593 Phase separation, aggregation and gelation in colloid-polymer mixtures and related systems Wilson CK Poon The study of phase transitions in model spherical colloids within a 'colloids as (super)atoms' paradigm continues to proceed apace. While interest in quasi-monodisperse systems remains on a high level, much recent attention has also focused on phase transitions in binary mixtures (colloid-polymer and colloid-colloid mixtures), as well as the effect of polydispersity on phase behaviour. At the same time, increasing attention is being paid to the study of metastability and the slow structural evolution ('ageing') of metastable states such as transient particle gels. The equilibrium phase behaviour and phase transition kinetics of protein solutions and other biological systems are increasingly being scrutinised under the light of insights gained from the study of model colloids. Addresses Department of Physics and Astronomy, University of Edinburgh, Mayfield Road, Edinburgh EH9 3J2, Scotland, UK; e-mail: w.poon@ed.ac.uk Current Opinion in Colloid & Interface Science 1998, 3:593-599 Electronic identifier: 1359·0294·003-00593 © Current Chemistry Ltd ISSN 1359·0294 Abbreviations BHS binary hard sphere PEO polyethylene oxide Introduction Colloidal aggregation and gelation was known to Graham [1], who cited these as characteristic phenomena in the col- loidal state of matter. The most recent phase of research began when the 'fractal' concept was applied to colloidal aggregates [2]. The investigation of equilibrium phase sep- aration in colloids also has a long history, many of the early studies being on biological systems (viruses, proteins, etc.) [3]. Partly because of the complexity of these systems, it was unclear for a long time whether the thermodynamic principles enunciated by Gibbs for atomic systems (espe- cially his famous phase rule) were applicable to colloidal phase separation [3]. The resolution of this issue in favour of Gibbs, and the emergence of the 'colloids as (superjaroms' paradigm [4], marked the beginning of the modern study of phase separation in colloids. .Colloidal phase separation research up to 1995 has been reviewed in [5-7]. A subsequent survey of both phase sep- aration and gelation (up to 1996) is also available from [8"]. Here, I focus on work not covered by these reviews, with particular emphasis on research in the past two years, and on uncharged (or highly screened) systems. During this period, the growth of interest in phase transitions in mixtures (colloid plus another species), evident since the early 1990s, shows no sign of abating, although increasing attention is now paid to nonequilibrium states, especially gels. A related, conceptual development is the attempt to place various nonequilibrium phenomena within a quasi- thermodynamic framework. A polydisperse colloid is, of course, the 'ultimate mixture'. Significant advances have been made on the theoretical description of polydisperse phase separation. Finally, a concerted effort has begun to apply the insights obtained from simple colloids to phase separation in complex biological systems; colloidal phase separation studies have come full circle [3]. These topics will be reviewed in detail below. Other themes such as anisotropic particles, purely coulombic systems, and direct determination of interparticle forces can only be touched upon here. Mixtures There is a growing body of literature on binary mixtures of a colloid plus another component, which may be a similar colloid shape (e.g. sphere plus sphere), a different colloid shape (e.g. sphere plus rod), or a different species all together (polymers, surfactants, etc.). When the second component is substantially smaller than the colloid, its effect is often discussed within the framework of 'deple- tion' [9]. For concreteness, consider a nonadsorbing polymer of radius of gyration 'i Polymer segments are depleted from a layer of thickness - r; surrounding each particle. Overlap of the 'depletion layers' from two parti- cles creates more free volume for the polymer and lowers the free energy. On the pair level, the polymer induces an attractive 'depletion potential', Udrpfr), of purely entropic origin and range - 2r g between the particles. For a recent direct determination of Udrpfr), see [10°]. Simple theories on polymers added to hard-sphere col- loids (radius R) suggest that the phase behaviour depends critically on the size ratio Y= r/R [11,12]. These theories were significantly confirmed some time ago by experi- ments using sterically-stabiliscd polymethylmcrhacrylute particles and non-adsorbing polystyrene in decalin [13]. Gas-liquid coexistence and triple coexistence of colloidal gas, liquid and crystal phases were observed in this system only when y> 0.24. Later, Meller and Stavans [14] studied sodium dodeeyl sulfate-stabilized silicone emulsion droplets in aqueous polyethylene oxide (PEa) solution with Y= 0.086 and 0.027; as predicted by theory [11,12], only gas and crystal phases of emulsion droplets were found. Encouragingly, there is reasonable agreement between the y = 0.08 and y = 0.086 phase diagrams in [13] and [14] despite a difference of a factor of three in absolute particle size. More recently, Leal-Calderon et 01. [15°] varied the degree of polyrnerisation (/I) of a linear polymer added to a glycerol-in-oil emulsion (stabilised by sorbitan monooleate).When 1/ > 100, the phase boundary was roughly independent of 1/, consistent with scaling