Search

• The densities and pH of partially carbonated aqueous ammonia solutions were measured at 20 °C. • The relation between CO 2 loading of the solution and its pH was derived from chemical equilibrium of the NH 3 -H 2 O-CO 2 system. • Using the fact that the increase in mass is proportional to CO 2 loading, a model derived from the definition of the density. • Presented method for determining CO2 loading using pH and density measurements proved successful on real process samples. The densities and pH of partially carbonated aqueous ammonia solutions were measured at 20 °C. With increasing carbon dioxide loading, increases in density and decreases in pH were observed. These results were used to produce correlations for aqueous ammonia solution loading as a function of ammonia concentration, density, and pH. The correlations are based on first principles to increase predictive capabilities. The predictive capacity of three correlations, having different complexity, was compared. The ensemble of models using both density and pH proved to be superior to correlations based on density or pH separately. [ABSTRACT FROM AUTHOR]

The densities of unloaded and partially carbonated piperazine (PZ) and 2-amino-2-methyl-1-propanol (AMP) aqueous solutions and their blends were measured at temperatures of 293.15 to 333.15 K. The results of these measurements were used to calculate excess molar volumes assuming that the volume of carbon dioxide dissolved in the solutions was zero. The excess molar volumes were correlated as a function of the CO2mole fraction using the Redlich–Kister equation. The results of the actual density are discussed in comparison with the case if no volume expansion occurs during the loading of the solutions. The no-volume-expansion model can be used to give an initial estimate of the density, while the Redlich–Kister equation is suitable for a precise density estimation. It must be noted that the no-volume-expansion model does not require any parameters except the density of the unloaded solution, unlike the Redlich–Kister equation, whose coefficients are regressed from experimental data. [ABSTRACT FROM AUTHOR]