I’ve always been drawn to pots with clean, elegant lines that can act as canvases, and I enjoy pairing them with glazes that mimic biological processes or the natural world. I spent several years working exclusively with zinc silicate macrocrystalline glazes, and I loved literally growing something in the glaze.
Recently, I spent six months developing a palette of microcrystalline glazes all created with one simple and affordable base recipe. To be honest, before last year, I didn’t know very much about glazes. I knew how to follow a glaze recipe, and had successfully mixed all my own crystalline glazes for years, but I lacked a comprehensive understanding of glaze development and how the different materials in a glaze worked together. If I ran into a glazing problem or my materials changed, I didn’t necessarily know what to do to fix it. I’m a self-taught ceramic artist and always hungry for knowledge, but it felt hard to find good resources to learn more about glaze chemistry. I would often fall down an internet rabbit hole searching for an answer to a basic glaze question. We’ve all been there.
Developing Glaze Knowledge
And then last year, in the midst of COVID-19-related craft show cancellations and with some spare time on my hands, I signed up for the semester-long, comprehensive glaze development course offered by Matt and Rose Katz of Ceramic Materials Workshop. When the class began, I selected a rutile glaze to experiment with because I love titanium (the active ingredient in rutile) and because it was incredibly simple—only five ingredients. Over the course of the semester, I mixed variation after variation of this five-ingredient rutile glaze (1), swapping out one flux for another, adding different colorants, or changing the silica and alumina levels. One lab instructed us to remove most of the silica from the recipe, and when I pulled the low-silica test tile out of the kiln, the glaze had formed very tiny, pink, glittery crystals. I was hooked. (I should mention here that I fired all of these tests to cone 10 in a crystalline cycle—a long crystalline hold at about cone 04 is what gave the microcrystals time to grow in the glaze matrix.)
Next, Matt instructed me to test systemic and very small variations in the silica and alumina content of the glaze (2). The first crystallized glaze was extremely matte, even dry, but I found that as I moved to the right on the Stull chart (3) (an X-Y-axis grid that shows the fired results of oxide mixtures like silica and alumina) and added some of the silica back in, I crossed the matte/gloss line and ended up with a glossy glaze that still contained microcrystals. I also discovered that as I added silica back into the formula, I needed an additional ingredient to flux the glaze and cause the crystals to separate a little bit, otherwise the crystals grew too large and became one solid mass. I found that most colorant oxides worked well for this purpose (with the exception of chrome, which is quite refractory). I have been using cobalt, manganese, and iron in small amounts (1–2%) with nice results (3–5). I also make a version with additions of both 1% cobalt carbonate and 1% copper carbonate that fires to a nice deep turquoise color. I found that adding textured slip gives the crystals a place to cluster (5).
I was very pleased with myself at this point—I had a lovely little palette of microcrystalline glazes that were incredibly simple and inexpensive to make. But when I mixed up a larger batch of the glaze (the glaze performs best at a specific gravity of 1.45), applied it to some mugs, and fired them, it crazed slightly on the porcelain I was using, which hadn’t been obvious on the test tiles. In the glaze class, we learned how crazing can decrease the durability and longevity of functional ware, and so I really wanted to create a version of this glaze that didn’t craze. I spent the next three months running dozens of glaze tests, but when I made the recipe changes necessary to get the glaze to stop crazing, I also lost the microcrystalline effect. During the same period, I also tested five additional clay bodies: three porcelains and two white/off-white stonewares. I found that the glaze fit the stoneware bodies but tended to become matte, and it crazed on the porcelain bodies. I almost gave up on a compatible porcelain clay body, but I finally got lucky with Coleman Porcelain from Aardvark. The glaze seems to fit this body fairly well (if I apply the glaze thickly, I’ll occasionally find a small craze line) and the higher flux level in the body means the glaze is particularly glossy and beautiful.
Expanding the Color Palette
Once I had figured out the crazing issue, I wanted to have more color options than traditional colorants like cobalt, copper, and iron provided. First I tried incorporating several different stains into the glaze, but they tended to ruin the microcrystalline effect. I had heard about some ceramic artists incorporating rare-earth elements (specifically neodymium, erbium (6), and praseodymium) into their glazes to achieve bright colors, and when the stains didn’t work, I decided to test these elements. The rare-earth metals are expensive, but US Pigment Corporation offers good prices and small quantities, so I ordered ¼ pound of neodymium to start with. It took a few tests to dial in the glaze, but I’ve found a really lovely, subtle purple color with nice crystallization results from the addition of 5% neodymium to the base glaze. (The 5% neodymium does seem to increase the crazing a little bit.) When adding rare-earth colorants, I also substitute pure titanium dioxide for the rutile—for some reason the iron in the rutile doesn’t play well with the neodymium and the glaze over-crystallizes. In the future I’m hoping to test the other two rare-earth elements (to follow along with the tests, find me on Instagram), but the colorants I’ve mentioned so far provide a variety of color options that pair well with each other.
*Note: The peak temperature will depend on your kiln and specific firing conditions. Test to determine whether you need to fire a little hotter or cooler to get the desired results.