Analysis of Low Element Concentrations in Quartz by Electron Microprobe

Abstract

Although quartz is one of the most abundant minerals in many rock types, it has not been the focus of in situ quantitative chemical analysis by electron microprobe for a long time. This was simply due to its high purity. Since cathodoluminescence observations reveal a great variety of complex structures within quartz, in situ chemical analysis methods like laser ablation inductively coupled plasma mass spectrometry (LA-ICPMS), secondary ion mass spectrometry (SIMS), and electron microprobe (EMP) applied to quartz have received increasing interest from geoscientists. Although the concentrations of many trace elements in quartz are far below the detection limits of an electron microprobe, Al, K, Ti, and Fe are, amongst others, suitable candidates for quantification. The advantage of EMP analysis over other methods is its high spatial resolution combined with high accuracy. Monte Carlo simulations of the elements listed above in a quartz matrix indicate sampling depths of <2.7 μm for 99% of the acquired X-ray photons. Sampling volumes range from 75 to 250 μm3, and depend on excitation energy and defocusing of the electron beam. Unfortunately, beam-induced damage of the quartz lattice limits the use of high beam currents and focused beams. Irradiation induced damage strongly influences the low energy X-ray lines like Al-Kα. The beam sensitivity of the various quartz samples needs to be frequently tested and the analysis protocol must be adapted according to this signal behaviour. To minimise the effect, we propose dividing the measurement of Al into several subsets. Furthermore, exact investigation of the curvature of the background signal is required to avoid systematic errors. Secondary fluorescence of adjacent minerals is an often-neglected problem of trace element analysis by EMP. Joined crystals of pure quartz connected to TiO2, ilmenite (FeTiO3) and sanidine (KAlSi3O8) were used to quantify the effect of secondary fluorescence in quartz. Depending on the location of the disturbing phase, along the “line of sight” of the spectrometer or perpendicular to it, measurable effects above the detection limits can be recognized at distances up to 40 μm for Al, 60 μm for K, 200 μm for Ti and 220 μm for Fe. Depending on the position of the spectrometer relative to an adjacent phase, secondary fluorescence effects vary for Ti and Fe even at larger distances, which has to be taken into account when very low concentrations need to be detected. This effect complicates the application of empirical corrections for secondary fluorescence near phase boundaries. Setting of specific elements on multiple different spectrometers will increase the statistical certainty and can point to secondary fluorescence effects. Using our analysis protocol, detection limits of <10–15 μg g−1 for the elements Al, K, Ti and Fe in quartz can be achieved.