Multiscale modeling and simulation of molecular devices and systems

点击翻译按钮获取中文摘要

Microelectronic integrated circuits are probably the most complex artificial systems ever created. Their `success story¿ is based on two facts: one is technological and the other one is designrelated. The technological base is commonplace: since electronic components and their interconnections are scaled down to almost atomic dimensions and it is possible to squeeze billions of devices on a single wafer, one can produce a very large (and complex) system for a reasonable price. The design- related base is equally important but often overlooked: this is the hierarchical, modular engineering approach. Circuit models of electronic devices are based on physical models, and circuit models are the foundations of architectural concepts. This clear separation of the levels of description facilitates the design of complex systems.It is now confirmed that single-molecule devices show electrical characteristics that could make them useful electronic components and a large amount of research is dedicated to the assembly and integration of molecules into circuit-like arrangements. On the other hand, no viable design methodology emerged so far for molecular circuits.The goal of our proposed research is to develop a multiscale modelling approach to the description of molecular systems and immediately apply it to the simulation of novel, promising computing architectures.Using Density Functional approaches for the molecular system (nanoscale level) and drift-diffusion and Monte Carlo tools for the whole devices (microscale level) we will investigate the behaviour of single-molecule devices in their environment, thus including the effect of contacts, electromagnetic interaction among molecular devices, thermal dissipation, fluctuations. From this, we will extract the required parameters to describe the system for an architectural point of view (macroscale level). In the architecture design we will concentrate on two promising molecular-scale architectures: crossbar-type circuits and field-coupled architectures. A large emphasis will be put on the interface of the molecules and the input / output circuitry. The key questions of molecular electronics whether molecular entities can be integrated in complex circuitry, like their electronic counterparts, and if the answer is yes, what are the most suitable architecture concepts for doing this. Our research should bring us closer to the answer of this problem.