Specialization: Chemical Engineering Degree: Master of Science Abstract: Athabasca Bitumen and other heavy hydrocarbon resources are high-mean-molar-mass and structured organic materials with complex phase behaviors at the nano- and micro- meter length scales. Diffusion of light hydrocarbons, and non-hydrocarbons in these resources is of significant interest as new production and refining concepts that envision, for example, addition of light hydrocarbons and non-hydrocarbons to reservoirs to enhance production have begun to emerge. Composition-distance profiles arising during free diffusion scale as a function of the joint variable (distance/time^nw). Simple fluids are governed by Fickian diffusion, where nw = 0.5. For nanostructured fluids the value of nw can be as low as nw = 0.25, known as the single file limit but more typically the value for the exponent falls between these two limits and is composition dependent. In this work, five published data sets comprising free diffusion composition profiles for Athabasca bitumen fractions and for Cold Lake bitumen + light hydrocarbons obtained using diverse apparatus, are probed from this perspective. Additional experimental results are provided for Athabasca bitumen + toluene mixtures over the temperature range 273 to 313 K, and results from positive and negative control experiments for two well-defined mixtures: (0.25 mass fraction carbon nanotubes + polybutene) + toluene, and polybutene + toluene, are also provided. The values of the exponent nw are shown to be light hydrocarbon dependent. They increase from nw ~ 0.25 at low light hydrocarbon mass fraction up to nw ~ 0.50 at high light hydrocarbon mass fraction. An approximate solution to the composition profile arising during Single File diffusion is presented and used along with a standard composition profile for Fickian diffusion to simulate sets of free diffusion composition profiles in these complex mixtures. Three parameters: a Fickian mutual diffusion coefficient, a coefficient H fit to composition profile, and a composition dependent weighting function for the two diffusion mechanisms, are employed in the model. Outcomes, including Fickian mutual diffusion coefficients, order of magnitude estimates for Single File mobility coefficients and the roles of diluent type and asphaltene content on the controlling diffusion mechanism, and industrial implications are discussed.