Plasmonic plastic optical fiber chips combined with artificial intelligence to identify water or alcoholic solutions

In this work, the classification and the identification of two different types of solvent solutions (i.e., isopropanol:water and ethanol:water) were achieved by exploiting the solvent solution properties combined with Artificial Intelligence. More in detail, a Surface Plasmon Resonance (SPR) sensor based on D-shaped plastic optical fibers (POFs) acted as an optical transducer to build a proper dataset. The plasmonic probe was used to monitor the bulk refractive index variations of the solvent solutions due to the evaporation phenomenon over time. The collected experimental data were used to train a machine learning-based algorithm useful for building a prediction model. In such a way, it was made possible to determine the presence of the solvent in the solution under test (water or alcoholic solutions) and, in addition, to recognize the type of solvent. Finally, the results obtained from the testing of unknown solutions testified to the goodness and suitability of the proposed simple sensing approach.