8 Photochemistry By H. A. J. CARLESS Department of Chemistry Birkbeck College Malet Street London WC1E 7HX The past year has seen an increased emphasis on the use of photochemical methods for organic synthesis. The well-known cis-trans isomerization of alkenes continues to provide results of interest. Direct irradiation (185 nm) of cyclohexene or cycloheptene in pentane solution produces the highly strained trans-isomers which can be trapped by acidic methanol to form the corresponding methoxycycloalkanes.' In this way trans- cycloheptene was found to be a reasonably long-lived species at -10 "C,with a lifetime of 23 minutes. A transient U.V. absorption spectrum produced on irradia- tion (248 nm) of 1-phenyl-cycloheptene solutions can be assigned to a twisted trans-like isomer of 1-phenylcycloheptene.2 The photochemical addition of alcohols to cycloalkenes proceeding via the trans-intermediate is generally non-stereospecific.However irradiation3 of the benzocycloheptadienone (1) 'in deuteriated methanol or acetic acid leads to the product (2) of a regio- and stereo-specific syn-addition to the trans-cycloheptadienone intermediate. It has proved possible to prepare a trans-doubly bridged ethylene (3)by xylene-sensitized photo-isomerization of the cis-precur~or.~ This photochemical route to (3) \ @J*&m \ H (1) (2) a; R=Me b; R =MeCO ' Y. Inoue S. Takamuku and H. Sakurai J.C.S. Perkin ZZ 1977 1635. R. Bonneau J. Joussot-Dubien J. Yarwood and J. Pereyre TetrahedronLetters 1977,235. 'E.Dunkelblum and H.Hart J. Amer Chem. SOC.,1977,99,644. M. Nakazaki K.Yamamoto and J. Yanagi J.CS. Chem. Comm. 1977,346. 165 H. A. J. Carless compares very favourably with the non-photochemical approach to such compounds recently reported by Mar~hall,~ who would classify compound (3) as a [10,8]betweenanene. Metathetic reactions of alkenes have recently become topical and these can now be induced in a route which involves photochemical addition followed by pyrolysis. Two groups have investigated this approach to medium-ring carbo- For example [2+21 photocycloaddition of the methyl cyclo-butenecarboxylate (4) to cyclopentene gives (5) which can be cleaved on ther- molysis to the cyclononadiene ester (7) or under less vigorous conditions to the divinylcyclopentane (Q6The reaction of a cyclobutene with a cyclohexenone may H hv -\ C0,Me C0,Me (4) (5) C0,Me (7) likewise provide a useful route to the ten-membered ring of germa~radienes.~,’ Photochemical cleavage of cyclobutanes is also possible and Kaupp’ has sum- marized the photolysis of substituted cyclobutanes in terms of a ‘&-effect’ in which there is a preference for breaking of the cyclobutane bond linked to the chromo- phore and also to the substituent which interacts most strongly with the chromo- phore.Simultaneous publications by two group^^"^ discuss the photochemical re- arrangement of 3-vinylcyclopropenes to cyclopentadienes in high chemical yield. For example (8)produces the cyclopentadienes (9) and (10) in the ratio 3 :1.9 The p& hvL QPh Ph Bu‘ Ph Ph Ph But (8) (9) (10) structure of the major product is more in accord with a biradical (11) pathway rather than a carbene (12) route.In confusing contrast 3-ally1 substituted cycle-’J. A. Marshall and M. Lewellyn J. Amer. Chem SOC.,1977,99 3508. ‘P. A. Wender and J. C. Lechleiter J. Amer. Chem. Soc. 1977,99 267. G. L. Lange M.-A. Huggins and E. Neidert Tetrahedron Letters 1976,4409. G. Kaupp and M. Stark. Chem. Ber. 1977,110,3084. H. E. Zimmerman and S. M. Aasen J. Amer. Chem. Soc.,1977,99,2342. lo A. Padwa T. J. Blacklock D. Getman and N. Hatanaka J. Amer. Chem. Soc. 1977.99 2344. Photochemistry 167 ‘ (1 1) (12) propenes do seem to undergo photorearrangement via a vinyl carbene inter- mediate.” The carbene gives cycloaddition to the allyl double bond as is also found in the photochemical reactions of allyl azirines (see ref.33). Kropp et aL” have used the technique of photolysis of alkyl iodides at 254nm as a convenient means for the generation of carbocations such as the 1-norbornyl cation from 1-iodonorbornane There has been a long-standing dispute as to whether the singlet excited state of alkanones possesses reactivity towards hydrogen abstraction comparable with that of the triplet state. Henne and Fischer13 have determined the yields of various deuteriated pinacols formed in the photoreaction of deuteriated acetone-propan- 2-01 mixtures and the multiplicities of the states which give rise to the pinacols.They conclude that singlet excited acetone has a rate constant for hydrogen abstraction at least as great as that of the acetone triplet. The second excited singlet (ww*) state of adamantanethione is apparently a very indiscriminate abstractor of hydrogen from alkanes since it shows only a small kinetic isotope-effect in abstrac- tion and has little preference for abstraction from tertiary rather than primary C-H bond^.'^ A variety of methods is being used to detect and intercept the 1,4-biradicals believed to be involved in the Norrish Type I1 photochemical reactions of dialkyl and alkyl aryl ketones. Thus laser photolysis (347.1 nm) of 5-methylhexan-2-one produces transient absorption spectra (lifetime ca. 100 ns) assigned to the 1,4- biradical (13).15 The paraquat dication (l,l’-dimethyl-4,4’-bipyridylium) is able to trap such biradicals yielding the easily detected paraquat radical cation whose rate of formation allows the first estimate of the lifetime of the y-methylvalerophenone biradical (14) (7 = 97 f15 ns) in methanol at room temperature.16 The spin-trap di-t-butyl selenoketone efficiently intercepts 1,4-biradicals such as (14) which yields PhCOCHZCH2C(Me),SeC(Bu‘),Has the final product.I7 0 OH RZ @ R OH (13) R’ = R2=R3=Me (14) R’ =Ph R2= R3= Me Irradiation of 5-methyl- 1,4-naphthoquinone yields the blue photo-en01 (15) which is stable at 77 K and provides the first unambiguous example of formation of “ A. Padwa and T. J. Blacklock J. Amer. Chem. SOC.,1977,99 2345.12 P. J. Kropp G. S. Poindexter N. J. Pienta and D. C. Hamilton J. Amer. Chem Soc. 1976,98 8135. ” A. Henne and H. Fischer J. Amer. Chem. Soc.,1977,99,300. l4 K.Y. Law,P. de Mayo and S. K. Wong J. Amer. Chem Soc. 1977,99,5813. Is R. D. Small and J. C. Scaiano Chem. Phys. Lerlers 1977,50,431. l6 R.D. Small and J. C. Scaiano J. Phys. Chem. 1977,81 828. ” J. C. Scaiano J. Amer. Chem. Soc. 1977,99 1494. H. A. J. Carless a (2)-dienol in a photo-enolization reaction.18 The Norrish Type I reaction of cyclic ketones can produce unsaturated aldehydes and this route has been used in the synthesis of some prostanoids by photolysis of the appropriately substituted bicycl0[2,2,l]heptan-2-ones,’~and in production of 5-carbomethoxyhex-4-en-1-a1 for terpene synthesis from 2-carbomethoxy-2-methylcyclopentanone.20 A notably short stereoselective synthesis of the insect sex-attractant em-brevicomin (18) relies on the selective excitation (using 1-methylnaphthalene as combined sensitizer and triplet quencher) of the carbonyl group of the dihydropyranyl ketone (16) followed by hydrogenation of the resulting bicyclic compound (17).21 The absorption of visible light by a dye photosensitizer in the presence of ground-state oxygen (302) is often used to generate singlet oxygen (lo,),and the importance of ‘0 in synthetic chemistry is increasing rapidly.Singlet oxygen generally gives a 1,4-addition to cisoid conjugated dienes producing endo-perox- ides which can easily rearrange to syn-1,3-diepoxides on heating.The use of this sequence on the annulene (19) can produce an intermediate (20) which on repeti- tion of these steps yields the remarkable naphthalene pentoxide (21)., The 1,4-cycloaddition of ‘0,may even occur when one of the diene double bonds forms part of an aromatic ring as in styrenes. An ingenious synthesis of the anti-tumour agent (f)-crotepoxide (24)relies on photo-oxidation of the styrene derivative (22) to produce a bis(endo-peroxide) (23) which can be rearranged and converted into cro tepo~ide.’~ Is E. Rommel and J. Wirz Helv. Chim. Acta 1977,60 38. l9 N. M. Crossland S. M. Roberts and R. F. Newton J.CS. Chem Comm. 1977 886. G. Bidan J. Kossanyi V. Meyer and J.-P. Morizur Tetrahedron 1977,33,2193. 21 P. Chaquin J.-P. Morizur and J.Kossanyi J. Amer. Chem. SOC.,1977 99,903. 22 E. Vogel A. Breuer C.-D. Sommerfeld R. E. Davis and L.-K. Liu Angew. Chem. Internat. Edn. 1977 16 169. 23 M. Matsumoto S. Dobashi and K. Kuroda Tetrahedron Letters 1977 336 1. Photochemistry 169 The immediate biological precursors of the prostaglandins are endo-peroxides having the 2,3-dioxabicyclo[2,2,l]heptaneskeleton. One approach to the synthesis of such peroxides involves the 1,4-~ycloaddition of ‘0,to cyclopentadienes such as 1,4-diphenylcyclopentadienewhich yields the peroxide (2S).24It has now been found that the double bond of (25) can be selectively reduced by di-imide without affecting the peroxidic link a process which should increase the synthetic uses of cyclic peroxides like (25).Thus this reaction has made possible a simple synthesis of the previously unknown 2,3-dioxabicyclo[2,2,2]octane from cyclohexa- 1,3-diene.” There is a rapidly growing interest in 1,2-dioxetans and their chemistry,26 and photo-oxidation of hindered alkenes by singlet oxygen presents a good method for their synthesis. In fact the preparation of the dioxetan (26) by sensitized photo- oxidation of bis(adamanty1idene) has become a routine experiment.*’ Thermal (26) R=R=H (27) R-R= (j /O decomposition of such dioxetans has been used because it provides a chemical route to electronically excited species; ring-fission yields excited carbonyl compounds with the subsequent emission of chemiluminescent light. Wynberg and his group2* have prepared the stable and optically active dioxetan (27) by photo- oxidation of the appropriate optically active alkene and theyz9 have observed that the thermal decomposition of this dioxetan (27) produces chemiluminescence which is circularly polarized.There has been dispute as to whether dioxetan formation from ‘0,attack on an alkene occurs concertedly or via a perepoxide intermediate as calculated by Dewar et uL30 The findings of McCapra and Beheshti3’ are especially relevant to 24 D. J. Coughlin and R. G. Salomon J. Amer. Chem. SOC. 1977 99 655. 25 W. Adam and H. J. Eggelte Angew. Chem. Intemat. Edn. 1977,16 713. 26 W. Adam Advances in Heterocyclic Chem. 1977,21 437. ’’ ‘Organic Photochemical Syntheses’ Vol. 2 ed. R. Srinivasan Wiley-Interscience New York 1976 p.10. 28 H. Wynberg and H. Numan J. Amer. Chem. SOC. 1977,99,603. 29 H. Wynberg H. Numan and H. P. J. M. Dekkers J. Amer. Chem. Soc. 1977,99,3870. 30 M. J. S. Dewar A. C. Griffin W. Thiel and I. J. Turchi J. Amer. Chem. SOC. 1975,97,4439. ” F. McCapra and I. Beheshti J.C.S. Chem. Comm. 1977 517. H.A. J. Carless this question because ‘0,reacts with camphenylidene-adamantane to give the rearranged 1,2-dioxolan (28) alongside the expected dioxetan. The formation of (28) becomes most easily understood in terms of the rearrangement of a dipolar intermediate such as the perepoxide (29). Turro Adam and m-worker~~~ have R (30) a; R=Me b; R=Ph now succeeded in isolating a-peroxylactones such as (30) from reaction of ‘0,with the double bond of ketenes.The dioxetanone (30a) is stable enough to with- stand distillation below room temperature but decarboxylates with intense chemiluminescence on warming. Padwa has continued his studies on the photochemical generation of nitrile ylides from azirines by examining the intramolecular cycloaddition of the ylide to a double bond. Either 1,l -(~arbene-like),~ have been found or 1,3-~ycloaddition~~ and for some azirines both pathways are competitive as with (31)yielding (32) and (33) re~pectively.~~ Padwa and his group believe the predominant photoproduct can be understood in terms of the electronic effect of substituents on the structure of the ylide as proposed by Houk et ~1.~~ -* \ h’ PY’J +M 0 The first example of a [2+21 photocycloaddition of an alkene to an azo-bond has been noted.37 Thus the rigid unsaturated azo compound (34) where a C=C bond and the N=N bond are parallel and close is claimed to undergo ring-closure to the (34) (35) 32 N.J.Turro Y.Ito M.-F. Chow W. Adam 0.Rodriquez and F. Yany J. Amer. Chem. SOC.,1977,99 5836. 33 A. Padwa and P. H. J. Carlsen J. Amer. Chem. Soc. 1977,99,1514. 34 A.Padwa and N. Kamigata J. Amer. Chem. Soc. 1977,99 1871. 35 A. Padwa P. H. J. Carlsen and A. Ku J. Amer. Chem. Soc.,1977,99 2798. 36 P.Caramella and K. N. Houk J. Amer. Chem. Soc.,1976,98 6397;P. Caramella R. W. Gandour J. A. Hall C. G. Deville and K. N.Houk ibid 1977,99 385. ’’W. Berning and S. Hunig Angew. Chem. Internat. Edn. 1977,16 777. Photochemistry 171 remarkable isomer (35) on irradiation in acetonitrile.An unusual photocyclization of N-allyliminium salts such as (36) to yield pyrrolidines [e.g. (37)] on irradiation in \ I methanol may well be the result of an intramolecular [2+2Jphotocycloaddition of the iminium chromophore to the olefinic bond followed by C-N bond cleavage and capture of a nu~leophile.~~ Photochemical [2 +21 cycloaddition reactions of carbonyl compounds to alkenes yielding oxetans are well known. Yang and Chiang3’ have now found that the photocycloaddition of aromatic aldehydes (38) to cyclohepta- 1,3,5-triene yields a [6+2] adduct (39) as a major product besides oxetans (40). The reaction is (38) a; Ar = 1-pyrenyl b; Ar =2-naphthyl believed to occur from the singlet excited state of the aldehyde and the fact that a [4+21 adduct could not be detected may be evidence for a concerted photochemi- cal mechanism since the [27rs+27rs] and [gS +2ms] additions are allowed on the basis of orbital symmetry rules whilst the [4~s +2~s] is not.Bryce-Smith Gilbert and co-~orkers~~ have previously published useful guide- lines which rationalize the orientation and stereochemistry of cycloaddition of photoexcited benzene to olefins but the exceptions to these rules which are appearing show that there are still factors to be understood. Thus the photoad- dition of ethylene or propene to benzene (in a U.V. autoclave) gives 1,3-addition to the benzene ring as the major process rather than the expected 1,2-additi0n.~’ 1,2-addition to the benzene ring generally gives an endo-adduct for donor-substi- tuted alkenes but an exo-adduct for acceptor-substituted alkenes.For example photoaddition of 2,3-dihydropyran is the most efficient (a=0.7) cycloaddition to benzene so far known and gwes a great preference for the endo-adduct (41).42 In contrast photoaddition of benzene to 2,2-dimethyl- 1,3-dioxolen gives an em-1,2-adduct (42).43 Photocycloaddition of benzene to methyl acrylate or methacrylate shows surprisingly little selectivity yielding a mixture of em-and endo- 1,2-ad duct^.^^ 38 P. S. Mariano J. L. Stavinoha and R. Swanson J. Arner. Chern. Soc. 1977,99 6781. 39 N. C. Yang and W. Chiang J. Amer. Chem Soc. 1977,99,3163. 40 a. Bryce-Smith A. Gilbert B.H. Orger and H. M. Tyrrell J.C.S. Gem. Cornrn. 1974,334. M. F. Mirbach M. J. Mirbach and A. Saw Tetrahedron Letters 1977,959. 41 42 A. Gilbert and G. Taylor Tetrahedron Letters 1977,469. 43 H.-D. Scharf and J. Mattay Tetrahedron Letters 1977,401. 44 R. J. Atkins G. I. Fray A. Gilbert and M. W. bin Samsudin Tetrahedron Letters 1977 3597 H. A. J. Carless The photocycloaddition of an acetylene to an aromatic compound usually results in the formation of a cyclo-octatetraene via a proposed bicyclo[4,2,0]octatriene intermediate. Two have presented new evidence on these intermediates. Sket and Zupan4' have been able to isolate stable bicyclics (43) from the photo- addition of phenyl-substituted acetylenes to hexafluorobenzene whilst Tinnemans Ph C0,Me ph*FR F I- R =But Pr" or Me (44) (43) and Necker~~~ have shown that the bicyclo[4,2,0]octatriene from methyl phenyl- propiolate and benzene can undergo intramolecular photocycloaddition to yield (44).Five-membered ring heterocycles often undergo ring-atom transpositions on U.V. irradiation [e.g. the 2-alkylindazole (45)rearranges to a benzimidazole (46) via the detectable tricyclic intermediate (47).47 Until now these processes had not been observed for carbocyclic rings. However an elegant study of the irradiation of a ['3C,]-labelled cyclopentadiene (48) shows that the phototransposition of cyclo- pentadiene [(48) +(50)]occurs probably by intramolecular [2 +21 addition to yield bicyclopentene (49) followed by a 1,3-sigmatropic shift4* This pathway would account for the fact that after irradiation the bicyclopentene is found to have a " B.Sket and M. Zupan J. Amer. Chem. SOC.,1977,99,3504. 46 A. H. A. Tinnemans and D. C. Neckers J. Amer. Chem. SOC.,1977,99,6459. 47 W. Heinzelmann M. Marky and P. Gilgen Helv. Chim. Acta 1976,59 1512. 48 G. D. Andrews and J. E. Baldwin J. Amer. Chem. SOC.,1977,99,4851. Photochemistry 173 greater ratio of non-vicinal to vicinal [”C,] label than is found in the recovered cyclopentadiene. Palensky and Morrison4’ have found that a similar skeletal re- arrangement occurs on irradiation (254 nm) of alkyl substituted indenes e.g. in the interconversion of (51) and (52) and they have independently proposed a similar R‘ (53) (51) R’=Me,R2=H (52) R’ = H R2= Me mechanism.In the case of irradiation of 1,l-dimethylindene the unstable inter- mediate isoindene (53) becomes the sole product because it cannot undergo a 1,Shydrogen shift which would normally reform an indene. Barltrop Day and Samuel who had already shown the intermediacy of a zwitterion [e.g. (55)] in the photochemical rearrangement of 4-pyrones to 2-pyrones in trifl~oroethanol,~~ have now been able to isolate a further intermediate the cyclopentadienone epoxide (56) in the conversion of (54) into (57).5’ Barton and Hulshof5’ have used the related photochemical ring contraction of a 4-pyrone (58) to cyclopentenedione (59) followed by in situ reduction with sodium cyanoborohydride in the elegant synthesis of the mould metabolite terrein (60).A similar rearrangement occurs for protonated 4-pyrones (i.e. hydroxy-pyrylium cations) on irradiation in sulphuric acid. Thus the 2,3-dimethyl-4- hydroxypyrylium cation rearranges in part to the corresponding 2-hydroxypyrylium cation; however in this case Pavlik et aZ.53have shown that the major reaction- product is a protonated fury1 cation (61). A mechanism for the curious formation of (61) remains to be established. 49 F. J. Palensky and H. A. Morrison J. Amer. Chem. SOC.,1977 99 3507. so J. A. Barltrop A. C. Day and C. J. Samuel J.C.S. Chem. Comm 1976,822. ” J. A. Barltrop A. C. Day and C. J. Samuel J.C.S. Chem. Comm. 1977 598. 52 D. H. R. Barton and L. A. Hulshof J.CS. Perkin I 1977 1103.53 J. W. Pavlik D. R. Bolin K. C. Bradford and W. G. Anderson J. Amer. Chem. Soc. 1977,99 2816. H. A. J. Carless (62) RF =CF(CF,) Irradiation of perfluorinated pyridines has given the first stable examples of 2-azabicyclic isomers such as (62).54 In contrast a study of the photochemical decomposition of 1,4-[ 15N2]-s-tetrazine to nitrogen and hydrogen cyanide shows that reaction occurs without any preliminary 1,4-nitrogen bonding.” Matrix isolation techniques applied to photochemical reactions have provided some interesting results such as the first isolation of thiirene (63) by photolysis at 8 K of 1,2,3-thiadia~ole,’~the production of unstable isocyanides [e.g. (64)] on irradiation of pyridine N-oxide at 10K,57 and evidence for a process (perhaps involving a ring-opened aldoketene) which scrambles C-2 and C-6 on irradiation of a [‘3C]-labelled a-pyrone at 8 K.58 The use of chiral solvents to induce asymmetric syntheses in photochemical experiments has received some attention this year.Irradiation of 2-styrylbenzo[c]-phenanthrene in chiral solvents gives the optically active hexahelicene in optical yields up to 2‘%0,~’ whilst photochemical rearrangement of nitrones to optically active oxaziridines occurs in optical yields up to 30% on irradiation in a (+)-or (-)-2,2,2-trifluoro-l-phenylethanol and fluorotrichloromethane solvent.60 The photochemical reduction of carbonyl compounds in the presence of the optically active (S,S)-1,4-bis(dirnethylamino)-2,3-dimethoxybutane gives both mew and optically active pinacols derived from the carbonyl.For example acetophenone gives an optical yield of up to 25%0 enriched in the (R,R)-pinacol enantiomer.61 ” R. D. Chambers and R. Middleton J.C.S. Chem. Comm. 1977 154. 55 D. S. King C. T. Denny R. M. Hochstrasser and A. B. Smith J. Amer. Chem. Soc. 1977 99 271. 56 A. Krantz and J. Laureni J. Amer. Chem. SOC.,1977,99,4842. 57 0.Buchardt J. J. Christensen C. Lohse J. J. Turner and I. R. Dunkin J.CS. Chem. Comm. 1977 837. ’* B.-S. Huang R. G. S. Pong J. Laureni and A. Krantz J. Amer. Chem. Soc. 1977 99 4154. 59 W. H. Laarhoven and T. J. H. M.Cuppen J.C.S. C’hem. Comm. 1977,47. D. R. Boyd and D. C. Neill J.C.S. Chem. Comm. 1977,51. ‘* D. Seebach H.-A. Oei and H. Daum Chem. Ber. 1977,110,2316.