Predicting the large scale properties of materials based on their chemical composition


The “Holy Grail” of materials science from a theoretical standpoint is to be able to predict the properties of a substance based on its atomic, molecular and chemical composition. Although collectively scientists are firmly committed to the notion that the macroscopic, engineering properties of matter are ultimately determined by their atomistic make up, it may come as a surprise to learn that relating properties such as strength, toughness, density and, indeed, melting and boiling points to the molecular composition of matter remains out of reach today. Indeed, it often takes many decades from the discovery of a new material to its use in the real world. Well known examples of this kind include carbon fibre based materials, light emitting diodes, high temperature superconductors, and carbon based nanowires. Many materials are discovered somewhat serendipitously during experimental investigations. When interesting properties are uncovered, following characterisation of the substance, extensive laboratory work is performed to try to develop a material which can be used in practical applications. There has been very little reliance on theory and computer based modelling and simulation. This is set to change in the future.

Our work is based on computational study of composites made up of clays and polymers. Clay minerals have been widely used since antiquity to construct materials such as ceramics and cements. Today, clays are also an essential constituent of paints, paper, rubber and cosmetics. Their widespread use is due in part to their global abundance: they are inexpensive and environmentally friendly. Clays naturally occur as stacks of thin aluminosilicate sheets, with lateral dimensions of about one micron, stacked like a rather disordered pack of cards, with anything between a few sheets and one thousand in any single pile; the separation between the sheets is of the order of a few nanometres.




Since the late 1980s, clays have been combined with synthetic polymers (such as nylon) to produce composite materials with superior thermo-chemical properties. Small amounts of clay (less than 5% by volume fraction) produce composites with substantially enhanced strength, stiffness, low density, fire retardency and barrier properties than the polymer with which they are combined. These properties make such clay-polymer nanocomposites attractive, among other things, for the design of automotive and aerospace components. However, the effectiveness of these composites depends on being able to coax the individual platelets of clay to organise in particular ways within the polymer. Because of our current inability to control these processes, the original promise of such composites has yet to be fully realised. One of these processes, called exfoliation, causes the constituent sheets—which are of order 1nm thick, similar to the separation between the clay layers prior to mixing—to be dispersed uniformly in the polymer matrix. Incomplete exfoliation can significantly reduce the properties from their theoretical maximum. For other applications, in which the clay volume fraction is much higher, one may be interested in the polymers intercalating and encapsulating the clay tactoids, to produce a bricks-and-mortar structure, mimicking the great strength, toughness and durability of naturally occurring materials such as mother of pearl and nacre.

Until now, the search for such diverse materials has proceeded by way of pure trial-and-error, laboratory based experimental approaches. Ideally, one would like to determine the most effective clay-polymer composite for a given application based on chemistry and processing conditions (i.e, how the two ingredients are physically combined, e.g. by melting the polymer, applying shear or other mechanical forces) so as to optimise property enhancement. If realised, this would lead to new high performance, high value nanocomposite materials. James Suter, Derek Groen and Peter Coveney have now published a major article in Advanced Materials in which they report their use of multiscale computer based modelling and simulation to examine how the chemistry of the clay and polymer at electronic, atomic and molecular scales affects the materials properties at much longer length and time scales. Their approach in which they have effectively built a virtual materials laboratory, opens up a route to computing the properties of complex materials based on knowledge of their chemical composition, molecular structure and processing conditions.




The simulation of layered nanomaterials, such as clay polymer nanocomposites is challenging; the microscopic structure and mechanisms operate over many different length scales, ranging from nanometers to microns. One of the key challenges addressed in the paper is how to efficiently sample these scales to understand how the microscopic structure affects the macroscopic properties of the composite. In their work, the multiscale simulation scheme was used to scan across length scales from electrons, through atoms to slabs of material many microns in extent, comprising many millions of atoms. This scheme ranges from the electronic structure level (to capture the polymer - clay interactions, especially those of the reactive clay edges) through classical atomistic molecular dynamics to coarse-grained dynamical models (to capture the long length scale structure and mechanical properties). Their multiscale simulation techniques give a quantitative picture of how the chemistry of the system affects how easily the clay layers can be separated, and how these layers further aggregate into larger structures. The authors compare various characteristics of the virtual nanocomposites they “synthesised” with experiments, including clay-layer spacings, the relative arrangements of polymers, clay sheets and tactoids, and the elasticity of these materials when stretched or compressed. They find properties very similar to those observed in the laboratory, but now with much finer control of the starting chemistry and ensuing properties in combination with specific processing conditions.

The authors believe that their research findings should set a precedent for future work, leading to increased innovation in high performance nanocomposites, as well as many other types of material. In the future, the authors expect to see a lot more focus on initial design of materials in the virtual world, using similar multiscale modelling and simulation methods, accelerating the search for and convergence on real world materials in combination with experimental work.

James Suter, Derek Groen, Peter Coveney, “Chemically specific multiscale modeling of clay-polymer nanocomposites reveals intercalation dynamics, tactoid self-assembly and emergent materials properties”; Advanced Materials, 9 December 2014. The graphic symbolises how computer-based simulations on multiple scales, from the electronic to the coarse-grained, have allowed the authors to predict materials properties of these complex systems in a systematic manner—a capability that should accelerate the discovery of real world materials.