R&D towards the CMS RPC Phase-2 upgrade

The high pseudo-rapidity region of the CMS muon system is covered by Cathode Strip Chambers (CSC) only and lacks redundant coverage despite the fact that it is a challenging region for muons in terms of backgrounds and momentum resolution. In order to maintain good efficiency for the muon trigger in this region additional RPCs are planned to be installed in the two outermost stations at low angle named RE3/1 and RE4/1. These stations will use RPCs with finer granularity and good timing resolution to mitigate background effects and to increase the redundancy of the system.

After the more than two years lasting rst Long Shutdown (LS1), the Large Hadron Collider (LHC) delivered its very rst Run-II proton-proton collisions early 2015.LS1 gave the opportunity to the LHC and to the its experiments to undergo upgrades.The accelerator is now providing collisions at center-of-mass energy of IQ e and bunch crossing rate of RH wrz, with a peak luminosity exceeding its design value.During the rst and upcoming second LHC Long Shutdown, the Compact Muon Solenoid (CMS) detector is also undergoing a number of upgrades to maintain a high system performance [1].
From the LHC Phase-2 or High-Luminosity (HL-LHC) period onwards, i.e. past the third LHC Long Shutdown (LS3), the performance degradation due to integrated radiation as well as the average number of inelastic collisions per bunch crossing, or pileup, will rise substantially and become a major challenge for the LHC experiments, like CMS that are forced to address an upgrade program for Phase-II [2].Simulations of the expected distribution of absorbed dose in the CMS detector under HL-LHC conditions, show in gure 1 that detectors placed close to the beamline will have to withstand high irradiation, the radiation dose being of the order of a few tens of qy.
The measurement of small production cross-section and/or decay branching ratio processes, such as the Higgs boson coupling to charge leptons or the B s 3 + decay, is of major interest and specic upgrades in the forward regions of the detector will be required to maximize the physics acceptance on the largest possible solid angle.To ensure proper trigger performance within the present coverage, the muon system will be completed with new chambers.In gure 2 one can see that the existing Cathode Strip Chamber (CSC) modules will be completed by Gas Electron Multipliers (GEM) and Resistive Plate Chambers (RPC) in the pseudo-rapidity region I:T < jj < P:R to complete its redundancy as originally scheduled in the CMS Technical Proposal [3].RPCs are used by the CMS rst level trigger for their good timing performances.Indeed, a very good bunch crossing identication can be obtained with the present CMS RPC system, given their fast response of the order of I ns.In order to contribute to the precision of muon momentum measurements, muon chambers should have a spatial resolution less or comparable to the contribution of multiple scattering [1].Most of the plausible physics is covered only considering muons with p T <IHH qe thus, in order to match CMS requirements, a spatial resolution of y(few mm) the proposed new RPC stations, as shown by the simulation in gure 3.According to preliminary designs, RE3/1 and RE4/1 readout pitch will be comprised between 3 and T mm and 5 -partitions could be considered.2 Irradiation tests at the CERN GIF++ In the very forward region of CMS, i.e. close to the LHC beam line, the third endcap disk will be subjected to a total expected background rate of the order of H:T krz=m 2 according to gure 4 requiring solutions for the new RPC stations with stable, high performance in time in a high radiation environment for the new RE3/1 and RE4/1 stations [2].Such rate is equivalent to an integrated charge of approximately I g=m 2 during the lifetime of the detectors assuming a mean charge deposition hqi per avalanche of PH pg and a safety factor of 2.
To ensure their robustness, CMS present muon system detectors and new R&D eorts will be irradiated and monitored.The GIF++ at CERN is a new gamma irradiation facility to test detectors for the HL-LHC program.High energy beam with momentum up to IHH qe= is provided and combined with a IR fq 137 Cs source.Performance of RPCs will be thoroughly tested with high radiation background.At HL-LHC, to integrate QHHH f 1 at a luminosity L a S ¢ IH 34 m 2 ¡ s 1 , an eective time T ef f a T ¢ IH 7 s is needed.
Using equations 2.1 and 2.2, and an acceleration factor AF a P, an irradiation time T irr of approximately 17 months will be necessary to reproduce a charge deposition similar Figure 4. Simulated particle hit rate as a function of radius at the RE3/1 station, assuming a 0.001 sensitivity to neutrons, 0.01 to photons and 1 to charged, ionizing particles.
to HL-LHC.RPC life time is dependant on the total integrated charge over time that can degrade the material and components inside the detector, and contribute to produce HF [4].Test in GIF++ will help us understand the HF production through time in RPCs using uoride-rich gas mixture as well as their overall ageing.

Investigated forward RPC technologies
As RPCs are resistive detectors, increasing the rate capability of the new generation via a reduction of their electrode's resistivity is a rst possibility to be investigated in order to accelerate the charge recombination at the level of the electrode's surface.
A reduction of resistivity should be coupled with a reduction of mean charge deposition per avalanche.Equipping RPCs with more sensitive front-end electronics and moving a part of the amplication to the electronics would allow to work at lower voltages.Another solution would be to improve the current CMS RPC design.For example, changing the number of electrodes or their thickness could yield a better ratio of induced signal to moving charge.
By combining these dierent techniques, a reduction of the voltage drop over the electrodes is expected with improvements on both rate capability and ageing of the detector.

Electrode materials
Materials that are being used in RPCs are listed into table 1.The materials presented here are the Bakelite and Low Resistive Silicate (LRS) glass considered by CMS together with other doped glass and ceramics.HPL being industrially produced oers lower costs and bigger material surfaces allowing to build wide area detectors, whereas glass and ceramics are still prototype materials produced locally at higher costs and small surfaces.IPNL-LLR-Tsinghua [6,7] Vanadate glass IH 4 to IH 16 Coe College-ANL-University of Iowa [8] SiC based ceramics IH 7 to IH 12 HZDR [9] Ferrite ceramics IH 6 to IH 13 CSIC-USC [10] Table 1.List of materials used for improved RPC prototypes.A resistivity range is given for Vanadate glass, SiC based ceramics and Ferrite ceramics for which the production process allows a ne tuning of the resistivity.
tunable range of resistivity values, while Bakelite has only limited resistivity range and with its porous surface usually requires a linseed oil treatment.

Front-end electronics
Only changing the electrode resistivity to make RPCs faster won't help in reducing the ageing eects to ensure smooth performance over a long time period.A main contribution to the ageing is due to the amount of deposited charge per avalanche created in the gas gap.Being able to reduce this mean charge deposition will lead to a reduction the total integrated charge over long periods.This can be achieved via more sensitive low-noise front-end electronics.
New prototypes of front-end electronics using SiGe technology have been proposed as shown in gures 5 & 6.The rst gure presents a prototype of new preamplier developed by Cardarelli and others [11].This low-noise preamplier can be fused into the current CMS front-end electronics ASIC as replacement of the actual preamplier [12].Tested on a CMS RPC, this electronics permitted a shift of the eciency curve of RTH to lower values corresponding to a average charge reduction from PH pg to Q p [2].
Figure 6 is a block diagram of the 16-channel PETIROC ASIC developed by OMEGA [13] for Time of Flight applications.This type of electronics is now being adapted for usage with multigap RPCs with high time resolution.The prototype oers the possibility to use 3 detection thresholds, enabling a semi-digital readout, and to read-out both strip ends.Coupled with the timing performance of multigap RPCs, a good position information along the strips can be obtained.

Chamber design
Choosing a better performing RPC design, with optimized electrode or gas gap thickness and various number of gaps, to combine with the previously presented front-end electronics prototypes can be investigated to further reduce the mean charge deposition per avalanche and thus the charge recombination time and the related detector dead time.
A bakelite based multigap solution is currently under development at KODEL [5]. Figure 8 shows a double bi-gap HPL RPC, composed of 2 bi-gap RPCs using a design similar to that of the CMS RPC with P mm electrodes and gas gaps and the results obtained with and without -irradiation.Using the standard CMS front-end electronics, a shift of    approximately PHH towards higher voltages can be observed under Q krz=m 2 irradiation rate.
Combining low resistivity electrodes with low noise electronics leads to the possibility to use single gap RPCs.Such a prototype made out of LRS glass and using a similar frontend electronics as the PETIROC called HARDROC, has been proposed and and has only a limited eciency drop compared to regular oat glass for increasing particle rate [6], as shown in in gure 9.
Finally, multigap prototypes made out of the same LRS glass have been made and tested with PETIROC electronics [7].The very high timing precision of such RPC designs is shown gure 10.

Figure 1 .
Figure 1.Absorbed dose in the CMS cavern after an integrated luminosity of 3000 fb 1 .R is the transverse distance from the beamline and Z is the distance along the beamline from the Interaction Point at Z=0.

Figure 2 .
Figure 2. A quadrant of the muon system, showing DT chambers (yellow), RPC (light blue), and CSC (green).The locations of new forward muon detectors for Phase-II are contained within the dashed box and indicated in red for GEM stations (ME0, GE1/1, and GE2/1) and dark blue for improved RPC stations (RE3/1 and RE4/1).

Figure 3 .
Figure3.RMS of the multiple scattering displacement as a function of muon p T for the proposed forward muon stations.All of the electromagnetic processes such as bremsstrahlung and magnetic eld eect are included in the simulation.
Glass and ceramics have smooth surfaces, naturally oering uniform electric elds, and oer a

Figure 5 .
Figure 5. Left: block diagram of INFN amplier where R1 is the input impedance resistance.C1C2 and R1R2 are the conditions for the circuit to work properly.Right: eciency plateaus for the cases of the standard CMS RPC front-end electronics and new INFN prototype electronics.A lower charge crossing the RPC gap is indicated by the shift of the plateau toward lower voltages leading to an improved rate capability of the RPC.

Figure 7 .
Figure 7. Picture of a strip board designed for a RE1/1 station using 16-channel PETIROC ASICs and 25 ns TDC.

Figure 8 .
Figure 8. Left: view of a multigap HPL RPC prototype.Right: eciency plateau and cluster size as function of operating voltage with (open symbols) and without (solid symbols) a 3 kHz=cm 2 -irradiation.

Figure 9 .
Figure 9. Left: schematic drawing of GRPC with electrodes made of low resistive silicate glass or oat glass.Right: eciency vs rate for dierent RPC.The orange line corresponds to the GRPC with oat glass.The semi-conductive chambers are represented with dierent colors.4ConclusionIn view of the HL-LHC, the CMS RPC group is designing and testing new RPC prototypes, using new dierent electrode materials, geometries or electronics.The nal technique adopted for the next upgrade of the CMS Muon System may combine the benets of the dierent approaches.Several options for new front-end electronics are being developed, and studies are ongoing to integrate the new electronics into the CMS Trigger and DAQ system.Ageing and rate capability tests of the new prototypes are ongoing at the CERN GIF++ irradiation facility.CMS is anticipating the installation of two new RPC stations during the upcoming third LHC Long Shutdown.

Figure 10 .
Figure 10.Structure of 6-gap (top) and 10-gap (center) resistive plate chambers from Tsinghua University.Time resolution of 6-gap and 10-gap resistive plate chambers (bottom) as a function of high voltage at an incoming muon rate of 1:2 kHz=cm 2 .