Optimization of Milk Sample Cleanup Using Response Surface Methodology

Abstract

Protein precipitation is a fast and simple methodology for sample preparation in milk, performed by addition of an organic solvent compatible with the analytical system. Therefore, the aim of this study is to compare four organic solvents efficiency: (i) acetonitrile, (ii) acetone, (iii) ethanol, and (iv) isopropanol for deproteinization. Acetonitrile was the most efficient at freezer temperature, while at room temperature the most efficient solvent was acetone. In order to optimize precipitation conditions, two central composite designs (CCDs) associated with response surface methodology (RSM) were performed using acetonitrile as precipitant. ANOVA was applied to the experimental designs and data were adjusted to quadratic model with high order of significance, suitable adjustment of the experimental data, with Adjusted-R2 > 0.98, and adequacy precision desirable. For models’ optimization and validation, the desirability function was performed, and under the optimized conditions, which were obtained with a total desirability value of 1.000, the experimental values are statistically equal to predicted values for protein. The volume of precipitant and the ultrasound time were significant on the response, whereas the vortex time was a non-signifcant factor. The RSM was applied efficiently in the optimization process allowing the simultaneous evaluation of the variables on the response.