9. ALICYCLIC COMPOUNDS By K. Mackenzie (Organic Chemistry Department University of Bristol) Small and Medium Rings.-Reviews cover the following subjects cyclo-propenes ;' cyclohexadienones ;'hexamethyl Dewar benzene ;3 norcaradiene compounds ;4 bicyclo[n,l,l]alkanes ;5 bullvalene ;6 cyclopropyl cyclobutyl and homoallyl systems ;' eliminations in cyclic cis-trans isomers ;8 topological approach to ring structure ;' free-radical reactions in bridged bicyclic systems ;lo cyclisation of free-radicals ;' photochemistry ;12 Diels-Alder reactions;' conformational abnormalities in cyclohexanes ;l4 and orbital symmetry principles.' Non-Bridged Systems.-Dialkoxycar bonylcarbene pho tochemica Ily gener-ated from diazomalonic ester has previously been observed only in C-H insertion reactions ;similarly prepared in stereoisomeric olefins it gives cyclo- propanes stereospecifically.Benzophenone sensitization destroys this specifi- city however (singlet and triplet states) ;some evidence for multiplicity change is evident from the decline in stereospecificity observed with cis-olefins in unsensitized ester photolysis in an inert solvent with increasing dilution after an increase in specificity in the initial stages of dilution (reduced collisional deactivation of primary singlet carbene). Intersystem crossing appears to be much less favourable than with fluorenylidene.I6 Singlet carbenes may also account for the stereochemistry of the cyclopropanes produced in photolytic decomposition of p-methoxyphenylsulphonyldiazomethane in cis- and trans- 2-butenes ; here the preponderance of anti-product from the cis-olefin is ascribed to steric mitigation of the favourable Van der Waals-London forces and charge sharing in the transition state previously invoked to explain the more usual preference for syn-adduct.' G.L. Closs. Adv. Alicyclic Chmi. 1966. I. 5.7 'I A. J. Waring Ado. Alicyclic Chem. 1966 I. I29 W. Schafer and H. Hellmann Angew. Clicrri. Irrternat. Edn. 1967 6 518. G. Maier Angew. Chem. Internat. Edn. 1967,6 402 J. Meinwald and Y. C. Meinwald Adv. Alicyclic Chem. 1966 1 2. G. Schroder and J. F. M. Oth Angew. Chem. Internat. Edn. 1967,6,414. M. Hanack and H. J. Schneider Angew. Chem. Internat. Edn. 1967,6,666. * W. Hiickel and M. Hanack Angew. Chem. Internat.Edn. 1967,6 534. R. Fugman V. Dolling and H. Nickelsen Angew. Chem. Internat. Edn. 1967,6 723. lo D. I. Davies and S. J. Cristol Adu. Free Radical Chem. 1965 1 155. l1 M. Julia Pure App. Chem. 1967 15 167. '' R. N. Warrener and J. B. Bremner Rev. Pure Appl. Chem. (Australia) 1966,16 117; P. J. Kropp Org. Photochem. 1967,1,1 see also R. Padwa p. 91; K. Schaffner Adu. Photochem. 1966,4,81. l3 J Sauer Angew. Chem. Internat. Edn. 1967,6 16. l4 D. L. Robinson and D. W. Theobald Quart. Rev. 1967,21,314. Is R. B. Woodward Chem. SOC. Special. Publ. No. 21 217. l6 M. Jones A. Kulczycki and K. F. Hummel Tetrahedron Letters 1967 183. '' A. M. van Leusen R.J. Mulder and J. Strating Rec. Trao. chim. 1967,86 225. K.M ackenzie 0 t Me2 S=CH Me Me (1) R'~ S=CR~ (2)R' = Me R' = H (3)R' = Ph R2 = Me R1y: R4 R5 (5) R' = R' = Me R3 = Ac R4 = R5 = H (7) R' = R' = C02Me R3= H R4 = R5= Me or R' = C02Me R2 = R3 = B R4 = Rs = Me (6) R = H or Me H (9) R = CHO (8) (10)R = CH20H (11) R = C02H In contrast to the reaction with carbonyl compounds to give epoxides the methylide (1) reacts with @-unsaturated ketones to give acylcyclopropanes (5);18" stereospecificity is observed in some cases depending on whether C-1-C-2 rotation is possible in the intermediate (4).'*' Sulphurane (2) reacts at carbonyl generally to give oxirans but the highly reactive analogue (3),like (l),gives cyclopropanes [(6)and (7)] in low temperature reactions with ap-unsaturated carbonyl comp~unds.'~ The enhanced s-character of C-H bonds in cyclopropanes in comparison with cyclobutanes prompted examina- tion of Cannizzaro reactions with the relevant aldehydes." Cyclopropane-carboxaldehyde appears normal in this respect save for the unexplained formation of its dimethyl acetal with methoxide anion in methanol.In the cyclobutane series dimeric products [(lo) and (1 l)] are largely formed perhaps via (9) with only very small yields of the normal products reported earlier. In this connection cyclopropanecarboxaldehyde is conveniently prepared by (a)C. Agami Bull. SOC.chim.France 1967 1391 ;(b)C. Agami and C. Prevost ibid. p. 2299. l9 E. J. Corey and M. Jautelat J. Amer. Chem. Soc.. 1967.89. 3912. (a)F. P. B. ban der Maeden H. Steinberg and Th. J. de Boer Tetrahedron Letters 1967,4521 ; (b)W.H. Perkin and H. G. Coleman J. Chem. SOC. 1887 228. Alicyclic Compounds ceric ion oxidation of the carbinol; the yield is less than that from established hydride reduction methods however.21 The technique for selective removal of chlorine in gem-chlorofluorocyclopropanesby use of tri-n-butyltin hydride is stereospecific ;2 2a chlorine is also selectively removed with sodium-liquid ammonia.22b The advantageous use of chromous ion for dehalogenating cyclopropanes appears to offer scope for further study ; gern-dibromocyclo-propanes are mono-debrominated the syn-isomers arising in the case of dibromobicyclo[n,1,0]alkanes (12) but further reduction can occur. Ring opening to cyclic allenes is observed for large values of n possibly via (15) produced from the initially formed cyclopropyl radical (13) through (14); the latter is the precursor of the monobromocyclopropane when n is too small for ring expansion through (15) but neither (14) or (15) can be trapped with 01efins.~~ Ring expansion of bridgehead alkoxybicyclo[n,l,0]alkanes containing a halogenocyclopropane is known ; rearrangement of 7,7-dibromonocarane to 2-bromo-3-ethoxycycloheptene or cyclohepten-2-one in the presence of silver is therefore not surprising.A similar reaction of the bis-adduct (16) gives the 1-tetralone (19) by the sequence (16) -f (17) -+(18) -,(19).24 Allenes and polycyclic compounds are mostly formed in the reactions of dibromocarbene-2' L. B. Young and W. S. Trahanovsky J.Org. Chem. 1967,32,2349. 22 (a)T. Ando F. Numigata H. Yamanaka and W. Funaska,J. Amer. Chem. Soc. 1967,89,5719; T. Ando H. Yamanaka S. Terabe A. Horike and W. Funasaka Tetrahedron Letters 1967 1123; (b)M. Schlosser and G. Heinz Angew.Chem. 1967 6 629. 23 H. Nozaki T. Aratani and R. Noyori Tetrahedron 1967,23,3645. 24 A. J. Birch G. M. Iskander B. I. Magboul and F. Stansfield J. Chem. SOC.(C) 1967 358. 314 K. M ackenzie olefin adducts with methyl-lithi~m,~~" but 2,2,2',2'-tetrabromobicyclopropyl derivatives give fulvenes e.g. (20) + (21) and (22) + (23) whilst cyclopenta- dienes are formed from gem-dibromovinylcyclopropanes. For the latter a route uia (24) seems possible with vinylcyclopropane rearrangement ; closure to a strained tricyclopentane structure and rearrangement or insertion at vinyl C-H could lead to bicyclopentenes but since reaction still ensues when there are terminal vinylic groups and further since (25) gives equal amounts of 1- and 2-methylcyclopentadienes,(28) and (29) whereas one might expect more of the 2-methyl isomer if bicyclopentenes were involved renders these routes unlikely.Whilst vinylcyclopropane rearrangements normally require elevated temperatures favourable interaction of the carbene site and the double-bond leads through (30) and (31) to (28) and (29). The formation of fulvenes from (20) and (22) is accomodated by initial opening of one ring to an allene and similar ring-expan~ion.~~~ 'b-7 ?R; R' R2 Br Br 9;. (20) R' = H R3 = Me R3 (24) R'-5 = H (22) R' = Me R3 = H (21) R' = R3 = Me R2 = H (25) R' 3-5 = H (23) R' = R' = Me R3 = H R2 = Me d V (26)R' = Me R2 = H (28)R' = H R' = Me (27)R' = H.R' = Me (29)R' = Me. R2 = H (30) The conjugative ability of cyclopropyl groups at unsaturated centres is well known ; factors governing photolysis of bicyclo[4,l,0]hexanones [derivatives of (32)] have been studied. Cyclopropyl ring-opening only competes with C-2- C-3 cleavage as the primary process when C-3 is unsubstituted or when C-7 is substituted for that bond which is best conjugated ; if both bonds conjugate equally well that bond with the highest degree of terminal substitution is cleaved.26 The direction of bond-migration in the decomposition of alkylcyclopropyl methyl ketone tosylhydrazone to cyclobutenes has received further attention.Vicinal alkyl groups in the cyclopropane appear at C-3 and C-4 in the cyclo- butene but cases of unsymmetrically substituted cyclopropanes have only now been examined. The predominant products are those formed by migration of 25 (a) L. Skattebel Chem. and Ind. 1962 2146; (b) L. Skattebel Tetrahedron 1967 23 1107; cf. C. G. Cardenas B. A. Shoulders and P. D. Gardner J. Org. Chem. 1967 32 1220. 26 W. G. Dauben and G. W. Shaffer Tetrahedron Letters 1967,4415; cf. R. E. K. Winter and R. F. Lindauer ibid. p. 2345. Alicyclic Compounds the less-substituted bond. The reaction may not be a true carbene rearrange- ment but a concerted loss of nitrogen from the intermediate diazo-compound and because the more substituted bond is subject to a strong steric interaction with the ‘carbene’ methyl during migration the alternative course is followed.If a carbene is involved rotation through the plane of the ring in the product- forming step will be easier if rotation is away from the ring-substituted carbon e.g. (42) -+ (43).27 (33)R’ = R’ = Me (36)-438) R3 = H R4 = Me and R2 as for (33t(35) (34)R’ = Me R’ = H (39)+41) R3 = Me R4 = H (35)R’ = H R’ = Me Cyclopropylcarbinylcarbene made from cyclopropanecarboxaldehyde tosyl-hydrazone gives mainly cyclobutene with only 11 % acetylene and ethylene whereas cyclopropyldiazomethane photolyses to give mainly the latter and butadiene. In contrast cyclopropane reacts with ‘cold’ carbon atoms to give methylenecyclopropane (C-H insertion) and no ethlyene or acetylene and with ‘hot’ carbon atoms to produce only these latter products.Different factors are clearly involved in the fate of the cyclopropylcarbinylcarbene produced under the various conditions but insofar as only the singlet species can give rise to acetylene and ethylene doubt is raised as to whether ‘hot’ carbon atoms are triplets; both states may be produced in the photochemical reaction.28 Isopropylidenecarbene is formed from 1,1 -dibromo-2-methylpropene and magnesium ; reactive olefins combine with it to give methylenecyclopropanes. If capable of extension the method has obvious advantages over methods employing lithium alkyl~.~~ Prototropic rearrangement of 2.3-di-n-alkyl-’’ C.L. Bird H. M. Frey and I. D. R. Stevens Chem. Comm. 1967 707. P. B. Shevlin and A. P. Wolf J. Amer. Chem. SOC. 1966,88,4735. 29 N. Wakabayashi J. Org. Chem. 1967,32.489. 316 K. M ackenzie cycloprop-2-ene- 1-carboxylic acids also affords methylenecyclopropanes.30 Surprisingly only one example of dehydrohalogenation of a halogenocyclo- propane to a cyclopropene appears to be known but significant yields of the cycloalkene can be obtained from dichlorocarbene adducts of olefins by their addition to base in aprotic media e.g. l,l-dichloro-2,2-dimethyl-3-t-butyl-cyclopropane gives 1-t-butyl-3-chloro-2,2-dimethylcyclopropene; the vinylic chlorine is replaced by thi~alkoxide.~ Similar methods give acetylcyclo- propenes and acetylmethylenecyclopropanes from dihalogenocarbene adducts of alk-3-en- 1-ynes ;the initial ethynyldihalogenocyclopropanesare hydrated converted into acetals dehydrohalogenated and finally hydroly~ed.~~ H (44)R = CO.CHN1 (45) R = CHZ.CO.CHN2 phxph Ph hPh (46) R = CON (47)R = CH,*CON (49)R = Ph PhJi? (49),A R' = HR2 = * Ph (49),B R' = *R2 = H (50) Photolysis of (44)was earlier shown to yield an a-keto-carbene which intra- molecularly adds to the double bond in competition with Wolff rearrangement.For ,the azides (46) undergoes Curtius rearrangement when heated to the isocyanate but low temperature photolysis in ether gives a nitrene-solvent insertion product. The difference between (44)and (46) may be due to the absence of an excess of vibrational energy in the keto-nitrene in comparison with the carbene.Thermolysis of (47) gives isocyanate but low temperature photolysis in ether and ethanolysis of the products gives among other com- pounds 5,6-diphenylpyridone which could arise from the intramolecular nitrene adduct (48); the latter could also be formed from an intermediate triazoline however. Irradiation of degassed solutions of triphenylcyclopropene (49) is ineffective 'O A. W. Herriott and W. M. Jones Tetrahedron Letters. 1967,2387. 31 T. C. Shields B. A. Loving and P. D. Gardner Chem. Comm. 1967,556. 32 N. Bertrand and H. Monti Compt. rend. 1967,264 C 998. 33 N. C. Castelluchi M. Kato H. Zenda and S. Masamune Chem. Comm. 1967,473. Alicyclic Compounds but with filtered light and benzophenone dimer (50) and compound (51) are formed ; the ratio is sensitizer independent implicating triplet (49).Hydrogen transfer is clearly involved in formation of (51) but no isotope effect is observed with deuterated (49). Intermediates (49),A and (49)2B can be visualized by addition of triplet (49) to (49) and whilst the former intermediate can easily close to give (50) the latter would have to assume an unfavourable (boat) conformation and hydrogen transfer therefore supervenes. Structure (50) is supported by the extremely low-field aliphatic proton n.m.r. signals.34 The stability of NN'-di-t-butyldiazacyclopropanone encouraged the hope that sym-di-t-butylcyclopropanone might be isolable ; indeed treatment of 2,2,6,6-tetramethyl-3-bromo-heptan-4-one with base affords the trans-isomer (ti 6 hr.at 150"); the readily formed benzyl hemiacetal exhibits non-equivalent t-butyl groups in the n.m.r. spectrum.35 Pure cyclopropanone has itself finally been isolated (vmax,1816 cm.-' ;z 8-28)and is stable at -196" for a few days.36 (52) R'R' = o (55) R = 1-hydroxycyclopropyl (53)R' = OH R2 = NMe (56) R = H (54)R' = R' = NMez X 07) (58) (59) X = CH, 0or S Whilst the reactions of the ketone with water and alcohols have been known for many years its behaviour with amines has only recently been explored. Dimethylamine gives the addition compounds (53) and (54); the former can also be made from cyclopropanol and dimethylamine (molecular sieves absorb the water produced) and hydrogen cyanide in ether gives the cyanhydrin.Heterocyclic products are formed if cyclopropanone is treated with methyl- amir~e.~' At low temperatures (55) and (56) are formed with an equivalent of aniline and the latter reacts with more ketone to form (59 which however reforms (56) with aniline;38" it had earlier been suggested that (56) might be 34 C. Deboer and R. Breslow Tetrahedron Letters 1967 1033; H. Diirr ibid. p. 1649. '' J. F. Pazos and F. D. Greene J. Amer. Chem SOC.,1967,89 1030. 36 S. E. Schaafsma H. Steinberg and Th. J. de Boer Rec. Trac. chim.,1966,85 1170. 37 W. J. M. van Tilborg S. E. Schaafsma H. Steinberg and T. J. de Boer Rec. Trav. chim. 1967 86,417,419. 38 (a)N. J. Turro and W. B. Hammond Tetrahedron Letters 1967,3085;(b)€1.H. Wasserman and D. C. Clagett J. Amer. Chem. Soc. 1966,88 5368. K.M ackenzie formed via cyclopropanone in the reaction of the ethyl hemiacetal with aniline.38b The ketone reacts with dry hydrogen chloride to form the l-chloro- 1-hydroxy-compound (at room temperature in the presence of acetyl chloride which acts as polymerization inhibiter) and with acetic acid the l-acetoxy-l- hydroxy-compound forms in a reversible reaction (even at -78”); both the acetoxy- and the chloro-hydroxy-compounds form the methyl hemiacetal of cyclopropanone quantitatively with methan01.~’ Solutions of cyclopropenone have now been made by hydrolysis of 1,l-dichlorocyclopropenepresent in the reduction products of tetrachlorocyclopropene ;the free ketone is extractable into organic phases (z 14&1.1; v,,, 1835 and 1870 cm.- I) and is also present in aqueous solution.The greater stability of the enone in comparison with cyclopropanone may reflect conjugative effects and if the carbonyl dipole is taken into consideration a 2~-aromatic system is present.40 The novel reaction of (57) with N-phenacylpyridinium salts to give high yields of coumalins by the sequence (57) +(58) -,(59) is reported.41 + ClNi I S’ L-Q I--. ‘‘ (61) (62) AH2 CH2-C’ @I mI I. ClFC-CF F2C-C CIF Cope rearrangement of divinylcyclobutane is believed to be the source of cyclo-octa-1,5-diene in the thermal and Nio-catalysed dimerization of butadiene. This conclusion is confirmed and extended since the rate of catalytic re- arrangement of divinylcyclobutane is strongly dependent on catalyst con-centration and the ligands present; the activity of the bis-rally1 form (60) as source of cyclo-octadiene increases with marked back-donation of metal to ligand whereas (61) increases in activity with reduction of this effect giving vinylcyclohex-3-ene.The reaction can be diverted to give other products e.g. by addition of ethylene to give cyclodeca-cis- l-tr~ns-5-diene.~~ Another interesting nickel-catalysed process is the conversion of (2 + 2) II 3y N. J. Turro and W. B. Hammond J. Amer. Chem. Soc. 1967,89 1028. 40 R. Breslow and G. Ryan J. Amer. Chem. Sac. 1967,8!3 3073. 41 T. Eicher and A. Hansen Tetrahedron Letters 1967 1169. 42 P. Heimbach and W. Brenner Angew. Chem.Internat. Edn. 1967 6 800. Alicy clic Compounds 319 butadiene4ichloromaleic ester adducts into cyclohexa- 1,4-diene- 1,2-dicar- boxylic esters ; ring scission-valence tautomerism reclosure and isomeri- zation could be envisaged but according to orbital symmetry requirements the primary step here would lead to a trans-1,3,5-hexatriene unfavourable for recyclisation besides which the formation of 4,5-dimethylcyclohexa-1,4-diene-1,2-dicarboxylic ester from 2,3-dimethylbutadiene in a similar reaction rules out this pathway. The rearrangement is pictured as abstraction of chloronium ion and ring expansion as in (62).43 The reaction of 2-alkoxy-2,3-dihydro- pyrans with Grignard reagents is useful in that it leads to cyclobutylalkyl- carbinols rather than to the alkyldihydropyrans or ring-opened products expected.44 The photochemical addition of olefins to allenes as a source of methylenecyclobutanes is not novel ; however 4-methylpenta-1,2,4-triene thermally equilibrates with methyl-3-methylenecyclobutene;the cyclic form is stable below 350°.45 In the expectation that the allylic assistance to rearrange- ment (13.7 kcal./mole) might appear in twofold measure for the degenerate rearrangement of the 1,Zbismethylene compound the thermolysis of bis-(dideuteriomethy1ene)cyclobutanehas been studied ; in fact the further assist- ance to resonance is only ca.3 kcal. The reaction does not appear to have any kinship with the dimerization of allene since it takes place in the temperature range where the latter does not dirneri~e.~~ Reaction of allene with trifluoro- chloroethylene gives an 85 :15 mixture of 2,2-difluoro-1-methylene-and 3,3-difluoro-l-methylene-cyclobutane rather than the single adduct expected by precedent ; this is consistent with initially formed biradical intermediates (63) and (64); the former intermediate can become allylically stabilized (0-rotation) whilst the latter cannot so that the product is formed from (63) rather than (64) which preferentially dissociates in a competing process.Similar results are observed for 1,l -dichlorodifluoroethylene.47 (65)X = Y = Z = C1 (70) X = H; Y = C1; Z = OEt (66)X = Z = CI ;Y = OEt (71) X = H ; Y = OEt ; Z = CI :nZ2 (67)X= Y = C1;Z = OEt (72)X = F;Y = C1;Z = H (68)X = H; Y = Z = C1 (73)X=F:Y=Z=H (69)X = H Y = F:Z = CI (74)X= H:Y = F:Z = C1 It is known that (65) reacts with alkoxide ion to give mainly (66) with a little (67); if this had been due only to the steric effect of the dichloromethylene group one might expect a similar distribution of products from (68) and (69) in similar reactions instead of their specific and quantitative conversion into (70) and (71).For (68) the intermediate carbanion is stabilized not only by a-chlorine but also by the P-difluoro-group but for (69) a-chlorine is a better 43 H.-D. Scharfand F. Korte Chem. Ber. 1966,99 3925. 44 R. Quelet and J. D’Angelo Compt. rend. 1967 264 C 216. 45 E. Gil-Av and J. Herling Tetrahedron Letters 1967 1. 46 W. von E. Doering and W. R. Dolbier. J. Amer. Chem.SOC. 1967,89,4534;cf.J. J. Gajewski and Chung Nan Shih J. .4mer. Chem. Soc. 1967,89,4532. 47 D. R. Taylor and M. R. Warburton Tetrahedron Letters 1967. 3277 320 K.M ackenzie stabilizing influence than a combination of a-fluorine with the P-difluoro- group in the carbanion ;for the former intermediate eclipsing interactions are significantly less than for the alternative possible anion but for carbanions from (69) there are probably only small differences in eclipsing interactions between the alternatives. The important effects of halogens in these compounds is apparent from the much greater reactivity of e.g. (65) in comparison with (68) and (69). The observation of mixed products e.g. (73) and (74) in hydride reduction of (72) but only a single product in alkoxide reactions has led to the suggestion that a different mechanism obtains for hydride reduction where steric effects associated with the approach of the reagent may be more im- portant so that the most stable carbanion may not necessarily be the one most easily formed.This is also apparent in the formation of (71) from (69) in comparison with the reaction of 1-chloro-2,3,3,4,4-pentafluorocyclopentene with hydride ion where vinylic chlorine is removed. Grignard reagents also remove vinylic chlorine from chloroperfluorocycloalkenes ; the yields of products resemble in ring-size dependence those from hydride reduction and only vinylic substitution occurs if vinylic fluorine is removed. Interestingly in this context 1,1'-bi-(2-fluorocyclobuten-3-one) gives the bis-enamine with piperidine by displacement of both vinylic flu~rines.~~ (76) R' = + RC-C .[CH,] * 0-SO,Ar (77) R' = 0.C0.CF3 R CO,Et R2 = Me C1 R' = Me C02Me R2 = Me (78) R' = Co2Me R2 = Me The solvolysis of alk-3-ynyl tosylates to cyclobutanones and cyclopropyl ketones was earlier described ;the yields of cyclobutanones are quantitative in reactions of trifluoroacetic acid with nitrobenzenesulphonates.Mercury salts divert the reaction to give cyclopropyl ketones viu solvolytic ring closure of the vinyl acetate produced precluding formation of the cyclobutanone by 48 J. D. Park Ci. Groppelli and J. H. Adarns Tetrahedron Letters 1967,103; J. D. Park R. Sullivan and R. J. McMurty ibid. p. 173; J. D. Park and W. C.Frank J. Org. Chem. 1967,32 1333. Alicyclic Compounds 321 homoallylic cyclisation of a solvent adduct and implicating direct participation of the triple bond. Consistently the solvolytic rate is only slightly less than that for saturated analogues in spite of the strong inductive effect of the acetylenic group ;cyclobutenyl vinyl cations e.g. (76),are therefore likely.49 Ring-scission to keten and ethylene or decarbonylation to propene or cyclopropane occurs in photolysis of cyclobutanones ; ring expansion to tetrahydrofurans via cyclic carbenes is an especially effective alternative pathway for 2,2,4,4-tetramethylcyclobutanones,analogous to cases in the steroid field." The quest for superior syntheses of bicyclobutanes continues ;a new inter- esting route is formation and decomposition of the bis-adduct (78) which on photolysis yields (among other products) the bicyclobutane (79) (ti.3hr. at 150").51 Dimethyl ester (81) has been obtained by reaction of dimethyl 3,3- dimethylcycloprop-1-ene- 1,2-dicarboxylate with diphenylisopropylidenesul-phurane. l9 The first synthesis of 1,3-dimethylbicyclobutanewas achieved in 3 % yield from diazomethane and 1,2-dimethylcyclopropene;superior methods are the reaction of methylacetylene or allene with hydrobromic acid followed by lithium amalgam dehalogenation (25") of the isomeric 1,3-dimethyl-1,3- dibromocyclobutenes produced to give 1,3-dimethylbicyclobutane(85%),52 and electrolysis of the dibromodimethylcyclobutanes. Control of the electrode potential in the electrolytic method allows preparation of halogenobicyclo-butanes; {e.g.1,1,3,3-tetrachloro-2,2,4,4-tetramethylcyclobutane gives 1,3-di- chloro-2,2,4,4-tetramethylbicyclo[ 1,l ,O]butane (80)). Polarographic evidence suggests that electrolytic ring-closure here involves addition of two electrons to the dihalogenocyclobutane generating a halide anion and a carbanion which undergoes 173-transannular halide displacement ; cyclobutane formed in the reaction may arise from intermolecular protonation of the carbanion by a second molecule of the halogenocyclobutane the derived carbanion ex- pelling halide anion i.e. it is effectively dehydrohalogenated. Support for the intermediacy of cyclic carbanions in these reactions comes from the observation that 3-bromopropyltriethylamine bromide gives cyclopropane on electr~lysis.~~ Several investigators have shown how sensitive bicyclobutanes are to acids in addition reactions in protic media; e.g.3-methylbicyclobutanecarbonitrile requires heating in water at 100" for 3 hr. to produce the same product as reaction at 25" in dilute hydrochloric acid (not an isomer as inadvertently stated last year).54 In cases where there is additional ring-strain rapid reaction with neutral solvents is observed; a carbonium ion is presumably formed by protonation of the ring and solvent attack follows. Methanolysis of tricyclic 49 M. Hanack I. Herterich and V. Viitt. Tetrahedron Letters 1967 3871. N. J. Turro and R. M. Southam Tetrahedron Letters 1967 545; H. U. Hostettler Heh.Chirn. Acta. 1966,49 241 7. M. Franch-Neumann Angew. Chem. 1967,79,98. K. Griesbaum and P. E. Butter Angew. Chem. Ititernat. Edn.. 1967. 6. 444. 53 M. R. Rifi J. Amer. Chem. SOC. 1967,89,4442. 54 E. P. Blanchard and A Cairncross J. Amer. Chem. SOC.,1966.88.487; A. Cairncross and E. P. Blanchard ibid. p. 496. 322 K.M ackenzie compounds (82)-(85) gives mixtures of cyclopropylcarbinyl and cyclobutyl methyl ethers; the product ratio in each case resembles that from known cyclobutyl and cyclopropylcarbinyl carbonium ions similar to those expected here from protonation as first step. The relative rates are (83) > (84) > (85). Strong acid considerably increases the rates but product ratios are insignifi- cantly changed. Increasing reactivity with ring strain is consequently not due due to a change in mechanism.55 Bicyclobutane reacts with benzyne mainly to give 3-phenylcyclobutene but some 1,3-cycloaddition product is also f~rmed.~ A central bond length of 1.49 A correlates well with observed microwave spectra of bicy~lobutane.~~ Stepwise synthesis of 2-methylcyclopentane- 1,3-diones often gives poor yields but they can be made by cyclisation of y-keto-carboxylic acids with aluminium chloride or by the reaction of succinic anhydride with 2-acetoxybut- 2-ene acetate.58 An unusual reaction in the halogenoperfluorocyclopentene field is the formation of (86) by the action of cyanide ion on 1,2-dichloro- perfluorocyclopentene perhaps by vinylic chlorine displacement to give a highly reactive vinylic nitrile.” 1-(l-Bromoethyl)-2-chloroperfluorocyclo-pentene (87) reacts with Grignard reagents to give stereoisomeric l-ethylidene- 2-chloroperfluorocyclopent-2-enes (88); their ratio is insensitive to the re-agents and prior exchange at the bromoethyl group and 1,4-elimination from the carbanion obtains.Each isomer formed can displace vinylic fluorine without stereomutation so that the carbanions involved resemble other simpler allylic charged species in their stereochemical integrity.60 In the field of cyclopentadiene chemistry the question of the nature of the cyclopentadiene-o-benzoquinone adduct has been settled ;6 lo the primarily formed adduct (90)very easily rearranges to the alternative structure suggested (91).616The structure of this and analogous adducts is supported by i.r.and n.m.r. spectral data.6 Cyclohexa-1,3-diene reacts similarly.61” F c1 c1 (88)R = F (89)R = OAlk 55 W. G. Dauben and C. D. Poulter Tetrahedron Letters 1967 3021. 56 M. Pomerantz J. Amer. Chem. SOC.,1966,88 5349. 57 M. D. Harmony and K. Cox J. Amer. Chem. Soc. 1966,88 5049. 58 H. Schick G. Lehmann and G. Hilgetag Angew. Chem. Internat. Edn. 1967,6 371 ;H. Schick G. Lehmann and G. Hilgetag Chem. Ber. 1967 100 2973; V. J. Grenda G. W. Lindberg N. L. Wendler and S. H. Pines J. Org. Chem. 1967,32 1236. ’’ W. R. Carpenter and G. J. Paienik J. Org. Chem. 1967,32 1219. 6o J. D. Park and R. J. McMurty Tetrahedron Letters 1967 1301. 61 (a)W. M. Horspool J. M. Tedder and Zia ud Din Chem. Comm.1966 775; cf. F. J. Evans H. S. Wilgus and J. W. Gates J. Org. Chem. 1965 30 1655; (b)M. F. Ansell A. F. Gosden and V. J. Leslie Tetrahedron Letters 1967,4537;(c) D. D. Chapman H. S. Wilgus and J. W. Gates jun. ibid. 1966 6175. Alicyclic Compounds 323 Cycloaddition reactions of perfluorocyclopentadiene resemble those of the hydrocarbon; the two types of Diels-Alder adducts that can be formed can be distinguished by the "F magnetic resonance spectra for bridge difluoro- methano-groups formed in cycloadditions give AB patterns at higher field than those due to allylic difluoromethylene groups which arise in compounds where the diene has reacted as dienophile.62 Magnetic resonance data on nitrocyclopentadiene liberated by acidification of Thiele's sodium salt indicate that it is the l-nitro-i~omer.~~ Ethylene undergoes cycloaddition with 5,5-dichloro-1,2,3,4-tetracyanocyclo-pentadiene at room temperature ! The astonishingly reactive diene is formed by the action of chlorine on chlorotetracyanocyclopentadienide-itself pre-pared by electrophilic substitution into the tetracyanocyclopentadienide ; so low is its basicity that reactions with the latter can be carried out in acids.Nitration reduction and diazotisation of the derived amine give the previously described diazotetracyanocyclopentadienide whilst Friedel-Crafts acylations are possible in trifluoroacetic acid or with the more conventional aluminium chloride.64 X xx + (92) R' = H R' = X = C02Me (94)C = K C5 H,N (93) R2 = H R' = X = C0,Me (95) = (96) R = C0,Et ; X = C0,Me x' R (97) R = CO,Et X = C0,Me (98)R = CO2.CH,-C6H,-Br.X = COzMe 62 R.E. Banks A. C. Harrison R. N. Haszeldine and K. G. Orrell J. Chem. SOC.(C) 1967 1608. 63 R. C. Kerber and M. J. Chick J. Org. Chem. 1967 32 1329. 64 0.W. Webster J. Org. Chem. 1967,32 39. L* 324 K.M ackenzie X X H (100) X = C0,Me Diels' condensation products of dimethyl acetylenedicarboxylate with methyl malonate (acetic acid-pyridine) are (92) and (93) for the high and low m.p. tautomeric constituents as earlier suggested. Both compounds give (94) with potassium acetate which liberates the strong acid (95) on acidification (7 5.8; O.O1N-solution pH2). The acid reacts with chlorine and bromine to give powerful halogenating agents whilst the reactions with tetracyanoethylene and maleic anhydride appear to give charge-transfer complexes.In a similar reaction with cyanoacetic ester and the acetylenic compound acyclic diene salt (96) is formed; this easily loses methanol to give blue cyclic oxyanion (97). The analogous (98) adds ethanol reversibly to give the colourless (99). Acidifica- tion of (97) gives a colourless dimer which is not however the expected cyclo- pentadienone dimer since the i.r. carbonyl frequency is too low. Compound (92) is probably converted into the cyclopentadiene compounds by 1,4- transannular Michael addition [(92) + (loo)] and extrustion of trimethyl- ethylenetricarboxylate ;in support of this heptaphenylcycloheptatrienetreated with potassium gives pentapheny lcyclopen tadienide.' A novel route to a cyclopentadienone is the ring contraction of 6-acetoxy- 2,4,6-triphenylcyclohexadienoneunder mildly basic conditions to give 2-benzoyl-3,5-diphenylcyclopentadienone dimer which is thermally dissociable and is converted into monomer adducts.Benzoyldiphenylcyclopentadienolcan be isolated under carefully controlled conditions.66" Other cyclopentadienone dimers have also been discussed and their structures clarified.66b In bicyclo[2,1,0]pentane chemistry routes to novel bridgehead ethoxy- carbonyl compounds have been explored; the best routes appear to be n-amylsodium metallation and carbonation to the 1-carboxylic acid and pyrolysis or preferably photolysis of pyrazoline adducts of diazomethane with l-ethoxycarbonylcyclobutene.67The photolytic method starting from 1-ethoxycarbonyl-3,3-dimethylcyclobutenegives a high yield of l-ethoxy-carbonyl-3,3-dimethylbicyclopentane;thermolysis of the intermediate pyrazo- line follows a different course to give l-ethoxycarbonyl-2,3,3-trimethylcyclo-butene and the tautomeric exocyclicmethylene compound.Thermolysis of the bicyclopentane compound obtained here gives an almost quantitative yield of dimethyl ethoxycarbonylcyclopentenes whilst hydride reduction of the ester 65 R. C. Cookson J. B. Henstock J. Hudec and B. R. D. Whitear,J. Chem. SOC.(C),1967,1986. 66 (a) K. Dimroth H. Perst and Kart Hans Muller Ber. 1967 100 1850; (b)M. Elliott S. H. Harper and M. A. Kazi J.Sci. Food Agric. 1967 18 167. " P. G. Gassman and K. T. Mansfield J. Org. Chem. 1967,32,915. Alicyclic Compounds 325 gives the expected bicyclo[2,l,0]pentylcarbinol which rearranges rapidly in the presence of acids to 1 ,l-dimethyl-4-methylenecyclopentanol, unlike the bridgehead carboxylic ester which is quite stable under acidic conditions even on warming6* The stereochemistry of deazetation of 2,3-diazanorbornenes as a source of bicyclo[2,1,0]pentanes has been studied ; the elimination of nitrogen occurs largely with inversion of the bridge-methylene group; uiz. (101) gives the anti-2,3-dideuteriobicyclo[2,1,O]pentane. This result suggested that the addi- tion of suitably reactive azo-compounds to bicyclopentanes might also occur with inversion of the methylene group which eventually forms the bridge; indeed reaction of the bicyclopentanes (102) and (103) in a mixture of known composition with N-phenyltriazoline-2,4-dionegives the products (104) and (105) with complete stereospecificity and inversion of the spiromethylene group.It is suggested that the decomposition and addition reactions are two step processes e.g. (101) gives (106) which ring-closes as indicated in (107) with inversion of the methylene group.690 I R'&J D N R' (101) (102) R' = D RZ = H (103) R' = H. R' = D R' R1y &+ ff. &$ I IN (104) R' = D R' = H (106) (107) i 11 (106) R' = H. R' = D N A novel route to bicyclo[ l,l,l]pentanes is the formation of 2-hydroxy-2- phenylbicyclo[ l,l,l]pentane-in the irradiation of cyclobutyl phenyl ketone by 1,5-hydrogen transfer to carbonyl oxygen formation of a 1,4-diradical and ring closure.69b Possible stereospecificity in the reaction of cyclohexanes stimulates much experimental work ;sulphur halides in pyridine convert cyclohexene oxide into cis-1,2-dichlorocyclohexane(99.5"/ stereoselectivity); earlier methods give trans-contaminated product.70 The epoxide is converted into the cis-diazide 68 J.H. Kinstle R. L. Welch and R. W. Exley J. Amer. Chem. SOC. 1967,89 3661. 69 (a) W. R. Roth and M. Martin Tetrahedron Letters 1967 4695; see however E. L. Allred and R. L. Smith J. Amer. Chem. SOC.,1967,89,7133;(b)A. Padwa and E. Alexander J. -4rner. Chem. SOC.,1967,8!3,6376. 70 J. R.Campbell J. K. N. Jones and S. Wolfe Conad. J. Chem. 1966,44 2339. 326 K.Mackenzie by aide ion by further reaction of the mesylate of the initially formed trans-hydroxy-azide ; catalytic reduction of the bis-azide gives the previously described cis-1,2-diamino-~ompound.~~ Catalytic reduction of aromatic compounds as a source of di-t-butylcyclohexanes generally gives the cis-isomers ;truns,truns-l,4-di-t-butylcyclohexan-2-01 is accessible via low tempera- ture hydroboration hydrolysis and oxidation of 1,4-di-t-butylcyclohexene. The possibility of replacing the boron group by halogeno- hydroxy- or amino-groups at an intermediate stage provides a general route to this group of corn pound^.^^ In the perfluorocyclohexane series examples of syn-clinal as opposed to anti-periplanar base elimination reactions are apparent from e.g.(108),which gives 1-chloroperfluorocyclohexene,but also 14% decafluorocyclohexene ; (109A)gives 18% of 1-chloroperfluorocyclohexene as well as the more normal product. It seems possible that these eliminations involve either the boat conformers or follow an Elcb pathway but whilst (108)is readily deuteriated in base (109B)is not.730 X An interesting photochemical transformation in this field is the formation of 1-trifluoromethyloctafluorocyclopenteneand the tautomeric difluoro-methylenecyclopentane from decafluorocyclohexene.7 3b The Claisen rearrangement of 2-vinyl-2,3-dihydropyranderivatives offers promise as a route to the less accessible isomers from Diels-Alder syntheses e.g.the transformations of (1 lo) (1 12) (1 14) and (116).Where stereoisomeric vinyl groups are present the stereo-relationships are retained in the product ; as expected from models the cis-isomer of (116)is much more stable owing to crowding in the transition state and the reaction is less ~tereospecific.~~ Diverse rearrangements of cyclohexadienones continue to excite much interest. Since the presence of alkyl groups at C-2 and/or C-5 in these dienones appears to effect critically the photochemical reaction pathway examples of pentamethylcyclohexadienones with vacant ring sites have been made e.g. (1 18)-(120),by use of the recently developed boron trifluoride-trifluoroacetic acid method with pentamethylbenzene. The preponderance of (1 18) and (119) 71 G.Swift and D. Swern J. Org. Chem. 1967,32,511. ’* D. J. Pasto and F. M. Klein Tetrahedron Letters 1967,963. 73 (a) S. F. Campbell F. Lancashire R Stephens and J. C. Tatlow Tetrahedron 1967 23 4435; (b)G. Gamaggi and F. Gozzo Chem. Comm. 1967,236. l4 G. Buchi and J. E. Powell jun. J. Amer. Chem. SOC.,1967,89,4559. Alicyclic Compounds in the oxidation products is due to involvement of (121) in which 1,2-methyl migration can occur equally well in either direction before proton loss to give the product. Pentamethylhexa-3,5-dienoicester is formed in the photolysis of (1 18) only in methanol together with bicyclic (122) whereas (120) gives only the dienoic ester under these conditions; (1 19) is photolysed both in ether and in methanol to give thermally labile (123).Both reactions are faster in methanol; bicycli- sation is the more rapid. Whilst the multiplicity of states involved here is uncertain the lack of any effect by methyl at C-5 on the reaction rate suggests that if charged dipolar intermediates are involved bicyclic zwitterion (124) is more likely than diene (125); the bicyclic products are stable under the reaction conditions and postulatidn of a keten dienoic ester precursor is untenable in the solvents used.75 Me (118) R' = R3 = Me R2 =H (121) (119) R2 = R3 = Me R' =H (120) R' = R2 = Me. R3 =H MeeMe Me Me \ (124) (1 25) Formation of 1,3,4,5,6,6-hexamethylbicyclo[3,1,0] hexen-2-one from hexa- methylcyclohexadienone in ether during photolysis was earlier shown to involve bond-crossing.More rapid photolysis occurs in ethanol ;the primarily '' P. M. Collins and H. Hart J. Chem. SOC. (C) 1967,895; H. Hart and D. W. Swatton J. Amer. Chem SOC. 1967,89 1874. 328 K.M ackenzie produced bicyclohexenone reacts with the solvent to give 3-ethoxy-l,3,4,5,6,6- hexamethylcyclohexa-1,4-dien-2-ol. When warmed this eliminates ethanol to give 1,4,5,6,6-pentamethyl-3-methylenecyclohexa-1,4-dien-2-ol (also by protic catalysis). The methylene-dienol is thermally isomerised to the 2-keto-tautomer but acid catalysis yields 1,2,4,5,6,6-hexamet hylcyclohexa-2,4-dien-3-one, appar-ently by 1,2-methyl shift and ketonisation. These transformations are confirmed with 2,4,6,6-tetramethyl-3,5-bistrideuteriomethylcyclohexadienone as starting material.75 Steric inhibition of migration to a carbon atom already bearing a t-butyl group prompted further experiments with e.g.(126); its rearrangement to (128) appears to involve as precursors the bicyclic compounds (127) ('isomers at C-6) which are in fact isolable and are converted on further photolysis by an unusual exocyclic cyclopropane ring scission into (128); formation of intermediate (129) is sterically facilitated by the gem-substituted methylene and the bridgehead t-butyl group; acyl bond switch and olefin formation complete the process according to one interpretation.76 More recently photoisomerisation of 2,4,6-trit-butyl-4-methoxycyclohexadienone to analo- gous bicyclohexenones in tungsten light has been described ;further transfor- mations at shorter wavelengths here give aromatic compounds or their keto- tautomers it is supposed from zwitterionic excited species derived by endo-cyclic 1,5-ring scission which suffer 1,2-t-butyl shifts [e.g.(130)].77 Photolysis of the tri-t-butylmethoxycyclohexadienonein aqueous acetic acid-methanol however gives stereoisomeric 2,5-di-t-butyl-3-trimethylacetylcyclopent-2-enones by hydrolytic ring scission of the protonated zwitterionic bicyclic R" R3&; (126) R = Pr" (132) R = n-C,H 0 OH (128) R = Pr" (129) (133) R = n-C,H 76 B. Miller and H. Margulies J. Amer. Chem. SOC.,1967,89 1678. T. Matsuura and K. Ogura J. Amer. Chem. SOC..1967.89,3846. 7' Alicyclic Compounds (130) R'T~ = Bu' R' = OMe R4 = H (134) R' = Me R2 = But R3 = n-C3H, R4 = H (137) R' = Me R' = Bu' R3= n-C,H, R4 = H (138) R' = But.R' = H.R3 = n-C,H,. R4 = Me intermediates first formed [cf. (131)l. Cleavage of the ring a-to both carbonyl groups and 1,3-bond switch on to the double bond gives bicyclo[2,1,0]pentan- 2-0nes.~~ A further stable keto tautomer is the compound (134) derived by heating the photoisomer (133) of the dienone (132); this compound is stable below loo" in the absence of acids or bases and is catalytically hydrogenated at the allyl group! Models of the dienone show that the allyl group can assume an orientation well away from the t-butyl and t-butyl-methyl steric conflict is reduced in comparison with the phenolic structure in which the allyl and methyl groups are compressed against the t-butyl group between them.79 Cyclohexadienone (137) appears to be involved in the photoisomerisation of syn anti isomeric allyldi-t-butylmethylbicyclo[3,1,0]hexenones (135) and (1 36) to syn-6-allyl-1,3-di-t-butyl-5-methylbicyclo[3,1,0]hexen-2-one; no aromatic compounds are formed! Irradiation of (138) gives an analogous product with a trans-but-2-enyl group at C-6 which clearly indicates 1,2-shift without inversion after the initial endocyclic bond-scission in the 6-alkyl-6-alkenyl- bicyclohexenones which leads to the cyclohexadienones e.g.(137) as inter- mediates. In contrast to reactions of bicyclo[2,1,0]pentanes and bicyclo[ l,l,O]butanes bridgehead chlorocarbon is formed from photochemical chlorination of bicyclo[2,2,0]hexane.* Unexpectedly the Oppenauer oxidation of exo-2- bicyclo[2,2,0]hexano1 in the presence of quinone as hydride acceptor gives bicyclo[2,l,l]hexan-5-one.The reaction resembles a similar reaction of quadricyclanol ; without the quinone insufficient bicyclo[2,l,l]hexan-5-01 is formed to allow of a common intermediate precursor for this compound and the 'ketone which does not therefore arise from a C(l)-C(2)-C(6) delocalised ion but possibly from a C(l)-C(2jC(4) ion derived by hydride abstraction from the bicyclohexyloxyaluminium compound.82 Diels-Alder reaction of the (2 + 2)n adduct of 1,l dichlorodifluoroethylene and cyclo-octatetraene with dimethyl acetylenedicarboxylate and thermolysis of the product gives 2,2-dichloro-3,3-difluorobicyclo[2,2,O]hex-5-ene ; this " T.Matsuura and K. Ogura J. Amer. Chern. SOC. 1967,89,3850. l9 B. Miller J. Amer. Chern. SOC.,1967,89 1685. B. Miller J. Amer. Chern. SOC. 1967,89 1690. 81 R. Srinivasan and F. I. Sonntag J. Amer. Chem. SOC. 1967 89 407; R. Snnivasan and F. 1. Sonntag Tetrahedron Letters 1967 603. R. N. McDonald and C. E. Reineke J. Org. Chern. 1967,32,1888. 330 K.M ackenzie gives explosive 2-chloro-3-fluorobicyclo[2,2,0]hexa-2,5-diene on treatment with lithium methyl. The Dewar benzene has a half-life of ca. 3 weeks in solution at 25°.83 Photodecarbonylations in bridged-ring compounds are not so well known as in larger cycloalkanones ; mercury-photosensitised decarbonylation of bicyclo[2,l,l]hexan-3-one to bicyclo[ l,l,l]pentane is therefore of interest.Other reaction products probably arise from the 1,4-pentadiene concomitantly formed.84 Novel syntheses of the bicyclo[3,1,0]hexane ring system utilise (a) the reaction of 1 -chlorocyclohexanone with piperidine to give 6,6-dipiperidino- bicyclo[3,1,0] hexane by intramolecular bond-closure dechlorination in the initially formed enamine giving the quaternary iminium compound which then reacts with more piperidine,” and (b)Grignard addition to sym-trinitro- benzene bromination to isomers of 1,3,5-tribromo-2,4,6-trimethyl-1,3,5-trinitrocyclohexane and ring closure to (139) with iodide ion or (140) with borohydride.86 R1 (141) R’ = alk/aryl. (142) R’ = alk/aryl R2 = OMe R3.4 = H (143) R3 = OMe R’s4 = H R’ = alk/aryl (1441 R’.’ = H.R4 = OMe. R3 = alk/aryl Photoisomerisation of tropolone methyl ether to the bicyclo[3,2,0] hepta- diene series (142) and thence to the isomeric series (143) is well known. Thermolysis of (142) regenerates the tropolone but at lower temperatures compounds of the (144) series are also formed by a previously unobserved Cope-type rearrangement.87 Thermal tropilidene rearrangements have al- ready received detailed attention whereas photochemical reactions have not although cases of 1,7-sigmatropic shifts are known. The primary photochemical products from 3,7,7-trimethyltropolidene appear to be the 2,6,7-trimethyl 83 G. Schroder and T. Martini Angew. Chem. Internat. Edn. 1967,6,806. 84 J. Meinwald W.Szkrybalo and D. R. Dimmel Tetrahedron Letters 1967 731. 85 J. Szmuskovicz E. Cerda M. F. Grostic and J. F. Zieserl jun. Tetrahedron Letters 1967 396 9. 86 T. Severin and M. Bohn Chem. Ber. 1967,100,2532. *’ T. Miyashi M. Nitta and T. Mukai Tetrahedron Letters 1967 3433. Alicyclic Compounds 331 isomer and 2,2,6-trimethylbicyclo[3,2,0]heptadiene,formed by 1,7-sigmatropic methyl shift and cyclisation respectively. It is incidentally considered that for the first excited state of a linear heptatriene radical the symmetry-allowed migrations are not those predicted by precise reversal of the Woodward- Hoffmann selection rules but rather 1,3-antarafacial 1,5-antarafacial and 1,7-suprafacial.* Thermal rearrangement of 2,5,7-triphenylnorcaradieneyields 1,3,5-triphenyl- tropilidene presumably by 1,5-sigmatropic shift in the cycloheptatriene in equilibrium with it ; further suprafacial 1,Shydrogen shift gives the 1,3,6- triphenyl isomer ;photochemical rearrangement of the norcaradiene however proceeds reversibly to the latter compound maybe by disrotatory ring-opening to the 2,5,7-triphenyltropilideneand 1 ,7-suprafacial hydrogen transfer.The reversal of the sequence consists of photochemical 1,7-hydrogen shift to the 2,5,7-triphenyltropilideneand thermal cyclisation to the n~rcaradiene.~~ Examples of sigmatropic hydrogen shift are also seen in the formation of cyclo-octen-4-one from cyclo-octa- 1,3-dien-5-01 present as rearrangement product in hium diethylamide treatment of 5,6-epoxycyclo-octene90 and the generation of cis,cis-3-oxacyclonona-1,4-diene from pyrolysis of the 3,4-epoxy- compounds an 'oxahomodienyl' hydrogen migration [(145) + (146)l." Low temperature n.m.r.studies indicate rapid pseudo-rotation of groups between 'above' and 'below' positions in cycloheptane even at -150" in the absence of steric effects ;however 4,5-trans-dibromo-l,l-difluorocycloheptane exhibits a single 9F resonance line above -114" but below -118" two lines appear corresponding to two conformers each with equivalent fluorines but in which the two bromines are either staggered or gauche.92 The cis,trans assignment to cyclo-octa-1,5-diene prepared by Willstatter's method from NN-dimethyl-4-cyclo-octenylaminehas been confirmed; earlier Raman spectra did not resolve the two lines now observed (1622 and 1635 cm.-') indicative of two different double bonds.The finding is conformed by double- resonance n.m.r. studies.93 In the expectation of achieving transannular cyclisation with cations derived from 1,5-cyclo-octadiene the olefin was treated with methyl methoxyacetate- L. B. Jones and V. K. Jones J. Amer. Chem. SOC.,1967,89 1880. 89 T. Mukai H. Kubota and T. Toda Tetrahedron Letters 1967,3581. J. K. Crandall and Luan-Ho Chang J. Org. Chem. 1967,32,532. 91 J. K. Crandall and R. J. Watkins Tetrahedron Letters 1967 1717. 92 R. Knorr C. Ganter and J. D. Roberts Angew. Chem. Internat. Edn. 1967,6 556. 93 A. C. Cope C. F. Howell J. Bowers R. C. Lord and G. M. Whitesides J. Amer. Chem. SOC. 1967,89,4024; see however Ann.Reports 1966 p. 433. 332 K.M ackenzie boron trifluoride ; the products indicate a preference for outside exo-attack leading to cis-6-endo-or exo-acetoxy-endo-2-methoxymethylbicyclo[3,3,0]-octanes in a non-concerted pathway since introduction of the acetoxy-group lacks stereoselectivity. Products of proton elimination and 8-acetoxy-endo- methoxymethylbicyclo[3,2,l]octanes are also formed. Interestingly analogous monoacetate by-products arise in catalysed reactions of the diene with diacetoxymethane presumably by protonation of the cyclo-octadiene by cationic intermediates involved in the main ~equence.’~ Aluminium chloride- catalysed addition of acetyl chloride to the diene also gives cyclised products the e.g. cis-6-chloro-2-exo-acetylbicyclo[3,3,0]o~tane;~~ formation of cis-bicyclo[3,3,0]oct-2-ene from the octadiene with potassium hydride is also rep~rted.’~ The probable involvement of homotropylium ions in the low temperature halogenation of cyclo-octatetraene suggested examination of the reaction of cis-7,8-dich~rocyclo-octatrienewith antimony pentachloride; the exo-8-chlorohomotropilium salt (147; X == SbC1,) is indeed formed from the exo,cis-dichloro-conformerat -15”.In contrast it is the endo,cis-dichloro- conformer which forms endo-8-chlorohomotropiliumfluorosulphonate (148 ; X == FSO,) (< 0); this has only a half-life of minutes at ca. 30”and rearrange- ment gives the 8-exo-chloro-salt (147; X = FS03). The effect of the ring current in the cation can be seen from the different chemical shifts for the 8-protons ‘T 8-2 in the em-chloro-ion and 2-51 for the endo-chloro-species.For the chlorination of cyclo-octatraene the intermediate (148; X = C1) is proposed since addition of tetraethylammonium chloride to (147; X = SbCl,) gives only the trans-7,8-dichloro-compound (n.m.r. signals in sulphur dioxide at -40”) whilst similar treatment of (148; X = FSO,) gives 94 % cis-7,8-dichlorocyclo-octatriene and 6 % trans-isomer. Hence endo-attack for both homotropilium ions appears to be kinetically preferred. The initial endo- chlorohomotropilium ion is probably formed from a complex of a chlorine molecule with the tub-like cyclo-octatetraene. Models indicate that C(1)-C(7) orbital overlap in the 8-halogenohomotropylium ion is substantially larger on the exo-side of the methylene bridge; this may be the reason for the preferred endo-reaction with anion.However in the low-temperature catalysed isomeri- sation of the cis-and trans-dichlorides since ring inversion is known not to occur at the temperatures used ionisation and recombination of the two dichlorides is not completely stereo~pecific.’~ Methylation of cyclo-octatetra- ene dianion with methyl iodide involves a slow first-step monomethylation.98 Dehydrocyclo-octatetraene occurs in the products of reaction of the bromo- tetraene with strong bases ;it reacts with dienes e.g. tetracyclone in the expected 94 I. Tobushi K. Fujita and R. Oda Tetrahedron Letters 1967,3815 3755. ’’ T. S. Cantrell J. Org. Chem. 1967,32 1669.96 L. H. Slough J. Org. Chem. 1967,32 108. 97 G. Boche W. Hechtl H. Huber and R Huisgen .I.Amer. Chem SOC.,1967,89,3344; R. Huisgen G. Boche and H. Huber ibid. 3345. 98 D. A. Bak and K. Conrow J. Org. Chem. 1966,31,3958. Alicyclic Compounds (147) ex0 R' = H R2 = C1 (148) endo R' = C1 R2 = H (149) manner and forms a vinyl ether by addition of t-butyl alcoh~l.~~ The chemistry of 'simpler' cyclo-octatetraene compounds is receiving increasing attention ; the hydrogen bromide adduct of the tetraene is a valence tautomeric mixture of mono- and (30%) bi-cyclic forms which is substituted by azide to give a mixture (70% bicyclic) which decomposes rather readily to 2-trans-butadienyl- pyrrole by way of nitrene (149) and prototropy of (150).'oo Bromocyclo- octatriene is convertible into the corresponding alcohol and acetate which are largely bicyclic.Rapid ring-chain tautomerism of the alcohol and analogous carbinols obtained from the derived ketone takes place (the l-phenyl- carbinol is not isolable). The appearance of the cis,cis,trans-structure (1 54) supports ring opening of the bicyclic form from (151d) as the monocyclic form might be expected to give an all-cis non-conjugated ketone by ring scission. Terminal cis-isomers of (154) give all-trans non-conjugated ketones by 1,3-prototropy. These results suggest that the known photolysis of cyclo- octatrienone in methanol to give methyl octatrienoate might involve a hemi-acetal of bicyclo[4,2,0]octa-2,4-dien-6-one rather than the keten previously suggested e.g.(155) -+(156)."' O -R = aR X X (151) (a) R = H,X = Br (153)X = OH (155) X = OH R = OMe (b) R = H X = OH (c) R = Me,X = OH (d) R = Ph X = OH R (154) R = alk/aryl (156) R = OMe 99 A. Krebs and D. Byrd Annalen 1967,707,66. loo M. Kroner Chem. Ber. 1967 100,3162. M. Kroner Chem. Ber. 1967,100,3172. 334 K.M ackenzie Catalytic cyclisations have been further developed 1,l-bischloromethyl-ethylene (157) is converted into 1,4,7-trimethylenecyclononane (158) by nickel carbonyl ;cyclisation of 1,9-dichloro-2,5,8-trimethylenenonane with the reagent also affords (158). The reaction of (1 57) with 1,6-dichloro-2,5-dimethyl-enehexane (159) also gives (158),and studies with labelled (159) indicate that (157) reacts with it much faster than (157) combines with itself.lo2 Nickel catalysis allows synthesis of quite large rings e.g.reaction of butadiene with cyclodecyne gives (160) which on reduction to (161) oxidation and Wolff- Kishner reaction gives cycloeicosane. Similar steps can be carried through with 1,8-~yclotetradecadiyne.~~~ Cyclic cumulenes are not well known but they can be made from the dibromocarbene adducts of cyclic allenes which when treated with lithium methyl ring-expand to the cumulene e.g. cyclonona- 1,Zdiene gives (162); this is stable at low temperature is reduced to cyclodeca- 1,3-diene with sodium- ammonia and absorbs iodine to give 2,3-di-iodocyclodeca- 1,3-diene. lo4 Improved methods for the synthesis of cycloalkadiynes from ao-dibromo- alkanes with sodioacetylene have been reported and previously inaccessible members described; the reduction-prototropic rearrangement of the derived di-cis-olefins is also discussed.Geometrical factors are more important for the smaller rings e.g. difficultly formed conjugated olefins lack coplanarity and are reduced slowly. O5 Chlorination and dehydrohalogenation of cyclodecane gives pure cis-and trans-cyclodecenes depending on the base used. lo6 The sequence (163)-1166) illustrates a new fragmentation applicable to the synthesis of medium and large rings e.g. (168) from (167) in 85 % yield.lo7 E. J. Corey and M. F. Semmelhack Tetrahedron Letters 1966,6237. Io3 P. Heimbach and W. Brenner Angew. Chem. Internat.Edn. 1966,5,961. W. R. Moore and J. M. Ozretich Tetrahedron Letters 1967 3205. A. J. Hubert J. Chem. SOC.(C),1967,2149. J. G. Traynham D. B. Stone and J. L. Couvillion J. Org. Chem. 1967,32 510. lo’ A. Eschenmoser D. Felix and G. Ohloff Helv. Chim. Acta 1967 SO 708; cf. M. Yanobe D. F. Grove and R. L. Dehu Tetrahedron Letters 1967 3943. Alicy cl ic Compounds (163) (164)X = 0 (165) X = =N.NH.SO,Ar Bridged and Caged Structures(en, n 3 7).-Reports have appeared recently on the synthesis of stereomechanistically interesting em-and endo-mono- methylene adducts of norbornenes (and dienes) reaction of 7-norbornadienyl benzoate with diazomethane gives a 5 :1mixture of em- and endo-syn-tricyclo- octenyl benzoates. lo* Corresponding bridge-carbonyl compounds have also been made but an alternative pathway is the synthesis of (169) via cyclo-propene-tetrachlorodimethoxycyclopentadieneadduct by dehalogenation and hydrolysis.log The stability of these compounds depends critically on the orientation of the cyclopropane ring and to a lesser extent on the substituents in the norbornene ring (169) has a half-life of ca. 90 min. at 35",whereas the em-isomer is comparably stable at 150". It seems possible that disrotatory cyclopropane ring-opening can assist bridge-elimination leading to the boat conformer of tropilidene directly from (169) whereas the em-methylene isomer must first give the norcaradiene. Fragmentation reactions of 7,7-disubstituted norbornadienes potentially pass through norcaradienes and this is seemingly supported by the rearrange- ment of 7-alkoxy- and 7-aryl-norbornadienes to equilibrated tropilidenes.However whilst 7,7-dialkoxynorbornadienesdecompose particularly easily the isomeric geminal dialkoxytropilidenes decompose only at much higher temperatures;the products are solvent dependent and consistent with elimina- tion of dialkoxycarbene as one pathway and methoxy-radical elimination (leading to tropone) as another.' ' The photochemical ring-chain rearrangement of cycloalkanones to alkenals has been postulated to involve intramolecular hydrogen transfer from the P-position during ring scission ;compatible with this the rearrangements of bridged-bicyclic ketones fall into two rate groups depending on whether or lo' B.Halton M. A. Battiste R. Rehberg C. L. Deyrup and M. E. Brennan J. .4mer. Chem. SOC. 1967,89,5864. lo9 S. C. Clarke and B. L. Johnson Tetrahedron Letters 1967 617. 'lo R. W. Hoffmann and J. Schneider Tetrahedron Letters 1967 4347; cf. G. Maier Angew. Chem. Internat. Edn. 1967,402. 336 K.Mackenzie not geometrical factors are suitable for 1,3-hydrogen transfer e.g. (170) belongs to the faster rate group. ''' Other photochemical studies of bridged bicyclic ketones include fragmentation to bicyclo[2,1,1] hexane and bridge-switching reactions of e.g. bicyclo[3,2,l]oct-2-en-8-one.''2uv ' ' 2b Me&Me 0 Me Me exo-Addition of hydrogen in di-imide reduction of norbornadienes syn to a bridge substituent is remarkable ;calculations based on steric and entropy factors suggest that the anti-double-bond ought to be favoured to the extent of 96 %! The 99 1 predominance of the syn-addition process points to a stabilisation of the transition state of ca.4.5 kcal./mole. '' exo-syn-Addition is also favoured for epoxidation and methylenation of 7-t-butoxynorborn- adiene but phenyl azide addition prefers endo-syn-approach.' 14' It is suggested that there may be polar interactions in these alkoxy-bridge compounds ; evidence for this is the higher reactivity of t-butoxynorbornadiene than norbornadiene towards lithium alkyls and the reaction of 7-syn-t-butoxy- norbornene with isopropyl-lithium under conditions leaving the anti-isomer and norbornene virtually unchanged. The well known tendency of organo-metallics to complex with ether oxygen groups may well account for these observations.The preference for endo-attack by dipolar phenyl azide at the syn-double-bond might be due to the proximity of the electron-rich oxygen atom which initiates electrophilic attack; other reactions which can be discussed in similar terms are the epoxidation of tropilidene and cyclo-octa- tetraene maleic anhydride adducts from the same side as the.anhydride ring. These effects may be much less important with very reactive reagents e.g. benzenesulphonyl azide.' 14a The preferred exo- mode of reaction in norbornyl compounds has been generally explained in terms of torsional effects between the bridgehead hydrogen and the C-2 substituent (usually hydrogen) at the point of attack ;endo-approach and bonding requires these groups to undergo eclipsing interaction absent in the case of em-attack.' 14' From the same laboratory as the hydroboration technique for anti-Markownikov hydration comes a simple technique for stereoselective synthesis of exo-norbornan-2-01 by oxymercuration-borohydride reaction.With 2- methylnorbornene endo-2-methylnorbornan-2-01 is formed ; the method complements the hydroboration technique. The absence of rearrangement ''I T. Matsui Tetrahedron Letters 1967 3761. (a) J. E. Baldwin and J. E. Gano Tetruhedron Letters 1967 2099; (b) W. F. Erman and H. C. Kretschmar J. Amer. Chem. SOC.,1967,89 3842. W. C. Baird jun. B. Franzus and J. H. Surridge J. Amer. Chem. SOC.,1967,89,410. 11' (a) G. W. Klumpp A.H. Veefkind W. L. de Graaf and F. Bickelhaupt Annulen 1967 706 47 57; (b) P. von R.Schleyer J.Amer. Chem SOC.,1967,89 701. Alicyclic Compounds products from methylated norbornenes and similar trideuteriomethyl com-pounds appears to preclude bridged-ion intermediates in these reactions.' l5 Only exo-methylene adducts are formed from pure methyl-lithium and methyl- ene chloride in reaction with norbornenes contrary to an earlier report,"6 although the smooth dkrotatory cyclopropyl-ally1 rearrangement of the syn-chloromethylene compound is confirmed. The major and minor products from the rhodium-catalysed dimerisation of norbornadiene (among other dimers) are related by hydroiodination-Wagner-Meerwein rearrangement and de- hydrohalogenation of the minor isomer which gives the major isomer; since the minor product is believed to be (172) the major reaction product is (173).Besides the rearranged hydroiodide the minor dimer gives a cage compound which is therefore (174) and not the analogous compound with parallel methylene bridges previously suggested. The bis-nortricyclane structure earlier proposed for a further dimer is now thought to be (175) rather than the bis-cis-endo-compound. With these assignments all the catalytic dimers of norbornadiene can be rationalised by the formation of one bond between two molecules on the catalyst surface with a subsequent product-forming step in the resulting di-radical; they can also be regarded as related to the exo-trans- exo- exo-trans-endo- and endo-trans-endo-2,3 :2',3'-dimers by scission-rota- tion-recombination steps e.g.the latter dimer gives (1 76) and hence (172) via (172').''' The dimer (172) or its stereoisomers are converted over alumina at 300" into (174)"* whilst purely thermal treatment affords (222) (R' == R2= H) in contrast to the endo-trans-exo-2,3 :2',3'-dimer which gives bicyclo- [4,2,1]nona-2,4,7-trieneand its 2,5-cyclised valence tautomer [exo-cyclobutene ring cf. (177)l. Sensitized irradiation of the bicyclononatriene gives the interesting tautomer (1 77) as well as the exo-isomer.' '' 'I5 H. C. Brown and W. J. Hamnar J. Amer. Chem. SOC. 1967 89 1524; H. C. Brown J. H. Kawakami and S. Ikegami J. Amer. Chem. SOC.,1967,89,1525. C. W. Jefford E. H. Yen and R. Medary Tetrahedron Letters 1966,6317 '" T.J. Katz and N. Acton Tetrahedron Letters 1967 2601. H.-D. Scharf G. Weisgerler and H. Hover Tetrahedron Letters 1967,4227. L. G.Cannell Tetrahedron Letters 1967 5967. 338 K.M ackenzie Novel 1 -perfluorobicycloheptyl Grignard reagents form from the 1 -halo- geno-compounds accessible from 1-undecafluorobicycloheptyl-lithium ; these reagents thermally eliminate magnesium fluorohalide to yield perfluoro-l- halogenonorbornan-2-enes seemingly via transient bridgehead olefins which add halide and expel fluoride anion from C-3; further reaction with magnesium can ensue to give bicyclohept-Zen- 1-yl Grignards. The analogous 1,4-dihalo- genoperfluoronorbornanes give bis-Grignards which are actually more stable than the mono-compounds since the latter are subject to an unfavourable 1,4-transannular interaction with the C-F dipole.12' Woodward and Katz observed Oppenauer oxidation of hydroxy-cyclo- pentadiene dimer (178) to give only the Cope rearrangement product (179) of the expected ketone; either rearrangement is very fast or catalysis is involved.Chromic oxide-pyridine oxidation of (178) does give the bridged ketone k3 (178) R'. = OH R2-4= H (I 79) R3'v4 = 0,R' = R2 = H Ph CI c1 (182) however; its rearrangement at the m.p. is very fast and it is catalysed by traces of protic or Lewis acids at lower temperatures. Thermolysis of the 8-keto- compound gives more dihydroindene than that of the 7-keto-compound which suggests that decarbonylation is faster than rearrangement.Ketone (179) photochemically cages endothermically ; the product cleaves at 450" whereas the cage product from irradiation of benzoquinone-cyclopentadiene adduct is stable above 500"!'21 An intermediate such as (180) formed from a 1,2,3,4-tetrasubstituted norbornen-7-one analogue of (178) could cyclise in a number of ways; indeed hydrolysis of acetal(l8 1) gives mainly (1 83)(vmax.1745 cm. -I) via an analogous intermediate (182). Photolysis of (183) gives the em-5-phenyl isomer of the S. F. Campbell J. M. Leach R. Stephens and J. C. Tatlow Tetrahedron Letters 1967 4269. R. C Cookson J. Hudec and R. 0.Williams J. Chem. SOC.(C) 1967 1382. Alikyclic Compounds 339 minor hydrolysis product of the acetal-the expected 7-ketone-through diradical(l84).The 7-carbonyl compound is not an intermediate for rearrange- ment of (181) to (183) since it rearranges more slowly than the acetal."*' Further rearrangements of 7-norbornadienyl cations have been elegantly followed by n.m.r. techniques ; these involve bridge-flipping and degenerate five-carbon scrambling by what are designated circumambulatory processes (see chapter 3 this Section).'22b The first examples of simple norcaradienes which probably exist by virtue of a positive enthalpy difference for the norcaradiene-tropilidene equilibrium rather than steric or other factors which preclude ring-opening as in earlier more complex examples are made by the decomposition of dicyanodiazo- methane in aromatic media; thus benzene gives (185) and p-xylene (186) and (187).The n.m.r. spectrum of (185) shows a complex multiplet T 3-2-3-9 quite different from that of the isomeric tropilidene which has three groups of signals z 3-5. The significantly larger dipole moment than that of malono- nitrile supports a charge-transfer explanation for the stability of these nor- caradienes but the diene U.V. absorption appears to preclude this.lZ3 Sigma- tropic 1,5-cyano-shifts are observed in thermolytic rearrangements of these compounds and Berson- Willcott rearrangements occur in norcaradiene taut- omers of 1,2-benzo-7,7-dicyanocycloheptatriene.1 24 '(185)R' = R3 = H '(186) R' = H R' = Me R2 (187) R' = Me R' = H A new synthesis of barrellene compounds comprises the unusual Diels- Alder addition of dicyanoacetylene to benzene; the reaction is catalysed by aluminium chloride which forms an isolable complex with the dienophile; separately formed the complex reacts with benzene to give the same product of 174-addition.The rates of reaction are increased by methylation of the ring which leads e.g. with xylene to isomeric 1,4- and 2,5-addu~ts.''~ A new synthesis of bicyclo[3,2,l]oct-2-ene is based on the catalytic activity of metal hydride-anhydrous cerous chloride which converts 4-vinylcyclo- hexene into the bicyclo-octene. '26 Autocondensation of acetone morpholino- enamine gives 2-[2-methylprop-l-enyl]-6,8,8-trimethylbicyclo[4,2,0]octen-2-one ; other derivatives of the bicyclo[4,2,0]octenone system are similarly accessible. Simple methods for the preparation of 1-halogeno-4-methyl- 122 (a) L.S. Besford. R. C. Cookson. and J. Cooper J. Chem. SOC. (C). 1967. 1385 (b) R. K. Lustgarten M. Brookhart and S. Winstein J. Amer. Chem. SOC.,1967 6350 6352 6354. 123 E. Ciganek J. Amer. Chem. SOC.,1967,89 1454. E.Ciganek J. Amer. Chem. SOC.,1967,89,1458. E.Ciganek Tetrahedron Letters 1967 3321. P.R.Stapp J. Org. Chem. 1966,31,4258. G. Bianchetti D. Pocar R. Stradi P. Dalla Croce and A. Vigevani Gazzetta 1967 97 872; G. Bianchetti P. Dalla Croce D. Pocar R. Stradi and G. G. Gallo ibid. p. 564. K. M ackenzie bicyclo[2,2,2]octanes are the reactions of the l-alcohols or methoxy-com- pounds with phosphoryl and sulphuryl halides in polyphosphoric acid128a or the reaction of the same starting materials with acyl halides and stannic chloride." 8b 1,4-Dihydroxybicyclo[2,2,2]octane made in a new synthesis based on tetrahydrobenzoquinonedienoldiacetate-maleicanhydride adduct is similarly converted into the dihalogeno-compound.In the field of charged bicyclic intermediates more details of base-catalysed deuterium exchange with bicyclo-octadiene (188) have been published. Other products of protonation of the proposed 67t non-classical carbanion (189) have been observed e.g. the tautomers tetracyclo[3,2,1,02~7,06~4]octane(190) and tricycl0[3,2,1,0'*~ Joct-3-ene (191). The anion itself prepared by cleavage with sodium-potassium alloy of the ether (188; R' = MeO) has also been observed.'30"*I3Ob A theoretical discussion of bicycloaromaticity has also appeared.Isomeric 3-chloro-8-azabicyclo[3,2,l~octanes (192) and (193) illustrate in their solvolytic behaviour the steric requirements of fragmentation reactions ; the latter fragments quantitatively with cyanide ion to give (194) whilst the former displaces chlorine probably via a cation or internal quaternary salt without fragmentati~n.'~~ Application of the orbital symmetry rules continues to serve structural elucidation; thus 2,3-homotropone (195) photochemically isomerises to (196) since the alternative mode of disrotatory ring closure is sterically hindered by the clash of methylene and vinyl protons.133 The past year has seen a number of interesting developments in bicyclononane chemistry. Addition of diphenyl- keten to olefins may involve non-symmetrical a-bond formation suggesting that cases of retro-reaction might also occur with selective scission of one of the bonds in the cyclobutanone ring; such is the case in the thermal rearrange- ment of 9,9-diphenylbicyclo[5,2,O]nona-3,5-diene-2,8-dione-the sole di-phenylketen adduct of tropone-which exists in equilibrium mainly with 12* (a)J.Kopecky and J. Smejkal Tetrahedron Letters 1967 1931 ;(b)Z. Suzuki and K. Morita J. Org. Chem. 1967,32 31. 129 J. Kopecky and J. Smejkal Tetrahedron Letters 1967 389. lJo(a)S. Winstein M. Ogliaruso M. Sakai and J. M. Nicholson J. Amer. Chem. Soc. 1967,89 3656; (b)J. M. Brown Chem. Comm. 1967,638. '" M. J. Goldstein J. Amer. Chem. SOC.,1967 89,6357. 132 A. T. Bottini C. A. Grob E. Schumacher and J.Zergenyi Helv. Chim Actu 1966 49 2516. IJ3L. A. Paquette and 0.Cox J. Amer. Chem. SOC.,1967,89 5633. Alicyclic Compounds 341 NC R 4Rl k2 (194) (192) R = H or Me R' = H R' = C1 (193) R = H or Me R2= €1 R' = C1 8,8-diphenyl-10-oxabicyclo[5,3,0]decan-9-one formed by C(7)-C(8) scission in the cyclobutanone ring closure of the acyl radcal on to ring carbonyl and subsequent 1,5-sigmatropic hydrogen transfer from C-7 to C-4. 134 The route to bicyclo[3,3,l]nonane compounds described by Stork and Landesman is by now well known; a similar reaction involving acryloyl chloride and 1-morpholinocyclohexene has also been described but the yields are low. The reaction has been further investigated; mixing the reactants at the b.p.gives the bicyclo[3,3,l]nonane-2,9-dione in good yield. Use of 3,333- tetramethylcyclohexanone enamine allows isolation of the intermediate 2,2,4,4- tetramet hyl-9-morpholinylidiniumbicyclo[3,3,1]nonan-8-one chloride (197) which when decomposed with water gives the tetramethylcyclononan- 2,9-dione and tetramethyloxocyclohexylpropiopicacid. Working with cinna- moyl chloride at low temperature allows isolation of the intermediate N-cinnamoyl chloride of an enamine whose decomposition to the keten (198a) [the type precursor of (197) via (198b)l in the presence of a different enamine demonstrates the intramolecularity of the reaction.' 35 (;I c1-c1-(197) (198a) (198b) 134 A. S. Kende Tecrohedron Letters 1967,2661. IJ5 P. W. Hickmott and J. R.Hargreaves Tetrahedron 1967,23,3151. 342 K.Mackenzie A 9 @ &OH --i 3 (200),(201) R 7 (199) R = H 1.r. evidence indicates significant C3-C7 transannular interactions in the bicyclo[3,3,l]nonane system but surprisingly few hydride transfer reactions in corresponding cations appear to have been described. Formolysis of exo-2,3-epoxybicyclo[3,3,l]nonane (199) however gives besides diol esters the em-6-or 7-en-2-01 (200) or (201) and a minor component of elimination from the initial cation exo-3-en-2-yl formate (202) which accords with the steric requirement for elimination of a proton from C-4 with a developing axial hydroxy-group at C-2; transannular hydride shift and elimination compete very well here.' 36 The complex hydride reduction of the enol lactone (203) was described a decade ago but only recently has the detailed stereochemistry of these re- actions been examined.The reduction gives preferentially the less thermo- dynamically stable axial epimer (204); the structure of the lesser component rests on the n.m.r. signals for the 2-proton and the smooth bridge cleavage with ethoxide ion which requires trans-coplanar arrangement of the C(9)-C( 1) and C(2jOTs groups. Use of model compounds e.g. (206),again gives mainly the axial product (207) with an increased ratio to minor component. The reaction appears to involve the lithium salts (208) and (209) (hydride transfer to carbonyl and ring scission); both of these possible structures can give rise to an axial oxygen function at C-2 in the bicy~lononane.'~~ Enol lactone (210) similarly gives the tricyclododecandiol (21 1; R = R' = H) presumably with an axial hydroxy-group.The X-ray data on (211 ; R = H R' = p-IC6H,.CO) shows the C-4 methylene group flexed outwards to a considerable extent.'38 0 X A0 & '6 R (204) R' = OH R2 = H R3= Me (203) R = Me (205)R' = H R2 = OH R3 = Me (202) R= H. X = O-CHO (206) R = H (207) R' = OH R2 = H R3 = H 13b R. A. Appleton J. R. Dixon J. M. Evans and S. H. Graham Tetrahedron 1967,23 805. 13' J. Martin W. Parker B. Shroot and T. Stewart J. Chem. SOC.(C) 1967 101. lS8 G. Ferguson W. D. K. Macrosson J. Martin and W. Parker Chem. Comm. 1967 102. Alicyclic Compounds H (208) 4 11 Attention has been drawn to the various products of dimerization of cyclo-hexenone with basic reagents ; one interesting product is (212) which is readily convertible into bridgehead halogeno-compounds and contrary to an earlier report the halides are solvolysed in appropriate media.' 39* Orbital symmetry rules have been applied to the products of thermal valence tautomerism of bicyclo[6,2,0]dec-9-ene (21 3) -deca-2,9-diene (214) and -deca- 4,9-diene (21 5); (214) gives trans,cis,cis,cyclodeca-l ,3,5-triene by conrotatory ring opening and then cyclises more slowly in a disrotatory process to trans-1,2,3,4,9,10-hexahydronaphthalene, in analogy with known cases.141 Methods for functionalising the more readily accessible tricyclooctane (216) have been explored; nitrene insertion at C-1 and C-3 usefully gives the amine compounds (217) and (218) when the hydrocarbon is heated with methyl azidoformate.Free radical chlorination gives the 3-chloro-compound and chromyl acetate gives ketone (219) and acetate (220). Solvolysis of the 3-tosylate occurs with significant C(lbC(2) scission to give bicyclo[3,3,0]octa-2,6- diene.142 Norbornadiene 2,3-methylene adduct undergoes intramolecular .Tc-cyclo-propyl addition and this has been extended to photochemical cyclisation of the acetylenedicarboxylate adduct (222; R' = R2= C0,Me) as a route to the thermally stable pentacyclo[4,3,0,02~403~80s~7]nonane system (223) in low ~ie1d.l~~" An alternative approach based on the carbene derived from ketone 13' R. C. Duffner and F. Kurzer Chem.and Ind. 1967 1642. 140 B. Furth J. Kossanyi J.-P. Morizur and M. Vandewalle Bull. SOC.chim. France 1967 1428; cf. J.-P.Morizur B. Furth and J. Kossanyi ibid. p. 1422. P. Rodlick and W. Fenical Tetrahedron Letters 1967,4901. J. Meinwald and D. H. Aue Tetrahedron Letters 1967 2317; J. Meinwald and B. E. Kaplan J. Amer. Chem. SOC.,1967,89,2611. 143 (a)H. Prinzbach and D. Hunkler Angew. Chem. Internat. Edn. 1967,6 247; (b)P. K. Freeman and D. M. Balls J. Org. Chem. 1967,32,2354; (c) E. Wiskott and P. von R. Schleyer Angew. Chem. Internat. Edn. 1967,6,694. 344 K.M ackenzie (221) uia the tosylhydrazone sodium salt not only gives the (223) system but fragmentation to 2-ethynylnorbornene also occurs presumably by retro-carbene addition.143h Prediction that thermal isomerization of (224) to (225) is precluded by orbital symmetry has been confirmed by its synthesis from (222; R' = R2 = H);above 550" (224) is cleaved but no 3,7-methanotriasterane (225) is 0b~erved.l~~~ (215) (216) R'= R'= H (217) R~=H,R~=NHCO~M~ (2IS) R'= H R' = NHC0,Me (219) R'=O,R'a H .G8 Rl R' (220) R'= OAc,R'= 0' (223) @@&hi%: (226) R = H (227) R=-(229) RE+ Dihydro-(222; R' = R2 == H) has been shown to solvolyse to exo-2-brend- any1 derivatives; brief low temperature treatment of the acetate or the hydro- carbon with concentrated sulphuric acid gives 2-noradamantanol in high yield,'44a whilst brexane the catalytic hydrogenation product of (222 ; R' = R2 = H) gives noradamantane on treatment with aluminium bromide 'sludge' at 25" very rapidly.'44b Further details of the rational synthesis of triasterane (226) have appeared ; diazomalonic ester gives mono-diethoxycarbonyl- methylene adducts with cyclohexa-l,4-diene and further transformations of the syn-isomer give the diazoketone the precursor of triasteran-9-one.Wolff- Kishner reduction of the ketone is accompanied by ring scission reactions rationalised on the basis of carbanions e.g. (227) and (228); the latter either protonates (minor product) or undergoes prototropic rearrangement-cyclo- propane ring opening to bicyclo[3,3,l]nona-2,6-(or -2,7-)dimes whose forma- tion suggests that stabilisation of the anion (228) by the homallylic double bond cannot be very significant in comparison with the C-4 allylic carbanion formed by 1,2-proton shift-cyclopropane ring opening.14' Triasteran-9-one is the precursor of the homosemibullvalene (230) via the cation (229) by de- protonation ring opening'46 and the fluxional molecule has also been made from 2-cycloheptatrienylethylidene(C7H7CH,CH:) the major product being 144 (a)A. Nickon G. D. Pandit and R. 0.Williams Tetrahedron Letters 1967 2851 ;(b)P. von R. Schleyer and E. Wiskott ibid. 2845; cf. B. R. Vogt and J. R. E. Hoover ibid. p. 2841. 14' H. Musso and U. Biethan Chem. Ber. 1967,100,119. 146 U. Biethan H. Klusacek and H. Musso Angew. Chem. Internat. Edn. 1967,6 176. Alicyclic Compounds bicyclo[4,2,l]nona-2,4,7-triene(23l),which appears to be formed in a separate reaction pathway perhaps via intermediate (232).14' Bicyclononatriene (231) is also formed as a major product in a similar thermolysis of bicyclo[ S,l,O)octa- 2,4-dienecarboxaldehyde tosylhydrazone sodium salt along with bicyclo- [5,2,0]nona-2,4,8-triene (233) which is not however the precursor of nonatriene (231) (by orbital symmetry allowed thermal 1,5-sigmatropy) since heating (233) gives cis-dihydroindene.The products of the reaction are best explained by various ring-closure modes of the diradical(234) formed from the expected carbene ; in confirmation of the formation of this tropilidene acetylene and bicyclo[3,2,2]nona-2,5-diene are also observed. A similar scheme explains the products of reaction of the analogous cyclo-octatetraene compound which ought to give bicyclo[6,2,0]deca-2,4,6,9-tetraene (235) but instead gives among other products both stereoisomers of 9,lO-dihydronaphthalene.The preponderance of the trans-isomer strongly suggests the intermediacy of cyclodecapentaene with one trans-double-bond -the expected product of thermally allowed conrotatory cyclobutene ring-opening of (235) -which by disrotatory ring closure gives the observed trans-dihydronaphthalene.14'' (231) X = CH (244) (236) X =.CH:CH-(24') H (243) (242) Indeed low temperature photolysis of the tosylhydrazone sodium salt gives (236) and only trans-9,lO-dihydronaphthalene;here decatetraene (235) can be shown to be formed and on warming the reaction mixture trans-9,lO- dihydronaphthalene appears.148b All-cis-cyclodecapentaene (among other products) has been observed in photolysis of truns-9,lO-dihydronaphthalene (excited state conrotatory ring opening); photolysis of the dihydronaphthalene at -190" (no cis-isomer formed) followed by warming of the mixture to 25" gives a product with the spectrum of the cis-i~omer;'~~ photolysis of the cis-isomer gives bicyclo[4,2,2]deca-2,4,7,9-tetraene (236) (thermally reversed) and 147 H.Tsuruta K. Kurabayashi and T. Mukai Tetrahedron Letters 1967 3175. 14' (a) M. Jones and S. D. Reich J. Amer. Chem. SOC.,1967 89 3935; M. Jones and L. T. Scott ibid.,p. 150; (b)S. Masamune C. G. Chin KOHojo,and R. T. Seidner ibid. p. 4804. 149 E.E. van Tamelen and T. L. Burkoth J. Amer. Chem. SOC. 1967.89 151. 346 K.Mackenzie subsequently bullvalene.' 50 The bicyclodecatetraene together with its previ- ously observed valence tautomer is therefore another occupant of the energy profile connecting cis-9,lO-dihydronaphthaleneand bullvalene ;further minima on this curve are (237) theoretically predicted last year,'" and the known (238).Formation of (237) from bullvalene is a concerted suprafacial 1,3-sigmatropic shift in excited state orbital ~yrnmetry.'~~ Thermolysis of (239) gives (240) presumably via a vinylogue of bullvalene formed by sterically unfavourable ring scission [raison d'etre for (239)] ; dihydrobullvalene (241) undergoes concerted 1,5-homodienyl hydrogen shift to the single product (242).' 53 A lucid account of the development of a rational synthesis of bullvalene and other fluxional molecules e.g.barbaralone (243) has been given. Reduction of the carbene insertion product of (243) and elimination through the acetate gives bullvalene and Grignard syntheses based on the ring-expanded ketone can clearly serve as routes to mono-substituted derivatives e.g. the methyl and phenyl compounds which (like the halogeno-and t- butoxy-compounds) appear to prefer vinylic substituents. Since bullvalone can enolise the forma- tion of a dideuterio-compound in basic deuteriated media is not surprising; eventually however all hydrogen is exchanged exemplifying the compound's fluxional character. In accord with the smaller ring system rearrangements in the barbaralone series are times faster than in the bullvalene Barbara101 obtained by hydride reduction of (243) is also made from bicyclo- [3,2,2]nona-2,6,8-triene with aluminium chloride.'54b Propellatriene (244) has been made from cis-9,10-bishydroxymethylhexa-hydronaphthalene dimesylate.' 55 Syntheses of [4,4,4]propellanes and deriva- tives of the [3,3,3]-series together with rules for their nomenclature have been discu~sed.'~~"~~ Photochemical routes to cage structures continue to enjoy popularity. An interesting thermal case is the reaction of tropone with tropilidene by consecu- tive (6 + 4)n cycloaddition and intramolecular Diels-Alder reaction to give pentacyclo[ 7,5,O,O29'O59' 306.1 '' 2]tetradeca-3,lO-dien-8-one. The accessibility of cubane and homocubane derivatives has stimulated mechanistic studies in this comparatively new area ;the solvolytic properties of syn-and anti-tosylates (245) and (246) are quite different for the former reacts without rearrangement and with retained configuration whereas the latter gives besides unrearranged alcohol (246; R2 = H R' = OH) mainly W.von E. Doering and J. W. Rosenthal Tetrahedron Letters 1967 349. W. von E. Doering and J. W. Rosenthal J. Amer. Chem. SOC. 1966,88 2078. 152 M. Jones jun. J. Amer. Chem. SOC. 1967,89,4236. J. N. Labows jun. J. Meinwald H. Rottele and G. Schroder J. Amer. Chem. SOC., 1967,89,612. 154 (a)W. von E. Doering B. M. Ferrier E. T. Fossel J. H. Hartenstein M. Jones jun. G. Klumpp R M. Rubin and M. Saunders Tetrahedron 1967 23 3943; see also J. B. Lambert Tetrahedron Letters 1963 1901 ;(b)M. J. Goldstein and B.G. Odell J. -4rner. Chem Soc.. 1967.89. 6356. 155 L. A. Paquette and J. C. Phillips Tetrahedron Letters 1967,4645. ls6 J. Altman D. Becker D. Ginsburg and H. J. E. LoewenthaI Tetrahedron Letters 1967 757; cf. R. L. Cargill and J. W. Crawford ibid. p. 169. 15' J. Altman E. Babad J. Itzchaki and D. Ginsburg Tetrahedron 1966 Suppl. 8 279. 158 S. lt8 Y. Fujise and M. C. Woods Tetrahedron Letters 1967 1059. Alicyclic Compounds R' (245) R' = H R' = OTs m = 0 (246) R' = H R' = OTS n = 1 (247) R' = OH RZ = H.m = 1 n = 0 the symmetrical bishomocubane (247). Since the rates are lo3-lo4 times faster than calculated values bridged ions are postulated.' 59 Syntheses of bridgehead hydroxy- and alkyl adamantanes have been de- scribed.160 2-Substituted adamantanes are more difficultly accessible 1-adamantanol however rearranges to the 2-ketone in sulphuric acid.' 61 Photolysis of 1-adamantyl azidoformate gives the C-2 nitrene insertion product which reduces to a useful yield of 2-aminoadamantan01.'~~1,2,3,4-Tetrabromoadamantane readily made in the presence of aluminium bromide is converted into the tetraol with sulphuric acid-silver ion and Hofmann degradation of the tetracarboxylic acid gives an excellent yield of the tetra- amino-compound.'63 A rational synthesis of 1,3-diethoxycarbonyladamantaneis based on 4,4-diethoxycarbonylcyclohexanone enamine and ethyl 2-bromomethylacrylate ; these react to give 3,3,7-triethoxycarbonylbicyclo[3,3,l]nonan-9-one Dieck- mann ring-closure leading to 1,3-diethoxycarbonyladamantane-2,6-dione.159 W. L. Dilling and C. E. Reineke Tetrahedron Letters 1967,2547; P. von R. Schleyer J. J. Harper G. L. Dunn V. J. DiPasque and J. R. E. Hoover J. Amer. Chem. SOC.,1967,89,699. S. Landa J. Vais and J. Burkhard Z. Chem. 1967,7,233; W. Hoek; J. Strating and H. Wynberg Rec. Trau. chim. 1966.85 1045. W. Hoek H. Wynberg and J. Strating ibid. p. 1054. H. W. Geluk and J. L. M. A. Schlatmann Chem. Comm. 1967,426. 16' W. V. Curran and R. B. Angier Chem. Comm. 1967,563. 163 H. Stetter and M. Kranse Tetrahedron Letters 1967 1841. 164 H. Stetter and H. G. Thomas Angew. Chem. Internat. Edn. 1967,6 554.