Chinese researchers have created a paper-based device that mimics electrochemical signalling in the human brain. The novel device could become a key component in the development of artificial neural networks for use in applications ranging from robotics to computer processing.


Developed by a team at Nanjing University and the Ningbo Institute of Material Technology and Engineering, the thin-film transistor

Qing Wan

(TFT) device was designed to replicate the natural junction, known as a biological synapse, between two brain cells or neurons. Neurons use the synapse to pass electrochemical signals and messages around the brain.

The synaptic TFT consists of an indium-zinc-oxide (IZO) channel and gate electrode, separated by a 550-nanometre-thick film of proton-conducting silicon dioxide electrolyte. The team chose to fabricate their TFT using paper because it is flexible, lightweight, inexpensive and environmentally friendly.

All neurons can generate an electric spike when their voltage changes by large enough amounts. These spikes cause signals to flow between neurons: the first neuron releases chemicals — called neurotransmitters — across the synapse, which are received by the second neuron, passing the signal on.

Similar to these spikes, the researchers applied a small voltage to the gate electrode in their device, causing protons from the silicon dioxide films to migrate towards the IZO channel opposite it. As protons are positively charged, this induced negatively charged electrons to be attracted toward them in the IZO channel, which subsequently allowed a current to flow through the channel — mimicking the movement of neurotransmitters from one neuron to another across a synapse.

As more and more neurotransmitters pass across a synapse between two neurons in the brain, the connection between the two neurons becomes stronger. This is the basis of how people learn and memorise things.

The researchers demonstrated this phenomenon, known as synaptic plasticity, in their own device. As reported in IOP Publishing’s journal, Nanotechnology, the team found that when two voltage pulses were applied to the device in a short space of time, the second pulse triggered a larger current in the IZO channel than the first voltage, as if it had remembered the response from the first voltage.

“A paper-based synapse could be used to build lightweight and biologically friendly artificial neural networks,” says co-author Qing Wan of Nanjing University. “At the same time, with the advantages of flexibility and biocompatibility, it could be used to create the flexible organism-machine interface for many biological applications.”



For further information contact:

Qing Wan
School of Electronic Science and Engineering, Nanjing University
Ningbo Institute of Material Technology and Engineering, Chinese Academy of Sciences
Peoples Republic of China