IMPROVED MOLECULAR MODEL GENERATION FOR SOOT, CHARS, AND COALS: HIGH-RESOLUTION TRANSMISSION ELECTRON MICROSCOPY LATTICE FRINGES REPRODUCTION WITH FRINGE3D
Open Access
- Author:
- Fernandez-Alos, Victor
- Graduate Program:
- Energy and Geo-Environmental Engineering
- Degree:
- Master of Science
- Document Type:
- Master Thesis
- Date of Defense:
- April 21, 2010
- Committee Members:
- Jonathan P Mathews, Thesis Advisor/Co-Advisor
Jonathan P Mathews, Thesis Advisor/Co-Advisor
Harold Harris Schobert, Thesis Advisor/Co-Advisor
Randy Lee Vander Wal, Thesis Advisor/Co-Advisor - Keywords:
- soot
molecular model
3D HRTEM lattice fringe images
char
low-rank coals - Abstract:
- Large-scale molecular models for carbon-rich structures are beneficial for advanced simulations and scientific progression. However, the generation of such structures is challenging and expensive, have considerable creator bias and often unrealistic simplifications are necessary. An improved and more desirable approach for carbon-rich structures should ease the construction process, enable large-scale model creation (>20,000 atoms) with reduced bias, improve accuracy, diversity capture, and greater utility. For this project models for carbonaceous materials were constructed by integrating HRTEM lattice fringe image analyses with computational techniques. The models constructed were the aromatic moieties of a diesel-soot “primary particle”, a coal gasification char, and an extensive model of the economically important Powder River basin, Wyodak-Anderson subbituminous coal, which included heteroatom, aliphatic carbons, and water forms. Image analyses obtained from published HRTEM lattice fringe images were utilized to directly duplicate aromatic structural features such as fringe lengths (aromatic layer dimensions), fringes separation distance (interlayer spacing), the number of fringes per stack (stacking distribution), and fringe orientation. Basically, the Fringe3D process creates three dimensional (3D) aromatic structural features directly from HRTEM lattice fringe micrographs and populated the appropriate aromatic structures in 3D molecular space. The resulting slice aromatic models are an improved molecular representation as it captures the microstructure and more importantly the distributions of structural features. This approach was used to construct a diesel soot “primary particle” model with a realistic diameter and periodicity and gasification char with variation in stacking and orientations. Also, an extensive model of the economically important Powder River basin Wyodak-Anderson subbituminous coal was created, including heteroatom, aliphatic carbons, and water. The soot model was produced due the fact that only small-scale models with limited utility were available in the literature. A 60,000 atom aromatic carbons model was generated as a base structure. This model could assist to explain reactivity/microstructure relationships for fuel and generation-condition specific soot “primary particle” structures, in addition to benefits for soot/environment/health issues. The gasification char model contains about 800 molecules based on graphene-like structure with various stacking extents and orientations. Such a large-scale model could aid in the process of exploring structure reactivity relationships, which are important in energy applications and offers a new cheaper and faster construction approach that are an improvement in comparison to current methods used such as Reverse Monte Carlo or manual manipulations. Modification to the base structure such as oxygen functionality, cross-links, and aliphatic components could be added to this method. Notwithstanding the economic importance to power generation in the U.S. and potential for coal-to-liquids production, few models of subbituminous coals exists. Around 40% of domestic coal production comes from the Powder River basin in Wyoming. Coal, especially low-rank coal, was a more extreme test of the model construction approach and it required some refinements to progress from aromatic carbons layers model to 3D representation including an extensive molecular weight distribution, heteroatom functionalities, aliphatic components, and water forms. The model created has ~25,000 atoms and was constructed utilizing a variety of published data, specifically nuclear magnetic resonance(NMR), fast neutron activation analysis, laser desorption ionization spectroscopy, and other mass spectroscopy for molecular weight distributions. The model showed good agreement with: a) NMR parameters, b) average molecular weight of a cluster, and c) molecular weight distribution.