Characterizing, Classifying and Manipulating Gas Hydrate Crystaline Interfaces and Associated Liquid-Like Layers to Understand Their Nucleation and Growth

Gas hydrates are inclusion compounds comprising a backbone of water molecules that enclose guest molecules in separate cages. Each volume of hydrate contains 160 volume equivalents of gas. The formation of these structures is a sophisticated crystallization process governed by heat and mass transfer considerations. It is composed of two steps: initial hydrate nucleation and subsequent growth. Water and gaseous guest molecules form small clusters, eventually reaching some critical size and yielding sustained growth. However, the exact mechanisms of the formation of small clusters, the diffusion of these nuclei, as well as the controllable phases of this process are unknown or poorly characterized. The labile cluster nucleation hypothesis maintains that water molecules form locally ordered arrangements that may form hydrate cage precursors. Pentagons, hexagons, and half-formed cages may form spontaneously and around gas molecules. The polygons may join to form cages, or the half-formed cages may close with another part of a cage before dissociation. These dynamics are too fast to reliably capture experimentally. With these challenges and dynamics in mind, the work uses molecular dynamics, graph theory, and geometric modeling to characterize, classify, and manipulate the different phases and elements of gas hydrate nucleation and growth to improve the understanding of crystallization and its potential applications in green science and engineering. Geometric modeling provides computationally inexpensive opportunities to create an analog of the energy landscape inside the gas hydrate cages. Examining the effect of the placements of spherical, linear, planar, and more complex guest molecules, we show that there are preferred orientations of guest molecules in the crystal structures. We resolve these orientations into points, lines, and planes of minimum energy and confirm the findings with rigorous density functional theory simulations. Additionally, we highlight the power of geometric modeling in providing accurate initial estimates of ground state atomic configurations but note the shortcomings of not including quantum effects and other electronic interactions. We subsequently use graph theory to provide an efficient and accurate cage recognition algorithm to obtain precise values of the lifetime of prenucleation clusters. Importantly, we identify the location of the formation of polygons, half-cages, and full cages in gas hydrate systems and their precursors. Combined with the results on minimum energy orientations of guest molecules, we can identify potential nucleation sites for hydrates in a simulation box. These results will lead to spatially and temporally resolved concentration profiles for the various subunits of critical hydrate nuclei relative to the interface of the system. With information regarding prenucleation clusters, energy landscape, and cluster distributions, we employ molecular dynamics simulations to control nucleation and growth behaviors. By examining the dipole moment distribution, we show that the water molecules in the novel liquid-like interfacial layer show strong parallel orientations with the surface of the hydrate. We then manipulate the interfacial layer and distribution of clusters by applying an external electric field. Preliminary results show that an electric field is a tool for controlling the spreading of the interface, the mass transfer within the interfacial layer, and disrupting the formation of prenucleation clusters in the bulk liquid phase as well as interrupting the organization in the liquid-like layer, tuning the nucleation kinetics of the gas hydrate system and providing avenues to go beyond classical nucleation theory and providing novel characterization of the multistep, multiphase crystallization process.