Normally I take the contribution of the silicate ion to the overall electrolyte balance to be unimportant. However, for some samples, the presence of significant amounts of silica is revealed following the acid dehydration steps employed in the work-up of samples for Ca/Mg analysis. Wall-wash samples impacted by geothermal outflows often show elevated levels.
For the purposes of this assay I report silicate as SiO2 mole equivalents, but I factor these amounts into the overall electrolyte balance as the ion [SiO4]4- . Given the tendency of silica to form poly-silicates, this treatment should be considered approximate only.
The choice of a suitable standard is likewise complicated by the fact that any alkali metal silicate is going to contain poly-silicates to some degree. I settled on a commercial sodium metasilicate that was claimed to be of high purity. My initial assumption was that it would be Na2SiO3.9H2O. However, dehydration using concentrated hydrochloric acid to form SiO2 showed that the material was more likely to be the anhydrous metasilicate. The stoichiometry of this dehydration step was far from perfect, and consequently I settled on the my experimentally determined SiO2 content as the basis for the standard value to be used in subsequent determinations.
The method uses the classic silicomolybdic acid spectrophotometric acid method. I followed the recipe of . These authors utilize a combination of two reducing agents, sulphite plus an aminophenol, to reduce the silicomolybdic acid to molybdenum blue. Colour development, which has a reputation for being inconsistent, proved to be very robust using this method.
- 2g ammonium molybdate tetrahydrate
- 6ml concentrated hydrochloric acid
- make up to 100ml with de-ionised water
- store in polyethylene bottle to avoid silica uptake from glass
- 5g oxalic acid (suppresses interference from phosphate)
- 1.7g 4-methylaminophenol sulphate
- 1.0g anhydrous sodium sulphite
- dissolve in 125ml of de-ionised water
- add 25ml concentrated sulphuric acid
- make up to 250ml with de-ionised water
- use 1ml aliquot of sample
- add 16ml de-ionised water
- add 1.5ml of reagent A
- stand 10mins for the conversion to silicomolybdic acid to occur. This time should be standardised because it strikes a balance between converting all monomeric silicate, while leaving such higher oligomers that might exist, unconverted.
- add 6ml of reagent B
- make up to exactly 25ml with de-ionised water
- stand for 4 to 6 hours. The time is not critical as the colour is very stable.
- measure the absorbance at 810nm against de-ionised water.
The recipe described above provides a useful analytical range of 0 to 50ug of SiO2, or 0 to 50ppm for the 1ml sample aliquot.
The response is linear and normally just one standard and blank is all that is required for calibration.
Method used to establish SiO2 content of commercial sodium metasilicate:
This method is one used to estimate SiO2 content of Portland cement.
- weigh-out in duplicate approx. 1g of the metasilicate
- add approx. 1g of ammonium chloride
- add approx. 10ml of de-ionised water
- add 2ml concentrated hydrochloric acid
- warm on steam-bath at 90ºC for 3 hours
- filter through Whatman #42 using vacuum
- wash x5 with 0.05M hydrochloric acid
- reduce to ash in a porcelain crucible and then heat at red heat for 3 hours
- heat to constant weight
The SiO2 recoveries obtained, although in agreement with one another, were higher than that expected if the starting material was in fact pure Na2SiO3. This is not altogether unexpected given the tendency of alkali metal silicates to be polymeric to some degree. Despite the composition of the metasilicate being uncertain, we can proceed to use it as a secondary standard, based on the amounts of silica recovered.
SiO2 recovered, sample A: 57.7%
SiO2 recovered, sample B: 57.5%
SiO2 expected for Na2SiO3: 49.2%