Azo-extrusion

Open-chain and cyclic compounds containing azo groups (-N2 —), such as azoalkanes, azoarenes, pyrazolines, triazolines, etc. may also eliminate N2, but these reactions are called azo-extrusions (IUPAC, 1989 a). The terms denitrogenation and nitrogen extrusion, both used by Adam et al. (1992, 1993) and by Adam and Sengelbach (1993) should not be used. They are superfluous and ambiguous. 

Dediazoniations of azides and azo extrusions are discussed in this book only occasionally. 

Arylazo-de-sulfonation 319 Arylazophenylsulfones, azo-extrusion 211 l-Aryl-2-benzenesulfonhydrazides, axo-extru-sion 211 Aryl cations... 

Extrusion of N2 from Pyrazolines. Pyrazoles, and Triazolines Azo-extrusion... 

There is interest in cycloaddition of diazoalkanes to cycloalkenes for various reasons. Small rings (cyclopropene and cyclobutene) react easily and often in good yield as a result of their angle-strained double bond. Their products with diazomethane, 2,3-diazabicyclo[3.1.0]hex-2-ene (6.56) and 2,3-diazabicyclo[3.2.0]hept-2-ene (6.57) and their substituted derivatives are interesting for synthetic purposes, e. g., by azo-extrusion leading to ring contraction. 

The reaction system (6-37) includes the thermal azo-extrusion of a cyclic azo compound to a cyclopropane derivative and the direct formation of cyclopropanes, catalyzed by metal complexes. Synthetic routes to cyclopropane derivatives became an important subject in the last two decades, and one frequently used method is the 1,3-dipolar cycloaddition of a diazoalkane to an alkene followed by thermal or photolytic azo-extrusion of the 4,5-dihydro-3//-pyrazole formed to the cyclopropane derivative (6-37 A). This route can be followed in many cases without isolation, or even without direct observation, of the 4,5-dihydro-3//-pyrazole. Therefore, it is formally very similar to cyclopropane formation from alkenes with diazoalkanes, in which a carbene is first formed by azo-extrusion of the diazoalkane (see Sect. 8.3). As shown in pathway (6-37 B), this step can be catalyzed by copper, palladium, or rhodium complexes (see Sects. 8.2, 8.7, and 8.8). There are cases where it is not clearly known whether route A or B is followed. Scheme 6-37 also includes... 

For examples of cyclopropanation of alkenes by diazoalkanes that are catalyzed by metal complexes [Mo(CO)6l and [Mo(AcO)4], but that follow the dipolar cycloaddition—azo-extrusion pathway see Doyle et al. (1982 a). 

An instructive example for pathway 6-37 A was found by Franck-Neumann and Miesch (1984). Reaction (6-38) demonstrates that the cycloaddition with 2-diazopropane takes place, as expected, at the double bond of methyl 5-methyl-hexa-2,4-dienoate (6.87) substituted by the (electron-withdrawing) ester group. The intermediate dihydropyrazole can be isolated. It is interesting, however, that in the subsequent azo-extrusion leading to the substituted cyclopropane carboxylate 6.88 (trivial name crysanthemic ester), the cis/trans ratios of the thermal, of the photolytic (high pressure Hg lamp), and of the benzophenone-sensitized photolytic methods are very different, in spite of the yield being the same (90%). 

In general, photolytic azo-extrusion from 4,5-dihydro-3/f-pyrazoles is superior to thermolysis. Photolysis was introduced as a method for synthesis by Jeger s group (Kocsis et al., 1960). The configurations of the products in the thermolysis, in the direct photolysis, and in the sensitized photolysis are often quite different. Cyclopropanation for the synthesis of alkyl cyclopropanes via dihydropyrazoles is preferred to the direct route via carbenes, because, in the latter, the C - H insertion of the carbene into the alkyl group is faster than the cyclopropanation. The dihydro-pyrazole pathway was used in particular for the formation of highly strained bicyclo[1.1.0]butanes (Franck-Neumann, 1967 Komendantov and Bekmukhametov, 1971 6.89) and bicyclo[2.1.0]pentanes (Vogelbacher et al., 1984 6.90). 

The nature of the cyclopropane-forming reaction is still controversial in the sense of a differentiation between a two-step cleavage of the two CN bonds and a concerted formation of a biradical in the azo-extrusion part of the reaction (for a review of the older literature, see Mackenzie, 1975, p. 354, and Engel, 1980, p. 118). This reaction does not strictly belong to the scope of this book. Therefore, we will not discuss it further except to refer to an investigation of Reedich and Sheridan (1988) who reported that, in contrast to earlier results (see, e. g., Cichra et al., 1980, and other references there), both thermal and photochemical processes take place by stepwise cleavage in a pair of isomeric dihydro-pyrazoles. 

Cyclopropanations by azo-extrusion of dihydro-pyrazoles may be accompanied by rearrangement of substituents. For instance, Hamaguchi and Nagai (1989) showed that, in 4-(arylseleno)- and 4-(arylthio)-4,5-dihydro-pyrazoles, these heteroaryl groups migrate into the 5-position with formation of cyclopropanes or ring opening to ethene derivatives. 

The dihydroarene-diazopropane adduct 6.98 is interesting because of very different product ratios in thermal and photolytic azo-extrusion, as shown in Table 6-4. 

Very active interest in a new addition reaction of aliphatic diazo compounds started in 1991 when WudPs group reported that diphenyldiazomethane forms diphenylmethanofullerene with buckminsterfullerene (C o Suzuki et al., 1991). Although this investigation showed that the reaction proceeds via the formation of a dihydro-pyrazole, i.e., in the mode of a 1,3-dipolar cycloaddition followed by an azo-extrusion, we shall discuss the syntheses of methanofullerenes in its entirety in the chapter on carbenes (Sect. 8.4) because Diederich s recent work (see review of Diederich et al., 1994b) shows that the methano bridge can also be obtained from a carbene. The question whether the dihydro-pyrazoles are intermediates or side-equilibrium products (see earlier in this section) is also open for the reaction of with diazoalkanes. 

For the synthesis of the sequiterpene (— )-cyclocopacamphene (6.112) an elegant application of an intramolecular cycloaddition, forming the annellated pyrazole derivative 6.111 followed by a photolytic azo-extrusion to the cyclopropane, was described by Piers et al. (1971). 

The synthesis of optically active cyclopropanes via formation of dihydro-pyrazoles by 1,3-cycloaddition and azo-extrusion has been studied since the late 1950 s. Modest success (10% ee) was achieved by cycloaddition of diazoalkanes to acrylic acid, esterified with ( —)-menthol, as studied by Walborsky s group (Impastato et al., 1959). Today, the use of chiral metal complexes as catalysts for the synthesis of chiral... 

The most important carbene reaction - mechanistically and for synthetic applications - is the cycloaddition to alkenes, i. e., the formation of cyclopropanes. This reaction was studied briefly by Buchner and Geronimus (1903) long before Hine s rediscovery of carbenes. They found ethyl 2-phenylcyclopropanecarboxylate in the reaction of ethyl diazoacetate with styrene. Besides the attention given by Staudinger et al. (1924), however, little attention was paid to that work, apart form the fact that it may be a 1,3-dipolar cycloaddition followed by an azo-extrusion (Sect. 6.5). 

As mentioned briefly in Section 6.5, it should be emphasized that there is no clear evidence available whether cyclopropanes, including these methanofullerenes, are formed via dihydropyrazoles, i.e., by a 1,3-dipolar cycloaddition, or by the primary dediazoniation of the diazoalkane to a carbene that subsequently reacts with 50-It may be that the mechanism is a dipolar cycloaddition followed by azo-extrusion at low temperature (20°C, i.e., Suzuki s conditions), but a carbene reaction in boiling toluene (Isaacs and Diederich), as shown in Section 6.5, Scheme 6-37, pathways C and A, respectively. In addition, the dihydropyrazole may be the product of a side-equilibrium only, but the reagents form the cyclopropane-type methanofullerene via pathway C. A mechanism via primary dediazoniation is, however, unlikely as dediazoniation of diazoacetate without C o in boiling toluene is much slower than it is in the presence of 50 (Diederich, 1994). 

---