Dichiara, AnthonyMoore, Gregory2024-09-092024-09-092024Moore_washington_0250O_26911.pdfhttps://hdl.handle.net/1773/52038Thesis (Master's)--University of Washington, 2024Rapid and quantitative recognition of trace amounts of water in organic solvents is of great importance in industrial operations. Currently the standard methods of determining water content in such cases include the Karl Fischer (KF) titration and fluorescence colorimetry, which are expensive, time consuming, and require skilled operators. In this work, a multifunctional liquid sensing and classification system was developed based on paper comprising pulp fibers adsorbed with multi-walled carbon nanotubes. Papertronics is a field of increasing interest due to the biodegradable nature, flexibility, and light weight of lignocellulosic paper and its ability to be produced at large scale using well-established and cost-effective manufacturing processes. The hygroscopic nature of cellulose fibers causes them to swell radially in response to water molecules, which alters the conductive pathway within the percolated carbon nanotube network present on their surface, leading to a significant change in the paper resistance when wet. Importantly, the swelling behavior strongly depends on the nature of the solvent, which makes it an exceptional material for liquid sensing applications. The electrical response of the sensor was tested in a classification problem to classify solvents, and a regression problem to quantify concentrations of water in solvents. Specific features were identified from the response profiles, generating over 600 data points which were implemented in the LDA and used to differentiate between the liquids. The sensing platform can classify five different solvents with 100% accuracy and quantify the amount of water in the organic solvent with a limit of detection of 250 ppm and comparable resolution which makes it competitive with the KF titration. The proposed paper-based electronic tongue demonstrates the ability to rapidly differentiate low concentrations of water in organic solvents, provides an economic alternative to current methods, and enables future discoveries in biosensing and bioelectronics.application/pdfen-USCC BY-NC-SALinear Discriminant AnalysisMachine learningPaper-based sensorsPapertronicsSensorsMaterials ScienceComputer scienceChemical engineeringForestryThesis