Cosmic Indifference: Galactic Indifference at Cone 6

Disclaimer: I do not recommend using Cosmic Indifference on functional ware. It may leach heavy metals and should not be applied to food surfaces.

Introduction

Galactic Indifference is a really pretty glaze. It shimmers in the light. It crystallizes when thick. I have not seen a glaze like it - it has quite the unique chemistry - yet it is as much trouble as it is desirable. Although claimed to work from cones 6 to 10, in typical firings it is a disappointing mass of over crystallization. The glaze is at its full glory in soda firing, but even then, uneven soda distribution can make pieces hit or miss.

I think it's possible to have all of Galactic Indifference's best aspects at cone 6. And I think I can make the glaze even better by fixing issues with the cone 10 soda version.

Mechanism

Galactic Indifference on Glazy

NC-4 is a soda feldspar and FF 3124 is a boron frit.

How does Galactic Indifference work? Looking at the recipe, most of the ingredients are found in other typical cone 10 glazes, but there are two things that jump out: the use of lithium carbonate and 12% titanium oxide.

Lithium carbonate is an uncommon glaze ingredient. It is a powerful flux and also crystallizes in runny glazes. Both of these properties are being used in Galactic Indifference. Lithium melts at cone 6 too, so it can be easily added.

What sets Galactic Indifference apart from other glazes is its inclusion of a whopping 12% titanium oxide. Titanium is hard to dissolve in glazes and just 2% can start to create crystals. Galactic Indifference (plus extra sodium from soda firing) contains an absurd amount of flux just to do one thing: dissolve all the titanium.

For my cone 6 version, I'll use boron as my main power flux. I'll increase boron as much as I can while keeping these key aspects of Galactic Indifference the same:

• 0.13 molar Li2O (for crystallization)
• 0.3 R2O : 0.7 RO flux ratio (for color response)
• 4.8 SiO2 : Al2O3 ratio (for surface texture)

1. Melt Test

For the initial test, I use glaze calculation software to come up with two formula that meet specific requirements. I think that these properties are important to how the glaze functions:

• The recipes target a 4.8 Silica-Alumina Ratio
  • This determines the surface texture
  • Likely to influence how much titanium dissolves and how it crystallizes
• Add 0.1 to 0.2 molar boron
  • Necessary to lower the melting temperature to cone 6
  • The original recipe has a small amount, so likely affects the visual character
• I don't initially add MgO as it doesn't melt as aggressively at cone 6
  • If necessary, I can add it back in
  • The original doesn't have that much (0.1 molar)
• Target around 20% kaolin for suspension

To start, I performed a linear blend between two recipes. One recipe melts too much and the other one melts way too much (theoretically). I expect some blend of the two to dissolve the titanium without being overly runny.

When mixing this glaze, I noticed that it had a very thick consistency. Adding water didn't help and caused the glaze to crack when drying. Water should be kept low and the glaze should be flocculated with something like sodium silicate.

100-0 Mix	67-33 Mix	50-50 Mix	33-67 Mix	0-100 Mix

Aggressive melt is on the left and blended with lax melt on the right. There is a smooth transition between the properties of the two end glazes. Surprisingly, the "lax melt" is runnier than "aggressive" melt. It ran more and had larger titanium crystals.

Aggressive melt has fine crystals and is matte. The surface is smooth and pleasant to touch but visually bland. The glaze moves little and doesn't run off the tile. The lax melt is more like the original Galactic Indifference but is over crystallized. The larger crystals mean that it melts more aggressively and is fluid at top temperatures.

This is quite strange as lax melt has less flux and less boron. Could it be because of a eutectic? Or caused by poor measuring of ingredients? If so, the most likely culprit would be lithium carbonate as small differences in weight adds large amounts of flux.

2. Ludicrous Melt

I made this glaze to push the boron content to the extreme, with over 50% of the recipe as boron frit. I simultaneously tested the amount of titanium dioxide added. Perhaps 12% is too much: maybe I just need a glaze that's highly fluid and contains just a little more titanium than can be dissolved. In which case, I should adjust the titanium content to the fluxing ability of the base recipe.

It's best to think of this glaze in three parts:

1. Titanium dioxide - The star of the show for reasons explained above. I keep the original recipe's 12%.
2. FF 3195 and Kaolin - I try to maximize frit content to maximize boron. I add kaolin to keep the original glaze's SiO2:Al2O3 ratio of 4.8.
3. Lithium carbonate and calcium carbonate (whiting) - These carbonates are additional sources of flux. I keep the lithium content at around 0.12 molar (see UMF) and add enough whiting to keep the flux ratio at around 0.3 R2O to 0.7 RO.

12 Titanium	9 Titanium	6 Titanium

This is strange. The titanium crystallized in clouds. It looks more like melting butter and is quite rough. All three tests ran off the tile, but didn't have the streaking found in the Lax melt recipe.

These glazes are successfully dissolving the titanium; most typical glossy glazes can't dissolve 6% titanium oxide, resulting in variegation. But here the glass is completely transparent.

The crystals feel nasty to touch. I can see that the crystals are long and needle like with a macro lens, unlike the flatter titanium crystals in Galactic Indifference and the Lax melt test. The only conclusion that I can come to is that the boron is negatively affecting the glaze. For the next test, I'll decrease the boron from over 0.4 molar to the more typical 0.15 molar found in cone 6 glazes.

3. More flux, less boron

+14 TiO2	+ 12 TiO2	+10 TiO2

This is based on the Lax Melt. EP Kaolin and silica are cut and neph sye and whiting is increased. As always, the Silica-Alumina Ratio is kept around 4.8. The titanium is added in excess of 100% to make adjusting titanium oxide easier.

In the electric kiln (which naturally cools faster), the +10 TiO2 test has a more viscous melt and within the glass bead there are large, beautiful titanium crystals. This is what the original Galactic Indifference has. However, the rest of the bead is almost completely transparent. The thicker the melt, the fewer crystals.

4. Altering the crystals

At this point, I can make a glaze that has crystals like the original Galactic Indifference, but the rest of the glaze is still clear. It's possible to dissolve 10% titanium in the glaze, but how the glaze crystallizes during cooling is unsatisfactory.

Why isn't the glaze phase separating due to the high titanium content? Why does it go from clear to large crystals with nothing in between? I think the answer is magnesium.

Since MgO stiffens the melt it can be used simply to check glaze fluidity (in a manner similar to alumina) and to prevent devitrification (the tendency to produce crystalline surfaces). When mixed with CaO, it is not as refractory.

— Magnesium Oxide on Digital Fire

The idea is that thickening the glaze by decreasing the flux content decreases the amount of titanium that can be dissolved into the glaze in the first place but magnesium can flux the glaze when melting and then check the over crystallization upon cooling.

Even more evidence to support this fact is that titanium blue glazes (in which titanium causes phase separation) often have a substantial amount of magnesium. Many glazes fall within the 0.2 to 0.4 molar range. Since I can't add talc (as it would increase the silica content), I can only use dolomite - the result is that I can only max out the MgO content at 0.35 molar.

I'll return to the Lax Melt formula because the increased flux test crazed on my porcelain. Lax Melt also has other desirable properties like good suspension due to its 20% kaolin content. The test is simple. I'll perform a line blend and gradually replace all the whiting with dolomite.

0 Dolomite	5 Dolomite	10 Dolomite	15 Dolomite	20 Dolomite

• Increasing magnesium did in fact alter the types of the crystals
  • White crystals increased
  • Crystal sizes increased
• The test with around 0.1 molar magnesium (5% dolomite) had the best results
  • Fewer yellow crystals, more white crystals
  • Most gloss out of all the tests in the series
  • This is similar to the amount of magnesium in Galactic Indifference
  • Slightly reducing the amount of titanium may produce the best match for Galactic Indifference
• Too much magnesium reduced the ability to dissolve titanium
  • The surface is rough and unpleasant to touch
  • Smaller (sub 100 mesh) chunks of titanium were dissolved in high calcium tests, but not in the high magnesium tests.

5. Tuning the titanium content

Well that didn't work. With the magnesium, decreasing the titanium content didn't increase the crystal size, it just decreased the amount of streaking. Where the glaze pooled, it created rough crystals which are unpleasant to touch.

I'm guessing there isn't enough flux to slowly grow the titanium crystals. The melt needs to be more fluid so only a few nucleation points grow into visible crystals.

6. Reducing the alumina content

The tiles from #More flux, less boron were the only tiles with "low" alumina (0.2 Al2O3). All others had an Al2O3 content of around 0.3. This might be crucial because glazes at 0.25 and lower alumina are fluid enough to create strong visual effects when recrystalizing.

I also try swapping out some of the calcium carbonate with strontium carbonate. This makes it similar to the glaze Strontium Crystal Magic (SCM). The difference between my glaze and SCM is that mine has a lower alumina content and enough boron to melt at cone 6.

• Strontium might
  • Make the glaze matte (mechanism: low silica content, strontium), see Strontium Matte glazes
  • Melt more aggressively
    • The strontium content is only 0.2 molar, which might not be enough to matte.
    • The diversity of fluxes may increase the melt and increase the amount of titanium dissolved.
• Strontium will definitely alter the color response to cobalt and copper, making brighter blues
  • Due to the lithium (which also makes brighter copper blues / greens), copper may create very bright blues and greens

100% Ca	66% Ca	33% Ca	0% Ca

Where the glaze ran off the tile and pooled, there were some occasional 1mm titanium crystals. On the front of the tile, the glaze was a rough to silky matte (with the strontium being silkier) where thin.

Reducing the alumina content did allow the glaze to dissolve more titanium, but only where the glaze pools. The top of all the test tiles were a dry matte. This make sense from a Silica-Alumina Ratio perspective, but why are pieces fired with the original Galactic Indifference in a soda kiln glossy?

7. Increased Gloss

• As learned before, high boron content negatively affects the crystallization of the titanium when the glaze cools. Therefore I'll keep boron low, at the 0.15 recommended for cone 6 glazes
• I don't think lithium is necessary for the crystal effect. Lithium is also quite expensive right now because apparently making electric vehicles is a thing

Low	Med	Hi

The low silica recipe has a Silica-Alumina Ratio of 7.16 and the high version has a ratio of 10.99.

The test tiles were a rough matte no matter what the silica-alumina ratio. The higher silica tiles were smoother, indicating smaller titanium crystals. The increase in silica also gave a bit of a pink blush. This has been observed in the past - titanium can give pinks in conjunction with chrome, similar to the more common tin-chrome pinks.

These have an alumina ratio of 0.24, but don't have the movement and streaking. This means the lithium content is likely crucial for the increased melt.

What does soda firing do, anyways?

My hypothesis is that in a Soda Firing, with large amounts of sodium deposited on the glaze surface, soda itself becomes a glass former. This is evident in how even high alumina clay bodies and slips can become glossy in a soda atmosphere. If the sodium were merely acting as a flux on the clay body, wouldn't it make the surface sealed but still matte? After all, clay bodies and slips used in soda firing have silica-alumina ratios far below the 7-8 value needed for a glossy glaze. Indeed, I've seen a mug glazed with Galactic Indifference that was glossy where hit with heavy soda and matte on the other side.

Therefore, to get the same effect outside of a soda atmosphere, the glaze itself needs increased silica to compensate.

In soda firings, the sodium carbonate (soda) is sprayed into the kiln. This additional soda interacts with the clay and glazes, producing the effects distinctive to the firing style.

Conventional wisdom says that this additional soda behaves just like the soda in sodium feldspars, fluxing the glaze. But is that true? Since the application of soda can be directional, we can observe two sides of the same pot.

If the soda simply acted additional flux for glazes, then that means the side of the pot hit with heavy soda should have a lower silica and alumina content. That should translate to runnier glazes. But that doesn't seem to happen.

Clay bodies can have very low silica-alumina ratios. Flashing slips of a known silica-alumina ratio, 4 for example, will go glossy in soda firings. Why? If the soda simply fluxes the clay body to produce a glaze, the low silica content should keep the surface matte.

The only conclusion I can come to is that in soda firings, the additional soda acts as a glass former. The firing builds up a layer of soda glass on top of the existing clay and glaze. Whether that additional glass can be dissolved into the glaze is another issue.

So what glaze goes over? I think any fluid but not runny glossy glaze should work. Titanium crystals should grow at the interface where the solubilities change.

Boron Test

Maybe it is possible to have titanium saturate glazes that don't over crystallize. The very high boron test could dissolve 6% titanium with no problem, something not possible in typical glazes.

Boron is very hard to crystallize

By controlling the alumina and boron levels, it should be possible to control the size of the crystals.

9. I'm running out of ideas

Ok, so I'm not getting the results I want so I think my understanding of the mechanisms are incorrect. After some digging, I found some similar titanium crystal glazes.

• Opalescence
• Opalescence Rx

These glazes are quite different from Galactic Indifference:

• No lithium
• No magnesium
• High boron (0.1 molar but fired to cone 10)
• Less titanium (0.25 molar, versus 0.37 in Galactic Indifference)
  • Does it even make sense to compare the titanium content to the flux content? Surely the flux ratio changes the solubility of the titanium?
• Lower alumina (0.25 molar versus 0.32 in Galactic Indifference)
• High silica ratio

I hypothesize that the mechanism behind Galactic Indifference is actually simpler than I initially believed.

• Low alumina content
  • Galactic Indifference's flux content is increased during a soda firing
  • Probably needs to target 0.20 to 0.25 alumina
    • Similar streaking glazes live in this area too
• Carefully controlled titanium content
  • The exact content is probably glaze dependent, but it looks like 0.25 is more realistic than 0.35
  • The titanium needs to be just a little over the solubility of the glaze so it crystallizes upon cooling
• Boron to control crystal growth
  • I think boron is necessary to place titanium crystallization in check.
  • Boron allegedly decreases crystallization (devitrification) in glass
  • Boron content needs to be high, but not too high
    • As seen in earlier tests, high boron seems to encourage unwanted needle shaped crystal growth

Created 2024-01-26. Updated 2024-07-13.

Home > Q Science > QD Chemistry > Cosmic Indifference: Galactic Indifference at Cone 6