Abstract :

Silicon nanonets are networks of randomly oriented silicon nanowires. Due to its flexibility, transparency and reproducibility, this material is highly attractive as an alternative to amorphous silicon or organic materials for various macroelectronic applications involving sensors and displays. Based on our original integration process simply relying on standard photolithography, we were already successful in demonstrating workability, reproducibility and excellent air stability along with interesting performance for device channel length ranging from the micrometer to the millimeter. Rigid transistors with millimeter channel length exhibit outstanding performances with high drain current up to 10-7A, IOn/IOff ratio as large as 105 and good mobility (0.004 m2V-1s-1) as compared to a-Si and organic materials. This presentation focuses on why NN-based devices are working well against all expectations and details the path to nice devices on flexible substrates. At first, results show the role played by the nanowire/nanowire junctions in the electrical properties of semiconductor percolating nanowire network by comparing transistors made by single nanowire, dual nanowire with one junction and network of nanowire. It shows the electrical characteristics of networks are strongly affected and enhanced by the nanowire junctions despite the disorder introduced by the increasing number of nanowires. Second, we demonstrate the robustness of the integration process and its nice adaptation for producing flexible resistors and transistors made of nanostructured material and using only standard microelectronic technology. As a result, the study of electrical performance under bending, with curvature radius in the 7-24mm range, evidences a crucial change in current for longer channel devices (>200µm) but stability for the shorter ones (<200µm). Thus, by choosing correctly the device geometry, we demonstrate that silicon nanonet is a suitable candidate for long-term electromechanical stable flexible devices (shorter devices) and for bending and pressure sensor (longer devices), allowing then to combine various geometry on one chip to produce simultaneously the sensors and the reading electronics, all based on Si nanonets.

Contact : annie.ducher@grenoble-inp.fr