The first attempt on fabrication of a nano-biosensing platform and exploiting first-order advantage from impedimetric data: application to simultaneous biosensing of doxorubicin, daunorubicin and idarubicin

Abstract

In this work, for the first time, we have developed a novel and very interesting electroanalytical methodology assisted by first-order multivariate calibration (MVC) for simultaneous determination of doxorubicin (DX), daunorubicin (DN) and idarubicin (ID) as three chemotherapeutic drugs at simulated physiological conditions. A sever overlapping was observed among signals of the three drugs which hindered us for simultaneous determination of them by conventional electroanalytical techniques. Therefore, we had to assist our method by chemometric approaches to develop a novel method for simultaneous determination of DX, DN and ID. Among the existing electroanalytical methods, electrochemical impedance spectroscopy (EIS) due to its high sensitivity was chosen. After individual calibration of the three drugs with the EIS data, a set of calibration samples was designed which was used to develop several first-order MVC models by partial least squares (PLS), continuum power regression (CPR), radial basis function-partial least squares (RBF-PLS), RBF-artificial neural network (RBF-ANN) and least squares-support vector machines (LS-SVM) as linear and non-linear chemometric algorithms. Then, performance of the developed MVC models in predicting concentrations of DX, DN and ID in synthetic samples was compared to choose the best model for the analysis of real samples. Our records confirmed more superiority of RBF-PLS algorithm than the other developed models which motivated us to choose it for the analysis of real samples. Fortunately, the results of the RBF-PLS in the analysis of real samples towards simultaneous determination DX, DN and ID was acceptable.