Transcript First year seminar

First Year Seminar:
Strontium Project
Graham Lochead
03/06/09
Outline
• Introduction and motivation
• Second generation cell
• Polarization spectroscopy and subDoppler DAVLL
• Strontium pyramid MOT
• 689 nm locking progress
Graham Lochead
03/06/09
Motivation: Rydberg physics
• Rydberg states are
states with large n
• Rydberg states have
2
large orbital radii  n
• We aim to trap
ultracold strontium in a
1-D optical lattice and
excite to Rydberg
states
Graham Lochead
03/06/09
Motivation: Ultracold plasmas
• Most plasmas are dominated
by their thermal energy
• Coulomb coupling parameter
 is ratio of Coulomb energy
to thermal energy
• Strong Coulomb interactions
lead to spatial corrrelations
• Cold plasma in a spatially
ordered lattice will be a first
T.C. Killian et al., Physics Reports 449, 77 (2007)
Graham Lochead
03/06/09
Strontium overview
• Alkaline-earth element (Group II)
• Atomic number 38
84 Sr
86 Sr
87 Sr
88 Sr
0.6%
9.9%
7%
82.5%
Graham Lochead
03/06/09
I=0
I=0
I=9/2
I=0
boson
boson
fermion
boson
Electronic structure
3D
3S
1
2
412 nm
1
1P
1
3P
2
1D
2
461 nm
/2p = 32 MHz
689 nm
/2p = 7.5 kHz
ground state – no
optical pumping
• Low decay rate to metastable state 3P2
1S
0
1
0
698nm
/2p = 1 mHz 87Sr
1S
0
•
3
• Broad linewidth for 1S0-1P1
transition
• Intercombination line for
further cooling
Graham Lochead
03/06/09
Why strontium?
• Singly ionised strontium has an optical
transition at ~ 422 nm for 2S1/2-2P1/2
• Ion transition can be used for:
– imaging
– observing charge transfer
– laser cooling
– Rydberg manipulation
T.C. Killian et al., Phys. Rev. Lett., 92:143001, 2004.
Graham Lochead
03/06/09
Laser frequency stabilization “locking”
Laser locking
requires an atomic
sample to investigate
the transition
And a detection
scheme that gives a
slope to lock to
Graham Lochead
03/06/09
• Introduction and motivation
• Second generation cell
• Polarization spectroscopy and subDoppler DAVLL
• Strontium pyramid MOT
• 689 nm locking progress
Graham Lochead
03/06/09
Problems with strontium
Locking to a transition
requires an atomic sample
Atomic strontium has very low
vapour pressure
Hot strontium reacts with
glass and copper
M. Asano, K. Kubo
J. Nuclear Sci. & Tech.
15 pp. 765~767 (1978)
Graham Lochead
03/06/09
Dispenser technology
Sr
Sr
• Sealed in argon with indium plug
• Directional source of atomic vapour
• Flux is dependent on current supplied
Graham Lochead
03/06/09
First generation cell
Dispenser
• Birefringent sapphire
windows not required
• No continual pumping
• No buffer gas
• Lifetime estimate ~ 10000 h
• Compact size for Sr
• Strontium acts as a getter
• Only 15-20% absorbtion
stable operation
A vapor cell based on dispensers for laser spectroscopy
E. M. Bridge, J. Millen, C. S. Adams, M. P. A. Jones
Rev. Sci. Instr. 80,013101 (2009)
Graham Lochead
03/06/09
Second generation cell
30 cm
Dispenser
Baffle
Designed by
Clementine Javaux
Graham Lochead
03/06/09
Second cell absorption
Doppler FWHM
of 1.7 GHz at
50% absorption
• Optically thick for 461 nm transition
• Wide Doppler profile due to dispenser type
Graham Lochead
03/06/09
Second cell saturated absorption
124.5 MHz
Probe = 0.14 mW
Pump = 7.3 mW
Pump
Laser
 
 probe  L  0  k  v
Probe
Cell
Frequency axis calibrated from 86Sr-88Sr splitting
88Sr
transition peak is ~5% of the optical depth
Graham Lochead
03/06/09
• Introduction and motivation
• Second generation cell
• Polarization spectroscopy and subDoppler DAVLL
• Strontium pyramid MOT
• 689 nm locking progress
Graham Lochead
03/06/09
Polarization spectroscopy theory
J=1
mJ = -1
5s5p
1P
1
0
σ-
5s2 1S0
I signal
p
J=0
mJ = 0
+1
σ+

1
(     )
2
2
1
(     ) Cell
2
x
 I H  IV  I 0 L 0
1 x2
x  2  0  
M. L. Harris, et al. Phys. Rev. A, 73:062509, 2006.
C. P. Pearman et al., J. Phys. B, 35:5141, 2002.
Graham Lochead
03/06/09
Polarization spectroscopy setup
Polarization spectroscopy
Frequency calibration
Laser
Metal
mirror
1
(     )
2
Cell 1

2


Cell 2
Differential
photodiode
4
Graham Lochead
03/06/09
Polarization spectroscopy results
Signal (mV)
• Gives a steep gradient
– easy to lock to
• 0.8 MHz rms offset
stability over an hour
20
0
-20
-40
-400
-200
0
200
Detuning (MHz)
40
20
(a)
0
0
0.2
0.4
0.6
Intensity (W cm-2)
0.8
Gradient (mV MHz-1)
Amplitude (mV)
60
1.2
0.8
0.4
00
(b)
0.2
Graham Lochead
03/06/09
0.4
0.6
Intensity (W cm-2)
0.8
400
DAVLL theory
Dichroic Atmoic Vapour Laser Lock (DAVLL)
5s5p 1P1
mJ = -1
E
J=1
+1
+1
0
mJ = -1
p
+
σ
σ+
σ-σ
5s2 1S0
Apply a uniform magnetic
 E field to atomic sample with
Helmholtz coils
J=0
mJ = 0
Creates a difference in frequency
between different transitions
E    mJ g J B B
Taking the difference of these
signals leads to a dispersion
signal with zero crossing at
the B=0 transition
B B
   0 

B B
   0 

Graham Lochead
03/06/09
Sub-Doppler DAVLL setup
Sub-Doppler
DAVLL DAVLL
To frequency
calibration
Laser
1
(     )
2
Metal
mirror

4

Coils

Differential
photodiode
1
(     )
2
M.L. Harris et al., J. Phys. B. Phys. 41 085401
Graham Lochead
03/06/09
Sub-Doppler DAVLL trace
• 3 MHz rms offset
stability over an
hour
Graham Lochead
03/06/09
Sub-Doppler DAVLL characteristics
Graham Lochead
03/06/09
Laser locking summary
• Polarization spectroscopy is used to lock the
461 nm laser with first cell as offset more stable
• These two locking schemes have been
characterized and written up
• Second cell will be used for thermal Rydberg
spectroscopy with pulsed dye laser
arXiv:0902.1430v1 [physics.atom-ph]
Graham Lochead
03/06/09
• Introduction and motivation
• Second generation cell
• Polarization spectroscopy and subDoppler DAVLL
• Strontium pyramid MOT
• 689 nm locking progress
Graham Lochead
03/06/09
What is a pyramid MOT?
Normal (6 beam) MOT
Pyramid MOT
K. I. Lee et al., Optics Letters, Vol. 21, Issue 15, pp. 1177-1179
Graham Lochead
03/06/09
Pyramid MOT function
Acts as a cold atom source
Graham Lochead
03/06/09
Benefits of a pyramid MOT
• Size – much smaller than a Zeeman slower
• Blackbody radiation effects reduced
peak  T
Graham Lochead
03/06/09
P T4
Chamber design
45 cm
Design considerations
• Trapping gradient
of 30 G/cm
• No water cooling
• Standard vacuum
parts
30 cm
Graham Lochead
03/06/09
Mirror mount design
Most pyramid MOTs are loaded
from background atomic vapour or
dispensers above pyramid
Problem
Low vapour pressure and mirrors get coated
Solution
• Dispensers below mirrors
• Slits where mirrors meet in corners
Graham Lochead
03/06/09
The mount design
45 mm
Mirror size needs to be small to avoid
pumping into meta-stable states
Graham Lochead
03/06/09
Atomic beam divergence measurement
Expanding lens

Atomic beam
4
Laser
Collimating lens
Light sheet
Cylindrical lenses
AOM
Light sheet
Graham Lochead
03/06/09
• Introduction and motivation
• Second generation cell
• Polarization spectroscopy and subDoppler DAVLL
• Strontium pyramid MOT
• 689 nm locking progress
Graham Lochead
03/06/09
Motivation for 689 nm laser
We will use a 532 nm optical lattice laser to
add periodic spatial confinement to our MOT

Doppler limited temperature TD 
2k B
1P
1
3P
1D
2
461 nm
/2p = 32 MHz
TD ≈ 1 mK
2
689 nm
/2p = 7.5 kHz
TD ≈ 0.2 μK
1S
0
Graham Lochead
03/06/09
1
0
Pound-Drever-Hall
FPD
Oscilloscope


2
PS

4
Laser
Graham Lochead
03/06/09
Pound-Drever-Hall setup
FPD
Strontium
cell
Filter
Slow
feedback
to piezo
Fast
feedback
to diode
Feedback
to cavity
piezo


2
PS

4
Laser
Graham Lochead
03/06/09
Slow lock
100k
*
Ramp
100k
1k
PDH signal
10nF
100k
1k
2k
100k
*
100
AD620
510
+15V REF02
1k
100k
100k
1k
1k
1k
1k
+15V
Unlocked laser
has linewidth of
~ 600 kHz
Locked laser
has linewidth
of ~ 350 kHz
Graham Lochead
03/06/09
To
piezo
Cavity lock to atomic transition
Going to use frequency modulation spectroscopy
as laser already modulated
AOM
Probe
80 MHz
20 kHz
10 MHz
Pump
Lock in
amplifier
Cell

4
Feedback
to cavity
piezo
Graham Lochead
03/06/09
Summary and future work
• Second cell and locking schemes characterized
• Pyramid MOT design
• 689 nm locking progress
Future work
• Build the pyramid MOT
• Achieve a red MOT
• Load 1-D lattice
Graham Lochead
03/06/09
Saturated absorption spectroscopy fit
Sat. spec. fit is achieved by minimizing sum of six Lorentzians in Matlab
6  a
Fit  A  I i
 i 1 
 (  ( sI is  0 ))2  
1 
   c  m
2
( s)

 
1% scaling accuracy for the frequency
Parameters
A
s
Amplitude of trace
Scaling factor
ω0
Centre frequency

Width
c
Offset
m
Gradient of background
Isotope
Abundance (%)
Shift (MHz)
Rel.
Strength
a
I
F
84Sr
0.56
0
-
-270.8
1
86Sr
9.86
0
-
-124.5
1
7/2
-9.7
4/15
9/2
-68.9
1/3
11/2
-51.9
2/5
-
0
1
87Sr
88Sr
Graham Lochead
03/06/09
I
7.00
82.58
9/2
0
I
s