OXYGEN AND CARBON ISOTOPE OF SERPENTINES, MAGNETITES, AND CALCITES
Serpentine and Magnetite
Oxygen isotope results for serpentine are reported in
Table 6. The oxygen isotope compositions for all the lizardite samples are strongly
enriched in 18O (between 8.8‰ and 12.2‰), in comparison to the normal mantle
18O values for fresh peridotite olivine and pyroxene minerals (5‰ to 6‰; Taylor, 1968;
Javoy, 1970) except for chrysotile from Sample 149-897D-16R-2, 35-42 cm, which is depleted in 18O. As oxygen is the principal constituent of silicate minerals in
peridotites, such a large modification of the initial 18O requires interaction with large quantities of hydrothermal fluids during the
serpentinization. High water-rock ratio conditions have been shown to be generally met in continental serpentinizing systems (Barnes and O'Neil, 1971) and oceanic serpentinized peridotites (Kimball and
Gerlach, 1986; Snow and Dick, in press).
For two samples (149-897C-72R-1, 40-47 cm, and 149-897D-23R-1, 62-68 cm) the coexisting magnetite and lizardite have been measured for their
18O values. In both cases oxygen isotope fractionation,
18Oserpentine–magnetite, is very large (11.0‰ and 12.5‰, respectively) and is similar to the previously published values for oceanic serpentinized peridotites
(Wenner and Taylor, 1973). Although the 18O fractionation between serpentine and magnetite has not yet been calibrated vs. temperature by laboratory experiments, some natural observations and regularities have led Wenner and Taylor (1971) to propose the following "tentative"
geothermometer:
1000 ln[
18Oserpentine–magnetite] = 1.69 (106/T2) + 1.68.
This calibration has been recently supported by the theoretical calculations of Zheng and Simon (1991) and Zheng (1993). Accordingly, the measured
18Oserpentine–magnetite values for Sample 149-897C-72R-1, 40-47 cm, and for Sample 149-897D-23R-1, 62-68 cm, correspond to temperatures of 150°C and 120°C (±50°C), respectively. These temperatures are in the range of those determined for the lizardites (and the chrysotiles also) of the oceanic serpentinized peridotites (70° to 250°C; Wenner and Taylor, 1971; Sheppard, 1980;
Bonatti et al., 1984; Hébert et al., 1990). As the associated uncertainties in these calibrations are probably considerable, these estimated temperatures mean that the serpentinization has occurred at low temperature, namely less than 200°C.
The other high 18O values of nonmagnetite paired samples are similar to the previous samples and are therefore compatible with low-temperature serpentinization processes (below 150°C) in seawater-dominated conditions. In this respect, these serpentines are similar to the serpentines from Hole 637A to the north
(Agrinier et al., 1988; Evans and Girardeau, 1988). No significant difference among the various serpentine types analyzed are observed: white serpentine from veins or from the background mesh, colorless, and pale-green fibrous serpentine from vein margins, all have
18O values around 10‰. As the
18O composition of serpentine is highly sensitive to temperature change, especially at low temperature
(Wenner and Taylor, 1971, 1973), this narrow range of 18O values suggests that serpentinization mostly affected these peridotites during a single (low-temperature) stage. Less equivocal than the low-temperature serpentinization conditions that can be found down to 5 km below seafloor (see estimates of the temperature gradient made from heat flux measurements; Louden and
Mareschal; this volume), the high 18O values of serpentines (around 10‰) necessarily imply that serpentinization occurred under very high seawater-rock ratios (Taylor, 1977). Such seawater-rock ratio conditions are unlikely to be met deep in the crust or in the mantle. This low-temperature serpentinization very likely started when they were brought to near seafloor position.
The contrasting vein chrysotile sample (149-897D-16R-2, 35-42 cm) with a low
18O value of +3.8‰ is lower than any observed in Hole 637A and in Holes 897C, 897D, and 899B. If it results from the serpentinization of the fresh
peridotite, we infer that it formed in the presence of significant amounts of fluid (high water-rock ratio) because the
18O value of this chrysotile is also strongly shifted relative
to its precursors (fresh olivine and orthopyroxene have
18O around 5.5‰). Since the serpentine-water oxygen isotope fractionation decreases with temperature, its low
18O value suggests that it formed at a significantly higher temperature (possibly 300°C). This chrysotile could be a remnant of the high-temperature serpentinization episode exemplified by the chlorites and the amphiboles. No other chrysotile vein sample has been recorded in Leg 149
serpentinites. The textural relationships of this sample suggest that chrysotile formed first in a vein and then was mantled by lizardite during the low-temperature serpentinization episode.
The carbon isotope values of calcites range from -0.7‰ to 2.4‰. They are similar to those commonly reported for seawater marine carbonates and vein carbonates from the upper oceanic crust
(Javoy and Fouillac, 1979; Bonatti et al., 1980; Stakes and O'Neil, 1982). In contrast to the calcites from Hole 637A peridotites
(Agrinier et al., 1988), there is no evidence for low 13C values, which would reveal some contribution from mantle carbon
(Javoy and Fouillac, 1979). Calcites from veins, cracks, or from the serpentine mesh have similar
13C and
18O values. The oxygen isotope values of these calcites range from 29.6‰ to 31.1‰. This narrow range corresponds to that of low-seawater-temperature marine carbonates. Assuming the calcites formed from seawater (18O = 0‰), calcite isotopic temperatures between 19° and 13°C can be derived (O'Neil et al., 1969). These inferred temperatures would not significantly change (less than 10°C) if these calcites formed from low-temperature 18O-depleted pore waters (18O between 0‰ and
–3‰, Lawrence et al., 1975).
