Mineralogy and geochemistry of Mg-phyllosilicates from Middle Valley, northern Juan de Fuca Ridge

We present results of a detailed mineralogical and geochemical study of the progressive hydrothermal alteration of clastic sediments recovered at ODP Site 858 in an area of active hydrothermal venting at the sedimented, axial rift valley of Middle Valley (northern Juan de Fuca Ridge). These results allow a characterization of newly formed phyllosilicates and provide constraints on the mechanisms of clay formation and controls of mineral reactions on the chemical and isotopic composition of hydrothermal fluids. Hydrothermal alteration at Site 858 is characterized by a progressive change in phyllosilicate assemblages with depth. In the immediate vent area, at Hole 858B, detrital layers are intercalated with pure hydrothermal precipitates at the top of the section, with a predominance of hydrothermal phases at depth. Sequentially downhole in Hole 858B, the clay fraction of the pure hydrothermal layers changes from smectite to corrensite to swelling chlorite and finally to chlorite. In three pure hydrothermal layers in the deepest part of Hole 858B, the clay minerals coexist with neoformed quartz. Neoformed and detrital components are clearly distinguished on the basis of morphology, as seen by SEM and TEM, and by their chemical and stable isotope compositions. Corrensite is characterized by a 24 Å stacking sequence and high Si- and Mg-contents, with Fe/(Fe+Mg) ratio of = 0.08. We propose that corrensite is a unique, possibly metastable, mineralogical phase and was precipitated directly from seawater-dominated hydrothermal fluids. Hydrothermal chlorite in Hole 858B has a stacking sequence of 14 Å with Fe/(Fe+Mg) ratios of ? 0.35. The chemistry and structure of swelling chlorite suggest that it is a corrensiteychlorite mixed-layer phase. The mineralogical zonation in Hole 858B is accompanied by a systematic decrease in d18O, reflecting both the high thermal gradients that prevail at Site 858 and extensive sediment-fluid interaction. Precipitation of the Mg-phyllosilicates in the vent region directly controls the chemical and isotopic compositions of the pore fluids. This is particularly evident by decreases in Mg and enrichments in deuterium and salinity in the pore fluids at depths at which corrensite and chlorite are formed. Structural formulae calculated from TEM-EDX analyses were used to construct clay-H2O oxygen isotope fractionation curves based on oxygen bond models. Our results suggest isotopic disequilibrium conditions for corrensite-quartz and swelling chlorite-quartz precipitation, but yield an equilibrium temperature of 300° C ± 30° for chlorite-quartz at 32 m below the surface. This estimate is consistent with independent estimates and indicates steep thermal gradients of 10-11°/m in the vent region.

Geochemistry at DSDP Leg 70 Holes

The hydrothermal mounds on the southern flank of the Galapagos Spreading Center are characterized by the following main features: 1) They are located over a young basement (0.5 to 0.85 m.y. of age) in a region known for its high sedimentation rate (about 5 cm/10**3 y.) because it is part of the equatorial high biological productivity zone. 2) They are located in a region with generally high heat flow (8 to 10 HFU). The highest heat-flow measurements (up to 10**3 HFU) correspond to mound peaks (Williams et al., 1979), where temperatures up to 15°C were measured during a dive of the submersible Alvin (Corliss et al., 1978). 3) They are often located on small vertical faults which displace the basement by a few meters (Lonsdale, 1977) and affect the 25- to 50-meter-thick sediment cover. Most of these characteristics have also been observed in the other three known cases of hydrothermal deposits with mineral parageneses similar to that of the Galapagos mounds. However, the case of the hydrothermal mounds south of the Galapagos Spreading Center is unique because of the unusual thickness of the hydrothermal deposits present. The mounds are composed of several, up to 4.5-meter-thick, layers of green clays which, in one case (Hole 509B), are overlain by about 1.4 meters of Mn-oxide crust. We suspect that such a large accumulation of hydrothermal products results from the "funnelling" of the hydrothermal solutions exiting from a highly permeable basement along the faults. This chapter reports a preliminary study of those green clays collected by hydraulic piston coring of the Galapagos mounds during Deep Sea Drilling Project (DSDP) Leg 70 of the D/V Glomar Challenger. Green clays have also been reported from three presently or recently active hydrothermal areas in or close to spreading centers.

Paleomagnetic of ODP Hole 197-1203A

On the basis of studies of Holocene samples,submarine basaltic glass (SBG) is thought to be an ideal paleointensity recorder because it contains unaltered single domain magnetic inclusions that yield Thellier paleointensity data of exceptional quality. To be useful as a recorder of the long-term geomagnetic field, older SBG must retain these optimal properties. Here, we examine this issue through rock magnetic and transmission electron microscope (TEM) analyses of Cretaceous SBG recovered at Ocean Drilling Program Site 1203 (northwestern Pacific Ocean). These SBG samples have very low natural remanent magnetization intensities (NRM <50 nAm**2/g) and TEM analyses indicate a correspondingly low concentration of crystalline inclusions. Thellier experiments on samples with the strongest NRM intensity (>5*10**-11 Am**2) show a rapid acquisition of thermoremanent magnetization (TRM) with respect to NRM demagnetization. Taken at face value,this behavior implies magnetization in a very weak (617 WT) ambient field. But monitoring of magnetic hysteresis properties during the Thellier experiments (on subsamples of the SBG samples used for paleointensity determinations) indicates systematic variations in values over the same temperature range where the rapid TRM acquisition is observed. A similar change in properties during heating is observed on monitor SBG specimens using low-temperature data: with progressive heatings the Verwey transition becomes more distinct. We suggest that these experimental data record the partial melting and neocrystallization of magnetic grains in SBG during the thermal treatments required by the Thellier method,resulting in paleointensity values biased to low values. We further propose that this process is pronounced in Cretaceous and Jurassic SBG (relative to Holocene SBG) because devitrification on geologic time scales (i.e., tens of millions of years) lowers the transition temperature at which the neocrystallization can commence. Magnetic hysteresis monitoring may provide a straightforward means of detecting the formation of new magnetic inclusions in SBG during Thellier experiments.