Application of ATR-FTIR for Green Arabica Bean Shelf-Life Determination in Accelerated Storage

Abstract

Coffee bean oxidation is associated with enzymatic and non-enzymatic browning, the degradation of desirable aromatic compounds, the development of undesirable flavors, increased susceptibility to microbial spoilage, and volatile compound losses. This study investigated natural dry process (DP) and honey process (HP) green coffee beans stored in GrainPro^® bags for 0, 5, 10, and 20 days under accelerated storage conditions at 30 °C, 40 °C, and 50 °C with relative humidity of 50%. A kinetic model was used to estimate the shelf life of the green coffee beans. DP recorded durability of 45.67, 29.9, and 24.92 days at 30 °C, 40 °C, and 50 °C, respectively, with HP 60.34, 38.07, and 19.22 days. Partial least squares (PLS) analysis was performed to build the models in order to predict the shelf life of coffee based on peroxide (PV) and thiobarbituric acid reactive substances (TBARS) values. In terms of prediction with leave-one-out cross-validation (LOOCV), PLS provided a higher accuracy for TBARS (R2 = 0.801), while PV was lower (R2 = 0.469). However, the auto-prediction showed good agreement among the observed and predicted values in both PV (R2 = 0.802) and TBARS (R2 = 0.932). Based on the variable importance of projection (VIP) scores, the ATR-FTIR peaks as 3000-2825, 2154-2150, 1780-1712, 1487-2483, 1186-1126, 1107-1097, and 1012-949 cm^-1 were identified to be the most related to PV and TBARS on green coffee beans shelf life. ATR-FITR showed potential as a fast and accurate technique to evaluate the oxidation reaction that related to the loss of coffee quality during storage.

Keywords: infrared spectroscopy; multivariate analysis; oxidation; rancidity.

Figure 1

Packages for green coffee bean storage. GrainPro^® bags (a) natural process, and (b) honey process.

Figure 2

Green coffee bean storage in GrainPro^® bags after 20 days under accelerated storage condition.

Figure 2

Green coffee bean storage in GrainPro^® bags after 20 days under accelerated storage condition.

Figure 3

Effect of storage conditions on changes in (a) moisture contents and (b) water activity during accelerated storage conditions. Different capital letters indicate significant differences in temperature at p < 0.05; different lowercase letters indicate significant differences in storage time at p < 0.05. DP-D30C: natural process storage at 30 °C; HP-D30C: honey process storage at 30 °C; DP-D40C: natural process storage at 40 °C; HP-D40C: honey process storage at 40 °C; DP-D50C: natural process storage at 50 °C; HP-D50C: honey process storage at 50 °C.

Figure 4

Changes in the oxidation reactions of green coffee beans during the natural and honey processes; (a) peroxide value, and (b) TBARS. Different capital letters indicate significant differences in temperature at p < 0.05; different lowercase letters indicate significant differences in storage time at p < 0.05. DP-D30C: natural process storage at 30 °C; HP-D30C: honey process storage at 30 °C; DP-D40C: natural process storage at 40 °C; HP-D40C: honey process storage at 40 °C; DP-D50C: natural process storage at 50 °C; HP-D50C: honey process storage at 50 °C.

Figure 5

ATR-FTIR spectra of DP green coffee bean storage at (a) 30 °C, (b) 40 °C, and (c) 50 °C and HP storage at (d) 30 °C, (e) 40 °C, and (f) 50 °C.

Figure 6

PCA score plot of PC1 against PC2 with the oxidative parameters during accelerated storage (a). The correlation graphs present the expected and predicted PV (b) and TBARS (c) values (circle symbols present the LOOVC, while triangle symbol presents the Auto-Prediction). The overlap VIP scores (highlighted zone means VIP ≥ 1) with FTIR spectrum of PV (d) and TBARS (e).