Vapour-liquid Equilibrium of Carbon Dioxide in Newly Proposed Blends of Alkanolamines

Abstract

In the backdrop of a major climatic change vastly due to the greenhouse gas emission, gas treating has become a significant area of interest as it has never been earlier. Carbon dioxide (CO2), one of the major greenhouse gases are treated mainly by absorption in alkanolamine solvents, though ionic liquids and Sodium and Potassium salts of primary or tertiary amino acids are given consideration presently for effective and energy efficient CO2 capture. Gas treating research is actually passing through a transition which has been chronicled in the present work. Among the different technologies available for CO2 mitigation, capture of CO2 by chemical absorption is the technology that is closest to get implemented commercially. In the present context, the role of alkanolamine solvents in sour gas treating research should be revered, hence, N-methyl-2-ethanolamine (MAE), and 2-(ethylamino) ethanol (EAE) has been prudently explored for CO2 capture. Apart from the knowledge of mass transfer and kinetics, the equilibrium solubility of acid gases over alkanolamines presumes an instrumental role in gas treating processes. To exploit poorer quality crude and natural gas twined with the enhanced environmental legislations, highly selective and economic acid gas treating processes are in an unprecedented need. In view of this, the present work was taken up to investigate the possibility of (EAE/MAE + MDEA/AMP) solvent as effective blends towards CO2 absorption.
Measurement of solubility of CO2 in aqueous single amine, MAE and EAE and aqueous amine blends MAE/AMP, MAE/MDEA, EAE/AMP and EAE/MDEA have been done in this work up to a maximum CO2 pressure of 600 kPa for various temperatures and amine concentrations using the VLE measurement set-up developed in this work. The deprotonation and carbamate reversion constants were found out. The VLE data generated on (CO2 + EAE + MDEA + H2O) system was correlated with rigorous thermodynamic model. The vapour phase non-ideality was taken care off in terms of fugacity coefficient calculated using Virial equation of state. Extended Debye-Hückel theory of electrolytic solution was used to address the liquid phase non-ideality. The experimental set-up and procedure has been validated with the systematic VLE data generated on CO2 solubility in
ix
(DEA + AMP/MDEA + H2O) systems. Some of the VLE data generated on the mentioned blends have been compared with the literature data. The generated data are also correlated with a rigorous; activity based thermodynamic model with minimal correlation deviations, thus indicating the robustness of our set-up and procedure. Moreover the new VLE data generated on (CO2 + DEA + AMP/MDEA + H2O) system have enhanced and enriched the data base.
Engineers and scientists usually refer excess Gibbs energy models for vapour- liquid equilibria calculations such as WILSON, NRTL, UNIQUAC, and UNIFAC. In order to describe the thermodynamics for mixtures, these methods compute the activity coefficient of the compounds using the information on binary interaction parameters that are derived from experimental results. Thus, these models have limited applicability in thermodynamics properties and VLE prediction for the new systems that have no experimental data. For solution of this problem, Solvation thermodynamics models based on computational quantum mechanics, such as the Conductor – like Screening Model (COSMO), provide a good alternative to traditional group-contribution and activity coefficient methods for predicting thermodynamic phase behaviour. This thesis provides the COSMO predicted thermodynamic properties of binary (MAE/EAE + H2O) and (MAE/EAE + CO2 + H2O) and systems. The COSMOtherm calculations have been performed the latest version of software that is COSMOtherm C30_1201.
The densities of aqueous blends of (MAE/MDEA), (MAE/AMP), EAE/MDEA, and EAE/AMP for various relative amine compositions have been measured over a wide range of temperature and relative amine composition, and useful correlations developed for prediction of densities of the amine blends. It is expected that the physico-chemical parameters thus generated will also be useful for the database for process design.