Digitalfire Ceramics Technical Articles
|Logged in as Level 2 access: Logout|
Reducing the Firing Temperature of a Glaze From Cone 10 to 6
Section: Glazes, Subsection: Adjustment, Adaptation
Moving a cone 10 high temperature glaze down to cone 5-6 can require major surgery on the recipe or the transplantation of the color and surface mechanisms into a similar cone 6 base glaze.
It is amazing how many potters or small scale manufacturers will push their clay bodies or kilns (and utility bills) to the temperature limit to get a special reactive glaze to melt the way they want. No industrial manufacturer could afford or would have the desire to do this, it only makes sense to fire to the lowest possible temperature and as fast as possible (typical industrial firings last as little as one hour these days). It takes much more energy, for example, in a typical poorly insulated kiln (any inexpensive kiln is poorly insulated) to increase the temperature by 1 degree at 1000C than it does to do the same at 500C. Obviously, lower temperatures also mean less wear and tear on kilns, especially electric. Not so obviously, it can be easier to keep the temperature even throughout the kiln at lower temperatures and thus either easier to soak at the end or shut it off in (with an even-temperature condition throughout the kiln chamber at shut-off time). In addition, at lower temperatures kilns have an easier time holding to a programmed schedule (and thus maintaining consistency firing-after-firing is easier).
At the same time it does not make sense to fire too low either since quality and strength issues come into play below about cone 1 using traditional bodies and glazes. It is important to realize that high temperature firing is not a requirement for strength and quality ware. Also, the range of achievable colors and color consistency is better at lower temperatures. Lower fired bodies can also be very dense and strong. Dental porcelain, albeit an extreme example, completely melts into a white pool of glass far lower than any typical functional porcelain would even begin to densify. Admittedly the mullite-glass-silica microstructure characteristic of high temperature porcelains is not as highly developed in middle-fire bodies, but other strength-imparting crystalline matrixes can be employed. But even with the simple glass-bonding-of-aggregates mechanism you need make no apologies for cone 4-6 porcelain strength (when I say 'cone 6' below I mean cone 4-6, or, middle temperatures).
The question is: Can you juggle the amounts (e.g. increase the fluxes) of ingredients in a typical cone 10 glaze recipe to make it melt at cone 6? The answer is almost always no. Middle and high fire glazes employ different fluxing oxides and they interact differently. To understand, let's fire lower for a moment. Low fire glazes (cone 06-02) are fundamentally different in a one principal way, they contain lots of boron, it is the key and principle flux. Cone 10 glazes contain none. Additionally, oxides that behave as fluxes at cone 10, like MgO and CaO, act as refractory matting agents at low fire! Cone 5-6 is a bit of an in-between land in this sense. Some boron is almost always needed because common cone 10 fluxes like feldspars, calcium carbonate, talc and dolomite do not melt well at all at cone 5. Thus, increasing the content of existing cone 10 fluxes to melt at cone 6 will not work, you must introduce new more powerful (better melting) ones and learn a new set of dynamics, interactions and tradeoffs. There are some special purpose middle-temperature glazes that will melt without boron (e.g. high zinc fast fire or bristol glazes) but these often require special expertise and lots of tolerance, I will not discuss them.
Cone 10 raw glazes contrast from those made only with frits. They contain a number of individual materials, most of which in turn contain a number of different components (or minerals). Thus, microscopically, raw glaze powders are mixtures of particles having very different 'individual' melt behaviors up to, and at the target temperature of the firing. Some have already, or are about to melt suddenly and flow while other more stable ones are caught somewhere along a softening process that spans range of temperatures. Some change crystal form. Some do nothing. However, their behavior together, in the 'orchestra' of firing, involves interactions that produce the 'music' of these glazes. The active ones get a melt started, that melt takes others into solution and changes chemistry and behavior with temperature rise. Each different particle type finds itself engulfed in an expanding melt and influences that melt differently. Does it sound like this should be easy to calculate? The inventory of these is quite interesting (we won't look at colorants, opacifiers, variegators):
The bottom line is that we cannot just reorganize a cone 10 recipe to melt at cone 6. We must add something new, a flux or fluxes not normally found in cone 10 glazes. We must also reduce the proportion of silica and alumina in the glaze, since even with added fluxes it is not possible to get a glaze to dissolve the amounts of high melting alumina and silica typically found in cone 10 glazes. I am not saying it is impossible, you might be the magician who has found a way. But remember, the challenge here is to adjust an existing cone 10 recipe to melt at cone 5- 6 and not craze or leach or scratch easily and still have good application and working properties. That is not easy.
The most obvious solution is to add powerful fluxes like zinc oxide or lithium carbonate (not commonly used, or needed, at cone 10). They melt very early and vigorously and can impart significant melting effects in small amounts in some circumstances. The only problem is that this is not one of those circumstances. A little zinc is not going to dissolve a lot of refractory particles. Zinc and lithium are like people "with personality issues", they participate at a cost if over employed. A lot of zinc and/or lithium to make the glaze melt at cone 5-6 will create a whole new animal and a whole new set of problems (especially with regard to color response, surface character, tendency to devitrify (crystallize), crawl, bubble, etc.)
I want to point out that when I say we need to 'add something new', I did not mean a material. I meant an oxide. Materials are composed of oxides, and most materials contribute more than one oxide (sometimes 6 or 8 different ones, e.g. feldspars). When you talk about a fired glaze and the way it melts and solidifies, you are talking mainly about chemistry (oxides), whether you like it or not. If we add or remove feldspar, for example, we are making a fundamental change in the balance of oxides in the glaze (because it contributes so many oxides), that does not jibe with the objective of making changes that will give the greatest reduction in melting behavior accompanied by the least change in overall glaze appearance and fired properties (e.g. color response, thermal expansion, surface character, working properties).
There are two main approaches we can take.
Method 1: Transplant Mechanisms
At the risk of negating most of what I have just said, identify the "mechanisms" in the glaze and transplant them into a good base glaze for cone 6 that has similar surface texture and degree of melting (compare with a flow tester if needed). Answer these questions:
There are other factors that are not as easy to transplant. These often relate to visual effects in glazes that are very fluid or runny. For example, glazes that form small or large crystals or a crystal mesh across sections of the surface during cooling need to have a very fluid melt. Often matte cone 10 glazes are actually over-fired cone 6 or 8 glossy glazes that have completely crystallized. For this type of glaze you will need to have an equally fluid melt in the cone 6 range to transplant things into (that means it is going to need to melt well at cone 2-3). This will not be easy since getting a glaze to melt that low means removing much of the low expansion silica and alumina, that in turn leads to crazing.
Another mechanism that is not easy to transplant is the character of the surface. Dolomite matte glazes, for example, have a pleasant silky surface that is a product of what CaO and MgO do around cone 10 in slow firing in the right host glaze (by slow firing I mean over a period of hours rather than minutes as is common in industry). Obviously you cannot duplicate this exact mechanism because MgO does not suddenly melt at cone 6 the way it does at cone 10. You must create or find a base that already has the surface you want (or adjust one that is close since matteness is a product of the glazes chemistry, that is, the mix of oxides it contains). Fortunately, at cone 6 various types of matte surfaces are also possible. Additionally, you can actually use a mechanical technical to duplicate the dolomite matte effect: Apply a thin layer of matte glaze over a glossy under layer, the upper layer can variegate in a similar manner.
Method 2: Add the Magic Oxide
The magic oxide is boron (B2O3). The reason boron is such an ideal additive is that it is more than a flux, it is a glass like silica, yet it melts low. This is a marvelous oxide! One of the beauties of boron is that it is compatible with most colorant, matte and variegation mechanisms. But there is another even more important reason: Boron has a low thermal expansion; additions of it will reduce glaze crazing. This is so important because most high temperature glazes contain far more high-expansion Na2O, K2O than can be tolerated at cone 6 without producing crazing. In addition the low-expansion silica and alumina they contain must be reduced to achieve melting at cone 6, thus moving the glaze toward crazing.
The next questions is: What source of boron should you use? The answer is: frits. There is a boron frit out there to meet almost any fluxing challenge. However each frit contributes other oxides besides boron. This means that to reach the goal of only increasing the boron in a glaze you need to be able to reduce other materials in the recipe that contribute the same non-boron oxides as the frit. Thus, to do what I am about to describe, you need a ceramic chemistry software program. The kinds of material juggling I am about to describe are dealt with in the videos on digitalfire.com.
The first thing to consider is this: Does the glaze need clay (does it settle in the bucket or lack dry hardness?) or does it have too much clay already (does it shrink and crack on drying?).
The second last question is: What level should the Al2O3 and SiO2 be at in the newly adjusted glaze? An obvious suggestion might be to adjust them to about 2/3 of what they were. However, a rule of thumb is to always have as much alumina and silica as you can in a glaze. But, since B2O3 (boron) performs the same function as silica you can afford to have lower SiO2 in the interests of good melting. Also, boron has very good alumina-dissolving abilities, thus it is often possible to leave Al2O3 at the same levels as it was in the cone 10 glaze. Boron is really great stuff!
The last question is: How much boron do I add? I would start with about 0.3 molar equivalents. If the glaze is reactive and has to be very runny to achieve the desired effect (like the ones in the book "Mastering Glazes"), you might have to take the boron as high as 0.8.
Some Other Points
If the cone 10 glaze is a stoney matte and is not melting enough, then to duplicate it faithfully I guess you have to make a cone 6 glaze that is not really melting enough either. However this is certainly not advisable for functional ware. You will want to fire tests to compare the degree if melting (consider using a melt flow tester). Be sure to try the glaze in thick and thin layers and watch for crazing and leaching. Different minerals release their oxides to the melt at different temperatures, even though they might have similar chemistries (for example wollastonite and whiting both contribute CaO yet the latter requires a higher temperature to release it to the melt).
Following are some glazes common to cone 10. Bear in mind that matte effects can be achieved by different mechanisms (e.g. Crystallization, incomplete melting, phase separation, prevention of surface levelout by rivulet formation or melt stiffening due to the presence of refractory or non-dissolving components or oxides like Al2O3 or zircon). Some of these mechanisms are a product of slow firing and thus not available in industry. Thus you need to select a matte of similar type to transfer properly, however bear in mind that colorants can affect matteness depending on the fragility of the effect (if you have used matte glazes for a long time you will know what I mean by this). One effect of cone 10 that can be difficult to duplicate is the dolomite or magnesia matte (the phase separation produces the micro rivulet surface of melting pockets of MgO) with its very pleasant silky surface.
Out Bound Links
In Bound Links
Copyright 2003, 2008 http://digitalfire.com, All Rights Reserved
INSIGHT is ceramic chemistry