Université de Nice Sophia Antipolis - CNRS I N L N Multiple scattering of light by cold atoms Robin KAISER INLN, Nice, France Waves and disorder, Cargèse, France, July 1 – 11, 2014
Université de Nice Sophia Antipolis - CNRS I N L N
Multiple scattering of light by cold atoms
Robin KAISER
INLN, Nice, France
Waves and disorder, Cargèse, France, July 1 – 11, 2014
Université de Nice Sophia Antipolis - CNRS I N L N
Lecture 1: 1.1 Two level atoms1.2 Diffusion
Lecture 2: 2.1 The case for Anderson2.2 The case for Dicke2.3 Anderson and Dicke
Université de Nice Sophia Antipolis - CNRS I N L N
INLN:
G. Labeyrie, D. Wilkowski, C. Miniatura, W. Guerin
N. Mercadier, Q. Baudouin, L. Bellando, T. Bienaimé, J. Chabé, T. Rouabah, G.L. Gattobigio, F.
Michaud, T. Chanelière, V. Guerrara, C. Müller, Y. Bidel
+
S. Tanzilli, J. Barré, B.Marcos , M. Faurobert , M. Lintz , F. Impens, F. Bouchet, D. Delande, R.
Carminati, S. Skipetrov, L. Froufe-Pérez, R. Pierrat, A. Picozzi
+
E. Akkermans, N. Piovella, L. Celardo, R. Bacherlard, P. Courteille, E. Perreira, M. Havey, T.
Ackemann,, J. Tabosa, M. Chevrollier, T.Pohl, J. Ott, T. Menconça, H. Tercas, G. Alvarez, G. Robb, A.
Arnold, W. Firth, G.L. Oppo
Université de Nice Sophia Antipolis - CNRS I N L N
DickeSuperradiance
Coherent Multiple Scattering
Multiplescattering
Interferences
Anderson Localisation
“Local” “Global”
mesoscopic transport(superconductors, insulators)
Cooperative emission (superradiant lasers, antennas,
quantum memories)
Université de Nice Sophia Antipolis - CNRS I N L N
AndersonLocalization
RadiationTrapping
DickeSubradiance
PhotonicCrystal
disorder
order
coherencedecoherence
slow/stopped light, quantum holography
How to trap a ‘photon’ with N atoms?
Université de Nice Sophia Antipolis - CNRS I N L N
The case for Anderson …
Université de Nice Sophia Antipolis - CNRS I N L N
θθθθ
((((1111))))kin
((((2222))))
I0
rin
routkout
θθθθ
Weak Localization => Coherent Backscattering
· uncorrelated paths add incoherently
∆ϕ=(kin+kout).(rin-rout) θ=0 ⇒ ∆ϕ = 0 for any path
<I(0)>
<I(θ)>>∆θCBS)>= 2
Coherent
Backscattering
· correlated (i.e. reciprocal) paths
add coherently
multiple self-aligned Sagnac interferometer
Université de Nice Sophia Antipolis - CNRS I N L N
Single
realization
Configuration Average
θ (mrad)
Université de Nice Sophia Antipolis - CNRS I N L N
Single
realization
Configuration Average
θ (mrad)
Configuration
average
Université de Nice Sophia Antipolis - CNRS I N L N
CCD
Lens
MOT
beam
splitter
Probe laserN ≈ 1010
T≈100µKkl ≈ 1000
Coherence afterresonant scattering
with atoms !
also : M. Havey et al.
Phys. Rev. Lett., 83, 5266 (1999)
Weak Localisation with resonant scattering by atoms
Université de Nice Sophia Antipolis - CNRS I N L N
Phys. Rev. Lett., 85, 4269 (2000)
Theory : • no “exact” solution• diagrammatic approach
Excellent agreement (no free parameter)
0 2 4 6-2-4-6
q
(mrad)
experimentMC simulation
h
^
CB
Sen
han
cem
ent
1.00
1.05
1.10
1.15
1.20
h //
lin //lin
^
Rb85 : F=3 - F'=4=4
Sr88 : J=0 - J'=1
=1
CB
Sen
han
cem
ent
1
2
0 2 4 6-2-4-6
q
(mrad)
h //
Université de Nice Sophia Antipolis - CNRS I N L NUniversité de Nice Sophia Antipolis - CNRS I N L N
-3 -2 -1 0 1 2 3
1,0
1,1
1,2
1,3
1,4
1,5
1,6
1,7
1,8
1,9
2,0
Rb
Sr
Sr88 : J=0 - J'=1
Rb85 : F=3 - F'=4
angle (mrad)
enhan
cem
ent
fact
or
Quantum fluctuations :inelastic scattering non linear response
Internal structure : Rb = quantum magnets Sr = classical dipole
Restoring two level atoms:(negative magneto-resistance)
0.0 0.2 0.4 0.6 0.81.5
1.6
1.7
1.8
1.9
2.0
CB
S e
nh
ance
men
t
saturation parameter
Sr88
COHERENT BACKSCATTERING = coherent probe
PRL, 85, 4269 (2000) PRE, 70, 036602 (2004)
PRL, 88, 203902 (2002)
PRL, 89, 163901 (2002)
PRL, 93, 143906 (2004)
-10 0 10 20 30 40 501,00
1,05
1,10
1,15
1,20
1,25
1,30
1,35
-60 -40 -20 0 20 40 60
CB
Sen
han
cem
ent
B (G)
angle (mrad)
B = 0
B = 43G
h //
MC
exp.
PRL, 93, 143906 (2004)
Temperature :‘fast’ atomic dynamics ‘slow’ transport of light
0,01 0,1 11,05
1,10
1,15
CB
S e
nhan
cem
ent
k vrms
/ Γ
T = 80µK T = 50mK
Rb85
PRL, 97, 013004 (2006)
Université de Nice Sophia Antipolis - CNRS I N L N
Towards strong localization of light : dense atomic clouds
Strong Localization
+BECBEC
10141012 1016 1018 10201010
10-10
10-8
10-6
10-4
10-2
10-0
Dynamical BreakdownDynamical BreakdownT
[K
]
Weak Localization
of Light
BEC
Ioffe-Regel :
Strong Localization
of Light
k l ≈ 1000N= 1010
k l ≈ 1
N= 107
k l ≈ 3
Dipole Trap(M. Havey)
Université de Nice Sophia Antipolis - CNRS I N L N
The case for Dicke …
Université de Nice Sophia Antipolis - CNRS I N L N
Single photon superradianceDicke : N ’’2-level’’ atoms
R. Dicke (1954)Reviews by : R,Friedberg, S.R.Hartmann, J.T;Manassah (1973)M.Gross, S.Haroche (1982)
1 excitation shared by many atoms
Regained interestTheory : M. Scully,
R,Friedberg, J.T.ManassahS.Prassad, R.Glauber
Experiments: R. RoehlsbergerC. Adams
Superradiant pair
Subradiant pair
|ee>
|eg>+ |ge>
|gg>
|eg>- |ge>
Université de Nice Sophia Antipolis - CNRS I N L N
Effective Hamiltonian
• Open System • Reminiscent of Anderson Hamiltonian• Heisenberg model with global coupling
Off diagonal :transport
Diagonal :On site energy
Université de Nice Sophia Antipolis - CNRS I N L N
Photon Escape Rate Distribution
∝ Im(Heff)
Université de Nice Sophia Antipolis - CNRS I N L N
Photon Escape Rate Distribution
Université de Nice Sophia Antipolis - CNRS I N L N
Photon Escape Rates : single parameter scaling
E. Akkermans, A. Gero, RK, PRL, 101, 103602 (2008)
cooperative effects dominate over disorder !no phase transition observed with P(Γ)
Single parameter scalingb0 = N/N⊥
Dicke > Anderson
Measure of long lived photons
Université de Nice Sophia Antipolis - CNRS I N L N
Two level + polarisation
no phase transition observed with P(Γhν)
Université de Nice Sophia Antipolis - CNRS I N L N
Cloud of Atoms = Large Molecule (with 1010 atoms)
‘dilute’ molecule‘dense’ molecule
molecular spectrum ?
Beyond Photon escape times :
proximity resonances
(Heller et al)
Mie-Debye model
(Sokolov et al)
Université de Nice Sophia Antipolis - CNRS I N L N
Eigenvalues
eikr/kr
Full vector model
Scalar model
eikr ( 1/kr + 1/kr2 + 1/kr3)
Two level
+ polarisation
Université de Nice Sophia Antipolis - CNRS I N L N
Eigenvalues
Université de Nice Sophia Antipolis - CNRS I N L N
What about Random Matrix Theory ?
No energy level repulsion … but in the complex plane
Dilute gas Dense gas
Université de Nice Sophia Antipolis - CNRS I N L N
Partition Ratio
N-1/3 N-2/3
Université de Nice Sophia Antipolis - CNRS I N L N
Level repulsion vs Level WidthResonance overlap in superradiant transition (doorway states) and Anderson transition (Thouless criterion)
metal (g>>1)
ln g
∂ ln g
∂ ln L
β(g)=
1D
2D
3D
g : dimensionless conductance
1<<g
insulating
metallic
1>>g
Université de Nice Sophia Antipolis - CNRS I N L N
Resonance overlap
Université de Nice Sophia Antipolis - CNRS I N L N
S. Skipetrov, I. Sokolov PRL 112, 023905 (2014)
Resonance overlap
metal (g>>1)
ln g
∂ ln g
∂ ln Lβ(g)=
1D
2D
3D
g : dimensionless conductance
1<<ginsulati
ng
metallic1>>g
g = 10-L/ξ
• M. Antezza, Y. Castin, D. Hutchinson
Phys. Rev. A 82, 043602 (2010).
Université de Nice Sophia Antipolis - CNRS I N L N
So far :
Effective Hamiltonian :
Escape rate : no phase transition
Eigenvalue analysis : phase transition in the scalar case
Observable?
Université de Nice Sophia Antipolis - CNRS I N L N
Hamiltonian(RWA)
Markov approx.
Low saturation
Heffcoherent
drive
Trace over environment
N coupled equations to be solved
Driven System of N Dipoles
γij ∝ exp(ikrij)/krji
Université de Nice Sophia Antipolis - CNRS I N L N
N coupled equations :A many body problem with 1 photon and N atoms
• Mean Field Ansatz (Timed Dicke Ansatz)
• Continuous field β(r)
• Numerical solution of the many body problem (‘exact’)• Experiments
index of refraction
Université de Nice Sophia Antipolis - CNRS I N L N
Experimental observable : Average force on center of mass (easier to measure):
“Mean Field” result (driven Timed Dicke Ansatz)
DisorderEmissionDiagram
(Mie)
Superradiance
Nat/Nmodes ∝∝∝∝ Nat / (L/λλλλ)2 ∝∝∝∝ b0Σ ≠ ∫
Université de Nice Sophia Antipolis - CNRS I N L N
PRL, 104, 183602 (2010)
A new experiment
Reduced radiation pressure force
explained by
cooperative scattering
Université de Nice Sophia Antipolis - CNRS I N L N
also works with ultra-cold atoms
in dipole trap
will work in all large ultra(cold) atom clouds !
Experiments in Tübingen (Ph. Courteille et al.)PRA 82, 011404 (2010)
Université de Nice Sophia Antipolis - CNRS I N L N
Beyond Mean Field :
• What to expect at resoance, when b(δ)>>1? • Multiple scattering … opaque sample …
more reflected light … larger momentum transfer Fc>Find ?
• No analytical result available • Numerical solution of the many body problem (‘exact’)• Experiments
Université de Nice Sophia Antipolis - CNRS I N L N
2
4
6
30
210
60
240
90
270
120
300
150
330
180 0
Many body
Timed Dicke
Incoherent
Spherical gaussian cloud : emission diagram
Université de Nice Sophia Antipolis - CNRS I N L N
How to probe coherent multiple scattering
Université de Nice Sophia Antipolis - CNRS I N L N
Cloud of Atoms = Large Molecule (with 1010 atoms)
‘dilute’ molecule‘dense’ molecule
molecular spectrum : narrow and broad lines
(see nuclear physics)
Broadband excitation vs coherent excitation
Time dependent experiments : coherent scattering
Université de Nice Sophia Antipolis - CNRS I N L N
Temnov, Woggon,PRL 95, 243602 (2005)
ΩΩΩΩ⊥⊥⊥⊥
| G >
| E >| D >
ΓΓΓΓN
ΓΓΓΓD
Inhomogeneous broadening in |ei> => coupling in |TD>
Superradiance = bright stateSubradiance = metastable state
Time dependent experiments : coherent scatteringInverted system
T. Bienaimé, N. Piovella, R.K. PRL (2012)
Université de Nice Sophia Antipolis - CNRS I N L N
Fano Coupling and controlled subradiance
Fano
DopplerRandom Light shift
T. Bienaimé, N. Piovella, R.K. PRL (2012)
Université de Nice Sophia Antipolis - CNRS I N L N
0 50 100 150
tran
smit
ted
dif
fuse
inte
nsi
ty
t (Γ−1
)
10−1
1
b = 19
b = 2.4
b = 34
probe
beamcold
atoms
0 t
PM
0
∝ e- t/τ0
t
IscI
in
L
Time dependent experiments : incoherent scattering
Université de Nice Sophia Antipolis - CNRS I N L N
probe
beamcold
atoms
0 t
PM
0
∝ e- t/τ0
t
IscI
in
L
Time dependent experiments : incoherent scattering
Experimental work in progress
Inelastic Raman scattering to avoid: Zeeman pumping in streched state
Université de Nice Sophia Antipolis - CNRS I N L N
Time dependant scattering : beyond diffusion
C. Aegerter et al., EPL, 2006
Anderson localisationt0∝∝∝∝ exp(-L)
?
…Dicke
subradiancet0∝∝∝∝1/L ?
Université de Nice Sophia Antipolis - CNRS I N L N
Transmission in dense media?
Milk
Coupled Dipoles(dense clouds)
Université de Nice Sophia Antipolis - CNRS I N L N
Combining Anderson and Dicke …
Université de Nice Sophia Antipolis - CNRS I N L N
Toy Model : Open Disordered System:A. Biella et al., EPL, 103, 57009 (2013)
3D Anderson model on 10x10x10 latticehoping (Ω) + disorder (W) + opening (γ)
All sites coupled to one single decay channel : Qij=1
Ω
γ
Université de Nice Sophia Antipolis - CNRS I N L N
Hybrid Subradiant States « decoupled » from outside world
Université de Nice Sophia Antipolis - CNRS I N L N
What’s next :
• Hybrid state with Heff
novel road vs localisation of light
• Subradiance experiments
• Atomic clocks (best clock)
• Dipole blockade• Quantum Optics• NMR: dipole-dipole coupling
probe
beamcold
atoms
0 t
PM
0
∝ e- t/τ0
t
IscI
in
L
Université de Nice Sophia Antipolis - CNRS I N L N