Researchers have created an event-driven organic photosensor that mimics biological vision by controlling signal amplitude within a single layer. The new sensor couples ionic and electronic transport in an organic mixed ionic-electronic conductor, enabling voltage-tunable control over the size of signal spikes. This functionality is typically achieved with complex multi-transistor setups. This design preserves motion-relevant contrast and reduces data load, establishing ionic-electronic coupling as a strategy for adaptive, low-power neuromorphic vision hardware.
Ionic-Mediated Inhibition in Organic Photosensors
Organic photosensors use ionic-mediated inhibition to achieve amplitude-controllable event-driven responses, a departure from conventional semiconductor designs that require complex multi-transistor architectures. The researchers specifically designed the sensors to move beyond simply modulating channel conductance, instead generating the paired excitation-inhibition dynamics necessary for an event-driven photoresponse. The developed sensors employ a two-terminal photoelectrode architecture, differing from the more common three-terminal organic electrochemical transistor geometries seen in previous OMIEC devices. This distinction leads to a unique operating mechanism and resulting photoresponse characteristics.
The operational cycle of these sensors comprises four stages, enabling a process where ionic doping mediates photocurrent inhibition within the bulk heterojunction. Supporting this, steady-state photoluminescence spectra of PgBDT-T:Y6 blends at different ratios demonstrate the material’s properties critical to the sensor’s function. Prior attempts using additional non-OMIEC blends failed to produce the necessary excitation-inhibition transients, emphasising the importance of ion-accessible mixed ionic-electronic conduction as a key requirement for this type of event-driven response.
OMIECs Enable Event-Driven Spike Generation
Event-driven organic photosensors achieve voltage-programmable amplitude modulation within a single active layer, a capability previously requiring complex multi-transistor architectures. This advance stems from integrating temporal-contrast detection with dynamic sensory-gain modulation, mirroring biological vision’s prioritization of stimuli. A bulk heterojunction composed of PgBDT-T and Y6 facilitates this process; illumination changes are converted into transient photocurrent spikes via fast electron extraction through Y6. Simultaneously, ion-compensated hole accumulation within the PgBDT-T donor screens the extraction field, promoting recombination and suppressing subsequent photocurrent.
This ionic-electronic coupling allows for both temporal-contrast detection and analogue amplitude control within the same pixel, effectively mimicking attentional gain modulation observed in biological systems. The result is spatially resolved, bias-programmable event weighting, which supports efficient data processing. These results establish ionic-electronic coupling in organic mixed ionic-electronic conductors as a materials strategy for adaptive, low-power neuromorphic vision hardware.
This approach moves beyond simply modulating channel conductance, offering a more subtle and efficient method for processing visual information. The work builds on previous advances in neuromorphic vision sensors, as demonstrated in the 2022 International Electron Devices Meeting and publications in ACS Nano, but distinguishes itself through its single-layer integration of spike generation and amplitude control.
PgBDT-T:Y6 Bulk Heterojunction Architecture
Fabricated on a gold working electrode and interfacing with a counter electrode through an electrolyte, the architecture of the new photosensor relies on a 130-nanometer-thick PgBDT-T:Y6 bulk heterojunction. This configuration allows for the interaction of both photocarriers and compensating ions within the same active layer, a key element in achieving the sensor’s unique functionality. Researchers synthesized PgBDT-T, a π-conjugated polymer with alkoxy side chains, and blended it with Y6 to form the photoactive heterojunction, ensuring both components retain their characteristic optical absorptions within the blend.
Electrochemical measurements confirmed p-type charging and bulk ionic doping of PgBDT-T, demonstrating its role in the sensor’s operation. Atomic force microscopy and in operando quartz crystal microbalance measurements revealed swelling-associated anion uptake in PgBDT-T, while Y6 exhibited negligible thickness and mass changes, indicating that ionic accessibility is primarily concentrated within the donor network.
The IEES operational cycle comprises four distinct stages, as detailed in the study, and is designed to integrate photodetection with ionic modulation within a single active layer. 1-second time resolution. Control experiments involved introducing P3HT interlayers with thicknesses of 10, 40, and 60 nanometers between the gold working electrode and the PgBDT-T:Y6 bulk heterojunction, further refining the understanding of charge transport dynamics.
Temporal-Contrast Detection via Electronic-Ionic Coupling
Fast electron extraction through the acceptor generates excitatory spikes, while ion-compensated hole accumulation within the donor material provides the inhibitory control over spike amplitude. This unique approach allows for in-sensor amplitude-temporal coding, a method that preserves motion-relevant contrast in optical events while simultaneously reducing redundant read-out processes. The sensor’s ability to bias-weight events effectively prioritizes visual information, mirroring how biological systems actively filter stimuli.
The study reports, highlighting the efficiency gained by integrating these functions into a single component. The results demonstrate that amplitude-coded event sensing combines sparse temporal sampling with analogue signal strength, preserving motion-relevant information and reducing redundant front-end processing.
In-Sensor Amplitude Coding of Optical Events
The newly developed sensor preserves motion-relevant contrast by directly encoding event amplitude within the active layer, a departure from conventional designs requiring complex multi-transistor setups to achieve tunable signal strength. This in-sensor amplitude-temporal coding reduces redundant read-out, a critical step toward more efficient data processing in dynamic scenes. The approach uses the unique properties of organic mixed ionic-electronic conductors, allowing for bias-weighting of optical events before external processing.
Researchers implemented a hybrid amplitude-temporal-coded neural network, termed IEES-HATC NN, where changes in illumination are converted into analogue events weighted by applied bias. Benchmarking on the Weizmann human-action dataset yielded 100% classification accuracy with a front-end energy consumption of 0.064 mJ, demonstrating a significant reduction in power requirements. The front-end convolution within the IEES-HATC NN pipeline assigns 3 × 3 × 16 convolutional kernel weights to pixel-specific bias voltages, directly modulating the amplitude response.
Accumulated weighted event amplitudes then generate feature maps for a lightweight LSTM classifier. This design draws inspiration from biological vision, where sensory gain is actively modulated according to contextual relevance, mirroring the efficiency of retinal bipolar cells and maximizing sensory information. The study highlights the parallels between the sensor’s operation and natural systems.
This approach contrasts with traditional frame-based sensors that generate continuous analogue signals regardless of scene dynamics, and event cameras that produce sparse binary events. The results establish ionic-doping-mediated inhibition in OMIECs as a materials strategy for adaptive, low-power neuromorphic vision.
The system’s front-end convolution and downstream LSTM protocols were evaluated with matched input resolution, enabling a direct comparison of performance metrics. The IEES front-end integrates temporal-contrast detection, analogue amplitude encoding and bias-programmed spatial weighting, creating a system that prioritizes relevant information before processing.
Adaptive Gain Modulation Mimics Biological Vision
Organic photosensors now use a mechanism mirroring biological vision to actively control signal strength within a single material layer, bypassing the need for complex multi-transistor circuits typically used to achieve tunable amplitude. The resulting system demonstrates a functional analogue to attentional gain modulation, a process where biological vision prioritizes stimuli by adjusting neural sensitivity.
The design establishes a new materials strategy for building adaptive, low-power neuromorphic vision hardware, moving beyond passive sensory adaptation found in earlier ionic-doping-based devices. But conventional semiconductor materials lack the intrinsic ionic feedback necessary to suppress photocarriers after initial excitation.
Electrolyte Interface Facilitates Ion-Carrier Interaction
Ultraviolet-visible measurements reveal a modest change in PgBDT-T absorption during continuous operation, a contrast to negligible spectral changes observed in control samples lacking organic mixed ionic-electronic conductors. This difference indicates a dynamic interaction between ions and photogenerated charge carriers within the active layer of the new sensor design, a key element in its ability to modulate signal amplitude. The observed spectral shift suggests the electrolyte interface isn’t merely a passive component, but actively participates in the sensor’s response to light.
Nitrogen purging confirmed that oxygen reduction isn’t the primary contributor to this behavior, isolating the effect of the ionic-mediated process. The speed of the photocurrent spike is directly linked to the sensor’s capacity for event-driven operation, minimizing redundant data transmission.
See today’s quantum computing news on Quantum Zeitgeist for the latest breakthroughs in qubits, hardware, algorithms, and industry deals.
