SIT Model - 2:2 electrolyte interaction parameters?

[Calcorn]

Hi all,

I've been using PHREEQC with the SIT model to calculate some solution concentrations, but I have no idea what the program is using for interaction parameters that aren't in the database, specifically, any 2:2 ion interaction coefficients.

From the SIT.DAT file:

Data gaps for estabilty, entropy/enthalpy and SIT coefficients have
# been filled in by estimations. More detailed information about some
# of the methodologies used for data estimation can be found in the
# following references. Three main estimation methods have been
# followed for SIT interaction coefficients: a) by analogy to similar
# compounds, b) by correlation with "epsilon" of the same complex but
# with other ionic media (nitrate or perchlorate) and c) by using the
# approach of (Ciavatta, L.. (1990)

As an example, I see epsilon(FeOH+2, ClO4-) = 0.38, but what would PHREEQC use for an interaction parameter between FeOH+2 and say, SO4-2?   From above, (a) I don't know what similar compounds there would be, (c) this paper seems to apply to only interaction parameters of neutral metal complexes with 1:1 electrolytes, so I'm guessing the answer lies in (b). 

I'm thinking that PHREEQC takes the known epsilon for the 2:1 interaction, and scales it by z² for the 2:2 interaction, but I have no idea.  It also remains an option it assumes that the 2:2 is zero (though I'd hope not).   I've dug through all the documentation I could find and have come up short.   Anyone know the answer here?

Cheers,

Chris

[dlparkhurst]

I think only the defined interaction parameters are used. Others would effectively be zero.

[Calcorn]

Thanks for the reply.  I've got no reason to believe that's not correct, but I'd certainly hope that this isn't the case - imagine the errors!

The text in the SIT.DAT file, "b) by correlation with "epsilon" of the same complex but with other ionic media (nitrate or perchlorate)" seems to hint at that they do some sort of estimation of unknown interaction coefficients, but I can't find out what that estimation might be.

[dlparkhurst]

I will say more definitively, PHREEQC does not estimate any parameters. It may be that the people that generated the SIT database used correlations to estimate some of the parameters, but PHREEQC only uses parameters the parameters in the sit.dat database to calculate activity coefficients for a solution.

[Calcorn]

Thanks again for the reply.  I suppose some sort of algorithm to estimate interaction parameters would cause more problems than it'd solve.  From the volume of literature I've read, I also suppose anyone using SIT has a large concentration of a 1:1 electrolyte in solution, and any unknown parameters can be assumed zero because of the relatively low concentrations.

Looks like I'm in the market for a new model to use!