For the past few decades, transistors have been continuously scaled. Dimensions are now at the nanoscale, and device performance has dramatically improved. Nanotechnology is also achieving breakthroughs in thermoelectrics, which have suffered from low efficiencies for decades. As the device scale enters the nanometer range and novel device structures are introduced, it becomes essential to revisit the device physics and develop new simulation frameworks to analyze the experimental data and project the device performance. In this thesis, a comprehensive theoretical study of nanoscale electronic and thermoelectric (TE) devices is presented. First, a ballistic one-dimensional (1D) MOSFET is explored to study the characteristic features of 1D transport compared with two-dimensional (2D) ballistic MOSFETs. Carrier scattering physics is compared across 1D and 2D transistors using Monte Carlo simulations and analytical derivations. It is shown that although the scattering physics is very different, the overall backscattering characteristics are surprisingly similar. For TE devices, performance is compared across 1D, 2D, and three-dimensional (3D) structures using the Landauer formalism, which quantifies possible advantages obtained by reducing dimensionality. It is found that the potential benefits of engineering the mode density are modest. The concepts of thermionic emission cooling and role of momentum conservation are also examined. A self-consistent electro-thermal transport simulation framework using the non-equilibrium Green's function (NEGF) method is developed to explore the TE properties of nanocomposite materials. We study the Seebeck coefficient of a 1D diffusive composite structure and relate it to the TE performance. We also explore the energy filtering effects in composite structures and show that they have the potential to improve the power factor.
Ra Seong Kim received her PhD in August 2011.
Cite this work
Researchers should cite this work as follows:
Purdue University, West Lafayette, IN
- Ph.D. thesis
- device physics
- Monte Carlo simulation
- NEEDS node
- device simulation