Thermodynamics and Statistical Mechanics
Neither of these closely related subjects
had been much cultivated by physical chemists in Oxford before the Second World
War, although Hinshelwood had written an engaging short textbook on
thermodynamics in 1926. Everett, some of whose early work at the PCL was in
chemical kinetics in solution, and Lambert, whose field was the study of the
rate of energy transfer between molecules, both undertook some thermodynamic
measurements on liquids and gases respectively in the late 1940's, initially as
an adjunct to their kinetic work. Everett's departure in 1948 removed an
important component of this work.
Brian Smith came to the laboratory as an ICI
Fellow in 1959 from Hildebrand's group at Berkeley and brought with him two
interests which were later to become major fields of work in the Department -
the physics of liquids and the use of computers for simulating matter in the
condensed states. His first research project was to measure virial
coefficients: this involved vast quantities of mercury, and Hinshelwood was
extremely tolerant of this very expensive operation. Soon after, he began to
tackle the question of the volume changes associated with the mixing of hard
spheres of different sizes. Having heard of the Molecular Dynamics method from
Berni Alder in Berkeley, he began a study using the Monte Carlo method, partly
with the help of Ken Lea who was then one of the very few experts on computing.
This study is believed to be the earliest European molecular simulation: the
first results were published in 1960. Shortly after, John Diamond, now a Reader
at Glasgow University, did further work on virial coefficients, and Maurice
Rigby, now a Senior Lecturer at King's College, London, began the long project
of measuring gas viscosities that continued until 1989. Leslie Sutton was
interested in the properties of polar gases, and work designed to describe and
account for their interactions, in both like and unlike pairs was done very
much in collaboration with David Buckingham (then briefly in the PCL as an 1851
Senior Student) and with Brian Smith. Kenneth Lawley (now Lecturer at Edinburgh
University) was a D.Phil student at the time. Measurements, including those of
the second virial coefficients, on like pairs could be interpreted
qualitatively in terms of a simple point dipole model, but this failed in more
complicated systems. Research in the early 1970's showed that much earlier work
in the 1930's and 40's to determine the viscosities of gases at high
temperatures was erroneous. The new measurements enabled progress to be made in
determining the intermolecular forces of simple gases, and shortly afterwards,
a series of inversion methods were discovered which allowed thermo-physical
measurements to be inverted to give intermolecular potentials directly. Geoff
Maitland played a large part in these discoveries, and much of this work is
described in Intermolecular Forces (OUP, 1981), by G.C. Maitland, M.
Rigby, E.B.Smith and W.A. Wakeham. More recently, the group has continued to
develop inversion methods, and to provide measurements of the viscosities of
simple gases and gas mixtures over a wide range of temperature.
The statistical mechanics of liquids and
interfacial systems, and the use of computers for simulation are Rowlinson's
interests and so these fields were re-introduced as a main stream of the
Department's work on his appointment in 1974. He had done his D.Phil. with
Lambert and so had become involved in the thermodynamics of gases. As a
lecturer in Manchester he had taken up the more difficult problems of liquid
and liquid mixtures, and had also become interested in the technical
applications of this work. He came to Oxford from Imperial College, where he
had been Professor of Chemical Technology. By the early 1970's many of the
problems of the properties of liquids were being solved, but their surface
properties were still a field almost untouched by modern theory. His years at
Oxford have, therefore, been devoted to the development of the statistical
mechanics of fluid interfacial systems and to their simulation on computers.
Much of the early part of this work was included in his Molecular Theory of
Capillarity (OUP, 1982), written in collaboration with B. Widom from
Cornell who was the third of the I.B.M. Visiting Professors of Theoretical
Chemistry.
In the last ten years much of his work has
been on inhomogeneous fluids of more complicated geometry – drops, bubbles, the
intersection of three surfaces in a line (as in a foam) and the state of fluids
adsorbed into cavities, such as those found in zeolites. Such systems have
proved to have extremely subtle mechanical and thermodynamic properties, and
the whole field is still one of lively controversy. He had, during much of this
period, the advantage of collaboration with two able research associates, J.R.
Henderson (now a Lecturer at Leeds) and F. van Swol (now an Assistant Professor
at Illinois), and of a collaborative programme with K.E. Gubbins of the
Chemical Engineering Department at Cornell, who has been a frequent visitor to
the PCL on sabbatical leaves. After his retirement in 1993, Rowlinson will
continue such collaboration as Andrew D. White Professor-at-Large at Cornell, a
part-time appointment which he can hold until he is seventy.
The rapid development of computers has led
to them finding ever growing areas of application in chemistry, and so when the
D.E.S. offered to fund a lectureship under its 1984 `new-blood' scheme, the
department seized the opportunity to appoint Paul Madden as a lecturer in
computational chemistry. (His formal title is, however, the conventional one of
lecturer in physical chemistry). He is principally concerned with problems of
molecular motion in liquids and solids, normally pursued via the methods of
computer simulation. One question of current interest, involving classical
simulations, is that of the the pyroelectricity (the change of macroscopic
dipole moment with temperature) of Langmuir-Blodgett films of aliphatic
molecules: another is that of the fluctuating polarisability in ionic systems.
Quantal simulations, in which knowledge of the motion of the electrons as well
as that of the molecules is required, are also being used in the attempt to
elucidate the metal/insulator transitions in alkali metal fluids. With the
appointment of David Logan, two years later, the theory of matter in the
condensed states became broadly established as one of the main fields of the
work of the PCL. David Logan's particular interests lie in the study of the
electronic properties of liquids and amorphous solids. The often subtle
interplay between structure, disorder and the effects of electron correlation,
in determining electronic properties and driving electronic phase transitions
in liquid metals and their alloys, is currently under investigation. So too are
related problems involving the dynamics of vibrational excitations and energy
flow in systems of coupled anharmonic oscillators.
