Two Reactions Start When the Leaf Is Bruised, Not One

Tumble a basket of oolong leaf and the bruise you cause sets off two separate reactions, not one. The reaction everyone can see, the reddening at the wounded edge, is polyphenol oxidase converting catechins into pigment, which this site's own piece on green, oolong, and black tea already covers in full (why one plant makes green, oolong, and black tea). The second reaction leaves the leaf's color alone. It runs on a different enzyme, acts on a different class of molecule, and is what actually manufactures the floral smell an oolong is bought for. Chemists have only mapped it in real detail in the last few years, and the mapping changes what "shaking the leaf" is actually doing.

One bruise exposes two substrates that were never in contact before

A tea leaf keeps its chemistry compartmentalized until something breaks the cells open. Catechins sit in the vacuole; polyphenol oxidase (PPO) sits elsewhere in the cell, walled off from its own substrate. Bruise the leaf edge by tossing it in a basket, and the two finally meet air and each other, starting the browning reaction this site's oxidation piece already covers in detail.

That is one substrate-enzyme pair. The same bruise also ruptures the leaf's cell membranes, which hold free polyunsaturated fatty acids, mainly alpha-linolenic and linoleic acid. A second, unrelated enzyme, lipoxygenase (LOX), meets those fatty acids for the first time in the same instant PPO meets the catechins. LOX has nothing to do with color. It is the first step in a chain that ends in the compounds a taster calls floral.

So "shaking" names one mechanical action, the bruise, that opens two separate locks at once, on two substrates that had no reason to meet before that moment.

Oxygen, not the enzyme, sets the pace of the reddening

The oxidation reaction's actual bottleneck was measured directly in 2023, and the finding cuts against the assumption that more bruising or more enzyme drives it faster. Chen, Wang, Ye, Wang, and Xu tracked seven catechins, four phenolic acids, and eleven flavonoid glycosides oxidizing under controlled conditions and found the reaction follows simple, predictable kinetics, with oxidation rate rising in a straight line against oxygen concentration (Chen et al., Food Chemistry: X, 2023). At the roughly 21 percent oxygen in ordinary air, the enzyme sat idle waiting on substrate exposure to air, not the other way round; raising the chamber to 35 percent oxygen sharply accelerated the browning of every compound class tested. Oxygen, not enzyme supply, was the limiting factor. The same study found catechins with a pyrogallol-type ring oxidize faster than the catechol-type ones, and a gallic-acid ester group actually slows a catechin down, which is one reason different cultivars redden at different rates under identical handling.

That single finding explains a piece of oolong processing that otherwise looks like superstition: the maker alternates tossing the leaf with resting it, covered, rather than just tumbling continuously. Tossing does two things at once, damaging more cells and exposing the leaf to fresh air; resting under cover slows the airflow and lets the reaction plateau at a level the maker can judge before the next toss. The craft term for reading that moment, zuoqing (做青), is usually translated flatly as "making green," which loses the fact that the maker is managing an oxygen-limited reaction by hand, section by section, without instruments.

A working Wuyi yancha workshop puts a number on the same principle from the other direction. Its own account of zuoqing practice states the guiding rule outright, kan qing zuo qing, kan tian zuo qing (看青做青，看天做青), "read the leaf to shape the leaf, read the sky to shape the leaf," and gives a concrete case: an evening session is held at 24 to 26 degrees Celsius (about 75 to 79 degrees Fahrenheit) and 70 to 85 percent relative humidity, and leaf picked in rain needs longer handling to compensate for the extra water it is carrying (513ypcy.com⁴). Neither figure comes from a lab. Both track exactly the variable the 2023 kinetics study isolated: how much oxygen and how much time the leaf's surface actually gets.

The floral compounds run on a wound-response pathway, not the browning one

Chen's oxygen-kinetics study was done on black tea fermentation, a fully oxidized end point, not oolong's partial one; its rate constants have not yet been tested directly on a partially oxidized leaf, and that gap is worth naming rather than papering over. What has been tested directly on oolong is the second, aroma-making reaction, and the picture there is sharper.

The fatty acids that LOX releases from the ruptured membrane are cleaved by a second enzyme, hydroperoxide lyase (HPL), into six-carbon aldehydes, principally hexanal and (Z)-3-hexenal, the compounds behind a cut leaf's raw, grassy smell. A third enzyme, alcohol dehydrogenase (ADH), can then reduce those aldehydes into six-carbon alcohols, among them (Z)-3-hexen-1-ol, sometimes called leaf alcohol. Which branch dominates depends on the same variable that governs the browning reaction: airflow. Zhou and colleagues measured ADH activity at roughly 0.16 micromoles per minute per milliliter in leaf turned over under oxygen-poor conditions, against about 0.09 under normal air, in a protocol that mechanically turned the leaf three times, five minutes each turn, an hour apart (Zhou et al., Frontiers in Plant Science, 2021). Starve the leaf of air during a turn and the pathway is pushed toward the greener alcohols; let it breathe and more goes to acids and their esters. The maker's choice of how hard, how often, and how long to toss is, at the molecular level, choosing which branch of this pathway runs.

That LOX-HPL chain only accounts for the grassy notes, though, and grassy is not what a finished oolong is sold on. A 2024 study followed the actual volatile chemistry of Tieguanyin through fresh leaf, withered leaf, and turned-over (shaken) leaf, identifying 579 distinct volatile compounds across the three stages. Withering alone shifted 173 of them measurably from the fresh-leaf baseline; the turning-over stage on top of that shifted 124 more, on top of withering's own changes, evidence that the shaking step is doing its own distinct chemistry, not just continuing withering's (Wu et al., Food Chemistry: Molecular Sciences, 2024). The compounds that rose specifically during turning-over were nerolidol, linalool oxides, indole, benzyl alcohol, and benzaldehyde, the backbone of what a taster would call floral rather than grassy; linalool itself, along with limonene and a few other terpenes, actually peaked at the withering stage and fell once turning-over began, so it is the oxidized derivatives, not the parent compound, doing the floral work here. The same study found that repeated mechanical damage raised expression of the gene for nerolidol synthase (NES), the enzyme that builds nerolidol from an upstream terpene precursor, directly linking the physical act of tossing to a specific gene turning on.

The mechanism behind that gene switch is where the finding gets genuinely strange. Wu and colleagues tie the shift to jasmonic acid (JA) signaling, the same wound-response chemistry a plant deploys generally after mechanical injury or drought stress: rising JA drives up expression of the genes for benzyl nitrile, jasmine lactone, and methyl jasmonate, and the result is more floral and fruity scent in the finished tea (Wu et al., 2024). Jasmonic acid is documented, mechanistic biochemistry, a signaling hormone with a known receptor and gene-expression pathway; describing it that way is not the same as saying the leaf minds being shaken. But the plain fact stands regardless of how it is described: the floral character oolong is prized for is manufactured through the same signaling chemistry a plant uses generally to respond to a wound, not added by processing, and not present in a leaf that is fixed before shaking ever happens.

Why a timer cannot replace the maker's nose

Line the two pathways up against each other and the reason zuoqing resists a fixed protocol becomes a matter of chemistry rather than tradition for its own sake. The reddening reaction is oxygen-limited and roughly linear: more air exposure, more browning, in a fairly predictable line. The aroma reaction is gene-regulated and threshold-driven: mechanical damage has to accumulate before NES and the jasmonate-linked genes switch on in earnest, and once they do, the floral compounds can keep climbing for a stage or two after the tossing itself has stopped. A single fixed number of tosses at a single fixed interval would let one curve run ahead of the other on every batch that varies even slightly in moisture, cultivar, or ambient humidity, which is exactly why kan qing zuo qing, kan tian zuo qing survives as working practice rather than as a rule a machine has fully replaced. A sensor can log oxygen and humidity. Judging when the floral chemistry has caught up to the color, on a specific basket of leaf, on a specific evening, is still a call a person makes by smell.

What is still open is whether the 2023 oxygen-kinetics numbers, measured on a fully oxidizing black tea leaf, hold at oolong's shallower oxidation levels, where the reaction is deliberately stopped early. Nobody has run that specific comparison yet. Until someone does, what exists is two well-mapped reactions on the same leaf under the same hands, studied by two separate teams that were not answering each other's question.