WIR SCHAFFEN WISSEN HEUTE FR MORGEN Workshop on

WIR SCHAFFEN WISSEN – HEUTE FÜR MORGEN Workshop on electromagnetic dipole moments of unstable particles, Milano 03. /04. 10. 19 Searches for electric dipole moments Philipp Schmidt-Wellenburg , Paul Scherrer Institute Switzerland

Outline State of the art EDM searches The neutron EDM (at PSI) Prospects of a muon EDM (at PSI)

EDM Just another form factor of in the electromagnetic interaction of a fermion magnetic-dipole Anapole - moment electric-dipole charge non-relativistic P, T conserving P, T violating (also CP-violating)

Page 4 The neutron EDM limit •

The mercury-199 limit • 5

The electron EDM limit • Thorium Oxide molecules • Laser excited superposed state • Fluoresence light indicates state • Giant internal electric field (partial Schiff screening) 6

n. EDM searches: no discovery so far EDM Baryon asymmetry of the Universe CPV Arises “naturally” in beyond SM theories Sensitive to QCD CP-violation

n. EDM searches: no discovery so far muon limit Neutron Electron A first non-zero result will be a major discovery Mercury

Complementarity of EDM searches 9 proton, neutron d, 3 He electron Scheme: courtesy Rob G. E. Timmermans Muon

Fresh from ESPP briefing book Page 10

The Ramsey technique Sensitivity: a T N Visibility of resonance Time of free precession Number of neutrons

Searching for an additional coupling to the spin /

Sensitivity for an EDM P: Initial polarization E: Electric field strength N: Number of particles T: Observation time A: Analyzing power 13

V D 2 O thermal moderator Materials used at PSI: DLC 230 ne. V Ni. Mo 220 ne. V d. PS 165 ne. V

Increasing sensitivity One cycle 300 s Many cycle 300 s • 2013

Sensitivity versus Stability • Many cycles sensitivity ideally: • • Require: alk mw o d n a r h wit Gaus sian n oise o nly

Mercury comagnetometer polarization cell B 0 ≈ 1μT PM τ = 140 s ¼ wave plate linear polarizer Hg light Hg. O source

Real data example (stability) Neutron R-ratio Neutron Mercury Allan deviations

Single Ramsey fit • One fit subsequence: combining all E-field states SF 2 ans SF 1 states o A total of 8 fit parameters o Less worry about not sufficient data points in E=0 cycles o Parameter errors are small due to high number of dof. 4 2 depends on SF 1 and SF 2 state one for each SF 1

Obtain d n from R •

Crossing point analysis l e r p y r a n imi y ar n i relim p Main features of this analysis Corrections for: • Field drift (199 Hg) • Gradient drift (Cs. M +Hg) • ABBA pattern • R versus (T, E) • Map corrected

Today’s status of n 2 EDM Status of setup: - MSR installed and commissioning has started - Installation of coil system, vacuum tank and precession chambers next - Area and environmental setup ongoing

The muon EDM • First dedicated experiment to search for EDM of second generation antimatter *Altmannshofer et al. EPJC(2017) **Hertzog DW, EPJ Conf 118(2016)

A relativistic charged particle in a strong B-field 24

Frozen spin technique for the muon EDM • An EDM signal is visible as growing vertical polarization Farley et al. , PRL 93(2004) Adelmann et al. , JPG 37(2010) • 26

Signal: asymmetry of upper to lower detector A. Adelmann et al. JPG 37(2010)085001 • 27 Upper detector Lower detector

HIPA -The proton accelerator facility at PSI PEN Mu. Lan/Mu. Cap Mu. Sun mu. Cool p. He m. H / m. He Lambshift, HFS FAST MUSE MEG-II Mu 3 e-I n. EDM 28

p -beam pion decay channel (8 m, 4 -5 T) 29

• 30 Detector prototype: Combination of scintillating tiles (timing) and thin MAPS (track, momentum) am tel e p sco e( g trig 0. 28 er) m ground Be 12 cm A. Adelmann et al. JPG 37(2010)085001 * H. Yamada et al. NIMA 467/368(2001) Inflector

Summary • The search for an n. EDM is an attractive complementary path for discover of new physics • Many groups worldwide compete for the next most sensitive n. EDM experiment • At PSI two projects searching for EDM of unstable particles • The n. EDM@PSI collaboration has taken sufficient data for a new result un-blinding of the blind analysis still in 2019 • The muon EDM welcomes interested groups and researchers

Thank you for your attention.

Example of complementarity 1 Hg Ra n n O Mh 2 = 400 Ge. V Future: dn x 0. 1 d. A(Hg) x 0. 1 d. Th. O x 0. 1 d. A(Ra) Farther future : dn x 0. 01 d. A(Hg) x 0. 1 d. Th. O x 0. 1 d. A(Ra) Inoue et al. PRD(2014)115023 Chen et al, JHEP 06(2015)056 SSP 2018, Aachen Th Present 34 1 1 P. Schmidt-Wellenburg Higgs Portal CPV (would also lead to strong first order phase transition)

neutron EDM: complementary to other searches SSP 2018, Aachen *J. M. Pendlebury et al. , PRD 92 (2015) 092003 **Guo et al. , PRL(2015)062001 P. Schmidt-Wellenburg 35 From lattice calculations** QCD theta term “natural” in neutron, proton Strongly suppressed in electron

Ph. D project: design prototype detector P. Schmidt-Wellenburg SSP 2018, Aachen 36 • Design a segmented detector • Build prototype • Test using existing dipole Magnet type “ASL” at PSI 36

Injection P. Schmidt-Wellenburg SSP 2018, Aachen 37 • Use weakly focusing magnet ½ integer resonance • 20 turn resonant injection: slowly ramp down perturbator (200 ns) • Start simulation with A. Adelmann and student this autumn • Measure beam emittance at mu 1 E for 125 Me. V/c and 200 Me. V/c • Need for trigger system T. Takayama et al, NIMB 24/25(1987)420 37

Beam trigger P. Schmidt-Wellenburg SSP 2018, Aachen 38 The challenge is to identify an “good” muon and don’t lose it due to detection • Strongly collimate beam and anticoincidence trigger 38

Main challenges P. Schmidt-Wellenburg • Injection • Cannot use conventional kicker due to very short turn-around time (<10 ns) • Possible solution: resonant injection using a non-linear magnetic field perturbation SSP 2018, Aachen • Injection trigger • Minimal latency between detection of “suitable” muon and ramp of injector • Detector • High B-field, minimal spacing, access difficult • Magnet • Conventional technology for fast field reversals (systematic) • Very good uniformity (<1 ppm) because of systematic 39 • 3 D Magnetic field characterization

Building a collaboration! P. Schmidt-Wellenburg SSP 2018, Aachen 40 Detectors • • Full Geant 4 Model Design + FE model Systematic studies DAQ simulator GEANT 4 FE calc. • Fast DAQ ( at 100 Mhz) • Experiment control • Environmental control DA Q • Scintillator / Silicon • Tracker / calorimeter • Trigger E/B field Beam & injection • • Magnet & PS E-field electrodes Sensors (NMR) Magnet shimming • Beam line • Injection scheme • Coils & power supplies Possible partners: PSI, EPFL, other Swiss universities, FZ Jülich, … 40

P. Schmidt-Wellenburg EFT analysis of contributions to F 2 and F 3 magnetic-dipole Anapole - moment charge electric-dipole SSP 2018, Aachen Effective Hamiltonian: 41 A. Crivellin, M. Hoferichter, PSW Ar. Xi. V: 1807. 11484 41

P. Schmidt-Wellenburg • SSP 2018, Aachen 42 A. Crivellin, M. Hoferichter, PSW Ar. Xi. V: 1807. 11484

Comsmological limits on axion plot P. Schmidt-Wellenburg SSP 2018, Aachen 43 We show that Big Bang Nucleosynthesis (BBN) significantly constrains axion-like dark matter. The axion acts like an oscillating QCD θ angle that redshifts in the early Universe, increasing the neutron–proton mass difference at neutron freeze-out. An axion-like particle that couples too strongly to QCD results in the underproduction of 4 He during BBN and is thus excluded. The BBN bound overlaps with much of the parameter space that would be covered by proposed searches for a time-varying neutron EDM. The QCD axion does not couple strongly enough to affect BBN. The supernova bound arises, as too strong coupling would result in lots of axions produced in supernovae which, in turn, would cause it to cool faster than observed. The "Galaxies" bound is dashed. If axions make up all of the dark matter, they need to be heavier than this so that they can reproduce observed distribution of dark matter (rotational curves). If they are only a part of dark matter, they can be lighter.

Overview of the data P. Schmidt-Wellenburg 2. 5% Issues which do not allow to use all data (no HV reversal, too short runs, …) 83. 1% SSP 2018, Aachen unblinded can't be used 13. 8% 3. 0% 44 A total of 54333 cycles to analyze 44

Crossing point analysis Me SSP 2018, Aachen rcu ry Off set erro Vis r ibili Sep ty erro a r Asy ration m. s tats div sta ts d rifts P. Schmidt-Wellenburg 1. 29999999 1. 89999999 1 1. 4 1 45 4. 69999999 Change in %

Least square spectral analysis P. Schmidt-Wellenburg SSP 2018, Aachen 46

Highest peaks P. Schmidt-Wellenburg SSP 2018, Aachen 47

Ramsey fit procedure P. Schmidt-Wellenburg A SSP 2018, Aachen f. RF / f. Hg 48 •

Mercury co-magnetometer SSP 2018, Aachen polarization cell ¼ wave plate linear polarizer Hg lamps 49 τ = 140 s B 0 ≈ 1μT P. Schmidt-Wellenburg PM Hg. O source •

Hg co-magnetometer P. Schmidt-Wellenburg 1. 6 p. T 100 p. T SSP 2018, Aachen 50 Extract B field from Larmor frequency and correct UCN frequency 3. 5 days

EDM and R-calculation P. Schmidt-Wellenburg SSP 2018, Aachen Earth rotation frequency correction : B-gradient fluctuation correction : 51

Three Periodograms SSP 2018, Aachen P. Schmidt-Wellenburg 52

Sensitivity versus field drifts P. Schmidt-Wellenburg SSP 2018, Aachen 53 • Sensitivity for many cycles ideal case: • Only if magnetic field is stable enough. (Good fit with orange, bad fit with purple) fn

Sensitivity versus Stability P. Schmidt-Wellenburg • Sensitivity for many cycles ideal case: SSP 2018, Aachen • Requires: 54 •

Sensitivity versus Stability P. Schmidt-Wellenburg • Many cycles sensitivity ideally: SSP 2018, Aachen • Require: alk mw o d n a r h wit Gaus 55 • sian n oise o nly

Stability and changing E-fields P. Schmidt-Wellenburg SSP 2018, Aachen 56 Options with field changes: • Change E-field with adequate period (e. g. 10 cycles) (loose time due to E ramps) • Use a stack of two neutron precession chambers • Use a comagnetometer Gatchina’s double chamber design Sussex’s co-magnetometer

Magnetic fields P. Schmidt-Wellenburg SSP 2018, Aachen optical pumped magnetometers (Cs. M/Hg. M/Xe. M…) 57

Frequency ratio R = fn/f. Hg P. Schmidt-Wellenburg • Center of mass offset • Non-adiabaticity SSP 2018, Aachen + further sys. 58 199 Hg UCN

Dominant systematic P. Schmidt-Wellenburg • SSP 2018, Aachen Result depends on how particle average the magnetic field: adiabatic (UCN) non - adiabatic (Hg) 59

Dominant systematic P. Schmidt-Wellenburg • Typical B-field gradients: ~10 p. T/cm • Dominant effect from mercury transferred to neutron by correction 60 Measure n. EDM as function of B-Field gradient Afach et al. , EPJD(2015)69: 225 SSP 2018, Aachen n. EDM strategy

Blinding P. Schmidt-Wellenburg How? • Shift the central value by adding an unknown offset EDM of -1. 5 to 1. 5 E-25 ecm to the data SSP 2018, Aachen with • Keep un-blinded data in a safe place (encrypted) 61