Renee Williams, Department of Chemistry and Biochemistry

Introduction

Particle induced X-ray emission (PIXE) accomplishes simultaneous multiple- trace element analysis with samples weighing less than one milligram. Having such rare capabilities, PIXE has applications in many disciplines. Environmental, industrial and medical fields span only a few of the areas where PIXE has been used. Brigham Young University has conducted PIXE research in these disciplines and others for over 25 years. One of the current research projects involves the elemental analysis of lichens from Federal lands in the Intermountain west. During this project, weaknesses in the existing PIXE system became more apparent which sacrificed the precision of the method. Research was conducted to identify and design the modifications necessary to increase the reproducibility of PIXE analysis at BYU.

Experiments with lichen samples in a powdered media showed a precision variance up to 16% in duplicate analyses; an inconsistency much higher than the 5% variance observed in liquid samples. Through experimentation and reading, homogeneity was identified to be the factor controlling precision. The literature further stated the best way to improve homogeneity in powdered media was to increase sample size. Instead of analyzing moderately thick samples of 2-4 mg/cm2, samples of more than 15 mg/cm2 in thickness improved precision of duplicates to 95% and better. The overall sample weight, for the thick target technique, would increase from 1 to 30mg.

To accurately analyze thick samples, several modifications had to be designed and implemented in the existing system. This progress report details the current status of those modifications, and the successes and failures which occurred since the initiation of the project.

Methods and Discussion

Three main concerns had to be addressed to achieve accurate quantitative thick target analysis: monitoring the beam current, prevention of charge buildup, and system re-calibration.

The thin target method allowed the proton beam to pass through the sample into a Faraday cup where it was measured. The current measurement was used to calculate the concentration of elements in the sample matrix. Increasing the sample mass caused the target to be thick enough to stop the beam. Consequently, the beam no longer entered the Faraday cup.

To resolve the problem, a small scatter foil will be placed in the beam line approximately three inches upstream from the target. The foil will scatter protons toward a small detector where the number of protons, being proportional to beam current, will be monitored. An arm holding the foil was fitted to the beam line, but the foil material has not been selected. One possibility is gold; literature has indicated gold to be resistant enough to sustain the beam intensity at the necessary foil thickness of 80 to 150 pg/ cm2.

The second concern arising from the imbedded beam was charge build-up. As more protons entered the sample, a positive charge grew on the target’s surface. The charge was sporadically released in the beam chamber as a continuum of X-rays. The extraneous X-rays entering the detector dramatically increased the background and decreased peak resolution in the spectrum.

Another foil, positioned directly in front of the target, was the solution proposed to eliminate the charging phenomenon. A 0.5 mm thick foil, when in contact with the beam, will spray electrons on the surface of the target and neutralize any positive charge. To properly position the foil, a new beam columnator was designed and machined. The foil material has not been selected yet, but initial studies with a foil in place show charging to be eliminated.

It was originally supposed that the charge neutralizing foil and the proton scatter foil could be combined. After further investigation, the geometry of the beam line was found to be Inadequate to allow the beam to be monitored by the existing detector. Therefore, the scatter foil and a separate detector was determined to be the best alternative to measure the beam current.

Calibration procedures have been initiated but can not be completed until after the two foils have been chosen and Installed. The theoretical basis for calibration changes dramatically between thin and thick targets. In thick targets, the beam decelerates and loses ionizing power according to depth traveled in the sample. A beam energy profile must be calculated by sample thickness to determine each element’s probability of ionization per charge unit of beam. Once an element Is ionized, the X-rays must exit the sample before entering the detector. As X-rays travel through the sample matrix, X-ray Intensity is lost and they can ionize other atoms as well. Intensity loss increases with distance traveled in the target, and must be compensated for by individual element calibration. If an X-ray Ionizes another element as it Is leaving the sample matrix, the signal of the first X-ray is diminished while the signal of the second X-ray is enhanced. Probabilities of such reactions must be calculated and considered in calibration coefficients for each element.

All of these calculations will be done once the foils are in place. Fortunately, a software package entitled GUPIX was purchased recently which will perform many of the complex calculations and has been designed specifically to analyze thick samples.

With the design, machining and Initial testing of the new beam columnator, and the preparations completed for the scatter foil, much progress toward thick target analysis has been accomplished. The necessary modifications to analyze thick targets will greatly improve the quantitative capability of BYU’s PIXE system as well as increase the variety of samples which can be analyzed. I have greatly appreciated the opportunity to work on this project and want to thank Dr. Nolan Mangelson and the Office of Research and Creative Work for the guidance and funding which made the project possible.