Study of the performance of the MicroMegas chambers for theATLAS Muon Spectrometer Upgrade

Marco Vanadia, on behalf of the ATLAS Muon Collaboration

Sapienza University of Rome & INFN

ANIMMA 2015, 23/04/2015

The ATLAS experiment

Biggest detector at acollider, working @ LargeHadron Colliderpp collisions @ 7-8 TeV(13 soon!) + heavy ionsKey detector systems forthis talk:

trigger system: 40MHz → . 1 kHz !muon spectrometerwith MDT, RPC, CSCand TGC chambersworking in a toroidalmagnetic field

Multi-purpose detector, able to measure precisely what we already know and to searchfor the unexpected!

2

The near and far future for ATLAS and the LHC

2010-2012: Run-1, 25 fb−1 @ 7-8 TeV, Higgsdiscovery (ATLAS+CMS), > 400 published papersNOW: energy → 13 TeV, new energy frontier, newphysics discovery potential2018: Phase-I upgrade, 2× design luminosity and along run scheduled (>300 fb−1)

main difficulties: dealing with trigger rate and higherenviromental background!→ if trigger thresholds become too high, we loseefficiency on physics (see plot on the right)

Long term: Phase-II upgrade and a long physicsprogram at ≈5× design luminosity!

Effects of trigger cuts onselection of WH events

3

The New Small WheelThe hit rate in the detector increases linearly with the luminosity. This is particularlyproblematic in the forward spectrometer

At the maximum Run-4 luminosity of5-7·1034 cm−2s−1 in the innermostpart of the forward spectrometer thehit rate can be as high as 15 kHZ/cm2

To achieve the demandingperformance goal for ATLAS, acomplete replacement of the SmallWheel is neededThe New Small Wheel will employMicroMegas (MM) and small-stripThin Gap Chambers (sTGC)ATLAS-TDR-020

MDT chambers performance vs hit rate

4

Triggering with the New Small Wheel

Currently trigger chambers are inMiddle station. In the figure on theright, tracks A, B and C would beaccepted by the trigger systemWith the New Small Whell only trackspointing to the interaction point willbe accepted (track A in the figure)

The NSW willemploy 8 sTGC and8 resistive MM layers

The trigger rate is significantly reducedwith the New Small Wheel

5

MicroMegas prototypes for the New Small WheelMicroMegas chambers are MicroPattern Gaseous Detectors introduced in the 90s1

Conversion/Drift Gap

Amplification Gap

Insulator

PCB Board

5 mm

128 μm

Readout StripsResistive Strips

e- e-

Drift Electrode

Micromesh

400 μm

-300 V

+500 V

E Field

E FieldPilla

r

Pilla

r

Pilla

r

Drift region: ≈5 mmwith low electric fieldGrounded mesh (325lines/inch)Amplificationregion: thin gap(128 µm) with highelectric field, fastions evacuation →high rate tolerance

Inclined tracks fire many strips, by measuring time of arrival of signal trackreconstruction inside MM is possible: µTPC reconstruction2

One of the most relevant breakthroughs achieved within the Muon AtlasMicroMegas activity: resistive strips for spark-protection (see NIM A640 (2011)110-118) capacitatively coupled with readout strips.NSW: 4 MM-layers chambers in Ar/CO2 93/7

The challenge: MM chambers with an unprecented size (up to 3 m2) and a highconstruction precision required to satisfy the demanding performance

1NIM A376 (1996) 29-352NIM A617 (2010) 161-165

6

Measurement of the performance of the MicroMegas chambers

Extremely wide performance analysis program ongoing within the MAMMA (MuonAtlas MicroMegas Activity) collaboration to

test the prototype chambers and the r/o electronics, measure the basic properties(gain, transparency ...), find theoptimal working point (HVs...)fully characterize the MicroMegas performance (resolution, efficiency ...) fortracks with varying inclination with/without a magnetic fielddevelop and optimize the track reconstruction strategies and software

with the final goal of obtaining a single-layer spatial resolution ≈100µm and areconstruction efficiency >95% for tracks up to 30° within a B field up to 0.3 T

Test beamSeveral test beam campaigns performed to studyMM prototypes performance10x10 cm2 bulk prototypes and 100x50 cm2

4-layers chamber have been studiedPerformed measurements with perpendicular andinclined tracks, with/without magnetic field, fordifferent working points (HV, gas mixture, ...)

Other testing activityCosmic rays test-standsX-ray guns for gain andmesh transparencymeasurementsIrradiation tests forageing and radiationhardness studies

7

Single-strip response

Strips signal is read out by APV25 hybrid cards 1 through the Scalable ReadoutSystem 2

Output is charge vs time distribution sampled every 25 nsMaximum of the distribution is the charge read by the stripsWith a fit on the rising time, time of arrival of the signal can be measuredAs a baseline, a fit wtih a Fermi-Dirac function is performed, the flex of the curvetFD is used as time of arrival of the signal

1Nuclear Instruments and Methods in Physics Research A 466 (2001) 359–3652JINST 6 C11028

8

Position reconstruction with a MicroMegas layer

Centroid methodselect a cluster ofneighbouring stripsreconstruct positionas average of strippositions weightedby their chargewell performing forperpendicular tracks

µTPC method1

time measured foreach stripknowing vdrift , trackreconstruction canbe performedwell performing forinclined tracks

1NIM A617 (2010) 161-1659

Spatial resolutionCentroid resolution for perpendicular tracks µTPC resolution for 30° inclined tracks

Spatial resolution has been measured forcentroid and µTPC reconstruction as thewidth of the difference between 2 chambers for10×10 cm2 ”T” (0.4 mm strip pitch) and”Tmm” (0.25 mm strip pitch) protoypesDouble gaussian fit performed to take intoaccount long tails (noise, events with multipletracks, residual cross-talk ...)In the NSW tracks will be inclined by 8°-32°>100 µm resolution achieved combiningcentroid and µTPC in this range (and beyond)

10

Reconstruction optimization: correction for capacitative coupling btw strips

Signal in first and last strip ina cluster is often just due tocapacitative coupling→ they introduce a bias inthe µTPC trackreconstructionA correction tehcnique hasbeen developed, by filteringlow-charged strips at thecluster endsFine tuning of the chargeposition assignment along thestrip width

11

Efficiency measurements

Left: centroid hit position in a MMprototype with 2D readout. The effectof the pillars is clearly visibleBottom: measurement ofreconstruction efficiency doneextrapolating track from telescope ofreference chambers, for perpendicular(left) and 30° inclined tracks (right)The impact of pillars is higher forperpendicular tracks, as expectedA very high efficiency is achieved

12

First 4-layers prototype: the MMSW chamber

MMSW chamber: 1.2 ×0.5 m2 × 80 mm built @CERN4 detection layers, 1024strips each (0.415 mmpitch, drift gap 5 mm)

2 back-to-back layersfor precision coordinate2 planes with -1.5° and+1.5° inclined strips(stereo)

The layers with stereo strips will be used in ATLAS to measure the second coordinate,with a resolution of 2-2.5 mm required by the trigger system

MMSW chamber will be installed in the ATLAS detector during Run-2!

13

Performance of the first 4-layers prototype

The prototype has been tested at CERN test beam facilities during the last year

Measured resolution for first and second coordinate are well within specificationsA ≈98% per-layer reconstruction efficiency has been measured for perpendiculartracks

14

MM in magnetic field: spatial resolution

For a given angle of inclination, thereis a “critical” value of the B fieldcausing a “focusing” effect on drift e−

The µTPC resolution degradation canbe compensated by combining withthe centroid reconstructionStudies are ongoing on recenttest-beam data to optimizereconstruction within B field

Note: resolutions are larger here due to lower amplification of the working point of these testbeam data

15

MM in magnetic field: measurement of vdrift

Test of prototypes in magnetic field isone of the main current topics ofactivity within our collaborationHere showing a measurementperformed on test beam data of thedrift velocity vdrift as a function of theB fieldFrom the width of the distribution oftime measurements from a chamber,since the gap size is known, vdrift canbe measuredIn magnetic field this must becorrected for the effect of the Lorentzangle, since vdrift is not parallel to theE fieldResults in good agreement withsimulation (Garfield)

16

Irradiation tests

Small prototypes (bulk MM) thoroughly tested with X-rays, neutrons, γ and αEquivalent of � 5 years of high luminosity LHC simulatedNo evidence of ageing effects have been measured

X-ray irradiation test

Exposure: 918 mC for 4 cm2

in 21.3 effective days. 225mC/cm2 (32 mC/cm2

estimated for 5 years ofHL-LHC)Mesh current stable and inline with referencenot-irradiated detector

Neutron irradiation

Flux: 3·1012 n/(hour · cm2),5-10 MeV, corresponding to 2years of HL-LHCMesh current stable duringthe whole testing period

17

Summary and outlook

After many years of R&D and breakthroughs in the technology in the context ofthe ATLAS upgrade, the MicroMegas are now mature detectors to be installedover large areas at HEP experimentsVery wide program of studies, measurements and tests ongoing within the MuonATLAS MicroMegas ActivityTests have been performed on small 10×10 cm2 prototypes, proving that thedemanding performance requirements for the ATLAS New Small Wheel are metDuring last year started to test chambers close to the final design, i.e. withfloating mesh and composed of 4 resistive MicroMegas planesThese tests have been very succesful as well, these chambers have highefficiency and very good resolutionMain focus is now on fully characterizing and optimizing the performance of theseMicroMegas chambers within a magnetic field4-layers prototype MMSW will be included in the ATLAS detector for Run-2!

18

BACKUP

20