An ultrasmall organic synapse for neuromorphic computing.

Area of Science:

• Materials Science
• Neuroscience
• Electrical Engineering

Background:

• Miniaturized organic neuromorphic devices are crucial for brain-inspired artificial intelligence.
• Challenges exist in downscaling organic devices and integrating them due to structural inhomogeneity.

Purpose of the Study:

• To design and fabricate high-performance organic synapses with reduced dimensions and enhanced integration.
• To achieve reliable device performance for neuromorphic computing applications.

Main Methods:

• Designed a semicrystalline polymer (PBFCL10) with an ordered structure to control conductive nanofilament formation.
• Fabricated organic synapses with a minimal device dimension of 50 nm and an integration density of 1 Kb.
• Implemented a mixed-signal neuromorphic hardware system with an organic neuromatrix and FPGA controller.

Main Results:

• Achieved the smallest organic synapse device dimension (50 nm) and highest integration size (1 Kb) to date.
• Demonstrated 32 linear conductance states with high cycle-to-cycle (98.89%) and device-to-device (99.71%) uniformity.
• Successfully executed a spiking-plasticity algorithm for decision-making tasks on the implemented hardware system.

Conclusions:

• The developed PBFCL10-based organic synapses represent a significant advancement in miniaturized and high-density neuromorphic devices.
• The high performance and uniformity of these organic devices are superior to existing organic counterparts.
• The successful implementation of a neuromorphic system highlights the potential for practical applications in artificial intelligence.