Particle-γ-Separation with CALIFA

The basic properties of a good scintillator material for low and medium energy γ-ray detection are a high and linear light output with energy. In Thallium-doped Cesium Iodine (CsI(Tl)) both requirements are nicely fulfilled. The scintillation is based on two different scintillating states with two significantly different lifetimes of (0.6 and 3.25 μs). It is known, that the ratio of light output from the two main components is dependent on the ionization density of the absorbed particle ([SJW58]) and therefore dependent on the type of particle. This feature will be used in the CALIFA calorimeter for particle-γ-separation. We have developed a new sensitive algorithm based on digital pulse shape analysis of the preamplifier signals from the photo sensors, called RPID (Reconstructive Particle IDentification) algorithm. Including the so called Moving Window Deconvolution (MWD) [GG93] this algorithm directly reconstructs the amplitudes of the two different time components in the light output. For a test of this new method at the energies relevant for CALIFA we have performed an experiment at the Munich Tandem accelerator (Maier-Leibnitz laboratory, Garching). Using a C(p, p′)12C reaction with a proton energy of Ekin = 21MeV . This allows to measure high energetic scattered protons as well as γ-rays from inelastic scattering with energies up to 15.1MeV at the same time. A 75 mg cm2 Carbon target and three 130 mm long crystals with APD readout at an angle of θ = 60 with respect to the beam axis were used. We recorded digital signal shapes of 10μs length and a granularity of 10ns, triggering on any energy deposit in the crystal. Figure 1 shows the result of the RPID for the full unbiased set of the experimental data. The γ-rays and the protons are separated accurately also for low energies. The signal width for elastically and inelastically scattered protons is dominated by the target thickness and the energy loss of particles in the crystals wrapping. The black line is fitted to the proton line and based on this fit the red line is calculated as the function of elastic scattered protons that have not been stopped completely in one crystal. So it should even be possible to determine the full energy of particles that lost part of their energy in the non active material between the crystals. An energy spectrum of all events identified as γ-rays from the proton-carbon reaction is shown in figure 2. While the 4.4MeV photo peak and its single and double escape peak can be identified, the 15.1MeV peak only represents an edge like distribution. Simulations of the experiment show that positrons and electrons that are mainly generated in the first interaction emit bremsstrahlung pho-