First Advisor

David Pringle

Second Advisor

Aaron Apawu

Degree Name

Bachelor of Science

Document Type


Date Created



College of Natural and Health Sciences, Chemistry and Biochemistry, Chemistry and Biochemistry Student Work


The research performed was to improve the precision and accuracy of sample measurements made with the Bruker Picofox X-ray fluorescence spectrometer. The initial method of sample preparation using a 10 μL drop spotted onto an acrylic plate and then drying in an oven at 80˚C produced inconsistent results even when processing replicates or the same sample multiples times. Multiple experiments were conducted to determine the effects of different sample preparation conditions on the resulting accuracy. The conditions tested were drop size, plate composition, and drying temperature. For each experiment an internal standard (gallium) was used at concentrations similar to that of the analytes. Trends in XRF signal intensity and concentration measurements were correlated against microscope images of crystal formations to identify why variations in detections were being seen. The results indicate that a sample drop size of 4 μL spotted onto an acrylic plate and dried at 60˚C on a hotplate produce the most precise measurements. This sample preparation was then used to prepare a calibration curve of 10, 30, 60, and 100 ppm copper and iron with a 50 ppm gallium internal standard. A control sample of 63 ppm iron and 76 ppm copper was also tested to compare calculated results to the predicted concentration from the XRF software. The calibration curve allowed for the concentrations of the metals to be calculated using their signal intensities with an error of less than 1% for both iron and copper while the XRF software predictions had 14% and 5% error respectively.

Abstract Format





XRF; spectroscopy; chemical analysis; optimization




24 pages

Rights Statement

Copyright is held by the author.

Included in

Chemistry Commons