2.5.4. Permselectivity measurements
The details of the ED performance of the membranes are described in our previous paper. 22 Briefly, ED experiments were performed at a current density of 10 mA cm−2. The selective layer faced the diluted chamber. First, 100 mL of 0.3 M Na2SO4 solution, 100 mL of a solution that contained 0.1 M NaCl (or 0.1 M LiCl) and 0.1 M MgCl2, and 200 mL of a 0.01 M KCl solution were added to the electrode compartments, diluted chamber, and concentrated chamber, respectively. All solutions were circulated at flow rates of 86 mL min−1 using peristaltic pumps. The ED tests were performed for 1 h. Samples were collected from the concentrated chamber for further analysis, and their ion concentrations were measured using ICP-OES.
Ion permeation through membranes (\(J_{N^{n+}}\), mol cm−2 s−1), which measures the changes in the concentration of ions in the concentrated chamber, was calculated as follows:
\(J_{N^{n+}}=\frac{\left(C_{t}\ -\ C_{0}\right)\ \bullet\ V}{A_{m}\ \bullet\ t}\), (1)
where \(C_{0}\ \)and \(C_{t}\) are the molar concentrations (M) of the ions (Nn+) in the concentrated chamber at the beginning (t = 0 min) and end (t = 60 min), respectively, of the ED test, \(V\) is the volume of solution in the concentrated chamber (200 mL), and \(A_{m}\) is the effective surface area of the membrane (7.07 cm2).
The permselectivity was calculated using the following equation:
\(P_{M^{+}{/D}^{2+}}=\frac{J_{M^{+}\ }\bullet\ C_{D^{2+}}}{J_{D^{2+\ }}{\bullet\ C}_{M^{+}}}\), (2)
where \(J_{M^{+}}\) and \(J_{D^{2+}}\) (mol cm−2s−1) are the permeations of monovalent and divalent cations, respectively, through the membrane after 60 min of testing and\(C_{M^{+}}\) and \(C_{D^{2+}}\ \)(M) are the average concentrations of monovalent (Li+ or Na+) and divalent (Mg2+) cations in the diluted chamber, respectively.
In this study, three sets of membranes were tested, and their separation performances were calculated as arithmetic averages. The testing device was thoroughly washed with DI water for 30 min after each test.
3. Results and discussion
3.1. Characterization of UiO-66(Zr)-NH2 andUiO-66(Zr/Ti)-NH2
The water stability of UiO-66(Zr)-NH2 was tested by soaking its powder in water, and the XRD results confirmed the excellent stability of UiO-66(Zr)-NH2 (Supporting Information, Figure S1). UiO-66(Zr/Ti)-NH2 nanoparticles were produced via the post-synthetic exchange of Zr4+ ions with Ti3+ ions using a TiCl3 solution. The SEM micrographs of UiO-66(Zr)-NH2 and UiO-66(Zr/Ti)-NH2(Figures 1a and b, respectively) revealed that the size of their particles was approximately 60 nm. However, the post-synthetic ion exchange process rendered the surface of the UiO-66(Zr/Ti)-NH2particles relatively rough. The XRD patterns of UiO-66(Zr)-NH2 and UiO-66(Zr/Ti)-NH2 (Figure 1c) presented sharp and intense diffraction peaks without visible peak shifting, which confirmed the highly crystalline structure of the UiO-66(Zr)-NH2and UiO-66(Zr/Ti)-NH2 nanoparticles. However, the intensities of the XRD peaks of UiO-66(Zr/Ti)-NH2 were lower than those of UiO-66(Zr)-NH2, which indicated the formation of defects during the post-synthetic process.30 FTIR spectra were used to analyze the structure of the UiO-66(Zr)-NH2 and UiO-66(Zr/Ti)-NH2nanoparticles. The negligible shifts of the characteristic FTIR peaks indicated that the post-synthetic process did not change the chemical structure of UiO-66(Zr)-NH2 (Supporting Information, Figure S2). The N2 adsorption–desorption isotherms (Figure 1d) indicated that the specific area of the nanoparticles increased from 764 m2 g−1 for UiO-66(Zr)-NH2 to 1168 m2g−1 for UiO-66(Zr/Ti)-NH2, which further confirmed the changes in the properties of the MOF nanoparticles after the post-synthetic process. The pore size distributions of the UiO-66(Zr)-NH2 and UiO-66(Zr/Ti)-NH2nanoparticles were calculated from the N2adsorption–desorption isotherms using the Saito–Foley (SF) and DFT models, and the results revealed the presence of inherent pores with sizes in the rage of 7-11 Å in the MOF nanoparticles (Supporting Information, Figures S3 and S4, respectively). This pore size distribution enabled the facile permeation and sieving of the Na+, Li+, and Mg2+ ions, which presented hydrated diameters of 7.2, 7.6, and 8.6 Å, respectively.29 The analysis of the SRPES profiles of the samples provided critical details on the post-synthetic process of the UiO-66(Zr)-NH2nanoparticles, such as the suppression of the Zr peak and emergence of the Ti peak (Figure 1e), which implied the successful replacement of Zr4+ ions with Ti3+ ions in UiO-66(Zr/Ti)-NH2. To analyze the XPS profiles of the nanoparticles, we used Ti4+ ions to exchange a fraction of the Zr4+ ions in UiO-66(Zr)-NH2 using the same post-synthesis process (Supporting Information, Figure S5). The 458.5 and 464.3 eV peaks in the Ti 2p XPS profile of UiO-66(Zr/Ti)-NH2-Ti4+ were ascribed to Ti4+ 2p3/2 and Ti4+ 2p3/2, respectively, (Supporting Information, Figure S5b). The SRPES profiles of UiO-66(Zr/Ti)-NH2 in the Ti 2p region shows two new peaks at 457.8 eV and 463.6 eV, corresponding to Ti3+2p3/2 and Ti3+ 2p3/2, respectively, revealing the coexistence of the Ti3+and Ti4+ions in the structure of UiO-66(Zr/Ti)-NH2, and that was ascribed to the oxidation of TiCl3 to TiCl4 in air (Figure 1f). These results indicated the successful replacement of Zr4+ ions with Ti3+ ions in UiO-66(Zr/Ti)-NH2. Furthermore, the Ti3+/Ti4+ molar ratio of UiO-66(Zr/Ti)-NH2 was determined to be approximately 6:5. In addition, the post-synthetic process was monitored using ICP-OES,31 and the results indicated that the Ti/Zr molar ratio of UiO-66(Zr/Ti)-NH2 was 9:4. Moreover, the zeta potentials of the UiO-66(Zr)-NH2 and UiO-66(Zr/Ti)-NH2 nanoparticles were tested (Supporting Information, Table S1). The decrease in zeta potential from 42.5 for UiO-66(Zr)-NH2 to 31.5 for UiO-66(Zr/Ti)-NH2 confirmed the exchange of Zr4+ ions with Ti3+ ions and successful conversion of UiO-66(Zr)-NH2 into UiO-66(Zr/Ti)-NH2.