Spectrophotometric Simultaneous Determination of Citral Isomers in Cyclodextrin Complexes with Partial Least Squares Supported Approach

This article describes the development and optimization of a UV spectrophotometric method to evaluate the isomers of citral in cyclodextrins complexes by a partial least squares (PLS) regression model. The central composite design (CCD) associated with the response surface methodology (RSM) was applied to select the wavelength range that provides the best prediction results. β-cyclodextrin (β-CD) and hydroxypropyl-β-cyclodextrin (HP-β-CD) complexes with citral were prepared in ethanolic solution and spray dried. These samples were analyzed by an optimized PLS-UV model and the results compared with HPLC-UV determinations to evaluate the predictive power of the PLS-UV model. The results from the PLS-UV model showed a relative standard deviation of prediction (RSEP) ranging from 0.87% for neral up to 1.42% for geranial. The relative standard deviation (RSD) obtained in analyses of four replicates of citral-cyclodextrins complexes ranged from 2.2% to 4.1%. Geranial showed higher affinity than neral for both types of cyclodextrins evaluated. Finally, β-CD was shown to be a better complexing agent for citral than HP-β-CD.