Transcript Search for temporal variation of a in RF transitions of

A Search for Temporal and
Gravitational Variation of a in Atomic
Dysprosium
Arman Cingöz
JILA/NIST Boulder, CO
University of California
at Berkeley
Variation of Constants & Violation of Symmetries, 24 July 2010
Partial support by:
Coworkers
Dmitry Budker, Nathan Leefer
Physics Department, University of California, Berkeley
Collaborators
Steve Lamoreaux
Yale University
Alain Lapierre
TRIUMF, Canada
A.-T. Nguyen
University of Pittsburgh
Justin Torgerson
Los Alamos National Laboratory
Valeriy Yashchuk and Sarah Ferrell
Lawrence Berkeley National Laboratory
Outline
•
•
•
•
•
Overview & motivation
Nearly degenerate levels in dysprosium
Experimental technique
Variation Results/ Status Update
Laser Cooling of Dy
Overview
•
Variation of a would signify physics beyond the Standard Model and
General Relativity.
• Violates Local Position Invariance (a component of Equivalence
Principle), which states that results of non-gravitational experiments
are independent of where and when they are performed
•
WHEN: Temporal variation of fundamental constants:

V. Dzuba et. al., Phys. Rev A 68, 022506 (2003)
V. Dzuba and V. V. Flambaum, Phys. Rev A 77, 012515 (2008)
• WHERE: Null gravitational redshift experiment: compare two different clocks
side by side at the same location
• Recast species dependent shift in terms of gravitational variation of a

V. V. Flambaum, Int. J. Mod. Phys. A22, 4937 (2007)
Search in Atomic Dy
D3 MHz – 1 GHz
B
A
a(t)
transitions
in 5 isotopes

dD/dt ~2.0 x 1015 Hz |a/a|
V. Dzuba et al, Phys. Rev A 77, 012515 (2008)
•Atomic dysprosium (Dy, Z=66) has two nearly degenerate levels that
are highly sensitive to a.

•For |a/a|
~ 10-15 /yr  dD/dt ~ 2 Hz/yr
Self-heterodyning Optical Comparison
• Opposite parity levels  can induce direct electric dipole transitions
between levels
•
DE ~ 3-1000 MHz  can induce transitions with an rf electric field
• Direct frequency counting  relaxed requirements on reference
clock [DE=1 GHz

requires Dn/n~10-12 for a mHz measurement (|a/a| ~ 10-18 /yr )]
• Essentially independent of other fundamental constants
A
n1 - n 2
B
n1
n2
G
Statistical Sensitivity
• Transition linewidth, g, is determined by the lifetime of
state A (t =7.9 ms)  g~20 kHz
• Counting rate ~ 109 s-1
• Statistical sensitivity:
dn ~ g/N1/2 ~ 0.6 Hz s1/2
T1/2
After 1 hour of integration time, dn ~10 mHz which corresponds
to a sensitivity of:

|a/a| ~ 5 x 10-18 yr-1
Additional Correlations
B
w1
A
A
w2
B
w1 + w2 insensitive to a variation
w1 - w2 a variation is twice as large
Currently we monitor:
3.1-MHz transition in 163Dy
235-MHz transition in 162Dy
Parity Nonconservation in Dy
• Degeneracy between levels A and B useful for enhancing mixing
due to the weak interactions
•
Detect quantum interference beat between Stark and PNC mixing
|Hw|=|2.3 ± 2.9 (stat) ± 0.7 (sys)| Hz
A. T. Nguyen et al., PRA 56, 3453 (1997)
• Theoretical calculations are difficult since dominant configurations do
not mix; effect due to configuration mixing and core polarization
Hw=70 (40) Hz
V. A. Dzuba et al., PRA 50, 3812 (1994)
•
Recently, improved calculations suggest Hw ~ 2-6 Hz
V. A. Dzuba and V. V. Flambaum, PRA 81, 052515 (2010)
•
Stay tuned for CW PNC experiment with improved statistical sensitivity
Population
3 step population scheme:
Step 1 and 2: cw laser excitation
Step 3: spontaneous decay with b.r. ~30%
1397 nm
669 nm
833 nm
Detection
• FM modulated rf field transfers population to state A
• State A decays to the ground state in two steps
• 564-nm light is detected
RF
4829 nm
564 nm
First Generation Apparatus Results
.
a/a= (-2.4 ± 2.3) x 10-15 yr-1 A.Cingöz et al., PRL 98, 040801 (2007)
ka=(-8.7 ± 6.6) x 10-6 S. Ferrell et al., PRA 76, 062104 (2007)
2nd Generation Apparatus
3
F
2
1
E
D
A
C
G
B
Differentially pumped
chambers
1. Oven chamber
2. Gate valve
3. Interaction chamber
A. Dy effusive oven
B. Collimator
C. Laser access port
D. Two-layer magnetic shield
E. 4p Optical collection system
F. PMT viewport
G. Rf electrodes
Current Status
• Operational for the past two years
• Collisional shifts reduced to ~ 10 mHz
• Shifts due to rf inhomogeneities
consistent with 0 at the 10 mHz level
• However there were unexpected
problems:
• DC Stark shifts due to stray
charge accumulation: problem
mostly for 3.1 MHz transition
• Zeeman shifts:
Dn/B=DgABmomFmax~2 kHz/1mG
• Zeeman shifts under control at the ~0.1
Hz level
• Stray electric fields mostly stabilized but
need further investigation
.a/a=
(-0.8 ± 2.1) x 10-15 yr-1
Future: Residual Amplitude Modulation
• RAM on top of FM creates
asymmetric sideband amplitudes
which leads to apparent shift of
zero crossing for 1st harmonic
• Due to the large linewidth, RAM is
a serious problem
 ~450 Hz/% RAM
• Measured value in our system
 ~1 x 10-4  4 Hz shifts
•Various ways to control:
• Choose proper phase angle
•Active stabilization
Laser Cooling of Dy
• Increase beam brightness
• A better control of beam density
 Study self collisions
 Reduce systematics due to
spatial inhomogeneities
•A strong cycling transition exists
 421 nm (t = 4.6 ns)
• However, many decay channels
• Calculations suggested B.R. of <10-4
V. A. Dzuba and V. V. Flambaum, PRA 81, 052515
(2010)
Laser Cooling of Dy
• 421 nm source: 1cm PPKTP in a bow tie cavity
90 mW out with 335 mW IR, 27% c.e.
• Transverse cooling experiment:
421 nm
• 3 cm interaction region: ~5000 cycles
• Probe velocity distribution w/ 658 nm
transition
658 nm
Rec. vel.
0.6 cm/s
Doppler limit 20 cm/s
Doppler temp. 0.8 mK
•Fit to Voigt Profile:
• Gaussian width of 0.8(5) MHz
• Lorentzian width of 4.2 (7) MHz
• Limit on branching ratio: < 5 x 10-4
• More stringent limit from MOT experiment
in Urbana-Champaign: 7 x 10-6
M. Lu et al., PRL 104, 063001 (2010)
N. Leefer et al., PRA 81, 043427 (2010)
Conclusion
• The nearly degenerate levels in dysprosium are highly
sensitive to a variation. Direct frequency counting
techniques allow for measurements without state-of-the-art
atomic clocks.
• First generation apparatus sensitivity is ~10-15 yr-1
• Second generation apparatus sensitivity is expected to be
~10 -17 yr-1. Actual data taking will commence soon.
• Transverse cooling of Dy to the Doppler limit has been
demonstrated for all isotopes with large abundance.
• XUV Frequency Combs: Monday Poster Session (Mo 89)
Systematic Effects
A.- T. Nguyen et al. PRA Phys. Rev. A 69, 022105 (2004)
• However, it is not the size but the stability of these effects that is important
 preliminary analysis showed that systematic effects may be controlled to
.
a level corresponding to |a/a| ~ 5 x 10-18 /yr
Search in Atomic Dy
Lock-in Detection Technique
• rf field is frequency modulated at 10 kHz with a modulation index
of 1
• Reduces asymmetries in the line shape caused by drifts (laser
and atomic beam fluctuations)
• Currently use the ratio of these two harmonics
First Harmonic
Second Harmonic
Laser Cooling of Dy
Stray B-fields
• If unresolved Zeeman sublevels are:
sym. populated  leads to broadening , but no shifts
asym. populated  leads to broadening and shifts
Dn/B=DgABmomFmax~2 kHz/1mG
• Nominal config.: linearly polarized pop. beams  aligned state; no shifts
• Systematic due to: spatially varying stress-induced birefringence on optics.
 run-to-run variations due to laser pointing variations.
RF Interaction Region
 Standing Wave
 Small radiative losses
(closed wave guide)
 Impedance matched
 Transparent to light
 Transparent to the atomic
beam
 Homogeneous electric
field (no phase shifts)
 Broadband: 3 MHz to 1
GHz
RF Interaction Region