Quantitative measurement of doping and surface charge in a ZnO nanowire using in‐situ biasing and off‐axis electron holography

Abstract

Semiconducting nanowires (NWs) are widely studied because the properties that stem from their three‐dimensional, nanoscale nature open new opportunities for device design. In particular ZnO NWs are widely studied for their interesting piezoelectronic properties. Though NWs can be readily grown today with increased carrier concentration due to doping, the measurement of the doping concentration at the nm scale remains challenging. We demonstrate that state‐of‐the‐art off‐axis electron holography in combination with electrical in‐situ biasing can be used to detect active dopants and surface charges quantitatively in ZnO nanowires. The outline of the contacted NW is described in Fig. 1. We have acquired series of holograms and averaged the phase to increase signal to noise but avoid blurring due to specimen drift. The 0V bias images were used to remove contrast not related to the varying electrostatic potential and to verify the nanowire was not electrically modified during the experiment. We analyzed the depletion width in the nanowire due to an applied reverse bias to a Schottky contact on the nanowire, using a fit to the data. Comparison of the experimental data with 3D simulations that were similarly treated indicates an n ‐type doping level of 1x10 18 at. cm −3 and a negative surface charge around −2.5x10 12 charges cm −2 . Fig. 2a shows the experimental vacuum corrected phase profiles converted to potential, and a fit to the data. The extracted depletion width is indicated with a cross. The inset shows the location of the phase profile in the NW core and two symmetrically defined phase profiles obtained in vacuum on either side of the NW. The average signal in vacuum was subtracted from the NW signal and the remaining phase signal was converted to potential using a thickness of 75 nm, much smaller than the 150 nm NW diameter. In Fig. 2b the experimental and simulated depletion length is compared for 3D simulations including varying doping and surface charge quantities. We expect that the real doping is between 1 and 2x10 18 at. cm −3 by comparison of experiment and simulation. The surface charge results in a surface depletion to a depth of 36 nm. We found an active/undepleted core thickness of 70–75 nm, providing excellent agreement between the simulated thickness of the undepleted core and the active thickness observed in the experimental data. Off‐axis electron holography thus offers unique capabilities for quantitative analysis of active dopant concentrations and surface charges in nanostructures with nanometer‐scale spatial resolution.