Extraction efficiencies for a series of model compounds representing heteroatom functional groups believed to be present in asphaltenes were determined in batch extractions with a variety of metal oxide nanoparticles (Fe3O4, TiO2, NiO, Co3O4, and MgO). Extraction efficiencies from toluene solution varied depending upon both the adsorbate and the type of metal oxide used for extraction. However, the adsorbate was found to be the most important factor governing selectivity, which generally followed the trend: benzoic acid >> pyridine approximate to phenol > pyrrole > thiophene approximate to diphenylsulfide approximate to benzophenone. An important exception to this trend was that MgO did not appreciably adsorb pyridine. The divergent adsorption behavior of pyridine on NiO (extraction efficiency = 82 +/- 1%) and MgO (extraction efficiency = 0 +/- 2%) was subsequently exploited to demonstrate a novel approach for profiling pyridine-containing molecules in an authentic asphaltene sample. Specifically, mass spectra of the asphaltene mixture were obtained before and after treatment with NiO or MgO and compared to identify peaks exhibiting reduced intensity after treatment with NiO but no appreciable change in intensity after treatment with MgO. Results of batch extraction studies with model compounds and elemental composition data deduced from accurate mass measurements support that these peaks likely correspond to (or minimally contain) a molecule(s) possessing a pyridyl functional group.