[Twigg]

I've been told that activity is a very empirical expression used to correct for deviations. Are the causes of deviation known? I'd think that if these deviations could be corrected for with a coefficient (the activity coefficient), then they'd have to be pretty regular and of an obvious cause. What would be the conceptual explanation for them (if it exists)?

If the answer is that the meaning of activity varies depending on the species, could you explain what activity would theoretically represent in a few cases? If you could explain it for HCl at least that would be great.

I have a feeling that the answer is going to be that it just fits the data. Unfortunately, I'm a realist and I believe that every quantity has a physical meaning. 

I came up with two explanations I'll share to show I have tried to think about it myself (I just don't know what's the accepted view). For HCl and other gaseous strong acids, I thought it may reflect some amount of unionized HCl in solution, detracting from the expected ion concentrations if you assumed that it would all dissociate. The problem with this is that if that were the case, the mathematical expression for activity would use an equilibrium expression, not a linear function (a=y*C). My second thought was that the activity reflected an effective or available concentration, kind of like the limits of the rates of enzyme catalyzed reactions with ever increasing concentrations of substrate. In other words, there would be so much ion in the area that it would make less and less of a difference if you added more because the frequency of collision was maxing out. Thus, you might say there's a function of concentration (C) expressing activity (a) whose derivative is constant (y) on the normal interval but right of the normal interval the function has a negative second derivative. The problem there is that the activity coefficient (y) is supposed to be a constant for all concentrations (but obviously that isn't true here since the first derivative must decrease as concentration becomes extreme).
Last thing I tried was to conjecture that some portion of the solute dissolves, and the rest enters a suspension. That works pretty well for salts, not so well for gases like HCl.

Any help?

[Borek]

Start googling Debye-Huckel theory, with some luck you will find pages where other, more advanced approaches will be discussed. Trick is, he more advanced approach, the m ore it relies on experimental factors.

For charged species activity coefficients can be calculated from electric interactions between ions. It works OK for diluted solutions, it stops to work for more concentrated ones. That's because there are many other interactions (not only charge/charge, but also dipole/dipole, charge/dipole, dipole/quadrupole and so on) present, plus there is plenty of interactions that are more "chemical" than electrostatic (like hydrogen bonds creation).

The problem there is that the activity coefficient (y) is supposed to be a constant for all concentrations

Is it? As far as I know, it is not.