Pulse Shape Fitting for Pulse Shape Discrimination and Light Output Applications

Pulse shape discrimination (PSD) is one of the most important features of organic scintillators because it allows for segregation of photon and neutron interactions [1,2]. PSD methods quantify the difference in photon and neutron pulse shape, more specifically the differences in the fast and slow components of the pulse. These components correspond to the processes that occur during and after an interaction of a particle in the detector [3]. The fast component is caused by excitations in the singlet state decaying exponentially, while the slow component is caused by triplet states recombining and forming the singlet states that decay exponentially [3,4]. The difference between the neutron pulse shape and the photon pulse shape is caused by the recoil particles produced in each interaction. For a neutron interaction on hydrogen, the recoil particle is a proton so a large portion of the singlets recombine and decay in a way that does not produce light and a large amount of the triplets recombine. For a photon interaction, only a small amount of the triplets recombine. Therefore, relative to a photon of equal energy deposition, the fast component is reduced (often called ionization quenching) and the slow component is enhanced in neutrons [1,2]. Commonly used charge integration techniques for PSD use this concept indirectly to discriminate between neutrons and gammas using ratios of the total integral of the pulse to the tail integral [5–8]. It is possible to be more direct by fitting the pulse and using the ratio of the fast component integral to the slow component integral. In this work, digitized waveforms are collected from neutron and photon pulses in a crystalline stilbene detector coupled to a photomultiplier tube (PMT) and are modeled as the response of an RC PMT circuit to a double exponential input signal [9]. The resulting pulse fits are used for a new approach to pulse shape discrimination. The entire pulse is fit, rather than just the decaying edge, while paying special attention to the fast and slow components. Then PSD is performed by taking a ratio of slow component integral versus fast component integral. We compare this approach to the charge integration method that is commonly used.