Published: Wednesday, August 24 2011 09:28
(1926 - 2005)
Richard Alan Ferrell, Professor Emeritus of Physics at the University of Maryland in College Park, died at his home in the arms of his wife Miriam on November 14, 2005, after a several year struggle with multiple myeloma. He was 79. During his long career as a theoretical physicist, Ferrell made major contributions to theoretical physics, especially condensed matter and statistical physics.
Ferrell was born in Santa Ana, California in 1926. After service in the Navy, he attended the California Institute of Technology, receiving the bachelor’s and master’s degrees. He is remembered as the only CalTech student of his time able to do all the problems in `Smythe’. He received his doctorate in 1952 at Princeton under Arthur Wightman, and then did post-doctoral work with Werner Heisenberg at the Max Planck Institute. In 1953 John Toll invited him to join the faculty of the nascent physics department at the University of Maryland in College Park. Toll was so convincing that Ferrell accepted the appointment before visiting College Park. He became full professor in 1959.
Ferrell’s earliest research was in quantum electrodynamics. Positronium had been discovered in 1950 by Martin Deutsch and several theorists competed to calculate the fine structure. Ferrell was the first to produce a completely correct result. Later, he published work on muonium and on positronium in helium.
Positronium led to his interest in positron annihilation in solids. His 1957 Reviews of Modern Physics article on this subject is still a standard reference. Positron annihilation is many-body problem, since it is necessary to find the response of the negatively charged electrons to the positively charged positron. At about the same time much of the general electronic response problem was solved by Lindhardt and his dielectric function. However, that work was published in the Proceedings of the Danish Royal Society, (Kgl. Danske Videnskab. Selskab…) and did not immediately become known to most Americans. So Ferrell had to invent for himself the required techniques.
He soon turned his attention to more general many-body physics. With John Quinn, he pioneered the self-energy approach to electron correlations. This was before the introduction of general Green’s function techniques into the many-body field. The zeros of the dielectric function give the plasmons. These can be observed as energy losses by energetic electrons passing through thin metallic films. An additional energy loss at lower energy was attributed to excitation of surface plasmons. However, this energy was too low to be explained by the existing theory. Ed Stern, an experimentalist colleague, and Ferrell showed that the oxide layer on the surface of the Al film accounted for the effect. At first this was controversial, but experiments measuring the energy loss during the oxidation of the film were conclusive. The nuclear physics analog to the plasmon is the giant dipole resonance. With his student Stavros Fallieros, Ferrell applied the `random phase approximation’ to this problem of nuclear physics.
Connecting electron self-energies and thin films to superconductivity Ferrell realized that with an appropriately prepared film the first order superconducting transition induced by a magnetic field can be turned into a continuous second order one. Ed Stern and his student verified this effect experimentally.
During this intermediate period, Ferrell made contributions to both nuclear and condensed matter physics. In the case of nuclei, he was particularly interested in correlation effects, corrections to the shell model, in light nuclei. He showed by a shell model analysis that an intermediate (between jj and LS) spin-orbit coupling provided the best description of beta decay FT values for light nuclei. He also established the connection between certain shell model wavefunctions and the Hill-Wheeler-Griffin collective wave functions. He and others utilized this subsequently to explain the enhanced excitation of the low lying 2+ and 0+ states in C12 in the inelastic scattering of energetic electrons. This is a consequence of the states’ collective quadrupole and dilatational characters. With William Visscher, he resolved the puzzle posed by the long C14 beta decay lifetime: he showed that the ground states of C14 and N14 implied by a shell model with mixing by (empirically fixed) two-body interactions produced a nearly total accidental cancellation of the beta decay matrix element for the decay.
Ferrell also employed a time dependent Hartree-Fock description of the dilatational mode in O16 , to show that the O16 6.06 MeV 0+ state was not primarily a collective dilation, as had been conjectured by others. In 1966 with William MacDonald he described the fine structure resonances observed in the Al28(P, γ)Si28 reactions in terms of “hallway” states which couple to the reaction continuum via the Feshbach “doorway” states. The latter determine the energy and width of the enveloping resonance. This seems to have been his last contribution to nuclear physics.
Ferrell’s work in condensed matter physics during this period up to 1966 is too varied and too copious to recount in detail. We give only a few examples. They are characterized by a close connection to experiment, an intuitive approach, followed by theoretical calculations.
A large part of this work was on superconductivity. The first paper was with Rolfe Glover III. It evolved in interaction with Michael Tinkham and resulted in the famous Ferrell-Glover-Tinkham sum rule. This is still very important for recent high temperature superconductivity analyses. It asserts that the finite frequency response which is lost in the superconducting transition reappears at the zero frequency `superconductivity’ delta-function. Ferrell’s ideas of the effect of spin-orbit scattering on the Knight shift in superconductors provided the key to the puzzle regarding the failure of the spin-susceptibility to vanish at low reduced temperature in some materials. Ferrell had several papers on the Josephson effect in which experiments were suggested.
One paper of 1963 with his student Peter Fulde proposed what is probably the first unconventional superconductor. This phase was independently proposed by Anatoly Larkin and Yuri Ovchinnikov and is now called the FFLO or LOFF phase for superconductivity in a strong spin-exchange field. After nearly forty years there has been a resurgence of interest in these ideas starting with the possible observation of the phase in CeCoIn5. Connections with states of `cold atoms’ are also being discussed. There may also be a subtle relationship between the FFLO state and d-wave superconductivity. These issues are by no means settled: it is an active field of research.
A new phase of his career started in 1966 when he teamed with a number of junior visitors from Hungary (Nora Menyhard and Peter Szepfalusy), Germany (Hartwig Schmidt) and Austria (Franz Schwabl). They produced important ideas which can be considered the birth of the field of statistical physics now called dynamical scaling theory for critical phenomena. These ideas are, of course, closely related to the scaling and renormalization group theories of Benjamin Widom, Leo Kadanoff, Michael Fisher and Kenneth Wilson. This work of “Ferrell and the United Nations” motivated the well-known papers of Pierre Hohenberg and Bertrand Halperin, a couple of years later.
The basic insight was that not only static quantities, e. g. susceptibilities, but dynamical quantities, such as thermal conductivity, will diverge at the phase transition and will be described by critical exponents. They predicted that as the superfluid transition in helium is approached from the normal phase, the thermal conductivity will diverge. If the transition is approached from the superfluid phase, the second-sound damping diverges. The experimental verification came soon after, although a 20% discrepancy remained. These was cleared up decade later, when Ferrell, Volker Dohm and Jay Bhattacharjee pointed out that the corrections to the asymptotic limit are rather large in this case, and made a quantitative estimate. Gunther Ahlers and Hohenberg, and Dohm and Reinhard Folk followed this up in every detail.
Ferrell’s style changed somewhat at this time. This was because he needed to control the severe migraine headaches which afflicted him. That required a nap in the dark each afternoon, and thus he could not continue his wide open interactions with all comers. Nevertheless, he maintained his prodigious output, much of which evolved from the dynamic scaling work. In fact, more than three-quarters of his papers date after this time. In particular, he has done much work with Jayanta Bhattacharjee, now professor at the IIT in Bangalore, who worked with Ferrell between Jay’s period as a graduate student until Ferrell’s death.
We give only a few examples. In the ‘70’s Ferrell began a three-decades effort to understand the shear-viscosity fully near the critical point. This was motivated by Kawasaki’s description of the dynamics near a liquid-gas transition, and Jan Senger’s observation of the divergent thermal conductivity. Ferrell became convinced that to confront theory and experiment properly, one should concentrate on quantities, like the shear viscosity, with small critical exponents. Experimentalists, many from the Maryland-DC area, worked on this problem in interaction with Ferrell. These include Jan Sengers, Herschel Burstyn, Robert Gammon, Sandra Greer, Michael Moldover, and Robert Berg.
For a weak divergence, it is necessary to have a reliable technique for separating the critical component from the noncritical. A crossover function to do this was found in 1981. The weak divergence of the shear viscosity had been ignored in the Kawasaki function. The light scattering experiments of Sengers, et al required a relook at this problem. The small but vital correction was found in 1983. See also the earlier theory of Eric Siggia, Halperin and Hohenberg.
What if the viscosity is measured by a vibrating wire viscometer? The finite frequency of the vibrating wire can become equal to the decay rate of fluctuations very close to the critical point. Then the viscosity is affected by the frequency and viscoelastic effects set in. Ferrell predicted this effect for a classical fluid in 1980. The definitive experiment followed only in the mid nineties. Measuring the predicted effect depended on getting very close to the critical point, normally impossible because of gravity stratification. Hence the experiments had to be carried out in the space shuttle. They were, and among other things, they finally suggested an exponent of 0.0690 with a possible error of 0.0006.
This called for a strong effort on the theoretical side and a three loop calculation, the only one for a dynamic exponent, eventually yielded a value of 0.0679 with a contribution from the missing loops estimated at 0.0007. This was the final paper, with Jay Bhattacharjee and Hong Hao, that Ferrell published. It appeared less than a year before his death.
Many other subjects attracted Ferrell’s attention during this time. He studied sound propagation near a critical point. Dynamic scaling was applied to binary liquid mixtures near the consulate point. Application of the techniques was extended to weakly first-order isotropic-nematic transition in liquid crystals. The predictions of this period were repeatedly confirmed in experiments over the next twenty years.
In addition to dynamic critical phenomena, Ferrell was fascinated by the small exponents of static critical phenomena. He focused on the anomalous dimension index. In the early seventies, he collaborated with Douglas Scalapino to develop a screening approximation to find this exponent. They followed this by a study of the analytic properties of the critical correlation function, which provided a convenient way of estimating the elusive exponent.
Ferrell also maintained his interest and published considerably on superconductivity and particularly the Josephson effect. The coexistence of spiral magnetic order and superconductivity in the Chevrel phase compounds inspired a paper jointly authored by theorists and experimentalists, including his colleague Jeffrey Lynn. His last contribution to the literature on superconductivity, in 1995, was a calculation of the sound attenuation as a way to distinguish s- and d-wave superconductivity.
In addition to research, Ferrell took very seriously his role as professor. He had a number of students whose achievements are exceptional. We mention again John Quinn, who until recently was Chancellor at the University of Tennessee, Peter Fulde, head of the Max Planck Institute for the Physics of Complex Systems in Dresden, and also Alan Luther, Professor at Nordita, who won the Buckley Prize a few years ago together with Vic Emery.
A somewhat unusual accomplishment in this respect concerned the large lecture halls which were built at Maryland. He and colleague Carroll Alley were the major faculty influence on the design which had rotating stages so that demonstrations can be prepared and practiced behind the scenes. This lecture-demonstration facility is internationally known and is still one of the few best in the world.
On the scale of the university community, Ferrell was very active. The University and City of College Park wanted to wait until they had enough money to plant trees big enough not to need watering. Ferrell persuaded them to let him raise money, personally plant many small trees, and personally water them. A famous story is that once he talked Rolfe Glover III into watering the trees in his absence. Rolfe, as a good experimentalist, watered the trees in the middle of US Rte 1 on the way home after working in the lab until 2 AM. He was unable to persuade the local police of his motives, and needed help from his wife to be released from custody the next morning.
Globally, Ferrell was very much a citizen of the world. For example, he made a point of learning the language of a country when he made an extended visit. In 1968 he visited both France and Russia, with his young family, driving from Paris to Moscow and back in a new Renault. While in Paris he learned both French and Russian, using a French-Russian dictionary to help him. On the way back, Ferrell brought back numerous preprints from the Soviet scientists he had visited. These were confiscated by the Russian police, who searched the car from under the fenders to inside the door panels. They probably thought the preprints were state secrets. However, his command of Russian was good enough that he persuaded the police to let him out of the country. This was a close call, as the Renault stayed only a few hundred miles ahead of the tanks coming to put an end to the `Prague Spring’ of that year.
Ferrell had other passions in his life besides physics. His wife, Miriam, and children, Rebecca and Robert, were certainly at the top of the list. Another passion was his ski house at Magic Mountain, VT. This is a magnificent edifice that he designed and built mostly by himself, with sporadic but crucial help from his family, his colleagues, his friends, his students and even an occasional local professional. Very much a conservationist, he insisted on using local materials. He had great confidence in his ability to figure out how to do any specific project. In fact, he was almost always successful, but, (as with his physics research), he was often too optimistic in his expectation of how longs things would take. An example is the small dam he built which created a pond for swimming or ice skating. Ferrell had seen and understood the Hoover dam, so it was just a question of scaling things down. However, building a curved frame of flat plywood took longer than anticipated, and when the concrete truck arrived the next morning, the frame was far from ready. This was not the first Ferrell project needing concrete, so the concrete company knew what to expect, and the driver waited several hours, and finally the dam was a great success.
Another story worth recounting concerns the installation of the roof of the chalet. Ferrell had succeeded in persuading Arnold Glick and Fred Koch to assist. Logically enough they started working from the top down. After all, it is important to have a good roof which fits together at the very top. Experienced carpenters, on the other hand, start from the bottom up because then there is something to stand on as the roof goes up. The story goes that there was a local retired carpenter who watched them at work all day, mightily amused at the way city folks build their houses.
Almost always, as with his physics research, Ferrell’s intuition and skills carried the day, even though some of the hard work went more slowly than anticipated. One exceptional case of failure concerns the paneling of the ski house. Ferrell thought that local Vermont maple would be beautiful and environmentally correct. Unfortunately, after a few months, the maple began to warp, and it was necessary to replace it with kiln dried wood from Oregon.
There are many more stories fondly remembered by Richard Ferrell’s friends and family. He was a person so alive that he needed to cram several lifetimes into the 80 years allotted him. Every day, he made a list of projects on a slip of paper so as to make the best possible use of his time.He will be missed by all who knew him.
We thank Jay Bhattacharjee, Michael Fisher, Pierre Hohenberg, Jim Griffin, John Quinn, Jan Sengers, Ed Stern, John Toll and others for reminiscences and for reminding us of some of Ferrell’s physics. Mistakes in the summary of the physics are ours, not theirs.