In the realm of physically informative neuromorphic perception, significant strides have been made in the research of artificial oscillatory neurons and receptors. Nevertheless, the exploration of biologically realistic neurons tailored for chemical perception remains relatively uncharted territory. These specialized neurons harness bionic chemical receptors or channels to produce spiking responses when exposed to chemical stimuli, such as ions or biomolecules, within biological environments. Although current chemical perception oscillation architectures, which are based on nanofluidic channels and organic electrochemical transistors, exhibit certain biomimetic qualities, they encounter hurdles in terms of material stability, miniaturization, integration, and overall system power consumption.

Traditional architectures predominantly rely on ion-sensitive field-effect transistors and transistor-assisted circuits. However, these setups are plagued by issues including a restricted neuronal dynamic range, intricate structural designs, and subpar biocompatibility. To overcome these obstacles, a research team helmed by Professor Wang Lili from the Institute of Semiconductors at the Chinese Academy of Sciences has put forth an innovative monolithic 'cell-on-memristor' chemical sensing architecture.

This groundbreaking architecture achieves in-sensor processing by three-dimensionally integrating volatile memristor units with battery units, eliminating the need for an external energy supply. As a result, it successfully pioneers the development of a microscale chemical receptor that boasts biologically realistic properties. The device showcases a range of impressive performance characteristics, encompassing an ultra-wide ion sensing range, notable voltage oscillation amplitude, consistent oscillation frequency, sustained spiking behavior, and remarkably low internal energy consumption.

Leveraging a CMOS-compatible manufacturing process, this system facilitates wafer-level array fabrication. It also demonstrates a high-density chemical receptor array system that achieves a 99.4% accuracy rate in classifying saltiness levels. This approach offers a versatile and scalable solution for advancing research in the field of chemical perception oscillation systems.