To study the impact of processing on quality of Arabica green coffee, the research was divided into two parts: I) the study of the effects of postharvest processing practices on the quality of coffee, and II) the study of the effects of heat pump drying conditions on physicochemical properties of coffee with focus on the determination of the key volatile compounds affected by drying treatments. In part 1, the subtopics of the study were: 1.1) the study of the effects of the selection of coffee cherries differing in maturity (ripe vs unripe) on the quality of green coffee based on volatile analysis and antioxidant activity; 1.2) the study of the effects of different processing methods (digestion by civet, wet and semi-dry processing) on the authenticity of green coffee quality based on volatile analysis, antioxidant activity and the application of an in-house electronic tongue (E-tongue) based on cyclic voltammetry and the principal component analysis (PCA); 1.3) the study of the effects of processing methods (processing with or without civet), subdividing processing (civet feeding conditions; caged vs free-range), the locations of processing (plantation area: Phahee, Doi-Chang, Doi-Tung and Loei), the countries of processing (Thailand, Indonesia and Vietnam), animal species (civet and weasel), and species (Robusta and Arabica) on volatile compounds of roasted coffee; 1.4) the study of the effects of processing methods, of subdividing processing, of the location of processing, of the country of processing, of animal species and of coffee species on overall difference in sensorial properties of coffee brew based on the overall difference method by pair comparison followed by sureness-rating tasks on a four-point scale. This experiment was divided into three sections and the evaluation was based on the difference between samples using E-tongue measurement. The results in the 1st part showed that, in terms of the relationship between the selection of coffee cherries and the quality of green coffee, the volatile compounds and antioxidant activity by FRAP assay were successful in the differentiation of ripe and unripe green coffee. With regard to volatile compounds, 17 vs 5 odorants were identified only in green beans from ripe and unripe berries, respectively. Total phenolic content and DPPH radical scavenging assay could not distinguish between both kinds of beans. The evaluation of the effects of different processing methods on the authenticity of green coffee revealed that there were 12 key volatiles that were successful in the discrimination of civet coffee from others. Total phenolic contents (TPC) and FRAP of civet green coffee (50.38 GAE/g coffee and 80.67 Trolox/g coffee) were significantly different (p<0.05) from the others. The rapid e-tongue also showed its potential to discriminate the profile of multi-processed green coffee into cluster group by the plot of the first principal component 1 (PC1) (94.5%) and the third principal component 3 (PC3) (1.6%). The following set of studies on the effects of processing methods on the quality of 8 roasted coffee samples (civet and without civet) based on volatile analysis indicated that 20 key odorants could be used to separate coffee according to the processing method chosen, subdivided processing, location and species. It was found that 2-furfurylthiol was the most characteristic impact compound with the highest odor activity value (OAV) determined in all coffee samples and it appeared to be successful in indicating the difference between roasted coffee samples from different species and having been subjected to different processing methods. The result of last section in part 1 of the study based on pair comparison/sensory test and the use of E-tongue measurement to evaluate the overall difference in 6 pair of coffee brew showed that there was a relationship between PCA separation via E-tongue and multidimensional scaling (MDS) solutions which revealed a strong correlation between e-tongue and human perception. Both were able to identify differences between civet and non-civet coffees, free-range and caged civet coffees, and civet and weasel coffees. The part 2 of the study, investigated the effects of the use of heat pump drying (40, 45, 50 ?C on quality of wet processed Arabica coffee in comparison with conventional drying process (tray drying at 50 ?C and sun drying), moisture content, water activity (aw), antioxidant activity measurement, color determination, non-volatile compounds (sugars, amino acids, fatty acids, chlorogenic acid), volatile compounds (40 selected key odorants) and sensory evaluation of the aroma by nose using R-index by similar-ranking test were applied. It was found that all heat pump dried samples had brighter colors than tray dried and sun dried coffee. The application of heat pump drying at 50 ?C could dry coffee to the safe moisture content <12 % w.b. with aw of less than 0.6 in shortest time (11 h). Heat pump dried coffee at 50 ?C had the highest total phenolic content and reducing antioxidant power of 53.19 mg GAE/g coffee and 65.27 mg Trolox/g coffee, respectively. it also had the highest content of chlorogenic acid isomers (3,4 and 4,5-dicaffeoylquinic acids) which indicated a good quality of coffee. There were no significant differences observed in sugar, free sulfur amino acids and fatty acids contents of all dried coffee. Heat pump dried coffee at 50 ?C had a closest R-index value to a bench marked coffee (sun dried coffee) at 31.67. Thus, this condition of heat pump dried coffee was selected for further volatile analysis compared with tray dried and sun dried coffee. 2-furfurylthiol was the key odorant that significantly discriminated coffee produced from heat pump drying and the commercial tray drying (p-value <0.05). In term of the calibration in the method used for the determination of high-boiling point volatile, Furaneol? between liquid-liquid extraction and solvent-assisted flavor evaporation, there was no significantly difference observed in the Furaneol? concentration evaluated in these two methods (p> 0.05). Heat pump dried coffee (50 ?C) had the closest similar odor properties to a bench marked coffee (sun dried coffee) with a value at 31.67. Thus, coffee samples subjected to this heat pump drying treatment were selected for further volatile analysis to be compared with tray dried and sun dried coffee. The compound 2-furfurylthiol was the key odorant that significantly differentiated coffee subjected to heat pump drying and the commercial tray drying (p-value <0.05). In terms of the calibration of the method used for the determination of high-boiling point volatile, Furaneol?, between liquid-liquid extraction and solvent-assisted flavor evaporation, there was no significantly difference observed in the Furaneol? concentration evaluated by these two methods (p> 0.05).
