Rare Earth Elements (REE)

The importance of Rare Earth Elements (REE) geochemistry in the interpretation of petrogenetic processes in mafic-ultramafic rocks is well known. These rocks, which are becoming important means for improving our knowledge on the composition, structure and evolution of the Earth’s upper mantle, are usually characterised by very low REE contents in the few ppm-ppb region.

We set up a SIMS procedure for mafic-ultramafic rock samples, which relies on the empirical approach of calibration curves and the use of international standard rock powders, fused as glass beads. The "energy filtering" is used to eliminate the most part of interfering molecular ions in the secondary-ion mass spectrum, and optimised instrumental conditions are adopted to achieve the best compromise between reduction of interferences and analytical sensitivity. Early measurements involved one isotope of each investigated REE, i.e., 140Ce, 146Nd, 149Sm, 163Dy, 167Er and 174Yb, for which molecular interferences were estimated to be negligible by detecting high-energy secondary ions, and "peak stripping" at masses 151 and 153 (amu) in order to separate Eu from residual BaO interference (Bottazzi et al., 1990). 30Si+ is monitored as the inner standard for the silicate matrix.

Such a procedure made it possible, moreover, to investigate matrix effects, when glasses of both basic and acidic composition are analysed (Bottazzi et al., 1991), and compare our results with those obtained from natural minerals and synthetic (Drake & Weill) glasses. In particular, a systematic study was performed on sputtering/ionisation using natural minerals (amphibole, clinopyroxene, garnet and plagioclase) and their equivalent glasses.

The results have shown that the ionisation efficiency of REE/Si is not significantly affected by the structure of the matrix when high-energy ions are analysed (Bonazzi et al., 1991; Bottazzi et al., 1992; Ottolini et al., 1993).

In case of REE analysis in Ba-rich silicates, or in the presence of light-REE enrichments relative to medium- or heavy-REE, and for high Gd/Yb ratios, special procedures were developed to deconvolute the secondary-ion mass spectrum at those mass numbers. They allow for the removal of the residual interferences (such as BaO+ on Eu+ isotopes; CeO+ and NdO+ on Gd+,… ). In the case of Er, in particular, we have improved the measurement reliability by evaluating the 151Eu16O contribution at mass 167 from the Eu+ signal (at mass 153, corrected for the isotopic abundance) and from the Eu oxidation ratio (i.e., EuO+/Eu+) as determined on suitable standards under the adopted analytical conditions. In the case of Yb, a similar procedure was followed to correct, through the oxidation ratio GdO+/Gd+ and Gd+ ion signal, the current intensity at mass 174 (amu) for the presence of interference due to 158Gd16O+.

A 10% rel. precision is currently achieved for REE concentrations in the range 0.1-0.7 ppm, with an overall integration time of 15 min for all REE (Bottazzi et al., 1994). With less stringent conditions on precision (~ 30% rel.), detection limits fall to 30-40 ppb. Since sensitivity can be further improved, increasing, for instance, the analytical time (counting time) or the primary-beam current intensity, we can estimate operating "detection limits" for REE less than 10 ppb (see Fig. 1).

Fig. 1- C1-chondrite normalised REE concentrations obtained by SIMS in feldspars. Experimental data for Sm are compared with those derived by extrapolation. The agreement is within analytical error (Bottazzi et al., 1994)

Factors as minimum sample consumption, fast preparation and high homogeneity of the fused glass pellets have shown the key role of SIMS relative to traditional techniques as well as its flexibility for both in-situ (on natural minerals) and microbulk analysis (on rock powders fused as glass beads). This approach was used, for example, to update the REE concentrations in an international gabbroic rock standard, GOG-1, which was characterised by means of SIMS (on its main mineral phases and on rock glass pellets), NAA and ICP-AES (Ottolini et al., 1992; Meloni et al., 1993) (Fig. 2).

Fig. 2- C1-chondrite normalised REE patterns for gabbro GOG-1 as determined by different analytical procedures. Literature values are reported for comparison (Ottolini et al., 1992)

 

 

Conventional Energy Filtering (CEF) vs. Specimen Isolation (SI)

Rare Earth Elements (REE) relative ion yields obtained by secondary ion mass spectrometry (SIMS) from amphibole and clinopyroxene samples are reported in MacRae et al. (1993) for the two most commonly described methods of energy filtering, i.e., the "conventional energy filtering" (CEF) technique described by Zinner and Crozaz (1986a), Shimizu et al. (1978) and others, and the "specimen isolation" (SI) technique described by Metson et al. (1983, 1984), MacRae (1987) and others.
We found that both techniques are equally quantitative, with particular advantages for each. Data obtained by SI are virtually free of molecular ion interferences, but the primary beam diameter is large (70-100 µm) and a slight loss of accuracy may occur due to critical beam placement within the unique geometry of the sample holder. Molecular ions of light REEs overlap peaks of heavy REEs in CEF operation, requiring a correction procedure involving determination of MO+/M+ yields and solution of simultaneous equations. The primary beam diameter, however, is approximately one quarter that for SI, an essential requirement for the analysis of small and chemically zoned crystals. Ion yields from crystals do not differ from those for glass of identical composition in CEF, but in SI are approximately 15 per cent lower (see MacRae et al., 1993 for details).

 

References cited

 

Bottazzi P., Ottolini L., Vannucci R.: Quantitative SIMS analysis of rare earth elements in mafic-ultramafic rock samples, VII International Conference on Secondary Ion Mass Spectrometry, Monterey (CA)-U.S.A., Sept. 3-8 (1989), SIMS VII Proceedings, edited by A. Benninghoven, C.A. Evans, K.D. McKeegan, H.A. Storms and H.W. Werner, John Wiley & Sons, Chichester (England), (1990), 413-416.

Bottazzi P., Ottolini L., Vannucci R.: Determination of Rare Earth Elements in Sixteen Silicate Reference Samples by Secondary Ion Mass Spectrometry Using Conventional Energy Filtering Technique, Geostand. Newslet., XV(1), (1991), 51-57.

Bonazzi P., Bottazzi P., Menchetti S., Ottolini L.: REE distribution in piemontite from Mt. Brugiana (Alpi Apuane, Italy) determined by means of EMPA and SIMS techniques, II Europ. Workshop EMAS ’91, 13-17 May (1991), Dubrovnik, abstr., 36.

Bottazzi P., Ottolini L., Vannucci R.: SIMS Analyses of Rare Earth Elements in Natural Minerals and Glasses: An Investigation of Structural Matrix Effects on Ion Yields, Scanning, 14, (1992), 160-168.

Ottolini L., Bottazzi P., Vannucci R.: Quantitative SIMS analyses of REEs in silicate minerals: choice of the standard for natural and glassy matrix, GEOANALYSIS 90 Proceedings, edited by G.E.M. Hall, Geological Survey of Canada, Bulletin 451, (1993), 131-138.

Bottazzi P., Ottolini L., Vannucci R., Zanetti A.: An Accurate Procedure for the Quantification of Rare Earth Elements in Silicates, IX International Conference on Secondary Ion Mass Spectrometry, Yokohama (Japan), Nov. 7-12 (1993), SIMS IX Proceedings, edited by A. Benninghoven, Y. Nihei, R. Shimizu and H.W. Werner, John Wiley & Sons, Chichester (England), (1994), 927-930.

Ottolini L., Bottazzi P., Vannucci R., Oddone M.: An Ion Probe Contribution to Rare Earth Element Investigation of Gabbro GOG-1 Using Secondary Ion Mass Spectrometry, Geostand. Newslet., 16 (1), (1992), 13-19.

Meloni S., Oddone M., Bottazzi P., Ottolini L., Vannucci R.: Neutron activation analysis investigations in updating REE composition of gabbro GOG-1 reference sample: a comparison with SIMS and ICP-AES data, J. Rad. Nucl. Chem., Articles, 168 (1), (1993), 115-123.

MacRae N.D., Bottazzi P., Ottolini L., Vannucci R.: Quantitative REE analysis of silicates by SIMS: Conventional energy filtering vs. specimen isolation mode, Chem. Geol., 103, (1993), 45-54.