مواد هیبریدی آلی- غیر آلی هترومتالیک مبتنی بر پلی اگزومتالات برای جذب سریع و تفکیک انتخابی متیلن بلو از محلول های آبی Polyoxometalates-based heterometallic organic– inorganic hybrid materials for rapid adsorption and selective separation of methylene blue from aqueous solutions

In this study, a series of new organic–inorganic hybrid monovacant Keggin-type silicotungstates (LnCu–POMs) containing 3d–۴f lanthanideIII–CuII heterometallic cations were hydrothermally synthesized with a mixture containing K8[a-SiW11O39]13H2O, CuCl22H2O, LnCl36H2O (Ln = Dy, 1 and 2; Er, 3 and 4) and en (1–۳) (or DETA (4)) [en = ethylenediamine, DETA = diethylenetriamine], and further characterized by PXRD, FTIR, and thermogravimetric analyses (TGA). The compounds 2 and 3 were obtained when oxalic acid was respectively added in the reaction mixtures used for 1 with en and 4 with DETA. It is well known in POMs systems that the reaction pH values strongly affect the formation of various structure types.13 Single-crystal X-ray diffraction analyses (Table S1, ESI†) reveal that compounds Dy-1 and Er-4 crystallize in the monoclinic space group Cm and P21c, respectively. Compounds Dy-2 and Er-3 are isostructural and crystallize in the orthorhombic space group Fdd2. Based on elemental analyses, single-crystal X-ray diffraction, thermogravimetric analyses (TGA), and charge-balance considerations (see the ESI†), their formula were confirmed as [K2(H2O)6.5][Cu(en)2]2[Dy(H2O)2SiW11O39]2[Cu(H2O)(en)2]210H2O (Dy-1), [Cu(en)2]3[Ln(H2O)SiW11O39]2(C2O4)[Cu(H2O)(en)2]310H2O (Ln = Dy-2, Er-3), [Cu(H2O)(DETA)2]H11[Er(SiW11O39)2]2DETA6H2O (Er-4). Their asymmetric units contain lacunary [a-SiW11O39] 8 POMs, in which a lanthanide cation occupies the vacant site and connects the terminal oxygen atoms of the POMs. Considering the charge balance, some protons should be added to Er-4. To locate the positions of these protons, bond valence sum (BVS) calculations14 have been performed on all the atoms of the POMs framework. The results show that all Si, W and Er atoms are in their formal valences of +4, +6 and +3, respectively. Except that the BVS values (1.60 for O5, 1.54 for O46, 1.60 for O58, 1.60 for O92, 1.57 for O93 and 1.54 for O95) in Er-4 are significantly lower than 2 (Table S2, ESI†), indicating they may be monoprotonated, the other O atoms are in the oxidation states of 2. The molecular unit of Dy-1 consists of one [K2(H2O)6.5] 2+ dimer, two unique [Dy(H2O)2SiW11O39] 5 polyanions, two coordinated [Cu(en)2] 2+, two discrete [Cu(H2O)(en)2] 2+ and ten water molecules of crystallization (Fig. S1, ESI†). Two coordinated [Cu(en)2] 2+ cations link to the Dy–POMs fragment via its terminal oxygen atoms. All Cu2+ ions locate on the special sites with 50% occupancy and adopt the square pyramid geometry with four N atoms from two en ligands and one oxygen atom from [Dy(H2O)2(SiW11O39)]5 or one aqua ligand. The neighboring [Dy(H2O)2(SiW11O39)]5 polyanion units are bridged into a one-dimensional chain structure by [Dy(H2O)2] 3+ and [K2(H2O)6.5] dimers (Fig. 1 and Fig. S1, ESI†). Each Dy3+ is seven-coordinate with two coordinated oxygen atoms from water molecules and five from two [SiW11O39] 8 fragments. The crystal structures of Dy-2 and Er-3 also contain two similar POMs fragments [Ln(H2O)(SiW11O39)]5, three linked [Cu(en)2] 2+ ions, three discrete [Cu(H2O)(en)2] 2+ ions, one {C2O4} 2 ion and ten lattice water molecules (Fig. S2, ESI†). The difference is that, in 2 and 3, such [Ln(H2O)(SiW11O39)]5 polyanion fragments were linked into a large Ln–POMs dimer cluster by {C2O4} 2 ligands in face-to-face mode (Fig. 1b). However, two monovacant [SiW11O39] 8 fragments in Er-4 were connected by an eight-coordinate Er3+ cation into a well-known sandwich-type {[SiW11O39] 8–Er–[SiW11O39] 8} structure (Fig. 1 and Fig. S3, ESI†). The Er3+ coordination cation adopts a distorted square antiprismatic geometry. Free water molecules are filled in the space among the POMs. Finally, 3D extended supramolecular architectures for 1–۴ were formed in view of hydrogen bonding interactions between nitrogen atoms of en or DETA molecules and surface oxygen atoms of POM fragments or water molecules. The experimental PXRD patterns for 1–۴ are in good agreement with the simulated patterns from the single-crystal X-ray diffraction, demonstrating the good phase purities for 1–۴٫ To evaluate the adsorption ability of Dy-1, Dy-2, Er-3 and Er-4 for the removal of dye from contaminated water, methylene blue (MB), methyl orange (MO) and rhodamine-B (RhB) with different charges and sizes as the typical organic pollutant targets were selected for experiments. The digital images and UV-vis spectroscopic results show that the removal of MO and RhB dyes is almost negligible for 1–۴ (Fig. S4, ESI†), but MB dye is almost completely adsorbed. To systematically investigate the adsorption behavior of the four compounds towards MB dye, the UV-vis absorption spectra of MB dye solution (20 mg L1 , 20 mL) in the presence of 1–۴ with different concentrations and constant time were compared. As shown in Fig. 2 and Fig. S5 (ESI†), the removal efficiency of Er-3 is the best and MB has been almost completely adsorbed with 1 mg of Er-3. The order of adsorption efficiency is Er-3 4 Dy-2 4 Dy-1 4 Er-4. The UV-vis spectra of pure MB solution with different concentrations was also obtained and used to determine the uptake capacity of four compounds (Fig. S6, ESI†). The uptake capacities of 193.6 mg g1 for Dy-1, 218 mg g1 for Dy-2, 391.3 mg g1 for Er-3, and 114.3 mg g1 for Er-4 were achieved at room temperature. These uptake capacities are considerably higher than that of commercial activated carbon6b and also considerably high compared to that of other materials reported up to now (Table 1). Fast adsorption rate, such that a material can remove most of the targeted dyes in a short time, is also a very important parameter to assess the efficiency and practicability of adsorbent in an economical wastewater disposal system. It is a very exciting discovery that the adsorption rate for Dy-1 reached to 90% in 1 min (Fig. 3a), and adsorption is almost complete in 15 min; compounds Dy-2 and Er-3 also completely remove MB molecules in 1 h (Fig. 3b and Fig. S7, ESI†). These results revealed that they were efficiently able to remove the MB molecules. As can be seen from the abovementioned experiments, compounds 1–۴ clearly exhibit superior adsorption properties for MB dyes compared with MO and RhB dye molecules. As expected, a large number of negative charges on the surface of the four LnCu– POMs may cause them to preferentially adsorb cationic MB-dye rather than anionic MO dye. Such a rapid adsorption rate for Dy-1 towards MB dyes was suspected to be due to the large amounts of dye adsorbed on the surface of sample particles. However, such an assumption does not explain why RhB with the same positive charge represents little uptake capacity. To deeply understand the removal process of dyes and clarify that they were adsorbed on the surface of sample particles or exchanged into the interior of crystals, a series of detailed studies have been performed. The solid particles and upper clear solution after the adsorption ofMB dye byDy-1were analyzedvia elemental analysis (EA) of S as well as inductively coupled plasma (ICP) spectrometry for Cu and K. EA results indicate an S amount of about 0.52% (slightly more than one MB cation) can be detected, which is considerably smaller than the experimental result (2.17%, 4.7 MB cations). ICP results (Cu, 16.6 mg mL1 ; K, 9.6 mg mL1 ) for the upper clear solution are considerably in accordance with the structural information of Dy-1 containing two isolated [Cu(H2O)(en)2] 2+ (a half of 36.7 mg mL1 based on all Cu) and two K+ ions (total 11.3 mg mL1 based on K). In contrast, Dy-1 was immersed into the aqueous solution for 24 h, and then the upper clear solution was also analyzed by ICP. The result is that Cu and K are not detected. The abovementioned analyses confirm that the adsorption for Dy-1 towards MB dye is indeed an ion-exchange process, where the free cations of K+ and [Cu(en)2] 2+ can be released during the adsorption of MB, and the adsorbed MB molecules may be exchanged into the interior of the Dy-1 crystal. At the same time, to balance the changes, about fiveMB cations in theorymay be adsorbed byDy-1,i.e., oneDy-1 molecule may adsorb five MB cations; thus, the adsorption quantity for Dy-1 is about 20.6%, which is very close to the experimental result (19.35%). However, the pore volume ratio for Dy-1 is 18.8% based on the calculation performed using the PLATON program,15 and the corresponding total potential solvent accessible void volume is 1166.8 Å۳ out of the 6209.3 Å۳ unit cell volume. However, the molecular volume of MB is B419.9 Å۳ , and thus most of the B2.7 mol MB cations may be ideally inserted into the void space of Dy-1. Combining the EA result of S in the interior of crystals after adsorption, it could be inferred that most of the adsorbed MB molecules should be on the surface of sample particles. From the experimental results and structural information of 1–۴, the amounts of negative charges on the surface for 1–۴ play a key role in the adsorption process. Omitting the isolated cationic part, the CuDy– POMs of Dy-1 represents 4 negative charges, CuEr–POMs of Er-3 contains 6 negative charges, and CuEr–POM of Er-4 contains 3 negative charges, which is consistent with the experimental order of adsorption efficiency and capacity of the target compounds (Er-3 4 Dy-1 4 Er-4) towards MB dyes. Thus, the anionic MO can be not naturally adsorbed. To further investigate the reason of negligible adsorption for RhB with positive changes, another cationic dye, basic red 2 (BR2) with similar size but different functional groups (Fig. S8, ESI†), has been selected for a comparative experiment. As shown in Fig. S9 (ESI†), the UV-vis spectra of the aqueous solution of BR2 (20 mg L1 ) dye with Dy-1 for the given time interval as well as with different compounds Dy-2, Er-3 and Er-4 indicate that the adsorption of BR2 with 1–۴ was negligible. The ICP data of the upper solution after adsorption for Dy-1 towards RhB and BR2 dyes show that K+ and Cu2+ from the [Cu(en)2] 2+ are almost not detected. The result further confirms that the ion-exchange is the key for the occurrence of adsorption, although most of the dyes are adsorbed on the surface of the samples. MB molecules with small size may be exchanged into the void space of the crystals, triggering the process, but RhB and BR2 are not due to their large sizes and steric hindrance. Because it is very difficult to in situ monitor this adsorption process, the removal mechanism of dyes is still not completely clear at this moment. To further confirm whether Dy-1, Dy-2, Er-3 and Er-4 materials have the ability to separate and recoverMB dyes frommixed dye solution, themixed dye solution ofMB andRhB,MB andMOwere prepared and used (Fig. 4 and Fig. S10a, ESI†). UV/vis spectra were measured to determine the separation capability of 1–۴, which shows that all the absorption peaks of MB molecules quickly disappeared, just leaving the characteristic absorption peaks of RhB and MO exposed to the corresponding dye-mixture (Fig. 4 and Fig. S10b–i, ESI†). The digital images of dye-selective adsorption clearly shows the color change from green for mixed MB–MO and purple for MB–RhB to pure orange forMO and pink forRhB solution,which can be readily observed by the naked eye (Fig. S10j and k, ESI†). Through the abovementioned discussion, it can be concluded that the materials exhibit excellent selective adsorption ability towards the cationic MB dyes in wastewater.