Previous   |   Next

doi:10.2204/iodp.proc.313.102.2010

Geochemistry

Offshore interstitial water sampling and analysis

Cores were sampled for microbiology and pore water immediately on recovery. Pore water was sampled using Rhizon samplers as long as the sediment was soft enough to insert the plastic tip of the sampler. A Rhizon sampler (CSS-F 5 cm; Rhizosphere Research Products, Netherlands) is a narrow elongated cylindrical filter (0.2 µm pore size; 5 cm long; 130 µL volume) with a stiff plastic core (Seeberg-Elverfeldt et al., 2005). Before use it is soaked for several hours in pure water, which is ejected under pressure by syringe and discarded prior to sampling. A Rhizon sampler is inserted into the core through a 3.8 mm hole, the same diameter as the Rhizon sampler to facilitate sealing, and drilled through the plastic core liner using a spacer on the drill bit to prevent it from going into the core itself. Rhizon samplers draw water from sediment under negative pressure, which is applied by attaching the Rhizon sampler to a 20 mL all-plastic syringe, pulling the plunger back, and bracing it there with a small wooden stick. Pore water is collected in the syringe. Rhizon samplers were left in a core until the rate of water collection slowed greatly, typically within 1–10 h. For long extractions, we removed the Rhizon samplers so that the core could be logged and then reapplied fresh Rhizon samplers to the same holes afterward. Usually one or two but up to four Rhizon samplers were used per sample, spaced 10 or 20 cm apart downcore, to obtain sufficient pore water; water from the separate syringes would then be combined into a single sample in a 50 mL syringe for analysis. If the sediment was too stiff to penetrate with the plastic tip of the Rhizon sampler, we made a 2.5 mm hole first with a steel rod and then pushed the Rhizon sampler into it.

Within a depth range of 190–280 mbsf, the sediment became too hard and dry to extract water with Rhizon samplers. We then switched to squeezing of whole rounds 5 cm in length in the Bremen Teflon-lined titanium squeezer of piston-cylinder design (described in detail at www.marum.de/​en/​Pore_water_squeezer.html). Whole rounds were nearly always cut from the bottom of the first section of core, as this material was most likely to be undisturbed; Rhizons were usually inserted into the first section well below its top for the same reason. Whole rounds were scraped clean on all surfaces to remove drilling mud and material from the outer surface that may have been smeared down the core liner. They were then shaved to <54 mm diameter to fit into the squeezer. Once in the squeezer, sediment was contained when under pressure (up to 15 tonnes, applied by a hydraulic press) by a circular titanium screen at the base of the cylinder overlain by nylon mesh and a paper filter (Whatman #1), through which the water was pushed to an online filter (0.45 µm; 25 mm; nylon) and through that into a 50 mL all-plastic syringe. All parts of the squeezer that touch sediment or pore water are either polytetra-fluoroethylene (PTFE), titanium (grade 2), or Polyamid plastic (Delrin). We measured the volume of pore water extracted, the downward distance the piston traveled during squeezing (Δh), and the height and weight of the final squeezed sediment sample, now compacted into a cylinder under the pressure of squeezing. We used these measurements to estimate the initial and final sediment densities before and after squeezing and thus the compressibility of the material, as well as the fraction of interstitial water that was squeezed out. Typically, sediment samples could be squeezed to a final dry density of 2.4–2.5 g/cm³.

Drilling mud was used throughout Expedition 313 to remove cuttings from the hole. Various formulations were used, generally including bentonite clay with a biodegradable biopolymer (starch) as a suspension enhancer. Mud was mixed in tanks with seawater from the ship's firepump system. During drilling, mud pumped downhole flushes the outside of the core barrel and exits the drill pipe through a small gap between the barrel shoe and the inner side of the drill bit. Sufficiently porous or coarse sediment can thus become contaminated with drilling mud, especially if disturbed, and must be avoided in sampling for pore water analysis. This was readily done when cutting whole rounds for squeezing, as these could be inspected in detail and cleaned before putting them into the squeezer. It was more difficult when Rhizon samplers were used, as it was sometimes difficult to assess the condition of the sediment through the clear plastic core liner. We did not sample pore water from any depth interval unless the material penetrated by the Rhizon was at least moderately stiff and therefore likely to be intact. We took samples of each of the seven mud formulations used during Expedition 313, separated the liquid from the solids using Rhizons, and analyzed this liquid along with pore water samples and ambient seawater to test for contamination. In filtering the drilling muds we found that they readily clogged the Rhizons, resulting in very slow flow rates. This rapid clogging effectively prevented us from inadvertently sampling drilling fluid instead of pore water, especially within the loose sands and disturbed sediment that were common within the upper 200 m of the sequence.

Once pore water was extracted from the sediment core into a syringe by Rhizon sampler or by squeezing, it was pushed through a 0.45 µm online filter (25 mm; nylon) into various containers for analysis. The water was analyzed immediately onboard for pH, alkalinity, ammonium, and salinity by refractive index. pH was measured by ion-specific electrode (±0.05 units; 1σ; based on 107 analyses of ambient seawater), alkalinity by single-point titration to pH < 3.9 with 0.01 N HCl using a 0.2–0.5 mL sample (±5.8%; 1σ; based on 109 analyses of ambient seawater), and ammonium by conductivity, after separation as NH₃ through a PTFE membrane in a flow-through system. In the latter technique, modified after Hall and Aller (1992), ammonia is stripped from a 100 µL sample by an alkaline carrier solution (0.2 M Na citrate in 10 mM NaOH), passed through a 200 mm × 5 mm PTFE membrane, and redissolved as NH₄⁺ in an acidic solution (1 mM HCl). Ammonium ion is then measured as the resulting conductivity signal in the acidic carrier, in a micro flow-through cell.

We monitored the drilling operation and cores for H₂S and flammable gases using the ESO gas "sniffer," a handheld unit equipped with a small pump for pumping ambient gas past a sensor. We detected H₂S only twice and flammable gases four times, only in Holes M0027A and M0028A. No H₂S odor was ever detected in any pore water sample, so no analyses were done for this species.

A total of 179 samples of sediment were collected onboard for headspace analysis, usually from the base of Section 1 adjacent to the interstitial water sample. A plug of sediment weighing 8.5 ± 1.6 g (1σ) was injected into a 20 mL glass vial prefilled with 10 mL of a solution of 0.5 M NaCl and 0.1 M HgCl₂ and shaken vigorously to mix for postcruise determination of the concentration of dissolved methane and its stable C isotopic composition.

Aliquots of interstitial water for OSP and postcruise analyses included an acidified aliquot (10 µL concentrated HNO₃/mL) for inductively coupled plasma–atomic emission spectroscopy (ICP-AES) (Li, Na, K, Mg, Ca, Sr, Ba, Mn, Fe, B, Si, S, P, Al, and Ti), inductively coupled plasma–mass spectrometry (ICP-MS) (Rb, Cs, I, Cu, Ni, Zn, Co, Mo, V, Cr, and U), and Sr isotopic composition; an unacidified aliquot for chlorinity, chloride, bromide, sulfate, nitrate, and fluoride; untreated aliquots for the stable isotopes of O and H in water and S and O in dissolved sulfate; and a 2 mL aliquot treated with 20 µL saturated HgCl₂ for stable isotopes of C and O in total dissolved carbonate species. Aliquots analyzed onboard for alkalinity represent an additional acidified sample, as the amount of 0.01 N HCl added is known exactly, so we kept and stored these for additional analysis if needed.

Onshore chemical analyses in Bremen

Water samples

A total of 222 filtered (0.45 µm) and acidifed (10 µL concentrated HNO₃/mL) pore water samples were analyzed by ICP-AES for the 15 elements listed above. Two sets of dilutions were run, the first adjusted for each sample to bring its chlorinity into the range of 40–50 mmol/kg and the second at a constant dilution of 100-fold for Na, K, Mg, Ca, Sr, S, and Si. The first set was run twice, the first time with a cross-flow nebulizer and the second time with an ultrasonic nebulizer. Additional trace elements (Be, V, Cr, Co, Ni, Cu, Zn, As, Mo, Cd, and Pb) were sought in the second run. In all cases, standardization was done against multielement solutions prepared from commercial standards, adjusted to a similar NaCl concentration of the diluted samples. For those elements measured in more than one run, we found that the more concentrated dilution yielded less scattered results. Our ICP-AES data for sulfur average 2.9% higher than our ion chromatography data for sulfate but show similar profiles in all three holes. Given the near-complete absence of H₂S in these pore waters, these measurements suggest that our accuracy for major elements is about ±3%.

Unacidified aliquots of the same samples were analyzed for chlorinity at the University of Hawaii (USA) prior to the OSP by automated electrochemical titration with silver nitrate (±0.53%; 1σ; based on 114 duplicate analyses of samples) using International Association for the Physical Sciences of the Oceans (IAPSO) seawater as a standard and by ion chromatography during the OSP at the University of Bremen (Germany) for chloride (±0.57%; 1σ; based on 7 duplicates of samples and 93 replicates of IAPSO seawater), bromide (±3.5%; 1σ; based on 7 duplicates of samples and 89 replicates of IAPSO seawater), sulfate (±1.6%; 1σ; based on 14 duplicates of samples and 90 replicates of IAPSO seawater), and nitrate (less than the detection limit of 3.2 µM in all samples). For chloride and sulfate, we standardized against dilutions of IAPSO seawater, whereas for bromide and for sulfate at concentrations <5 mM, we used a series of artificial seawater standards with variable amounts of bromide and sulfate. Results for chlorinity by titration were higher than those for chloride plus bromide by ion chromatography by, on average, 0.42% (1σ). We preferred the chlorinity values because they yielded ratios of Br, Na, K, Mg, and Ca to chlorinity in our seawater samples that were closer to the accepted values for seawater. We preferred the ion chromatography sulfate data over the ICP-AES sulfur data for the same reason. For both chloride and sulfate, the two sets of data show identical depth profiles in all holes, and no interpretation was changed by using one set or the other.

We calculated Na concentrations from charge balance and used these values rather than those measured by ICP-AES. The measured values averaged 5.9% higher but were much more scattered, consistent with the precision of this technique of ~5% on 100-fold diluted samples. The calculated values have essentially the same precision as the chloride data, ~0.6%. The two data sets for Na yielded similar depth profiles in each hole.

Sediment samples

About 150 samples of sediment (5 cm³) were freeze-dried and finely ground by hand in an agate mortar. Sediment samples were analyzed for contents of organic carbon, carbonate, and sulfur using a LECO CS-125 carbon-sulfur analyzer at the University of Bremen. Approximately 50 mg of dried, ground sample was weighed in a ceramic cup and heated in a furnace. The evolved CO₂ and SO₂ were then measured with a nondispersive infrared detector. A second aliquot of ~90 mg was weighed in a ceramic cup, reacted with 12.5% HCl twice, washed with deionized water twice, and reanalyzed as above. The CO₂ measured in the second run was assumed to come from organic carbon. The analytical precision is about ±0.02% absolute.

X-ray diffraction

A subset of the same freeze-dried and ground samples was analyzed for mineral content by XRD. XRD measurements were performed at the Crystallography Department of Geosciences, University of Bremen, on a Philips X'Pert Pro X-ray diffractometer equipped with a Cu tube (Kα λ 1.541), a fixed divergence slit (¼°2θ), a 15 sample changer, a secondary monochromator, and the X'Celerator detector system. Measurements were made from 3° to 85°2θ with a calculated step size of 0.016°2θ (see also Vogt, 2009, and Expedition 302 Scientists, 2006, for IODP Expedition 302). The calculated time per step was 100 s. Peak identification was done graphically using the Macintosh program MacDiff (version 4.5), available at servermac.geologie.uni-frankfurt.de/​Staff/​Homepages/​Petschick/​Petschick.html (Petschick et al., 1996).

For mineral identification, integrated intensities for the investigated mineral peaks were calculated by MacDiff. Ratios of these intensities to the total intensity for a given scan were then used to estimate the proportions of the various minerals present. The fraction of the total intensity represented by the 35 peaks quantified for 29 different minerals or mineral groups varied from 90% to 93%. To provide an easy comparison to published data on surface samples from the potential source regions (Andersen et al., 1996; Vogt, 1997; Vogt et al., 2001), the fixed divergence was changed to automatic divergence using an algorithm integrated in MacDiff.

Quality assurance and quality control

Care was taken throughout to obtain uncontaminated sediment and interstitial water for analysis, as documented above. Sampling and analytical procedures that have long been standard practice within ODP and IODP were used throughout.

Top of page   |   Previous   |   Next