Predicting the Properties of Polymer Nanocomposites Without Forgetting the Atomistic Details
 |
Landry Khounlavonglandry@che.utexas.edu Office: CPE 3.402 Phone: 512-471-6754 |
Polymer nanocomposites (PNCs) have existed for a long time in nature, yet it was not until the 1980s that scientists could synthetically produce them. Once accomplished, many discovered that PNCs can display extraordinary properties besting traditional polymer composites of the past, which have fillers with a length scale on the order of microns whereas fillers in PNCs have at least one dimension on the order of nanometers. Such a minor difference has the ability to produce superior mechanical, barrier, electrical, and optical properties for some PNCs at very low filler loadings. How can a minor difference lead to such property enhancements? The answer seems to reside at the polymer-filler interface. At low filler loadings the interface in PNCs is quite substantial due to the nanometer size of the filler. Typically, experimentalists have seen that the most exfoliated PNCs show the greatest enhancements in mechanical properties. One way to interpret this observation is homogenously dispersed fillers lead to more polymer-filler contacts which allow for stress to transfer to the filler. In addition, it has been shown that altering the polymer-filler interaction by dressing the filler with amphiphiles can greatly influence the filler’s arrangement in the polymer matrix. From these discoveries we notice two key points: (1) the structure of the fillers influences the properties of the PNC and (2) the surface chemistry plays a vital role in determining how the fillers disperse in the polymer matrix. The latter requires an atomistic lens to resolve while the former demands large length scales and long time scales to allow one to obtain macroscopic properties. Thus, in order to accurately predict PNC properties we use a multiscale approach that utilizes atomistic information to predict filler structure, which is then used to determine the macroscopic properties of PNCs.
Projects:
Linking atomistic details and nanoparticle structure
In order to accurately simulate PNCs and predict the filler structure we must include the important structural details at the polymer-filler interface. Such details require a fine resolution. Unfortunately, to predict filler dispersion with atomistic simulations the required simulation box would be too large and simulation time too long to accommodate. Therefore, we take an alternative approach and decide to view a PNC from a slightly lower (mesoscopic) resolution but not low enough to wipe out the details of the polymer-filler interface, i.e. we coarse-grain (CG) the PNC (see Figure 1). We determine the minimal amount of atomistic information required to generate CG interaction potentials for mesoscopic simulations of a two component PNC. To CG the individual interactions in the system we borrow existing CG methods that have been developed for single component fluids and polymers. From the mesoscopic interactions we are able to successfully predict the filler structure (see Figure 2), despite eschewing some of the atomistic details, for filler volume fractions up to 20% while significantly lowering the computational resources required to predict the filler structure. This is accomplished regardless of how the polymer interacts with the filler.
Determining how the polymer-filler interface influences PNC rheology
While the polymer-filler interaction can dictate the morphology of the PNC, it also can influence the dynamics of the polymer near the surface of the filler. In addition to this, the arrangement of the filler in the polymer can affect the polymer motion. Such influence undoubtedly leads to changes in rheological properties of the PNC relative to that of the pure polymer. Via CGing (see above), we have the capability of accurately predicting filler structure with mesoscopic simulations. Yet, CG potentials defining the mesoscopic simulation tend to be smoother than their atomistic counterparts, and thus ruin the hydrodynamics in the mesoscopic simulations. To appropriately account for hydrodynamics in a CG framework dissipative particle dynamics (DPD) can be used to simulate the system. The trick in this instance is how to connect the terms accounting for hydrodynamics with the effects produced by the polymer-filler interactions. To accomplish this, we will incorporate information from the motion of the polymer near the filler’s surface (see Figure 3) into DPD simulations.
Revealing how interfacial layers influence PNC barrier properties
When fillers are introduced into polymers, they can perturb the polymer that surrounds it resulting in a region of polymer with nonuniform local properties different from that of the unperturbed (or bulk) polymer. The extent of this region from the surface of the filler can have a length scale on the order of the filler size. If the fillers are well dispersed in the matrix, then they have the opportunity to overlap creating even more regions of varying local properties. This behavior has proven useful for membranes used for gas separation. These interfacial layers have allowed for high selectivity for large molecular weight molecules by generating more free volume in the polymer regions near the filler’s surface. We developed a numerical approach that elucidates these effects and builds upon previous theoretical works. Our approach (1) uses a microscopically based semi-flexible polymer model to determine the penetrant diffusivity characteristics in interfacial layers and (2) embeds the so-determined interfacial characteristics into a numerical homogenization procedure that accounts for the overlap of interfacial layers (approximately) and multibody interactions (exactly). Ultimately, we will link this step to our CG methods by using the interfacial layers given by mesoscopic simulations as input to our homogenization procedure.
Publications:
L Khounlavong and V Ganesan, “Influence of interfacial layers upon the barrier properties of polymer nanocomposites”, J Chem Phys 130, 104901 (2009).
V Ganesan, L Khounlavong, and V Pryamitsyn, “Equilibrium characteristics of semiflexible polymer solutions near probe particles”, Phys Rev E 78, 051804 (2008).
About Me:
I graduated from the University of Florida with a BS in Chemical Engineering and joined the Ganesan group in 2005. Austin is a great place for outdoor activities so I take advantage of that by running and playing basketball, soccer, and tennis. Also, the music scene is huge here, leaving few to resist participating in South by Southwest and ACL, including myself. Though Austin is a great place to live, I do like to get out and travel around the U.S. and by far my favorite place to visit is Chicago. The fun part of traveling for me is getting to enjoy many different types of local cuisines.