The importance of factorial design on the optimization of biosensor performance: immobilization of glucose oxidase as a case study.

Conventionally, the optimization of glucose biosensors is achieved by varying the concentrations of the individual reagents used to immobilize the enzyme. In this work, the effect and interaction between glucose oxidase enzyme (GOx), ferrocene methanol (Fc), and multi-walled carbon nanotubes (MWCNTs) at different concentrations were investigated by a design of experiments (DoE). For this analysis, a factorial design with three factors and two levels each was used with the software RStudio for statistical analysis. The data were obtained by electrochemical experiments on the immobilization of GOx-Fc/MWCNT at different concentrations. The results showed that the factorial DoE method was confirmed by the non-normality of the residuals and the outliers of the experiment. When examining the effects of the variables, analyzing the half-normal distribution and the effects and contrasts for GOx-Fc/MWCNT, the factors that showed the greatest influence on the electrochemical response were GOx, MWCNT, Fc, and MWCNT:Fc, and there is a high correlation between the factors GOx, MWCNT, Fc, and MWCNT:Fc, as shown by the analysis of homoscedasticity and multicollinearity. With these statistical analyses and experimental designs, it was possible to find the optimal conditions for different factors: 10 mM mL GOx, 2 mg mL Fc, and 15 mg mL MWCNT show a greater amperometric response in the glucose oxidation. This work contributes to advancing enzyme immobilization strategies for glucose biosensor applications. Systematic investigation of DoE leads to optimized immobilization for GOx, enables better performance as a glucose biosensor, and allows the prediction of some outcomes.