1. Jerzy Kosinski and Andre Anderko
Modeling Silicate Chemistry for Water Treatment Applications Phase I: Solubility Behavior
Jerzy Kosinski and Andre Anderko
OLI Systems Inc.
March 9, 2017

2. Objective
Predict the formation of silicates in short- and medium-term experiments
Ensure that the predictions are consistent with long-term behavior of minerals
Based on a combination of Sandia report and literature data
Scope
Calcium silicate chemistry
Magnesium silicate chemistry
Zinc silicate chemistry
Zinc-calcium silicates
Aluminum silicate chemistry

3. Calcium silicate chemistry: Metastable phases
Solubility map: SiO2 vs. CaO
Key phases: amorphous tobermorite and afwillite
Predicted solubility of amorphous tobermorite and amorphous afwillite is consistent with Sandia experiments
NaCl and NaOH shift solubilities (cf. thin and dashed lines), consistent with literature data (Glasser, 2005)

4. Calcium silicate chemistry: Metastable phases
Solubility map: SiO2 vs. CaO
Lower temperature: 70 C
Key phases: amorphous tobermorite and afwillite

5. Calcium silicate chemistry: Metastable phases
A different solubility map: SiO2 vs. pH
Key phases: amorphous tobermorite and afwillite

6. Calcium silicate chemistry: More stable phases
Over a longer period of time, more stable phases form
Key phases: crystalline tobermorite and afwillite
Solubility map: SiO2 vs CaO
Numerous data are reported in the literature for C-S-H phases (not necessarily identified), partially dependent on the adopted procedure (hydration, precipitation)
Model identifies the phases

7. Calcium silicate chemistry: More stable phases
A different solubility map: SiO2 vs pH

8. Magnesium silicate chemistry: Metastable phases
Complication: pH and complexants strongly shifts the lines
Solubility isotherms are simpler in the SiO2 vs. log(Mg)+2pH variables, which correspond to dissolution equilibria
Key phases: amorphous sepiolite, amorphous antigorite and poorly crystalline antigorite
Poorly crystalline antigorite is based on data of Gunnarson et al. (2005)
Crystalline phases are included for comparison

9. Simulation of Sandia’s experiment F
Ca(OH)2 is added stepwise to artificial geothermal water
Assumption: amorphous calcium silicates (tobermorite and afwillite) precipitate fast
Precipitation of amorphous tobermorite is predicted after the addition of 0.3 g Ca(OH)2 per liter of solution, followed by precipitation of amorphous afwillite

10. Simulation of Sandia’s experiment F - continued
Ca(OH)2 is added stepwise to artificial geothermal water.
Assumption: amorphous calcium silicates (tobermorite and afwillite) precipitate fast
Precipitation of amorphous tobermorite is predicted after the addition of 0.3 g Ca(OH)2 per liter of solution, followed by precipitation of amorphous afwillite.

11. How to simulate non-equilibrium behavior using a thermodynamic model?
Sandia experiments A and B
Ca(OH)2 is added to artificial geothermal water. Mg(OH)2 (A) or MgO (B) is added in the next step
Assumption: amorphous calcium silicates (tobermorite and/or afwillite) precipitate fast, while magnesium silicates precipitate slowly
Formation of layers of magnesium silicates on the grains of added Mg(OH)2 or MgO is proposed as a possible mechanism of separating the solution from a part of the added magnesium compounds.

12. How to simulate non-equilibrium behavior using a thermodynamic model?
Sandia experiments A and B continued
Gradual formation of different phases explains the change in dissolved SiO2 and MgO concentration

13. How to simulate non-equilibrium behavior using a thermodynamic model?
Sandia experiment I
MgO is added to artificial geothermal water. Ca(OH)2 is added in the next step
Assumption: magnesium silicates precipitate slowly
Formation of Mg(OH)2 limits the availability of magnesium in the first step of experiment
Predicted solubility of magnesium hydroxide is very low at the high pH of the solution, so the availability of magnesium can be much lower after the addition of Ca(OH)2.

14. Zinc silicate chemistry
Amphoteric character
Various silicate phases: hemimorphite (Zn4Si2O7(OH)2.H2O), willemite (Zn2SiO4), layered 1:2 (Zn2SiO4.2H2O) and 1:1 zinc silicates (ZnSiO3)
Solubilities reported by McPhail et al. (2006) are consistent with Zn2SiO4.H2O and Zn2SiO4.2H2O phases

15. Zinc silicate chemistry
Solubilities reported by Palmer and Anovitz (2009) are consistent with predicted incongruent solubility of willemite with precipitation of ZnO except for one point

16. Zinc silicate chemistry
Data of Brady et al. (2011): zinc nitrate and sodium hydroxide are added to artificial geothermal water.
Massive precipitation of zinc hydroxide is expected
Reported zinc concentrations are consistent with those predicted for zinc hydroxide
Reported silica concentrations are bracketed by those predicted for Zn2SiO4.2H2O coprecipitating with Zn(OH)2 and those predicted for Zn2SiO4.1H2O coprecipitating with Zn(OH)2

17. Status Calcium silicate and magnesium silicate chemistry
Modeling has been finalized
Parameters will be included in the upcoming release (9.5.3)
Report is available
Zinc silicate chemistry
Modeling is being finalized
Aluminum silicate chemistry
Work is in progress