Steven Chu
Nobel Physicist

PAGE 2 OF 6

Steven Chu
Laser Super-Cooler

GS: Because of the size aspect?
SC: There are fundamentally four different forces. Two are nuclear forces. Strong and weak nuclear forces, so short range that they only exist at nuclear distances on the order of a hundred thousand times -- I'm doing a little calculation -- so the size of a nucleus is 10 to the minus 13 centimeters, the size of an atom is 10 to the minus 8 centimeters, so we're talking eight orders of magnitude, a 100 million times smaller than an atom. So those forces you can't use. The other is gravity. You would need earth-sized objects in order to use gravitational forces. So what is left is electricity, magnetism and electromagnetic. And one can prove a theorem that says that you cannot hold onto an atom, a ground-state atomm using static forces. You can use static magnetic forces and the only [other] thing left is light. So you're left with two options, a magnetic field taylored in a clever way or light.

GS: Just so I can understand the setup -- you were directing the six lasers as though you're trying to enclose the atoms within the six sides of a box?
SC: No, the six lasers weren't used to trap the atoms. Let me back up. Historically what the world was focused on at the time was, first you trap an atom and perhaps then you can use light to cool the atom down. The reason people were focused on that sequence -- this sounds really trivial but that's how life works --- people were able to trap ions using electromagnetic fields and in particular combinations of electric fields or magnetic fields to trap particles with net charge. After they trap these particles with net charge in these electromagnetic traps, then there were proposals to cool the atoms down to very low temperatures. In fact, a nobel prize had been given to the developers of the so-called ion trap in something like 1989, maybe earlier. So people with ion traps who had done this work starting in the 1970s were leading this charge. Atoms were much harder to deal with because they were neutral. There's no net charge. And remember we're confined to electric, magnetic or electromagnetic fields. So the idea was to trap them and then think about cooling them. Steve Chu

GS: So cooling is not synonymous with trapping?
SC: No.

GS: So you trapped them using the ion method, then you cooled them?
SC: No, quite the opposite. The little insight was, cool them first. Not only cool them but you cool them with these six beams going along six different directions.

GS: That's counter-intuitive. How do you cool an object by directing a beam of energy at it.
SC: The way you do it is to take account of the fact that atoms only interact with light in a very narrow frequency range, so-called resonance. When an atom is traveling toward the laser beam there's another effect called the Doppler shift. You tune the light so that when the atom moves in the direction opposing its motion, it's automatically tuned into resonance. Meanwhile it's running away from a laser co-propagating with the motion of the atom so it's running away from that laser beam. So the doppler that the atom sees on the laser beam pushing it faster is actually tuned out of resonance, whereas the frequency the atom sees on the laser beam opposing its motion is tuned into resonance. So it's actually the local motion of the atom which enables one to break this symmetry so an atom moving toward the laser beam scatters many more protons from the beam opposing the motion than co-moving with the motion. So think of the atom as absorbing a particle of light, a proton. Every time it absorbs a proton it gets a little impulse telling it slow down. It gets more impulses on the front side than the back side. And you do this in three dimensions, the same thing occurs. Does that make sense?

GS: Slowing it down is cooling it?
SC: Yeah. I forgot to tell you that temperature is defined to be the average kinetic energy of a whole group of atoms and the kinetic energy is proportional to the velocity squared. So going lower in temperature a factor of two lower in average velocity means a factor of four lower in temperature.

GS: You have to be able to determine what direction the atom is traveling in and then apply the laser beam.
SC: If it weren't for the doppler shift, that wouol be absolutely true. You would have to say the atom is moving to the right, so the we're going to hit it with a beam going to the left. But the beauty of the doppler effect is you don't really have to know which way a particular atom is going. If an atom goes to the right it gets hit more from a laser beam opposing its motion and if an atom goes to the left, the same thing occurs because the frequency shift due to the atom's own velocity.

GS: So it would always absorb more of the laser that's opposing it.
SC: It's very analogous to having some particle in some molasses, syrup. If you push this little marble or something in the syrup, if you push to the right, if you push to the left, up or whatever, the thing wants to slow down. In fact the equations look very similar. It looks like what we call viscous damping force. The force is actually proportional to the speed of the motion of the particle. And that's why I called it optical molasses. It's important to realize that optical molasses is not a trap. A trap is something that says, go to the center, go to a point, the trap center. Because of what we call Browning motion, because the atoms are actually scattering photons all the time, it ends up analogous to Browning motion where you have a small particle in a cup of water. The bouncing of this atoms on this particle make it jostle around and it has this little jiggly motion. But there's also this viscous drag that tells it to slow down and stop. So there's eventually some equilibrium temperature which is the balance of the viscous drag forces telling it to stop and this jiggling due to the random scattering of the water molecules. This is actually a very deep concept in the sense that the dissipation due to the water molecules is the same thing which causes the fluctuations and there is a theorem that says that dissipation or drag is intimately connected to the fluctuations. That enables one to figure out what the temperature of the atom should be due to these forces.

GS: Was there a eureka moment when you felt you had succeeded?
SC: There were a number of eureka moments, meaning when you're working on a problem it doesn't all come as one Eureka. Typically what happens is, there is a succession of realizations that are days, weeks, months apart, where you begin to realize that some initial idea was much better, much more powerful than one thought. For me the Eureka was two things. First it was reversing the order. Instead of first trapping, then cooling, what happens if you just cool the atoms. People were worried about cooling means you decrease the average volocity, the temperature. A moment for me in hindsight -- it sounds utterly simple because it's undergraduate physics -- but for me it was, once they're cold in this optical molasses, they begin to do a random walk in space. The nature of a random walk means that it doesn't really know which way it goes. That means that it actually lingers in certain position in space for a very long period of time. And the optical molasses without the actual trap suddenly looked better than any trap optical trap that I and others were considering. Even without the cooling, just the optical molasses would keep an atom around for a good fraction of a second in a small volume of space a centimeter on a side.
     That was a eureka moment. It was not in the consciousness of most physicists at the time. Once you say, ah this optical molasses is so powerful it can get them cold, so cold that you can trap them, and it can keep them around long enough so you have time to load the trap. At that point you say ah ha, this is gonna work.

GS: Is that when you were high-fiving your post-doc or did that come later?
SC: I think there were two steps. One is, I went to him and said, This is going to work. Then I went to my boss who is the director of the laboratory -- I was the department head -- and I said, This is so simple, it can't fail. He kind of looked askance at me. [laughs] The reason is previously Art Ashkin had assembled a group of Bell Labs scientists that were working on the problem and they had been working on it for about five or six years. Then they cut the funding off. Because he said, Look you've done some nice science here, you've shown that these forces exist but you're nowhere near beginning to trap atoms, so let's go on to other things.

GS: What was the other eureka moment?
SC: So the first moment was forget about trapping, just cool first. And after you get them cooled you got a shot at trapping. The other eureka moment was, we still have this problem about the large volume. We tried for eight months in my laboratory to make the trap of the kind that Art Ashkin designed. First we designed a large-vo;lume trap but theoretical calculations showed that's not going to work.

GS: What do you mean by a large-volume trap?
SC: Suppose you had atoms in an optical molasses confined to a region in space on the order of a half centimeter or centimeter in diameter. You need a trap that can hold onto a large fraction of those atoms. That means the volume over the trap can exert substantial forces would be on the scale of a half a centimeter on a side. That's what we mean by a large volume. I tried six or eight months to make a trap with such a large volume but couldn't. The density of atoms we had in optical molasses was high by laser-cooling standards, [laughs] enormously high by laser-cooling standards because it was the first. It was about a million per cubic centimeter. So those traps didn't work.
     There was one trap that we thought might work which is just a simple focused laser beam but the volume of that trap in order to get the forces to come out right it would have to be something like 10 to the minus seven cubic centimeters. That's very small. It's tens of microns on a side. If the density is only 10 to the 6th atoms per cubic centimeters, that means on average you would only find one tenth of one atom in the trap at any given time. So that type of trap was automatically rejected. No one in his right mind would ever consider that because you had a background of a million atoms and on average you only get one-tenth of an atom, maybe one atom.
     So the second Eureka moment came only after working with the bigger trap failed. I again thought of this random walk problem, and the idea here is that as the atoms randomly walked about, if they would land in the trap because you simultaneouly had cooling and trapping, they would roll down this hill into this little well picking up speed but since you're cooling, you would take away that energy and those atoms would be trapped. But atoms nearby that weren't initially in that volume, well they're randomly walking about, so they also would have opportunity to find this little dip in the energy. PAGE 3

Page 1 | 2 | 3 | 4 | 5 | 6

“The density of atoms we had in optical molasses was high by laser-cooling standards, enormously high by laser-cooling standards because it was the first. ”