1. DISCRETE 2010
Rome, 6-10 December 2010 + -
Review of EDM experiments Yannis K. Semertzidis Brookhaven National Laboratory
Motivation
Neutron EDM experiments
Mercury EDM 199Hg
Electron EDM experiments
Storage Ring EDM experiments (proton & deuteron) …
Many thanks to Antonio Riotto and Josh Long for excellent talks on the subject

2. Physics at the Frontier, pursuing two approaches:
Energy Frontier
(FNAL, LHC,…)
Precision Frontier
(gμ-2, flavor, EDM,…)
which are complementary and inter-connected. The next SM will emerge with input from both approaches.

3. Physics reach of magic pEDM (Marciano)
Sensitivity to new contact interaction: 3000 TeV
Sensitivity to SUSY-type new Physics:
The proton EDM at 10-29e∙cm has a reach of >300TeV or, if new physics exists at the LHC scale, < rad CP-violating phase; an unprecedented sensitivity level.
The deuteron EDM sensitivity is similar. 3

4. Purcell and Ramsey: “The question of the possible existence of an electric dipole moment of a nucleus or of an elementary particle…becomes a purely experimental matter”
They knew it would violate
Parity.
Phys. Rev. 78 (1950)

5. Short History of EDM
1950’s neutron EDM experiment: a search for parity violation (before the discovery of P-violation)
After P-violation was discovered it was realized EDMs require both P,T-violation
1960’s EDM searches in atomic systems
1970’s Indirect Storage Ring EDM method from the CERN muon g-2 exp.
1980’s Theory studies on systems (polar molecules) with large enhancement factors
1990’s First exp. attempts w/ polar molecules. Dedicated Storage Ring EDM method developed
2000’s Proposal for sensitive srEDM exp. approved; Magn. Opt. Traps developed, progress w/ polar mol.

6. In Quantum Mechanics: a non-degenerate system with Spin is defined by the spin vector+ -
If the particle has an EDM, its vector needs to be aligned with the spin vector, locked to its direction, i.e. it needs to choose either along or opposite but not both (non-degenerate). “CP-Violation Without Strangeness”, Khriplovich/Lamoreaux.

7. A Permanent EDM Violates both T & P Symmetries:+ - + - T P + -

8. A Permanent EDM Violates both T & P Symmetries:
EDM physics in degenerate systems is not important
(batteries are allowed!)
Yannis Semertzidis, BNL. ICHEP02, 31 July 2002

9. T-Violation CP-Violation
CPT
Andrei Sakharov 1967:
CP-Violation is one of three conditions to enable a universe containing initially equal amounts of matter and antimatter to evolve into a matter-dominated universe, which we see today….
Yannis Semertzidis, BNL. ICHEP02, 31 July 2002 9

10. CP-violation was discovered at BNL in 1964

11. CP-violation is established
The SM CP-violation is not enough to explain the apparent Baryon Asymmetry of our Universe by ~10 orders of magnitude (see excellent talk by Antonio Riotto).
A new, much stronger CP-violation source is needed to explain the observed BAU.

12. The great mystery in our Universe: matter dominance over anti-matter.
EDMs could point to a strong CP-violation source capable of creating the observed asymmetry.

13. EDM Searches are Excellent Probes of Physics Beyond the SM:
Most models beyond the SM predict values within the sensitivity of current or planned experiments:
SUSY
Multi-Higgs
Left-Right Symmetric …
The SM contribution is negligible…
Yannis Semertzidis, BNL
Yannis Semertzidis, BNL. ICHEP02, 31 July 2002

14. The Electric Dipole Moment precesses in an Electric field
The EDM vector d is along the particle spin direction + -
Yannis Semertzidis, BNL

15. A charged particle between Electric Field plates would be lost right away.- + +
Yannis Semertzidis, BNL. ICHEP02, 31 July 2002 15

16. Neutron EDM Vs Year Purcell and Ramsey started a long effort
“…at 3 x e cm, it is analogous to the Earth's surface being smooth and symmetric to less than 1 µm” (John Ellis).
Yannis Semertzidis, BNL. ICHEP02, 31 July 2002 16

17. Spin precession at rest+ - E
Compare the Precession Frequencies
with E-field Flipped:
Caution is needed applying this equation to obtain
the statistical accuracy…
Yannis Semertzidis, BNL

18. Measuring an EDM of Neutral Particles
H = -(d E + μ B) ● I/I d E B
mI = 1/2 ω2 ω1
mI = -1/2
d = e cm
E = 200 kV/cm
δw = 10-7 rad/s 
~1 turn/year 

19. Main Systematic Error: particles have non-zero magnetic moments!
For the nEDM experiments a co-magnetometer or SQUIDS are used to monitor the B-field: cancellation level needed for 10-28e-cm is of order 3pG. (See Josh Long’s talk for application of 199Hg co-magnetometer in the nEDM.)

20. Important Stages in an EDM Experiment
Polarize:state preparation, intensity of beams
Interact with an E-field:the higher the better
Analyze:high efficiency analyzer
Scientific Interpretation of Result! Easier for the simpler systems
Yannis Semertzidis, BNL
Yannis Semertzidis, BNL. ICHEP02, 31 July 2002

21. EDM methods (they all are sensitive to different combinations of CPV sources)
Neutrons: Ultra Cold Neutrons, apply large E-field and a small B-field. Probe frequency shift with E-field flip
Atomic & Molecular Systems: Probe 1st order Stark effect
Storage Ring EDM for charged particles: Utilize large E-field in rest frame-Spin precesses out of plane (Probe angular distribution changes)

22. EDM method Advances
Neutrons: advances in stray B-field effect reduction; higher UCN intensities (~106/cycle)
Atomic & Molecular Systems: high effective E-fields (up to ~1-100GV/cm)
Storage Ring EDM for d, p: High intensity polarized sources well developed (108/cycle); High electric fields made available; spin precession techniques in SR well studied

23. EDM method Weaknesses
Neutrons: Intensity; High sensitivity to stray B-fields; Motional B-fields and geometrical phases
Atomic & Molecular Systems: Low intensity of desired states; in some systems: physics interpretation
Storage Ring EDM: sensitive to vertical E-fields or radial B-fields; some systematic errors different from g-2,…

24. EDM Experiments
Statistics is of the essence (number of neutrons, spin coherence time,…)
Systematics is close second (magnetic fields, geometrical phases,…)

25. nEDM experiment at ILL (Grenoble):
Yannis Semertzidis, BNL
Yannis Semertzidis, BNL. ICHEP02, 31 July 2002

26. Yannis Semertzidis, BNL
New UCN experiment at ILL: Expect a factor of ~100 improvement in sensitivity due to
Neutrons in 0.5 K
He bath
~50 more neutrons
E-field: 4-6 at cryo temp.
Longer coherence times
Yannis Semertzidis, BNL
Yannis Semertzidis, BNL. ICHEP02, 31 July 2002

27. Neutron EDM at PSI

28. SNS nEDM Experiment - Vertical Section View
Lower Cryostat
Upper Cryostat
~6 m
DR LHe Volume
~450 liters
Upper Cryostat Services DR
Central LHe Volume
~400 mK, ~1000 liters
Reentrant Neutron Guide
ABS Line
3He Injection Volume
4-layer
m-metal shield
Cosq magnet
ABS
The current value is < 3 x e•cm (90% C.L.)
Hope to obtain roughly < 8 x e•cm with UCN in superfluid He

29. Strength: Some systematic errors are different
from ILL and PSI experiments.
Applying spin dressing techniques to equalize and further reduce
the stray B-field sensitivity (see talk by Josh Long).

30. Yannis Semertzidis, BNL
Neutron EDM Timeline
ILL, n
1E-24
PSI, n
SNS, n
1E-26
1E-28
EDM Limits [ecm]
1E-30
1E-32
Future
Yannis Semertzidis, BNL
Yannis Semertzidis, BNL. ICHEP02, 31 July 2002 30

31. An electron or a nucleus in a neutral atom feels a zero average E-field:+
The atom is polarized and
the average total E-field
is zero: - +
(Schiff theorem):
Exceptions to the rule: a) relativistic effects b) size effects, …
Yannis Semertzidis, BNL. ICHEP02, 31 July 2002 31

32. Schiff Theorem: A Charged Particle at Equilibrium Feels no Force… …An electron in a neutral atom feels no force either. However, the average interaction energy is not zero because the EDM in the lab frame is velocity dependent
E. Commins et al., Am. J. Phys. 75 (6) 2007

33. The mercury EDM experimental results published last year

34. The apparatus and parameter values
B=22 mG
V=±10 KV, E=~2 MV/m (height of cell ~1cm)
SCT = 102 s

35. The data
The drift in frequency is taken out by taking the frequency difference between the cells.
Runs with micro-sparking are taken out.

36. Systematic errors
The systematic error is ~60% of the statistical error

37. The results and best limits
It now dominates the limits on many parameters
They expect another improvement factor ~3 - 5.

38. Storage Ring EDM experiments
(or how to create a Dirac-like particle in a storage ring) 38

39. A charged particle between Electric Field plates would be lost right away…- + +
Yannis Semertzidis, BNL. ICHEP02, 31 July 2002 39

40. …but can be kept in a storage ring for a long time
Yannis Semertzidis, BNL
Yannis Semertzidis, BNL 40

41. Yannis Semertzidis, BNL
The sensitivity to EDM is optimum when the spin vector is kept aligned to the momentum vector E
Momentum
vector
Spin vector
Yannis Semertzidis, BNL
Yannis Semertzidis, BNL 41

42. Yannis Semertzidis, BNL
The spin precession relative to momentum in the plane is kept near zero. A vert. spin precession vs. time is an indication of an EDM (d) signal. E
Yannis Semertzidis, BNL
Yannis Semertzidis, BNL 42

43. Yannis Semertzidis, BNL
The spin precession relative to momentum in the plane is kept near zero. A vert. spin precession vs. time is an indication of an EDM (d) signal. E E E E
Yannis Semertzidis, BNL
Yannis Semertzidis, BNL 43

44. Freezing the horizontal spin precession
The spin precession is zero at “magic” momentum (0.7 GeV/c for protons, 3.1GeV/c for muons,…)
The “magic” momentum concept was first used in the last muon g-2 experiment at CERN and BNL.

45. When P=Pmagic the spin follows the momentum
No matter what the E-field value is the spin follows
the momentum vector creating an ideal Dirac-like
particle (g=2)
Yannis Semertzidis, BNL
Yannis Semertzidis, BNL 45

46. There is an asymmetry between E and B-fields and g-2 precession
Independent of velocity!
E-fields:
Depends on velocity

47. Freezing the horizontal spin precession for the proton a=1
Freezing the horizontal spin precession for the proton a=1.79, and deuteron a=-0.143
For particles with negative anomalous magnetic moment there is no …magic.
A dipole B-field is required to cancel the g-2 precession in the deuteron case

48. Storage Ring EDM experiments:
High beam intensities ( ), with high polarization (>0.8), and low emittance are currently available
Large electric fields are possible (10-20MV/m)
Spin coherence time ~103s are possible
High efficiency, large analyzing power (~0.5) polarimeters are available for the proton and deuteron at ~1 GeV/c momentum making possible the next sensitivity level
Direct access to charged particle EDMs 48

50. EDMs of hadronic systems are mainly sensitive to
Theta-QCD (part of the SM)
CP-violation sources beyond the SM
A number of alternative simple systems could provide invaluable complementary information (e.g. neutron, proton, deuteron,…).

51. EDMs of different systems
Theta_QCD:
Super-Symmetry (SUSY) model predictions:

52. Two different labs to host the S.R. EDM experiments
BNL, USA:
proton “magic” ring
COSY/IKP, Germany: deuteron ring

53. Physics strength comparison (Marciano)
System
Current limit [ecm]
Future goal
Neutron equivalent
Neutron
<1.6×10-26
~10-28
10-28
199Hg atom
<3×10-29
<10-29
129Xe atom
<6×10-27 ~
Deuteron nucleus
~10-29
3× ×10-31
Proton nucleus
<7×10-25
10-29 53

54. Yannis Semertzidis, BNL
Hadronic EDM Timeline
ILL, n
1E-24
PSI, n
SNS, n
1E-26
BNL, p
COSY, d
1E-28
EDM Limits [ecm]
199Hg
1E-30
129Xe
129Xe,
Rn, Ra
1E-32
Future
Yannis Semertzidis, BNL
Yannis Semertzidis, BNL. ICHEP02, 31 July 2002 54

55. Electron EDM experiments
Current limit: 1.6×10-27 (90%), from Tl (paramagnetic) atom
Future:
Paramagnetic atoms: Fr, Cs
Polar molecules: YbF, PbO, ThO, PbF, HBr, BaF, HgF, ThO, …
Solids: GGG, …
unpaired
electron －　 +
paramagnetic
atom
diamagnetic
internal electric field　
10～100 GV/cm
paramagnetic molecule (radical) Fr Sr

56. Yannis Semertzidis, BNL
Electron EDM Timeline
Tl, e-
1E-24
Polar
mol.
1E-26
1E-28
EDM Limits [ecm]
1E-30
1E-32
Future
Yannis Semertzidis, BNL
Yannis Semertzidis, BNL. ICHEP02, 31 July 2002 56

57. Sensitivity to Rule on Several New Models
Baryogenises
Electroweak
GUT SUSY
Gray: Neutron
Red: Electron
n current
e current
n target
p, d target
e target
e-cm
(Adopted from J. Long’s slide)
J.M.Pendlebury and E.A. Hinds, NIMA 440 (2000) 471

58. Summary The 199Hg Exp. sets many of the current limits
Better sensitivity is expected within 5-10 years for 199Hg, 129Xe, Fr, Rn, Ra, YbF,… and neutron
The proton EDM experiment for 10-29e-cm has been recently approved at BNL and preparing a proposal for DOE-NP. Results >2015.
By the end of the decade we could have a number of new improved results and either we are discovering or setting severe constraints on BSM CP-Violation. If found, they will help explain the large BAU mystery. In any case we’ll have a lot of fun and learn a great deal!

59. Brookhaven National Laboratory
Home of the only polarized proton collider in the world up to 500 GeV on 500 GeV
Long experience with polarized sources: a few times 1011 protons with ≥80% polarization
Our needed proton beam parameters are readily available
Yannis Semertzidis, BNL
Yannis Semertzidis, BNL. ICHEP02, 31 July 2002

60. Clock-wise (CW) & Counter-clock-wise (CCW) storage60

61. Experimental needs Two Proton bunches 0.7 GeV/c 90% polariz.;
<100m base length
Rep. rate: 103s
Beam energy: ~1J
Average beam power: ~1mW
Beam emittance:
95%, unorm.
Horizontal: 3mm-mrad
Vertical: 10mm-mrad
(dp/p)rms~ 2.5×10-4
CW & CCW injections: Average beam parameters: same to ~10% 61

62. E-field plate module: The (26) FNAL Tevatron ES-separators would do
Beam position
0.4 m
3 m
Vertical plates are placed everywhere around the ring to minimize vertical electric/radial B- fields from image charges

63. pEDM polarimeter principle: probing the proton spin components as a function of storage time
Micro-Megas TPC detector
and/or MRPC
“defining aperture”
polarimeter target
extraction adding white noise to slowly increase the beam phase space
carries EDM signal
increases slowly with time
carries in-plane precession signal

64. Is the polarimeter analyzing power good at Pmagic? YES!
Analyzing power can be further optimized
Yannis Semertzidis, BNL

65. Certain (main) systematic errors easier to handle if CW & CCW is done at the same time (Coincident BeamS: CBS)
In a ring with Electric field bending it is possible to store protons CW & CCW at the same time in the same place

66. OSCILLATING VERTICAL FOCUSING OSCILLATES SIGNAL AT SAME FREQUENCY

67. The EDM signal: early to late change
Comparing the (left-right)/(left+right) counts vs. time we monitor the vertical component of spin
M.C. data
(L-R)/(L+R) vs. Time [s]

68. Proton Statistical Error (230MeV):
p : 103s Polarization Lifetime (Spin Coherence Time)
A : Left/right asymmetry observed by the polarimeter
P : Beam polarization
Nc : 21010p/cycle Total number of stored particles per cycle
TTot: 107s Total running time per year
f : 0.5% Useful event rate fraction (efficiency for EDM)
ER : 17 MV/m Radial electric field strength (65% azim. cov.)
Yannis Semertzidis, BNL 68

69. Sensitivity to Rule on Several New Models
Baryogenises
Electroweak
GUT SUSY
Gray: Neutron
Red: Electron
p current
n current
e current
n target
p,d target
e target
e-cm
J.M.Pendlebury and E.A. Hinds, NIMA 440 (2000) 471

70. Technically driven pEDM Timeline07 08 09 10 11 12 13 14 15 16 17
Spring 2008, Proposal to the BNL PAC
Fall 2009 Conceptual Technical Review at BNL
December 2009, the pEDM experiment is approved
R&D phase; ring design
Fall 2012, Finish R&D studies: a) Develop BPMs, 10 nm, 1 Hz BW resolution, <1pm syst.
b) spin/beam dynamics related systematic errors c) Polarimeter detector development and prepare for testing
d) Finalize E-field strength to use
e) Establish Spin Coherence Time, study systematic errors, optimize lattice
FY 2013, start ring construction (two years) 70

71. From the September 2008 run at COSY
Polarimeter team
Resonance crossing
(full spin flip)
Vector asymmetry V+ T−
Unp V− T+
Slowly extracting the beam while monitoring
its polarization as a function of time

74. EDMs: New CPV? sinCP ~ 1 -> M > 5000 GeV
SM “background” well below new CPV expectations
New expts: 102 to 103 more sensitive
CPV needed for BAU?
3.1 x 10-29
Mass Scale Sensitivity
sinCP ~ 1 -> M > 5000 GeV
M < 500 GeV -> sinCP < 10-2

75. Main polarimeter systematic errors
Polarimeter team
Main polarimeter systematic errors

76. Gas cluster ion beam surface treatment

77. AGS Complex at BNL AGS m g-2 experiment NSRL Linac TTB pEDM @ 17 MV/m
pEDM Ring
30.8m
Linac
Booster
AGS
100 m
TTB
C-AD Admin

78. Storage Ring EDM Polarimeter Development Polarimeter
Technical Review – 12/7/2009
Polarimeter Development
Polarimeter
(Half) Polarimeter in the ring:
Absorber to remove
low analyzing power
particles. (Detector
choice can also give
discrimination.)
Quadrupoles here
are larger aperture
for clearance.
5° to 20°
acceptance
Equal rate readout pads
One target is
shown. We
want a target
available from
at least the
left, right, up
and down
directions. cm
Generic detector:
(?) Multi-resistive plate chamber
(?) Micro-megas
(?) Gas electron multiplier
(?) …other
Rate = 800 /s/pad
In one store:
Edward J. Stephenson, IUCF 15

79. Storage Ring EDM Polarimeter Development COSY tests COSY ring:
Technical Review – 12/7/2009
Polarimeter Development
COSY tests
COSY ring:
EDDA detector:
Rings and bars to determine angles.
Use EDDA
detector UP
LEFT
DOWN
RIGHT
Azimuthal angles yield two asymmetries:
Edward J. Stephenson, IUCF 4

80. Storage Ring EDM Polarimeter Development Target Concept
Technical Review – 12/7/2009
Polarimeter Development
Target Concept
Target solution found at COSY:
expanded vertical phase space
beam core
end
view
White noise
applied to
electric field
plates.
maximum
allowed at
COSY
15 mm
Do enough particles penetrate far enough into the front
face to travel most of the way through the target?
This requires a comparison of the efficiency with model values.
Edward J. Stephenson, IUCF 5

81. Polarimeter rates:
Beam intensity with 2×1010 pol. protons/ ~103s and a detection efficiency of 1%  200KHz for ~3000cm2 area, or ~100Hz/cm2 on average but much higher at small radius. Design: ~1KHz/pad.
70 cm

82. Applying Micro-Megas TPC
Differentiate between protons and deuterons
Have a pointing accuracy of 0.5mrad/event
Yannis Semertzidis, BNL

83. Electric Dipole Moment
EDM T + +
T violation
~ CP violation under CPT invariance
Electric Dipole Moment ー ー
Standard Model
Lowest order where EDM appears
Estimation ~ e・cm
Small SM background
EDM～sensitive to the phenomena beyond the standard model
Super Symmetry

84. Francium EDM Fr Tl Xe Cs Rb
Small e-EDM ~ magnified in the heavy paramagnetic atoms Fr Rb Xe Cs Tl Fr
Enhancement factor
Francium :
Heaviest alkali element
e-EDM enhancement ~ largest ~ 1150
Atomic structure ~ simple
Radioactive atom ~ τ~ 3 min.

85. EDM search with laser cooled atoms
1 nK
1 mK
1 K
1000 K
Room temp.
liq. 4He
liq. 3He
Laser cooling
Evaporative cooling
Temp.
BEC FD
Trap laser cooled Fr atoms with ~uK
→ long coherence time, realize the accurate measurement
Laser cooled and trapped atoms
Atomic beam
210Fr
Tl（world record）
1150
K: enhancement
585
~102 sec
t： coherence time
~10-3 sec
> 10-29
EDM (e・cm)
~10-27
Laser
Magneto-Optical Trap
Radioactive atom 210Fr with largest enhancement of e-EDM
Produce the 210Fr using nuclear fusion reaction
Long coherence time ～ laser cooled and trapped 210Fr atoms
Achieve the high accuracy measurement

86. e-EDM experiment with atom in the world
EDM of stable alkaline atoms
Cesium Fountain EDM experiment
(Harvey Gould et al., LBNL, California, USA)
Projected accuracy de~1×10-29 e cm
EDM experiment of Cs and Rb in 1-D Optical Lattice
(David Weiss et al., Penn. State Univ., USA)
Projected accuracy de~5×10-30 e cm
Cesium EDM Experiment using Optical Dipole Force Traps
(Heinzen et al., University of Texas, Austin, USA)
Projected accuracy de~10-29 e cm
Cesium EDM Experiment using Optical Lattice
(K. Honda, Tokyo Institute of Technology, Tokyo, Japan)
Many activities in the world ….

87. Possibility of ultracold polar molecules
e-EDM sensitivity
unpaired
electron －　 +
paramagnetic
atom
diamagnetic
internal electric field　
10～100 GV/cm
paramagnetic molecule (radical) Fr Sr
P : polarization of the system
Eeff : internal electric field
t : interaction time
N : number of molecules
T : integration time
Preparation method of ultracold molecules
・Direct laser cooling of molecules is difficult.　
（interaction time of molecular beam t is short, ～1 ms or ms）
・We laser cool and trap Fr atoms and Sr atoms, then associate molecules
from Fr and Sr mixture through magnetically tunable Feshbach resonance/
photoassociation technique.
　　　　 longer interaction time t ～ 1 s in trapped ultracold polar molecules
～10-29 e 1 day integration
if Eeff = 10 GV/ cm, N = 104

88. Coil and Shield Nesting
Inner-Dressing & Spin-Flip Coil
50K Shield
Outer Dressing Coil
4K Shield
Superconducting Lead Shield
Ferromagnetic Shield
B0 cosθ Magnet

89. EDMs: What We May Learn Present n-EDM limit Proposed n-EDM limit
Slide by M.J. Ramsey-Musolf
EDMs: What We May Learn
Present n-EDM limit
Proposed n-EDM limit
Refined theory ? New theory ? Leptogenesis ?
Matter-Antimatter
Asymmetry in
the Universe ?
“n-EDM has killed more theories than any other single experiment”

90. Yannis Semertzidis, BNL3
Deuteron EDM, UoR, 18 April, 2006
Yannis Semertzidis, BNL

91. Ramsey’s method of separated fields
Yannis Semertzidis, BNL
Yannis Semertzidis, BNL. ICHEP02, 31 July 2002

92. Matter-Antimatter Asymmetry in our universe
Pie chart of energy share of our universe
Matter-Antimatter Asymmetry in our universe
4% of our universe is made out of matter. Apparently this is too much according to SM.
The CP-violation observed within the SM can only account for mass equivalent to ~ galaxies compared to ~102 billion visible ones.
A new, much larger, source of CP-violation is needed; probably due to New Physics.

93. The great mystery in our universe: matter dominance over anti-matter