Amino Acid Imprinted UiO-66s for Highly Recognized Adsorption of

Jun 5, 2019 - Solvothermal synthesis of defect-free UiO-66 is based on the .... In the case of amino acid imprinted UiO-66, Phe and Trp with rigid ...
0 downloads 0 Views 4MB Size
Download PDF
Research Article Cite This: ACS Appl. Mater. Interfaces 2019, 11, 23039−23049

www.acsami.org

Amino Acid Imprinted UiO-66s for Highly Recognized Adsorption of Small Angiotensin-Converting-Enzyme-Inhibitory Peptides Long Liu,†,∥ Zhiwei Qiao,‡,∥ Xinfang Cui,† Chunjiao Pang,† Hong Liang,‡ Peng Xie,† Xuan Luo,† Zuqiang Huang,† Yanjuan Zhang,† and Zhongxing Zhao*,†,§ †

School of Chemistry and Chemical Engineering, Guangxi University, Nanning 530004, China Guangzhou Key Laboratory for New Energy and Green Catalysis, School of Chemistry and Chemical Engineering, Guangzhou University, Guangzhou 510006, China § Guangxi Key Laboratory for Electrochemical Energy Materials, Guangxi University, Nanning 530004, China Downloaded via BUFFALO STATE on July 24, 2019 at 03:37:56 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.

‡

S Supporting Information *

ABSTRACT: Introduction of targeted defects into microporous UiO-66s for manipulating their three-dimensional size and surface properties can endow them with adsorption and separation areas involving angiotensin-converting-enzyme-inhibitory (ACE-inhibitory) peptides. Three hydrophobic amino acids (AAs) (i.e., proline (Pro), phenylalanine (Phe), and tryptophan (Trp)) having different physical/chemical properties were applied to in situ tailor defects in UiO-66 through targeted incoordination of missing linkers or missing nodes. Characterization results revealed a uniform oval shape of the developed defects with lengths ranging from 1.8 to 3.1 nm, which was also highly consistent with our molecular simulation. Among these three defective UiO-66s, Phe and Trp imprinted UiO-66s significantly promoted the adsorption affinity of small ACE-inhibitory peptides (uptake: 1.25 mmol g−1 for DDFF and 1.37 mmol g−1 for DDWW) and ultrahigh selectivity for DDFF (249) or DDWW (279) from inactive KKKK solution based on a lock-and-key mechanism. As a result, the imprinted UiO-66 showed an enrichment capacity for ACE-inhibitory peptides about eight times higher than that of pristine UiO-66. Therefore, the amino acid imprinting strategy endorsed by its facile and discerning ability can be envisioned to be of great value for small functional peptide separation and oriented enrichment in biomedicines. KEYWORDS: defective UiO-66, amino acid imprinting, ACE-inhibitory peptides, recognized adsorption, separation, oriented enrichment

1. INTRODUCTION

having similar 3-dimensional shape or size coexist in the same system.14,15 Therefore, designing porous materials to realize highly specific adsorption and efficient separation for targeted peptides is a significant, urgent, and hot research topic in this area.11,16,17 Recently, porous crystalline metal organic frameworks (MOFs) have attracted general interest in the fields of adsorption and separation18−20 attributed to their high surface area, uniform structure, and tunable chemical functional-

Hypertension is one of the major cardiovascular and fatal diseases.1,2 For controlling hypertension, many studies have focused on searching for deriving some small bioactive peptides from edible protein fragments as antihypertensive agents due to their minimal side effects, absorbability,3,4 and antidigestive performances.5−7 Among these, angiotensinconverting-enzyme-inhibitory (ACE-inhibitory) peptides have attracted greater attention in recent years.8−10 In general, purifying and separating target peptides from hydrolysis system require complex procedures and long processing times.11−13 This separation via adsorption or membrane approaches becomes much more difficult when some small peptides © 2019 American Chemical Society

Received: April 29, 2019 Accepted: June 5, 2019 Published: June 5, 2019 23039

DOI: 10.1021/acsami.9b07453 ACS Appl. Mater. Interfaces 2019, 11, 23039−23049

Research Article

ACS Applied Materials & Interfaces ity.21−23 They are reported to have ultrahigh adsorption capacity and specific affinity toward many light molecules, such as CO2, olefin, alkane, and dutrex.24−26 However, few MOFs are reported to adsorb or separate larger molecules because of their limited pore tuning in the mesoporous regime.27,28 Design of defective pores with specific surface properties can create nanoregions in MOFs and endorse them with high affinity or recognition toward nanometer size specific molecules, such as some small peptides. In this regards, many studies have been reported on defect engineering to expand the application scope of MOFs.29−31 For example, Ravon used 2-methyl-toluic acid as a modulator to engineering large defects in MOF-5, which open the door to the alkylation of very large polyaromatic compounds.32 Zeng used a molecular imprinting strategy to create defects in HKUST-1 and attained anionic framework properties and mesoporosity,33 and this represents a new way to rationally design and functionalize MOFs and their derived products according to targeted applications. Zhong synthesized various hierarchicalpore MOFs with structural defects by a self-assembly template strategy. The obtained MOFs showed rich porous properties and potential applications for large molecule adsorption.34 Creating large pores in MOFs can be highly desirable to widen the adsorption region in real application. However, pure pore expansion in MOF structures may hamper their selectivity.34 Thus, constructing selective adsorption sites on MOF defects is extremely valuable and worth studying. Molecular-imprinting technology (MIT) in the cross-linked matrix has been used to enhance selectivity toward target organic molecules who possess identical three-dimentional structure to imprinting molecules.35 Therefore, we tried to fabricate some defects in MOFs via a molecular-imprinting approach, which may realize a relatively high selective adsorption for small ACE-inhibitory peptides with larger sizes. In this work, three hydrophobic amino acids are proposed to fabricate defects in UiO-66s, i.e. Pro, Phe, and Trp (UiO-66-P, UiO-66-F, and UiO-66-W), through a monodentate carboxylate ligand with metal sites for specific adsorption/separation of small ACE-inhibitory peptides. An in-situ AA imprinting approach concurrently enlarged pore size from microporous to mesoporous scale, also imprinting multiple adsorption sites in the defective cavity. Based on a lock-and-key mechanism, the AA imprinted UiO-66s exhibited a significant promotion in the adsorption affinity and ultrahigh selectivity from small ACEinhibitory peptide mixtures.

were synthesized by GL Biochem. Co., Ltd. Acetonitrile and methanol for HPLC analysis was supplied by Thermo Fisher Scientific Co., Ltd. The water was purified via the Smart-S15UV (Hitech Instruments Company, Ltd., Shanghai, China) system, and the resistivity was 18.2 MΩ cm. All the chemicals were analytically pure and used without further purification. 2.2. Synthesis Procedure. Solvothermal synthesis of defect-free UiO-66 is based on the previous report with slight modifications.36 Briefly, a stock solution (A) comprised of zirconium chloride (0.117 g, 0.50 mM) was prepared by dissolving in N,N-dimethylformamide (DMF) at 30 °C. Then, another stock solution (B) comprised of 1,4benzenedicarboxylic acid (H2BDC) (0.083 g, 0.50 mM) was dissolved in DMF. Solution A was dropwise added into solution B under constant magnetic stirring at 30 °C. A 2.0 mL portion of glacial acetic acid was added as a modifier to the mixture of both the solutions, and they were transferred into a Teflon-lined steel autoclave. The autoclave was sealed and placed in an oven, and crystallization was carried out at 120 °C for 24 h with a heating rate of 1 °C min−1. After reactions, the as-synthesized white product (defect-free UiO-66) was separated via centrifugation and washed with DMF, diluted HCl (pH = 1.0) solution, and anhydrous methanol, repeatedly. After several times of separation and washing, these powders were dried in oven at 150 °C for 24 h under vacuum. Imprinted UiO-66s were synthesized via in situ AA imprinting approach using three AAs (Pro, Phe, and Trp). The synthesis procedure of the imprinted UiO-66 samples was the same as that of defect-free UiO-66 expect for adding AAs. The mole ratio of AA and zirconium chloride was set to about 15 during reaction. The imprinted UiO-66s were designated as UiO-66-P, UiO-66-F, and UiO-66-W, corresponding to the molecular imprinter of Pro, Phe, and Trp, respectively. 2.3. Characterization. The crystallinity of all the products was identified by powder X-ray diffraction (PXRD, RIGAKU, Japan) with Cu Kα radiation (λ = 1.5406 Å). The surface morphologies and microstructural analyses of the materials were characterized by field emission scanning electron microscope (FESEM). The surface area and pore size distribution parameters were measured by adsorption− desorption of nitrogen using a Micromeritics ASAP 2460 instrument. Thermogravimetric analysis (TGA) was carried out in a STA 449 F3 Jupiter DSC-TGA of NETZSCH. Temperature-controlled desorption (TPD) was carried out in a Rheometric Scientific STA1500 instrument. The element components of the samples’ surface were analyzed by X-ray photoelectron spectroscopy (XPS). Spherical aberration corrected environmental transmission electron microscopy (Cs-corrected ETEM) was carried out by a TITAN ETEM G2 80-300 (FEI, USA) microscope at 300 kV. 1H NMR characterization was recorded on a Bruker AVANCE III HD600 spectrometer at 600 MHz. 2.4. Peptide Adsorption Experiments. A series of peptides with different dimensions and chemical properties (2-D, 4-D, 6-D, 8-D, 12D, 4-Y, 4-K, DDFF and DDWW) were used to evaluate the adsorption performance of defect-free and imprinted UiO-66s (UiO66-P, UiO-66-F or UiO-66-W). In the adsorption experiment, 10.0 mg of UiO-66 or AA imprinted UiO-66s was added to a 20 mL of 2.0 mmol L−1 peptide aqueous solution. After the adsorption processes for 24 h, adsorbent samples were isolated via centrifugation at 8000 rmp with 2 min. Then, the supernatant solutions were collected and the peptide content was analyzed via RP-HPLC (Agilent 1260) on a Polaris 5 C8-A column (250 × 4.6 mm, Agilent). The column was eluted with 25% acetonitrile in water (v/v) containing 0.1% trifluoroaceric acid (TFA), at a flow rate of 0.5 mL min−1, and monitored at 214 nm by a diode-array detector (DAD). Linear regression of these peptides’ concentration (Y, mmol) and absorbance (X, mAU) values was performed to obtain a standard curve (Figure S12), and adsorption capacity Qe (mmol g−1) of the peptide was calculated according to eq 1:

2. EXPERIMENTAL SECTION 2.1. Materials. Zirconium chloride (ZrCl4, 99.5%) was purchased from J & K Chemical Technology Co., Ltd. (Beijing, China). The hippuryl-L-histidyl-Lleucine (HHL) and ACE-inhibitor from rabbit lung were supplied by SigmaAldrich Chemical Co. (St. Louis, MO). Anhydrous methanol (CH3OH, 99%) and glacial acetic acid (C2H4O2, HAc, 99.5%) were purchased from Guangdong Guanghua Sci-Tech Co., Ltd. (Guangdong, China). Hydrochloric acid (HCl, 48%) was purchased from Kelong Chemical Reagent Co., Ltd. (Chengdu, China). N,N-Dimethylformamide (HCON(CH3)2, DMF, 98%) was obtained from Shanghai Shenbo Chemical Co., Ltd. (Shanghai, China). The proline (C5H9NO2, Pro, 99%), phenylalanine (C9H11NO2, Phe, 99%), tryptophan (C11H12N2O2, Trp, 99%), and 1,4-benzenedicarboxylic acid (H2BDC, 99%) were purchased from Aladdin Industrial Co., Ltd. (Shanghai, China). All of the adsorbed peptides (DD (2-D), DDDD (4-D), DDDDDD (6-D), DDDDDDDD (8-D), DDDDDDDDDDDD (12-D), YYYY (4-Y), KKKK (4-K), DDFF, and DDWW, 99% purity) in this experiment

Qe = 23040

(C0 − Ce) V M

(1) DOI: 10.1021/acsami.9b07453 ACS Appl. Mater. Interfaces 2019, 11, 23039−23049

Research Article

ACS Applied Materials & Interfaces

Figure 1. XRD patterns (a), nitrogen adsorption isotherms (b), DFT pore-size distribution (c), and 1H NMR spectra (d) of UiO-66, UiO-66-F, and UiO-66-W (the samples of 1H NMR spectra were dissolved in 1:6 mixture of HF and DMSO-d6 at 600 MHz, 298 K. *The location of these peaks in samples’ XRD pattern of Figure 1a). × 150 mm, particle size 5 μm; Agilent) to detect hippuric acid (HA) from HHL. The column was eluted with 15% methanol (in water, v/ v) containing 0.1% trifluoroacetic acid (TFA) with a flow rate of 1.0 mL min−1 and measured at 228 nm. ACE-Inhibition activity (I, percent) was calculated by using the following equation):

Where C0 and Ce are the initial and equilibrium concentrations of peptide (mmol L−1), respectively; M is the mass of adsorbent (g), and V is the volume of the solution (L). The equilibrium adsorption capacity qe (mmol g−1), transient fractional adsorption capacity qt (mmol g−1), and selectivity α were calculated according to eqs 2−4), respectively.

I=

(C − Ct )V qt = 0 m

(2)

(C − Ce)V qe = 0 m

(3)

αI /(NI ) =

(5)

Where Ae, Af, and Ab are the relative area of HA peak generated without ACE inhibitors, in the presence of purified peptides and without ACE and purified peptides, respectively. The IC50 value is defined as the concentration of inhibitor (millimolar) required to inhibit 50% of ACE activity, determined by regression analysis of ACE inhibition versus peptide concentration. 2.6. Molecular Computation Parameters. The calculation models of different peptides (2-D−12-D) were set in an octahedral water box using the TIP3P water model for calculating the threedimension size of molecules.38 The conflicts overlaps with atoms in these systems were removed by energy minimization. Then, these systems were heated from 0 to 300 K at 200 ps according to Langevin temperature controls with weak restrain (10 kcal mol−1) on the solute at constant volume periodic boundary conditions. Finally, all calculations were performed without restraints at 300 K and 1.0 bar for 6 ns with AMBER parm 10 force field. To provide microscopic insight into the defect formation of UiO66-P, UiO-66-F, and UiO-66-W, MD simulation were conducted to estimate the energy barriers of three amino acids in UiO-66s with different defects. Three amino acids (Pro, Phe, and Trp) and UiO-66s were described by the Lennard-Jones (LJ) and electrostatic potentials. The LJ potential parameters for UiO-66s were adopted from the universal force field, and the LJ parameters for amino acids were taken from the CHARMM27 force field.39,40 The atomic charges of UiO-

qt , I /Ct , I qt , NI /Ct , NI

Ae − A f Ae − Ab

(4)

Where C0, Ce, and Ct (mmol L−1) are the concentrations of different peptides at the initial time, equilibrium and time t (min), respectively; V is the volume of solution (L), I represents peptides with AA fragments, DDFF and DDWW; while NI represents peptide without AA fragments, 4-K. 2.5. ACE-Inhibitory Activity Tests. The assay of ACE-inhibitory activity in vitro was measured by Cushman’s method37 with slight modifications. First, HHL was dissolved in a 100 mM borate buffer (pH = 8.3) containing 300 mM NaCl. Rabbit lung ACE was dissolved in the same buffer solution at a concentration of 10 mU mL−1. After that, 40 μL of ACE solution and a certain concentration of peptide were diluted with water to 200 μL and then preincubated at 37 °C for 10 min. The mixture was subsequently incubated with 10 μL of HHL solution at 37 °C for the next 30 min. The reaction was then terminated by the addition of 50 μL HCl (1.0 M). Finally, 20 μL solution was injected directly into a Zorbax SB C18 column (4.6 mm 23041

DOI: 10.1021/acsami.9b07453 ACS Appl. Mater. Interfaces 2019, 11, 23039−23049

Research Article

ACS Applied Materials & Interfaces 66-P, UiO-66-F, and UiO-66-W were estimated using the MEPO-Qeq method which has been proven to be fast and accurate to evaluate the electrostatic interaction of MOFs.41,42 To calculate the energy barrier of each amino acid in the imprinted UiO-66s, first, each amino acid molecule was inserted into different UiO-66s with different types of defect, and followed by optimization using the Forcite module in Materials Studio; then, NVT-MD simulation was performed at 298 K for 150 ns using GROMACS 5.0.6 software.43 During MD simulation, the UiO-66 structures were assumed to be rigid during this simulation except both the hydrogen and oxygen atoms in the defects, but all of atoms of amino acids were flexible. Finally, the binding energies between amino acid molecules and the imprinted UiO-66s were estimated for the final structures after MD simulation.

In the case of amino acid imprinted UiO-66, Phe and Trp with rigid benzene group facilitated competitive coordination with Zr sites in UiO-6627,46 and thus generated more defective mesopores in UiO-66. Conversely, Pro with a polar pyrrole group may have no competitive merit in coordination compared to the ligand of poly(carboxylic acid) (H2BDC); hence, only Phe and Trp were selected for imprinted UiO-66 in onward experiments. To further insight the competitive coordination relationship between hydrophobic AAs and H2BDC ligand toward Zr(IV) sites, different mixed modes of three chosen AAs (Pro, Phe, and Trp) and H2BDC with ZrCl4 solution were designed, and their adsorption wavelengths were detected by using UV−vis spectrophotometry (Figure S1). ZrCl4 shows a sharp adsorption peak at 269 nm (black line), which was not altered by the addition of H2BDC. However, upon the addition of excessive AAs into ZrCl4 and mixture of H2BDC/ZrCl4 solution, all the adsorption peaks of Zr(IV) were observed great variation. For example, the adsorption peaks of Zr(IV) without H2BDC all drifted from the original position after AAs addition (red line), indicating that monocarboxyl acid may have coordinated with Zr(IV) to form [Zr−O] clusters.46,49 However, this scenario changed when AAs was added to H2BDC/ZrCl4 solution (purple line). The addition of Pro did not affect Zr(IV) peak (around 269 nm), while those Phe or Trp did change the peak position of Zr(IV) toward 271 or 263 nm, respectively. These results concluded that Phe and Trp have the competitive advantages in coordination with Zr(IV) in the presence of H2BDC ligand, while Pro may not superior in bonding for Zr(IV) compared to H2BDC and hence difficult to imprint defects in UiO-66. Based on these grounds, only Phe and Trp were chosen to imprint UiO-66 in the next sections. To further confirm the role of AAs for imprinted UiO-66s, the defect-free UiO-66 and imprinted UiO-66s were analyzed by 1H NMR spectroscopy (Figure 1d). Two highest peaks were attributed to solvent (DMSO-d6) and H2BDC ligand, respectively. In addition, some small peaks appeared in yellowcolored zoom of Figure 1d, which was magnified on the right. Compared with standard 1H NMR spectra of Phe and Trp, these spectra (Figure S2a, b) were confirmed to Phe and Trp and thus stamping on the presence of Phe and Trp in UiO-66F/W frameworks. However, no Pro was detected from its 1H NMR spectrum in Figure S2c. Combined with UV−vis characterization, it can be concluded that Phe and Trp were successfully imprinted into UiO-66 which caused some missing-linker-type defects in UiO-66 and developed some larger mesopores.27,36,47 This is consistent with their corresponding N2 isotherms and pore size distribution (Figure 1b, c). Furthermore, thermal stability of UiO-66-F/W was measured via thermogravimetric analysis (TGA) (Figure S3). Decomposition inflection points of UiO-66-F and UiO-66-W in TGA curves were located at 544 and 548 °C, very similar to that of defect-free UiO-66. Thus, a portion of AA coordination and linker missing in UiO-66 did not influence their thermal stability, favoring their prospects for industrial applications. 3.2. Morphological Characterization. To determine the existence of defective pores in imprinted UiO-66s and their three-dimentional morphology and distribution, transmission electron microscopy (TEM) measurements of dark field (DF) images and bright field (BF) images were performed. For comparison, we first measured HADFI-TEM images of defectfree UiO-66 and imprinted UiO-66-F/W within small

3. RESULTS AND DISCUSSION 3.1. Structure, Surface Area, and Porosity Analyses. Powder X-ray diffraction (PXRD) patterns depicting the crystal structures of UiO-66, UiO-66-P, UiO-66-F, and UiO66-W are shown in Figure 1a. As seen, three imprinted UiO66s exhibited very similar diffraction peaks to pristine UiO-66, reflecting a similar crystal structure and good crystallinity after the incursion of hydrophobic AAs. The discrepancies are in the intensity and the full width at half-maximum of characteristic peaks in PXRD patterns. Three diffraction peaks of imprinted UiO-66s at 2θ = 7.4, 8.5, 17.0, 25.8, and 30.8° became slightly weaker and broader than that of pristine UiO-66. The data in Table S1 showcased the crystalline size of UiO-66s calculated from Scherrer equation for the five main diffraction peaks.44,45 The decrease in the crystalline sizes of imprinted UiO-66s suggested that AAs can disturb the crystal growth of UiO-66 and thus decrease crystalline sizes of MOFs being imprinted with AAs.46,47 N2 adsorption−desorption approach and pore size distributions of imprinted UiO-66 samples are shown in Figure 1b, c. Some changes displayed in their pore structure caused from the incursion of AAs in regular MOF frameworks. As shown, different types of hydrophobic AAs had different effects on porous structure of UiO-66. Pristine UiO-66 exhibited a type I isotherm characteristic of a microporous material, and almost no N2 uptake at high relative pressure. For UiO-66-P, its adsorption profile was very similar to the pristine one but with low N2 uptake. That means no many defects were created in UiO-66 with proline. However, UiO-66-F (benzene ring modified) and UiO-66-W (indolyl modified) exhibited a different N2 isotherm shape from pristine UiO-66 and UiO66-P, showing a slight increase of N2 uptakes at medium relative pressure. It indicated a hierarchical porosity including a portion of mesopore structure.47,48 Figure 1c showed the density functional theory (DFT) calculated pore size distribution of these two UiO-66-F/W. It suggested that they possessed the original micropore as defectfree UiO-66 in the range of 9 and 11 Å.34 In addition, some larger pores in the range of 18−31 Å appeared in imprinted UiO-66s, which had contributed appropriate 50% of the total pore volume. Thus, it can be regarded that Phe and Trp have constructed amount of large micropores and mesopores in UiO-66, which will affect their adsorption performance. Moreover, porous textural parameters of pristine UiO-66 and imprinted UiO-66s are summarized in Table S2, and it suggested that implantation of hydrophobic AAs lead to loss of a portion of surface area. However, the ratio of mesopores in the total pore volume of UiO-66-F and UiO-66-W were increased to varying degrees. 23042

DOI: 10.1021/acsami.9b07453 ACS Appl. Mater. Interfaces 2019, 11, 23039−23049

Research Article

ACS Applied Materials & Interfaces

Figure 2. HAADF-STEM images of UiO-66 (a), UiO-66-F (b), and UiO-66-W(c) and TEM-BF images of UiO-66 (d), UiO-66-F (e), and UiO66-W (f).

Figure 3. XPS survey spectra of imprinted UiO-66s: (a) N 1s core energy levels, (b) O 1s core energy levels, and (c and d) TPD spectra of (c) CO2 and (d) NH3 on defect-free and imprinted UiO-66s.

the high-magnified HAADF-STEM. These pinholes were oval shaped with a diameter of roughly 2−4 nm, which is consistent with N2 adsorption analysis (Figure 1c). These exposed oval holes can be considered as AA imprinted defects in UiO-66s. The corresponding crystal lattices of defect-free UiO-66, UiO-66-F, and UiO-66-W were also corroborated by TEM-BF images (Figure 2d−f). As we know, it is hard to capture the

magnification as shown in Figure 2a−c. The defect-free UiO66 exhibited the octahedral morphology with a uniform particle size of ∼150 nm. After in situ adding Phe- and Trp-, the morphology of imprinted UiO-66s changed from well octahedra to intergrown cubes with smaller particle sizes of 50−100 nm. Moreover, many defective pinholes (yellow squares) were densely arranged throughout each particle in 23043

DOI: 10.1021/acsami.9b07453 ACS Appl. Mater. Interfaces 2019, 11, 23039−23049

Research Article

ACS Applied Materials & Interfaces

Figure 4. Adsorption capacity of small peptides, selectivity, and ACE-inhibitory activity on defect-free UiO-66, UiO-66-F, and UiO-66-W: (a) small peptides with different molecular dimensions, (b) small ACE-inhibitory peptides with MIT adsorption, (c) selective adsorption of small ACEinhibitory peptides, and (d) ACE-inhibitory activity of different mixed peptides solution in UiO-66-F/W.

at. % compared to defect-free UiO-66. It was due to the substitution of AA for a portion of BDC ligands. Moreover, O 1s XPS spectra (Figure 3b) were deconvoluted into three peaks, which were centered at 530.7/530.5 (Zr−O), 531.8 (CO), and 532.9 eV (O−H) respectively in both pristine and imprinted UiO-66s (Table S4). By comparison, imprinted UiO-66s exhibited 2.9−3.5 times higher intensity of O−H peak than that of defect-free UiO-66, which suggested that monodentate coordination happened between carboxyl groups of AAs and Zr(IV) of Zr−O clusters, leading to more freecoordinated −OH groups in imprinted UiO-66s.52,53 The monodentate coordination of Zr−O cluster can also be reflected from the decrease in Zr−O binding energy from 530.7 to 530.5 eV as compared to bidentate coordination between COOH and Zr(IV).54,55 The intensity of Zr−O peak in imprinted UiO-66s decreased by 7−9% of defect-free UiO66 which could be due to the missing-linker-type defect sites in UiO-66 leading to the loss of segmental Zr−O clusters.56 To insight the surface property of pristine and imprinted UiO-66s, gaseous CO2 and basic NH3 temperature-programmed desorption (TPD) measurement were performed at the heating rate of 7 °C min−1. Figure 3c, d show that all UiO66s exhibited two separated peaks in CO2- and NH3-TPD curves. It directly reflects the surface energy heterogeneity for CO2 and NH3 adsorption. The two peaks in CO2-TPD curves in Figure 3c indicated a weak basic and a strong basic site respectively on UiO-66s.57,58 Both of the peaks were shifted to higher temperatures upon the anchoring of Phe and Trp into UiO-66, indicating an increase in surface basicity. On the contrary, NH3-TPD curves in Figure 3d shows that the two successive peaks were shifted to lower temperatures, suggesting a decrease in surface acidity.59,60 These results echoed each well with other. In light of 1H NMR and XPS analysis, exposure of free-coordinated NH2- groups intensified surface alkalinity of UiO-66-W and UiO-66-F, while the former being equipped with more basic groups endowed higher surface alkalinity than the latter.

crystal lattice fringes of MOFs via TEM due to their low thermal stability.50 Our synthesized UiO-66s would decompose to ZrO2 nanoparticles under strong bombardment by electron beams. Defective UiO-66 will be reflected in generated ZrO2 crystalline. Based on this concept, decomposed ZrO2 lattice was proposed to detect pore defects in UiO-66s. Figure 2d−f shows crystal lattices of ZrO2 in both defect-free UiO-66 and imprinted UiO-66-F and UiO-66-W. Under the persistence energy irradiation, a large amount of ZrO2 nanoparticles as clear lattice fringes of 1.53, 1.80, and 3.04 Å size appeared in all of the UiO-66 frameworks which were ascribed to a stack of (121), (112), and (011) planes of ZrO2 lattice.51 By comparison, UiO-66-F and UiO-66-W revealed the atom deficiency, lattice distortion, and irregular arrangement of ZrO2 nanoparticles in the imprinted UiO-66s, which verified the existence of defective pores. Thus, DF and BF images demonstrated and elaborated the three-dimensional shapes and size of the existing defective pores. 3.3. Surface Composition Analysis. To better understand the coordination mechanism of hydrophobic AAs with Zr(IV) in the Zr−O clusters, the defect-free and imprinted UiO-66s were characterized by XPS. The XPS survey spectra of UiO-66-F and UiO-66-W in Figure S4 verified the presence of N element on the surface of imprinted UiO-66, whose percentages were about 3.3 and 4.7 at. %, respectively (Table S3). The N 1s XPS spectra of imprinted UiO-66s shown in Figure 3a suggest N 1s peaks in UiO-66-F and UiO-66-W can be assigned to N atom from coordinated-free −NH2 and −NH groups in Phe and Trp, respectively. Furthermore, these Phe and Trp were confirmed to be embedded into UiO-66s through successful coordination of Phe or Trp with Zr(IV), which is accordance with 1H NMR results in Figure 1d. In addition, this indicated that the anchoring Phe and Trp into the imprinted UiO-66 were coordinated with zirconium clusters through COO− groups in AAs instead of their NH2− groups. Moreover, the content of carbon element in imprinted UiO-66s was slightly decreased from 60.5 to ∼56.0 23044

DOI: 10.1021/acsami.9b07453 ACS Appl. Mater. Interfaces 2019, 11, 23039−23049

Research Article

ACS Applied Materials & Interfaces 3.4. Adsorption and Selective Separation Performance for ACE-Inhibitory Peptides. Adsorption performance of polypeptides with same AA (aspartic acid, D) but different numbers from 2 to 12 peptides (i.e., dipeptide (2-D), tetrapeptide (4-D), hexapeptide (6-D), octapeptide (8-D), and dodecapeptide (12-D)) were carried out on defect-free and imprinted UiO-66s at 298 K and shown in Figure 4a. As shown, all of UiO-66s displayed a ladder-like decrease in adsorption capacity with the increase of peptide chain length. Defect-free UiO-66 showed a very low uptake for all of these peptides, and its peptide uptake had plunged occurred to 4-D and 6-D peptides. On the contrary, UiO-66-F/W showed much higher uptakes for 2-D, 4-D and 6-D peptides, while dropped between 8-D and 12-D peptides. The enhanced uptake by imprinted UiO-66s could be attributed to the larger pore volume, which can accommodate more peptides as compared to defect-free UiO-66 as suggested by pore structure data (Table S2). Apart from this, the higher uptake can also be accredited to the three-dimensional shape of defective pores in imprinted UiO-66s. According to molecular dynamics (MD) simulations (Figure S5), the length of peptides increases gradually, though not proportionally due to their curling to round configuration. The 4-D peptide with three-dimensional size of 14.0 Å × 7.7 Å × 5.2 Å cannot be adsorbed by the micropores of defect-free UiO-66 and hence is excluded from adsorption leading to lower uptake. Upon AA in situ imprinting, pore sizes of UiO-66s are significantly expanded and thus can adsorb 8-D with three-dimensional geometry of 16.8 Å × 14.7 Å × 7.4 Å. Furthermore, the shape of 8-D peptide has transited from rods to ellipsoid, which indirectly suggested that the pore shape in imprinted UiO-66s can probably be an ellipsoid. Their pore size, judged from their limitation adsorption for 12-D, may be identified in size of 18.7 Å × 16.6 Å × 10.5 Å for UiO-66-F/W. From these data, it can be imaged that these imprinted defects are more like a nearspherical shape, whose width was expanded dramatically from ∼8 to ∼15 Å. Effect of surface property of imprinted materials on selective adsorption behavior was investigated using three tetra-peptides (acidic DDDD (4-D), aromatic YYYY (4-Y) and basic KKKK (4-K)) possessing unique functional groups. Their adsorption capacities on defect-free UiO-66, UiO-66-F, and UiO-66-W are shown in Figure S6. As shown, the defect-free UiO-66 exhibited very low adsorption capacities (