Unwanted Calcium Carbonate in Refractory Cement

[Matcon]

I have noticed that most castable refractories effervesce if mixed with acid. In context, this is orthophosphoric acid added to promote the formation of a calcium-aluminium phosphate bonding phase. In one case, I estimated that 40% of the calcium aluminate cement had become carbonated by reaction with the atmosphere. Suppliers don't see it as a problem

It is impractical for me to roast large quantities of refractory to drive off the carbon dioxide, so I considered chemical ways of decarbonation that would not ultimately weaken the refractory and found the problem much more challenging than expected.

Treatment with formic acid vapour decarbonates OK but the formate ions don't contribute to strength. The same applies to all the obvious anions. Chemically, chromic acid might work but the environmental hazards don't bear thinking about.

 

[Puma]

Oxalic acid - calcium oxalate is a hard substance? I am not sure what percentage conversion you will get from the reaction.

 

[Matcon]

@Puma
Thanks for the contribution. The percentage conversion was the reason I originally rejected oxalic acid but your post made me re-think.

The reason calcium carbonate is unwanted is that the converted calcium is no longer available for the hydrolysis reaction that results in the cement setting. In an extreme case, a calcium aluminate particle could be completely surrounded by carbonate and unavailable for hydrolysis, with obvious consequences for refractory strength.

The reason oxalic acid may work is the difference in density between the carbonate and oxalate (relative density 2,7 down to 2,2). This expansion may be sufficient to promote physical interlocking between adjacent particles. If the oxalate layer cracks and allows the oxalic acid to penetrate to unreacted carbonate, the reaction may go as far as completion. It's certainly worth doing a gravimetric experiment.

A variation on your suggestion, if the oxalic acid alone doesn't work, would be to vapour-phase treat the cement with formic acid to convert the carbonate to soluble calcium formate and add oxalic acid to the water when working the refractory. The oxalate would precipitate in the interparticle space, hopefully in sufficient quantities to form a bond. This could be further enhanced by adding a soluble metal oxalate instead of oxalic acid. For example, ferric oxalate is soluble in water, while ferrous oxalate is not. Precipitation of two oxalates (calcium and ferrous) yields more solids in the interparticle space, which is what you want for a bond.

 

[ashember]

The major difference is in solubility, calcium carbonate is almost soluble in water vapour, to insult it.
Oxalate, on the other hand should be very, very stable. How about adding magnesium compound into the reaction?
I mean: prevent CaCO3 from even forming, magnesium compounds have insanely low solubility due to being Mg2- compounds in reality.

remember: oxalic acid: two bonds, anything that bonds to it 1:1 will be insoluble.

Ideally, you would want organic somewhat-stable oxalate, like ester for example, so that the oxalic acid is released after you cast your blocks. etc...

 

[Borek]

The major difference is in solubility, calcium carbonate is almost soluble in water vapour, to insult it.
Oxalate, on the other hand should be very, very stable. How about adding magnesium compound into the reaction?
I mean: prevent CaCO3 from even forming, magnesium compounds have insanely low solubility due to being Mg2- compounds in reality.

remember: oxalic acid: two bonds, anything that bonds to it 1:1 will be insoluble.

Ideally, you would want organic somewhat-stable oxalate, like ester for example, so that the oxalic acid is released after you cast your blocks. etc...

Sorry to be blunt, but half of that is a pure nonsense. The other half is less pure, but still doesn't make much sense.

 

[Puma]

Fascinating matcon. I hope it works out.

 

[Matcon]

Thanks to everyone for their contributions; they helped me to reformulate the problem and come up with some new thoughts.

If the problem is reformulated to "How to reduce carbonate in a solid without water?" with the proviso that the product of the reaction must either be soluble in water, hydrolysable or polymerisable (you can't form a bonding phase without at least one mobile species).

Formic acid, or any reducing vapour, as mentioned previously, is one solution.

Another possibility is to use a solid reductant. What makes this reasonable is that it is simple to screen out the fine cement phase, mill in the reductant and remix with the aggregate. Among the more readily available reductants are calcium carbide, silicon carbide, aluminium metal powder and silicon metal powder. Calculating the free energy associated with the reaction of each reductant suggests that, while the mixtures would still need heating to drive off carbon dioxide, the temperatures required would be much lower.

Silicon carbide looks particularly promising because the reaction starts at 300°C and, if the mixture is right, the product is C3S, one of the hydrolysable products found in Portland cement.

Even more promising is aluminium carbide. Theoretically, that reaction starts at 100°C. Of course, the bugbear with free energy calculations is the reaction kinetics. Fortunately, the ductility of aluminium may come to my rescue. With high energy milling it is conceivable that a sufficiently intimate mixture of cement, aluminium powder and lampblack would produce the desired result, especially if the local temperature were to exceed the threshold for the reaction to proceed.

The oxalic gravimetric experiment is under way. The oxalate is drying over the weekend. I was surprised by the amount and duration of effervescence and am looking forward to Monday's weighing.

 

[Matcon]

Just in case anyone thinks I overlooked the obvious solution of topping up the calcium aluminate content of the refractory, I didn't. I suspect that the root of the problem is the calcium aluminate, as supplied. The quality may not be up to standard and/or it may not have been stored properly and/or it may carbonate more quickly than it should and/or it is stored for too long.

I tested Portland cement with acetic acid and found that it didn't effervesce.

 

[Puma]

My knowledge is very basic compared to yours but I had a look around and thought this might inspire some new ideas, it shows how the calcium carbonate can be reacted in a dry state in a dry state at 200 degrees c. I don't know if it would be useful in the production of screeds/ water proof renders... Anyhow it might be interesting for you!

https://www.google.co.uk/url?sa=t&r...NrXb85_UdhOR5xg&bvm=bv.90491159,d.d2s&cad=rjt

 

[Matcon]

@Puma
Thank you for yout continued interest in this subject. Your link certainly added a new dimension: oxidation of carbonate rather than reduction. The product, calcium chloride, is not particularly desirable because it doesn't have the desired physical properties but it could be converted to calcium silicate or calcium aluminium silicate by addition of sodium silicate. I know of a case where a refractory with an anorthite binder phase (melting point approx. 1550°C) handles a ferrous alloy at over 1700°C.

The small amount of sodium chloride co-product could be allowed to vaporise in service, provided suitable precautions were taken.

A fluoride analogue might be more useful, as calcium fluoride melts at just over 1400°C. That would be suitable for a refractory rated for service at 1350°C and maybe even a bit higher.

Another possibility would be to go for carbo-chlorination that would proceed at even lower temperatures but may be less selective and chlorinate the remaining (uncarbonated) calcium aluminate.

It's definitely worth following up. My intuition tells me there's something waiting to be found down that road.

At risk of being criticised for going off topic, I am delighted to see 1960s chemistry being applicable to 21st century problems. I have huge respect for work done around that time and have drawn on it many times to solve problems.