Jean S. Kane

Reference samples for use in analytical geochemistry: their availability, preparation, and appropriate use

Abstract A large number of certified and other reference samples are available for use in analytical geochemistry. Certified materials are preferred, but of much more limited availability than other reference samples for most geochemical applications. The availability of rock, sediment, soil, water, and plant reference samples is outlined; ore and mineral separate reference samples are not included in the discussion. The preparation of these materials, including the establishment of certified or recommended concentrations, is then reviewed. It is shown that comparable quality can be achieved for both certified and recommended concentrations, though it has not always been achieved in the past. Finally, the most appropriate ways to use reference samples in quality control and instrumental calibration are discussed.

Determination of nanogram amounts of bismuth in rocks by atomic absorption spectrometry with electrothermal atomization

Abstract Bismuth concentrations as low as 10 ng g-1 in 100-mg samples of geological materials can be determined by atomic absorption spectrometry with electrothermal atomization. After HF—HClO4 decomposition of the sample, bismuth is extracted as the iodide into methyl isobutyl ketone and is then stripped with ethylenediaminetetraacetic acid into the aqueous phase. Aliquots of this solution are pipetted into the graphite furnace and dried, charred, and atomized in an automated sequence. Atomic absorbance at the Bi 223.1-nm line provides a measure of the amount of bismuth present. Results are presented for 14 U.S. Geological Survey standard rocks.

GeoPT - A Proficiency Test for Geoanalysis†

GeoPT is a new proficiency test for silicate rock analysis. Participants’ results are processed by the formation of a z-score, according to the Harmonised Protocol on Proficiency Testing. Two classes of performance criteria are available for participants to work to. In the second round conducted in 1997, 44 analytes were given z-scores and 13 reported without z-scores. Of the results to which z-scores were assigned, 74.4% were classified as satisfactory.

Multi-element analysis of manganese nodules by atomic absorption spectrometry without chemical separation

Abstract Five manganese nodules, including the USGS reference nodules A-1 and P-1, were analyzed for Co, Cu, Fe, K, Mg, Mn, Na, Ni and Zn without prior chemical separation by using a simultaneous multi-element atomic absorption spectrometer with an air—cetylene flame. The nodules were prepared in three digestion matrices. One of these solutions was measured using sixteen different combinations of burner height and air/acetylene ratios. Results for A-1 and P-1 are compared to recommended values and results for all nodules are compared to those obtained with an inductively coupled plasma. The elements Co, Cu, Fe, K, Mg, Mn, Na, Ni, and Zn are simultaneously determined with a composite recovery for all elements of 100 ± 7%, independent of the digestion matrices, heights in the flame, or flame stoichiometries examined. Individual recoveries for Co, K, and Ni are considerably poorer in two digests than this composite figure, however. The optimum individual recoveries of 100 ± 5% and imprecisions of 1–4%, except for zinc, are obtained when Co, K, Mn, Na and Ni are determined simultaneously in a concentrated digest, and in another analytical sequence, when Cu, Fe, Mg, Mn and Zn are measured simultaneously after dilution. Determination of manganese is equally accurate in the two sequences; its measurement in both assures internal consistency between the two measurement sequences. This approach improves analytical efficiency over that for conventional atomic absorption methods, while minimizing loss of accuracy or precision for individual elements.

Homogeneity of Reference Materials

Homogeneity is that attribute of a sample which assures that analysis of all subsamples taken for measurement will produce the same analytical result, within measurement uncertainty. In reality, all geological samples are heterogeneous to some degree, for some constituent elements. The nugget effect for gold and the platinum group elements is the most obvious example. The fundamental issue is not whether heterogeneity exists, but whether it is detectable, given the number of measurements made, the sample size taken for each, and the inherent repeatability of measurement for the method used to assess heterogeneity. In establishing reference values for reference materials, the contribution of detectable heterogeneity to the overall uncertainty of individual reference values must be quantified.

Review of geochemical reference sample programs since G-1 and W-1 : progress to date and remaining challenges

Abstract A brief history of programs to develop geochemical reference samples and certified reference samples for use in geochemical analysis is presented. While progress has been made since G-1 and W-1 were issued, many challenges remain.

Comparison of several analytical methods for the determination of tin in geochemical samples as a function of tin speciation

Abstract Accurate and precise determinations of tin in geological materials are needed for fundamental studies of tin geochemistry, and for tin prospecting purposes. Achieving the required accuracy is difficult because of the different matrices in which Sn can occur (i.e. sulfides, silicates and cassiterite), and because of the variability of literature values for Sn concentrations in geochemical reference materials. We have evaluated three methods for the analysis of samples for Sn concentration: graphite furnace atomic absorption spectrometry (HGA-AAS) following iodide extraction, inductively coupled plasma atomic emission spectrometry (ICP-OES), and energy-dispersive X-ray fluorescence (EDXRF) spectrometry. Two of these methods (HGA-AAS and ICP-OES) required sample decomposition either by acid digestion or fusion, while the third (EDXRF) was performed directly on the powdered sample. Analytical details of all three methods, their potential errors, and the steps necessary to correct these errors were investigated. Results showed that similar accuracy was achieved from all methods for unmineralized samples, which contain no known Sn-bearing phase. For mineralized samples, which contain Sn-bearing minerals, either cassiterite or stannous sulfides, only EDXRF and fusion ICP-OES methods provided acceptable accuracy. This summary of our study provides information which helps to assure correct interpretation of data bases for underlying geochemical processes, regardless of method of data collection and its inherent limitations.