The thermometer gave an awkward answer
In 1988, Paul Lett, William Phillips and their colleagues released sodium atoms from a region where six laser beams had slowed their motion. The atoms spread much more slowly than the simplest cooling theory predicted. Their paper reported temperatures as low as 43 microkelvin, with an uncertainty of 20 microkelvin. The expected lower limit was about 240 microkelvin.
A temperature this low could signal a faulty thermometer or a missing piece of physics. The group compared four ways of measuring the released motion, including timing atoms in flight and releasing them before attempting to recapture them. The consistent low temperatures made the discrepancy worth taking seriously.
Read the evidence: NBS (1988), Center for Atomic, Molecular, and Optical Physics technical activities, cooling below the cooling limitP. D. Lett et al. (1988), Observation of Atoms Laser Cooled below the Doppler Limit, PRL 61, 169–172
Why shining light can make atoms colder
Light carries momentum. An atom that absorbs a photon gets a small push in the direction the photon was traveling. With two beams aimed against each other, the pushes would balance for a stationary atom if both beams were equally easy to absorb.
Tune the beams a little below an atomic absorption frequency. A moving atom sees the beam ahead shifted upward in frequency and closer to the absorption peak. It scatters more photons from that direction and receives more pushes against its motion. This is Doppler cooling. Three opposing pairs can slow the random motion in three dimensions.
The word random matters. All the atoms could drift in the same direction together without being hot. Temperature describes how widely their velocities differ around the average. Following the cloud’s center and following its expansion answer different questions.
Read the evidence: Lett et al. (1988), Observation of atoms laser-cooled below the Doppler limitNBS (1988), Center for Atomic, Molecular, and Optical Physics technical activities, cooling below the cooling limit
The simplest theory balanced braking against recoil
Each spontaneous emission sends a photon in a random direction and kicks the atom the other way. The kicks produce heating, even as the average absorption force produces braking. A theory with only one lower and one upper atomic level balances those effects and predicts a minimum temperature in weak light.
For the sodium transition used in these experiments, that Doppler minimum is roughly 240 microkelvin. It is a definite prediction made under definite assumptions. At that temperature the atoms still move. The experiment’s slower expansion asked whether those assumptions included all the processes operating in real sodium.
Read the evidence: Lett et al. (1988), Observation of atoms laser-cooled below the Doppler limit
The result survived changing the thermometer
A small, bright cloud does not directly reveal its temperature. On release, its size, the directions of the initial velocities, gravity and the detector geometry all affect the signal. Atoms that start farther from a probe arrive at different times even when their velocities match. An initially compact sample can also hide a wide velocity spread.
The four ballistic methods sampled the released atoms in complementary ways. Agreement across them made a mistaken reading of one geometry less plausible. The temperatures were also sensitive to magnetic fields and to the polarizations of the cooling beams. Those sensitivities pointed toward details of the atomic states and of how the light coupled to them.
Read the evidence: NBS (1988), Center for Atomic, Molecular, and Optical Physics technical activities, cooling below the cooling limitP. D. Lett et al. (1988), Observation of Atoms Laser Cooled below the Doppler Limit, PRL 61, 169–172
The atom had internal routes for losing more energy
Sodium’s ground state contains several magnetic sublevels, or different internal angular-momentum orientations. Polarized light can move atoms between them through repeated absorption and emission. The distribution of atoms among these states takes time to respond as an atom crosses a changing polarization pattern.
That extra internal motion opens cooling mechanisms missing from the two-level picture. In the Sisyphus mechanism, for example, atoms lose speed climbing state-dependent energy hills, then change state near the top and begin climbing again. The unexpectedly cold atoms had exposed an oversimplified description of the atom.
Optical molasses supplies braking without a force toward a fixed center. A magneto-optical trap adds position-dependent restoring forces using a field gradient and polarized light. Cooling and storing atoms became related experimental tools with distinct jobs. The 1997 Nobel Prize recognized Chu, Cohen-Tannoudji and Phillips for developing methods to cool and trap atoms with laser light.
Read the evidence: NBS (1988), Center for Atomic, Molecular, and Optical Physics technical activities, cooling below the cooling limitNIST, The Nobel Moment: Bill PhillipsLett et al. (1988), Observation of atoms laser-cooled below the Doppler limit