The concomitant and concurrent presence of cellular mechanisms, such as the activity-dependent Long Term Potentiation (LTP) and Depression (LTD), and the Short Term Plasticity (STP) is believed to be at the basis of encoding memories in brains. Thus, the best possible emulation of these fundamental brain activities is essential for developing bioinspired circuits and novel adaptive technologies, and, looking in perspective, an efficient brain-machine interface. In this framework, memristive devices are increasingly considered as key elements in the emerging fields of neuromorphic engineering and computing because of their synaptic-like plasticity properties. Here we demonstrate the realization of organic memristive devices (OMDs), based on polyaniline (PANI), mimicking controlled neuromorphic functions. In particular our OMDs exhibit LTD or LTP by varying the polarity or, more interestingly, by varying the frequency of the incoming stimuli, according to the typical biological patterns. In both cases, OMDs show also an effect analogous to the transition between short term to long term memory, as a function of the total number of received pulses. We validate that organic memristors represent an important step toward an "intelligent" neuroprosthetics demonstrating through the hardware implementation of OMDs neuromorphic functionalities, the possibility of pairing the frequency dependence of synaptic signals with the non-volatile evolution of the internal memory states of the device.
Frequency driven organic memristive devices for neuromorphic short term and long term plasticity
Battistoni S;Erokhin V;Iannotta S
2019
Abstract
The concomitant and concurrent presence of cellular mechanisms, such as the activity-dependent Long Term Potentiation (LTP) and Depression (LTD), and the Short Term Plasticity (STP) is believed to be at the basis of encoding memories in brains. Thus, the best possible emulation of these fundamental brain activities is essential for developing bioinspired circuits and novel adaptive technologies, and, looking in perspective, an efficient brain-machine interface. In this framework, memristive devices are increasingly considered as key elements in the emerging fields of neuromorphic engineering and computing because of their synaptic-like plasticity properties. Here we demonstrate the realization of organic memristive devices (OMDs), based on polyaniline (PANI), mimicking controlled neuromorphic functions. In particular our OMDs exhibit LTD or LTP by varying the polarity or, more interestingly, by varying the frequency of the incoming stimuli, according to the typical biological patterns. In both cases, OMDs show also an effect analogous to the transition between short term to long term memory, as a function of the total number of received pulses. We validate that organic memristors represent an important step toward an "intelligent" neuroprosthetics demonstrating through the hardware implementation of OMDs neuromorphic functionalities, the possibility of pairing the frequency dependence of synaptic signals with the non-volatile evolution of the internal memory states of the device.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.