Effect of Internal Donors on Raman and IR Spectroscopic Fingerprints of MgCl/TiCl Nanoclusters Determined by Machine Learning and DFT.

To go deep into the origin of MgCl2 supported Ziegler-Natta catalysis we need to fully understand the structure and properties of precatalytic nanoclusters MgCl2/TiCl4 in presence of Lewis bases as internal donors (ID). In this work MgCl2/TiCl4 nanoplatelets derived by machine learning and DFT calculations have been used to model the interaction with ethyl-benzoate EB as ID, with available exposed sites of binary TixCly/MgCl2 systems. The influence of vicinal Ti2Cl8 and coadsorbed TiCl4 on energetic, structural and spectroscopic behaviour of EB has been considered. The adsorption of homogeneous-like TiCl4EB and TiCl4(EB)2 at the various surface sites have been also simulated. B3LYP-D2 and M06 functionals combined with TZVP quality basis set have been adopted for calculations. The adducts have been characterized by computing IR and Raman spectra that have been found to provide specific fingerprints useful to identify surface species; IR spectra have been successfully compared to available experimental data.