Puttering Around in the Basement
on the Road to Stockholm

Early Experiences

This experimentation went on side by side with my education in science and math throughout grade school and junior high school. While I excelled in my science and math courses, my two major preoccupations were very separate activities. I was, however, deeply influenced by some excellent and carrying teachers. In seventh and eighth grade I first had science as a classroom subject. My teacher, Mr. Miller, had a very unfair reputation as a strict disciplinarian. Rumor was that he beat students who disrupted his class. For the first week of class I was absolutely petrified of Mr. Miller, and failed my first weekly exam cold. Then something magic happened. Mr. Miller came to me and very gently asked why I had done so badly on the exam. He said he was certain I was much smarter than that, and asked if there was anything he could do. Clearly he was not a beast, but someone who was deeply concerned about my education! For the next two years, no student in any of Mr. Miller's five sections of science class ever beat me on a single one of our weekly exams.

Even in high school I regarded my experimentation as at least as important as my studies. For instance, I always took a study hall so that I could get my homework out of the way, thus allowing me to spend more time experimenting in the evenings. It was not until my junior year in high school that these two activities began to overlap, when I was fortunate enough to have a chemistry teacher, Mr. William Hock, who took the time to explain to his students what physical research was all about. At least in those days such an endeavor was rare, and indeed, to my knowledge, he was the only one of my teachers who actually had research experience himself. I cannot say that I was ever as fascinated by chemistry as I am by physics, but Mr. Hock brought to my madness a method: the scientific method. I remember quite well one class assignment: to record our own observations of a burning candle. I knew pretty well how a candle worked, and simply wrote down an explanation of how radiant heat from the flame melted the wax, which was then drawn up into the wick by capillary action, etc. Mr. Hock read my explanation, and then came to me and pointed out that what I had written could not possibly have been drawn from my own observations. He asked me to do the assignment over again. Perhaps more importantly, however, Mr. Hock condoned a strategy I developed to insure that my chemistry labs went well: I would calculate out ahead of time what I should expect, so that if things went wrong, I would know it, hopefully before it was too late. I know that many teachers regard this as a form of cheating, but I can tell you this: without a firm knowledge of what to expect in the experiments which led to my Nobel Prize winning discovery, I very much doubt I would be speaking to you here today.

I have from time to time kept in touch with Mr. Hock over the years, and am pleased that he is able to be here today. William, please stand up. And would everyone join me in thanking an excellent teacher who has indeed made a difference!

At this point, William Hock, sitting at the front of the audience, rose in acknowledgement of Prof. Osheroff's recognition.

Background to Superfluid ³He

Let me first say a little about the nature of the new phases of matter which I and my two professors at Cornell University, David Lee and Robert Richardson discovered. They are neutral analogs to conventional form of superconductivity, a state of matter in which conduction electrons in metals form a macroscopic quantum wave function. This quantum wave function is similar in many respects to the wave functions which describe electrons in the charge cloud around the nuclei of atoms, but exists over a length scale a billion times as large. One of the many unusual properties of macroscopic quantum condensates is the ability of the condensed fluid to flow without the dissipation of energy. In the case of superconductors, where the condensed particles are charged, this leads to a state of zero electrical resistance. In the case of superfluidity in ³He, where it is neutral atoms which form the condensate, what is seen instead is a state with zero viscosity. I hasten to add, however, that all macroscopic quantum states exhibit other, more subtle, unusual behavior, such as the interference of matter waves, and the quantization of angular momentum and magnetic flux on a macroscopic scale.

Superconductivity was discovered by Kamerlingh Onnes in the Netherlands in about 1910, and yet it was not until 1957, with the development of the BCS theory of superconductivity by John Bardeen, Leon Cooper, and Robert Schrieffer that we achieved a microscopic understanding of how superconductivity comes about. The problem for a long time had been that electrons are Fermi particles, and must obey the Pauli exclusion principle: No two Fermi particles can occupy the same quantum state. So how could electrons condense into a single macroscopic quantum state? The answer to this question, provided by the BCS theory, was that the electrons formed weakly bound pairs called Cooper pairs. A bound pair of Fermi particles is a Bose particle, and Bose particles love to be in the same quantum state. That's why a laser works!

Shortly after the development of the BCS theory, physicists began to speculate that a similar condensation might occur for other fluids of 'cold' Fermi particles, however just two substances were expected to remain in a fluid state to sufficiently low temperatures: liquid ³He, and the neutron matter in neutron stars. While we still believe that the matter in neutron stars exists in a BCS state, it is impossible to tell for sure. In the case of ³He, however, physicists tried repeatedly to cool liquid ³He sufficiently low in temperature that it would undergo a BCS phase transition throughout much of the early to mid-1960's with no success. Despite the success of the BCS theory, it was not clear what mechanism would provide the attractive interactions between ³He atoms to bind them together in Cooper pairs, and even with knowledge of the relevant interaction, it is extremely difficult to predict the phase transition temperature using the BCS theory. As low temperature physicists succeeded in cooling liquid ³He to successively lower and lower temperatures but saw no superfluidity, theorists kept changing their predictions, suggesting lower and lower temperatures for the transition to the ordered state. Eventually, the experimentalists had gotten down to about one thousandth of a degree Kelvin above absolute zero, but the theorists then predicted a transition temperature of fifty millionths of a degree! At that point, everyone gave up the search.

Choosing College

If I went to Stanford, I felt I would go into something which would allow me to become a journalist, as I was fascinated by the power and responsibility of the press. If I went to Caltech, I would undoubtedly go into physics. Finally, I realized that if I went to Stanford, I would have to endure having my academic record compared to my very bright older brother's for four more years, and thus chose Caltech. I must confess, not a very sound decision.

Choosing a Major

Eventually I decided I should major in electrical engineering. The next term I took a course on transistors, since I had always wanted to know how they worked. For those who are not aware of the fact, John Bardeen, who led the effort which explained the nature of superconductivity, also participated in the development of the transistor. He is the only person to have shared two separate Nobel Prizes in physics. Now this class on the transistor met at 8:00am, a very early hour. I had effectively excluded myself from a career in geology by sleeping through a geology class which met the previous term at 8:00am. The professor would promptly turn off the lights after entering the room, and spend almost the entire hour showing slides of rock formations. I hadn't realized how much of the course I had slept through until the final exam!

The course on transistors did not tell me how transistors worked, but only how to use them in electronic circuits. This was to me a big disappointment, and I soon decided that I would prefer being a mediocre physicist than a frustrated engineer.

Undergraduate Research Experiences

While I had always liked astrophysics, I found that as a topic of research it was for me too constraining. One doesn't do experiments in astrophysics, but rather makes observations of distant objects whose behavior one cannot hope to influence. Neugebauer was an early pioneer in infra-red astrophysics, and mine were actually the third set of eyes to see the center of our galaxy (on a strip chart recorder), but I eventually decided I wanted something else.

I should add that I my experience in astrophysics had not only allowed me to keep my sanity, but it had well prepared me for future research studies. I learned how to design and build instrumentation, how to manage large amounts of data, and how to manage my own limited physical stamina. In astrophysics, one would go to an observatory to take data, typically only at night. This business of switching from a day time schedule to a night time schedule and back was to play an important part in my later work in low temperature physics, where experiments had to continue until they were done, often 16 to 32 hours straight.

In my senior year, in order to get out of a third term of senior physics lab, I signed up to work in a low temperature group which was trying to reach a temperature of 0.5 Kelvin by pumping on a bath of liquid ⁴He. At about 2.2 Kelvin, ⁴He condenses into a macroscopic quantum state which is also characterized by superfluidity, but ⁴He atoms are Bose particles, not Fermi particles, and this very interesting state of matter had been explored thoroughly beginning in the 1930's. One of the unusual properties of superfluids is that they consist of interpenetrating mixtures of normal fluids and superfluids. If one part of the fluid is warmer than the rest, the superfluid component, which possesses no entropy, or disorder, will be driven toward the source of heat by what is called the fountain pressure. This unique behavior also occurs even in thin helium films, and had for a long time limited man's ability to use liquid ⁴He as a refrigerant below about 0.7 Kelvin.

In the end, the group did not reach 0.5 Kelvin by pumping on superfluid ⁴He, but by that time my mind was filled with all the bizarre behavior which can be studied at low temperatures. So much so, that when I went to Cornell for graduate school, I went with the intention of studying solid state physics, rather than astrophysics.

Cornell

Bob Richardson had just been made an assistant professor at Cornell, and that fall he gave a physics seminar on a new refrigeration device called a ³He-⁴He dilution refrigerator, capable of reaching and maintaining temperatures as low as 15 mK. This seemed almost beyond belief for someone who had worked so hard to reach 0.5 Kelvin. Then I learned about another potential cooling device which might reach temperatures as low as 2 mK. My mind began to reel with excitement as I imagined all the new and exciting physics which might exist in the temperature range which these new devices would make available for study. By the beginning of my second semester, I joined the low temperature group.

Career Choices

First, when considering a career, look back at what has motivated you in the past. Often, but not always, you will find you have done things because you enjoyed them. Also, however, careers are complex entities which involve a myriad of different skills and activities. Do not be attracted to a career because of a single high ideal, but because it involves things which you enjoy doing.

Second, choose a career which will satisfy your needs. In particular, choose something which you will find intellectually stimulating and emotionally satisfying. If you are not stimulated by your work, you soon find it a bore, and you a drudge. By emotionally satisfying, I mean something that makes you feel good about what you have done. Usually, this means something that you can do very well, at least compared to your peers, but it also involves the value. What is the value of what you are doing, both to yourself and to mankind? Some people need a career in which the value is instantly recognizable, such as in medicine, while some are willing to realize that the benefits to their work may be slow in coming, and that the value exists in the probability that the work will provide benefit to mankind, rather than the demonstration of that benefit.

Finally, when choosing a subfield of study, assess your strengths and weaknesses. Choose a sub-field which allows you to best utilize your strengths and hopefully does not depend crucially on areas in which you are weak. In today's world, one cannot avoid competing with others both for job opportunities, and for the resources necessary to carry out those jobs. While you may believe that theoretical physics is at the real heart of the field, if you are weak at math (compared to the brightest people around) and find abstract theories difficult to understand, you will be unlikely to contribute to the development of the field were you to enter it, and would probably not be given many opportunities to fail.

I often tell my students at Stanford that the most important things they will learn as students is not something that a professor will tell them, but things they learn about themselves. What motivates them, how hard they can work, and what they find satisfying.

Getting Started in Research

Soon after joining the low temperature group at Cornell, I built with a senior graduate student, Jim Sites, a ³He-⁴He dilution refrigerator to be used in Jim's studies of nuclear spin ordering in solid ³He. Both ³He and ⁴He will not freeze under their own vapor pressure even as they are cooled arbitrarily close to absolute zero, but if pressurized to slightly over 34 atmospheres for ³He and 25 atmospheres for ⁴He, they will form solids near T = 0. It was expected that the nuclear spins in solid ³He would order into an unknown arrangement below about 2 mK due to the actual exchange of atoms between lattice sites in the solid at rates as high as 40 million times each second.

During my second year Jim built an additional cooling device called a Pomeranchuk, or compressional, refrigerator which relies on the fact that ³He has a negative latent heat of solidification below 0.3 Kelvin. This device was expected to cool slightly below the temperature at which the solid spin system ordered. Then Jim and I did a series of NMR experiments on the solid above the ordering transition to better estimate the rate of atom-atom exchange. Jim graduated that summer, and I inherited all his apparatus.

I spent my third year building and testing a new Pomeranchuk device, based on a design I had conceived while lying in a hospital bed after knee surgery following a skiing accident. This new device looked very promising, and I spent my entire fourth year of graduate study working with another student, Linton Corruccini, using this device to cool liquid ³He-⁴He mixtures to about 6 mK for some NMR studies of normal state properties of liquid ³He (which certainly aren't what you would regard as a normal liquid!).

David Lee had told me that for my Ph.D. thesis, he wanted me to measure some property of solid ³He through the temperature range at which the solid spin system ordered. The problem was, we didn't know in those days how to make a thermometer which would accurately measure the temperature of the ³He in our cell. Below about 3 mK every thermometer we had tried lost thermal contact with the sample. But, toward the end of my fourth summer, Dave showed me a pre-print of an article from another group, our chief competitor in these studies of the solid ³He spin system. They had observed a very strange thing: When they applied large magnetic fields across a sample of liquid and solid ³He, the melting pressure at constant temperature dropped by an amount which was over a factor of ten larger than could be accounted for by existing theory. These results were seen at 5 mK, where my thermometer, based on NMR on copper wires, worked very well.

I decided to reproduce these strange results, using my NMR thermometry. However, it soon became clear that the previous strange results were an artifact of the other group's thermometry scheme. The actual effect was entirely consistent with existing theory, but so small, it was hard for me to measure.

After trying to measure the effect for over two months, another group of graduate students argued that I should give up the lab's only NMR magnet, which weighed several tons and whose power supply sounded like a B-52 taking off. Reluctantly I complied. However, back in those days, a low temperature apparatus was not very reliable, and so I kept my apparatus cold, hoping that the other students would have problems and have to warm their probe up for repairs.

The Discovery of Superfluid ³He

A Pomeranchuk cell should cool about 1 mK for every percent of the liquid which was converted to solid. I decided to convert liquid to solid at a very steady rate, and measure the melting pressure in the cell as a function of time. We had measured the melting pressure down to 2.7 mK, and I felt I could extrapolate below that.

The first time I tried this experiment, I found everything went fine down to about 2.6 mK, but at about that temperature the rate of cooling suddenly slowed abruptly. Four days later I tried the experiment a second time, starting the compressional cooling at a lower temperature, and found a similar decrease. I had worried that if I was unlucky, the moving piston which solidified the liquid might deform solid ³He previously formed, and this would lead to heating. However, in this second experiment, I found the decrease in cooling rate occurred at almost exactly the same temperature as in the previous run, even though the starting conditions were very different. The melting pressure at this change reproduced itself to one part in 50,000.

To me, it seemed like too big a coincidence that the cell would develop heating at exactly the same pressure twice in a row, particularly with very different amounts of solid in the cell. It seemed much more likely that this behavior resulted from some very highly reproducible phase transition in our cell at about 2.6 mK. But none was expected.

The rest is a story for another day. If anyone is interested in reading about the next seven months, during which time we began to realize that what we had discovered was indeed three BCS transitions in liquid ³He, the Reviews of Modern Physics will publish the Physics Nobel Lectures for 1996 in what I believe is their July issue. I encourage you to read about the discovery, and the subsequent three years during which we began to realize just how unusual these BCS states really were. Before writing my Nobel Lecture, I spent an entire weekend reading 400 pages of old lab notebooks, and even I was surprised at how difficult the task had been.

Concluding Thoughts

What if, after two months of very difficult and somewhat disheartening work, I had looked at the loss of my NMR magnet as an opportunity to take a much needed and well deserved rest rather than the opportunity to do something fun, different, and hopefully revealing?

The second is that one cannot set out to make a major discovery, because, by the nature of un-discovered things, they are difficult to anticipate. Indeed, the more strongly they change the way we think of things, the less likely we are to anticipate them. In the four discoveries I have made in my research career, two have resulted when I have deviated from my focused research program to try something crazy and new. Perhaps we all need to do this more often. The trick is to not waste too much time at it if nothing turns up!

At the conclusion of the Address, Prof. Osheroff answered questions from the audience publicly for about 20 minutes. Afterwards, he spoke with many participants at the foot of the Award Ceremony stage.