Abstract
Experimental and computational results of a ΔE/ΔE diamond detection system are presented. The ΔE/ΔE detection system was evaluated using energetic proton and iron beams striking thick polyethylene targets at the NASA Space Radiation Laboratory (NSRL) at Brookhaven National Laboratory (BNL). The measured data for diamond sensor A show good agreement with the Geant4 simulation. In addition, simulations have demonstrated the ability to identify hydrogen isotopes using a diamond detection system.
1. Introduction
Diamond is an attractive material because of its many electrical, chemical, and mechanical characteristics [1–4]. Its consistent application in radiation sensing occurred after synthetic chemical vapor deposition (CVD) growth of diamond was demonstrated to produce consistent and suitable properties [5, 6]. Since then, diamond detectors have been explored for use in extreme environments [7–11], microdosimetry [12, 13], and thermal [10, 14, 15] and fast neutron sensing [10, 16–20], including deuterium-deuterium (DD) and deuterium-tritium (DT) fusion plasma diagnostics [21, 22]. One extreme environment under consideration for diamond detectors is high-rate particle tracking sensors at the LHC [7, 23–26]. Its large displacement energy (43 eV) results in a high radiation hardness [27], and its 5.5 eV band gap enables operation at room temperature without the need of heat sinks [1]. Its high saturation velocity (mobility) of both holes and electrons yields an excellent timing resolution [5, 28, 29]. These essential attributes make diamond sensors a potential candidate for current silicon detector replacement technology in high-energy physics tracking experiments [30–32].
In this paper, we present results on a ∆E/∆E diamond detection system for secondary particles produced by proton and iron energetic beams striking thick polyethylene targets. Geant4 simulations are used to benchmark and understand experimental observations, with a focus on particle identification using time of flight (ToF) [25, 33–36].
2. Experimental Setup
A schematic view of the ∆E/∆E diamond detection system is shown in Figure 1. Two parallel electronic-grade single-crystal CVD diamonds (diamonds A and B) were separated by 2 cm, where the top sensor was diamond A. Both diamond detectors are 4 mm × 4 mm × 0.45 mm in dimension. Each detector was cleaned using aqua regia followed by chromium etchant. After that, sensors were metalized with an Au/Cr contact scheme via RF magnetron sputtering. Following metallization, each detector was annealed in 10 mTorr of argon gas for 20 minutes. Finally, each diamond detector was mounted onto a printed circuit board (PCB), both mounted within an aluminum enclosure containing two CR-110 charge-sensitive preamplifiers.
The analog signal from each detector, after preamplification, was divided into two pulses using an NSRL splitter. The first pulse was delayed before being fed into the digital acquisition system. The other pulse from each detector was fed into a constant fraction discriminator (CFD), and each output was fed into a fast-coincident module to trigger true minimum ionizing particle (MIP) events for collection by the digital acquisition system. Data were analyzed online through a modified acquisition framework [36]. A schematic diagram of the electronics is shown in Figure 2.
The experiment was conducted at the NASA Space Radiation Laboratory (NSRL) at Brookhaven National Laboratory (BNL) in December 2016. The experimental arrangement is shown in Figure 3. Two thick polyethylene targets were situated along the beamline. The upstream target had a thickness of 20, 40, or 60 g/cm2, while the thickness of the downstream target was 60 g/cm2. These targets were separated by 3.5 meters down the beam. The high-energy beams striking the target were 1 cm in diameter, consisting of either protons (400 MeV, 800 MeV, and 2.5 GeV) or iron (400 AMeV). Two detector positions were selected based on beam species and energy. In the first position, the detection system was located along the beam axis (0°) and behind the backstream target for 800 MeV and 2.5 GeV protons since they penetrate through both targets. In the second position, the detection system was placed 45° off the beam axis to measure scattered protons/iron and secondary particles at 400 AMeV.
3. Geant4 Simulations
The simulated interaction of the proton and iron beams with polyethylene targets and production of secondary ions were implemented using Geant4 [35]. The Bertini intranuclear cascade with a high precision neutron (lower than 20 MeV) model was used to simulate energetic ion interaction for all projectiles [37]. The NSRL room geometry was modeled including targets, concrete floor, walls [38], and the ∆E/∆E diamond detection system. The beamline rail system and other equipment within the room were not included in the simulation. The incident beam (or particle) was located 70 cm upstream from the front target and was simulated as a monoenergetic and monodirectional point source. The simulation results were normalized and plotted using the same bins width of the measured response of the diamond detection system. The CFD threshold was set to prevent low-energy false coincidences, which was accounted for in the Geant4 simulation to provide adequate agreement with experimental spectra.
4. Results
The simulated and experimental energy spectra for the 800 MeV proton beam are shown in Figure 4. The response of diamond sensor A agrees well with computational results. However, diamond sensor B exhibits significantly less pulse height for the primary peak and an overall different shape, where simulations indicate that the signal in the two detectors should be nearly identical. The detection system operates in coincidence between diamond detectors A and B, and both diamonds recorded the same number of events. We further investigated the observed difference in the spectra by changing the orientation of the detection system with respect to the beam vector, and the same result was observed, indicating that the observed differences are not due to secondary particle generation in the front diamond detector striking the back diamond detector. We hypothesize that the observed differences are due to poor metal contact-diamond interface properties along with natural differences in gain between the two preamplifiers.
(a)
(b)
(c)
(d)
The response of diamond detector A, compared with the simulated results with different upstream target thicknesses, is provided in Figure 5. Geant4 simulations observed proton, pion, positron, and electron interactions for all simulations. The pions, positrons, and electrons manifest themselves in the lowest energy region of the energy spectrum (on the order of tens of keV), while the protons comprise the primary peak(s), which is responsible for >90% of the observed counts (Figure 6). The measured spectra at 800 MeV show a few peaks due to the different thicknesses of the upstream target, which is due to the higher rates of energy deposition from slower moving ions exiting the thicker targets [39]. For the 2.5 GeV proton beam, simulations indicate that the energy deposition is almost constant for all thicknesses since the secondary particles are located on the minimum stopping power region in the proton stopping power graph [40]. Nevertheless, the measured spectra exhibit two peaks versus one peak in the simulation (Figure 5(a)), which is likely a consequence of the simple geometry simulated and slight variation in the location of a diamond sensor.
(a)
(b)
(c)
(d)
(e)
(f)
(a)
(b)
(c)
For 400 MeV proton beam, the detection system was located at position 2, resulting in significantly lower observed counts, requiring deactivation of the coincidence mode. The simulations indicate that fast neutron-induced alpha particles and scattered carbon nuclei are created, contributing mainly in the lowest energy region of the spectrum, as shown in Figure 6(c). However, the small fast neutron cross section with carbon resulted in a minute contribution to the overall observed response [41]. In addition, protons, pions, positrons, and electrons were observed from proton interactions with the polyethylene target (Figure 5(f)). The number of detected pions and electrons is higher for the 400 MeV proton beam than for the other proton beams because it deposited a higher energy within the polyethylene target and consequently liberated more secondary particles than 800 MeV and 2.5 GeV proton projectiles.
From the Geant4 simulation of the 400 AMeV iron beam on the 20 g/cm2 upstream target, isotopes of hydrogen were identified using stopping power (dE/dx) in detector A and kinetic energy of the particle, as shown in Figure 7. Kinetic energy was reconstructed using relativistic kinematics and the simulated ToF between the two sensors [42]. Comparing computational stopping power of proton in the diamond detector and NIST graphite target shows that the general slope shape of the stopping power curve is achieved [40]. However, the diamond detector shows higher stopping power for energies greater than 100 MeV due its higher density.
(a)
(b)
5. Conclusion
The measured spectra for diamond sensor A exhibits good agreement with Geant4 simulations for proton beams. The diamond detector B exhibited poor response, perhaps due to poor electrical connections and/or front-end electronics. A wide energy range of produced proton particles made it impractical to distinguish between different ions using only ∆E1. However, increasing the separation distance between the two diamonds with appropriate fast electronics could lead to particle discrimination between energetic light ions using ToF.
Data Availability
The data used to support the findings of this study are available from the corresponding author upon request.
Disclosure
Part of this research was conducted in the Micro-Processing Research Facility, a core facility of the University of Tennessee.
Conflicts of Interest
The authors declare that there are no conflicts of interest regarding the publication of this paper.
Acknowledgments
The authors would like to express their gratitude to Mike Sivertz and Adam Rusek of the NASA Space Radiation Laboratory at Brookhaven National Laboratory for their assistance. Also, the first author would like to thank the King Abdulaziz City for Science and Technology (KACST) for sponsoring his Ph.D. studies at the University of Tennessee.