LIBRARY Michigan State IJrflanxdty ~q V This is to certify that the thesis entitled . . ~ 1. Synthetic Uses of High Surface Sodium and Pota551um 1n Alkylation and Reduction of Ketones and Nitriles 2. Synthesis and Photochemistry of substituted 2-cyclooctenones 3. Synthesis and Photochemistry of 1,5-Dimethy-4-methy1ene- bi cyclo(3.3 .0)octa- presented by dienes Bing-Lin Chen has been accepted towards fulfillment of the requirements for Ph .0. degree in Orqanic Chemistry Jigsaw 41$“ 7 1 Major professor Date Jan- 24, 1928 _. 0-7639 .l.fll' Iivly :lril}‘ V J! Ill: 4) L9 6‘12; PART I SYNTHETIC USES OF HIGH SURFACE SODIUM AND POTASSIUM IN ALKYLATION AND REDUCTION OF KETONES AND NITRILES PART II SYNTHESIS AND PIIOI‘OGIEMISTRY OF SUBSTITUTED 2-CYCIDOCI‘ENONES PART III SYNTHESIS AND PHOTOGiFi‘IISTRY OF 1,5-DIM3’I’IiYL-4-ME'ITIYLENEBICYCLOB.3.0]OCI‘ADIENES By Bing-Lin Chen A DISSERTATION submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1978 ABSTRACT PART I SYNTHETIC USES OF HIGH SURFACE SODIUM AND POTASSIUM IN ALKYLATION AND REDUCTION OF KETONES AND NITRILES PART II i SYNTHESIS AND PHOTOCHFMISTRY 0F SUBSTITUTED 2—CYCLOOCTENONES PART I I I SYNTHESIS AND PllOTOCl [EMISTRY OF 1 , S-DIl~1E'I'llYL-4 -I‘-lE'I'l-IYLENEBICYCLO[3 . 3 . 0]OCTADIENES By Bing -Lin Chen In part I of this thesis, synthetic uses of high surface sodium (H55) and potassium (USP) , particularly for the alkylation and reduction of cyclic ketones and nitriles, have been explored. Alkylations using HSS (or IISP) have been performed with 2-methy1cyclohexanone, cyclohexanone, 4-t-butylcyclohexanone, cycloheptanone and cyclooctanone. The results show that monoalkylation usually predominates; where regioselectivity is involved, the more highly substituted products are obtained. The main competing reactions to alkylation are reduction to alcohols and to pinacols. For example, 2—methylcyclohexanone was alkylated with IISS Bing-Lin Chen and allyl brmnide in hexane to give a 68% yield of Z—allyl—Z-methyl— cyclohexanone along with some starting ketone and reduction product. HSS alkylation was also carried out on a,s-unsaturated ketones. An interesting result was obtained for the methylation of isophorone with 20% HSS-C and methyl iodide in hexane. In this case reduction competes very well with alkylation, and a substantial amount of 1,6- diketone (i.e. 1,1',3,3,3',3'-hexamethylbicyclohexyl-5,S'-dione, 38% yield) resulting from the reductive coupling at B-position of isophorone was obtained. The only alkylated product was 2-methylisophorone (32% yield). The methylation of phenylacetonitrile with H58 and methyl iodide gave a 74% yield of 2-methy1phenylacetonitrile with no reduction product. High surface sodium has aslo been used to study the reduction of cyclohexanone and 2-methy1cyclohexanone. The reduction of 2- methylcyclohexanone with 2 molar equivalents of high surface sodium on graphite in THF afforded a very good yield of 2-methylcyclohexanol (83%). The more stable alcohol (trans/cis = 85/15) predominated. A mechanism involving enolate anions is proposed to account for the formation of alkylation products in ”SS alkylation of cyclic ketones. The intermediate leading to reduction products is probably a radical anion. The synthesis and photochemistry of Z-methyl-Z-cyclooctenone I, 3-methy1—2-cyclooctenone II, and 2,3-dimethyl-2-cyclooctenone III is Bing-Lin Chen described in part II of this thesis. A method was found to prepare Zymethylcyclooctanone in quantitative yield, involving the methylation of cyclooctanone via its dimethylhydrazone. Bromination-dehydro- bromination of Z-methylcyclooctanone gave Z-methyl-Z-cyclooctenone in good yield. S-Methyl-Z-cyclooctenone and 2,3-dimethyl-2-cyclooctenones were prepared by oxidizing the tertiary allylic alcohols generated by the 1,2—addition of methyllithium to 2-cyclooctenone and Z-methyl- Z-cyclooctenone respectively. Irradiation of 2-methyl-2-cyclooctenone and 3-methyl-Z- cyclooctenone in methanol resulted in the fonnation of adducts cis-Z- methyl-3-methoxycyclooctanone IX and 3-methyl—3-methoxycyclooctanone X respectively, whereas methanol did not add photochemically to 2,3— dimethyl-Z-cyclooctenone. The base-catalyzed Michael addition of methanol to I gave an equilibrium mixture of I, trans-Z-methyl-S- methoxycyclooctanone XI, and TX in 45 : 35 : 20 ratio. Moreover, the base-catalyzed exchange reaction of IX led to the same product mixture. These results are rationalized by a photoisomerization of IN (or II) to its trans isomer which then thermally adds methanol in a regio— and stereospecific syn manner. The reason that no methanol adduct was found in the photolysis of III is probably because other energy dissipation processes compete efficiently with the double bond iSOmerization. The synthesis and attempted photochemistry of 1,5-dimethyl-4o methylenebicyclo[3.3.0]octa-2,6-diene, l,S-dimethyl-4-methylenebicyclo- Bing-Lin Chen [3.3.0]octa-2,7-diene, and 1,S-dimethyl-3,7-diphenyl-4-methylenebicyclo- [3.3.0]octa-2,6-diene is described in part III of this thesis. However, these compounds on irradiation did not give detectable amounts of di-n- methane rearrangement products or the products corresponding to intra- molecular cycloaddition of the two endocyclic double bonds. To my parents and my wife, Yu-I (Yolande) ACKNOWLEDGEMENTS I wish to express my sincere appreciation to Professor Harold Hart for his guidance, support, and encouragement throughout the course of this study. Appreciation is also extended to Michigan State University, National Science Foundation, and National Institute of Health for financial support in the form of teaching and research assistantships. Finally, the typing of the entire manuscript by my wife is grateful acknowledged. TABLE OF CONTENTS PART I SYNTHETIC USES OF HIGH SURFACE SODIUM AND POTASSIUM IN ALKYLATION AND REDUCTION OF KETONES AND NITRILES INTRODUCTION.... ................................................. RESULTS AND DISCUSSION ........................................... 1. The Alkylation of Cyclic Ketones......... ................ (A) Alkylation of Z-Methylcyclohexanone..................... (B) lbthylation of Cyclohexanone .......... ....... ........... (C) Methylation of 4-t-Buty1cyclohexanone................... (ID th-ylation of Cycloheptanone and Cyclooctanone... ...... (E) Mechanism................... ............. ....... ........ 2. bbthylation of Isophorene ....... . ....................... . 3. Methylation of Phenylacetonitrile ............ ... ......... 4. Reduction of Cyclic Ketones ...... . ......... ... ........... S. Smnnary ....... . ..... ..... .............. .................. EXPERIMENTAL ........... .... ...................................... 1. General Procedures..... ................ ......... ........ . 2. Preparation of High Surface Sodium ...... ..... ....... ..... 3. hbthylation of Cyclohexanone with HSS—C and Methyl Iodide... .................. . ........ ..................... 4. Methylation of Z-Methylcyclohexanone 9..... ............ .. 5. Methylation of Enolate Mixture IN and 15 in the presence and Absence of Charcoal. ... .......................... ... iv Page 2 15 15 15 21 25 26 27 32 36 37 4o 41 41 42 43 44 45 TABLE OF (DNI‘ENTS (continued) ‘0 oo \1 0‘ C 10. 11. 12. 13. 14. 15. Page AllYlation 0f Z'MethYICYCIOhexanone go o o o o o o o t o o o o o u o n o o a 46 Preparation of 2 , Z ,6,6-Tetradeuterocyclohexanone 59 . . . . . . 46 bbthylationof'ZWOOQDOOCIDIDO ...... OO’CCOOOOOCOOOOQOUIIOCOU 47 Methylation of 4-t -Butylcyclohexanone a . . . . . . . . . . . . . . . . . 48 Methylation of Cycloheptanone finnw ....... 49 Methylation of Cyclooctanone 3g 49 Methylation of Isophorone «SAZ. ........... . . . . . ............. 50 Methylation of Phenylacetonitrile @ ............ . ........ 51 Reduction of Cyclohexanone u with HSS ............. . ..... 51 Reduction of 2-1‘ Iethylcyclohexanone 2 with 1188 ..... . . . . . . . 51 PART II SYNTHESIS AND PIIOI'OGIENISTRY OF SUBSTITUTED Z-CYCLOOCI'ENONES INTRODUCTION ................................................. g. . . . 53 RESULTS AND DISCUSSION. . ......................................... 65 1 . Synthesis of 2-Methyl—2-cyclooctenone %. . . . . ....... . . . . . 65 2 . Synthesis of 3-Methyl- and 2 ,3-Dimethyl-2 -cyclooctenones ngde. ........... . ..................... ........ 73 3 . Photochemistry of 2 -Methyl-2-cyclooctenone 33 in Methanol 76 4 . Photochemistry of 3-Methyl- and 2 ,-3 -Dimethyl— —-2 ~cyclooctenones 38am ,lNQ ....... .. ........................ 80 5 . Summary. .................... . . . . . .................. . ..... 82 EXPERIl-ENI‘AL ................... ..................... 83 l . Gas Chromatography ............................. . . . . ...... 83 2 . Preparation of l -l=lethylcyclooctene ,8]; ........... . ........ 83 V TABLE OF CONTENTS (continued) 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. Bromination of 8% ...... . ............. . ....... ...... ...... Allylic Oxidation of 82....... ........... . ........... .... Acid-catalyzed Rearrangement of l-bbthylcyclooctene Oxide§fé ...... .0... ...... O ........... 00.00.00... ......... Preparation of Cyclooctanone-N,N-dimethylhydrazone 99.... lbthylation of £9 ..... . .................................. Hydrolysis of 2- -Methylcyclooctanone-N, N- -dimethylhydrazone ooooooooooooooooooooooooooooooooooooooooooooooooooooo Bromination of Z-Methylcyclooctanone 82 .................. Dehydrobromination of 2-Bromo-2—methylcyclooctanone 92... Preparation of 2,3-Dimethylcyclooctanone 24 ......... ..... Bromination of 94............ ......................... ... Dehydrobromination of 2 -Bromo- 2, 3- -dimethylcyclooctanone no.0 00000000 o ...... o oooooooooooooooo ooooooooooovuinuot Preparation of l-Methylcyclooct-2-en—l-ol 27 .......... .... Preparation of l,Z-Dimethylcyclooct-2-en-l-ol 22 ...... ... Oxidation of 21.......... ................................ Oxidation of 22 ................ . ......... . .............. . Photolysis of Z-Methyl-Z-cyclooctenone 23 in Methanol... Europiwn Shift Study of 3-Methoxy-2-methylcyclooctanone 0 Decoupling Experiment of lggg and Eu-complexed 1225...... RIiC}Ia.el Addition of ”ethanOl to 2g. 0 a o n o I o o D ooooooooo o n n o Base-catalyzed Exchange Reaction of lggg.... ............. Photolysis of 3-Methyl-2—cyclooctenone 98 in Methanol.... Europium Shift Study of 3-Methoxy-3-methylcyclooctanone ooooooooooooooooooooooo .000.000000...OOOODIOIDOIOQOOIQ \ vi Page 84 84 85 85 86 86 87 88 88 89 89 90 91 91 92 93 93 94 95 96 96 TABLE OF CONTENTS (continued) 25. SYNTHESIS AND PI-IOI‘OCHEMISTRY OF 1 ,5-DIl«lETllYL-4-METHYLENE-BICYCLO[3.3.0] - Page Photolysis of 2, 3- -Dimethyl- -2- -cyclooctenone 100 in “'iethaltlol 000000000 0"... ....... ..........IOIOOIOOIOIOOOOIO 97 PART III OCTADIENES INTRODUCTION ....... ........ .............. ........ ......... . ...... 99 RESULTS AND DISCUSSION .......................................... 107 1. Synthesis of l ,5- -Dimethyl —4-methylenebicyclo[3. 3. 0]- octa-Z, 6 and 2 ,7 dienes lmg and 159 ....... . ..... . ........ 107 2. Synthesis of 1,S-Dhnethyl-S,7—diphenyl-4-methylenebicyclo- [3.3.0]0Cta'2,6'diene m ooooooooooooooooooooooooo ...-00. 113 3. Photolysis of 1 ,5 -Dnnethyl- -3 ,7 -dipheny1bicyclo[3. 3. 0]octa— 2 ,6 dien- 4- -one 192 ...................................... 115 4. Photolysis of 1,5-Dhnethylbicyclo[3.3.0]octa-2,6-dien-4-one W IIIIIIIIIIIIIIIIIII ' ....... OI. ..... OIQIIIOO IIIIIIIIIII 116 5. Photolysis of 1,S-Dimethylbicyclo[3.3.0]octa-2,7-diene 4 000000 O OOOOOOOOO D OOOOOOOOOOOO U. ....... O 0000000 $000.... 117 6. Photolysis of 138, 159, and lgé ................... . ...... 118 EXPERIlfiflTDAL ..................................................... 120 1. Gas Chromatography ............ . ...... . ....... . ........... 120 2. l,5-Dimethy1-2,4,6,8-tetramethoxycarbonylbicyclo[3.3.0]- OCta-3,7‘dione We 0 o ooooooooooooooooooo h oooooooo u o a o o o o o 120 3. 1,5-Dimethy1bicyclo[3.3.0]octan-3,7—dione 138 ..... . ...... 120 4. 1,S-Dimethylbicyclo[3.3.0]octan-3,7-diol 132 ...... . ...... 121 TABLE OF CONTENTS (continued) Page 5. 1, 5- -Dimethy1bicyclo[3. 3. 0]octa- 2 ,6 and 2 ,7- diene 1mg and 111..................... ........... .. ............... 121 6. Allylic Oxidation of 119 and 111 ....... . ........... . ..... 122 7. 1,5-Dimethyl-4-methylenebicyclo[3.3.0]octa-2,6-diene 111. 124 8. l,5-Dimethyl-4-methy1enebicyclo[3.3.0]octa-2,7-diene 11g. 124 9. 1,5-Dimethy1-3,7-diphenylbicyclo[3.3.0]octan-3,7-diol 112 125 10. l, 5- -Dimethyl- 3, 7- -diphenylbicyclo[3. 3. 0]octa- 2 ,6- diene 160 and 2 ,7 diene 1Q1.. .................................. 125 11. l,5-Dimethyl-3,7-dipheny1bicyclo[3.3.0]octa-2,6-dien-4- one 1Q1 ............. ... ..................... ... ......... . 126 12. 1,5-Dimethyl-3,7-dipheny1-4-methylenebicyclo[3.3.0]octa- 2,6-diene 1g1 ............................ . ...... . ..... ... 126 13. Photolysis of 1Q1........................... ..... . ...... . 127 14. Photolysis of 111 ........... . ..... . .............. ........ 128 15. Photolysis of 111.. ..... ...... .......... ................. 128 16. Photolysis of 1g3 ........ ...... ........... .. ........ ..... 129 17. Photolysis of 118 and 11g ....... . ................ ~........ 129 BIBLIOGRAPHY ..... . ................. . ............... . ............. 132 viii LIST OF TABLES TABLE 1 . Average Film Thickness vs . Percent Sodium on Alumina and CarbonOIOOODOIIIOOO.....0.00.0.0...OOIOIODOO'OIOOOOOOOIIOI0. 2. fligh Surface $dilmsy5t61ns ..... 0.0."OIIOOOOODOOOOIIOIOODO. 3. Methylation of Z-Methylcyclohexanone with High Surface Alkali Metals and Methyl Iodide in Hexane. . . . . . . . . . . ....... . 4 . Methylation of Cyclohexanone with High Surface Alkali hktals alld biethyl IodideOOOIOOO OOOOOOOOOOOOOOOOOO 00.0.0.0... 5 . Methylation of Cycloheptanone and Cyclooctanone ............ . 6. Reduction of cyclohexanone with High Surface Sodium in the Absence ofanadded Proton Donor...... ..... ..... 7. Reduction of Z-Methylcyclohexanone with High Surface Sodium in the Absence of an added Proton Donor. . . . . . . . . . . . . . 8. Ir Data of cis and trans Z-Cycloalkenones. . . . . . . . . . . . . . . . . . . 9. L18 Shift Data forWOOOCI. .................. .....OOIIOIOO 10. L18 shift Data form ........ O. ..... ......OOOOIO‘I ......... 16 22 27 38 38 60 77 LIST OF FIGURES FIGURE 1. Equilibration of Enolate Ions 11 and 11 on Charcoal........... 2. Adsorption of Enolate Ion 11 on Charcoal............. 0.0.0.... 3. Energy for the Twisting of various Electronic States of a 1,2'DiSUbStitUted Olefin ooooooo a o o o o u o 0 Q o ........ o oooooooooo 4 o Blantiolnorphic Trans-CYCIOOCteneS . o ......... a o I o o a u o o I I o Page 20 26 PARTI SYNTHETIC USES OF HIGH SURFACE SODIUM AND POTASSIUM IN ALKYLATION AND REDUCTION OF KETONES AND NITRILES IhHTKYXKTTION High Surface Sodium and Potassium Sodium metal and its derivatives have been widely used as reagents in organic synthesis.10’11a The rate at which sodium reacts depends on its chemical activity and surface area available for reaction. With bulk sodium, insoluble reaction products may coat the surface and effectively remove the underlying sodium from the reaction zone. This often gives undesirable results. High Surface Sodium (HSS) offers a means of overcoming some of the problems associated with the use of bulk sodium, The term high surface sodium1 is applied to films of sodium on inert solids of high surface area. The ease and simplicity of generating HSS make it desirable to prepare sodium in this fonm at the point of use. Although it seems likely that conditions could be developed for storing HSS without appreciable loss of acitvity, this has not yet been done. The preparation of HSS is accomplished simply by mixing molten sodium with suitable inert solid materials having very large surface areas. Usually, sodium spontaneously spreads over the surface of hot dry solids at temperatures between 100°C and 200°C. Among the substances which may be used as sodium carriers are salt, sodium carbonate, charcoal, metal powders, aluminum oxide, etc. Potassium can also be coated on inert solids to give High Surface Potassium (HSP) which is expected to be more reactive than HSS (see experimental section). 2 In HSS, the sodium films produced approach atomic dimensions in thickness, the particle size being even smaller than in sodium dispersions. Thus the surface area available for reaction or for catalytic effects is increased tremendously. Table 11 shows the calculated sodium film thickness versus the percent of sodium adsorbed on both activated alumina and activated carbon. Table 1 Average film thickness vs. percent sodium on alumina and carbon Alumina (1) Carbon (2) % Sodium (160 sq.m/g) (750 sq.m/g) 5% 1 atomic layer 1 atomic layer 10% 1.5 atoms thick 1 atomic layer 15% 2.5 atoms thick 1 atomic layer 20% 3.5 atoms thick 1 atomic layer 25% 5 atoms thick about 1 atom thick * Sodium atom diameter = 4.2A (1) Grade F-ZO, Aluminum Co. of America. (2) Columbia 1W, Carbide 6 Carbon Co. The effective surface area of the solid carrier determines the amount of sodium which can be absorbed. Salt may carry 2 to 10% sodium; soda ash, 10% sodium; alumina, 20% to 25% sodium; and activated or colloidal carbons, over 30% sodium (Table 1)1. Within these concentrations a free flowing solid is obtained, while above these concentrations the mixture becomes a pasty mass. This free flowing characteristic can be 4 maintained at temperatures up to the boiling point of sodium (883°C) depending, of course, on the temperature stability of the carrier. Table 1 High surface sodium systems fl Solid Temp Sodium Carrier oC % Appearance Activated 140-160 20-25 Black or mottled Alumina (1) white granules Activated Carbon 120 35+ Silver to black Coconut (2) pyrophoric powder Colloidal 170 30 Black pyrophoric Carbon (3) powder Sodium Chloride 150 2-10 Gray to black 40-80 mesh colored salt ** Optimum dispersing temperature for the system. (1) Alcoa F-l Aluminum Co. of America. (2) Columbia L-W. Carbide & Carbon Co. (3) hbnarch 71, Godfrey L. Cabot, Inc. It has been shown that high surface sodium is advantageous for the preparation of finely divided metals, for the purification of hydrocarbons and ethers, and for the preparation of inorganic and organo- sodium derivatives.1 There are only a few isolated references in the literature concerning the use of high surface sodium and potassium. Most of the original literature is in patents. For instance, HSS has been used to isomerize and hydrogenate cyclododeca-1,5,9-triene 1 (Eq. 1).3 The potential of H58 and HSP as reagents to effect the alkylation and reduction of ketones will be discussed in this thesis. HSS-A HSS-A a H ’ (Eq. 1) 2 + + 1 other isomers other isomers Alkali Metal-graphite Intercalation Compounds Potassium and the higher alkali metals readily fornilamellar compounds4 with graphite, whereas sodium (or lithium) has a lesser tendency to intercalate in graphite. The best known intercalation compound is CSK in which all carbon-layers are separated by a layer of potassium of atanic size. With such an arrangement, there is a partial transfer of the 4s electron of potassium to the n—electronic system of graphite. This transfer can markedly influence the reactivity of the 4s potassium electron. Potassium-graphite (C8K), though easily prepared,5 has until now found only a few applications in the synthesis of organic compounds. It has been used as a catalyst in polymerization reactions,4 in the nuclear and side-chain alkylation of aromatic compounds with ethylene,6 and in certain base-catalyzed reactions.4 For example, the Claisen condensation of ethyl acetate to ethyl acetoacetate is catalyzed by C8K. C8K acts essentially as a base (Eq. 2).7 , 0 g Cgk (II H CH (:3— l E 2 013012—0- -GiS———-—) 3—C_ 2- ”O CIIZG 3 ( q. N) Pertinent to this part of the thesis is the reaction of C8K with ketones, where it functions primarily as a reducing agent by an electron transfer process. For instance, benzophenone was reduced to benzhydrol in 98% yield (Eq. 1).5 Recently, however, CSK has been shown to have Lewis base properties toward weak acids,8 and it also has been used as base to generate anions from nitriles and activated esters.9 Thus ethyl phenylacetate was metallated with CBK and subsequently methylated to give ethyl Z-methylphenylacetate (Eq. 1). C H OH 8K (L (are —————’ one . THF H N @. cnzcozczfls __, @glooz c2115 (Eq. 4 THE W Due to reports that the sodiumegraphite intercalation compound is unstable, essentially no studies concerning its structure and chemical properties have been made.4 Sodium-graphite was prepared in very much the same way as HSS, although no attempt was made to determine if it is an intercalation compound. Some of its chemical properties will be discussed in this thesis. Alkylation of Ketones and Nitriles 11a constitutes a Alkylation reactions of ketone enolate anions very important, well—established and synthetically useful method of elaborating complex molecules. Due to the weak acidity of ketones and nitriles (pKa2= 19), it is apparent that a stronger base than sodium ethoxide and a solvent less acidic than ethanol must be used in order to obtain an appreciable concentration of their anions. Ketone enolate anions have been generated quantitatively by the action of strong bases or in equilibrium amounts by using weaker bases. A comprehensive list of commonly used strong bases is available.11a Nonnal monoalkylation of ketones has always been plagued by undesirable side reactions such as di-, and polyalkylation, aldol condensation and, to a much lesser extent, O-alkylation and reduction of the carbonyl group. The importance of these side reactions is largely dependent on the nature of the ketone and the reaction conditions employed. When bases like sodium hydride or sodium amide (or sodium alkoxides) are used, the enolate is formed relatively slowly in solution, and aldol condensation may occur due to the presence of appreciable concentrations of both the free ketone and its enolate anion. Aldol condensation can easily be avoided by slow addition of the ketone to a solution of a strong base in a suitable aprotic solvent. In this way, no excess ketone is present to react with the enolate. Among the strong bases, the lithium dialkylamides (e.g. lithium diisopropylamide) are particularly convenient to prepare and use. These amides possess bulky alkyl groups, and therefore preferentially abstract an alpha proton from a ketone rather than attack the carbonyl group. However, these amides slowly attack ethereal solvents such as THE and 1,2-dimethoxyethane at elevated temperatures. Consequently the ethereal solutions of lithium dialkylamides must be prepared at 25°C or less and used promptly. When an equivalent amount of a strong base is used in the monoalkylation of ketones, the amount of di- and polyalkylation can be reduced but not eliminated. Polyalkylation is due to a proton exchange between the alkylated product and the starting enolate (reaction k2 and k3) as illustrated in the following reaction with cyclohexanone.12 If reaction k2, k3 and k4 compete with the initial alkylation k1, complex product mixtures will be formed. O'Li+ base mm,o’c “‘_‘> (56%) ether, 0°C Cl 3I k1 k2 011' kg - + _ 0 Li 0 L1 0 . (11.51 + E k4 (13%) In order to minimize the side reactions encountered in the alkylation of ketone enolates a variety of cations and complexing agents have been used. For example, the problem of dialkylation can be considerably reduced by the use of triethanolamine borate (TEAB) or triethylborane as complexing agents.13 9 Li“o'B(0R)3 Ra _ . R/N L1 cm I c113 ————> —3a 13(012)3 R.T. or TEAB Often in the course of a synthesis the need arises to introduce an alkyl group selectively at one of the two alpha positions of an unsymmetrical ketone. Selectivity in the alkylation of unsymmetrical ketones may be achieved by using activating or blocking groups,113 or by taking advantage of the fact that kinetic control generally favors creation of the less substituted enolate anion, whereas under equilibrating conditions the more substituted anion usually predominates. Other means of generating a specific enolate ion, such as reduction of a,8- unsaturated ketones,14 or reaction of enol ethers and esters with organometallic reagents,123 have also been used to accomplish specific alkylations. The alkylation of nitriles usually presents no problem other than the separation of the monoalkylated product from starting material or dialkylated product. Thus the alkylation of nitriles and subsequent 10 hydrolysis often provides a good synthetic route to a -substituted acetic acids in spite of the rather vigorous conditions required for the hydrolysis (Eq. (p.113 NaNHZ 112504 (80%) n—C4H9—GIZ-CN —————> (n—C4H9)3C—CN —-—-————-—> n—C4H9Br H20,reflux IDCHS , reflux 81% n—C4H9—ONO (n—C H ) C—CONH -——-———) (n—C H ) C—CO I (Eq. 6) 1493 2 HCl,HOAc 493 zI v. 90% 79% Alkylation of QLEfUnsaturated Ketones 11a Alkylations of a,B-unsaturated ketones having enolizable gamma hydrogens proceed almost exclusively at the alpha position to form the or alkyl - B, Y-uns aturated ‘ ketone . 23 This initial product may isomerize to an a-alkyl-a, B-unsaturated ketone or may undergo further alkylation. Dialkylation has often been the major reaction in such cases because a proton is abstracted more readily from the intermediate B,Y- unsaturated ketone than from the starting material or the alkylated a,8-unsaturated ketone. Dialkylation, however, may be diminished either by the slow addition of the alkylating agent or by the use of a less reactive alkyl at ing agent . 11 C) ___.___has_e_. ____. RZGICH—(‘llmR R21“. CH—CllCOR<——> slow (ll 1 3 base : (113 ('113 RZC = CH-C-CO-R 4——-> RZC—Gl =: C—CO-R (Eq. Z.) G + (“311 H J [slow ‘5”3 (.113 RZC=CH—C—CO—R RZ-CH—Ol-C-(D—R CH3 When lithium-amnenial4 was used as the base for the alkylation of a,B-unsaturated ketones, the only monoalkylation product obtained was derived from alkylation of the specific lithium enolate produced in the reduction step (see Eq. 8) . This reduction -a1kylation procedure provides an excellent method for directing alkylation to the relatively inaccessible 0‘-position of an unsymmetrical ketone. The success of this method depends upon the now well established fact that alkylation 'Li+ Me 2 Li IIe E181 Die e (Eq. 5) ——> ——————> N”3 EtZO Me Me Me Me Me Me of specific lithium enolates of unsymmetrical ketones with relatively reactive alkylating agents occurs faster in a variety of solvents than does equilibration among the structurally isomeric enolates via proton transfer reactions . .Jn» 12 Reduction of Ketones Ketones can be reduced to alcohols by a wide variety of reducing agents such as lithium aluminum hydride, diborane, hydrogen over a platinum catalyst and dissolving metals.11b Pertinent to this thesis are the dissolving metal reductions. These reductions of aliphatic ketones to alcohols are believed to follow the reaction path indicated below. 0 ' on“ M- OH ROH m (113 (113 2 3 H l a. II T - + n on n (11 on n- 3 $2 ROH n or u or C—O ‘ R2. - 013 on H .3 e13 H 4b ’V\a An electron is transferred from the metal to the ketone to form a radical anion 2 which is then protonated to give a free radical intermediate such as 1 with the hydroxyl group in an equatorial position. In a medium containing excess reducing species, rapid reduction of the neutral radical 3 to form an anion 4 or an organometallic intermediate is to be expected. This species would also be expected to adopt the indicated more stable geometry 4. The final protonation of the anionic intennediate with retention of configuration at carbon leads to the observed more stable product. 13 Radical anion 1 may also exist as a dimeric ion pair 6, especially in relatively nonpolar media when the cation is a relatively small alkali metal cation such as lithium or sodium. Pinacol products may be formed under these conditions. The fonnation of dianionic intennediates such as 1 from aliphatic ketones, or even from a,B-unsaturated ketones, seems unlikely, since the reduction potential for forming such species is too great, and the commonly used reducing systems2 just do not have sufficient reduction potential for the second reduction unless the negative charge is first neutralized. - + 2le .Nai _ 'l “ . O R \ . " ‘ - / R c —- o, o -c R ’ \ R 013 Na“ H 9, Z We have seen that ketone alkylations, when carried out in solution using conventional strong bases, often suffer from low to modest yields, lack of regiospecificity, considerable dialkylation and other problems which clearly limit synthetic utility. High surface sodhnn and potassium are easily prepared, constitute a highly reactive fonn of alkali metals, and high surface area of solid support. Therefore, it was thought worthwhile to examine and explore the synthetic applicability of high surface sodium and potassium, particularly in the alkylations and reductions of ketones. It was the hope that in these alkylations, the enolate ions generated by HSS would be adsorbed specifically on the surface. In this way, monoalkylated product might be fonned regioselectively as compared to the conventional alkylations in solution. The high surface 14 sodium reduction of ketones might also occur stereoselectively. Results of these studies are presented in this part of my thesis .. RESULTS AND DISCUSSION High surface sodium on charcoal (HSS-C) , high surface sodium on graphite (HSS-G) , high surface potassium on charcoal (HSP-C) and high surface sodium on alumina (USS-A) were prepared for this study. Procedures for these preparations are given in the experimental section. 1. The Alkylation of Cyclic Ketones A. Alkylation of Z—Methylcyclohexanone It was originally thought that an ideal substrate for this study would be a simple, unsymmetrical ketone which could give us information as to whether alkylation using 1188 is regioselective. 'Ihus Z-methyl- cyclohexanone was chosen. Hexane was chosen as the solvent simply because it is inert to alkali metals and relatively easy to dry. Methylation of Z-methylcyclohexanone was carried out with high surface sodium and potassium under various conditions. Results of these methylations are presented in Table ,3. The products were isolated by VPC and identified by comparison of their IR,NMR and “mass spectra with those of the literature. The product mixture contained Z-methylcyclohexanol 8, Z-methylcyclohexanone 9 , 2 ,2- dimethylcyclohexanone 1,0, 2 ,6-dimethylcyclohexanone 1,1, and 2,2,6- trimethylcyclohexanone 12. In one reaction, 2,2'-dimethy1bicyclohexyl-l,1' - ~ diol 13 was isolated in trace amounts. 15 16 Table 1 lbthylation of Z-Methylcyclohexanone with High Surface Alkali Metals and Methyl Iodide in Hexane.a’b o o 0 OH 0 0 0, (j —-> (7+ (3+ (1630+ d—O'b 8 9 10 11 12 1 2 N '\a ’V\a 'VV ’\/\4 ME Reaction metal(%) 8(%) 9(%) 10(%) 11(%) 12(%) 13(%) time q, q, 'Vb 'VM ’Vb '\/\a 1 hr HSS—C(10) 15 22.5 28.1 3.0 1.3 - 1 hr USS-C(15) 18.5 10.1 38.3 4.2 4.0 - 2 hr HSS-C(20) 11.8 19.7 45 5 5.7 - 1 hr HSS-C(25) 20.4 14.2 43 4.6 2.8 - 1 hr HES-C(30) 20.1 16.9 42 4.5 2.7 — 2 hr ass-0(20) 8.9 2.3 57.6 6.4 14.5 0.5C 1 hr ESP-C(10) 6.7 36.5 18.0 1.8 2.3 - overnight HSS—A(15) 10.2 19.8 45 4.4 7.2 - (a) Equimolar amounts of ketone, metal, and methyl iodide were used. (b) Yields are calculated by VPC analysis using tridecane as an internal standard. (c) Isolated yield. 17 The presence of 2,6-dimethylcyclohexanone 11 was confirmed by comparing its VPC retention time (using a capillary column to effect separation) and its NMR spectrum with those of an authentic sample. The NMR spectrum of the alcohol 8 showed that it consisted of about 85% of trans form and 15% of cis ferm. The IR spectrum of compound 13 in CCl4 solution shows absorptions at 3620 and 3550 cm'1 (Av = 70 cm‘l) indicating strong intramolecular hydrogen bonding15 between the l- and l'-hydroxyl groups. It is apparent from the data in Table 3 that the alkylation of Z-methylcyclohexanone using high surface alkali metals does work, and that monoalkylation predominates, usually to a greater extent than for comparable reactions in solution (for example see Eq. 9). The monoalkylaton product in all cases studied is mainly (= 90%) 2,Z-dimethylcyclohexanone indicating that the surface may provide some selectivity with regard to 'different ketone enolates. In solution, the predominance of 2,2-dimethyl- cyclohexanone over 2,6-dimethylcyclohexanone is usually not so large (see .lla,12b The main competing reaction to alkylation is reduction to Eq. 9) the corresponding alcohol, and in one case to a pinacol. Methylation with high surface sodium on graphite (in place of charcoal) seems to give more dialkylation product, probably because HSS-G is more reactive than HSS-C. In general, 20% HSS seems to give the best yield of monoalkylation product. 18 e - {tree done A: 03CK (1131 we 22 9 - 41 21 6 (% yield) B: NaH,G-131 [ME 48 13 13 18 8 - (% yield) (Eq. 2,) To see if the solid support would exercise a selectivity between different enolate ions, a mixture of lithium enolate 1,3. (99%) and 15 (1%) was generated by the reaction of 2—methy1cyclohexanone with lithium diisoprOpylamide under conditions of kineticcontrol,‘ according to 12a Charcoal was then added and the resulting enolate House's procedure. solution was alkylated with methyl iodide. Surprisingly, the monoalkylation product found was mainly 2,2-dimethylcyclohexanone, which is the alkylation product of enolate ion 15 (Eq. 10) . On the contrary, the same enolate ion mixture 14 and 15, when reacted with methyl iodide directly, gave 2,6- dimethylcyclohexanone as the major monoalkylated product. Apparently charcoal (and other solid supports) shift the enolate composition from predominantly the less substituted enolate ion 14 to predominantly the more substituted enolate ion 15, leading to the geminally substituted ketone. 19 Charcoal U (j 0' '+ L1 9(20. 2%) 10(29%) 0 O O “(99%) s 6 >919 . , __DME_1_, + __ q110.9.) q1,202.2.) ' o’Li+ 9 N (21131 15(10 I 9(2%) + 11(57. 4% ) + v» °) 1M3 12(24.4%) 'W (139.39) A plausible explanation for this result can be envisioned as follows. In nonpolar solvents, metal ions are mostly on the surface and most of the enolate ions are adsorbed on the surface. Models of enolate lg show that the methyl group is in the double bond plane, and can be adsorbed parallel to the surface, whereas in enolate ii the C-2 methyl group is either in a quasi-equatorial or quasi-axial position which would hinder adsorption on the surface to some extent. Charcoal may contain some proton donor (HR, such as trace of water in this case, and starting ketone in the case of HSS-alkylation) which would catalyze the equilibration between enolate l1 and £5 (Fig. 1) with the result that the thermo- dynamically more stable enolate lg predominates and the geminally substituted alkylation product is obtained. 20 £fi. 15 (013) -n“ H v— V—" /<7’ .. “ 013 .1 (H) 013 .9. 3 4- Li IIéR LiR R—H Li / 7777777777777 ll//////////////7 Surface surface Surface Figure l. Equilibration of Enolate Ions l4 and IS on Charcoal The adsorption of enolates on the surface can also account for the predominance of monoalkylation product over dialkylation product. Since the monoalkylated product has to diffuse to another place on the surface to fbrm enolate again for dialkylation, this process is relatively slow and noncompetitive with.monoa1kylation. Therefore only a small amount of the dialkylation product would be expected. In contrast to the conventional methods for the alkylation of ketones, side products like aldol condensation and O-alkylation products are not observed in HSS alkylation of Z-methyl- cyclohexanone. The reactivity of the alkylating agent is also important in the a-alkylation of ketones. When the alkylating agent is more reactive, as with allyl bromide, competing dialkylation and reduction are decreased. Thus, Z-methylcyclohexanone was alkylated with 15% HSS-C and allyl bromide to give a 68% yield of monoalkylated product lg along with some starting ketone and reduction product. 21 15% 1183-6 / g g (11.5%) 901%) . l2(68%) B . Methylat ion of Cyclohexanone High surface alkali metals were also used to study the methylation of cyclohexanone. “The results are smnmarized in Table 4'. Alkylation still predominates over the reduction. In this case dialkylation constitutes a serious side reaction, particularly when 'IIIF was used as solvent. 'One possible explanation is that the enolate ion remains adsorbed on the surface in hexane but is somewhat soluble in THF . Thus in THF equilibration of the initially formed enolate anion with the monoalkylated product [will compete with monoalkylation and result in dialkylated product as shown in equation 3.1)ialkylation was more serious when USS-G was used in place of IISS-C. o —-—-» moody. a 9% 2,2212 1% a 22 Table fi Methylation of Cyclohexanone with High Surface Alkali Metals and Hethyl Iodidefa‘;h Reaction Solvent Alkali l§(%) lZ(%) 3(%) 12(% ll(%) 12(%) time 2 hr hexane ”SS-C(20) 10.5 7.5 49 1.5.0 1.7 10.5 2 hr hexane USS-C(20) 3.3 5.0 38 23 2.4 18.5 2 hr 11”: “SS-C(20) 26.8 1.3 22.7 26 2.9 5.0 1 hr hexane RSV-C(10) 8.5 18.9 33.3 3.6 0.4 - (a) Equimolar amounts of ketone, metal, and methyl iodide were used. (b) Yields are calculated by VPC analysis using p-diisopropylbenzene as an .45.; L standard. (c) Isolated.vie1d. Pinacol is fonned in a greater amount than alcohol when hexane is used as the solvent, whereas in THF more alcohol is formed than pinacol. It has been suggested that pinacols nonnallv arise from the dimerization of . . . . . h 0 the ion pair dimer 69 formed form radical anion $§°1J ’1 Radical anion 45 M is probably not soluble in a nonpolar solvent like hexane and remains adsorbed on the surface as an ion pair dimer which would favor the formation of pinacol. Radical anion is somewhat soluble in THF: it can either diffuse hack to the metal surface for fiirther reduction or abstract a proton from the starting ketone in solution, and subsequent hydrolysis would yield an alcohol (see page 27 for mechanism). . 23 + ' Na 0 -Na+ 0:. ..\‘ . ‘\ 5‘ ’1'! Oil H .__, Na" a 4 3. 4% 1% In order to gain some understanding about the mechanism, particularly the source of hydrogen for the unimolecular reduction, a labeling experiment was performed. Treatment of cyclohexanone with 020 and potassium carbonate )2” gave 2,2,6,6-tetradeuterocvclohexanone 22 (=- 92% d which has only a 4 singlet at 6 1.80 in its .‘NP. spectrum. Compound 22. was then methylated with 20% USS-C and methyl iodide in hexane for 2 hr. When the reaction mixture was quenched with 1120 and (11.501), there was obtained a mixture of deuterated compounds which was analvzed by \”’C (using tridecane as an internal standard) to give the product composition indicated in Eq. 11. Among the products, 1,2,2,6,6-pentadeuterocvclohexanol-0-d z)» is worth noting. Compound 2’1» was clearlv deuterated at the C-1 position because in its 1le spectrum the area of the peak corresponding to the C-1 methine proton was reduced by 9091; also, the mass spectrum showed a parent peak at m/e 106. When the reaction mixture was ouenched with I120 and ethanol, the alcohol obtained was 1,2,2,6,6—pentadeuterocvclohexanol 2,6 which also had deuterium at the C-1 position according to its NMR spectrum and to the parent peak at m/e 105 in its mass spectrum. However, the recovered starting ketone in this case had most of its a—deuteriums exchanged with hydrogen (69% d0, 25% d 6% d2) and the monomethylated ketone also 19 24 had some of its a—deuterium exchanged with hydrogen (19%d 38% d d 0’ 1,11% 9 2, 120 d3). To be sure that the deuterated compound 23 was the 2,2-dimethy1ated product instead of 2,6-dimethy1ated product, it was treated with K200.5 and water, to give after workup the non-deuterated 2,Z-dimethylcyclohexanone (z 90%, = 10% of it is 2,6-isomer). D O 2 D D2 -—-————-————> K2003 1 7 «220 20% USS-C CH31 .hexane lDZO H7O 101300 ~ CHSOH o e e 2552 21(6. 7%) 20(12. 4%) 22(35. 7%) 26 ’VD + D D OD D D exchanged starting ketone, 2. [::f:j———JE:::i] alkylated product, pinacols. D2 D2 2306) 9 25(17.30) (Eq. 1,1») .53”.- 25 C. Methylation of 4-t-Buty1cyclohexanone In the case of 4-t-hutylcyclohexanone, high surface sodium methylation gave the product composition shown in liq . 1,2. The monoalkylated product cis and trans-Z-metbyl~4-t-butv1cyclohexanone were obtained in 93 : 7 relative ratio which is the equilibrium rather than the kinetic ratio.21 The predominant formation of cis-Z-methyl-4-t-buty1cyclohexanone may be because in order to have better adsorption on the surface, enolate a has to admit a conformation (possibly half chair, Fig. ,2) with its 4-t-buty1 group pointing away from the surface: as a result, methyl iodide would have to approach the enolate anion from the side where t-butyl group is sticking up, to give the cis product 258’ The reduction products found were trans-4-t-butylcyclohexanol awhich is the more stable alcohol , and pinacol 32 . 20% IISS-C ———+ + (‘HSI hexane «2’3 21(172 29,) ano) 29(32. 5%) i "’4’: it) OH: 30(2. 4%) 31(4 %) 32(25. 1%) (Eq. 3,3) 26 OLI D \\ \ ~ 33 'Vh Na+ III/IIIIIII/r Surface Figure 2. Adsorption of Enolate Ion QR on Charcoal D. Methylation of Cycloheptanone and Cyclooctanone In order to see if ring size would have any effect on the high surface sodium methylation, cycloheptanone and cyclooctanone were studied. The results are summarized in Table 3. They show that monoalkylation still predominates over dialkylation, and that the dialkylated products are 2,2-isomers instead of 2,6—isomers. Reduction to alcohol is less important for cycloheptanone than for cyclohexanone, and even less so for cyclooctanone. The reason for this is not clear. 0“ o oeeeww 0 CS C1600 a"? $2 $1 42 27 Table 5 Methylations of Cycloheptanone id and cyclooctanone éfi Ketone Reaction Alcohol Starting Monoalkylated Uiallylated Pinacols time (%) ketone(%) ketone(%) ketone(%) (%) ' 9 9 A 9 .9 79 éfi overnight ££(6.So) §i(21.8u) §2(.5.8 ) §Z(5.8 ) Zifl“ ) $2 1.5 hr - 39(36.2%) 40(41%) 41(l4%) 42(5%) ’W M 'V): ’V): (a) Methylations were done with 20% HSS—C and methyl iodide in hexane. (b) Product composition calculated by VPC employing triangular method. (c) nialkylated products are predominantly 2,2-isomer in each case. B . Mechanism The fact that alkylated products and reduction products are both formed in the alkylation of saturated cyclic ketones with high surface alkali metals indicates that the formation of both the intermediates (enolate ions) leading to the alkylated products and intermediates (radical anion)16’17 leading to the reduced product are occurring simultaneously. A possible general mechanism to account for the results of the USS alkylation of cyclic ketones is shovm in Scheme 1 (cyclohexanone is taken as an example). High surface sodium (or potassium), due probably to its atomic dimension, may abstract an alpha hydrogen from ketone lg to form radical Scheme 1 N 'Na+ Na - ‘ Na. NaH or 1220—02 17 43 44 '\/\a 'VM ’VM 4. NaH 1, O O ()4 22 2—-— 2-- 10 9 ’\/\o N + - O O Na Na ~ ___._2. or RZC—O_ 17 45a 45b Iv» W» W ’Na" 0': ,0- Na+o‘ O'Na+ (a) \Ié" 45b --—--’ —-———-> MA; $8 inzo OH OH Scheme 1 (continued) 0 — + - + 0 Na 0 Na Cb) Na+ (j H H 0 OH Na~ 2 ——_, —————-—’ '—'—_‘7 or R294)- 48 18 47 “A: 'VV 'V): + 44 'V): O Dimerize i i) 0” 19 'Vb 421: ~——’ 226°” (C) Na' ”20 -——-»18 f}: or ch—o_ m + 50 M: M . o . - + + 6 II 0 0N3 (d) Na, 1120 45b e—-—~ ~———> ———> > 18 NNN or R2943- ”N 51 48 ’\/\1 'VV + 44 30 fig/band sodium hydride. Radical 4; thus formed may further be reduced to the enolate ion 2% which then reacts with methyl. iodide to give the monoalky- lation product 2. Subsequent enolization and alkylation of 2 should give rise to the diallylation product ’12. Enolate ion 3’13, may also be produced by the abstraction of an alpha proton from ketone £5 with sodium hydride or radical anion in situ. The reason for the predominant formation of the more substituted enolate ions from the unsymmetrical ketone with 188 is presented on page 19. Since dissolving metals and alkali atoms have about the same reduction potentials,2 the mechanism to account for, the formation, of reduction products of cyclic ketones by high surface alkali metals may be similarily envisioned as those for the reduction of ketones by dissolved metals.11b Transfer of an electron from the high surface sodium to the antibonding u*-orbita1 of the carbonyl group can give rise to a radical anion % (or ketyl)16 having the greater unpaired electon density on carbon. This radical anion has several competitive reaction pathways open to it . When proton donors are absent, formation of the tight ion pair 41/5}; and the ion pair dimer 3’6, derived from it would be expected.' The ion pair dimer $6 in a nonpolar solvent can couple to form pinacol 13. however, in the case of Z-methylcyclohexanone, steric hindrance due to the C-2 methyl retards pinacol formation to the extent that only trace of it was observed, thus pathway 3 would seem to be noncompetitive. Although it is unlikely that the reduction potential of atomic sodium 31 is sufficiently negative to fonn free dianions from aliphatic ketones, tight ion pair 352 might be capable of adding a second electron to form dianion £1, for the negative charge in fiéh is already partially neutralized by association with metal ions. The dianion $1, in the absence of added proton donor, can abstract a proton from the weakly acidic starting ketone to form alkoxide ion 38 and enolate 3i; Alkoxide ion 38 upon hydrolysis should give alcohol 18. '\/\I Radical anion £52, in a less likely reaction, might also abstract a proton from starting ketone to give rise to an hydroxy radical Q2 which could then be further reduced to hydroxy anion 52 in the presence of excess reducing species. The hydroxy radical i3 may also couple to yield pinacol 12. Protonation of radical anion 152 can also occur at the carbonyl carbon as suggested by House19 and.hhrphy18 to give alkoxy radical 51 which could be further reduced to an alkoxide ion Q8, and finally hydrolyzed to alcohol 18. It is difficult to tell which pathways would be favored for 152 in the reduction part of the mechanism. However, on the basis of our results some suggestions can be made. Pathway b may not be competitive because no excess reducing agent was used in almost all cases studied. Pathway 2 may be ruled out based on a labeling experiment which showed that the C-1 methine proton of the reduced product (alcohol) came from the starting ketone (see page 23). Therefore, pathway d seems most likely to be operative. In general, the results showed that the alkylation products were formed in greater amounts than the reduction products in the HSS alkylation of saturated cyclic ketones. Hence, the rate of fonnation 32 of enolate ions would be faster than that of the radical anions. 2. Methylation of Isophorene A high surface sodium alkylation was also done on an 11,8 -unsaturated ketone. Interesting results were obtained for the methylation of isophorone with 20% HSS-C (10% excess of Na) as indicated in Eq. 13. It is apparent that in this particular case reduction competes very well with alkylation. A substantial amount of 1,6-diketone reSulting from the reductive coupling at the B-position of isophorone was obtained. 20% 1158- C C113 I hexane 9,2 92 (12.6%) 9,; (10%) 34 (32.1%) 9,9 (38.3%) (Eq. 1,3) In contrast to the metal—anmonia reduction of isophorone, no corresponding unsaturated pinacol was found.37 And unlike the reductive alkylation of a ,6 -unsaturated ketones with lithium-almonia, no saturated alpha alkylated product was observed in this reaction.14 The only alkylated product was Z-methylisophorone 5i which presumably arises from alpha alkylation of the dienolate anion formed by abstraction of a y-proton or y'-proton from isophorone, and subsequent isomerization. There have been many intensive studies concerning the reactions of a,B-unsaturated ketones, particularly reductions and alkylations with '33 metal-ammon1a. 7 Several possible mechan1sms 1nvolv1ng a radical anion as a key intermediate have been proposed fer these reactions. Similar mechanisms for the high surface sodium alkylation can also be envisioned as shown in Scheme 2. Scheme 2 '\a 07951" O O or R2943, + 1 52 O Na 0 54 57 59 ’V‘a ’\/\a ,Na ‘ O'Na+ 0:" 2 o' 52 M 34 Scheme 2, (continued) 0 OH 2 i couple 2’1" __ on on 39 ——)—-—> - HO (b +e 2 L————~) + ‘—', 62 Na + 56 + 57 63 641 «A. M. M’ 1’ '53 «A. o no - + O Na+ fl 0 Na 2 ,__> 53 - o 60 L90 ”—7 0 fl CH I (C) - L__, 65 66 ’V» M + 2’6" «1» a $7» 0 . 0. O’Na+ I k ' H o ((1) +6- , 3—9 5 60 -—-—+ "—_’ M a 22 + 56 + 57 £0131 Mb «I» 67 35 Formation of dienolate anions £2 and 51 from starting isophorone may be effected by either high surface sodium (in the same way as with saturated cyclic ketones) or by radical anion. Alkylation of the dienolate anions at the alpha positions would fonn the B,y~unsaturated ketones $8 and 52. Isomerization of these initial products then gives the observed product éfi' This result suggests that the dienolate anion formed by abstraction of a y proton (or y' proton) from an a,B-unsaturated ketone is more stable than its cross conjugated isomer generated by abstraction of a a'-proton. The initial step in the reduction part of the mechanism is the transfer of an electron from high surface sodium to an antibonding n*- orbital of the conjugated system, to produce radical anion Q2. This radical anion may have four possible pathways for further reaction. In a nonpolar solvent, it may exist as an ion pair dimer which then dimerize to give, after hydrolysis, the intermediate dienol 21 and tautomerization of $1 eventually affords 1,6-diketone 55. In pathway 2, in the absence of added proton donor, radical anion 60 can abstract an acidic proton (Y or y') from isophorone to produce the hydroxyallyl radical 33 and dienolate anion 5p or 51. The hydroxyallyl radical g2 thus generated may be further reduced either by M88 or another radical anion to give a hydroxyallyl anion Q§_which would undergo protonation to afford enol pi, and subsequent tautomerization should give a saturated ketone éé' Radical 32 can also couple to give pl. Dienolate anions 52 or 57 would be alkylated to yield the u-alkylated product 53. 'V\a Pathway 2 involves the fonnation of dianionic intennediate 25 by 36 adding a second electron to the radical anion 2Q. Dianionic species have been suggested as intennediates in the reduction of enones with metals 17 Protonation at the B-position of the dianionic in liquid ammonia. intermediate Qé then gives the enolate anion 66. The hydrogen introduced at the B°position is derived from a proton donor and, in the absence of added proton donor, starting enone can serve as the proton donor. Enolate anion 66 can be alkylated at the a-position to yield the alkylated product Qz or hydrolyzed to give the reduced product ég. Pathway d, which seems least likely, involves protonation of the N radical anion 28 at the B-position17 to produce the enolate radical Qfi which adds an electron to give the enolate 69. Alkylation or hydrolysis of enolate 66 would yield 67 or 53. Since no ketone 67 was found in the ’\/\o Ivy Ivy ’\/\a “SS methylation of isophorone, pathways g and d may be ruled out. The result that 1,6-diketone £3 was fonned in greater amount than the monomeric reduction product is indicative of pathway 3 being faster than pathway h as far as the radical anion 60 is concerned. ’VV 3. hbthylation of Phenylacetonitrile Recently, phenylacetonitrile has been alkylated using potassium- graphite (C8K) as base by a group of Italian chemists.9 They employed a molar ratio of nitrile : CSK : alkyl halide of l : 2 z 2, and the yield of monoalkylation products obtained were about 60%. High surface sodium has proved to be useful in the alkylation of nitriles and the yield is better than that reported for CBK‘ Thus the methylation of phenylacetonitrile 1 with high surface sodium gave good yield of monoalkylation product as 37 indicated in Eq. lie; only small amount of the dialkylated product was isolated. The molar ratio of nitrile : IISS : methyl iodide employed was 1 : l : 1. The products were isolated by VPC and identified by canparing their spectroscopic and mass spectral data with those of authentic samples. Since no reduction product was found, the a-carbanion is likely to be the only intermediate leading to the formation of alkylated products. ‘ 3 3 l 1 © 01201 209. ass-q @0120: © CIICN @L-CN . + + 0131 _ a, 9 9 9 6"]9" 6v9\’(7060) 7IV0\’(74 O) 1%,(404 0) (Eq. {1) 4 . Reduction of Ketones Since the high surface sodium methylations of cyclic ketones are always accompanied by a certain amount of reduction, it was interesting to see to what extent ketones would be reduced in the absence of an alkyl at ing agent . Thus cyclohexanone and 2 -methyl cyclohexanone were reduced with HSS-C and HSS-G under various conditions, and the results are given in Tables 6 and Z. No proton donor was added in all cases studied . 38 Table 2 Reduction of Cyclohexanone with High Surface Sodium in the absence of an added Proton Donor. a Metal(%) Solvent Reaction 18(8)d gland 12(8)° time Alcohol S .M. Pinacol HSS-C(20) ' hexane 2 hr 16.8 66.5 6.3 ass-C(20) hexane. 2 hr 9.4 63.2 20.6 HSS-G(20) 1111: 4 hr 52.4 33.3 5.3 HSS-C(20) nu 4 hr 55.5 31.5 3.6 ass-(X30)b mp 2 hr 67.6 20.5 3.0 (a) No proton source was used except at workup. Ketone : metal = 1 : 1. (b) Ketone : metal = l : 2 molar ratio. (c) Isolated yield. (d) Yields determined by VPC analysis using p-diiSOpropylbenzene as internal standard. Table z Reduction of Z-l-Iethylcyclohexanone with High Surface Sodium in the absence of an added Proton Donor.a Metal(%) Solvent Ratio 8(8)d 2(9.) 13(8)C figmetalj A1 cohgl S . M . Plnacol BBS-C(20) hexane 1 : 1 ' 25.8 62.6 2.0 HSS-C(20) 1111: 1 : 1 33.1 54.4 4.0 HSS-G(20) hexane l . 2 30 66 3 HSS-G(ZO) 1m: 1 : 2 83.4 8 3.5 (a) No proton source was added except at workup . (b) All reactions were stirred at 60°C for 6 hr. (c) The yields indicated refer to isolated yields. (d) Yields determined by VPC analysis using p-diisopropylbenzene internal standard, and trans / cis z: 85 / lS. 39 It is clear from the results of both Tables .9, and 1 that the reduction of ketones with high surface sodium in the absence of an added proton donor would not go to completion even after prolonged reaction time. Apparently, the starting ketone has served as a proton donor, becoming converted to an enolate anion which, after workup , gave back the starting ketone. Reductions carried out in THF- seem to yield more alcohol and less pinacol. This is probably because the radical anion intermediates formed during reduction are somewhat soluble in 'I'IIF and subsequent diffusion back to the metal surface or an encounter with another radical anion (or ketone) occurs more readily. The reduction of 2- methylcyclohexanone with 2 molar equivalents of high surface sodium on graphite afforded a very good yield of alcohol, the most successful of the reductions studied. Thus, it seems necessary to use the molar ratio of ketone : metal of 1 : Z for the reduction of ketones to alcohols without an added proton donor. The mechanism of the reduction of ketones with high surface sodium should be considered the same as the reduction part of the mechanism for the alkylation of cyclic ketones with HSS (see Scheme 3) . When ketone : metal = 1 : 1 molar ratio. is employed, pathways 9 and g are probably competitive processes. In the case of ketone : metal =' l : 2 molar ratio, dianionic intermediate might be produced predominantly and subsequent hydrolysis would yield an alcohol. Although reduction of ketones to alcohols can be achieved with high surface sodium in good yield, very little is lmown about its detailed mechanism, and more research remains to be done in this field. 40 5m In this part of the thesis, evidence of the metallating and reducing properties of high surface sodium have been provided. In the alkylations of ketones or nitriles with HSS monoalkylation usually predominates: where regioselectivity is involved, the more highly substituted.product is Obtained. As to the HSS reduction, alcohol can be obtained in high yields with H88 (2 molar) in 'I‘HF. The simplicity of workup, good yield of the desired product and the inexpensiveness of the reagent may make high surface sodium a synthetic useful reagent. For instance, HSS has been used to effect the monoalkylation of 1-tetralone and Z-tetralone in excellent yield.24 EDG’ERIMBNI‘AL 1 . General Procedures The general procedures described here apply to all parts of the thesis. Analytical gas chromatography (VPC) was carried out on a Varian Aerograph Model 1400 (flame ionization detector), and preparative VPC was performed with a Varian Aerograph Model 90 P instrument (thermal conductivity detector). Except where otherwise noted, all I-MR spectra were measured in CDCl or CCl4 solutions using TMS as an internal standard on a Varian 3 T-60 spectrometer. The 180 MIz spectra were recorded on a Bruker spectrometer. The small number placed next to protons in the structures in the discussion sections are the NMR chemical shifts of those protons relative to tetramethylsilane. The numbers in parentheses beside the chemical shifts are the normalized europium shift numbers. These were obtained by adding small increments of tris- (1,1,1,2,2,3,3-heptaf1uoro- or CDCl solution of the 7,7-dimethyl-4,6-octanedione) Eu (III) to a CC14 3 compound being investigated. Infrared spectra were recorded on a Unicam SP-ZOO or a Perkin Elmcr 167 grating spectrophotometer and were calibrated against a polystyrene film. Ultraviolet spectra were obtained with a Unicam SP-800, using 95% ethanol as the solvent unless otherwise noted. Mass spectra at 70 eV were obtained from a Hitachi-Perkin Elmer RMJ-o Operated by Mrs. Ralph Guile. High resolution mass spectra were done in Biochemistry Department, MSU. Melting points were determined with a Thomas-Hoover Melting Point 41 42 Apparatus and are uncorrected. Analyses were performed by Spang Microanalytical Laboratories, Ann Arbor, Michigan, or Clark Microanalytical Laboratories, Urbana, Illinois. 2. Preparation of High Surface Sodium (HSS) High surface sodium (HSS) and high surface potassium (HSP) of various percentages on different solid supports were prepared according to the following procedure (20% HSS on charcoal is taken as an example): 4.6 g of activated charcoal (predried under vacuum at 200°C for 15-30 min) was placed in a 250-ml 3-necked round-bottomed flask (equipped with mechanical stainless steel stirrer, thermometer, nitrogen inlet and oulet, and sodium dropping port) and heated to 120°C with stirring under nitrogen. Sodium (1.15 g, 0.05 mol) in small pieces was added through the port over a period of 15 min. As soon as the sodium melted, the mixture was stirred more rapidly for about 15 min. Upon cooling to room temperature, the HSS is then ready for use. The abbreviations used are as follow: HSS-C, HSS-G, and HSS-A for high surface sodium on charcoal, graphite or alumina respectively, HSP-C for high surface potassium on charcoal. HSS-C, HSS-G, and HSP-C are highly reactive, pyrophoric powders, black to dark purple depending on the percentage of the metal used. A small sample of each, when exposed to air, becomes red hot (sometimes even catches fire, HSP-C in particular) A. 2A. Chromatographic Colunns VPC columns that were used for analysis or preparative work in this section are designated as follows: 43 _ . 10' x 0.125 in colunn, 15% D.C. silicone oil 710 on firebrick. . 10' x 0.25 in column, 20% D.C. silicone oil 710 on firebrick. A B C. 8' x 0.25 in column, 15% Ucon on firebrick. D 5' x 0.25 in colum, 5% FFAP on chromosorb G. E . 150 ft D.C. silicone oil 550 capillary column. 3. Methylation of Cyclohexanone with HSS-C and Methyl Iodide The methylation of cyclohexanone can be considered as typical for the procedure used with most ketones . Except where otherwise noted, all reactions were carried out under nitrogen or argon. A condenser and dropping funnel were connected to the flask which was used to prepare the 20% ass-c (0.05 mol). Hexane (50 ml) was added to the cooled HSS-C to make a slurry. A solution of 4.9 g (0.05 mol) of cyclohexanone in 30 m1 of hexane was added slowly with constant stirring at room temperature, and the mixture was stirred for about 30 min (the reaction is exothermic; if bubbling became too vigorous or the reaction mixture became too hot the flask was cooled briefly in an ice bath). A solution of 7.1 g ( 0.05 mol) of methyl iodide in 30 m1 of hexane was introduced, and the reaction mixture was first stirred at room temperature for 30 min then refluxed for 2 hr. After the reaction mixture was cooled to room temperature, the unreacted sodium was carefully destroyed first with 10 ml of the methanol, then with 10 m1 of water (it is essential to make sure that the unreacted sodium is completely destroyed, or it might catch fire during diltration) . A mixture of solvents (ether : methylene chloride : methanol = 1 : 1 : 1; total of 100 ml) was added to the flask. After further stirring for 15 min, the reaction mixture was filtered, and the charcoal was washed with 30 m1 of the above solvent mixture . The 44 combined filtrates were washed with saturated sodium chloride solution, and the organic layer was separated and dried (1132804) . Removal of solvent gave 5.0 g of the product mixture which was analyzed by VPC (column A, 140', p-diiSOpropylbenzene was used as an internal standard) to give the following product composition: cyclohexanol (10.5%), cyclohexanone (7.5%), 2-methylcyclohexanone (49%) , 2,2-dimethylcyclohexanone (15%) , and 2,6-dimethylcyclohexanone (1.7%). Each of these compounds was collected by preparative VPC (column B, 155°C) and gave satisfactory data (IR,1\MR, and mass spectra) as compared to those in the literature. A sample of 2,2-dimethylcyclohexanone “1’9 collected on the preparative column was found to contain about 10% of 2 ,6-dimethylcyclohexanone 1,1, by both analytical VPC (column B at 65°C, retention time 13 min for 12, 14.6 min for 1,1) and by NMR (for 1,0 singlet at 6 1.08, for 1,1 doublet at 0.96). After distilling the product mixture at reduced pressure (45"5575 mm Hg), there was left in the flask 0.5 g of solid which was recrystalized from petroleum ether and identified as bicyclohexyl-l,1'- diol (10.5%): MP 126'c; IR (c014) 3620 (m), 3570(m), 2950 (s),’l450 (s), 965 (s), 910 (m) and; NMR (CC14) 6 1.0 1.8 (broad); mass spectrum, m/e (rel. intensity) 198 (0.3), 180 (2), 162 (3), 137 (3.7), 99 (100), 81 (47). The results of methylations of cyclohexanone using 10%_HSP—C, 20% HSS-G in hexane, and 20% HSS-G in THF are given in Table 4". 4 . Methylation of Z-Iwiethylcyclohexanone 9 'b The procedure and workup were as described for cyclohexanone. The results are summarized in Table ,1. Yields were obtained by VPC analysis (colunn A, l45'C) using tridecane as an internal standard. Dialkylated 45 product in these cases is 2,2,6—trimethylcyclohexanone 12. When methylation was conducted with HSS—G (20%) and methyl iodide , reductive coupling of ketone 2 gave 2,2'-dimethy1bicyclohexyl-1,l'—diol 1,3 which was isolated in very small amount by distillation. For 12: IR (neat) ‘ 'VV 1 1706 cm- ; NMR (cm 8 0.96 (311, d, J = 7 Hz), 1.0 (3H, s), 1.2 2.25 4) (6H, broad), 2.54 (1H, 111); mass spectrum, m/e 140 (21%, parent). For 13: MP 138~140°; IR (0014) 3620 (w), 3550 (w), 2940 (s), 1465 (m), 1390 (m), 1140 (m), 970 (m), 890 (w) on1 ; NMR (CDClS) 6 1.14 (6H, d, J = 7 Hz), 1.25 2.3 (20H, broad); mass spectrum, m/e (rel. intensity) 226 (1.6), 108 (2.8), 190 (6.6), 175 (3.7), 133 (3.7), 123 (8), 113 (100), 95 (48) . 5. Methylation of Englate Mixture 14’ and 1,5 in the Presence and Absence of Charcoal An enolate mixture 1}» and 145 was prepared from 2.8 g (0.025 mole) of Z-methylcyclohexanone and 0.025 mole of lithium diisopropylamide in 20 m1 of dimethoxyethane (DME) at 0°C according to House's proceduralza Predried charcoal (10 g) was then added and 3.6 g of methyl iodide in 5 ml of IMF. was added all at once. After stirring for 20 min, 20 ml each of NaHCl)3 solution and ether were added. The charcoal was filtered and the filtrate was extracted with ether. The organic layer was then washed successively with 5% HCl, sat. NaHm.5 solution and dried (NaZSO4) . The concentrated reaction mixture was analyzed by VPC (column A, at 145°C) to give the following composition: 9 (20.2%), 10 (29%), 11 (2.9%), 'b ’Vb ’V'b 1’2" (12.2%) . 46 When the same enolate mixture if» and 1,5" (prepared as above) was methylated with methyl iodide in DVIE at 0°C directly, the reaction product analyzed by VPC (colunn A, 14 5° C) was found to have the following composition: 9 (2%), 1}» (57.4%, both cis and trans isomers), k2» (24.4%). 6. Allylation of Z-Methylcyclohexanone The procedure and workup were as described for the methylation of 2- methylcyclohexanone. Thus , Z-methylcyclohexanone (2.8 g, 0.025 mol) reacted with 15% HSS-C (3.88 g, 0.025 mol) and allyl bromide (3.02 g, 0.025 mol) in hexane for 4 hr to yield the following products (determined by VPC, column A, at 145°C): 2-methy1cyclohexanol (11.5%), 2-methylcyclohexanone (11%), 2- allyl-Z-methylcyclohexanone 1,6 (68%). For 12: IR (neat) 1700 (s), 1640 (m), cm'l; NMR (001,) 6 1.0 (3H, s), 1.2-1.95 (6H, broad), 1.95-2.5 (4H, broad), 4.8 (1H, m), 5.0 (1H, m), 5.58 (1H, 111); mass spectrum, m/e (rel. intensity) 152 (19%, parent), SS (100%). 7. Preparation of 2,2 ,6,6-Tetradeuterocyclohexanone ,qu Cyclohexanone (9.8 g, 0.1 mol) was refluxed overnight with 8 g of D O 2 and 1.5 g of potassium carbonate in a lOO-ml flask equipped with a drying tube. Ether (8 ml) was added to the cooled reaction mixture. The organic layer was separated, dried (Na2504) , and concentrated to give 8.6 g of deuterated cyclohexanone which according to NMR had 60% of the four a- hydrogens exchanged. The deuterium exchange react ion was then repeated several times. After the 4th exchange, 92% of deuterium exchange was a achieved and 6 g of the 2,2,6,6-tetradeuterocyclohexanone 59 was obtained. 1 IR (neat) 1700 cm- ; NMR (CC14) 6 1.80 (singlet); mass spectrum, m/e (rel. intensity) 102 (40), 56 (100). 47 8. Methylation of 2,2,6,6-Tetradeuterogclohexanone 2,0 The procedure as described above for cyclohexanone was followed except at the point of workup where D20 and (1130D were used to destroy the unreacted sodium, and tridecane was used as an internal standard for VPC analysis (column A, 140°C). Thus 5.1 g (0.05 mol) of 22 was methylated with 5.75 g of 20% HSS—C (0.05 mol) and 7.1 g'of methyl iodide (0.05 mol) to give the following products: 1,2,2,6,6-pentadeutero- cyclohexanol-O-d 2,1 (6.7%), 22 (12.4%), Z-methyl-Z,6,6-trideutero- cyclohexanone ’22., (35.7%), 2,2-dimethyl-6,6-dideuterocyclohexanone 2,3 (6.3%), 2,6-dimethyl~2-6-dideuterocyclohexanone 2/4 (= 0.7%), 2,2,2',2',6,- 6,6',6'-octadeuterobicyclohexy1-l,1'-diol-O-d2 2’5 (17.3%). For 21: IR 1 (neat) 3350 (s), 2920 (s), 2200 (w), 1450 (m) cm“ ; NMR (c2014) 8 1.0~1.9 (broad); mass spectrum, m/e 106 (parent). For 52: IR (neat) 1703 011-1; NMR (CC14) 6 0.98 (3H, s), l.l7~2.50 (6H, broad); mass spectrum, m/e 1 115 (parent). For 23: IR (neat) 1700 cm" ; NMR (CC14) 6 1.06 (6H, s), 1.68 (6H, broad); mass spectrum, m/e 128 (parent). For 25: MP 123~125°C; IR (cc14) 3620 (w), 3570 (w), 2930 (s), 2200 (w), 2100 (w) cm'1 ; NMR (CC14) 6 1.0-l.9 (broad); mass spectrum, m/e (rel. intensity) 188 (3%, M-ZO), 168 (6%, 111-40), 149 (1.7), 131 (4), 114 (S), 103 (100); high resolution mass spectrum showed 208 (25) , 207 (90), 206 (100), molecular formula (312111le00? When the quenching was done with 1le and ethanol, the alcohol obtained was 1,2,2,6,6-pentadeuterocyclohexanol 26: IR (neat) 3340 (s), 2940 (s), 1 2200 (w), 2100 (w), 1450 (m) cm” ; NMR (0014) 8 1.0~2.0 (broad); mass spectrum, m/e 105 (parent). For recovered ketone 2,2,: mass spectrum 48 showed 69% (10’ 25% d1, 6% d2. For 23: mass spectrum showed 19% d0, 38% d 31% dz, 12% d 1’ 3° 9. Methylation of 4-t-Butylcyclohexanone a 4-t-Butylcyclohexanone (7.5 g, 0.0487 mol) was methylated with 20% HSS-C (5.75 g, 0.05 mol) and methyl iodide (7.0 g, 0.049 mol) in hexane fer 1.5 hr to give 8 g of crude product which was distilled at 60~65'C/0.5 mm.Hg to yield 4.8 g of distillate and 2.0 g of solid residue. The distillate was analyzed by VPC (column B, at 170°C and.column C at lSO'C) to give the fellowing products: 4-t-buty1cyclohexanol 28 (8%), 4-t-butylcyclohexanone 27 (17.2%), cis-2-methyl-4-t-buty1cyclohexanone 32 (32.5%), trans-2-methy1-4-t-butylcyclohexanone 39 (2.4%), 2,6-dimethy1- 4-t-butylcyclohexanone 31 (4%). The solid residue was recrystallized from petroleum ether to give 4,4'-di-t~butylbicyclohexyl-1,1'-diol 32 1 (24.1%). For 28: IR (nujol) 3280 (m) cm' ; NMR (00013) 6 0.82 (9H, s), 0.9 2.2 (10H, broad), 3.4 (1H, broad). For 33: IR (neat) 1710 cm‘l; NMR (CC14) 6 0.89 (9H, s), 1.0 (3H, d, J = 7 Hz), 1.20-2.60 (8H, broad); mass spectrum, m/e 168 (parent). For 32: IR (neat) 1710 011-1; NMR ((1314) 6 0.92 (9H, s), 1.13 (3H, d, J = 7 Hz), 1.25-2.65 (8H, broad); mass spectrum, m/e 168 (parent). For 31: NMR (CDC13) 6 0.90 (9H, s), 1.10 (6H, d, J = 7.5 Hz), 1.25-2.50 (7H, broad); mass spectrum, m/e 182 (parent). For 32: MP 252~253°C (sealed tube, sublimed); IR (nujol) 3500 on), 2930 (s), 1460 (s), 1380 (m), 950 (w) cm'1 ; NMR (CDC13) 6 0.86 (18H, 5), 1.0 2.0 (20H, broad); mass Spectrum, m/e 310 (0.5%, parent), 155 (100%, base). 4.9 10. Methylation of Cycloheptanone “3’4 cycloheptanone (2.8 g, 0.025 mol) was methylated with 20% HSS-C (2.9 g, 0.025 mol) and methyl iodide (3.6 g, 0.025 mol) in hexane overnight to give after distillation (80~90'/20 mm) 2.3 g of distillate and 0.34 g of solid residue. The distillate was analyzed by VPC (column B, 175° C) and found to contain the following compounds: cycloheptanol 3,5 (6.5%), cycloheptanone 34 (21.8%), Z-methylcycloheptanone 36 (45.8%), 2,2-dimethylcycloheptanone 3,7 (3.8%). The solid residue was recrystallized from hexane to give bicyclohepty1-1,1'-diol 38 (12%). For 35: NMR (CC14) 6 2.10-0.90 (13H, broad), 3.65 (1H, broad). For 32: NMR (CC14) 6 1.02 (3H, d, J = 7 Hz), 1.68 (8H, broad), 2.40 (3H, broad). For 37: NMR (CC14) 6 1.02 (6H, s), 1.56 (8H, broad), 2.40 (2H, broad). For 88‘ MP 75.5'c; IR (0014) 3620 (m), 3560 (m) cm"1 ; NMR ((1214) 6 1.53 (broad); mass spectrum, m/e (rel. intensity) 226 (0.36), 208 (2.2), 190 (5.5), 151 (6), 133 (46), 113 (100), 95 (50). ll . Methylation of Cyclooctanone 3,9 Cyclooctanone (3.2 g, 0.025 mol) was methylated with 20% HSS-C (3.14 g, 0.028 mol) and methyl iodide (4.0 g) in hexane for 1.5 hr to give after distillation 3.0 g of distillate and 0.16 g of solid residue. The distillate was analyzed by VPC (column B, at 185’C) and contained the following canpounds: cyclooctanone 32 (36.2%) , 2-methy1cyclooctanone 4’0 (41%), 2,2-dimethy1cyclooctanone 4’1 (14%). The solid residue was recrystallized from‘hexane to give bicycloocty1-1,1'-diol 4,2 (5%) . For ’40: NMR (C014) 6 0.98 (3H, d, J = 7 Hz), 1.1-2.05 (10H, broad), 50 2.05-2.7 (3H, broad). For 4’1: NMR (001,) 8 1.0 (6H, 8), 1.1-2.0 (1011, broad), 2.0-2,5 (211, broad). For 412: MP 86-88‘0, IR (01:14) 3630 (w), 1 3560 (w) cm- ; NMR (CC14) 6 1.3-2.0 (broad); mass spectrum, We (rel. intensity) 254 (1.5), 236 (2.2), 218 (2.2), 165 (6), 127 (100). 12 . Methylat ion of Isqahorone 5,2 ISOphorone (3.5 g, 0.02536 mol) , was methylated with 20% HSS-C (3.1 g, 0.028 mol) and methyl iodide (4.0 g, 0.028 mol) in hexane for 1.5 hr to give after distillation (64-69’C/1.5 11m Hg) 2 g of distillate and 1.33 g of residue. The distillate was found by VPC (column B, 185°C) to have the following compositon: 3,3,5-trimethylcyclohexanone 5,3 (10%) , isophorone 53 (12.5%); Z-methylisophorone 54 (32.1%). The residue was recrystallized from petroleum ether to giv:1,1' ,3,3,3' ,3'-hexamethy1- bicyclohexy1-5-5'-dione 55 (38.3%). For 52: IR (neat) 1715 (s) cm'l; NMR (CC14) 6 0.95 (3H, (1:01] = 6 Hz), 1.02 (6H, s), 1.15-1.75 (3H, broad), l.75-2.4 (4H, broad); mass spectrum, m/e (rel. intensity) 140 (23), 125 (13), 109 (5), 97 (7), 83 (100), 69 (57). For 54:117. (neat) 2950 (s), 1665 (s), 1380 (s), 1320 (s) cm'l; LN (95%, ethanol) 1 max 252 nm'(e 6870); 10112 (C014) 8 0.98 (6H, s), 1.68 (311, s), 1.83 (311, s), 2.07 (411, 5); mass spectrum, m/e (rel. intensity) 152 (30), 137 (S), 109 (8), 96 (100), 83 (7), 68 (27). For 55’: MP 156-157°C; IR (CC14) 2960 (s), 1710 (s), 1460 (w), 1395 (w), 1280 (m) cm'1 ; Til-"IR (CDC13) 6 1.05 (1811, s), 1.35-1.80 (4H, broad), 2.0-2.3 (8H, broad); mass spectrum, m/e (rel. intensity) 278 (0.2), 263 (0.7), 207 (l), 153 (11), 139 (100), 125 (33). 51 13. Methj'lation of Phenylacetonitrile 69 Mb Phenylacetonitrile (5.85 g, 0.025 mol) was methylated with 20% HSS-C (5.75 g, 0.05 mol) and methyl iodide (7.1 g, 0.05 mol) in a solvent mixture of hexane and THF for 2 hr to give 5.6 g of product mixture which was analyzed by VPC (column C, 170°C) and contained the following products: phenylacetonitrile 39 (7.6%) , 2—methy1pheny1acetonitrile a) (74%) , 2,2- dimethylphenylacetonitrile '71 (4.4%). For $9: NMR (CC14) 6 1.6 (3H,. d, J = 7 Hz), 3.75 (111, q, J = 7 Hz), 7.2 (5H, s); mass spectrum, m/e 131 (parent, 31%), 116 (100%). For 311.: NI‘IR (CC14) 6 1.67 (6H, s), 7.2 (SH, broad); mass spectrum, m/e 145 (parent, 27%), 130 (100%). 14 . Reduction of Cyclohexanone with High Surface Sodium The procedure and workup were essentially the same as that described for-the methylation of cyclohexanone, except in these cases no methyl iodide was used. The results are summarized in Table ,6. 15. Reductions of 2-Met_hy£;yclohexanone with I—Iigh Surface Sodium The procedure and workup were as described for the reductions of cyclohexanone and the results are summarized in Table Z. PART II SYNTHESIS AND PIIOTOCHFMISTRY OF SUBSTITUTED Z-CYCLOOCI'ENONES INTRODUCTION In recent decades the photochemistry of cyclic a,B-unsaturated ketones has received considerable attention.253 Particularly interesting is the fact that the reaction course is profbundly influenced by the ring size. The major photochemical reaction of cyclopropenones is the loss of carbon monoxide. For example, irradiation of a 3% solution of di-t-butylcyclopropenone at 2537A. results in the 25b formation_of di-t-butylacetylene; apparently, decarbonylation is sufficiently fast that other processes cannot compete. hv w l I -F 8 Cyclobutenone derivatives open to vinyl ketenes thermally and 26b photochemically, but the stereochemistry of the opening is different in the thermal and photochemical processes.‘ The ketene from perchloro- cyclobutenone has been observed spectroscopically.26a 53 54 hv ‘_ //’ ‘\\/JH_ 1 ' ---—~ 2 “'"i‘Rz R1 (:43 A / V \/R2 R 1 The five and six membered homologs can.undergo intermolecular 27a photoreactions. 2-Cyclopentenone, upon irradiation, leads to the cyclobutane dimers in good yield, while in the presence of excess cyclopentadiene (2 t 2) cross cycloaddition takes place.27b hv ; + 9 ' Q T / Z-Cyclohexenone shows a reactivity qualitatively parallel to that of Z-cyclopentenone in cycloaddition. Certain 4,4-disubstituted 2- cyclohexenones are known to rearrange to the bicyclo[3.1.01hexan- 55 2-ones with group migration. A dipolar species such as 72 has been suggested as an intermediate.28 Although it is well known that the 1) phenyl. migration + 2) 2 ,4 -bondlng ph "ph ph : : ph ph 72 'VV 0 H R\. __ /C-CH3 11v \ R\ _ /R2 /C-C\ \ /.C—-C\ R1 R2 R1 fi—CHS 0 primary photochemical reaction of most acyclic enones is cis-trans isomerization about the double bond, no good evidence has yet been found for the cis-trans isomerization of a cyclopentenone or cyclohexenone. The process is probably not geometrically permissible. 29 29b It has been shown by Eaton aand Corey that both cis-2- cycloheptenone and cis-Z-cyclooctenone can be isomerized photochemically to their trans isomers, and IR spectra of these trans isomers were observed at low temperatures. Although dimerization and cycloaddition reactions do occur, these have been shown unequivocally to be reactions of trans-Z-cycloheptenone and trans-Z-cyclooctenone that occur readily in the dark.30 Thus irradiation of a mixture of cis-Z-cycloheptenone and cyclopentadiene at -50°C afforded a single 1 : 1 adduct which is 56 29b Irradiation of the cis isomers in the presence of formulated as 2,3. piperylene or cyclopentadiene leads neither to sensitized isomerization of piperylene nor to sensitized dimerization of cyclopentadiene, in contrast to the corresponding experiments with five and six membered cyclic enones. The intramolecular cis-trans isomerization of cyclo- heptenone and cyclooctenone, presumably via the triplet state, is very much faster than either intermolecular energy transfer or cyclo- additions. What is notable is that this reaction dominates the photochemistry of medium-ring cycloalkenones and provides exceptionally easy access to compounds of great interest. w 73(95%) mm It is clear that the flexibility of the ring determines the photochemical behavior of simple cycloalkenones. Since the ethylenic portion of the triplet state of an a,B-unsaturated ketone can be treated- in a useful simplification, rather like the triplet state of an unconjugated olefin, the relationship between the reactivity and ring size of cyclic enones can be inferred from the potential energy curves of the twisted olefins (Fig. 3).303’313 The twist angle, 0, indicates the deviation from the normal C (spz) = C (spz) plane. An electronically excited olefin (either S1 or T1) should prefer a rotation about the 57 C (spz) -C (spz) single bond to afford an orthogonal geometry (0 = 90°), thus minimizing the mutual repulsive interactions of the n and 11* electrons. Decay by either internal conversion or intersystem crossing leads to the cis or trans ground state, in which 0 is 0' or 180° respectively . Energy . ‘ . A. 0 90 180 cis trans RD 11 RI 11 R (a R Li R ~"11 ] 11 - 2 2 9 - C (SP )‘C (SP) Figure 55' Energy for the Twisting of Various Electronic States of l , 2 -Disubstituted Olefin . 303 With rather small ring olefins, however, a complete twisting (0 > 90°) is sterically impossible; consequently, their cis-trans isomerization is precluded. In contrast, open chain or sufficiently 58 large membered cyclic olefins undergo the geometrical isomerization readily. -Seven and eight membered cyclic olefins are located intermediate between the two extremes. Obviously, a coplanar trans double bond (0 = 180’) can not be accommodated, but a 90°twisted confermation can be adopted and further twisting to a ground state isomer (90’< 0 <180’) with minimized energy could occur. Such molecules are conventionally referred to as trans-isomers. The eight membered ring is the smallest cycle capable of incorporating a double bond of trans configuration, while trans-cycloheptene and trans- cyclohexene have both been preposed as fleeting intermediates.31b Models suggest that a rigid, planar trans double bond cannot be built into an eightnmembered ring. In accordance with this, trans-cyclooctene has dipole manent of 0.8 D caused by out of plane bending and rehybridization of the strainedlr-bond.323 Since trans-cycloalkenes cannot have a planar structures, it was been suggested that trans cyclic olefins of intermediate size should be capable of existence in stable enantiomorphic confermations (Fig. g). The molecular asymmetry of the trans-cycloalkenes results from.steric barriers to rotation of double bond substituents past the carbon chain. Trans-cyclooctene has been resolved by Cope et a1,33 whereas trans- cyclononene can be resolved only at low temperature and trans-cyclodecene has not yet been resolved. Ho” ----.—-—-- Figure ,4. Enantiomorphic Trans-cyclooctenes Interest in the properties of trans-cycloalkenes led to speculation on a novel and unknown class of bicyclic trans -cycloalkenes wherein the two rings share a common double bond. Such compounds have been named as (a,b)betweenanenes because the double bond is sandwiched between two alkyl chains. It was not until recently that the first known betweenanene, i.e. (10,10)betweenanene was prepared, by Marshall.32a In their elegant synthesis, they have sucessfully introduced a trans double bond bearing two ester substituents in a lZ-membered ring olefin; subsequent reactions led to the (10,10)betweenanene (Eq. 15) . While this work was in progress, another example of a betweenanene was reported by Nakazaki.32b They prepared (10,8)betweenanene by photochemical isomerization of its cis precursor (Eq. 16). Since it is known that median ring cycloalkenones are capable of isomerization to their trans- isomers, it seems likely that if one could photochemically generate a stable trans-cycloalkenone bearing suitable substituents at the trans- double bond, trans -cycloalkenones might be key intermediates leading to the synthesis of betweenanenes . This was one objective of the present research. 60 3002(01234 ) 312%310 @2)10 m————9’-—€> -—4> -——€> d 01a. (Eq. L59 : H : h (”238 (1:112)10 ((112)8 " .... . / . ‘ @5310 (Eq. 16) Table 8 Ir data of cis and trans-Z-Cycloalkenones Z-Cycloalkenones IR cm- -C- cis-cycloheptenone 1669 trans-cycloheptenone 1715 cis-cyclooctenone 1664 trans-cyclooctenone 1715 cis-cyclononenone 1667 trans-cyclononenone 1692 Table 8 compares the IR spectral data of certain cis and trans 'b 303 The degree of conjugation of C = C bond with C = 2-cycloalkenones. group in the seven and eight-membered trans-Z-cycloalkenones is markedly decreased as compared with that of the cis isomers. These trans double 61 bonds suffer enough torsional strain to provide effective strain releasing reactions. Hewever, the IR spectrum shows that the trans- 2-cyclononenone is sterically less strained, and in fact is stable enough to be isolated.303 The photochemical polar addition of alcohols to cycloalkenes (C6-C8) has been extensively studied.31 The reaction has been proposed to proceed by the fellowing sequence: 1. a photochemical cis-trans isomerization of the olefins; 2. protonation of the strained trans- olefins to produce carbocations; 3. nucleOphilic attack by the solvent. For instance irradiation of 1-methy1cycloheptene in a.mixture of ‘methanol and xylene solution affords a 62% yield of an adduct (Eq. 13). These photoadditions are not stereospecific, since mixtures of cis and trans adducts are usually formed13 (Eq. 18). 01 3 (113 0c113 hv \ . 013011 _ (Eq‘ 31.) xylene 62% ..H ,1) CO Co 01 hv ‘D + ‘n H * ROD r——-’ “11 fl (Eq. if?) 01 g ' 62 Similarly, various examples of photochemical polar addition of protic solvents to a,B-unsaturated ketones have been reported. The irradiation of Z-cyclcheptenone and Z-cyclooctenone in various solvents (alcohols, acetic acid, water, diethylamine) results in the formation of Michael-type addition products. The mechanism of these reactions has been suggested to consist of a prior photochemical isomerization to the trans isomer and a subsequent thermal reaction with the 30 It is still not clear, nucleophilic solvents to give the adducts. however, if the addition of protic solvents across the trans double bond is a concerted or a stepwise process. I IN / ”Y > (c1 ' (CH ),,_3 (c112) _3 I)n--3 C 000C“ Y= 007 2, .5 3, 011, Mt Recently, Hart and Dunkelblum3 4 found that irradiation of 2,3-benzo-2,6-cycloheptadienone 14 in.methanol-d give 6-methoxy-2,3- benzo-Z-cycloheptenone 75 in which the methoxyl at C-6 and deuterium at C-7 are trans. The results are rationalized by a photoisomerization of'Zg to the A6’7-trans isomer 4t which thermally adds methanol in a regio- and stereospecific syn manner to give a single adduct. Thus the addition of methanol to the trans double bond may be a concerted' process. In connection with stereochemical studies on the photochemical addition of methanol to 2-cycloa1kenones, llmkelblwn and Hartd'5 have observed a dramatic confermational change between the free and europium 63 coordinated forms of these canpounds when the rings were sufficiently large so that the substrate could act as a bidentate ligand toward the europium. Hence, Eu-complexed '76 has a sufficiently rigid structure that the gem coupling between the H2 protons (12 .Hz) was readily observed. This observation was useful in assigning the stereochemistry of the (1130D adducts . O 4% So far, there have been no studies done on the photochemical addition of methanol to a Z-cycloalkenone with substituents at the C-2 or C-3 positions. It is the purpose of this part of my thesis to further examine the generality of the photo-induced addition of methanol to substituted 2-cyclooctenones and to see if the substituents would have any effect to these photochemical reactions . Compounds that were 64 synthesized for this study are as follows: Z-methyl-Z-cyclooctenone 9,3, 3 -methyl - 2 -cyclooctenone 38 , and 2 , 3 —dimethyl -2 -cyclooctenone 1,0,0 . RESULTS AND DISCUSSION 1. Synthesis of Z-Methyl-Z-cyclooctenone Our initial approach to the synthesis of 2-methy1e2-cyclooctenone was essentially the one used by Whitham and Heap35 for the synthesis of medium ring Z-cycloalkenones. For example, Z-cyclooctenone can be made according to Eq. .19. Allylic bronination of cyclooctene using N-bromosuccinimide gives 1-bromo-2-cyclooctene. Hydrolysis of allylic bromide 47/7 in aqueous acetone buffered with sodium bicarbonate produces allylic alcohol 7‘8”, and oxidation of the allylic alcohol with 6N chromic acid in acetone affords Z-cyclooctenone. Br NaI ICO3 CC14 Acetone Acetone 1.2 [9 (Eq. 12) In order to synthesize Z-methyl-2-cyclooctenone employing the reaction sequence outlined in Eq. '19, we had to start with l-methyl- cyclooctene. The procedure leading to the synthesis of l-methylcyclooctene by Brown36 was followed. Reaction of cyclooctanone with methyl- magnesium iodide and subsequent dehydration of the alcohol 81" with p-toluenesulfonic acid gave 1—methylcyclooctene 82 in 76% yield. ' 65 66 Bromination of 1-methylcyclooctene 3‘2 with N-bromosuccinimide in carbon tetrachloride afforded in 44% yield, an allylic bromide which was identified as l-brononethylcyclooctene. Allylic bromide 84 was characterized through its NMR spectrum. Apparently, allylic bromination of 8% occurred at the side chain instead of in the ring. Therefore this approach to the synthesis of Z-methyl-Z-cyclooctenone, by way of Eq. 1,9, is not feasible. Br . -'H-) 011 . (“3“31 p-TsOH NBS , ._____, ———-—-- 83 EtZO (:c14 M. (100) 0 2 2 o B 80 81 82 2 r M 'Vb ’V‘u ""—, 84 VD 37 that allylic oxidation It has been shown by Dauben and Shaffer of cyclic olefins with chromium trioxide-pyridine complex can give the 01,8-unsaturated ketone in fairly good yield. For example 1- methylcyclohexene can be oxidized with chromium trioxide -pyridine complex in methylene chloride to afford 15% of Z-methyl-Z-cyclohexenone and 60% of 3-methyl-2-cyclohexenone (Eq. 22) . If l-methylcyclohexene is replaced by l-methylcyclooctene, a route similar to Eq. 20’ could be used for the synthesis of substituted-2-cyclooctenones. However, when l-methylcyclooctene was treated with chromium trioxide -pyridine complex in methylene chloride, the desired substituted 2—cyclooctenones 67 were not obtained. Instead, l-methylcyclooctene oxide 82 was isolated along with some recovered starting olefin. 6:: 8 8 :6 d CrO3° Zpy 'Vi V8» CHZCl 2 82 t____,. [:::::::]f5[) + 82 'V» M 85 V» It has been reported by Cohen38a that epoxides can be oxidized by dimethyl sulfoxide to a-hydroxy ketones if catalytic amounts of boron trifluoride are present. For example , when cyclohexene oxide is heated 'with a catalytic amount of boron trifluoride etherate in dimethyl sulf6xide on a steam bath for 22 hr, Z-hydroxycyclohexanone 82 is isolated in 76% yield (Eq. 8;). Thus, oxidation of epoxides by dimethyl sulfoxide to a-hydroxy ketones and.subsequent dehydration of the a-hydroxy ketones might constitute a way of synthesizing Z-cyclooctenones. However, when a solimion of l-methylcyclooctene oxide 88 and a catalytic amount of boron trifluoride etherate in dimethyl sulfbxide was heated at 68 110°C for 20 hr, 2-methylcyclooctanone was isolated as the only major product (Eq . 8’2") . OH ’. BFS-EtZO \ .,O mso (59- a) 82 (766) C81 0 ... (Eq. 2 ) ‘0 111150- «2% 110°C 0 . —» 0 M 89 ”VD It has been demonstrated by Garbisch39 that a convenient and expedient synthesis of many Z-cycloalkenones may be accomplished by monobrominating the cyclic ketone in methanol or ethylene glycol followed by dehydro- bromination and subsequent ketal hydrolysis (Eq. 23) . However, the synthesis of Z-substituted-Z-cycloalkenones from X’substituted cyclic ketones by the same sequence of reactions has not been tried. Although Garbisch also formd that monobromination of dimethyl or ethylene ketals of Z-allqlcyclohexanones occurs with predominant substitution at C-6 , and brominations of Z-alkylcyclohexanones in ether lead to predominant substitution at C-2 (Eq. 83), dehydrobromination of 2~alkyl-2- bromocyclohexanones to Z-alkyl-Z-cyclohexenones has not been effected. 69 Since Z-alkylcycloalkanones can be brominated selectively at the C-2 position in ether, bromination and subsequent dehydrobromination might provide a good route to 2~substituted—2-cycloalkenones. Q/Eii] ——-———9 l i B—————9im iziifl [(iEE)] (593 éé) Cllz)n C112) n=0~3 (113011 . 1f Br2 . .___...: (Eq. 21) EtZO £ Br2 In order to prepare Z-methyl-Z-cyclooctenone by this method, we would need.2-methylcyclooctanone. It has been known that direct alkylation of ketones using conventional bases gives a complex product mixture which is often difficult to separate. Literature search showed that to date the best way to synthesize 2-methylcyclooctanone was that developed by C'onia.4O Cyclooctanone is converted to the trimethyl silyl enol ether 81. Q'clopropanation of 81 by way of an improved Simmons- Smith reaction gives siloxycyclopropane compound 88 and hydrolysis 'V’u of 88 with sodium hydroxide in.methanol afferds Z-methylcyclooctanone 70 (Eq. 25). However, the Simmons-Smith reaction seems to be not always 'VM reproducible and is quite dependent on the quality of the zinc-silver couple. Therefore we decided to look for a simple and better way of synthesiz ing 2 -methylcyclooct anone . OSil~Ie.. OSiMe3 J ClSiMe3 l) GIZIZ+Zn/Ag ————> NEt3 2) pyridine IMF 9 80 87 (87%) ii (84°) M’ M NaOH (Eq. 25) MeOH ”“ 89 (90%) M Recently, Corey413 found that N,N-dimethylhydrazone derivatives (i.e. IMI's) of enolizable aldehydes and ketones can be metallated cleanly by lithium diisopropylamide (LDA) in THF at 0°C, and that these a-lithiated TNH's can serve as equivalents of enolate ions in synthesis (e.g. alkylation). For instance, the IMH of Z-methylcyclohexanone can be methylated to give trans-2,6—dimethylcyclohexanone in 95% yield (Eq. 82). Since Dl-H's of ketones are easily available in quantitative .yield from the corresponding carbonyl compound and N,N-dimethylhydrazine, methylation of cyclooctanone via its IMH derivative may constitute a simple and efficient method of preparing Z—methylcyclooctanone. 71 N—N(CH$) 1) LDA,20 hr, 0°C 3) Nan4 3 2) (1131, 10 IT; I)“ 7 ; (Eq. 3,9) 95% When cyclooctanone 82 was refluxed with N,N-dimethylhydrazine in absolute ethanol overnight, dimethylhydrazone 39 was obtained in quantitative yield. Methylation of the dimethylhydrazone 39 with lithium diisopropylamide and methyl iodide in 'IHF at 0°C gave cleanly the monoalkylation product 31. Hydrolysis of dimethylhydrazone 31’ with 10% sulfuric acid afforded after distillation a 95% yield of Z-methyl- cyclooctanone «8,8. The structure of 82 followed from its spectra. Clearly our overall yield of 2-methy1cyclooctanone starting from cyclooctanone is much better than that obtained using Conia's method, and the procedure is simpler. Therefore, alkylation via the dimethylhydrazone derivative provides a very ‘good method for the a -alkylation of medium ring cyclic ketones. N—N(CH3)2 N7N((113)2 HZNN(CH3)2 1) LDA,'IHF, 0 C ——> EtOH 2) (1131 reflux (= 100%) §,Q 22 (= 100%) 72 Bromination of Z-methylcyclooctanone in ether with bromine gave 2-bromo—2-methylcyclooctanone in quantitative yield. The bromoketone 33 was Characterized by the appearance of a singlet methyl resonance at 6 1.75 in its NMR spectrum and the disappearance of the doublet methyl resonance corresponding to the starting ketone 82. Dehydro- bromination of bromoketone 23 with refluxing methanolic sodium hydroxide yielded only rearranged product which was not identified. After several trials we finally arrived at the proper conditions for the dehydro— bromination. Thus bromoketone 28 was dehydrobrominated with lifllhml carbonate in refluxing dimethyl formamidc to give after column chromatography a 66% yield of 2-methy1-2-cyclooctenone 28. The structure of 88 was deduced from its spectra. The IR spectrum, with a strong band at 1660 cm'1 with a shoulder at 1685 on_1, and the UV maxima at 242 nm (6800) and 285 nm (750) support a conjugated enone structure. Theldfll spectrum showed a doublet at 6 1.76 (J = 1.5 Hz) for the C-2 methyl, a region at 6 1.33-2.0 for six methylenic protons, a broad region at 6 2.0-2.66 for two allylic and two alpha protons, and a triplet of quartets (J = 1.5 and 7 Hz) at 6 5.67—6.07 for the C-3 vinylic proton. The molecular fonmlla C9H140 was confinned by a mass spectrwn (parent peak m/e 138) and elemental analysis. BrZ r lef‘D3 EtZO IMF o A 0°-lo c 89 92 (a 100%) 93 (66%) m '\/\a ’VM 73 2. Synthesis of 3-Methy1- and 2,3-dimethyl-2-cyclooctenones Scheme 3 N Glgflgl CuBrZ _____} _.___.__) r CuCl 0 IC13+EtOAC 93 Et2O 94 A 95 ’\/\J 'VV 1 100 96 ’V\/\; W Our first approach to the synthesis of 2,3-dimethyl-2-cyclooctenone 100 is shown in Scheme 8. 1,4—Afldition of methyhnagneshnn iodide to 2- methyl-Z-cyclooctenone 28 in the presence of cuprous chloride in ether produced an essentially quantitative yield of 2,3-dimethylcyclooctaone 28. Bromination of 2% with either bromine in ether or cupric bromide in refluxing chlorofbnn and ethyl acetate gave bromoketone 28 in quantitative yield. Bromoketone 28 was characterized by a strong carbonyl band at 1710 cm‘l. Its 3002 spectrum showed a methyl doublet at 6 1.28 (J = 7 Hz) and a methyl singlet at 5 1.6. Finally, dehydrobromination of bromoketone 28 with lithium carbonate (or lithium chloride) in refluxing rhfi'afforded only 2-methylene-3-methylcyclooctanone 28 in good yield. The structure of 22 was deduced from its spectra. The IR spectrum showed a carbonyl band at 1690 cm-1 which means that the conjugation is very weak. The NMR spectrum of 22 had a methyl doublet (J = 7 Hz) at 5 1.15, eight methylenic protons at 6 1.2-2.0, one allylic proton and two alpha protons at 74 6 2.0-3.0, and two-proton multiplets at 6 5.05 and 5.6 for the vinyl protons. Apparently the dehydrobromination of bromoketone «9’5! occurred at the side chain instead of in the ring. ——9 —-—> (Eq. 2],) \ H R R DaubenA‘.5 recently developed a procedure by which tertiary allylic alcohols, generated by the 1,2-addition of an organometallic reagent to an a,B-1msaturated ketone, are converted with pyridinium chlorochromate (PCC) in a single step to a new, transposed B-alkyl-a,B-unsaturated ketones in good to excellent yield (Eq. a). This method for allxylative 1,3- carbonyl transposition provided us a simple and effective way of synthesizing the desired CYclooctenones 2’5: and 122. Thus, 3-methyl and 2,3—diinethyl- 2-cyclooctenones were synthesized according to Scheme i. Scheme 4 N OH 013L1 PCC E cu C1 Et20 2 2 9 9. 2’9 91 (82.) 9’53 (40 ) o n o (31131.1 pcc Et 0 ’ (n 01’ + 2 2 2 93 9’9 (52%) we (30%) m (27%) 75 2-Cyclooctenone 12 was prepared as reported by Whitham.35 1,2-Addition of methyllithium to 2-cyclooctenone produced l-methylcyclooct- 2-en~1-ol (82%) which was oxidized with pyridinium chlorochromate in methylene chloride to give S-methyl-Z-cyclooctenone gfi in 40% yield after purification. Compound g§ was assigned the structure shown on the basis of the following spectral data. The molecular fonmula C9H14O was confirnwd by the high resolution mass spectrum (parent peak m/e 138). The IR absorption at 1650 cm-1 showed that there is conjugation between the carbonyl group and the carbon-carbon double bond. The UV spectnmni showed a maximum at 245 nm (e 7500) which indicates that compound 2% is an a,B;unsaturated ketone." Its NMR spectrum showed six methylenic protons at 6 l.3-2.1, a vinyl methyl doublet at 6 1.84 with a coupling constant of 1.5 Hz, a four proton broad region at 6 2.2—2.8 and a vinyl proton at 6 5.87. Treatment of 2-methy1-2—cyclooctenone 33 with methyllithium in ether afforded l,2-dimethylcyclooct-2-en—l-ol 83 in 52% yield. Oxidation of alcohol 33 with PCC in methylene chloride gave a 30% yield of 2,3- dimethyl-Z-cyclooctenone 189 and 27% yield of epoxyketone 18%. Compound $99 was assigned the structure shown on the basis of the following spectral properties. The molecular formula C10H160 was confirmed by the high resolution mass spectrum (parent peak m/e 152). Its IR spectrum showed a strong absorption at 1685 cm-1 with a shoulder at 1650 cm-1. The UV spectrum had maxima at 250 nm (e 4360) and 206 nm (890) indicative of conjugation. Its bbfll spectrum showed a slighten broadened singlet at 6 1.7 corresponding to two vinyl methyls, a six-methylenic proton resonance at 6 1.2-1.9, a fbur-proton resonance at 6 2.0-2.6 and no vinyl protons. 76 3. Photochemical Addition of Ibthanol to 2-Methyl-2-cyclooctenone 93 The photolysis of Z-methyl-Z-cyclooctenone 22 in methanol through Pyrex was followed by analytic VPC. As the reaction proceeded, the peak due to 23 began to diminish while that corresponding to the major product 1825 began to rise. The reaction was essentially complete with 90% conversion in 4 hr. 0 0 II (igon , 2 -—> hV ' H3 3 93 102c OC”3 M W The product 3-methoxy-2-methylcyclooctanone 1355 was assigned the structure shown on the basis of its spectral properties. The molecular fonnula C10H180 was confirmed by a mass spectrum (parent peak m/e 170) and elemental analysis. In the IR spectrwn, absorption at 1705 cm'1 indicates that 1385 must be a saturated cyclic ketone. The NMR spectrum (60 MHz) showed a methyl doublet at 6 0.98 (J = 7 Hz), a methyl singlet at 6 3.27 corresponding to the methoxyl group, a methine proton at 6 3.7-4.1, eight methylenic protons at 6 1.1—2.1 and three a-protons at 6 2.1-3.0. To obtain stereochmnical infonnation about 102c, a 1802dlz proton 'VVVMWD spectrum was taken in CDC13. In this spectrum, the C-2 methyl appeared at 6 1.09 as a doublet (J = 7 Hz), Hz was a quartet of doublets at 5 2.90 (J = 7 Hz and 3.5 Hz) which indicates that H2 was split by the C-2 methyl into a quartet and each of these quartets was split further by 77 H3 into a doublet. H3 appeared as a multiplet at 5 4.04. Irradiation at 6 1.09 (C-Z methyl) caused the quartet of doublets at 6 2.90 to become a doublet with J = 3.5 Hz, and irradiation at 6 1.52 (C-4 methylene) caused the multiplet at 6 4.04 to become a doublet with J = 3.5 Hz. Thus the coupling constant between H2 and H3 is 3.5 Hz and H2, H3 are assigned as cis to each other based on the premise that Jtrans > 44 Jcis' To obtain further support for this assignment, we studied the LIS spectra of compound 1225 with Eu (fod)3. The resulting chemical shift differences for compound 1228 are summarized in Table 3. Table 9 Lis Shift Data for Eu-complexed 102c N W A (ppm) 11 (n OGI J (Hz) H2 ,H3 2 3 3 3 w 102c ’VVW 10.4 7.4 8.7 4.4 8.2 3.0 As we can see from Table 2 that there was only a very slight change in Hz-H3 coupling constant when shift reagent was added. Therefore, there would be only a small change in dihedral angles for HZ—H3 from free 1825 to the Eu-complexed 1825. The coupling constant JIi H3 of 3 Hz 45 The fact was reported for Eu-complexed 19 by Hart and Ennkelblum. that C-Z methyl has a large A value suggests that C-2 methyl is cis to the C-3 methoxyl. Also, the greater A value for H2 than for the w protons would be consistent with having europium coordinated between the carbonyl oxygen and the methoxyl oxygen, above the C-2 position. The stereochemistry of photoadduct 1228 is thus certain with methyl and 78 methoxyl cis to each other, and there is no appreciable confonnational change in free 122g and Eu-complexed 102C. Although the photochemical addition of methanol (and other protic solvents) to cyclic enones has been suggested to proceed through a photochemical isomerization to the trans isomer which then adds methanol (or other protic solvent) thennally to give the products, little is known about the stereochemistry of the addition of methanol to the trans intermediate. Our stereochemical findings outlined above seem best accommodated by a mechanism as shown in Scheme 5; that is, compound 32 photoisomerizes to its trans isomer 2§£ which then adds methanol in a concerted, regio- and stereospecific syn manner to give adduct 1255. The addition process deserves some consideration. A conventional Michael type addition mechanism is ruled out on the basis of the following observations. Treatment 0f.8§ with 0.01 N methanolic sodium methoxide solution for 8 days gave a mixture of 93, 102t.and 102c in 45 : 35 : 20 'V\4’\NV\« WW ratio (only very small amount of lOZt and 102c were formed in a shorter WW 'VVW reaction time). Interestingly, when the photoadductllgge was treated with 0.01 N methanolic sodium methoxide solution for 1 day, the same mixture of,gé,,lgge, and $925 was obtained. The stereochanistry of 1925 (C-2 methyl and methoxyl groups are trans to each other) was confirmed from its I002 spectrum and europium shift study (see experimental section). These results indicate that the rate of base-catalyzed Michael addition ofxnethanol to %Z is slower then that of the base-catalyzed exchange 2C, and that both the Michael addition and exchange reaction of 10° WW reactions are stepwise processes involving the smne intermediate (i.e. 79 carbanion). It is apparent that in the Michael type addition the enolate is not protonated stereospecifically. Scheme 5 'b 01011 h --CH 3 ._\’_—) / —--——-? 22 as NaC)CH3 OISOH O H NaOCH3 CH3 ~-CH +——-— OCH «93 + or; + W 013% 3 102. was, MM (45 35 : 20) O 6 o m 9 0 CH - . CH3 ———> ——> H 5+ H I \ H H 6) CH3 0 3 93t $94 we; § 80 The carbon-carbon double bond in trans intermediate 93% is highly polarized by the electron-withdrawing carbonyl group, although it is not appreciably conjugated. Consequently, if methanol addition is stepwise, the intermediate carbanion would be fonned with charge localized on the u-carbon (in lgg), and a proton must be transferred to it before any conformational change necessary for charge delocalization can occur, to account for the formation of only the trans adduct. In the concerted addition process, methanol adds to the carbon-carbon double bond in trans intermediate 23% probably via a four center transition state (i.e. syn addition in 103) to give the trans adduct 102c. ’VW MN» 4. Photochemistry of 3-Methyl and 2,3-dimethyl-2-cyclooctenones in Methanol The irradiation of 3-methyl-2-cyclooctenone 2§ in methanol through Pyrex with a Hanovia 450 W lamp afforded 3-methoxyl-3-methylcyclooctanone 125 in 72% yield after 5 hr. Compound $23 was assigned the structure shown on the basis of its spectral properties. It has a parent peak at m/e 170 and a peak at 138 corresponding to the loss of a methanol molecule in mass spectrum. The IR absorption at 1700 cm‘1 showed that there is no conjugation within the molecule. The NMR spectrum showed two methyl singlets at 6 1.16 and 3.12, two broad regions at 6 1.2-2.0 (8H) and 2.2-2.4 (2H), two doublets at 5 2.24 (1H) and 2.7 (1H) respectively with a coupling constant of 11 Hz. LIS spectra of compound $05 with Eu (fod)3 were also studied and the results are shown in Table 12. O O Q ——"hv God 98 105 - mmm 81 Table 12 LIS Shift Data for 105 W ACMm) l I! H or our m J ' 0h) 2c 2t 3 3 HZtHZC 105 17.4 12.3 6.9 11.6 8.3 12 The gem coupling constand J”2tIlzc is essentially the same in both the free and complexed 125, indicating that there is no appreciable confonnational difference between these two fonns of compound 125. Free 12% already has a sufficiently rigid structure with the C-3 methyl in an equatorial position that the gem coupling between the C-2 protons (J = 11 Hz) is readily observed. The mechanism for the formation of 105 would be the same as that described for 122;. W When 2,3-dimethyl-2-cyclooctenone 122 was irradiated in methanol through Pyrex with a Hanovia 450 W lamp for 8 hr, only starting material was recovered. Prolonged irradiation did not give a methanol adduct and only resulted in more polymeric material. A possible explanation for this result is as follows. Since intramolecular cis-trans isomerization is considered to proceed vis the triplet state, the presence of methyl groups at both C-2 and C-3 positions would hinder the fonnation of trans isomer £92} from its triplet state due to the presence of steric crowing during the process of rotation around the C2-C3 single bond in the triplet .state. In both 93 and 9%, there is still one way open for the rotation about the C2-C3 single bond to occur readily to form the trans isomer. Thus the reason that no methanol adduct was found in the photolysis of £22,i5 probably because no trans isomer was formed during irradiation. 82 = r=d§ «1’99 triplet state m 5 . Suxmnary A method was found to prepare 2-methylcyclooctanone in quantitative yield, and Z-methyl-Z-cyclooctenone 9;: was synthesized by bromination- dehydrobromination of 2-methylcyclooctanone in good yield. 3-Methy1- and 2,3-dimethyl-2-cyclooctenones 99 and 999 were prepared by the oxidation (with PCC) of the tertiary allylic alcohols generated by the 1,2—addition of methyllithium to 2-cyclooctenone ’99 and Z-methyl-Z-cyclooctenone 33 respectively. Irradiation of 2-methyl- and 3-methyl-2-cyclooctenones ’99 and 99 in methanol resulted in the formation of adducts 9999 and 995, whereas no methanol adduct was found for the photolysis of 2,3-dimethyl- 2—cyclooctenone 999. The addition of methanol to the trans intermediate of 9;) is stereospecific. EXPERIMENTAL 1. Gas Chromatography VPC columns that were used in this section are as follows: 5'9x 0.25 in column, 6% SE30 on chromosorb W. 5' x 0.25 in column, 10% SE30 on chromosorb w, 5' x 0.125 in column, 5% SE30 on chromosorb G. 5' x 0.25 in column, 10% FFAP on chromosorb W. F'PPE"? 5' x 0.125 in column, 5% FFAP on chomosorb W. 2. Preparation of l-Methylcyclooctene The procedure of Brown36 was employed. A solution of 7.56 g (0.06 mol) of cyclooctanone in 70 ml of anhydrous ether was added slowly to a methylmagnesium iodide solution (prepared from 4.86 g of magnesium turnings in ZO-ml ether and 28.4 g of methyl iodide in 30-ml ether) under nitrogen over about 20 min, followed by refluxing for 1.5 hr. The reaction mixture was then quenched carefully with 20 m1 of saturated mnnonium chloride solution, and 30 m1 of ether was added. The ether layer was separated, washed with saturated sodium chloride solution, and dried over NaZSO4. Removal of the ether gave a light yellowish liquid which was the crude l-methylcyclooctanol. An IR spectrum of the crude alcohol showed the disappearance of the carbonyl group and the presence of strong -OH absorption at 3440 cm'l. 83 84 The crude alcohol 9% was refluxed.with 2 g of p-toluenesulfonic acid in 100 ml of benzene for 2 hr using a Dean-Stark trap for continuous removal of water. The reaction mixture was cooled, and 40 ml each of ether and dilute NaCl solution was added. The organic layer was separated, washed with sat. NaZCO3 and NaCl solutions, and dried over Na2504. Removal of the solvent with an evaporator gave, after distillation (62°/25 mm), 5.7 g of pure 1-methylcyclooctene 9% (76% yield based on starting ketone): 1001 (CC14) 6 1.44 (8H, s), 1.64 (3H, d, J = 1.5 Hz), 2.08 (4H, broad), 5.2 (1H, broad triplet, J = 7 Hz). 3. Bromination of l-methylgyclooctene 9% AInixture of 5.6 g (0.045 mol) of l-methylcyclooctene 82, 9.1 g of N-bromosuccinimide and 0.04 g of benzoyl peroxide in 50 ml of CCl4 was refluxed under nitrOgen. The reaction was essentially complete in 20 min, with succinimide floating on the top. Succinimide was removed by filtration, the filtrate was washed with 20 m1 of 5% NaHCOS, 40 ml of water and dried over Na2504. The solvent was removed to give a crude product which upon distillation (93-960/10 mm Hg) afforded 4.0 g of l- bromomethylcyclooctene 84 (44% yield): IR (neat) 2950 (s), 1690 (w), 1655 (M0, 1480 (s), 1460 (s), 1210 (s) cm-l; NMR (CC14) 5 1.2-1.8 (8H, broad), 1.8-2.4 (4H, broad), 3.83 (2H, s), 5.7 (1H, t, J = 8 Hz). 4. Allylic Oxidation of l-Methylcyclooctene The procedure of Shaffer37 for the allylic oxidation of olefins was used. Thus 7.4 g (0.06 mol) of l-methylcyclooctene was oxidized with 85 diromimntrioxide-pyridine comples (80 g of CrO3 and 126.4 g of pyridine) in 800 m1 of methylene chloride. The reaction mixture after workup and column chromatography (silica gel, ether : hexane = l : 1) gave 1.2 g of starting olefin and 2.4 g (29%) of 1-methylcyclooctene oxide 99:38b IR (neat) 2920 (s), 1470 (m), 1450 (m), 1380 On), 910 on), 830 on) cm‘l; NMR (CC14) 6 1.22 (3H, s), 1.24-2‘.3 (12H, broad), 2.3-2.7 (1H, broad); mass spectra, m/e 140 (parent). 5. Acid-catalyzed Rearrangement of l-Methylcyclooctene Oxide 9‘5. A solution of 0.23 g (0.0016 mol) of epoxide 33 and 0.05 ml of boron- trifluoride etherate in 10-ml INSO was heated at llO'C for 20 hr. The reaction mixture was poured into ice water and extracted with chloroform. The combined organic layers were washed with sat. NaHOO3 and NaCl solutions, and dried over MgSO4. Evaporation of solvent gave a brown liquid. Preparative VPC (column B, 135'C) gave Z-methylcyclooctanone as the major product which had the correct spectra data (see page 86) . 6. Preparation of cyclooctanone-N,N—dimethylhydrazone 9p A mixture of 50.4 g of cyclooctanone (0.4 mol) and 72 g of anhydrous N,N-dimethylhydrazine (1.2 mol) in 100-ml absolute ethanol was refluxed overnight. Excess dimethylhydrazine and ethanol were removed by distillation at aspirator pressure (25 mm Hg). The crude product was then distilled at 55°/0.5 mm Hg to give 67.6 g of pure 9,9 as a colorless liquid (quantitative yield): IR (neat) 2960 (s), 1630 (m, C = 'N-), 1475 (m), 1455 (m), 1030 (m), 980 (m) cm-l; NMR (CC14) 6 1.2-2.0 (1011, broad), 86 2.26 (6H, s, dimethyl), 2.0-2.5 (4H, broad, allylic). 7. Methylation of Cyclooctanone-N,N-dimethylhydrazone 9,9 A solution of 67.2 g of cyclooctanone-N,N-dimethylhydrazone (0.4 mol) in SO-ml 'I'HF was added slowly to a lithium diisopropylamide (LDA) solution (prepared from 63 m1 of diisopropylamine (0.45 mol) and 281 m1 of 1.6 N n-butyllithium (0.45 mol) in lOO-ml THF) at 0°C under nitrogen, resulting in a milky—pink solution. After the mixture was stirred for 4 hr, 59 g of methyl iodide (0.41 mol) in 50 ml of THF was added slowly at 0°C, and the reaction mixture was stirred overnight. Upon workup, 200 m1 of water was added dropwise to destroy the unreacted diisopropyl amide. The organic layer was separated, and the aqueous layer was extracted with 200 ml of methylene chloride. The combined organic solutions were washed with sat. NaCl, and concentrated to give 92 g of crude Z-methyl- cyclooctanone-N,N-dimethylhydrazone 91 which was used directly for hydrolysis without further distillation (quantitative yield). Data for 91: IR (neat) 2960 (s), 1632 (m, C = N-), 1475 (m), 1455 (m), 1380 (w), 1030 (m), 970 (m) on‘l; NMR (cc14) 5 0.93 (3H, d, J = 7 Hz, c-z methyl), 1.1-2.0 (10H, broad), 2.0-2.4 (3H, broad, allylic), 2.25 (6H, s, -N(CHS)2). 8. Hydrolysis of Z-Methylcyclooctanone-N,N-dimethylhydrazone 9’1 A mixture of 92 g of crude 2-methylcyclooctanone-N,N-dimethylhydrazone 9,1, and 300 ml 10% sulfuric acid was refluxed overnight. The solution was cooled and the organic layer separated. The aqueous layer was extracted with 200 ml ether. The combined organic layers were washed with saturated NaHCI)3 and NaCl solution, and dried over Na 504. Removal of the solvent 2L 87 gave 74.5 g of crude 89 which upon distillation (at 38°/0.4 mm Hg or 720/51nn Hg) afforded 53.3 g of pure Z-methylcyclooctanone 89 (95% yield) as a colorless liquid: IR (neat) 2950 (s), 1698 (s), 1470 (s), 1455 (m), 1380 (w) cm'l; NMR (CC14) 6 0.98 (3H, d, J = 7 Hz, methyl), l.l-2.05 (1m1, broad), 2.05~2.7 (3H, broad). 9. Bromination of Z-Methylcyclooctanone 89 'W A solution of 19 g of Z—methylcyclooctanone (0.136 mol) in 100 ml of ether was placed in a 200401 flask (equipped with a magnetic stirrer, thermometer, dropping funnel) and cooled in an ice bath. Bromine (22.4 g, 7.5 ml, 0.14 mol) was added dropwise. At the end of the addition, a faint bromine color persisted in the solution for several minutes (if not, a little additional bromine was added). After the reaction mixture was stirred for another 15 min, 20 ml each of sat. NaHCO3 and NaZSO solution 3 were added to remove the excess bromine and HBr produced. The reaction mixture was then extracted with 200 ml of ether. The organic layer was washed with saturated NaHCO3 and NaCl solution, and dried over Na2S04. Removal of the solvent gave 37.3 g (quantitative yield) of crude 2- bromo-Z-methylcyclooctanone 9%. Since the NMR spectrum of the crude product showed the disappearance of the doublet corresponding to methyl protons in 89 and the apperance of a singlet at 6 1.75 (C-Z methyl), the crude product was used for dehydrobromination without purification. 88 10. Dehydrobromination of 2-Bromo-2-methy1cyclooctanone 912‘ A solution of 37.3 g (0.17 mol) of crude 2-bromo-2-methylcyclooctanone 9’2 and 70 g of lithium carbonate in 300 m1 of dry IMF was refluxed overnight under nitrogen. After the mixture was cooled, 100 m1 of water was added. The reaction mixture was then extracted with 300 m1 of ether, and the organic layer was dried (NaZSO4) and concentrated to give crude product which upon distillation at 73°/2.75 mm Hg yielded 14 g distillate and 10 g residue. Column chromatography of the distillate on silica gel (chloroform / hexane = 3 : 1) gave 12.4 g of pure 2-methyl-2-cyclooctenon 93 (66%) as a yellowish liquid: IR (neat) 2940 (s), 1685 (m), 1660(5), 1450 On), 1380 On) cm'l; UV (95% ethanol) Am 242 nm (e 6800), 285 nm ax (750); NMR (CC14) 6 1.76 (3H, d, J = 1.5 Hz C-Z methyl), 1.33—2.0 (61, broad), 2.0-2.66 (4H, broad), 5.67-6.07 (1H, triplet of quartets, J = 7 Hz, 1.5 Hz, vinyl proton); mass spectrum, m/e (rel. intensity) 138 (parent, 18), 123 (3), 110 (14), 96 (35), 95 (100), 81 (40), 67 (72), 54 (41). A231;_Calcd. for C9H14O: C 78.21; H 10.21. Found : C 78.06; H 10.10. 11. Preparation of 2,3-Dimethylcyclooctanone 9% A solution of 0.108 g (0.7 mmol) of 93 in S-ml ether was added to a methylmagnesium iodide solution (prepared from 0.05 g of magnesium and 0.284 g of methyl iodide in 20 ml ether) with 0.05 g of cuprous chloride added as catalyst. After addition, the reaction mixture was stirred for 30 min and refluxed fer 20 min. Wbrkup involved addition of aqueous NH4C1 solution and extraction with ether. The ether layer was separated, dried 89 (NaZSO4) and concentrated to give 0.1 g of crude product. Preparative VPC (column A, 155°C) gave 2,3-dimethylcyclooctanone 9% as the major product (quantitative yield). The crude product was used for bromination directly without further purification. For 9%: IR (neat) 2950 (s), 1692 (s), 1385 (m) cm'l; NMR (c014) a 0.94 (311, d, J = 7 Hz), 1.00 (3H, d, J = 7 Hz), l.l-2.0 (9H, broad), 2.0-3.0 (3H, broad); mass spectrum, m/e (rel. intensity) 154 (30), 139 (9), 125 (27), 112 (37), 93 (90), 83 (100). 12. Bromination of 2,3-Dimethy1cyclooctanone 94 'VD A solution of 0.694 g (4.5 mmol) of 9% in 50 m1 of a mixture of chlorofonn and ethyl acetate was refluxed with 2.1 g (9 mmol) of cupric bromide for 3 hr. The reaction mixture became dark green with white cuprous bromide suspending in the solution. The cuprous bromide was filtered and the filtrate was washed with saturated NaHCO3 and NaCl solution, and dried over NaZSO4. Removal of the solvent gave 1.2 g of crude 2-bromo-2,3-dimethylcyclooctanone 99 (quantitative yield). The crude bromo ketone 99 was used for dehydrobromination directly. Bromination of 93 with bromine in ether also gave the same result. For 99: IR (neat) 1710 cm'l; mm (CC14) a 1.28 (3H, d, J; 7 Hz), 1.6 (3H, s), 0.7—2.0 (8H, broad), 2.0-3.1 (3H, broad). 13. Dehydrobromination of 2-Bromo-2,3-dimethylcyclooctanone 92 The procedure as described for the dehydrobromination of bromoketone 9% was followed. The crude bromoketone 92 (1.2 g) was dehydrobrominated with 1.0 g of lithium carbonate and 0.5 g lithium chloride in 15 ml of 9f) UMP to give 0.3 g (44%) of the product which was identified as 2-methylene- 3-methylcyclooctanone 99: IR (neat) 2940 (s), 1690 (s), 1615 (m), 1470 On) cm-l; 1011 (CC14) 6 1.15 (3H, d, J = 7 Hz), 1.2-2.0 (8H, broad), 2.0-3.0 (3H, broad), 5.05 (1H, m), 5,6 (1H, m). 14. 1,2-Addition of Methyllithium to 2-Cyclooctenone 99 To a stirred solution of 2-cyclooctenone (2.5 g, 0.02 mol) in 50 m1 of anhydrous ether at roan temperature under nitrogen was added, dropwise, an ethereal solution of methyllithium (14 ml, 1.7 M ethereal solution). The resulting solution was stirred for 2 hr, refluxed 1 hr, and quenched by the dropwise addition of 20 ml water. The phases were separated and the aqueous layer was extracted with two 20—ml portions of ether. The combined organic layers were washed with 20 ml of saturated NaCl solution and dried over NaZSO4. The solvent was removed at reduced pressure to give a yellow liquid which after column chromatrography (neutral alumina, hexane : ether = 1 : 1) afforded 2.3 g of l-methylcyclooct-2-en-l-ol gz (82%) and 0.4 g of starting material (16%). Pure samples 0f.31 were collected via VPC (column D, 175°C) for identification: IR (neat) 3370 (s), 2940 (s), 1450 (m), 1375 (m), 1240 (w), 1040 On), 920 (m), 895 On), 805 (m), 735 (m) cm'l; NMR (cc14) 6 1.08 (111, s, -on), 1.23 (311, s, c-1 methyl), 1.3-2.0 (8H, broad), 2.0-2.9 (2H, broad), 5.4 (2H, m, vinyl proton); mass spectrum, m/e (rel. intensity) 140 (2.5), 122 (47), 107 (40), 93 (8), 81 (39), 79 (100), 77 (34), 67 (31). 91 15. 1,2-Addition of Methyllithium to Z-Methyl-Z-cyclooctenone 99 The procedure and workup were the same as that described above for 99. Thus 2 g of 2-methyl-2-cyclooctenone (0.015 mol) in 50 m1 ether reated with 10 ml of methyllithium to yield after column chromatography 1.2 g of 1,2-dimethy1cyclooct-2-en-l-ol 99 (52%) and 0.7 g of starting material (35%). For 99: IR (neat) 3410 (s), 2930 (s), 1450 (s), 1375 (m), 1035 On), 950 (m), 885 On), 850 On) cm-l; 0002 (CC14) 6 1.06 (1H, s, 411), 1.23 (3H, s, C-1 methyl), 1.76 (3H, s, C—2 methyl), 1.2-2.6 (10H, broad), 5.12 (1H, m, vinyl proton); mass spectrum, m/e (rel. intensity) 154 (7), 136 (70), 121 (49), 107 (85), 93 (100), 91 (44), 79 (84), 67 (53). 16. Oxidation of l-lbthylcyclooct—2-en-1-ol 99 To a magnetically stirred slurry of pyridinium chlorochromate (PCC, 8.6 g, 0.04 mol) in 60 ml of dichloromethane was added in one portion a solution of 99 (2.3 g, 0.0164 mol) in 10 ml of dichloromethane at room temperature. The resulting dark red-black mixture was allowed to stir for 3 hr, and was diluted with 30 m1 of ether. The ethereal solution was decanted from the black resinous polymer, which in turn was washed with three 20-ml portions of ether. The combined ethereal phases were washed successively with two lOO-ml portions of 10% aqueous NaGI, 100 m1 of 10% aqueous HCl, and two 50-m1 portions of saturated NaHCO , and dried over MgSO Removal of the solvent gave a yellow liquid which after chromato- 4. graphy (silica gel, hexane : ether = l : 1) yielded 0.9 g (40%) of 3- methyl-Z-cyclooctenone 99. Pure samples of 99 were collected via VPC 92 (column 0, at 165°C) for spectral data: IR (neat) 2940 (s), 1650 (s), 1450 (m), 1380 On), 1340 on), 1265 (m), 1145 (w), 1040 (w), 885 (w), 845 (w) cm'l; UV (95% EtOH) "max 245 nm (e 7500); mm ((7014) 6 1.35-2.1 (611, broad), 1.84 (3H, d, J = 1.5 Hz, C-3 methyl), 2.2-2.8 (4H, broad), 5.78 (1H, broad); mass spectrum, m/e (rel. intensity) 138 (20), 123 (7), 109 (6), 95 (100), 82 (34), 67 (34), 55 (18), 41 (23). High resolution mass spectrum mol wt. 138.10384 (calcd. for C O, 138.10446). 9”14 17. Oxidation of l,Z-Dimethylgyclooct—2—en-l-ol 99 The procedure described above for the oxidation of 99 was followed. A solution of 1.2 g of 99 (0.0078 mol) in 10 m1 of methylene chloride was oxidized with PCC (4.2 g, 0.0195 mol) in 40 m1 of methylene chloride to give after chromatography (neutral alumina, hexane : ether = 1 : 1) 0.36 g (30%) of 2,3-dimethyl-2-cyclooctenone 999 and 0.36 g (27%) of 2,3- dimethyl-Z,3-epoxycyclooctanone 999. Compounds 999 and 999 were also separable by VPC (column D, at 165°C). For compound 999:. IR (neat) 2930 (s), 1685 (s), 1650 (m), 1450 (m), 1380 (w), 1275 (w) cm—l; UV (95% EtOH) )‘max 206 nm (e 890), 250 (4360); NMR (CC14) 6 1.7 (6H, s, C-2 and C-3 methyls), 1.2-1.9 (61, broad), 2.0-2.6 (4H, broad); mass spectrum, m/e (rel. intensity) 152 (29), 147 (8), 137 (11), 123 (8), 109 (100), 96 (22), 81 (42), 67 (32). High resolution mass spectrum mol wt. 152.12029 (calcd. for C10“ 0, 152.12012). For compound 99%: IR (neat) 2940 (s), 16 1715 (s), 1465 (s), 1390 (m), 1110 (m), 1075 (m), 895 On), 850 On), 830 (m) cm-l; 0002 (CC14) 6 1.3 (3H, s, C-3 methyl), 1.43 (3H, s, C-2 methyl), 1.0-2.0 (8H, broad), 2.1-2.8 (2H, broad); mass spectrum, m/e (rel. intensity) 168 (11), 140 (10), 126 (83), 125 (100), 112 (22), 111 (62), 98 (23), 93 97 (63), 85 (39), 84 (57). 18. Irradiation of Z-Methyl-Z-cyclooctenone 93 in Methanol A solution of 0. 139 g (1 mmol) of 93 in 10 m1 of methanol was placed in a Pyrex tube, flushed with nitrogen and irradiated externally using a 450 W Hanovia Mercury lamp at room temperature. The photolysis was followed by analytic VPC (column C, at 165°C). As the reaction proceeded, the peak with a retention time of 0.6 min (starting material) decreased in area and a major product peak appeared at 1.3 min. After 4 hr of irradiation, the conversion was about 90% (a longer reaction time gave more of other undesired products) and the major product 102c was formed in 80% yield (55% isolated yield). 3-1bthoxy—2-methy1cyclo- octanone 1025 was collected by preparative VPC (column A, at 165 C) and examined: IR (neat) 2940 (s), 1705 (s), 1470 (s), 1450 (s), 1384 On), 1330 On), 1095 (s), 950 On), 910 (m) cm-1; 1101 (CC14) 6 0.98 (3H, d, J = 7 Hz, C-Z methyl), 1.1-2.1 (8“, bread), 2.1—3.0 (3H, broad), 3.27 (3H, s, methoxyl), 3.7-4.1 (1H, broad, C-3 methine); mass spectrum, m/e (rel. intensity) 170 (6), 138 (24), 123 (7), 109 (25), 98 (25), 81 (26), 71 (100), 57 (71). Anal; Calcd. for C H O ’10 18 2‘ Found : c 70.57; H 10.57. C 70.54; H 10.66. 19. Europium Shift Study of 3-Methoxy-2-methy1cyclooctanone[192$ LIS spectra were recorded by gradually adding weighed amounts of Eu (fod)3 (Aldrich Chemical Company) to 14 mg of 1923 in CC14. The LIS 94 chemical shifts were plotted against the weight of Eu (fod)z and A is the extrapolated value of the chemical shift difference in p.p.m. for a molar ratio 1 : 1 of shift reagent : substrate. A's: 4.4 (0013), 7.4 (H3), 8.7 (013), 10.4 012), 8.2 (w proton). 20. Decoupling Earp, eriment of 102c and‘IZu—complexed 1192’s Proton spectra of 02c were measured in CDCl3 with 1118 as an internal standard using a 180 Mlz Bruker spectrometer: 1.09 (3H, d, J = 7 Hz, C-Z methyl), 1.25-2.07 (8H, broad, C4-C7 methylenes), 2.24-2.76 (2H, m, C-8 methylene), 2.90 (1H, quartet of doublets, J = 7 Hz, 3.5 112, C-2 methine), 3.4 (3H, s, methoxyl), 4.04 (1H, m, C-3 methine). Irradiation at 6 1.09 (C-Z methyl) caused the quartet of doublets at 6 2.90 to become a doublet with J = 3.5 Hz. Irradiation at 6 1.52 (C-4 methylene) caused the multiplet at 6 4.04 to become a doublet with J = 3.5 112. A solution of molar ratio of Eu (fed) : 102c = 0.71 in CDCl ('I'IVS 3 NNNN 3 as an internal standard) was placed in a 5-mm NMR tube which was in turn placed in a 10-mm NMR tube containing CDCl solution for taking a 3 180MHz spectrum. Shifted spectra (CDC13) 2.68-4.8 (6H, m, C5-C7 methylenes), 5.37-6.0 (211, m, C-4 methylene), 6.1 (3H, d, J = 7 Hz, C-Z methyl), 6.33 (3H, s, methoxyl), 6.6-7.22 (2H, m, C-8 methylene), 8.83 (1H, m, C-3 methine), 9.28 (1H, m, C-2 methine). Irradiation at 6 6.1 (C-Z methyl) caused the multiplet at 6 9.28 to become a doublet, J = 3.0112. 95 21. Base-catalyzed Addition of Methanol to 913 A solution of 76 mg of 9,3 in 15 m1 of 0.01 N methanolic sodium methoxide was stirred at room temperature and the reaction course was followed by VPC (column C at 170° C). The reaction rate was quite slow; only very small amount of the adduct was formed after 1 day. The maximum yield of adduct was obtained after stirring for 8 days and VPC showed the composition of the reaction mixture (equilibrium mixture) to be 33 : 19.31: ; 1925 = 45 : 3S : 20. This ratio remained about the same for a prolonged reaction time (2 weeks). For 182/t: IR (CC14) 2940 (s) , 1705 (s), 1465 (m), 1100 (5); NMR (CDC13) 6 1.14 (3H, d, J = 6112), 1.25- 2.2 (8H, broad), 2.35 (211, broad), 2.95 (2“, bread), 3.28 (3H, 5); mass spectrum, We (rel. intensity) 170 (5), 138 (31), 123 (6), 109 (23), 98 (21), 81 (29), 71 (100). LIS shift data were obtained (as that described for 10,25) with 4.5 mg of 41192/1; in CDCl3 solution to give the following results. Pblar Ratio A (ppm) 102t : Eu (fod)3 1 : 1 11m 112-#113 CH:5 0013 4.9 4.6 3.2 1.4 1 : 2 HZ ”m 11.5 (113 0013 1021: 7.14 6.87 6.70 4.80 2.16 LIS shifted spectrum (180 M12, 4.5 mg of 102t and 54.8 mg Eu (fed).5 in CDCl:5 solution): 6 3.3-5.16 (8H, C4-C7 methylene), 5.26 (311, s), 5.47 96 (311, d, J = 6.5 Hz), 8.48 (211, C-8 methylene), 8.9 (111, C-3 methine), 9.29 (111, C-2 methine); irradiation at 6 9.29 caused the doublet at 6 5.47 to become a singlet. These results indicated that C-2 and C-3 methine protons in 1,921.; are trans to each other. 22. Base-catalyzed Exchange Reaction of 102c in Methanol A solution of 25 mg of 1325 in 5 ml of 0.01 N methanolic sodium methoxide was stirred at room temperature and the reaction course was sollowed by VPC (column C, 170' C) . After 1 day, VPC showed the composition of the reaction mixture to be ,9/3 : 02t : 19216:, = 45 : 35 : 20. This ratio remained about the same for a few days. 23. Irradiation of 3-Methyl-2-cyclooctenone 38 in Methanol A degassed solution containing 70 mg of 38 in 10 m1 of methanol in a Pyrex tube was irradiated with a 450 W Hanovia lamp. The photolysis was followed by analytical VPC (colunn E, 1800 C) . The reaction was essentially complete in about 5 hr. VPC showed that the major product 4195 was formed in 72% yield (retention time 7 min). Preparative VPC (column D, 1750 C) gave pure 3-methoxy-3~methylcyclooctanone 195: IR (neat) 2940 (s), 1700 (s), 1470 (m), 1380 (w), 1310 (m), 1280 (w), 1220 (w), 1130 (m), 1080 (m) 011-1 ; NMR (CC14) 6 1.16 (3H, s, C-3 methyl), 1.2-2.0 (811, broad), 2.0-2.4 (2H, broad, C-8 methylene), 2.24 (111, d, J = 11 Hz, Ht)’ 2.7 (1H, d, J = 11 Hz, Hc)’ 3.12 (311, s, methoxyl); mass spectrum, m/e (rel. intensity) 170 (3), 155 (5), 140 (7), 138 (7), 123 (3), 114 (18), 99 (22), 95 (31), 85 (77), 72 (100), 55 (45). 97 24 . Europium Shift Study of 3—Methqu;3 -met11ylcyclooctanone 199 The procedure was the same as that described for 192. The chemical shift differences found for a molar ratio 1 : 1 of shift reagent to substrate are as follows: A'S in ppm, 6.9 (C-3 methyl), 11.6 (0013), 12.3 (Ht), 17.4 (11C), 8.3 (11m). 25. Irradiation of 2,3-Dimethyl-2-cyclooctenone 199 in Methanol In a Pyrex test tube, 46 mg of 199 was dissolved in 10 ml of methanol. After being deoxygenated for 15 min with a stream of nitrogen, the solution was irradiated with a Hanovia 450 W lamp. The course of the reaction was followed by VPC (column B, 180'" C) . After 8 hr, VPC showed that the peak corresponding to starting material remained and the intensity decreased only a little bit. The solvent was removed under reduced pressure. Preparative VPC (column C, 175°C) gave only the starting material . PART III SYNTHESIS AND PHOTOCI 1EMISTRY OF 1 , 5-D11 IETI'IYL-4 JETHYLENEBICYCLO [3 . 3 . 0] OCTADIENES kg INTRODUCTION It was first noted by Zimmenman and coworkers46 that molecules having the di-w-methane moiety, (i.e. having two n-systems bonded to a single sp3 carbon) undergo a general photochemical transfonnation to vinylcyclopropanes. This reaction has been termed the di-w—methane rearrangement. This rearrangement can be accounted fer by the gross mechanism shown below, which involves vinyl-vinyl bridging to give diradical 192, cleavage of the cyclopropane ring at either bond 2 or bond 9 and subsequent ring closure of the radical to afford the vinyl- cyclopropane. a . (WI—e ——»n-)—>r<1 106 MA; \ Zimmerman's47 study of the photochemistry of cis- and trans-1,1- diphenyl-3,3-dimethyl—l,4-hexadienes 191 and 199 constitutes an example of a di-n-methane rearrangement which occurs from a singlet excited state. Upon direct irradiation, rearrangement of the cis and trans isomers (192 and 199) occurred with extreme regiospecificity and striking stereo- specificity. The cis isomer 191 yielded exclusively the cis-propenyl product 111, and the trans isomer 199 only the trans-propenyl product 119. No product 111 was fonmed in either case. However, when 192 and 199 were separately irradiated in a solution containing benzo- phenone as a sensitizer, no vinylcyclopropanes were produced but the cis-trans isomerization interconverting 191 and 199 was observed. Hence 99 100 the triplet excited states of w and m do not undergo the di-n- methane rearrangement . 114 «N» 101 The selectivity in the opening of cyclopropane intennediates $93 and llg by the cleavage of only bond 3 in each case indicates that the reaction proceeds by the route which allows maximum odd-electron delocalization. Routes b lead to loss of benzhydryl delocalization while routes 3.do not. Hence, ll£ would be expected to be a higher energy intermediate than either ill or iii which afford the observed products. It has been proposed that the di-n-methane rearrangement, especially from the singlet excited state, can best be rationalized by a concerted mechanism.46 Therefore, intermediates like $93 - gig may not be discrete species and should be considered only as working models of the system. It has been generally observed that nonconstrained acyclic and methylene monocyclic di—n-methane systems undergo the di—w-methane rearrangement exclusively from singlet excited states, whereas constrained bicyclic systems undergo the rearrangement from triplet excited states. The rationale for this observation is that bicyclic or constrained triplets are unable to dissipate their excitation energy by rotation about a double bond and thus undergo the di-n-methane rearrangement. Alternatively, triplet states of acyclic systems such as $21 or l9§ do not have this constraint, and cis-trans isomerization is observed. The di-n-methane rearrangement was further generalized by Givens49a in 1969 when he noted that a carbon-oxygen w-bond (in 8,Y~unsaturated ketones) can also participate in the same manner as a carbon-carbon double bond. Dauben49b tenned this the oxa-di-n-methane rearrangement. In constrast to the di-w-methane rearrangement which has been found to occur from both singlet and triplet states, the oxa-di-n-methane 102 rearrangement appears to take place solely from the triplet state. The singlet photochemistry observed under direct photolysis of 8,y-unsaturated ketones consists mainly of 1,3-acy1 migration and/or decarbonylation, although other processes are sometimes also observed. For example, sensitized irradiation of bicyclo[2.2.1]hepta—5-ene-2-one llz48 yields the oxa-di-n-methane product ll§, whereas direct irradiation of llz affords an isomeric 3,y-unsaturated ketone llg. hv ._________, C) acetone ““" 118 hv a]. ,___, 0 direct 119 W The [n2 + «2] cycloaddition is one of the most widely observed photochemical reactions, with numerous examples reported in the literature.50 Although intramolecular [n2 + n2] cycloadditions are photochemically allowed, factors such as bond distance and orbital dihedral angle may make these cycloadditions unfavorable. For example, compounds lgg and lgk whose double bonds are separated by about 3.5; do not undergo intramolecular cycloaddition on irradiation. In contrast, compound kgg upon irradiation (acetonitrile, Pyrex) affords compound kid'SI 103 ph W» .123 Recently, however, Hart and Kuzuya52 found that.$§€ on irradiation (ether, vycor) gave an essentially quantitative yield of the crystalline isomer 1 , whose structure corresponds to that of a cycloaddition of the two endocyclic n-bonds. The structural feature present in lge but absent in $39 and lgl is the di-n-methane moiety. This reaction which on the surface corresponds to an intramolecular ["2 + “2] cycloaddition has been proposed to proceed in two steps, involving initial di—n- methane rearrangement to $33 and subsequent [02 + n2] cycloaddition to give lgé as shown below. 104 hv ——5 \ - a2 124 ,h . -1? T” Et2o on H A k0 p—a U" 0 § § 109 Scheme Z Ie c MeOZC I MeO2 Me MeOZC :0 \ 0:: "’ :0 ——> :0 ____, 0 he l-on MeOZC Meo2 Me MeOZC Me 136 135 151 152 ‘VW 'VV'V 'VW 'VW cone 0 , COzMe MeOZC Me MeOZC Me 00 Me O I 2 0 OH 0 Me MeO c MeO r: “e CO “e 2 2 153 154 MA: 'VW 00 Me Meo (.7 co Me MeO c 2 2 2 2 ie a O __._—-—-) 0 0 Meo C Me cone Meo o co Me 2 2 137 2 155 m .1 LCC14 OMe 11002".- \C 3 0~ “II 0 0 IL, orc‘ H cone OMe enol form 110 and the methylene protons at 6 2.34 was well resolved into two doublets with J = 19 Hz. 1.2(1.0) CH3 H 2.34 (2.6) ~41 2.34(3.0) 0 Reduction of diketone léfi with lithium aluminum hydride in ether yielded 88% of diol lég which upon dehydration with potassium bisulfate gave the isomeric dienes 140 and 141 in only 30% yield, along with cyclic ether 14% as a major undesired product (50% yield). Dienes‘lgg and 14% were separated by VPC in SS : 45 ratio. Allylic oxidation of the mixture of 141 and 142 with chromiun trioxide-pyridine complex resulted in a complicated mixture of a,B-unsaturated ketones which was first _separated into three fractions by preparative TLC (with 14; and 144 being in the first fraction, and 14§ and 146 constituting the second and third fractions respectively). The mixture of 14% and 1&4 was further separated by VPC to give each compound pure. Compounds 14% and 144 were distinguished on the bases of different chemical shifts and europium ahift slopes of the methylene protons. This signal is most affected by shift reagent when the methylene protons are at C-6 (i.e. in igg). The IR (vc=o 1720 cm‘l) and UV spectra (Amax 238 nm,e 1740 for rag; 241 nm, 1600 for 144) support the presence of a cyclopentenone moiety in each isomer. The NMR spectrum of each isomer showed two bridge-head methyls, four vinyl protons, and two methylene protons. The mass spectra of 14; and 124 were nearly identical, and they both had a correct elemental analysis. 111 1 1 1.08(1.22) 1.ll(1.28) (° 7)2}38 H or, n 7.20(1.20) 5'42(1'0) n of; II 7.36(1.35) (LMSJSH H semspm H “senses a1 0 (2.2)2.24 n : 0 2.42 5.38 ” 0‘ ( ) 1.l8(2.38) (3 63)2 so ' ‘ ' 1.15(2.67) ea 12% The distinction between 145 and 146 is based.on their IR and NMR 'VW'VVV spectra. The IR spectrum of 142 showed one carbonyl absorption at 1700 cm-1, whereas that of 146 showed two absorptions at 1720 and 1680 cm‘l. An infrared absorption at 1700 cm-1 was reported for 152 by Shih.57 The structure of 145 was also suggested by the presence of two equivalent methyls at 6 1.30 (6H, s) in the NMR spectrum as compared to the two different methyl singlets at 6 1.28 and 1.35 for 149. Both Add and $49 had parent peaks at m/e 162 in their mass spectra. 0 156 -~ Compounds 143 and 144 were finally converted to the desired trienes 'VVV’VV» 44§ and 152 reSpectively in good yield upon treatment with methyllithium followed by dehydration with dilute sulfuric acid. Compounds 148 and 'VVD 112 152 were assigned structures on the basis of their spectral data. The IR absorptions at 1635 cm-1 for both compounds indicated the presence of conjugated diene moiety. The UV spectrum of 148 at Amax 234 nm (E 14800, cyclohexane) and of $39 at 238 nm (12620) also support the presence of a conjugated diene moiety. The mass spectra of both 148 and $59 showed a parent peak at m/e 146. The NMR spectrum of 148 showed a singlet for the two bridgehead methyls at 6 1.08 (6H), two methylene protons at 6 2.26 (2H), two exo-methylene protons at 6 4.56 (1H, s) and 4.72 (1H, 5), four vinyl protons at 5 5.2 (2H), 5.63 (1H, d, J = 6 Hz), and 5.8 (1H, d, J = 6 Hz). The NMR spectrum of 152 showed the expected resonances: two methyl singlets at 6 1.05 and 1.10, two methylene protons at 5 2.32 (2“), two exosnethylene protons at 6 4.57 (1H, s) and 4.72 (1H, 5), four vinyl protons at 6 5.38 (2H, m) and 5.80 (2H, m). All of the spectra of,13§ and,139 resemble closely the published spectra of 157 and 158 by Hart and Kuzuya.58 WVWV 1.03 H 1.01 2.26 H H Me II 5.8 538 Me II 5.80 II H 5.63 H II 5.2 H II Me I H 4-56 2.32 ”Me “14°57 1.03 ” 1.05 II 4.72 4 72 «145 4&9 113 4.60 0.85 4.50 0.95§ 0.93 1.63 1.45 1.73 1.02 1.55 157 158 W» MA. 2. Synthesis of l,5-Dinethyl-3,7-diphenyl-4-methylenebicyclo[3.3.0]- octa-2,6-diene lQ3 The synthesis of 123 was accomplished according to Scheme 8. Treatment of diketone 138 with phenyllithium or phenylmagnesium ’VVM bromide in ether gave diol 159 in 30% yield. Dehydration of diol 159 W» 'VVV with refluxing p-toluenesulfonic acid in benzene afforded a mixture of dienes 160 and 161 (in a 75 : 25 ratio, calculated from NMR) in 46% 'VVV'VVV yield. Allylic oxidation of the diene mixture 120 and 121 with chromium trioxide-pyridine complex in methylene chloride produced after chroma- tography on alumina, a 24% yield of an a,8-unsaturated ketone 12% which was assigned the structure shown on the basis of its spectral data. The molecular formula CZZHZOO was confirmed by its mass spectrum (parent l.32(l.6) (1.0)2.32 H aI II 6.42 ,,~ 3 ph ph 7.1 egg H 04 (3.0)S.8 3 1.22(3.6) 114 peak m/e 300) and elemental analysis. Compound '16}, showed infrared bands at 1695, 705, and 770 cm'1 which are indicative of the presence of cyclo- pentenone moiety and a monosubstituted phenyl group. The UV spectrum showed Xmax 223 (I: 15550) and 258 nm (18550) consistent with extended conjugation between the phenyl ring and the a ,B-unsaturated carbonyl chromophore. The NMR spectrum showed two methyl signals at 6 1.22 and 1.32, two methylene protons at 6 2.82 (211), two vinyl protons at 6 5.80 (111) and 6.42 (1H) , and ten phenyl protons centered at 6 7.1. Scheme 8 ph phLi p h p- -TsOH CH 161 6 6 “f 138 159 (30%) ph h 'VW WW 41640 CrO3'2py 1 1) (313m E ph pl 2) p-TsOH ph ph (IIZCl2 0 \ 9 1’62 (240) 163 '\/V\; Compound 1% on treatment with methyllithium followed by dehydration with p-toluenesulfonic acid in benzene gave the desired triene 122 in quantitative yield. Compound 1’6": was assigned the structure shown. The molecular formula C231122 was confirmed by a mass spectrum (parent peak m/e 298) and elemental analysis. The IR absorptions at 1630, 765 115 1.25 II H (“3 115.77 ph ph 7.10 oI 4.79 [I 34.90 163 «AA. and 706 cm—1 are consistent with the conjugated olefin and the presence of a phenyl ring; UV maxima at 222 nm (e 7950) and 255 nm (7000) indicated the presence of conjugation within the molecule. The NMR spectrum showed the two bridgehead methyls at 6 1.25 (6H), methylene protons at 2.73 (2H), exo-methylene protons at 6 4.79 (1H) and 4.90 (1H), two vinyl protons at 6 5.77, and ten phenyl protons at 6 7.10. 3. Photolysis of 1,5-Dimethyl-3,7-dipheny1bicyclo[3.3.0]octa-2,6- dien-4-one kg; The irradiation of 162 in ether with 2537 A light resulted in the fonnation of 164, which is the product of a 1,3 acyl migration. Compound leg was assigned the structure shown on the basis of the following spectral data. The molecular formula CZZHZOO confirmed by a mass spectrum (parent peak m/e 300) and elemental analysis. IR absorption at 1690 cm-1 and a UV maximum at 225 nm (a 15600) are consistent with the conjugated structure shown. In its NMR spectrum, irradiation at 6 5.73 caused the doublet at 6 1.86 to become a singlet indicating that it is due to a vinyl methyl group. 116 164 (37%) 4. Photolysis of l,5-Dimethylbicyclo[3.3.0]octa-2,6—dien-4-one“133 When the mixture of 6,8-unsaturated ketones 1,4,3 and 1,441 was irradiated in ether through a vycor filter for 14 hr, compound 1,4,4 was recovered and 1,4,3 underwent 1,3 acyl migration to give compound 1,6,5. The structure of 16,5 followed from its spectral data. The molecular formula ClOHlZO was confirmed by a mass spectrum (parent peak Isl/C 148) and elemental analysis. Infrared absorption at 1685 00—1 and UV maxima at 244 nm (6 2570) and 281 (1230) are consistent with the conjugated cyclic enone. The NMR spectrum and europium shift data are also consistent with the Structure shown. (1.41)2.4 II II 2.4(3.33) C113 1.26(l.l6) (l.0)l.76 013 I 6.93(l.75) $35 * -———.>’“’ .511 * / vycor II 5.56 “I; o H 5.36(6.21) iii/3. (2'52) (6.58) 165 MA. 117 5. Photolysis of l,§;Dimethylbicyclo[3.3.0]octa-2,7-diene 141 When the diene mixture of 140 and 141 was irradiated (ether, quartz, MAW followed by VPC) for 38 hr, diene 142 remained unchanged, whereas a new product l,S-dimethyl[3.3.01’S .02’8.03’71quadricyclooctane 122 was formed from diene INl. This result was further confinned by the irradiation of diene 141 alone under identical condition to give the same product 122 in 10% isolated yield. Hence, diene 142 did not undergo intramolecular photocycloaddition due presumably to the unfavorable bond distance between the two double bonds. The formation of 166 from 141 is similar to what was observed by Kaupp51 (see introduction). Compound 162 was assigned the structure shown on the basis of the following spectral data. The absence of infrared bands in the region of 1500-1680 cm—1 together with the absence of signals corresponding to vinyl protons in the NMR spectrum indicated that 199 must be a saturated compound. The mass spectrum showed a parent peak at m/e 134 and a base peak at m/e 119 corresponding to a loss of Gig. TheITHISpectrum showed two methyl singlets at 6 1.06 and 1.32, four methine protons at 6 0.95 (2H, d, J = 9.5 Hz) and 1.24 (2H, d, J = 9.5 Hz) and four methylene protons at 6 1.8 (2H) and 2.28 (2H). quartz H J:- H H 0‘ 0" § 118 6. Photolysis of 1,5-Dimethyl-3,Z-diphenyl-4-methy1enebicyclo[3.3.0]- octa-2,6-diene 123 and 1,5-Dimethyl-4-methylenebicyclo[3.3.0]octa- 2,6 and 2 7-dienes 148 and_150 «MW» Compound 1&8, 150, and 163 all contain di-n—methane moieties, and would theoretically be capable of undergoing the di-n-methane rearrangement. However, when compound 163 was irradiated in acetone with 2537 A light and the reaction was followed by VPC for 22 hr, no detectable new product was formed. Irradiation of 148 in ether through a vycor filter for 29.5 hr only resulted in recovery of starting material. Photolysis of 150 under identical conditions gave some recovered starting triene 150 and an unidentified product, according to VPC and NMR analysis. All three of the above compounds may have suffered significant hv 10 Ph . N R 2537A acetone 163 MA: h " N. R ether 148 MA: hv 150 + an unidentified product ether ”m“ 150 «AA. 119 polymerization because substantial amounts of starting materials were lost and some polymeric material was found in each case. The failure of these compounds to undergo the di-n-methane rearrangement is probably because the polymerization and other energy dissipation processes (involving no chemical change) of the excited trienes are more efficient than the di-n-methane rearrangement. In summary, compounds 14%, 150, and 163 containing the di-r—methane moiety have been successfully synthesized. However, these compounds on irradiation did not give detectable amount of di-n—methane rearrangement product or the product corresponding to cycloaddition of the two endocyclic double bonds. FXPERIMENTAL 1. Gas Chromatography VPC columns that were used in this section are as follows: 10' x 0.25 in column, 21% 8,8'oxydipropionitrile on firebrick. 5' x 0.25 in column, 20% FFAP on chromosorb W. 5' x 0.25 in column, 5% SE30 on chromosorb W. boa? 5' x 0.125 in column, 5% SE30 on chromosorb W. 2. l,5-Dimethyl-2,4,6,8-tetramethoxycarbonylbicyclo[3.3.0]octan—3,7- dione 13753 -———'VW Reaction of dimethyl e-ketoglutarate 136 (59.12 g, 0.34 mol) and 2,3-butanedione 135 (14.62 g, 0.17 mol) at room temperature in 300 mol of aqueous buffer (pH 5) for 1 day afforded tetraester 131 in 90% yield, MP 155-159o (sublimed), recrystallized from methanol. It is 100% enolized in CDCl3 and CCl solution as shown by NMR. NMR (CDCl3 + 4 0014) 6 1.25 (6H, s), 3.68 (6H, s), 3.8 (411, s), 10.4 (2n, 5, enol). IR (CDClS) 1730 (s), 1665 (s), 1630 (s) cm-l; mass spectrum, m/e 398 (parent). 3. 1,5-Dhmethylbigyclo[3.3.0]octan—3,7—dione 13853 Hydrolysis of 131 (139.3 g, 0.35 mol) in 600 ml of 6 N HCl at reflux for 1 day resulted in an 82% yield of 138 (after recrystallization 120 121 from ethanol), MP 166- 167° (sealed tube). In (c0c13) 1738 tm'l ; hmflR (CDC13) 6 1.2 (6H, s), 2.34 (8H, 5); Eu (fed)3 shift reagent resolved the peak at 6 2.34 into two doubles with J = 19 Hz; for europium shift slopes see its structure in the text. 4. 1,5-Dimethylbicyclo[3.3.0]octan-3,7-diol 139 A solution of diketone 138 (9.96 g, 0.06 mol) in 150 ml ether was added under nitrogen over 40 min to a stirred refluxing solution of LiAlH4 (2.6 g, 0.068 mol) in 75 m1 ether. After addition, the reaction mixture was stirred at reflux for 4 hr, then cooled and quenched through the sequential addition of 2.4 ml H20, 2.4 ml 1.5% N301, and 7 ml H20. The precipitate was filtered, and filtrate was dried over NaZSO4. Concentration of the filtrate gave 8.5 g of a white powder. Extraction of the salts precipitated through the quenching sequence with ether led to an additional 0.5 g of diol. Recrystallization from chlorofonn gave a white crystalline solid in 88% yield, MP 148-151o (sublimed); IR (KBr) 3300 cm'l; mass spectrum, m/e 170 (parent). 53b 5. l,5-Dimethylbicyclo[3.3.0]octa-2,6 and 2,7-dienes 140 d 141 'VW—‘VV‘M A mixture of 6.8 g (0.04 mol) of diol 122 and 6 g of fused KHSO4 was placed in a lOO-ml round-bottomed flask equipped with a distillation apparatus. The mixture was heated to 200°C for 2 hr. The dehydrated product was condensed in the condenser as a semisolid material which is mostly diene. The residue left in the flask was then cooled, dissolved in 30 m1 of water, extracted several times with ether. Ether layer was 122 washed with saturated 11311003 and NaCl solution, dried over MgSO4 and decolorized with Norit A. The ether extract was then combined with the distillate and concentrated to give about 20 ml of crude product which upon distillation at 85-90°C/120 mm Hg gave 1.64 g of colorless liquid (30% yield) consisting of dienes 140 and 141 in 55 : 45 ratio (judged by VPC). The residue (3 g, 50% yield) left behind in the flask contained mainly cyclic ether 14%. The mixture of 140 and 141 was further separated by VPC (column A, 80°C). For 142: retention time 10 min: 11m1 (CC14) 6 1.02 (a1, 5), 2.18 (4H, s), 5.29 (4H, 5); mass spectrum, m/e 134 (40%, parent), 119 (100%). For 141: retention time 12.5 min; bflflk (CC14) 6 0.95 (3H, s), 0.97 (3H, s), 2.18 (4H, s), 5.35 (4H, 5); mass spectrum, m/e 134 (45%), 119 (100%). For 143%: MP 140-14211; IR (CC14) 2950 (s), 1460 (s), 1320 (s), 1190 (m), 1160 (s), 1130 (m) cm-l; 1002 (CC14) 6 1.07 (6H, s), 1.2-1.7 (8H, m), 4.12 (2H, broad); mass spectrum, m/e 152 (14%, parent), 95 (100%). High resolution mass spectrum mol wt 152.11969 (calcd. for O, 152.12012). C10" 16 6. Allylic Oxidation of Dienes 142 and 141 ——--'\/V\: The procedure of Shaffer and Pesero37b was followed. Thus 0.373 g (2.8 mmol) of diene mixture 142 and 141 was oxidized by chromium trioxide— pyridine complex (6.7 g pyridine and 4.18 g Cr03) in 70 m; methylene chloride. The reaction mixture after workup and preparative TLC (silica gel, hexane : ether = l : 1) gave 0.168 g of a mixture of 1,5-dimethy1- bicyclo[3.3.0]octa-2,6—dien—4-one 143 and 2,7-dien-4-one 144, 0.0117 g 'VW 'VVV of 1,S-dimethylbicyclo[3.3.0]octa-2,6-dien—4,Q-dione iii and 0.0077 g of 1,S—dimethylbicyclo[3.3.0]octa—2,7—dien-4,6-dione 146. The mixture of 123 "1,4,3 and «119/3 was further separated by VPC (column B, 105°C) . Therefore, the yields for these compounds are as follows: 143 : 1,4,4 : «MN‘S, : «1,4,9 = 20% : 20% : 2.6% : 1.8%. Spectroscopic data for compound 143: IR (neat) 1702 cm'l; UV (95% EtOH) xmax 238 (c 1740); NMR (CDC13) a 1.08 (3H, s), 1.18 (3H, s), 2.38 (2H, s), 5.38 (2H, s), 5.86 (1H, d, J = 6 Hz), 7.20 (1H, d, J = 6 Hz); for Europium shift data see structure 1,4,3; mass spectrum, m/e (rel. intensity) 148 (59), 133 (57), 120 (23), 105 (100), 91 (26), 79 (29). Anal. Calcd. for ClOHIZO: C, 81.04; 11, 8.16. Found: C, 80.78; H, 8.05. For compound 1%: IR (neat) 1702 cm-1 ; UV (95% EtOH) )‘max 241 (c 1600); NMR (CDC13) 6 1.11 (3H, s), 1.15 (311, s), 2.24 (1H, d, J = 18112), 2.60 (1H, d, J = 18 Hz), 5.42 (211, broad), 5.84 (111, d, J = 6 Hz), 7.36 (1H, d, J = 6 Hz); for Europium shift data see structure 144; mass spectrum, m/e (rel. intensity) 148 (71), 133 (61), 120 (17), 105 (100), 91 (27), 79 (30). Anal. Calcd. for cldl o: c, 81.04; H, 8.16. 12 Found: C, 81.09; H, 8.14. For compound 143: MP 161—162°c (sublimed); IR (KBr) 1700 cm'l; NMR (CDClS) 6 1.30 (6H, s), 5.88 (2H, d, J = 6 Hz), 7.30 (2H, d, J = 6 Hz); mass spectrum, m/e (rel. intensity) 162 (48), 147 (35), 134 (44), 119 (8), 106 (24), 91 (100). For compound tfié= mp 170°C (sublimed); IR (KBr) 1720 (s), 1680 (5) 011-1; NMR (CDClS) 6 1.28 (3H, s), 1.35 (3H, s), 5.74 (2H, d, J = 6 Hz), 7.40 (2H, d, J = 6112); mass spectrum, m/e (rel. intensity) 162 (50), 147 (38), 134 (56), 119 (11), 105 (26), 91 (100). 124 7. Synthesis of 1,5-Dimethy1-4-methy1enebicyclo[3.3.0]octa-2,6-diene 128 To a solution of 0.567 g (3.8 mmol) of 142 in 30 ml ether was added 5 ml of methyllithium (1.8 M in ether) with a syringe at 0°C under nitrogen. The mixture was then refluxed for 2 hr. After cooling, the reaction mixture was poured into an ice-NH4C1 solution (= 100 ml) and extracted with 150 m1 ether. The ether layer was concentrated to give 0.63 g of crude alcohol 147 which showed hydroxyl absorption at 3460 cm-1 and no carbonyl absorption. The crude alcohol 881 was refluxed with 30 m1 of 20% sulfuric acid for 1 hr. The reaction mixture was extracted with 50 ml of ether, the ether layer washed with saturated NaHCO3 and NaCl solution, and dried over MgSO4. Removal of the solvent left 0.56 g of crude diene 148 as a yellow liquid. A pure sample of compound 148 was collected in 50% yield by VPC (column B, 130 C, retention time 3.1 min). IR (CC14) 2980 (s), 1635 (s), 1380 (s), 880 (s) cm_1; UV (cyclohexane) Amax 234 nm (a 14800); 0011 (CC14) 6 1.08 (6H, s), 2.26 (2H, s), 4.56 (1H, s), 4.72 (1H, s), 5.2 (2H, m), 5.63 (1H, d, J = 6 Hz), 5.8 (1H, d, J = 6 Hz); mass spectrum, m/e (rel. intensity) 146 (31), 131 (100), 116 (19), 105 (9), 91 (38), 77 (9). High resolution mass spectrun mol wt. 146.10854 (calcd. for C 146.10954). 11”14’ 8. Synthesis of l,S-Dimethyl~4-methylenebicyclo[3.3.0]octa—2,7-diene 150 The procedure and workup were as described for diene 148. From RM 0.64 g of 144 there was obtained 0.7 g of crude alcohol 142, and dehydration of the crude alcohol 149 gave 0.58 g of crude diene 150. A 'VW ’VW pure sample was collected by VPC (same conditions as for 148, except 125 retention time = 2.8 min) in 40% yield. IR (CC14) 2980 (s), 1635 (s), 1380 (s), 880 (s) cm‘l; 11v (cyclohexane) 1mm, 238 mm (c 12620); MR (cc14) 5 1.05 (3H, s), 1.10 (3H, s), 2.32 (2H, m), 4.57 (1H, s), 4.72 (1H, s), 5.38 (2H, m), 5.80 (2H, m); mass spectrum, m/e (rel. intensity) 146 (35), 131 (100), 116 (16), 105 (7), 91 (35), 77 (12). High resolution mass spectrum mol wt. 146. 10881 (calcd. forC 146. 10954). ‘11} 14’ 9. 1,5-Dimethy1-3,7-diphenylhicyclo[3.3.0]octan-3,7-diol 152 A solution of 0.83 g of (5 mmol) of diketone 138 in 50 ml THF was added solwly at 0°C under nitrogen to a solution of 8 ml of 2.0 M phenyllithium (16mmol) in 20 ml THF. The reaction mixture was refluxed for 2 hr, then cooled and quenched with 50 ml ice water. Ether (100 ml) was used to extract the diol. The organic layer was dried over NaZSO4 and the solvent was stripped to give crude diol 153. Recrystallization from methanol gave colorless needle—like crystals (0.5 g, 30% yield), 0? 155-157°: IR (KBr) 3250 (s), 2950 (s), 1600 (m), 760 (s), 700 (s) cm'l. Compound 152 was also obtained.from diketone 138 and phenylmagnesium bromide in a similar yield. 10. l,5-Dimethyl-3,7-diphenylbicyclo[3.3.0]octa-2,6 and 2,7—diene 122 56 and 161 ——--'\/V\; A mixture of 17.6 g of (55 mmol) of diol 152 and 12 g of p-toluene- sulfOnic acid in 350 m1 benzene was refluxed for 2 hr with continous removal of water. The reaction mixture was then cooled, neutralized with 350 m1 5% IJaHCD3, and extracted with 200 ml benzene. The benzene 126 layer was washed with water and dried (MgS04). Removal of solvent and column chromatography (neutral alumina, cyclohexane) gave 7.2 g (46% yield, recrystallized from petroleum ether) of colorless crystals which was a mixture of dienes 160 and 161 in a 75 : 25 ratio (calculated franrdfld, 1P 64-6S°; IR (KBr) 2920 (s), 1500 (s), 1450 (s), 760 (s), 700 (s) cm'l; UV (cyclohexane) Amax 220 nm (e 1770), 247 (21200), 253 (23100), 258 (24200), 263 (24200); 1118 (cc14) 6 1.2 (6H, n0 , 2.5 (4H, broad), 5.8 (2H, broad), 7.1 (10H, broad). 11. 1,S-Dimethyl-S,7-diphenylbicyclo[3.3.0]octa-dien-4-one 182 The procedure as described for making 133 and 134 was followed. A diene mixture of 199 and 131 (75 : 25 ratio, 8 g, 28 mmol) was oxidized by chromium trioxide-pyridine complex (72 g pyridine, 45 g Cr03) in 1000 ml methylene chloride. Product 162 was separated on alumina (hexane : ether = l : 4 initially, then 1V: 1) in 24% yield: MP 94-95°C; IR (KBr) 2950 On), 1695 (s), 1500 On), 145- (m), 770 (s), 705 (s) cm'l; UV (95% EtCXU Amax 223 nm (c 15550), 258 (18550); NMR (CC14) 5 1.22 (3H, s), 1.32 (3H, d, J = 2 Hz), 2.82 (2H, m), 5.80 (1H, m), 6.42 (1H, m), 7.1 (10H, broad); mass spectrum, m/e (rel. intensity) 300 (100), 238 (20), 272 (20), 257 (23), 183 (58), 170 (64). Anal. Calcd. for C22“ 0: C, 87.96; H, 6.71. 20 Found: C, 87.91; H, 6.78. 12. 1,S-Dimethyl-3,7-diphenyl-4-methylenebicyclo[3.3.0]octa-2,6-diene 123 The procedure described for making compound 148 was followed except 127 that p-toluensulfonic acid was used fer the dehydration in this case. Thus compound 16; was obtained in quantitative yield from 122 (0.3 g, 1 mmol) by treatment with methyllithium (1.5 ml, 1.8 M in ether) followed by dehydration with p-toluenesulfonic acid (0.24 g) in 150 ml benzene. A sample of pure 123 was obtained by VPC (column C, 225°) or TLC (silica gel, ether + hexane). IR (neat) 3000 On), 1630 On), 1505 On), 1455 On), 765 (s), 706 (s) cm'l; UV (cyclohexane) Amax 223 nm (e 7950), 255 (7000); NMR (CC14) 6 1.25 (61, singlet with side peak), 2.73 (2H, broad), 4.79 (1H, s), 4.90 (1H, s), 5.77 (2H, broad), 7.10 (10H, broad); mass spectrum, m/e (rel. intensity) 298 (100), 283 (61), 268 (16), 252 (12), 205 (25), 195 (55), 165 (53). Anal. Calcd. for C C, 92.62; H, 7.83. 2f52‘ Found: C, 92.70; H, 7.71. 13. Photolysis of 1,5-Dimethyl-3,7—Dipheny1bicyclo[3.3.0]octa-2,6-dien- 4-one 162 _-’\/W A degassed solution containing 100 mg of 122 in 12 m1 ether was irradiated with 2537 A light in a Rayonet photochemical reaction apparatus for 18 hr. The photolysis was followed by VRC. A major product 124 was separated by VPC (column C, 215°) in 37% yield. Data for 164: IR (CC14) 2980 (m), 1690 (s), 1510 (m), 1455 On), 705 (s) cm-l; UV (95% EtOH) Amax 225 nm (c 15600); MIR (CC14) 6 1.38 (3H, s), 1.86 (3H, d, J = 2 Hz), 2.40 (1H, d, J = 10 Hz), 2.92 (1H, d, J = 10 Hz), 5.73 (1H, broad), 6.95 (1H, broad), 7.06 (10H, 5); irradiation at 6 5.73 caused the doublet at 6 1.86 to become a singlet: mass spectrun, m/e (rel. intensity) 300 (95), 285 (20), 262 (19), 257 (27), 170 (100), 128 155 (56). Anal. Calcd. for CZZHZOO: C, 87.96; H, 6.71. Found: C, 88.02; H, 6.76. 14. Photolysis of 1,5-Dimethylbicyclo[3.3.0]octa-2,6-dien-4-one 143 A quartz tube containing 100 mg of the mixture of 143 and 144 ’VW'VW (1 : 1 ratio) was flushed with nitrogen and irradiated for 14 hr with a Hanovia 450 watt lamp using a vycor filter. Compound 144 was recovered, whereas compound 14% (50% conversion) gave a new product 125 which was collected in 12% yield by VPC (column B, 130°). Data for 125: IR (neat) 2930 On), 1685 (s) cm'l; UV (958 Etdi) Nu 244 nm (e 2570), 281 (1230); ax rent (CDC13) 6 1.26 (3H, s), 1.76 (3H, d, J = 2 Hz), 2.4 (2H, broad), 3.12 (1H, broad), 5.36 (1H, d x d, J = 10 Hz, 2 Hz), 5.56 (1H, broad), 6.93 (1H, d x d, J = 10 Hz, 2 Hz); Europium shift data see structure 165; mass spectrum, m/e (rel. intensity) 148 (59), 133 (51), 120 (19), 105 (100), 91 (28). Anal. Calcd. for c 1 0: c, 81.04; H, 8.16 10’12 ' . Found: c, 81.13; H, 8.07. 15. Photolysis of 1,5-Dimethylbicyclo[3.3.0]octa-2,7-diene 141 A quartz tube containing 41 mg of diene 141 in 12 ml ether was degassed and irradiated with a Hanovia 450 watt lamp (no filter used) for 46 hr. The progress of the photolysis was followed by VPC (column 1,5 2 8 02,7 A, 80°C). The new product 1,5-dimethyl[3.3.0 .0 ’ ]quadri- cyclooctane 166 was fonned with retention time of 5.4 min in 45% yield 129 (10% isolated yield from VPC). IR (CC14) 2960 (s), 1460 On), 1315 On) cm’1;NMR (cc14) 6 0.95 (2H, d, J = 9.5 Hz), 1.06 (3H, s), 1.24 (2H, d, J = 9.5 Hz), 1.32 (3H, s), 1.8 (2H, broad), 2.28 (2H, broad); mass spectrum, m/e (rel. intensity) 134 (16), 119 (100), 105 (22), 91 (71), 77 (27). Photolysis of a mixture of 142 and 141 under the same conditions gave recovered 142 and product 122 in the same yield. 16. Photolysis of 1,5;Dimethyl-3,7-dipheny1-4-methylenebicyclo(3.3191; octa-2,6-diene 163 -—-——-——-—-— MA. A quartz tube containing 55 mg of 163 in 10 ml of acetone was degassed and irradiated with 2537 A light. The photolysis was followed by VPC (column D, 200°). After irradiation for 22 hr, VPC showed that starting material remained there with no detectable volatile new product fonned. Preparative VPC gave only starting material. Prolonged irradiation produced only polymeric material. 17. Photolysis of l,S-Dimethyl-4-methy1enebicyclo[3.3.0]octa-2,6-diene 148 and 2,7—diene 150 ————————MA. A degassed solution containing 97.5 mg of 148 in 12 ml ether was irradiated through vycor filter with a 450 watt lamp. VPC (column D, 85°) showed no reaction even after irradiation for 29.5 hr. Preparative VPC gave only the starting diene 148. Dienellég (91 mg in 12 m1 ether) was irradiated under identical condition. VPC showed a new peak corresponding to the new product in 130 12 hr; however, the ratio of the new product to diene 1,5/Q was small and did not change even after irradiation for 29.5 hr. After removal of solvent, there was left a light brown viscous liquid (containing polymeric material). Preparative VPC gave some starting diene 1,5}?! and a very small amount of an unidentified new product. BIBLIOGRAPHY 10. 11. 12. 13. 14. 15. 16. BIBLIOGRAPI [Y National Distillers Chemical Co., Ashtabula, Ohio, Technical Bullentin, "High Surface Sodium". K. W. Bowers, R. W. Giese, J. Grimshaw, II. 0. House, N. H. Kolodny, K. Kronbager, and D. K. Roe, J. Am. Chem. gm” 8%, 2783 (1970). Thor Alvik and Johannes Dale, Acta. Chem. Scand. , 25, 1153 (1971). M. A. M. Boersma, Cat. Rev. Sci. Eng., 1,9 (2), 243 (1974). J. M. Lalancette, G. Rollin, and P. Dimas, Can. .1. Chem., 50, 3058 (1972). M’ H. Dodall and W. E. Foster, J. Org. Chem., “23, 401 (1958). W. Rochus and R. Kichuth, German Patent, 1,095,832 (1957). D. B. Bergbreiter and J. M. Killough, J. Chem. Soc.me. Comm., 913 (1976). D. Savoia, C. Tromhini and A. Dmani-Ronchi, Tetrahedron Lett., 653 (1977). Marshall Sittig "Sodium, its manufacture, properties and useS" Reinhold Publishing Co., N. Y. 1956, Chap. 7. H. 0. House, "Modern Synthetic Reactions", W. A. Benjamin, Inc., California, 1972, (a) Chap. 9; (b) Chap. 3. (a) H. 0. House, M. Gall and H. D. Olmstead, J. Org. Chem., «89’ 2361 (1971). (b) H. 0. House and V. Kramar, J. Org. Chem., «28, 3362 (1963). Andreas Lindert, Ph. D. Thesis, Michigan State University, 1972. H. A. Smith, B. J. L. Huff, W. J. Powers III, and D. Caine, J. Org. Chem., 32, 2851 (1967). J. Kuhn, J. Am. Chem. Soc., 33, 2492 (1952) and 76, 4323 (1954). 'VD S. P. McManus "Organic Reactive Intermediates" Academic Press N. Y., 1973, Chap. 6. 132 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 133 Drury Caine, Organic Reactions, 2}, 1 (1976); D. H. R. Barton and C. H. Robinson, J. Chem. Soc., 3054 (1954). W. S.hiuphy and D. F. Sullivan, J. Chem. Soc. Perkin I, 999 (1972). H. 0. House, H. C. Muller, C. G. Pitt, and P. P. Wickham,1. .Org. Chem, 28, 2407 (1963); J. W. Huffman and I. T. Charles, J. Am. Chem. Soc. , 90, 6486 (1968). Alan Francis Thomas "Deuterium Labeling in Organic Synthesis" Appleton-Century-Croft, N. Y., 1971, Chap. 4. B. J. Huff, F. N. Tuller, and D. Caine, J. 0r .Chem ,34, 3070 (1969); J. M. Conia and P. Briet, Bull. Chim. Fr., 3881, 3888 (1966). G. D. Meakins, R. K. Per E. E. Richards, and R. N. Young, J. Chem. Soc. , (C), 1106 (1968). (a) R. A. Lee, C. McAndrews, K. M. Patel, and W. Reusch, Tetrahedron Lett., 965 (1973). (b) H. Hart, C. M. L5ve, and I. C. Wang, Tetrahedron Lett., 1377 (1977). H. Hart, B. L. Chen, and C. T. Peng, Tetrahedron Lett., 3121 (1977). (a) O. L. Chapman '(kganic Photochemistry" Marcel Dekker Inc. N. Y., 1973, Vol. 3, Chap, (b) J. Ciabattoni andp E. 3C. Nathan, III, J. Am. Chem. Soc., 91, 4766 (1969). (a) O. L. Chapman and J. D. Lassilla, J. Am. Chem. Soc., 39, 2449 (1968 . . 1. a 1n an . . c 1e , . Am. em. oc., (b) J E B ldw' d M C M Dan' 1 J Ch S 39, 6118 (1968). (a) P. E. Eaton, Accounts. Chem. Res. cited therein. (b) P. E. Eaton, J. Am. Chem. Soc., 83, 2344 and 2454 (1962). 50 (1968) and other papers Iv, H. E. Zimmerman, R. K. Rieke, and J. R. Scheffer, J. Am. Chem. Soc., 89, 2033 (1967). V» (a) P. E. Eaton and Kang Lin, J. Am. Chem. Soc. , 86, 2087 (1964) and 87, 2052 (1965). (b) E. J. Corey, M. Tada, R. LeMahieu, and L. Libit, J. Am. Chem. Soc. , 87, 2051 (1965). (a) R. Noyori and M. Kato, Bull. Chem. Soc. Jap., 47, 1466 (1974). (b) R. Noyori, A. Watanabe, ma H. Kato, Tetrahedron Lett. ,5443, (1968). 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 134 (a) P. J. Kropp, Pure 1e. Chem., 24,, 585 (1970). (b) J. A. Marshal , ccounts em. Res., 2, 33 (1969). (a) J. A. Marshall and M. Lewellyn, J. Am. Chem. Soc., 99, 3508 (1977). (b) M. Nakazaki, K. Yamamoto, and J. anagi, J. C. S. Ch’én. Comm., 346 (1977). A. C. Cope, K. Banholzer, H. Keller, B. A. Bawson, J. J. Whang, and H. J. S. Winkler, J. Am. Chem. Soc., 81, 3644 (1965). E. Dmkelblum and H. Hart, J. Am. Chem. Soc., 39, 645 (1977). N. Heap and G. H. Whitham, J. Chem. Soc., (B), 164 (1966). H. C. Brown and M. Borkowski, J. Am. Chem. Soc., 74, 1900 (1952). (a) W. G. Dauben, M. Lorber and D. S. Fullerton, J. Ora. Chem., 34 3587 (1969). "“"““““ N“ (b) G. W. Shaffer and M. Pesaro, J. Org. Chem., 32, 2489 (1974). (a) T. Cohen and T. Tsuji, J. Or . Chem., 26, 1681 (1961). (b) J. K. Crandall and L. H. C. fim, J. Org. Chem., 32, 2375 (1968). E. W. Garbisch, Jr., J. Org. Chem. , '32 2109 (1965). J. M. Donia and C. Girard, Tetrahedron Lett., 2767 (1973). (a) E. J. Corey and D. Enders, Tetrahedron Lett., 3 (1976). (b) G. R. Newkome and D. L. FisHeI, 3. (fig. Chem., 31, 677 (1966). H. 0. House and R. W. Bashe,,II, J. Org. Chem., 39, 2942 (1965). W. G. Dauben and D. M. Michno, J. Org. Chem., 42, 682 (1977). S. Kabuss H. G. Schmid H. Friebolin, and W. Faisst, Org. Magn. Resonance: g, 451 (1969). E. Dunkelblum and H. Hart, J. Org. Chem.. in press. S. S. Hixson, P. S. Mariano, and H. E. Zimmerman, Chem. Rev., 73, 531 (1973) and references cited therein. M’ H. E. Zimmerman and A. C. Pratt, J. Am. Chem. Soc., ’32, 6267 (1970). J. Ipaktschi, Tetrahedron Lett., 2153 (1969). (a) R. S. Givens and W. F.0ettle, Chem. Comrmm., 1164 (1969). (b) W. G. Dauben, M. S. Kellog, J. I. Seeman, and W. A. Spitzer, J. Am. Soc., 32, 1786 (1970). 50. 51. 52. 53. 54. SS. 56. 57. 58. 135 R. B. Woodward and R. Hoffmann " The Conservation of Orbital Symmetry" Verlag Chemie 1970, Chap. 6 and references cited therein. G. Kaupp and K. Krieger, Angew. Chem. Int. Ed. Engl., 1,1 719 (1972). H. Hart and M. Kuzuya, J. Am. Chem. Soc., 36, 3709 (1974). (a) U. Weiss and J. M. Edwards, Tetrahedron Lett. 4885 (1968). (b) J. B. Baldwin and M. S. Kaplan, J. M. Chen. Soc., 33, 3969 (1971). S. Y. Lan, M. M. Johnson, J. Oehldrich, D. Wichman, and J. M. Cook, J. 03g. Chem., 41, 4053 (1976). Pelayo Camps, Tetrahedron Lett. , 4067 (1974). R. Askani, Tetrahedron Lett., 447 (1971). E. M. Shih, Ph. D. Thesis, Michigan State University, 1975. H. Hart and M. Kuzuya, Tetrahedron Lett., 3891 (1973). ; “'11111111111111111111ES