Search

The Si‐based self‐powered broadband photodiode (SSBP) is prized for its ability to swiftly detect light across a wide spectrum without requiring an external voltage. However, boosting its efficiency remains challenging due to its high refractive index and limited UV light penetration. A combination of Si with ZnGa2O4, an ultra‐wide‐bandgap spinel material, can bring new opportunities to address these shortcomings of SSBP. In this study, a ZnGa2O4/p‐Si heterojunction photodiode is presented, which is capable of detecting UV to near‐infrared light autonomously. Operating without bias, this device exhibits excellent rectification and detects wavelengths from 265 to 1000 nm, achieving impressive performance metrics such as a photo‐to‐dark current ratio of 5.8 × 104, response speed of less than 3 ms, responsivity of 117 mA W−1, and specific detectivity of 5.5 × 1012 Jones while the photodiode demonstrates exceptional stability and durability under harsh conditions. The versatility of this device is demonstrated by applying it to the optical imaging sensors and physically unclonable security devices. This study provides new inspirations for the development of the energy‐efficient and emerging optical sensing technologies. [ABSTRACT FROM AUTHOR]

Memristive logic-in-memory circuits can provide energy- and cost-efficient computing, which is essential for artificial intelligence-based applications in the coming Internet-of-things era. Although memristive logic-in-memory circuits have been previously reported, the logic architecture requiring additional components and the non-uniform switching of memristor have restricted demonstrations to simple gates. Using a nanoscale graphene oxide (GO) nanosheets-based memristor, we demonstrate the feasibility of a non-volatile logic-in-memory circuit that enables normally-off in-memory computing. The memristor based on GO film with an abundance of unusual functional groups exhibited unipolar resistive switching behavior with reliable endurance and retention characteristics, making it suitable for logic-in-memory circuit application. In a state of low resistance, temperature-dependent resistance and I-V characteristics indicated the presence of a metallic Ni filament. Using memristor-aided logic (MAGIC) architecture, we performed NOT and NOR gates experimentally. Additionally, other logic gates such as AND, NAND, and OR were successfully implemented by combining NOT and NOR universal logic gates in a crossbar array. These findings will pave the way for the development of next-generation computer systems beyond the von Neumann architecture, as well as carbon-based nanoelectronics in the future. [ABSTRACT FROM AUTHOR]

In this study, a hybrid organic‐inorganic field‐effect transistor (FET) is proposed with n‐type zinc‐tin oxide (ZTO) and p‐type dinaphtho[2,3‐b:2′,3′‐f] thieno[3,2‐b]thiophene (DNTT), presenting two applications: (1) random number generator (RNG) and (2) spiking neuron. Interface challenges lead to operational instabilities such as bias stress and hysteresis due to trap site formation from oxide surface hydroxyl groups. However, these trap sites are utilized to produce unstable noise for RNG. Also, the impact of an internal interlayer is explored to mitigate instability in the negative transconductance (NTC) effect. This interlayer enhances material compatibility, improving turn‐on voltage, on‐off current ratio, and reducing hysteresis in the FET. These improvements highlight and maximize the robustness of NTC characteristics. Utilizing this behavior, a spiking neuron is demonstrated that emulates neuronal spiking and generates neuronal spike signals. [ABSTRACT FROM AUTHOR]