In all the laser types that we have discussed so far stimulated electromagnetic transitions have lead to a coherent stream of photons, which we call laser light. The transitions that we have considered so far are atomic, molecular, or involve electron-hole re-combination in semiconductors. We have examined how pumping gases, liquids, or solids with electrical, chemical, or optical energies can produce population inversion to make laser action sustainable. While some of these laser systems require optical resonators with high Q-factors, others can produce laser light even without an optical resonator. The laser systems that we have examined range in power from a few mW to tera-Watts, in wavelength from the IR to the UV, and in operating modes from cw to femtosecond pulses. There are two other laser types, however, which are substantially different from all of these lasers: free-electron lasers and atom lasers.
The Free-Electron Laser (FEL) requires an electron accelerator for its operation. The electron beam is first accelerated to reach velocities comparable to the speed of light. At these velocities the so-called classical laws of physics need to be modified, according to Einstein's Theory of Relativity, in order to describe the electrons. So, the beam is said to contain "relativistic" electrons. The accelerating relativistic electrons are then passed over a series of permanent magnets that create a region of oscillating magnetic field. These magnetic field regions cause slight oscillations in the path of the electron beam; so the magnets are referred to as "wigglers". Because accelerating charges radiate, as the electron beam passes over the wigglers it radiates. By proper design of the magnetic field an external "seed" of light causes the relativistic electron beam to generate coherent laser light in the direction of travel of the beam. An optical cavity, similar to those in most lasers, amplifies the light emitted from the electron beam and produces the output laser light.
In the above diagram the accelerating electron beam, shown with solid blue lines, is bent to pass over the wiggler field. The seed light is injected from the left through the high reflector and the output laser light emerges from the output coupler, traveling to the right.
Unlike the light output from all other types of lasers, the laser radiation from the FEL is not the result of an electromagnetic transition in a bound system (atom, molecule, or a solid). Here, the interaction of the (Free) electron beam with the external (seed) input light creates a coherent beam of light. So, in the strict sense of the word, the FEL is not a laser, because it is not the result of stimulated emission of radiation. Of course, there is also no population inversion in an FEL. In the most commonly pulsed electron beams it is the double Doppler effect that generates the coherence in the output light of an FEL. The primary Doppler effect arises when the traveling electron "sees" the wiggler frequency to be shifted from its value in the stationary laboratory frame. Because of the relativistic speed of the electron the centimeter spatial periodicity of the field appears to it to be comparable to the wavelength of the light. The second Doppler shift is in the wavelength of the seed photons. These two Doppler shifts along with variations in the energy of the electron pulse are the design parameters that allow the FEL to generate coherent radiation over a wide range of wavelengths.
In a typical FEL the electron beam is about a millimeter in diameter and made of picosecond pulses of electrons separated by a few nanoseconds. The electrons are accelerated in the MeV range, typically around 50 MeV, but in some accelerators GeV electrons are used to make the FEL. What makes FELs especially important sources is that they can generate a wide range of tunable laser light, both in the far IR, from 3 to 100 mm, to the far UV well below 150 nm. FELs are not just unique lasers, but unique sources of light as well. As a result, FELs have developed a wide variety of applications from industrial, to scientific, to medical, to military. But because these lasers require a particle accelerator and possibly storage ring facilities, they are very expensive to build and operate. Still, because any electron accelerator capable of generating sufficiently fast electrons could be used for an FEL, many such facilities around the world are now generating this kind of laser light. ( For a comprehensive, albeit technical, review of FEL see the article by W. B. Colson, E. D. Johnson, M. J. Kelly, and H. A. Schwettman in the January 2002 issue of Physics Today , published by the American Institute of Physics. )
An even more exotic laser that has recently been developed (circa 1997) is the atom laser. What makes this laser type very special is that its output is not a coherent beam of photons, but instead, an atom laser is a source of coherent atoms, i.e. coherent matter.
Since the early 1980's scientists have managed to trap and slow down a variety of atomic species in magneto-optic traps. These traps use a spatially variable magnetic field along with six intersecting laser light beams to confine atoms in a small region of space. The effective interaction of light and the magnetic field with the atom not only traps it, but it can also be used to drastically slow its speed, and therefore lower its temperature.
Above diagram shows the two electric current coils (in blue) that generate the magnetic field for the atom trap, as well as the six intersecting laser beams (red arrows). The trapped atom cloud is shown at the point of intersection of the laser beams.
According to the laws of thermodynamics, temperature is a measure of "motion". The faster molecules of a gas move about, the higher the temperature of the gas. So, the lowest possible temperature is the so-called "absolute zero," a state of essentially completely frozen motions. In the Kelvin temperature scale (K), for example, room temperature (20 oC, or 68 oF) is 313 K. Ice-water is at 293 K and liquid air is at 77 K. The background cosmic radiation that is leftover from the Big Bang heats up the empty space to a toasty temperature of about 4 K! Magneto-optic traps, in comparison, routinely attain mK temperatures (that is 0.000,001 K). At these ultra-cold temperatures, atoms of the vapor (or gas, if you like) that are trapped move around at very slow speeds; about a few cm/s. At temperatures still closer to absolute zero some of he atomic species can exhibit a very peculiar effect. This effect, which is quantum mechanical in origin just like stimulated emission, and was also first proposed by Einstein, is referred to as Bose-Einstein Condensation (BEC).
Statistics and the laws of quantum theory classify all particles into two separate camps. In one, any number of particles can have identical properties so that they become indistinguishable from one another. Photons fall into this camp and it is because of this that we can have laser light. In laser light, as we have seen, all photons have the same wavelength, phase, direction of motion, etc. That is to say, coherence is a consequence of the fact that indistinguishable photons can occupy the same space. The particles, like photons, that belong to this "indistinguishability camp" are called bosons, named after the Indian born physicist Bose who first calculated the statistics of these particles. Photons are not the only bosons. Other particles and even atoms fall into this camp. In a cloud of cold atoms held in a trap even though the atoms are very far from each other they interact to reach a steady-state value of speed, i.e. equilibrium temperature. Because of this, different atoms (even the bosonic atoms) end up with slightly different speeds, and therefore lack coherence. Here is where Einstein (and separately, Bose) predicted the novel phenomena of the BEC, namely that at every temperature there is a critical density for these non-interacting bosons, so that once this density is attained all motion comes to a stop; i.e. the gas condenses into a new phase of matter. In the condensate, then, all constituents occupy the same space and so the vapor reaches coherence.
In atom lasers the trapped atoms are cooled low enough in temperature so that they form a Bose-Einstein Condensate from the vapor cloud. In analogy to the conventional photon lasers the medium for this laser is the atomic vapor, portions of which form the condensate. Instead of the optical cavity the magneto-optic trap plays the role of amplifier by confining the laser's atoms in the condensate. By partially de-confining the condensate portions of it are allowed to leave the trap, similar to using the partially reflecting (confining) output coupler for the photon laser output.
Several experiments to date have demonstrated the coherence and "laser-like" properties of atom lasers. New developments in the technology of nanofabrication are suggesting the promise of an "atom laser on a chip" in the foreseeable future. In the early years of the development of photon lasers no one predicted the many applications they now have. In fact, the laser was initially called "a solution looking for a problem". It took roughly 20 years of research and development before semiconductor lasers became marketable. Are we now seeing the beginning of a new generation of lasers leading to now unthinkable applications? It is certain that as laser technologies have given us an indispensable tool for better understanding optics, atom lasers will allow us to investigate many interesting properties of matter waves.
( For a recent review, albeit technical, of atom lasers see the article by E. W. Hagely, Lu Deng, W. D. Phillips, K. Burnett, and C. W. Clark in the May 2001 issue of Optics & Photonics News, a publication of Optical Society of America. )
