Title: Advancing Software for Modeling Complex Molecular Systems: Investigating Surface Roughness Effects on Ionic Liquid-Silica Interfaces

Abstract: Current software tools for generating molecular systems (i.e., trial configurations) present significant limitations in exploring the extensive domain of chemical space. These tools predominantly cater to constructing molecular systems of smaller scale (<1000 atoms) and are constrained to simple geometries (e.g., boxes, cylinders, spheres). Their applicability becomes cumbersome or even unfeasible for generating larger systems (>100,000 atoms) or those with complex geometries (e.g., stars, rings, hyperbolic surfaces). Such large and geometrically diverse systems are, nevertheless, ubiquitous across various scientific endeavors (e.g., viruses, protein motors, imperfect surfaces). In my proposed research, I am investigating hydrophilic-oleophobic coatings employing innovative ionic liquids, such as HFILOH, on various silica substrates that vary in their surface roughness. Preliminary findings have demonstrated the efficacy of these coatings, yet the precise mechanisms underlying molecule diffusion through these coatings remain a subject of contention. Given the complexity of this system, existing tools are inadequate to even generate the input files for a molecular simulation package, let alone run the simulation itself. Consequently, the objectives of my proposal are as follows:

The first objective entails the creation of an open-source Python package, which will be integrated with Avogadro, to facilitate the generation of these complex systems. The short-term goal is to create software that can fill arbitrary volumes with chosen patterns of atoms, which can be used to model various silica surface morphologies, for example. The long-term goal is to deliver a software solution capable of rapidly assembling large molecular systems in any desired configuration while giving the user full control over the important descriptors of the system such as density, molecular orientation, or the distance between molecules. This could look like designing system geometries in other software such as Blender or SOLIDWORKS, and then turning those shapes into molecules. While no optimizations would occur, the package will also feature the ability to introduce more chaos into a system by randomly rotating or translating molecules. This package aims to be user-friendly, seamlessly integrated with Avogadro, and versatile enough to handle matter as solids (both crystalline and amorphous), liquids, gases, or any combination of the three.

The second objective involves employing this Python package to generate initial structures for a simplified ionic liquid system, specifically, BMIM-NTF2 coated on a crystalline silica substrate. The application of LAMMPS for molecular dynamics simulation will enable the study of the ionic liquid's atomic configuration on the surface and its diffusion mechanism into the substrate. The ultimate goal is to incrementally introduce complexities, such as variable surface textures, including a native oxide layer, using a more accurate amorphous silica substrate, or using a more complex ionic liquid (i.e., HFLIOH) to these models to better approximate real-world conditions.

Chair:

Christopher E. Wilmer

Department of Chemical and Petroleum Engineering, University of Pittsburgh

Committee:

Geoffrey R. Hutchison

Department of Chemistry, University of Pittsburgh

Lei Li

Department of Chemical and Petroleum Engineering, University of Pittsburgh

J. Karl Johnson

Department of Chemical and Petroleum Engineering, University of Pittsburgh