Cellular Automation on an organic monolayer

Using molecular neurons we have grown bilayer on atomic flat metal surface. On this monolayer we have written molecular conducting states and patterns in terms of conducting states.

Then we have analysed how this pattern evolves on the surface, identified particular rules. Most importantly we have encoded these rules in a program so that one can mimic any operation taking place on the surface.

We have named this program "Bramha", whose primitive version one can download freely from this website. We wish to make this software perfect, but there may be couple of bugs, so users are requested to give us feedback on this bio-processor software.

See Movie: Blinking of bits on the organic monolayer in real time as seen in STM

Understanding ripples around molecule for information transport

While investigating large number of molecules, we have found that the surface of the neighborhood, changes very interestingly for the multi-level switching molecules.

We have found that charge symmetry and conformation, three very important parameters of a molecule could be determined from the STM images directly.

Currently, we are increasing our statistical database and developing fundamental theoretical formulation for this invention.

Swarm intelligence in molecular electronics

Schools of fish, flock of birds, together in a team exhibit an excellent example of intelligence, in spite of the fact that alone they can not produce intelligence.

Several research groups are already carrying out experiments with tiny little robots which can form unique architectures for each set of rules.

Using multilevel molecular switches that moves on the surface following particular modes of oscillation, we have generated particular set of architectures, similar to swarms on a surface.

See Movie: Atomic scale motion of molecules

Room temperature Kondo effect

To build neural network zero bias conductance based devices could be a very possible alternative. Cellular neural network and cellular automata scientists have long shout for such devices.

We have determined a way to regulate spin in a molecule so that we can realise zero bias conductance at room temperature.

Currently, we are generating artificial neural network using this molecule.

Standalone Vertical Parallel Processor or 3 D nano brain

Our current 2 D molecular assembly is an architecture realised in an ultra high vacuum chamber. To use this assembly in an ambient atmosphere we need to convert this 2 D assembly into a 3 D architecture.

However, even though necessity is clear but not the way to do it. We have successfuly resolved all critical design issues related to such 3 D assembly.

There is one additional problem. It may be easy for a chemist to synthesize such an assembly, however, it is an extremely challenging task to interface multiple electrodes with this 20 nm device at the maximum scale.

We are working on modifying this architecture such that this device dimension increases to 20 micrometer scale, or of the order of a single cell in size.

See Movie 3 D nano brain in the making

Cellular neural network in cross-bar architecture

In the cross-bar architecture we fabricate cells, that can process more than one conducting states. Due to an inherent fluctuation or noise most of the cells in a cross bar architecture are not able to survive their own state. In the classical definition of a memory chip this is simply a mal-manufactured chip.

However, the complete architecture may behave as an integrated system and potential profile on the complete cross-bar assembly may vary following particular rules. These rules could be used particularly for the cellular neural network based computations.

The device architecture has 25-50 nm size cells with particular benzoquinone derivative molecules

Cellular neural network in spherical miceller assembly

We use 2 D surface assembly of 100 nm miceller fluorescent balls which can switch between four colours, red, green, yellow, and blue. Therefore on a surface they can change colors depending on interaction between balls. We attempt to take series of optical images of these surfaces and try to identify generalised equation expressing next state of a cell.

Alternately we also study possible cellular automata rules for expressing the logical pattern evolution on the surface.

Finally we try to build an optical bio-processor.

Contact: Anirban Bandyopadhyay, National Institute for Materials Science, Tsukuba, Japan-305-0037