3.1 Characterization
PXRD patterns of MIL-101(Cr) and IL/MIL-x were carried out. Figure 1a indicates that the PXRD patterns of MIL-101(Cr) correspond well with their simulated ones, confirming the successful synthesis28. Besides, the IL/MIL composites maintain similar diffraction peaks to pristine MIL-101, indicating that the materials retain the framework stability after loading IL. However, the characteristic peaks of IL/MIL at 5-7° are weakened, which may be ascribed to an alteration in electron density, morphology, as well as crystallization 29-31. Then, the porosity of the adsorbents was evaluated by analyzing the N2adsorption/desorption isotherms at 77 K. Figures 1b and 1c suggest that the N2 uptake and the pore size of samples gradually decrease with increasing the content of loaded IL. As listed in Table S1, the BET surface area of the sample decreases from 3111 m2·g-1 to 3 m2·g-1 and the total pore volume decreased from 2.0 cm·g-1 to 0.14 cm3·g-1 with the increase of IL content, which can be ascribed to the pore blockage resulted by the IL entering the pore of MIL-101(Cr). Furthermore, FT-IR spectra of MIL-101(Cr) and IL/MIL-x are shown in Figure 1d. The asymmetric and symmetric stretching vibrations of the dicarboxylate linker O–C–O can be observed at 1624 and 1401 cm-132. In addition, other bands on the benzene ring of the ligand can be found at 1507 cm-1 (C=C stretching) and 1158, 1107, 882 and 748 cm-1 (C-H bending)31,33. These demonstrate the MOF framework is well preserved after loading IL, indicating that the loaded IL does not affect the structural integrity. Besides, new peaks were observed in IL/MIL composites. For instance, the peak situated at 1579 cm-1 and the weak peaks at 2951 and 2852 cm−1 can be ascribed to the stretching vibration of C-N 31,34 and the C-H stretching vibration of the alkyl chain in IL32,35. The above evidence indicates that IL is successfully loaded in MIL-101(Cr).
To further illustrate the existence of IL, the XPS spectra and EDS mapping were investigated by using IL/MIL-0.7 as a case study. Compared with that of MIL-101(Cr) (Figure 2), the spectrum of IL/MIL-0.7 displays the characteristic peaks corresponding to Cl and N elements. EDS mapping under SEM mode of IL/MIL-0.7 shows the homogeneous distribution of N and Cl elements (Figures S2c and S2d), which shows the excellent dispersion performance of IL in MIL-101(Cr). SEM images further confirmed the maintenance of crystal morphology. IL/MIL-0.7 has similar particle sizes (200–500 nm) and octahedral crystals typical of MIL-101(Cr) as shown in Figures S2a and 2b. Additionally, TEM analysis (Figure 3) reveals a layer of transparent substance with a non-uniform thickness (more than 15 nm) on the outside of IL/MIL-0.7 compared with pristine MIL-101(Cr), suggesting that IL was deposited on the outer surface of MIL-101(Cr). The above characterization data demonstrate the successful loading of IL.
3.2 SO2adsorption performance
To evaluate the properties of the composites, the adsorption isotherms of SO2 over four adsorbents were determined by the volumetric method at 298 K and 1 bar. Figure 4a indicates that the maximum adsorption uptake of IL/MIL-x decreases compared with MIL-101(Cr) due to the reduction of pore volume and the different adsorption sites caused by the introduction of IL 9,19. Besides, the saturated adsorption capacity of SO2 changes with the increase of loaded IL content. As shown in Figure 4a and Table S2, the SO2 uptake capacity of IL/MIL-0.7 and IL/MIL-0.4 is (13.18 and 11.36 mmol·g-1, respectively) lower than that of IL/MIL-0.2 (13.51 mmol·g-1) at 298 K and 1 bar. Due to the BET surface area of the composite decreasing with the increasing IL content, the adsorption capacity of the porous carrier drops somewhat. Interestingly, the SO2 adsorption capacity of IL/MIL-0.7 (13.18 mmol·g-1) is higher than IL/MIL-0.4 (11.36 mmol·g-1), but the BET and pore volume of the former material is smaller than those of the latter. This indicates that the introduction of IL induces additional adsorption sites for the adsorbent, prompting the capture of SO2. Therefore, it can be concluded that the saturated adsorption capacity is not only determined by the specific surface area and pore volume but also depends on the action of IL. Although the adsorption capacity of IL/MIL-0.7 is subequal to that of MIL-101(Cr) at 298 K and 1 bar, its performance is better than some other materials previously reported, including PI-COF-m10 (6.3 mmol·g-1) 36, and other typical porous materials, such as SIFSIX-1-Cu (11.01 mmol·g-1)19, Ph-4MVIm-Br (8.12 mmol·g-1)18, MFM-300(In) (8.28 mmol·g-1)15, MIL-160 (7.2 mmol·g-1)16, ECUT-111(11.6 mmol·g-1)37, SIFSIX-2-Cu-I (6.90 mmol·g-1) 20, SIFIX-3-Ni (4.30 mmol·g-1) 19, CPL-1 (1.999 mmol·g-1) 38, ELM-12 (2.73 mmol·g-1) 39, NPC-1 (2.45 mmol·g-1) 40, and MFM-601 (12.3 mmol·g-1) 20 (Table S2). Notably, the low-pressure adsorption capacity is increased for IL/MIL with the increase of the loaded IL content, as shown in Figure 4b. The SO2 adsorption capacity of IL/MIL-0.7 significantly exceeds that of pristine MIL-101(Cr) at low partial pressures (below 0.1 bar). Moreover, the SO2 uptake of IL/MIL-0.7 (0.67 mmol·g-1) at 0.002 bar exceeds that of MFM-300(In) (0.43 mmol·g-1) 15 and MFM-601 (0.24 mmol·g-1) [20], only lower than [TMEDA][DES]@BN (0.82 mmol·g-1, 293 K) 41, and P(Ph-4MVIm-Br) (1.55 mmol·g-1) 18. However, these two kinds of materials exhibit lower adsorption capacity compared to IL/MIL-0.7 at 1 bar. SO2 adsorption performance is up to 1.68 mmol·g-1 when the partial pressure reaches to 0.01 bar, which is more than 3.3 times as high as MIL-101(Cr) (0.5 mmol·g-1). As the pressure further increases to 0.1 bar, the adsorption amount of SO2 is up to 4.7 mmol·g-1. Compared with the other reported porous materials, this material has a much higher low-pressure adsorption capability, including P(Ph-4MVIm-Br) (4.14 mmol·g-1) 18, SIFSIX-2-Cu-i (4.16mmol·g-1) 19, and HNIP-TBMB-1(3.54 mmol·g-1) 42. These suggest that IL/MIL-0.7 can provide excellent SO2capacity from low pressures to atmospheric pressures, which is of great importance from a practical point of view.
3.3 SO2 separationperformance
Since SO2 is usually involved in many industrial gases as a contaminant, for example, in flue gases (CO2 and N2)43, it is crucial to explore the separation performance of SO2from gas mixtures. The CO2 and N2adsorption isotherms of IL/MIL-0.7 at 298 K were obtained to evaluate the separation selectivity. Although both CO2 and SO2 are acid gases, their adsorption isotherms on the composites have different adsorption behaviors. As shown in Figure 5a, the adsorption capacity of SO2 (13.18 mmol·g-1) was significantly higher than that of CO2 (0.27 mmol·g-1) and N2 (0.07 mmol·g-1) in IL/MIL-0.7. This variation in the adsorption capacity can be attributed to the introduction of IL which induces unique adsorption sites for SO2. In detail, the strong adsorption interaction between SO2 and IL/MIL-0.7 pushes the SO2 to overcome the large mass transfer resistance of the IL layer to enter the material, while CO2 and N2 cannot penetrate the adsorbent due to their weak force between them. Therefore, IL/MIL-0.7 exhibits larger SO2 capacity and lower N2 and CO2 uptake. Furthermore, from the comparison of the SO2 adsorption capacity of IL/MIL-0.7 with reported porous materials in Figure 5b and Table S2, IL/MIL-0.7 shows superior SO2 adsorption capacity. More importantly, the CO2 and N2 uptakes of IL/MIL-0.7 are significantly lower than other adsorbents reported.
To further evaluate the separation efficiency, SO2/CO2 selectivity was calculated based on the ideal adsorption solution theory (IAST) (Section 2 in SI). Firstly, the dual-site Langmuir (DSL) model was used to fit the adsorption isotherm (eq. S1), and then the separation selectivity was calculated by eq. S2 from the single-component gas adsorption isotherms. Fitted curves and fitted parameters can be found in Figure S3 and Table S2, respectively. As expected, the composite exhibits an ultra-high SO2/CO2selectivity over 104 (about 11925) for a mixture of 10% SO2 and 90% CO2 at 298 K and 1.0 bar (Figure 6a). The excellent adsorption selectivity towards SO2 for IL/MIL-0.7 can be explained by a molecular-sieve effect. The introduction of IL led to selective permeation, whereas SO2 can be accessible to the interlayer spaces IL/MIL-0.7 while the CO2 and N2molecules are excluded by their weak adsorption interaction. Then, the SSO2/CO2 of IL/MIL-0.7 was compared with the reported adsorbent. Figure 6b indicates that the separation performance of IL/MIL-0.7 is significantly higher than those in the benchmark materials (Figure 6b and Table S2), such as Mg-gallate (325) 14, and P([allyl-TMG]Br-DVB)(452) 8. Furthermore, for a pressure swing adsorption (PSA) process, selectivity and working capacity are typically regarded as two crucial criteria to identify a new adsorbent, while the sorbent selection parameter (S sp), combining selectivity and working capacity in a single parameter, can be used to better evaluate the potential of adsorbent. Therefore, the value ofS sp of IL/MIL-0.7 for SO2/CO2 is calculated by eq. S3 and compared with the value of reported adsorbents. Figure 6c indicates that IL/MIL-0.7 exhibits a higher value of S sp (about 472131) than the majority of other materials. To summarize, IL/MIL-0.7 has a good potential to capture trace SO2 with high selectivity during the flue gas desulfurization (FGD) process.
It is well known that the content of SO2 in industrial flue gas (such as coal combustion flue gas) is very low (about 2000 ppm)41,44. To confirm the practical ability of IL/MIL-0.7 to capture trace SO2 and the feasibility of composites in the application process, the breakthrough experiment was carried out on the simulated flue gas at 298 K and 1.0 bar, that is, a ternary gas mixture SO2/CO2/N2 (2000 ppm/15%/84.8%, v/v/v) containing 2000 ppm SO2 with a flow rate of 40 mL·min-1. Figure 7 indicates that CO2 and N2 rapidly elute through the column at the beginning, while the breakthrough time of SO2 on IL/MIL-0.7 could reach up to 130 min·g-1, which means that the material can realize the deep and selective removal of trace SO2. The breakthrough selectivity of SO2/CO2 in such ternary gas mixture SO2/CO2/N2 can reach up to 668 calculated by eq. S4, which is at least 5–9 times higher than those in other materials reported in the literature (Figure 7b).
To assess the binding energy between IL/MIL-0.7 and different gas molecules, single-component adsorption isotherms were compared at different temperatures. As shown in Figures 8a and 8b, the adsorption capacity of both decreases significantly with the increase in temperature, indicating an exothermic adsorption process. The heats of adsorption (Qst ) of SO2 and CO2 were calculated using the Clausius-Clapeyron equation (eq. S5) (Figures 8c and 8d), which is from 20 to 40 kJ·mol-1 for SO2 while from 0 to 4 kJ·mol-1 for CO2. This indicated that the affinity of IL/MIL-101 for SO2 is strong than that for CO2. To further explore the interaction between SO2 and composites, FT-IR and XPS spectra were used to evaluate possible SO2 adsorption sites in IL/MIL-0.7. Firstly, the IL/MIL-0.7 before and after SO2 absorption was characterized by FT-IR. Figure S4 demonstrates that the intensity of the hydroxyl peaks of IL/MIL-0.7 after adsorbing SO2 is weakened, indicating that SO2 interacts with the hydroxyl groups45. No symmetric or asymmetric stretch band (between 1389 cm-1 and 1085 cm-1) was observed for SO2, probably because of the overlap with the strong vibrational band of IL/MIL-0.7. At the same time, the characteristic peak of the S element appears in the XPS spectrum, demonstrating that SO2 molecules were adsorbed in the composites (Figure S5a). The binding energy of Cl increases with a deviation of 0.14 eV after IL/MIL-0.7 adsorption of SO2(Figure S5b), which suggests that Cl plays a key role in the adsorption of SO2 4,46,47. Acidic hydrogens in [BOHmim]+ such as alkyl side chain C-H and 2-CH can also form two hydrogen bonds with SO2: H···O(SO2)-S(SO2)7,48,49. As a result of the synergistic effect of multiple adsorption sites and the inner surface of the composite, IL/MIL-0.7 has a high adsorption capacity for SO2.
3.4 Stability and regeneration properties of composites
Considering the strong corrosive nature of SO2, the regenerative prosperities and thermostability of IL/MIL-0.7 were investigated. Firstly, the regeneration property of IL/MIL-0.7 was evaluated. As shown in Figure 9a, 4 cycles of SO2adsorption were performed to test the reversibility of IL/MIL-0.7, which still maintains good adsorption performance for the fourth time at 1 bar. In addition, FT-IR spectroscopy was employed to assess the completion of SO2 desorption. Compared with the fresh composites, the characteristic peaks of FT-IR spectra did not change significantly (Figure 9b) in recycled IL/MIL-0.7, which means that the material can completely desorb SO2. Furthermore, Figure S6 suggests that the structure of IL/MIL-0.7 can maintain stability until 450 K, demonstrating that IL/MIL-0.7 is stable enough for industrial applications 14. All these results indicate that the composites have good enough reversibility for SO2 adsorption.
4. Conclusion
In conclusion, an IL/MOF composite with multiple adsorption sites has been prepared and used as an efficient adsorbent for SO2capture. IL/MIL-0.7 exhibits a high SO2 capacity at 298 K and 1 bar, and an adsorption capacity of 4.7 mmol·g-1 at 298 K and 0.1 bar, which is higher than most known adsorbents in the literature. Meanwhile, IL/MIL-0.7 excludes N2 and shows high SO2/CO2 selectivity, up to 11925 for the 10/99 mixture at 298 K and 1 bar. Finally, SO2, CO2 and N2 molecular sieving can be achieved. In addition, the breakthrough experiment of SO2/CO2/N2 further proves that IL/MIL-0.7 has excellent trace SO2 capture capacity in its practical application. IL/MOF also exhibits good structural stability and sufficient reversibility. Therefore, loading ILs on MOFs to synthesize new adsorbents shows great potential for a practical flue gas desulfurization process.