Assessment of CO2 desorption from 13X zeolite for a prospective TSA process

In this study, two configurations of Temperature Swing Adsorption (TSA) were assessed with the aim of evaluating their efficacy on CO2 capture on commercial adsorbent zeolite 13X within a post-combustion scenario. A fixed bed setup was employed to measure breakthrough curves from the adsorption and desorption steps. Considering a dry desulfurized flue gas stream, breakthrough curves for CO2–N2 (15/75 % v/v) in Helium were performed at 25, 50 and 75 °C. The desorption step was carried out following two TSA regeneration strategies: a two-step desorption arrangement (configuration 1) consisting of a purging phase followed by a heating-purging phase, and a one-step desorption arrangement (configuration 2) involving only the heating-purging phase. Adsorption equilibrium isotherms were also obtained for pure CO2 (25, 50, 75, 100 and 125 °C) and N2 (25, 50 and 75 °C) on zeolite 13X in the range of 0–1 bar. Finally, a mathematical model considering mass and energy differential balances was used to predict the whole adsorption-desorption history. The results obtained on breakthrough curves showed that CO2 separation from N2 on zeolite 13X is accomplished by adsorption under the studied conditions with a marked selectivity for CO2. In regards to the desorption phase, configuration 1 may not be adequate for an integration of adsorption-desorption steps once only the purge phase duplicates the desorption time as compared to the adsorption stage. On the other hand, configuration 2 is more likely to synchronize the whole adsorption—desorption process since the regeneration time was significantly reduced by this strategy. However, configuration 1 managed to obtain full CO2 recovery with all the temperatures tested during the heating step, whereas configuration 2 reached recovery values around 92%. Moderate temperatures (e.g. 125–150 °C) are feasible to be used for configuration 2 regeneration strategy so as to avoid energy penalties. Simulations were able to reproduce well the experimental breakthrough curves, even though some discrepancies were observed in the desorption histories.