Nonaqueous phase liquids (NAPL, e.g., hydrocarbons and chlorinated compounds) are common groundwater pollutants. Electrokinetic remediation of NAPLs uses electric fields to draw them toward electrodes and remove them from groundwater. The treatment requires NAPL mobility. Emulsification increases mobility, but at a risk for downstream receptors. We propose using alkaline aqueous solutions of zein and graphene nanoparticles (GNP) to form conductive materials, which could also act as barriers to control NAPL migration. Alkaline zein–GNP solutions can be injected in the polluted soil and solidified by neutralizing the pH (e.g., with glacial acetic acid, GAA). Shear rheology experiments showed that zein–GNP composites were cohesive, and voltammetry showed that GNP increased electrical conductivity of zein-based materials by 3.5 times. Gas chromatography–mass spectroscopy (GC–MS) demonstrated that the electrokinetic treatment of model sandy aquifers yielded >60% and ∼47% removal of emulsified toluene in freshwater and in salt solutions, respectively (with 30 min treatment using a 10 V differential voltage between a zein–GNP and an aluminum electrode. NaCl was used as model salt contaminant. The conductivity of surfactant solutions was lower in saline water than in freshwater, explaining differences in toluene removal. Toluene–water emulsions were stabilized using the natural surfactants lecithin and saponin. These surfactants acted synergistically in stabilizing emulsions in either freshwater or salt solutions. Lecithin and saponin likely interacted at toluene–water interfaces, as indicated by the morphology, interfacial tension and compressional rigidity of toluene–water interfaces with both components (relative to interfaces of either lecithin or saponin alone). The compressional behavior of interfacial films was well-described by the Marczak model. Electrokinetic treatment of saturated model sandy aquifers also decreased the turbidity of emulsions of water and either tricholoroethylene (TCE, by ∼41%) or diesel (by ∼75%), in the presence of a bacterial biosurfactant. This decrease was used as semiquantitative indicator of NAPL removal from water.