Universal 3D‐Printing of Suspended Metal Oxide Nanowire Arrays on MEMS for AI‐Optimized Combinatorial Gas Fingerprinting

作者：Yu Liu, Kichul Lee, Hongjun Liu, Chenwei Li, Xiaoyi Zeng, Rongyue Liu, Yulong Chen, Zijun Chen, Jihyuk Yang, Huan Xiao, Inkyu Park, Ji Tae Kim, Xing Cheng · 发表于：Advanced Science · 年份：2025 · DOI：10.1002/advs.202511794 · 被引用次数：5 · 研究领域：Gas Sensing Nanomaterials and Sensors、Advanced Sensor and Energy Harvesting Materials、ZnO doping and properties

Abstract Additive manufacturing technology has the potential to provide great versatility in the design and fabrication of sensing devices. This prerequisite necessitates further technological improvement in precision and material diversity. Here, a universal meniscus‐guided 3D printing method is reported that can fabricate freestanding metal oxide semiconducting nanowires with programmed compositions and compatible substrate options at the single‐entity level. By studying printing process and ink compositions, polycrystalline metal oxide (MOX) nanowires with controlled shapes and tunable diameters down to 180 nm are achieved. The method enables a high‐precision, mask‐free printing of MOXs nanowire arches array on a 1.5 µm thick suspended membrane, paving the way for integrating a micro‐electromechanical systems (MEMS) chemiresistive gas sensor. The diversity of 3D printable materials demonstrated in this study covers 24 types of combination MOX nanowires, including TiO 2 , ZnO, SnO 2 , In 2 O 3 , WO 3 , and CeO 2 , doped with noble metals of Au, Ag, Pd, and Pt. Their sensing performances for CH 4 , NH 3 , CH 3 CH 2 OH, CO, and H 2 S gases are quantitatively investigated, while artificial intelligence (AI)‐driven analysis of multi‐sensor responses achieves 98% gas classification accuracy via sliding time window‐based convolutional neural networks (CNN). The ability to 3D print semiconducting materials opens the possibility to freely design and realize new‐concept electronic d...