Intermolecular O–O Bond Formation between High-Valent Ru–oxo Species

Despite extensive research on water oxidation catalysts over the past few decades, the relationship between high-valent metal-oxo intermediates and the O–O bond formation pathway has not been well clarified. Our previous study showed that the high spin density on O in RuV=O is pivotal for the interaction of two metal-oxyl radical (I2M) pathways. In this study, we found that introducing an axially coordinating ligand, which is favorable for bimolecular coupling, into the Ru-pda catalyst can rearrange its geometry. The shifts in geometric orientation altered its O–O bond formation pathway from water nucleophilic attack (WNA) to I2M, resulting in a 70-fold increase in water oxidation activity. This implies that the I2M pathway is concurrently influenced by the spin density on oxo and the geometry organization of the catalysts. The observed mechanistic switch and theoretical studies provide insights into controlling reaction pathways for homogeneous water oxidation catalysis.


General Procedures
All chemicals were purchased from commercial suppliers and used as received.All the solvents were purchased from Fisher Scientific Sweden and used as received.High resolution mass spectrometry was performed at the Dalian University of Technology and Westlake University (UPLC-HRTOF, Synapt-G2-Si, Waters). 1 H NMR and 13 C NMR spectra of the compounds were recorded with either a Bruker Ascend 400 or a Bruker Avance DMX 500 NMR spectrometer.
Electrochemistry measurements were carried out with a CHI660c potentiostat, with glassy carbon as working electrode, Pt wire as auxiliary electrode and measured versus Ag/AgCl/Sat.KCl as a reference electrode.All potentials reported herein are converted to their corresponding values versus NHE.Kinetics studies were carried out with an Omega PXM409 pressure transducer.
Single-crystal growth and measurement.Single crystal of 1 was obtained by slow diffusion of diethyl ether into a methanolic solution of 1 at room temperature.The diffraction data of 1 was measured at 293 K by using Mo Kα radiation on a Bruker D8 VENTURE single crystal Xray diffractometer equipped with a kappa geometry goniometer.The dataset was reduced, and absorption correction was applied in APEX3 suite.The crystal structure was solved and refined by SHELXL-2018 and OLEX2, respectively.(G.M. Sheldrick, Acta Crystallogr.Sect.A: Found.Crystallogr., 2007, 64, 112-122) The crystal structure was refined using full-matrix least-squares based on F 2 with all non-hydrogen atoms anisotropically defined.Hydrogen atoms were placed in calculated positions by means of the "riding" model.A summary of the crystallographic data, the data collection parameters, and the refinement parameters are given in Table S1.

Kinetic isotope effect (KIE) values calculation from electrochemical experiments
The CVS were recorded with a scan rate of 20 mV s −1 in 0. which is negligible.Therefore, we reported the KIE using Ag/AgCl as the reference to maintain the same driving force.
Computational details.All density functional theory (DFT) calculations for estimating Gibbs free energies were performed using the Jaguar 8.3 program package from Schrödinger LLC. 1 We employed Becke's three-parameter hybrid functional combined with the LYP correlation functional (B3LYP), 2 along with D3 dispersion corrections as proposed by Grimme et al. 3,4 The LACVP** core potential and basis set were used for the optimization of molecular geometries. 5To identify the transition states for O-O bond formation, we scanned the potential energy surface by varying the terminal O-O bond distance in the [ antiferromagnetic open-shell singlet state.The thermochemical corrections required to estimate the Gibbs free energy barrier from the prereactive dimers were computed at the B3LYP-D3/LACVP** level for both the prereactive dimers and the transition-state structures.
Single-point energy corrections were applied using the B3LYP-D3 functional with the LACV3P**++ basis set, which includes additional f-functions on the metal.Based on these gasphase optimized geometries, implicit solvation energies were estimated through single-point calculations employing the Poisson-Boltzmann reactive field (PBF) method in Jaguar, simulating water as the solvent.The Gibbs-free energy was defined by the following equation: G = E(B3LYP-D3/LACV3P**++ 2f on Ru) + G solv + ZPE + H 298 -TS 298 + 1.9 kcal mol - 1 (The value 1.9 kcal mol -1 is a concentration correction to the free energy of solvation, which by default is calculated at 1 M (g) to 1 M (aq) in Jaguar).
The geometries of both the reactant monomer and the product dimer were optimized using the DFT method outlined above.For the force field parameters of the Ru(pda)(isoq) complex, we utilized data from previous research 6 , supplemented with standard OPLS-AA (all-atom optimized molecular potential for liquid simulation) 7 bonded and van der Waals parameters for the isoquinoline ligand.Partial charges were determined using ESP charges, following the methodology described in earlier work. 6These parameters were then converted into the GROMACS topology format.The Fourier coefficients of the dihedral potential term were transformed into the Ryckaert-Bellemans form to facilitate molecular dynamics (MD) simulations.

8
MD Simulations.MD simulations were performed with the GROMACS 5.0.4MD software package. 9We performed a 100 ns MD run in a 52 × 50 × 50 Å 3 periodic box filled with TIP3P 10 water molecules and a chloride ion to neutralize the charge.In MD simulations, the resulting systems were subject to 100,000 steps of steepest descent minimization.The periodic boundary condition was applied in the simulation.The cutoff radii for the Lennard-Jones and electrostatic interactions were set to 10 Å.For accurate evaluation of long-range Coulombic interactions, the particle mesh Ewald 11 summation method is used for electrostatic interactions beyond the cutoff.The system was gradually heated to 300 K over 100 ps using a v-rescale thermostat 12 for canonical ensemble (NVT) simulations.During this phase, the linear constraint solver algorithm 13 was utilized to constrain all bond lengths.Subsequently, isothermal-isobaric ensemble (NPT) simulations were performed, maintaining the pressure at 1 bar over 100 ps, controlled by a v-rescale thermostat and the Parrinello-Rahman barostat.The systems were then simulated for an additional 100 ns.
Three independent MD simulations with different initial velocities were also carried out.A time step of 2.0 fs was used for all simulations.Trajectories were recorded every 200 ps (100,000 timesteps), and the final 50 ns of each trajectory were used to calculate the mean square displacement and obtain the diffusion coefficient of the catalyst.
Potential of Mean Force.The PMF calculations were performed for the two species of Ru(pda)(isoq) in the water phase using the umbrella sampling method with the GROMACS 5.0.4MD software package.Equilibrations were performed for 100 ps under an NVT ensemble, using the same methodology described above.Two Ru complexes were pulled apart using a spring constant of 30 kJ mol -1 Å -2 and a pull rate of 0.05 Å ps -1 .From these trajectories, snapshots were taken to generate the starting configurations for the umbrella sampling windows.Then, we sampled the distances using roughly 0.5 Å spacing.Such spacing allowed for increasing details at a smaller distance and resulted in windows.In each window, a 100 ps isothermal-isobaric ensemble (NPT) simulation, with the pressure set to 1 bar in, was used.Followed by a 10 ns MD run for a total simulation time utilized for umbrella sampling.Analysis of results was performed with the weighted histogram analysis method. 14All PMF simulations resulted in smooth dissociation curves, and the mean value was obtained from three repeated simulations.
EVB Simulations.The EVB simulations were conducted in both vacuum and aqueous phases to investigate the influence of the interface on the dimerization of the Ru(pda)(isoq) complex.Each molecular dynamics (MD) simulation involved sequential equilibration of the catalyst under conditions consistent with prior studies.The EVB simulations and mapping procedures followed the strategy detailed in previous work.b] All values are calculated with EVB in the TIP3P water environment.c] Gibbs free energy needed to rearrange the complex to the correct pre-reactive complex from the front-to-back geometry.
Table S3 Atom types and partial charges assigned to atoms in the complex Ru(pda)isoq 2 .The labelled atoms are shown in Figure S12.

Figure S1 .
Figure S1.Single crystal structures of 1 with thermal ellipsoids at 50% probability.Hydrogen atoms are omitted for clarity.

Figure
Figure S6 (a) CVs of complex 1 at different scan rates from 0.02 to 1 V s −1 , working electrode: GC.(b) Reaction order between peak current (E OX1 and E RE1 ) and the scan rate.The slope of 1indicates that this process is surface-controlled, while the slope of 0.5 indicates that this process is diffusion-controlled.

Figure S7
Figure S7 LSVs of 0.7 mM complex 1 in a 0.1 M phosphate buffer solution in H 2 O and D 2 O, (pH = 7 and pD = 7.87) containing 10% CF 3 CH 2 OH, scan rate = 100 mV s −1 , working electrode: GC.Each curve is the mean of the data fitting of three replicate experiments, and the shaded areas show the standard deviations for each series.

Figure S8 TOFs of complex 1 Figure S9
Figure S8TOFs of complex 1 at various concentrations.

Figure S10
Figure S10 DFT optimized geometries of the transition state in the I2M (left) and WNA (right) reaction pathway for Ru(pda)isoq 2 .

Figure S11
Figure S11DFT optimized geometries of the Ru(pda)isoq 2 with labelled atoms.

Figure S12
Figure S12Hydrogen bonds formed between oxygen atoms and water molecules from MD simulations.

Figure S13
Figure S13Radial distribution function between the oxo atom and water molecules from MD simulations.

Figure S14
Figure S14Free energy profile for the binding of two [Ru(pda)isoq 2 ] catalysts in the water phase from three replicated umbrella sampling simulations.The reaction coordinate corresponds to the distance between the Ru atoms of the two [Ru(pda)isoq 2 ] catalysts.

Figure S15
Figure S15Histogram of the reaction coordinates from three replicated umbrella sampling simulations for calculating binding free energy with the WHAM.

Figure S16
Figure S16 Free energy profile for the O-O bond formation of [Ru(pda)pic 2 ] (a) and [Ru(pda)isoq 2 ] (b) catalysts via the WNA reaction pathway.The snapshots presented in the profile are a configuration of a reactant (1), the transition state (TS), and a product (2) which were obtained from DFT simulations.

Figure S17
Figure S17 Free energy profile for the O-O bond formation of two [Ru(pda)pic 2 ] catalysts in the water phase.The snapshots presented in the profile are a configuration of two catalysts positioned relatively far apart (1), a front-to-back configuration (2), a prereactive dimer (3), the transition state (TS), and a product (4) which were obtained from umbrella sampling and EVB-MD simulations.The TIP3P water molecules have been omitted to enhance clarity.

Table S1
Summary of the crystallographic data for 1(CCDC 2133354)