Previous   |    Next

doi:10.2204/iodp.proc.340.207.2016

Analytical methods

Pore fluid extractions

Pore fluids were obtained from all sediment cores presented in this report (Table T1). Sampling occurred every 10 m unless the sediment was unsuitable for pore fluid extraction (i.e., because of sands, debris, or other mission-specific rationales). Sampling procedures are described in detail in the “Methods” chapter (Expedition 340 Scientists, 2013b); however, we present a brief description for completeness. A 10 to 15 cm section of whole-round core was removed to begin the squeezing process in the laboratory. Whole-round sections were processed within a nitrogen-filled bag at room temperature and then transferred to a hydraulic press for pore fluid extraction (Manheim, 1966). Following extraction, pore fluids were filtered and subsampled for various dissolved constituents (see the “Methods” chapter [Expedition 340 Scientists, 2013b]).

Pore fluid ICP-OES and ICP-MS methods

As part of this report, we provide some previously published major element data (Ca, Mg, and SO₄^2–) (see the “Expedition 340 summary” and “Methods” chapters [Expedition 340 Scientists, 2013a, 2013b]) along with previously unpublished minor element data (Li, B, Sr, Si, Mn, Rb, and Cs; Table T1). For the minor elements (Li, B, Si, and Mn), samples were diluted to a 0.25:5 ratio with 1% quartz-distilled nitric acid. Strontium was measured twice, once with the major elements at 100-fold dilution and once with the minor elements at 20-fold dilution. Together, these two runs capture the full dynamic range of Sr in these samples with greater precision than either single analysis alone. Minor element standards were matrix-matched to the samples with an artificial salt water solution made to approximate seawater concentrations of Na, Mg, and Cl from ultrapure salts (Sigma Aldrich). Instrumental drift was monitored by repeated runs of the standard curve, but was not found to influence the inductively coupled plasma–optical emissions spectrometer (ICP-OES) results significantly. Individual sample values represent a mean of three replicate analyses for each element in each sample tube. Uncertainties reported here are 1σ uncertainties derived from two sources: (1) the regression uncertainty, calculated using the standard error of the regression, and (2) the “internal” uncertainty calculated from the standard deviation of the three replicate analyses. The reported uncertainties are combined as the square root of the sum of squares. Detection limits reported here are the point at which the regression uncertainty is one-third of the concentration. Above this point, the concentration measured is >3σ above zero.

Cs and Rb samples were run on a Thermo X-Series II inductively coupled plasma–mass spectrometer (ICP-MS). ICP-MS signals can drift significantly over the course of an analytical session; therefore, instrumental drift was corrected by spiking each sample tube with an internal standard solution. This solution consisted of ~50 ppb each of Rh and Re. Measured counts throughout the run were corrected to the value of the first acid blank run of the day. Although Re was not used to correct either of the target elements, it was used as an assessment tool to monitor instrumental drift. Samples were diluted to a 0.25:5 ratio with 1% quartz-distilled nitric acid. Standards were spiked with the same artificial salt water solution as used for the minor element analysis to match the matrix of the samples. Though ultrapure salts were used, we found this solution contained a Cs blank. We determined that the solution contained ~2.5 ± 0.1 nmol/L in the undiluted salt water solution, and this value was used to correct the standard concentrations.

Solid-phase total digestion methods and analyses

Total sediment digestion methods are described in detail elsewhere (Muratli et al., 2012; Muratli et al., 2015). Briefly, prior to digestion sediments were dried and ground. Total sediment digestion was accomplished through a microwave-assisted (CEM MARS-5 microwave oven) digestion process that utilized the inorganic acids HCl, HNO₃, and HF (Muratli et al., 2015). Samples were diluted with 5% HNO₃ and were heated for ~24 h prior to analysis to remove residual fluoride complexes (Muratli et al., 2012).

To assess precision, laboratory standards (PACS-2 and an in-house standard RR9702A-42MC) were digested on multiple occasions, and ~10% of the samples during a run were replicated. RR9702A-42MC is an in-house Chilean margin sediment standard that our group has been using to assess long-term method precision, and PACS-2 is a marine sediment reference material (National Research Council [NRC] Canada) from the Harbour of Esquimalt, Canada.

For the solid-phase major element data (Table T2), we analyzed samples on a Leeman Laboratories Prodigy ICP-OES at the K.W. Keck Collaboratory at Oregon State University (USA), and our approach was similar to that employed for pore fluids as discussed above. Specific emission wavelengths for analysis of solid-phase digests are reported in Table T3. Samples and standards were typically diluted 20-fold with 1% quartz-distilled nitric acid. For elements affected by instrumental drift, a correction was applied. Data values are an average of three replicate analyses for each element. We report uncertainties at 1σ, which are calculated as indicated for the dissolved phases.

Dithionite-extractable Fe and Mn

A single-step mild chemical leach was performed on samples to extract the labile or “reactive” fraction of Fe and Mn from the sediments (Mehra and Jackson, 1960) (Tables T4, T5, T6). We added ~0.25 g of dried ground sediment to centrifuge tubes and then added ~10 mL of dithionite reagent. We used a sodium acetate, sodium citrate solution as a buffer (Kostka and Luther, 1994; Roy et al., 2013). Centrifuge tubes were agitated with a vortex stirrer and placed in a heating block for 4 h at 60˚C. Each sample was mixed every 15 min, and at the end of 4 h samples were centrifuged at 4000 rpm for 5 min. Leachate was then transferred into a labeled bottle, and sample mass was recorded. For some of the samples (Sites U1399 and U1400), a precipitate or gel-like substance precipitated from solution over time. Samples that showed the presence of a precipitate were reextracted and then diluted immediately following extraction.

Analyses for reactive Fe and Mn samples (Sites U1394–U1396) were analyzed on a Leeman Labs Prodigy ICP-OES at Oregon State University (Table T4), and samples from Sites U1399 and U1400 were analyzed on a Perkin Elmer Atomic Absorption Spectrometer, AAnalyst 700, at the University of Akron (Tables T4, T5). Additional samples for Site U1396 were also analyzed at the University of Akron on an Agilient Technologies 700 series ICP-OES (Table T6). For the samples from Sites U1394–U1396 that were measured at Oregon State University, we assessed precision by using the standard reference material, PACS-2 (NRC Canada), as well as our Chilean margin laboratory standard. Reactive Fe in these two standards was measured to be 0.79 ± 0.02 wt% for the PACS-2 standard and 0.93 ± 0.02 wt% for the Chilean margin standard. For reactive Mn, the values are 0.0029 ± 0.0001 wt% and 0.0024 ± 0.0001 wt%, respectively. These values compare favorably with previous runs from our laboratory with the exception that the reactive iron values are slightly lower than the longer term average reported in Roy et al. (2013), where the reactive Fe values for PACS-2 and Chilean margin samples are 0.88 ± 0.08 wt% and 1.08 ± 0.08 wt%, respectively. The values reported for this communication are, however, consistent with those reported in Muratli et al. (2015). To assess precision for the samples from Sites U1399 and U1400 (Tables T4, T5), we used an in-house carbonate-rich laboratory standard, PACS-3 (NRC Canada), and the standard reference material 2702 (National Institute of Standards and Technology), which is a marine sediment from Baltimore Harbor. Reactive Fe for these samples was determined to be 0.71% ± 0.03% (in-house carbonate), 0.85% ± 0.04% (PACS-3), and 3.11% ± 0.08% (2702), and reactive Mn is 0.16% ± 0.05% (in-house carbonate) and 0.15% ± 0.05% (2702). Reactive Mn for PACS-3 was not detectable.

Reproducibility for samples (Table T6) was accessed from standard reference materials PACS-3, MESS-3, and MESS-4 that were analyzed during extractions with unknown samples. Iron, manganese, and aluminum in PACS-3 are 0.89% ± 0.03% (N = 5), 0.0038% ± 0.0008% (N = 5), and 0.121% ± 0.010% (N = 3), respectively. Fe, Mn, and Al for MESS-3 are 1.68% ± 0.05% (N = 5), 0.0182% ± 0.0009% (N = 5), and 0.131% ± 0.001% (N = 3), respectively. Fe, Mn, and Al in MESS-4 are 1.51% ± 0.07% (N = 3), 0.0171% ± 0.0003% (N = 3), and 0.143% ± 0.001% (N = 3), respectively.

Organic carbon, nitrogen, and inorganic carbon

Organic carbon and nitrogen for most of the samples presented here were analyzed at Oregon State University following the methodology outlined in Goñi et al. (2003) with minor modifications (Tables T4, T5). This technique uses 20–30 mg of sample placed into silver boats and exposed to hydrochloric vapors to remove all inorganic carbon (Hedges and Stern, 1984). Complete removal of inorganic carbon involves pipetting one to two drops of 10% HCl into the samples after vapor acidification is complete. Samples are then dried and run on a Thermo Quest EA2500 elemental analyzer. Blanks containing minor amounts of carbon but no nitrogen were analyzed. Organic carbon measurements were corrected for the amount of carbon recorded by the blanks. Inorganic carbon measurements were completed at The University of Akron on a UIC Coulometrics coulometer. The analysis required ~40 mg of sample placed in gelatin capsules. A purge of the sample flask with CO₂-free carrier gas removes atmospheric CO₂. Following the purge, 5 mL of perchloric acid (HClO₄) was added to the flask to acidify the inorganic carbon in order to determine the amount of CO₂ within the sample.

Hole U1396C tephra samples (Table T6) were freeze-dried (24 h), crushed using a granite pestle and mortar, and decarbonated using excess hydrochloric acid (10% v/v) until any visible sign of reaction had ceased. This process was followed by repeated washing with deionized water until a neutral solution was obtained, followed by freeze drying (24 h). Approximately 10 mg of decarbonated and unaltered sample was weighed into a tin cup and then was crushed and analyzed for total organic carbon (TOC) and total carbon (TC) content, respectively, using a Carlo Erba 1110 elemental analyzer with L-cysteine as the calibration standard.

Top of page   |    Previous   |    Next