Abstract
The nature of the Ru bonding environment in a set of Grubbs catalysts has been studied by means of density functional theory (DFT). On the one hand, for a set of 20 [Ru(═CH2)(NHC)(PCy3)][Cl]2 second-generation adducts, the results show that calculated 31P NMR shieldings exhibit a good correlation with the calculated R(Ru–P) bond lengths, which are in turn strongly correlated with the calculated PCy3 ligand dissociation energies. Bond energy decomposition analysis (EDA) also indicates that there is a strong correlation between the σ and π orbital interaction energies for the Ru–PCy3 bond, while no correlation was found for the case of the bond between the Ru moiety and the N-heterocyclic carbene (NHC) ligands, Ru–NHC. Furthermore, π orbital interaction energies of the Ru–PCy3 bond were found to be strongly correlated with the calculated PCy3 ligand dissociation energies, as well as with the R(Ru–P) bond lengths, confirming the significance of the π back-donation component from Ru to PCy3 in determining the lability of the PCy3 ligand in the studied adducts. On the other, for a set of 17 [Ru(═CHR)(NHC)x(PCy3)2–x][Cl]2 first (x = 0)- and second-generation (x = 1) complexes, DFT results show that changes occurring in the 13C NMR shielding of the Ru═ylidene bond are mainly due to σ(Ru═C) → π*(Ru═C) molecular orbital (MO) transitions. Good correlations are observed between σ(13C) and the energy gaps of the MOs involved in such transitions, between the binding energies of the ylidene moiety and the rest of the Ru fragment, as well as with the R(Ru═C) bond lengths. Finally, our novel preliminary results suggest that, once the metallacycle intermediate is formed by reaction with ethylene, 13C′(β) NMR shielding retains the NMR information from σ(13C) in the 16e species, in contrast to what happens with the 13C(α) NMR shielding.
Original language | English (US) |
---|---|
Journal | Organometallics |
DOIs | |
State | Published - Nov 12 2020 |
Bibliographical note
KAUST Repository Item: Exported on 2020-11-17Acknowledgements: A.P. is a Serra Hunter Professor and thanks ICREA Academia ́2019 and the Spanish MINECO for project ref PGC2018-097722-B-I00. The authors acknowledge the Universidad de Las Palmas de Gran Canaria (ULPGC) and the King Abdullah University of Science and Technology (KAUST) for support. Gratitude is also due to the KAUST Supercomputing Laboratory using the supercomputer Shaheen II for providing
the computational resources.