Voltammetric Determination of Diclofenac in the Presence of Paracetamol and Naproxen by an Artificial Neural Network Model Using a Carbon Paste Electrode

• Published in: ECS transactions 2018-01, Vol.84 (1), p.195-205
• Main Authors: Aguilar-Lira, Guadalupe Yoselin, Gutiérrez-Salgado, Juan Manuel, Rojas-Hernández, Alberto, Rodríguez-Ávila, Jose Antonio, Páez-Hernández, María Elena, Álvarez-Romero, Giaan Arturo
• Format: Article
• Language: English
• ISSN: 1938-5862; 1938-6737; 1938-6737; 1938-5862
• DOI: 10.1149/08401.0195ecst

This work presents the quantification of diclofenac in presence of naproxen and paracetamol using differential pulse voltammetry and an artificial neural network model. Cyclic voltammograms of paracetamol, diclofenac and naproxen showed a strong interference in the anodic peaks usually used for its quantification. By means of an artificial neural network model, the interpretation of the voltammograms at different concentrations of the drugs allowed the calibration of the system; each voltammogram was pre-treated with the Discrete Wavelet Transform so the data were compressed. The architecture of the mathematical model is based on a multi-layer perceptron network and a Bayesian training algorithm. With the trained model, R2 values about 0.96 were obtained for the test data when quantifying the three drugs, even allowing quantification outside the linear interval of the calibration curve. Five pharmaceutical samples were tested, obtaining an R2 of 0.98 and a recovery percentage of 98.1% for diclofenac.