Nano-Electronics (what is it and why is it important)

The tremendous impact of information technology (IT) has become possible because of the progressive downscaling of integrated circuits and storage devices. IT is a key driver of todays information society which is rapidly penetrating into all corners of our daily life. Integrated circuits are also known as microelectronics. The term micro derives from microfabrication technology, which embraces all highly sophisticated techniques like optical- and electron-beam lithography, metallization, implantation and etching that allow to generate structures on the scale of one micrometer.

What is "Nano" going to change? The answer appears to be straightforward: microelectronics will gradually evolve into nanoelectronics. In fact, this has already happened as can be seen from the smallest feature size of todays integrated circuits which is below of one micrometer. There is no fundamental reason why the continuous size reduction should stop in the next couple of years. It is very likely that it will continue for the next decade. Although, each time the structure size is reduced, technical problems -- often very demanding ones -- arise, but conceptually not much happens. The electronics will be based on silicon and functions the way it functions today. The rule of the game is smaller, faster and better. One must not be an expert to immediately see that this "straightforward" downscaling must sooner or later stop. Will it be when we reach single atoms, small molecules or earlier at the level of supramolecular structures? We do not know yet for sure, but we know that the basic operation principles of todays electronics cannot be scaled below approximately 10 nm. Charge carriers in semiconductors are due to doping. Even if we would make use of highly doped material alone, only by chance could we find a single dopant in a 10 nm-sized transistor. Naturally, we will be dealing with a very few numbers of charge carriers, if at all, and control of charge and electrical current on a single electron level will be required. Moreover, quantum phenomena will increasingly start to dominate the overall behaviour of such structures. Finally, tinny structures have a large surface-to-volume ratio which is deadly for conventional semiconductor devices. Taking these facts, it is for certain that new concepts need to be developed!

Two possible approaches for obtaining structures within the size-regime of 1-100 nanometers are generally discussed. The first one is the top-down approach based on lithography, which is currently used to fabricate integrated circuits. It has been highly sucessful until now, but lacks control on the single atom level. The second one is the bottom-up (or chemist's) approach in which complex structures are assembled from single atoms and molecules into supramolecular structures. This supramolecular chemistry embodies single atom precision, at least in principle. The disadvantage is that we have no idea yet on how to assemble these molecules into useful operational devices. Though these two possible routes were known, most researchers in the field of nanoelectronics enjoyed their own training in traditional top-down technologies and no paradigm change was in sight.

Nanoelectronics (the future)

Nanoelecetronics has witnessed a shift towards molecular systems in recent years. Though the term molecular electronics is a pretty old one, it is only since very recent that single molecules have become the focus of interest. To a great extent this was triggered by research on carbon nanotubes. Before the carbon nanotubes entered the scene, molecular electronics was the science of organic polymers, their synthesis, processing and doping. With carbon naotubes, we finally have a model system at hand that is equally of interest for chemists, material scientists and physicists. While carbon nanotubes are supramolecular objects for a chemist, they are one-dimensional solids for a physicist. In the future, more of these supramolecular structure will be studied on a single molecule level. Companies like, for example Motorola, IBM and Hewlett Packard, are starting to take an active part in this development. Single supramolecular structures are invisaged as switches and storage media. As has been shown already with DNA molecules, the trend towards molecule will include biological macromolecules as well. The ability to manipulate and characterize single molecules is an important first step for the exploration of suitable molecular functions. A fully functional chip, however, requires the ability to assemble the moelcules with high prescission into a functional network.

Quantum computing and quantum communication is another development that has recently taken place. Until now, these computers only exist as proposals on paper, but the theory has already been pushed very far. We know today that a quantum computer will outperform any classical computer. We are now faced with the challenge to realize a demonstrator of such a machine, if possible one that can be scaled up using integration and mass fabrication. A new territory has been discovered for the science of micro- and nanoelectronics. Unlike classical computers, in which bits are either one or zero, quantum computers make use of coherent superpositions of "zero" and "one". In contrast to a classical "switch", the quantum switch can be both open and close, because it may be in a coherent superposition of the two possibilities. The quantum computer derives its power from this unambiguity. However, this peculiar feature is lost after some time by interaction with the environment. This so-called "dephasing" time determines the number of operations that can be performed by a quantum computer before it has to be "reset". We have to learn how to manipulate coherent superpositions in the solid and to understand the limiting factors leading to depasing.

Finally, there is a third direction in nanoelectronics which will receive more attention in the future. This new field is called "spintronics". Spintronics is concerned with electromagnetic effects in nanostructures and molecules caused by the quantized angular momentum (the spin) that is asscociated with all fundamental particles like, for example, the electron. The magnetic moment of a particle is directly proportional to its spin. Hence, if we learn to manipulate not only charge, but also spin on a single electron level, information may be stored and transported in the form of quantized units of magnetism in the future.

part of a text written for the closing booklet of the research program NFP36 on Nanoscience by Christian Schönenberger, June 2000