Accurate and Precise Analysis of Trace Levels of Cu, Cd, Pb, Zn, Fe and Ni in Sea Water by Isotope Dilution Mass Spectrometry

Abstract

Accurate and precise determinations of natural levels of trace metals in sea water are highly reliant upon the size and variability of the analytical blank, the method for determining the yield, and, to a lesser extent, the inherent precision of the instrument used. Thermal source, isotope dilution mass spectrometry together with ultra-clean room techniques were successfully used in the determinations of Cu, Cd, Pb, Zn, Ni and Fe in sea water. Multi-element analyses were performed in a single experiment owing to the differing release with filament current for each element. A single Re filament loaded with a substrate of silica gel and phosphoric acid gave high precision (0.1% to 0.5%) for the determination of a single ratio, and allowed low detection limits (from 0.02 fmol Cd to 0.07 pmol Fe). Yields were accurately and uniquely determined in the same sample by the addition of two isotopically enriched spikes, one before, and one after an extraction. Blanks were assessed as the summation of individual contributions determined in separate experiments prior to sample analysis. Simply obtaining one value for a total blank contribution can be misleading and can generate larger errors. Rigorous clean room procedures allowed very low blanks; 0.0002 nmol Cd kg⁻¹, 0.0005-0.002 nmol Pb kg⁻¹, 0.02 nmol Cu kg⁻¹, 0.03 nmol Zn kg⁻¹ and 0.7 nmol Fe kg⁻¹.

Sea water, collected from relatively low-lead, coastal waters was stored acidified in rigorously cleaned, conventional polyethylene carboys at pH 2.0, pH 1.6 and at pH 1.1. Strict, clean room conditions allowed accurate comparisons to be made between carboys with varying PH’s of sample pretreatment and to the stability of the concentration with time. Extraction of sea water stored at pH 1.6 to 2.0 by dithizone and by APDC/DDDC in Freon TF yielded values representing the dissolved fraction. Acidification to pH 1.1 yielded higher Pb and Zn results but appeared insufficient for ‘total’ results as defined by aqua regia digestion of sea water. Isotopic composition work on Pb extracted or treated in the above manner supports the definition of soluble and total. The precision in the determination of soluble Cd and Zn with time was ca 1%, a level approaching that attained in measuring a single ratio. The Pb results were less precise, 5%, reflecting the low levels encountered, 0.04 nmol kg⁻¹. Work at higher Pb concentrations indicated a much higher precision was possible. The determination of Cu, Ni and Fe were complicated by isobaric interferences. These could be minimized or eliminated by an ion-exchange purification of the final aqueous extract, and by a stabilization of the ratio with time and with filament current. Analysis of comparable amounts of Cd, Pb and Zn in NBS standard 1643a corroborated the accuracy of the IDMS approach/dithizone extraction procedure. Comparison of the sea water results with FAA and ASV methods indicated that for these metals, the IDMS approach was the most precise and the most accurate.