PNNL

Abstract

Three-dimensional models of atmospheric inorganic aerosols need an accurate yet computationally efficient parameterization of activity coefficients, which are repeatedly updated in aerosol phase equilibrium and gas-aerosol partitioning calculations. In this paper, we describe the development and evaluation of a new mixing rule for estimating multicomponent activity coefficients of electrolytes typically found in atmospheric aerosol systems containing H(+), NH4(+), Na(+), Ca(2+), SO4(2-), HSO4(-), NO3(-), and Cl(-) ions. The new mixing rule, called MTEM (Multicomponent Taylor Expansion Model), estimates the mean activity coefficient of an electrolyte A in a multicomponent solution from a linear combination of its values in ternary solutions of A-A-H2O, A-B-H2O, A-C-H2O, etc., as the amount of A approaches zero in the mixture at the solution water activity, aw, assuming aw is equal to the ambient relative humidity. Predictions from MTEM are found to be within a factor of 0.8 to 1.25 of the comprehensive Pitzer-Simonson-Clegg (PSC) model over a wide range of water activities, and are shown to be significantly more accurate than the widely used Kusik and Meissner (KM) mixing rule, especially for electrolytes in sulfate-rich aerosol systems and for relatively minor but important aerosol components such as HNO3 and HCl acids. Because the ternary activity coefficient polynomials are parameterized as a function of aw, they have to be computed only once at every grid point at the beginning of every 3-D model time step as opposed to repeated evaluations of the ionic strength dependent binary activity coefficient polynomials in the KM method. Additionally, MTEM also yields a non-iterative solution of the bisulfate ion dissociation in sulfate-rich systems, which is a major computational advantage over other iterative methods as will be shown by a comparison of the CPU time requirements of MTEM for both sulfate-poor and sulfate-rich systems relative to other methods.