Our computational demonstration of synaptic plasticity in a 2D porous membrane is now available online! We are deeply grateful for this achievement.

The human brain uses a variety of ions and molecules to perform its sophisticated functions in an energy-efficient manner. In contrast, state-of-the-art neuromorphic technology relies on solid-state memristors that utilize a single type of charge carrier—the electron. While the importance of ionic diversity in brain function is well acknowledged, our ability to harness it to implement artificial neuromorphic functionalities remains a key open question. Our study presents the first computational demonstration of synaptic-like plasticity in two-dimensional porous membranes that directly arises from multi-ion effects. In particular, due to the atomic thinness of the membrane, the conductive states are tunable by nanosecond-scale voltage pulses. This results in attojoule (aJ)-scale energy dissipation per spike, which is several orders of magnitude lower than that of a biological synapse (0.1–10 fJ/spike).

Visit Link for the full manuscript.