Evaluation of Peak Deconvolution Techniques for Analyzing Am-241 and Am-243 Alpha Spectrum
Part 1: Results for electroplated samples
Beth Cummings1,
Zhichao Lin2, K. G. W. Inn2, and Anna Berne3
1Brigham Young University
2National Institute of
Standards and Technology
3Environmental Measurements
Laboratory
Am-241 is a low-level alpha emitter that contaminates the environment. Accurate determination of Am-241 in a sample requires extensive chemical purification and careful alpha spectrometric measurement. To monitor the amount of Am-241 lost during chemical purification, Am-243, the only other suitable alpha emitter is added to the samples as a tracer in the beginning of analysis. Since the energies of the alpha particles emitted by Am-241 and Am-243 differ by only by 0.2 MeV, they cannot be fully resolved by an alpha detector when the sample spectrum is collected under conditions most typically employed by commercial laboratories. Very often, the Am-241 peak tails under the Am-243 peak and a significant bias may result in the peak area ratio of Am-241 and Am-243, which is used to quantify Am-241 activity in the sample. There are several techniques available in the scientific community that can be used to resolve such a spectrum and make correct assignments for the peak areas. We have evaluated the adequacy of five different peak deconvolution techniques that are currently in use in the Am-241 determination by alpha-spectrometry. The evaluation was performed using spectra collected under a number of different experimental conditions, such as Am-241/Am-243 ratio, detector size, source-to-detector distance, sample thickness, source preparation method, and energy per channel. Because of a large number of variables involved, the concept of Experimental Design (DEX) was applied to minimize the number of experimental runs.
In part 1 of this presentation, the experimental design, and the competence of each technique revealed by statistical analysis of the experimental data will be discussed for electroplated samples.