The NUMASS Experiment BNL E 952 Presented by

The NUMASS Experiment: BNL E 952 Presented by B. L. Roberts for P. Cushman A Direct Measurement of the Muon Neutrino Mass From 170 ke. V => 8 ke. V and beyond… Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Kyoto, Japan Sept. 27 -29, 2002

Why do a DIRECT Measurement ? It is a fundamental constant! Its uncertainty also affects our knowledge of other fundamental constants d. GF dtm 5 dmm 4 m 2 nm e. g. Gfermi GF 2 tm 2 mm 0. 5 ppm Mu. Lan 0. 38 ppm m 2 m 10 ppm Direct access to mass rather than Dm 2, no oscillation hypothesis It closes the intermediate mass (10 – 170 ke. V) loophole e. g. Neutrino Decay: nm -> ne ne ne via DLo (minimal LR symmetric model) is consistent with Mass Density of Universe, Primordial nucleosynthesis, CMB, Diffuse g-ray bkgd, SN 1987 a as long as m(nm) > 35 ke. V (from Z-width) Supernovae: For m(n. X) > 10 ke. V, Dt ~ day => pulse is below background. Results are only valid for Dirac Neutrinos. Check out all those 17 ke. V papers for more exotic loopholes Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Kyoto, Japan Sept. 27 -29, 2002

Direct Measurements of Neutrino Mass Current Limits m(ne) < 4. 35 - 15 e. V < 23 e. V < 0. 5 - 9 e. V Tritium b-decay endpoint TOF spread from SN 1987 A Double b-decay for Majorana n’s m(nm) < 170 ke. V p -> mn (stopping p’s) m(nt) < 18. 2 Me. V Inv. Mass of t -> n + hadrons (e+e- Colliders) Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Kyoto, Japan Sept. 27 -29, 2002

Muon Neutrino Direct Limits from Pion Decay at Rest 1979: Daum et al. (Phys Rev D 20 p. 2692) Solution A Solution B m 2(nm) = + 0. 13 +- 0. 14 (Me. V/c 2)2 m(nm) < 570 ke. V/c 2 1984: Abela et al (Phys Lett B 146 p. 431) Solution A Solution B m 2(nm) = - 0. 163 +- 0. 080 (Me. V/c 2)2 m(nm) < 250 ke. V/c 2 1996: Assamagan et al. (Phys Rev D 53 p. 6065) Solution A Solution B m 2(nm) = - 0. 143 +- 0. 024 -0. 016 +- 0. 023 (Me. V/c 2)2 m(nm) < 170 ke. V/c 2 Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Kyoto, Japan Sept. 27 -29, 2002

Muon Neutrino Direct Limits from Pion Decay in Flight 1982: Anderhub et al. Phys Lett B 114 p. 76 m 2(nm) = - 0. 14 +- 0. 20 (Me. V/c 2)2 m(nm) < 500 ke. V/c 2 200? Neutrino Mass Experiment m(nm) < 8 ke. V/c 2 Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Kyoto, Japan Sept. 27 -29, 2002

Muon Neutrino Direct Limits Graphical Form If you believe atmospheric neutrino result: nm => nt with only Dm 2~. 002 Then this experiment reduces the t neutrino mass limit by 3 orders of magnitude! Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Kyoto, Japan Sept. 27 -29, 2002

In a perfectly uniform B-field Any charged particle returns to origin independent of B, p, q * Origin can produce a range of angles and momenta * Uniformity is more important than value of B * 1 st harmonic (and other nonuniformities) are always monitored using residuals of prescaled pions and undecayed protons “origin” Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Kyoto, Japan Sept. 27 -29, 2002

G-2 Storage Ring becomes a Precision Spectrometer G-2 Experiment Weak-focussing Storage Ring: Muons stored for 800 ms Quadrupoles Muon Kicker Nu. Mass Experiment Spectrometer: p -> mn observed evt-by-evt No Quads Pion kicker Same Momentum - 3 Ge. V (re-examine at new facility) retain excellent shimming and B-field uniformity 0. 1 ppm over orbital path range Trolley runs in vacuum to map field Fixed probes to track changes Active shimming and thermal insulation to minimize change Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Kyoto, Japan Sept. 27 -29, 2002

Put pions on orbit using d. E/dx Injection 5. 2 cm Beryllium p orbit without degrader p orbit with degrader “Pion Kicker” D p = -16. 2 Me. V/c X/Xo = 14. 7 % q (rms) = 1. 56 mr Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Kyoto, Japan Sept. 27 -29, 2002

Conceptual Design Forward-going decay muons orbit a larger diameter by DD CM nm p m q = 29. 7 Me. V/c DD D pm - pp pp 0. 7 Me. V/c 3 Ge. V/c decay m’s undecayed pions 3. 26 mm 14 m D DD d. D depends on m(n) However, if mn is non-zero, then DD shrinks by d. D D -mn 2 2 q mp ( d. D = 0. 04 mm for current limit ) Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Kyoto, Japan Sept. 27 -29, 2002

Concept Continued Radial distribution of forward-going muons has sharp outer edge. Auxiliary detectors veto non -forward going muons J-Veto undecayed pions daughter muons Edge of muon distribution for zero-mass neutrinos g-2 Cal’s R (muon) – R (pion) in mm Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota S 1 S 2 Kyoto, Japan Sept. 27 -29, 2002

Experimental Method p Injection Beam counter J-veto: restrict early m‘s at large angles J-cal: 2 nd turn electron id 24 g-2 calorimeters restrict late decays identify electron bkg initial beam tuning C-veto: restrict incoming p’s decay m p orbit S 1 S 2 2. 56 cm S 2 6. 4 cm Trigger pads Beryllium S 1 Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Kyoto, Japan Sept. 27 -29, 2002

Highlights of the Experimental Technique • Translate Dp to Dr in 0. 1 ppm uniform B-Field no multiple scattering no need to measure decay angle or location • Reference each m to parent p slow extraction • In situ alignment protons (7 ns/turn late) prescaled undecayed pions remote positioning of active vetoes remote angular adjustment of detector • Position resolution from silicon with upgrade to emulsion 1. 4 mm SSD => 0. 5 mm emulsion reel • Time resolution from scintillators and PMT’s tight triple coincidence trigger TDC’s on all vetoes and embedded hodoscope Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Kyoto, Japan Sept. 27 -29, 2002

32 strips per Viking chip serial readout into 1 ADC @40 MHz = 0. 8 ms Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Readout System Kyoto, Japan Sept. 27 -29, 2002

225 ns 150 ns Beam Counter Hodoscope Trigger: latch data S 1 (strip 6) p 1 st turn Detect parent pion in S 1 (e. g. in strip 6 of S 1) O S 2 (strip 6) p S 1 (strip 71) m S 2 (strip 71) m 2 nd turn O O Detect daughter muon in S 2 (3. 2 mm away from strip 6) O Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Kyoto, Japan Sept. 27 -29, 2002

Proposed Parasitic Running with AGS Crystal Extraction E 949 Running Conditions E 952 Parameters 25 Gev protons 70 TP in a 4. 1 s spill / 6. 4 s cycle 2. 8 x 106 p+ into g-2 ring/TP 5. 4 x 1012 p+ for an 8 ke. V result Running Time 5% of SEB beam => 492 hrs (crystal extr. eff. ) Instantaneous rates (100% extr. eff. ) Detector p-p Entering Ring 8 x 106 part/s Triggers p-m Offline (p-m)+vetoes 1 x 106 part/s 1. 8 x 105 s-1 910 s-1 400 Hz/strip 55 ms/SSD 11 ms/SSD Prescale in trigger 100 MB/s 0. 5 MB/s Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota 42 s-1 Kyoto, Japan Sept. 27 -29, 2002

Scintillator Hodoscope Radial segmentation = 2 mm Vertical segmentation = 12. 8 mm • 4 ns gate for 3 -fold coincidence trigger Accidentals at 0. 004, flagged by beam counter • Veto events Dr < 2 mm to enrich p-m events x 50 prescale => 0. 5 MB/s or 37 DLT tapes • Select readout SSD 0. 7% dead time 1/10 data volume • 1 ns timing resolution (TDC) + 2 mm segmentation reject accidentals offline (another factor of. 002) Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Kyoto, Japan Sept. 27 -29, 2002

Sources of Background • Beam-gas scatters vacuum is 10 -6 torr • Injected p (27%) 7 ns/turn slower • Injected e (12%) lose 1 Me. V/turn from SR (4. 7 mm inward) identify in J-Veto calorimeter (or position) • m => enn (gt = 64 ms) injected m (1%) and p =>mn < 10 -4 of good p -m events rejected by g-2 calorimeters • p => en (BR=1. 2 x 10 -4) low tail out to ~ 5 mm calorimeter at inner J-Veto Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Kyoto, Japan Sept. 27 -29, 2002

PROOF OF PRINCIPLE Test of a prototype silicon microstrip detector in the g-2 ring SSD Sandwich SSD 2 PSD tiles SSD 1 12. 8 mm Removable Copper sheets Vacuum Flange with fiber feedthru Viking Readout Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Kyoto, Japan Sept. 27 -29, 2002

Test successfully measured 1 st and 2 nd passage of pion 1 mm 1 st turn 2 nd turn The right side of ring has less average B-field than the left. 1 st harmonic ~ 50 ppm Online traces from silicon strip detectors triggered by embedded scintillator. Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Kyoto, Japan Sept. 27 -29, 2002

And lots of other important parameters as well • Establish pion injection and orbit parameters in g-2 ring and steer pions to detector • Pion kicker (degrader) in vacuum returns pions to origin measure 1 st harmonic of g-2 magnet • Read out the SSD/serial Viking system through g-2 WFD’s • Run SSD in Vacuum and 1. 45 T B-field and establish characteristics MIP response, Charge-sharing, alignment position resolution signal: noise Charge (ADC) S&H delay curve charge-sharing • Trigger tile-fiber scintillator hodoscope on MIP’s (with vacuum feedthru’s) • Timing and linearity study of SSD hold time • Multiple scattering in degrader Double-hit residuals No degrader double degrader • Resolution for both single and double hits Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Kyoto, Japan Sept. 27 -29, 2002

Conclusions … Physics There are mysteries in the neutrino mass spectrum which a complementary, direct measurement can help unravel. Oscillation Exp. only sensitive to Dm 2 n disappearance => oscillation => mass No direct access to the 3 flavor (or more? ) mass hierarchy Astrophysics/Cosmology assumes no sterile n, standard model interactions, stable n Supernovae only applies to Dirac neutrinos model-dependent at supernuclear densities … Experiment G-2 storage ring: state-of-the-art spectrometer at bargain prices Beamline can be parasitic with SEB, easily switched to RHIC, FEB Upgrade path uses the SSD to locate track and emulsion to further refine position: Can improve limit to below 1 ke. V Nuclear and Particle Physics at 50 Ge. V PS Priscilla Cushman University of Minnesota Kyoto, Japan Sept. 27 -29, 2002