Our research is focused on bridging the gap between atomic/nano-scale and meso/macro-scale scientific and engineering problems using applied mathematics, computational materials science, and thermodynamic modeling. We supplement my computational work with verification and validation studies that are based on advanced analytical, spectroscopic, and electrochemical techniques. This framework has applications in surface and interface science, corrosion science, electrochemistry, thin films/oxides, durability of materials, transport in porous media, cement/concrete research, inverse modeling, and non-destructive testing and evaluation.
In the atomic scale we use Reactive Force Field Molecular Dynamics (ReaxFF-MD) using the Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS) framework. Unlike the atomic force field models that are used in conventional molecular dynamics simulations of non-reactive processes, ReaxFF-MD allows modeling of chemical reactions that include breaking and forming of bonds, where bond order is determined empirically from interatomic distances. Unlike computationally expensive explicit quantum mechanical calculations, the reactive force field approach treats chemical bonding implicitly, which yields significant computational efficiency. In other words, ReaxFF-MD bridges the gap between the traditional MD and the computational quantum-chemistry; therefore, it is well suited to study phenomena at metal-electrolyte interfaces such as corrosion, passivation, and film formation.
Although ReaxFF-MD can model relatively large domains that can simulate surface phenomena, it cannot answer fundamental questions that take place in the quantum mechanical scale. Among these are the determination of critical force field parameters such as interatomic distances, bond angles, bond energies, atomic charges, adsorption energies, surface energies, and activation energies. These properties are critical to develop fundamental understanding of the studied phenomena and serve as input for the ReaxFF-MD simulations. We perform these calculations using spin-polarized Density Functional Theory (DFT) via Vienna Ab initio Simulation Package (VASP).
The verification and validation data for the atomic scale simulations are critical thrusts of our research. For this purpose we use advanced nano-scale analytical/spectroscopic tools such as x-ray photoelectron spectroscopy (XPS), electron energy loss spectroscopy (EELS), Auger and Raman spectroscopy, transmission electron microscopy (TEM), focused ion beam (FIB) sampling, and low energy diffraction. These analytical techniques provide temporal and spatial snapshots of the processes we study; however, in this scale they do not allow kinetic observations. To overcome this problem, we use the electrochemical quartz crystal nanobalance (EQCN) technique to link our computational and analytical observations by monitoring nano-scale mass changes during processes that occur at surfaces. We are also exploring additional in-situ techniques that can study kinetic processes on nano-scale surfaces and interfaces in real time.
We link our atomic/nano-scale studies to meso/macro scale using thermodynamic and continuum modeling approaches and electrochemical/spectroscopic methods. Finite element modeling is one of the major tools that we use in modeling continuum phenomena such as reactive-transport processes in porous media, corrosion, and damage. Our group has made a major contribution in modeling reactive-transport processes by coupling thermodynamic simulations with non-linear finite element modeling algorithms. For this purpose we developed an interface between COMSOL Multiphysics finite element analysis software and (geo)chemical thermodynamic modeling platform, which is based on Gibbs Free Energy Minimization (GEM) technique, for the reactive-transport modeling of (geo)chemical processes in variably saturated porous media. The two standalone software packages are managed from the interface that uses a non-iterative operator splitting technique to couple the transport (COMSOL) and reaction (GEMS) processes. The interface allows modeling media with complex chemistry (e.g. cement) using GEMS thermodynamic database formats.
Currently we are expanding this framework to include modeling of corrosion and corrosion-induced damage processes by replacing the commercial FEM modules with generic FEM codes that are based on the FEniCS Project framework. The research involves four main tasks toward the development of comprehensive modeling framework for corrosion-induced damage in reinforced concrete structural components: (1) generation of the discretized finite element domain of the analysis framework; (2) development of coupled reactive-transport-corrosion modeling framework; and (3) coupling reactive-transport-corrosion framework with chemo-damage modeling.
We generate kinetic and thermodynamic data from micro- to macro-scales for verification and validation purposes using localized and bulk electrochemical methods to collect. Electrochemical impedance spectroscopy (EIS), electrochemical noise (EN), polarization resistance, and cyclic voltammetry techniques allow us to test the hypotheses that are developed in simulations. Localized measurements using Scanning Kelvin Probe, Scanning Vibrating Electrode Technique, localized EIS, and Scanning Electrochemical Microscopy allow us to link nanoscale observations to bulk scale.
Current and past funding sources for our research include: