r i IHII‘HH ' THE SYNTHEStS A245 {El-{02‘3“ a {NDUCED— EiEARRANGEMEN? Q? 8; 8 f 6, 6m TET'Ré'kikfiETf-fiiai! éeCY'CLQE'EEXAQ {ME- 108 655 THS Thesis €03! {fin Doqsm. {If M. 5. MICELGAH STATE UNIYEEEETY Bonaid W Frey 3.966 THESIS LIBRARY Michigan State University w .L w nun ABSTRACT THE SYNTHESIS AND PHOTO-INDUCED REARRANGEMENT OF 3,5,6,6-TETRAMETHYL-1,4-CYCLOHEXADIENE by Donald W. Frey Investigations in the chemistry of norbornadiene and 1,5—cyclooctadiene have revealed the formation of products resulting from interaction of the non-conjugated double bonds. The chemistry of the simplest cyclic non-conjugated diene with interacting double bonds, 1,4-cyclohexadiene, is compli- cated by double bond isomerization and dehydrogenation to benzene. The substituted analog, 5,5,6,6-tetramethyl-1,4- cyclohexadiene(I) has been prepared to avoid these difficulties. Dimedone was alkylated with methyl iodide and the substi- tuted diketone was subsequently reduced, using lithium aluminum hydride, to the corresponding diol. Treatment of the diol with p—toluenesulfonyl chloride in pyridine gave a ditosylate which, in the presence of potassium tfbutoxide in dimethyl sulfoxide, yielded I. Irradiation of I in the vapor phase led to the formation of a new compound, 4,4,6,6-tetramethylbicyclo[5.1.0]hex-2-ene (VII). The structure has been assigned on the basis of Spectral observations. Oxidative degradation of VII, Donald W. Frey followed by esterification, has given an ester with the expected Spectral properties. When subjected to hydrogenation VII yielded a saturated hydrocarbon with the appropriate spectral characteristics for 2,2,6,6-tetramethylbicyclo[5.1.0]- hexane. IHES THE SYNTHESIS AND PHOTO-INDUCED REARRANGEMENT OF 3,3,6,6-TETRAMETHYL-l,4-CYCLOHEXADIENE BY Donald W. Frey A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of MASTER OF SCIENCE Department of Chemistry 1966 IHESIS To My Wife, Susan ii ACKNOWLEDGMENT Guidance and encouragement from Dr. Reusch are gratefully acknowledged. iii TABLE OF CONTENTS Page INTRODUCTION AND HISTORICAL . . . . . . . . . . . . . 1 EXPERIMENTAL. . . . . . . . . . . . . . . . . . . . . 4 I. General Procedures and Apparatus. . . . . . . 4 II. Synthesis of 3,5,6,6-Tetramethyl-1,4-cyclo- hexadiene . . . . . . . . . . . . . . . . . . 4 A. 2, 2, 5, S-Tetramethyl-1, 3-cyclohexanedione (III)................... 4 B. 2, 2, 5, 5- -Tetramethyl- -1, 5- -cyclohexanediol (Iv) 6 C. 2, 2, 5, 5- -Tetramethylcyclohexane 1, 3- di— tosylate (V) . . . . . . . . . . . . . . 7 D. 3, 3, 6, 6- -Tetramethyl-1, 4- -cyclohexadiene (I) 7 III. Vapor Phase Photolysis of 5,3,6,6-Tetramethyl- 1,4-cyclohexadiene. . . . . . . . . . . . . . 8 IV. Determination of Photoproduct Structure . . . 10 A. Ozonolysis of Photoproduct . . . . . . . 10 B. Diazomethylation of Ozonolysis Product . . 11 C. Hydrogenation of Photoproduct. . . . . . . 11 RESULTS AND DISCUSSION. . . . . . . . . . . . . . . . 13 LITERATURE CITED. . . . . . . . . . . . . . . . . . . 55 iv was LIST OF TABLES TABLE Page 1. Major Peaks in the Mass Spectrum of VII. . . . . 20 2. Major Peaks in the Mass Spectrum of IX . . . . . 27 5. Major Peaks in the Mass Spectrum of X. . . . . . 29 FIGURE 1. 2. 5. 4. LIST OF FIGURES The infrared spectrum of I. . . . The n.m.r. spectrum of I. . . . . . The infrared spectrum of VII. . . . The n.m.r. spectrum of VII (a) neat benzene . . . . . . . . . . . . . . The infrared spectrum of IX . . . . The n.m.r. spectrum of IX . . . . . The infrared spectrum of X. . . . . The n.m.r. spectrum of X. . . . . . vi Page 17 18 21 23 26 28 50 51 yfiesu INTRODUCTION AND HI STORICAL Extensive investigations of norbornadiene, a compound possessing non-conjugated but interacting double bonds, have disclosed many unusual reactions (1). Although the ultra- violet Spectrum of the Simplest hydrocarbon with this struc— tural feature, 1,4-cyclohexadiene, is also indicative of v-electron system interaction (2,3), this compound‘s chemistry is complicated by double bond isomerization (4) and dehydro- genation to benzene (5). In order to avoid these difficul- ties, the substituted analog, 5,3,6,6-tetramethyl-1,4—cyclo- hexadiene (I), was prepared by Reusch (6) and by other workers (7) through pyrolysis of 2,2,5,S-tetramethylcyclohexane 1,5-diacetate. Unfortunately this synthesis is complicated by concurrent pyrolytic decomposition of I to pfxylene and ethane. In an effort to find an improved route to I, we decided to employ a relatively new technique of elimination, treatment of a tosylate with potassium t-butoxide in dimethyl sulfoxide (8). Although the ultraviolet spectrum of I, showing only intense end absorption (6), is indicative of relatively little interaction of the double bonds, and attempts to achieve homoconjugate addition to I have proved unsuccessful (7), the possibility of photochemical isomerization or rearrange- ment has not been investigated. In particular, reports of IHESI photochemical transannular cycloaddition of the olefinic functions in norbornadiene (9) and 1,5-cyclooctadiene (10) (Equations A and B) stimulated us to study similar reactions with I. Using Srinivasan's technique for the mercury sensitized vapor phase photolysis of 1,5-cyclooctadiene, we anticipated _1:W (A) O 3. ( hv 4!!!!) (B) I the possible formation of 5,5,6,6-tetramethyltricyclo- All...» (C) I I IN) a \ + Caz-He (D) / .... I hv ; + HC=CH (E) \\ I _ hv E (F) I __hv 3 (G) I hv (H) Chart 1. (2.2.0.02'5)hexane (C); however, this ring system would be highly strained and the photochemical reaction might well take another course. Some other possibilities for photorearrangement are given in Chart 1. The products in (D) are those observed ‘in the thermal decomposition of I, and those in (E) would result from a retro-Diels—Alder reaction. Reaction (F) parallels the norbornadiene transformation to quadricyclene, and similar products were obtained by Stechl (11) in the photodimerization of 1,3,S-trimethylcyclopropene. The products in (G) and (H) are possible if alkyl migration occurs. Similar compounds were reported by Crowley (12) and Meinwald (13) from the photolysis of allgocimine and d-phellandrene respectively, although the unsaturated ring system in (H) is still unknown. Product identification using infrared and ultraviolet absorption spectra and nuclear magnetic resonance spectra was not expected to pose a serious problem. EXPERIMENTAL I. General Procedures and Apparatus Infrared spectra were obtained with a Perkin-Elmer 327B Grating Infrared Spectrophotometer. Sodium chloride cells or discs were used for all determinations. Ultraviolet spectra were measured using 1 cm. silica cells and a Beckman DB Spectrophotometer. Nuclear magnetic resonance Spectra were determined on a Varian A-60 high resolution Spectrometer, using tetramethyl- silane as a standard. Vapor phase chromatography, analytical and preparative, employed an Aerograph A-90-P Gas Chromatograph. Melting points were determined on a Kofler hot stage microscope and are uncorrected. Samples were irradiated using a Srinivasan-Griffin Photo- chemical Reactor. This consists of a circular bank of Sixteen mercury resonance lamps in a reflective well. The apparatus was air cooled by a fan; the temperature in the chamber dur- ing irradiation was 40—450 C. II. Synthesis of 5,5L6,6—Tetramethyl— 114—Cyclohexadiene A. 2,2,S,S-Tetramethyl—l,5-cyclohexanedione (III) Dimethyl sulfoxide was distilled from calcium hydride at ca. 15 mm. into a 500 ml. three-neck flask. After 100 ml. had 4 been collected, 20 g. of dimedone (II) was added and the flask was fitted with a mechanical stirrer, an addition funnel and a condenser with a drying tube. A solution of 5.5 g. of sodium metal in 60 ml. of absolute ethanol was added drOpwise, with stirring, to the solution of the diketone. When addition was complete, the mixture was heated to 700 (a thermometer was suSpended in the condenser) and stirred at that temperature for one hour. ’After cooling to 50, 18 ml. of methyl iodide was added dropwise and the mixture was then allowed to warm to room temperature, followed by stirring for 2 hours at 750. The reaction mixture was cooled to room temperature; a solution of 5.0 g. of sodium metal in 55 ml. of absolute ethanol was added, and the resulting mixture was again heated to 750 with stirring for one hour. After cooling to 50, 20 ml. of methyl iodide was added dropwise and the mixture was heated to 750 with stirring for two hours. After being chilled in an ice bath, the reaction mixture was poured into ice and water. About 100 ml. of ether was added. The two phase mixture was shaken with enough 5% sodium bisulfite, in small portions, to remove the iodine color. The layers were separated and the aqueous layer was extracted with three 100 ml. portions of ether. The ether extracts were washed with three small portions of 10% sodium carbonate, and then dried over anhydrous magnes— ium sulfate. Evaporation of the ether solvent yielded a white crystalline residue. Recrystallization of this solid from petroleum ether (50-600) gave 12.48 g. of 2,2,5,5-tetramethyl- 1,5-cyclohexanedione (III), m.p. 95-960. was: B. 212L5,5-Tetramethyl-1,5~cyclohexanediol (IV) To 150 ml° of dry tetrahydrofuran (distilled from lithium aluminum hydride) in a 500 ml. three-neck flask fitted with a mechanical stirrer, an addition funnel and a condenser with a drying tube, was added 5.7 g. of lithium aluminum hydride. A solution of 10 g. of 2,2,5,5-tetramethyl-1,5-cyclohexanedione (III) dissolved in 50 ml. of dry tetrahydrofuran was placed in the addition funnel and was dropped Slowly into the vigorously stirred hydride suspension. When addition was complete, the mixture was stirred at reflux for four hours. After cooling to room temperature, excess hydride was destroyed by dropwise addition of ethyl acetate. The reaction mixture was then poured into 500 ml. of ice and water, and the resulting sus- pension was acidified to pH 1-2 by adding small portions of concentrated sulfuric acid. After the suspension was extracted with several portions of ether, the extracts were concentrated to a colorless oily liquid. The residue was dissolved in ether and the solution was washed with ice-water and dried over anhydrous sodium sulfate. Evaporation of the solvent yielded a white solid, which upon recrystallization from chloro- form gave 1.51 g. of fine white needles. Further crystalliza— tion of the mother liquor from hexane and acetone gave 6.99 g. of white crystals of mixed forms. The infrared Spectrum of each crop showed no carbonyl absorption. C. 2,2,5,5-Tetramethylcyclohexane 1,5-ditosylate (V) To a 250 ml. three-neck flask fitted with a magnetic stirrer, a stopper and a condenser with a drying tube was added 6.9 g. of 2,2,5,5-tetramethyl—1,5—cyclohexanediol (IV), 70 ml. of pyridine (reagent grade), and 17.1 g. of p-toluene- sulfonyl chloride (purified by recrystallization from hexane). The resulting yellow solution was warmed to 500 and stirred at that temperature for 60 hours. After cooling to room temperature, the mixture was poured onto ice and acidified by stirring in small portions of concentrated hydrochloric acid. The mixture was then extracted with four portions of chloro- form and the chloroform extracts were washed once with 5% hydrochloric acid. The extracts yielded a heavy oil from which a white solid precipitated when hexane was added. The solid (15.7 g.) was collected and washed with hexane. Recrystallization from hexane and acetone gave an amorphous white solid which showed characteristic tosylate absorptions at 1600, 1465, 1560, and 1170 cm'1 and no hydroxyl absorption in its infrared Spectrum. D. 5,5,6,6-Tetramethyl-1,4-cyclohexadiene (I) A 500 ml. three-neck flask was fitted with a magnetic stirrer, a stopper, a nitrogen bubbler, and an outlet tube to carry the nitrogen from the flask through a cold trap chilled in crushed ice and brine. A solution of 15 g. of 2,2,5,5- tetramethylcyclohexane 1,5-ditosylate (V) in 200 ml. of dry dimethyl sulfoxide was added to the flask. Potassium IHE .t-butoxide (29) (14.9) was added in one portion to this solu- tion and the mixture was stirred vigorously and warmed in an oil bath to 1500 over a period of one and one-half hours. A slow stream of nitrogen was passed through this mixture throughout the course of the reaction. After stirring for one hour at 1500, the mixture was allowed to cool to room temperature. The volatile liquid which had condensed in the trap was washed with two portions of ice-water and one portion of saturated sodium chloride solution. The remaining non- aqueous liquid was purified in 100 ul portions by v.p.c. (6 ft., 20% Apiezon "L" on Chromosorb~W at 1550). A total of 2.96 g. of a colorless liquid was collected. This proved to be identical with authentic 5,5,6,6-tetramethyl-1,4-cyclohexadiene (I) in v.p.c. retention time, infrared absorption Spectrum (Figure 1), and nuclear magnetic resonance spectrum (Figure 2). III. Vapor Phase Photolysis of 5,5,6,6-Tetramethyl- 1,4-chlohexadiene A Spherical Vycor flask with a single standard taper fit- ting and a total capacity of 550 ml. was used as the photolysis vessel. For sensitized reactions a small drop of clean mercury and 0.5-0.8 g. of 5,5,6,6-tetramethyl~1,4—cyclohexadiene (I) were added to the flask. The flask was immersed in a dry ice- isopropanol cooling bath and evacuated; it was then filled with purified nitrogen (1 atm.) and allowed to warm until the solid in the flask melted. After being immersed in the cooling bath, the flask was evacuated again and the flushing procedure was repeated two more times. The flask was evacuated a final time to 0.1 mm. and closed. The outer surface of the vessel was carefully cleaned and it was then suspended in the photo- chemical reactor and irradiated for sixteen hours. After be- ing removed from the reactor, the flask was set into a dry ice-isopropanol bath so that the bottom 5 mm. of it was cooled. Dry air was Slowly admitted and the product mixture was drawn out with a capillary pipette, when it melted. The walls of the flask were rinsed with a few ml. of pentane and this was added to the product. Volatile components of the mixture were separated from non—volatiles in a micro-distillation apparatus (1200 at 20 mm.). Unchanged starting material and photOproduct were obtained from the volatile fraction by preparative v.p.c. (6 ft., 20% Apiezon "L" on Chromosorb-W, 1200). The major photoproduct was a colorless liquid of molecular weight 156, determined from its mass spectrum. It exhibited infra-red absorption (Figure 5) at 5048, 1590, and 755 cm‘l. An ultra- violet spectrum showed an absorption maximum at 208 mu (6 = 5,400). The n.m.r. Spectrum of the compound (Figure 4) exhibited a multiplet at T 4.45 and complex resonance around T 8.9 with an integral ratio of 1:7.1. One minor photoproduct was observed, but in insufficient quantity to permit character— ization. 10 IV. Determination of Photoproduct Structure A. Ozonolysis of Photoproduct A 0.1 9. sample of the photoproduct was dissolved in 2 ml. of ethyl acetate (reagent grade) in a 10 ml. pear-Shaped flask fitted with a condenser. The solution was chilled in a dry—ice acetone bath and a stream of ozone was bubbled through it. After fifteen minutes, when the solution had turned blue, the ozone bubbler was removed and the solution was allowed to warm until the blue color dissipated. The solution was then immersed in the cooling bath and ozone was bubbled through again until a distinct blue color persisted. The bubbler was removed and the solution was allowed to warm as before. This sequence was repeated two more times. The solution was then allowed to warm to room temperature and the solvent was evaporated by a stream of nitrogen. A color— less viscous liquid remained. To this was added 2 ml. of a mixture of one part 50% hydrogen peroxide and three parts methanol. The resulting mixture was warmed to 600 and kept at that temperature for twelve hours. After cooling to room temperature, it was added to a solution of 2.1 g. of ferrous sulfate heptahydrate in 50 ml. of water. This mixture was extracted with four portions of ether, and the ether extracts were washed with two portions of ice—water and dried over an~ hydrous sodium sulfate. Evaporation of the ether left a viscous, slightly yellow oil. An infrared spectrum of the oil showed intramolecular hydrogen bonded hydroxyl absorption at 11 5650-2400 cm“1 and strong carbonyl absorption at 1710 cm‘l. B. Diazomethylation of Ozonolysis Product Freshly distilled diazomethane solution (50) was added to a solution of the crude diacid obtained by ozonolysis (VIII) in 10 ml. of ether until a distinct yellow color per- sisted. The mixture was then allowed to stand until the color dissipated, and was concentrated to 1 ml. The components of the mixture were separated by v.p.c. (5 ft., 4% QF-1 on Chromosorb-G, 1750). The predominant product, a light yellow liquid, was collected. Its infrared spectrum (Figure 5) showed no hydroxyl absorption and strong carbonyl absorption at 1755 cm-1. An n.m.r. spectrum (Figure 6) displayed sing- lets at T 6.57, 8.61, 8.65, and 8.80 with some complex sig- nals lying under the upfield resonance peaks. The integrated ratio of methoxy to upfield signal was 5:7.5. The mass spectrum of the diester (IX) exhibited a low intensity parent peak at m/e 228. C. Hydrogenation of Photoproduct A solution of 0.1 g. of photoproduct in 2 ml. of absolute ethanol was stirred with 0.1 g. of 10% palladium on charcoal under hydrogen at atmOSpheric pressure. After two hours ab- sorption stopped (one molar equivalent of hydrogen was taken up). The suSpension was filtered, added to 15 ml. of water and extracted with 15 ml. of pentane. The pentane solution was concentrated to 1 ml. and the reaction products were 12 isolated by v.p.c. (6 ft., 20% Apiezon “L" on Chromosorb-W, 1250). No peak appeared for unhydrogenated starting material. Two new peaks of longer retention time were present; the second peak was ca. 1% of the total. The major component was collected and determined to be a colorless liquid of molecu- lar weight 158 (mass spectrum). Its infrared absorption spectrum (Figure 7) showed no absorption in the fundamental region (4000 to 1600 cm-1) except for saturated C-H stretching. The n.m.r. Spectrum of the compound (Figure 8) exhibited com- plex methylene resonance above T 8.1, methyl singlets at 8.85, 8.95, 8.98, and 9.08 with some complex underlying signal, and resonance at 9.25 and 9.57. The integrated ratio of methylene to methyl region to higher field signal was 5.5:12.5:2. RESULTS AND DISCUSSION The synthesis of 5,5,6,6-tetramethyl-1,4-cyclohexadiene (I) was carried out following the scheme outlined in Chart 2. O O O HO H TSO OTS ———+> -———+» ———S> ——-a> II III IV V I Chart 2. Although 2,2,5,5-tetramethyl-1,5-cyclohexanedione (III) has been known for some time as a product from the methyl- ation of dimedon (14) (II), the compound has been obtained only in poor yields. In the course of this research, differ- ent reaction conditions were used in an effort to find a method of synthesizing the dimethylated material in higher yield. Reactions were run in dry dimethyl sulfoxide using potassium t—butoxide in tfbutanol, sodium ethoxide in ethanol and dry sodium ethoxide as bases. In technical dimethyl sulfoxide, sodium ethoxide in ethanol and aqueous sodium hydroxide were used as bases. Sodium ethoxide in ethanol was used as the base for an experiment in dry glyme, and aqueous sodium hydroxide was used in an aqueous ethanol reaction medium. The best results (52%) were obtained in anhydrous dimethyl sulfoxide using an ethanolic solution of sodium 15 14 ethoxide as the base. When the alkylation was stopped at a point where only monosubstitution Should have occurred, 2,5,5-trimethyl-1,5-cyclohexanedione (VI) was isolated (yields of 50% or less) as white plates from aqueous ethanol, m.p. 165-1650. On standing VI decomposes. High yields (84%) of III were obtained by subsequent methylation of VI with methyl iodide. The reduction of III with lithium aluminum hydride pro- ceeded readily in tetrahydrofuran solution (6), using 2.5 moles of reducing agent per mole of diketone. Allen, Sneeden and Colvin (15) have reported the separation of the isomers of 2,2,5,5-tetramethyl-1,5-cyclohexanediol (IV) obtained in this manner. The gig—diol is a meso compound and the trans- diol exists as a mixture of enantiomers. Recrystallization of the crude reaction product from chloroform gave nearly pure gig-diol as fine needles, m.p. 2010, subliming above 1700. After repeated fractional crystallization from acetone and hexane the trans-diol was obtained as octahedral crystals, m.p. 1060. The isomers were not separated effectively by sublimation or column chromatography. The purity of the isomers was checked by thin layer chromatography on micro slides. Adsorption of iodine by the substances proved to be an unsatisfactory developing technique since the trans-diol adsorbed iodine appreciably but the gig-diol did not. The slides were thus visualized by Spraying with an ethanolic solution of anisaldehyde and concentrated sulfuric acid. 15 In subsequent work the gigf and transfdiols were not sepa- rated and reactions were carried out on mixtures of the isomers. The possibility of going from IV directly to I by a bigfdehydration was theoretically attractive, but experi— mentally unattainable. Acid catalyzed dehydration was ruled out because of the probability of rearrangement. Unsuccess- ful attempts were made to effect dehydration of the diol with iodine, both neat and in solution, and by treatment with pyridine and thionyl chloride. The synthesis of 2,2,5,5-tetramethylcyclohexane 1,5- ditosylate (V) from IV was effected by warming a pyridine solution of pftoluenesulfonyl chloride and IV for an extended period. Reactions run at higher temperatures were troubled by decomposition of the products. The esterification was slowed by steric effects since both hydroxyl groups of IV are secondary. Additional hinderance probably resulted from the ggmfdimethyl groups in the diol. Work by Snyder and Soto (8) on relatively simple com- pounds showed that alkenes could be prepared in high yield from the corresponding tosylates by treating them with potas- sium t—butoxide in dry dimethyl sulfoxide. More vigorous conditions were required in order to obtain an efficient conversion of V to I. Volatile products were swept from the heated and stirred reaction mixture by a stream of nitrogen, and condensed to a cold trap. Additional product might have 16 been obtained from the reaction mixture, but it was not worked up because of its foul odor. Nevertheless, I was iso- lated in good yield (70%). In agreement with a previous report (6), the infrared Spectrum of I (Figure 1) exhibits absorptions at 5010 and 1655 cm‘l, and a very strong vinyl hydrogen deformation at 765 cm‘l; the n.m.r. Spectrum (Figure 2) consists of two sharp singlets at T 4.7 and 9.0 with an area ratio of 1:5. Irradiation of I in the vapor phase with mercury sensi- tization, according to Srinivasan's method of photolyzing 1,5—cyclooctadiene, yielded starting material and a new com- pound having a longer retention time in v.p.c. analysis. Microdistillation of the crude reaction products gave a clear liquid and a viscous yellow residue. A sample of the non- volatile material was analyzed by v.p.c. (6 ft., 20% SE-50 on Chromosorb-W, 1400) and proved to be a mixture of higher molecular weight components, probably dimers and other polymers. The structures of these products were not investigated. Unexpectedly, the photoreaction was observed to proceed with approximately the same efficiency in the absence of mercury. The ultraviolet Spectrum of the diene shows insig- nificant absorption at 2557 A (the major line in the resonance lamps employed in the photochemical reactor), however, weak higher energy bands are known to be present in this light source . 17 Hi EU 00m 000d .H mo Esuuommm Umumumcfl one OONd oo¢fi .fi musmflm Good 00mm 000m A _ _ . I . 18 Cd .H mo Esuuommm O.m .H.E.c mLB .m musmflm O.m 1 q _ _ 19 During the irradiations, the starting material was not entirely vaporized, and a liquid phase reaction could not be ruled out. To determine in which phase the reaction was taking place, samples of I were irradiated in a 12x100 mm. quartz test tube. In these experiments the ratio of I present in the vapor phase to that present as liquid was much less than for irradiations carried out in the Vycor flask, and intended to contain only vapor. After irradiation in the same manner as the vapor phase samples, yellowing of the sample was negligible and v.p.c. analysis showed only trace amounts of the photoproduct. Hence, the reaction appears to proceed mainly in the vapor phase. In subsequent work with 1,5-cyclooctadiene (16), Srinivasan observed that much better yields of the tricyclic product could be obtained if the diene were irradiated in an ether solution saturated with a complex of the diene and cuprous chloride. The catalytic effect of the cuprous ion is not understood, and it exerted no effect in solution photolysis of some other dienes. No photoproducts could be detected by v.p.c. after an ether solution of I, saturated with cuprous chloride, was irradiated for 24 hours. The use of benzene as a combination solvent and sensi- tizer appeared attractive because the W'fi> W* (V+N) transition at 256 mu (em 160) would absorb almost all the energy from ax the mercury resonance lamps. Energy transfer from the cor- responding triplet would then lead to an excited state of I, 20 provided the transfer is exothermic. When a degassed solution of I in benzene was irradiated for 24 hours a solid, insoluble in common solvents, was deposited on the walls of the photolysis vessel, but no volatile photoproducts were detected by v.p.c. and the quantity of I present did not decrease appreciably. The photoproduct (VII) was isolated from the volatile fraction of the vapor phase reaction products by preparative v.p.c., and proved to be a colorless liquid exhibiting a parent peak at m/e 156 in the mass spectrum. Other ions in the mass spectrum and their intensities are given in Table 1. Table 1. Major Peaks in the Mass Spectrum of VII Relative Relative m/e abundance m/e abundance 156 (P) 42.4 79 44.9 121 100.0 78 11.0 106 15.6 77 57.6 105 41.2 65 12.4 95 52.5 55 15.0 91 59.0 55 21.4 80 10.7 51 18.5 In addition to the usual absorptions of the methyl group and a ggmfdimethyl doublet at 1565 cm‘l, the infrared spectrum of VII (Figure 5) showed absorptions at 5048, 1590 and 755 cm'l. These are characteristic of the vinyl carbon-hydrogen bond, the carbon-carbon double bond and the vinyl hydrogen deformation of a cis-disubstituted double bond. Consequently, at least one double bond, weakly conjugated if at all, must 21 EU oom oooa .HH> mo Eduuommm Umumumcfl one coma oowfi .m musmflm OOmfi 00mm Doom _ 1 _ _ . _ 22 remain in the photoproduct. Chemical support for this con— clusion was found in the reaction of VII with dilute potassium permanganate solution (the Baeyer test). The n.m.r. Spectrum of VII (Figure 4) exhibited vinyl hydrogen resonance as a complex singlet at T 4.45 and the methyl hydrogens appeared as a strong multiple signal around T 8.92. Weaker complex resonance was present from T 8.2 to 8.4 and from the methyl region to T 9.15. The integrated ratio of vinyl to combined upfield signal was 1:7.1. The area of the signal from 8.2 to 8.4 was one-half the area of the vinyl signal. The weaker upfield signals are probably from one allylic cyclopropyl hydrogen, appearing downfield from the Inethyl region, and one cyclopropyl methine hydrogen at high :field. The configuration of the methyl resonance peaks sug— gfiested that VII did not have a plane of symmetry passing through tile gem-dimethyl groupings (as in H). When the Spectrum was rWJn as a 50% benzene solution, the methyl resonance was changed, as; illustrated in Figure 4, to a different unsymmetrical signal anypearing in the same region. This effect, caused by complex- aILion with the aromatic rings, is an additional indication of tile unsymmetrical structure of the photoproduct. An absorption maximum at 208 mu (6 = 5,400) was observed ill the ultraviolet spectrum of VII. Absorption in this region i4; characteristic of a weakly conjugated vinyl cyclopropane. Crrmwley (12) has reported that 5,4,6,6vtetramethylbicyclo(5.1.0)- hex—Z-ene exhibits )‘max 212 mu (E- = 5,000) . 25 Ofi .mcmNch CH ADV Ucm umm: O.m Amy HH> mo Esuuummm .H.E.c q a ’1- C3 24 The structure which best fits all the Spectral data is 4,4,6,6-tetramethylbicyclo(5.1.0)hex-2-ene. The infrared Spectrum is consistent with this structure and the ultraviolet and n.m.r. spectra are in good agreement with corresponding data from other examples of this ring system reported by Freeman (17), Crowley (12), and Meinwald (15). The bicyclo(5.1.0)hexene structure present in the photo- product has been observed in products isolated after photolysis of certain trienes (18-20) and 1,5—cyclohexadienes (21,22). It is also similar to the lumi-products formed in the well- known cyclohexadienone photorearrangement (25). Unpublished work by Hart's group (24) concerning the photolysis of cross- conjugated methylene cyclohexadienes has demonstrated the formation of analagous products. The conversion of I to VII can be rationalized by a Zimmerman type mechanism (Chart 5). I VII * 1 Chart 5. 25 Additional evidence supporting structure VII was ob- tained from the reactions outlined in Chart 4. HOgC MHSCOBC I><—-> HOgC H3C02C VIII ‘ / VII ———9- x Chart 4. When VII was ozonized, followed by oxidative work-up, the crude product showed broad intramolecular hydrogen bonded hydroxyl absorption at 5650-2400 cm“1 and strong carbonyl absorption at 1710 cm-1 in the infrared Spectrum. These features were very similar to corresponding absorptions in the Spectrum of 8,6-dimethylglutaric acid (25). The crude diacid (VIII) obtained from the ozonolysis of VII was esterified by reaction with diazomethane. The major product from the esterification was isolated by preparative v.p.c.; several minor products of Shorter retention time were present but were not collected. The purified diester (IX) was a light yellow liquid exhibiting carbonyl absorption at 1755 cm‘1 in the infrared spectrum (Figure 5). No carbon-hydrogen 1 stretch was observed above 5015 cm‘ , indicating that olefinic hydrogens were no longer present. As expected, no cyclopropyl 26 IEU .NH mo Esupommm Umumuwca $38 coma oowa .m musmflm coma coma comm 000m 9) oooa _ H 4 _ _ H a 1 q 27 carbon-hydrogen stretching was observed at 5050 cm-1, since this absorption is lacking unless a cyclopropyl methylene is present (26). The n.m.r. spectrum (Figure 6) of IX showed a strong singlet at T 6.57 and methyl resonance at T 8.61, 8.65 and 8.80. A complex resonance signal, attributable to the cyclo- prOpyl methine hydrogens, was present above T 8.5 and ex- tended through the methyl region; the hydrogen a to the carbomethoxy group appeared in the low field section of this range. The area ratio of methoxy to upfield signal was 5:7.5. The mass spectrum (27) of IX had a very low intensity parent peak at m/e 228. Low intensity peaks also appeared at 215, 197 and 169, representing parent minus methyl, methoxy and carbomethoxy ions respectively. A list of the most prominent ions in the spectrum is given in Table 2. Mass Spectra of this nature are not unusual for diesters (28). Table 2. Major Peaks in the Mass Spectrum of IX Relative Relative m/e abundance m/e abundance 127 100 59 17 109 54 55 20 102 21 55 12 95 60 44 45 85 11 45 19 75 22 41 45 67 57 59 25 28 ..r oa .XH mo Esuuummm .H.E.c msfi .m mndmflm O.m .> 29 The products obtained from the ozonolysis of VII ap- peared to be quite sensitive to acid, since products of a dif- ferent nature, probably lactonized esters, were isolated when mild acidic conditions were employed during the work-up. One molar equivalent of hydrogen was taken up when VII in ethanol was stirred with palladium on charcoal under one atmosphere by hydrogen. Analysis of v.p.c. showed only one minor impurity in the product. The purified hydrogenation product (X) had a very simple infrared Spectrum (Figure 7) in which no functionality of any sort was apparent. Its mass Spectrum exhibited a parent peak at m/e 158 and other abundant ions presented in Table 5. Table 5. Major Peaks in the Mass Spectrum of X Relative Relative m/e abundance m/e abundance 158 5.8 79 10.2 124 10.2 77 10.0 125 100.0 75 60.5 95 44.5 69 54.5 85 14.4 67 44.5 82 51.4 55 24.5 81 58.6 55 19.1 The n.m.r. Spectrum (Figure 8) of X showed complex methylene resonance starting at T 8.1 and continuing upfield into the methyl region where four singlets of nearly equal intensity appeared at T 8.85, 8.95, 8.98 and 9.08. Signals were also 50 HI EU 00m OOOH .N mo Esuuommm Umumnmcfl one coma oowa .> musmflm oomfi 00mm ooom . L A _ _ A _ _ _ (— 51 .x mo Esuuowmm O.m .H.E.: one o.m . .m musmflm 52 present at T 9.25 and 9.57 and are attributed to the cyclo~ prOpyl methine hydrogens. The integrated ratio of methylene to methyl to higher field Signal was 5.5:12.5:2. The sym~ metrical structure H (Chart 1) can thus be eliminated as an alternative formulation for VII. 1. 2. 6. 7. 9. 10. 11. 12. 15. 14. 15. 16. 17. LITERATURE CITED A summary of pertinent publications is found in reference 6. C. F. Wilcox, Jr., S. Winstein, and W. McMillan, J. Am. Chem. Soc., 82, 5450 (1960). L. w. Pickett and E. Sheffield, ibid., ggj 216 (1946). E. E. van Tamelen, ibid., ll, 1704 (1955). . D. T. Longone and G. Smith, Tetrahedron Letters, 205 (1962). W. Reusch, M. Russell, and C. Dzurella, J. Org, Chem., $2; 2446 (1964). F. W. Grant, R. W. Gleason, and C. H. Bushweller, ibid., L9, 290 (1965) . “— . C. H. Snyder and A. R. Soto, J. Org. Chem., 29 742 (1964). W. G. Dauben and R. L. Cargill, Tetrahedron, 15, 197 (1961) . — R. Srinivasan, J. Am. Chem. Soc., 85, 819 (1965). H. H. Stechel, Angew. Chem., 15, 1176 (1965). K. J. Crowley, Tetrahedron Letters, 2865 (1965). J. Meinwald, A. Eckell, and K. L. Erickson, J. Am. Chem. Soc.,-81, 5552 (1965). R. D. Desai, J. Chem. Soc., 1079 (1952). A. Allen, R. Sneeden, and J. Colvin, ibid., 557 (1957). R. Srinivasan, J. Am. Chem. Soc., 85, 5048 (1965‘. P. K. Freeman, M. F. Grostic, and F1 A. Raymond, J. Org. Chem., 50, 771 (1965). 55 18. 19. 20. 21. 22. 25. 24. 25. 26. 27. 28. 29. 50. 54 w. G. Dauben,_§£‘ 1., J. Am. Chem. Soc., 0, 4116 (1958). w—-—4 — W. G. Dauben and P. Baumann, Tetrahedron Letters, 565 (1961). D. H. R. Barton and A. S. Kende, J. Chem. Soc., 688 (1958). G. R. Evanega, W. Bergmann, and J. English, Jr., J. Org: Chem., El, 15 (1962). H. Prinzbach and J. H. Hartenstein, Angew Chem- (Int. Ed.), 477 (1965). H. E. Zimmerman, gt 1., J. Am. Chem. Soc., fig, 1965 (1966). A. Sheller, Private communication, 1966. Sample provided through the courtesy of C. Markos. K. Nakanishi, Infrared Absorption Spectroscopy, Nankodo Company Limited, Tokyo, 1962, p. 20. Gratitude is expressed to Professor C. Djerassi for obtain- ing this spectrum. D. Bak, Private communication, 1966. Organic Syntheses, John Wiley & Sons, Inc., New York, 1961, Vol. 41, p. 16. Organic Syntheses, John Wiley & Sons, Inc., New York, 1965, coll. vol. 4, p. 152. "71111711lefillilflffl’l‘flil'lliluflflfflflfllflm