Low-voltage electron-probe microanalysis of Fe-Si compounds using soft X-rays.

Conventional electron-probe microanalysis has an X-ray analytical spatial resolution on the order of 1-4 µm width/depth. Many of the naturally occurring Fe-Si compounds analyzed in this study are smaller than 1 µm in size, requiring the use of lower accelerating potentials and nonstandard X-ray lines for analysis. Problems with the use of low-energy X-ray lines (soft X-rays) of iron for quantitative analyses are discussed and a review is given of the alternative X-ray lines that may be used for iron at or below 5 keV (i.e., accelerating voltage that allows analysis of areas of interest <1 µm). Problems include increased sensitivity to surface effects for soft X-rays, peak shifts (induced by chemical bonding, differential self-absorption, and/or buildup of carbon contamination), uncertainties in the mass attenuation coefficient for X-ray lines near absorption edges, and issues with spectral resolution and count rates from the available Bragg diffractors. In addition to the results from the traditionally used Fe Lα line, alternative approaches, utilizing Fe Lß, and Fe Ll-η lines, are discussed.

Non-Sequential Spectral Acquisitions and Data Reconstruction to Remove Time-Dependent Effects from X-Ray Spectra.

X-ray spectroscopy techniques may require prolonged exposure of a sample to an electron beam to generate X-rays. With typical spectroscopic methods the measured signal is acquired while varying the independent parameter in a systematic way, for example, stepping a photon detector in a series of energy steps from one end of the range of interest to the other incrementally or varying the energy of the incoming excitation incrementally. This can be a time-consuming process when signal strength is low, and if the sample is affected by prolonged beam exposure it could potentially change the shape and position of the obtained X-ray spectrum. Hydrocarbon contamination is of particular concern in electron beam instruments because of the interaction between the hydrocarbons and the electron beam at the point of impact on the sample surface. The authors suggest that hydrocarbon contamination does affect the generated spectra and present methods for removing this (and other) time-dependent artifacts for low-voltage X-rays. The software is freely available as a compiled Windows executable and as source code (http://geoscience.wisc.edu/geoscience/?p=16417).

Diversity in early crustal evolution: 4100 Ma zircons in the Cathaysia Block of southern China.

Zircons are crucial to understanding the first 500 Myr of crustal evolution of Earth. Very few zircons of this age (>4050 Ma) have been found other than from a ~300 km diameter domain of the Yilgarn Craton, Western Australia. Here we report SIMS U-Pb and O isotope ratios and trace element analyses for two ~4100 Ma detrital zircons from a Paleozoic quartzite at the Longquan area of the Cathaysia Block. One zircon ((207)Pb/(206)Pb age of 4127 ± 4 Ma) shows normal oscillatory zonation and constant oxygen isotope ratios (δ(18)O = 5.8 to 6.0‰). The other zircon grain has a ~4100 Ma magmatic core surrounded by a ~4070 Ma metamorphic mantle. The magmatic core has elevated δ(18)O (7.2 ± 0.2‰), high titanium concentration (53 ± 3.4 ppm) and a positive cerium anomaly, yielding anomalously high calculated oxygen fugacity (FMQ + 5) and a high crystallization temperature (910°C). These results are unique among Hadean zircons and suggest a granitoid source generated from dry remelting of partly oxidizing supracrustal sediments altered by surface waters. The ~4100 Ma dry melting and subsequent ~4070 Ma metamorphism provide new evidence for the diversity of the Earth's earliest crust.