--- title: 'What Teaware Is Made Of: Porosity, Heat and Vitrification' url: 'https://howtotea.me/science/material-and-firing/teaware-materials' markdown: 'https://howtotea.me/science/material-and-firing/teaware-materials.md' lang: en date: '2026-10-25' description: 'How clay, porcelain, glass and metal change the brew: porosity, water absorption, heat capacity and vitrification. What tests confirm and what stays myth.' --- # What Teaware Is Made Of: Porosity, Heat and Vitrification Editors · October 25, 2026 · 10 min read In lab tests teaware has a moderate effect on tea, through heat, surface and ions: sencha from iron-rich clay had about a fifth fewer tannins than from porcelain, with the same amino acids. Preheating has the most repeatable effect: a cold 100 ml vessel takes about 20°C off the water in a minute. ![An unglazed clay cup and a porcelain cup with the same tea](https://howtotea.me/images/c/o/v/e/r/cover-8e065d1e.webp) Illustration Teaware materials Zisha porosity: 8–13% vs 2.4–2.6% for porcelain Water absorption, GB/T 10816-2024: up to 6.0% for teapots and cups Thermal conductivity, W/(m·K): zisha 1.28–1.34; porcelain 1.55–1.67 The main lever: preheating and wall mass Weakest in the lore: a "breathing" wall and a teapot "memory" A teapot holds the water while the leaf steeps, and the liquor touches its walls the whole time. Labs in Taiwan, Japan and mainland China have tested whether the material matters, comparing Yixing clay, porcelain, glass, tin and steel and measuring catechins, amino acids, minerals and volatiles. The effects are smaller than tea lore claims, but they exist. Three routes hold up: heat, surface and ion exchange. A fourth, the "breathing" wall, rests on interpretation. ## Porosity: What Was Measured Yixing zisha is a ceramic with an open porosity of roughly 8–13% after firing (11 samples dated 1573–1911, fired at 1070–1170°C). White porcelain, vitrified at 1280–1300°C, comes in at 2.4–2.6%, and plain earthenware fired at 930–950°C at 17–18%.[1](#fn:1) That is the scale of **vitrification**: the hotter the firing, the denser the body. Zisha sits in the middle. Porosity varies even within one teapot wall. The outer layer measures 5–9%, the middle 9–14%, the inner 15–17%. The gradient comes from the craft: the maker beats the clay slab with a mallet, which presses grains into the outer surface and leaves it smooth and dense. Makers of [Yixing teapots](https://howtotea.me/teaware/brewing-vessels/yixing) call this the "mingzhen" step: they smooth the outer skin with a paddle while the inside stays loose. Mercury porosimetry on four zisha clays found two pore ranges: 0.02–3 microns and above 40 microns. This is what people call "double porosity." The structure itself can be measured, but conclusions like "breathes yet does not leak" are interpretation. The same manufacturer that describes the two pore types says they are there to hold heat, not to breathe: if the wall let air through, the aroma would escape. Porosity depends on the clay. In one set of measurements zini (紫泥) came in at 7.2%, qingshuini at 13.6% and qinghui duan at almost 15%, while red benshanzhu in a Chinese study measured 2.85%. The last two figures come secondhand, from a 2025 review. Clay names are a weak guide: a Hong Kong dealer recalls that fifteen years before 2014, nobody told "dicaoqing" from "qinghuini." Water absorption on paper The GB/T 10816-2024 standard allows zisha teapots and cups up to 6.0% water absorption, and a teapot filled to the brim must show no water outside after 24 hours. Japanese unglazed Banko kyusu measure 0.08–0.24% when fired at 1150–1200°C. The word "unglazed" covers a wide gap: the standard allows tens of times more than was measured for Banko. ## Heat: Heat Capacity Beats Conductivity Ceramics differ in thermal conductivity by tens of percent: zisha 1.28–1.34 W/(m·K), earthenware 1.09–1.18, porcelain 1.55–1.67. Against metal the gap is two orders of magnitude, with silver at about 430. But for a teapot it is not the number that matters most. Heat leaves mainly by radiation and convection, and there polished silver (emissivity under 0.05) beats glazed porcelain (about 0.9). In a test by the physicist Rutter, water in silver and porcelain pots fell from 84°C to 70.5 and 67°C over 30 minutes. A metal pot does not cool "instantly." The first minutes belong to the **heat capacity of the wall**. A pint of boiling water holds about 2400 J/K, a 600 g porcelain pot about 600 J/K, a 400 g steel pot about 200. Ignoring losses, water in a cold porcelain pot settles at 84°C, in a steel one at 94°C. That is why ISO 3103, the standard behind professional [tea tasting](https://howtotea.me/science/evaluation/tea-tasting), fixes the mass of the pot: 200 ± 10 g for the large one, 118 ± 10 g for the small. Cooling speed has been measured too. Water at 95°C went into vessels that had not been preheated, in a 22.5°C room, and after 11 minutes it was at 70°C in a 185 ml silver pot, 68°C in a 165 ml zhuni pot and 64°C in a thick-walled 150 ml porcelain [gaiwan](https://howtotea.me/teaware/brewing-vessels/gaiwan). Preheated clay catches up with silver, while glazed ceramic loses the most. But the ranking shifts over time. In one Chinese test the gaiwan cooled fastest in the first two minutes, then from minute four to ten a glass tumbler overtook both it and the zisha. A Taiwanese study found the reverse for glazed versus unglazed pots: the white-glazed pot held heat better, which the authors put down to its wall thickness. Taken together, measurements by different authors show that thickness, mass and preheating matter more than the material. What a cold vessel costs Pots of the same clay in a 25°C room, no preheating: after a minute the water was at 79°C in a 100 ml pot, 83°C in 140 ml and 85°C in 300 ml. One [preheat](https://howtotea.me/brewing/basics/preheating-and-rinsing) raised that to 81.5°C (100 ml) and 89°C (300 ml), two to 89 and 91°C.[2](#fn:2) A double wall pulls the other way. Green Huangshan Maofeng kept in a vacuum mug for up to three hours scored higher for "cooked" and "metallic" aroma and lower for "fresh" than the same tea in a glass tumbler, and the gap grew with time. Hours of heat change the liquor itself. ## Surface and Ions: What Really Changes in the Cup Unglazed ceramic adsorbs catechins on its rough surface and barely touches amino acids. In Japanese sencha tests (2 g per 100 ml, 85°C, 1 minute 30 seconds) tannins came to 73.2 mg/100 ml from shudei (iron-rich Tokoname kyusu), 86.9 from black clay and 88.9 from Arita porcelain. The author ties the drop to iron in the material rather than to ceramic in general.[3](#fn:3) The Tokoname saying that tea from their kyusu is softer held up in the lab, though mildly. On a taste sensor, astringency ranged from −2.33 to +1.99, five steps on a scale where most people can tell one step apart. More on these pots in [Tokoname and Banko](https://howtotea.me/teaware/by-region/japan-and-korea/tokoname-banko). ![The same oolong in porcelain, clay and glass cups](https://howtotea.me/images/t/r/i/-/m/tri-materiala-53092f01.webp)Illustration For oolong the data disagree. A Taiwanese study of six materials found more EGC, EGCG and total catechins and less [caffeine](https://howtotea.me/science/chemistry/caffeine) in oolong from zisha and zhuni; a Chinese test found more amino acids and sugars with red clay. The gap with the Japanese figures comes down to method: the Taiwanese poured ready-made liquor into the pots, while the Japanese brewed the leaf in the kyusu itself. The results cannot be compared head to head. Zisha does release calcium, potassium, magnesium, sodium and silicon into ultrapure water, but no iron or manganese, and no heavy metals showed up in 27 days. Yet in oolong from Yixing pots, total minerals and potassium were lower than in other vessels. No study has shown that clay "enriches tea with trace elements." ## What Goes With What In a comparison of five materials on five types of tea (1:50, water at 100°C, 5 minutes), different vessels won. Tin scored best for green and white teas, with more glutamic acid and theanine. Porcelain won for [black tea](https://howtotea.me/teas/black), with more sweetness and freshness. Glass keeps white and green teas fresh but with less body. Zisha scored best with [oolong](https://howtotea.me/teas/oolong) and dark teas, giving a thicker taste and more catechins. Fresh green and black teas do best with a quick temperature drop, well-roasted oolongs and shu puer with the long heat of clay. This is one sample from one Chinese group, not a law: nobody has run blind studies with matched shape, mass and preheating. Vitrification is a separate factor. The Taiwanese potter Wu Mingyi cast ten teapots identical in shape and thickness from a porcelain body and fired them at temperatures from 1280 down to 1170°C. Oriental Beauty brewed in the 1280°C pot scored highest for strength of aroma and flavor, and in the 1170°C pot lowest. A porous vessel tames the "fiery" edge of a fresh roast and mutes the aroma of fragrant, lightly oxidized oolongs, which suit a dense pot. Jason Cohen calls this "muting" in underfired zisha, and ties the defect "tuhei" (吐黑, "black vomiting") to absorption into the pores of badly underfired ware. That is an expert's opinion, not a measurement. If a pot made of [ceramic](https://howtotea.me/teaware/materials/chinese-ceramics) has a dense, ringing body and water does not sink into the wall, it is likely better vitrified than the clay name on the label suggests. ## Myths: What Did Not Hold Up **"The teapot breathes."** A 2016 study found no bactericidal effect in zisha: after a day of soaking, no zone of E. coli inhibition formed around the shards. The authors credit the design for tea keeping overnight, not the clay: a small spout, a tight lid, boiling water. **"An empty teapot smells of tea."** After six months of steeping oolong in the pots (about 250 times), the coating was made of palmitic and stearic acids, the lipophilic part of the liquor. It was nearly identical for fresh and aged oolong, and the nitrogen compounds that give roast aroma wash off with a rinse. Control pots that held only boiling water had far fewer of these acids. A counterpoint: the authors of the study on 1573–1911 shards allow that the inner layer holds organics, but they did not measure it. For cleaning and storage, see [teaware care](https://howtotea.me/teaware/care-and-collecting/care). **"The stain in a kyusu is tannin."** In both unglazed and glazed kyusu, the sencha stain is calcium oxalate, crystals a few microns long. The rainbow sheen at a layer of 400–800 nm is interference, like a soap bubble, and household dish bleach removes it. **"A tea cosy barely helps."** In a 0.8 liter pot, tea under a cosy cooled from 90 to 78°C in 15 minutes, and 4°C more without one. For short gongfu infusions that makes no difference; for a big British pot it does. No blind taste test of a cosy has turned up. Two questions remain open. Is the measured difference noticeable in a gongfu session, where the infusions decide everything? And how large is it next to the way water temperature and steep time drive [extraction](https://howtotea.me/science/water-and-extraction/extraction)? One test on one tea is a hypothesis, not a fact. ## Common questions Does a Yixing teapot soak up tea and smell of it when empty? In a well-fired body there are few open pores, so almost nothing soaks into the wall and most of the interaction happens on the inner surface. The coating people call a teapot's "memory" is mostly palmitic and stearic acids, and roast aromas rinse off with water. This rests on a Taiwanese study and an expert's assessment; nobody has measured absorption into the wall directly. Does clay change flavor more than temperature and steep time? Nobody has compared them directly; none of the material studies set out to. Unglazed iron-rich clay slightly lowers tannins and leaves amino acids alone, while temperature, time and dose shift concentrations far more. Should porcelain and glass be preheated too? Yes. Preheating works for every material, because the heat capacity and mass of the wall outweigh the differences between materials. The smaller the vessel, the more it matters. Notes 1. Wu J. et al. An analysis of the chemical composition, performance and structure of China Yixing Zisha pottery from 1573 A.D. to 1911 A.D. Ceramics International 39 (2013) 2589–2595. [Full text at sic.cas.cn](https://www.sic.cas.cn/zt/lijiazhi/lwzz/201905/P020190513376144195423.pdf). [↩](#fnref1:1) 2. Hojo A. Cooling speed of water differs depending on the volume of teapot, 2018. [hojotea.com](https://hojotea.com/en/posts-191/). Measurements by a tea shop and expert, not a peer-reviewed study. [↩](#fnref1:2) 3. Yamazawa K. Effects of kyusu material on the taste components and color of green tea infusion. Journal of Cookery Science of Japan 55(3) (2022) 168–171. [J-STAGE](https://www.jstage.jst.go.jp/article/cookeryscience/55/3/55_168/_pdf). [↩](#fnref1:3) --- ## Navigation - Parent: [Teaware material and firing](https://howtotea.me/science/material-and-firing.md) - Next: [Kilns and Firing: Reduction, Wood and Kiln Change](https://howtotea.me/science/material-and-firing/kilns-and-firing.md)