Previous   |   Next

doi:10.2204/iodp.proc.301.202.2008

Analytical methods

Dilution high-resolution inductively coupled plasma–mass spectrometry analysis

Rb, Mo, Cs, Ba, and U were analyzed in samples diluted to 1% in a solution of 10 mL/L that was made from diluting subboiled, concentrated (14.7 N) nitric acid (Optima grade) in 18.2 MΩ milli-Q (mQ) water (1% Q-HNO₃). Standard stock solution was made by combining aliquots of Claritas PPT standards in 1% Q-HNO₃. A Finnigan Element 2 high-resolution inductively coupled plasma–mass spectrometer (HR-ICP-MS), located at Moss Landing Marine Laboratories (MLML) (Moss Landing, California), was used for the analyses. The samples were introduced into the HR-ICP-MS through a CETAC Aridus desolvating nebulizer system fitted with a 100 μL/min nebulizer. Normalization of the data over varying plasma conditions was achieved by addition of internal standards (Rh and Tl).

Standard curves prepared in a 1% dilution of Pacific deep water exhibited negligible differences from those prepared in a salt-free solution. Recovery was within 95% with different matrix conditions and nearly 99% in Pacific deep water. Accuracy of these results was determined by analysis of two sets of previously analyzed samples: Mo and U measured in Baby Bare hydrothermal spring water (Wheat et al., 2002) and Rb, Cs, and Ba measured in pore fluids from ODP Hole 1200F (Shipboard Scientific Party, 2002; Mottl et al., 2004). The average percent error between the two methods for each set of samples ranged between 0.7% for Rb and 8.2% for Mo. Precision was calculated as the standard deviation of Pacific deepwater subsamples that were analyzed periodically throughout a sample run. Detection limits, defined as three times the standard deviation of the blank, were calculated from procedural blanks.

Standard addition HR-ICP-MS analysis

Standard addition was performed on 10% dilutions of three 200 μL aliquots of samples diluted in 1% Q-HNO₃. This method was used to increase the detected counts per second for the transition metals V, Cr, Co, Ni, Cu, and Zn. Standard stock solution was created based on preliminary scans of the samples and diluted to 20% and 50% in 1% Q-HNO₃. A matrix-matched blank was produced by subjecting surface seawater to a column filled with 8-hydroxyquinoline (8-HQ), removing these metals from solution. Rhodium was added to all of the dilutions as an internal standard.

Analysis was conducted at MLML on a Finnigan Element 2 HR-ICP-MS with a 100 μL/min nebulizer fitted into a Teflon perfluoroalkoxyl copolymer resin spray chamber. A series of 10% dilutions of the metal-free seawater blanks were introduced into the HR-ICP-MS for 1 h prior to the start of the series to condition sample inlet cones on the HR-ICP-MS.

Accuracy of the standard addition analysis was determined by comparing the measured values of Pacific deep water (Wheat et al., 2002) for V and NASS-4 (National Research Council of Canada) standard reference material (SRM) for the remaining metals. The accuracy of Cr was the best (1.5% at 2.2 ng/kg), whereas the poorest was Ni (58% at 3.9 ng/kg). Precision was determined by calculating the standard deviation of results from repeated analyses of Pacific deep water. Detection limits were determined by multiplying the standard deviation of the metal-free seawater procedural blank by three.

Extraction HR-ICP-MS analysis

The extraction procedure for the rare earth elements (REEs), Y, and Cd revolved around the 8-HQ functional group immobilized onto TSK Fractogel (Fig. F1) (Landing et al., 1986; Measures et al., 1995; Dierssen et al., 2001). TSK Fractogel AF-Epoxy-600M was used to immobilize the 8-HQ ligand inside a column of Teflon wool within a 5 cm long section of 2 mm inner diameter silicon peristaltic pump tubing. The extraction apparatus, illustrated in Figure F1, sits inside a class-100 laminar flow clean bench to prevent contamination of the samples and reagents during extraction. The apparatus consists of two four-port and one eight-port, two-position injection valves (VICI Valco, Cheminert) controlled by microelectric actuators. Valve switches are controlled by a graphical user interface that dynamically adjusts the parameters controlling sample uptake volumes, column rinse times, and elution volumes. A peristaltic pump placed outside the clean bench pushes the samples and reagents through the apparatus at a constant rate, which is monitored gravimetrically.

Acidified samples were buffered to a pH of 5.3 (Sohrin et al., 1998) by in-line addition of an acetic acid and ammonium hydroxide solution (buffer solution: 84 mL ammonium hydroxide [Optima], 45 mL acetic acid [Optima], and 121 mL 18.2 MΩ mQ water) before the mixture was injected into the 8-HQ sample column. Once 4 mL of sample was introduced, the column was rinsed with 3 mL of 10% buffer solution diluted in mQ water to remove residual salts. After rinsing, the column was purged with air to minimize dilution of the eluate. A 0.5 mL volume of 1.12 N Q-HNO₃ eluant was used to extract elements off the 8-HQ column. Rh and Tl were added for internal standardization to the eluate, and the mass of samples and extracts were recorded. Between each sample extraction, the apparatus was reconditioned by rinsing the column and outflow tubing with 1.12 N Q-HNO₃ for a minimum of 10 min, while the remaining section of the apparatus was rinsed with 0.3 N Q-HNO₃. The column was then rinsed with the dilute buffer solution for 1 min prior to the next sample extraction to restore the pH of the column to 5.3.

Standards of Y, Cd, and REEs were added to filtered, acidified Pacific deep water in order to maintain a salt matrix within the standard solution during extraction. Standards were extracted in the same manner as the samples at the beginning and end of an analysis series, typically around 30–40 samples. Procedural blanks were produced every seventh sample by extracting 4 mL of mQ water into a vial containing an internal standard in the same method used for samples and standards. Analysis of the eluates by HR-ICP-MS was conducted on the Element 2 HR-ICP-MS at MLML through the Aridus desolvating nebulizer with a 100 μL/min nebulizer.

Column recovery within the first 0.5 mL of eluate was generally >97%, ranging between complete recovery of Ho and only 95% recovery of Ce. Accuracy of the method was determined by analyzing a Pacific deepwater sample that had been previously analyzed at Oregon State University (Wheat et al., 2002) and was within the margin of error of the differing analyses. Precision was determined by the standard deviation of Pacific deep water analyzed every seventh sample. Detection limits are calculated by multiplying the standard deviation of daily procedural blanks by three. Additional details of the method are included in Hulme (2005).

Sr isotopic analysis

Strontium isotopic compositions were determined by thermal ionization mass spectrometry (TIMS) at the National Oceanography Centre, Southampton. Strontium was separated from pore fluid samples by standard ion-exchange procedures. Aliquots of pore fluid were evaporated to incipient dryness and redissolved in 0.2 mL of 3 M HNO₃. Strontium was isolated with 80 μL Sr-Spec columns and eluted with 3 M HNO₃. Sr samples were loaded onto outgassed Ta filaments using a Ta activator solution and analyzed in multidynamic mode using a VG sector 54 TIMS. The average value of ⁸⁷Sr/⁸⁶Sr for National Institute of Standards SRM (NIST)-987 on this instrument was 0.710255 ± 0.000026 (2σ) for the period of the analyses (N = 54), and Sr-Spec column blanks were <0.07 ng.

Pore water Ge analysis

Inorganic Ge (germanic acid) was measured using an isotope dilution technique (Mortlock and Froelich, 1996; Hammond et al., 2000) at Oregon State University (OSU) W.M. Keck Collaboratory on a VG ExCell quadrupole ICP-MS. Samples were diluted (0.1–1 mL) and spiked with ⁷⁰Ge to a final sample volume of 20 mL. After sample equilibration, germanic acid is then converted to germane (GeH₄) in a reaction chamber using sodium borohydride, which is transferred using a He carrier to a glass trap submerged in liquid nitrogen. Upon subsequent warming the germane is injected into an ICP-MS. The 7% relative error posted is the average error from running duplicates of samples at varying concentrations; however, because it is difficult to predict, a priori, the Ge concentration of these samples, we use a range of sample/spike volumes. As a result, our variability was >7% (up to 15%) for samples with high concentrations of Ge when the range of sample/spike fluid ratios that were used were not in reasonable proportion to the sample Ge concentration (Mortlock et al., 1993).

Bulk sediment analysis

Analysis of sediment elemental compositions was conducted at OSU. Because Si can be lost via volatilization during HF digestions, sediment samples were digested using a “fusion” technique (Murray et al., 2000). This technique uses ~800 mg of Li metaborate and ~200 mg of sample. Samples are then fused at 1100°C for 15–20 min and then dissolved in 2 N HNO₃. This technique was compared to two other hot acid digestion techniques to ascertain compatibility with other methods (McManus et al., 2006). Results of the comparison were found to be analytically indistinguishable among the three techniques, and reasonable agreement was obtained for SRMs (McManus et al., 2006).

Top of page   |   Previous   |   Next