Detection of Adulteration in Australian Orange Juices by Stable Carbon Isotope Ratio Analysis (SCIRA) Michael Antolovich, Xia Li, and Kevin Robards* School of Science and Technology, Charles Sturt University, Wagga Wagga, NSW 2678, Australia Stable carbon isotope ratio analysis (SCIRA) was used to determine the authenticity of commercial Australian orange juices. Thirty-five samples of Valencia (δ 13 C values from -23.8 to -24.7 ppt) and eight samples of Navel juices (δ 13 C values from -24.1 to -24.5 ppt) of known origin were used to establish a decision level before analysis. No significant seasonal variations in 13 C/ 12 C ratio were observed. Variations in combustion temperature in the method were also found to be insignificant. Keywords: Authenticity; adulteration; SCIRA; stable carbon isotope ratio; orange juice; Valencia; Navel INTRODUCTION Adulteration of orange juice (1) by the addition of high-fructose corn syrup (HFCS) or beet invert syrup plus water is common as sugar represents a major ingredient of juice. High-precision isotope measurement was developed as a method to detect this form of juice adulteration. Stable carbon isotope ratio analysis (SCIRA) measures small variations in the 13 C content of the carbon in different plants. Methods based on measure- ment of the 13 C/ 12 C ratio are used for the detection of C 4 sugars. This technique is based on the constancy of the 13 C/ 12 C ratio in plants using a common photosyn- thetic pathway for carbon dioxide fixation. Natural variations in this ratio have been immensely useful in detecting adulterated foods. The measurement requires high precision because 13 C represents only 1% of the total carbon present, and the maximum relative variation in the 13 C/ 12 C is ∼5% (2). The application of 13 C/ 12 C ratio analysis to the detection of food adultera- tion began in the early 1970s when the relatively inexpensive high-fructose corn syrup (HFCS) was being widely added to honey. Thus, SCIRA was used to determine sugar addition to Israeli orange juices (3) and adulteration of honey with cane sugar or corn syrup (4). Since then, the technique has been successfully used in assessing the authenticity of apple juice (5), maple products (6), orange juice (7, 8), grape juice (9), cran- berry juice (10), and various fruits (11) and juices (12). Other kinds of additives such as organic acids, essence, and flavor constituents can also be detected by SCIRA. Byrne et al. (13) reported the use of 13 C/ 12 C ratio analysis to distinguish the ethyl butyrate which was naturally present in orange juice from that produced by fermentation. There is a natural variation in δ 13 C values in orange juice. In a comprehensive study of orange juice samples from four countries (United States, Mexico, South Africa, and Spain) Doner and Bills (7) reported a mean δ 13 C of -24.5 ppt (range )-23.3 to -25.6 ppt) of all samples from the different growing areas, whereas in a more restricted study (3) of 42 Israeli and French orange juices the mean δ 13 C values were -24.3 and -25.0 ppt, respectively. The mean δ 13 C values (14) for orange juice sugars and orange pulp derived from Israel, United States, and Brazil were -25.1 and -25.6 ppt, respectively. The most negative values have been re- ported (15) for orange juices from Central America, averaging ∼-27 ppt. Thus, analysis of orange juices from different locations suggests that δ 13 C values for oranges are relatively uniform. This facilitates the detection of C 4 adulterants (HFCS and cane invert syrup with δ 13 C values near -10 ppt) in orange juice. In 1982, Doner and Bills (16) conducted an interlaboratory study of δ 13 C values in orange juices and HFCS mixtures. The agreement among different laboratories was excellent. Nonetheless, Doner (17) reported that detection of orange juice adulteration using 13 C/ 12 C ratio measure- ment is qualitative rather than quantitative because the δ 13 C value of a given pure juice sample before it is adulterated cannot be determined. Using a conservative approach, they concluded that any orange juice having a δ 13 C value more negative than -22.1 ppt (4 standard deviations from the mean) was pure and unadulterated. An interlaboratory trial (18) using an internal refer- ence method (19), in which the 13 C/ 12 C ratio of free sugar was compared to the 13 C/ 12 C ratio of the pulp in each sample, found that smaller quantities of adulterants could be detected. Good correlation between laboratories was found, but the method is more complicated, and great care is needed to prepare the pulp samples to prevent interference from lipids. MATERIALS AND METHODS Reagents. Milli-Q water was used for carbon isotope ratio analysis. Standards for carbon isotope ratio analysis included CSR white cane sugar (CSR Ltd. Australia) and ANUC4 sugar (C 4 sugar standard supplied by Alan Chivers, Australian National University, Canberra, Australia). Citrus Juice Samples. Orange juice samples were sup- plied by Leeton Citrus Juice Limited, Leeton, NSW, Australia. Sample collection was performed under the supervision of the authors. These juice samples were made from two cultivars (Valencia and Navel oranges), harvested at different times from sampling locations in New South Wales Riverina and Victorian Sunraysia regions. Navel oranges were sampled 8 * Author to whom correspondence should be addressed (telephone +61 2 6933 2547; fax +61 2 6933 2737; e-mail krobards@csu.edu.au). 2623 J. Agric. Food Chem. 2001, 49, 2623-2626 10.1021/jf001384f CCC: $20.00 © 2001 American Chemical Society Published on Web 05/03/2001