Characterization of the Hamamatsu 8-inch R14688-100 PMT
1 Introduction
Large-area photomultiplier tubes (PMTs), with sensitive areas larger than 300 cm2, have been used in dozens of neutrino and dark-matter experiments as an effective and cost efficient way of effectively covering large surface areas with photo-sensitive sensors. Detailed understanding of the PMT response is crucial for future experiments selecting sensors, as well as for accurately simulating the PMT response in existing detectors.
Eos is a tonne-scale detector designed to perform demonstration of key technology for future advanced neutrino detector [Anderson:2022lbb] that plan to distinguish Cherenkov and scintillation light. This technology includes spectral sorting of photons , novel types of liquid scintillators , and fast-timing photodetectors [Theia:2019non]. The R14688-100 PMT from Hamamatsu Photonics was primarily selected for use in the Eos experiment due to its excellent timing resolution of around 1 ns FWHM . This can be compared to other PMTs of this size that typically have timing resolutions around XXX to XXX . This fast-timing is an important aspect in Cherenkov and scintillation separation as well as helping to improve vertex reconstruction in large-scale scintillation-based neutrino detectors.
A total of 206 of these PMTs were purchased to be used in Eos. The experiment requires precise characterization of each PMT to verify they each meet the detector standard in a variety of metrics. In this manuscript, we characterize the properties of these 206 PMTs. In detectors the size of Eos and larger, the PMTs most commonly are detecting single photons within a given event window, and thus the single photoelectron (SPE) response is critical to characterize. In addition to the SPE response, we investigate the after-pulsing rate, the dark-rate and how the PMT response changes with magnetic shielding.
2 R14688-100 PMT
The R14688-100 Hamamatsu PMT is 202 mm in diameter with a high quantum efficiency, super-bialkali photocathode and an expected transit time spread (TTS) of around 1 ns full width at half maximum (FWHM) [hamamatsu_datasheet]. The neck and base of the PMTs are water-proof potted by Hamamatsu and a 20 meter water proof SHV cable is connected. A picture of the PMT is shown in Figure 1.
3 Experimental Setups
3.1 SPE Characterization
The experimental setup to measure the SPE response of the PMTs consists of a central UV-transparent, cylindrical acrylic vessel (AV) that is 3 cm tall and 3 cm in diameter. A 90Sr source is placed above the acrylic vessel. The particles enter the acrylic and produce Cherenkov light. A Hamamatsu H11934-200 1-inch square PMT is optically coupled to the acrylic vessel using Eljin Technology EJ-550 optical grease.
Two R14688-100 ‘measurement’ PMTs are placed 20 cm on either side of the AV. The distance is selected to limit the coincidence rate between the trigger PMT and the measurement PMTs to below 5%, which ensures we are primarily detecting SPEs. The PMTs are wrapped in a one layer of magnetic shielding that spans from the base of the PMT (where the cable exits) to the equator. The magnetic shielding protects against the Earth’s magnetic field, which can deflect the photoelectrons as they travel inside of the PMT, affecting the overall collection efficiency. Further discussion of the magnetic shielding is provided in Section 6. A picture of the SPE testing setup is shown in Figure 2.
The signal from the trigger PMT is used to start the data acquisition (DAQ). Custom software reads out a V1742 CAEN waveform digitizer, which digitizes the signals from each of the PMTs. The sampling rate of the digitizer is 5 GHz over 1024 samples and it has 12-bit ADC over a 1 V dynamic range. The waveforms for each channel are saved to a hdf5 files.
4 Data Analysis
The data analysis code processes the hdf5 files output by the DAQ. First, a per-channel baseline is calculated by averaging over a 20 ns pre-trigger window. Then, for each channel,
5 Characterization Results
5.1 Pulses
Probably don’t need this.
5.2 Timing
The TTS and late-pulsing of the PMTs. The dark-rate of several PMTs.
TO-DO: data-taking every hour for twenty four hours and calculating the dark-rate.
5.3 Charge
5.4 TO-DO: Voltage Scan
High voltage sweep for one of the PMTs going into the detector. Steps of 50V from about 1500V to 2300V, for a PMT running around 1900V.
5.5 TO-DO: 1D Scan?
5.6 TO-DO: Afterpulsing
6 Magnetic Shielding Tests
These should probably be performed by comparing the efficiency with and without the magnetic shielding in the assembly. Will need (a) SPE tests comparing the timing and charge with and without the magnetic shielding. (b) LED tests comparing the total amount of light detected with and without the magnetic shielding.