A novel artificial synapse with dual modes using bilayer graphene as the bottom electrode

Area of Science:

• Materials Science
• Neuroscience
• Computer Engineering

Background:

• Resistive Random Access Memory (RRAM) is a promising technology for artificial synapses in neuromorphic computing.
• Current RRAM devices exhibit limited potentiation states, hindering the mimicry of the learning process and leading to depression-only behavior.
• Oxygen vacancies are key factors in RRAM device operation and state retention.

Purpose of the Study:

• To develop a novel RRAM device capable of mimicking both excitatory and inhibitory synaptic behaviors.
• To overcome the limitations of abrupt set operations in conventional RRAM for enhanced neuromorphic applications.
• To increase the number of potentiation states for more effective synaptic plasticity.

Main Methods:

• Fabrication of an Al/AlOx/graphene RRAM stack utilizing oxygen vacancies as trapping centers.
• Employing a bilayer graphene bottom electrode as the channel material.
• Utilizing pulsed stimuli to gradually create oxygen vacancies for controlled potentiation and depression.

Main Results:

• Introduction of a new operating mode (Mode II) in the RRAM device.
• Achieved 166 potentiation states, significantly exceeding conventional RRAM limitations.
• Demonstrated the capability for both gradual potentiation and depression of post-synaptic current (PSC).
• Successfully realized inhibitory synapse functionality using Mode II.
• Enabled a single device to function as both an excitatory and inhibitory synapse by combining Mode I and Mode II.

Conclusions:

• The novel RRAM device with Mode II offers enhanced capabilities for artificial synapses.
• This advancement significantly improves the learning and reconfigurability of neuromorphic systems.
• The developed device paves the way for more sophisticated and efficient neuromorphic computations.