Imagine a world where materials can heal themselves and electronics are built to learn the way your brain does, all using wires tinier than a human hair.  My research lab at NUI Galway is focused on taking that world from science fiction to science fact! We take materials that are very small, or nanoscale, in one dimension, spray them onto a variety of surfaces, and try to make circuitry from them. Networks of nanowires  are memristive, which means that their electronic behavior depends on their measurement history.

We use metal nanowires that can be sprayed onto a variety of surfaces to create random networks. Light or electricity can change these networks of nanowires in useful ways. Where two nanowires cross to form a junction, light or electricity can change the strength of the junction. These individual changes change the overall behaviour of the network. The changes are self-healing, so electrical currents can route around damaged sections of network. The network also becomes ‘smarter’ controlled changes create new meaningful patterns of response to particular complex stimuli. When they are built, traditional silicon architectures for computing have fixed structure and implicit fixed digital modes of computation. Networks of nanomaterials can create flexible computational modes that can adapt on the fly, which makes these devices more like brains than like silicon chips.

By electrically stressing networks of nanomaterials, we can tune them to create electronic materials that learn from their history. Our nanoelectronics can thus be exploited to create an adaptive material suitable for making artificial skin, transparent conductors, or neuromorphic electronics.

Read more about neuromorphic electronics, the most exciting application of memory in nanostructured materials, in a piece I wrote for Physics World here.