Quantitative EPMA of Nitrogen in Silicate Glasses

Abstract

Nitrogen is the dominant gas in the Earth's atmosphere and the key to planetary habitability. However, exchange processes of nitrogen between deep Earth reservoirs (crust, mantle and core) and the surface (atmo-, hydro-and biosphere) are still not well understood. Accurate determination of nitrogen at low concentrations is important to place constraints on N partitioning during planetary differentiation and to address its role in the evolution of the early atmosphere. The experimental determination of N partitioning and solubility at high pressure and temperature conditions requires the use of microbeam techniques due to the size of the experimental charges. In addition, silicate glasses at relatively high fO 2 can have N concentration levels as low as a few tens of ppm. Determination of ultra-light elements such as nitrogen by EPMA has consistently been fraught with problems due to its low fluorescence yield, resulting in low count rates and poor peak-to-background ratios [1, 2]. Accordingly, few studies on N in silicate glasses have been carried out by EPMA [3-7]. These studies unfortunately only report very limited information on analytical conditions, detection limits and uncertainties. One study [4] report a practical detection limit of 1500 ppm, limiting the compositional range that can be accurately constrained in samples, while another [7] achieves lower detection limits using a calibration curve approach limiting its application to very simple systems. We developed an analytical protocol on our JEOL JXA-8900 at the University of Minnesota that resolves concentrations in silicate glasses down to a detection limit of 0.04 wt% N with a reasonable analysis time. Analytical errors were usually better than 3% for concentrations > 0.6 wt%.