Charge-Coupled Device

CCD comprises a series of coupled Metal-Oxide-Semiconductor (MOS) capacitors as its fundamental structure. Electrons are produced when photons hit MOS capacitors because of the photoelectric effect, and will be collected as signal charge packets by potential wells of each MOS capacitor, which are formed by applying positive voltages to gate electrodes. Subsequently, the packets stored in each MOS capacitor are transferred from one pixel to its neighbor through sequential adjusting gate voltages of each MOS capacitor. The last capacitor dumps its charge into a charge amplifier, which converts the charge into a voltage, and then readout. MOS capacitors share readout circuit, so the transferring process will be repeated continuously until all the signal charge packets are read out.

CCD-based detectors can be implemented in several different architectures. The most common are full-frame, frame-transfer, and interline. In a full-frame device, all of the imaging area is active. While in frame-transfer CCDs, half of the silicon area is covered by an opaque mask, typically aluminum. It is necessary to quickly transfer input signals from the image area to the opaque or storage area to minimize smear in this design. Unlike frame-transfer, the interline architecture involves only a pixel shift from the imaging area to the storage area, enabling shutter times of less than 1 μs and virtually eliminating smear.

The advantages of CCD are apparent. It has excellent spatial resolution because its pixel size can be very small. Pixels as small as 1.56 μm × 1.56 μm have appeared as early as 2006. Also, the CCD has low noise because its readout process does not produce noise, making it well-suited for small signal detection. In addition, shared readout provides good pixel-to-pixel consistency. What’s more, it’s worth noting that modern CCD technology allows for manufacturing large wafer-level devices, simplifying the mechanical design of detectors, which is crucial for experiments requiring a large sensitive area [8].

The primary utilization of CCD is in astronomical observation. For instance, the Skipper-CCD, designed for the Oscura experiment [9] and employed in low-energy neutrino studies and dark matter searches, is crafted from high-resistance (HR) n-type silicon wafers. The sensor has a thickness of 675 µm, an effective area of 1.9 cm × 1.6 cm, and charge can be read through amplifiers at the four corners. Meeting the Oscura’s stringent requirements, including a readout time of less than 2 hours and a radioactive background rate of 0.01 event/kg/day/keV, the sensor exhibits a dark current of approximately 10^-6 e^-/pix/day. Test results conducted at Fermilab (as depicted in Fig. 1) reveal that 71% of the sensors exhibit sub-electron noise, with a surface dark current of about 0.031 e^-/pix/day [9].

Fig. 1Full-screen View

(Color online) Left: Skipper-CCD in Cu module. Center and right: Setups at FNAL for testing packaged sensors [9].

The "Mozi" Wide Field Survey Telescope (WFST) is a new generation survey telescope that is being built in China. It is equipped with a mosaic CCD camera with 0.73 gigapixels on the primary focal plane for high quality image capture over a 6.5-square-degree field of view [10]. Comprising nine E2V CCD290-99 chips, each with a resolution of 9 K × 9 K and a pixel size of 10 µm, the camera supports full-frame or split full-frame readout modes. The 3D model of the CCD detector system is illustrated in Fig. 2. The readout can occur through 8 or 16 output channels, with a typical readout noise of 4 e^-@0.5MHz. To meet the scientific objective of rapid sky surveying, the maximum readout speed reaches 3 MHz, and the readout noise is maintained below 15 e^-@1MHz [11]. Nowadays, CCD also finds applications in fields such as spectral analysis and heavy ion Computed Tomography (CT) [12].

Fig. 2Full-screen View

(Color online) 3D model of the CCD detector system on “Mozi"[11].

Silicon Strip Detector

The SSD is the first detector device using the lithographic capabilities of microelectronics, mainly used as a vertex and tracking detector. As shown in Fig. 3, the SSD is commonly fabricated through selective etching of microstrips on the surface of N-type silicon wafers. The highly doped P+ microstrip forms a PN junction on the surface and extends its depletion layer across the entire N-type substrate with applied reverse bias voltage. After covering aluminum on the top, a microstrip that acts as a charge-collecting electrode is completed. Multiple strips form a single-sided SSD, which provides one-dimensional position detection. When particle hits, electron-hole pairs are generated in the depletion region because ionization caused by deposited energy. Electrons, driven by an external electric field, are collected by the strips, converted into electrical signals, and processed by the readout circuit. The double-sided SSD featuring heavily doped P+ junction strips that collect the electrons and the N+ ohmic strips that collect holes can realize two-dimensional position detection. Since SSDs have low noise, good linearity, a high count rate, and excellent position and energy resolution, they are widely used in high-energy physics, nuclear physics, and space detection facilities [13, 14].

Fig. 3Full-screen View

(Color online) Working principle diagram of AC coupled single-sided SSD [15].

In high-energy physics, the radiation-hardened Silicon Microstrip Vertex Detector (SVX’) was incorporated into the CDF experiment during the Tevatron $\overline{p}p$ collider run 1B in 1993. As the first silicon vertex detector in a hadron collider environment, SVX’ modules (also referred to as barrels) consist of four layers of SSDs segmented into twelve 30 wedges. Three wafers are bonded to form a 25.5 cm long "ladder" with strips aligned for γϕ coordinate measurements. The AC-coupled readout is facilitated by the radiation-resistant SVXH3 chip[16], utilizing a 1.2 μm CMOS process with radiation tolerance exceeding 1 Mrad. The chip provides a gain of 21 mV/fC and an equivalent input noise of 1300 e^- at a strip input capacitance of approximately 30 pF [17], supporting data sparse and serial readout modes [18]. When a single strip is hit, the positional resolution can reach 13 μm [16]. The SSD has also been widely used in high-energy experiments, such as the Semiconductor Tracker (SCT) in ATLAS for two-dimensional track measurement[19], the Silicon Strip Tracker (SST) in CMS for readout of the γϕ coordinate and measuring three-dimensional information for the separation of particle tracks[20], and the NA11 spectrometer at CERN SPS for flight paths reconstruction [21], and so on.

SSD plays a vital role in nuclear physics experiments. The Institute of Modern Physics (IMP) has constructed a gas-filled recoil separator called Spectrometer for Heavy Atom and Nuclear Structure (SHANS) to study the heavy and superheavy nuclei properties. A silicon semiconductor detector box (Si-box) is installed at the focal plane position of the separator, as illustrated in Fig. 4. The Si-box incorporates three Position Sensitive Silicon Detectors (PSSD) at the back as implantation detectors, with a thickness of 300 μm and a collective effective area of 150 mm × 50 mm, offering 3 mm horizontal position resolution. The PSSD demonstrates a typical energy resolution of 50 keV FWHM for 5–10 MeV α particles. Leveraging this PSSD detector, it can be used to measure α particles of 1–20 MeV, implantation and fission fragments with energies of 5–200 MeV, and detect the escaping radioactive decay events [22]. Similar applications of SSD in nuclear physics can also be found in MUST, TIARA, and HIRFL-RIBLL. The MUST ("MUr à Strips") detector consists of 8 SSD–Si (Li) telescopes used to identify recoiling light charged particles through time of flight, energy loss and energy measurements and to determine precisely their scattering angle through X and Y position measurements [23]. The Transfer and Inelastic All-angle Reaction Array (TIARA) consists of 8 resistive charge division detectors forming an octagonal barrel around the target and a set of double-sided silicon-strip annular detectors positioned at each end of the barrel, designed to study direct reactions induced by radioactive beams in inverse kinematics [24]. The detection system consists of five Δ E - E telescopes, consisting of one double sided silicon strip detector (DSSD) as the Δ E detector, and a square silicon detector as the E detector, designed to measure the angular distributions of both elastic scattering and breakup simultaneously, on the Radioactive lon Beam Line in Lanzhou at Heavy lon Research Facility in Lanzhou(HIRFL-RIBLL) [25].

Fig. 4Full-screen View

(Color online) Schematic view of the focal plane detection array [26].

The SSD has been a "standard" component for space detection [27]. The Dark Matter Particle Explorer (DAMPE), launched in 2015, incorporates silicon-tungsten tracker-converter (STK) to collect the electron-positron pairs converted from the incoming photons by the tungsten converters. As shown in Fig. 5, the STK contains of 6 tracking planes each consisting of 2 layers of SSD arranged orthogonally. Each single layer consists of 16 ladders, and each ladder comprises 4 modules bonded end-to-end. The SSDs are glued on the flex part of the Tracker Front-end Hybrid (TFH) board to form a ladder. Each layer of SSD has an active area of 0.55 m², with a total radiation length of 0.976% X₀, spatial resolution < 80 μm within 60 incidence, and a total weight of 154.8 kg.

Fig. 5Full-screen View

(Color online) Left: Exploded view of the STK; Right: The STK single ladder, made by four SSDs[28].

The readout is performed on one every other strip (corresponding to 384 channels per ladder) in the SSD. The signal shaping and amplification is performed by six VA140 ASIC chips mounted on the TFH[28]. The VA140 is a 64-channel low-noise charge-sensitive amplifier ASIC from the Norwegian IDEAS company. Each channel of the VA140 comprises a charge-sensitive preamplifier, a filter shaping circuit, and a sample and hold circuit. The input charge dynamic range spans from -200 fC to 200 fC, with a shaping time of only 6.5 μs. This ASIC employs serial readout through a multiplexing circuit and the power consumption is 0.29 mW/channel.

Based on the successful development and operation of China’s first DAMPE, the Purple Mountain Observatory of the CAS, together with several institutions, proposed the Very Large Area Gamma-ray Space Telescope (VLAST).As shown in Fig. 6, the VLAST plans to use SSD to build Silicon Tracker and low Energy gamma-ray Detector (STED). The main function is to realize the conversion of high-energy gamma photon-electron pairs, and to achieve high-angular resolution observation of gamma photons through the track measurement of electron pairs. The main technical indicators of STED are [29]:

Fig. 6Full-screen View

(Color online) The schematic plot of the payload of VLAST [29].

• Number of detector layers: 8 super layers; (each super layer includes a CsI detection layer and two large silicon micro-strip detection layers);

• Active detection area of each layer: ≥ 2.8 m × 2.8 m;

• Detection energy range: 1 MeV –100 MeV;

• Spatial resolution < 0.1 (@50 GeV).

The VA140 from IDEAS company is also a potential candidate for the readout of the SSD on VLAST. In addition, the IMP is now developing a readout chip called SiReadout. The first version of the chip is shown in Fig. 7. The preliminary single-channel test indicates that it has a gain of 2.01 mV/fC, power consumption of approximately 220 μW, a peaking time of 3.12 μs, linearity error below 1%, and an Equivalent Noise Charge (ENC) of 992 e^- at zero F plus 13.5 e^- per pF.

Fig. 7Full-screen View

(Color online) Microphoto of the first version of SiReadout.

Also, the ALPHA Magnetic Spectro-meter (AMS) experiment similarly utilizes SSD to build a magnetic spectrometer to measure momenta, charges and mass of the particles [30], the Fermi Large Area Telescope (Fermi-LAT) builds the LAT tracker to reconstruct the incoming photon direction, etc [31].

Silicon Pixel Detector

Silicon pixel detector is currently one of the main detectors and is expected to play a significant role in the future due to their exceptional spatial resolution, reduced material budget, low noise, and rapid readout rates. It mainly consists of a densely packed pixel array and peripheral readout circuits, with each pixel containing a charge-sensitive unit for collecting electrons and outputting electrical signals and its front-end signal processing circuit. The distinct signals generated by particles interacting with different pixels offer precise two-dimensional positional information. Silicon pixel detectors can be classified architecturally into monolithic and hybrid types.

Monolithic Active Pixel Sensor The monolithic type integrates the charge-sensitive unit and front-end circuit on a single substrate, minimizing complexities in interface designs. This integration ensures high granularity, a low material budget, and a relatively small equivalent input capacitance. Monolithic Active Pixel Sensor (MAPS) is currently one of the most promising and well-developed technologies among monolithic types.

MAPS shares the similar detection principle as SSD. As depicted in Fig. 8, a lightly doped P-type epitaxial (EPI) layer grows on a substrate heavily doped with P-type impurities, the Nwell/P-EPI diode is the charge-sensitive unit and the front-end circuit is also on the same substrate connecting to it directly. MAPS typically features pixel sizes ranging from 10–30 μm, resulting in a position resolution of 3–7 μm. The readout speed can achieve hundreds of kilohertz, and noise is maintained at approximately 10e^- level. MAPS demonstrates a good balance in resolution, material budget, radiation hardness, readout speed, and power consumption.

Fig. 8Full-screen View

(Color online) Left: The array structure of MAPS; Right: The principle of MAPS.

The EPI process in MAPS significantly impacts charge collection. The standard EPI process has a thickness of approximately 10 μm which causes a small depletion layer, and an electrical resistivity of around 10 Ω cm, making thermal diffusion the dominant way for charge collection. Therefore, it needs a long time to collect electrons and the collection efficiency is very low. In contrast to that is an HR EPI process with a thickness of up to 40 μm and an electrical resistivity that potentially can reach ∼1000 Ω cm. The depletion layer is thicker, and charges can also be collected by drifting. The process of collecting is faster and more efficient. Moreover, it reduces the probability of electron-hole pair recombination, enhancing radiation hardness and improving the Signal-to-Noise Ratio (SNR). In the standard twin-well process (Fig. 9 left), PMOS cannot be used in pixel design because the Nwell where PMOS resides competes with the Nwell of the diode, leading to a reduction in Charge Collection Efficiency (CCE). Hence, the deep Pwell is needed to isolate the two. In addition, applying a reverse bias voltage to the substrate could expand the depletion layer region which means increasing CCE and improving radiation hardness, but it may also cause the peripheral readout circuit on the same substrate not to work properly, so deep Nwell is needed for protection. Simulations can give the preliminary study of charge collection in certain process for specific particles in the early phase of design [32-34]. In summary, the prevailing approach for MAPS involves the utilization of a complete quadra-well process (Fig. 9 right) with HR EPI.

Fig. 9Full-screen View

(Color online) Left: twin-well process; Right: quadra-well process.

In experimental physics, MAPS finds primary applications as vertex and track detectors in collider experiments. The Heavy Flavor Tracker (HFT) detector in the STAR experiment is the world’s first detector to use MAPS to enhance vertex resolution and extend measurement capabilities in the heavy flavor domain [35]. The HFT detector incorporates MAPS in its innermost two layers (at radii of 2.8 and 8 cm from the beamline) among four concentric cylinders near the STAR interaction point, and these layers feature 400 Ultimate (MIMOSA-28) chips, covering a total silicon area of 0.16 m², as illustrated in Fig. 10. Every layer costs a global material budget of only 0.5% X₀ because the chip on it operates normally with air cooling at room temperature[36].

Fig. 10Full-screen View

(Color online) Left: Schematic view of the HFT inside the TPC inner field cage [36]; Right: MIMOSA28 on its PCB [37].

The Ultimate chip, designed by the IPHC, fabricated by 0.35 μm HR OPTO process (400 Ω cm), features a pixel array of 928 × 960 pixels with a pixel pitch of 20.7 μm, providing a sensitive area of approximately 3.8 cm². Its pixel includes a sensing diode for charge collection as well as an amplification circuit and a Correlated Double Sampling (CDS) circuit for signal extraction and noise removal. Each pixel column is terminated with a high precision discriminator and is read out in a rolling shutter mode at 5 MHz. The discriminator’s outputs are processed through an integrated zero suppression logic and the sparsified data are sent out to the acquisition via two 160 Mbps LVDS outputs. The Ultimate demonstrated a MIPs detection efficiency near to 100%, a fake hit rate of less than 10⁴, and power consumption maintained at 150 mW/cm²[37].

The ALICE Inner Tracking System 2 (ITS2) stands as the world’s largest detection system employing MAPS. As illustrated in Fig. 11 left, the detector layout features concentric 7 layers with radii ranging from 23 mm to 400 mm. These layers consist of over 24120 ALICE Pixel Detector (ALPIDE) chips (Fig. 11 right) [38], forming a 10 m² active detection area with a total of 12.5 × 10⁹ pixels.

Fig. 11Full-screen View

(Color online) Left: Layout of the ITS2 detector [39]; Right: The ALPIDE chip [40].

ALPIDE is fabricated using the TowerJazz 180 nm quadra-well HR CMOS process (>1 kΩ cm), and features 512 × 1024 pixels with the pitch of 28 μm, providing a sensitive area of 30 mm × 13.8 mm [41]. Its pixel includes an octagon collection diode which achieves nearly 100% CCE and a distinctive front-end circuit consisting of a continuously active discriminating amplifier and a multiple-event memory, which yields an ENC of less than 10 e^-. The in-pixel multiple-event memory is read out asynchronously through a priority encoder circuit in each double column. This is fast and power efficient as the expected occupancy is low and only hit pixels are read out in a hit-driven mode. Data is collected at the periphery and shipped off the detector by means of a high-speed serial link [42]. Extensive beam tests confirm that ALPIDE achieves a detection efficiency exceeding 99%, a false hit probability below the required 10^-6 level, and a spatial resolution better than the desired 5 μm. Even after exposure to a TID of 2.7 Mrad and a non-ionizing energy loss (NIEL) flux of 1.7×10¹³ 1 MeV neq/cm², surpassing the anticipated detector lifetime, it sustains performance and provides a readout rate of 100 kHz with a power density of less than 40 mW/cm² [43].

MAPS also plays a crucial role in high-resolution beam telescopes for tracking charged particles, exemplified by the Desy II beam telescope. The first generation adopted the EUDET-type framework, employing MIMOSA-26, which provided outstanding spatial resolution and a lower material budget. Achieving a distribution resolution of 1.83 μm for a 6 GeV electron or positron beam on the Device Under Test (DUT), the rolling shutter readout operated at a frequency of approximately 10k frames per second. As MIMOSA-26 production ceased and updating became unfeasible, a new beam telescope was constructed based on ALPIDE technology (Fig. 12). This telescope features a sensitive area of 414 mm², a trigger rate of up to 100 kHz, and significantly reduced noise levels compared to the previous telescope. This enhancement allows for more efficient beam utilization and a cleaner tracking environment. The spatial resolution remains in the micron range (approximately 5 μm at 6 GeV), with a low fake hit rate [44].

Fig. 12Full-screen View

(Color online) Design of the ALPIDE beam telescope system [44].

MAPS is increasingly pivotal in proton and heavy-ion imaging, particularly in the burgeoning field of ion beam radiotherapy. Unlike traditional photon radiation, ion beams exhibit a Bragg peak, a concentrated dose deposition, whose depth can be finely adjusted by altering the beam energy. This feature enhances protection for healthy tissues [45]. Proton and heavy-ion imaging research contributes to refining treatment plans for proton/heavy-ion therapy [46]. Current methods of making particle therapy treatment plans rely on converting X-ray CT HU value into Relative Stopping Power (RSP) of proton/heavy ions, which led to inevitable range uncertainties. However, the Bergen proton CT (PCT) system (Fig. 13), incorporating a Digital Tracking Calorimeter (DTC), can directly measure the RSP to address this problem.

Fig. 13Full-screen View

(Color online) Principle of the Bergen proton CT [47].

The DTC comprises 43 layers, each equipped with 108 ALPIDE chips, which feature the first two layers as tracking layers for acquiring the position information of incoming particles and the subsequent 41 layers as a digital calorimeter. As a particle traverses the detector, it deposits energy in the sensitive layer, generating a 3D digital hit map along its path. The final prototype has a measuring area of 27 × 16.6 cm², which is adequate for imaging the human head [47].

Another compact full digital system that utilizes the MAPS with a similar structure to Bergen PCT but for heavy-ion imaging is the Hi’CT proposed by IMP (Fig. 14). The simulation results achieve less than 1% deviation in Relative Stopping Power (RSP) and better image quality of CATPHAN phantoms. The real-time imaging capability of Hi’CT may reduce target deviations caused by organ or respiratory movement, achieving optimized treatment effectiveness and improved quality of life for patients [48].

Fig. 14Full-screen View

(Color online) Conceptual design of the Hi’CT [48].

MAPS technology has advanced with the adoption of stitching technique in wafer-level chip manufacturing. The stitching technique is the way to continue the scaling of technology in the semiconductor industry because of the reticle size limitation [49]. It is achieved by dividing the reticle into sections that could individually be exposed and aligned with neighboring sections [50]. As Fig. 15 left shows, the design reticle gets subdivided in sub-frames that correspond to sub-frames of the photomasks. During the photolithographic patterning of wafers, these are selectively exposed to adjacent locations according to a pre-established pattern. This requires very accurate translations and alignment of the wafers between each exposure. The peripheral structures along the outer edges of the core array and at the four corners are designed in dedicated sub-frames of the reticle. The red lines in the figure indicate the dicing lanes. With a judicious design of the geometries in the reticle sub-frames, large chips with diagonals approaching the wafer diameter become feasible, provided their core area can be constructed as an array of sub-units regularly repeating in space. Interconnections across the sub-units are made with the joining of wiring geometries at the abutment boundaries [51].

Fig. 15Full-screen View

(Color online) Left: Principles of stitching [51]]; Right: Layout of the ITS3 Inner Barrel [52].

The upgrade of ALICE’s ITS3 (Fig. 15 right) is prepared to utilize stitching technology to manufacture a kind of single MAPS on a 12-inch silicon wafer, achieving dimensions of 27 cm × 9 cm. ITS3 is expected to significantly reduce the material budget from 0.35%X₀ to 0.05%X₀ because the new MAPS will be capable of self-bending without the need for complex support and cooling systems. Simultaneously, ITS3 can place the innermost layer at a radial distance of only 18 mm from the interaction point. These features will enhance the resolution of impact parameters to twice that of all momenta and significantly improve tracking efficiency at low transverse momentum [52]. In addition, large scientific facilities such as the Electron-Ion Collider in China (EicC), the Electron-Ion Collider (EIC), and the Nuclotron-based Ion Collider Facility (NICA) are also preparing to adopt wafer-scale MAPS implemented with stitching technology.

As mentioned before, reverse bias can enlarge the depletion layer region, but normal CMOS processes typically have a maximum bias substrate voltage of 5 V [53]. High voltage (HV) CMOS technology is starting to be used in MAPS to improve sensor properties. Unlike the quadra-well process mentioned earlier, in the HV CMOS process, the deep Nwell is used to protect the entire circuit except for the charge-sensitive unit, which is another kind of diode composed of EPI and the deep Nwell itself as shown in Fig. 16. Such a design can even achieve a 100% fill-factor [54]. A typical example is ALTASpix3,originally designed for the upgrade of the ATLAS tracking experiment and a option for the silicon tracking detector of the Circular Electron Positron Collider (CEPC). Its deep Nwell connects to the positive power supply rail at 1.8 V and the maximum voltage of the substrate is around 66 V, corresponding to a depleted region of approximately 30 μm [55].

Fig. 16Full-screen View

(Color online) The principle of HV MAPS [55].

The Silicon-On-Insulator (SOI) CMOS technology has also been used to significantly enhance radiation hardness for MAPS design. SOI wafers are composed of a thin Silicon film (device layer) on top of an insulating layer (buried oxide) that is manufactured on a standard Silicon substrate. In traditional SOI technologies, the substrate serves solely as a mechanical support for the silicon film. In detector applications, as shown in Fig. 17, the substrate has high resistivity and is utilized for creating the sensor (a matrix of fully depleted diodes) monolithically coupled to the readout electronics integrated in the silicon film above the insulating layer [56]. Laboratory laser testing of SOI CMOS pixel detectors has demonstrated excellent TID tolerance up to 1 Gray [57].

Fig. 17Full-screen View

(Color online) The principle of HV MAPS [56].

One consideration to address limitations in pixel size is constructing circuit chips in three-dimensional (3D) structures [58]. One method is 3D stacking, illustrated in Fig. 18, employing Through-Si Via (TSV) for vertical communication.

Fig. 18Full-screen View

(Color online) Schematic cross-section of the two tiers (sketch not to scale) [59].

Another method, known as 3D detectors, decouples electrode distance and substrate thickness. It allows to place the electrodes perpendicular to the silicon wafer penetrating through the substrate. Therefore, in contrast to standard planar detectors, the depletion region laterally grows between the electrodes [60], as depicted in Fig. 19.

Fig. 19Full-screen View

(Color online) Left: Schematic cross-sections of planar sensor; Right: 3D sensor. emphasizing the decoupling of active thickness (Δ) and collection distance (L) in 3D sensors [61].

Nuclear physics experiments also require MAPS to have the capability for time and energy measurements beyond position only. Time measurement is needed to address the high counting rate requirements, such as the China Hyper-Nuclear Spectrometer (CHNS) and precise monitoring of heavy ion beams, which may require counting rates as high as 10⁶ to 10⁸ counts per second, corresponding to a time precision of 10 ns to 1 μs. Energy measurement contributes to position resolution and particle identification. For example, the Bρ-DE-TOF (magnetic rigidity-ionization energy loss-time of flight) method, which combines momentum generated by trajectory curvature and the Bethe-Bloch formula, is mainly used in the identification of particles in medium to high energy radioactive beams and the identification of reaction products on secondary targets. According to the relationship between ionization energy loss Δ E and particle charge number Z, and velocity v: $\Delta E\propto Z^2/v^2$, using Δ E combined with TOF measurement can realize particle identification [62]. Meanwhile, for real-time monitoring of heavy ion beams at facilities such as the Heavy Ion Research Facility in Lanzhou (HIAF), the spatial resolution can be improved by calculating the center of mass of large clusters using energy information [63]. It can also provide calibration for time measurements.

MAPS development in China started in the 2010s and has grown rapidly. The representing ones for MIPs detection are the MIC series [64] from Central China Normal University (CCNU), the Taichupix [65] and Jadepix [66] from the Institute of High Energy Physics (IHEP) of the CAS, the SuPix from Shandong University [67], etc. The domestic MAPS with multi-dimensional measurement for nuclear physics includes the Nupix [68, 69] from the Insitute of Modern Physics (IMP) of the CAS, and the Topmetal-M [70] from IMP and CCNU, both designed with China’s own CMOS process.

Hybrid Hybrid pixel detectors feature the charge-sensitive unit and front-end readout circuit on separate substrates connected by flip-chip bonding. The hybrid pixel detectors can use different materials for the charge-sensitive unit and readout circuit, which can be optimized separately. For example, the GaAs, CdTe, and other non-silicon materials are used for the sensor part in X-ray imaging, and the readout part is designed with silicon [6]. The simplified design of the charge-sensitive unit also enables good yields at larger sizes. Therefore, hybrids generally offer flexible readout capabilities and broader applications compared to MAPS. However, the complex and labor-intensive production process and numerous production steps make hybrid pixel detectors relatively more costly [57].

The reliability and excellence of the flip-chip bonding techniques significantly impact overall chip performance. The solder bumps and bonding technique is the most commonly used (Fig. 20). Before producing the solder bumps, the sensor and chip surfaces undergo treatments, such as cleaning and de-oxidizing. After accurately locating the devices to be connected, the solder is heated to melt and then cooled to form a strong connection. IBM introduced solder balls in the C4 (Controlled Collapse Chip Connection) process in 1969. However, the bumps’ pitch limit of 170 μm or greater is often too large for fine-pitch and high-granularity pixel detectors in particle physics. Eutectic solder bumps are considered to be the primary soldering technology for hybrid pixel detectors, with high connection density of as low as 25 μm and bump sizes of 15 μm. Higher yields can be achieved using Sn/Pb and lead-free Sn/Ag or Au/Sn alloys, and the final pitch limit is around 5–10 μm. Eutectic solder bumps have been used to realize hybrid pixel detectors in ATLAS, achieving high bonding rates of more than 99%. Other technologies include In-In bonding, Cu-Cu direct bonding, oxide-oxide direct bonding, and Solid-Liquid Interdiffusion (SLID) bonding. Overall, bonding technology requires adding small capacitance to pixel preamplifiers through bonding connections, good yield and defect rate of less than (open or shorted), good contact (<100 mΩ), and robustness to temperature cycling (-40 ℃ to +60 ℃) [71].

Fig. 20Full-screen View

Eutectic solder bumps and the flip-chipping process [71].

Hybrid pixel detectors have been widely used in collider physics experiments as vertex and track detectors. The first hybrid pixel detector was developed in the RD19 project WA97 particle physics experiment. The hybrid pixel detector has a 64 × 96 pixel array, with a sensitive area size of 53 mm × 4.73 mm and a thickness of 300 μm. Six Omega2 readout chips, each one containing 16 × 64 front-end circuits, are bump-bonded to one pixel array. Each pixel is connected to a virtual ground via a front-end circuit by a Pb-Sn solder bump with a diameter of 38 μm. Each front-end circuit contains a low noise amplifier followed by a comparator with an adjustable threshold (from 4000 to 15000 e^-), an adjustable delay line (from 100 ns to 1 μs), and some coincidence and memory elements, which store the bit indicating a particle hit during the time frame selected using an external strobe signal. During the proton run of WA97 in November 1993, the intrinsic precision of the pixels was about 24 μm, and the efficiency of good chips was 98–99%, while the average efficiency of the entire detector (including bad areas) was about 80% [72]. Subsequently, from 1998 to 2006, the hybrid pixel detectors were installed in the four experiments (ALICE, ATLAS, LHCb, and CMS) at CERN.

In recent years, hybrid pixel detectors have seen further development in synchrotron experimental facilities, primarily for Hybrid Photon Counting (HPC). Since the initial concept validation for synchrotron sources in 1999, HPC has had a profound impact on nearly all X-ray applications in synchrotrons and laboratories, significantly affecting data quality, data collection speed, and the development of complex data acquisition schemes. A recent example is the EIGER2 readout chip, with a pixel size of 75 μm × 75 μm and an array size of 256 × 256, featuring an amplifier and two energy discriminators (Fig. 21). It has four functionalities: a 8-bit readout mode that doubles the maximum frame rate, a dual-gating mode for parallel measurements of two time-delays in pump-probe experiments, a multi-image stream mode and a line region of interest (lines-ROI) mode with a high frame rate of up to 98 kHz. The charge-sensitive unit that goes with it can be Si or CdTe for photon energies of 3.5–40 keV and 8–100 keV. Since 2018, EIGER2 has been used in the production of various detectors. In the field of macromolecular crystallography (MX), they have been used to employ high-energy X-rays for reducing radiation damage, and in the biomedical imaging field, they are used to advance scanning techniques [73].

Fig. 21Full-screen View

(Color online) Left: EIGER2 readout pixel electronic-circuit scheme; Right: The principle of photon counting detection with two thresholds [73].

The High Energy Photon Source (HEPS) in Beijing also needs a photon counting pixel detector and here comes the HEPS-BPIX series. The HEPS-BPIX chip is a dedicated pixel readout chip that operates in single photon counting mode for X-rays. Fabricated by SMIC 0.13 μm process, it contains a 104 × 72 matrix of 150 μm square pixels. Bump bonded with a 300 μm thick silicon charge-sensitive unit, the measurement results showed that noise after bonding was 115.8 e^- rms, the threshold dispersion was 55.1 e^- rms after calibration and trimming, and the frame rate was 1.2 kHz with a negligible dead time of 175 ns per frame at 20 MHz clock frequency. All the performance was found to satisfy the project’s requirement [74].

Hybrid pixel detector technology has been widely adopted in various scientific fields, such as medical X-ray imaging, synchrotron radiation applications, low-energy electron microscopy, materials analysis using X-ray diffraction, adaptive optics, dosimetry, and space dosimetry [75]. The Medipix and Timepix series, developed at CERN, are the most common hybrid pixel detectors applied in medical imaging.

The Medipix is primarily used for X-ray photon counting. The latest iteration, Medipix4, is designed to facilitate large-scale applications utilizing high-z materials to achieve high rates (up to 5 × 10⁸ photons/mm²/s) at fine spacing [76]. Medipix4 can be tiled on all 4 sides, making it ideal for building large-area detectors with minimal dead zones, and it is characterized by energy grouping of individual photons while maintaining spectral accuracy by implementing an inter-pixel architecture that corrects for the effects of charge sharing. The Medipix4 chip has three modes of operation: High Dynamic Range Mode (HDRM), Low Noise Mode (LNM), and Ultra-Fast Mode (UFM). In HDRM, the chip can use a CdTe sensor to process X-ray photons with energies up to 154 keV [77]. In LNM, with a fine-pitch configuration without charge-sharing correction, the measured electronic noise is 72 e^- rms [78]. In UFM, the front-end’s post-layout analog count-rate capability is 19 × 10⁶ photons/mm²/s [77]. The chip can read out charge-sensitive units with 320 × 320 pixels measuring 75 μm × 75 μm (Fine Pitch Mode, FPM) or 160 × 160 pixels measuring 150 μm × 150 μm (Spectral Mode, SM). The Medipix ASICs have been applied in X-ray CT, prototype systems for digital mammography, CT imagers for mammography, and beta and gamma radiation autoradiography of biological samples. In addition, they can also be used for commercial X-ray material analysis [79].

The Timepix, introduced as a development based on the Medipix2 chip in 2005, addresses the demand for high-resolution time measurement scenarios. It operates in three modes: the hit counting mode, which tallies signals exceeding the threshold, incrementing the count for each signal; the Time of Arrival (ToA) mode, where the counter starts when the signal arrives and stops when the signal ends; the Time over Threshold (ToT) mode, during which the counter tallies the duration the signal exceeds the threshold. The latest Timepix4 is a 24.7 mm × 30.0 mm hybrid pixel detector readout chip with a 448 × 512 pixel array, specifically designed for tiling the detector on all four sides, and can count incoming hits at rates maximum of up to 5 GHz/mm²/s. The block diagram of the Timepix4 pixel cell and super pixel architecture is shown in Fig. 22 [80]. The chip can read out charge-sensitive units at a pitch of 55 μm.

Fig. 22Full-screen View

(Color online) A block diagram of the Timepix4 pixel cell and super pixel architecture [80].

In addition, the Topmetal [70], designed by Central China Normal University, and the IMPix series [81], currently in development at the IMP, are also hybrid pixel readout chips developed by Chinese researchers. It’s worth noting that these chips can also directly collect charge with their exposed topmost metal layer, which can be used for TPCs [70].

Low Gain Avalanche Diode

The high luminosity physics experiments increase the number of collisions in each bunch crosses dramatically. The LGAD, invented by the Centro Nacional de Microelectronica (CNM), Barcelona, Spain, with the support of the CERN-RD50 community [82], can provide a time resolution better than 40 ps and becomes an excellent choice of solid-state timing detectors [83]. Combined with the 3D position information, LGAD-based detectors are considered suitable for tracking.

As an innovative charge-sensitive unit, LGAD enables rapid measurements of MIPs combined with the readout circuit. Its fundamental principle leverages the avalanche gain effect to enhance the sensitivity and resolution. As shown in Fig. 23, an n⁺/p/p^- junction is created along the center of the LGAD’s electrodes. With a reverse bias voltage, a high electric field region is generated in the junction region, which can lead to a multiplication of electrons reaching the n⁺ electrode. The p-type multiplication layer is to enhance the electric field in that region, and its impurity profile adjusts the gain [84].

Fig. 23Full-screen View

(Color online) Schematic cross-section of the LGAD pad design [84].

An incident particle to junction generates many carriers, which are then accelerated by an electric field, producing additional secondary carriers, culminating in an avalanche effect that yields a substantial charge signal. By controlling the doping of the P-type multiplier layer, an appropriate multiplication of around 10 allows the fabrication of thinner devices with the same output signal of standard thick substrates [84], which means a decrease in noise and better SNR. On the other hand, the signal shape is mainly governed by the holes escaping toward the backside. Hence, the LGAD with thin active thickness makes the signal rise faster and the collection time shorter [85]. Better SNR means more negligible time jitter and faster signals mean smaller time walks. Therefore, a very high time resolution is achievable.

This LGAD technology has become an emerging detector for physics experiments. The High Granularity Timing Detector (HGTD) based on LGAD [86] from IHEP is a crucial component of the ATLAS phase II upgrade to cope with the extremely high pile-up. This HGTD detector is located in the gap region between the end-cap calorimeter and the cylinder, covering a pseudo-fastness range of 2.4 $\left|\eta \right|$ 4.0. Complementing the internal tracker by providing high-precision temporal measurements for charged particles, and improving the reconstruction performance of physical objects. HGTD designed a low-noise readout circuit called ALTIROC (Fig. 24) bump bonded to a LGAD with a size of 1.3 mm × 1.3 mm × 50 μm. ALTIROC is produced by 130 nm TSMC technology, and has a similar architecture with readout circuits developed for pixel detectors. The front-end circuits include a preamplifier with a bandwidth of about 1 GHz and a fast discriminator with two Time to Digital Converters (TDC) followed for TOT and TOA measuring. TOT is used as an estimate of the signal amplitude to correct the TOA for the time walk effect offline. The digital front end is used to store the time data up to the reception of a trigger and buffers the data in order to be read by the End Of Column (EOC) cells. At last, a complex peripheral circuit is used to read the data from the chip. Beam test during 2016–2020 indicated that operated with a bias voltage of 230 V, the LGAD achieved a MIP charge deposit of about 20 fC, and the measured time resolution of HGTD is about 39 ps containing contributions from the electronics jitter, TDC and clocks [83].

Fig. 24Full-screen View

(Color online) The global architecture of the ALTIROC [83].

It is also expected to use LGAD to upgrade the internal layers of the CMS detector to optimize the detection performance of energetic particles. In the EiC and EicC projects, LGAD is also under consideration for developing new detectors.

One of the most significant drawbacks of LGADs is their poor spatial resolution. A development trend is using capacitively coupled LGAD (AC-LGAD) to address the dead zone issue at the pixel boundary of LGADs. This technology can achieve a spatial resolution of less than 10 μm. Other silicon detectors based on LGADs, such as deep-junction LGAD (DJ-LGAD) and trench-isolated LGAD (TI-LGAD), are also under development [82].

All Silicon Tracker

Recently, all trackers become a hot topic. This approach usually utilizes a radially compact structure, leaving more space for external particle identification (PID) detectors [87]. It also offers faster readout speed and significantly reduces the workload for calibration and alignment. Additionally, it can provide a higher resolution of momentum and single-track.

The ALICE collaboration has proposed an all-silicon tracker capable of covering the rapidity region |η| < 4 [88]. The detector concept comprises a barrel and two end caps constructed from layers of ultra-thin Si-sensors. As shown in Fig. 25, this all-silicon tracker includes the Inner Tracker (IT), outer tracker (OT), Time-Of-Flight (TOF) detector, and electromagnetic Shower Pixel Detector (SPD). The inner tracker comprises 3 layers within the beam pipe and 4 disks at both ends. It will employ wafer-level ultra-thin CMOS MAPS, manufactured using the stitching technique mentioned earlier, with 10 μm × 10 μm pixels providing better than 3 μm position resolution, with a material budget of 0.05% X₀ per layer. The outer tracker has 7 layers and 6 disks at each end, utilizing the same technology as the vertex detector. The pixel size is increased to 30 μm × 30 μm to reduce power density, yielding a spatial resolution of about 5 μm and a material budget of 0.5% X₀ per layer. The TOF detector is planned to be a silicon detector with a temporal resolution of 20 ps, based on LGAD. It is utilized for identifying hadrons and electrons at very low transverse momentum (p_T < 500 MeV/c). Additionally, the SPD can identify electrons and photons (p_T > 500 MeV/c) with a disk at each end. In summary, with unprecedented low mass, the all-silicon tracker would allow reaching down to an ultrasoft region of phase space, to measure the production of very-low transverse momentum lepton pairs, photons and hadrons at the LHC [88].

Fig. 25Full-screen View

(Color online) Longitudinal view of the experimental apparatus [89].

Meanwhile, the Electron-Ion Collider (EiC) plans to construct a fully silicon-based tracking detector, projected to be 242 cm long and 43.2 cm in radius. As shown in Fig. 26, six layers will provide tracker coverage for low values of pseudorapidity. They are laid out in three double layers to provide redundancy, and the middle double layer is placed equidistantly between the inner and outer double layers to measure hits in the vicinity of the sagitta to optimize the momentum resolution. In addition, five disks in each direction provide coverage at larger absolute values of pseudorapidity. With the current configuration, the material budget they contributed is less than 5% X₀. This geometry was implemented in GEANT4 and studied within the full Monte-Carlo framework for detector simulation. Preliminary results show that the design meets the physics requirements over the entire 0<p 30 GeV/c range for -2.0 < η < 3.0 at a 3.0 T field [87]. Furthermore, fully silicon-based detectors have also been proposed for the CHNS and EicC.

Fig. 26Full-screen View

(Color online) All-silicon tracker geometry of EiC [87].

The SiC Detector

Among the wide bandgap semiconductor materials, silicon carbide material has better quality and the most mature device technology, so that SiC radiation detector has a great development. There are more than 200 isomers in SiC crystals, and the C/Si atoms in each isomer are stacked in a different order. The three most common isomers are 3C-SiC, 4H-SiC and 6H-SiC, and the number before H represents the number of C/Si atomic layers in the stacking cycle. In the above various types of SiC, usually 4H-SiC is the most one in the common application [90].

The wide bandgap semiconductor material SiC has a wide bandgap with 3.26 eV at room temperature and has good radiation resistance. With extremely high breakdown voltage characteristics, this device can withstand high voltage and high current density. The saturation velocity of electron can improve the charge response speed of the detector. At the same time, it has high thermal conductivity, which is three times that of Si, and it has good heat dissipation properties and stable chemical properties. The device has the characteristics of high resistivity and small dark current. Especially in the three materials of gallium nitride, silicon carbide and diamond, SiC can obtain larger crystal size (6 inch and 8 inch), so SiC radiation detector is also the most deeply studied detector in the wide bandgap semiconductor.

The research of charged particle detector of silicon carbide can be traced back to the 1950s, and the neutron detector of silicon carbide was reported as early as 1957 [89]. In 1973, V.A. Tikhomirova et al. used a silicon carbide device as a fission fragment counter at a high injection rate (4 × 10¹⁴n/cm²), and compared with the silicon detector, it is confirmed that the silicon carbide device has good radiation hard characteristics [91]. However, in the next 20 years or so, due to the limitation of growing process technology, the quality of SiC material is relatively poor, which makes the research progress of SiC radiation detector is very slow.

Until the late 1990s, researches carried out a lot of fruitful research work, significantly reduced the defects (such as dislocation, microtubule, etc.) produced in the growth process of SiC crystals, and at the same time, the controllable doping process has also been significantly improved, which makes the manufacturing technology of high-performance SiC devices have been developed by leap. In 1997, F.H. Ruddy’s research team of Westinghouse Electric developed two SiC detectors, SBD and PN junction, with the energy resolution of 5.8% and 6.6% for α particles, respectively, and no effect on the performance of the detector in the temperature range of 22–89 ℃ [92]. With the continuous development and improvement of SiC thin film epitaxial growth technology, the research team reported the results of high energy resolution in 2006. The SBD detector with the epitaxial thickness of 100 μm was prepared, and the energy resolution of 5.449 MeV α particles reached 0.37% [93]. In 2012, the United States, a team of scientists at the university of South Carolina were compared the performance of the detector of an insulator SiC with n-type epitaxial layer on single-crystal SiC, the results show that the performance of the epitaxial film detector is better than single crystal detector performance, n-type epitaxial detector on the best bias is 100 V, to 5.486 MeV alpha particle energy resolution was 2.7%, within the scope of soft X-ray also have very good response. In 2015, B. Zat ’ko’s team at the Slovak Academy of Sciences reduced the epitaxial layer doping concentration to 1.0 × 10¹⁴ cm^-3, 15 nm Ni/Au (0.3 mm) was used as a Schottky electrode, the leakage current of the detector is only 2 pA at the reverse voltage of 700 V, and the energy resolution of 5.486 MeV α particles reaches 0.25%, which is one of the best reported result [94].

Heavy particle identification in nuclear physics experiments, there are Italian c. Ciampi and s. Tudiso researchers such as more than 40 scientific research personnel of SiCILIA (Silicon Carbide Detectors for Intense Luminosity Investigations and Applications) related research report of the project, they are based on 10 mm and 100 mm thick SiC telescope structure Δ E - E detector, The feasibility of SiC detector for particle identification under high fluence was verified [95]. In addition, proton beam monitoring in laser plasma experiments [96], α particle diagnosis in nuclear fusion [97] also have important applications. In addition, in recent years, some complex SiC detectors, such as SiC microstrip detectors, have been developed for the identification of heavily charged particles and isotopes [98].

Neutron detection technology has a wide application prospect in many fields such as space radiation environment detection, contraband detection, scientific experiment, medicine, military and industry [99, 100]. Experiments show that the SiC detector in proton 1.4 × 10¹⁶ p/cm² and 7 × 10¹⁵ n/cm² can still be work [101, 102]. In silicon carbide, ¹²C and ²⁸Si can react with fast neutrons to release α particles, which interact with SiC to generate electron hole pairs, this can be applied to detect fast neutrons [103]. However, since the neutron reaction cross section ¹²C and ²⁸Si is small, so SiC neutron detector generally uses solid neutron conversion layer material, such as ¹⁰B, ⁶Li and its composite materials, which convert neutrons into charged particles for detection. At the same time, appropriate transfer layer thickness should be simulated to achieve the best detection efficiency. Usually SiC detectors use a flat structure, only part of the charged particles incident into the semiconductor is detected, so the efficiency is very low, usually no more than 5% [104].

Due to the deep penetration of X and γ rays, a thickness sensitive zone of several hundred microns is needed. The device with epitaxial thin sensitive zone can only detect low energy and X and γ-rays [105-107]. In order to realize high-energy ray detection, the quality of the bulk single crystal used needs to be improved, and the energy spectrum test is difficult. However, in synchrotron radiation, SiC detectors are also used in X-ray energy detection and beam monitoring [108]. The semi-insulating bulk 4H-SiC detector shows moderate resolution of γ ray spectra, and the 4H-SiC n-type epitaxial layer detector shows good resolution of 59.5 keV γ-ray spectra.

In addition, with the development of device technology, the 3D structure which can be used for ultra-fast detection and the micro-structure silicon carbide detector which can be used for efficient neutron detection have also become an inevitable demand [109]. These studies are bound to facilitate ultrafast radiation testing with silicon carbide detectors at high fluence [110], in-core monitoring [111], neutron monitoring of space nuclear reactor power supply, fusion [97], free electron laser [112] and so on.

The GaN Detector

Gallium nitride is a direct bandgap semiconductor with a hexagonal wurzite structure (α-GaN). GaN thin films are usually grown on non-primary substrates, i.e., heteroepitaxy. Considering the lattice constant, crystal structure, chemical, thermal and electrical properties of the material, silicon, sapphire (Al₂O₃) and SiC are the most popular substrates [113]. However, due to the mismatch between lattice and thermal expansion coefficient, there is still a high dislocation density in GaN on heterogeneous substrate (10⁸ – 10¹⁰ cm^-2).

GaN thin film growth techniques are varied. The most used techniques are metal organic chemical vapor deposition (MOCVD) and molecular beam epitaxy (MBE). MBE has the advantages of precise process control, low growth temperature and low working pressure. However, the growth rate of GaN in MBE is relatively slow and the operation cost is high [114], so it is mostly confined to laboratory research. MOCVD, especially in combination with the epitaxial transverse overgrowth technique (ELOG), can produce high quality GaN films, for example, in some studies, the linear dislocation density can be reduced to 10⁵ –10⁷ cm^-2 [115, 116]. In addition, the high throughput capability of MOCVD also determines its extensive application in commercial large-scale GaN epitaxy growth [114]. At the same time, hydride vapor phase epitaxy (HVPE) is the main technology for the production of thick GaN wafers due to its high growth rate (hundreds of microns per hour) [116]. The dislocation density decreases with the increase of GaN thickness [117]. Therefore, the thick GaN grown by HVPE has higher crystal mass and lower impurity concentration; for example, Xu et al. were able to produce less than 1 × 10⁶ cm^-2 of thick GaN dislocation density. However, HVPE is difficult to achieve controlled film doping and low background carrier concentration (about 10¹⁴/cm³) without compensation, and is rarely used to prepare GaN pin thin film devices.

GaN doping has a certain effect on the performance of the detector [118]. Undoped GaN usually exhibits n-type characteristics due to unintentional background doping, such as silicon and oxygen. These impurities can be released from the reaction chamber, usually produced by quartz, and act as shallow donors in GaN [116]. In addition, some donor point defects in GaN, such as nitrogen vacancies, are also responsible for the n-type conductivity of undoped GaN [116, 119].

The first GaN α-particles realized by Vaitkus et al. are based on this device structure[120]. MOCVD method was used in Al₂O₃ The N-GaN epitaxial layer grows 2 µm continuously on the surface, and Au acts as the Schottky contact surface, forming a double Schottky structure. The detector is detected in the reverse bias range of 0 ∼ 28 V (breakdown voltage is 29 V). The charge collection efficiency (CCE) of 5.48 MeV alpha particles emitted by ²¹⁴AM source was measured to close to 100%, as shown in Fig 27. This work proves that GaN devices can be used in particle detection. The double Schottky structure is simple to manufacture and does not need to optimize the ohmic contact, but the long distance between the two Schottky contacts leads to parasitic capacitance and resistance affecting the performance of the detector [121].

Fig. 27Full-screen View

Energy spectra of double Schottky GaN detectors at different bias voltages [121].

Compared with GaN Schottky devices, PIN devices have more lower leakage current. Lu et al. applied this structure to α particle detection in 2012[122]. At -30 V operating voltage, the reverse leakage current of the detector is in the nA range, and the pulse height spectrum of ²⁴¹Am alpha particles is measured. The CCE at this voltage is about 80%, but the energy resolution (FWHM) is about 40%. The authors attribute this high energy resolution to the thick dead layer and the absence of a collimator. In addition, Zhu et al. studied the radiation performance of GaN α particle detector with temperature variation based on the thin film p-i-n structure [123]. As the temperature changes from 290 to 490 K, the peak position of ²⁴¹Am Alpha particle pulse height spectrum is slightly shifted towards the lower energy direction, and the change of half peak width is almost negligible, as shown in Fig. 28. This work proves that GaN-based alpha particle detector has good stability in high temperature environment.

Fig. 28Full-screen View

(Color online) Energy spectra of α particles measured by thin-film GaN PIN detectors at different temperatures [123].

Grant et al. proposed the application of bulk GaN to nuclear radiation detection in 2005 [124]. The bulk GaN Schottky structure detector developed by Xu et al. has the lower leakage current (7.53 ± 0.3 nA at -200 V) and the highest operating voltage (550 V) among all GaN α-particle detectors reported. Due to the high operating voltage and low background carrier concentration, the depletion width can reach 27 mm at -550 V, and the energy resolution can be improved to 2.2% [125], as shown in Fig. 29. Sandupatla et al. also applied this structure to particle detection. Metal-organic gas phase epitaxy (MOVPE) method was used to grow the sensitive zone with a thickness of 15 μm, compensated by Mg impuries, and the carrier concentration was as low as 7 × 10¹⁴ cm^-3. High CCE (65%) can be obtained at low voltage (-20 V) and also observed at -300 V (96.7%) [126].

Fig. 29Full-screen View

(Color online) Energy resolution and peak position of gallium nitride schottky detector at different reverse voltages[125].

Recently, Liang et al. have done a lot of work on sapphire based GaN pin detectors, and simulated and optimized the performance of GaN pin devices [127, 127].GaN pin detectors with 10 μm and 20 μm thickness sensitive zones were realized by means of isoelectron doping process. The leakage of 10 μm detector at reverse 40 V is 10 pA, and the leakage at reverse 200 V bias is only 10 nA. using the hybrid ²⁴¹Am and ²³⁹Pu α source was tested, and the energy resolution reached about 3%. The actual device is shown in Fig. 30.

Fig. 30Full-screen View

(Color online) GaN pin detector with 10 μm sensitive zone on sapphire substrate [128].

It has also been noted that GaN detectors can be used for X-ray detection because of their high radiation resistance and insensitivity to visible light. Duboz et al. firstly studied the absorption coefficient of GaN to 6 ∼ 40 keV photon energy. The large absorption coefficients observed between 10 keV and 20 keV indicate that GaN is difficult to use for detecting X-rays with photon energies greater than 20 keV, except that GaN layers are very thick, such as GaN substrates [129]. Due to the thin epitaxial layer, X-ray spectroscopy measurement based on GaN detectors still has great challenges.

The application of GaN in particle detection shows its great potential in the field of neutron detection. However, research on GaN neutron detectors is still in its infancy. Detecting neutrons with GaN detector is mainly to cover the detector with neutron transfer layer material, and to detect neutrons indirectly by detecting the α particles emitted after the interaction between neutrons and transfer layer. The commonly used neutron transfer layer material is containing ⁶Li and ¹⁰B of material. When a neutron is absorbed by the transfer layer, two charged particles are produced, thus achieving neutron detection. However, the detection efficiency of this method is still affected by the long neutron mean free path and the neutron mean free path [121] Limits on the short distances at which charged particles are produced in a detector. Gd (thermal neutron capture cross section 49000 Barns) is also used to doping GaN in order to obtain high detection efficiency [130, 131].

The Diamond Detector

Diamond detectors have a lot of research in radiation detection, especially detection of alpha particles, X-rays and neutrons. In terms of energy resolution, Iwona Wodniak et al. used 2.5 mm × 2.5 mm × 50 μm diamond single crystals to make X-ray detectors with a maximum half-width of 21 keV and a minimum energy resolution of 0.36%, comparable to that of silicon detectors [132]. A. V. Krasilnikov et al. developed three kinds of spectral determination systems based on natural diamond detectors, and measured D-T neutron spectrum and flux on the tokamak fusion test reactor [133]. D-T neutrons pass through ¹²C (n, α)⁹Be reaction interacts with the natural diamond detectors reaction and produces a narrow peak in the pulse height distribution, with an energy resolution of 2% ∼ 3%, which is well isolated ∼ 2 MeV from other reactions. L. Giacomelli et al. show that the energy resolution of 20 MeV incident neutrons is 1%, and the detection efficiency is 1%, which opens up a new prospect for the study of fast ion physics in high-energy fusion devices, such as the S-D neutron spectrum of deuterium-tritium plasma heated by neutral beam injection [134].

In recent years, diamond thin film detectors, especially CVD diamond polycrystalline and single crystalline thin film detectors, have been widely studied. Due to the difficulty of preparation and high cost, there are few reports on the research of high quality thick single crystal diamond. For low energy α-rays and X-rays, complete deposition can be achieved in diamond films of tens of micron. However, for the detection of high-energy rays, such as X-rays of 10 keV, diamond films of dozens of microns can only deposit about 3% of their total energy, and even diamonds of 200 ∼ 300 μm thickness can only deposit about 15%.

The diamond nuclear radiation detector has good energy resolution. In a large energy range, the amplitude of the pulse signal is proportional to the energy of the incident particle. The high radiation resistance makes the detector particularly suitable for neutron measurement; Diamond material detector has a long lifetime, especially suitable for strong radiation environment particle testing. The main problem of diamond detector is the material preparation of large size and low cost. On this basis, further studies are needed on n-type doping, Ohmic and Schottky contacts and micromachining processes. In addition, in the aspect of signal processing, how to identify particles is also a problem.

The Ga₂O₃ And BN Detector

Compared with the SiC, Ga₂O₃ and BN detectors are still relatively immature. This is mainly due to the lack of breakthrough in the growth technology of Ga₂O₃ with high quality and low background carrier concentration. The advantages of both detectors are a wide bandgap and the ability to be largely free of interference except for the longer wavelength of the UVC, making it possible to work on the ground even in daylight. We have made some progress in the particle detector of Ga₂O₃ alpha detector, using HVPE technology to grow high-quality gallium oxide epitaxial layer, and realize the particle detector on the surface of the earth to resist light interference.

Hexagonal boron nitride (h-BN) is the simplest Group III nitride semiconductor material, with a ¹⁰B percentage content of about 20%, ¹⁰B (n, α) reaction cross-section in the thermal neutron energy region of about 3840 barns, theoretically can achieve 100% thermal neutron capture rate. Neutron detection can be realized by using charged particles generated by interaction with neutrons to excite electron hole pairs. It has many advantages such as high energy resolution, high charge collection rate and good sensitivity, and is the most ideal material for direct thermal neutron detection. h-BN is both a neutron conversion layer and a carrier collection area, which can directly capture neutrons and generate charge carriers in the layer, avoiding the self-absorption effect common in indirect conversion semiconductor detectors, and significantly improving detection efficiency and energy resolution. With very high resistivity, it can significantly reduce the dark current of the detector, improve the signal-to-noise ratio, and enhance the detection efficiency. The Q value of ¹⁰B (n, α) reaction reaches 2.79 MeV, and the outgoing α energy of the reaction reaches 1.7 MeV, which can achieve high energy resolution. Low density and atomic number, weak response to γ, can achieve high n/γ screening ability [135]; Due to its solid detector nature, it is easier to achieve sub-millimeter position resolution, which can improve the position resolution of existing detectors while maintaining high detection efficiency. Therefore, h-BN has significant material performance advantages for neutron detection, among which the sub-detector has many advantages such as fast response time, high sensitivity and high detection efficiency. In particular, h-BN has a wide bandgap (> 6 eV), high thermal conductivity, low thermal expansion coefficient, high breakdown field strength, radiation resistance, chemical and thermal stability, etc., which can ensure the stable operation of the detector in extreme environments and harsh conditions such as high temperature and strong radiation.

The average absorption length of natural h-BN for thermal neutrons is theoretically about 277 μm [136], and efficient neutron detection requires a sensitive region thick enough to ensure full collection of thermal neutrons. High melting point of h-BN and low vapor of boron make it difficult to obtain high quality and large size single crystal and epitaxial materials. At present, chemical vapor deposition (CVD) method is commonly used to grow heterogeneous epitaxy on sapphire substrate [137, 138], but the obtained epitaxy film thickness is generally between 2.5–15 μm, which does not meet the detector’s thickness requirements. After technical improvement, Jiang et al. from Texas Tech University in the United States used MOCVD to increase the thickness of the epitaxial film to 100 μm, and the detection efficiency of the self-supported film after stripping can reach 59%, which is the best international level at present [139]. However, due to the limitation of growth methods, it is difficult to obtain boron nitride films with larger thickness by MOCVD technology. In 2023, Z. Alemoush et al. proposed to use HVPE method to prepare h-BN thick film in order to overcome the growth problem of MOCVD thick film, but the thermal neutron detection efficiency of the prepared 100 µm thick boron nitride is about 20% [140]. Further using pure ¹⁰B to grow boron nitride is an important way to reduce the thickness of sensitive region and achieve efficient neutron detection.

Conclusion

Wide band semiconductor has important application potential in high temperature and radiation resistant nuclear radiation detection. Compared with silicon, wide bandgap semiconductors still have a lot of room for improvement in both material quality, thickness and device technology, but they have shown potential application in strong radiation fields, high-temperature particle and ray detection, and efficient neutron detection.

Main Structure And Readout Mode

The Time Projection Chamber (TPC), first proposed in the 1970s by researcher David R. Nygren [141] at Lawrence Berkeley Laboratory in the United States, is electronically readable gaseous detector that can directly provide high-resolution three-dimensional track information: x, y, and z coordinates are measured simultaneously along the track. The TPC, filled with gas, mainly comprises three parts: the field cage, the end cap detector, and the readout plane, as depicted in Fig. 31. The field cage consists of a series of electrodes, each connected to a separation resistance of a specific gradient potential, which creates a uniform and stable electric field (E) in its internal region and parallel to the magnetic field (B). When incident particles enter the cage, they ionize the gas medium within the TPC, generating an electron-ion pair. These original electrons drift uniformly toward the end cap detector. The end cap detector is mainly used for the multiplication of electrons. The electrons drift toward the signal readout plane after an avalanche at the end cap detector, which consists of a pad or strip. The fired position on the readout plane records the position of the particles in x and y coordinates. Measuring the drift time t of the electrons with electrons moving at constant velocity v (defined by electric field E of field cage) gives the position of the particle in z coordinate. In addition, the ionization density along the particle trajectory depends on the momentum and the type of the particle. The ionization energy loss (dE/dx), measured by the signal amplititude on each pad or strip, contributes to Particle identification (PID). By arranging the magnetic (B) and electric (E) fields in parallel within the drift volume, the curvature of the trajectory of a charged particle in a known magnetic field can be used to measure the momentum of the particle. Thus, the maximum magnetic field strength is obtained to the extent technically feasible and the gas pressure is kept at a feasible level, resulting in improved resolution by significantly reducing diffusion problems.

Fig. 31Full-screen View

(Color online) Schematic of the time projection chamber.

Initially, TPC primarily employed MWPC [142] as the end cap detector. The structure and equipotential lines of the MWPC are illustrated in Fig. 32. This wire chamber [143, 144] consists of two flat cathodes with anode metal wires equidistantly arranged in the middle plane, filled with mixed gas and operating in the proportional zone. The anode wire with positive voltage creates a strong electric field in its vicinity, where electron avalanche amplification only occurs near the wire, inducing a signal (one-dimensional position information of the incident particle) on the anode readout pad close to the filament. Traditional readout structures based on MWPC and pad, such as ALEPH TPC, DELPHI TPC, STAR TPC, and ALICE TPC, have been widely adopted. However, using MWPC as an end cap readout detector has some inevitable drawbacks.

Fig. 32Full-screen View

(Color online) (left) Schematic of MWPC. (right) equipotential lines in MWPC. Reproduced from [145].

Firstly, the directional nature of the anode wire affects the position resolution and the angle between the track and the wire, resulting in non-isotropic behavior. In addition, the electric and magnetic fields are no longer parallel around the wire, causing electrons to spread in the direction of E × B. It leads to a degradation of position resolution by approximately 600 μm. Furthermore, a positive ion feedback effect occurs, where ions generated by the avalanche cannot be immediate absorbed by the cathode in the drift region, leading to uneven distribution of the drift field. A gate plane is added above the cathode plane to mitigate ion feedback, albeit with a limited count rate. Moreover, due to process technology limitations, achieving wire spacing of less than 1 mm in MWPC is challenging, restricting the position fraction of MWPC to the order of millimeters. Finding new high-performance detectors capable of replacing MWPC is imperative.

With the increasing maturity of printed circuit board technology and microelectronic photoetching technology, Micro pattern structure gas detector (MPGD) [146], e.g., Micromegas and GEM have overcome the early-stage shortcomings of low gain and small sensitive areas while retaining the advantages of high counting rate and high position resolution. As a result, many experiments and studies have optimized and applied these detectors [147, 148].

Micromegas [149, 150] were developed by the French physicist Giomataris Yannis in 1996. The detector is a gas parallel plate detector with a narrow amplification space rely on a new thin metallic micromesh electrode design, as depicted in the Fig. 33. The micromesh structure divides the detector into a drift region of about 3 mm and an avalanche amplification region of only 100 μm. This asymmetrical design achieves a high electric (∼100 kV/cm) in the avalanche amplification region and a low electric field (∼1 kV/cm) in the drift region. As the incident particles pass through the Micromegas detector, energy will be deposited in the drift region and ionized electrons will be released. The resulting electron beam is finally collected by the anode plane of the split strip or pad at the end of the amplification region. Meanwhile, most positive ions drift to the wire mesh and are collected, while the electrons move towards the anode in a narrow funnel due to the high amplification region compared to the drift region. Significantly reducing diffusion and eliminating positive ions. Micromegas offers several benefits, including a high position resolution of ∼20 μm, a high-count rate of up to 10⁸ mm^-2s^-1 due to its short response time, and a high gain ranging from 10⁴ – 10⁶. Furthermore, it boasts a large detection area, stability in operation, good reliability and uniformity.

Fig. 33Full-screen View

(left) Conceptual design of Micromegas. (right) Electrical field lines. Reproduced from [151].

Fabio Sauli from CERN proposed the concept of the GEM [152] detector, which is manufactured by chemically etching small holes, with a diameter of 50–100 μm at high density, on a thin polymer foil coated with metal (e.g., copper). Figure 34 illustrates an electron microscope image of a typical GEM electrode cross-section and an electric field map of the electrode pore area. By applying a voltage of several hundred volts to the metal coating, a strong electric field can be generated within the hole. The electrons released in the upper region drift towards the pore and gain sufficient energy to ionize and collide with the gas molecules. Most of the electrons generated by the avalanche exit the multiplication zone and move to the lower part of the structure, where they can be collected by the electrodes or injected into a second multiplication zone. Compared to wire chambers, GEM technology offers several advantages. It is easier to support, eliminates wire drooping and instability, and can be positioned close to the readout pads, reducing diffusion after amplification. The high-density environment of high-altitude caves ensures uniform amplification over a large area. Additionally, the positive ions generated in the avalanche naturally drift away from the amplification area, preventing the accumulation of space charges. Detector stability and inhibition of positive ion feedback, gain multiplication can be obtained by cascading two or three layers on top of each other. As a result, GEMs and Micromegas are getting popular for TPC [153-155].

Fig. 34Full-screen View

(left) Electron microscope picture of a section of typical GEM electrode. (right) Electric field in the region of the holes of a GEM electrode. Reproduced from [156].

The dimensions of the amplification structure can be reduced from millimeters to about 10 μm by utilizing the micro-structure of the readout anode. However, the pad size of the readout plane does not match the high granularity due to limitations on the integration level of readout electronics. To address this, using silicon pixel sensors with a pixel size of a few microns to read MPGDs is under research. In the first study, combining the Medipix2 chip with a triple GEM stack or Micromegas, a single electron with an efficiency of 90% can be detected [158]. The recent GridPix [159], as shown in Fig. 35 integrates micro-gas amplification and high-density pixelation of ASIC Timepix [160]. With the photolithographic post-processing techniques, GridPix achieves high accuracy in hole position, size, and gap size. As a result, GridPix can effectively detect and separate single primary electrons[161]. The IMPix[81] series pixel sensor, similar to Timepix, can simultaneously measure the charge’s energy, arrival time, and position. Figure 36 is the microscopic photographs of IMPix-S1. The Topmetal series pixel sensor [162, 163] can also collect charge from the sursroundisng media through the exposed part of its topmost metal layer. Figure 37 shows a photo of the Topmetal-II-[164] bonded to a PCB base with gold bonding wire. Both sensors are promising candidates for future TPC applications.

Fig. 35Full-screen View

Scanning Electron Microscope picture of a GridPix detector. The aluminum grid is partially removed for better visibility. Reproduced from [157].

Fig. 36Full-screen View

Microscopic photograph of IMPix-S1. Reproduced from [81].

Fig. 37Full-screen View

(Color online) A photograph of the Topmetal-II- bonded to the PCB pedestal with gold bonding wire. Part of the pixel matrix is zoomed in, and the rest includes peripheral circuitry. Reproduced from [163].

TPC Developments And Frontiers

The first experiment in particle physics using a TPC was the PEP (Fig. 38) experiment at the Stanford Collider in Berkeley Laboratory for the e⁺ e^- PEP-4 collider [165]. The main tracking device in the PEP-4 is a TPC [166] with a drift cavity measuring two times 1 m in length with a diameter of 2 m, as shown in Fig. 39. It consists of a gas-filled sensitive volume, usually with a central cathode that divides the volume into two identical halves. Each side has an anode with a readout system. It is filled with an 80% argon (Ar) and 20% methane (CH₄) mixed gas at 8.5 atmospheres and is put in a solenoid magnetic field of 0.32 Tesla. The central membrane of the TPC contains a series of equipotential rings at a voltage of -75 kV, and the position ring at the inner and outer radius of the TPC generates an axially uniform electric field in both parts of the drift chamber. The length of electrons drifting to the two readout planes at each end is 1 meter. As charged particles pass through the cylinder, they ionize the gas along their trajectory, and the resulting ionized electrons drift parallel to the cylinder axis to the end cap sector. Each end cap contains six MWPC sectors for detecting drift ionization, with each sector having 183 anode wires spaced radially by 4 mm. The process of proportional amplification at the wire induces signals on the 2–3 cathode pads nearby [167]. All anode wire and cathode pad signals are sampled through the Charge-Coupled Device at 10 MHz [168] to provide 3-D information (along the drift direction). The Charge-Coupled Device is an analogue shift register capable of storing associated time and pulse-height information [165]. The Charge-Coupled Device record 455 pulse height information and read it at approximately 20 kHz. Thus, for a track traversing the entire radius, there are 15 points of 3D information and 183 (dE/dx) samples for PID. In 1982, this TPC gathered data from the SLAC PEP storage ring and studied 29 GeV e⁺ e^- collisions. When the TPC with the sense-wire spacing 0.4 cm filled with 80% argon (Ar) and 20% methane (CH₄) at 8.5 atmospheres pressure, the dE/dx resolution of 6.9% [169] FWHM is reached for isolated tracks, and the tracks in particle jets are measured with a resolution of 8.3% [169] FWHM. It achieves the momentum resolution of about 1.0 p% [165], and the excellent PID was achieved with a resolution of 3.0% [170] for minimum ionizing particles and 2.7% [170] for Bhabha electrons, the excellent value of TPC allowing for the recognition of pions, kaons, and protons in most kinematics.

Fig. 38Full-screen View

(Color online) The PEP facility at SLAC (PEP-4). Reproduced from [165].

Fig. 39Full-screen View

(Color online) Schematic of PEP-4 TPC structure. Reproduced from [165].

For the same physics, the TOPAZ TPC [171-173] of KEK TRISTAN was proposed in 1986, and the ALEPH TPC [174, 175] and the DELPHI TPC [176] in 1990 and in 1992, respectively, for the LEP at CERN. These TPCs mentioned were used for particle identification and track reconstruction of high-energy particles, capable of handling low event rates with multiplicities of 2–30 charged tracks.

In the early 1980s, as relativistic heavy ion collision research developed, the TPC began to be used in heavy ion experiments, leading to the necessity of addressing higher particle density. The first TPC used to investigate high-energy events in heavy ion physics experiments was the Equation-of-State (EOS) of the Bevalac accelerator Heavy Ion Spectrometer System (HISS) experiment at Lawrence Berkeley Laboratory (LBL) in the United States. The HISS experiment is a target experiment. At the Bevalac, the EOS TPC [177, 178] is greatly expand detector capabilities. It is enable the complete measurement of most charged particles emitted from heavy ion collisions. 3-D tracking will allow the analysis of high multiplicity events with up to 200 charged particles [177]. Accurate tracking resolution in the HISS dipole field and dE/dx information provide momenta and particle identification for the majority of the charged particles of interest. With other components of the HISS system, EOS can observe and identify nearly all charged particles generated in the collisions. LBL designed a 4-channel integrated preamplifier [179] specifically with CMOS technology for EOS TPC. This low-noise, high-dynamic range preamplifier has a switching current source capable of resetting the integral capacitor at the end of each TPC readout cycle. Stuart Kleinfelder also developed a switching capacitor array (SCA) [180, 181] for analog signal sampling and storage for the project. This approach reduces the cost and power consumption of the waveform digitizer, making TPC a practical tool for heavy-ion collision research. Nearly all electronic devices of EOS TPC are located on the detector, a unique design feature that significantly reduces the number of cables leading outside and permits the entire detector to be installed in Bevalac’s HISS magnet. EOS, using full pad coverage, achieved a spatial resolution of 300 μm [170]. When it was filled with P10 gas (10% CH₄, 90% Ar) at 1 atmosphere pressure, 128 dE⁄dx samples for PID and the powerful mass separation by dE/dx is demonstrated in [182]. Furthermore, the detector represents an important step towards complete integration and miniaturization of the electronic equipment required by the detector.

Subsequently, Brookhaven’s BNL 810 [183] and CERN’s NA35 [184], NA36, and NA49 [185, 186] experiments all utilized TPC to investigate heavy ion collision experiments. The NA49 advanced the development of these devices by deploying four large TPCs, with the largest pair measuring 3.8 m × 3.8 m × 1.3 m. Two Main TPCs (MTPCL. MTPCR), are situated downstream of the magnets, and two Vertex TPCs (VTPCI. VTPC2), are situated within large dipole magnets [187]. A total of 182,000 SCAs are utilized for readout, which integrates ADCs. At Brookhaven’s RHIC STAR experiment, which operates at higher collision energies of heavy ions, the TPC [188, 189] has a diameter of 4 meters and a length of 4.2 meters. The STAR detector structure is illustrated in Fig. 40. It is filled with P10 gas (10% methane, 90% argon) regulated at 2 mbar above atmospheric pressure [190]. This gas has long been used in TPCs [188]. The STAR TPC achieves a resolution of 350 µm, and an energy loss resolution of 8% for a track that crosses 40 pad-rows with the magnetic field of 0.25 T [188].

Fig. 40Full-screen View

(Color online) Perspective view of the STAR detector, with a cutaway for viewing inner detector systems. A TPC lies at the heart of STAR. Reproduced from [191].

The primary objective of the ALICE experiment at the CERN Large Hadron Collider (LHC) is to investigate Pb-Pb collisions at a center-of-mass energy of 5.5 TeV per nucleon pair. Its TPC [192] aims to handle the exceptionally high charged particle multiplicity resulting from central Pb-Pb collisions at LHC energy levels. It provides charged particle momentum measurement with sufficient momentum resolution, particle identification by dE/dx and by decay topology analysis, and vertex determination in the region pt < 10 GeV/c and pseudorapidities |η| > 0.9. Figure 41 shows a cut-away view of the ALICE detector and a schematic of ALICE TPC. The ALICE TPC, largest of its kind, is a 90m³ cylinder filled with a 90% Ne - 10% CO₂ - 5% N₂ gas mixture, suited in a 0.5 Tesla solenoid magnetic field. It is divided into two drift zones by the central electrode along its axis, with the field cage ensuring a uniform electric field along the z-axis. The readout plates at two ends are based on the MWPC technique with a total of ∼ 557,568 readout pads. Each readout plate is divided into 18 sectors, each housing an internal readout room (IROC) and an external readout room (OROC). The PASA (preamplifier shaper chip) [193] and ALTRO (ALICE TPC readout chip [194, 195] constitute the ALICE front-end electronics for LHC Run1. Each pad connects to the one of the 16 channels in the PASA chip. Each ALTRO consists of 16 × 10 bit ADCs and the digital logic realizing programmable digital baseline subtraction, tail elimination filter, zero suppression, and multi-event buffer. The use of CMOS ASIC in ALICE TPC results in an extremely compact readout electronics with reduced power consumption, enabling a large number of electronic channels to be positioned close to the detector, thus creating a highly compact and cost-effective system. As the main tracking device of ALICE, this TPC provides the dE/dx resolution ranges between 5–8% depending on the track inclination angle and drift distance, the energy loss, and the centrality in p-Pb and Pb-Pb collisions due to the differing detector occupancy [196].

Fig. 41Full-screen View

(Color online) (Up) Cut-away view of the ALICE detector. (Down) Schematic of ALICE TPC. Reproduced from [197] and [198].

In LHC Run1, the readout electronics suffered from radiation-induced issues. Thus, in the first long shutdown, the TPC replaced its 216 Readout Control Units [199, 200] to handle the increased radiation load and readout speed in Run2. It was the first time that Flash-based FPGA was employed in physics experiments. Then, in the second long shutdown of the LHC, the ALICE upgraded its TPC detector to accommodate the 50 kHz Pb-Pb collisions in Run3. The upgraded TPC employs GEM technology and continuous readout [201, 202]. A new customized mixed-signal front-end chip called SAMPA will process the GEM signals [203, 204]. The SAMPA chip employs 130 nm CMOS technology with a rated voltage of 1.25 V, featuring 32 channels with optional input polarity and five possible combinations of forming time and sensitivity. Each channel includes a charge-sensitive amplifier, a semi-Gaussian shaper, and a 10-bit ADC, followed by a digital signal processor.

Figure 42 shows the structure of the CSR External-target Experiment (CEE) [207], a spectrometer for research on the properties of nuclear matter at high baryon density regions, which will start physics runs in 2025. The CEE spectrometer [208] includes a TPC [209] at the center of a uniform magnetic field created by a large acceptance superconducting dipole to measure the trajectories of the light-charged particles. The sensitive volume of this TPC is 0.9 m (length) × 0.8 m (height) × 1 m (width), filled with the gas mixture of 90% Ar and 10% CO₂ at atmospheric pressure. A large-area GEM foil will amplify the signal to replace the traditional anode wires and address space charge issues arising from positively charged ions drifting back into the sensitive volume. The pad size of the readout will designed as about 5 mm × 12 mm. The total number of channels will be about 15000. These readout pads are connected to the SAMPA ASICs, enabling the CEE-TPC to handle data at an event rate of 10 kHz. This CEE TPC is still under development and commission. The prototype readout electronics provide an energy resolution better than 10% with an input charge of 1.4 fC. Preliminary cosmic ray measurements on the engineering prototype of this TPC with readout electronics show a position resolution of approximately 462 μm (σxz = 461.6 μm) [210].

Fig. 42Full-screen View

(Color online) (Up) Overall layout of the HIRFL-CSR complex. (Down) The concept design of the CSR External target Experiment (CEE). Reproduced from [205] and [206].

Small-sized TPCs have a broad range of applications in diverse nuclear physics experiments [211]. The ACtive TARget and Time Projection Chamber (ACTAR TPC) [212-214] at GANIL aim to study the excitation functions with very exotic nuclei, to perform spectroscopic studies beyond the drip line, and the excitation of very collective states such as giant resonances or cluster states. Figure 43 shows the structure of The ACTAR TPC. The active volume is 295 mm × 295 mm × 255 mm encased by a two-wire drift cage connected to the cathode powered via a 20 kV HV feedthrough. The drift cage comprises 20 µm diameter wires with 1 mm spacing in the inner planes and 2 mm spacing in the outer planes, connected by 4.7 MΩ resistors. In the presence of a drifting electric field, ionized electrons generated in the active zone drift toward the pad plane, which is highly segmented into 128 × 128 square pads, each with a side length of 2 mm. The resulting high density of channels (25 channels per square centimeter, for a total of 16,384 pads) was a major challenge in designing the mechanical and pad connections, as well as ensuring that the mechanical deformation of the flange was minimized when applying differential pressures of up to 1 bar. The TPC was successfully produced two gain regions with an incident beam intensity of 20 kHz while kept the capablily of charge measurement. The ¹H(¹⁸O, ¹⁸O)¹H and ¹H(¹⁸O, ¹⁵N)⁴He channels were open and could be distinguished with the scattered heavy-ion identified as being in its ground state [213]. The excitation functions in both channels were reconstructed and fit to obtain the center-of-mass energy resolutions. Resolutions of 38(4) keV FWHM for the (p, p) channel and 54(9) keV FWHM for the (p, α) channel were achieved [213].

Fig. 43Full-screen View

(Color online) A 3D drawing of the ACTAR TPC. Reproduced from [213].

The ACTAR TPC utilizes the General Electronics for TPCs (GET) as a data acquisition system [217, 218]. Figure 44 shows a GET system with a single Micro TCA chassis, which mainly consists of the front-end AsAd (ASIC and ADC) board [216] and the back-end CoBo (concentration board) board. Each AsAd board contains 4 AGET chips consisting of an FPGA (ACTEL ProASIC3E model A3PE1500) and a 12-bit 4-channel ADC. Each AGET chip has 64 input signal channels and 4 fixed-pattern noise (FPN) channels [218]. The AsAd board processes the detector output signals with the AGET [215]. Then, it digitizes the sampled values stored in the analog memory of the AGET SCA with the ADC with a sampling clock of 25 MHz synchronized to the SCA [215, 216]. The longest processing time for reading out the 68 AGET channels is 1.44 ms [216]. The CoBo is a Micro TCA-compatible module (custom PCB and firmware package) developed for the GET system using the Xilinx Virtex-5 System-on-Chip, which contains an FPGA and dual PowerPC440 CPU cores. The CoBo board receives the digital data from the AsAd through the VHDCI connections at a maximum data rate of approximately 1.2 Gb/s [215], processes it, and transfers it to the storage field. Each Cobo can handle up to 4 AsAds boards with 1024 detector signal channels [216]. The MUTANT module (MUltiplicity ANd Trigger) synchronizes the boards and handles the triggers and timestamps. Given the generic approach, the system has been used in many nuclear, particle, subsurface, and medical physics applications.

Fig. 44Full-screen View

(Color online) (Up) Overview of the GET system. (Down) Photograph of a single AsAd board. Reproduced from [215] and [216].

Plenty of small-scale TPCs have been built across the world. The TexAT (Texas Active Target) detector [220-222] was constructed at the Cyclotron Institute Texas AM University for nuclear structure and astrophysics experiments with rare isotope beams [219]. Figure 45 shows the structure of the TexAT. It is designed for general use in nuclear structure and astrophysics experiments with rare isotope beams. It combines a highly segmented TPC with a two-layer solid-state detector, offering high accuracy and flexibility for experiments with low-intensity exotic beams and enabling 3D trajectory reconstruction of incoming and outgoing particles involved in nuclear reactions and decay. The MATE (Multipurpose time projection chamber for nuclear AsTrophysical and Exotic beam experiments) at HIRFL studies the reactions at stellar energies [223, 224]. Figure 46 hows the schematic diagram of the MATE-TPC with a gating grid. MATE possesses an active volume of 300 mm × 300 mm × 200 mm [224] in the form of a triple-wire field cage and a pair of thick GEMs mounted on top of the readout plate with 4000 channels. The versatile multipurpose time projection chamber (MTPC), with a cylindrical sensitive volume with a 140 mm diameter, has been proposed at Back-n [225], designed primarily for light-charged particle emission reactions but also adaptable for other measurement types such as fission reactions, neutron beam profiling, neutron imaging, etc. The 1519 [225] signal channels are read out with micromegas connected to a scalable readout system. The GET system reads the signals in TexAT and the MATE, and the MTPC readout system is developed with AGET ASIC.

Fig. 45Full-screen View

(Color online) (Up) Sketch of the TexAT Assembly. (Down) 3D-drawing of TexAT scattering chamber. Reproduced from [219].

Fig. 46Full-screen View

(Color online) The schematic diagram of the MATE-TPC with a gating grid. Reproduced from [224].

In recent years, TPC has been introduced for studying rare physics, especially in searching dark matter and neutrino-less double beta decay (0νββ), due to the good performance in tracking capabilities, stability, radiopurity and scalability.

Weakly interacting massive particles (WIMPs) remain one of the most promising directions for dark matter research. Dual-phase (liquid and gas) TPCs are popular for dark matter search by detect the WIMP [227]. Figure 47 shows the principle of the dual-phase TPC. When a dark matter particle interacts with liquid argon or xeno [228, 229], it generates two signals: one from the prompt scintillation (S1) and the other one from ionized electrons (e^-). The electrons drifted into gas, creating a delayed secondary light signal (S2) by proportional scintillation. The S1 and S2 are detected by photomultiplier tubes (PMTs) installed at the top and bottom of the TPC. Combining the S2 signal observed on the top PMT and the time interval between the S1 and S2 reconstructs the three-dimensional coordinates of the interaction position. Background events are rejected by the ratio between S2 and S1 and the number of S2 signals. The XENON [230] is a project for direct dark matter search at the Laboratori Nazionali del Gran Sasso (LNGS). The XENONnT is shown in Fig. 48, the latest member in this project, has a cylindrical TPC, 148 cm long and 137 cm diameter, filled with 5.9 tons of liquid xenon. With an exposure of 20 ton × year, XENONnT is expected to reach the highest sensitivity of 1.4 × 10^-48 cm² for a WIMP of 50 GeV/c² [231], which is an order of magnitude below the limits set by the XENON1T [232].

Fig. 47Full-screen View

(Color online) The principle of the dual-phase TPC. Reproduced from [226].

Fig. 48Full-screen View

(Color online) CAD rendering of the XENONnT cryostat and TPC. Reproduced from [227].

The PandaX-4T experiment [233], located at China Jinping Underguoun Laboratory (CJPL) [234, 235], is the first multi-tonne liquid xenon detector operating in the world. Its large cylindrical TPC which is shown in Fig. 49, with a transverse diameter of 1.2 m and a height of 1.3 m, holds 3.7 tons of liquid xenon. It has 169 and 199 photo-multipliers (PMTs) on the top and bottom of the TPC, respectively. The data from PandX-4T set a stringent limit to the dark matter-nucleon spin-independent interactions, with the lowest excluded cross-section (90% CL) of 3.8 × 10^-47 cm² a WIMP of 40 GeV/c² [236].

Fig. 49Full-screen View

(Color online) Layout of the PandaX-4T detector. Reproduced from [237].

There have been several experiments aimed at searching for 0νββ in a number of different candidate isotopes [238] such as ¹³⁶Xe and ⁸²Se, which has also been suggested as a promising isotope in the search for 0νββ. The PandaX-III employ high-pressure Xe TPC and EXO-200 employ liquid Xe TPC looking for 0νββ in the ¹³⁶Xe. The NνDEx experiment utilizes a high-pressure Se TPC searching 0νββ in the ⁸²Se. The prototype experiment of the Enriched Xenon Observatory (EXO) EXO-200 currently operating at WIPP (Waste Isolation Pilot Plant) and using a TPC in a cylinder shape with 40 cm diameter and 44 cm length and contains ∼ 175 kg liquid Xe enriched to 80.6%[239]. Cutaway view of the EXO-200 setup as shown in Fig. 50. The TPC is held in a low-background cryostat system filled with high-purity HFE7000 fluid to keep the temperature of liquid xenon at 167 K and shield the radiation. To improve energy resolution, this TPC simultaneously records the ionization and scintillation signals in the liquid. The TPC cylinder contains two identical drifting sectors formed by a biased cathode grid in the middle. The end of each sector consists of two wire grids crossed at 60^∘ and an Large-area avalanche photodiodes (LAAPD) array. The incident particle ionizes the xenon atoms and generate electrons. The ionized electrons drift towards the wire grid, providing two-dimensional localization. Some xenon ions recombine with the electrons before they drift away, inducing excited states in the xenon atoms. Upon relaxation of the excited atoms, they emit ultraviolet light, referred to as scintillation, which the LAAPD captures. The time interval between the light and ionization signals reconstructs the (longitudinal) coordinate. In addition, combining the light and the ionization signal also provides better energy measurement. In August 2011, the EXO-200 first observed double beta decay in ¹³⁶Xe [240]. In 2014, the EXO-200 experiment set a lower limit on the 0νββ half-life in ¹³⁶Xe at 1.1×10²⁵ years at 90% CL [241]. The next-generation Enriched Xenon Observatory (nEXO), with 5000 kg of isotopically enriched liquid xenon, is expected to reach a half-life sensitivity of 1.35×10²⁸ yr at 90% confidence level in 10 years of data taking [242].

Fig. 50Full-screen View

(Color online) Cutaway view of the EXO-200 setup. Reproduced from [243].

The PandaX-III [244] (particle and astrophysical xenon experiment III) experiment uses gas TPC to search 0νββ in ¹³⁶Xe at the China Jin-Ping underground Laboratory II (CJPL-II). The schematic of the PandaX-III TPC is shown in Fig. 51. The gas TPC can reconstruct the 3D trajectory and the energy deposition for every single track, which significantly improves the background suppression. The first phase of PandaX-III uses a cylindrical TPC of 1.5 m in diameter and 2 m in length, filled with 140kg of 99% ¹³⁶Xe and 1% TMA (trimethylamine) at 10 bar [245]. Tine pitch micro-pattern gas detector (Microbulk Micromegas), employed at both ends of the TPC, incorporating a cathode in the middle, accomplishes the readout of the drift charge. The 140kg level PandaX-III experiment expects to achieve an energy resolution of 3% FWHM, a signal efficiency of 35%, reaching a half-life sensitivity of 0νββ (90% CL) of 8.5 × 10²⁵ years after three years of operation [245]. The second phase of the PandaX-III, in ton-scale, will consist of five TPCs in the same water tank, with improved energy resolution and background shielding [246].

Fig. 51Full-screen View

(Color online) Schematic of the PandaX-lIl TPC. Reproduced from [246].

As depicted in Figure 52, the NνDEx[247] (No neutrino double-beta-decay Experiment) uses high-pressure gas TPC with silicon pixel sensors-based readout to search 0νββ in ⁸²Se. If double beta decay happens in ⁸²Se, two emitted electrons ionize the gas, creating sinuous trails with a Bragg peak at each end. In 0νββ, the total energy of the two electrons is 2.996 MeV. Due to the electronegativity of Se, the rapid formation of negative ions between electrons and surrounding Se molecules makes avalanche amplification impossible, it is necessitating very low noise in the readout plane. The charge readout plane consists of a self-developed Topmetal [248] pixel sensor that directly collects and reads the initial ionized charge without amplification. This approach circumvents the fluctuations introduced by the traditional TPC avalanche amplification process. The specific design of the electric field structure achieves 100% charge collection efficiency. Hence, the NνDEx aims to reach 1% energy resolution near the Q value of 2.996 MeV. The time interval between different ions determines the drift distance. The NνDEx experiment can reconstruct each event’s trajectory and ionization deposition. The NνDEx-100, which is being built with 100 kg of SeF₆ gas, is scheduled to be installed at CJPL around the year 2025. In the future, the NνDEx will be scaled to a ton-scale.

Fig. 52Full-screen View

(Color online) Schematic design of the main part of the NνDEx-100 experiment. Reproduced from [247].

The rapid advancement of TPC has extended its significance to space exploration. The Low-Energy X-ray Polarization Detector (LPD) [249] is one of the payloads in the POLAR-2 experiment, designed as an external payload for the China Space Station (CSS) deployment in early 2024. Figure 53 illustrates the mechanical structure of the LPD. The LPD utilizes a charge-TPC structure with pixel sensors as its readout to observe the polarization of Gamma-Ray Bursts (GRBs) prompt emission in the energy range of 2-10 keV, with a wide field of view (FoV) of 90 degrees. The detector comprises 81 units, including a cathode, a gas microchannel plate (GMCP), a pixel readout chip (Topmetal-L), and a gas-sealed chamber.

Fig. 53Full-screen View

(Color online) Mechanical structures of LPD. (a) The LPD payload. (b) The detector array module. (c) The detector unit. Reproduced from [249].

The Electron-Tracking Compton Camera (ETCC) reconstructs the three-dimensional tracks of the scattered electron in the Compton process for both sub-MeV and MeV gamma rays. Figure 54 illustrates diagram of the ETCC. The ETCC typically comprises the gaseous electron tracker and the surrounding position-sensitive scintillators. Gamma-ray incident to the ETCC makes a Compton scattering in the gaseous electron tracker, which detects the scattering direction and energy of the Compton-recoil electron. The scintillators detect the absorption point and energy of the Compton-scattered gamma ray. The ETCC has been installed on the Sub-MeV gamma-ray Imaging Loaded-on-balloon Experiment (SMILE) series [250]. The ETCC on the SIMLE series uses a TPC with a micro-pixel chamber (μ-PIC) as the gas electron tracker. In the SMILE-2+ balloon, the 30 cm³ cubic TPC fills with the gas ratio of A_r: CF_A: iso-C₄H₁₀=95 : 3 : 2, and the position-sensitive detector is made of GSO scintillators.

Fig. 54Full-screen View

(Color online) The diagram of the ETCC. Reproduced from [251].

The silicon pixel sensor has shown great potential in real time beam monitoring. The researchers from IMP proposed the Hi’Beam detection systems [252-255] based on silicon pixel sensors designed to fulfill the beam monitoring requirements at the HIRFL and HIAF. Figure 55 illustrates the general principle of the Hi’Beam-SEE, which is composed of two detector systems with silicon pixel sensors acting as the anode. When charged particles pass through the detector field, they ionize the gas and produce electron-ion pairs. The anode then captures the charge in the presence of an electric field to reconstruct the trajectory of the particles. Two detector systems are positioned perpendicular to each other to reconstruct the 3-D profile of the beam. The HiBeam series includes the HiBeam-A [252, 253] for accelerators, the HiBeam-T [254] or physics terminals, and the HiBeam-SEE [256] for precise single event effects locating. The HiBeam-A prototype employs a microchannel plate (MCP) to amplify the rare ionized charge in the high vacuum chamber and one pixel sensor to collect the charge, demonstrating a spatial resolution of ∼ 5 μm. The HiBeam-T prototype features a gas chamber with P10 gas (composed of 90% Ar and 10% CH₄) and utilizes a gating grid to prevent the pixel sensors from becoming saturated with a high particle rate. The preliminary test results shows a resolution ∼ 30 μm [254]. The HiBeam-SEE reconstructs the trajectory of individual particles in the beam to infer their point of impact on the device under test. It uses air directly as the gas, primarily capturing the charge from the oxygen ion that combines the ionized charge. The first test with its prototype indicate a spatial resolusion better than 2 μm for Ta particle [255]. The HiBeam-A and HiBeam-T prototypes use the Topmetal-II sensor, while the HiBeam-SEE uses Topmetal-M. In the near future, the IMPix-S sensor will be integrated into the HiBeam system. With a similar principle as Hi’Beam, the HIRLF-CSR CEE experiments are building a silicon pixel sensor-based beam monitor, providing a spatial resolution better than 50 µm and a timing resolution better than 1 μs [257]. The first prototype of this beam monitor uses the readout based on the Topmetal-II- chip [258]. The second prototype is using the Topmetal-CEE pixel sensor for the readout [259].

Fig. 55Full-screen View

(Color online) The general principle of the Hi’Beam. Reproduced from [256].

The Time Projection Chamber (TPC) was initially employed as a tracking detector. It has since found widespread use in various particle physics experiments. As new scientific challenges emerge and technology and science continue to advance, the TPC has been continuously developed to enhance its capabilities in terms of high resolution, high count rate, and low noise. Consequently, it has found increasingly extensive applications in nuclear physics experiments, rare physics phenomena experiments, space detectors, etc.

Introduction

To reveal the physics in the high energy particle collisions, it is important for spectrometers to reconstruct and identify the secondary particles (i.e. Particle Identification, PID)[264]. Figure 56 shows the π/K separation powers of TOF systems given a fixed flight distance of 8 meters, where σ_TOF is the time resolution of the TOF system. Simply distancing the detectors is not practical since the coverage area and the cost rise quadratically. Thus, high time precision is crucial to recognize high momentum particles.

Fig. 56Full-screen View

(Color online) π/K separation power of TOF system with different time resolution, with a fixed flight distance of 8 m. Reproduced from [265].

However, timing performance is not the only consideration for building a TOF system. Due to the momentum measurement by the tracking detector, the spectrometer must be operated in high magnetic field, so the material and signal propagation of the TOF detectors must not be affected, especially for the detectors installed within the magnet. Moreover, the total active areas of the TOF detectors, in order to cover the phase space at several meters’ distance, are usually in tens of m², even over 100 m², making the cost of the detectors nonnegligible [266].

Since the 1990s when the novel structure Multigap Resistive Plate Chamber [267] (MRPC) was invented, it has been widely applied to many high energy physics experiments, such as BESIII [268], ALICE [268, 269], CBM [270], STAR [270, 271], and MPD [272]. Compared with other technologies such as fast scintillator with photomultipliers (PMT or SiPM), the competitive edge of the MRPC has been verified: the theoretic intrinsic time resolution is about 10 ps [273], and the best reported resolution is as low as 16 ps [274]; in large experiments MRPCs works stably with time precisions in 60 ps level and show good tolerance in magnetic fields; the components of the MRPCs are in low prices and can be easily produced in large area. Table 1 lists the representative high energy physics experiments around the world which apply MRPCs for time measurements.

Working Principle And Performance Of MRPC

MRPC is a gaseous detector with resistive plate electrodes and multigap structure. The common working gas for MRPCs is a mixture of Freon, iC₄H₁₀ and SF₆ with a fraction of 90/5/5. High electric field is applied between the gas gaps so that the primary ionization by an incident particle will immediately start a Townsend avalanche[279] which has an exponential grow-up of charge and a quench by the electronegative gas and space-charge effect. The usage of the resistive electrodes is to limit the voltage drop to a local area (mm² magnitude), ensuring the sensitivity for another region. The multigap structure results in the ordered gas gaps of hundred-micron level thickness. The narrow gap limits the avalanche spatially and allows MRPCs to be operated in even higher fields, e.g., above 10⁵ V/cm. As a result, the electron drift velocity v_d and effective Townsend coefficient α^* increases and the time resolution is improved significantly. The intrinsic time resolution of the MRPC is estimated by the following equation[280]: $\sigma \left(t\right)=\frac{1.28}{{\alpha }^{*}{v}_{\text{d}}}\mathrm{.}$ (5) Thanks to the innovative structure, it is not difficult for common MRPC TOF systems to reach a time resolution of better than 100 ps, while maintaining the high efficiency.

Figure 57 shows the structure of the MRPC. Usually, nylon fish lines with specific width are used to sustain the gas gaps. Carbon layers with surface resistivity in MΩ/$\square$ level are sprayed or pasted on the outermost electrodes for High Voltage (HV) application. It is noted that the resistive layers in the middle are electrically floating, meaning that no connection is needed for maintaining the potential of these electrodes. In principle, their potentials are stable in the electric field through a dynamically balanced charge neutralization between the electrode surfaces.

Fig. 57Full-screen View

(Color online) Structure diagram of MRPC.

When the electron-ion pairs are avalanched and drifted in the gas gaps, they will induce a current signal on to the readout electrodes which are usually laid on a PCB in shape of strips or pads. For the case of corresponding readout strips in Fig. 57, differential signals are transmitted to the electronics on both ends. If the readout strips are long with nonnegligible signal propagation time, both-end readout is essential for reconstructing both the hit time and position.

The behavior of current through the resistive electrodes constitutes the basic principle of the ’Ohmic Model’ [281], which is very useful to understand and evaluate the rate capability of an MRPC. The resistivity of the electrodes leads to a voltage drop ΔV_el given a current I which is determined by the incident flux Φ. It can be described as the following equation: $\Delta V_{\text{el}}=R_{\text{b}}I={\rho }_{\text{b}}\frac{2d}{s}\Phi S\langle q_{\text{aval}}\rangle \mathrm{,}$ (6) where ρ_b is the bulk resistivity of the electrode material, while $\langle q_{\text{aval}}\rangle$ is the average charge of an avalanche. The higher the counting rate, the higher the voltage drop on the electrode. If the rate exceeds the rate capability of the MRPC, the efficiency will decline. For most MRPCs which use glass as resistive electrodes, the rate capability constraints to no more than 1 kHz/cm². Using thin electrodes or heating the detector can improve the rate capability, but not in orders of magnitude. New experiments, with unprecedented luminosity and requirements of particle resolution, call for MRPCs with evolutional performances. Up to now, the development of MRPCs can be categorized into three generations [282]: the first generation MRPCs work in common situations with a counting rate of > 100 Hz/cm² and a time resolution of 80 ps level; the second generation MRPCs work in high counting rates of > 20 kHz/cm² and a time resolution about 80 ps; the third generation will be operated in an extremely high time precision of 20 ps level, as well as a high counting rate of > 20 kHz/cm². In the following section, we will describe the main challenges in the research and development on the next-generation MRPCs, while showing the typical representatives and applications.

Typical Applications Of MRPCs

One of the most successful applications of the first generation MRPCs is that on the STAR-TOF system [283]. A barrel-shape array of MRPC detectors surrounds the central Time Projection Chamber (TPC). The total active area for STAR-TOF reaches about 60 m², including 3840 MRPCs with 23040 readout channels. The STAR-TOF MRPC has 6 gas gaps of 0.22 mm thickness. The electrodes are made of float glasses of 0.7 mm thickness whose bulk resistivity is in 10¹² Ω cm level. Each MRPC contains 6 readout pads of 3.1 × 6.0 cm². The working gas is the Freon and SF₆ mixture in a fraction of 95/5.

The readout chain for STAR-TOF MRPCs is the NINO-based [284] front-end electronics (FEE) and the high-precision time-to-digital converter (HPTDC) [285]. NINO is an ASIC chip developed by the group in CERN for the amplification and discrimination of the differential MRPC signals. To realize the precise timing purpose, the designed LVDS output of discriminated signal has a rise time within 1 ns. The power consumption of NINO is as low as about 45 mW per channel. The overall time jitter of the STAR-TOF readout chain is better than 25 ps. NINO discriminates the time when a signal exceeds or descends to the threshold. The time-over-threshold (TOT) positive correlated to the signal amplitude or charge and is very useful to correct the time-walk caused by the amplitude difference. With time-slewing correction, the time resolution for STAR-TOF MRPCs reached 60 ps during the beam test.

The second generation MRPC, featured of an expanded rate capability over 20 kHz/cm², is developed for high-rate conditions. The interaction rate of the existing and proposed experiment has been increasing for better statistics and rare physics probe finding. The compressed baryonic matter spectrometer (CBM) [286] is one of these typical experiments for which the designed interaction rate will reach 10 MHz. As a fixed target experiment, CBM measures the products of heavy ion collisions with beam energies up to 11 GeV/u. Thus, even though the TOF wall is placed at 10 m from the target with a total active area of 120 m², the counting rate for MRPCs at the central region still reaches 30 kHz/cm², which is a huge challenge for MRPC. As equation (2) indicates, the most effective way to improve the rate capability is developing low resistivity novel materials for the resistive plates. Since 2008, the high energy physics detector group in Tsinghua University has been dedicated to the low-resistive glass. By many efforts of tests and optimizations on the components and techniques, the TUYK-LRG10 low-resistive glass has been successfully developed [284, 285], whose typical features are listed in Table 2. The real-size MRPC prototypes for CBM-TOF have been tested under the high intensity electron beam in Helmholtz-Zentrum Dresden-Rossendorf (HZDR), Germany, showing a capability as high as 70 kHz/cm² (with efficiency > 90%) and a time resolution better than 80 ps [287].

The CBM-TOF wall is divided into three regions according to the rate conditions [288]. The inner region is built with MRPC1a, 1b, 1c, operated under the rate condition between 5–30 kHz/cm². The MRPC2, developed by Tsinghua, is for the intermediate rate region with condition about 5 kHz/cm². The MRPC1 series and MRPC2 are assembled with low resistive glass. The MRPC3 and 4 take up the low rate region and are made up of thin float glass. Figure 58 shows the picture of the MRPC2 detector. It consists of double stacks with 8 gas gaps of 0.25 mm thickness, 32 readout strips of 27 × 1 cm². Each TOF module houses 5 MRPCs with PADI [289], the FEE board, plugged directly on the detectors. Then the discriminated signals are transmitted to the GET4-TDC [290]. 108 MRPC2 detectors were first installed on the endcap of the STAR thanks to the FAIR-Phase0 collaboration [191]. They have played an important role in expanding the phase space η from 1.1 to 1.6 for STAR during the BES-II runs [292].

Fig. 58Full-screen View

Picture of an MRPC2 detector. Reproduced from [265].

As high rate MRPCs being developed and tested, the gas related effect becomes an important issue for the stable and sustainable operation of the second generation MRPCs. Firstly, Freon and SF₆, as main components of the working gas, have serious greenhouse effects. The regulations to the greenhouse gases in EU, China and other countries may lead to uncertain situations on the availability and cost. There have been continuous efforts [288-290] on the substitutes of the MRPC working gas, such as HFO. However, it is also important to decrease the gas consumption especially for large area MRPC-TOF system. Secondly, gas pollution effects characterized by the increase of dark current and noise rate under persistent irradiation were observed at mini-CBM, another phase0 program of CBM dedicated for high rate tests. The gas pollution effect has been studies by means of both the simulation and test, and we find that the main cause is the accumulation of ionized gas pollutants in gas gaps due to the inefficient gas exchange [293]. Without a delicate structure for pressure-driven gas exchange, the resistance of thin gas gaps in traditional MRPCs significantly restrict the gas exchange to a diffusion level. The gas pollutants, if not expelled in time, may result in unexpected noise discharge in gas gaps or even the creepage along the fishline spacer. There have been observations showing the noise patterns and glass microscopic damages at the fishline region, as displayed in Fig. 59 [294].

Fig. 59Full-screen View

(Color online) (a) The noise hit position distribution (right) shows a pattern very correlated to the fihline layout (left). (b) Under the Scanning Electron Microscope, it is seen on the electrodes after the aging runs an influenced area within the fishline width (left) and a cracked area where fishlines contact with glasses(right). Reproduced from [294].

To solve this essential problem, a novel structure called the sealed MRPC has been developed [295]. For each sealing gas chamber, a 3D-printed sealing frame is spiced with the PMMA panels, as shown in Fig. 60(a). The gas tubes and fishline layout are carefully designed so that the gas can be flushed directly through the gas gaps. The sealed MRPC prototype has been tested under cosmic rays and X-rays, showing promising performances: the efficiency reaches 97%; the time resolution is around 60 ps at working point; the operational currents are constant at fixed rate conditions and decays in seconds to a low level after the irradiation.

Fig. 60Full-screen View

(Color online) (a) Structure of the sealed MRPC with fishline spacers. (b) Picture of the mylar spacer layout.Moreover, an ongoing effort is made to eliminate the effects of creepage along the spacers. Reproduced from [293, 294].

A prospective candidate of new spacer is the cylindrical mylar pads. The rectangular cross section may have minor distortion to the neighboring electric fields, and the resistivity of mylar is higher than the fishline. Figure 61(b) shows the layout of the mylar spacers pasted on the electrodes, it is estimated that the total contact area of spacers on the glass is a half of that of fishlines, implying that smaller dead zone is achieved at the same time. A first prototype has been tested, showing promising first results, see Fig. 36.

Fig. 61Full-screen View

(Color online) Current behavior of the mylar spacer sealed MRPC. The estimated counting rate is 7 kHz/cm². Left: current during the irradiation. Right: current relaxation after the irradiation. Reproduced from [294].

The experiments with low rate conditions can also benefit from the application of the sealed MRPC, since the gas consumption can be saved significantly. The external TOF system (eTOF) [296] of the CSR External-target Experiment (CEE) [208] in Lanzhou, China will be the first application of the sealed MRPC. For eTOF MRPCs, the outermost glass electrodes are glued with the sealing frame instead of the PMMA plate because the float glass is of higher mechanical strength. Performances of the real-size prototypes have been validated in cosmic and beam tests. The efficiency reaches 98% with a time resolution better than 60 ps [297]. It is very exciting to see in Fig. 62 that the prototype works in stable performance with a gas flow as low as 1 sccm, lasting for 21 days. It is estimated that no more than 70 sccm gas flow is needed for the 3.2×1.6 m² active area of the eTOF wall, which is a reduction to about 1/10 compared with the typical flow rate if using traditional MRPCs.

Fig. 62Full-screen View

(Color online) Long-term performance of the sealed MRPC for CEE-eTOF. Reproduced from [295].

The third generation MRPC is proposed for the future applications when the requirements are drawn to the ultimate limit of the detector physics. A typical example is the proposed SoLID [277] experiment at JLab, which is expected to realize high energy collisions between electrons and target for nucleus structure studies. The TOF system should be able to separate kaons and pions at 7 GeV/c moments, requiring a time resolution in 20 ps level. Besides, the rate condition is estimated as 20 kHz/cm². To reach a time precision less than 20 ps, both the time jitter of MRPC and the electronics must be less than 14 ps. With narrow gas gaps, MRPC can reach a 10 ps intrinsic resolution according to the simulation. However, time jitter of the NINO (PADIX) + HPTDC (GET4) used in the first and second generation TOF spectrometers is generally greater than 20 ps. Thus, a new generation technology of fast Switched Capacitor Array (SCA) chip based waveform sampling is proposed with typical bandwidth of 650 MHz and sampling rate over 5 GHz [298]. Diagram of this high resolution technology is shown in Fig. 63. With details of signal waveforms available, more information can be excavated for the reconstruction of time. For example, a study [274] has successfully outperformed the traditional time-slewing correction using the deep learning models, as shown in Fig. 64. It trained a model called ComLSTM using the simulated MRPC signals for the recognition of start time of the avalanche, and tested the model performance with test data. Structure of this model can be seen in Fig. 65. The simulation was carried out by a standalone MRPC simulation framework based on GEANT4. Two identical MRPC prototypes and the relevant electronics have been designed and produced. The MRPC has 32 gas gaps of 104 um thickness. Accordingly, the working point rises to about 150 kV/cm. Time resolution is analyzed by the ComLSTM model, showing result of 16.8 ps, see Fig. 66. Timing performance has fulfilled the requirements for both methods.

Fig. 63Full-screen View

(Color online) Diagram of the high resolution MRPC and readout electronics. Reproduced from [298, 299].

Fig. 64Full-screen View

(Color online) Performance analysis of the high resolution MRPC with simulated(left) and tested (right) signals. With simulation data, only the intrinsic time resolutions are included. Reproduced from [265, 274].

Fig. 65Full-screen View

(Color online) The network structure of the ComLSTM model used for MRPC time reconstruction. Reproduced from [274].

Fig. 66Full-screen View

(Color online) Time spectrum of MRPC in cosmic test analyzed with LSTM network. Reproduced from [274].

Summary

This work reviews the main characteristics and developments of MRPC for TOF systems. With the improvement of beam energy, intensity and the requirement of particle identification, performance of the MRPC has been developed from the initial ∼ 100 ps resolution and < 1 kHz/cm² rate capability, to a level of 16 ps and 70 kHz/cm². Great efforts have been made on the detector physics, materials, electronics, and time reconstruction methods. It is believed that MRPC with higher performances will continue to show up, promoting the application of MRPC to modern high energy physics experiments.

In addition to the research on the time resolution and the rate capability, there are other issues and frontiers that attract much attention in recent years:

1. Find ecological gas substitutes. As mentioned, the environmental impact from the operation of large MRPC-TOF systems must be eliminated. Related works have started for large experiments such as ATLAS and CMS [300, 301];

2. Application to industry and health care. With excellent time and position resolution, MRPC shows potential as futural detectors in technologies like PET [302, 303] and muongraphy [304];

3. Fast waveform sampling techniques. The readout chain with high precision waveform sampling is proved to be essential for reaching the extreme time resolution of MRPC. Nowadays SCA and FPGA technologies are the feasible solutions as next-generation electronics [305]. In the future new technologies and better performance may emerge with unremitting work.

Introduction of Electromagnetic Calorimeter

Electromagnetic Calorimeter [306] is a specialized and critical technique used to measure the energy of high energy particles in particle physics, such as electrons and photons, as they interact with and pass through a detector in high-energy physics experiments. The accurate determination of particle energies is fundamental to understanding the properties of particles, the dynamics of particle interactions, and the search for new physics phenomena beyond the Standard Model.

In the realm of particle physics, calorimeters are designed to capture and quantify the energy deposited by charged and neutral particles in the form of electromagnetic showers and hadronic cascades. These detectors play a central role in experiments conducted at particle accelerators like the Large Hadron Collider (LHC) [307] at CERN, Fermilab’s Tevatron [308], and other high-energy physics facilities worldwide.

Particle physics calorimeters are characterized by their high precision, large coverage, and ability to withstand extreme radiation and particle fluxes. They are typically composed of layers of dense and scintillating materials, which emit light when charged particles traverse them. The emitted light is then detected and converted into electrical signals for analysis.

Electromagnetic Interaction Of Particle With Material

The electromagnetic shower begins when a high-energy electron or gamma ray enters a dense material, such as a calorimeter detector or the Earth’s atmosphere, the electron primarily loses energy through Bremsstrahlung radiation, producing photons, as shown in Fig. 67. If the energy of these produced photons exceeds twice the rest energy of the incident electron 2m_ec², they can convert into electron-positron pairs through interactions with atomic nuclei. If the generated electron-positron pairs have energies greater than a critical energy E_c, in turn, they can undergo Bremsstrahlung radiation. As the depth of penetration increases, these processes iterate, and the initially high-energy electron or photon gradually transforms into numerous low-energy electrons and photons. This phenomenon is known as the electromagnetic shower process, an electromagnetic shower in GEANT4 is shown in Fig. 68.

Fig. 67Full-screen View

(Color online) Fractional energy lost per radiation length in lead as a function of electrons and positrons energy. Reproduced from 520

Fig. 68Full-screen View

(Color online) The electromagnetic shower development of a 3 GeV electron in CsI from the Geant4 simulation.

The critical energy E_c is defined as the energy of electron when energy loses by bremsstrahlung is equal to ionization loss. The critical energy E_c is approximately given by [309]: $E_{\text{c}}=\frac{610(710)\text{MeV}}{Z+1.24(0.92)}\text{solids}\left(\text{gases}\right)\mathrm{.}$ (7) In order to describe the shower longitudinal development ability of material, radiation length X₀ is introduced, which equal to the distance over which the energy of an electron decreases by a factor of 1/e due to bremsstrahlung radiation. The value of radiation length could be approximately calculated as [309]: $X_0=\frac{716\left({\text{g/cm}}^2\right)A}{Z\left(Z+1\right)\ln \left(287/\sqrt{Z}\right)}\mathrm{,}$ (8) where Z and A are the atomic number and weight of the material. In order to deposit most energy of high energy electrons and photons, the total radiation length of calorimeter should larger than 10X₀ ∼ 15X₀.

To describe the transverse spread of an electromagnetic shower in a calorimeter, which mainly due to multiple scattering of electrons and positrons, Molière radius is introduced. It’s defined as the radius of transverse size containing approximately 90% of the energy of the electromagnetic shower. R_M can be approximated by [309]: $R_{\text{M}}\left({\text{g/cm}}^2\right)=21\text{MeV}\cdot X_0/E_{\text{c}}\mathrm{.}$ (9) This calculation shows the transverse size is roughly energy independent, and R_M is related to the calorimeter material that implicitly embedded in X₀ and E_c. To reconstruct the particle hit position based on the shower shape, the cells of a segmented calorimeter must be comparable in size to one R_M or smaller.

The energy resolution (σE/E) is the key parameter of a calorimeter, and determines the accuracy of incident particle energy measurement. It improves with increase of energy, and could be written as [309]: ${\sigma }_E/E=a/\sqrt{E}\oplus b/\sqrt{E}\oplus c\mathrm{,}$ (10) where symbol ⊕ indicates a quadratic sum, the parameter a is the stochastic term, which is the main parameter of energy resolution and associated with statistical fluctuations [310], the parameter b is the noise term that comes from electric noise and not related to energy E, and the parameter c is the constant term due to energy leak, the contribution of which become more significant with increasing energy.

Techniques Of Calorimeter And Application In Experiments

There are two main techniques of calorimeter style: homogeneous calorimeter and sampling calorimeter. The homogeneous calorimeter has only one dense material that have both absorbing and sensitive property. The sampling calorimeter has at least two material in which the material that produces the particle shower as absorber is distinct from the material that measures the deposited energy. The generally used absorber is lead and iron, and the sensitive material is scintillator.

Each technique has its advantage and drawback. Since the homogeneous calorimeter measure all the energy deposited in calorimeter, which means low stochastic term a in Eq. (4), it has a better resolution than sampling calorimeter. However, the homogeneous calorimeters, such as inorganic crystals, are more expensive and constrained in shape and maximum size. On the other hand, sampling calorimeters are cost-effective and can be tailored easily to various shapes based on specific requirements, such longitudinal and lateral segmentation, and therefore they usually offer better space resolution and particle identification than homogeneous detectors.

Homogeneous Calorimeters Homogeneous calorimeters broadly divided into four types that characterized by its material [310]: semiconductor, Cherenkov, scintillation and Nobel-liquid calorimeter. These different types vary in their construction, sensitivity, and application, but the entire detection volume made of a uniform material ensure complete absorption and precise measurement of incident particle energies. Here we will introduce the undoped CsI scintillation calorimeter in Mu2e experiment as an example, and it has good energy resolution and fast response time that works for electron detection with high pile-up.

The Mu2e experiment [311-313] at Fermilab searches for the charged-lepton flavor violating (CLFV) [314] process of a muon-to-electron conversion in an ²⁷Al nucleus field: $\mu +N\left(Z\mathrm{,}A\right)\to e+N\left(Z\mathrm{,}A\right)\mathrm{.}$ (11) The calorimeter in the Mu2e experiment plays an important complementary role by providing particle recognition, a fast online triggering filter and a trajectory reconstruction seed. The requirement for the calorimeter is to have a high acceptance blue for the converted electron energy at 105 MeV, with a time resolution of better than 0.5 ns, an energy resolution of less than 10%, and a positional resolution of 1 cm.

The Mu2e calorimeter is comprised of an array of undoped CsI crystals organized into two annular disks (see Fig 69), with front-end electronics positioned on the rear of each disk, and voltage distribution, slow controls, and digitizer electronics mounted in crates. Each disk has an internal (external) radius of 374 mm (660 mm) and is filled with 674 (34 × 34 × 200) mm³ CsI crystals. Each CsI crystal [315] features a transverse dimension of 34 mm×34 mm and a length of 200 mm, and the readout is done by two large-area SiPMs. Notably, to capture the fast scintillation peaked at 310 nm in CsI, the calorimeter utilizes UV-extended SiPMs and inserts a band-pass filter to eliminate the slow scintillation light peaked at 450 nm. The front-end electronics (FEE) comprises two discrete and independent chips (Amp-HV), and provides the amplification and shaping stage, local linear regulation of the bias voltage, the monitoring of current and temperature on the sensors and pulse tests.

Fig. 69Full-screen View

(Color online) The Mu2e calorimeter[312].

The test results [312] of a large-sized prototype in the Mu2e calorimeter, dubbed as Module0, show that the time resolution at 100 MeV electrons is better than 120 ps, and at 22 MeV (energy deposited by a minimum ionizing particle in a CsI crystal) is 250 ps. The energy resolution of electrons is about 7%, mainly influenced by the shock leakage and beam energy spread.

Sampling Calorimeters The sampling calorimeters have typically have the energy resolution 5–20%/$\sqrt{E(\text{GeV})}$ [310]. For the experiments that energy resolution is not the key requirement, the sampling calorimeter will be a better choice for its cost-effective, segmentation design and good radiation hardness. The Shashlyk style sampling calorimeter has been developing for decades, and the Shashlik calorimeter in the KOPIO experiment will be described as an example for sampling calorimeter.

The KOPIO experiment [313-315] in Brookhaven National Laboratory (BNL), aimed at measuring the branching ratio of the very rare decay model [316] for K ${}_L^0\to {\pi }^{0}v\overline{v}$, places great demand on the calorimeter. The mechanical structure of a Shashlyk module (see Fig. 70) comprises a lead-scintllator sandwich with 300 layes where each layer consists of a 0.275 mm lead plate and 1.5 mm scintillator plate. The read out is by means of Wavelength-Shifting (WLS) [317] fibers passing through holes in the scintillator and lead, and collected by an APD.

Fig. 70Full-screen View

(Color online) The mechanical structure of a Shashlyk module [313].

The calorimeter applies the "LEGO"-type locks to improve module’s mechanical property. The lock maintains an air gap between scintillator and lead, and removes 600 paper tiles to reduce effective radiation length from 4 cm to 3.5 cm and the insensitive area from 2.5% to 1.6%. To improve light collection efficiency, a new BASF143E-based scintillator is used in the sampling calorimeter. This scintillator has a high light reflection efficiency of 97% from its surface. In addition, with the BASF143E-based scintillator, the effective light yield in the Shashlyk module (at the entrance to the photo-detector) is 60 photons per MeV.

Based on photon beam test [313], the KOPIO Shashlyk calorimeter employed the APD+WFD readout can reach the energy resolution: ${\sigma }_E/E\left(\%\right)=\left(1.96\pm 0.1\right)\oplus \left(2.74\pm 0.05\right)/\sqrt{E\left(\text{GeV}\right)}\mathrm{.}$ (12) And the time resolution of the calorimeter is ${\sigma }_{\text{T}}\left(\text{ps}\right)=\left(72\pm 4\right)\sqrt{E\left(\text{GeV}\right)}\oplus (14\pm 2)/\sqrt{E\left(\text{GeV}\right)}\mathrm{.}$ (13) The dense multi-layer design of KOPIO Shashlyk sampling calorimeter, which has a 3%/$\sqrt{E\left(\text{GeV}\right)}$ energy resolution, shows potential advantage in the performance improvement of Shashlyk sampling calorimeter, and it is used as a alternative for electromagnetic calorimeter in more and more experiments now, such as PANDA [318] and MPD [319].

New Techniques And Trend Of Calorimeters

Current improvements in electromagnetic calorimeters are mainly focused on three aspects [310, 320, 321]: time measurement, materials and dual readout (DRO). The time resolution at the level of 10 ~ ps or better will allow the electromagnetic calorimeters to improve the performance of shower reconstruction, energy correction, particle identification and so on. Precise timing data can be collected by all calorimeter cells or a specialized timing layer integrated in the electromagnetic calorimeter to achieve high resolution for the smallest ionizing particles.

Calorimeter materials are divided into inorganic scintillators and organic plastic scintillators. Inorganic scintillators are characterized by fast decay time, with the commonly used inorganic material BaF₂ which has a decay time of 0.5 ns. Organic scintillators, such as plastics [322], pure liquid scintillators (LS) [323], and water-based liquid scintillators (WbLS), exhibit fast pulse response (1 ∼ 2 ns) and adequate light yield (10⁴ photons/MeV). When coupled with SiPMs readouts, plastic scintillators are widely utilized due to lower cost, ease of use and flexibility. LS and WbLS are insensitive to radiation, possess long optical transparency, and can be deployed in high-dose environments.

In the DRO method [324], Cerenkov light is used to estimate the electromagnetic component of the shower, and the relative amount of Cerenkov and scintillation light is used to correct the energy of each shower. DRO can be added to a homogeneous calorimeter in two ways: using two filter + SiPM assemblies (sensitive to Cerenkov or scintillation light) at the rear of the crystal; or developing different time structures for fast emission of Cerenkov light and slow emission of scintillation light. Recently, the development of uniform crystal ECAL with DRO has become possible due to the availability of SiPMs with good sensitivity in the red portion of the spectrum. In the long run, research on new optical materials, such as photonic crystal fibers, quantum dots, and semiconductor nanoparticles wavelength shifters, as well as fibers that maintain polarization information, can improve performance and/or reduce the cost of DRO fiber calorimeters.

Inorganic Scintillators

Inorganic Scintillators Containing ⁶Li Typical inorganic scintillators containing ⁶Li include Li glass, LiI(Eu), LiF/ZnS:Ag, and potassium cryolite crystals. Li glass detectors can measure neutrons below several hundred keV with high detection efficiency, but they have low light yield and are also sensitive to γ [358]. LiI(Eu) scintillators, on the other hand, can measure neutron energy spectrum in the MeV range, but with poor resolution and low light yield. Recent studies have shown that appropriate doping levels and thermal treatment of pure LiI crystals doped with Eu2+ can address the light yield issue in LiI(Eu) for neutron detection, and improve its energy resolution and n/γ-discrimination capability to some extent [359]. LiF/ZnS:Ag, which exhibits excellent n/γ-discrimination characteristics, necessitates thin detectors due to the low transparency of ZnS(Ag), resulting in lower detection efficiency [360]].

Potassium cryolite materials, as a newly developed type of inorganic scintillator, are considered as potential alternatives to LiI(Eu). Various derived products have been developed in recent years, including Cs₂LiYCl₆:Ce (CLYC), Cs₂LiLaCl₆:Ce (CLLC), Cs₂LiLaBr₆:Ce (CLLB), Cs₂LiLaBr₆-xClx:Ce (CLLBC), Cs₂NaYCl₆ (CNYC), Cs₂LiYBr₆:Ce (CLYB), etc. Potassium cryolite materials exhibit high light output (>20000 photons/MeV), good energy resolution (Eγ = 662 KeV, < 5%), and excellent n/γ discrimination capability [361]. However, they have some limitations in the detection of fast neutrons[362]. A. Gueorguiev and others have developed a new technology that combines inorganic scintillator (CLYC) with an organic scintillator matrix (methyl methacrylate, PMMA) to achieve simultaneous detection of fast neutrons, thermal neutrons, and γ-rays. Fig 85 illustrates the structure of this composite detector [363].

Fig. 85Full-screen View

(Color online) Photograph of ∅1"x1%" composite CLYC-Plastic detectors produced in RMD’s (left), schematic of its structure (middle) and CLYC crystallites used in the composite detector[364].

Inorganic Scintillators Containing ¹⁰B The latest developments involving scintillators containing ¹⁰B are primarily focused on neutron imaging. In 2014, Nakamura developed a low-afterglow, thermal neutron-sensitive ZnS/¹⁰B₂O₃ scintillator with a decay time of 60 ns, which offers potential for neutron imaging [365]. In 2020, Miller conducted neutron imaging research using a ¹⁰B/Cs(Tl) scintillator screen, directly depositing ¹⁰B onto Cs(Tl) thin film using electron beam deposition to create a layered scintillator structure, enabling a spatial resolution of 9 μm [366]. Additionally, William found that boron-based scintillators exhibit excellent neutron detection capabilities in applications involving cold and ultra-hot neutrons, making it highly likely to achieve higher-quality neutron images with shorter imaging time [364].

Organic Scintillators

Organic scintillators, rich in light elements such as H and C, are commonly utilized for neutron detection using the nuclear recoil method. Organic crystals, liquid scintillators, and plastic scintillators all fall under the category of organic scintillators. They have short decay times for light emission, often on the order of tens of nanoseconds, making them suitable for high neutron flux detection. However, organic scintillators generally exhibit sensitivity to gamma rays as well.

Organic Crystals Anthracene and stilbene crystals are the basic types of organic crystals used for fast neutron detection. These two crystals have high light yields and good discrimination capabilities between neutrons and gammas. However, they exhibit anisotropic light response and significant temperature effects. Additionally, it is challenging to obtain crystals with large size and high cost. In the first decade of this century, a team led by Natalia Zaitseva at Lawrence Livermore National Laboratory (LLNL) in the United States developed new crystal growth methods [367]. The new crystal growth methods developed by the team at LLNL allowed for the production of larger and more uniform organic crystals, addressing the previous challenges in obtaining large crystals. These advancements in crystal growth have contributed to an increased interest in organic crystals for various applications, including fast neutron detection.

Liquid Scintillators Liquid scintillators are also used for fast neutron detection. However, their solvents mainly consist of flammable and toxic substances such as toluene, xylene, and trimethylbenzene. Therefore, extra caution is required during their use to prevent incidents such as breakage, leaks, or flash explosions. The conditions for using liquid scintillators are more demanding and strict in order to ensure safety.

Plastic Scintillators Plastic scintillators have higher detection efficiency for fast neutrons compared to other detectors. They are also cost-effective, high mechanical strength, and can be processed into various shapes, making them suitable for use in various harsh environments. However, plastic scintillators cannot be used for thermal neutron measurements, and most commercially available plastic scintillators, except for EJ-276, have poor n-γ discrimination capabilities. Therefore, the development of doped plastic scintillators that have simultaneous detection capabilities for fast and thermal neutrons, as well as improved n-γ discrimination, is one of the current directions in scintillation detector research.

Natalia Zaitseva has produced plastic scintillators with enhanced n-γ discrimination capabilities, surpassing the commercially available EJ-309 liquid scintillator [368]. Figure 86 shows the pulse shape discrimination spectrum of this scintillator compared to EJ-309 for ²⁵²Cf. Furthermore, Natalia Zaitseva and others have also developed plastic scintillators containing benzyl salicylate lithium, which exhibit good discrimination capabilities for thermal neutrons, fast neutrons, and γ-rays[369]. Figure 87 neutron energy range.

Fig. 86Full-screen View

(Color online) ²⁵²Cf n/γ discrimination plot of a ∅ 51 and 76 mm long PSD 36% PPO plastic in comparison to the same size liquid scintillator obtained under the same conditions of measurements [368].

Fig. 87Full-screen View

²⁵²Cf PSD profiles obtained with ∅2.5 cm × 2.5 cm-thick plastics containing 30% of PPO, 5% of ⁶Li-3-PSA, and secondary dyes: (a) DPA; (b) Bis-MSB. 9 mm of HDPE moderation. FOMs are calculated in the near-thermal-neutron-spot energy range by Gaussian approximation of individual peaks for gamma, fast neutrons, and thermal neutrons[369].