No abstract

An attenuated association model describing the aggregation of asphaltenes in solution is extended to derive an equation for the weight-average degree of association and account for phase behavior. The weight-average molecular weight is calculated to be higher than number average, as it must be for a polydisperse material, but not by enough to explain the very large differences in these quantities reported in the literature. Binodals and spinodals are calculated using expressions derived previously, but modified to account for free volume (thermal expansion) differences. The phase behavior of asphaltene solutions is examined in more detail, particularly in the dilute solution regime. It is shown that the formation of nanoaggregates significantly affects the critical value of the χ interaction parameter. The phase diagram is highly asymmetric and the phase boundary approaches the pure solvent composition limit. This has a number of implications in terms of asphaltene solution characterization and the nature of asphaltene solutions. The results indicate that there are toluene insoluble asphaltene components, but these could exist as microphase-separated clusters stabilized against further aggregation by steric and kinetic factors. This would explain the large difference between observed number and weight-average molecular weights. In addition, because of the shape of the binodal curve at low concentrations, experimental data that have previously been interpreted in terms of a critical cluster or micelle concentration are shown to be consistent with a microphase separation.

A guide to the solubility of asphaltenes in a range of solvents is constructed through the use of an association model to account for asphaltene nanoaggregation and its effect on phase behavior and Hildebrand solubility parameters to model interactions between aggregates and solvent. Solvents are classified according to their polarity and ability to self-associate (e.g., through hydrogen bonds). In addition, estimates of the contribution of free volume terms to interaction parameters indicate that a further distinction must be made between solvents with flexible molecules (such as the n-paraffins) and those that are relatively inflexible (such as toluene). A "bare" interaction parameter (χ 0 ) is calculated, and it is this parameter that is related to Hildebrand solubility parameters. For nonpolar and weakly polar solvents, a critical value of the solubility parameter difference (Δδ c ) between an asphaltene or asphaltene component and solvent is calculated to be ±3.5 MPa 0.5 at 25 °C for a nonpolar or weakly polar solvents with largely inflexible molecules and a molar volume of 100 cm 3 /mol. For flexible solvents such as the n-paraffins, free volume effects are larger and Δδ c is about ±2.8 MPa 0.5 . For strongly polar solvents that have limited flexibility, the equivalent critical value of the solubility parameter difference is also calculated to be ±2.8 MPa 0.5 . Hydrogen bonded solvents like methanol are calculated to be immiscible with asphaltenes, with miscibility being defined as forming a single phase across the composition range (at 25 °C). Miscibility maps are constructed in terms of the calculated phase boundary at the critical point, δ A ± Δδ c , where δ A is the asphaltene component solubility parameter, plotted as a function of solvent molar volume. The solubility of asphaltenes and asphaltene components in various solvents is discussed. The solvent that defines asphaltene identity, toluene, is predicted to dissolve only a limited range of asphaltene components. This is consistent with various reported experimental observations. However, solubility is often defined in terms of an absence of a visible precipitate, and on the basis of recent work in the literature, toluene solutions may contain some microphase-separated material stabilized against further aggregation by steric and kinetic factors.

No abstract