| Abstract
|
:
|
Currently single crystal Si dominates the high performance (∼ GHz), low footprint micro-electronics applications such as computing and communication, while amorphous Si (a-Si) dominates the lower performance (∼ 1-10 kHz), large area macro-electronics applications such as displays, large area sensors, etc. However, as the transistor scaling slows over a period of next 10 years or so (∼ 2020), new technologies for classical applications like active matrix flat-panel displays, solar panels, and novel applications like biochemical sensors, electronic paper, flexible and wearable electronics may drive the future semiconductor industry. These applications need higher performance than a-Si and require lower cost, lower temperature (< 250 C) manufacturing processes for flexible substrates like plastic. Nanonet thin film transistors (NN-TFTs), which consist of networks of carbon nanotubes (CNTs) or Si nanowires (SiNWs) as channel materials, are emerging as a promising alternative to a-Si technology. In addition to applications in electronics, recently CNT nanonets are also emerging as highly transparent and flexible alternative for the conventional ITO electrode in photovoltaic applications. Here the density and alignment of tubes is a crucial optimization criterion to obtain the best combination of transparency and conductivity. Although recent research activities have demonstrated the versatility of these TFTs and PV cells, there has been little theoretical modeling effort to systematically organize and interpret the large amounts of experiments. The 2D stick percolation theory needs to be generalized and several key aspects like high voltage behavior, short channel characteristics, 3D electrostatics, co-percolation of metallic/semiconducting subnetworks, optical/sensing properties, etc. are yet to be successfully modeled within this theoretical framework. Here, we develop physics-based, predictive models for Nanonet-TFTs using a 'bottom-up' approach based on nonlinear percolation theory.
|