Well, upon firing, the "muffin", though looking great straight out of the kiln, crumbled when handled.
The supplier of the silicon assures me it is not tungsten, that he would not make that mistake as tungsten is more expensive. Looking at other sites/companies selling silicon the stuff does resemble very high purity silicon.
I tried, however, putting some in a very strong sodium hydroxide solution and it did nothing. No reaction. It is supposed to release hydrogen gas. Other online references confirm this.
To make sure the drain cleaner was not to blame, I dropped in a small bit of aluminium foil, which should have a similar reaction. That "boiled" like crazy and the reaction went on and on and on and on until the aluminium was totally dissolved.
The patent does mention something about silicon normally requiring some kind of acid wash to remove oxidation but his formula supposedly does not require that.
The patent also mentions he used impure silicon.
The silicon I got is an extremely fine powder. Perhaps it is easily oxidized. Don't know.
But, if silicon is less than 100% reliable at making hydrogen bubbles, why not use aluminium if that is the only purpose of the silicon; to make hydrogen bubbles?
Re-reading the patent, that would appear to be the case:
When mineral foam compositions are made according to the process set forth above, expansion of the mineral foam is believed to result from the reaction of silicon with the aqueous solution of alkali metal hydroxide, thereby generating hydrogen gas. Because this reaction is exothermic, the temperature of the expanding foam composition is significantly increased, and water vapor or steam is driven from the composition together with some escaping hydrogen gas.
While it is well known that hydrogen can be highly explosive when combined with oxygen or air, the wet gas liberated from the compositions produced according to the formulations and process disclosed herein is not believed to present an explosion hazard, and in fact burns with a low flame when ignited.
The aqueous sodium silicate, as used in the disclosed process, aids in retaining the gas bubbles generated within the composition without overly limiting expansion. Then, once expansion is complete, the bonding properties of the sodium silicate are believed to contribute substantially in achieving a preliminary "set" that will, for example, permit forms to be stripped away from the expanded material.
Where the mineral foam compositions have, either through overexpansion or through excessive application, exceeded the boundaries of the forms or containers into which they are poured, excess material can be easily trimmed away by means of a common hand saw or the like.
Luckily, I think, hydrogen dissipates rapidly and is virtually impossible to contain, so I don't think exploding displacers would be a problem.
I think, though, I might give the general methodology another go, but substituting aluminium powder in place of the silicon.