Graphene-enabled electrochemical fingerprint profiling for differentiating authentic and adulterated honeysuckle (Lonicera japonica)

This study presents a novel sensing platform that integrates modified electrodes with advanced computational analysis for rapid quality assessment of honeysuckle (Lonicera japonica) and other herbal products. A cost-effective electrode was prepared by drop-casting 10 µL of a 5 % nanomaterial ink onto a 0.126 cm² screen-printed electrode, followed by a 2-h drying process at ambient temperature. Differential pulse voltammetry measurements conducted in 0.1 M PBS (pH 7.0) and ABS (pH 4.5) yielded distinct oxidation peaks; high-quality samples exhibited peaks at approximately 0.48 V in PBS and 0.42 V in ABS, whereas substandard specimens produced broader signals at around 0.52 V and 0.46 V, respectively. An optimized extraction protocol using 0.1 g of powdered sample in 2 mL ethanol with a 10-min sonication step ensured efficient recovery of electroactive constituents. Stability testing over a 12-h period demonstrated current fluctuations confined within ± 2 %, and reproducibility assessments across five independent electrodes revealed a variation of less than 3 %. Multivariate statistical techniques, including PCA and LDA, enabled clear separation of sample clusters, while classification algorithms—MLP, NLSVM, and RF—achieved laboratory accuracies up to 92 % and 90.3 % in large-scale market tests. These performance metrics confirm that the integrated platform can reliably discriminate complex voltammetric profiles, offering a rapid and scalable solution for on-site quality control in resource-limited environments. The proposed method provides a promising alternative to conventional approaches and holds significant potential for revolutionizing quality assurance practices in the herbal medicine industry.