Bacterial detection and classification are critical challenges in healthcare, environmental monitoring, and food safety, demanding selective and efficient methods. This study presents a novel, label-free approach for E. coli detection using ultrathin Langmuir-Blodgett films of octadecylamine functionalized (ODA)-functionalized graphene on gold electrodes, with a detection range spanning \(10^{1}-10^{6}\) colony-forming units/mL (CFU/mL). Electrochemical impedance spectroscopy (EIS) was performed on six bacterial strains, representing Gram-negative and Gram-positive classes, to evaluate selectivity. The method achieved a remarkably low detection limit of 2.5 CFU/mL for E. coli, with its EIS spectra exhibiting distinct features compared to other bacterial strains. The pronounced differences enabled perfect classification using the Bagging Classifier, achieving no false positives. Machine learning (ML) algorithms applied to raw impedance data improved detection precision and reliability, enabling automated and accurate analysis. These findings establish a robust framework for rapid and selective E. coli detection, crucial for ensuring food and water safety. The integration of ML significantly improves detection accuracy, reduces analysis time, and minimizes human error, paving the way for scalable, cost-effective diagnostic tools for diverse biological and environmental applications.
