On thermodynamic properties of aqueous ethylamine mixtures and modelling of excess properties
Ethylenediamine (EDA) is a bifunctional amine used for several different purposes such as for production of the Ethylenediaminetetracetric acid (EDTA) that is used as a chelating agent in detergents but also for fungicides, bleaching activators etc. In many applications very high purity is needed and the removal of the byproduct Ethylethylenediamine (EtEDA) has proven to be difficult. Experimental data presented in this thesis showed that the relative volatility between EDA and EtEDA in the region close to pure EDA is very low. EDA and water form a maximum boiling azeotrope at atmospheric pressure and below. In both its binary mixtures with water and with EDA, EtEDA was the heaviest component at all compositions. In the ternary mixture however, EtEDA is the most volatile component in an area of composition close to the binary azeotrope. A ternary saddle point azeotrope is found at at a molar fraction of 0.24 of water, 0.62 of EDA and 0.14 of EtEDA and a temperature of 392.6 K at atmospheric pressure.
Experimental data for the density of mixtures of water and EDA are also presented in this thesis. At temperatures below 300K, there is a local maximum in density in the water rich region. Thus, measurements of density cannot be used to determine the composition of EDA water mixtures in that area. The excess molar volume is negative and has unusually high amplitude. This indicates strong intermolecular forces.
In chemical engineering, reliable models for vapor-liquid equilibria (VLE) are important. There are two different approaches to VLE: with the equation of state, the one and same model is used for both phases and several different properties can be calculated with the model. It is used for calculations with non-polar compounds. For polar compounds at low pressures a Gibbs excess energy (gE) model is used to describe the liquid phase and the vapor phase is often considered to be an ideal gas. The gE mixing rules attempt at incorporating the information from the gE model into the equation of state. In this thesis, the ability of some models to cross-predict excess enthalpy (hE) and VLE was investigated. Among the gE models included, the UNIQUAC equation was found to be the most reliable for prediction of hE from VLE data. Flexibility has to be added to the equation by letting the binary parameters be linearly dependent on temperature. If the UNIQUAC equation is to describe both VLE and hE reliably, both kinds of data have to be included in the estimation of the parameters. The UNIQUAC equation was incorporated in the MHV2 and Soave mixing rules for the Soave-Redlich-Kwong equation of state. For VLE temperature extrapolations it had been observed that a cancellation of errors improved the results; A lack of fit caused by an approximation done in the MHV2 mixing rule, cancelled out a common error in the temperature dependence of the UNIQUAC equation. The same phenomenon was observed for the predictions of hE in this study.