by Stephen M. Seddio¹ and Sarah N. Valencia^2,3

1. Thermo Fisher Scientific, Fitchburg, Wisconsin, USA.
2. University of Maryland, Department of Astronomy, College Park, Maryland 20742, USA. 
3. NASA Goddard Space Flight Center, Planetary Geology, Geophysics, and Geochemistry Laboratory, Greenbelt, Maryland 20771, USA.

Introduction

The electron microprobe has long been the stalwart tool of mineralogical and petrographic investigation using WDS (Wavelength-Dispersive Spectroscopy), electron imaging, and EDS (Energy-Dispersive Spectroscopy). EPMA (Electron-Probe Microanalysis) is a particularly useful analytical technique for the study of planetary and geological materials given its broad analytical range (i.e., B – Pu) and non-destructive nature.

Scanning electron microscopes (SEM) are much more common, owing to their relatively low cost, but are typically associated only with imaging and perhaps some microanalysis using EDS which may lack some of the analytical rigor typically associated with WDS. However, modern SEM-based X-ray microanalysis systems are able to achieve EPMA-like analysis using standards-based EDS. Additionally, modern X-ray microanalysis systems enable WDS analysis to be performed on an SEM. Here, we compare quantitative WDS analyses of pigeonite, ferroaugite, and pyroxferroite in lunar meteorite NWA (Northwest Africa) 2727 acquired using an electron microprobe and an SEM.

Experimental Method

A thin section of NWA 2727 was analysed by EPMA using the 5-spectrometer JEOL JXA-8200 Superprobe at Washington University in St. Louis, USA. A 15 kV accelerating voltage was used. The beam current was set to 25 nA and measured at the beginning of each acquisition with an in-column Faraday cup. Na, Mg, Al, and Si were analyzed using a TAP diffractor. Ca was analyzed using a PET diffractor. Ti, Cr, Mn, and Fe were analysed using an LiF diffractor. The 5 spectrometers were used to count multiple elements concurrently with an average acquisition time of approximateky 120 seconds.

The same thin section was then analysed using a Thermo Scientific MagnaRay parallel beam WDS spectrometer mounted on a field emission SEM operating at 15 kV accelerating voltage. The beam current was set to 25 nA and measured at the beginning of each acquisition with an in-column Faraday cup. Mg, Al, and Si were analysed using a TAP diffractor. Ca was analysed using a PET diffractor. Ti, Cr, Mn, and Fe were analysed using an LiF diffractor. On-peak count times for each element was determined by counting until a 1% error from counting statistics resulting in an average acquisition time of approximately 250 seconds using a single WDS spectrometer measuring each element sequentially.

In both instances, natural and synthetic mineral standards were used for calibrations.

Results and Discussion

Pyroxene analyses acquired using the electron microprobe (red) and the SEM (blue) are plotted on a pyroxene quadrilateral in Fig. 1. Both instruments yield similar results for pigeonite, ferroaugite, and pyroxferroite compositions. Average pyroxene compositions obtained using each instrument are in Table 1.

Although WDS systems on SEMs currently do not fully replicate the analytical capabilities that exist for electron microprobes (e.g., interference corrections and time-dependent intensity corrections), the data produced in this study indicates that the two instruments can yield similar quantitative results. A more practical approach for both electron microprobe and SEM-based X-ray microanalysis would be to rely on EDS standards-based analysis for many of the elements and analyse trace or minor elements by WDS.

Table 1. Average Pyroxene compositions from NWA 2727.

Conclusion

In this study we directly compared data generated by an EPMA and an SEM equipped with WDS. The quantitative data produced by the SEM/WDS system compares very favourably with the EPMA. While EPMA is a standardised procedure for characterising geological and planetary materials, SEM/WDS has been shown to produce equivalent data at a fraction of the cost. Furthermore, with the large number of SEMs now installed, the addition of a WDS detector like the Thermo MagnaRay could be an extremely cost effective alternative to a dedicated EPMA system

References

[1] North, S. N. et al., LPS XLIV (2013), Abstract #3013.