Solar Neutrino Results from Phase III of the Sudbury Neutrino Observatory

Transcription

1 Solar Neutrino Results from Phase III of the Sudbury Neutrino Observatory Alan Poon Berkeley Lab

2 Solar Neutrinos Bahcall et al.

3 Solar Neutrino Problem (~Y2K) Deficits were seen in all terrestrial solar ν detectors (which were sensitive primarily to ν e ). SSM Prediction Gallium Chlorine

4 Sudbury Neutrino Observatory (SNO) 2 km to surface 1 kt D 2 O 12.01m dia. acrylic vessel 1700 t inner shielding H 2 O 17.8m dia. PMT support structure cm PMTs; 56% coverage 5300 t outer shielding H 2 O Nucl. Inst. Meth. A449, 127 (2000) Image courtesy National Geographic

5 Detecting ν at SNO CC ν e + d p+ p+ - e Measurement of ν e energy spectrum Weak directionality: 1! 0.340cos" NC ν + d p+ n +ν x x Measure total 8 B ν flux from the sun σ(ν e ) = σ(ν µ ) = σ(ν τ ) e ES ν ν - x e x Low Statistics σ(ν e ) 6 σ(ν µ ) 6 σ(ν τ ) Strong directionality: " e # 18 o (T e = 10 MeV)

6 Smoking gun for flavor transformation Does the total flux of solar neutrinos equal the pure ν e flux? Measure: CC NC = Alternatively! e! e +! µ +! " CC ES =! e! e (! µ +! " ) Transformation to another active flavor if:! CC (" e ) <! NC (" x )! CC (" e ) <! ES (" x ) Flavor transformation can be demonstrated without any assumption on the Standard Solar Model prediction of the total neutrino flux.

7 SNO Phase I: Pure Heavy Water Pure D 2 O target - Ended May 2001 n + 2 H 3 H + γ(6.25 MeV) Low neutron detection efficiency (~14%) σ = 0.5 mb 2 H+n 6.25 MeV 3 H assumed Max. an undistorted 8 B Likelihood spectrum; Fit but neutrino oscillation can have energy dependence " CC " NC " ES OR a null hypothesis test " e " µ# large NC uncertainties when the energy constraint is removed

8 SNO Phase II: Salt Phase II (D 2 O + 2 tonnes NaCl) - Ended Sep n + 35 Cl 36 Cl + γʼs(σe γ =8.6 MeV) High neutron detection efficiency (~41%) σ = 44 b 35 Cl+n 8.6 MeV 36 Cl use of light isotropy removed assumption of 8 B shape in physics extraction NC: multiple e CC, ES: single isotropy total NC flux uncertainty ~8.4% Strong CC-NC anti-correlation (-0.52)

9 Phase III : 3 He counters Construction / com. D 2 O Salt D 2 O Com. 3 He counters n + 3 He t + p kev 3 He σ = 5330 b ε n = 21 % 4 He n NCD p t Different systematics Reduce CC-NC correlation Better CC flux measurement

10 SNO Detector: Current Status since 28th November, 2006

11 Getting the Last Drop of D 2 O The acrylic vessel was completely emptied at 14:45 (Sudbury time) on 28th May, 2007.

12 NCD String 360 µm Ni wall High purity CVD nickel End cap 3 He-CF 4 Cu anode wire Fused silica connector 9-11 m ~1/100 background of previously cleanest PC Delay line Ni + 4CO Ni(CO) 4 50 C at pressure Chemical Vapor Deposition (CVD) Raw Ni 175C Anchor line U, Th carbonyls Ni(CO) 4 Aluminium mandrel

13 NCD Deployment ROV NCD ROV ball NCD Anchor ball ROV AV Anchor block

14 NCD String 360 µm Ni wall 3 He-CF 4 Cu anode wire Fused silica connector Delay line Digital scopes Shaper-ADCs Multiplexer trigger on amplitude slow readout ( 8 B neutrino) trigger on integral charge fast readout ( 8 B, SN) Digital oscilloscope Shaper-ADC Current Preamp Anchor line NCD

15 NCD Signals 360 µm Ni wall 3 He-CF 4 Cu anode wire Current Neutron n p Fused silica connector Time (µs) t Delay line Current 20 Alpha α Anchor line Time (µs)

16 Shaper-ADC 360 µm Ni wall Neutrino data 3 He-CF 4 Cu anode wire Po alphas Fused silica connector Delay line n t 573 kev Neutron 764 kev Energy (MeV) 191 kev p p t Anchor line n n Energy (MeV)

17 Calibrations Source PMT NCD Laserball nm 16 N 6.13 MeV γ 8 Li e - spectrum Optics Energy Source-manipulator system capable of 2-5 cm positional accuracy AmBe n 252 Cf n 24 Na n Neutron eff. Ropes Th low E γ Rn low E γ Bkgrd. PDFs Source Distributed source

18 Optical Calibration Detector y (cm) 1 / (rel. PMT occupancy) PMTs NCD PMTs y Laserball runs x Detector x (cm)

19 Energy Calibration - PMT Salt phase NCD phase T eff (data / MC) T eff (data - MC) / Data) ρ = (r/600) 3 ρ = (r/600) 3 ΔE/E = 1.1% Position and energy resolutions are comparable to the salt phase

20 Neutron Calibration 24 Na method Monte Carlo method Mimic the signal with mixed 24 Na, which generates neutrons by γ + 2 H n + p Calibrate the Monte Carlo with point AmBe and 252 Cf sources ε n = ± ε n = ± Na mixing in the heavy water Typical source run locations ( )

21 24 Na Mixing 2005 injection points 2006 injection point 24 Na mixing during the 2006 spike

22 24 Na Mixing Neutrons Gammas After injection 1.4 Ratio of 2006 to Mixing Steady state x (cm)

23 Neutron Backgrounds n NCD Cable PD n n n Hot spot PD NCD Ni PD D 2 O PD Hot spots K5 K2 n (α, n) Acrylic vessel ( PD = photodisintegration ) z (cm) z (cm)

24 Instrumental Background Cuts Energy spectrum before and after cuts Time domain cuts Frequency domain cuts Burst cuts Raw shaper spectrum Shaper and scope trigger After instrumental cuts Current (µa) Time (µs) Energy (MeV) Current (µa) Time (µs)

25 Instrumental Backgrounds Hot spot Gas leak into counter inter-space Electrical disconnect Electrical micro-discharge Gain instability Electrical micro-discharge J3-like Neutron peak N4-like Energy (MeV) Low energy cut 3 He 4 He Energy (MeV)

26 Problems with Other Strings Hot spot Gas leak into counter inter-space Electrical disconnect Electrical micro-discharge Gain instabilities Electrical disconnect P NCD string number Resistive coupler 3 He Anode wire 4 He

27 Simulating an NCD Pulse Energy deposition, electron drift Charge multiplication, ion drift, pulse propagation, electronics Noise Proton Triton Time (ns) Pulse simulation : Time (ns) α energy loss, α straggling, α multiple scattering electron-ion pair generation electron drift, diffusion electron multiple scattering ion mobility electron avalanche space charge signal generation, electronics, noise Time (ns)

28 Comparisons with Data Data Monte Carlo Pulse width (ns) Pulse width (ns) Energy (MeV) Energy (MeV) Wire Po αs Wire U/Th αs Neutrons Wall Po αs Wall U/Th αs Endcap Po αs

29 Alpha Pulse Simulation Width Kurtosis Energy Blind data Total alphas Po alphas U alphas Th alphas FWHM Kurtosis Energy (MeV) Relative contributions of U, Th and Po alphas fit using data above the neutron (signal) energy window.

30 Alpha Energy Spectrum Po U/Th Total 0.4 Energy 1.4 (F sys -F CV ) / F sys 0.5 Po α depth Bulk α depth Electron drift SC gradient SC offset Po/bulk α fraction Ion mobility Instrumental cuts Relative contributions of these different systematics are constrained by the neutrino data Energy

31 Blind Analysis First month of neutrino data open Then only 20% open to Dec to finalize instrumental background cuts (instrumental cut bias) Thereafter include hidden fraction of neutrons that follow muons (change S/B ratio) AND Omit an unknown fraction of candidate events (change S/B ratio) Detailed internal documentation, review by topic committees Blind Un-blind E ADC (MeV) Box Opened May 2, 2008

32 Neutrino Signal Extraction Live time days NCD raw triggers 1,417,811 PMT raw triggers 146,431,346 NCD υ candidates 7,302 PMT υ candidates 2,381 PDFs and observables Systematic uncertainties Backgrounds 62-parameter likelihood function - 13 CC flux energy bins - 13 ES flux energy bins - NC flux - 35 systematic parameters 3 independent algorithms to determine the neutrino fluxes

33 Markov Chain Monte Carlo (MCMC) Try to sample parameter space (instead of a 62-parameter MINUIT fit) Initial step i parameter guesses p i calculate likelihood L i Add random amounts to all p i : p i+1 = p i + Norm(0,σ i ) calculate likelihood L i+1 if ( Uniform(0,1) > min(1,l i /L i+1 ) ): No. steps else: keep p i+1 keep p i ; start again Metropolis-Hastings method Flux A posterior distribution After burn-in the start point is forgotten and the algorithm samples the function correctly.

34 Systematics Table NC Detection efficiency 3.3% CC PMT energy scale 2.7% NCD energy resolution 2.7% PMT radial scale 2.7% NCD alpha background 2.7% Neutron background 2.3%

35 Opening the Box ES 5 th energy bin posterior Before fix Three algorithms : Markov Chain Monte Carlo (MCMC) Maximum Likelihood with randomly sampled systematics Maximum Likelihood with floating and shift-re-fit systematics After fix Post box opening : (1) 10% difference in NC flux uncertainty between analyses (2) MCMC ES flux low by 0.5 σ

36 Results

37 Compare to Salt Phase Salt NCD CC NC CC ES NC CC ES NC CC ES NC CC ES NC

38 Comparisons CC ES D 2 O con. (306 d) SuperK (1496 d) Salt (391 d) D 2 O con. (306 d) Salt (391 d) NCD (385 d) NCD (385 d) Flux (x 10 6 cm -2 s -1 ) Flux (x 10 6 cm -2 s -1 ) Statistical unc. Total unc CC 1.67 (stat) (sys) 10 6 υ cm -2 s -1 ES 1.77 (stat) (sys) NC 5.54 (stat) (sys)

39 Comparisons NC Corrected to Winter 8 B spectrum D 2 O (306 d) Agreement with past measurements (estimated p-value = 0.328) Agreement with standard solar models Salt (391 d) NCD (385 d) Φ µτ (x10 4 cm -2 s -2 ) Flux (x 10 6 cm -2 s -1 ) Φ e (x10 4 cm -2 s -2 ) φ SSM = 569(1±0.16) x 10 4 cm -2 s -1 (BSB05-OP: Bahcall, Serenelli, Basu Ap. J. 621, L85, 2005).

40 MSW Contours 2-neutrino mixing model. Marginalized 1-σ uncertainties. Solar t-y KamLAND: SNO only degree Solar Solar + KamLAND Cl-Ar Super-K SAGE Gallex GNO SNO Borexino (first result) 766 t-y KamLAND

41 Summary A model independent measurement of the 8 B flux Improved precision on mixing angle θ Reduced correlation between CC and NC Different systematics Agreement with previous measurements More from SNO LETA (Low E Threshold Analysis) of Phases I and II (T=3.5-4 MeV) Muons, atmospheric ν Three-phase solar neutrino analysis Three-neutrino mixing analysis Three-phase hep flux Three-phase solar neutrino Day-Night Asymmetry arxiv: v1 [nucl-ex]

42 Expect the Unexpected Found at the bottom of the cavity:

43 The SNO Collaboration The SNO collaboration University of Alberta, University of British Columbia, Carleton University, University of Guelph Laurentian University, Queen s University SNOLAB, TRIUMF Brookhaven National Laboratory, Lawrence Berkeley National Laboratory, Los Alamos National Laboratory, Louisiana State University, MIT, University of Pennsylvania, University of Texas at Austin, University of Washington LIP (Lisbon) University of Oxford