ISSN:
1434-4475
Keywords:
Absorption equilibrium
;
Data regression
;
Ionogene substances
;
Maximum likelihood principle
;
Reaction equilibrium
Source:
Springer Online Journal Archives 1860-2000
Topics:
Chemistry and Pharmacology
Notes:
Abstract The solubility of ionogen substances in water and aqueous ionic solutions is important for calculation of absorption processes. Aqueous solutions with complex reaction systems behave themselves extremely nonideal. In simple cases equilibria can be determined with the concept of nonideal thermodynamics. The model used in this work is based on ideal calculation of reaction equilibria and gas solubility. The model parameters (equilibrium constants andHenry constants) for the systems SO2-H2O,MEA-H2S-H2O,DEA-H2S-H2O andMEA-CO2-H2O are computed by regression of experimental data. Equilibrium reactions are selected according toBrinkley's method. The selection of the reacting species has decisive influence on the accuracy of the data fitting. Data regression is done numerically and leads to the formulation of nonlinear systems of equations, which have to be solved for each data point. This solutions are performed in an inner loop. By using the maximum-likelihood-principle the model parameters are optimized in the superior regression loop. Experimental data for the regression are the partial pressure and the total concentration of gas in the liquid phase. The used model is able to fit these data satisfactoryly. The model parameters, which are calculated from simultaneous data regression for different temperatures, ensure a simple correlation ofvan't Hoff. However, for similar reactions equilibria in different reaction systems, it is impossible to compute the same values for the equilibrium constants.
Type of Medium:
Electronic Resource
URL:
http://dx.doi.org/10.1007/BF00808937
Permalink
