Feist’s ester

Analogous ligand exchange reaction of cis- and Pww-Feist s esters with the dirhodium complex [/i-ClRh(ethylene)2]2 in pentane gave the corresponding Feist s esters rf-complexes [/(-ClRhL, (L = cF, tF) (equation 310). Further reaction of the latter complex [p-ClRhL2]2 (L = tF) with cyclopentadienylthallium (CpTl) in CH2C12 afforded the monorhodium complex CpRh(tF)2, and reaction with a mixture of both CpTl and dirhodium... 

Likewise, mononuclear complexes of rhodium and platinum containing only one meth-ylenecyclopropane ligand are prepared by ligand exchange reactions of the Feist s esters with (acac)Rh(CO)2 and trans-C 2(pyr)Pt(et hylene), giving complexes (acac)(CO)Rh(tF) and franj-Cl2(pyr)PtL (L = cF, tF), respectively (equation 311). 

CpCo(mcp)2, which in turn can be further transformed to the mono-complex CpCo(PPh3)(mcp) by exchange of one methylenecyclopropane ligand with PPhj (equation 312). Although both complexes are isolable crystals, they are thermally less stable than the analogous Feist s ester complexes. CpCo(mcp)2 readily undergoes thermal isomerization at 110 °C, to give cyclopropyl-substituted -butadiene complexes (see below). 

X-ray structural analysis of 2,2-dimethyl-3-phenyl-l-methylenecyclopropane tungsten pentacarbonyl reveals an octahedral complex with characteristic W—C bond distance of 238 pm. The typical bond distances within the organic ligand are 138 (complexed C=C), 148 (proximal C—C), 154 (distal C—C) pm, compared e.g. with 140, 148 and 154 pm, respectively, for the Feist s ester iron complex analogue (see above). 

Unlike the alkyl and aryl-substituted methylenecyclopropanes discussed above, both cis- and franj-Feist s esters undergo chloropalladation with proximal 1,2-ring opening, to give isomeric n3-[3-chloro-l,2-bis(methoxycarbonyl)but-3-enyl]palladium complexes (equation 331)397. Formation of the but-3-enyl complexes is rationalized by sequential... 

Proximal ring opening also occurs when the Feist s esters undergo hydroplatination400. However, the reaction apparently stops at the early but-3-enyl stage (isolated as the... 

The ozonation of Feist s ester yields three products 2,3-dicarbomethoxytetrahydrofuran-4-one (III), 3,4-dicarbome-thoxy-5-hydroxy-2-oxa-2,3-dihydropyran (IV), and 4,5-di-carbomethoxy-l-oxaspiro[2.2 pentane (V). Structures were assigned primarily through spectroscopic evidence (mass, NMR, and infrared spectra). The last of these compounds is probably not a direct ozonation product (stoichiometry and trapping studies). A mechanism is proposed in which the initial step is formation of a primary ozonide. From that stage, the breakdown is abnormal. 

T,he structure of Feists acid, la, was determined with certainty some 60 years after its preparation (1-6). More recently, the proposal that the ozonation product of Feist s ester (lb) had structure II was shown to be incorrect (7). Our initial report not only showed that no II was present in any ozonation of Feists ester but offered chemical and spectroscopic proof for III as a major ozonation product. 

Product Analysis. Quantitative analysis for each of the products of the ozonation of Feist s ester was carried out by an NMR method (Varian HA60). The solvent for the ozonation was evaporated in vacuo, a suitable internal standard (usually f erf-butyl benzoate), was added and the spectra were determined and integrated in chloroform-d. All analyses were reproducible to 10%. 

The epoxide is not a primary product both the stoichiometry and the tetracyanoethylene experiments indicate that it is formed from attack of an intermediate on unreacted Feist s ester. For example, less than 1 mole of ozone per mole of Feists ester is required for all ozonations—with the less reactive solvents, such as methylene chloride and Freon 11, the deficiency in ozone required is approximately equal to the amount of epoxide formed. Addition of tetracyanoethylene before ozonation results in an ozonation of normal stoichiometry with no epoxide being formed. 

If the reaction is stopped before completion, the relative amounts of IV formed are greatly decreased. This suggests to us that the initial ozonation product has several modes of decomposition available, and that reaction with Feist s ester (or tetracyanoethylene) is more rapid than unimolecular decomposition. 

We were unable to detect an intermediate [e.g., Criegee s primary ozonide (9)] by low temperature NMR (to — 130°C) yet the above-mentioned results and the fact that IV does not react with Feist s ester strongly suggest that a highly reactive intermediate does exist. 

Story et al. have suggested that VIII might be the important intermediate in the ozonation of Feist s ester (11). 

If this is the intermediate, other methylenecyclopropanes should give ozonation products similar to those found in the ozonation of Feist s ester. Since they do not and because our evidence indicates that the epoxide is not a primary product, the mechanism of Story et al. does not seem to apply to this reaction. 

In 1932 Kon and Naji reported that pyrolysis of trans-2,3-ethoxycarbonyl-methylenecyclopropane ( Feist s ester ) resulted in an isomerization to give compounds later identified as the stereoisomeric 2-carbethoxy-l-(carbethoxy-methylene)cyclopropanes (Figure 36). Subsequent studies by Ullman " using the optically resolved reactant showed that the products were not racemic and that the planar trimethylenemethane biradical could not, therefore, be the sole intermediate in the... 

---