Comparison of Pulse Shape Discrimination Methods for Phoswich and CsI:Tl Detectors

Abstract

Comparison of four pulse shape discrimination (PSD) methods was performed on the same digitized pulses from two different detector configurations. Beta/gamma discrimination of <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">90</sup> Sr and <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">137</sup> Cs was performed using a phoswich detector made of BC400 and BGO scintillators. Alpha/gamma discrimination of <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">210</sup> Po and <sup xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">137</sup> Cs was performed using a CsI:Tl scintillation crystal. The pulse waveforms were digitized with a DGF-4C and analyzed offline with IGOR Pro software. The four pulse shape discrimination methods that were compared include: constant fraction discrimination (CFD), rise time discrimination (RTD), constant time discrimination (CTD), and charge comparison (CC). The CTD method resulted in a spillover of 9.2% (Figure of Merit, FOM=3.9) and 0.25% (FOM=3.2), while the CC method resulted in a spillover of 9.9% (FOM=3.3) and 0.033% (FOM=3.7) for the phoswich and CsI:Tl detectors, respectively. Although the CTD results in the lowest spillover for high signal-to-noise pulses, the optimized CC method, as implemented here, appears to be the best PSD method overall. Analysis of the reciprocal of the pulse shape data typically resulted in a significantly higher FOM than conventional methods with no reduction in spillover. This simple mathematical transformation of the pulse shape data illustrates that the FOM may not be a good scheme for the quantification of PSD