Previous   |    Next

doi:10.2204/iodp.proc.338.102.2014

Appendix A

Evaluation of GRIND and leaching methods for extracting interstitial water as an alternative to the standard squeezing method

An alternate method of pore water extraction from sediment with porosity <40% was used during Expedition 315. This method was initially developed by Cranston (1991) and later used by Wheat et al. (1994) to assess pore fluid composition under conditions where it was impossible to extract a sufficient volume of interstitial water by the standard squeezing method for chemical analysis (Expedition 315 Scientists, 2009a). In this method, an appropriate aliquot of indium standard solution is added to the sediment and ground in a ball mill to dilute the interstitial water. Because the water volume increases, the mixture of interstitial water and added standard solution is easier to squeeze from the sediment-fluid mixture than from the intact sediment. During Expedition 338, Hole C0002H planned to drill through the accretionary prism, which is older than 5.6 Ma (see “Biostratigraphy” in the “Site C0002” chapter [Strasser et al., 2014b]). These sediments were expected to have low porosity and permeability. Thus, the GRIND method would be more suitable to extract interstitial water instead of the standard squeezing method.

Before applying this method, the chemical analyses of interstitial water by the standard squeezing and GRIND methods were compared. The validity of the use of In- and rare earth element (REE)–spiked samples was checked to estimate the dilution of interstitial water by the additional water. A leaching test was also conducted to simplify the extraction procedure of soluble salts from interstitial water. Here, the results of these methods are described and compared.

Experiment 1: extraction methods

Eight different methods of extracting interstitial water from whole rounds are compared:

1. Standard squeezing (Method 1-A)
2. Previously developed GRIND method
• a. Spike with In standard (Method 2-A)
• b. Spike with Lu standard (Method 2-B)
3. Leaching method
a. Dry sample leaching (Method 3-A)
1. Dry in air at 105°C (Method 3-A1)
2. Dry in a vacuum chamber (Method 3-A2)
b. Wet sample leaching (Method 3-B)
1. Adding Milli-Q water (Method 3-B1)
2. Adding 20 μM CH₃COONH₄ solution (Method 3-B2)
3. Adding 100 μM CH₃COONH₄ solution (Method 3-B3)

Standard squeezing (Method 1-A)

The standard squeezing method for extracting interstitial water from whole-round core samples is described in “Geochemistry.”

Previously developed GRIND method

GRIND Method 2-A. A 5 mL aliquot of Milli-Q water spiked with a 500 ppb In standard is added to 40 g of sediment, which is then ground in a ball mill. The sediment-water sample is squeezed using the conventional squeezing method to produce a spiked sample of interstitial fluid.

GRIND Method 2-B. A 5 mL aliquot of Milli-Q water spiked with a 500 ppb Lu standard is added to 40 g of sediment, which is then ground in a ball mill. The sediment-water sample is squeezed using the conventional squeezing method to produce a spiked sample of interstitial fluid.

Leaching method

The sample sediment is suspended in a centrifuge tube to dissolve the soluble salts. The combination of sample treatment and solutions are listed in Table AT1. The following sections describe the processes for each leaching treatment.

Leaching Method 3-A. The sediment is dried in a convection oven at 105°C overnight (same as that for MAD analysis) (Method 3-A1) or dried in a vacuum oven (Method 3-A2). The latter was conducted to evaluate the effect of oxidation during drying in air. A 4 g (weighed accurately) aliquot of the dried sample is powdered and suspended in 40 mL Milli-Q water (weighed or volume measured accurately) in a centrifuge tube by hand shaking.

Leaching Method 3-B. Sediment is suspended without any treatment after being roughly crushed, placed in a centrifuge tube, and added to 40 mL of leachate. Leachates tested are as follows:

1. 40 mL Milli-Q water (Method 3-B1),
2. 40 mL of 20 μM CH₃COONH₄ (ammonium acetate) solution (Method 3-B2), and
3. 40 mL of 100 μM CH₃COONH₄ solution (Method 3-B3).

The centrifuge tube is shaken for 1 min using a tube mixer to mix the sediment and solution well and dissolve the soluble salts. After shaking the mixtures of sediment and solution, the tube is centrifuged at 10,000 rpm for 10 min to separate the solids and solution. After separating the solids and solution by centrifugation, the solution is stored in two plastic bottles until analysis; an appropriate amount of HCl is added to one of the bottles to dissolve redox-sensitive metal ions. If the water content of sediment is ~20%, the interstitial water is diluted 50 times.

Analytical procedures

Chemical analysis of the liquid. Analytical procedures to quantify the dissolved elements follow those routinely applied in the geochemistry laboratory on board the Chikyu (see “Geochemistry”). All the extracted water samples were filtered through a 0.45 μm filter and stored in two plastic bottles; HCl was added to one of those bottles to a concentration of 0.01 M HCl. Just after the sample solution was stored in the plastic bottles, pH was measured using a glass electrode, and alkalinity was determined by titration using the same electrode. Chlorinity was determined by titration. The other major anions (Br^– and SO₄^2–) and cations (Na⁺, K⁺, Ca²⁺, and Mg²⁺) were analyzed by ion chromatography using anion and cation exchange columns, respectively. Nutrients PO₄^3– and NH₄⁺ were analyzed by colorimetry. Minor elements B, Si, Fe, Li, Mn, Ba, and Sr were analyzed by ICP-AES. Trace metals V, Cu, Zn, Rb, Mo, Sc, Pb, and U were analyzed by ICP-MS. Indium (In), Y, and Lu were analyzed by ICP-MS only for the spiked samples and squeezed using Method 1-A as a reference (Table AT2).

Calculation of dilution ratio. The dilution ratio can be obtained by

Densities of the added solution were close to 1, and the volume (mL) was assumed to be the same as weight (g). The water weight in the sediment is determined by

The water content was measured by drying sediment in a convection oven at 105°C in the same manner as water content analyses routinely operated on board for MAD analyses and in a vacuum chamber.

Results

The water content determined by the two different ways of drying was almost the same (Table AT3). Thus, the sample used for the second experiment was dried in a convection oven at 105°C for determination of water content as usual.

The results of chemical analyses are shown in Table AT4 and Figures AF1, AF2, and AF3. The extracted solutions by the GRIND method give, in general, comparable values to those by the standard squeezing method. Among the major elements, chlorinity is in good agreement (within 5% error). Na⁺ and Mg²⁺ concentrations are also within the range of 5% error, which is acceptable for major element analysis. SO₄^2– and Ca²⁺ concentrations of the extracted solution by the GRIND method are larger and Br^– is smaller than those by the standard squeezing method, but SO₄^2–, Ca²⁺, and Br^– concentrations are within the range of 10% error. Among the minor and trace elements, both GRIND and standard squeezing methods give consistent values for NH₄⁺, B, and Sr within 5%–10% error. The concentrations of other elements determined by the GRIND method are not useful when compared to those by the standard squeezing method. The extracting solution in Method 2-A included an In standard and the solution in Method 2-B included Lu. Analytical results of these elements together with Y are listed in Table AT2, which indicates that those elements were almost completely lost from the extracted solutions. This is probably because those elements were adsorbed onto mineral surfaces, especially onto clay minerals, during grinding. Thus, it is clear that the spikes are not a valuable addition to the solution before extraction. As described above, the dilution rate can be calculated from the water volume in sediment and added solution. Thus, it is not necessary to add elemental spikes to solutions prior to the extraction procedure.

Solutions prepared using the leaching method generally showed large variations. The anion concentrations of leaching solutions from dried sediment (Methods 3-A1 and 3-A2) give much higher values than those of the standard squeezed water. Alkaline element concentrations are much higher in the solutions prepared using the dried sediment, except Li. Mg and Sr concentrations of those solutions are much lower, whereas Ba in those solutions are much higher than those from the standard squeezed water. Ca concentration is comparable to that of standard squeezed water. Leaching solution from wet sediment (Methods 3-B1, 3-B2, and 3-B3) generally contained lower concentrations of major anions and cations and much higher concentrations of minor and trace elements. The concentrations of trace elements V, Mo, Cu, Zn, and U, especially, are more than three orders of magnitude higher than those in the standard squeezed water. Such differences must be attributed to the strong adsorbability of those elements onto the mineral surface; those are released into the solution when the water/sediment ratio becomes high. Therefore, the leaching method is not suitable to extract dissolved elements because the solution condition drastically changes from the original interstitial water, which largely modifies the chemical composition.

Conclusion

Indium and REE used as spikes for determining dilution ratios are adsorbed onto mineral surfaces during grinding, so it is better to use water content for the determination of dilution ratios. In the leaching method, leachates drastically changed the solution condition from the original interstitial water, so this method is not suitable for extraction of dissolved elements from sediment.

Experiment 2: comparison of extracting solutions in the GRIND method

As described in “Experiment 1: extraction methods,” the spiked elements were not detected or mostly lost in the extracted water. Thus, the chemical condition of the extraction solution was checked in an additional test. An 80 g sediment sample was ground with 10 g of the following solutions:

1. Milli-Q water (Method 2-C),
2. 100 µM CH₃COONH₄ (ammonium acetate) solution (Method 2-D),
3. Diluted HNO₃ solution with pH adjusted to 3 (Method 2-E), and
4. Diluted HNO₃ solution with pH adjusted to 5 (Method 2-F).

The weights of sediment and solution were twice those used in the previous studies. The extracting water was prepared using Milli-Q water flushed with N₂ for >48 h to remove dissolved oxygen. After addition of the extracting solution, interstitial water was extracted using the standard squeezing method (Method 1-B).

Results

Analytical results are given in Table AT4 and Figures AF4, AF5, and AF6. Chlorinity obtained using the GRIND method with four different extracting solutions (Milli-Q, ammonium acetate solution, and diluted HNO₃ solutions with pH adjusted to 3 and 5) is within 1% of that determined using the standard squeezing method. Alkalinity, SO₄^2–, and PO₄^3– concentrations tend to be higher in the extracted solutions by the GRIND method, and Br^– concentration is lower than in the standard squeezed water. For SO₄^2–, the Milli-Q water extracted solution gave an extremely high concentration compared with that in the other solutions, for which concentrations are comparable (within 5% error) to that in the standard squeezed water. Alkalinity of solution extracted by ammonium acetate was higher than those of the other extracted solutions because of the addition of ammonium ion to increase alkaline component. It is also clear that NH₄⁺ increased in the extracted solution using ammonium acetate. Na⁺, Mg²⁺, and Ca²⁺ concentrations became higher in the solutions extracted using the GRIND method than those using the standard squeezing method (again within 5% error). Among the major cations, K⁺ concentration exceeds the 5% error range and is higher than that of the standard squeezing method. Minor alkaline and alkaline earth elements tend to give higher concentrations in the extracted solutions using the GRIND method than those using the standard squeezing method, except Sr, which has a concentration similar to that using the standard squeezing method. It is notable that the different extracting solutions do not seriously affect the analytical results of those elements.

Minor and trace elements that behave as oxoacid (neutral molecules and anions) increase in the extracted solutions by the GRIND method (Si, V, Mn, Mo, and U; Fig. AF6). Transition metals Fe, Cu, Zn, and Pb tend to increase in the extracted solutions by the GRIND method, and their behavior is complicated compared with the above elements. Among the analyzed transition metals, Mn concentration decreases in the extracted solutions by the GRIND method. In general, minor and trace element concentrations determined by the GRIND method cannot be used as alternatives to those of the standard squeezing method, except Sr and B, which differ slightly (2% error) from those values obtained by the standard squeezing method.

Conclusion

For the GRIND method, Milli-Q shows the best performance for accuracy of IW chemistry as an additional solution (Method 2-C), but only some components allow for use of the GRIND method instead of the standard squeezing method: alkalinity, chlorinity, Li, Na⁺, Mg²⁺, Sr, and B. As an exception, only SO₄^2– concentration using Milli-Q was not consistent with that using the squeezing method. Therefore, for SO₄^2– concentration, in addition to Method 2-C we adopt Method 2-E using HNO₃ solution adjusted to pH 3.

Top of page   |    Previous   |    Next