OTIMIZAÇÃO E VALIDAÇÃO DE METODOLOGOGIAS ANALÍTICAS PARA DETERMINAÇÃO DE 5-(HIDRÓXIMETIL)-2-FURALDEÁ DO (HMF) EM ALIMENTOS COM ALTOS TEORES DE AÇÚCARES

Abstract

This work presents the validation of HPLC-UV method for the determination of HMF in honey, syrup and cane molasses samples and optimization and validation studies of a new spectrophotometric method for the analysis of this contaminant in syrup. Furthermore, a work was also conducted to evaluate if the chemical composition of these foods had influence on HMF levels found. To confirm the validation of HPLC-UV method were evaluated parameters of selectivity, linearity, precision and accuracy obtained by different statistical techniques. The results indicated that the investigated method presents accuracy (recovery rates between 89-105 %) and precision (RSD for repeatability and intermediate precision between 0.57 % and 6.43 %) appropriate. The limits of detection (LOD) and quantification (LOQ) of 0.09 and 0.26 mg L?¹, respectively, were low, demonstrating the sensitivity of the HPLC-UV method. It was also observed that the method shows good linearity in the concentration range from 1.0 to 5.0 mg L?¹, with values F regression (62063.04) > Fcritical (4.067) and without lack of fit of the linear model ((Flof (1.57) < Fcritical (2.839), p > 0.05)). Furthermore, a comparison of chromatograms and UV-PDA spectra of the HMF standard and of the samples indicates that the method shows good selectivity. Therefore, the validation study demonstrated that the chromatographic methodology can be considered validated. The spectrophotometric was optimized with regard to the concentration of sodium metabisulphite and wavelength. The results showed the concentration of sodium metabisulphite 0.1 % (w/v) and the wavelength of 285 nm as optimal conditions. The validation study was performed calculating the same parameters adopted for the HPLC-UV method, with the exception of selectivity. The spectrophotometric method was linear in the range from 2.0 to 10.0 mg L- 1, with Fregression (12732.73) > Fcritical (4.667), without lack of fit of the linear model ((Flof (0.04) < Fcritical (3.708), p > 0.05)). The values of LOD (0.101 mg L?¹) and LOQ (0.338 mg L?¹) were adequate, nevertheless higher than ones observed for the HPLC-UV method, demonstrating that this analytical methodology presents a lower sensitivity. The results of the precision study (RSD for repeatability and intermediate precision between 0.296 % and 0.362 %) were suitable for the determination of HMF in syrup being within the recommended limits. The accuracy was assessed by comparing the levels of HMF obtained by the spectrophotometric method with those obtained by HPLC-UV method by applying a paired t-test and linear regression analysis at 95 % confidence. It was found that according to the two statistical tests there are not significant differences between the compared methods, indicating that the spectrophotometric method shows good accuracy. Thus, the analytical methods developed in this work for the determination of HMF in honey, syrup and cane molasses were efficient and could be applied in the analysis of quality control, due to ease of implementation, speed and why these not make use of sample preparation steps. To check whether there was a relationship between the chemical composition of the investigated foods (honey, cane molasses and syrup) and the levels of HMF determined by HPLC-UV has been applied Principal Component Analysis (PCA) for reducing sugars, moisture, pH, acidity, conductivity and degree of non-enzymatic browning measured by absorbance at 420 nm. The results of the PCA showed that PC1 could satisfactorily explain the variability in the HMF contents in the foods investigated and the factors evaluated for the three matrices analyzed. The storage time was not important for the formation of HMF in none of foods studied. However, all the matrices presented influence of chemical composition (reducing sugar, pH and acidity) and of degree of non-enzymatic browning, that in this work, was used as a measure of the extent of heat treatment applied, which it suggest that there is not a quality control of foods investigated. The levels of HMF were also influenced by moisture contents in syrup samples and by the electrical conductivity values in cane molasses samples.