Event Reconstruction for a DIRC

where the two protons will be measured and the η ′ will be identified via missing mass technique. The cross section for this process is small, because no nucleon resonance couples strongly to the η ′ p channel, in contrary to η production. The η ′ peak will be on top of a huge multi-pion background. The only chance to identify the η ′ is to measure the four momentum vectors of the two protons with high accuracy. Although their emission angles (θ, φ) are measured with sufficient precision, their kinetic energies are poorly determined by the present setup. A possible option to achieve sufficient resolutions seems to be a DIRC detector (Detection [of] Internally Reflected Cherenkov [light]) [3] in the forward detection system of WASA. Cherenkov detectors are widely used in high energy physics. They can have different applications: particle identification, energy measuring, threshold detector for detecting specific type of particles in a definite energy range.

Our aim is to measure the velocity of protons in the kinetic energy range from T = 400 MeV to T = 1000 MeV with a precision better than ∆β/β = 0.5%

With the Monte Carlo package GEANT3 [4] different radiator materials were studied: LiF, CaF 2 , quartz (fused silica) and plexiglas. All materials have different refractive indices which are smooth functions of the wave length in the visible range (see Fig. 1). Quartz has the largest values for the refraction index n at any given wavelength λ compared to the two other materials. n varies from 1.60 in the UV range to 1.45 in the IR range. The transmittances are rather independent from the wavelength above a certain “threshold” wavelength as is shown in Fig. 2 (data taken from Ref. [5] and [6]). The radiator should have a large refractive index and high transmittance to measure particles with “low” kinetic energy. LiF has higher transmittance in the UV range compared to CaF 2 and quartz. This gives the possibility to build a Cherenkov detector with LiF as radiator and a proper photocathode for detecting photons in the UV range. However, in the visible range, where the transmittance of LiF and quartz are compatible, quartz has the larger refractive index. We therefore omit CaF 2 and LiF in the further studies. [5] and quantum efficiencies of photocathodes (dotted curves) [7][8][9][10] as function of the wave length. The bulk transmittances are given for 10 mm thickness and reflection losses at the surface result in an additional reduction of the transmittance by 10%.

The quantum efficiencies for different photocathodes are also shown in Fig.

2 (from Refs. [7][8][9][10]). CsI and CsTe are used for photon detecting in the UV range. Bi-alkali (KCsSb) and multi-alkali (GaAsP) photocathodes are used for the near UV and the visible light range. These types of photocathodes have larger and broader distributions of the corresponding quantum efficiency function. This could compensate a lower number of emitted photons in this range of wavelengths. In a first step, simulations were performed for a RICH type Cherenkov detector with different radiators and photon detectors. However, a DIRC type detector was found to be favorable due to following reasons: energy losses in the detector, the broad range of velocities to be measured in the proximity to the Cherenkov threshold and the availability of very fast photon detectors. Based on the above discussed properties of the materials we limit ourselves in the simulations of DIRC detectors to fused silica and UV transparent plexiglas as radiator and photomultipliers with bialkali photocathodes. We will now concentrate on the DIRC detector. The design differs from the Babar DIRC [3,11]. It consists of radiator bars with dimensions 2x6x50 cm 3 and a focusing element with a cylindrical mirror (Figs. 3 and4). The dimensions were for fused silica whereas for plexiglas an increased thickness of 4 cm was assumed.The individual modules are arranged in two planes, one above and one below the beam axis. The normal axis to each plane is tilted by 20 • away from the beam axis (see Fig. 3). Emitted Cherenkov photons create an image on a pixelised screen. Each pixel is a bi-alkali photocathode with dimension 5x5 mm

A single event image consists of two arcs, which are the results of the internal reflection from the left and right side of the radiator bar. The shape of the arcs depend on the kinetic energy and particle species. The relative positions of the arcs depend on the horizontal angle of the detected particle track: the larger the horizontal angle, the larger the distance between arc centers [3]. We have simulated Cherenkov images for protons with 600 MeV and 1000 MeV kinetic energy. We applied a left handed coordinate system with the The average number of emitted photons depends on the path through the radiator end the kinetic energy. Fig. 6 shows the d

…(Full text truncated)…