Hysteresis, Rectification, and Relaxation Times of Nanofluidic Pores for Neuromorphic Circuit Applications

Abstract

[EN] Based on the emergence of iontronic fluidic components for brain-inspired computation, the general dynamical behavior of nanopore channels is discussed. The main memory effects of fluidic nanopores are obtained by the combination of rectification and hysteresis. Rectification is imparted by an intrinsic charge asymmetry that affects the ionic current across the nanopores. It is accurately described by a background conductivity and a higher conduction branch that is activated by a state variable. Hysteresis produces self-crossing diagrams, in which the high current side shows inductive hysteresis, and the low current side presents capacitive hysteresis. These properties are well captured by measurements of impedance spectroscopy that show the correspondent spectra in each voltage wing. The detailed properties of hysteresis and transient response are determined by the relaxation time of the gating variable, that is inspired in the Hodgkin-Huxley neuron model. The classification of effects based on simple models provides a general guidance of the prospective application of artificial nanopore channels in neuromorphic computation according to the measurement of complementary techniques.; Rectifying fluidic nanochannels are attractive for neuromorphic computation with fully liquid circuits. A model is presented that explains the measured properties of hysteresis, transient response, and synapse volatility, necessary for neuromorphic operation in the context of multichannel devices. image