So far, individual activity coefficients of ions could not be measured directly in electrochemical experiments. Their knowledge is important, however, for the solution of various physical chemical problems, and is particularly of high relevance for understanding and estimating the pH of seawater and related climatological studies. In this paper we present a summary of available statistical methods of electrolyte theory, and their applications to simplified seawater models for the theoretical calculation of individual activity coefficients of ions in aqueous solutions. We start with exact results for the model of hard charged spheres obtained from the statistical cluster expansion theory. Then we develop consistent analytical approximations such as generalized Debye-Hückel (DHX) and Mean-Spherical approximations (MSA), as well as the Justice method for including weak ionic association effects (DHXJ). These approximations are consistent with the exact statistical theory but still sufficiently simple to allow calculations on standard personal computers. A simplified seawater composition model is suggested for this purpose. The main adjustable input parameters of this model are the mutual contact distances of ion pairs, from which the individual ion activities can be derived. We show that our new results, exploiting parameters from numerical Monte-Carlo (MC) and Hypernetted-Chain (HNC) calculations, are comparable with the semi-empirical Pitzer methods as well as with a few available experimental data. The basic buffer solution of our model contains the six most relevant seawater ions Na+, K+, Mg2+, Ca2+, Cl− and SO42- in concentration relations similar to the 2008 Reference Composition of IAPSO Standard Seawater. Some of the remaining ions such as H+, OH− and HCO3− are considered as tracers. We study the dependence of individual and mean ion activities on marine solute concentrations ranging from typical Baltic-Sea to Atlantic salinities.
Buy this Article