DeMille
Group
Dave DeMillePhysics Department
Yale University
The electric dipole moment of the electron
FundingNSF
ꞏ Electric dipole moments (EDMs) and new particle physicsꞏ How to detect an EDMꞏ Upper bound from the ACME experimentꞏ What it means, and where we go next
+ d =
CPT theorem T-violation = CP-violation
T
An EDM violates time-reversal symmetry
If all electrons look like this…
This does not existT not a good symmetry of nature
PurcellRamseyLandau
Pn.b.
inversion symmetry also broken by EDM
Q. How can a T-violating electron EDM arise?
A. From radiative corrections, with CP-violating interactions
e e
W t
bWc
W
Standard Model
4-loops for nonzero effectAND
severe cancellations in sum over all diagrams
[GIM mechanism]
CP-violating phasesfrom
CKM matrix
sg
e e
e e
~ ~
~
Supersymmetry1st order perturbation
ANDcancellations not built in
ßde de (Std. Model)
CP-violating phasefrom
SUSY-breaking
Q. How can a T-violating electron EDM arise?
A. From radiative corrections, with CP-violating interactions
Dimensional estimate for size of EDM
e
= -field
(new heavy particle)
gg
Dimensionless coupling constant T‐violating
phase
“natural” assumptionsg2/ħc sin() ~ 1
m ~ 1 TeVei
de ~ 30× current limit
(for N = 1 loop)
typical e-EDM
N = # loops
22
~2
sinN
ee B
g mdm
Artist’s impression of an electron EDM…
Note the resemblance to 0…
any EDM must be VERY small
QED radiative correction scale[/2] B/e ~10-13 cm
:Size of earth ~104 km
::
eEDM charge displacement:
<20 nanometers
General method to detect an EDM
Energy level picture:
Figure of merit: int1
1/ /coh
coh
shift d N Tresolution S N
+2d
+2d
Bw=2mB
w=2mB
-2d
-2d
S
Amplifying the electric field with a polar molecule
eff
Th+
O–
ext
Inside polarized molecule, eEDM acted on by eff ~ P 2Z 3e/a0
2 due to relativistic motion
Small energy splittings(rotational levels)
enables polarization ~ 100% with ext ~ 10 V/cm
Requires unpaired electron spin(s): chemical free radical
P. Sandars1965
eff 80 GV/cm for ThO* [near theoretical maximum]Meyer & Bohn (2008); Skripnikov, Petrov & Titov (2013, 2015); Fleig & Nayak (2014)
Advanced Cold‐Molecule Electron EDM
Yale University Harvard University
The ACME team 2009-2014BenSpaun
Chris Overstreet Nick Hutzler Cris Panda
Jacob Baron Paul Hess Elizabeth PetrikBrendon O’Leary
John Doyle DPD
Adam West
Emil Kirilov
Yulia Gurevich
Ivan Kozyryev
Max Parsons
New molecular beam technology:Cryogenic Buffer Gas-cooled Beam (CBGB)
Liquid helium bath
or pulse‐tube refrigerator
Beam brightness [=flux/divergence] ~ 103 ‐104 larger vs. other sources for refractory/free radical species
typically ~ 21011 mol/sr/state/pulse @ 10 Hz rep. rate
Used for SrF, ThO, BaF, YO, CaF, O2, NH3, Yb, …
• Inject hot molecules (e.g. via laser ablation)• Cool w/cryogenic buffer gas @high density• Efficient extraction to beam via “wind” in cell: 10‐4 10%‐40%
• “Self‐collimated” by extraction dynamics• Rotational cooling in expansion: T ~ 1 ‐ 4K•Moderately slow: v ~ 130‐180 m/s
[Maxwell et al. PRL 2005; Patterson & Doyle J Chem Phys 2007;Barry et al. PCCP 2011; Hutzler et al. PCCP 2011]
“New” molecular species: ThO* • Largest effective internal : ~80 GV/cm [Titov et al. 2013/2015, Fleig et al. 2014]
• Sufficient coherence time (~1.9 ms max) lifetime of metastable state H 31
• Suppressed magnetic moment<0.01 B in H 31 reduces B‐field systematics
[Idea: Meyer, Bohn, Cornell et al. (JILA);Measured: A.C. Vutha et al., PRA 2011]
• Omega‐doublet co‐magnetometersuppresses many possible systematics
• All spectroscopic data previously known
• State preparation and readoutw/standard, robust diode & fiber lasers
• Blue‐shifted fluorescence from probe laserno problem with backgrounds
• High beam source yieldground state
X 1
A 30
H 31
C 11
943 nmPump
1090 nm prep/probe
ThO
690 nm detect
[A.C. Vutha et al. J. Phys B 2010]
EDM measurement with -doublet states
m = 0ext
+
-eff
+
-
eff
+
-eff
+
-
eff
m = -1 S m = +1 SJ=1-
J=1+
+
- +
-+
+
- +
-
-
S S
S S
many dangeroussystematic errors cancel
in comparisonbetween
(near-)mirror image = 1 states
[DD et al., AIP Conf. Proc.
596, 2001]
= -1
= +1
ACME experimental schematic
ACME apparatusMolecular
BeamSource
Pulsed ablation
laser
Prep Lasers
Probe Lasers
Pulse Tube Cryo Frig
Interaction Region: internal -field plates & external
B-field coils (not shown here)
"
ITO Coated Borofloat glass
Molecule Beam E, B,
Transparent Field Plates
μ-metal shield endcaps(5 layers)
ACME apparatusMagnetic field coils
(3 orthogonal components& all first-order gradients)
.001 km
Completebeam source
& magnetic shields& last-stage optics
ACME apparatus
One of several optical tables w/~ten lasers, dozens of modulators,
hundreds of meters of optical fiber, etc.spread over two buildings
“control room”
Typical spin-rotation fringe signal
Take EDM dataat any of these points
B=0 ‐B
+Bsets = /2
( )cos 22f q= -
Contrast ~ 95%
Take data here
Final spin angle dvs. probe laser polarization angle
Spin‐rotation sig
nal
Primary machine settings to extract the EDM
--psuedo-random (pair-wise), interleaved reversals
Data sorting & systematic error analysis
Switch-correlated phases measure physical contributions:
0
1 1...
2 2nr
e eff N H leak N H nrd g gf m m+ D+µ D +
EDM
Symmetries ensure systematic errorsdue ONLY to experimental imperfections e.g.
--leakage current-induced -field leak --non-reversing -field nr
--etc.
Rewrite phase as components correlated w/switches:
EDM phaseSuperscipt means
“odd under this reversal” other phases to diagnose systematics
Spurious terms
0
1 1...
2 2
nr
e eff H lea rH nk Nd g gf m m+ D + +µ D
Data analysis: diagnosing imperfections
EDMExperimental imperfections
12 H leakgf mµ
0
12
nrN Hgf mµ D
Most imperfections also appear in other correlated phasesBUT GREATLY AMPLIFIED
/ ~ 1000N
g gD / ~ 1000nr
“Other” correlated phases diagnose imperfections
Switch-correlated phases isolate physical contributions:
Search strategy for systematic errorsSwitch-correlated phases isolate physical contributions:
0
1 12 2
nre eff N H leak N nrHd g gf m mµ D+ + +D ...
EDMExperimental imperfections
Strategies:
Change all possible parameters that shouldn’t affect EDM ( magnitude, magnitude, global polarization, etc. etc.)
but might couple to unanticipated imperfections
Deliberately amplify imperfections, understand any changes in correlated phases
But… what about terms we don’t anticipate?
stat 3.7 ∗ 10 cm
stat
ACME electron EDM data
Blind analysis: randomly chosen offset added to data until analysis complete
???
stat 3.7 ∗ 10 cm
stat
ACME electron EDM data
Blind analysis: randomly chosen offset added to data until analysis complete
Consistent with zero set upper limit
J. Baron et al.,Science (2014)
New upper bound on electron EDM from ACME
10-4110-4010-3410-3310-3210-3110-3010-2910-2810-2710-2610-25
de (ecm)
Heavy sFermions
Model
Left Right Symmetric
Accidental Cancellations
Berkeley(2002)
ExactUniversality
Seesaw Neutrino Yukawa Couplings
Lepton FlavorChanging
Approx. CP
SO(10) GUT
10-10-10-3410-3310-3210-3110-3010-2910-2810-2710-2610-25
de (ecm)
Standard Model
ExactUniversality
Approx. Universality
Alignment
~
Left Right Symmetric
Lepton FlavorChanging
Accidental Cancellations
SO(10) GUT
Approx. CP
Heavy sFermions
Seesaw Neutrino Yukawa Couplings
MultiHiggs
Split SUSY
ve SUSYïNa
Imperial(2011)
ExtendedTechnicolor
ACME(2013)
YalePbO*2013
-
--
• ~10-fold improvement• extends deep into expected ranges• limited by statistics
Impact of EDMs in particle physicsJ. Feng: “Naturalness and the status of SUSY”, Annu. Rev. Nucl. Part. Sci. (2013)
EDM results push SUSY scale into “unnatural” regions--potential for imminent discovery…?
“All of the constraints shown are merely indicative and subject to
significant loopholes and caveats”
ACME (electron), 2014Seattle
(199Hg)2016 199Hg
ACME
Reece & NakaiarXiv: 1612:09080
th
Diagrams withknown SM particles
can rule outnon-SM couplings
Example:CP-violating
Higgs-top couplingBrod et al.,
arXiv: 1310.1385
CP-odd/CP-even Higgs-top coupling <1% from ACME
LHCConstraint
ACME Constraint
What does the eEDM limit mean for particle physics?
Impact of EDMs in particle physics(A very generic view)
Plot courtesy of M. Reece (Harvard)
The ACME II team
Brendon O’Leary
John Doyle
Adam West
CrisPandaZack
Lasner
Daniel AngVitaly Andreev
Gerald Gabrielse
DD
Elizabeth Petrik
Jonathan Haffner
NickHutzler
(Caltech)
ColeMeisenhelder
XingWu
ACME II statistical sensitivity
ACME II statistical sensitivity
~10x improved EDM statistical sensitivity/unit timeSystematic error studies ~complete, final data taking underway
Near-future discovery potential with the electron EDM
ACME II
ACME Ultimate