Catalysis of Diels-Alder Reactions by Low Oxidation State Transition-Metal Lewis Acids: Fact and Fiction

Catalysis of Diels-Alder reactions between the dienes cyclopentadiene, butadiene, isoprene, and piperylene and the enones acrolein, methyl vinyl ketone, and methyl acrylate is induced by 0.1-2.5 mol % of mer-(cis-Me₃P)(trans-NO)-(CO)₃W(μ-F)SbF₅(1), (Cy₂PCH₂CH₂PPh₂)(CO)₂(NO)W(μ-F)SbF₅ (2), Cp(CO)₂FeL⁺X⁻ (L = THF, X⁻ = BF₄⁻, 3a; X⁻ = SbF⁻₆, 3b; L = η¹-acrolein, X⁻ = PF₆⁻, 3d), or Cp(CO)₂L'ML⁺PF₆⁻ (L' = CO, L = acrolein, M = Mo, 4a; L = PPh₃, L = THF, M = Mo, 4b; L' = CO, L = THF, M = W, 4c). Enhancement of rates and regio-and stereoselectivity is observed compared to the thermal reactions; the order of apparent catalytic activity is 1 > 2 ≈ 3a > 4a, 4c. The order of Lewis acidity is 1 > 2 > 4a > 3a, casting doubt on the role of Cp(CO)₂Fe⁺ in catalysis. The potential impurity Ag⁺BF₄⁻ is similarly reactive, although not in lower concentrations. Use of 2, 6-di-tert-butylpyridine (5) and l-(n-butyl)-2, 2, 6, 6-tetramethylpiperidine (6) as hindered bases to trap Ag⁺ and H⁺ in the presence of transition-metal Lewis acids is described. Substoichiometric use of 5 demonstrates that Ag⁺BF₄⁻ is not the real catalyst and that the true activity of 3a is low. Use of 5 with stronger acids, namely the acrolein adduct of 1 (la) and 4a, or of the stronger base 6 with 3a leads to catalyst destruction, via a pathway proposed to involve deprotonation of coordinated methylene chloride. The reactivity of other potential impurities (HBF₄.Et₂O, BF₃.Et₂O, Ph₃C⁺PF₆⁻, and NO⁺SbF₆⁻) is briefly examined, as is that of analogues of 3a that have different counterions. Kinetic analysis of stoichiometric reactions of metal-acrolein adducts with isoprene shows that the relative rates of cycloaddition for la, the acrolein adduct of 2, 4a, and 3d are 68:20:8:1 and that the rate-determining step in the catalytic reactions is the rate of aldehyde turnover. The calculated rate constants are used to predict catalytic yields and demonstrate that 1 and 2 can be the real catalysts. For 3a and possibly 4a as well, the observed catalytic activity is significantly greater than expected on the basis of the stoichiometrically determined rate constants, so the real catalysis in these cases apparently is due to the presence of much more reactive materials present as impurities.