Development and application of inductively coupled plasma-mass spectrometric techniques for the precise measurement of trace elements and lead isotopic compositions in geological materials

Inductively coupled plasma-mass spectrometry (ICP-MS) is a powerful analytical technique, regarded as one of the most successful in atomic spectrometry, and widely used for measurement of a wide range of trace elements in geological materials. In general, this established technique has the merits of high sensitivity, fast analysis capacities, true multi-element abilities, and a very large dynamic linear concentration range. Sample preparation is also relatively simple. This study has systematically investigated analytical performance characteristics of a high resolution (HR)-ICP-MS instrument (Finnigan Mat Element, Germany). This instrument has been found to be more sensitive, with less spectral interferences (due to the pre-defined setting of different resolutions) than quadrupole instruments. It has been found that only a single internal standard is sufficient to compensate for the instrumental drift during an analytical sequence. The detection limits for some trace elements studied were found to be as low as pg g-1 levels in solution. To obtain accurate analytical results using ICP-MS, complete sample dissolution is required. Decomposition techniques for the dissolution of different types of geological materials have been studied. Savillex Teflon beaker HF/HN03 digestion has been found to be sufficient for the complete dissolution of basaltic samples. High pressure HF/HC104 digestion (PicoTrace¬¨vÜ TC-805 digestion system, Bovenden, Germany) can be used for the dissolution of ultramafic rocks such as dunite and peridotite, in which many geologically useful trace elements are at very low abundance, and also granites (which may contain difficult to dissolve zircon) and magnetite-rich samples. It has been found that a lithium tetraborate fusion method allows complete dissolution of different types of geological materials, but this method precludes analysis of volatile elements due to the high fusion temperature (approximately 1000 ¬¨‚àûC). The sodium peroxide sinter method may enable fast determination of Y, Sc and REE in different types of geological materials. However, neither the lithium tetraborate fusion nor sodium peroxide sinter method can be used for the measurement of low abundance geological samples, due to the lack of ultra-pure reagents. In general, the microwave digestion method has been found to be unsuitable for the total decomposition of geological materials. Systematic investigation of chelates between several complexing reagents (citrate, tartrate, EDTA and DTPA) and Zr, Nb, Ta, Hf, Th and U (\high field strength elements\" - HFSE) has shown that DTPA can form stable HFSE complexes in solution. These prevent the hydrolysis of HFSE and hence allow the accurate determination of HFSE in geological materials. A complexing method has been developed which coupled with an appropriate digestion method and ICP-MS can lead to the accurate determination of HFSE at different levels (particularly higher than 1 ¬¨¬µg g-1 of Nb and Ta) in geological reference materials. As a comparison study and a geological application of the ICP-MS techniques developed in this study trace element concentrations of seventeen oceanic basalt glasses from Macquarie Island have been measured by both solution-based HP-ICP-MS & quadrupole ICP-MS and solid sampling-based laser ablation ICP-MS. The majority of elements measured by the three analytical protocols showed discrepancies of <10% and most showed discrepancies of less than 5% with the exception of Tl and Cd (showing unacceptably large variations of more than 20%). The new high-precision data for a wide range of elements in a suite of primitive mid-ocean ridge basalts (MORB)-like suite allow evaluation of the relative incompatibility levels of these elements and comparison of the order evident from the Macquarie Island suite with the widely accepted incompatibility 'order' proposed for mantle melting by Sun and McDonough (1989). Significant discrepancies have been observed. For the most incompatible elements those primitive glasses representing the lowest-degree partial melts show significantly higher Ba K Th and P contents relative to Nb (and thus slightly different orders of incompatibility than that proposed by Sun and McDonough 1989) than glasses representing higher degree partial melts. This is taken to reflect involvement of one or more K- and P-bearing phases in the earliest stages of partial melting which are eliminated early in the melting history. Subsequent partial melts do not reflect these anomalous enrichments. Sulfur is shown to have a bulk Kd=1 during partial melting that generated the Macquarie Island basalt magmas suggesting that a sulfide buffers melt S contents across the melting range represented. Lead is highly incompatible (close to La) compared to Zn and Cu and Cu is close to Ni in being among the most compatible elements in this mantle-melt system. This may indicate that the mantle sulfide is Cu-bearing pentlandite on a Cu-(Ni-)-bearing monosulfide solid solution. Zinc shows a moderate compatibility level close to that of V whereas Mo behaves similarly to La and Sb is similar to Ce both metals being significantly more incompatible than suggested by Sun and McDonough (1989) Analytical methods for the precise measurement of lead isotopic compositions using HR-ICP-MS and the latest quadrupole ICP-MS (HP 4500 plus) have been successfully developed in this study. Based on a series of measurements of geological reference materials and lead-bearing samples the precision (%RSD) values for 208Pb/204Pb 207Pb/204Pb 206Pb/204Pb 208Pb/206Pb and 207Pb/206 Pb ranged from 0.05 to 0.12 for HR-ICP-MS and from 0.06 to 0.15 for quadrupole ICP-MS. By comparison with lead isotope data obtained by thermal ionisation mass spectrometry (TIMS) the accuracy of both methods was found to be generally better than 0.20%. The ICP-MS methods tested here are simple and cost-effective compared to TIMS suggesting greater potential for application to geological problems in the future. Finally lead isotope ratios in base metal sulfide samples collected from the Lady Loretta deposit in northern Australia have been successfully measured by the quadrupole ICP-MS technique. The results are in excellent agreement with the lead isotope data previously measured by TIMS. The wide variation in the lead isotopic compositions from the Lady Loretta deposit suggests multiple sources for lead in the ores."