A strontium

1
A strontium detective story
Ions?
James Millen
Strontium detective Group meeting 19/10/09
2
The scene
Create aMOT
Strontium detective Group meeting 19/10/09
3
Rydberg excitation
5s18d 1D2 or 5s18s 1S0
Oc
420nm
?c 2px0.3MHzfor D state 1
5s5p 1P1
460.7nm
Op
?p 2px32MHz
5s2 1S0
Strontium detective Group meeting 19/10/09
4
The crime
Integrated ion signal at each coupling laser
frequency
Ion signal or Ion signal height
Loss fraction or Resonant loss fraction
Fraction of atoms lost during Rydberg excitation
Strontium detective Group meeting 19/10/09
5
Clues
Coupling laser power
Strontium detective Group meeting 19/10/09
6
Clues
Probe laser power (D state only)
Strontium detective Group meeting 19/10/09
7
Clues - timing
4µs probe andcoupling pulse
Time evolution of the ion signal
4µs 3.6Vcm-1E field pulse
?t 0
?t 4µs
Strontium detective Group meeting 19/10/09
8
The suspects
Electric field ionization 1
For 5s18d state require 5.4kVcm-1
With our field of 4Vcm-1 should only be able to
field ionize n97 and above
Strontium detective Group meeting 19/10/09
9
The suspects
Blackbody ionization
Gallagher (the book!) gives a rough estimate
This yields tbb 5µs for n 18 and T 900K.
More recent work by some Russians 1 yields a
lower rate. The lifetime of the 5s18d 1D2 state
is 640ns 2
Strontium detective Group meeting 19/10/09
10
The suspects
Auto-ionization (two electron excitation) 3
5s18s state is closest to the auto-ionizing
resonanceover 300GHz away! However, the width
of the resonance is 200GHz! So only 1.6
linewidths away
s
5s18s
s
5s5p
p
s
5s2
s
Strontium detective Group meeting 19/10/09
11
The suspects
Hot atoms
Collisions with hot Rydberg atoms could cause
ionization 4
The Zeeman slowed atoms have been ruled out
Could a small fraction of our hot Sr atoms get
excited to the Rydberg state? How do we test this?
Strontium detective Group meeting 19/10/09
12
The suspects
Collisions
Rydberg atoms collide (with resonant
dipole-dipole interactions)
This leads to ionization, plasma formation,
avalanche effect. It has also been observed to be
extremely fast (sub µs) 5 6
Tentative calculations suggest were exciting
40 of our sample to the Rydberg state
Strontium detective Group meeting 19/10/09
13
The Culprit?

• Electric field Unlikely, unless population is
  being transferred to higher n states
• Black body Looks like its too slow, though it
  can redistribute population
• Auto-ionization need to look in more detail. Is
  scaling with power as youd expect?
• Hot atoms pass
• Collisions not seen a signature of plasma
  formation yet needs more work

Strontium detective Group meeting 19/10/09
14
References
1 Ionization of Rydberg atoms by blackbody
radiation N. N. Bezuglov et. al.
arXiv0807.2535 (2009) 2 Natural radiative
lifetimes in the interacting 1,3D2 sequences in
SrS. Svanberg et. al. Phys. Rev. A 27 947
(1983) 3 Sr 5p1/2ns1/2 and 5p3/2ns1/2 J1
autoionizing statesT. F. Gallagher et. al.
Phys. Rev. A 33 2401 (1986) 4 Spontaneous
evolution of Rydberg atoms into an ultracold
plasmaP. Pillet et. al. Phys. Rev. Lett. 85
4466 (2000) 5 Ionization due to the
interaction between two Rydberg atomsRobicheaux
J. Phys. B At. Mol. Opt. Phys. 38 S333
(2005) 6 Autoionization of an ultracold
Rydberg gas through resonant dipole couplingM.
Weidemüller et. al. arXiv0903.3147 (2009)
Strontium detective Group meeting 19/10/09