Hazardous Color

*notice: no kiln shelves were irrevocably damaged in these tests*

*notice: no kiln shelves were irrevocably damaged in these tests*


Above: Installation images of numbered donut tests and blind multiple color tiles.

In ceramic work, color is predominately reserved for surface decoration. Through centuries of glaze research, nearly any color[1] can now be achieved through fine-tuning the variables of a glaze recipe’s ingredients and percentages, kiln firing atmosphere, glaze firing temperature and base clay body color. Glaze experimentation is a well-trodden path and one that constitutes a significant portion of many studio potters’ practice. Wanting to take my research project into new territory, I decided to test the same colorants used in glazes but mixed directly into a clay body recipe.

When considering why coloring clay isn’t a more popular practice, I first speculated that it was some antiquated Bernard Leach sensibility stemming from his “truth to materials, truth to process” mantra.  By necessity some area of a fired ceramic form always reveals its material “truth” since any glazed surface touching a kiln shelf would irrevocably seal the clay object to said kiln shelf. This exposure is often the foot ring of a functional pot or the bottom plane of a ceramic sculpture. Centuries ago, ceramics had a history of disguising itself as other materials when it was less costly to create in clay rather, than say, bronze. However, this kind of camouflaging is a less common modern pursuit (especially by studio potters) and my hypothesis of a material honesty may have some ground but I ultimately think there are larger considerations for why clay isn’t often colored beyond its native state. Fully saturating a clay body with colorant is not the most cost efficient way of adding hue to work, it means you are handling toxic materials throughout the entirety of the making process and some colorants are more volatile in the firing when mixed in a clay.

A secondary concern in my research was comparing the strength and volatility of proprietary inclusion stains to raw oxides. I used stains from the company Mason, who like all ceramic stain suppliers, prefires combinations of colorant oxides and grinds them down to powder. This is for 2 reasons: it makes it so that toxic oxides are less dangerous to handle and the expected stain color is more reliable and consistent (the color quality is more homogenous than the varied finish of an oxide). I tested 11 Mason stains and 10 variants of 6 transition metals (Chromium, Manganese, Iron, Copper, Cobalt and Nickel). I mixed up 4 dry ingredient batches of each colorant:

·      10% colorant in porcelain throwing clay body

·      20% colorant in porcelain throwing clay body

·      10% colorant in porcelain sculptural clay body (additions of perlite and mulcoa acting as burnout filler and gritty grog)

·      20% colorant in porcelain sculptural clay body (additions of perlite and mulcoa acting as burnout filler and gritty grog)


As expected, the raw oxides behaved more erratically, a few even fluxed out (lowered the melting point of the clay body) and lost their donut form or oozed down in the multiple color test tiles. There was also no indication of what the final color would be after firing from the color of the pre-fired material. By contrast, every mason stained clay stayed true to form, only dulled or intensified in color a matter of a few degrees and were safe to handle without gloves.

The three most dramatic oxides from my tests were Manganese Dioxide, Copper Carb and Black Copper Carb. I looked to see if I could find any indicator of that prior to testing, thinking that perhaps their molar mass was heavier, but that didn’t line up. Manganese dioxide has a molar mass of 86.9368 g/mol and was unstable but Chromium oxide has a greater molar mass of 151.99 g/mol and was completely stable and Cobalt oxide had a small molar mass of 74.9326 g/mol and was stable too (just a small amount of off-gassing of color to kiln shelves but no distortions to form). So I was unable to place oxides of a spectrum of mass as indicator of volatility.

My scientific testing standard broke down pretty quickly when I combined unmarked scraps of the clay into press mold test tiles, throwing all caution to the wind and painting with what I already knew to volatile clay bodies but this is where I found the results to be most compelling. Arranging the finished tiles from tamest to gnarliest, I discovered that the collection was stronger with this range of results all present. In any given tile, a combination of Mason stained and oxide colored clays was more compelling than just one type on its own. While this feels like a tidy “best of both worlds” kind of conclusion, it did make me reflect on the emotions that color can evoke – this tendency to view color as dangerous and unpredictable – and realize that making that hazardous nature explicit (color being the variable that could completely distort form) was exciting.


[1] An exciting recent discovery in glaze technology is the use of Neodymium oxide as a colorant that gives a subtle lavender color when fired in oxidation. Surprisingly, when viewed in certain artificial lighting, this glaze changes color from lavender to a blue, darker purple or green. Neodymium is a lanthanide metal or a “rare earth oxide” (the color experiments in this project focused on the transition metals most commonly used as colorants in glaze).

Below: Blind color clay test tiles before firing