Nanocomposite stretches performance of artificial synapse

Researchers have developed a stretchable transistor based on a polymer-perovskite quantum dot nanocomposite that functions as an artificial synapse [Chen et al., Materials Today (2023), https://doi.org/10.1016/j.mattod.2023.10.010].

“[This work] targeted the emulation of photonic synapses, offering a promising avenue for developing low-energy consumption soft electronics, neurologically inspired robotics, and neuromorphic network computation,” explains Chi-Ching Kuo of National Taipei University of Technology, who led the study with Wen-Chang Chen and colleagues at National Taiwan University and National Cheng Kung University.

Attention is turning to neural-inspired approaches to get around two major shortcomings of traditional computing – high energy consumption and slow computation speeds. The brain, by contrast, uses synapses to receive information from external stimuli, process signals, and generate motor responses at high speed and with minimal energy consumption. The result is an extraordinarily efficient system able to recognize patterns, self-learn, and process information. In a bid to mimic the brain’s operation, organic semiconductors are one of the most promising materials for artificial synapses because of their low cost, easy processability, mechanical flexibility, and biocompatibility.

Now Chen and his coworkers have fabricated fully stretchable photo-stimulated synaptic devices using poly(d-decanolactone) (PDL) conjugated block copolymers (BCPs) in conjunction with perovskite quantum dots (PeQDs). The researchers found that carefully choosing the right solvent for PDL-based BCPs controls the distribution of PeQDs in the composite, leading to improved self-aggregation, larger grain size, and better interfaces.

“Regulating the assembly of conjugated polymers and improving the self-aggregation of PeQDs… is beneficial to low energy consumption and high photosensitivity devices,” points out Kuo.

Devices based on the BCP/PeQD nanocomposite show light-stimulated artificial synaptic functions including paired-pulse facilitation, spike-dependent plasticity, and short/long-term memory with low energy consumption, fast response time, and adjustable wavelength response.

“We [also demonstrated] various light stimuli for the devices, producing different excitatory postsynaptic currents (EPSC)… with [which] we can achieve switching functions and data recognition,” says Kuo.

The fully stretchable BCP/PeQD photosynaptic device mimics the actions of human synapses in perceiving and propagating signals, maintaining accurate responses even under strain. The researchers believe the new device could hold potential for innovative technologies such as soft electronics and robotics, neurobotics, pattern recognition, and artificial intelligence. Attention now needs to be focused on materials optimization by controlling the interaction of the two components, using functional groups, hydrogen bonding, and surface ligand engineering, and improving recognition response.