How Asphaltenes Aggregate: Role of Chemistry and Solvent

Abstract

Asphaltenes were separated into several subfractions by sequential precipitation from n-heptane and toluene. Three distinct solubility regimes were indicated where the subfraction chemical and colloidal behaviors varied significantly with total precipitated asphaltenes. The earliest fractions precipitated had lower than average aromaticity and atomic N/C ratios and contained significant inorganic solids contents. Subfractions isolated in the second regime varied significantly in aromaticity and had systematically decreasing N/C ratios and increasing O/C ratios with increasing asphaltene yield. The most aromatic subfractions formed the largest aggregates in solution. Subfractions in the most soluble regime were more 'resin-like' in chemical composition and aggregation behavior.

Application of various geometric form factors to the SANS scattering spectra of asphaltenes suggested the aggregates are polydisperse radius oblate cylinders. A polydisperse cylinder model provided ranges of average particle thicknesses (5-32Å), radii (25-125Å), and polydispersity (~30%). Calculation of aggregate molar masses suggested solvent entrainment within the aggregates from 30-50% (v/v) that was consistent with previous viscosity measurements. Changes in the apparent aggregate mass with concentration indicated deviations from ideal solutions that were quantified through the calculation of second virial coefficients (A2). A2 values varied significantly with solvent conditions, concentration, and chemical composition of the solute. Results suggested that interactions of asphaltenes, resins, and solvent are dominated by dispersion and p-bonding interactions. Experimentally measured A2 values under-predicted those calculated on an excluded volume basis, suggesting energetic interactions of the solute and entrained solvent are significant.

UV-vis spectroscopy was used to determine the solubility of poly-nuclear aromatics in binary solvent mixtures with the intention of extending the methods to asphaltenic systems. Binary solvent mixtures were selected to probe specific intermolecular solvent-solute interactions (i.e., dispersion, polar, and hydrogen bonding). Solubility data were fit to a three-dimensional solubility parameter model that provided accurate solubility predictions in some solvents with less than 30% error. The predictive capability of the model decreased as one or more contributions to the solvent solubility parameter deviated from the solute. Experiments measuring solubility in a multi-component solute mixture suggested a need to incorporate aggregation in the model.