Search

We describe the temperature-dependent characterization of resistive switching behavior in bulk silver nanowire−polystyrene composites between 10 and 300 K. We propose that the resistive switching behavior is caused by the electroformation of silver filaments between adjacent nanowire clusters, resulting in an extension of the electrical percolation network in the on state. This process is reversible above 200 K, and irreversible below 100 K. The switching fields are shown to depend strongly on sample composition (i.e., proximity to the electrical percolation threshold), as well as measurement temperature.

Traditionally, bulk nanocomposites of electrically conducting particles and insulating polymers have been categorized as either insulating or conducting when the nanoparticle concentration is below or above the percolation threshold, respectively. Meanwhile, thin-film polymer nanocomposites can exhibit resistive switching behavior appropriate for digital memory applications. Here, we present the first report of reversible resistive switching in bulk, glassy polymer nanocomposites. At compositions close to the electrical percolation threshold measured at low voltage, silver nanowire-polystyrene nanocomposites demonstrate reversible resistive switching with increasing voltage at room temperature. Nanocomposites with compositions outside of this range exhibit either irreversible switching, or no switching at all. We propose that resistive switching in these materials is the result of the field-induced formation of silver filaments that bridge adjacent nanowire clusters, extending the percolation network and decreasing the sample’s bulk resistivity. These findings break from the usual dichotomy of insulating or conducting properties in polymer nanocomposites and could inspire new devices that capitalize on this responsive behavior in these versatile materials.

The design and preparation of isotropic silver nanowire-polystyrene composites is described, in which the nanowires have fi nite L/D ( < 35) and narrow L/D distribution. These model composites allow the L/D dependence of the electrical percolation threshold, φ c , to be isolated for fi nite- L/D particles. Experimental φ c values decrease with increasing L/D , as predicted qualitatively by analytical percolation models. However, quantitative agreement between experimental data and both soft-core and core–shell analytical models is not achieved, because both models are strictly accurate only in the infi nite- L/D limit. To address this analytical limitation, a soft-core simulation method to calculate φ c and network conductivity for cylinders with fi nite L/D are developed. Our simulated φ c results agree strongly with our experimental data, suggesting i) that the infi nite-aspect-ratio assumption cannot safely be made for experimental networks of particles with L/D < 35 and ii) in predicting φ c , the soft-core model makes a less signifi cant assumption than the infi nite- L/D models do. The demonstrated capability of the simulations to predict φ c in the fi nite- L/D regime will allow researchers to optimize the electrical properties of polymer nanocomposites of fiL/D particles.