Cold and slow molecular beams: Application to electron EDM

Report
Cold and slow molecular beams:
Application to electron electric
dipole moment (EDM)
measurements
Katsunari Enomoto, Univ. of Toyama
Fundamental Physics Using Atoms 2010
2010/Aug/9
Osaka U.
Electron electric dipole moment
spin
electron
spin
EDM
T, P
EDM
related with CP, T violation physics
Standard model
electron EDM de  10-38 e cm
SUSY, left-right, multi-Higgs
de < 10-24 e cm
Experiment (Tl atomic beam)
de < 10-27 e cm
PRL 88, 071805 (2002).
Table-top experiment for the physics beyond the standard model.
EDM measurement using atoms
E // B
Typical atomic beam method
E // B
or
S
precession
Eappl0.1 MV/cm
B1 nT
 t
Due to the relativistic effect, heavy atoms have large enhancement factor R.
(Cs: R110, Tl: R590, Fr: R1150)
h
B
Eeff = R Eappl  0.1 GV/cm,   de Eeff / h  10 Hz
m=1/2 m=1/2
False EDM signal (systematic error)
Leak current loop
v  E induced field
v
I
E
EDM measurement using molecules
atom
molecule
Induced dipole
|p
Eeff
elec.
mix
with
Eappl
permanent dipole
|J=1
Emol
rot.
|J=0
mix
with
Eappl
|s
Eeff = R Eappl  0.1 GV/cm,
with Eappl  0.1 MV/cm
Eeff = P Emol  10100 GV/cm,
P  1 with Eappl  0.01 MV/cm
(Eappl is needed just for aligning molecule)
Sensitivity  1001000
Systematic error  0.1
Atoms vs molecules
Tl beam experiment
PRL 88, 071805 (2002).
de < 10-27 e cm
Why is it not so good?
Vibration (1000 K)
Rotation
( 1K)
YbF beam experiment
vs
PRL 89, 023003 (2002).
de < 10-25 e cm
…. because radical molecular beams
are difficult to produce,
and molecules have
many internal levels
(especially rotation).
Cold (large population in the ground state) and slow (long interaction time)
molecular beam will improve greatly the sensitivity.
In this talk, after reviewing cold molecule experiments,
I will present our recent results and ongoing projects.
Ultracold molecules
Ultracold molecules are one of the hottest topics in
atomic/molecular/optical (AMO) physics in this decade.
High resolution
spectroscopy
New
condensed matter
Test of fundamental
physics
Quantum simulator
Direct
cooling of
molecules
(mK)
Ultracold
chemistry
Control of chemical
reaction
Laser cooling
of atoms and
associating to
molecules (nK)
Direct cooling methods (1)
Supersonic expansion is a conventional method for molecular spectroscopy,
and it generates cold (1 K) but fast (supersonic) molecular beams.
How to slow down?
Stark decelerator & electrostatic trap
Bethlem et al., Nature 406, 491 (2000).
Counter-rotating nozzle
Gupta et al,
J. Phys. Chem. A 105, 1626 (2001)
Direct cooling methods (2)
Laser ablation can generate molecular gases in
cryogenic helium gas (1 K).
Effusive molecular beam
Maxwell et al., Phys. Rev. Lett. 95, 173201 (2005)
Buffer-gas cooling & magnetic trap
Weinstein et al., Nature 395, 148 (1998)
Hydrodynamically enhanced-flux
(but boosted to 160 m/s)
molecular beam
Patterson et al., J. Chem. Phys. 126, 154307 (2007)
Control of translational motion
Now, molecules can be cooled/decelerated down to 1 K.
Many tools are available to control molecular translational motion,
e.g. electric & magnetic static field, optical field, …
Our approach: using microwave field
Stark shift of diatomic molecules
Advantage of microwave:
DeMille et al, Eur. Phys. J. D 31, 375 (2004)
Energy
High-field-seeking (HFS)
ground state can be trapped.
LFS
J=1, m=0
HFS
J=1, |m|=1
J=m=0
HFS state cannot be trapped with
static fields due to Earnshaw’s theorem.
Electric field
Microwave trap for molecule
It has been proposed to a microwave field enhanced in a Fabry-Perot
cavity to trap polar molecules.
For static field (dc Stark shift)
(J=0,1 states)
 2B
dE / 3 


H 

dE
/
3
0


DeMille et al, Eur. Phys. J. D 31, 375 (2004)
Electric field E  (P  Q)1/2
For microwave field (ac Stark shift)


dE / 2 3 
r


H 

0
 dE / 2 3

2B: rotational splitting
: detuning
d : dipole moment of molecule
Assuming power P 2 kW, quality factor Q 105,
Electric field E  30 kV/cm ( 3 K trap depth) is possible.
Microwave Stark decelerator
We proposed that HFS state molecules can be decelerated
by using time-varying standing wave of microwave.
Enomoto & Momose, PRA 72, 061403 (2005)
Current plan: to use circular waveguide resonator TE11 mode
TE11
Potential
w/ microwave
Alternate gradient focusing decelerator
Bethlem et al., PRL 88, 133003 (2002)
Tarbutt et al., PRL 92, 173002 (2004)
More powerful,
but dynamical radial confinement
Radial confinement
for HFS state
Potential
w/o microwave
Simulation of deceleration
L
1000
Molecule :
Initial velocity : 21 – 24 m/s
Center molecule : 22.5 m/s (5.8 K)
Deceleration : 93 cm, 80 ms
P[W]  Q : 107
Molecules
174YbF
800
600
400
200
0
23.0
0.2
0.4
0.6
0.8
1.0
Distance (m)
20
40
Time (ms)
60
80
Initial velocity (m/s)
Velocity (m/s)
0
25
20
15
10
5
0
0.0
25
20
15
10
5
0
0
5
10
15
20
25
Final velocity (m/s)
22.8
22.6
22.4
22.2
22.0
-4
-3
-2
Initial position (mm)
Microwave Stark decelerator can be used for molecular beams
pre-cooled to about 5 K.
-1
First experimental step: microwave lens
w/o microwave
Molecular beams can be
focused with a microwave
field.
w/ microwave
Performed in Fritz-Haber institute by using a decelerated NH3 beam
Odashima et al., PRL 104, 253001 (2010)
Next plan for microwave control
Electric field E2  (power P)  (quality factor Q)
High P needs expensive amps and causes heating.
So we are planning to use a superconducting cavity for high Q.
(Q factor is mainly determined by the surface resistance.)
Lens exp.
(Cu cavity)
Power P[W]
3
SC cavity
(Nb or Pb/Sn)
<3?
Q-factor
5000
3106 ?
PQ
1.5 104
107 ?
Limited by cooling power
(Note that only 0.1 s is
needed for deceleration.)
Q > 106 is typically easily
obtained, but we have to
rapidly switch microwave.
This limits the Q factor.
We will test the superconducting cavity soon in U. British Columbia
(Momose lab.)
Project in Univ. of British Columbia
We are constructing a Stark decelerator in UBC.
We will combine the Stark decelerator with superconducting cavity.
Testing a microwave resonator
Firstly, we tested a copper resonator
with a loop antenna.
loop antenna
Transmission
room temperature
FWHM = 2.92 MHz
0.8
0.6
0.4
Cool down
with L.N2
Q factor
 3
1.0
QL  16000
L. N2 temperature
FWHM 1.07 MHz
0.8
Transmission
QL  5000
1.0
0.6
0.4
0.2
0.2
0.0
785
790
795
Frequency - 14000 (MHz)
0.0
680
685
690
Frequency - 16000 (MHz)
We will test a Pb/Sn-coated superconducting cavity soon.
Project in Univ. of Toyama
We are making cold molecular beams based on He buffer-gas cooling.
L. He bath
He gas
line
To mass
spectrometer
Laser ablation
(pulsed green laser)
exit hole
sorption
pump
We have observed Pb and O atoms produced
by laser ablation of a PbO target with mass spectrometer.
EDM measurement project
We are starting the EDM measurement project in Univ. of Toyama
from this year.
Only the project plan is presented here.
What molecules?
How to produce molecules?
How to cool them to a few K?
How to enhance the flux?
What more?
Choice of molecule
unpaired
electron
Heavy atom
Large
electronegativity
To obtain high beam flux in a single internal state
Low boiling point (even for laser ablation)
Small nuclear spin (simple hyperfine structure)
large natural abundance
From experimental point of view
Less toxic
Not radioactive
Tentative plan: to use YbF (like E. Hinds group, Eeff = 26 GV/cm)
or BaF (Eeff = 8 GV/cm)
low
density and directionality
high
Cooling procedure
Supersonic jet
room T
Initial velocity is determined by
carrier gas
e.g. YbF in Xe 300 m/s corresponds to 1000 K for YbF
Hydrodynamic He buffer-gas-cooled beam
Initial velocity is determined by
He gas (160 m/s  300 K for YbF)
4 K, high He density
Effusive He buffer-gas-cooled beam
Initial velocity is determined by
the cell temperature ( 4 K)
4 K, low He density
We will use He buffer-gas-cooled beam close to effusive regime.
Improvement of flux
How to generate molecules?
Laser ablation
Injection from oven
oven
1012 /pulse
poor reproducibility
1015 /s ?
(like J. Doyle group)
How to improve directionality?
Microwave lens
Laser cooling
(SrF: Shuman et al., PRL 103, 223001 (2009).)
They also help isotope selection  suppression of background noise
Future possibility
Microwave deceleration and trap
Combination of alternate gradient decelerator and microwave decelerator
Conclusion
Microwave enhanced in resonators is available to control
molecular translational motion (such as deceleration and trap).
As a first step, we demonstrated the microwave lens.
Odashima et al., PRL 104, 253001 (2010)
We will test soon a high-Q superconducting resonator.
For electron EDM measurement, we are making He-buffer-gasbased cold molecular beam (YbF or BaF).
EDM measurement with molecular beams with cold molecule
technologies developed in this decade is promising.
Acknowledgments
Microwave lens experiment
H. Odashima, S. Merz, M. Schnell, G. Meijer (Fritz-Haber-Institut)
Superconducting cavity project
O. Nourbakhsh, P. Djuricauin, T. Momose,
W. Hardy and his students
(Univ. of British Columbia)
Buffer-gas cooled beam project
Y. Kuwata, H. Noguchi, H. Hasegawa, S. Tsunekawa,
K. Kobayashi, F. Matsushima, Y. Moriwaki (Univ. of Toyama)
And courtesy of D. DeMille
77K shield
分子ビーム
4K shield
チャコール
セル
マイクロ波定在波
TE11
TE01
マイクロ波定在波あり
マイクロ波定在波なし
Fabry-Perot
TEM00
光ポンピング
B state
Diode
J’=1
laser
J=1
X state J=0
マイクロ波
トラップ
pump
He
gas
L. He
Q-mass
PbO
pump
pulse
YAG
pump
pump
シュタルク
ガイド
LFS
HFS
Bethlem et al., PRA 65, 053416 (2002).
Stark
UBC
Microwave
Lens
deceleration
(collimation)
Buffer gas
Cold slow beam
Toyama
trap
EDM measuremen
Acknowledgment
FHI
UBC
Toyama
Atoms or molecules?
atom
vib.
molecule
Induced dipole
elec.
mix
with
Eappl
rot.
Large internal electric field
(Eeff  10 GV/cm)
Easy to handle
High electric field Eappl is needed
(causing systematic error)
Eeff  500 Eappl, Eappl  100 kV/cm
mix with Eappl
to align molecule
Rotation and vibration exist
(small population in the ground state
at room temperature, which reduce
statistical certainty)
Experiment (Tl atomic beam)
de < 10-27 e cm
Experiment (YbF molecular beam)
de <
PRL 88, 071805 (2002).
10-25 e cm
PRL 89, 023003 (2002).
Cold molecular beam (or trapped molecules) will improve much more.

similar documents