Some multivalent metal ion is necessary for a sodium silicate cement to cure. In industry and antiquity, copper and calcium have both been used. I hear that the Egyptians used a variety of copper accelerated sodium silicate techniques (glazes, formed objects). Calcium accelerated sodium silicate cements are discussed later. Both of these reactions produce free sodium hydroxide.
Aluminium is an ideal candidate because in pH 14 solution, and with available alkali metal ions, in the presence of silica, aluminium is capable of precipitating in four-fold coordination with oxygen. This is a more tightly packed molecular arrangement, and the pair (sodium/aluminium) is closer to the oxygen. the bond, being shorter, becomes more covalent. In this system, Alumina and silica will co-precipitate as a covalently bonded polymer. Solids of all other metal/acid systems (metal: copper, calcium, sodium. acid: sulfate, chloride, phosphate, silicate) like copper silicate, sodium carbonate, calcium sulfate (gypsum), and calcium silicate which have a predominantly ionic/crystalline character. (see Pauling, “nature of the chemical bond”).
In the zeolite industry, sodium aluminate or meta kaolin clay are used as sources of alumina. Meta kaolin is raw kaolin (porcelain) clay that has been strongly heated in a kiln. For geopolymers, kaolin is superior to sodium aluminate because it (1) does not add excess alkali and (2) has alumina and silica already intimately mixed (and dissolves as an alumino-silicate monomer). In the case of kaolin, only alkali is necessary for polymerization because both silica and alumina are present (see discussion of Na-PS). All naturally occurring aluminosilicates may also be used, provided that they can be destabilized so that dissolution can occur (not always possible, or reasonable).
Many natural aluminosilicates can be destabilized by calcination (strong heating) at between 500 and 800 degress celcius (the same calcination temperature range used to convert limestone to quick lime) or higher depending on the starting material. My experience has been with kaolin which I calcined at 750 C for 6-12 hours in a kiln. A large flowerpot is my crucible. An unadorned pot must be used, as ones with designs on them have been slip cast and will not survive being refired. Calcined kaolin clay is called meta kaolin (see the discussion on Na-PS for more about calcined kaolin), and kaolin calcined at 750 C is termed MK750 by Davidovits (”Geopolymer Chemistry and Applications”, an essential book which I highly recomend reading. I have read most of the sections four or five times).
It is also possible to destabilize kaolin, and possibly other aluminosilicates, by leaching them with a dilute alkali solution for 24 hours. I have tried this several times and been unsucessful, but I will relate what I know about it. One group (^1) used .1 M (4g NaOH per L water) to make a watery slip of kaolin, which they stirred for 24 hours, and then washed the excess alkali out with several changes of pure water. They then must have dried the slip, probably by the same methods ceramicists use to recycle dried clay, though this was not related to me in communication. Finally, this clay was combined with the appropriate sodium silicate solution and allowed to cure at room temperature. Apparently, the product was crystaline rather than amorphous as metakaolin based GPs are. For my experiment, I used .5M in a thick slip that I let sit for 24 hours (did not stir except two or three times). I did not wash out the alkali. When combined with an appropriate sodium silicate solution, the sample did not cure. I intend to repeat the experiment more precisely, now that I have all the information.
(^1) C. Nicholson, B. Murray, et al. Towards an understanding of the synthesis mechanisms of geopolymer materials. Geopolymer 2005