Vapor-liquid equilibrium experiments using MEA

https://doi.org/10.1016/j.cesx.2021.100096

“A schematic drawing of the VLE apparatus employed in this work is shown in Fig. 2. This apparatus and its standard operating procedure are the same as used in past studies by our group (Bernhardsen et al., 2019, Hartono et al., 2017). What follows is a quick recapitulation of what has been reported previously.”

“Fig. 2. Schematic of the vapor–liquid equilibrium setup.”

“The setup consists essentially of a steel cylinder containing pressurized CO₂ and a stirred glass autoclave coupled to a Julabo temperature controller. Fresh solution containing water, amine and salt is prepared gravimetrically and fed to the reactor. This reactor is then vacuumed, so that the vapor phase in the autoclave should consist only of solvent molecules in equilibrium with the liquid phase. The temperature controller is set to its designed setpoint and the stirrer is turned on to approximately 500 rpm. After this setpoint is reached and enough time is given for temperature and pressure to reach equilibrium both inside the CO₂ cylinder and the reactor, the valve connecting these two equipments is opened. Gas flows from the cylinder to the autoclave. Once enough CO₂ is injected, the valve is closed. Once again, one must wait until temperature and pressure reach an equilibrium in both equipments. If required, the CO₂ cylinder must be refilled. A new injection is made when temperature and pressure are stable again, and this cycle is repeated until the autoclave reaches its maximum safe pressure of 600 kPa.

The volume comprised between the two valves and the CO₂ cylinder shown in Fig. 2 has been previously calibrated, and so has the volume comprised between the second valve and the autoclave. The volume of the liquid phase is considered constant throughout the experiment, being estimated by knowledge of the amount of solvent fed to the reactor plus its density, here assumed to be the same density of aqueous MEA 30 %wt. at 40 °C as obtained by Amundsen et al. (2009). With volumes, temperatures and pressures for both the cylinder and the vapor phase of the autoclave before and after each injection, mass balance calculations can be performed to evaluate CO₂ loadings versus total pressure. In our work, the Peng-Robinson equation of state has been chosen for estimating the molar quantity of CO₂ at all times.

This methodology does not allow for the direct evaluation of CO₂ partial pressures, though these can be obtained by subtraction if one assumes that the solvent vapor pressure is kept constant throughout the experiment. After the procedure is finished, the solvent left in the reactor is titrated so that one can compare if the loading calculated by mass balance is consistent with the loading evaluated analytically. A match of ±3% between these final loading values is considered sufficient for validating our methodology. Additionally, our data is reported in the Appendix with confidence intervals obtained by following the procedure outlined by Wanderley et al. (2020).”