Materials for Tomorrow

An introduction to the Molecular Dynamics (MD) simulation of chemically realistic models for undercooled fluids and glasses is given, emphasizing silicatic materials such as molten silicon dioxide and its mixtures with sodium oxide and aluminium oxide, and comparing the simulation results to experimental data whenever possible.

A key ingredient to the computer simulation of materials is a sufficiently accurate description of the force fields with which the atoms interact. The need to simulate large systems for sufficiently long times makes the use of effective potentials for classical MD methods desirable. The validation of such effective potentials is best done studying the corresponding crystalline states of the material. As an example, the use of the so-called BKS-potential studying the structural and thermal properties of quartz crystals is described, and a comparison to other potentials is discussed.

When one studies undercooled fluids and glasses, a second problem enters, the disparity between experimental cooling rates and the much larger rates of the simulation. The extent to which a meaningful comparison to experiments is nevertheless possible is discussed. It is shown that the simulations can reproduce the structural and dynamic properties of molten silica (including self diffusion coefficients, viscosity, sound velocity, etc.) and its mixtures with other oxides. Evidence for the formation of sodium-rich channels responsible for anomalously large diffusion constants of sodium in mixtures containing sodium oxide will be discussed. No enhanced diffusion of aluminium occurs, however, due to “tricluster” formation in such mixtures.

Motivated by the high application potential of carbon nanotubes, the search for other quasi one-dimensional nanostructures has been pursued both by theoretical and experimental approaches. The investigations soon concentrated on layered inorganic materials, which may be exfoliated and rolled up to tubular and scroll-type forms. The present chapter reviews the basic design principles, which govern the search for novel inorganic nanostructures on the basis of energy- and strain-related stability criteria. These principles are then applied to the prediction and characterisation of the properties of non-carbon, elemental and binary nanotubes derived from layered boride, nitride, and sulfide bulk phases. Finally, the present chapter introduces examples, where one-dimensional nanostructures such as tubes and scrolls have successfully been constructed from non-layered materials, especially from oxides. Examples for the experimental verification of the predicted structures are given throughout the discussion and impressively underline the predictive power of today’s materials modelling.

Nanotechnology plays a decisive role in information technology. However the rapid increase (doubling of the Internet traffic every 6 months, of the wireless capacities every 9 months and of the magnetic information storage every 15 months) cannot be compensated by a simple downscaling of the semiconductor devices, as it was done in the past 30 years. To keep up with the demands, completely new devices have to be invented, operating on the nanoscale and exploit quantum effects. A very promising option is to use the spin of the electron in addition to its charge for information transmission and storage, i.e. going from the conventional electronics to spintronics. The foundations of this technique and the broadest application areas today, exploiting the giant magnetoresistance and the tunneling magnetoresistance are discussed from the experimental and theoretical point of view.

The proper treatment of defects is one of the major tasks in materials design, because defects are responsible for the either desirable or detrimental deviations between the characteristics of the material to be tuned and the well-known properties of an ideal crystal. Microelectronic devices work because of clever point defect engineering, line defects govern plastic deformation processes, and interfaces determine the mechanical stability of composite materials. Especially interfaces gain importance with the current trend towards nanoscale materials; first, the surface-to-volume ratio is strongly increased in nanocrystalline material, and, second, stable arrangements of point or line defects require a minimum crystallite size, which can be larger than the actual nanocrystallites. Thus, the present chapter gives an introduction into the most common approaches for modeling interface properties. We introduce the basic concepts of interface symmetry, structure and analysis with a strong focus on the theoretical methods and give an overview of currently available techniques for the modeling and simulation of the interface properties at an atomic-scale level. Two fundamentally different interface types are distinguished: The discussion of the homophase boundary properties is focussed on oxide grain boundaries, which we studied extensively in comparison with amply available experimental observations. For the heterophase boundaries examples of non-reactive, reactively doped, and inherently reactive boundaries are presented. A special focus lies on the interfaces between metals and oxides where the discrepancy of the material properties across the interface is most prominent and all three bonding situations can occur: weak adhesion between inert fragments, activated adhesion upon doping, and strong adhesion.

In this chapter, we discuss the physical models that are commonly used for the quantum simulation of electronic states and currents in nanostructures. Since most of these structures are too large for an atomistic description, we focus here on continuum models with empirically adjusted material parameters. In specific, after a short introduction into the band structure theory of crystalline solids, we first present the k·p-equations for semiconductors. Next, we discuss the envelope function approximation for heterostructures and consider the effects of elastic deformations and strain. Furthermore, we also examine carrier densities at non-zero temperature and consider the interplay of the Poisson and Schrödinger equation. After describing the electronic structure, we now discuss the Boltzmann equation and the numerically more tractable drift-diffusion equations as semi-classical models for carrier transport in semiconductors. We then extend the drift-diffusion model to take quantum corrections for size quantization into account, and we outline the principles of ballistic quantum transport. Finally, we present nextnano ³, a software package for the simulation of nanostructures that has been developed by the authors, and give an example application.

Metallic nanocrystals have numerous technical applications, especially because of their favourable magnetic and mechanical properties, surpassing often by far those of the corresponding polycrystals. As applies to most of the nano-materials, their importance in applications will grow further and wider in the years to come.

Because of their small extension of only a few nm in all or some of their dimensions, the properties of nanocrystals are generally characterized by two main facts: a competition of length scales due to the fact that their extension is smaller than some physical relevant length scale like e.g. the correlation length, screening length, free path length. For small nanocrystals, the network of grain boundaries between the crystallites can occupy up to 50% of the sample volume and thus there is a strong influence of the network of grain boundaries on the overall properties of nanocrystalline materials.

After an introduction into the production and characterization methods of metallic nanocrystals and some characteristic properties of their grain boundary network, in this lecture the consequences of both of the two characteristic facts just mentioned for some of the properties of (mainly fcc) metallic nanocrystals are discussed: the melting temperature, the magnetic and the mechanical properties and finally in detail their atomic dynamics.