Developing an XRF Analysis Method for Identifying Trace Metal Contamination in Natural Waters using Rigaku and Whatman Filters

Session: 37th Annual Undergraduate Research Exhibition Sponsored by Sigma Gamma Epsilon (Posters)

Presenting Author:

Authors:

Abstract:

X-ray fluorescence (XRF) spectrometry is an excellent analytical method for determining the chemical composition of different geological samples. Traditionally, XRF is used to analyze solid samples, but it can also be used to analyze water samples to detect trace contaminants. The use of XRF to analyze liquid samples is frequently faster and cheaper than other common water analysis methods. This study aims to develop a quick, low-cost XRF method for identifying trace metal contamination in natural waters.

We used a Rigaku ZSX Primus II to analyze the NIST 1643e,f standards across four different filters: Rigaku Ultra Carry (UC), Rigaku Ultra Carry Light (UCL), Rigaku Mirco Carry (MC), and Whatman Hardened Filter Paper cut into 38 mm diameter disks. The maximum amount of sample was pipetted onto each Rigaku filter (500 microliters for UC and UCL; 100 microliters for the MC), with two Whatman filters receiving the same amounts. Each filter was initially given 24 hours to dry, after which the UC and UCL filters required another 24 hours. A preliminary run was performed using default measuring size (30 mm for the UC and UCL, 20 mm for the MC and Whatman) and time (approximately 3-4 minutes per element) in the ZSX program. Analysis of this run showed that every detected element returned results well above the amount in the standard. Additional background corrections improved the readings; however, most elements still measured at least one order of magnitude higher than the standard. To improve data quality, Fe, Cu, and Pb were chosen for calibration, as they are common metal contaminates in water systems. A second run was performed with additional background corrections, including filter-specific corrections, and the specific measurement ranges for Fe, Cu, and Pb to decrease runtime. Overall, measured values were closer to accepted values. We found that Pb peaks were undetected, despite the presence of Pb in the standard. This suggests that the Pb measuring conditions need to be modified. Future experiments will attempt to maximize accuracy (e.g., adjust counting times, correct peak overlap, measure different lines), enhance efficiency (e.g., use materials appropriate for preparation in the field), and minimize cost (e.g., affordable sample preparation materials) to detect metal contaminants in water systems.