Incorporating Sulfur Atoms into Palladium Catalysts by Reactive Metal–Support Interaction for Selective Hydrogenation

Open AccessCCS ChemistryRESEARCH ARTICLE5 Sep 2022Incorporating Sulfur Atoms into Palladium Catalysts by Reactive Metal–Support Interaction for Selective Hydrogenation Zhen-Yu Wu†, Hang Nan†, Shan-Cheng Shen, Ming-Xi Chen, Hai-Wei Liang, Chuan-Qi Huang, Tao Yao, Sheng-Qi Chu, Wei-Xue Li and Shu-Hong Yu Zhen-Yu Wu† Division of Nanomaterials and Chemistry, Hefei National Laboratory for Physical Sciences at the Microscale, CAS Center for Excellence in Nanoscience, Institute of Energy, Hefei Comprehensive National Science Center, Department of Chemistry, Institute of Biomimetic Materials and Chemistry, Anhui Engineering Laboratory of Biomimetic Materials, University of Science and Technology of China, Hefei 230026 , Hang Nan† Division of Nanomaterials and Chemistry, Hefei National Laboratory for Physical Sciences at the Microscale, CAS Center for Excellence in Nanoscience, Institute of Energy, Hefei Comprehensive National Science Center, Department of Chemistry, Institute of Biomimetic Materials and Chemistry, Anhui Engineering Laboratory of Biomimetic Materials, University of Science and Technology of China, Hefei 230026 , Shan-Cheng Shen Division of Nanomaterials and Chemistry, Hefei National Laboratory for Physical Sciences at the Microscale, CAS Center for Excellence in Nanoscience, Institute of Energy, Hefei Comprehensive National Science Center, Department of Chemistry, Institute of Biomimetic Materials and Chemistry, Anhui Engineering Laboratory of Biomimetic Materials, University of Science and Technology of China, Hefei 230026 , Ming-Xi Chen Division of Nanomaterials and Chemistry, Hefei National Laboratory for Physical Sciences at the Microscale, CAS Center for Excellence in Nanoscience, Institute of Energy, Hefei Comprehensive National Science Center, Department of Chemistry, Institute of Biomimetic Materials and Chemistry, Anhui Engineering Laboratory of Biomimetic Materials, University of Science and Technology of China, Hefei 230026 , Hai-Wei Liang *Corresponding authors: E-mail Address: email protected E-mail Address: email protected E-mail Address: email protected Division of Nanomaterials and Chemistry, Hefei National Laboratory for Physical Sciences at the Microscale, CAS Center for Excellence in Nanoscience, Institute of Energy, Hefei Comprehensive National Science Center, Department of Chemistry, Institute of Biomimetic Materials and Chemistry, Anhui Engineering Laboratory of Biomimetic Materials, University of Science and Technology of China, Hefei 230026 , Chuan-Qi Huang *Corresponding authors: E-mail Address: email protected E-mail Address: email protected E-mail Address: email protected Hangzhou Institute of Advanced Studies, Zhejiang Normal University, Hangzhou, Zhejiang 311231 , Tao Yao National Synchrotron Radiation Laboratory, University of Science and Technology of China, Hefei 230026 , Sheng-Qi Chu Institute of High Energy Physics, Chinese Academy of Sciences, Beijing 100049 , Wei-Xue Li Department of Chemical Physics, School of Chemistry and Materials Science, iCHeM, CAS Center for Excellence in Nanoscience, Hefei National Laboratory for Physical Sciences at the Microscale, University of Science and Technology of China, Hefei 230026 and Shu-Hong Yu *Corresponding authors: E-mail Address: email protected E-mail Address: email protected E-mail Address: email protected Division of Nanomaterials and Chemistry, Hefei National Laboratory for Physical Sciences at the Microscale, CAS Center for Excellence in Nanoscience, Institute of Energy, Hefei Comprehensive National Science Center, Department of Chemistry, Institute of Biomimetic Materials and Chemistry, Anhui Engineering Laboratory of Biomimetic Materials, University of Science and Technology of China, Hefei 230026 https://doi.org/10.31635/ccschem.021.202101428 SectionsSupplemental MaterialAboutAbstractPDF ToolsAdd to favoritesDownload CitationsTrack Citations ShareFacebookTwitterLinked InEmail Developing highly active and selective catalysts for the hydrogenation of nitroarenes, an environmentally benign process to produce industrially important aniline intermediates, is highly desirable but very challenging. Pd catalysts are generally recognized as active but nonselective catalysts for this important reaction. Here, we report an effective strategy to greatly improve the selectivity of Pd catalysts based on the reactive metal–support interaction, by which the sulfur atoms doped in the carbon supports are extracted out and react with Pd at 500 °C to form Pd4S nanoparticle catalysts. The Pd4S catalysts that are formed display high selectivity of >98% at complete conversion to diverse nitroarenes (with more than 20 examples) under mild conditions. Density functional theory calculations reveal that the incorporation of S atoms into Pd breaks up the Pd ensembles, which results in the preferential adsorption of nitro groups on the Pd4S surfaces and thus selective hydrogenation of substituted nitroarenes to corresponding anilines. Download figure Download PowerPoint Introduction Hydrogenation reactions have attracted continuing interest due to their great significance in the chemical industry and scientific research.1–7 A typical but important example is the hydrogenation of substituted nitroarenes, which produces substituted anilines as key intermediates for the manufacture of pharmaceuticals, agrochemicals, dyes, and functional polymers.8–14 A big challenge for this reaction is the selective reduction of the nitro group when more than one reducible group (e.g., –Cl, –C=C, –C=O, –OH) is present in the nitroarenes, as conventional supported noble metal catalysts (e.g., Pd or Pt) cannot discriminate different functional groups for selective hydrogenation.8,12,13,15–20 Developing new hydrogenation catalysts with high intrinsic chemoselectivity is a straightforward and effective strategy to tackle this tough problem.8,21–27 Some important progress has been achieved recently by employing Au-based, Ag-based, and pyrolyzed Fe(or Co)/nitrogen/carbon catalysts for selective reduction of the nitro group of substituted nitroarenes.8,21,25,28–30 Nevertheless, these catalysts always suffer from low activity and thus often require harsh reaction conditions, such as high temperature (≥120 °C), high H2 pressure (≥5.0 MPa), and long reaction time (≥12 h).8,21,25,28,31 Another effective strategy is improving chemoselectivity of conventional hydrogenation catalysts (mainly Pt-group catalysts), which possess poor catalytic selectivity but high intrinsic activity.13,15,20,31–35 A traditional approach is the modification of active metal species with inorganic or organic modifiers. However, those modifiers not only decrease the activity via coverage of the active sites, but also raise environmental problems.20,34,36 Other ways of improving selectivity of Pt-group catalysts mainly include size control of active components (even at the single-atomic level),11,12,18,31 formation of alloys or intermetallic compounds,17,32,37–40 and regulation by the pore structure of supports.33,35,41 Particularly, construction of intermetallic compounds, featuring atomically ordered and thermodynamically stable structures with defined stoichiometry, can change the adsorption/desorption properties of the relevant reaction species, which makes it possible to catalyze the nitroarene hydrogenation reaction in the desired direction.32,37,40 Furthermore, well-defined atomic arrangements provide an excellent platform to study the catalytic mechanism, enabling better performance optimization and catalyst design.32,37 Inspired by these promising works, we propose that incorporating nonmetallic elements (i.e., sulfur) into active Pd metal lattices to form compounds with well-defined structures is a feasible way to create selective nitroarene hydrogenation catalysts that will work well under mild conditions. Herein, we report a novel nitroarene hydrogenation catalyst, that is, crystalline Pd4S nanoparticles supported on S-doped carbon (Pd4S/SC), which is prepared through the reactive metal–support interaction (RMSI). RMSI, a chemical reaction that generally occurs on the metal–oxide support interface under high temperatures, was recently employed to prepare bimetallic structures.42–44 However, RMSI has never before been explored in the carbon-supported metal catalysts. Here, we demonstrate an example of carbon support-based RMSI, in which the sulfur atoms doped in the carbon supports are extracted out at 500 °C and incorporated into Pd to form crystalline Pd4S nanoparticle catalysts. In contrast to conventional Pd/C catalysts with poor selectivity, the prepared Pd4S catalyst gives high selectivity of >98% at complete conversion for structurally diverse nitroarenes (with more than 20 examples) under mild conditions. Density functional theory (DFT) calculations disclose that the geometric effect induced by the S incorporating into the Pd lattice results in the preferential adsorption of the nitro group on the Pd4S surfaces and thus selective hydrogenation of nitroarenes to corresponding anilines. Experimental Methods Chemicals and materials SiO2 fumed powder and commercial Pd/C catalyst were purchased from Sigma-Aldrich (St. Louis, MO). Commercial Pt/C catalyst was obtained from Alfa Aesar (Ward Hill, MA). Other chemicals were commercially procured from Sinopharm Chemical Reagent Co. Ltd. (Shanghai, China) and used as received without further purification. Synthesis of SC supports and a-PdxS/SC-b catalysts In a typical experiment for synthesizing SC, 2.0 g of 2,2′-bithiophene, 1.0 g of Co(NO3)2·6H2O, and 2.0 g of SiO2 fumed powder (7 nm; Sigma-Aldrich S5130) were first added into 120 mL of tetrahydrofuran and stirred for ca. 2 h. Then, the solvent was removed by rotary evaporation. The obtained dried powder was subsequently carbonized under flowing N2 for 2 h at 800 °C. The carbonized product was then leached in 2.0 M NaOH for 3 days and 0.5 M H2SO4 at 90 °C for 8 h to remove SiO2 templates and metallic species, and finally afforded SC. An inductively coupled plasma-atomic emission spectrometer (ICP-AES) test showed that the residual Co content in SC was very low (0.24 wt %). To synthesize a-PdxS/SC-b catalysts, PdCl2 was first dissolved in HCl aqueous solution to generate a H2PdCl4 solution with a concentration of 2.815 mg mL−1. Then, a certain amount of H2PdCl4 aqueous solution and 60 mg of SC was added into 30 mL deionized water. After stirring for 6 h and sonication for another 0.5 h, the water was removed by rotary evaporation. The obtained powder was treated by thermal reduction at 300–700 °C under flowing 5% H2/Ar for 2 h to afford final a-PdxS/SC-b catalysts, where a and b are the weight ratios of Pd to SC and reduction temperatures, respectively. Catalyst characterizations Transmission electron microscopy (TEM) images were taken with a Hitachi H7700 transmission electron microscope with a charge-coupled device (CCD) imaging system and an accelerating voltage of 100 kV. High-angle annular dark-field scanning TEM (HAADF-STEM) and high-resolution TEM (HRTEM) images were acquired by using a JEM-ARM 200F atomic resolution analytical microscope (JEOL, Tokyo, Japan) operating at an accelerating voltage of 200 kV. X-ray diffraction (XRD) data were collected on a Philips X'Pert PRO SUPER X-ray diffractometer (Almelo, The Netherlands) equipped with graphite monochromatic Cu Kα radiation (λ = 1.54056 Å). The energy-dispersive spectroscopic (EDS) line-scan and EDS elemental mapping was carried out on a Talos F200X transmission electron microscope (FEI, Hillsboro, OR) at an accelerating voltage of 200 kV equipped with an energy dispersive detector. N2 sorption analysis was recorded on an ASAP 2020 accelerated surface area and porosimetry instrument (Micromeritics, Norcross, GA), equipped with automated surface area, at 77 K using Barrett–Emmett–Teller (BET) calculations for the surface area. X-ray photoelectron spectroscopy (XPS) was performed on an X-ray photoelectron spectrometer (ESCALab MKII, Thermo Scientific, Waltham, MA) with an excitation source of Mg Kα radiation (1253.6 eV). The X-ray absorption near-edge spectroscopy (XANES) and extended X-ray absorption fine structure (EXAFS) spectra of Pd K-edge were measured at BL14W1 beam line of Shanghai Synchrotron Radiation Facility (SSRF, Shanghai, China). The XANES tests of S K-edge were carried out at the 4B7A beamline of Beijing Synchrotron Radiation Facility (BSRF, Beijing, China). The XANES tests of S L-edge were performed at the BL11U beamline of National Synchrotron Radiation Laboratory (NSRL, Hefei, China). The S XANES spectra in the figures have been normalized to the background before and after the main features. The gas chromatograph (GC) instrument was equipped with a Restek-5 capillary column (5% diphenyl and 95% dimethylsiloxane, 0.32 mm diameter, 60 m length), and a flame-ionization detector (FID). GC (Shimadzu GC-2014, Nakagyo-ku, Kyoto, Japan) was used to calculate the conversion and selectivity of the catalytic products. Pd K-edge XAFS analysis The obtained EXAFS data were processed according to the standard procedures utilizing the ATHENA module in the IFEFFIT software packages. The EXAFS spectra were obtained by subtracting the postedge background from the overall absorption and then normalizing with respect to the edge-jump step. Subsequently, the χ(k) data were Fourier-transformed (FT) to real (R) space using a hanning windows (dK = 1.0 Å−1) to separate the EXAFS contributions from different coordination shells. Least-squares curve parameter fitting was carried out using the ARTEMIS module of the IFEFFIT software packages to obtain the quantitative structural parameters around central atoms. Catalytic reaction experiments The hydrogenation of 4-chloronitrobenzene was carried out in a stainless steel autoclave equipped with a pressure control system and a magnetic stirrer (1200 rpm). In a typical hydrogenation experiment, 1.0 mmol of 4-chloronitrobenzene and 0.094 mol % Pd of catalysts were mixed with 1 mL of ethyl acetate in a reaction glass vial. The reaction vial was then placed into a 100 mL steel Parr autoclave. The autoclave was purged three times with H2 and then charged with 0.6 MPa of H2 before the reaction mixture was stirred for 1 h at 80 °C. After the reaction was completed, the autoclave was placed into a cool water bath to stop the reaction. Then, the H2 was released slowly, and 60 μL of o-xylene was added to the system as the internal standard. Next, 10 mL of ethyl acetate was added to dilute the reaction solution. The products were analyzed by GC. In the recyclability test, all reactions were carried out for 0.75 h according to the procedures above. After each run, the used catalyst was separated by vacuum filtration and washed with deionized water and ethanol. Then the obtained sample was dried at 80 °C overnight before being tested in the next run. For the gram-scale hydrogenation of the 4-chloronitrobenzene test, the reaction was conducted by using 25 mmol of 4-chloronitrobenzene, 0.0188 mol % Pd of catalysts, and 15 mL of ethyl acetate as solvent under 1.0 MPa of H2 at 100 °C. The hydrogenation experiment procedure remained the same. After the reaction was performed for 1 h, the autoclave was placed into a cool water bath to stop the reaction. Then the remaining hydrogen gas was discharged, and 0.1 mL liquid mixture was extracted to analyze the product by GC. Afterward, the autoclave was charged with pure H2 again to perform the next run. In the experiment of hydrogenation of substituted nitroarenes, the substrates and products were determined by GC or GC-mass spectrometry (GC-MS). The conversion and selectivity were calculated by GC analysis, which were determined by GC peak areas. DFT calculation method DFT calculations were performed using the Vienna ab initio Simulation Package (VASP) based on the projected augmented wave (PAW) method.45–48 The nonlocal correlation functional of van der Waals density functional (vdW-DF) scheme proposed by Dion et al.49 was used to account for vdW interaction. The exchange effect of the electron was described with the optB86b exchange functional.50 This combined optB86b-vdW functional was verified to have a good description of the vdW interaction between molecules and metal surfaces.50,51 Optimized lattice constants of 3.941 Å for Pd and a = 5.18 Å, c = 5.65 Å for Pd4S were used. For Pd(111), Pd(100), and Pd(110) surfaces, slabs of 4, 4, and 5 layers, with supercells of (5 × 5), (4 × 4), and (3 × 4) size were used, respectively. For Pd4S(110), Pd4S(100), Pd4S(001), Pd4S(102), and Pd4S(112) surfaces, (2 × 2) super cells of about 9 Å atomic layer thickness were used, respectively. Slabs were separated with 20 Å of vacuum. The upper two layers of each slab were relaxed. A plane wave basis set with energy cutoff of 400 eV and (2 × 2) k-points were used to ensure the convergence of total ground-state energy. All geometrical optimization was performed until the force acting on relaxing atoms was less than 0.02 eV Å−1. Results and Discussion Synthesis and characterization of the catalysts The first step of the synthesis is to prepare the SC support by carbonization of molecular precursors with SiO2 nanoparticles as hard templates (see Experimental Methods for details).52,53 The SC support possessed a high specific surface area of 1164 m2 g−1, a large pore volume of 3.24 cm3 g−1, and a particularly high S content of 13.2 wt % ( Supporting Information Figures S1 and S2a and Table S1). The Pd4S/SC catalyst was then prepared by the wet impregnation of H2PdCl4 (5 wt % Pd) and subsequent H2 reduction at 500 °C. The S-enriched carbon supports enabled the reaction of Pd with S atoms to form Pd4S nanoparticles based on the RMSI (Figure 1a). As far as we know, it is the first example of synthesizing metal sulfide catalysts by RMSI on the metal–carbon interface. Figure 1 | Synthesis and characterization of the catalyst. (a) Schematic illustration for the preparation processes of Pd4S/SC catalyst based on RMSI. (b) Low-magnification HAADF-STEM image of the Pd4S/SC catalyst. Inset shows the particle size distribution of Pd4S. (c) High-resolution HAADF-STEM of a Pd4S nanoparticle. (d) EDS elemental mapping images of the Pd4S/SC catalyst. (e) Line-scan EDS elemental distribution curves of an individual Pd4S nanoparticle. (f) XRD pattern of the Pd4S/SC catalyst. (g) Crystal structure model of Pd4S. Download figure Download PowerPoint We first made HAADF-STEM observations to characterize the microstructure of the prepared catalysts. HAADF-STEM images and the corresponding particle size distribution showed that the nanoparticles with an average diameter of 5.8 ± 1.8 nm were homogeneously distributed on the SC supports (Figure 1b, the inset in Figure 1b, and Supporting Information Figure S3). The high-resolution HAADF-STEM image of an individual nanoparticle revealed well-defined lattice fringes with spacings of 0.243 and 0.220 nm, which were consistent with (102) and (112) planes of Pd4S (Figure 1c). Additionally, HRTEM also showed two sets of lattice fringes from Pd4S ( Supporting Information Figure S4). EDS elemental mapping images indicated that the S element was distributed over the whole carbon matrix, while line-scan curves confirmed that both of Pd and S elements were concentrated in individual particles (Figures 1d and 1e). Subsequently, we performed XRD to identify crystalline phase compositions of the prepared catalyst. The XRD pattern displayed main diffraction peaks at 35.2°, 36.6°, 39.5°, 40.8°, 42.8°, 48.2°, and 51.8° (Figure 1f), which corresponded well to the (200), (102), (210), (112), (211), (202), and (212) planes, respectively, of tetragonal Pd4S (JCPDS 73-1387). The space group of Pd4S is P-421c with eight palladimn atoms in general positions and two sulfur atoms in positions (0, 0, 0) and (1/2, 1/2, 1/2) (Figure 1g).54 Overall, the above characterizations unambiguously proved the formation of highly crystalline Pd4S nanoparticles on the SC supports based on RSMI. We then explored the effect of reduction temperature and Pd contents on RSMI for the catalyst preparation. All of the catalysts are referred to as a-PdxS/SC-b, where a and b are the weight ratio of Pd to SC and reduction temperatures, respectively. XRD analyses clearly indicated the vital role of reduction temperature on RSMI for the catalyst synthesis ( Supporting Information Figure S5). Unlike the highly crystalline Pd4S phase of (i.e., diffraction peaks were for the sample prepared at °C ( Supporting Information Figure structure or the size of the For the °C sample we the of diffraction peaks corresponding to Pd4S and metallic Pd ( Supporting Information Figure the of Pd4S to form metallic Pd phase at °C. with the of Pd structures on temperatures, the particle size from ± nm for to ± nm for when the temperature from to °C ( Supporting Information Figures and we the temperature at 500 °C and the Pd contents in the of wt nanoparticle size were ( Supporting Information Figures and However, we also the of the metallic Pd phase for the wt sample at 500 °C ( Supporting Information Figure the of S atoms in the SC supports to form Pd4S for a high Pd Additionally, N2 tests revealed that all of the a-PdxS/SC-b catalysts possessed high specific surface of ca. m2 and large pore of cm3 ( Supporting Information Figures and Table S1). properties of the catalysts We further conducted XAFS to reveal the of reduction on the and structures of the catalysts based on RSMI. The Pd K-edge XANES in Figure showed that the near-edge of was to that of while those of and were to that of Pd that the of Pd in and were to and respectively. The and peaks or with the reduction that the of the Pd element in Figure shows the EXAFS curves of as well as those of and Pd a peak at ca. 1.8 Å, well with coordination of the and peak for was the reduction temperature was to 500 and a new peak at ca. Å to coordination was the of coordination with the reduction the from to Figure 2 | properties of the catalysts. (a) XANES spectra at the Pd K-edge of the catalysts, and Pd The inset is the image of XANES (b) at the Pd K-edge of catalysts, and Pd (c) of catalysts, and Pd (d) Pd spectra of catalysts. Download figure Download PowerPoint analysis was further conducted to the atomic of Pd in (Figure The of showed only one at 1.8 Å to while Pd displayed only one at Å from the of coordination in of catalysts when the reduction temperature from to the of and the of Furthermore, we carried out EXAFS fitting to obtain the quantitative chemical of the Pd atoms in the catalysts ( Supporting Information Figure and Table For was only one coordination (i.e., at ca. Å with a coordination of ca. very to the with an average of Å and a coordination of about As and possessed two coordination that is, and The coordination of and were and in and and in respectively. Pd EXAFS results in for catalysts, and Pd are in Supporting Information Figure We also performed tests to further characterize the properties of catalysts (Figure The Pd of of Pd and peaks at and respectively, which are the of The Pd and Pd peaks for and and two peaks at and eV which were of The ratios of were 0, and for and respectively. In the S L-edge XANES and we only the from the SC support and not the from ( Supporting Information Figures and due to the low content of with that of in catalysts. the above and we proposed the process of Pd species on the SC supports with the reduction temperature based on RMSI. the low reduction temperature of the Pd precursors react with S atoms that are extracted out from SC supports to form with coordination 500 with the the Pd and S atoms in the particles will into well-defined crystalline Pd4S the temperature to °C will S atoms to the Pd4S that is, the of Pd4S into metallic Catalytic performance in hydrogenation of 4-chloronitrobenzene We hydrogenation of 4-chloronitrobenzene as a model reaction to the catalytic performance of a-PdxS/SC-b catalysts, in which the selective hydrogenation of the nitro group to produce is desired (Figure The catalytic reactions were performed under a mild = 80 = 0.6 for all the a-PdxS/SC-b catalysts. Commercial Pd/C and Pt/C were also under the for As the commercial Pd/C and Pt/C catalysts showed the high conversion of after 1 h reaction but poor selectivity of and (Figure as on both catalysts to form the aniline as The SC support not catalytic activity the 4-chloronitrobenzene the catalyst a selectivity of with the complete conversion with 1 h reaction (Figure Additionally, we conducted the hydrogenation reaction under harsh reaction conditions, reaction time and reaction to further demonstrate the selectivity of the catalyst. As in Supporting Information Figure excellent selectivity of when the reaction time was to h, while the selectivity to for the Pd/C catalyst. Additionally, also selectivity above at the high temperature of an poor selectivity of was obtained for Figure 3 | Catalytic performance in hydrogenation of (a) Hydrogenation reaction of and and product for the hydrogenation of