Electronic tongue and metabolomics trace how Xinyang Maojian green tea builds its sweet taste

Xinyang Maojian, one of China’s ten most celebrated green teas, is prized for a flavor that seems almost paradoxical: a mellow sweetness and lasting sugary aftertaste emerging from leaves that begin life loaded with bitter polyphenols. A new study published in Food Chemistry: X has now traced exactly where that sweetness comes from, following the tea leaf by leaf through every stage of an industrial manufacturing line and pairing a high-resolution metabolomics platform with an electronic tongue capable of quantifying taste the way human sensors do. The result is a molecular map of green tea quality formation, one that identifies fixation and drying as the two decisive steps that turn fresh, grassy leaves into a cup with clear soup, green leaves, and a smooth, sweet finish.

The research team, led by Lanlan Peng and Xinlin Wei, sourced fresh leaves of one bud and one leaf from a demonstration tea garden in Chenwan Village, Guangshan County, in Xinyang City, Henan Province, on April 10, 2024. The samples then passed through the full Xinyang Maojian protocol: a four-hour spreading period, four minutes of fixation at 200 degrees Celsius in a roller-type machine, thirty minutes of rolling, fifteen minutes of scattering at 86 degrees, six minutes of shaping at 115 degrees, initial drying at 90 degrees, and finally re-drying at 80 degrees after a thirty-minute rest. Samples were collected at all seven stages, from fresh leaf to finished tea, giving the researchers a complete chemical time series of the manufacturing process.

What distinguishes this study from earlier green tea metabolomics work is the analytical database behind it. Conventional untargeted metabolomics in tea research suffers from incomplete metabolome coverage and uncertain compound identification. Here, the team used a self-established tea mass spectrometry database containing more than 3,000 authentic standard compounds, ensuring that every identification was verified against consistent equipment parameters and reference retention times. Samples were extracted with 70 percent methanol, separated on an ultra-performance liquid chromatography system, and analyzed with a quadrupole time-of-flight tandem mass spectrometer operating in positive ion mode at a resolution of 60,000. Peak picking, alignment, and normalization were performed in MS-DIAL, and final compound confirmation required matching accurate mass within 0.03 daltons, retention time within 0.05 minutes, and fragment ions within 0.10 daltons against the in-house database.

In parallel, the researchers deployed an electronic tongue system, the TS-5000Z from Insent in Japan, which uses an array of specific lipid-membrane taste sensors to quantify nine taste attributes: sourness, astringency, bitterness, saltiness, umami, sweetness, richness, bitter aftertaste, and astringent aftertaste. Precisely three grams of crushed tea from each stage were extracted with 90-degree ultrapure water at a ratio of 1:30 for five minutes, cooled, and filtered before testing. A five-member sensory panel, working under the Chinese national standard GB/T 23776-2018, evaluated shape, liquor color, aroma, taste, and infused leaves. The finished tea after re-drying delivered the best overall sensory performance, with uniform tender strips, a bright tender-yellow liquor, a fresh sweet honey aroma, and a mellow, smooth taste with a lasting sweet aftertaste.

The metabolomics survey detected 1,335 metabolite ion features across all samples and identified 90 non-volatile metabolites, including 27 flavonoids, 14 amino acids, 13 catechins and catechin derivatives, 7 organic acids, 6 lipids, 5 alkaloids, 4 nucleotides, 3 vitamins and cofactors, and 12 other compounds. Orthogonal partial least-squares discriminant analysis separated the seven stages into three distinct clusters: fresh leaves alone; fixation, rolling, scattering, and shaping together; and initial drying with re-drying together. Thirty-six key differential metabolites with variable importance in projection values above 1 were pinpointed as the chemical drivers of these transitions. Compared with fresh leaves, the mid-processing cluster showed 37 metabolites up-regulated and only 3 down-regulated, while the drying cluster showed 45 up-regulated and 1 down-regulated relative to fresh leaves, confirming that the transformation begins decisively at fixation.

Among the most striking findings was the fate of theanine, the amino acid most associated with the umami and freshness of green tea. D-theanine itself declined slightly through processing, consistent with the Maillard reaction between theanine and sugars under heat. But theanine glucoside, its glycosylated derivative, rose dramatically during the two drying stages, reaching 7.39 times the fresh-leaf level after initial drying and 13.57 times after re-drying. Correlation analysis revealed that theanine glucoside was positively associated with umami intensity, suggesting that this previously under-characterized compound may be a hidden contributor to the savory depth of well-processed green tea. Glutamic acid, a classic umami compound, rose consistently from fixation onward, reinforcing the electronic tongue’s record of increasing umami through the manufacturing chain.

Sweetness, the signature attribute of Xinyang Maojian, was traced through a targeted gas chromatography-mass spectrometry analysis of 24 sugars and derivatives. Total soluble sugar content climbed steadily to a peak of 49.77 milligrams per gram at the re-drying stage. Sucrose, the dominant disaccharide, accounted for more than 90 percent of total disaccharides throughout processing and reached 38.86 milligrams per gram by re-drying, constituting 78.09 percent of all soluble sugars. Monosaccharides such as glucose and fructose, by contrast, dropped sharply during fixation, likely because they are rapidly consumed in Maillard and caramelization reactions at high temperatures. The accumulation of sucrose and the trisaccharide raffinose appears to reflect heat inactivation of their hydrolytic enzymes, which suppresses breakdown while allowing biosynthesis to continue.

Bitterness told the opposite story. Theobromine, a confirmed contributor to the bitter taste of tea infusions, declined progressively through processing, suggesting that manufacturing either inhibits its synthesis or promotes its metabolism. The bitter aftertaste and astringent aftertaste recorded by the electronic tongue fell in parallel, with initial drying identified as the key step for reducing both. Meanwhile, epiafzelechin, a non-methylated catechin, rose significantly during the drying stages to 1.23 times the fresh-leaf level, and the researchers propose that this compound may contribute a distinct sweet aftertaste. Camellianin A, a flavonoid glycoside, was the only metabolite to decline consistently throughout the entire process, falling to 0.27 times its fresh-leaf abundance by re-drying.

Adenosine offered another dramatic example of processing chemistry. Its content surged during fixation to 5.83 times the fresh-leaf level, accompanied by concurrent rises in adenine and 5′-methylthioadenosine, a sulfur-containing nucleoside whose direct sensory role in thermally processed green tea remains unclear and warrants further study. Gamma-linolenic acid, a common fatty acid, dropped sharply during fixation to 0.42 times its starting abundance before gradually recovering through subsequent stages. Pheophorbide B, a chlorophyll degradation product, climbed steadily to 4.54 times the fresh-leaf level after initial drying, reflecting the progressive breakdown of leaf pigments under heat.

The study concludes that fixation and drying are the two governing steps in the formation of Xinyang Maojian’s taste profile. The increase in sweetness is associated with the accumulation of soluble disaccharides, primarily sucrose; the enhancement of umami is linked to theanine glucoside and glutamic acid; and the reduction in bitter and astringent aftertastes correlates with declining theobromine. By anchoring every sensory observation to a verified molecular measurement, the work provides a theoretical foundation for refining the manufacturing protocol of one of China’s most esteemed green teas, and it elevates theanine glucoside, a compound whose flavor attributes remain largely unexplored, as a promising target for future quality research.

Subject of Research: food chemistry

Article Title: Integrated metabolomics and electronic tongue analysis reveal dynamic changes in taste characteristics and non-volatile metabolites during Xinyang Maojian green tea processing

Article References: Peng, L., Cheng, L., Wei, K., & Wei, X. (2026). Integrated metabolomics and electronic tongue analysis reveal dynamic changes in taste characteristics and non-volatile metabolites during Xinyang Maojian green tea processing. Food Chemistry: X, Article 104601. https://doi.org/10.1016/j.fochx.2026.104601

Image Credits: AI Generated

DOI: Not provided

Keywords: green tea, metabolomics, electronic tongue, Xinyang Maojian, food science, umami, sweetness, Integrated, electronic, tongue, analysis, reveal