Phototransistor Mimics the Synaptic Behavior of Neurons

Technology #34446

Key Points

  • Enables neuromorphic computing for real-time machine learning applications
  • Made of graphene-perovskite quantum dot (G-PQD) superstructures
  • Synchronizes charge generation and transport on a single platform

Abstract

Researchers at the University of Central Florida have developed a phototransistor device that can act as an artificial photonic synapse for neuromorphic computing. Ultrathin and highly efficient, the device comprises a superstructure with the fast charge transport capability of graphene (G) and the efficient photogeneration features of perovskite quantum dots (PQDs). The UCF team also devised a unique method for using the device to mimic the synaptic behavior and energy efficiency of the human brain.

Biologically, a synapse acts as a channel of communication between two neurons. One neuron (a presynaptic cell) transmits information to a receiving neuron (a postsynaptic cell). With the UCF invention, the presynaptic signal consists of external stimuli—optical pulses, electrical pulses, or both. The postsynaptic signal is the current obtained across the G-PQD channel. Experimental results show that the device exhibited excellent responsivity of 1.4 × 108 AW–1 and specific detectivity of 4.72 × 1015 Jones at 430 nm.

Technical Details

In one embodiment, the UCF device comprises a substrate with a silicon dioxide layer and a patterned graphene source-drain channel. Grown on the graphene source-drain channel is a perovskite quantum dot layer of methylammonium lead bromide material. The new approach can extend to other 2D materials, including transition metal dichalcogenides and other heterostructures. A method of operating the device as an artificial photonic synapse includes applying a first fixed voltage to a gate of the phototransistor and a second fixed voltage across the graphene source-drain channel. In this example, the presynaptic signal comprises one or more pulses of light or electrical voltage. The postsynaptic signal is a measurement of the current across the graphene source-drain channel. The artificial synapses can strengthen (potentiate) or weaken (depress) based on the appropriate triggers of optical pulses.

Partnering Opportunity

The research team is looking for partners to develop the technology further for commercialization.

Benefit

  • Potential candidate for high-performing phototransistors that detect blue wavelength light
  • Provides good photonic memory without the need for an external gate voltage
  • Needs only low light intensity to significantly increase light response and photonic memory

Market Application

  • Image and pattern recognition, artificial eyes, artificial intelligence
  • Speech processing
  • Autonomous vehicles, drones