摘要:
The speaker will describe the seven month period when, as a graduate student, he participated in the discovery of three superfluid phases of liquid He3. These are macroscopic quantum states of matter first predicted to exist by the BCS theory which explains the origins of superconductivity in metals. We now understand these phases as p-wave BCS states, and hence complex generalizations of the order existing in most superconductors.
講題內容：
As a graduate student in the mid-1960's, I was attracted to low temperature physics by the new and powerful cooling technologies which were then being developed. I felt certain that they would allow us to view nature's behavior in a new realm, and that important new science would follow. Many of these new technologies involved the unusual properties of liquid and solid He3. While He4 is a reasonably abundant isotope, at least in the United States; 3He has a very low natural abundance. It only became available for study in the mid 1950's with the advent of growing nuclear arsenals in the United States and the Soviet Union. The tritium used in the triggers of hydrogen bombs decays with a half life of about 12 years, and what is left is one of the most wondrous materials known to mankind: He3.
Helium 3 is wondrous for many reasons. Both isotopes of helium share the unique ability to remain in a fluid state even as they are cooled to temperatures arbitrarily close to absolute zero. However, the properties of liquid He3 and He4 are remarkably different at low temperatures, owing to the spin which only the He3 nucleus possesses. This spin makes He3 atoms Fermi particles, in the same family of particles as the electron, and as such He3 atoms must obey the Pauli Exclusion Principle, which forbids any two indistinguishable (identical and sharing the same volume) Fermi particles from occupying the same quantum state. By contrast, He4 atoms are Bose particles, as are quanta of light, and Bose particles have a built in propensity to accumulate in the same state at very low temperatures through a process named 'Bose-Einstein condensation'. Indeed, below about 2.2 degrees Kelvin, liquid He4 undergoes a phase transition to a superfluid state, in which a large fraction of all the atoms in the fluid condense into a single quantum state, and hence can be described by a macroscopic quantum wave function. In the superfluid phase, liquid He4 can flow forever, without the dissipation of any energy. To temperatures almost three orders of magnitude closer to absolute zero, however, liquid He3 exhibits no such transition.
The low temperature behavior of Fermi fluids is rather unusual, however, as is illustrated by conduction electrons in metals. Even very close to absolute zero, these electrons whiz around at remarkably high velocities, approaching one percent of the velocity of light! Because 3He atoms are so much more massive than electrons, their velocities are not much more than those achievable on the German autobahns. In many metals, the conduction electrons undergo a phase transition to a superconducting state, in which electrical currents can flow without the dissipation of energy. This is also a macroscopic quantum state in which some fraction of all the electrons can be described by a single quantum wave function. How this can be true for a collection of Fermi particles was first understood by Bardeen, Cooper and Schrieffer in their BCS theory, published in 1957, in which the conduction electrons formed pairs, called Cooper pairs, which then behaved like Bose particles.
Shortly after the publication of the BCS theory, physicists began to wonder if other Fermi fluids at low temperatures might also undergo a BCS phase transition. Only two other Fermi fluids were known which would remain in a fluid state to low enough temperatures: liquid 3He, and the dense matter in neutron stars. Thus began a search for superfluidity in He3 which continued until almost the mid-1960's. Theorists would predict transition temperatures, experimentalists would, through heroic efforts, cool liquid He3 to below this temperature and see nothing, and then the theorists would construct a new theory which predicted a much lower transition temperature. By the time I entered graduate school in 1967, most people believed that the BCS transition in He3 would be below 50 microkelvin, well below the temperature range accessible for study.
This did not cause low temperature physicists to lose interest in He3, however, for it was clear that the solid also possessed some rather interesting properties. For example, the atoms in the solid were only marginally localized to their lattice sites, and in the lowest density solid (in equilibrium with the liquid) these atoms would exchange lattice sites with their neighbors up to 40 million times each second. For silicon, a more typical element, this rate is slower than one exchange over the age of the universe. This particle exchange leads to an effective spin-spin interaction between the exchanging He3 atoms, and it was believed that this would lead to nuclear spin ordering to an antiferromagnetic state at a temperature of about 0.002 K ( 2mK).
It was this solid spin-ordering transition which I was looking for in 1971, as a graduate student with Professors David Lee and Robert Richardson, when I by chance discovered the existence of a phase transition in a mixture of liquid and solid He3 at a temperature of about 2.6 mK, where no transition had been expected. Indeed, what I was doing at the time was not the experiment I was supposed to be working on. This lecture describes these events, and the subsequent seven month period during which we slowly began to learn the true nature of the new state of matter we had discovered. What we had found were three new superfluid phases. These are the only known BCS states to this day in which the Cooper pairs are magnetic, and they all exhibit liquid-crystal-like textures in which the properties of the fluids are different in different directions, but change smoothly with distance. Studies of these phases have vastly expanded our understanding of macroscopic quantum states, and have shown the broad applicability of the BCS theory. Today, these states represent model systems for testing our understanding of remarkably diverse aspects of nature. |
