Comparison of univariate and multivariate strategies for the determination of sucrose in fruit juices by automated flow injection analysis with Fourier transform infrared detection

An automated technique for the determination of sucrose by flow injection analysis with Fourier transform infrared detection is presented, Two approaches to the quantitation of sucrose in synthetic and real sample matrices have been explored, The first was based on the on-line enzymatic hydrolysis of sucrose to alpha-D-glucose and beta-D-fructose by beta-fructosidase, with difference measurements being performed by using a flow injection manifold with two internally coupled injection valves, The manifold was optimized with time-resolved software and then automated with a custom-built TTL interface to transfer signals from the controlling program to the external apparatus, Univariate calibration was performed on the difference signals between the sucrose samples with and without enzyme hydrolysis, The second approach, using multivariate data analysis, negated the need for sample treatment and produced similar errors for the prediction of sucrose in different matrices within the established linear range of 0 to 100 mM. Both approaches have been successfully applied to the determination of sucrose in fruit juices.