Study On The Thermodynamic Models Of Ionic Liquid Systems And Process Simulation Of Carbon Capture

Ionic liquids (ILs) have received burgeoning academic and industrial interests in energy, catalysis, and separation field due to their unique properties, such as extremely low vapor pressure, high thermal stability and designability. However, there is a lack of physicochemical properties and thermodynamic models for IL systems, which hindered the design and application of IL-containing new processes. Although accurate experimental measurement is indispensably important, it is always not practical to obtain all the data through experiment for the large amounts of IL systems. Thus generalized predictive methods are fundamentally requisite. In this work, group contribution method and corresponding states principle were employed to develop predictive models for thermodynamic properties and phase equilibria of ionic liquids system. Moreover, the simulation and assessment approach of IL-containing process was developed in this paper. Taking CO2 capture by IL-based solvent as an example, techno-economic assessment was performed to provide information for the development of energy-saving and cost-efficient carbon capture. The main contents and results of the dissertation are as follows:(1) Physicochemical properties prediction of pure and mixed ionic liquids. Considering the ionic features of ILs,46 fragments were obtained on the basis of the fragment-based classification method proposed in this paper. Combining with corresponding states principle (CSP), the fragment increments were determined using the experimental density data of 490 ILs. Thus a new fragment contribution-corresponding states (FC-CS) method was successfully developed to predict the properties of ILs. The critical properties of 538 ILs were calculated by the FC-CS method, which provides the necessary parameters for the application of CSP and equation of state into ILs. As extended applications of the FC-CS method, density and surface tension of pure and mixed ILs were predicted by original CSP equations, and the average absolute relative deviation (AARD) of 6150 data points was less than 4%. CSP correlations for heat capacity and thermal conductivity prediction of ILs were further proposed with AARDs less than 4%. All the above results indicate that the fragment classification method and the FC-CS method are reasonable and widely available for pure and mixed ILs.(2) Physicochemical properties prediction of mixtures of ionic liquids with molecular solvents. Based on the FC-CS method, density equation without adjustable parameter was developed through combining mixing rules with CSP. The AARD of predicted density of 63 binary and ternary mixtures (1985 data points) was merely 0.92%. Semi-empirical models with only one characteristic parameter were proposed to predict the surface tension, heat capacity and thermal conductivity of IL mixtures, and the AARDs were all less than 2%. Besides the developed methods based on thermodynamic theory, models based on artificial neural network were also developed with the AARDs less than 0.5%, which presented higher accuracy for property prediction of mixtures of ionic liquids with molecular solvents.(3) Prediction of vapor liquid phase equilibrium of IL systems. Based on the partial dissociation of ILs and the classification of ionic fragment (IF), the IF-UNIFAC model was developed to predict the phase equilibrium of ionic liquid systems, which described the long range electrical interaction by the PDH (Pitzer-Debye-Huckel) model and the short range interaction by the UNIFAC model. The prediction activity coefficient of molecular solvents in 8 binary IL-containing systems showed an AARD of 2.2%, which indicates the fragment classification approach and the IF-UNIFAC model are reasonable and available for vapor liquid phase equilibrium prediction of IL systems.(4) Phase equilibrium study of CO2 absorption system with IL-based solvents. In order to model phase equilibrium of CO2 absorption system with IL-MEA hybrid solvents, RK (Redlich-Kwong) equation of state and NRTL model were employed to correlate the gas liquid equilibrium data of CC^in three different ILs ([Bmim][BF4], [Bmim][DCA] and [Bpy][BF4]). The Henry constants and NRTL binary interaction parameters between CO2 and ILs were obtained with the AARDs less than 0.9%. A set of experimental apparatus was established to measure the vapor liquid equilibrium data of IL+H2O (+MEA). NRTL binary interaction parameters between IL and H2O or MEA were determined by correlating the measured experimental data, which presented an AARD of 0.8%.(5) Process simulation and assessment of IL-containing clean process. Based on the thermodynamic model established in this paper, simulation and assessment approach of IL-containing process was developed. CO2 capture processes using three different IL ([Bmim][BF4], [Bmim][DCA] and [Bpy][BF4])-MEA hybrid solvents were simulated in Aspen Plus. Among these, the [Bpy][BF4]-MEA process (IL process) presents the minimum regeneration heat duty of 3.17GJth/ton CO2 at the lean loading of 0.2, which is 15% lower than the conventional MEA process. After screening the solvent, the process structure of the IL process was modified via adding inter-cooling and lean vapor recompression. The Modified IL process saved about 31% regeneration heat energy consumption and 13.5% total capture cost compared to the MEA process. The simulation and assessment results provide design information for carbon capture process using IL-based solvent.