minerals

(see Table 2 for selected mineral analyses), but the extent of alteration is typically N90 vol.%. Olivine is replaced by serpentine-group minerals and magnetite in hour glass and mesh textures (Fig. 2a). Brucite has not been identified by means of microscopy and EPMA. Moreover, XRD patterns of whole-rock powders (see Appendix A) lack the characteristic (001) reflection of brucite (cp. Hostetler et al., 1966). Spinel is commonly rimmed by magnetite. Orthopyroxene is almost completely replaced by serpentine-group minerals in bastite texture (Fig. 2a,b), where locally higher interference colors as well as higher SiO2 or Al2O3 contents indicate that small amounts of talc or chlorite might be intergrown with serpentine. Under crossed nicols, the straight optical extinction of the fine-fibrous bastite confirms that it pseudomorphously replaces orthopyroxene. In contrast to mesh texture, magnetite is apparently absent from bastite. The serpentinite is crosscut by several generations of transgranular serpentine veins. Textural observations indicate that older veins consist of very finegrained serpentine with some magnetite, whereas the youngest veins are made up of optically isotropic serpentine (picrolite; Fig. 2a,c) with no microscopically visible magnetite. Picrolite is yellowish to brownish in color and shows a zonation parallel to the elongation of the vein (Fig. 2c), indicating a polyphase growth on episodically opening cracks. Composite serpentine veins are comparatively thick (locally mm-scale) where they crosscut altered melt veins on shear zones (Fig. 2a); elsewere they occur as several narrow subparallel veins crosscutting the mesh texture (Fig. 2b). In proximity of bastite, they can contain additional talc or chlorite. Based on petrographical observations, it can be inferred that most late-stage serpentine veins formed post-kinematically; however, a few cases have been observed where a vein is offset by a shear zone, pointing to syn-kinematic formation. Locally, carbonate veins are present, which consist of

200

N. Jöns et al. / Chemical Geology 274 (2010) 196–211

Table 2 Selected electron microprobe (in wt.%) and LA-ICP-MS (in μg·g−1) analyses of spinel, olivine and serpentine from serpentinites. Sample† Analys. no. Mineral Comment

2 61 Spl Mesh

3 645 Spl Mesh

2 21 Ol Mesh

2 39 Ol In bastite

2 47 Ol Recryst.

3 657 Ol Mesh

2 262 Srp Mesh

2 348 Srp Bastite

2 313 Srp Picrolite

3 475 Srp Picrolite

[wt.%] SiO2 TiO2 Al2O3 Cr2O3 Fe2O*3 FeO MgO MnO CaO Na2O K2O Total

0.08 0.10 27.74 38.99 3.41 17.73 12.25 0.26 0.04 n.d. n.d. 100.26

0.04 0.06 28.49 37.80 3.42 19.27 11.30 0.26 0.01 n.d. n.d. 100.30

40.53 0.02 0.03 0.00 n.c. 10.76 48.41 0.22 0.01 n.d. n.d. 99.98

40.96 0.00 0.03 0.00 n.c. 8.61 50.18 0.14 0.03 n.d. n.d. 99.95

40.78 0.02 0.00 0.00 n.c. 9.36 49.37 0.17 0.00 n.d. n.d. 99.70

40.90 0.01 0.02 0.00 n.c. 8.74 49.82 0.12 0.00 n.d. n.d. 99.61

42.62 0.03 0.06 0.00 n.c. 2.50 40.60 0.07 0.03 0.02 0.01 85.94

42.36 0.01 0.09 0.02 n.c. 2.35 40.55 0.06 0.01 0.02 0.00 85.48

42.66 0.00 0.21 0.00 n.c. 5.64 37.32 0.10 0.02 0.07 0.00 86.01

43.09 0.01 0.38 0.00 n.c. 5.41 36.33 0.14 0.06 0.16 0.00 85.58

Si Ti Al Cr Fe3+* Fe2+ Mg Mn Ca Na K Total

0.00 0.00 0.98 0.93 0.08 0.45 0.55 0.01 0.00 n.d. n.d. 3.00

0.00 0.00 1.01 0.90 0.08 0.49 0.51 0.01 0.00 n.d. n.d. 3.00

1.00 0.00 0.00 0.00 n.c. 0.22 1.78 0.01 0.00 n.d. n.d. 3.01

1.00 0.00 0.00 0.00 n.c. 0.18 1.82 0.00 0.00 n.d. n.d. 3.00

1.00 0.00 0.00 0.00 n.c. 0.19 1.80 0.00 0.00 n.d. n.d. 2.99

1.00 0.00 0.00 0.00 n.c. 0.18 1.82 0.00 0.00 n.d. n.d. 3.00

2.01 0.00 0.00 0.00 n.c. 0.10 2.86 0.00 0.00 0.00 0.00 4.97

2.01 0.00 0.01 0.00 n.c. 0.09 2.87 0.00 0.00 0.00 0.00 4.98

2.04 0.00 0.01 0.00 n.c. 0.23 2.66 0.00 0.00 0.01 0.00 4.95

2.07 0.00 0.02 0.00 n.c. 0.22 2.60 0.01 0.00 0.01 0.00 4.93

oxygens Mg/(Mg + Fe2+) Cr/(Cr + Al) Fe3+/(Fe3++ Cr + Al)

4.00 0.55 0.49 0.04

4.00 0.53 0.47 0.04

4.00 0.89 n.c. n.c.

4.00 0.91 n.c. n.c.

4.00 0.90 n.c. n.c.

4.00 0.91 n.c. n.c.

7.00 0.97 n.c. n.c.

7.00 0.97 n.c. n.c.

7.00 0.92 n.c. n.c.

7.00 0.92 n.c. n.c.

[μg·g−1] Li Ca Ti V Cr Co Ni Cu Zn Rb Sr Y Zr Nb

b.d.l. 20.4 826 864 n.d. 324 635 1.47 2007 0.934 b.d.l. 0.003 0.073 0.036

0.149 431 176 809 n.d. 415 537 b.d.l. 2975 0.670 0.236 0.020 0.023 0.054

7.67 252 2.47 0.332 14.5 148 3186 0.172 99.7 0.017 0.028 0.011 b.d.l. 0.007

b.d.l. 249 10.0 1.92 145 146 3308 0.115 42.7 0.048 0.042 0.019 0.008 b.d.l.

0.413 99.1 17.6 0.963 58.7 138 3176 3.37 27.8 0.175 0.153 0.356 0.310 0.345

b.d.l. 48.6 1.54 0.110 10.6 60.8 1604 0.003 17.4 0.004 0.006 0.002 0.014 0.019

11.6 254 7.07 0.583 20.8 110 2987 0.434 16.3 1.36 b.d.l. 0.017 0.037 0.046

6.08 5400 52.6 73.7 4810 45.4 630 b.d.l. 24.3 0.098 0.233 0.227 0.006 0.005

20.2 595 28.0 5.28 0.316 90.3 1543 b.d.l. 38.4 9.02 0.934 0.011 1.38 1.04

1.47 117 35.3 1.25 0.306 15.2 482 0.187 22.0 1.44 1.48 0.086 0.143 0.454

† sample number corresponds to Table 1; * calculated assuming stoichiometry; n.d.: not determined; n.c.: not calculated; b.d.l.: below detection limit.

dolomite or calcite or of both carbonate minerals (Fig. 2d). Crosscutting relationships indicate that they generally represent the latest alteration stage, i.e. they locally follow the direction of pre-existing picrolite veins or crosscut them. Olivine is Mg-rich (XMg = 0.88–0.91; Table 2). The intense serpentinization leads to a xenomorphic crystal shape. MnO contents are 0.10–0.25 wt.% and the maximum CaO content is 0.05 wt.%. The most important trace elements are Ni (up to ca. 3200 μg g−1) and Co (up to 155 μg g−1). Spinel is Fe-rich (XMg = 0.48–0.55) and Cr-rich (XCr = Cr/(Cr+ Al) = 0.47–0.55; see Table 2). Assuming stoichiometry, 13–15 wt.% of the iron seem to be trivalent. The spinel contains TiO2 (0.05–0.25 wt.%) as well as Co (300–500 μg g−1), Ni (500–700 μg g−1), Zn (1800–3500 μg g−1) and V (750–900 μg g−1). Orthopyroxene is characterized as XEn = Mg/(Mg + Fe + Ca) = 0.86– 0.91, XFs = 0.08–0.09 and XWol ≤ 0.06. Al2O3 contents range from 1.3 to 3.5 wt.%. The maximum content of Na2O is 0.15 wt.%. Chondritenormalized REE patterns are flat and approximately chondritic, without

a pronounced negative or positive Eu anomaly. Chromium contents range from 500 to 6200 μg g−1. Serpentine pseudomorphously replaces olivine and orthopyroxene, and it forms different generations of veins; however, it is difficult to use mineral chemistry to distinguish between these serpentines, because the chemical variation is dependent on the texture (i.e. replacement of olivine or orthopyroxene) and the timing of growth (i.e. growth at different temperatures and/or w/r ratios). Therefore, the trace element concentrations are highly variable, even on thin section scale. In mesh textures after olivine, serpentine generally shows the highest XMg ratios (0.95–0.98; Table 2), whereas in bastite textures it reveals a larger variation in XMg (0.89–0.97). Late-stage banded and picrolitic veins locally have XMg ratios as low as 0.86 (max. 0.92). Alumina contents of serpentine are similar to that of the precursor minerals: 0.05–0.12 wt.% Al2O3 in mesh textures and up to 2.5 wt.% in bastite; however, it is likely that microprobe analyses do not represent pure serpentine, but fine-fibrous intergrowths with talc and/or chlorite. Vein serpentine has intermediate alumina contents

N. Jöns et al. / Chemical Geology 274 (2010) 196–211

(0.2–1.2 wt.% Al2O3). The precursor mineral chemistry is also reflected in the amounts of Ca (mesh: 180–300 μg g −1 ; bastite: 700– 5500 μg g−1), Ti (mesh: 6–13 μg g−1; bastite: 50–360 μg g−1), and Cr (mesh: 10–45 μg g−1; bastite: 4500–6700 μg g−1). 5. Whole-rock geochemistry 5.1. Major and trace element composition Serpentinite samples containing abundant altered melt veins were taken from the drillcores. Prior to chemical analyses, melt-bearing portions were removed from the host rock using a small cutting wheel. The most prominent chemical change compared to unaltered

Table 3 Whole-rock geochemical data of serpentinites. No.† [wt.%] SiO2 TiO2 Al2O3 FeOT MnO MgO CaO Na2O K2O P2O5 LOI Total

1

2 40.88 0.02 0.73 6.66 0.12 36.95 0.09 0.02 0.02 0.01 13.31 98.81

[μg·g−1] Co 105 Cu 3.30 Ni 1945 Zn 44.4 Cr 2542 Sc 7.69 V 40.1 Mo 0.689 Ga 1.28 Sn 0.253 Sb 0.100 W 0.043 Li 6.16 Rb 0.151 Sr 4.69 Cs 0.003 Ba 8.57 Y 1.12 Zr 0.530 Nb 0.364 Hf 0.016 Ta 0.009 Pb 0.339 Th 0.053 U 0.950 La 0.748 Ce 1.96 Pr 0.204 Nd 0.789 Sm 0.155 Eu 0.238 Gd 0.163 Tb 0.025 Dy 0.158 Ho 0.034 Er 0.101 Tm 0.016 Yb 0.113 Lu 0.018 * ρ 2.54

3

4

5

6

41.15 0.02 0.63 6.94 0.13 37.34 0.11 0.01 0.01 0.01 13.08 99.43

44.30 0.03 1.27 6.35 0.12 35.09 0.87 0.14 0.03 0.01 10.90 99.11

40.86 0.06 1.21 6.98 0.13 36.38 0.09 0.04 0.06 0.01 13.53 99.35

42.72 0.05 1.13 6.14 0.14 35.95 0.81 0.09 0.02 0.02 13.61 100.68

101 3.95 2054 46.6 1937 7.63 39.4 0.764 1.19 0.181 0.088 0.058 5.22 0.083 3.76 0.002 4.46 1.43 1.39 0.289 0.026 0.009 0.335 0.093 1.02 0.763 1.57 0.193 0.766 0.165 0.182 0.186 0.031 0.197 0.044 0.127 0.020 0.138 0.022 2.70

95.8 2.45 1878 50.3 2806 8.98 38.7 0.452 2.07 0.251 0.038 0.082 9.92 0.392 6.75 0.010 6.30 2.91 6.46 1.82 0.232 0.118 0.239 0.327 0.324 1.77 4.11 0.511 1.96 0.406 0.151 0.420 0.070 0.454 0.098 0.291 0.047 0.334 0.052 2.60

103 2.25 2073 43.6 1645 6.92 27.6 0.626 1.88 0.249 0.006 0.035 6.99 0.825 4.49 0.012 8.90 5.77 20.1 3.20 0.502 0.234 0.217 0.154 0.211 1.95 4.33 0.596 2.70 0.711 0.287 0.841 0.144 0.938 0.199 0.558 0.083 0.554 0.083 2.52

94.2 3.52 1939 42.0 1552 7.15 27.1 0.561 2.03 0.308 0.063 0.043 9.79 0.251 7.11 0.007 6.47 9.27 20.8 1.56 0.490 0.066 0.405 0.106 0.696 2.59 5.98 0.905 4.34 1.19 0.367 1.37 0.235 1.49 0.313 0.869 0.127 0.822 0.121 2.51

† sample number corresponds to Table 1; * density ρ in g·cm−3.

201

peridotite is the addition of H2O, with values for LOI ranging from 10.90 to 13.61 wt.%. In terms of major elements, the serpentinites show just a small variability (Table 3). The contents of SiO2 vary between 40.88 and 44.30 wt.%, MgO between 35.09 and 37.35 wt.% and FeOT (total iron calculated as ferrous) between 6.14 and 7.47 wt.%. Al2O3 contents range from 0.63 to 1.27 wt.%. Furthermore, the samples have low contents of alkali metals (Na2O b 0.14 wt.%; K2O b 0.06 wt.%; Rb = 0.083–0.825 μg g−1; Li = 3.36–9.92 μg g−1) and other fluid mobile elements (Sr = 2.71–7.11 μg g−1; Ba = 1.28– 8.90 μg g−1). Chondrite-normalized REE concentrations show a slightly U-shaped pattern (Fig. 3a) with about one order of magnitude variation. The presence and intensity of a positive Eu anomaly is dependent on the overall amount of REEs: the sample with the lowest REE concentration pffiffiffiffiffiffiffiffiffiffiffiffiffiffi ffi (heavy REEs ≈ 0.4 times chondritic) has a Eu/Eu* (Eu = Sm⋅Gd) of 4.32, whereas higher REE concentrations (heavy REE ≈ 7–10 times chondritic) correspond to lower Eu/Eu* of 1.06 or even a slightly negative Eu anomaly (Eu/Eu*= 0.83). When trace element contents are normalized against an unaltered harzburgite (using the composition of the JP-1

41.19 0.01 0.70 7.47 0.09 37.35 0.13 b.d.l. 0.01 0.01 12.76 99.72

106 3.16 2116 44.5 2034 7.51 32.1 0.600 1.19 0.092 0.006 0.070 3.36 0.096 2.71 0.002 1.28 0.525 1.51 0.159 0.034 0.004 0.330 0.020 0.138 0.495 0.859 0.100 0.392 0.083 0.087 0.085 0.014 0.083 0.018 0.053 0.009 0.066 0.011 2.47

Fig. 3. Geochemical composition of serpentinite samples from Site 1270 and adjacent altered melt-bearing veins (Chl-Amph veins; Jöns et al., 2009). (a) Chondritenormalized (McDonough and Sun, 1995) REE patterns showing a slightly U-shaped pattern with a positive Eu anomaly. For comparison, analyses of the least-altered peridotites from Site 1274 (Paulick et al., 2006) are also shown. (b) Harzburgitenormalized (JP-1 standard; Imai et al., 1995) trace element patterns of serpentinites and neighboring chlorite-amphibole-bearing veins.

202

N. Jöns et al. / Chemical Geology 274 (2010) 196–211

Fig. 4. Illustration of the main features of isocon diagrams in Fig. 5). Element concentrations of serpentinite samples (axis of the ordinate) are plotted against two rock compositions (on the abscissa): the least-altered harzburgite composition and a brucite-bearing serpentinite (data from Paulick et al., 2006) from Leg 209/Site 1274 on the outer and inner quarter circle, respectively. The isocons are calculated for the assumption of constant mass (❶ for the harzburgite and ❷ for the serpentinite).

standard; Imai et al., 1995), an enrichment in REE (both light and heavy REE) is evident, whereas the picture for high field strength elements (HFSE) is ambiguous, e.g., zirconium and hafnium may be either slightly depleted or enriched (Fig. 3b). As the latter elements are generally believed to be comparatively fluid immobile, the variability likely results from addition of a melt phase. This is confirmed by a significant positive linear correlation of HFSE and LREE concentrations (e.g., RLa–Nb = 0.766, RCe–Nb = 0.747, RLa–Zr = 0.911, RCe–Zr = 0.886). 5.2. Mass balance calculations Mass balance calculations have been conducted to quantify elemental changes and to characterize the conversion of a harzburgite protolith into a serpentinite. We applied the method of Gresens (1967), which accounts for both volume and composition changes during metasomatic alteration. By means of isocon analysis (Grant, 1986, 2005) the mass transfer can be graphically illustrated (Fig. 4). Grant (1986) gives the following main equation: A

Ci =

 M  O Ci + ΔCi A M O

For each element such an equation can be written, where CAi is the concentration of element “i” in the altered rock, CO i is the concentration of element “i” in the original rock, ΔCi is the change in concentration of “i”, and MO and MA are the masses before and after alteration. By plotting CAi against CO i and defining an isocon (i.e. a line through the origin defined by e.g., an immobile element), elemental gains and losses are readily seen. However, instead of an arbitrary scaling as proposed by Grant (1986), we have calculated different scaling factors for all elements to achieve that all elements plot at the same distance from the origin (normalizing so that the sum of squares

equals 1 or 1.2; see Humphris et al., 1998, for a more detailed description). The difficulty is the absence of any unaltered protolith, the composition of which therefore had to be inferred. Close to shear zones, original textures are generally obliterated by deformation and thus the proportions of primary mineral phases cannot be estimated undoubtedly. Therefore, in a first step we compare our serpentinite samples with the JP-1 harzburgite standard (Imai et al., 1995) as unaltered precursor (see Fig. 3b); however, this harzburgite composition might not be a typical protolith for our case. Figure 3 shows an enrichment in REE, Ti and K for all samples. In contrast, Rb and Ba are depleted or equal to the JP-1 concentration. For other elements, especially for the high field strength elements (HFSE) Nb, Ta, Zr and Hf, either a depletion or an enrichment is observed. Patterns of all samples in Fig. 3b show almost parallel shifting, which might be an effect of mechanical mixing with melt at the border of the shear zones. To compensate for a chemical discrepancy of the true (but unknown) precursor composition from the JP-1 harzburgite standard and to disentangle melt-related and fluid-related changes, the serpentinite samples are compared with other rock compositions from Leg 209 (analyses taken from Paulick et al., 2006): a) with the least-altered harzburgite (ca. 60 vol.% secondary minerals) from Site 1274A, a core segment that is almost free of any obvious melt influence (Core 6, Section 2, 32.73 mbsf [mbsf = meters below seafloor]); b) with a completely serpentinized harzburgite that is brucite-bearing and thus indicative of low silica activity (melt absent) conditions (1274A, core 15, Section 2, 75.86 mbsf). Using whole-rock compositions from Leg 209 Site 1274 for comparison is, on the one hand, justified by proximal occurrence in the same geodynamic setting and the observation of a regionally homogeneous mantle; on the other hand, similarities in the compositions of chromian spinel (cp. Table 2 and Seyler et al., 2007), which is an alteration-resistant but host rock sensitive petrogenetic indicator (e.g., Dick and Bullen, 1984), make Site 1274 rocks a good choice for comparison. The isocon diagrams (main features explained in Fig. 4) in Fig. 5 show a comparison with both chosen rock compositions. This manner of data representation allows to compare the sample compositions with those of melt-unaffected serpentinites, and thus to distinguish between changes that are characteristic for the fluid-driven conversion of harzburgite to serpentinite and melt-related chemical changes. However, some uncertainty of not knowing the true protolith composition remains. For all samples an enrichment in REE compared to both serpentinized harzburgite samples is observed. This enrichment is more pronounced compared to the fully serpentinized sample than compared to the least altered harzburgite, indicating REE mobility during serpentinization. The latter is also supported by the more pronounced enrichment of fluid mobile light REE compared to heavy REE. The REE behaviour is contrasting that of the HFS elements Zr and Nb (Ta, Ti and Hf cannot be displayed in Fig. 5, because their concentrations are below detection limit in the 1274 peridotites), which might also be strongly enriched (Fig. 5c,d,e), but apparently the enrichment compared to both assumed protoliths is of similar extent, indicating fluid immobile behaviour of these elements during serpentinization. The content of transition metals largely clusters around the isocons in Fig. 5, which can be explained by slight differences in the protolith composition. The major elements Mg, Fe and Si do not show a significant deviation from the constant mass isocon. While Al is generally enriched in the examined serpentinites, these samples show either a loss or a gain in Ca. It is also worth noting that the Al enrichment compared to the fully serpentinized sample is more pronounced,

Fig. 5. Isocon diagrams for six abyssal serpentinites. Element concentrations in serpentinites are plotted against compositions of the least altered harzburgite of Site 1274 (outer quarter circle) and a brucite-bearing serpentinite (inner quarter circle; data from Paulick et al., 2006). See Fig. 4 for details about the isocon diagrams. Further discussion is given in the text (sample numbers correspond to Table 1).

N. Jöns et al. / Chemical Geology 274 (2010) 196–211

203

204

N. Jöns et al. / Chemical Geology 274 (2010) 196–211

Fig. 6. Mineral assemblages as well as composition of minerals and fluids resulting from interaction of seawater with harzburgite and harzburgite–plagiogranite mixtures (a) pure JP1 harzburgite (Imai et al., 1995); b) 96% harzburgite and 4% plagiogranite; c) 90% harzburgite and 10% plagiogranite). For all diagrams the water/rock mass ratio is 1 at T = 100 °C, but varies at higher temperatures due to different amounts of hydrous mineral phases. At temperatures above 600°C, orthopyroxene and olivine are the most abundant minerals. Formation of serpentine starts at temperatures below ≈420 °C. The amount of brucite within the alteration assemblage is strongly dependent on the amount of plagiogranite that is mixed with the harzburgite. See text for further discussion.

which might indicate Al mobility during serpentinization. The fluid mobile elements Sr and Ba are generally strongly enriched. 6. Geochemical reaction path modeling Reaction path model were compared with the observed mineral assemblages. The predicted mineral assemblage for alteration of pure harzburgite is shown in Fig. 6a. At T N 600 °C, orthopyroxene (XMg = 0.93–0.95) and olivine (XMg ≈ 0.90) are the major mineral phases. Calcium-bearing minerals are clinopyroxene and plagioclase; however, they make up less than 5 mol% and their former existence cannot be proven in almost completely serpentinized and deformed samples. At lower temperatures, orthopyroxene breaks down, and talc and chlorite coexist with olivine of slightly more Fe-rich composition (Fig. 6a). Formation of serpentine and magnetite starts when tempera-

tures drop below ≈420 °C. Serpentine and magnetite, at T b 330 °C also brucite, replace olivine. The amount of magnetite is largest when brucite formation starts; with decreasing temperature the iron contents of brucite and serpentine increase and thus the amount of magnetite is reduced (cf. Klein et al., 2009). Secondary garnet and clinopyroxene are also predicted to form; this discrepancy between model and samples might result from the fact that the modeling does not account for Ca incorporation in other mineral phases (e.g., in olivine and orthopyroxene). The composition of the coexisting fluid is characterized by a pH between 5 and 6 at T N 400 °C, which becomes more alkaline at lower temperatures (Fig. 6a). A decrease in the Fe content of the fluid during cooling corresponds with formation of more Fe-rich secondary hydrous minerals (e.g., serpentine, brucite and chlorite). The influence of silica-rich melt veins on the mineralogy and fluid chemistry of harzburgite serpentinization has been evaluated

N. Jöns et al. / Chemical Geology 274 (2010) 196–211

using different bulk-rock compositions, which are mixtures between JP-1 harzburgite (Imai et al., 1995) and average plagiogranite (composition provided by J. Koepke and given in Jöns et al., 2009). Fig. 6b and c shows results for 4 wt.% and 10 wt.% plagiogranite, respectively. At high temperatures an increasing amount of plagiogranite results in larger amounts of orthopyroxene and plagioclase in the assemblage. Upon cooling the formation of talc and chlorite starts at similar temperatures than in the pure harzburgite model (T ≈ 600 °C), but as they form from breakdown of orthopyroxene, they make up a larger proportion of the rock. Although olivine breaks down to form serpentine and magnetite at T b 420 °C, which is comparable to the temperature predicted in the pure harzburgite model, the temperature range of coexisting olivine and serpentine is narrower. The most obvious influence of bulk chemistry on the predicted secondary mineralogy is the cutback of the amount of brucite formed from olivine breakdown. In the model involving 10 wt.% plagiogranite, only very small amounts of brucite are expected to form at temperatures b170 °C. Mineral compositions are influenced by the bulk chemistry, resulting in formation of slightly more Fe-rich olivine (XMg ≈ 0.88 compared to 0.90 at T = 440–580 °C), and more Fe-rich serpentine is stable at higher temperatures (XMg ≈ 0.93–0.94 compared to 0.93– 0.98 at T = 100–360 °C). A similar observation can be made for the chlorite chemistry. The modeled chemical composition of the intergranular fluid shows a shift of the low pH region towards higher temperatures (Fig. 6b) with increasing proportion of siliceous material. In addition, the harzburgite– plagiogranite mixing models predict fluids containing more Al and Si, but less Fe, Mg, Ca and K. The differences are predicted over the whole temperature range from 100 to 900 °C, but they are more pronounced in the low-temperature region of the reaction path. Assuming that melt-bearing shear zones crosscutting the peridotites are the main fluid pathways, which is indicated by oxygen isotope data (Jöns et al., 2009), the fluid/rock mass ratio should be comparatively high on shear zones and should decrease with increasing distance from the shear zone. The effect of variable w/r ratios on mineral assemblage, mineral compositions and fluid chemistry has been evaluated in another model run. Fig. 7 shows the modeled mineral proportions, mineral compositions and fluid chemistry resulting from titrating a pure harzburgite into an assumed shear zone fluid (fluid in equilibrium with a 70:30 harzburgite–plagiogranite mixture) at 300 °C. The results of this model are evaluated against increasing logξ, where ξ is a variable of reaction progress. As logξ values are internal parameters that are not of relevance for real w/r ratios, no scale is added to the x-axis of diagrams in Fig. 7. Serpentine is by far the most common mineral phase in the serpentinized harzburgite (N80 mol%), with larger amounts of other phases at lower w/r ratio. Besides serpentine, these minerals are (with increasing logξ): Amph ⇒ Amph + Chl ⇒ Amph + Chl + Cpx ⇒ Chl + Cpx ⇒ Chl + Cpx+ Grt⇒ Chl + Grt⇒ Chl + Grt + Brc. In addition, the amount of magnetite steadily increases and just the emergence of brucite results in a slight decrease in magnetite abundance (Fig. 7a). These mineral zones can be seen as an equivalent of metasomatic mineral zoning around fluid-bearing veins. The prediction is in agreement with chlorite and amphibole being present on the hydrated melt-bearing shear zones and brucite lacking in serpentinites that are adjacent to such zones. Mineral compositions of amphibole, serpentine, chlorite and clinopyroxene are most Fe-rich at low log ξ and increasing with increasing distance from a fluid vein. Just the occurrence of brucite leads again to a slight decrease in XMg of coexisting chlorite and serpentine (Fig. 7b). At high w/r ratios, i.e. near the shear zone, mineral compositions deviate significantly from values derived from the reaction path model presented in Fig. 6a (at 300 rcC), where harzburgite equilibrated with a seawater-derived fluid. In addition, the chemistry of the coexisting fluid in proximity of the plagiogranite-bearing shear zones is characterized by higher pH as well as higher Si and Al contents (Fig. 7c).

205

Fig. 7. Modeled mineral assemblage, mineral compositions and fluid chemistry resulting from interaction of shear zone fluid (equilibrated with a 70:30 harzburgite– plagiogranite mixture) with harzburgite at T = 300 °C and p = 725 bar. ξ is a variable of reaction progress during titration, i.e. the left side of the diagrams represents conditions of high fluid/rock mass ratios closer to the melt-influenced fluid-bearing shear zone, whereas w/r ratios are decreasing towards the right of the diagram. See text for further details and discussion.

7. Discussion 7.1. Results from petrography, whole-rock geochemistry and mass balance calculations Brucite is an abundant mineral forming during serpentinization of olivine (e.g., Hostetler et al., 1966; Moody, 1976; D'Antonio and

206

N. Jöns et al. / Chemical Geology 274 (2010) 196–211

Kristensen, 2004; Beard et al., 2009), and it is also commonly found in serpentinized dunites and harzburgites of ODP Leg 209 (Kelemen et al., 2004; Paulick et al., 2006; Klein et al., 2009). The most obvious mineralogical difference, when comparing serpentinites in vicinity of altered melt veins and those occurring farther away, is the absence of brucite. In general, brucite-bearing rocks are easily affected by low-temperature alteration at the sea floor (e.g., Snow and Dick, 1995). However, increased silica activity would also lead to breakdown of brucite and formation of serpentine (usually taking place at higher temperatures, i.e. prior to exposition to weathering). In the case of the examined samples it can be ruled out that the absence of brucite is related to low-temperature alteration. Firstly, petrographic studies at Leg 209 have shown that brucite dissolution due to weathering affects only the top few meters below seafloor (mbsf; e.g., at Leg 209 Site 1271; Kelemen et al., 2004), whereas samples from depths N1 mbsf are brucite-bearing. Secondly, both processes impose different chemical changes on the rocks: when the silica activity increases, brucite will react to form serpentine, whereas brucite dissolution would just remove magnesium from the rock. Formation of serpentine does not affect the Mg# (100·MgO/[MgO + FeOT]) of the whole rock; in contrast, removal of brucite would lead to a lowering of the Mg# of the rock. Brucitefree samples from 0.7 to 0.8 mbsf at Site 1271A of Leg 209 have Mg# as low as ca. 76 (Paulick et al., 2006); the studies samples from Site 1270 have Mg# of 83–85 (Table 3), irrespective of their depth below sea floor. Thirdly, and last, weathering alters the trace element characteristics by scavenging effects, which have a strong influence on the concentration of light REEs, but less influence on heavy REE concentrations (Byrne and Kim, 1990; German et al., 1990). Samples chosen for the present study are strongly enriched also in heavy REEs (Fig. 3). High concentrations of all REEs reflect most likely the magmatic influence of a melt phase, whereas some samples that are less enriched in REEs (Fig. 3) indicate a more serpentinization fluiddominated alteration history. Similar findings have been presented by Paulick et al. (2006). The variable influence of melt veins on the host serpentinite is also displayed by the overall large variability in trace element concentrations, such as Zr/Hf = 27.9–53.3 and Nb/ Ta = 13.7–39.7. Isocon diagrams are useful for illustrating chemical changes; however, interpretation of these diagrams has to be made carefully, as chemical changes resulting from several different processes may interfere. On the one hand, the nearby evolved melt veins have certainly influenced the chemical composition and, on the other hand, elements might have been mobilized during serpentinization. Apart from the uncertainty in choosing the protolith composition, the plotting position (below or above the isocon, meaning either element loss or gain compared to an unaltered precursor) is also dependent on how the isocon has been constructed. Several approaches can be used: constant mass, constant volume or immobility of certain elements. The elements Al and Zr are often believed to be almost immobile during fluid-driven processes (e.g. Floyd and Winchester, 1978), but their concentrations may be easily changed by melt supply or melt extraction. In the present case of both hydration of the precursor harzburgite and melt influence from plagiogranitic veins, the choice of an immobile element defining the isocon is difficult. However, drawing isocons for the assumption of constant mass, an enrichment in REE compared to both the harzburgite and the serpentinite composition is striking (Fig. 5). The degree of enrichment in REE is larger compared to brucite-bearing serpentinite than compared to the least-altered harzburgite, which clearly indicates that a fluid-related process influenced the REE content. This is corroborated by the light REEs being more enriched than the heavy REEs, reflecting the increasing fluid mobility from heavy to light REE. Contrasting behaviour of HFS elements Nb and Zr is observed: in all cases, additional isocons in

the diagrams of Fig. 5 would match both data points for both rock compositions, e.g., if one draws an isocon through the Zr data point of the harzburgite (on the outer quarter circle) the Zr point of the brucite-bearing serpentinite (on the inner quarter circle) also matches this line. Minor discrepancies from this behaviour can be explained by the slightly different “constant mass isocons” for both rock compositions. This finding indicates that Zr and Nb remain immobile during the serpentinization and thus their enrichment must originate from the interaction with the melt vein. A similar behaviour is also expected for Ti, Ta and Hf; however, in the serpentinite samples from Site 1274 (that were used for comparison) concentrations of these elements are below detection limit and thus no data points can be plotted in Fig. 5. While Nb and Zr show the aforementioned fluid-immobile behaviour during serpentinization, the situation for Al2O3 is less clear: from Fig. 5 a slight enrichment during serpentinization might be deduced. As Al transport in a fluid is controlled by the difference in Al solubility between the vein assemblage and the harzburgite assemblage (see e.g., Rosing and Rose, 1993), the differential Al enrichment can be interpreted as a consequence of the proximity of a melt vein. However, as Al-rich samples are also enriched in REEs and show less positive Eu anomalies (see Table 3 and Fig.3), a mechanical mixing of vein material and host rock cannot be completely ruled out. The same holds true for the behaviour of Ca; however, it might be enriched or depleted, which is likely resulting from the variable intensity of fluid-related alteration. The enrichment in Sr and Ba is potentially related to high concentrations of these elements in evolved melts, but also alteration by seawaterderived serpentinization fluids would be a possible explanation. Finally, transition metal elements like Ni, Co, Cr and Zn plot near the isocons in Fig. 5, meaning either minimal depletion or enrichment, which is reasonable, as these elements are contained in comparatively large amounts within ultramafic rocks. Alternatively, this slight deviation from the isocon can be explained by variability in protolith compositions. 7.2. Results from reaction path models The calculated reaction path model (Fig. 6) correlates well with petrographical observations, despite the absence of brucite in rocks we investigated. The absence of brucite in the examined samples might point to serpentinization temperatures of 330–420 °C for pure harzburgites from site 1270 (cp. Fig. 6a); however, published temperature estimates for serpentinization at Leg 209 are, although variable, mostly b350 °C (e.g., Bach et al., 2004; Alt et al., 2007; Klein and Bach, 2009; Klein et al., 2009). An alternative explanation for the absence of brucite might be the elevated silica activity in immediate proximity of melt-bearing veins. High aSiO2 lithologies (gabbroic and granitic rocks) would affect phase assemblages in the adjacent serpentinites, i.e. with increasing aSiO2 brucite occurring in melt-unaffected (low aSiO2) serpentinites is first replaced by serpentine minerals and then by talc (Frost and Beard, 2007; Klein et al., 2009). Fig. 6b and c illustrates this influence of nearby plagiogranitic melt veins by variable proportions of mixed harzburgite and plagiogranite. The main changes compared to the pure harzburgite are decreasing amounts of brucite with increasing amount of plagiogranitic material mixed with harzburgite. When the plagiogranite proportion exceeds 10 wt.%, absence of brucite in the secondary mineralogy is predicted. Instead, larger amounts of serpentine and chlorite would form. Although we do not observe chlorite within the mesh textures of the serpentinites, its presence in the nearby alteration products of the hydrated melt veins indirectly confirms the model results. Mineral compositions are predicted to remain almost unchanged compared to the pure harzburgite alteration. The most pronounced changes in fluid chemistry are observed for aluminium: while fluid coexisting with

N. Jöns et al. / Chemical Geology 274 (2010) 196–211

harzburgite has Al concentrations of ca. 10 −5 mol kg−1 (for T = 200―350 °C; Fig. 6a), a harzburgite―plagiogranite mixture (90:10; Fig. 6c) coexists with a fluid with ca. 10–2 mol kg–1 Al in the same temperature range. Also concentrations of Si are higher, whereas contents of K, Ca, Mg and Fe are predicted to be markedly lower in the harzburgite―plagiogranite mixture model. In the context of the presented reaction path models, slight enrichment in Al compared to an unaltered protolith, which was discussed above, has likely occurred at the high temperature section of the reaction path, because Al contents of the fluid are highest here (Fig. 5a). An isothermal (T = 300 °C) reaction path model (Fig. 7) illustrates the influence of variable w/r ratios on the mineral chemistry. As fluidrelated alteration seems to be most intense on the melt-bearing shear zones, the model results are representative of metasomatic zoning from the shear zone into the less influenced serpentinite. Predictions from this model run fit well petrographic observations: amphibole + chlorite assemblages are found on the shear zone, the amount of magnetite in immediate vicinity of the shear zone is very low, and brucite is only stable more distal to shear zones, where lower SiO2 contents in the coexisting fluid are present. Although petrographic observations and mineral chemical measurements are in general agreement with the model results, the variability in XMg ratios of serpentine depending on the microtextural setting illustrates that even in the case of complete hydrous alteration local equilibria might persist. Therefore, the presented model results might only provide a rough estimate for the mineralogical and chemical changes in serpentinites adjacent to melt-influenced veins. Another major difference in mineral chemistry between reaction path model and rock samples from Leg 209 is the olivine composition below 400 °C (Fig. 6). The models predict that the XMg of olivine is dramatically lowered (as low as XMg = 0.72 in the harzburgite–plagiogranite model; Fig. 6c) as highly magnesian secondary minerals form. In the examined rock samples this is never found, because no growth of olivine takes place in this temperature range due to slow kinetics. Alternatively, rims of former Fe-rich olivine were replaced by serpentine minerals during subsequent equilibration at lower temperatures, and thus measured compositions are only representative of olivine from the original peridotite. 7.3. Metasomatic influence of nearby melt veins During the past decades much effort has been made to unravel the process of serpentinization (e.g., Page, 1967; Miyashiro et al., 1969; Evans and Trommsdorff, 1970; Labotka and Albee, 1979; Prichard, 1979; Frost, 1985; Cannat et al., 1992; Alt and Shanks, 1998; Alt and Shanks, 2003; Bach et al., 2004; Früh-Green et al., 2004; Bach et al., 2006; Frost and Beard, 2007; Boschi et al., 2008; McCollom and Bach, 2009), for which e.g., temperature and pressure as well as mineralogy and chemical composition of the protolith have been found to be of importance. In recent years, the availability of analyses from fluids discharging from ultramafichosted hydrothermal systems (e.g., Charlou et al., 2002; Douville et al., 2002; Schmidt et al., 2007) has furthered the understanding by combining data from rocks and fluids and thus led to a more comprehensive picture. In addition, results of several serpentinization experiments (e.g., Janecky and Seyfried, 1986; Berndt et al., 1996; Seyfried et al., 2007) yielded chemical information that now can be used for thermodynamic modeling of the whole serpentinization system. These examinations have mostly dealt with comparatively simple model systems, which are sufficient to describe many natural systems, but it has not been accounted for small-scale chemical heterogeneities that are frequently present in abyssal peridotites. However, small-scale structural and chemical heterogeneities provide conditions that are favorable for having a comparatively

207

large impact on the system: structural heterogeneities such as shear zones can act as highly effective fluid pathways, and large elemental concentration gradients present at e.g., peridotite–gabbro or peridotite–plagiogranite interfaces impose the driving force for metasomatic reactions. Rodingitization of gabbroic rocks is a typical example for the metasomatic influence of steep chemical gradients, where the metasomatic front advances from the contact with ultramafic rocks into the gabbro (e.g., Barriga and Fyfe, 1983; Austrheim and Prestvik, 2008; Frost et al., 2008; Bach and Klein, 2009). Another example is metasomatic high-pressure whiteschists that are locally observed to form at the interface between marbles and quartzites, where large chemical potential gradients exist (e.g., Kulke and Schreyer, 1973; Jöns and Schenk, 2004). In the present case of small-scale melt impregnation veins hosted in peridotites the metasomatic zoning is most likely not present on a thin section scale. The often millimeter-scaled melt veins, the occurrence of which is frequently bound to shear zones, are hydrated during progressive cooling to temperatures where the surrounding peridotites still remain largely unaltered (ca. 500 °C; Jöns et al., 2009). Nevertheless, during this stage chemical changes emanating from the shear zone fluids might be imposed upon the adjacent serpentinite by fluid migration on grain boundaries. As temperatures are higher than during typical serpentinization processes, the solubilities of elements like Al and Si in the fluid are increased, which facilitates metasomatic changes that would not or only to lesser extent be possible under lower temperatures (T b 400 °C typical for serpentinization). This holds also true for the HFSE: their mobility is enhanced by high pH (Jiang et al., 2005), which is common for serpentinization fluids (e.g., Barnes et al., 1967)) and also our reaction path calculations indicate pH N 7 (at T b 500 °C) for hydration of pure harzburgite (Fig. 6a). The influence of nearby melt veins leads to an even higher pH of N8 in the model involving 10 wt. % plagiogranite (at T b 500 °C; see Fig. 6c) or pH N 9 (at T b 300 °C; see Fig. 7c) in the model of shear zone fluids hydrating a pure harzburgite, thus allowing increased HFSE mobility at temperatures where olivine is still stable. Also at other localities mobility of HFSE during serpentinization/rodingitization has been inferred from growth of e.g. zircon (King et al., 2003; Dubińska et al., 2004) or Ti-rich hydrogarnet (Beard and Hopkinson, 2000) in metasomatic lithologies. From our isocon calculations (Fig. 5) we conclude that the HFSE variability and Al enrichment in serpentinites adjacent to melt veins developed mainly under temperatures higher than that of the main serpentinization, and suggest that this likely involved a magmatic fluid. This high-temperature hydrothermal process is overprinted by intense serpentinization leading to an almost complete hydration of the primary mineralogy. The fluid influx is emanating from shear zones, which act as fluid pathways, as evidenced by oxygen isotope data from altered melt veins (δ18O = + 3.0–4.2‰ VSMOW) and adjacent serpentinites (δ18O = + 2.6–3.7‰; Jöns et al., 2009). Values for these lithologies are somewhat lower than for serpentinites more distal from shear zones (δ18O = + 3.2–5.2‰; Alt et al., 2007) and are interpreted as an indicator for increasing influence of a seawater-derived fluid towards the shear zone. Compared to meltunaffected serpentinites from Site 1274 (Paulick et al., 2006), the samples examined in this study show a distinct enrichment in REE. This enrichment could also, at least in part, be explained by melt– rock interaction; however, a more intense addition of comparatively fluid mobile light REE as well as increasing enrichment with increasing degree of serpentinization (Fig. 5) indicates that even in the temperature range typical for serpentinization (ca. 200– 350 °C) element transport is still effective. Although mobility of Al might be deduced from comparison of serpentinites with different degrees of alteration, this process was not able to homogenize the serpentinite completely: It is still observed that serpentine in bastite textures after orthopyroxene is more aluminous than

208

N. Jöns et al. / Chemical Geology 274 (2010) 196–211

serpentine in mesh textures after olivine. This can be explained by stability of Al-bearing serpentine and chlorite in the temperature range where alumina is mobile. In addition, the fluid–rock ratio and Al content in the examined serpentinite samples seem to be too low to form other Al-bearing minerals such as garnet or zoisite. Alternatively, the deduced enrichment in Al results from the inappropriate assumption of a common protolith. We cannot rule out this possibility and have to admit that conclusions drawn from isocon diagram calculations have to be accepted with caution, because the unaltered protolith of the serpentinites is not preserved and assuming a protolith composition might introduce large errors. This uncertainty is emphasized by reconstructed protolith trace element patterns from mineral analyses of olivine, orthopyroxene, clinopyroxene and spinel using deduced mineral modes (using data of Site 1274 rocks given in Seyler et al., 2007): trace element contents of clinopyroxenes are strongly variable and thus chondrite-normalized REE pattern show large variability especially in the light REEs (M. Seyler pers. comm.). Assuming such protolith compositions instead for mass balance calculations, a less pronounced enrichment of light REEs would be deduced. Fluids in serpentinization systems are not only found to be alkaline, but also generally rich in Ca (e.g., Barnes and O'Neil, 1969; Palandri and Reed, 2004). In our harzburgite–plagiogranite mixture models the fluids are alkaline as well; however, the Ca content of the coexisting fluid is considerably lower than for alteration of a pure harzburgite, resulting from stabilization of secondary clinopyroxene and tremolitic amphibole. In pure harzburgite, calcium is present only in minor amounts in orthopyroxene and olivine and thus the coexisting fluid is comparatively rich in Ca. Although the addition of plagiogranitic material to the harzburgite increases the Ca content of the system, it also enables formation of Ca-bearing minerals, e.g. clinopyroxene, Grs-rich garnet or tremolitic amphibole (also typically found in rodingites), the presence of which in turn lowers the Ca content of the fluid. In this respect the model predictions cannot fully reproduce the reality, because calculations were done assuming simple Fe–Mg solid solution models for olivine and orthopyroxene. Although our calculations are in general agreement with the observations on mineral chemistry and whole-rock geochemistry, future work involving enhanced solid solution models is desirable to confirm our results and to improve the understanding of chemical fluxes at peridotite–plagiogranite interfaces.

8. Conclusions We have shown that the presence of evolved melt veins influences the mineralogy and chemistry of adjacent peridotites by both hightemperature melt–rock interaction and lower-temperature fluid–rock interaction processes. Serpentinites adjacent to altered plagiogranite melt veins are exposed to increased silica activity and thus lack brucite. In general, brucite-bearing rocks are easily affected be seafloor weathering; however, by inhibiting brucite formation during serpentinization the influence of nearby melt veins may thus have an indirect effect on the weathering behaviour at ambient temperatures. The presence of melt veins furthermore causes a shift of the high-pH range of the serpentinization fluid to higher temperatures, permitting increased mobility of high field strength elements. While mobility of the latter group of elements is restricted to higher temperatures (Nca. 500 °C), rare earth elements are increasingly enriched in serpentinites with increasing degree of serpentinization, i.e. in the temperature range of ca. 350–200 °C. Also alumina seems to be mobile during serpentinization; however, the hydration of serpentinites is not effective enough to erase differences in Al content of serpentine minerals in different textural settings (e.g., bastite serpentine and mesh texture serpentine). Finally, uncertainties in the quantification of element transport remain, because unaltered protoliths might be variable in composition, but are not preserved.

Acknowledgements We thank P. Appel and B. Mader for help with the electron microprobe, D. Garbe-Schönberg for whole-rock ICP-MS measurements, H. Anders and A. Klügel for assistance during LA-ICP-MS analyses, and W. Hale for help during sampling. M. Hentscher provided support during EQ3/6 database compilation and modeling. The quality of the paper benefited from reviews of J. Alt, M. Seyler and an anonymous reviewer. We thank R. Rudnick for editorial handling. This research used samples supplied by the Ocean Drilling Program (ODP), which is sponsored by the U. S. National Science foundation (NSF) and participating countries under management of Joint Oceanographic Institutions (JOI), Inc. Funding was provided by the Deutsche Forschungsgemeinschaft through grant BA 1605/2-1 and through DFG-Research Center/Excellence Cluster ‘The Ocean in the Earth System’.

Appendix A. Whole-rock X-ray diffraction patterns

Fig. A.8. X-ray diffraction patterns prepared from powder pellets of whole-rock sample powders (sample numbers correspond to Table 1). Backgrounds have been subtracted and multiples of 1000 counts have been added to samples 2–6 for better display. The major mineral phase within the serpentinites is lizardite, whereas brucite cannot be identified.

N. Jöns et al. / Chemical Geology 274 (2010) 196–211

209

Appendix B. Analyses of blank, duplicate and standards by ICP-MS

Table B.4 Analyses of blank, duplicate and standards by ICP-MS.

Li Sc V Cr Co Ni Cu Zn Ga Rb Sr Y Zr Nb Mo Sn Sb Cs Ba La Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Hf Ta W Pb Th U

Blank

Sample no. 3†

BHVO-2

UB-N

N=1

N=4

1σ

RSD %

N=1

N=1

0.001 0.000 0.041 0.232 0.000 0.044 2.422 0.187 0.000 0.001 0.013 0.000 0.087 0.001 0.007 0.004 0.000 0.000 0.009 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.000 0.002 0.000 0.003 0.292 0.000 0.000

10.1 9.24 39.7 2492 97.7 1914 2.49 51.4 2.09 0.397 6.79 2.93 6.48 1.83 0.460 0.247 0.037 0.010 6.36 1.78 4.12 0.510 1.95 0.405 0.151 0.417 0.070 0.450 0.097 0.288 0.047 0.332 0.052 0.229 0.117 0.082 0.235 0.323 0.322

0.100 0.154 0.574 37.3 1.19 25.3 0.023 0.635 0.016 0.003 0.042 0.014 0.017 0.009 0.005 0.002 0.001 0.000 0.043 0.006 0.006 0.001 0.005 0.002 0.001 0.002 0.000 0.003 0.001 0.002 0.000 0.002 0.000 0.002 0.001 0.001 0.002 0.003 0.003

0.99 1.66 1.44 1.50 1.22 1.32 0.92 1.24 0.77 0.79 0.62 0.48 0.26 0.47 1.12 1.01 2.91 1.26 0.68 0.35 0.14 0.25 0.23 0.45 0.63 0.40 0.31 0.57 0.54 0.71 0.25 0.65 0.75 0.84 0.49 1.05 1.04 0.78 0.82

4.62 31.3 308 193 44.4 118 126 105 22.3 9.49 393 25.6 161 17.1 4.52 1.67 0.178 0.097 128 15.1 36.1 5.39 24.9 6.17 2.07 6.26 0.950 5.39 0.988 2.45 0.327 2.01 0.277 4.30 1.07 0.252 4.90 1.19 0.406

28.7 12.7 63.8 1905 94.3 1742 22.6 80.1 2.66 3.46 7.57 2.58 3.35 0.055 0.526 0.262 0.270 10.8 25.9 0.305 0.744 0.117 0.600 0.216 0.080 0.307 0.060 0.422 0.094 0.272 0.043 0.292 0.045 0.121 0.013 16.5 12.8 0.064 0.056

Concentrations in μg g−1; RSD = relative standard deviation; † sample number corresponds to Table 1.

Appendix C. Analyses of standards by LA-ICP-MS

Table C.5 Analyses of standards by LA-ICP-MS. NIST 612*

Li Ti V Cr Co Ni Cu Zn Rb† Sr Y Zr Nb Ba La

NIST 610*

BCR2-G

GOR132-G

N = 15

1σ

RSD %

N = 16

1σ

RSD %

N=4

1σ

RSD %

N=4

1σ

41.6 38.4 39.3 35.6 35.3 38.7 36.7 37.9 31.8 77.7 38.6 35.7 35.7 37.8 38.8

0.80 1.18 0.85 1.79 0.57 1.11 0.94 1.36 3.58 1.12 0.992 0.733 0.699 0.634 0.589

1.93 3.06 2.16 5.03 1.61 2.86 3.18 0.96 3.00 1.44 2.57 2.06 1.96 1.68 1.52

n.a. 433.0 n.a. 404.7 n.a. n.a. n.a. n.a. n.a. 496.8 449.6 439.5 418.6 424.3 458.0

– 5.82 – 4.77 – – – – – 4.79 5.550 7.200 5.480 6.424 6.926

– 1.35 – 1.18 – – – – – 0.96 1.23 1.64 1.31 1.51 1.51

9.30 14,548 403 16.4 35.3 12.1 15.5 129 42.5 301 30.1 154 10.8 580 22.0

0.942 1288 24.7 1.49 2.38 1.15 0.985 25.7 4.04 28.4 2.91 14.9 0.999 63.0 2.24

10.1 8.85 6.15 9.07 6.73 9.50 6.36 19.9 9.51 9.42 9.64 9.62 9.29 10.9 10.1

9.61 1972 225 2653 96.7 1244 204 68.7 2.12 14.7 12.8 9.36 0.040 0.671 0.076

0.228 32.9 6.99 54.6 3.45 40.8 13.6 2.70 0.051 0.554 0.557 0.277 0.021 0.221 0.031

RSD % 2.37 1.67 3.10 2.06 3.57 3.28 6.64 3.93 2.41 3.77 4.36 2.95 53.4 32.9 41.1

(continued on next page)

210

N. Jöns et al. / Chemical Geology 274 (2010) 196–211

Table C.5 (continued) NIST 612*

Ce Pr Nd Sm Eu Gd Tb Dy Ho Er Tm Yb Lu Hf Ta Pb Th U

NIST 610*

BCR2-G

N = 15

1σ

RSD %

N = 16

1σ

RSD %

39.1 37.6 36.3 38.2 37.7 36.7 39.0 36.0 39.3 36.8 36.5 40.3 37.6 36.7 32.4 37.4 38.0 37.6

0.635 0.678 0.592 0.915 0.735 0.751 0.811 0.832 0.822 0.836 0.912 1.15 1.07 1.01 0.912 1.61 1.11 0.789

1.62 1.80 1.63 2.39 1.95 2.04 2.08 2.31 2.09 2.27 2.50 2.85 2.85 2.76 2.82 4.30 2.92 2.08

447.6 429.7 430.8 450.9 461.0 420.3 443.1 426.5 449.7 426.5 420.7 462.1 434.7 417.7 376.5 414.1 450.5 457.3

5.048 4.679 5.887 6.891 7.402 8.319 7.413 7.343 8.534 9.607 10.08 11.53 10.07 10.79 9.095 11.70 11.38 9.505

1.13 1.09 1.37 1.53 1.61 1.98 1.67 1.72 1.90 2.25 2.40 2.49 2.32 2.58 2.42 2.83 2.53 2.08

References Aldiss, D.T., 1981. Plagiogranites from the ocean crust and ophiolites. Nature 289, 577–578. Alt, J.C., Shanks III, W.C., 1998. Sulfur in serpentinized oceanic peridotites: serpentinization processes and microbial sulfate reduction. Journal of Geophysical Research 103, 9917–9929. Alt, J.C., Shanks III, W.C., 2003. Serpentinization of abyssal peridotites from the MARK area, Mid-Atlantic Ridge: sulfur geochemistry and reaction modeling. Geochimica et Cosmochimica Acta 67, 641–653. Alt, J.C., Shanks III, W.C., Bach, W., Paulick, H., Garrido, C.J., Beaudoin, G., 2007. Hydrothermal alteration and microbial sulfate reduction in peridotite and gabbro exposed by detachment faulting at the Mid-Atlantic Ridge, 15°20′N (ODP Leg 209): a sulfur and oxygen isotope study. Geochemistry, Geophysics, Geosystems 8, Q08002. Anonymous, 1972. Penrose field conference on ophiolites. Geotimes 17, 24–25. Armstrong, J.T., 1995. CITZAF: A package of correction programs for the quantitative electron microbeam X-ray analysis of thick polished materials, thin films, and particles. Microbeam Analysis 4, 177–200. Austrheim, H., Prestvik, T., 2008. Rodingitization and hydration of the oceanic lithosphere as developed in the Leka ophiolite, north-central Norway. Lithos 104, 177–198. Bach, W., Klein, F., 2009. The petrology of seafloor rodingites: insights from geochemical reaction path modeling. Lithos 112, 103–117. Bach, W., Garrido, C.J., Paulick, H., Harvey, J., Rosner, M., 2004. Seawater–peridotite interaction: first insights from ODP Leg 209, MAR 15°N. Geochemistry, Geophysics, Geosystems 5, Q09F26. Bach, W., Paulick, H., Garrido, C.J., Ildefonse, B., Meurer, W.P., Humphris, S.E., 2006. Unraveling the sequence of serpentinization reactions: petrography, mineral chemistry, and petrophysics of serpentinites from MAR 15°N (ODP Leg 209, Site 1274). Geophysical Research Letters 33, L13306. Barnes, I., O'Neil, J.R., 1969. The relationship between fluids in some fresh alpine-type ultramafics and possible modern serpentinization, western United States. Geological Society of America Bulletin 80, 1947–1960. Barnes, I., LaMarche Jr., V.C., Himmelberg, G., 1967. Geochemical evidence of presentday serpentinization. Science 156, 830–832. Barriga, F., Fyfe, W.S., 1983. Development of rodingite in basaltic rocks in serpentinites, East Liguria, Italy. Contributions to Mineralogy and Petrology 84, 146–151. Beard, J.S., Hopkinson, L., 2000. A fossil, serpentinization-related hydrothermal vent, Ocean Drilling Program Leg 173, Site 1068 (Iberia Abyssal Plain): some aspects of mineral and fluid chemistry. Journal of Geophysical Research 105, 16527–16539. Beard, J.S., Frost, B.R., Fryer, P., McCaig, A., Searle, R., Ildefonse, B., Zinin, P., Sharma, S.K., 2009. Onset and progression of serpentinization and magnetite formation in olivine-rich troctolite from IODP Hole U1309D. Journal of Petrology 50, 387–403. Berndt, M.E., Allen, D.E., Seyfried Jr., W.E., 1996. Reduction of CO2 during serpentinization of olivine at 300 °C and 500 bar. Geology 24, 351–354. Bogdanov, Y.A., Bortnikov, N.S., Vikent'ev, I.V., Gurvich, E.G., Sagalevich, A.M., 1997. A new type of modern mineral-forming system: black smokers of the hydrothermal field at 14°45′N latitude, Mid-Atlantic ridge. Geology of Ore Deposits 39, 58–78. Boschi, C., Früh-Green, G.L., Delacour, A., Karson, J.A., Kelley, D.S., 2006. Mass transfer and fluid flow during detachment faulting and development of an oceanic core complex, Atlantis Massif (MAR 30°N). Geochemistry, Geophysics, Geosystems 7, Q01004. Boschi, C., Dini, A., Früh-Green, G.L., Kelley, D.S., 2008. Isotopic and element exchange during serpentinization and metasomatism at the Atlantis Massif (MAR 30°N): insights from B and Sr isotope data. Geochimica et Cosmochimica Acta 72, 1801–1823. Byrne, R.H., Kim, K.-H., 1990. Rare earth element scavenging in seawater. Geochimica et Cosmochimica Acta 54, 2645–2656. Cannat, M., Bideau, D., Bougault, H., 1992. Serpentinized peridotites and gabbros in the Mid-Atlantic axial valley at 15°37′N and 16°52′N. Earth and Planetary Science Letters 109, 87–106.

N=4 46.3 5.72 24.8 5.66 1.59 5.34 0.859 5.51 1.08 3.13 0.427 2.91 0.450 3.72 0.639 7.10 4.97 1.39

GOR132-G 1σ

RSD % 4.64 0.566 2.27 0.464 0.124 0.542 0.080 0.791 0.091 0.506 0.056 0.249 0.088 0.570 0.090 2.90 0.802 0.202

10.0 9.91 9.19 8.20 7.79 10.2 9.37 14.4 8.48 16.2 13.1 8.53 19.5 15.3 14.1 40.8 16.1 14.6

N=4 0.304 0.075 0.624 0.472 0.197 1.16 0.266 2.04 0.466 1.51 0.223 1.55 0.214 0.297 0.021 17.7 0.002 0.049

1σ 0.072 0.008 0.150 0.061 0.013 0.048 0.022 0.049 0.041 0.088 0.007 0.107 0.006 0.048 0.007 0.589 0.004 0.015

RSD % 23.5 10.3 24.0 12.9 6.78 4.15 8.18 2.41 8.81 5.85 2.99 6.87 2.86 16.2 31.7 3.34 173 29.9

Charlou, J.L., Bougault, H., Appriou, P., Nelsen, T., Rona, P., 1991. Different TDM/CH4 hydrothermal plume signatures: TAG site at 26°N and serpentinized ultrabasic diapir at 15°05′N on the Mid-Atlantic Ridge. Geochimica et Cosmochimica Acta 55, 3209–3222. Charlou, J.L., Donval, J.P., Fouquet, Y., Jean-Baptiste, P., Holm, N., 2002. Geochemistry of high H2 and CH4 vent fluids issuing from ultramafic rocks at the Rainbow hydrothermal field (36°14′N, MAR). Chemical Geology 191, 345–359. Coleman, R.G., Peterman, Z.E., 1975. Oceanic plagiogranite. Journal of Geophysical Research 80, 1099–1108. D'Antonio, M., Kristensen, M.B., 2004. Serpentine and brucite of ultramafic clasts from the South Chamorro Seamount (Ocean Drilling Program Leg 195, Site 1200): inferences for the serpentinization of the Mariana forearc mantle. Mineralogical Magazine 68, 887–904. de Wit, M., Stern, C.R., 1976. A model for ocean-floor metamorphism, seismic layering and magnetism. Nature 264, 615–619. Delacour, A., Früh-Green, G.L., Bernasconi, S.M., Kelley, D.S., 2008. Sulfur in peridotites and gabbros at Lost City (30°N, MAR): implications for hydrothermal alteration and microbial activity during serpentinization. Geochimica et Cosmochimica Acta 72, 5090–5110. Dick, H.J.B., Bullen, T., 1984. Chromian spinel as a petrogenetic indicator in abyssal and alpine-type peridotites and spatially associated lavas. Contributions to Mineralogy and Petrology 86, 54–76. Douville, E., Charlou, J.L., Oelkers, E.H., Bienvenu, P., Jove Colon, C.F., Donval, J.P., Fouquet, Y., Prieur, D., Appriou, P., 2002. The rainbow vent fluids (36°14′N, MAR): the influence of ultramafic rocks and phase separation on trace metal content in Mid-Atlantic Ridge hydrothermal fluids. Chemical Geology 184, 37–48. Dubińska, E., Bylina, P., Kozłowski, A., Dörr, W., Nejbert, K., Schastok, J., Kulicki, C., 2004. U–Pb dating of serpentinization: hydrothermal zircon from a metasomatic rodingite shell (Sudetic ophiolite, SW Poland). Chemical Geology 203, 183–203. Escartín, J., Hirth, G., Evans, B., 2001. Strength of slightly serpentinized peridotites: implications for the tectonics of ocean lithosphere. Geology 29, 1023–1026. Evans, B.W., 2004. The serpentinite multisystem revisited: chrysotile is metastable. International Geology Review 46, 479–506. Evans, B.W., Trommsdorff, V., 1970. Regional metamorphism of ultramafic rocks in the central Alps: paragenesis in the system CaO–MgO–SiO2–H2O. Schweizerische Mineralogische und Petrographische Mitteilungen 50, 481–492. Ferry, J.M., Watson, E.B., 2007. New thermodynamic models and revised calibrations for the Ti-in-zircon and Zr-in-rutile thermometers. Contributions to Mineralogy and Petrology 154, 429–437. Flanagan, F.J., 1976. Description and analyses of eight new USGS rock standards. USGS Professional Paper 840, 1–192. Floyd, P.A., Winchester, J.A., 1978. Identification and discrimination of altered and metamorphosed volcanic rocks using immobile elements. Chemical Geology 21, 291–306. Frost, B.R., 1985. On the stability of sulfides, oxides, and native metals in serpentinite. Journal of Petrology 26, 31–63. Frost, B.R., Beard, J.S., 2007. On silica activity and serpentinization. Journal of Petrology 48, 1351–1368. Frost, B.R., Beard, J.S., McCaig, A., Condliffe, E., 2008. The formation of micro-rodingites from IODP Hole U1309D: key to understanding the process of serpentinization. Journal of Petrology 49, 1579–1588. Früh-Green, G.L., Connolly, J.A.D., Plas, A., Kelley, D.S., Grobéty, B., 2004. Serpentinization of oceanic peridotites: implications for geochemical cycles and biological activity. The subseafloor biosphere at Mid-Ocean ridges. : Geophysical Monograph Series, vol. 144. American Geophysical Union, pp. 119–136. Fujiwara, T., Lin, J., Matsumoto, T., Kelemen, P.B., Tucholke, B.E., Casey, J.F., 2003. Crustal evolution of the Mid-Atlantic Ridge near the Fifteen-Twenty Fracture Zone in the last 5 Ma. Geochemistry, Geophysics, Geosystems 4 (3), 1024. Garbe-Schönberg, C.-D., 1993. Simultaneous determination of thirty-seven trace elements in twenty-eight international rock standards by ICP-MS. Geostandards Newsletter 17, 81–97.

N. Jöns et al. / Chemical Geology 274 (2010) 196–211 German, C.R., Klinkhammer, G.P., Edmond, J.M., Mitra, A., Elderfield, H., 1990. Hydrothermal scavenging of rare-earth elements in the ocean. Nature 345, 516–518. Govindaraju, K., Roelandts, I., 1989. 1988 compilation report on trace elements in six ANRT rock reference samples: diorite DR-N, serpentine UB-N, bauxite BX-N, disthene DT-N, granite GS-N and potash feldspar FK-N. Geostandards Newsletter 13, 5–67. Gràcia, E., Charlou, J.L., Radford-Knoery, J., Parson, L.M., 2000. Non-transform offsets along the Mid-Atlantic Ridge south of the Azores (38°N–34°N): ultramafic exposures and hosting of hydrothermal vents. Earth and Planetary Science Letters 177, 89–103. Grant, J.A., 1986. The isocon diagram—a simple solution to Gresens' equation for metasomatic alteration. Economic Geology 81, 1976–1982. Grant, J.A., 2005. Isocon analysis: a brief review of the method and applications. Physics and Chemistry of the Earth 30, 997–1004. Gresens, R.L., 1967. Composition–volume relationships of metasomatism. Chemical Geology 2, 47–65. Hostetler, P.B., Coleman, R.G., Mumpton, F.A., Evans, B.W., 1966. Brucite in Alpine serpentinites. American Mineralogist 51, 75–98. Humphris, S.E., Alt, J.C., Teagle, D.A.H., Honnorez, J.J., 1998. Geochemical changes during hydrothermal alteration of basement in the stockwork beneath the active TAG hydrothermal mound. In: Herzig, P.M., Humphris, S.E., Miller, D.J., Zierenberg, R.A. (Eds.), Proceedings of the Ocean Drilling Program, Scientific Results, vol. 158. College Station, pp. 255–276. Ildefonse, B., Blackman, D.K., John, B.E., Ohara, Y., Miller, D.J., MacLeod, C.J., Integrated Ocean Drilling Program Expeditions 304/305 Science Party, 2007. Oceanic core complexes and crustal accretion at slow-spreading ridges. Geology 35, 623–626. Imai, N., Terashima, S., Itoh, S., Ando, A., 1995. 1994 compilation values for GSJ reference samples, “Igneous rock series”. Geochemical Journal 29, 91–95. Janecky, D.R., Seyfried Jr., W.E., 1986. Hydrothermal serpentinization of peridotite within the oceanic crust: experimental investigations of mineralogy and major element chemistry. Geochimica et Cosmochimica Acta 50, 1357–1378. Jiang, S.-Y., Wang, R.-C., Xu, X.-S., Zhao, K.-D., 2005. Mobility of high field strength elements (HFSE) in magmatic-, metamorphic-, and submarine–hydrothermal systems. Physics and Chemistry of the Earth 30, 1020–1029. Jochum, K.P., Willibold, M., Raczek, I., Stoll, B., Herwig, K., 2005. Chemical characterisation of the USGS reference glasses GSA-1G, GSC-1G, GSD-1G, GSE-1G, BCR-2G, BHVO-2G and BIR-1G using EPMA, ID-TIMS, ID-ICP-MS and LA-ICP-MS. Geostandards and Geoanalytical Research 29, 285–302. Jochum, K.P., Stoll, B., Herwig, K., Willbold, M., Hofmann, A.W., Amini, M., Aarburg, S., Abouchami, W., Hellebrand, E., Mocek, B., Raczek, I., Stracke, A., Alard, O., Bouman, C., Becker, S., Dücking, M., Brätz, H., Klemd, R., de Bruin, D., Canil, D., Cornell, D., de Hoog, C.-J., Dalpé, C., Danyushevsky, L., Eisenhauer, A., Gao, Y., Snow, J.E., Groschopf, N., Günther, D., Latkoczy, C., Guillong, M., Hauri, E.H., Höfer, H.E., Lahaye, Y., Horz, K., Jacob, D.E., Kasemann, S.A., Kent, A.J.R., Ludwig, T., Zack, T., Mason, P.R.D., Meixner, A., Rosner, M., Misawa, K., Nash, B.P., Pfänder, J., Premo, W.R., Sun, W.D., Tiepolo, M., Vannucci, R., Vennemann, T., Wayne, D., Woodhead, J.D., 2006. MPI-DING reference glasses for in situ microanalysis: new reference values for element concentrations and isotope ratios. Geochemistry, Geophysics, Geosystems 7, Q02008. John, T., Klemd, R., Gao, J., Garbe-Schönberg, D., 2008. Trace element mobilization in slabs due to non steady-state fluid–rock interaction: constraints from an eclogitefacies transport vein in blueschist (Tianshan, China). Lithos 103, 1–24. Johnson, J.W., Oelkers, E.H., Helgeson, H.C., 1992. SUPCRT92: a software package for calculating the standard molal thermodynamic properties of minerals, gases, aqueous species, and reactions from 1 to 5000 bar and 0 to 1000 °C. Computers and Geosciences 18, 899–947. Jöns, N., Schenk, V., 2004. Petrology of whiteschists and associated rocks at Mautia Hill (Tanzania): fluid infiltration during high-grade metamorphism? Journal of Petrology 45, 1959–1981. Jöns, N., Bach, W., Schroeder, T., 2009. Formation and alteration of plagiogranites in an ultramafic-hosted detachment fault at the Mid-Atlantic Ridge (ODP Leg 209). Contributions to Mineralogy and Petrology 157, 625–639. Kaczmarek, M.-A., Müntener, O., 2008. Juxtaposition of melt impregnation and hightemperature shear zones in the upper mantle; field and petrological constraints from the Lanzo Peridotite (Northern Italy). Journal of Petrology 49, 2187–2220. Kelemen, P.B., Matsumoto, T., Shipboard Scientific Party, 1998. Geological results of Mode 98, Leg1: JAMSTEC/WHOI Shinkai 6500 cruise to 15°N, Mid-Atlantic Ridge. EOS Transactions 79 (Fall Meeting Suppl.) (U22A-18). Kelemen, P.B., Kikawa, E., Miller, D.J., et al., 2004. Proceedings of the Ocean Drilling Program, Initial Reports, 209 [CD-ROM]. Ocean Drilling Program, Texas A&M University, College Station TX 77845-9547, USA. Kelley, D.S., Karson, J.A., Blackman, D.K., Früh-Green, G.L., Butterfield, D.A., Lilley, M.D., Olson, E.J., Schrenk, M.O., Roe, K.K., Lebon, G.T., Rivizzigno, P., AT3-60 Shipboard Party, 2001. An off-axis hydrothermal vent field near the Mid-Atlantic Ridge at 30°N. Nature 412, 145–149. Kelley, D.S., Karson, J.A., Früh-Green, G.L., Yoerger, D.R., Shank, T.M., Butterfield, D.A., Hayes, J.M., Schrenk, M.O., Olson, E.J., Proskurowski, G., Jakuba, M., Bradley, A., Larson, B., Ludwig, K., Glickson, D., Buckman, K., Bradley, A.S., Brazelton, W.J., Roe, K., Elend, M.J., Delacour, A., Bernasconi, S.M., Lilley, M.D., Baross, J.A., Summons, R.E., Sylva, S.P., 2005. A serpentinite-hosted ecosystem: the Lost City hydrothermal field. Science 307, 1428–1434. King, R.L., Kohn, M.J., Eiler, J.M., 2003. Constraints on the petrologic structure of the subduction zone slab–mantle interface from Franciscan Complex exotic ultramafic blocks. Geological Society of America Bulletin 115, 1097–1109. Klein, F., Bach, W., 2009. Fe–Ni–Co–O–S phase relations in peridotite–seawater interactions. Journal of Petrology 50, 37–59. Klein, F., Bach, W., Jöns, N., McCollom, T.M., Moskowitz, B., Berquó, T., 2009. Iron partitioning and hydrogen generation during serpentinization of abyssal perido-

211

tites from 15°N on the Mid-Atlantic Ridge. Geochimica et Cosmochimica Acta 73, 6868–6893. Koepke, J., Berndt, J., Feig, S.T., Holtz, F., 2007. The formation of SiO2-rich melts within the deep oceanic crust by hydrous partial melting of gabbros. Contributions to Mineralogy and Petrology 153, 67–84. Kulke, H., Schreyer, W., 1973. Kyanite–talc schist from Sar E Sang, Afghanistan. Earth and Planetary Science Letters 18, 324–328. Labotka, T.C., Albee, A.L., 1979. Serpentinization of the Belvidere Mountain ultramafic body, Vermont: mass balance and reaction at the metasomatic front. Canadian Mineralogist 17, 831–845. McCollom, T.M., Bach, W., 2009. Thermodynamic constraints on hydrogen generation during serpentinization of ultramafic rocks. Geochimica et Cosmochimica Acta 73, 856–875. McDonough, W.F., Sun, S.S., 1995. The composition of the Earth. Chemical Geology 120, 223–253. Melchert, B., Devey, C.W., German, C.R., Lackschewitz, K.S., Seifert, R., Walter, M., Mertens, C., Yoerger, D.R., Baker, E.T., Paulick, H., Nakamura, K., 2008. First evidence for high-temperature off-axis venting of deep crustal/mantle heat: the Nibelungen hydrothermal field, southern Mid-Atlantic Ridge. Earth and Planetary Science Letters 275, 61–69. Miyashiro, A., Shido, F., Ewing, M., 1969. Composition and origin of serpentinites from the Mid-Atlantic Ridge near 24° and 30° North Latitude. Contributions to Mineralogy and Petrology 23, 117–127. Moody, J.B., 1976. Serpentinization: a review. Lithos 9, 125–138. O'Brien, D., Carton, M., Eardly, D., Patching, J.W., 1998. In situ filtration and preliminary molecular analysis of microbial biomass from the Rainbow hydrothermal plume at 36°15′N on the Mid-Atlantic Ridge. Earth and Planetary Science Letters 157, 223–231. O'Driscoll, B., Petronis, M.S., 2009. Oxide mineral formation during the serpentinization of a Cr-spinel seam: insights from rock magnetic experiments. Geochemistry, Geophysics, Geosystems 10, Q01008. Page, N.J., 1967. Serpentinization at Burro Mountain, California. Contributions to Mineralogy and Petrology 14, 321–342. Palandri, J.L., Reed, M.H., 2004. Geochemical models of metasomatism in ultramafic systems: serpentinization, rodingitization, and sea floor carbonate chimney precipitation. Geochimica et Cosmochimica Acta 68, 1115–1133. Passchier, C.W., Simpson, C., 1986. Porphyroclast systems as kinematic indicators. Journal of Structural Geology 8, 831–843. Paulick, H., Bach, W., Godard, M., De Hoog, J.C.M., Suhr, G., Harvey, J., 2006. Geochemistry of abyssal peridotites (Mid-Atlantic Ridge, 15°20′N, ODP Leg 209): implications for fluid/rock interaction in slow spreading environments. Chemical Geology 234, 179–210. Pearce, N.J.G., Perkins, W.T., Westgate, J.A., Gorton, M.P., Jackson, S.E., Neal, C.R., Chenery, S.P., 1997. A compilation of new and published major and trace element data for NIST SRM 610 and NIST SRM 612 glass reference materials. Geostandards Newsletter 21, 115–144. Prichard, H.M., 1979. A petrographic study of the process of serpentinisation in ophiolites and the ocean crust. Contributions to Mineralogy and Petrology 68, 231–241. Rosing, M., Rose, N.M., 1993. The role of ultramafic rocks in regulating the concentrations of volatile and non-volatile components during deep crustal metamorphism. Chemical Geology 108, 187–200. Schmidt, K., Koschinsky, A., Garbe-Schönberg, D., de Carvalho, L.M., Seifert, R., 2007. Geochemistry of hydrothermal fluids from the ultramafic-hosted Logatchev hydrothermal field, 15°N on the Mid-Atlantic Ridge: temporal and spatial investigations. Chemical Geology 242, 1–21. Schroeder, T., Cheadle, M.J., Dick, H.J.B., Faul, U., Casey, J.F., Kelemen, P.B., 2007. Nonvolcanic seafloor spreading and corner-flow rotation accommodated by extensional faulting at 15°N on the Mid-Atlantic Ridge: a structural synthesis of ODP Leg 209. Geochemistry, Geophysics, Geosystems 8, Q06015. Seyfried Jr., W.E., Foustoukos, D.I., Fu, Q., 2007. Redox evolution and mass transfer during serpentinization: an experimental and theoretical study at 200 °C, 500 bar with implications for ultramafic-hosted hydrothermal systems at Mid-Ocean Ridges. Geochimica et Cosmochimica Acta 71, 3872–3886. Seyler, M., Lorand, J.-P., Dick, H.J.B., Drouin, M., 2007. Pervasive melt percolation reactions in ultra-depleted refractory harzburgites at the Mid-Atlantic Ridge, 15°20′N: ODP hole 1274A. Contributions to Mineralogy and Petrology 153, 303–319. Smith, D.K., Cann, J.R., Escartín, J., 2006. Widespread active detachment faulting and core complex formation near 13°N on the Mid-Atlantic Ridge. Nature 442, 440–443. Snow, J.E., Dick, H.J.B., 1995. Pervasive magnesium loss by marine weathering of peridotite. Geochimica et Cosmochimica Acta 59, 4219–4235. van der Straaten, F., Schenk, V., John, T., Gao, J., 2008. Blueschist-facies rehydration of eclogites (Tian Shan, NW-China): implications for fluid–rock interaction in the subduction channel. Chemical Geology 255, 195–219. Wessel, P., Smith, W.H.F., 1991. Free software helps map and display data. EOS Transactions 72, 441. Wessel, P., Smith, W.H.F., 1998. New, improved version of generic mapping tools released. EOS Transactions 79, 579. Wilson, S.A., 1997. Data compilation for USGS reference material BHVO-2, Hawaiian basalt. Tech. rep. U.S. Geological Survey. Wolery, T.J., 2002. EQ3/6—software for geochemical modeling. Wolery, T.J., Jove-Colon, C.F., 2004. Qualification of thermodynamic data for geochemical modeling of mineral–water interactions in dilute systems. Tech. rep. U.S. Department of Energy.