9 Photochemistry By A. GILBERT Chemistry Department The University Reading THEvolume of literature concerned with the many and varied aspects of photo-chemistry continues to expand and this is reflected by the now annual increase in size of the Chemical Society’s Specialist Periodical Report’ on this topic. Ac- cordingly the present review is restricted to some of the publications from the 1971 literature which in the author’s opinion are significant in their own special- ized area or are of general interest. It may have been true in the past that organic photochemists did not pay sufficient attention to the mechanistic features of reactions. This criticism is rapidly becoming less true as judged by the detailed study which is now evident in a good proportion of reports from this area of the effects of solvent concentra- tion atmosphere quenchers and temperature on quantum yields.Concern over the possible involvement of ground-state complexes and/or excimers and exci- plexes in such processes also is increasing. The effect of heavy-atom solvents in enhancing the yield of dimeric products has been previously attributed to an increase in the rate of intersystem crossing. Work on the dimerization of coumarin now suggests that this is not the case as the enhanced formation of the cis-anti-cis cyclobutane dimer in CC14 is not accompanied by the expected increase in intersystem crossing.2a Formation of this dimer is a triplet process but the solvent effect seems to be that of increasing the fraction of coumarin triplets which dimerize.The cis-s~v~-cis dimer arises by a singlet process involving an excimer as previously postulated. Methanol is a favoured solvent in photochemical reactions but a change in media to this protic solvent can have a marked effect on the reaction pathways as several workers have reported. For example the formation of (1) from (2) in benzene has been previously noted but in methanolic solution the product is (3).2b This change in reaction is undoubtedly related to the ability of methanol to solvate the developing charge centres in the intermediate (4). Excitation of charge-transfer bands in molecular complexes is often a fruitful source of photochemical reaction and such a process has been studied with some ’ ‘Photochemistry,’ ed.D. Bryce-Smith (Specialist Periodical Reports) The Chemical Society London vol. 1 1970 470 pp.; vol. 2 1971 817 pp.; vol. 3 1972 876 pp. (a) R. Hoffman P. Wells and H. Morrison J. Org. Chem. 1971 36 102; (b) A. Padwa and E. Glazer Chem. Comm. 1971 838. 32 1 322 A. Gilbert Ph &Ph NXN H Ph toluene c~mplexes.~ Formation of (5) from the toluene-1,2,4,5-tetracyano-benzene complex is accounted for in terms of a proton transfer from a cation- radical to an anion-radical in the excited state of the complex by a process some- what reminiscent of the first step in the Birch reaction this is followed by the loss of HCN. There is seemingly no evidence for a ground-state complex in the toluene-tetramethylpyromellitate system but reaction ultimately resulting in the formation of (6) is suggested to involve an exciplex and a mechanism of proton transfer to the carbonyl group similar to the first step in the acyloin condensation.NcDcH2Ph Meo2cG NC CN Me0,C H Ph (5) (6) Consideration of orbital symmetry factors now seems to be a very necessary part of the interpretation of photoprocesses and a short elementary review dealing with this topic has been p~blished.~ An earlier suggestion concerning such considerations that the simple rule of ‘ground state forbidden = excited state allowed’ was more readily applicable than the more precise Woodward- Hoffmann postulatzs has been criticized by Michl who considers that this is an overgeneralization and cites the ground-state-forbidden (7) to (8) conversion which does not arise from the first excited singlet or triplet states but involves a higher tri~let.~ Cookson and co-workers have reported further examples of A.Yoshino M. Ohashi and T. Yonezawa Chem. Comm. 1971 97. H. Katz J. Chem. Educ. 1971 48 84. J. Michl J. Amer. Chem. SOC.,1971 93 523. Photo chemistry light-induced 1,3-allylic migrations with such systems as (9),6 but have pointed out from subsequent work with the optically pure dinitrile (10) the danger in uncritical extension to strongly perturbed derivatives of orbital symmetry rules which were devised from parent systems.' (7) Ph As may be expected the photochemistry of the carbonyl group has again attracted numerous workers and a multitude of reports concerned with reduction oxetan formation Norrish Type I and I1 reactions and sensitization processes have appeared.The original observation that the relatively short triplet lifetime of benzophenone in benzene indicates a possible specific quenching process which does not lead to significant reduction,' has received further consideration. It also appears that in this system not all the benzophenone consumed had been accounted for in terms of product. Complementary studies of flash photolysis and continuous irradiation of benzophenone in ultra-pure benzene have now been made and a primary hydrogen abstraction by the ketone triplet to yield ketyl and phenyl radicals is consistent with the observed products and inter- mediate~.~," Biphenyl had been previously observed but now benzpinacol and 4-biphenyldiphenyl carbinol have been detected and again the presence of the unidentified yellow compound is noted.Experiments with deuteriated solvents confirm the solvent origin of the biphenyl and an isotope effect consistent with primary hydrogen abstraction from the solvent is also observed. It is known that di-p-t-butyl substituents in benzophenone cause complication and a deviation from parent ketone behaviour in propan-2-01. Thus the effect on the ketone ' ' R. C. Cookson J. Hudec and M. Sharma Chem. Comm. 1971 107 108. R. C. Cookson and J. E. Kemp Chem. Comm. 1971 385. J. A. Bell and H. Linschitz J. Amer. Chem. SOC.,1963 85 528. A. V. Buettner and J. Dedinas J. Phys. Chem. 1971,75 187.lo J. Dedinas J. Phys. Chem. 1971 75 181. 324 A. Gilbert photochemistry of varying the size of the p-alkyl substituents has been investi- gated.' Short irradiation of (1 1) with R = Me Et or Pr' results in a clean con- version into the transient intermediate observed for the parent ketone and for which structure (12) has been suggested. At conversion of <50% this inter- mediate reacted with the ketone to yield acetone and the benzpinacol for R = Me R but at higher conversions for R = Et or Pr' a stable yellow species as with R = Bu' was formed in a competing process. It has been assumed that the products (except oxetan) from irradiation of acetone in olefins particularly cyclohexene may be accounted for by the formation of free ketyl and cyclohexenyl radicals.The new products (13) and (14) have now been found in this fundamental system and a different primary course of reaction has been proposed.12 A dual mechanism is now suggested which involves both the acetone triplet and an acetone<yclohexene exciplex. Seemingly the reaction of acetone with cyclo- octene and cyclo-octa-1,3-diene is not so complex and in both cases the cis and trans fused oxetans are formed.' The principal difference between the two reac- tions is that the olefin process occurs from the triplet state whereas that of the diene involves a singlet state as had been previously reported for the reaction of acetone with conjugated acyclic dienes.14 The involvement of the oxetan (15) in the formation of (16) from the reaction of benzophenone with 2,3-dimethylbuta- 1,3-diene had earlier been postulated,' and has now been verified by its isolation I' N.Filipescu L. M. Kindley and F. L. Minn J. Org. Chem. 1971 36 861. l2 P. Borrell and J. Sedlar Trans. Faraday SOC.,197 1 66 1670. l3 K. Shima Y. Sakai and H. Sakurai Bull. Chem. SOC.Japan 1971,44 215. l4 J. A. Barltrop and H. A. J. Carless Chem. Comm. 1970 1637. l5 J. Saltiel R. M. Coates and W. G. Dauben J. Amer. Chem. SOC.,1966 88 2745. Photochemistry 325 from the process.I6 The lack of detection of (15) earlier is attributed to its sen- sitivity towards acid. Oxetan formation has thus been studied with a wide range of ketones and generally triplet state reactions are encountered except in reactions with dienes (refs.13 and 14) and relatively few cases involving electron-deficient olefins. The oxetan-forming process has now been investigated with esters of aromatic carboxylic acids and olefins and dienes by a group of Japanese wor- kers.17,18 With 1,l-diphenylethylene tetramethyl pyromellitate yields a mono- oxetan as an intermediate which loses formaldehyde to form (17) in 95% yield. Ph / Similar reaction occurs with butadiene but with cyclo-octene as olefin the product is reported to be an unstable oxetan and reaction of this olefin with the triester of trimesic acid results in addition to the aromatic ring. The mechanism of the ester-oxetan reactions has been studied with dimethyl terephthalate and isophthalate with trimethyl ethylene and the involvement of exciplexes is clearly indicated.l8 Although @-unsaturated ketones readily form oxetans tropone is found to undergo a [8 + 21 cycloaddition forming (18) via the nn* triplet state.” The Norrish Type I reaction continues to be investigated with a wide range of carbonyl compounds. The hydrocarbon tetrahedrane remains elusive but forma- tion of acetylene propylene and but-2-yne from the tricyclopentanone (1 9) suggests that it may have been formed following the decarbonylation.20 A general photodecarbonylation of phenol lactones and carbonates (20) has been reported to yield the quinone methides o-quinones or o-thioquinones.21 Evi- dence for formation of the more unstable of these intermediates was obtained from low-temperature i.r. studies and their Diels-Alder reaction with electron- rich olefins.Type I reactions have also been observed for the naturally occurring l6 J. A. Barltrop and H. A. J. Carless J. Amer. Chem. SOC.,1971 93 4794. Y. Katsuhara Y. Shigemitsu and Y. Odaira Bull. Chem. SOC.Japan 1971 44 169. Is Y. Shigemitsu Y. Katsuhara and Y. Odaira Tetrahedron Letters 1971 2887. l9 T. S. Cantrell J. Amer. Chem. SOC.,1971 93 2540. lo H. Ona H. Yamaguchi and S. Masamune J. Amer. Chem. SOC.,1970 92 7495 0. L. Chapman and C. L. McIntosh Chem. Comm. 1971 383. 326 A. Gilbert OH-OX0 coronopilin (21),22 and siduloses (22) too are reported to yield decarbonylated products.23 Carbonyl compounds with y-hydrogens undergo a Norrish Type I1 process to given an enol and olefin.The reaction is one of the most widely studied mecha- nistically and a review of recent advances in this area was published within the year.24 The 1,4-biradicals produced by this process continue to stimulate interest and the properties of such radicals produced by different methods have been c~mpared.’~ In the Type I1 process alkyl ketones react from both the nn* singlet and triplet states whereas phenyl alkyl ketones do so only from the triplet state. It was thus of interest to examine the process with P-naphthyl alkyl ketones and with the n-butyl derivative the reaction was found to arise exclusively from the nn* singlet state.26 Surprisingly cyclobutanol formation which is the major process with corresponding phenyl and alkyl ketones was not observed in the present case it may thus be that the cyclobutanol arises mainly by a triplet process.Both singlet and triplet states are however involved in the inefficient Norrish Type I1 elimination from aromatic esters (23).27 Results from quenching studies show that certain of the esters [e.g. (23a)l react from the singlet state whereas others [e.g. (23b)l involve triplet intermediates. It is concluded that the main reason for such low efficiency of the process with aromatic esters is that a back hydrogen transfer occurs from the initially formed biradical. The analogy between the Type I1 photoelimination of carbonyl compounds and their McLaf- ferty rearrangement under electron impact has been commented upon in the past 0 (23) (a) R’ = R3 = H R2= Me (b)R’ = R2= Me,R3 = OH 22 H.Yoshioka T. H. Porter A. Higo and T. J. Mabry J. Org. Chem. 1971 36 229; J. Kagan S. P. Singh K. Warden and D. A. Harrison Tetrahedron Letters 1971 1849. ’’ P. M. Collins J. Chem. Soc. (0,1971 1960. 24 P. J. Wagner Accounts Chem. Res. 1971 5 168. ” L. M. Stephenson and J. I. Brauman J. Amer. Chem. Soc. 1971,93 1988. 26 N. C. Yangand A. Shani Chem. Comm. 1971 815. ’’ J. A. Barltrop and J. D. Coyle J. Chem. Soc. (B),1971 251. Photochemistry 327 (but see ref. 28). A further example of these comparative processes has been observed with P-ketoanilides which although in principle could yield the elimina- tion process or a Fries reaction photochemically give a quantitative yield of acetophenone and phenylisocyanate on photolysis this result is paralleled by electron impact studies.29 Many other elimination processes continue to be investigated and it is especially interesting to read of a useful photochemical route to ben~yne.~' The inter- mediate is generated by photolysis of phthaloyl peroxide and undergoes stereo- specific [2 + 41 and non-stereospecific [2 + 21 cycloadditions :such reactions are identical in symmetry properties with those found by conventional decomposi- tion of benzenediazonium-2-carboxylate.Di-n-methane rearrangements of non-conjugated dienes [i.e.(24)-+(25)]are still stimulating interest and Zimmerman and co-workers who have added much of the useful data to this area have continued their studies on the factors which affect or divert the process.Photolysis of 1,1,4-triphenyl-3,3-dimethyl-penta- 1,4-diene is found to yield 1 ,a-styryl-2,2-diphenyl-3,3-dimethylcyclopro-pane as the major product and the alternative di-n-methane rearrangement prod- uct 1-(2,2-diphenylvinyl)- l-phenyl-2,2-dimethylcyclopropane ' is not f~rmed.~ This rearrangement is discussed in terms of control by electron delocaliza- tion during the reaction and it is found to involve the excited singlet state. Reaction of the methylene analogues of 4,4-diphenyl- and 4,4-dimethyl-cyclo- hexadienones in the present process has also been considered and the observations are compared with those of the acyclic di-n-methane rearrangement^.^^ The dimethyl derivative yields (26) superficially by a process analogous to that of the dienone but with the hydrocarbon reaction occurs from the singlet rather than the triplet state.The diphenyl derivative differed in that phenyl migration oc-curred to yield cis- and trans-isomers (27) but again via a singlet pathway. Photolysis of the allylic alcohol (28)has been studied and a novel rearrangement elimination reaction is observed to yield o-terphenyl via it is suggested the 6,6- diphenyl bicyclo[3,1,0]hexene (29). 33 The irradiation of 1,1,2,2-tetraphenyl ethane has led to the discovery of a new reaction involving a 'di-n-ethane' *' M. M. Bursey D. G. Whitten M. T. McCall W. E. Punch h4. K. Hoffman and S. A. Benezra Org. Mass. Spectrometry 1970 4 157. 29 W. R. Oliver and L. R. Hamilton Tetrahedron Letters 1971 1837. 30 M. Jones and M.R. Decamp J. Org. Chem. 1971,36 1536. H. E. Zimmerman and A. A. Baum J. Amer. Chem. SOC., 1971.93 3646. 32 H. E. Zimmerman P. Hackett D. F. Juers J. M. McCall and B. Schroeder J. Amer. Chem. SOC.,1971,93 3653. 33 W. G. Dauben W. A. Spitzer and R. M. Boden J. Org. Chem. 1971 36 2384. 328 A. Gilbert R3 (26) R' = R2 = Me,R3 = H (27) R' = R3 = Ph,R2 = H Biphenyl and cis-and trans-stilbenes are the minor products and have been shown to result by such a rearrangement whereas formation of (30) the major product has possibly the same origin but studies at present are in- conclusive. The conversion of bicyclo[2,2,l]hepta-2,5-dienesinto quadricyclanes (31) is a well-known and general process largely because of the research efforts of Prinzbach and co-workers.Such conversion with (32)is reported to be a singlet process even in the presence of trans-~iperylene,~~ and the rearrangement is found to be non-~oncerted.~~ Direct irradiation of such phenyl-substituted diene esters as (33)yields the appropriate quadricyclane whereas acetone sensiti- zation leads to tricyc10[3,2,0,0~~~]heptene derivatives uiua di-n-methane (31) (32) R' = R2= H,R3 = R4= Ph (33) R' = R2 = Ph,R3 = C02Me,R4 = H Formation of 'caged' compounds continues to be of interest and has been reported for two Diels-Alder adducts of cy~lo-octatetraene.~~,~~ Diene-p-quinone adducts have always been favoured compounds for such investigations 34 J. A. Ross W. C. Schumann D. B. Vashi and R.W.Binkley Tetrahedron Letters 1971 3283. 35 G. Kaupp and H. Prinzbach Chem. Ber. 1971 104 182. 36 G. Kaupp Angew. Chem. Internat. Edn. 1971 10 273. 37 H. Prinzbach and M. Thyes Chem. Ber. 1971 104,2489. 38 W. G. Dauben C. H. Schallhorn and D. L. Whalen J. Amer. Chem. SOC.,1971 93 1446. 39 L. A. Paquette J. Amer. Chem. SOC.,1970 92 5765. Photochemistry but it seems that a cyclic diene must be used in the original thermal process to ensure a successful light-induced ‘caging’ reaction since the non-bridged adduct from p-benzoquinone and butadiene only yields polymer on photolysis. With light of A > 340nm however the process is more specific leading to (34) and (35).40 Compound (34) possesses a previously unknown carbon skeleton but formation of (35) provides a facile entry into copaborneol ring systems.Intra- molecular photocyclization is also reported for the Diels-Alder adduct of cyclopentadiene and 1,4-naphthoquinone :41 in this case the product (36) shows that a [6 +21 1,2-cycloaddition to the aromatic ring has occurred. The photochemistry of aromatic molecules generally has again this year attracted much attention. The photoisomerization of benzene into the ‘Dewar’ isomer is now reported to arise from the S state of benzene by a symmetry- allowed process and this appears to provide the first example of a non-dissociative reaction from an upper singlet state.42 The involvement of the valence bond isomer benzvalene in the photochemistry of benzene in the presence of acids and water has been commented upon by two groups.Berson and Hasty have ration- alized the observations and products from their work in acidic solution as arising from acidolysis of ben~valene,~~ and workers in the Argonne Laboratories have again sorted out a problem in benzene photochemistry by re-investigating its reactions in aerated water.44 The product from this latter reaction is now found to be penta-1,3-diene-l-carboxylateand not 2-formyl-4H-pyran as previously reported and oxygen is not essential to the process. In the presence of the strong acid CF3C0,H the photochemistry of benzene is further modified and irradiation of this system is reported to yield am-trifluoroacetophenone as the primary product which subsequently yields the meso-pinacol and trifluoro- methyl diphenyl carbin01.~~ Until recently the photoaddition of simple olefins to benzene had essentially only been found to yield products (37) of a formal 1,3-addition.Three groups of workers have carried out further studies on this topic and describe that as well as ‘O J. R. Scheffer J. Trotter R. A. Wostradowski C. S. Gibbons and K. S. Bhandari J. Amer. Chem. Sor. 1971 93 3813. 41 A. S. Kushner Tetrahedron Letters 1971 3275. 42 D. Bryce-Smith A. Gilbert and D. A. Robinson Arzgew. Chem. 1971 83 803. 43 J. A. Berson and N. M. Hasty J. Amer. Chem. SOC.,1971 93 1549. 44 L. Kaplan L. A. Wendling and K. E. Wilzbach J. Amer. Chem. SOC., 1971,93 3821. 4s D. Bryce-Smith G. B. Cox and A. Gilbert Chem. Comm. 1971 914. 330 A.Gilbert 1,3-type products both 1,2-46,47 and 1,4-46,48 cycloaddition processes occur. Formation of the 1,2-adducts is favoured by high olefin concentrations and with some olefins quantum yields for the process are high but the adducts are generally photolabile which limits their isolation.46 With tetramethylethylene as olefin the major product (38)is that resulting from an ‘ene type’ process and its formation is favoured in proton donor solvent~.~’ More and more workers are concerning themselves with the effect of solvents and acids on the course of photo-reactions. Thus formation of (39) from tertiary amines and benzene is greatly accelerated in the presence of a proton donor,49 and the analogous product (40)from diethyl ether and benzene is only formed in the presence of acid.50 (39) x = NR; (40)X = OEt R = Me In the past there has been some disagreement as to the structure of the photo- dimer of p-alkoxynaphthalenes.The fact that the dimers revert to the monomers on dissolution has suggested to some workers that crystal forces contribute significantly to dimer ~tability.~’ This aspect has been investigated and naphtha- lene-2-carbonitrile has been dimerized in hexane-benzene solution. Unlike the other naphthalene derivative dimers the present dimer is suficiently stable for its n.m.r. spectrum to be recorded and only products of 1,4-1,4-type dimerization are consistent with the data. Such a mode of reaction had been originally deduced for the P-alkoxy-derivatives. Topochemical control in photo-reactions has been studied for both trans-cinnamic acid and anthracene dimerizations.Formation of the trans-dimer from 9-cyanoanthracene is however not in accord with such an approach. A detailed examination of this system has led Cohen and co-workers to suggest that dimeriza- tion occurs within stacking fault regions (bounded by dislocations) in which the monomer molecules are in a trans ~rientation.~~ Some sixteen years ago the photolability of the nitrobenzene-olefin system was described and now de Mayo and co-workers following their preliminary 46 K. E. Wilzbach and L. Kaplan J. Amer. Chem. Soc. 1971 93 2073. 47 D. Bryce-Smith B. E. Fougler A. Gilbert and P. J. Twitchett Chem. Comm. 1971 794. 48 R. Srinivasan I.B.M. J. Res. Develop. 1971 15 34.49 D. Bryce-Smith M. T. Clarke A. Gilbert G. Klunklin and C. Manning Chem. Comm. 1971 916. D. Bryce-Smith and G. B. Cox Chem. Comm. 1971 915. 5’ T. W. Mattingly J. E. Lancaster and A. Zweig Chem. Comm. 1971 595. ’’ M. D. Cohen Z. Ludmer J. M. Thomas and J. 0. Williams Proc. Roy. SOC.,1971 A324,459. Photochemistry 331 communication in 1968 of this interesting process have published full details and have described the isolation at low temperature in a pure crystalline form of the previously postulated 1,3,2-dioxazolidine intermediates (41).53 Evidence is presented to show that the addition proceeds via the nn* triplet state in a two- step electrophilic process. Modification of the photochemistry of nitrobenzene by complexation has been studied by Trotter and Testa who find that although the uncomplexed species yields phenylhydroxylamine in propan-2-01 and is inert in cyclohexane the complex with boron trichloride in cyclohexane yields nitrosobenzene H3BO3 and chlorocyclohexane.54 N.m.r.studies indicate that the co-ordination is confined to the nitro-group. Nitrobenzenes with ortho bulky substituents provide interesting compounds for study since intermolecular hydrogen abstraction by the nitro-group should be hindered and an alternative pathway of intramolecular abstraction is pro- vided. With 2,6-dialkyl derivatives two types of process are reported the photoreduction and nitro-nitrite rearrangement are suggested to arise from the triplet state whereas singlet intermediates are postulated for the intramolecular oxygen transfer involving CI-and 0-attack on the alkyl ~ide-chain.’~ Dopp has studied the o-t-butylnitrobenzenes in detail and has expanded his earlier results as well as reporting on mechanistic considerations.56 The processes have been examined in a series of common solvents when the indolene is formed in the presence of amines solvent-derived products are obtained. Crystalline 173,5-tri-t-butylnitrobenzene yields a variety of product^.^' The field of photo-oxidation processes is large and is the subject of numerous publications each year involving a wide variety of organic compounds. Reports of the formation of dioxetans from oxidation of electron-rich oiefins last year however must have interested all photochemists particularly because of the involvement of such compounds in certain chemiluminescence reactions.58 Inevitably the process has now been extended to other systems and dioxetans have been isolated from p-dioxan and 173-dioxole at -78 “C,and their n.m.r. spectra have been recorded.59 Interest in the use of photochemistry in synthesis continues of course par- ticularly with natural products. Seemingly however even the most well-known 53 J. L. Charlton C. C. Liao and P. de Mayo J. Amer. Chem. Soc. 1971 93 2463. 54 W. Trotter and A. C. Testa J. Phys. Chem. 1971 75 2415. 55 Y. Kitaura and T. Matsuura Tetrahedron 1971 27 1583. 56 ’ D. Dopp Chem. Ber. 1971 104 1035 1043 1058. D. Dopp and K. H. Sailer Tetrahedron Letters 197I 276 1. ’* P. D.Bartlett and A. P. Schaap J. Amer. Chem. Soc. 1970 92 3223; S. Mazur and C. S. Foote ibid. p. 3225. 59 A. P. Schaap Tetrahedron Letters 1971 1757. 332 A. Gilbert photosynthetic procedure of oximation is not fully understood and further evidence against a free-radical chain mechanism involving C1. has been re-ported. ' It is again encouraging to see the amount of effort which is being devoted to the sociological aspects of photochemistry and research into the degradation of herbicides pesticides etc. continues as does the concern over the medical and meteorological aspects of photochemical air pollution and smog formation6 M. W. Mosher and N. J. Bunce Cunud. J. Chcm. 1971,49 28. J. F. McKellar and P. H. Turner Phorochem. and Phorobiol.1971 13 437; P. E. Joosting Chem. Weekblad 1971 67 1; J. A. Wisse ibid. p. 19; J. Van-Ham and H. Nieboer ibid. p. 15.