Phenolic Composition, Quality and Authenticity of Grapes and Wines by Vibrational Spectroscopy

ABSTRACT

Phenolic compounds have a strong influence on the quality and authenticity of grapes, and they have been the main objective of several studies due to its importance on the structure, flavour and color of wines. These compounds have been determined along time by different instrumental techniques combining physical and chemical methodologies. Nowadays, innovative analytical systems, based on spectroscopy, have the ability to analyse complex chemical samples. These systems are increasing on the wine industry since they are fast, accurate and eco-friendly methods. This work reviews current applications of vibrational spectroscopy techniques for the measurement and prediction of different phenolic compounds in grapes and wines. The employed chemometrics tools and the statistical strategies have also been revised and included.

Abbreviations: ANN: artificial neural network; ATR: attenuated total reflection; CANDECOMP: canonical decomposition; CDA: canonical discriminant analysis; CSPT: computer screen photo-assisted technique; CVA: canonical variate analysis; CVM: characteristic vector method; CZE: capillary zone electrophoresis; DA: discriminant analysis; DAD: diode array detector; DPLS: discriminant partial least squares; FDA: factorial discriminant analysis; FLD: fluorescence detection; FT: Fourier-transform; HCA: hierarchical cluster analysis; HPLC: high-performance liquid chromatography; HPTLC: high-performance thin-layer chromatography; IBECSI: iterative backward elimination of changeable size intervals; ICA: independent component analysis; ICC: inter-class correlation coefficients; ICP-OES: inductively coupled plasma optical emission spectroscopy; IF: infrared; iPLS: interval partial least squares; ISO: international organization for standardization; MIR: mid-infrared; MLS: multiple linear regression; MPLS: modified partial least squares regression; MS: mass spectrometry; NIR: near infrared; NMR: nuclear magnetic resonance; OPLS: orthogonal partial least square; PARAFAC: parallel factor analysis; PCA: principal component analysis; PCR: principal component regression; PLS: partial least-squares; PRM: Partial robust M; R: coefficient of correlation; R²: coefficient of determination; RMSEC: root mean standard error of calibration; RMSECV: root mean squared error cross validation; RMSEP: root mean standard error of prediction; RPD: residual predictive deviation; RR: ridge regression; RSQ: coefficient of determination (square of Pearson product-moment correlation coefficient); SEC: standard error of calibration; SECV standard error of cross-validation; SEP: standard error of predictionSIMCA: soft independent modeling of class analogy; SNV: Standard normal variate; UV-vis: ultraviolet-visible; VIP: variable importance in the projection; WILMA-D: wavelet iterative linear modelling approach-discriminant version.

KEYWORDS:

• Food quality
• polyphenols
• spectroscopy
• traceability

Acknowledgments

Thanks are due to the Spanish MINECO (Project ref. AGL2017-84793-C2 co-funded by FEDER, and Project IR-WINE: RTC-2015-4501-2). IGE also thanks Spanish MICINN for postdoctoral contract Juan de la Cierva-incorporación (Ref. IJCI-2017-31499). Proyecto del Programa Operativo FEDER Andalucía 2014-2020 (Ref: US-1261752).