Selectivity and stability of bismuth-based metal-organic frameworks for CO2 separation

https://doi.org/10.1016/j.micromeso.2021.111548

“The calculated gas selectivities for four hypothetical flue gas or industrially relevant gas mixtures; CH₄/N₂ (50:50), CO₂/CH₄ (50:50), CO₂/N₂ (15:85), and SF₆/N₂ (10:90), can be seen in Table 2 and Fig. 6a–c. The selectivities are calculated from the data obtained at 293 K (closest to room temperaure) only, as the calculated selectivities are considered to be qualitatively indicative selectivities for real-life applications. The CO₂/N₂ IAST selectivities at 100 kPa for all samples were found to vary between 25 and 35, where SU-101 showed the highest CO₂/N₂ selectivity of ∼35, which was comparable to that of UU-200 (∼35), while CAU-17 had the lowest selectivity of ∼25. The CO₂/N₂ selectivity of UU-200 was also found to be lower than that of porous sorbents such as Mg(H₂gal) (377 at 303 K 100 kPa) [64], Fe(Hgal) (612 at 303 K and 100 kPa) [64], zeolite 13X (981 at 298 K and 100 kPa) [65], and ZIF-78 (396 at 298 K and 100 kPa) [66]. But comparable to other MOFs like HKUST-1 (∼30 at 273 K and 100 kPa), HNUST-1 (∼30 at 298 K and 100 kPa), Co-MOF-74 (∼40 at 296 K and 100 kPa), and MIP-202 (∼30 at 298 K and 100 kPa) [55]. Although all samples were found to have Langmuir-shaped CO₂ isotherms, indicating an enhanced molecular interaction between the CO₂ molecules and the surface of the pores as compared to e.g. CH₄ and N₂, the comparably high N₂ uptakes on CAU-17 and CAU-33 result in an overall lower CO₂/N₂ selectivity on these samples. The SF₆/N₂ IAST selectivities showed a similar trend; the highest selectivity was observed on UU-200 (∼45) followed by SU-101 (∼40), and then CAU-17 (∼35). These values were comparable to corresponding literature values for MOFs such as UiO-66 (∼40 at 298 K and 100 kPa) [67] and CAU-17 (∼30 at 293 K and 100 kPa) [29]. The SF₆/N₂ IAST selectivity was not calculated for CAU-33 due to its SF₆-induced flexibility, as the model is typically not applicable for flexible materials [68]. However, modelling of the SF₆ isotherm using a number of isotherm models (e.g. modified dual-site Langmuir [69]) is possible for other purposes. It has also been documented in recent literature that the IAST model has a number of shortcomings; IAST assumes an ideal gas mixture and is particularly problematic for gases with low interaction with the sorbent (as integration over a large pressure range is needed if the selectivity of the sorbent is high). As highlighted by Cheung et al. [70] and Björnerbäck et al. [71], in order to utilize the IAST model, binary or multi-component gas adsorption isotherms are needed. We therefore also calculate the CO₂ and SF₆ selectivity using the Henry’s law model and the numbers are listed in Table 2 for comparison.”

“Fig. 6. IAST selectivities for hypothetical gas mixtures composed of (a) CO₂/CH₄ (50:50), (b) CO₂/N₂ (15:85), (c) SF₆/N₂ (10:90) and (d) pressure-swing CO₂ adsorption cycling on UU-200, CAU-17, CAU-33, and SU-101.”