ABSTRACT THE THERMAL AND ACID CATALYZED REARRANGEMENTS OF THE PULEGONE OXIDES by Donald Frederick Anderson During the past twenty-five years, a vast amount of research has been conducted to study the behavior of strained systems. Much attention has been devoted to investigations involving the chemical, photochemical, and thermal reactions of the highly strained three-membered ring compounds. These compounds have been found to be highly reactive and present a wide variety of attractive research investigations. Included among the three—membered ring heterocyclic systems are the highly reactive epoxides. Many investi- gations have been undertaken with a,B—epoxy ketones and interesting photochemical, thermal and acid catalyzed re- actions have been observed. The acid catalyzed reactions of d,B-epoxy ketones often give B-diketones as rearrangement products resulting from an 1,2-acyl migration. The thermal reactions give products resulting from epimerization, frag- mentation, and rearrangement. The products obtained from the thermal reactions do not generally parallel the products from the photochemical processes. Donald Frederick Anderson In this study of the diastereoisomeric pulegone oxides, identical products were obtained from both the thermal and photochemical reactions. The diastereoisomers of 2,5-dimethyl- 2-acetylcyclohexanone were isolated and identified as the major products of the neat liquid phase thermal rearrangement. This thermal rearrangement is believed to proceed through a diradical intermediate. The course of reaction in solution depended on the nature of the solvent. Oxygen-containing sol- vents, capable of stabilizing an electron deficient diradical intermediate, directed the rearrangement toward the formation of the diastereoisomeric 2,5-dimethyl—2-acetylcyclohexanones. In hydrocarbon solvents, predominant formation of 2,2,5-tri methylcycloheptane-l,5-dione was observed; this is explained by a heterolytic cleavage and a subsequent 1,2-acyl migration. The vapor phase thermal rearrangement of both diastereo- isomeric pulegone oxides gave 2,2,5-trimethylcycloheptane-l,3- dione as the major product; the reaction was found to be sub- stantially catalyzed by an increase in surface area. The mechanism for this rearrangement probably parallels the re- arrangement in hydrocarbon solvents. Acid catalysis with boron- trifluoride etherate converted the pulegone oxides stereo- specifically to diastereoisomeric fluorohydrins, upon heating at 1500, the fluorohydrins decomposed to give 2,2,5-trimethyl- cycloheptane-1,5-dione. Treatment of the pulegone oxides Donald Frederick Anderson with p—toluenesulfonic acid gave diastereoisomeric unsat- urated hydroxy ketones. N.M.R..and infrared spectroscopy were used to establish the identity of each of the rearrange— ment products. THE THERMAL AND ACID CATALYZED REARRANGEMENTS OF THE PULEGONE OXIDES BY Donald Frederick Anderson A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1966 . r V’,’ Dedication To the Late Mr. and Mrs. H. B. Mans ACKNOWLEDGMENTS The author wishes to express his appreciation to Professor William H. Reusch for his continued interest, counsel, and encouragement throughout the course of this investigation. Grateful acknowledgment is extended to the National Science Foundation for personal financial assistance from March, 1964 through June, 1965 and from September, 1965 through March, 1966. Special thanks are also extended to my mother for her encouragement throughout the course of my graduate study. ii TABLE OF CONTENTS Page INTRODUCTION AND HISTORICAL . . . . . . . . . . . . 1 RESULTS AND DISCUSSION. . . . . . . . . . . . . . . 7 I. Products from the Rearrangements of the Pulegone Oxides. . . . . . . . . . . . . . 7 II. Mechanisms for the Thermal Rearrangements of the Pulegone Oxides . . . . . . . . . . 19 A. Liquid Phase Reaction. . . . . . . . 19 B. Gas Phase Reaction . . . . . . . . . 27 C. Solvent Study. . . . . . . . . . . . 31 EXPERIMENTAL. . . . . . . . . . . . . . . . . . . . 55 I. General Procedure. . . . . . . . . . . . . 55 A. Apparatus. . . . . . . . . . . . . . 55 B. Preparation of Pyrex Vials . . . . . 35 C. Melting Points . . . . . . . . . . . 56 D. Microanalysis. . . . . . . . . . . . 56 E. Furnace. . . . . . . . . . . . . . . 36 F. Pulegone . . . . . . . . . . . . . . 56 G. Synthesis of the Pulegone Oxides . . 37 H. Isolation of Pulegone Oxide Isomers I and II . . . . . . . . . . . . . . 57 I. Purity and Purification of Solvents. 58 1. Solvent and Reagent Purity . . 38 2. Solvent Purification . . . . . 38 II. Acid Catalyzed Rearrangements. . . . . . . 40 A. Rearrangement of Pulegone Oxide Iso- mer I with Borontrifluoride Etherate 40 B. Rearrangement of Pulegone Oxide II with Borontrifluoride Etherate . . . 43 C. Reactions of Products from I and II with Base. . . . . . . . . . . . . . 45 iii TABLE OF CONTENTS - Continued D. Thermal Decomposition of Fluorohy- drins XI and XII . . . . . . . . . . E. Attempted Deuterium Exchange of Compound V . . . . . . . . . . F. Esterification of IXa and IXb with Diazomethane . . . . . . . . . . . . G. Preparation of Diazomethane. . . . . H. Rearrangement of Pulegone Oxide I with p-Toluenesulfonic Acid. . . . . I. Rearrangement of Pulegone Oxide II with p—Toluenesulfonic Acid. . . . . III. Gas Phase Pyrolysis of the Pulegone Oxides A. Procedure. . . . . . . . . . . . . B. The Effect of Oxygen on the Reaction Rate . . . . . . . . . . . . . . . . C. The Effect of Increased Surface Area on the Reaction Rate . . . . . . . . IV. Liquid Phase Pyrolysis of the Pulegone OXides O O O O O O O O O O O O O O O O O A. Procedure. . . . . . . . . . . . . . B. Thermal Rearrangement of Pulegone Oxides . . .'. . . . . . . . . . . . C. Solvent Effects on the Thermal Re- arrangement of Pulegone Oxide II . . D. Rearrangement of Pulegone Oxide II in Presence of Base. . . . . . . . . E. The Effect of Free—Radical Initiators on the Thermal Rearrangement of the Pulegone Oxides I and II . . . . . . V. Miscellaneous Experiments. . . . . . . . . A. Preparation of 2,2,5,5-Tetramethyl- 1,5-cyclohexanedione (VII) . . . . . B. 2,2,5,S-Tetramethyl-l,5—cyclohexane- dione Bisethylenedithioketal (VIII). C. Raney Nickel Reduction of VIII . . . D. Attempted Preparations of 1,1,4,4- Tetramethylcyclohexane . . . . . . . 1. Lithium Aluminum Hydride Re- duction of 2,2,5,5—Tetramethyl- 1,3-cyclohexaneditosylate. . . 2. Wolff-Kishner Reduction of VII iv Page 46 50 52 54 57 6O 61 61 65 65 66 66 66 69 7O 71 72 72 76 76 78 78 80 TABLE OF CONTENTS - Continued Page E. Attempted Preparations of 1,1,4—Tri- methylcycloheptane . . . . . . . . . 81 1. Synthesis of Tetrahydro- eucarvone. . . . . . . . . . . 81 2. Tetrahydroeucarvone Ethylene- dithioketal. . . . . . . . . . 81 5. Reduction of Tetrahydro- eucarvone Ethylenedithioketal. 81 4. 2,2,5-Trimethylcycloheptane- 1,5-dione Bisethylenedithio- ketal (Synthesis and Raney Nickel Reduction). . . . . . . 82 5. Preparation of the Tosylhydra- zone of Tetrahydroeucarvone. . 82 6. Reduction of Tetrahydro- eucarvone Tosylhydrazone with Sodium Borohydride . . . . . . 85 7. Semicarbazone of Tetrahydro- eucarvone. . . . . . . . . . . 85 8. Reduction of the Semicarbazone of Tetrahydroeucarvone . . . . 84 F. Effect of Lead Tetraacetate on Diols of the 2,5-Dimethyl-2-acetylcyclo- hexanones (III and IV) and 2,2,5-Tri- methylcycloheptane-1,5-dione (V) . . 84 1. Lithium Aluminum Hydride Re- duction of a Mixture of III andIVOOOOOOOOOOOO 84 2. Lithium Aluminum Hydride Re— duction of V . . . . . . . . . 85 5. Reaction of the Diols from III and IV with Lead Tetraacetate. 85 4. Reaction of the Diol from V with Lead Tetraacetate . . . . 86 SUMMARY...................... 88 LITERATURE CITED. . . . . . . . . . . . . . . . . . 90 TABLE 1. 2. 5. LIST OF TABLES Solvent Dielectric Constants . . . . . . . . . Gas Phase Rearrangement of the Pulegone Oxides The Thermal Rearrangement of Pulegone Oxide Isomer I . . . . . . . . . . . . . . . . . . . The Thermal Rearrangement of Pulegone Oxide Isomer II. . . . . . . . . . . . . . . . . . . The Thermal Rearrangement of Pulegone Oxide Isomer II (In Air) . . . . . . . . . . . . . . Solvent Effect on Thermal Rearrangement of Pulgone Oxide Isomer II. . . . . . . . . . . . The Effect of Base on the Thermal Rearrange- ment of Pulegone Oxide Isomer II . . . . . . . The Effect of Radical Initiators on the Therm- al Rearrangement of Pulegone Oxides I and II . vi Page 52 66 68 68 68 7O 71 72 LIST OF FIGURES FIGURE Page 1. Energy Diagram . . . . . . . . . . . . . . . 25 2. Mass Spectra of the pulegone oxides I and II 59 5. Infrared Spectrum of fluorohydrin XII. . . . 41 4. N.M.R. Spectrum of fluorohydrin XII. . . . . 42 5. Infrared spectrum of fluorohydrin XI . . . . 44 6. N.M.R. Spectrum of fluorohydrin XI . . . . . 45 7. Infrared spectrum of 2,2,5-trimethylcyclo- heptane-1,5-dione. . . . . . . . . . . . . . 48 8. N.M.R. spectrum of 2,2,5-trimethylcyclo- heptane-1,5-dione, V . . . . . . . . . . . . 49 9. Mass spectrum of 2,2,5-trimethyl-1,5-cyclo- heptanedione (V) . . . . . . . . . . . . . . 5O 10. Infrared spectrum of keto—acid mixture, IX . 55 11. Infrared spectrum of keto-ester mixture Xa and Xb O O O O O O O O O O O O O O O O O O O 55 12. N.M.R. Spectrum of keto—ester mixture Xa and Xb O O O O O O O O O O O O O O O O O O O C O 56 15. Infrared spectrum of 2—hydroxy-2-isopropenyl- 5-methylcyclohexanone, XIV . . . . . . . . . 58 14. N.M.R. Spectrum of 2-hydroxy1—2-isopropenyl- 5-methylcyclohexanone, XIII. . . . . . . . . 59 15. Infrared Spectrum of 2-hydroxy-2—isopropenyl- 5-methylcyclohexanone, XIII. . . . . . . . . 62 16. N.M.R. spectrum of 2-hydroxy-2—isopropenyl- 5-methylcyclohexanone, XIV . . . . . . . . . 65 17. Mass spectra of the 2,5-dimethyl-2—acetyl- cyclohexanones III and IV. . . . . . . . . . 67 vii OIII- v .8 hill-.191...— LIST OF FIGURES - Continued FIGURE 18. 19. 20. 21. 22. Infrared spectrum of 2,2,5,5-tetramethyl- cyclohexane—1,5-dione, VII . . . . . . . . . N.M.R. Spectrum of 2,2,5,5-tetramethylcyclo- hexane-1,5-dione . . . . . . . . . . . . . . N.M.R. spectrum of 2,2,5,5-tetramethylcyclo- hexane-1,5-dione bisethylenedithioketal. . . N.M.R. spectrum of 1,1,4,4-tetramethylcyclo- hexane, XV . . . . . . . . . . . . . . . . . N.M.R. spectrum of 4—methylisopulegone, XVI. viii Page 74 75 77 79 87 CHART II. III. LIST OF CHARTS Favorski Reactions of the Pulegone Oxides. Mechanism for the Liquid Phase Rearrange- ment of the Pulegone Oxides. . . . . . . . Mechanism for Gas Phase Thermal Rearrange- ment . . . . . . . . . . . . . . . . . . . ix Page . 6 . 21 . 5O INTRODUCTION AND HISTORICAL During the past forty years, reactions of strained molecules have attracted the attention of many investigators. The synthesis of a variety of small ring compounds has stimu~ lated investigations of their chemical and photochemical behavior. The three-membered ring carbocyclic and hetero- cyclic systems have undergone extensive study; this is not surprising since the three-membered ring structures.are particularly strained and are potential sites for bond rupture upon thermal and photochemical excitation. Among the hetero- cyclic three-membered ring structures which have been widely- studied are thewaziridines and oxiranes. Thermal transfor- mation of oxiranes to aldehydes and ketones have been reported (1,2,5). Much of the published literature on d,8-epoxy ketones concerns acid- and base-catalyzed rearrangements (4), whereas comparatively few investigations have been reported concerning the thermal and photochemical behavior of these substrates. House (8,9,10,11) has investigated the acid catalyzed rearrange- ments of several a,8—epoxy ketones with borontrifluoride etherate; each of these rearrangements resulted in the forma- tion of a B-diketone. In recent years, H. E. Zimmerman has studied photochemical transformations of a,8—epoxy ketones and offered mechanistic interpretations. Recently, he re- ported that 5,4-epoxy-4—phenyl-2-pentanone was converted to 1—phenyl—2-methyl-1,5—butanedione upon irradiation and postu- lates a novel.1r2rmethyl migration (5). Reusch, Dominy, and Johnson (6) have reported.photochemical rearrangements of 5,4-epoxy-4-methyl-2-pentanone and isophorone oxide to give 5-methylpentane-2,4-dione and 2-acetyl-4,4—dimethylcyclo- pentanone, respectively, as the major products. They also reported that upon irradiation of the phenyl analog of iso- phorone oxide 5-phenyl-5,5-dimethyl-2,5-epoxycyclohexanone and the pulegone oxides, 2-benzoyl—4,4-dimethylcyclopentanone and the diastereoisomeric 2,5-dimethyl—2-acetylcyclohexanones, respectively, were isolated. This thesis reports the behavior of the pulegone oxides under thermal and acidic conditions. The thermal reactions were studied with the neat liquid and also in the vapor phase. A pronounced solvent effect was found in the thermal re— arrangement and will be discussed in detail. The pulegone oxides were first prepared by Prilejaeff (12) in 1927 by the oxidation of commercially available pulegone (obtained from oil of pennyroyal) with perbenzoic acid. Treibs (15) also synthesized pulegone oxide in 1955 by the treatment of pulegone with hydrogen peroxide in the presence of alkali. A few years later, Prilejaeff reported that upon heating a mixture of pulegone oxides at 2000 a "liquid isomer" was produced. In 1957, Pigulevski and Mironova (14) observed a trans- formation of an optically active pulegone oxide mixture, while heating under nitrogen at 2000, to a more strongly dextrorotatory liquid; the Raman Spectra was similar to that of the original epoxide after 8 hours of heating. They postulated from this limited data that the change in optical rotation can be rationalized by epimerization of the oxirane ring.. This interpretation was shown later to be erroneous. Reusch and Johnson (7) synthesized pulegone oxide by the method of Treibs (15) and reported the presence of stereo- isomers I and II. These isomers were isolated by vapor phase chromatography and identified by infrared and n.m.r. spectroscopy; an effective separation and purification was achieved by fractional crystallization. Upon heating at 2000, two liquid diastereoisomeric products, III and IV, were isolated and identified as reported in the doctoral thesis of Calvin K. Johnson (16). III IV In 1962, Pigulevski (15) reinvestigated the thermal rearrangement of the pulegone oxides and reported that when heated in a stream of nitrogen at 2000 for 44 hours, optical rotation of the liquid product increased. After fraction- ation, two compounds-were isolated and identified as stereoisomers of VI formed by rearrangement of pulegone oxide. This interpretation is also erroneous. Boo VI The configurations of isomers I and_II were assigned by Johnson (16) on the basis of three arguments. Two of the arguments used involved the interpretation of the n.m.r. spectra and optical rotatory diSpersion measurements. The isomers were found to have specific rotations of opposite sign as well as opposite Cotton effects near 500 mu. Isomers I and II showed positive and negative Cotton effects, respectively, as determined from the rotatory dispersion curves. Application of the octant rule (17) led Johnson to assign structures Ia and 11a: Ia IIa Recently, Djerassi et al. (18) discovered that cyclo- propane and oxirane rings make contributions to the Cotton effect which are opposite in sign to those made by alkyl groups; consequently, the true configurations of the pulegone oxides are probably represented by I and II. Chemical evidence has also been found to substantiate these structural assignments. Cavill (19) has recently re- ported a Favorski rearrangement of pulegone oxide. In his study he uses pulegone oxide having a m.p. 490-50O which corresponds to neither of the pulegone oxides I nor II des- cribed in this thesis; thus, it may be assumed that he was working with a mixture of the pulegone oxides. Cavill has recently carried out the same reactions with the pure isomers (20). Reusch and Mattison have also investigated the Favorski reaction using pulegone oxide isomers I and II; the reaction scheme is described in Chart I. Compounds Ib and 11b are formed stereospecifically via a cyclopropanone Favorski intermediate. This data lends credence to the configurational .assignments of I and II. CHART I Favorski Reactions of the Pulegone Oxides 1)NaOMe,Glyme 2)CH2N2 ---C02R + Other Products I ,H ’ H ,H o COgMe J, 1)NaOMe,Glyme ,.” _¥ > \ + o 2)CH2N2 \\m ' o + C02M8+ + Other Products RESULTS AND DISCUSSION I. Products From the Rearrangements of the Pulegone Oxides When heated in Pyrex vials at 2000, pulegone oxide isomers I and 11 formed products resulting from rearrange— ments, fragmentation, and epimerization. Johnson (16) has discussed the isolation and identification of the two major products, III and IV, which were formed in 25% and 50% yield, respectively. The same compounds were isolated when the thermal reaction was performed in oxygen-containing solvents and were identified by a compariSon of the infrared and n.m.r. spectra with authentic samples (22). Pure samples of III and IV were obtained by preparative gas chroma- tography. >13: /_ III IV In the vapor phase rearrangement of both pulegone oxide isomers, the major product was found to be an isomer of III and IV as indicated by the elemental analysis and the mass spectrum. The compound was a white crystalline solid, V, melting at 500-510 and yielding a disemicarbazone having a melting point at 2150-2170. The n.m.r. spectrum of V indi- cated the presence of three methyl groups; one appeared as a doublet at T 8.96 (J = 6.0cps) and the gem-dimethyl group as a singlet at T 8.86. O V VI Pigulevski and co—workers reported (25) the isolation of two isomeric liquids from a thermal rearrangement of pulegone oxide; it was suggested that the compounds were dediketones and structure VI was assigned. The experimental evidence which was presented to substantiate these views involved lithium aluminum hydride reduction to the vicinal diol and subsequent cleavage with lead tetraacetate. The products from the thermal reaction formed disemicarbazones (m.p. 2150-2170; 2480-2500). The melting points of the semicarbazone of III (211.50-212.50) and the disemicarbazone of IV (2220-2250) were reported by Johnson (16). The lack of published experimental and spectral data prohibits one from making a rigorous comparison of the work of Pigulevski and the data presented in this thesis. The infrared carbonyl absorption for III, IV, and V (1704, 1701, and 1682 cm‘l) were too low to substantiate the presence of an abdicarbonyl grouping. After lithium aluminum hydride reduction of III, IV, and V, the respective diols were treated with lead tetraacetate; the infrared and n.m.r. spectra of the products gave no indication of any cleavage reaction. One approach to demonstrating the presence of a seven- membered ring in V involved reduction to the hydrocarbon 1,1,4-trimethylcycloheptane. This hydrocarbon would also be available by the reduction of tetrahydroeucarvone. The reduction was tested, using as a model compound 2,2,4,4-tetra- methyl-1,5-cyclohexanedione, VII, which was synthesized from dimedone. The bis—ethylenedithioketal (VIII) was formed in good yield and upon reduction with Raney nickel, the hydro- carbon, 1,1,4,4-tetramethylcyclohexane was isolated in a 75% yield. When the reaction was attempted with V, none of the hydrocarbon was isolated. However, in View of the sensitivity of cycloheptane—1,5—diones to cleavage (24 and 25), it seems likely that ring opening has occurred during thio- ketal formation. Several variations of the Woff-Kishner reduction were used to effect the formation of 1,1,4-tri- methylcycloheptane but none were successful. VII VIII 10 In order to simplify the complex n.m.r. Spectral pattern in the low methylene region of the spectrum, a base catalyzed deuterium exchange was attempted with V (replacement of the hydrogens as to the carbonyl groups should simplify the n.m.r. spectrum). Unfortunately, only a cleavage product and partially deuterated V were isolated; the cleavage product was identified as a keto acid, IX, whose infrared spectrum is shown in Figure 10. 9 (CH3)2CD(H)—C—CH2-gH-CH2CH2-C02H H3 IX An independent method for the synthesis of V was devised using the method of House (9) who has reported the rearrange- ment of several d,8-epoxy ketones to B-diketones in the presence of borontrifluoride etherate. These B-diketones are apparently formed by a 1,2—acyl migration to a carbonium ion center (e.g., equation 1). 0 CH0 / 0 Befiggne >: O (1) Some alkyl (aryl) migration was also observed. When these reactions were run in ether solutions, the major products isolated were identified as fluorohydrins which are believed to be intermediates in the formation of the B—diketones. ‘l 11 A mixture of the pulegone oxides I and II, when treated with borontrifluoride etherate, formed two major products; these products were colorless liquids whose infrared spectrum showed a strong hydroxyl absorption at 5580 cm‘1 and a carbonyl absorption at 1715 cm'l. Upon standing for a period of six months, the product mixture turned dark brown with an accompany- ing pungent irritating odor. When isomer I was treated with borontrifluoride etherate in benzene, a single product was formed as indicated by v.p.c. analysis. In a similar reaction isomer 11 gave an isomeric substance as the major product. These compounds were found to be thermally unstable when heated to 1500, and rapidly de— composed at this temperature to a dark brown tar with evolu- tion of hydrogen fluoride. The tarry material, when passed through a silica-gel chromatographic column, yielded a white crystalline solid identified as V (50%§yield) by n.m.r. and infrared analysis, and a mixed melting point determination. Upon treatment of each of the fluorohydrins with methanolic potassium hydroxide, the respective starting materials, I and II, were isolated in almost quantitative yield. A fluorine resonance signal was observed with the Varian HR-60 research spectrometer. Pure samples of each of the fluorohydrins were obtained by preparative v.p.c. and the following structures and configurations are assigned: u 12 BFB'OEtg \ Benzene (2) BF3 '0Et2 Benzene \/ (5) The n.m.r. spectrum of XI (Figure 6) Shows the Cl-methyl doublet centered at T 9.07 (J = 5 cps) which is overlapped with the methyl doublet at T 8.92. The Ca-methyls are each split by fluorine into two doublets centered at T 8.92 and 8.56 (J = 15 cps), respectively. In structure XII, the C1- methyl group doublet appeared at T 8.92 (J = 6 cps); the two CB-methyl doublets were observed at T 8.88 and 8.55 (J = 18 cps). Thus, one methyl group seems to be in the deshielding region of the carbonyl group (26). The configurations XI and XII were assigned due to the stereospecific formation of the fluorohydrins from the parent pulegone oxides I and II. The conformations XI' and X11' are fixed by hydrogen-bonding between the carbonyl and the hydroxyl group which functions as a stabilizing factor. 15 XII' It is interesting to note the difference in chemical shift of the Cl-methyl groups in the two isomers; this dif- ference suggests that one is mainly axial and the other equatorial. It is known in the hydrocarbon androstane that the Clg-methyl proton resonance signal for A/B trans (axial methyl) is observed at a higher field than the equatorial methyl (A/B cis) (27). This data lends credence to the conformational assignments of XI and XII. The infrared spectra of XI (XI') and XII (XII') indi- cated strong intramolecular hydrogen bonding; this fact was evident when there was no change in the hydroxyl band absorp- tion frequency with change in concentration (28). Both XI and X11 exhibited strong -OH stretching vibrations at 5440 and 5450 cm‘l, respectively. Since XI has an axial methyl group, it would have a greater tendency to exist partly in the opposite chair form and one should observe a free —OH stretch- ing mode. Indeed a weak-OH band was found at 5580 cm‘1 in the spectrum of XI. Suitable mechanisms for the formation of the fluoro— hydrins are described in equations 4 and 5. 14 BF3 (4:). BF3 (5) In these reactions, the oxirane oxygen is first coordinated with BF3 followed by the attack of a nucleophilic fluoride ion to open the complexed oxirane ring with stereospecific formation of the respective fluorohydrins. When the fluorohydrins were decomposed thermally, re- arrangement occurred to form V and can be rationalized by the following mechanism, 6: 7 Km (6) The reaction of the pulegone oxides with p-toluenesulfonic acid in benzene formed products which were identified as 15 unsaturated hydroxy ketones, XIII and XIV. Infrared analysis of both showed strong hydroxyl stretching at 5470 cm’l, indi- cative of hydrogen bonded species. Medium absorption bands located at 5080, 1820, 1645, and 910 cm-1 indicate the presence of.a.terminal methylene. A pure sample of XIII was obtained by the rearrangement of.pulegone oxide 11 on an acid-surfaced silicone v.p.c. column; the eluted material was collected and found to contain a single component, XIII. The hydroxyl stretching frequency at 5470 cm‘1 contained a shoulder at 5560 cm‘1 which is characteristic of free hydroxyl. The carbonyl absorption band was observed at 1712 cm-1 (Figure 15). If one assumes that the unsaturated hydroxy ketones are formed stereospecifically from the parent pulegone oxide then structure XIII can be assigned to the product obtained from pulegone oxide II (equation 7). (c) H 6 _fl_; \ 0 <7) H0 \ [Ts (b) (a)x111 The n.m.r. data (Figure 14) supported the assignment of the configuration XIII and furthermore allows one to make a possible conformational assignment. In the 2-isopropenyl group, the methyl signal (a) appeared at T 8.28 as a broad singlet which, when expanded at 50 cps, was resolved into a 16 pair of doublets. The terminal methylene hydrogens (b) were observed as a broad peak at T 4.82; at 50 cps it was a poorly resolved quartet (J = 1 cps). The Cl-methyl signal (c) was quite complex and appeared as two poorly resolved doublets at T 8-88 and 9.01 which remained unresolved at 50 cps. One would expect a doublet from the Cl-methyl; however, the presence of two doublets seem to suggest the occurrence of an equilibrium mixture of two possible chair conformations, XIIIa and XIIIb. HO \ \ T ....-////\/ )\ XIIIa XIIIb Preparative v.p.c. was also used to obtain a pure sample of XIV as a product of the p-toluenesulfonic acid catalyzed rearrangement of I. The infrared spectrum (Figure 15) was almost identical with that of XIII with the exception of the presence of overlapping carbonyl peaks at 1710 and 1695 cm'l, and no free hydroxyl was observed. Similarly, if one assumes a stereospecific formation of the unsaturated hydroxy ketone, XIV, then the following configuration (equation 8) would be reasonable: (8) 17 The n.m.r. spectrum of XIV (Figure 16) exhibited a broad 2-propylene methyl (a) signal at T 8.50 which was seen as two doublets (J = 0.8 cps) at 50 cps. The terminal methylene protons (b) appeared as two distinct broad peaks at T 4.80 and 4.90; when expanded at 50 cps, the lower field signal appeared as a complex multiplet in which 8 lines (separation 25 0.8 cps) could be seen. The higher field signal (T 4.90) ap— peared as a complex overlapped multiplet when expanded at 50 cps. The Cl-methyl group exhibited two poorly resolved doublets; the chemical shifts are estimated as T 8.95 and 8.86. With these spectral data, one would again suggest an equilibrium between chair conformers XIVa —--)~XIVb. The mechanisms for the stereospecific rearrangement catalyzed by p—toluenesulfonic acid are described in equa- tions 10 and 11. (11) 18 The initial step involves the protonation of the oxirane oxygen with subsequent ring opening by heterolytic cleavage of the Ca-oxygen bond forming the stable tertiary carbonium ion. With subsequent loss of.a proton, compounds XIII and XIV are formed from pulegone oxides II and I respectively. In order to further investigate the existence of equilibration between conformers (vide infra) of XIII and XIV, a known model compound was chosen, 4—methylisopulgone XVI. Djerassi (22) prepared a mixture of the 4-methyliSOpulegoneS by methylating (+)-pulegone to obtain an 85/17 ratio of XVI/XVII and assigned the configurations shown in equation 12. The n.m.r. Spectrum of XVI (Figure 22) exhibited the following signals: T 5.15, 5.25 (two multiplets- '221 cps); T 8.55 (unresolved quartet— Jtt?1 cps); T 8.90 (singlet); T 8.98-9.20 (complex multiplet). Again, the single expected doublet for the Cl-methyl protons was not observed (T 8.98- 9.20); instead, a complex multiplet appeared which suggests an equilibration of conformers between the opposite chair conformations. 19 II. Mechanisms for the Thermal Rearrangements of the Pulegone Oxides A. Ligyid Phase Reaction Thermal reactions of the pulegone oxides I and II have been interpreted by other investigators as yielding products resulting from epimerization and molecular rearrangements (12,14,15). The predominant products of the liquid phase re- arrangements have been identified as the diastereoisomers of 2,5-dimethyl-2-acetylcyclohexanone, III and IV (16). The ratio of III and IV formed in the liquid phase rearrangement was found to be 1:2, respectively, with either isomer I or II as the reactant. Reusch et al. (6) have reported that photolysis of the pulegone oxides yields the acetyl compounds III and IV. An investigation of the thermal chemistry of pulegone oxide suggested the possibility that these photochemical reactions proceed by way of vibrationally excited species. Zimmerman and Schuster (29) propose that n—-€>T* photochemical reactions do not generally parallel thermal reactions. Johnson (16) has discussed the rearrangement of the pulegone oxide isomers I and II and suggested two mechanistic possibilities to explain the formation of III and IV. The first mechanism involved a concerted process (e.g. equation 15); however, a concerted mechanism would appear to be highly unlikely, because the products formed would result from a stereospecific rearrangement (e.g., from isomer I, one Should 20 II IV isolate only III and similarly from isomer II, only IV should be formed). This does not agree with the common product ratio, III/IV (1:2), obtained regardless of the starting iso- mer. The second mechanism is outlined in Chart II; the pule- gone oxide isomers I and II are thermally cleaved to a di— radical intermediate which can epimerize, revert to starting material, or decompose with the formation of two free—radicals which could couple within the solvent cage to give the re— arrangement products III and IV. The 1,2-methyl radical migration has been described with the methyl radical as a discrete intermediate, which agrees with Zimmerman's postulate (50) that there is no bonding in the transition state of a 1,2-alkyl shift to a radical center. Reports of such re- arrangements are rare and often ambiguous (5,51,52,55). 21 CHART II Mechanism for the Liquid Phase Rearrangement of the Pulegone‘Oxides _%H _ AC ‘ ( 1° :3° II 1 1 0° L L /\\ p x /«\\ '/ r- z/H ( K g . O , O /:0 CH3 \ L 1 III IV Several mechanisms can be visualized to explain the thermal rearrangement and isomerization of the pulegone oxides. A homolytic cleavage of the C4-oxygen bond (path a, equation 14) leads to a resonance stabilized diradical inter- mediate (C); whereas, homolytic cleavage of the Ce—oxygen bond produces a stable tertiary alkyl and an alkoxy di- radical (D). Walling (55,54) has described some free- radical intermediates as being polar in nature with the 22 free-radicals being considered an electron-deficient species. (14) a b \ 7 (15) ,’H :7 (16) O H 42* a b \ 0 I7 (17) G (I) to be less stable than (H) One might predict intermediate (C) (D), due to the inductive effect of the carbonyl group adja- cent to the C4 radical. However, intermediate (D) neither accounts for the formation of the rearrangement products nor does it explain the observed epimerization Since the assym- metric center at C4 remains intact. Path (a) is therefore 25 the preferred mode of homolytic cleavage. There appears to be a strong solvent effect in the presence of electron- donating solvents which results in the stabilization of (C); in the liquid phase reactions, the pulegone oxide is its own solvent. The solvent effects are discussed in consider— able detail in section C of this thesis. An energy diagram for the homolytic cleavage mechanism is suggested in Figure 1. Formation of (C) Formation of (D) ---- Formation of (C) (solvated) \ / Energy P.O. I or II Products Figure 1. Energy Diagram The activation energy required for the transformation of I or II to the diradical intermediate (C), A Ec, appears to be lowered due to solvation by electron donating species. 24 Therefore, path (a) becomes the more favorable path for the decomposition. Another mechanism which involves the homolytic or heterolytic cleavage of the C4-C8 bond to form the diradical (path a) or the ionic (path b) intermediates (E) and (F) is illustrated in equation 15. These intermediates adequately explain the isomerization products but do not provide a suitable explanation for the rearrangement process. Also, the C4-oxygen bond would be expected to break in preference to the C4-C8 bond inasmuch as the carbon—oxygen bond of ethers normally have lower bond dissociation energies than carbon-carbon bonds (56). The mechanism suggested in equation 16 is not especially attractive because it involves homolytic cleavage of a strong bond to form radicals which are not particularly stable (G). The heterolytic cleavage of the C4—oxygen bond (equation 17a) to form (H) is also an unlikely possibility since it results in the formation of a carbonium ion adjacent to a carbonyl group; nevertheless, this intermediate would certainly explain the formation of the observed thermal rearrangement products by a 1,2—methyl migration. The epimerization is also ade- quately rationalized by a rotation around the C4-C8 bond and subsequent recombination. The acid-catalyzed rearrangements of several a,8—epoxy ketones to form B—diketones are known to proceed by the formation of a positive site 8 to the carbonyl group (11). In this laboratory, we have shown that 25 acid-catalyzed reactions using borontrifluoride and p- toluenesulfonic acid produce compounds XI, XII, XIII, and XIV; the products III and IV were not observed. These products were formed by opening the oxirane ring at C8; thus, one can assume that the thermal reaction.iS :not acid catalyzed. Furthermore, reactions of the pulegone oxides in the presence of bases remained relatively unchanged and the only products observed were those found in the neat liquid phase reaction (Table 7). This accumulation of experimental facts would lead one to argue the absence of ionic species as intermediates. When the thermal reactions were run in the.presence of oxygen or t-butyl hydroperoxide, little effect on the reaction rates was noted, and what little was observed could be in- terpreted as retarding; there was no formation of tars or other new products. In the presence of radical initiators such as 2,4-dimethyl-2-formylcyclopentanone and 2,2'-azoisopropane, again there was no obvious effect on the reaction rates (Table 8). These data tend to rule out the occurrence of a radical chain mechanism. Also, one would not normally predict a radical chain mechanism here, since the pulegone oxides lack the reactive oxirane hydrogens. Further evidence for the absence of acid catalysis was reported (15) when the addition of powdered Pyrex glass slowed the reaction indicating that acid sites on the Pyrex surface were not causing the reaction. The experimental facts presented in the preceding sections do not permit the unequivocal assignment of a mechan- ism for the rearrangement to III and IV. Using the general 26 representations [A] and [B] for the reactive intermediates, the following mechanisms are formally possible: [B] V‘_‘ II k 2 gun—+9 k3 This is logically equivalent to having I be the source of the rearrangement. II is chosen to explain the mechanism because the experimental data indicate it to be more reactive than I. Starting with pure I it is found that the rate of con- version to g_is significantly faster than isomerization to II by a factor of ~’2 [Table 5]. For this to be true in case 1, 27 the rate constant for the transformation of II ---9- [A] would necessarily have to be quite large compared with the other rate constants (e.g. k3 >> kl and k2). However, it has been observed that isomerization of II ---> I (starting with pure II) occurs about as rapidly as the rearrangement of II --e>P (e.g. kgiv k3); therefore, this inconsistency suggests that case 1 fails to explain the experimental re- sults and thus can be ruled out. Mechanism 2_is consistent with the facts, i.e., k2 > k1, k3;3=k_1 > k-2; these inequalities do not appear to be large (e.g.‘v’2x). Mechanism 5 is also consistent with the facts provided that the amount of isomerization occurring via [B] does not exceed that via [A]. The facts presented above do not clarify the reaction pathway nor the exact nature of the intermediate involved; one can only conclude that this is a true "thermal reaction." B. Gas Phase Reaction Thermal rearrangements of I and II effected in the gas phase formed V as the predominant product (65%) along with small amounts of III and IV. In order to rationalize the formation of product V, formation of reactive intermediates [C] and [C‘], and subsequent rearrangement of the acyl group is proposed. In the gas phase thermal reaction no isomerih zation was detected regardless of the starting isomer. Experimental results Show also that the Slower reacting iso- mer I produced only traces of III and IV, whereas, with 28 [C] [C'] isomer 11 about 20% of the acetyl isomers III and IV were fbrmed. The formation of III and IV is rationalized in the preceding section of this thesis. The following formal mechanisms logically explain the path of the reaction. I _£is. [C] k 1 \a ‘ _ 41, III + Ivé— [A] {K v (18) k;3 k II —ie [C'] k-2 When the gas phase reaction was carried out with pure II, it was found that the rate of conversion to V was markedly greater than the formation of III and IV (”V5x); thusly, the rate constant for the transformation of II --9—C is much greater than the conversion of II --9»A, i.e., k2 >> kg. The slower reacting isomer, 1, was observed to produce V at a significantly slower rate than I, and formed only traces of III and IV. Therefore, to be consistent with the observed experimental results, the rate of the conversion II -—9-A is 29 greater than I —-->A and II —-->C is greater than I --->C, (k3 > k1 and k2 > k1). The rate of isomerization of I --9’ II or II -->-I appears to be negligible in the gas phase reaction, thus k_3 95 k_4 £5 0. The reaction pathway illus— trated in equation 18 appears to be consistent with these facts although the true nature of an intermediate has not been clearly illucidated. The thermal reaction in the vapor phase was Slightly retarded in the presence of oxygen. When the reaction was performed without degassing, the oxygen present decreased the rate of reaction by 10% (Table 2). The reaction was also found to be surface catalyzed. When the surface area was increased 100% by the addition of Pyrex tubing, the re- action rate was increased by 25%. These data are insufficient to permit a choice between a diradical or an ionic inter- mediate for the formation of V; however, two mechanistic possibilities are presented in equations 19 and 20. (19) 50 Equation 19, involves a homolytic cleavage mechanism with the formation of a stable diradical intermediate with subse- quent 1,2-acyl radical migration. The latter mechanism pro- ceeds via a heterolytic cleavage path to form the stable ionic intermediate with a 1,2-acyl migration to the stable positive center. Rondestvedt (55) has reported catalytic decomposition of 1,5-dioxanes by acidic sites on silica gel. House (11) has shown that the borontrifluoride etherate catalyzed rearrangement of isophorone oxide gives almost exclusively 2-formyl-2,4,4—trimethylcyclopentanone (equation 21). O H C-H O BF3'0(Et)2 > (21) Isophorone 2-Formyl-2,4,4-tri— Oxide methylcyclopentanone House and Wasson (8) have also observed the isomerization of 2-benzalcyclohexanone oxide with borontrifluoride etherate to 2-phenyl-1,5-cycloheptanedione (equation 22). Both re- arrangements involve 1,2-acyl migration. O o H BF3-0(Et)g_ \\ (22) Benzene \K\V/ O 2-Benzalcyclohexanone 2-Phenyl-1,5-Cyclo- Oxide heptanedione 51 When either pulegone oxide isomer was treated with borontrifluoride etherate, isomeric fluorohydrins XI and XII were formed, and these rearranged to V upon heating. Pyrex glass contains boric acid, and alumina which provide acid sites on the Pyrex glass; hence, the observed surface effect is consistent with an acid catalyzed reaction. A schematic representation of the proposed acid catalyzed surface re- arrangement is described in Chart 5; however, this mechanism is also consistent with the free radical mechanism shown in equation 19. CHART 5 Mechanism for Gas Phase Thermal Rearrangement \ \ \ (566 I Pyrex Glass \ -K :]\\((LL§ surfacetda " \Vp / ] _\._\T\\ \ \\\\ \\ C. Solvent Study In an attempt to delve further into the nature of the reactive intermediate [A], thermal rearrangements were con— ducted in a series of solvents of varying polarity. Table 1 32 lists the solvents used and their dielectric constants. Table 1. Solvent Dielectric Constants Solvent Dielectric Constant (6) Cyclohexane 2.02 Cyclohexene 2.20 p-Cymene 2.50 Dioxane 2.2 Cyclohexanone 18.5 t-Butyl Alcohol 10.9 If the reactive intermediate were an ionic species, then one would predict an increase in reaction rate with an increase in solvent polarity. When solutions of 50% pulegone oxide in the various solvents were subjected to thermal re- arrangement, only products III and IV were formed; all attempts to correlate reaction rates were unsuccessful because of the erratic nature of the data. The reaction in dioxane, which would be expected to exhibit reaction rates similar to the hydrocarbons was found to react as fast as cyclohexanone. When solutions of 5% pulegone oxide were submitted to thermal reaction conditions, there was a pronounced solvent effect operating to determine the mechanistic pathway of the reaction. The reactions in the oxygen-containing solvents led to products III and IV exclusively; in hydrocarbon solvents, the only rearrangement product observed was V. 55 Russell (56) has reported profound solvent effects in free-radical reactions. These effects were observed in the photochlorination of 2,5-dimethylbutane where he found that aromatic solvents form a complex with the chlorine atom; the complexed chlorine atom is much more selective than the free atom and consequently the major product is the tertiary chloride. These solvent effects were noticeably absent when aliphatic solvents were used. Walling (57 and 58) also ob- served solvent effects in the decomposition of t-butoxy radicals; the radical is stabilized by polar solvents and thus survives long enough to give increased fragmentation. Arguing from these facts, the electron-deficient free— radical at C4 may be stabilized by solvent complexation (equation 25). The activation energy for the formation of Q, A BC, then becomes lower than that leading to B, A Eb, Figure 1 (equations 14 and 25). W c \\ //, JA . Solvent stabilized Previous photochemical studies suggest that thermal rearrangement is closely related to the photochemical re— arrangement where it is known that alkyl groups migrate in 54 preference to phenyl groups (6 and 50). This fact can be explained by using the Reusch—Zimmerman postulate that there is more bond—breaking than bond-making in the transition state. Several equivalent structures may contribute to the transition state of the solvated species (equation 24), each of which could be stabilized by the participation of oxygen- containing solvents. (24) In hydrocarbon solvents, there is no possibility for the stabilization of Q; consequently, A Ea becomes greater than A Eb, and the reaction pathway leading to Q_becomes the pre- ferred mode of decomposition (equation 14). The surface mechanism described in Chart 5 appears to be a better repre- sentation than those of equation 14b or equation 20. Further— more, a free radical surface mechanism is more consistent with the results observed in the solvent studies. From the preceding discussion, it becomes readily appar- ent that the thermal reactions involving 50% pulegone oxide II solutions mentioned at the beginning of section C obviously did not show true solvent effects, but are related to the neat liquid reactions. EXPERIMENTAL I. General Procedure A. Apparatus 1. The infrared Spectra were run on the following recording Spectrophotometers: Perkin Elmer, Model 21 Perkin Elmer, Model 257 Unicam, SP-200 Beckman, IR-5 Sodium chloride cells were used for each Spectral determin- ation. 2. Proton magnetic resonance spectra were measured using a Varian, A-60, high resolution spectrometer. All spectra were obtained at 60 mc using tetramethylsilane as an internal standard. 5. Vapor phase chromatographic analyses were performed with an Aerograph A-90-P gas chromatograph. B. Preparation of Pyrex Vials For the liquid phase reactions, Pyrex tubing (8 cm.-OD; 4 cm.-ID) was cut into 20 cm sections and placed in a large cylinder, covered with concentrated nitric acid and heated for 15 hours on a steam bath. The tubes were rinsed five 55 56 times with distilled water and covered with ammonium hydrox— ide for 50 minutes. The tubes were again rinsed five times with distilled water and dried in an oven at 1200 for 24 hours. After drying, the tubes were sealed at one end and placed in a desiccator. For the gas phase reaction, Pyrex combustion tubes (length- 55.5 cm; ID- 1.9 cm; OD- 2.5 cm) were used. The tubes were cleaned as described above and dried at 2000 in a combustion furnace (Presicion Scientific Co.) just prior to use . C. Melting Points Melting points were determined on a Reichert hot-stage and are uncorrected. D. Microanalysis All microanalyses were performed by the Spang Micro- analytical Laboratory, Ann Arbor, Michigan. E. Furnace The sample vials for liquid phase reactions were heated in an aluminum block furnace. The combustion tubes used in the gas phase reactions were heated in a combustion tube furnace (Precision Scientific Co.). The temperature was within $.20 of the reported values. F. Pulegone Oil of Pennyroyal obtained from Fritzsche Brothers, Inc. was distilled under nitrogen to obtain the pulegone used in 57 the synthesis of the pulegone oxides. The fraction boiling at 760-820/15 mm was collected and used. G. Synthesis of Pulegone Oxides To a 1500 ml three—neck flask equipped with a stirrer, a dropping funnel, and a thermometer were added 90 gms of pulegone, 480 ml of methanol, and 105 gms of 50% hydrogen peroxide. The mixture was cooled to 150 in an ice bath and a solution of 5.0 gms of sodium hydroxide in 50 ml of water was added dropwise with vigorous agitation over a period of 15 minutes. Stirring was continued for 4 hours with the temperature maintained between 200-250. The reaction mixture was poured into 1200 ml of saturated salt solution and ex- tracted with 2400 ml of ether. The ether extract was washed with 500 ml of water, dried over anhydrous magnesium sulfate, and removed at reduced pressure. The residue was distilled through a vigreux column and the material boiling between 850-94O was collected and analyzed by vapor phase chroma- tography. The v.p.c. analysis indicated a 67% yield of the diastereoisomeric pulegone oxides (29% I and 71% II). Infrared and n.m.r. Spectra are shown in the doctoral thesis of C.K. Johnson (16). Mass spectral data are shown in Figure 2. H. Isolation of Pulegone Oxide Isomers I and II Isolation of the pure diastereoisomers was accomplished as described by C. K. Johnson in his doctoral thesis (16). 58 I. Purity and Purification of Solvents 1. Solvent and Reagent Purity 2,6-Lutidine- Reagent Grade, Eastman Organic Chemi- cals Co. 1,4—Dioxane- Spectroquality Reagent, Matheson Coleman & Bell Cyclohexane- Spectroquality Reagent, Matheson Coleman & Bell - Cyclohexene- Chromatoquality Reagent, 99+%, Matheson Coleman & Bell p-Cymene— (Terpene-free)-Matheson Coleman & Bell t-Butyl Hydroperoxide- Wallace and Tiernan Methyl Iodide- Columbia Organic Chemicals 2,2'-Azoisopropane- Merck, Sharp, and Dohme (Canada Ltd.) 2. Solvent Purification (a) Cyclohexanone Technical grade cyclohexanone was purified through preparation of the bisulfite addition compound (59) and de- composition with 10% sodium hydroxide solution. The liberated cyclohexanone was separated and the aqueous layer was satur- ated with salt and extracted with ether. The ether extract was combined with the ketone layer and dried with anhydrous magnesium sulfate. After removing the ether the residual cyclohexanone was distilled, collecting the fraction which boiled between 1550-1550. (b) N,N-Dimethylaniline A crude sample of commercial dimethylaniline was refluxed for 1 hour with acetic anhydride and then allowed to 59 cool. The mixture was distilled and the fraction boiling at 1020-1040/12 mm was collected; v.p.c. analysis showed a single peak- (c) Triethylamine Triethylamine was purified by distillation from potassium hydroxide pellets. The portion distilling at 890-900 was collected and used immediately. (d) 2-Formyl-2,4,4-Trimethylcyclopentanone A sample of this compound was synthesized by the method of House (11) and distilled. The fraction boiling between 500-510/2 mm was collected. v.p.c. analysis showed one peak. Figure 2. Mass Spectra of Pulegone Oxides I and II m/e % Base Peak m/e % Base Peak Pulegone Oxide I 55 26.6 84 50.0 55 100.0 86 60.0 56 56.7 97 55.5 67 65.4 98 50.0 68 25.5 111 60.0 69 80.0 125 50.0 70 100.0 126 45.5 81 50.0 155 86.5 82 50.0 168 45.5 85 60.0 Pulegone Oxide II 55 94.2 97 41.5 56 55.5 98 55.5 67 70.7 111 55.0 69 55.5 125 76.5 70 100.0 155 76.5 85 58.8 168 55.0 86 55.5 Intensity of P+1 Peaks: P+2 Peaks Calcd. for C10H1502: 11.14% Calcd. - 0.96 Found: Isomer I — 11.20% Found (I)- 1.08 Isomer II— 11.25% Found (II)— 1.28 40 II. Acid Catalyzed Rearrangements A. Rearrangement of Pulegone Oxide Isomer I With Borontrifluoride Etherate Two grams of the pure isomer I (m.p.-57o-59O) was dis- solved in 50 ml of dry benzene in a 100 ml Erlenmeyer flask. To this solution was.added 1.9 ml of freshly distilled boron— trifluoride etherate and the resulting solution was allowed to stand for five minutes at room temperature. The benzene solution was washed with 5% sodium bicarbonate solution and distilled water until the wash layer was neutral. The ben- zene layer was dried over anhydrous magnesium sulfate and flash evaporated to leave a pale amber residue (1.90 gms). Vapor phase chromatography (6 ft., 5% PDEAS, 1580) indicated the absence of unreacted isomer I and showed the presence of one major component. The product was distilled through a short-path distillation apparatus (b.p.-75°—780/2 mm). A pure sample was collected by preparative v.p.c. and had the following properties: V55i4 5450, 1708, 1115 cm‘l. The infrared Spectrum is shown in Figure 5. The n.m.r. spectrum displayed a pair of doublets centered at T 8.5 and 8.88 (J = 18 cps); a doublet at T 8.92 (J = 6 cps); a Singlet at T 6.00 and a complex multiplet centered at T 7.70. A fluorine resonance signal was apparent from a measurement on the HR-60 Spectrometer using a fluorine probe. The above data indicate a fluorohydrin. The n.m.r. Spectrum is shown in Figure 4. 41 00m 000d .HHX CHHUSQOHOSHM mo Esupommm pmumumcH AHIEUV mocmsvmum OONH oowfi Good 000m .m musmflm 00mm _ _ _ 42 .HH% :HHUSSOHOSHM mo Ednummmm .m.2.z mwSHm> I P .w musmflm 43,, E 4\ A h e — _ — 45 B. Rearrangement of Pulegone OxideyII With Boron Fluoride Etherate The method described in (A) was used to effect the re- arrangement of isomer II (m.p.- 550-550). After evaporation of the benzene solvent, the product mixture was distilled through a short—path distillation apparatus (b.p.— 750-780/2 mm). Vapor phase chromatographic analysis of the product showed one predominant peak, accounting for 80% of the re- action products. The colorless liquid was collected by preparative v.p.c. (6 ft., 5% PDEAS, 1580) and exhibited the following infrared spectrum: v§§i4 5580 (w), 5440 (s), 1705 (S), 1715 (shoulder), 1120 cm‘1 (8). See Figure 5. The n.m.r. Spectrum exhibited the following Signals: 15 cps), T 8.56 (doublet; J = 15 cps), T 8.92 (doublet; J T 9.07 (doublet; J = 5 cps), T 8.00 (multiplet), T 6.58 (singlet). Ratio 5:5:5:6.8:1. The n.m.r. spectrum is shown in Figure 6. C. Reaction of Products From I and II With Base (1) Two and one-half gms of the fluorohydrin XI from II was dissolved in a solution of 25 ml of methyl alcohol, 1.0 gm of sodium hydroxide and 5 ml of distilled water. The alcoholic solution was refluxed for 4 hours, cooled to room temperature and neutralized with 5% sulfuric acid. The neutralized reaction products were transferred to a separatory funnel and extracted with ether. The ether extract was washed with distilled water and dried over anhydrous sodium 44 .Hx cfluphcouosHm mo Ezuuommm pmumnmsH .m wudmflm AHIEOV mocmdqmum oom 000d coma oowd coma ooom 00mm 1 4 1 i 1 (.1 7 _ i u r x 1 - --\ _ .4 X \ ,/L\_.r1 45 .HX aflupchHOSHm wo Esuuommm mwsHm> I e .m.2.z .m musmflm A a ._.Lm 46 sulfate. After filtering off the drying agent and removing the solvent, v.p.c. analysis of the viscous residue showed one peak, which had the same retention time as pulegone oxide isomer II. A pure sample of the isomer was collected by preparative v.p.c. (5 ft., 4% QF-1 on Chromosorb-G, 1250) and appeared as a white crystalline solid (m.p.- 550-550). The infrared and t.l.c. analysis was identical to that of authentic oxide isomer II. (2) Two gms of the fluorohydrin XI from oxide isomer I were treated as described in the above procedure. Vapor phase chromatographic analysis exhibited one peak which had an identical retention time with that of oxide isomer I. A pure sample was collected by preparative v.p.c. and formed a crystalline white solid (m.p.- 570-590). The infrared spectrum was identical to that of oxide isomer 1. D. Thermal Decomposition of Fluorohydrins XI and XII To a 25 ml pear-Shaped flask equipped with a reflux con- denser was added 5.0 gms of a 50-50 mixture of the fluoro- hydrins XI and XII. The neat mixture was heated to 1500 and after heating for 20 minutes the solution turned pink and then dark brown, accompanied by vigorous foaming and rapid evolu- tion of a gas with a pungent, irritating odor. The dark brown tarry residue was dissolved in ether and transferred to a separatory funnel. The ether solution was washed with one 50 m1 portion of sodium bicarbonate and subsequently with 47 distilled water until neutral. The ether solution was dried over anhydrous sodium sulfate and evaporated by a rotary evaporator. The dark brown viscous residue was shown to be wv/80% of one major component by v.p.c. (4% QF-1, on Chromo- sorbTG, 5 ft-, 1250). The pure material was collected by preparative v.p-c. and was found to be a white, low-melting, crystalline solid (m.p.- 500-510) . The dark brown residue was chromatographed through a silica gel column with methylene chloride as the eluent. After several fractions were collected, the solvent was evaporated and each fraction containing solid material was analyzed by thinhlayer chromatography. With methylene chloride as the eluting solvent, t.l.c. microslides indicated only one component. The off-white solid material was then sublimed to give a volatile white crystalline solid, V, (m.p.- 500;51o). This compound had the following properties: vCC14 max 1682 (shoulders at 1658 and 1702), 1582, and 1262 cm-1; disemicarbazone, m.p. 2150-217O (recryst. from ethanol-water) (40). The infrared spectrum showed no trace of alcohol, as indicated in Figure 7, and is typical of a seven-membered ring nonenolic 8—diketone (41). The n.m.r. spectrum shown in Figure 8 displays a doub- let at T 8.96 (J = 6.0 cps, 5-H), a singlet at T 8.86 (6-H), and a multiplet centered at T 7.65. Anal. Calcd. for CloHlsog: C, 71.59; H, 9.59 Found: C, 71.15; H, 9.58. ooh .msoaplm.filmcmummLoHoonhnumEHuulm.N.N mo Esuuommm UmumumsH Adlzov mosmskum oom OOfia coma coma GOSH ooom .S musmflm 00mm 48 1 _ _ d 7 . _ 1 1 _ 1 11 111. . 11.1. _ ,. ’. 1 11 1 11. 1 11 1 1 11. 1 N. 1.1 11 11. 11 1 1. .1 1.1? .1 1 .. ‘11... . e . i 1 1.. 11 1 11 1 v 1 11 11 1 1. 1 1 1 1 1. ... . v. 1 .1. m 1.1 . w u 1 1 11 1W 11 1 1 1 1’ ._ ._ 1. . . . 1 1 11 11 1.1 1 1 1 1 . 1 .1 _ .1. 1 1 1 1. 1 w 1.1.1.1111 .11. 1.111 11 1 1 . 1.1 1.1 111111 1... 1.111 1 1 1 _ C 1 .. , 1 . . . 1 . .11 .111 1:111 1 . 1 1 . 1 1 1.11...." 1.1. .. 1.1 1 1 ,1 1 1 .1 1., 1:11.111; 1 1 1 1 1 1 1 .1711 .1. 1... 1.12.121 .1 1 1 . 1 g: . _ 1 1 1 . . r _r.1.. 1 1 1 . 1 I Y H .S t. (1* fl d . .1 ..... 1 A 49 .> .mcoflplm.almsmummcoHowoawcumEHuulm.N.N Mo Eduuommm .m.2.z mmSHm> I e m m n .m musmflm a 1 A _ 50 The compound is designated as 2,2,5—trimethyl-1,5-cyclo- heptanedione, V. The mass spectrum, Figure 9, showed several metastable ions which appear to be characteristic of seven- membered ring ketones (42). Figure 9. Mass Spectrum of 2,2,5-trimethyl-1,5-cyclo— heptanedione (V). Egg % Base Peak m1§_ %4Base Peak 52 24.4 85 24.5 54 58.0 95 51.5 55 19.4 97 50.0 55 24.2 110 50.5 58 72.5 122 27.5 59 100.0 125 19.4 70 18.5 124 97.2 80 24.5 159 18.4 81 25.5 158 24.2 82 51.5 Intensity of P + 1 Peak: (42) Calcd. for C10H1502: 11.14% P Found: 10.70% p E. Attempted Deuterium Exchange of Compound V Several attempts were made to obtain 2,2,5-trimethyl-f 1,5-cycloheptanedione-4,4,7,7—d4 by exchange with deuterihm oxide in the presence of potassium carbonate as a catalyst. The reaction products were found to be predominantly base cleavage products of V. The procedure finally used was a variation of those previously employed (44,45). To 10 ml of 99.9% deuterium oxide (U.S. Nuclear Corp., Burbank, Calif.) was added 0.6 gm of anhydrous potassium carbonate (Reagent grade). Two gms of V was dissolved in 10 ml of cyclohexane (Baker, Reagent grade) and transferred to the D20 solution contained in a 50 ml round-bottom flask 51 equipped with a condenser and drying tube. The heterogeneous mixture was allowed to reflux for three days while being stirred with a magnetic stirring bar. The mixture was then transferred to a separatory funnel where the aqueous layer was removed. After extracting the aqueous layer with two 50 ml portions of ether, the ether extracts were combined with the cyclohexane layer, and dried over anhydrous magnesium sulfate. Thin layer chromatography of the ether extract dis- played two Spots and an infrared spectrum of the crude product showed traces of a carboxylic acid. The ether solution yielded a semi-solid off-white residue (0.55 gm) which was chromatographed through a silica-gel column using methylene chloride as the eluent. Those fractions containing the white solid were combined for analysis. The infrared spectrum showed a doublet at 1585 cm‘l(m) and a strong carbonyl stretch at 1690 cm'l. The n.m.r. spectrum was almost identical to that of V. The mass spectrum of deuterated V showed 1.2% D4, 8.2% D3, 29.7% 92, 58.0% D1 and 22.9% unlabeled V. The aqueous deuterium oxide layer was acidified with dilute hydrochloric acid and extracted with two 50 ml portions of ether. After washing the ether extract with 5% sodium bicarbonate and water, the solution was dried over anhydrous magnesium sulfate. The ether was removed yielding an amber- colored liquid (0.7 gm). The infrared Spectrum of the crude material showed a strong broad —OH band at 5100-5600 cm"1 and 52 1 which are typical of carbonyl stretching at 1700 cm- carboxylic acids. The n.m.r. Spectrum displayed two doublets at T 9.16 and 9.02 assigned to the secondary methyl groups on IXa and IXb indicating a mixture of isomeric keto acids. A broad singlet was observed for the geminal dimethyl grouping where there was deuterium substituted at the isopropyl carbon (T 8.90); with hydrogen at the isopropyl carbon, a doublet was observed at T 8.92 (J = 6.5 cps). The acidic proton ap- peared as a singlet at T 1.56. The suggested structures are shown below: 0 H3 3 (CH3)CH(D)-&-CH2(D)- H-CH2CH2(D)- -0H IXa 8 8““ 8 (CH3)CH(D)- -CH2(D)-CH2- H-CH2(D)- -0H IXb F. Esterification of IXa and IXb with Diazomethane A freshly prepared solution of diazomethane was added dropwise to 0.1 gm of the mixture of keto acids dissolved in 10 ml of ether until the faint yellow color of diazomethane persisted. The excess diazomethane was destroyed by adding 2 drops of glacial acetic acid. The ether solution was washed with 5% sodium bicarbonate and distilled water and subse- quently dried over anhydrous sodium sulfate. The ether was removed by a rotary evaporator and a pale amber viscous liquid remained in the flask. An infrared spectrum of the crude ester 55 801 moswsvmum .musuxHE Uflom oumx mo Eduuommm UmumumsH .oa musmflm >_..._ ._1 - .9“. ~.- _‘. I 54 showed a trace of acid and it was thus chromatographed through a silica-gel column using chloroform as the eluent. The chloroform was evaporated from the fractions collected and an infrared spectrum was taken of each residue. The fractions were combined and subjected to a v.p.c. analysis which showed the presence of two products X (80%) and X' (20%). A sample of X was collected by preparative v.p.c. and an infrared spectrum (Figure 11) indicated a partially deuterated ester. X' was not identified. The n.m.r. Spectrum, Figure 12, ex- hibited doublets at T 9.12, 9.08, and 8.96; these data sug- gest a mixture of keto esters Xa and Xb. The doublets at T 9.12 and 9.08 can be assigned to the secondary methyl group of isomer IXb and IXa, respectively. The doublet at T 8.96 is assigned to the geminal dimethyls being split by hydrogen on the isopropyl carbon. A broad singlet observed at T 8.94 is assigned to the geminal dimethyl grouping with deuterium substituted at the isopropyl carbon. The -OCH3 resonance appears at ”r6.40. 3 (EH3 9 (CH3)2CH(D)- -CH2(D)-CH-CH2—CD2(H)-C—0CH3 Xa 9 9H3 9 (CH3)2CH(D)-C-CH2(D)-CH2—CH—CH2(D)—C-OCH3 Xb G. Preparation of Diazomethane (46) A mixture of 1.25 gms of N,N'-dinitroso-N,N'-dimethyl terephthalamide (Dupont's EXR—101 with inert filler), 57 ml .QXH cam mXH mHSDXaE Hmummloumx mo Esuuummm UmumumcH .aa Guzman AHIEUV mucmsvmum com coca coma coaa coma coma coon comm 55 1 1 1 1 4 1 1 1 56 .Qx paw ax musuxafi “mummlouox mo Eduuommm mzz .ma musmam m05am> I e m m _ fl _ _ _ as: 57 of ether, and 12.5 ml of 40% sodium hydroxide were stirred vigorously at 00-50 for 15 minutes and then allowed to warm to 200 over a 15 minute period. The yellow ether-diazomethane mixture was then distilled (water bath) and used immediately. H. Rearrangement of Pulegone Oxide I with p-Toluenesulfonic Acid To a 125 ml Erlenmeyer flask was added a solution of 2.0 gms of oxide isomer I in 50 m1 of benzene. To this benzene solution was added 0.1 gm of p-toluenesulfonic acid and the solution was swirled to dissolve the acid. The homogeneous acidic mixture was allowed to stand at room temperature for 1 hour with occasional swirling. After being washed with 5% sodium bicarbonate solution and water, the benzene solution was dried over anhydrous magnesium sulfate, filtered, and the solvent was removed at reduced pressure. The residue was analyzed by v.p.c. (4% QF-1 on Chromosorb-G, 5 ft., 1250) and was found to contain about 90% of a new compound XIII and several other products in small amounts. A pure sample of XIII was collected by preparative v.p.c. and the infrared spectrum suggested a terminal methylene (5080, 1820, and 905 cm“1), and a saturated ketone (1710 cm'l), Figure 15. The n.m.r. Spectrum (Figure 14) gave the following sig- nals: T 9.02, two poorly resolved doublets; 8.28, doublet (J = 1 cps); multiplet 7.5-8.4; singlet 6.18; doublet 4.82 (complex multiplet when expanded at 50 cps). 58 .>Hx .mcocmch0aoxoawcumEIm Iahsmmonmomalthxoucthm mo Esnuommm omnmnwca .ma musmam AHIZUV hocmsvmum ooh com coaa coma coma oowa oocm comm 1 1 a 1 a a s m 1 59 .HHHx .mcocmmeOHO>0ahcumfilmIahcmmoumomalmlmxouc>clm mo Esnuommm .m.z.z _ — monam> I m 1 P 1 a: ma.mle m n so.n-e .aa madmam aO.mIe 60 9.01 /H // / H0 \\ éflZZ 4.82 8. 8 XIII I. Rearrangement of Pulegone Oxide II with p-Toluenesulfonic Acid Two gms of isomer II was dissolved in 50 ml of benzene contained in a 125 ml Erlenmeyer flask. To this solution was added 0.1 gm of p-toluenesulfonic acid and the mixture was stirred for 1 hour at room temperature. The reaction mixture was worked up as described above, and the crude product Showed five peaks when analyzed by v.p.c. The major product comprised approximately 60% of the total rearrange- ment products and two of the peaks appeared as shoulders on the main peak. Several attempts were made to collect the major product on a preparative scale using several different columns but none were successful. By coincidence, while using a fluorinated silicone column (4%;QF-1 on Chromosorb- G, 5 ft., 1250) which was used previously to collect XI and XII, pulegone oxide isomer II was found to rearrange to give a Single compound which had a retention time identical to the major product of the acid catalyzed rearrangement of isomer II. Apparently, the fluorohydrins XI and X11 undergo 61 partial decomposition on the column leaving an acid surface on the stationary phase. A sample was collected and analyzed by infrared and n.m.r. The infrared Spectrum showed a terminal methylene (5090, 1645, and 1820 cm‘l), hydroxyl absorption (5470 cm'l), and a saturated carbonyl, (1712 cm-l), Figure 15. The n.m.r. Spectrum (Figure 16) exhibited the following signals: T 8.50 (quartet or two doublets); 7.4-8.8 (multiplet); 8.95 and 8.86 (poorly resolved doublets); 4.80 and 4.90 (broad). (D L0 01 00 CD CD “ 0 N0” \ \‘4.80 and 4.90 8.50 XIV III. Gas Phase Pyrolysis of the Pulegone Oxides A. Procedure To a clean Pyrex combustion tube (volume = 150 ml) was added 0.2 gm of isomer I or II. The combustion tube was degassed, sealed under 0.5 mm pressure of nitrogen and placed in a combustion furnace for the desired length of time at 2000. After the indicated reaction period, the tubes were removed from the furnace and immersed in cold water to allow 62 .HHHX .msocmmeOHoxoamcumeImIawsmmommomaImImxoucchm mo Esuuowmm pmumumsa .ma musmam AHISUV mocmdvmum 00> com coaa coma coma ocna ocom comm 1 1 1 1 1 1 1 1 65 .>HX .maocmX0£OHUxoaxnumEImIamcmmoumomaININxouomclm mo Esuuommm modam> I a m m S m m .m.2.z .ma madmam — 1 1 1 . - 1 — 64 condensation of the sample prior to opening. The contents, a pale-amber liquid, were taken up with 50 ml of pentane and the residue was analyzed by v.p.c. (5% PDEAS, 5 ft, 1280) after evaporating the pentane. The v.p.c. of the pulegone oxide 11 reaction mixture showed five peaks including some unreacted starting material and a material found also in the liquid phase rearrangement. The major peak, which comprised 75% of the product mixture, was collected by preparative v.p.c. A t.l.c. analysis of the semi-solid product Showed two spots which indicated the presence of two compounds with identical v.p.c. retention times. Several v.p.c. columns were used to attempt a separation of the two compounds and the best separation was achieved with 4% QF-1 on Chromosorb—G. The following compounds were isolated from the crude pyrolysis mixture. 0 / 0 2:0 III (10%) IV (12%) (0) $76 The remainder of the products was unreacted pulegone oxide II. Compound C was shown to have the exact retention time as V and was isolated as a white, volatile, crystalline solid (m.p.-500-510). A mixture of V and C showed no melting point depression. The infrared and n.m.r. Spectra of the two com— pounds were identical. The gas phase pyrolysis of pulegone 85 oxide isomer I appeared to give almost exclusively V and un- reacted.starting material. Table 2 shows the results of the gas phase reactions of both I and II. B. The Effect of Oxygen on the Reaction Rate Two-tenths gm of isomer I was placed in a Pyrex com- bustion tube and sealed at 0.5 mm without degassing. The tube was placed in a combustion furnace, heated at 2000 for 4.5 hours and then cooled to condense the vapors. The pale amber liquid was taken up with 25 ml of pentane and the residue after the evaporation of pentane was andlyzed by v.p.c. and only V (25%) and unreacted pulegone oxide isomer I (77%) were indicated. This reaction rate was somewhat slower than the rate of the degassed samples. The results are tabulated in Table 2. C. The Effect of Increased Surface Area on the Reaction Rate One-tenth gm of pulegone oxide isomer II was placed in a Pyrex combustion tube and Pyrex tubing was added to in- crease the surface area by 100%. The reaction tube was de- gassed and sealed under nitrogen at 0.5 mm pressure and subsequently placed in a tube furnace at 2000. The reaction mixture was worked up as described in section B and analyzed by v.p.c. (5% PDEAS, 6 ft., 1250) to give IV + V (70%) and III (24%) and the remainder unreacted starting material. The data are shown in Table 2. The increase in surface area increased the reaction rate considerably. 66 Table 2. Gas Phase Rearrangement of the Pulegone Oxides Condi- Number Isomer Time V(%) III(%) IV(%) II I tions of Runs _II 12 66.2 11.5 12.0 10.5 -- N2 5 11 12 66.1 15.0 11.9 7.0 -- N: 3 II 4 55.2 16.7 5.1 25.0 -- N2 2 11 4 66.0 24.0 10.0 Trace -- N: 2 I 4.5 59.0 Trace -— -— 41 N: 2 I 4.5 25.0 Trace -- -- 77 Air 2 I 4.5 50.0 Trace -- -- 70 N2 2 a . . . . Reactions run With 100% increase in surface area. IV. Liquid Phase Pyrolysis of the Pulegone Oxides A. Procedure The experimental techniques used in these experiments are discussed in Johnson's thesis (16). A small sample (0.2 ‘ ‘ gm) of pulegone oxide was placed in each tube, degassed, and sealed under nitrogen at atmospheric pressure. The samples were heated in an aluminum block furnace at 2000 for the desired length of time. B. Thermal Rearrangement of Pulegpne Oxides Thermal reactions were run in duplicate and analyzed by v.p.c. using 4% QG-1 on Chromosorb-G at 1250. No high 67 boiling products or tars were formed from the neat samples of pulegone oxide. Both I and 11 formed the two isomers of 2,5-dimethyl-2-acetylcyclohexanone III and IV in the ratio of 1:2. The mass Spectrum of III and IV are Shown in Figure 17. Figure 17. Mass spectra of the isomers of 2,5- dimethyl-Z-acetylcyclohexanone. Isomer III Isomer IV gig %yBase Peak gig % Base Peak 55 46.5 56 59.4 56 41.4 70 59.4 67 8.6 84 15.2 69 45.0 85 59.4 70 56.2 99 15.2 82 25.9 112 59.4 84 41.4 127 100.0 85 54.5 128 9.1 97 58.6 168 5.5 98 59.5 111 42.2 125 51.1 Intensity of P + 1 peak: 126 100.0 127 9.5 Calcd. for clonlsogz 11.14% 168 5.4 Found: (III) 10.05% (IV) 11.50% Isomer I appeared to react at a considerably slower rate than isomer II in the formation of III and IV, and there was less epimerization in the reaction of isomer I. Several samples were sealed without degassing to study the effect of air on the rearrangement rates and there was no obvious inhibition by the presence of oxygen. The data for these experiments are tabulated in Tables 5, 4, and 5. 68 Table 5. The Thermal Rearrangement of Pulegone Oxide Isomer I Reaction % Decom- Low Time position I/II % III % IV IV/III Boilers (Hours) (%) 2 5.8 -- 1.5 5.2 2.41 1.0 4 14.8 -- 5.0 5.5 1.75 5.4 6 28.8 29.0 7.4 15.5 1.80 7.2 8 58.4 25.1 10.0 17.7 1.77 8.8 10 55.0 12.6 15.1 24.2 1.88 15.4 Table 4. The Thermal Rearrangement of Pulegone Oxide Isomer II Reaction % Decom- Low Time position I/II % III % IV IV/III Boilers (Hours) (%) 2 67.5 0.95 11.0 22.7 1.76 5.5 4 88.5 2.18 18.5 58.4 2.08 6.7 6 94.5 2.56 25.0 48.0 2.09 10.1 8 94.5 2.21 25.5 50.2 2.16 8.5 10 98.6 5.06 28.0 54.0 1.95 12.7 Table 5. The Thermal Rearrangement of Pulegone Oxide II (In Air) Reaction % Decom- Low Time position I/II % III % IV IV/III Boilers (Hours) (%) 2 57.0 0.47 11.1 20.0 1.80 5.9 4 75.1 0.88 17.5 50.4 1.77 5.6 6 84.5 1.50 20.7 54.8 1.68 9.0 8 92.9 1.85 25.6 45.5 1.69 10.5 10 95.7 2.25 24.7 48.5 1.95 15.0 69 C. Solvent Effects on the Thermal Rearrange— ment of Pulegone Oxide II A 5% solution of pulegone oxide II in various solvents was prepared and 0.5 ml portions of this solution was placed in Pyrex vials. The samples were degassed and sealed under nitrogen at atmospheric pressure. The solvents used in this experiment were cyclohexane, cyclohexene, decalin, p-cymene, cyclohexanone, 1,4-dioxane, and t-butyl alcohol. The purity and preparation of the solvents was described in a previous section of this thesis. The solutions were heated in an aluminum block furnace at 2000 for 8-10 hrs.. The reaction vials were then removed from the furnace and placed in a cold water bath. Vapor phase chromatographic analysis of the solvent reactions showed that the oxygen-containing solvents gave compounds III and IV in addition to low-boiling products and unreacted starting material. The hydrocarbon solvents gave exclusively compound V in addition to some low-boiling products and unreacted starting material. The ldw-boiling products were not identi- fied. Compounds III and IV were identified by comparison of infrared spectra with those of authentic samples. An infra- red comparison with V and a mixed melting point determination proved its identity. No epimerization was detected in these reactions. These data are tabulated in Table 6. 70 Table 6. Solvent Effect on Thermal Rearrangement of Pulegone Oxide Isomer II Reaction Low Solvent Time Boilers % v % III % IV % II (Hours) (%) Cyclohexane 10 55.0 15.4 -- a— 55.6 Cyclohexene 10 15.0 41.1 -- -- 45.9 p-Cymene 8 a 44.4 -- -- 55.6 1,4-Dioxane 8 49.2 -- 9.7 18.5 22.6 Cyclohexa— none 6 a -- 14.4 51.4 54.2 t-Butyl Alcohol 6 a -- 14.5 26.1 59.4 aLow boilers could not be measured due to overlap of solvent peak on the v.p.c. analyses. D. Rearrangement of Pulegone Oxide II in Presence of Base Samples were prepared by placing 0.2 gm of 10% of the designated base and 90% of isomer II in a Pyrex tube. The samples were degassed by the usual procedure and sealed under nitrogen at atmOSpheric pressure. After heating in the alumi— num block furnace for 6 hours the samples were analyzed by v.p.c. and revealed the same products which were formed in the neat liquid phase reaction. A considerable amount of epimerization was observed in the course of these reactions. The results are summarized in Table 7. 71 Table 7. The Effect of Base on the Thermal Rearrangement of Pulegone Oxide Isomer II Reaction % Low Base Time Reaction Boilers (Hours) (fo) 2’5 III % 1v % II j I 2,6-Lutidine 6 75.8 2.8 14.6 55.6 26.2 20.8 Triethylamine 6 87.7 4.8 16.7 41.0 12.5 25.2 N,N—Dimethyl 6 68.8 7.5 15.5 54.4 51.2 11.6 Aniline E. The Effect of Free Radical Initiators on the Thermal Rearrangement of the Pulegone Oxides I and II The samples were prepared by adding 0.2 gm of 90% pulegone oxide and 10% of the radical initiator to a Pyrex reaction tube. The mixture was degassed and the reaction vials were sealed at atmospheric pressure under a nitrogen atmosphere. After heat- ing at 2100 from 5-6 hrs., the tubes were removed from the furnace, immersed in cold water, and opened for a v.p.c. analy- sis. The analysis Showed the presence of the products found in the neat liquid phase rearrangement. The reaction data are shown in Table 8. 72 Table 8. The Effect of Radical Initiators on the Thermal Rearrangement of Pulegone Oxides I and II Isomer and Reaction % Decom- Low Initiator Time position % III % IV % II % I Boilers (Hours) (%1 t-Butyl Hydro- peroxide (II) 5 51.0 9.1 20.4 49.0 11.5 10.0 2,2'-Azoiso— propane (I) 6 46.6 7.4 16.0 5.2 55.4 20.0 2,4,4-Tri- methyl-2- formylcyclo- pentanone (II) 5 57.8 1.7 15.9 42.2 -- 42.2 V. Miscellaneous Experiments A. Preparation of 2,245,5-Tetramethyl- 1y5-cyclohexanedione (VII) (47) To a 500 ml three-neck round—bottom flask equipped with a mechanical stirrer, condenser, thermometer, and dropping funnel was added 20 gms (0.12 mole) of dimedone and 60 ml of dry dimethyl sulfoxide and the mixture stirred until all the dimedone had dissolved. To this solution was added a solu- tion of 5.5 gms of sodium metal in 60 ml of absolute ethanol. The resulting solution was heated to 700 with vigorous stirring for a period of one hour, after which the solution was cooled to 50. To this cooled solution 12 ml of methyl iodide was added dropwise with vigorous stirring over a fifteen minute period. After warming to room temperature, the reaction mix- ture was heated to 750 and stirred for two hours. 75 After cooling the reaction mixture to 250, 5.0 gms of sodium metal.in 50 ml of absolute ethanol was added and the DMSO solution was heated to 750 with continued vigorous agi— tation for 1 hour and then cooled again to 50. Methyl iodide (10 ml) was added dropwise and the mixture was then heated to 750.and maintained at that temperature for two hours with continuous stirring. The mixture was again cooled to room temperature, poured over ice in a 600 ml beaker and extracted three times with 50 ml portions of ether. The ether extract was washed with five 25 ml portions of a 10% solution of sodium carbonate, one 50 ml portion of water and dried over anhydrous magnesium sulfate. The ether solvent was evaporated by using a rotary evaporator and a white crystalline residue formed (10.50 gms) which was identified by infrared, n.m.r., and mass spectral determinations as 2,2,5,5-tetramethyl-1,5- cyclohexanedione (VII). The white solid was recrystallized from petroleum ether (60-90), m.p.— 960-980. The infrared Spectrum of VII which is shown in Figure 18 appears to substantiate the assigned structure. The n.m.r. spectrum shown in Figure 19 exhibited three singlets at T 9.02, 8.85, and 7.48 in the ratio of 5:5:2. 8.85 /0 VII 74 .HH> .msoaclm.almcmch0ao>0a>£umamuumulm.m.N.N mo Esuuommm Umumuwca AHIZOS hosmsgmnm OOH 00m ooaa coma Coma 005a OOOM .oa one... comm 1 a 1 1 1 1 1 1 1m. 1 .1. 1 1 v 1 1 1 1 1 1.1 1 . 1... . _ 111 1-1 1 1 1.11 1 1 11 1 1 1 1 _ o _ .1. _ ‘ .1 .1 .. 1. , 1 1 1 a. . . 1 a . . H m . . .. .. . . .. . n u _ 1 n a .w . n .m \. 1 .. w . 1 . . x 1 . . . a .. . 1 . m1 ., ... n. — .1 1 1 “U... w I .. .. . . .1 my C — n .. 1. n .1. 1 .Il _ . . 1 . r 1 1 .. .1 1 1 1 I. w \ x v y a K / x V . ,f I 1 .\ 1 .mGOHUIm.HimcmmeOHUhuawzumfimuumulm.m.N.m mo Ezuuommm .m.2.z .ma musmflm mwsam> I 9 0a m m 1 1 r 75 _ a —‘ —I 1x; 76 B. 2,2,5,5—Tetramethyl-1,3-cyclohexanedione Bisethylenedithioketal VIII (48) A solution of 2,2,5,5-tetramethyl—l,3-cyclohexanedione (0.10 gm) in ethane dithiol (1.0 ml) was treated with freshly distilled borontrifluoride etherate (0.8 ml). The reaction mixture deposited a white solid after standing for 10 minutes and was then worked up by addition of 20 ml of 50:50 methanol: water mixture followed by ether extraction. The ether extract was washed with several portions of water and subsequently dried over anhydrous magnesium sulfate. The ether was evapor- ated and the residual solid was recrystallized from acetone (yield-0-13 gm, 65%), m.p.-155-156O. The infrared Spectrum showed the absence of carbonyl absorption; the n.m.r. spectrum (Figure 20) exhibited four sharp singlets at T 8.87, 8.53, 7.85, and 6.83 respectively (area ratios- 3:3:2z4). 6.83 Anal. Calcd. for C14H24S4: C, 52.46; H, 7.53 Found: C, 52.38; H C. Raney Nickel Reduction of MIII A 0.5 gm sample of VIII in 75 ml of freshly purified dimethylformamide was placed in a 250 ml pressure bottle. To this solution was added 6.0 gm of Raney nickel W42 (49), and the resulting mixture was shaken for 20 hrs at 850 under 77 .Hmemoneeeememflmnemmem mcoflplm.HumcmxonoHUwoamnumEmnumulm.m.m.m mo Eduuommm .m.z.z .ON musmfim mmSHm> I m m .1. 1 1 1 E 78 50 psi of hydrogen. After cooling, the mixture was filtered and the filtrate was diluted with 200 ml of water and then extracted with three 100 ml portions of pentane. The com- bined pentane extracts were washed and dried, and after care- fully removing1most of the solvent by distillation, the residue was examined by v.p.c. analysis (6 ft., 20% Apiezon "L" on Chromosorb-W, 1080). Three volatile hydrocarbons were isolated by preparative v.p.c. and the major product, 1,1,4,4- tetramethylcyclohexane (XV), was present in 70% yield. The n.m-r. spectrum of XV (Figure 21) showed two sharp singlets at T 9-14 and 8.76 having an area ratio of 3:2. The infrared spectrum indicated the presence of only C and H stretching and bending frequencies. D. Attempted Preparations of 1,1,4,4-Tetra- methylcyclohexane (1) Lithium Aluminum Hydride Reduction of 212,515-Tetramethyl-i,3-cyclohexane- ditosylate A solution of 0.8 gm of 2,2,5,5-tetramethyl-l,3—cyclo- hexaneditosylate (prepared by reaction of 2,2,5,5-tetramethyl- cyclohexane—1,3-diol with p-toluenesulfonyl chloride in pyridine) in 20 ml of dry tetrahydrofuran (distilled from calcium hydride) was added dropwise to a solution of 50 ml of dry tetrahydrofuran and 0.72 gm of lithium aluminum hydride. The reaction mixture was stirred by a magnetic stirrer for 4 hours and subsequently decomposed with a saturated solution of sodium sulfate, filtered and dried over anhydrous magnesium 79 .>x .mcmxmnoHumoHMSumEmnumule.w.H.a mo Esuuummm mmsHm> I e .m.2.z .em meeeee m 1 TI .4 80 sulfate. The solvent was distilled through a short Vigreux column. The residue was analyzed by v.p.c. and no hydro- carbon.was detected. Infrared analysis showed the presence of an alcohol which indicated reduction to the diol. (2) WOlff—Kishner Reduction of VII (50) A mixture of 10 gms of VII, 16 gms of potassium hydrox— ide, 60 ml of triethyleneglycol, and 12 ml of 85% hydrazine hydrate was refluxed for one hour following which the water was removed kura Dean-Stark takeoff. The reflux temperature rose to 1150 and refluxing continued for 5 hours. The reaction mixture and the aqueous distillate were combined and extracted with ether.I The neutral organic fraction was distilled over sodium and the distillate analyzed by vapor phase chroma- tography. There were five components in the product mixture, which was not analyzed further. E. Attempted Preparations of 1,1,4-Tri- methylcycloheptane (1) Synthesis of Tetrahydroeucarvone A solution of 10.0 gms of eucarvone (prepared by the method of Corey and Burke, 51) in 75 ml of absolute ethanol was placed in a pressure bottle and 6.0 gms of Raney nickel, W-2, was added. The mixture was then shaken at room temper— ature under 30 lbs. of hydrogen pressure; the reaction was complete within 30 minutes as indicated by no further uptake of hydrogen. The Raney nickel was removed by filtration and the filtrate distilled through a short-path distillation 81 column to give 7.0 gms of tetrahydroeucarvone (b.p.- 550/10 mm). Vapor phase chromatographic analysis showed only a single peak and the infrared spectrum showed no unsaturation or hydroxyl.absorption. (2) Tetrahydroeucarvone Ethylenedithioketal To a solution of 1.0 ml of tetrahydroeucarvone in 1.0 ml of ethanedithiol was added 0.8 ml of borontrifluoride etherate. The mixture was allowed to stand for 1 hour. The reaction mixture was diluted with 10 ml of water and 5.0 ml of methanol and extracted with ether. The ether layer was dried over anhydrous.sodium sulfate and the.ether evaporated using an aspirator. The oily residue was allowed to stand for one week but no crystals were deposited. v.p.c. analysis of the residue showed five peaks (20% SE-30, 1500, 5 ft). The infrared spectrum showed no carbonyl absorption. (3) Reduction of Tetrahydroeucarvone Ethylenedithioketal The oily material isolated from the above experiment was dissolved in 50 ml of dimethylformamide and transferred to a pressure bottle. To the resulting solution was added, 6.0 gms of Raney nickel (W-2) and the mixture was shaken for 8 hours under 50 psi of hydrogen at 700. After cooling to room temperature the deactivated Raney nickel was removed by filtration and the dimethylformamide solution was diluted with 50 ml of water and extracted with pentane. The pentane extract was washed with water, dried over anhydrous magnesium sulfate and finally distilled through a 24-inch Vigreux column. 82 v.p.c. analysis of the residue showed four components, the major one being 50% of total. An n.m.r. spectrum of the major component showed only methyl and methylene protons. (4) 2,2,5-Trimethylcycloheptane-i,3-dione Bisethylenedithioketal (Synthesis and Raney Nickel Reduction) To a solution of 0.2 gm of 2,2,5-trimethylcycloheptane- 1,3-dione, (V), and 0.2 ml of ethanedithiol was added 0.1 ml of borontrifluoride etherate and uniformly mixed. The mixture was allowed to stand for 1 hour at room temperature and worked up as previously described in section (2) above. The residue was a viScous oil whose infrared spectrum showed the complete absence of a carbonyl absorption. The isolated oil was dissolved in 50 ml of absolute ethanol and reduced over Raney Nickel as described in section (3). After completing the work-up, the residue was an oil which was shown to have at least five components by v.p.c. analysis. None of the v.p.c. peaks corresponded to the products described in section (3). The products were not further identified. (5) Preparation of the Tosylhydrazone of Tetrahydroeucarvone (52) A methanol solution of tosyl hydrazine was prepared by dissolving 1.23 gms of tosylhydrazine in 2.0 ml of boiling dry methanol. The resulting solution was cooled to 500 and added at once to 1.0 ml of tetrahydroeucarvone. The reaction mixture was allowed to stand at room temperature for 30 minutes and was diluted with water until the solution became cloudy. 85 The cloudy suspension was cooled in an ice-water bath for 2 hours and deposited a viscous oil which crystallized after vigorous stirring for several minutes. The material was re- crystallized from a methanol—water mixture (Yield- 50%; m.p.— 1080-1100), and dried in a vacuum desiccator for 24 hours. (6) Reduction of Tetrahydroeucarvone Tosylr hydrazone with Sodium Borohydride (55) To a solution of 1.0 gm of tosylhydrazone in 50 ml of dry methanol (distilled.from magnesium turnings) was added 6.0 gms of sodium borohydride, and the reaction mixture was refluxed for 8.hours. After refluxing, the mixture was diluted with 100 ml of water and extracted with two 50 ml portions of pentane. The pentane extracts were washed with two 50 ml portions of water, 5% sodium bicarbonate solution, and final- ly with water again. After drying the extracts, the solvent was distilled through a vigreux volumn and the residue was analyZed by v.p.c., which indicated the presence of three components(20% Apiezon "L", 1500, 5 ft.). The retention times did not agree with any products formed in section (4) or (6). The investigation of these products was discontinued. (7) Semicarbazone of Tetrahydroeucarvone A solution of 1.0 ml of tetrahydroeucarvone in 10 ml of ethanol was prepared and water was added dropwise until the solution became faintly turbid; the turbidity was then removed with a few drops of ethanol. To this reaction mix- ture was added 1.0 gm of semicarbazide hydrochloride and 2.5 gms of sodium acetate ~5H20. The reaction flask was placed 84 in a beaker of boiling water and permitted to slowly cool to room temperature. Upon cooling, the semicarbazone precipi- tated and after further cooling in an ice-water bath was filtered. The solid semicarbazone (Yield-1.2 gms) was re- crystallized from ethanol-water (m.pa-1BOO-1620). (8) Reduction of the Semicarbazone of Tetrahydroeucarvone (54) A mixture of 0.5 gm of tetrahydroeucarvone semicarba- zone, O-5 gm potassium hydroxide and 3.0 ml of diethylene glycol was heated cautiously and then distilled until no further.material distillate was obtained. The distillate was taken up with 5-0 ml of pentane and the residue extracted with 5.0 ml of pentane. The pentane fractions were combined and washed with 2.0 ml of 5% hydrochloric acid and 2.0 ml of water before being dried over anhydrous sodium sulfate. The pentane was distilled until most of the pentane was removed and the oily residue was shown to contain two prominent products and several minor components (20% Apiezon "L", 5 ft, 1200). The product mixture was not investigated further. F. Effect of Lead Tetraacetate of Diols on v, III, and IV (15) (1) Lithium Aluminum Hydride Reduction of a Mixture of III and IV A solution of 1.46 gms of a 1:2 mixture of III and IV in 5.0 ml of ether was added to a suspension of 0.70 gm of lithium aluminum hydride in 20 ml of ether. The mixture was stirred while refluxing for 4 hours; the complex was decom- posed with moist ether, water, and finally with concentrated 85 sodium hydroxide solution. The ether layer was separated and the aqueous layer extracted with ether. The ether extracts were combined, washed with water and finally dried over anhydrous sodium sulfate. The ether was evaporated and the.residue (1.16 gms) was a pale amber gel which formed a white crystalline material on standing overnight. The iso- mers were not separated. An infrared analysis showed the presence of a strong —OH band and the absence of any carbonyl absorption- ((2) Lithium Aluminum Hydride Reduction of 2,2,5- Trimethylcycloheptane-i,S-dione (V) A.solution of 0.96 gm of V in 5.0 ml of anhydrous ether was added to a solution of 0.45 gm of lithium aluminum hydride in 20 ml of ether. The mixture was stirred and heated under a gentle reflux for four hours. The complex was decomposed with moist ether, water, and finally with concentrated sodium hydroxide solution and the work-up was completed as described in section (1). After evaporating the ether, the pale-amber viscous residue (0.85 gm) was allowed to stand overnight but did not solidfy. Infrared analysis showed a strong -0H band and the absence of any carbonyl absorption. (5) Reaction of the Diols from III and IV with Lead Tetraacetate With stirring, a chloroform solution of 5.0 gms of freshly prepared lead tetraacetate was added dropwise to a solution of 1.16 gms of the diol mixture in 55 ml of chloro- form. The mixture was allowed to stand overnight. The pre- cipitate of lead diacetate was separated and washed with 86 chloroform. The filtrate was treated with sodium carbonate solution, washed with water and dried over anhydrous sodium sulfate. An infrared analysis showed a strong -OH band and a small trace of carbonyl which is probably due to the presence of some acetates. The n.m.r. spectrum exhibited only hydroxyl, methylene, and methyl proton absorptions. There was no trace of an aldehydic proton. (4) Reaction of the Diol From V with Lead.Tetraacetate The diol of V (0.85 gm) was treated as described in section (5) above. After work-up, the infrared spectrum showed a strong -0H band and only a trace of carbonyl absorp— tion. The n.m.r. spectrum again indicated complete absence of any aldehydic protons. 87 O3 .H>X .mcommasmomfllawnumelw mo Esuuommm .m.2.z .Nm musmflm mmsHm> P w n \ iii SUMMARY (1) Both pulegone oxide isomers I and II rearranged to the diastereoisomeric 2,5-dimethyl-2-acetylcyclohexanones III and IV in the neat liquid phase thermal rearrangement (2000); isomerization and fragmentation was also observed. Regardless of the starting isomer, it was shown that a constant ratio of III and IV was obtained. The identity of the rearrangement products was confirmed by comparison of infrared spectra and v.p.c. retention times with authentic samples. A radical cleavage recombination mechanism was proposed. (2) The vapor phase rearrangement (2000) of both I and II gave the ring-expanded B-diketone, 2,2,5-trimethylcyclo- heptane-1,5-dione (V), a white crystalline solid, m.p.-500- 510, along with minor amounts of III and IV. The reaction was found to be strongly accelerated by an increase in surface area. Compound V was identified by infrared and n.m.r. spectra and confirmed by an unequivocal synthesis. A mechan— ism for the rearrangement has been proposed as proceeding through an ionic intermediate with initial cleavage of the Ce-oxygen bond and a subsequent 1,2-acyl migration. (3) When the thermal reactions were investigated with 5% solutions of II in various solvents, significant results were obtained. In oxygen-containing solvents, only III and IV were isolated as the major rearrangement products. 88 89 The mechanism is again suggested as a radical cleavage re— combination type in which a diradical intermediate is stabilized by some type of complex formation between the electron-deficient radical and the non-bonded p-electrons of the oxygen in the solvent; this stabilization tends to favor homolytic cleavage at the C4-oxygen bond. In hydrocarbon solvents, only (V) was isolated from the reaction mixture; no trace of III or IV was observed. An ionic mechanism can be rationalized by forming a.C8-carbonium ion followed by a 1,2—acyl migration since there appears to be no solvent participation in the transition state; the hydrocarbons are incapable of stabilizing a diradical intermediate. (4) The acid-catalyzed rearrangements of I and II at room temperature proceeded by a heterolytic cleavage of the Cg-oxygen bond to form a tertiary carbonium ion at C8. The products from the Lewis acid catalyzed rearrangement (boron- trifluoride etherate) were the fluorohydrins XI and XII formed stereospecifically from I and II, respectively. The protonic acid, p-toluenesulfonic acid, effected the stereospecific formation of the diastereoisomeric unsaturated hydroxy ketones XIII and XIV. 1. 2. 3. 9. 10. 11. 12. 13. 14. 15. 16. 17. LI TERATURE CI TED M. Tiffeneau and J. Lévy, Comp. Rend., 1 2, 391 (1926). M. Tiffeneau and J. Lévy, Bull. Soc. Chim. (Fr.), 42, 1806 (1931). S. Winstein and R. B. Henderson, "Heterocyclic Compounds," John Wiley and Sons, New York, N. Y., 1950, Vol. 1, p. 1. R. Parker and N. Isaacs, Chem. Rev., 59, 737 (1959). . H. E. Zimmerman, B. R. Cowley, C. Y. Tseng, and J. W. Wilson, J. Am. Chem. Soc., 86, 947 (1964). W. Reusch and C. K. Johnson, and B. Dominy, J. Am. Chem. Soc. .§§, 3894 (1963). W. Reusch and C. K. Johnson, J. Org. Chem., 28, 2557 (1963). H. 0. House and R. L. Wasson, J. Am. Chem. Soc., 18, 4394 (1956). H. 0. House, J. Am. Chem. Soc., 16, 1235 (1954). H. 0. House and G. D. Ryerson, J. Am. Chem. Soc., §§J 981 (1961). H. 0. House and R. L. Wasson, J. Am. Chem. Soc., 19, 1488 (1957). M. N. Prilejaeff, Bull. Soc. Chim., (iv), 41, 687 (1927). W. Treibs, Ber., 66, 1492 (1933). G. V. Pigulevsky and I. K. Mironova, Zhur. Obs. Khim., EZJ 1101 (1957). S. A. Kozhin, S. I. Yakimovich, and G. V. Pigulevsky, Zhur. Obs. Khim., 32, 2368 (1962). C. K. Johnson, Doctoral Thesis, Michigan State University, 1963. C. Djerassi, "Optical Rotatory DiSpersion," McGraw Hill, Inc., New York, N. Y., 1960. 90 91 18. C. Djerassi, w. Klyne, T. Norrin, G. Ohloff, and E. Klein, Tetrahedron, 21, 163 (1965) 19. G. W. K. Cavill, Aust. J. Chem., 858 (1963). 20. G. W. K. Cavill, Personal Communication. 21. W. H. Reusch and P. Mattison, Unpublished Results. 22. C. Djerassi, J. Osiecki, and E. J. Eisenbaum, J. Am. Chem. Soc., §§, 4433 (1961). 23. S. A. Kozhin, S. I. Yakimovich, and G. V. Pigulevsky, Zhur. Obs.thim , ég, 3392 (1962). 24. I. Maclean and R. P. A. Sneeden, Tetrahedron, 21, 31 (1965). 25. B. Eistert, et al., Ann., 665, 55 (1963). 26. L. M. Jackman, "Nuclear Magnetic Resonance Spectroscopy," Pergamon Press, New York, N. Y., 1962. 27. N. S. Bhacca and D. H. Williams, "Applications of Nuclear Magnetic Resonance Spectroscopy in Organic Chemistry," Holden-Day, Inc., San Francisco, 1964, p. 16. 28. K. Nakanishi, "Infrared Absorption Spectroscopy," Holden- Day, Inc., San Francisco, 1962. 29. H. E. Zimmerman and D. I, Schuster, J. Am. Chem. Soc., §$. 1485 (1962). 30. H. E. Zimmerman and A. Zweig, J. Am. Chem. Soc., 83, 1196 (1961). 31. H. E. Zimmerman, Abstracts of the 17th National Organic Chemistry Symposium, June, 1961, Bloomington, Indiana, p. 31. 32. M. S. Kharasch, A. Fono, and W. Nudenberg, J. Org, Chem., .;§. 113 (1951). 33. C. Walling and M. F. Mayahi, J. Am. Chem. Soc., 81, 1485 (1959). 34. J. M. Tedder, Quart. Rev., 14, 336 (1960). 35. C. S. Rondestvedt, Jr., J. Am. Chem. Soc., 84, 3307, 3319 (1962). 36. G. A. Russell, J. Am. Chem. Soc., 19, 2977 (1957). 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 92 C. Walling and A. Padwa, J. Am. Chem. Soc., _8, 1593 (1965). C. Walling and P. Wagner, J. Am. Chem. Soc., 88, 2333 (1963). A. I. Vogel, "A Textbook of Practical Organic Chemistry," 3rd ed., Longmans, Green and Co., Toronto, Canada, 1956. R. L. Shriner, R. C. Fuson, and D. Y. Curtin, "The Systematic Identification of Organic Compounds," John Wiley and Sons, Inc., New York, N. Y., 4th ed., 1958, p. 218. L. J. Bellamy, "The Infrared Spectra of Complex Molecules," John Wiley and Sons, Inc., New York, N. Y., 2nd ed., 1958, p. 147. K. Bieman, "Mass Spectrometry," McGraw-Hill, New York, No Yo: 1962: pp. 155-156. R. M. Silverstein and G. C. Bassler, "Spectrometric Identi— fication of Organic Compounds," John Wiley and Sons, Inc., New York, N. Y., 1963. R. Srinivasan and S. E. Cremer, J. Am. Chem. Soc., 81, 1648 (1965). A. C. Cope and D. M. Gale, J. Am. Chem. Soc., 88, 3570 (1963). E. Baumgarten, R. Levine, and C. R. Hauser, J. Am. Chem. Soc., fig. 862 (1944). R. D. Desai, J. Chem. Soc., 1079 (1932). S. M. Kupchan, S. McLean, G. W. A. Milne, and P. Slade, J. Org. Chem., 21. 147 (1962). R. Mozingo, "Organic Synthesis," Coll. Vol. 3, John Wiley and Sons, Inc., New York, N. Y., 1955, p. 181. Huang—Minlon, J, Am. Chem. Soc., 88, 2487 (1946). E. J. Corey and H. J. Burke, J. Am. Chem. Soc., 18, 177 (1956). G. L. Closs and R. A. Moss, J. Am. Chem. Soc., _8, 4050 (1964). 53. 54. 55. 56. 93 L. Caglioti and P. Grasselli, 153 (1964). K. B. Wiberg, B. R. Lowry, and T. H. Colby, Soc., 83, 4005 (1961). Chem. and Ind., Jan. 25, J. Am. Chem. H. Kristinsson and G. W. Griffin, J. Am. Chem. Soc., 378 (1966). R. A. Nelson and R. S. Jessup, Standards, 48, 206 (1952). J. Research Nat'l Bur. fifl: HICHIGQN srnr: UNIV. LIBRARIES \lV(IWIWill)lllWl‘)NIWW|W||W||H|||NH| 31293001578628