Muon simulation : status & plan Partha Pratim Bhaduri Subhasis Chattopadhyay VECC, Kolkata

2 CBM Physics – keywords physics program complementary to RHIC, LHC rare probes What does theory expect? → mainly predictions from lattice QCD: crossover transition from partonic to hadronic matter at small  B and high T critical endpoint in intermediate range of the phase diagram first order deconfinement phase transition at high  B but moderate T However... deconfinement = chiral phase transition ? hadrons and quarks at high  ? signatures (measurable!) for these structures/ phases? how to characterize the medium?

3 Physics of CBM : Observables physics topics Deconfinement at high  B ? Equation of State at high  B ? order of phase transition ? Critical point ? in-medium properties of hadrons  onset of chiral symmetry restoration at high  B observables strangeness production: K,  charm production: J/ , D flow excitation function event-by-event fluctuations   l + l - open charm CBM: rare probes → high interaction rates! CBM: detailed measurement over precise energy bins (pp, pA, AA) FAIR beam energy range AGeV (protons 90 GeV)

4 Charm production at threshold [W. Cassing et al., Nucl. Phys. A 691 (2001) 753] HSD simulations CBM will measure charm production at threshold → after primordial production, the survival and momentum of the charm quarks depends on the interactions with the dense and hot medium! → direct probe of the medium! charmonium in hot and dense matter? relation to deconfinement? relation to open charm?

5 screening of pairs in partonic phase anomalous J/  suppression observed at top-SPS and RHIC energies Sequential suppression - signal of deconfinement? OR Co-mover absorption? Deconfinement : charmonium suppression no J/ψ, ψ' → e+e- (μ+μ-) data below 158 AGeV Measure excitation functions of J/ψ and ψ' in p+p, p+A and A+A collisions Still an open issue

6 [Rapp, Wambach, Adv. Nucl. Phys. 25 (2000) 1, hep-ph/ ]  -meson couples strongly to the medium vacuum lifetime  0 = 1.3 fm/c dileptons = penetrating probe  -meson spectral function particular sensitive to baryon density connection to chiral symmetry restoration? p n  ++  p   e +, μ + e -, μ -  In medium effects:  -meson

7 In-medium modifications :  mesons (II) no ρ,ω,φ → e + e - (μ + μ - ) data between 2 and 40 AGeV Data: In+In 158 AGeV, NA60 Calculations: H.v. Hees, R. Rapp Low mass excess well established by CERES (dielectrons). Clear discrimination between different theoritical explanations is still missing. Latest NA60 data shows a clear evidence for broadening of width- no mass shift Data: CERES Calculations: R. Rapp  broadening  mass shift

8 Detector requirements Systematic investigations: A+A collisions from 10 to 45 (35) AGeV, Z/A=0.5 (0.4) (up to 8 AGeV: HADES) p+A and p+p collisions from 8 to 90 GeV observables detector requirements & challenges strangeness production: K,  charm production: J/ , D flow excitation function event-by-event fluctuations   e + e - open charm tracking in high track density environment (~ 1000) hadron ID lepton ID myons, photons secondary vertex reconstruction (resolution  50  m) large statistics: large integrated luminosity: high beam intensity (10 9 ions/sec.) and duty cycle beam available for several months per year high interaction rates (10 MHz) fast, radiation hard detector efficient trigger rare signals!

9 Dipole magnet The Compressed Baryonic Matter Experiment Ring Imaging Cherenkov Detector Transition Radiation Detectors Resistive Plate Chambers (TOF) ECAL Silicon Tracking Station Tracking Detector Muon detection System

10 Di-muon measurement : De-confinement transition (charmonia) Medium modification ( LMVM) Major Indian participation Building of Muon chambers : Detector simulation for feasibility measurement R & D with Chambers

11 Standard Muon Chambers low-mass vector meson measurements (compact setup) 125 cm. Fe≡ 7.5 λ I 225 Cm. Fe≡ 13.5 λ I Fe cm W-shielding  J/  

12 Challenges in muon measurement Dimuons from vector meson decays are notoriously difficult to measure : Low multiplicity at FAIR energies Very small branching ratios in di-muon channel (Yield per event = multiplicity ×branching ratio) Large combinatorial background in heavy-ion collisions due to  Weak decays ,  decays into   Hadron punch through  Secondary electrons (  electrons ) compact layout to minimize K,p decays → use excellent tracking to reject p,K decays in the STS by kink detection → absorber-detector sandwich for continous tracking → use TOF information to reject punch through K,p → Increase Air gap between detector-absorber to reduce delta electrons → Incerase number of stations after each absorber

13 Simulation Framework : CbmRoot Input : Pluto event generator for signal UrQMD event generator for background HSD for multiplicity GEANT3 for transportation of the particles through detector materials Cellular Automata (CA) for track finding Kalman Filter (KF) for track fitting Super Event Analysis (SE) technique for estimation of signal to background ratio

Much working group at different countries GSI, Germany India Russia 14

15 Muon GSI Time measurements for the muon identification LMVM trigger J/ψ p T reconstruction Muon simulations with reduced detector acceptance

16 Background rejection via mass determination (L, t) → β simple design of MuCh Muon ToF TOF gives velocity Measure mass of incoming particle

17 Full reconstruction ω→μ + μ - + central Au+Au collisions at 25 AGeV time information withoutwith time resolution 30 psec50 psec80 psec S/B ratio ε, %

18 Invariant mass spectra time information: — without with time resolution: — 80 psec — 50 psec — 30 psec ω→μ + μ - + central Au+Au collisions at 25 AGeV

19 Trigger strategy z target x target,y target ∆x,∆y 1.Find events with min. 12 hits in 6 detector layers, which might correspond to two tracks (hit selection in muon ToF: velocity value) 2.Straight line fit 3.Track selection: fit criteria Remark: if track passes cuts, its hits will not used for second track searching Muon ToF

20 Trigger 1000 central events (Au+Au collisions at 25 AGeV) min. 12 hits in 6 detector layers + β  [0.96; 1.02] (β cut) β cut + χ 2, X Z=0, Y Z=0 cuts β cut + χ 2, X Z=0, Y Z=0 cuts + Z X=Y=0 cut ω→μ + μ - + central Au+Au collisions at 25 AGeV ε all trigger cuts /ε without trigger cuts 40% background suppression factor ~35

21 P t  [0.0, 0.2] GeV/cP t  [0.2, 0.4] GeV/cP t  [0.4, 0.6] GeV/cP t  [0.6, 0.8] GeV/c P t  [0.8, 1.0] GeV/cP t  [1.0, 1.2] GeV/cP t  [1.2, 1.4] GeV/cP t  [1.4, 1.6] GeV/c P t  [1.6, 1.8] GeV/cP t  [1.8, 2.0] GeV/cP t  [2.0, 2.2] GeV/cP t  [2.2, 2.4] GeV/c Invariant mass spectra for different P t J/ψ

22 Spectra of extracted J/ψ for different P t J/ψ P t  [0.0, 0.2] GeV/cP t  [0.2, 0.4] GeV/cP t  [0.4, 0.6] GeV/cP t  [0.6, 0.8] GeV/c P t  [0.8, 1.0] GeV/cP t  [1.0, 1.2] GeV/cP t  [1.2, 1.4] GeV/cP t  [1.4, 1.6] GeV/c P t  [1.6, 1.8] GeV/cP t  [1.8, 2.0] GeV/cP t  [2.0, 2.2] GeV/cP t  [2.2, 2.4] GeV/c

23 Reconstruction results STS acceptancefullreduced S/B ratio ε J/ψ (%) Cuts STS: –  2 prim.vertex –N of STS hits MuCh: –N of MuCh hits TRD: –N of TRD hits TOF: –hit in ToF –  cut STS acceptance:  full  reduced J/ψ  μ + μ - + Au+Au collisions at 25 AGeV

24 Muon India Optimization of muon detection system Detector in-efficiency study Development of charmonium trigger J/Psi p T reconstruction

We have to decide upon : – Total number of stations (layers) – Total absorber thickness, total no. of absorbers & the absorber material – Number of layers (2/3) in between two absorbers – Distance between stations & absorber to station distance Present constraints : – Absorber material (Fe, Pb, W ) – Layer to layer distance >= 10 cm. – Layer to absorer distance >= 5cm. Much Geometry optimization

Comparative study between two extreme cases: SIS100 geometry: 9 detector layers; (proposed by collaboration meeting) SIS 300 geometry: 18 detector layers; (existing in SVN) Total absorber thickness in both the cases is same (225 cm. of Fe)

Optimization should be done with low mass vector mesons (lmvms) rather than J/ψ and at the lowest available energy. J/ψ measurements due to low background after more than 2 m of Fe are not so sensitive to the muon setup as the measurements of muons from LMVM. Issue is to reconstruct the soft muons ( eg: ω→μμ ) Use the same set-up for in simulation for J/ψ & LMVM. For LMVM use information from stations just before the last thick absorber. Run full simulation & obtain signal reconstruction efficiency & S/B ratio. Simulate both lowest (minimum boost) & highest energy (maximum multiplicity). Few facts to remember …

Much Geometries: specifications Standard Geometry # of stations : 6 # of layers : 3*6 =18 Total absorber thickness : 225 Cm ( ) Distance between layers : 10 cm. Detector to absorber distance : 10 cm. Reduced Geometry: # of stations: 3 # of layers : 3*3 = 9 Total absorber thickness: 225 cm. ( ) Distance between layers : 10 cm. Detector to absorber distance: 10 cm.

Simulation : Transport : Central 10 AGeV, 25 AGeV & 35 AGeV Signal : Pluto (ω→μμ) Background : UrQMD Reconstruction : Segmentation scheme : Manual segmentation Segmentation 1: minimum pad size: 4mm. ; maximum pad size : 3.2 cm. Segmentation 2: minimum pad size: 5mm. ; maximum pad size : 5 cm. Simple Much hit producer w/o cluster & avalanche Ideal (STS) & Lit (Much) tracking

Implementation of detector in-efficiency 5% in-efficiency w/o in-efficiency ~ 5% change in average number of hits

Implementation of detector in-efficiency No loss 5 % hit loss

No hit loss 5% hit loss Effect of hit loss on reconstructed tracks Global tracks Much tracks

Invariant mass spectrum (ω→μμ ) Cuts : 1. No. of Muchhits>=4 2. No. of STS Hits >=4 3. chi2primary < 3 Super event (SE) analysis for bkg (combine all the positive tracks with all the negative tracks over all the events excluding only tracks from same event). Gaussian fit to signal Polynomial fit to bkg. 10k central embedded events for Au + 10GeV/n Reduced Geometry

Results for various pad sizes (ω→μμ ) Pad size ( station #1 ) Total PadsReconstruction efficiency (%) S/B 2 mm mm mm mm mm k central embedded events for Au + 10GeV/n

Invariant mass spectra (ω→μμ) Cuts : 1. No. of Muchhits>=15 2. No. of STS Hits >=4 3. chi2primary < 3 Super event (SE) analysis for bkg (combine all the positive tracks with all the negative tracks over all the events excluding only tracks from same event). Gaussian fit to signal Polynomial fit to bkg. Standard Geometry Central embedded events for Au + 25GeV/n

Invariant mass spectrum Standard Geometry Central embedded events for Au + 25GeV/n Super event (SE) analysis for bkg (combine only urqmd the positive tracks with urqmd negative tracks over all the events excluding only tracks from same event). Gaussian fit to signal Polynomial fit to bkg. Cuts : 1. No. of Muchhits>=15 2. No. of STS Hits >=4 3. chi2primary < 3

Results of full reconstruction Energy (GeV/n) Segmentatio n Total padsReconstructi on efficiency (%) S/B S/B (UrQMD) Segmentation 1: Minm. Pad size: 4 mm. Maxm. Pad size: 3.2 cm. Segmentation 2: Minm. Pad size: 5 mm. Maxm. Pad size : 5 c,m. Standard geometry

38 Development of charmonium trigger Charmonia (J/ ,  ’ are rare probes i.e. they have very low multiplicity(~10 -5 or ). For example for central Au+Au AGeV beam energy multiplicity of J/  is 1.5*10 -5 and that of  ’ is 5* They have very low branching ratio (~5-6%) to decay into dimuon channel. Their detection requires an extreme interaction rate. For example to detect one J/  through its decay into di-muons it requires around 10 7 collisions. Online event selection based on charmonium trigger signature is thus mandatory, in order to reduce the data volume to the recordable amount.

Wednesday, April 15, 2009 CbmRoot Version: Trunk version Much geometry : Standard Geometry 2 layers in 5 stations Distance between layers 10 cm. Gap between absorbers 20 cm 3 layers at the last trigger station Total 13 layers Total length of Much 350 cm Signal : J/  decayed muons from Pluto Background : minimum bias UrQMD events for Au+ Au at 25 GeV/n Much Hit producer w/o cluster & avalanche L1(STS) & Lit (Much) tracking with branching Input : reconstructed Much hits Simulation Absorber thickness (cm):

Wednesday, April 15, 2009 Trigger algorithm Take 3 hits from the trigger station with one from each of the 3 layers & fit with st. line both in X-Z & Y-Z plane passing through the origin (0. 0) i.e. X = m 0 *Z ; Y=m 1 *Z Make all possible combinations Find  2 & apply cut on both  2 X &  2 y Hit combination satisfying the cuts is called a triplet. Hits once used for formation of a triplet is not used further. Find m 0 & m 1 of the fitted st. lines Define a parameter α=√(m 0 2 +m 1 2 ) Apply cut on α Magnetic field (0,0,0) (0,0.0) Trigger station

Wednesday, April 15, 2009 Specification of cuts Cut 1: at least 1 triplet/event Cut 2 : at least 2 triplets/event Cut 3 : at least one of the selected triplets satisfy alpha cut Cut 4 : at least two of the selected triplets satisfy alpha cut Events analyzed: 80k minimum bias UrQMD event for background suppression factor & 1k embedded minimum bias events for J/  reconstruction efficiency

Wednesday, April 15, 2009 Background suppression factor (B. S. F) Cut Events survived Statistical Error B. S. F % ~ % ~ % ~ % ~1430 B. S. F = Input events (80,000) / events survived

Wednesday, April 15, 2009 Reconstructed J/  Trigger cut Reconstruction efficiency (%) no cut 29.3 % Cut % Cut % Cut % Cut % 1k embedded minimum bias events

Motivation: Physics performance analysis for SIS-100. Developed a “close-to-standard” version of Much for SIS-100. p T & Y dependent J/  reconstruction efficiency First step towards physics case study. J/Psi p T reconstruction

Methodology  In cbmroot framework J/Psi’s are generated and decayed into di-leptons employing the event generator PLUTO.  Pluto generates J/Psi’s following gausian rapidity & thermal p T distribution.  Generated J/Psi’s are decayed into di-leptons isotropically in the rest frame of mother (J/Psi) & the decayed leptons are lorentz boosted in lab frame.  J/Psi yield is low at high p T (exponential p T spectra); not suitable for studying p T dependent efficiencies.  Either huge increase in statistics (exponential distribution) or use flat distribution with moderate statistics.  Modify the Box generator to generate J/Psi’s with specified rapidity (2.0<Y<4.0) & p T (up to 4 GeV with steps of 100 MeV).distribution.  Generated J/Psi’s are decayed following isotropic angular distribution into two muons.

Simulation : Transport : Central 8A GeV Signal : Box generator J/  with given kinematic range : rapidiy (y) =2-4; pT : up to 4 GeV with steps of 100 MeV 1k embedded events for each step Background : UrQMD 8 GeV/n Reconstruction : Segmentation scheme : Manual segmentation Station 1 (layers 1, 2, 3) : 2 regions (pad size in the central region : 0.4 cm.) Station 2 (layers 4, 5, 6) : one region with pad size 3.2 cm * 3.2 cm. Station 3 (layers 7, 8, 9) : one region with pad size 5 cm.*5 cm. Implementation of detector in-efficiency at hit producer level. Simple Hit producer w/o clustering

p T dependent reconstruction efficiency Small bin size (100 MeV) ; Low statistics (1k in each bin) Large statistical fluctuation Cuts : No. of Muchhit>=7 No. of STS Hits >=4 Track MCId <2 Track pdgcode 13

p T dependent reconstruction efficiency Rebin the previous plot to reduce statistical fluctuation

Discussion p T dependent reconstruction efficiency does not show any monotonic variation. Higher be the p T of J/Psi, easier should be the reconstruction. Reconstruction efficiency should monotonically increase with p T. Results do not show such increasing trend; instead a large fluctuation (even though 1k input J/Psi’s per p T bin). Re-binned results decrease the fluctuation but does not show the increasing nature with p T. Generate J/Psi’s in the entire p T range & look at the reconstructed J/Psi p T.

Distribution for J/Psi  Pair pT distribution does not show any trend  Pair Y distribution show a dip in the middle  Look at the distribution of single muons

Distribution for single muons Input muons Reco. muons

Discussion Significant loss in the single muon level Input muons are distributed over a large rapidity interval. Some input muons are even at negative rapidity in the lab frame (backward scattering??) Input muons are even lost at mid-rapidity & high pT. Recheck the decay kinematics employed in the box generator.

Summary For SIS-100 we have a ‘close to standard’ geometry for muon detection system. Full simulation with different segmentation (varying pad size) shows we can use 4mm. /5mm. Pads in the first station. Issues with occupancy & rate needs to be fixed. Use the ‘reduced’ geometry for J/Psi simulation for 30 GeV p +Au collisions. Detector in-efficiency has been implemented in the hit producer level. J/Psi p T reconstruction needs to be completed

Future plans To complete the comparative study with more statistics & with other particles. Repeat the same simulations with an intermediate geometry with number of layers 12/15. Gap study & absorber study (change the air gap between layers; change the absorber material /thickness). Physics performance simulation : J/Psi p T & rapidity distribution. J/Psi flow study. Physics simulation of different observables (following Peter Senger’s list)

Thank you

56 Quarkonium dissociation temperatures: (Digal, Karsch, Satz) Measure excitation functions of J/ψ and ψ' in p+p, p+A and A+A collisions ! rescaled to 158 GeV Probing the quark-pluon plasma with charmonium J/ψψ'ψ' sequential dissociation?

57 Dipole magnet The Compressed Baryonic Matter Experiment Ring Imaging Cherenkov Detector Transition Radiation Detectors Resistive Plate Chambers (TOF) ECAL Silicon Tracking Station Tracking Detector Muon detection System

Signal Tracks Background tracks Distribution of chi2vertex

59 Summary: CBM physics topics and observables Onset of chiral symmetry restoration at high  B  in-medium modifications of hadrons ( , ,   e + e - (μ + μ - ), D) Deconfinement phase transition at high  B  excitation function and flow of strangeness (K, , , ,  )  excitation function and flow of charm (J/ψ, ψ', D 0, D ,  c ) (e.g. melting of J/ψ and ψ')  exitation function of low-mass lepton pairs The equation-of-state at high  B  collective flow of hadrons  particle production at threshold energies (open charm?) QCD critical endpoint  excitation function of event-by-event fluctuations (K/π,...) CBM Physics Book (available online)

60 Observables: Penetrating probes: , , , J/  (vector mesons) Strangeness: K, , , , , Open charm: D o, D  Hadrons ( p, π) Experimental program of CBM: Systematic investigations: A+A collisions from 10 to 45 (35) AGeV, Z/A=0.5 (0.4) p+A collisions from 10 to 90 GeV p+p collisions from 10 to 90 GeV Beam energies up to 2 to 8 AGeV: HADES Large integrated luminosity: High beam intensity and duty cycle, Available for several month per year Detector requirements Large geometrical acceptance good particle identification excellent vertex resolution high rate capability of detectors, FEE and DAQ

61 CBM setup with muon detector Muon system TRD ToF STS track, vertex and momentum reconstruction Muon system muon identification TRD global tracking RPC-ToF time-of-flight measurement STS

62 Tracks of one central collision (GEANT3) Central Au+Au collision at 25 AGeV: 160 p, 400  -, 400  +, 44 K+, 13 K-,....