TECHNIQUES FOR ELEMENTAL QUANTIFICATION OF LASER ABLATION INDUCTIVELY COUPLED PLASMA MASS SPECTROMETRY (LA-ICP-MS). RESULTS FOR MELT AND FLUID INCLUSIONS - PROBLEMS AND SOLUTIONS

Multielement data on single melt and fluid inclusions is revolutionizing the understanding of metal transport and deposition and is essential for the development of improved assessment models for mineral resources. Inclusion analysis is one of the most challenging applications of LA-ICP-MS due to: (1) lack of appropriate matrix matched reference materials, (2) difficulties with the ablation of a fluid in a mineral matrix resulting in large signal intensity mismatch, (3) corrections for host mineral contributions, and (4) internal standardization. Our efforts focus on the optimization of LA-ICP-MS for quantitative analysis of single inclusions and address each issue with an emphasis on calibration. Results obtained using multiple standards (geological glasses, e.g. BCR-2G, NIST 612, GSD-1G, glass and quartz capillaries filled with aqueous solutions, and frozen solutions) highlight the advantages and drawbacks of each method.

The LA-ICP-MS system is composed of a petrographic microscope with 193 nm eximer laser (20 ns pulse width) capable of spot sizes from 5 µm to 160 µm providing energy densities of < 1 J/cm² to 40 J/cm². The ablated material is detected using a quadrupole ICP-MS operated in a low positive extraction lens mode (equipped with a collision cell) that allows low detection limits combined with low backgrounds for most ions. Internal standardization for fluid inclusion analysis was performed by using salinity measurement obtained by microthermometry and/or by Raman microspectroscopy and using electron microprobe Al data for glassy melt inclusion or whole rock magma fractionation trends for crystalline inclusions. Detection limits based on a 24 µm spot of BCR-2G for melt inclusions are on the order of 10 µg/g for low masses (Mg), 1 µg/g for mid-mass region (Sr, Mo) and 0.5 µg/g for heavy masses, (Tl, U). Calculation of limits of detection for fluid inclusions is problematic and will be discussed in this presentation. A discussion of analytical uncertainty and comparison of calibration and quantification strategies will also be presented. While our detection limits for most elements meet the needs for petrogenetic and ore genesis research, we find it necessary to continually tailor our methods for the mineral matrix, inclusions, and elements of interest for each study.