Narrow Results

Cluster analysis of apple juice is an important method in the quality identification of fruit juice. However, the traditional principal component analysis (PCA) is to reduce the dimension of high dimensional data and to achieve clustering by way of score value. Experiments have found that this clustering method is not suitable for real samples with fluorescence overlap. In this paper, we propose a processing mechanism to optimize the derivative of the fluorescence spectrum. When the derivative is calculated, we can amplify the spectral difference of the similar matter and improve the ability of the PCA to achieve the best effect of clustering. Experiments show that the optimized method for the actual sample classification accuracy rate of 100%, which can fully realize the type of commercially available apple juice beverage distinction and achieve the practical application of quality testing.

Three methods for determination of patulin in apple juice have been studied comparatively. The selected procedure for liquid samples implies liquid-liquid extraction with ethyl acetate/hexane (96:4, v/v). The collected patulin was then analyzed by liquid chromatography with UV-Visible detection. The obtained recovery was 82.9±3.2 % and the limit of detection was 2.0 µg/kg of juice, which are according to the valid guidelines established by the European Union. The developed analytical method has been used for the determination of patulin in various Spanish apple juices, where this mycotoxin has not been detected.