Oxacarbene intermediate

A fifth unimolecular reaction involves electronic reorganization in [1] to afford a ring-expanded oxacarbene intermediate [9]. If photolysis is carried out in alcoholic solvents, [9] is efficiently trapped to afford acetal [10]. Yates has recently reviewed the photochemical ring-expansion reaction (9) and has found it to... 

More work on the nature of product formation from cyclobutanones, such as [79a], in inert solvents is required in order to better understand the fate of the presumed oxacarbene intermediate. At any rate, it is clear that ring expansion can occur in the absence of hydroxylic solvents, thereby eliminating a special role for protonic solvents. 

An interesting question which has not been clarified concerns whether formation of the oxacarbene intermediate is reversible, i.e., whether the reaction [31] [26a]... 

The postulation that the "biradical-like" transition state [135] (and not a freely rotating acyl alkyl biradical intermediate) is the precursor of the oxacarbene intermediate [136] is made primarily to accomodate the fact that the ring-expansion reaction is stereoselective. Transition state [135] could also decay by S-cleavage and/or decarbonylation [both stereospecific (23), although definitive evidence concerning this point is not available in the solution phase. Finally [135] could decompose back to the starting cyclobutanone, which would explain the observed lack of efficiency in the previously described photolyses. (See Section II.E for a further discussion of this point.)... 

Ring-expansion of the 1,3-dione (27a) at —70 °C gives the acetal (28) in low yield, probably via an oxacarbene intermediate.38 The main product, the keto-ester (29) (55%), is formed by the Norrish Type I reaction and is photolabile, forming the ester (30) as well as methyl cyclopentanecarboxylate. The ring-expansion is dependent upon the ring size of the substituent groups and does not occur when the dione (27b) is used. 

Switlak, K., He, D., and Yates, P, Intramolecular trapping of an oxacarbene intermediate in the photochemical ring expansion of a cyclopentanone,/. Chem. Soc., Perkin Trans. 1, 2579, 1992. Staab, H. A. and Ipaktschi, J., Photochemische reaktionen des benzocyclobuten-1, 2-dions, Chem. 

Scheme 10.14 rationalizes the divergent behavior of the two catalytic systems in these selective transformations of pent-l-yn-ols. The presence of phosphine ligands promotes the formation of ruthenium vinylidene species which are key intermediates in both reactions. The more electron-rich (p-MeOC6Fl4)3P phosphine favors the formation of a cyclic oxacarbene complex which leads to the lactone after attack of the N-hydroxysuccinimide anion on the carbenic carbon. In contrast, the more labile electron-poor (p-FC6H4)3P) phosphine is exchanged with the N-hydroxysuccinimide anion and makes possible the formation of an anionic ruthenium intermediate which liberates the cyclic enol ether after protonation. 

An attempt to affect the ratio of products from [47] by inclusion of ethyl iodide led to negative results. This experiment was run with the hope that intersystem crossing of some intermediate (e.g., S, biradical, or oxacarbene) might lead to a change in the product distribution and the reaction stereospecificity. 

These results can be interpreted as true "solvent effects".on the partition of some intermediate or, more specifically, as evidence for reversible formation of an oxacarbene, and its potential participation in decar-bonylation. In any case, some revealing information might be given by experiments in which the yields for all products and quantum yields for cyclobutanone disappearance are measured in an "inert" solvent as a function of added methanol. 

It is also possible to conceive of less likely paths, such as Sq being a precursor to a biradical intermediate or an oxacarbene. Path (a) is rejected as a general pathway since the triplet of cyclobutanone apparently cleaves to yield a triplet biradical capable of scrambling starting material, stereochemistry, and product stereochemistry. Path (b) is rejected because the emission spectrum and fluorescence "lifetime" of cyclobutanone... 

As a final suggestion for future research, cyclobutanones have also provided the organic photochemist with the opportunity of investigating the existence of unusual and reactive intermediates oxacarbenes, trimethylene biradicals, trimethylenemethane biradicals, acyl alkyl biradicals, and ketenes. Evidence for the intervention of oxacarbenes in the ring-expansion reaction is quite compelling however, their unusual behavior relative to "typical" carbenes (e.g., failure to form cyclopropane adducts with some olefinic substrates) makes them prime subjects for further study and characterization. Unlike oxacarbenes, the existence of acyl alkyl biradicals (e.g., [30]) is tenuous at best. Ideally,... 

The propensity of cyclopropylideneamines to give stable adducts with alcohols was also reflected in reactions in which the strained iminocyclopropanones were generated as nonisolable intermediates. Photolysis of 3-iminocyclobutanones 13 in alcohols resulted in decarbonylation giving 2,2,3,3-tetramethylcyclopropylideneamines 14 which were isolated as the stable alcohol adducts 15. This reaction was accompanied by the formation of an a-oxacarbene 16, which was transformed into the functionalized tetrahydrofurans 17. In an inert solvent, no photolytic decarbonylation was observed, which precluded possible isolation of the corresponding tetramethyl-cyclopropylideneamine 14. ... 

The photochemistry of carbonyl compounds still continues to be a major general area of interest, and physical methods, especially e.s.r., C1DNP, and C1DEP (electron polarization), continue to be widely applied for detection of radical-like transient intermediates. Quinkert and Jacobs have provided further evidence that ring expansion of cyclobutanones to oxacarbenes, and thence to tetrahydrofurans, occurs in a concerted fashion without the intermediacy of biradicals formed by Norrish Type I fission of a C—CO bond. Medary et al. report that Norrish Type I cleavage of 2-ethylcyclopentanone is non-stereospecific, and gives the cis- and /ra/z.y-hept-4-enals previous reports (Srinivasan and Cremer, 1965) that reactions of this type are stereospecific appear to require revaluation. 

As frequently noted before, the formation of the silicon-oxygen bond occurs here also with virtually complete retention of configuration based on studies with (+ )-l-naphthylphenylmethylbenzoylsilane. Similar photochemical formation of oxacarbenes have recently been reported by several workers 65-68). The role of base and acid in leading to these apparently unrelated pathways is not yet understood. Qualitative rate studies indicate that conversion of the mixed acetal to the alkoxysilane and acetal under comparable acid conditions is much slower than their direct photochemical formation so that the mixed acetal does not appear to be an intermediate in the reactions containing acid. 

Lee-Ruff and co-workers have continued their interest in the photochemistry of cyclobutanones. Specifically, they have profitably extended the chemistry of the oxacarbenes by trapping these intermediates with selected amines, to good effect Thus, the Canadian team photol) ed the ketones 4 and 5 in the presence of purines as well as water and alcohols to produce the corresponding nucleoside analogs, lactols, and acetals, respectively (Scheme 6 and Table 49.1). 

Best results were obtained with Rh(PR3)3Cl and [Rh(COD)Cl]2 catalysts in the presence of an excess of electron-poor triaryl phosphines to avoid undesirable dimerization/oligomerization processes. The proposed reaction mechanism involves the formation of the Rh-vinylidene complex followed by the intramolecular endo-dig cyclization. The protodemetallation of intermediate I seems to be the more plausible path, whereas the formation of the Rh-oxacarbene complex II was excluded because all attempts to generate lactones by using N-hydroxysuccinimide failed and the cycloisomerization product was the only product obtained (Scheme 10). 

---