From micro to macro

Insights on the formation of particle networks hold potential for engineering new and improved materials.

Confocal micrograph of fluorescently labeled polymeric colloids forming a model colloidal gel

Confocal micrograph of fluorescently labeled polymeric colloids forming a model colloidal gel

Image courtesy of the researchers

From the perspective of a chemical engineer, particulate gels are the stuff of modern life. These materials, in which small pieces of one kind of substance are suspended or distributed within another, can be found in such construction products as concrete, inks, and paints; foods like cheese, yogurt, and ice cream; and in a range of cosmetic- and health-related staples including shampoo, toothpaste, and vaccines. In sum, says James W. Swan, the Texaco-Mangelsdorf Career Development Professor in Chemical Engineering, “a massive variety of real-world, everyday things bear particles.”

Many of these ubiquitous gels, creams, emulsions, and compounds evolved through “trial-and-error experimentation,” says Swan. Engineering such materials often proves to be a prolonged and sometimes inefficient hit-or-miss process.

But now Swan and collaborators from other universities have devised a framework that will help guide the design of new materials involving such particulate compounds. An account of their research, which began in 2015, appears in the May 20 issue of Nature Communications. The experimental studies were conducted at the University of Delaware and the University of Michigan. Swan and MIT doctoral student Zsigmond Varga (now a process development engineer at ExxonMobil) were responsible for the computational side of the work.

Combining laboratory experiments and computational simulations, the team has analyzed the formation of networks of particles that determine the microstructure of a wide range of materials. This enables the researchers to predict the macroscopic mechanical properties conferred by these networks. Their new approach will make it possible to “seek out new materials or engineer systems better optimized for tasks in terms of properties or costs, for a lot of different technologies,” says Swan.

The key is elasticity

Predicting the elasticity of materials – how soft or hard they will be – has proved a longstanding challenge in the area of chemical engineering that deals with particle networks. This property, called the elastic modulus, is central to designing new things. But, says Swan, “We have had no equations to make predictions, and these equations would be really helpful in creating new formulations.”

To develop such useful mathematical models, though, the science required a platform of informative experimental data that could fill in “fundamental gaps in understanding how networks of particles are built, and where mechanical properties come from,” explains Swan. So his colleagues set out to conduct these essential studies.

In order to investigate the formation and properties of particle networks, Swan’s collaborators devised a set of nifty methods. They created particles that were effectively translucent except when illuminated by a special kind of fluorescent light. This permitted the researchers to image particle networks in real time under a microscope. They also fashioned the particles so they could be manipulated by laser tweezer – a device that can exert forces on small particles. With their novel tools, researchers could directly measure the force each particle exerted on another. “My colleagues could observe the particles’ motion, how they stuck together, and gained insight into the nature of the particle network’s elasticity,” says Swan.

These very precise experimental techniques, says Swan, “made it possible for our lab to build simulations that quantitatively reproduced the experimental results.” But Swan also faced challenges: The distribution of particles in a network has often seemed disordered to scientists – as if a distracted builder had laid bricks haphazardly in mortar. “It’s as if there are no discernible patterns,” says Swan.

His lab’s unique solution was to apply methods from graph theory, a field of mathematics often used to understand computer or social networks, to deconstruct and model the discrete elements and connections among particles in chemical networks.

By employing “the same tools useful for finding cliques in social networks – tightly connected groups loosely connected to each other – we are able to discriminate the bricks from the mortar in these particle networks, see how the particles are positioned relative to each other, and infer the strength or weakness – the elasticity – of the overall network.”

From networks to new materials

Two scientists not involved in the studies behind the journal article find the work extremely promising.

“This task of discerning connections within a huge collection of particles has historically been extremely puzzling and challenging,” says Thibaut Divoux, a research scientist with a joint appointment at MIT and France’s National Center for Scientific Research. “The use of graph theory to identify clusters, verify experimental results, and to predict properties is really elegant, as well as groundbreaking,” he says.

Randy H. Ewoldt SM ’06, PhD ’09, an associate professor of mechanical science and engineering at University of Illinois at Urbana-Champaign, found “the agreement between simulation and experiment impressive.” He believes the work of Swan and his collaborators “represents excellent progress toward the goal of predicting and engineering properties of these materials.”

There might be myriad applications of this research to improve existing particle-based products and to formulate new ones, suggests Swan. Take the abundant substance of concrete. Since “we now know how changing different chemical or physical factors in these networks will affect properties,” he says, concrete’s particle network – comprised of aggregate rock and cement – could be engineered for enhanced strength while using less material.

There is potential as well for advancing grid-scale energy storage, where particle networks are deployed as electrodes for flow batteries capturing energy from wind or solar power; and for developing new pharmaceuticals such as protein-based solutions used in drug delivery.

And some scientists hungrily envision applications in food manufacturing: “Imagine designing yogurt, cheese, or other dairy colloidal products to determine the mouthfeel, so that the product crumbles or breaks just the way you want,” says Divoux. “This work gives us an experimental key we can use to control the microstructure properties of a material to tweak its macrostructure.”

Funding for the experimental portion of this research came from the International Fine Particle Research Institute, and funding for the computational work was made possible by the American Chemical Society Petroleum Research Fund, and by a National Science Foundation Young Investigator Award.

/University Release. View in full here.