Dynamic Headspace Gas Chromatography Combined with Multivariate Analysis to Classify Fresh and Processed Orange Juices

Abstract

Sixty-two commercially processed and fresh orange juice samples were analyzed by headspace gas chromatography to quantify 46 volatile constituents in each sample. Computer multivariate analysis was then used to classify these juices into three types based on degree of processing employed. The three types were unpasteurized, pasteurized not from concentrate and reconstituted to single strength from frozen concentrated juice. Principal components (PC) analysis using the first three PC's in a three dimensional graphics plot was the most effective means of classification for these samples. The computer program also provided information regarding the constituents most important to this classification technique. Such classification has potential for use in quality control and in improving quality of processed orange juices.