Summer Research Fellowship Programme of India's Science Academies

Fabrication and characterization of resistive plate chamber

P M Patil

Shivaji University, Vidya Nagar, Kolhapur 416004

Vivek Datar

Senior Professor, India-based Neutrino Observatory Cell, Tata Institute of Fundamental Research, Homi Bhabha Road, Mumbai 400005 India


The Resistive Plate Chamber (RPC), introduced in 1981 by R. Santonico and R. Cardarelli, as a practical alternative to localized discharge spark counters, utilizing a constant and uniform electric field produced by two parallel electrode plates which are made up of a material with high bulk resistivity. Its working principle is based on the detection of gaseous ionization produced by charged particles traversing the active area of the detector, large avalanche of electron under a strong uniform electric field. A gas mixture of R134a, iso-C4H10 and SF6, which is ionized by charged particles traversing the detector, is flown through the gap between the electrodes. The electrodes which we are using in this experiment are commercially used float glass electrodes, coated with a layer of graphite on the outer side. Here, I fabricated one single gap glass RPC prototype of 30cm x 30cm area and studied its operation in avalanche mode. I did its characterization viz- I-V characteristics, noise rate measurement and efficiency test. In TIFR Mumbai, there is a detector stack of 12 RPCs of 100cm x 100cm in area along with a state of art, front end electronics, trigger, data acquisition, control and monitoring systems. There is also a gas system for the mixing and the distribution of gas in controlled manner inside the gap of RPCs. This is a Mass Flow Controller (MFC) based on-line gas mixing and multichannel distribution system. It mixes all three gases in specified amounts and distribute it simultaneously to 12 RPCs in controlled manner. I did the calibration of that gas system by water displacement method and using these plots, we obtain MFC calibration equations for individual gases, using which we set their required proportions in the gas mixture according to the total flow rate of the mixture.

Keywords: INO, RPC, ICAL, MFC, data acquisition


INO India Based Neutrino Observatory
RPC Resistive Plate Chamber
ICAL Iron- Calorimeter
MFC Mass Flow Controller
DAQ Data Acquisition
NIM Nuclear Instrumentation Module



The resistive plate chamber is a particle detector utilizing a constant electric field produced by the two parallel plates having high bulk resistivity of about 1012 Ω-m, whose working is based on the process of ionization. When the charged particle travels through the detector it will ionize the gas between the plates, subsequently an avalanche is produced, which results in the local discharge of the electrodes. This discharge is limited to a tiny area of about 0.1cm2, due to high resistivity of the glass electrode and the quenching characteristics of the gas. The discharge induces an electrical signal on external pickup strip on both sides, orthogonal to each other, which can be used to record the location and time of detection. The two Mylar sheets are placed in between the pickup panels and the electrode, for the electric signal to be processed by the read-out electronics and analyzed. 

The RPC’s are preferred over scintillators because of the following advantages [2]:-

1) They can be made to have a large area but at a minimal material cost.

2) These are easy to assemble and they possess simple read-out electronics.

3) They exhibit better time resolutions than scintillators and long term stability.

4) Moderate position resolution and give good detection efficiency.

Objectives of the Research

* Fabrication of RPC

* Characterization of RPC

* Gas Calibration


The resistive plate chamber is a particle detector and will be used as an active element in the Iron-Calorimeter at the India based Neutrino Observatory, ICAL which will be used to study of atmospheric neutrinos [2]. RPCs are fast, planar, rugged and low-cost gas detectors which are and being used extensively in a number of high energy and astro-particle physics experiments. They find applications for charged particle detection. It has excellent position and time resolutions. Almost all big high energy physics experiment presently operational use RPCs either for trigger or timing measurements, whereas INO is going to use this for both triggering and tracking of charged particles [2]. Single and double gap RPCs have also found application in cosmic ray experiments as well as in Astro-particle physics [2].



The glass RPC is consisting of two parallel electrodes made up of float glass with a volume resistivity of about 1012 Ω-cm [2]. The two electrodes, 2-3mm thick, are mounted 1-2mm apart by means of highly insulated spacers. A suitable gas mixture is flown at the atmospheric pressure through the gap while an appropriate electric field is applied across the glass electrodes through a resistive coating on their outer surfaces [2]. An ionizing charged particle traversing the gap initiates an avalanche or streamer in the gas volume that results in a local discharge of the electrodes. This discharge is limited to a tiny area of about 0.1cm2 due to the high resistivity of the glass electrodes and the quenching characteristics of the gas. The discharge induces an electrical signal on external pickup strips on both sides orthogonal to each other, which can be used to record the location and time of ionization. The discharge area recharges slowly through the high resistivity glass plates and the recovery time is about secs [2]. The duration of discharge is typically ~ ns. This discharge is quenched by the following mechanisms:

 1.) Prompt switching off of the field around the discharge point, due to the large resistivity of electrode.

2.) UV photon absorption by the quencher (Iso-butane is a common quencher) preventing secondary discharges from gas photo-ionisation.

gas gap (2).jpg
    Construction schematic of a Resitive Plate Chamber.
    avalanche (2)_1.jpg
      Schematic of the avalanche formation with variation in the electric field caused by the avalanche charge carriers inside the RPC E0 is uniform electric field applied across the electrodes. E1,E2 and E3 are the electric field at tip, centre, and tail of the avalanche charge distribution.
      -ve pulse_1.png
        Signal at X strips
        +ve pulse.png
          Signal at Y strips

          3) Capture of outer electrons of the discharge due to the gas with high electron affinity (e.g. SF6), which reduces the size of the discharge and its transverse dimensions [2].

          Because of the large difference between the duration of discharge and recovery time, the electrode plates behave like insulators so that only a limited area of ~0. 1cm2 around the discharge point remains inactive for the dead time of the detector. The motion of electron during avalanche process induces an electrical pulse which is picked up by the pick-up panels, made up of copper strips, through graphite painted high voltage electrodes. The velocity of positive ions is very low and consequently the induced signal is negligible and for all practical purpose induced signal due to the motion of positively charged ions are neglected. The electrons from the same event produces induced signal in both pickup panels and that is opposite due to the motion of electron opposite with respect to two plates. The copper strips are insulated against the high voltage by thin insulator layer, e.g., Mylar sheets. The electric signal is then picked up by the “read-out electronics” and analysed.

          Fabrication of RPC

          Resistive Plate Chambers (RPC) can be fabricated and tested easily. The following steps followed to fabricate the RPC gas Gap.

          a. Cutting and Cleaning of glass

          The RPCs are fabricated using commercially available float glass of thickness 3mm having graphite coating at the outer side [1]. Here I use the glass which is already cut in size 30cm x 30cm with conductive coating on its one side. Its edges are chamfered to make 45° angle. The conductive coating not only increases the conductivity of the glass but also allows for the uniform application of high voltage over the area of the RPC. The conductivity is chosen in such a way that it can produce nearly uniform electric field throughout the chamber, without much screening the induced signal, which is picked up by pickup panel and passed to electronics. The glass sheets are cleaned by alcohol followed by labolene and distilled water. The glasses are then left for drying. The next step is to introduce the button and side spacers. The edge spacers are as shown figure a. The edge spacers are fitted in with a nozzle creating a gas inlet. The button spacers and the edge spacers ensure a gap of 2mm between the glass plates.

            From left Side Spacer, Edge Spacer, Button Spacer and Nozzel.
            ino3 (2).jpg
              Squarer Jig.

              The four chamfered edges are fitted with corner edge spacers. Side gas nozzles used to create one inlet and one outlet. The spacers were cut according to the size of the RPC. Thus before going to make the glass gap, it is important to measure resistivity of both top and bottom glass, this will be done using a square jig of copper and brass (5cm x 5cm) as shown in figure b. The measured resistivity in a 5 by 5 grid is shown in the following two graphs one for vertical and other for horizontal values. Optimum surface resistivity is ~1-1.2MΩ/ÿ.V

                Surface Resistivity Plots

                b. Gluing the glass with the spacers

                The glue is used for stabilizing the glass plates with the spacers was 3M Scotch-weld epoxy adhesive DP190 Gray in a duo-pack cartridge. Since the size of the RPC is small, it was decided to use four button spacer in the between the plates. A small drop of glue binds the buttons with one sheet of glass. The other glass plate is rested on top of the buttons so that glass sits neatly in line with the one under it. The glue is then applied to the gap between the spacer and glass. After about 8 hours, the RPC is turned over and glue is applied on other side. After another 8 hours RPC glass gap is ready for the leak test.

                  RPC after gluing

                  c. Gas leak test

                  To test if there are any leaks during the application of glue a pressurized gas leak test was done. The RPC is filled with R134a/N2 gas which is connected to the RPC with a needle valve which helps in slowly increasing the pressure in the RPC. A pressure excess of 30 mm of water column was maintained using Adriano based leak test setup; it was found that the RPC has to keep at this pressure for almost 8 hours revealing that there are no leaks [3].

                  0001 (1).jpg
                    Plots of temperature, pressure and pressure difference between atmospheric pressure and pressure inside the RPC with time.

                    The first graph shows that there is the variation in temperature with time, because the room in which we tested our RPC has an AC which creates such temperature behavior. Due to that variation, outside pressure also varies as shown in second graph, red line shows the variation in atmospheric pressure and black line shows pressure variation inside the RPC and third plot shows the variation in pressure difference between atmospheric pressure and inside RPC pressure. From the third graph pressure difference is almost constant for 12 hours, shows that leak test of RPC is successful, there is no leak in RPC.

                    d. Pickup Strips and characteristic impedance

                    The RPC is now sandwiched between two honeycomb pickup panels placed orthogonal to each other and then packed in an aluminum case. The pickup panel consists of 8 strips copper foil on one side of a layer of 5mm of foam and aluminum on the other side. Each strip is of width 2.8 cm with a gap of 0.2cm in between two adjacent strips. Each strip is terminated with a 50Ω impedance to match the characteristic impedance of the preamplifier. A layer of Mylar sheet of thickness 100 μm is placed between the resistive coating and the pickup panel to provide insulation.

                    Characterization of the RPC

                    The RPC is then kept on an aluminum plate and is connected to the pre amplifier board. The RPC is introduced into a system of continuous gas flow for more than a day so as to flush out the air from it [2]. The gas mixture used is given in the table.

                    Gas Constituents in Gas Mixture.
                    Gas Constituents Percentage
                    R134a 95.2
                    Isobutane 4.5
                    Sulphur Hexafluoride (SF6) 0.3  

                    Tetrafluoroethane (known as Freon), which is widely used, has shown these specifics. But here we use R134a (as Freon) which is eco-friendly and used as a medium for ionization by charged particles. Isobutane acts as the quenching gas which absorbs the extra photons that are generated. We use SF6 (Sulphur-hexafluoride) to control the excess number of electrons and help to contain and localize the signal to have precise position information of the charged particle.

                    a. I-V characteristics and noise rate

                    In this setup, the gas mixture is fixed for the operation of RPC in avalanche mode. To have the best performance, first we need to find the operating HV. This can be done using information of dark current in the RPC and the noise rate of RPC signal as a function of applied HV. The gain of the preamplifier is ~80 and the discriminator threshold of signal is~30mV. Variation of dark current and noise rate in a few strips with applied HV will be used to decide whether the chamber is useful for any experiment or there is a need to change/correct a component. The typical operational HV of this chamber is ~10kV.

                      I-V Characteristics of RPC.

                      Figure 10 shows the IV Characteristics of the fabricated RPC. From that graph, it is seen that the operational HV for that RPC is ~11kV

                      b. Measuring Efficiency

                      The final test of the RPC is done by measuring the detection efficiency of charged particles, e.g., cosmic ray muon. Typically, efficiency is a measured with the help of scintillator paddles. Scintillator paddles have wavelength shifting optical fibers/wave guide which are connected to a photomultiplier tube. The photons generated in the scintillator are propagated in all directions. All sides except the side of the PMT, was covered with highly reflective material, e.g., Tyvek, thus eventually all photons, except the loss due to self-absorption and loss at surface, are propagated towards the photomultiplier tube which converts them into electrical signals. These paddles are arranged in line with a particular strip of the RPC pickup panel. The presence of a muon is ensured by the trigger paddles and efficiency of the RPC is determined by the fraction of events, where RPC signal is above the threshold value of the discriminator. Before doing this we must ensure that the noise rate of the RPC is small. During this experiment, both noise rate/dark current and efficiency are measured together. The following setup is used to do that.

                      IMG_20190705_011708 (2).jpg
                        Set up for efficiency test.

                        Three scintillator paddles are arranged on top and below of the RPC under test as mentioned above. Care was being taken that the alignment is accurate to avoid inefficiency due to wrongly triggered events where the muon does not pass through the RPC. These paddles PMTs are supplied with high voltage from the HV distribution box. HV for RPC is supplied separately. Signal outputs of paddels are directly taken to the discriminator module. Threshold adjustment for the discriminator and width adjustment for the pulse shaper can be done in the same module. RPC pulses are small - of the order of few mV. So, RPC pulses are amplified by using a preamplifier with a gain of about 80 and then fed to the discriminator. The following steps were adopted [2]:

                        1. Before starting the experiment make sure that the RPC is flushed with appropriate gas for sufficient time.

                        2. Ramp Up the Voltage of the RPC detector in steps of 100V for both the polarities.

                        3. At each step, wait for few minutes so that the current stabilises.

                        4. Connect the raw signal output from RPC detector to oscilloscope and look for appropriate pulse and noise level.

                        5. Remove cables from oscilloscope and connect to AND logic module to get following logical signals.

                        a. 1f = RPC

                        b. 3f = P1 . P2 . P3

                        c. 4f = RPC . P1 . P2 . P3

                        Delayed_pulse (2).jpg
                          Typical signals seen on oscilloscope during measurement of efficiency of RPC

                          6. Efficiency is calculated as (4f) ×100/(3f) at each HV operation and is shown in the following plot; initially the efficiency is zero and gradually increased to nearly 100%.

                          7. Noise Rate is calculated as number of counts measured by RPC divided by the corresponding counting time. This noise rate is plotted against high voltage as shown below. From that graph it is seen that noise rate increases with increase in high voltage.

                            Efficiency of RPC as a function of applied voltage
                              Noise rate as function of applied voltage .
                              prajakta_patil (1)_1.png
                                Pixel wise efficiency of RPC

                                From that plot it is seen that most of the area of RPC has an efficiency of 40 to 60%.

                                Gas calibration

                                The proper and efficient functioning of the RPC detector requires pre-mixing or on-line mixing of individual gases, in an appropriate proportion, together with a controlled flow of mixed gases through the detector [1]. This is achieved by using gas mixing and multichannel distribution system. In TIFR Mumbai, there is a detector stack of 12 RPCs of 100cm*100cm in area. For that stack a Mass Flow Controller (MFC) based on-line gas mixing and multichannel distribution system is used. MFCs are calibrated to control a specific type of gas in a particular range of flow rate. I did its calibration again to check whether the gas inside the stack flow properly or not. Water displacement method is used for its calibration.

                                ino7 (2).jpg
                                  Photograph of second generation gas mixing and multi-channel distribution system
                                  ino5 (2).jpg
                                    Schematic diagram of MFC calibration by water displacement method

                                    The MFC need a 0-5 volts DC signal for their operation, and flow through them increases linearly with the applied voltage [1]. Its calibration is done by using inverted water filled measuring jar, in a container filled with water. The gas tube from the each MFC output is carefully inserted, and the gas is collected in the jar from one known level to another known level and time required for that is recorded [1]. From that we calculated the flow rate of the gas in sccm (standard cubic centimeters per minute) for different set values of gas flow rate on the gas unit while also noting, the set actual flow rate and corresponding control voltage given to the MFC and its output voltage from the MFC. Following are the plots of calibration for three gases: R134a, SF6, and Isobutene

                                    1. For R134a

                                      Plot of Actual Flow Rate Vs Set Value for R134a.
                                        Plot of Output Voltage Vs Set Voltage for R134a.
                                          Comparision with 2016th data.

                                          *The plot of actual flow rate vs. set value shows linear behavior.

                                          *Output voltage also varies linearly with set voltage as shown in 2nd graph.

                                          *From that comparison plot, it is seen that recent data and 2016th data matches.

                                          This shows that MFC used for R134a works properly.

                                          2. For Isobutane

                                          ISO1 (2).png
                                            Plot of Actual Flow Rate Vs Set Value for Isobutane.
                                            ISOBUTANE2 (3).png
                                              Plot of Output Voltage Vs Set Voltage for Isobutane.
                                                Comaparision Plot.

                                                *The plot of actual flow rate vs set value shows linear behavior.

                                                *Output voltage also varies linearly with set voltage as shown in 2nd graph.

                                                *From that comparison plot, it is seen that recent data and 2016th data matches.

                                                This shows that MFC used for Isobutene works properly.

                                                3. For SF6

                                                  Plot of Actual Flow Rate Vs Set Value for SF6
                                                    Plot of Output Voltage Vs Set Voltage for SF6
                                                      Comparison Plot.
                                                        Plot of Actual Flow Rate Vs Set Value in cc/mm for SF6 (bottom reading).

                                                        *The plot of actual flow rate vs set value is not properly linear.

                                                        *Output voltage varies linearly with set voltage as shown in 2nd graph.

                                                        *From that comparison plot, it is seen that recent data and 2016th data doesn’t match.

                                                        *That’s why I recalibrate it again.

                                                        For lower values, shown in 4th plot, that plot shows linear behavior.

                                                        This shows that MFC used for SF6 sometimes works properly, sometimes not.

                                                        Hence, there is a need to replace that MFC.

                                                        Using the plots shown above we obtain the MFC calibration equations for individual gases as shown in following table, using which we set their required proportions in the gas mixture according to the flow rate of the mixture.

                                                        MFC Calibration equations.
                                                        Gas Equations Percentage Required Flow Rate in SCCM Set Values in SCCM
                                                        R134a y=1.0319x+0.6143 95.2% 17.13 16.000
                                                        Isobutene y=0.9247x+0.5179 4.5% 0.75 0.8832
                                                        SF6 Y=1.0716x+0.0041 0.3% 0.054 0.0465


                                                        [1] S. Bheesette, PhD thesis (2009), Design and Characterisation Studies of Resistive Plate Chambers.

                                                        [2] R. Shinde, E. Yuvraj, Fabrication and Characterization of Glass Resistive Plate Chamber RPC.

                                                        [3] S. Mondal, V. Datar , G. Majumder, N.K. Mondal, K.C. Ravindran , B. Satyanarayana, Leak Test of Resistive Plate Chamber Gap by Monitoring Absolute Pressure.


                                                        I would like to express my sincere gratitude to my guide Prof. Vivek Datar for giving me an opportunity to do this summer project under his guidance. I also want to give my thanks to Prof. Satyanarayana Bheesette and Prof. Gobinda Majumder who also guided me during my project work.

                                                        My sincere thanks also goes to all scientific officers: Mandar Saraf sir, Ravindra Shinde sir and E. Yuvraj sir who have been always supportive and helpful for my project work. I also thankful for the help given from Neha Panchal and Mamta Jangra

                                                        Last, but not the least, I would like to thank my parents who support me every time throughout my life.

                                                        Written, reviewed, revised, proofed and published with