QUANTUM MECHANICS

MOT Lab
Guide

ANDREW M. C. DAWES
C HAPTER 1

Getting
Started

The MOT system is a complex apparatus

that requires many steps to operate cor-
rectly. The first time through it will seem
cumbersome, but as you get a sense of
how each knob and mirror fits into the
puzzle, you will move through it easily.
S ECTION 1 Most of the controls used to operate the MOT system are lo-
cated on the equipment rack shown here. There are three la-
Instrument Rack sers in the system, and each laser has a current controller, a
temperature controller, and a scan controller. In some cases,
these are in the same physical box. The first laser you need to
learn is the trap laser. It generates the light used to confine at-
oms in the MOT. While it is the most important, it is also the
most straightforward to operate. All controls for the trap laser
are on the moglabs box in the middle of the rack.

C ONTROL INSTRUMENTS LOCATED HERE

1. Trap laser control (current, temp. and scan)

2. Repump laser temperature
3. Repump laser current
4. Repump laser scan
5. Probe laser current & temperature
6. Probe laser scan
7. CCD monitor for trap image
8. Oscilloscope for trap laser scan

2
S ECTION 2 scan appears on the scope as shown to the right. Use the fre-
quency (C) and span (D) knobs to shift and zoom respectively.
Trap Laser When the scan matches the photo here, use the error offset
knob (E) to bring the blue line into the scan as shown. The
blue line should cross the zero-voltage axis just to the left of
the resonance blip. At this point, the laser is ready to lock, so
flip switch (F) from scan to lock.

The trap laser needs to be tuned to the F = 2 F = 3 reso-

nance in Rubidium-87. This resonance is a cycling transition
which means the atoms excited to the F = 3 state primarily re-
lax to the F = 2 ground state.

The laser power is enabled by switch (A) and the current ad-
justed by knob (B). Usually the current is close to the correct
value from day to day. Adjust the current until the resonance

B C D
F
A E
3
S ECTION 3 The repump laser has a
dedicated box for each
Repump Laser control (temperature,
current, and scan). The
temperature control
should not need adjust-
ment. To tune this laser,
adjust the current until a
shallow dip is visible on
the scope.

If the dip is not centered, use the offset knob on the scan con-
Even in the cycling transition, not all atoms excited by the trol to shift the scan onto resonance. Finally, turn the ampli-
trap laser decay back to the F = 2 ground state. Those that re- tude fully CCW to park the laser on the resonance.
lax to the F = 1 state must be pumped back to the F = 2 state
to interact with the trap laser again. This repumping action is
provided by the repump laser. To operate correctly, it must be
tuned to the F = 1 F resonances.

Repump Current Control Repump Scan Control

4
 Probe Temp. and Current Control
S ECTION 4

Probe Laser

The probe laser is used to perform measurements of the trans-

mission (and absorption) of cold atoms within the MOT. This These absorption dips correspond to the resonances in Rb-87
laser does not need to be adjusted in order to trap atoms in and Rb-85 (left to right).
the MOT. We will use it for experiment III later in the lab.

As you did with the repump D OPPLER B ROADENING

laser, adjust the current until
The wide, smooth lines on these laser scans are
a scan with two dips appears
due to doppler broadening that arises when atoms
on the scope.
in a sample move fast enough to experience
This may require some in- frequency-shifted laser light. As an example, an
structor assistance as the la- atom moving toward the laser will absorb light
ser tends to drift more from day to day. that is actually below resonance for stationary
Use the ramp amplitude and atoms.
manual knobs to scale and
shift the scan to zoom in on the
left (shallow) dip.

5
6
S ECTION 5 This is the hardest part of the lab, but you can do it.

Making a MOT Turn the Rb Getter current on (switch located on the mini-
MOT). The current should be between 3.15 and 3.3 A. This pro-
vides Rb atoms for the trap, and it takes a few minutes to load
the cell with atoms. Instead of waiting, make sure the trap and
repump lasers are on resonance, and practice locking the trap
laser.

When the lasers are all tuned to resonance, and the trap laser
is locked, turn on the magnet power supply (near the MOT).
Watch the camera for a mot cloud as you lock the trap laser
and gently adjust the ERROR OFFSET. This is what makes it
hard, the knob is sensitive so you dont need to turn it much
(small, slow turns by about as little as you can manage).

The ERROR OFFSET knob shifts the lock point for the laser
and you should see the MOT cloud start to form. As you tune
the lock point, wait whenever you see the mot form as it takes
a few second for the cloud to grow to a reasonable size.

O BJECTIVES

1. Lock trap laser and park repump laser on

resonance
2. Achieve a stable MOT cloud

7
C HAPTER 2

Lab activities

Trapping atoms in the MOT is an excel-

lent accomplishment but it is also just
the beginning. The following lab activi-
ties provide information about the atoms
trapped in the MOT and allow you to
characterize the trap itself. You will meas-
ure the number of atoms trapped in the
MOT, the fill rate of the trap, and the ab-
sorption spectrum of the cold atoms.
S ECTION 1 The number of atoms held in the MOT cloud can be deter-
mined by measuring the fluorescence of the cloud using a
Lab I: Atom Number nearby detector.

The number of photons scattered per atom per second into 4

steradians (i.e. the full spherical surface around the MOT) is
given by the following expression:

I/Is
R=
2 1 + I/Is + 4(2 / 2)

with the atomic linewidth = 2 6 MHz, laser detuning

2 14 MHz, total trap laser intensity I and saturation in-
tensity Is 4.1 mW/cm2.

These scattered photons are emitted in every direction so you

must take into account the fact that you only collect a certain
fraction of them with the detector. You will also have to con-
vert a measured power level into a reading of photons per sec-
ond. To complete this calculation, you need to remember the
wavelength of the light we use (780 nm) and relate that to fre-
quency and hence to energy.

O BJECTIVES Making the measurements

1. Observe MOT fluorescence with an amplified
photodiode.
2. Measure total trap and repump laer power
3. Calculate the number of trapped atoms
The detector is already in place near the MOT. Be sure the de-
tector power is turned on and the detector output appears on
the oscilloscope. This detector is amplified and will max out at
10V under normal room lights. Room lights will not harm the
detector, but you will have to make the actual measurement in
the dark.

9
The most important part of this measurement is an accurate
assessment of the background light (i.e. light that does not
come from the trap cloud). There are several ways to do this,
but keep in mind that blocking the laser beams eliminates the
trap fluorescence and a source of background light.
Using the magnets is a more reliable way to turn off the trap
without affecting the amount of measured light. The figure be-
low shows a data trace where the detector sees the MOT and
the background light, then the magnetic field is switched off
and on and the trap begins to re-form. You are interested in
the difference in optical power from with-the-trap to without-
the-trap.
Scatter Detector (V)
Analysis
The detector calibration has been performed already:
0.92 0.01 V/W
You can use this calibrated output to calculate the detected
power based on the measured voltage. Compare the observed
power to the expression for the expected power scattered by
the atom cloud and solve for the number of atoms in the
cloud. Be sure to take into account the fact that we only detect
a small portion of the light scattered from the MOT.

0.06

0.05

5 10 15 20
Time (s)

10
S ECTION 2 The trap fill rate is the rate at which atoms accumulate in the
trap. This rate indicates a variety of things but the parameter
Lab II: Fill Rate it depends on most strongly is the background vapor pressure
in the trap cell. If there is a lot of Rubidium in the vapor cell
(i.e. a high background pressure) the trap fills quickly as at-
oms are readily caught. However, this also means atoms
leave the trap quickly and it is not always desirable. On the
other hand if the background pressure is low, atoms accumu-
late slowly and tend to remain in the trap for a longer period
of time. For these reasons, the fill rate is a reasonable measure
of the background pressure in the vapor cell.

We do not have a readily-accessible measure of the vapor pres-

sure in our MOT as the system is sealed and self-contained.
This technique is an excellent way to determine the conditions
in the cell.

Making the measurements

Use the same detector that you used for the Fill Rate experi-
ment. It should already be connected to the scope. You can of-
ten use the same set of data that you collected for Lab I.

O BJECTIVES
1. Measure the trap fill rate
2. Calculate the background vapor pressure
With the MOT running and the trap full, set the scope to col-
lect at least 25 s of data (it will be in untriggered mode). With
a bright MOT, turn the magnet power supply off and back on
while the scope is near the beginning of a full trace. The detec-
tor signal should drop several mV (hopefully at least 20 mV)
and begin to increase along an exponential growth curve. This
growth is the brightness of the MOT increasing as atoms fill
the trap.
11
 t=?s

Analysis
In your graphical data program of choice, fit or model an expo-
nential growth function to the data. You only need to recover
the time constant from the data, but it is important to allow
other parameters in your fit or model in order for it to be accu-
rate. Expect a result on the order of 10 seconds.

12
S ECTION 3 One question that often comes up is why do we want to trap
atoms? The main reason is that if we dont trap them, they
Lab III: Cold atom absorp- tend to fly all over the place. Room-temperature atoms move

tion spectrum at several hundred meters per second. This means they inter-
act with each other and with beams of light very briefly. If we
want to have a sample of atoms to interact with repeatedly, or
for a long time, we need to hold them still.

The spectrumtransmission vs. laser frequencytells us what

laser frequencies are absorbed by an atom sample. You have
used spectroscopy already to get the lasers on resonance. In
those cases, the atoms the laser hits are moving around and in
equilibrium at room temperature. We will compare the spec-
trum measured in those atoms to the spectrum measured in
the MOT cloud.

Making the measurement

O BJECTIVES
1. Tune the probe laser to resonance
2. Measure probe laser transmission through the
active MOT cloud
3. Calibrate the frequency axis of the spectrum
data
4. Compare warm-vapor spectrum to cold-vapor
spectrum
The probe laser is already aligned to the MOT cloud, but you
first need to tune it to resonance. Using the current control
and the scan control shown in Section 4, scan the probe laser
across resonance and observe the transmission on the scope.
For this measurement the scope should be connected as
follows:
CH1 CH2 EXT TRIG
MOT probe PROBE sat abs PROBE SCAN
and set to show CH1 and CH2 while triggering on the rising
edge of the EXT trigger. The horizontal scale should be set to

13
10 ms per div., adjust the vertical scales of each channel to see
the full signal. The screenshot below, should be your goal.

Once the MOT cloud is well formed, three distinct dips should
appear in your transmission scan. These correspond to the
three resonances in Rubidium vapor. Note they also corre-
spond to three of the features in the saturated absorption
scan.

0.5
0.7
Probe signal (V)
Sat. Abs. (V)

0.4 0.6

0.5
0.05 0 0.05
Time (s)
14
C HAPTER 3

Appendix 5 Data Tables

193.7407(46) MHz
gF o=o2/3
(0.93 MHz/G)
F=3
25

266.6500(90) MHz

52P3/2
72.9112(32) MHz

gF o=o2/3 F=2
229.8518(56) MHz (0.93 MHz/G)
302.0738(88) MHz 156.9470(70) MHz

gF o=o2/3 F=1
72.2180(40) MHz (0.93 MHz/G)
F=0

780.241 209 686(13) nm

384.230 484 468 5(62) THz
12 816.549 389 93(21) cm-1
1.589 049 462(38) eV

gF o=o1/2 F=2
(0.70 MHz/G)
2.563 005 979 089 109(34) GHz

52S1/2
6.834 682 610 904 290(90) GHz

4.271 676 631 815 181(56) GHz

gF o=o-1/2 F =1
(-o0.70 MHz/G)

Figure 2: Rubidium 87 D2 transition hyperfine structure, with frequency splittings between the hyperfine energy
levels. The excited-state values are taken from [9], and the ground-state values are from [29]. The relative hyperfine
shifts are shown to scale within each hyperfine manifold (but visual spacings should not be compared between
manifolds or to the optical splitting). The approximate Lande gF -factors for each level are also given, with the
corresponding Zeeman splittings between adjacent magnetic sublevels.

Additional information relevant to the

lab activities.
16