The J-PARC KOTO Experiment Yau WAH Fermilab Project-X Workshop June 2012 1
K0 at To kai (KOTO) for the rare decay 2
A short history • Earlier searches before E391a were crippled by limited veto abilities. • Searches with better kinematics constraint via 0 Dalitz decay (1.2% br) but with degraded sensitivity. 3
The Basic Strategy • Pencil beam • Hermetic veto • Reconstruction of 0 vertex and P t assuming 0 mass 4
Pilot E391 at KEK CsI Calorimeter Front Barrel with CC02 Main barrel Charged Veto • Decay region in high vacuum(10 -7 Torr) 5
MC signal & neutron background after Reconstruction BG Estimation CC02 ‘pi0’ 0.6+/-0.4 CV-eta 0.2+/-0.1 CV-pi0 <0.3 Sensitivity of 1.1x10 -8 6
E391a neutron related background BG Estimation Neutron interactions with detector close to the beam were the main background CC02 ‘pi0’ 0.6+/-0.4 sources. CV-eta 0.2+/-0.1 CV-pi0 <0.3 7
8 Pilot E391a Result • Data plot of blind analysis
Pilot E391 result • Data plot of blind analysis, open the box • BR( ) < 2.6x10 -8 (90% CL) 9
How do we improve: • More K L ,less halo neutron Beam (JPARC, new beamline) • Background and acceptance Calorimeter and electronics (1% acceptance to ~ >5%) • Background New veto detectors 10
Japan-Proton Accelerator Research Center (KEK/JAEA ） Lina c South to North 3 GeV Synchrotron Neutrino Beams (to Kamioka) Materials and Life Experimental 50 GeV Facility Synchrotron CY2007 Hadron Exp. Beams JFY2008 Beams Facility JFY2009 Bird’s eye photo in January of 2008 Beams
Hadron Experimental Hall
The Beam 16 o To KOTO Ni To beam dump J-PARC E14 KOTO KEK-E391a • Primary proton energy 30 GeV 12 GeV • Proton intensity(/spill) 2x10 14 2.5x10 12 • Spill-length/repetition 0.7s / 3.3s 2s / 4s • Extraction angle 16 deg. 4 deg. • K L yield(/spill) 8.1x10 6 3.3x10 5 • Average P KL 2.1 GeV/c 2.6 GeV/c • n/K L ratio 6.5 45 13
14 The Beamline Design
15 Target, collimators and sweeping magnets
Beam Survey K. Shiomi et. al. Nucl. Inst. Meth. A664, 264 (2012): KL yield consistent with FLUKA; 2.4* GEANT4 QGSP-BERT-CHIPS (proposal) 3 Snowmass year -> <1.5 16
Comparison of halo neutron momentum E391a KOTO Neutron Momentum MC study of particle production The halo neutron background is expected to be 0.2 for KOTO. 17
To Further Reduce Neutron Background E391a KOTO Move CC02 upstream and make it fully active This configuration of CC02 and CV and the new beamline reduce the halo neutron background by a factor of >200. 18
19 Detector Upgrades
The CsI Calorimeter E391a KOTO E391 CsI crystal 7cmx7cmx30cm 16X 0 20 KTev crystals
The effect of longer CsI crystals E391a Crystal Reconstructed Energy in CsI as vertex of MC 0 a fraction of sample with ncident photon KTeV Crystal 0 generated with energy fixed position. Main mechanism for K L ->2pi0 background becomes ‘even- paring’ events with two photon missing in veto detectors 21
KTeV Calorimeter KOTO Calorimeter(Chicago) • New in KOTO: high rate, different energy range, vacuum, higher timing requirement • CW base ->low power • Differential outputs to the 125MHz waveform digitizer 14bit 125MHz pulse shaping FADC 22
The Gaussian Filter in the FADC CsI pulse(inverted). Yellow: filtered Filter circuit 23
Timing Measurement with an ADC • Full pulse sampling allows fitting to determine the time of the pulse • Timing measurement is important for veto and reducing acceptance loss. A good timing measurement also have the potential to find the angle of the photon. The angle measurement suppresses a large class of background. Smaller pulses have irregular shapes ns 24 Timing resolution ~110ps at E=100MeV
Readout and Architecture Optical fiber readout Data Flow FADC • Two level trigger system • Fully pipelined system with no delay cable • With 48 samples for a pulse, the FADC board can send out data at a trigger rate of >100kHz. Data throughput: 20GByte/second 25
Improvements on the Charged Veto (CV) • 0.8mm CFRP (carbon fiber reinforced) mechanical support; • Scintillator fiber MPPC both ends readout; • Prototype measured light yield 8-10 p.e./100keV deposited; inefficiency is expected to be <1x10 -3 ; • June 2012 verification with beam
• 1 p.e. • MPPC output/1p.e. 50ns pulse 7.5x10 5 gain • Pulse height : 0.24mV/1p.e. • Amp out: 12mV/1p.e. • FADC: 10mV/1p.e. ~ 80 counts • MIP peak •25-50 p.e./single end (FADC:~2000-4000 counts) •50-80 p.e./sum both end •2.3nsec timing resolution achieved Data: Simulation:
Catching photons down the hole • Challenge: ~10 8 neutrons/0.7 second, and more low energy beam photons(<10MeV) Lead/Aerogel Cerenkov counter Readout fast pulses: 500MHz FADC board 25 modules in total 28
Expected Performance • Energetic photons go down the hole 29
KOTO Main Barrel(MB) • With the improvement in the beam and Calorimeter, the main ->2 background comes from K L with missing photons. • Adding another 5X 0 to the 14 X 0 E391 MB reduces 2 background by a factor of 2. Photon Detection Inefficiency E391 MB KOTO MB With angle convoluted 30
How can we go further? • MB Inefficiency Mechanism MC Study Low energy sampling effect dominate For perpendicularly incident photons, 19X 0 isn’t enough for punch though. Not many such photons in MB, but many such photons in CsI. This explains why E391a K->2pi0 background comes from ‘punch through’. 31
Background and Sensitivity KEK-E391a improvement J-PARC KOTO KL yield/spill x30/sec 8.1x10 6 3.3x10 5 Run time 12 months 2 months x6 Decay prob. 3.6% 2.1% x2 Acceptance 4.7% 1% x3.6 Sensitivity x1300 0.8x10 -11 1.1x10 -8 32
Feb 2012 KL π + π - KL 3 π 0 Event display of a KL 3 π 0candidate
Summary • KOTO(step 1) is designed to make the first observation of • Construction of other detectors is on-going and will be completed in Fall 2012. • Physics run will start in Fall 2012 with ~30kW for few (?) months and will reach the Grossman-Nir limit. • Start preparing beamline for Stage-II. 34
Basic line : 5degee & 52m from target
36 Backup
39 Beam Survey With 1-3kW beam
Beam Survey in 2009 Halo suppression meets expectation K L yield measurement K L yield: 1.94*10 7 /spill ! 3 snowmass year <1.5 consistent with FLUKA; 2.5*GEANT4 QGSP-BERT-CHIPS 40 K. Shiomi et al, Nucl. Instr. and Meth. A664, 264 (2012)
Fusion • To reject fusion events, E391a suffered ~60% acceptance loss. In KOTO, the fusion background is reduced with a small acceptance loss. KTeV CsI photon 2 photons 5cm distance between two photon to identify fusion 41 using KOTO calorimeter; 15cm for E391a
Recommend
More recommend
