In situ analysis of multiple biomarkers in the body provides better diagnosis and enables personalized health management. Since many of these biomarkers are redox-active, electrochemical sensors have shown promising analytical capabilities to measure multiple redox-active molecules. However, the analytical performance of electrochemical sensors rapidly decreases in the presence of multicomponent biofluids due to their limited ability to separate overlapping electrochemical signals generated by multiple molecules. Here we report a novel approach to use charged chitosan-modified electrodes to alter the diffusion of ascorbic acid, clozapine, L-homocysteine, and uric acid-test molecules with various molecular charges and molecular weights. Moreover, we present a complementary approach to use chemometrics to decipher the complex set of overlapping signals generated from a mixture of differentially charged redox molecules. The partial least square regression model predicted three out of four redox-active molecules with root mean square error, Pearson correlation coefficient, and R-squared values of 125 µM, 0.947, and 0.894; 51.8 µM, 0.877, and 0.753; 55.7 µM, 0.903, and 0.809, respectively. By further enhancing our understanding of the diffusion of redox-active molecules in chitosan, the in-situ separation of multiple molecules can be enabled, which will be used to establish guidelines for the effective separation of biomarkers.
Keywords: ascorbic acid; chemometrics; chitosan; clozapine; electroanalysis; electrochemical sensors; homocysteine; multi-sensor arrays; partial least squares; uric acid.