High-sensitivity sensor array base on molecular design and machine learning for amine differentiation in exhaled vapor

•A sensor array which can classify pathological biomarkers of amines is provided.•By molecular backbone rigidity together with multiple hydrogen-bonding acceptor and aromatic terminal structure tuning, the difference interactions between molecules and multiple amines are realized.•Through molecular design and machine learning, the detection and recognition of ppb concentration of amine/ammonia at room temperature with an accuracy of 90% by the sensor array are achieved.