Super-radiant%20light%20scattering%20with%20BEC

1
Super-radiant light scattering with BECs a
resource for useful atom light entaglement?
Jörg Helge Müller, Quantop NBI Copenhagen

• Motivation quantum state engineering
• Light-atom coupling in Rubidium
• Sample preparation BEC setup
• First light Superradiance revisited
• Dynamics in simple models
• Counting atoms and photons
• Future directions

QCCC Workshop, Burg Aschau, October 2007
2
Light-Atom interaction seen from both sides
Spectroscopy light is modified by atoms (e.g.
polarization rotation)
Laser manipulation Atoms are modified by
light (laser trapping, optical pumping,...)
Both things happen at the same time
We want to study and exploit the regime where
quantum effects matter to prepare interesting
quantum states!
Quantum State Engineering
3
Coupling at the microscopic level
...plain dipole scattering
Use a high finesse cavity! or Use many
atoms/photons!
In free space this coupling is small
Our strategy

• mix quantum modes with strong orthogonally
  polarized local oscillator
• light quadratures show up as polarization
  modulation
• use ensemble of many polarized atoms ?
  macroscopic spin/alignment
• phase matched scattering into forward direction
  
• polarization modulation modifies the
  macroscopic spin/alignment

4
Rb F1 ensembles and polarized light
Local atom light interaction
birefringence
phase shift
polarization rotation
Raman coupling
Larmor precession
level shift
5
Reduction to forward scattering

1. Transverse light propagating along z-direction
2. Atoms prepared initially in mF -1 , 1 , (0)

J Bloch vector of the 2-level system (one
classical, two for quantum storage) S Stokes
vector for light (one classical, two for the
quantum mode)
b coefficients can be tuned with the choice of
laser frequency!
Vector coefficient Faraday interaction (single
quadrature, QND coupling) Tensor coefficient
Raman coupling (two quadratures, back-action)
Now we need to add propagation effects....
6
Application to Quantum memory

1. Quantum memory

Negative feedback (back-action cancellation) in
both quadratures
Single-pass Optimized geometry
(Tune bV to zero)
Output light for coherent state input in the
quantum mode oscillating response
Feedback during propagation leads to spatial
structure Spin waves
7
Application to light atom entanglement
2. Parametric Raman amplifier
Positive feedback (back-action
amplification) EPR-type entanglement between
light and atoms Super-radiant Raman
scattering Our detour Super-radiant Rayleigh
scattering
Input/Output relations can be calculated and
decomposed into mode pairs for atom and light
Wasilewski, Raymer, Phys.Rev. A 73, 063816
(2006) Nunn et al., quant-ph/0603268 Gorshkov
et al., quant-ph/0604037 Mishina et al.,
Phys.Rev. A 75, 042326 (2007) Efficient
optimization of memory performance by tailored
drive pulses possible
8
Important parameter for collective coupling
On-resonance optical depth of the sample
Single atom spontaneous scattering
Coupling strength bigger than 1 (usually) means
quantum noise of atoms becomes detectable on
light and vice versa. Optical depth should be as
high as possible!!
9
Sample preparation BEC setup
10
BEC setup (2)
QUIC trap (inspired by Austin group,
good thermal stability) Ioffe coil with
optical access
Imaging along vertical direction Ioffe axis free
for experiments
11
Evaporation and trap performance
Slope ? 1.3
Slope ? -3
Radial frequency ? 116 Hz Aspect ratio ? 12 Atom
number ? 6 ? 105
12
First light Super-radiance revisited
3-level system with total inversion initially
Build-up of coherence enhances scattering

• Example Coherence in momentum space
• photons and recoiling atoms created in pairs
• atom interference creates density grating
• enhanced scattering off density modulation
• runaway dynamics until depletion sets in

Super-radiant emission
Ordinary spontaneous emission
R.H.Dicke, Phys.Rev. 93, 99 (1954)
13
Sample shape and mode structure
High gain in directions of high optical depth
L
2w
Diffraction angle Geometric angle
Flt1 single mode dominant
14
Modes and competition

• Backreflected light and recoiling atoms
• Forward scattering with state change

State change constrained by dipole pattern
15
Rayleigh scattering dominant

• Favor Rayleigh scattering by choice of detuning
  and polarization
• Backward reflected light and recoiling atoms
• Forward scattering with state change suppressed

First experiments in end-pumped geometry
16
End-pumped superradiant scattering(first
experiments)

• in-trap illumination
• - 1.8 GHz detuning from F1 ? F1
• 2 1011 photons/s through BEC cross section
• immediate release after pump pulse

Rayleigh scattering dominant for these
parameters! Threshold expected after 103
incoherent events Dynamics slower than Dicke
model prediction Possible reasons collisions,
longitudinal structure, photon depletion,
misalignment,
17
Dynamics in experiment and simple models the
light side
Detect reflected light to observe dynamics
directly
Simulated pulse shape from modified rate equation
model
Backscattered light for different pump powers

• Setup for reflected light detection
• balanced detector
• shot-noise sensitive at 105 photons
• focused pump beam

Reasonable but not yet satisfactory
agreement Refined model needed
Comparison experiment and model
18
Dynamics in experiment and simple models the
atom side

• clearly observable but poorly understood
  structures in original and recoiling cloud
• separation of the clouds does not match photon
  recoil
• 3-D modeling of expansion urgently needed!
• high population of scattering halo

Modeling the role of collisions

• decoherence
• gain reduction

19
Can we use it?

• Backscattered photons and atoms should be fully
  correlated
• (in fact, entangled) but we need to show it!
• Challenges
• count backscattered photons to better than N1/2
• count recoiling atoms reliably
• keep atom-atom collisions during expansion low
• quantitative modeling of the dynamics

Atom-detection
Photo-detection

• high Q.E. CCD detector implemented
• pump geometry changed to avoid stray light
  background

• Cross calibration with different methods
• more atoms than initially estimated

20
Counting atoms and photons...the hard work
recoiling atoms
passive atoms
with atoms
without atoms
Need to improve background reduction in light
detection
Need to reduce noise level in atom detection
21
What do other people do?
arXiv/cond-mat/0707.1465v1
Atom-Atom entanglement by super-radiant light
assisted collisions
Also here the challenge is actually detecting the
entanglement
22
Future directionsQuantum memory
Access to internal atomic degrees of freedom
Use of light polarization degree of freedom
Forward scattering with state change
Funnily enough, we might need to
suppress Super-radiance as a competing channel
23
Under construction Optical dipole trap

• state insensitive trapping potential
• matched aspect ratio for easier transfer

Achromat lens f60mm
Trap beam

• diode lasers at 827 nm (P 100 mW)
• shared optics with probe beam
• stable confinement without magnetic fields
• scattering into probe mode below 100 ph/s
• compatible with magnetic bias field control
• flexible trap geometry

Probe beam
Trap beam
Collaboration with Marco Koschorreck (ICFO)
24
Who did the actual work?
Andrew Hilliard Franziska Kaminski Rodolphe Le
Targat Marco Koschorreck Christina
Olausson Patrick Windpassinger Niels
Kjaergaard Eugene Polzik
Funding by Danmarks Grundforsknings Fund,
EU-projects QAP and EMALI