Biologically Switchable Volatility and Nonvolatility Toward Real Neurotransmitter Mediated Aqueous Reservoir Computing.

The study on aqueous reservoir computing (RC) has just come into being. Nevertheless, aqueous RC still faces significant challenges. Recently, with the participation of real biological receptors, aqueous neuromorphic engineering is promising to emulate the intricate human brain more closely due to their synaptic resemblance in electrolytes. In this work, biologically switchable volatility and nonvolatility are first explored for real neurotransmitter glutamate-mediated aqueous RC application, which is based on rational synergy between glutamate oxidase catalytic chemistry and porphyrin-based metal-organic framework/tungsten oxide (PCN-224/WO3) photogate in the newly emerged organic photoelectrochemical transistor. As a proof of concept, the glutamate-mediated RC is then developed and explored for image recognition. These results have of potential for advancing real neurotransmitter-mediated aqueous RC toward in-biology application, which are also envisaged to promote bioinspired algorithm-hardware codesign in future aqueous neuromorphic engineering.