Despite substantial progress achieved in understanding the microstructural basis of transport and rheology in many classes of complex fluids, including polymers, colloids, glasses, liquidcrystals and others, the flow behavior of one of the most important classes of complex fluids, surfactant solutions, remains mysterious. In particular, it has been known for at least two decades that translucent solutions of thread/rod like surfactant micelles undergo a phase transition to form viscoelastic gels under sufficiently strong shear or extensional flows. Lacking an understanding even of what these structures are, they are simply given the name Flow-Induced Structures, or FIS. While progress has been made towards simulating thread/rod like(nanoscale) micelles at the molecular level, and towards simulating the macroscopic consequences of the presence of FIS, such as banded structures in the shear flow, there is an almost complete lack of theoretical connection between molecular structures and the possibility and conditions under which such nano structures might manifest.
Intellectual Merit. This is a collaborative investigation between two universities to help close gaps in understanding through both experiments and a multi scale set of simulations encompassing length and time scales ranging from atomic (and nano) to continuum. A set of experiments exploring the regimes of transient and novel permanent flow-induced structures, induced by extensional deformation in micro channels, will be carried out. In parallel, the project proposes four different simulation methods. 1) Atomistic Molecular Dynamic Simulations. These can capture the structure and interactions of one or two thread like micelle fragments in a periodic box roughly 10 nm on a side, at the atomic level on timescales of 10 nanoseconds. This is long enough and big enough to determine ionic effects on micellar structure and intermicellar interactions. 2) Coarse Grained (CG) Molecular Dynamics Simulations. Using the Marrink MARTINI model that lumps around four heavy atoms into each bead, a 1000 fold speed up relative to atomistic simulations is attained, reaching nearly to the millisecond time scale, while preserving molecular scale properties through suitably chosen CG potentials. The CG model will allow for the determination of micelle persistence lengths and the stability of thread like micelles as a function of salt concentration. 3) Brownian Dynamics Simulations using pearl necklace micelle model. This model, pioneered by Ryckaert and coworkers, treats the wormlike micelle as a string of beads that can break and fuse end-to-end, and is fast enough to allow for the equilibration of micelle length distributions, with and without flow. We will incorporate into this model the potential for micelle junctions or cross links, and bundling, thereby allowing for the first time a molecular scale simulation of flow induced gel formation. 4) Kinetic Model and Constitutive Equation. We will attempt to draw from the simulations the ingredients necessary to build a kinetic model and, if possible, a full nonlinear constitutive equation for flow of thread like micelles. Through this set of interlocking simulations, each aimed at different length and time scales, complemented by experiments, the investigators have developed a roadmap to bridge between molecular properties and macroscopic flow effects such as flow induced gelation and shear banding.
Broader impacts include a collaboration with scientists at Proctor and Gamble, whose nterest is in understanding, modeling, and controlling the properties of thread like micellar solutions. They plan annual meetings between P&G scientists and our team of graduate and undergraduate students and faculty as well as month long student interships at P&G. This will lead to fruitful exchange of ideas, bringing practical commercial concerns to the attention of students, and carrying novel fundamental ideas and new modeling methods into the corporate world. They plan to also recruit UG (REU) as well as school students including minority students (through STARS program at Washington University) and involve them in developing modules driven by fast GROMACS and MARTINI engines with coarse grained potentials to help learn self assembly in surfactant solutions.
Surfactants are molecules that are "amphiphilic" - literally "loving both" water and oil. They are therefore the basis of all soaps, shampoos, cleansers, and detergents. Since they dissolve in water, but can attract, and pick up oily, greasy molecules and sequester them in small compartments call micelles that disperse the oil into the water, to be flushed away. Formulating good surfactant solutions is the basis of large industries, including the consumer product industry, for whom the combination of pleasing fragrance, strong cleansing ability, and the right "body" - or viscosity - is required to please the customer. A very successful strategy for makers of shampoos and body washes is to make surfactant molecules that assemble themselves into long cylindrical micelles, such as that depicted in the figures. These structures kill two birds with one stone. The oil can go into the interior of the cylinder, and the cylinder itself acts like a giant thread that entangles with other such cylindrical threads, forming a viscous solution in water, with great "body" so the shampoo doesn't drip out of one's hand. However, salts and perfumes which must be added to the solution can degrade the cylinders, breaking them into small cylinders or even spheres. Or, the added ingredients can make the cylinders turn into large sheets, which form a "scum," like what forms at the bottom of a soap dish. Our research seeks to use powerful new computational tools to help formulate these products faster and more precisely than ever before, so that we can make more perfect cylinders that do not break down or turn into scum when perfumes and salts are added. With NSF support, and with our collaborators at Procter and Gamble, we have developed new tools to predict how surfactant can be made into micelles with the right properties. The work is of such interest to the company that the Ph.D. student Xueming Tang working on this project has spent two summers as a paid intern at Procter and Gamble, allowing her to gain industrial experience and leverage the NSF support with industrial support as well. The figure shows an example of the kind of a simulation of cylindrical micelles that Xueming, and others in our group, are producing. One of our computer simulation methods allows the lengths and other properties of cylindrical micelles to be determined from a measurement of their flow properties. This simulation code has been transferred to Procter and Gamble, and is now publically available. It should help manufacturers to more quickly design surfactant solutions for new products.
