Diamond pixel detector for beam profile monitoring in COMET experiment at J-PARC

We present the design and initial prototype results of a pixellized proton beam profile monitor for the COMET experiment at J-PARC. The goal of COMET is to look for charged lepton flavor violation by direct muon to electron conversion at a sensitivity of $0^{-19}$. An 8 GeV proton beam pulsed at 100 ns with $10^{10}$ protons/s will be used to create muons through pion production and decay. In the final experiment, the proton flux will be raised to $10^{14}$ protons/sec to increase the sensitivity. These requirements of harsh radiation tolerance and fast readout make diamond a good choice for constructing a beam profile monitor in COMET. We present first results of the characterization of single crystal diamond (scCVD) sourced from a new company, 2a systems Singapore. Our measurements indicate excellent charge collection efficiency and high carrier mobility down to cryogenic temperatures.


Motivation and test setup
Diamond has emerged as an attractive alternative to silicon for detection of ionizing radiation. Properties that make diamond a superior sensor material are summarized in Table 1. Diamond detector structures require no dopants and can be operated at higher fields for faster charge collection. By far the most attractive property of diamond is its high radiation tolerance. We performed two types of tests to determine the signal characteristics in diamond: 1. We measure the current induced as a function of time by a stopping α particle with the transient current technique (TCT) [1]. This is done as a function of temperature down to 4 K.
2. Charge collection efficiency of signal deposited by minimum ionizing β particles is measured as a function of the bias voltage.  Figure 1a shows the setup used to perform TCT measurements. Bias voltage (with a current limiting MΩ resistor) is provided as pull-up in the signal path. The setup is arranged so that α particles enter from the opposite, grounded side of the sample. The penetration depth of α in diamond is ≈ 10µm. The profile of current induced by primary e-h pairs as they traverse the bulk can be measured as a function of time.
The readout scheme for measuring the charge collection efficiency shown in Figure 1b is similar. A collimated 90 Sr β source is aligned with the diamond and a trigger silicon diode. A charge shaping amplifier integrates the deposited charge and measures dE/dx in the sample.  Table 2 shows results of the TCT current profile measurements as a function of bias voltage and temperature. Figure 2 highlights the results for electrons at room temperature.

Results: current profile
The signal rises fast due to the plasma of primary e -h pairs created by the α, followed by a flat phase during which charge carriers drift in the uniform electric field, inducing a constant current in the signal electrodes. There is an exponentially decaying tail when the charge carriers are collected at the electrode. At lower temperatures, there is a pronounced slope in the drift phase of the signal indicating the presence of space charge. High carrier mobility and carrier recombination in the charge cloud at low temperature [2] reduces the pulse duration and amplitude.

Comparison with TCAD calculations
We have performed TCAD calculations in the Sentaurus tool suite to cross-check our experimental measurements. A simplified geometry shown in Figure 3a with 300µm diamond and two dimensional calculation was used to improve convergence time. With a space charge of 3 × 10 11 cm −3 specified in the input, the results obtained are shown in Figure 3b. The TCAD calculation results are in qualitative agreement with the trend of the TCT measurements in Table 2 but the signal magnitude is significantly smaller. There are a few possible reasons for the discrepancy: a) in simulation, the ionizing α enters into the crystal at the electrode to which high voltage is applied; b) the simplified two dimensional geometry and c) the space charge setting in the calculation. Reason (a) is most likely since we have found that in our experimental setup (Figure 1a) reversing the direction of the α entry reduces the signal magnitude. Figure 4a shows a characteristic Landau distribution of signals recorded from passage of minimum ionizing β particles through the diamond. We have shown a comparison of signals recorded in the 2A sample and the E6 sample in the same setup.

Results: MIP signal distribution, Charge collection efficiency
Using the benchmark figure of ρ MIP = 36 eh pairs/µm in diamond and sample thickness d = 460µm for 2A (d = 538µm for E6) the Charge collection efficiency (CCE) is calculated as the ratio of the measured charge Q mp to the total charge deposited in the sample: This also gives a measure of the charge collection distance δ c = ε Q × d to be 428 µm in 2A and 509 µm in the E6 sample.
Our measurements of ε Q as a function of the bias voltage shown in Figure 4b indicate that the sample reaches close to maximum CCE already at 0.2V/µm bias.

COMET: Coherent Muon to Electron detection experiment at J-PARC
The COMET experiment at J-PARC [3] will look for physics beyond the standard model through charged lepton flavor violation (CLFV). Muons are captured in an aluminum target to create muonic atoms. After waiting for ≈ 1.2µs which allows for natural muon to electron decay, the experiment looks for the emission of an electron with kinetic energy exactly equal to the rest mass of the muon minus its (small) binding energy in the aluminum atom. This would indicate a coherent conversion of the muon to electron − a CLFV process not allowed by the standard model. The expected signal stands out at the end of the continuum as shown in Figure 5a. The sensitivity of the experiment is determined by the muon statistics. Hence a beam profile monitor is necessary to ensure the purity and time structure of the beam.

Radiation and timing requirements
The beam structure is indicated in Figure 5b. At startup, we expect to have an 8 GeV beam of intensity 10 10 protons/sec. In stable operation, the beam intensity is ramped up to 10 14 protons/sec to increase muon flux and improve the sensitivity of the experiment.
The primary challenge in monitoring the beam profile is high proton flux and the necessity to ensure that the beam is completely extinguished in 100 ns. Any stray protons entering the beamline outside the 100ns bunch would contribute to background during the detection period highlighted in red in Figure 5b. The combination of high radiation dose and fast timing make diamond an ideal choice for this application.

Conclusions
We have measured signal properties and charge collection efficiency in scCVD diamond from a new vendor 2A SYSTEMS. The TCT current response is excellent at 5 ns signal fwhm with no indication of space charge at room temperature, i.e. absence of any charge trapping centers. The charge collection efficiency in a our 460 µm sample at 1 V /µm bias is 92.9%. Both these properties combined with the high radiation tolerance (to be studied) make this scCVD diamond an excellent choice to construct a beam profile monitor for the COMET experiment at J-PARC.