Spectral Multi-peak Analysis Methodology for Eliminating Effects of NaI Temperature Drift

Year
2010
Author(s)
Jose March-Leuba - Oak Ridge National Laboratory
Belle R. Upadhyaya - University of Tennessee
Taner Uckan - Oak Ridge National Laboratory
Patrick Brukiewa - The University of Tennessee
Abstract
This paper presents a methodology of spectral multi-peak analysis for eliminating temperature drift effects of NaI scintillation gamma-ray detectors. Properties and cost of NaI make it an excellent choice for low energy (< 1-MeV) spectroscopic studies. However, the light output of NaI scintillators has high temperature dependence, and the common practice of selecting a region of interest (ROI) for the peak of study in the measured spectrum requires that NaI temperature drift be monitored and corrected, typically by gain control of the measurement electronics. Moreover, to determine the peak area of interest, the conventional approach is to define ROIs and draw a line from the left ROI point to the right. Everything under the line is considered “background,” and everything above is related to the peak area. Unfortunately, this commonly used practice may yield inconsistent results because each user may select a different ROI for the same peak. Furthermore, not everything under the ROI line is necessarily background; therefore, valuable spectral information might be missed. In addition, the statistics of the ROI background depend on the statistics of the ROI anchor points, which induce an additional measurement uncertainty, especially when the goal is measurement of small-amplitude peaks in the presence of large backgrounds. The approach taken for the proposed methodology is to fit the spectrum to a model from the known spectral lines of the gamma source under study. The multi-peak method eliminates the effects of temperature drift because the model automatically tracks any temperature-related changes in the peak energy and its width. This technique also provides two main benefits with regard to the measurements: (a) consistency—the user is removed from any subjective decision; and (b) increased accuracy—more gamma ray counts are used to define the count rate, which leads to reduced measurement variance. Details of the methodology will be described and applied to the 235U measurements at different temperatures. This analysis technique can be applied to obtaining the high accuracy enrichment determination required for nuclear material safeguards and material balance verifications in UF6 processing facilities.