Dual-Modal Memory Enabled by a Single Vertical N-Type Organic Artificial Synapse for Neuromorphic Computing

Area of Science:

• Materials Science
• Neuroscience
• Electronics

Background:

• Organic artificial synapses are crucial for advanced applications like image cognition and prosthesis control.
• Integrating dual-modal memory (volatile and nonvolatile) into a single synaptic transistor remains a significant challenge.

Purpose of the Study:

• To propose and demonstrate a single vertical n-type organic synaptic transistor (VNOST) capable of dual-modal synaptic learning and memory behaviors.
• To achieve high performance in both volatile and nonvolatile operating modes within a single device.

Main Methods:

• Utilized a novel polymeric organic mixed ionic-electronic conductor as the core channel material in the VNOST.
• Achieved dual-modal operation through electric double-layer formation and reversible ion doping at different current densities.
• Fabricated and characterized the VNOST for its synaptic functionalities and performance metrics.

Main Results:

• The VNOST demonstrated unprecedented volatile operating current density in the MA cm⁻² range.
• As a nonvolatile synapse, it achieved 150 analog states, symmetric conductance modulation, and 100s state retention.
• Artificial neural networks utilizing the VNOST's nonvolatile feature achieved a 94% handwritten digit recognition rate.

Conclusions:

• The developed VNOST successfully integrates dual-modal memory characteristics into a single organic synaptic transistor.
• This provides a promising platform for high-performing n-type organic synapses in complex neuromorphic network circuits.
• The device's capabilities pave the way for sophisticated organic neuromorphic applications.