The structure of the electrical double layer at the interface of planar electrodes and aqueous solutions is investigated. Electrical impedance spectroscopy is used to measure the impedance of aqueous solutions of sodium chloride and two different surfactants over a wide range of concentrations. The electrode capacitance is directly inferred from the admittance spectra, as well as by regression of the impedance spectra to an equivalent circuit. It is found that the electrode capacitance remains on the same order of magnitude over the entire range of investigated concentrations. This is contradictive to the predictions of the Gouy-Chapman-Stern theory which predicts that at low concentrations, the electrode capacitance should be determined by the diffuse layer. It is concluded that the Stern layer capacitance always dominates the electrode capacitance, even at very low concentrations, and the establishment of a diffuse layer capacitance 1 requires an ionic strength of around 1 mM.
The addition of surfactants can considerably impact the electrical characteristics of an interface, and the zeta potential measurement is the standard method for its characterization. In this article, a comprehensive study of the zeta potential of poly(methyl methacrylate) (PMMA) in contact with aqueous solutions containing an anionic, a cationic, or a zwitterionic surfactant at different pH and ionic strength values is conducted. Electrophoretic mobilities are inferred from electrophoretic light scattering measurements of the particulate PMMA. These values can be converted into zeta potentials using permittivity and viscosity measurements of the continuous phase. Different behaviors are observed for each surfactant type, which can be explained with the various adsorption mechanisms on PMMA. For the anionic surfactant, the absolute zeta potential value below the critical micelle concentration (CMC) increases with the concentration, while it becomes rather constant around the CMC. At concentrations above the CMC, the absolute zeta potential increases again. We propose that hydrophobic-based adsorption and, at higher concentrations, the competing micellization process drive this behavior. The addition of cationic surfactant results in an isoelectric point below the CMC where the negative surface charge is neutralized by a layer of adsorbed cationic surfactant. At concentrations near the CMC, the positive zeta potential is rather constant. In this case, we propose that electrostatic interactions combined with hydrophobic adsorption are responsible for the observed behavior. The zeta potential in the presence of zwitterionic surfactant is influenced by the adsorption, because of hydrophobic interactions between the surfactant tail and the PMMA surface. However, there is less influence, compared to the ionic surfactants. For all three surfactant types, the zeta potential changes to more-negative or less-positive values for alkaline pH values, because of hydroxide adsorption. An increase of the ionic strength decreases the absolute value of the zeta potential, because of the shielding effects.
We introduce a new neural network architecture, Multimodal Neural Graph Memory Networks (MN-GMN), for visual question answering. The MN-GMN uses graph structure with different region features as node attributes and applies a recently proposed powerful graph neural network model, Graph Network (GN), to reason about objects and their interactions in an image. The input module of the MN-GMN generates a set of visual features plus a set of encoded region-grounded captions (RGCs) for the image. The RGCs capture object attributes and their relationships. Two GNs are constructed from the input module using the visual features and encoded RGCs. Each node of the GNs iteratively computes a questionguided contextualized representation of the visual/textual information assigned to it. Then, to combine the information from both GNs, the nodes write the updated representations to an external spatial memory. The final states of the memory cells are fed into an answer module to predict an answer. Experiments show MN-GMN rivals the state-of-the-art models on Visual7W, VQA-v2.0, and CLEVR datasets.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.