Aplicação de espectroscopia no Infravermelho médio para detecção de adulteração em produto a base de frango

• Ano de defesa: 2014
• Autor(a) principal: Onara Jandafet Sampaio Martins
• Tipo de documento: Dissertação
• Tipo de acesso: Acesso aberto
• Idioma: por
• Instituição de defesa: Universidade Federal de Minas Gerais; UFMG
• Palavras-chave em Português: Frango- adulteração; Soja; Miúdos; Espectroscopia no Infravermelho com Transformada de Fourier (FTIR); Pele de frango; Alimentos de origem animal Adulteração e inspeção; Carne de ave Industria; Alimentos Controle de qualidade; Espectroscopia de infravermelho
• Link de acesso: http://hdl.handle.net/1843/BUBD-A8YRK5

Chicken burgers are food products susceptible to fraud, either in the form of adulteration (adding ingredients that are not allowed by legislation) or non-conformity (adding ingredients above limits that are allowed by legislation). Given the extensive use of mid-infrared spectroscopy in food adulteration studies, and the non-existence of studies regarding the application of this technique to the analysis of chicken burgers, this study evaluated the potential of mid infrared spectroscopy for detection and quantification of fraud in chicken burgers. The evaluated adulterants were chicken gizzard, liver and heart. The non-conformity analysis was based on the addition of textured soy protein and chicken skin. Initially the proximate analysis of all samples was performed, in order to establish basic composition differences among the samples (free range chicken, broiler chicken, and textured soy protein). Discrimination between free range and broiler chicken samples, and between chicken and its adulterants, according to their infra-red spectra, was based on Principal Components Analysis (PCA). Perishability of the chicken burgers was also evaluated. Discrimination of adulterated or non-compliant samples was based on Partial Least Squares Discriminant Analysis (PLS-DA). Quantitative analysis was based on Partial Least Square Regression (PLS). The model for discrimination between samples that were adulterated (with chicken gizzard, or liver) or not was based on 37 samples, with all but one sample being incorrectly classified. The model for discrimination between samples that were non-compliant because of high levels of chicken skin was based on 18 samples (15 non-compliant) and provided 100% correct classification. The model for discrimination between samples that were non-compliant because of high levels of textured soy protein was based on 18 samples (15 non-compliant), and all but one sample were incorrectly classified. The model for quantification of chicken skin added to chicken burgers (in concentrations ranging from 5 to 59%) presented RMSEC (Root Mean Square Error of Calibration) and RMSEP (Root Mean Square Error Prediction) values of 1.37202% e 3.06618%, respectively. The model for quantification of soy protein added to chicken burgers (in concentrations ranging from 4 to 53%) presented RMSEC and RMSEP values of 0.16254 % and 1.05653%, respectively. The generic model for quantification of adulteration of chicken burgers (gizzard, liver and heart, pure or mixed, in concentrations ranging from 4 to 47%) presented RMSEC and RMSEP values of 0.16254% and 1.05653%, respectively. The results herein obtained confirm the potential of mid-infrared spectroscopy for the development of a fast and reliable methodology for detection of adulteration or non-compliance in chicken burgers.