The ICARUS experiment Between LNGS and FNAL
Christian Farnese Università di Padova and INFN on behalf of the ICARUS Collaboration
ICNFP 2018, Crete, July 12th Outline Present ICARUS Collaboration l LAr-TPC technology: ICARUS T600 performance and results @ LNGS. Brookhaven National Laboratory (BNL), USA Colorado State University, USA l Search for sterile neutrinos at FNAL: Fermi National Laboratory (FNAL), USA the Short Baseline Neutrino University of Houston , USA Experiment. INFN Sez. di Catania and University, Catania, Italy INFN GSSI, L’Aquila, Italy l Generalities on the ICARUS T600 INFN LNGS, Assergi (AQ), Italy overhauling. INFN Sez. di Milano Bicocca, Milano, Italy l T600 current status. INFN Sez. di Napoli, Napoli, Italy INFN Sez. di Padova and University, Padova, Italy l Conclusions. INFN Sez. di Pavia and University, Pavia, Italy Los Alamos National Laboratory (LANL), USA University of Pittsburgh , USA University of Rochester , USA SLAC, Stanford, CA, USA University of Texas (Arlington), USA
+CERN and others INFN groups involved in the SBN program Spokesman: C. Rubbia , GSSI Slide# : 2 ICARUS T600: the first large Liquid Argon TPC (760 t of LAr)
l ICARUS-T600 LAr TPC is a high granularity uniform self-triggering detector with 3D imaging and calorimetric capabilities, ideal for ν physics. It allows to accurately reconstruct a wide variety of ionizing events with complex topology. l Exposed to CNGS beam, ICARUS concluded in 2013 a very successful 3 years run at Gran Sasso INFN underground lab, collecting 8.6x1019 pot event statistics, with a detector live time >93%, and cosmic ray events. Two identical modules: 476 t total active mass: 1 T600 module Wire planes (anode) l 2 TPC’s per module, with a common
© 2016-2018 CERN central cathode: EDrift= 0.5 kV/cm, vDrift~1.6 mm/µs, 1.5 m drift length; l 3 ‘’non-destructive’’ readout wire planes per TPC, ≈54000 wires at 0°, ±60° w.r.t. horizontal: Induction 1, Induction 2 and Collection views; l Ionization charge continuously read (0.4 µs sampling time); l 74 8” PMT’s, coated with TPB wls, Cathode for t0, timing and triggering. Field cage PMTs Slide# : 3 ICARUS LAr-TPC performance (CNGS ν’s and cosmics) l Tracking device: precise 3D event topology, ~1 mm3 resolution for any ionizing particle; Edep distrib. for l Global calorimeter: full sampling homogeneous νµ CC events calorimeter; total energy reconstructed by charge integration with excellent accuracy for contained events; momentum of non contained µ by Multiple Coulomb Scattering (MCS) with Δp/p ~15%; l Measurement of local energy deposition dE/dx: Momentum resolu on for different track remarkable e/γ separation (0.02 X0 sampling, µ lengths X0=14 cm and a powerful particle identification by dE/dx vs range):
dE/dx (MeV/cm) vs. Low energy electrons: residual range (cm) σ(E)/E = 11%/√ E(MeV)+2% for p,π,µ compared to Electromagnetic showers: Bethe-Bloch curves σ(E)/E = 3%/√ E(GeV) Hadron showers: σ(E)/E ≈ 30%/√ E(GeV) Validation on pMCS of stopping µ’s, compared with calo estimate.
Slide# : 4 νe CC identification in CNGS beam: Electron/gamma separation Three “handles” to separate e/γ and reject NC background: l reconstruction of π0 invariant mass l dE/dx: single vs. double m.i.p. l γ conversion separated from primary vertex 1 m.i.p. 2 m.i.p. Collection view (mainly 35 cm γ (pair Compton) production) γ 105 cm
Proton Shower Electron starts
Collection view
Evolution in Collection view from single m.i.p. to e.m. shower evident from dE/dx (MeV/cm) on individual wires. Atmospheric neutrino events @ LNGS
ICARUS collected @ LNGS also atmospheric νe and ν CC interactions µ These events are particularly suitable to emulate the ν interactions expected with FNAL beams (more on this later) because of the similar energy range
Example of upward-going νµ CC even with a Muon deposited energy ~ 1.7 GeV: l 4m escaping µ, 1.8±0.3 GeV/c from MCS; Proton Pion l Two pions (Edep~80 MeV) and a proton (Edep~250 MeV) at vertex.
l Reconstructed Eν~2 GeV with ~78°zenith angle
Proton Collection 23 cm Downward-going, quasi elastic νe event. view deposited energy: 240 MeV l dE/dx ~ 2.1 MeV/cm measured on first wires corresponds to a m.i.p. Electron 30 cm l Short proton track recognized. Atmospheric neutrino events @ LNGS – preliminary l Cosmic ray events recorded in ~0.48 kton y exposure (2012-2013 run) analyzed to identify atmospheric ν events. Cosmic µ tracks collected in the l Incoming c-rays identified and WEST and EAST module; the rejected by factor ~100 different rate due to the different numbers of PMTs in the 2 cryostats l ν CC candidates with Edep>200 MeV are automatically pre-selected (~80%/~25% efficiency for νe/νµ), then validated by visual scanning; MC expectations l Globally 6 νµCC and 8 νeCC atmospheric neutrino events Identified identified, to be compared with 18 neutrino events expected events, evaluated taking into account of the trigger, filtering and scanning efficiencies;
see C.Farnese’s poster “Atmospheric neutrino search in the ICARUS T600 detector”
ICARUS results and search for an LSND-like effect ICARUS run at LNGS allowed reaching several physics/technical results demonstrating the maturity of the LAr-TPC technology: l An exceptionally low level ~20 p.p.t. [O2] eq. of electronegative impurities in LAr; the measured
e- lifetime τele >15 ms ensured few m long drift path of ionization e- signal without attenuation; l Demonstrated detector performance especially 0 in νe identification and π bkg rejection in νµ-νe study to unprecedented level; l Performed a sensitive search for LSND-like anomaly with CNGS beam, constraining the LSND window to narrow region at: Δm2 < 1 eV2, sin22θ ~ 0.005 where all positive/ negative experimental results can be coherently accommodated at 90% C.L., confirmed by OPERA. Eur. Phys. J. C (2013) 73:2599 Success of LAr-TPC technology with large impact on neutrino and astro-particle physics projects: SBN short base-line neutrino program at FNAL with 3 LAr- TPC’s (SBND, MicroBooNE and ICARUS) and the multi-kt DUNE LAr-TPC. Slide# : 8 3
TABLE I: The expected (unconstrained) number of events for QE the 200
20 1.8 energyLSND range for the total 12.84 10 POT data. Each ν : 12.84×1020 POT QE ⇥ 1.6 e bin of reconstructed E corresponds to a distribution MiniBooNE 20 ⌫ ν : 11.27×10 POT of “true” generated neutrino energies, which can overlap 1.4 low-energy excess e adjacent bins. In neutrino mode, a total of 1959 data 1.2 381.2 ± 85.2 ν events
Excess Events/MeV e events pass the ⌫e CCQE event selection requirements 1 (4.5 σ evidence) QE with 200 with mnew < 0.24 eV; l No evidence of νµ disappearance in MINOS and IceCube in 0.32-20 TeV; l Recent reactor data (especially NEOS) Reactor Antineutrino Anomaly best fit value Δm2~2.4 eV2, sin2(2θ)~0.14 are intriguing but still not conclusive. l New results are expected from ongoing and upcoming experiments at reactors. free=no assumption on flux norm. & spectra fixed=prediction & Allowed regions published values used @ 95% CL THE EXPERIMENTAL SCENARIO CALLS Result presented Slide#Slide# : :10 10 FOR A DEFINITIVE CLARIFICATION! @ Neutrino2018 Short Baseline Neutrino (SBN) in a nutshell L/Eν ~ 600 m / 700 MeV ~ O(1 m/MeV) BNB spectrum T600 also off-axis on NUMI beam: Asset for DUNE TARGET 110 m 600 m 470 m FAR DETECTOR: MicroBooNE T600 – 476 ton 89 ton NEAR DETECTOR: SBND – 82 ton 13/03/2017 3D MODEL Sensitivity to Sterile neutrino @ SBN SBN can clarify the issue by exploiting similar νe appearance: LSND 99% CL LAr-TPCs at different distances from the target region covered at 5 σ level l SBND will give the “initial” BNB flux/composition l ICARUS, as far detector, will characterize the ν oscillation parameters. Δm2 = 0.43 eV2 sin2 2θ = 0.013 SBND @ T600 @ 100 m 600 m Both plots: 6.6x1020 pot (3 years) 3-5 σ νµ disapp. SBN sensitiv. Δm2 = 1.1 eV2 sin2 2θ = 0.1 SBND @ 100 m T600 @ 600 m Slide# : 12 Taking data @ shallow depth ICARUS at FNAL is facing a more challenging experimental condition than at LNGS, requiring the recognition of ν interactions amongst 11 KHz of cosmic rays. l A 3 m concrete overburden will remove contribution from charged hadrons/γ’s. l ~11 µ tracks will occur per triggering event in 1 ms TPC drift readout: associated γ’s represent a serious background source for νe search since e’s produced via Compton scatt./ pair prod. can mimic a genuine νe CC. Cosmic rays (Pavia test) low energy CNGS neutrino events Rejecting cosmic background, i.e. reconstructing the triggering events, requires to precisely know the time of each track in the TPC image: l A much improved light detection system, with ~ns time resolution; l An external cosmic ray tagger (CRT) to detect incoming particles and measure their direction of propagation by time-of-flight: ü Scintillating bars surrounding T600 (aim: 98% coverage) Top equipped with optical fibers to convey light to SiPM arrays. ü Top coverage under INFN/ CERN responsibility. FNAL is recovering modules by MINOS/Double Chooz for side/bottom. Slide# : 13 ICARUS T600 Overhauling at CERN (WA104/NP01) To face the new experimental situation at FNAL (shallow depth data taking with higher beam rate) ICARUS T600 detector underwent an intensive overhauling at CERN in 2015/17 in the framework of CERN Neutrino Platform (WA104/NP01 project) before being shipped to FNAL: l New cold vessels, purely passive insulation; T600 leaving from l CERN June 12th 2017 Renovated cryogenic / LAr purification equipment; l Flattening of TPC cathode: few mm planarity; l Upgrade of light collection system; l New higher performance TPC read-out electronics T600 in Antwerp: unloading from barge from Basel and loading into T600 arriving at SBN Far site ship to Burns Harbor (Michigan lake) building @FermiLab, July 26th 2017 Slide# : 14 The light collection system In ICARUS, light collection is used to: l Identify precisely the time of occurrence (T0) of each interaction; l Identify the event topology for fast selection purposes l Generate a trigger signal to enable the event read-out by combining: Ø Pattern/majority of hit PMT signals 2 ns Ø BNB/NuMI bunched beam spill Ø Veto from CRT The light collection system is based on 360 PMT’s, 90/chamber, to have: 1. High detection coverage, to be sensitive to the lowest-expected neutrino energy deposition in the TPC (approximately 100 MeV), also using the light fast-component only; 2. High detection granularity, longitudinal resolution is better than 0.5 m (effective Q.E. = 5%). 3. Fast response time/ high time resolution (≈1 ns), with a PMT timing calibration provided by a laser system (Hamamatsu PLP10, λ~450 nm, FWHM<100 ps, peak power ~400 mW) + 50 µm optical fiber. Slide# : 15 PMT layout l 90 PMT’s per TPC layout: 5% cathode coverage area, 15 phe/MeV deposited energy collected. PMT ~ 3m ~ vertical direction ~ 18 m Beam direction l Hamamatsu R5912-MOD (8”, 10 dynodes) Optical Fiber rated for cryogenic temperature PMT Base Optical (cathode with platinum under-layer). Fiber P l Each PMT is enclosed in a wire screening Exit Signal M cage to prevent induction of PMT pulses T Cable on the facing TPC wires. Screening l PMT glass windows coated by ~200 HV Cage µg/cm2 of Tetra-Phenyl-Butadiene wavelength shifter to detect the Possible cosmic µ‘s identification λ = 128 nm scintillation light in LAr. provided by pattern/time recognition of PMT signals Slide# : 16 The new TPC read-out electronics ICARUS readout electronics l One mini-crate, ICARUS mounted electronics on the flange, at can LNGS host was9 boards based serving on analogue 576 channels, low noise64 channels “warm” each. front- The boards areend directly amplifier, connected a multiplexed to the proprietary 10-bit flanges.2.5 MHz ADC and a digital VME module Each board hostsfor local64 front storage,-end low data noise compression, charge sensitive trigger pre-amplifiers information:, 64 serial 12 bit ADC (2.5 MHz), 0-3.3V railS/N to rail ~9 input in Collection, range, FPGA, ~0.7 memory, mm singleoptical hitlink resolution,interface… resulting in a precise The backplane ofspatial the crate event distributes reconstr the. powerand µ supply momentum and local measurement control signals. by MCS. Flange l Improvements concern: CAEN ExternalA2795 side board, 64 chs Internal side Ø Serial 12 bits ADC, one per ch, 400 ns sampling synchronous on the whole detector; ICARUS readout electronics Ø Serial bus architecture with Gbit/s optical One minilinks-crate, mounted on the flange, can host 9 boards serving 576 channels, 64 channels each. to increase the bandwidth (10 MHz); The boards are directly connected to the proprietary flanges. Each board hosts 64 front-end low noise charge sensitive pre-amplifiers, 64 serial 12 bit ADC (2.5 Ø Both analogue/digital electronics are housedMHz), 0in-3.3V rail toFlange rail input range, FPGA, memory, optical link interface… a single board inserted in a new mini-crate The backplane of the crate distributes the power supply and local control signals. directly installed on ad-hoc signal feedthrough Backplane connector Custom made UHV thigh technology (no through-going holes, electrical H.V. (<±500 V) flanges REG REG A2795 connections through staggered PCB layers) Trigger Liquid argon Gas 18 connectors (32 channels each) for a total of 576 channels. ExternalTT side: connectors for CAEN boards insertion. Sense wires - Twisted pair cables link (~1.8-6.2 m, 50pF/m) (4-9m, 20pF/m) ADC Internal side: SMD connectors to receive decoupling boards or cables. 8 additional SMA connectors for test pulse distribution and wires biasing Fibers optical … … FPGA link Induction(32ch) (~ ±500V) . 10 liters mini-crate ADC Slide: 5 Test IN hosts 9 boards (576 RAM channels) Collection(32ch) Decoupling board UHV Front-end 12bit serial ADC Feed-through amplifiers 400ns sampling Backplane connector (18x32ch) (64/board) Slide# : 17 Slide: 2 H.V. (<±500 V) REG REG A2795 Trigger Liquid argon Gas Sense wires TT - Twisted pair cables link (~1.8-6.2 m, 50pF/m) (4-9m, 20pF/m) ADC Fibers optical … … FPGA link Induction(32ch) ADC Test IN RAM Collection(32ch) Decoupling board UHV Front-end 12bit serial ADC Feed-through amplifiers 400ns sampling (18x32ch) (64/board) Slide: 2 Improved front-end electronics for T600 l In the newT600 analogue front-end the adopted improvements concern: Ø A faster shaping time ~0.6 µs of analogue signals to match electron transit time in wire plane spacing; Ø A drastic reduction of undershoot in the preamp response as well as of the low frequency noise while maintaining a same or better S/N; Ø A same preamp for Induction1,2 and Collection wires allowing dE/dx measurement in Induction2 too. l In addition the full 400 ns synchronous signal sampling on the whole detector will allow slightly improving the resolution on µ momentum by MCS. Ø Same ~2ADC counts (~1500 e-) noise for Collect. & Induct.; Ø Unipolar Coll signal: ~ 25 ADC counts; Symmetric bipolar Ind. Signal Ø After signal integration by a running sum and baseline restoring, a S/N ~ 10 is recognized in Induction view Better event reconstruction provided! arXiv:1805.03931 Single 45° m.i.p. µ track Paper submitted to JINST Simulation of ICARUS electronic response !! Electronic response has been simulated, using a common response function for all views: *366/7831( -t/" Inner light t/simulation" e (" =1.5andµs )reconstruction - 1 ICARUS!! Approximate event signalreconstruction normalization according @ SBN to the 1/2 l Common SBN framework (LarSoft) used, providing Typical wire shape for test at CERNSimulation (Collection of ICARUS measurements); electronic response1 m.i.p. µ tools to simulate (Geant4), reconstruct/identify events (cosmic!! Ongoing µ’s, activity e.m .in showers, view!!!! ElectronicTypicalof neutrinos, signals response for has a… muon been). simulated,(m.i.p.): using a common response function for all views: *366/7831(;1<27831(>( l Experimental"! Complete a geometrymore realistic setup and configurable is described readout in simulation LarSoft module. !! t/" e-t/" ("=1.5µs) "! Noise distributions extracted from ICARUS l ScintillationProvide input light to triggerin LAr study is parameterized to simulate tdrift/400ns !! Approximatefrequency spectra signal normalization measured at according LNGS (toto thebe updates PMT" ! signalsPrepare pattern for any recognition MCtest event, at CERNthrough to(Collection lightstudy measurements); trigger and event recognition. "! Algorithms for handling!! Typicalas lightsoon signals hitsas we and for get light a noisemuon flashes (measurementsm.i.p.): (multiple hits) at FNAL l MC simulations include new wire electronic response, ;1<27831(>(;1<27831(=( !! Noise distributions extracted from ICARUS realistic noise, as well as PMT scintillation light signals. tdrift/400ns frequency spectra measuredSimulated at LNGS signal (to on abe PMT updates as soon as we get noise measurements at FNAL LNGS Noise Frequency spectra ;1<27831(=( !! New electronics simulation tdrift/400ns Detectorcode available (more ;1<27831(=( *366/7831( hall LNGS Noise Frequency spectra Ind.2 signal simulation, including noise detailed description of the Ind.2 signal, including noise Simulated PMT signal Overburdenadopted electronics and ;1<27831(=( amplitude *366/7831( Ind.2 signal simulation, including noise trigger{ will Warmbe soon vessel included) TPC CRT Cryostat Time (µs) 0 0.6 1.2 Time (µs) May 18 ICARUS collaboration meeting 0 0.6 1.2 0 0 0.6 0.6 1.2 1.2 Slide# : 18 Slide# : 19 F (MhZ) F (MhZ) May 18 ICARUS collaboration meetingF (MhZ) F (MhZ) Slide: 9 May 18 ICARUS collaboration meeting Slide: 9 •! First implementation of Optical hit finding for ICARUS ICARUS event reconstruction @ SBN 2/2 Some advanced tools already ported in Larsoft: Reconstructed λ vs simulated one, in l LAr purity λ=1/τele (τele: electron lifetime) 2-15 ms e-lifetime measurement from charge attenuation of range cosmic µ’s tracks along the drift Work in Ø Track selection at shallow depth difficult progress! Deconvoluted Reconstruction of e.m. showers in ICARUS signal hits due to crowded events and lower energy µ’s • The electromagnetic shower identificationl Particle ID, and basedmeasurement on dE /dxis a key vs range ingredient for the νeCC analysis Raw signal hits l Electromagnetic shower axis identification • Example of reconstruction of the shower axis based on an algorithm already developed for ICARUS at LNGS Ø Provides 3D reconstruction of shower dE/dx vs range for µ’s Δθ between reconstructed and true axis • First results for 800 MeV electrons Particle ID Preliminary isotropically oriented Track inclination: 0.35 ÷ 2.4 cm per wire " The average angle between the true and reconstructed direction is ~20 800 MeV isotropic e- " Only 56% efficiency: needs Δθ=20 ---- Expected dE/dx improving the hit/cluster reconstruction! Δθ • shower identification tools need toSoftware be prepared! is mature enough to realistically simulate events with BNB beam Slide# : 20 • dedicated regular meetings to share progress on these items already started: very welcome participation and contributions! May 18 ICARUS collaboration meeting Slide: 16 Proton Proton Electron Electron Electron Electron 45 cm cm 45 1 m m 1 BNB (MC) and real atmospheric atmospheric real BNB (MC) and E E 80 cm cm 80 dep ν = 1.26 1.26 = = 1.2 1.2 = ν e Proton Proton CC GeV GeV 76 cm MC SBN MC SBN BNB beam (MC) LNGS LNGS ν ν e e ATMOSPHERIC EVENT ATMOSPHERIC CC interactions interactions ν e CC l l l showering single dE/dx E Electron identified by by identified Electron by identified Proton Quasi-elastic events comparison events Dep l l = 0.9 GeV. 0.9 = interactions for (below) @LNGS events atmospheric Similar results hold Similar very alike to typical . m.i.p νµ . before . before CC ν e CC ν eCC Slide# : 21 21 : Slide# ICARUS @ FNAL 1/2 - Status l Warm vessel floor/walls were assembled/installed in the pit in the Far Detector (FD) building in 2017. l Bottom CRT modules (200 m2 total area) already installed in 2017. l Assembly of cold shields completed (May 2018), now under leak test. l Installation of detector supports is in progress. Should be done by July 2018. l Main vessels doors sealed. Helium leak tests ongoing… ICARUS Vessel 2 parked west of SBN Far Detector Building Warm vessel, cold shield (June 2018) ICARUS @ FNAL 2/2 – 2018 plans and commissioning l Detectors insertion in Warm Vessel –July-August 2018 l Top part of cold shield will be then installed and tested, followed by installation of top part of warm vessel (August – September 2018). l From Fall 2018, activities on top of detector will start (cryo, purification and vacuum systems, ext. cabling, read-out & decoupling boards, feedthrough flanges, optical fibers,…) l Vacuum pumping should start by late 2018 / early 2019 and last until ready to start cool-down. Detector commissioning will consist of three phases: l Cryogenic commissioning: Vacuum (1 month), Cooling (15 days), Filling (15 days), Purification (1 month), Stabilization (1 month). l TPC and PMT system commissioning (2 months in total): HV system, PMT’s supply, calibrations, DAQ & trigger commissioning. l CRT commissioning: in parallel with the activities for the completion of cryo, TPC and PMT system commissioning. Slide# : 23 Conclusions l LAr-TPC detection technique taken to full maturity with ICARUS-T600. l ICARUS completed in 2013 a successful continuous 3-year run at LNGS exposed to CNGS neutrinos and cosmic rays, and performed a sensitive search for a potential νe excess related to the LSND–like anomaly defining a narrow region at (Δm2, sin2 2θ) ~ (1 eV2, 0.005). No excess evidence, as confirmed by OPERA. l The T600 underwent a major overhauling at CERN and has been transported to FNAL to be exposed to Booster and NuMi neutrinos. l SBN experiment will provide a clarification of the sterile neutrino issue, both in appearance and disappearance modes. l Installation in the Far Site building @ FNAL is in progress Ø vacuum pumping should start at the earliest in winter 2018. Ø Detector commissioning while waiting for clearance by FNAL (by Jan - Feb 2019) to start cool-down.Then data taking for physics! Slide# : 24 Thank you ! LNGS_May2011 Slide 25