Research

Metal-Ceramic Interfaces Metal machine parts (such as gears and pistons) often break down due to corrosion, along with thermal and mechanical wear. Modern parts are coated with ceramics to reduce these and increase their lifetimes (notice how an automotive transmission lasts much longer than it used to). However, understanding the interactions between metals and ceramics are not well understood. These calculations required parameterizaing modified embedded atom method (MEAM) models for metal and ceramic interactions. This particular case on the right investigated the impact of small to medium amounts of Al to the Ti-TiN interface, finding significant enhancement in shear strength. We are waiting on our collaborators from LSU to do these experiments to confirm, but this can be a way to tailor the interfacial strength of metal-ceramic interfaces.

Complex Conjugated Alloys We are focused on specific types of complex conjugated alloys that have phase separation at lower temperatures into two similar phases (in this case FCC). Part of this work is developing computational tools to do automated DFT calculations of binary and ternary interactions, along with automatically parameterize MEAM models. The parameterization method for the MEAM models uses a genetic algorithm. Once the models are parameterized to binary and ternary interactions, any combination of them (up to any number of different elements) can be simulated. We use a Monte Carlo minimization procedure to identify stable structures for these systems. The example shown is for CrNiCu, which separates into to FCC phases, and was confirmed with experimental measurements.

Shape Memory Polymers Shape memory polymers allow a change in conditions to return a polymer to its initial shape. During this, a force is released, called recovery stress, that can be used to carry out useful work. Furthermore, we study vitrimers, wich are polymers that can self-heal. We are studying these with computational means to determine how best to optimize recovery stress. This project requires modeling of step-wise polymerization, free radical chain polymerization, and vitrimer reactions to form and reform polymer networks. A goal is to determine which atomistic and topological (of the network) "fingerprints" can be used to predict the recovery stress of the polymer network.

Li-Ion Battery Electrodes The goal is to devise battery electrodes that hold the most lithium (allowing greater capacity) at a reasonable voltage, but also that this capacity is retained for many cycles. To achieve this goal, an understanding of the charging and discharging mechanism in these materials is needed. In these calculations, the lithiation mechanism for a variety of materials is investigated with plane-wave DFT. This is in collaboration with Dr. Ramachandran and Dr. Meda at Xavier University, who carries out experiments of these systems. The experimental discharge voltage curves are calculated and compared with experiment to verify the computational results. With this verification, insights into the mechanism are provided. Notice how much of the systems include interfaces. Our published work on this gave strong evidence that extra capacity can be stored at the interface between metallic and Li2O phases.

Metal Alloy Interfaces Selective laser 3d printing is the process in which metal powders are used to mold machine parts of specific shapes, sizes, and properties. To optimize this process requires knowledge of the interfacial properties of the alloys used. For instance, at the solid/liquid and air/liquid interfaces of TiAl6 alloys, is there more Al at the these surfaces than in the bulk, and how does the segregation of certain metals influence the interfacial and mechanical properties. To study this requires large scale simulations, not achievable with DFT calculations, requiring embedded atom (MEAM, specifically) models. Unfortunately, these models are generally designed for solid properties, and others, such as melting points and surface tensions are often poorly reproduced. We are developing new MEAM models with a self designed Python optimization code that works with the LAMMPS software. This code uses a combination of Monte Carlo and Genetic Algorithm to fit the MEAM parameters to solid, liquid, phase change, and interfacial properties of metals and alloys.

Aqueous Interfaces When anything comes in contact with water, it first has to interact with the interface. For aerosol's and microscopic droplets, the interface can become a significant part of the available phase space. As such, the unique properties of interfaces need to be understood. I have carried out many calculations of aqueous interfaces, focusing particuraly on ions, including halides, alkali, hydronium, and hydroxide. The figure to the right shows a hydroxide anion (which is yellow) at the instantaneous air-water interface, which was devised by Chandler and coworker. The mesh makes it difficult to see that the hdyroxide anion accepts hydrogen bonds from multiple water molecules, which share its anionic charge. In fact that hydroxide anion has a charge around -0.7, versus the expected -1.0 because it shares its negative charge with adjacent water molecules. The MS-EVB model (devised by Voth and coworkers) allows one to model this efficiently and accurately. The interfacial pH is influenced by this, which if it differs from the bulk, can provide a unique environment for acid or base catalyzed reactions.