Laser cooling of atoms

The concept of radiation-pressure cooling of atoms was independently suggested in 1975 for the case of a gas of neutral atoms by Hänsch and Schawlow, and for atomic ions bound in an electromagnetic trap by Wineland and Dehmelt. While the notion that momentum exchange from a photon moving in the opposite direction could slow an individual atom was well understood, until this time no one had come up with a means for producing an aggregate cooling of a larger ensemble of atoms (a gas). If all atoms in a hot gas absorb photons, then some will be heated and some cooled, and the ensuing equilibrium temperature is not lowered. The general feature of the cooling concepts is that a gas of atoms or ions can be cooled by ensuring that photon absorption takes place preferentially when the atoms or ions are moving against the flow of photons from a laser. In 1978, following these ideas, Wineland, Drullinger, and Walls performed their seminal experiment [1] in which they demonstrated the very first radiationpressure cooling below ambient temperature of any atomic species. The key to the experiment was the variation in photon absorption associated with the Doppler frequency shift. They used a collection of positive magnesium ions contained in an electromagnetic trap subjected to laser radiation near the 280 nm resonance of the magnesium ion. When this laser radiation was tuned slightly below resonance, cooling to below 40 K was observed. For this particular tuning, those ions with motions opposing the laser radiation are Doppler shifted toward resonance and are more likely to absorb photons, thus slowing their motions. Ions moving away from the source are Doppler shifted further from resonance and are thus less likely to absorb photons. Since the re-radiation from this excited state is symmetric, the net effect averaged over the ensemble of atoms is a cooling of the gas of ions. The very next year, Wineland and Itano [2] published a paper providing the first detailed theoretical analysis of laser cooling, which served as the foundation for the rapid development of this field. In ensuing years, they improved their methods and soon cooled ions to millikelvin temperatures. This experimental demonstration stimulated the development of a large number of ion-cooling groups around the world and encouraged others to attempt to cool neutral atoms. In fact, it was only a few years later (in 1982) that a beam of neutral atoms was cooled by Bill Phillips and his collaborators at NIST (as described elsewhere in this volume). These ideas have contributed significantly to atomic-clock technology. Clocks using both trapped ions [3] and cooled neutral atoms [4] have now demonstrated frequency uncertainties of a few parts in 10, an improvement of an order of magnitude over conventional atomic-clock technology. Further improvements will certainly be demonstrated over the next few years. The potential of the cooled-ion standards can be appreciated by recognizing that, for a small group of ions, the systematic frequency shifts are now understood at an uncertainty level of 1 part in 10. Based on this early work, NBS established an Ion Storage Group in Boulder; the Group now includes five full-time staff members and a number of postdoctoral associates, guests, and students. After the initial cooling experiments, the methods were improved, but particularly striking results were obtained by cooling at the sideband frequency created by the periodic motion of the trapped ions. Using this method, the Group achieved, for the first time, cooling to the zero-point energy of motion [5,6], the fundamental limit for any cooling technique. Fig. 1. A schematic diagram of a linear ion trap using alternating and static electric fields to confine linear arrays or “strings” of ions. The expanded ultraviolet image at the bottom shows the fluorescence image of an array of positive mercury ions.