Amadeu Sum and David Wu, Colorado School of Mines (CSM), Gregg Beckham, CSM and National Renewable Energy Laboratory (NREL), Baron Peters , University of California, Santa Barbara and Valeria Molinero, University of Utah are supported by a CDI-TYPE II award to develop novel path sampling methods, based in statistical mechanics and information theory in order to collect and analyze data from large scale molecular simulations. These new techniques are being developed with the ultimate goal of elucidating the mechanism of clathrate hydrate nucleation. The specific aims are 1) to develop order parameters to identify ice and clathrate hydrate structures; 2) develop the Packet Evolution Path Sampling (PEPS) method; and 3) apply PEPS to simpler systems to demonstrate the utility of the method. The order parameters are a key component this effort as they provide physical insight and quantify the mechanisms of nucleation. The new sampling method (PEPS) provides a more efficient method of sampling and extracting mechanistic details on nucleation by harvesting the evolution along a 'reaction coordinate.' Lennard-Jones liquid-to-solid nucleation and water-to-ice nucleation are the testing systems to demonstrate the first application of the methods which will then be applied to clathrate hydrates nucleation. The methods themselves, however, are generally and may be used to gain understanding from molecular simulations of a variety of complex phenomena.
Understanding transition mechanisms in real systems is a daunting problem for molecular simulation, both in terms of efficient data collection and analysis. There is a need for transformative new methods that can be applied to realistic systems and huge data sets that are now commonplace with increasing computing power. One specific motivating problem where new path sampling methods can be applied is to the mechanism of clathrate hydrate nucleation. Clathrate hydrates, ice-like compounds formed from hydrogen-bonded water cages surrounding small non-polar molecules such as methane or carbon dioxide, have gained interest for their importance in energy (hydrates are present at the ocean floor in quantities estimated to be at least twice that of known oil and gas reserves; conversely, plugging of pipelines by gas hydrate formation poses the primary flow assurance problem to the oil and gas industry, as well as an impedance in the containment of a deepwater oil/gas blowout) and the environment (carbon sequestration and methane release from natural hydrate deposits). In a broader context for the application to hydrate nucleation, fundamental knowledge will be gained in nucleation theory, understanding arising from hydrophobic and hydrophilic interactions, and new approaches for structural quantification. Insight into the molecular mechanism for hydrate nucleation may be transformative because it will be constructing knowledge from microscopic interactions to macroscopic behavior, as opposed to current approaches where the reverse is predominant.
Understanding transition mechanisms in real systems is a daunting problem for molecular simulation, both in terms of efficient data collection and analysis. We have used path sampling and statistical data mining tools to tackle these challenges for simple and complex systems, in particular to the mechanism of clathrate hydrate nucleation. Clathrate hydrates, ice-like compounds formed from hydrogen-bonded water cages surrounding small non-polar molecules such as methane or carbon dioxide, have gained interest for their importance in energy and the environment. Large-scale microsecond simulations have produced direct, unbiased simulation of hydrate nucleation. While visual observation of these trajectories produced qualitative insight into the complex, multistep nucleation mechanism, this project addressed the quantitative understanding is hindered by the expense and difficulty both in the collection and analysis of the vast data contained in such long trajectories over rough energy landscapes. To tackle problems such as these, which are ubiquitous in chemical physics, we developed algorithms to address these challenges and apply the methods to elucidate the nucleation mechanism of clathrate hydrates (knowledge) from simulations (data). Intellectual Merit. The following has been accomplished in this project: 1) developed a quantitative order parameter to identify the small building block (mutually coordinated guest) in the formation of hydrate structures, 2) used advanced sampling techniques (equilibrium path sampling) to obtain a rigorous reaction coordinate and mechanism for methane hydrate nucleation, 3) obtained enhanced methane hydrate nucleation by freeze-concentration, 4) developed a first-of-a-kind stochastic model for solute precipitate nucleation in the freeze-concentration boundary layer, 5) obtained heterogeneous nucleation of ice on graphitic surfaces, 6) studied the thermodynamics of melting, vitrification and crystallization of binary water-salt mixtures, 7) studied novel phases of binary mixtures of water and spherically symmetric solutes, 8) studied solubility/supersaturation calculations for nucleation simulations, and 9) obtained nucleation rate at moderate driving forces. These accomplishments produce two transformative outcomes: 1) novel and much-needed techniques for extracting mechanisms from large-scale simulations of complex phenomena, and more generally to stochastic dynamical systems, and 2) quantitative determination of the hydrate nucleation mechanism, providing the basis for advances in the control of hydrate formation. Broader Impact. This project had a large impact in the training of postdocs, graduate students, and undergraduate students. 17 publications have been generated from the funding of this project, and over 50 presentations have been given with material from this project. All the personnel involved in the project (PIs, postdocs, and students) have been numerous presentations at technical conferences, workshops, and short courses, exposing the methods and results generated from the project. Some of the lectures and short courses given by the PIs have been made available in YouTube. The PIs have also worked with high school students and teachers to teach them about clathrate hydrates and demonstrate the importance of the research being performed.