SERS nose arrays based on a signal differentiation approach for TNT gas detection

TNT, a well-known explosive, is highly toxic and difficult to decompose, making the detection of trace amounts of residual TNT in the environment a topic of significant research importance. Label-free surface-enhanced Raman spectroscopy (SERS) has been demonstrated to be capable of capturing rich compositional information from the sample being tested. Here we show a SERS nose array that contains six individual SERS substrates composed of different components based on a signal differentiation approach (SD-SERS arrays). In this strategy, the SD-SERS arrays integrate differentiated signal structures, physically enhanced structures, and structures with varied adsorption capabilities. Through the differentiated information obtained from SD-SERS arrays, further integration with machine learning algorithms demonstrates the high accuracy of SD-SERS arrays in classifying TNT and structurally similar 2,4-DNPA, as well as in distinguishing between gases at different concentrations. The SERS nose based on SD-SERS arrays presents a convenient and broadly applicable technology with great potential for substance classification and concentration categorization.