The physical limits of electronic circuit integration may be overcome by assembling elementary processors at the atomic or molecular scale, that operate in parallel and interact only with their nearest neighbours. New strategies and technologies are therefore necessary in order to satisfy the ever increasing request of higher performance from information processing systems. A first requirement, borrowed from biological paradigms, is that a molecular circuit can be obtained by assembling elementary processors made from nanometer size particles, molecules or atoms. Such molecular systems are noteworthy as they accomplish in the physical world the concept of cell automaton, i.e. neural networks at short range of interaction, and hence they are potential candidates for molecular information processing.
In this framework, we consider a system where nanometer-size gold particles, thought as elementary cells, are embedded within an organic dielectric medium in order to obtain an ultrathin mixed film. This film lies onto a conductive substrate in such a way that organic molecules constituting the medium behave as an insulating barrier between adjacent conductive particles or between a particle and the substrate; in this system an electron may jump by tunnel effect either from one particle to another one or to the substrate.
Different methods have been taken into account to build up this system.
A first approach is to assemble the gold nanoparticles taking advantage of the self-ordering properties of Langmuir-Blodgett films. Dipalmitoyl-Phosphatidyl-Ethanolamine (DPPE) molecules, functionalised with gold atomic clusters, are mixed in suitable proportions with cadmium arachidate and alkanes, and hence deposited onto the substrate.
Conformational properties of films obtained on different substrates, e.g. mica, glass and silicon, have been investigated by means of scanning probe microscopy.
Another approach is to exploit the self-assembling properties of alkyldithiol molecules onto gold monocristalline terraces; by choosing an appropriate length of the alkyl chain, the molecules can assemble having one end onto the gold and the other end capable of covalently binding to gold nanoparticles dissolved in solution and brought in contact.
The choice of the proposed general structure appears quite promising since it is characterised by an affordable technological complexity and it is part of the wider class of metallic nanometer size particles systems. Moreover, it harmonises with the recent rise of interest for single electron devices and memories, even operating at room temperature.
Acknowledgement: work supported by the National Research Program "Langmuir-Blodgett quantum superlattice for parallel information processing"