A sensor array based on a nanozyme with polyphenol oxidase activity for the identification of tea polyphenols and Chinese green tea

Green tea is popular among consumers because of its high nutritional value and unique flavor. There is often a strong correlation among the type of tea, its quality level and the price. Therefore, the rapid identification of tea types and the judgment of tea quality grades are particularly important. In this work, a novel sensor array based on nanozyme with polyphenol oxidase (PPO) activity is proposed for the identification of tea polyphenols (TPs) and Chinese green tea. The absorption spectra changes of the nanozyme and its substrate in the presence of different TPs were first investigated. The feature spectra were scientifically selected using genetic algorithm (GA), and then a sensor array with 15 sensing units (5 wavelengths × 3 time) was constructed. Combined with the support vector machine (SVM) discriminative model, the discriminative rate of this sensor array was 100% for different concentrations of typical TPs in Chinese green tea with a detection limit of 5 μM. In addition, the identification of different concentrations of the same tea polyphenols and mixed tea polyphenols have also been achieved. Based on the above study, we further developed a facile and efficient new method for the category differentiation and adulteration identification of green tea, and the accuracy of this array was 96.88% and 100% for eight types of green teas and different adulteration ratios of Biluochun, respectively. This work has significance for the rapid discrimination of green tea brands and adulteration.

Comparative Metabolomics Study of Four Kinds of Xihu Longjing Tea Based on Machine Fixing and Manual Fixing Methods

China Xihu Longjing tea is famous for its good flavor and quality. However, information on its related metabolites, except for flavonoids, is largely deficient. Different processing methods for China Xihu Longjing tea fixing-by machines at both the first and second step (A1), first step by machine and second step by hand (A2), first step by hand and second step by machine (A3), and by hand at both the first and second step (A4)-were compared using a UHPLC-QE-MS-based metabolomics approach. Liquid chromatography-mass spectrometry was used to analyze the metabolic profiles of the processed samples. A total of 490 metabolites (3 alkaloids, 3 anthracenes, 15 benzene and substituted derivatives, 2 benzopyrans, 13 coumarins and derivatives, 128 flavonoids, 4 furanoid lignans, 16 glycosides and derivatives, 5 indoles and derivatives, 18 isocoumarins and derivatives, 4 chalcones and dihydrochalcones, 4 naphthopyrans, 3 nucleosides, 78 organic acids and derivatives, 55 organooxygen compounds, 5 phenols, 109 prenol lipids, 3 saccharolipids, 3 steroids and steroid derivatives, and 17 tannins) were identified. The different metabolic profiles were distinguished using PCA and OPLS-DA. There were differences in the types and contents of the metabolites, especially flavonoids, furanoid lignans, glycosides and derivatives, organic acids and derivatives, and organooxygen compounds. There was a positive correlation between flavonoid metabolism and amino acid metabolism. However, there was a negative correlation between flavonoid metabolism and amino acid metabolism, which had the same trend as prenol lipid metabolism and tannins. This study provides new valuable information regarding differences in the metabolite profile of China Xihu Longjing tea processed based on machine fixing and on manual fixing methods.

A machine learning method for juice human sensory hedonic prediction using electronic sensory features

This study proposed a method that combines fused electronic sensory analysis technology with artificial neural network to predict the human sensory hedonic of fruit juice. Quantitative descriptive analysis (QDA) and the scoring test method were utilized for human sensory evaluation. The first step involved modeling the fused e-sensory features with human sensory attributes, followed by establishing a fitting model of human sensory attributes and acceptance. The R2 and RMSE values obtained were 0.77 and 0.42 (QDA method), and 0.63 and 0.63 (scoring test method). Finally, the relationship between the fusion e-sensory features and the human sensory hedonic was established. Model-1 achieved an R2 of 0.95 and an RMSE of 0.04, while model-2 achieved an R2 value of 0.88 and an RMSE value of 0.21. This study demonstrates the potential of fusing e-sensory technologies to replace human senses, which may lead to the development of devices with simultaneous multiple senses.

An instance-based deep transfer learning method for quality identification of Longjing tea from multiple geographical origins

For practitioners, it is very crucial to realize accurate and automatic vision-based quality identification of Longjing tea. Due to the high similarity between classes, the classification accuracy of traditional image processing combined with machine learning algorithm is not satisfactory. High-performance deep learning methods require large amounts of annotated data, but collecting and labeling massive amounts of data is very time consuming and monotonous. To gain as much useful knowledge as possible from related tasks, an instance-based deep transfer learning method for the quality identification of Longjing tea is proposed. The method mainly consists of two steps: (i) The MobileNet V2 model is trained using the hybrid training dataset containing all labeled samples from source and target domains. The trained MobileNet V2 model is used as a feature extractor, and (ii) the extracted features are input into the proposed multiclass TrAdaBoost algorithm for training and identification. Longjing tea images from three geographical origins, West Lake, Qiantang, and Yuezhou, are collected, and the tea from each geographical origin contains four grades. The Longjing tea from West Lake is regarded as the source domain, which contains more labeled samples. The Longjing tea from the other two geographical origins contains only limited labeled samples, which are regarded as the target domain. Comparative experimental results show that the method with the best performance is the MobileNet V2 feature extractor trained with a hybrid training dataset combined with multiclass TrAdaBoost with linear support vector machine (SVM). The overall Longjing tea quality identification accuracy is 93.6% and 91.5% on the two target domain datasets, respectively. The proposed method can achieve accurate quality identification of Longjing tea with limited samples. It can provide some heuristics for designing image-based tea quality identification systems.

Adsorbent-SERS Technique for Determination of Plant VOCs from Live Cotton Plants and Dried Teas

We developed a novel substrate for the collection of volatile organic compounds (VOCs) emitted from either living or dried plant material to be analyzed by surface-enhanced Raman spectroscopy (SERS). We demonstrated that this substrate can be utilized to differentiate emissions from blends of three teas, and to differentiate emissions from healthy cotton plants versus caterpillar-infested cotton plants. The substrate we developed can adsorb VOCs in static headspace sampling environments, and VOCs naturally evaporated from three standards were successfully identified by our SERS substrate, showing its ability to differentiate three VOCs and to detect quantitative differences according to collection times. In addition, volatile profiles from plant materials that were either qualitatively different among three teas or quantitatively different in abundance between healthy and infested cotton plants were confirmed by collections on Super-Q resin for dynamic headspace and solid-phase microextraction for static headspace sampling, respectively, followed by gas chromatography to mass spectrometry. Our results indicate that both qualitative and quantitative differences can also be detected by our SERS substrate although we find that the detection of quantitative differences could be improved.

Collaborative Analysis on the Marked Ages of Rice Wines by Electronic Tongue and Nose based on Different Feature Data Sets

Aroma and taste are the most important attributes of alcoholic beverages. In the study, the self-developed electronic tongue (e-tongue) and electronic nose (e-nose) were used for evaluating the marked ages of rice wines. Six types of feature data sets (e-tongue data set, e-nose data set, direct-fusion data set, weighted-fusion data set, optimized direct-fusion data set, and optimized weighted-fusion data set) were used for identifying rice wines with different wine ages. Pearson coefficient analysis and variance inflation factor (VIF) analysis were used to optimize the fusion matrixes by removing the multicollinear information. Two types of discrimination methods (principal component analysis (PCA) and locality preserving projections (LPP)) were used for classifying rice wines, and LPP performed better than PCA in the discrimination work. The best result was obtained by LPP based on the weighted-fusion data set, and all the samples could be classified clearly in the LPP plot. Therefore, the weighted-fusion data were used as independent variables of partial least squares regression, extreme learning machine, and support vector machines (LIBSVM) for evaluating wine ages, respectively. All the methods performed well with good prediction results, and LIBSVM presented the best correlation coefficient (R² ≥ 0.9998).

Metabolomics Based on UHPLC-Orbitrap-MS and Global Natural Product Social Molecular Networking Reveals Effects of Time Scale and Environment of Storage on the Metabolites and Taste Quality of Raw Pu-erh Tea

Raw Pu-erh tea (RPT) needs ageing before drinking. However, the influence from environment and time of storage on chemical profile and flavor of RPT is unclear. In this study, the RPTs stored in wet-hot or dry-cold environment for 1-9 years were assessed using metabolomics based on UHPLC-Orbitrap-MS and global natural product social (GNPS) feature-based molecular networking as well as electronic tongue measurement. The results exhibited that the chemical profiles of RPTs were similar at an early stage but started to differentiate from each other at the 5th and the 7th year in wet-hot and dry-cold environments. The discriminating features including N-ethyl-2-pyrrolidinone-substituted flavan-3-ols (flavoalkaloids), unsaturated fatty acids, lysophosphatidylcholines, flavan-3-ols, amino acids, and flavonol-O-glycosides among the three chemical profiles were discovered and analyzed by means of multivariate statistics, GNPS multilibraries matching, and SIRIUS calculation. The metabolomic data were consistent with the results obtained through electronic tongue measurement.

The qualitative and quantitative assessment of tea quality based on E-nose, E-tongue and E-eye combined with chemometrics

Electronic nose (E-nose), electronic tongue (E-tongue) and electronic eye (E-eye) combined with chemometrics methods were applied for qualitative identification and quantitative prediction of tea quality. Main chemical components, such as amino acids, catechins, polyphenols and caffeine were measured by traditional methods. Feature-level fusion strategy for the integration of the signals was introduced to integrate the E-nose, E-tongue and E-eye signals, aiming at improving the performances of identification and prediction models. Perfect results with an accuracy of 100% were obtained for qualitative identification of tea quality grades, based on fusion signals by support vector machine and random forest. Quantitative models were established for predicting the contents of the chemical components based on independent electronic signals and fusion signals by partial least squares regression, support vector machine and random forest. Random forest based on the fusion signals achieved the best performance in predicting the concentration of those chemical components.

Identification of the Rice Wines with Different Marked Ages by Electronic Nose Coupled with Smartphone and Cloud Storage Platform

In this study, a portable electronic nose (E-nose) was self-developed to identify rice wines with different marked ages—all the operations of the E-nose were controlled by a special Smartphone Application. The sensor array of the E-nose was comprised of 12 MOS sensors and the obtained response values were transmitted to the Smartphone thorough a wireless communication module. Then, Aliyun worked as a cloud storage platform for the storage of responses and identification models. The measurement of the E-nose was composed of the taste information obtained phase (TIOP) and the aftertaste information obtained phase (AIOP). The area feature data obtained from the TIOP and the feature data obtained from the TIOP-AIOP were applied to identify rice wines by using pattern recognition methods. Principal component analysis (PCA), locally linear embedding (LLE) and linear discriminant analysis (LDA) were applied for the classification of those wine samples. LDA based on the area feature data obtained from the TIOP-AIOP proved a powerful tool and showed the best classification results. Partial least-squares regression (PLSR) and support vector machine (SVM) were applied for the predictions of marked ages and SVM (R² = 0.9942) worked much better than PLSR.