. _ I1, ._' o I n' res-4' ‘ I q: a ‘4 Wk ....,‘..‘.... .Vu PART 3. ' SYNTHESIS AND awn-ammo ~- mmemm or Epoxvames PART" > r assmsmmcm m, m w-MHHANE momsomawmus mm for the Degreeot-Pn.xn. * meg-um STATE UNIVERSITY ' , ‘PLAU‘L’BgLAVRM. ‘ /fi\/ 6K\ ABSTRACT LN , PART I SYNTHESES AND ACID-CATALYZED REARRANGEMENT OF EPOXYENONES PART II REGIOSPECIFICITY IN DI-n-METHANE PHOTOISOMERIZATIONS BY Paul B. Lavrik The epoxidation of several conjugated dienones and diene esters with molecular oxygen was studied in Part I of this thesis. Compounds containin the moiet re resented b 1 are easil e oxidized at the 8 Y P Y N Y P C===C‘—-C===C‘-C==H) -———————> C-—-C-—-C:::C-—C:::O RN 1, y,6 double bond to give % when heated in the presence of air or oxygen in a solvent which has readily abstractable hydrogen atoms. Thus, 4,5-epoxy—2,3,4,5,6,6-hexamethy1-2-cyclohexenone (g4), 4,5-epoxy—3,4,6,6- tetramethyl-Z-cyclohexenone (gé), 4,5-epoxy-2,6,6-trimethy1-2-cyclohept- enone (gg), 4-(l,2-epoxy-2,6,6-trimethy1-l-cyclohexyl)—3-buten-2-one (%Z)’ 3a,4a-epoxy-S-cholesten-7-one (fig), and methyl 4,5-epoxy-2-hexenoate (22) were prepared from their corresponding unsaturated precursors. The acid—catalyzed rearrangement of epoxide fig, 3,5-epoxy-2,3,4,S,6,6- hexamethyl-Z-cyclohexenone, was also studied in Part I of this thesis. In aqueous acid epoxide gg rearranged quantitatively to S—hydroxy-4- methylene-2,3,5,6,6-pentamethy1-2-cyclohexenone (§£)' In neat trifluoro- acetic acid (TFA), ii rearranged to 5-isopropeny1-4-methy1ene-2,3,5- trimethyI-Z-cyclopentenone (§2) which, on longer treatment with TFA, was dealkylated to 4-methy1ene-2,3,S-trimethyl-Z-cyclopentenone (ii) and acetone. When gfi was treated directly with neat TFA, there was formed, in addition to £2 and £1, a small yield of 2-acety1-2,3,4,5,5-penta- methyl-S-cyclopentenone (ii) and a larger amount of 4—acety1-2,3,4,5,5- pentamethyl-Z-cyclopentenone (£§)- Appropriate deuterium-labeling experiments allowed fairly complete mechanistic schemes to be postulated. In Part II of this thesis, the acetone-sensitized di-n-methane photoisomerization of syn- and anti-Q-methyl-l,4-dihydro-l,4-ethano- naphthalene (Egg and Egg) was studied. Of the two possible products which might arise from such rearrangement, the product resulting from the diradical in which the unpaired electrons are close to rather than remote from the methyl substituent on the saturated bridge was preferred. The extent of the preference depended on the geometry of the methyl-substituent and the aromatic ring. When the methyl was anti to the aromatic ring the reaction was regiospecific and yielded anti-6- 2.8 methy1-3,4-benzotricyclo[3.3.0.0 ]oct-3-ene (3Z3). When the methyl was syn to the aromatic ring, some syn-7-methy1-3,4-benzotricyclo 2.8 [3.3.0.0 ]oct-3-ene (figs), was obtained, but the reaction was still highly regioselective yielding mainly syn-o-methyI-S,4-benzotricyclo [3.3.0.02’8]oct—3-ene (QZR)’ In addition to the di-n-methane products, photoreduction products of ééfi and éfia were also obtained. The structures of all photoproducts were determined by independent syntheses. PART I SYNTHESES AND ACID-CATALYZED REARRANGEMENT 0F EPOXYENONES PART II RBGIOSPECIFICITY IN DI-n-METHANE PHOTOISOMERIZATIONS BY Paul B? bavrik A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1974 ACKNOWLEDGMENTS I am deeply indebted to Professor Harold Hart for his guidance, encouragement and leadership throughout the course of this study. I gratefully acknowledge financial support from the National Science Foundation, National Institute of Health and Michigan State University in the form of research and teaching assistantship. To my family, I express appreciation for their relentless support and encouragement. Also, I thank Nancy for her patience, encouragement and many other contributions among which the typing of the manuscript was but a small fraction. ii TABLE OF CONTENTS PART I SYNTHESES AND ACID-CATALYZED REARRANGEMENT OF BPOXYENONES INTRODUCTION I O O O O O I O O O O O O O O O O O I O O O O O 0 RESULTS AND DISCUSSION . . . . . . . . . . . . . . . . . . . . A. B. Autoxidation of Conjugated Dienones and Diene Esters to Y , 6 -Epoxyenones o o o o o o o o o o o o o o o o o o Acid-Catalyzed Rearrangement of 4,5-Epoxy-2,3,4,5,6,6- hexamethyl-Z-cyclohexenone ($1) . . . . . . . . . . . Thermal and Photochemical Rearrangements of 7,6- EP oxyenone S O O O I O O O O O O O O O O O I O 0 O O O EXPERIWNTAI‘ O O O O O O O O O O O O O O O O O O O O I O O O C General Procedures. . . . . . . . . . . . . . . . . . 4,S-Epoxy-2,3,4,S,6,6-hexamethy1-2-cyclohexenone (fig) 4,5-Epoxy-3,4,6,6-tetramethy1-2-cyclohexenone (££)° . 4,5-Epoxy-2,6,6-trimethyl-2-cycloheptenone (g2) . . . 4'(1,2-Ep0xy-2,6,6-trimethy1-l-cyclohexyl)-3—buten- 2"one (a) o o o o o o o o o o o o o o o o o o o o o 0 3a, 4a-Epoxy-S-cholesten-7-one (%£) . . . . . . . . . Methyl 4,5-epoxy-2-hexenoate (£2) . . . . . . . . . . Autoxidation of Methyl 2,4-Hexadienoate in the Presence of a,a'-Azobis-iso-butyronitri1e. . . . . . . . . . . Autoxidation of Methyl 2,4-Hexadienoate in the Presence Of 4‘t-BUty1C3teCh01. o o o o o o o o o o o o o o o 0 iii Page 11 11 20 35 39 39 4O 41 42 43 44 45 45 46 TABLE OF CONTENTS (continued) Page J. Autoxidation of 2,3,4,5,6-6-Hexamethyl-2,4-cyclo- hexadienone ( ) and 2,3,4,4,S,6-Hexamethyl-2,S- cyclohexadienone (£3). . ......... . . . . . . 46 K. Autoxidation of Methyl 2-Butenoate (3%). . . . . . . . 47 L. 4-Methylene-5-hydroxy-2,3,5,6,6-pentamethy1-2- cyclohexenone (3%) . . . . . . . . . . . . . . . . . . 47 M. Rearrangement of 4,5-Epoxy-2,3,4,S,6,6-hexamethy1-2- cyclohexenone (£3) in Neat Trifluoroacetic Acid. . . . 48 N. Cleavage of 2-Acetyl-2,3,4,5,5-pentamethyl-3-cyclo- pentenone ($3) With Base . . . . . . . . . . . . . . . 51 0. Treatment of 5-Isopropenyl-4-methylene-2,3,S-trimethy1- 2-cyclopentenone ($3) With Trifluoroacetic Acid. . . . 52 P. Rearrangement of S-Hydroxy-4-methylene-2,3,5,6,6- pentamethyl-Z-cyclohexenone (fig) in Trifluoroacetic Acid 52 Q. Rearrangement of Labeled 4,5-Epoxy-2,3,4,S,6,6-hexa- methyl-Z-cyclohexenone in Trifluoroacetic Acid . . . . 53 R. Thermal Rearrangement of 4,5-Epoxy-2,3,4,5,6,6- hexamethyl-Z-cyclohexenone (£1). . . . . . . . . . . . 53 S. Photochemical Rearrangement of 4,5-Epoxy-2,3,4,S,6,6- hexamethyl-2-cyclohexenone (gfi). . . . . . . . . . . . 54 T. Irradiation of 4-Acety1-2,3,4,5,5-pentamethy1-2- cyclopentenone (£3). . . . . . . . . . . . . . . . . . 54 U. Thermal Rearrangement of 4,5-Epoxy-2,6,6-trimethyl- 2-cycloheptenone (£2). . . . . . . ...... . . . . 55 PART II REGIOSPECIFICITY IN DI-fl-METHANE PHOTOISOMERIZATIONS INTRODUCTION ..... . . . . . . . . . . ....... . . . . 57 RESULTS AND DISCUSSION. . . . . . . . . . . . . . . . . . . . . 64 A. Synthesis of syn- and Anti-Q-methyl-l,4-dihydro-l,4- ethanonaphthalene, 2&2 and £32 . ....... . . . . 64 iv TABLE OF CONTENTS (continued) Page B. Photolysis of syn- and Anti-Q-methyl-l,4-dihydro-l,4- ethanonaphthalene, £33 and éga. . . . . . . . . . . . . 67 C. Conclusions . . . . . . . . . . . . . . . . . . . . . . 83 EXPERIMENTAL . . . . . . . . . . . . ....... . . . . . . . 87 A. General Procedures. . . . . . . . . . . . . . . . . . . 87 B. 9-Methylene-l,4-dihydro-1,4-ethanonaphthalene (22). . . 87 C. syn-Q-methyl-l,4-dihydro-l,4-ethanonaphthalene (£83). . 88 D. Dimethyl 9-methy1ene-l,2,3,4-tetrahydro—l,4-ethanonaph- thalene-z,3-dicarboxy13te (2%). o o o o o o o o o o o o 90 B. Dimethyl Anti-9-methyl-1,2,3,4-tetrahydro-l,4-ethano- naphthalene-2,3-dicarboxylate (8%). . . . . . . . . . . 91 F. Anti-9—methyl-1,2,3,4-tetrahydro-l,4-ethanonaphthalene— 2,3-dicarboxylic Acid (2%)' . . . . . . . . . . . . . . 92 G. Anti-9-methy1-l,4-dihydro-l,4-ethanonaphtha1ene (£83) . 93 H. Irradiation of syn-Q-methyl-l,4-dihydro-1,4-ethano- naphthalene (gsvsh) O I O O O O O O O I O O O O O O O O O 94 I. Irradiation of Anti-Q-methyl-l,4-dihydro—l,4-ethano- naphthalene (38v?) 0 O O O O O O O O O O O O I O O O O O 96 J. syn-Q-methyl-l,2,3,4-tetrahydro-1,4-ethanonaph- thalene (w) 0 O O O O O O O O O O O O O O O O O O O O 9 7 K. Anti-9-methy1-l,2,3,4-tetrahydro-l,4-ethanonaph- thal ene (fl) 0 O C O O O O O O O O O O O O O O O O O O 97 L. 7-Methy1ene-3,4-benzotricyclo[3.3.0.02’8]oct-3-ene (22) 98 M. syn—7-methyl-3,4-benzotricyclo[3.3.0.02’8]oct-3-ene (223) 98 N. 3,4-Benzotricyclo[3.3.0.02’8]oct-3-ene-6-one (ZR) . . . 99 O. syn-é-hydroxyl-o-methyl-3,4-benzotricyclo[3.3.0.02’8]- OCt-s-ene (z%)o o o o o o o o o o o o o o o o o o o o o 101 P. syn-6-methyl-3,4-benzotricyclo[3.3.0.02’8]oct- S-ene (gig) o o o o o o o o o o o o e o o o o o o o o o 102 Q. Hydrogenation of Anti-6-methyI-3,4-benzotricyclo- [3.3.0.02’8]oct-3-ene (21%) . . . . . . . . . . . . . . 103 TABLE OF CONTENTS (continued) Page R. Anti-o-methyl-S,4-benzotricyclo[3.3.0.02’8]oct- s-ene'7’0ne (a) o o o o o o o o o o o o o o o o o o o 103 S. Anti-G-methyl-S,4-benzobicyclo[3.3.0]oct- 3-ene-7—One (a) o o o o o o o o o o o o o o o o o o o 104 T. Anti-6-methyl-3,4-benzobicyclo[3.3.0]oct-3-ene (73a) . 105 U. 3,4-Benzobicyclo[3.3.0]oct-3-ene-6-one (lg). . . . . . 106 V. Anti-7-methyl-3,4-benzobicyclo[3.3.0]oct- 3'ene-6-One (a) o o o o o o o o o o o o o o o o o o o 107 W. Anti-7-methy1-3,4-benzobicyclo[3.3.0]oct-3-ene (78). . 107 X. syn-7-hygroxy- -anti- -6-methyl- -3 ,4-benzotricyclo- [3.3.0. 0 ' 8]oct- 3- -ene (79) . . . . . . . . . . . . . . 108 Y. Anti-o-methyl-S,4-benzobicyclo[3.2.1]oct- 3-8118-7-0118 (a) o o o o o o o o o o o o o o o o o o o 109 Z. Anti-6-methyl-3,4-benzobicyclo[3.2.0]oct-3-ene (73b) . 110 AA'. Hydrogenation of syn-6-methy1- -3, 4- -benzotricyclo- [3. 3.0. 02,8]oct——3-ene (67s). . . . . . . ..... . . 111 APPENDIX. 0 O O O O O O O O O O O O O O O O O O O O O O O O O O 113 BIBLIMWHY. O O O O O O O O O O O O O 0 O O O O O O O O O O O 137 vi LIST OF TABLES TABLE Page I. The Effect of Solvent on the Autoxidation of 1,2,3,4,5, 6,6-Hexamethyl-2,4-cyclohexadienone (£3). . . . . . . . . 13 II. Epoxyenones Prepared Using Molecular Oxygen . . . . . . . 14 III. The Effect of a Radical Initiator on the Autoxidation of Methyl Sorbate in Xylene. . . . . . . . . . . . . . . . . 17 IV. 270 MHz NMR and Spin gegoupling of Anti-6-methy1-3,4- benZOtriCYCIO [3 o 3 o 0 o 0 ’ ]OCt-S'ene (m) o o o o o o o o o 74 vii FIGURE 10 11 12 13 14 15 LIST OF FIGURES NMR Spectrum of 4,5-Epoxy-2,6,6-trimethy1-2-cyclo- A heptenone (£9) 0 O O O I O O O O O O O O O I O O I O 0 NMR Spectrum of Methyl 4,5-epoxy-2-hexenoate (£2) . . NMR Spectrum of syn-Q-methyl-l,4-dihydro-1,4-ethano- naphthalene (ISV‘Bli) O O O O O O I O O O O O O O O O O 0 NMR Spectrum of Anti-Q-methyl-l,4-dihydro-1,4-ethano- naphthalene (22:) . . . . . . . . . . . . . . . . . . NMR Spectrum of syn-Q-methyl-l,2,3,4-tetrahydro-1,4- ethanonaphthalene (6&3) . . . . . . . . . . . . . . . NMR Spectrum of syn-7-methy1-3,4-benzotricyclo- [3.3.0.02’8]OCt-3‘ene (22%) o o o o o o o o o o o o 0 NMR Spec§r§m of syn-6 -methyl- -3 ,4-benzotricyclo- [3.3.0.0 ]oct- -3- -ene (67s) . . . . . . . . . . . . . NMR Spectrum of Anti-Q-methyl-l,2,3,4-tetrahydro- 1,4-ethan0naphthalene (653) o o o o o o o o o o o o 0 NMR Spectrum of Anti-G-methyl- -3 ,4-benzotricyclo- [3. 3. 0. 02’8 ]oct- -3- -ene (67a) . . . . . . . . . . . . . 270 MHz NMR Specs? of Anti- 4-methy1- 3, 4-benzo- tricyclo[3.3.0. 0 ]oct- 3-ene (67a) . . . . . . . . . Double Irradiatign8 of Anti-6-methy1-3, 4-benzo- tricyclo[3.3.0. 0 ’ 8]oct- 3- -ene (67a) . . . . . . . . . NMR Spectrum of 7-Methy1ene-3,4-benzotricyclo- [3.3.0.02’8]00t“3-en6 (22) o o o o o o o o o o o o 0 NMR Spectrum of -6-hydroxyl—6-methy1-3,4-benzo- triCYC10[3.3.0.0 , ]0Ct-3-ene (Z&)o o o o o o o o o 0 NMR Spectrgm of Anti-6-methy1-3,4-benzotricyclo- [30 30 0. 02 ]OCt-3‘6n6-7-One (Z3). 0 o o o o o o o o 0 NMR Spectrum of Anti-G-methyl-S,4-benzobicyclo- [3.3.0]0Ct-3-ene-7'One (Z3) 0 o o o o o o o o o o o 0 viii Page 113 114 115 116 117 118 119 120 121 122 123 125 126 127 128 LIST OF FIGURES (continued) FIGURE Page 16 NMR Spectrum of Anti-6-methyl-3,4-benzobicyclo- [3.3.0]0Ct'3'6ne (13%). o o o o o o o 0-. o o o o o o 129 17 NMR Spectrum of 3,4-Benzobicyclo[3.3.0]oct- S-ene-6-0ne (i2). 0 o o o o o o o o o o o o o o o o o 130 18 NMR Spectrum of Anti-7-methyl-3,4-benzobicyclo- [3.3.0]0Ct-3-ene-6'One (lg) o o o o o o o o o o o o o 131 19 NMR Spectrum of Anti-7-methyl-3,4-benzobicyclo- [3.3.0]oct-3-ene (g) I C O O O C O O O O O O O O O O 132 20 NMR Spectrum of gyg-7-hydroxy-anti-6-methyl-3,4-benzo- tricyclo[3.3.0.0 ’ ]oct-3-ene (72). . . . . . . . . . 133 21 NMR Spectrum of Anti-6-methyl-3,4-benzobicyclo- [3.2.1]0Ct'3‘ene'7-One ($2) a o o e o o o e o o o o o 134 22 NMR Spectrum of Anti-6-methyl-3,4-benzobicyclo- [3.2.1]oct-3-ene (73R). . . . . . . . . . . . . . . . 135 23 NMR Spectrum of syn-6-methyl-3,4-benzobicyclo- [3.2.1]oct-3-ene (£1) . . . . . . . . . . . . . . . . 136 ix PART I SYNTHESES AND ACID-CATALYZED RBARRANGEMENT OF EPOXYENONES INTRODUCTION The preparation of y,6-epoxyenones from conjugated dienones can be carried out by a number of methods, the most frequently employed and generally applicable of which is peracid oxidation.1 The versa- tility and selectivity of this reaction makes it the method of choice in most conversions of a double bond to the corresponding epoxide in a variety of monounsaturated or polyunsaturated systems. The most common peracids used fbr the conversion of dienones 2 represented by l to the epoxide K have been perbenzoic, monoper- phthalic,3 and m-chloroperbenzoic acids.4 Of these peracids, ’,0 c:c-c:c-—c:o + RC< _.___. c— — :c—c=o 00H 1 2 'b ’b m-chloroperbenzoic acid is the most convenient oxidizing agent. It is commercially available, reacts at a somewhat faster rate than perbenzoic acid, and is ideally suited for epoxidations which require long reaction times due to its excellent stability. The preferential oxidation of the 7,6-double bond over the a,B- double bond in conjugated dienones reflects the electrophilic nature of organic peracids. Epoxidation reactions proceed by an electrophilic attack of the peracid upon the double bond; thus the rate of epoxidation . . . . . . . 1 13 very sen51t1ve to the electron dens1ty at the olef1n1c Slte. In a 3 system represented by l the y,6-doub1e bond has a greater electron density than the a,B-double bond,5 and selectivity is easily achieved. The generality of epoxidation with organic peracids is impaired when the olefin or the epoxide is unstable to the carboxylic acid formed as a by-product.6 Also, it is essential that other functional groups present in the molecule be stable to the peracid employed. Finally, it should be noted that fOr systems such as A, selectivity towards oxidation of the y,6-doub1e bond is not achieved if a third double bond with similar or greater reactivity towards the peracid is present in the molecule. A number of modifications have been employed to overcome these major difficulties encountered with normal peracid epoxidation. For instance, epoxidation with a peracid in dichloromethane-aqueous sodium bicarbonate biphasic system permits, in some cases, the epoxidation of acid-sensitive olefins or the epoxidation of olefins yielding acid- sensitive epoxides.7 Another useful variant of the peracid method which shows promise in the preparation of acid-sensitive y,6-epoxyenones is the reaction of benzonitrile with alkaline hydrogen peroxide. The probable path, illustrated in the fbllowing equations, involves a peroximidic acid . . . . . . . . . 8 1ntermed1ate wh1ch read11y epox1d1zes olef1ns under neutral cond1t10ns. H O K CO NH 2 2 » 2 3 ¢ C H CEEN i; C H -C . 6 5 6 S O 4% ll"2 In addition to organic peracids, other methods of epoxidation have shown great promise in the synthesis of 7,6-epoxyenones. A very effec- tive route involves the dehydrohalogenation of bromohydrins or chloro- hydrins.9 This process is illustrated in the fbrmation of a steroidal 10 68,78-epoxide 3 from the corresponding bromohydrin g. 8W 8-: A particularly significant aspect of this method, besides its effective— ness in the synthesis of acid-sensitive epoxides, is that it often leads to epoxides isomeric in configuration to those obtained by peracid epoxidation.10 A method frequently employed fOr epoxidation of double bonds conjugated with a strong electron withdrawing group such as carbonyl is 1d,9 However, the selectivity of such a alkaline hydrogen peroxide. method is the reverse of organic peracids when applied to systems such as I. Thus, when B-ionone i is treated with alkaline hydrogen peroxide, it yields only the a,B-epoxide 2. In contrast, when B-ionone is treated with perbenzoic acid it gives the 7,6-epoxide 7.3a Similarly, treatment 11 of eucarvone or 4-benzoy1-1-pheny1buta-1,3-diene12 with alkaline hydrogen peroxide, even in excess, yields only the a,8-monoepoxides. \c C6H5C03H CH2C12 é Z Olefins have also been epoxidized by reaction with peroxides in 13 The neutral and mild the presence of vanadium and molybdenum salts. conditions of this method offers a potentially useful route into acid- sensitive epoxyenones. Besides the major epoxidation methods discussed above, other less general processes are available.9 Among these is the autoxidation of 14 olefins. This autoxidation process is not well understood but probably involves a free radical mechanism of the following type:15 R0- + RH —-I- ROH + R- There are several examples of epoxide formation by this approach; however, because of competition among other free radical processes, high yields desirable for laboratory preparations are rare.14 Nevertheless, given a system where these competitive processes are minimized, the autoxidation of alkenes to epoxides may be of some synthetic and preparative value. The autoxidation of'axo,exo-3,4,6- triphenylbicyclo[3.1.0]hex-2-ene g to give epoxide a in 76% yield exemplifies this method.16 H Ph Ph Vw :2 Ph P V Ph Ph § 2 Among the synthetic uses of epoxidation products, the acid-catalyzed rearrangement of epoxides are of special interest, since they provide a 6,17 simple means of converting olefins to carbonyl compounds. Thus, in the presence of aluminum trichloride, trimethylethylene oxide *2 re- arranges to methyl isopropyl ketone llzls 0 “3C\ A /H A1c13 “3C\ a /c-c , /CH— -CH3 H3O \CH3 H3C 10 ll 'Vb 'Vb In general, Lewis acids (i.e., aluminum trichloride, magnesium bromide or boron trifluoride etherate) are the usual catalysts, although aqueous mineral acids and non—nucleophilic carboxylic acids such as trifluoroacetic acid are also used. Rearrangement reactions of this kind become especially useful if one can predict in any particular caSe what the major product will be. The major product formed from the rearrangement of an epoxide is governed by two main factors: the direction of ring opening, and the relative migratory aptitudes of the different sub- stituents. The direction of ring opening in acid-catalyzed rearrangements may in general be predicted simply on the basis of the relative ease of ionization of the two carbon-oxygen bonds in question. By analogy with other situations in organic chemistry, the sequence is expected to be tertiary > secondary > primary. The rearrangement of trimethylethylene oxide lg to methyl isopropyl ketone ll illustrates this rule. Similarly, isobutylene oxide gives largely isobutyraldehyde and propylene oxide yields propionaldehyde.18 Aryl groups are superior to alkyl groups in increasing the ease of heterolysis of the carbon-oxygen bond. For example, in indene oxide l2 cleavage of the C-0 bond which leads to an incipient benzylic carbonium ion is preferred, and 2-indanone 12 results.19 In fact, monoaryl-sub-' stituted epoxides invariably rearrange to give nonconjugated ketones. '0 ‘ MgBrz "' = [ o W] Et20 Br' .13. a 9‘”? 8 Vinyl substitution, like phenyl substitution, weakens 3 C-0 bond with respect to ionization. However, there are a number of reports of rearrangement of epoxides involving cleavage of the bond to a vinyl rather than a phenyl substituted carbon atom.6b For example, the cyclohexenyl epoxide l1 rearranges to aldehyde 1i? whereas the cyclo— hexyl analog l2 rearranges with cleavage of the other C-O bond to give ketone 17.20 Mb 0 r: I E r: 0‘ :2 U1 0 C H (311/ij MgBrz § 6 5 ,/ Et20 H O / \ MgBr2 ' C6H5CH"CH O 5 C6H5CH2(! EtZO 17 M m Once the direction of ring opening has been decided, in epoxide 12, for example, by the circumstance that a tertiary carbonium ion is formed, there remains a choice between hydrogen migration or alkyl migration. This will be decided by the relative migratory aptitudes of the groups and in general this order is aryl > acyl > H > ethyl > methyl.‘ Among the few exceptions to this order of migratory aptitudes, are those which occur when fer steric reasons hydrogen migrates in preference to phenyl.6 In certain cases, the rearrangement may constitute a ring expansion or ring contraction, as shown by the following two exampleszn’22 As illustrated by epoxide £2, acid-catalyzed rearrangement of a,B-epoxyketones is an extremely useful route to dicarbonyl compounds. This rearrangement has been intensively studied by House and co- workers.23’24 As with other epoxides, the rearrangement of an a,B- epoxyketone (22) under the influence of acids can also proceed in two directions: 10 However, with only few exceptions,25 the rearrangement proceeds through the more stable carbonium ion A and the observed products depend only on the migratory aptitudes of the groups attached to the epoxide ring. In contrast to o,B-epoxyketones, the acid-catalyzed rearrangement of y,6-epoxyenones such as 2 have been completely neglected in the literature. A detailed study of this rearrangement should yield valuable information concerning its synthetic utility as well as information concerning the mode of epoxide ring opening and group migration in such systems. It is the purpose of this part of this thesis to examine the gener- ality of autoxidation of conjugated dienones to 7,6—epoxyenones and to study the acid catalyzed rearrangement of such epoxides. RESULTS AND DISCUSSION A. Autoxidation of Conjugated Dienones and Diene Esters to y,6-Epoxyenones In an attempt to obtain a Diels-Alder adduct between 2,3,4,S,6,6- hexamethyl-Z,4-cyclohexadienone £3 and hexamethyl-Dewar benzene by refluxing the two components in xylene, a white solid, mp 48-49.S°, was isolated which was identified as epoxide 21. The infrared spectrum showed a band at 840 cm-1 which is characteristic for epoxides26 and a carbonyl band at 1680 cm-1, indicating conjugation with a double bond. The conjugation of the carbonyl group was corroborated by the compound's ultraviolet spectrum, which showed Amax (cyclohexane) 253 nm (e 7230), 324 (240). The nmr spectrum (assignment and europium shift data are given below) was also consistent with structure 33. Finally, epoxide 2% was conclusively identified by independent synthesis using m-chloro- perbenzoic acid. The unexpected formation of epoxide it led to further investigation of this reaction. The hexamethyl-Dewar benzene obviously had nothing to do with the reaction and was omitted from further studies of the epoxi- dation. Additional attempts to effect a Diels-Alder addition of hexa- methyl-Dewar benzene to dienone 23 in the absence of air were unsuccessful. When dienone 23 was heated in xylene at 120-130° for 21 hr in the presence of air, epoxide 21 was fermed in 65-75% yield. No reaction was observed when dienone 2% was heated in xylene at 120-130° under a 11 12 CD 1.38 2.96 I 1.82 C ) 02, xylene (3.00) 1.01 (2.33) 4. 120-1300 2.10 1.48 (1.12) (1.30) 1.57 (1.00) g; 24 ‘Vh nitrogen atmosphere for 48 hr or when dienone 23 was refluxed in chlorobenzene for 48 hr in the presence of air. When dienone 23 was heated in cyclooctane, cumene or mesitylene at 120-130°, epoxide g: was formed, generally more rapidly and in better yield than when xylene was used as solvent. However, xylene yielded relatively smaller amounts of side products, thus making it the most practical solvent for the reaction from a preparative viewpoint. These results are summarized in Table I. When dienone 23 was heated in xylene for longer than 21 hr (36-48 hr), the yield of epoxide 23 decreased and a significant amount of polymer was formed. Also, there was very little advantage in reaction time, yield or reproducibility by bubbling air or oxygen through the reaction mixture. To explore the generality and scope of this epoxidation method, the reaction was carried out using a variety of dienones. Table II lists several epoxides which have been prepared by autoxidation of the corresponding unsaturated precursors. In all cases, the dienone or diene ester was heated in xylene at 120-130° for 24 hr. No attempt 13 Table I. The Effect of Solvent on the Autoxidation of l,2,3,4,S,6,6- Hexamethyl-Z,4-cyclohexadienone (2%). Solvent Reaction Time Yield of Epoxide 243 Xylene 21 hr 65-75% Cyclooctane 13 70-75 Mesitylene 14 80 Cumene 8 90-95 Chlorobenzene 48 0 aAbsolute yields obtained by nmr versus an internal standard. was made to optimize yields. The reactions of 23-28 were perfbrmed on a 100 mg scale. When the oxidation of 23 was scaled up from 100 mg to 3 g, the yield improved from 49% to 90%. Epoxides 2E and 22 have not been previously described. Their structures were established by their spectral properties and independent synthesis using m-chloroperbenzoic acid. Epoxide £2, 4,5-epoxy-3,4,6,6- tetramethyl-2-cyclohexenone, was obtained as colorless crystals, mp 48.0-50.0°, from the corresponding dienone.27 The infrared spectrum showed a carbonyl absorption at 1680 cm'l, indicating conjugation with a double bond, and an epoxide band at 890 cm-1. The ultraviolet spectrum also showed that the carbonyl and carbon-carbon double bond were con- jugated; Amax (95% ethanol)247 nm (e 7200), 326 (143). The nmr spectrum was consistent with the assigned structure. Epoxide £2, 4,5-epoxy-2,6,6-trimethy1-2-cycloheptenone, was obtained 2 . . from 2,6,6-trimethyl-2,4-cycloheptadienone (eucarvone) 8 as a clear l1qu1d. 14 Table II. Epoxyenones Prepared Using Molecular Oxygen Epoxide No. Yield, %a o I 25 51 'VD o o | 26 52 ’VD C) 27 72 m 23 31 «A; H3C‘ A H 33 9° 3Yield of isolated epoxide, corrected for recovered starting material. 15 The infrared spectrum showed a carbonyl band at 1670 cm.1 , indicating conjugation with a double bond, and a band at 875 cm'l, due to the asymmetric stretching of the epoxide ring. The ultraviolet spectrum showed a Amax (95% ethanol) at 246 nm (e 7920). The nmr assignment, based primarily on the splitting pattern and europium shift data, is shown below. The long range coupling between HD and HF and the assign- ment of protons HE and “F was confirmed by deuterium labeling. Prepar- ation of 26 from eucarvone deuterated at C-729 gave 26-d2, lacking ’VM WW methylene signals at 6 2.85 and 2.08. In addition, the signal at 6 3.00 changed from a "doublet of doublets" to a doublet. Fl 2.85 (2.85) (1.51) 1.83 HQC """" HE 2.08 (2.49) CH3 0.98 (1.22) (1.22) 6.43 H “CH3 1.25 (1.00) 3.22 (2.42) H Hr) 3.00 (2.42) 53% 16 HA = d, JAB = 1.6 Hz HB = q of d, JBC = 6.1 Hz, JAB = 1.6 Hz HC = d of d, JBC = 6.1 Hz, JCD = 4.2 Hz HD = d of d, JCD = 4.2 Hz, JDE = 1.4 Hz HE = d of d, JEF = 13.0 Hz, JDE = 1.4 Hz H = d, JEF = 13.0 Hz Epoxides 21-22 have been previously described in the literature.30-32 However, in order to confirm their structures, these epoxides were syn- thesized independently using m-chloroperbenzoic acid. In every case, the nmr and infrared spectra of the epoxide obtained with molecular oxygen was identical to the spectra of the epoxide obtained with m-chloro- perbenzoic acid. The spectral properties of these epoxides are given in the experimental section. To probe the effect of radical initiators on the epoxidation.with molecular oxygen, the autoxidation of methyl sorbate in the presence of 0,0'-azobis-iso-butyronitrile was studied. The results are summarized in Table III. There is some advantage in reaction time and yield by using a radical initiator. However, such a procedure is of little practical value since it increases the amount of undesired side-products making isolation of the epoxide more difficult. The epoxidation with molecular oxygen can best be explained by a free radical mechanism as illustrated in Scheme 1. The autoxidation may be initiated by reaction of triplet oxygen with the hydrocarbon 17 Table III. The Effect of a Radical Initiator on the Autoxidation of Methyl Sorbate in Xylene. Concentration Reaction Reaction Epoxide 29, % Unreacted of AIBN, M Temp. Time, hr. (% yield)a Methyl Sorbatea 0.01 80-90 6.0 45.0 50.0 0.02 80-90 6.0 47.0 45.0 0.04 80-90 6.0 50.3 41.5 none 120-130 6.0 24.0 70.0 0.01 80-90 20.0 71.2 18.8 none 120—130 20.0 61.0 28.0 aAbsolute yields obtained from the mixture. nmr spectrum of the crude reaction solvent to form a peroxy radical which in turn adds to the conjugated dienone. The epoxide can then be SCHEME 1 formed by an 5H2 reaction on the 0-0 R. + 02 --—-9 R02. r ‘ R02. + C:C—C:C—C:0 -——. RO-O-C-C‘J-l 2; 2:. :0 A a C m RO-O-(P-C'JCZ-"C'L'C'J-‘O —’ R0‘ + C-IC-C=C-C=O C m RO° + RH ——. ROH+ R- 18 bond in radical g. Consistent with this mechanism is the lack of oxidation in chlorobenzene and the formation of typical oxidation products from the solvent (p-methylbenzyl alcohol from xylene, cumyl alcohol and acetophenone from cumene, etc.). Also, epoxide fermation was completely suppressed by addition of 4-t-buty1catechol, a radical inhibitor, to the reaction mixture. As a further probe into the nature of this autoxidation process, the reaction between t-butyl hydroperoxide and dienone 23 in chloro- benzene was examined. Under these conditions two products were observed, o o o I l l + t-BuOOH “1 = + 100°, 12 hr 0 (Md 23 24, 35% 32, 48% 'Vb ’V'b ’Vb epoxide 24 and a hydroxy ketone assigned structure 3%. A plausible mechanism for epoxide formation with an alkyl hydroperoxide is presented in Scheme 2. The formation of epoxide 24 with t-butylhydroperoxide clearly supports the mechanism presented in Scheme 1 in which the peroxy radical is the epoxidizing agent. A detailed discussion of the structure and formation of hydroxy ketone 32 will be presented later. As illustrated by the reaction of 28 and 22, the autoxidation of dienones and diene esters to y,6-epoxides may be highly stereoselective. Thus, with 3,5-cholestadien-7-one the product is the a—epoxide and with methyl sorbate the E,E-epoxide $2 is the predominant product. In the case of methyl sorbate, however, the nmr spectrum shows traces of another isomer, probably cis at the epoxide ring. 19 SCHEME 2 RO-OH —’ RO° + °OH ROO-H + ~OH -—b ROO- + HOH . i 1 R00- + c=C~C=C-C=O --+ RO—O—C—C"—= '-" == 2:0 C . m r A ’W RO-O—C—CgC‘J—‘C'é'C’g ——D RO' + C"lC-"C=C'-C=O C '1. RO’ + ROOH —9 ROH + ROO- In all cases studied, the epoxides obtained by autoxidation are identical to those obtained using m-chloroperbenzoic acid. However, isolated double bonds or singly conjugated systems such as 22 and 31 are not oxidized by molecular oxygen, so that the method provides a selectivity not always available with a peracid. Thus, though dienone 32 is easily epoxidized with m-chloroperbenzoic acid even at 0°,33 NCOZCHs $8 20 an equimolar mixture of £3 and 32, heated in xylene for 24 hr, gave a 70% yield of 24 and essentially quantitative recovery of unreacted 32. Clearly, epoxidation of dienones and diene esters with molecular oxygen offers a valuable route to 7,5—epoxyenones and y,5-epoxyenoates. The selectivity towards oxidation at the y,5-double bond, the neutral and mild reaction conditions, and the simplicity of this reaction makes it a potentially useful alternative to epoxidation of certain classes of compounds with organic peracids. B. Acid-Catalyzed Rearrangement of 4,5-Epoxy-2,3,4,S,6,6—hexamethyl- 2-cyclohexenone (24). ’VD The epoxyenone 24, described in the previous section, was obtained in high yield by treating the corresponding dienone 23 with m-chloro- perbenzoic acid. Epoxide 21 could be purified by chromatography over Florisil or neutral alumina but it was unstable to weakly acidic or basic conditions. For instance, chromatography on silica gel or treatment with a little aqueous acid, resulted in nearly quantitative rearrangement to a hydroxy ketone assigned structure 32, based on its spectral properties and further rearrangements in stronger acid (vide infra). Similarly, if after reacting dienone £3 with m-chloroperbenzoic acid, the reaction mixture is washed with 10% NaZSO3 and saturated NaHCO3 solutions to remove excess peracid and acid, epoxide 24 is converted to hydroxy ketone 32. The infrared spectrum of 32 showed a carbonyl band at 1670 cm‘1 which indicates conjugation with both double bonds, 3 terminal methylene group band at 960 and 930 cm"1 , and hydroxyl group bands at 3620 (sharp, free OH), 3590 (sharp, intramolecular n-H bond), 3500 cm'1 (broad, intermolecular H bond). The conjugation of the carbonyl 21 group with two double bonds is corroborated by the compound's uv spectrum which showed lma at 282 nm (a 11,700), 273 (15,200) and 266 (14,500). X The nmr data given below are also consistent with structure 32. ’Vb i 1.85 1.10 (2.82) H+ (trace) (3-29) 0.96 (2.37) " 1.22 (1.75) 2.08 (1.35) OH 1.85 5.43 (1.00) 5.64 (1.22) 24 32 Hydroxy ketone 32 is undoubtedly formed from 24 by proton loss from the intermediate cation 2 (Scheme 3). The alternative epoxide ring-opening mode to give E would lead to structure 33, which is also reasonably consistent with the nmr spectrum, but which is less con- sistent with the uv and ir data and is conclusively eliminated by labeling results. Preparation of 24 from dienone 23 containing CD3 at C-3 and C-534 gave the epoxide Efigfi6 lacking methyl signals at 6 1.48 and 2.10. Treatment of 22826 with dilute acid gave the hydroxy ketone 3%;26 lacking methyl signals at 6 1.22 and 2.08. If the hydroxy ketone had structure 33, the product would have contained only five deuteriums and would have lacked the vinyl proton signals. Treatment of epoxyenone £4 with neat trifluoroacetic acid gave four products, only two of which were isomers (mass spectrum) of the starting epoxide. These products were assigned structures 34-31. The 22 SCHEME 3 O O 0 11+ -———+ ——> 0 9H CH 24 P. «A. ‘\ “\ l_H+ V (3 C) CN4 product ratios depend on reaction time and temperature, those shown being for 20 hr reaction time at room temperature. The product structures were established primarily by their spectral properties. Diketone 3435 showed two carbonyl absorptions at 1740 and 1710 cm'l, and only weak uv absorption, indicating that neither carbonyl group was conjugated with the double bond. The nmr spectrum, with eurOpium shift data, is consistent with structure 34. 23 .____. O I TFA 34, 6% 35, 27% __________ mm mm 0 I i 14 a 0 *0 -————. 36, 33% 37, 34% M» M Cleavage of 34 with base supports the structural assignment. With sodium methoxide in methanol, 34 afforded the conjugated cyclopentenone 38. Cleavage of the acetyl group should give the allylic anion E, which 0 O (1.36) 1.05 1.98 (4.22) (1.61) 1.08 1.27 (3.58) 1.72 (1.00) 1.60 (2.09) SE 24 is protonated in the y-position to give 38. Structure 32, which could be formed by a-protonation, is eliminated from consideration by the ir and uv data. The infrared spectrum of 38 showed a carbonyl band at 1700 cm'l, indicating conjugation of a cyclopentanone carbonyl group with a double bond, and a band at 1650 cm'1 due to carbon-carbon double bond absorption. The uv spectrum showed a A at 235 nm (e 6000) max also indicating conjugation with the carbonyl group. The nmr data given above are consistent with structure 38. NaOCH3 CH30H 8’11 (o——— 0 (2-00) 0-90 1.63 (2.76) I H (1.95) 0.99 9 1. . ( 00) 1 01 +4 1.95 (1.23) J = 8.0 R233 (1.90) 38 39 "Vb m Diketone 32 also showed two carbonyl absorption bands at 1710 and 1660 cm'l, but unlike 34, 35 showed a uv absorption at 237 nm (e 7720), 25 0.90 (2.25) (2.52) 1.75 1.03 (2.08) 1.95 (1.26) 3.5 indicating that one of the carbonyls was conjugated with the double bond. The nmr data are in agreement with the assigned structure; the shift reagent appears to coordinate primarily at the cyclopentenone carbonyl group. The base peak in the mass spectrum of 32 appeared at M-42 (loss of CH2=C=0) and the next most intense peak (rel. intensity 60) was at M-S7 (loss of CH2=C=0 and CH3). The product assigned structure g6 corresponded in analysis to loss of water from the epoxyenone 24. The infrared spectrum showed a carbonyl 'VV 0 4.95 (1.07 and 1.28) (3.04) 1.80 1.46 (1.90) 1.21 (2.28) (1.30) 2.13 4.95 (1.03) 5.30 (1.00) 58% 26 absorption at 1710 cm"1 which is consistent with a cyclopentenone carbonyl, 1 strong carbon-carbon double bond absorptions at 1645 and 1620 cm" , and a strong terminal methylene band at 915 cm'l. The uv maxima [340 nm (e 63), 270 (13,700)] were consistent with a conjugated dienone. The nmr data for structure 36 are given above. The product assigned structure 37 corresponded in analysis not only to loss of water from epoxyenone 24, but to loss of a C3H4 fragment as well. The uv spectrum was very similar to that of 36 [267 nm (2 15,200)], as was the ir spectrum (”C=O 1710, V=CH2 910 cm'l). The nmr chemical shifts and europium shift data are consistent with structure 31. The proton at 6 2.70 was readily exchanged at room temperature in NaOCH3- CH30D, causing collapse of the doublet at 6 1.17 to a singlet. O (3.04) 1.80 F4 2.70, q (4.37) J = 7.5 Hz 1.17, d (2.90) (1.30) 2.10 5.10 (1.00) 5.27 (1.00) S‘x’i 27 A plausible mechanism for the formation of 31 from $6 is given in Scheme 4. When the reaction was monitored by nmr, a sharp singlet, which increased in intensity as the reaction progressed, appeared at 6 2.33. This singlet was later shown to correspond to acetone in trifluoroacetic acid. SCHEME 4 n 0130013 + (012300) 20 28 Products 36 and 31 were also formed by acid—catalyzed rearrangement of hydroxy ketone 32. Thus, independent treatment of 32 with trifluoro- acetic acid at room temperature gave only 36 and 31, in ratios which depended on the reaction time. After 2 hr the product was 94% £2 and 6% 3g, whereas after 46 hr it was 3% 36 and 97% 31. The fOrmation of these products can be rationalized as in Scheme 5. Either 3% is a discrete intermediate as in the route 24 + 32 +'36 + 37 or it may be ’VD’V'U'VM bypassed through the intermediacy of E. If the reaction proceeds via SCHEME 5 PU + H ”“20 G 'b l) 1,2 acyl E shift m 2) -H+ Scheme 4 5% A 1 S84 29 intermediate g, dehydration may precede 1,2-acyl shift as shown in Scheme 5, or the order of these steps may be reversed. Unfortunately, no simple experiment can distinguish between the two possible routes. It is clear, however, that the remaining two rearrangement products 34 and 35 are not produced from hydroxy ketone 32. MD M ’Vb Possible routes to 34 and 32 are shown in Scheme 6. The first formed intermediate is once again 2. Ring contraction and proton loss lead to 32, whereas attack of a nucleophile at the carbon 0 to the SCHEME 6 o I H+ 1,2-alkyl 24 __.. 9 i ”W shift 014 D N -H+ TBA 35 '\/\a O o I CF-E—o TFA 3 \~’ _____.. 34 Mb 014 30 carbonyl group followed by an acyl shift and loss of hydroxyl lead to 31. The nucleophile shown in Scheme 6 is trifluoroacetic acid but it may also be water or even an intramolecular attack of the hydroxyl group at the C-5 position. This alternate route to 34 is shown in Scheme 7. Still another plausible route from £4 to 34, involving a cyclo- propylcarbinyl rearrangement, is shown in Scheme 8. This mechanism, however, was eliminated as a consequence of deuterium labeling experiments (vidé infra). SCHEME 7 Ch I/ QC 1,2 acyl shift 31 SCHEME 8 i 1 ___.:i:__. —_————' C) 014 14 H-—/C; 2‘3 N J: 8111 Several deuterium labeling experiments were carried out to test the mechanisms in Schemes S-8. The results are shown in Scheme 9. Experiments were conducted with trideuterioepoxide (24*) containing a CD3 group at C-3 and hexadeuterioepoxide (24*,f) containing CD3 groups at C-3 and C-5. The mechanisms leading from epoxyenone 24 to 35-37 are fairly obvious, ’VM’V); and the labeling results support the proposals in Schemes S and 6. 32 SCHEME 9 ——D o 3% 33 I 0 TFA , + O 313. I i H , + + + 'k * as: .31 The route from 21 to 34 is less obvious, and in order to ascribe meaning to the labeling results it was necessary to establish unequi- vocally the nmr assignments of 34. The europium shift data support the assignment shown in structure 34; however, there is some uncertainty because the molecule contains two functional groups with which coordination can occur. Since the nmr assignment of 28 is unambiguous from both chemical and europium shift data, the base cleavage of 34 to 38 was conveniently used in the nmr assignment of 34. Thus, when diketone 34 labeled with a C03 group at 6 1.27 (obtained from 34 containing a CD3 group at C-3) was cleaved 33 with base, the resulting cyclopentenone 38 lacked the allylic methyl signal at 6 1.63. Consequently, the methyl in 24 corresponding to the signal at 6 1.27 must be the methyl between the two carbonyl groups. The signals at 6 1.05 and 1.08 must therefore correspond to the gem-dimethyl group. 1.63 SE 52% A similar procedure was used to assign the allylic methyl signals. Treatment of'gé containing CD3 groups at C-3 and C-5 with acid gave a sample of 34 lacking the methyl signals at 6 1.27 and 1.72. Cleavage of this labeled 3% with base gave 38 lacking the allylic methyl signal at 6 1.63 and the doublet (J = 8.0 Hz) at 6 1.01. This result estab- lishes unequivocally that the allylic methyl furthest from the acetyl group is at 6 1.72, and that all the other assignments for 34 are correct as shown in the first structure. The labeling results clearly eliminate the mechanism in Scheme 8 as a possible route to 31. For the mechanism shown in Scheme 8 to be consistent with the labeling results, acid treatment of epoxyenone 34 containing CD3 groups at C-3 and C-5 should yield 34 lacking allylic methyl signals at 6 1.72 and 1.60. 34 1.63 NaOCH3 V CHSOH 1.01 SE 366' The labeling results fully support both the mechanisms in Schemes 6 and 7. A possible method to distinguish between these two routes to 34 would involve labeling the oxygen (180) in the epoxide ring of 21. The presence of 180 in 34 would support the mechanism in Scheme 7 whereas the absence of 180 would support the mechanism in Scheme 6. In summary, the epoxyenone £1 rearranges quantitatively in dilute acid to the hydroxy ketone 32 through the allylic cation R (Scheme 3). In less basic solvents (such as trifluoroacetic acid) the same inter- mediate may rearrange by a 1,2-alkyl shift to give 35 or may, following nucle0philic attack a to the carbonyl group, undergo ring contraction to give 34 (Scheme 6 or 7). Product 36 may arise from protonation of 32 or may be formed by the alternate method of epoxide ring opening (Via,§’ Scheme 5); product 31 is formed by dealkylation of’gg (Scheme 4). Although this is the first study of an acid-catalyzed rearrangement of a conjugated y,6-epoxyenone such as 24, it is clear that the prefer- ential (if not exclusive) mode of epoxide ring opening is to give the allylic cation. Rearrangement invariably leads to ring contraction 35 products; consequently, such reactions can be useful for the synthesis of cyclopentenones. However, the manner in which the various methyl groups or other substituents, may determine the mode of the acid- catalyzed rearrangement of conjugated epoxyenones remains to be explored. C. Thermal and Photochemical Rearrangements of y,6-Epoxyenones. Several exploratory experiments were conducted on the thermal and photochemical rearrangement of conjugated y,6-epoxyenones. The results of these experiments are described in this section. In general, thermal and photochemical rearrangements of epoxy- ketones parallel the acid-catalyzed rearrangement to the corresponding diketone.9’17 This proved to be the case with epoxyenone 24. When epoxide 24 was heated in a sealed tube at 200° for 2 hr, either neat or in carbon tetrachloride, it rearranged to three products. The two major products were identified as hydroxy ketone 32 and diketone 35. The minor product was not isolated. Since hydroxyketone 32 is easily o I O A 2000 ‘ 2’4 .32 3.5 36 formed from 24 by traces of acid or base (see previous section), it is not certain whether 32 formed under these conditions is truly the result of a thermal rearrangement. The photochemistry of epoxide it is more complex. Irradiation of 21 in cyclohexane with pyrex-filtered light gave diketone 35. However, the a,B-unsaturated diketone 35 is photolabile and under these same conditions rearranges further. Diketone 35 could be isolated in 10-20% yield using short photolysis times. On prolonged irradiation at approxi— mately 10°, the diketone 35 seems to undergo an oxa-di-fl-methane rear- rangement to give a compound which was tentatively assigned structure 39 based on its spectral properties. The nmr spectrum of 32 showed no allylic methyl signals; instead it had five sharp methyl singlets, at 6 0.80, 1.12, 1.23, 1.27, and 1.33, and a sharp acetyl methyl singlet at 6 2.20. The infrared spectrum showed two carbonyl bands, at 1760 and 1710 cm-1, corresponding respectively to the cyclobutanone and acetyl moieties.36 Unfortunately, due to its thermal instability, isolation of pure 39 was not possible. Additional work is necessary if structure 39 is to be unambiguously characterized. 37 This type of photochemistry is not new. A diketone similar to 35, 2,S-di-t-butyl-4-pivaloy1-2-cyclopentenone, has been reported to undergo a similar oxa-di-w-methane rearrangement to give a bicyclo- [2.l.0]pentan-2-one derivative.37 Nevertheless, the potential synthetic route from conjugated 7,6-epoxyenones to strained systems such as 39 warrants further investigation of the photochemistry of 24 and other related epoxides. An unusual thermal rearrangement was detected in the case of epoxide 22, 4,5-epoxy-2,6,6-trimethyl~2-cycloheptenone. When epoxide 22 was heated in a sealed tube at 180° or merely gas chromatographed at 170° (5 ft 10% FFAP column) it rearranged quantitatively to a new compound which was an isomer of 26 (mass spectrum). This isomer of 22 was assigned structure 42 based on its spectral properties. The infrared 1, indicating no conjugation with the double bond, and a strong C-O-C stretch band at 1075 cm'l. The spectrum showed a carbonyl band at 1720 cm" nmr data given below support structure 40. Analogous isomerizations are common with vinyl cyclopropanes; however, these isomerizations are rare with vinyl epoxides. 38 HA = d, JAB = 6.0 Hz HB = d of d, JAB = 6.0 Hz, JBC = 2.0 Hz HC = d of d, JBC = 2.0 Hz, JCD = 1.3 Hz HD = d of d, JDE = 17.0 Hz, JCD = 1.3 Hz HE = d, JDE = 17.0 Hz The exploratory experiments described in this section clearly indicate the need for further research in this area. A more thorough investigation of the thermal and photochemical rearrangements of systems analogous to 24 may create new synthetic routes into cyclopentenones and highly strained compounds such as 39. Furthermore, a study of the photochemical and acid-catalyzed rearrangements of eucarvone oxide 26 as well as its thermal rearrangement product (40) might be of considerable synthetic and mechanistic interest. EXPERIMENTAL A. General Procedures Except where otherwise noted, all nmr spectra were measured in CDC13 or CC14 solutions using TMS as an internal standard. The 60 MHz spectra were recorded on a Varian T-60 spectrometer and the 100 MHz spectra were recorded on a Varian HA-lOO spectrometer. The small number placed next to protons in the structures in the discussion section are the nmr chemical shifts of those protons. The numbers beside the chemical shifts in parentheses are the normalized europium shift numbers. These were obtained by adding small increments of tris-(l,l,l,2,2,3,3,-heptafluoro-7,7-dimethy1-4,6-octanedione)Eu(III) to the CC14 or CDC13 solution of the compound being investigated. After each addition the nmr spectrum was scanned and the new frequency of each absorption was recorded. The shift for each absorption is the difference between the frequency of the shifted absorption and the original one. The normalized shift numbers are ratios obtained by dividing the shift of each signal in the spectrum by the shift of the least shifted signal. Infrared spectra were recorded on a Perkin Elmer 237 grating spectrophotometer and calibrated against a polystyrene film. Ultra- violet spectra were obtained with a Unicam SP-800 in 95% ethanol, unless otherwise noted. Mass spectra were obtained from a Hitachi- 39 40 Perkin Elmer RMU-6 operated by Mrs. Ralph Guile. Melting points were determined with a Thomas-Hoover Melting Point Apparatus and are uncor- rected. Varian Aerograph gas chromatographs were used. Analyses were performed by Spang Microanalytical Laboratories, Ann Arbor, Michigan. B. 4,S-Epoxy-Z,3,4,5,6,6-hexamethy1-2-cyclohexenone (24). ’VD Epoxidation With Molecular Oxygen. A solution containing 178 mg (1.0 mmol) of 2,3,4,S,6,6-hexamethyl-2,4-cyclohexadienone (23)34 in 10 m1 of xylene was heated for 24 hr at 120-130° in a flask equipped with a reflux condenser. After removal of the solvent in vacuo, the residue was chromatographed on a Florisil column using 5% ether in hexane as the eluent, to afford 39 mg (22%) of recovered dienone £3 and 115 mg (0.60 mmoles, 77%) of epoxide 24: mp 48-49.S°; ir (KBr) 2970 (m), 2920 (m), 2860 (m), 1680 (s), 1465 (m), 1380 (m), 1310 (m), 1070 (m), 840 cm'1 (s); uv (cyclohexane) Amax 253 nm (e 7230); nmr (CDC13) see structure; all peaks had equal area; all were sharp singlets except those at 6 1.82 and 2.10, which were mutually coupled quartets, J = 1.0 Hz; mass spectrum (70 eV) m/e (rel intensity) 194 (14), 179 (14), 178 (19), 163 (18), 152 (46), 151 (70), 147 (21), 137 (29), 135 (25), 126 (29), 124 (29), 123 (27), 109 (31), 91 (16), 81 (30), 43 (100). Anal. Calcd for C12H1802: C, 74.10; H, 9.34 Found: C, 74.06; H, 9.32 The autoxidation of 23 in cyclooctane, mesitylene, cumene and chlorobenzene was carried out following the procedure described above. The yield of epoxide 24 was obtained by nmr for integration of the "UM 41 methyl signal at 6 1.01 and comparing it against an internal standard (CH2C12)' The results are given in Table I. Epoxidation With m-Chloroperbenzoic Acid. To a solution of 0.600 g (3.37 mmol) of dienone 23 in 10 m1 of methylene chloride was added, at 0°, a solution of 0.620 g (3.60 mmol) of m-chloroperbenzoic acid in 10 m1 of methylene chloride. The mixture was stirred for 2 hr at 0°, during which time m-chlorobenzoic acid precipitated from solution. The solvent was evaporated, petroleum ether (bp 30-60°) was added to the residue, and the m-chlorobenzoic acid was removed by filtration. Evaporation of the solvent from the filtrate left 0.648 g of a light yellow oil; an nmr spectrum of the crude material showed it to be >90% 21. The crude product was chromatographed on a Florisil column using 5% ether in hexane as the eluent, to give 0.523 g (2.70 mmol, 80%) of epoxide 24: mp 48-49.S°. The nmr and ir spectra were identical to the epoxide obtained by autoxidation of 23. Similar oxidation of 23 with a CD group at C-334 gave 24* whose 3 nmr spectrum lacked the quartet at 6 2.10, and with the quartet at 6 1.82 sharpened to a singlet. Oxidation of 23 with CD3 groups at C-3 and V1; 34 C-5 gave 21*,+ whose nmr spectrum, in addition to the changes just cited, lacked the singlet at 6 1.48. C. 4,5-Epoxy-3,4,6,6—tetramethy1-2-cyclohexenone (25). A solution of 166 mg (1.1 mmol) of 3,4,6,6-tetramethy1—2,4-cyclo- hexadienone27 in 10 ml of xylene was autoxidized following the same procedure used for the autoxidation of dienone £3 (vidb supra). After column chromatography on Florisil using 5% ether in hexane as the eluent, the reaction yielded 39 mg (0.26 mmol, 26%) of recovered dienone and 42 70 mg (0.42 mmol, 51%) of epoxide 25: mp 48-50°; ir (CC14) 2960 (s), 2920 (m), 2840 (w), 1680 (s), 1640 (w), 1470 (m), 1455 (m), 1440 (m), 1430 (m), 1360 (w), 1315 (m), 1235 (w), 1220 (w), 1170 (m), 920 (m), 890 cm'1 (m); uv (95% ethanol) lmax 247 nm (e 7200), 326 (143); nmr (CC14) see structure; the peaks at 6 2.08 (3H) and 5.72 (1H) were mutually coupled doublet and quartet respectively, J = 1.6 Hz, all other peaks were sharp three protons singlets with the peak at 6 3.03 being a sharp one proton singlet; mass spectrum (70 eV) m/e (rel intensity) 166 (8), 150 (39), 135 (37), 123 (36), 119 (32), 107 (100), 96 (35), 91 (72), 79 (44), 67 (51), 65 (26), 43 (46). Anal. Calcd for C H140 2: C, 72.26; H, 8.49 Found: C, 72.20; H, 8.56 10 Epoxide 25 was also prepared with m-chloroperbenzoic acid in 89% ’VD yield following the same procedure for the epoxidation of dienone £3 (vide supra). The nmr and ir spectra were identical to the epoxide obtained by autoxidation. D. 4,5-Epoxy-2,6,6-trimethy1-2-cycloheptenone (22). A solution of 113 mg (0.755 mmol) of 2,6,6-trimethy1-2,4-cyclo- heptadienone (eucarvone)28 in 7 m1 of xylene was autoxidized following the same procedure for the autoxidation of dienone £3 (vidé supra). After preparative thick-layer chromatography on neutral alumina using 10% ether in hexane as the eluent, the reaction yielded 33 mg (0.22 mmol, 29%) of recovered eucarvone and 48 mg (0.29 mmol, 52%) of epoxide 22, a clear liquid. The ir spectrum (CC14) showed bands at 2960 (s), 2920 (m), 2860 (w), 1670 (s), 1470 (m), 1440 (m), 1380 (m), 1310 (w), 1250 (w), 1220 (w), 1110 (w), 1070 (w), 1030 (w), 935 (m), 43 875 cm'1 (m); uv (95% ethanol) Amax 246 nm (e 7920); nmr (CC14) see structure; mass spectrum (70 eV) m/e (rel intensity) 166 (50), 123 (65), 109 (33), 107 (30), 82 (600), 81 (100), 54 (96), 53 (52), 43 (96), 41 (S9), 39 (86). Anal. Calcd for C10H1402: C, 72.26; H, 8.49 Found: C, 72.20; H, 8.50 Following the same procedure for the epoxidation of dienone £3, epoxide 22 was prepared with m-chloroperbenzoic acid in 76% yield. The nmr and ir spectra were identical to the epoxide obtained by autoxidation. 29 with m-chloroperbenzoic acid as described Treatment of dz-eucarvone above gave 26-d2 with the following nmr spectrum: 6 0.98 (s, 3H), 1.25 (s, 3H), 1.83 (d, 3H, J = 1.6 Hz), 3.00 (d, 1H, J 4.2 Hz), 3.22 (d of d, 1H, J = 6.1 Hz,.J = 4.2 Hz), 6.43 (q of d, 1H, J 6.1 Hz, J = 1.6 Hz). E. 4-(1,2-Epoxy-2,6,6-trimethy1-l-cyclohexy1)-3-buten-2-one (21). A solution of 105 mg (0.55 mmol) of 4-(2,6,6-trimethy1-l-cyclo- hexen-l-yl)-3-buten-2-one (B-ionone) in 4 m1 of xylene was autoxidized following the same procedure for the autoxidation of dienone 23 (vide supra). After preparative thick-layer chromatography on neutral alumina using 10% ether in hexane as the eluent, the reaction yielded 32 mg (0.17 mmol, 30%) of recovered B-ionone and 57 mg (0.27 mmol, 72%) of epoxide 27: mp 45.5-47° (11t.3° mp 46°); ir (KBr) 2950 (s), 2920 (s), 2860 (w), 1690 (m), 1680 (s), 1625 (m), 1455 (m), 1380 (w), 1365 (m), 1255 (s), 1175 (w), 1070 (w), 1050 (w), 990 (m), 910 cm"1 (w); uv (95% ethanol) 1 233 nm (a 11,200); nmr (CC14) 6 0-90 (5, 3H, methyl), max 1.12 (s, 3H, methyl), 1.15 (s, 3H, methyl), 1.15-2.10 (m, 6H, cyclohexyl H's), 2.20 (s, 3H, methyl a to carbonyl), 6.18 (d, 1H, J = 16.5 Hz, 44 vinyl H), 6.95 (d, 1H, J = 16.5 Hz, vinyl H); mass spectrum (70 eV) m/e (rel intensity) 208 (1.5), 193 (7), 165 (16), 135 (50), 123 (400), 109 (28), 107 (28), 95 (33), 91 (26), 81 (21), 79 (28), 77 (21), 69 (28), 67 (24), 65 (14), 55 (52), 53 (23), 41 (70), 39 (39). Following the same procedure for the epoxidation of dienone 23, epoxide gz was prepared with m-chloroperbenzoic acid in 76% yield. The nmr and ir spectra were identical to the epoxide obtained by autoxidation. F. 3a,4a-Epoxy-5-cholesten-7-one (28). A solution of 205 mg (0.54 mmol) of 3,5-cholestadiene-7-one38 in 7 ml of xylene was autoxidized following the same procedure for the autoxidation of dienone 23 (vidé supra). Thus, after column chromato- graphy on neutral alumina using 25% ether in hexane as the eluent, the reaction yielded 45 mg (0.12 mmol, 22%) of recovered dienone and 55 mg (0.14 mmol, 31%) of epoxide 28: mp 129-130° (lit.31 mp 127-128°; ir (CC14) 2950 (s), 2860 (m), 1675 (s), 1470 (m), 1380 (m), 1340 (w), 1280 (m), 1175 (m), 960 (w), 880 (w), 860 cm‘1 (w); uv (95% ethanol) Amax 239 nm (6 12,100); nmr (CC14) 6 0.66-2.50 (m, 39H), 3.30 (broad s, 2H, epoxide ring H's), 5.96 (s, 1H, vinyl H); mass spectrum (70 eV) showed a base peak of m/e 174 and a parent peak of m/e 398. Epoxide 28 was also prepared with m-chloroperbenzoic acid in 60% yield f0110wing the same procedure for the epoxidation of dienone 23 (vide supra). The nmr and ir spectra were identical to the epoxide obtained by autoxidation. 45 G. Methyl 4,5-epoxy-2-hexenoate (29). ’1/11 A solution of 3.00 g (23.5 mmol) of methyl 2,4-hexadienoate in 150 ml of xylene was autoxidized following the same procedure for the autoxi- dation of dienone 23 (vide supra). After column chromatography on neutral alumina using 10% ether in hexane as the eluent, the reaction yielded 1.12 g (8.89 mmol, 37.3%) of recovered dienoate and 1.62 g (11.4 mmol, 32 The infrared spectrum (neat) showed bands at 90.5%) of epoxide 29. ’Vb 2980 (m), 2940 (m), 2920 (w), 1725 (s), 1660 (m), 1440 (s), 1380 (m), 1350 (m), 1320 (s), 1275 (s), 835 cm'1 (s); nmr (CC14) 6 1.27 (d, 3H, J = 5.3 Hz, epoxide ring methyl), 2.88 (d of q, 1H, J = 5.3 Hz, J = 1.9 Hz, C-S H), 3.08 (d of d, 1H, J = 6.7 Hz, J = 1.9 Hz, C-4 H), 3.72 (s, 3H, carboxylate methyl), 6.07 (d, 1H, J = 16.0 Hz, C-2 H), 6.68 (d of d, 1H, J = 16.0 Hz, J = 6.7 Hz, C-3 H), the nmr spectrum also showed several small peaks at 6 1.23, 3.17-3.54 and 6.75 which are due to traces of another isomer, probably 023 at the epoxide ring. Epoxide 22 was also prepared with m-chloroperbenzoic acid in 85% yield following the same procedure for the epoxidation of dienone 23 (vidb supra). The nmr and ir spectra were identical to the epoxide obtained by autoxidation. H. Autoxidation of Methyl 2,4-Hexadienoate in the Presence of a,a'-Azobis-iso-butyronitrile. A solution of 300 mg (2.38 mmol) of methyl 2,4-hexadienoate (methyl sorbate) in 10 ml of xylene containing a,a'-azobis-iso-butyronitri1e was heated in a flask equipped with a reflux condenser. The reaction temperature, time and the concentration of 0,0'-azobis-iso-butyronitri1e 46 in xylene are given in Table 111. After removal of the solvent in vacuo, the crude reaction mixture was analyzed by nmr. The yield of epoxide 22 was obtained from integration of the methyl doublet at 6 1.27 and comparing it against an internal standard (CH2C12)' Similarly, the yield of re- covered dienoate was obtained by integration of the allylic methyl doublet at 6 1.90. The results are given in Table III. I. Autoxidation of Methyl 2,4-Hexadienoate in the Presence of 4-t- Butylcatechol. To a flask equipped with a reflux condenser was added 300 mg (2.38 mmol) of methyl 2,4-hexadienoate and 10 m1 of 0.01M solution of 4-t- butylcatechol in xylene. The mixture was heated at 120-130° for 24 hr, after which time, the solvent was removed in vacuo. Analysis of the crude reaction mixture by nmr showed no peaks at 6 1.27 and 2.88-3.08 indicating no formation of epoxide 29. M J. Autoxidation of 2,3,4,5,6,6-Hexamethy1-2,4-cyc1ohexadienone (23) ’\/\4 and 2,3,4,4,5,6,-Hexamethy1-2,5-cyclohexadienone (30). ’Vb A solution containing 178 mg (1.0 mmol) of conjugated dienone £234 and 178 mg (1.0 mmol) of cross-conjugated dienone 3834 in 15 m1 of xylene was heated at 120—130° in a flask equipped with a reflux condenser, for 24 hr. After removal of the solvent in vacuo, the residue was chromatographed on a Florisil column using 5% ether in hexane as the eluent to afford 45 mg (25%) of conjugated dienone 23, 102 mg (0.52 mmol, 70% of epoxide 34 and 155 mg (87%) of cross-conjugated dienone 38. 47 K. Autoxidation of Methyl 2-Butenoate (31). A solution of 176 mg (2.0 mmol) of methyl 2-butenoate (methyl crotonate) in 10 m1 of xylene was heated at 120-130° in a flask equipped with a reflux condenser, for 33 hr. After removal of the solvent in vacuo, analysis of the crude mixture (nmr and vpc) showed no epoxide formation. Preparative thick layer chromatography on neutral alumina yielded 144 mg (82%) of recovered 31. L. 4-Methy1ene-S-hydroxy-2,3,S,6,6-pentamethy1-2-cyclohexenone (32). To a solution of 24 (0.250 g, 1.29 mmol) in ether (15 m1) at 0° was added a solution of trifluoroacetic acid (0.5 ml) in water (5 ml). After the mixture was stirred for 0.5 hr, the layers were separated, and the ether layer was washed successively with saturated NaHCO3 solution, water, saturated NaCl solution, and dried (MgSO4). Evaporation of the ether left 0.238 g (95%) of the hydroxy ketone 32 as a colorless liquid which was not purified. Ir (CC14) 3620 (w, sharp), 3590 (w, sharp), 3500 (m, br), 2980 (s), 2930 (m), 2870 (w), 1670 (s), 1590 (w), 1460 (w), 1380 (5), several weak bands from 1350-1170, 1150 (m), 1130 (w), 1070 (m), 1040 (m), 1010 (w), 960 (m), 930 cm.1 (m). The bands at 3620 and 3590 cm'1 did not change in relative intensity as a function of the concentration of 32 in CCl whereas the intensity of the band at 3500 mm 4’ cm.1 decreased drastically with decrease in concentration of 32. Uv (cyclohexane) Amax 282 nm (a 11,700, sh), 273 (15,200), 266 (14,500, sh); nmr (CC14) see structure; the peaks at 6 1.85 and 2.08 were broadened, all other peaks being sharp singlets; mass spectrum (70 eV) m/e (rel intensity) 194 (29), 179 (31), 176 (8), 161 (16), 151 (100), 137 (30), 48 133 (44), 123 (15), 121 (15), 109 (26), 105 (15), 91 (23), 83 (16), 79 (23), 78 (25), 67 (13), 65 (15), S6 (18), 54 (26), 51 (15), 43 (80), 39 (42). Anal. Calcd for C12H1802: C, 74.19; H, 9.34 Found: C, 74.05; H, 9.36 Treatment of 24*,f with aqueous trifluoroacetic acid as above gave ég$sé with the following nmr spectrum (CC14): 6 0.96 (s, 3H), 1.10 (s, 3H), 1.85 (s, 3H), 5.43 (s, 1H), 5.64 (s, 1H). M. Rearrangement of 4,5-Epoxy-2,3,4,S,6,6-hexamethyl-2-cyclohexenone (24) in Neat Trifluoroacetic Acid. A solution of 0.100 g (0.515 mmol) of 24 in 2 m1 of ice-cold trifluoroacetic acid was stirred at 0° for 1 hr, then at room temperature for 20 hr. The reaction was monitored by nmr after the spectrum of each product had been determined. The reaction was quenched by pouring the mixture into ice and saturated NaHCO3 solution. The products were extracted with ether, and combined ether layers were washed successively with saturated NaHCO3 solution, water, saturated NaCl solution and dried (MgSO4). Evaporation of the solvent left 0.092 g of a light yellow liquid which was analyzed by vpc (5 ft x 1/8 in, 10% FFAP on chromosorb W, AW-DMCS 80/100, 160°, 30 ml/min). Four products (ret time in min, % yield) were observed: 34 (1.5, 6%), 35 (4.4, 27%), 36 (1.8, 33%), 31 (1-2: 34%). At 150° the ret times were, respectively, 3.5, 10.9, 4.0 and 2.5 min. After only 30 min reaction time, the yields were 34 (7%), 35 (32%), 32 (53%), 37 (8%). After 20 hr, the respective yields were 6, 27, 33 and 34%. The ratio of 34/35/(36 + 37) was time independent, but the yield ’Vb’Vb’Vbfl/b of 37 increased at the expense of 36. 'VD ’V'b 49 The products from this and larger scale experiments were separated by preparative vpc (10 ft x 0.25 in, 20% FFAP on chromosorb W P/G 30/60, 160°). 2-Acety1-2,3,4,5,5-pentamethyl-3-cyclopentenone (34): ir (neat) 2960 (m), 2920 (m), 2860 (w), 1740 (s), 1710 (s), 1470 (m), 1450(m), 1380 (m), 1360 (m), several weak bands from 1300 to 900 cm'l; uv (95% ethanol) 1 282 nm (e 81), 203 (35,660); nmr (CC14) see structure; max the peaks at 6 1.60 and 1.72 were mutually coupled quartets, J = 1.5 Hz; mass spectrum (70 eV) m/e (rel intensity) 194 (<1), 152 (94), 137 (100), 123 (22), 109 (12), 91 (12), 81 (27), 67 (14), 43 (55). Anal. Calcd for C12H1802: C, 74.19; H, 9.34 Found: C, 73.94; H, 9.41 4-Acety1-2,3,4,5,5-pentamethy1~2-cyclopentenone (35): ir (CC14) 2970 (s), 2940 (m), 2860 (w), 1710 (s), 1660 (5), several medium intensity bands from 1475-1400, 1380 (m), 1350 (m), 1320 (m), 1215 (m), 1150 (m), 1075 (m), 1025 (m), 960 cm"1 (w); uv (cyclohexane) Amax 237 nm (a 7,720), 209 (6,000); nmr (CC14) see structure; the bands at 5 1.75 and 1.95 were mutually coupled, J = 1.0 Hz; mass spectrum (70 eV) m/e (rel intensity) 194 (1.5), 179 (1.0), 166 (1.0), 152 (100), 137 (60), 123 (S3), 109 (9), 95 (8), 93 (9), 91 (12), 81 (35), 67 (12), 55 (12), 53 (12). Anal. Calcd for C12H1802: C, 74.19; H, 9.34 Found: C, 74.25; H, 9.29 S—Isopropenyl-4-methy1ene-2,3,S-trimethyl-2-cyclopentenone (36): ir (CC14) 3080 (w), 2965 (s), 2910 (s), 2865 (m), 1710 (s), 1645 (s), 1620 (s), 1455 (s), 1400 (s), 1370 (m), 1340 (w), 1290 (m), 1175 (w), 1160 (w), 1120 (w), 1030 (m), 1010 (w), 915 cm“1 (s); uv (cyclohexane) Amax 278 (a 10,030, sh), 270 (13,700), 262 (11,070, sh); nmr (cc14) 50 see structure; the band at 6 1.46 was a doublet, J = 1.7 Hz, those at 6 1.83 and 2.13 were broadened by mutual coupling, that at 5 4.95 was a multiplet (3 vinyl protons) and that at 6 5.30 was a broadened singlet. The peak at 6 1.21 was a sharp singlet; mass spectrum (70 eV) m/e (rel intensity) 176 (13), 161 (34), 148 (24), 133 (100), 105 (31), 91 (35), 79 (19), 77 (26), 65 (16), 53 (14), 51 (19), 41 (37), 39 (41). Anal. Calcd for C12H160: C, 81.77; H, 9.15 Found: C, 81.72; H, 9.11 4-Methy1ene-2,3,S-trimethyl-2-cyclopentenone (37): ir (CC14) 3080 (w), 2960 (m), 2925 (m), 2860 (w), 1710 (s), 1640 (s), 1620 (5), several medium intensity bands from 1460-1375, 1340 (w), 1310 (m), 1265 (m), 1145 (w), 1120 (m), 1050 (m), 990 (m), 905 cm'1 (s); uv (cyclohexane) Amax 275 nm (a 10,020, sh), 267 (15,200), 258 (13,450, sh); nmr (CC14) see structure; the peaks at 6 1.80 and 2.10 were broadened by mutual coupling, and the singlets at 6 5.10 and 5.27 were also broad; mass spectrum (70 eV) m/e (rel intensity) 136 (35), 121 (13), 93 (100), 91 (37), 79 (19), 77 (36), 67 (16), 65 (13), 55 (11), 54 (23), 53 (25), 52 (12), 51 (23), 50 (10), 41 (18), 39 (44). Anal. Calcd for CgHIZO: C, 79.37; H, 8.88 Found: C, 79.15; H, 9.04 A solution of 37 (50 mg, 0.368 mmol) in 4 m1 of CHsoD containing 54 mg (1.0 mmol) of sodium methoxide was stirred for 1 hr at room temper- ature, then quenched with 15 ml of 020 and extracted 3x with pentane (10 ml). The combined organic layers were washed with water (2x), saturated NaCl solution, and dried (MgSO4). Evaporation of the solvent left 45 mg of £1Rg1 with the following nmr (CC14): 6 1.17 (s, 3H), 1.80 (br s, 3H), 2.10 (br s, 3H), 5.10 (br s, 1H), 5.27 (br s, 1H). 51 N. Cleavage of 2-Acetyl-2,3,4,5,S-pentamethyl~3-cyclopentenone (34) With Base. A solution of’éfi (100 mg, 0.515 mmol) and sodium methoxide (20 mg, 0.555 mmol) in 3 m1 of methanol was stirred at room temperature for 8 hr. The mixture was poured into ice-water and extracted with ether, the combined ether extracts were washed with water, saturated NaCl solution, and dried (MgSO4). After evaporation of the solvent, the residue was analyzed by vpc (5 ft x 0.125 in, 20% FFAP, 120°) and showed that all of 34 was consumed, the sole product being 2,3,4,S,5-penta- methyl-Z-cyclopentenone (38), ret time 1.4 min. Pure 38 was collected by preparative vpc. Ir (neat) 2980 (m), 2880 (w), 1700 (s), 1650 (s), 1460 (m, br), 1400 (m), 1335 (m), 1040 cm“1 (m, br); uv (95% ethanol) 1 ax 235 nm (a 6,000); nmr (CC14) see structure; the peaks at 6 1.01 m (3H) and 2.33 (1H) were a mutually coupled doublet and quartet respectively, J 8.0 Hz and the peaks at 6 1.63 and 1.95 were mutually coupled quartets, J 1.0 Hz. Treatment of 34 with sodium methoxide in excess methanol-d for 8 hr at room temperature, followed by workup, gave éi$g4 whose nmr spectrum lacked the quartets at 6 1.95 (3H) and 2.33 (1H), the three proton peaks at 5 1.01 and 1.63 now becoming sharp singlets. Anal. Calcd fer C10“ C, 78.89; H, 10.59 160: Found: C, 78.75; H, 10.71 Cleavage of giads lacking the singlet at 6 1.27 (vidé infra) gave ifiags lacking the signal at 6 1.63 and with the peak at 6 1.95 a sharp singlet. Cleavage of éfiRgb lacking the methyl signals at 6 1.27 and 1.72 (vide infra) gave éfiRgb lacking the quartet at 6 1.63 and the doublet at 6 1.01, and having the signal at 6 1.95 a sharp singlet and that at 6 2.33 a broadened one-proton singlet. 52 0. Treatment of S-Isopropenyl-4-methy1ene-2,3,S-trimethy1-2-cyclo- pentenone (36) With Trifluoroacetic Acid. A solution of 36 (60 mg, 0.34 mmol) in 1 m1 of trifluoroacetic acid was allowed to stand at room temperature for 20 hr, the reaction being monitored by nmr. During the reaction, a sharp singlet appeared at 6 2.33, shown to correspond to acetone in trifluoroacetic acid. The reaction was quenched by pouring it into ice and saturated NaHCO3 solution. The mixture was extracted with ether, and the ether extract was worked up and analyzed as in the rearrangement of 24. The sole components (determined by vpc) after 20 hr were 36 (55%) and 37 (45%). When the reaction was carried out at higher temperatures, conversion to 37 was quantitative. ’Vb P. Rearrangement of 5-Hydroxy-4-methy1ene-2,3,5,6,6-pentamethy1-2- cyclohexenone (32) in Trifluoroacetic Acid. mm A solution of 32 (0.100 g, 0.515 mmol) in 1.5 m1 of trifluoroacetic acid was stirred at room temperature. Aliquots were withdrawn at various time intervals, quenched and worked up as usual, and analyzed by vpc (5 ft x 0.125 in, 10% FFAP on chromosorb W, AW DMCS 80/100, 150°, N2 flow rate 30 ml/min). Only two components were present, 36 (4.0 min) and 31 (2.5 min), identified by nmr and ir. The relative amounts at various time intervals were: 2 hr, 94% 36, 6% 31; 26 hr, 32% 36, 68% 37; 46 hr, 3% 36, 97% 37. «A. an Ivy ’Vb 53 Q. Rearrangement of Labeled 4,5-Epoxy-2,3,4,5,6,6-hexamethy1-2- cyclohexenone in Trifluoroacetic Acid. A solution of 24* (lacking the signal at 6 2.10; 500 mg, 2.58 mmol) in 5 ml of trifluoroacetic acid was stirred at room temperature for 21 hr, then quenched and worked up as described for unlabeled 24. ’X/b The products had the following nmr spectra (CCl ' 31, 6 1.05 (s, 3H), 4)' 1.08 (s, 3H), 1.60 (m, 3H), 1.72 (m, 3H), 1.98 (s, 3H); 35, 6 0.90, 1.03, 1.25, 1.75 and 1.95, all s, 3H; 36, 6 1.21 (s, 3H), 1.46 (d, 3H, J = 1.7 Hz), 1.83 (s, 3H), 4.95 (m, 3H), 5.30 (br s, 1H); 31, 6 1.17 (d, 3H, J = 7.5 Hz), 1.80 (s, 3H), 2.70 (q, 1H, J = 7.5 Hz), 5.10 (br s, 1H), 5.27 (br s, 1H). A solution of 24f,+ (lacking the signals at 6 2.10 and 1.48) in trifluoroacetic acid was allowed to rearrange in the amounts and manner described for 24f. The products had the following nmr spectra (CC14): 34, 6 1.05, 1.08, 1.60, and 1.98, all s, 3H; 35, 6 0.90, 1.03, 1.25 and 1.75, all s, 3H; i2} 6 1.46 (d, 3H, J = 1.7 Hz), 1.83 (s, 3H), 4.95 (m, 3H), 5.30 (br s, 1H); 31: 6 1.80 (s, 3H), 2.70 (br s, 1H), 5.10 (br s, 1H), 5.27 (br s, 1H). R. Thermal Rearrangement of 4,5-Epoxy-2,3,4,5,6,6-hexamethy1-2-cyclo- hexenone (24). "Vb A solution of 24 (55 mg, 0.31 mmol) in 0.5 ml of CCl4 was sealed ’Vb in an nmr tube and heated at ZOO-210°. After 2 hr no starting epoxide was detected by nmr. Analysis of the crude reaction mixture by tlc showed three products. The mixture was separated by preparative thick- layer chromatography on neutral alumina using 10% ether in hexane, 54 yielding two products, 25 and 32, which were identified by nmr. Only traces of the third product could be isolated, no attempt was made to identify it. A similar rearrangement of epoxide 24 in the absence of solvent yielded the same products. S. Photochemical Rearrangement of 4,S-Epoxy-Z,3,4,5,6,6-hexamethy1- 2-cyclohexenone (24). mm A solution of 250 mg (1.29 mmol) of 24 in 400 m1 of anhydrous ether was flushed with nitrogen and irradiated under nitrogen (Hanovia 450 W, Pyrex) for 1.5 hr. Analysis of the crude reaction mixture by tlc showed two products. After evaporation of the solvent, the mixture was separated by preparative thick-layer chromatography on neutral alumina using 10% ether in hexane yielding 126 mg (51%) of unreacted starting material and 26 mg (10%) of diketone 35, identified by nmr. Similar irradiation of epoxide £4 at 10-15° for 8 hr followed by evaporation of the solvent in vacuo yielded a light yellow oil which was not purified. All attempts to purify the crude product resulted in decomposition. Nmr analysis of the crude oil showed it to consist mainly of one product which was tentatively assigned structure 3%. Ir (CC14) 1760 (s) and 1710 cm’1 (s); nmr (CC14) 6 0.80, 1.12, 1.23, 1.27, 1.33 and 2.20, all sharp three protons singlets. T. Irradiation of 4-Acety1-2,3,4,5,S-pentamethyl-2-cyclopentenone (35). ’Vb A solution of 28 mg (0.14 mmol) of 35 in 20 ml of ether was placed in a Pyrex tube which was then stoppered with a serum cap. The solution 55 was flushed with nitrogen (using syringe needles) for 15 min and then irradiated for 2 hr with a 450 W Hanovia lamp. The solvent was removed in vacuo and the residue analyzed by nmr which showed small amounts of unreacted starting material and six sharp methyl peaks corresponding to 39 (vide supra). mm U. Thermal Rearrangement of 4,5-Epoxy-2,6,6-trimethy1-2-cycloheptenone (26). mm A solution of 26 (50 mg, 0.30 mmol) in 0.5 ml of CCl4 was sealed in an nmr tube and heated at 180-185° for 4 hr, the reaction being monitored by nmr. After 4 hr the reaction mixture consisted of only one product which was assigned structure 40 based on the following spectral data: ir (neat) 2960 (s), 2930 (s), 2870(m), 1720 (s), 1470 (m), 1440 (m), 1410 (m), 1375 (m), 1365 (m), several small bands from 1325- 1100, 1075 (s), 1010 (m), 975 (w), 915 (m), 880 (m), 860 (m), 760 (s), 720 cm'1 (s); nmr (CC14) see structure peaks at 6 1.30, 1.27 and 0.90 were sharp three proton singlets; mass spectrum (70 eV) m/e (rel intensity) 166 (17), 151 (1.5), 123 (28), 109 (17), 95 (9.0), 82 (100), 54 (36), 43 (43), 39 (44). Anal. Calcd for C H 0 ° C, 72.26; H, 8.49 10 14 2' Found: C, 72.20; H, 8.50 PART II REGIOSPECIFICITY IN DI-n-METHANE PHOTOISOMERIZATIONS INTRODUCTION The generality of the di-n-methane photorearrangement was first recognized by Zimmerman.39 The reaction involves the conversion of two N moieties bonded to a single saturated carbon atom into a n-substituted cyclopropane. The rearrangment is illustrated in the scheme below. -—r / R R R 7' 1 R1 R2 2 R1 R1 R2 2 a b hv I 423 I I ° m mmm R1 R1R2 R2 R1 R1Rz R2 \ 41 H \ ———» \ ’1!» ’b c R R R1R1R2R2 1R1R2 2 J 42b ’b ’VW Although two vinyl groups are used for simplicity, the n moiety may alternatively include phenyl or carbonyl groups.40 In unsymmetric di-fl-methane systems, i.e., when R1 # R2, two products (42% and/or 422) are possible depending upon whether bond a or bond b is broken along the reaction route. Several cases are known in which the reaction follows only one of these two possible routes.4O For example, 57 58 direct irradiation of trans-l,l-dipheny1-3,3-dimethy1-l,4-hexadiene (43) gave 1,l-diphenyl-Z,2-dimethy1-3-propenylcyclopropane (44), but no 1,1,2- trimethyl~3~(2,2-dipheny1viny1)cyclopropane (45).41 Thus, "intermediate" «~76 V Ph § cleaved exclusively at bond a. a I I hu 141 Ph I»; Ph b , 43 K ‘\\>kk m m A ' A PhPh Phph 45 'V'b Another type of regiospecificity has been observed in bicyclic di-n-methanes.42 Sensitized irradiation of carbomethoxydibenzobarrelene 46 gave dibenzosemibullvalene 41, but no 48. In this case, the mode of initial bonding determines the structure of the reaction product. Thus, the observed product results from "intermediate" k not M. These two types of regiospecificity can be easily rationalized. The reactions proceed along routes which maintain maximum odd—electron stabilization. In the first example, cleavage of bond b leads to loss of benzhydryl delocalization while cleavage of bond a does not. Similarly, in the second example, odd-electron stabilization by the carbomethoxy group makes L the favored bridged species. "h 60 The substituent effect does not seem to be associated with a particular excited state multiplicity, since the first reaction proceeds from a singlet state, and the second from a triplet.39’42 It should be noted that this rationale to account for the regio- specificity in the di-fl-methane rearrangement does not require a non~ concerted mechanism. In fact, it has been proposed that the di-w-methane rearrangement, especially from the singlet state, may be concerted.43 Hence, formulations such as H-M do not mean to imply that the species along the reaction route necessarily are intermediates and correspond to energy minima. In some instances such discrete diradicals may be involved; in other cases, these may just be points along the hypersurface leading from excited state of reactant to ground state of product. Never- theless, such qualitative valence-bond structures are useful in depicting odd-electron disposition along the reaction coordinate and one can expect predictions based on such models to correlate with experiment. Still a third and less well understood type of regiospecificity in di-W-methane photoisomerizations has been observed by Hart and Murray.44 Acetone-sensitized irradiation of secondary alcohol 49s gave 52:, but no 51$. This process is shown in detail in Scheme 10. In this case also, the mode of initial bonding determines the reaction product. Thus, the observed product results from intermediate N not 2. Irradiation of the epimeric alcohol 422 was also regioselective, but the directive effect was less dramatic than with the syn isomer 42%. Since the hydroxyl group in the saturated bridge is not bonded to any of the atoms directly involved in the rearrangement, this directive effect may be a result of "through space" rather than a ”through bond" interaction. Hart and Murray44 suggested that this apparent stabilization of N over 0 could arise from charge transfer and/or hydrogen bonding 'b 61 H0 495 HO H 82 515 W SCHEME 10: The di-n-methane rearrangement of (4,93. HO 62 hv acetone interaction of the hydroxyl moiety with the delocalized n—system "beneath" it. Subsequent results,45 however, indicated that the corresponding acetates show similar selectivity, thus eliminating the hydrogen bond explanation. The same substituent effect was observed in the photochemistry of the epimeric alcohols 525 and 523.45 Irradiation of 525 in acetone gave mm W); HO H hv O -——-—> but no acetone 525 53s 54s mmm mmm H OH hv -———————» but no 0 acetone 52a 53a 54a 63 §tzi and an unidentified minor product in a ratio of 7:1. Similarly, irradiation of 52a gave a single photoproduct 53a. m M In another case, Murata and Sugihara46 observed a similar substituent effect. Sensitized irradiation of secondary alcohol 55 gave 56 but no 57. “A: ’Vb V): In this case also, of the two possible di-n—methane intermediates P and Q, N ’1: p m 56 OH hv acetone 0 / £3 57 the reaction proceeds via the former. The explanation given to account for this substituent effect was also hydrogen bonding or charge-transfer interaction between a developing radical and the oxygen. The purpose of this part of this thesis was to examine this unusual substituent effect with the hope of learning more about the nature of the stabilization of triplet excited states by groups not directly bonded to any of the atoms with odd electron density. RESULTS AND DISCUSSION One of the proposals of Hart and Murray for the selective formation of intermediate N rather than 2 in the photolysis of alcohol 423 was a charge-transfer interaction between a lone pair on the oxygen and the proximal unpaired electrons. To test this explanation two epimeric methyl substituted 1,4-dihydro-l,4-ethanonaphtha1ene 58% and 582 were prepared and irradiated. HC H H CH 3 3 58 58a mmz mmm A. Synthesis of syn- and Anti-9-methy1-l,4-dihydro-l,4-ethanonaphthalene, 585 and 58a. MAW): The initial phase of this investigation was concerned with the trans- formation Of ketone E245 to both epimers of 58 (Scheme 11). Treatment of 59 with methylenetriphenylphosphorane by the modified Corey procedure47 proved to be highly efficient, leading to the isolation of 9-methy1ene- l,4-dihydro-1,4-ethanonaphthalene (62) in 84.0% yield. However, selective reduction of 60 with disiamylborane followed by treatment with propionic 64 65 SCHEME 11 CH /O / 2 Ph P=CH O 3 2’ 0 8% 81 l. SiazBH H3C H 2. Propionic acid, A @ *—‘ 585 W acid unexpectedly gave only one product, syn-9-methy1-l,4-dihydro-l,4- ethanonaphthalene (583) in 63.0% yield. Several attempts to synthesize the anti-isomer 582 from 22 by other reduction methods were unsuccessful. For example, catalytic hydrogenation with palladium on charcoal or treatment of 60 with tris-(triphenylphos- phine)rhodium chloride catalyst lead to preferential reduction of the AZ-double bond, yielding 9-methy1ene-l,2,3,4-tetrahydro-l,4-ethanonaph- thalene. The synthesis of anti-Q-methyl-l,4-dihydro-l,4-ethanonaphthalene (582) was finally accomplished by the sequence of steps described in Scheme 12. 66 SCHEME 12 0 CH2 / CO Me i C02Me ph3p=CH2 _ 2 Cone CO2Me 61 62 M W H2; Pd/C MeOH H H CH3 CH3 CO Me CO2H Agl. NaOH ,,. 2 2. H30+ COZH COZMe 64 63 V): H CH3 electrolysis; 3 58a Reaction of keto-diester 6148 with methylenetriphenylphosphorane resulted in the formation of methylene-diester 62 in 70.7% yield. Reduction of 62 with palladium on charcoal in methanol was completely stereospecific, giving only the isomer with the methyl substituent anti to the aromatic ring, 63, in 91.5% yield. Saponification of diester 63 followed by ’\/\a 67 electrolytic decarboxylation of the diacid 64 in pyridine-water-tri- 'Vla ethylamine solution conveniently gave only the anti-isomer 58a in W 34.5% yield. The structure of 585 and 58a follow from their method of synthesis W W and spectroscopic properties. Their spectral properties are given in the experimental section. The configuration assignments of 585 and 58a ’VW MA; are based on their nmr spectra. In general, the methyl substituent in 2,3-benzobicyclo[2.2.2]octa-2,S—dienes appears at higher field when the methyl is syn to and "over" the aromatic n system, than when the methyl has the anti geometry.4S Thus in 585 the methyl doublet appears at 6 0.62, W whereas in 58a it appears at 6 0.98. W B. Photolysis of Syn- and Anti-Q-methyI-l,4-dihydro-l,4-ethanonaphthalene, 585 and 58a. WWW mmm Epimers 58s and 58a were photolyzed in acetone through a pyrex filter, mm the reaction being monitored by vpc. The photoproducts were separated by preparative vpc and their yields determined by vpc using a calibrated internal standard. Irradiation of the syn-epimer 58: gave three products which were identified from their 100 MHz nmr spectra and by independent syntheses. These photoproducts together with their respective absolute yields are given in Scheme 13. In contrast to 58$, irradiation of the anti-epimer 58a gave only MA; ’VVM two products which were also identified from their nmr spectra and by independent syntheses. These photoproducts together with their respective absolute yields are given in Scheme 14. 68 SCHEME 13 H C 3 H 585 MW hv acetone H3C H lélily + + 65s, 22.7% 675, 63.8% mmm mmm Photoproducts 655 and 65a were determined to result from photore- WMWM duction of 585 and 58a, respectively. Their nmr spectra were consistent 'VWWM with the structures, and the assignment was corrobarated by their mass spectra which displayed a parent ion peak at m/e 172, an increase of two mass units compared to 58s and 58a (parent peak m/e 170). The WW spectral properties of 655 and 65a are given in the experimental section. ’VWWD Finally, 655 and 65a were conclusively identified by comparison with 'VVb'VVb authentic speciments obtained by separate hydrogenation of 583 and 522 with 10% Pd/C in methanol. Photoproduct 66s was identified from its 100 MHz nmr spectrum which ’VW showed, in addition to the four aromatic protons at 6 6.73-7.17 and the methyl doublet at 6 0.48, a complex spectrum fOr an additional seven protons. The nmr assignments were based on double irradiation experiments; 69 SCHEME 14 H CH3 58a mmm hv acetone H CH3 0 * 653, 11.7% 67a, 81.8% mmm mmm however, due to long range coupling and the complexity of the spectrum, a complete analysis of the coupling constants could not be made. The double irradiation results and other spectral properties are given in the experimental section. In order to confirm its structure, photoproduct 223 was synthesized from 3,4—benzotricyclo[3.3.0.02’8]oct-3-ene-7—one (9%)49 as outlined in Scheme 15. Treatment of ketone 98 with methylenetriphenylphosphorane by the modified Corey procedure47 gave 7-methy1ene-3,4-benzotricyclo[3.3.0.02’8] oct—3-ene (23), in 70.0% yield as colorless crystals, mp 55-56°. Its 60 MHz nmr spectrum showed in addition to the four aromatic protons at 6 6.83-7.28 and a multiplet for two vinyl protons at 6 4.73-4.97, a complex spectrum for an additional six protons with a pattern similar to ketone 38. Hydroboration of 69 with BZH6 in tetrahydrofuran followed by ’Vb 7O SCHEME 15 1. 82H 6 2. Propionic acid, A treatment with propionic acid gave a single product (60.5% yield) with ir and nmr spectra identical to photoproduct 663. The major photoproduct from irradiation of 38s was assigned as syn-6-methy1-3,4—benzotricyclo[3.3.0.02’8]oct-3-ene (672) also by analysis of its 100 MHz nmr spectrum. Thus, the nmr spectrum showed four aromatic protons at 6 6.85-7.37, a three-proton doublet for the methyl substituent at 6 0.74 and a complex spectrum for the remaining seven protons. A complete assignment was made based on double irradiation experiments. The double irradiation results and other spectral properties are given in the experimental section. In addition to the double irradiation results, a comparison of the methyl chemical shift of isomers 623 and 21: clearly indicates their respective positions. As the three dimen- sional drawings indicate, the methyl substituent in 22: is expected to be shielded by the aromatic ring current to a greater extent and, therefore, 71 66s ' 67s absorb at higher field than the methyl substituent in 212. Thus in 633 the methyl doublet appears at 6 0.43, whereas in gzs it appears at 5 0.74. Finally, the structure of 67g was conclusively established by comparison with an authentic sample obtained from 3,4-benzotricyclo[3.3.0.02’8]oct-3- ene—6-one (ZR)45 as outlined in Scheme 16. Reaction of ketone ZR with methyllithium resulted in the formation of a single alcohol (97.0% yield) with the hydroxyl group syn to the aromatic ring. The stereochemical assignment is based on the nmr and infrared spectra. The 60 MHz nmr showed a methyl doublet at 5 0.95 which is predictably at lower field than if the methyl were syn to the aromatic ring. The infrared spectrum showed a band at 3575 cm.1 which after concentration studies was attributed to internal hydrogen bonding with the aromatic ring. Such intramolecular hydrogen bonding would not lye expected if the hydroxyl were anti to the aromatic ring. Treatment of alcohol 71 with thionyl chloride followed by reduction with tri-n-butyltin 72 SCHEME 16 CH3Li D SOCIZ, Py Benzene LCD-BLOAgan 8‘ Toluene 67s 72 hydride in toluene gave a single product (85% yield) with nmr and ir spectra identical to photoproduct 21:. The formation of a single alcohol from ketone ZR and also the forma- tion of a single product from reduction of chloride 72 is not unreasonable, since attack of methyllithium and also of the hydrogen atom from tri-n- butyltin hydride is expected to take place from the convex face of the cupped molecule to produce, respectively, syn-alcohol 11 and the syn- isomer 27:. The same stereospecific attack occurred in the reaction of alkene 62 with borane (vidé supra). 73 The major photOproduct from irradiation of 38% was assigned as anti- 6-methy1-3,4-benzotricyclo[3.3.0.02’8 ]oct-3-ene (67a) based on its spectral WM properties and by degradation to a relay compound which was synthesized independently. The 100 MHz nmr spectrum was considerably more complex than the nmr spectra of photoproducts 665 and 67s and showed, in addition WWW to the four aromatic protons at 5 6.92-7.37 and the methyl doublet at 6 1.08, a half-proton multiplet at 6 3.46—3.60, a half-proton doublet at 6 3.17, one—proton multiplets at 6 2.70-2.93 and 2.26-2.48, and an extremely complex four proton multiplet at 6 1.25-2.16. The complexity of the spectrum prevented a complete proton assignment, similar to those of photoproducts 66s and 67s. However, similarities between the nmr 'VW’VW spectrum of 67a and the nmr spectra of 66s and 67s in the aromatic WM WWW region, in the low field multiplet at approximately 6 3.5 (benzylic methine proton at C5), and in the "quartet-like" multiplet at 6 2.8 (cyclopr0py1 methine proton at C1) indicated that photoproduct $12 was also a methyl-substituted 3,4-benzotricyclo[3.3.0.02’8]oct-3-ene. This gross structural assignment was suggested by the similarity in the infrared spectra of 67a, 66s and 67s, and by their identical mass spectra. Wb’VWW'b However, photoproduct 67a could have the anti-methyl substituent located W at either C6 or C7. In an attempt to establish the position of the methyl substituent in 67a, a 270 MHz nmr spectrum was determined and double irradiation was WM performed.50 The results are given in Table IV and shown in the Appendix section (Figures 10 and 11). Examination of the spectrum clearly indicates that it still remained second-order and unambiguous analysis was not possible. For instance, with the exception of the signals at 6 1.03, 2.72-2.82 and 2.26-2.38 all other signals integrate to fractions of a proton. The total integration agrees with structure 612. However, as ‘1‘ ll I'll-Ill- l i: ) \l' .6 74 Table IV. 270 MHz NMR and Spin Decoupling of Anti-6-methy1-3,4-benzo- 2,8 tricyclo[3.3.0.0 ]oct-3-ene (67a).a ’VW Signal Rel. Signal description Signal b Decoupling position, 6 area and assignment irradiated resultC 3.47-3.55 0.4 apparent triplet, 2.72—2.82, H-l doublet benzyl methine, H-S 1.66-1.86, anti H-7 doublet 3.14 0.6 doublet, benzylic 2.72-2.82, H-l singlet cyclopropyl H, H-2 2.72-2.82 1.0 apparent doublet of a 3.47-3.55, H-S triplet quartet, cyclopropyl 3.14, H-2 triplet methine, H-l 2.26-2.38, H-8 triplet 1.41-1.60 quartet 2.26-2.38 1.0 two partially over- 2.72—2.82, H—l triplet lapping "triplets", 1.66-1.86, anti H-7 d of d cyclopropyl adj to 1.41-1.60 d of d methylene, H-8 1.97-2.08 0.5 apparent quintet, 1.18-1.37, syn H-7 quartet methine, H-6 1.03, methyl doublet 1.66-1.86 1.4 multiplet, anti H 3.47-3.55, H-S incomp: at C-7 2.26-2.38, H-8 incomp 1.18-1.37, syn H-7 incomp 1.41-l.60 0.7 "triplet" super- 2.72-2.82, H-l doublet: imposed on a 2.26-2.38, H-8 doublet multiplet, 1.18-1.37, syn H-7 doublete unassigned 1.66-1.86, anti H-7 incomp 1.03, methyl incomp 1.18-1.37 1.4 multiplet, syn H 1.97-2.08, H-6 incompd at C-7 1.66-1.86, anti H-7 incompd l.41-l.60 incomp 1.03 3.0 doublet, methyl l.97-2.08, H-6 incomp: 1.41-l.60 incomp aRun in CDCl with TMS as internal standard. bSignal irradiated that affects signal in column 1. CResult of irradiation in spin decoupling. Incomplete decoupling. eMultiplet remained unchanged. 75 shown in Table IV, if one assumes that each multiplet corresponds to one proton, an assignment based on spin decoupling seems to place the anti-methyl substituent at C6. The position of the methyl substituent was clarified mainly by the long range coupling between H-5 and anti H-7.4S These protons are geometrically fixed in the "W" configuration for coupling through four bonds.51 Thus, irradiation at anti H-7 showed it to be coupled to H-8, H-5 and syn H-7. If the anti-methyl substituent were at C such long 7 range coupling would not be observed. Furthermore, the splitting pattern for H-8 would be simplified considerably since it would have one less vicinal hydrogen. The presence of two half-proton signals at 6 3.47-3.55 and 3.14 still remained a puzzling problem. Since these signals are far removed from the rest of the spectrum and do not seem to be mutually coupled, one would expect this region of the spectrum to differ only slightly from a first order spectrum. This problem raised the possibility that photoproduct 61% is a mixture of two isomeric products (anti-methyl substituent at C6 and C7). However, since no other product was detected by vPc using three different columns (3% SE-30, 10% FFAP and 20% 0568 on chromosorb W) at various temperatures and flow rates, and only one methyl doublet was observed in the 270 MHz nmr spectrum, this possibility seemed unlikely. Nevertheless, in View of the spectral ambiguities, it was important that structure 61% be established more firmly, preferably by independent synthesis. Since reduction of chloride 12 afforded only photoproduct 672, several attempts were made to synthesize 212 from methyl ketone 74 (Did? infra). Unfortunately, Wolff-Kishner reduction of ketone 74 or conversion 'Vb of 74 to the respective chloride followed by tri-n-butyltin hydride 76 reduction resulted in opening of the cyclopropane ring. This difficulty was overcome by hydrogenation of photoproduct 212 and comparing its product with an authentic sample obtained by independent synthesis. Thus, hydrogenation of $73 with 10% palladium on charcoal resulted in the uptake of one equivalent of hydrogen. The mass spectrum showed a parent peak at m/e 172, corroborating the addition of one equivalent of hydrogen to photoproduct 21%. The nmr spectrum of the dihydro photoproduct showed four aromatic protons at 6 6.68-7.08, a methyl doublet at 6 1.00 and an extremely complex spectrum for the remaining nine protons. Analytical gas chromatography showed the hydrogenation of photoproduct 272 to consist of a single product; however, the nmr spectrum showed a small doublet at 6 1.17 and a multiplet at 6 6.95-7.10 corresponding to approximately 5% of a second product. The presence of a second product is not surprising in view of the fact that three isomeric products can result from catalytic hydrogenation of photoproduct 67% (Scheme 17). Of these three possible hydrogenation products, 13: and 132 are most likely to be formed since the C1-C2 and C2-C8 bonds are activated by the adjacent phenyl group. Furthermore, based on steric considerations, one might expect product 722 to be the major product. As depicted in the three dimensional drawing below, molecular models of 613 clearly indicate that the C7 methylene restricts the approach of the C2-C8 bond to the catalyst surface, while the C1-C2 bond is noticeably less hindered to such an approach. A convenient synthetic entry to one of the possible hydrogenation products (132) is outlined in Scheme 18. Addition of ketone 68 to a solution of hexamethyldisilazane and n-butyllithium followed by treatment with methyl iodide proved highly stereospecific, leading to the formation 2,8 of anti-6-methy1-3,4-benzotricyclo[3.3.0.0 ]oct-3-ene-7-one (14) in SCHEME 17 77 C2C8 cleavage L 73a 'VV'b H2, Pd/C C1C2 =: CHSOH cleavage CH ; 3 67a H “W“ 73b ’VVb C1C8 _ cleavage 7 :CH3 H 73c WM 98% yield. As with other reactions in this ring system, attack of methyl iodide on the enolate ion of 68 occurred only from the convex face of ’Vb the cupped shape molecule. Reduction of ketone 14 with 10% palladium on 78 SCHEME 18 1- HMDS/n-BuLi 2. CHI 7 3 68 74 mm mm H2, Pd/C CHSOH V AHZNNHZ, KOH ‘VTriethylene glycol §\l LN 0.: SE charcoal in methanol gave only one product which was assigned structure 15 based on an analogous reduction of ketone 68.49 Also, in analogy 9 to the parent ketone (less methyl substituent),4 Wolff-Kishner reduction of 75 gave 80% of 73a. Although this compound (73a) had an identical ’Vb W W» vpc retention time with the material obtained from reduction of photo- product 67a, their nmr and ir spectra differed considerably. W In order to eliminate the possibility that the anti-methyl substi- tuent in photoproduct 67a is at C7, anti-7-methy1-3,4-benzobicyclo[3.3.0]- WM oct-3-ene (72) was synthesized from ketone 10 by a sequence of steps similar to the synthesis of 73% (Scheme 19). In this case, however, hydrogenation preceded methylation due to the instability of the enolate 79 SCHEME 19 H2; Pd/C CH30H 0 a 2,2 1. HMDS/ n-BuLi 2. CH3I 1! CH3 )1 H NNH , KOH I 2 2 Triethylene glycol 78 77 ’\/\J ’Vb ion resulting from ketone ZR. Thus, hydrogenation of ketone 72 with 10% palladium on charcoal in methanol resulted in the uptake of one equivalent of hydrogen. Vapor-phase chromatography showed the presence of two products in a ratio of 3:1. The major product gave a 2,4-dinitropheny1- hydrazone derivative whose melting point was in close agreement with the literature value for the 2,4-dinitropheny1hydrazone derivative of ketone 12.52 The minor product was not identified. Methylation of ketone 12 was stereospecific and gave only the anti-methyl ketone 11 in 67% yield. Finally, Wolff-Kishner reduction of ZZ afforded Z3 as the sole product in 74% yield. This compound (18) differed in every respect 80 (ir, nmr, and vpc retention time) with the sample obtained from reduction of photoproduct 67a. ‘V'W . Authentic anti—6-methy1-3,4-benzobicyclo[3.2.l]oct—3-ene (73b) was ’VW conveniently produced by a sequence of steps similar to the synthesis of 73a (Scheme 20). In this sequence, however, ketone 74 was reduced W ’\/\a to syn-alcohol 79 in order to control the direction of hydrogenation. The syn-hydroxyl substituent at C7 impedes approach of the catalyst to the C2C8 bond so that hydrogenation takes place preferentially at the clcz bond to yield the bicyclo[3.2.0]octane derivative. Thus, reduction SCHEME 20 LiAlH4 : 3ft If: 1. H2; Pd/C 2. Jones oxidation V : CH3 H 73b ’V'V'b \ HZNNHZ , KOH 0 Triethylene CH glycol 5 3 H 80 ’Vl; H CH3 81 of 79 over palladium on charcoal in methanol followed by Jones oxidation ’Vb of the crude hydrogenation mixture yielded ketones 7S and 80 in a ratio 'Vb ’Vb of 1:3. Separation of these isomeric ketones by preparative thick-layer chromatography was followed by Wolff-Kishner reduction of ketone 80 to ’Vh afford two hydrocarbons in approximately 3:2 ratio. Separation of the two components by vapor-phase chromatography gave 73b and 81. The W "Vb infrared spectrum and vpc retention time of the major component (132) was identical to that of the material obtained from hydrogenation of photoproduct 67a. The nmr spectrum of 73b lacked the small doublet at mmm mmm 6 1.17 and the multiplet at 6 6.95-7.10 observed in the dihydro photo- product, but the remaining spectrum was identical. This comparison establishes the structure of the dihydro photOproduct which in turn, then, firmly establishes the structure of the major photoproduct from 2,8 58a as anti-6-methyl-3,4-benzotricyclo[3.3.0.0 ]oct-3-ene (67a). ’VVb MA: The structures of compounds 72-81 are based on their method of synthesis and on their spectral properties which are given in the experi- mental section. The configuration of the methyl substituent is based on its nmr chemical shift. In general, methyl substituent syn to the aromatic n system is expected to be more shielded and appear at higher field than when the methyl has the anti geometry. The configuration of the hydroxyl group in 19 is based on its infrared spectrum which showed bands at 3640 and 3590 cm_1 attributed to free O-H stretching and internal hydrogen bonding with the aromatic ring. As in the case of alcohol 71, such intramolecular hydrogen bonding is not expected when the hydroxyl is anti to the aromatic ring. Additional evidence supporting structures 73b and 81 was obtained W ’\/\J from the previously identified photoproduct 67s (vidé supra). As shown ’VW 82 in Scheme 20, Wolff-Kishner reduction of ketone 80 conveniently caused ’Vb epimerization of the methyl substituent giving rise to syn-methy1-3,4- benzotricyclo[3.2.l]oct-3-ene (81). In analogy to photoproduct 612, then, 81 should be identical to the product resulting from hydrogenation of photoproduct 675. Thus, as with photoproduct 67a, hydrogenation of MA; W 613 over 10% palladium on charcoal in methanol resulted in the uptake of one equivalent of hydrogen. The mass spectrum showed a parent peak at m/e 172, corroborating the addition of one equivalent of hydrogen to photoproduct 61%. The nmr spectrum showed four aromatic protons at 6 6.65-7.07, a methyl doublet at 6 0.63 and a complex spectrum for the remaining nine protons. This nmr spectrum was identical to that of 81. However, similar to the reduction of 61:, the nmr spectrum of the material obtained from hydrogenation of 212 showed a small doublet at 6 0.88 and a multiplet at 6 6.95-7.02 corresponding to approximately 15% of a second product. This additional doublet and multiplet were not present in the nmr spectrum of the material obtained from the Wolff-Kishner reduction. 83 The discrepancies between the nmr spectrum of 73b and dihydro WM photoproduct 612 and similar discrepancies between the nmr spectrum of 81 and dihydro photoproduct 67s are most likely due to small amounts of ’VW product formed by reduction of either the C1-C2 or CZ'C8 bonds in the respective photoproducts (vide supra). A comparison between the product obtained from hydrogenation of 67a and 73a showed them to have identical WD’VW vpc retention times and the small methyl doublet at 6 1.17 in the nmr spectrum of dihydro photoproduct 67a had the same chemical shift as W the methyl doublet of 73a, thus indicating possible formation of small 'V'Vb amounts of 73a during hydrogenation of photoproduct 67a. A similar 'VV'b W1; reduction could have taken place during hydrogenation of 675. W C. Conclusions The results of the experiments reported here may be summarized by examining the two benzo—vinyl intermediates R and S. The di-n-methane ’b ’b photoisomerization of compounds S85 and 58a proceed preferrentially MA; 'VW through intermediate S in which the unpaired electrons are closest in ’b space to the methyl substituent. The extent of the preference depends CH 8” 8m 84 on the relation between the methyl substituent and the aromatic ring. When the methyl was anti to the aromatic ring the reaction was regio- specific. When the methyl was syn to the aromatic ring some product corresponding to intermediate R was obtained, but the reaction was still rather highly regioselective. Clearly, since the methyl substituent has no lone pair of electrons to share, this uneven distribution of di-n-methane photoproducts rules out any charge transfer interaction between the substituent and the proximal unpaired electrons as originally suggested by Hart and Murray.44 As with the systems containing hydroxyl and acetoxyl substituents at C9,45 the methyl substituent can be the only directive influence on the reaction since it is the only dissymmetric element in the reaction of £83 and 382. Furthermore, the influence must be through space rather than through bonds. If the directive effect of the methyl substituent occurred through bonds one might have expected the syn and anti epimers (Egg and 38%) to react similarly. The precise nature of this substituent effect is still not under- stood, and explanations remain speculative. One possibility would be that the substituent affects the transfer of triplet energy from the acetone sensitizer to the molecule. Such a mechanism would require that the functionality at C9 direct the excited acetone molecule to transfer its triplet energy asymmetrically to the substrate. This directive effect could be steric in the case of the methyl substituent or both steric and polar in the case of the hydroxyl and acetoxyl substituents. The sequence of events during electronic energy transfer clearly requires diffusion together of a triplet donor molecule and an acceptor molecule followed by an interaction between the two molecules. The 85 nature of the interaction is still unclear. However, Schenk53 has taken the viewpoint that during energy transfer the excited sensitizer interacts with the substrate to form a reactive complex or even a covalent bond. The activated substrate may then undergo isomerization or rearrangement. The sensitizer is eventually excised unharmed. There is no case for which this mechanism has been clearly demonstrated. However, if such a mechanism is operating, steric and polar effects may well influence the energy transfer process. Photolysis of 583 in the presence of a ketone with inherent steric strain such as diisopropyl ketone or other bulkier ketones as sensitizer could give further insight regarding this problem. If, for example, an increase in the ratio of photoproducts 212 to 26s with an increase in the size of the alkyl group in the ketone sensitizer is observed, one would have a strong case for the explanation presented in this thesis. Perhaps a careful kinetic study of the reaction of 583 and 38% using different alkyl ketones as sensitizer would also be fruitful. The formation of photoreduction products 222 and 222 is not without precedent. The same type of photochemically-induced reduction was observed by Sauers54 during irradiation of acetone solutions of several norbornene . derivatives, and by KroppSS during irradiation of alcoholic solutions of several cycloalkenes in the presence of aromatic hydrocarbon photosensi- tizers such as benzene or toluene. The reductions probably involve the triplets produce via energy transfer. Abstraction of hydrogen from acetone would lead to the l,(2 or 3),4-trihydro-l,4-ethanonaphthalene radicals (58aH- and 58sH-) and acetonyl radicals. In dilute solution, 58aH° and 58sH- can abstract a second hydrogen atom in competition with other radical recombination reactions. 86 It should be noted that the same type of reduction products were observed during the acetone sensitized irradiation of 525 and 52a. ’VW’VW Although 823 was previously detected, it was not characterized.45 Alcohol 82% was not previously detected (see Part N of Experimental Section). H0 H0 H H + hv 36?, acetone 525 825 'VW W» H H OH OH hv Agily + £32 acetone 52a 82a MA; W In conclusion, although the results presented here do not permit a conclusive explanation for the observed regioselectivity of the reactions, it clearly rules out the possibility of a charge transfer complex. At present a steric interaction between the substituent and the sensitizer best explains the observed results. EXPERIMENTAL A. General Procedures Except where otherwise noted, all nmr spectra were measured in CDC13 or CC14 solutions using TMS as an internal standard. The 60 MHz spectra were recorded on a Varian T-60 spectrometer and the 100 MHz spectra were recorded on a Varian HA-lOO spectrometer. Infrared spectra were recorded on a Perkin Elmer 237 grating spectrophotometer and cali- brated against a polystyrene film. Ultraviolet spectra were obtained with a Unicam SP-800 in 95% ethanol, unless otherwise noted. Mass spectra were obtained from a Hitachi-Perkin Elmer RMU-6 operated by Mrs. Ralph Guile. Melting points were determined with a Thomas-Hoover Melting Point Apparatus and are uncorrected. Varian Aerograph gas chromatographs were used. Analyses were performed by Spang Micro- analytical Laboratories, Ann Arbor, Michigan. B. 9-Methylene-l,4-dihydro-l,4-ethanonaphthalene (99). To a 250 m1 three-necked flask equipped with a magnetic stirrer, condenser, addition funnel and gas inlet was added 1.47 g (41.0 mmol) of 57% NaH in mineral oil. The flask was flushed with nitrogen and the NaH was washed free of mineral oil with hexane. To the NaH was added 15 ml of DMSO and the solution was heated to 60° in an oil bath until 87 88 hydrogen gas evolution ceased. The reaction mixture was cooled to room temperature and to it was added 20.3 g (45.0 mmol) of triphenyl- methylphosphonium bromide in 60 m1 of DMSO followed by addition of 7.00 g (41 mmol) of 9-oxo-l,4-dihydro-l,4-ethanonaphtha1ene (39)45 in 10 ml of DMSO. The reaction mixture was stirred at room temperature for 4 hours,di1uted with 500 m1 of water and extracted with pentane. The pentane extracts were combined, washed with water, saturated NaCl solution and dried over anhydrous MgSO4. The pentane was rotary evaporated leaving behind 6.80 g of a clear liquid. Chromatography on silica gel using hexane as the eluent yielded 5.80 g (34.5 mmol, 84.0%) of 62: ir (neat) 3060 (m), 3010 (w), 2960 (m), 2810 (w), 1650 (m), 1475 (m), 1465 (w), 1345 (w), several weak bands from 1300 to 900, 880 (m), 845 (m), 780 (m), 760 (s), 700 cm'1 (s); nmr (CDCIS) 6 2.13-2.33 (m, 2H, bridge methylene H's), 3.85-4.13 (m, 1H, bridgehead H), 4.25-4.45 (m, 1H, bridgehead H), 4.63-4.78 (m, 1H, =CH2), 4.93-5.08 (m, 1H, =CH2), 6.43-6.75 (m, 2H, vinyl H's at C-2 and C-3), 7.00-7.23 (m, 4H, aromatic H's); mass spectrum (70 eV) m/e (rel. intensity) 168 (89), 167 (56), 153 (35), 128 (100), 115 (15), 102 (8), 63 (10), 51 (10). Anal. Calcd for C13H12: C, 92.81; H, 7.19 Found: C, 92.85; H, 7.18 C. Syn-Q-methyl-l,4-dihydro-l,4-ethanonaphthalene (38$). In a dry three-necked flask equipped with a condenser, magnetic stirrer, thermometer, nitrogen gas inlet and a pressure-equilibrated funnel were placed 50 ml of a 1.08M solution of 82H6 (52.0 mmol) in tetrahydrofuran. The reaction flask was immersed in an ice salt bath and cooled to -S°. From the dropping funnel 11 ml (104 mmol) of 89 2-methy1-2-butene was added dropwise to the stirred solution over a period of 30 minutes. The disiamylborane was permitted to remain an additional 15 hours at 0-5° under a nitrogen atmosphere, then used for the hydroboration of 9-methy1ene-l,4-dihydro-l,4-ethanonaphthalene (62). To a lOO-ml three-necked flask equipped with a condenser, magnetic stirrer, thermometer, nitrogen gas inlet and a serum cap was added 2.50 g (14.9 mmol) of 60 in 15 ml of tetrahydrofuran. The flask was immersed in an ice-salt bath, the solution cooled to 0°, and 18.2 ml (14.9 mmol) of disiamylborane in tetrahydrofuran was added dropwise by means of a syringe to the stirred solution. After being stirred for one hour at 0° the reaction mixture was allowed to remain fer approximately 3 hr at room temperature. To the reaction mixture was added 5.0 m1 of pro- pionic acid, then 30 ml of diglyme. The reflux condenser was substituted by a distillation head and enough tetrahydrofuran was distilled to main- tain a reflux temperature of 130°. The reflux condenser was replaced and the residual reaction mixture was refluxed at 130° for 8 hr. The reaction mixture was then diluted with 200 ml of water and extracted with pentane. The pentane extracts were combined, washed with saturated NaHCOS, water, saturated NaCl solution and dried over anhydrous MgS04. The pentane was evaporated leaving behind a yellowish liquid. Chroma- tography on silica gel using hexane as the eluent yielded 1.59 g (9.35 mmol, 63.0%) of 585, a clear liquid: ir (neat) 3040 (m), 3010 (w), 2950 (s), 2860 (m), 1475 (m), 1460 (m), 1380 (w), 1350 (w), several weak bands from 1275-800, 770 (m), 690 cm'1 (s); nmr (cnc13) 6 0.62 (d, 3H, J = 7.0 Hz; methyl), 0.72-1.05 (m, 1H, methine), 1.53-2.08 (m, 2H, methylene), 3.42—3.65 (m, 1H, bridgehead), 3.67-3.92 (m, 1H, bridgehead), 6.27-6.77 (m, 2H, vinyl), 7.13 (broad s, 4H, aromatic); mass spectrum (70 eV) m/e (rel. intensity) 170 (25) 165 (16), 155 (32), 153 (44), 90 152 (49), 141 (55), 128 (2330), 115 (77), 102 (100), 89 (31), 77 (94), 63 (107), 51 (147); uv (95% ethanol) Amax 270 nm (6 411), 263 (411), 256 (213, sh), 219 (3730, sh), 203. Anal. Calcd for C13H14: C, 91.71; H, 8.29 Found: C, 91.60; H, 8.11 D. Dimethyl 9-methy1ene-l,2,3,4-tetrahydro-l,4-ethanonaphthalene-2,3- dicarboxylate (62). 'V‘b Following the procedure for the synthesis of 62, a solution of sodium methylsulfinyl carbanion was prepared from 4.25 g (0.10 mol) of 57% NaH in mineral oil and 60 ml of dimethyl sulfoxide. To this solution was added 35.7 g (0.10 mol) of triphenylmethylphosphonium bromide in 100 ml of dimethyl sulfbxide. The reaction mixture was stirred for 30 minutes at room temperature. To a 500-ml three-necked flask equipped with a magnetic stirrer, condenser, addition funnel, and nitrogen gas inlet was added 28.8 g (0.10 mol) of dimethyl 9-oxo-l,2,3,4-tetrahydro-l,4-ethanonaphtha1ene- 2,3-dicarboxy1ate (61)48 in 100 m1 of dimethyl sulfOxide. By means of a U-tube the ylid solution prepared above was transferred to the addition funnel and added dropwise over a period of 30 minutes to the keto-diester. The reaction mixture was stirred overnight, diluted with 1000 m1 of water and extracted with methylene chloride. The methylene chloride extracts were combined, washed with water, saturated NaCl solution and dried over anhydrous MgSO4. Evaporation of the methylene chloride gave a dark-brown oil which on vacuum distillation, bp 135-140° at 0.05 mm Hg, yielded 20.19 g (70.7 mmol, 70.7%) of 22, a clear viscous liquid: ir (neat) 3070 (w), 3015 (m), 2940 (m), 2840 (m), 1725 (doublet, s), 1650 (W), 91 1480 (m), 1465 (m), 1375 (w), 1310 (m), 1280 (m), 1205 (broad s), 1115 (m), 1030 (s), 895 (m), 875 (w), 765 cm.1 (s); nmr (CDC13) 6 1.85-2.78 (m, 2H, C-10 methylene H's), 3.02-3.30 (m, 1H, C-2 or C-3 methine), 3.60 (broad s, 5H, C-2 or C-3 methine, bridgehead and carboxylate methyl), 3.80 (s, 3H, carboxylate methyl), 3.93-4.08 (m, 1H, bridgehead), 4.68-4.85 (m, 1H, =CH2), 4.93-5.22 (m, 1H, -CH2), 7.23 (s, 4H, aromatic H's); mass spectrum (70 eV) m/e (rel intensity) 286 (3), 255 (3), 227 (3), 195 (3), 167 (9), 164 (4), 145 (13), 143 (13), 142 (100), 141 (32), 128 (7), 115 (7), 113 (10), 59 (3). Anal. Calcd for C17H1804: C, 71.31; H, 6.34 Found: C, 71.26; H, 6.40 E. Dimethyl Anti-9-methyl-l,2,3,4-tetrahydro-l,4-ethanonaphthalene-2,3- dicarboxylate (63). m A solution of 62 (6.00 g, 21.0 mmol) in 100 ml of methanol was hydrogenated in the presence of 0.8 g of 10% Pd on charcoal at room temperature and 1 atm of pressure. The reduction was stopped after one equivalent of hydrogen was absorbed. The catalyst was removed by filtration, and the solvent was evaporated leaving 5.50 g (19.3 mmol, 91.5%) of 23 as a clear oil which solidified on standing. Recrystallization from chloro- form yielded a white solid mp 70-72°; ir (neat) 3060 (w), 3010 (w), 2940 (m), 2870 (w), 1730 (s), 1480 (w), 1460 (m), 1430 (m), 1375 (w), 1305 (m), 1280 (m), 1235 (s), 1195 (s), 1120 (w), 1080 (m), 1025 (m), 945 (w), 920 (w), 865 (w), 795 (w), 760 cm‘1 (m); nmr (coc13) c 1.03 (d, 3H, J = 6.5 Hz, C-9 methyl), 1.15-1.98 (m, 3H, C-9 methine and methylene), 2.87-3.10 (m, 1H, C-2 or C-3 methine), 3.37-3.53 (m, 3H, C-2 or C-3 methine and bridgeheads), 3.58 (s, 3H, carboxylate methyl), 3.78 (s, 3H, 92 carboxylate methyl), 7.20 (s, 4H, aromatic H's); mass spectrum (70 eV) m/e (rel. intensity) 288 (22), 256 (26), 228 (21), 213 (9), 186 (61), 169 (26), 155 (45), 143 (69), 128 (100), 114 (57), 87 (11), 77 (9), 59 (25). Anal. Calcd for C17H2004: C, 70.81; H, 6.99 Found: C, 70.90; H, 7.11 F. Anti-9-methy1-l,2,3,4-tetrahydro-1,4-ethanonaphthalene-2,3-dicar- boxylic Acid (64). mm A solution of 80 m1 of approximately 20% aqueous sodium hydroxide was placed in a 250-ml Erlenmeyer flask and to it was added 5.50 g (19.3 mmol) of diester 23. The mixture was then heated to boiling for 1.5 hr during which time the diester completely dissolved. After the solution had been boiled vigorously for 30 minutes, it was cooled to room temperature and concentrated hydrochloric acid was added with stirring until the solution was acidic to litmus paper. The white solid which precipitated after cooling was filtered, washed with water and dried. The yield of dicarboxylic acid 64, mp 179-183° (recrystallized from methanol) was 5.00 g (19.1 mmol, 98.5%): ir (KBr) 3650-2300 (s), 2950 (m), 2870 (m), 1700 (s), 1485 (w), 1470 (w), 1420 (m), 1310 (m), 1290 (m), 1260 (s), 1225 (m), 1085 (w), 925 (m, broad), 775 (m), 755 cm‘1 (m); nmr (acetone-d6) 6 1.25 (d, 3H, J = 7.0 Hz, methyl), l.30-1.85 (m, 3H, C-9 methine and methylenes), 2.82-3.02 (m, 1H, carboxylic methine), 3.17-3.73 (m, 3H, carboxylic methine and bridgeheads), 7.20 (s, 4H, aromatic H's); mass spectrum (70 eV) m/e (rel intensity) 260 (40), 242 (9), 214 (19), 172 (38), 153 (94), 129 (100), 118 (23), 115 (18), 100 (16), 91 (7), 89 (4), 77 (11), 72 (11), 57 (18). 93 Anal. Calcd for C15H1604: C, 69.21; H, 6.20 Found: C, 69.23; H, 6.20 G. Anti-9-methy1-l,4-dihydro-l,4-ethanonaphthalene (58g). A solution of 5.0 g (19.1 mmol) of 24 in a mixture of 150 m1 of pyridine, 3 m1 of triethylamine and 50 m1 of water was placed in a rectangular jar which contained two Pt electrodes (5 x 7.5 cm) separated by 4.5 cm, a water cooling coil and a nitrogen dispersion tube. A constant unfiltered DC potential (110 V) was placed on the electrodes; the initial current was 800 mA. After 15 hr of electrolysis the current had decreased to 200 mA and the solution was black. The solvents were evaporated in vacuo and the black residue was extracted with hot pentane. The pentane extracts were combined, washed successively with 15% hydro- chloric acid, saturated NaHC03, water, saturated NaCl solution and dried over anhydrous MgS04. Evaporation of the solvent yielded 0.842 g (4.96 mmol, 26%) of 582. Column chromatography on alumina using hexane as the eluent yielded a clear liquid: ir (neat) 3040 (m), 3010 (w), 2950 (s), 2860 (m), 1475 (m), 1375 (w), 1350 (w), several weak bands from 1275-900, 775 (w), 755 (s), 705 cm‘1 (s); nmr (cuc13) 5 0.98 (d, 4H, methyl superimposed on C-9 methine, J = 6.5 Hz), 1.48-2.00 (m, 2H, methylene H's), 3.57-3.97 (m, 2H, bridgehead H's), 6.37-6.75 (m, 2H, vinyl H's), 7.15 (broad s, 4H, aromatic H's); uv (95% ethanol) and mass spectra are identical with those of 58g. Anal. Calcd for C13H14: C, 91.71; H, 8.29 Found: C, 91.50; H, 8.27 94 H. Irradiation of Syn-9-methy1-1,4-dihydro-l,4-ethanonaphthalene (585). MA; A solution of 285 mg (1.68 mmol) of 58: in 470 m1 of acetone was photolyzed through pyrex with a 450-W Hanovia lamp, while a slow stream of nitrogen passed through the solution. The progress of the reaction was followed by vpc. After 3.5 hr the acetone was evaporated and the residue was extracted with pentane. The mixture was analyzed by vpc (5 ft x 0.125 in., 10% FFAP on chromosorb W, AW-DMCS 80/100, 128°, N2 flow rate 30 m1/min). Three products [retention times in minutes, % yield versus a calibrated internal standard (pentadecane)] were observed: 22: (10.0, 22.7%), 262 (20.2 10.8%), 27: (24.0, 63.8%). The products were isolated by preparative scale vpc (10 ft x 0.25 in., 20% FFAP on chromosorb W P/G 30/60, 150°, He flow rate 40 m1/min). syn-9-methy1-l,2,3,4-tetrahydro-l,4-ethanonaphthalene (652): ir (neat) 3060 (w), 3025 (w), 3010 (w), 2925 (s), 2850 (m), 1480 (m), 1450 (m), 1375 (w), several weak bands from 1350 to 840, 760 cm-1 (s); nmr (CDCIS) 6 0.58 (d, 3H, J = 6.5 Hz, methyl partially superimposed on C-9 methine), 0.60-1.07 (m, 1H, C-9 methine), 1.15-2.28 (m, 6H, methylene H's), 2.60-2.77 (m, 1H, bridgehead), 2.83-3.05 (m, 1H, bridgehead), 7.20 (s, 4H aromatic H's); mass spectrum (70 eV) m/e (rel. intensity) 172 (33), 143 (14), 130 (100), 115 (20), 102 (3.5), 77 (6), 63 (S), 51 (6). Anal. Calcd for C H C, 90.64; H, 9.36 13 16‘ Found: C, 90.92; H, 9.37 Syn-7-methyl-3,4-benzotricyclo[3.3.0.02’°]oct-3-ene (62$); ir (neat) 3040 (m), 3020 (m), 2930 (s), 2860 (m), 1480 (s), 1465 (m), 1380 (w), several weak bands from 1350 to 850, 810 (m), 775 (m), 760 (s), 745 (m), 725 cm‘1 (w); HA-100 nmr (0001 5 0.48 (d, 3H, J = 7.5 Hz, 3) methyl), 1.39 (d, 1H, J = 11.0 Hz, syn-H at C-6), 1.76-1.98 (m, 1H, 95 cyclopropyl H adj. to methyl), 2.16-2.34 (m, 1H, benzylic cyclopropyl H), 2.46-2.77 (m, 2H, C-7 methine and anti-H at C-6), 2.78-2.99 (m, 1H, cyclo- propyl methine), 3.43-3.60 (m, 1H, benzylic methine), 6.73-7.17 (m, 4H, aromatic H's); mass spectrum (70 eV) m/e (rel. intensity) 170 (32), 169 (8), 155 (68), 141 (16), 128 (100), 115 (58), 102 (7), 91 (9), 89 (6), 77 (12), 65 (8), 63 (8), 55 (22). Double resonance was performed; only those bands which were affected are included. Irradiation of the doublet at 6 0.48 partially collapsed the multiplet at 6 2.46-2.77; 3.43-3.60 irradiated, 2.78-2.99 and 2.46-2.77 partially collapsed; 2.16- 2.34 irradiated, 2.78-2.99 and 1.76-1.98 partially collapsed; 2.46-2.77 irradiated, 3.43-3.60, 1.76-l.98, 1.39 and 0.48 collapsed. Anal. Calcd for C H 13 14‘ Found: C, 91.68; H, 8.33 C, 91.71; H, 8.29 Syn-6-methy1-3,4-benzotricyclo[3.3.0.02’8]oct-3-ene (272): ir (neat) 3040 (m), 3020 (m), 2930 (s), 2860 (m), 1480 (s), 1465 (m), 1380 (w), several weak bands from 1350 to 825, 790 (w), 775 (w), 755 (s), 720 cm"1 (w); HA-100 nmr (CDC13) 6 0.51—0.91 (m, 1H, syn-H at C-7), 0.74 (d, 3H, J = 7.0 Hz, methyl superimposed on syn-H at C-7), 1.60-1.93 (m, 2H, cyclopropyl H adj. to methylene and anti-H at C-7), 2.12-2.38 (m, 1H, benzylic cyclopropyl H), 2.44-2.95 (m, 2H, cyclopropyl methine and C-6 methine), 3.17-3.33 (m, 1H, benzylic methine), 6.85-7.37 (m, 4H, aromatic H's); mass spectrum (70 eV) m/e (rel. intensity) 170 (27), 169 (5), 155 (4), 141 (12), 128 (100), 115 (32), 102 (5), 91 (3), 89 (4), 77 (6), 63 (7), SS (9). Double resonance was performed; only those bands which were affected are included. Irradiation of the doublet at 6 0.74 partially collapsed the multiplet at 6 2.44-2.95 and 1.60-1.93; 1.60—1.93 irradiated, 2.12-2.38, 2.44-2.95 and 0.51-0.91 partially collapsed; 96 2.12-2.38 irradiated, 2.44-2.95 and 1.60-1.93 partially collapsed; 2.44-2.95 irradiated, 3.17-3.33, 2.13-2.38 and 1.60-1.93 partially col- lapsed; 3.17-3.33 irradiated 2.44-2.95 partially collapsed. Anal. Calcd for C13H14: C, 91.71; H, 8.29 Found: C, 91.91; H, 8.23 I. Irradiation of Anti-9-methy1-1,4-dihydro-l,4-ethanonaphthalene 58a). A solution of 258 mg (1.56 mmol) Of.§§€ in 470 m1 of acetone was irradiated for 3.5 hr in the same manner as 5mg. The mixture was analyzed by vpc (5 ft x 0.125 in., 10% FFAP on chromosorb W, AW-DMCS 80/100, 125°, N2 flow rate 30 ml/min). Two products [retention time in minutes, % yield versus a calibrated internal standard (pentadecane)] were observed: 65a § (16.2, 11.7%), 632 (24.8, 81.8%). The products were isolated by pre- parative scale vpc (10 ft x 0.25 in., 20% FFAP on chromosorb W P/G 30/60, 150°, He flow rate 40 ml/min). Anti-9-methy1-l,2,3,4-tetrahydro-1,4-ethanonaphthalene (252): ir (neat) 3060 (w), 3040 (w), 3020 (w), 2940 (s), 2860 (m), 1480 (m), 1465 (w), 1455 (w), 1380 (w), several weak bands from 1350 to 800, 765 (s), 750 cm‘1 (w); nmr (c0c13) 5 1.10 (d, 4H, J = 6.5 Hz, methyl super- imposed on C-9 methine), 1.16-2.20 (m, 6H, methylene H's), 2.52-2.68 (m, 1H, bridgehead), 2.73-2.93 (m, 1H, bridgehead), 6.98 (s, 4H, aromatic H's); mass spectrum (70 eV) is identical to 653. Anal. Calcd for C13H16: C, 90.64; H, 9.36 Found: C, 90.86; H, 9.43 Anti-6-methy1-3,4-benzotricyclo[3.3.0.02’8 ]oct-3-ene (61%): ir (neat) 3060 (w), 3040 (w), 2940 (m), 2920 (m), 2860 (w), 1475 (m), 1460 (m), 1375 (w), several weak intensity bands from 1310-810, 800 (m), 775 (m), '97 760 em'1 (5); HA-100 nmr (00013) 5 1.08 d, 3H J = 7.0 Hz, methyl), 1.25- 2.16 (m, 4H), 2.26-2.48 (m, 1H), 2.70-2.93 (m, 1H), 3.17 (d, 0.5H, J = 5.5 Hz), 3.46-3.60 (m, 0.5H), 6.92-7.37 (m, 4H, aromatic H's); mass spectrum (70 eV) is identical to 67%. A 270 MHz nmr (Bruker) was deter- d;50 mined and double resonance was performe the results are given in Table IV and shown in the Appendix section (Figures 10 and 11). Anal. Calcd for C H C, 91.71; H, 8.29 13 14‘ Found: C, 91.56; H, 8.20 J. syn-9-methy1-l,2,3,4-tetrahydro-l,4-ethanonaphthalene (655). A solution of 58% (100 mg, 0.588 mmol) in 4 m1 of MeOH was hydro- genated in the presence of 10 mg of 10% Pd on charcoal at room temper- ature and 1 atm of pressure. The reduction was stopped after 1 equivalent of hydrogen was absorbed. The catalyst was removed by filtration, and the solvent was evaporated, leaving behind 91 mg (5.29 mmol, 90%) of 652 which was not purified. The nmr and ir spectra were identical to those of the product obtained from irradiation of syn-9-methyl-l,4-dihydro-1,4-ethanonaphthalene (585). MA: K. Anti-9-methy1-1,2,3,4-tetrahydro-l,4-ethanonaphthalene (223). Hydrogenation of 58a (98.2 mg, 0.588 mmol), as described for 585 m 'VV’o (vide supra), gave 82.1 mg (0.477 mmol, 81.2%) of 252. The nmr and ir spectra were identical to those of the product obtained from irradiation of anti-9-methy1-l,4-dihydro-1,4-ethanonaphthalene (582). 98 L. 7-Methy1ene-3,4-benzotricyclo[3.3.0.02’8]oct-3-ene (69). Following the procedure given for 60, 69 was prepared in 70% yield from 0.395 g (6.90 mmol) of 57% NaH in mineral oil, 2.46 g (6.90 mmol) of triphenylmethy1phosphonium bromide, and 1.17 g (6.90 mmol) of 3,4-benzotricyclo[3.3.0.02’8]oct-3-ene-7-one (g§)49 in dimethyl sulfoxide. Chromatography of the crude product on alumina using hexane as the eluent yielded 0.810 g (4.83 mmol, 70.0%) of 62, a clear liquid which solidified on standing: mp 55-56°; ir (neat) 3060 (m), 3030 (m), 2970 (w), 2930 (m), 2875 (w), 2810 (w), 1650 (m), 1475 (s), 1455 (m), 1425 (w), 1310 (w), several weak bands from 1200 to 925, 875 (s), 800 (m), 720 cm" (s); nmr (cc14) 5 1.97-2.67 (m, 3H, methylene H's and cyclopropyl H adj. to double bond), 2.80-3.28 (m, 2H, benzylic cyclo- propyl H and cyclopropyl methine, 3.48-3.75 (m, 1H, benzyl methine), 4.73-4.97 (m, 2H, vinyl H's), 6.83-7.28 (m, 4H, aromatic H's); mass spectrum (70 eV) m/e (rel. intensity) 168 (100), 167 (96), 153 (62), 152 (39), 141 (10), 128 (58), 115 (21) 102 (8), 83 (21), 77 (8), 76 (10), 63 (7), 51 (7). Anal. Calcd for C H 13 12‘ Found: C, 92.87; H, 7.08 C, 92.81; H, 7.19 M. syn-7-methy1-3,4-benzotricyclo[3.3.0.02’8]oct-3-ene (223). To a dry lS-ml three—necked flask equipped with a condenser, serum cap, magnetic stirrer, and nitrogen gas inlet was added 3 m1 of freshly distilled tetrahydrofuran. The flask was immersed in an ice- salt bath, cooled to approximately 0° and 1.10 ml (1.19 mmol) of 1.08M B2H6 in tetrahydrofuran was added by means of a syringe followed 99 by the addition of 200 mg (1.19 mmol) of 69 in 3 m1 of tetrahydrofuran. After being stirred for 0.5 hr at 0°, the reaction mixture was allowed to remain for 2 hours at room temperature. To the reaction mixture was added 1.0 m1 of propionic acid followed by 7 m1 of diglyme. The tetrahydrofuran was distilled and the reaction mixture was heated at 130° for 8 hours. The mixture was then diluted with 100 ml of water and extracted with pentane. The pentane extracts were combined, washed with saturated NaHCO , water, saturated NaCl solution and dried 3 over anhydrous MgSO Chromatography on alumina using hexane as the 4. eluent yielded 0.122 g (0.718 mmoles, 60.5%) 0f.38§° The nmr and ir were identical to those of the product obtained from irradiation of syn-9-methyl-1,4-dihydro-l,4-ethanonaphthalene (383). N. 3,4-Benzotricyclo[3.3.0.02’8]oct-3-ene-6-one (39). A solution of 1.15 g (6.70 mmol) of a mixture of syn and anti-9- hydroxy-1,4-ethanonaphthalene (Egg and 523)45 in 470 ml of acetone was flushed with nitrogen for 20 minutes and photolyzed through pyrex with a 450-W Hanovia lamp, while a slow stream of nitrogen passed through the solution. The progress of the reaction was fellowed by vPc (5 ft x 0.125 in, 10% FFAP on chromosorb W, AW-DMCS 80/100, 190, N2 flow rate 15 ml/min). After the mixture was irradiated for 3 hr the vpc trace showed complete disappearance of starting material and the fermation of four products in a ratio of 2:1:14:7 with retention times of 12.8, 17.0, 18.0 and 23.7 minutes. Comparison of retention time by spiking, indicated that the products with retention times of 12.8 and 17.0 minutes correspond to syn and anti-9-hydroxy-1,2,3,4- tetrahydro-l,4-ethanonaphthalene (825 and 82a, obtained by catalytic mmm mmm 100 hydrogenation of 523 and 52%) and products with retention time of 18.0 and 23.7 minutes correspond to alcohols 533 and 53%. The photolysis was repeated three times, the products combined and the solvent evapor- ated leaving behind a dark yellow oil which was oxidized as described below. Chromium trioxide, 20 g (0.20 mol), was added to a mechanically stirred solution of 32 g (0.40 mol) of pyridine in 500 ml of methylene chloride at 0°. The flask was stoppered with a drying tube containing drierite, and the deep burgundy solution was allowed to warm to room temperature and stirred for 0.5 hr. At the end of this period, a solution of the alcohols obtained above in 15 ml of methylene chloride was added in one portion. A tarry black deposit separated immediately. After being stirred for one additional hour at room temperature, the solution was decanted from the residue. The residue salts were washed with methylene chloride, the organic solutions were combined and washed successively with 5% NaOH, 5% HCl, saturated NaHCOS, saturated NaCl solution and dried over anhydrous MgSO The methylene chloride was 4. rotary evaporated leaving 3.73 g of a yellow oil. Chromatography on a 4.5 ft x 1 in column packed with silica gel using 10% ether in hexane as the eluent afforded 0.300 g (1.76 mmol, 6.6%) of a clear oil which was crystallized from hexane to give white crystals of 9-oxo-1,4- dihydro-l,4-ethanonaphthalene, mp 54-56° (111;.56 mp 56.5-58°), 0.283 g (1.65 mmol, 6.2%) of 9-oxo-1,2,3,4-tetrahydro-l,4-ethanonaphthalene which was crystallized from hexane to give white crystals, mp 33-34 56 mp 30-31°), and 2.50 g (14.7 mmol, 55%) of white crystals which were recrystallized from hexane to give pure ketone 724$, (11:. mp 57-59°. 101 O. syn-G-hydroxy1-6-methy1-3,4-benzotricyclo[3.3.0.02’8]oct-3-ene (71). To a 50-ml three-necked flask equipped with a condenser, serum cap, magnetic stirrer and nitrogen gas inlet was added 0.500 g (2.94 mmol) of 3,4-benzotricyclo[3.3.0.02’8]oct-3-ene-6-one (10) in 20 m1 of tetra- hydrofuran. The solution was cooled to approximately 0° there was added, by means of a syringe, 2.33 ml (3.50 mmol) of a 1.50M solution of methyllithium in ether. The reaction mixture was stirred under N atm 2 at room temperature for 3 hr, then quenched with saturated NH Cl solution 4 and extracted with ether. The ether extracts were combined, washed with water, saturated NaCl solution and dried over anhydrous MgSO4. Evapor- ation of the solvent yielded 0.529 g (2.84 mmol, 97%) of a yellowish oil which solidified on standing. Recrystallization from cold pentane gave a white solid, 71: mp 52-53°; ir (CC14) 3575 (m), 3450 (m, broad), 3040 (s), 2980 (s), 2950 (s), 2870 (m), 1475 (s), 1460 (m), 1445 (m), 1380 (s), 1355 (s), 1300 (m), 1215 (s), 1150 (s, broad), 100 (w), 1 955 cm" (m). The band at 3575 cm'1 did not change in relative intensity as a function of the concentration of 11 in CC14, whereas the intensity of the band at 3450 cm.1 decreased drastically with a decrease in con- centration of $1. Nmr (00013) 5 0.95 (d, 1H, J = 11.5 Hz, syn c-7 H), 1.50 (s, 3H, methyl), 1.57-1.98 (m, 3H, hydroxyl H superimposed on anti C-7 H and cyclopropyl H adj. to methylene), 2.18-2.53 (m, 1H, benzylic cyclopropyl H), 2.58-2.92 (m, 1H, cyclopropyl methine), 3.00-3.17 (m, 1H, benzylic H), 6.83-7.22 (m, 4H, aromatic H's); mass spectrum (70 eV) m/e (rel intensity) 186 (10), 168 (16), 167 (12), 153 (10), 143 (25), 128 (100), 115 (14), 101 (4), 71 (12), 43 (42). Anal. Calcd for C H O: C, 83.83; H, 7.58 13 14 Found: C, 83.77; H, 7.50 102 P. syn-6-methyl-3,4-benzotricyc1o[3.3.0.02’8]oct-3-ene (27:). To a solution of 154 mg (0.83 mmol) of alcohol 71 in 10 ml of anhydrous benzene, cooled to approximately 5°, was added 130 mg (1.66 mmol) of pyridine followed by 0.10 ml (160 mg, 1.35 mmol) of thionyl chloride. The mixture was stirred at approximately 5° for 15 minutes then added to ice-water. The organic layer was separated and the aqueous layer was extracted with benzene. The benzene extracts were combined and washed successively with cold 15% HCl, saturated NaHCO , saturated 3 NaCl solution and dried over anhydrous MgSO4. The benzene was evaporated leaving behind 145 mg (0.71 mmol, 85.4%) of chloride 72 which was not purified further. The infrared spectrum (neat) showed no VO-H band at 3450 cm'l, but the appearance of a vC-Cl band at 900 cm-1. To a 145 mg sample (0.709 mmol) of chloride 72 prepared as above, placed in a dry test tube, was added 254 mg (0.850 mmol) of tri-n-butyltin hydride and 3.0 m1 of a 5.0 mm solution of azobisisobutyronitrile in toluene. The mixture was flushed with nitrogen, the test tube sealed with a serum cap andirradiated at room temperature in a Rayonet-Chamber with 3500 A° lamps for 6 hours. The mixture was analyzed by vpc (5 ft x 0.125 in., 10% FFAP on chromosorb W, AW-DMCS 80/100, 125°, N2 flow rate 30 m1/min) and showed a main product with the same retention time as 2Zi° Thick-layer chromatography on alumina using hexane as the eluent yielded 40 mg of starting chloride and 75 mg (0.44 mmol, 85%) of 87:. The nmr and ir spectra were identical to those of the product obtained from irradiation of syn-9-methy1-l,4-dihydro—1,4-ethanonaphthalene (58$). mmm 103 Q. Hydrogenation of Anti-6-methyl-3,4-benzotricyclo[3.3.0.02’8]- oct-3-ene (67a). 'VVb A solution of 212 (34 mg, 0.20 mmol) in 4 m1 of methanol was hydrogenated in the presence of 7 mg of 10% Pd on charcoal. The reduction was stopped after one equivalent of hydrogen was absorbed. The catalyst was removed by filtration and the solvent was evaporated, leaving 30 mg (0.17 mmol, 87%) of a clear liquid. Analysis of the crude product by vpc (5 ft x 0.125 in, 10% FFAP on chromosorb W, AW-DMCS 80/100, 125°, N2 flow rate 30 ml/min) showed only one product. Pre- parative gas chromatography (5 ft x 0.25 in, 20% SE-30 on chromosorb W, 30/60, 130°, He flow rate 30 m1/min) afforded the pure anti-6-methyl- 3,4-benzobicyclo[3.2.0]oct-3-ene (122); ir (CC14) 3060 (w), 3010 (w), 2940 (s), 2860 (m), 1490 (m), 1460 (m), 1380 (w), several weak intensity bands from 1350-825 cm"1 ; nmr (CDC13) 6 1.00 (d, 3H, J = 7.0 Hz, methyl), 1.08-2.25 (m, 5H), 2.26-3.25 (m, 4H), 6.68-7.08 (m, 4H, aromatic H's). The nmr spectrum also showed a small doublet at 6 1.17, J = 7.0 Hz, and a multiplet at 6 6.95-7.10 which probably correspond to traces (5-10%) of an isomer of 722. The mass spectrum (70 eV) m/e (rel intensity): 172 (24), 157 (1.5), 143 (6), 129 (100), 117 (4), 116 (10), 115 (15), 91 (4), 89 (2), 77 (4), 63 (3), 51 (4). Anal. Calcd for C13Hl6‘ C, 90.64; H, 9.36 Found: C, 90.80; H, 9.36 R. Anti-6-methyl-3,4-benzotricyclo[3.3.0.02’8]oct-3-ene-7-one (74) In a dry three-necked flask equipped with a condenser, magnetic stirrer, nitrogen gas inlet and a serum cap was added 424 mg (2.63 mmol) 104 of hexamethyldisilazane in 7 m1 of dry tetrahydrofuran. The flask was immersed in an ice-salt bath, the solution was cooled to approximately 0°, and 1.17 ml of 2.06M n-butyllithium in hexane was added dropwise by means of a syringe to the stirred solution. After the mixture was stirred for 2 hr at 0°, 406 mg (2.39 mmol) of ketone 6849 in 4 m1 of tetrahydrofuran was added and stirring was continued for 0.5 hr. To the reaction mixture was then added 0.5 m1 of methyl iodide. Stirring in the cold was continued for 1 hr followed by stirring at room tempera- ture for 8 hr. The solution was diluted with 100 m1 of water and extracted with ether. The ether extracts were combined, washed with water, saturated NaCl solution and dried over anhydrous MgSO The 4. solvent was evaporated leaving 412 mg (2.24 mmol, 98%) of a yellowish oil. Column chromatography on alumina using 15% ether in hexane as the eluent yielded pure 1:: ir (neat) 3040 (m), 2960 (s), 2920 (m), 2860 (m), 1720 (s), 1480 (m), 1465 (m), 1380 (w), several medium intensity bands from 1350-850, 830 (s), 820 (s), 770 cm'1 (s); nmr (CDC13) 6 1.32 (d, 3H, J = 7.0 Hz, methyl), l.90-2.43 (m, 2H, C-6 methine and cyclopropyl H adj to carbonyl), 2.68-3.02 (m, 1H, benzylic cyclopropyl H), 3.12-3.50 (m, 2H, benzylic and cyclopropyl methines) 6.80-7.28 (m, 4H, aromatic H's); mass spectrum (70 eV) m/e (rel intensity) 184 (7), 169 (3), 155 (5), 141 (18), 128 (100), 115 (12), 100 (3), 91 (1.5), 89 (2), 77 (5), 63 (5), 51 (5). Anal. Calcd for C H 0: C, 84.75; H, 6.57 13 12 Found: C, 84.61; H, 6.54 S. Anti-6-methy1-3,4-benzobicyclo[3.3.0]oct-3-ene-7-one (15). Hydrogenation of 74 (235 mg, 1.28 mmol), as described for 67a ’Vb ’VW 105 (vide supra), gave 209 mg (1.12 mmol, 87.8%) of a clear liquid. Analysis of the crude product by vpc (5 ft x 0.125 in, 10% FFAP on chromosorb W, AW-DMCS, 80/100, 185°, N2 flow rate 30 m1/min) showed only one product. Preparative thick-layer chromatography on alumina using 10% ether in hexane afforded pure 75: ir (neat) 3065 (w), 3015 (w), 2960 (m), 2930 (s), 2840 (w), 1735 (s), 1480 (m), 1460 (m), 1410 (w), 1380 (w), 1175 (m), several weak bands from 1150-800, 760 cm.1 (s); nmr (CDC13) 6 1.33 (d, 3H, J = 7.0 Hz, methyl), 2.10-3.47 (m, 7H), 7.08 (s, 4H, aromatic H's); mass spectrum (70 eV) m/e (rel intensity) 186 (73), 171 (10), 157 (15), 143 (16), 129 (100), 115 (57), 91 (10), 89 (6), 77 (10), 65 (10), 63 (10), 57 (18). T. Anti-6—methy1-3,4-benzobicyclo[3.3.0]oct-3-ene (73a). mmm A solution of 150 mg (0.807 mmol) of ketone 72, 2.5 m1 of hydrazine hydrate, and 2.0 g of KOH in 15 m1 of triethylene glycol was refluxed for 30 hr. The reaction mixture was then cooled, added to 100 m1 of 15% HCl and extracted with ether. The ether extracts were combined and washed successively with saturated NaHCOS, water, saturated NaCl solution and dried over anhydrous MgSO4. Evaporation of the solvent yielded 111 mg (0.646 mmol, 80%) of a light yellow liquid. Analysis of the crude product by vpc (5 ft x 0.125 in, 3% SE-30 on chromosorb W, 80/100, 120°, N2 flow rate 30 ml/min) showed only one product. Preparative thick-layer chroma- tography on alumina using pentane as the eluent afforded pure 132: ir (neat) 3065 (w), 3015 (w), 2940 (s), 2860 (m), 1480 (m), 1465 (m), 1380 (w), several weak bands from 1350-775, 750 (s); nmr (CDClS) 6 1.15 (d, 3H, J = 7.0 Hz, methyl), 1.22-2.23 (m, 5H), 2.37-3.22 (m, 4H), 6.98 (s, 4H, aromatic H's); mass spectrum (70 eV) m/e (rel intensity) 172 (43), 106 143 (8), 129 (100), 117 (12), 116 (54), 115 (32), 91 (5), 89 (3), 77 (5), 63 (5), 51 (5). Anal. Calcd for C H ° C, 90.64; H, 9.36 13 16‘ Found: C, 90.74; H, 9.20 U. 3,4-Benzobicyclo[3.3.0]oct-3-ene-6-one (76). Hydrogenation of 3,4-benzotricyclo[3.3.0.02’8]oct-3-ene-6-one (ZR) (292 mg, 1.72 mmol), as described for 67a (vidé supra), gave 256 mg (1.49 mmol, 86.5%) of a clear oil. Analysis of the crude reaction mixture by vpc (5 ft x 0.125 in, 10% FFAP on chromosorb W, AW-DMCS 80/100, 190°, N2 flow rate 30 ml/min) showed two main products in a ratio of 1:3 with retention times of 6.2 and 7.6 minutes, respectively. These two products were separated by preparative thick-layer chroma- tography on neutral alumina using 15% ether in hexane as the eluent to give 48 mg of an unidentified product and 129 mg (0.750 mmol, 43.6%) of 76. The 2,4-dinitrophenylhydrazone of 72 was prepared57 which after crystallization from chloroform-methanol (50:50) gave orange crystals mp 190-192° (lit.52 mp 195°). The infrared spectrum (neat) of 12 showed bands at 3040 (w), 2990 (w), 2995 (m), 2880 (w), 1740 (s), 1490 (m), 1460 (m), 1440 (w), 1405 (w), 1170 (s), 1130 (m), 1050 (w), 1030 (m), 875 (m), 760 (s), 730 cm’1 (m); nmr (CDC13) 6 1.67-2.37 (m, 4H), 2.45-2.92 (m, 2H), 3.00-3.47 (m, 2H), 6.98 (s, 4H, aromatic H's); mass spectrum (70 eV) m/e (rel intensity) 172 (36), 151 (S), 143 (5), 141 (11), 129 (100), 128 (98), 116 (19), 115 (20), 104 (4), 102 (4), 91 (4), 89 (3), 77 (7), 51 (8). 107 V. Anti-7-methy1-3,4-benzobicyclo[3.3.0]oct-3-ene-6-one (ll). Methylation of ketone 76 (129 mg, 0.745 mmol) as described for the preparation of 74 (vidb supra), yielded 110 mg of a pale yellow oil. Preparative thick-layer chromatography on neutral alumina using 10% ether in hexane as the eluent afforded 30 mg of unreacted starting material and 65 mg (0.347 mmol, 67.4%) of 77, a clear oil: ir (neat) 3060 (w), 3010 (w), 2950 (s), 2925 (s), 2860 (m), 1740 (s), 1490 (m), 1460 (m), 1435 (w), 1380 (w), 1175 (m), 1130 (m), 1120 (w), 1005 (m), 875 (m), 790 (m), 765 (m), 740 cm“1 (m); nmr (c0c13) 6 1.20 (d, 3H, J = 7.0 Hz, methyl), 1.65-2.90 (m, 5H), 3.05-3.43 (m, 2H), 7.00 (s, 4H, aromatic H's); mass spectrum (70 eV) is identical to that of 75. ’VD Anal. Calcd for C13H140: C, 83.83; H, 7.58 Found: C, 83.75; H, 7.65 W. Anti-7-methy1-3,4-benzobicyclo[3.3.0]oct-3-ene (78). MN Wolff-Kishner reduction of 17 (50 mg, 0.269 mmol) as described for the preparation of Zia (vide supra), gave a yellowish oil which on analysis by vpc (5 ft x 0.125 in, 3% SE-30 on chromosorb W, 80/100, 125°, N flow rate 30 m1/min) showed one product. Preparative thick- 2 layer chromatography on neutral alumina using pentane as the eluent yielded 34 mg (0.198 mmol, 74%) of 78. The infrared spectrum (neat) showed bands at 3060 (w), 3010 (w), 2940 (s), 2860 (m), 1490 (m), 1460 (m), 1380 (w), several weak bands from 1350-800, 780 (m), 760 (m), 730 cm‘1 (m); nmr (c0c13) 6 1.00 (d, 3H, J = 7.0 Hz, methyl), 1.15-2.48 (m, 6H), 2.65-3.22 (m), 3H), 6.90 (br s, 4H, aromatic H's); mass spectrum (70 eV) m/e (rel intensity) 172 923), 155 (3), 143 (4), 141 (3), 108 129 (100), 115 (13), 91 (4), 89 (2), 77 (4). Anal. Calcd for C13H16: Found: C, 90.58; H, 9.51 C, 90.64; H, 9.36 8 X. syn-7-hydroxy-anti-6-methy1-3,4-benzotricyclo[3.3.0.02’ ]- oct-3-ene (79). To a 25-ml three-necked flask equipped with a condenser, magnetic stirrer, nitrogen gas inlet and serum cap was added 0.4 g of lithium aluminum hydride and 15 m1 of anhydrous ether. The suspension was cooled in an ice-bath and to it was added, dropwise by means of a syringe, 275 mg (1.49 mmol) of ketone 14 in 3 ml of ether. After the mixture was stirred at room temperature for 2 hr, the excess lithium aluminum hydride was destroyed by adding, dropwise, 2 ml of saturated NH4C1 solution. The white precipitate which formed after 0.5 hr stirring was filtered and the filtrate was dried over anhydrous MgSO4. Evaporation of the ether afforded 235 mg (1.26 mmol, 85.0%) of alcohol 19. Analysis of the crude reaction mixture by vpc (5 ft x 0.125 in, 10% FFAP on chromosorb W, AW-DMCS 80/100, 200°, N2 flow rate 30 ml/min) and tlc (neutral alumina, 20% ether in hexane as the eluent) showed only one product which was used for the following reaction without further purification. Purification of a small sample of alcohol 19 by thick-layer chromatography on neutral alumina using 20% ether in hexane as the eluent gave a clear oil which failed to crystallize on standing or in cold pentane: ir (CC14) 3640 (m), 3590 (m), 3440 (br, s), 3060 (w), 3035 (m), 3010 (w), 2950 (s), 2920 (s), 2860 (m), 1480 (s), 1460 (m), 1055 (br, s), 1 several weak bands from 875-780, 760 (s), 745 cm- (s). The bands at 109 3640 and 3590 cm.1 did not change in relative intensity as a function of the concentration of 12 in CC14, whereas the intensity of the band at 3440 cm.1 decreased drastically with a decrease in concentration of 12; nmr (CDC13) 6 1.12 (d, 3H, J = 7.0 Hz, methyl), 1.55-3.08 (m, 5H), 3.22 (d, 1H, J = 5.0 Hz), 3.57 (d, 1H, J = 5.0 Hz), 6.75-7.20 (m, 4H aromatic H's); mass spectrum (70 eV) m/e (rel intensity) 186 (7), 172 (12), 168 (22), 141 (16), 128 (100), 115 (30),104 (10), 91 (6), 89 (5), 76 (15), 46 (15). Anal. Calcd for C H 0: C, 83.83; H, 7.58 13 14 Found: C, 83.81; H, 7.62 Y. Anti-6-methy1-3,4-benzobicyclo[3.2.1]oct-3-ene-7-one (80). mm Hydrogenation of 19 (100 mg, 0.538 mmol), as described for 612 (vidé supra), gave a clear liquid which on analysis by vpc (5 ft x 0.125 in, 10% FFAP on chromosorb W, AW-DMCS 80/100, 200, N2 flow rate 30 ml/min) showed no starting material but showed the formation of two products. The mixture was not analyzed further. The crude hydrogenation product was dissolved in 5 ml of acetone and placed in a 25-ml round-bottomed flask equipped with a magnetic stirrer. The stirred solution was cooled in an ice-water bath to 10°, and Jones' reagent (2.67 M) was added dropwise until the orange color of the reagent persisted for about 15 minutes. The excess oxidizing agent was destroyed by the addition of 0.5 ml of isopropyl alcohol. The mixture was decanted into a separatory funnel, the residual green salts were washed with 5 m1 of acetone and the washing added to the main acetone solution. The acetone solution was diluted with 100 m1 of water and extracted with ether. The ether extracts were combined 110 and washed successively with saturated NaHCOs, water, saturated NaCl solution and dried (MgSO4). Evaporation of the ether yielded 75 mg (0.40 mmol, 75%) of a light yellow oil. Analysis of the crude product by vpc (5 ft x 0.125 in, 10% FFAP on chromosorb W, AW-DMCS 80/100, 180°, N2 flow rate 30 ml/min) and tlc (neutral alumina, 10% ether in pentane as the eluent) showed two products in a 3:1 ratio with vpc retention times of 6.5 and 7.5 minutes respectively. Preparative thick-layer chromatography on neutral alumina using 10% ether in pentane as the eluent yielded 15 mg (0.081 mmol, 15%) of 75, identified by its nmr spectrum, and 50 mg (0.269 mmol, 50%) of 83: ir (neat) 3050 (w), 3005 (w), 2945 (s), 2860 (m), 1740 (s), 1485 (m), 1470 (m), 1455 (m), 1430 (w), 1370 (w), several medium intensity bands from 1300-850, 760 (s), 745 (s), 720 cm'1 (m); nmr (CDC13) 6 1.17 (d, 3H, J = 7.0 Hz, methyl), 1.82-2.82 (m, 4H), 2.87-3.08 (m, 3H), 6.97 (s, 4H, aromatic H's); mass spectrum (70 eV) m/e (rel intensity) 186 (55), 171 (8.0), 157 (11), 143 (12), 129 (100), 115 (27), 102 (4), 91 (6), 89 (3), 77 (8). Anal.Ca1cd for C H 0: C, 83.83; H, 7.58 13 14 Found: C, 83.90; H, 7.52 Z. Anti-6-methy1-3,4-benzobicyclo[3.2.0]oct-3-ene (132). Wolff-Kishner reduction of 82 (45 mg, 0.242 mmol), as described for the preparation of 132 (vide supra), gave a pale yellow liquid which on analysis by vpc (5 ft x 0.125 in, 3% SE-30 on chromosorb W, 80/100, 125°, N2 flow rate 30 m1/min) showed two products in approximately 2:3 ratio with retention times of 8.0 and 8.7 minutes, respectively. Comparison of retention times on 3% SE-30, 10% FFAP, 5% TCEP and 20% DEGS columns (5 ft x 0.125 in columns on chromosorb W, 80/100, 125°, 111 N2 flow rate 30 ml/min) showed that the product with a retention time of 8.0 minutes had the same retention times as the product obtained from the catalytic hydrogenation of photoproduct 673, and the product with a retention time of 8.7 minutes had the same retention times as the product obtained from catalytic hydrogenation of photoproduct 612. The two components were separated by preparative vpc (10 ft x 0.25 in, 10% FFAP on chromosorb W, 30/60, 140°, He flow rate 15 ml/min). The minor product (with a retention time of 8.0 minutes) had an nmr spectrum identical to that of 31, obtained from the hydrogenation of photoproduct 67:, but lacked the small doublet at 6 0.88 and the multiplet at 6 6.95-7.02. The major product (with a retention time of 8.7 minutes) had an ir spectrum identical to the product obtained from the hydro- genation of photoproduct 672. The nmr spectrum was also identical to that of the dihydro photoproduct 612, but lacked the small doublet at 6 1.17 and the multiplet at 6 6.95-7.10. AA'. Hydrogenation of Syn-6-methyl-3,4-benzotricyclo[3.3.0.02’8]- oct—3-ene (675). W Hydrogenation of 612 (91 mg, 0.535 mmol), as described for 672 (vide supra), gave 80.5 mg (0.468 mmol, 87.5%) of a clear liquid. Analysis of the crude reaction mixture by vpc (5 ft x 0.125 in, 3% SE-30 on chromosorb W, 80/100, 125°, N2 flow rate 30 ml/min) showed only one product with a retention time of 8.0 minutes, 81. The infrared spectrum (neat) showed bands at 3060 (w), 2010 (w), 2930 (s), 2860 (m), 1490 (m), 1460 (m), 1375 (w), several weak intensity bands from 1350 to 810, 760 (m), 730 (m), 720 cm"1 (w); nmr (coc13) 6 0.63 (d, 3H, J = 7.0 Hz, methyl), 1.77 (br s, 2H), 1.90-3.30 (m, 7H), 112 6.65-7.07 (s, 4H, aromatic H's). The nmr spectrum also showed a small doublet at 6 0.88, J = 7.0 Hz and a multiplet at 6 6.95-7.02 which probably correspond to a small amount (approximately 15%) of an isomer of 81. Mass spectrum (70 eV) is identical to that of 732. Anal. Calcd for C H C, 90.64; H, 9.36 13 16‘ Found: c, 90.68; H, 9.40 APPENDIX 113 V V 3:1...- - "ve‘ . —_.-——...——~ co - Figure 1. NMR Spectra of 4,S-Epoxy-Z,6,6-trimethy1-2-cycloheptenone (26) [bottom] and 7,7-Dideuterio-4,5-epoxy-2,6,6-trimethy1-2-cyclo- hgptenone (26-d2) [top]. 114 .fiwxmv opmocoxonumuxxommumé T390: mo 5530mm mzz .N 093w: 6.. o." 3" 3. E 2.... 66 o... 2. a... 1 ‘ ‘ ‘ S ‘ 4 — ‘ ‘ q C — q ‘4 11 M q ‘ q ‘ 1.‘ d ‘ J 1‘ — ‘ 4 q ‘ — ‘ 4H4 1" ‘ [m M ‘ 1‘ _ 1 . . fl 4 _ — _ q _—° F b P f . P P _ D k D b 1— b b b I h b b D L 115 g .flmwmv ocoflwaucmmaosmnuouv.Huonvxcfiuuv.H-Hx:pos-m-th mo anhuoomm mzz .m ousmwm en a. .:L?: ad ed at d 1 id I. —1 1‘ 4 1 ”III .1 M «‘1 M1‘1 ‘I‘dlq _ 4‘ d I a UH ‘d 1_ 1‘ 1‘ d ‘ .1 __ “fl 1 ‘ 7 ‘1 ‘1 ‘M‘ 1‘ 1. d ‘ a n _ _ Q; 1 g 1 I um: 47: 21 a ow. emu . 8» J. Ha I n P P n — F n D D — D b k n F b n ‘D _ D n b D D P F D h D D D D — D n b n _ D ANNE ocofimcunmmcocmfioue.793833616.Hugfiosdunfirv. mo Sfiuoomm «22 .6 6.3m: 116 6 6.. 6.6 6.» 6.6 CF 2.: 6.6 6.6 6.5 6.6 I ‘ d ‘ 1 _r d 4 1 I .— 1 1 1 1 — I 4 ‘ 1 H ‘ ‘ d q — q ‘ ‘ ‘ 1F 1 1 I q is d ‘1 ‘ ‘ 1‘ J‘ H l‘ 1 I 1] 4 J I 1 II I I ._1 I 1 J 4‘ d1 1 4 q I — I 1‘ 1 J I T 1 J 4 - I I I l I‘ 1 I 1‘1! 1-‘ 4 1 [III‘ I1 I 1 I 1 l1 1- I 2.)). . ammov mcgmfizmwcocmnuouv.Tchuxcwhpmuéfi.mJugfioEnmusmm mo 65.30QO mzz .m madman 6 6. 6... 6.6 66 .3 5.: 66 6.6 6.5 6.6 1‘1‘111 1111 4111 1111 1111 1 1 11 1111 11 _ a — _ _ _ _ _ 4... .._.....«..._........._ .1.._........_ . ‘- 4.. 411-11-? \l’ 7 l l _ _ . _ m m . u: _ _ 8... m. ......_ ...4.....w. Ib~nblhfi h_h h-_-h-—hh 118 €38 . n33 mcoumupoomw.mo.o.m.2oHoxUEuouconé£ufixnuofinnusmm. mo 65.65on «:2 .0 0.53m. 6.. 6.6 6.6 63 E 2.: 6.6 1 1 1 1 T W A— 1 1 1 1 — 1 I 1 11 H 4'1 1 1 w 1 I] 1 1 1 I 1 1 I - 1 1 1 1 1 III 1 1 1 _ 1 1| 1 1 1 1 1 1 I . .. 119 2).); Annoy m:m-m-uuofiw.mo.o.m.2oHuxowhuoNaopuv.71505-0..sz mo guoomm mzz .n 933... 6.. 64. 3” 6.. E at 6.... 6.6 6.. 6.6 — I 11 1 —\ 1 1 1 1 I F I I 1 I H 1 I I [I —r 1 I q I .— 1 1 1 1 — 1 1 1 I 1— J 1 I 1 1 1 111 1 _ I 11 1 I 1 1 1 1 1 — 1 I 1 d 1 [T 1 — 1 1 1 1 fl I — 1 1 1 1 T 1 1 1 - 1‘ .‘Illt'lll'.n ‘ g 933 66322666666564.78626534.m.m.725266-326. mo 6366on «22 .6 6.8mm. 6.6 .3 2.: 6.6 6.6 I 1 II 1 I — 1W 1 1 120 1‘ TI‘ 1 121 .1“wa oco-m-pooflm.mo.6.m.23656362664.?626066-326. mo 568% 6:2 .6 936: 122 .-..Z_‘+ ."I '."..".".'T.."P\r" “ " WW3 4.0 3.0 2.0 1.0 PPM (6) Figure 10.270 MHz NMR Spectrum of Anti- 6— “methyl —3, 4- b—enzotricyclo- [3.3.0. o2 8]oct- 3- —ene (67a). 123 ‘l Irradiated at .., W'MW 5| I J J...) . ' 7 .41} u..‘.q-.. . Irradiated at ~2.7ppm A JUUUF"? 5% ' ‘ Irradiated at ~3.1ppm 'E f i. . I Irradiated at . 2L, j. A A , 1v *. A A 1 L A A g l A a A A ' L A A A l A A L L4 A L 1 A I 3.0 2.0 1.0 PPM (5) Figure 1%.8 Double Irradiation of Anti-6-methy1-3,4-benzotricyclo- [3.3.0.0 ’ ]oct-S-ene (21%). 124 Irradiated at ' $1.0ppn Irradiated at m1.3ppfl Irradiated at «1.5ppfl Irradiated at ~1.80vu Irradiated at ~2.0ppn ALIJLAI IJLALJL L4 [HA41A4AAI 3.0 2.0 1.0 PPM (6) Figure 11 (continued). Double Irradiation of Anti-6-methy1-3,4-benzo- tricyclo[3.3.0.02’8]oct-3-ene (33%). 125 88 .nmov ocoumupoomw.~o.o.m.mHoHu»UfiHuoN:onuv.Mumcoflxcuozun mo ashuoomm «22 .NH whomHm 0 ca 0." . 0.9 06 .6. It 0.. 0.0 0.5 0.. 1 1 1 1 1 1 1 1 1 1 1 1 1 1 — 1 1 1 1 F 1 1 1 1 —V J 1 1 1 A 1 1 1 1 _ 1 1 1 6 6 6 q _ id 6 - J1 m 6 6 A J Alhd 6 - 1 6 W 1 — 6 d 6 6 — .1 6 6 1 w I . 6 6 - 4 d . 6 m d . d . a . . a n. M ,. 6 . . . _ i . _ 6 _ 6 ., _ 6 6. . - , . .l--- t... “ J D a . . . . w 6 m m T . _ . a h . .H _ a . . .. _ . ., . I . u u . ’ . .. . . — 6. . . . . n . . . . . . 6 6 .. I l :1... J. I ma .. . . ’7 6 w 6 As; 86 o oo- Su 80 00' 03 p — 6 P p b _ b FL P b w L _ 6 6 I . — p P r n _ 6 r b r b L ~ b . h — F P 6 h p n n 6 b b P _ I D P P _ F P — PL h Ll — P D h I — P D b P b L L n L b F b r“ n I h h H b n b 126 . $665 ocwumnuoomw.mo.o.m.3oHoonuoNconuv.mugfimauolHAKOHPEIoISmm mo 65.33% 6622 .2 0.3me a n o n 3 2 6 2.: o m 3 o 6 36 1 1 H 1 1 1 1 — 1 1 1 1 — 1 1 1 16 _ 61 6‘ 1 6 1 I 16 _r 41 6 1 6 1 16 _ 61 1 1 16 _ 61 1 1 1 6 — 6 1 6 6 6 6 6 < 6 — 6 6 6 . a — — 127 at? .6656 m:o-n-o:m-m-uoomw.mo.o.m.66oHUxu6Hpoacon-6.m-6666ms-o-6uzq mo asyuommm 622 .66 065666 o. 6 ad ad 0.. _ 1 1 1 1 — 1 1 1 1 — 1 1 1 1 — 1 1 1 1 ‘ 1 1 1 11 1 1 1 by F p h ~P> —IDbDH\PDPI—Ibbh_DID.‘*LPD~Db-P~bhhl_bbL 128 88 Amt mcofiumcoumuuuosoé.23983868664.magnuosnonwutw mo Eauuomw 6622 .3 963m: E at q. 3 3. 9.6 1 1 1 — 1 .. 1 1 1 q 1 it 1 1 _ 1 1 1 1 fl 1 1: l1 1 1 + 1 _ 1 111 1 1 1 1 1 1 1g 1 1 1 1 1 14 1 1 1 q [11 129 .hwmwv o:m-m-uoo_o.m.mHOHUxufiaoNcmn-6.6-6»;6os-o-6626 mo asauuogm «:2 .96 ou5666 36. 3” 3 E 3.... 9.6 1 1 1H 1 1 1 1 H1 1 1 1—1: 1 1 1 — 1 1 1 1H1 11 1 _r 1 1 1 1 — 1 111‘l’1‘ 1| 1 1 1 _\ 1 1 1 1 1 1 1 11‘ 1 _ 1 1 1 1 _ 1 1 1 1 1 rrrlr D _ P rib L r.r>|> rQLL FL'IL r b b b “'PIF b PL lb + FPL H bf pit [ 7 up r ‘ b b up p- + b y. u- HF PIPLhIanPIPLL rlIInrh’lPL DIIPIPILVHFIF S . .AOBV oco-o-e=o-m-poo_o.m.mHoHoxofiaoucom-v.m mo esppuomm mzz .5H 093m“; o.~. , a.» 3 Sr at a.» o... o 5 ed 1 1 1 _ 1 ‘1 1 1 — J I q [1‘ _ 11 1 1 1 1 1 1 1 — 1 1 1 1 H 141 1 1 1 — 1 1 1 1 — 1 1 1‘ 1 fl 1 1 - 1 1| 1 1 1‘ 1 11 1 1 1 1 1 J1 1 11 11 - 1 1 1 1 1 1 1 1 1‘ _‘ 1 til E r b . I p b’ It ’ I I: I} L t i, 1 11 { ‘ 130 131 S .mnnv m:o-ouo:o-mupuo_o.m.mHoHuxufinoucmnuv.muHxnuoE-nnw«:w mo sapwoomm mZZ .3 0.8m: . 05 06 0.1 2. SE; 0.0 0.0 0.5 0d 1 ‘1 ‘1 _‘| 1 1 1 1 —H 1 1 —‘r 1 111 < 1 1 — _ 1 u 1 < _ 4 1‘ 1 4 d1 1 l4 1 i ‘1 + q q 1 1 1 1. 1 d o — _ q 1 H . ‘ .. _ _ . . . . . m _. m _ m _ m n v _ . z n g m. A... __ . u: o o8 N_ P - F L L {r ? & D P L h b L ’ P L b *L’ I F b hI L t. h 3‘ . I _ M h F ilk {r .— P p — P b P h F P h h — P h L 132 ”=22 .mH mgamfim 133 134 S . 88 unannécoumuuoo:.N.239333534.?acumsuouwurq mo 52.":on ~52 .Hm chum: an ad 3 2. at a... ed .2 ed 4 1 . ‘I—~“”' ~ .~. g...- .0.-.- 1 —- 135 V V V I V V 1 "l v v j 7 Ti 7 rflt'rvlvrv v1 1 YVT vvvvvv Om T—v—V v I v I 7 f7 'ffiifir] v v v v V v v *7 v I v v f v v v ifi v r V V l V V V T] K'vv v vvr'vl NMR Spectra of Anti-o-methyl-s,4-benzobicyclo[3.2.1]oct- Top: material obtained from hydrogenation of photo- Figure 22. Bottom: material obtained from Wolf-Kishner reduction 3-ene (73b). product a. of ketonemgg. 136 T V r Lia V V r V V V W I V V f V V V V f V V V V V V r V V V V V v v T V v *7 v v v’ ‘7 v 7 iv v v r v. v *v V’fiv ff‘v V irva 1 ! v v v 'fi ! v v v v E v fifv r I v .1 .L IN N. 4W1 1H1” v Ll A A AA A A L A A l A A A A A L A A A J A A AA A AAA L AA AA A A A A A A A l A A AA A A A A A A A A A l A A A A A AA A A I A A A A l A A L I A A A A I A A A A l A A A A j A A A A P _A I0 1.. 00 I. "a "i u u 20 I. 0 Figure 23. NMR Spectra of Syn-o-methyl-S,4-benzobicyclo[3.2.1]oct- 3-ene ( ). Top: material Obtained from hydrogenation of photo- product . Bottom: material Obtained from Wolf-Kishner reduction of ketone . BI BLIOGRAPHY 10. BIBLIOGRAPHY (a) D. Swern, Chem. Rev., 45, 1 (1949); (b) D. Swern, Org. Reactions, Z, 378 (1953); (c) D. Swern in D. Swern, ed., "Organic Peroxides," Vol. 2, Wiley-Interscience, New York, N.Y., 1971, Chapter 5; (d) H. 0. House, "Modern Synthetic Reactions," 2nd ed., W. A. Benjamin, Menlo Park, Calif., 1972, Chapter 6; (e) M. S. Malinovskii, "Epoxides and Their Derivatives," Daniel Davey, New York, N.Y., 1965. (a) Y. R. Naves, 0. Schwarzkopf and A. D. Lewis, HeZv. Chim. Acta, 39, 880 (1947); (b) P. Heinanen, Suomen Kemistilehti, llfi, 2 (1938). [C. A. g3, 2903 (1938)]; (c) M. Okada and Y. Saito, Steroids, g, 645 (1965). (a) H. Sturzinger and P. Karrer, HeZv. Chim. Acta, 22, 1829 (1946); (b) Y. Suhara and T. Minami, Bull. Chem. Soc. Jdp., 32, 1968 (1966); (c) S. Akagi and K. Tsuda, Chem. Pharm. Bull. (Tokyo), 9, 464 (1961); (d) L. H. Knox, J. A. Zderic, J. P. Ruelas, C. Djerassi—and H. J. Ringold, J. Amer. Chem. 300., 82, 1230 (1960). (a) P. A. Diassi, S. D. Levine and R. M. Palmere, J. Med. Chem., l_0_, 551 (1967); (b) E. Farkas and J. M. Owen, J. Med. Chem., 2, 510 (1966). P. Schuster, Mbnatsh. Chem., 29, 1624 (1968); D. J. Bertelli and T. G. Andrews, Jr., J. Amer. Chem. 800., 91, 5280 (1969); G. Payette and C. Sandorfy, J. Chem. Phys., 39, 168 (1959). (a) R. B. Parker and N. S. Isaac, Chem. Rev., 52, 737 (1959); (b) S. Winstein and R. B. Henderson in R. C. Elderfield, ed., "Heterocyclic Compounds," Vol. 1, Wiley-Interscience, New York, N.Y., 1950 3 pp 1'60. W. K. Anderson and T. Veysoglu, J. Org. Chem., 38, 2267 (1973). G. B. Payne and P. H. Williams, J. Org. Chem., 29, 651 (1961); G. B. Payne, P. H. Deming and P. H. Williams, J. Org. Chem., 29, 659 (1961); G. B. Payne, Tetrahedron, 18, 763 (1962); Y. Ogata and Y. Sawaki, Tetrahedron, 29, 2065 (1964). A. Rosowsky in A. Weinberger, ed., "Heterocyclic Compounds with Three and Four Membered Rings," Part 1, Wiley-Interscience, New York, N.Y., 1964, Chapter I. H. Laurent, G. Schulz and R. Wiechert, Chem. Ber., 102, 2570 (1969). 137 11. 12. 13. 14. 15. 16. 17. 13. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 138 Z. Chabudzinski and M. Skwarek, Rocz. Chem., 93, 1907 (1971) [C. A. 29, 125448q]. E. Weitz and A. Scheffer, Chem. Ber., 54B, 2327 (1921). K. B. Sharpless and R. C. Michaelson, J. Amer. Chem. Soc., 93, 6136 (1973) and references cited therein. (a) F. R. Mayo, Accounts of Chem. Res... 3, 193 (1968); (b) W. G. Lloyd in E. S. Huyser, ed., "Methods in Free Radical Chemistry," Vol. 3, Marcel Dekker, New York, N.Y., 1973, Chapter 1. G. H. Twigg, Chem. Eng. Sci., 3, Suppl., 5 (1954); F. R. Mayo, J. Amer. Chem. Soc., 99, 2497 (1958); C. Walling, "Free Radicals in Solution," Wiley-Interscience, New York, N.Y., 1957, p. 439; W. G. Lloyd, Reference 14b, p. 49. A. Padwa and L. Brodsky, Tetrahedron Lett., 1045 (1973). This example appeared while the work reported in this thesis was in progress. M. S. Malinovskii, "Epoxides and Their Derivatives," Daniel Davey, New York, N.Y., 1965; H. House, "Modern Synthetic Reactions," 2nd ed., W. A. Benjamin, Menlo Park, Calif., 1972, p. 318. E. L. Eliel and D. W. Delmonte, J. Amer. Chem. Soc., 99, 1744 (1958). B. Tchoubar, Cbmpt. Rend., 214, 117 (1942); J. E. Horan and R. W. Schiessler, Org. syn., 93, S3 (1961). Tiffeneau and Kuriaki, Cbmpt. Rend., 209, 465 (1939). C. Dermer and A. M. Durr, J. Amer. Chem. Soc., 19, 912 (1954). 0. House and R. L. Wasson, J. Amer. Chem. Sbc., 29, 4394 (1956). S. Malinovskii, Reference 1e, p. 338. 0. House, D. T. Reif and R. L. Wasson, J. Amer. Chem. Soc., 7_9_, 0. H. A. Rosowsky, Reference 9, p. 254 and references cited therein. M. H. l 88, 2490 (1957). R. T. Conley, "Infrared Spectroscopy," Allyn and Bacon,.Inc., Boston, Mass., 1966, p. 130. H. Hart and R. M. Lange, J. Org. Chem., 33, 3776 (1966). E. J. Corey and H. J. Burke, J. Amer. Chem. Soc., Z9, 174 (1956). A. A. Bothner-By and E. Moser, J. Amer. Chem. Sbc., 99, 2347 (1968). P. Karrer and H. Sturzinger, HeZv. Chim. Acta, 99, 1829 (1946). S. Akagi and K. Tsuda, Chem. Pharm. Bull. (Tokyo), 9, 464 (1961). 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 139 Y. Suhara and T. Minami, Bull. Chem. Soc. Jap., 99, 1968 (1966). H. Hart, M. Verna and 1. Wang, J. Org. Chem., 99, 3418 (1973). H. Hart, P. M. Collins and A. J. Waring, J. Amer. Chem. 800., 99, 1005 (1966). Assignment of structure 9g_was done in conjunction with Dr. Irene Huang of these laboratories. R. T. Conley, Reference 26, p. 157. T. Matsuura and K. Ogura, J. Amer. Chem. 300., 99, 3850 (1967). Purchased from International Chemical Corp., New York, N.Y. H. E. Zimmerman, R. W. Binkley, R. S. Givens, and M. A. Sherwin, J. Amer. Chem. Soc., 99, 3932 (1967). S. S. Hixson, P. S. Mariano, and H. E. Zimmerman, Chem. Rev., 13, S31 (1973). H. E. Zimmerman and A. C. Pratt, J. Amer. Chem. Soc., 99, 6267 (1970). E. Ciganek, J. Amer. Chem. Sbc., 99, 2882 (1966). H. E. Zimmerman and P. Mariano, J. Amer. Chem. 300., 99, 1718 (1969). H. Hart and R. K. Murray, J. Amer. Chem. 500., 91, 2183 (1969). H. Hart and G. M. Love, J. Amer. Chem. 500., g, 4592 (1973). I. Murata and Y. Sugihara, Tetrahedron Lett., 3789 (1972). R. Grenwald, M. Chaykovsky and E. J. Corey, J. Org. Chem., 99, 1128 (1963). Prepared by the method of R. C. Cookson and N. S. Wariyar, J. Chem. Sbc., 2302 (1956). Prepared by the method of R. S. Givens, W. F. Oettle, R. L. Coffin, and R. G. Carlson, J. Amer. Chem. 500., 99, 3957 (1971); J. Ipaktschi, Tetrahedron Lett., 215 (1969). I thank Mr. Ravinder Sdnu (University of Chicago) for running the spectrum and Professor Philip Eaton for making his facilities available. M. Barfield, J. Chem. Phys., 99, 3825 (1964). S. Julia, C. Huynh and J. Olivie, Bull. Soc. Chim. Fr., 147 (1966). 53. S4. 55. 56. 57. 140 For a discussion of the Schenck mechanism, see: A. A. Lamola in P. A. Leermakers and A. Weissberger, editors, "Technique of Organic Chemistry," Interscience, New York, N.Y., 1969, p. 80 and references cited therein. R. R. Sauers, W. Schinski and M. M. Mason, Tetrahedron Lett., 4763 (1967). P. J. Kr0pp, J. Amer. Chem. Soc., 9;, 5783 (1969). K. Kitahonoki and Y. Takano, Tetrahedron Lett., 1597 (1963); H. E. Zimmerman, R. S. Givens and R. M. Pagni, J. Amer. Chem. Soc., 99, 6096 (1968). R. L. Shriner, R. C. Fuson and D. Y. Curtin, "The Systematic Identification of Organic Compounds," 5th Ed., John Wiley and Sons, New York, N.Y., p. 147.