Comparative analysis of prediction algorithms for sodium chloride concentration using a liquid crystal microfluidic system

• A novel, precise microfluidic system using chiral liquid crystals for fast NaCl concentration measurements in water is presented. • Robust colorimetric measurements using image processing were performed to achieve high accuracy with low sample consumption. • The presented method significantly improves the speed, accuracy and reliability of measurements compared to traditional techniques. This research study delves into the investigation of utilizing chiral liquid crystals (CLCs) within microfluidic systems as a novel method for precisely determining the concentration levels of sodium chloride (NaCl) present in aqueous solutions. The proposed system employs the variations in color exhibited by CLC material to serve as a gauge for NaCl concentration levels, with these levels being predicted through the application of sophisticated machine learning models. Following a 30-minute chemical reaction period, the solution’s colors achieve equilibrium, exhibiting a direct correlation between the resulting color and the concentration of NaCl present. In addition to the previous analyses, three distinct machine learning models—specifically, linear regression, random forest, and XGBoost—were rigorously evaluated for their capacity to predict the concentration of NaCl, utilizing RGB values as the predictive features. Analysis of the results revealed that, among the models tested, the Random Forest model achieved the highest predictive accuracy, exhibiting an R² value of 0.78, a result that was substantially superior to the performance demonstrated by its counterparts. This system provides a rapid, precise, and economical way to measure NaCl concentration, proving to be an invaluable asset across diverse fields such as food safety, medical diagnostics, and environmental monitoring, where accurate and efficient NaCl measurements are crucial.