Zinc oxide, the essential component of crystalline glazes is the source of one particular type of problem that can drive the potter to distraction. This is the familiar situation: a crystalline glaze recipe is mixed and enough water added to achieve a brushing consistency, the mixture goes on tolerably well but within minutes the thick glaze layer has cracked. Very likely some of the glaze will drop off in the firing, ruining the piece and damaging the kiln shelf. The problem can be mitigated to some extent using CMC gum but not eliminated.
Some batches of zinc oxide do not cause this problem. It is well known that heating zinc oxide first to around 800ºC (calcining) can prevent it from happening. This is strange as zinc oxide is unaffected chemically at this temperature, nor are its physical properties greatly changed: there is only a little sintering to a larger particle size at 800ºC
So what is going on? The answer seems to be that the zinc oxide is reacting with air and carbon dioxide to some extent creating a very small amount of basic zinc carbonate. This is not present in sufficient amounts to appreciably affect the fired composition of the glaze (although, this may not always be true) but it acts as a potent flocculant to the suspension of glaze in water. This means that more water is required to give the glaze consistency suitable for application. On drying, the glaze, lacking the tensile strength that would be imparted by a significant clay content, inevitable cracks with results that are all too familiar.
Calcining works by eliminating the basic zinc carbonate and converting it back to zinc oxide. There is an alternative. Simply counter with a powerful deflocculant such as dispex or darvan. The results have to be seen to be believed. A crystalline glaze too thick to pour can be transformed to low viscosity with a few drops. A recipe without deflocculant might be too thick to brush even at a density as low as 1.6 g/ml whereas the same with dispex can be freely brushed with a density approaching 2.0 g/ml. The resulting laydown is very dense and cracking is eliminated. Even with this, it is sometimes still desirable to use some CMC gum to slow down the rate of water absorption into the body and to harden the dried glaze layer.
The required amount of dispex varies with each recipe. Typically 4g per 1000g of dry material is enough. Dispex is very forgiving of overdosing so the amount is not critical. Thus prepared, some glazes will remain stable for months but others gradually thicken in a way that cannot be countered with more dispex, more water is the only way to get the viscosity back down and that leads to cracking. The thickening seems to be due to certain materials slowly dissolving in the glaze. Ferro 3134 frit and many other calcium borate frits produce the effect (Ferro 3195 is better than 3134 in this respect but it contains more alumina which further limits the scope for adding kaolin to the recipe). In such cases it is best to use the glaze within a few weeks of mixing.
Reaction rate
The reaction between zinc oxide and atmospheric carbon dioxide requires the presence of water. O’Connor [1] found that the reaction proceeds with appreciable rate only when the air is close to being saturated with water vapour. Thus, if zinc oxide is stored in a dry environment, the formation of the basic carbonate should be very limited.
My own experience is that only a few hours of exposure to damp air can cause problems with zinc oxide. This might explain why sucessive batches from the same supplier can behave differently, any significant exposure to air during handling could result in some reaction. My advice it to keep zinc oxide in a air-tight container and resort to calcining only if a batch resists deflocculation.
References
[1] M.F. O’Connor; ‘A study of the kinetics of the basic carbonate formation reaction’; Institut für Anorganische Chemie der Universität München; 1975.