No abstract

No abstract

The oxidation of 1-(3,8-dimethylazulen-1-yl)alkan-1-ones 1 with 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (¼ 4,5-dichloro-3,6-dioxocyclohexa-1,4-diene-1,2-dicarbonitrile; DDQ) in acetone/H 2 O mixtures at room temperature does not only lead to the corresponding azulene-1-carboxaldehydes 2 but also, in small amounts, to three further products (Tables 1 and 2). The structures of the additional products 3 -5 were solved spectroscopically, and that of 3a also by an X-ray crystal-structure analysis ( Fig. 1). It is demonstrated that the bis(azulenylmethyl)-substituted DDQ derivatives 5 yield on methanolysis or hydrolysis precursors, which in a cascade of reactions rearrange under loss of HCl into the pentacyclic compounds 3 (Schemes 4 and 7). The found 1,1'-[carbonylbis(8-methylazulene-3,1-diyl)]bis [ethanones] 4 are the result of further oxidation of the azulene-1-carboxaldehydes 2 to the corresponding azulene-1-carboxylic acids (Schemes 9 and 10).1. Introduction. -More then ten years ago, we applied a procedure of Okajima and Kurokawa [1], just published at that time, to the smooth oxidation of the Me group of 1-(3-methylazulen-1-yl)alkan-1-ones with 2,3-dichloro-5,6-dicyano-1,4-benzoquinone (¼ 4,5-dichloro-3,6-dioxocyclohexa-1,4-diene-1,2-dicarbonitrile; DDQ) in aqueous acetone to yield the corresponding azulenecarboxaldehydes. The authors reported for the oxidation of 3-acetylguaiazulene (¼ 1-(5-isopropyl-3,8-dimethylazulen-1-yl)-ethanone; 1a) an attractive yield of 91% for carboxaldehyde 2a (Scheme 1). In our hands, the reaction gave, in a ten times higher concentration of the reactants, also 2a as the main product, however, in yields ranging from 40 to 60%; and to our surprise, on TLC beside the dark red spot of 2a, at least two additional faint spots were present, a blue one, moving distinctly faster, and a red one, moving clearly slower than 2a, which stood for two additional products of unknown structure and in estimated yields of ca.

Short Total Syntheses of (&)-Sativene and (+)-cis-SativenediolOur approach to (A)-sativene (7) and (f)-cis-sdtivenediol (9) involves: a) reaction of 3-methylhutanoyl chloride with Et,N/cyclopentadiene to give the endo-isopropyl-ketone 1 (here improved to 71 %), b) NBS hromination of 1 to a 5 : 1 mixture (87 YO) of the hromo-ketones 2 and 3, c) NFD-reaction sequence initiated by the attack of 1.2-hutadienyl titanate (complex of 15, obtained from 2-hutine) on 2/3 to afford 52% of the hrexenone derivative 4 (along with 8% of its epimer 16), d) addition of dibromomethane to 4 forming 63% of the diene-alcohol 5 (along with 13% of the diene-carbaldehyde 38), and e) carhenoid ring-expansion with MeLi applied to 5 resulting in 41 YO the diene-ketone 6 (along with 15 % of a 1 :3 mixture of the diene-ketones 32 and 33). Wolff-Kishner reduction of 6 led to 81 % of (f)-sativene (7), when enough 0, was present, but to 97% of the diene 8 in the strict absence of 0,. (+)-cis-Sativenediol (9) was obtained (86%) by OsO, hydroxylation of 8. The brexenone derivatives 4 and 16 (6: I , 50%) were also produced when the NFD-reaction sequence was applied to the isomeric bromo-ketone mixture 13/14 (1 :3). The latter was obtained by NBS hromination of 10, which in turn was avdilahk by base epimerization of 1, followed by destructive removal of unreacted 1 by repeated gas-flow thermolysis. An analogous (less convenient) route to (f)-sativene (7) passed through a series of dihydro compounds (the ene series); it started with the methylidene-ketone 36, which was the product (97%) of a partial hydrogenation of 4. Addition of dibromomethane to 36 led to 62% of the methylidene-alcohol 39 (along with a little tetracyclic ether 40). Carhenoid ring expansion of 39 with MeLi afforded cu. 42% of the methylidene-ketone 41 (along with 7% of the methylidene-ketone 43 or, under slightly different conditions, along with 9 % of the methylidene-ketone 42 and 10% of the methylidene-carbaldehyde 44). The methylidene-alcohol 39 and the methylidene-ketone 43 were also obtained by partial hydrogenation of 5 and 33, respectively. Wo(ff-Kishner reduction converted 41 into (5)-sativene (7,99%); the same conditions applied to 42 afforded only ca. 8 % 7 (along with three other hydrocarbons, one of them (ca. 21 YO) probably being (f)-copacamphene (45)). In the diene series, the two succeeding reactions (4-5 and 5-6) competed with the same side reaction, a rearrangement leading to the brendene-aldehyde 38. In the ene series, the corresponding dihydro-by-product 44 was found in the reaction 39-41, hut not during 36-39. These side reactions could largely be suppressed by keeping the reaction temperature low. An explanation is proposed.