2.5 Nanofiltration performance of hierarchical MOF lamellar membranes
Nanofiltration performance was evaluated after immersing hierarchical MOF lamellar membrane in corresponding solution for 3 h to reach a fully equilibrium state. Based on the home-made device, solvent permeance and dye rejection were tested. Dyes dissolved in methanol (10 mg L-1) were used for rejection measurement and analyzed by UV-vis spectrophotometer. The solvent permeance P (L m-2 h-1 bar-1) was calculated by using the following equation:
\(P=\frac{V}{P\times A\times t}\) (1)
where V , ΔP , A and t , represent the permeate volume (L), operating pressure (bar), effective membrane area (m2), and testing time (h), respectively. The following expression was obtained for the calculation of dye rejection (R , %):
\(R=\left(1-\frac{C_{p}}{C_{f}}\right)\times 100\) (2)
where C f and C p are the concentration of feed solution and permeate solution, respectively. The obtained data were average of three parallel tests.
3 RESULTS AND DISCUSSION
3.1 Preparation and characterization of hierarchical MOF lamellar membranes
MOF nanosheets were solvothermally synthesized from a mixed solution of metal ions (Ni2+, Co2+) and organic linkers under continuous ultrasonication.[37]Here, three kinds of organic linkers with distinct functional groups (–CH3, –NH2, and no extra groups) were selected (Figure S1), which are designed to construct MOF nanosheets bearing intrinsic pores with close size but different groups. These nanosheets were denoted as MOF-CH3, MOF-NH2, and MOF-BDC, corresponding to the constructed organic ligands, respectively. AFM and SEM images in Figure 1a and Figures S2, S3 show that the synthesized MOF samples display typical sheet morphology, with uniform transverse size of ~ 1.5 μm and thickness of ~ 3.7 nm. And the XRD spectra (Figure 1b) show that three MOF nanosheets possess the same lattice structure, where the typical (200) peak at 2θ = 8.7° reflects the intrinsic pores.[37] This agrees with the HRTEM images in Figure S4, which visually show the lattice spacing of ~ 1.05 nm for these nanosheets.[38,39] Furthermore, N2sorption/desorption spectra, Figures 1c and S5, reveal that the average pore sizes for MOF-CH3, MOF-NH2 and MOF-BDC nanosheets are 1.12 nm, 1.05 nm, and 1.29 nm, respectively. It should be noted that the pore size of MOF-CH3 and MOF-NH2 nanosheets is smaller than that of MOF-BDC nanosheet, which is originated from the steric effect of –CH3 and –NH2 groups on the pores.[40,41]
FTIR result (Figure S6) shows that, compared with MOF-BDC nanosheet, the typical C=O peak gives a blue shift for MOF-CH3 and MOF-NH2 nanosheets at 1360 ~ 1380 cm-1, owing to the presence of –CH3/–NH2 groups adjacent to C=O group. Meanwhile, the grafting of –NH2 groups bring a new strong peak at 1253 cm-1, which assigned to Ar–N on FTIR and a typical characteristic peak of N element at 400 eV in XPS spectra of MOF-NH2 (Figure S7).[42] While the decorating of –CH3groups is supported by the higher C element content for MOF-CH3 relative to MOF-BDC. Specifically, the content of N element on MOF-NH2 nanosheet is tested to be 2.95%, matching well with the theoretical value calculated from lattice structure. This also implies that there is one –NH2 group in a unit cell,[41]and the condition is identical for MOF-CH3 nanosheet (Figures 1g and i), this implies the topological structure of MOF is explicitly constructed during synthetic process. Note that there is no extra group in the pore of MOF-BDC nanosheet, which is designed as a comparison.