Organic synapses with programmable linearity for neuromorphic computing.

Charge-separated-buffered adaptive luminescence enables linearity-programmable organic all-photonic synapses with v = 0.0093, 99% uniformity, and 97% repeatability.

• Why it matters: Nonlinear, poorly reproducible signal transmission in organic synapses and nonlinear photochemical/photoisomerization in all-photonic designs limit energy efficiency and reliable neuromorphic processing.
• What they did: The team engineered guest molecular structures to tune charge-separation kinetics, then quantified synaptic linearity (v) and device performance metrics including uniformity, repeatability, trigger energy, and response time.
• The result: An all-photonic sensor system using these synapses achieved high-quality image acquisition and high image-classification accuracy, supporting a molecularly programmable photophysical platform for low-energy interfaces.

AI summary of the abstract · may contain errors · not medical advice · read the paper.