Synaptic-like plasticity in 2D nanofluidic memristor from competitive bicationic transport

Area of Science:

• Materials Science
• Computational Neuroscience
• Nanotechnology

Background:

• Synaptic plasticity is crucial for learning and memory, underpinning the brain's energy efficiency.
• Biological synapses exhibit remarkable adaptive signal transmission capabilities.
• Understanding and replicating synaptic plasticity in artificial systems is a key goal in neuromorphic engineering.

Purpose of the Study:

• To demonstrate synaptic-like plasticity in a synthetic subnanoporous two-dimensional membrane.
• To investigate the molecular mechanisms and energy efficiency of this artificial plasticity.
• To compare the energy dissipation of the artificial system with biological synapses.

Main Methods:

• Utilized molecular dynamics simulations to model ion transport and membrane property changes.
• Applied voltage spikes to induce and observe changes in ionic permeability.
• Analyzed competitive bicationic transport and its role in plasticity induction.

Main Results:

• Demonstrated repeatable, voltage-spike-induced synaptic-like plasticity in a 2D membrane.
• Observed dynamic modification of ionic permeability via competitive bicationic transport.
• Achieved ultra-low energy dissipation (0.1-100 aJ/spike), significantly lower than biological synapses (0.1-10 fJ/spike).

Conclusions:

• The subnanoporous 2D membrane exhibits functional synaptic-like plasticity.
• The system's atomic thinness and subnanometer pores enable unprecedented energy efficiency.
• This work provides a molecular-level understanding of artificial synaptic plasticity and its potential for energy-efficient computing.