Postsynthetic Modification of UiO-66 with Perfluoroalkyl for Adsorbing Organic Pollutants
Being the most important resource in the world, water plays the main role in our living conditions. Up to now, various approaches, including superhydro-phobic and superoleophilic membrane separation[2-3], photocatalysis degradation, powder generation, and biological separation are used to remove pollutants from water. Among all these methods, adsorption is a promising and effective way to deal with pollutants in water, both soluble and insoluble[7-8]. At the same time, most adsorbent materials have the advantages of low cost, good recyclability, and environmentally friendly feature[9-11]. Common adsorbents in industry are natural materials, such as corn straw and cotton, and synthetic materials, such as polyurethane foam and graphene[12-16]. In general, porosity and stability are essential for adsorbents. Metal-organic frameworks (MOFs), being a kind of composite material with high porosity, are a class of potential adsorbents.
Since their advent, MOFs have been widely applied in many fields owing to their special connection and abundant pore structure. The as-synthesized MOFs can be applied in the fields of gas separation[17-18], dye adsorption[19-20], fluorescence, sensing, catalysis, drug delivery, and supercapa-citors due to the different properties of their nodes and organic ligands. In addition to the direct use of MOFs, their properties can also be altered by treatment like carbonization, combining with polymer or graphene materials. However, as adsorbent, the weak water stability of MOFs is an obvious defect. To improve the water resistance of MOF materials, different methods were adopted. The first method is using polystyrene as the template and surface-modifying material to construct macropores and hydrophobic surface of MOFs. The water contact angle of this material with hollow structure is 97°, which shows hydrophobicity. The second one is coating MOFs′ surface with hydrophobic polydimethysiloxane (PDMS) by a facile vapor deposition technique. In this way, the BET specific surface area of coated MOFs decreases slightly, while hydrophobicity is greatly improved. The third method is using the modified ligand to improve the water stability of MOFs. For example, the frameworks with ligand containing hydrophobic fluorine are water stable[30-31]. However, the selection of specific or modified ligands generally results in reconsidering the synthetic conditions of MOFs. To solve this problem, the modification of ligand can also occur after the formation of the frameworks. The frameworks through postsynthetic amidation modification with alkyl chains are also hydrophobic. In order to have considerable hydrophobicity, this kind of modification occurs mainly on the surface. Considering the integrity and convenience of the decoration, postsynthetic modification (PSM) is the best way to make MOFs hydrophobic, which can not only effectively and accurately control the pore size of MOFs, but also introduce specific functional groups. The material obtained in this way has better uniformity, which makes the material not affected by the decomposition or the loss of effective components in recycling. In order to get the higher substitution ratio, the modified molecules should have a suitable size.
In this work, we use postsynthetic modification on 2-aminoterephthalic acid in UiO-66, which is with high water stability. White and high crystalline UiO-66-NH2 powder was functionalized with hydrophobic perfluoroalkyl groups. This modification has been proved to be the most efficient method for amidation of amino-tagged MOFs, both in speediness and simplicity. Via this approach, the frameworks with large proportion of oleophilic ligands were obtained. The ratio of substitution decreased with the carbon chain of perfluoroalkyl lengthening, but still reaching over 70%. The contact angle of the modified materials reached around 90° and endowed the material with certain hydrophobicity. As a result, the adsorption amount of the obtained materials for non-polar organic solvents increased significantly, while the adsorption capacity for water reduced sharply. Thus, the hydrophobic treatment also improved the water stability of the materials, so that it could remove aromatic pollutants in water.
Zirconium tetrachloride (ZrCl4), trifluoroacetic anhydride (TFAA), and pentafluoropropionic anhydride (PFPA) were obtained from Alfa Aesar Chemical Co. 2-Aminoterephthalic acid (H2ATA) was obtained from J & K Scientific Ltd. Benzene, toluene, pentane, hexane, N, N-dimethylformamide (DMF), methanol (MeOH), acetone, and formic acid were obtained from Beijing Chemical Factory. All the reagents and solvents are commercially available without further purified.
1.2 Instrumental characterization
The 1H NMR and 13C NMR spectra of all frame-works and organic compounds were recorded on a Bruker Advance Ⅲ 400 NMR spectrometer (Bruker Daltonics Inc., Germany). The thermogravimetric analy-sis was investigated from room temperature to 800 ℃ with 10 ℃·min-1 ramp in air and nitrogen atmosphere (Diamond TG/DTA, PerkinElmer, U.S.A.). X-ray diffraction (XRD) patterns were obtained from D/MAX-TTRIII diffractometer (Rigaku, Japan) equipped with Cu Kα radiation (λ=0.154 056 nm, operating voltage=40 kV, operating current=200 mA, the scan rate=10°·min-1, 2θ=5°~50°). The IR spectra as KBr pellets were recorded on a Spectrum One FTIR spectro-meter (PerkinElmer, U.S.A.) in a range of 4 000~400 cm-1. The ultraviolet-visible (UV-Vis) curves were collected by Lambda 950 UV-Vis-NIR spectrometer (PerkinElmer, U.S.A.). Nitrogen adsorption-desorption isotherms were measured with 3-FLEX surface area and porosity analyzer (Micromeritics Instrument Corporation, USA) at 77 K. The sample was degassed for 12 h at 120 ℃.
1.3 Synthesis of fluoramide modified UiO-66
UiO-66-NH2 was synthesized following a typical report. Amidation was reacted between amino-tagged MOF and fluoric anhydride as described below. UiO-66-NH2 (100 mg) was dispersed in TFAA (5 mL) in a 25 mL flask. The mixture was stirred for 24 h by magnetic stirring at room temperature. The suspension was collected by centrifugation and washed by acetone for three times. The light yellow powder of UiO-66-F1 was obtained by drying at 60 ℃ under vacuum situation. The light yellow powder of UiO-66-F2 was synthesized by similar procedure using PFPA as the source of fluoric anhydride with reaction time extending to 72 h.
2. Results and discussion
2.1 Synthesis and characterization
The most effective approach to amidate amino-tagged MOFs is using anhydride without additional solvents at elevated temperature, which has been proved by previous research. UiO-66-NH2 was obtained by solvothermal reaction of H2ATA and ZrCl4. The modifications of UiO-66-NH2 using TFAA and PFPA are shown in Scheme 1. The acid resistance of UiO-66-NH2 is a prerequisite for modification because the reaction is carried out under strong acidic condition. The white crystals turn light yellow after amidation. This was due to the introduction of perfluorocarbon chains. With the carbon chain of anhydride extending, it is necessary to raise the temperature and prolong the reaction time to ensure the considerable substitution.
The yields of amidation modified samples were acquired by NMR with the method of decompounding in CsF/D2O/DMSO-d6 solution (24 mg of CsF dissolving in 450 μL DMSO-d6 and 250 μL D2O mixed solution). The specific operation procedures are listed in the Supporting Information. The PSM yields of UiO-66-NH2 modified with fluoric anhydride at different reaction conditions are shown in Fig. 1. For parent MOF and the corresponding modified frameworks UiO-66-F1 and UiO-66-F2, the appearance of peaks for benzene ring protons clearly testifies that the original and modified frameworks were equivalent. After formation of amide structure, the peaks of benzene ring protons display significant downfield shifts. Especially, the signal of the proton at o-position of amino group moved to lower chemical shift, which appears around 8.77 for independent ligand and 8.33 for ligand connecting with metal ions after decoration with the perfluoroalkyl group. Compared with free ligands (BDC-NH2, BDC-F1, and BDC-F2, BDC=dicarboxylbenzene), upfield shift appears in the spectra of decomposed frameworks, whether modified or not. This may be due to that the different electric influence of carboxylate radical and carboxyl group. The ratio of substitution was calculated by comparing the ratio of the hydrogen atom (corresponding to the C-3 position of aromatic ring) chemical shifts associated with the unmodified and modified ligands in the 1H NMR spectra of each independent acid-digested material (Fig. 1). The yields of PSM reaction were 96.1% and 74.3% for TFAA and PFPA, respectively. In spite of further raising the reaction temperature or prolonging the reaction time, no significant improve-ment in the PSM yield was obtained for the PFPA reaction. The molecule size increasing from TFAA to PFPA, listed in Table S1 (Supporting Information), may be the main cause for PSM yield (Calculated and visualized projection diameters can be found in Fig.S1).
UiO-66-F1 and UiO-66-F2 were characterized by XRD and FT-IR techniques. The XRD patterns of UiO-66 are exhibited in Fig. 2a. The patterns of UiO-66-F1 and UiO-66-F2 are similar to the one before modification and the one simulated by software. It means that the original UiO-66-NH2 crystal structure was still maintained after modification, although the crystallinity of the modified frameworks slightly decr-eased. Furthermore, the significant peak at 1 020 cm-1 (Fig. 2b) belongs to fluoroalkyl group, which indicates that the functional groups have been introduced into MOFs. Some peaks also changed due to the formation of amides. The small peak at 1 780 cm-1 was ascribed to amides in amidated frameworks, while the peaks at 1 720 and 1 610 cm-1 were assigned to stretching vibration of carbonyl group in carboxylic acid. The broad peak at 3 180 cm-1 was attributed to the stret-ching vibration of N-H bond in amide group, while the characteristic absorption bands arising from unmodified N-H were observed at 3 400 and 3 500 cm-1. The elemental mapping substantiated the expe-cted homogenous distribution of Zr, C, N, F, and O elements throughout the sample (Fig.S2).
2.2 Porosity properties
The porosity of these MOFs was characterized through the nitrogen sorption tests. Nitrogen adsorption-desorption isotherms of the frameworks measured at 77 K are shown in Fig. 3a. All the samples showed similar nitrogen sorption isotherms, which can be classified to type Ⅰ according to the IUPAC classification. The three frameworks are the typical microporous materials, whose porous features can be inferred from the rapid uptake at a low relative pressure (P/P0=0~0.10). Especially, the adsorption curves coincided with the corresponding desorption curves very well, which is consistent with literatures. The adsorption volumes of the modified materials at the low relative pressure (P/P0=0.05) were obviously smaller than that of the original material. This may be attributed to the fact that fluoramide group occupied the pore volume to some degree. With the increase in relative pressure, the adsorbed volumes of UiO-66-NH2 and UiO-66-F1 grew up slowly, while that of UiO-66-F2 almost remained unchanged. However, the adsorbed volumes of the three materials increased significantly at high relative pressure (P/P0=0.90~1.00), which was attributed to the macroporous structure formed by aggregation of the particles. The specific surface area of all three frameworks was characterized by BET model. As discussed in previous report, the BET specific surface area was calculated with the relative pressure (P/P0) ranging from 0.01 to 0.10 for microporous materials. Like other postsynthetic modified MOFs, the shape of the sorption isotherms of the modified materials was basically the same as that of the original materials, but the specific surface area was significantly reduced. As the size of the modified functional group increased, from trifluoroa-cetyl to pentafluoropropionyl group, the BET specific surface area decreased from 1 120 to 810 and 610 m2·g-1. The pore volume also decreased by introducing of the substituents. To compare the changes in the pore structure of the MOFs before and after the modification, pore size distribution (PSD) profiles of all the frameworks were calculated by the nonlocal density function theory (NLDFT) approach (Fig. 3b). There were three kinds of pore size corresponded to separated channel structure. Taking UiO-66-NH2 for example, pore width at 0.6 nm was the characteristic width for triangular window, while 0.8 nm was for tetrahedral cage and 1.1 nm was for octahedral cage. All the pore sizes calculated by density functional theory (DFT) approach were well coincided with single crystal structure. The PSD profiles of modified MOFs had not much difference from that of UiO-66-NH2, but the pore size decreased slightly. The selected NLDFT model is not sensitive to the modified functional groups. In addition, perfluoroalkyl occupied a certain pore volume, resulting in a sharp decrease in pore volume, and the capacity of gas adsorption decreased while the weight of the material increased, causing the specific surface area of BET changed[41-42].
2.3 Thermal stability
Thermogravimetric analyses were carried out on all three frameworks in air and nitrogen atmosphere. In both condition, the powder of samples was heated from room temperature to 800 ℃, with the heating rate at 10 ℃·min-1. The results summarized in Fig.S3 investigated the thermal decomposition and degradation behavior of different MOFs. In air atmosphere, the modified frameworks showed the first mass loss about 25% at the temperature of 200 ℃, which is much lower than the mass loss appeared in UiO-66-NH2 (370 ℃), indicating the modified functional group dissociated from the frameworks firstly. Furthermore, both UiO-66-F1 and UiO-66-F2 decomposed at the same temperature as UiO-66-NH2, resulting in a second weight loss. Their final mass remnants are 49.89%, 37.21%, and 31.23% for UiO-66-NH2, UiO-66-F1, UiO-66-F2 at 600 ℃, respectively. Under air conditions, organic components are oxidized to gases, while metal ions are oxidized to deposit in the form of ZrO2. The percentage of deposition was in agreement with the yield of PSM given by NMR (43.03%, 32.23%, and 28.50% for UiO-66-NH2, UiO-66-F1, and UiO-66-F2, calculated from the weight percentage of transformed metal oxide in different frameworks). Under a nitrogen atmosphere, we can observe that the weight loss of the three materials ranged from 200 to 600 ℃, and the final weight remained slightly more than 50%, indicating the similar thermal stability of materials. The mass loss was due to the rich oxygen element in the materials.
Water contact angles of these three frameworks were measured by using the pellets of different MOFs. By compressing with 25 MPa, the similar micromor-phology of sample pellets helps to exclude the effect of roughness on the wettability of materials (Fig.S4)[43-45]. The results proved the hydrophobicity of the modified materials (Fig. 4). The water contact angles of UiO-66-F1 and UiO-66-F2 were 90.5° and 88.6°, respectively, whereas UiO-66-NH2 had a water contact angle of 11.2°. The increase in water contact angle was due to the introduction of hydrophobic functional groups into the frameworks. Although the fluorine content of pentafluoropropionyl group is higher and its hydrophobicity should have been better than that of trifluoroacetyl group, the contact angles of the two modified MOFs are almost same. This is because the substitution ratio of UiO-66-F2 is lower, so its hydrophobicity is not superior to that of UiO-66-F1.
2.5 Adsorption capacities
The hydrophobicity of these frameworks makes them possess application prospects in the adsorption of organic liquids[46-47]. As it has been reported[8, 48], the adsorption capacity of different liquids was tested by differential gravity method. The specific approach was mentioned in the Supporting Information. Parent MOF and its modified frameworks were applied to study their adsorption of water and several aromatic and non-aromatic non-water-soluble solvents, with the method of sank in different solvent in powder form (Fig. 5). After modification, the frameworks, which were decorated with hydrophobic functional groups, display much lower water capacity less than 15%(w/w) compared with aminated MOF (around 110%(w/w)). Contrarily, UiO-66-F1 and UiO-66-F2 exhibited excellent adsorption capability for organic solvents, including oil as well. By virtue of their hydrophobicity, the adsorption capacities of fluorinated materials were almost twice higher than that of original structure. It can be easily understood that those water-insoluble compounds were absorbed heavily.
The adsorbents in pellet form are more convenient to be transferred and used in practice. In order to evaluate the modified materials more comprehensively, the solvent adsorption experiments of three MOFs in the form of pellets were also carried out (Fig.S5~Fig.S7). The adsorption capacities of UiO-66-NH2 in powder form for one-minute adsorption were comparable to those in pellet for one-hour adsorption. As test time increasing, UiO-66-NH2 in pellet form adsorbed more organic solvent. This elucidated that within seconds UiO-66-NH2 could not reach adsorption equilibrium. However, the adsorption capacities of UiO-66-F1 and UiO-66-F2 under three conditions were very close to each other. Albeit extending test time, UiO-66-F1 and UiO-66-F2 showed adsorption capacity similar to those for short-time adsorption, which indicated that the modified frameworks could achieve adsorption equilibrium within seconds and endowed better adsorption capacity for organic solvents.
According to the pore volume (Table S2), the calculated adsorption capacities of organic solvents by the three frameworks were lower than the experimental values. Therefore, the amount of adsorption can hardly exclude the contribution of macropore formed by particle stacking for liquid adsorption and the surface wetting of the materials. Taking advantage of the low surface energy of C-F bonds, the modified materials with hydrophobicity were more wettable in organic solvents. The maintaining on organic solvents of hydrophobic MOFs was much higher than that of unmodified one.
The recycling adsorption tests using benzene show that the powder of hydrophobic MOFs was still recyclable after ten cycles (Fig. 6). There was no significant capacity decrease after all cycles, indicating that the materials could be regenerated after simply vacuum drying at 70 ℃ for 1 h. The XRD measure-ments of the frameworks after recycling tests confirmed no evident change (Fig.S8). According to the nitrogen adsorption-desorption isotherms of modified frame-works, the porosity of hydrophobic MOFs was maintained (Fig.S9).
The adsorption capacity and cycling ability for benzene is remarkable, then, the adsorption of benzene in aqueous solution by MOFs was carried out. The powder of hydrophobic MOFs (2.0 mg) were dispersed in solution of benzene (10 mL) with different concentration (36 to 1 800 μg·g-1). After more than 2 h, the adsorbent was separated by centrifugation. The content of benzene in remaining solution were obtained by UV analysis (the specific UV spectra were listed in Fig.S10~Fig.S13). With the same temperature and humidity, after enough equilibration time, the modified MOFs can capture much more benzene at a lower equilibrium concentration (Fig. 7). The saturated adsorption capacities of aqueous solution by hydro-phobic MOFs were basically two times as that of UiO-66-NH2. The adsorption property of UiO-66-F1 was better than UiO-66-F2, no matter the saturated adsorption (302 mg·g-1 for UiO-66-F1 and 285 mg·g-1 for UiO-66-F2) or the equilibrium concentration. The above phenomena can be explained from the internal structure of the material. The amount of hydrophobic modified ligand unit in UiO-66-F2 was less than that in UiO-66-F1, because of its lower substitution ratio. Thus, even if pentafluoropropionyl group contains larger hydrophobic unit, UiO-66-F1 showed higher hydrophobicity, which is consistent with the results of contact angles. Besides, the pore volume of UiO-66-F1 was also larger than that of UiO-66-F2 (Table S2). UiO-66-NH2, without hydrophobic modification, could only capture 124 mg·g-1 of benzene, much lower than UiO-66-F1. For capturing the aromatic compounds in water, the substitution ratio of the functional group plays a major role. Fluorinated anhydride modified materials have potential applications in the removal of aromatic compounds from water.
We have reported a simple, convenient, and efficient route to post-synthesize hydrophobic MOFs that were modified with fluorinated anhydride. The hydrothermal stability and acid-resistance of UiO-66-NH2 help maintain the structure of the frameworks after modification. Compared with the parent MOF, the modified MOFs with hydrophobicity achieve higher adsorption capacity and excellent recycling perfor-mance for pure organic solvents, because these materials can reach equilibrium faster in the adsor-ption process. Significantly, the modified materials exhibit better adsorption behavior for aromatic comp-ounds in aqueous solution. There is a large room for the development of MOFs obtained by easy and energy-efficient postsynthetic modification for various applications in oil spill clean-up, removal of aromatic compounds, and even drug extraction.
Supporting information is available at http://www.wjhxxb.cn
Huang J K, Yan Z F. Langmuir, 2018, 34:1890-1898
Gu J C, Xiao P, Chen J, et al. J. Mater. Chem. A, 2014, 2:15268-15272
Cao C X, Yuan J Y, Cheng J P, et al. Sci. Rep., 2017, 7:3101(9 Pages)
Chong M N, Jin B, Chow C W K, et al. Water Res., 2010, 44:2997-3027
Datta S, Samanta S, Chaudhuri D. J. Mater. Chem. A, 2018, 6:2922-2926
Gupta S, Tai N H. J. Mater. Chem. A, 2016, 4:1550-1565
Huang Y P, Zhang W J, Ruan G H, et al. Langmuir, 2018, 34:3661-3668
Jing P, Zhang S Y, Chen W J, et al. Chem. Eur. J., 2018, 24:3754-3759
Stolz A, Floch S L, Reinert L, et al. Carbon, 2016, 107:198-208
Chu Y, Pan Q M. ACS Appl. Mater. Interfaces, 2012, 4:2420-2425
He Y Q, Liu Y, Wu T, et al. J. Hazard. Mater., 2013, 260:796-805
Bayat A, Aghamiri S F, Moheb A, et al. Chem. Eng. Technol., 2005, 28:1525-1528
Li H, Wu W, Bubakir M M, et al. J. Appl. Polym. Sci., 2014, 131:40081-40089
Radetic M M, Jocic D M, Jovancic P M, et al. Environ. Sci. Technol., 2003, 37:1008-1012
Zhang Y L, Wei S, Liu F J, et al. Nano Today, 2009, 4:135-142
Zhu Q, Chu Y, Wang Z K, et al. J. Mater. Chem. A, 2013, 1:5386-5393
Wang Z G, Ren H T, Zhang S X, et al. J. Mater. Chem. A, 2017, 5:10968-10977
Chen Z J, Feng L, Liu L M, et al. Langmuir, 2018, 34:14546-14551
Zhang W, Yang J M, Yang R N, et al. J. Mol. Liq., 2019, 283:160-166
Hu P, Zhao Z X, Sun X D, et al. Chem. Eng. J., 2019, 356:329-340
Wang X M, Wang P, Fan R Q, et al. Dalton Trans., 2015, 44:5179-5190
Li R, Ren X Q, Ma H W, et al. J. Mater. Chem. A, 2014, 2:5724-5729
Zhao J, Wang Q, Sun C Y, et al. J. Mater. Chem. A, 2017, 5:12498-12505
Wang H, Xu J, Zhang D S, et al. Angew. Chem. Int. Ed., 2015, 54:5966-5970
Wang L, Feng X, Ren L T, et al. J. Am. Chem. Soc., 2015, 137:4920-4923
Bhadra B N, Song J Y, Lee S K, et al. J. Hazard. Mater., 2018, 344:1069-1077
Zhang Y Y, Feng X, Li H W, et al. Angew. Chem. Int. Ed., 2015, 54:4259-4263
Sui Z Y, Zhang P Y, Xu M Y, et al. ACS Appl. Mater. Interfaces, 2017, 9:43171-43178
Zhang W, Hu Y L, Ge J, et al. J. Am. Chem. Soc., 2014, 136:16978-16981
Mukherjee S, Kansara A M, Saha D, et al. Chem. Eur. J., 2016, 22:10937-10943
Ji P F, Drake T, Murakami A, et al. J. Am. Chem. Soc., 2018, 140:10553-10561
Rubin H N, Reynolds M M. Inorg. Chem., 2017, 56:5266-5274
Liu T, Li D Q, Wang S Y, et al. Chem. Commun., 2014, 50:13261-13264
Abdelhameed R M, Simes M M Q, Silva A M S, et al. Chem. Eur. J., 2015, 21:11072-11081
Hintz H, Wuttke S. Chem. Mater., 2014, 26:6722-6728
Schaate A, Roy P, Godt A, et al. Chem. Eur. J., 2011, 17:6643-6651
Sing K S W, Everett D H, Haul R A W, et al. Pure Appl. Chem., 1985, 57:603-619
Servalli M, Ranocchiari M, Van Bokhoven J A. Chem. Commun., 2012, 48:1904-1906
Walton K S, Snurr R Q. J. Am. Chem. Soc., 2007, 129:8552-8556
Dau P V, Kim M, Cohen S M. Chem. Sci., 2013, 4:601-605
Britt D, Lee C, Uribe-Romo F J, et al. Inorg. Chem., 2010, 49:6387-6389
Doonan C J, Morris W, Furukawa H, et al. J. Am. Chem. Soc., 2009, 131:9492-9493
Bellanger H, Darmanin T, de Givenchy E T, et al. Chem. Rev., 2014, 114:2694-2716
Su B, Tian Y, Jiang L. J. Am. Chem. Soc., 2016, 138:1727-1748
Sun Q, Aguila B, Verma G, et al. Chem, 2016, 1:628-639
Wang B, Lv X L, Feng D, et al. J. Am. Chem. Soc., 2016, 138:6204-6216
Xie L H, Liu X M, He T, et al. Chem, 2018, 4:1911-1927
Jayaramulu K, Geyer F, Pert M, et al. Adv. Mater., 2017, 29:1605307
- PDF下载量: 3
- 文章访问数: 523
- HTML全文浏览量: 2