Template Method for Synthesizing Hierarchically Porous MIL-101(Cr) for Efficient Removal of Large Molecular Dye

As one of the most important prototypical chromium-based MOFs, MIL-101(Cr) is well-studied and widely employed in various scientific fields. However, due to its small capture window sizes and curved internal apertures, its application in large molecular removal is quite limited, and given its high stability and high synthetic temperature (>200 °C), it is difficult to achieve hierarchically porous MIL-101(Cr). In our study, hierarchically porous MIL-101(Cr) involving a high macro-/meso-/micropores ratio was designed and synthesized using acetic acid as an additive and silicon dioxide (SiO2) nanoparticles as a template. The optimal hierarchically porous MIL-101(Cr) (A-4) possessed a high specific surface area (2693 m2 g−1) and an abundant macro-/mesoporous structure with the addition of SiO2 of 200 mg. Compared with the control sample (A-0) with a less macro-/mesoporous structure, A-4 showed good adsorption properties for both coomassie brilliant blue R-250 (CBB, 82.1 mg g−1) and methylene blue (MB, 34.3 mg g−1) dyes, which were 1.36 times and 9.37 times higher than those of A-0. Moreover, A-4 also had good recyclability, and the removal rate of CBB was still higher than 85% after five cycles of adsorption.


Introduction
Different synthesis methods greatly influence the morphology, specific surface, and crystallinity of metal-organic framework materials (MOFs) [1]. The common synthesis methods of MOFs include hydrothermal synthesis [2,3], the solvent heat method [4,5], the microwave-assisted method [6], and the template method [7]. The template method is an important strategy for controlling the crystals' morphology and size. It is divided into hard and soft templates, depending on the characteristics of the template itself and its domain-limiting ability [8,9]. In general, the template method is commonly used to prepare hierarchically porous MOFs (HP-MOFs), which improve the performance of MOF materials and expand their application fields [7,10]. The prefabricated template is directly added to the precursor solution of the MOFs material, and the MOFs crystals are then grown directly on the surface of the template [11]. The template occupies space in the crystal framework of MOFs, and the removal of the template results in the HP-MOFs containing extra porous structures such as mesopores or macropores [12,13].
The hierarchically porous materials are comprised of a multimodal hierarchically porous structure made of interconnected pores with different lengths, ranging from micro-(<2 nm) to meso-(2-50 nm) to macropores (>50 nm). Compared with conventional MOFs and microporous structures, HP-MOFs have a specific morphology and size, larger specific surface area, and greater pore volume [14,15]. HP-MOFs not only maintain the basic properties of the related pristine MOFs but also construct macro/mesoporous structures on the basis of micropores, which reduce macromolecular diffusion barriers and optimize the deficiencies of microporous pristine MOFs [16,17]. The HP-MOFs demonstrate excellent performance in the utilization of gas storage and separation [18], catalysis [19], drug mitigation [20], and adsorption [21][22][23]. Thus, the design and synthesis of HP-MOF materials purchased from Aladdin Chemical Reagent Co. All chemicals were used as obtained from commercial sources without further purification.

Characterization and Analysis
X-ray diffraction tests were performed on an X-ray diffractometer (D8 Advance, Bruker, Karlsruhe, Germany). A scanning electron microscope (Gemini 300, Zeiss, Jena, Germany) and transmission electron microscope (Talos F200X, FEI, Hillsboro, OR, USA) were used to characterize the samples' microscopic morphology and particle size. The BET-specific surface area and pore volume of the samples were determined via nitrogen adsorption/desorption experiments with a fully automated specific surface area and porosity analyzer (NOVA-4200e, Quantachrome, Boynton Beach, FL, USA). A UV-Vis spectrophotometer (UV-2600, Shimadzu, Suzhou, China) was employed to monitor the concentration of dyes during the adsorption experiments. All samples followed the same pretreatment before measurements.

Synthesis and Purification of Samples
To synthesize MIL-101 (Cr), we added 800 mg of Cr(NO 3 ) 3 ·9H 2 O (2 mmol), 332 mg of H 2 BDC (2 mmol), 8 mmol of acetic acid, and 0~200 mg SiO 2 (15 nm) to 10 ml of deionized water and stirred for 30 min. The mixture was then transferred into a polytetrafluoroethylene (PTFE) autoclave (20 mL, Youmai, Shanghai, China) and placed in the oven at 200 • C for 8 h. After the reaction, the autoclave was slowly cooled down to ambient temperature, and the green product was collected for purification. Samples of 0 mg, 50 mg, 100 mg, 150 mg, 200 mg, and 250 mg of SiO 2 were named A-0, A-1, A-2, A-3, A-4, and A-5, respectively. However, adding 250 mg of SiO 2 into the autoclave would have filled the whole container and overflowed when stirred. Thus, for security consideration, an A-5 experiment was not conducted, and 200 mg of SiO 2 was the largest sample in our study.
For purification, the green products were gathered and washed with hot DMF (70 • C) and hot ethanol (60 • C) twice (1 h each). The samples were the washed with HF aqueous (1%) for 12 h to remove SiO 2 . Finally, the solids were rinsed with deionized water at least 5 times to remove residual HF. The final products were dried in a vacuum oven at 120 • C for 12 h.

Dye Adsorption Experiments
We added 2.5 mg of adsorbent to 10 mL of dyes (MB and CBB) with different initial concentrations at ambient temperature. The mixture was then placed into an ultrasonicator and shaken for an appropriate time (~10 min). After centrifugation, the supernatant was taken for an absorbance test using a UV spectrophotometer. The remaining concentration of dyes in the supernatant was measured using a UV spectrophotometer.

1.
Adsorption capacity The adsorption capacity q e (mg g −1 ) of the adsorbent for the dyes was calculated using the formula shown in Equation (1) [49].
where c 0 and c e are the initial and equilibrium concentrations of the dye's aqueous solution (mg L −1 ), respectively. v is the volume of the dye solution (L), m is the mass of the adsorbent (g), and q e is the adsorption amount of the adsorbent on the dye molecules at adsorption equilibrium (mg g −1 ). The pseudo-first-order kinetic model considers a physical adsorption process between the adsorbent and the adsorbate [50,51]. It assumes that a diffusion step controls the arrival of the adsorbate from the solution to the adsorbent surface and that there is only one active site on the adsorbent surface. The linear expressions of the pseudo-first-order kinetic mode model are shown in Equation (2) [52].
where t is the adsorption time (min), q e is the adsorbed amount of adsorbent on the adsorbent's surface (mg g −1 ), q t is the adsorption amount of the dye molecule by the adsorbent at moment t (mg g −1 ), and k 1 (min −1 ) is the adsorption rate constant for the pseudo-first-order kinetic model.
(2) Pseudo-second-order kinetic model The pseudo-second-order kinetic model considers the chemisorption process between the adsorbent and the adsorbate [53,54]. It assumes that the chemisorption mechanism controls the adsorption rate and that there are two binding sites on the adsorbent surface. The linear expressions of the pseudo-second-order kinetic model are shown in Equation (3) [53].
where t is the adsorption time (min), q e is the adsorbed amount of the adsorbent on the adsorbent's surface (mg g −1 ), q t is the adsorption amount of the dye molecule by the adsorbent at moment t (mg g −1 ), and k 2 (g mg −1 min −1 ) is the adsorption rate constant for the pseudo-second-order kinetic model.

Isothermal adsorption model
(1) Langmuir model The Langmuir model is one of the commonly used adsorption models. It assumes that adsorption occurs uniformly on all active sites of the adsorbent and that the adsorbate is adsorbed on the surface of the adsorbent as a single molecular layer [55]. The linear expression of the Langmuir model is shown in Equation (4) [56].
where q m is the maximum adsorption of the adsorbent on the adsorbent's surface (mg g −1 ), c e is the concentration of the solution at the adsorbent's surface when it reaches adsorption equilibrium (mg L −1 ), and k L is the Langmuir equilibrium constant (L mg −1 ).
(2) Freundlich model The Freundlich model is based on the facts that the surface energy of the adsorbent is inhomogeneous and that the adsorption of the adsorbent on the adsorbent's surface is not a single monomolecular layer adsorption, but a multimolecular layer adsorption [57]. The linear expression of the Freundlich model is shown in (5) as follows [58].
where q e is the adsorbed amount of adsorbent on the adsorbent's surface (mg g −1 ), c e is the concentration of the solution at the adsorption equilibrium on the adsorbent's surface (mg L −1 ), k F is the Freundlich equilibrium constant representing the adsorption capacity (mg g −1 ), and n is the Freundlich equilibrium constant for the adsorption strength.

Results
X-ray diffraction (XRD) plots showed that all of the experimentally prepared samples were highly consistent with the simulated XRD pattern generated from the deposited X-ray data file at the Cambridge Structure Database (CSD-Refcode OCUNAK) using the Mercury program. Thus, all the samples were confirmed to be pure MIL-101(Cr) [59] (Figure 1). The nitrogen adsorption-desorption curves of the samples are shown in Figure 2, and the adsorption and desorption curves of the samples did not completely overlap. The hysteresis loops indicated that the pore structures of the samples were irregular and contained two or more types of pore structures [60]. The pore structure information of each sample is shown in Table 1. The specific surface area of the MIL-101(Cr), synthesized using only acetic acid as the modulating agent, was 3006 m 2 g −1 . However, the addition of SiO2 decreased the specific surface area of MIL-101(Cr). In particular, when the addition amount of SiO2 was 200 mg, the BET-specific surface area of A-4 was 2693 m 2 g −1 . Interestingly, adjusting the amount of SiO2 increased the ratio of macro-mesopores to micropores (Smacro-meso/Smicro). The Smacro-meso/Smicro ratio increased from 4.62:1 to 12.8:1 when the addition amount of SiO2 was raised from 50 mg to 200 mg (see Table 1). Combined with the Barret-Joyner-Halenda (BJH) pore size distribution curves of the samples (Figure 2b), we found that involving the addition of SiO2 as a template created more macro-/mesopores. Thus, A-4 had the highest Smacro-meso/Smicro ratio among all samples. The specific surface area was influenced by many factors, such as particle size and pore structure. In general, the smaller the particle size was and the simpler the pore structure of MIL-101(Cr) was, the higher its specific surface area was. In our study, A-4 contained the most abundant macro-/mesoporous structure and the most relatively large particle size; thus, it possessed relatively low specific surface area. The nitrogen adsorption-desorption curves of the samples are shown in Figure 2, and the adsorption and desorption curves of the samples did not completely overlap. The hysteresis loops indicated that the pore structures of the samples were irregular and contained two or more types of pore structures [60]. The pore structure information of each sample is shown in Table 1. The specific surface area of the MIL-101(Cr), synthesized using only acetic acid as the modulating agent, was 3006 m 2 g −1 . However, the addition of SiO 2 decreased the specific surface area of MIL-101(Cr). In particular, when the addition amount of SiO 2 was 200 mg, the BET-specific surface area of A-4 was 2693 m 2 g −1 . Interestingly, adjusting the amount of SiO 2 increased the ratio of macro-mesopores to micropores (S macro-meso /S micro ). The S macro-meso /S micro ratio increased from 4.62:1 to 12.8:1 when the addition amount of SiO 2 was raised from 50 mg to 200 mg (see Table 1). Combined with the Barret-Joyner-Halenda (BJH) pore size distribution curves of the samples (Figure 2b), we found that involving the addition of SiO 2 as a template created more macro-/mesopores. Thus, A-4 had the highest S macro-meso /S micro ratio among all samples. The specific surface area was influenced by many factors, such as particle size and pore structure. In general, the smaller the particle size was and the simpler the pore structure of MIL-101(Cr) was, the higher its specific surface area was. In our study, A-4 contained the most abundant macro-/mesoporous structure and the most relatively large particle size; thus, it possessed relatively low specific surface area. an error of ± 50 m 2 g −1 . b Calculated from the N 2 adsorption/desorption curves at 77 K and a relative pressure of p/p 0 = 0.95. c Calculated from the BJH desorption curves by integrating over the micropore and mesopore intervals S macro-meso /S micro .  a Calculated from the N2 adsorption/desorption curves in the relative pressure range of 0.05 < p/p0 < 0.2, with an error of ± 50 m 2 g −1 . b Calculated from the N2 adsorption/desorption curves at 77 K and a relative pressure of p/p0 = 0.95. c Calculated from the BJH desorption curves by integrating over the micropore and mesopore intervals Smacro-meso/Smicro.
Scanning electron microscope (SEM) and transmission electron microscope (TEM) images of the samples showed that the SiO2-free sample A-0 presented a uniform particle size of ~138 nm with a relatively regular morphology (Figure 3a,d and Figure S1a in the Supplementary Section). There was a weak nanofusion phenomenon at the boundary of the crystals and a trace of mesoporous structure (red dashed area). Samples A-2 and A-4 showed irregular polyhedral morphology, with the adjacent crystals showing nanofusion and forming mesoporous structures (Figure 3b,c,e,f). The difference was that A-4 had a higher degree of nanofusion and formed more abundant mesoporous structures (red dashed area). Scanning electron microscope (SEM) and transmission electron microscope (TEM) images of the samples showed that the SiO 2 -free sample A-0 presented a uniform particle size of~138 nm with a relatively regular morphology (Figure 3a,d and Figure S1a in the Supplementary Section). There was a weak nanofusion phenomenon at the boundary of the crystals and a trace of mesoporous structure (red dashed area). Samples A-2 and A-4 showed irregular polyhedral morphology, with the adjacent crystals showing nanofusion and forming mesoporous structures (Figure 3b,c,e,f). The difference was that A-4 had a higher degree of nanofusion and formed more abundant mesoporous structures (red dashed area). The adsorption properties of A-0, A-2, and A-4 were tested using aqueous solutions of methylene blue (7.9 Å × 16.3 Å, Figure S2a The adsorption properties of A-0, A-2, and A-4 were tested using aqueous solutions of methylene blue (7.9 Å × 16.3 Å, Figure S2a) and Coomassie Brilliant Blue (15.1 Å × 24.1 Å, Figure S2b in the Supplementary Section). A-4 showed the best adsorption capability toward CBB and MB among all of the samples, with the adsorption amounts of 82.1 mg g −1 and 34.3 mg g −1 (Figure 4), respectively, attributable to its high S macro-meso /S micro ratio structure. The more mesopores that a sample had, the better that the large molecular dyes were adsorbed. Hence, we were not surprised that A-4, which possessed the highest S macro-meso /S micro ratio, exhibited the best adsorption capacity for the large molecular dyes in CBB and MB. The adsorption properties of A-0, A-2, and A-4 were tested using aqueous solutions of methylene blue (7.9 Å × 16.3 Å, Figure S2a) and Coomassie Brilliant Blue (15.1 Å × 24.1 Å, Figure S2b in the Supplementary Section). A-4 showed the best adsorption capability toward CBB and MB among all of the samples, with the adsorption amounts of 82.1 mg g −1 and 34.3 mg g −1 (Figure 4), respectively, attributable to its high Smacro-meso/Smicro ratio structure. The more mesopores that a sample had, the better that the large molecular dyes were adsorbed. Hence, we were not surprised that A-4, which possessed the highest Smacromeso/Smicro ratio, exhibited the best adsorption capacity for the large molecular dyes in CBB and MB.  The effects of time and dye concentration on the adsorption capacity of A-4 were investigated, and the results are presented in Figure 5. During the first 60 min was the rapid adsorption phase, when the adsorption amount increased rapidly from 0 mg g −1 to 154.5 mg g −1 for CBB and from 0 mg g −1 to 50.0 mg g −1 for MB. This was followed by a slow increase until it finally reached the adsorption equilibrium, after which the adsorption amount was essentially unchanged. This was considered to be the maximum adsorption. Figure 5b shows the adsorption capacity of A-4 for different initial concentrations of CBB and MB (100 mg L −1~5 00 mg L −1 with an adsorption time of 24 h). As the dye concentration increased, the adsorption of A-4 on the dye molecules also increased, while the adsorption increment decreased. It is noteworthy that the adsorption capability of A-4 for CBB is much higher than that of MB ( Figure 5). This phenomenon may be caused by the positive charge of A-4 ( Figure S3 in the Supplementary Section) and the negative charge of the anionic dye CBB, which induces a strong electrostatic interaction between them. In contrast, since there was no such electrostatic interaction between the cationic dye MB and A-4, A-4 had better adsorption of CBB.
The results of fitting the above adsorption data using adsorption kinetics and adsorption isotherms are shown in Figure 6 and Table S1 in the Supplementary Section. The correlation coefficients R 2 of the pseudo-second-order kinetic fits of A-4 for the adsorption processes of CBB and MB were 0.993 and 0.989, which were higher than those of the pseudo-first-order kinetic fits (0.936 and 0.956). Furthermore, the theoretical adsorption amount of the pseudo-second-order kinetic was close to the actual adsorption amount, indicating that the adsorption processes of A-4 on CBB and MB were more consistent with the secondary kinetic model. For the adsorption isotherms, the correlation coefficients R 2 of the Langmuir model were 0.999 (CBB) and 0.982 (MB), respectively, much higher than those of the Freundlich model (0.689 and 0.888); the theoretical values were also very close to the actual values. This indicated that the Langmuir model could more accurately describe the adsorption process of A-4 on CBB and MB. Additionally, the cyclicity of A-4 was conducted using CBB as the model, and the results are presented in Figure S4 in the Supplementary Section [21,48,[61][62][63][64][65]. After five cycles, the removal efficiency of A-4 for CBB still reached more than 85%, indicating that A-4 had good reusability. Figure 5b shows the adsorption capacity of A-4 for different initial concentrations of CBB and MB (100 mg L −1~5 00 mg L −1 with an adsorption time of 24 h). As the dye concentration increased, the adsorption of A-4 on the dye molecules also increased, while the adsorption increment decreased. It is noteworthy that the adsorption capability of A-4 for CBB is much higher than that of MB ( Figure 5). This phenomenon may be caused by the positive charge of A-4 ( Figure S3 in the Supplementary Section) and the negative charge of the anionic dye CBB, which induces a strong electrostatic interaction between them. In contrast, since there was no such electrostatic interaction between the cationic dye MB and A-4, A-4 had better adsorption of CBB. The results of fitting the above adsorption data using adsorption kinetics and adsorption isotherms are shown in Figure 6 and Table S1 in the Supplementary Section. The correlation coefficients R 2 of the pseudo-second-order kinetic fits of A-4 for the adsorption processes of CBB and MB were 0.993 and 0.989, which were higher than those of the pseudo-first-order kinetic fits (0.936 and 0.956). Furthermore, the theoretical adsorption amount of the pseudo-second-order kinetic was close to the actual adsorption amount, indicating that the adsorption processes of A-4 on CBB and MB were more consistent with the secondary kinetic model. For the adsorption isotherms, the correlation coefficients R 2 of the Langmuir model were 0.999 (CBB) and 0.982 (MB), respectively, much higher than those of the Freundlich model (0.689 and 0.888); the theoretical values were also very close to the actual values. This indicated that the Langmuir model could more accurately describe the adsorption process of A-4 on CBB and MB. Additionally, the cyclicity of A-4 was conducted using CBB as the model, and the results are presented in Figure S4 in the Supplementary Section [21,48,[61][62][63][64][65]. After five cycles, the removal efficiency of A-4 for CBB still reached more than 85%, indicating that A-4 had good reusability.

Conclusions
In summary, hierarchically porous MIL-101(Cr) containing a rich macro-/mesoporous structure can be obtained by using SiO2 as a template, whose Smacro-meso/Smicro ratio could be up to 12.8:1. Under optimal synthesis conditions, the obtained HP-MIL-101(Cr) (A-4) possessed a relatively high specific surface area and a large molecular dyes' adsorption capacity toward CBB and MB. Compared with non-template synthesized MIL-101(Cr) (A-0), the significantly enhanced adsorption capability of A-4 for dyes was attributed to the richer macro-/mesoporous structure, which could notably increase the

Conclusions
In summary, hierarchically porous MIL-101(Cr) containing a rich macro-/mesoporous structure can be obtained by using SiO 2 as a template, whose S macro-meso /S micro ratio could be up to 12.8:1. Under optimal synthesis conditions, the obtained HP-MIL-101(Cr) (A-4) possessed a relatively high specific surface area and a large molecular dyes' adsorption capacity toward CBB and MB. Compared with non-template synthesized MIL-101(Cr) (A-0), the significantly enhanced adsorption capability of A-4 for dyes was attributed to the richer macro-/mesoporous structure, which could notably increase the mass transfer capacity of MIL-101(Cr) in aqueous solutions. The demonstrated template synthesis broadened the application of MIL-101(Cr) in effluent treatment fields, especially in terms of large molecular organic dyes' removal.