Big, Fast, and Cheap: Precision Timing in the Next Generation of Cherenkov Detectors LAPPD Collaboration Meeting 2010 Matthew Wetstein - Enrico Fermi Institute, University of Chicago HEP Division, Argonne National Lab
LAPPD Collaboration: Large Area Picosecond Photodetectors The Challenge in Designing Detectors for HEP Experiments: Need to bridge the huge gap between what is possible 06/11/10 LAPPD Collaboration Meeting 2
LAPPD Collaboration: Large Area Picosecond Photodetectors The Challenge in Designing Detectors for HEP Experiments: and what is feasible (scalable) 06/11/10 LAPPD Collaboration Meeting 3
The Good News is: Sometimes it s worth the effort Georges Charpak If successful, we stand to greatly improve the quality of physics we can study in a given experiment. Because we are also forced to really understand that science behind the device, we end up Making the device better Enabling new possible spinoffs for other scientific or commercial applications In the R&D process, we are forced to imagine entirely new analysis techniques and new physics capabilities. 4
Water Cherenkov Basics Main Neutrino Interactions: Charged Current (signal) ν l Neutral Current (bkgd) ν l W ± Z 0 l ± ν l An shockwave of optical light is produced when a charged particle travels through a dielectric medium faster than the speed of light in that medium: c/n This light propagates at an angle θ C = 1/nβ with respect to the direction of the charged particle Using photodetectors, we can measure the emitted light and reconstruct the track Recoil(π 0 ) Credit: Mark Messier \ 5
LAPPD Collaboration: Large Area Picosecond Photodetectors Ring Counting Particle ID Hough transformation converts ring counting into peak counting \ 06/11/10 LAPPD Collaboration Meeting 6
Tracking in Water Cherenkov 1-D Track Multiple Tracks/ Showers t hyp = s 1 /c 0 + ns 2 /c 0 7 Fit pameters: x v,y v,z v,t 0,θ track,φ track,β Also: Particle ID? How many tracks? v=(x v,y v,z v,t 0. ) s 1 s 2 θ C In addition to t hyp, we can also use information from: Light intensity VS θ track Intensity of emitted light VS track position \ 7
Possible Neutrino Applications As a successor to photomultiplier tubes in water-cherenkov based neutrino experiments Provide better coverage Use timing information to improve tracking and vertex separation suppress largest reducible background - neutral pion fakes an electron γ π 0 π 0 boost better able to separate forward gammas with angular separations typically smaller than 15º Typical vertex resolutions in Super K O(10) cm γ γ π 0 γ π 0 boost improved sensitivity to low momentum gammas with better coverage 100 psec TOF for light in water corresponds to roughly 3 cm 8
Photomultiplier Tubes Phototubes: ~2-3 nsec time resolution Spatial resolution cannot exceed tube radius Total coverage offered is typically less than 40% Typical photocathode efficiency ~25% \ 9
Timing in Water Cherenkov Aren t time and space information degenerate? Timing Information: Can provide an extra lever arm for constraining time and space information, placing greater weight on detector hits that contain useful information (ie, those hits that haven t scattered) \ 10
Timing in Water Cherenkov Timing Information: Can help to disentangle early light of one track. π 0 γ γ \ 11
Timing in Water Cherenkov Timing Information: Can help to disentangle early light of one track. From spatially overlapping late light of another track π 0 γ γ shower \ 12
Timing in Water Cherenkov Timing Information: On average, this amounts to separating the two vertices from which the Cherenkov cones radiate... ~1 radiation length ~37 cm π 0 vertices are separated: at 7 degrees: ~4.5 cm at 15 degrees: ~9.7 cm Finding a single event vertex is limited by our ignorance of T0. Vertex separation is not... \ 13
Timing in Water Cherenkov Timing Information: On average, this amounts to separating the two vertices from which the Cherenkov cones radiate... ~1 radiation length ~37 cm π 0 vertices are separated: at 7 degrees: ~4.5 cm at 15 degrees: ~9.7 cm Imagine a likelihood fit for ring counting that uses both time and space information (cut the Hough transform into slices in time) \ 14
Understanding Chromatic Dispersion photon arrival times at 50 m Chromatic dispersion/scattering/absorption present a problem We probably won t need the same time resolution as collider applications Still, even at 50 meters, we can expect to do much better than 2 nanosecond resolution in water Typical PMT timing resolutions > 1ns ~2ns rise time Uncertainty (psec) 600 450 300 150 Uncertainty on Arrival Time 100% Coverage 30% Coverage 10% Coverage J. Felde, B. Svoboda: UC-Davis Resolution losses over large distances in water can be recovered with more coverage (photon statistics) 0 20 50 80 Distance from Detector (meters) 15
Understanding Timing in Water Cherenkov TrackFit_x Package for analytic track-fitting based on Cherenkov geometry Currently optimizing multi-parameter fitting and smoothness of likelihood curve Goal: to study identification of π 0 backgrounds as a function of time resolution To better understand analysis using largearea, picosecond photodetectors Expected arrival time dist at fit minimum Measured arrival time dist M. Wetstein(ANL/UofC), M. Sanchez (Iowa State/ANL), B. Svoboda (UC Davis) 16
Understanding Timing in Water Cherenkov TrackFit_x RMS of Vertex Fits In Muon Direction vs. Detector Resolution (50 trials) RMS of Fitted Vertex in Muon Direction (mm) 140 120 100 80 60 40 20 0 0 1 2 3 4 5 6 Timing Resolution (nsec) Simple test of the refined track fitting: No chromatic dispersion straight muon trajectory 100% covereage T0 is known Proof of methodology M. Wetstein(ANL/UofC), M. Sanchez (Iowa State/ANL), B. Svoboda (UC Davis) 17
180 160 140 120 100 80 60 40 20 14000 12000 10000 8000 6000 4000 2000 3 10 Hits per PMT for 50, 5 GeV Muons 0 0 5 10 15 20 number of hits True Vs Measured PMT Hit Times 0 0 100 200 300 400 500 hit time (nsec) M. Wetstein(ANL/UofC), M. Sanchez (Iowa State/ANL), B. Svoboda (UC Davis) Next Steps Generate data directly on the official, LBNE Geant model. Apply tracking algorithms directly to PMT data. Test tracking algorithms in more realistic scenarios (include chromatic dispersion, etc). Coordinate with the LBNE algorithms group. build tools to do particle ID/ring counting keep the algorithms flexible over variations in detector resolution, granularity, and geometry 18
Possible Neutrino Applications Finer spatial resolution = Ability to resolve dn γ /dω within a single module better particle ID (ability to resolve the sharpness of ring-edge) better able to reconstruct events close to the wall Could imagine new tank geometries (building walls closer together in direction orthogonal to beam?) Other Possible Advantages: Better magnetic susceptibility (applied magnetic field?) Further cost reductions by requiring less bulk mass for the same physics cheaper excavation costs Will require detailed simulation 19
Summary of Potential Payoffs Improved π 0 /electron separation. Better vertex resolution. Additional forward ring separation. Lowered threshold for lower energy gamma detection. Reduced magnetic field susceptibility compared to PMTs. Increased fiducial volume by designing flat photodetectors. Lessened constraints on cavern height, thanks to geometric design. Effort is under way to study each of these possibilities. 20