Physics for the 21st Century logo

Section 4: Making an Optical Molasses

Adjustment of one of the dye lasers.

Figure 9: Adjustment of one of the dye lasers.

Source: © Hau Laboratory. More info

We have two laser systems on the table, each generating an intense yellow laser beam. The lasers are "dye lasers," pumped by green laser beams. A dye laser has a solution of dye molecules circulating at high pressure. This dye was originally developed as a clothing dye; but if illuminated by green light, the molecules emit light with wavelengths covering a good part of the visible spectrum: from red to greenish yellow.

In our apparatus, a narrow ribbon of dye solution is formed and squirted out at high velocity. We then hit the dye with a green laser beam inside a laser cavity where mirrors send light emitted by the dye molecules right back on the dye to stimulate more molecules to emit the same kind of light, at the same wavelength and in the same propagation direction (see Figure 9).

For the right length cavity, we can build up a large intensity of light at a particular wavelength. The cavity length should be an integer number of the desired wavelength. By intentionally leaking out a small fraction of the light, we generate the laser beam that we use in the experiments. How do we know that the wavelength is right for our experiments—that it is tuned on resonance with the atoms? Well, we just ask the atoms. We pick off a little bit of the emitted laser beam, and send it into a glass cell containing a small amount of sodium gas. By placing a light detector behind the cell, we can detect when the atoms start to absorb. If the laser starts to "walk off" (in other words if the wavelength changes), then we get less absorption. We then send an electronic error signal to the laser that causes the position of a mirror to adjust and thereby match the cavity length to the desired wavelength. Here—as in many other parts of the setup—we are using the idea of feedback: a very powerful concept.

So, perhaps you've started to get the sense that building an experiment like the one described here requires the development of many different skills: We deal with lasers, optics, plumbing (for chilled water), electronics, machining, computer programming, and vacuum technology. That is what makes the whole thing fun. Some believe that being a scientist is very lonely, but this is far from the case. To make this whole thing work requires great teamwork. And once you have built a setup and really understand everything inside out—no black boxes—that's when you can start to probe nature and be creative. When it is the most exciting, you set out to probe one thing, and then Nature responds in unanticipated ways. You, in turn, respond by tweaking the experiment to probe in new regimes where it wasn't initially designed to operate. And you might discover something really new.

Now, back to optical molasses: In a matter of a few seconds, we can collect 10 billion atoms and cool them to a temperature of 50 microkelvin (50 millionths of a degree above absolute zero). At this point in the cooling process, we turn off the lasers, turn on an electromagnet, and trap the atoms magnetically. Atoms act like small magnets: They have a magnetic dipole moment, and can be trapped in a tailored magnetic field. Once the laser-cooled atoms are magnetically trapped, we selectively kick out the most energetic atoms in the sample. This is called evaporative cooling, and we end up with an atom cloud that is cooled to a few nanoKelvin (a few billionths of a degree above absolute zero).

Evaporative cooling into the ground state.

Figure 10: Evaporative cooling into the ground state.

Source: © Sean Garner and Lene V. Hau. More info

According to the rules of quantum mechanics, atoms trapped by the magnet can have only very particular energies—the energy is always the sum of kinetic energy (from an atom's motion) and of potential energy (from magnetic trapping). As the atoms are cooled, they start piling into the lowest possible energy state, the ground state (see Figure 10). As a matter of fact, because sodium atoms are bosons (there are just two types of particles in nature: bosons and fermions), once some atoms jump into the ground state, the others want to follow; they are stimulated into the same state. This is an example of bosons living according to the maxim, the more the merrier. In this situation, pretty much all the atoms end up in exactly the same quantum state—we say they are described by the same quantum wavefunction, that the atoms are phase locked, and move in lock-step. In other words, we have created a Bose-Einstein condensate. Our condensates typically have 5–10 million atoms, a size of 0.004", and are held by the magnet in the middle of a vacuum chamber. (We pump the chamber out with pumps and create a vacuum because background atoms (at room temperature) might collide with the condensate and lead to heating and loss of atoms from the condensate).

It is interesting to notice that we can have these very cold atoms trapped in the middle of the vacuum chamber while the stainless-steel vacuum chamber itself is kept at room temperature. We have many vacuum-sealed windows on the chamber. During the laser cooling process, we can see these extremely cold atoms by eye. As described above, during the laser cooling process, the atoms absorb and reemit photons; the cold atom cloud looks like a little bright sun, about 1 cm in diameter, and freely suspended in the vacuum chamber.

Now, rather than just look at the atoms, we can send laser beams in through the windows, hit the atoms, manipulate them, and make them do exactly what we want...and this is precisely what we do when we create slow light.