Benzo[a]heptalenes from Heptaleno[1,2‐c]furans. Part 2

Abstract

AbstractIt is shown in this ‘Part 2’ that heptaleno[1,2‐c]furans 1 react thermally in a Diels–Alder‐type [4+2] cycloaddition at the furan ring with vinylene carbonate (VC), phenylsulfonylallene (PSA), α‐(acetyloxy)acrylonitrile (AAN), and (1Z)‐1,2‐bis(phenylsulfonyl)ethene (ZSE) to yield the corresponding 1,4‐epoxybenzo[d]heptalenes (cf. Schemes 1, 5, 6, and 8). The thermal reaction of 1a and 1b with VC at 130° and 150°, respectively, leads mainly to the 2,3‐endo‐cyclocarbonates 2,3‐endo‐2a and ‐2b and in minor amounts to the 2,3‐exo‐cyclocarbonates 2,3‐exo‐2a and ‐2b. In some cases, the (P*)‐ and (M*)‐configured epimers were isolated and characterized (Scheme 1). Base‐catalyzed cleavage of 2,3‐endo‐2 gave the corresponding 2,3‐diols 3, which were further transformed via reductive cleavage of their dimesylates 4 into the benzo[a]heptalenes 5a and 5b, respectively (Scheme 2). In another reaction sequence, the 2,3‐diols 3 were converted into their cyclic carbonothioates 6, which on treatment with (EtO)3P gave the deoxygenated 1,4‐dihydro‐1,4‐epoxybenzo[d]heptalenes 7. These were rearranged by acid catalysis into the benzo[a]heptalen‐4‐ols 8a and 8b, respectively (Scheme 2). Cyclocarbonate 2,3‐endo‐2b reacted with lithium diisopropylamide (LDA) at −70° under regioselective ring opening to the 3‐hydroxy‐substituted benzo[d]heptalen‐2‐yl carbamate 2,3‐endo‐9b (Scheme 3). The latter was O‐methylated to 2,3‐endo‐(P*)‐10b. The further way, to get finally the benzo[a]heptalene 13b with MeO groups in 1,2,3‐position, could not be realized due to the fact that we found no way to cleave the carbamate group of 2,3‐endo‐(P*)‐10b without touching its 1,4‐epoxy bridge (Scheme 3).The reaction of 1a with PSA in toluene at 120° was successful, in a way that we found regioisomeric as well as epimeric cycloadducts (Scheme 5). Unfortunately, the attempts to rearrange the products under strong‐base catalysis as it had been shown successfully with other furan–PSA adducts were unsuccessful (Scheme 4).The thermal cycloaddition reaction of 1a and 1b with AAN yielded again regioisomeric and epimeric adducts, which could easily be transformed into the corresponding 2‐ and 3‐oxo products (Scheme 6). Only the latter ones could be rearranged with Ac2O/H2SO4 into the corresponding benzo[a]heptalene‐3,4‐diol diacetates 20a and 20b, respectively, or with trimethylsilyl trifluoromethanesulfonate (TfOSiMe3/Et3N), followed by treatment with NH4Cl/H2O, into the corresponding benzo[a]heptalen‐3,4‐diols 21a and 21b (Scheme 7).The thermal cycloaddition reaction of 1 with ZSE in toluene gave the cycloadducts 2,3‐exo‐22a and ‐22b as well as 2‐exo,3‐endo‐22c in high yields (Scheme 8). All three adducts eliminated, by treatment with base, benzenesulfinic acid and yielded the corresponding 3‐(phenylsulfonyl)‐1,4‐epoxybenzo[d]heptalenes 25. The latter turned out to be excellent Michael acceptors for H2O2 in basic media (Scheme 9). The Michael adducts lost H2O on treatment with Ac2O in pyridine and gave the 3‐(phenylsulfonyl)benzo[d]heptalen‐2‐ones 28a and 3‐exo‐28b, respectively. Rearrangement of these compounds in the presence of Ac2O/AcONa lead to the formation of the corresponding 3‐(phenylsulfonyl)benzo[a]heptalene‐1,2‐diol diacetates 30a and 30b, which on treatment with MeONa/MeI gave the corresponding MeO‐substituted compounds 31a and 31b. The reductive elimination of the PhSO2 group led finally to the 1,2‐dimethoxybenzo[a]heptalenes 32a and 32b. Deprotonation experiments of 32a with t‐BuLi/N,N,N′,N′‐tetramethylethane‐1,2‐diamine (tmeda) and quenching with D2O showed that the most acid CH bond is HC(3) (Scheme 9).Some of the new structures were established by X‐ray crystal‐diffraction analyses (cf. Figs. 1, 3, 4, and 5). Moreover, nine of the new benzo[a]heptalenes were resolved on an anal. Chiralcel OD‐H column, and their CD spectra were measured (cf. Figs. 8 and 9). As a result, the 1,2‐dimethoxybenzo[a]heptalenes 32a and 32b showed unexpectedly new Cotton‐effect bands just below 300 nm, which were assigned to chiral exciton coupling between the heptalene and benzo part of the structurally highly twisted compounds. The PhSO2‐substituted benzo[a]heptalenes 30b and 31b showed, in addition, a further pair of Cotton‐effect bands in the range of 275–245 nm, due to chiral exciton coupling of the benzo[a]heptalene chromophore and the phenylsulfonyl chromophore (cf. Fig. 10).