Electron field emission in nanostructures: A first-principles study.

摘要

The objective of this work was to study electron field emission from several nanostructures using a first-principles framework. The systems studied were carbon nanowires, graphene nanoribbons, and nanotubes of varying composition. These particular structures were chosen because they have recently been identified as showing novel physical phenomena, as well as having tremendous industrial applications. We examined the field emission under a variety of conditions, including laser illumination and the presence of adsorbates. The goal was to explore how these conditions affect the field emission performance.;In addition to the calculations, this dissertation has presented computational developments by the author that allowed these demanding calculations to be performed. There are many possible choices for basis when performing an electronic structure calculation. Examples are plane waves, atomic orbitals, and real-space grids. The best choice of basis depends on the structure of the system being analyzed and the physical processes involved (e.g., laser illumination). For this reason, it was important to conduct rigorous tests of basis set performance, in terms of accuracy and computational efficiency. There are no existing benchmark calculations for field emission, but transport calculations for nanostructures are similar, and so provide a useful reference for evaluating the performance of various basis sets. Based on the results, for the purposes of studying a non-periodic nanostructure under field emission conditions, we decided to use a real-space grid basis which incorporates the Lagrange function approach.;Once a basis was chosen, in this case a real-space grid, the issue of boundary conditions arose. The problem is that with a non-periodic system, field emitted electron density can experience non-physical reflections from the boundaries of the calculation volume, leading to inaccuracies. To prevent this issue, we used complex absorbing potentials (CAPs) to absorb electron density before it could reach the boundaries and reflect. The CAPs were zero in the region of the emitting structure and in regions were measurements were made. Still, we wanted to make sure that complex potentials were an accurate and efficient solution to the boundary problem. To evaluate this, we once again turned to benchmarks from transport calculations. The results showed CAPs to be an extremely accurate and efficient computational tool, and were incorporated into all of this dissertation's calculations.;Our calculations show that adsorbate atoms significantly increase the field emission current of carbon nanotubes. Adsorbate atoms also cause strong differences and nonlinearity in the Fowler-Nordheim plot. The source of the enhanced current is electronic states introduced by the adsorbate atom.;We have investigated the emission of electrons from nanostructures induced by short intense laser pulses in the presence of a weaker uniform static field. Based on the results of our simulations, two important qualitative features of this process have been determined: (1) a significant enhancement of the emission when a laser pulse is applied, and (2) the field emission current has a peak of some duration and the position of this peak correlates with the time of the pulse arrival. These two features suggest the possibility of using short laser pulses for making few-electron emitters of nanoscale size [212]. Such emitters could have many desirable properties, especially very high spatial and time resolutions.;The field emission from nanotubes of various composition has been studied. The calculations predict that the GaN, SiC, and Si nanotubes are particularly good field emitters. The highest-current nanotube, Si, is predicted to produce a current an order of magnitude higher than BN or C nanotubes.;The calculations predict that carbon nanotubes with various adsorbates can be used as spin-polarized current sources. These are the first first-principles calculations to show spin-polarized field emission for carbon nanotubes with iron adsorbates. Also, there is currently no experimental data with which to compare these results. However, the predicted strong spin-polarized field emission from these structures will hopefully spur experimental tests, potentially leading to applications within spintronics.;We also studied field emission from carbon nanowires and graphene nanoribbons. Since there is not yet any experimental work on field emission from isolated carbon nanowires or graphene nanoribbons, we cannot compare these results to experiment. These results are still useful in demonstrating the ability of our framework to study non-nanotube structures.;In addition, our nanoribbon results make interesting predictions of high currents. The currents increase by several orders of magnitude when increasing the applied electric field. This is a much greater response than with the nanotube results. The time to reach a relatively stable current is also much shorter. These features suggest that these nanoribbons could be extremely good field emitters, and experiments are needed to further explore this.