A "flexible cluster" model approach to Ziegler− Natta catalysts for the production of isotactic polypropylene, allowing the use of realistically sized MgCl 2 monolayer clusters (up to 38 MgCl 2 units) without any constraints, was employed to investigate the formation of adducts between the MgCl 2 support and three industrially relevant internal donor classes, namely phthalates, succinates, and 1,3-dimethoxypropanes. The calculated adsorption modes and thermochemical data for adducts of single-donor molecules confirmed earlier literature trends only in part. Results for adducts with multiple donor molecules, in turn, did not confirm the indications of periodic models about steric repulsion between neighboring adsorbates hampering high degrees of surface coverage; as a matter of fact, such repulsions seem to be largely traceable to unnecessary constraints inherent in periodic calculations.

In this paper we introduce a DFT‐D ‘flexible‐cluster’ approach to the computational modeling of MgCl2‐supported Ziegler‐Natta catalysts. The adsorption of TiCl4 and probe donor molecules on plain and defective lateral terminations of monolayered MgCl2 clusters was studied allowing for full structure relaxation by means of modern DFT methods including dispersion corrections. The approach offers a good compromise between speed and accuracy, and has great potential for realistic simulations of the catalytic species in industrially relevant formulations. In particular, compared with periodic DFT‐D descriptions, it is better able to capture the features of real nano‐sized and highly disordered primary catalyst particles with complex adsorbate compositions.

No abstract