SI for brucite from PhreeqC interactive using wateq4f, and T dependency of logK

[zaki]

Hello all!
I have two questions related to solving some speciation in seawater with PhreeqC interactive 3.7.3(15968) using wateq4f database.

Q1. Is SI for Brucite in solution calculated correctly?
In wateq4f database the reaction related to Brucite/Mg(OH)2 is written as
Mg(OH)2 + 2H+ = Mg2+ + 2H2O with H0=-27 Kcal/mol and logK=16.84
From this info, we have -log(Ksp) for Brucite = 11.16.
I use PhreeqC interactive, where after an aqueous speciation the IAP for Brucite is correctly evaluated and reported as [Mg2+][OH-]2. However, the SI for Brucite is evaluated as log(IAP)-logK , instead of log(IAP)-log(Ksp), which I believe is not correct.
Could anyone please clarify?

Q2. How is the temperature dependence for logK evaluated?
This question is related to the evaluation of the K value of the same neutralization reaction at temperatures other than 25C. H0 is reported  on the database as a single value (-27 Kcal/mol) without giving any relation for temperature dependence. On aqion website, it is mentioned that, in such cases, van’t Hoff with constant enthalpy is used. If we assume constant enthalpy, with the aid of Van’t Hoff we have for logK, 17.62 @ 20C, 19.26 @ 10C. However, the output file for speciation evaluates 17.18 @ 20C and 17.89 @ 10C.

[dlparkhurst]

(1) The log Ks defined in all databases are thermodynamic equilibrium constants, not solubility constants. The saturation index is calculated as

Code: [Select]

Mg(OH)2 + 2 H+ = Mg+2 + 2 H2O
SI =  log[ (Mg+2)(H2O^2) / (H+)^2] - log(K)

where parentheses indicate activity.

(2) The calculation for log K of brucite with data from wateq4f.dat is as follows:

Code: [Select]

log10(K) = 16.84 - (-27.1) * (298.15 - 283.15) / (2.303 * 0.00198726 * 283.15 * 298.15) = 17.89

where 0.00198726 is R in kcal per degree mole.

[zaki]

Thank you so much! I inserted the reference logK value in van't Hoff equation without converting that to lnK, and then left the result alone without converting it back to logK. The difference was too small to suspect a mistake.