Research On Charged Particle Identification Methods Based On Pulse Shape Analysis

Charged particle detection has been widely used in the field of radiation protection, environment monitor, particle event forecast, element analysis, and so on. Because of the varieties of particles, it is essential to identify these particles, i.e., confirming the mass number A and atomic number Z of particles, in the detection process. A lot of identification methods have been proposed recently, in which Pulse Shape Analysis(PSA) method has been received more attention because of its utilizing all the information contained in the detector output pulse.In this paper, a method of PSA, i.e., Frequency Ratio Analysis(FRA), which brings frequency analysis method into charged particle identification based on the experiment results, was proposed. The experiment platform consisted of a totally depleted Si detector, a charge sensitive preamplifier(CANBERRA Model 2003BT), a waveform digitizer(CAEN V1729A), a high voltage supplier(ORTEC 556) and a fast amplifier(ORTEC 9302). There were two kinds of particles i.e., the alpha particle and the proton particle, measured in the experiment. The alpha particles were produced from a uranium-plating copper slice; while the proton particles were got as a result of neutrons emitting from 252 Cf bombing the polytene. Having finished the measurement, a series of data processing were carried out, consisting of the rejection of useless data, the elimination of baseline shift, the truncation of pulse shape, the normalization and the energy calibration. Then the frequency spectrum of the processed pulse was obtained with the Fourier transform. By analyzing the frequency spectrum, the ratio of the first frequency component to the zero frequency component, is used as the characteristic parameter to discriminate the alpha particle and the proton particle.The experimental results show that the figure of merit(FOM) of the FRA method is higher than 1.7 and the error rate is lower than 1%. Compared with two time-domain methods i.e., the current peak method and the second moment method, and with another three frequency-domain methods, the FRA method gets the best performance.The FRA method proposed in this paper will have a quite wide application prospect in charged particle identification because of its excellent performance, good noise resisting property, and acceptable calculation work.