Sta$s$cs, Systema$cs and Run Phases

Sta$s$cs, Systema$cs and Run Phases DOE Nuclear Physics MOLLER Science Review UMass, Amherst! September, 2014 Kent Paschke University of Virginia

α Rates, Noise Budget, and Sta$s$cal Precision 0.4 0.35 0.3 0.25 0.2 0.15 0.1 0.05 90 o CM Acceptance Moller 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02 0.022 θ [rad] 2 mrad 23 mrad ep Noise Budget Parameter Noise (75 μa) Noise (60 μa) Sta$s$cal Width (0.5 ms) ~79 ppm ~89 ppm Target Density FluctuaXon 26 ppm 21 ppm Beam Intensity ResoluXon ppm ppm Beam PosiXon Noise 7 ppm 7 ppm Detector ResoluXon (25%) 3.1% 3.1% Electronics Noise ppm ppm Measured Width 88 ppm 95 ppm High Rate, high precision 150 cm cell, up to 85 μa 38% 70% (CM) acceptance @ 75 μa: 144 GHz Mollers, 159 GHz total Quiet JLab beam, high- flow target, and low- noise electronics keeps precision high A expt = p pair N N pair = (8.3x 7 ) N days A PV 1 P b 1 (1 f bkgd ) A expt A PV ~ 35 ppb 9.5% irreducible background fracxon 80% polarizaxon A expt ~ 25 ppb DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 2

Run Phases Mul$ple Run Phases Op$mizes Sta$s$cal and Systema$c Precision Hardware changes (beamline, collimaxon, electronics, detectors ) Detailed analysis Improved run planning / test planning PublicaXon of intermediate results E158 Run 1: 2001 collimator upgrade Run 2: 2002 slice bpm upgrade 1st physics publicaxon Run 3: 2003 75% of total staxsxcs Qweak Commissioning: Oct 20 BCM electronics detector radiators, new collimators background detectors 1st publicaxon Run 1: Oct 20 - May 2011 Dump beamline rebuild BCM electronics upgrade Run 2: Nov 2011 - May 2012 >50% of staxsxcs DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 3

Run Phases Plan for Approaching 2% Precision 1 khz Width PAC Days (prod) Stat Error (ppb) Stat Error (%) Eff % Calendar Weeks (prod.) Comm Weeks Total Weeks I 5 14 3.09.9 40 5 6 11 II 0 95 1.13 3.9 50 27 3 30 III 95 235 0.68 2.4 60 56 4 60 Total 344 0.57 2.01 13 1 Run Phase 1 Spectrometer opxcs, acceptance, alignment First look at backgrounds Test sufficiency of beam correcxon tools and analysis beam quality (asymmetry and halo) Result: near precision of SLAC- E158 60 μa, 90% polarization Run Phase 2 staxsxcal behavior of beam asymmetries, measured asymmetry quality of slow reversals (Wien, g- 2) precision on background, normalizaxon, beam correcxons, polarizaxon Result: 2.5x beyond SLAC- E158, δ(sin 2 θ W )=0.00044 (stat), 0.00047 (stat+syst) Run Phase 3 ulxmate precision, ulxmate systemaxc uncertainty Result: δ(sin 2 θ W )=0.00024 (stat), 0.00028 (stat+syst) DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 4

Physics Impact of Run 2 result sin 2 eff (e) 0.235 0.235 LHC events 0.234 0.234 E158 MOLLER (Ul$mate) 0.233 0.233 0.232 0.232 MOLLER 0.231 0.231 MOLLER (Run II) 0.23 0.23 0.229 APV 0.229 0.228 0.228 0.227 1 0 00 0.227 M H [GeV] DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 5

Systema$c Uncertainty Uncertainty Source Ul$mate Frac$onal Error (%) Sta$s$cal 2.0 kinemaxc normalizaxon 0.5 Beam PolarizaXon 0.4 Transverse beam polarizaxon 0.2 beam (2nd moment) 0.4 Beam (posixon/angle/energy) 0.4 Beam (intensity) 0.3 e+p (+γ) e+x (+γ) 0.4 Designed with experience with E158, Qweak, and PREX! Plans for managing systemaxc uncertainxes in normalizaxon backgrounds beam asymmetries e+p (+γ) e+p (+γ) 0.3 γ + p (π,μ,k) + X 0.3 e+al (+γ) e+al (+γ) 0.3 neutral backgrounds 0.1 Total systema$c 1.1 DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 6

A P V = me G F p 2 Kinema$c Normaliza$on 4sin 2 (3 + 2 cos 2 ) 2 Qe W Need 4sin 2 (3 + 2 cos 2 ) 2 to within 0.5% of itself Map acceptance using GEM chambers between the hybrid magnet and integraxng detector stack Acceptance 11 GeV ep 0.4 0.35 Moller ep p 11 GeV ee OpXcs calibraxon using lower beam energy, sieve slit, opxcs target α 0.3 0.25 0.2 0.15 0.1 0.05 θ 0 0.002 0.004 0.006 0.008 0.01 0.012 0.014 0.016 0.018 0.02 0.022 θ [rad] CalibraXons and tracking studies will recur through all run phases UlXmately esxmated using simulaxon, benchmarked to tracking runs DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 7

Beam Asymmetries beam posixon, angle, energy variaxon with helicity causes an experimental asymmetry Injector source laser e- beam delivery (adiabaxc damping) cancellaxon with slow reversals detector symmetry correcxon calibraxon (beam modulaxon) 20-50 nm in 5MeV injector factor of - 0 from adiabaxc damping factor of 2- Factor > in sensixvity % precision X position difference 6 0.56 ± 0.53 nm 5 HAPPEX- II: Zero posixon differences RMS = 2.77 µm Y position difference 6 5 1.69 ± 1.83 nm RMS = 9.50 µm 4 3 2 4 3 2 HAPPEX- II: posixon difference convergence 1 40 30 20 0 20 30 40 50 micron X angle difference 6 5 0.26 ± 0.24 nrad RMS = 1.23 µrad 1 0 50 0 50 0 micron Y angle difference 6 5 0.21 ± 0.25 nrad RMS = 1.29 µrad micron 4 3 2 4 3 2 1 30 20 0 20 30 µ rad 1 40 30 20 0 20 30 40 1 day averages DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 8

Beam Asymmetries - 2nd order Helicity- correlated Beam Spot Size variaxon creates false asymmetry Source configura$on Bounds < - 4 from laser configuraxon must be maintained throughout run Adiaba$c Damping Good beam match keeps variaxon small Slow Reversals Laser opxcs reversals (e.g. IHWP) do not cancel expected sources of spot- size differences Helicity reversal on e - beam will be incoherent with spot- size differences Net factor ~ suppression of beam asymmetries Injector Spin Manipula$on Solenoids + 2 Wien rotaxons ~80 reversals during run phase 2&3 (weekly) PREX- II showed ISM cancella$on of posi$on differences micron g- 2 rota$on Beam energy (ΔE~0 MeV) ~few reversals during run phases 2 and 3 DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 9

Beam Polariza$on Unimpeachable credibility for 0.4% polarimetry Two independent measurements which can be cross- checked Con$nuous monitoring during produc$on (protects against drigs, precession...) Sta$s$cal power to facilitate cross- normaliza$on (get to systema$cs limit in about 1 hour) High precision opera$on at 11 GeV Compton Møller Detection of backscattered photons and recoil electrons High-Gain Optical Cavity 532 nm (green) or 64 nm (IR) Microstrip electron detector Photon calorimeter continuous measurement with high precision state-of-the-art: 0.5% (SLD), 0.8% (JLAB) laser polarization to 0.2% Independent electron/photon analyses, each expected to reach 0.4% Pure Iron at High Field Magnetized perpendicular to foil 3-4 T applied field - magnetization saturated Spin polarization, known to 0.25% 0.5% precision already claimed on Hall C polarimeter Low-current, invasive measurement Spin polarization can not be independently verified DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014

Beam Polariza$on - Evolu$on to 0.4% Projected Performance: Analysis improvements, beam tests, hardware evoluxon needed to improve from baseline upgrade Run Phase 1: projecxng to completed upgrades in progress, from current state- of- the- art: Compton electron - 0.8% Compton photon - 0.7% Compton correlated error: 0.2% Moller: 1% Run Phase 2: Compton electron: 0.4% Compton photon: 0.4% Compton correlated error: 0.2% Moller: 0.5% Detailed, high- precision cross- check in Run 2 Alternative: Atomic H Møller Moller polarimetry from polarized atomic hydrogen gas in an ultra-cold magnetic trap Brute force polarization Development at Mainz underway for P2 at MESA Pe =0% ± -4 high beam currents Non-invasive, continuous E. Chudakov and V. Luppov, IEEE Transactions on Nuclear Science, v 51, n 4, Aug. 2004, 1533-40 possible alternative if needed to resolve controversial results in Compton and high-field Moller DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 11

Transverse Asymmetry Interference between one- and two-photon exchange electron beam polarized transverse to beam direction AT (ppm) For identical particles: magnitude of asymmetry must be odd around 90 degrees in the center of mass Potential systematic error in APV. Suppressed by - small transverse polarization - azimuthal acceptance symmetry - acceptance symmetry in c.m.s. polar angle 0 2.75 5.5 8.25 E (GeV) 11 Measured at E158 Theory References: 1. A. O. Barut and C. Fronsdal, (1960) 2. L. L. DeRaad, Jr. and Y. J. Ng (1975) 3. Lance Dixon and Marc Schreiber:hep/ph-0402221 DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 12

Transverse Beam Polariza$on Average transverse asymmetry (ppm) simulated: ~1 hour at P T =0% A T 5 0 5 expected grand average for the simulated experimental acceptance 0 5 15 20 25 detector number IniXal beam setup ~ 1-2 degrees Unique signature of transverse beam polarizaxon 50 ppb error on A T *P b in 4 hours: 1 degree precision Over enxre run: feedback will hold transverse polarizaxon small (<<1 degree) Rate (GHz/0.1GeV) 3 2.5 2 1.5 1 0.5 Moller E' vertex Distributions All Open Transition Closed Run Phase 1: A T measurement Feedback technique tested Run Phases 2 and 3: RouXne feedback 0 0 2 4 6 8 (GeV) E' vertex DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 13

ep Backgrounds black: Møller red: ep elas$c green: ep inelas$c ep elas$cs: 8.9% under Møller peak, asymmetry well known ep inelas$cs: <0.5% of signal but asymmetry not well known Al elas$cs: ~0.3% of signal, elasxc from target windows Radial and Azimuthal binning - measure asymmetries under the Møller peak R4, R3, R2: inelasxc contribuxon dominant R4: best measure. W 2 distribuxon matches signal region R5 at low W 2 Test using radial and azimuthal variaxons open transi$on closed DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 R5 Azimuthal segments black: Møller red: ep elas$c green: ep inelas$c 14

Other Backgrounds π/μ: direct scaling from E158 suggests ~0.1% fracxon, simulaxon confirms 0.13% asymmetry ~500 ppb hyperon decay could contribute significantly direct measurement with pion detector required A}er run 3: known to within ~150ppb. GEM GEM quartz assembly centerline pion detectors luminosity monitor Neutrals: suppressed by thin quartz, lead- shielded PMT assemblies blinded tube, spectrometer- off runs Expected contribuxon ~0.1%, will be well measured Expected asymmetry negligible, will be well measured with auxiliary detectors DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 15

Background Summary Background FracXon of total signal (%) Asymmetry (ppb) correcxon (ppb) Note e+p (+γ) e+p (+γ) 8.9 ~40 ~3.6 A ep e+p (+γ) e+x (+γ) <0.5 ~300 ~1.4 Measured in azimuthal/radial dependence γ + p (π,μ,k) + X ~0.1 ~500* ~0.7 EsXmated from E158. Hyperon decay would show up here. Direct measurement in pion detector. Al elasxc (target) ~0.3 ~440 1.3 simulaxon, QWeak measurement neutral backgrounds <0.1 0 0.1 measured with blinded tube and other auxiliary measurements DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 16

Summary of Systema$c Uncertainty Uncertainty Source Run Phase1 Frac$onal Error (%) Run Phase 2 Frac$onal Error (%) Ul$mate Frac$onal Error (%) Sta$s$cal.9 3.9 2.0 kinemaxc normalizaxon 3 0.7 0.5 Beam PolarizaXon 1 0.4 0.4 Transverse beam polarizaxon 2 0.2 0.2 beam (2nd moment) 4 0.4 0.4 Beam (posixon/angle/energy) 4 0.4 0.4 Beam (intensity) 3 0.3 0.3 e+p (+γ) e+x (+γ) 2 0.4 0.4 e+p (+γ) e+p (+γ) 1 0.3 0.3 γ + p (π,μ,k) + X 1 0.4 0.3 e+al (+γ) e+al (+γ) 0.3 0.3 0.3 neutral backgrounds 0.5 0.1 0.1 Total systema$c 8.0 1.3 1.1 DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 17

Summary of Precision 60 μa, 90% polarization Run Phases Stat Error A Stat Error Q Syst Error Q Tot Error Q Stat Error sin Syst Error sin Tot Error sin I 3.09.9 8.0 13.5 0.00129 0.00096 0.00161 Through II 1.06 3.71 1.3 3.9 0.00044 0.00015 0.00047 Total (III) 0.57 2.01 1.1 2.3 0.00024 0.00013 0.00028 DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 18

Run Phases 1 khz Width Run Phase Scheduling PAC Days (prod) 60 μa, 90% polarization Stat Error (ppb) Stat Error (%) JLab FY = 30 weeks, usually 2 separated periods with ± 5 weeks separaxon Run Phase 1 Run Phase 2 Run Phase 3 1/3 FY running 1 full FY 2 full FY Breaks of 2-3 months (consistent with full JLab operaxon schedule) provide opportunity for analysis, modificaxons, upgrades A significant break (4 months+) is desirable, especially between Run 1 & 2 or Runs 2 & 3. More complete analysis, hardware improvements, run plan opxmizaxon Eff % Calendar Weeks (prod.) Comm Weeks Total Weeks I 5 14 3.09.9 40 5 6 11 II 0 95 1.13 3.9 50 27 3 30 III 95 235 0.68 2.4 60 56 4 60 Total 344 0.57 2.01 13 1 Scheduling flexibility to run other experiments between MOLLER run phases DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 19

Modular Experimental Design Modular design and moderate alignment tolerances provide scheduling flexibility HRS/BigBite experiments need: pivot region upstream beamline open aperture to beam dump room for HRS in forward direcxon MOLLER: modular assemblies protect crixcal alignments Target chamber collimator assemblies Hybrid Magnet assembly Alignment tolerances for posi$oning of assemblies ~ 1mm HRS snouts and target pivot Upstream Magnet, front collimators \ Collimator Assemblies Hybrid Magnet Assembly Target Chamber Modular shielding Tracking detectors IntegraXng detector assembly DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 20

Cri$cal Alignments in Fixed Assemblies Hybrid Magnet Assembly Coils in assembly, relaxve alignment preserved to simplify re- commissioning Posi$on Tolerance ~1mm Machining tolerance on acceptance defining collimator: 200 μm Upstream Magnet, front collimators Collimator DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 21

Summary Plans for achieving systema$c and sta$s$cal error goals in 3 run phases Experiment opxmized for high precision and accuracy - At design current: <90 ppm uncertainty at 960 Hz. SystemaXc uncertainxes must be correspondingly small. - Building from experience and experxse of E158, Qweak, and PREX, design for 1.1% (0.4 ppb) systemaxc uncertainty! Run phases are designed to fit into JLab annual running cycles - Fit a natural schedule for hardware and configuraxon opxmizaxon - Phased plan, modular detector design, and moderate tolerance requirements provides scheduling flexibility! First run (commissioning): result compares to SLAC- E158 precision First physics run: - SystemaXc error control at high level - high- impact physics result: 2.5x improvement over E158 Final physics run: push precision to level of collider measurements Plan shows path to full precision result in 3 years of physics running DOE Nuclear Physics MOLLER Science Review UMass, Amherst September, 2014 22