Mechanistic Insight for Improved Catalytic Conversion of Fatty Acids to Linear α-Olefins : A Density Functional Theory Study

Abstract

One of the greatest challenges in our time is the replacement of fossil fuels with more sustainable alternatives for energy and chemical production. Arguably one of the most important classes of petrochemical intermediates is linear α-olefins (LAOs), which are used in the production of polymers, surfactants and lubricants, and are mainly obtained through oligomerization of ethylene. Alternatively, they can be produced by ethenolysis or deoxygenation of the renewable feedstock triglycerides and their unsaturated and saturated fatty acids respectively. While there exist homogeneous, heterogenous and biocatalysts for the deoxygenation reactions, LAOs can most selectively be produced through the homogenous transition-metal catalyzed reaction decarbonylative dehydration. However, the reaction typically requires high temperature, excess of ligands, fatty acid activation and distillation or a toxic solvent to achieve activity and selectively, and the development of catalysts that can compete with the fossil-based alternatives has been stifled by the lack of mechanistic insight. Only recently has the first computational investigations of the reaction arrived, and together with the first well-defined precatalyst for the transformation, Pd(cinnamyl)Cl(DPEphos) (A1), there is now potential for rational catalyst design. Within this thesis we have investigated, by density functional theory (DFT) calculations, how different factors of the reaction in general, and of the precatalyst A1 in particular, affect the reaction mechanism. The reaction mechanism for a Rh catalyst is derived and compared to that of Pd, and it is found that while the ratedetermining step is olefin-formation for both, the rate-determining intermediate for Rh is the starting complex, (PPh3)2RhI(CO)Cl. The main reason for the overall higher barriers for Rh, and thus the observed higher activity of Pd, stems from the greater stability of this RhI-CO bond and the preference Rh has for a higher coordination number which leads to increased steric hindrance. For Pd a low coordinated metal center is favored for the rate-determining olefin formation, but the classic P-O-P ligand DPEphos facilitates good α-olefin production and catalyst stability by hemilabile coordination. The change from bidentate κ2 to monodentate κ1- coordination prior to decarbonylation creates a coordination site for CO deinsertion, and recoordination of the phosphine arm promotes CO displacement and dissociation which is key for catalyst activity. The escaping CO is modelled using a reduced pressure in the calculation of the thermochemical corrections and the effect of this correction is that the reaction is found to be catalytic as in accordance with the experiments. The main benefit of the precatalyst A1 beyond this hemilabile behavior of DPEphos, which ensures that the ligand does not dissociate from the complex and dismisses the requirement of ligand excess, is the faster initiation compared to in situ systems. A major drawback to the reaction is the fatty acid activation by a sacrificial anhydride, which forms a mixed anhydride system, and is required to cleave the strong C(=O)-O bond. However, the resulting carboxylate from the following oxidative addition is important for the reaction because it readily accepts the hydrogen transferred from the alkyl and thus prevents isomerization. Polar aprotic solvents are found to be beneficial for the reaction, even though a cationic pathway is not preferred, likely due to the destabilization of rate-determining intermediates. In conclusion the mechanism of decarbonylative dehydration has been investigated and key parameters for catalysis has been identified. Improved catalysts may be developed in the future by exploring and designing new asymmetric hemilabile ligands and faster initiating precatalysts, perhaps even automatically.