Design, Identification, and Evolution of a Surface Ruthenium(II/III) Single Site for CO Activation

Abstract RuII compounds are widely used in catalysis, photocatalysis, and medical applications. They are usually obtained in a reductive environment as molecular O2 can oxidize RuII to RuIII and RuIV. Here we report the design, identification and evolution of an air‐stable surface [bipy‐RuII(CO)2Cl2] site that is covalently mounted onto a polyphenylene framework. Such a RuII site was obtained by reduction of [bipy‐RuIIICl4]− with simultaneous ligand exchange from Cl− to CO. This structural evolution was witnessed by a combination of in situ X‐ray and infrared spectroscopy studies. The [bipy‐RuII(CO)2Cl2] site enables oxidation of CO with a turnover frequency of 0.73×10−2 s−1 at 462 K, while the RuIII site is completely inert. This work contributes to the study of structure–activity relationship by demonstrating a practical control over both geometric and electronic structures of single‐site catalysts at molecular level.


Experimental Procedures Materials Synthesis
To prepare PPhen-bipy, 1,2,4,5-tetrabromobenzene (1.316 g, 3.34 mmol), 5,5'-dibromo-2,2'-bipyridine (0.342 g, 1.09 mmol), benzene-1,4-diboronic acid (1.290 g, 7.78 mmol), K2CO3 (2.0 M aqueous solution, 15 mL) and Pd(PPh3)4 (0.30 g, 0.26 mmol) were added into 120 mL dimethylformamide (DMF). The mixture was degassed through three freeze-pump-thaw cycles. The mixture was then purged with Ar and refluxed at 150 °C for 20 h under stirring. After the heating, the mixture was poured into 600 mL water under fierce stirring to make K2CO3 and DMF dissolve in water. The product was collected by filtration and was washed with water and methanol. Approximately 1 g of dark-green product was obtained in each batch. To remove the Pd in the crude product, the collected product was dispersed in 200 mL 30% H2O2 aq./37% HCl aq./H2O/EtOH solution (volume ratio 1:5:5:10) and kept stirring at room temperature overnight and then heated at 60 °C for 4h. The bulk Pd cluster was oxidized to Pd 2+ and dissolved into solution. After the heating, the product was collected by filtration and was washed with water and methanol, dried at 60 °C overnight. To further remove the Pd 2+ coordinated with the bipyridine ligand, the 1 g of light-yellow product was dispersed in 200 mL 32% NH3 aq./H2O/EtOH solution (volume ratio 1:15:30) with 10 g Na2EDTA·2H2O and kept stirring at 70 °C for 24h. After the heating, the product was collected by filtration and was washed with water and methanol. The final product was earthy yellow and dried at 60 °C in a vacuum oven overnight.
PPhen-[bipy-RuCl4]H was synthesized via the following steps. 1.0 g of PPhen-bipy product obtained via the previous method and 150 mg RuCl3·xH2O were added into 250 mL round bottom flask with 50 mL 10% HCl aq. and 50 mL ethanol. The mixture was stirred at 60 °C for 5 hours, during this time the colour of solution changed from dark brown to reddish brown. Then the product was collected by filtration and washed with water and methanol until the filtrate become colourless. The final product was dried at 60 °C in a vacuum oven overnight.

Calculation of theoretical bipy density in PPhen-[bipy-RuCl4]H
1,2,4,5-tetrabromobenzene (Mw. = 393.70 g·mol -1 , 1.316 g, 3.34 mmol), 5,5'-dibromo-2,2'-bipyridine (Mw. = 313.98 g·mol -1 , 0.342 g, 1.09 mmol) and benzene-1,4-diboronic acid (Mw. = 165.75 g·mol -1 , 1.290 g, 7.78 mmol) are used for synthesis of PPhen-bipy support. Such ratio of these precursors ensures the -Br/-B(OH)2 ratio is equal to 1. We assume the Suzuki C-C coupling is 100% efficient, which stoichiometrically subtract all the -Br and -B(OH)2 groups, leaving only the phenyl groups in the final product. Hence the yield of PPhen-bipy product is: (PPhen-bipy) = The ratio of actual Ru loading to the maximum bipy density is: The possible explanations for this 67% ratio are: (1) The micro-structure of PPhen-bipy is made by random connections between phenyl groups and bipy groups, accordingly the distribution of bipy within the PPhen-bipy is homogeneous at macro level but inhomogeneous at micro level. Thus, the pore size and shape for each bipy ligand varies, and it is possible that some of the bipy ligands are inaccessible for the metal cations. (2) The polyphenylene framework is hydrophobic, which to some extent will prevent the infiltration of the aqueous solution and inhibit the sufficient transportation of metal cations from liquid phase to solid phase.
(3) Not all the bipy ligands are converted into PPhen-bipy, leading to the decreased amount of actual bipy loading.

STEM and EDS study
High resolution aberration-corrected Annular Bright Field (ABF) and High Angle Annular Dark Field Scanning Transmission Electron Microscopy (HAADF-STEM) studies were performed at the electron Physical Science Imaging Centre (ePSIC). ABF and HAADF-STEM images were acquired simultaneously on probe-corrected (JEOL-COSMO) JEM ARM 300CF (JEOL, Japan) operated at 300 kV. All samples were prepared by sprinkling a small amount of dry powder on 400 mesh gold/copper grids with lacey carbon support film. A 30 μm probe-forming aperture was used, resulting in 22.4 mrad probe convergence semi-angle. A probe size of 8C was chosen to maximize the spatial resolution. The HAADF signal was gathered at 9.0 cm STEM camera length, integrating the scattered electron intensity between 77 and 209 mrad. In order to mitigate accumulation of carbon contamination during STEM imaging, the regions of interest were exposed to an intense electron "beam shower" for 15 min.
Energy-Dispersive X-ray Spectroscopy (EDS) results were obtained on the probe-corrected (CEOS) JEM ARM 200CF (JEOL, Japan) operated at 200 keV, which is equipped with large solid-angle dual EDX detectors for X-ray spectroscopy and elemental mapping. The EDX data acquisition was carried out in STEM imaging mode. Each EDX spectrum image is 80 × 80 pixels in size, with 0.1 second exposure time per pixel. The mapping procedure was performed with spatial drifting correction every 30 seconds. Gatan Microscopy Suite Software was used for EDS spectrum imaging data acquisition. The in situ XAFS study of PPhen-[bipy-RuCl4]H catalysts was performed in a plug-flow microreactor with same X-ray beam setting-up and data acquisition parameters. The catalysts powder was packed into a kapton foil tube (diameter 5 mm) with quartz wool at both ends. A controlled flow of mixed gases via 4 mass flow controllers (pure He, 1% CO in He, 2% H2 in He and 5% O2 in He, respectively) was introduced into the reaction tube at ambient pressure. A hot air gun was placed under the reaction tube to heat the catalysts bed (heating and cooling ramp rate of 10 °C·min -1 and 20 °C·min -1 , respectively). The heating zone was sheathed with a ceramic drivepipe to improve the heat conductivity and prevent non-uniform heating. There are two 3 mm × 15 mm windows on both side of the ceramic drivepipe to let X-rays passing through. A K-type thermal couple was positioned inside the catalysts bed to monitoring the temperature. During the reaction, XAFS data was acquired continuously with spectra obtained every 135 seconds (kmax = 17, step size 0.3 eV). The outlet gas composition was analysed using a Quadrupole Mass Spectrometer Quantitative Gas Analyser (Hiden Analytical, UK). The Hiden QGA could continuously sampling and scan atomic mass range from 1 to 200 AMU with 500 times/second for measurement speed. The analysis sensitivity is 100% to 100 PPB subject to spectral interference. The gas or vapour spectrum were simulated and analysed by automatic subtraction of spectral overlaps.

Ex situ and in situ XAFS measurements
XAFS study of Cl K-edge and Ru L3,2-edge were conducted at BL6N1 beamline of Aichi Synchrotron Radiation Centre (Aichi Science & Technology Foundation, Japan). Soft X-ray beam was introduced via InSb (111) monochromator to a surface XAFS experiment chamber equipped with an electrostatic hemispherical photoelectron spectrometer (SPECS PHOIBOS 150). XAFS data were acquired in both conversion electron yield (CEY) mode and partial fluorescence yield (PFY) mode with energy range of 2730-3150 eV at 0.2 eV step size. Ru foil and RuO2 were measured as standard reference for energy calibration.
XAFS data was analysed by Demeter (including Athena and Artemis methods, version 0.9.25) [2] . Athena software was used for data extraction and XANES analysis. Artemis software was used to fit the Ru K-edge EXAFS data (fitting range 3.3 Å −1 < k < 13.7 Å −1 and 1.0 Å < R < 3.5 Å). The amplitude reduction factor S0 2 for each edge was calculated from EXAFS analysis of the metal foil and used as fixed parameter for EXAFS analysis.

DRIFTS, ATR-FTIR and Far-FTIR study
Diffuse reflectance infrared Fourier transform spectroscopy (DRIFTS) data were collected on an Agilent Carey 680 FTIR spectrometer (Agilent, UK) equipped with liquid nitrogen cooled MCT detector and Harrick reaction chamber. PPhen-[bipy-RuCl4]H powder was filled in a sample cup and placed inside the reaction chamber. The IR beam was directed into it by the Praying Mantis accessory. The time resolution of each spectrum is 60s (400-4,000 cm −1 ), while 64 scans were taken and merged with a resolution of 4 cm -1 per minutes. The outlet gas was analysed by the same Hiden QGA mass spectrometer.
Attenuated Total Reflection Infrared (ATR-IR) spectroscopy was measured with Bruker ALPHA FTIR spectrometer. The spectra were taken within 4000-400 cm -1 range.

Other characterisations
Powder X-ray diffraction (PXRD) measurement was performed using Bruker D8 diffractometer with a voltage of 40 kV, at 30 mA, using a Cu source with Kα1 = 1.540562 Å and Kα2 = 1.544398 Å. The contributions of Kα2 line in the XRD patterns were subtracted.
Nitrogen adsorption-desorption isotherms were recorded at 77 K using a Micromeritics 3Flex surface characterization analyser. The samples were degassed in vacuum at 200 °C overnight for removal of any contaminates. Specific surface areas were determined according to the BET model.
X-ray photoelectron spectrum (XPS) analysis were performed on a Thermo-scientific XPS K-alpha surface analysis machine using an Al source. The sample powder was immobilized on silicon chips for measurement. C 1s electron at 284.8 eV was used as standard reference to calibrate the photoelectron energy shift. XPS spectra in C 1s and Ru 3d region from 270 eV to 300 eV was collected (0.1 eV step size) and deconvoluted to identify the oxidation states of Ru. All the data analysis was performed on the CasaXPS software (version: 2.3.18PR1.0).
Thermogravimetry and Differential Scanning Calorimetry (TG-DSC) were used to measure the temperature-dependent mass changes and energetic effects. The experiment was carried out on STA 449 F5 Jupiter simultaneous thermal analyser (NETZCH, Germany). The thermal stability tests of PPhen-[bipy-RuCl4]H and PPhen-bipy in synthetic air and argon were performed. About 5 mg of sample powder was placed into a Pt-Rh crucible (with pierced lids) for the measurement. A gas flow of 70mL/min was employed for either synthetic air or argon (protective gas: 20mL/min & purge gas: 50mL/min). The temperature range was set from R.T. to 1000 °C with a ramp speed of 10 °C/min. The temperature was monitored via a type S thermocouple.

Computational Methods
The simulated structures were based on a simplified octahedral model complex: a bipyridine (bipy) with the two nitrogen atoms co-ordinated to the Ru and Ru further co-ordinated to CO and Cl. Calculations were done in Gaussian09 [3] using the B3LYP [4] functional and LANL2DZ basis set for Ru and 6-31G(d,p) basis set for all other elements.  There is no obvious Pd 3d signals in the XPS spectra, suggesting the Pd concentration is lower than the typical detection limits of XPS (0.1 atom%). [6] (c) XPS fine scan at C 1s and Ru 3d region. The C 1s peaks are overlapping with Ru 3d peaks due to the similar binding energies.  For PPhen-bipy, the TG curve showed slight mass loss of 3.2% at 400 °C, which was caused by the release of residual moisture. At peak temperatures of 474°C and 546 °C, distinctive overlapped exothermic effects were detected. The origin of these pronounced effects is due to the combustion of the organic content within the sample. 14.5% = x (Ru loading) ·Mw (RuO2) /Aw (Ru) +·(1-x (Ru loading) ·(Mw (RuCl3) + Mw (HCl) )/Aw (Ru) )·8.0%  Table S4.       Table S6.  Figure 3. Peak F, G and H are attributed to Ru 2p 1/2 → t 2g , Ru 2p 1/2 → eg and Ru 2p 1/2 → CO π*transition respectively. The peak F of 2p 1/2 → t 2g transitions are dipole forbidden in octahedral complexes, so it has been less pronounced compares to peak A at Ru L 3 -edge. [7] For Ru-bipy complex, it has been found that Ru 2p → bipy π* transitions exhibit nearly the same energy levels as eg orbitals, so it will overlap with peak B and G in both L 3 and L 2 edges. [7][8] Table S7. The shift of Ru K-edge position is strongly affected not only by oxidation states of Ru but also ligands and coordination structures. [9] The energy shift of Ru K-edge XANES is proportional to the oxidation state of Ru to a certain extent, but this is not the only dominant factor. The XANES features also contribute from the electron configurations, coordination geometries and ligand types. Such approach which to determine the oxidation states just based on the edge shift is priori limited to compounds with similar electronic configuration, local coordination and geometry.   Table S8. For WHSV = 60,000 mlh -1 g -1 (1% CO, 5% O 2 , and N 2 balancing), the conversion of CO at 428 K is 10.04 %. Therefore, the TOF value is: For WHSV = 600,000 mlh -1 g -1 (2.0% CO, 2.5% O 2 , and N 2 balancing), the conversion of CO at 479 K is 6.82 %. Therefore, the TOF value is:       Table S3. In sum, the selection of ligands for single sites is also controllable with this PPhen-bipy platform. It is worth mentioning that, the scattering features in the EXAFS between 2-3 Å are mainly due to the scattering from the four N-bonded carbon atoms of the bipy ligand. The other scattering features above 3 Å are attributed to multiple scattering. In comparison with 3d metals, the intensity of long path single scattering between 4d (or 5d) metal centre with these light back scatterers (especially far aromatic carbon atoms) is significantly reduced. This is the reason that the M-C scattering between 2-3 Å and multiple scattering above 3 Å decreased while the atomic number increased. Supporting Tables   Table S1. [a] The coordination number of Ru-N is set to 2.0 as a fixed parameter for EXAFS fitting. Table S2. Bond length calculated from the structures in Fig. S4 and Fig. S21. [a] The (ax) and (eq) are initials for axial position and equatorial position respectively.  [21] B Ru (III) 330 v(Ru-Cl) Ref [22] C Ru (III)  372 Ref [23] D Ru (III) 2064 v(C≡O) Ref [24] 2040 v(C≡O) 320 v(Ru-Cl)