Transcript Optical Trapping of Atoms

Optical Trapping of Atoms:
Characterization and Optimization
Charlie Fieseler
University of Kentucky
UW REU 2011
Subhadeep Gupta
So… why?
• Studying superfluid/degenerate gas properties
– Two species, Lithium and Ytterbium, can use one as
a probe
– Condensed matter simulations
• Molecule formation, specifically polar
– Quantum computing in a lattice
• Fundamental physics, of course
– Fine structure constant measurement through Yb
BEC atom interferometry
– Electric dipole moment of electron
Getting Cool Atoms: Laser cooling
• Zeeman Slower
– absorbs on-resonance light in the atom’s frame of
reference, with a magnetic field to counteract Doppler
• MOT (also using Zeeman effect)
– A 3D trap that catches the cooler atoms
– Oppositely polarized light is preferentially absorbed
• After compression, ends up:
– >10^6 atoms at ~20 μK
Getting Cold Atoms
• Using an optical dipole trap (ODT), cool
evaporatively (two step)
• >2*10^4 Yb atoms below ~170nK (critical
temperature), and can go below 30nK
• >10^4 Li atoms below ~300nK, and can go
below 100nK
• With both species, can cool
sympathetically (different trap depths)
Optical Dipole Trap
ODT (cont.)
• The atoms are high-field seeking, i.e.
optical tweezers
• None of the other measurements make
any sense unless you know the waist and
position of the focus
• Cameras are usually used, but they can
be quite expensive and (firsthand) very
unreliable
• Do something simple: razorblade
A different method of beam
profiling
• A more conventional method scans
perpendicular to the beam
– This is not very sensitive to small waists
– Hard to know where the minimum is
• For the proof of concept, the beam is
single-mode with a Gaussian shape
– A scan along the axis of propagation can
measure small waists with <5% error
The shape of the beam
• This method can also measure deviations from a
Gaussian shape
• A Gaussian intensity function gives the power by
integrating:
Setup: single-mode fiber
Setup: “razor” blade
3D Power as a function of razorP
position
z
x
It really does fit well!
Progression of plots
Modeling the trap geometry
• What do you want the waists to be?
• In reaching degeneracy, trapping
frequencies (i.e. of an harmonic oscillator)
are key:
Optimizing (or at least a first guess)
• Symmetrical makes the most sense: same
power, circular, same size
• But then gravity… poof, nonlinear
• Harder to model, but there are some
theoretical benefits
– Weaken dependence of frequency on trap
depth: if gravity were tunable, could get down
to .075 from .5
Effects of Gravity
• To the right: Trap at
10W and .25W
• The trap becomes
dominated by gravity
at low power: two
effects
– Lower exponent
– Smaller curvature and
therefore coefficients
Trap depth vs. frequency
The aforementioned first guess
• The trap disappears in one dimension
before the others
• The power in that beam should be held at
a minimum, while the other beam
continues the evaporation
• Gravity is not a large enough effect to
break the symmetry earlier
Next steps
• Actually build this setup!
– Will be used for a Ytterbium BEC
interferometry experiment
• The curves shown do not really show a
benefit, but other tweaks need to be
tested.
References
• http://lanl.arxiv.org/PS_cache/arxiv/pdf/11
05/1105.5751v1.pdf
• NWAPS 2010 (Walla Walla, WA) Invited
Talk by Deep Gupta
• http://grad.physics.sunysb.edu/~fdimler/ind
ex1.html