Nuclear magnetic resonance and high performance liquid chromatography chemical analysis of peruvian roasted coffee beans

Abstract

Coffee is the most traded agricultural commodity in the world. Currently, Peru is considered the third principal producer of Coffea arabica in South America, and the sixth worldwide, accounting for 6 % of the global production. However, most of the coffee (~ 99 %) is exported as green beans due, in part, to the fact that the local quality control of the roasting process is not yet optimal. Hence, it is crucial to develop a more standardized process for the quality control of roasted coffee beans that will allow local Peruvian coffee farmers to introduce to the market a more valuable product. In the thesis presented here, work associated with the project FINCyT-PIPEI-PUCPCENFROCAFE- 2012 on the quantitation of the main compounds developed during the roasting process was embraced using NMR and HPLC-DAD methodologies. Attention was focused on the secondary metabolites that, from a flavor-aroma perspective, are of interest: caffeine, 5- caffeoylquinic acid, trigonelline, 1-methylpyridinium ion and nicotinic acid, and 5- hydroxymethylfurfural as a marker of deterioration. One- and two-dimensional NMR techniques allowed the simultaneous identification of eleven compounds known to be associated with the flavor and aroma of coffee. The NMR quantitation of five compounds was performed using ERETIC2 and Standard Calibration Curves and the results were validated by a new HPLC methodology, which constitutes the only validated methodology currently available for the simultaneous quantitation of these five compounds. The percentage difference among this three methods was within acceptable values (1 – 20%) for most of the compounds. It was demonstrated that these numbers were sample dependent. In addition, PCA analyses of quantitative data (NMR and HPLC-DAD) allowed the discrimination of coffee samples acording to the degree of roasting, as well as to their origin (instant coffee, speciality coffees from different regions in Peru). Hence, these preliminary results indicate that NMR and HPLC can be used as quality control tools to optimize the roasting conditions of Peruvian specialty coffee. Recommendations are included in this work to improve further the error percentages between the NMR and HPLC data that in some cases (for 5-caffeoylquinic acid and nicotinic acid) were high.