binary interactions in SIT model

[Laurent]

Hello,
We are currently adjusting an aqueous model using the SIT model and sit.dat. The system contains uncharged species, and we did not manage to fit our experimental data (speciation and water activity data), unless adding a SIT interaction between an uncharged specie and OH- (as we would do with the LAMBDA parameter of the Pitzer model). It is specified in the SIT keyword data block description that only cation-anion interactions should be used, but it seems that calculations are running well with an uncharged-anion interaction. Do you think it's safe to proceed with such additionnal SIT interactions?
Thanks in advance,
Regards,
Laurent

[dlparkhurst]

It's been too long since I looked at the theory of SIT. In the end, it seems like it is just a model, and if it fits better, that is a justification. My guess is that it is not quite thermodynamically correct.

[Laurent]

Thanks David for this feedback.
From thermodynamic point of view, it seems OK to add such interactions (see OECD_NEA guidelines from I. Grenthe 2013), as long, as you said, that it improves the simulation. I was mainly wondering if the calculations performed by Phreeqc are correct if we add such interaction in the SIT datablock. From our current tests it seems ok.
Regards,
Laurent