An elaborate response in molecular random network -Approach for information processing functionality-
Megumi Akai-Kasaya
An animate brain consists of interconnected complex biological networks of locally active components. The informational processing exploits massive parallelism with low energy consumption, utilizing the multiplicity and stochasticity of signals. Recently, utilization of desiglees random networks of inanimate matter for computation or informational processing has started to be proposed. Here we introduce two nanomaterial networks, which show complex response showing potency for advanced future signal processing.
Abstract
We succeeded in showing experimental evidence to prove Coulomb blockade taking place on Two-dimensional organic conducting polymer films. Electrical conductivity of hexylthiophene monolayer showed temperature-dependent threshold voltage in temperature range of 150K—4K and increased in power law of the current-voltage beyond the threshold. The onset of two-dimensional Coulomb blockade in the organic thin film was theoretically verified through the calculation of the delocalization of electric charge in the molecular film. This system meets the criteria to be a suitable building block for the physical realization of a cellular neural network. We further found a neuromorphic device that consist of extremely dense and random single-walled carbon nanotube (SWNT) complexed with polyoxometalate molecule generating spontaneous spikes similar to nerve impulses of neurons. A simulation calculation of the random molecular network model, which are able to store electric charges, replicate spikes generated from the network. This molecular model would very likely become a component of reservoir computing that is anticipated as next-generation artificial intelligence. These results indicate the possibility that complex functional networks can be constructed using molecular devices, and contribute to the development of neuromorphic devices.
Keywords: Nanomaterial Molecular Device Neuromorphic Neural Network
Topic: Information Technology