stoichiometric coefficient of the KINETICS data block

[sama.k]

Hello;
I am trying to simulate alite or C3S (3Cao.Sio2) hydration reaction kinetically using KINETICS data block. According to PHREEQC manual, stoichiometric coefficient defines the mole transfer coefficient for formula per mole of reaction progress (evaluated by the rate expression in RATES). The product of the coefficient times the moles of reaction progress gives the mole transfer for formula relative to the aqueous solution; a negative stoichiometric coefficient and a positive value for reaction progress gives a negative mole transfer, which removes reactants from the aqueous solution.

Alite or C3S (3Cao.Sio2) hydration reaction: 2C3S (s)+ 6H (l)= C3S2H3 (s) + 3CH (s)
(H=H2O, CH=Ca(OH)2, C3S2H3= defined as CSH solid solution)

Is my approach (about the definition of kinetic reactions and especially stoichiometric coefficients) correct in the below code?

Any help or advice will be greatly appreciated.
Many thanks

Here is the partial input data for my model:

Code: [Select]

KINETICS 1
C3S
    -formula  (CaO)3SiO2  1 CSH  -0.5 Ca(OH)2  -1.5
    -m        0.2912
    -m0       0.2912
    -tol      1e-008
-steps       8640000 in 100 steps # seconds
-step_divide 1
-runge_kutta 3
-bad_step_max 5000


[dlparkhurst]

I think it might be OK. I'm a little concerned that CSH is ambiguous in that it could mean a "CSH" phase defined in PHASES or it could be a formula 1 carbon, one sulfur, and one hydrogen. I think you should define the PHASE as "CSH(s)" and use it in -formula to be clear. Alternatively, you could define -formula with true chemical formulas (stoichiometric combinations of Ca, O, H, Si, etc, rather than using PHASE names).

Note, the -formula in KINETICS would not contain a true SOLID_SOLUTION (variable composition), it must be a fixed composition of elements.

I should also add that only the net reaction defined in -formula is used in the calculation. So, if you are conserving Si in your reaction, that is dissolving from C3S and identically precipitating in C3S2H3, Si would not appear in the net formula, and its concentration in solution would not change. In fact, provided

2C3S (s)+ 6H (l)= C3S2H3 (s) + 3CH (s)

balances for Ca, H, O, and Si (you should probably write the reaction with the explicit chemical formulas), no element concentration would change except through the consumption of water. The net reaction is simply -formula H2O -6. Your reaction would have little impact on the solution composition. The amounts of each phase would not be explicitly accounted for; the amounts would be calculated from how much of your kinetic reactant [KIN("C3S")] remains.

Another way to approach the problem would be to dissolve C3S kinetically (-formula C3S 1), and include C3S2H3 and CH in EQUILIBRIUM_PHASES. This would explicitly track the amounts of each phase by KIN("C3S"), EQUI("C3S2H3"), and EQUI("CH"), plus the solution composition (pH, concentrations of elements) would vary as the C3S dissolves and the other minerals precipitate. Another wrinkle would be to make C3S2H3 and CH kinetic reactions instead of equilibrium reactions.

[sama.k]

Thanks for your comment

All the intended phases such as (CaO)3. (SiO2) or C3S, (CaO)3. (SiO2)2.(H2O)3 or C3S2H3 and  Ca(OH)2 or CH  are defined in the database, but the hydration reactions (E.g. 2(CaO)3. (SiO2)  + 6H2O = (CaO)3. (SiO2)2. (H2O)3  + 3 Ca(OH)2) and other intended chemical reactions of my model aren't considered in the database. Should I define hydration reactions using SOLUTION_SPECIES keyword block, If I want to dissolve C3S kinetically (-formula C3S 1), and include C3S2H3 and CH in EQUILIBRIUM_PHASES (as you suggest)?

[dlparkhurst]

Sorry, I don't know enough to help you decide on a modeling approach for concrete minerals.

If you are looking for overall equilibrium among hydrated and unhydrated phases, then I think you would need to define log Ks and reactions for all of the phases in PHASES (not SOLUTION_SPECIES). If you want to consider SOLID_SOLUTIONS, you would still need PHASES definitions for the component phases.

Conceivably, you could model hydration as simply a kinetic reaction unrelated to thermodynamic equilibrium, in which case, perhaps you do not need equilibrium constants.

I'm going to stop here because I don't know enough to give you a reasonable modeling approach.