A specialist in atomic physics and spectroscopy, the German-born Ketterle, 41, studied physics at the University of Heidelberg and the Technical University of Munich. His doctoral studies, undertaken at the Ludwig Maximilians University of Munich and the Max Planck Institute for Quantum Optics, Garching, were completed in 1986. Among his early accomplishments at Garching, Ketterle lists the first observation of discrete spectra of HeH, a complete analysis of the electronic structure of HeH, measurements of the lifetime of triatomic hydrogen, and spectroscopy of all the stable isotopes of H₃. Following this fundamental work, he returned for a short time to the University of Heidelberg to work on immediately applicable physics: laser diagnostics of combustion in the diesel engine. But his ambition to work on fundamental topics led to Ketterle joining the Physics Department of MIT as a research associate in 1990, following which he was appointed assistant professor in 1993.

   MIT recognized Ketterle's experimental genius in July 1997 with promotion to full professor. In 1997 he shared with Eric Cornell (University of Colorado, Boulder) the Rabi Prize of the American Physical Society. The German Physical Society awarded him the 1997 Gustav-Hertz Prize. Both societies cited Ketterle's achievement in the realization and subsequent study of BEC.

Science Watch Physics correspondent Simon Mitton spoke with Ketterle at his MIT office.

You’ve contributed to MIT becoming one of the leading centers in the world for the study of atom trapping, laser cooling, and Bose-Einstein condensate. What interested you in atomic physics as a research field?

   Ketterle: My career started out in condensed matter physics and molecular spectroscopy. I turned to atomic physics only later. My switch from molecular spectroscopy to atomic physics was influenced by the fact that I knew spectroscopy very well and looked for a new field in which to apply it. At that point I decided that atomic spectroscopy, laser cooling, and trapping were very promising fields, and I wanted to join in. I came here eight years ago as a postdoc with Dave Pritchard, where I learned a lot of atomic physics.

At that time, what problems in atom trapping and laser cooling needed solutions?

   Ketterle: Laser cooling was in pretty full swing, and people had demonstrated wonderful cooling and trapping techniques. However, it was clear that some dreams had not been fulfilled. Cooling and trapping were both hitting limitations. When laser cooling was invented and pioneered in the late 80s, people were already dreaming of Bose-Einstein condensation, where new collective phenomena would show up. BEC requires extremely low temperatures and high densities, and laser cooling was, at that point, at best five orders of magnitude in phase space density away. Early on in my work with Dave Pritchard we started working first theoretically and later experimentally on how could we overcome the limitations of laser cooling.

Why didn't laser cooling get closer to the desirable conditions?

   Ketterle: The limitation of laser cooling is reaching a high enough density for interesting physics. Laser cooling is great for low density, when the laser light can reach the atoms. But once you have a very high density, the laser light just gets absorbed. So although the technique was very powerful for atomic gases at low density, we couldn’t explore the denser parts of phase space, which you must do to create Bose-Einstein condensate.

    However, in another field of physics, the community studying spin polarization of hydrogen—which is more traditional low-temperature physics—had developed another cooling technique which does not require laser light. Evaporative cooling was pioneered here at MIT, and it simply requires elastic collisions to thermalize the gas you want to cool. This, of course, is the same phenomenon that causes bathwater or the coffee in your cup to get cold. The hot water molecules escape as steam and the remaining water gets colder and colder. It works great for atoms too, but this cooling scheme requires high densities because you need frequent collisions between the particles in order to reach thermal equilibrium. Many people perceived a gap between the densities that can be achieved by laser cooling and the (higher) densities needed to get evaporative cooling started. In 1994 we finally closed the gap between the techniques. Once we had achieved the combination of laser and evaporative cooling, it was really amazing. Within little more than a year, BEC was realized. The moment we had bridged the gap, everything just took off, at a rate of progress that took everyone’s breathe away, including myself!

   In the set-up at MIT we use sodium atoms and lasers operating in the visible-yellow light. Our multistage process starts with a hot atomic beam at 600 K. We cool the atomic beam by Zeeman slowing, in which the atom velocities are reduced by a counter-propagating laser beam. Then we use the standard magneto-optic trap; we employ the dark version of a technique that Dave Pritchard and I introduced. This technique involves higher densities than other traps. After additional laser cooling, we transfer the atoms to a magnetic traps. We now use a novel cloverleaf trap as a magnetic trap, something we introduced in 1996. Finally, we use evaporative cooling to achieve the density and low temperature needed for BEC.

Once you thought you had reached the critical conditions for BEC, how did you prove that the atoms were behaving coherently, which is the crucial test of the physics?

   Ketterle: In the first experiment on BEC we mainly showed that the gas condensed into an extremely cold form of matter—or, to be more precise, the gas had extremely small energy content. When we released the cloud all of a sudden we could see a dense core which was almost not spreading out at all. This was the condensate!

   In addition to being ultracold, the BEC gas has the property that its atoms are coherent, a feature I describe as "atoms marching in lockstep." This coherence is a different property to simply being ultracold, although the two are related. The next step was to show the coherence property directly to show that we had in effect made a matter wave. The trick to prove that was an interference experiment. We followed traditional examples from optics where you demonstrate the coherence of light by recording an interference pattern.

   If you shine two coherent light beams on a screen, you get the bright and dark lines of an interference pattern. So eventually we did the same experiment with Bose condensates. We created two condensates in a special atom trap and then made them overlap, and what we saw was a regular pattern of dark and bright fringes. That was immediate and direct confirmation that we had created coherent atoms. continued

Search | Jan/Feb 1999 Index | Archives | Contact | Home