Biomimetic artificial olfactory systems based on metal-organic frameworks: from material design to intelligent pattern recognition

Abstract

Artificial olfactory systems (AOS) capable of mimicking biological olfaction are emerging as powerful tools for complex gas analysis. However, conventional sensing materials often face limitations regarding selectivity, response speed, and high operating temperatures. Metal-organic frameworks (MOFs) offer an ideal solution to these challenges due to their ultra-high porosity and tunable chemical functionality. This review presents a comprehensive overview of strategies to implement high-performance MOF-based AOS. First, we systematically examine three material categories—pristine MOFs, MOF-based composites, and MOF-derivatives—highlighting how their structural design enhances sensitivity and response diversity. Beyond hardware, this review bridges the gap between material synthesis and intelligent pattern recognition. We discuss the integration of machine learning algorithms, ranging from unsupervised clustering to advanced deep learning techniques, which are essential for decoding high-dimensional odor data. Finally, critical challenges for real-world deployment, including humidity stability, large-area fabrication, and sensor drift, are addressed, along with potential solutions. By converging innovative MOF design with advanced pattern recognition, this review provides a roadmap for developing next-generation intelligent machine olfaction for applications in healthcare, environmental monitoring, and smart industries.