Insertion and carbonylation reactions of styrene promoted by [HRh(CO)_x‐ (PMe₃)_3−x] (x = 1, 2) compounds: A theoretical investigation

Abstract

ABSTRACT:In this work we have applied Quantum Mechanical calculations to investigate the first two elementary steps (olefin insertion and carbonylation) in the hydroformylation of styrene, using the model catalysts of the type [HRh(CO) x (PMe 3 ) 3Ϫx ] (x ϭ 1, 2), which are supposed to be the catalytic species that will be present depending on the CO pressure. The migratory styrene insertion reaction and CO insertion reaction into the metal-alkyl bond were investigated at the MP4(SDQ) level using the BP86 optimized geometries. Additionally the Spin Component Scale procedure on the MP2 and MP3 energies (SCS-MP2 and SCS-MP3) was also applied. It is shown that, at normal hydroformylation conditions and normal CO and H 2 pressure, the active catalytic species is preferentially formed in trans arrangement, trans-[HRh-(CO)(PMe 3 ) 2 ], 2b. Because of the greater steric hindrance around the rhodium atom, the electronic effects of the phosphine do not contribute significantly to the stability of the catalyst. The MPn calculations overestimate the stability of the -complexes, in comparison with the BP86 value, as much as 25 kcal/mol, with large fluctuations along the perturbation series. The activation energies predicted by the BP86 method, in comparison with the MPn results, are underestimated about 5 kcal/mol. The reaction and the coordination energies are very sensitive to the theoretical level employed. Our results indicate that the competitive trapping of the styrene by different catalytic species, one leading to the branched and the other leading to the linear species, as was experimentally proposed to explain the selectivity, only seems to hold if we consider the subsequent CO insertion step. This assumption based only on the olefin insertion reaction, as the selectivity determining step, does not apply since the branched product will always preferentially be formed. All methods employed here predict the selectivity in good agreement with the experimental selectivity of 70-95% in favor of the branched product.

Propene Hydroformylation Reaction Catalyzed by HRh(CO)(BISBI): A Thermodynamic and Kinetic Analysis of the Full Catalytic Cycle

“…The formation of linear and branched Rh‐acyl products, ( 6L ) and ( 6B ), are both favorable processes with branched species around 1.3 kcal mol –1 more stable than linear intermediate. Similar behavior was also observed in other computational studies involving the hydroformylation reaction catalyzed by Pt‐Sn [ 55,56 ] and Rh [ 39,57 ] catalysts. The overall analysis of these results shows that the carbonylation step is thermodynamically controlled.…”

“…According to this proposal, after the in situ formation of the active catalytic species, the catalytic cycle is composed mainly by five elementary steps: ( A ) coordination of the olefin to catalyst, ( B ) insertion of the olefin into Rh‐H bond, ( C ) carbonylation of Rh‐Alkyl intermediate generated in step B, ( D ) oxidative addition of H 2 into Rh‐Acyl complex formed in C and ( E ) reductive elimination followed by the formation of products and regeneration of the catalyst. This mechanism involving different Rh catalysts has been extensively investigated in several computational studies, [ 25,39,50–53 ] which have provided quite a lot of detailed information about the structure of the intermediates and transition states of each elementary step. Thus, in this work, we will not elongate much about the discussion of the structural features of each species located along the catalytic cycle.…”

“…Our experience involving theoretical studies of homogeneous catalytic process promoted by transition metal catalysts has shown that DFT methods provide energetic results deviating 4–6 kcal mol –1 from ab initio methods. [ 39,40,55,75 ] In the absence of accurate experimental data, the way to evaluate the accuracy of DFT methods is usually to make a comparison with ab initio correlated methods. In this context, the emergence of new correlated methods, for instance, the newly developed version of the coupled‐cluster method with explicit correlation [CCSD(T)‐F12] [ 76 ] and local‐pair natural orbital coupled‐cluster [LPNO‐CCSD(T)], [ 77 ] together with the increased processing capacity of the computers and more efficient implementation schemes have allowed the accurate determination of kinetic and thermodynamic properties present in homogeneous catalytic processes.…”

“…We have showed in previous studies that this combination of basis set with B3LYP functional provides reliable energies and structural parameters for rhodium‐phosphine complexes. [ 38–40 ] All structures were characterized as true minima or first‐order transition states by analytical determination of harmonic frequencies. Transition states were obtained using the Synchronous Transit‐Guided Quasi‐Newton (STQN) method developed by Schlegel and co‐workers.…”

DFT study of the ligand effects on the regioselectivity of the insertion reaction of olefins in the complexes [HRh(CO)2(PR3)(L)] (R = H, F, Et, Ph, OEt, OPh, and L = propene, styrene)

“…The use of these “real” phosphines imposes restriction on the level of theory to be used in the calculation and, sometimes, low accurate methods are used. Despite of this, computational studies have been carried out aiming to understand the origin of the regioselectivity in the hydroformylation reaction9, 17–21. Most of these studies employ unmodified catalysts (metallic carbonyls) or phosphine‐modified catalysts, using small models such as phosphine, PH 3 , or trimethylphosphine, P(Me) 3 .…”