Electrical percolation in metal wire network based strain sensors

Abstract

Metal wire networks rely on percolation paths for electrical conduction, and by suitably introducing break-make junctions on a flexible platform, a network can be made to serve as a resistive strain sensor. Several experimental designs have been proposed using networks made of silver nanowires, carbon nanotubes and metal meshes with high sensitivities. However, there is limited theoretical understanding; the reported studies have taken the numerical approach and only consider rearrangement of nanowires with strain, while the critical break-make property of the sensor observed experimentally has largely been ignored. Herein, we propose a generic geometrical based model and study distortion, including the break-make aspect, and change in electrical percolation of the network on applying strain. The result shows that when a given strain is applied, wire segments below a critical angle with respect to the applied strain direction end up breaking, leading to increased resistance of the network. The percolation shows interesting attributes; the calculated resistance increases linearly in the beginning and at a higher rate for higher strains, consistent with the experimental findings. In a real scenario, the strain direction need not necessarily be in the direction of measurement, and therefore, strain value and its direction both are incorporated into the treatment. The study reveals interesting anisotropic conduction features; strain sensitivity is higher parallel to the strain, while strain range is wider for perpendicular measurement. The percolation is also investigated on direct microscopic images of metal networks to obtain resistance-strain characteristics and identification of current percolation pathways. The findings will be important for electrical percolation in general, particularly in predicting characteristics and improvising metal network-based strain sensors.