Javascript must be enabled in the security or content
settings of your browser for this site to work properly

To learn more about cookies search for HTTP cookie at

Home | Overview | Description | Substitutes | Boraq | Glaze Recipes | Learning | Contact

INSIGHT Software, the best way to deal with this: Watch the Gerstley Borate video here.

Iron Red GB Glazes

Gerstley borate was used widely in bright red and maroon high iron glazes. These contained as much as 25% iron oxide and many had more than 50% GB. The red colors can be quite stunning (red is a very difficult color in ceramics). Needless to say, they were very messy to work with. These glazes are temperamental also, they have to be the right thickness (they run easily and bubble/blister if applied too thickly, color does not develop if they are too thin). Although it seems logical that they not be fired too quickly to give the crystals time to grow, one respondent with 18 years of iron red experience said this: "cone six firings require a medium fast fire to maximum temperature in a vented kiln and the same for cooling. Venting assures adequate oxidation. If maximum temperature is not reached or the elements are going and it takes a much longer time to reach temperature the color will be brown". 

Boraq 1 does not easily produce intense color in iron red glazes. It tends toward dark red and maroon colors. We got some opinions from visitors on this problem and it appears that it has been solved in the form of an adjustment to the Boraq 1 recipe. Some thought that silica or potash might be lacking, that B2O3 might be excessive in Boraq 1, or that CaO or MgO are reacting to produce green mag/calcium-iron-silicate crystals. However one comment tipped us off:

'Doesn't the different formula weight of Boraq mean that you would not substitute it on a gram-for-gram basis with GB?'

Sure enough, Boraq 1 has a much higher formula weight, that means less is required. This is why glazes have been running more with Boraq. We compared Boraq 1 and GB calculations again (using non-unity and percentage analysis numbers this time) and it became clear that CaO was lacking much more than the unity formula we used to develop Boraq 1 suggested. This confirms that a minimum (although not high) amount of calcia is needed for the development of the red iron crystals. High magnesia content also appears to be an important factor. Magnesia is being delivered by significant amounts of talc or dolomite in the recipes so it is not as important that we supply it in the Boraq.

To the left is one example of a cone 6 iron crystal red that employs 31% Gerstley Borate. This is a horizontal surface and the two are applied close to the same thickness. The red crystals lie on a dark glass on both and they are fairly similar.  Click here to see vertical surfaces that are less similar, the Boraq version being darker because its crystals are larger and more isolated whereas the GB version has a very fine mesh of crystals that cover the entire surface.

Below is a worst-case example of the 2826E recipe using Boraq 1. It is applied thinly and only a few red crystals have managed to form. However a best case example is the Boraq 3. The person who submitted this glaze for testing said that 'green flecks' would sometimes appear with new batches of Cornwall Stone.

G2826E GB 
Boraq 1 Boraq 3
CaO 0.41 * 8.11 6.62 8.71
MgO 0.46 * 9.10 8.43 8.29
K2O 0.03 * 0.64 0.71 0.69
Na2O 0.11 * 2.10 1.76 1.63
TiO2 0.00 0.04 0.05 0.05
Al2O3 0.17 3.39 3.67 3.61
B2O3 0.41 8.16 9.68 8.61
SiO2 3.14 62.59 63.19 62.55
Fe2O3 0.29 5.86 5.84 5.82

This chart compares the mole% formula of the Gerstley Borate version of the glaze with the Boraq versions (mole% is the formula in percent rather than unity format, it is being popularized by Richard Eppler).

With GB

With Boraq 1

With Boraq 3

Another demonstration of the benefits of adding dolomite and whiting to the Boraq 1 recipe for iron reds (making Boraq 2 and 3) are the fired results with G2826D recipe. The original glaze contains more than 50% Gerstley Borate. The fired samples of the Boraq 2 and GB versions here look more different than the glazed tiles.

 G2826DGB FormulaGB Mole%Boraq 2Boraq 3
CaO 0.51 * 13.84% 13.72% 14.86%
MgO 0.40 * 10.86 10.71 9.39
K2O 0.00 * 0.10 0.19 0.19
Na2O 0.09 * 2.57 1.72 1.72
Al2O3 0.01 0.35 0.77 0.78
B2O3 0.54 14.69 15.42 15.41
SiO2 1.79 48.91 48.83 48.98
Fe2O3 0.32 8.68 8.55 8.58

With Boraq 2

With GB

We also did a firing with four versions of this glaze, one with GB, one with Boraq2, one with Laguna Borate and one with our Laguna Borate extrapolated recipe. All of them fired almost identical.

The Boraq 3 version gave best results. Small magnified sections of tiles with an incised line are shown below. As in the above glaze, the samples look identical.

INSIGHT 5.3 Full Report on Boraq 3 Version

2826D Iron Red with GB
2826D Iron Red with Boraq 3
Analysis Mole%
CaO 0.51 9.9 13.8
MgO 0.40 5.6 10.9
Na2O 0.09 2.0 2.6
Al2O3 0.01 0.5 0.4
B2O3 0.54 13.0 14.7
SiO2 1.79 37.5 48.9
Fe2O3 0.32 17.7 8.7
  LOI 13.7  

Analysis Mole%
CaO 0.57 10.8 14.9
MgO 0.36 4.9 9.4
K2O 0.01 0.2 0.2
Na2O 0.07 1.4 1.7
Al2O3 0.03 1.0 0.8
B2O3 0.59 13.8 15.4
SiO2 1.87 38.0 49.0
Fe2O3 0.33 17.7 8.6
  LOI 12.1  

As a cone 5 glaze, it is apparent that iron red glazes can produce stunning surfaces, we were amazed and the vibrancy and variegation of the color, especially on porcelain. However we were also amazed at the amount they run, this must be considered or kiln shelves will be ruined. Generally you will have to fire ware on disposable tiles to protect the shelves.

Randy's Red

This is a popular iron red and it employs the standard GB 50, kaolin 20, silica 30 base recipe and adds 15% iron. Our tests using Boraq 2 show it to fire very similar. It does go on thicker so you must dip quickler.

What to do about the runny nature of iron reds

Here is a comment from Paul Lewing: "I've tested and used dozens of versions of iron red glazes ... They are all runny and very sensitive to temperature and even more to rate of cooling, and are very affected by the brand of iron you use. And they won't be red if the Si/Al ratio is under about 18:1, or if there's more Mg and Ca. Food-safe? maybe. Durable? no way. I've given up on them in favor of a lead glaze I call Drop Dead Red, which has all the same post-firing problems but does at least come out consistently iron red and glossy."

Here is a comment from Timoty Joko-Veltman: I have to strongly disagree with Mr. Lewing on this one. Below are the unity formulas for two recipes I fire to 1270C (oxidation), and which seem to me to contradict what Mr. Lewing says.

Kaki #1
0.44 CaO
0.23 MgO
0.23 K2O
0.11 Na2O
0.18 P2O5
0.40 Al2O3
2.98 SiO2
0.20 Fe2O3
7.42 Si:Al Ratio

Kaki #2
0.56 CaO
0.22 MgO
0.15 K2O
0.07 Na2O
0.27 P2O5
0.45 Al2O3
3.88 SiO2
0.23 Fe2O3
8.57 Si:Al Ratio

My personal preference is for the latter of the two, but both are reliably red, glossy, non-runny, and contain significant amounts of CaO and MgO.

However there is an answer:

Use a less fluxed and therefore less fluid underbase version of the glaze and a thin application of the fluid overtop. How? Use INSIGHT to increase the silica/alumina in the iron red glaze while maintaining their ratio, apply this less melting version in a thick layer, then apply a thin layer of the original glaze. The result will be the visual character of the thicker version without the runniness. Of course, this means you need to have two glazes and double dip, but it is worth it for iron reds. For this to work your glaze cannot have excessive shrinkage during drying or it will crack and peel after double dip. Here is an interesting parallel: we found that Randy's Red (15% iron) gives the brighter color associated with a thick layer and yet does not run when applied in a thin layer over Ravenscrag slip on porcelain and whiteware bodies.

Consistent Iron Oxide, We Take it for Granted

Here is an interesting note from one reader: "For a long time iron oxide (as a coloring or pigment) was found in natural deposits. While these materials worked well in paints they were not chemically as consistent as they physically appeared. This led commercial ceramics to condemn it as an unwanted tramp ingredient or contaminant.  That was a long time ago and today's technology in processing iron by companies like Elimetis and Pfizer (before them) has made some very pure forms available to us. Parmelee, Seger, Binns and many of the other ceramic pioneers did not have access to consistent supplies of iron oxides. So, times have changed and company A's brand 9098 is the same in Portland as it is in London. Iron no longer has to be considered a fugitive material - it can still vary a lot from source to source, just like clays and feldspars. But, even though its primary use is still as a pigment, it can also be chemically consistent as long as we stay with the same reputable manufactures brand and grade.

Tony Hansen

Follow/Like me on Facebook
Google+ Add me to your Google+ Circles
Follow me on Twitter

DigitalFire Logo

Privacy Policy