Fluorinated Self-Assembled Monolayer Ion Receptors for Retentive Analog Synaptic Behavior

Area of Science:

• Materials Science
• Neuroscience
• Electrical Engineering

Background:

• Ionic transistors are key for retentive artificial synapses in neuromorphic computing due to analog modulation and low energy use.
• Self-discharge of ions at the channel-electrolyte interface currently limits the retention characteristics of these devices.
• Engineering interfacial electrostatic interactions is crucial for controlling ion discharge kinetics and improving device performance.

Purpose of the Study:

• To enhance ion interactions and improve the retention characteristics of ionic transistors for artificial synapses.
• To introduce a novel ion receptor material for electrolyte-gated transistors (EGTs).
• To address the limitations of ion self-discharge in neuromorphic computing applications.

Main Methods:

• Introduction of a fluorinated self-assembled monolayer (F-SAM), specifically Heneicosafluorododecyl phosphonic acid (F21-DDPA), as an ion receptor in EGTs.
• Investigation of ion-dipole interactions between lithium ions and the F21-DDPA layer at the channel/electrolyte interface.
• Utilizing chemical analysis and first-principles density functional theory (DFT) calculations to understand ion dynamics.

Main Results:

• The F21-DDPA layer effectively traps lithium ions at the interface via strong ion-dipole interactions.
• The modified EGTs exhibit stable, near-linearly tunable multimodal conductance states, indicating enhanced retention.
• F21-DDPA facilitates sequential ion trapping, providing a fundamental solution for ion dynamics at the interface.

Conclusions:

• The proposed F-SAM approach significantly enhances ion trapping at the channel-electrolyte interface, overcoming self-discharge limitations.
• This method enables ionic transistors to achieve stable, long-lasting charge retention essential for artificial synapses.
• The findings offer a pathway for developing advanced neuromorphic computing devices with improved training acceleration and inference capabilities.