Muon g-2/EDM experimentat J-PARCSchwingerFest, UCLA

Dec 4, 2018Tsutomu Mibe (KEK)

for the J-PARC g-2/EDM collaboration

1

www.g-2.kek.jp

28 Julian Schwinger

Fig. 3.1 Memorial lecture of Professor Julian Schwinger for Professor Sin-itiro Tomonaga (July8, 1980, Tokyo)

Two shakers of physics

2

Memorial lecture of Professor Julian Schwinger for Professor Sin-itiro Tomonaga (July 8, 1980, Tokyo)

“Shin (�) : to wave, shakeSchwing : to swing, shake”

Lecture note on physics 746, pp 27, 2008

Why new g-2 experiment?

• An important discovery must be confirmed by independent experiments.– a(SM) : e+e- expt’s, theory groups, lattice groups– a(exp) : BNL à Fermilab, J-PARC

• New experimental method to explore further precision.

3

Three steps of g-2 measurement

1. Prepare a polarized muon beam.

2. Store in a magnetic field (muon’s spin precesses)

3. Measure decay positron

4

úúû

ù

êêë

é÷÷ø

öççè

æ+´+

´÷÷ø

öççè

æ-

---=cEB

cEaBa

me

!!!

!!!!

bhbg

w µµ 2112

π+ μ+νμ

spin 0

neutrino� left handed

helicity�−� helicity�−�

μ+

e+

spin

μ+

B

spin

Ideal g-2 experiment?

• Beamline phase space matched at injection • All muons have same lifetime• All muons have same spin direction• No muon losses

5

μ+

B

Y. Semertzidis, g-2 school at BINP (2018)

https://indico.inp.nsk.su/event/14/

6

Thermal muoniumproduction,Ionization laser

Muon storagemagnet(3 T)

MLF muon experimentalfacility (H-line)

Positron trackingdetector

Proton beam (3 GeV)

Surface muon (4 MeV)

Ultra-slow muon (25 meV)

Reaccelerated muon(212 MeV)

3D spiral injectionMuon LINAC

Muon g-2/EDMexperimentat J-PARC

Features:• Low emittance muon beam (1/1000)• No strong focusing (1/1000) & good injection eff. (x10)• Compact storage ring (1/20) • Tracking detector with large acceptance• Completely different from BNL/FNAL method

7

The Collaboration 102 members (Canada , China, Czech,France, Japan, Korea, Russia, USA)

New Ph.Ds

8

Awards

9

PASJ poster award

LINAC 2018 poster award

KEK student day poster award

muon g-2 and EDM measurements

10

úúû

ù

êêë

é÷÷ø

öççè

æ+´+

´÷÷ø

öççè

æ-

---=cEB

cEaBa

me

!!!

!!!!

bhbg

w µµ 2112

( )úûù

êëé ´+-= BBa

me !!!!

bhw µ 2

In uniform magnetic field, muon spin rotates ahead of momentum due to g-2 = 0

ω = −

em

aµB+η

2

β ×B+Ec

#

$%

&

'(

)

*+

,

-.

BNL E821 approachγ=30 (P=3 GeV/c)

J-PARC approachE = 0 at any γ

J-PARC E34

general form of spin precession vector:

FNAL E989

Conventional muon beam(BNL, Fermilab)

proton π+ μ+

pionproduction

decay

emittance~1000π mm�mrad

Strong focusingMuon lossBG π contamination

11

Muon beam at J-PARC

Reacceleratedthermal muon

proton π+ μ+

pionproduction

decay

cooling μ+

emittance~1000π mm�mrad

emittance1π mm�mrad

Strong focusingMuon lossBG π contamination

Free from any of these

12

( + - ) + + -( ) 2 5

2 5 7( 5( 5 7)

.A

HA A

0 1 C(

2E A E7 . 7 C 4 103,-

+ -2 5

( 2 5 7) )

µ

µ µ

/

Re-accelerated thermal muon

Experimental sequence

14

40ms (25 Hz)

~1μs

40 μs

Surfacemuon beam(28 MeV/c)

Thermalmuonium(3keV/c)

Ultra-slow muon(3keV/c)

Acceleration + injection(300 MeV/c)

Storage and detection(300 MeV/c)

~1ns

~3ns

laser ionization

μ+àe+

μ+

μ+

Mu

μ+

s]µ) [aTime (modulo T0 0.2 0.4 0.6 0.8 1 1.2 1.4 1.6 1.8 2

Up-

dow

n as

ymm

etry

0.06-

0.04-

0.02-

0

0.02

0.04

0.063-10´

1110´ = 5.7+e = 50% N+µP < 275 MeV+e200 MeV < Ecm× e-2010´ = 1 µd

s]µTime [0 5 10 15 20 25 30

cou

nts

/ 5 n

s+

Dec

ay e

710

810

9101110´ = 5.7+e = 50% N+µP

< 275 MeV+e200 MeV < E

15

Expected time spectrum of e+ in µàe+nn decay

w

ω = −

em

aµB+η

2

β ×B( )#

$%&

'(

e+ decay time (µsec)

∝ EDM

Mod(e+ decay time, Tg-2) (µsec)

Expected uncertainties

16

Table 5 Summary of statistics and uncertainties

Estimation

Total number of muons in the storage magnet 5.2× 1012

Total number of positrons 0.57× 1012

Effective analyzing power 0.42

Statistical uncertainty on ωa [ppb] 450

Statistical uncertainty on ωp [ppb] 100

Uuncertainties on aµ [ppb] 460 (stat.)

< 70 (syst.)

Uncertainties on EDM [10−21 e·cm] 1.4 (stat.)

0.36 (syst.)

Once the ωa and ωp are extracted from the experimental data, aµ is obtained from Eq. 5.435

Table 5 summarizes statistics and uncertainties for 2× 107 seconds of data taking. The436

estimated statistical uncertainties on ωa and ωp are 450 ppb and 100 ppb, respectively.437

Thus, the statistical uncertainty of aµ would be 460 ppb.438

Systematic uncertainties on ωa are estimated as follows. A timing shift due to pile up of439

hits in the tracking detector is estimated as less than 36 ppb in the detector simulation by440

taking into account time responses of readout electronics. A correction for pitch angle is not441

necessary in the case of the muon storage in the perfect weak magnetic focusing field [53]. A442

difference in the actual field distribution from the perfect case leads to a systematic uncer-443

tainty of 13 ppb which is estimated from a precision spin-tracking simulation of the muon444

beam storage. Residual electric field modifies ωa through the β × E term. With 1 mV/cm445

monitoring resolution for an E-field, the error on ωa is 10 ppb. Other effects, such as delayed446

high-energy positrons and differential decay, are of the order 1 ppb. In the ωp measurement,447

absolute calibration of the standard probe has an uncertainty of 25 ppb. Positioning res-448

olution of the field mapping probe at the calibration point and the muon storage region449

leads to 20 ppb and 45 ppb uncertainties, respectively. Other effects, such as field decay and450

eddy current from kicker are less than 10 ppb. In summary, we estimate that the combined451

systematic uncertainties on aµ is less than 70 ppb.452

A muon EDM will produce muon spin precession out of the horizontal plane that is defined453

by the ideal muon orbit. This can be seen from Eq. 4 where the second term is the EDM454

term that is perpendicular to the aµ term. Due to the fact that the EDM term generates455

vertical motion of the spin, one can extract the EDM term from the oscillation of the up456

and down asymmetry AUD(t) in number of positrons detected,457

AUD(t) =Nup(t)−Ndown(t)

Nup(t) +Ndown(t)=

PAEDM sin (ωtott+ φ)

1 + PA cos (ωtott+ φ), (8)

where P and A are the polarization of the muon and an effective analyzing power of muon458

decay, respectively. AEDM is an effective analyzing power associated with the EDM. A459

simulated up-down asymmetry in the case of dµ = 1× 10−20 e · cm is shown in Fig. 14 (right).460

The estimated statistical sensitivity for EDM is 1.5× 10−21 e · cm (See Table 5).461

19/21

Bird’s eye photo in Feb. 2008 17

Proposed experimental site

18

Parking lot

muonproductiontarget

sparation neutronsource

g-2/EDMstorage magnet

muon linac

H-Line

MuSEUMDeeMe

Material and Life science Facility in J-PARC

New ElectricPower substation

H-line being constructed!

19

Photo by T. Yamazaki

To g-2/EDM

New power station being constructed

20Photo by T. Yamazaki

21

Production target

(20 mm)

3 GeV proton beam

( 333 uA)

Surface muon beam

(28 MeV/c)Muonium Production

(300 K ~ 25 meV⇒2.3 keV/c)

SiliconTracker

66 cm

Super Precision Storage Magnet

(3T, ~1ppm local precision)

Resonant Laser Ionization of Muonium (~106 µ+/s)

Δ(g-2) = 0.1 ppm

ΔEDM = 10-21 e�cm

22

muon beam

Muonium(µ+e-)

Silica aerogel�SiO2, 30 mg/cc�

Laser-ablated holes

8 mm

� no hole� w/ holes

P. Bakule et al., PTEP 103C0 (2013)G. Beer et al., PTEP 091C01 (2014)

Muon LINAC

30

406172.

b

-

5

0

0

0

0

5 5 58 8

-

23

x (mm)8− 6− 4− 2− 0 2 4 6 8

x’ (m

rad)

8−6−4−2−02468

y (mm)8− 6− 4− 2− 0 2 4 6 8

y’ (m

rad)

1.5−

1−

0.5−

0

0.5

1

1.5

(deg.)φ ∆10− 8− 6− 4− 2− 0 2 4 6 8 10

w (M

eV)

∆

1−0.8−0.6−0.4−0.2−

00.20.40.60.8

1

p (MeV/c)298 300 302

Even

ts

01000200030004000500060007000

RMS/Mean: 0.04%Mean: 299.7 MeV/c

Phase space distributions after muon LINAC (simulation)

Muon RF acceleration for the first time!

24

J-PARC MLF D2 area, October 2017 Slide by M. Otani

Muon RF acceleration for the first time!

25

J-PARC MLF D2 area, October 2017 Slide by M. Otani

S. Bae et al.,Phys. Rev. AB 21, 050101 (2018).

26

An accelerating structure (IH-DTL cavity)to 1.3 MeV Mar 2017, Photo by M. Otani

Design: M. Otani et al., Phys. Rev. AB 19, 040101 (2016)

Muon beam injection and storage

Horizontal injection + kicker

(BNL E821, FNAL E989)

3D spiral injection + kicker

(J-PARC E34)

Injection efficiency : 3-5%(*)

(*) PRD73,072003 (2006)

Injection efficiency : ~85%

H. Iinuma et al., Nucl. Instr. And Methods. A 832, 51 (2016)27

Muon storage magnet and detector

28

Cryogenics

e+ trackingdetector

2900 mm

Muon storage orbit

Iron yoke

Super conducting coils

666 mm

Drawn by Hitachi Co.

M. Abe et. al., Nuclear Inst. and Methods in Physics Research A 890, 51 (2018)

• Requirements– Detection of e+ (100<E<300 MeV)

– Reconstruction of momentum vector

– Stability over rate changes�1.4 MHz à 14 kHz�

• Specifications– Sensor: p-on-n single-sided strip

– Number of vanes: 40

– Number of sensors : 640

– Number of strips : 655,360

– Area of sensors : 6.24 m2

Positron tracking detector

29

¼ vane

φ590 mm

750

mm

200 mm

First test module tested with muon at J-PARC

30June 2017

e+ track reconstructionwith positron tracking detector

31

Y. SatoLow rate (0.6 track/ns)

High rate (6 track/ns)

Simulation

Simulation

Reconstruction efficiency8. DETECTOR AND RECONSTRUCTION 35

Positron energy [MeV]0 50 100 150 200 250 300 350

Find

ing

effic

ienc

y

0

0.2

0.4

0.6

0.8

1 0.06 tracks/ns

0.6 tracks/ns6 tracks/ns

Positron energy [MeV]0 50 100 150 200 250 300 350

Rel

ativ

e fin

ding

effi

cien

cy

0

0.2

0.4

0.6

0.8

1 0.6 tracks/ns

6 tracks/ns

Figure 27: Track finding efficiency as a function of positron energy in different conditions onthe positron track rate (left) and track finding efficiency respect to the sample of track rate of0.06 tracks/ns (right). The detector with 48 vanes are used and the effect of readout ASIC is consid-ered.

and fitting separately the different energy-sorted time histograms, one can increase thestatistical power of the data and reduce the sensitivity to a change in efficiency vs time. Inthat case, an energy-sorted histogram having a rate-changing efficiency would only look likea varying slow term, which can be removed by the ratio analysis or the spin-flip technique.This presumes that any track-assigned momentum – once found and fitted – does not dependon any rate-related background; it seems to be a good working assumption.

We studied energy-sorted counting method and estimated statistical uncertainty. To demonstrate it,histograms of the number of positrons are separated into several energy bins. Each histograms arefitted and the resulting uncertainties on ωa are combined statistically. Figure 28 shows this result. Asincreasing the number of energy bins, the statistical uncertainty on ωa is reduced. We are going toadopt this method as an alternative counting method (see Section 12.1.5 of TDR).

8.7 Number of vanes

Recommendation-36: It was not clear to the committee that the number of vanes inthe proposed detector configuration was optimized. Given the high cost of the tracker, avalue engineering exercise should be carried out. There are many other implications besidescosts; e.g., heat load, data load, current draw.

We have reviewed the number of vanes in various aspects since the focused review. Figure 29 shows theestimated track finding efficiency as a function of the number of vanes. In terms of track reconstructionefficiency, we need at least 24 vanes to have efficiency above 80% in the highest pileup condition. Wewould like to realize 48 vanes at which more than 90% efficiency is expected even in the highest pileupcondition. Other quantities for 48-vane detector are summarized in Table 5 and they are manageable.

Design of 48-vane detector has been started and potential issues were identified (weight and mechanicalsupport of thin vanes). We build 24 vanes first to gain experiences. Then, we will extend to 48 vanesto complete the detector construction.

Responses to recommendations given by the focused review committee, Dec. 2017

Yamanaka + Sato

32

used forg-2 analysis

Simulation

Reconstruction of muon beam distribution

33300− 200− 100− 0 100 200 300

x [mm]

300−200−

100−0

100200

300y [mm]

250−200−150−100−50−

050

100150200250

z [m

m]

300− 200− 100− 0 100 200 300x [mm]

300−

200−

100−

0

100

200

300

y [m

m]

0 1 2 3 4 5 6 [rad]φ

250−

200−

150−

100−

50−

0

50

100

150

200

250

z [m

m]

Fig. 13 All reconstructed hits from 25 muon decays obtained from simulation projected

onto horizontal (x-y) plane (bottom left) and in φ-z plane (bottom right), and perspective

view in three-dimensional space (top) are shown. There are two positron tracks in the energy

range of 200 MeV< E <275 MeV. Track candidate hits are shown by colored dots and the

other hits are shown by white dots. Reconstructed track orbits are shown by colored curves

(top and bottom left) and straight lines for track finding are shown (bottom right).

17/22

Y. Sue, et al., JPS meeting (Sep 2018)Muon beam (truth)

������

������

������

e+ trackback

������

after unfolding

MC truthtrack backafter unfolding

������

Simulation data

34

B-field shimming test with the MuSEUMmagnet (1.7 T) at J-PARC

No shim�2015/05/19!

3!

!  860 ppm p-p�

in 50 cm DSV �

Field shimming by iron arraysAfter shimming #3�

2015/05/19!

!  Measured by single probe system !  Spheroid : r=100 mm, z=300 mm�

22! 35

After shimming

Before shimming

Plots by K. Sasaki

B(μT)

azimuthal angle (rad)

axial position (mm)

r = 140 mm

B(μT)

azimuthal angle (rad)

axial position (mm)

400 ppm

0.4 ppm

-600 ppm

-0.5 ppm

Cross calibration of B-field probes

36ANL, March 27 2018

J-PARCProbe

FermilabProbe

Workshop on muonium

37

Osaka University, Dec 10-11, 2018http://www-kuno.phys.sci.osaka-u.ac.jp/muonium2018/

• Topics• BSM in muonium (M. Yoshimura, M. Endo)• Muonium HFS (M. Eides, D. Nomura, K. Shimomura)• Muonium 1S-2S (Y. Liu, P. Crivelli, S. Uetake)• Muonium Lamb shift (A. Olin)• Muonium for cLFV, g-2/EDM …

Summary• The J-PARC muon g-2/EDM experiment will measure g-2 and

EDM with completely different method.

• Features:• Low emittance muon beam (1/1000)• No strong focusing (1/1000)• Good injection eff. (x10)• Compact storage ring (1/20)• Tracking detector with large acceptance• Completely different from BNL/FNAL method

• We are moving to R&D to construction phase.• In coming years,

– Construction of muon beamline (H-line)– Demonstration of “muon cooling”– Re-acceleration to 1 MeV 38