Understanding 1,000 year old synthesis conditions used by Chinese potters could lead to an easier, more reliable synthesis of epsilon-phase iron oxide, enabling better, cheaper magnetic materials - including those used for data storage.
Jian wares, such as tea bowls, are famous for their shiny black glaze and variable brown and silvery surface patterns known as “oil spot” and “hare’s fur.” The ceramic bowls, produced thousands at a time in giant kilns, were made during the Song dynasty in the Fujian Province of Southeast China between 960 and 1279 AD. Today, surviving Jian bowls are highly prized.
Modern replica of a Tenmoku tea bowl (top) with “oil spot” surface patterns. Close-up of the oil spot pattern from an ancient Jian ware provided by the musum of Jian province (bottom). Photo courtesy of Weidong Li and Zhi Liu.
To make the pottery, ancient artisans used local iron-rich clay coated with a mixture of clay, limestone, and wooden ash. Kiln temperatures of 1300 degrees Celsius (nearly 2400 degrees Fahrenheit) hardened the clay, melted the coating, and bubbled oxygen within the glaze, pushing iron ions to the surface. As the glaze cooled, molten iron flux flowed down the sides of the ceramics and crystallized into iron oxides imparting characteristic patterns.
Based on coloration of oxides in the hare’s fur pattern, experts assumed it only contained the mineral hematite. The oil-spot pattern was thought to be made of the mineral magnetite.
The new analysis reveals hare’s fur patterns contain small quantities of epsilon-phase iron oxide mixed with hematite, while oil spot patterns boast large quantities of highly pure epsilon-phase iron oxide, a finding no one expected. The analysis team used optical microscopy, electron microscopy, Raman spectroscopy, and synchrotron x-ray techniques to assess the chemical composition and crystalline structures within the glaze.
a) Le Bail fit of the HF (“hare’s fur”) Fe-rich surface region, showing the presence of hematite and ε-Fe2O3; b) Le Bail fit of the OS (“oil spot”) Fe-rich surface region, showing the presence of ε- Fe2O3 . For a) and b), powder patterns were recorded with a Bruker D8 Advanced and a beam of ~100 μm in diameter. The high intensity of the 002 reflection at 18.84° (2θ CuKα) indicates preferred orientation. c) 2D diffraction pattern obtained from the surface of the HF sample, from a low content Fe zone (HF-Surf-G) and from a Fe-rich region (HF-Surf-Fe). Theoretical positions for the ε- Fe2O3 are indicated in red; d) 2D diffraction pattern obtained from 2 Fe-rich regions at the surface of the OS sample. Theoretical positions for the ε- Fe2O3 are indicated in red. For c) and d), 2D patterns were obtained at ALS beamline 12.3.2 using a ~15 μm*15 μm microbeam. While the 2D diffraction rings look homogeneous for HF-Surf-Fe, they are highly textured for OS-Surf-Fe. Credit and link: doi:10.1038/srep04941
Epsilon-phase iron oxide was first identified in 1934, and only within the last ten years has it been fully characterized. Marked by extremely persistent magnetization, epsilon-phase iron oxide could hold the key to better, cheaper permanent magnets used in data-storage and other electronics. Moreover, the epsilon phase is non-toxic and highly resistant to corrosion. Unfortunately, modern synthesis techniques have only managed to grow tiny, crystals often contaminated with hematite.
The oil spot pattern, with its pure epsilon phase of iron oxide, could hold the key to better synthesis techniques, says Catherine Dejoie, scientist at Berkley Lab’s Advanced Light Source and ETH Zurich. “The next step will be to understand how it is possible to reproduce the quality of epsilon-phase iron oxide with modern technology and to identify and extract synthesis conditions and other factors to obtain large crystals of pure epsilon phase.”
Catherine Dejoie, Philippe Sciau, Weidong Li, Laure Noé, Apurva Mehta, Kai Chen, Hongjie Luo, Martin Kunz, Nobumichi Tamura, Zhi Liu, 'Learning from the past: Rare ε-Fe2O3 in the ancient black-glazed Jian (Tenmoku) wares', Scientific Reports 4, Article number: 4941 doi:10.1038/srep04941