The replacement of the conventional field effect transistor (FET) by single electron transistors (SET) would lead to high energy savings and to devices with significantly longer battery life. There are many production approaches, but mostly for specimens in the laboratory. Most of them suffer from the fact that they either only work at cryogenic temperatures, have a low production yield or are not reproducible and each unit works in a unique way. A room temperature (RT) operating SET can be configured by inserting a small (few nm diameters) Si-Nanocrystal (NC) into a thin (<10 nm) SiO2 interlayer in Si. Industrial production has so far been excluded due to a lack of manufacturing processes. Classical technologies such as lithography fail to produce structures in this small scale. Even electron beam lithography or extreme ultraviolet lithography are far from being able to realize these structures in mass production.
However, self-organization processes enable structures to be produced in any order of magnitude down to atomic sizes. Earlier studies realized similar systems using a layer of Si-NCs to fabricate a non-volatile memory by using the charge of the NCs for data storage. Based on this, it is very promising to use it for the realization of the SET. The self-organization depends only on the start configuration of the system and the boundary conditions during the process. These macroscopic conditions control the self-formed structures. In this work, ion beam irradiation is used to form the initial configuration, and thermal annealing is used to drive self-organization. A Si/SiO2/Si stack is irradiated and transforms the stack into Si/SiOx/Si by ion beam mixing (IBM) of the two Si/SiO2 interfaces. The oxide becomes metastable and the subsequent thermal treatment induces selforganization, which might leave a single Si-NC in the SiO2 layer for a sufficiently small mixing volume. The transformation of the planar SiOx layer (restriction only in one dimension) into a small SiOx volume (restriction in all three dimensions) is done by etching nanopillars with a diameter of less than 10nm. This forms a small SiOx plate embedded between two Si layers. The challenge is to control the self-organization process. In this work, simulation was used to investigate dependencies and parameter optimization.
The ion mixing simulations were performed using binary collision approximation (BCA), followed by kinetic Monte Carlo (KMC) simulations of the decomposition process, which gave good qualitative agreement with the structures observed in related experiments. Quantitatively, however, the BCA simulation seemed to overestimate the mixing effect. This is due to the neglect of the positive entropy of the Si-SiO2 system mixing, i.e. the immiscibility counteracts the collisional mixing. The influence of this mechanism increases with increasing ion fluence. Compared to the combined BCA and KMC simulations, a larger ion mixing fluence has to be applied experimentally to obtain the predicted nanocluster morphology. To model the ion beam mixing of the Si/SiO2 interface, phase field methods have been applied to describe the influence of chemical effects during the irradiation of buried SiO2 layers by 60 keV Si+ ions at RT and thermal annealing at 1050°C. The ballistic collisional mixing was modeled by an approach using Fick’s diffusion equation, and the chemical effects and the annealing were described by the Cahn Hilliard equation. By that, it is now possible to predict composition profiles of Si/SiO2 interfaces during irradiation. The results are in good agreement with the experiment and are used for the predictions of the NCs formation in the nanopillar.
For the thermal treatment model extensions were also necessary. The KMC simulations of Si-SiO2 systems in the past were based on normed time and temperature, so that the diffusion velocity of the components was not considered. However, the diffusion of Si in SiO2 and SiO2 in Si differs by several orders of magnitude. This cannot be neglected in the thermal treatment of the Si/SiO2 interface, because the processes that differ in speed in this order of magnitude are only a few nanometers apart. The KMC method was extended to include the different diffusion coefficients of the Si-SiO2 system. This allows to extensively investigate the influence of the diffusion. The phase diagram over temperature and composition was examined regarding decomposition (nucleation as well as spinodal decomposition) and growing of NCs.
Using the methods and the knowledge gained about the system, basic simulations for the individual NC formation in the nanopillar were carried out. The influence of temperature, diameter, and radiation fluence was discussed in detail on the basis of simulation results.
| Identifer | oai:union.ndltd.org:DRESDEN/oai:qucosa:de:qucosa:70979 |
| Date | 16 June 2020 |
| Creators | Prüfer, Thomas |
| Contributors | Faßbender, Jürgen, Nordlund, Kai, Technische Universität Dresden |
| Source Sets | Hochschulschriftenserver (HSSS) der SLUB Dresden |
| Language | English |
| Detected Language | English |
| Type | info:eu-repo/semantics/publishedVersion, doc-type:doctoralThesis, info:eu-repo/semantics/doctoralThesis, doc-type:Text |
| Rights | info:eu-repo/semantics/openAccess |
Page generated in 0.0022 seconds