1. Time-of-Flight Working Group From Space to Spacetime Conveners: Bob Wagner (Argonne) & Mike Albrow (Fermilab) Presentations in our sessions Bob Wagner: Introduction in plenary session Mike Albrow : Introduction to parallel sessions Fritz DeJong : μ  eγ, μ  eee Andrew Brandt: ATLAS forward proton timing Mike Albrow: CMS forward proton timing Andrey Elagin: Timing properties of MCP with Atomic Layer Deposition (ALD) Erik Ramberg: Fermilab test beam and timing detector development Vic Scarpine: Ultra fast lasers fro beam diagnostics Bob Velghe: GigaTracker (Low mass Si pixel tracker with timing) Mike Albrow: Fast EM calorimeter (energy, direction, timing) concept Yuri Kamyshkov: Timing in n-nbar conversion Thanks to all speakers! Other parallel session talks concerned good timing, but some we missed (unfortunate time overlap). Possibly I miss something important (tell me!) PROJECT X PHYSICS STUDY WORKSHOP (PXPS 2012)

2. X credit:

3. Uses for timing, classes, mostly not PX, not all needing super-precision): Together with p or E : identification of particle type (PID) in beam (small area ~ cm 2 ) OR in large area (~10 m 2 ) e.g. CDF/D0 size OR very large area (~100s m 2 ) e.g Large water Cherenkov Together with another time (or reference time): Position of origin in space (spacetime) Again: cm 2 areas (e.g. forward protons at LHC, p + X + p, Brandt, Albrow) ~ m 2 (PET) to 100’s m 2 areas (e.g. photons in K 0  π 0 υ υ, Littenberg) Actual time an event occurred Examples: Time of a neutrino interaction (wrt production) for speed (mass?!) measurement Supernova neutrinos Background reduction (referred to a pulsed beam (20 ps!) / source, e.g.) Within a < 1ns bunch crossing at LHC Directionality Example: timing shower front in AUGER for cosmic ray direction, source search. c = 300,000 km/s = 3 mm/10 ps

4. Large area MCP-PMTs for timing on EM shower in water Cherenkov (direction, size, etc.) Same thing, different scale! AUGER Cosmic ray observatory Could Auger get better pointing with better timing? Don’t think so, but it could be improved if it helps

5. Detector Techniques: Cherenkov light (solids, liquids, gases) : prompt, directional, speed-dependent (both amount and direction) Very large masses (ICECUBE, HYPER-K) possible with solid & liquid radiators. Gases for low-mass, e.g. timing intense beam, few mm low pressure gas (rad hard) Detecting the Cherenkov light: Classic PMTs, MCP-PMTs (faster) – Large area development, SiPMs good for small areas. Scintillation light (solids, liquids, gases) : Classic, not as fast (but maybe fine), Wavelength shifting collection also “slow”, inefficient (< 1/6), but area reduction can be crucial. Ionization (gases) : E.g. Drift chamber (to wires) main purpose tracking, but multiple time measurements for free. or TPC, timing  track position reconstruction from time of drift. Solid state trackers: E.g. Silicon pixels or SiPMs with added timing

6. Parameters of importance (sometimes) Time resolution: ps, 10ps, 100ps, ns,..? Size of unit area: cm 2 ? m 2 ? # channels and Cost per unit area or per channel Electronics (speed, cost, location) to read out to DAQ Thinness, physical and probably in radiation lengths X 0 Radiation hardness? Rates... from 10 10 /s to 10 -8 /s ! Calibration Electronics issues Crucial issue... speed, #channels, integration Calibration Time information required (or useful) in 1 st level triggers? Into fast trigger processors perhaps. Does it need to be rad hard? compact? Reference time signals, <ps clock and transmission (PLL?) Time stretcher electronics x10, x50...

7. Erik Ramberg: Fermilab test beam and timing detector development Facility: beams up to 120 GeV/c p ~ 40K / min in 2 sec spill over ~ 1 cm 2 User groups come for a week or more. Excellent infrastructure (see ER slides) Series of timing detector studies: Cherenkov radiators (quartz, aerogel, gas) in developing geometries (Brandt, Albrow talks) with MCPs and SiPMs.

8. MicroChannel Plate photomultipliers: MCP-PMT Faster than conventional PMTs, compact, gains ~ 106. Commercial are more expensive with modest areas (<~ 10 cm2) Large area flat thin being developed (talks here) Rise time 68 ps! From Bob Wagner’s Intro PILAS laser signal

9. Time of Flight p/π separation: 2 angled-bar QUARTICs + PMT240 over 8m e/μ/π aligned at 0 Protons SiPM signals count photoelectrons proton quartz bar: 30mm  16 ps SiPM Ramberg’s talk

10. Timing two far-apart (430m) protons to determine origin (z) and reject pile-up background Goal is σ(1-2) ~ 10ps (  σ(z) ~ 2 mm). Small area (2 cm 2 ), edgeless, close (4mm) to 1ry beam radiation hard (~ 10 15 protons), high rates (~ 50 MHz), 25ns/X + reference time “clock” @ 400 MHz (LHC RF) Not discussed here, but < 2 ps with PLL is standard (ILC developments) Andrew Brandt ATLAS Forward Protons: Timing at 10 ps level required for PU rejection

11. QUAR TZ TI MING C HERENKOV (QUARTIC) Principle: Cherenkov light at θ ch ~ 48 o propagates along N bars (if ~ good φ) (wavefront; isochronous ) hitting separate anode pads in PHOTONIS 8x8 channel MCP-PMT. Segmented (x not y) Edgeless (on side) Only quartz (rad hard) close to beam Fermilab test beam Jan 12: Bottom line: 27ps /bar implying <~ 10 ps for 8 bars. BUT: Photocathode lifetime limited, + ion feedback. Ion barrier development, or Arradiance coatings. Work in progress From Brandt’s talk

12. AB Univ. Texas Arlington set up for timing studies, especially QUARTICFrom Brandt’s talk

13. Requirements of timing system: Time resolution ~ 10 ps Area ~ 8 mm (V) x 20 mm (H) Segmentation for > 1 proton/bunch Edgeless, active to ~ 200 μm from pipe Radiation hard, ~ 10 12 /cm 2 Lifetime > ~ 1 year at LHC at 10 34 Rate: 25 ns sensitivity Reference time system (“clock”) with < few ps jitter at stations 500m apart Mike Albrow: CMS Forward proton timing requirements (as for ATLAS AFP) L-Bar Quartic: different geometry from angled bars: Principle: ALL Cherenkov light to back and up LG bar (1 st pass or later). SiPM not rad hard but no lifetime issue. Far from beam (low rad). Repeat N times (N = 4 – 8) Good x,y segmentation (even 1mm x 1mm possible) 4 in-line at Fermi Test Beam (Feb) : 30 & 40 mm bars

14. 1 Lbar vs PMT240 (after time slewing corr)  32 ps intrinsic  16 ps 4 in-line 70V on SiPMs, Clip signals (90pF) & x 20 preamp.  DRS4 5 GHz waveform digitiser Read 8 traces (200 ps/point) 20 mV/div. & 2 ns/div GEANT simulation of time of photoelectrons:fast peak and then after back – front – back trip Good total internal reflection is crucial Faster (4x) SiPMs, STMicroelectronics N-on-P available, other improvements on way. Multiple measurements with segmentation: “time-track”. Beam halo monitor application. MGA talk

15. Idea: Beam (intense) monitor (position and time) with similar principle Array of SiPMs (e.g.) 1mm x 1mm Amplitude & time (waveform e.g.)... CCD or digital camera (no time info) BEAM Al Mirror (v. thin) GAS Cherenkov radiator, e.g. He He @ 1 atm refr index (n-1) = 35.10 -6 sin θ ch = 0.0084 (β = 1) (very narrow cone) # detected photons ~ 90.sin 2 θ ch /cm ~ 0.0006/proton/mm. Window (v. thin) Too much light!! Can : reduce radiator length (subject to clear light path) reduce gas pressure (subject to keeping very thin window+mirror) Filter light (neutral density or color filters) Tilted entrance window so path length independent of x,y in beam Could get I(x,y,z,t) for intense beam (Complement classic BPMs) Remember sodium vapor curtain method? MGA VAC Could flip out of beam when not in use.

16. Timing properties of MicroChannel Plate PMTs with Atomic Layer Deposition Andrey Elagin (LAPPD = Large Area Picosecond Photo Detectors) Argonne-Chicago-Fermilab collaboration Only ~ 0.5 X0 Can be inside calorimeters BIG NEWS! also R&D 100 AWARD announced today Water Cherenkov? K 0  π 0 νν ?

17. Henry Frisch:

18. Ron Lipton: Power reduction maintaining fast timing? SiPMs with tracking too?

19. Bob Velghe Adding timing to pixel tracker & time-over-threshold  PH Made, tested, now making full system Important advance for beams & modest areas, but not (or not yet) for large multi-m 2 detectors (electronics coverage & power consumption). But a dream; central tracking in central LHC with “4π” detectors with fast timing too!

20. Timing depends on: GOOD NEWS Bob Velghe

21. Fritz DeJong : μ  eγ, μ  eee Huge background from μ  eνν (100%) (Michel), want 10 -16,-15 level rejection. Timing can help. Photons only ~ 53 MeV. Convert them, time e+e- in pixel tracks... Gigatracker with Si pixels with timing FDJ Timing add-ons to Si tracking: major applications, but power and thinness issues for large areas. In μ  eγ, huge fluxes, combinatorial issues from Michel, 2-3 points/track, can this work ???

22. Ron Lipton (from tracking session) Timing with tracking: 3D devices

23. Henry Frisch: Large area MCP-PMTs in radiator sandwich for PET-TOF : concept On to CALORIMETRY

24. M. Albrow Just a concept, not optimized. Probably good (cf scintillator, WLS, PMT) for few GeV γ’s, e’s... probably not for 500 MeV γ’s. Higher energies (multi-GeV) ?or not?

25. Motivation: ~ 200 m neutral beam: E & t  Mass. K0L, n/nbar, Λ/Λbar, X0, strangelets?? Borer, Pretzl, MGA et al 90 ps resolution with 30 GeV π+, with energy, shower shape and position MGA

26. Jet timing at LHC ? Rare p + (H  Jet+Jet) + p events with ~ 30 – 40 inelastic interactions in bunch crossing Pile-up (accidental coincidences) are a major problem. Time difference between protons is one method, but does not resolve two events at same place, different times. Want central event (jets) time. Also help matching jets to collisions if several jetty events.

27. Yuri Kamyshkov: timing in n-nbar mixing Slow reactor-n beam test setup Detector end Biggest challenge Huge n background

28. MGA, tx Vic Scarpine GigaHertz Streak camera? Just a concept: TIME  POSITION at few ps level Problem: velocity spread of electron bunch. Solutions???

29. R&D DIRECTIONS for Project X? Scale of order 0.2M$ - 2M$ ? Large area fast photodetectors, e.g. Photocathodes and MCPs (Elagin’s talk) Material budget minimization (inside tracking?) (Done in ANL-Chicago-FNAL Coll.) Timing with “Si”-tracking a la GigaTracker (Done in NA62 eg) Cherenkov radiator materials (Interesting but not $$$) Solid state photodetectors e.g. SiPM and others (Big industry) Geometries (experiment specific) (Thinking, simulating, testing... not $$$) Very fast electronics (e.g. CERN HPTDC 25ps  10 ps) ( Maybe... did not address) New ideas: streak camera detector? jet timing ring? X? Y ? (Who knows?) Applications not to forget (keep in mind): Non-HEP applications, e.g. PET (of course) Other medical? Industrial (source location “bomb in truck”) You name it. Opinion: Probably any PX major research effort in timing would be in combination with calorimeter or tracking

30. Summary Most experiments use timing at some level. Could your experiment benefit from more and/or better timing? If so, technology is advancing... perhaps there is a solution. Challenges drive innovation. Good progress in several areas: Small (~ cm 2 ) Cherenkov detectors with MCP or SiPM readout at 10 ps level Medium (~ 10 cm 2) Timing of Silicon pixel trackers (Gigatracker)... breakthrough! Large area (~ 400 cm 2 ) thin MCP’s... breakthrough! “Wallpaper” large detectors, insert as layers inside calorimeters (shower timing)! Make your detector relativistic! Think SPACETIME not just space!

31. Thank you! Back Ups  & special thanks to co-convener Bob Wagner

32. Gas detectors timing characteristics (shown by Bob Wagner)