High rate MRPC technology for TOF systems and beyond

1 Introduction

In modern high energy physics experiments, particle identification (PID) is essential for signal extraction and background suppression. A time-of-flight (TOF) detector measures the flight time of a charged particle over a known path length and, together with the momentum reconstructed by tracking detectors, provides mass information for π / K / p separation. The basic principle of TOF PID and the corresponding derivation have been discussed in previous studies (Llope and Collaboration, 2012; Williams, 2012). In the high-momentum approximation, the separation power between two particle species can be expressed as Equation 1:

n σ TOF ≡ | Δ t | σ TOF ≈ L c 2 p 2 σ TOF | m 1 2 − m 2 2 | . (1)

This relation shows that, for fixed flight length and momentum, the separation power increases as the system time resolution improves. For example, for π / K separation at p = 4 G e V / c and L = 8 m, a system time resolution of about 60 p s is required for a 3 σ separation.

Since the late 1990s, the multi-gap resistive plate chamber (MRPC) has been developed as a gaseous timing detector that provides high time resolution with large-area coverage at a relatively low cost (Zeballos et al., 1996), and it has been adopted in several large time-of-flight systems in high energy physics experiments, such as RHIC–STAR, LHC–ALICE, and BESIII (Wang Y. et al., 2010; Wu et al., 2005; Alici and ALICE Collaboration, 2013; Li et al., 2017). An MRPC is assembled by stacking multiple resistive plate electrodes with a bulk resistivity of 1 0 10–1 0 12 Ω c m. High voltage is applied to the outermost electrodes, while the inner resistive plate electrodes are left electrically floating and acquire their potentials via capacitive coupling. The gas volume is divided into multiple thin gas gaps separated by the resistive plate electrodes, with a typical gap thickness of 100–300 μ m. Such thin gaps are important for achieving good timing performance, but they also impose stricter mechanical requirements for large-area detector construction. In particular, precise control of the gap uniformity and a sufficient density of spacers are usually required, which may slightly reduce the effective active area. MRPC operates at an electric field of about 100 k V / c m in avalanche mode. Common MRPC gas mixtures are based on R-134a ( C 2 H 2 F 4 ) with iso-butane as a quencher and an electron-attachment component such as S F 6. Although S F 6 is a minor component compared with the main gas, its fraction in MRPC mixtures can reach several percent, for example, about 5 %, which is usually higher than that used in conventional RPC gas mixtures. This conventional gas choice also brings an important environmental concern. R-134a and S F 6 are fluorinated greenhouse gases with high global warming potential, and S F 6 has an especially large GWP. Therefore, the use of R-134a/S F 6-based mixtures is becoming a long-term sustainability challenge for both RPC and MRPC technologies, especially under stricter F-gas regulations. Recent RPC studies have investigated eco-friendly mixtures based on HFO-1234ze and C O 2, which can reduce the environmental impact, but they may require higher operating voltage and can show larger current or aging-related effects under irradiation (Abbrescia et al., 2024; Quaglia et al., 2024). For MRPCs, such replacement is particularly challenging because the gas mixture must still provide stable avalanche multiplication, high efficiency, and excellent timing resolution. Therefore, future MRPC development should consider both low-GWP gas mixtures and detector-level mitigation methods, such as low gas flow, gas recirculation, and sealed or direct-flow structures. The avalanche signal is capacitively induced on external readout pads or strips. Depending on the readout geometry, MRPCs can provide position information in different ways. In pickup-pad designs, the hit position is determined by the fired pad, as in the ALICE TOF detector (Hatzifotiadou and ALICE Collaboration, 2003). In detectors employing two orthogonal strip views, the two transverse coordinates are obtained directly from the crossing strips, as demonstrated in a double-stack MRPC developed for medical imaging applications (Sharifi and Saramad, 2018). In long strip MRPCs with double-ended readout, which is widely used in TOF applications, one coordinate is determined by the strip index, while the coordinate along the strip is reconstructed from the signal arrival-time difference between the 2 strip ends; a representative example is the MRPC system developed for the STAR experiment at RHIC (Llope and STAR Collaboration, 2012). Figure 1 shows a schematic structure of a long strip MRPC detector with double-ended readout. As a specific example, under typical operating conditions with a gap thickness of 250 μ m, an operating electric field of about 110 k V / c m, and a gas mixture of 90 % R-134a, 5 % iso-butane, and 5 % S F 6, MRPC-based TOF systems have demonstrated system time resolutions better than 100 p s and efficiencies typically better than 95 %. These characteristics make MRPCs a practical choice for TOF-based π / K / p identification in modern collider and fixed-target experiments.

With the increase in accelerator luminosity and event rate, TOF systems are moving from traditional low-rate operation to conditions that require excellent time resolution, high efficiency, and improved rate capability. Two representative cases are the CBM experiment at FAIR and the proposed SoLID experiment at Jefferson Lab. In the CBM (Compressed Baryonic Matter) heavy-ion collision experiment at FAIR (Friman et al., 2011), rare probes need to be collected at interaction rates up to about 10 M H z. Under this condition, the peak local rate on the CBM-TOF inner wall can reach 30 k H z / c m 2 (Herrmann et al., 2014; Deppner and Herrmann, 2019). For the proposed SoLID experiment, high-current electron scattering also requires TOF detectors with both good time resolution and high rate capability. The MRPC-TOF is expected to work with high efficiency and stable performance at local rates of about 20 k H z / c m 2 (Wang Y. et al., 2013; 2019). These examples show that future TOF systems require rate capability far beyond that of traditional float-glass MRPC detectors.

At such high particle fluxes, the MRPC performance is limited by several rate-related effects. First, the increased current through resistive plate electrodes leads to a rate-induced voltage drop, reducing the effective electric field in the gas gaps and worsening efficiency and time resolution (Wang J. et al., 2010). Second, high rate operation puts strong demands on gas flow and exchange. The larger avalanche charge increases the production of gas dissociation by-products in the working gas. If these by-products are not removed efficiently, they may accumulate on glass surfaces and spacer regions, modify local surface conditions, and lead to increased dark current, noise rate, and aging effects (Akindinov et al., 2004). In this work, the term “sealed MRPC” refers to a self-sealed detector structure with continuous gas flow directly through the active gas gaps. Third, spacers between resistive plate electrodes can introduce local field distortions, which may increase discharge probability, dark current and noise rate. These electrode, gas, and spacer-related effects are discussed separately in the following sections. Studies on high rate MRPCs are not only for higher instantaneous rate capability and efficiency, but also for long-term stable operation under continuous irradiation.

To meet these requirements, many design optimizations have been studied for MRPCs, including low-resistive or thin electrodes (Akindinov et al., 2007; Wang J. et al., 2010; Forster et al., 2016), direct-flow or sealed gas structures (Akindinov et al., 2004; Liu et al., 2019; Aprodu et al., 2023), and improved spacer designs (Wang et al., 2023; Yao et al., 2025; Sun et al., 2025). In this review, we summarize these developments and group them into three directions to provide a comprehensive overview of the research and progress in high rate MRPC technologies:

• Electrode-related optimizations: to reduce the rate-induced ohmic voltage drop and to shorten the local field-recovery time, i.e., the time needed for the local electric field to recover after avalanche charge deposition;
• Gas flow and exchange optimizations: to improve gas flow uniformity and gas exchange efficiency inside the gas gaps, and to reduce rate-related gas aging;
• Spacer-related optimizations: to reduce local field distortion and decrease noise rate and dark current in spacer regions.

The following sections will discuss these three directions, focusing on the MRPC performance optimization under high particle flux.

2 Resistive plate electrode optimizations using low resistive glass

In a high particle flux environment, typically 1 – 100 k H z / c m 2, the rate capability of an MRPC is primarily limited by the ohmic voltage drop across the resistive plate electrodes. Under continuous irradiation, avalanche charges are deposited on the electrode surfaces and are removed through the finite bulk conductivity of the electrodes, producing an average voltage drop and therefore reducing the effective voltage across the gas gaps. The average current over an irradiated area A can be written as I ̄ ≃ ϕ A q, where ϕ is the particle flux density and q is the average transported avalanche charge per event. For flat resistive electrodes, the equivalent resistance scales as R ∝ ρ d / A, where ρ is the electrode bulk resistivity and d is the electrode thickness. Therefore,which shows that the average voltage drop V ̄ drop increases with ϕ, q, ρ, and d. Equation 2 should be understood as a simplified average estimate where the avalanche charge q is treated as a constant. In practice, however, q depends strongly on the local electric field and thus on the voltage drop itself, leading to a dynamic feedback effect. Furthermore, the instantaneous local voltage drop can fluctuate due to variations in the avalanche charge and the incident particle rate. Depending on the definition of the effective resistive thickness, the proportionality for the average voltage drop may contain a constant factor of order unity (Carboni et al., 2003; Collaboration et al., 2024).

V ̄ drop ≡ V ap − V ̄ gap ≃ I ̄ R ∝ ϕ q ρ d , (2)

Electrode-related optimizations to improve the rate capability mainly aim at reducing the rate-induced voltage drop across the resistive electrodes. According to Equation 2, for a given particle flux and avalanche charge, this voltage drop is governed by the product of the electrode bulk resistivity ρ and thickness d. Therefore, reducing the glass thickness can reduce V drop, but the rate capability is not determined by thickness alone. If the electrode material still has a large bulk resistivity, the voltage drop can remain significant at high flux. MRPCs built with 280 μ m thick soda lime glass plates with a bulk resistivity of ρ ≈ 5 × 1 0 12 Ω c m were evaluated in a CERN test beam at different fluxes (Forster et al., 2016). The MRPC prototype had six gas gaps with a gap thickness of 220 μ m. At a working voltage of 15 k V, it achieved a time resolution of 80 p s with an efficiency of about 95 % at a particle flux of 1.5 k H z / c m 2. However, when the flux increased to 16 k H z / c m 2, the efficiency dropped to about 70 % at the same working voltage. This degradation should be understood as a consequence of the increased voltage drop across the resistive electrodes, which is related to ϕ q ρ d. These results show that reducing the electrode thickness is useful for improving the rate capability, but it may not be sufficient for stable operation at particle fluxes of order O ( 10 k H z / c m 2 ) when standard high-resistivity glass is used. For higher-rate applications, reducing the bulk resistivity ρ becomes essential, because it reduces both the ohmic voltage drop and the dielectric relaxation time of the electrode. For standard glass electrodes, the intrinsic dielectric relaxation time is typically on the order of seconds, while low-resistive materials can reduce this recovery time to the millisecond level. This can be estimated from the dielectric relaxation time (Equation 3):where ϵ 0 is the vacuum permittivity and ϵ r is the relative dielectric constant of the glass. Taking a typical value of ϵ r = 8.5, for the soda-lime glass discussed above with ρ ≈ 5 × 1 0 12 Ω c m = 5 × 1 0 10 Ω m, one obtains τ ≈ 3.8 s. For the low-resistive glass with ρ ≈ 1 0 10 Ω c m = 1 0 8 Ω m, the relaxation time is τ ≈ 7.5 m s. Therefore, reducing the bulk resistivity naturally shortens the intrinsic field-recovery time from the second scale to the millisecond scale. Therefore, reducing the bulk resistivity naturally shortens the intrinsic field-recovery time from the second scale to the millisecond scale. Furthermore, compared to traditional RPCs, the multi-gap structure of MRPCs limits the avalanche charge per gap, resulting in a faster effective restoration of the local electric field. One early candidate material was low-resistive phosphate glass. An RPC prototype with phosphate glass electrodes, with a bulk resistivity on the order of 1 0 10 Ω c m, was developed. At a particle flux of 18 k H z / c m 2, the RPC maintained a time resolution of 120 p s and an efficiency above 95 % (Akindinov et al., 2007).

τ = ρ ϵ = ρ ϵ 0 ϵ r , (3)

To satisfy the CBM-TOF requirement of maintaining > 95 % efficiency and a system time resolution better than ≈ 80 p s under particle fluxes spanning 0.1 – 30 k H z / c m 2, the CBM collaboration developed a TOF wall concept (Deppner et al., 2014). The CBM-TOF wall is divided into two parts, the inner wall and the outer wall. The layout of the outer wall is shown in Figure 2. In order to meet the local rate capability and granularity constraints, different MRPC types are deployed in regions with different particle fluxes. For the central part of the outer wall, where the expected particle flux is 1 – 5 k H z / c m 2, a specific high rate MRPC design using low resistive silicate glass electrodes was developed by Tsinghua University. This detector is designated as MRPC2 in the CBM experiment. A photograph of the MRPC2 detector is shown in Figure 3. It features a double-stack structure with 2 × 4 gas gaps; each gap has a thickness of 250 μ m. The geometrical parameters of MRPC2 are listed in Table 1.

TABLE 1. Parameters of the MRPC2 detector.

From a general point of view, this low-resistive-glass MRPC design illustrates an effective route to improve the rate capability of MRPC detectors. Compared with conventional float-glass MRPCs, the lower bulk resistivity reduces the rate-induced voltage drop across the resistive electrodes and helps to maintain the electric field in the gas gaps under high particle flux. However, this approach also has practical limitations. The low resistive silicate glass used in this design is mechanically more fragile than standard float glass, which limits the available plate size and the maximum module area, and also requires special care during detector assembly, transportation, and installation. It is also difficult to use very thin electrodes, for example, glass plates of 0.4 m m thickness or below, while keeping sufficient mechanical reliability. Because of this limited mechanical robustness, sufficient spacer support is also required to prevent glass deformation and to maintain uniform gas gaps; however, the spacer structures may reduce the effective active area and introduce additional local field distortions near the spacers. Thus, low-resistive glass is an important component for high rate MRPCs, but the final detector performance still depends on the combined optimization of electrode material, mechanical structure, spacer design, gas exchange, readout granularity, and system integration.

The low resistive glass features ohmic behavior with a bulk resistivity of the order of 1 0 10 Ω c m. This value was measured on ground and polished glass samples in a constant-temperature dry box at 25 ° C and 30% relative humidity with an applied DC voltage of 1 kV. Furthermore, the surface roughness of the polished glass was evaluated using a 3D surface profiler, yielding a ten-point-height roughness of less than 10 nm (Wang J. et al., 2013), ensuring the microscopic smoothness of the glass electrodes. In contrast, the thickness uniformity, obtained through macroscopic measurements at various positions on the glass plate, is essential to maintain a uniform gas gap. In addition, previous studies have demonstrated the robust aging properties of this glass (Wang J. et al., 2010; 2014). During a stability test conducted at 28 ° C for 34 days with an applied DC voltage of 1 kV, the total transported charge across the glass reached 1 C / c m 2. This value is approximately equivalent to the transported charge expected for 5 years of operation in the highest flux region of the CBM-TOF wall. These material-level DC tests indicate good electrical stability of the low-resistive glass under the tested conditions. The accumulated-charge result mainly supports the suitability of this glass as a candidate electrode material for long-term high rate MRPC operation, while the detector performance is discussed below based on beam test results. The main material parameters of this low-resistive glass, collected from the original development studies, are summarized in Table 2 (Wang J. et al., 2010; Wang et al., 2013 J.; Wang and Yu, 2020). This table is included to define the electrode material used in the CBM-MRPC2 design and to provide the basis for the following discussion on high rate detector performance. The general MRPC performance and TOF applications have been summarized in previous reviews (Wang and Yu, 2020), while the present work focuses on high rate MRPC technologies, including low-resistive electrodes, gas flow and exchange optimization, and spacer-related stability. In terms of detector fabrication, the dimensions of a single MRPC module are constrained by the available size and mechanical robustness of the glass electrodes, the manufacturing limits of the PCBs, and the strict requirement to maintain a uniform gas gap over the entire active area. At the TOF system level, the detector system must also satisfy constraints imposed by readout granularity, the total number of electronics channels, and the geometric acceptance of the experiment. Therefore, the design of an MRPC module represents a careful balance between fabrication feasibility, detector performance, and system integration. As listed in Table 1, the active area and readout-strip geometry of the CBM-MRPC2 module were specifically optimized to meet the local rate and granularity requirements of the 1 – 5 k H z / c m 2 region of the CBM outer wall, while preserving high efficiency and excellent timing performance. The geometric coverage required by the experiment is achieved by assembling multiple MRPC modules into a large-scale TOF wall.

TABLE 2. The performance of low resistive glass.

Following the above discussion on the material properties and detector design of the low-resistive-glass MRPC, representative beam-test results from multiple facilities are summarized here to show its performance under high particle flux conditions. A deuteron-beam test at JINR demonstrated that, even at a particle flux of around 70 k H z / c m 2, the MRPC efficiency remained 93 % while the time resolution stayed below ≈ 80 p s (Wang et al., 2014). In addition, a separate high-intensity electron-beam test was carried out at the Helmholtz-Zentrum Dresden-Rossendorf (HZDR) in Dresden. Here, both Pad-readout (PMRPC) and Strip-readout (SMRPC) modules were evaluated under non-uniform irradiation using a narrow Gaussian electron beam. Accidental coincidences were negligible due to a strict hardware coincidence trigger. The specific results of this HZDR test are shown in Figure 4. The data show that the detector achieves an efficiency better than 95 % and a time resolution better than 70 p s at a particle flux of 20 k H z / c m 2. When the particle flux increases to 35 – 60 k H z / c m 2, the performance degrades but remains within an acceptable range, with an efficiency better than 90 % and a time resolution better than 80 p s (Wang J. et al., 2013).

Because the cited studies used different detector geometries, beam conditions, readout structures, and operating voltages, the comparison should be regarded as qualitative. Nevertheless, the results indicate that the electrode contribution to the rate-induced voltage drop is a central factor in high-rate operation. Reducing the glass thickness can improve the rate capability, but high bulk resistivity still leads to significant efficiency loss at fluxes of O ( 10 k H z / c m 2 ). Low-resistive glass electrodes with ρ ∼ 1 0 10 Ω c m reduce this voltage drop and can extend efficient operation to several tens of k H z / c m 2. Therefore, low electrode resistivity is a key requirement for high-rate MRPC operation, although the final performance still depends on the combined optimization of electrode material, gap structure, readout segmentation, gas flow, spacer design, and mechanical stability.

3 Gas flow and exchange optimizations using sealed MRPC

In high particle flux environment, the avalanche charge crossing the MRPC electrodes per unit time increases significantly. For typical MRPC working gases such as R-134a, the avalanche process produces various reaction by-products, including F −. These by-products may accumulate on the resistive plate electrode surfaces, as well as in regions such as narrow slits near the edges of spacers. Scanning Electron Microscope (SEM) analysis using a Hitachi SU8000 series instrument on resistive plate electrodes before and after long-term irradiation under the highest flux beams provides direct evidence of such depositions. As shown in Table 3, compared with a fluorine content below 0.1 % on the surface of a new resistive glass electrode, the fluorine fraction on an irradiated glass surface exceeds 10 %, and the local fraction in the fishline spacer region can even exceed 40 % (Wang et al., 2023). Additionally, the intrinsic N a + from the glass migrates towards the cathode surface under the strong electric field, showing significant variations and readily bonding with the accumulated F −. Such gas-related contamination can lead to an increase in dark current and noise rates during high energy physics operations. Therefore, for MRPCs targeting high particle flux applications, improving the gas exchange efficiency should be considered one of the key design goals. To provide a complete picture of how gas related aging is mitigated in high rate environments, this section focuses on detector level gas flow and gas exchange optimizations for MRPCs, especially direct flow and sealed structures that reduce the inactive gas volume and force the working gas to pass through the active gaps.

TABLE 3. Summary of the SEM elemental analysis (Wang et al., 2023).

Research related to the ALICE TOF MRPC indicates that the effective sensitive volume of the MRPC, which consists of the narrow gas gaps between the resistive plates, accounts for only about 0.7 % of the total volume inside the large external gas box housing the detector (Akindinov et al., 2004). To improve the gas exchange efficiency, a direct-flow Multi-Strip Multi-Gap Resistive Plate Chamber (MSMGRPC, following the notation used in Ref (Aprodu et al., 2023)) was developed as an alternative to the traditional gas-box MRPC. In this novel design, each stack is tightly sealed with gas guides and epoxy glass fiber ledges to enclose the gas volume. This configuration forces the gas to flow directly through the gaps between the resistive plate electrodes and allows for a reduced number of spacers in the electric field. During aging investigations using a high-intensity X-ray flux, a continuous decrease in the dark counting rate was observed for the direct-flow MRPC after successive exposures. This performance is significantly superior to that of the conventional diffusion-based chamber, where the dark counting rate increased towards a plateau, confirming that the direct-flow design effectively mitigates gas pollution effects (Aprodu et al., 2023). Following the same idea of forcing the gas to exchange more efficiently in the active gaps, a sealed MRPC configuration was developed (Liu et al., 2019). Here, “sealed MRPC” refers to a self-sealed detector structure without a large external gas box, rather than a fully closed system without gas exchange. The working gas is still continuously supplied and exhausted through gas tubes, but it flows more directly through the active gas gaps with a much smaller internal gas volume. By applying silicon glue to the PCB edges and positioning gas tubes directly at the glass boundaries, this design reduces the internal gas volume by a factor of 50 compared to traditional gas boxes. Beam tests showed that for MRPCs with small gap width (e.g., 140 μ m), this reduced gas exchange volume configuration helps to improve the gas exchange efficiency in the active gaps. The resulting faster removal of avalanche by-products is beneficial for maintaining the time resolution under high particle fluxes. At a particle flux of 30 k H z / c m 2, the sealed MRPC reached a time resolution of about 60 p s and an efficiency better than 80 % (Liu et al., 2019).

To mitigate gas contamination and aging in high particle flux environments, Tsinghua University developed a new sealed MRPC detector. Unlike traditional designs where the MRPC is placed inside a gas-tight aluminum or plastic box, the sealed MRPC detector body itself is completely self-sealed, with gas flowing directly through the gaps.

The structure of the sealed MRPC is shown in Figure 5. In a traditional MRPC, each stack contains several resistive plate electrodes acting as electrodes, which are separated by spacers to form gas gaps. A carbon layer is coated on the outer surfaces of the top and bottom resistive plate electrodes, to which high voltage is applied to create an electric field across the gas gaps. Based on this stacked structure, an integrated sealing frame is manufactured using 3D printing. The design of the sealing frame is shown in Figure 6. Designed according to the dimensions of the resistive plate electrodes, this frame perfectly encloses the existing stack. For the top and bottom surfaces of the sealed stack, two sealing methods are employed depending on the rigidity of the glass. For relatively rigid glass (e.g., the 0.7 m m float glass used in the CEE experiment) (Wang et al., 2020), the outermost resistive plate electrodes serve as the top and bottom surfaces of the sealed stack. For more fragile glass (e.g., the 0.7 m m low resistive silicate glass used in the CBM experiment) (Wang et al., 2021), 1 m m thick PMMA panels are added to seal the top and bottom surfaces to increase the structural strength of the detector, with the high voltage introduced via copper strips. In both sealing methods, grooves are reserved on the sealing frame to accommodate the glass or PMMA panels, ensuring a flat surface for the final sealed stack. Additionally, shallow grooves are designed on the outer part of the frame overlapping areas for applying epoxy resin to ensure gas tightness. Gas inlets and outlets are located on the two shorter side surfaces of the sealed stack for gas exchange. Consequently, each sealed stack becomes an independent unit. Combined with readout PCBs and nylon screws, these units can be assembled into double-stack (Wang et al., 2020) or four-stack (Sun et al., 2023) MRPC detectors as needed.

The sealed design significantly enhances the gas exchange capability of the MRPC detector. For an MRPC operating in a traditional gas box, a gas flow rate of 50 m l / m i n for 1.5–2 days is required to completely flush out the air. In contrast, the internal gas volume of a sealed MRPC is only about 250 m l. A gas flow rate of 20 m l / m i n for 1 hour can exchange more than four times this volume, allowing the detector to reach its working state within 2 hours. The self-sealed MRPC prototype for the CEE experiment (detailed in (Wang et al., 2020)) successfully underwent a long-term cosmic ray test. With an operating gas flow rate as low as 1 m l / m i n, the detector maintained a dark current below 10 n A, a time resolution of 80 p s, and an efficiency of 95 % over a 24-day test period. A low resistive glass sealed MRPC designed for the CBM experiment was tested at the CERN SPS beamline in November 2015. At a particle flux of 10 k H z / c m 2, the sealed MRPC achieved an efficiency of 98 % and demonstrated operational stability (Lyu et al., 2017). Furthermore, during the relaxation period following the beamtime, the dark current of sealed MRPC recovered much faster than the dark current of unsealed one. These results demonstrate that the sealed MRPC accelerates gas exchange, provides sufficient stability in high particle flux experiments, and reduces the required gas flow rate by one to two orders of magnitude, leading to lower costs. The sealed MRPC has been successfully used in high energy physics experiments. For instance, the CEE-eTOF wall was constructed using this new generation of sealed MRPCs, and the system has successfully completed mass production and installation.

Overall, these studies show a common design tendency for high-rate MRPCs: the gas system should be optimized according to the effective active gap volume rather than the total mechanical volume of the detector housing. The SEM observations of fluorine-rich deposits indicate that aging is closely related to the local accumulation of avalanche by-products, especially near spacers and poorly exchanged regions. Therefore, reducing dead gas volume, shortening the gas path, and forcing the gas to flow directly through the gaps are effective ways to suppress contamination. These results provide practical design guidelines for future high-rate MRPC systems, where timing performance, aging tolerance, gas consumption, and detector modularity must be considered together.

4 Spacer-related optimizations using improved spacer design

Improvements in resistive plate electrodes and in gas flow and exchange methods have effectively enhanced the performance of MRPC detectors under high particle flux by mitigating rate-induced voltage drop and gas contamination. However, new evidence from long-term operation in the mini-CBM (mCBM) experiment, a full-system test setup for the CBM experiment (collaboration, 2017), indicates that high beam intensity can still lead to high noise rates, large dark currents, and long recovery times in local regions around conventional fishline spacers, which are typically nylon strings used to maintain a uniform gas gap between the resistive plates (Wang et al., 2023). By directly comparing the physical layout in Figure 7A with the noise-hit distribution in Figure 7B (Wang et al., 2023), the spatial correlation between the spacer geometry and the high noise rate regions is clearly demonstrated. In addition, comparative X-ray tests were carried out using MRPC prototypes with different total lengths of fishline spacers. It was found that, with other key variables (including detector structure, gas mixture, applied high voltage, and X-ray irradiation conditions) kept constant, a longer total fishline length led to a larger working current during irradiation, as well as a higher dark current after the irradiation ended (Wang et al., 2023). Meanwhile, as already summarized in Table 3, SEM analysis before and after X-ray irradiation showed increased Na and F concentrations in the fishline spacer regions. These results suggest that the presence of spacers induces more intense local avalanches, which leads to a stronger accumulation of chemically active by-products in their vicinity.

Based on these observations, several improvements to the traditional fishline spacer have been proposed (Wang et al., 2023; Yao et al., 2025; Sun et al., 2025). In this section, we summarize the recent progress in these spacer-related optimizations to illustrate how different spacer designs mitigate the local instability issues. Spacer based gap supports have already been widely used in conventional RPCs. Therefore, the novelty of the pad spacer MRPC is not the general concept of using spacers, but the implementation of isolated cylindrical pad spacers in a thin gap MRPC stack to replace the long continuous fishline spacers. This geometry is expected to reduce the spacer-induced local field distortion, surface charge accumulation, and non-uniform gas exchange observed around fishlines in high rate MRPC operation. Tsinghua University developed the pad spacer MRPC. In this type of detector, cylindrical pad spacers made of Mylar are used to replace the traditional nylon fishline spacers. Their structures are shown in Figure 8. The pad spacer scheme was motivated by several simulations and physical considerations. First, electrostatic simulations showed that a local electric-field distortion occurs in the gas gap near the narrow slit at the interface between the fishline spacer and the resistive plate electrode, with a maximum field strength of 276.1 k V / c m. This value is more than three times the peak field strength in the pad spacer case of about 78 k V / c m. The fishline spacer surface also exhibits higher normal and tangential electric fields. The normal field may promote the adsorption of avalanche charges onto the fishline spacer surface, while the tangential field may increase electrical creepage along the spacer surface. By contrast, the pad spacer does not create the narrow slit present in the fishline design, and therefore avoids field distortions. Owing to its geometry, the normal electric field on the pad spacer surface is also smaller, which reduces charge adsorption. On the other hand, computational fluid dynamics (CFD) simulations showed that the sealed MRPC with pad spacers has faster and more uniform gas exchange, which helps remove by-products. Based on these considerations, pad spacers with a diameter of 3 m m and a height equal to the gas-gap thickness were finally adopted for the production of the CBM-TOF MRPC2 detector. The detailed parameters can be found in (Sun et al., 2025).

Subsequently, multiple rounds of irradiation tests were carried out for both fishline spacer and pad spacer MRPCs under X-ray conditions with a tube voltage of 45 k V, a tube current of 0.1 m A, and a flux of 3 k H z / c m 2. In terms of current behavior, the pad spacer MRPC showed reproducible response and recovery over four irradiation cycles, each consisting of 10 h of irradiation followed by 14 h of relaxation, as shown in Figure 9. During each X-ray-on period, the current increased over several minutes before reaching a plateau. This indicates that the detector current under irradiation is not purely instantaneous, but contains a transient component. Such behavior may be associated with the gradual establishment of equilibrium among avalanche charge production, charge transport through the resistive electrodes, charge accumulation or polarization on insulating surfaces, and charge removal or neutralization in the gas volume. The plateau current shows a slight decrease over successive cycles, from about 4.8 μ A to 4.5 μ A. The present data do not allow us to identify a unique origin of this small variation. Nevertheless, after each X-ray-off transition, the current rapidly returned to the dark-current level, and no cumulative increase of the residual dark current was observed. Therefore, within the tested irradiation cycles, the observed behavior does not suggest progressive aging or instability of the pad spacer MRPC. Further dedicated long-term irradiation studies are needed to clarify the microscopic mechanism of the transient current evolution. After the X-ray test, the noise counting rates of the fishline and pad spacer MRPCs were measured using self-trigger electronics. The noise counting rate of the pad spacer MRPC was found to be 0.33 H z / c m 2, much lower than the 34 H z / c m 2 of the fishline spacer MRPC, corresponding to a reduction by about two orders of magnitude. Its dark current was only 0.13 μ A, about one-10th of that of the fishline spacer MRPC (Sun et al., 2025). Another study from USTC also reported experimental results based on thermal-bonding spacers (TBS) (Yao et al., 2025). A TBS MRPC and a fishline spacer MRPC were exposed to an X-ray flux of 700 H z / c m 2 for 7 days. The working current during irradiation, as well as the dark current and background noise before and after exposure, were monitored. The noise of the fishline spacer MRPC increased by two orders of magnitude after irradiation, while the noise of the TBS MRPC remained stable. Earlier MRPC developments had also pointed out possible advantages of optimizing spacer geometry for gas exchange, although these studies were not specifically aimed at high rate aging (Devi et al., 2016). Spacer-related noise and aging have been investigated from both theoretical and experimental aspects, and results show that improved spacer designs can reduce noise rate, lower working current during irradiation and dark current after irradiation, and improve long-term stability in MRPCs operated under high particle flux. The MRPC2 detector with pad spacers has been confirmed for application in the CBM-TOF experiment and has already completed multiple beam tests in mCBM. In general, the improved spacer designs mainly reduce the local noise rate, dark current, and recovery behavior after irradiation. For the pad spacer MRPC, the basic detector performance was also compared with that of the fishline spacer MRPC in cosmic-ray tests. Both prototypes showed comparable efficiency and time resolution, and no obvious trade-off in the main timing performance was observed within the tested conditions (Sun et al., 2025). Therefore, the main advantage of the pad spacer design is not an improvement of intrinsic time resolution, but a reduction of spacer-induced instability under irradiation. It should also be noted that the available studies are not yet a complete systematic before-and-after comparison using the same detector. They are mainly based on prototype comparisons and irradiation tests under controlled conditions.

Overall, the currently available studies consistently point to the spacer/interface as a critical local factor in high rate MRPC stability. A possible physical picture is that the spacer changes the local electric field and gas flow. In the fishline design, the narrow slit at the fishline-glass interface enhances the normal and tangential electric fields. The normal field may help avalanche charges and ions to be adsorbed on the insulating surface, while the tangential field may drive surface charge transport or creepage. Such surface charging and creepage can increase the local current and induce noise signals. In addition, the fishline geometry can make the local gas exchange less uniform, which may enhance the accumulation of irradiation-induced chemical by-products near the spacer. SEM studies after irradiation showed increased Na and F concentrations in the fishline spacer regions, suggesting that fluorine-containing deposits may be involved in the aging process. However, dedicated gas analysis has not yet been performed in these studies, and specific molecular species have not been identified. Therefore, the microscopic physical and chemical mechanisms are still not fully understood. It remains unclear which factors among spacer geometry, surface properties, electric-field distortion, gas flow, and chemical deposition play the dominant role. Further local irradiation tests, gas analysis, and surface studies are still needed.

5 Conclusion

High particle flux conditions in the next-generation TOF systems require MRPC detectors to provide not only excellent time resolution and high efficiency, but also stable long-term operation under continuous irradiation. In this review, recent progress in high rate MRPC technology has been summarized from three points of view: resistive plate electrode optimization, gas flow and exchange optimization, and spacer-related optimization.

Representative results from the reviewed studies show clear performance improvements. For electrode optimization, low resistive glass with a bulk resistivity of about 1 0 10 Ω c m enables MRPCs to maintain an efficiency better than 95 % and a time resolution better than 70 p s at 20 k H z / c m 2. Even at 35 – 60 k H z / c m 2, the efficiency remains above 90 % with a time resolution better than 80 p s. For gas flow optimization, the sealed MRPC design reduces the internal gas volume by a factor of about 50 compared with a traditional gas-box MRPC, which significantly improves gas exchange efficiency and helps mitigate gas-related aging effects. In long-term cosmic-ray tests, at an operating gas flow rate of 1 m l / m i n, the sealed MRPC maintained a dark current below 10 n A, a time resolution of 80 p s, and an efficiency of 95 % over 24 days. In beam tests, the sealed MRPC also achieved an efficiency of 98 % at 10 k H z / c m 2. For spacer optimization, the maximum local electric field is reduced from 276.1 k V / c m in the fishline spacer case to about 78 k V / c m in the pad spacer case. After irradiation, the noise counting rate of the pad spacer MRPC is 0.33 H z / c m 2, much lower than 34 H z / c m 2 for the fishline spacer MRPC, while the dark current is reduced to 0.13 μ A, about one-10th of the fishline design.

These results show that high rate MRPC performance can be effectively improved through the combined optimization of electrode materials, gas flow structure, and spacer geometry. However, no single optimization can solve all limitations of high rate operation. Low-resistive electrodes reduce the rate-induced voltage drop and improve field recovery, but they also require careful control of material quality, dark current, current consumption, and mechanical robustness. Sealed or direct-flow gas structures improve gas exchange and reduce gas consumption, but they require reliable sealing and uniform gas distribution. Improved spacer designs can suppress local field distortion, noise rate, and dark current, although the microscopic mechanism of spacer-related aging still needs further study. It should also be noted that prototype or beam test performance does not directly represent the final performance of a full TOF system. In HEP experiments, the TOF system performance also depends on front-end electronics, calibration, tracking, pile-up, environmental conditions, and long term operational stability. Therefore, efficiency, time resolution, rate capability, gas consumption, and detector stability should be considered together when designing future MRPC-TOF systems. In addition, the environmental impact of conventional R-134a/S F 6-based gas mixtures remains an important issue. The development of low-GWP gas mixtures, together with low-flow or recirculating gas systems and sealed detector structures, will be necessary for improving the long-term sustainability of large-area MRPC systems.

Beyond high energy physics, some of the high rate MRPC developments reviewed here may also benefit societal applications. In medical imaging, MRPCs have been investigated for TOF-PET as a potentially low cost technology with good time resolution, 2D readout capability, and large-area scalability. However, current studies show that the main challenges are still the limited detection efficiency for 511 keV gamma rays and low-noise self-triggered readout, rather than rate capability alone. For muography, where the particle flux is much lower, the most relevant advantages are low gas consumption, long-term stability, and large-area coverage, which make sealed MRPC designs especially attractive.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. During the preparation of this work the authors used ChatGPT by OpenAI in order to improve language and readability. After using this tool, the authors closely reviewed and edited the content as needed and take full responsibility for the content of the publication.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Diego Gonzalez-Diaz, University of Santiago de Compostela, Spain

Reviewed by: Marcello Abbrescia, University of Bari Aldo Moro, Italy

Crispin Williams, European Organization for Nuclear Research (CERN), Switzerland

Raveendrababu Karnam, National Institute of Science Education and Research (NISER), India