Influence of slow or fast surface traps on the amplitude and symmetry of the piezoelectric response of semiconducting-nanowire-based transducers

ZnO nanowires are excellent candidates for energy harvesters, mechanical sensors, piezotronic and piezophototronic devices. These nanowires are usually non-intentionally n-doped during their growth. The essential role of doping, surface traps and surface Fermi level pinning in the actual response of piezoelectric semiconductors has already been demonstrated. In order to go further, this paper investigates the influence of the density and of the dynamics of surface traps on such important parameters as the output generated potential and the effective piezoelectric coefficient. We implemented numerical simulations based on the finite element method by combining the mechanical, piezoelectric, and semiconducting characteristic of ZnO nanowires array based nanocomposites (the so-called vertically integrated nanogenerator configuration) operated in compression. It was found that a certain amount of surface traps was required to obtain a usable generated output potential from the studied devices in the range of dimensions and doping level reported in most experimental results. Moreover, the surface traps influence was strongly dependent on their dynamics. As a first step towards the analysis of traps dynamics, we compared the two extreme cases of ultra-slow and ultra-fast traps. The symmetry and asymmetry of the piezoelectric response and a comparison to thin film was also discussed. This study demonstrates that the realistic modelling of the piezoelectric response of semiconductor based transducers should account for traps dynamics effects.

Automated summary

This study investigates the influence of the density and dynamics of surface traps on the output potential and effective piezoelectric coefficient of ZnO nanowire array based nanocomposites. It is found that a certain number of surface traps are necessary to achieve usable output potential, and their influence is strongly dependent on their dynamics.