sub-gev DM Searches at FNAL

Transcription

1 V p π 0 γ N sub-gev DM Searches at FNAL R. T. Thornton MiniBooNE-DM Collaboration LDMA17 - Isola d Elba

2 Fermi National Accelerator Laboratory Booster Neutrino Beamline (8 GeV) SeaQuest (10 GeV) NuMI (10 GeV) R. T. Thornton LDMA17 - Isola d Elba

3 tion of mesons the p-be process, and transport of Booster Neutrinoresulting Beamline particles through the horn and decay volume. A full description of the calculation with associated uncertainties is provided in Ref. [19]. MiniBooNE neutrino data Beam: 8 GeV protons Target: Be Distance to MiniBooNE: 541 m Stable and well-understood beam FIG. 1 (color online). Schematic overview of the MiniBooNE experiment including the booster neutrino beamline and MiniBooNE detector. ν-mode ν-mode Th lati me rel in sam þ þ fit the res HA int dat dic eff inc R. T. Thornton LDMA17 - Isola d Elba 3

4 Beam Dump (Off-Target) Mode Can reduce νs by steering beam past target Target is air cooled can run beam so target is not hit p Target Be Decay Pipe Air Beam Dump MiniBooNE Detector V π 0 γ Steel Earth 50 m 4 m 487 m N Charged mesons are absorbed in the steel beam dump before decaying reduces the measured ν flux R. T. Thornton LDMA17 - Isola d Elba 4

5 MiniBooNE Detector Signal Region Veto Region Neutral-Current Nucleon (NCE) 800 tons of mineral oil (CH ) Cherenkov tracking detector with scintillator component 180 inner and 40 veto PMTs. Ran for a decade in ν/ ν Modes and has obtained/published 7 papers The detector is well understood This analysis focuses on NCE interactions R. T. Thornton LDMA17 - Isola d Elba 5

6 Neutrino Interaction Channels Charge-Current Quasi-Elastic (CCQE) 1 Neutral-Current Elastic (NCE) Cherenkov 1 e µ 1 Cherenkov 1 C C νµ-beam νµ-beam n p (Scintillation) p(n) Scintillation -39 ) /GeV (cm dσ/dq MiniBooNE NCE cross-section with total error Monte Carlo NCE-like background First ν double-differential CCQE cross section measurement Q (GeV ) QE Produced absolute and relative cross sections 1 A. A. Aguilar-Arevalo et al., Phys. Rev. D81, (0), arxiv:0.680 [hep-ex] A. A. Aguilar-Arevalo et al., Phys. Rev. D8, (0), arxiv: [hep-ex] R. T. Thornton LDMA17 - Isola d Elba 6

7 Off-Target Run Stability ν/pot ν/pot = (.1 ± 0.1) /ndf = /86 ν-mode ν-mode Off-Target 6.7e+0 ν POT 1.13e+1 ν POT 1.56e+0 bot POT -17 ν/pot = (0.79 ± 0.05) /ndf = /78-17 ν/pot = (.00 ±0.04) /ndf = 1.81/ /Jan/04 31/Dec/04 31/Dec/05 31/Dec/06 01/Jan/08 31/Dec/08 31/Dec/09 31/Dec/ 01/Jan/1 31/Dec/1 31/Dec/13 Beam dump run Ran 9 months, Nov 013 to begining of Sept. 014, collected POT CCQE ν event /POT decreased by 50 compared to ν-mode R. T. Thornton LDMA17 - Isola d Elba 7

8 Off-Target Neutrino Flux Events/(1e0 POT) CCQE Off Data (stat errors) (sys errors) ν det /GeV) ) (ν/pot/cm ν Φ Off (E ν µ ν e ν µ ν e CCQE Off /CCQE ν ) ν )/Φ ν (E ν Φ Off (E (GeV Q QE ) E ν (GeV) Event rate reduced by 50 Flux rate reduced by 30 R. T. Thornton LDMA17 - Isola d Elba 8

9 Vector Portal Kinetic Mixing Theory Four Free Parameters m mass of dark matter () m V mass of dark vector mediator ( dark photon ) (V ) ɛ kinetic mixing angle between V and SM photon α D = g /4π coupling constant D between V and Annihilation Cross Section SM V ɛ g D DS ( ) 4 σv α D ɛ α m Y ɛ m α mv 4 D m V σv Y α m Y Dimensionless parameter controlling cross section C. Boehm and P. Fayet, Nucl.Phys. B683, 19 (004), arxiv:hep-ph/ [hep-ph], C. Boehm et al., Phys.Rev.Lett. 9, 1301 (004), arxiv:astro-ph/ [astro-ph] B. Batell et al., Phys.Rev.Lett. 113, (014), arxiv: [hep-ph] R. T. Thornton LDMA17 - Isola d Elba 9

10 Light Dark Matter: Production and Detection Production (O ( ɛ g D ) ) Detection (O ( ɛ g D ) ) Neutral-Meson Decay Elastic Bound Nucleon γ π 0, η V n, p 1 C X V Proton Bremsstrahlung V p γ p p p Elastic Free Nucleon V p p R. T. Thornton LDMA17 - Isola d Elba

11 N-DM event selection Single p/n Track 1 track (single recoil) in beam timing window Event is centralized contained No activity in the veto Within tank fiducial volume Signal above visible energy and number of hits threshold PID: nucleon or electron Based on the ν-nce cross section analysis 1 1 A. A. Aguilar-Arevalo et al., Phys. Rev. D91, (015), arxiv: [hep-ex] R. T. Thornton LDMA17 - Isola d Elba 11

12 Fit Strategy Do a combined fit as a function of Q QE (reconstructed four momentum squared) to reduce systematic effects on NCE Off Include nuisance parameters to better fit NCE ν Mode data Events/(1e0 POT) Events/(1e0 POT) CCQE ν Data (stat errors) Total Bkg (sys errors) ν det ν dirt Q (GeV ) QE NCE ν Data (stat errors) Total Bkg (sys errors) Beam unrel. bkg ν det ν dirt Events/(1e0 POT) Events/(1e0 POT) Q (GeV ) QE Q QE (GeV ) Q QE (GeV ) R. T. Thornton LDMA17 - Isola d Elba CCQE Off Data (stat errors) Total Bkg (sys errors) ν det ν dirt NCE Off Data (stat errors) Total Bkg (sys errors) Beam unrel. bkg ν det ν dirt

13 Null Fit Results NCE Off-Target Events Ratio to Fit Results 500 Data (stat error) Predicted total bkg. 400 Beam unrelated bkg. ν-related detector bkg. ν-related dirt bkg. 300 Fit results, no DM Fit results + DM1 00 Fit results + DM reco T (GeV) #events uncertainty Beam unrel. bkg 697 Beam rel: ν det bkg 775 Beam rel: ν dirt bkg 7 Total Bkg % (pred. sys.) Data % (stat.) Fit Results % (fit effective error) Data consistent with background only Systematics dominated Constraint samples reduce systematics to 13% Detailed simulation to predict DM interactions in detector (arxiv: ) DM1 DM m V = MeV m = 1 MeV ɛ 4 α D = m V = 769 MeV m = 381 MeV ɛ 4 α D = R. T. Thornton LDMA17 - Isola d Elba 13

14 ε 4 90% Confidence Limits Results(arXiv: ) /m V) K π +invis. J/ψ invis. Y = ε αd (m 8 α µ favored NA64 BaBar017 BaBar 9 XENON - -e E137 Direct Detection Nucleon LSND m V = 3m, α D = 0.5 MB 90% CL MB 90% Sensitivity 11 Relic Density ± ± 1 σ σ 1 1 m (GeV) BaBar α µ favored NA64 LSND + + K π +invis. Only considered on-shell decay (m V > m ) CL in ɛ 4 α D for a given m V, m Slice space to compare to other experiments 3 4 XENON - -e Relic Density E137 BaBar017 m = MeV, α D = 0.1 MB 90% CL MB 90% Sensitivity ± 1 σ ± σ 1 1 m V (GeV) R. T. Thornton LDMA17 - Isola d Elba 14

15 Confidence Limit Results (arxiv: ) 4 /m V) Y = ε αd (m 7 8 α µ + + K π +invis. favored NA64 BaBar017 J/ψ invis. BaBar ε α µ BaBar favored NA64 LSND + + K π +invis. 9 XENON - -e E137 Direct Detection Nucleon 3 E137 BaBar LSND Relic Density m V = 3m, α D = 0.5 MB 90% CL MB 90% Sensitivity ± ± 1 σ σ 1 1 (GeV) m 4 XENON - -e Relic Density m = MeV, α D = 0.1 MB 90% CL MB 90% Sensitivity ± 1 σ ± σ 1 1 (GeV) m V Relic density line is where this model satisfies the thermal relic Dip in confidence limit is from increase production around mρ Exclude where this model solves muon g- discrepancy in most parameter space Exclude where this model matches relic density in some parameter space Exclude new parameter space regions Cover most of the gap between 1 MeV < m < direct detection R. T. Thornton LDMA17 - Isola d Elba 15

16 Future: MiniBooNE N-DM Analysis: Model Independent Fit Results in efficiency corrected excess nuclear recoil events as a function of true nuclear recoil energy This will be in a data release with covariance matrix Integrated = 130 ± 70 events/1e0 POT (last bin goes out to GeV) Used by theorist to test other DM models Leptophopic Model α B Neutron Scattering 0 π γ +invis. + K + π +invis. Monojet (CDF) Preliminary m V = 3m, ε = 0 MB 90% CL 1 1 (GeV) m J/ψ invis. Direct Detection Nucleon R. T. Thornton LDMA17 - Isola d Elba 16

17 Future: MiniBooNE Electron Scattering -6 e e mv=3m α'=0.5 POT=x 0 π 0 Production -6 N Nπ 0 mv=3m α'=0.5 POT=x 0 Y=ϵ α'(m/mv) MiniBooNE LSND E137 BaBar K + π + +invisible Excluded Electron/Muon g- >1 Event J/ψ invisible > Events Relic Density > 3 Events Direct Detection Preliminary Y=ϵ α'(m/mv) MiniBooNE LSND E137 BaBar K + π + +invisible Excluded Electron/Muon g- >1 Event J/ψ invisible > Events Relic Density > 3 Events Direct Detection Preliminary m(gev) m(gev) e (p) V e (p) 1 C V X π 0 N Future MiniBooNE analysis is promising Electron and π 0 sensitivity plots provided by Patrick deniverville R. T. Thornton LDMA17 - Isola d Elba 17

18 Future: Possible New Beam Configuration Target Decay Pipe Beam Dump MiniBooNE Detector p Be Air V π 0 γ Steel Earth 50 m 4 m 487 m N Target Decay Pipe Beam Dump MiniBooNE Detector p V π 0 γ Air Steel Earth N R. T. Thornton LDMA17 - Isola d Elba 18

19 Future: Possible New Beam Configuration p Target Decay Pipe Beam Dump MiniBooNE Detector V π 0 γ Air Steel Earth N Initial look shows neutrino event rate suppressed by 0 compared to Off-Target running NCE Events/1e0 POT 1 Off-target BUB Off-Target ν det PreliminarySteel-Target νdet Q (GeV 1. ) QE Prediction at MiniBooNE for illustration R. T. Thornton LDMA17 - Isola d Elba 19

20 Future: Short Baseline Near Detector Predicted sensitivities for SBND electron and π 0 channels with 6 0 POT and steel target Plot from Patrick deniverville R. T. Thornton LDMA17 - Isola d Elba 0

21 MiniBooNE-DM Collaboration A.A. Aguilar-Arevalo 1 M. Backfish 3 A. Bashyal 15 B. Batell B.C. Brown 3 R. Carr 4 A. Chatterjee 15 R. Cooper 5,1 P. deniverville 6 R. Dharmapalan 7 Z. Djurcic R. Ford 3 F.G. Garcia 3 G. T. Garvey 8 J. Grange 9, J.A. Green 8 W. Huelsnitz 8 I. L. de Icaza Astiz 1 G. Karagiorgi 4 T. Katori 11 T. Kobilarcik 3 W. Ketchum 8 Q. Liu 8 W.C. Louis 8 W. Marsh 3 C.D. Moore 3 G.B. Mills 8 J. Mirabal 8 P. Nienaber 13 Z. Pavlovic 8 D. Perevalov 3 H. Ray 9 B.P. Roe 14 M.H. Shaevitz 4 S. Shahsavarani 15 I. Stancu 7 R. Tayloe 5 C. Taylor 8 R.T. Thornton 5 R. Van de Water 8 W. Wester 3 D.H. White 8 J. Yu 15 1 Instituto de Ciencias Nucleares, Universidad Nacional Autónoma de México, D.F. México University of Pittsburgh, Pittsburgh, PA 1560, USA 3 Fermi National Accelerator Laboratory, Batavia, IL Columbia University; New York, NY 07 5 Indiana University, Bloomington, IN University of Victoria, Victoria, BC, V8P 5C 7 University of Alabama, Tuscaloosa, AL Los Alamos National Laboratory, Los Alamos, NM University of Florida, Gainesville, FL 3611 Argonne National Laboratory, Argonne, IL Queen Mary University of London, London, E1 4NS, UK 1 New Mexico State University, Las Cruces, NM Saint Mary s University of Minnesota, Winona, MN University of Michigan, Ann Arbor, MI University of Texas (Arlington), Arlington, TX R. T. Thornton LDMA17 - Isola d Elba 1

22 Extra Slides

23 Lessons Learned Data (stat errors) Bkg Prediction (sys. errors) Bkg Only Fit Result Understand Backgrounds PRD 8, (0) Beam related backgrounds: decade of data used to reduce systematics Dirt events small because MiniBooNE is huge with veto region Beam unrelated backgrounds: Measured with 1 ()Hz trigger for off-target (ν Mode) QE PRD 81, (0) PRL 0, (008) PRL 3, (009) PRD 83, (011) Include Nuclear Physics in DM Prediction Stripping a nucleus of a proton involves complex nuclear physics: e.g. binding energy, Pauli blocking, etc Honest sensitivity estimate must include a decent nuclear model threshold effects Include sideband/other channels Introduces correlations between the samples Results: Smaller systematic correlations on signal sample Results: Increases sensitivity Events/(1e0 POT) Rel. Diff. To Data CCQE ν NCE ν CCQE Off Q QE (GeV^) NCE Off NUANCE Events/1e0 POT Total From NUANCE Free Bound Extrapolated Total true T p (GeV) R. T. Thornton LDMA17 - Isola d Elba

24 Reducing Beam Unrelated Background Limited Sample Off-Target-NCE (NCE Off ) Events/(35.8 MeV) Data Total Bkg Beam Unrel. Bkg Beam Rel. Bkg: ν det Beam Rel. Bkg: ν dirt Initial look showed Beam unrelated bkg. = 69% of data Using previous trigger info. to reject events that had hit info. before the beam trigger Nucleon Recoil Energy (MeV) Reduced Beam unrelated bkg. to 58% of data with BRB efficiency of 95% Off-target Beam unrelated bkg. measured with random trigger at Hz R. T. Thornton LDMA17 - Isola d Elba 3

25 Extension to sub-gev Dark Matter Add baryonic current 1 L = L 1 4 F µν D F µν + 1 D m V V µ V µ ɛ F µνf µν + q D B g D V µ J µ B + Jµ B = 1 q i γ µ q i sum over all quark species 3 i When ɛ = 0 g D g B (α B = gb /4π) L is solely dependent on baryonic current This scenario is called leptophobic MiniBooNE NCE analysis is sensitive to this model (proton production and proton/neutron detection) 1 B. Batell et al., Phys. Rev. D90, (014), arxiv: [hep-ph], P. deniverville et al., (016), arxiv: [hep-ph] R. T. Thornton LDMA17 - Isola d Elba 4

26 Model Independent Fit Added fit parameters, one for each TN true bin Efficiency corrected predicted # events in true nuclear recoil energy Integrated = 130 ± 70 events/1e0 POT This will be in a data release with covariance matrix Events/(1e0 POT)/GeV Rel. Error true T True radius is 6.6 cm N Last bin is [0.55, ) GeV (GeV) R. T. Thornton LDMA17 - Isola d Elba 5

27 Leptophobic Model α B Monojet (CDF) Used Model Independent Fit to produce 90% CL 1 LSND, E137, etc. do not constrain due to lepton final state interaction MiniBooNE covers a lot of new parameter space Neutron Scattering 0 π γ +invis. + K + π +invis. m V = 3m, ε = 0 MB 90% CL 1 1 (GeV) m J/ψ invis. Direct Detection Nucleon 1 Thanks to P. deniverville for doing the analysis R. T. Thornton LDMA17 - Isola d Elba 6

28 Detector and Cut Effects on DM Pred. True Reco DM Generator Events/1e0 POT ε = 0.001, α' = 0.1 m V = MeV m = 1 MeV m V = 767 MeV m = 381 MeV Preliminary Events/1e0 POT ε = 0.001, α' = 0.1 m V = MeV m = 1 MeV m V = 767 MeV m = 381 MeV Preliminary true T N (GeV) Q (GeV ) QE Detector and Cut efficiencies reduces the number of low energy DM scatters MiniBooNE sensitive to TN true = O (0 MeV ) after efficiencies applied Expect to see higher sensitivity around m ρ than below m π 0 R. T. Thornton LDMA17 - Isola d Elba 7

29 Signal Best Fit Fit Statistic (GeV) m m V (GeV) Fit Test Statastic Percent Diff. To Best Fit ɛ 4 α D (GeV) m m V (GeV) Negative of Fit ε 4 α' Little change in test statistic when changing V, Huge change in ɛ 4 α D R. T. Thornton LDMA17 - Isola d Elba 8

30 Next Analyses in this Data Set Off Limited NCE Beam with stat. errros Total Bkg. Pred. with sys. errors Beam unrel. bkg. ν det ν dirt /NDF = 7.8/9 RF spill-event timing Massive DM will be delayed relative to ν backgrounds Predictable timing spectrum vs. dirt which is flat in time Timing is even applicable to ν-oscillation data to separate e-γ DM-Electron Scattering ν-e scattering as background Expect better sensitivity than NDM Limited NCE Off RF Timing Spectrum ) 0 Events/(POT Preliminary Bunch Time (ns) N + X, π 0 NCπ 0 background π 0 is a clean detection signal BUB small and expect better sensitivity than NDM R. T. Thornton LDMA17 - Isola d Elba 9

31 Future sub-gev DM Searches What would more data do? NDM is systematics limited more data will further optimization of analysis π 0 DM and edm are currently statistics limited and more data is crucial Recent BNB Upgrade: Dual beam on-/off-target running Fast trim magnet to switch proton trajectory is installed BNB horn operates at 5 Hz but 7 Hz can be delivered to BNB when NuMI is down for maintenance or repair Gates around MI are installed; more opportunities for BNB beam Short Baseline Neutrino Program Entire program can search for low-mass DM Short Baseline Near Detector DM signal is 0 the MiniBooNE rate R. T. Thornton LDMA17 - Isola d Elba

32 References I [1] A. A. Aguilar-Arevalo et al., Phys. Rev. D81, (0), arxiv:0.680 [hep-ex]. [] A. A. Aguilar-Arevalo et al., Phys. Rev. D8, (0), arxiv: [hep-ex]. [3] C. Boehm and P. Fayet, Nucl.Phys. B683, 19 (004), arxiv:hep-ph/ [hep-ph]. [4] C. Boehm, D. Hooper, J. Silk, M. Casse, and J. Paul, Phys.Rev.Lett. 9, 1301 (004), arxiv:astro-ph/ [astro-ph]. [5] B. Batell, R. Essig, and Z. Surujon, Phys.Rev.Lett. 113, (014), arxiv: [hep-ph]. [6] A. A. Aguilar-Arevalo et al., Phys. Rev. D91, (015), arxiv: [hep-ex]. [7] B. Batell, P. deniverville, D. McKeen, M. Pospelov, and A. Ritz, Phys. Rev. D90, (014), arxiv: [hep-ph]. [8] P. deniverville, C.-Y. Chen, M. Pospelov, and A. Ritz, (016), arxiv: [hep-ph]. R. T. Thornton LDMA17 - Isola d Elba 11