Wolfgang Ketterle, "The Coldest Matter in the Universe" - 2001 Nobel Laureate Lecture in Physics

PRESENTER: Glad to see the usual audience at our physics colloquium.

[AUDIENCE LAUGHS]

You may think that I'm going to introduce Wolfgang Ketterle, but you'd be wrong. I'm introducing Dan Kleppner. Dan Kleppner came to MIT after getting his PhD, and serving for some years on the faculty at Harvard.

And he's really famous for many of the things he's done in atomic physics. For Rydberg atoms, making atoms as large as 1,000 angstroms in diameter. For many other things, and for his work with Tom Greytak on Bose condensation, and spin-aligned hydrogen.

Dan, when he came to MIT, began the establishment of what is now a truly great school of atomic physics. And the test of a great school is the students, and several of those students have won Nobel Prizes. Bill Phillips in 1997 was a student of Dan's, and one of this year's prizewinners, Carl Wyman, did his undergraduate thesis research with Dan.

Dan is a great physicist, a great educator. Those of you who've taken 8012 know his wonderful textbook. He's a great writer. He writes columns in physics today that are wonderful. And he's a great colleague. And it's my great pleasure to introduce Dan.

[APPLAUSE]

KLEPPNER: So success has many fathers, and some grandfathers. And today I am a very happy grandfather. And I suspect that Norman Ramsey over here is a very happy great-grandfather.

[AUDIENCE CHUCKLES]

[APPLAUSE]

I had the great good fortune to have Dave Pritchard as one of my first graduate students, and to have the experience of him as a colleague at MIT for many years. Dave has many accomplishments. This year he has mentored, directly, two of the Nobel Prize winners. Wolfgang Ketterle, and Eric Cornell was a graduate student of his.

He has made many seminal contributions to physics. He invented the field of atom optics. And he made seminal contributions to the work, which garnered the Nobel Prize in 1997 for atom cooling, and trapping. He invented the Ioffe-Pritchard trap. He did the seminal work on what's called the MOT, which is the workhorse in that area. And in 1990 he invited Wolfgang Ketterle to come as a post-doc to work with him on the atom cooling, even though Wolfgang had no background in that area. He spotted Wolfgang's qualities, and the two of them together did such fabulous work that it was clear that Wolfgang was headed for great things.

But not at MIT, because we could not appoint a young faculty member who would be in competition with a senior faculty member, or who would really be in close collaboration. Because we expect our faculty to be able to work independently. Dave overcame that obstacle by withdrawing from the field, and turning over his equipment to Wolfgang. So Wolfgang could start out his work towards Bose condensation at top speed.

And the rest of that is history, and you'll hear more about it. But to me this example of mentoring, which is to leave a field in which you're very interested in doing outstanding work to make room for young talent, is in the very best tradition of science and of teaching. And I hope in the very best tradition of MIT. So with this those thoughts, I would like now to introduce to you, Dave Pritchard.

[APPLAUSE]

PRITCHARD: Well, thank you very much Dan. By my account, this is the fourth time that Wolfgang has given an MIT physics colloquium here in the six years since he's been on the faculty, and up to the seventh now. This is a sign that, well, it's actually in the past six years. This is a sign that MIT recognized his worth somewhat before the Nobel Prize committee. Since I've had the honor of introducing him three times, I thought I might just summarize those introductions to give you an idea of his growth.

Now I know some of you were saying, wait a minute, three introducers, and now three introductions. But it's okay because we discussed it. And we realized that if each one of us talked a factor of three longer than the first, then the previous speaker, and Wolfgang followed suit, he'd still talk for more than 2/3 of the time. You can work that out.

So in the first introduction I said, Wolfgang was my post-doc, and we were fortunate to get him here to take over running my cold atom apparatus because he can run it so much better than I could. As expected, he has made fantastic progress. Now having made an observation of BEC, but with orders of magnitude more atoms than the condensate that Eric Cornell made at JILA.

And the second introduction not too much, but a year later I said, Wolfgang is absolutely courageous. With the best BEC, the biggest BEC in the world, and with his phone ringing off the hook with invitations. Neither of these is an irrelevant concern for a professor who doesn't have tenure. He decided to scrap his magnetic trap. He felt that a new cloverleaf magnetic trap that he designed would allow for better scientific studies of the condensate. As usual, he was right, and so on. So you want to hear something like that. That's the second one.

The third time I said, Wolfgang Ketterle needs no introduction. He--

[AUDIENCE LAUGHS]

[CHUCKLES]

He was just promoted from assistant professor to a full professor in a single jump, and he justifies this by doing spectacular research, and so on, and so forth. Now when I hear the trite phrase, needs no introduction, I often know the speaker myself. And I then agree, of course, with the introducer. But it always raises in my mind the question, well who's this guy who's doing the introduction? Doesn't he need an introduction?

And at this fourth introduction I see that the relative stature of Wolfgang has grown so disproportionately that many other people are now concerned with this issue. And that obviously explains today's format where there were two guys who introduced me.

[AUDIENCE LAUGHS]

Occasionally I feel that I should try to retain some of my mentorship. A vestige of that, and so I try to challenge Wolfgang to get even better. This is really difficult to do, because of all the people that I've worked with, which so far includes six Nobel laureates.

He's the only one who was A-plus at absolutely everything, scientific tastes, design, organization, group management, writing, scientific talks, public lectures, A-plus. So challenging them is like saying, well very good about jumping over all those tall buildings with a single bound, but if you could learn to jump over whole cities you wouldn't crush so many automobiles.

Now previously I've challenged him by guaranteeing that he would give a spectacular talk. And actually today I can raise the bar simply by challenging him to do that again, because I know that this is a week where he hasn't had much sleep. And he's had numerous distractions from the press, and from the many interviews. But I'm sure that he's up to it. One final note, just from my personal perspective. I want to say that your most important, A-plus ability is as a colleague and friend. My world has been tremendously brightened since you walked into my lab. And I'm glad to see that MIT is the same way.

[APPLAUSE]

[SIGHS]

KETTERLE: What a place, and what an introduction. I think you heard from the previous people who introduced me what nice a place this is. What great mentors I've had, and sometimes I feel I was privileged to finish what those people have started. And you will see that in my talk. But you, I think you've also heard from those introduction that this place, and the atomic physics community, or the physics department, is an inspiring and friendly environment. And it is this environment which brings the best out of us. Now Dave was raising the bar by saying, by sort of announcing an A-plus talk. I think this talk for me has two superlatives.

Number one, it is the best-attended lecture I've ever given.

[AUDIENCE LAUGHS]

[HUFFS]

But now I'm embarrassed to say number two. It's a worse prepared one. Because I was given very short notice, and I thought, so well, I'll just put a few note cards together. But the closer this talk came, the more I felt it's a very emotional talk for me. Because this is my family. This is my place. And I wanted to tell a story, and not just give a talk on our recent results. So I try to prepare something special, but I couldn't really polish it.

So let me tell this story of Bose-Einstein condensates, which is the coldest matter in the universe. And since I know here is a part audience, I want to be explicit in one of the concepts. So the ball's condensate, you can regard it as metal made of matter waves. It's a form of matter where the quantum mechanical nature, the wave nature of matter, manifests itself at a microscopic scale. So you're all familiar with the, or most of you I hope, with a particle-wave duality for light, for photons. But we have the same particle-wave duality for all particles.

We know from the early days of quantum mechanics that a particle, which is classically characterized by position and velocity, should either be described as a quantum mechanical wave, a de Broglie wave, and the wavelengths of the de Broglie wave of the quantum mechanical wave follows the famous de Broglie relation. It's inversely proportional to the velocity of the particle. And I know when we learn physics, we all struggled with the concept that particles are all objects. You, and me, and microscopic objects, and macroscopic objects are waves and particles at the same time. Particles propagate as waves, but when we detect them it makes click in a detector.

So well, if you and me feel all waves, why don't we perceive the wave nature of matter in everyday life? So let me raise you, let me raise the question, how do we perceive if something is a wave or not? So if I would challenge you and say, explain me why sound is a wave. A very reasonable answer would be, if there are two people who talk to each other, they can hear each other even if they don't see each other. Because the wave propagates, deflects around corners.

If I would ask you, what happens in the case of light? You all know that light is an electromagnetic wave. It's not so obvious, because the wavelengths of light is so short that light is just deflecting a little bit. And you can't really perceive the wave nature of light very easily. But now in the case of matter waves, the wavelengths are even shorter. The wavelengths is given by the de Broglie relation, and the velocity of a particle is determined by its temperature.

So from that we know that the colder the temperature, the longer is the de Broglie wavelengths. Now talking about atoms, at room temperature, the matter wavelengths, the de Broglie wavelengths, is smaller than the size of an atom. So therefore the wave nature of atoms is not manifest.

But now we can cool down, and the title of my talk is, The Coldest Matter in the Universe. We cool down to micro and nano Kelvin temperatures. More than a million times colder than interstellar space. And the de Broglie wavelengths becomes longer and longer. And this value is 30 micrometer is a fraction of the diameter of a human hair. So it's really getting microscopic. And this long de Broglie wavelengths, they are at the heart of the phenomenon which I want to describe. Bose-Einstein condensation.

What happens if you take a gas in a container, and cool it down? The particles slow down. As I just mentioned, we shouldn't describe the particles as sort of little billiard balls. They are wave packets. They are quantum mechanical waves. But as long as the wavelengths is very short, those wave packets are very localized. And we can follow the motion of those wave packets as if they were particles.

But now we can cool down. The de Broglie wavelengths becomes longer and longer. And when we reach the point where the wave packets, where the de Broglie waves overlap, then we cannot follow individual particles anymore. They become sort of a quantum soup of wave packets, and it's exactly at this point when the wavelengths, the de Broglie wavelengths is comparable to the spacing of particles that a new form of matter forms. And this is the Bose-Einstein condensate.

Or to say it again, if the de Broglie wavelengths become longer and longer, and those matter waves overlap, then all the particles in the gas, they start to oscillate in concert. And what they form is one giant matter wave. And this is the Bose-Einstein condensates. Just use an analogy. The difference between atoms in random motion, and the Bose condensate is exactly the same difference as the light from a light bulb, and the light coming out of a laser. So what Bose-Einstein condensate is about, it's about the creation of atoms with laser light properties.

And this is a nice artist's conception of how you can imagine the Bose-Einstein condensate. First of all, these are indistinguishable atoms. You see they have all the same facial expression, and they march in lockstep, which means they are in phase. It's a single wave function. And this is how you can imagine what happens in a Bose-Einstein condensate. You also see sort of those other guys. They walk in random direction. They're not Bose condensed. They are in different quantum states, and are therefore distinguishable from the atoms in the condensate.

Now of course, what I just said, that all the atoms go into one quantum state, and fall in lockstep is only possible for atoms which we'll refer to as bosons. Particles with have integer spin. I don't want to go into details here, but there is another class of particles called fermions. The have half integer spin, and their behavior at low temperature is completely different. So the phenomenon of Bose-Einstein condensation can be regarded as the most striking manifestation of quantum statistics, of the difference between bosons and fermions.

And the name Bose-Einstein condensation is still alive from those two well-known people who predicted it in the 1920s. However, Bose-Einstein condensation in a gas was not realized until '95 for reasons which will become obvious in a moment. So the prediction of Bose-Einstein's statistics was 1925. And for many, many years, the only manifestation of Bose-Einstein condensation was in liquid helium. The superfluity of liquid helium is a consequence of Bose-Einstein condensation, but it's a liquid. It's not a gas. And there are high density effects, which modify this phenomenon greatly.

But then there was an attempt to realize Bose-Einstein condensation dilute gases, and a lot of credit for that should go to Tom Greytak, and Dan Kleppner, who are here on the faculty. They really put Bose-Einstein condensation, and dilute atomic gases on the agenda, and made it a goal. And it's, I'm very pleased to mention that after 20 years of efforts they realized Bose-Einstein condensation in hydrogen, and achieved a longstanding goal.

However, a few years earlier there were new cooling techniques developed. Laser cooling, I would briefly comment on them. And new systems, alkali atoms, could be cooled to ultra low temperatures. And it's really the alkali atoms which have created this whole new field of Bose-Einstein condensation with many, many activities. And the success here came in '95. And I will tell you the story today.

So well, I think from most of the concepts I've presented, if you want to achieve Bose-Einstein condensation, quote, unquote. All you have to do is you have to cool down the gas to micro and nano Kelvin temperatures. You have to cool it down until the de Broglie wavelengths is comparable to the spacing between atoms, and the matter waves overlap, and start to oscillate in concert. So, I mention that the de Broglie wavelengths depends on temperature. The distance between atoms depends on density. So we haven't, so the transition to Bose-Einstein condensation is characterized by a combination of temperature and density.

So let's just look at this relation for something of the density of water. It would predict that the transition temperature is 1 Kelvin, which is easy to reach. However, tough luck at 1 Kelvin everything is liquid or frozen, and you can't observe the phenomenon Bose-Einstein condensation. So, in order to prevent that, and also to prevent formation of molecules out of the atomic gas, we had to work at extremely low density. So the system I'm referring to is at a density which is a billion times more dilute than water. It's 100,000 times thinner than air. And in those gases, we can observe Bose-Einstein condensation without self-destruction by molecule cluster formation, and solidification.

But there is a price to be paid. If you work at ultra low density, you need ultra long de Broglie wavelengths to make those matter waves oscillate in concept. So we needed the lowest temperatures ever achieved. So we needed cooling methods. And what I'm actually talking about is a combination of two different cooling methods, which we had developed in two different subareas.

One is laser cooling. the pioneers of laser cooling. Some of them were recognized with a Nobel Prize in '97. And major developments to laser cooling. Were also done by Dave Pritchard at MIT. The principle of laser cooling is fairly simple. You shine laser light on atoms, the atom skate on light. And if you play some tricks, which I don't want to you explain, that the light, which is scattered, which is emitted, has a shorter wavelength, is more energetic than the absorbed light, than the scattering of light removes energy from the system, and the system cools down.

It works great. You just shine laser light on atoms, and you create micro-Kelvin samples. And a very important configuration to do that is the one co-invented by Dave Pritchard. And I think this is probably one of the most important singular techniques which were ever introduced to the field of atomic physics, the magneto-optical trap. It is the workhorse of the whole field of ultra cold atoms now.

So well, and if you have this configuration of laser light, and shine light into a vacuum chamber, and add some atoms. You see, this is in one of our lab upstairs. You see a small ball, a few millimeters in diameter, yellow glow. So you can see with the naked eye, ultra cold sodium atoms. But at the fairly hot temperature of a fraction of a milli-Kelvin. So this is the starting point for the next stage of cooling towards Bose-Einstein condensation.

Laser cooling has so far not been able to carry atoms down to Bose-Einstein condensation, because it has limitations. And some of them are very fundamental. If you scatter light from atoms, each photon is the, photons are grainy. It's like sort of sand corns which transfer a finite of equal momentum to the atoms. So therefore, the temperature of the standard laser cooling methods is limited to micro-Kelvin. So well, if you can't get very cold you should get very dense, because it's a density-temperature combination which makes Bose condensation happens.

But unfortunately, there's also a density limitation in laser cooling, which, without going into details, is simply set by the process. If you have a very dense sample, all the laser light is absorbed, and laser cooling doesn't work. So laser cooling, at least in the beginning of the 90s, fell short by four orders of magnitude in density, or three orders of magnitude in temperature, to reach Bose-Einstein condensation.

So well, this is something which Dave Pritchard developed. And I was a post-doc with him. But next door is Dan Kleppner, who together with Tom Greytak, and Harold Hess, developed evaporative cooling. And this is a different cooling method, which is not suffering from all the limitations of laser cooling.

So let me explain what evaporative cooling is. Actually, before I explain the slide, evaporative cooling is what happens in everyday life. If you sit in a bathtub, the water cools down. And it cools down because the hottest particles escape as steam. And what remains behind are the less energetic particles. This is called evaporation.

So well, if you do it in an atom trap, if you can find atoms with magnetic forces, we create such a potential. It's like an invisible container, the walls of which are magnetic fields. I don't want to say more about it right now. But so if particles in this container, and if two such particles collide, one can lower its energy, come to rest. These are the particles we want to keep. They're really cold. And the more energetic one can escape.

And this was suggested by Harold Hess, and realized in the group of Dan Kleppner and Tom Greytak. So if we have a method to kind of eliminate the hottest particles in many, many steps, and the remaining particles we thermalize. We create a thermal distribution which gets colder, and colder, and colder.

So the question is how to implement it? And Dave Pritchard had a brilliant idea. We sort of X which, so well, it's even dripping with blood here. Okay, well what is a suitable X for atoms? So well, if you're in a magnetic trap, there's a very simple X. And this is just switch the magnetic moment of the atoms so that the attractive force, the confinement force turns into a repulsive one. You just have to shine a radio frequency onto the atoms, and you have your X, and you can do the cooling. So this was Dave Pritchard's suggestion in '89.

So well, here are two cooling methods developed by the people who had their office next to me, laser cooling and evaporative cooling. Laser cooling works great initially, has certain limitations. But evaporative cooling does not suffer from them. So it probably doesn't require a rocket scientist to realize it would be a good idea to combine the best of two worlds.

So this was sort of the idea in the early '90s. But there was a problem, which some people regarded as fundamental. And the problem is that, I mentioned that laser calling doesn't work when the sample becomes opaque, when the light gets absorbed. On the other hand, evaporative cooling requires collisions between particles. One particle gets colder, and one particle gets hotter.

And now, if you look at the collision cross-section, you face a dilemma. We don't want light to be absorbed, but we want collisions to happen. And the difference, the ratio of those collision cross-sections effect of 1,000. So it seemed absolutely impossible to have many collisions for evaporative cooling, and at the same time avoid the complete absorption of laser light.

A solution, which Dave Pritchard developed together with me was, so well, we can modify. We can push optical cooling to higher densities by just playing a simple trick. If we hide the cold atoms, in a different quantum state so that the light is not absorbed, we can push laser cooling to higher densities. I don't want to go into details, but by modifying the standard knot, which Dave co-invented, we realized something which is called the dark spot And it was a key technique to achieve Bose condensation both at MIT, and at Boulder.

And I just showed you the result, which at that point in '92, published in '93 was fairly impressive. We were shining light at a cloud of atoms, and the shadow was really black. We could infer that the transmitted light was 10 to the minus-80. This was absolutely black. And so in '92 what we had in our hands was an unprecedented combination of high density and large number of atoms.

And then, you told stories about me, Dave. Now I have to tell a story about you. Namely I still, I mean I remember those dramatic days. This was the summer of '92, and Dave Pritchard, he had funded work, funded proposals, and he had great ideas to study cold collisions, to study cold molecule formation, and eventually many people got a lot of credit for that. But we had, so this was the agenda, and this was the proposed work, and this is what the funding agencies had supported.

But then we sat together, and decided maybe we can do something else with those high density sample. Maybe we can go for the very speculative, and challenging goal of Bose-Einstein condensation. And it was for me just this signature of a great scientist that Dave said, okay, I give up all my ideas, which were really great. And which other people got enormous credit for that. Let's do what you suggest. Let's go for the long haul. Let's try to go further, and use this as a starting point for the quest for Bose-Einstein condensation.

And it was very dramatic. Within a few weeks, even before publishing the paper, we placed all the orders for additional equipment necessary to do the next step. To combine laser cooling with evaporative cooling. So eventually we wanted to embark on the agenda to use laser cooling as the first step. And then lower the temperature by evaporative cooling, and get to Bose-Einstein condensation.

So well, we had to build a complete machinery for that. And it took one or two years to assemble it. And then we were sort of ready. And tried to see evaporative cooling. And this, actually May of '94 was an important step. I went to a meeting, and announced for the first time that laser cooling, and evaporative cooling had been combined. By using this high density sample of atoms, we could bridge the gap between this factor of thousands difference in cross section of light absorption, elastic collision.

We had seen evaporative cooling, and now we were hoping to push further to lower temperature. But still, we were five orders of magnitude away from BEC. At the same meeting, the Boulder group announced that they accomplish something similar. So the race went on, and it was one of the most exciting races of my lifetime. And Eric Cornell, Carl Wieman, and myself, we remained friends. And we talked to each other on the phone on Tuesday. And people would say to each other, you got the best out of me because if you have competition you work harder, you think harder, and you do better work.

Actually, at that point I thought it could take the rest of my life to push further incrementally, and explore the new physics, which would come about at higher densities, and lower temperatures. And I knew that the quest for Bose-Einstein condensation in Tom Greytaks, and Dan Kleppner's lab had already lasted 15 years. And I was expecting, so well we trying on something alternative. And let's see which method works better.

And to the big, big surprise, and it came completely unexpected, just one year later Bose-Einstein condensation was achieved. It's one of the rare examples, I think, where you think you do an incremental step, and you step forward by a factor of a million. So these were very dramatic moments.

Let me mention a little bit about the last obstacles on the finish line to Bose-Einstein condensation. So the combination of evaporative cooling, and laser cooling was accomplished. But in order to have enough collisions between the atoms to have good evaporation, we had to keep the gas tightly confined. So we needed tightly confining containers, or magnetic traps.

Now this may be a little bit technical, but some people in the audience who probably are interested in that. There are two types of magnetic traps. Some have a pointy potentially, a linear re-shape potential. And others have a round potential at the bottom. Those traps are more tightly confining, and that's what both the group in Boulder and at MIT used to demonstrate initial evaporative cooling because it makes it easier to get many collisions, and see evaporation.

But they have a problem. There is a point of zero magnetic field in the center. And if the field is zero, the atoms don't know where to spin up, and spin down. They lose their orientation, and virtually flip out of the trap. So we lost the atoms. It's like a hole where atoms leak out. And the months before BEC was achieved, both groups had to invent methods. And this, how it was dubbed, how to plug the hole.

Eric Cornell's group in Boulder used an ingenious method with rotating magnetic fields. And we decided to use a different method. We wanted to plug the hole with a laser beam, which was keeping the atoms away from this dangerous region where they would just fall out of the magnetic trap. So what we created was a potential, which has a hole in the middle. And you see here the shape of our clouds. And it worked great. It was really dramatic. We could push forward orders of magnitude towards Bose-Einstein condensation.

And then, this was in June of '95, we heard about the Boulder results. They had succeeded to get Bose-Einstein condensation. And that's maybe something, I don't know if I've shared that publicly. If you're an assistant professor, you put all your money, all your resources, all your efforts in one goal, and then you're notified that somebody else has accomplished it. This is a difficult situation.

[AUDIENCE LAUGHS]

You wonder about promotion. You wonder about your scientific career, and of course I know at MIT the bar is very high. MIT is not keeping losers. So I was really scratching my head. I woke up early in the morning. What to do? And people ask me, why don't you just copy the design in Boulder? And I mean this seems to be a successful way to do it. And then a whole new field opens up. And we had fierce discussion in my group. And what we did is actually a two-fold strategy.

We designed a new magnetic trap. I will talk about it. And this turned into the workhorse of the field. And this has given us an enormous productivity over the last few years. But, and I was sort of willing to say let's just put all our eggs into this basket. Let's pursue that, because this new magnetic trap is a really good idea. And indeed, it turned out to be a good idea.

But I work with a good group of people, and they told me, so well, this trap with the plug, maybe it will work. Or even if it doesn't work, because maybe the laser beam is jittering, and heating up the atoms. Let's just give it one, or two more tries. And that's really what, let's just figure out how far we've come, because the Boulder group has a big success. And we should at least document how close we came.

So well, I try to listen to my collaborators, and said, okay, but just a few times. We can't waste time. We don't want to miss the boat. And it was one of the very last attempts to see how far we could push with our original idea of this optically plugged trap that we saw Bose-Einstein condensation. This is a picture, which doesn't look nice, but I want to show it to sort of the experts. This was our first signal of Bose-Einstein condensation on September 30th in '95.

We were not really ready to observe. We just wanted to see where are the limits of the plugged trap. And for the experts, we couldn't switch off the laser beam. So in time of flight, the atoms were just pushed away by the laser beam. So it's not a pretty time of flight picture. It doesn't look as pretty as the other ones. But we saw something black. Sharp, black spots on the screen. And if a cloud expands ballistically. And stays compact, that means there's very little velocity. There is very little motion. It's very, very cold.

And we got enormously excited. And within a few hours, we worked throughout the whole night. We improved the set up, and we got clear evidence that we had observed Bose-Einstein condensation. We were running the experiment only one more time to get final data. And this is actually, for me, still amazing. This is the quality of the data. Excuse me. Which was obtained the second time we were ever running the machine for Bose-Einstein condensation.

Let me just explain what it means. We let the gas go, and it expands out. And then we take a snapshot, and the larger the cloud is, the faster is the velocity, the hotter are the atoms. And now you see what happens when we cool down. You see how this cloud is sort of shrinking because the atom slowed down. And this is sort of the last factor of two. The last fraction of a second to achieve Bose-Einstein condensation, and then this new form of matter. This ultra cold gas, which is a Bose-Einstein condensate forms. It's very visual. It's very dramatic. And we obtained that without really optimizing our machine.

Okay, so I mentioned that this was a night on September 30th. And so well, at MIT you have to teach. And October 1st I had to teach 8012.

[AUDIENCE LAUGHS]

I received an email message from one of my students now, six years later. As a freshman in the fall of '95, I took your 8012 class. I remember one day when you were unusually late for your lecture, and you came rushing in very excited, yet very tired from the previous night's experiment. I think I told the students I didn't get any sleep at all. And please, if I fall asleep during my class, ask me a question.

[AUDIENCE LAUGHS]

So well, a few weeks later there was some short report in Tick Tock, but this was the night that Bose-Einstein condensation was realized at MIT. Okay, but immediately with our approach, we had 100 times more atoms. And we had a system, which was more versatile to pursue the new physics of Bose-Einstein condensation. This is a picture of the group a few months earlier. It doesn't show Ken Davis, but it shows the people, the other people who were on board. Dylan Durfee, Mark Mavis, Michael Andrews, Ben Stamper-Kurn, Klaas-Jan van Druten as a post-doc, and Chris Townsend joined a few months after we saw BEC.

Anyway, we were playing, and I'll tell you a story of sometimes you do the right thing for the wrong reason. And this is when we changed our magnetic trap, which is really the core of the apparatus. We were one of the two or three groups that was a little bit ubiquitous observation by a third group, who had seen Bose-Einstein condensation. And everybody was brimming with excitement, and I made a wrong assessment. I said, hey it wasn't really so complicated, what we do. In the next few months, many groups around the world will repeat it. And now we have to be really careful. Maybe with our jittery laser beam, we're not competitive. We really have to go forward, and make an improvement.

And we thought, let's build a new magnetic trap. What happened, actually, is it took us five months to build it. We couldn't get any Bose condensation for the next five months. But when we got it, we were really going like gangbusters. But my assessment was wrong, because nobody else got Bose condensation in the meantime. Indeed, to build such a complex apparatus it took two years for the other groups. So if we had known that, we would have probably just stuck to our original set up.

But we went forward, and created what turned out to be a good trap, and which gave us enormous productivity. So we realized this trapping configuration, which is not suffering from this leak, it's sort of a round bottom trap. It actually carries the name Ioffe-Pritchard because Dave Pritchard suggested that in the mid '80s. And we found a new way to wind the trap. It's now referred to as a cloverleaf trap. And so well, this is how we have produced Bose condensates ever since.

This shows you the actual winding path. And now I show you a few examples of what we were able to do in this trap. It was an enormously productive time, and one of my graduate students, who was with me between '95 and '99, he graduated with 20 papers in [INAUDIBLE], Nature, or Science.

So these were one of the first experiments we did in this novel trap. We could now visualize the condensate inside the trap using a novel light scattering technique, and now we could really observe what was going on inside the condensate. And here you see the cool down. Our magnetic container was elongated, so it's an elongated cloud, and it shrinks as we cool. And when we reach the phase transition in the middle of that trap we see this formation of a dense core that Bose-Einstein condensate. And if you just take a cut through this cloud you can see, you know, this. You remember the blue guys who march in lockstep? This is this peak, and then you have a thermal component. These were the colorful guys which were moving and whizzing around in all directions.

And we did a series of experiments. Here you see a condensate set in oscillation. People call it the study of sound, and some of those results were just the frequency of oscillation. What is its damping was really causing major challenges to many body theory, and led to hundreds of theoretical papers.

On something which I felt was very dramatic, those nano-Kelvin atoms, you think they are so fragile. It's colder than anything else, but by using this light scattering technique and just stroboscopic illumination we could really see how, color-coded in red, how condensate formed. It was for me like watching how nature is giving birth to something, which is very fragile, but nevertheless we were able to observe it in its natural environment. And actually, the way how the condensate forms was challenging theory, and over many, many years now, quantum kinetics theory has been developed, to some extent, to explain the dynamics of the formation of the condensate.

But let me now talk about two examples of our research, which really demonstrates what makes Bose condensate so special. Now, if I talk to a lay audience, so if I want to help you. If you meet your friends, and you want to tell them, what is really special about Bose-Einstein condensates? I want to now give you two examples how you can convince people that this is something that has unusual properties. And it's really related to what I call it here, the magic of matter waves.

Now I want to show you two examples. The interference of condensates, which allow you to directly photograph matter waves. And I want to show you the example of rotating condensates, which have a Swiss cheese-like appearance. This is very different from how any ordinary matter behaves. Okay. What I want to explain to you now, the interference between condensates was actually a key experiment which we did. And it was also actually one of our key contributions, which was emphasized by the Swedish Academy. Because what we could do in '96, and publish in '97 was we could show for the first time that the condensate is not really cold. So well, coldness is relative, but that it has these special properties that all the atoms march in lockstep. That they're just one quantum mechanical matter wave.

Okay, if you want to show that something, if you want to show, for instance, that this laser beam is an electromagnetic wave, the best way to do it is you take two laser beams, and interfere them. And then you see an interference pattern. The interference is the clear evidence for the wave nature. So we needed to condensates, and so well by using a laser knife, and cutting a condensate into two pieces, foregoing all the details. We created two condensates. And so now we had two condensates, and we switched off our magnetic container, and the condensates, so well, if you switch it off, the condensates fall down due to gravity. But they also expand because of zero point motion, and atomic repulsion.

So what we are hoping for is as the condensates expand into each other, and overlap to see the wave nature of those atoms. Certainly showed you what we expected to see. So let me just give you an example for the interference between two sources. So assume that this is the antenna of our favorite radio station in Boston, and it radiates radio waves. And now we bring in a second radio station, and now the waves of the two radio stations interfere.

And now you see the interference pattern. And of course, the interference pattern becomes more microscopic the closer the distance is. And at least those who took 803, they should realize that this interference pattern has the form of parabolas, because a parabola is a mathematical curve for which the path length difference to the two origins is constant. So well, we're not interfering radio waves, we're interfering matter waves.

So we have our container, the atoms spread out, and I have to remind you probably of pretty much the simple equation I showed you, the de Broglie relation. The faster the atoms are, the shorter is the de Broglie wavelengths. So if you have a puff of gas, and let it go, and then take a snapshot, the atoms which are further out they have moved there faster. And the atoms which are closer in, they are slower.

So the wave pattern of such a pulsed matter wave looks like this. The matter wavelengths is longer where the atoms haven't traveled so far, and it is shorter where they have travelled further. So now we interfere our two Bose-Einstein condensates. And you again see now, as we bring them closer together, that there is a distinct interference pattern.

But in contrast to the two radio station, this interference pattern consists of perfectly straight lines. And when we observe the overlap of the two condensates, and we saw that, we were really jumping up from our chairs. And I think I remember the first night we saw that. Again, we had worked into the wee hours in the morning. I think it was you, Dan, you came in at 8:00 or 9:00, at 8 or 9 o'clock and I was almost drunk, and I came to Dan's office. Dan, I have to show you something. And Dan came to the office, and Dan came to the lab and we had this pattern on the screen. So we knew the Bose condensate is a coherent form of matter.

Actually, just to talk in terms of layperson, what you actually see here is the-- well, two things. One is, this is a direct observation of the wave nature of matter. We simply take two puffs of atoms, they overlap, and then we illuminate the atoms with a laser beam. And we take a shadow image. There is, it's the most direct observation of the nature of matter.

But secondly, this is also demonstrating in a fairly dramatic way, this equation that atoms plus atoms, gives vacuum. Because where there is no shadow, there are no atoms. So the two matter waves interfere each other away. Atoms plus atoms gives vacuum. I mean, I sometimes get fan mail from people with their own theories of physics, and.

[AUDIENCE LAUGHS]

And I got some questions about, yeah, what happens if you annihilate atoms by interference, but I think you all know. The atoms which are missing here, they appear in the dark fringes because atom plus atom equals four atoms. If you have coherent interference.

Okay, so this is just a different, more fancy representation of that. At about the same time, we found a trick to release atoms out of the magnetic trap. We were just sort of spin-flipping atoms. They were accelerated down by gravity, and so within a few months we had done two things. We had shown that the atoms are laser-like, and secondly we were able to create pulse beams of atoms.

And this together is now recognized as the first realization of an atom laser. An atom laser, which emits atomic matter with laser light properties. And other groups have followed, and improved on that. So this is sort of an atom laser gallery. Continues atom lasers and in different ways its a rich field, and there's a lot to be done.

Let me now talk about this other example, and where the wave nature of matter really becomes manifest. And this is actually related to a very special property of the Bose condensate. The Bose condensate is sort of one giant matter wave, and its possible for wave-like matter to move, to propagate without any friction. This one similar system which has enormous technical importance, and these are super conductors. In superconductors electricity can flow without any dissipation. You can switch on a ring current, and the ring-- you can go home, and months later the ring current still flows without any active power.

Those phenomena are usually dubbed, "super." Super conductivity for electrons, super fluidity for liquid helium. And here we have a system which was predicted to be super fluid, but its a gas. Its eight orders of magnitude more dilute the liquid helium. So we have the most fundamental system in our hands to understand details of super fluidity or macroscopic random phenomena.

I don't want to go into details, and rather emphasize one aspect, which is fairly directly connected to superfluity. But I'm not really explaining it, and this is the physics of vortices. Now, I think you are familiar with vortices. This as a vortex on a very large scale. You see how they can, off the coast of Florida. I think we also encounter vortices in our everyday life. At least, I could used one today.

[AUDIENCE LAUGHS]

You simply push the button, and you create a vortex. So what is special about vortex physics? Okay, there's something important difference. And that's what I want to explain. Let's consider this rotating bucket experiment. And I want to show you that there's a dramatic difference between rotating, and ordinary fluid. Let's say a bucket full of water, and rotating what I call a quantum fluid. And this is the system where all the particles are just one big wave.

First of all, if you rotate a normal fluid, you know that upon the rotation it forms this parabolic surface because of centrifugal forces. If you take, for instance, superfluid helium, and you do the same. I just told you there's a big difference, but the same happens. It's a parabolic surface. And it has to be like this, because on large scales there is a correspondence equivalence between a classical system, and the quantum system. But if you now look a little bit closer onto the quantum system, you will find that the system is littered with tiny little holes. Tiny little vortices, so there's a big difference, and I want to explain that to you.