摘要： Overcoming the low selectivity issue of semiconductor oxide (SMO)-based gas sensors at room temperature and realizing the accurate detection of trace disease biomarkers are highly desirable for widespread deployments of sensors in exhaled breath. Here, a self-assembly strategy is proposed to create a graphene quantum dot (GQD) functionalized porous and hierarchical SnO2 quantum nanoparticles (SnO2QNP)/ZnO nanostructure. SnO2QNP/ZnO nanosheets self-assembled directly on the digital integrated electrodes with a post-synthetic humidity treatment (psHT), the construction of GQD and SnO2QNP loaded ZnO nanosheet heterostructure is highly controllable and reproducible. The strong synergistic effect and p-n heterojunction between the p-type GQD and n-type SnO2 and ZnO effectively enlarged the resistance variation due to the change in oxygen adsorption. In comparison with pristine ZnO and SnO2/ZnO sensors, the GQD modified hierarchical SnO2QNP/ZnO nanostructure exhibited a remarkably high response (S=15.9 for 0.1 ppm H2S) and rapid response/recovery time (14/13s), and good selectivity towards H2S against other interfering gases. In particular, we applied principal component analysis (PCA) to analyze the sensing performance of GQD-SnO2QNP/ZnO sensor and found that the combined effects of GQD/SnO2QNP/ZnO heterointerfaces contributed to the improvement of selectivity of sensors. The results demonstrate that the GQD modified SMO with the hierarchical structure has a high potential in the non-invasive exhaled diagnosis.