Article Cite This: Inorg. Chem. XXXX, XXX, XXX−XXX
pubs.acs.org/IC
Peroxo-Cerium(IV)-Containing Polyoxometalates: [CeIV6(O2)9(GeW10O37)3]24−, a Recyclable Homogeneous Oxidation Catalyst Hafiz M. Qasim,† Wassim W. Ayass,† Patrice Donfack,§ Ali S. Mougharbel,† Saurav Bhattacharya,† Talha Nisar,§ Torsten Balster,§ Albert Solé-Daura,∥ Isabella Römer,† Joydeb Goura,† Arnulf Materny,§ Veit Wagner,§ Josep M. Poblet,∥ Bassem S. Bassil,*,†,‡ and Ulrich Kortz*,† Downloaded via NOTTINGHAM TRENT UNIV on August 16, 2019 at 03:35:18 (UTC). See https://pubs.acs.org/sharingguidelines for options on how to legitimately share published articles.
†
Jacobs University, Department of Life Sciences and Chemistry, Campus Ring 1, 28759 Bremen, Germany University of Balamand, Faculty of Arts and Sciences, P.O. Box 100, 1300 Tripoli, Lebanon § Jacobs University, Department of Physics and Earth Sciences, Campus Ring 1, 28759 Bremen, Germany ∥ Universitat Rovira i Virgili, Departament de Química Física i Inorgànica, c/Marcel·lí Domingo 1, 43007 Tarragona, Spain ‡
S Supporting Information *
ABSTRACT: The class of peroxo-cerium-containing polyoxometalates has been discovered via the synthesis of the 9-peroxo-6cerium(IV)-containing 30-tungsto-3-germanate, [CeIV6(O2)9(GeW10O37)3]24− (1). Polyanion 1 consists of a cyclic [Ce6(O2)9]6+ assembly that is stabilized by three dilacunary [GeW10O37]10− Keggin fragments. The title polyanion 1 is solution-stable, on the basis of 183W nuclear magnetic resonance, and was shown to act as a recyclable homogeneous catalyst for the selective, microwave-activated sulfoxidation of the model substrate methionine to the sulfoxide in the absence and to the sulfone in the presence of hydrogen peroxide. Solution and solid-state Raman as well as solid-state infrared studies of 1 demonstrated the complete loss (and regain) of the nine peroxo groups in situ during the catalytic cycle, suggesting that the peroxo-free {Ce6(GeW10)3} skeleton remains most likely intact during the catalytic cycle. Solid-state X-ray photoelectron spectroscopy measurements showed that peroxo loss is accompanied by reduction of the cerium ions from +4 to +3, which is fully reversible. Density functional theory calculations are in complete agreement with all of these observations and furthermore suggest that the reduction of the six cerium(IV) ions is accompanied by the formation of molecular dioxygen.
■
sulfone.9 Herein, we report on the discovery and characterization of the first peroxo-cerium(IV)-containing POM.
INTRODUCTION
■
One of the main challenges in homogeneous oxidation catalysis is the ability to synthesize and test “recyclable” catalysts, with reproductive efficiency and quantitative recuperation after the catalytic cycle.1 There has been much interest in transition metal complexes for the catalytic oxidation of organic substrates.2 In this regard, polyoxometalates (POMs), as discrete polynuclear metal-oxo complexes with a large structural and compositional diversity, are also attractive candidates, as well as in other areas such as magnetism, material science, and biomedicine.3 Peroxocontaining POMs are known, and some are highly active in oxidation catalysis, 4 including the Venturello ion [PW4O8(O2)8]3−.5 Peroxo-containing polyoxomolybdates and -niobates as well as lanthanum-containing peroxopolytungstates have also been synthesized.6−8 Our group has reported dimeric and trimeric peroxo derivatives of zirconium- and hafnium-containing POMs, which were active in the stoichiometric oxidation of methionine to the sulfoxide or © XXXX American Chemical Society
EXPERIMENTAL SECTION
General Information. All reagents and chemicals were of high purity and were used as purchased without further purification. The FT-IR spectra (KBr pellet) were recorded on a Nicolet-Avatar 370 spectrometer. Thermogravimetric analysis (TGA) was carried out with a TA Instruments Q 600 device at a heating rate of 5 °C/min under a nitrogen atmosphere. The 183W and 1H NMR spectra were recorded on a JEOL ECS400 instrument. Elemental analysis was performed at CREALINS (Villeurbanne, France). Synthesis of Na-1. CeCl3·7H2O (0.066 g, 0.177 mmol) was dissolved in 20 mL of a 2 M NaCl aqueous solution at pH 5.0. Subsequently, Na10[A-α-GeW9O34]·18H2O (0.500 g, 0.177 mmol) synthesized according to the published procedure was added.10 The reaction mixture was stirred for 30 min at 50 °C, and then 1 mL of 30% hydrogen peroxide was added to the solution, which was kept at 50 °C for an additional 1 h, at which point a color change to yellow Received: April 23, 2019
A
DOI: 10.1021/acs.inorgchem.9b01164 Inorg. Chem. XXXX, XXX, XXX−XXX
Article
Inorganic Chemistry was observed. The reaction mixture was cooled to room temperature and then filtered. The solution was left for 1 day in a closed vial. On the following day, 0.5 mL of 30% H2O2 was added and the mixture heated to 50 °C for 15 min. The reaction mixture was then filtered at room temperature. After the solution had been left for 1 week in a closed container, a dark red crystalline product was obtained, which was collected and air-dried. Yield: 0.030 g (10%). IR (Figure S1). Elemental Anal. Calcd for Na24[CeIV6(O2)9(GeW10O37)3]·100H2O (Na-1): Na, 5.03; Ce, 7.65; Ge, 1.96; W, 50.0. Found: Na, 5.02; Ce, 7.76; Ge, 1.97; W, 50.1. The number of crystal waters was confirmed by TGA (Figure S2). We have also made the silicon analogue of 1, [CeIV6(O2)9(SiW10O37)3]24− (2), on the basis of single-crystal XRD and TGA. Na24[Ce6(O2)9(SiW10O37)3]·∼100H2O (Na-2) is isomorphous with Na-1: triclinic, P1̅, a = 22.2638(7) Å, b = 23.5142(7) Å, c = 23.5205(6) Å, α = 119.8558(13)°, β = 102.2457(17)°, γ = 99.0349(18)°, V = 9922.5(5) Å3, R = 0.0546. See also Table S1. Single-Crystal X-ray Diffraction. A single crystal of Na-1 was mounted on a Hampton cryoloop in light oil for data collection at 100 K. Indexing and data collection were performed on a Bruker D8 SMART APEX II CCD diffractometer with κ geometry and Mo Kα radiation (graphite monochromator; λ = 0.71073 Å). Data integration was performed using SAINT.11 Routine Lorentz and polarization corrections were applied. A multiscan absorption correction was performed using SADABS.12 Direct methods (SHELXS97) successfully located the tungsten atoms, and successive Fourier syntheses (SHELXL14) revealed the remaining atoms.13 Refinements were full matrix least-squares against |F2| using all data. In the final refinement, all nondisordered heavy atoms (W, Ge, Ce, and Na) were refined anisotropically; oxygen atoms and disordered countercations were refined isotropically. No hydrogen atoms were included in the models. For the sake of overall consistency, we show in the CIF file the same formula unit as in the text, with the exact number of countercations and crystal waters on the basis of elemental analysis and TGA, as this reflects the true bulk composition of the compound. Single crystals of Na-1recharged were obtained after the catalytic reaction and recharging of peroxo-depleted 1 with fresh H2O2, and the XRD measurements demonstrated that 1 and 1recharged are identical. See Tables S1 and S2 for crystallographic details. SO Formation. The homogeneous oxidation of DL-methionine (S) by 1 was performed by dissolving S (10 mg, 0.067 mmol) in 2.5 mL of water, and then Na-1 (100 mg, 0.009 mmol) was added before the solution was heated by microwave irradiation (power of 30 W) at 50 °C for 50 min. SO2 Formation. The homogeneous oxidation of DL-methionine (S) by 1 and hydrogen peroxide was performed by dissolving S (13 mg, 0.087 mmol), Na-1 (92 mg, 0.0084 mmol), and 0.05 mL of 30% H2O2 (0.75 mmol) in 2.5 mL of water. The solutions were then heated by microwave irradiation at 50 °C (power of 30 W) for 35 min. Catalyst Recovery. The depleted POM catalyst was recharged by addition of 0.05 mL of 30% H2O2 to the solution. The formation of charged 1 can be seen by a change in the color of the solution from light yellow (SO formation pathway) or light red (SO2 formation pathway) to dark red (Figure S5). Polyanion 1 can be easily recrystallized after several catalytic cycles by addition of 2 mL of a 2 M NaCl solution to the reaction mixture. The turbid solution turns transparent upon being heated at 50 °C for 5 min. Single-crystal XRD of these crystals confirmed the presence of polyanion 1. Solution Raman Studies. The 514.5 nm line of an argon ion laser source (Inova 308, Coherent Inc.) was focused using a microscope setup equipped with a long-working distance 50× objective (MPlan, Olympus) to excite the sample contained in a quartz glass cell with a maximum laser power of ∼10 mW at the sample cell. Then the scattered light collected in a 180° backscattering geometry through the same objective was filtered by a Notch filter for Rayleigh scattering rejection, dispersed in a single monochromator (Triax 550, Jobin Yvon) using an entrance slit width of 100 μm and a high-resolution diffraction grating with 2400 grooves/mm, and detected on a high-pixel density liquid nitrogen-
cooled and ultraviolet-enhanced back-illuminated CCD detector having a chip size of 2048 × 1024 pixels (CCD 3500, Jobin Yvon). As the single-crystal XRD data suggested not significantly different bond strengths for the peroxo groups in the POM, high-spectral resolution optics (grating and detector) were needed for the Raman experiments. With these settings, the most characteristic Raman spectral region of the POM (240−1226 cm−1) required twice the CCD spectral window and was recorded with an exposure time of ≤120 s and over three averaged signal accumulations per spectral window. Toluene was used for both wavenumber position and intensity calibration of the Raman setup. The measurement control software NGSLabSpec (Jobin Yvon) was also used for Raman data analysis. The raw spectra were noise-filtered by being processed five times through a 4-degree and 7-size Savitsky−Golay least-squares smoothing algorithm, and then the baseline caused by a fluorescence background in each measured spectrum was approximated by a line fit (with a set of manually fixed fitting nodes for all spectra) and subtracted. XPS Measurements. To measure the photoelectron spectra of differently treated layers of the Ce-containing polyanion salt Na-1, a 50 nm thick Au film was deposited by means of an electron beam evaporator on a piece of a silicon wafer. The as-prepared, the depleted, and the regenerated POM were dispersed in acetone and deposited by drop casting. After deposition, the samples were introduced into the XPS vacuum vessel, which was equipped with a photoelectron spectrometer consisting of a hemispherical analyzer (Specs Phoebos 100) and a Mg/Al X-ray gun (Specs XR-50) with an angle of 45°. For all measurements, Mg Kα1,2 radiation (E = 1253.6 eV) was used as source of excitation, whereas the analyzer was operated in fixed analyzer transmission mode with a pass energy of 50 eV. The energetic shift in binding energy positions due to charging of the samples was referenced to the C 1s peak. Overview (not shown here) and Ce 3d spectra were measured for the differently treated POM samples. The measured data were analyzed using the CASAXPS software. The background was subtracted using Shirley’s method. The different Ce 3d doublets were fitted using two simplified Voigt functions with the same full width at half-maximum and an intensity ratio of 0.6 between the 3d3/2 and 3d5/2 peaks. Spectra of (NH4)2Ce(NO3)6 and CeCl3 were taken as references for the CeIV and CeIII oxidation states, respectively (see the Supporting Information). Computational Details. DFT calculations were carried out at the B3LYP level14 using the Gaussian 09 package.15 The MWB28 basis set16 was used to describe Ce, while the LANL2DZ basis set17 was used for the remaining atoms, with the exception of oxygen atoms directly bonded to Ce, which were described by a Pople-type double-ζ basis set supplemented with polarization functions.18 Solvent effects of water were included in the calculations by means of the IEF-PCM implicit solvation model19 as implemented in Gaussian09. Geometry optimizations were performed without any symmetry restriction. Due to the complexity of the system, we set the optimization convergence criteria to a maximum step size of 0.01 au and a root-mean-square force of 0.0017 au. However, it must be pointed out that the maximum force at the last step of geometry optimization is
