Tailoring Asymmetric Sites to Promote Sensitive Detection of Pneumonia-Related Gases.

Exhaled breath analysis, situated at the intersection of nanomaterial science, analytical chemistry, and clinical diagnostics, is a transformative noninvasive diagnostic technology. However, the ultratrace concentration (ppb-ppm scale) and complex matrix of biomarkers such as nitrogen oxides (NOx) in exhaled breath impose stringent challenges for detection technologies. In this work, we successfully synthesized In2O3-N nanorods with well-defined O-In-N asymmetric active sites via a hydrothermal method and subsequent calcination. The effects of nitrogen doping concentration on the microstructure and gas-sensing properties were systematically investigated, and the optimal doping content was determined. At 80 °C, the In2O3-N sensor exhibits excellent sensing performance toward nitrogen dioxide. Its response reaches 4.5-2 ppm of NO2, approximately 2.2 times that of pristine In2O3. Additionally, it displays excellent selectivity toward NOx (negligible response to non-NOx interfering gases) and good stability (response fluctuation <5% in five cycles). Considering the high humidity in practical detection environments, we further evaluated its humidity resistance. The sensor operates stably over a wide humidity range and possesses strong capability against humidity interference. DFT calculations reveal that the enhanced performances result from two key factors. One is that nitrogen treatment increases the concentration of oxygen vacancies, thereby providing abundant adsorption sites. The other is that in the O-In-N asymmetric sites, electrons transfer from the N atom to the In atom, strengthening the orbital interactions between the In atom in In2O3-N and the N atom in NO2. The In2O3-N-based sensors were further integrated into a portable device for noninvasive pneumonia detection. Clinical tests on 40 exhaled breath samples show that the sensor can effectively distinguish between the two groups of people, with 100% accuracy for pneumonia patients and 80% accuracy for healthy individuals, which is further verified by 3D principal component analysis (PCA) with good separation of sample points. This work not only provides a generalizable strategy for designing high-performance gas sensors via asymmetric active site engineering but also highlights the great potential of In2O3-N sensors in clinical noninvasive diagnosis bridging the gap between nanomaterial design and translational breath analysis.