ACS PRF | ACS
All e-Annual Reports
44793-AC9
Understanding Wax Formation from a Molecular Perspective
This project is concerned with the study of wax formation in petroleum fuels from a molecular perspective. Petroleum fuels are modeled using n-alkanes as representative fuels and oils and the conformation of individual molecules and their packing in crystal structures are studied in order to build an understanding of stable states such as low temperature ordered (LO) solid phases and rotator (R) or liquid-like phases as well as the transitions between them. Physically observed, stable molecular conformations and crystal structures are assumed to correspond to stationary points (local and global minima) on an appropriate model of the potential energy surface while transition states are saddle points. Unfortunately, many models of potential energy surfaces for molecular conformation and crystal structure determination contain large numbers of minima at the small length scale but only a handful of low energy minima at the large length scale. Moreover, these two length scales often have significantly different geometry. To reliably determine important minima and saddle points, an in-house, multi-scale optimization method is used. This multi-scale optimization method is called the terrain/funneling method, is deterministic, and builds models of the geometry of the potential energy surface at two separate length scales. At the small length scale, the terrain methodology of Lucia and Yang (2003) is used to gather average gradient and curvature information, which is communicated to the large length scale. The funneling method of Lucia, DiMaggio, and Depa (2004) is used at the large length scale to make large changes in the geometric structure and drive the optimization calculations to the global minimum.
Results to date include the determination of the molecular conformations of diesel, home heating and residual fuel oils modeled as pure n-dodecane, n-hexadecane, and n-tetracosane respectively – as well as a number of other important stationary points on the potential energy surface. These results were computed using a united atom model of the potential energy with bond length, bond angle, torsion, and van der Waals contributions and have been recently published in the Journal of Global Optimization. We have also compared our multi-scale methodology with the basin hopping method of Doye and Wales (1997). Here numerical results clearly show that the terrain/funneling method is both more reliable and more efficient than basin hopping. Subsequent studies have been focused on comparisons of united atom and all atom models of the potential energy surface for n-alkanes because there are significant discrepancies in the structure predictions made by these two models. Of particular interest here is the identification of the model that is most suitable for both molecular conformation and crystal structure determination. Very recent results show that the all atom model yields planar zigzag molecules within crystal structures while a united atom model gives molecular conformations with considerable wrapping at the ends. Our findings suggest that the all atom model is more suitable for LO phases because it agrees very closely with experimental results but that the united atom model may be appropriate for R phases. Numerical results at the larger length scale of our multi-scale approach also clearly identify major potential energy levels or plateaus. This is an important observation because it suggests that other important conformations or crystal structures (R or other LO phases) are contained within these energy plateaus. If this conjecture is true, it makes the task of identifying important conformations and crystal structures, as well as the dynamics of transitions between them, considerably easier because it significantly narrows the conformations that need to be investigated. All of these facts are depicted in the accompanying graphical illustration for the molecular conformation of n-tetracosane using an all atom potential energy model.
The impact of this research on the PI's career has been significant. The present studies of wax formation in petroleum fuels and the surrounding results to date have provided strong evidence that the terrain/funneling methodology developed by the PI is a powerful tool for molecular conformation and crystal structure determination in petroleum fuels as well as other materials. This work has also led to a number of invited presentations by the PI and the invitation to join the editorial board of the Journal of Global Optimization. The results to date also clearly suggest that the main objective of this research – to build an understanding of wax formation of petroleum fuels from a molecular perspective – is clearly achievable and within reach.
The work thus far has also had a significant positive impact on the training and development of the graduate student participating in this research. It has enabled the student to present his work at one American Institute of Chemical Engineers meeting, to co-author a presentation at a conference on Frontiers in Advances in Global Optimization, held in 2007, and to prepare a manuscript of the first phase of this work for publication that has subsequently been published.
