IODP Proceedings    Volume contents     Search

doi:10.2204/iodp.proc.322.103.2010

Inorganic geochemistry

The main geochemical objective at Site C0011, located seaward of the trench, was to document the variations in interstitial water chemical composition. Such data may be used to elucidate the origins, volume, and nature of fluid and chemical transport and the fluid-rock interactions that may affect the state and geotechnical properties of the strata at the décollement and their evolution after subduction. A total of 46 interstitial water samples were squeezed from selected whole-round sections for chemical and isotopic analyses. Samples depths ranged from 360.0 to 867.5 m CSF. One sample per core was collected when possible. Because of poor core recovery and/or intense disturbance, no samples of interstitial water were recovered from Cores 322-C0011B-1R, 2R, 17R, 18R, 20R, 22R, 29R, 34R, 41R, 46R, 49R, 50R, 54R, 60R, and 61R.

Fluid recovery and contamination

To obtain enough interstitial water for shipboard and shore-based analyses, 30 to 40 cm long sections were squeezed in Unit II (340–479.60 m CSF). Longer sections (50 to 56 cm in length) were collected for interstitial water extraction in Units III–V (below 479.60 m CSF, starting at Core 322-C0011B-17R) because these sediments are more consolidated. The interstitial water volumes, which were recovered from whole-round sections by squeezing at a maximum pressure of 25,000 psi, are presented as a function of depth in Figure F55. A drastic change in interstitial water recovery occurs between Units II and III. Interstitial water volume ranges from 7.5 to 82.5 mL between 359.61 and 478.21 m CSF (Cores 322-C0011B-3R through 16R). The higher recovered volumes correspond to sandy lithologies. Below a tightly cemented terrigenous sandstone, from which no interstitial waters were recovered (Cores 322-C0011B-17R and 18R at 479.0 and 481.4 m CSF, respectively), the average interstitial water volume fell to 7 ± 5 mL, even after increasing the whole-round section length by 25%. Because of the lithified nature of the formation at this site and technical problems during coring operations, many of the cores were very disturbed. Removing the outer layers of these highly fractured and friable samples was not always possible; thus, many interstitial water samples are contaminated with drilling fluid (Figs. F55, F56). We began coring operations in Hole C0011B at 340 m CSF, and thus, we did not sample the sulfate–methane transition (SMT) zone; however, we can assume that this transition lies well above our shallowest sample and use the sulfate concentration to identify and quantify contamination as indicated in "Inorganic geochemistry" in the "Methods" chapter. The degree of drilling fluid contamination of the highly disturbed samples reaches 30% (Fig. F55); we chose to discard all data from samples in which contamination is >15%. There is no apparent correlation between the degree of contamination and volume of interstitial water recovered (Fig. F55).

Interstitial water data collected at Site C0011 are listed in Table T17. In addition, in Table T18 we list sulfate-corrected data, which represent the composition of the interstitial water corrected for drilling fluid contamination. In these samples, contamination ranges from 2% to 14%, with the higher values (>10%) corresponding to disturbed, friable samples that were hard to clean properly. The sulfate-corrected data are illustrated in Figures F57 and F58. These contamination-corrected distributions reflect the combined effects of organic matter diagenesis, hydration and dehydration reactions, precipitation of authigenic phases, and alteration of volcanogenic sediments and oceanic basement.

Biogeochemical processes

Because the upper 340 m of this site was not cored, we do not have information on the shallow diagenetic processes. Data from sites drilled on this margin during Leg 190 (Shipboard Scientific Party, 2001b) and IODP Expeditions 315 and 316 (Tobin et al., 2009) show that the base of the sulfate reducing zone occurs in the upper 20 m CSF. The observed ammonium and phosphate distributions at Site C0011 (Fig. F57) reflect the production of these metabolites at shallower sections than those drilled, although the actual production and consumption pathways cannot be defined at this site. Alkalinity is also probably produced in the shallower sections, and the observed low values reflect consumption by authigenic carbonate formation (Fig. F57).

Halogen concentration (Cl and Br)

In the cored section recovered from Site C0011, Cl concentration generally decreases from ~550 mM to values that are ~7% lower than those of seawater. This freshening trend is superficially consistent with that observed at ODP Site 1177, drilled seaward of the deformation front in the Ashizuri transect during Leg 190 (Fig. F59). In general, some possible causes for the freshening include dehydration of smectite at depth and Cl uptake into an authigenic hydrous phase. As originally postulated for the profiles observed at Site 1177 (Shipboard Scientific Party, 2001b), hydration of volcanic ash layers combined with dehydration reactions may have contributed to the observed Cl variability. But subsequent shore-based analyses of clay minerals and numerical modeling of thermal conditions and diagenetic reaction progress (Steurer and Underwood, 2003; Saffer et al., 2008; Saffer and McKiernan, 2009) indicate that the freshened interstitial water at Site 1177 originates from greater depth arcward; fluid migration updip probably occurred along highly permeable sand beds. This possibility also exists at Site C0011. Chlorinity increases in the deeper sections of Hole C0011B (below ~700 m CSF), and this trend is probably the result of hydration reactions during basement alteration and diffusional exchange with a more seawater-like fluid in the basalt (see "Inorganic geochemistry" in the "Site C0012" chapter). There is no significant variability in the Br/Cl ratio downcore, but observed values (1.6 to 1.8 mmol/mol) (Fig. F57) are slightly higher than seawater.

Chemical changes due to alteration of volcaniclastics

The dominant lithology in Unit II is an alternation of green-gray silty claystone and medium- to thick-bedded volcaniclastic and tuffaceous sandstone. To evaluate the chemical exchange due to volcanic glass alteration, we consider here the alteration of anorthite, one of the major and most reactive mineralogical constituents of glass. Dissolution of anorthite leads to a release of calcium, aluminum, and silica to the interstitial water following Reaction 1. The presence of magnesium and potassium in the interstitial water permits the precipitation of montmorillonite-(Ca, Mg, K), a dioctahedral smectite, as a secondary mineral as described by Reaction 2. By combining anorthite dissolution and montmorillonite precipitation and assuming that aluminium is not mobilized, we can write the overall chemical Reaction 3.

CaAl2(SiO4)2 + 8H+ =
Ca2+ + 2Al3+ + 2H4SiO2(aq).

(1)

4H4SiO2(aq) + 1.67Al3+ + (0.33 + x)Mg2+ + (2y)K+
+ (0.165 – x y)Ca2+ = (Ca0.165 – × – y, Mgx, K2y)
Mg0.33Al1.67Si4O10(OH)2.

(2)

Anorthite + 2.8H4SiO2(aq) + (0.33 + x)Mg2+
+ (2y)K+ = montmorillonite-(Ca, Mg, K)
+ (0.8 + x + y)Ca2+.

(3)

In basaltic glass, olivine and pyroxene alteration will provide extra aqueous silica for montmorillonite precipitation. Thus, the expected effect of anorthite alteration, and more generally of volcanic glass, will be an uptake of magnesium and potassium from interstitial water and a release of calcium. These changes are common in ash-bearing sediments (e.g., Gieskes et al., 1990) as well as in deep formations where exchanges with underlying basement rocks are known to occur (e.g., Lawrence and Gieskes, 1981). It is worth noting that interstitial water from Site C0011 bears a significantly more extensive alteration fingerprint (Fig. F59) than that from Site 1177, with almost complete removal of potassium and magnesium below 400 m CSF. The observed concomitant change in dissolved silica suggests that ash alteration is also controlling the silica profile. Alternatively, the silica distribution may reflect temperature-dependent opal diagenesis as postulated for Sites 1173 and 1177 by Spinelli et al. (2007).

In an attempt to elucidate these equilibria, we performed calculations of the interstitial water saturation state using the PhreeqcI software package (Parkhust and Appelo, 1999) and integrated Lawrence Livermore National Laboratory database, "thermo.com.V8.R.230" (Johnson et al., 1992), using values for pure end-members of mineral phases and assuming temperatures of 5° and 25°C. The results shown in Figure F60 are a first-order approximation of the thermodynamic equilibrium and do not include any kinetic effects. The results for the formation of montmorillonite illustrate supersaturation of this mineral above 400 m CSF and an envelope of values that approximate equilibrium at depths >400 m CSF, suggesting that this reaction controls the distribution of silica, potassium, and magnesium in interstitial water. The resulting removal of potassium from interstitial water observed at this site raises the question of the source of potassium for smectite to illite conversion, even after this sediment sequence is deeply buried and warmed to sufficiently high temperatures.

Dissolved calcium values at Site C0011 are much higher than those measured at Site 1177. A very sharp increase in dissolved calcium with depth occurs in Units IV and V, probably as a result of calcium supply from basement alteration reactions. The high levels of dissolved calcium at this site support carbonate precipitation, even at very low alkalinity (<2 mM). This expectation is consistent with calculated saturation values for calcite and dolomite (Fig. F60) and the presence of carbonate throughout the cored section as vein filling, carbonate cements, and discrete layers as thick as 2.7 cm (see "Lithology"). Bulk powder XRD shows that analcime (NaAlSi2O6·H2O), heulandite ([Ca,Na2][Al2Si7O18]·6H2O), and/or clinoptilolite ([Na,K,Ca]2–3Al3[Al,Si]2Si13O36·12H2O) are present in Unit V. These zeolites form as the devitrification product of glass in tuffs and volcanic rocks. The decrease in Na in interstitial water in the lower sections sampled is consistent with the formation of these minerals, which may also be favored by high calcium concentration in the basal fluids. As shown in Figure F60, Na-clinoptilolite is the best candidate to control Na concentration in interstitial water, as it reaches equilibrium below 500 m CSF, and the lack of potassium in the fluids probably hinders formation of K-clinoptilolite.

Strontium and lithium profiles also show significant enrichment at depth relative to seawater. The increase in strontium in interstitial water as a product of reaction with volcanoclastic material is well established (e.g., Gieskes et al., 1990), and shore-based isotopic characterization of dissolved strontium will aid in constraining this Sr source. In contrast, authigenic clay formation at low to moderate temperatures incorporates Li, with preference to the light 6Li isotope, such that there is a decrease in the Li content of interstitial water, and Li should become isotopically heavier because of this process (e.g., Chan and Kastner, 2000). Field and laboratory observations document remobilization of lithium from aluminosilicates at temperatures ranging from 70° to 100°C (Edmond et al., 1979; Seyfried et al., 1984), and the isotopic composition of the mobilized Li depends on the temperature-dependent isotopic fractionation. The deeper the sediment source, the lighter the dissolved Li should be; this is because at higher temperatures there is less of an isotopic fractionation. Light Li isotopic signature has been observed at the décollement zone of ODP Sites 808 (Nankai) and 1040 (offshore Costa Rica), which signifies migration of deeply generated fluids (You et al., 1995; Chan and Kastner, 2000).