The memristive implementation of the hippocampus: A hypothesis

We present a hippocampus-inspired neuromorphic system based on a stochastic poly-crystalline nano-fibre mesh. This system takes advantage of the inherent randomness of the material structure to create a dynamically evolving connectivity pattern that mimics the probabilistic nature of biological synaptic networks. The memristive architecture is based on polycrystaline memristive materials derived from acetylsalycilic acid (aspirin). Three different compounds were obtained by adding axial ligands to [Cu2(asp)4]: benzimidazole (bimi), pyridine (py), and 1,4-diazabicyclo[2.2.2]octane (DABCO). The results showed that the memristive properties of devices based on copper complexes can be modulated with various axial ligands. Moreover, theoretical analysis predicts the possibility of mimicking hippocampal architecture via cocrystallization of hydrophobic organic molecular semiconductors with selected copper complexes. Additionally, we demonstrated that controlling crystallization temperature and solvent composition allows precise tuning of resistive switching in methylammonium lead iodide perovskites. These approaches together provide a versatile foundation for implementing dynamic, brain-like memory functions in hardware.