Submitted by Sandra Pereira (srpereira@ufrrj.br) on 2018-08-20T12:37:22Z No. of bitstreams: 1 2010 - Raquel Duarte da Costa Cunha Bandeira.pdf: 1478855 bytes, checksum: f7637b4df6dd802e4e0003f4fb48d8cd (MD5) / Made available in DSpace on 2018-08-20T12:37:22Z (GMT). No. of bitstreams: 1 2010 - Raquel Duarte da Costa Cunha Bandeira.pdf: 1478855 bytes, checksum: f7637b4df6dd802e4e0003f4fb48d8cd (MD5) Previous issue date: 2010-09-24 / Coffee is an extremely complex food matrix and has an important role in the world?s economy, especially in producing and exporting countries like Brazil. However this product may suffer from technical barriers imposed for exportation because of the possible presence of ochratoxin A, which is nefrotoxic and carcinogenic mycotoxin found in many foods including coffee. The aim of this study was to implement chemical metrology concepts in the development and validation of Liquid Chromatography with Mass Spectrometry in tandem (CLAE-EM/EM) method for identification and quantification of ochratoxin A in green and roasted coffee estimating uncertainty of measurement according to directive 2002/657/EC and Inmetro guidelines (DOC-CGCRE-2010). The extraction method was based on Pittet (1998) and chromatographic parameters were: flow rate of 0.3 mL/min, mobile phase 80:20 water trifluoracetic acid 0.05 %: methanol trifluoracetic acid 0.05 %, injection volume of 50 PL, injection mode Full loop, isocratic mode. The column was Synergi Hydro C18. The mass spectrometry parameters were optimized and transitions selected based on the colision energies monitored were m/z 404 >358 (-10.5 V) and m/z 404 >239 (-20.5 V). From the validation procedure, methods were considered seletive. The evaluation and verification of matrix effect was performed by comparing variances and averages using F and t test. Value of Fcalculated for green coffee (25.2152) and roasted coffee (104.0353), were higher than Ftable (4.0426). Value of t calculated for green (5.0214) and roasted coffee (10.1997) were higher than ttable (2.0106). Both methods were considered linear in the working range of calibration curve with linear correlation coefficients (r) of 0.98188 and 0.91754 for green and roasted coffee, respectively.The quantification and detection limits were 1.2 Pg/kg and 3.0 Pg/kg; 0.36 Pg/kg and 1.0 Pg/kg, for green and roasted coffee respectively. The average recoveries, RSDr and RSDR were in range of 90.45 ? 108.81 %, 5.39 ? 9.94 % and 2.20 ? 14.34 % for green coffee and 89.02 ? 108.85 %, 2.43 ? 13.73 % and 12.57 ? 17.84 % for roasted coffee. All results obtained were considered within acceptable levels according to literature. Measurement value and expanded uncertainties (U) for ochratoxin A were mass fraction w = (11.50 ? 1.11) and w = (4.63 ? 0.63) for green coffee and roasted coffee. Both methods developed and validated using a high sensitivity technique, that allowed detection, confirmation and quantification of ochratoxin A in green and roasted coffee with a estimated uncertainty of measurement, and in the future these methods can be used to help overcome possible technical barriers imposed for exportation of Brazilian coffee. / O caf? constitui uma matriz extremamente complexa e tem importante papel na economia mundial, especialmente nos pa?ses produtores e exportadores como o Brasil. No entanto tem sido alvo de barreiras t?cnicas devido a uma subst?ncia denominada ocratoxina A, micotoxina potencialmente nefrot?xica e nefrocarcinog?nica encontrada em muitos alimentos inclusive o caf?. O presente trabalho tem como objetivo implantar os conceitos da metrologia qu?mica no desenvolvimento, e valida??o do m?todo para identifica??o e quantifica??o de ocratoxina A no caf? verde e caf? torrado estimando a incerteza da medi??o e utilizando a t?cnica de Cromatografia L?quida acoplada a Espectrometria de Massas em s?rie (CLAE-EM/EM) seguindo os crit?rios da diretiva EC-657/2002 e o documento orientativo do Inmetro (DOCCGCRE- 2010). A metodologia de extra??o baseou-se em Pittet (1998) e os par?metros cromatogr?ficos foram: fluxo de 0,3 mL/min, fase m?vel 80:20 ?gua ?cido trifluoroac?tico 0,05%: metanol ?cido trifluoroac?tico 0,05 %, volume de inje??o de 50 PL, com o modo de inje??o Full loop e sistema de elui??o isocr?tico. A coluna utilizada foi Synergi Hydro C18. As condi??es do espectr?metro de massas foram otimizadas e a transi??o selecionada de acordo com suas energias de colis?o foram m/z 404 >358 (-10,5 V) e m/z 404 >239 (-20,5 V). A partir da valida??o os m?todos propostos foram considerados seletivos, a avalia??o e comprova??o do efeito matriz foi realizada atrav?s da compara??o das vari?ncias e das m?dias atrav?s do teste F e teste t. O Fcalculado para o m?todo caf? verde (25,2152) e caf? torrado (104,0353), apresentaram valores maiores que o Ftabelado (4,0426). O tcalculado para o caf? verde (5,0214) e torrado (10,1997) apresentaram valores superiores ao ttabelado (2,0106). Os m?todos foram considerados lineares em toda a faixa de trabalho da curva de calibra??o com os coeficientes de determina??o linear (r) de 0,98188 e 0,91754 para matriz caf? verde e caf? torrado, respectivamente. O limite de quantifica??o e detec??o para os m?todos propostos foram de 1,2 Pg/kg e 3,0 Pg/kg para caf? verde e 0,36 Pg/kg e 1,0 Pg/kg para caf? torrado. Os valores das recupera??es m?dias, DPRr e DPRR variaram na faixa de 90,45 - 108,81 %, 5,39 - 9,94 % e 2,20 - 14,34 % para caf? verde; e de 89,02 - 108,85 %, de 2,43 - 13,73 % e 12,57 - 17,84 %, para caf? torrado. Todos os resultados obtidos encontram-se dentro dos limites comumente aceit?veis na literatura. Todos os resultados de medi??o e as incertezas expandidas (U) para ocratoxina A foram as fra??es m?ssicas W = (11,50 ? 1,11) Pg/kg e W = (4,63 ? 0,63) Pg/kg para caf? verde e caf? torrado, respectivamente. Os m?todos desenvolvidos e validados utilizaram t?cnica de elevada sensibilidade, permitindo a detec??o, confirma??o e a quantifica??o de ocratoxina A no caf? verde e caf? torrado com c?lculo da incerteza, podendo auxiliar futuramente na supera??o das barreiras t?cnicas para exporta??o do caf? brasileiro.