1. Future electron EDM measurements using YbF
Ben Sauer

2. Recent electron EDM measurements
10-22
Tl experiment (2002)
de < 1.6 x e.cm (90% c.l.)
10-24
MSSM f ~ 1
YbF experiment (2011)
de < 1 x e.cm (90% c.l.)
10-26
Multi Higgs
Left -Right
MSSM f ~ a/p
10-28
ThO* experiment (2013)
de < 8.7 x e.cm (90% c.l.)
Predicted values for the electron edm de (e.cm)
10-30
10-32
10-34
10-36
Standard Model

3. An EDM experiment E
Precess
time T
Analyze
Polarize

4. Sensitivity of an EDM experiment
Uncertainty:
size of E field
coherence time
number of molecules
polarization
contrast

5. Why polar molecules? -hde E• hde  E Interaction energy
Analogous to magnetic dipole interaction -gem B.s but violates P&T
hde  E
Factor h includes both relativistic interaction Z3, and polarization
electric field
system containing
electron
© Imperial College London

6. YbF: really large internal field
Parpia
Quiney
Kozlov
Titov
18 GV/cm
15 GV/cm
(2011) 15
Effective Field |hE| (GV/cm) 10 5 5 10 15 20 25 30
Applied Electric Field (kV/cm)
© Imperial College London

7. ThO*: huge internal field
Effective field Eeff in YbF is 26 GV/cm
when molecule is fully polarized
For ThO* Eeff is about 84 GV/cm (factor of
3.2 more sensitive)
Mostly relativistic:
(also depends on structure)
ThO* can be fully polarized!

8. Comparing some atomic and molecular systems
YbF, 2011: |Eeff|= 14.5 GV/cm (h = 0.56)
|de|<1.0 x e.cm (90% c.l.)
Tl, 2002: |Eeff|= 72 MV/cm (Eeff = -582 Eapplied)
|de|<1.6 x e.cm (90% c.l.)
PbO*, 2013: |Eeff|= 25 GV/cm
|de|<1.7 x e.cm (90% c.l.)
Eu0.5Ba0.5TiO3, 2012:
|de|<6 x e.cm (90% c.l.)
ThO*: |Eeff| = 84 GV/cm (factor of 6 on 2011 YbF)
|de|<8.7 x e.cm (90% c.l.)

9. Upgrades since 2011 In total, a factor of 3 in sensitivity
3rd layer of magnetic shield
(less noise)
Longer inner magnetic shield
(reduce end effects)
Separate rf, high-voltage plates
(reduce end effects, higher voltage, less leakage)
1kW/1ms rf pulses
(reduce gradient effects from both movement and linewidth)
Longer interaction region
In total, a factor of 3 in sensitivity

10. Our plans for YbF More molecules - increase beam intensity
- better detection
Slower molecules
Made possible by new technology
- solid state lasers
- buffer gas beam sources

11. © Jony Hudson

12. A rough guide to YbF 552nm A 2P½ (v=0, N=0) F = 1 170MHz
mF = -1
mF = 0
mF = +1
170MHz
X 2S+ (v=0, N=0)
F = 0
mF = 0

13. YbF eEDM measurement Measure population in F = 0 E, B Precess Polarize
time T
Analyze
Measure population
in F = 0

14. Signal vs. magnetic phase
F=0 population
F = 1
© Jony Hudson
F = 0

15. Scheme increases population by a factor of 7, sensitivity by 2.6
More molecules: Initial pumping
Use cycling transition to optically pump molecules into ground rotational state.
(-)
F=0, 1 A2P1/2 (v=0, J=1/2)
Optical pumping (N=2 rotational state)
F=2+
Scheme increases population by a factor of 7, sensitivity by 2.6
F=1
N=2 (J=3/2, 5/2) (+)
F=3
F=2-
F=1+
rf mixing (~100 MHz)
N=1 (-)
F=2
Microwave mixing (14 GHz)
F=0
F=1-
F=1
N=0 (+)
F=0

16. More molecules: Better detection
Fluorescence detection is only about 0.7% efficient
Probe laser beam

17. More molecules: make them cycle
F=0
F=0, 1 A2P1/2 (v=0, J=1/2)
F=1
F=1+
F=1-
F=2
F=2+
F=2-
F=3
N=0 (+)
N=1 (-)
N=2 (J=3/2) (+)
(-)
Molecules cycle until they escape to v=1 vibrational state (14 photons/molecule)

18. F = 0 and F = 1 are the two output ports of the interferometer
A flaw: measuring the eEDM
F = 0
F = 1
Detector count rate
F = 0 and F = 1 are the two output ports of the interferometer
-B0 B0
Applied magnetic field

19. Shelve population in N=1
(-)
F=0, 1 A2P1/2 (v=0, J=1/2)
Sensitivity gain of 5
F=2+
F=1
N=2 (J=3/2) (+)
F=3
F=2-
F=1+
N=1 (-)
F=2
F=0
F=1-
F=1
N=0 (+)
F=0

20. High fidelity shelving
The problem is the YbF beam is larger than l at 14 GHz.
Cross section of simulated parallel plate transmission line.
plate
uniform (integrated) field
plate
microwaves in

21. High fidelity shelving
Transition probability over cross section of YbF beam
Average transition probability N=0 Þ N=1 over YbF beam >98%

22. Slow source of YbF
4K copper cell
4K helium flow
Nick Bulleid (PhD thesis, 2013)
YbF formed by laser ablation, cools to 4K, forward velocity is 150 m/s. Flux is similar to current beam (5x109 YbF /str/pulse)

23. Slow source of YbF
(20 sccm)
YbF signal
YbF velocity (m/s)
time after ablation (ms)
time after ablation (ms)
2011 supersonic beam had forward velocity of 600 m/s

24. “Traditional” YbF eEDM
Compared to 2011 measurement:
Factor 3 for longer plates
Factor 2 for N=2 population pumping
Factor 5 for cycling detection
Factor 4 for slower beam
Two orders of magnitude improvement is underway.
We have a lot of experience and a fairly sophisticated data analysis scheme, so should be able to control systematic effects.

25. Is there more?
YbF eEDM experiment takes place in the ground state, so why not coherence times of 1s?

26. Building the YbF fountain4K
Fantastically inefficient: 10-8 from cell to detector. But T = 300ms, so 60h of data gives sd = 3x10-31 e.cm!

27. The YbF team Mike Tarbutt Ed Hinds
Jony Hudson Joe Smallman Isabel Rabey
B.E.S Jack Devlin
YbF fountain:
James Bumby
James Almond
Jongseok Lim
Noah Fitch

28. Everything
clear?