# Self-Assembled Porous Media

Principal Investigator: Jens Harting, Department of Applied Physics, Technische Universiteit Eindhoven (The Netherlands)
HPC Platform: JUQUEEN of JSC
JSC Project ID: compflu1

A porous medium is a material characterized by the presence of holes, or “pores“. These pores are usually surrounded by a solid and can be filled with a gas or a liquid. In nature one can find many examples of porous media, such as many types of rock (e.g. in oil reservoirs or aquifers). Many types of man-made porous media exist too, as they can be designed to have very useful properties. For example, porous media form the basis of many types of reactors, filters, or fuel cells.

Scientists of the Eindhoven University of Technology and the Institute for Computational Physics of the University of Stuttgart studied the possibility of using the emulsification process of particle-stabilized emulsions to create self-assembled porous media and it was shown that this type of self-assembly can result in highly tuneable systems, which can be optimized for various purposes. To this end, the researchers employed the lattice Boltzmann algorithm to simulate the fluid phases, coupled to a molecular dynamics solver to treat the presence of solid particles. These particles generally diffuse towards fluid-fluid interfaces and stabilize those by reducing the interfacial free energy–a process commonly used for example in the food industry. In view of the different length scales involved in the problem, very large three-dimensional lattices are required to attain the necessary resolution; these can then only be treated by a massively parallel implementation of the algorithms involved. The researchers apply a simulation code called lb3d’’ which was developed at University College London, the University of Stuttgart and Eindhoven University of Technology over the last 12 years. This implementation has demonstrated excellent scaling behaviour on several available supercomputing platforms including HERMIT in Stuttgart and JUQUEEN in Jülich and is perfectly suited to attack the scientific problem by harnessing the power only available on Tier-0 HPC resources.

It is of advantage to be able to create a catalyst or filter directly at a location that is hard to access and where fine-grained control over an assembly process is necessarily absent, e.g. in underground pipes. In a two-phase flow with suspended particles, one finds all ingredients for making an emulsion. If one of the fluid phases contains an additional component that can be triggered in some fashion to solidify, a geometry can be created in a remote location by supplying the ingredients (and trigger) from far away. As shown by the researchers, this can also be achieved by modifying the stabilizing particles to be able to adhere to each other after a trigger, such that the covered interfacial area transforms into a particle scaffold. In this way a different type of structure is created, with greatly enhanced porosity. The scientists explored the parameter space that governs the development of particle-stabilized emulsions for regions that result in geometries with desirable properties, such as large permeability yet small domain sizes (filters), or large surface-to-volume ratio, porosity, and large permeability (reactors). A particularly striking result is that by changing the contact angle a particle forms with it surrounding fluids alone, the permeability of the resulting porous medium can be varied by at least two orders of magnitude. Scientific journal articles summarizing the obtained results are currently being written up and expected to appear soon.

The techniques described in this chapter can be combined with the study of the effects of ions in oil-water mixtures on the flow through porous media, which is of particular interest when improving enhanced oil recovery techniques such as brine injection.

A particle-stabilized emulsion is used as a basis for creating a self-assembled porous medium. For clarity, only part of the sample is represented by a 3D volume, the rest of the image shows planar cross-sections: blue represents pore space while red represents the solid.

Scientific Contact:

Prof. Dr. Jens Harting
Department of Applied Physics
Technische Universiteit Eindhoven
P.O. Box 513, NL-5600 MB EINDHOVEN/The Netherlands
e-mail: j.d.r.harting@tue.nl

July 2014