Search

A recent advance in the synthesis of alkenylated arenes was the demonstration that the Pd(OAc)₂ catalyst precursor gives >95% selectivity toward styrene from ethylene and benzene under optimized conditions using excess Cu(II) carboxylate as the in situ oxidant [ Organometallics 2019, 38(19), 3532−3541]. To understand the mechanism underlying this catalysis, we applied density functional theory (DFT) calculations in combination with experimental studies. From DFT calculations, we determined the lowest-energy multimetallic Pd and Pd–Cu mixed metal species as possible catalyst precursors. From the various structures, we determined the cyclic heterotrinuclear complex PdCu₂(μ-OAc)₆ to be the global minimum in Gibbs free energy under conditions of excess Cu(II). For cyclic PdCu₂(μ-OAc)₆ and the parent [Pd(μ-OAc)₂]₃, we evaluated the barriers for benzene C–H activation through concerted metalation deprotonation (CMD). The PdCu₂(μ-OAc)₆ cyclic trimer leads to a CMD barrier of 33.5 kcal/mol, while the [Pd(μ-OAc)₂]₃ species leads to a larger CMD barrier at >35 kcal/mol. This decrease in the CMD barrier arises from the insertion of Cu(II) into the trimetallic species. Because cyclic PdCu₂(μ-OAc)₆ is likely the predominant species under experimental conditions (the Cu to Pd ratio is 480:1 at the start of catalysis) with a predicted CMD barrier within the range of the experimentally determined activation barrier, we propose that cyclic PdCu₂(μ-OAc)₆ is the Pd species responsible for catalysis and report a full reaction mechanism based on DFT calculations. For catalytic conversion of benzene and ethylene to styrene at 120 °C using Pd(OAc)₂ as the catalyst precursor and Cu(OPiv)2 (OPiv = pivalate) as the oxidant, an induction period of ∼1 h was observed, followed by catalysis with a turnover frequency of ∼2.3 × 10⁻³ s⁻¹. In situ1H NMR spectroscopy experiments indicate that during the induction period, Pd(OAc)₂ is likely converted to cyclic PdCu₂(η²-C₂H₄)₃(μ-OPiv)₆, which is consistent with the calculations and consistent with the proposal that the active catalyst is the ethylene-coordinated heterotrinuclear complex cyclic PdCu₂(η2-C₂H₄)₃(μ-OPiv)₆.

We show how to improve the yield of MeX from CH₄ activation catalysts from 12% to 90% through the use of “capping arene” ligands. Four (FP)Rh^(III)(Me)(TFA)₂ {FP = “capping arene” ligands, including 8,8′-(1,2-phenylene)diquinoline (6-FP), 8,8′-(1,2-naphthalene)diquinoline (6-^(NP)FP), 1,2-bis(N-7-azaindolyl)benzene (5-FP), and 1,2-bis(N-7-azaindolyl)naphthalene (5-^(NP)FP)} complexes. These complexes and (dpe)Rh^(III)(Me)(TFA)₂ (dpe = 1,2-di-2-pyridylethane) were synthesized and tested for their performance in reductive elimination of MeX (X = TFA or halide). The FP ligands were used with the goal of blocking a coordination site to destabilize the Rh^(III) complexes and facilitate MeX reductive elimination. On the basis of single-crystal X-ray diffraction studies, the 6-FP and 6-^(NP)FP ligated Rh complexes have Rh–arene distances shorter than those of the 5-FP and 5-^(NP)FP Rh complexes; thus, it is expected that the Rh–arene interactions are weaker for the 5-FP complexes than for the 6-FP complexes. Consistent with our hypothesis, the 5-FP and 5-^(NP)FP Rh^(III) complexes demonstrate improved performance (from 12% to ∼60% yield) in the reductive elimination of MeX. The reductive elimination of MeX from (FP)Rh^(III)(Me)(TFA)₂ can be further improved by the use of chemical oxidants. For example, the addition of 2 equiv of AgOTf leads to 87(2)% yield of MeTFA and can be achieved in CD₃CN at 90 °C using (5-FP)Rh(Me)(TFA)₂.

We report a combined experimental and computational study focused on the mechanism of oxidative conversion of benzene and ethylene to styrene using [(η²-C₂H₄)₂Rh(μ-OAc)]₂ as the catalyst precursor in the presence of Cu(OPiv)₂ (OPiv = pivalate). Using [(η²-C₂H₄)₂Rh(μ-OAc)]₂ as the catalyst precursor, ∼411 turnovers of styrene are observed after 1 h, giving an apparent turnover frequency of ∼0.11 s⁻¹ (calculated assuming the binuclear structure is maintained in the active catalyst). We identify the catalyst resting state to be [(η²-C₂H₄)₂Rh^I(μ-OPiv)₂]₂(μ-Cu), which is a heterotrinuclear molecular complex in which a central Cu^(II) atom bridges two Rh moieties. At high Rh concentration in the presence of Cu(OPiv)₂ and pivalic acid (HOPiv), the trinuclear complex [(η²-C₂H₄)₂Rh^I(μ-OPiv)₂]₂(μ-Cu) converts to the binuclear Rh(II) complex [(HOPiv)Rh^(II)(μ-OPiv)₂]₂, which has been identified by ¹H NMR spectroscopy and single crystal X-ray diffraction. The binuclear Rh(II) [(HOPiv)Rh^(II)(μ-OPiv)₂]₂ is not a catalyst for styrene production, but under catalytic conditions [(HOPiv)Rh^(II)(μ-OPiv)₂]₂ can be partially converted to the active catalyst, the Rh–Cu–Rh complex [(η²-C₂H₄)₂Rh^I(μ-OPiv)₂]₂(μ-Cu), following an induction period of ∼6 h. Using quantum chemical calculations, we sampled possible mononuclear and binuclear Rh species, finding that the binuclear Rh(II) [(HOPiv)Rh^(II)(μ-OPiv)₂]₂ paddle-wheel is a low energy global minimum, which is consistent with experimental observations that [(HOPiv)Rh^(II)(μ-OPiv)₂]₂ is not a catalyst for styrene formation. Further, we investigated the mechanism of styrene production starting from [(η²-C₂H₄)₂Rh^I(μ-OAc)₂]₂(μ-Cu), [(η²-C₂H₄)₂Rh(μ-OAc)]₂, and (η²-C₂H₄)₂Rh(κ²-OAc). For all reaction pathways studied, the predicted activation barriers for styrene formation from [(η²-C₂H₄)₂Rh(μ-OAc)]₂ and (η²-C₂H₄)₂Rh(κ²-OAc) are too high compared to experimental kinetics. In contrast, the overall activation barrier for styrene formation predicted by DFT from the Rh–Cu–Rh complex [(η²-C₂H₄)₂Rh^I(μ-OPiv)₂]₂(μ-Cu) is in agreement with experimentally determined rates of catalysis. Based on these results, we conclude that incorporation of Cu(II) into the active Rh–Cu–Rh catalyst reduces the activation barrier for benzene C–H activation, O–H reductive elimination, and ethylene insertion into the Rh–Ph bond.