Earlier work has shown that the cyclo-addition of a ketene to a conjugated diene is always (a) 2 + 2, (b) polarily directed, and (c) suprafacial with respect to the diene C=C. The adducts of ketenes and cyclopentadiene are thus always 7-substituted bicyclo [3.2.0] hept-2-ene-6-ones. New evidence is presented to show that unsymmetrically substituted ketenes add to cyclopentadiene in such a manner that the larger substituent has a greater tendency to take up the endoposition in the adduct. This is interpreted to mean that a ketene participates in such reactions antarafacially. Thus the ketene approaches cyclopentadiene (a) with its functional plane perpendicular to that of the ring, (b) with the carbonyl carbon over the middle of the ring, and (c) with the larger of the two substituents oriented preferentially away from the ring (transition state 11).This endo-specificity for the larger ketene substituent is demonstrated by the indicated endo/ exo ratios observed in the cyclo-adducts from ketenes with the following substituent pairs :C,H,/CH, = >95/<5, CH,/Cl = SO/ZO, CH,/CH=CH, = -65/35, C,H,/CH, = -60/40, n-C,H,/CH, = -60/40, CH,/Br = 56/44. These ratios enable a list to be compiled indicating the endo-specificity of the ketene substituents. The order closely parallels the space filling capacity as derived by other methods. The establishment of such ratios required reliable configurational assignments at carbon 7. These were derived by five methods based upon the following effects: (1) Both H-C7 and CH3-C7 cause nmr. signals at higher field in endo-position (compared with e m ) . (2) The CH,-C7 group in exo-position gives rise to a nuclear Overhauser effect with the vicinal H-C1, and in one case also with the trans-annular H-C5. (3) The nmr.-coupling constants of H-C7-exo (observed at H-C7) with H-C1 is always iarger than of H-C7-endo. (4) The coupling constant of H-C7-exo with H-C6(known to be exo-) of the LiAlH, reduction products of the cyclo-adducts (observed at H-C6) is always larger than that of H-C7-endo. (5) The nmr. signals of most protons in the cyclo-adducts are a t higher field in benzene than in chloroform solution; this "benzene shift" is larger for H-C7 or for CH,-C7 when in exo-than when in endo-position.Ketene (1) reagieren mit Olefinen (2) gewohnlich unter Bildung von Cyclobutanonen (3) [l] [a]. Die kinetischen Parameter und Losungsmitteleffekte zeigen [3], dass es sich -I -A-)=c=o .+ )={ + 4-4 1 2 3 dabei um eine Mehrzentren-Cycloaddition handelt . Zur Erklarung der Beschleunigung und der sehr wirksamen konstitutione118n Spezifitat bei Olefinen mit Donator-Substituenten (4) nimmt man einen polaren ubergangszustand (5) an [3]. Sorgfaltige x 1 4 5 Analysen der Cycloaddukte weisen auf eine ebenfalls sehr wirksame Stereospezifitat im Sinne einer syn-(= suprafacialen) Addition an das Olefin (6) [4] [5] hin. Besonders ---_ _---: f 6 interessant, und bis vor kurzem immer wieder zu Widerspruchen Anlass gebend, ist die offenbar stark bevorzugte ( 2 + 2)-Cycloaddition (7) auch in Fallen (8) wo eine (2 + ...

rearrangement of acyl-stabilised ammonium ylides has been investigated with regard to stereoselectivity, intramolecularity and the formation of products in addition to the [l ,Z] rearrangement 1012

Syntheses of Enantiomerically Pure Violaxanthius and Related CompoundsThe epoxides 16 and ent-16, prepared by Sharpless-Katsuki oxidation of 15 in excellent yield and very high enantiomeric purity, were used as synthons for the preparation of (+)-(S)-didehydrovomifoIio1 (45), (+)-(6S,7E,9E)-abscisic ester 46, (+)-(6S,7E,9Z)-abscisic ester 47, (-)-(3S,7E,9E)-xanthoxin (49), (-)-(3R,7E,9E)xanthoxin (50), (3S,5R,6S,3'S,S'R,6S, all-E)-violaxanthin (1) (3R,SR,SS,3'R,SR,6S, all-,?)-violaxanthin (55) and their (9Z) (see 53,57), (1 32) (see 54,58), and (152) (see 60) isomers. The novel violadione (61) was prepared from I by oxidation with DMSO/Ac,O. By base treatment, 61 was converted into violadienedione (62), a potential precursor of carotenoids with phenolic end groups. Schema 3 5 R'= H, R'=CH=CHC(CH,)=CHCOOCH, 6 8 R'=Ac, R2=C=CCH(CH3)0Ac 9 11 R'=Ac, R'=CH=CHCOCH, 12 7 1 : 6 [8] 10 1.2 [9] 13 1.4 [lo] eine ergiebige Synthese von 1 anstrebten. Fur unsere Syntheseplanung war die Tatsache wichtig, dass Epoxydierungen an der 3-Hydroxy-j3-Endgruppe stets zu einem Uberwiegen der cis-Stereoisomeren gefuhrt hat; s. 5-+6/7, 8-+9/10 und 11+12/13 in Schema 3. Besonders ungunstig sind die Ergebnisse auf der Carotinoid-Stufe, wie Resultate an Lutein [ll], Antheraxanthin [12] und Zeaxanthin (s. Kap.8) zeigen. Wir haben deshalb einen andern Weg eingeschlagen und die Epoxid-Funktion als erstes Chiralitatszentrum und die OH-C(3)-Gruppe nachtraglich eingefiihrt. Damit kam die enantioselektive Epoxydierung nach Sharpless et al. [13] als entscheidender Schritt an den Anfang der Synthese. Synthese der diastereoisomeren (SR,6S)-3-Hydroxy-5,6-epoxy-ionyliden-alkohole 38 und 39 (Schema 4) '). ~ AusIsophoroncarbonsaure-ethylester (14) wurde nach einer verbesserten Vorschrift nach [ 141 der Dioxolanylalkohol 15 erhalten'). Anschliessend wurden nach dem katalytischen Verfahren von [I31 bei -70" die enantiomeren Epoxide 16 und ent-16 in Ausbeuten von 93 bzw. 95% und ee-Reinheiten von 97,4 bzw. 94,8% hergestellt2). Hydrolyse der Acetale 17 und ent-17 ergab die sehr labilen Ketone 18 und ent-18, welche unter sauren und basischen Bedingungen rasch in die Enone 19 bzw. ent-19 ubergingen. Diese unerwunschte Reaktion liess sich durch Acetal-hydrolyse mit Montmorillonit ('clay und nachfolgender Reduktion rnit (NaBH,/Et,O/MeOH/ 1,2-Dimethoxyethan ( = Monoglym)) weitgehend vermeiden. Allerdings gelang uns trotz vieler Versuche nicht, eine diastereoselektive Reduktion zu erreichen: 20 und 21 entstanden nebeneinander im Verhaltnis 1 :0,7. Nach Schutz von OH-C(3) durch Veresterung rnit Pivaloyl-chlorid (+22 bzw. 23) wurde die AcO-Gruppe verseift (+24 bzw. 25), der Alkohol mit DMSO/Oxalyl-chlorid [16] oxydiert und der Aldehyd 26 bzw. 27 rnit einer Wittig-Reaktion zum geschutzten Epoxyionon 28 bzw. 29 oder nach Wittig-Horner zum C,,-Ester 30 bzw. 32 verlangert. Dabei entstand stets auch das (9Z)-Isomere 31 bzw. 33, und zwar im Fall von reinem 26 zu ca. 40%, bei reinem 27 zu ca. 20%. Alle Isomeren wurden rein hergestellt und spektroskopisch charakterisiert; s....

A new synthesis of 2‐pyrones has been developed. Two molecules of α, β‐unsaturated acid chlorides (8, 12 and 18) condense, with loss of two molecules hydrogen chloride, to pairs of substituted 2‐pyrones (9 and 10, 13 and 14, 19 and 20) when treated with triethyl amine in chloroform or methylene chloride at room temperature. In the case of 18, two additional products were obtained, namely the resorcinol derivative 21 and traces of the 1, 3‐cyclobutanedione derivative 22. Under the same conditions the α, β‐unsaturated acid chlorides 8, 15, 18 and 41 were condensed with trichloroacetyl chloride to give 6‐trichloromethyl‐2‐pyrones (42, 43, 44 and 46). These 2‐pyrones are valuable intermediates for the synthesis of 6‐carboxy‐2‐pyrones and 6‐methyl‐2‐pyrones. A methyl group in β‐position of the α, β‐unsaturated acid chloride appears to be essential for the described condensations, for the acid chlorides 16 and 17 did not yield defined products and the acid chloride 40 reacted with trichloroacetyl chloride in a very low yield. It is considered that the described reactions proceed via the 1, 4‐addition of an acid chloride to a vinyl ketene or through the acylation of an intermediate anion by an acyl derivative as outlined in reaction scheme 1. The structures of the 2‐pyrones were confirmed by their spectroscopic properties, summarized in table 3, and by some of their chemical transformations.

No abstract