Expressions derived from the regular solution equation and the generalized cubic equations-of-state (EOS) were employed to model the infinite dilute activity coefficient (y-) of organic-aqueous systems. A database of 1 19 organic species in aqueous systems, divided into six classes, was gathered for models evaluations. The use of the molar density, the first order molecular connectivity, and the dipole moment appear to be adequate correlating variables for describing the structural dependence of the residual part in the modified regular solution equation (MRS). The Soave-RedlichKwong (SRK) and Peng-Robinson (PR) were specified and employed in the generalized cubic EOS form. The excess activity coefficient concept was used in the SRK and PR EOS models to account for the deficiency of EOS in dealing with highly non-ideal or polar systems. The abilities of two existing correlation models along with two modified versions of the UNIFAC model were also assessed. The proposed models provide creditable equations for correlating y-of organic-aqueous systems. tringent environmental regulations in the fields of water S reuse and wastewater management have increased the industrial significance of predictive models for the phase behavior of organic species in aqueous systems. The fate of trace organic species in the environment is typically described in terms of thermodynamic properties such as Henry's constants, octanol-water partition coefficients, and aqueous solubilities. Such properties are needed, for instance, to predict: (1) water-air phase exchanges (Henry's constant), (2) the ability of the organic solute to bioaccumulate and its ability to sorb in soils and sediments (octanolwater partition coefficient), and (3) trace water-organic phase distribution (mutual solubility). When experimental measurements for Henry's constant, octanol-water partition coefficient, and mutual solubility are not available, they can be calculated from y-. However, all experimental methods employed to measure y" suffer from some limitations (Tiegs et al., 1986;Bader, 1993; Bader and Gasem, 1996a). The current experimental methods are inapplicable for a large number of organic-aqueous systems; systems that are challenging to characterize. Thus, generalized models for predicting ym of organic-aqueous systems based on pure component properties and limited binary mixture data are often needed.*Author to whom correspondence may be addressed. E-mail address: This work is devoted to the modeling of yw of organicaqueous systems. The objectives of this work are to: (1) modify the regular solution equation; (2) apply and amend the generalized form of cubic EOS (SRK and PR); and (3) test the abilities of two existing correlation models as well as two modified versions of the UNIFAC model.
Literature reviewTwo different types of models are used to correlate andor predict yw. These models are either based on component properties or group contribution methods.
COMPONENT-BASED MODELSThe regular solution theory (Scatchard, 193 1;Hildebrand and Wood, 1933...