Sodium silicate refractory firing

A range of cheap DIY refractories can be made with sodium silicate as a bonding agent. These can be made with various mineral, metal oxides and even fine soil when mixed with a little sodium silicate (alkaline water glass).

The current post describes the careful but simple drying and ‘ramped-heat-curing’ that can be used to prepare the DIY refractories for their final ‘birth in fire’ during high-temperature firing or their use in very hot stoves or furnaces. “This prevents stuff-ups and forms rock-like, dense, functional refractories and prevents cracking, puffing, blowing, foaming and blistering during use or the subsequent firing step.”

On the other hand, you may be interested in making the insulating, low-density and expanding types of refractories (DIY refractories). In this case, you can use special restricted curing preparation methods. They can, with care and practice, allow you to make expanding or foaming refractories during the sodium silicate refractory firing.

Both high and low density types of refractories can be made and bonded together with more DIY refractory render if required.

Drying

This step should be done slowly, particularly for thick moldings. It simply removes the easily removable water and can be done in the temperature range of 20-130C.

A cooking oven can be used for drying and it is easy to set and even ramp the temperature up as the molding dries.

I begrudge using gas or electricity for this task, particularly if it will take a long time. Consequently, I like to use heat from the sun or the coktop on my wood stove (that burns waste wood) for this step.

With a low fire, I put an inverted can (insulated with layers of newspaper) on the cooktop and put the molding underneath it. To prevent direct contact of the molding with the cooktop, I put a DIY refractory disk on the cooktop and place the molding on it before putting the can over them. I periodically monitor the molding temperature with an infrared thermometer.

The inside of the firebox can also be used. However, it can be tricky to keep the temperature below 150C, above which the curing and blowing conditions can start. Perversely, the carbon dioxide-rich vapour within the firebox can accelerate the curing process.

Candleing

For some small projects, the dual benefits of heat for drying and CO₂ for curing can come from burning candle/s below the moulding with a suitable cover to trap the heat and the CO₂ gas from the wax combustion. It makes a cheap, easy and slow way of drying the refractory and starting the conversion of the sodium silicate to

Curing

This second step should drive off any remaining tightly bound water from the sodium silicate and start the conversion of the silicate into silica that is dispersed between the refractory mineral particles as a bonding agent. It will also harden the refractory. The temperature required is ~200C and ramping up between 150-200C will provide adequate curing. As for drying, a cooking oven, cooktop or inside wood stove can all be used for this step as described above.

Firing

The temperature of this last step is largely dependant on the end use of the refractory. At about 500C the dispersed silica will glass and the surface may develop a smoother texture.

If an excessive amount of sodium from the silicate is in a mix there is a risk that the refractory mix it may slump at high temperatures. This is particularly the case when there is too much low melting point mineral in the mix. I think this is caused by the excess sodium fluxing or softening the glass, as in salt-glazed pottery.

Firing may be as simple as just putting the refractory molding into its high-temperature service in a stove or furnace.

For example, some of my stainless steel or titanium stoves are coated (painted) with layers of refractory render. They must be slowly pre-cured to ~200C. Then the use of the stove simply finishes off the firing process in situ.

As I turn up the heat, the hottest parts (400-600C) change permanently to a dull red colour and the smooth texture show obvious signs of glassing. Other parts that do not get so hot (250350C) just look like the cured refractory, but a chemical change has occurred all the same.

Formal curing is not advisable in most domestic electric ovens because 250C is usually the upper temperature limit.

“If the oven has a circulation fan the high temperature will not be good for the fan. Even 200C is a bit harsh”

Consequently, using the inside of a wood stove (or similar) is a good option. I start with a small fire on one side of the firebox and put the molding on the other side as far away as possible. The fire can be progressively built up to ramp up the firing temperature. Also, the molding can be moved closer to the fire. For a high temperature finish the molding can be put within a big bed of charcoal. This method can also be done outside with a fire pit or drum fire.

A closed chamber of burning charcoal with forced air from a blower is another high-temperature firing option. All these ‘in fire methods’ have the advantage of high carbon dioxide concentration during the process, the benefit of which is described below.

Thin refractory coatings

I have previously described sodium silicate cleaners, primers, foamers and renders that are used in preparing thin protective coatings for metal surfaces such as stoves. These can be finished by the above methods. However, a flame from a simple gas burner can be used to rapidly do the, curing and firing if you can get the flame distance and timing right.  

Chemistry of silicate during drying, curing and firing

Not all sodium silicates are equal. There can be a wide range in the proportion of alkali to silicate. The DIY sodium silicate that I make and use to make refractories is strongly alkaline and has a weight ratio of Silica to sodium oxide (Na20) of about 2.4 (calculated crudely from the weight of silica gel and sodium hydroxide used. It is a concentrated viscous (goopy) fluid that appears to be ready to form crystals. It may be much more concentrated and alkaline than silicate from other sources.

Carbon dioxide curing

During drying and curing the sodium silicate in the refractory reacts with carbon dioxide. Carbon dioxide without heat is used to prepare temporary foundry molds from sand and sodium silicate. The foundry application only uses a small quantity of silicate and they also add another agent to weaken the molding to make the sand easy to remove from the casting.

Enriching the carbon dioxide concentration will also speed curing of permanent refractories as discussed here. Drying and curing in a controlled firebox is excellent in this regard because of the high carbon dioxide concentration in the combustion gases.

I think the reaction can be described this way:

Na₄SiO₄.H₂O+2CO₂=SiO₂ (gel or amorphous dispersed silica polymer or glass] +2Na₂CO₃+H₂O [steam]

The carbon dioxide (CO₂) is a weak acid that can neutralise the alkaline silicate to provide the silica gel-forming conditions. The bound water from the silicate or other entrapped water is heated to make steam in the forming silica polymer and makes it puff or form bubbles.

In the cured and highly dried situation, the sodium silicate is anhydrous (no bound water). The reaction generates no gases and forms no bubbles:

Na4SiO₄+2CO₂=SiO₂ (gel) [or amorphous dispersed silica polymer] +2Na₂CO₃ (no water or steam bubbles)

This description of this reaction may add to mine:

“A large number of applications of silicates are based on the ability of in-situ formation of a silica hydrogel. When silicates react with acid-forming products such as organic esters, the alkalinity of the silicate solution is consumed by the hydrolysis of these esters over an extended period of
time. The gel forms an adhesive bonding with the surrounding substrate (e.g. sand, fly ash, cement and wood).”

Sodium and Potassium Silicates- Versatile compounds for your applications

Refractory thickness and drying/curing time

There is no fixed rule for refractory drying/curing times. For the thin film on stainless steel (above), the careful drying/curing with a gas flame can be done in a few seconds. By contrast, the crabhole soil test puck (above) was dried for several hours and even so it was inadequately dried or cured before firing and it puffed slightly during firing.

Greater refractory thicknesses will slow the heat transfer into the refractory. It also slows the removal of water from the mix and also the entry of carbon dioxide to finish off the chemical conversion of the silicate to gel.

Using the most concentrated sodium silicate ( Homemade DIY sodium silicate, without adding extra water to make mixing easy, is another way of reducing drying/curing time. My post on refractory mixing describes how such low water ‘stiff-mixes’ can be mixed easily.

The experimental stick burner shown below is thick and chunky. It was made with clay free soil and sodium silicate. It was very slow to dry, but I think it would have dried a lot quicker if it had some the paper additive described below.

Refractory with paper and improve drying and much more…….

Lastly, I have added two more refractories to my list of tinkerers ceramic tricks. These are paper clay ceramic (without sodium silicate) and paper soil and sodium silicate- a dirt-cheap ceramic (without clay). They both have significant amounts of paper fibre in the mix.

The paper can be toilet paper (new) or cheaper newspaper (secondhand). “Don’t get them mixed up.

Others report that paper clay in the moist form will develop a bad smell over time. However, I think I have conquered this little problem by adding a trace of quaternary ammonium compound in the form of laundry antifungal/bacterial rinse liquid. It stays sweet! I use this liquid for odourless boots and socks for walking and skiing.

The paper fibre imparts amazing workability and joining properties to the refractories that would be good for novice ceramic tinkering.

Importantly, the paper fibres greatly speed the drying and the evenness of drying of thick refractories with greatly reduced risk of cracking. “It would have been a great addition to the above chunky experimental stick burner.

Lastly, the paper makes the ceramic easy to fire and after it burns away during the firing it leaves a heat-tolerant ceramic that is resistant to thermal shocks.

The proportion of paper in the mix can vary greatly. When the paper proportion is low the refractory is denser, less insulating and strong. I also expect it will be less resistant to thermal shock than the mixes with more paper. Conversely, when there is a high proportion of paper the refractory is lighter, more insulating, somewhat weaker and better able to withstand thermal shock. “This shock resistance can be tested by dropping the red hot refractory into a bucket of cold water if you have the courage!

“Looking back into ceramic history this should not really be not a surprise. As a research colleague used to tell me: “Tim,…… Tim,…… Tim….. (looking over his glasses). Nothing is really new, it has just been forgotten.” Baldosa has been made this way forever for building houses, stoves and ovens in most subsistence communities around the world. The organic fibre that is used may vary. We have just swapped the cow pooh with toilet paper.

Dispersed amorphous silica gel flexibility

The alkaline silicate in the refractory mix is converted into silica gel in the cured and fired refractory. We mainly experience silica gel as hard rigid beads or crystals of drying material that is used for example in packing or kitty litter. Its physical properties are far removed from quartzite crystals, but it is chemically identical (SiO₂).

Similarly, dispersed silica gel that is precipitated in the refractory matrix is different again and the resulting SiO₂ polymer has amazing flexibility.

This flexibility is most clearly demonstrated with the thin film of refractory that was applied to stainless steel foil shown above. The coating stays intact when the foil is curved. It also stays intact when the foil is repeatedly heated to temperatures of 500+C which is the peak operating temperature in my tiny backpacking tent stoves. With this change in temperature, stainless steel changes in length by about 0.8%. “A terrible metal property in my preferred stove building material and titanium is not better.” Nevertheless, the refractory persistently hold on!

Conclusion

It is interesting to think that my silicate refractory liquid starts as cheap crunchy and rigid ‘crap adsorbing kitty litter’ crystals. They are dissolved in ‘caustic soda drain cleaner’ to become an alkaline liquid sodium silicate. That it can be mixed with all sorts of soil, minerals grits, metal oxide powders and even dunny paper. Then it can be dried and neutralised by carbon dioxide during curing and heating. Finally to go back to silica gel, but in a new flexible dispersed polymer form that turns the minerals back into a rock that can ‘laugh at fire’.

Addendum:

I have subsequently found a reference about the need for systematic drying before high-temperature treatment in The oxychem sodium silicate handbook.

Tim