Lithium carbonate is now ultra-expensive. Yet the reactive glaze on the left needs it. Spodumene has a high enough Li2O concentration to be a possible source here. It also has a complex chemistry, but the other oxides it contains are those common to glazes anyway. I did recipe rescue calculations and got a pretty good match in the formulas (lower section in the green boxes). Then I made 10-gram balls and did a GLFL test at 2200F.
Not surprisingly, this recipe is very runny; that's why the tiny yellow crystals grow during cooling. The new version fires very similarly, perhaps better. My calculated cost to mix these in 2022 was $17.84/kg vs. $10.40/kg. In 2026, the difference is even greater! But there is a practical cost: Poor slurry properties. The spodumene sources so much Al2O3 that 70% Alberta Slip had to be dropped to accommodate it! How does one use this type of glaze without ruining kiln shelves? Using a catcher glaze is one answer.

This picture has its own page with more detail, click here to see it.
Almost all ceramic glazes are a base recipe with additions of colors, opacifiers, variegators, etc. Our traditional G3933 oatmeal glaze is a good example (recipe on the left). It can produce rich brown silky matte surfaces, especially on dark burning bodies. But problems have emerged, especially a tendency to crawl. Much testing has yet to reveal the reason. Would it be possible to base the recipe on Ravenscrag Slip and achieve the same chemistry? Yes. And some unexpected benefits accrue. In the recipe on the right, I sourced MgO (the key to the matte surface) from dolomite and Ferro Frit frit 3249 (earlier tests sourcing from talc were unsuccessful, off-gassing from the talc was puffing up the glaze with micro-bubbles). This G3933E recipe has the same chemistry (I derived it in my account at insight-live.com). It is not likely to be without problems, but it looks identical (with richer color from a little more iron oxide), it does not crawl, and it's recipe and chemistry are flexible. It is glossy when cooled fast and silky matte when cooled slowly. The mix of calcine and raw Ravenscrag Slip also enable control over the slurry and application properties. I later tuned this recipe a little more and settled on G3933EF.

This picture has its own page with more detail, click here to see it.
Of course, if a recipe only calls for 1-5% lithium carbonate either of these might be candidates to supply the Li2O. However, Petalite is eight times less and Spodumene five times less concentrated than lithium carbonate so to make either worthwhile the prices would need to be eight and five times cheaper. But if a recipe calls for more there is another problem: Petalite is extremely high in silica, which means supplying the needed Li2O from it is almost certainly going to oversupply SiO2. Spodumene will likely do the same. Both are also high in Al2O3 and likely to oversupply that (or at minimum supply the bulk preventing the presence of kaolin in the recipe).

This picture has its own page with more detail, click here to see it.
The crystals that form as this cone 6 fluid-melt glaze solidifies are one of the minerals in the recipe: Spodumene. Actually β-spodumene. But it is so expensive, and there is 42% here! What would it take to reduce the cost of making this glaze ($52/kg in 2026)? Another flux could certainly make it just as melt-fluid, but would not grow spodumene crystals.
β-spodumene is Li2O - Al2O3 - 4SiO2 So the 0.22 Li2O in this glaze would require 0.22 Al2O3 and 0.88 SiO2 to crystallize as a β-spodumene-type lithium aluminosilicate. The glaze contains 0.24 Al2O3 and 1.41 SiO2. There is almost exactly enough lithium to combine with nearly all of the alumina as β-spodumene. The 2% rutile supplies TiO2; it acts as a nucleating agent. Replacing Li2O with another flux could maintain the melt fluidity, but Na2O, K2O or B2O3 cannot simply take the place of Li2O in β-spodumene.
It is possible that less spodumene could still do this. That could be tested by reducing Li2O to 0.1 and increasing Na2O to 0.23 (while maintaining the other oxides as-is). B2O3 could also be tried (it has a far lower thermal expansion). Of course, effecting changes in specific oxides requires some glaze chemistry (e.g. in an Insight-live.com account).

This picture has its own page with more detail, click here to see it.
Spodumene is super expensive; one dunk in a dipping glaze can now cost dollars. But lithium carbonate is even more so. However, both are used as a source of Li2O. The latter is 6+ times more concentrated. And spodumene is more troublesome in glazes (issues with settling, dusting, slurry properties, consistency). And it can be hard to get. So using the pure lithium material might be a better choice for you.
This side-by-side calculation seems to indicate that 15 Lithium Carbonate, 55 Kaolin and 30 Silica can substitute for 100 parts by weight of Laguna Spodumene Substitute. However, that is not quite correct. Note that my substitute recipe calculates to an LOI of 16.3. If I assume the Laguna material has no LOI (for convenience, it actually has 0.2%), that means each of the amounts need to be divided by (100-16.3)/100=0.84. Thus, to substitute for 100 parts of spodumene, you need ~120 of this mix (18 lithium carbonate, 66 kaolin and 36 silica).
| Typecodes |
Glaze Chemistry
Case studies where glaze chemistry was used to solve a problem. |
| Materials |
Spodumene
Spodumene is a lithium sourcing feldspar, an alternative to lithium carbonate to supply Li2O to ceramic glazes. Contains up to about 8% Li2O. |
| Materials |
Lithium Carbonate
A powerful melter very valuable in ceramic glazes. It is 40% Li2O and has an LOI of 60% (lost as CO2 on firing). This material in now incredibly expensive. |
![]() PayPal | No tracking, No ads, No paywall, No transient content! Just organized, concise information constantly updated and improved. Was this helpful? Consider supporting me. |
Buy me a coffee and we can talk