https://doi.org/10.1016/j.seppur.2020.118193
“In loose terminology terms, from the bulk of the vapor phase to the bulk of the liquid phase, each CO2 molecule has to overcome resistances coming from the gas and liquid sides. This means that the three equations below describe this system.NCO2=kg·pCO2-pCO2,iNCO2=kl·CO2i-CO2pCO2,i=HCO2·CO2i
Applying the three equations shown before, the molar flux of CO2 can be calculated by the following equation.NCO2=pCO2-HCO2·CO21kg+HCO2kl
Considering the equation above, we must notice that the numerator is given by the difference between the CO2 partial pressure in the vapor phase (pCO2) and the CO2 partial pressure in equilibrium with the solvent at a given loading. This driving force is the ΔpCO2 of a process and is not a function of the solvent itself, but merely of how reversibly one chooses to operate the absorption of gas. Similarly, the vapor phase mass transfer coefficient kg is not generally dependent on the solvent, though solvent volatility does have an effect on kg in very particular experimental applications [143]. Conversely, the Henry’s coefficient of CO2 and the liquid phase mass transfer coefficient are strong functions of the solvent inherent properties.
Supposing one wants to develop a water-lean solvent that provides faster absorption rates than an aqueous amine, the following equation should hold.ΔpCO21kg+HCO2,wlkl,wl>ΔpCO21kg+HCO2kl
Song et al. [159] have empirically demonstrated in a packed column that kl is a function of liquid velocity, CO2 diffusivity and solvent viscosity. By normalizing the effects of velocity and applying the consideration that diffusivities themselves vary with viscosity, they show that kl and viscosity are correlated following a slope of approximately − 0.7. In other words, the equation given previously could be rewritten as the one shown below. The equation below completely ignores the effects that organic diluents have on kinetic rates, since the expression for kl obtained by Song et al. [159] is valid only for liquid phase mass transfer coefficients in the absence of chemical reaction.ΔpCO21kg+HCO2,wlηwl-0.7>ΔpCO21kg+HCO2η-0.7
As we have seen, kinetic rates are typically depressed in water-lean solvents. Therefore, the expression above delimits a best-case scenario for these mixtures, i.e. one cannot reasonably assert that water-lean solvents that obey this relationship will absorb CO2 faster, but that water-lean solvents that do not obey it will probably absorb CO2 slower.
Reorganizing this expression, one finally ends up with the relationship below between the Henry’s coefficient and the viscosity for a fast water-lean solvent for CO2 capture.HCO2,wlHCO2<ηwlη-0.7
This relationship, though far from perfect, has some interesting implications. Fig. 10 has been obtained by plotting this equation. The locus for finding fast absorbing organic solvents seems smaller than one would initially think, being entirely delimited by the area below the red line. It is imaginable that a similar locus should also be observed in the context of mixing these diluents with amines.”
“Fig. 10. Relationship between Henry’s coefficient and viscosity for a fast absorbing organic solvent. The relative Henry’s coefficients and viscosities of some typical diluents, pure and at 25 °C, are plotted for comparison. Their CO2 solubilities were obtained in Hansen [64] while their viscosities were obtained in Yaws [160].”
“In Fig. 10 one can also see the relative properties of some pure diluents at 25 °C. These diluents are shown to illustrate the fact that, though there are several organic solvents that deliver CO2 solubilities higher than water, their viscosities might often be problematic if one is looking for high mass transfer rates.
Another consequence of that relationship is that variations in temperature will affect the relative advantage of shifting from aqueous to water-lean solvents. This is because the temperature dependency of Henry’s coefficients and viscosities are not alike for water and organic diluents, though they might be similar. Fig. 11 illustrates this fact. One can observe that increases of temperature move sulfolane towards the ‘right direction’ in terms of enabling a faster CO2 absorption than water.”
“Fig. 11. Dependency on temperature of relative Henry’s coefficients and viscosities for sulfolane. Henry’s coefficients at different temperatures for CO2 in sulfolane were obtained in Murrieta-Guevara et al. [23], whereas those for CO2 in water were obtained in NIST. Viscosities and their variations with temperature were obtained in Yaws [160].”
“Fig. 11 might perhaps help explaining the observations of Garcia et al. [144], who have experimentally identified a steeper increase in mass transfer rates with temperatures while absorbing CO2 in water-lean solvents than in aqueous ones.
However, one should be careful to note that both Fig. 10 and Fig. 11 show the relative Henry’s coefficients and viscosities only for pure diluents, with no addition of amines nor loadings of CO2. As we shall see in Section 6.2, the viscosity dependency on CO2 loading is quite an important aspect in water-lean solvents.”