Photon-veto counters at the outer edge of the endcap calorimeter for the KOTO experiment

Abstract

The Outer-Edge Veto (OEV) counter subsystem for extra-photon detection from the backgrounds for the ${\mathrm{K}}_L^0\to {\mathrm{\pi }}^0\mathrm{\nu }\overline{\mathrm{\nu }}$ decay is located at the outer edge of the endcap CsI calorimeter of the KOTO experiment at J-PARC. The subsystem is composed of 44 counters with different cross-sectional shapes. All counters are made of lead and scintillator plates and read out through wavelength-shifting fibers. In this paper, we discuss the design and performances of the OEV counters under heavy load $(~8\mathrm{tons}/{\mathrm{m}}^2)$ in vacuum. For 1-MeV energy deposit, the average light yield and time resolution are 20.9 photo-electrons and 1.5 ns, respectively. Although no pronounced peak by minimum-ionizing particles is observed in the energy distributions, an energy calibration method with cosmic rays works well in monitoring the gain stability with an accuracy of a few percent.

Introduction

The KOTO experiment is dedicated to observe the ${\mathrm{K}}_L^0\to {\mathrm{\pi }}^0\mathrm{\nu }\overline{\mathrm{\nu }}$ decay using the 30-GeV proton beam at Japan Proton Accelerator Research Complex (J-PARC) [1]. The ${\mathrm{K}}_L^0\to {\mathrm{\pi }}^0\mathrm{\nu }\overline{\mathrm{\nu }}$ decay is a direct CP-violating and flavor-changing neutral current process. The branching ratio (BR) is proportional to the square of the CP-violation parameter $\eta$ in the CKM matrix [2] and is predicted to be $2.43\times {10}^{-11}$ in the Standard Model (SM) [3]. The most attractive feature of this decay is the exceptionally small theoretical uncertainty of the BR of only 2–3%. Therefore measurement of the BR of this decay mode is highly sensitive to the contribution of new physics beyond the SM. The experimental upper limit on the branching ratio was set to be $2.6\times {10}^{-8}$ at 90% confidence level by the E391a experiment at KEK [4]. The KOTO experiment [5], [6], [7], [8] aims to reach a sensitivity below ${10}^{-11}$ by utilizing the high intensity beam at J-PARC and with additional E391a detector upgrades to improve the acceptance and the efficiency for background rejection [6], [7], [8].

In the KOTO experiment, the ${\mathrm{K}}_L^0\to {\mathrm{\pi }}^0\mathrm{\nu }\overline{\mathrm{\nu }}$ decay is identified by detecting two photons from the ${\mathrm{\pi }}^0$ decay with an electromagnetic calorimeter, while the hermetic veto system ensures that no other particles are present. The endcap CsI calorimeter located downstream of the decay region (see Fig. 1a) serves the function of detecting the two photons. All the remaining detectors are photon veto and charged-particle veto counters. The well-collimated ${\mathrm{K}}_L^0$ beam at J-PARC [9] allows us to reconstruct the ${\mathrm{\pi }}^0$ momentum and decay position from the energies and positions of the two photons measured with the CsI calorimeter. The main background source is expected to be the ${\mathrm{K}}_L^0\to {\mathrm{\pi }}^0{\mathrm{\pi }}^0$ decay ($\mathrm{BR}=8.64\times {10}^{-4}$ [10]). Because the decay can mimic a signal candidate if two of the four photons from the ${\mathrm{\pi }}^0{\mathrm{\pi }}^0$ decay are missing, the performance of the photon veto counters and the CsI calorimeter are crucial.

The Outer-Edge Veto (OEV) counter subsystem is one of the photon-veto-counter subsystems and is located around the outer edge of the endcap CsI calorimeter. The 2716 undoped CsI crystals are stacked in a cylindrical support structure of the vacuum vessel. The OEV counters fill the narrow space between the CsI crystals and the cylindrical support structure. The main role of the OEV counters is to reject photons passing through the outer-edge region of the CsI calorimeter before entering the inactive material of the support structure (see Fig. 1b). In particular, photons from the ${\mathrm{K}}_L^0\to {\mathrm{\pi }}^0{\mathrm{\pi }}^0$ decay with an energy around 600 MeV must be detected with high efficiency. In fact the kinematics allows one of the two photons from the ${\mathrm{\pi }}^0$ to hit the barrel detector (MB in Fig. 1a) with an energy about 10 MeV. However this low energy photon may not be detected with a 20% probability due to sampling fluctuations of MB [5]. To keep a short veto time window under high rate condition, the time resolution of a few nanoseconds is required for the OEV counters. In addition, they need to operate stably in vacuum under the heavy load of the CsI crystals.

In consideration of the requirements mentioned above, we adopted a technology based on lead–scintillator sandwich calorimetry with wavelength-shifting (WLS) fiber readout. This type of detector is efficient for photons with an energy higher than 100 MeV [11], [12]. Moreover, it is characterized by a fast response due to the short decay times of plastic scintillator and WLS fiber.

We describe the design of the OEV counters (Section 2) and the construction process (Section 3) in detail. The following topics on the required performance are discussed in Section 4: mechanical robustness of the counters located under the CsI crystals, discharge characteristic of photomultiplier tubes (PMTs) in vacuum conditions, light yield, and time resolution. Finally, we discuss an energy calibration method with cosmic rays as well as its performance and validity (Section 5).