What Steeping Time Does to the Metals in Your Oolong

What Steeping Time Does to the Metals in Your Oolong

The Order measures the leaf. Regulators do too: a mass spectrometer, a sample bag, a limit in milligrams per kilogram. But no one drinks the leaf. Between the dry leaf and the cup sits a step this Order holds sacred and the contamination literature, before November 2025, had almost never tested: how long the water and the leaf are actually left together.

In November 2025, a team publishing in Frontiers in Nutrition¹ tested it. Wu, Peng, Ou, Mhungu and their colleagues took 108 oolong teas sourced from across China and measured six toxic elements (lead, cadmium, aluminum, copper, manganese and nickel) plus fifteen rare earth elements, then brewed them under systematically varied conditions: two water temperatures, 90C and 100C, and a range of contact times running from five seconds to two hours.

The result that matters is simple. Time did most of the work. At 90C, the lead content of the liquor rose from 0.179 to 0.662 micrograms per liter as the steep ran from five seconds out to 120 minutes. That is roughly a 270 percent increase, driven by nothing but how long the leaf sat in the water.

Temperature refused to behave

The intuitive model of extraction is a single dial: hotter water pulls out more of everything. That model is wrong here, and it is wrong in an interesting way.

Cadmium leached less at 100C than at 90C. Scandium leached more. Across the six toxic elements and fifteen rare earths, the temperature effect was inconsistent and element-specific, with no clean rule holding across the set. Solubility, the chemical form the element takes in the leaf, and what else in the matrix competes for it all move independently, and the sum of those movements does not resolve into "hotter equals more."

This is the finding that should adjust how a careful drinker thinks. A hot brew is not simply a strong brew scaled up. It is a different extraction with a different profile, and the elements you might most want to know about do not all travel in the same direction when you raise the kettle.

The curve flattens early

Leaching for most elements rose fastest in the first 5 to 10 minutes. After that the rate of increase slowed and the concentration plateaued.

Read that curve backward and it says something useful about where the sensitivity lives. The difference between a five-second infusion and a ten-minute one is large. The difference between a ten-minute infusion and a two-hour one is comparatively small, because by ten minutes most of what was going to come out has come out. The steep part of the curve sits precisely in the range where ordinary Western brewing operates and where careless brewing (a mug with the leaf left in, a pot forgotten on the counter) runs well past.

The samples that scored worst are ours

The 2025 study did not find contamination spread evenly across its 108 teas. Tieguanyin-variety samples and Fujian-province samples carried the highest levels of any group tested, and the Fujian samples significantly exceeded China's safety standards for lead and aluminum specifically.

We will not soften that. Tieguanyin is Anxi's cultivar, Anxi is in Fujian, and this Order's own practice is built on both. A finding is not less true because it lands on the thing you love.

What the study does not say

It does not say the tea is dangerous.

The measurement here is leaching potential: how much of an element moves out of the leaf and into the liquid under a given set of brewing conditions. That is not the same as gastrointestinal bioaccessibility, which asks how much of what is in the cup can actually cross into the body, and it is not an absorbed-dose health-risk figure. The 2025 paper establishes what brewing method does to concentration. It leaves the dose question to the literature that has already asked it.

That literature is worth reading beside it, because it reached a different-shaped answer. Guo, Zhang, Lai and Wang, writing in the Journal of Food Composition and Analysis² in 2015, measured rare earth oxide leaching from oolong, the same family of elements behind China's deleted rare-earth standard, at roughly 9.86 to 17.4 percent of the leaf's total content on a single infusion, rising to roughly 25.8 to 32.6 percent across five successive infusions. Their conclusion was that typical tea drinking does not pose a meaningful rare-earth health risk, even for habitual drinkers.

The conventional metals got the same treatment in 2022. Yao and colleagues, publishing in Foods³, tested Tieguanyin specifically, from Fujian. They found mean lead of 0.84 mg/kg (range 0.21 to 2.00) and mean cadmium of 0.04 mg/kg (range non-detect to 0.11), and calculated estimated daily intakes for all five metals far below tolerable limits, with a hazard index between 1.1x10^-2 and 1.7x10^-2 and no cancer-risk threshold exceeded.

But they flagged something inside that reassurance. Lead, nickel and cadmium deserve attention, they wrote, precisely because of how efficiently they extract into the liquid during brewing. Lead in the brewed liquor can approach the drinking-water limit of 0.01 mg/L even when the dry leaf sits nowhere near a food-safety limit. Those are two different standards written for two different purposes, and a leaf that passes one comfortably can produce a liquor that presses against the other. Brewing is what moves it between them.

Two literatures, one variable neither controlled

Line the three papers up and the shape becomes clear. One body of work asked whether rare earth elements in tea were dangerous and said no. A separate body of work, years later and about a different class of elements entirely, asked whether conventional toxic metals in Tieguanyin were dangerous and also said no, while pointing at extraction efficiency as the part that merits watching.

Neither treated brewing parameters as the object of study. Both held brewing roughly constant and varied the tea. The 2025 paper inverted that, and in doing so found that the variable sitting between the leaf and the drinker moves the number in the cup by 270 percent on its own, without any change in the leaf at all.

The seventh steep, and an honest question

This Order brews in short infusions, poured off in a single breath, many of them rather than one long one. A first steep may run five to ten seconds. The seventh, brewed patiently, carries aromatics the first never released.

That is exactly the brewing shape that keeps any single infusion's contact time in the flat, early part of the leaching curve.

We are not going to claim more than that. Gongfu brewing was not invented as a safety practice. It exists because short infusions produce a sequence of distinct, legible cups where one long steep produces a single muddled one, and because a small vessel and a fast pour give the drinker control over a leaf that rewards control. Nobody in the lineage was thinking about lead. The 2015 data also cuts against any tidy claim of safety-by-technique: five successive infusions extracted 25.8 to 32.6 percent of the leaf's rare earth content, more than double a single infusion's 9.86 to 17.4 percent. Many short steeps still add up to real total extraction. What they change is that no single steep sits long enough to reach the flat top of the curve.

So the honest statement is narrow. A tradition built entirely around flavor happens to land where a safety-minded brewer would land, and the newest data on why that positioning matters exists only because a team finally varied the kettle and the timer instead of varying the tea.

The next question, the one nobody has yet answered, is what fraction of that leached lead the body can actually take up. Until someone runs the bioaccessibility work on a gongfu-style infusion series, what we have is a curve, a plateau at 5 to 10 minutes, and a number: 0.179 micrograms per liter at five seconds, 0.662 at two hours.