0 avis
Development of 3D organic polymer dendrites as neuromorphic device
Archive ouverte : Communication dans un congrès
Edité par HAL CCSD
Oral. International audience. During recent years, neuromorphic engineering has gained significant attraction due to its endeavor to reproduce attractive brain features such as high computational efficiency, low power consumption, functional and structural adaptability onto hardware devices for competitive computing development. Currently, most of these implementations are achieved with either standard silicon-based technologies (complementary metal–oxide–semiconductor) or more emerging material and devices (iono-electronic materials and memristor devices). Mainly, these technologies are produced by means of top-down approaches. In contrast, brain computing largely relies on self-assembling processes to interconnect cells and form pathways for neural networks communication. Thus, to benefit more from neuromorphic features, there is a strong need to explore such devices that can rely on the same bottom-up approach as in the brain. Promising solutions for this question are gathered in organic electronic materials, more precisely in organic mixed ionic electronic conductors. This study is focused on the development of bottom-up fashioned 3D organic devices that are able to mimic biological neural network branching. By employing Alternating-Current (AC) bipolar electropolymerization, we show how one can synthesize polymer dendritic structures and tune its morphologies depending on various applied AC signals . Additionally, we show how dendritic devices may exhibit synaptic plasticity properties (short-term and long-term memory effect). Finally, we demonstrate implementation of structural plasticity through spike-event activation AC-electropolymerization and the possibility to modify the weight of obtained dendritic connections with respect to spike rate intensity of the applied signals.
Consulter en ligne
Suggestions
Du même auteur
Du même sujet
