Epsilon-phase iron oxide is one of the rarest polymorphs of Fe₂O₃. It is metastable — thermodynamically unfavored, persisting only because the energy barrier to transformation traps it in place. Modern synthesis requires precise nanoparticle confinement or carefully controlled sol-gel routes. The compound is of interest for its unusually strong magnetic coercivity at room temperature, a property that makes it a candidate for high-density recording media.
Song dynasty potters in Fujian Province made it routinely — in tea bowls.
Jian ware, produced between 960 and 1279 AD, is prized for its black glaze decorated with patterns called “hare's fur” and “oil spot.” These are not painted. They form during firing: iron-rich clay oxidizes at approximately 1,300°C, molten iron flux flows down the sides of the bowl, and crystallization during cooling produces the streaked or spotted surface. Dejoie and colleagues at the Advanced Light Source found that these patterns contain ε-Fe₂O₃ mixed with hematite — more in the brown oil spots, less in the silver hare's fur.
Reader note (2026-11-17): Checked against the open full text of the Scientific Reports paper (Dejoie and colleagues, 2014, DOI 10.1038/srep04941). The finding matches: epsilon-Fe2O3 found on the glaze surfaces of Jian wares, with synchrotron diffraction at the Advanced Light Source. The essay says there is more epsilon phase in the brown oil spots and less in the silver hare's fur. The paper's labels are the other way round: the Hare's Fur (HF) pattern is the brownish one and the Oil Spot (OS) pattern is the silvery one, and the paper reports that the surface crystals of the OS sample are up to several micrometers in size and exclusively epsilon-Fe2O3, arranged in a periodic two-dimensional fashion, while epsilon-Fe2O3 is also present in the HF sample. The paper gives the firing temperature as 1200 to 1300 degrees Celsius and says the glaze's high calcium and manganese content may change the interface energy of the epsilon phase in the glass matrix; it presents the earlier accepted view as hematite on HF surfaces and magnetite-hematite mixtures on OS surfaces.
The potters were not optimizing for a rare iron oxide polymorph. They were optimizing for beauty. But the conditions that produce the most visually striking patterns — the temperature window, the iron concentration, the cooling rate — happen to be the conditions that stabilize epsilon-phase iron oxide. The aesthetic constraint located a synthesis route that the rational approach, working forward from the desired compound, had difficulty reaching.
The lesson is not that ancient artisans were secretly doing materials science. It is that optimization for one property can inadvertently satisfy the conditions for another when the parameter spaces overlap. The potters searched the space of beautiful surfaces. That search passed through the region where ε-Fe₂O₃ is stable. They didn't need to know it was there.