Spiking neural network (SNN), widely known as the third-generation neural network, has been frequently investigated due to its excellent spatiotemporal information processing capability, high biological plausibility, and low energy consumption characteristics. Analogous to the working mechanism of human brain, the SNN system transmits information through the spiking action of neurons. Therefore, artificial neurons are critical building blocks for constructing SNN in hardware. Memristors are drawing growing attention due to low consumption, high speed, and nonlinearity characteristics, which are recently introduced to mimic the functions of biological neurons. Researchers have proposed multifarious memristive materials including organic materials, inorganic materials, or even two-dimensional materials. Taking advantage of the unique electrical behavior of these materials, several neuron models are successfully implemented, such as Hodgkin–Huxley model, leaky integrate-and-fire model and integrate-and-fire model. In this review, the recent reports of artificial neurons based on memristive devices are discussed. In addition, we highlight the models and applications through combining artificial neuronal devices with sensors or other electronic devices. Finally, the future challenges and outlooks of memristor-based artificial neurons are discussed, and the development of hardware implementation of brain-like intelligence system based on SNN is also prospected.

Emerging optoelectronic memristors are promising candidates to develop neuromorphic computing, owing to the combined advantages of photonics and electronics. However, the reversible modulation on device conductance usually requires complicated operations involving hybrid optical/electrical signals. Herein, we design a fully light-modulated memristor based on ZnO/MoOx heterojunction, which exhibits potentiation and depression behaviors under the irradiation of ultraviolet and visible light, respectively. Several basic synaptic functions have been emulated by utilizing optical signals, including short-term/long-term plasticity and spike-number-dependent plasticity. Based on the all-optical modulation characteristics, low-level image pre-processing (including contrast enhancement and noise reduction) is demonstrated. Furthermore, logic operations (“AND,” “NOTq,” and “NIMP”) can be performed by combining various optical signals in the same device. The memristive switching mechanism under optical stimulus can be attributed to barrier change at the heterojunction interface. This work proposes a fully light-modulated memristor based on ZnO/MoOx heterojunction that may promote the development of neuromorphic computing with high efficiency.

A neuromorphic computing chip that can imitate the human brain’s ability to process multiple types of data simultaneously could fundamentally innovate and improve the von-neumann computer architecture, which has been criticized. Memristive devices are among the best hardware units for building neuromorphic intelligence systems due to the fact that they operate at an inherent low voltage, use multi-bit storage, and are cost-effective to manufacture. However, as a passive device, the memristor cell needs external energy to operate, resulting in high power consumption and complicated circuit structure. Recently, an emerging self-powered memristive system, which mainly consists of a memristor and an electric nanogenerator, had the potential to perfectly solve the above problems. It has attracted great interest due to the advantages of its power-free operations. In this review, we give a systematic description of self-powered memristive systems from storage to neuromorphic computing. The review also proves a perspective on the application of artificial intelligence with the self-powered memristive system.

Temperature is known as an important factor in biological synaptic transmission. In this study, temperature-modulated switching behaviors are reported in an amorphous carbon (a-C) diffusive memristor device to emulate biorealistic synaptic plasticity. The devices exhibit memory switching and threshold switching behaviors depending on the compliance current and ambient temperature. As confirmed by conducting atomic force microscopy, the thermal effect can promote the electrochemical formation of a stable metallic conductive filament. A series of timing-controlled pulse experiments are carried out to study the temperature effect on the switching characteristics, and the device shows second-order memristive behaviors. Frequency-dependent synaptic plasticity and timing-controlled spike-time-dependent plasticity are demonstrated in the device, which are analogous to the synaptic strength in a biological synapse at elevated temperatures. As a proof of concept, the forgetting behavior of numerical images learned at different temperatures and different pulse durations is conceptually emulated with synaptic device arrays. It is expected the present device with second order memristive behaviors provides alternatives for biorealistic synaptic applications.

True random number generator (TRNG) that cannot be arbitrary attacked with predictable software algorithm is a promising data security solution. Memristors, possessing specific intrinsic stochasticity, are just appropriate to be the random sources for encryption applications. In this work, a TRNG system based on the stochastic duration time of double threshold-switching (TS) memristors is proposed. The reliable stochasticity of this system is mainly attributed to the gradual dissolution of Ag conductive channels and the synergistic effect of these two TS memristors. A digital circuit system based on the micro-controller unit is designed to produce stable random “0” and “1” bitstreams. The random data key generated by this double TS memristor TRNG passed the National Institute of Standards and Technology randomness tests without post-processing procedure. Taking advantage of the true random data key, the binary image can be encrypted and decrypted by executing the XNOR logic operation. In addition, the Gaussian noise effect on the decryption stability of image information secure communication is also analyzed. Although 90% Gaussian noise was applied to the original image, the pristine binary image still can be roughly recognized. This work demonstrates that the double TS memristors-based TRNG system has great potential for hardware security applications.

Neuromorphic multimodal perception of sensory systems can integrate the stimulation from different senses, thus enhancing the perception accuracy of organisms to understand the external environment. An optoelectronic memristor with the capability to combine multidimensional sensing and processing functions is highly desirable for developing efficient neuromorphic multimodal sensory systems (MSSs). In this work, a tellurene (Te) nanoflake‐based optoelectronic memristor relying on solution plasma process (SPP) treatment is demonstrated for the first time, which is capable of combining infrared (IR) optical and electrical stimuli in a single synaptic device for a multisensory integration function. The multimode switching mechanisms can be attributed to the migration of SPP‐induced defects under an electric field and the charge trapping/detrapping in those defects under light illumination. Furthermore, a multimode reservoir computing (MRC) system is realized to perform pattern recognition tasks using the IR optical (1022 nm) and electrical stimulations, including the fusion of visual and haptic multimode sensory and physiological recognition tasks. The accuracy of pattern recognition using MRC is obviously enhanced compared to that of individual sensory. This work provides a new approach to developing multimodal 2D Te‐based MRC for the construction of highly efficient neuromorphic MSSs.