Detectors for Particle Physics Lecture 2: Showers and calorimeters Particle flow D. Pitzl, DESY DESY summer students lecture 3.8.2011 Particles Signatures and detectors non-destructive destructive tail measurement measurements catcher D. Pitzl, DESY Detectors 2.2 DESY summer students lecture 3.8.2011 Photon interactions in matter (A, Z) A A*+ e- photo + + photo A* A effect fluor A A ' Rayleigh e e ' Compton Z Z e+ e- Pair coherent + - scattering e e e e Pair pair prod. on Z e- 42 barn e giant dipole res. + pair prod. on e e ee Pair production Z Ze+e- dominates above ~4 MeV D. Pitzl, DESY Detectors 2.3 DESY summer students lecture 3.8.2011 Radiation length High energy photon cross section in lead: = 42 barn/atom Convert to photon absorption coefficient: 2 3 [1/cm] = [cm /atom] · NA [atoms/mol] / A [g/mol] · [g/cm ] Define interaction length: = 1/ Define radiation length: X0 = 7/9 · Plug in some numbers: 1 barn = 10-24 cm2 23 NA = 6 · 10 atoms/mol A = 207 for lead 3 = 11.35 g/cm for lead ⇒ radiation length of lead: X0 = 0.56 cm. D. Pitzl, DESY Detectors 2.4 DESY summer students lecture 3.8.2011 Radiation Loss for electrons in matter e ● Bremsstrahlung: e Z Z e electromagnetic radiation produced by the deceleration of an electron, when deflected by an Z atomic nucleus. dE −E −x / X0 = ⇒ E (x) = E0 e dx X 0 716 A 1 X [cm]= 0 ρ [g/cm3] Z (Z +1) ln (287/√ Z) Pb: X0 = 0.56 cm Si: X0 = 8.9 cm D. Pitzl, DESY Detectors 2.5 DESY summer students lecture 3.8.2011 Pair production and Bremsstrahlung Pair production Bremsstrahlung e+ e- e- e- Z Z Very similar Feynman diagrams Just two arms swapped. At high energy: = 7/9 e D. Pitzl, DESY Detectors 2.6 DESY summer students lecture 3.8.2011 Muons radiate only at extreme energies A charged particle of mass m radiates while being influenced the Coulomb field of a nucleus Z: d Z 2 ln E ∝ dE m2 E F 2 classically, radiation power P ∝ a2= ( m ) m2 e = ⋅ −5 2 2.3 10 mμ critical energy for muons is 450 GeV in copper, 200 GeV in lead. D. Pitzl, DESY Detectors 2.7 DESY summer students lecture 3.8.2011 Electromagnetic Showers e+ x 0 X0 2X0 3X0 4X0 N 1 2 4 8 16 0 <E> E0 E0/2 E0/4 E0/8 E0/16 <Ec D. Pitzl, DESY Detectors 2.8 DESY summer students lecture 3.8.2011 Shower in a cloud chamber ● Cloud chamber image of a shower between lead plates. D. Pitzl, DESY DESY summer students lecture 3.8.2011 A simple shower model E 0 Start with a high energy electron: E0 - ⇒ After 1X0 : 1 e and 1, each with E0/2 k - + [ X ] ⇒ After 2X0 : 2 e , 1 e and 1 , each with E0/4 0 . k k ⇒ After kX0 : total N = 2 , each with <E> = E0/2 At <E> = Ec pair production and bremsstrahlung stop. Compton- or photoeffect and ionization take over. The shower ranges out. Ec = 0.6 GeV / (Z+1.24) = 7 MeV for lead. (empirical fit by the PDG) ⇒ kmax = lg2(E0/Ec). Shower depth grows logarithmically with E0. kmax ⇒ Nmax = 2 = E0/Ec. Number of particles grows linearly with E0. D. Pitzl, DESY Detectors 2.10 DESY summer students lecture 3.8.2011 Electron energy loss and critical energy relative energy loss for electrons: Critical energy: energy loss due to Bremstrahlung and (Bethe-Bloch does not ionization are equal: apply for electrons) 610 MeV E ≈ c Z +1.24 7 MeV in Pb. D. Pitzl, DESY DESY summer students lecture 3.8.2011 A sophisticated shower simulation Energy particle profile flow dE (bt)a−1 e−bt x longitudinal shower profile: = E0 b , t = dt Γ (a) X0 D. Pitzl, DESY Detectors 2.12 DESY summer students lecture 3.8.2011 Shower simulation 1 GeV e- in lead +5 cm -5 cm 10 cm photons electrons electrons positrons positrons interactive at http://www2.slac.stanford.edu/vvc/egs/basicsimtool.html D. Pitzl, DESY Detectors 2.13 DESY summer students lecture 3.8.2011 Energy measurement Total number of particles in the shower in the simple model: k Ntot = ∑k 2 = 2 kmax – 1 ≈ 2 E0 / Ec + - 2/3 of Ntot are charged (e + e ). ⇒ Nch ≈ 4/3 E0 / Ec Each e travels 1 X0 between interactions. total path length: Lch ≈ 4/3 X0 E0 / Ec Electrons and positrons also ionize the medium. Collect the charge or the fluorescent light signal: S ~ X0 E0 / Ec After calibration, S is an energy measurement! Shower fluctuations: particle production is a Poisson process. (N) = N (S) / S = 1 / S The relative energy resolution improves as 1/E0. D. Pitzl, DESY Detectors 2.14 DESY summer students lecture 3.8.2011 CMS PbWO Crystals Charged particles create scintillation light: ~120 /MeV fast: 95% < 25 ns. D. Pitzl, DESY DESY summer students lecture 3.8.2011 CMS ECAL CMS φ=85mm PbWO4 crystal calorimeter crystals: 95% lead by mass ● Barrel: ● 62k crystals, 2.2 × 2.2 × 23 cm3 ● End-caps: ● 15k crystals, 3 × 3 × 22 cm3 X0 ≈ 0.9cm Photomultiplier Tube ● Light falls on a photocathode and a photoelectron is emitted (photo effect) ‣ Quantum Efficiency depends on cathode material and wavelength ( QE ~ 25% ) ● Photoelectron focused and accelerated towards the first dynode by electric field. ● Photoelectron strikes dynode and several electrons are emitted (on average n ~ 5) ● Several dynodes ( ~ 10-15 ) give high gain ( 107) ● High speed: few ns transit time. ● Gain can be much lower in a magnetic field, depending on orientation. Source: Cutnell and Johnson, 7th edition image gallery Vacuum photo-triodes ● ~20% quantum efficiency Light ● Single-stage photomultiplier ● Gain ~ 10 at B = 4 T Photocathode 0V Grid anode 1000V Dynode 800V radiation-resistant UV glass window used in the CMS endcap ECAL. D. Pitzl, DESY DESY summer students lecture 3.8.2011 Avalanche Photodiode 85% quantum efficiency 300-400 V reverse bias: photoelectrons create cascade of electron-hole pairs in the bulk. Gain ~100 in linear mode. Low sensitivity to magnetic field. APD gain decreases by 2.3%/OC. Crystal light yield decreases by 2.2%/OC Need temperature stabilization within 0.1OC in the ECAL! 2 0 2 avalanche photodiodes per crystal in the barrel: CMS EM Calorimeter Readout Crystal D. Pitzl, DESY DESY summer students lecture 3.8.2011 Test beam calibration Response of a PbWO4 calorimeter to a 120 GeV e- test beam: CMS ECAL Test beam with final electronics Energy Position ) ) m % m ( ( n n o o i i t t u u l l o o s s e e R R Energy (GeV) Energy (GeV) σ (E) 2.4 % 142 MeV σ µ = 5040 ⊕ = ⊕ ⊕ 0.44 % Y ( m) 430 E E E E 0.6% at 50 GeV. 0.85 mm at 50 GeV. D. Pitzl, DESY DESY summer students lecture 3.8.2011 Energy resolution terms ● The intrinsic shower fluctuations give σ(E) ~ √E ● Fluctuations in the photo-electron yield also give σ(E) ~ √E ● Noise (electronics, radiation) gives a constant term: σ(E) = c ● Inhomogeneities and leakage give σ(E) ~ E σ (E) 2.4 % 142 MeV = ⊕ ⊕ 0.44 % E E E D. Pitzl, DESY DESY summer students lecture 3.8.2011 Summary so far ● Lecture 2: ▶ Interactions of electrons and photons in matter. ▶ Radiation length X , critical energy E . 0 c ▶ Electromagnetic showers ▶ Crystal calorimeters ▶ Photon detectors and readout Questions? Detectors 2.24 Higgs decay into two photons (simulation) D. Pitzl, DESY DESY summer students lecture 3.8.2011 Photon conversions in the CMS tracker + - Detector Data A A e e simulation 30% - 50% of all photons convert before reaching the calorimeter! LHC tracker material budget D. Pitzl, DESY DESY summer students lecture 3.8.2011 Electron tracks radiate radiative tail J/ J/ ee e e in front of the tracker, e track has lower momentum. D. Pitzl, DESY DESY summer students lecture 3.8.2011 Sampling calorimeters Absorber and detector are separated as passive and active layers: Absorber: Lead, or Tungsten, Uranium Detector: gas, or scintillator, silicon pads, noble liquid The active detector material samples a fraction F of the shower. The detector signal is still proportional to the incident energy. Allows longitudinal segmentation, good for hadrons. Energy resolution is degraded ~1/F ('sampling fluctuations'). Less expensive. D. Pitzl, DESY DESY summer students lecture 3.8.2011 varieties of sampling calorimeters Absorber + scintillor with photomultiplier readout Fixed target experiments Absorber + scintillator with light guide to PMT ZEUS at HERA CDF at Tevatron Pb or Fe absorber and liquid argon readout of ionization signal H1 at HERA, D0 at Tevatron, ATLAS at LHC. Gas ionisation chamber: DELPHI and ALEPH at LEP D. Pitzl, DESY DESY summer students lecture 3.8.2011 ATLAS LAr ECAL Back Cell Middle Cell Strip Cell e 3 sections: ● Pb absorber in LAr ● strips for position resolution ● Accordion geometry for ● middle for energy routing of readout signals to measurement the back ● back for leakage control ● Allows dense packing and fine granularity. ATLAS LAr ECAL Cu electrodes at +HV Spacers define LAr gap 2 × 2 mm 2 mm Pb absorber clad in stainless steel. ATLAS LAr Barrel ECAL Linearity D. Pitzl, DESY DESY summer students lecture 3.8.2011 ATLAS LAr Barrel ECAL resolution 2002 test beam data D. Pitzl, DESY DESY summer students lecture 3.8.2011 Hadronic showers Muon Tracker EM cal Hadronic calorimeter tracker electron EM showers should be fully contained in the ECAL.