The Barrel DIRC detector of PANDA

TL;DR

The paper presents the design, implementation, and test results of the Barrel DIRC detector for the PANDA experiment, which provides particle identification capabilities crucial for high-precision QCD studies at FAIR.

Contribution

It introduces the novel Barrel DIRC detector design and demonstrates its performance through test beam results, highlighting its particle separation capabilities.

Findings

Achieves pion-kaon separation of 3 standard deviations up to 3.5 GeV/c

Test beam results validate the detector's photon detection efficiency

Demonstrates effective particle identification in a compact design

Abstract

The PANDA experiment is one of the four large experiments being built at FAIR in Darmstadt. It will use a cooled antiproton beam on a fixed target within the momentum range of 1.5 to 15 GeV/c to address questions of strong QCD, where the coupling constant $\alpha_s \gtrsim 0.3$. The luminosity of up to $2 \cdot 10^{32} cm^{-2}s^{-1}$ and the momentum resolution of the antiproton beam down to \mbox{$\Delta$p/p = 4$\cdot10^{-5}$} allows for high precision spectroscopy, especially for rare reaction processes. Above the production threshold for open charm mesons the production of kaons plays an important role for identifying the reaction. The DIRC principle allows for a compact particle identification for charged particles in a hermetic detector, limited in size by the electromagnetic lead tungstate calorimeter. The Barrel DIRC in the target spectrometer covers polar angles between $22^\circ$ and $140^\circ$ and will achieve a pion-kaon separation of 3 standard deviations up to 3.5 GeV/$c$. Here, results of a test beam are shown for a single radiator bar coupled to a prism with $33^\circ$ opening angle, both made from synthetic fused silica read out with a photon detector array with 768 pixels.

Full text

1 Introduction

The PANDA experiment [1] employs several particle identification (PID) systems for its physics program [2]. Two of them are Cherenkov counters using the DIRC (Detection of Internally Reflected Cherenkov light) principle which allow for building compact PID systems. This is accomplished by using the radiators as light guides for the produced Cherenkov photons to a photon detector outside of the active volume. The first of such detectors was built and successfully operated by the BaBar experiment [3, 4]. The Barrel DIRC described here, separates pions from kaons with at least 3 standard deviations (s.d.) up to momenta of $3.5$ GeV/$c$.

2 Design

The design of the PANDA Barrel DIRC [5, 6] is based on the design of the BABAR DIRC with several improvements. The radius of the 16-sided polygonal barrel is 476 mm.

To eliminate photon background from neutrons hitting the large water tank between the radiators and photon detectors observed by BaBar it was replaced by 16 smaller $30$ cm-deep prisms made from synthetic fused silica. Coupled to each prism are three radiator bars and a photon detector array of eleven lifetime-enhanced Microchannel-Plate (MCP) PMTs [7] on the opposite readout side. Each radiator bar has a thickness 17 mm, a width of 53 mm, and a length of 2400 mm. The DIRC detector is partitioned in 16 independent segments with radiators and an expansion volume with the photon detector that can be dismounted from the radiators. The separation into smaller ”cameras” and the usage of synthetic fused silica was already favored by the SuperB FDIRC [8] and the Belle II TOP [9]. The reduction in size and the finite radiator bar size requires focusing the photons onto the photon detector. Space limitations within the magnetic yoke of the target spectrometer solenoid excludes the usage of focusing mirrors. A three-layer spherical compound lens between radiator bar and expansion prism is used to get a sharp Cherenkov image. The readout electronics is based on the HADES Trigger Readout Board (TRB) [10] and front end amplification and discriminator cards directly mounted on the MCP-PMTs[11].

3 Beam tests

After several test beam campaigns at GSI and CERN in the recent years, the campaign at the CERN PS 2017 served the validation of the baseline design together with slight changes of the setup compared to previous beam times. The prototype consisted of the important elements for one bar box: A fused silica bar (17.1 $\times$ 35.9 $\times$ 1200.0 mm3) was coupled on one end to a flat mirror, on the opposite end to a focusing lens and the fused silica prism as optical expansion volume. Latter had an opening angle of 33∘ and allowed, different from previous beam times [5, 12], for coupling 12 MCP-PMTs with 64 pixels each. Two focusing lenses, a spherical and a cylindrical one, made from LaK33B [13] glass were used for tests. Proton and pion beam particles were tagged by a time-of-flight system. The reconstructed particle type comes from the comparison of the arrival time of the photons at each MCP-PMT pixel and the expected arrival time from simulation (time-based imaging). The log-likelihood difference distribution obtained using time-based imaging for a polar angle of 25∘ and a momentum of 7 GeV/c is shown in Fig. 2 together with Gaussian fits. A $\pi/p$ performance of 4.0 $\pm$ 0.1 s.d. is observed. It deteriorates to 3.0 $\pm$ 0.1 s.d. when using a three-layer cylindrical lens for focusing. Simulation [6, 14] are used to extrapolate the result for the three-layer spherical lens to the expected $\pi/K$ separation power. With the expected photon detection efficiency and timing precision of the MCP-PMTs in the Barrel DIRC, a $\pi/K$ separation power of about 6.6 s.d. at 25∘ and 3.5 GeV/$c$ is achieved. This excellent result allows for studying the performance of the setup with a reduced number or photon detectors in coming experiments to achieve a cost optimization.

Acknowledgments

This work was supported by HGS-HIRe, HIC for FAIR, BNL eRD14. We thank the GSI and CERN staff for the opportunity to use the beam facilities and for their on-site support.