Foreword

The idea of stabilizing liquid domains with interfacially trapped particles dates back at least as far as the work of Ramsden and Pickering in the early 1900s. For most of the intervening century, attention was focussed almost exclusively on droplet structures – often called Pickering emulsions. It has long been known that the presence of a coating of colloidal particles can protect these droplets against coalescence and also, in part, against Ostwald ripening – the two main sources of instability in conventional, surfactant-stabilized emulsion systems.

For micrometre-sized colloidal particles (but not nanoparticles) at the interface between typical immiscible fluids, detachment energies are vastly larger than the thermal energy k_BT. This means that the interfacial structure is far from equilibrium, unlike the case with surfactants, which can adsorb and desorb to maintain a fixed surface density despite changes in area. This irreversibility is simultaneously the cause of the strong stability, and the origin of significant complexity, in Pickering emulsion formulation: since details of the processing history (such as previous changes in droplet shape) are remembered by the interfacial structure, one needs close control of the process to get reproducible results. In consequence, for many years, Pickering emulsions were considered too wayward for widespread use in formulating high-functionality products.

This started to change in the 1990s and early 2000s, when experiments in many laboratories showed that careful control of process history could lead, after all, to predictable material properties. Part of the background to this progress was an improved general understanding of arrested soft matter systems, such as colloidal gels, whose nonequilibrium nature has – thanks to many advances in basic science – gradually shifted from ‘bug’ to ‘feature’ from a formulation perspective. Gels come in many forms, but one particular type, in which colloidal particles are sequestered into a foam-like structure by expulsion from growing domains of a liquid crystal, was of particular interest in the Edinburgh Soft Matter Physics Group in the early 2000s. The idea emerged, first as a pure thought experiment, of starting with colloids suspended in a single-phase, binary solvent, and then demixing the solvent. Particles would perhaps be swept up by the coarsening interface between domains, and be trapped there, causing structural arrest once they reached a close-packed, jammed layer. Rather than the foam that had separated liquid-crystal droplets, the interfacial layer would divide space, templated by the interface between bicontinuous fluid domains that would themselves become locked in place at the point of jamming. The overall structure would be a solid gel, even for purely repulsive colloids, by virtue of the convoluted, spanning monolayer of jammed interfacial particles, held in place by interfacial tension.

A simple and elegant idea – but would it ever work? Rather than pursue this question immediately in the laboratory, we chose to address it first by large-scale computer simulations. This was a good step, because simulations could cut many corners and jump directly to a best-case scenario: two fluids of exactly equal viscosity and volume fractions, with particles of equal affinity for both solvents. These simulations, spearheaded by Kevin Stratford of EPCC (Edinburgh high performance computing centre) indeed showed proof-of-concept for arrest of the spinodal structure under these idealised conditions. (This was far from obvious: for instance, it would be possible for the coarsening interfaces to move too fast and leave the particles behind, as happens in some of the liquid crystal systems.) With the simulations as a roadmap, the experimental team set to work, and before long confirmed that these new materials were indeed realisable in the laboratory. I believe this laboratory work might have taken much longer, or possibly been abandoned altogether, without the confidence from simulations that ‘it ought to be possible’. This is a good example of how simulation and theory can inform the experimental quest for new materials.

Patents were duly filed – initially before the experimental verification – for ‘fluid-bicontinuous particle stabilized gels’ (FBPSGs): an accurate if pedestrian name. A journal editor handling the first simulation paper (Stratford et al., Science, 2005) asked for some more memorable branding, so we came up with ‘bicontinuous interfacially jammed emulsion gels’, whose acronym – with some poetic licence – became ‘Bijels’.

The subsequent development of these scientifically fascinating, and potentially very useful, materials is described in this book. A solid structure with two independently continuous domains of chemically different fluids, each percolating across the entire sample, is not commonplace. Allowing for the interfacial layer, one has, in fact, a tricontinuous material. Such structures can be made by other means, but generally not by simply mixing the ingredients – two fluids and one species of colloid – required to create the three phases. More typically these elements need to be expensively and covalently bonded together, in the form of triblock copolymers, for example.

Simple ingredients do not necessarily make for a simple process, and, alongside its exploration of the underpinning science, much of this volume reports the major steps forward in the past 15 years towards the robust and adaptable creation of bijels themselves, and of bijel-related structures more generally. Some of these steps were speculatively anticipated in our early patent filings, such as the in situ self-assembly of Janus particles, which can impart vastly greater interfacial stability than chemically homogeneous particles, allowing a shift from the colloidal to the nanoscale for the particles themselves. This step also lessens the need to closely approach the neutral wetting condition – which has been achieved with increasing robustness but remains challenging in general.

Similarly, the original idea of demixing a binary fluid mixture spinodally is only one way of imprinting the required bicontinuous structure. It is limited to a narrow range of fluid pairs (albeit with the possibility of replacing these afterwards) and, while it can give arguably the most beautiful and uniform domain structure, routes based on direct mixing have other advantages. The same applies to solvent-transfer demixing, which gives a rich and intriguing set of morphological controls in the case where a quasi-one-dimensional fibre geometry is required. We did not anticipate that approach, nor the combination of bijel-type interfacial stabilization with additive manufacturing methods, creating another powerful suite of techniques to create functional materials.

The repertoire of interfacially stabilized bicontinuous structures is increased further by a variety of post-processing options, which are also described in this volume. While the complete solidification of one of the two fluid domains (e.g., replacing it with a metal) creates a relatively conventional porous medium, albeit with a narrower pore-size distribution than other routes to such structures, bijel templating also allows many other options, such as multi-scale porosity, and for either or both fluid phases to become a cross-linked gel that remains fully permeable to fluid transport. This retains the original concept of fluid bicontinuity while enhancing mechanical integrity. Similar improvements can be achieved via the permanent solidification of the particle layer.

It is a long path from a thought experiment, through computational proof-of-concept, to making things work in the laboratory. As I theorist I have particular admiration for the skill, patience and imagination of all the experimental colleagues, especially the recent contributions of Alexandra Bayles (ETH Zurich), Joe Forth (UCL) and Martin Haase (Utrecht), who have taken this field forward by developing a working set of design principles, and creating an array of exciting new materials at proof-of-concept level. Thanks to these efforts, the field of bijels has progressed from a purely hypothetical suggestion to the threshold of commercial exploitation.

Whether it can now cross that threshold is the next challenge. This may require further work on scale-up – which is certainly improving, but not yet what is needed for bulk chemicals applications – or a focus on bespoke, high-value areas that can make use of relatively small samples (perhaps in film or fibre geometries), several of which are suggested in this book. Achievement of such goals will take the bijel adventure to the next level, allowing the practical utility of these materials to match the scientific excitement of their discovery.

Michael Cates FRSLucasian Professor of Mathematics,

Department of Applied Mathematics and Theoretical Physics,

University of Cambridge, UK