Small-pad resistive Micromegas for high rate environment: Performance of different resistive protection concepts

Nuclear Inst. and Methods in Physics Research, A 936 (2019) 408–411

Contents lists available at ScienceDirect

Nuclear Inst. and Methods in Physics Research, A journal homepage: www.elsevier.com/locate/nima

Small-pad resistive Micromegas for high rate environment: Performance of different resistive protection concepts M. Alviggi a,b , V. Canale a,b , M. Della Pietra a,b ,∗, C. Di Donato c,b , E. Farina d , S. Franchino e , P. Iengo f , M. Iodice g , L. Longo i , F. Petrucci g,h , E. Rossi g,h , G. Salamanna g,h , G. Sekhniaidze b , O. Sidiropoulou f,j a

Università di Napoli Federico II, Naples, Italy INFN Napoli, Naples, Italy c Università di Napoli Parthenope, Naples, Italy d Università di Pavia and INFN Pavia, Pavia, Italy e Ruprecht-Karls-Universität Heidelberg - Kirchhoff-Institut für Physik (KIP), Heidelberg, Germany f European Center for Nuclear Research, CERN, Meyrin, Switzerland g INFN Roma Tre, Rome, Italy h Università di Roma Tre, Rome, Italy i INFN Sezione di Lecce and Dipartimento di Matematica e Fisica, Università del Salento, Lecce, Italy j Bayerische Julius Max. Universität Wuerzburg, Würzburg, Germany b

ARTICLE

INFO

Keywords: Gas detectors Tracking detectors Micromegas Micro-pattern gas detectors

ABSTRACT We present the development of resistive Micromegas with small pad readout aiming at precision tracking without efficiency loss up to several MHz/cm2 . Several prototypes have been built with the spark protection resistive layer deposited with different techniques: a pad-patterned embedded resistor layout with screen printing, and a uniform layer by sputtering (Diamond Like Carbon structure). All detectors consist of a matrix of 48 × 16 rectangular shaped pads with a pitch of (1 × 3) mm2 . The active surface is (48 × 48) mm2 with a total number of 768 channels, routed off-detector for readout. Characterization and performance studies of all prototypes have been carried out by means of radioactive sources, X-rays, cosmic rays and high energy particle beams. A comparison of prototypes with different resistivity layout will be presented.

Contents 1. 2. 3. 4. 5.

Introduction .................................................................................................................................................................................................... Prototypes construction details .......................................................................................................................................................................... Detector characterization and performance ......................................................................................................................................................... Test beam results ............................................................................................................................................................................................. Conclusion ...................................................................................................................................................................................................... Acknowledgments ............................................................................................................................................................................................ References.......................................................................................................................................................................................................

1. Introduction Micromegas (MM) [1] is now a mature technology for High Energy Physics experiments. Resistive strips MM [2] will be used in the New Small Wheel ATLAS upgrade operating at moderate hit rate estimated to be of the order of several tens of kHz/cm2 [3]. Nevertheless, upgrades of forward muon detectors of LHC experiments as well as in experiments at future colliders will require rate capability up to few MHz/cm2 . In

408 409 409 410 411 411 411

order to improve the rate capability up to this requirement a careful optimization of the resistive protection system and a finer read-out (pixel/pad) are necessary. We present the development of resistive micromegas with O(mm2 ) pad readout aiming at precision tracking without efficiency loss up to several MHz/cm2 . The proposed detector has been originally inspired by a similar R&D project considered for the COMPASS experiment [4] and sampling calorimetry [5] and based on a solution proposed earlier by R. De Oliveira et al. [6].

∗ Corresponding author at: Università di Napoli Federico II, Naples, Italy. E-mail address: [email protected] (M. Della Pietra).

https://doi.org/10.1016/j.nima.2018.10.052 Received 30 June 2018; Received in revised form 7 October 2018; Accepted 8 October 2018 Available online 19 October 2018 0168-9002/© 2018 Elsevier B.V. All rights reserved.

M. Alviggi, V. Canale, M. Della Pietra et al.

Nuclear Inst. and Methods in Physics Research, A 936 (2019) 408–411

Fig. 1. Top: schematic view of the Embedded-Resistor prototype. Bottom left: side view of the schema for the DLC-Layer prototype. Bottom right: top view of the DLC-Layer prototype with the pattern of the conducting links matrix.

2. Prototypes construction details

Fig. 2. 55 Fe source spectra for different prototypes as recorded with a multi channel analyzer (MCA).

Several prototypes have been built with different spark protection resistive layers, deposited with different techniques: a pad-patterned embedded resistor layout with screen printing, and a uniform DLC (Diamond Like Carbon structure) layer by sputtering. In the first case (Embedded-Resistor) [7] the anode pads are overlaid by resistive pads, both interconnected by intermediate ‘‘embedded’’ resistors, as shown in top Fig. 1. The signal is transmitted to the readout pad by capacitive coupling, while the charges are evacuated through the intermediate resistors, having a resistance in the range 3–7 MΩ. The second technique (DLC-Layer) uses, instead, two continuous resistive DLC layers, with a mean sheet resistivity of about 50 MΩ/□, interconnected between them and to the readout pads with a network of conducting links with a few mm pitch, filled with silver past, to evacuate the charge, as referred in bottom part of Fig. 1. Actually, the detector was divided in two halves, each one having a different pitch of the conducting vias through the DLC layers: 6 mm and 12 mm respectively. This spark protection mechanism with a double DLC layer has been inspired by the technique used for the developing of 𝜇-RWELL detectors prototypes [8]. All detectors have in common the anode, that consists of a matrix of 48 × 16 copper pads, each pad having a rectangular shape with a pitch of 1 and 3 mm in the two coordinates. The active surface is (48 × 48) mm2 with a total number of 768 channels, routed off-detector for readout. The detector is completed with a bulk-Micromegas process [9], defining the 128 μm amplification gap with a metallic micro-mesh supported by Pyralux® insulating pillars, and with a drift cathode defining the 5 mm wide conversion gap.

Fig. 3. Detector gain as a function of the amplification voltage for different intensities of the iron source. Red dots indicate the curve measured for Embedded-Resistor prototype with ‘‘High Activity’’ source.

• high rate measurements obtained using 8 keV X-rays peak from a Cu target tuning the intensity of the X-rays gun by varying the excitation current. Fig. 2 shows the charge spectra measured for both Embedded-Resistor and DLC-Layer prototype using ‘‘High activity’’ 55 Fe source. The energy resolution of the DLC-Layer detector, measured as the ratio between the width and the peak value of the fitted gaussian, is about 14%; this value has to be compared with the one measured for the Embedded-Resistor prototype of about 36%. The different behavior of the two detectors is due to their difference in the spark protection resistive layer: the DLC-Layer detector has a more uniform electric field in proximity of the resistive layer, while the pad-patterned layout of the other detector produces strong variations of the electric field along the border of the resistive pads, thus getting the energy resolution worse. Detector gain as a function of amplification voltage has been measured for the two prototypes by means of the two different iron sources, the ‘‘Low activity’’ one and the ‘‘High activity’’ one as defined earlier in the text. Several measurements have been performed in different periods and using different methods. Results are shown in Fig. 3. All measurements are consistent among each other except for the gain curve for the Embedded-Resistor prototype when it is exposed to the more intense source, where a gain drop of about 20% is observed. This effect is due to the charge-up of the dielectric that takes place when the intense source is used (higher particle rate), resulting in a local reduction of the

3. Detector characterization and performance All the produced detectors have been extensively tested in the GDD (Gas Detector Development) laboratory of the RD51 collaboration [10] at CERN, using different radiation sources and measuring their performances in different conditions. All the measurements have been carried out using Ar/CO2 (93/7) gas mixture. Both detectors have been operated with the anode pads connected to ground: hence the mesh and the cathode have been powered with negative voltages. In the following, if not explicitly stated, all the voltages applied to the electrodes have to be considered as negative values. For both prototypes the gain has been measured with two methods: reading the detector current from readout pads with a pico-ammeter and counting signal rate from the mesh or analyzing signals from mesh with a MultiChannel Analyzer (MCA). Two different sets of measurements have been performed on both prototypes using two different radiation sources: • low rate measurements performed using two 55 Fe sources with different activities, referred in the text as ‘‘Low activity’’ (with a measured rate of 1.3 kHz) and ‘‘High activity’’ (total rate of 128 kHz); 409

M. Alviggi, V. Canale, M. Della Pietra et al.

Nuclear Inst. and Methods in Physics Research, A 936 (2019) 408–411

Fig. 4. Gain as a function of the rate for a fixed amplification value [11]. Fig. 5. Normalized gain for the DLC-Layer prototype as a function of the incoming particle rate for different shielding holes: 10 mm (red markers), 3 mm (blue markers) and 1 mm (green markers). The open symbols refer to the side of the detector with 6 mm pitch for grounding vias, the full ones refer to the other side with 12 mm pitch . (For interpretation of the references to color in this figure legend, the reader is referred to the web version of this article.)

amplification field. In Fig. 3 is also reported the same measurements performed on DLC-Layer prototype using both sources: in this case no evidence of the gain reduction is present, accordingly with the fact that no dielectric is present on the amplification gap surface. To study the detector response to increasing incoming particle flux, we used an Xray gun as reported earlier in the text. In order to fix the impinged area, photons have been focused by using a 1 or 3 mm diameter collimator, or by shielding the detector active area with an absorber hosting a 10 mm, 3 mm and 1 mm diameter hole. In this way we were able to explore a large rate range up to more than 150 MHz/cm2 . Photon rate per surface unit has been calculated supposing that the X-rays are uniformly hitting a surface corresponding to the hole/collimator diameter. Fig. 4 reports the gain of the Embedded-Resistor prototype measured for an amplification voltage of 530 V as a function of the photon rate: a gain reduction of about 20% is reached with X-rays at ≃10 MHz/cm2 , but gain is still about 4000 for rate well above 150 MHz/cm2 , corresponding to a gain reduction of about 40%. Fig. 5 shows the results of the same X-rays measurements performed on the DLC-Layer prototype. In order to focus on the gain drop, gain curve as a function of the rate has been normalized to the mean value of the gain measured at low rates (below 80 kHz/cm2 ). The amplification voltage is 530 V and the absolute value of the gain is about 12000. A gain drop of about 20% is reached with X-rays at few MHz/cm2 with a sheet resistivity of the DLC layer of 50 MΩ/□, but with 1 mm wide photon spot no gain drop is observed up to several MHz/cm2 . Moreover, gain drop slightly depends from the pitch of the conducting vias between the two DLC layers and the readout pads: as expected the farer is the connection to the ground, the higher is the resistance of connection and the voltage drop of the anode with respect to the mesh, thus producing a lower amplification electric field.

Fig. 6. Experimental setup of the 2017 test beam, when both prototypes have been tested simultaneously.

rotation corrections were applied. Fig. 7 shows the residuals distribution for the Embedded-Resistor prototype, on the left for the precision coordinate (1 mm pitch) and on the right for the secondary one (3 mm pitch), obtained during 2016 test beam campaign. From a gaussian fit to the residuals distribution we measure a space resolution of about 190 μm for the most accurate coordinate; from the box distribution of the second coordinate we got a resolution of about 720 μm. The same analysis has been performed for the DLC-Layer prototype and the results are shown in Fig. 8. As a preliminary result we measure for the latter detector a better resolution of about 110 μm in the precision coordinate, even if the residuals distribution has larger tails, that can be sized with a double gaussian fit. This effect has to be further investigated. The resolution in the second coordinate is compatible with the one measured with the Embedded-Resistor prototype. Three types of efficiency have been defined: (a) cluster efficiency: an event is classified as efficient if a good reconstructed cluster is found; (b) software efficiency: an event is efficient if a good reconstructed cluster is found within an acceptance window of 1.5 mm around the position of the extrapolated track; (c) 5𝜎 efficiency: an event is efficient if a good reconstructed cluster is found within five times the width of the residuals distribution around the extrapolated track [11]. Fig. 9 shows the Embedded-Resistor detector efficiency as a function of the amplification voltage. Plateau values of 99.8%, 99.4% and 99.0% were measured for the cluster, the software and the 5𝜎 efficiency, respectively. Preliminary results for the DLCLayer prototype show that the detector resolution is compatible with the Embedded-Resistor prototype.

4. Test beam results Detector performance, including efficiency and spatial resolution have been studied in two dedicated testbeam campaigns at SPS facility at CERN, where high energy muon and pion beams are available. Data have been acquired with APV-25 [12] hybrids and RD51 SRS system [13]. A scintillator hodoscope has been used for triggering incoming particles and two double view bulk resistive Micromegas detectors with strip segmented anode have been used as tracking chambers. Fig. 6 shows the experimental setup used in 2017, where both Embedded-Resistor and DLC-Layer prototypes were tested together. Instead, during 2016, only the Embedded-Resistor prototype was present. The clusters on the pad detector have been reconstructed combining the charge information from neighboring fired pads. The spatial resolution has been evaluated comparing the cluster position reconstructed on the detector and the expected hit position obtained from tracking detectors. Alignment and 410

M. Alviggi, V. Canale, M. Della Pietra et al.

Nuclear Inst. and Methods in Physics Research, A 936 (2019) 408–411

Fig. 7. Residuals distribution between expected and reconstructed cluster position for precision (left) and second (right) coordinate of the Embedded-Resistor prototype. The amplification voltage is 530 V.

Fig. 8. Residuals distribution between expected and reconstructed cluster position for precision (left) and second (right) coordinate of the DLC-Layer prototype. The amplification voltage is 510 V.

by optimizing the sheet resistivity of the DLC layer and the pitch of conducting links to ground. Acknowledgments We would like to thank the CERN MPT Workshop (R. De Oliveira and A. Teixeira for ideas, discussions and construction of the detectors), the RD51 Collaboration for support with tests at the GDD Lab and at H4 SPS beam line. This research hase been founded by I.N.F.N within MPGD_NEXT R&D program. References [1] Y. Giomataris, P. Rebourgeard, J.P. Robert, G. Charpak, MicroMegas: A high granularity position sensitive gaseous detector for high particle flux environments, Nucl. Instrum. Methods A 376 (1996) 29. [2] T. Alexopoulos, et al., A spark-resistant bulk-MicroMegas chamber for high-rate applications, Nucl. Instrum. Methods A 640 (2011) 110. [3] ATLAS collaboration, New Small Wheel Technical Design Report, CERN-LHCC2013-006, 2013. [4] F. Thibaud, et al., Performance of large pixelised MicroMEGAs detectors in the COMPASS environment, JINST 9 (2014) C02005. [5] M. Chefdeville, et al., Resistive Micromegas for sampling calorimetry, a study of charge-up effects, Nucl. Instrum. Methods A 824 (2016) 510–511. [6] R. de Oliveira, Resistive protection in MicroMegas, talk given at the RD51 Mini Week Workshop, CERN, Geneva Switzerland, 22–24 February, 2010. [7] M. Alviggi, et al., Small-pads resistive micromegas, JINST 12 (2017) C03077. [8] G. Bencivenni, et al., The 𝜇-RWELL detector, JINST 2 (2017) C06027. [9] I. Giomataris, et al., MicroMegass in a bulk, Nucl. Instrum. Methods A 560 (2006) 405, [physics/0501003]. [10] RD51 collaboration, R&D Proposal Development of Micro-Pattern Gas Detector Technologies, CERN-LHCC-2008-011, 2008. [11] P. Iengo, et al., Small-pad resistive Micromegas for Operation at Very High Rates PoS EPS-HEP2017, 2017, p. 793. [12] M.J. French, et al., Design and results from the APV25, a deep sub-micron CMOS front-end chip for the CMS tracker, Nucl. Instrum. Methods A 466 (2001) 359. [13] S. Martoiu, H. Muller, J. Toledo, Front-end electronics for the Scalable Readout System of RD51, in: The Proceedings of Nuclear Science Symposium and Medical Imaging Conference (NSS/MIC), 23–29 October, 2011.

Fig. 9. Efficiency vs amplification voltage for the Embedded-Resistor prototype [11].

5. Conclusion Two different versions of Micromegas with small resistive pads readout have been designed, built and tested. They differ from each other by the layout of the resistive spark protection layer: one with a pad patterned embedded resistors and the other with a uniform DLC layer. Both configuration show a very good efficiency and a good spatial resolution, less than 200 μm, in the more accurate coordinate. The Embedded-Resistor prototype has shown a moderate energy resolution and a gain drop of about 20% at a rate of about 10 MHz/cm2 ; a gain reduction related to the charge-up of the dielectric that is present on the amplification gap surface has been evidenced. A better energy resolution (Fig. 2) and a similar rate capability (Fig. 5) has been measured for the DLC-Layer prototype and no charge-up effect was visible during the tests. An higher rate capability for second prototype can be reached

411