Given the fact that coffee is highly hygroscopic (Ortalá et al., 1998), it is probable that the water adsorbed in the samples was the major cause for TAG hydrolysis during storage (Fig. 2), and therefore could have blunted the effects of temperature and atmosphere on TAG reduction during storage. On the other hand, the roasting process promotes
free radical formation and is associated with pyrolysis reactions (Morrice, Deighton, Glidewell, & Goodman, 1993) that can accelerate degradation. Possibly, H 89 order free radicals initially present in all the fresh coffee samples might explain the absence of significant differences between inert and oxidizing atmospheres. The interaction between storage time and atmosphere influenced the total TAG content in the 1st, 3rd, and 4th months of storage of light-medium samples (Fig. 2). During these months, the highest contents of TAG were observed in samples under oxidant atmosphere (Fig. 2 and Table 1). It is possible that losses of more thermolabile compounds in oxidant atmosphere, as previously mentioned (Pérez-Martínez et al., 2008; Toci, 2010), have caused this apparent increment in TAG contents. Sigmoidal kinetic curves were obtained for TAG degradation in both roasting degrees (Fig. 2). This indicates a two-step hydrolysis process. In Fig. 2, two periods of stability may be observed in total contents
of TAG during storage, from 2 to 3 months and from 4 to 6 months of storage for the light-medium sample, and from 1 to 2 months and from 3 to 5 months of storage for the dark-medium sample. Tanespimycin datasheet These results suggest a decrease in hydrolysis in these periods. Ortalá et al. (1998) also observed a slow kinetic of lipid degradation during the first 100 days (≈3 months) of storage, followed by 100 days of stability. The classical molecular model for lipid oxidation (Frankel, 2005) establishes that reactions occur through
a chain mechanism controlled DNA ligase by free radical formation, with three typical steps: initiation, propagation, and termination. The main factor affecting the reaction rate was the initiation reaction. On the basis of the model of Koelsch, Downes, and Labuza (1991), as well as on the basis of the present data, it appears that a monomolecular or bimolecular reaction can be responsible for the initiation step of the oxidative chain in coffee, through hydroperoxide decomposition. It depends on the initial concentration of these compounds, as observed in other products. So, during the first months, the initially low hydroperoxide concentration, as also observed by Ortalá et al. (1998) for roasted coffee, favors the monomolecular initiation and, when a critical value is attained, in line with the reaction progress, the bimolecular mechanism becomes more controlled. In the light-medium control sample, FFA content was 0.