https://doi.org/10.1016/j.ijggc.2018.10.008
“Before performing regeneration analysis of CO2-loaded MEA solutions, fresh solutions were first submitted to microwave irradiation to gauge their behaviour to this specific heating mode. This preliminary work allowed to select an adequate initial microwave heating power of 100 W which was sufficient to heat the solutions to high temperatures (80 °C) in a relatively short time frame (from around 30–50 seconds depending on the MEA concentration). Several MEA concentrations were tested and their heating profile can be seen in Fig. 2a along with the heating curve of pure water. All experiments were performed at least twice and the reproducibility was excellent, giving the same trend every time. It should be noted that all data included in figures in this work can be found in a dataset available online (Bougie and Fan, 2018). From Fig. 2a, it is first possible to see that the heating rate increases with MEA concentration but up to 50 wt% MEA. The 60 wt% MEA solution has a similar heating profile than the 50 wt% one while the 70 wt% MEA solution take more time to reach the temperature set point and heat up at a rate similar to the 40 wt% MEA solution.

Fig. 2. Heating curves for fresh (a) and CO2-loaded MEA (b) solutions with a microwave power of 100 W.
These heating rates can be partly explained by the heat capacity of each solution. Data reported in the literature (Chiu and Li, 1999) show that the heat capacity of MEA solution decreases continuously with an increase in concentration at a given temperature (e.g. 4.19, 3.76, 3.43 and 3.14 J/g.K for water and 30, 50 and 70 wt% MEA respectively at 25 °C). Therefore, as the solutions need less energy, the heating rates should increase continuously in the same way, which is true up to 50 wt% but not for higher concentrations. One possible explanation for the different behaviour at higher concentrations may come from the viscosity of the solution that significantly increases above 50 wt% (e.g. 0.89, 2.48, 5.51, 12.46 mPa.s for water and 30, 50 and 70 wt% MEA respectively at 25 °C) (Amundsen et al., 2009). It was mentioned that molecules in a highly viscous media, or in a media with a high number of intermolecular interactions (which can explain the high viscosity), have a slower response to an oscillating electric field like microwave (Mishra and Sharma, 2016; Salvi et al., 2009). In such media, molecule oscillations will be hindered resulting in a slower rate of volumetric heating. On the other hand, the variation of the heating rate with MEA concentration does not seem to be related to density as this physicochemical property is quite constant (e.g. 0.997, 1.010, 1.021, 1.026 g/cm3 for water and 30, 50 and 70 wt% MEA respectively at 25 °C) (Tseng and Thompson, 1964).
The heating rates of CO2-loaded solutions were also determined and results can be seen in Fig. 2b. Several differences can be found in comparison to Fig. 2a; the most obvious one being the longer time needed to reach 80 °C (from 50 to 80 s, depending on the concentration). Again, lower heat capacities for CO2-loaded solutions in comparison to fresh ones (Weiland et al., 1997) would have alone decreased these time. It means that the viscosity increase after CO2 absorption (Zhang et al., 2015) has a more significant effect on heating these solutions. For illustration, a 30 wt% MEA solution experiences a decrease of 10% of its heat capacity after reaching a loading of 0.5 mol CO2/mol amine at 25 °C (Weiland et al., 1997) while its viscosity increases by 46%. The high viscosity of the loaded 60 and 70 wt% solutions was also responsible for the different shape of their heating curve as microbubbles were trapped in the liquid during absorption and released during the heating step, modifying the temperature quickly.
Besides, the slower heating rate of CO2-rich solutions in comparison to fresh ones can certainly be explained by the endothermic nature of the CO2 desorption; some microwave energy being diverted to strip the CO2 instead of heating the solution. One last parameter that may influence the microwave heating rate of these solutions while variating the MEA concentration or the CO2 loading is the change in their dielectric constants (ε′ and ε″). However, as the regeneration of spent aqueous amine solutions is a new field of research, no data were found in the literature to analyse this phenomenon.
Therefore, a maximal heating rate can then be explained to be around 50 wt% MEA based mainly on the opposite trend of the solution heat capacity and viscosity with increasing MEA concentration. At concentrations higher than 50 wt%, the viscosity may become too much detrimental for high microwave efficiency. From a process point of view, a maximal heating rate is interesting as if the solution take less time to reach its temperature set point, a lower residence time in the stripper will be required and the latter can then be smaller.”