A ceramist wants to arrive as an artist, a ceramic glaze wants to arrive at the right glaze temperature. Greg Daly calls this “arriving at temperature” in his glaze travels. Who dares to say that there is no poetry in ceramic glazes?
Daly provides methods to investigate at which temperature glazes melt, but no (theoretical) tools to predict this for a given glaze. However, you do need this if you want to design a glaze with a predetermined temperature. If you don’t know which corner to look in, it is also hard to get there.
Glaze temperature is a mechanism, as there are several in glaze technology. To be able to decipher this mechanism you have to look from the right angle (more about this in this blog).
For glaze temperature this is the oxide level. If we dissect a glaze into oxides (the different molecules in a glaze) it is possible to discover trends.
I am still researching how to arrive as an artist, but I have guidelines to arrive at the right melting temperature for any glaze. In this blog I explore the three main ratios:
- SiO2 -Al2O3
- RO2 -RO
- B2O3
Glaze is glass
Before I start with the guidelines for the firing temperature first a short summary of what a glaze is.
A glaze is a glass that does not run from the pot. More accurately said, a glaze is an aluminum silicate. A silicate (like glass), but with alumina (which isn’t in glass). The silica (quartz or SiO2 in chemistry) is the glass former. The addition of alumina ensures a high viscosity, (syrupy) so that it does not run from a vertical surface.
Such a silica-alumina glaze is fine, but only melts at a very high temperature*, much higher than the clay on which it is applied can be fired (also much higher than my ceramic kiln can handle). That’s why every glaze has a third type of ingredient: a melting agent or flux.
By adding a flux, the temperature can be lowered for:
- Porcelain (1250-1320 oC)
- Stoneware (1200-1280 oC)
- “Cone 6” middle temperature (1180-1220 oC)
- Earthenware (1000-1080 oC)
- Raku (900-1000 oC)
What is glaze temperature?
“The” firing temperature of a glaze does not exist, a glaze has a melting range based on kiln temperature (read on a pyrometer) and time. This is called “heat work“.
To measure temperature and time ceramists use cones. A cone that slowly bends and melts at a certain time and temperature (just like a glaze). For example, a glaze is fired at cone 10 (see also this blog).
In the last 100 to 150 degrees C at top temperature of the firing, the raw materials become softer, sinter and finally melt. If you fire too high (or too long) the glaze will drip or run. A little drip is not bad (even pretty), but when it flows from your pot on the kiln shelves it becomes a problem…
If the glaze is not completely melted it is rough and stony matte. If you fire higher, the glaze will become glossy. Such a (not fired at “maturity”) matte glaze is not suitable for functional pottery. It is not strong and will, for example, wear out in the dishwasher.
You also have “real” matte glazes (glazes with a lot of alumina compared to silica). These remain matte even if you fire them higher. Firing them too high they will run (and stay matte) just like glossy glazes.
Predict glaze temperature
A lot of research has been done to predict the melting temperature of a glaze. One of the first calculation methods was proposed by Martin Lengersdorff. He published in 1964 “Practical Berechnung von Massen und Glasuren“. In this book he described a method to use the “melting agent factor” (“flussmittelfactor F”) to estimate the melting temperature of a given glaze.
His approach has been followed by among others Linus Pauling‘s “Ionic potential” **. And more recently in a glaze calculation program “Glasurenspiel” by Gustav Weiss. In this program a different approach was chosen. Based on the calculation of the viscosity curve (Vogel-Fulcher-Tammann equation), a temperature “tendency” was displayed.
With the development of computers it has become easier to make these complex calculations. For example, Lawrence Ewing (former ceramics/glaze technology professor at New Zealand’s School of Art) has experimentally added both the Lengersdorff (Flux Factor) and Pauling (Ionic Potential) methods in his glaze calculation software Matrix.
Doesn’t work in practice
I have also used these methods and made the calculations in Excel for my own glazes. I spent many hours behind my pc to understand the calculation, enter data and compare it with the actual firing results. Although nice to make this kind of calculations (you understand I don’t have a social life anymore), it didn’t help me on a practical level.
The various calculations indicated, for example, that the melting temperature would be between 1200 and 1300 oC. Usually this was the case for “regular” glazes. But I didn’t found that of much use. I used to fire at 1220 oC, so I still didn’t know if the glaze would melt at “my specific firing temperature”. Also the calculation methods didn’t account for the less used oxides. So for the more “exotic glazes” the calculations were even less accurate.
In short, due to the complexity of the firing process, there is no unequivocal way to predict the melting temperature of a glaze. Although I find the theoretical approach very interesting, on a practical level they are not (sufficiently) reliable and certainly not accurate enough to really benefit from it.
Glaze ratio UMF
But how then can you estimate a glaze temperature? By looking at specific ratios of a glaze. But before we go there, first we have to calculate the UMF (Unity Molecular Formula).
That’s also the first ratio we are looking at. The glaze recipe is divided into three groups that I have already mentioned:
- Flux (approx. 10 oxides, sometimes called base or RO/R2O)
- Stabilizer (Al2O3 = alumina or aluminium oxide, sometimes neutral or R2O3)
- Glass former (SiO2= quartz, sometimes called acidic or RO2 )
In the UMF the total of the fluxes is always calculated as one. The stabilizer (Al2O3) and the glass former (SiO2) is thus proportional to the fluxes.
If you have calculated the UMF (glazy.org) of a specific glaze, the first thing you can look at is the amount of quartz (SiO2) and alumina (Al2O3) in that glaze. Simply put, the more SiO2 and Al2O3 in the glaze, the higher the firing temperature will be.
The UMF is a wonderful ratio, but how should you interpret it?
Ratio SiO2 -Al2O3
In 1914, Stull and Howatt*** created a chart that plotted the amount of Al2O3 (vertical axis) and SiO2 (horizontal axis) on a graph. The area to the above left “Unfused” and under to the right “Devitrified” indicate where the glaze will not melt. This gives a good indication of how high (or low) both quantities should be.
But how high is high and at what temperature does the glaze melt? You can’t exactly say… But no worries there are indications!
The first one I found came from Gustav Weiss. In his book “Freude an Keramik” 1972 (Reprint Neue KERAMIK Workshop IV, 1998). He made a table that allows you to estimate the temperature. For a given quantity Al2O3 (in the table R2O3) and SiO2 (in the table RO2) you can read a temperature.
| Temp C | Temp F | RO+R2O | R2O3 | RO2 |
|---|---|---|---|---|
| 800 | 1472 | 1 | 0,060 – 0,225 | 1,000 – 2,100 |
| 825 | 1517 | 1 | 0,060 – 0,275 | 1.000 – 2,150 |
| 850 | 1562 | 1 | 0,075 – 0,300 | 1,025 – 2,200 |
| 875 | 1607 | 1 | 0,080 – 0,300 | 1,050 – 2,275 |
| 900 | 1652 | 1 | 0,085 – 0,325 | 1,060 – 2,350 |
| 925 | 1697 | 1 | 0,090 – 0,340 | 1,075 – 2,450 |
| 950 | 1742 | 1 | 0,095 – 0,350 | 1,100 – 2,550 |
| 975 | 1787 | 1 | 0,100 – 0,375 | 1,150 – 2,700 |
| 1000 | 1832 | 1 | 0,100 – 0,390 | 1,200 – 2,825 |
| 1025 | 1877 | 1 | 0,100 – 0,410 | 1,300 – 3,000 |
| 1050 | 1922 | 1 | 0,100 – 0,450 | 1,375 – 3,150 |
| 1075 | 1967 | 1 | 0,120 – 0,475 | 1,500 – 3,500 |
| 1100 | 2012 | 1 | 0,150 – 0,500 | 1,600 – 3,500 |
| 1125 | 2057 | 1 | 0,175 – 0,525 | 1,750 – 3,750 |
| 1150 | 2102 | 1 | 0,200 – 0,550 | 1,950 – 4,000 |
| 1175 | 2147 | 1 | 0,250 – 0,600 | 2,150 – 4,350 |
| 1200 | 2192 | 1 | 0,275 – 0,650 | 2,400 – 4,700 |
| 1225 | 2237 | 1 | 0,325 – 0,700 | 2,600 – 5,150 |
| 1250 | 2282 | 1 | 0,375 – 0,750 | 3,000 – 5,750 |
| 1275 | 2327 | 1 | 0,450 – 0,825 | 3,500 – 6,400 |
| 1300 | 2372 | 1 | 0,500 – 0,900 | 4,000 – 7,200 |
| 1325 | 2417 | 1 | 0,575 – 0,975 | 4,700 – 8,200 |
| 1350 | 2462 | 1 | 0,650 – 1,050 | 5,400 – 9,200 |
| 1375 | 2507 | 1 | 0,725 – 1,150 | 6,250 – 10,200 |
| 1400 | 2552 | 1 | 0,800 – 1,250 | 7,200 – 11,300 |
| Gustav Weiss | (Neue Keramik Workshop IV, 1998, p.181) | |||
It is not very precise (and there are exceptions), but you have an idea in which corner you should look when designing a glaze. In his calculation program Matrix , Ewing makes a similar estimation ****. With a given glaze (the small yellow square) and a firing temperature the graph shows in what region the quantity Al2O3 and SiO2 should be.
Still not very precise. But that is to be expected because other ratios also have there influences.
Ratio RO2 – RO
For the next ratio we look at the flux ratio; Also referred to as R2O/RO. This group can be divided into two: the alkali (R2O) and the alkaline earth (RO). Why make this distinction? Because there is a big difference in both types of fluxes.
- The alkali: Lithium, Sodium and Potassium (Li2O, Na2O and K2O) are strong fluxes at low temperatures.
- The alkaline earths: Magnesium, Calcium, Strontium and Barium (MgO, CaO, SrO and BaO) are relatively weaker at lower temperatures, but do their work at higher temperature.
By looking at the ratio of these two groups you can estimate the melting temperature. For example, if the ratio R2O (total alkali) is 0.8 and RO (total alkaline earth) 0.2 (the total of the melting agents is always 1 in the UMF) the glaze will melt at a low temperature.
The French brother and ceramist Daniel de Montmollin has beautifully mapped the influence of the R2O-RO ratio at the melting temperature in his book “The Practice of stoneware glazes“. In 60 graphs he shows how the melting area is enlarged by increasing the proportion R2O. Note that the axis for SiO2 and Al2O3 have been reversed! (the French always do things just a little different :-))
Can you use this ratio? No, at least not for glazes that are used for functional pottery!
Flux ratio should be 0.3 : 0.7
Rose and Matt Katz of Ceramic Materials Workshop have done a lot of research on the relationship of this ratio and the resistance of a glaze. If the proportion of alkali is too high, the glaze wil not be strong. Acids (e.g. lemon juice) or bases (dishwasher) break down the glaze.
Therefore a strong glaze (for functional ware) should have a ratio near 0.3 : 0.7 (R2O : RO). Higher (or lower) makes a glaze less resistant. So varying this ratio affects the melting temperature of a glaze, but you shouldn’t use that on functional pottery.
Ratio B2O3
Okay, so if R2O-RO ratio is a given, how can you make a strong glaze at a lower temperature? By using a special flux: boron oxide (B2O3).
Boron oxide is not really a flux and is therefore not grouped under the R2O/RO group in the UMF. Boron oxide is a glass former, like silica. However, it has a (relatively) very low melting temperature (510 oC). As a result, it functions in a glaze as a flux (without affecting the resistance).
At 1280-1300 oC a glaze can melt without boron oxide. But if you want to make a glaze at a lower temperature, for example an earthenware glaze at 1080 oC, then you need boron oxide. Probably about 0.2 up to 0.4 B2O3. If you want to know exactly then you will have to do tests.
Matt Katz also has a lot of research into the role of B2O3 in glazes. On his website you can download an Excel calculator in which you can find a graph with B2O3 need for a given temperature.
Other ratios
Are these all the methods to make a glaze arrive at temperature? I would like to say yes, but nature is endlessly more beautiful. There are other ways to make a well balanced glaze at lower temperatures without boron oxide.
Such as:
- Zinc oxide (ZnO) at specific quantity
- Lead oxide (PbO)
To start with lead oxide, this is a very poisonous material. Even in fritted form you have to beware of this, so I can’t recommend it. But….. you can make nice glazes with lead at low temperatures. So I had to mention it…
When the dangers of lead glazes became evident at the end of the 19th century ,ceramists were looking for alternatives. Probably in Bristol (UK) they discovered that with a specific amount of zinc oxide in a glaze, the temperature could be lowered to approx. 1200 oC. This type of glaze is also known as a “Bristol glaze”.
Other factors…
Nature is complicated, and glazes more so. I hope that with this schematic overview I have made some insights clear.
There are more factors that determine the temperature of a glaze, such as the fineness (“mesh size”) of the raw materials, whether or not they are “already melted” (like frits) and the various fluxes. And I am sure I forgot a few more.
Looking at existing (historical) recipes and specific “target” formulas (from a reliable source) can guide you in to right direction. But no matter how well you understand the theory, you will eventually be able to find the melting range of the glaze through testing.
Theory guides, practice proofs…
Learned something from this blog? Give me a cup of coffee so I can write the next with new energy!
Footnotes:
- 1595 oC, Source: Phase Diagrams for ceramists, 1964, p. 123 Figure 314.
** Ionic Potential is calculated according to the valency divided by the radius of relevant atoms in the enamel. The method was published in Cardew’s “Pioneer Pottery” (original edition, p. 302) in 1969.
*** Original publication: Stull, R.T. and Howatt, W.L., “Deformation Temperatures of Some Porcelain glazes”, Trans. Am. Cer. Soc. XVI, 454 (1914). Also in C.W. Parmelee “Ceramic glazes”, 1948, (p. 156) and D. Green “A Handbook of Pottery glazes”, 1979 (p. 52).
**** These are originally from 1989 Monash University, Melbourne, Australia.
About Post Author
So looking at Montmollins fuse diagrams being available in glazy don’t you think we have some pretty good tools to estimate glaze temperature?
His axes are reversed because they follow phase diagrams typical setup. So not so french 😀
Hi Lauge,
Yes, Derek Au did a marvelous job converting the original fuse diagrams in Montmollin book to the conventual axes 🙂
And yes these fuse diagrams show how to manipulate glaze temperature by just changing the flux ratios. They are a great tool to predict glaze temperature. But not so precise that you can just make the glaze. I think testing is still necessary… but what the future holds I don’t know, maybe with a little help from an AI (but what’s the fun in that 😉
Regards,
Daniel