Eco-Friendly Encapsulation of Metal Clusters in Porous Organic Cages for Engineerable Microenvironment and Enhanced Catalysis

Abstract

Open AccessCCS ChemistryRESEARCH ARTICLES5 Feb 2024Eco-Friendly Encapsulation of Metal Clusters in Porous Organic Cages for Engineerable Microenvironment and Enhanced Catalysis Siyu Ren†, Liangxiao Tan†, Jun-Hao Zhou, Jian Sun, Peng Zhang, Xingzhong Cao, Yun-Hong Zhang and Jian-Ke Sun Siyu Ren† MOE Key Laboratory of Cluster Science, Beijing Key Laboratory of Photoelectronic/Electrophotonic Conversion Materials, School of Chemistry and Chemical Engineering, Beijing Institute of Technology, Beijing 102488 , Liangxiao Tan† MOE Key Laboratory of Cluster Science, Beijing Key Laboratory of Photoelectronic/Electrophotonic Conversion Materials, School of Chemistry and Chemical Engineering, Beijing Institute of Technology, Beijing 102488 , Jun-Hao Zhou MOE Key Laboratory of Cluster Science, Beijing Key Laboratory of Photoelectronic/Electrophotonic Conversion Materials, School of Chemistry and Chemical Engineering, Beijing Institute of Technology, Beijing 102488 , Jian Sun MOE Key Laboratory of Cluster Science, Beijing Key Laboratory of Photoelectronic/Electrophotonic Conversion Materials, School of Chemistry and Chemical Engineering, Beijing Institute of Technology, Beijing 102488 , Peng Zhang Institute of High Energy Physics, Chinese Academy of Sciences, Beijing 100049 , Xingzhong Cao Institute of High Energy Physics, Chinese Academy of Sciences, Beijing 100049 , Yun-Hong Zhang MOE Key Laboratory of Cluster Science, Beijing Key Laboratory of Photoelectronic/Electrophotonic Conversion Materials, School of Chemistry and Chemical Engineering, Beijing Institute of Technology, Beijing 102488 and Jian-Ke Sun *Corresponding author: E-mail Address: [email protected] MOE Key Laboratory of Cluster Science, Beijing Key Laboratory of Photoelectronic/Electrophotonic Conversion Materials, School of Chemistry and Chemical Engineering, Beijing Institute of Technology, Beijing 102488 https://doi.org/10.31635/ccschem.024.202303577 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Developing artificial catalysts that mimic the functionality of enzymes and adapt to the surrounding microenvironment to achieve specific activity and selectivity is a fascinating research area yet remains a great challenge. In this work, we present a meticulously designed strategy for the successful encapsulation of ultrasmall metal clusters (MCs) within an amine-type porous organic cage (POC) through electrostatic complexation, phase transfer, and alcohol reduction processes. The amine cage showcases an intriguing and customizable feature that allows for the regulation of the surrounding microenvironment of the confined MCs through a feasible postmodification approach. This functionalization of cage skeleton further facilitates precise adjustment to the surface electronic state of Pd cluster, thereby influencing the adsorption behavior of substrate. Consequently, this controlled regulation leads to modified activity and chemoselectivity in the catalytic hydrogenation of halogenated nitrobenzene. Importantly, the investigation of the correlation between the surrounding microenvironment, substrate adsorption, and catalytic performance in the POC-immobilized MCs system has not been previously reported. We anticipate that our research will provide valuable insights in this field. Download figure Download PowerPoint Introduction Enzymes are renowned for their ability to catalyze reactions with exceptional efficiency and selectivity under mild conditions. This is primarily due to the confined reaction condition and favorable microenvironment such as the hydrophobic pocket surrounding the active site.1,2 The development of artificial analogs that can mimic enzyme catalysis has attracted increasing attention. Among various reported catalysts, supported metal clusters (MCs) represent an important class of catalysts that are extensively used in industry.3–6 In general, the catalytic activity of MCs can be effectively modulated by tailoring their physicochemical properties, such as size, shape, and interactions with supports.7–10 Recently, there has been a growing interest in modulating the surrounding microenvironment of MCs sites that facilitates the recognition, enrichment, preorganization, and activation of substrates, ultimately leading to enhanced activity.11–19 Despite great progress that has been achieved thus far, the precise construction of pockets featuring customized microenvironments around active sites for molecular understanding of the catalytic mechanism remains highly desired yet a great challenge. Porous organic cages (POCs) are a novel type of crystalline porous material composed of discrete molecules with intrinsic, guest-accessible cavities and excellent solution processability.20–32 The enzyme-like binding pockets make them ideal platforms for the immobilization of MCs in heterogeneous catalysis.33–40 Thus far, most studies have focused on the creation of MCs sites and the investigation of catalytic performance by tailoring the pore matrix and components of POCs.41–47 However, the effect of microenvironment modulation on POC-immobilized MCs catalysis and the underlying structure-performance relationship remain largely elusive. An alternative approach is to develop a simple and eco-friendly method for encapsulating MCs inside cage cavities, followed by chemical modification of the skeleton to interact with the MCs centers and modify their electronic structure without affecting the cluster size. Given the unique structural features of POCs, intriguing biomimetic behaviors are highly expected via microenvironment engineering in POC-based heterogeneous catalytic systems. In this work, we describe a novel approach for encapsulating ultrasmall MCs (e.g., Pd, Au, Pt) into amine-based POCs and investigate the relationship between catalyst structure and activity through microenvironment engineering of MCs. Initially, the anionic metal precursor (AMP) was encapsulated into a quaternary ammonium cage (QA-Cage+) through electrostatic complexation in aqueous solution (AMP⊂QA-Cage+). This was followed by neutralization of QA-Cage+ with base (resulting in amine cage, denoted as A-Cage), leading to the formation of AMP⊂A-Cage. Subsequently, an oil phase was added into the system, which facilitated the phase transfer of AMP⊂A-Cage, and excluded peripheral AMP outside the cage cavity. A mild and environmentally-friendly alcohol-reduction route was then employed to generate the MCs⊂A-Cage (Scheme 1). Interestingly, the size and spatial location of MCs can be precisely controlled by selecting differently sized alcohol reductants. In particular, a small-sized alcohol, such as methanol, can diffuse inside the cage and achieve overwhelming encapsulation through an in-situ reduction procedure. Furthermore, the Pd cluster surrounding the microenvironment can be readily regulated by postmodifying the amine group of cage skeleton. This enables the regulation of inner Pd surface electronic state, leading to the alteration of substrate adsorption as well as consequential reaction activity and chemoselectivity in the model catalysis of hydrogenation of halogenated nitrobenzene (NB). To our knowledge, this is the first study to explore the influence of surrounding microenvironment regulation on substrate adsorption and catalytic performance for POC-immobilized MCs system. Our findings may open new avenues for the design and application of cage-based biomimetic heterogeneous catalysts. Scheme 1 | Schematic illustration of the three-step strategy for encapsulating ultrasmall MCs involving (1) the electrostatic complexation induced AMP inclusion into the QA-Cage+ cavity, (2) the base treatment/phase transfer to exclude unencapsulated AMP, and (3) an alcohol in-situ reduction procedure. Download figure Download PowerPoint Experimental Methods Synthesis of MCs⊂A-Cage-X (X stands for different alcohols) The following three-step approach is elaborately designed for the efficient encapsulation of ultrasmall MCs into amine-type A-Cage cavities. First, 15.8 mg (10 μmol) of QA-Cage+ was readily dissolved in 5 mL of deionized water. Then, an aqueous solution (10 mM) of AMP (K2PdCl6 was used for Pd cluster, NaAuBr4 for Au cluster and K2PtCl6 for Pt cluster) was added in small portions (10 μL for each time, 100 μL in total) and the mixture was vigorously vortexed for 2 min. The yellow solution quickly faded to colorless, implying that the AMP was likely to be included in the cationic QA-Cage+ cavity due to strong electrostatic complexation. The addition of AMP was repeated 10 times, during which the colorless solution gradually changed to light yellow (the final colors varied slightly depending on the specific AMP used). After the final addition, the solution was vigorously vortexed for another 10 min and left to age for 30 min. Second, a NaOH solution (1 M, 180 μL, 18 equiv) was added into the above-prepared AMP⊂QA-Cage+ solution. A pale-yellow precipitate formed immediately indicating the successful conversion of QA-Cage+ to neutral A-Cage. Then, 5 mL of dichloromethane (DCM) was added into the dispersion and the mixture was gently shaken for 30 s, accompanied with the disappearance of the precipitate and the color of DCM changed to light yellow. The upper water phase containing any unreacted components was then discarded, leaving behind the AMP⊂A-Cage in DCM solution. Third, a certain amount of alcohol (1 mL for methanol/isopropanol/cyclohexanol and 1 g for 9-anthracenemethanol, respectively) was added into the AMP⊂A-Cage DCM solution and the mixture was stirred at room temperature (298 K) for 8 h to realize the in-situ reduction process. Finally, the DCM and excess alcohol were removed by rotary evaporation operated at room temperature to obtain MCs⊂A-Cage-X (X stands for different alcohols, i.e., Me refers to methanol, Isop refers to isopropanol, Ch refers to cyclohexanol and An refers to 9-anthracenemethanol). Since 9-anthracenemethanol is a high boiling point alcohol, it was removed by acetonitrile wash (1 mL × 3 times) after rotary evaporation. The obtained samples showed similar yields of ∼11.0 mg (∼89%) for different alcohol reduction. Catalytic hydrogenation of p-CNB The Pd⊂A-Cage-Me catalyst is selected as an example, while the catalytic procedures of other catalysts are similar. Specifically, 5.0 mg of Pd⊂A-Cage-Me (containing 1.55 μmol Pd) was dissolved in a 3 mL of DCM/ethanol solvent (v/v = 1:2), then 7.9 mg (0.05 mmol) of p-chloronitrobenzene (p-CNB) and 7.7 mg (0.05 mmol) of biphenyl (as internal standard) was added to the solution, which was further stirred for 10 min to obtain a homogeneous solution. Next, 7.6 mg (0.20 mmol) of NaBH4 was added all at once to the solution, and the reaction was performed at room temperature with stirring (600 rpm) for 10 min. Once the reaction was completed, 1 mL of deionized water was added to the mixture to quench the excess NaBH4. Subsequently, 2 mL of ethyl acetate was introduced to thoroughly extract the substrate and product under vigorous vortexing. The mixture was then centrifuged and the organic phase was collected for gas chromatography analysis. Further details for synthetic procedures, catalytic experiments, special characterizations, and computational simulations may be found in the Supporting Information. Results and Discussion Encapsulation of ultrasmall MCs In order to encapsulate ultrasmall MCs, a well-designed three-step approach was implemented, starting with electrostatic complexation-induced inclusion of AMP into the QA-Cage+, followed by phase transfer and direct alcohol reduction. The QA-Cage+ host was derived from a classical CC3 cage followed by a reduction and quaternarization procedure ( Supporting Information Figures S43–S45).48 Here, the encapsulation for Pd cluster is selected as a representative to elucidate the whole process, while the experimental procedure and characterization for other MCs (e.g. Au and Pt) can be found in the Supporting Information. The electrostatic complexation procedure was carried out by continuously adding AMP (PdCl62−) droplets into an aqueous solution containing QA-Cage+ under vigorous vortexing (denoted as PdCl62−⊂QA-Cage+). Such a process was monitored by a surface charge density decrease from +38.3 to +25.0 mV, as determined by Zeta potential analysis (Figure 1a). Meanwhile, a significant reduction in the characteristic absorption of PdCl62− ions in PdCl62−⊂QA-Cage+ was observed in UV–vis spectroscopy, further supporting the successful inclusion (Figure 1b). To exclude the peripheral AMP outside the QA-Cage+, a phase transfer procedure was introduced. The water-soluble PdCl62−⊂QA-Cage+ underwent a transition to its neutral amine state and immediately precipitated upon treatment with a base (NaOH, 1 M). This resulted in the formation of PdCl62−⊂A-Cage, which was then transferred into the oil phase by adding DCM to yield a light-yellow solution, while the unencapsulated PdCl62− ions remained in the aqueous solution (Figure 1b and Supporting Information Figure S1). The existence of Pd (IV) in PdCl62−⊂A-Cage was confirmed by X-ray photoelectron spectroscopy (XPS) (Figure 1c). Proton nuclear magnetic resonance (1H-NMR) spectra exhibited slight upfield shifts of 0.01–0.02 ppm for certain peaks, accompanied by broadening, indicating the interaction between the encapsulated PdCl62− ions and A-Cage ( Supporting Information Figure S2).42,49 In contrast, direct mixing of pure A-Cage or A-Cage DCM solution with K2PdCl6 aqueous solution showed negligible absorption of PdCl62− ions, highlighting the unparalleled encapsulation effect achieved through the phase transfer strategy ( Supporting Information Figure S3). Figure 1 | Characterizations for the successful encapsulation of Pd cluster into the cage cavity. (a) Zeta potential curves of QA-Cage+ (blue) and PdCl62−⊂QA-Cage+ (green). (b) UV–vis spectra of PdCl62− (red), QA-Cage+ (blue), PdCl62−⊂QA-Cage+ (green) in aqueous solution, as well as the aqueous solution of PdCl62−⊂QA-Cage+ after base treatment and phase transfer (gray). (c) XPS spectrum of Pd 3d in PdCl62−⊂A-Cage, showing the presence of Pd (IV) at 339.2 eV for 3d5/2 and 344.5 eV for 3d3/2, respectively. (d) XPS spectrum of Pd 3d in Pd⊂A-Cage-Me indicating the presence of metallic Pd at 335.8 eV for 3d5/2 and 341.1 eV for 3d3/2, respectively. (e) HAADF-STEM image and (f) corresponding statistical size distribution histogram of encapsulated Pd clusters in Pd⊂A-Cage-Me (data calculated from 100 counts), scale bar: 10 nm. (g) 1H-NMR and 2D DOSY spectra of A-Cage (orange) and Pd⊂A-Cage-Me (blue) exhibiting similar diffusion coefficient values to demonstrate the preserved size and shape of A-Cage after Pd cluster encapsulation. The chemical structure of A-Cage is inserted, with assigned NMR signals for reference. (h) PAL spectra of A-Cage (orange) and Pd⊂A-Cage-Me with three times Pd cluster content (blue) revealing a distinguished positronium lifetime distribution. Download figure Download PowerPoint Next, we employed an in-situ alcohol reduction method to produce ultrasmall Pd clusters. The alcohol reduction has been recognized to be a mild and eco-friendly method, leveraging the reducing ability of hydroxyl group with active α hydrogens.50–52 At first, methanol as the smallest alcohol (size: 0.44 × 0.40 × 0.40 nm) was utilized which could readily diffuse into the cage cavity (window diameter: 0.70 nm)53 and directly reduce PdCl62− (Eθ HCHO/CH3OH = +0.24 V vs standard hydrogen electrode (SHE); Eθ PdCl62−/Pd = +1.88 V vs SHE; ΔG = −746.70 KJ mol−1, detailed Gibbs free energy calculation is presented in Supporting Information Section S3, Figures S4 and S5, and Movie S1). The resulting material is denoted as Pd⊂A-Cage-Me. XPS analysis revealed the characteristic peaks of metallic Pd at 335.8 eV for 3d5/2 and 341.1 eV for 3d3/2, respectively, providing conclusive evidence for the complete reduction of PdCl62− precursor (Figure 1d). The size of Pd clusters is estimated to be 0.64 ± 0.12 nm, as observed by the high-angle annular dark-field scanning transmission electron microscopy (HAADF-STEM), which matches well with the cavity size of A-Cage, indicating possible encapsulation (Figure 1e,f). To rule out the possibility of Pd cluster stabilized by multiple cages, two-dimensional diffusion-ordered spectroscopy (2D DOSY) NMR experiments were performed on both A-Cage and Pd⊂A-Cage-Me under the same conditions. The results display very similar diffusion coefficient values of 5.16 × 10−10 m2 s−1 for A-Cage and 5.08 × 10−10 m2 s−1 for Pd⊂A-Cage-Me, demonstrating the almost unchanged cage size and shape after encapsulation (Figure 1g and Supporting Information Figures S6 and S7).43 Furthermore, a positron annihilation technique was employed to evaluate the location of Pd cluster relative to the A-Cage host. Since the positrons are highly sensitive to pore or void on the atomic scale, any subtle volume change of space caused by molecule inclusion would be probed, thus providing insight into their accurate space occupation/location.54–56 To emphasize the space discrepancy caused by encapsulation, a higher content of Pd cluster (about three times) was introduced (to occupy more cage cavities) specifically for this measurement, which hardly affect the cluster size and surface electronic state ( Supporting Information Figures S8 and S9, the normal content was used for all other characterizations and catalysis experiments). The positron annihilation lifetime (PAL) spectra were analyzed by fitting them to three exponentially decayed components with respective lifetime (τ1, τ2, and τ3) and relative intensities (I1, I2, and I3) using a LT 9.0 program (Figure 1h and Supporting Information Table S1).57 The shorter lifetimes of τ1 and τ2 correspond to the annihilations of para-positronium (singlet state, with a vacuum lifetime of ∼125 ps) and free positron (lifetime of ∼400 ps), respectively.58,59 On the other hand, the long-lived lifetime of τ3 indicates the "pick-off" annihilation of ortho-positronium (triplet state, lifetime of >142 ns), which is likely trapped within the cage cavities.60 By correlating the spatial volume with PAL data through the Tao-Eldrup model, a fitted diameter of 0.51 nm was estimated (from τ3) for the cage cavity, which is consistent with the reported values.61,62 As clearly displayed in Supporting Information Table S1, the lifetimes of τ1, τ2, and τ3 are similar for both A-Cage and Pd⊂A-Cage-Me, while the relative intensities of I1 and I3 show notable difference. Specifically, I3 significantly decreases from 16.01% to 1.77%, suggesting a nonnegligible reduction in free space ratio of the A-Cage cavity after Pd cluster encapsulation. Meanwhile, I1 increases from 43.50% to 55.29%, indicating that the presence of Pd clusters in the cage cavities accelerates the annihilation of positronium, resulting in a larger proportion of the shorter lifetime (τ1).63 Therefore, the unchanged τ indicates that the main structure and the type of pores or cavities (probed by positrons) remain intact after the Pd cluster inclusion. However, the increase of I1 and decrease of I3 provide strong evidence for the embedment of Pd clusters within the A-Cage cavities. The above evidence strongly suggests that Pd clusters are most likely formed and confined within the cage cavities rather than being stabilized by multiple cages. To further investigate the in-situ reduction process and its influence on Pd cluster size and location, a control experiment was conducted using a much larger 9-anthracenemethanol (size: 1.13 × 0.94 × 0.40 nm) as a reductant that can hardly diffuse into the cage cavity ( Supporting Information Figures S10 and S11). As expected, only a negligible amount of the encapsulated PdCl62− ions was reduced, as evidenced by the XPS analysis ( Supporting Information Figure S12). Furthermore, both methanol and 9-anthracenemethanol produced larger Pd nanoparticles with severe aggregation in the absence of the A-Cage host, demonstrating the crucial role of nanoconfinement from the cage cavity ( Supporting Information Figure S13). It is noteworthy that the in-situ reduction step is critical and opens up the possibility of controlling size and location of Pd clusters by adjusting the reducing activity and molecular size of the alcohols. Building upon these results, moderate sized isopropanol (size: 0.67 × 0.59 × 0.49 nm) and cyclohexanol (size: 0.79 × 0.65 × 0.51 nm) reductants were utilized to produce Pd clusters (denoted as Pd⊂A-Cage-Isop and Pd⊂A-Cage-Ch, respectively). The successful diffusion of reductive alcohols into the cage cavities was confirmed by alcohol capture experiments and molecular dynamic simulations ( Supporting Information Figures S14–S17 and Movies S2 and S3). However, the obtained Pd clusters displayed larger average sizes of 1.83 ± 0.86 nm for Pd⊂A-Cage-Isop and 2.28 ± 0.92 nm for Pd⊂A-Cage-Ch with broader size as by HAADF-STEM and XPS spectra ( Supporting Information Figures findings can be that alcohols have a higher to occupy more space in the cage cavity. As a the sites of MCs, through are more readily from the cavities. to the of these clusters to resulting in the formation of larger which are stabilized by multiple cages as from the DOSY NMR results ( Supporting Information Figure this strategy is and by which other ultrasmall MCs such as Au and Pt clusters have been encapsulated with similar cluster sizes ± nm for Au and 0.67 ± 0.12 nm for Pt) using methanol as the reducing ( Supporting Information Figures Microenvironment regulation by postmodification of cage host After the successful encapsulation of ultrasmall Pd we to the surrounding microenvironment of the inner cluster through postmodification of cage skeleton. The amine in A-Cage was to amine and quaternary ammonium through the and quaternarization resulting in the formation of and (Figure and Supporting Information Figure Importantly, these the Pd cluster size ( Supporting Information Figure significantly the surface electronic state of the encapsulated Pd which was confirmed by density adsorption, and XPS analysis. To obtain the atomic of the encapsulated Pd the model with modified cage were and through the As in the electron density distribution electron transfer behaviors from Pd clusters to the cage were with the transferred electron density on the cage as while the electron density on the Pd clusters as Figure on the the electron transfer were calculated to be and respectively, following the order of Pd⊂A-Cage-Me which the on Pd cluster surface (Figure Figure 2 | Microenvironment regulation to the surface electronic state of encapsulated Pd clusters. (a) Schematic illustration to show the regulation of electron transfer from Pd cluster to cage host by postmodifying the cage skeleton. (b) revealing the distinguished electron transfer behaviors from clusters to different modified cage model model Pd are by electron density to be hardly electron model and color represent the and reduction of electron respectively). (c) The corresponding electron transfer calculated from with the order of (orange) Pd⊂A-Cage-Me (blue) (green). (d) spectra of adsorption and (e) XPS spectra of Pd 3d for Pd⊂A-Cage-Me and respectively, clearly demonstrating the Download figure Download PowerPoint To further the different electron transfer behaviors and investigate the surface electronic of Pd cluster, adsorption spectra were collected using spectroscopy The spectra revealed characteristic adsorption peaks of in the of demonstrating the strong binding between Pd clusters and molecules (Figure and Supporting Information Figure with the adsorption of Pd⊂A-Cage-Me a to was observed for while a to was observed for the The higher with the of the and of the which a electron from on the Pd the electron density on Pd clusters the of Pd⊂A-Cage-Me In the shifts of adsorption could be to the effect of on the cage skeleton to the Pd clusters. The of electron the order of amine amine quaternary which a charge from the to Pd for the order of electron transfer observed in the Furthermore, this in the binding energy in the Pd 3d peaks of XPS spectra (Figure with an 3d5/2 from eV to 335.8 eV to eV Therefore, the modification of cage skeleton with amine could effectively the electronic state of Pd clusters. Catalytic hydrogenation of halogenated It can be that the electronic of the encapsulated Pd clusters would significantly influence their catalytic a model reaction involving the hydrogenation of p-CNB to was to investigate the of microenvironment a catalytic complete conversion of p-CNB and high selectivity for within 10 min 1). However, the Pd⊂A-Cage-Me only conversion and selectivity for with only conversion and selectivity for a significant of selectivity is implying its catalytic results that the crucial regulation of the Pd cluster the group reduction group in the hydrogenation of which can both the catalytic activity and Table 1 | of p-CNB Conversion 1 2 Pd⊂A-Cage-Me 3 5 QA-Cage+ 3 mL DCM/ethanol

Topics & Concepts