Spiro transition structure

This controversy concerning the use of MP2 calculations for epoxidation reactions was rather short-lived since more efficient density functional calculations (DFT) came into general use and generally produced symmetrical spiro transition structures. Consequently, the use of MP2 theory for 0—0 bond cleavage reactions has been largely discontinued. Most have assumed that the question of symmetrical versus asymmetrical approach of the peracid had been resolved. Recall that this same problem with MP2 calculations existed for the early calculations for dioxirane epoxidation (see Section V.D). 

Both experimental and theoretical studies of the peracid epoxidation of ethene and alkyl-substituted olefins have suggested that a spiro transition structure is favored over a planar one. For example, DFT calculations indicated that the transition structure for the reaction of peroxyformic acid and ethene has the geometry shown in Figure 9.45. In this case the two C-0 distances are identical, indicating synchronous formation of the two C-O bonds. If the olefin is substituted with a methyl, methoxy, vinyl, or cyano group, however, the transition structure can become asynchronous, meaning that one C-0 bond has been formed to a greater extent than the other in the transition structure. ... 

The similarity of olefin epoxidation by TM peroxo and hydroperoxo complexes with epoxidation by dioxirane derivatives R2CO2 and percar-boxylic acids RCO(OOH) was confirmed by computational studies [73-79]. This similarity holds in particular for the spiro-type transition structure. 

The very large value of the pre-exponential factor indicates a transition complex which is very loose compared with the highly strained spiro-pentane structure. Inspection of the models of the reactant and product makes it clear that considerable distortion of the reactant must occur on going to the transition complex. A minor reaction path results in the formation of allene and ethylene. These products are primary. 

For the epoxidation of propylene with HO—ONO, both the QCISD and CISD calculations result in a Markovnikov-type transition structure, where the electrophile is slightly skewed toward the least substituted carbon, with a small difference in the bond lengths between the spiro-oxygen and the double-bond carbons (0.106 and 0.043 A, respectively Figure 8). The B3LYP calculations also lead to an unsymmetrical Markovnikov-type transition structure (b). However, the MP2/6-31G geometry optimization results in an anti-Markovnikov-type structure (c) (Figure 8). The CCSD(T)/6-31G and BD(T)/6-31G barriers for ethylene epoxidation with HO—ONO calculated with the QCISD/6-31G ... 

The transition structures for the epoxidation of ethylene and propylene with peroxyformic acid and of ethylene with dioxirane and dimethyldioxirane calculated at the B3LYP, QCISD and CCSD levels are symmetrical with a spiro orientation of the electrophilic oxygen, whereas the MP2 calculations favor unsymmetrical transition structures. The geometries of the transition structures calculated using the B3LYP functional are close to those found at QCISD, CCSD, CCSD(T) levels as well as those found at the CASSCF(10,9) and CASSCF(10,10) levels for the transition structure of the epoxidation of ethylene. 

The epoxidations of propylene and isobutylene with peroxyformic acid proceed in a concerted way via slightly unsymmetrical Markovnikov-type transition stmctnres where the differences in the bond distances between the donble-bond carbons and the spiro oxygen are only 0.021 and 0.044 A at the QCISD/6-31G level. In contrast, the more polarizable natnre of the carbon-carbon double bond of o ,/ -unsaturated systems results in a highly nnsymmetrical transition structure for the epoxidation of 1,3-butadiene with an order-of-magnitnde difference in the carbon-oxygen bond distances of 0.305 A at the QCISD/6-31G level. A highly unsymmetrical transition structure has been also found for the epoxidation of acrylonitrile. 

Alkenes strained by twist or r-bond torsion, such as E-cyclooctene, exhibit much lower barriers due to relief of strain in the TS for the oxygen transfer step. While the epoxidation of symmetrically substituted alkenes normally involve a symmetrical approach to the TT-bond, the TSs for epoxidation of E-cyclooctene and E-l-methylcyclooctene exhibit highly asymmetric transition structures. The AAE = 3.3 kcalmol" for E- versus Z-cyclooctene is clearly a reflection of the relative SE of these two medium ring alkenes (16.4 vs 4.2 kcalmol ) ". The classical activation barrier (AE ) for the highly strained bicyclo[3.3.1]non-l-ene is also quite low (Table 10, Figure 26). In these twist-strain alkenes, the approach of the peracid deviates markedly from the idealized spiro approach suggesting fliat this part of the potential energy surface is quite soft. 

FIGURE 25. Transition structures for the epoxidation of cyclopropene, cyclobutene and cyclopen-tene with peroxyformic acid (PFA), optimized at the B3LYP/6-31+G(d,p) level of theory. The classical activation barriers are given at B3LYP/6-311+G(3df,2p)//B3LYP/6-31+G(d,p). Dihedral angles (deviation from an ideal spiro approach (c n, is 90°) of the HO group in PFA onto the C=C bond of the alkene... 

The transition state for erythro epoxidation, syn-(2A,3 A,45 )-TS, is only 0.9 kcal mol 1 higher in energy and possesses a nonplanar peracid approaching the C=C bond in a manner intermediate between spiro and planar. The relative energy and nonplanarity of this syn transition structure is highly sensitive to the basis set applied. It was shown that... 

A concerted, spiro-structured, oxenoid-type transition state has been proposed for C-H oxidation by dioxiranes (Scheme 5). This mechanism is based mainly on the stereoselective retention of configuration at the oxidized C-H bond [20-22], but also kinetic studies [29], kinetic isotopic effects [24], and high-level computational work support the spiro-configured transition structure [30-32], The originally proposed oxygen-rebound mechanism [24, 33] was recently revived in the form of so-called molecule-induced homolysis [34, 35] however, such a radical-type process has been experimentally [36] and theoretically [30] rigorously discounted. 

---