Photocatalytic proton reduction with ruthenium and cobalt complexes immobilized on fumed reversed-phase silica

Non-covalent binding of water reduction catalysts and photosensitizers to hydrophobic silica particles represents a general approach for heterogenizing molecular water splitting components.


General
All dry and air free reactions were performed on a standard nitrogen/vacuum line in glassware dried at 140°C at least overnight. THF was dried over Na/Benzophenone and dry DMF purchased from Acros. All other solvents were of analytical grade and used without further purification. Water for catalytic reaction was doubly distilled before use and extractions and column chromatographies were done with normally distilled technical solvents. Chemicals: Trichlorooctadecyl silane, sodium hydride (60 % dispersion), 1-iodooctadecane, 1-bromooctadecane, nitric acid, hydrophilic fumed silica, porous hydrophobic silica (45-70 µm, pore size: 7 nm, pore volume: 0.7-0.9 cm 3 /g), sodium 4-dodecylbenzenesulfonate (Na[C 12 -PhSO 3 ]), sodium ascorbate, ammonium hexafluorophosphate and lithium diisopropylamine (LDA, 2 M in THF) were purchased from sigma aldrich, porous hydrophilic silica (d = 10 µm) from Marchery Nagel, ascorbic acid from Acros, acetic acid and ethanol from Merck and 4,,N,N, and sodium trifluoromethanesulfonate (NaOTf) from TCI. All products were directly used without further purification and syntheses were monitored by HPLC measurements. 1 H-and 13 C-NMR measurements (200 MHz or 300 MHz) were performed on a Varian Mercury or a Varian Gemini-2000 spectrometer. Chemical shifts are reported relative to the residual solvent peaks. Solid and solution NMR of PS 5 shown below were run on a Bruker AV3-500 MHz NMR spectrometer (500.25 MHz 1 H-and 125.78 MHz 13 C-frequency). The solid state 13 C measurements have been carried out under Hartman-Hahn condition and magic angle spinning ( 13 C-CPMAS) with high power 1 H-decoupling. Elemental analyses were performed on a Leco CHNS-932 elemental analyzer. ESI-MS were recorded on a Brucker HCT ultra mass spectrometer. UV/VIS absorption spectra were measured on a Varian Cary50Scan UV-Visible Spectrophotometer. Luminescence spectra were recorded on a Perkin Elmer LS 50B. HPLC measurements were performed on a VWR Hitachi Elite LaChrome with a C 18 -Phenomenex Core Shell column and H 2 O/MeOH as eluent (gradient starting with 10 % MeOH, 0.1 % TFA in H 2 O to pure MeOH). Lifetime and transient absorption measurements were recorded on an Edinburgh LP920 Laser Flash Photolysis transient absorption spectrometer using a flash lamp pumped Qswitched Nd:Yag laser (355 nm) as excitation source. Samples were degassed by bubbling with N 2 prior to measurements. The laser power was adjusted to ca 1.2mJ/shot in order to avoid double excitation of Ru PS as observed by Brettel et al. 1 Relevant components of a singular value decomposition (SVD) analysis were exponentially fitted to determine the corresponding reaction rates. Transmission electron microscopy (TEM) was done on a FEI Tecnai G2 Spirit operated at 120 kV. One drop of a dilute aqueous silica suspension was placed onto a formvar-coated copper grid, left for ca 30 s followed by removing residual liquid with a tissue and drying the sample overnight at rt. Fluorescence microscopy was performed on a Leica SP5 Confocal laser scanning microscope (Objective: 63 x 1.4; oil immersion). Dynamic light scattering (DLS) measurements were performed on a DynaPro Titan machine equipped with a temperature controlled microsampler from Wyatt Technology Corporation. Data were analyzed with the Dynamics V6 software.
The surface area was determined by the Brunauer−Emmett−Teller (BET) method on a Quadrasorb SA 6 in N 2 adsorption mode. Samples (ca. 300 mg) were degassed at 423 K for 24 h in vacuo prior to nitrogen adsorption measurements with a Quantachrome FLO VAC degasser. Figure SI1: Green: 13 C-CPMAS solid state NMR of f-SiO 2 -C 18 without adsorbants. Black: 13 C CPMAS solid state NMR of PS 5 adsorbed on f-SiO 2 -C 18. The additional peaks between 40 and 80 ppm are assigned to surfactant 7, which was adsorbed during immobilization of PS and WRC. Red: 13 C-CPMAS solid state NMR of pure PS 5. Blue: 13 C-NMR of PS 5 in MeOD.

Structures of used surfactants
Scheme SI1: Structures of the used surfactants.

DLS measurements
Stock suspensions for DLS measurements were prepared at the same concentrations as catalytic suspensions (7-8 mg/mL) and diluted 1:1000 with water prior to measurements.

Fluorescence microscopy:
Suspensions for FM measurements were prepared the same way as for catalysis but without ascorbate buffer. One drop (ca 10 µL) was placed between two glass plates separated by a sticking mask (d = 100 µm) with a measurement window (r = 3 mm).   Figure SI3, the length of the scale bar is 10 µm and the distance between each picture 3 µm. Figure SI5: Representative fluorescence micrograph of loaded glowed, non-porous silica in water with 0.1 mM surfactant 6 and 0.1 M NaOTf electrolyte at the same particle concentration as used in catalysis (50 mg/mL). The length of the scale bar is 10 µm. Figure SI6: Representative fluorescence micrograph of a selected non-porous silica particle at different depths (z-direction). The conditions are the same as in Figure SI5 the distance between each picture is 2 µm. Figure SI7: Representative fluorescence micrograph of loadedporous silica in water with 3 mM surfactant 6 and 0.1 M NaOTf electrolyte at the same particle concentration as used in catalysis (4 mg/mL). The length of the scale bar is 50 µm. Figure SI8: Representative fluorescence micrograph of a selected porous silica particle at different depths (z-direction). The conditions are the same as in Figure SI7 and the distance between each picture is 5 µm.

Transient absorption measurements
Quenching of Ru* with WRC 1 ([CoBr(appy)]Br) was examined in pure water with 100 µM [Ru(bpy) 3 ]Cl 2 and varying concentrations of 1. No formation of Ru III or Co I was observed, but non-productive quenching of Ru* by 1. The quench rate was estimated by plotting k obs (1/τ) against the Co concentration ( Figure SI3). The quench rate of Ru* by ascorbate was examined in 0, 0.5 and 0.95 M aqueous ascorbate buffer (pH 4) with 50 µM [Ru(bpy) 3 ]Cl 2 . Transient absorption spectroscopy clearly showed formation of reduced PS ( Figure SI4). Reaction rates (k obs ) were obtained by exponential fit of the first 2 components of a SVD analysis. The quench rate was determined by plotting k obs against the ascorbate concentrations ( Figure SI5)  The electron rate transfer rate from reduced PS to WRC (PS -+ Co II → PS + Co I ) was determined in 0.95 M aqueous ascorbate buffer with 50 µM [Ru(bpy) 3 ]Cl 2 and 0, 100 and 200 µM [CoBr(appy)]Br (1). Transient absorption spectroscopy clearly showed formation of PSfollowed by formation of Co I ( Figure SI6). The reaction rates (k obs ) were obtained by exponential fit of the first 3 components of a SVD analysis. The electron transfer rate was determined by plotting k obs against the WRC concentrations ( Figure SI7).