IODP Proceedings    Volume contents     Search

doi:10.2204/iodp.proc.343343T.103.2013

Geochemistry

Interstitial water geochemistry

Twelve whole-round core samples from 178.89 to 831.45 mbsf were collected for interstitial water analysis (Table T10). A single spot core (343-C0019E-1R) taken at the top of this interval is ~465 m shallower than the other cores and will not be presented in the figures. The remaining 11 cores were recovered from 689.44 to 831.45 mbsf, spanning the two major fault zones identified in LWD data and reaching into mudstone 1.2 m above lithologic Unit 6. Core sections processed for interstitial water were 15.5 to 31 cm long, and the volumes of interstitial water obtained were 4.2–23 mL except for the volume from Core 1R (63 mL). The interstitial water whole round from Core 19R was taken from two intervals on either side of a whole-round sample selected for anelastic strain recovery measurement. The whole-round sections with good condition for interstitial water sampling were limited, and the lengths of the whole rounds were generally shorter than ideal. Only three samples had sufficient water to measure alkalinity and pH (Table T10).

Possible contamination by drilling fluids

The 11 deepest interstitial water samples contained detectable dissolved sulfate at depths that should be below the sulfate–methane transition (SMT) zone. Sulfate concentrations ranged from 3.41 to 20.24 mM (Fig. F58; Table T10), compared to a seawater value of 28.9 mM. Generally, the presence of sulfate here would be taken as evidence for contamination of the whole rounds by drilling fluids. However, the observed interstitial water sulfate concentrations, if derived from contamination by drilling fluids, would suggest contamination of 12%–70%. The largest sulfate concentration was measured in a whole-round sample (Core 343-C0019E-20R) that was thoroughly cleaned by scraping, and only intact unfractured pieces were used for interstitial water extraction, so contamination is considered unlikely. Results from experiments using the perfluorocarbon (PFC) tracer (see “Microbiology”) indicate minimum contamination by drilling fluids. Further evidence against contamination is that a correction for drilling fluid using the sulfate proxy would result in negative magnesium and potassium concentrations in some cases. Finally, cross-plots of strontium-sulfate and cesium-lithium (Fig. F59) provide evidence for multiple fluid reservoirs and indicate interstitial water samples with high sulfate concentrations represent waters with in situ dissolved sulfate. For the data plotted here, no sulfate correction is applied, and the presence of dissolved sulfate is interpreted as evidence for a sulfate-bearing fluid reservoir at depth.

Sulfate, alkalinity, pH, phosphate, and ammonium

Because of the deep coring plan with incomplete core recovery through the hole, the sulfate profile to the SMT zone could not be established. Sulfate concentrations (SO42–) from 689 to 816 mbsf generally vary in a range of <10 mM (Fig. F58; Table T10). Sulfate concentrations trend to lower values from 787 to 816 mbsf, the latter sample being the last taken from lithologic Unit 3 above the faulted horizon in Core 343-C0019E-17R. A large sulfate concentration is observed in each of the two interstitial water samples below Core 17R in Units 5 and 6 and above the chert horizon in Unit 7. These last two samples were obtained as relatively intact pieces from the whole round and it is unlikely that sulfate found in the interstitial water is due to drilling contamination, which suggests a deep source of sulfate-bearing fluids. Such high sulfate concentrations at these depths are rarely observed in accretionary prisms, although there is evidence at one of the Nankai reference section sites for dissolved sulfate entering the base of the sedimentary section, perhaps through oceanic basement (Expedition 322 Scientists, 2011).

Because alkalinity and pH could be measured in only two deep interstitial water samples (Fig. F58), no useful trends can be discerned. Measured alkalinity of the sample from 720.42 mbsf is relatively high (35 mM) compared to that of the deeper sample (7.9 mM). Given the low calcium concentrations and lack of carbonate found in the cores, it is unlikely that alkalinity values are substantially affected by carbonate precipitation. Measured pH is also relatively high in the sample from 720.42 mbsf (7.91) compared to the deeper sample (7.64). However, both alkalinity and pH from 720.42 mbsf are comparable to values measured on the sample from 178.89 mbsf.

The highest phosphate (PO4) concentration of 96 µM is observed at 179 mbsf in Unit 1. Deeper values are substantially lower and generally decrease with depth (Fig. F58), possibly related to removal by oxide phases. Ammonium (NH4) concentrations decrease systematically in the deeper core intervals (>688 mbsf). Phosphate and ammonium concentrations are likely related to production of these metabolites during degradation at shallower depths.

Salinity, chlorinity, and bromide

Salinity values are similar to seawater within most of Unit 3 (688.50–820.01 mbsf). However, salinity decreases to low values of 33.5 parts per thousand (ppt) at 810.53 mbsf and 33.6 ppt at 816.50 mbsf (Fig. F58). The low salinity values suggest either an exotic fluid source or possibly advanced dehydration of hydrous minerals at the base of Unit 3, immediately above the zone of highly sheared sediment. Below Unit 4, the plate boundary décollement zone (see “Structural geology”), salinity returns to near-seawater values in Cores 343-C0019E-19R and 20R.

Chlorinity (Cl) is similar to seawater concentrations at the upper target horizon, has slightly higher than seawater values at 787 and 802 mbsf, and then monotonically decreases toward the bottom of the hole (Fig. F58). The decrease suggests a reservoir of low-chlorinity fluids deeper than the maximum depth reached by drilling. A slight minimum at 699 and 706 mbsf in Unit 3 might reflect a similar low-chlorinity fluid source. The minimum is also located at the horizon with the highest hydrogen concentration (see below). The general similarity of Cl to seawater concentrations in much of the hole is consistent with the low amounts of observed volcanic matter in the dominant dark gray mudstone of Unit 3. Alteration of ash to clay minerals would have resulted in increases in Cl concentrations. Br concentrations are higher than seawater throughout the section, but show a general decrease with depth (Fig. F58). The lowest concentrations are observed in the bottom two samples of Hole C0019E. Elevated Br concentrations might reflect release from organic matter.

Major cations (Na, K, Mg, and Ca)

Major cations are lower than seawater values throughout most of the hole and show little variation in concentration in Unit 3 (Fig. F60). The very low calcium concentrations in Units 2 and 3 are consistent with the generally low abundances of observed carbonate minerals in visual core description. The lack of cation variation in these units likely reflects minimal alteration of sparse volcanic matter to clay minerals. Near the bottom of the hole (below 820 mbsf) in Units 5 and 6, calcium markedly increases and both potassium and magnesium decrease. The observed profiles below 780 mbsf are consistent with the formation of clay minerals (illite, chlorite, or mixed-layer clays) associated with fluid-rock interactions. The narrow depth range suggests this process is only important in the deepest part of the hole. The increase in calcium concentrations might be expected to promote carbonate formation, but little carbonate was observed in the deep sediment horizons. Sodium concentrations, calculated from charge balance, decrease from Units 3 to 5 and could reflect minor alteration of volcanic matter to Na-bearing clays (smectite).

Minor and trace elements

The depth dependence of rubidium and cesium is consistent with that of other alkali metals, sodium and potassium, and decreases from Units 3 to 5 (Fig. F61). The decrease of cesium in the lowest part of the hole is remarkable. These observations are consistent with the involvement of fluids associated with Na-K–rich clay mineral formation. Silica concentrations are high throughout the hole, possibly related to the high abundance of siliceous microfossils (Fig. F60). There is a minimum in silica at 713 mbsf. Deeper than 780 mbsf, silica values increase with depth, negatively correlated with the values of sodium, potassium, magnesium, rubidium, and cesium. This suggests a supply of silica associated with mineral reactions such as clay mineral formation. Lithium and boron correspond with silica and are likely controlled by a similar mechanism to that which governs silica concentrations (Fig. F60). Manganese concentrations increase abruptly below the inferred décollement. Manganese values that are well correlated with calcium values are also likely to result from such mineral reactions.

Barium shows a complicated depth profile but is negatively correlated with sulfate values (Fig. F60). This strongly suggests the control of barium concentration by sulfate species in fluids and rocks. In contrast to barium, strontium shows relatively constant concentrations with a slight increase at the bottom of the hole, which is rather consistent with variations in silica, boron, and lithium (Figs. F60, F61). Decoupling of these alkali-earth metals indicates the formation of a barium-specific phase, possibly barite, in the deeper part of the hole.

Depth profiles for some of the trace metals such as zinc, copper, iron, molybdenum, lead, and uranium show a strong positive peak at ~700 mbsf near the upper fault horizon (Fig. F61). Zinc and molybdenum are also enriched in the lowermost part of the hole. Although the cause of trace metal enrichment at ~700 mbsf is unknown, it is possible that these anomalies result from fluid transfer associated with a seismic or postseismic event as suggested by hydrogen data.

Control of pore fluid compositions by mixing

The variations in pore fluid chemistry described above suggest the involvement of fluid-mineral interactions at various depths. Such fluid-mineral interactions may have produced two or more end-component fluids that mixed to form the observed pore fluid compositions. The relationship between sulfate and strontium concentrations and lithium and cesium concentrations for seawater and high-density drilling mud fluid (Table T11) are shown in Figure F59. The lack of clear correlation in these diagrams indicates that at least three end-components are controlling the pore fluid compositions, one of which is a seawater-like fluid. The data from fluids collected at depths shallower than 730 mbsf exhibit clear linear trends in the mixing diagrams, indicating mixing of seawater-like fluid and fluid characterized by low sulfate–high strontium and high lithium–high cesium concentrations. Some of the pore fluids collected from depths greater than 780 mbsf exhibit a different mixing trend, for which the contribution of the seawater-like component largely disappears, and another end-component fluid likely characterized by high sulfate–high strontium and high lithium–low cesium is involved. It should be noted that the seawater-like end-member component requires sulfate concentrations distinctly lower than seawater. Thus, this end-member component is not seawater itself and might represent deeper fluids of modified seawater composition. Although the drilling mud fluid apparently fits the end-component with high sulfate and low strontium, this scenario fails to explain the observed low lithium and cesium. Therefore, neither pure seawater nor drilling mud fluid can be end-component fluids, excluding the probability of pore fluid contamination by drilling. Therefore, two unidentified end-member fluids control the composition of deep pore fluids.

Carbon, nitrogen, and sulfur concentrations

The total carbon (TC), total nitrogen (TN), and total sulfur (TS) data are summarized in Table T12 and Figure F62. Calcium carbonate (CaCO3) values range from 0.05 to 32.1 wt%. The highest values are at 688.82 and 689.41 mbsf. The average calcium carbonate value is 2.3 wt% and most samples have values <1 wt%. Total organic carbon (TOC) values remain low throughout the section, averaging 0.38 wt%. Some of the lowest TOC values are in Units 5 and 6 of the lower plate sedimentary section. Like TC, TN is relatively uniform in sediments of Unit 3 above the décollement. The TOC/TN average ratio of 5.47 is on the low end of that expected for marine origin, but most samples from Unit 3 (mean TOC/TN = 6.85) fall within the range of 6–9 expected for marine organic matter. TS content is mostly low, ranging from 0.01 to 3.12 wt%, with an average of 0.30 wt%. There are two local sulfur maxima at 705.83 and 810.50 mbsf. Excluding the two highest values, the mean sulfur content is 0.17 wt%.

Gas chemistry

Dissolved gas components in the sediments, such as H2, CO, CH4, and C2H6 were measured at the bottom of each section of the recovered core in Hole C0019E (Table T13), and the depth profile of concentration of each gas component is shown in Figure F63. It should be noted that the gas component concentrations in Cores 343-C0019E-2R and 3R are likely reduced from potential in situ concentrations because of heavy fragmentation and mixing of the core material during recovery (see “Lithology”). Thus, the following description and discussion based on the abundance of gas components excludes data from Cores 2R and 3R (Table T13).

A large and spatially restricted increase in H2 concentration was found in the bottom of Section 343-C0019E-5R-1 (Fig. F63) at 697.4 mbsf. This sharp spike of H2 includes neighboring sampling points in Sections 4R-CC and 5R-2 (Table T13). H2 concentrations in this maximum were significantly higher than the other representative background levels of dissolved H2 concentrations in Hole C0019E sediments, or so-called “drill bit–induced” H2 concentrations.

Based on field observations and laboratory experiments, mechanochemical H2 is expected to be produced with the creation of new surface area in rocks during friction work, and thus large amounts of H2 is likely generated on major faults during earthquakes in natural environments (e.g., Wakita et al., 1980; Wiersberg and Erzinger, 2008; Hirose et al., 2011). Because many geophysical observations suggest that the Tohoku-oki earthquake induced a large amount of slip along one or more faults at depth beneath Site C0019 (Fujiwara et al., 2011), we predicted that a possible earthquake-induced H2 concentration anomaly was generated and may still be observable in the sediments of Site C0019 14 months after the Tohoku-oki earthquake. The maximum H2 concentration at ~700 mbsf may represent a signature of this predicted earthquake-induced H2. A similar anomaly (minimum value) was also observed in the depth profile of the methane/H2 ratio (CH4/H2) (Fig. F63). This value reveals a more standardized index of relative H2 abundance in the subseafloor sediments and may constrain the duration of anomalous H2 concentrations after the time of generation. In the LWD measurements, significant low gamma ray and resistivity signals were found at ~700 mbsf (see “Logging while drilling”). In addition, low chlorinity values were detected in the interstitial water obtained from the closest section of Core 5R (Section 5R-2) (Fig. F58). Together with the structural characterization of Section 343-C0019E-5R-1 (see “Structural geology”), these measurements suggest the existence of a major fault and may be associated with fault activity during a large earthquake. Due to incomplete core recovery from ~700 to ~705 mbsf, a detailed H2 concentration anomaly profile from this section was not obtained. However, the H2 anomaly is the first potential example of earthquake-induced H2 production discovered in deep subseafloor sediments. Future isotopic and chemical-microbiological analyses will provide important insights into the earthquake-induced environmental impacts and the postearthquake biogeochemical and microbiological processes in the deep subseafloor environments of subduction zones.

The CO concentration depth profile is generally similar to the H2 concentration profile but the pattern is less clear (Fig. F63). The stable C1/C2 ratio with depth indicates that the methane and ethane profiles are quite similar (Fig. F63). Methane concentrations are high (>1000 µM), except in a few samples, and exhibit some scatter (Table T13). Unusually low methane concentrations were found in Core 343-C0019E-20R. With increasing depth in Core 20R, methane concentration decreases to 9.6 µM (Fig. F63). The lowest methane concentration, at Section 20R-2 (833.5 mbsf), is two to three orders of magnitude lower than at other sampling points in Hole C0019E. This steep depletion might result from two processes: (1) very little in situ production of methane and/or (2) rapid diffusion of methane at this depth range in Core 20R. In particular, the rapid diffusion of methane points to the possible occurrence of relatively high fluid flow beneath Core 20R, likely within the oceanic basement basalt. The deep depletion of methane concentration is consistent with the observed decrease of chlorinity. These results strongly suggest the presence of another fluid end-member that is chemically distinct from seawater or possible faulting-enhanced fluid flow. Future compositional and chemical analyses and microbiological characterizations will clarify the detailed chemical properties of and the origin of potential deep-sourced fluids.