Table 3.

Applications of spectroscopic techniques in authenticity studies of fruit juices.

Aim of Study	Analytical Technique		Type of Study		Chemometric Approach		Sensitivity & Accuracy	Authenticity Markers	Ref.
Authentication of pomegranate juice – Detection of adulteration with apple, grape - exogenous addition of water	UV-Vis		Untargeted		PCA		Lowest level of detected adulteration: 10%	UV-Vis Spectra	[52]
Analysis of variety, adulteration, quality and ageing of apple juices	UV spectroscopy		Untargeted & Targeted		PCA, PCR, PLSR		Prediction of adulteration: PLSR: Root-mean-square error < 0.7783%, R2 > 0.9980	UV spectra Physicochemical parameters	[53]
Authentication of orange juice – Detection of adulteration with grapefruit juice	3D-Front-Face Fluorescence Spectroscopy		Untargeted		ICA		Lowest level of detected adulteration: 1%	3D fluorescence spectra Antiradical Activity, Total Flavonoid Content	[54]
Authentication of apple juice categories (fruit juice vs reconstituted from concentrate)	Synchronous Scanning Fluorescence		Untargeted		PLS-DA, PCA		PLS-DA: Cross-validation error rate: 0.05–0.14 External validation error rate: 0.05–0.44	Total and Specific Synchronous Fluorescence Spectra	[55]
Authentication of bayberry juice – Detection of the adulteration with the addition of water	NIR spectroscopy		Untargeted		PCA, RBFNN		Lowest level of detected adulteration: 10% Recognition Rate: 97.6%	NIR spectra	[56]
Authentication of grape juice – Varietal Differentiation	VIS-NIR & MIR spectroscopy		Untargeted		PCA, LDA, PLS-DA		Correct Classification Rate LDA: 86% (MIR)–80% (NIR) PLS-DA: 86%	VIS-NIR & MIR spectra	[57]
Authentication and differentiation of apple, blueberry, cranberry, concord grape and plum juice	FTIR		Untargeted		PCA, HCA, SIMCA		100% of correct classification	Phenolic-rich fraction spectra	[58]
Authentication of Concord grape juice in different grape juice blends	FTIR		Untargeted		SIMCA, PLSR			FTIR Spectra (phenolic fraction)	[59]
Authentication of pomegranate juice concentrate – Detection of adulteration with grape juice concentrate	FTIR		Untargeted		PCA, PLSR		Lowest level of detected adulteration: 2%	FTIR Spectra 1700–1800 cm-1 region, C=O stretching mode.	[60]
Authentication of mango juice - Detection of the addition of sucrose solution	FTIR		Untargeted		PCA, PLSR		Lowest level of detected adulteration: 3–5%	FTIR spectra	[61]
Authentication of orange juice - Detection of the addition of sugar adulterants	FTIR spectroscopy		Targeted		PC-DFA, PLSR		Lowest level of detected adulteration: 0.5–20%	Sucrose, glucose & fructose	[62]
Authentication of orange juice – Detection of the adulteration of freshly squeezed with concentrated orange juices	FTIR-ATR		Untargeted		PCA, LDA		Lowest level of detected adulteration: 10%	FTIR spectra	[63]
Quality Control of fruit juices - Estimation of fruit content of juices	1H-NMR		Targeted & Untargeted		PCA, SIMCA		Accuracy: 10%	NMR Spectra	[64]
Authentication of orange juice – Detection of adulteration with clementine juice	1H-NMR		Untargeted		PLSR		Lowest Error Rate: 3.47% ofmisclassification	NMR spectra	[65]
Authentication of sour cherry juices – Clones/cultivars discrimination	1H-NMR Spectroscopy, Quantitative DescriptiveSensory Analysis,		Untargeted		SIMCA, PCA, PLS-DA		Prediction ability: 82–91.2%	NMR spectra Malic acid, malic acid/glucose ratio	[66]
Authentication of mango juices – Cultivars discrimination		1D and 2D-NMR Spectroscopy		Untargeted		PCA	-	band-selective 1D and 2D NMR arginine, histidine, phenylalanine, glutamine, shikimic acid, trigonelline	[67]
Quality and authentication of orange juice – influence of storage conditions and variety discrimination		1H-NMR spectroscopy		Targeted & Untargeted		PCA, PCR, PLSR	-	a-glucose, b-glucose, fructose, ethanol, acids (citric, formic, fumaric, succinic, acetic, lactic, malic, pyruvic)	[68]
Grape juice Quality Control (storage conditions, commercial juice authenticity)		1H-NMR spectroscopy		Untargeted		PCA	-	Ethanol, acetate, aminoacids, citric acid, sucrose, sodium benzoate	[69]
Authentication of sweet cherry juice – Geographical Origin Characterization		1H-NMR		Untargeted		PCA, LDA, PLS-DA	Prediction ability: 94.9% for PLS-DA 92.3–94.9% for LDA	NMR spectra malate, glucose, fructose, glutamine, succinate	[70]

DA: Discriminant Analysis, FTIR: Fourier Transform Infrared Spectrometer, HCA: Hierarchical Cluster Analysis, ICA: Independent Components Analysis, k-NN: k-Nearest Neighbors, LDA: Linear Discriminant Analysis, NIR: Near-infrared spectroscopy, NMR: Nuclear Magnetic Resonance, PC-DFA: Principal Components-Discriminant Function Analysis, PCR: Principal Component Regression, PLS-DA: Partial Least Squares Discriminant Analysis, PLSR: Partial Least Squares Regression, RBFNN: Radial Basis Function Neural Networks, SIMCA: Soft Independent Modelling by Class Analogy.