Título:	Visible and Near Infrared Spectroscopy (Vis/NIRS) application for rapid total phenolic and antioxidant activity analyses in wines.
Autoria:	MARQUES, E. J. N. BIASOTO, A. C. T. FREITAS, S. T. de PEREIRA, G. E. OLIVEIRA, W. P. de MEDEIROS, E. P. de
Afiliação:	EMANUEL JOSÉ NASCIMENTO MARQUES, UNIVERSIDADE FEDERAL DE PERNAMBUCO; ALINE TELLES BIASOTO MARQUES, CPATSA; SERGIO TONETTO DE FREITAS, CPATSA; GIULIANO ELIAS PEREIRA, CNPUV/CPATSA; WALKIA POLLIANA DE OLIVEIRA, UNIVERSIDADE FEDERAL DA BAHIA; EVERALDO PAULO DE MEDEIROS, CNPA.
Ano de publicação:	2014
Referência:	In: CONGRESSO BRASILEIRO DE CIÊNCIA E TECNOLOGIA DE ALIMENTOS, 24., 2014, Aracajú; CONGRESSO DO INSTITUTO NACIONAL DE CIÊNCIA E TECNOLOGIA DE FRUTOS TROPICAIS, 4., 2014, Aracajú. Inovação e sustentabilidade em ciência e tecnologia de alimentos: anais. Aracajú: sbCTA, 2014.
Conteúdo:	Near infrared spectroscopy (NIRS) has been widely used for analysis of wine quality due to its simplicity, speed, versatility, reliability and low cost. The objective of this study was to use Visible and Near Infrared Spectroscopy (Vis/NIRS) to obtain robust models that can be used to determine total phenolic content and antioxidant activity in Brazilian red wines. samples produced in the São Francisco Valley in Brazil were subjected to transflectance analysis in the spectral range from 400 to 2500nm. After the Vis/NIR analysis, wine samples were analyzed for total phenolic content and antioxidant activity following the standard Folin- Ciocalteu and DPPH methods, respectively. Predictive models were obtained from 48 wine samples, which were divided into calibration (n=28) and validation (n=20) sets using the calibration SPXY algorithm. The Partial Least Squares (PLS) method with cross validation was used to obtain the predictive models. The model performance was evaluated in terms of the standard error of cross validation (SECV) in the calibration process. The performance of the prediction set was evaluated using the standard error of prediction (SEP). In addition, the coefficient of correlation was used in the process of calibration (Rcal) and prediction (Rpre). The model obtained for total phenolic content used two latent variables and yielded the following values for the statistical parameters SECV= 0.24 g.L-1, Rcal=0.70, SEP=0.21 g.L-1, and Rpre=0.61. The model applied to predict the antioxidant activity used three latent variables and provided the following values of SECV=0.64 μMol GAE.mL-1, Rcal=0.83, SEP=0.39 μMol GAE.mL-1, and Rpre=0.77. Our results suggest that Vis/NIR spectroscopy provides an easy and precise approach to determine total phenolic content and antioxidant activity in Brazilian red wines produced in the São Francisco Valley.
Thesagro:	Uva Vinho Composto Fenólico
NAL Thesaurus:	Vitis vinifera subsp. vinifera Red wines Chemometrics
Palavras-chave:	Vis NIRS
Tipo do material:	Resumo em anais e proceedings
Acesso:	openAccess
Aparece nas coleções:	Resumo em anais de congresso (CPATSA)