Mathematical modeling of CO₂ absorption with ionic liquids in a membrane contactor, study of absorption kinetics and influence of temperature

Abstract

BACKGROUND This work presents a comprehensive study of carbon dioxide (CO2) capture from CO2/N2 mixtures in a hollow fiber membrane contactor (HFMC). The absorbents considered were three different ionic liquids (ILs): 1‐ethyl‐3‐methylimidazolium acetate ([emim][OAc]), 1‐ethyl‐3‐methylimidazolium ethylsulfate ([emim][EtSO4]) and 1‐ethyl‐3‐methylimidazolium methyl sulfate ([emim][MeSO4]). A comprehensive two‐dimensional (2D) mathematical model for the transport of CO2 was developed and solved for steady‐state and pseudo steady‐state modes. For pseudo steady‐state mode, the 2D model was linked with a dynamic model applied on a tank from which the IL was recirculated into the shell side of the HFMC. RESULTS The model showed excellent agreement between the simulations and experimental data within the range of 2 to 5% standard deviation. Initially, overall mass transfer coefficients of 7.6 × 10−6 (m s−1), 1.6 × 10−6 (m s−1) and 3.7 × 10−6 (m s−1) were obtained for [emim][OAc], [emim][EtSO4] and [emim][MeSO4], respectively. Solubility, diffusivity and absorption efficiency of CO2 in ILs showed strong dependence on the temperature. Furthermore, concentration drop in the wetted portion of the membrane, effect of porosity, tortuosity, gas and absorbent flow rates were systematically studied. CONCLUSION Findings of both experimental and modeling work suggests that in spite of the difference of extraction efficiency, these ILs are potential absorbents for CO2 in membrane contactors. Moreover, the modeling approach was found very effective to predict the absorption behavior of CO2 in ILs and to study the wetting phenomena and parametric effects. © 2019 Society of Chemical Industry

CO2 Desorption Performance from Imidazolium Ionic Liquids by Membrane Vacuum Regeneration Technology

“…The sensitivity analysis indicated that at higher vacuum applied in the MVR system, which corresponds to higher desorption efficiencies, the Henry constant variability was less important, as the sensitivity of the model to this parameter was more critical when the desorption efficiencies were low. into account e tchemical equilibrium between CO2 and [emim][Ac] [39,40]. Our reference was the Henry constant calculated in our work ( = 8.8) by a correlation proposed by Hospital-Benito et al [44], which was known to be very successful for the estimation of Henry constant for CO2 absorption in imidazolium ILs.…”

“…Additionally, a sensitivity analysis of Henry’s constant ( ) of CO 2 in [emim][Ac] (Equation (11)) and pre-exponential factor (B) of the first-order reaction rate constant (Equation (2)) were performed since these parameters were the ones with more uncertainty based on the literature. Just as an example, Henry’s constant is in the range of 0.0051 to 9.3 MPa because some studies do not take into account e tchemical equilibrium between CO 2 and [emim][Ac] [ 39 , 40 ]. Our reference was the Henry constant calculated in our work ( = 8.8) by a correlation proposed by Hospital-Benito et al [ 44 ], which was known to be very successful for the estimation of Henry constant for CO 2 absorption in imidazolium ILs.…”

“…Additionally, the pre-exponential factor value was reported in a range from 0.11 to 1545 s −1 since different approaches to estimate pre-exponential value were described in the literature [ 35 , 40 ]. However, Figure 13 shows that B value changes are slight in spite of the large range considered (from 0.11 s −1 to three orders of magnitude higher) because the mass transfer of CO 2 in the proposed model mainly depends on the driving force in the gas-liquid interface.…”

“…The sensitivity analysis indicated that at higher vacuum applied in the MVR system, which corresponds to higher desorption efficiencies, the Henry constant variability was less important, as the sensitivity of the model to this parameter was more critical when the desorption efficiencies were low. Additionally, the pre-exponential factor value was reported in a range from 0.11 to 1545 s −1 since different approaches to estimate pre-exponential value were described in the literature [35,40]. However, Figure 13 shows that B value changes are slight in spite of the large range considered (from 0.11 s −1 to three orders of magnitude higher) because the mass transfer of CO2 in the proposed model mainly depends on the driving force in the gas-liquid interface.…”

“…However, Figure 13 shows that B value changes are slight in spite of the large range considered (from 0.11 s −1 to three orders of magnitude higher) because the mass transfer of CO2 in the proposed model mainly depends on the driving force in the gas-liquid interface. Additionally, the pre-exponential factor value was reported in a range from 0.11 to 1545 s −1 since different approaches to estimate pre-exponential value were described in the literature [35,40]. However, Figure 13 shows that B value changes are slight in spite of the large range considered (from 0.11 s −1 to three orders of magnitude higher) because the mass transfer of CO 2 in the proposed model mainly depends on the driving force in the gas-liquid interface.…”