Table of contents

The Third Liquid Matter Conference took place at the University of East Anglia, Norwich, UK from 6 to 10 July 1996. This event, which followed the successful conferences in Lyon (1990) and Firenze (1993), was organized under the auspices of the Liquids Section of the European Physical Society and was co-sponsored by the University of East Anglia and the Institute of Physics. The aim of these conferences is to provide a forum for exchange of the most recent results and ideas between physicists, chemists, chemical engineers, material scientists and biologists working in a wide spectrum of liquid matter science. This special issue comprises the majority of the papers presented orally at the conference, i.e. as plenary or symposium lectures. A forthcoming issue of the journal will contain several papers based on poster contributions.

The Norwich Conference attracted some 460 delegates, with significant representation from each continent. Eleven plenary lectures, 84 symposium talks and three poster sessions, in which approximately 350 posters were presented, formed the scientific programme. In addition there was a very lively Round Table on Perspectives for Computer Simulation chaired admirably by G Ciccotti and D Frenkel. In keeping with the tradition of the Conference Series, the eleven symposia were chosen to reflect the breadth of the subject: simple liquids and solutions; ionic and conducting liquids; reaction dynamics; quantum phenomena; liquid crystals; polymers, polyelectrolytes and gels; colloids, amphiphiles and emulsions; interfacial phenomena and membranes; waves, instability and turbulence; supercooled liquids and glasses; and phase transitions. Interactions between delegates were encouraged by the social activities which took advantage of the 270 acres of beautiful parkland in which the University of East Anglia is situated, the fine city of Norwich, and the richly agricultural county of Norfolk. The social programme included an enjoyable welcome reception in an enormous marquee, a concert of music by Mozart and Brahms, conference excursions to Blickling Hall and the Norfolk Broads, and culminated in a City Reception and Banquet at St Andrew's and Blackfriars' Halls, long to be remembered for the splendid regalia and spontaneous post dinner dancing! A liquids theme for the promotional picture, initiated at the Lyon and Firenze meetings, was maintained by adopting the 1824 painting of the Thorpe Water Frolic by Joseph Stannard who was a leading member of the Norwich School of Artists.

We do not attempt to select highlights of the conference, rather we encourage the reader to browse through these proceedings and sample the excellent science that was presented. It is our strong impression that the conference achieved its goal of bringing about stimulating interactions between experimentalists, theoreticians and simulators from the various liquid matter subject areas and the different disciplines contained therein. Thus the basic philosophy of the Conference Series was perpetuated, which bodes well for the Fourth Conference to be held in Spain in 1999.

It is a great pleasure to acknowledge all those individuals and organizations without whose help the Conference would not have been such a great success. Special thanks go to the supporters listed overleaf, to the programme, advisory and local committee members, and to the faculty, students and other personnel at the University of East Anglia who assisted in innumerable ways.

P S Salmon (University of East Anglia) J E Enderby, R Evans (University of Bristol) P N Pusey (University of Edinburgh) The Editors

Supporters

The organizers gratefully acknowledge the support of the following individuals and organizations:

Academic Press Air UK Barclays Bank plc Berol Ltd City of Norwich Collaborative Computational Project (CCP) 5 Current Science Group Mr Dennis Lister Engineering and Physical Sciences Research Council (EPSRC) European Physical Society (EPS) Institute of Physics Publishing Ltd International Science Foundation Leybold Ltd The Numerical Algorithms Group (NAG) Ltd Pilkington plc The Royal Society Springer-Verlag Unilever plc University of Bristol University of East Anglia Waterstones

International Programme Committee

Conference Chairman J E Enderby, Bristol

Programme Committee Chairman P N Pusey, Edinburgh

A C Barnes, Bristol J Piasecki, Warsaw A Castellanos, Sevilla L F Rull, Sevilla V Degiorgio, Pavia P S Salmon, Norwich W Freyland, Karlsruhe A Sanfeld, Paris J F Joanny, Strasbourg J Teixeira-Dias, Coimbra D Langevin, Pessac L M Torrell, Göteborg P A Madden, Oxford J Troe, Göttingen

Local Organizing Committee, Norwich

Local Committee Chairman P S Salmon, Physics

G C Barker, Food Research D C Champeney, Physics N E Cusack, Physics K W R Gilkes, Physics E J A Lea, Biology B H Robinson, Chemistry D C Steytler, Chemistry R M Wood, Physics

International Advisory Committee

P S Belton, Norwich J A Janik, Kraków T Boublik, Prague A J Leadbetter, Grenoble S Bratos, Paris J C Leyte, Leiden D Chandler, Berkeley A E Merbach, Lausanne S F Edwards, Cambridge P Migliardo, Messina P A Egelstaff, Guelph M-L Saboungi, Argonne D J Evans, Canberra L Sjögen, Göteborg R Evans, Bristol S Tamaki, Niigata J L Finney, London M M Telo da Gama, Lisboa G Fytas, Crete M P Tosi, Pisa F Hensel, Marburg J W White, Canberra I B Ivanov, Sofia B Widom, Ithaca

Experiments on critical behaviour in certain electrolytes pose puzzling theoretical challenges. Some recent progress by the author and his co-workers in meeting these challenges through the study of the restricted primitive model (hard spheres carrying charges ) is summarized briefly. However, basic questions regarding the universality class(es) of ionic criticality and the factors determining cross-over from mean-field to Ising behaviour remain unresolved.

Recent precise experimental results of neutron elastic and inelastic scattering in simple fluids, like noble gases in various density ranges, are briefly summarized. The comparison of these results with available theoretical calculations and the relevance of this experimental method, with the present state of the art, for the study of the microscopic properties of these systems are discussed.

There are five mechanisms which can lead to mesoscopic interfacial structures: large interfacial fluctuations, the appearance of new length scales, the divergence of the bulk correlation length, long-range forces and the presence of massless Goldstone modes. The importance of the first four mechanisms for roughening, wetting, critical adsorption, and the orientational order in dipolar fluids, respectively, is discussed.

We will discuss the meaning of piezo-, pyro-, ferro- and antiferroelectricity (defined for solids) and their possible appearance in the liquid state. We will do this using general symmetry principles as a starting point. The arguments will also be used for the prediction of non-linear optical properties in the liquid state.

Many viscoelastic surfactant solutions contain giant, self-assembled micelles. These can be described as `living polymers', whose chains are subject to reversible scission and recombination. Their dynamics in the entangled regime is accordingly modified from the reptation picture for conventional polymer chains. For rapid scission kinetics, the linear viscoelastic spectrum approaches a single-exponential (Maxwell) behaviour: small departures from this can be measured, and the model used to deduce information both on the micellar kinetics (the lifetime of a typical micelle before breaking) and on the structure (the mean micelle length). These ideas work for several systems, but for others, unreasonable trends for these quantities are found. The most likely reason for this is micellar branching effects, which (as far as the reptation - reaction model is concerned) introduce an effective micellar length equal to the mean distance between branch points. Another possible discrepancy comes from the breakdown of mean-field averaging for the micellar reactions. The reptation - reaction model yields a non-linear constitutive equation which shows a non-monotonic dependence of stress on strain rate, in simple steady shear. This leads one to expect flow instabilities, and (with further assumptions) suggests that steady shear-banded flows should arise, in which macroscopic layers of fluid of different shear rates coexist. Several experimental observations support this general picture, although the same instability could instead lead to wall slip, or unsteady flows.

The polymer dynamics near the glass transition can be classified into three different categories: (i) anomalous low-frequency vibrations, the so-called boson peak, which are accompanied by a fast relaxational process; (ii) secondary and higher-order relaxations, which are assumed to involve only local molecular motions and (iii) the structural relaxation underlying the flow processes which freezes at the glass transition. This paper presents inelastic and quasielastic neutron scattering data on these dynamical processes, emphasizing polybutadiene. In the light of these data, current models on the dynamic glass transition are inspected. In a second step, recent experiments on the coherent dynamic structure factor are exploited in order to learn about the molecular nature of the various polymer motions.

The classic problem of evaluating the free energy of an electric double layer around a planar electrode or polyion through an appropriate charging process is reconsidered, within the Poisson - Boltzmann framework. After a brief consideration of an infinite planar electrode, we examine finite-size effects. An application is made to swollen arrays of disc-shaped clay platelets.

Experiments made possible by new laser methods explore ultrafast features of solution phase reaction dynamics. The results from two examples are given. Photodissociation of small molecules such as and in solution yields highly excited vibrational state distributions and narrow bond length distributions. The dynamics in the transition state region and the ensuing dynamics can be followed by femtosecond laser experiments that probe the phase and energy loss processes.

The mercuric iodide experiments show that the phase relaxation is dominated by coherence transfer processes which have a strong analogy to classical motions. The population relaxation dynamics gives results similar to those from classical molecular dynamics simulations which suggest, notwithstanding the significant polarity of the molecules and solvents, that the relaxations are dominated by Lennard - Jones repulsion at sufficiently low frequencies.

The hierarchical reference theory (HRT) is a unified theory of fluids: in the dense regime it has the accuracy typical of a good theory of the liquid state. At the same time, close to a critical point, it develops the structure of a renormalization group theory in which all non-universal quantities can be deduced from knowledge of the interatomic interaction. The HRT can be applied above as well as below the critical temperature so that the phase diagram, thermodynamic properties and distribution functions are obtained in a unified way. Similar information can be extracted also from models of binary mixtures. In this case, the HRT allows for the determination of the order parameter along lines of critical points and provides an explanation of the strange crossover phenomena found in mixtures.

Drops and bubbles are ubiquitous. Stokes, Rybczynskii, Hadamard, Boussinesq and others provided the drag law experienced by a drop. Young, Goldstein and Block and others studied the motion due to interfacial stresses induced on a drop immersed in a medium where an imposed thermal gradient exists. They demonstrated the possibility of its levitation in the presence of gravity. Levich, Sanfeld and others pointed out the role played by surfactants in affecting the drag law and the possible fluid motion inside the drop. All those works refer to passive drops, i.e. drops experiencing at most the interfacial stresses due to variation of interfacial tension with temperature or surfactant concentration in the surrounding fluid near the drop surface. After providing a succinct account of the results of the earlier theories and some relevant experiments, we consider the behaviour of an active drop, i.e. a drop with chemical reaction at its surface or with an internal heat generation source, etc. Attention is focused on the case of a drop immersed in a homogeneous surrounding when due to surface stresses (the Marangoni effect) and, consequently, due to thermo/soluto-hydrodynamic instability there is spontaneous breaking of the radial symmetry of the temperature and/or concentration distributions, hence overcoming the drag and originating self-sustained translational drop motion. Moreover, the autonomous motion may offer a multiplicity of steady values for a given external (weak) force like buoyancy, and levitation is possible for multiple (weak) buoyancy levels.

Liquid is a unique liquid because of its Bose nature but it can be used to study general problems. A number of such properties are reviewed including quantum evaporation, interactions between excitations and wetting studies with Cs.

Recent neutron diffraction experiments, which exploit hydrogen isotope substitution techniques to extract the HH, OH and OO site - site radial distribution functions for water, indicate that as the temperature of water is raised above the critical point, the hydrogen-bonding network, as measured by the height of the first peak in the OH distribution function, collapses. Several computer simulations, however, dispute the accuracy of the experimentally determined distribution functions: they show that the measured data cannot be obtained from any physical arrangement of water molecules. By applying additional constraints on the small-radius behaviour of the radial distribution functions, and by repeating some of the experiments under different experimental conditions, the accuracy of the extracted radial distribution functions has now been improved. The new distribution functions, while not qualitatively different from what has already been published, satisfy the fundamental objection to the previous results, namely they can be simulated with physical assemblies of water molecules. These simulated distributions of molecules indicate a weak degree of hydrogen bonding in water at 673 K which is greatly reduced compared to that in the hydrogen-bond network of ambient water. The analysis is supported by a new computer simulation of water structure using an `empirical' water interaction potential, which contains none of the long-range features of traditional charge models, such as SPC/E. This short-range potential is able to reproduce most features of the experimental data to good accuracy, even under supercritical conditions.

The collective dynamical properties of water are discussed with reference to recent inelastic scattering data as well as to simulation results reported here for the first time. Some striking features (notably the so-called `fast sound') which aroused in the past a considerable debate now appear to be rather well understood. In contrast, there are other additional aspects (in particular, the character of a nearly flat mode at present both in water and in ice) which undoubtedly require a better understanding.

The methods of neutron diffraction and isotopic substitution (NDIS) provide a direct means to probe the structure of a complex fluid mixture. The method is easily extended to studies under non-ambient conditions for which large sample containers can be used to withstand temperatures of up to and pressures of 2000 bar. Within the past few years NDIS experiments have been performed on aqueous solutions of several salts at temperatures up to and beyond their critical states. The results show that both the aqua ion and hydrogen bond structure change appreciably with increases in temperature.

We describe molecular dynamics simulations of supercritical water and supercritical aqueous solutions using simple non-polarizable models of water and a new polarizable model for water developed by our research group. We compare the simulation results to neutron diffraction studies where available and to experimental measurements of ion pairing in the case of supercritical aqueous electrolyte solutions. Simulation results obtained on massively parallel supercomputers are used to evaluate size effects in the simulations and to speed up the CPU-time-consuming polarizability component of the simulation.

The geometrically based fundamental-measure free-energy density functional for hard spheres is briefly reviewed. The corresponding bridge functional can be successfully applied for a large variety of quite disparate systems, including classical plasmas. The application of the Gauss - Bonnet theorem enables one to generalize the theory to molecular fluids.

High-pressure neutron diffraction measurements on fluid hydrogen chloride, carbon dioxide, ammonia, sulphur hexafluoride and methane at different supercritical states are reported. Some aspects of the experiment and the data evaluation are mentioned.The measurements were performed to assess the usefulness of some pair potentials reported in the literature for statistical mechanical calculations, based on the extended reference interaction site model (RISM) in combination with the HNC closure. The density-dependent measurements proved to be useful for tests of the individual potentials.

In the present work, we study the dynamical properties of fluids confined at molecular scales. First, we show that, for a fluid confined between two parallel solid walls, the phenomenological hydrodynamic description of bulk fluids holds down to nanoscopic scales, once appropriate boundary conditions at the solid walls are applied. Then, we investigate the influence of confinement on the self-diffusion coefficient (in the direction parallel to the walls) in a fluid slab. We have computed the effect of confinement on the mode-coupling contribution to the diffusion coefficient. These finite-size corrections are shown to reduce the diffusion constant by an amount , where h is the thickness of the fluid slab and the atomic size. This behaviour can be interpreted in terms of the suppression of long-wavelength modes in the backflow effect, due to confinement.

The self-diffusion coefficient of tetramethylammonium counterion has been measured in aqueous solutions of polymethacrylic acid (PMA) of 100 and 1000 monomers as well as in persistence-length DNA fragment solutions. The concentration ranged from to 1 M (monomolar) and no low-molecular-weight salt was added. The counterion self-diffusion coefficient was found to be independent of the kind of polyion and of molecular weight over a large concentration range. In the low-concentration region the self-diffusion coefficient is a function of polyion length and species. In this concentration region the self-diffusion coefficient is the smallest in solutions of the polyion with the largest persistence length (TMADNA). The self-diffusion coefficients in solutions of the two PMAs start to deviate upon dilution at the crossover concentration from semidilute to dilute for the shorter PMA (100 monomers). In the dilute solutions of the short PMA (100 monomers) the counterion diffusion coefficient is lower than that in semidilute solutions of the higher-molecular-weight polyelectrolyte at the same monomer concentration.

At elevated temperatures (about ) liquid alkali metal - alkali halide solutions transform continuously from the nonmetallic to the metallic state (NM - M transition) as a function of the metal mole fraction . In this study we present results of new experiments on spectroscopic ellipsometry and on absorption spectroscopy across the transition regime. The data indicate that on both sides of the NM - M transition localized and mobile electronic states may coexist

The structural and the electronic properties of liquid alkali metals along the liquid - vapour coexistence curve and along the melting curve are investigated by the ab initio molecular dynamics (MD) simulation, in which the Kohn - Sham energy functional is minimized for each ionic configuration of the MD step using the preconditioned conjugate-gradient method.

It is shown for the expanded liquid Rb that the calculated structural functions are in good agreement with the experiments and that the electronic states are strongly correlated with the ionic configuration and tend to localize due to the large spatial fluctuation of ionic density with decreasing density.

As for the compressed liquid Rb under high pressure, the structure obtained by the simulation agrees well with the recent experiment; that is, for the pressure less than 3 GPa the liquid contracts uniformly but above 3 GPa it starts to deviate from the simple uniform compression. This structural feature is related to the electronic s - d transition due to the high pressure, which is clearly seen in the calculated density of states.

Turbidity measurements showing crossover from mean-field to Ising criticality have been reported by Narayanan and Pitzer for the liquid - liquid phase transition in ionic solutions of alkyl-ammonium picrates in higher alcohols. The Ising region was found to increase with the dielectric permittivity D for solvents with 4 < D < 8. It was conjectured that the Ising region becomes too small to be observed for lower values of D, which is in accordance with the finding of mean-field criticality in the system triethylhexylammonium triethylhexylborate in diphenyl ether , where . In order to check this hypothesis, we investigate solutions of salts in non-protonating solvents with D<2.5. The systems are tetrabutylammonium naphthyl sulphonate in toluene and tributylheptylammonium dodecyl sulphate in cyclohexane. The location of the critical points in the corresponding state diagram is in general agreement with the model system of charged hard spheres in a dielectric continuum, i.e. the restricted primitive model (RPM). However, changes of by minute variations of the salt and of the solvent (toluene, xylene, ethylbenzene) cannot be explained by the RPM. We report measurements of the phase diagram and light-scattering results. The amplitudes of the correlation length are up to an order of magnitude larger than those typically found in non-ionic fluids. For the new systems, but also for the solution of in , Ising criticality is found in the region of .

The restricted primitive model is a complex fluid, compared to simple-fluid models such as Lennard-Jonesium. This is a result of the strong ionic association between anions and cations. Some ways in which this ionic association may contribute to the distinctive features of phase separation observed in RPM-like ionic fluids are noted.

X-ray absorption spectroscopy can be applied to the study of liquid matter yielding quantitative structural information on the short-range order. Recent experimental advances and a new single-energy detection technique suitable for monitoring phase transition and supercooling phenomena are described. Examples of the change in absorption at the melting of Ge and Ag are presented. The general theoretical framework for the interpretation of the spectra is reviewed.

An overview of recent EXAFS measurements on binary liquids, including molten ionic and superionic salts, is given. Multiple-edge EXAFS data analysis using multiple-scattering theory is shown to be able to give accurate and reliable information on short-range structure, complementary to that obtained using diffraction techniques. Structural results on alkali bromides (KBr, RbBr) are illustrated and shown to be in agreement with computer simulations based on well established interatomic potentials. EXAFS results of liquid CuBr, for which a neutron diffraction study using the isotopic substitution method is available, are presented. Evidence for a narrower Cu - Br bond-length distribution is found. The usefulness of the performance of EXAFS measurements on binary liquid systems is addressed.

To investigate the structural change in the semiconductor-to-metal transition in expanded fluid selenium, x-ray diffraction and small-angle x-ray scattering (SAXS) measurements at high temperatures and pressures up to and 1500 bar have been carried out. The results of the x-ray diffraction study show that the twofold-coordinated structure is largely preserved in the metallic fluid. The covalent bond becomes short at around , where the metallic character starts to be enhanced. These results suggest that the local structure of chain molecules changes with the transition. Since the semiconductor-to-metal transition occurs near the liquid - gas critical point, it is important to study how the critical density fluctuations are concerned with the transition. The SAXS spectra at 400 bar show that a broad peak appears at around above , which may suggest density fluctuations with a correlation length of about 50 Å in the metallic fluid.

Ab initio molecular dynamics based on density functional theory has been used to study liquid Se at three temperatures: 570, 870 and 1370 K. The calculated g(r) is in very close agreement with neutron diffraction data, except in the region of the first minimum at . We have examined the effect of including gradient corrections in the density functional description, and we find that they give a substantial improvement in the agreement with experiment near the first minimum. We analysed the bonding topology and we find a significant fraction of onefold- and threefold-coordinated Se atoms at the highest temperature.

We present a lattice-gas model with attractive interactions arising from valence electron delocalization from single-site orbitals. Such effects cannot be described by a pair-interaction potential and produce a strongly asymmetric phase diagram for a nonmetallic molecular vapour coexisting with a liquid metal. Although the model is extremely simplified, it reproduces features of the alkali fluids which contrast with those of simple fluids. A unified treatment of the ionic and electronic structures allows obtaining the electrical conductivity at chosen thermodynamic conditions. We have found that the metal - nonmetal transition in this model is driven by the break in the percolation of the ionic structure, which takes place in the vapour phase close to the critical temperature, rather than by electronic localization induced by other effects of disorder.

Electromagnetic levitation is a useful tool to process high-temperature and highly reactive melts without a container. It provides a pure environment and contamination of the melt is reduced to a minimum. With the application of non-contact measurement techniques, structures and properties of pure samples can be investigated. One major advantage of this approach is that it allows one access to the metastable state of the undercooled melt. This paper reports on measurements of thermophysical properties of undercooled liquid metals and recent investigations of their structure. We have performed measurements of the surface tension, electrical conductivity, density and thermal expansion of a number of pure metals and alloys. Some experiments were performed under microgravity conditions which reduce the strength of the required electromagnetic fields by orders of magnitude and lead to a much higher precision. We have recently used electromagnetic levitation in combination with synchrotron radiation to obtain EXAFS spectra of an undercooled Co - Pd alloy.

The dynamic structure factor, , and the static structure factor, S(Q), of liquid lithium were investigated by means of inelastic x-ray scattering and small-angle x-ray scattering. Comparing our results to different molecular dynamics simulations we found a good agreement with MD simulations using an ab initio pair potential derived from the neutral pseudoatom method. However, measurements of the absolute cross section indicate that, in addition to coherent scattering, there is an incoherent scattering contribution for energy transfers of several meV.

The solid - liquid - solid-phase transitions of a alloy have been investigated by means of the recently developed method of time-resolved photoelectron spectroscopy, using dye laser pulses to melt the sample surface on a microsecond scale. Starting at a temperature of the liquid sample just above the melting point (740 K), the laser pulse heating allowed us to obtain, for the first time, valence band spectra of an alloy at temperatures of up to 1620 K, distinctly above the vapour pressure limit of steady-state photoelectron spectroscopy. The photoemission results of the liquid at different temperatures are compared with data from the corresponding amorphous alloy, prepared by vapour condensation at liquid-nitrogen temperature. By increasing the temperature of the disordered phase, the Au 5d band is found to shift continuously toward a lower binding energy, showing a linear behaviour over the whole temperature range. Our measurements confirm earlier results obtained with conventional photoemission within a restricted temperature range. The results presented here clearly show the potential of the new technique to investigate the electronic structure of alloys at conditions under which standard photoemission experiments cannot be performed.

The role of the solvent in the liquid phase dynamics of unimolecular processes involving small energy barriers is studied by ultrafast laser spectroscopy. Different regions of the interaction potential between the solvent environment and the reactant are characterized by investigating photoisomerization reactions of trans-stilbene and related compounds in liquid n-alkanes and n-alkanols over a wide pressure range. By varying the physical properties of the solvent environment in such a controlled way, aspects of the reactant - solvent interaction which dominate the different density regimes are identified. In particular, the kind of friction acting along the reaction path, the role of solvent-induced changes of the effective potential energy surface for reaction, and the coupling of dielectric solvent shell relaxation and motion along the reaction coordinate are addressed.

The dynamics of the photodetachment of an electron from a chloride ion in water induced by excitation of the lowest charge-transfer-to-solvent state is explored by using quantum molecular dynamics simulations. The ejected electron is described in terms of floating Gaussian orbitals, and solvent electronic polarization effects are accounted for in a fully self-consistent way. The simulation results point to a two-step photodissociation mechanism: the formation of a metastable electron - atom pair on a subpicosecond time-scale followed on a picosecond time-scale by the competition between two different reaction channels, (i) a diffusive barrier-impeded dissociation of the pair, yielding a solvated halogen atom and a free electron, and (ii) a non-radiative quantum recombination, eventually leading to the chloride ion in its ground state. The computed transient absorption spectra are compatible with the experimental data either at early times during the formation of the electron - atom pair or at longer times when dissociation - recombination occurs.

Bi-chained surfactants, e.g. sodium dialkylbenzenesulphonates, can spontaneously form vesicles when salts (e.g. NaCl) are added to water. Critical vesicle concentrations can be readily determined. The kinetics and mechanism for the breakdown and formation of vesicles will be discussed. A mechanism for assembly/disassembly is proposed. Organic dyes can be encapsulated inside the vesicles and their release rates can be monitored using flow experiments. It is found that the vesicle bilayer provides a rather low energy barrier to the transport of the dye from the vesicle core to the external aqueous medium.

The application of ab initio molecular dynamics simulation to bulk molecular systems is reviewed with emphasis on the density functional treatment of intermolecular interactions. Examples discussed are water, hydrogen fluoride, benzene, and sulphuric acid.

Molecular dynamics is applied to the title problem. It is shown how the methodology for solution reactions easily allows for a detailed study of the interchange, which has a rather complex character with unexpected dynamical features. The methodology can be applied with equal ease to cases with very different lifetimes including those with strongly bound shells.

Systematic high-resolution neutron inelastic scattering measurements have been made of the collective phonon - roton excitations in normal and superfluid as a function of both pressure (density) and temperature. The IN6 time-of-flight spectrometer at the Institut Laue - Langevin has been used to obtain accurate and continuous data over a wavevector range with a resolution of 0.12 meV. We present the experimental data and discuss the energies and widths of the single excitations; changes concomitant with the transition are discussed. The structure of the multiphonon continuum at pressure is also discussed and is shown to be consistent with theory.

This paper gives a brief outline of the theory and practice of kinetic energy measurements made on the electronvolt spectrometer at the ISIS spallation neutron source. The usefulness of the technique is illustrated by measurements made on as a function of temperature and density, on mixtures as a function of concentration and on liquid neon. Future developments and applications are discussed.

The electronic structure factors for liquid light methanol and heavy methanol have been measured at under their normal vapour pressures using a -ray beam from an radio-isotope source. Observable differences between the electronic structure factors of the two liquids reveal small, but real changes in their r-space correlation functions. This is due to quantum effects in the intermolecular and intramolecular electron density correlation functions and will be discussed in terms of the structures of light and heavy liquid methanol.

Computer simulations, using the molecular dynamics and Monte Carlo techniques, and employing simple molecular models, yield insight into general features of phase equilibria, structure, and dynamics of liquid crystals. Here, results are reported from extensive simulations of the Gay - Berne family of molecular models, in which potential parameters are adjusted to vary the molecular length-to-width ratio in a systematic way. Attention is paid to the characterization of nematic, smectic-A and smectic-B phases as functions of these parameters.

A simulation study of the approach to the isotropic - nematic phase transition, using a large system size and lengthy runs on the T3D parallel supercomputer, is described. Spatially long-ranged collective orientational correlations develop in the isotropic phase, close to the transition. The direct correlation function has been calculated for these systems, and remains short-ranged, as expected, as the transition is approached. The simulation results are compared with the density functional analysis of isotropic instability relative to the nematic phase.

We outline the elements of a mean-field density functional theory of inhomogeneous liquid crystals which is able to account for surface-enhanced smectic ordering (SESO) at a free surface. The theory generates SESO without requiring an external anchoring potential, and depending only on the properties (i.e., strengths and ranges) of the anisotropic intermolecular forces. Application of the theory to explaining recent experimental findings of layer-thinning transitions in freely suspended smectic-A films is briefly summarized.

A model suitable for simulating lyotropic polymer liquid crystals (PLCs) is described. By varying the persistence length between infinity and 25, the effect of increasing flexibility on the nematic - smectic transition of a PLC with a length-to-width ratio L/D = 6 is investigated. It is found that increasing flexibility shifts the formation of a smectic phase to higher densities. Comparison is made with a recent theory of the nematic - smectic transition of slightly flexible rods.

The isotropic - nematic (I - N) phase transition in dispersions of sterically stabilized rod-like boehmite (AlOOH) colloids is studied. We have examined the influence of the steric stabilizer, the dispersion medium and the presence of non-adsorbing polymer on the phase transition process. Dispersions in cyclohexane show an I - N phase separation that proceeds by a slow sedimentation of a pinned structure, shrinking from the meniscus, finally forming the nematic phase after weeks or months, depending on the steric stabilizer used. In toluene the onset of the I - N has shifted to higher volume fractions where individual nematic droplets grow and sediment, forming the nematic phase after one week. By adding non-adsorbing polymer to dispersions in cyclohexane the onset of the I - N phase separation shifts to lower colloid volume fractions. At polymer concentrations just above the phase boundary the same scenario as in the toluene dispersions without added non-adsorbing polymer is observed. At slightly higher polymer concentrations an abrupt change in scenario occurs. Now an interconnected network is formed which starts to sediment, resembling the process in pure cyclohexane dispersions. Clearly small variations in colloidal interactions engendered by the changes in dispersion characteristics considered in this work, have a strong influence on the I - N phase transition process.

Small-angle neutron scattering data from dense (30% by mass) gelling 7 nm colloidal silica spheres are presented. The coarsening process that occurs during gelation exhibits temporal self-similarity, and the time-dependent structure factor obeys the dynamic scaling relation . Here, q is the scattered wavevector, is the location of the low-angle peak in is a time-independent characteristic structure function which has a maximum at x = 1, and is the fractal dimension. Connections between the silica gelation and spinodal decomposition in a simple fluid are reviewed.

We present simulations of the basic model of weakly charged polyelectrolyte chains of variable intrinsic stiffness within the Debye - Hückel approximation. For intrinsically flexible chains the persistence length shows a sublinear dependence on the screening length in strong contrast to all known analytical approaches which propose an effective exponent of either y = 2 or y = 1. The observed exponent y varies as a function of the system parameters. With increasing intrinsic stiffness the corresponding effective exponent y crosses over to values of up to 2 when the intrinsic persistence length exceeds the electrostatic one . We find a pronounced minimum of with increasing intrinsic stiffness due to a reduction of entropy. The concept of a unique persistence length is not applicable for these systems.

By linearizing the electro-hydrodynamic equations and using general arguments, we have recently described the deformation and drift of a polyelectrolyte in solution under the simultaneous action of an electric field and a non-electric force, and obtained results qualitatively different from previous theories. We show here how one can adapt the Zimm model to obtain a more operational description for such problems, which allows us to recover our previous results in a simple way and could be used to describe more general situations such as transient phenomena or the electrophoresis of a polyampholyte.

We report a study of the dynamical behaviour of a polystyrene latex sphere in a telechelic poly(ethylene oxide) solution using optical tweezers. With this new technique, we use a position-sensing detector and a lock-in amplifier to measure the displacement magnitude and phase responses of one latex sphere driven sinusoidally by optical tweezers.

For a single particle in solution, the equation of motion of the particle is simply that of the forced oscillation problem with damping from viscous drag and the restoring force from the elasticity of the solution medium and that of the optical tweezers. Because the system is overdamped, it is not feasible to probe the high-frequency regime. Thus we cannot measure the viscosity and elastic moduli separately from frequency-dependent measurements alone. At low polymer concentration, measurements of the viscosity have been achieved. We compared the measured viscosity with that obtained with other measurements. Key issues for further development of this technique, such as measuring the elastic modulus, are briefly discussed.

This paper describes an experiment designed to measure surface and hydrodynamic forces between a mercury drop and a flat mica surface immersed in an aqueous medium. An optical interference technique allows measurement of the shape of the mercury drop as well as its distance from the mica, for various conditions of applied potential, applied pressure, and solution conditions. This enables a detailed exploration of the surface forces, particularly double-layer forces, between mercury and mica. A theoretical analysis of drop shape under the influence of surface forces shows that deformation of the drop is a sensitive indicator of the forces, as well as being a very important factor in establishing the overall interaction between the solid and the fluid.

It was earlier shown that capillary measurements may be interpreted in favour of hydrophobic slippage. Here, the possible implications of this phenomenon as regards the dynamic technique for the hydrophobic attractive force measurements are discussed. We demonstrate that under experimental conditions discernible deviations from the Reynolds theory due to slippage may occur. Misuse of the Reynolds theory may lead to overestimation of the hydrophobic attractive force. This apparent `extra attraction' depends critically on the driving speed, and the types of force present in the system.

By studying the time-dependence of the depolarized field scattered in the forward direction (zero angle), we have measured the rotational diffusion coefficient of hard-sphere colloidal particles. The results are compared to those we previously obtained by using depolarized dynamic light scattering at finite scattering angle. The experimental concentration-dependence of the rotational diffusion constant is discussed in relation to existing theoretical results and numerical simulations.

We have performed optical transmission and synchrotron small-angle x-ray scattering (SAXS) experiments on colloidal crystals with optical refractive index ratios as large as possible over a wide range of volume fractions. These conditions push colloidal crystals into the regime where strong coupling of photonic crystals with light occurs. The optical transmission spectra reveal minima corresponding to stop gaps on the edges of the Brillouin zone of the photonic band structures. The positions of the optically measured stop gaps agree well with lattice spacings measured by SAXS. The stop gap in the 111 direction of crystals of polystyrene in water has a width of up to 5% of the gap frequency as a function of volume fraction, in agreement with theoretical band-structure calculations. A maximum of the relative width confirms the notion that the strength of the interaction between photonic crystals and light has an optimum as a function of volume fraction. The detailed structural information from SAXS data greatly assists in the interpretation of optical experiments on photonic crystals.

The rheological properties of colloidal suspensions of spheres and rods have been studied using dissipative particle dynamics (DPD). We have measured the viscosity as a function of shear rate and volume fraction of the suspended particles. The viscosity of a 30 vol% suspension of spheres displays characteristic shear-thinning behaviour as a function of shear rate. The values for the low- and high-shear viscosity are in good agreement with experimental data. For higher particulate densities, good results are obtained for the high-shear viscosity, although the viscosity at low shear rates shows a dependence on the size of the suspended spheres. Dilute suspensions of rods show an intrinsic viscosity which is in excellent agreement with theoretical results. For concentrated rod suspensions, the viscosity increases with the third power of the volume fraction. We find the same scaling behaviour as Doi and Edwards for the semidilute regime, although the explanation is unclear. The DPD simulation technique therefore emerges as a useful tool for studying the rheology of particulate suspensions.

A series of experiments has been performed in order to analyse the shear-induced structures for a micellar cubic phase using small-angle neutron and x-ray scattering techniques. Steady shear was applied in a Couette cell to the (EO) triblock copolymer system dissolved in water. At rest, the system crystallizes into a long-range ordered mesophase of face-centred symmetry (lattice parameter ). The good resolution of the x-ray technique enables us to study in detail the transition between shearing flows dominated by oriented textures at low shear rates and flows mediated by the mechanisms of layer sliding at higher rates .

Some recent progress in the study of liquid foams is reviewed in outline. Calculations of foam conductivity are presented, which further improve upon the approximation recently proposed by us, in accounting for the nonlinearity in the dependence of conductivity on liquid fraction.

We present an entirely new experimental device dedicated to the study of the mechanical behaviour of freely suspended soap films. We find that the measured surface viscosity is comparable to the surface viscosity of Gibbs monolayers, but surprisingly, we found that soap films have a small shear modulus.

When an oil-continuous dispersion is frozen two microdomains are formed: one, the pure oil solvent which is selectively solidified (I), the other, a concentrated `liquid' dispersion of particles in oil (II). These two domains are intimately mixed within the frozen colloid and exist in a state of equilibrium determined by the system pressure and temperature. The position of equilibrium controls the proportion of the solvent which is solidified, and thereby the concentration of particles within the fluid microdomains (II). Combined with SANS measurements, to determine the inter-particle separation in these microdomains, an analysis based on osmotic pressure provides a measure of the inter-particle repulsion forces presented by the surfactant layers.

The linear viscoelastic behaviour of molecularly thin OMCTS (octamethylcyclotetrasiloxane) films has been studied as a function of film thickness using a modified surface forces apparatus. The frequency spectra of viscoelastic relaxation measured at different film thicknesses superposed onto a master curve when each spectrum was shifted horizontally and vertically with respect to the frequency spectrum at a reference film thickness. The master curve indicates a wide separation between two sets of viscoelastic relaxations and suggests a gradual glass-like transition to solidity.

The free liquid surface of the binary system has been studied by reflection ellipsometry in the temperature range below and near the upper miscibility gap. The ellipticity of samples of concentrations below and above the critical composition was measured as a function of temperature T and compared with calculations of based on model refractive-index profiles. The pronounced increase of with T at concentrations is consistent with a model profile that entails there being a water-enriched layer just below the surface. The thickness of this water-rich layer increases as one approaches the phase-separation temperature, as is to be expected for preferential wetting of the surface of the concentrated micellar solution by the coexistent dilute micellar solution.

Reflectivity experiments on fluid mercury against an optically transparent sapphire window at high temperatures and high pressures close to the liquid - vapour critical point reveal clearly the existence of a prewetting transition of mercury on the sapphire substrate. The prewetting line intersects the coexistence curve at the wetting temperature , and terminates at the prewetting critical temperature and prewetting critical pressure lying close to the bulk critical point.

It is shown that the coupling between capillary waves and dilatational fluctuations at a liquid surface can lead to the appearance of mixed modes. Calculations are presented for various cases, and related to recent experiments.

The buoyancy-driven boundary layer convection on a heated vertical plate was investigated in critical xenon employing phase-contrast microscopy. In a boundary layer of about thickness vapour tubes with a diameter of the same size and up to several mm in length are observed to form a pattern with a period . The pattern formation is restricted to the fluid temperature interval and to the region below the position of the meniscus. The vapour tube formation is attributed to spinodal decomposition due to a local temperature lowering of the boundary layer relative to the bulk fluid.

A stability analysis of the contact line at the bottom of vapour stems is undertaken in order to find out the dominant parameters responsible for the transition from nucleate boiling to film boiling. For strong constraints, the increase in the evaporation rate depletes the macrolayer and, as consequence, there is an enlargement of the dry areas. The second step of the boiling crisis study is to establish the relationship between the macrolayer depletion and the vapour column instability. The contact angle dynamics is very crucial to the occurrence of the crisis. When the macrolayer becomes unstable, the columns are cut off and consequently break down. We expect that before the macrolayer has been completely consumed the relative speed will have reached the critical value at which the Kelvin - Helmholtz instability of the vapour columns appears. It is clearly demonstrated that many kinds of instability participate in the boiling crisis.

We explore the role of disorder-induced localization in the dynamics of glasses and supercooled liquids using instantaneous normal mode (INM) analysis. This study is motivated by the fact that such localized excitations (tunnelling states and soft harmonic vibrations) are believed to be important in the thermodynamics and dynamics of amorphous systems at very low temperatures. The results are presented for two simple model systems that show the existence of a temperature below which all unstable INMs become localized. The relationship of this temperature to the glass transition is discussed.

Brillouin and Raman experiments have been performed in hydrogenated silica sol - gel and vitreous silica samples made by sol - gel processes (xerogels) and by the melting powder method. The aim of this work has been the study of the low-frequency dynamics of sol - gel systems in comparison with that of amorphous silica. By thermal treatment of the gel a progressive densification of the system was obtained: at we obtained densification (temperature well below the usual temperatures); at higher temperature () the pores were eliminated and the dynamics (Brillouin shifts, quasielastic scattering, and Boson peaks) became equivalent to that of fused silica.

The - and -relaxations in a disparate-size binary liquid near the glass transition have been investigated within a mode-coupling theory. We focus our attention on the - and -peaks in the frequency-dependent susceptibility and their dependence on the concentrations of big and small particles. For the 1:1 mixture of big and small particles (size ratio = 0.2), a large peak appears in the susceptibility of the small particles in the frequency range of the -relaxation, which corresponds to the fast relaxation of the small particles within the random potential produced by the big particles. The intensity of this peak grows further as the concentration of the small particles is decreased. For large , on the other hand, the peak becomes lower than the -peak and the susceptibility is similar to that of a one-component liquid.

An analysis of light scattering spectra of different glass-forming liquids is presented. It is shown that at high temperatures one finds a scenario for the temperature variations of dynamic spectra at high frequencies which is common for strong and fragile liquids. This scenario breaks down at some crossover temperature . One conclusion is that the difference between strong and fragile glass-forming liquids is the temperature range between and : it is small for fragile and large for strong glass-forming systems.

We describe recent light scattering measurements performed on two small molecules, metatoluidine and salol, and one polymer, 1-4 cis - trans polybutadiene, in their undercooled phase. For the small molecules, the -relaxation times deduced from the isotropic and anisotropic spectra are different and we propose, through different experiments, that the latter spectra are related to the reorientation motion of the molecules, the dynamics of which follows some of the MCT predictions. For polybutadiene, a broad and rather flat spectrum dominates the anisotropic spectrum, while the damping of the acoustic phonons is possibly due to their coupling to intra-chain conformation motions.

This paper briefly reviews recent results of extensive Monte Carlo simulations for the glass transition of polymer melts. The simulation used the bond-fluctuation model, a lattice model, which exhibits glassy behaviour due to the development of a competition between packing constraints and chain stiffening at low temperatures. For this model static and dynamic properties were analysed, such as the influence of the cooling rate and of the chain length on the glass transition temperature, physical aging effects, the time-dependences of various mean-square displacements and structural relaxation functions and the temperature-dependences of structural relaxation times and of the diffusion coefficient. Besides an outline of these results we discuss in some detail a quantitative comparison between the incoherent intermediate scattering function and the extended mode-coupling theory and between the entropy of the melt and the Gibbs - Di Marzio theory.

The reorientation of a nearly spherical paramagnetic tracer dissolved in supercooled o-terphenyl is studied by electron spin resonance in the range 200 K < T < 370 K. For T > 300 K the rotational correlation time follows the Debye - Stokes - Einstein law. On decreasing the temperature, the rotational motion of the probe and the viscosity decouple. In particular, in the region around T = 290 K it is found that with .

Silica aerogel is a highly porous glass consisting of a tenuous network of strands interconnected at random sites. When a mixture is placed inside the aerogel, the coexistence region is found to be detached from the superfluid transition line, giving rise to a new miscible superfluid mixture at high concentration and low temperatures.

We summarize the results of a recent positron annihilation study of the phase behaviour of confined in Vycor glass. Particular emphasis is placed on the phase diagram of the confined fluid and on the usefulness of the positron annihilation technique in determining the mechanisms underlying phase transitions of fluids in porous solids.

We present the calculation of phase diagrams for a single-component fluid adsorbed in disordered porous material using integral equation theories. The model consists of a Lennard-Jones 12 - 6 fluid confined in a rigid matrix of spheres. In most cases a vapour - liquid coexistence curve is obtained. It is similar to that observed for the bulk fluid although displaced towards the phase that adsorbs preferentially. The approximation also predicts the appearance of a second fluid - fluid phase transition at low temperature that may be consistent with the narrowing of the coexistence curve observed in experiments.

We have established the solid - fluid coexistence region for a system of polydisperse hard spheres with near-Gaussian diameter distributions, as a function of polydispersity. Significantly, we observe a `terminal' polydispersity above which there can be no fluid - solid coexistence. At the terminus the polydispersity is 5.7% for the solid and 11.8% for the fluid while the volume fractions are 0.588 and 0.547 for the solid and fluid respectively. Substantial fractionation observed at high values of the polydispersity (> 5%) implies that the `constrained eutectic' assumption made in previous theoretical studies is not generally valid. Our results for the terminal polydispersity are consistent with experiments performed on polydisperse colloidal suspensions.

In analogy with the well-known Van der Waals theory for disordered fluid phases we propose a simple analytic expression for the Helmholtz free energy of an ordered crystalline phase. It is based on a free-volume approximation to the entropic contribution due to the hard-core repulsions and a static lattice energy approximation to the contribution of the attractions. In this way we are able to describe in a simple way phase transitions between two disordered phases, between two ordered phases and between an ordered and a disordered phase in systems of any spatial dimensionality.

We consider the application of finite-size scaling methods to simulations of tricritical phenomena in a two-dimensional symmetrical binary fluid. A simulation strategy is described which together with the scaling framework enables the accurate determination of both the universal and non-universal tricritical point properties of the model. The results also provide insight into the character of the tricritical fluctuations.

The vapour - liquid equilibrium of short n-alkanes is considered by using perturbation theory. This requires as a previous step obtaining an equation of state (EOS) for hard flexible models. An EOS for hard-n-alkane models which shows excellent agreement with computer simulation of hard-n-alkane models with up to 100 carbon atoms is proposed. This EOS is combined with a mean-field term and the vapour - liquid equilibrium of n-alkanes is computed. When theoretical critical densities and pressures are plotted as a function of the number of carbon atoms, a maximum is found. This is in agreement with experiment. The conditions for the appearance of such a maximum in general chain models are established. Some ways of improving the mean-field theory are suggested.

We examine the effect of dipolar interactions on the liquid-crystalline phase behaviour of L/D = 5 hard spherocylinders with a terminal point dipole. The hard spherocylinder consists of a cylinder of length L and diameter D with hemispherical caps on each end; the point dipole is located at the centre of the hemispherical cap (2.5D from the centre of the spherocylinder) and is oriented along the principal molecular axis. The phase transitions exhibited by this system are studied using the isothermal - isobaric Monte Carlo (MC-NPT) technique. As for systems with central dipoles, the terminal dipole is seen to slightly destabilize the nematic (orientationally ordered) phase relative to the isotropic phase when compared with the non-polar hard spherocylinders. More interestingly, the smectic (layered) phase is destabilized in the terminal dipole case, and is only seen at the very highest densities. This is in stark contrast to what is seen for systems with central point dipoles in which the smectic phase is stabilized relative to the nematic phase due to the strong anti-parallel dipolar interactions. We do not find any evidence for ferroelectric or anti-ferroelectric ordering in these systems.

Density functional approaches have become well known in applications of equilibrium statistical mechanics to phase diagrams, interfacial structures, and free energies. This paper describes how these methods can be extended to the dynamics of phase transitions. Two aspects are emphasized: the calculation of free-energy barriers in nucleation and of growth velocities for crystallization from the melt. Significant deviations from classical nucleation theory and from diffusion-based theories of crystal growth are found. The continuum density functional approach is effective in describing the dynamics of collective-mode fluctuations, which dominate single-particle dynamics in these cases. The role of body-centred ordering at the interface between a face-centred crystal and the melt is described.

A new model is proposed for the study of porous media and complex fluids using morphological measures to describe homogeneous spatial domains of the constituents. Under rather natural assumptions a general expression for the Hamiltonian can be given extending the model of Widom and Rowlinson for penetrable spheres. The Hamiltonian includes energy contributions related to the volume, surface area, mean curvature, and the Euler characteristic of the configuration generated by overlapping sets of arbitrary shapes. Phase diagrams of the model are calculated and discussed. In particular, we find that the Euler characteristic in the Hamiltonian stabilizes a highly connected bicontinuous structure, resembling the middle phase in oil - water microemulsions.

We present a novel molecular dynamics simulation technique, which accounts for both two- and three-body dispersion interactions. This technique is a unified approach of molecular dynamics and quantum mechanical variational methods, in the spirit of the Car - Parrinello method (1985 Phys. Rev. Lett. 55 2471). We use a highly simplified model for the electronic structure of the atoms, which is, nevertheless, sufficient to correctly reproduce the London two-body, and the Axilrod - Teller three-body dispersion forces in an appropriate limit. The advantage of this new method is that it allows for a consistent treatment of both dispersion damping and periodic boundary conditions at the pair and three-body levels.