Protocols and Pitfalls of Electron Microprobe Analysis of Apatite

Abstract

A suite of oriented apatite samples, including the well-known Durango (Mexico) and Wilberforce (Canada) apatites, was used to evaluate optimal analytical conditions for electron microprobe analysis (EMPA) of apatite. X-ray count rates as a function of beam exposure time were acquired for a variety of electron beam settings, in order to select beam conditions yielding stable or predictably varying count rates. However, significant anisotropy in the F count rate due to sample orientation resulted in overestimation of F concentrations on samples analyzed with the crystallographic c-axis parallel to the electron beam, even when careful zero time count rate extrapolation routines were used. Other techniques, such as cryogenic cooling, use of Al and Au conductive coatings and sample motion during analysis were also evaluated. A second suite of synthetic apatites (Schettler et al. 2011; provided by D. Harlov) of known composition was analyzed to confirm accuracy of the analytical procedure. BSE, CL and X-ray images of many of the apatite samples show elemental zoning, which may affect their use as primary microanalysis standards. Evaluation of the results indicates that F (and other elements) in apatite can be analyzed accurately by electron microprobe under any of the following three sets of conditions: (i) apatite crystals where the orientation of the electron beam is perpendicular to the c-axis of the crystal and use of a defocused or rastered beam of ≥ 5 μm diameter; (ii) use of sample motion at a velocity of ≥ 5 μm/s with a defocused or rastered beam of ≥ 5 μm diameter on a large, homogenous apatite grain; and (iii) use of cryogenic cooling of the sample to about -80°C with a rastered or defocused beam and an extrapolated time zero F counting rate. Quantitative results from the oriented apatites allowed calculation of (OH) contents, which were in good agreement with direct IR measurements.