What are common gases in the universe

Quantum Physics: The Coldest Gas in the Universe

Suppose you are purring to the size of a molecule by magic and you can observe the movement of individual atoms in a gas. The particles in front of you are like unbreakable glass marbles that flit around in an almost empty room and keep bouncing off one another. Perhaps when you look at it, you will remember the description of the ideal gas from your school days.

Suddenly you notice that the marbles are racing back and forth less hectically. Aha! Some process gradually cools the gas down. At first the marbles only lose speed and move closer together: The density of the gas increases as it cools. But then, to your surprise, you see that the marbles themselves change. The size of the slowest is growing a thousandfold, and its previously mirror-clear surface has now become completely blurred. These ever more shadowy atoms penetrate each other, sometimes without deflection, then again with a recoil, as if they had collided with something hard inside.

Right in front of you, two of the slowest, fuzzy atoms overlap and appear to merge into a single larger bubble. This ellipsoid absorbs more atoms individually, in pairs or by the dozen, and with astonishing suddenness only one thing remains of all the confusion: a huge motionless zeppelin. What happened to all the individual atoms? What kind of mysterious object is that?

In front of you is a purely quantum mechanical structure, a so-called Bose-Einstein condensate, the coldest form of a gas in the universe. And although the atomic components in this gas still exist, they have lost all individuality.

Usually the bizarre properties of quantum mechanics remain hidden behind the facade of classical physics. We confuse this facade with reality itself, and this is where our everyday ideas of how the world works arise: for us, every object has its well-defined position, movement and identity, and its behavior is precisely determined by deterministic laws.

On the other hand, the nature of quantum mechanics opposes our view: the location and movement of particles are fundamentally uncertain and determined by probabilities. Even the idea that things have an unmistakable identity has to be given up for quantum particles. A Bose-Einstein condensate is a collection of matter that follows a purely quantum mechanical behavior pattern with a clarity that has hardly been observed before.

Particularly noteworthy is the enormous size of such condensates - 100,000 times more extensive than the largest ordinary atoms, even larger than human cells. Therefore, physicists can observe the quantum behavior of a condensate in a normally unthinkable vividness. Steven L. Rolston from the National Institute of Standards and Technology (NIST) in Gaithersburg, Maryland, emphasizes: "Our images of Bose-Einstein condensates are real images of quantum mechanical wave functions - we can really see quantum mechanics in action experience."

Gaseous Bose-Einstein condensates were first generated in the laboratory in 1995 - after all, only seventy years after Albert Einstein, based on ArbeBose, predicted the phenomenon (see "The Bose-Einstein Condensation" by Eric A. Cornell and Carl E . Wieman, Spectrum of Science 5/1998, p. 44). The experimenters produce these condensates in so-called atom traps - arrangements of laser beams and magnetic fields that trap, hold and cool a highly diluted cloud of atoms in a vacuum chamber (see box on page 55). The renowned atomic physicist Daniel Kleppner from the Massachusetts Institute of Technology (MIT) calls the generation of these condensates "the most exciting individual development in atomic physics since the development of the laser".

Research groups around the world, some led by Nobel Prize winners and future laureates, have been exploring the exotic territory made accessible by this breakthrough for the past five years. They scan the condensates with laser beams, vary the traps that hold them together and watch the gas bounce, rock and vibrate according to the quantum laws.

But condensates are not only exemplary quantum systems, they also embody a peculiar mixture of several large research areas: atomic physics (individual atoms), quantum optics (laser beams and their interaction) and many-particle physics (solids, liquids and gases) including the technologically important research into the flow of electrons in metals and Semiconductors.

Super liquid helium

This article is only able to give a few examples of the astonishing experimental feats that physicists have achieved with Bose-Einstein condensates. The results shed light on some facets of these quantum objects: their behavior as a superfluid like in liquid helium, as a precisely controllable atomic gas and as a kind of laser beam that consists of matter instead of light.

If liquid helium is cooled to less than 2.2 Kelvin (degrees above absolute temperature zero), it adopts a very strange behavior. As the Soviet physicist Pyotr Kapiza and the Canadian John F. Allen discovered in 1938, helium becomes superfluid below this temperature: It flows without any viscosity and is able to crawl up and out of the edge of an open container. These effects are based on the Bose-Einstein condensation (see box on this page).

The experimental physicists would like to know whether the gaseous condensates also show superfluidity, but finding the answer is proving to be extremely difficult. Superfluid helium can be produced in such large quantities that its strange behavior can be observed with the naked eye. The new condensates, on the other hand, are tiny gas clouds, hardly more substantial than a vacuum, and are held together by magnetic fields for a few minutes at most. What would it mean for such a delicate vapor to be superfluid anyway?

The quantum vortices created in a rotating superfluid are a unique effect. If you let a bucket of ordinary liquid helium rotate on a turntable, after a while all of the helium - much like water - follows the rotation of the bucket. On the other hand, separate, regularly arranged quantum vortices are formed in superfluid helium. The smallest possible rotation corresponds to a single vortex that rotates quickly in the middle of the helium and only slowly at the edge of the container. If you try to make the superfluid rotate even more slowly, it will remain completely motionless.

These effects occur because the atoms in the condensate are always in one and the same quantum state and must therefore all have the same angular momentum. But the angular momentum can only assume discrete quantum values. In the motionless state, all atoms have zero angular momentum; in a vortex they just have a quantum of angular momentum.

In 1999, Carl E. Wieman and Eric A. Cornell at the JILA (Joint Institute for Laboratory Astrophysics) in Boulder, Colorado, produced vortices in Bose-Einstein condensates for the first time using a method proposed by James E. Williams and Murray J. Holland had. They initially created a double condensate: two overlapping condensates from the same element (rubidium), but in slightly different quantum states.

The researchers irradiated the double condensate with microwaves and a laser and achieved that one of the condensates was impressed with exactly the quantum mechanical phase required for the formation of a vortex. This process, which to a classical physicist doesn't look like even an atom is moving, creates the rotating vortex. By observing how the two condensates interfere with each other, the researchers were able to directly demonstrate the quantum properties of the vortex; this had never been achieved in sixty years of working with superfluid helium.

Later that year, at the École Normale Superieure in Paris, Jean Dalibard was able to create vortices for the first time using the rotating bucket method. Dalibard's group moved a laser beam around the edge of the atom trap, creating a kind of rotating distortion of its shape. These researchers were able to map arrangements of up to 14 vertebrae. In September 1999 they published angular momentum measurements: In accordance with the theory, the value is zero until the first vortex appears and then immediately jumps to a whole angular momentum quantum.

The quantum dynamics of such eddies is not only interesting for basic research, but also for the technology of high-temperature superconductivity: Magnetic fields penetrate these materials by creating an arrangement of electrical current eddies. The movement of such flux eddies leads to power losses and thus destroys the most attractive property of superconductors - that they do not have any electrical resistance. Investigations on Bose-Einstein condensates could help to get this problem under control.

Manipulable interactions of atoms

In superfluid helium, the vortex nuclei are only a tenth of a nanometer (millionth of a millimeter) in diameter and can therefore hardly be examined in detail. But the nuclei of the eddies observed in Colorado and Paris are about 5000 times larger, because gaseous condensates have an extremely low density compared to liquid helium, and their atoms only interact very weakly with one another.

Almost nothing can be changed in the density and interactions of liquid helium, but the density of gaseous Bose-Einstein condensates can be varied by tightening or loosening the magnetic traps that hold the gas in place. It is also possible to change the interactions in gaseous condensates literally by turning an adjusting knob. This ability is every experimenter's dream: imagine what chemical research would be like if we could weaken or strengthen the bonds between atoms at will.

The atoms in a gaseous condensate experience weak mutual repulsion or attraction, depending on the type. For example, sodium, rubidium-87 or hydrogen atoms repel their own kind. Lithium-7 and rubidium-85 atoms, on the other hand, attract each other. Although these forces are tiny, they modify numerous properties of a condensate - for example its internal energy, its size, its vibration modes and the speed at which it is formed. Above all, repulsion stabilizes a condensate, while attraction has a destabilizing effect. Therefore, in experiments with repulsive rubidium-87 or sodium, millions of atoms are always condensed at the same time, and the condensates can be twenty times larger than if there were no repulsion. Conversely, the attraction limits the lithium-7 condensates from Randall G. Hulet's group at Rice University in Houston, Texas to about 1500 atoms. Above this size, the condensate contracts and becomes so dense that the atoms are thrown out of the trap by collisions. These results can now be easily explained by sophisticated theoretical models, but at the beginning of the 1990s physicists still doubted whether atoms that were attracted to one another were able to form a condensate at all.

The interactions of the atoms can be changed by so-called Feshbach resonances; the nuclear physicist Herman Feshbach from MIT investigated an analogous phenomenon in colliding atomic nuclei in the 1960s. In an ultra-cold gas, a strong magnetic field deforms the atoms and, at certain field strengths, causes a resonance between two colliding atoms. In a condensate the atoms feel the effect of these resonances continuously because their wave functions overlap one another; the resonances modify the forces between the atoms, with the strongest effects occurring in the vicinity of the resonance magnetic field strength.

One difficulty, of course, is that a strong magnetic field can destroy the magnetic confinement of the atoms. Wolfgang Ketterle's group at MIT solved this problem in 1998 by transferring sodium condensate from a magnetic trap into a laser trap. But although the MIT group succeeded in observing the effect of the Feshbach resonances, more detailed investigations were impossible: When the magnetic field was raised to a value close to a resonance, the sodium condensate disintegrated within a few thousandths to the great surprise of the researchers Seconds.

Long-lived condensates with variable interaction were developed in early 2000 by Cornell and Wieman using rubidium-85 and a conventional magnetic trap. Normally, the attractive interaction of rubidium-85 prevents the condensate from growing to more than a paltry 80 atoms. But by making the forces repulsive with the help of the Feshbach resonances, the group from Colorado succeeded in producing stable condensates from up to 10,000 atoms and with a lifespan of up to ten seconds.

The most spectacular effect occurred when the group gradually reduced artificial rejection. As theoretically predicted, the large condensate shrank and became denser. Eventually - about five milliseconds after the interaction became attractive again - the condensate exploded; Wieman jokingly dubbed this phenomenon "Bose Nova" because of its vague resemblance to the implosion that drives exploding stars. The explosions hurled about a third of the condensate atoms out of the trap and left a residual condensate surrounded by a hot atomic cloud - if you can call a temperature of a ten-millionth of a degree hot.

One possible application of interaction tuning in condensates is the generation of special atomic beams, so-called atomic lasers. Ordinary atomic beams are already used in all kinds of scientific and industrial applications, for example in atomic clocks, in the precision measurement of natural constants and in the production of computer chips. But all these rays lack the intensity and coherence of an atomic laser - just as ordinary light lacks the strength and coherence and thus the versatility of a laser beam. With coherence it is meant that all atoms or photons of a beam move in quantum physical mode: The associated waves oscillate in phase.

From condensate to atomic lasers

It took decades before the laser - in 1960 just an esoteric experimental device - became an almost ubiquitous part of consumer electronics. Some researchers believe that the atomic laser has a similarly diverse future in store for decades to come. However, enormous obstacles pile up on this hypothetical path, mainly because atomic beams - unlike laser beams - need a vacuum to propagate.

The first atomic lasers generated their pulses and rays very differently than optical lasers; therefore some even said that the term laser was misleading in this case. In essence, an atomic laser is nothing more than a coherent and freely moving piece of Bose-Einstein condensate. The atoms of a condensate are held in the magnetic trap because, due to their spin, they themselves act as tiny magnetic dipoles. When precisely tuned radio waves overturn the atomic spins, the atoms become immune to the surrounding field. Ketterle used this effect at MIT in 1997 to realize the first atomic laser. He irradiated a sodium condensate with pulsed radio waves. The atoms with their spin flipped over simply fell out of the trap; This resulted in sickle-shaped condensate packets that were only set in motion by gravity.

At the end of 1998, Theodor Hänsch's group at the University of Munich constructed a similar system that emitted a continuous beam of rubidium atoms (see Spektrum der Wissenschaft 7/2000, p. 23). The Munich group estimated that their atomic beam was more than a million times more intense than similar - but non-coherent - atomic beams produced by other techniques.

Around the same time, William D. Phillips and Steve Rolston built an atomic laser at NIST that did not only work in the downward direction. Optical laser pulses drove atoms out of the condensate and out through a circular hole on the edge of the trap. A sequence of laser pulses precisely synchronized with the rotation of the hole created a tightly bundled and practically continuous beam; one report spoke of an "atomic beam cannon with laser-like precision".

The "a" in laser stands for "amplification"(Amplification), but in the case of the atomic lasers described so far, the only significant amplification in the initial generation of the Bose-Einstein condensates takes place when the atoms" amplify "the common quantum state in the course of the condensation. A real amplification of the atomic laser beams - a so-called matter wave amplification - was only achieved at the end of 1999 by a group led by Ketterle and Pritchard at MIT and, independently of that, by Takahiro Kuga at the University of Tokyo.

Matter wave amplification does not mean that the amplification process creates matter from energy.Rather, a small atomic laser pulse is generated in a Bose-Einstein condensate, and this pulse is amplified when additional condensate atoms obey their Bose nature and join them. The simultaneous scattering of light from a pumped laser beam guarantees that momentum and energy are retained.

The MIT group recognized in early 1999 that matter wave amplification is possible in this way, when they aimed a polarized laser beam on one of their cigar-shaped condensates; To the surprise of the researchers, piles of atoms emerged below 45 degrees, and rays of light fell from both ends of the "cigar". These were scattering processes with a certain reinforcement effect; in this respect they resembled the so-called super radiance, a form of avalanche-like amplification of radiation.

Non-linear optics and braked light

In these processes, the condensates behave in a very similar way to light - in contrast to their behavior as a superfluid. A particularly lively research area in the past decade has been nonlinear optics, in which light interacts with itself. Researchers at the American Bell Laboratories, for example, used non-linear light pulses, so-called solitons, to send huge data packets through glass fibers.

Normally light hardly interacts with itself; therefore extremely high light intensities or special media are necessary to achieve non-linear effects. Since the weak interactions of the atoms in condensates automatically cause non-linear effects, the Bose-Einstein structures are ideally suited for investigating such processes. The classical idea of ​​atoms as particles colliding like tiny marbles completely fails in the interpretation of these experiments.

A feat of nonlinear optics is the enormous slowing down of light. In a vacuum, electromagnetic waves - whether radio, X-ray or light rays - propagate at an absolute top speed: 300,000 kilometers per second. Light travels more slowly in a medium: in water at about three quarters and in normal glass at two thirds the speed of light in a vacuum.

In 1999, Lene Vestergard Hau from the Rowlands Institute for Science in Cambridge (Massachusetts) braked a beam of light using an ultra-cold and optically modified gas to 17 meters per second - the speed of a fast bicycle. In November 2000, Ketterle's group reported that a beam of light had crossed a condensate at one meter per second, literally at walking pace. In and of itself, no condensate is required to achieve such effects, but the enormous coldness of the condensed gases creates ideal conditions for this.

Ulf Leonhardt and Paul Piwnicki from the Royal Institute of Technology in Stockholm speculated in 1999 that slowed light grazing a vortex in a condensate could serve as a miniature model for processes around rotating black holes. For example, the light could be drawn into the vortex core - especially if the beam moves against the rotational flux.

In work that has not yet been published, Peter Zoller and Ignacio Cirac from the University of Innsbruck show that it should be possible to build sound models of black holes with technology that is already available today - that is, those in which sound waves take on the role of light. According to their calculations, such structures explode and emit multitudes of sound quanta, so-called phonons, in the process. These explosions would simulate the evaporation of microscopic black holes, in which a thermal mixture of particles, known as Hawking radiation, escapes as a result of quantum effects.

In an August 2000 article, Wayne Hu and his co-workers at Princeton University speculate that invisible dark matter, which apparently makes up around ninety percent of the mass of the universe, could exist in the form of a Bose-Einstein condensate of extremely low-mass particles. If this bold hypothesis is correct, the coldest gases in the universe would also be the most abundant.


Experimental Studies of Bose-Einstein Condensation. By Wolfgang Ketterle in: Physics Today, Vol. 52, p. 30, December 1999.

Bose Condensates Make Quantum Leaps and Bounds. From Yvan Castin et al. in: Physics World, Vol. 12, p. 37, August 1999.

Relatives of the Bose-Einstein condensate

P> The condensates produced in 1995 were not the first examples of Bose-Einstein condensation. Her long-known relatives include:

Super liquid helium. Liquid helium-4 becomes superfluid when cooled to less than 2.2 Kelvin. The liquid flows without any viscosity and offers, among other things, the amazing spectacle of the helium fountain (right). The reason is that some - up to 10 percent - of the helium atoms Bose condensation occurs. Because of the strong cohesion of the atoms in the liquid, it is hardly possible to investigate the quantum properties of the condensate fraction theoretically and experimentally.

Laser. Laser radiation has a lot in common with a Bose-Einstein condensate. In ordinary light - like that of a light bulb - the light waves are not synchronized; but in a laser all waves oscillate in phase, that is, their mountains and valleys coincide exactly. Expressed in quantum physics, the light quanta - the photons - belong to the particle type of bosons, which strive to assume the same quantum state. The amplification process that creates a laser beam takes advantage of this tendency of bosons.

Superconductor. The Bose condensation of electron pairs is the basis of superconductivity - the resistance-free flow of current. Since unpaired electrons are not bosons but fermions, they cannot form a Bose condensate. Weakly bound electron pairs are only created under certain conditions, for example in aluminum below 1.2 Kelvin. Such pairs are bosons and will readily form a quantum condensate. The pairing mechanism and the electrical charge of the pairs make superconductors very different from neutral, dilute condensate. A similar pairing and condensation process occurs in superfluid helium-3, the atoms of which are fermions.

Excitons. In semiconductors, a missing electron can behave like a positively charged particle - a "hole". If a hole and an electron are generated by a laser pulse, both can form a common pair state for a short time, a so-called exciton. In 1993 it was observed that excitons can form a short-lived gaseous condensate in a copper oxide semiconductor.


Fermions are quantum particles that avoid their own kind: two fermions can never assume the same quantum state in the same place. They include electrons, protons and neutrons.

Bosons are the opposite: they strive to gather as many as possible in the same quantum state. The photons (light quanta) belong to them. Compound particles, especially atoms, are either bosons or fermions. An atom made up of an even number of protons, neutrons, and electrons is a boson.

Bose-Einstein condensation occurs when a collection of similar bosons is sufficiently cooled and compressed without forming a solid. In the condensate, all bosons have one and the same quantum state.

Devices for cooling and trapping atoms

Laser cooling. In order to generate a gaseous Bose-Einstein condensate, a dilute atomic gas has to be cooled to extremely low temperatures in a vacuum chamber. The first step is almost always laser cooling: laser beams slow down the movement of the atoms so much that their temperature is only around 50 microkelvins (millionths of a degree above absolute zero).

Magneto-optical trap. It combines laser cooling and trapping atoms with magnetic fields. The magnetic fields compress the gas. Often two such traps are used one after the other - the first primarily for trapping, the second specifically for cooling the atoms.

Evaporative cooling. The final cooling stage in experiments on Bose-Einstein condensates is similar to cooling a cup of coffee: while a magnetic field holds the atoms together, the hottest atoms are continuously removed, leaving cooler and colder gas behind. In contrast to laser cooling, evaporative cooling works best at higher densities.

TOP trap. The time-averaged orbiting potential trap, which Eric A. Cornell and Carl E. Wieman used to generate the first gaseous condensate in 1995, has been adopted by various groups. The magnetic coils generate a field, which however disappears at one point; therefore atoms could escape from the trap from there. By rapidly rotating the field, however, the atoms are trapped within the circular path of the leak and form an ellipsoid when they condense.

Ioffe-Pritchard traps. Such devices - named after the Russian physicist M. S. Ioffe, whose Ioffe trap was used to trap plasma from charged ions, and David Pritchard from MIT - create a trap field without a leak. They are the most important alternative to the TOP traps. Their magnetic fields are generated by means of electricity flowing through four parallel rods or by coils shaped like the letter D, the seams of a baseball or four-leaf clovers.

Permanent magnet trap. This type of Ioffe-Pritchard trap creates the fields with permanent magnets. Randall G. Hulet's group at Rice University in Houston, Texas uses this type of trap to produce condensates in lithium. Since the permanent magnets cannot be switched off, the condensate can only be observed at the point of origin.

The godfather of the Bose-Einstein condensate

A year or two after the first condensates were produced, Daniel Kleppner was introduced at conferences as the "godfather of the Bose-Einstein condensates". After all, he couldn't pass as their "father" because his own group had unfortunately still not produced any condensate. And yet he held his fatherly hand over the field - as a pioneer, as an active participant and as a mentor to the young climbers who had snatched the Holy Grail from him.

The three groups that created the first quantum condensates in 1995 and 1996 were led by Kleppner's students and their students. Wieman had worked as a student in Kleppner's laboratory in the early 1970s. Cornell was a PhD student from Pritchard who was in turn a PhD student from Kleppner. Ketterle initially worked on cold atoms under Pritchard. Hulet was a PhD student in Kleppner's group, as was Nobel Prize winner Phillips, whose group produced a Bose-Einstein condensate in 1998.

When Kleppner's earlier students created their spectacular condensates from the alkali atoms rubidium, sodium, and lithium, Kleppner was still struggling with the atom of his choice: hydrogen. He had already started doing this in the late 1950s as a doctoral student and postdoc at Harvard University. There he was involved with Norman Ramsey in the invention of the hydrogen measles; This relative of the laser works in the microwave range and is used, among other things, for high-precision measurements, for example when testing Einstein's theory of relativity. In 1996, Kleppner moved from Harvard University to MIT, where he is now the executive director of the electronics research laboratory.

Kleppner got involved with the Bose-Einstein topic around 1976 when he was working with so-called spin-polarized hydrogen. "I thought the idea was crazy," remembers Kleppner, but a young professor named Thomas Greytak changed his mind. The two have been working together ever since.

In spin-polarized hydrogen, the spins of all atoms are aligned in the same way; The spin can be thought of as a tiny magnetic compass needle that carries every atom with it. Such a gas is as inert as helium, since two hydrogen atoms must have exactly opposite spins in order to form a molecule. As the only one of all elements, this form of hydrogen should remain gaseous down to absolute zero.

In the late 1970s, Kleppner and Greytak at MIT and competitors at the University of Amsterdam hopefully tried to create a Bose-Einstein condensate in spin-polarized hydrogen; They never dreamed how long the search would take and that condensates made of metallic atoms would steal the show from them.

Even if Kleppner's group did not finish first, they made several decisive advances, such as the proof of evaporative cooling on spin-polarized hydrogen in 1987 - a masterpiece that the alkali atom groups could only repeat seven years later. By 1991, the Kleppner-Greytak group had reached the temperature and density required for condensation by a factor of three, while the alkali atoms were back then by a factor of a million. Unfortunately, there were some tricky properties of hydrogen standing in the way at this point; Among other things, it turned out to be difficult to measure important properties of the gas in order to detect a condensate. In the case of the alkali atom gases, visible light and normal laser technology can be used for this; but the corresponding light for hydrogen is ultraviolet and requires much more cumbersome methods.

In June 1998, two of Kleppner's students called him late at night: he should come to the laboratory quickly. A Bose-Einstein condensate in hydrogen had finally been observed. A month later, Kleppner announced the success of his group at a conference in Varenna, Italy. The experts present - colleagues, competitors and former students - celebrated the proud pioneer with a standing ovation.

From: Spectrum of Science 2/2001, page 50
© Spektrum der Wissenschaft Verlagsgesellschaft mbH

This article is included in Spectrum of Science 2/2001