Today's artificial neural networks face challenges of increasing complexity, ever-larger training data and rapidly increasing energy consumption. While a large number of contemporary networks are based on conventional silicon technology and CMOS-compatible top-down processes, neuron assemblies in the biological context rely on bottom-up network evolution in soft matter and continuous stimulus-dependent optimisation through dynamic reconfigurations, ultimately achieving exceptional energy efficiency and robustness. To further explore brain-inspired electronics requires complementing conventional deposition approaches in nano- and microtechnology with bottom-up processes, self-organization, organic matrices and flexible substrates.

This contribution showcases bio-inspired electronic features observed in nanogranular matter, which is made up of self-organized nanoobjects, such as silver-based nanoparticles prepared by gas phase synthesis. In particular, two arrangement principles of nanoparticles into nanogranular structures are highlighted: On the one hand, Ag-based nanoparticles are organized as percolated nanoparticle networks via nanoparticle beam deposition under continuous monitoring of the resistance between planar electrodes on rigid substrates. Brain-like avalanche dynamics and criticality have been found in Ag nanoparticle networks by analysis of the inter-spike-intervals, avalanche size and avalanche duration. Threshold behaviour of collective resistive switching, criticality and avalanche dynamics in the current response are showcased using a Ag/CxOyHz@Ag composite network at the percolation threshold. On the other hand, Ag/polyethyleneglycol nanofluids are obtained by directly depositing Ag nanoparticles from gas phase synthesis and nanoparticle beam deposition into the reservoir of a vacuum compatible liquid. These nanofluids, after application onto rigid substrates with lithographically pre-structured electrodes, show the capacity for resistive switching upon formation of long-range reconfigurable conductive bridges via electrophoretic rearrangements of Ag nanoparticles in a liquid polymer matrix.