Transcript Ultracold Atom and Molecule Laboratory

Exploration of the Ultracold
World
Ying-Cheng Chen(陳應誠), Institute of Atomic & Molecular Sciences,
Academia Sinica
IAMS
12 October, 2009, NDHU
Outline
•
•
•
•
Overview of Ultracold Atoms
Introduction to Ultracold Molecules
Exploration I: Molecular cooling
Exploration II: Nonlinear optics with
ultracold atoms
Studying, Research and Life:
Adventure & Exploration
Temperature Landmark
Core of sun
L He
surface of sun
3He
L N2
Sub-Doppler
superfluidity cooling
2003 MIT
Na BEC
0
106
103
Room temperature
1
10-3
Rb MOT
10-6
Typical TC
of BEC
10-9
(K)
What is special in the ultracold world?
• A bizarre zoo where Quantum Mechanics governs
– Wave nature of matter, interference, tunneling, resonance
 h
–
–
–
–
2mkBT
~1μm for Na @ 100nk
Quantum statistics
Uncertainty principle, zero-point energy
System must be in an ordered state
Quantum phase transition
Matter wave interference, MIT
Fermi pressure, Rice
Vortex Lattice, JILA &MIT
Superfluid-Mott insulator t
Ransition, Max-Planck
Laser Cooling & Trapping
• Cooling, velocity-dependent force: Doppler effect
• Trapping, position-dependent force: Zeeman effect
Atom
Laser
fv
v
Magnetic Trapping & Evaporative Cooling
U (r )     B(r )
Microwave transition
Modern Atomic Physics : Science & Technology
Quantum simulation of
condensed-matter physics
Precision measurement
BEC/Degenerate Fermi gas
Atomic clock
Superfluidity/superconductivity
Test of particle physics (EDM)
Quantum phase transition
Test of nuclear physics (parity violation)
BEC/BCS crossover
Test of general relativity
Antiferromagnetism/
Variation of physical constants
Core technology
high Tc superconductivity
Atom
manipulation
Opto-mechanics
& Nano-photonics
Laser cooling of mirror
/mechanical oscillator
Coupling of cold atom with
mesoscopic(nano) object
Quantum limit of detection
Near field optics
Laser
advancement
Quantum information science
Weakness:
Molecule manipulation
Extreme nonlinear optics
Atom/molecule under intense short pulse
High harmonic generation
X-ray laser
Attosecond laser
Quantum control
Quantum teleportation
Quantum network
Quantum cryptography
Quantum computing
Double Helix of Science & Technology
Technology
Better understanding of science helps
technology moving forward
Science
Better technology helps to explore new science
It is a tradition in AMO physics to extend new
technology to explore physics at new regime.
• Atom cooling
Core Technology
U (r )     B(r )
Microwave transition
atom trapping
/optical lattice
Laser cooling
Magnetic-tuned
Feshbach resonance
evaporative cooling
• Laser technology
Ultra-short
Sub-Hz
Ultra-stable
250 as
Ultra-intense
Lasers
Sub-Hz
Ultra-narrow
-linewidth
Non-classical
(single photon,
entangled photon pairs)
100TW
Cold Molecules: Why ?
• Test of fundamental Physics.
– Search for electron dipole
moment…
• Quantum Dipolar Gases
– Add new possibility in quantum
simulation.
• Cold Chemistry
– Chemistry with clear
appearance of quantum effects
– Controlled reaction
• Quantum Computation
– Long coherence time and short
gate operation time
d S
P


r  r
+
+
-
T
t  t
+
-
Cold molecules : How ?
Coherent transfer
from Feshbach
molecule
Enhanced PA?
Laser cooling?
Sympathetic cooling?
Evaporative cooling?
Photoassociation
Direct
approach
Indirect
approach
Buffer gas
cooling
Electric,
magnetic,
optical
deceleration
＋
Breakthrough in Indirect Approach
•
The door to study quantum degenerate dipolar gases and quantum information with
polar molecules is opened by JILA’s recent experiment with indirect approach.
K.-K. Ni et al
Science, 18,1(2008)
Laser Cooling of Molecule ?
Not so cool !
• Its impractical to implement
laser cooling in molecules due
to the lack of closed transition
with their complicated internal
structures.
See, however, Di Rosa, Eur.Phys. J. D
31,395 (2004) for molecules with nearly
closed transition.
The ying and yang (dark/bright) sides of molecules. You have to pay the price !
Our approach ? General considerations
• Choose the direct approach to make cold molecules in order to have
more impacts in other fields as well.
• Generate a large number of molecules in the first stage.
• Build an AC trap in order to avoid the inelastic collision loss.
• Use sympathetic cooling with laser-cooled atoms in the ac trap to
overcome mK barrier for direct cooling.
• What advantages to take? What disadvantages to live with?
Molecules
precooling
loading
Trapping
loading
Laser-cooled
atoms
sympathetic
cooling
Ultracold
Inelastic collision? Molecules
Reaction?
Routes Towards Ultracold Molecules
1 mK
1K
Buffer gas cooling
plus magnetic
guiding
Radiative
damping
& trap loading
Sympathetic cooling in a
microwave trap by
ultracold cesium atoms.
1 μK
Evaporative cooling in
a microwave trap.
hotter molecules
colder molecules
SrF molecule
Cs atom
Recent Ideas
1K
Buffer gas cooling
plus magnetic
guiding
1 μK
1 mK
Direct laser cooling
Evaporative cooling in
an optical dipole trap.
hotter molecules
A2Π1/2
v’
0
ω00
A00
A01
A01
 10  2
A00
A02

X2Σ1/2
A02
 10  4
A00
v’’
2
1
0
colder molecules
What molecule? SrF, Why?
• Alkali-like electronic structure with strong transitions at visible
wavelengths. Easy to be detected by convenient diode lasers.
• Large electric dipole moment, 3.47 D and many bosonic and
fermionic isotopes . More possibilities in the future.
• Microwave trapping consideration. Available microwave high power
amplifier at its rotational transition (2B~ 15 GHz).
• With nearly diagonal Frank-Condon array that allow direct laser
cooling with reasonable number of lasers.
• Suitable for test of fundamental physics and quantum information
science.
• Radical molecules. Disadvantages in molecule generation.
• What advantages to take? What disadvantages to live with ?
Buffer Gas Cooling
X2Σ,v=1→A2Π1/2,v’=1
Q12(7.5)
P11(8.5)
P11(7.5)
Q12(6.5)
P11(6.5)
Q12(5.5)
P11(5.5)
SrF molecules generated by
laser ablation of SrF2 solid.
Q12(4.5)
Development of an intense SrF Molecular
Beam
2B+3 SrF2(high-temperature~1500K)→BF3+Sr+2SrF+BF
N+2
+
CO+2BF 2(neutral
BF3)
BF+
Sr+
SrF+
RGA Trace
Electron-bombardment heating
If one want to work with (cold) molecules then he need to learn some chemistry !
SrF Beam Characterization
Laser beam
5cm
13cm
10cm
Light baffle
ψ3mm skimmerψ2mm
PMT
oven
Residual gas
analyzer
Turbo pump
Brewster window
chopper
Toptica WS-7
Wavelength meter
ECDL laser
New Focus 6009/6300
Setup for laser-induced fluorescence
Typical Spectrum
(0,0) vibrational band of
A2Π1/2- X2Σ+ transition
of 88SrF
Laser intensity ~5 00mW/cm2
FWHM linewidth ~ 130MHz
S/N ratio >200
Even near the congested band edge,
all hyperfine lines are well resolved !
Laser intensity ~ 5mW/cm2
FWHM linewidth ~ 15 MHz
S/N ratio > 50
Hyperfine lines resolved
(I=1/2 for 19F)
Beam Characterization
Flux v.s. oven temperature
Flux stability ~ 20% / one hour
Highest flux of 2.1×1015 /(steradian．sec)!
Even stronger and more stable beam is possible by resistive heating
and is under development!
“An intense SrF radical beam for molecule cooling experiment” submitted to Phys. Rev. A.
Better Spectroscopy of SrF
The rotational/hyperfine lines of (0,0) A2Π1/2- X2Σ+ band 88SrF have been
recorded to 10-4 cm-1 precision with a fitting accuracy of ±10-3 cm-1 to the
effective Hamiltonian.
Theoretical Modeling
•
•
Effective Molecular Hamiltonian
H  H rot  H spin rot  H spinorbit  H lamda doubling
2
4
  D 2  

H rot  BN  DN ; H spin rot   ( N  S )  [ N , N  S ] ; H spinorbit  ALz S z  Ad N 2 ( Lz S z )
2
1
1
H lambda doubling  ( p  2q)(e  2i J  S   e 2i J  S  )  q( J  2 e  2i  J  2 e 2i )
2
2
2
2
1
1
 2  2 i
 2 2 i
 2 i  
2 i  
 qD [ J e  J e , N ]  ( pD  2qD )[e J S  e J S , N ]
4
4
Better molecular constants have been determined !
parameter
T00
B
D
A
p
q
Value(cm-1)
15216.33978(19)
0.2528325(12)
2.5274(28)x10-7
281.46333(34)
-0.13353(9)
9.32(3.8)x10-5
“High-resolution laser spectroscopy of the (0,0) band of A2Π1/2- X2Σ+ transition of 88SrF ”
submitted to J. of Mol. Spec.
Buffer-Gas-Cooled Molecular Beam &
Guiding
• On-going work
Dewar
cryostat
Magnetic guide
oven
Helium
SrF
Turbo pump
Spectroscopy
or laser cooling
UHV Chamber
Estimation of Flux (6.6×1015/s) × (9×10-4)x(2.9×10-3)=1.7×1010/s @ ~5K
Already very intense for a radical beam! Higher flux is possible with modified oven.
Routes Towards Ultracold Molecules
1 mK
1K
Buffer gas cooling
plus ac electric
guiding
Radiative
damping
& trap loading
Sympathetic cooling in a
microwave trap by
ultracold cesium atoms.
1 μK
Evaporative cooling in
a microwave trap.
hotter molecules
colder molecules
SrF molecule
Cs atom
Development of the Microwave Trap
J=1
Rotational
transition
Red-detuned
microwave
AC Stark shift
U(x)
x
J=0
Trapping state
DeMille, Eur.Phys.J D 31,375(2004)
Advantages of microwave trap
1. High trap depth ( ~ 1K)
2. Large trap volume (~ 1cm3)
3. Good optical access. Allow overlap of MOT with trap for sympathetic
cooling.
4. It can trap molecules in the absolute ground states and thus immune to
inelastic collisions loss at low enough temperature.
Observation of standing wave pattern
by thermal-sensitive LCD sheet
E0  4
Q=11000
η=0.87
Pin=1060W
R=0.217m
D=0.2m
QPin z0
D (2 R  D) D
E0=0.45 MV/m
Trap depth ~ 0.1 K
for SrF ground state
“ A high-power microwave Fabry-Perot
resonator for molecule trapping experiment”
Rev. Sci. Inst. In preparation.
Routes Towards Ultracold Molecules
1 mK
1K
Buffer gas cooling
plus ac electric
guiding
Radiative
damping
& trap loading
Sympathetic cooling in a
microwave trap by
ultracold cesium atoms.
1 μK
Evaporative cooling in
a microwave trap.
hotter molecules
colder molecules
SrF molecule
Cs atom
Sympathetic Cooling of Molecules by
Ultracold Atoms
• Conceptually easy but depends on unknown collision properties.
Tempature
TM( t)
Tm
Teq
Ta
N aTa  N mTm
Na  Nm
3N a N m
 th 
2( N a  N m )
• Equilibrium temperature Teq 
• Thermalization time
• Collision rate
  n pa n pm amv c(Ta , Tm )
2
c: a geometry factor and   2M a M m ( M a  M m )
τth
time
Larger number of cold atoms,
colder atom temperature
and higher atom density
implies
lower molecular temperature
and shorter thermalization time.
Large-number Ultracold Atom
System
• Initially developed for molecule sympathetic cooling (with N~ 1010).
• Found its application in low-light-level nonlinear optics based on
electromagnetic-induced transparency (EIT).
7cm
trapping
Coils&cell
Absorption Spectrum
Atom cloud
probe
trapping
trapping beam
“An elongated MOT with high optical density”
Optics Express 16,3754(2008)
Optical density=105
for Cs D2 line F=4 →F’=5
Quest of Second Stage Cooling to overcome
the mK Barrier for Direct Approach
• Sympathetic cooling with ultracold
atoms
– Not so promising due to strong
inelastic loss
cavity-enhanced
Scattering rate
– AC trap is necessary
Rayleigh scattering

cavity linewidthκ
• Cavity laser cooling
N
atomic linewidthΓ
– Haven’t been demonstrated.


 ,
• Direct laser cooling




– Being demonstrated
A2Π1/2 v’
0
– Limited to a few species
ω00
A01
• Single-photon (information) cooling
 10  2
A00
A00
– In combination with magnetic
A01 A
A02 02  10  4
trapping
A00
 v’’
– May be demonstrated soon
2
1
M.Raizen
• ...
X2Σ1/2 0
p
p
p
c
a
Laser Cooling of SrF : to overcome the
mK barrier!
• Di Rosa, Eur.Phys. J. D, 31,395 (2004)
A2Π1/2
v’
0
ω00
A01
 10  2
A00
A00
A01
A02

X2Σ1/2
A02
 10  4
A00
v’’
2
1
0
state
X2Σ,v=0
v=1
v=2
v=3
A2Π, v=0
0.9895
0.0103
1.33x10-4
1.57x10-6
J Phy Chem A, 102,9482,1998
By repumping the v=1 population back to v=0,
the transition is closed to 10-4 level
mv
~ 3600, T  1K
k
0.9998673600=62%
Considering to hyperfine states, it is
necessary to generate two
frequencies differed by ~50 or 107
MHz by acousto-optical modulator
for each laser.
J’
2.5
Nearest>14GHz away
1.5
0.5
2
~45GHz 1
0
X2Σ1/2(v’’=0)
663.1nm
(0,1)P12(1.5)
(0,1)Q11(0.5)
(0,0)R12(1.5)
nearest
interferencerepumping
(0,0)Q12(1.5)
(0,0)P12(1.5)
N’’
(0,0)Q11(0.5)
main
J’’
2.5
1.5
1.5
0.5
0.5
2
X Σ1/2(v’’=1)
685.1nm
parity
＋
－
－
＋
＋
－
N’
0
N’’
parity
＋
＋
－
－
＋
2
J’
F’ parity
A2Π1/2,v’=0
1
0.5
J’’
112.19M
Hz
1.5
0.5
0
X2Σ1/2(v’’=0)
663.1nm
Small
~ few
MHz
(0,0)P12(1.5)
A2Π1/2,v’=0
(0,0)Q11(0.5)
Considering to rotational states, four
lasers (two @ 663nm and two
@685nm ) required to close the
transition to 10-4 level.
29.72MHz
21.75MHz
80.38MHz
－
0
F’’ parity
1
2
＋
26.79MHz1
＋
0
Nonlinear optics with ultracold atoms
- Detour of my planned journey but back to my old track !
Electromagnetically-induced Transparency
Transparent!
|3>
Probe
laser
32 32
p
c

2>
Coupling laser
|1>
 c  0, c  0.53 ,   0
Physical origin: destruction interference between different transition pathways!
|3>
probe
coupling
＋
＝
＋…
＋
|2>
|1>
Path i
Path ii
Path iii
2
Atot

EIT, Propagation Effect
vg 
d p
dk p
Vg<17m/s, Hau et.al.
Nature397,594,1999
| p 0 
c
n   p (dn / d p ) | p 0
Slow light !
Lng


1 1
32
 d  L(  ) 
 N(
) L 2 31
 OD 2 31
vg c
c
2
( c  31 )
( c  31 )
• Large optical density and small ground-state decoherence rate are
two crucial factors in EIT-based application, e.g. optical delay line.
Nonlinear Optics with Ultracold Atoms
• With on-resonance signal, one can control the
absorption/transmission of probe photon by signal photon.
Photon switching.
• With off-resonant signal, one can control the phase of probe photon
by signal photon. Cross phase modulation.
4
3
probe
signal
coupling
1
γ
With signal beam
Without
signal
2
Schmidt & Imamoglu Opt. Lett. 21,1936,1996
XPM Application: Controlled-NOT gate for
Quantum Computation
•
•
CNOT and single qubit gates can be used to implement an arbitrary unitary
operation on n qubits and therefore are universal for quantum computation.
Single photon XPM can be used to implement the quantum phase gate and
CNOT gate
Truth table for CNOT gate
0 C 0T
 0 C 0T ; 0 C1T
1C 0T
 1C1T ; 1C1T
PBS
Control qubit
1  
0 or 1
 1C 0T
PBS
Signal
0  
Probe
 0 C1T
Atoms



Target qubit
For a good introductory article, see 陳易馨&余怡德 CPS Physics Bimonthly, 524, Oct. 2008
Reduction of Ground-state decoherence rate
Reduction of mutual laser linewidth
Coupling
ECDL
RF
Bias-Tee
Idc
Reduction of inhomogeneity of stray
magnetic field
Faraday rotation
as diagnosis tool.
Three pairs of coils
for compensation.
VCSEL
L  B
PBS
λ/2
350kHz/Gauss for Cs
Probe
DL
coupling
Without
compensation
~9GHz
VCSEL
FFT
frequency
probe
Beatnote between coupling & probe laser
With
compensation
~10Hz
δB<2mG limited by 60Hz AC magnetic field!
Good EIT Spectrum
Obtained EIT with ~50% transmission at 200kHz width for OD~ 60 for Cs
D2 F=3 →F’=3 transition.
The Slow Light
10μs for ~2cm
atomic sample !
Vg~2000m/s
XPM with Group-Velocity-Matched Double
Slow Light Pulses
• Both probe & signal pulses becoming group-velocity-match slow
light in a high OD gas for longer interaction time. M. Lukin Phys.
Rev. Lett. 84, 1419 (2000).
4
3
probe
signal
signal
coupling
2
Atom A
1
probe
medium
signal
Atom B
Double EIT Spectrum
mF=
0
1
2
3
P2
4
F=4, gF=4/15
C2 P2
F=3, gF=0
P1
(a)
P1
F=4, gF=1/4
C1
F=3, gF=-1/4
Cs 6S1/2 -6P3/2 (D2-line)
• Photon-switching with on-resonance signal field has been observed.
• XPM work is underway !
Matching the Group Velocity
Probe 1
Probe 2
No atoms
Group velocity matched !
IC1
fixed
Td(P1)
Td(P2)
decrease
IC2
Future Work : Cavity Enhanced Cross Phase
Modulation
•
A “holy grail” in nonlinear optics is to realize a
mutual phase shift of πradian with two light
pulses containing a single photon.
It can be applied to the implement of controlledNOT gate for quantum computation and to
generate quantum entangled state.
Few-photon-level XPM is challenging !
•
•
–
–
–
–
•
Large Kerr Nonlinearity
Low loss
Strong focusing to increase the atom-laser
interaction strength
Long atom-laser interaction time
We are working on cavity-enhanced XPM. The
technology may also be applied to cavity laser
cooling of molecules in the future.
cold atom
Coupling&
probe
Signal beam
The Setup
Acknowledgement
• Financial support from NSC, IAMS.
• Helps from many colleagues,
WY Cheng, KJ Song, J Lin, K Liu, SY Chen…
• Current member:
–
–
–
–
Chih-Chiang Hsieh
Ming-Feng Tu
Jia-Jung Ho
Wen-Chung Wang
• Former member
–
–
–
–
–
S. -R. Pan (now in Colorado state University)
H.-S. Ku (now in Univ. of Colorado/JILA)
T.-S. Ku (now in Univ. of Colorado/JILA)
Prashant Dwivedi (now in Germany’s Univ.)
P.- H. Sun (now in industry)
Keep walking !
Molecule cooling
Nonlinear optics with ultrcold atoms
Welcome to join us !
Ultracold Atom and Molecule Lab
IAMS, Academia Sinica
Slow Light : Dark-State Polariton
coupling
coupling
probe
coupling
probe
|2>
|2>
|2>
1>
|3>
|3>
|3>
1>
1>
Light component
 ( z , t )  cos E p ( z , t )  sin 
tan   n g ; k  k c  k p
N  21 ( z , t )e ikz
Matter component:
atomic spin coherence


 c cos 2 
] ( z , t )  0
t
z
v g  c cos 2 
Lukin&Fleischhauer, PRL 84,5094,2000
[
EIT and the Photon Storage
•
•
•
By adiabatically turn off the coupling light, the probe pulse can completely
transfer to atomic spin coherence and stored in the medium and can be
retrieved back to light pulse later on when adiabatically turn on the coupling.
This effect can be used as a quantum memory for photons.
The photon storage and retrieved process has been proved to be a phase
coherent process by Yu’s team.
coupling
probe
Hau et.al. Nature, 409,490,2001
Y.F. Chen et.al. PRA 72, 033812, 2005
Q-Value Measurement Under
High-Power Operation
Quality-factor
PUnlocked
U (t )  U (0)e  t /
Q
 
microwave OFF
0
Coupling efficiency
PLocked
PLocked
  1
PUnlocked
Cavity Frequency Locking
•
•
•
Pound-Drever-Hall Scheme to obtain error signal
Feedback by vacuum linear translation stage
Locked to better than 50 kHz (linewidth ~ 700kHz)
Locked
Fabry-Perot Cavity Coupling
•
•
Coupling by a circular horn through mirror with mesh.
Obtained optimum coupling through systematic study by varying mesh
parameters.
Reflection signal
Observed Line narrowing effect
for large OD gas
Increasing the
OD of atom cloud