1 Introduction

The compressed baryonic matter experiment (CBM) [1, 2] will be one of the main research pillars of the Facility of Antiproton and Ion Research (FAIR) which is currently under construction in Darmstadt/Germany. CBM aims to investigate strongly interacting matter at large baryon densities by colliding heavy ion beams provided by the SchwerIonenSynchotron SIS100 accelerator on a gold target, enabling center-of-mass energies \(\sqrt{s_{NN}}\) up to 4.9 AGeV. The main research interests are the exploration of the QCD phase diagram in the baryon-rich region and the determination of the equation of state of QCD matter. These goals can be reached by measuring not only all produced bulk particles (protons, kaons, pions) but also a variate of rare probes like multi-strange (anti-)hyperons and (multi-)strange hypernuclei. A further statistic hungry observable intended to be measured extensively by CBM is the invariant mass spectra of di-lepton pairs (both, in the electron and muon channel) which provides information at essentially each moment of time during the collision. For this purpose, an extremely high interaction rate of up to 10\(^7\) Hz is anticipated (conf. right panel of Fig. 1). Figure 1 shows on the left panel the setup of the CBM experiment. On the right panel, the interaction rate of several present and future heavy ion experiments is compared. The location of the CBM experiment is highlighted in the plot by the red ellipse. Inside a 1 Tm dipole magnet, tracking stations for vertex reconstruction and momentum determination will be available. For charged hadron identification, a time-of-flight (TOF) wall of about 120 m\(^2\) will be installed at about 8 m from the interaction point. For clear electron separation, a ring imaging detector (RICH) and a transition radiation detector (TRD) are foreseen. The TRD delivers additionally further tracking information which increases the pointing accuracy of tracks on the TOF wall. The RICH can be exchanged by the Muon Chamber (MuCH) when the di-muon program is going to be measured. For event plane determination, a fragment spectator detector (FSD) is installed in the back of the experiment. It is evident that all subsystems require radiation hard detectors with large rate capability.

Fig. 1
figure 1

Left panel: setup of the CBM experiment. Right panel: comparison of the interaction rate for present and future heavy ion experiments [3]. The red ellipse guides the eye to the anticipated interaction rate of CBM

Full size image

Timing multi-gap resistive plate chambers (from now on called MRPC) are gaseous detectors [4] providing nowadays an excellent time resolution in the order of 50 ps [5] with an efficiency of more than 95%. The major advantage of MRPCs compared to scintillator bars, read out by PMTs, is their extremely low cost per readout channel, making systems with more than 100,000 channels affordable. The CBM-TOF system [6,7,8] will be composed of about 1500 MRPCs with different granularity starting from 5 cm\(^2\) strip area up to 53 cm\(^2\), respectively. Counters with higher granularity are placed in the center of the wall where the track multiplicity and the charged particle flux are highest (up to 50 kHz/cm\(^2\)) while counters with the largest granularity are located in the outermost regions of the TOF wall where the flux gradually decreases to about 100 Hz/cm\(^2\). Several MRPCs are grouped in gas tight aluminum boxes called modules (230 modules in total), with a total gas volume of about 25 m\(^3\).

In order to reach such a high performance as stated above, MRPCs operate nowadays typically with gas mixtures composed of 90– 95% tetrafluoroethane (TFE) C\(_2\)H\(_2\)F\(_4\) (also known under the name R134a), 0–5% isobutane i-C\(_4\)H\(_{10}\) and/or 0–5% sulfur hexafluoride SF\(_6\). SF\(_6\) is introduced as an electron quencher, and consequently, it inhibits the streamer production. Isobutane is known to act as a photon quencher, stabilizing the avalanche dynamics. Since isobutane is a flammable gas, in some cases experiments like ALICE abandoned it lately. Measurements have shown that gas mixtures without isobutane reach similar detector performances [9].

The main issue with the currently used RPC [10] and MRPC gas mixtures is their high global warming potential (GWP over 100 years period) of 1430 for TFE, 3.3 for isobutane (negligible) and 23900 for SF\(_6\) [11]. Other literature quotes slightly lower values [12]. In order to reduce the environmental impact during the operation of the detectors to an acceptable level, the following possibilities are conceivable [12]:

  • As a first step, experiments introduced a closed loop gas system in their setup in order to lower the emission. This is nowadays standard.

  • Usage of alternative gases with lower GWP as a replacement for TFE and SF\(_6\) also known as eco-friendly gas mixtures. Finding proper replacements is currently a very active research topic.

  • The 100% recuperating/recycling of all gasses and their re-injection in the gas circuit. This possibility is discussed in the community, and first recuperation prototypes are in use at CERN.

  • Disposal of the gas at the exhaust. This possibility is not seriously considered in the community.

The strategy to reduce the global warming impact during operation in CBM-TOF will be elaborated in the following chapters.

2 Investigation on eco-friendly gas mixtures

During the last several years, an enormous research effort has been made to find eco-friendly gas mixtures as a replacement for the usually used “standard” gas mixtures in the RPC [13,14,15,16,17,18,19,20,21,22,23,24] and MRPC [25,26,27] community. As a replacement for TFE, the hydrofluoroolefin (HFO-1234ze) with a GWP of 7 turned out to be the most suitable candidate. It shows similar behavior as TFE, however, shifts the working point of RPCs and MRPCs by typically a few thousand volts to higher values. This might be a non-overcoming problem for some existing experiments, since present installations are not designed to operate at higher voltages and a replacement is not feasible. Therefore, a huge effort is ongoing to find complementary gases which lower the working point to acceptable values [26]. As a replacement of SF\(_6\), many new alternative gasses with all their advantages and disadvantages were presented and discussed here [28].

The TOF group of CBM investigated via cosmic tests five different eco-friendly gas mixtures based on other results reported in the literature:

  1. 1.

    Pure HFO-1234ze

  2. 2.

    HFO/isobutane 95%/5%

  3. 3.

    HFO/isobutane 97%/3%

  4. 4.

    HFO/isobutane/SF\(_6\) 90%/5%/5%

  5. 5.

    HFO/SF\(_6\) 95%/5%

The aim was to replace the TFE while keeping the usual quencher gases and verify the counter performance in terms of efficiency, cluster size and time resolution. With our standard mixture (TFE/isobutane/SF\(_6\) 90%/5%/5%), an efficiency of 97% was reached. The cluster size, which is a proxy for the produced charge, was in the order of 1.35. A counter time resolution in the order of 60 ps could be obtained at a working point of 5400 V. Figure 2 shows the results obtained with cosmic muons with the 5 different eco-friendly gas mixtures. The efficiency (left panel) requirements of 96% indicated by the upper red dotted line were reached for all mixtures but with a shifted working voltage of up to 2 kV. The cluster size shows slightly higher values for the mixtures without SF\(_6\) which is identical with TFE mixtures without SF\(_6\). However, for the time resolution (right panel) only the mixtures containing SF\(_6\) could show somewhat convincing results.

Fig. 2
figure 2

Left panel: efficiency and cluster size as function of applied high voltage for various gas mixtures based on HFO. Right panel: time resolution as function of applied high voltage for various gas mixtures based on HFO

Full size image

In conclusion: based on the presented results, a change to a more eco-friendly gas mixture could be considered, even though a better time resolution would be desirable.

3 Aging of MRPCs operated at standard gas mixtures at high charged particle flux

As stated in the introduction, CBM aims to operate at 10 MHz interaction rate. This demands stable operating detectors at charged particle fluxes up to 50 kHz/cm\(^2\). During high-rate beam tests (with fluxes above 30 kHz/cm\(^2\)) at the SIS18 machine at GSI/Germany, two kinds of phenomena leading to an increase in dark rate and dark detector current were observed. First a temporal aging effect emerged as a rapid increase (within several hours) in dark current and rate of up to five times the original value, which decreases exponentially and returns back to its almost nominal value within a few weeks. Second, a steadily increasing dark rate and current was observed, which we refer to as permanent aging. After each high-rate beam time, we noticed that the chambers do not recover fully and show more and more detector noise. Since we read out the detector on both ends of the readout strips, the noise source could be localized. From the middle panel of Fig. 3, it became evident that the noise is generated homogeneously very close to the spacers (fishing lines) as is visible when compared to the left panel of Fig. 3. In order to investigate the noise source in more details, a few MRPCs were opened, and a white substance was found on the touching surface of the spacers (conf. right panel of Fig. 3). The chemical analysis of the substance revealed a high concentration of fluorine and sodium. The fluorine clearly has to come from reaction products (F-radicals and F\(^-\)) of the gas, while the sodium is a compound of the glass material. That these impurities, which are responsible for the dark current, are preferentially developed close to the spacers is not surprising or unexpected. Close to the touching point of the round spacer (on the flat glass surface) where the gas gap becomes very small the electric field is substantially higher (simulations show up to a factor of 4), and therefore, the chemistry in this location is enhanced. In addition, a rapid exchange of the gas close to the spacers is suppressed, which generates a large retention time for radicals close to this area. In order to investigate the permanent aging in more details, a gamma irradiation test was performed at IRASM Bucharest/Romania [29] where similar chemical composition of the deposits was found [30].

Fig. 3
figure 3

Left panel: open MRPC counter-spacers are visible arranged in a triangular manner. Middle panel: position reconstruction of the dark rate hits. A clear pattern of the fishing lines is visible. Right panel: traces of a white substance below spacers were found. Chemical analysis revealed the composition of the substance

Full size image

The speed of permanent aging seems to correlate linearly with the received dose during operation, which is a severe problem when operating MRPCs at high particle fluxes. A mitigation strategy could be to guide the gas flow primarily inside the gaps in order to enhance the gas exchange more rapidly. On the other hand, it is well known that bonds of eco-friendly gases break up more easily (which is actually the reason why these gases have a lower GWP) which leads to an enhanced F\(^-\) and radical production [31].

In conclusion: it is very likely that eco-friendly gas mixtures will lead to a strongly increased permanent aging of the detector at high particle fluxes and a shortening of its lifetime.

4 Gas recuperation system for CBM-TOF

A very promising method to reduce the environmental impact during operation of MRPC systems could be the recycling of each gas of the mixture. The idea here is to separate the gases first, eventually clean them from other impurities and re-inject them into the gas circuit. This implies high investment costs for the recuperation system in the beginning but saves running costs later on by reusing the recycled gas and only the losses due to gas leaks in the system have to be procured. First thoughts in this direction were done at CERN [32] and even a first prototype unit recuperating the TFE is in operation at CMS [33]. However, the separation of gasses from the standard mixture is not straight forward. A combination of TFE and isobutane forms an azeotropic mixture, which means that these gasses cannot be separated by distillation to a purity level of 100% [34]. Furthermore, SF\(_6\) has its triple point at a temperature of \(-\)49.60 \(^\circ \)C and a pressure of 2.314 bar [35], i.e., it cannot be separated from TFE by fractional distillation at normal pressure. Therefore, a pressurized distillation unit or a solid phase cold trap has to be considered if a recycling of SF\(_6\) is desired. For these reasons, the CBM-TOF group decided to abandon the isobutane and operate the MRPCs with a gas mixture of 97.5% TFE and only 2.5% SF\(_6\). Figure 4 shows the first layout of a closed loop gas system with a recuperation system attached designed for CBM-TOF.

Fig. 4
figure 4

First layout of the CBM-TOF gas system including a gas recuperation system

Full size image

The aim is to separate first the TFE from the SF\(_6\) and potentially other impurities via a pressurized condenser and in a second stage freeze out the SF\(_6\) in the solid phase via cold traps. The system is designed to operate three traps in a consecutive manner to allow for loading and emptying the traps simultaneously. The recycled gasses are stored in vessels (VT6 and VT7) and re-injected in the system.

In conclusion: a recuperation system could be an alternative to eco-friendly gasses in order to minimize the environmental impact. However, it should be considered that the industrial production of TFE and SF\(_6\) is going to be reduced or possibly even stopped in the future, which will lead to an increase in price. Nevertheless, since eco-friendly gasses will generate similar operating costs, a recuperation system could be considered for these gasses as well.

5 Summary

The CBM experiment is one of the main research pillars at FAIR. It aims to operate at an interaction rate of up to 10 MHz, which demands radiation hard and rate capable detectors. For charged hadron identification, a 120 m\(^2\) TOF wall composed of MRPCs with rate capabilities of up to 50 kHz/cm\(^2\) is planned to be installed at 8 m from the target. Even though detector performance results obtained with eco-friendly gasses mixtures were promising in terms of efficiency and time resolution, it is a measured fact that gas molecules with low GWP get destroyed much more effectively under RPC operation conditions and the fluorine-ion (F\(^-\)) production is increased in comparison with the standard gasses. Similar chemical processes are expected to be present during the MRPC operation, i.e., at higher electric fields as well. Keeping in mind the observed permanent detector aging at high particle fluxes (which turned out to be caused by fluorine compound deposits close to the spacers) which would be clearly strongly enhanced using eco-friendly gasses, the TOF group decided to continue to use a mixture of 97.5% TFE and 2.5% SF\(_6\). The TOF group aims to recuperate both gasses and re-inject them in the circuit. This should reduce the operating costs and lead toward an environmental friendly operation of the CBM-TOF system.