PART! THE THERMAL REARRANGEMENT OF. f w 1 ~ METHYLENE - 2,3, 4,4,5; '6 - HEXAMETHYL- 2, 5 -~ cvczowuqmma PART H 7' . THE PHOTOCHE’MICAL REARRANGEMENT 0F 1 - METHYLENE - 2,3, 4, 4, 5, 6 - HEXAMEFHYL - 2, 5, - CYCLOHEXADIENE Thesis for the Degree of Ph. D. MICHIGAN STATE UNIVERSWY JERRY DEAN DE VRIEZE 1968 l’ H QNC tun-f, This is to certify that the thesis entitled PART I THE THERMAL REARRANGEMENT OF 1—METHYLENE-2,3,4,4,5,6- HEXAMETHYL-Z,5-CYCLOHEXADIENE PART II THE PHOTOCHEMICAL REARRANGEMENT OF l-METHYLENE- 2,3,4,4,5,6-HEXAMETHYL-2,5-CYCL0HEXADIENE presented by Jerry Dean De Vrieze has been accepted towards fulfillment of the requirements for Ph. D. degree in Chemistry ¥'L7LA~{Et1: J'l<*=:<1’—fl——— Y Major professor Date December 19, 1968 0-169 ABSTRACT PART I THE THERMAL REARRANGEMENT OF l-METIIYLENE-Z ,3 , 4 I 4 I 5 I 6-HEMTHYIJ’ 2.5-CYCLOHEXADIENE BY Jerry Dean De Vrieze The purpose of the first part of this thesis was to in- vestigate the mechanism of the thermal rearrangement of l-methylene-Z,3,4,4,5,6-hexamethyl—2,5-cyclohexadiene (£1) to ethylpentamethylbenzene (12). The apparent similarity of this rearrangement to the Auwer's rearrangement - in which a dichloro- or trichloromethyl group migrates from the 4-position of a 1-methy1ene—2,5-cyclohexadiene to the eggfmethylene carbon(1)-led us to investigate the reaction in some detail. Pyrolysis of 11 in the gas phase (420°) produced l§ and decamethylbibenzyl (El). Pyrolysis of 11' in dioxane (100°) and in decalin (160°) also produced lg; a small amount of g; was formed in the latter case. a E are 17 18 21 W "W W 2 Jerry Dean De Vrieze Pyrolysis of lz'in cumene (150°) produced lg’and 2-phenyl-2-pentamethylbenzylpropane (fig). Lower initial concentrations of 11 led to more gg and less l§x Pyrolysis of 11.1“ cumene containing thiophenol gave hexamethylbenzene (4g), phenyl pentamethylbenzyl sulfide (42) and diphenyl disulfide (Q2). CH3 ¢-c-CH cpscn2 ¢ss¢ CH3 46 48 49 50 The thermal rearrangement Of.l1 is postulated to pro- ceed by a radical chain mechanism in which the reaction is initiated when 11 is cleaved to produce methyl radicals and pentamethylbenzyl radicals (42) (step a) or when it reacts with a radical initiator (g) to produce methyl radicals (step a'). A general scheme which accounts for the observed products as a function of reaction conditions is shown below. 3 Jerry Dean De Vrieze 41 + 17-——> g;,+ 'CH3 (b) 'CH3 + 11,-”> 1§,+ 'CHa (c) CH3 CH3 'CH3 + ¢-< ¢———<: + CH4 Id) CH3 CH3 51 2.1. + u > 22 + as (e) ‘CH3 + ¢SHg > ¢S° + CH4 (f) g; + CDSH > + ¢S' (9) 2,8: ¢S' + 11’ > 52’ + ’CH3 (h) 2 ws > ¢ss¢ (i) 29. At high temperatures, the difference in reactivity of the two radicals formed in step a is small, and nearly equimolar amounts of 1Q and 21 are formed (steps b and c). In solution, at lower temperatures, step a is slower and step a' becomes important in initiating the reaction. In dioxane, in decalin, and in cumene the methyl radical (which is more reactive than 41) leads to products by steps c, d, and e, and the only product derived from gl'is a trace of gl'which is formed in decalin. In the presence of thiophenol 4 Jerry Dean De Vrieze the methyl radical is trapped by hydrogen transfer. This leads to thiophenoxy radical (step f) which either adds to £1 (step h) or couples (step i). The pentamethylbenzyl radical (21) formed in step a leads to 4Q (step g) by hydrogen abstraction. PART II THE PHOTOCHEMICAL REARRANGEMENT OF 2,5-CYCLOHEXADIENE The second part of this thesis is concerned with the photochemical rearrangement of 11; Irradiation of 11' through a Vycor filter produced 4—methylene-1,2,3,5,6,6- hexamethylbicyclo[3.1.0]hex-2-ene (22). The reaction could not be sensitized by irradiation of 11.1“ acetone nor was it quenched by piperylene. *4 17 74 W W The irradiation of 11 which contained CD3 groups at the C3- and C5-positions produced 22 which had CD3 groups at the C1- and Cz-positions. This established that the isomerization occurred gig a bond-crossing mechanism and that methyl migration was not involved. 5 Jerry Dean De Vrieze It was also found that diene 24 underwent a facile thermal rearrangement to 5—i§gpropenylpentamethylcyclopenta- diene (22). This isomerization was shown to proceed by the cleavage of the C1-C6 bond of 22 (at 200°) by using 14 which contained a deuterium label in various positions. 74 75 Dehydration of the alcohol 4-hydroxy-1,2,3,4,5,6,6- heptamethylbicyclo[3.1.0]hex-2-ene (§§) by a number of methods produced 22 with no evidence for the intermediacy MU W of 24; REFERENCES 1. Auwers, K., Chem. Ber., §§/ 2167 (1922), and earlier references in this paper. PART I THE THERMAL REARRANGEMENT OF 2.5-CYCLOHEXADIENE PART II THE PHOTOCHEMICAL REARRANGEMENT OF 1*METHYLENE-2,3,4.4,5,6-HEXAMETHYL- 2.5-CYCLOHEXADIENE BY Jerry Dean De Vrieze A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1968 ACKNOWLEDGMENT The author wishes to express his sincere appreciation to Professor Harold Hart for his interest, guidance, and encouragement during the course of this study and for his help in perfecting this thesis. Appreciation is also extended to the National Science Foundation, to the Army Research Office (Durham), and to Minnesota Mining and Manufacturing for providing financial support. ii TABLE OF CONTENTS PART I THE THERMAL REARRANGEMENT OF 1-METHYLENE-2,3,4,4,5,6-HEXAMETHYL-2,5-CYCLOHEXADIENE Page INTRODUCTION . . . . . . . . . . . . . . . . . . . . 2 RESULTS AND DISCUSSION . . . . . . . . . . . . . . . 7 A. Gas Phase Thermal Rearrangement of 1-Methylene- 2,3,4,4,5,6—hexamethyl—2,5-cyclohexadiene . 7 1. Pyrolysis on a Gas Chromatograph . . . 7 2. Gas Phase Pyrolysis of 17 . . . . 10 3. Gas Phase Pyrolysis of Labeled Trienes. 11 4. Mechanism of the Thermal Rearrangement of 11 in the Gas Phase . . . . . . . . 18 B. Pyrolysis of 1-Methylene-2,3,4,4,5,6—hexa— methyl-2,5-cyclohexadiene (11) in Solution . 20 1. In Dioxane . . . . . . . . . . . . . . 20 2. In Decalin . . . . . . . . . . . . . . 21 3. In Cumene . . . . . . . . . . . . . 22 4. In Cumene Containing Thiophenol . . . . 25 5. Pyrolysis of Labeled Triene 38 in Cumene and in Decalin . . . . 27 6. Attempted Kinetic Study of the Thermal Rearrangement of 17 in Decalin . . 29 7. The Effects of Oxygen and a Free-Radical Source on the Rate of Reaction of 17 in Decalin . . . . . . . . . . . 31 8. The Mechanism of the Thermal Rearrange— ment of 12.1“ Solution . . . . . . . . 33 iii TABLE OF CONTENTS (Continued) Page EXPERIMENTAL . . . . . . . . . . . . . . . . . . . . 42 A. General Procedures . . . . . . . . . . . . . 42 B. Preparation of 1-Methylene—2,3,4,4,5,6-hexa- methyl-2,5-cyclohexadiene (11) . . . . . . . 43 C. Preparation of Ethylpentamethylbenzene (18) from Pentamethylbenzene (19) . . . . . . . . 44 D. Preparation of Decamethylbibenzyl (21) from Chloromethylpentamethylbenzene (22)w . . . 45 B. Preparation of 2—Phenyl—2—pentamethylbenzyl- propane (46) from Phenyldimethylcarbinyl Chloride 737) and Chloromethylpentamethyl- benzene (22) . . . . . . . . . . . . . . . . 46 F. Pyrolysis of 1—Methylene—2,3,4,4,5,6-hexamethyl- 2, 5- -cyclohexadiene (17) on a Vapor Phase Chromatograph . . . . . . . . . . . . . . . 48 G. Gas Phase Pyrolysis of 1-Methylene-2,3,4,4,5,6- hexamethyl—Z,5-cyclohexadiene (11) . . . . . 49 H. Gas Phase Pyrolysis of l-Dideuteromethylene— 2,3,4,4,5,6-hexamethyl-2,5—cyclohexadiene (£1) 51 I. Gas Phase Pyrolysis of 1-Methylene—4,4-bis~ (trideuteromethyl)—2,3,5,6-tetramethyl-2,5- cyclohexadiene (34) . . . . . . . . . . . . 52 J. Gas Phase Pyrolysis of 1-Methylene—4—trideutero— methyl—2,3,4,5,6-pentamethyl-2,5-cyclohexa- diene (3Q) . . . . . . . . . . . . . . . . . 54 K. Pyrolysis of 1-Methylene—2,3,4,4,5,6—hexamethyl— 2,5-cyclohexadiene (11) in Dioxane . . . . . 55 L. Pyrolysis of 1-Methylene-3, 5, -bis(trideutero- methyl)- -2, 4, 4, 6— tetramethyl— —2, 5- -cyclohexa— diene (44) in Dioxane . . . . . . . 56 M. Pyrolysis of 1—Methylene-2,3,4,4,5,6—hexamethyl- 2,5-cyclohexadiene (11) in Decalin . . . . 57 N. Pyrolysis of 1-Methylene- -2, 3, 4, 4, 5, 6—hexamethyl- 2, 5- -cyclohexadiene (17) in Cumene . . . 59 iv TABLE OF CONTENTS (Continued) SPECTRA SUMMARY Pyrolysis of l-Methylene—Z, 3, 4, 4, 5, 6—hexa- methyl- -2, 5- -cyclohexadiene (17) in Cumene Containing Thiophenol . . . . . . . . . . Pyrolysis of Phenyl Pentamethylbenzyl Sulfide (42) in Cumene Containing Thiophenol . . . Pyrolysis of 1—Methylene-4— trideuteromethyl— 2, 3, 4, 5, 6-pentamethyl- -2, 5- -cyclohexadiene (38) in Cumene and in Decalin . . . . . . . . . . Attempted Kinetic Study of the Thermal Rear— rangement of 11.1“ Decalin . . . . . . . . . Effect of Oxygen and Di-tebutyl Peroxide on the Rate of Reaction of 11 in Decalin PART II THE PHOTOCHEMICAL REARRANGEMENT OF INTRODUCTION RESULTS AND DISCUSSION . . . . . . . . . A. Photochemical Rearrangement of 1-Methylene- 2,3,4,4,5,6—hexamethyl-2,5-cyclohexadiene (11) 1. Isolation and Structure Determination of the Product . . . . . . . . . 2. Solvent Effect, Quenching, Sensitization, and the Effect of Different Wavelength Light . . . . . . . . . . . . 3. Photolysis of Labeled Trienes . . . 4. Mechanism of the Photochemical Rearrange— ment of 11 . . . . . . . . . . . . . . Thermal and Chemical Formation of 5— Iso Opropenyl— pentamethylcyclopentadiene (75) Mechanism of the Thermal Rearrangement of a Homofulvene (Z4) to a Cyclopentadiene (15) V Page 60 63 63 65 66 68 74 77 83 83 83 87 88 92 94 102 TABLE OF CONTENTS (Continued) Page EXPERIMENTAL . . . . . . . . . . . . . . . . . . . . 107 A. Photochemical Rearrangement of 1-Methylene- 2,3,4,4,5,6—hexamethyl-2,5—cyclohexadiene (11) 110 B. Photochemical Rearrangement of l—Methylene— 3,5-bis(trideuteromethyl)-2,4,4,6-tetramethyl— 2,5-cyclohexadiene ($4) . . . . . . . . . . 110 C. Photochemical Rearrangement of 1-Dideutero- methylene— 2, 3, 4, 4, 5, 6-hexamethyl— —2, 5— —cyclo- hexadiene (23) . . . . . . . . . . . . 111 D. Preparation of 5-—Isopropeny1pentamethy1cyclo- pentadiene (75) . . . . . . . . 112 E. Maleic Anhydride Adduct of 15 . . . . . . . 113 F. Tetracyanoethylene Adduct of Z5 . . . . . . 114 G. Preparation of 5— —Iso opropylpentamethylcyclo— pentadiene (89) . . . . . . . 115 H. Tetracyanoethylene Adduct of 89' . . . . . . 116 I. Thermal Rearrangement of 74 to 75’ . . . . . 117 J. Thermal Rearrangement of 4—Methylene——1,2—bis- (trideuteromethyl)- -3, 5, 6, 6— —tetramethylbicyclo— [3. 1. 0]hex- -2—ene (79) . . . . . . . 118 K. Thermal Rearrangement of 4-Dideut.eromethylene— 1, 2, 3, 5, 6, 6-hexamethylbicyclo[3. 1. O]hex- -2— ene (82) . . . . . . . . . . 119 L. Attempted Preparation of 4—Methylene—1,2,3,5,6,6— hexamethylbicyclo[3.1.0]hex—2—ene (14) . . . 119 SPECTRA 123 SUMMARY . . . . . . . . . . . . . . . . . . . 137 LITERATURE CITED . . . . . . . . . . . . . . . . . . 139 vi TABLE II. III. IV. VI. VII. VIII. LIST OF TABLES NMR spectra of trienes and the ethylpenta- methylbenzenes derived from them . . . Product distribution from the pyrolysis of ll’in cumene . . . . . . . . . . . . . Ratio of products 48 and 49 from the pyrolysis of 0.09 M 17 in cumene cont.aining thiophenol. Mass spectra of labeled ethylpentamethyl- benzenes from triene 38 . . . . . . . Reaction of 17 in the presence of oxygen and di- t- -butyl peroxide . . . . . . Spectral data of homofulvenes . . . . . . Solvent effect on the photochemical rearrange— ment of 11’ . Spectral data of cyclopentadienes . . . . . . vii Page 13 24 27 29 32 85 89 96 FIGURE 10. 11. 12. 13. LIST OF FIGURES Page Kinetics of the thermal rearrangement of 11 in decalin . . . . . . . . . . . . . . . . 30 Partial mass spectrum of ethylpentamethylbenzene (1.8.) 69 Partial mass spectrum of labeled ethylpenta— methylbenzene (39) from the gas phase pyrolysis of triene 38 . . . . . . . . . . . . . . . . 69 Partial mass spectrum of labeled ethylpenta- methylbenzene (39) from t.he pyrolysis of 0.10 M 38 in cumene atmTSOO . . . . . . . . . . . 70 Partial mass spectrum of labeled ethylpenta— methylbenzene (39) from the pyrolysis of 0.82 M 38 in cumene at~T50° . . . . . . . . . . . . 70 Partial mass spectrum of labeled ethylpenta- methylbenzene (39) from the pyrolysis of 0.97 M 38 in decalin at ~~150°. . . . . . . 71 NMR spectrum of ethylpentamethylbenzene (18) . 72 NMR spectrum of labeled ethylpentamethylbenzene (32) from the gas phase pyrolysis of triene 34' 73 NMR spectrum of 4-methylene-1,2,3,5,6,6—hexa- methylbicyclo[3.1.0]hex—2—ene (14) . . . . . . 124 NMR spectrum of 4—met.hylene-1,2—bis(trideutero— methyl)- -3, 5, 6, Gauaramethylbicyclo[3. 1. O]hex' 2- —ene (79) . . . . . . . . . . . . . . . 125 NMR spectrum of 4—dideuteromethylene—1,2,3,5,6,6- hexamethylbicyclo[3.1.0]hex-2—ene (82) . . . . 126 NMR Spectrum of 5—iso opropenylpentamethylcyclo— pentadiene (75). . . . . . . . . . 127 NMR spectrum of 5- —isopropenyl-1,2—bis(tri— deuteromethy1)- -3, 4, 5—trimethylcyclopentadiene (106) . . . . . . . . . . . . . . . . . . . 128 viii LIST OF FIGURES (Continued) FIGURE 14. 15. 16. 17. 18. 19. 20. 21. Page NMR spectrum of 5—isopropenyl-1-dideutero— methyl-2,3,4,5—tetramethylcyclopentadiene (108) 129 NMR spectrum of 5-isopropenylpentamethylcyclo- pentadiene from the pyrolysis of 106 at 2750 . 130 IR spectrum of 4-methylene—1,2,3,5,6,6-hexa- methylbicyclo[3.1.0]hex—2—ene (Ii) . . . . . . 131 IR spectrum of 4-methylene-1, 2- -bis(trideutero- methyl)- 3, 5, 6, 6— —tetramet.hylbicyclo[3. 1. 0]hex— 2—ene (79) . . . . . . . 132 IR spectrum of 4-dideuteromethylene—1,2,3,5,6,6- hexamethylbicyclo[3.1.0]hex—2—ene (82) . . . . 133 IR spectrum of 5— —isopropenylpentamethylcyclo— pentadiene (75) . . . . . . o . . . . 134 IR spectrum_of 5-iso opropenyl- -1, 2-bis(tri- deuteromethyl}3, 4, 5trimethylcyclopentadiene (106) . . . . . . . . . . o . o . . . . . . . 135 IR spectrum of 5-isopropenyl—1-dideuteromethyl- 2,3,4,5-tetramethylcyclopentadiene (108) . . . 136 ix PART I THE THERMAL REARRANGEMENT OF 1-METHYLENE-2,3,4,4,5,6-HEXAMETHYL-2,5-CYCLOHEXADIENE INTRODUC TI ON The thermal rearrangement of 1-methylene—2,5—cyclo— hexadienes to aromatic compounds was first reported by Auwers in 1903 (1). Over the years Auwers and coworkers published several papers on the subject (2—8) and more recently other examples have appeared (9-11). The product arises from a 1,5-shift of a dichloromethyl or trichloro- methyl group. In no case did the other group in the 4- position, usually a methyl, migrateo Some examples are shown below. CHC12 A O > CHC12 1' 2 Auwers and Kiel, (1) (1903) COZH CleC COZH A > O CHC12 4 ~ Auwers, (7) (1911) 200 2 cc13 U A —"—_'_> cc13 g 6 Auwers and Julicher, (8) (1922) CHC12 ‘“ 1 3 ~ CHC12 Fuson and Miller, (9) (1952) coast c13c COzEt A > cc13 9 10 ~ rw Newman and Eberwein, (10) (1964) $co13 U A O ——"—> cc13 2.1. 12 w Patel and Schuster, (11) (1967) Eto2 Eto2 4 The first mechanism proposed for the 1,5—shift invol- ved ionic intermediate 122.(12)' This unusual mechanism was presumably fashioned after that of the Claisen rearrange- ment, and apparently was designed to avoid the postulation of an ion pair -— CC13- and benzyl cation -— because in this case the positive charge would have to end up a to a carbonyl function. ¢(//AEC12 EtOZC cc13 C I K1 ‘ /C1 > H > O KJ¢C12 13 13a 14 W W ’W Tse and Newman, (12) (1956) Later work by Bird and Cookson (13) showed that the rearrangement probably involves a free-radical chain pro- cess with a dichloromethyl or trichloromethyl group as the chain carrier. Bird and Cookson found that the rearrange- ment of g'to 2 involved a temperature dependent induction period, after which good first order kinetics were obeyed. Added benzoyl peroxide reduced the induction period and duroquinone inhibited the reaction. In addition, ultra— violet irradiation of §’at room temperature produced 2; The postulated reaction path is shown below. COZH ' C02H > <:) + 'CHC.12 c11012 (00 CleC COZH ’3‘, + oCI‘IC].2 —_> CHc12 ClZHC COZH 3a _—__> O + oCHCIZ 2, After the present work was completed, a paper appeared by Newman and Layton (14) who found that the thermal rear- rangement of optically active 9 gave racemic 12; This re— sult eliminated the original concerted Newman mechanism which would require that the product be optically active; it was, however, consistent with a free-radical chain mech- anism. This rearrangment also had an induction period after which first order kinetics were obtained. The reaction was inhibited by radical traps such as benzoquinone. Exposure to light also converted g'to 12; 6 The 1,3—shift of trichloromethyl groups in 1-methylene- 2,4-cyclohexadienes has also been proposed (15-17). The conversion of 15 to 16 is postulated to proceed via the m intermediate 15a. r— —I Pc15 > > ccl3 ccl3 cc13 _ .1 12 12%. 12, Newman and Wood, (15) (1959) In the present work it was discovered that the hydro- carbon 1-methylene-2,34n4,5,6—hexamethyl-2,5-cyclohexadiene (11) rearranged in the heated injector port of a gas chroma- tograph to ethylpentamethylbenzene (lg). Since the reaction appeared to involve simple methyl migration, whereas pre- viously reported rearrangements of this type involved a di- chloromethyl or trichloromethyl migrating group, the reac— tion was investigated further, both in the gas phase and in solution. Evidence is presented to show that the thermal rearrangement of lZ,tO lgtoccurs by a free-radical chain mechanism. 17 18 RESULTS AND DISCUSS ION A. Gas Phase Thermal Rearrangement of l-Methylene— 2,3,4,4,5,6-hexamethyl-2,5-cyclohexadiene (11) 1. gyrolysis on a Gas Chromatograph During the course of an investigation of the photo- chemical rearrangement of 12 (Part II of this thesis) it was discovered that ll rearranged thermally. When solutions of 11 were injected into a gas Chromatograph two peaks were detected. The first was shown to be due to $1, The com- pound which caused the second peak accounted for 57% of the volatile material at an injector temperature of 325°. At lower temperatures, e.g. 200°, the second peak was absent. The new compound was collected as a white solid and was in— vestigated. The infrared spectrum showed the absence of carbon-carbon double bonds or olefinic hydrogens. An ethyl group was indicated by the nmr Spectrum which had a quartet at r 7.38 (2H) coupled to a triplet at T 8.97 (3H), J - 7.4 cps. Two singlets at T 7.83 and 7.86 integrated for 15 pro- tons. The peak at 7.86 was larger but the two peaks could not be successfully integrated independently. The positions of the singlets are very similar to the peak for hexamethyl- benzene which appears at T 7.85. This indicated that the 7 8 singlets were due to five methyl groups on an aromatic ring. Based on this information the compound was considered to be ethylpentamethylbenzene (£8). Compound 18 was also synthesized by the Friedel-Crafts ethylation of pentamethyl- benzene (12). . o o — < > A1c13 17 18 19 w "W w The melting point of 18 prepared in this way was 59-610. The reported melting point of 18 in the literature was 125° (18). After the present work was completed, a note appeared by Berezina and coworkers (19) who also observed the thermal rearrangement of lz'to 18, after heating 11 for five hours at 150° in heptane. The physical and spectral properties of lg'reported by these authors were in complete agreement with those obtained in this work (see the Experimental section). They did not report any mechanistic study of the reaction. The rearrangement of lz'to 18 can be viewed most simply as a direct migration of one of the C4 methyl groups to the gngmethylene carbon. As such it would be analogous to the Auwers rearrangement with the exception that in this case an alkyl (methyl) radical migrates whereas the previously ob- served migrating groups were dichloromethyl and trichloro— methyl radicals (see the Introduction). Other mechanistic 9 paths are also feasible. Consecutive 1,3-methyl shifts, with conjugated triene gg_as an intermediate, would also lead to 18; i W# 0 17 20 18 m rw ”W The second step in this path is analogous to that proposed by Newman and Wood (15) for the conversion of lg’to 16, 0 cl 1 PCl - 5> 0 > O CC13 CC13 CCl3 L _J 1.3. ms 12 Equilibration of the ring methyls through intermediate 17a with the final product being formed from 12” or 22/ or both, is also possible. 9H2 Io ‘\ < > < 11’ 17a k////// 10 In order to determine which, if any, of these mechanisms was Operative, the rearrangement was investigated further under a variety of conditions and with the aid of deuterium labeling. 2. Gas Phase Pyrolysis of 12‘ Triene 11 was pyrolyzed in the gas phase by passing it through a heated (420°) Pyrex tube containing Pyrex beads. A nitrogen atmosphere was maintained throughout the pyrol- ysis and the contact time in the heated portion of the tube was approximately six seconds. Two products were obtained from the pyrolysis of 11 under these conditions. Ethyl- pentamethylbenzene (18) was obtained in 52% yield along with 48% of a high melting solid. This new compound ex- hibited three singlets in its nmr Spectrum, at T 7.10 (4H), 7.58 and 7.72 (30H total)° The simplicity of this spectrum suggested a highly symmetrical molecule. The most likely structure was considered to be decamethylbibenzyl (2}). Compound gl'had been previously prepared by Humphlett and Hauser by the Grignard coupling of Chloromethylpentamethyl- benzene (22) (20)° A sample of 21 prepared in this manner ‘was identical in all respects to the product obtained from the pyrolysis of 11. 11 0 420 > + CHZ-CHZ Cl M, CH)o O 9(252 2’1» D {l 22 m 3. Gas Phase Pyrolysis of Labeled Trienes Three deuterium-labeled trienes, 22/ £2” and 88“ were prepared and pyrolyzed in order to gain an insight into the mechanism of the thermal rearrangement. 1-Dideuteromethylene-2,3,4,4,5,6—hexamethyl-2,5-cyclo— hexadiene (triene 22) was prepared from the reaction of hexamethyl-Z,5-cyc10hexadienone (24) (21) with trideutero- methylmagnesium iodide. The nmr spectrum of g§'(Table I) showed 0.7 protons at T 5.26 for the gngmethylene group, indicating that some label had been lost, probably during workup. This can happen because 11 is unusually basic for a hydrocarbon (22). Exchange can occur yia_the heptamethyl- . . + benzenonium cation (17—H ). 12 + H + -H 17 17-H rw + Pyrolysis of 22 gave the labeled products 22 and 22' (see Table I; the asterisked positions on these and follow- ing figures denote the position of the deuterium label). The nmr spectrum of 22 differed from that of 18 in that the methylene quartet at T 7.38 integrated for only 0.7 protons. In addition the methyl peak at T 8.97 was broad, as expected for the presence of 0, 1, and 2 protons on the adjacent methylene. The nmr spectrum of 22 differed from that of 21,1“ that the methylene peak at T 7.10 integrated for 1.4 instead of 4 protons. O * .1 CD 3MgI A —-———-—> ---—-—-> 2.2. 2.52, A * * O + CHz—CHZ 13 mum was .ssonm mum maaouw woumufipcw map you mxwma msu mo muwfism HmUHEwnu use .xmaHOumm an %n kuoc msam> mcu cu cowumNHHmEuoc co momma mum cacao mwsdm> 0:90 .oaoauu «cu CH mucuoud mo amass: «Lu no woman uwwzm HusuoE m.a m How kuuacwud mmosu mum wmwwcuamumd CH mm=Hm> use A .UHNUQNUQ HQGHGUCfi Gm mm ma Ou ObnfiuwHOH .3Hoo a« mum muuooam HHwumc monouampahnumemucoaaznum unu can mucmaua mo mmuuoudm «:2 .H «Home 14 The fact that the deuterium label in 25 is found exclusively at the methylene position shows that the engmethylene group of the triene does not become equilibrated with the methyls during the pyrolysis. A means by which the gngmethylene could become equilibrated with the methyls is a 1,5-hydrogen shift (gg'to 232). If this processxmxe reversible, or if gga'rearranged to the aromatic product, the deuterium label would not be found exclusively in the methylene position, i.e., label would be found in ring methyls (25a). * * * 1,5 H~ —-—-—-> > 22' 23a 25a Since the deuterium label in 26 is also at the methylene position, no hydrogen shifts can occur prior to coupling. It was of interest to pyrolyze triene containing two deuterated methyls at the C4 carbon. This could distinguish the mechanism involving two successive 1,3-methy1 shifts from that of a simple 1,5-methyl shift. The possibility of deuterating the methyls at the 4-position arises out of the work of Hart and Swatton (21). They found that irradi- ation of bicyclic ketone 21 in acid—free methanol produced enol gg'which could be converted to dienone 22 by treatment with acid. The bicyclic ketone 21 was obtained from dienone 32 (21) by irradiation. Labeled dienone 22 was prepared 15 following their procedure. Dienone 22 was then heated at reflux with CH30D and sodium methoxide to give 31; The process of irradiation and treatment with acid was repeated to give a mixture of dienones 32 and 322, These were then back exchanged with methanol and sodium methoxide to give 32 and 33a. The scheme is shown below. O CH3O 'H hv —-—5> > * ‘X X- * 32’ 27 28 O O + H CH3OD 1 hv > CH3ONa > 2 H > ‘X’ * 'X' * x» 22. 31 O i + CHBOH CH30Na * ‘X- * * * * 32 32a 0 O CHIM X + 3 g > * ‘X' * * 33 33a 34 16 The mixture of dienones 32 and gga'was then converted to triene 34;(§2'actually contained a mixture of material with one and two deuterated methyls at the 4-position but for convenience it is written as above). The nmr spectrum of 34'showed 1.9 protons at T 8.92 for the geminal methyl groups at C4 (Table I). This compares reasonably well with the expected value of 1.5 protons. The exchange procedure was not repeated because sufficient label was present to give meaningful results in the subsequent experiment. Pyrolysis of 34 gave labeled product 35. O A> O * * 34 35 W (‘W The nmr spectrum of 32'(Table I and Figure 8) was different from that of lfi'in that the methyl triplet at T 8.97 inte- grated for 1.0 protons and the singlets at T 7.83 and 7.86 integrated for 13.0 protons. In addition the methylene peak at T 7.38 was broad. The predicted areas under these peaks are 1.0 and 13.0 protons for a 1,5—methyl shift. For consecutive 1,3-methyl shifts, which would give a mixture of 32 and 32“ the predicted areas for these peaks correspond to 2.0 and 12.0 protons respectively. 17 x- x- > > * * * 34 37 35 36 w ’W "W W The results obtained from the pyrolyses of trienes g3 and 34'clearly show that the rearrangement involves a 1,5- methyl migration. In order to determine whether the methyl migration is inter- or intramolecular, triene 38 was syn— thesized and pyrolyzed. An intramolecular rearrangement of 38 would give products 322 and 322“ each of which contains one deuterated methyl group. An intermolecular methyl transfer would give not only 322 and 392/ but also 322 which contains two deuterated methyls and 392 which contains no deuterated methyls. Mass Spectral analysis of the product would then determine which of these mixtures was actually obtained. *- Intra- _ + molecular) 3(- gg 39a 39b *- Inter- §§. molecular > 39a + 39b +fi +i 9(- 39c 39d 18 Triene 38 was prepared from dienone 42 which was in turn prepared from dienone 22 by back exchange with methanol and sodium methoxide. CH3OH CH3M9X *CH3ONa > * X- 29 4O 38 w ’W m Pyrolysis of triene 38 gave a labeled product whose nmr spectrum was in agreement with a 1,5-methyl Shift (see Table I). The mass spectrum of the product gave conclusive evidence for an intermolecular methyl migration. The rela- tive intensity (54%) of the parent peak at m/e 179 (322 and 322) was nearly equal to the sum of the intensities at m/e 176 (30%, 39d) and 182 (22%, 39c). See Figure 3. 4. Mechanism of the Thermal Rearrangement of 1“in the Gas Phase A mechanism consistent with the data presented is shown below. 19 2.2. 1% > CHf‘CHZ + °CH3 2,1. 4.3.. 22. > Q + CH3 18 or 43+17—————>18+g 20 The formation of products (18 and 21) can also take place by radical coupling, i.e., CH3' +-41'——¢ 18” 41'+ 41-—> 21; Ethane would be formed by the coupling of methyl radi— cals, but no attempt was made to isolate it. Although step 6 is not necessary, it is possible that lg'could be formed from this reaction. If the reaction in step 6 is operative 43 is the chain carrier whereas for step 5 a methyl radical iS the chain carrier. The driving force of aromatization in step 5 should be sufficient to make it unnecessary to invoke step 6, especially since the necessary energy for bond cleavage is readily available at the temperature of the pyrolysis. B. Pyrolysis of 1-Methylene—2,3,4,4,5,6-hexamethyl—2,5— cyclohexadiene (11) in Solution The pyrolysis of 11 was studied in solution to determine if the thermal rearrangement would take place at lower tem— peratures, and if so, to determine whether the mechanism was the same, and in particular, to determine whether it was still intermolecular. 1. In Dioxane Pyrolysis of 11.1“ dioxane at 100° under a nitrogen atmosphere produced ethylpentamethylbenzene (18) which was isolated in 12% yield. No other products were isolated. Despite this low yield of isolated product, no significant amount of impurities were observed in the nmr spectrum of 21 the crude product. Deuterium labeled triene gé'was pre- pared from the reaction of dienone 32'with methylmagnesium iodide at low temperature. The nmr spectrum of 44'showed that 0.4 protons remained on the C3-C5 methyls (Table I). Pyrolysis of 42 in dioxane gave labeled product 45. CH3MgI 100° -200 ' Dioxane 7 O «)6 * x» «)6 30 44 45 m I'W The nmr Spectrum of 42'(Table I) differed from that of l§‘in that the singlets at T 7.83 and 7.86 integrated for only 9.0 protons. The low field Singlet at 1 7.83 was now larger than the Singlet at 7.86. This permits the assign— ment of the peak at T 7.83 to the g££h2_methyls and the peak at T 7.86 to the meta and para methyls. 2. In Decalin Since the isolated yield of 18 was low in dioxane, the pyrolysis of lz'in decalin was studied so that higher tem- peratures could be attained. A 1.00 g solution of 11 in decalin was heated at 165° under a nitrogen atmOSphere. Ethylpentamethylbenzene (18) was isolated in 72% yield after workup of the pyrolysis solution. In another experiment, an 0.21 g solution of lz'in decalin was similarly pyrolyzed at 160°. Vpc analysis of the solution after pyrolysis showed that lg had been formed in 90% yield. In addition a trace of decamethylbibenzyl (21) was found. 22 160° Decalin 17, 0.21 g; 18, 90% 2‘1, trace I‘w In still another experiment, an 0.06 g solution of 11 in decalin was pyrolyzed at 165°. The disappearance of 11 was followed by removing aliquots of the pyrolysis solution and noting the decrease in intensity of the 256 nm band of 11.1” the uv Spectrum. A plot of ln (Absorbance) yg time was linear after an initial rapid disappearance of 11, i.e., approximately 50% of 11 reacted in the first 20 minutes. The half-life of lz'was 49 minutes after the initial fast reaction, under these conditions. 3. In Cumene With the hope of trapping radical intermediates, 11 was pyrolyzed in cumene. An 0.08 g solution of $1 in cumene was heated at 150° under a nitrogen atmosphere. Work— up of the pyrolysis solution provided ethylpentamethyl- benzene (18) and a new product which was subsequently identi- fied as 2-pheny1-2-pentamethylbenzylpropane (42). The infrared spectrum of 46 had characteristic bands for a gem: dimethyl group (1365 and 1385 cm-1) and a mono-substituted benzene ring (700 cm_1). The nmr spectrum of 46 had singlets 23 at T 2.98, 7.02, 7.84, 7.92, 8.17, and 8.70 in the ratio 5:2:3:6:6:6, consistent with the assigned structure. These are assigned (low to high field) to the aromatic pro— tons, methylene, pa£§_methyl,(mg£a_methyls, orthg methyls, and geminal methyls. The upfield Shift of the grthg methyls is due to the shielding effect of the other phenyl ring. Compound gg'was synthesized independently in low yield from the cross-coupling reaction of Chloromethylpentamethyl- benzene (22) with phenyldimethylcarbinyl chloride (41). Samples of 46 from the two sources were identical in all respects. CH 150° ' 3 Cumene , CH3 17 18 46 m m m CH3 M ( > ' I C H O + ¢-C-Cl g 2 5 2 > 46 A W CH3 2%. 2.2. In Table II the yields of 18 and 42 are shown for the pyrolyses of different concentrations of lZ'in cumene. The result for a pyrolysis at a lower temperature is also Shown. - , tin-VJIMIN .- flaillbfiflfl p 24 It can be seen that at low concentrations of 11“ more 46' and less 1§'is formed. The temperature effect appears to be negligible, at least for the 15 degree temperature dif- ference which was studied. No hexamethylbenzene was ob- tained in these reactions. A peak with the same retention time as 12 accounted for the remaining starting material in each case. This material was shown to be a complex mix— ture of decomposed 11 by the ir and nmr spectra of a VpC collected sample. The mechanistic conclusions which can be drawn from these results are discussed in Section 8. Table II. Product distribution from the pyrolysis of 11' in cumene Initial Temp.,°C Pyrolysis % Yielda’b Conc° of 11 Time, hr. 18' 46 1.00 §_ 150 24 87 6 0.45 g’ 150 36 78 10 0.09 g. 150 24 54 32 0.09 M. 135 70 58 32 aYields were determined by vpc; conditions are given in the Experimental section. bDecomposed 11 accounted for the remaining material; see the Experimental section. 25 4. In Cumene Containing Thiophenol Although 42 must arise from hydrogen abstraction by radicals formed in the pyrolysis of 11 (see below), all the radicals were not trapped since ethylpentamethylbenzene was still produced in substantial amounts. The pyrolysis of 11 in the presence of thiophenol was studied with the hope of completely trapping radical intermediates. A cumene solution which contained an 8.521 mole ratio of thiophenol to 11 was pyrolyzed at 150°. None of the nor- mal products (58 and 46) were obtained from the pyrolysis. The only products derived from 11 were hexamethylbenzene (48) and phenyl pentamethylbenzyl sulfide (42). Diphenyl disulfide (52) was also isolated. Cumene , Thligggnol + ¢SCH2 + ®SS¢ 17 48 49 50 Hexamethylbenzene was identified by comparison of its nmr spectrum and vpc retention time with those of an authentic sample. Sulfide 42 was identified on the basis of its spec— tral and physical properties and elemental analysis. The nmr spectrum of 42 had a broad peak centered at T 2.84 and singlets at r 5.94, 7.75, and 7.84 in the ratio 5:2;15 (for the last two peaks). These are assigned to the aromatic 26 protons, methylene, and the orthg_methyls and meta and BEES methyls respectively (low to high field). Diphenyl diSul- fide (52) was identified on the basis of its physical prop- erties and its ir Spectrum which was identical to that re- ported in the literature (23). In Table III the mole ratio of the products derived from 11 (48 and 42) in the presence of thiophenol is Shown. The ratio of 48 to 42 remained essentially unchanged with varying initial ratios of thiophenol to 11 at 150° and at 135°. The actual yield of products accounted for only 73% of the Starting material. Since no unreacted lz'was found and no products derived from 12 except 48 and 42 were de- tected, some 11 must have polymerized in the presence of thiophenol. The isolation of products 48 and 42 provides evidence for the radical nature of the pyrolysis of 11; the implica- tions will be discussed more thoroughly below. However, it could be argued that 48 arises from an ionic reaction be— tween 11 and thiophenol. This possibility was checked by examining a solution of 11 and thiophenol in cumene which had been kept at room temperature for three days. No 48' was found. The formation of diphenyl disulfide (52) during the pyrolysis is also consistent with a radical reaction but not with an ionic reaction. 27 Table III. Ratio of products 48 and 49 from the pyrolysis of 0. 09 M 17 in cumene containing thiophenol Mole Ratio of Temp.,°C Pyrolysis Mole Ratioa Thiophenol to 11 Time, hr. of 48 to 42' 1:1 150 36 38:62b 2:1 150 36 38:62 3:1 150 36 37:63 2:1 150 24 30:70C 2:1 135 70 31:69 3:1 135 70 36:64 aRatios were determined by vpc; conditions are given in the Experimental section. bApproximately 3% of 18 and a trace of 46 were also formed. CThe actual yields were 22% of 48 and 51% of 49 as deter- mined by vpc; conditions are given in the Experimental section. 5. Pyrolysis of Labeled Triene 22 in Cumene and in Decalin The results of the pyrolyses of 11 in cumene and in cumene containing thiophenol indicated a radical mechanism for the reaction. In order to determine whether the ethyl— pentamethylbenzene formed in the pyrolysis was produced by an intra- or intermolecular methyl transfer, labeled triene 38 was pyrolyzed in cumene and in decalin at 150°. As pre- viously discussed, an intermolecular methyl transfer would lead to a mixture of product with no deuterated methyls, one deuterated methyl, and two deuterated methyls. 28 Solutions of 0.10 _r_4_ and 0.82 M £38.. in cumene and 0.97 g gg'in decalin were pyrolyzed at 150°. AS expected the nmr spectra of the labeled ethylpentamethylbenzenes which were recovered in each case indicated that a 1,5-methyl migration had occurred (Table I). IN Cumene or * O Decalin > 3‘23 532113 150° + 39c 39d Table IV shows the relative intensities of the major parent peaks at m/e 176, 179, and 182 for each run. Figures 4, 5, and 6 Show partial mass spectra for the labeled ethylpenta- methylbenzenes which were obtained from each run. The re- sults clearly show that the methyl transfer was intermolecular. 29 Table IV. Mass spectra of labeled ethylpentamethylbenzenes from triene 38' Relative Intensity 8f Major Parent Peaks m/e, 176 m/e, 179 m/e, 182 Pyrolysis Method 0.10 g_§§ in cumene, 150° 26 46 19 0.82 g 38 in cumene, 150° 24 43 18 0.97 g_§§'in decalin, 150° 26 45 18 aThe intensities are relative to the base peak at m/e 164 in each case. See also Figures 4, 5, and 6. 6. Attempted Kinetic Study of the Thermal Rearrange- ment of 11 Decalin was chosen as the solvent for the kinetic study of the thermal rearrangement of 11 since the reaction was quite clean and gave a good yield of one product in this solvent. The reaction course could be followed by examining diluted aliquots of the reaction mixture by uv spectroscopy and noting the decrease in intensity of the 256 nm band of 11. The results of several runs under a variety of condi- tions were very erratic (see the Experimental section for the conditions which were used). In some cases straight lines were obtained from a portion of a plot of ln (Absorb- ance) y§_time. In all cases it was found that lz'disappeared rapidly at the beginning of the pyrolysis. The results of two runs are Shown in Figure 1. 30 .8325. assaizea .5 5 a fine no n 85.: who a n6: a... mu: n10 no .8325 385..er 23 5 a t we a 35... find $3.: .2 a «do .9 .3338 3 go oiled-.58 1.55. .5 go 8325. A 25m: 5.35. p b dofil v N ,3 I T o ,3; t c; ‘2 ‘? (”mac-w) at HI! ‘03 u no“ I X 31 7. The Effects of Oxygen and a Free-Radical Source on the Rate of Reaction of lz'in Decalin It is known that triene lz'reacts with oxygen, even at room temperature (Experimental, Part B and reference 22). Although precautions were taken to remove oxygen from the pyrolysis solutions by purging them with nitrogen, it ap— peared that a small amount of oxygen which remained caused 11.t° react rapidly when the solution was heated. In order to test this possibility, three reactions were carried out. A sample of 11 in decalin (0.10 g) was sealed in a tube without taking any precautions to remove air. A set of' solutions of 0.10 §_11 in decalin were purged thoroughly with nitrogen and the samples sealed. An identical set of solutions was degassed to 10—3 mm of Hg and sealed. The samples were then pyrolyzed at 147°. Table V Shows the percent of 11 which had reacted after the samples were heated for the times indicated. The initial disappearance of 11 was rapid in all cases, with 88, 71, and 22% respectively of 11 having reacted after 30 minutes. Longer pyrolysis times for the nitrogen- purged and degassed samples caused only a small additional amount of lz’to react. These results indicate that the amount of oxygen in the system is important. The effect of a radical source on the reaction of 11 in solution was investigated by pyrolyzing 17 in the presence Of' o + 17 41 17 + 2w 'cn. + .13. ——> O (4) 43 'CH3 5.2. (138 35 CH3 CH3 . /’ + ¢-C-H > CH4 + w-c- I CH3 \ CH3 51 CH3 u > (D‘C-CHZ Q I CH3 2,2. CH3 I > ¢-c-CH2 + 'CH3 u CH3 46 .CHa + ¢SH > CH4 + $8 + .ll > ¢SCH2 + °CH3 49 2 <1>S ' > $834) 50 (9) (10) (11) g+¢sn > @ + 23' (13) 2.8. In cumene, hydrogen abstraction from the solvent by methyl radicals (step 7) forms 52 which leads to 25 (steps 8 and 9). At lower concentrations of 11 more hydrogen abstraction occurs and more 22 is produced (Table II). The pentamethylbenzyl radical (22) does not readily abstract a hydrogen atom from cumene since no detectable amount of hexamethylbenzene (22) was produced. In the presence of thiophenol the methyl radicals pro- duced in steps 1 and 1' abstract hydrogen atoms from thio- phenol to produce thiophenoxy radicals (step 10). These add to ll'which leads to phenyl pentamethylbenzyl sulfide (22) (step 11), or they couple to produce diphenyl disul- fide (52) (step 12). Because of the stability of the thio- phenoxy radical its rate of reaction with 11.15 slower than that of a methyl radical and the rate is more nearly the same as the rate of formation of 22 in step 1. Hydrogen atom abstraction from thiophenol by gl'then gives hexamethyl— benzene (22) (step 13). If the rate of attack of thiophenoxy radical on 11 was considerably faster than step 1, triene l1 would be converted entirely, or nearly so, to 22; Subse- quent cleavage of g2'to thiophenoxy and pentamethylbenzyl radicals could then lead to hexamethylbenzene (22). This pathway for the formation of g2'was Shown to be incorrect 37 when it was found than no 22'was formed after pyrolysis of a cumene solution of thiophenol and 22 at 149° for 24 hours. In order to determine the stability of 22 relative to that of 11“ the heats of combustion of the two compounds were calculated. A slightly modified version of Klages' method was used to calculate the heat of combustion of lZ'(25). Eggg_ No. in 17 Kcal/bond . C——H 20 54.0 0 C—C 10 49 .3 >C=CH2 1 115 .5 .11. >c==c< 2 112 .0 Total Correction for quaternary carbon Correction for Six-membered ring Total 1080.0 493.0 115.5 224.0 1912.5 -4.5 +1.0 1908.8 This calculated value for the heat of combustion of £1 does not take into consideration the stabilization due to" conjugation of the double bonds. Hfickel molecular orbital calculations Show that the delocalization energy for a conjugated hexatriene such as cycloheptatriene (Q) is 0.996 and for a cross-conjugated triene (2) it is 0.905 (26). 3::- 1w 38 The stabilization of Q'due to conjugation is 6.7 kcal based on the heat of hydrogenation (reference 25, p. 80). The stabilization in §'(£Z) due to conjugation can therefore be calculated to be 6.0 kcal. The calculated heat of com— bustion of ll’is therefore 1902.8 kcal. The heat of combustion of hexamethylbenzene (22) is 1726.3 kcal (reference 25, p. 98). Addition of the heats of combustion of one C-C bond and two C-H bonds gives a calculated value of 1883.6 kcal for the heat of combustion of‘yi. @ + C—C + ZC— H > @ 2.5}. 132. 1726.3 kcal + 49.3 kcal + 108.0 kcal = 1883.6 kcal From the differences of the calculated values for £1 (1902.8 kcal) and l2'(1883.6 kcal) it is seen that 12'is predicted to be more stable than lz'by 19.2 kcal/mole. Calculation of the heat of combustion of lz'accord- ing to Franklin's method (reference 25, p. 94) gives the results shown below. 39 Group No. in_lZ. Kcal/group -CH3 6 186.41 = 1118.46 0 -C- 1 94.85 = 94.85 >C= CH2 1 273.31 = 273.31 \c-=c/ 2 212 .67 = 425 .34 / \ ——_—— £1 Total 1911.96 Correction for six—membered ring -0.45 Total 1911.51 Consideration of the delocalization energy of 6.0 kcal in lz'gives a predicted heat of combustion of 1905.5 kcal. The heat of combustion of 22 plus that due to one -CH2- group gives a calculated value of 1883.7 kcal for 12; 22. ~ 22. 1726.3 kcal + 157.4 kcal = 1883.7 kcal The results of these calculations predict that 22 is more stable than lz'by 21.8 kcal/mole. The agreement between the two methods is quite good. Some information can also be obtained about the two postulated chain—carrying steps, steps 4 and 5, above. Hfickel molecular orbital calculations indicate that AZ) 22” and l2'have delocalization energies (DE) of 0.90, 1.46, 40 and 2.00B respectively (26). Based on these values, steps 4 and 5 would take place with almost equal facility. > o 2.2, 22. DE = 0-90B DE = 1.466 ADE = 0.566 52. > 6 + .CH3 (5) DE = 1.466 22, DE2= 2.006 ADE = 0.546 These differences in DE oversimplify the question of whether the reaction in step 4 or that in step 5 is thermo- dynamically more favorable. Common values of B vary from 16 to 20 kcal (reference 26, p. 242). These values should only be used for benzenoid aromatic hydrocarbons, however, since the average value of B obtained from conjugated ole- fins was only 6 kcal (reference 26, p. 242). In part, the difference is due to the alternation of bond distances that occur in polyenes, but for the most part the discrepancy is undoubtedly due to the effects of electron correlation; the conjugation 7- energy between adjacent double bonds in a polyene is less than that given by HMO theory... (reference 26, p. 242). If B is taken as 6 kcal for 11 and 22 and as 16 kcal for £2” step 4 is thermodynamically favored in the direction Shown by approximately 3 kcal and step 5 is favored by 41 approximately 23 kcal in the direction shown. The rate limiting step for the formation of 12'18 thus predicted to be step 4, the first of the two chain-carrying steps. EXPERIMENTAL A. General Procedures Nmr spectra were obtained on a Varian A—60 or JEOLCO C-60H Spectrometer in CC14 solution (unless otherwise stated) with tetramethylsilane as an internal reference, given the value 1 10.00. Infrared spectra were taken with a Unicam SP-200 or a Perkin-Elmer 237-B spectrometer in CCl4 solu- tion (unless otherwise stated). Ultraviolet spectra were taken on a Unicam SP—800 Spectrometer. Mass spectra were determined by Michael Petschel with a Consolidated Electro- dynamics Corporation 21-103C instrument. VarianeAerograph gas chromatographs were used. Melting points were deter- mined with a Gallenkamp Melting Point Apparatus and are uncorrected. Analyses were performed by Spang Microanaly- tical Laboratories, Ann Arbor, Michigan. Pre-purified nitrogen was further purified by passing it through a sequence of traps containing Fieser's solution, saturated lead acetate, and concentrated H2804, and it finally was passed through a tower containing Drierite and sodium hydroxide pellets. 42 43 B. Preparation of 1—Methylene-2,3,4,4,5,6-hexamethyl-2,5- cyclohexadiene (17) 1. From Hexamethyl-Z,4-cyclohexadienone (55) Dienone 55'was prepared according to the method of Gray and Hart (27). A solution of 10.0 g of 55'in 75 ml of ether was added slowly to a magnetically stirred solution of 26 ml of 3glmethylmagnesium bromide which was diluted with 150 ml of ether. When the addition was complete the reac- tion mixture was stirred and heated at reflux for 2 hr, then coOled in an ice bath and hydrolyzed with 10% NH4C1 solution. The organic layer was washed with water, dried over M9804, and the ether was evaporated to give a yellow oil. Recrys- tallization of the oil from 95% ethanol gave 6.7 g of white crystalline 12’, mp 41-30 (.lit. mp (28) 36-440). The infra- red spectrum was identical to that reported for lz'(28); the nmr spectrum Showed the peaks due to 11 (see Table I), but also showed a small peak at T 7.85 due to hexamethyl— benzene (22) impurity. In all preparations of 11 from 55 a small amount of g2'was found. It was found that this could be avoided by using the following method. 2. From Hexamethyl-2,5-cyclohexadienone (24) Dienone gg'was prepared according to the method of Hart and Swatton (21). The reaction of dienone 22 with methyl- magnesium bromide or iodide, carried out as above, gave 11) mp 43.5-450. Typical yields were 70-80% of 12 based on 22; 44 The infrared and nmr spectra were identical to those of $1 in the literature (28), and no hexamethylbenzene impurity was evident. The nmr spectrum of 11 is reported in Table I. As previously noted by Doering and co-workers (22), 11’ is sensitive to air. .It was found that lz'could be stored for extended periods of time (3 to 4 months) by suspending it in 95% ethanol and storing it in a refrigerator. C. Preparation of Ethylpentamethylbenzene (12) from.Penta- methylbenzene (12) Anhydrous aluminum chloride, 0.36 g, was added to a solution of 2.0 g of l2'in 15 ml of ethyl bromide. The re- action mixture was stirred and heated at reflux overnight, cooled, and diluted with 100 ml of ether. The solution was washed with 25 ml of 10% HCl and then with water. The organic layer was dried over MgSO4 and the ether and ethyl bromide were evaporated to give a white solid which was analyzed by Vpc (5' x 1/4" SE-30 column at 160°, He flow rate of 82 ml/min). Two peaks with approximately equal areas with retention times of 5.1 and 13.0 min were found. The first peak was Shown to be due to 12 by comparison of its retention time and the ir spectrum of a collected sample with those of authentic 12; A total of 137 mg of the com- pound with a retention time of 13.0 min was collected by preparative vpc (same conditions as above). «The collected material was then sublimed at 80° and atmospheric pressure to give a white solid, mp 59-61° in a sealed tube. Clement 45 reported a melting point of 1250 for 18 (18). More re- cently however, Berezina and coworkers reported a melting point of 61.5-63o for 1§'(19). The nmr spectrum of the col- lected material in CC14 had a quartet at 1 7.38 (2H) coupled to a triplet at T 8.97 (3H), J = 7.4 cps, and singlets at T 7.83 and 7.86 (15H total). The reported (19) nmr spectrum of l§,in C014 had a quartet at T 7.33 (2H) coupled to a triplet at T 8.94 (3H), J = 7.5 cps, and singlets at T 7.79 (6H) and 7.81 (9H). The infrared spectrum of 18 had major peaks at 2980 and 1460 cm-1. The nmr spectrum of $8 is also reported in Figure 7. D. Preparation of Decamethylbibenzyl (2;) from Chloromethyl- pentamethylbenzene (22) The method of Humphlett and Hauser was used to prepare 21'(20). ~Accordingly 5.0 g of 22 (prepared by R. W. Fish in this laboratory (29)) in 30 ml of ether was added to 0.7 g of magnesium turnings in 5 m1 of ether. The reaction mixture was stirred and heated at reflux for 20 hr. The resultant white slurry was diluted with 100 ml of ether and poured onto dry ice. The ethereal slurry was washed with 50 ml of 10% HCl. The ether was evaporated from the organic layer and the crude product was recrystallized from dioxane to give 3.36 g of pale yellow crystals. A second recrystal- lization from dioxane gave pure 2}, mp 240-2o (lit. mp (20) 241-20). The infrared spectrum of 2}'(CHC13) was nearly -1 featureless, showing only broad bands at 2925 and 1465 cm 46 and a medium band at 1390 cm-1. The nmr spectrum (CDCla) had singlets at T 7.10 (4H, methylene protons) and at T 7.58 and 7.72 (30H total, QEEEQ methyls and meta and p§£§ methyls respectively). ‘E. Preparation of 2-Phenyl-2—pentamethylbenzylpropane (46) from Phenyldimethylcarbinyl Chloride (4;) and Chloro- methylpentamethylbenzene (222 A cross-coupling reaction between 41 and gz'was used to prepare 46. Phenyldimethylcarbinyl chloride (41) was pre— pared as follows: From 10 ml of 3g_phenylmagnesium bromide and 2.05 g of acetone, followed by NH4C1 workup, there was obtained 3.42 g of crude phenyldimethylcarbinol (broad hydroxyl band at 3400 cm.1 in the ir spectrum, neat, and aromatic protons at T 2.63-3.05, hydroxyl proton at T 6.10, and a singlet for the methyls at T 8.60 in the nmr spectrum). Hydrogen chloride was bubbled into the crude carbinol until a second layer formed, and the flow of hydrogen chloride was then continued for a few more minutes. The lower, aqueous layer was removed with a capillary pipette and anhydrous CaC12 was added to the organic layer. The flask which con- tained the organic layer was then evacuated for a short time to remove dissolved hydrogen chldride. The organic chloride was distilled at reduced pressure (41-40 at 0.3-0.4 mm of Hg) to give 2.31 g of 31" (lit bp (30) 56-80 at 1.5 mm of Hg). The nmr spectrum of 41,had aromatic protons at T 2.40—3.00 and a singlet for the methyls at T 8.14. 47 Equimolar amounts of 41 (2.18 g) and 22 (2.76 g) were treated with 0.75 g of magnesium turnings as in the prepara- tion of 21x above. The yellow solid which was obtained after the ether was evaporated was taken up in CCl4 and dried over M9804. After some of the CCl4 was evaporated and the solution was cooled, a white solid precipitated. It was shown to be 24 by its melting point and nmr spectrum which were identical to those of 24 prepared above. The CC14 solution was examined by vpc (5' x 1/4" SE-30 column at 245°, He flow rate of 100 ml/min). Three peaks with re- tention times of 1.0, 2.6, and 7.8 min were found. The first peak was shown to be due to hexamethylbenzene (48) by com- parison of its vpc retention time and the nmr spectrum of a collected sample with those of authentic 48; The peak with a retention time of 2.6 min was shown to be due to dicumyl (the coupling product of 41) by comparison of its vpc reten- tion time and the nmr spectrum of a collected sample with those of authentic dicumyl. The nmr spectrum of a collected sample of the material with a retention time of 7.8 min in- dicated that it was 46; The mixture of products was chromato- graphed on 25 g of silica gel using hexane as the eluant. This provided fractions which contained mostly 46. These fractions were combined and 46 was further purified by pre- parative vpc (1.5' x 1/4" SE-30 column at 245°, He flow rate of 80 ml/min) and sublimation at 90° and water aspirator vacuum. -Approximately 80 mg of pure 46, mp 82-4°, was 48 obtained. The mass spectrum of 46 had a parent peak at m/e 280, consistent with the molecular formula, C21H23. Anal. Calcd for C21H23: C, 89.92; H, 10.08. Found: C, 89.83; H, 10.15. The infrared spectrum of 42 had bands at 1365 and 1385 cm.1 (ggmfdimethyl) and at 700 cm-1 (mono-substituted benzene). The nmr spectrum of 46 consisted of singlets at T 2.98 (5H, aromatic protons), 7.02 (2H, methylene), 7.84 (3H, EEEE methyl), 7.92 (6H, meta methyls), 8.17 (6H, ortho methyls), and 8.70 (6H, geminal methyls). F. Pyrolysis of 1-Methylene—2,3,4,4,5,6-hexamethyl-2,5- cyclohexadiene (41) on a Vapor Phase Chromatograph The ether was evaporated from a solution of 0.26 g of 41 (prepared by method B-1) in 10 ml of ether until the volume was approximately one ml. Samples of this solution were injected into a vpc instrument under the following conditions: 10' x 1/4" Apiezon-L column at 175°, He flow rate of 100 ml/min, injector temp. 325°. Three peaks with retention times of 5.7, 11.2, and 12.3 min were observed. The first peak was shown to be due to 41 by comparison of the nmr spectrum of a collected sample with that of authen— tic £1. The peak with a retention time of 11.2 min was shown to be due to hexamethylbenzene (48) by comparison of the nmr spectrum of a collected sample with that of authentic 2.8.. The infrared and nmr spectra of a collected sample of the compound with a retention time of 12.3 min were identical 49 to those of ethylpentamethylbenzene (48) prepared inde- pendently (part C of the Experimental). With the vpc con- ditions above, 43% of 11 and 57% of 48 were found. Hexa- methylbenzene (48) was present in the ether solution and was not formed by pyrolysis of 41 (see below). The ether was evaporated from a solution of 0.13 g of 41 in 5 ml of ether which had been kept at room temperature for three days and the residue was examined by nmr. Only the peaks for 41 and a small peak for 48 (see part B of the Experimental) were present. G. Gas Phase Pyrolysis of l—Methylene-Z,3,4,4,5,6-hexa- methyl-2,5-cyclohexadiene (£1) 1. General The center 10-inch length of a 0.5" x 24“ Pyrex tube was filled with 3 mm Pyrex beads and this portion of the tube was placed in a Sargent tube furnace. A pressure equi- librating dropping funnel with a nitrogen inlet at the top was fitted to the tOp of the pyrolysis tube. A receiver containing glass wool and open to the air through a drying tube which contained Drierite was attached to the bottom of the tube. The receiver was cooled in a dry ice—acetone bath. A nitrogen flow of 60 ml/min was maintained throughout the pyrolysis. This corresponded to a contact time of about 6 seconds in the heated portion of the tube. The temperature of the furnace was maintained at 420°. The sample was 50 introduced into the heated tube dropwise by dissolving it in a small volume of benzene or by first melting it, in which case a small amount of benzene was rinsed through after all the triene was added. After the reaction was completed the cooled tube was rinsed with chloroform and the rinsings were combined with the material in the receiver. 2. Pyrolysis of 41' A solution of 4.66 g of 41 in 5 ml of benzene was pyro- lyzed as above. A total of 4.37 g of crude product was collected (94% recovery of material). Sublimation of the crude product at 80° and atmospheric pressure afforded 2.27 g of ethylpentamethylbenzene (48) (52%, based on recovered material). Ethylpentamethylbenzene (48) was further puri- fied by column chromatography on silica gel with hexane as eluant and by another sublimation to give 1.57 g of pure product, mp 63-5° in a sealed tube. The mass spectrum of 18'had a parent peak at m/e 176 (54% of the base peak at m/e 161), consistent with the molecular formula, C13H20. A partial mass spectrum is shown in Figure 2. Anal. Calcd for C13H20: C, 88.56; H, 11.44. Found: C, 88.46; H, 11.47. The ir and nmr spectra were identical to those of 48’ prepared from pentamethylbenzene. The material which failed to sublime at 80° was analyzed -by vpc (1.5' x 1/4" SE-30 column at 280°, He flow rate of 85 ml/min). Only one peak, with a retention time of 2.4 min, 51 was observed. A portion of this material was recrystal- lized three times from dioxane to give pure decamethyl- bibenzyl (84), mp 238-42°. The ir and nmr spectra were identical to those of authentic 84; H. Gas Phase Pyrolysis of 1-Dideuteromethylene-2,3,4,4,5,6- hexamethyl-Z,5-cyclohexadiene (23) 45a— 1. Preparation of 88 Trideuteromethylmagnesium iodide was prepared from 4.00 g of trideuteromethyl iodide and 0.68 g of magnesium turnings in 75 ml of ether. A solution of 4.66 g of hexa- methyl-Z,5-cyclohexadienone (84) in 85 ml of ether was added to the solution of Grignard reagent and the reaction was carried out as in part B, above. After recrystallization 2.61 g of product was obtained, mp 44.5-460. The ir and nmr spectra of the product indicated the presence of some remaining dienone. The nmr spectrum also showed 0.7 protons for the methylene group at T 5.26 (see Table I). 2. Pyrolysis of 88 A 1.00 9 sample of 28'(which included some dienone im- purity) was pyrolyzed as in G, above, to give 0.79 g of crude product. The crude product was sublimed at 80° and atmospheric pressure to give 0.49 g of labeled ethylpenta— methylbenzene (82) which was contaminated with dienone 84' as shown by the nmr spectrum and vpc analysis (5' x 1/4" SE-30 column at 155°, He flow rate of 40 ml/min; the dienone 52 had a retention time of 19.2 min and 2§.a retention time of 22.0 min). The mixture of 84 and 88 was chromatographed on silica gel with hexane as eluant to give 88 which was sublimed at 80° and atmospheric pressure to give 0.18 g of pure product, mp 63-4° in a sealed tube. The nmr spectrum of 88 is reported in Table I. The labeled decamethylbibenzyl (88) was purified by preparative vpc (the conditions in G—2, above, for unlabeled decamethylbibenzyl were used). The'nmr spectrum (CDClg) of 88 had the expected peaks at T 7.10, 7.58, and 7.72 in a ratio of 1.4 protons for the methylenes to 30.0 protons for the ring methyls. I. Gas Phase Pyrolysis of 1-Methylene-4,4-bis(trideutero- methyl)—2,3,5,6-tetramethyl—2,5—cyclohexadiene (34) 1. Preparation of 84 3,S-Eigjtrideuteromethyl)-2,4,4,6-tetramethyl-2,5- cyclohexadienone (82) was prepared according to the method of Hart and Swatton (21). A solution of 10.6 g of 82'in 350 ml of methanol which was freshly distilled from sodium methoxide was cooled in an ice bath and irradiated with a 450 watt Hanovia lamp through a Pyrex filter. After 4.5 hr all of the dienone had reacted as shown by the absence of the 246 nm band of 82'in the uv spectrum. Two drops of cone HCl was added to the photolysis solution and the solution was stirred at room temperature for 30 min. The methanol 53 was then evaporated on a rotary evaporator. The crude product was dissolved in 200 ml of CH2Clzand was washed with 40 ml of ice-cold 5% NaHCOa. The organic layer was dried over M9804 and the CHZClz was evaporated on a rotary evaporator. The recovered material was recrystallized from hexane to provide 4.48 g of 3,4v§i§(trideuteromethyl)- 2,4,5,6—tetramethyl-2,5—cyclohexadienone (88). Dienone 88 was then refluxed with CH3OD and sodium methoxide until the C3- and C5-methyls were completely deuterated, as in the preparation of 88 (21), to give 3,4,543£i§(tri— deuteromethyl)—2,4,6—trimethyl—2,5-cyclohexadienone (84). The nmr spectrum of 84 showed a singlet at T 8.17 (6.0H) for the C2- and 06—methyls and a singlet at T 8.79 for the geminal methyls which integrated for 3.7 protons. The peak for the C3— and C5-methyls at T 8.05 was absent. Dienone 88’was irradiated in acid-free methanol, the pho- tolysis solution treated with acid, and the reaction mix- ture worked up as above to give 1.38 g of product. This dienone (88'and 888'in the text) was then back-exchanged with methanol and sodium methoxide to give, after workup and recrystallization, 1.18 g of dienone 88, 4,4figi§(tri- deuteromethyl)-2,3,5,6-tetramethyl-2,5—cyclohexadienone. The nmr spectrum of 88 showed peaks at T 8.05, 8.17, and 8.79 in the ratio 6.0:6.0:1.8 for the C3-C5, C2-C6, and geminal methyls respectively. Triene 84 was prepared from 88'and methylmagnesium bromide as in B—2, above. 54 The nmr spectrum of 84 showed 1.9 protons at T 8.92 for the geminal methyls (see Table I). 2. PyrolySis of 334‘ An 0.73 9 sample of 84'was pyrolyzed as above and 0.59 g of crude product was obtained. Sublimation of the crude product at 80° and atmospheric pressure gave 0.31 g of labeled ethylpentamethylbenzene 88) mp 63-4° in a sealed tube. The nmr spectrum of 88'is reported in Table I and Figure 8. The material which failed to sublime was not in- vestigated. J. Gas Phase Pyrolysis of 1-Methylene-4-trideuteromethyl- 2,3,4,5,6-pentamethyl-2,5-cyclohexadiene (38) 1. Preparation of 38. 3,4-§i§(trideuteromethyl)—2,4,5,6-tetramethyl-2,5-cyclo- hexadienone (88) was prepared from dienone 82'as in I—1, above. Dienone 88'was then back-exchanged with methanol and sodium methoxide to give 4-trideuteromethyl—2,3,4,5,6- pentamethyl-2,5-cyclohexadienone (48). The reaction of 42' with methylmagnesium bromide, as in B-2, above, gave 88) mp 43.5-45°. The nmr spectrum of 88'showed 3.3 protons at T 8.92 for the geminal methyls (see Table I). 55 2. Pyrolysis of 88' A 1.00 g sample of 88 was pyrolyzed as above to give 0.89 g of crude product. The crude product was sublimed at 80° and atmospheric pressure to give 0.49 g of labeled ethylpentamethylbenzene (88) which was chromatographed on silica gel with hexane as eluant and then sublimed again to give 0.28 g of pure 88) mp 60.5—62° in a sealed tube. The nmr Spectrum of 88 is reported in Table I. The mass spec- trum of 88 had major parent peaks at m/e 176, 179, and 182 with intensities 30, 54, and 22% respectively of the base peak at m/e 164. A partiaI"mass Spectrum is shown in Figure 2. The unsublimed material was not investigated. K. Pyrolysis of 1-Methylene-2,3,4,4,5,6-hexamethyl—2,5- cyclohexadiene (17) in Dioxane Triene 41) 1.30 g, was dissolved in 60 ml of dioxane in a three-necked pear—shaped flask equipped with a nitrogen inlet tube, a condenser with a drying tube attached, and a septum cap. Nitrogen was bubbled through the solution for 30 min and then the solution was heated at reflux (100°) with nitrogen bubbling through the solution continuously. The reaction course was followed by removing aliquots of the reaction mixture and noting the decrease in intensity of the 256 nm band of 12.1“ the uv spectrum. After five hours less than 10% of lz'remained and the dioxane was evaporated. The resultant orange oil was triturated with 4 ml of methanol to give a white crystalline solid which was sublimed at 80° 56 and atmospheric pressure. This gave 0.15 g of 88) with ir and nmr spectra identical to those of synthesized £8. The uv spectrum of 1.5.3. had xiii“ 270 m (e 354) (lit. (19) Alsooctane 270 nm (e 200)) max ° In a separate experiment which was carried out as above, the nmr spectrum of the crude reaction mixture showed that the only materials present were ll_(approximately 30%) and 48'(approximately 70%) with small amounts of impurities. L. Pyrolysis of 1-Methylene-3,5-bis(trideuteromethyl)- 2,4,4,6-tetramethyl-2,5-cyclohexadiene (44) in Dioxane 1. Preparation of 44' 3,S—Qigjtrideuteromethyl)-2,4,4,6-tetramethyl-2,5- cyclohexadienone (82) was prepared as previously described (21). -Nine ml of 3g methylmagnesium bromide was added to a solution of 1.50 g of 82'in 100 ml of ether which was cooled in a dry ice-CCl4 bath. The mixture was stirred for 7 hr at -15 to -20°. The resultant white slurry was hydrolyzed with 30 ml of 10% NH4Cl while still cold, and the aqueous layer was rapidly separated from the organic layer. The organic layer was washed twice with saturated NaCl solution, dried over M9804, and the ether was evaporated. The nmr spectrum of 44 is reported in Table I. _. I. .5 .I.~.|.In1h.u.o'|:.h.rvl!‘lu - _ 57 2. Pyrolysis of 44' The pyrolysis of 1.30 g of 44 in 75 ml of dioxane was carried out as described for the unlabeled triene. Tritura- tion of the crude product with methanol failed to give a solid, but sublimation at 80° and atmospheric pressure pro- vided 0.15 g of labeled ethylpentamethylbenzene (48). The nmr spectrum of 48 is reported in Table I. M. Pyrolysis of 1-Methylene-2,3,4,4,5,6-hexamethyl-2,5- cyclohexadiene (11) in Decalin An 0.35 9 sample of 41 in 2.0 ml of decalin (1.00 M), in a 13 x 100 mm Pyrex test tube which had been drawn out for sealing, was purged with nitrogen, frozen in liquid nitrogen, and sealed. The sample was heated in an oil bath at 165° for 3.5 hr. The tube was cooled and opened and the decalin was evaporated at room temperature with a water as- pirator. The crude product was chromatographed on silica gel with hexane as the eluant to give 0.25 g of 48; The nmr spectrum of 48 was identical to that of synthesized 48; Vpc analysis (5' x 1/4" SE-30 column at 205°, He flow rate of 85 ml/min) showed one peak with a retention time of 3.4 min, identical to that of authentic 48; An 0.21 g_solution of 41 in decalin was prepared as above and was heated at 160° for 24 hr. Vpc analysis (5' x 1/8" SE-30 column programmed from 145° to 260° at 12°/min, He flow rate of 60 ml/min) of the pyrolysis solution 58 showed a large peak with retention time of 3.0 min and a very small peak with a retention time of 13.6 min. Authen- tic 48'had a retention time of 3.0 min and decamethylbi- benzyl (84) a retention time of 13.6 min. The yield of 18 was shown to be 90% by comparison of the peak area of £8 for the pyrolysis solution with the peak area of 48 for a 0.21 32 solution of 48 in decalin. An 0.06 5 solution of 41 in decalin in a three—necked pear-shaped flask equipped with a nitrogen inlet tube, thermometer, and a reflux condenser with an attached drying tube was heated in an oil bath. Nitrogen was passed through the solution continuously during the pyrolysis and the temperature of the solution was kept constant at 165°. The reaction course was followed by removing 3 ul aliquots from the reaction solution and diluting them to 3 ml in cyclohexane and recording the uv spectrum. The intensity of the 256 nm band of 41 decreased as the reaction progressed. Only a small amount of 81 remained after a 3 hr pyrolysis time. A plot of 1n (absorbance) ys time gave a straight line after an initial rapid disappearance of 50% of 42; The half-life of 41 was determined to be 49 min and the rate — —1 constant for the disappearance of 41 was 2.4 x 10 4 sec . 59 N. Pyrolysis of 1-Methylene-2,3,4,4,5,6-hexamethyl-2,5- cyclohexadiene (4;) in Cumene 1. Product Identification A 1.00 9 sample of 41 in 75 ml of cumene (0.08 g) was heated at reflux for 24 hr. Nitrogen was passed through the solution continuously during the pyrolysis. After the pyrolysis most of the cumene was distilled and the remaining cumene was evaporated at room temperature with a water aspir- ator. The solid material which remained was examined by vpc (5' x 1/4" SE-30 column at 245°, He flow rate of 100 ml/min). Two major peaks with retention times of 1.0 and 7.6 min and two very minor peaks with retention times of 1.7 and 2.6 min were found. The retention time of authentic 48 was 1.0 min. Sublimation of the crude product at 70° and water aspirator vacuum provided 48 which after recrys- tallization from ethanol had a mp 62.5-64° and ir and nmr spectra identical to those of synthesized £8. The unsub— limed material was recrystallized twice from ethanol to give pure 2-phenyl-2—pentamethylbenzylpropane (48), mp 83-40. The ir and nmr spectra were identical to those of 48_syn- thesized from phenyldimethylcarbinyl chloride and chloro— methylpentamethylbenzene in E, above. 2. Concentration and Temperature Effect Solutions of varying mole ratios of 48 and 48 were ex- amined by vpc (5' x 1/8" SE-30 column programmed from 170 60 to 225° at 6°/min, He flow rate of 60 ml/min). From the peak areas of 48’and 48’a plot was made of mole percent 88’y§_the percent of the peak areas due to 88; Cumene solutions which contained 1.00 g, 0.45 g and 0.09 gllz'were prepared and examined by vpc under the fol- lowing conditions: 5' x 1/4" XF-1150 column at 145°, He flow rate of 60 ml/min. The area of the peak due to 48 was then determined for each concentration. In both cases the areas of the peaks were determined with a K and E COmpen- sating Polar Planimeter. Solutions of 1.00 g, 0.45 g_and 0.09 g 41 in cumene were prepared. Samples of each of these solutions were placed in 13 x 100 mm Pyrex test tubes which had been drawn out for sealing. These were purged with nitrogen, frozen in liquid nitrogen, and sealed. Samples of 1.00 g and 0.09 5:41 were heated in an oil bath at 150° for 24 hr and a sample of 0.09 g 12 was heated in an oil bath at 135° for 70 hr. A 0.45 5:11 sample was heated at 150° for 36 hr. Each of the samples was then examined by vpc using the two sets of conditions given above. The yields of 48’and 48' obtained from these data are shown in Table II. No deca- methylbibenzyl (84) was detected in any of these reactions. In addition to the peaks for 48 and 48) a peak with the same retention time as 41 was found in all cases. In an- other experiment in which this material was collected by preparative vpc, the nmr spectrum showed broad peaks from T 7.5-8.0 and from T 8.4—9.2 and the ir spectrum showed several peaks in the carbonyl region (1640-1750 cm-l). 60/ O. Pyrolysis of 1-Methylene-2,3,4,4,5,6-hexamethy1—2,5— cyclohexadiene (17) in Cumene Containing Thiophenol 1. Product Identification A 1.00 9 sample of 11,1“ 70 ml of cumene and 5 m1 of thiophenol (8.5:1 mole ratio of thiophenol to 11) was heated at reflux for 24 hr. Nitrogen was passed through the solu- tion during the pyrolysis. After the pyrolysis most of the cumene was distilled from the pyrolysis solution. -After the remaining solution had cooled to room temperature, a white solid precipitated and was collected (0.39 g). A sodium fusion test on this material indicated the presence of sulfur (31). ‘Recrystallization of this material from 4:1 methanol-benzene gave white crystalline product, mp 140-2°. This was identified as phenyl pentamethylbenzyl sulfide (48). Anal. Calcd for C18H22S: C, 79.93; H. 8.22; S, 11.85. Found: C, 79.81; H, 8.13; S, 11.81. The ir spectrum of 48 had strong bands at 2920, 1485, and 690 cm-1. The nmr spectrum had a broad peak centered at T 2.84 (5H, aromatkzprotons) and singlets at T 5.94 (2H, methylene), and at 7.75 and 7.84 (15H, grthg methyls and mggg and p§£§_methyls respectively). The remaining solution was analyzed by vpc (5' x 1/4" SE—30 column at 245°, He flow rate of 100 ml/min). Three peaks (in addition to cumene and thiophenol) with retention times of 0.9, 3.0, and 8.6 min were found. 61 Hexamethylbenzene (48) had a retention time of 0.9 min and 48 had a retention time of 8.6 min under these conditions. The cumene and thiophenol were evaporated and the solid which reamined was sublimed at 90° and atmospheric pressure to give 0.36 g of 48 which was identified by comparison of its vpc retention time and nmr spectrum with those of authentic 48; The solid material which failed to sublime at 90° was sublimed at 100° and 2 mm of Hg to give 0.45 g of material which was shown by vpc (same conditions as above) to be a mixture of 48’and the unidentified product with a retention time of 3.0 min. Two recrystallizations of this mixture from 4:1 methanol—benzene provided 0.20 g of 48; The re— maining solid was recrystallized from ethanol to remove re- maining traces of 48 and water was added to the filtrate to crystallize out the remaining compound. This procedure was repeated two more times to give a white crystalline product which was identified as diphenyl disulfide (88), mp 55-6° (lit. mp (32) 61°). The mass spectrum of 88 showed a par- ent peak at m/e 218, 96% of the base peak at m/e 109, con- sistent with the molecular formula C12H1052- The nmr spec— trum of 88 had only a broad peak from T 2.50-3.10. In ad- dition the ir spectrum (nujol) was identical to that of 88 as reported in the literature (23). 2. Effect of Temperature and Thiophenol Concentration Solutions of varying mole ratios of 48 and 48’were ex— amined by Vpc (5' x 1/8" SE-30 column programmed from 160 to 245° at 8°/min, He flow of 60 ml/min) as in N—2, above. 62 From the peak areas for 48 and 48'a plot was made of mole percent 48'ys_the percent of the peak areas due to 48, Three samples of 0.09 g lz'in cumene which contained 1:1, 2:1, and 3:1 mole ratios of thiophenol to 41,were pre- pared for pyrolysis as in N—2, above. These samples were then heated at 150° for 36 hr. Similarly prepared samples 'which contained 2:1 and 3:1 mole ratios of thiophenol to 41 were heated at 135° for 70 hr. These samples were ana- lyzed using the vpc conditions above and the mole ratio of 48’to 48'was obtained. The results are reported in Table III. No peaks corresponding to 41 or 84 were found in any of these cases. An 0.09 g cumene solution of 41 which contained a 2:1 mole ratio of thiophenol to 41 was prepared as above and heated at 150° for 24 hr. The sample was analyzed by the vpc conditions above and the mole ratio of 48 to 48' was found to be 30:70. A 0.09 g solution of 48 in cumene was prepared and the peak area due to 48 was determined by vpc analysis under the following conditions: 5' x 1/4" XF- 1150 column at 145°, He flow rate of 60 ml/min. The pyrol— ysis solution was also analyzed under these conditions and the yield of 48'was found to be 22%. The yield of 48'can therefore be calculated to be 51%. A sample of 0.09 8:41 in cumene which contained a 2:1 mole ratio of thiophenol to lziwas prepared as above and kept at room temperature for three days. Vpc analysis (5' x 1/8" SE-30 column at 150°, He flow rate of 65 ml/min) 63 of this sample showed that no 48'had been formed (the re- tention time of 41 was 2.4 min and that of 48 3.8 min under these conditions). P. Pyrolysis of Phenyl Pentamethylbenzyl Sulfide (48) in Cumene Containing Thiophenol A cumene solution which contained 0.05 g 48'and 0.15 g thiophenol was prepared for pyrolysis as in the pyrolysis of 11,1“ cumene (N-2, above). The sample was heated in an oil bath at 149° for 24 hr and then examined by vpc (5' x 1/4" Carbowax column at 198°, He flow rate of 150 ml/min). A 0.10 M_solution of 48’in cumene was examined under the same conditions and the retention time of 48 was 3.4 min. No peak at a retention time of 3.4 min was found in the pyrolysis solution. Q. 'Pyrolysis of 1-Methylene-4—trideuteromethyl-2,3,4,5,6- pentamethyl-2,5-cyclohexadiene (88) in Cumene and Decalin Triene 88 was prepared as in J-1, above. The nmr spec- trum of 88'is reported in Table I. 1. Pyrolysis of 0.10 g 88'in Cumene To 27 ml of cumene heated at reflux (150°) and through which nitrogen was passed continuously there was added 0.53 g of 88 in 3 ml of cumene. The solution was heated for 14 hr after which most of the cumene was distilled. The remaining cumene was evaporated at room temperature with a 64 ‘water aspirator, leaving 0.56 g of crude product. The crude product was sublimed at 75° and water aspirator vacuum to give labeled ethylpentamethylbenzene (88) which ‘was further purified by column chromatography on silica gel with hexane as eluant. The nmr Spectrum of 88 is re- ported in Table I. A partial mass spectrum is Shown in Figure 4 and Table IV shows the relative intensities of the major parent peaks. 2. Pyrolysis of 0.82 g 88 in Cumene J— A solution of 0.43 g of 88 in 3 ml of cumene was pre- pared for pyrolysis as in N-2, above. The sample was then heated at 150° for 24 hr. The cumene was then evaporated at room temperature with a water aspirator to give 0.48 g of crude product. Sublimation of the crude product at 75° and water aspirator vacuum gave 0.31 g of 88 which was chromatographed on silica gel with hexane as eluant. The nmr spectrum of 88,15 reported in Table I. A partial mass spectrum is Shown in Figure 5 and Table IV Shows the rela— tive intensities of the major parent peaks. 3. Pyrolysis of 0.97 g 88'in Decalin An 0.51 9 sample of 88 in 3 ml of decalin was prepared for pyrolysis as in N-2, above, and was heated at 150° for 14 hr. The decalin was then evaporated at room temperature with a water aspirator to give 0.57 g of crude product. The crude product was chromatographed on Silica gel with hexane 65 as eluant to give 0.40 g of 88 which was then sublimed at 75° and water aspirator vacuum. The nmr spectrum of 88 is reported in Table I. A partial mass spectrum is Shown in Figure 6 and Table IV shows the relative intensities of the major parent peaks. R. Attempted Kinetic Study of the Thermal Rearrangement of 41 in Decalin The disappearance of 48 was monitored by uv Spectros- copy. Two or three ul aliquots of the reaction solution ‘were diluted to 3 ml with cyclohexane in a quartz uv cell and the uv spectrum was recorded. The intensity of the 256 nm band of 41 decreased as the reaction progressed. Method 1. Solutions of 41 in decalin (0.06 to 0.15 All) were placed in a 3-necked pear-shaped flask equipped with a thermometer, reflux condenser with a drying tube, nitrogen inlet, and a septum cap. The system was flushed thoroughly with nitrogen and then placed in a constant temperature bath. The nitrogen flow was continued during the heating period. Aliquots were removed periodically and the uv Spectrum was recorded, as above. A plot of ln (absorbance, 256 nm) y§_time was made from this data. The result of one such run at 145.3 i 0.2° is shown in Figure 1. In other runs no part of the plot was linear. Method 2. This method involved nitrogen purging of solutions of 4Z'in a test tube stoppered with a septum cap. 66 The sample was then heated in a constant temperature bath and aliquots of the reaction mixture were removed periodically and examined by uv Spectroscopy. The data thus obtained was treated as above. The result of one such run, at 146.5 i 0.2°, is Shown in Figure 1. AS in the first method, no linear plot was obtained from some runs. The values of the rate constants from those runs which gave a linear plot for part of the heating period varied by i 10%. Method 3. Samples of 0.10 g 41 in decalin were prepared as in N—2, above, and heated in a constant temperature bath at 147°. Samples were removed periodically and the reaction was quenched by cooling the sealed tube in an ice bath. Each sample was then examined by uv spectroscopy as above. The results are shown in Table V. S. Effect of Oxygen and Di-Efbutyl Peroxide on the Rate of Reaction of 41 in Decalin A sample of 0.10 M 42 in decalin was sealed in a Pyrex test tube and heated in a constant temperature bath at 147° for 30 min. The solution was then examined by uv spectros- copy as in R, above. It was found that 88% of 41 had re- acted. Samples of 0.10 M 4Z'in decalin were degassed to 10-3 mm of Hg by at least three freeze-thaw cycles. These were then heated in a constant temperature bath at 147°. Samples were removed periodically and treated as in Method 3 of part R, above. The results are shown in Table V. 67 Samples which contained 0.10 _1~_d_ 41 and 0.005 M di-_t_:_- butyl peroxide were degassed to 10-3 mm of Hg and sealed. These were heated in an oil bath at 148-9°. Samples were removed periodically and treated as in Method 3 of part R, above. The results are Shown in Table V. Vpc analysis (5' x 1/4" Carbowax column at 187°, He flow rate of 90 ml/min) of a pyrolyzed sample Showed that 48 was formed in 14% yield. No other low molecular weight materials were detected. SPECTRA Massggpectra The partial mass spectra presented here Show the peaks with relative intensities of greater than 4%. The Spectra were taken with an ionizing voltage of 56 volts and an ionizing current of 10 uamps. NMR Spectra The nmr spectra presented here were obtained using CC14 solutions. 68 69 Relative Intensity ..-... ........'. ......... -..... I _ ... 1.61- ...-..-... .... .. ”WI ......- 80"” ... .......... ._ i .- ”-i- A i- ' I 60 «4 - ....... ..-...- - ”...--.-. ...... ...“... 1.76 . ..... . ........ - g4. - _- - -. -_fm - (F. ..-.-- 3”-- 40. ........... 44 _ -- gm.-,-lg 20 ...... 145 155 165 175 m/e Figure 2. Partial mass spectrum of ethylpentamethyl- benzene (48). , 164 f. .. - ... _ o .. . . y...o .. - » . . . . ,. .I .. . ... ... . . n. 1-- ..-..- .. -....a---4 - o... . . .. . . ...... ................................................... on O 60 d ;-.,. . ........... ........ ...-... .. .- . .......4 ........ ......H 179’“ ... ........ Wk 0 L o X f i ................. .. .. . . . ‘ .. .. . . .. . . .~ . . . .--- .-...-... -.- -. .. . ., . . , , , . .. - . . -. . .. . . _, . ......... . .. 1" . . , ,_ .. .... . ,,,,,,, .. .. .. . . , . ‘ . . . . . ... . .. .. .. ..... .. . ..... . y . A . . ..4 l .4 . ‘-.. . . . , . .... a ‘~ ‘I . .- . , ... ...” .. . .. , . . , . » l... Mo-h— v .4 ”.....- : - ~o- v: T— ‘ 1‘” . , ... . . , . . . . .. .1 . ....g . ... !. ... . ~ - - - , . . .... . ., - - ..... UH... ....... "I” . A. ~ . -> ~»a... A... H... ... ... . . . . 4 - « - ----- .... ... ... ...... . . . , » ... . . . .. . . . . - 7.1.. .... .... . ‘ » . . . ... ... .. ....u . .. . .. .. » .. ~- . , ... .. 145 ' 1 5 165 175 Relative Intensity N O : m/e Figure 3. Partial mass spectrum of labeled ethylpentamethyl- benzene (88) from the gas phase pyrolysis of triene 88; 7O 165 164‘ 1 . .................. ...-....____..._.1,..-,.,- ,- ,__,,_ ......l.......-... . ,, .-. .161 E >180 4 --------- ~ .. .. _ .................. ... .|. . u ._ 1 1 "-1 ......................................... g1 . U} 8 1 .3.) 60 -1[ ............. i C 1 .. E H ........ ............... _. .- - ...... .179 ... 1 EB .4) ... 40 ~ ,. -— ~ - ... ............ .. .= g H ~-§3:?ii'j‘. ”- H" ""T" """"" 77176::f'f'f 3" 2° (Iii...- " “ I ’ 1771321 “““ ..;;i;::i-1z. II I 1 I I 1 J..- 145 m/e Partial mass spectrum of labeled ethylpenta— methylbenzene (39) from the pyrolysis of 0.10 M 88 in cumene at“350°. Figure 4. _ 1&4 _ -- _ 161 ............. _ ..... - _ --- ......... :Ngo. ...................................... -g ................ _ ................ .p H .................................................... ,. m | c: 1 1 2360‘} .ww. -mW-mm W. -m-fl- .4. 1* a - l H .......... 5.. ...... .- .- : . 1" g . j} M 179 I . 40“: ”i? ”“ *‘ ‘ “ " “t ‘ " 73 ‘ 176 a: 20 ..tf'fir"? "“1"““45 “- ~"~ — “ “"‘f'f‘ff‘af'JjSiZj-M 1 l1 1 ! L1? I' [1 145 Partial mass spectrum of labeled ethylpenta— methylbenzene (88) from the pyrolysis of 0.82 M Qfi’in cumene at 150°. Figure 5. 175 165 /e -.--... ‘-..-. .. . . -.'.. . .. I - . -- .. . . .. . . . >uovoo- ... .... ..... ... ......... . ..... ,-. .... .. . . .... .... . ... .... .... . . . ...... ........ .....o.-n ...—-... u..---... .........‘.. . ... .- ...-.-- --oo.... 1-.o--o~-- ...-9...: ......oo. coo-Ono- .-oso... --.-....x........-..........Q..-. .. . -..,.... - ..--—-- uooooooo ...¢o.’.' .OOOOOI-I .D-u>.q .. ... . . ... .. . 0 . --.--.o .o-.- -.. .I o --- - .- CI...- ..-.cv..o .00.... a ... ~... . . . ...,.......... . . . . 9. ...... . .u-o-n .{...-.ouon IOOIoO-CI o... ...-n - .. . . . .. ... . n - o 3 .- o ... . ...-. ~ -n—... .u ......«a ...----- ---..o-o ‘~-.......,.........'-.---..-~...-....-.’ .........o....o-.- ....Oot n .o-¢- . ...—. ..-......g. .. . .. .. ...... ----.- .- »-. a. . ... .. ........ ......-. .......-. I _‘A v v ...-n .1........ '..... . ... .9 . A. I. ' .... - A - - ...t-. m...: I a. .v...-.-: -..-.. .. .. .... ....... .:-. ...-.... ...- I». on. -. ...... .A-o.~ , ova.~..on no.0... o -.-m---t.------.—.‘.........'o.--oo-- .0 ------.t......... .... ..-o o .0...”acco¢.ooo4oooooon-O - o--v- .. - .............o ,,_,... . . .... .. cots... . ...... .....3-00 cooooooo. . 0...... -- -. .1 .....-.'...-....1........{. .. .. ........ . .-.. .- o.o-c.o. 0.0.0500. --—‘ -~ 1 - ---~ - - - - - - .. - -- ... -- . -o---.o-- ...-otucc u~¢ooocoo ova-oo'-‘ ~ -- -... - _. . .. .-..... . ... .. .. .-. cocooo- .0000... 00...... olnoooono cocooouoo ---. ~§ . . .- ... . .... . ......1........ ..-.-.... .....-- . ...--ooo. ......... ....... . A A v V v o-o--—--Q----.-.. 1.... -..-.. ..-... . -.... . ......u.........1..-....-- . o-o...- o.n----o- - --- -- -- -- - .0 -- - -‘o - .-----I. -o.---¥-O§-.-.- 5......--. ...-.... ...-..... ....... - - -~~ -- -..»--.-...... .. -...........-..... .... ...I ......~ 1 .. ... ....o..- ... nov-O -.---.-.-. .--.-.-.. -..-.—... -.....o-..oo.-..ooooo -ho‘—‘ ~---o-oo~ -.-...o-.:..-o--.oo!.--oo.-..t....-.noot-ooooo-o- .....n... ......ooo uoo-oooo- .. n-cco' v-IO-OO—nov ....--.....-.-.. . ............ . .... .H.-~o.- ......o- --oo-.--. co-ooac-o -- --¢-'° -o---o..-‘..-....-..-----....' ... .. .... . -..‘..-.o.-.- ......o-- ...-II... ...-us... can 10" .-..--.-... -. .. . ...... . . . . .. .9... ........ .-....-. .-.-.-.. m --- ..-.. . .. . . . . . . -. ...-...... .... . .‘n..-.-o -.o-.o. - ~ , A ......-._.1 .------.'..-..... ' .. .. . .. ............ .....-.. . ...... .-... ... .— .-------. ...-...- ..--.--.. . . . .... ...-... . --..... ...-...I- ......n-- .. ~~o--.- » --«-.---- -.-.-. ... . .--..'--- ----- --.o-...o ooovo..~ w-.--.- 0.0. u0-4-- t 0' ' ---.o--.- o----o-».i..-.-.... ..---... oooo-uu-o ...oao-oa ........o 000‘... ICC. o'o‘ Otanoa’ I .-.- .-- .--..--.. ...---..-‘..-o. ...‘.. - .... 009...... ...-cocoa N --.-.---- --.-..-»' .... . .. ... ' ....a- 0......n. . .....o. .o '0‘ol n 9...... co..co‘-. ---...-. ‘ .. .. . .. . .. .. - o~--- ......... ......... £0.04. nod-H‘s- coco-o-n . . , ... ..1 ..n ... at“..unun..ruu A ---.-'*- -~ - -U -----I--' ~--~o -oT-I A-cll-IOIII' 00- p.-..-.v- IO’OOOOOC ..‘OOOOI. III... 0‘ --------.)..--....--- 3.-. .... . ....-- --....... ......o-- ...-cc... o- co...- -..-oo-n- ...-«.... ......-.- ..-.....- -...... . .--.. . ..--..... . - .00.: .....no-- ......... ......aao ..uoo-’-o ......D.. -.-..--I- I..'I.l-- on-tounu~ ICDIII.. 0...... - --I-..C¢ ......lol IOOCOOOO .....". ----... . ....-.. .... ... --..-.-. . .- .... .- ... .. - -.....< ...-noun. n-go-oooc o-n-o---l -.--.- - .. -.--.‘-........ ......... ...-...u. o .....I- .--.-oo-- 0.0.3-¢- ......g... .ooo-ua-n -.n<~-- .----.-. ...-.....i“....... --->o-.bo OOOOOIUCO I O OI... ......O'. 00.00.... ......C.‘ -.. «--- -~o~..o.l)-o-...-.. -.o--¢... ......uo. 000.00... ooa-OOOCC .«ocaauoo Iboooooo¢ Oooe~o u- . . ....... . ... ...... .... .... .... o...-... .....A... ...Ho-:o 44.....oo Dalton-t A ---..---- ...-.... ...-oo-o-I-u- --—.- ~-0---cpo -.-.-oo-- van-....-}--o-ponoo .oooooooo coa-oo--- - >-.-.-» . , --. ....n.-. I. -. . - .. . III-IA u: ...-.n-. on-Olddol ...-l... '9. -"" ..--- ...-.---. ....-.---.--- .-)....-... --~..--o- ..----.-- ...-o....}o...-oooo ......o-o .. ..---. .-.-.... . ......' .. .... - ...-... .c-u-u... ..--..oo. .C-QQ-OOO 0.900.000 Oolccnoa. .------- - -.-.-.. . ... .. . ...-. - -. ---. ---oooo-->-oo-c---. o-Oio‘... .ooo-ooooo a-«.-c-- --.--c-. - .-. .. ...... ... -. .-.. . ...---- .-...an- .......-. ...-co.-. ccoooocl- IO-oOII-o .-.. ... .. _. . . . o . ... . 4 un......- .....oou. ,,,.. -- o~1...'.O.. Cocoon... .a ao-unu . .. .... .... .....-. ... .... ....... ......... ......... uooocnuo. .....o.-. ...... o. . ...... . . .. .. ...... ». .. ... --.-.-.-. ......... -.-.-.-oo - C‘s-b: e-OOOOAC- .....-.-n ......-- ......-. .--..-.. .....-.- .. ..... ...-u .. noooo.... no.0...ou ...oaoaoo ...-scoo- ... .. . ..-... .. ....a...- ........- ......co- ......... ......... ....oagu- ~- ...... ...-too. . - -. . -.. . .- ...o-an- nun-.4... :1. Ir... 11.....l. 4...... I A v . . . . .. . .. .. . .- ...-.... -......o . 0...... ...1.-- . ....-. .... . . . 1. . .. .... ........ ... .... .. our". IoOquQOI I o-OA . . . . .. . . . -. a ... . ... .o. .... ..aogo... n4-no.:1 4.. . . .. .. .. .... .-~ ......... I-ooll 6 ......o.. .noaa.-.. . ...... .. . .. . ... . . .. . ...... ....... - -. .5... too-o0... do... 0.. ...-on... ......-.. -...--- .. ...... . .... . ...... - ...-.... ...-0C... Icooonooo son. .00 -c H .. - . .n ---a‘ -..uu-~ oil-7 I» -.o a.- 4vII-o n- II... .1. I 0...... I QIIU-II 3.... “- - ... ...... ........ .. .. . ..... . . ...o. .... .- co .....~ ......... r. »---.- --. a-- 1 1---~.-- .. an... ....-.. . .n-.~-. goos¢>o~o 0.00.0004 on... .0. ..oo--o'- H .. . . . ... . . . . .. ......u . .... .... ......o-n .. .... . . . ..-. .... . A .. - u..- on u ...... ...-l - ......‘I. 000...... I'D. to IOIa t h-CQOVO—LQQIQO 0.0 0.0.0... lit-0.... ...-ill. I ~00” G I~OQOOIOO n 0...».. II. .0 .....l'l\ . ..-;a. .... .... -. a. .. ......... ......... ... ... . -..o.4ooc ...Aooo-. o 00-... . .1 . o . . . . .... . ..... ........ .....n- ..-...o-s ...-o-A-o on. - . ~’-' - c -. .. . .... ..5 .... .. ...-....a ... ... ... - .- A A A A -t O : o t I.- Acconctun 3-. 00-0. -- vv - n ...... ... D . . ... .... ... .-... - . ......... .... .. . .. . .. ... .... . ....-.- o. . . . . . . . . .... . .. . . .... .. . ---.. ..- -fi- .. .. ...... ...-... . .... . ........ ... .... ... . . ..oo-.... .,-.. .- ..-.._.- .. - ....‘JQCO-Aaua -.-¢. o..1..n.-onoo ...-on... coo-oo-c‘ to... ~ on on..-- .....,.,. - I-.¢ . . . v I -- -o I .a I- ~ g. . ~l| .- ......JCI .. Clio. ... .C‘lu -‘ooaool| ...o- ~ - -- ‘~-~-- ...-9--.. ...-..... ......u. ......o.. ... ..... - ...-on r.. too-o oun-p-cn. »...--.- . ...-o - no.0.oo.v ........ ......4-- .....a.. ...-u... no... - no. u-o» u a... ,. -... _..,..... . . .... . ....... . ... ...-... .. .- .. ulonooooo -...--. - ...-n . . ...-..ooo -- -.-.. ---¢----- c.....-.- ..a...... . ... - ~...--a-« noun-n... ......... b..~....n ....-..-a a. o.-.- ...-9.... ...-.... ......... ....o . ..not. o- UC-oooool ...-an... o.n... .- ... ... ........ -........ .. .... . ......o.. . ...-... ...-o-yo 410...... none-coo. ...-0.... ...-.... ..-o-- on ......... ---- 0 -. . --...u- ogooooo-- -aun‘cq-- -¢-.-uo¢a $.0'O,~. ...-OI... ...-... ~ 0... .n-.....- ---.¢..- .....s... ......a.. ....... - ...Aognuo oucauqovo aaaooaotu .. .. . ..... .. ... . ........ . .. .. ........ . ..... ......... ...on.... ...... .. ..--... .... . . .... ...-no... ......... . .. . ....--..- - ao-nc- Stoo-O-u 0.0.0.0.. .....-o- 001...... -oooooooo ~n-o.-¢.. .- nnoooo ...‘O.nlc .onn-uoon toa~O-uuo uooo-ouo. cos-....a . .. . ...-..... ......oo. .. ...... cocoon-o9 ...o-ooa. ......nat 4ob$¢~ooo a.|40000 id.a.¢n.. ....H.-oo ....oooo. ......... o. .0... .--ooo. - ...ooa... ....ol-oo o‘ooQ-o-o -¢Oo-oooo oo-ooa-no ...-on. . putt-.... .-....- - ...-.-o¢. ......oo. ......4ao ......oac nto-odoooo ouooooocnr...oag.o. --n~-v-- u—ouooO.‘ -~.~..ll ho IIII‘O 0.0.0.! I no. bhiul tullio I. i ,.._,._ ,_,.._ ,, ,,,.,_... , ..--... .__.,.... ,....... ......co- c.4060... on 0'000‘ -...-.... .....-.. -~-....-- .0. ...-n ...... -- ......ono -u-oon .. - .‘4¢o . 0.06 '0 6a.. no 0 ....... . .....--- ..-.... ...-pa... ... ...-.).4.....-- ......... .....oa-n .o¢-vO-‘o na-o..... -...-..-. . ...... ......o.. .....a-.- ..-...--.4. 5...... .otcouv o .....oo o ..oaoonno ......aoo loos-gn-o .-...n. -.-~.-... o-oooooo.- .... .. . .....—... ---4..n-. .....nooo ...-4.... 09-04.... .....ooun .-.--... ..-..-..-(-........ ,,....... ....o...o lung-e-o-(coaaoooooj0.0-0.... ......o-o ......... ....-n.-. ~.----...-(-.—....-.o ...o-oooo .o-----.- ...:o.-. ....n..-. ...-on.-- ......... nao.-.... .a..~-- .-- o-oo- .........4......... ........- sooalo-O‘ --- ----- Oo-IO“-- -00000o6¢ u........ 4.,.‘..._ .-_H-7-_. .__.,,,,_ _,..‘,.. _...,. .- ..-...... .......o. .....ooc‘ ......... ..... ...' . . . ... ..v.... .... ...... .... . .... ...-n u- .no-- -0 coco-00¢: too-IOI- - ... .- .-.-.-., _.-..-._........_.,1,.....,..'......... ....-¢--- uo-Aol--¢ ‘Ils'.‘- 00- hit-D --o-ou... .... . .. . - 1....----. .. ...... ......... ......-. ......... .... ........ .... ... .....-.-. .-.......... ...... ......... ..-..ooo. -.-~.. .I ....oaooo ages-u... .-.. ...-.... .. ......I......... .....-... ......... ......a.- -.---.n.- ooOQOOOQO ...-.000. 40......- .,._.-__ ,_ ...._. .._._... , ...-... ...... .. ......a . .. ...... - ...... «.0400... «...-.... . H. ._ -._-.__1.__. ...- _...._-_. ...-_.._. ,........ .......-- ...-on... ....o-aoo ......-.. -....--.. ...-......,-.,..,,_ , ....... ,..-..--. .....--.- -..-IO... ...-0*... ~oa. o... O -.- - - u-u. 0.1....OO.QOO-iovoc O. 0 -¢ 0 .400 ODD."' ‘ '.....‘ ...'...‘. .' ...... .‘."".‘ AL ...-o . ......... ......-.. .....noonfi‘.o---<... .... coo o-0400ncc o ...... .....a. .. ...... ... .~... ....... .1......... . .... . ---aa ... 000-0.... o o ’--'-..- .-..--.- -...-... .........1.-...--o0 -...II.‘0 ......‘Ol I COD... ...-OQ-‘g ___.--,_ ___ ..---‘_.-_, _,,i . _...-........... ...‘...-.1.....'... 0 0000-! Ito-o-o - .... .--.-.-, ........ ..................- canoe-o-o C.......- . Ooo¢—o¢ .‘OQ.0¢.- .--.-o-»-4..o---... -.-.--.---3.---..--.‘---.....- --.o..-.oo4‘o-.4coonlo 4“... 4...- ---“--. .---....- ..-..--.. ._......-§---..-.ou‘..-..ooco o-..‘.~oc¢{o ......4..-...... ..-.--o--4 >o-c--.--.-.-4----01...o.¢o--Q--— -..-.I-.-.----. .-.-00*‘0-[O 00-.“ n“... n. .........1 -------,..-.-.-.-. -......--1.-.--- -9 ot‘nouno- -...704... o ...... .....lovl .......oo4.oo-o--o- ...-...-.p..........-o-- ..-- ~----'-“ ""°°"'°" 40'.-. o........ .n-oo--.- ..--..oo->0..-0>¢0001-.o-..¢¢. -.--...oo o-v-o-~-- --‘*‘°“‘ ‘OCficolooocaaoon 0000-0-.. ooonooocn ng‘ooofi9. coo-.-.o. o-uoonco. .-ocooaoo ---ooa.-o - ¢t----o---H¢ooo-oo-O‘ o—a .¢-¢.o... -.--~-Q-}‘--,_‘----{..‘---~-- .. ..-...}....o -'.-".‘.4‘ 50940-0... 0H“....* .-.--,--- __-_-_--. ..‘-_-.-_ _.--..... ....-. -- -.------- ~ ...--..o-ouooooo4 HH.-.4...---... .---“0"-H_.H_ ----._.__ __‘-_'-___ .._,_---. .....-..- ---o—---4~00:--.—o4.4¢-.—-54._. H.---.,. --Ho----4o--.‘-oan ..‘-o.oo-(t-oo-.-o.- --—>~---- --°“"“ "“""‘"‘ “""‘.‘ “""“‘ 900...... > + Hexane 56 57 58 W w M Barton and Quinkert, (38) (1958) O h «.11: \ | 92 A Q? 2?. ‘\ \ ONMe2 61 Griffiths and Hart, (39) (1968) 00 m rw Zimmerman, et. al., (40) (1965) 79 0 OH Q h V + > O cc13 64 65 W m Schuster and Patel, (41) (1965) O O -——-—> CD CD (13 66 w 5.1 Zimmerman, et. al., (42) (1966) The photochemical reactions of conjugated olefins have also been reviewed thoroughly (43—46). Several examples of the photochemical isomerization of 1,3—cyclohexadienes and hexatrienes to bicyclo[3.1.0]hex-2—enes have been reported However, the photochemistry of compounds which have the same fixed geometry and the same w-systems as cyclo- hexadienones §4 and §§'(6§'and 62), but which cannot be excited to n-#* states, have received only little attention. R 80 In 1958 Barton and Kende (48) reported that dehydro- ergosteryl acetate (22, R = CH3CO) rearranged photochemically to photodehydroergosteryl acetate (1;). Irradiation of the C10 epimer of ZQ'(dehydrolumisteryl acetate) gave the C10 epimer of Zl'(49). These results indicate that the isomeri- zation was stereospecific. C7H13 C7313 D. h, . > R0 R0 70 ll 22. Ii Bird and Cookson (13) found that the cross-conjugated triene g'rearranged to 4'when irradiated. Similarly Newton and Layton (14) found that irradiation of 9 gave 12; These reactions involve a radical chain mechanism and also occur (thermally (see the Introduction to Part I of this thesis). CleC COZH o or A 7 cozn CHC12 3w 2th 81 C13C COzEt C02Et hv > or A CC13 2’ 10 Recently it was reported that 22 isomerizes with phenyl migration to give 28 when irradiated (50). The correspond- ing dienone (62, above) rearranges photochemically by a "bond-crossing" mechanism. hv 72. 73 Zimmerman, et. al., (50) (1967) It was of interest to study systems where the 4-sub- stituents of Qg'were alkyl rather than aryl or halomethyl groups to determine whether similar rearrangements would occur. Consequently the photochemical reaction of 1-methyl- ene-2,3,4,4,5,6—hexamethyl-2,5—cyclohexadiene (11) was in- vestigated. It was found that 11 isomerized to 4-methylene- 1,2,3,5,6,6-hexamethylbicyclo[3.1.0]hex—2-ene (Z4) upon 82 irradiation. The bicyclic diene (14) rearranged thermally to 5-isopropenyl-1,2,3,4,5-pentamethylcyclOpentadiene (15) and this reaction was also investigated. The mechanisms of these isomerizations were investigated, using labeling and other techniques. 17 74 75 V RESULTS AND DISCUSSION A. Photochemical Rearrangement of I‘Methylene-2,3,4,4,5,6- hexamethyl-Z,5-cyclohexadiene (11) 1. Isolation and Structure Determination of the Photo- product A solution of 11 in ether (0.23 g” 5% by weight) was irradiated through Vycor with a 450-watt Hanovia lamp. The sample was maintained at 200 or below by cooling in a water bath. The reaction course was followed by vpc. As the re- action proceeded the peak for lz became smaller and a new peak which appeared became larger. A second new peak also appeared and became larger at the expense of the primary photoproduct. The photolysis was stopped when the primary photoproduct accounted for 45% of the volatile material (polymeric material formed on the sides of the photolysis vessel). The primary photoproduct was collected by prepara- tive vpc as a colorless oil. Elemental analysis of the photoproduct showed that it was isomeric with the starting material. The photoproduct was identified as 4-methylene— 1,2,3,5,6,6—hexamethylbicyclo[3.1.0]hex-2-ene (24) on the basis of its spectral properties and its facile thermal con- version to 5—isopropenylpentamethylcyclopentadiene (12). Several initial attempts to isolate 24 from the photolysis 83 84 of lz'had in fact given 15; It was found that this could be avoided by keeping the temperature of the gas chromato- graph at or below 150°. The thermal rearrangement of Z4 to 25 is discussed below, in Section B. a so 17 74 75 ’w ’W ’w 95%Et0H max The uv spectrum of 24 had a x at 253 nm (8 11,300) and the infrared spectrum (neat) had bands at 3080 (vinyl C-H), 1625 (conjugated c=c), and 850 cm.1 (terminal c=CH2). The nmr spectrum of 24'(Figure 9) showed vinyl hydrogens at 1 5.32 and 5.47 (1H each) and allylic methyls at T 8.34 (3H, Cz-methyl) and 8.44 (3H, C3-methyl) which were broadened by homoallylic coupling. A singlet at T 9.35 (3H) was assigned to the ggdg_C6—methyl (the upshield shift is due to the v-electrons of the double bonds; see compounds 16 and ZZ,in Table VI). Two singlets at T 8.89 (6H) and 8.97 (3H) were assigned to the C1-C5 methyls and the 2x2 Cs—methyl respectively. The spectral data of zg'are in ex- cellent agreement with those of similarly substituted homo— fulvenes. Some examples are shown in Table VI. In an at- tempt to prepare 24 by an independent route, hexamethylbi— cyclo[3.1.0]hex-3-en-2-one (62) (28) was treated with 85 Table VI. Spectral data of homofulvenes Compound and NMRa UVb, nm (a) IRC, cm_1 Refeé- ence 5.32 5.47 r*w 9.35 253 (11,300) 3080 8 4 8.97 13:3 8 34 8/28.89 a i E 22, i“ 1 5,29 9.30 g 5*“ H 255 (10,800) 3090 51,52, I 1622 54 '1 848 i 9.03 8.3 ‘\ 8.38 “/2 8.85 8.93 22‘. 9.34 9.25 53,54 12, 5.26 r-A—x H H 252 (10,000) 3085 55 8.27 "" 8.44 \\ 854 78 86 Table VI. (Continued) aThe chemical shift of each group (or groups, where definite assignments were not made) is relative to TMS as T 10.00. bThe uv spectra of other homofulvenes are reported in refer— ences 48, 49, 52, and 56 and the values are similar to those shown here. Paquette and Krow (57) reported xmax 255 (a 10,600) for a compound which was either zg'or 11, CThe numbers shown refer to the bands for vinyl C-H, con- jugated c=c, and terminal c=CH2 respectively. dThe references refer to where the compound was first re- ported and to where the spectral data were presented since in some cases no spectral data were given for the compound in the original paper. 87 methylene triphenylphosphorane. No reaction occurred, presumably due to steric hindrance in 62; Attempted dehy- dration of the alcohol obtained from 62 and methylmagnesium iodide gave 22; This reaction is discussed in Section B. O + CH2 = P03 > No Reaction :52. 1) CH3MgI 29. > 2) -H20 22. 2. Solvent Effect, Quenching, Sensitization,and the Effect of Different Wavelength Light Triene 11 was irradiated through a Corex filter (no light is transmitted below 260 nm). No products were formed as determined by vpc analysis. When the triene was irradiated at 254 nm several new products were formed as determined by vpc analysis. These were not investigated. The failure of 11.t° react when irradiated through a Corex filter is con- sistent with the fact that ll'has a kmax at 256 nm. The formation of several products when $1 is irradiated at 254 nm is reasonable since the initial product (74) absorbs light in this spectral region (xmax 253 nm). The further 88 photochemical reactions of Zé'would account for the large number of products formed. An attempt to photosensitize the rearrangement of 11 failed. Irradiation of an acetone solution of lz'through Pyrex produced no products although acetone senSitization of photochemical reactions is well established (several ex- amples are given in reference 58). The reaction of lz'was not quenched when piperylene was present in the photo1ysis solution (see below). In fact the photochemical isomeriza— tion of lz'to Z4 seemed to be slightly more efficient with piperylene present than without it. These results indicate that the rearrangement of lz'takes place from a singlet rather than a triplet excited state. Solutions of 12.1“ benzene, hexane, methanol, ether, and ether-piperylene (9:1) were irradiated and the photolysis solutions were analyzed by vpc° The results are shown in Table VII. The data presented in Table VII show that lz'is con- verted to zg'in polar as well as non-polar solvents although the reaction is slower in non-polar solvents. The presence of piperylene inhibited polymer formation to some extent, possibly by quenching some side reaction which leads to polymer. 3. Photolysis of Labeled Trienes The formation of Zé'from lz'could be considered to arise by a “bond-crossing" mechanism or by a methyl migration. .In 89 Table VII. Solvent effect on the photochemical rearrange- ment of 11? 11“ Solvent Irradiation % of Volatile Materialb‘e mg (5 ml) Time, hr .£1 14' OtherC 200 Benzene 24 59 41 --- 200 Methanol 24 51 49d 200 Ether 24 45 53 2 100 Hexane 19 73 27 --— 100 Ether 20 45 (29) 47 (31) 8 (5) 100 Ether-Piperylene 20 43 (35) 47 (38) 10 (8) (9:1) . . a b C All samples were irradiated with a 450 watt Hanovia lamp through a Vycor filter. Determined by vpc analysis; conditions are given in the Experimental section. Polymer was formed in all cases. Except with methanol, only one "over-photolysis" product was observed. dThree products were formed, one of which was 74, but 74 e and one of the other products had nearly the gzhe retgfition time. v The values in parentheses for the last two entries are the actual yields, as determined by vpc. Polymeric, non- volatile material makes up the material balance. 90 order to determine which of these mechanisms was operative, deuterium labeled triene ég'was irradiated. "Bond-crossing" should give 12” whereas methyl migration should give a mixture of §Qjand 8;” In the event, irradiation of éé'gave ybond- 5', //crossing> ‘\ * as 79 * *- * .. \ r 0’ * methyl + migration) ‘ \ \ * *- 29. £1. 22; .The nmr spectrum of 22'(Figure 10) differed from that of Zé'in that the peak at'f 8.89 for the Cl- and C5-methyls now integrated for only three protons. In addition the peak at T 8.34 for the Cz-methyl was absent and the peak at T 8.44 for the C3-methyl was sharpened as expected. If a mixture of 82'and gl'had been obtained the peaks at T 8.97 and 9.35 (CG-methyls) would have integrated for 1.5 protons each and the peak at T 8.89 (C1- and C5—methyls) for six protons. .In conjunction with the thermal rearrangement of 74 to Z§'(below), it was of interest to have bicyclic diene with the exo-methylene group deuterated. Triene gg'was prepared 91 from hexamethyl-2,5-cyclohexadienone (22) and trideutero- methylmagnesium iodide as described in Part I of this thesis except that the Grignard reaction was hydrolyzed with D20 to avoid loss of the deuterium label during workup. Triene Eg'thus prepared showed only a trace of hydrogen for the gngmethylene group at T 5.26. .The nmr spectrum of the labeled diene (82) obtained from irradia— tion of gg'was identical to that of Zé'except that now only traces of the peaks at 1 5.32 and 5.47 for the vinyl protons were present (Figure 11). O x» 1 )CD3MgI M f > > 2) 020 \ 24. 252. 5%. 4. Mechanism of the Photochemical Rearrangement of $1 The photochemical isomerization of 11 to 74 apparently arises from a singlet excited state, as discussed above. This is analogous to the photochemical isomerization of Z2 to zg'which also occurs from a singlet excited state (50). .A mechanism for the rearrangement which accounts for the “bond—crossing" which was observed is shown below. 92 ; o O 2 Hz hv > > > 4/ 1,2, 522, 8.4. 'CH2 i \ > \ l- 85 74 . or 84' > 74. The conversion of ZZ,tO 72 takes place with phenyl migration (50) whereas the analogous dienone (66) rear- ranges photochemically to 61 without phenyl migration (42). ¢ hv ‘ \ H \\Ii ¢ ¢ ¢ 12. 2.12. 0 0 0 ¢ hv > \ H 0 ¢ H 66 :51. 93 The migration of a phenyl during the rearrangement of:§g to zg'is rationalized by the fact that the singlet energy of the triene moiety and that of the phenyl group are similar and that distribution of energy between the two moieties may occur and thus facilitate migration (50). Triene £1 has no phenyl groups and the distribution of energy in the excited state which facilitates phenyl migration in the case of zg'is not possible. The isomeri- zation of 11 to 74 could be considered to be an example of the photochemical di-w-methane to vinylcyclopropane rear- rangement, although these reactions normally occur from the triplet excited state (59). For example, the mercury sensi- tized photochemical isomerization of 82'to 81 has been Of \H H 86 87 rw m reported (60). Irradiation of protonated 11'(17-H+) to give hepta- methylbicyclo[3.1.0]hexenyl cation (74-H+) has recently been described (61). | ‘ / hv > Q _H+ 17-H 17 Ill 94 Orbital symmetry considerations show that lZ'——¢ §4' and 4Z;Ef'-—> Z4zgf are photochemically allowed for a disrotatory process (62). Intermediate 84 can then go on to product (Z4) or collapse back to 41; B. Thermal and Chemical Formation of 5-Isgpropenylpenta— methylcyclopentadiene (72) Initial attempts to isolate the photoproduct of lZ.bY preparative vpc (see above) gave a colorless oil which was identified as 5-isopropenylpentamethylcyclopentadiene (22). The cyclopentadiene was shown to arise from the thermal re- arrangement of Z4 when the latter was passed through the gas Chromatograph. This is discussed in more detail below, in Section C. Vpc > (thermal) 17 74 75 W m m 95%Et0H at 248 nm max The uv spectrum of 12 had a A (e 4250) and the ir spectrum (neat) had bands at 3080 (vinyl C-H), 1640 (conjugated c=C), and 885 cm.1 (terminal C=CH2). The nmr spectrum of zg’had a distorted quartet at T 5.25 (2H, vinyl protons) coupled to a triplet at 8.93 (3H), J ::1.0 cps. The coupling constants of the cis and trans vinyl protons with the methyl of the isopropenyl group are 95 slightly different and this causes the "distortion" in these two peaks. A singlet at T 9.03 (3H) for the C5-methyl and quartets, homoallylically coupled, J = 0.7 cps, at T 8.27 (6H, C1-C4 methyls) and 8.45 (6H, C2—C3 methyls) made up the remainder of the spectrum. The shielding of the v—system causes the upfield shift of the isopropenyl methyl protons. A similar shift is noted for the methyl a to the methoxy group in compound 22,1n Table VIII. The nmr spectrum of Z8'is displayed in FigUre 12. Cyclopentadiene 28 was also obtained by dehydration of the alcohol obtained from 82'and methylmagnesium iodide. The alcohol (88) was readily dehydrated by passing it through a gas Chromatograph, even under mild conditions (see below). This reaction was initially attempted as a means of producing 24. 0 HO CH3MgI -H20 \ > \ > 88 60 75 ““V "W NV Catalytic hydrogenation (1 mole H2, Pt, ethanol) 0f,Z§ gave 5-isopropylpentamethylcyclopentadiene (88). The nmr" spectrum of 82'had a peak at T 8.29 (12H) for the allylic methyls, a singlet at T 9.12 (3H) for the Cs-methyl, and a doublet with peaks at T 9.26 and 9.37 (6H) for the iso- propyl group. The seven line signal for the methine proton 96 Table VIII. Spectral data of cyclopentadienes 90 Compound and NMRa UVb, nm (8) IRC, cm-1 Reference E53? 248 (4250) 3080 1640 9'03 885 8.93 8.27 8.45 22. 9.26,9.37 (/,/~ 9 12 260 (3620) -- H . 1650 8.29 22. 4.96 242 (6600) 3080, 3055 53,54 r’*" 9 00 — 1625 63 ”' 906 w ‘ -——-—--—~— --y-.o-——. .' b.- \... .' A “'1. ‘ ‘ *‘-‘;1=~z;r... 97 Table VIII. (Continued) a b c —1 Compound and NMR UV , nm (8) IR , cm Reference 9.09 252 (4140) —- 64 r‘J“”‘ 1640 Em, 8.30 i l 91 6.93 ‘_7 11 0CH3 6.73 3 9.45 ‘ 9.02 260 (3800) 54 VNII’? 8.33 2.2.. 3The numbers refer to the position of the signal for that group (or groups, where definite assignments were not made) relative to TMS. The signal for the methine hydrogen of 89 was not detected. Nmr values of other substituted cyclkov pentadienes are given in references 57 and 65. bUV data for similarly substituted systems are also reported in references 57, 65, and 66 and are similar to those reported here. CThe values given here are those for vinyl C-H, conjugated C=C, and terminal C=CH2. 98 of the isopropyl group was not detected. The spectral data for 28 and 88 and similarly substituted cyclopentadienes are shown in Table VIII. The tetracyanoethylene and maleic anhydride adducts of 28'(88 and 84 respectively) and the tetracyanoethylene ad— duct of 88'(88) were prepared. The adducts are shown below, along with the nmr assignments. The values in parentheses refer to the multiplicity of the signal and the coupling constant (b = broad, bd = broad doublet). 8.32(bd,1.1) 8.35(bd.1.1) [{45L\\ 8.56 011)} A. 0‘01“) V6301 CN CN CN 95 m The signal of the methine proton of 88 was not detected. The realtively low field positions of the methyls in the tetracyanoethylene adducts (88'and 88) are similar to those of the analogous methyls in 88'(27). 8.19 CN O 96 W In the adducts 88 and 84 one of the vinyl protons is strongly coupled to the methyl of the isopropenyl group whereas the other proton is only slightly coupled to the methyl, thus causing the methyl to appear as a broad doub- let. The similarity of the chemical shifts of the isopro- penyl groups in adducts 88 and 84 indicates that these groups are 3g£l_to the tetracyano and anhydride moiety in each case. This is consistent with attack of the dienophile from the least hindered side of the diene in each case. Maximum overlap of the w—orbitals in the transition state of the addition of maleic anhydride to 28 would give 24. with the anhydride group 22921 as shown. The unexpected facile thermal rearrangement of Z4’to Zé'indicated that the dehydration of alcohol 88 by passing it through a gas Chromatograph had been carried out under conditions where, if Z4'were formed, it would have been con- verted to 28; Consequently the alcohol (88) was dehydrated on a gas Chromatograph at a temperature where Z4 was known 100 to be stable -—-Z4'was recovered without rearrangement to 18 at 150°; alcohol 88'Was dehydrated at 1400. Even under these mild conditions cyclopentadiene zg'was obtained from the dehydration of 88; Dehydration of 88'with pyridine- pyridine hydrochloride, with iodine, and by column chroma- tography on basic alumina also gave 18. Dehydration of 88 to 18 can be rationalized by the fol- lowing mechanism. We 0 —> " 74—H+ 75—H+ + + -H -H 74 75 M m It seems that 74-~H+ opens up to 75--H+ faster than it loses a proton from a C2- or C4-methyl (which would give Z4). The reverse of this step has been shown by de Vries (55) who found that 81'solvolyzed with first order kinetics to give Z8'in pyridine; in acetic acid 97-OAc was obtained. The formation of 97—OAc when 81 is solvolyzed in acetic acid indicates that ion 88'(an ion in which formation of I ' “" liaq:u»€-fl'fl5¢_ii:-fl‘l‘u4‘ ’wrfin: ‘7'! l 'l 101 I the cyclopropane ring is not complete) is formed in the solvolysis. In pyridine cyclOpropane ring formation is completed and loss of a proton from a C2- or C4-methyl of 88 gives 28. H H OAc / HOAC 5 7 > 5 / 5 78 OBs QZKQAE. "' 2 _H q i " + . 5 91 F H2 i I yriding ' > Q ' 98 ‘ 9.9. g In this case the ring-opened carbonium ion (88) is primary whereas ZEZEf is a tertiary carbonium ion. The equilibra- tion of Z8 and 21 in acid (attained starting from either 28 or 22) indicates that ring opening to the secondary car- bonium ion (188) is not as facile as the ring opening of 24:8) to 18:8) (54). Both Z8 and 21 are converted to 88 by acid under more severe conditions, presumably gig 180 (54). 102 + H _— + ‘iil!" -H > 100 82' C. Mechanism of the Thermal Rearrangement of a Homofulvene (14) to a Cyclopentadiene (18) The acid—catalyzed rearrangement of homofulvenes to cyclopentadienes has been reported (53,54). Compounds 18 and zz'rearrange to 88'when treated with acid (53). vThe probable mechanism for this reaction was discussed above. H CH3 H+ \ + . —H 76, exo-CH3 9,9, ZZ, endo-CH3 Photochemically the homofulvenes 101-a and 101-b were converted to the aromatic compounds 102-a and 102—b 103 ! respectively whereas when 101-a, 101-b, and 101-c were heated in the absence of solvent, there was no change until polymerization occurred (52). R R \\ H L R' R' 101—a 102-a 101-b 102—b 101-c m w | - R' = CH3 by R = CH3, R = H R' = H 0 23 II The conversion of 101-a and 101-b to 102-a and 102—b is postulated to proceed by a suprafacial sigmatrOpic [1,7] hydrogen shift (52) which is photochemically allowed (62). The conversion of homofulvene 24 to cyclopentadiene 28 was shown to be a thermal rearrangement. Homofulvene 24'was injected into a gas Chromatograph (injector and detector temperatures 160°, column temperature 125°) and was recovered unchanged. When the injector temperature was increased to 210° (all other conditions were the same) and 24 was injected into the gas Chromatograph, the collected material was shown to be 28, The rearrangement therefore must have taken place in the injector port and not on the column. 104 74 75 Mechanistically the reaction 24 -—> 28 can be considered to arise from either C1-C6 or C5-C6 bond cleavage. Cleavage of the C5—C6 bond of 28'yig a thermally allowed sigmatropic [1,5] hydrogen shift (62) (path A) or 323 diradical inter- mediate 288 (path B) would give 224; Cleavage of the C1-C6 bond of 28'yig a [1,5] sigmatropic hydrogen shift to give 228 (which then goes to 288) path C) or gig diradical inter- mediate 107 would give 106 (path D). W W 1,5 IV (6:753? O (A) 79 104 > 104 (B) 105 * ‘X- 105 106 CH2 ' O 79 > ——————> 106 (D) rw iCI’C6; a W *- 107 Homofulvene 28) obtained from the photolysis of triene 22x was pyrolyzed at 200° in a gas Chromatograph. The nmr spectrum of the collected product showed that the peaks for the allylic methyls at T 8.27 and 8.45 now integrated for only three protons each, and were still broadened by homoallylic coupling, consistent with structure 228'(see Figure 13). When the homofulvene was similarly pyrolyzed at 275° extensive label scrambling occurred. See Figure 15. Further evidence for C1-C6 bond cleavage was obtained by pyrolysis of homofulvene 82 which was obtained by photol- ysis of triene 28; The nmr spectrum of the pyrolysis product (288) showed that the ratio of the peak for the allylic methyls at T 8.27 (C1-C4) to the peak for the allylic methyls at T 8.45 (C2-C3) was 4.1:6.0 (see Figure 14). This is also consistent with C1-C6 bond cleavage. ‘ ‘ ‘ ‘ fimamfln‘u'n.fi'~u&:x 934‘.”an I. 106 82 108 W From the available data no decision can be made as to whether path C or path D is followed for the thermal rear- rangement of 24'to 28, Cleavage of the C1-C5 bond rather than the C5-C6 bond is favored for either a diradical (path B or D) or a concerted mechanism (path A or C). Formation of the conjugated diradical 282’instead of the cross-con- jugated diradical 288 favors C1-C6 bond cleavage. Molecular models show that the hydrogens of the 22Q2_C5-methyl are very close to the C3-carbon but not to the gégfmethylene carbon. A sigmatropic hydrogen shift would therefore pre- sumably follow path C rather than path A. ET” “7 T “At-9" ji‘i :7 '0 1 I l EXPERIMENTAL A. Photochemical Rearrangement of 1—Methylene-2,3,4,4,5,6- hexamethyl-2,5-cyclohexadiene (£2) 1. Product Identification A solution of 1.00 g of 22 in 25 ml of ether was pre- pared (0.23 g, 5% 12 by weight). Six ml samples of this solution were placed in each of four quartz test tubes which were then stoppered with septum caps. Three of these samples were placed approximately one inch from a quartz cooling well which contained a 450 watt Hanovia lamp in— side a Vycor filter. The samples were maintained at 20° or below by means of a water bath. The fourth sample was kept in the dark. The reaction course was followed by vpc under the following conditions: 5' x 1/8" SE—30 column at 125°, 90 ml/min He flow, injector temp. 160°, detector temp. 150°. A new peak with a retention time of 1.9 min slowly increased in size as the photolysis proceeded (42' had a retention time of 6.6 min). A second new peak with a retention time of 1.1 min also appeared and slowly in- creased in size, apparently at the expense of the primary photoproduct. After 56 hr of irradiation the primary photo— product accounted for 45% of the volatile material and the 107 _— ~—__._. A—d- in h; jaunt—.9“; 108 photolysis was stopped. Polymeric material was found on the sides of the test tubes. Vpc examination of the sample which had been kept in the dark showed that no reaction had taken place. The irradiated solutions of 22 were combined and the major photoproduct was collected by preparative Vpc (same conditions as above) to give 157 mg of a pale yellow '3' .1 oil. This was collected again by preparative vpc and was identified as 4-methylene-1,2,3,5,6,6-hexamethylbicyclo— [3.1.0]hex—2-ene (24). Anal. Calcd for C13H20: C, 88.56; H, 11.44. 7.01-me x n ..-: A -.m x: 3.. n ‘0')"- . Found: C, 88.53; H, 11.45. 95%EtOH max The uv spectrum of 24 had A 253 nm (3 11,300) and the ir spectrum (neat) had bands at 3080 (vinyl C-H), 1625 (conj. c=c), and 850 cm-1 (terminal C=CH2). The nmr spectrum had peaks for the vinyl protons at T 5.32 and 5.47 (1H each), for allylic methyls, broadened by homo- allylic coupling, at T 8.34 (3H, Cz-methyl) and 8.44 (3H, C3-methyl), and singlets at T 8.89 (6H, C1- and C5-methyls), 8.97 (3H, EgngG—methyl), and 9.35 (3H, ggdngG-methyl). These spectral data are also presented in Table VI and the nmr spectrum is shown in Figure 9, and the ir spectrum is shown in Figure 16. 2. Sensitization A sample of 200 mg of 22 in 5 ml of acetone was irradi- ated through Pyrex with a 450 watt Hanovia lamp. After six hr of irradiation Vpc analysis (conditions above) 109 showed that no new products had been formed and no 22 had reacted. 3. Quenching and Solvent Effect Samples which contained 200 mg of 22 in 5 ml of ether, benzene, and methanol in quartz test tubes were attached to a quartz well which contained a 450 watt Hanovia lamp inside a Vycor filter. These samples were then irradiated and analyzed by vpc (5' x 1/8” SE-30 column at 125°, 90 ml/min He flow, injector temp. 155°, detector temp. 150°). The results are shown in Table VII. Samples which contained 100 mg of 22 in 5 ml of hexane, ether, and in 4.5 ml of ether plus 0.5 ml of piperylene were similarly irradiated and then analyzed by vpc (5' x 1/4" SE-30 column at 125°, 100 ml/min He flow, injector temp. 150°, detector temp. 160°). The results are reported in Table VII. In all of these cases polymer formed on the walls of the test tubes. 4. Effect of Different Wavelength Light A 200 mg sample of 22 in 5 ml of ether was irradiated as above through a Corex filter (no light transmitted below 260 nm) for 2 hr. Vpc analysis, as above, showed that no new products were formed. The Corex filter was replaced by a Vycor filter (75% transmission of light at 260 nm) and the sample was irradiated for two hr. Vpc analysis showed that 10% of the volatile material was 24, E"mnu"fi‘w‘. l, mun . s..::m.. .- Ann-...? I A. \ n 110 A dilute solution of 22 in ether (0.2% by weight) in a quartz vessel was irradiated at 254 nm with a Rayonet Photochemical Reactor and the reaction was monitored by vpc, as above. After 5 hr of irradiation, vpc analysis of the solution showed a new peak. Continued photolysis caused the formation of several new products according to vpc ) analysis. The ether was evaporated and the crude product was analyzed by TLC (silica gel, hexane eluant) which also showed that several materials were present, including start- ing material. ‘ _ ‘— ‘ Timmy-4;. 1“.“ - u; "‘1 mrqur .‘r‘ . ”I B. Photochemical Rearrangement of 1-Methylene—3,5-bis(tri- deuteromethyl)—2,4,4,6-tetramethyl-2,5-cyclohexadiene (44) Previous work in this laboratory indicated that unless the Grignard reaction of dienone 88 (21) with methylmagnes- ium halide was carried out at -20°, in the preparation of 44xdeuterium label was lost. In the present work it was found that the Grignard reaction could be carried out at ice—bath temperature without appreciable loss of the deuter- ium label in 44 if the reaction was carefully hydrolyzed with ice-cold NH4Cl solution. The nmr spectrum of triene 44'prepared in Ufis manner showed 6.1 protons at T 8.21 for the allylic methyls, i.e., 0.1 protons for the C3— and C5- methyls. Four quartz test tubes, each of which contained 150 mg of 44) 4.5 ml of ether, and 0.5 ml of piperylene, were at- tached to a quartz well which contained a 450 watt Hanovia 111 lamp inside a Vycor filter. These samples were irradiated for 29 hr. Vpc analysis of the photolysis solution showed that 38% of 44 was unreacted (5' x 1/4" SE—30 column at 125°, 100 ml/min He flow, injector and detector temp. 150°). The major photoproduct (28) was collected by preparative Vpc (conditions above) and then was further purified by gas chromatographing it again. The nmr spectrum of 28 (Figure 10) showed the vinyl protons at T 5.26 and 5.41 (1.0H each) and the 2292 and Egg Cg-methyls at T 9.33 (2.9H) and 8.95 (3.1H) respectively. The peak for the C1- and C5-methyls at T 8.87 now integrated for 3.0 protons, the peak at T 8.40 for the C3-methyl (3.0H) was now a sharp singlet, and the peak for the Cz-methyl was absent. The ir spectrum (neat ) of 28 differed from that of 24 in that it contained bands for C—D stretching at 2050-2225 cm.1 and had minor differences in the fingerprint region (the ir spectrum of 28 is also shown in Figure 17). C. Photochemical Rearrangement of 1—Dideuteromethylene- 2 , 3 , 4 , 4 , 5 , 6-hexamethyl-2 , 5-cyclohexadiene (28) Triene 28'was prepared as in part H—1 of the Experi- mental section in Part I of this thesis except that the Grignard reaction was hydrolyzed with 5 ml of D20 which contained 5 drops of D2804. The nmr spectrum of 28 thus prepared showed only a minute peak for the vinyl hydrogens at T 5.26. 112 I The irradiation of 28'was carried out as in part B, ‘ above, and the photoproduct (82) was collected by prepara- tive vpc. The nmr spectrum of 82’(Figure 11) was identical to that of 24 except that now only a trace of the peaks for the vinyl hydrogens at T 5.32 and 5.47 were present. The ir spectrum (neat) of 82 showed that the band for vinyl i- C-H stretching at 3080 cm-1 was absent. In addition the band for the conjugated double bonds was at 1590 cm-1 and the band for the terminal C=CH2 at 850 cm"1 was absent (this peak presumably shifted out of the range of the Spec- trometer; the ir spectrum of 82 is shown in Figure 18). I 1 ' “ VIM“ "rmzuun'd :41. ‘mflmgo. "xv D. Preparation of 5—Isopropenylpentamethylcyclopentadiene (75) _g. Hexamethylbicyclo[3.1.0]hex-3-en-2-one (88) (28), 0.80 g, was added to the Grignard reagent prepared from 0.22 g of magnesium and 1.42 g of methyl iodide. The re- action mixture was stirred and heated at reflux for 2 hr, then was cooled in an ice bath and was hydrolyzed with 10% NH4Cl. The organic layer was washed with water, dried over M9804, and the ether was evaporated. The crude product was then analyzed by vpc under the following conditions: 5' x 1/4" FFAP column at 170°, 90 ml/min He flow, injector temp. 260°, detector temp. 250°. A small peak due to 88 and one major peak were found. The product was collected as a nearly colorless oil (0.56 g) and was identified as 5-isopropenylpentamethylcyclopentadiene (28). The mass 113 Spectrum of 28 had a parent peak at m/e 176, consistent with the molecular formula, C13H20. The uv Spectrum (neat) had bands at 3080 (vinyl C-H), 1640 (conj. c=c), and 885 cm- (term. c=CH2). The nmr spectrum of 28 had a distorted quartet at T 5.25 (2H, vinyl protons) allylically coupled to a triplet at T 8.93 (3H), J 211.0 cps. Two quartets at T 8.27 and 8.45 (6H each, J = 0.7 cps) were assigned to the C1- and C4-methyls and to the C2- and C3—methyls re- Spectively and a singlet at T 9.03 (3H) was due to the C5- methyl. The spectral data for 28 are also reported in Table VIII. The nmr spectrum of 28 is shown in Figure 22 and the ir spectrum in Figure 19. E. Maleic Anhydride Adduct of 28 A solution of 0.80 g of 28 and 0.46 g of maleic an- hydride in 15 ml of ether was heated at reflux overnight. The ether was then evaporated from the reaction mixture and the crystalline adduct (84) was recrystallized from petro— leum ether (30-600) to give 0.91 g of fine white needles. An analytical sample had a mp 183.5—185°. 532;. Calcd for C17H2203: C, 74.42; H, 8.08. Found : C, 74.60; H, 8.20. The ir spectrum of 84 had strong bands at 1850 and 1770 cm.1 (anhydride carbonyl). The nmr spectrum (CDC13) had a broadened quartet (J = 1.1 cps) at T 5.10 (1H) and a broader peak at T 5.57 (1H) for the vinyl protons and a broadened 3 proton doublet (J = 1.1 cps) at T 8.35 for the 114 methyl of the isopropenyl group. The remainder of the spec- trum had singlets at T 6.78 (2H, protons a to the anhydride moiety), 8.41 (6H, allylic methyls), 8.63 (6H, bridgehead methyls), and at 9.06 (3H, C7-methyl, syn to the anhydride moiety). F. Tetracyanoethylene Adduct of 75 w A solution of 0.59 g of Zé'in 10 ml of tetrahydro- furan (THF) was added to a solution of 0.37 g of tetracyano— ethylene in 10 ml of THF. The reaction mixture was heated at reflux for 2 hr, then was cooled, and the THF was evapor— ated at reduced pressure. The adduct (93) was recrystal- lized from isopropyl alcohol and then from ether-petroleum ether to give 0.33 g of tan crystals, mp 154-50. éfléln Calcd for C19H20N4: C, 74.97; H, 6.62; N, 18.41. Found: c, 75.07; H, 6.53; N, 18.30. The ir spectrum of 2§’(CHC13) had bands at 1630 and 1 1 1660 cm_ (c=C) and at 2250 cm- (CEN). The nmr spectrum (CDC13) had broad peaks for the vinyl protons at T 4.95 and 5.53 (1H each) and a broad 3 proton doublet (J = 1.1 cps) at 8.32 for the allylic methyl of the isopropenyl group. The remainder of the spectrum had singlets at T 8.17 (6H, allylic methyls), 8.40 (6H, bridgehead methyls), and at 8.58 (3H, C7-methyl, syn to the tetracyanoethylene moiety). 115 G. Preparation of 5-Isopropylpentamethylcyclopentadiene (89) 5-lg2propenylpentamethylcyclopentadiene (72) was hydro- genated according to the method of Brown and Brown (67). A three-necked 300-ml flask was equipped with a magnetic stirrer, a nitrogen inlet, a septum cap, and a three-way stopcock, one side open to the air and the other side sealed with a balloon as a means of monitoring the pressure. To the flask there was added 1 ml of 0.20 g BthCle solution and 40 ml of absolute ethanol. The system was flushed thoroughly with nitrogen and then 5 ml of 1.0 fl_NaBH4 solu- tion in ethanol was added. This and the remaining addi— tions were made by injecting the material into the flask gig the septum cap. Four ml of 6 g HCl was then added to provide a hydrogen atmosphere. The excess pressure was released and then 2.0 g of 25 (11.4 mmole) in 5 ml of ethanol was added. A NaBH4 solution (lflu 3 ml = 12 mmole Hz) was then added at such a rate as to maintain approximately atmOSpheric pressure in the reaction flask. After the ad- dition was completed the reaction mixture was stirred un- til the hydrogen had been consumed (approximately 30 min). The platinum was filtered from the solution and most of the solvent was evaporated at reduced pressure. The residue was taken up in hexane and washed with water to remove the remaining ethanol. The organic layer was dried over MgSO4, and the hexane was evaporated. The crude product was ex- amined by vpc (10' x 1/4" Apiezon-L column at 145°, 85 ml/min He flow). A peak with a retention time of 2.8 min was shown 116 to be due to zg'by comparison with the retention time of an authentic sample. The second peak (retention time 3.6 min) was collected by preparative vpc as a pale yellow oil (340 mg) and identified as S-isgpropylpentamethylcyclo- pentadiene (82). The ir spectrum of 82 (neat) had a c=c band at 1650 cm-1. The absence of bands at 3080 and 885 cm- indicated that the terminal double bond of 25 had been re— 95%Et0H max 260 nm (e 3620). duced. The uv Spectrum had A The nmr spectrum of 82 is reported in Table VIII. A crystalline material which separated from the crude product mixture was recrystallized from aqueous ethanol and shown to be hexamethylbenzene by its nmr Spectrum which had only a singlet at T 7.81. This was probably formed by the acid catalyzed rearrangement of 75 and accounts for the low yield of g2, H. Tetracyanoethylene Adduct of 89 To 200 mg of 82 in 10 ml of THF there was added 141 mg of tetracyanoethylene, which caused the solution to turn brown. The brown color disappeared almost immediately. The reaction mixture was stirred at room temperature for 30 min and then the THF was evaporated at reduced pressure. The brown crystals were recrystallized twice from ether- petroleum ether to give 219 mg of adduct 22/ mp 142-30, turning brown at 125°. Anal. Calcd for C19H22N4: C, 74.48; H, 7.24; N, 18.29. Found : C, 74.36; H, 7.26; N, 18.30. 1 117 The ir spectrum of 9§'(CHC13) showed weak bands at 2270 (CEN) and 1665 cm'1 (c=c). The nmr Spectrum (CDCl3) had singlets at T 8.05 (6H, allylic methyls), 8.35 (6H, bridgehead methyls), and at 8.73 (3H, C7-methyl, gyg_to the tetracyanoethylene moiety). The doublet for the iso- propyl methyl groups appeared at T 9.09 and 9.20 (6H), but the methine proton of the isopropyl group was not detected. I. Thermal Rearrangement of 24 to 25’ A sample of 24 was injected into a vpc with the follow— ing conditions: 5' x 1/8" SE-30 column at 125°, 80 ml/min He flow, injector temp. 160°, detector temp. 160°. The material was collected and shown to be unchanged 14 by its ir spectrum which was identical to that of authentic 14. The injector temperature was increased to 210° and a sample of 24 was injected into the vpc. The ir spectrum of the collected material was identical to that of zg'prepared in part D, above. No other materials were observed. -The re— tention time of both zg’and zg'was 2.4 min under these vpc conditions. Before 14 had been successfully isolated, the major photoproduct from the photolysis of 17 was collected by preparative vpc with the following conditions: 10' x 1/4" Apiezon—L column at 175°, injector and detector temp. 250° The collected material had ir and nmr spectra identical to those of 25 prepared in part D, above. The ir and nmr spec- tra of the tetracyanoethylene adduct of this material were identical to those of gg'prepared from zg'as in part F, above. 118 l J. Thermal Rearrangement of 4—Methylene-1,2-bis(trideutero- methyl)-3,5,6,6—tetramethylbicyclo[3.1.0]hex42-ene (12): The labeled bicyclic diene (12) prepared in part B, above, was injected into a vpc with the following conditions: 5' x 1/4” SE—30 column at 155°, 60 ml/min He flow, injector temp. 200°. The collected material was shown to be Seigg- propenyl-l,2—bisjtrideuteromethyl)-3,4,5-trimethylcyclo- a pentadiene (126). The nmr spectrum of 122 (Figure 13) had the peak for the vinyl protons at T 5.25 (1.7H) and the peaks for the isopropenyl methyl at T 8.93 (3.0H, triplet) g and for the C5-methyl at T 9.03 (2.8H). The peak for the ="T C1- and C4-methyls at T 8.27 integrated for 3.2 protons and i the peak for the C2- and C3—methyls at T 8.45 integrated for 3.2 protons (these peaks were still broadened by homo- allylic coupling). Peaks due to impurities were also pres- ent. The ir spectrum of 126'(neat) had C-D stretching bands at 2100—2250 cm”1 and small differences in the fingerprint region compared to 22, The ir spectrum is shown in Figure 20. Labeled bicyclic diene 22 was also pyrolyzed with the injector temperature at 275°. At this temperature extensive 1In previous work in this laboratory (by R. M. Lange) the labeled cyclopentadiene 106 was incorrectly identified as zg'after the major photop?gduct of 44 was collected by pre— parative vpc under unspecified conditions. The integration of the nmr spectrum in this case gave better values for structure 106. The values were 2.09 protons for the vinyl group (T 5723), 2.95 protons for the isopropenyl methyl (T 8.93), 3.10 protons for the C1- and C4-methyls (T 8.27), 3.05 protons for the C2- and C3-methyls (T 8.45), and 3.00 protons for the C5-methyl (T 9.03). 119 label scrambling must have occurred since no meaningful integration could be obtained from the nmr spectrum of the collected cyc10pentadiene. See Figure 15. K. Thermal Rearrangement of 4-Dideuteromethylene—1,2,3,5,6,6- hexamethylbicyclo[3.1.0]hex-2-ene (82) Labeled bicyclic diene gz'was pyrolyzed using the condi— tions in part J, above, and the collected product was shown to be 5-ig2propenyl—1-dideuteromethyl-2,3,4,5-tetramethyl- cyclopentadiene (128). The nmr spectrum of 128 (Figure 14) had peaks at T 5.25 (1.9H) for the vinyl protons, at T 8.93 (3.1H) for the Eggpropenyl methyl, and at T 9.03 (3.1H) for the Cb-methyl. The peak for the C1- and C4-methyls at T 8.27 integrated for 4.1 protons and the peak at T 8.45 for the C2- and C3—methyls integrated for 6.0 protons. The ir spectrum of 12§'(neat) had weak bands at 2150-2250 cm-1 (C-D stretching) and small differences in the fingerprint region compared to 12: The ir spectrum is shown in Figure 21. L. Attempted Preparation of 4-Methylene—1,2,3,5,6,6-hexa— methylbicyclo[3.1.0]hex—2—ene (74) 1. Wittig Reaction Methylene triphenylphosphorane was prepared according to the method of Corey and coworkers (68). The Wittig reagent prepared in this manner was found to react readily with the relatively unreactive ketone camphor. Accordingly 0.82 g of a 52.8% suspension of NaH (0.018 mol) in mineral R" I“ .-: I II at 54‘ .‘Afiinfi "5.1.734 'a - ”\‘11 120 T oil was placed in a three—necked 100-ml flask and the mineral oil was removed by repeated washing with pentane. The flask was equipped with a magnetic stirrer, reflux condenser with a three—way stopcock at the top, and a septum cap. The sys- tem was then alternately evacuated and flushed with nitro- gen to provide a nitrogen atmosphere. Ten ml of dimethyl sulfoxide (DMSO) (distilled from CaHz) was injected into the flask and the mixture was heated at 75-800 for one hr. The reaction mixture was then cooled in an ice bath and 6.43 g of methyltriphenylphosphonium bromide (0.018 mol) in 20 ml L of warm DMSO was injected into the flask. The resultant . dark orange reaction mixture,was stirred for 10 min at room temperature and then 2.57 g of hexamethylbicyclo[3.1.0]hex- 3-en-2-one (62) (28) in 5 ml of DMSO was injected into the flask. The reaction mixture was heated at 56—600 for 18 hr and then cooled and poured into 30 ml of water. The two phases which resulted were each extracted with three 30-m1 portions of pentane and the combined pentane fractions were washed with 70 ml of 1:1 DMSO-H20 and then with 70 ml of 50% saturated NaCl solution. The organic layer was dried over M9804 and the pentane was evaporated to give 2.23 g of crude product. The ir spectrum of this material Showed it to be unreacted Q0. 121 2. Dehydration of the Alcohol (88) Obtained from Bi- cyclic Ketone 82'and Methylmagnesium Halide To the Grignard reagent prepared from 0.24 g of magnes- ium and 2.13 g of methyl iodide in 60 ml of ether there was added 0.89 g of hexamethylbicyclo[3.1.0]hex-3-en-2-one (82) in 25 ml of ether. The reaction mixture was stirred and heated at reflux for 45 min, cooled in an ice bath, and then hydrolyzed with 100 ml of ether saturated with water. The organic layer was decanted from the precipitated magnesium salts, dried over MgSO4, and the ether was evaporated to . 4 _ snug-mm “at.“ r... P . u give 0.90 g of a yellow oil. The ir spectrum of this mater- ial (neat) had bands at 3650 and 3500 cm'1 (hydroxyl) which indicated that it was the expected alcohol, 4—hydroxy- 1,2,3,4,5,6,64mptamethylbicyclo[3.1.0]hex-2-ene (88). A 400 mg sample of 88'was dissolved in 3 ml of pyridine to which there was added 2 ml of pyridine which contained 3 drops of conc HCl. The solution was then heated at 65° for 12 hr after which most of the pyridine was evaporated at reduced pressure. The remaining material was dissolved in ether to remove pyridine hydrochloride. The remaining sol- vent (pyridine and ether) was then evaporated to give an oily residue. The ir Spectrum of this material (neat) showed that it was 5-isopropenylpentamethylcyclopentadiene (Z8) and not 13. The remaining crude alcohol was column chromatographed on 20 g of basic Woelm Alumina with hexane as eluant. The hexane was evaporated from the chromatographed material 122 and the ir spectrum of the residue (neat) showed that it was 18. I In another experiment alcohol 88 was prepared as above from methylmagnesium iodide and 1.00 g of 88. The crude alcohol was dissolved in 15 ml of ether and an iodine crys- tal was added to the solution which was then stirred at room temperature for two hr. The reaction mixture was washed with dilute sodium thiosulfate solution, dried over M9804, and the ether was evaporated. The ir Spectrum of the re- covered material (neat) showed that it contained a mixture of 88'and 28; The procedure was repeated without solvent. The ir spectrum of the recovered material (neat) now showed only zg'and no 22; In another experiment alcohol 88'was dehydrated on a vpc with conditions such that if 28 were formed, it would not rearrange thermally (5' x 1/4" SE—30 column at 120°. 60 ml/min He flow, injector temp. 140°, detector temp. 150°). The ir Spectrum of the collected product (neat) showed that it was 18' 341...! .TJ‘hL-flmi.‘ 9 C I L. ' 4 "__‘-1V. SPECTRA Nflngpectra The nmr spectra presented here were obtained using CCl4 solutions. Ir Spectra f L_ The ir spectra presented here were obtained from liquid films on NaCl plates. 123 124 .Awmc memnm wanmo.fl.«HOHU%UAQHm£ummemQI®.®.m.m.N.Hlmcwamnumfilv mo Esnuommm mzz .m wusmflm is. 0 o h o n. V _q_g____Ae_e__._q«_q__;quIv mo ESuuummm HH .mH munmflm 0 00m 000. 003 003 000— com. OOON oonw 000m 8mm 131 132 1.. Po $511....) (M4545). 1%. . ...! IMF I T .1 ..QMV mamumuxmio.afloaomfiflsfimg umuumuuo m m mnflamnumaoumusmoeupvmenum.fiumcmamnumauv mo asuuommm HH .SH muswfim 00m 000— 00m. 003 000. com. 000m oonm 000m 000m .-.; r-_,,r,4 ._,__T . .0 a 9 T).Ilo Io .flIllolllllo. 1'. ‘10))!‘ll1Tl l T v--.- . A A . \ ‘Tx l O 1 .0 It _ A 00 om ..... lll‘lllll'l‘ll"! Ila); ll) 1 v 13 134 15 w- .. o. to u. . ......... . 4...-.-.. . . ....... . ......o.. 33', 03>. .-9-- . ... ..... . ecotcvcol . ....»oc-u o v--‘-vbnu . o. u p 00 p . . .... .—.. 200 WEXELENGTH HICROIS 2 ...-....A. < tour->0,- - u».>u-... . ....n ... . ......... u .....v... . ......-.. . ....~:ou. i wavenumber UIICIH SP. 1 650 nubfinnolfl 1200 8 00 00 3000 2 1300 5000 (82). Ir spectrum of 4-dideuteromethylene-1,2,3,5,6,6-hexamethylbicyclo- [3.1.0]hex-2-ene 18. Figure 077 r I--- e 71" “mm can”: 134 _ _ a F‘T .... ALE-(I .... .r.’~ .. .0... {fiatlhmj'rn .Ammv mcmflpmucwmoHUmoawgumamucmmaMCmQOHQOmAIm mo Ednuommm HH .mH musmflm com 000— 009 8! ooo. oom— ooo~ can u ooom comb 0 I. . ‘ . (OI. I . . . . A . . e e . a e . . . o . . . . , . _ . — _ .. m , . W I . . g 7 om ”-2..--e. .. ... 1 ..A . . . 4 o . c h r A . . . _ . .. a u r . v H . A . . . f e «.1 . e 4 o . .HI - .. 5 . _ _ _ , , . . . _, . . _ . . . 4 . . _ oo _ -- .1 r . A . . . , . . . . _ . . _ . . . . w 5 Q A .6 t l I . J. All I) .o e o . u n o n . . . . . _ . . . P l w H . . _ . . . . . r . . . . _ . . . , . . . . . . . - . . . 1 h - T k v . -.-) . x , . . r . . . . . ., ._ . v . . . . . . T'llilltl‘ 4 A )r 4 e . _ a . r 1 . a I . _ . 4 . _ . . 1 . . . . . . . H . . Ill) ¢ 6 Q A e > 4 . o l .v L ) l)..-‘ t 0 a . . . . . . . . . . . . . H . w . . . , . . , ‘ e . . . . , H .YltolL ‘ gr .I . P 4 14'. . . . . 4 a r . . u . . . . . 8— p . . . . . . r H . r 4 _ i 4 4 1 1 4 4 d 1 1) 4 1 nII'illl l' I |‘ II II I'll III III!) 11"“) 135 “.... y. . ..n . . ....u )5? ./GCHV mcmflomucwmoaomoamnuma Ifluulm.w.mlflahnumfioumusmpfluuvmflQIN.HIchwmoudomfllm mo Esuuowmm HH .om musmflm com 000— oom_ oov_ ooo_ 00m. oooN CV .....- ..---.-- a —-—. -8H-H._._x'._..------ 0v oo om LonESCQ>m3 OOOn 000v 000m 0 a . a." int-3.3 l r 11. r r -- r m l w l r k u I; n r r H r. T r M <1 --- H __1 V > > All}. .7, .x|.. I!||..V, II -1 v n #265... Evictij 0 9 R 8 aauengwsuen 9 0x. 136 .Mwofiv mcmflpmucmmoaowo nahLumEmuumulm.v.m.NIHwsumfioumusmpflplfiuamcmmou 0mfi|m m0 Ednuommm HH .Hm mnsmflm ... .. 2...... . 503E353; Ono 000 009 00m. 00! 009 009 000m 000m 009. 000... con .3 13:3. 9 OF 19 J B u 5 0m 0mm M. 19 2 0n 0mm a 0v 9. 0m 0m 00 00 0k 0k 0m 00 0m _ 0m 00—... i ) -.-: fl . , m 001 m:& 0 m. e 9 m m. k e -m., m n WLS§Fflfimzfitlm SUMMARY - Irradiation of 1-methylened2,3,4,4,5,6-hexamethyl-2,5- cyclohexadiene (ll) produced 4-methylene-1,2,3,5,6,6- hexamethylbicyclo[3.1.0]hex—2—ene (22). Quenching and sensitization studies indicated that the photochemical rearrangement of 11 to 22 proceeds from the singlet excited state. By the use of 11 which contained CD3 groups the forma- tion of zg'was shown to proceed by a "bond-crossing" mechanism and not by methyl migration. Bicyclic diene 22 rearranged thermally to Seiggpropenyl' pentamethylcyclopentadiene (Z8) when passed through a gas Chromatograph with an injector temperature of 200°. By the use of zg'which contained CD3 groups it was shown that cleavage of the C1-C6 bond of Zé'takes place at 200°, although at higher temperatures (275°) the C5—C6 bond may also be cleaved. 74 75 137 138 Dehydration of 4-hydroxy-1,2,3,4,5,6,6-heptamethyl— bicyclo[3.1.0]hex-2-ene (88) by various methods produced Z8 rather than the bicyclic diene 28. _,__,..T e \ 88 75 HO -_v .. p.-.£d L"_ h.‘i“. 2T!" 1' H Loooqmowuswto 10. 11. 12. 13. 14. 15. 16. LITERATURE CITED Auwers, K. and G. Keil, Chem. Ber., 6, 1861 (1903). m Auwers, K. and G.-Keil, Chem. Ber., 88) 3902 (1903). Auwers, K., Chem. Ber., 88) 1697 (1905). Auwers, K., A333, 888/ 219 (1907). Auwers, K. and M. Hessenland, Ag33, 888) 273 (1907). Auwers, K. and A. Kbchkritz, Agg,, 888) 288 (1907). Auwers, K., Chem. Ber., 22) 588 (1911). Auwers, K. and W. Jfilicher, Chem. Ber., 88) 2167 (1922). Fuson, R. C. and T. G. Miller, J. Org. Chem., $1, 316 (1952). Newman, M. S. and J. A. Eberwein, J. Org. Chem., 88) 2516 (1964). Patel,, D. J. and D. I. Schuster, J. Am. Chem. Soc., 82) 184 (1967). Tse, R. L. and M. S. Newman, J. Org. Chem., 8;) 638 (1956) Bird, C. W. and R. C. Cookson, J. Org. Chem., 82) 441 (1959). rNewman, M. S. and R. M. Layton, J. Org. Chem., 88) 2338 (1948). -Newman, M. S. and L. L. Wood, Jr., J. Am. Chem. Soc., 8;) 6450 (1959). Newman, M. S., D. Pawellek, and S. Ramachandran, J. Am. Chem. Soc., 82) 995 (1962). 139 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 140 Newman, M. S. and F. Bayerlein, J. Org. Chem., 28, 2804 (1963). ’”‘ Clement, H., Bull. Soc. Chim. France, 8* 1011 (1938). Berezina, R. N., V. G. Shubin, and V. A. Koptyug, Journal of the D. I. Mendeleev All-Union Chemical Society, No. 3, 356 (1968). Humphlett, W. J. and C. R. Hauser, J. Am. Chem. Soc., 22, 3289 (1950). Hart, H. and D. W. Swatton, J. Am. Chem. Soc., 89, 1874 (1967). "” Doering, W. von E., M. Saunders, H. G. Boyton, H. W. Earhart, E. F. Wadley, W. R. Edwards, and G. Laber, Tetrahedron, 8) 178 (1958). "Sadtler Standard Spectra," Vol. 1A, Sadtler Research Laboratories, Philadelphia, N0. 318. Walling, C., "Free Radicals in Solution,“ John Wiley and Sons, Inc., New York, 1957, p. 469. Wheland, G. W., "Resonance in Organic Chemistry," John Wiley and Sons, Inc., New York, 1955, p. 88. Streitwieser, A., Jr., "Molecular Orbital Theory for Organic Chemists,9 John Wiley and Sons, Inc., New York, 1961, p. 44. Gray, A. C. Gripper and H. Hart, J. Am. Chem. Soc., 88) 2569 (1968). Hart, H., P. M. Collins, and A. J. Waring, J.-Am. Chem. Soc., 88) 1005 (1966). Fish, R. W., Ph.D. Thesis, Michigan State University, 1960, p. 89. Okamoto, Y. and H. C. Brown, J. Am. Chem. Soc., Z8) 1903 (1957). Shriner, R. L., R. C. Fuson, and D. Y. Curtin, "The Systematic Identification of Organic Compounds," Fourth Edition, John Wiley and Sons, Inc., New York, 1962, p. 57. "Handbook of Chemistry and Physics," 44th Edition, The Chemical Rubber Publishing Co., Cleveland, 1962, p. 1162. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 141 Zimmerman, H. E., in "Advances in Photochemistry," edited by W. A. Noyes, Jr., G. S. Hammond, and J. N. Pitts, Jr., Interscience Publishers, New York, Vol. 1, 1963, p. 183. Chapman, 0. L., in "Advances in Photochemistry," edited by W. A. Noyes, Jr., G. S. Hammond, and J. N. Pitts, Jr., Interscience Publishers, New York, Vol. 1, 1963, p. 323. Waring, A. J., in "Advances in Alicyclic Chemistry," edited by H. Hart and G. J. Karabatsos, Academic Press, New York, Vol. 1, 1966, p. 241. Kropp, P. J., in "Organic Photochemistry,” edited by O. L. Chapman, Marcel.Dekker, Inc., New York, Vol. 1, 1967, p. 1. Neckers, D. c., "Mechanistic Organic Photochemistry," Rheinhold Publishing Corporation, New York, 1967, p. 217. Barton, D. H. R. and G. Quinkert, J. Chem. Soc., 1 (1960); Proc. Chem. Soc., 197 (1958). Griffiths, J. and H. Hart, J. Am. Chem. Soc-, 82) 3297 (1968). Zimmerman, H. E., R. C. Hahn, H. Morrison, and M. C. Wani, J. Am. Chem. Soc., 81) 1138 (1965). Schuster, D. I. and D. J. Patel, J. Am. Chem. Soc., 81) 2515 (1965). Zimmerman, H. E., R. G. Lewis, J. J. McCullough, A. Padwa, S. W. Staley, and M. Semmelhack,.J. Am. Chem. Soc., 88, 1965 (1966). DeMayo, P., in "Advances in Organic Chemistry," edited by R. A. Raphael, E. C. Taylor, and H. Wynberg, Inter- science Publishers, Inc., New York, Vol. 2, 1960, p. 367. -DeMayo, P. and S. T. Reid, anrt. Rev., 18) 393 (1961). Chapman, 0. L., in "Advances in Photochemistry,” edited by W. A. Noyes, Jr., G. S. Hammond, and J. N. Pitts, Jr., Interscience Publishers, New York, Vol. 1, 1963, p. 323. Dauben, W. G. and W. T. Wipke, Pure and App. Chem., 8, 539 (1964). 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60. '61. 62. 63. 64. 142 Fonken, G. J., in "Organic Photochemistry,” edited by O. L Chapman, Marcel Dekker, Inc., New York, Vol. 1, 1967, p. 197. Barton, D. H. R. and A. S. Kende, J. Chem. Soc., 688 (1958). Barton, D. H. R., R. Bernasconi, and J. Klein, J. Chem. Soc., 511 (1960). Zimmerman, H. E. P. Hackett, D. F. Juers, and B. Schrbder, J. Am. Chem. Soc., 88) 5973 (1967). Hogeveen, H. and H. C. Volger, Chem. Commun., 1133 (1967). Rey, M., U. A. Huber, and A. S° Dreiding, Tetrahedron Letters, 3583 (1968). Shafer, W., and H. Hellmann, Angew. Chem. Internat. Ed., 8, 518 (1967). Criegee, R. and H. Grflner, Angew. Chem. Internat. Ed., 2, 467 (1968). DeVries, L., J. Am. Chem. Soc., 88) 5242 (1960). Paquette, L. A., Tetrahedron Letters, 2133 (1968). Paquette, L. A. and G. R. Krow, Tetrahedron Letters, 2139 (1968). Schonberg, A., "Preparative Organic Photochemistry," Springer-Verlag, New York, 1968, p. 559. Zimmerman, H. E., R. S. Givens, and R. M. Pagnie, 8; Am. Chem. Soc., 88) 6096 (1968). Reusch, W. and D. W. Frey, Tetrahedron Letters, 5193 (1967). Childs, R. F. and S. Winstein, J. Am. Chem. Soc., 88 (1968), in press. Hoffman, R. and R. B. Woodward, Accounts Chem. Res., L, 17 (1968). Reinheimer, H., H. Dietl, J. Moffat, D. Wolff, and D. M. Maitlis, J. Am. Chem. Soc., 82) 5321 (1968). DeVries, L., J. Org. Chem., 88) 1838 (1960). 65. 66. 67. 68. Skattebol, L., Tetrah McLean, S. and P. Hay Brown, H. C. and C. A 1495 (1962). Greenwald, R., M. Cha Chem., 88) 1128 (1963 143 edron, 88) 1107 (1967). nes, Tetrahedron, 88) 2343 (1965). . Brown, J. Am. Chem. Soc., 84’, ykovsky, and E. J. Corey, J. Org.