Rapid quality determination of mulberry leaf tea during storage using electronic nose and electronic tongue

Abstract

This study was aimed to develop rapid methods based on electronic nose (e-nose) and electronic tongue (e-tongue) to determine the quality attributes and phytochemicals of mulberry leaf tea. This study was divided into three parts: (1) to determine the effect of cultivar on the quality attributes and phytochemicals of mulberry leaf tea, (2) to evaluate the effect of packaging and storage time on quality attributes and phytochemicals of mulberry leaf tea, and (3) to employ e-tongue combined with chemometrics for prediction of phytochemicals and antioxidant activity of mulberry leaf tea.

Mulberry leaf tea prepared from three different mulberry cultivars including Burirum 60 (BR), Khunphai (KP), and Sakonnakhon (SK) were compared in terms of total phenolic content (TPC), total flavonoids content (TFC), antioxidant activity (FRAP and DPPH), volatile components, and sensory attributes. Volatile components of the samples were analyzed using SPME-GC/MS and e-nose. Quantitative descriptive analysis was performed to evaluate sensory attributes (e.g. greenness, brownness, green odor, dried-leaf odor, and astringency). The results showed that cultivar significantly affected (p<0.05) the quality attributes and phytochemicals of the mulberry leaf tea. KP cultivar had the highest values of TPC, TFC, FRAP and DPPH. The distinct patterns of volatile components including aldehyde, ketone, alcohol, ester, and acid as a factor of different mulberry cultivars were observed. The results of sensory analysis showed that color and odor attributes were significantly different (p<0.05) among three mulberry cultivars. SK cultivar had the highest greenness score (7.63±0.51) but the lowest brownness score (4.27±0.52). The highest score of green odor (7.81±0.67) and dried-leaf odor (8.89±0.59) were observed in BR cultivar. Principal components analysis (PCA) was used to analyze the e-nose response data and with 2 components (PC1 and PC2) could explain 94% of the total variance.

In order to investigate the effects of packaging and storage time on the quality of mulberry leaf tea, the sample prepared from mulberry leaves of Burirum 60 cultivar was packaged in a linear low-density polyethylene laminated aluminum (AL) bag and polypropylene (PP) bag and stored at 30±1°C (75% RH). The changes in tea qualities including water activity (aw), color (L*, a*, b*), phytochemicals (TPC and TFC), volatile components, and sensory attributes were determined at three-month interval for 18 months. The results showed that packaging material and storage time significantly (p<0.05) affected quality attributes and phytochemicals of the mulberry leaf tea. As the storage time increased, aw and a* values of the samples in PP increased but L*, b*, TPC, TFC, FRAP and DPPH values decreased, whereas the quality attributes of mulberry leaf tea were not significantly different throughout the storage. After 18 months of storage, TPC, TFC, FRAP and DPPH values of the samples in PP lost by 28.84, 34.39, 33.31, and 30.34%, respectively. The intensity of volatile compounds produced from lipid oxidation such as hexane and 4-oxo-2-nonenal of the mulberry tea in PP significantly (p<0.05) increased. The greenness, green odor, and dried-leaf odor scores of mulberry leaf tea in PP significantly (p<0.05) decreased while the brownness scores significantly (p<0.05) increased. PCA was performed to discriminate the quality of the samples by e-nose response data and the first two PCs (PC1 vs. PC2) could explain 98% of the total variance. The results suggested that e-nose is able to discriminate quality of the samples as a result of different storage periods.

In this research, e-tongue was used to monitor phytochemicals and antioxidant activity of mulberry leaf tea during storage. Partial least squares (PLS) regression was applied to predict phytochemical values from the e-tongue response data. The PLS models for TPC, TFC, FRAP and DPPH achieved good predictability with high values of coefficient of determination (r2>0.9) and gave low mean square error of prediction. The results of this study indicated that e-nose and e-tongue have proved to be a non-destructive, rapid, and reliable method for quality discrimination of the mulberry leaf tea and have potential as an alternative to the conventional analyses of quality attributes and phytochemicals of the mulberry leaf tea.