Abstract

The ratio of a substance's chemical activity to its molar concentration is known as the activity coefficient in chemistry. The activity coefficient, which is 1 in an ideal solution where each molecule's efficiency is equal to its theoretical effectiveness, is a measure of how much a solution deviates from the ideal state.

Some properties of ions in solution are very non-ideal. Even at a low concentration, the activity coefficient is different from one. In basic chemistry, it is common to use concentration or molality instead of activity to calculate equilibria. As long as concentration is small, the activity coefficient is assumed to be one. However, for ions, it can be observed that even at very low concentration, e.g., 0.01 mol/L or mol/kg the activity coefficient is already very different from one (e.g., 0.903 for NaCl). That is why it is important to calculate activity coefficients for chemical equilibrium calculations with ions. Examples of applications are like hydrometallurgy, solubility of metal compounds, scaling, and any crystallisation process of electrolytes. In these cases, it is important to predict the equilibrium and so activity coefficient prediction is needed.
The Specific Ion Theory (SIT) model and Pitzer model are the most popular models to predict activity coefficients of ions in solutions. However, neither of these models account for ion pairing. Ion paring is the formation of dissolved molecules from the ions (e.g., MgSO4(aq)). The purpose of this thesis is to build a model that accounts for ion pairing and predict activity coefficient. The SIT model was used as a basis. The model successfully predicted the activity coefficients of most 1-1 and 2-2 electrolytes to about 1 % accuracy. Including ion pairing improved the accuracy of the SIT model, particularly in the case of 2-2 electrolytes (from about 10 % to about 1 %).