Complete dissolution is essential to obtain accurate analytical results using ICP‐MS. In this study, decomposition techniques (i.e. acid digestions using Savillex Teflon vials, a high pressure digestion system and microwave oven, a combined lithium tetraborate fusion ‐ HF/HNO3 acid decomposition and sodium peroxide sinter) for the total dissolution of different types of geological reference materials have been investigated. Savillex Teflon vial HF/HNO3 digestion is effective for basaltic samples. The high pressure HF/HClO4 digestion (PicoTrace TC‐805 digestion system, Bovenden, Germany) allows dissolution of basalts and ironstones. Granites and magnetite‐rich samples can be dissolved using a high pressure HF/H2SO4 method. Geological samples cannot be effectively attacked by microwave acid digestion. A combined lithium tetraborate fusion ‐ HF/HNO3 acid digestion method allows complete dissolution of many different types of geological materials; however, this method precludes the determination of volatile elements due to the high fusion temperature (1000 °C). A sodium peroxide sinter method at 480 °C has the potential for the rapid determination of Y, Sc and REE in different types of geological materials. However, the lack of ultra‐pure reagents precludes the use of lithium tetraborate fusion and sodium peroxide sinter methods for the measurement of geological samples with low trace element abundances.

No abstract

The high sensitivity, minimal oxide formation and single internal standard capability of high resolution inductively coupled plasma‐mass spectrometry (HR‐ICP‐MS) is demonstrated in the direct determination of Sc, Y and REE in the international reference materials: basalts (BCR‐1, BHVO‐1, BIR‐1, DNC‐1), andesite (AGV‐1) andultramafics (UB‐N, PCC‐1 and DTS‐1). Time consuming ion exchange separation or preconcentration were found to be unnecessary. Smooth chondrite normalized plots of the REE in PCC‐1 and DTS‐1 were obtained in the range 0.8‐50 ng g‐1 (0.01‐0.1x chondrite). Method precision was found to be digestion dependent with an average external repeatability of 2‐4% for the basalts, AGV‐1 and UB‐N, and 10% for PCC‐1 and DTS‐1. The mass peak due to 45Sc was completely resolved from 29Si16O and 28Si16O1H spectral interferences using medium resolution, which casts doubt on the accuracy of Sc determinations using quadrupole ICP‐MS. Literature values for Y in rock reference materials were found to be approximately 9% high after HR‐ICP‐MS and XRF analysis.

Major and trace element compositions of fifteen silicate rock reference materials have been determined by a combined XRF and laser ablation ICP‐MS (LA‐ICP‐MS) technique on glasses prepared by fusing the sample with a lithium borate flux (sample:flux = 1:3). Advantages of this technique include the ability to measure major and trace element abundances on a single sample using a quick and simple preparation that attacks resistant phases such as zircon without the need for acid dissolution. The method is suitable for a wide variety of bulk compositions including mafic, intermediate and silicic rocks. Abundance‐normalized mass response patterns (the ratio of signal intensity to element concentration) of the LA‐ICP‐MS analyses vary systematically with major element composition, demonstrating the presence of a matrix effect that cannot be compensated by normalisation to a single internal standard element. Increasing the sampling distance between the ICP‐MS cone and the torch reduces the magnitude of this effect, suggesting that a mechanism related to residence time of ablated particles in the plasma may be at least partially responsible for the observed variations in mass response patterns. When using a matrix‐matched calibration, agreement of the LA‐ICP‐MS results with published reference values or those obtained by solution ICP‐MS is 10% relative. Analytical precision based on replicate analyses is typically 5% RSD. Procedural detection limits that include contributions from gas background and flux are 0.01‐0.1 μg g‐1 for the heavy mass trace elements (Rb‐U). Major element analyses by XRF show excellent agreement with results obtained using a conventional heavy element absorbing flux. High quality major and trace element data for silicate rocks can be achieved by a combined XRF and LA‐ICP‐MS analysis of Li2B4O7/LiBO2 fused glasses provided an appropriate matrix‐matched calibration is adopted.

A range of independently characterised reference materials (RMs) for LA‐ICP‐MS, used for the determination of the platinum‐group elements (PGE) and Au in a sulfide matrix, were analysed and compared: 8b, PGE‐A, NiS‐3, Po727‐T1, Po724‐T and the Lombard meteorite. The newly developed RM NiS‐3 was used as the RM for the calibration of all LA‐ICP‐MS analyses and the measured concentrations of the other RMs compared against their published concentrations. This data were also used to assess the consistency of concentrations calibrated against the different RMs. It was found that Po727‐T1 and 8b produced results that were comparable, within uncertainty, for all elements. Po727‐T1 also produced consistent results with NiS‐3 for all elements. All other RMs showed differences for some elements, especially Ru in Po724‐T, and Os, Ir and Au in PGE‐A. The homogeneity of the PGE and Au in each RM was assessed, by comparing the precision of multiple LA‐ICP‐MS spot analyses with the average uncertainty of the signal. Po724‐T, Po727‐T1 and the Lombard meteorite were found to be homogeneous for all elements, but 8b, PGE‐A and NiS‐3 were heterogeneous for some elements. This is the first direct comparison between a range of independently characterised PGE and Au LA‐ICP‐MS RMs.