Recent Progress in Organic Optoelectronic Synaptic Devices

Abstract

Organic semiconductors hold immense promise in the field of optoelectronic synapses due to their tunable optoelectronic properties, mechanical flexibility, and biocompatibility. This review article provides a comprehensive overview of recent advancements in organic optoelectronic synaptic devices. We delve into the fundamental concepts and classifications of these devices, examine their roles and operational mechanisms, and explore their diverse application scenarios. Additionally, we highlight the current challenges and emerging opportunities in this field, outlining a forward-looking path for the future development and application of these materials and devices in next-generation artificial intelligence (AI). We emphasize the potential of further optimizing organic materials and devices, which could significantly enhance the integration of organic synapses into biointegrated electronics and human–computer interfaces. By addressing key challenges such as material stability, device performance, and scalability, we aim to accelerate the transition from laboratory research to practical applications, paving the way for innovative AI systems that mimic biological neural networks.

1. Introduction

1.1. Memristor Presentation

With the advancement of information technology, the efficiency of data capture, encoding, and processing in computers has led to increasingly demanding requirements. However, due to the separation of processing and storage units, the central processor must first retrieve data from the storage unit during command execution, leading to processing delays and high power consumption for data retrieval [1,2], which does not meet practical demands. In contrast, the human brain governs nearly all complex physiological functions and is capable of simultaneous computation and information storage, offering superior time efficiency. Figure 1a illustrates a schematic representation of the information processing mechanisms employed by conventional computers and the human brain. when addressing intricate problems, in neuroscience, the efficiency of the nervous system is largely attributed to the extensive network of neuronal connections within the cerebral cortex, which enables highly parallel processing [3]. The brain contains tens of billions of neuronal cells, which are interconnected to form synapses—the fundamental units of signaling and regulation in neural networks [4,5,6,7]. A key feature of synapses, synaptic plasticity, serves as the molecular foundation for learning and memory. Therefore, the dual emulation of synaptic mimicry and plasticity emerges as a fundamental prerequisite for constructing neuromorphic computational architectures, serving as a crucial pathway to achieve energy-efficient brain-inspired artificial neural systems. Synapses play critical roles in energy exchange, material transport, and information transfer between neurons. To address the pressing demands of computer development, researchers have sought to simulate information transfer between neurons in the human brain [8,9], aiming to create brain-like systems with integrated memory and signal processing capabilities. In 1971, Professor Shao-Tang Tsai from the University of California, Greeley, proposed the “disappearing circuit element”, the memristor, which, due to its low power consumption, is highly beneficial for designing and optimizing neuromorphic circuits [10]. Research has shown that memristors and synapses exhibit similar transmission properties, with memristors possessing unique, synapse-like nonlinear electrical characteristics. A single memristor can replicate the fundamental function of a synapse [11,12,13,14]. In 2008, Hewlett–Packard Labs successfully developed a nano-memristor using titanium dioxide (TiO₂). They presented a fundamental physical model, the ion migration model, which has since inspired many researchers to fabricate a variety of memristors based on this model [15]. In 2013, Dr. Thomas developed a memristor integrated into a chip 600 times thinner than human hair, using it as a key component of an artificial brain [16]. As a result, efforts have been made to utilize memristors for simulating various synaptic functions, thereby enabling effective neuromorphic computation [17,18], which is essential for the development of efficient brain-like visual systems. In recent decades, numerous research groups have successfully modeled various synaptic functions using memristors.

1.2. Photoelectric Synapses

The human nervous system is comprised of three primary components: the sensory system (SS), the central nervous system (CNS), and the effector system (ES). Specialized receptor cells enable the transmission of various external stimuli as electrochemical signals to the CNS, which processes these signals and sends corresponding commands to the effector systems for response. The sensory system encompasses several critical components, including sight, hearing, smell, and taste. Among these, the visual sensory system (VSS) is particularly significant, as nearly 80% of the external information received by humans is derived from vision. The human eye efficiently captures visual scenes and converts them into neuroelectrical signals, which are subsequently processed and interpreted by the retina and the cerebral cortex. Through vision, humans are capable of perceiving the size, brightness, color, and motion of external objects. Consequently, light has been incorporated into artificial synaptic devices as a widely used and easily modulated form of stimulation. Light’s high bandwidth, robustness, and parallelism help mitigate conflicts between integration and signal integrity. Additionally, regulating the optical signal’s power can effectively reduce the Joule heating generated by the device. We are currently in the era of artificial intelligence (AI), with one of its most attractive features being its capacity for intelligent perception. In response to these advancements, the development of extremely low-energy, highly efficient storage-computing devices—integrating detection, information processing, and memory—has demonstrated significant potential in the fields of artificial vision systems, brain-like neural computation, and beyond. The key advantage of brain-like artificial synapses lies in their ability to emulate the internal conductivity associated with “renewal” and “memory” (referred to as “synaptic weights”). Consequently, optoelectronic materials and devices that emulate the functionality of biological synapses have garnered significant attention, and an increasing number of researchers are dedicating their efforts to this field.

1.3. Organic Photoelectric Synapse (OPS)

A broad range of materials, including metal oxides, two-dimensional materials, chalcogenides, quantum dots, and organics, have been explored for photoelectric synapses. Metal oxides offer notable advantages in memristor applications, particularly with respect to tunability and compatibility with CMOS systems. However, they also pose several challenges, particularly concerning stability, reproducibility, and integration with read-out integrated circuits (ROICs) [20,21,22,23,24,25,26]. Two-dimensional materials, including transition metal sulfides [27,28,29,30], hexagonal boron nitride (hBN) [31], graphene [32,33], and phase-change materials such as Ge₂Sb₂Te₅ [34], exhibit unique structural and electronic properties. These properties, which can be modulated through phase transition engineering, make them promising candidates for neuromorphic devices. However, these materials still face significant challenges, including mass production, long-term stability, and consistent performance. Halide chalcogenides have been utilized in various di- and tri-terminal synaptic devices [35,36,37,38,39] and show significant potential in neurocomputing. However, challenges remain, particularly in terms of stability, toxicity, and biocompatibility. Quantum dots have emerged as promising candidates for resistive functional materials in resistive memories, owing to their high electron mobility and tunable energy band structures. However, they face several challenges, including unclear resistive mechanisms, reduced stability and controllability, and a lack of effective modulation techniques. Organic semiconductors are among the most promising candidates due to their superior physical and chemical properties compared to inorganic materials. These advantages enable the design and fabrication of bionic sensors with enhanced sensing performance, biocompatibility, and wearability.

In biosensors and neural synaptic devices, substrates serve as both physical supports and functional carriers, with their material properties directly influencing device performance, process compatibility, and application scenarios. Substrates can be classified into polymer substrates, inorganic flexible substrates, composite substrates, and rigid substrates based on material types. Polymer substrates (e.g., PET, PEN) exhibit advantages such as high flexibility, light weight, low cost, and suitability for roll-to-roll fabrication processes, making them ideal for all-solution-processed synaptic arrays. Their high surface flatness facilitates uniform transfer of two-dimensional materials and interfacial contacts [40]. However, their poor thermal stability restricts high-temperature processing. Inorganic flexible substrates (e.g., ultra-thin glass) combine flexibility with high-temperature stability, featuring low thermal expansion coefficients. Nevertheless, their mechanical brittleness limits conformal contact with dynamic biological tissues, reducing practical applicability. Composite substrates (e.g., PI/AgNW [41]) achieve a balance between flexibility and functionality through hybrid designs, utilizing AgNW networks to enhance conductivity, support low-power synapses, and improve temperature resistance. However, they suffer from high process complexity due to the need for precise interface compatibility control. Rigid substrates (e.g., silicon wafers, glass) offer advantages such as high thermal stability and surface flatness, enabling high-temperature processing and CMOS integration to ensure high carrier mobility and low noise performance. Their lack of flexibility, however, restricts conformal applications on biological tissues or dynamic environments, limiting use in wearable and implantable devices.

In particular, organic materials can be processed using low-cost techniques, such as rotary coating, blade coating, roll-to-roll printing, and inkjet printing. These materials can be directly applied to a variety of flexible substrates, making them suitable for the large-scale fabrication of flexible bionic devices [42,43,44,45,46]. The key advantage of organic semiconductors in electronics lies in their ability to serve as highly balanced, high-mobility bipolar active layer materials, facilitating the simultaneous transport of holes and electrons through molecular-level modification strategies. Additionally, organic synapses feature a simple device structure and low energy consumption, comparable to biological systems [2,47,48,49,50,51,52,53]. This has sparked significant interest in next-generation intelligent sensing systems based on neural networks, such as robotics, wearable devices, disease diagnosis, the Internet of Things, and human–computer interaction systems. Figure 1b illustrates the working mechanism of the biological nervous system in conjunction with artificial photonic synapses. Artificial organic photoelectric synapses can respond to light stimuli, perform real-time preprocessing and short-term memory storage of imaging information, and enhance the efficiency and accuracy of subsequent image recognition tasks. The image preprocessing function of photoelectric synapses is realized through light intensity-dependent and light time-dependent plasticity, enabling processes such as image noise reduction, contrast enhancement, and sharpening. The human VSS can be modeled and extended to incorporate imaging memory, damage receptors, and perceptual applications.

To provide a comprehensive understanding of organic optoelectronic synapses, this review systematically summarizes recent advancements in the field, aiming to offer multifaceted guidance for the design and integration of highly biomimetic intelligent sensory devices. Based on device architectures, organic optoelectronic synapses are categorized into two primary classes: organic photodiodes and organic phototransistors. The classification of organic semiconductor material systems is further elucidated, with an in-depth exploration of the underlying operational mechanisms. Key performance metrics for organic optoelectronic synapses are critically analyzed (as illustrated in Figure 2). This review also addresses critical challenges hindering practical applications. With continued innovations in materials, device architectures, and fabrication processes, organic optoelectronic synapses are poised to enable autonomous perception–learning–decision systems.

2. Basic Concepts of Organic Photoelectric Synapses and Their Classification

Organic photoelectric synapses are photoelectric synapses that incorporate organic materials. These synapses can mimic the behavior of biological nervous system synapses and simultaneously detect and retain light signals. The devices exhibit varying conductance changes in response to different types of light signals. A stand-alone photoelectric memristor device can be classified into two-terminal and three-terminal structures based on its design. These devices can also be classified as vertical (also known as stacked or sandwich structures) or planar, depending on the cross-sectional shape. From a biological perspective, the two-terminal memristor closely resembles biological synapses. In terms of circuit integration, the two-terminal memristor is smaller and simpler, allowing for higher integration, and it consumes less energy. The three-terminal memristor is derived from a field-effect transistor (FET), where the drain current flowing through the channel is modulated by the gate voltage at a fixed source–drain bias. This enables parallel functions of transmission and learning, making it a generalized type of optoelectronic memristor.

2.1. Photodiode

A photodiode is a two-terminal electronic device with a light-sensitive layer sandwiched between two electrodes, as shown in Figure 3a. The photosensitive layer is composed of a mixture of donor and acceptor materials, forming a bulk heterojunction (BHJ). When a bulk heterojunction film is irradiated with light, it generates excitons, which are bound electron–hole pairs. The excitons then travel to the donor–acceptor interface, where they separate into individual electrons and holes. After separation, electrons are transported through the acceptor material network, while holes move through the donor material network. Ultimately, these carriers are collected by the device’s external electrodes, generating a photocurrent. Applying a pulsed voltage across the device simulates the neuron’s stimulus signal, which captures the interfacial charge through a high adjacent energy barrier and charge trap, thereby triggering synaptic properties. The conducting state of the memristor simulates changes in synaptic weights. By varying parameters such as pulse voltage frequency and duration, different synaptic functions can be simulated, leading to corresponding changes in the conducting state.

2.2. Phototransistors

Among various device types, photoresponsive organic field-effect transistors (OFETs) have emerged as one of the most promising options for developing photocrystal synaptic electronics, owing to the tunable optical and electrical properties of organic materials, their solution processability, and excellent mechanical flexibility [49]. These devices consist of three electrodes: a source, a drain, and a gate, as shown in Figure 3b. OFETs primarily achieve biomimetic synaptic functions under light stimulation by capturing and releasing photogenerated charges within the organic semiconductor layer or at the interface between the semiconductor and dielectric materials [54,55]. Specifically, transistor devices with a three-terminal configuration utilize additional gate electric fields to adjust their response to optical stimulation and synaptic weights [56], thereby enabling these devices to mimic optical signaling and learning behaviors in the biological visual nervous system [8,57]. In 2022, Liu et al. [58] pioneered a multi-sensory artificial synapse based on electrolyte-gated vertical organic field-effect transistors (VOFETs). By employing a crosslinking strategy to minimize the channel length of organic transistors and leveraging electric double-layer capacitance (EDLC), they achieved significant energy efficiency reduction and successfully emulated biological synaptic plasticity. In 2023, Liu et al. [59] further developed an all-polymer electrochemical transistor (AECT) to mimic human tactile and gustatory receptors. The core innovation lies in using UV lithography to create microchannels in PMMA, which are filled with PVA electrolyte to enable independent gate control. This device exhibits outstanding advantages in low energy consumption, high biocompatibility, optical transparency, and mechanical flexibility. Critically, it operates effectively in electrolyte media, addressing the limitations of conventional solid-state technologies. For the first time, the integration of tactile and gustatory functions within a single device was realized, overcoming the functional limitations of traditional sensors.

In summary, this chapter provides a systematic comparison of the distinct advantages of two-terminal and three-terminal structures in bionic functionality and integration potential. Two-terminal devices exhibit unique strengths in emulating fundamental synaptic plasticity through their synapse-mimetic structural simplicity, high-density integration capability, and ultralow energy consumption. In contrast, three-terminal devices facilitate optoelectronic multimodal signal co-processing via gate-mediated modulation, enabling precise regulation of synaptic weights. Future research should focus on developing optimized architectures (e.g., hybrid configurations) to establish an optimal balance between key performance metrics.

3. Types of Organic Materials Used in Photonic Synaptic Devices

Organic semiconductors are materials that use π-conjugated molecules as the fundamental structural unit and exhibit semiconductor properties. These materials include small molecules and conjugated polymer semiconductors. The main structural backbone of these materials consists of elements such as acenes, thiophene, triphenylamine, and pyrrole, including examples like pentacene, poly(3-hexylthiophene-2,5-diyl) (P3HT), poly[bis(4-phenyl)(2,4,6-trimethylphenyl)amine] (PTAA), and poly [2,5-(2-octyldodecyl)-3,6-diketopyrrolopyrrole-alt-5,5-(2,5-di(thien-2-yl)thieno-[3,2-b]thiophene)] (DPPDTT). These materials enable charge transport through π-π electron jumps in conjugated systems. Small-molecule organic semiconductors possess stable molecular structures and strong intermolecular interactions, which contribute to the formation of regularly ordered solid stacks, often resulting in high charge carrier mobility. Conjugated polymer semiconductors consist of a series of semiflexible molecular chains of varying lengths, offering good film-forming ability and charge-transfer properties, with performance comparable to or exceeding that of amorphous silicon. Compared to rigid inorganic semiconductors with strong covalent bonds, organic semiconductors offer advantages such as low-temperature solution processability, high mechanical flexibility, tunable optoelectronic properties, and deformability. The tunable photovoltaic property is the design of the molecular structure, including the addition or substitution of functional groups. The bandgap of an organic semiconductor is tuned to span the light absorption range from ultraviolet (UV) to near-infrared (NIR) through appropriate molecular design. The design of the organic material influences the charge transport properties, photoconductivity, and interfacial effects of the channel material. The organic semiconductor materials used in optoelectronic synaptic devices are summarized in Figure 4, categorized by their optical absorption ranges.

3.1. Small-Molecule Organic Semiconductors

Pentacene is a material with a wide range of absorption properties, especially in the 500–700 nm range. Owing to its planar structure and tight intermolecular packing behavior, pentacene has a hole mobility as high as 11 cm²/V·s. It achieves a high mobility of 30.6 cm²/V·s on specific substrates. Wang et al. [60] constructed photonic synapses using pentacene and CsPbBr₃ quantum dots and achieved multiple nonvolatile memory states through the combined modulation of optical and electrical signals, simulating synaptic functions at different stages of maturation. In this structure, light irradiation causes charge carriers to be generated in the CsPbBr₃ QD, and the holes are able to move through the PMMA layer to the HOMO energy level of pentacene, while the electrons are retained in the quantum dots. This leads to a continuous rise in current. In addition to pentacene, other high-mobility small-molecule semiconductors, such as DNTT and C8-BTBT, have been used to increase the performance of phototransistors. Han et al. [61] further combined pentacene with black phosphorus (BP)–zinc oxide (ZnO) nanoparticles to realize a variety of synaptic functions, including EPSC, spike rate-dependent plasticity (SRDP), long-term potentiation (LTP), long-term depression (LTD), and spike timing-dependent plasticity (STDP), and pentacene can also be used as a trap layer in photonic synapses.

C8-BTBT is a highly efficient hole-transport semiconductor material with a maximum hole mobility of 43 cm²/V·s. Huang et al. [62] applied it in combination with polyacrylonitrile (PAN) to photonic synapses in 2018. Upon ultraviolet illumination, electron–hole pairs are generated at the C8-BTBT and PAN interfaces, with a portion of the holes becoming trapped. Once the UV light is extinguished, these trapped holes are slowly released back into the channel, resulting in the decay of the EPSC, which aligns with the characteristics of Short-Term Plasticity (STP). In addition to the STP, the system also exhibits paired-pulse facilitation (PPF), spike-number-dependent plasticity (SNDP), spike-timing-dependent plasticity (STDP), and light-pulse-intensity-dependent PPF. In subsequent studies, the team improved the mobility and resolution of the device by replacing thermally evaporated C8-BTBT with self-assembled 2D crystals, which helped to address the short-channel effect.

Y6, a high-performance small-molecule non-fullerene acceptor (NFA), features a benzotrithiophene–pyrrolidone (BTP) core structure with cyano electron-withdrawing end groups and fluorinated side chains to enhance intermolecular interactions. Its narrow bandgap enables strong light absorption in the near-infrared (NIR) region. Pu et al. [63] integrated Y6 with the polymer donor PM6 to construct an Au/PM6:Y6/ITO optoelectronic synaptic memristor in 2025. The energy level alignment between Y6 and PM6 facilitates efficient charge separation. Photogenerated excitons dissociate at the PM6:Y6 interface into electrons and holes, which migrate toward the ITO and Au electrodes, respectively, forming a photovoltaic field. By applying positive/negative voltage pulses, the device modulates hole trapping and release processes, achieving multiple continuous and reversible photoconductive states that emulate synaptic LTP and LTD. Furthermore, the device successfully demonstrated adaptive optical pattern recognition in a 5 × 5 pixel array. Y6’s high electron mobility and synergistic optoelectronic response mechanism provide a critical material foundation for neuromorphic devices integrating sensing, memory, and computing functionalities.

3.2. Conjugated Polymer Semiconductors

Poly(3-hexylthiophene-2,5-diyl) (P3HT) is a conjugated polymer containing thiophene rings, commonly used in the fabrication of organic solar cells and field-effect transistors due to its excellent solubility and film-forming ability. In 2023, Chen et al. integrated a photosensitive bulk heterojunction layer (P3HT:PCBM) into the channels of organic electrochemical transistors, thereby forming a novel optoelectronic synapse [64]. Upon illumination, this layer generates charge carriers that induce an influx of anions from the electrolyte to balance the charge within the channel, thereby increasing the drain current of the transistor. Once the light is extinguished, the anions surrounding the P3HT delay charge recombination, leading to a gradual decrease in current and enabling nonvolatile memory capabilities.

PEDOT:PSS is a highly conductive, solution-processable conductive polymer. Its electrical conductivity can be significantly enhanced by doping with ethylene glycol (5 wt.%), while maintaining high optical transparency and mechanical flexibility. Park et al. [65] fabricated a flexible transparent organic memristor by integrating this material with poly(methyl methacrylate) (PMMA) and silver (Ag) nanoclusters. Under an electric field, conductive filaments form via the electrochemical metallization mechanism of Ag nanoclusters. The conductivity difference of PEDOT:PSS electrodes (pristine vs. doped) governs the device’s memory modes. Pristine PEDOT:PSS electrodes exhibit high contact resistance, which suppresses stable conductive filament growth, resulting in volatile memory behavior that emulates STP. In contrast, doped electrodes reduce resistance to facilitate robust filament formation, enabling non-volatile bipolar memory with long-term data retention. The device maintains stable resistive switching characteristics under mechanical bending and demonstrates promising applicability in transparent electronic systems.

Small-molecule semiconductors, defined by high crystallinity and strong intermolecular interactions, demonstrate superior carrier mobility coupled with narrow bandgap characteristics, rendering them ideal candidates for high-precision photoresponsive synaptic devices. Conjugated polymer semiconductors attain a balance among solution processability, mechanical flexibility, and broadband spectral response via molecular chain engineering combined with side-chain optimization, demonstrating distinct advantages in large-area flexible memristive systems. However, current material platforms remain constrained by challenges, including limited environmental stability and exciton recombination losses. Future investigations should focus on implementing application-specific material selection strategies while employing interface engineering approaches to tailor organic layer properties, thereby substantially improving device operational longevity.

4. Mechanism of Operation of Organic Photoelectric Synapses

The primary function of organic photoelectric synapses is to mimic the signaling and processing capabilities of biological synapses. Their operating principles include light signal capture, charge generation and separation, and charge transport and storage under the influence of an electric field. The common function of these mechanisms is to simulate the dynamic behavior of biological synapses in response to various light stimuli. In the following, we explore the working mechanisms of organic photoelectric synapses and analyze how these mechanisms interact to achieve complex synaptic functions.

4.1. Organic Ferroelectric/Electret-like Synaptic Devices

Ferroelectric materials determine the resistance state based on the polarization state of the internal electric dipole. Under an external electric field, the dipole shifts in the direction of polarization, enabling multistate resistance changes and enhancing device reliability. This polarization state is non-volatile. In 2018, Liu et al. fabricated a ferroelectric/electrochemical dual-modulated organic synapse, demonstrating multimodal synaptic plasticity by integrating ferroelectric and electrochemical mechanisms [66]. At low gate voltages, the electrochemical doping mechanism induces STP and electrochemical LTP, with signal durations ranging from milliseconds to minutes. At gate voltages exceeding the coercive field strength of the ferroelectric material, dipole switching events within the material result in ferroelectrically-induced LTP. This mechanism enables signal retention for up to 10,000 s, as illustrated in Figure 5. Unlike the volatile nature of the electrochemical mechanism, the ferroelectric mechanism provides reliable non-volatile memory performance, ensuring device stability over extended periods. The dual-modulation mechanism offers novel insights for the development of artificial neuromorphic systems. These systems are particularly advantageous for emulating the dynamic signaling responses and long-term memory of biological synapses.

Based on the polarization switching mechanism of ferroelectric materials, the device achieves ultra-long-term synaptic weight modulation, significantly surpassing the second-level storage capability of conventional electrochemical transistors. This characteristic provides a hardware foundation for simulating long-term memory consolidation in biological neural systems. The device demonstrates remarkable advantages in non-volatile memory and multimodal plasticity, offering novel insights for artificial visual perception systems. However, triggering ferroelectric LTP requires a high driving voltage, and the integration of multilayer heterostructures involves complex fabrication processes. Therefore, further improvements in low power consumption, process simplification, and durability are essential.

4.2. Heterojunction-like Synapses

The integration of organic semiconductors with diverse active materials to form heterojunctions can significantly expand the functional capabilities of devices. Specifically, these heterojunctions can enhance the photosensitivity of detectors [67,68] and broaden the spectral response range of devices [69,70]. Through meticulous design of the energy band structure of these heterojunctions, it is possible to achieve robust memory functionality, which holds substantial significance for the advancement of synaptic transistors and optical storage devices. [71,72]. The arrangement of the energy bands in the heterojunction promotes efficient separation of electrons and holes, as electrons naturally migrate from regions of higher energy to regions of lower energy. This configuration leads to the formation of a built-in electric field. The built-in electric field facilitates the separation of photogenerated charge carriers (electron–hole pairs).

For instance, Li et al. proposed a vertical photosynaptic device based on poly [2-methoxy-5-(2-ethylhexyloxy)-1,4-phenylethynyl]]:[6,6]-phenyl C61 butyric acid methyl ester (MEH-PPV:PCBM) to enhance light-induced charge separation and transfer [73,74]. In MEH-PPV:PCBM bulk heterojunction devices, under a high electric field, electrons can attempt to penetrate from the high-potential region to the low-potential region through the potential barrier. Organic semiconductors contain numerous trap states that can capture and release charge carriers. The trapped charges have greater difficulty in jumping, leading to the formation of a memory charge and the development of a hysteresis window. When the electric field is sufficiently high, carriers trapped in the potential well can be released, increasing the number of free carriers and, consequently, the current. The conductivity of the device gradually increases with the rising concentration, resulting in a corresponding increase in the response current as a function of the applied voltage.

Wu et al. [75] fabricated phototransistor arrays based on chalcogenide quantum dots (CsPbBr₃ QDs) heterojunctions with organic semiconductors, specifically poly(2,5-(2-octyldodecyl)-3,6-diketopyrrolopyrrole-alt-5,5-(2,5-di(thien-2-yl)thieno[3,2-b]thiophene) (DPPDTT), which demonstrated excellent performance with remarkable photosensitivity, rapid response, and high light detection efficiency. As shown in Figure 6a,b, the photoresponse performance of the pure DPPDTT-based samples is characterized, with the device exhibiting higher photoresponsivity in air. It is likely that moisture and oxygen in the air induce electron trapping at deeper energy levels, thereby enhancing the photoresponse. However, a significant photoresponse was observed under vacuum in phototransistors with CsPbBr₃ QDs/DPPDTT heterojunctions (Figure 6c), indicating that the incorporation of CsPbBr₃ enhanced the devices’ photoresponsiveness. The UV–Vis–NIR absorption spectra in Figure 6d also show that the addition of QDs enhances light absorption at wavelengths below 520 nm. This enhancement is attributed to the formation of a heterojunction between DPPDTT and CsPbBr₃ QDs (Figure 6e), which improves the electrical conductivity of DPPDTT. As shown in Figure 6f, the steady-state photoluminescence (PL) spectra indicate that the luminescence intensity of the hybrid film formed from CsPbBr₃ QDs and DPPDTT is significantly lower at 520 nm compared to that of the QD film. This phenomenon suggests that the heterojunction formed between DPPDTT and CsPbBr3 QDs facilitates the efficient separation of photogenerated excitons. After light irradiation ceases, electrons within the CsPbBr₃ QDs gradually combine with holes in the conduction channel, resulting in a gradual weakening of the photocurrent and thus exhibiting a persistent photoconductive effect. The device thus replicates key functions of a wide range of biological synapses, including EPSC, PPF, and STP to LTP. Notably, the system achieves energy consumption as low as 27.9 aJ at an operating voltage of −0.0001 V.

Heterojunction-based synaptic devices, leveraging the synergistic effects between organic semiconductors and active materials, combined with optical microlithography for high-precision micropatterning, demonstrate remarkable advantages in photoresponsivity, ultralow power consumption, and ultrahigh integration density. These attributes provide a bio-inspired hardware foundation for neuromorphic computing. Benefiting from a broad spectral response range, the devices enable multi-wavelength optical signal processing and hold potential for integration with CMOS circuits or flexible substrates, thereby advancing applications in biomimetic vision chips and wearable sensing systems. However, current challenges persist: (1) the optical microlithography process relies on multi-step photolithography and solvent compatibility control, resulting in high fabrication complexity; (2) the dynamic response capability remains limited, hindering the detection of ultrafast motion signals. Future research could focus on heterojunction designs to enhance charge carrier mobility, coupled with gradient energy band engineering to optimize interfacial charge transfer kinetics, thereby addressing the response speed bottleneck. Furthermore, the development of universal patterning strategies may reduce fabrication complexity and facilitate scalable manufacturing of large-area flexible electronics.

4.3. Schottky Barrier Synaptic Devices

A Schottky barrier is formed at the interface between a metal and a semiconductor, restricting the flow of electrons and holes and thereby creating a unidirectional conductive channel. The height of the Schottky barrier can be modulated by the density of trapped electrons, as the injected current exhibits an exponential dependence on both the Schottky barrier variation and the trapped electron density. In an organic photodiode, light-induced trapped charges play a key role in modulating the Schottky barrier. Upon light irradiation, photons are absorbed by the organic crystalline semiconductor, resulting in the generation of numerous excitons (electron–hole pairs). Under an external electric field, these photogenerated holes migrate rapidly into the depletion region of the Schottky junction, lowering the barrier height and increasing channel conductance, thereby inducing a photocontrolled change in conductivity.

Yang et al. [76] demonstrated an optically modulated organic planar diode, with the device structure schematically illustrated in Figure 7a. The conducting layer comprised a bilayer organic crystalline semiconductor, C8-BTBT. Under a negative bias, the resting current is significantly suppressed due to the high Schottky barrier. Upon light irradiation, the bilayer C8-BTBT absorbs photons, generating numerous excitons (electron–hole pairs). An external electric field subsequently drives the holes to migrate into the depletion region of the Schottky junction, thereby reducing the barrier height and enhancing channel conductance. Consequently, the Schottky barrier diode exhibits distinctive characteristics, as illustrated in Figure 7b. The I–V curve demonstrates the typical rectifying behavior of a diode, achieving a rectification ratio of approximately 10⁵ in the absence of light and an extremely low reverse current of approximately 10–11 A. Under illumination at 9 µW/cm², the device achieves a light-to-dark current ratio (I light/I dark) of approximately 100 at an operating voltage of −2 V, demonstrating effective photoswitching behavior. After the light is removed, the current gradually decays, as shown in Figure 7c, effectively mimicking synaptic plasticity. Additionally, the device exhibits notable photosensitivity of approximately 9 µW/cm² and ultralow energy consumption of around 13.6 pJ.

Schottky barrier-based synaptic devices leverage the unique optoelectronic properties of organic crystals, enabling precise emulation of bio-inspired synaptic behaviors through light-modulated Schottky barrier dynamics. These devices exhibit exceptional performance in photosensitivity, energy efficiency, and memory retention. The planar diode architecture offers simplicity and CMOS compatibility, facilitating integration into flexible electronics. However, challenges remain, including limited spectral response and interface defects introduced during multilayer organic film fabrication. Future research should focus on hybrid material heterojunctions and interface engineering to broaden spectral coverage and suppress trap states. Such advancements will expand their applicability in light-controlled intelligent systems, such as neuromorphic vision chips and adaptive robotic skins.

4.4. Trap-Modulated Synaptic Devices

Charge transport is influenced by the energy levels of traps within the material. At low voltages, charge transport is predominantly governed by shallow traps, a process referred to as shallow trap-modulated space charge-limited current. With increasing applied voltage, charge begins to occupy deeper trap energy levels, causing a transition in the transport mechanism from shallow trap modulation to deep trap modulation. The trap energy levels in the memristor capture and release charges, and this process induces changes in the device’s conductance state. When charges are captured by the traps, the device’s conductance increases; conversely, when charges are released, the conductance decreases. This mechanism enables the memristor to emulate changes in biological synaptic weights, demonstrating synaptic plasticity. Due to the presence of traps, the conductance–voltage relationship in a memristor is typically nonlinear. This nonlinear characteristic complicates signal processing in circuit applications but simultaneously provides opportunities for modeling synaptic functions.

Pei et al. [77] utilized the organic semiconductor material bis [1]-benzothieno[2,3-d; 2′,3′-d’]naphtho[2,3-b;6,7-b’]dithiophene (BBTNDT) with structural inhomogeneities (nano-budding structures) to develop flexible nonvolatile optical memory devices based on organic field-effect transistors. This approach successfully reduced fabrication complexity and cost. The resulting device achieved an average mobility of up to 7.7 cm²/V·s, a photoresponse of 433 A W⁻¹, a retention time exceeding 6 h, and a current-switching ratio greater than 10⁶. Under the combined effects of light and a positive gate voltage, the nano-budding structure acts as a charge capture center. Electrons are trapped in deep energy-level states, while holes accumulate in the channel. This process induces a positive shift in the threshold voltage (VTH), activating the device’s storage state. The deep energy-level traps significantly inhibit the release of trapped electrons, enabling the device to maintain its storage state during the holding phase. The erasure phase is initiated by applying a negative gate voltage, which releases the trapped electrons from the deep traps and reinjects them into the channel. This decreases the hole concentration in the channel, weakens the device’s conductivity, restores the threshold voltage to its initial state, and deactivates the storage state. Furthermore, the device was extended to fabricate a 16 × 16 optical storage transistor array on a 12 µm-thick flexible polyethylene naphthalene dicarboxylate (PEN) substrate. This array demonstrated the capability to capture and store two-dimensional optical images. Notably, it exhibited excellent mechanical stability, with mobility decreasing by only 8% after enduring 10,000 bending cycles with a bending radius of 3.2 mm. These results highlight the potential of highly integrated optical storage devices for flexible electronics, paving the way for future wearable devices and intelligent electronic systems.

Trap-engineered synaptic devices achieve high-performance non-volatile optical memory by leveraging intrinsic charge trap states induced by structural defects in organic semiconductor thin films. By eliminating the need for conventional floating-gate structures, these devices feature a simplified architecture, low fabrication costs, mechanical flexibility, and scalability for large-area array integration. However, performance optimization faces critical challenges: limited operational flexibility—memory effects require combined optical and gate stimuli, leading to elevated power consumption; high process sensitivity—defect formation critically depends on thin-film deposition temperatures, necessitating stringent control over processing conditions; and circuit optimization—leakage currents must be suppressed to enhance reliability. Addressing these challenges will facilitate their practical applications in flexible optoelectronics and optical encryption systems.

4.5. Electrolyte-Gated Layer-like Synaptic Devices

The electrolyte layer in this device functions as a gating medium, facilitating the regulation of ion migration through variations in gate voltage. Upon applying a voltage to the gate electrode, ions in the electrolyte migrate and accumulate at the interface with the organic semiconductor, altering the charge carrier concentration and the conductivity of the semiconductor layer. Coupled with photochemical control of ion permeability, this mechanism enables dynamic tuning of the device’s conductivity and memory characteristics.

For instance, researchers have precisely modulated ion permeability in vertical organic channels using UV-initiated photochemical reactions with photoreactive cross-linking agents, such as 2Bx, in combination with the organic semiconductor poly(3-hexylthiophene) (P3HT) [78]. Upon UV irradiation, a crosslinked structure forms, altering the intermolecular free volume of P3HT and reducing ion penetration pathways, thus limiting ion mobility and regulating permeability throughout the channel. By adjusting the duration and intensity of UV irradiation, the ion permeability in vertical organic channels can be precisely tuned, enabling meticulous control of synaptic weight modulation. STP simulates a transient change in synaptic weight, as shown in Figure 8a (top), where the current quickly decays to its baseline level due to the high mobility of ions in the substantial free volume of P3HT. LTP simulates a durable change in synaptic weight, as shown in Figure 8a (bottom), where the current gradually decays due to the crosslinked structure formed by 2BX and P3HT, which restricts ion mobility. The retention characteristic differences between STP and LTP become increasingly evident with a higher number of applied pulses. Figure 8b illustrates the currents of short-term and long-term synaptic devices subjected to 10 consecutive −3 V pulses. Short-term devices exhibit higher peak currents that rapidly return to baseline within 60 s, while long-term devices show lower peak currents but more pronounced retention. Selective UV exposure through photolithography enables precise definition of crosslinker-irradiated regions, creating synaptic regions with distinct memory properties on the same organic semiconductor channel, thus facilitating controlled modulation of synaptic weights. This represents an innovative strategy for developing advanced neuromorphic computing devices and processing extensive biometric datasets.

This study presents an electrolyte-gated organic synaptic device based on photochemically regulated ion permeability, which enables the physical programmability of synaptic weights without relying on complex circuit designs or software algorithm adjustments. The device exhibits excellent performance in dynamic range, high linearity, and cycling stability under low operating voltages. While it demonstrates significant advantages in low power consumption, dynamic reconfigurability, and biomedical compatibility, the following challenges require attention. Organic semiconductor materials are susceptible to moisture and oxygen under prolonged environmental exposure, potentially leading to degradation of ion migration pathways; photolithography and selective exposure processes may increase the complexity of scalable fabrication. Future research will focus on developing encapsulation technologies to enhance environmental stability, thereby advancing its practical applications in flexible electronics and edge computing.

4.6. Photonically Modulated Electrochemically Doped Synapse-like Devices

To date, field-effect transistors are among the main platforms for organic optoelectronic synapses [79,80,81]. However, there are some inherent barriers to realizing such devices for neuromorphic computing. For example, realizing linear multilevel conductance states in FETs remains difficult because of the charge shielding effect at the channel–dielectric interface. Photon-modulated electrochemical doping incorporates a photoactive layer consisting of a donor–acceptor (DA) heterojunction interface in the channel of organic field-effect transistors (OECTs). When light strikes the photoactive layer, the light is absorbed by the donor–acceptor heterojunction, generating charge carriers (electron–hole pairs). The photoinduced charge carriers perturb the electrochemical doping process in the channel, i.e., the charge distribution in the channel is altered by the photoinduced charge carriers. To maintain electroneutrality, the production of photogenerated charge carriers induces charge compensation by injecting the opposite type of ions from the electrolyte into the channel, leading to an increase in the drain current. When the light is turned off, the presence of anions around the doped organics prohibits immediate charge recombination, leading to slow current decay and contributing to a nonvolatile memory current. A schematic diagram of photon-modulated electrochemical doping is shown in Figure 9b.

Chen et al. [64] proposed constructing an artificial retina consisting of a series of optoelectronic synaptic devices (schematically shown in Figure 9a). The current increases significantly when the device is illuminated by a white light pulse; after the light pulse is turned off, the current gradually decreases to approximately 6.0 microamps and remains at that level, as shown in Figure 9c. The magnitude of the difference between the steady-state currents before and after optical programming is defined as the nonvolatile photonic memory, which can be gradually erased by applying a reverse gate voltage pulse, demonstrating that the device is capable of optical signal writing and electrical signal erasing. To assess the linear multilevel conductance states, we subjected the device to three cycles of approximately 280 successive light pulses. Each pulse had a duration of 0.4 s, with an interval of 0.23 s between pulses, as illustrated in Figure 9d. This figure demonstrates the stability of the write and erase cycles. Upon zooming in for a detailed view, the distinct nonvolatile conductance states are clearly discernible. Hence, light acts as a presynaptic stimulus, triggering an electrical response in the postsynaptic element. The consequent diffusion of ions, driven by photons, results in synaptic activity and memory effects.

This device functions via a photo–ion–electron coupling mechanism. The fundamental mechanism involves photoexcited charge carriers perturbing the electrochemical doping process, which induces ion migration from the electrolyte into the organic channel to realize non-volatile conductance modulation. It exhibits breakthrough advantages in ultralow-power operation, biocompatible interfaces, and synergistic optoelectronic signal processing capabilities. However, practical implementation requires addressing persistent challenges, including organic material stability under prolonged operation, cost-effective manufacturing processes, and response time optimization. Future developments should focus on designing chemically robust non-fullerene acceptors and integrating scalable fabrication technologies, thereby accelerating practical deployment in neuromorphic vision processors and flexible bioelectronic systems.

This chapter conducted a systematic investigation of the core working mechanisms, merits, and constraints of six device categories. Each architecture achieves biomimetic synaptic plasticity emulation through simulated dynamic responses and plasticity modulation mechanisms, exhibiting significant potential for neuromorphic computing and bionic systems. However, practical implementation remains constrained by challenges, including high operational voltage thresholds, fabrication complexity, material degradation, and limited temporal resolution. Notably, Schottky barrier synaptic devices replicate biological synaptic plasticity via light-modulated barrier height adjustment, combining high photosensitivity with ultralow energy consumption. Their non-volatile memory behavior inherently supports long-term memory functionality. Furthermore, these devices employ a simplified planar diode configuration compatible with established organic crystal processing techniques, drastically reducing manufacturing complexity while maintaining CMOS-compatible operating voltages for direct silicon circuit integration. Unlike the environmental susceptibility of alternative architectures, Schottky barrier devices strike an optimal equilibrium among synaptic functionality, manufacturing feasibility, and system compatibility, establishing them as prime candidates for neuromorphic vision chips and flexible electronics. Future advancements may focus on heterojunction integration for spectral range expansion and interface defect mitigation to enhance performance and enable practical deployment.

5. Performance Metrics

The threshold voltage and switching ratio are the basic parameters for evaluating the performance of a transistor device. The lower the threshold voltage is, the lower the voltage at which the transistor can turn on, which usually means lower energy consumption and higher integration. The switching ratio is the ratio of the transistor’s current in the on state to that in the cut-off state. A high switching ratio means better stability, immunity to interference, lower leakage current, and greater load-driving capability. For organic optoelectronic synapses, a lower switching threshold voltage, faster response time, better retention performance, and lower power consumption are needed, and even the ability to realize biological synaptic function and plasticity, of which the memory function is the most important. Therefore, the criteria for evaluating the performance of synaptic devices are the EPSC, inhibitory postsynaptic current (IPSC), STP, LTP, PPF, and conversion from the STP to LTP.

Relaxation times, divided into rising and falling components, are crucial metrics for photoconductance performance. A short relaxation time during the photoconductance increase indicates a strong photoresponse, enabling the photoelectric synapse to respond rapidly to external light stimuli. Conversely, the downward relaxation process describes the return of the photoelectric synapse from the photoexcited state to the initial dark state after light pulse stimulation ends. The relaxation time constant (τ) quantifies this process, reflecting the memory capacity of the device. A long τ indicates strong memory retention, as the synapse maintains state changes induced by light for extended periods. In contrast, a short τ suggests weaker memory, with the synapse quickly returning to its initial state.

5.1. Transition from STP to LTP

The efficiency of synaptic signaling is influenced by factors such as the synaptic gap size, neurotransmitter release levels, and receptor density on the postsynaptic membrane. Synaptic strength, which remains constant without external stimuli, changes in response to signals from pre- and postsynaptic neurons. This ability to modify synaptic strength, termed synaptic plasticity, forms the biomolecular basis of memory and learning in the brain. Modeling synaptic plasticity is thus critical for enabling neuromorphic computing.

Synaptic plasticity is categorized into STP and LTP. Memory training with optical and electrical signals can induce a transition from STP to LTP, allowing sustained increases in synaptic strength and facilitating long-term information storage [6,82,83]. As shown in Figure 10, LTP forms the foundation of memory. Human memory transitions from short-term to long-term with repeated exposure, extended duration, or higher frequency of stimuli. This transition was simulated by adjusting the pulse number (Figure 10a), pulse width (Figure 10b), and pulse frequency (Figure 10c). These adjustments resulted in a dramatic increase in EPSC, with prolonged decay times to the original state, indicating a transition from STP to LTP. The slow decay process enhances information retention, demonstrating the device’s capability for long-term memory and its potential to simulate complex memory functions under optical signal control.

5.2. Learning Memory Behavior Simulation

The simulation of learning and memory under light stimulation replicates key processes such as learning, forgetting, and relearning, effectively mimicking synaptic behaviors associated with memory formation and decay. A reduction in the time required to reach the same current level during subsequent light stimuli demonstrates the device’s optical memory capabilities. As shown in Figure 11a, after initial learning using 200 excitation pulses, the device undergoes a forgetting phase. During the second learning phase, the current is restored to its previous state with only 52 pulses. This simulation not only replicates the learning–forgetting–relearning behavior of biological synapses but also highlights that the required relearning time decreases with repeated learning cycles. This finding underscores the device’s effective learning ability and strong memory retention under light stimulation.

PPF, a hallmark of STP, represents the enhancement of synaptic response following two consecutive stimulus signals. The amplitude of the first current peak is denoted as A1, while the amplitude of the second current peak is A [84]. The PPF index is defined as the ratio of A2 to A1. In artificial photoelectric synaptic devices, continuous neuronal stimulation can be simulated by applying two successive light pulses. Following consecutive presynaptic stimuli, the postsynaptic response exhibits enhancement, a mechanism integral to learning and memory processes involving dynamic synaptic strength modulation. As shown in Figure 11b, the EPSC observed after the 10th light pulse (A10) exceeds that of the 2nd (A2) and 1st (A1) pulses. This current accumulation arises because photon-induced charge carriers in the channel cannot fully decay to the baseline before the next light pulse is applied. Consequently, a larger number of applied light pulses results in greater charge accumulation. The interval between successive light pulses plays a crucial role in determining the degree of current enhancement.

This chapter systematically analyzed the core performance metrics of organic optoelectronic synaptic devices. By optimizing threshold voltage (as low as −2 V) and on-off ratio (e.g., 10⁶), these devices exhibit significant advancements in ultralow power consumption, operational stability, and interference immunity, fulfilling the stringent energy efficiency and reliability requirements of neuromorphic computing systems. Through precise relaxation time regulation, an optimal equilibrium is achieved between rapid photoresponse and prolonged memory retention. For instance, the light-controlled current decay process accurately emulates biological synaptic plasticity spanning STP to LTP, enabling complex memory functions while validating their potential for adaptive intelligent systems. Future research should prioritize the co-optimization of three key aspects: novel material design (e.g., thermally stable non-fullerene acceptors, interface engineering, and scalable fabrication processes, collectively advancing device performance toward large-scale integration in biomimetic vision chips.

6. Advances in the Application of Organic Photoelectric Synapses

Organic optoelectronic synapses have emerged as a promising platform to emulate and even functionally extend the VSS in applications such as imaging memory, artificial nociceptor simulation, and motion perception.

6.1. Imaging Memory: Mimicry of the Visual System

In their 2022 study, Shi et al. successfully fabricated an 8 × 8 array of optical synaptic organic field-effect transistors (OFETs) on a flexible substrate using a full-solution printing technique [85]. The photosynapse consumes minimal energy (0.07–34 pJ). As illustrated in Figure 12a, the array demonstrates mechanical compatibility with biological tissues and can conformably adhere to an ocular model. The researchers systematically measured all 64 pixels in the low-operating-voltage OFET array under both planar and curved configurations using the characterization system depicted in Figure 12b. To evaluate the image recognition capability, an optical pattern “8” was projected onto the optical synaptic OFET array while monitoring the current response of individual pixels. Through quantitative conversion of the acquired current values into thermal maps, the researchers successfully reconstructed the “8” pattern in both physical states of the array, as visually confirmed in Figure 12c. Notably, the progressive enhancement of pattern clarity with increasing numbers of optical pulses (Figure 12d) reveals the array’s inherent capacity for image refinement and adaptive learning—a functional analog to the neural processes underlying familiar face recognition in biological visual systems. This evidence substantiates the potential of optical synaptic technology to emulate human ocular functionalities, thereby establishing a novel paradigm for developing bio-inspired artificial vision systems.

Organic semiconductor materials with narrow bandgaps (<1.6 eV) exhibit significant potential for near-infrared optoelectronic applications, such as fluorescence imaging, medical monitoring, remote detection, and optical communications. For instance, Yokota et al. [86] successfully fabricated flexible organic near-infrared optoelectronic devices with a thickness of just 15 µm, utilizing PMDPP3T:PCBM materials in conjunction with polycrystalline silicon circuits for signal readout. This innovation enabled imaging of fingerprints and vein patterns in the 850 nm near-infrared wavelength range.

In 2023, Liu et al. [87] demonstrated a hybrid optoelectronic transistor utilizing C8-BTBT and PM6, in which the channel layer exhibited broad spectral response characteristics via molecular heterojunction design. Compared with pure C8-BTBT devices, the hybrid structure expanded the optical response range to encompass red, green, and blue primary color bands, achieving dynamic switching between STP and LTP through synergistic control of light intensity and pulse duration. The experimental results revealed that the device maintained excellent stability with non-volatile memory capabilities under prolonged light stimulation. Moreover, an 8 × 8 synaptic array implemented with this device accomplished multicolor light-controlled letter encoding and tricolor information storage, with photocurrent spatial distribution showing precise correspondence to the input light patterns.

6.2. Sensing Function: Motion Sensing

The optical sensor array encodes and interprets the trajectory of an object’s motion in real time. By converting optical signals into electrical signals and emulating the response mechanism of biological synapses, the array enables efficient processing of dynamic visual information. Wu [75] developed and evaluated motion-aware neural networks using a custom image dataset. Object motion trajectories were encoded into the color depth and shape of current heatmaps, providing a visual representation of motion information. The neuromorphic photosensor demonstrates varying current responses to distinct light stimulation sequences by compiling the instantaneous currents of multiple light pulses into unified current sequences (Figure 13b). Images derived from the sensor array encode information regarding the timing of light stimuli. A motion-perception neural network determines an object’s direction of motion by analyzing these images and decoding the embedded information (Figure 13a). To train and validate the neural network, a custom image dataset was constructed, encompassing motions in upward, downward, leftward, and rightward directions, as well as stationary states. Following 25 training sessions, the recognition accuracy of the neural network surpassed 90% (Figure 13c). After 100 iterations (Figure 13d), the classification accuracy further improved to 97%, highlighting the potential of neuromorphic optoelectronic sensor arrays.

6.3. Harm Receptors

Harm receptors play a vital role in the human body, serving as early warning systems against potentially damaging stimuli, such as extreme temperatures and pressures. Replicating the functionality of injury receptors in electronic devices remains a significant challenge for researchers developing neuromorphic systems. Kumar M et al. [88] proposed a UV-activated artificial injury receptor that mimics the response mechanism of the human visual system. The injury receptors demonstrated behaviors resembling “hyperalgesia” and “allodynia”, represented by the horizontal and vertical arrows in Figure 14a, respectively. High-intensity ultraviolet (HUV) irradiation induces a damage state in pain receptors. In Figure 14b, the black and red lines represent the photocurrents measured after HUV irradiation (damage state) and before irradiation (no damage state), respectively. HUV irradiation induces a damage state that alters the response profile of nociceptors. Prior to HUV exposure, undamaged nociceptors exhibit lower sensitivity to low-intensity UV light, resulting in reduced photocurrent values. However, under HUV irradiation, nociceptors exhibit significant changes in their response. Even under low-intensity UV exposure, damaged nociceptors display heightened responses, characterized by increased photocurrent. This occurs because HUV irradiation fills the trap states, enabling a rapid UV response that reduces the threshold and enhances nociceptor responses, as reflected by increased photocurrent values. Under self-biased conditions, the device demonstrated diverse injury receptor-like properties in response to UV stimuli, such as threshold responses, attenuation, anomalous nociception, and nociceptive sensitization, offering a novel strategy for energy-efficient neuromorphic device operation.

7. Conclusions and Outlook

Organic optoelectronic synapses, as emerging neuromorphic devices, have demonstrated immense potential in emulating the complex functionalities of biological neural systems. However, despite significant advancements, several critical challenges remain unresolved.

Although various organic semiconductors (OSCs) have been employed in optoelectronic synapses, the development of novel OSCs tailored to diverse application scenarios is still necessary. In-depth exploration of structure-property relationships in OSCs could unlock new opportunities leveraging their exceptional optoelectronic properties. Future efforts should focus on identifying narrow-bandgap materials with high carrier mobility and broad spectral response (covering ultraviolet to near-infrared regions), such as non-fullerene acceptors. Integration with quantum dots or perovskite heterojunctions may further enhance light absorption efficiency and charge separation capabilities.

Practical applications require reducing fabrication complexity and transitioning from discrete devices to integrated, arrayed systems. Simplified manufacturing techniques, such as inkjet printing or roll-to-roll (R2R) techniques, should be prioritized. Multimodal sensing fusion (e.g., light–electricity–force synergy) and hybrid CMOS/flexible integration (e.g., bio-inspired skin with multifunctional perception) are critical for scalable deployment.

Environmental sensitivity (e.g., oxygen/moisture degradation) remains a bottleneck. Future research should enhance device stability through interface engineering, such as atomic layer deposition for interface trap passivation and gradient bandgap engineering to improve cycling endurance.

Applications in healthcare (e.g., minimally invasive neural interfaces), bio-inspired robotics, and human–machine interfaces require further exploration. For instance, light-controlled neuromodulation via optoelectronic synapses could enable precise, low-invasive neuroprosthetics, offering new paradigms for brain-machine interfaces.

While organic optoelectronic synapses are still in their nascent stage, their potential to underpin next-generation, energy-efficient, and adaptive brain-inspired systems is undeniable. Bridging the gap between laboratory prototypes and real-world applications will require concerted efforts in material innovation, process scalability, and cross-disciplinary collaboration.

Author Contributions

Writing original draft preparation: M.H.; Writing review and editing: M.H. and X.T. Funding acquisition: X.T. All authors have read and agreed to the published version of the manuscript.

Institutional Review Board Statement

Not applicable.

Informed Consent Statement

Not applicable.

Data Availability Statement

Not applicable.

Conflicts of Interest

The authors declare no conflicts of interest.

Footnotes

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Figures

Figure 1 (a) Schematic diagram of the flow of information processing in the von Neumann and human brains. (b) Comparison of the working mechanism of a biological nervous system with that of an artificial photonic synapse [19].

Figure 2 The field of optoelectronic synapses is summarized in terms of device structure types, material types, device performance metrics, and innovative applications.

Figure 3 Schematic diagram of different device structures. (a) Two-terminal device structure. (b) Three-terminal device structure.

Figure 4 A summary of organic semiconductor materials applied in optoelectronic synaptic devices is presented, with the molecular structure formulas arranged according to their optical absorption ranges.

Figure 5 Mechanisms of STP (left), electrochemical doping-induced LTP (center), and ferroelectric switching-induced LTP (right) in synaptic transistors [66]. STP signals are generated through the electrochemical doping process when the applied gate field strength remains below the coercive field strength (Ecd) of P(VDF-TrFE). Repeated presynaptic pulses lead to the gradual accumulation of the electrochemical doping effect, resulting in LTP. When the gate electric field strength exceeds Ecd, the dipole moments within the ferroelectric material undergo directional rearrangement, thereby triggering ferroelectric-induced LTP. The parameter d, labeled in the figure, represents the thickness of the P(VDF-TrFE) layer (Source and drain electrodes are denoted in yellow, while the gate dielectric layer is color-coded in blue).

Figure 6 Light response mechanism of CsPbBr₃ QD/DPPDTT PHJ-based OPTs [75]. (a) Transmission curve of the DPPDTT-based device in air. (b) Transmission curve of the DPPDTT-based device in vacuum. (c) Transmission curve of the CsPbBr₃ QD/DPPDTT PHJ-based OPT in vacuum. (d) UV–Vis–NIR absorption spectra of DPPDTT, CsPbBr₃ QD, and CsPbBr₃ QD/DPPDTT PHJ-based OPTs. (e) Simplified energy band diagrams of DPPDTT and CsPbBr3 QD films. (f) Steady-state PL spectra of CsPbBr₃ QD films and CsPbBr₃ QD/DPPDTT hybrid films.

Figure 7 Device structure and electrical behavior [76]. (a) Schematic representation of the device structure, featuring a bilayer C8-BTBT crystalline semiconductor film. (b) I-V characteristics measured under three conditions: in the dark, under illumination (9 μW cm⁻²), and post-illumination (30 s and 90 s). (c) I–T responses recorded under identical illumination conditions and at an operating voltage of −2 V.

Figure 8 Characterization of the ion gel-gated vertical crossbar synapse array [78]. (a) Excitatory current responses of the STP (top) and LTP induced by the application of various VWC. (b) Real-time current responses of the STP (black line) and LTP (red line) under 20 consecutive potentiation pulses (VWC = −3 V).

Figure 9 Electrochemical doping for photonic modulation [64]. (a) Schematic diagram of the photon-modulated OECT. (b) Schematic diagram of the photon-modulated electrochemical doping mechanism. (c) Instruments with optical write (purple region) and voltage erase (gray region) devices with characteristic photonic responses. (d) Optical storage write (VG= −0.6 V, VDS = −0.8 V) and voltage erase (VG = 0.4 V, source–drain voltage (VDS) = −0.8 V) for three representative cycles.

Figure 10 Transition of the memristor from short-term to long-term plasticity [64]. (a) Transition induced by increasing the number of light pulses (P = 7.0 mW/cm², pulse duration tp = 0.4 s, interval time Δt = 0.2 s). (b) Transition achieved by lengthening the pulse width. (c) Transition facilitated by decreasing the pulse interval. In comparison to short-term memory (STM), long-term memory (LTM) exhibits a slower decay of EPSC, taking longer to return to its original level.

Figure 11 Behavioral simulation of learning memory in the presence of light [64]. (a) Optical learning–forgetting–optical relearning behavior. (b) PPF characteristics.

Figure 12 Image recognition and reinforcement learning [85]. (a) An image of a flexible optical synaptic OFET array attached to a human eye model, along with a detailed design of the array. (b) Schematic diagram of the measurement process. (c) Measurement results of the “8” pattern in both the flat and curved states. (d) Measurement results of the “8” pattern in the initial state and after 1, 5, 25, and 50 pulse cycles.

Figure 13 Potential applications of neuromorphic optoelectronic sensor arrays in motion sensing [75]. (a) Schematic representation of a neuromorphic photosensor array integrated with a neural network. (b) Histogram depicting the response currents of neuromorphic photosensors under four distinct light pulse sequences. (c) Correlation between classification accuracy and the number of training epochs in the neural network. (d) Confusion matrix illustrating classification results after 100 training iterations.

Figure 14 Analogy between injury receptors and the human eye [88]. (a) Schematic representation of allodynia and nociceptive hypersensitivity features as a function of increasing stimulus intensity under both normal (uninjured) and injured conditions. (b) Experimental measurements of the device’s photocurrent before and after high-intensity ultraviolet (HUV) irradiation, demonstrating allodynia and nociceptive hypersensitivity features.

References

1. Zidan, M.A.; Strachan, J.P. The future of electronics based on memristive systems. Nat. Electron.; 2018; 1, pp. 22-29. [DOI: https://dx.doi.org/10.1038/s41928-017-0006-8]

2. Van, B.Y.; Melianas, A. Organic electronics for neuromorphic computing. Nat. Electron.; 2018; 1, pp. 386-397.

3. Tang, Y.; Nyengaard, J.R. Total regional and global number of synapses in the human brain neocortex. Synapse; 2001; 41, pp. 258-273. [DOI: https://dx.doi.org/10.1002/syn.1083] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/11418939]

4. Alibart, F.; Pleutin, S. An organic nanoparticle transistor behaving as a biological spiking synapse. Adv. Funct. Mater.; 2010; 20, pp. 330-337. [DOI: https://dx.doi.org/10.1002/adfm.200901335]

5. Lai, Q.X.; Zhang, L. Ionic/Electronic hybrid materials integrated in a synaptic transistor with signal processing and learning functions. Adv. Mater.; 2010; 22, pp. 2448-2453. [DOI: https://dx.doi.org/10.1002/adma.201000282] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/20446309]

6. Ohno, T.; Hasegawa, T. Short-term plasticity and long-term potentiation mimicked in single inorganic synapses. Nat. Mater.; 2011; 10, pp. 591-595. [DOI: https://dx.doi.org/10.1038/nmat3054]

7. Josberger, E.E.; Deng, Y.X. Two-terminal protonic devices with synaptic-like short-term depression and device memory. Adv. Mater.; 2014; 26, pp. 4986-4990. [DOI: https://dx.doi.org/10.1002/adma.201400320]

8. Yang, J.Q.; Wang, R. Neuromorphic Engineering: Neuromorphic Engineering: From Biological to Spike-Based Hardware Nervous Systems. Adv. Mater.; 2020; 32, 2070392. [DOI: https://dx.doi.org/10.1002/adma.202070392]

9. Choi, S.; Yang, J. Emerging memristive artificial synapses and neurons for energy-efficient neuromorphic computing. Adv. Mater.; 2020; 32, 2004659. [DOI: https://dx.doi.org/10.1002/adma.202004659]

10. Qi, J.; Li, C. Stability of delayed memristive neural networks with time-varying impulses. Cogn. Neurodyn.; 2014; 8, pp. 429-436. [DOI: https://dx.doi.org/10.1007/s11571-014-9286-0]

11. Indiveri, G.; Chicca, E. A VLSI array of low-power spiking neurons and bistable synapses with spike-timing dependent plasticity. IEEE Trans. Neural Netw.; 2006; 17, pp. 211-221. [DOI: https://dx.doi.org/10.1109/TNN.2005.860850] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/16526488]

12. Rachmuth, G.; Poon, C.S. Transistor analoguesgs of emergent iono-neuronal dynamics. HFSP J.; 2008; 2, pp. 156-166. [DOI: https://dx.doi.org/10.2976/1.2905393] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/19404469]

13. Liu, Y.H.; Zhu, L.Q. Freestanding artificial synapses based on laterally proton-coupled transistors on chitosan membranes. Adv. Mater.; 2015; 27, pp. 5599-5604. [DOI: https://dx.doi.org/10.1002/adma.201502719]

14. Lont, J.B.; Guggenbuhl, W. CMOS implementation of a multilayer perceptron with nonlinear synapses. IEEE Trans. Neural Netw.; 1992; 3, pp. 457-465. [DOI: https://dx.doi.org/10.1109/72.129418]

15. Strukov, D.B.; Snider, G.S. The missing memristor found. Nature; 2008; 453, pp. 80-83. [DOI: https://dx.doi.org/10.1038/nature06932]

16. Münchenberger, J.; Reiss, G.; Thomas, A. A memristor based on current-induced domain-wall motion in a nanostructured giant magnetoresistance device. J. Appl. Phys.; 2012; 111, 07D303. [DOI: https://dx.doi.org/10.1063/1.3671438]

17. He, W.; Huang, K.J. Enabling an integrated rate-temporal learning scheme on memristor. Sci. Rep.; 2014; 4, 4755. [DOI: https://dx.doi.org/10.1038/srep04755] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/24755608]

18. Li, Y.; Zhong, Y.P. Activity-dependent synaptic plasticity of a chalcogenide electronic synapse for neuromorphic systems. Sci. Rep.; 2014; 4, 4906. [DOI: https://dx.doi.org/10.1038/srep04906]

19. Zhang, Q.; Jin, T. Organic field effect transistor-based photonic synapses: Materials, devices, and applications. Adv. Funct. Mater.; 2021; 31, 2106151. [DOI: https://dx.doi.org/10.1002/adfm.202106151]

20. Yi, W.; Tsang, K.K. Biological plausibility and stochasticity in scalable VO2 active memristor neurons. Nat. Commun.; 2018; 9, 4661. [DOI: https://dx.doi.org/10.1038/s41467-018-07052-w]

21. Oh, S.; Shi, Y. Energy-efficient Mott activation neuron for full-hardware implementation of neural networks. Nat. Nanotechnol.; 2021; 16, pp. 680-687. [DOI: https://dx.doi.org/10.1038/s41565-021-00874-8] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33737724]

22. Boyn, S.; Grollier, J. Learning through ferroelectric domain dynamics insolid-state synapses. Nat. Commun.; 2017; 8, 14736. [DOI: https://dx.doi.org/10.1038/ncomms14736] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/28368007]

23. Shi, J.; Ha, S.D. A correlated nickelate synaptic transistor. Nat. Commun.; 2013; 4, 2676. [DOI: https://dx.doi.org/10.1038/ncomms3676]

24. Mou, X.; Tang, J. Analogueg memristive synapse based on topotactic phase transition for high-performance neuromorphic computing and neural network pruning. Sci. Adv.; 2011; 7, eabh0648. [DOI: https://dx.doi.org/10.1126/sciadv.abh0648]

25. Zhang, H.T.; Park, T.J. Perovskite neural trees. Nat. Commun.; 2020; 11, 2245. [DOI: https://dx.doi.org/10.1038/s41467-020-16105-y]

26. Ryu, H.; Kim, S. Implementation of a reservoir computing system using the short-term effects of Pt/HfO₂/TaOx/TiN memristors with self-rectification. Chaos Solitons Fractals; 2021; 150, 111223. [DOI: https://dx.doi.org/10.1016/j.chaos.2021.111223]

27. Xu, R.; Jang, H. Vertical MoS2 double-layer memristor with electrochemical metallization as an atomic-scale synapse with switching thresholds approaching 100 mV. Nano Lett.; 2019; 19, pp. 2411-2417. [DOI: https://dx.doi.org/10.1021/acs.nanolett.8b05140] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30896171]

28. Bessonov, A.A.; Kirikova, M.N. Layered memristive and memcapacitive switches for printable electronics. Nat. Mater.; 2015; 14, pp. 199-204. [DOI: https://dx.doi.org/10.1038/nmat4135]

29. Zhang, F.; Zhang, H. Electric-field induced structural transition in vertical MoTe2- and Mo1–xWxTe2-based resistive memories. Nat. Mater.; 2019; 18, pp. 55-61. [DOI: https://dx.doi.org/10.1038/s41563-018-0234-y]

30. Seo, S.; Jo, S.H. Artificial optic–neural synapse for colored and color-mixed pattern recognition. Nat. Commun.; 2018; 9, 5106. [DOI: https://dx.doi.org/10.1038/s41467-018-07572-5]

31. Shi, Y.; Yuan, B. Electronic synapses made of layered two-dimensional materials. Nat. Electron.; 2018; 1, pp. 458-465. [DOI: https://dx.doi.org/10.1038/s41928-018-0118-9]

32. Sharbati, M.T.; Du, Y. Low-power, electrochemically tunable graphene synapses for neuromorphic computing. Adv. Mater.; 2018; 30, 1802353. [DOI: https://dx.doi.org/10.1002/adma.201802353]

33. Kireev, D.; Liu, S. Metaplastic and energy-efficient biocompatible graphene artificial synaptic transistors for enhanced accuracy neuromorphic computing. Nat. Commun.; 2022; 13, 4386. [DOI: https://dx.doi.org/10.1038/s41467-022-32078-6] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/35902599]

34. Abdollahramezani, S.; Hemmatyar, O. Electrically driven reprogrammable phase-change metasurface reaching 80% efficiency. Nat. Commun.; 2022; 13, 1696. [DOI: https://dx.doi.org/10.1038/s41467-022-29374-6]

35. Raifuku, I.; Chao, Y.P. Halide perovskite for low-power consumption neuromorphic devices. EcoMat; 2021; 3, 12142. [DOI: https://dx.doi.org/10.1002/eom2.12142]

36. Hao, D.; Yang, Z. Recent developments of optoelectronic synaptic devices based on metal halide perovskites. Adv. Funct. Mater.; 2023; 33, 2211467. [DOI: https://dx.doi.org/10.1002/adfm.202211467]

37. Hong, X.; Liu, X. Review on metal halide perovskite-based optoelectronic synapses. Photon. Res.; 2023; 11, pp. 787-807. [DOI: https://dx.doi.org/10.1364/PRJ.480057]

38. Satapathi, S.; Raj, K. Halide-perovskite-based memristor devices and their application in neuromorphic computing. Phys. Rev. Appl.; 2022; 18, 017001. [DOI: https://dx.doi.org/10.1103/PhysRevApplied.18.017001]

39. Chen, S.; Huang, J. Recent advances in synaptic devices based on halide perovskite. ACS Appl. Electron. Mater.; 2020; 2, pp. 1815-1825. [DOI: https://dx.doi.org/10.1021/acsaelm.0c00180]

40. Chae, K.; Lee, S.; Phung, T.Q. Mechanical failures of Two-Dimensional materials on polymer substrates. Appl. Surf. Sci.; 2022; 605, 154736. [DOI: https://dx.doi.org/10.1016/j.apsusc.2022.154736]

41. Wang, J.; Liang, X.; Xie, J. Complete Solution-Processed Semitransparent and Flexible Organic Solar Cells: A Success of Polyimide/Ag-Nanowires-and PH1000-Based Electrodes with Plasmonic Enhanced Light Absorption. Nanomaterials; 2022; 12, 3987. [DOI: https://dx.doi.org/10.3390/nano12223987]

42. Li, W.; Guo, F. Solution-processed wide-bandgap organic semiconductor nanostructures arrays for nonvolatile organic field-effect transistor memory. Small; 2018; 14, 1701437. [DOI: https://dx.doi.org/10.1002/smll.201701437] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/29165914]

43. Yu, X.; Zhang, H. Highly stretchable, ultra-soft, and fast self-healable conductive hydrogels based on polyaniline nanoparticles for sensitive flexible sensors. Adv. Funct. Mater.; 2022; 32, 2204366. [DOI: https://dx.doi.org/10.1002/adfm.202204366]

44. Wang, K.; Dai, S. Light-stimulated synaptic transistors fabricated by a facile solution process based on inorganic perovskite quantum dots and organic semiconductors. Small; 2019; 15, 1900010. [DOI: https://dx.doi.org/10.1002/smll.201900010] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/30740892]

45. Zhang, H.; Guo, X. Interface engineering of semiconductor/dielectric heterojunctions toward functional organic thin-film transistors. Nano Lett.; 2011; 11, pp. 4939-4946. [DOI: https://dx.doi.org/10.1021/nl2028798] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/22011136]

46. Niu, H.; Zhang, H. Micro-nano processing of active layers in flexible tactile sensors via template methods: A review. Small; 2021; 17, 2100804. [DOI: https://dx.doi.org/10.1002/smll.202100804]

47. Lee, Y.; Lee, T.W. Organic synapses for neuromorphic electronics: From brain-inspired computing to sensorimotor nervetronics. Acc. Chem. Res.; 2019; 52, pp. 964-974. [DOI: https://dx.doi.org/10.1021/acs.accounts.8b00553]

48. Xu, W.; Min, S.Y. Organic core-sheath nanowire artificial synapses with femtojoule energy consumption. Sci. Adv.; 2016; 2, e1501326. [DOI: https://dx.doi.org/10.1126/sciadv.1501326]

49. Wang, X.; Lu, Y. Highly sensitive artificial visual array using transistors based on porphyrins and semiconductors. Small; 2021; 17, 2005491. [DOI: https://dx.doi.org/10.1002/smll.202005491]

50. Leydecker, T.; Herder, M. Flexible nonnon-volatile optical memory thin-film transistor device with over 256 distinct levels based on an organic bicomponent blend. Nat. Nanotechnol.; 2016; 11, pp. 769-775. [DOI: https://dx.doi.org/10.1038/nnano.2016.87]

51. Lee, C.; Lee, S. A reconfigurable binary/ternary logic conversion-in-memory based on drain-aligned floating-gate heterojunction transistors. Nat. Commun.; 2023; 14, 3757. [DOI: https://dx.doi.org/10.1038/s41467-023-39394-5] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/37353504]

52. Sarkar, T.; Lieberth, K. An organic artificial spiking neuron for in situneuromorphic sensing and biointerfacing. Nat. Electron.; 2022; 5, pp. 774-783. [DOI: https://dx.doi.org/10.1038/s41928-022-00859-y]

53. Harikesh, P.C.; Yang, C.Y. Ion-tunable antiambipolarity in mixed ion–electron conducting polymers enables biorealistic organic electrochemical neurons. Nat. Mater.; 2023; 22, pp. 242-248. [DOI: https://dx.doi.org/10.1038/s41563-022-01450-8]

54. Hou, Y.X.; Li, Y. Large-scale and flexible optical synapses for neuromorphic computing and integrated visible information sensing memory processing. ACS Nano; 2020; 15, pp. 1497-1508. [DOI: https://dx.doi.org/10.1021/acsnano.0c08921] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/33372769]

55. Hua, Z.; Yang, B. Monolayer molecular crystals for low-energy consumption optical synaptic transistors. Nano Res.; 2022; 15, pp. 7639-7645. [DOI: https://dx.doi.org/10.1007/s12274-022-4372-9]

56. Dai, S.; Zhao, Y. Recent advances in transistor-based artificial synapses. Adv. Funct. Mater.; 2019; 29, 1903700. [DOI: https://dx.doi.org/10.1002/adfm.201903700]

57. Chiang, Y.C.; Hung, C.C. High-performance nonvolatile organic photonic transistor memory devices using conjugated rod–coil materials as a floating gate. Adv. Mater.; 2020; 32, 2002638. [DOI: https://dx.doi.org/10.1002/adma.202002638]

58. Liu, G.; Li, Q.; Shi, W. Ultralow-power and multisensory artificial synapse based on electrolyte-gated vertical organic transistors. Adv. Funct. Mater.; 2022; 32, 2200959. [DOI: https://dx.doi.org/10.1002/adfm.202200959]

59. Liu, G.; Wen, W.; Zhao, Z. Bionic tactile-gustatory receptor for object identification based on all-polymer electrochemical transistor. Adv. Mater.; 2023; 35, 2300242. [DOI: https://dx.doi.org/10.1002/adma.202300242]

60. Zhang, Z.D.; Zhong, Y.N. Soft memtransistor with ion transfer interface. Flex. Print. Electron.; 2022; 7, 014015. [DOI: https://dx.doi.org/10.1088/2058-8585/ac60b3]

61. Mao, J.Y.; Hu, L. Artificial synapses emulated through a light mediated organic–inorganic hybrid transistor. J. Mater. Chem. C; 2019; 7, pp. 48-59. [DOI: https://dx.doi.org/10.1039/C8TC05030A]

62. Dai, S.; Wu, X. Light-stimulated synaptic devices utilizing interfacial effect of organic field-effect transistors. ACS Appl. Mater. Interfaces; 2018; 10, pp. 21472-21480. [DOI: https://dx.doi.org/10.1021/acsami.8b05036]

63. Pu, X.; Shu, F.; Wang, Q. Visual synapse based on reconfigurable organic photovoltaic cell. J. Semicond.; 2025; 46, 022403. [DOI: https://dx.doi.org/10.1088/1674-4926/24080018]

64. Chen, K.; Hu, H. Organic optoelectronic synapse based on photon-modulated electrochemical doping. Nat. Photonics; 2023; 17, pp. 629-637. [DOI: https://dx.doi.org/10.1038/s41566-023-01232-x]

65. Park, H.L.; Choi, M. Flexible transparent memory systems based on solution-processed organic memristors. J. Korean Phys. Soc.; 2022; 81, pp. 285-289. [DOI: https://dx.doi.org/10.1007/s40042-022-00537-4]

66. Wang, H.; Zhao, Q. A ferroelectric/electrochemical modulated organic synapse for ultraflexible, artificial visual-perception system. Adv. Mater.; 2018; 30, 1803961. [DOI: https://dx.doi.org/10.1002/adma.201803961]

67. Yang, B.; Lu, Y. Bioinspired multifunctional organic transistors based on natural chlorophyll/organic semiconductors. Adv. Mater.; 2020; 32, 2001227. [DOI: https://dx.doi.org/10.1002/adma.202001227] [PubMed: https://www.ncbi.nlm.nih.gov/pubmed/32500583]

68. Wang, C.; Zhang, X. Organic photodiodes and phototransistors toward infrared detection: Materials, devices, and applications. Chem. Soc. Rev.; 2020; 49, pp. 653-670. [DOI: https://dx.doi.org/10.1039/C9CS00431A]

69. Rim, Y.S.; Yang, Y. Ultrahigh and broad spectral photodetectivity of an organic–inorganic hybrid phototransistor for flexible electronics. Adv. Mater.; 2015; 27, pp. 6885-6891. [DOI: https://dx.doi.org/10.1002/adma.201502996]

70. Zhang, J.; Shi, Q. Spectrum-dependent photonic synapses based on 2D imine polymers for power-efficient neuromorphic computing. InfoMat; 2021; 3, pp. 904-916. [DOI: https://dx.doi.org/10.1002/inf2.12198]

71. Lin, Y.C.; Yang, W.C. Recent advances in organic phototransistors: Nonvolatile memory, artificial synapses, and photodetectors. Small Sci.; 2022; 2, 2100109. [DOI: https://dx.doi.org/10.1002/smsc.202100109]

72. Kim, Y.; Zhu, A. A hemispherical image sensor array fabricated with organic photomemory transistors. Adv. Mater.; 2023; 35, 2203541. [DOI: https://dx.doi.org/10.1002/adma.202203541]

73. Wang, S.; Ryan, J.W. Encapsulation of MEH-PPV: PCBM hybrids in the cores of block copolymer micellar assemblies: Photoinduced electron transfer in a nanoscale donor–acceptor system. Langmuir; 2016; 32, pp. 329-337. [DOI: https://dx.doi.org/10.1021/acs.langmuir.5b04053]

74. Zhu, M.; Meng, K. Lateral photovoltaic effect in ITO/PEDOT: PSS/MEH-PPV: PCBM/Al organic photovoltaic cells. Org. Electron.; 2020; 78, 105585. [DOI: https://dx.doi.org/10.1016/j.orgel.2019.105585]

75. Wu, Y.; Dai, S. Optical Microlithography of Perovskite Quantum Dots/Organic Semiconductor Heterojunctions for Neuromorphic Photosensors. Adv. Funct. Mater.; 2024; 34, 2315175. [DOI: https://dx.doi.org/10.1002/adfm.202315175]

76. Yang, C.; Qian, J. An optically modulated organic Schottky-barrier planar-diode-based artificial synapse. Adv. Opt. Mater.; 2020; 8, 2000153. [DOI: https://dx.doi.org/10.1002/adom.202000153]

77. Pei, K.; Ren, X. A high-performance optical memory array based on inhomogeneity of organic semiconductors. Adv. Mater.; 2018; 30, 1706647. [DOI: https://dx.doi.org/10.1002/adma.201706647]

78. Choi, Y.; Ho, D.H. Physically defined long-term and short-term synapses for the development of reconfigurable analogueg-type operators capable of performing health care tasks. Sci. Adv.; 2023; 9, 5946. [DOI: https://dx.doi.org/10.1126/sciadv.adg5946]

79. Wang, S.; Chen, C. A MoS2/PTCDA hybrid heterojunction synapse with efficient photoelectric dual modulation and versatility. Adv. Mater.; 2019; 31, 1806227. [DOI: https://dx.doi.org/10.1002/adma.201806227]

80. Qian, C.; Oh, S. Solar-stimulated optoelectronic synapse based on organic heterojunction with linearly potentiated synaptic weight for neuromorphic computing. Nano Energy; 2019; 66, 104095. [DOI: https://dx.doi.org/10.1016/j.nanoen.2019.104095]

81. Deng, Z.; Zhou, B. Recent progresses of organic photonic synaptic transistors. Flex. Print. Electron.; 2022; 7, 024002. [DOI: https://dx.doi.org/10.1088/2058-8585/ac66ac]

82. Abbott, L.F.; Regehr, W.G. Synaptic computation. Nature; 2004; 431, pp. 796-803. [DOI: https://dx.doi.org/10.1038/nature03010]

83. Kim, M.K.; Lee, J.S. Synergistic improvement of long-term plasticity in photonic synapses using ferroelectric polarization in hafnia-based oxide-semiconductor transistors. Adv. Mater.; 2020; 32, 1907826. [DOI: https://dx.doi.org/10.1002/adma.201907826]

84. Deng, W.; Zhang, X. Organic molecular crystal-based photosynaptic devices for an artificial visual-perception system. NPG Asia Mater.; 2019; 11, 77. [DOI: https://dx.doi.org/10.1038/s41427-019-0182-2]

85. Shi, J.; Jie, J. A fully solution-printed photosynaptic transistor array with ultralow energy consumption for artificial-vision neural networks. Adv. Mater.; 2022; 34, 2200380. [DOI: https://dx.doi.org/10.1002/adma.202200380]

86. Yokota, T.; Nakamura, T. A conformable imager for biometric authentication and vital sign measurement. Nat. Electron.; 2020; 3, pp. 113-121. [DOI: https://dx.doi.org/10.1038/s41928-019-0354-7]

87. Liu, W.; Chen, W.; Jin, C. Organic optoelectrical synaptic transistors for color information processing. Appl. Phys. Lett.; 2023; 123, 193703.

88. Kumar, M.; Kim, H.S. A highly transparent artificial photonic nociceptor. Adv. Mater.; 2019; 31, 1900021. [DOI: https://dx.doi.org/10.1002/adma.201900021]