The potential of NMR spectroscopy and multivariate analysis methods to detect the adulteration of orange juice with pulp wash is demonstrated. Principal component analysis has been applied to (1)H NMR spectra of >300 orange and pulp wash juices, and stepwise linear discriminant analysis was used to classify the samples. A model with six principal components gave a high success rate of classification (94%) for both training and validation sets. An important principal component loading showed that dimethylproline played a key role in the discrimination between the two types of juice, with higher levels in pulp wash. Dimethylproline was not previously known as a marker compound for orange juice adulteration. An ANOVA test revealed at least 21 other NMR signals that differed significantly between the authentic and pulp wash groups. The compounds they represent could be seen as potential marker compounds in addition to dimethylproline. This makes NMR with chemometrics an attractive screening tool with advantages in terms of rapidity, simplicity, and diversity of information provided.

Discrimination between orange juice and pulp wash by 1H nuclear magnetic resonance spectroscopy : Identification of marker compounds

Books, Reviews, and Symposia Proceedings 515R Theory and Optimization 516R Theory 516R Optimization 517R Data Analysis 520R Stationary Phase Classification 520R Retention/Solute-Stationary Phase Interactions 521R Modeling/Peak Characterization/Smoothing 521R Calibration and Curve Fitting 521R Deconvolution 521R Factor Analysis 521R Miscellaneous Fundamental Studies 521R Pharmacological and Biological Applications 522R Normal Phase 522R Reversed-Phase 523R Mobile Phase Studies 524R Stationary-Phase Studies 525R Biopolymer Separations 527R Column Packings 527R Reversed Phase 529R Hydrophobic and Hydrophilic Interaction Chromatography 531R

Liquid chromatography: theory and methodology.

This paper describes an approach for studying collections of near-infrared spectra by using multivariate analyses. The method is illustrated with the use of two sets of spectra of gelatinized starch, recorded in the transmission mode between 650 and 1235 nm. The first set consisted of 99 spectra of partly gelatinized samples (from 24.5 to 100% gelatinization). Application of principal component analysis (PCA) made it possible to identify an outlying sample and to identify the importance of spectral variations due to the effect of scattering. Hence, it was possible to eliminate the scatter variations. From principal component regression (PCR), it was shown that the relationship between corrected spectra and gelatinization was not linear. Discriminant analysis was applied to seven classes of starch gelatinization. Only five samples out of 98 were incorrectly identified. The second set of samples was designed for studying the effect of temperature variation on the spectra of fully gelatinized starch samples. It was possible to show from PCR that the relationship between the spectra and temperature was linear. The "spectral patterns" assessed from discriminant analysis of starch gelatinization and from the PCR of temperature were compared.

Application of Multivariate Analyses to NIR Spectra of Gelatinized Starch

Discrimination on latent components with respect to patterns. Application to multicollinear data

Multivariate strategies for classification based on NIR-spectra—with application to mayonnaise

Continuous digitalized signals such as spectra, electrophoregrams or chromatograms generally have a large number of data points and contain redundant information. It is therefore troublesome performing discriminant analysis without any preliminary selection of variables. A procedure for the application of canonical discriminant analysis (CDA) on this kind of data is studied. CDA can be presented as a succession of two principal component analyses (PCAs). The first is performed directly on the raw data and gives PC scores. The second is applied on the gravity centres of each qualitative group assessed on the normalized PC scores. A stepwise procedure for selection of the relevant PC scores is presented. The method has been tested on an illustrative collection of 165 size-exclusion high-performance (SE-HPLC) chromatograms of proteins of wheat belonging to 55 genotypes and grown in three locations. The discrimination of the growing locations was performed using seven to nine PC scores and gave more than 86% accurate classifications of the samples both in the training sets and the verification sets. The genotypes were also rather well identified, with more than 85% of the samples correctly classified. The studied method gives a way of assessing relevant mathematical distances between digitalized signals according to qualitative knowledge of the samples.

Regis Meritan

Papers

Stepwise canonical discriminant analysis of continuous digitalized signals: Application to chromatograms of wheat proteins