Detection of adulteration in orange juices by a new screening method using proton NMR spectroscopy in combination with pattern recognition techniques

This paper describes the application of proton NMR spectroscopy as a screening tool for the determination of the authenticity of orange juices. Principal component and discriminant analyses were used to discriminate between authentic and nonauthentic (suspect) samples. In one set of profiles, additions of sucrose, beet medium invert sugar, sodium benzoate could easily be detected. In another set of data, K-nearest neighbor classification based on the principal component scores was used to correctly classify at least 94% of samples known to deviate from authentic samples when measured with analytical techniques such as high pressure liquid chromatography and atomic absorption spectroscopy. Principal component loading plots and factor spectra were an effective tool in the interpretation of the differences between the profiles.