Kinetics and mechanism of sequential ring methyl C–H activation in cyclopentadienyl rhodium(iii) complexes

We have studied activation of the methyl C–H bonds in the cyclopentadienyl ligands of half-sandwich Rh(iii) complexes [η5-CpXRh(N,N′)Cl]+ by observing the dependence of sequential H/D exchange on variations in CpX = Cp* (complexes 1 and 2), Me4PhCp (CpXPh, 3) or Me4PhPhCp (CpXPhPh, 4), and chelated ligand N,N′ (bpy, 1; phen, 2–4). H/D exchange was fastest in d4-MeOD (t1/2 = 10 min, 37 °C, complex 1), no H/D exchange was observed in DMSO/D2O, and d4-MeOD enhanced the rate in CD3CN. The proposed Rh(i)–fulvene intermediate was trapped by [4 + 2] Diels–Alder reactions with conjugated dienes and characterized. The Rh(i) oxidation state was confirmed by X-ray photoelectron spectroscopy (XPS). Influence of solvent on the mechanisms of activation and Diels–Alder adduct formation was modelled using DFT calculations with the CAM-B3LYP functional and CEP-31 g basis set, and influence on the reaction profile of the dimiine ligand and phenyl substituent using the larger qzvp basis set. The Rh(iii)–OH intemediate is stabilised by H-bonding with methanol and a Cp* CH3 hydrogen. The Rh(i)(Me4fulvene) species, stabilised by interaction of methanol with a coordinated water, again by two H-bonds H2O–HOMe (1.49 Å) and fulvene CH2 (1.94 Å), arises from synchronous transfer of the methanol OH proton to a Rh(iii)–OH ligand and Cp* methyl hydrogen to the methanol oxygen. Additionally, the observed trend in catalytic activity for complexes 1–4 was reproduced by DFT calculations. These complexes form a novel class of catalytic molecular motors with a tunable rate of operation that can be stalled in a given state. They provide a basis for elucidation of the effects of ligand design on the contributions of electronic, rotational and vibrational energies to each step in the reaction pathway at the atomic level, consideration of which will enhance the design principles for the next generation of molecular machines.

Positive ion electrospray mass spectra were obtained on a Bruker Daltonics Esquire 2000 ion trap mass spectrometer. All samples were prepared in methanol. Data were processed using Data- The x-ray photoelectron spectroscopy (XPS) data were collected at the Warwick Photoemission Facility, University of Warwick. The samples investigated in this study were attached to electrically-conductive carbon tape, mounted on to a sample bar and loaded in to a Kratos Axis Ultra DLD spectrometer which possesses a base pressure below 1 x 10 -10 mbar.
For XPS measurements, reaction mixtures were used directly without further purification. The solvents (d4-MeOD/D2O) were evaporated by freeze-drying. XPS measurements were performed in the main analysis chamber, with the sample being illuminated using a monochromated Al Kα x-ray source (hν = 1486.7 eV). The measurements were conducted at ambient temperature and at a take-off angle of 90° with respect to the surface parallel. The core level spectra were recorded using a pass energy of 20 eV (resolution approx. 0.4 eV), from an analysis area of 300 m x 700 m. The work function and binding energy scale of the spectrometer were calibrated using the Fermi edge and 3d5/2 peak recorded from a polycrystalline Ag sample prior to the commencement S4 of the experiments. In order to prevent surface charging, the surface was flooded with a beam of low energy electrons throughout the experiment and this necessitated recalibration of the binding energy scale. To achieve this, the main C-C/C-H component of the C 1s spectrum was referenced to 284.6 eV. The data were analysed in the CasaXPS package, using Shirley backgrounds and mixed Gaussian-Lorentzian (Voigt) lineshapes. For compositional analysis, the analyser transmission function was determined using clean metallic foils to determine the detection efficiency across the full binding energy range.  d4-MeOD/D2O (3:2) or in 1.56 mL d4-MeOD in the dark following our reported protocol. [3] The reaction was stirred for 12 h at ambient temperature. Then the reaction mixture was freeze-dried immediately, or the solvents were evaporated. The product was characterized by LCMS, HRMS and XPS.

Applied methods
The calculations were performed using Gaussian 16 package with use of the CAM-B3LYP functional and CEP-31g or QZVP basis set (see text). IEFPCM method was used to model the methanol solvent. The corresponding structures of the reactants and products were optimized and the frequency calculations were subsequently performed in order to show that the true minimum was achieved. The saddle (transition) points were calculated using the quadratic synchronous transit method (QST3 keyword of Gaussian) with calculations of the forces (CalcAll keyword of Gaussian). The subsequent frequency calculations were performed revealing one negative frequency mode in case of both models (with and without methanol molecule). Finally, the internal reaction coordinate (IRC) calculations were performed that additionally confirmed the transition state character of the saddle point found within the QST3 approach.

E-and Z-isomerism in complexes of monosubstituted R-Cp(Me)4 complexes
For the Cp*-substituted ligand systems under considerations, two conformations of the complex are possible, with the phenyl or biphenyl substituent located either on the side of N,N'-Rh plane with coordinated oxygen ligand, or on the opposite one, resulting in E and Z conformers for details. These conformers are illustrated in Fig. S13. They may be described as E and Z conformers with respect to the O-Rh-Cp*-R fragments. The isomers are shown in Fig. S13. Note that this isomerism occurs also for the square-planar species.

Modelling the reactivity of the catalyst precursor towards for forming hydroxide-species
In order to obtain insight into possible sources of this dependence, other than the electronic effect of the ligands and substituents we examined the reactivity of the catalyst precursors [Rh(N,N')(Cp*-R)Cl] + towards forming the hydroxido-species [Rh(N,N')(Cp*-R)(OH)] + . We modelled the process of the subsequent aquation of the precatalyst followed by the deprotonation of the coordinated water leading to formation of the hydroxido-species. The following reaction was therefore modelled: For the sake of simplicity, we excluded the methanol molecule from the model. As the previously mentioned calculation for the substituted Cp* ligands show that the E and Z conformers of all investigated stages of the proton transfer differ by less than 3 kJ/mol in energy, we performed the calculations for the aquation and deprotonation process only for the E conformations of the Cp*-Ph and Cp*-biph complexes. The results are collected in Table S2 and shown schematically in Fig. 13. Due to the general problem of estimation of the explicit solvent interaction with DFT mentioned above and discussed in ref. 4, the absolute energies shall be treated as the estimates.