1. Introduction

Having been a professional potter for thirteen years or so, I can think of many occasions when my good humor, and perhaps even my sanity, smashed up against a glaze problem that I could not quickly solve. One instance several years ago occurred in late summer as I was rushing to fill pre-Christmas orders for shops and galleries while also getting ready for several shows. That time of year is our busiest, so it was with a little shock, and a mildly queasy feeling in the pit of my stomach, that I opened a kiln-load of pottery to find my two main glazes covered with pits, pinholes, and craters. These glazes were not only unattractive and unsellable, they were completely unsafe to use. I tried re-firing a few pieces, but that only smoothed the smallest of the defects. While a few pots were salvageable, most of the firing went under the hammer. I did a little reading in my glaze texts, made a change or two, and fired another load of pottery confident that I had remedied the problem. The same thing happened, and another load of pottery fell under the hammer. Orders stopped going out, and I had to place some embarrassing calls to gallery owners explaining the problem. I did a little more research, made some new changes, fired another load. Again, most of the kiln load was no good. This went on for weeks. My mood darkened, my confidence slumped, and I began to seriously think of giving up pottery as a career. I finally discovered the source of my glaze problem after the better part of eight loads of pottery were ruined.; The culprit was a new material I had started using a couple of months before. Once that material was removed from new glazes batches, things got back to normal, and I was able to complete orders and prepare for shows. The episode left a bit of a mark on my psyche, though

During this period I delved into glaze chemistry like never before, desperate to find a solution for my problem. Even after the immediate issue was resolved, I realized it was only a matter of time. before something like it happened again. My partial ignorance of the processes from which I made my living left me completely exposed to such occurrences. My livelihood was sitting on a house of cards. So I resolved to get better, to learn more This became true of every aspect of my business: the clay, the glazes, the processes, the studio, and the business itself. This period is the point at which I changed from a slacker potter to a professional.

A few years later, I began teaching a glaze testing class at the local community college, which provided me the opportunity to organize all of the glaze knowledge I had accumulated along the way. It also forced me to delve even deeper in the subject so that I would be prepared to answer students’ questions. I sat down to organize my notes, and this glaze primer is the result.

For students and novice potters, glaze formulation can seem like a daunting body of knowledge. This primer is designed to introduce potters who have little or no glaze formulation experience with a basic understanding of the “how” and “why” of glazes. While this article assumes the reader has already advanced to mixing and using glazes from established recipes, little else is taken for granted. In addition to an explanation of glaze chemistry, this primer outlines several simple experiments that will give the reader first hand experience with glaze materials and their affect on color and texture in glaze.

 4. An Overview of Glaze

When most substances solidify from a liquid state to a solid state, they form a crystalline structure (a regular recurring pattern). Ice crystals, sugar, and salt are all examples of solids that have a crystalline structure. Glass and glaze, on the other hand, are solids that have no crystalline structure, so-called “super-cooled liquids.” As glass and glaze turn from a liquid state to a solid state in the furnace or kiln, the quickly cooling glass does not have the opportunity to form crystals. If a glaze is cooled slowly in a controlled way, some crystallization can take place forming a category of glaze known as “crystalline glazes.”

 5. The Oxide Groups

As described in the last section, glazes are comprised of three groups of oxides: glass-forming oxides (exp. silica), fluxing oxides (exp. magnesia), and refractory oxides (exp. alumina). These groupings have also been described, respectively, as the acid, alkaline, and neutral groups (owing to the pH of the substances), and as the RO2, RO, and R2O3 groups (which describes the numbers of radicals and oxygen atoms in each group). A more poetic way of describing the groups are as the bones, blood, and muscle of a glaze. The glass-formers form the structure of the glaze, the fluxes bring fluidity, and the refractories hold the glaze on the pot. The most common ceramic oxides are listed below in their appropriate RO, R2O3, and RO2 categories:

 6. The Fluxing Oxides

The effect of Silica and Alumina in the glaze have already been discussed, and the role of coloring oxides and opacifiers will be discussed later, so lets take a look at the effects of the fluxing oxides. Each of the fluxing oxides has a slightly different effect on the glaze melt and the particular color that the coloring oxides produce. Despite its position under the R2O3 column, Boric Oxide is included in this section due to its influence on the melt and character of glazes.

Boric Oxide (B2O3)

Character: Boric Oxide is a unique oxide in that it acts as both flux and glass-former. It has no specific melting point, but rather softens over a broad temperature range. Materials rich in Boric Oxide will melt even at bisque temperatures. Due to its low melting range, Boric Oxide can be used at any firing temperature and can be used as the primary or secondary flux of a glaze. Boric Oxide has a low rate of expansion so is helpful in eliminating crazing. It promotes a milky or mottled texture in the glaze melt and helps to produce more intense colors from the coloring oxides.

Color Effect:

Sources: Gerstley Borate, Gillespie Borate, Laguna Borate, Various Frits (3124, 3134), Borax

Barium Oxide (BaO)

Character: Barium is a less active flux and is best used at the higher temperature ranges in small quantities. Barium tends to produce a matte finish due to the formation of micro-crystals in most instances, but will develop dry, rough surfaces if used in too high an amount. In combination with a sufficient amount of Boron, it will create glossy, runny glazes even at lower temperatures.

Color Effect: Barium helps to develop celadon blues in reduction, and can produce colors ranging from pink to purples from cobalt in high fired oxidation. It encourages copper to fire turquoise in oxidation. With nickel, plums, reds, and purples are possible in oxidation above Cone 4. With small amoutns of chromium, acid-yellows, yellow-greens, and chartreuse are possible.

Sources: Barium Carbonate

Toxicity: Toxic (inhalation, ingestion) in material form, and will leach out of fired glazes if not properly fused.

Calcium Oxide (CaO)

Character: Calcium lends hardness, durability, and scratch resistance to glazes. Calcium is the chief flux in most high-fire glazes due to its reliability and low cost, but must be accompanied by one or more secondary fluxes. Used alone, or in high concentrations, calcium will produce dull, dry surfaces.

Color Effect: Minimal color effect in most glazes, but does favor the the development of celadons.

Sources: Whiting, Wollastonite, Dolomite, Various Feldspars, Various Borates, Various Frits, Wood Ash

Lithium Oxide (LiO3)

Character: Lithium produces effects similar to Sodium and Potassium. Lithium is an active flux even at lower temperatures. It is limited to use in lower percentages due to its low rate of expansion, but even in amounts of 1 - 2%, Lithium can significantly assist the melt of a glaze. Substituting Lithium for portions of Sodium and Potassium in a glaze can reduce crazing without a significant affect on glaze color.

Color Effect: Can produce blue effects from copper, and pinks and warmer blues from cobalt.

Sources: Lithium Carbonate, Petalite, Spodumene

Magnesium Oxide (MgO)

Character: Magnesium Oxide is typically a high temperature fluxing oxide, but can be used at lower temperatures as an opacifier. Magnesium will produce smooth matte or satin-matt surfaces, while too much Magnesium will cause dry, rough surfaces. Magnesium has a low rate of expansion that can be helpful in reducing crazing.

Color Effect: Can produce purples, pinks, and lavenders from cobalt. With small amounts of nickel, will produce lime to olive greens.

Sources: Talc, Dolomite, Magnesium Carbonate, Various Feldspars

Potassium Oxide (K2O)

Character: Potassium Oxide, or Potash, is a powerful flux at any temperature. It has the effect of intensifying colors, but has a high rate of expansion that leads to significant crazing in glazes where high amounts of Potassium are present. Potassium produces glazes that are softer and less resistant to wear.

Color Effect: With Barium, can produce pink from manganese. Encourages the celadon blues and greens from small amounts of iron oxide in reduction. Promotes reds from copper in reduction, and blues from copper in oxidation. Can produce purples, violets, burgundys, and red-blues from manganese.

Sources: Custer and G-200 Feldspars (discontinued), Nepheline Syenite, Cornwall Stone, Potassium Carbonate, Various Frits

Sodium Oxide (Na2O)

Character: Sodium Oxide, or Soda, is a powerful flux at any temperature. It has the effect of intensifying colors, but has a high rate of expansion that leads to significant crazing in glazes where high amounts of Sodium are present. Sodium produces glazes that are softer and less resistant to wear.

Color Effect: Effects are similar to that of Potassium.

Sources: Minspar 200 and other Feldspars, Nepheline Syenite, Sodium Frits

Strontium Oxide (SrO)

Character: Strontium is usually used as a flux in the mid to high temperatures, but can be used in small quantities at lower temperatures to good effect. Strontium is often used as a replacement for Barium in glazes in order to reduce toxicity. Strontium has a similar textural effect as Barium, and will create a matte surface if used as the primary flux. Strontium is a slightly more powerful flux than Barium so is generally used in smaller amounts.

Color Effect: No significant color effect.

Sources: Strontium Carbonate

Zinc Oxide (ZnO)

Character: Zinc Oxide is helpful across all temperature ranges in small amounts. At stoneware temperatures, zinc is a potent secondary flux that can have a pronounced effect on the glaze melt. Higher amounts of Zinc tend to produce rough, dry surfaces. Zinc also tends to produce crystalline formations in the glaze and is used in the formulation of “crystalline glazes.”

Color Effect: With copper, can produce turquoise greens. With chrome, produces brown instead pink, and zinc must absent to produce chrome-tin pinks.

Sources: Zinc Oxide, Calcined Zinc Oxide

For more detailed and thorough descriptions of the fluxing oxides and their source materials, check any one of the texts listed in the resource section. Also check the reference and article sections of the Digitalfire website.

 7. What is the use of “calculating” a glaze?

The typical glaze recipe is expressed as a percentage recipe, meaning that the total amount of materials in the base glaze total up to 100%. While this kind of recipe is necessary for actually mixing up a batch of glaze, it is not so helpful in comparing different glazes or determining the relative amounts of oxides in a glaze. Two different recipes can appear to be very different, but their oxide formulas could turn out to be quite similar. A cursory glance at the two recipes below may lead one to believe the glazes are quite different. They use different materials in different amounts, and yet their formulas are nearly identical.

By converting a glaze percentage recipe into a glaze formula, a potter can get a clearer view of what oxides are present and in what amounts. Glaze formulas can be adjusted more easily and accurately, then converted back into a new percentage recipe for actual testing. Also, in the event that a specific material becomes unavailable (such as Gerstley Borate or Albany Slip), understanding a glaze at the formula level makes it possible to use a replacement material to achieve the same oxide makeup in the glaze, allowing the potter to continue using a favorite glaze virtually unaltered.

Calculating the formula for a glaze can also make the development of new glazes quicker and easier. Say, for example, that I have a glaze that has most of the qualities that one would want in a glaze: the glaze melts easily at the proper temperature without flowing too much, the glaze is fairly resistant to scratches, and the glaze does not craze or have any other common faults. In short, its a wonderful, trouble-free glaze that is reliable and easy to use. But what if I decide to develop a glaze that gives me a different color or texture that my original cannot produce? Do I start from scratch with my testing? No. I first calculate the formula for my original glaze and see what oxides are present and in what amounts. Here is one of the recipes we compared a moment ago:

The glaze above is rich in Calcium, low in Sodium and Potassium, and has a good balance of Alumina and Silica. From experience, I know that this glaze “fits” my pots, meaning it does not craze or shiver. Let us say, then, that I want to produce a glaze that is rich in Magnesium because of a particular color effect I read about. The quickest, simplest way to produce an initial Magnesium-rich test is to use the original formula above, but increase the amount of Magnesium in the flux column while reducing one or more of the other fluxes, probably Calcium, while leaving everything else the same. Using a glaze calculation spreadsheet, I shuffle the ingredients and percentages and produce a new formula and its resulting recipe:

Thus, using the calculated formula of a glaze helps to direct one’s glaze research more precisely and results in a large savings in time, effort, and materials over the long term.

Another reason to calculate the chemistry of glazes is to help control their rate of contraction in comparison to the clay that the glaze is on. By aligning the contraction of the glaze with the contraction of the claybody during firing, one can not only eliminate some glaze faults but also strengthen the finished pottery by quite a bit. There is more on this topic in Section 14.

8. Molecular Weight

Calculating the formula for a glaze begins with using the molecular weight of each oxide to arrive at a comparative amount of each oxide in the fired glaze. This may sound confusing, but think of it the following way. To compare glazes, one wants to compare apples to apples, but a since all of the oxides have unique masses, or molecular weights, then trying to compare an unconverted percentage recipe is like trying to compare apples to oranges, or grapefruits, or even watermelons. By taking the molecular weight of each oxide into account, then all the different oxides (an entire basket of fruit) can be successfully compared in apple-to-apple terms.

So what is molecular weight? Looking at the periodic table, one finds each element has been assigned a specific mass. This mass, or “atomic weight,” has nothing to do with pounds or ounces, grams or kilograms. The “weight” of an atom is really an expression of its mass relative to the mass of Hydrogen. Hydrogen has been assigned an atomic weight of one (1) because it is the lightest element. All of the other elements were then assigned weights as multiples of the weight of Hydrogen. Oxygen, for example, has an atomic weight of 15.9998 because it is 15.9998 times heavier than Hydrogen.

So this explains the atomic weights of each element, but how does one determine the weight of a molecule, or molecular weight? Determining molecular weight is simply a matter adding up the atomic weights of the elements in the molecule. To figure the molecular weight of Iron Oxide (Fe2O3), for example, one would refer to the atomic weights of Iron (55.8457) and Oxygen (15.9994) on the Periodic Table and multiply these weights by the numbers of atoms present for each element. Since there are two atoms of Iron and three atoms of Oxygen, the molecular weight would look something like this:

Iron (2 x 55.8457) + Oxygen (3 x 15.9994) = 111.6914 + 47.9982 = 159.6896

This is how one arrives at the molecular weight of any molecule, however simple or complex.

9. Glaze Calculation, Step-by-Step

The following describes the procedure for converting a percentage recipe into a flux unity formula. It is a bit of math, but it’s not difficult math. A later section describes some software that makes this job easier, but it is still important to understand calculation first-hand.

1. Obtain or produce a glaze percentage recipe.

2. Obtain or calculate the molecular formula (unity formula) for the materials used in the glaze (See Appendix C).

3. Construct a table (see example below) in which to enter the oxide contributions of each glaze material.

4. Enter materials and percentages used in the table.

5. Calculate moles by dividing the percentage in the recipe by the material’s molecular weight (also Appendix C), and enter results in table.

6. For each material, multiply moles by the amount of each oxide present in the material (according to the unity formula) and enter the result in the appropriate column.

7. Total the amounts of oxides in each column. This gives you the molecular equivalent formula.

8. Express the molecular equivalent formula as a Flux Unity formula by totaling the fluxing oxides and dividing each oxide in the formula by the sum of the fluxes.

Referring to the material analyses for these materials in Appendix C: Unity Formulae of Common Glaze Materials, one can see that the following oxides are present in the glaze: Calcium (CaO), Magnesium (MgO), Potassium (K2O), Sodium (Na2O), Titanium (TiO2), Iron (Fe2O3), Silica (SiO2), Alumina (Al2O3), and Boron (B2O3). Using these oxides, construct a chart that includes a column for each oxide present, such as the one below.

Calculation Example

We start with the following glaze recipe:

Referring to the material analyses for these materials in Appendix C: Unity Formulae of Common Glaze Materials, one can see that the following oxides are present in the glaze: Calcium (CaO), Magnesium (MgO), Potassium (K2O), Sodium (Na2O), Titanium (TiO2), Iron (Fe2O3), Silica (SiO2), Alumina (Al2O3), and Boron (B2O3). Using these oxides, construct a chart that includes a column for each oxide present, such as the one below.

Having constructed the chart, now look up the molecular weight for each material used in the recipe. Again, this information can be found in Appendix C. The molecular weights for the materials are:

Now divide the percentage of the material in the recipe by the molecular weight of that material and enter the result under the Moles column. The chart should then look like so:

Next, multiply the Moles number for each material by the amount of each oxide present in that material’s unity formula. The unity formula for G-200 is shown below (and in Appendix C). Each oxide in G-200 is multiplied by the mole number, 0.0897, and entered into the appropriate column in the chart below (Note: All calculations in this example are rounded to four decimal places in order to fit in the space provided. For that reason, a few entries will appear as zero amounts even though a trace amount of the oxide is present.)

Now perform the same calculation with the other materials in the recipe. Then total the columns and enter the results at the bottom. The final chart should look like this:

The totals at the bottom of the columns are a molecular equivalent formula for the glaze recipe, however this formula is not all that helpful in analyzing this glaze or comparing it to others. To make this formula more “readable,” it needs to be expressed as a flux unity formula, meaning that the total of the fluxing oxides in the formula should add up to one. Expressing the molecular equivalent formula as a flux unity formula is simply a matter of totaling the fluxing oxides in the formula (in this example, calcium, magnesium, potassium, and sodium) and dividing each oxide in the formula by the total of the fluxing oxides.

The resulting flux unity formula, arranged in the appropriate RO, R2O3, and RO2 columns looks like this:

10. The Limit Formula

A limit formula is a formula in which all of the oxides present have the minimum and maximum amounts recommended for a glaze temperature range or a specific glaze type. The formula below is an example from Clay and Glazes for the Potter, 3rd edition, by Daniel Rhodes:

A limit formula can be helpful in originating new glaze bases and provides a standard for comparison with other glazes. If one knows that the maximum amount of calcium in a typical functional stoneware glaze is .7, then a glaze with a calcium level of almost .9 should raise a red flag and inspire thorough testing. Glazes that do not have the proper balance of fluxes, silica, and alumina may not mature and fuse properly. This can cause a variety of problems, the most serious of which is the leaching of toxic materials from the glaze into food or drink. Sometimes a potter may have a very good reason for breaking the guidelines of a limit formula, perhaps for a specialty glaze or a glaze intended for sculptural work, but glazes for functional ware should generally fall within the limits of the limit formula.

With the issue of safety and toxicity at hand, is it responsible for a potter or ceramicist to develop a glaze that goes outside the boundaries of the limit formula for functional ware? In the quest to develop glazes with specific or unique characteristics, a potter may be tempted to use more or less of an oxide than is recommended by limit formulas. While glazes such as these may turn out to have stunning effects, they are likely to be unsafe for use with food or drink. Even if a potter or sculpture develops a glaze strictly for visual effect, there is some responsiblity to keep that glaze out of circulation. As Tony Hansen, owner of the Digitalfire Corporation and a proponent of the formula and oxide viewpoint in developing glazes, describes:

“if a glaze formula has twice the maximum of an oxide (I see this all the time with manganese, barium, lithium, fluxes, metallics) or half the minimum (i.e. SiO2 or Al2O3) eyebrows should raise! It is true that ‘visual’ glazes often defy the norms (i.e. fluid variegated glazes often lack alumina) but no one expects them to be food safe, fit the clay perfectly, and fire to a very hard and abrasion resistant surface either. We also expect that such glazes not be trafficked on the functional glaze market where they quickly develop amnesia about who and what they are.”2

11. Glaze Calculation Software

Now that the reader has some understanding of calculating the formula for a glaze, it may be a good point to discuss the usefulness of commercially available glaze calculation software. The two products that I am most familiar with are Digitalfire Insight and Glazemaster. Both of these software packages provide the basic function of calculating the unity formula of any glaze recipe, which in itself would make them worth the money. However, the software also allows the user to conveniently store recipes within the program, compare different recipes, make easy adjustments to formulas and recipes, calculate the thermal expansion of glazes, and input new materials. Both products offer a 60-day free trial download from their respective web sites, so trying out the features of these packages is easy. Another software package is The Glaze Calculator, available as a free download. I have not had the opportunity of use this last package. The web sites for these programs are included in the Resource section.

Lest the reader think the author is peddling glaze calculation software, I’d like to point out that it is possible to design a rudimentary calculation program using a spread sheet such as MS Excel®. Looking at the chart one constructs to perform the calculations by hand, one can see that the procedure for calculation is relatively simple but requires both a lot of repetitious calculating and referencing of oxide data, so automating the calculations in a spreadsheet makes perfect sense. I have created such an Excel® program so that my glaze chemistry students would have a free, consistent way to calculate their test glazes throughout the semester. My program is limited in that it uses only forty-six of the most common glaze materials, but it serves its purpose fairly well and has become the program I turn to most often when working with my glazes.

12. Development of Color

One should note that in all of the glaze calculations thus far, the coloring oxides have not been included. As mentioned in section four, the coloring oxides do not offer, for the most part, any qualities other than color and opacity in the melted glaze. Most of the time, one will develop a glaze base that melts at the proper temperature and has the desired fired properties and then introduce different coloring oxides. Before focusing too much on the coloring oxides, though, first consider all of the variables that might affect color in a glaze. These are:

1. Coloring oxides

2. Fluxing oxides

3. Texture of the glaze (matt, semi-matt, gloss, etc.)

4. Atmosphere during firing

5. Color of the clay body

While five variables may not seem like a lot to work with, a surprising number of colors and variations can be achieved through persistent and vigorous testing. Many beginning glaze chemists will already be familiar with the differences in an oxidation firing versus a reduction firing as well as the dramatically different affects of the coloring oxides. The changes that can occur from the color of the underlying clay and the texture of the glaze may also be familiar. The color variable that often goes under-appreciated among students and others new to glaze chemistry is the color affect of the fluxing oxides. Understanding the color affect of the fluxes is really at the heart of color development because manipulation and blending of the fluxing oxides in a glaze base is what achieves the variety and intensity of color possible in glazes. Copper will produce a significantly different color in the presence of sodium than it will in presence of magnesium. While reading some glaze texts will reveal many descriptions of the effect of fluxes on colorants, only through experimentation can one really appreciate the variety and subtlety of these differences. The different tests in Appendix C quickly lead up to producing a wide variety of color and texture in glazes.

13. Opacity

As mentioned in Section 4, some oxides are responsible for making glazes less transparent and more opaque. Some glazes are essentially clear and one can clearly see the claybody underneath the layer of glaze. These glazes can be said to be completely melted, as all of the components of the glaze have vitrified leaving little or no unmelted particles suspended in the glassy layer. If a glaze is underfired, leaving some particles out of the melt, then a glaze becomes more opaque.

Even when a glaze is fully matured at the right temperature, opacity can be caused by a few other factors. The presence of tiny bubbles in the glaze can produce a opalescent effect, such as in a blue celadon. The presence of Phosphorus, contributed by bone ash, can produce this effect. Opalescence can also be produced by “a mixture of glasses of differing indexes of refraction.”3 Boron can encourage this quality in a glaze, producing a milky blue-white texture typical in rutile blue glazes. Solid crystals in the glaze, both tiny and large, can refract light and also create an opaque effect. Glazes rich in iron oxide, such as iron red or tenmoku, are an example of this effect. Some fluxing oxides -Calcium, Magnesium, and Barium- have an opacifying effect on a glaze if they are present in high enough concentrations. The opacity derived from all of these effects, though, can be driven out if the glaze if fired to a high enough temperature.

The opacifying oxides can generally produce opaque or semi-opaque qualities in the glaze regardless of firing temperature (within reason). These are Tin Oxide, Zirconium Oxide, and Titanium Oxide. Tin Oxide has been in use the longest and produces perhaps the best opacifying and textural effect in a glaze, but the high price of Tin Oxide makes Zirconium Oxide a more affordable alternative. Zirconia is sold under a variety of brand names, such as Zircopax, Superpax, and Opax. Titanium Oxide contributes opacity to glazes, but also has a softening effect on some colorants. Titanium Oxide is often introduced into glazes as part of the material, Rutile, and is the chief contributing factors to the creation of rutile blues.

14. Other Considerations

Once a potter has produced a glaze that melts at the correct temperature and has a desirable color and appearance, two additional areas can become a concern. The first is the mechanical nature of the glaze batch. Calculating the chemistry of a glaze becomes fairly simple when one gets comfortable with the process, but looking at a glaze from a strictly chemical point of view fails to address the challenge of getting the glaze on the pot successfully and consistently. For example, a glaze recipe needs to include enough “plastic” material, such as kaolin or ball clay, in its make-up in order for the glaze batch to remain in suspension in the bucket during the glazing process. Without this plastic component, the glaze materials would simply settle out of the water to the bottom of the bucket. In addition to the help with the suspension of the glaze, clay provides both alumina and silica to the glaze chemistry. From the chemistry standpoint, the alumina portion of a glaze could be provided exclusively with alumina hydrate and the silica provided exclusively with silica, however neither of these materials are plastic and would fail to provide the needed suspension for applying the glaze to the pot.

Another mechanical factor to consider is the soluability of the materials one uses to make a glaze. Some common glaze ingredients such as lithium carbonate, soda ash, and gillespie borate are soluable or partially soluable in water, a feature which can cause problems in the glaze batch. Materials that dissolve in water may soak into porous bisqueware during glazing, then leach back out of the clay as the piece dries. During glaze firing the soluable material will volatize and run the risk of flaws in the fired glaze surface. Pitting, pinholing, and cratering can all be caused by volatized gases erupting through the molten surface of glaze during firing. By avoiding the use of soluable materials whenever possible one can often solve some of these glaze faults before they begin. If one needs lithium in a glaze, then consider the use of spodumene instead of lithium carbonate. When boron is required, Frit 3134 may be a more reliable source than Gerstley Borate. Sometimes one must use a soluable material to satisfy the oxide requirements of a glaze, but most often glazes can constituted from insoluable alternatives.

Another source for pitting and pinholing problems are materials that are not soluable, but do have components such as carbonate and sulphates that similarly volatize out of the glaze during firing. Whiting, a source for calcium, loses more than 40% of its mass in the form of carbonate during firing. Dolomite, another source for calcium, loses almost half of its mass during firing. If a glaze that included these materials demonstrated a problem with pinholing, a possible solution might be eliminating whiting or dolomite in favor of wollastonite, which loses only 2% of its mass during firing. A simple solution to a problem which can best be accomplished by calculating the formula for the glaze.

Having just commented on reasons to avoid the use of soluble materials, let me now point out one case where a bit of soluble material may be desirable. Anyone who has ever attempted to glaze a pot with a glaze slurry that has flocculated (become thick and gummy), knows that it is an unpleasant, if not impossible, task. Flocculation occurs when particles in the glaze slurry cling together, forming clumps, due to electric attraction. These clumps, or “flocks,” do not flow across one another as fluidly as single particles do, and therefore the glaze slurry requires more water to attain the same viscosity as a non-flocculating glaze4. Materials such as gerstley borate, gillespie borate, and laguna borate seem to be particularly culpible in producing such glaze batches. This condition can often be remedied with a small addition of a soda ash solution, which introduces an electrolyte into the slurry that breaks up the “flocks.” Three tablespoons of soda ash dissolved in one quart of water seems to be just enough of a deflocculant to treat two 10,000 gram batches of glaze. The small amount of soluable material does not seem to have a significant effect on the quality of the fired glaze.

Beyond the mechanical properties of the glaze batch, the second area of concern is the expansion of the glaze during firing. All glazes will contract, or shrink, a bit as they fuse. The rate of contraction varies among different glazes according the amounts of the different oxides present in the glaze. The expansion (or contraction) of a glaze can become a problem when the rate of expansion of the glaze differs from the rate of expansion of the clay that the glaze is applied to, described as the coefficient of expansion. When the clay and the glaze contract at different rates, glaze faults such as crazing and shivering can result. Crazing describes the tiny network of cracks that form when the fired glaze contracts more than the fired clay, while shivering describes the shearing off of glaze from the clay when the glaze contracts less than the clay. Both conditions can cause the fired clay to become weaker than it would be without any glaze at all. A glaze that fits well, however, can strengthen the clay quite a bit.

A glaze chemist can control the expansion of the glaze by changing the chemistry of the glaze slightly. Some oxides, such as sodium and potassium, have a very high rate of expansion, while other oxides have a comparatively low rate expansion, such as silica, zinc, and magnesium. By reducing oxides with a high expansion rate and increasing oxides with a low expansion rate, one can eliminate crazing. Similarly, one can solve a shivering problem by increasing the level of oxides with higher rates of expansion. Experience shows that a small addition of lithium, with its very low rate of expansion, can be very helpful in reducing crazing in a glaze without a significant effect on the color of the glaze. Again, familiarity with glaze materials, and their oxide makeup, will be most helpful in solving expansion issues. The expansion rates of some oxides are available in Appendix B.

15. Resources

Recommended Glaze Tests:

Clay and Glazes for the Potter, 3rd edition, by Daniel Rhodes; ISBN 0-87341-863-8

This is the glaze text that I came through school with (well, actually the 2nd edition) so it would be difficult for me not to give it high praise. With that being said, the chapter on glaze calculation has some problems that I’m surpised have not been corrected in this latest edition. However, it is still a very valuable text and has very good descriptions of materials and oxides as well as many other topics.

The Ceramic Spectrum, Robin Hopper, ISBN 0-80197-275-2

Another good glaze formulation text, but one that does not dwell on calculation. Robin Hopper takes an experimental approach with developing glaze bases. I especially like his “glaze saturation” tests. I did a similar type of experiment during college.

Glazes for the Craft Potter, Harry Frasier, ISBN 1-57498-076-9

This book is at the top of many studio glaze chemists’ “must have” lists. I think it is an excellent text with a good section on calculation.

Mastering Cone 6 Glazes, John Hesselberth and Ron Roy, ISBN O-97300-630-7

With so many potters relying on electric kilns for their glaze firings, it is important to have a text that specifically addresses Cone 6 glazes. I highly recommend this text for anyone interested in developing their own electric kiln glazes.

AUTHOR'S UPDATE: Since this primer was written in 2006, great new chemistry books have come out (or come to the author's attention). There are numerous books dealing with all facets of glaze chemistry available!

References:

1. The American Heritage College Dictionary, 3rd edition, ISBN 0-395-67161-2, p. 525

2. “Limit Formulas and Target Formulas”, http://ceramic-materials.com/cermat/education/206.html (2019 Note: now a dead link)

3. Clay and Glazes for the Potter, 3rd edition, Daniel Rhodes, ISBN 0-87341-863-8, p. 216

4. Clay and Glazes for the Potter, 3rd edition, Daniel Rhodes, ISBN 0-87341-863-8, pp. 99-100

Appendix A:

Weights of Atoms and Molecules

Appendix B:

Expansion of Oxides

Appendix C:

Unity Formulae of Common Glaze Materials

Appendix D:

Experiments

The following experiments are designed to provide the reader with direct, relevant information and experience with glaze materials and glazes bases. These experiments were initially designed as part of a science fair project for my nine-year-old niece, Claire. She had no trouble performing the experiments, largely unassisted, and drawing excellent conclusions from the results. My hope is that the reader will find them equally helpful.

The first experiment looks at the melting properties of single materials, while the second experiment demonstrates how to develop an original glaze base from basic materials. The third experiment adds coloring oxides to the base to begin producing a palette of glaze colors, and the final experiment shows how to develop additional glaze bases that will offer a wide variety of textures and color effects with the fluxing oxides.

Experiment 1: Melting of Single Materials

This experiment demonstrates the melting of single materials at the two critical temperatures for a potter, bisque temperature and glaze temperature. All one needs to do this experiment is to make a tile of roughly 5 inches by 8 inches. While the clay is still soft, use a finger to make small depressions in the tile, one for each material you have available to you (excluding colorants), and number each depression. My studio currently has twenty-one different materials so I would make a tile with that many depressions. Once the tile has dried, place a tiny amount (1/4 teaspoon) of each material in the depressions, one material for each depression. Be sure to make note of which material is in which depression. Fire the tile to your normal bisque temperature (for me, Cone 09). Once fired, which of the materials, if any, melted or fused in some way? Record your observations and move on to the next part. Fire the same tile (or a duplicate) up to glaze temperature (Cone 6 - 10).

Which materials have melted or fused at the higher temperature? Again, record your observations. What conclusions can you draw from this simple test? Of the materials that melted at bisque temperature, what oxides might they have in common? What oxides are common to the materials that melted at the higher temperature? What materials might be helpful in lowering the melting temperature of a glaze?

Experiment 2: Creating a Base Glaze

The first experiment demonstrated the melting of single materials. This experiment begins to blend materials in an effort to find a combination that melts at a glaze temperature appropriate for the type of work that you do (or want to do.) This experiment, while simple, is a bit more involved, so the procedure is laid out more formally.

The first part of the experiment is really a bit of research. Look through various glaze recipes designed for Cone 6 - 10. What materials are included in just about every stoneware glaze recipe? The answer is Feldspar, Whiting, Silica, and some kind of clay (EPK, Ball Clay). So, the experiment is a methodical testing of each variation possible of these four materials in amounts ranging from 0% to 50% at 10% intervals. To make sure each variation is tested, its helpful to lay out the tests on a quadraxial blend grid (Fig. 1). On the grid, any point where two lines meet represents a variation in the blend of the four materials. (For example, Test Variation #7 would be: EPK 40%, G-200 Feldspar 50%, Silica 0%, Calcium Carbonate 10%.) For this experiment, one should use specific brand name materials (when possible) so that an accurate analysis of the tests can be made later. In my test, I used Silica (brand name unknown), G-200 Feldspar, Whiting (brand unknown), and OM-4 Ball Clay, but one could use any of the brands effectively.

Tools Needed:

37 plastic cups or other containers, a gram scale, a cup-size 60 mesh screen (or a blender), some spoons or scoops, plastic wrap, and rubber bands

Procedure:

1. Create four 9” x 9” tiles approximately 1/2” thick. (or one large 18” x 18” tile.)

2. Make nine 2” diameter, 1/8” to 1/4” deep impressions (3 rows of 3) in each tile using the bottom of a cup (I used a plastic measuring cup). Number the impressions 1-36 across the four tiles. If you are using the 18”x18” tile, make 36 impressions (6 rows of 6) Note: What you are doing here is creating cavities in which to test the separate variations. By testing in these cavities, one avoids the risk of variations that are too runny and cause kiln shelf damage.

3. Allow tile(s) to dry thoroughly, and bisque.

4. Label 36 of the cups 1 - 36 (one for each of the variations) and retain one cup for use on the scale.

5. Write out the recipe for each variation for a twenty gram batch. Twenty grams of glaze is plenty for carrying out this test and avoids unnecessary waste.

6. Calibrate the scale. Since you are working with very small batches of glaze, precision is critical for achieving accurate results.

7. Weight out the materials for each variation and deposit them in the appropriate cups. One may find it easier to weight out one material for each variation before moving on the next material (i.e. weight out the feldspar for all 36 variations, then weight out the silica for all 36 tests, etc.)

8. Once the dry batch of each variation is complete, mix each batch with a small amount of water in order to produce a usable glaze slurry. Mix thoroughly and pour through a cup-sized 60 mesh screen. If you have one available, using a blender for mixing the glaze slurry is a lot faster and easier. A blender will thoroughly blend the dry materials and water without the need for screening. Just be sure not to let any dry material cling to the damp inside walls of the blender (that would make your tests less accurate by unevenly removing some materials) and rinse the blender between variations. If, after creating the wet batches, you realize you have used too much water, simply allow the batches to settle overnight and sponge off some of the clear water before performing the next step in the test.

9. Pour a bit of each variation into the numbered depression on the tile(s) that the variation corresponds to. It is not necessary to fill the depression or to use all of the wet batch.

10. After all of the variations have been applied to the appropriate depressions, allow the tile(s) to dry. Then place the tile(s) flat into your kiln and fire to glaze temperature.

After the fired tile(s) emerge from the kiln, study the results. How many of the variations in this test melted? How many did not fuse at all? Which variations have a melted consistency that interest you? Notice that some of the tests that are very high in calcium carbonate did not melt any more than some of the tests that have little or no calcium. Why might this be so? If any of the results from this test did not come out the way you expected, then this is a good moment to do some additional reading (or re-reading) from a glaze text.

Since this test is designed to give you direct experience with glaze chemistry and developing glazes, take some time to really analyze the results. Calculate the unity formula for each of the tests (or at least the tests that melted). What seem to be the minimum and maximum amounts of calcium that is useful in melting the glaze? How much variation is there in the alumina and silica content among the tests that melted?

Experiment 3: Color Testing

Select the coloring oxides that you are interested in using in your glazes. The coloring oxides I tend to use are: Copper, Cobalt, Iron, Rutile, Manganese, Chromium, and Yellow Ochre, but you could add others to your list. For the purposes of outlining the experiment, I’ll use my short list of seven. This series of tests is simpler than the previous series, so a specific procedure is not necessary. With the feldspar, clay, silica, and whiting (calcium carbonate) base that you have developed, mix up a 50 gram test batch for each coloring oxide you wish to use and add that colorant to the base before mixing with water in the blender (or screening). How much of each colorant do you use? If you want to be truly methodical, then mix up a base with 1%, then one with 2%, then one with 3%, etc. for each colorant you want to test. If, like me, you don’t want to waste expensive coloring oxides on unnecessary tests, then refer to a glaze text to see what percentage of each coloring oxide is recommended. Appendix 7 on page 339 of Clay and Glazes for the Potter, 3rd edition, by Daniel Rhodes has a good chart of recommended colorant percentages for achieving different colors. I have had success with the following percentages:

Note that Red Iron Oxide is tested at two different percentages, which provides an opportunity to develop a celadon-type glaze and a iron saturation glaze.

Once you have mixed up your base/colorant tests, apply the glazes to a test tile with a sponge or dip a free-standing test piece into the cup of glaze, or do both. The test tile is nice because it provides a broader surface for examining color and is easier to store for future reference while the free-standing test piece is better for showing the fluidity of the melted glaze. Fire the test pieces to temperature in reduction and/or oxidation. Ideally, you should test all of your color tests in both atmospheres, but some students and potters only have access to one or the other.

Experiment 4: Developing Other Bases

Having now developed a calcium-rich base glaze and added a variety of coloring oxides, the next step is to develop other base glazes that are rich in the other fluxing oxides in order to see their effect on the colorants. The simplest way to create new bases is to use the base you have already created and perform a line blend between your base and a different material. A line blend is much like a quadraxial blend in that it is a methodical way to blend substances but differs in that it only blends two variables rather than four. In this case, a line blend is a way to introduce a new material, and its constituent oxides, to the base in increasing amounts to test their effect. The list of materials that one could select for this line blend is nearly endless; from Nepheline Syenite and Alberta Slip to more exotic materials to your own local materials such as surface clay or crushed sea shells. The likeliest candidates for this test, though, are standard ceramic materials that are easily and reliably available. Any of the materials you used in Experiment 1 would make a good start, but a list of recommended materials is given below.

All one needs to do to perform a line blend is to lay out the incremental amounts by which one wants to blend the two materials. A very brief line blend might blend two materials in increments of 25% :

But for this test a more helpful blend would be increments of 5% or 10%:

In the above example, I have dropped the mixes that would test 100% of the base glaze and the variable material since these have been tested in earlier experiments. As the variable material increases in relation to the base glaze, one gets higher and higher concentrations of the oxides contained in that material. By testing in this fashion, one should be able to create at least one new base glaze per line blend with desirable properties.

To perform this test, assemble your supplies (materials, cups, labels for the cups, blender, etc.) and work out the amounts for each individual test to create a 50 gram batch. If you plan to mix several line blends for different materials, you may consider mixing a large dry batch of the base glaze (mixing and dry screening thoroughly to ensure homogeneity) in order to speed the mixing process. Mix the individual tests, clearly marking the cups the tests are in, add some water, and mix in the blender (or mix and screen through a cup-sized 60 mesh screen). You can apply and fire the tests on a tile or on free-standing test pieces, however the free-standing test pieces will show you more about the flow of the test glaze. Once the tests are fired, you should find that one or more of the tests has melted to a functional and attractive surface. Now perform the color testing for your new base just as you did in Experiment 3, preferably using the same colorants in the same percentages so that you can get a good comparison of the color effect of the different oxides.

One can see that a chemist could spend a lot of time mixing all of the possible tests, not to mention the tests one could perform to blend different oxides and different colorants. This is when working in a group with other potters or students and sharing results could be of great benefit. At some point, you will have to make decisions about what tests you care to perform and which tests to set aside. Simply follow whatever avenue of testing interests you, but always remember to keep your testing notes (observations, conclusions, and ideas for follow-up tests) where you can find them.

Suggested Materials:

The following materials are recommended because they will provide a cross section of the different oxides available for ceramics. There are plenty of other materials that could give good results. The primary oxide that the material is intended to provide is given in parentheses.

• Spodumene (Lithium)

• Alberta Slip (Iron)

• Nepheline Syenite (Sodium/Potassium)

• Gerstley, Laguna, or Gillespie Borate (Boron)

• Talc (Magnesium)

• Dolomite (Magnesium)

• Barium Carbonate (Barium)

• Strontium Carbonate (Strontium)

• Zinc Oxide, calcined (Zinc)

• Borax (Boron & Sodium)

• Wood Ash (Different Oxides)

One could also try Soda Ash and Potassium Carbonate for Sodium and Potassium, respectively, but these oxides generally show up in lower amounts than most of the tests in this line blend will provide. One may consider combining Soda Ash or Potassium Carbonate with Nepheline Syenite (50% / 50%) to achieve a more functional test for Sodium and Potassium.

Return to the Beginning