Diradical trapping

Thiirene intermediates in the photolysis of 1,2,3-thiadiazoIes readily undergo ring opening to a diradical which can be trapped by reaction with an alkyne (Scheme 25) (79CB1769). Significant polymerization is not observed. 

The mechanism of the Patemo-Biichi reaction is not well understood, and while a general pathway has been proposed and widely aceepted, it is apparent that it does not represent the full scope of reactions. Biichi originally proposed that the reaction occurred by light catalyzed stimulation of the carbonyl moiety 1 into an excited singlet state 4. Inter-system crossing then led to a triplet state diradical 5 which could be quenched by olefinic radical acceptors. Intermediate diradical 6 has been quenched or trapped by other radical acceptors and is generally felt to be on the reaction path of the large majority of Patemo-Biichi reactions. Diradical 6 then recombines to form product oxetane 3. 

Diradical species 4 is more stable than diradical 5, and the oxetane 6 is thus formed preferentially oxetane 7 is obtained as minor product only. Evidence for diradical intermediates came from trapping experiments, as well as spectroscopic investigations. ... 

Reaction of 1 with semicarbazide hydrochloride gives the semicarbazone 4, in 74 % yield, which can be oxidized by selenium(IV) oxide to provide dibenzo[2,3 6,7]thiepino[4,5-rf][l,2,3]selenadi-azole (5) in 80 % yield. Thermolysis of selenadiazole 5 leads, with subsequent release of nitrogen, to diradical 6, which can either dimerize to 7 or lose selenium to give the intermediate cycloalkyne. The latter can be trapped by dienes as cycloadducts.93 Thus, the thermolysis of 5 in the presence of 2,3,4,5-tetraphenylcyclopenta-2,4-dienone gives the cycloadduct 1,2,3,4-tetraphenyltribenzo-[/ ,<7,/]thiepin (8) in 14% yield. 

Dimer and trimer byproducts have been isolated from MMA polymerizations and these are suggestive of 1,4-diradical intermediates.323 28 Lingnau and Mcycrhoff523 found that rates of spontaneous polymerization of MMA were substantially higher in the presence of transfer agents (RH). They were able to isolate the compound (98) that might come from trapping of the biradical intermediate (Scheme 3.65). 

The photoadditions proceed through 1,4-diradical intermediates. Trapping experiments with hydrogen atom donors indicate that the initial bond formation can take place at either the a- or (3-carbon of the enone. The excited enone has its highest nucleophilic character at the (3-carbon. The initial bond formation occurs at the (3-carbon for electron-poor alkenes but at the a-carbon for electron-rich alkenes.191 Selectivity is low for alkenes without strong donor or acceptor substituents.192 The final product ratio also reflects the rate and efficiency of ring closure relative to fragmentation of the biradical.193... 

The dimer showed no ESR signal as a solid but when it was dissolved in pure solvent, it gave rise to a strong characterless ESR signal, which was virtually identical to that shown by the original photolysis solution. Attempts to trap the putative diradical 75 failed, but do not rule out its possible existence as an intermediate in low concentration in the system. 

In addition to the photoxygenation/diimide (equation 6),16) silver salt (Eq. 22), 36) and triflate (Eq. 44)s6> routes, 9 has also been prepared by benzophenone-sensitized photodecomposition of the corresponding azo compound 59 and trapping of the resultant triplet diradical with oxygen (Eq. 45) 57). 

Although no other examples have been reported, oxygen trapping of azo-derived triplet diradicals provides a potentially versatile strategy for the synthesis of bicyclic peroxides under neutral conditions. 

Use of mild conditions was crucial and the development of diimide reduction of singlet oxygenates, silver-salt-assisted displacement of halide by peroxide nucleophiles, peroxymercuration and demercuration, peroxide transfer from organotin to alkyl triflates, and oxygen trapping of azoalkane-derived diradicals have all played a part in providing the rich harvest of new bicyclic peroxides described herein. 

The cycloadditions in entries 1-3 are still believed to occur via a diradical stepwise pathway, as confirmed by obtaining a thermodynamic 78 22 trans/cis mixture of dispirooctanes 536 from frans-dicyanoethylene (533) (entry 3) [13b, 143], The cycloaddition to tetracyanoethylene (131) in the absence of oxygen gives only low yields of the [2 + 2] adduct, due to the simultaneous formation of products 542 and 543 (Scheme 74) [13b]. Still, the formation of the cyclobutanes 537 and 542 is noteworthy, since the reactions of TCNE with phenyl substituted MCPs exclusively afford methylenecyclopentane derivatives [37,144], The reaction is thought to occur via dipolar intermediates 539-541 formed after an initial SET process (Scheme 74) [13b]. The occurrence of intermediates 540 and 541 has been confirmed by trapping experiments [13b]. 

DMAZD also proved an effective trapping agent for the 1,3-diradical intermediate derived by thermal extrusion of N2 from 81, which was itself prepared from DMAZD and 6,6-dimethylfulvene (Scheme 9). The final product was the cyclopentapyrazole 82.131... 

The mechanism of oxetane formation is similar to the one discussed for cyclobutane formation in chapter 4.3.3. The 1,4-diradicals can be efficiently trapped with molecular oxygen. The resulting 1,2,4-trioxanes are interesting synthetic intermediates (4.81) 495>. 

RIES from diazoalkanes is also sensitive to the dihedral angle between the migrating a-H and the C-N bond of the diazo moiety.57 For example, the A values for the pyridine capture of the photolytically generated carbenes from 46 and 47 are in the ratio of 1.7 1. Similarly, the carbene from 46 is more efficiently generated and trapped in methanol, whereas the photolysis of 47 in methanol affords twice as much olefin (by 1,2-H RIES) compared to the photolysis of 46. These phenomena are attributed to conformational factors that favor RIES during the photolysis of 47, with the proximal excited state represented as a pyramidalized 1,3-C-N=N diradical.57... 

Numerous synthetic and mechanistic studies were done to investigate this reaction further, and a variety of enediynes have been thermalized in the presence of radical traps such as 1,4-cyclohexadiene. Even though large excesses of radical traps were employed, the yields of the substituted benzenes were often moderate at best. Most important of all, Tour et al.50 demonstrated that 1,4-naphthalene diradicals generated in solution couple to eventually form a polymer [Eq. (9)]. 

Much experimental and theoretical work has been performed with the two allenes 1,2,6-heptatriene (32) and 1,2,6,7-octatetraene (34). Thermal isomerization of 32 leads to 3-methylene-l,5-hexadiene (346), a process that at first sight looks like a typical Cope rearrangement. However, trapping experiments with either oxygen or sulfur dioxide have shown that at least half of the rearrangement passes through the diradical 345 (Scheme 5.52) [144],... 

In addition to the tert-butyl enol ethers mentioned above (15% yield), the action of KOtBu on l-iodo-4-methylcyclohexene in DM SO furnished the dimers 85 and tri-mers of 81 in 30 and -25% yield (Scheme 6.24). As in the case of 6 (see Scheme 6.10), the formation of oligomers of 81 was completely suppressed on performance of this reaction in the presence of (tBu)2NO, whereas theenol ethers (86 and its 5-methyl isomer, with the former originating in part and the latter totally from 4-methylcydohex-yne) were observed as in the reaction in the absence of the stable radical. Instead of the dimers 85 and the trimers of 81, a mixture of the hydroxylamine derivatives 87 was isolated in 38% yield. These findings indicate that 81 has no diradical character, in contrast to its immediate dimer 84, which is hence trapped quantitatively by (tBu)2NO [61]. 

Scheme 6.24 Formation of the enolether 86 and of oligomers from 5-methyl-l,2-cyclohexadiene (81) and trapping of the diradical 84 by di-tert-butyl nitroxide, according to Bottini and co-workers.

The dimer 352 of 351 was isolated from the product mixtures of two experiments conducted to trap 351 by alkenes, one with 350 and the other with 354 as substrate. Although no cycloadduct with the alkene was observed in one case, the yield of 352 amounted to only 0.8%. Nevertheless, the structure of 352 is interesting, since it suggests that the tetramethyleneethane diradical assumed to be the intermediate undergoes ring closure preferentially between two different allyl-radical termini. 

---