Paper No. 36
Presentation Time: 9:00 AM-6:00 PM

TRACE ELEMENT ANALYSIS OF DIAMONDS BY ELECTRON MICROPROBE: SI CONTENT


HOLSING, Nicholas A., Department of Geological Sciences and Environmental Studies, Binghamton University, PO Box 6000, Binghamton, NY 13902-6000 and JENKINS, David M., Geological Sciences and Environmental Studies, Binghamton University, Binghamton, NY 13902-6000, nholsin1@binghamton.edu

Diamonds from the Udachnaya Pipe, Russia, are being used as seed material in developing a Si-in-diamond geothermobarometer. Because of the excellent spatial resolution (~1 µm) and non-destructive nature of the electron microprobe, different configurations were considered for analyzing elements at ppm levels. This study explored the possibility of using peak areas instead of background-corrected peak heights. For all analyses of Si, the JEOL JXA-8900M electron microprobe at Binghamton University was run in the qualitative mode by moving a conventional TAP crystal from 75 to 80 mm (x-ray wavelengths of 6.8973 to 7.3571 Å) with a step size of 50 µm (0.0046 Å), yielding Si Kα and Kβ peaks. A linear background was interpolated between 75 to 75.5 mm and 79 to 80 mm. Peak area (AP) was calculated using Philips ProFit 1.0, which solves for intensity, FWHM, asymmetry, and Lorentzian-Gaussian distributions. Dead time of the gas-flow detector was determined by calculating peak area of Si in quartz from C. M. Taylor Corp., with currents (I) ranging from 10 to 200 nA at 15 kV. The y-intercept of dAP/dI vs. I is the ideal slope of AP vs. I if there is no dead time. Comparing the observed peak area to the ideal peak area without dead time yielded a linear dead time of 8.61% per 106 cts·mm·sec-1. A working curve of peak area vs. weight percent Si was constructed using a variety of silicates on the standard block. The analyses were conducted at approximately 200 nA with a variety of dwell times, normalized to dwell time, then corrected for dead time. Peak areas were then corrected for any minor changes in current and multiplied by the ZAF factor averaged from several (n ≥ 3) quantitative analyses on each standard. This produced a linear curve, with r2 = 0.99931. This curve was confirmed with BSM SRM160a Cr-Ni-Mo alloy steel with a Si content of 6,050 ppm. A calibration curve (26,367 cts·mm·sec-1 per wt. % Si, s.e. = 145.11) was then created using graphite as a blank and BSM SRM160a metal. Diamond seeds, crushed and sieved between 100 and 325 mesh, were epoxy mounted and polished to 0.5 µm diamond grit size for characterization by this method. Observed Si content of the seeds ranged up to 653 ppm, and consecutive analyses on the same analysis site are reproducible within the confidence limits.