English
Swedish

15 October 1997

The Royal Swedish Academy of Sciences has decided to award the 1997 Nobel Prize in Physics jointly to

Professor Steven Chu, Stanford University, Stanford, California, USA,

Professor Claude Cohen-Tannoudji, Collège de France and École Normale Supérieure, Paris, France, and

Dr. William D. Phillips, National Institute of Standards and Technology, Gaithersburg, Maryland, USA,

for development of methods to cool and trap atoms with laser light.

Atoms floating in optical molasses

At room temperature the atoms and molecules of which the air consists move in different directions at a speed of about 4,000 km/h. It is hard to study these atoms and molecules because they disappear all too quickly from the area being observed. By lowering the temperature one can reduce the speed, but the problem is that when gases are cooled down they normally first condense into liquids and then freeze into a solid form. In liquids and solid bodies, study is made more difficult by the fact that single atoms and molecules get too close to one another. If, however, the process takes place in a vacuum the density can be kept low enough to avoid condensation and freezing. But even a temperature as low as -270°C involves speeds of about 400 km/hr. Only as one approaches absolute zero (-273°C) does the speed fall greatly. When the temperature is one-millionth of a degree from this point (termed 1 µK, microkelvin) free hydrogen atoms, for example, move at speeds of less than 1 km/hr (= 25 cm/s).

Steven Chu, Claude Cohen-Tannoudji, and William D. Phillips have developed methods of using laser light to cool gases to the µK temperature range and keeping the chilled atoms floating or captured in different kinds of “atom traps”. The laser light functions as a thick liquid, dubbed optical molasses, in which the atoms are slowed down. Individual atoms can be studied there with very great accuracy and their inner structure can be determined. As more and more atoms are captured in the same volume a thin gas forms, and its properties can be studied in detail. The new methods of investigation that the Nobel Laureates have developed have contributed greatly to increasing our knowledge of the interplay between radiation and matter. In particular, they have opened the way to a deeper understanding of the quantum-physical behaviour of gases at low temperatures. The methods may lead to the design of more precise atomic clocks for use in, e.g., space navigation and accurate determination of position. A start has also been made on the design of atomic interferometers with which, e.g., very precise measurements of gravitational forces can be made, and atomic lasers, which may be used in the future to manufacture very small electronic components.

Slowing down atoms with photons
Light may be described as a stream of particles, photons. Photons have no mass in the normal sense but, just like a curling stone sliding along the ice they have a certain momentum. A curling stone that collides with an identical stone can transfer all its momentum (mass times velocity) to that stone and itself become stationary. Similarly, a photon that collides with an atom can transfer all its momentum to that atom. For this to happen the photon must have the right energy, which is the same as saying that the light must have the right frequency, or colour. This is because the energy of the photon is proportional to the frequency of the light, which in turn determines the latter’s colour. Thus red light consists of photons with lower energy than those of blue light.

What determines the right energy for photons to be able to affect atoms is the inner structure (energy levels) of the atoms. If an atom moves the conditions change because of what is termed the Doppler effect – the same effect that gives a train whistle a higher pitch when the train is approaching than when it is standing still. If the atom is moving towards the light, the light must have a lower frequency than that required for a stationary atom if it is to be “heard” by the atom. Assume that the atom is moving in the opposite direction of the light at a considerable speed and is struck by a stream of photons. If the photons have the right energy the atom will be able to absorb one of them and take over its energy and its momentum. The atom will then be slowed down somewhat. After an extremely short time, normally around a hundred-millionth of a second, the retarded atom emits a photon. The atom can now immediately absorb a new photon from the oncoming stream. The emitted photon also has a momentum, which gives the atom a certain small recoil velocity. But the direction of the recoil varies at random, so that after many absorptions and emissions the speed of the atom has diminished considerably. To slow down an atom an intensive laser beam is needed. Under the right conditions effects can be achieved with a strength corresponding to what would be seen if a ball was thrown upwards from the surface of a planet with a gravity 100,000 times the Earth’s.

Doppler cooling and optical molasses
The slowing down effect described above forms the basis for a powerful method of cooling atoms with laser light. The method was developed around 1985 by Steven Chu and his co-workers at the Bell Laboratories in Holmdel, New Jersey. They used six laser beams opposed in pairs and arranged in three directions at right angles to each other. Sodium atoms from a beam in vacuum were first stopped by an opposed laser beam and then conducted to the intersection of the six cooling laser beams. The light in all six laser beams was slightly red-shifted compared with the characteristic colour absorbed by a stationary sodium atom. The effect was that whichever direction the sodium atoms tried to move they were met by photons of the right energy and pushed back into the area where the six laser beams intersected. At that point there formed what to the naked eye looked like a glowing cloud the size of a pea, consisting of about a million chilled atoms. This type of cooling was named Doppler cooling.

At the intersection of the laser beams, atoms move as in thick liquid, and the name optical molasses was coined. To calculate the temperature of the atoms cooled in the optical molasses the lasers were switched off. It was found that the temperature was about 240 µK. This corresponds to a sodium atom speed of about 30 cm/s, and agreed very well with a theoretically calculated temperature – the Doppler limit – then considered the lowest temperature that could be reached with Doppler cooling.

The atoms in the above experiment are cooled, but not captured. Gravity causes them to fall out of the optical molasses in about one second. To really capture atoms, a trap is required, and a highly efficient one was constructed in 1987. It was called a magneto-optical trap (MOT). It uses six laser beams in the same sort of array as in the experiment described above, but has in addition two magnetic coils that give a slightly varying magnetic field with a minimum in the area where the beams intersect. Since the magnetic field affects the atoms’ characteristic energy levels (the Zeeman effect) a force will develop which is greater than gravity and which therefore draws the atoms in to the middle of the trap. The atoms are now really caught, and can be studied or used for experiments.

Doppler limit broken
Magnetic fields had already been used at the beginning of the 1980s by William D. Phillips and his co-workers in a method of slowing down and completely stopping atoms in slow atomic beams. Phillips had developed what was termed a Zeeman slower, a coil with a varying magnetic field, along the axis of which atoms could be retarded by an opposed laser beam. With his device Phillips had in 1985 stopped and captured sodium atoms in a purely magnetic trap. Enclosure in this trap, however, is relatively weak, for which reason the atoms within it must be extremely cold to remain inside. When Chu managed to cool atoms in optical molasses Phillips designed a similar experiment and started a systematic study of the temperature of the atoms in the molasses. He developed several new methods of measuring the temperature, including one in which the atoms are allowed to fall under the influence of gravity, the curve of their fall being determined with the help of a measuring laser.

Phillips found in 1988 that a temperature as low as 40µK could be attained. This value was six times lower than the theoretically calculated Doppler limit! It turned out that the Doppler limit had been calculated for a simplified model atom that had previously been considered sufficiently realistic. However, Claude Cohen-Tannoudji and his co-workers at the École Normale Supérieure in Paris had already in theoretical works studied more complicated cooling schemes. The explanation of Phillips’ result lay in the structure of the lowest energy levels of the sodium atom. What happens can be likened to Sisyphus’ endlessly rolling his stone up the slope, but in this case finding that the slope beyond the crest is also an uphill one. The comparison has led to the process being termed Sisyphus cooling.

The recoil velocity an atom gains when it emits a single photon corresponds to a temperature termed the recoil limit. For sodium atoms the recoil limit is 2.4 µK and for the somewhat heavier cesium atoms about 0.2 µK. In collaboration with Cohen-Tannoudji and his Paris colleagues Phillips showed that cesium atoms could be cooled in optical molasses to about ten times the recoil limit, i.e. to about 2 µK. It first appeared that in optical molasses it was generally possible to reach temperatures only about ten times higher than the recoil limit. In a later development both Phillips and the Paris group have showed that with suitable laser settings it is possible to trap the atoms so that they group at regular intervals in space, forming what is termed an optical lattice. The atom groupings in the lattice occur at distances of one light wavelength from each other. Atoms in an optical lattice can, as has been shown, be cooled to about five times higher temperature than the recoil limit.

Recoil limit also broken
The reason why the recoil velocity an atom obtains from a single photon sets a limit to both Doppler cooling and Sisyphus cooling is that even the slowest atoms are continually being forced to absorb and emit photons. These processes give the atom a small but not negligible speed and hence the gas has a temperature. If the slowest atoms could be made to neglect all the photons in the optical molasses, perhaps lower temperatures could be reached. One mechanism through which a stationary atom can be caused to assume a “dark” state in which it does not absorb photons, was known. But a difficulty was to combine this method with laser cooling.

Claude Cohen-Tannoudji and his group between 1988 and 1995 developed a method based on use of the Doppler effect and which converts the slowest atoms to a dark state. He and his colleagues showed that the method functions in one, two and three dimensions. All his experiments use helium atoms, for which the recoil limit is 4 µK. In the first experiment two opposed laser beams were used and a one-dimensional velocity distribution was achieved which corresponded to half the recoil limit temperature. With four laser beams a two-dimensional velocity distribution was achieved, corresponding to a temperature of 0.25 µK, sixteen times lower than the recoil limit. Finally with six laser beams a state was attained in which the whole velocity distribution corresponded to a temperature of 0.18 µK. Under these conditions helium atoms crawl along at a speed of only about 2 cm/s!

Applications just round the corner
Intensive development is in progress concerning laser cooling and the capture of neutral atoms. Among other things, Chu has constructed an atomic fountain, in which laser-cooled atoms are sprayed up from a trap like jets of water. When the atoms turn at the top of their trajectory and start falling again, they are almost stationary. There they are exposed to microwave pulses that sense the atoms’ inner structure. With this technique it is believed that it will be possible to build atomic clocks with a hundredfold greater precision than at present. The technique rewarded this year also forms the basis for the discovery of Bose-Einstein condensation in atomic gases, a phenomenon that has attracted great interest.

Steven Chu was born 1948 in St. Louis, Missouri, USA. American citizen. Doctoral degree in physics 1976 at the University of California, Berkeley. Theodore and Frances Geballe Professor of Humanities and Sciences at Stanford University 1990. Among other awards Chu received the 1993 King Faisal International Prize for Science (Physics) for development of the technique of laser-cooling and trapping atoms.

Professor Steven Chu
Physics Department
Stanford University
Stanford, CA 94305
USA

Claude Cohen-Tannoudji was born 1933 in Constantine, Algeria. French citizen. Doctoral degree in physics 1962 at the École Normale Supérieure in Paris. Professor at the Collège de France 1973. Member of, among other institutions, the Acadèmie des Sciences (Paris). Among many prizes and distinctions Cohen-Tannoudji received the 1996 Quantum Electronics Prize (European Physical Society) for, among other things, his pioneering experiments on laser cooling and the trapping of atoms.

Professor Claude Cohen-Tannoudji
Laboratoire de Physique de École Normale Supérieure
24, Rue Lhomond
F-75231 Paris Cedex 05
France

William D. Phillips was born 1948 in Wilkes-Barre, Pennsylvania, USA. American citizen. Doctoral degree in physics in 1976 at the Massachusetts Institute of Technology, Cambridge, USA. Among other awards Phillips has received the 1996 Albert A. Michelson Medal (Franklin Institute) for his experimental demonstrations of laser cooling and atom trapping.

Dr. William D. Phillips
National Institute of Standards and Technology
Gaithersburg, MD 20899 USA