Status and objectives Physics Beyond Colliders CERN, 02 March 2017 Matthew Moulson INFN Frascati For the NA62-KLEVER project
K πνν: an overview Extremely rare decays with rates very precisely predicted in SM: SM predicted rates* Experimental status K + π + νν BR = (8.4 ± 1.0) 10 11 7 evts from BNL787 BR to 10% from NA62 by 2018 K L π 0 νν BR = (3.4 ± 0.6) 10 11 Only limits at present KOTO (JPARC): ~few SM events by 2021 New physics affects K + and K L differently Measurements of both can discriminate among NP scenarios sensitivity 5 years starting Run 4 60 SM K L π 0 νν events S/B ~ 1 BR(K L π 0 νν) ~ 20% BR(K L π 0 νν) 10 11 Min. flavor viol. Z/Z, LHT Randall-Sundrum Buras, Buttazzo, Knegjens JHEP 1511 Buras et al, JHEP 1511* BR(K + π + νν) 10 11 2
3 m An experiment to measure K L π 0 νν AFC LAV 1-12 LAV 13-17 LAV 18-21 LAV 22-26 LKr 80 m IRC SAC FV CPV 105 m 155 m 241.5 m For 60 SM events, need: 5 10 19 pot E.g. 2 10 13 ppp/16.8 s 5 yrs p K = 70 GeV for decays in FV Photons from K L π 0 π 0 boosted forward for easier vetoing Much higher energy than KOTO: Complementary approach Main detector/veto systems: AFC Active final collimator/upstream veto LAV1-26 Large-angle vetoes (26 stations) LKr NA48 liquid-krypton calorimeter IRC/SAC Small-angle vetoes CPV Charged-particle veto 3
Summary: PBC-phase project goals Simulation Beam & beamline Small-angle calorimetry Technical development of new MC/framework Study backgrounds from broader variety of sources Evaluate feasibility of high-intensity P42/K12 beam Develop reliable simulation of neutral beamline Beam test of crystal pair enhancement Validate concept for W/Si SAC or choose alternative technology LKr calorimeter Study LKr performance with NA62 data and MC Evaluate long-term reliability of LKr (2018 2030) Readout upgrades to improve time resolution Large-angle vetoes Upstream veto Charged particle veto Additional detectors Develop & simulate detector design in more detail Develop & simulate detector design in more detail Study MUV1/2 response to µ/π in NA62 data and MC Examine feasibility of adding tracking to experiment 4
Work in progress: simulation Results presented so far have been based on fast simulations implementing mainly geometry + parameterized response Next step is to develop full MC with realistic detector simulation: NA62 analysis framework (MC + reconstruction + analyzers) Make use of existing simulation for LKr, HAC (MUV1/2) Optional use of parameterized response to continue fast simulation work 5
Work in progress: simulation Add new detectors to MC: In parallel with refinement and optimization of detector design Active final collimator/upstream veto Large-angle vetoes Charged particle vetoes Specific issues: Simulation of neutral beamline Development of small-angle calorimeter with crystalline absorber Complete study backgrounds from a broader variety of sources: Validation of results obtained with fast simulations for π 0 π 0, π 0 π 0 π 0 Perform general survey of backgrounds Decays to charged particles, including esp. radiative channels and K l4 Add new generators to MC Short-lived components: Λ nπ 0, K S π 0 π 0 6
Primary beam requirements 10 19 pot/yr 5 years 2 10 13 ppp/16.8s 6 increase in intensity relative to NA62 Advantages of siting in K12/ECN3: Long beam cavern and experimental hall needed Experimental infrastructure & NA48 LKr calorimeter already in place Can intensity requirement be met in P42/K12? Need to evaluate feasibility and estimate cost of upgrades to target area and transfer lines Minimization of consequences of beam loss Additional shielding against continuous losses, e.g., on splitters Study issues of equipment survival, e.g., TAX motors Ventilation, zone segmentation, etc. Feasibility study is a primary goal of our involvement in the PBC Conventional Beam WG 7
Neutral beam simulation FLUKA simulation of 400 GeV p on 400-mm Be target Geant4 simulation of beamline 3 collimators, Δθ = 0.3 mrad 30-mm Ir photon converter in dump collimator K L in beam: 280 MHz 35% scattered by converter Photons in beam E > 5 GeV: 230 MHz E > 30 GeV: 20 MHz Neutrons in beam E > 1 GeV: 3 GHz Effect of γ converter 8
Neutral beam simulation Currently working on a series of improvements: Refine and improve current studies with FLUKA & Geant4 Optimize beamline parameters, e.g., photon converter thickness 35% of K L are absorbed or scattered by converter Make converter thinner by using metallic crystal? Understand neutral beam halo from photons and neutrons Impacts on rates and background Simulate beam-gas interaction (e.g. nn nnπ 0 ) Evaluate background, level of vacuum necessary in decay volume Complete muon halo studies Optimize beamline elements to reduce 9
Small-angle calorimeter 90% of γs from K L π 0 π 0 on SAC have 30 < E γ < 250 GeV 3133 evts total E γ < 30 GeV 313 evts K L π 0 π 0 photons on SAC Exactly 2 γs on LKr No γs on LAV or IRC Cuts on z FV, r min (LKr), p Required inefficiency 1 ε E γ [GeV] Rate (MHz) 1 ε γ, E > 5 GeV 230 10 2 γ, E > 30 GeV 20 10 4 n 3000 Reject high-energy γs from K L π 0 π 0 escaping through beam hole Must be insensitive as possible to 3 GHz of beam neutrons Baseline solution: Tungsten/silicon-pad sampling calorimeter with crystal metal absorber 10
Efficient γ conversion with crystals Coherent effects in crystals enhance pair-conversion probability Rel. pair production Courtesy of V.V.Tikhomirov Tungsten (W) E γ = 200 GeV for different φ inc and temp Rel. pair production E γ = 10 GeV E γ = 100 GeV E γ = 5 GeV Tungsten (W) θ inc [μrad] θ inc [μrad] Use coherent effects to obtain a converter with large effective λ int /X 0 : 1. Beam photon converter in dump collimator Effective at converting beam γs while relatively transparent to K L 2. Absorber material for small-angle calorimeter (SAC) Must be insensitive as possible to 3 GHz of beam neutrons while efficiently vetoing high-energy γs from K L decays 11
Test program for crystal studies Using a high-energy tagged photon beam: 1. Observe pair conversion enhancement in a practical experimental setup, with a real tungsten crystal 2. Measure pair conversion vs. E γ, θ inc for 5 < E γ < 200 GeV Confirm theoretical predictions for tungsten 3. Obtain information for comparison with MC, to assist development of beam photon converter and SAC 12
KLEVER Example of test beam setup in H4 AXIAL Study of axial phenomena in crystals for beam steering and generation of intense electromagnetic radiation CERN INFN FE-LNL-MIB-TIFPA Experimental setup: Scheduled July 2017 SD1 SD2 Similar to other experiments in H4 (e.g., NA63) SD3 Calorimeter for emitted radiation γ CAL GC Double-sided silicon detectors 300 μm thick 5 μm spatial resolution e CAL Calorimeter for deflected beam Crystal mounted on high-precision goniometer 13
Pair conversion test with AXIAL AXIAL group interested in collaboration with KLEVER on test beam measurement of pair production enhancement in crystals Modifications for tagged photon beam setup: 1. Replace crystal with bremsstrahlung target 2. Move crystal into photon beam GC SD1 SD2 SD3 γ CAL Exploring possibilities to reuse most of AXIAL instrumentation Need 1 week of beam time in 2018 3. Move SD3 to measure electron beam deflection e CAL 14
LKr calorimeter Study and confirm LKr performance with NA62 data Two-cluster resolution In parallel with Photon detection efficiency efforts by NA62 Effect of dead cells, etc. Explore possibilities to improve time resolution with faster readout Signal π 0 candidates all have E γγ > 20 GeV σ t = 2.5 ns/ E (GeV) 500 ps or better Needs improvement SAC may have ~100 MHz accidental rate Simulating readout upgrades to estimate effect on time resolution: Shorter shaping time, faster FADCs Evaluate long-term reliability of LKr (2018 2030): Identify support systems needing replacement or upgrade Catalog of dead cells, prospects for repair 15
Charged particle rejection & tracking Use simulation to add detail to design of charged particle vetoes Charged particle rejection with hadron calorimeters Study of HAC (MUV1/2) response in NA62 data Parameterization of HAC response for fast simulation Add a tracking system for charged particles? Expand physics scope of experiment Facilitate calibration and efficiency measurements Potential complications for K L π 0 νν Simulate impact of material budget on photon veto efficiency Evaluate impact of magnet on photon veto coverage Add a preshower detector in front of LKr? Place γ converter in front of tracking planes, measure impact angle Potentially adds redundancy for background elimination Require at least 1 conversion for signal events cost in signal? Same complications as for adding tracking 16
Summary and outlook Project timeline: 2017-2018 Project consolidation and proposal Beam test of crystal pair enhancement 2019-2021 Detector R&D 2021-2025 Detector construction Possible K12 beam test in cooperation with NA62++ 2024-2026 Installation during LS3 2026- Data taking beginning Run 4 Major goals in context of Physics Beyond Colliders: Develop full simulation of experiment and use to add detail to design of new detectors Establish feasibility of high-intensity P42/K12 beam in ECN3 Test pair conversion in crystals and validate concept for SAC Verify suitability of LKr calorimeter 17
Additional information Physics Beyond Colliders CERN, 02 March 2017 Matthew Moulson INFN Frascati For the NA62-KLEVER project
K πνν and the unitarity triangle Dominant uncertainties for SM BRs are from CKM matrix elements Buras et al., JHEP 1511 Intrinsic theory uncertainties ~ few percent Measuring both K + and K L BRs can determine the unitarity triangle independently from B inputs Overconstrain CKM matrix reveal NP? η Hypothetical CKM fit to K πνν 10% mmts for K + and K L ρ 19
K πνν and new physics New physics affects BRs differently for K + and K L channels Measurements of both can discriminate among NP scenarios BR(K L π 0 νν) 10 11 Buras, Buttazzo, Knegjens JHEP 1511 BR(K + π + νν) 10 11 Models with CKM-like flavor structure Models with MFV Models with new flavorviolating interactions in which either LH or RH couplings dominate Z/Z models with pure LH/RH couplings Littlest Higgs with T parity Models without above constraints Randall-Sundrum 20
K πνν and new physics General agreement of flavor observables with SM invocation of MFV Long before recent flavor results from LHC But NP may simply occur at a higher mass scale Null results from direct searches at LHC so far Indirect probes to explore high mass scales become very interesting! Tree-level flavor changing Z LH+RH couplings Some fine-tuning around constraint from ε K K πνν sensitive to mass scales up to 2000 TeV Up to tens of TeV even if LH couplings only Order of magnitude higher than for B decays K πνν is uniquely sensitive to high mass scales BR(K L π 0 νν) 10 11 Buras et al., JHEP 1411 m Z = 500 TeV BR(K + π + νν) 10 11 21
K πνν and other flavor observables Simplified Z, Z model used as paradigm Buras, Buttazzo, Knegjens, JHEP 1511 CMFV hypothesis: Constraints from B and K observables BR(K L π 0 νν) 10 11 20 15 10 5 BR(K L π 0 νν) 10 11 Modified Z 5 TeV Z 10 8 6 4 2 LH and RH couplings allowed: Constraints from K observables: ε K, ΔM K ε /ε, K µµ (for modfied Z) BR(K L π 0 νν) 10 11 0 0 0 5 10 15 20 25 30 0 2 4 6 8 10 12 14 BR(K + π + νν) 10 11 BR(K + π + νν) 10 11 10 8 6 4 2 Modified Z 0 0 5 10 15 20 25 30 0 0 5 10 15 20 25 BR(K + π + νν) 10 11 BR(K + π + νν) 10 11 22 BR(K L π 0 νν) 10 11 20 15 10 5 5 TeV Z
K πνν and large-scale SUSY SUSY with m stop ~ 10 TeV Tanimoto & Yamamoto, PTEP 2016 BR(K L π 0 νν) 10 11 10 8 6 4 2 BR(K L π 0 νν) 10 10 4 3 2 1 with constraint from ε K 0 0 1 2 3 4 ε /ε 10 3 0 0 1 2 3 4 BR(K + π + νν) 10 10 Possible to choose Zsd couplings such that gluino Z penguin simultaneously enhances ε /ε and BR(K L π 0 νν) BR(K L π 0 νν) up to 1 10 10 possible with experimental bounds on ε /ε satisfied 23
Required intensity for K L π 0 νν Assumptions: BR(K L π 0 νν) = 3.4 10 11 Acceptance for decays in FV ~ 10% Beam parameters: 400 GeV p on 400 mm Be target Production at 2.4 mrad to optimize (K L in FV)/n Fiducial volume acceptance ~ 2% 3 10 13 K L decay in FV for 100 signal evts 2.8 10 5 K L in beam/pot Norm. counts K L in beam p K = 97 GeV K L decays in FV p K = 68.5 GeV Abs. counts K L in beam K L decays in FV p(k L ) [GeV] p(k L ) [GeV] Required total proton flux = 5 10 19 pot 10 19 pot/year (= 100 eff. days) E.g.: 2 10 13 ppp/16.8 s 24
Neutral beamline layout Beam acceptance Δθ 0.3 mrad : ΔΩ 0.283 μsr Just fits inside LKr central vacuum tube (r = 80 mm) Collimators: 1. Dump collimator: TAX1/2 (r = 5 mm) moved forward to z = 15 m 2 vertical sweeping magnets upstream of TAX 3 horizontal muon sweeping magnets downstream of TAX 2. Defining collimator: r = 42 mm at z = 60 m Keep background from TAX/converter within LKr bore (r < 120 mm) 3. Final collimator: z = 105 m to remove upstream decay products Regenerated K S reduced to 10 4 between defining and final collimators Final collimator is also an active detector to veto upstream decays Photon converter between TAX1/2 Explore use of crystal converter to optimize K L transmission Pair production enhanced by coherent effects in crystals 25
Neutral beamline layout y [mm] 150 100 50 0 50 100 150 400 GeV proton beam 2.4 mrad Target T10, r = 1.0 mm Dump collimator TAX 1&2, r = 5 mm γ converter 2 MTRV sweeping magnets 3 MBPL muon sweepers r = 18 mm Defining collimator ± 0.5 mrad r = 42 mm Active final collimator Fiducial decay region Limit of background from TAX/γ absorber through defining collimator Long beamline and detector layout, like NA62 Ideally sited in NA62 hall z [m] 0 50 100 150 200 250 26 LKr Cold bore LKr beam pipe Cold bore LKr
3 m Detector layout for K L π 0 νν AFC LAV 1-12 LAV 13-17 LAV 18-21 LAV 22-26 LKr 80 m IRC SAC FV CPV 105 m 155 m 241.5 m Vacuum tank layout and FV similar to NA62 90-m distance from FV to LKr significantly helps background rejection Most K L π 0 π 0 decays with lost photons occur just upstream of the LKr π 0 s from mispaired γs are mainly reconstructed downstream of FV K L π 0 π 0 2γ on LKr only True decay z Reconstructed z FV 27
Upstream decays KLEVER Background source: KL π0π0 upstream of final collimator 80 m < z < 105 m Upstream veto: Make final collimator active, with acceptance for r < 1 m Outer ring (10 cm < r < 1 m): Shashlyk calorimeter, Pb/scint in 1:5 ratio 1/3 of total rate. Blind for E < 50 MeV, 1 ε = 10 5 for E > 1 GeV Inner ring (4.2 cm < r < 10 cm): LYSO collar counter, 80 cm deep, shaped crystals 2/3 of total rate. Blind for E < 100 MeV, 1 ε = 10 4 for E > 1 GeV Vacuum tank starts at z = 80 m Assume veto blind to photons upstream of vacuum tank Residual background from upstream decays: 15 events/5 years 28
Large-angle vetoes 26 new LAV detectors providing hermetic coverage out to 100 mrad Need good detection efficiency at low energy (1 ε ~ 0.5% at 20 MeV) Baseline technology: CKM VVS Scintillating tile with WLS readout Good efficiency assumptions based on E949 and CKM VVS experience 1 ε E949 barrel veto efficiencies Same construction as CKM Tests for NA62 at Frascati BTF 1 ε Tagged e Parameterization: E γ [MeV] 1-129 MeV: KOPIO (E949 barrel) 203-483 MeV: CKM VVS E γ [MeV] Tests at JLAB for CKM: 1 ε ~ 3 10 6 at 1200 MeV 29
Small-angle vetoes Proof-of-concept simulation for baseline solution: W-Si pad calorimeter, 14 layers 1 mm crystal absorber, θ inc = 2 mrad Depth = 14X 0 for E γ = 30 GeV, but only 4X 0 for E γ = 5 GeV Naïve simulation of pair-conversion enhancement with Geant4: Increase overall density as function of E γ, instead of X 0 Photons E γ (GeV) ρ/ρ 0 1 ε 350 GeV 3.5 5 10 5 30 GeV 3.5 1 10 4 10 GeV 1.5 4.5% 5 GeV 1.0 20% Neutrons 50-300 GeV 1 ε = 20% E vis thr. = 16 MeV chosen for E γ = 30 GeV Inefficiency at small E γ from punch through Need better treatment of coherent effects Need additional handles for γ/n separation Work in progress: Better simulation with X 0 for photons a function of E γ and θ γ Benefit from effort by UA9 collaborators to introduce into Geant4 detailed simulation of coherent effects in crystals Optimize transverse and longitudinal segmentation to increase γ/n separation 30
Charged-particle rejection K e3 most dangerous mode: e easy to mistake for γ in LKr Acceptance π 0 νν/k e3 = 30 Need 10 9 suppression! Item Charged particle veto planes: 5-mm thick scintillator strips Cluster RMS Or other LKr shower profile cuts HAC/LKr energy ratio Re-use MUV1/2 from NA62 Rej. factor 0.005/track 2 planes 0.05 0.07 All πe from K e3 reconstructed as π 0 have z true < 200 m z reco [m] 155 m 200 m z true [m] LKr cluster RMS [cm] 1.10-1.35 cm ε γ = 0.95 ε π = 0.05 < 0.05 ε γ = 0.99 ε π = 0.07 LKr cluster energy [GeV] HAC/(HAC + LKr) energy [GeV] 31
Future test with neutral hadron beam Main K12 modification for neutral beam: Remove achromat/raise SCR1 Possibilities for beam test in K12 with NA62 during Run 3? K12 is the only high-energy neutral beam available Use very low intensity avoid reconfiguring straws & RICH? Easy setup at zero angle, but still several weeks of work Make changes to beamline during winter shutdowns Neutral beam running compatible with beam-dump program? Explore possibilities with NA62 for a short test 32
K L π 0 π 0 rejection K L π 0 π 0 simulated with fast MC (5 yr equivalent statistics) Accept only events with 2 γs in LKr and no hits in AFC, LAV, IRC/SAC Distinguish between even/odd pairs and events with fused clusters 1. Require z rec (m γγ = m π0 ) in fiducial volume (105 m < z < 155 m) p (π 0 ) [GeV] Even Odd Fused upstream decays 2. Require r min > 35 cm on LKr and p (π 0 ) > 0.12 GeV E(γγ) [GeV] p (π 0 ) [GeV] Even Odd Fused 22 π 0 π 0 evts/year About 50% with 1γ with 100 < θ < 400 mrad, E < 50 MeV E(γγ) [GeV] 33
K L π 0 νν acceptance Cut stage Cut eff. Cuml. eff. K L π 0 νν evts with 2γ on LKr 2.0% 2.0% z rec (m γγ = m π0 ) in FV 31% 0.62% r min > 35 cm on LKr 42% 0.26% p (π 0 ) > 0.12 GeV 78% 0.20% z rec (m γγ = m π0 ) in FV r min and p (π 0 ) cuts Alternatively: 2.2% K L decay in FV 27% π 0 νν with 2γ on LKr π 0 in π 0 νν has large E kin V A matrix element With: 10 19 pot/year 2.8 10 5 K L /pot BR = 3.4 10 11 ε total = 0.20% 19.4 π 0 νν evts/year excluding transmission losses from γ converter 34
K L π 0 νν at J-PARC Primary beam: 30 GeV p p K = 2.1 GeV Run 62-2015 Current status: Reached 42 kw of slow-extracted beam power in 2015 Preliminary results: 10% of 2015 data SES = 5.9 10 9 Expected background = 0.17 events Background estimate still under study, signal box not yet unblinded Expect to reach SM sensitivity by 2021 Beam power will increase to 100 kw by 2018 Continuing upgrades to reduce background: New barrel veto (2016) Both-end readout for CsI crystals (2018) 35
KL 0 π νν at J-PARC KOTO Step-2 upgrade: Increase beam power to >100 kw (Originally 450 kw) New neutral beamline at 5 p(kl) = 5.2 GeV KOTO Increase FV from 2 m to 11 m Complete rebuild of detector Requires extension of hadron hall Long-term future: Strong intention to upgrade to O(100) event sensitivity No official Step 2 proposal yet (plan outlined in 2006 KOTO proposal) Scaling from 2006 estimates: ~10 SM evts/yr per 100 kw beam power Exploring machine & detector upgrade possibilities to increase sensitivity Indicative timescale: data taking starting 2025? 36