ABSTRACT PART I STRUCTURAL EFFECTS ON 2.4-CYCLOHEXADIENONE PHOTOCHEMISTRY PART II MECHANISM FOR ELECTROPHILIC ADDITION TO METHYLCYCLOPROPANE PART III THE PHOTOCHEMISTRY OF ARYLNITRILE OXIDES BY Richard J. Bastiani In the first part of this thesis the effect of ring constraints on 2,4-cyclohexadienone photochemistry was in— vestigated. Oxidation of tetramethylbenzocyclobutene with peroxy— trifluoroacetic acid-boron fluoride etherate produced 2,4,4,5-tetramethylbicyclo[4.2.0]octa-1,5-dien-3—one (l§)° Irradiation of lg’in methanol resulted in five photOpro— ducts, a_nonconjugated ester 3}.(55%) and a mixture of the O ' COzMe COZMe hv ' I + methanoI E6 Ex 15 31 32-35 w W WW (4 geometric isomers) geometric isomers §gf§§'(45%) of a conjugated ester. In Richard J. Bastiani contrast, irradiation of 2,4,4,5-tetramethylbicyclo[4.4.0]— deca-1,5-dien—3-one (12), prepared by the oxidation of tetramethyltetralin with peroxytrifluoroacetic acid—boron fluoride etherate, provided only tricyclic enone fig. 0 O ——>*“’ l.> methanol - ll 40 m The cis-diene ketene intermediate ég is unaffected by the cyclohexane ring, since photoisomerization occurs to give ¢0 ¢O /’ x’ 22. 22, enones analogous to those obtained from other hexaalkyldi- enones. Ketene Qfi'however, is prevented from similar cycli- zation by the geometry of the system. Furthermore, it is distorted by the cyclobutane ring in such a way as to allow facile nucleophilic attack on the carbonyl carbon. The second part of this thesis is concerned with the mechanism of electrophilic addition to methylcyclopropane. The mechanism of electrophilic addition to cyclopropane involves a protonated cyclopropane intermediate, which is 2 Richard J. Bastiani more stable than the primary propyl cation. To determine if the reaction of methylcyclopropane with electrophiles involves an analogous intermediate, additions to methyl- cyclOpropane-l-dl (lg) were studied. Compound Z§ was prepared by the cyclization of 1-bromo—3-chloro-Z—methyl- propane-2-d1 (22) with zinc. Acylation of zg'with acetyl chloride-aluminum chloride in methylene chloride produced 5-chloro—2—hexanone (g2) D o o AlCla W \ 7 AcCl Cl D Cl 78 80 83 m m NV which contained deuterium only at C-5. No §§J which is a predicted product from a protonated methylcyClopropane intermediate, was found. Reaction of zg‘with hydrogen bromide produced gg'with 98% deuterium at C-2. A protonated D HBr 3; D /' /\|/ Br 78 96 m AN methylcyclopropane or hydride transfer in the secondary butyl cation may explain the 2% hydrogen at C-2. The ad- dition of hydrogen chloride to zg'produced similar results. The photochemistry of arylnitrile oxides was the subject of the third part of this thesis. The generality of the 3 Richard J. Bastiani photochemical rearrangement of nitrile oxides to isocyanates and of the photochemical deoxygenation to nitriles was CEN—>O N =C =O CEN R1 R3 R1 R3 R1 R3 hv ; + R2 R2 R2 R2=Cl, R1=R3=H 144 149 1ng R1=R2=Cl, R3=H 145 154 155 R1=R2=R3=CH3 128 131 157 MW I'M established, as shown by the examples in the equation. The fate of the liberated oxygen atom has not yet been ascertained. The photoisomerization of nitrile oxides to isocyanates has been postulated to involve an oxazirine intermediate. Attempts to observe this type of intermediate Spectro- scopically by irradiating 2,4,6-trimethylbenzonitrile oxide (128) in 2-methyltetrahydrofuran at -1700 have thus far been unsuccessful. PART I. STRUCTURAL EFFECTS ON 2,4-CYCLOHEXADIENONE PHOTOCHEMISTRY PART II. MECHANISM FOR ELECTROPHILIC ADDITION TO METHYLCYCLOPROPANE PART III. THE PHOTOCHEMISTRY OF ARYLNITRILE OXIDES BY { . \ Richard J. Bastiani A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1970 To My Wife, Janice ii ACKNOWLEDGMENT The author wishes to express his appreciation to Professor Harold Hart for his enthusiasm, encouragement, and guidance throughout the course of this study. Appreciation is extended to Michigan State University for a Graduate Teaching Assistantship from September, 1967 to June, 1968 and from September, 1968 to March, 1969. Appreciation is also extended to the National Science Foundation for financial support from June, 1968 to August, 1968 and to the National Institute of Health for financial support from March, 1969 to September, 1970. iii TABLE OF CONTENTS PART I STRUCTURAL EFFECTS ON 2,4-CYCLOHEXADIENONE PHOTOCHEMISTRY Page INTRODUCTION . . . . . . . . . . . . . . . . . . . . . 1 RESULTS AND DISCUSSION . . . . . . . . . . . . . . . . 5 A. Preparation and Characterization of 2,4,4,5-Tetra- methylbicyclo[4.2.0]octa-1,5-dien-3—one (15) . . 5 B. Irradiation of 2,4,4,5-Tetramethylbicyclo[4.2.0]— octa-1,5-dien-3-one (15) . . . . . . . . . . . . 11 c. Oxidation of 5,6,7,8-Tetramethyl-1,2,3,4-tetra- hydronaphthalene (42) and Characterization of the Products . . . . . . . . . . . . . . . . . . 19 D. Irradiation of 2, 2, 4, 5-Tetramethylbicyclo[4. 4. O]— deca-1(6), 4-dien-3- -one (18), 3, 3, 4, 5— —Tetramethyl- bicyclo[4 .4.0]deca- -1(6), 4“aien- 2-one (16), 2, 4, 4, 5- Tetramethylbicyclo[4. 4. 0]deca- -1, 5 -d1en:§;one 17) and 3, 4, 5, 5-Tetramethylbicyclo[4. 4. 0]deca-1(6 :3? dien-Z-one (43) . . . . . . . . . . . . . . . . 25 EXPERIMENTAL . . .'. . . . . . . . . . . . . . . . . . 32 A. B. General Procedures . . . . . . . . . . . . . . . 32 Preparation of Tetramethylbenzocyclobutene (23). 32 Oxidation of Tetramethylbenzocyclobutene (23). . 33 Identification of the Products from the Oxidation of Tetramethylbenzocyclobutene (23) . . . . . . 34 Preparation of 2, 4, L 5-Tetramethy1bicyclo[4. 2. O]— octa-l, 5-dien-3- -one- 8-d2 (15-d ) . . . . . . . 35 Catalytic Reduction of 15: Preparation of 2, 4, 4 ,5-Tetramethylbicyclo[4. 2. O]octa- -1—ene- 3-0118 (28) o o o o o o o o o o o o o o o o o o o 35 iv TABLE OF CONTENTS (Continued) Page G. Preparation of 2, 4, 4 ,5-Tetramethylbicyclo[4. 2. O]— octa-l-ene-B-one— —6—d1-8-d2 (29) . . . . . . . 36 H. Irradiation of 2, 4, L 5-Tetramethylbicyclo[4. 2. 0]- octa-l, 5-dien-3- -one (15) . . . . . . . . . . 37 I. Identification of the Products from the Irradiation of (15) . . . . . . . . . . . . . . 37 J. Preparation of 5,6,7,8-Tetramethyl—1,2,3,4-tetra— hydronaphthalene (42) . . . . . . . . . . . . . 40 K. Oxidation of 5,6,7,8-Tetramethyl-1,2,3,4—tetra- hydronaphthalene (42) . . . . . . . . . . . . . 41 L. Identification of Products from the Oxidation Of Eta O O O O O O O O O O O O O O O I C O O O O 41 M. Preparation of 2, 2, 4, ,5—Tetramethylbicyclog4. .4 O]— deca-1(6), 4-dien-3—one 5-methyl—d3 (18-d . . . 43 N. Preparation of 3,3,4,5—Tetramethylbicyclo[4.4.0]— deca—1(6),4-dien-2-one—7-d2 (16-d2) . . . . . . 44 0. Preparation of 3, 4, 5 ,5-Tetramethylbicyclo[4 .4 O]- deca-1(6), 3-dien— 2-one— —4-methyl- d3- 7—d2 (43-d ). 44 P. Irradiation of 3,4,5,5-Tetramethylbicyclo[4.4.0]- deca-1(6),3-dien—3-one (42) . . . . . . . . . . 44 Q. Irradiation of 2,2,4,5-Tetramethylbicyclo[4.4.0]- deca-1(6),4-dien-3-one (18) . . . . . . . . . . 45 R. Irradiation of 3,3,4,5-Tetramethylbicyclo[4.4.0]- deca-1(6),4-dien—2-one (1g) . . . . . . . . . . 46 S. Irradiation of 2,4,4,5-Tetramethylbicyclo[4.4.0]- deca-1,5-dien-3—one (11) . . . . . . . . . . . . 47 SUMMARY . . . . . . . . . .'. . . . . . . . . . . . . 49 TABLE OF CONTENTS (Continued) Page PART II MECHANISM FOR ELECTROPHILIC ADDITION TO METHYLCYCLOPROPANE INTRODUCTION . . . . . . . . . . . . . . . . . . . . 52 RESULTS AND DISCUSSION . . . . . . . . . . . . . . . 59 A. Consequence of Deuterium Label . . . . . . . . 59 B. Preparation of Methylcyclopropane—l-d1 . . . . 61 C. Reaction of Methylcyclopropane—l-d1 with Electrophilic Reagents . . . . . . . . . . . . 63 1. Acylation of Acetyl Chloride in the Presence of Aluminum Chloride . . . . . . . . . . . 63 2. Reaction of Z§’with Hydrogen Bromide . . . 65 3. Reaction of zg'with Hydrogen Chloride . . 67 D. Recent Related Studies and Mechanistic Conclusions . . . . . . . . . . . . . . . . . 68 EXPERIMENTAL . . . . . . . . . . . . . . . . . . . . 73 A. Preparation of Methylcyclopropane-l-dl . . . . 73 1. Preparation of Deuterium Bromide . . . . . 73 2. Preparation of 1-Bromo-3-chloro—2—methyl- propane-Z-dl o o o o o o o o o o o o o o o 73 3. Preparation of Methylcyclopropane-l—dl (12) 74 B. Reaction of Methylcyclopropane—l-d1 with Hydrogen Bromide . . . . . . . . . . . . . . . . . . . 75 C. Reaction of Methylcyclopropane-l-d1 with Hydrogen Chloride O O I O O O O O O I O O O O O C O O O 76 D. Acylation of MethylcycloprOpane-l—d1 with Acetyl Chloride . . . . . . . . . . . . . . . . . . . 77 SUMMARY . . . . . . . . . . . . . . . . . . . . . . 80 vi TABLE OF CONTENTS (Continued) PHOTOCHEMISTRY OF ARYLNITRILE OXIDES INTRODUCTION . . . . PART III RESULTS AND DISCUSSION . . . . . . . . . . . . . A. B. C. Preparation of pfchlorobenzonitrile Oxide (144) Irradiation of prhlorobenzonitrile Oxide (144) Preparation and Irradiation of 2,4-Dichloro- benzonitrile Oxide (145) . . . . . . . . . . . . Preparation and Irradiation of 2,4,6—Trimethyl— benzonitrile Oxide (128) . . . . . . . . . EXPERIMENTAL . . . . A. B. H. Preparation of Preparation of (148) . . . . Irradiation of Preparation of (145) . . . . Irradiation of (145) . . . . Preparation of Oxide (128) . Irradiation of Oxide (128) . Irradiation of Oxide (128) at S UWY . O O O O . LITERATURE CITED . . APPENDIX . . . . . . EfChlorobenzonitrile Oxide (144) 3,4-Di(pfchlorophenyl)furoxan prhlorobenzonitrile Oxide (144) 2,4—Dichlorobenzonitrile Oxide 2,4-Dichlorobenzonitrile Oxide 2,4,6-Trimethylbenzonitrile 2,4,6—Trimethylbenzonitrile 2,4,6-Trimethylbenzonitrile Liquid Nitrogen Temperature vii Page 94 96 99 100 100 101 102 103 104 105 107 109 113 LIST OF TABLES TABLE Page I. Spectral data of dienone 15 and enone gfix . . . 9 II. Spectral data for the products obtained upon irradiation of dienone l2 . . . . . . . . . . 13 viii LIST OF FIGURES FIGURE Page 1. A mechanism for the oxidation of tetramethyl- benzocyclobutene . . . . . . . . . . . . . . . 7 2. A mechanism for the products obtained upon photolysis of dienone (’12) . . . . . . . . . . 16 3. A mechanism for the oxidation of 5,6,7,8-tetra- methyl-1,2,3,4-tetrahydronaphthalene (42) . . 24 4. Labelling consequences in the acylation of methylcyclopropane-l—d1 . . . . . . . . . . 6O 5. Nmr spectrum (CC14) of 2,4,4,5—tetramethyl- bicyclo[4.2.0]octa-1,5—dien-3-one (22) . . . . 113 6. Nmr spectrum (CC14) of 2,4,4,5-tetramethyl- bicyclo[4.2.0]octa—1-ene-3-one (22) . . . . . 114 7. Nmr Spectrum (CCl4) of 2,4,4,5-tetramethyl- bicyclo[4.2.O]octa-l-ene-3-one—d1-8-d2 (22) . 115 8. Nmr spectrum (CC14) of methyl 2-(1,2-dimethyl- propenyl)-a—methy1-1-cyclobutene—1-acetane (22) 116 9. Nmr spectrum (C014) of (Z,Z)-methyl 2-(1,2- dimethylpropylidene)-a-methyl-A1'a-cyclo— butane acetate (22) . . . . . . . . . . . . . 117 10. Nmr spectrum (CClg) of (Z,E)-methyl 2—(1,2- dimethylpropylidene)-d-methyl-A1'Q—cyclo- butane acetate (22) . . . . . . . . . . . . . 118 11. Nmr spectrum (CC14) of (E,E)-methyl 2-(1,2- dimethylpropylidene)-a—methyl-A1:O—Cyclo- butane acetate (22) . . . . . . . . . . . . . 119 12. Nmr spectrum (CClé) of (E,Z)-methyl 2—(1,2- dimethylpropylidene)-a-methyl-A1'Q-cyclo— butane acetate (22) .. . . . . . . . . . . . . 120 ix LIST OF FIGURES (Continued) FIGURE Page 13. Nmr spectrum (CCl4) of 2,2,4,5-tetrameth l— bicyclo[4.4.0]deca-1(6),4-dien-3-one (l2) . . 121 14. Nmr Spectrum (CCl4) of 3,4,5,5-tetrameth l- bicyclo[4.4.0]deca—1(6),3-dien-2—one (42' . . 122 15. Nmr spectrum (CC14) of 3,3,4,5-tetrameth l- bicyclo[4.4.0]deca-1(6),4—dien-2—one (£2 . . 123 16. Nmr spectrum (CCl4) of 3,4,4,5-tetramethyl- tricyclo[4.4.0.03:5]deca-1(6)-ene-2-one (44) . 124 17. Nmr spectrum (CCl4) of 8,9,10,10—tetramethyl- tricyclo[4.3.1.0]deca—8-ene-7-one (42) . . . . 125 18. Nmr spectrum (CC14) of 2,2,3,5—tetramethyltri— cyclo[4.4.0.01:3]deca—5-ene-4—one (42) . . . . 126 19. Nmr spectrum CCl4) of 1—bromo-1-Chloro-2-methyl- propane-2-d1 g2). . . . . . . . . . . . . . 127 20. Nmr spectrum (CCl4) of methylcyclopropane-l—dl (233.) 128 21. Mass spectrum of methylcyclOpropane-l—d1 (Z2). 129 22. Nmr Spectrum (CCl4) of 2—bromobutane—2-d1 (22) 130 23. Nmr spectrum (CC14) of 2-Chlorobutane-2—d1 (101) 131 24. Nmr spectrum (CCl4) of 5-chloro-2—hexanone-5-d1 (g2) 132 25. Nmr spectrum (CC14 of partially deuterated 2-chlorobutane (24' . . . . . . . . . . . . . 133 INTRODUCTION As shown by Griffiths and Hart (1) the only photo— chemical reaction of alkyl substituted 2,4—cyclohexadienones is ring opening to a ketene. The ketene may thermally re— convert to dienone, cyclize to a bicyclo[3.1.0]hexenone, or in the presence of nucleophile, react to form an acid deriva— tive. The particular path that the ketene follows is ap- parently a function of solvent and structural effects. Cyclization to a bicyclohexenone is favored by polar sol- vents (1) and may even occur in a strongly nucleophilic solvent such as methanol. The structural effects which influence the reaction path are more subtle. Completely alkylated 2,4—cyclohexa— dienones, such as the hexamethyl— or hexaethyl—cases yield only bicyclohexenones. Thus, irradiation of dienone L in ether or methanol gave only bicyclic enone 2; However, in O O éR R R th :/| R R RMeOH \\. R R R R R = Me or Et 1 ~ ([0 3w 2 the isomeric pentamethyl cases, two types of products were obtained. Irradiation of 4'in methanol yielded only bicyclo- hexenone 2'(2). hv :> MeOH 2. :2, In contrast, irradiation of 2 under identical condi- C02Me hv ¥:_ ' MeOH / \ 5 Z. tions resulted in the exclusive formation of the unsaturated ester Z'expected from a Cis-diene ketene intermediate (2), analogOus to ketene 2; As an intermediate case, isomer 2' gave a mixture of ester 2 and bicyclic enone 42 in a 1:5 300 ratio. 3 Less substituted 2,4-cyclohexadienones, such as the tetramethyldienone lloand trimethyldienone 42) gave only esters upon irradiation in methanol (3,4). O H hv MeOH H 1.1, 12 o hv 3; H H MeOH H 13 In order to further study the structural effects on the ,photochemistry of 2,4-cyclohexadienones, compounds 42) 42) ’21 and 42'were synthesized. In these highly substituted' (1% 5:5 B NV m 'uv rvv cases, a second ring system is fused to the dienone. Although all pueviously studied hexaalkyl-Z,4-cyclohexadienones s ’0’“ r. R . 7 ‘ .—" ~.V .v“'.- . , ' . ‘_--- '. ‘ ,,. .L. n a-." "“ ' 1 . a p.- .chH uv-q . ."" ~.. 0......i l H! {H .. fl. ‘ y i-u. _. 4 formed bicyclo[3.1.0]hexenones (1,9), the structural re— straints imposed by the fused ring system in enone 42'pro- hibits this reaction path. 0 > I.» 15 19 W NV Likewise, the more flexible, but nevertheless con— straining, cyclohexane ring in dienone lz'is expected to be an important influence in determining the reaction product. Thus, Part I of this thesis will examine the effects that a constraining ring system has upon the photochemistry of 2,4-cyclohexadienones. ‘ ~ V “ D‘O-C "'r-o D - F-.—‘ ‘ t ‘v.q ’ ' ill “on . \lv- .» to. h RESULTS AND DISCUSS ION A. Preparation and Characterization of 2,4,4,5-Tetramethyl— bicyclo[4.2.0]octa—1,5-dien-3-one (15) The preparation of dienone 42 was accomplished by the oxidation of tetramethylbenzocyclobutene 22’with peroxytri— fluoroacetic acid in the presence of boron trifluoride etherate. As the hydrocarbon 22 has previously been pre- pared by Hart and Hartlage (6), their synthetic route was followed with one modification. Pentamethylbenzene was C1 .C1 >+<—-> . 20 21 22 CC13 Cl 0 a 0 3 aq BF3'Et20 m” ’5 o 22 23 15 6 trichloromethylated with carbon tetrachloride in the presence of aluminum chloride to give trichloromethylpentamethyl- benzene 24, Heating 22'to 125O resulted in a cyclization to tetramethyldichlorobenzocyclobutene 22'in 90% yield. Reduction of 22'was accomplished according to the improved procedure of Hart (7), whereby treatment with lithium in liquid ammonia cleanly gave an 88% yield of tetramethyl- benzocyclobutene 22; The oxidation of 22 was performed ac— cording to the method of Hart and Gray (8). In order to minimize polymerization, it was necessary to lower the reac— tion temperature to -30°. Even at this temperature only a 34% yield of volatile product could be isolated. Analysis of the distilled product by vpc showed the presence of two major components. The compound composing 7% of the mixture was identified as hexamethyl-2,4-cyclo- hexadienone, which was formed from a small amount of hexa— methylbenzene present in crude 22; The component accounting for greater than 90% of the mixture was identified as 2,4,4,5-tetramethylbicyclo[4.2.0]octa—1,5-dien-3-one 42. While there are three possible isomers (42) 22 and 21) which can be formed in the oxidation of 22) only one of these was found. A mechanism rationalizing the formation of a single oxidation product is presented in Figure 1. Dienone lé'is formed by electrophilic attack of peroxytrifluoro- acetic acid at C-3, resulting in the Carbonium ion 222, A 1,2-shift of the methyl group to C-4 gives intermediate 222' and loss of a proton yields dienone 42; The reaction paths C-Z Attack C-3 Attack HO OH s \ I I I g \ 4," \4' a " v 24a 26a inCH3 25 27 15 FiSMLre 1. A mechanism for the oxidation of tetramethyl— benzocyclobutene (42) . 8 which lead to dienones 22 and 22 apparently are not competi— tive with the path which leads to 42, It may be that the cyclobutene ring which is present in 22'and 2Z'results in a more strained ring system than the dimethylenecylobutane system of 22. Although only one of the three possible isomers was formed in the oxidation, the spectral data of 42'(Table I) were not sufficiently characteristic to distinguish it from isomers 22'and.212 Consequently, several derivatives of 42 were made in order to unequivocably establish its structure. To distinguish between a methyl or methylene group at- tached to C-1, 42 was treated with 1M_sodium methoxide in methanol—§_at room temperature for one hour. The result was the exchange of two hydrogens with deuterium. The nmr spectrum of the labeled compound 42:22 was marked by a loss of two protons (by integration) of the resonance at T 7.14. Accordingly, the mass spectrum of 12:22 was marked by a gain of two amu in the molecular weight, indicating the presence of two deuterium atoms. Since alkyl groups at C—1 exchange much faster than those at C-5 (5), there must be a methylene attached to C-1. This result rules out 21, as it has a methyl group at this position. Structure 22'was eliminated by partial hydrogenation of 42'with hydrOgen absorbed on platinum. If the reduction is halted after one mole of hydrogen is absorbed, compound 22'is isolated. Spectral data of 22'(Table I) show it to be an a,B-unsaturated ketone, which has one allylic and one Table I 9 Spectral data of dienone 15,and enone 22. O 3’ O 15 28 Infrared Spectrum (CC14) ‘52 22’ _1 . . .. VC=O cm (conj.) 1685 1655 _1 . VC=C cm (con3.) 1630 UltraVIolet Spectrum (MeOH) xmax nm (e) xmax nm (e) 320 (4800) 247 (4800) Nuclear Magenetic Resonance Spectrum (CCl4) 1.5.. .231 T area 'assignment T area 'assignment 8.90 6H gem-dimethyl 9.14 3H C-5 methyl group group (d. QF7 HZ) 8.36 3H allylic methyl 9.0 3H group gem-dimethyl group 8.31 3H allylic methyl 8.83 3H group 7.14 4H cyclobutyl 8.45 3H allylic methyl protons group 7.8—8.3 3H C-7 and C-6 protons 7.1-7.5 2H C-8 protons 6.34 1H C—5 proton 10 aliphatic methyl group. The fact that there is still one allylic methyl group eliminates 22) since it would require that no allylic methyl groups be present upon reduction to an a,B-unsaturated ketone. It further supports the previous conclusion which eliminated 22) as this dienone would give a dihydro product with two allylic methyl groups. As further verification, deuterium exchange of 22're— sulted in two products, each of which contained three' deuterium atoms, as determined by their nmr and mass spectra. The structures of these two compounds were assigned as 935 22'and trans 22'isomers, since isomerization is expected in the base catalyzed exchange. 0 NaOMe MeOD 28 29 2,9, m Analysis of the nmr spectrum of 22 showed that the ‘broad allylic methyl signal at T 8.45'(g_= 0.9 Hz), which Ivas homoallylically coupled to the cyclobutyl protons, be- cane a sharp singlet. This fact thus verifies the position <3f the cyclobutane ring. 11 B. Irradiation of 2,4,4,5-TetramethylbicycloI4.2.0]octa— 1,5-dien-3-one (15) A 1% solution of 42'in methylene chloride was irradi— ated for eight hours through Pyrex with a 450 watt Hanovia lamp. The reaction was monitored by ultraviolet spectroscopy. After eight hours no net reaction had occurred, since the recovered material had ir and uv spectra identical to 42; However, upon irradiation under identical conditions but in methanol, lé'disappeared in two hours yielding five photo- products. The product ratio varied with the intensity of the light and the irradiation time. When the photolysis was performed using a Rayonet photochemical reactor with 3000 R lamps the major product constituted 55% of the reaction mix— ture. The remaining 45% contained the four minor products in nearly equal amounts. The major product was identified as methyl 2-(1,2- dimethylpropenyl)-a—methyl—1—cyclobutene—1-acetate 24; COzMe H l \\ 31 The compound was identified by analysis of its spectral properties. The uv spectrum of 24'had a A:::H at 239 nm (6 8100). This absorption, together with an ir spectrum 12 (CCl4) with bands at 1715 and 1630 cm-1, showed the presence of a nonconjugated carbonyl group. The nmr spectrum (Table II) showed an aliphatic methyl at T 8.80 (doublet, 3H), allylic methyls at T 8.30 (9H), cyclobutyl protons at T 7.61 (4H), a tertiary proton at T 6.73 (quartet, 1H), and the ester methyl group at T 6.40 (3H). These data distin- guished 22'from the other four products, since it is the only nonconjugated ester. The other photoproducts are geometric isomers. The COzMe COzMe O/ 0/ £32, .353, MeOZC MeOZC II .// ,‘II"// 24. .19. particular geometry of each isomer was assigned by examin- ation of the uv and nmr spectra of the series (Table II). It is well established that the magnitude of the ex— tinCtion coefficient of a conjugated system depends on the degree of overlap in the 1T electron system. Consequently, 13 Table II. Spectral data for the products obtained upon ir— radiation of dienone 42; MeOH nmr (CCl4) Compound Kmax (e) T area assignment C 02Me /’ 33 (W 02Me 8.46 3H propylidene methyl 8.21 3H a-methyl group 7.54 4H cyclobutyl protons ’/ 6.36 3H carbomethoxy protons 32 m 239 (8100) 8.80 3H d-methyl group CO Me 2 8.30 9H allylic methyl groups H I 7.61 4H cyclobutyl protons \\ 6.73 1H a-proton 6.40 3H carbomethoxy group 21, 289 (6700) 9.12 3H 9 01 3H gem—dimethyl group 289 (8100) 3:3; fig ggmfdimethyl group 8.63 3H propylidene methyl 8.20 3H a-methyl group 7.50 4H cyclobutyl protons 6.36 3H carbomethoxy protons 233 (17000) 3'33 33 gem-dimethyl grOUP MeOZC 8 .42 3H propylidene methyl | 8.12 3H a-methyl group /’ 6.95-7.94 4H cyclobutyl protons 6.35 3H carbomethoxy group 34 Meo2 ‘II'V/l ’W 35 ’W 290 (14000) 3:33 33 ggmfdimethyl group 8.32 3H propylidene methyl 8.13 3H a-methyl group 6.93-7.65 4H cyclobutyl protons 6.31 3H carbomethoxy group 14 _g$§_isomers, which are more likely to be forced into a non— planar configuration, have lower absorption intensities than the less strained, coplanar §£§g§_isomers (10). Consequently, the magnitude of the extinction coeffi— cient of the four geometric isomers should vary. Steric repulsion varies from maximum in 22) where the bulky carbo— methoxy and isopropyl groups are gig to each other, to a minimum in 24) where these two groups are Egggg. Accordingly, the absorption intensity of 22 is 6700, whereas in 24 it is 17,000. Compounds 22 and 22 fall between these two extremes and their structures are asSigned accordingly. These structural assignments are also consistent with the chemical shifts of the various methyl groups in the molecule. The methyl alpha to the carbomethoxy resonates at T 8.21 in 22'and 22; In both of these compounds the methyl group is direCted away from the remainder of the molecule. In 24 and 22 the same methyl resonates at T 8.12. In these two cases the methyl is directed toward the other groups, and, since this is a rigid molecule, it interacts with the methyl or isopropyl g;§_to it. Likewise the methyl group delta to the carbomethoxy in 22‘resonates at significantly higher field (T 8.63) than it does in 22'(T 8.46), 35, (T 8.42), or 24) (T8.32). The reason for this shift is that the methyl group is directed by the geometry of the molecule into the shielding region of the carbomethoxy group. Similar arguments may be advanced for the isopropyl group. 15 The mechanism for the formation of the five photopro- ducts involves the intermediacy of a Eggfdiene ketene 22' analogous to the ketene 2'observed by Griffiths and Hart (1). In the presence of methylene chloride no net re- 0 O 0 ¢ hv /’ A ——> —> I. Q 6T 5f CH2” l> 12, S22 12. action results, as the intermediate 22 can only thermally re— cyclize to form 12 or cyclize to the tricyclic compound 42, which is prevented by strain. However, if the irradiation of lfi'is performed in methanol, the ketene has an additional option, as it may suffer nucleophilic attack by the solvent to give unsaturated esters (see Figure 2). Two sets of products are obtained in approximately equal amounts. Com- pounds 22) 22) 22) and 24 compose 45% of the mixture. These four geometric isomers result from 1,6-addition to ketene 22; To date, this is the only example of 1,6-addition to a ketene produced from photolysis of a 2,4-cyclohexadienone. The remainder of the mixture is composed of ester 24) which is formed from 1,2-addition to the ketene. The nucleophile normally adds 1,2 to ketenes as evi- denced by the numerous reports of nonconjugated acid deriva— tives (1,2,3,4). For example, the highly substituted O O hv ’,r<‘f?, ——; ID <——, \l\ 0 42' MeOH 1,2-Addition 1,6-Addition OzMe COZMe ' / \ I ~2~ 21. + COzMe R 2.3. . MeOZC l 1’ 34 Figure 2. A mechanism for the products obtained upon pho- tolysis of dienone 42, 17 O he : 2H ([0 00 q hexamethylcyclohexadienone 4 gives nonconjugated amide 21, which results from 1,2-addition of dimethylamine to the ketene. Examples of 1,4—addition to a ketene intermediate are rarer. In their pioneering work, Barton and Quinkert (4) report several examples of 1,4-addition products. Irradia- tion of acetoxydienone 22'in the presence of cyclohexylamine H2 ' O OAC \\ / OAc 1,4 additiK / H H 3332. £32, results in the formation of a,B—unsaturated amide 22; No cases of 1,4-addition have been reported for 2,4-cyclohexa- dienones with only saturated alkyl substituents. Thus, when compared to the equally substituted but less strained hexamethylcyclohexadienone 1/ compound 42 reacts 18 differently thermally and towards nucleophiles. When ir- radiated in ethanol, the only reaction of photoproduct 2’is the thermal bond reorganization to bicyclohexenone 2 (1). No ketene reverts to starting dienone 4) nor is it trapped by the moderately nucleophilic solvent. In the case of O éo O .__JLL_;> ’/ ‘ ‘E_J;____ \\ 3. ethanol" 1, 2. Si photoproduct 22) reversion to starting material does occur, as no net reaction is observed in a non-nucleophilic solvent. In the presence of methanol, strained intermediate 22 reacts slowly with the solvent to form unsaturated esters. Appar- ently, the intramolecular bond-crossing path to form bi- cyclic enone systems is the preferred process. However when this pathway is blocked, the slower, less favorable nucleo— philic attack of a solvent molecule becomes a predominant operation. Also, compared to ketene 2) the carbonyl group 19 in 22 is less hindered as a result of an increase in bond angles caused by the rigid cyclobutane ring. Consequently only ester formation is observed for ketene 22; C. Oxidation of 5,6,7,8-Tetramethyle4i2,3,4-tetrahydro- naphthalene (42) and Characterization of the Products The results obtained from the photolysis of dienone 42, suggested that the photochemistry of a dienone containing a larger fused ring be studied. Since the isomerization of ketene 22,to tricyclic enone 42 is prevented by the strain induced by the cyclobutane ring, substitution of a more flexible ring might allow the isomerization to occur. ¢° 0 /l A9 n} l 9.2. 19 m Molecular models established that a cyclohexane ring is sufficiently pliant to make product 42'a reasonably stable 0 leg 17 :29, enone o 20 A synthetic route for the preparation of dienone 44 was devised. Gregorovitch (9) has recently developed a' procedure for tetramethylating tetralin in one step. Treatment of an acetic anhydride solution of tetralin with hydriodic acid and paraformaldehyde results in iodomethyla— tion. In situ reduction of the iodomethyl group with hypo- phosphorus acid allowed complete methylation to occur in (11:) HI > O. H3P02 41, 42 CF3CO3H BF3-etherate W” m 42 72% yield. Oxidation of hydrocarbon 42’with peroxytri- fluoroacetic acid in the presence of boron trifluoride etherate resulted in an 84% yield of isomeric dienones 42, 41) 42) and 42; The four isomers were isolated by column chromatography and vpc. The product composition consisted Of 2,2,4,5-tetramethylbicyclo[4.4.0]deca-1(6),4-dien-3-one 42' (20%), 3,3,4,5-tetramethy1bicyclo[4.4.0]deCa-1(6),4-dien-2- one 42'(36%), 3,4,5,5-tetramethylbicyclo[4.4.0)deca-1(6),3- diensz-one 4§,(20%), and the desired isomer, 2,4,4,5-tetra— rmfiflhylbicyclo[4.4.0]deca-1,5-dien-3-one, 41 (24%). The 21 structures were determined from their spectral properties analyses, chemical derivatives and photochemical reactions. Dienone 18 was an oil which showed conjugated carbonyl and double bond absorptions in the infrared region at 1645 and 1580 cm_1 respectively. Ultraviolet absorption ob- served at 329 nm (e 4600) is characteristic of substituted 2,4-cyclohexadienones. The nmr spectrum of lg'contained a six-proton singlet at T 8.90 (ggmfdimethyl group), two broad three-proton singlets at T 8.18 and 8.03 (allylic methyls), and two broad four-proton multiplets at T 8.12— 8.53 and 7.60-7.95 (cyclohexyl protons). The spectral properties of dienone 18 were not suf- ficiently characteristic to distinguish it from isomers 1Q, and 11; A base catalyzed deuterium exchange was performed on lgzwhich resulted in the uptake of three deuterium atoms 0 O . MeOD D e D 18 m 18--d3 (mass spectrum). The resonance at 1 8.03 disappeared from its nmr spectrum, indicating that there is a methyl group attached at the diene-carbon beta to the carbonyl. The only isomer which can satisfy these data is structure 18, as both isomers 1Q and lz’have methylene groups at this pOSition. Dienone gg'was a white solid which showed conjugated (EHHDOnyl and double-bond absorptions in the infrared region 22 at 1600 and 1630 cm-1 respectively. Ultraviolet absorptions observed at 250 nm (€ 12,000) and 275 nm (e 4500) are char— acteristic of substituted 2,5-cyclohexadienones. The nmr spectrum of gg’consisted of a six—proton singlet at T 8.82 (ggmfdimethyl protons), two three—proton triplets (g'= 0.9 Hz) at T 8.20 and 8.07 (allylic methyls) and two four- proton multiplets at T 8.25-8.55 and 7.57-7.88 (cyclohexyl protons). Spectral examination or chemical degradation of isomers 1g and 11 did not result in an unambiguous characterization of their structures. Consequently, it is necessary to assign these structures indirectly. The expected photo— o 0 11 2.9. O o O h h OgifiéMe 12 1:1. 252, IIrOduct of 1Q is enone 22, which contains no allylic methyl 1a 51b ' 49 50a and 51a, which may isomerize to the cyclopropyl inter- mediates 522 and 5123 Migration of the cyclopropyl group followed by ketonization results in the observed enones 44 and 42; The irradiation of a 1% solution of dienone l§.in a Pyrex test tube with a Rayonet photochemical reactor equipped with 3000 X lamps was monitored by ultraviolet spectroscopy. Complete conversion of dienone lg'to product was accomplished in 28 hours. Analysis of the photolysate by vpc estab- lished the presence of a single photoproduct 423 28 49 Spectral analysis (ir, nmr, uv spectroscopy) of enone 42 documented that it was identical with the second isomer isolated from the photolysis of dienone 42; This coinci- dence was an expected result, as the accepted mechanism for the photochemical isomerization of 2,4-cyclohexadienones predicts that enone 49 will be formed (1). O ¢‘;C) O *1>—>l.> 52 18 49 ‘UU m m Likewise, photolysis of a solution of 12 under identi- cal conditions employed for the irradiation of 18 gave the gm. 44 m expected photoproduct 44; A mechanism similar to that used to explain the photoisomerization of lg'can also be applied to‘LQ. AI:‘V' .H . '6- 'I. V (II 329 Dienone lz'could not be completely separated from 16' and 18; However, purification was not necessary since the photoproducts of 16 and 18 could easily be obtained and characterized in separate reactions. These photoproducts are easily distinguished from the expected product of L1 by nmr spectroscopy. Consequently, irradiation of a mix- ture of the three isomers would yield two known products. Any additional products can thus be attributed to 11. Ir- radiation for 28 hours of a 3% methylene chloride solution of a mixture of isomers 16, 11, and l§,in a Pyrex test tube with a Rayonet Photochemical reactor equipped with 3000 8 lamps resulted in three photoproducts. Purification of the products was accomplished by vpc. Spectral analysis showed that the two expected isomers 44' and 42 were present. The third isomer, which composed 30% of the mixture, was the desired enone 42, It was inferred from this result that dienone lz'comprised 24% of the orig- O // 0 hv _____g> /’ ‘\ engslz -- > I.» O O . ll 53 40 m I‘W inal oxidation product mixture. Compound 42 was a colorless oil whose infrared absorptions at 1690 and 1630 cm.1 are characteristic of a conjugated carbonyl and double bond) 30 respectively. The enone system is substantiated by ultra— violet absorption at 231 nm (e 6200). The nmr spectrum of 42 showed singlets at T 9.10, 9.05 and 8.92 with a total band area of nine protons (aliphatic methyl groups), a three-proton signal at T 8.23 (allylic methyl group) and a broad eight-proton multiplet at T 7.47—7.90 (cyclohexyl protons). The ketene intermediate 52 can undergo two competing processes, intramolecular cyclization to enone $2.0r sol- vent attack to form esters. The observed product is de— pendent upon the relative rates of these two reactions. Apparently, cyclization to enone is a more rapid process than solvent attack, since hexa-substituted dienones yield enones exclusively. However, attack of nucleophiles such as methanol can be facilitated by structural effects which distort the ketene. As previously described, irradiation of dienone $2 in methanol resulted only in a mixture of methyl esters (13). However, irradiation of lz'in methanol gave only photo- isomerized enone 42; No esters could be detected. Thus, the fused cyclohexane ring of dienone 11.15 sufficiently flexible to allow the formation of photoproducts analogous to those of other hexa-substituted cyclohexadienones, as evidenced by the exclusive production of tricyclic enone fig'upon irradiation in methanol. Apparently, the fused ring has no retarding effect upon the intramolecular cycli- ‘Zation rate of ketene intermediate 53, since nucleophilic 31 attack from the solvent (methanol) does not occur. This result supports the rationale advanced earlier to explain the facile reaction of ketene gg'with methanol. The com- pact configuration of the molecule which is caused by the cis, cis arrangement of the diene system is distorted somewhat by the cyclobutane ring resulting in facile nucleo- philic attack at the carbonyl carbon. The pliant cyclohexane ring exerts no such force on ketene 52; Consequently, the molecule does not suffer attack by the solvent, but rather cyclizes to form enone 42; EXPERIMENTAL A. General Procedures All infrared spectra were obtained using a Unicam Model SP-200 infrared spectrometer and were calibrated with polystyrene. The ultraviolet spectra were measured with a Unicam Model SP-800 spectrophotometer. Nuclear mag— netic resonance spectra were recorded with a Varian A-60 spectrometer using tetramethylsilane as an internal refer- ence. The mass spectra were performed by Mrs. Guile using a Hitachi Perkin-Elmer RMU-6 instrument. Melting points were obtained with a Thomas apparatus and are uncorrected. Elemental analyses were performed by Spang Microanalytical laboratories, Ann Arbor, Michigan. B. Preparation of Tetramethylbenzocyclobutene (222 A solution of 30 g (0.122 mole) of dichlorotetramethyl— benzocyclobutene in 225 ml of anhydrous ether was added to 900 ml of liquid ammonia which contained 7.4 g (1.07 mole) of lithium. Dry ice temperature was maintained during the One hour addition period and for one hour thereafter. The excess lithium was destroyed by the addition of ammonium Chloride and the ammonia was allowed to evaporate. The COntents of the flask were dissolved in a 1:1 mixture of 32 33 water and ether. The organic layer was washed with water (3 x 200 ml), dried with magnesium sulfate and the solution concentrated. The light yellow solid that remained was analyzed by vpc (5' x 1/4" SE-30, 180°, 60 ml/min of He). The analysis showed that two compounds were present. The major component (93%) was the desired tetramethylbenzocyclo— butene {231, mp 125-38° (lit. val. (6) 138-1390). The ir and and nmr Spectra corresponded to those in the literature. The minor component (7%) was identified as hexamethylbenzene. Its mp, ir, and nmr spectra were indentical to those of an authentic sample. The mixture was oxidized without further purification. C. The Oxidation of Tetramethylbenzocyclobutene (23) A solution of peroxytrifluoroacetic acid was prepared from 2.28 ml (0.083 mole) of 90% hydrogen peroxide and 17.7 g (0.083 mole) of trifluoroacetic anhydride in 24 ml of methylene chloride. The solution was maintained at 00 as it was added with stirring over 45 min to a solution of 9.0 g (0.057 mole) of 22.1“ 400 ml of methylene chloride which had been previous1y cooled to -300 with a 2-propanol- Dry Ice bath. Boron trifluoride etherate (27 ml of 48% BF3'Et20) was added concurrently with the addition of the peracid. The temperature was maintained at -300 during the addition and stirred for two hours thereafter. The reaction mixture was hydrolyzed with 100 ml of water and the organic layer was separated. The organic layer was washed with 34 water (2 x 200 ml), saturated sodium bicarbonate (3 x 100 ml), 5% aqueous sodium hydroxide (3 x 100 ml) and again with water (3 x 100 ml). The organic layer was dried over mag— nesium sulfate and concentrated to give a deep red viscous oil. Vacuum distillation resulted in 3.41 g (34%) of a yellow oil, bp 80—900 (0.2 mm). The distillate was analyzed by vpc (5' x 3/8" OV—225, 190°, 200 ml/min of He), which showed that only two components were present in approximately a 10:1 ratio. Separation of the two components was effected by preparative vpc employing the above conditions. D. Identification of the Products from the Oxidation of Tetramethylbenzocyclobutene (23) The major product, which was found to be 2,4,4,5-tetra— methylbicyclo[4.2.0]octa-1,5-dien-3-one 12» had xfing at 320 nm (e 4800), ir bands (CC14) at 1685 (c-o conjugated) and 1630 cm-1 (C=C conjugated), and an nmr spectrum (CC14, Figure 5) with bands at r 8.90 (s, 6H, ggmfdimethyl), 8.31, 8.36 (br s, 6H, allylic methyls) and T 7.14 (br s, 4H, cyclo— butyl protons). The mass spectrum showed a parent peak at m/e 176. ,éflél- Calcd for C12H160: C, 81.77; H, 9.15 Found: C, 81.89; H, 9.26. The minor product was found to be hexamethyl—2,4-cyclo- hexadienone. The compound had identical ir and nmr spectra 35 as an authentic sample of the dienone. Its mass spectrum showed a parent peak at m/e 178. E. Preparation of 2,4,4,5—Tetramethylbicyclo[4.2.0jocta— 1,5-dien~3-one-8-d2 (15-dg) To 100 mg of 15 was added 0.5 ml of 1M sodium methoxide in methanol-d. The solution was allowed to stand for 1.5 hours at room temperature, then was poured into 2 ml of carbon tetrachloride and washed with water (3 x 5 ml). The solution was dried with anhydrous magneisum sulfate and concentrated. The ir and uv spectra of recovered lézdz were identical with those of the unlabeled compound 12 except for an ir absorption at 2200 cm-1 (C~D stretching). The nmr spectrum of 15-d2 showed signals at T 8.90 (s, 6H, gem—dimethyl), 8.37 and 8.31 (br s, 6H, allylic methyls), and 7.18 (br S, 2H, cyclobutyl protons). A parent peak at m/e 178 was ob- served in the mass spectrum of 15-d2 F. Catalytic Reduction of Ag,: Preparation of 2,4,4,5- Tetramethylbicyclo[4.2.0]octa-1-ene-3-one (28) To 50 mg of prereduced platinum oxide suspended in 13 ml of absolute ethanol was added 100 mg (5.7 x 10‘4 mole) of 15 dissolved in 1 ml of absolute ethanol. The mixture was stirred under one atmosphere of hydrogen for 13 min, at which time a definite decrease in the rate of hydrogen ab- sorption was observed, as one mole equivalent (12.7 ml) of 36 hydrogen was absorbed. After removal of the catalyst and solvent, a yellow oil remained, which consisted of 95% of the desired compound (vpc analysis). The product was puri— fied by preparative Vpc (10' x 1/4", SE-30 at 180°, 80 ml/min of He). The colorless oil had a Afi§3H at 247 nm (e 4800), ir band (CCl4) at 1655 cm"1 (c=o conjugated) and an nmr spectrum (CCl4, Figure 6) with bands at T 9.14 (d, 3H, i= 7 Hz), 9.0, 8.83 (s, 6H, ggmfdimethyls), 8.45 (br s, 3H, g = 0.9 Hz, allylic methyl), 7.8-8.3 (m, 3H, C-7 and C—6 pro— tons), 7.1-7.5 (m, 2H, C-8 protons) and T 6.34 (q, 1H, C-5 proton). The mass spectrum showed a parent peak at m/e 178. Anal. Calcd for C12H130: C, 80.85; H, 10.18 Found: C, 80.91; H, 10.23. G. Preparation of 2,4,4,5-Tetramethylbicyclo[4.2.0]octa— 1—ene-3-one-6-d1-8-d2 (22) The procedure outlined in section B of the experimental part was followed to obtain the labeled isomer. Analysis by vpc (10' x 1/4", 53-30 at 1800, 80 ml/min of He) showed the presence of two compounds, which proved to be gi§_and trans isomers. The isomers were separated by vpc (above conditions). The gi§_isomer was not studied further. The trans isomer had ir and uv spectra similar to the unlabeled compound 28” except that the ir spectra had a band at 2360 cm'1 (C-D stretching). The nmr spectrum (CCl4. Figure 7) consisted of bands at T 9.16, 9.05 (d, 3H, g.= 6,5 Hz, c-5 methyl), 9.11 (s, 3H, gem-dimethyl), 8.98 (s, 3H, 37 gem-dimethyl), 8.45 (s, 3H, allylic methyl) and T 8.0—8.6 (m, 3H). The mass spectrum showed a parent peak at m/e 181. H. Irradiation of 2,4,4,5-Tetramethylbicyclo[4.2.0]octa— 1,5-dien-3—one (15) A solution of 200 mg of 15 in 18 ml of methylene chloride was irradiated using a 450 watt Hanovia lamp with a Pyrex filter for 8 hours. Only starting material was recovered. A solution of 200 mg of lé'in 18 ml of methanol was irradiated using a Rayonet photochemical reactor with 3000 R lamps. The photolysis was monitored by Vpc (10' x 1/4" SE~30, 200°, 80 ml/min of He), which indicated that five photoproducts were formed. All of the starting material was consumed in 22 hours. The retention times of the products were 10.5, 14.7, 17.7, 20.5, 22.5 minutes (above conditions). The five compounds were separated by preparative vpc using the above conditions. I. Identification of the Products from Irradiation of (13) 1. The compound with a retention time of 10.5 min, which constituted 50% of the mixture, is considered to be methyl 2-(1,2-dimethylpropeny1)-a-methyl—1~cyclobutene—1- acetate 31; The compound had a AgggH at 239 nm (e 8100), ir bands (CC14) at 1715 (0:0) and 1630 cm.1 (c=c) and an nmr spectrum (CC14, Figure 8) with signals at T 8.80 (d, 3H, g,= 7 Hz), 3.30 (s, 9H, allylic methyls), 7.61 (br d, 4H, 38 cyclobutyl protons), 6.73 (q, 1H, g_= 7 Hz), and T 6.40 (s, 3H, ester methyl). The maSS spectrum showed a parent peak at m/e 208. 5221. Calcd for C13H2002: C, 74.96; H, 9.68 Found : C, 74.79; H, 9.67. 2. The compound with a retention time of 14.7 minutes, which constituted approximately 12% of the mixture, is considered to be (Z,Z)-methyl 2-(1,2-dimethylpropylidene)- a—methyl-Al'a-cyclobutaneacetate 32; The compound had a MeOH xmax at 289 nm (e 6700), ir bands (CC14) at 1700 (c=o con- 1 (C=C conjugated and an nmr Spectrum jugated) and 1650 cm- (CC14, Figure 9) with signals at T 9.07 (d, 6H, g.= 7 Hz, ggmfdimethyls), 8.46 (br s, 3H, allylic methyl), 8.22 (s, 3H, allylic methyl), 7.57 (br s, 4H, cyclobutyl protons). and T 6.36 (s, 3H, ester methyl). The mass spectrum showed a parent peak at m/e 208. Anal. Calcd for C13H2002: C, 74.96; H, 9.68 Found: C, 75.04; H, 9.78. 3. The compound with a retention time of 17.7 minutes constituted approximately 12% of the mixture. It is con— sidered to be (Z,E)-methyl 2—(1,2-dimethylpropylidene)—d- methyl-A1'a-cyclobutaneacetate 32; The compound had a ifiggH at 289 nm (e 8100), ir bands (cc14) at 1702 (c=o conjugated) and 1650 cm.-1 (C=C conjugated), and nmr spec— trum (CC14, Figure 10) with signals at T 8.99 (d, 6H, g_= 7 Hz, gem-dimethyl), 8.63 (s, 3H, allylic methyl), 8.20 (s, 39 3H, allylic methyl), 7.50 (br s, 4H, cyclobutfi_protons), and T 6.36 (s, 3H, ester methyl). The mass spectrum of the compound had a parent peak at m/e 208. ‘Aggl. Calcd for C13H2002: C, 74.96; H, 9.68 Found: C, 74.92; H, 9.63. 4. The compound with a retention time of 20.5 minutes, which is considered to be (E,E)-methyl 2-(1,2-dimethyl— propylidene)-a-methyl-A1'a4cyclobutaneacetate 32; The compound had a xfing at 288 nm (6 17,000), ir bands (CC14) at 1700 (c=o) and 1640 cm-1 (C=C), and nmr spectrum (CCl4, Figure 11) with bands at T 8.96 (d, 6H, g_= 7 Hz, gem: dimethyl), 8.42 (s, 3H, allylic methyl), 8.15 (s, 3H, allylic methyl), 6.95-7.95 (br m, 5H, cyclobutyl and iso- propyl protons), and T 6.35 (s, 3H, ester methyl). A parent peak in the mass spectrum was observed at m/e 208. Anal. Calcd for C13H2002: C, 74.96; H, 9.68 Found: C, 74.90; H, 9.65. 5. The compound with a retention time of 22.5 minutes, which constituted approximately 125%of the mixture, is con- sidered to be (E,Z)-methyl 2-(1,2-dimethylpropylidene)—a- methyl-A1'a-cyclobutaneacetate 35; The compound had a Afi§3H at 290 nm (€ 14,000), ir bands (CC14) at 1700 (c=0) and 1640 cm"1 (c=c), and an nmr spectrum (CC14, Figure 12) with signals at T 8.99 (d, 6H, g_= Hz, gem-dimethyl), 8.32 (s, 3H, allylic methyl), 8.13 (s, 3H, allylic methyl), 6.93— 7.65 (br m, 5H, cyclobutyl and isopropyl protons), and 40 T 6.36 (s, 3H, ester methyl). The mass spectrum showed a parent peak at m/e 208. .éflélf Calcd for C13H2002: C, 74.96; H, 9.68 Found: C, 74.80; H, 9.65. J. Preparation of 5,6,7,8-Tetramethyl-1,2,3,4-tetrahydro— naphthalene (42) Compound 42 was prepared according to the method of Gregorovitch (9). To a 2 liter, 3-necked flask was added 800 ml of acetic anhydride and 40 g of paraformaldehyde. To the stirred solution 200 ml of hydriodic acid was cautiously added. As the solution was maintained at 90- 100°, 20 ml (19.4 g, 0.147 mole) of 1,2,3,4-tetrahydro— naphthalene was added over an hour period. The temperature was held at 90-100° for an additional four hours and then raised to reflux (116°) for five hours. During the 10 hour reaction period, the solution was cooled to 60° three times and decolorized with hypophosphorous acid (140 ml). The warm solution was added to 2500 ml of ice water and the product was separated by filtration. The collected solid was boiled for five minutes in 150 ml of pyridine and the hot solution was poured into 2500 ml of water containing 200 ml of acetic acid. The mixture was cooled and the product was collected. Further purification was accomplished by dissolving the product in 400 ml of methanol, concentra- ting the solution to half volume, and cooling. A yield of 20 g (72%) of white plates, mp 78-81° [lit. (9) mp 80-810] was collected. n...) --o‘ ‘r .v'...‘ a. . . OD! ’ V - b. O ' P”. i. /' -o h‘v I - V. l h. ‘. \- I“ i,“ h I h.. H“, n I 41 K. Oxidation of 5,6,7,8-Tetramethyl-1,2,3,4-tetrahydro- naphthalene (42) Oxidation of hydrocarbon gz'was performed according to the procedure described in Section C with one modifica— tion, a change of reaction temperature to -10°. The re- action mixture was vacuum distilled to yield 77% of a yellow oil, pr-me 108-1140. The oil was analyzed by vpc (10- x 1/4", 0v-25, 200°, 60 ml/min of He), which showed the presence of four major compounds with retention times of 20, 22.8, 24 (approximate), and 28 minutes. L. Identification of Products from the Oxidation of Cgi)° 1. The compound with a retention time of 20 minutes, which constituted 20% of the reaction mixture, is considered to be 2,2,4,5-tetramethylbicyclo[4.4.0]deca-1(6),4-dien— 3-one 18, The compound had a xnggg at 329 nm (e 4600), ir bands (CC14) at 1645 (c-o conjugated) and 1580 cm-1 (C =C conjugated), and an nmr spectrum (CC14, Figure 13) with signals at T 8.90 (s, 6H, ggmfdimethyl), 8.12-8.53 (m, 4H, cyclohexyl protons), 8.18 (br s, 3H, allylic methyl), 8.03 (br s, 3H, allylic methyl, and T 7.60-7.95 (m, 4H, cyclo— hexyl protons). The mass spectrum showed a parent peak at m/e 204. ‘Aggl, Calcd for C14H200: C, 82.30; H, 9.87 Found: C, 81.78; H, 10.12. 42 2. The compound with a retention time of 22.8 minutes, which constituted 40% of the reaction mixulre, is con- sidered to be 3,3,4,5-tetramethylbicyclo[4.4.0]deca-1(6),4— dien-2-one 42, The compound had a i332? at 329 nm (e 3000), ir bands (CC14) at 1640 (C-O conjugated) and 1580 cm.1 (c=c conjugated), and nmr spectrum (CC14, Figure 15) with sig- nals at T 8.89 (s, 6H, ggmfdimethyl), 8.18 (s, 6H, allylic methyls), 8.07—8.53 (m, 4H, cyclohexyl protons), and T 7.48- 7.88 (m, 4H, cyclohexyl protons). The mass spectrum showed a parent peak at m/e 204. Anal; Calcd for C14H200: C, 82.30; H, 9.87 Found: C, 82.42; H, 9.83. Under identical conditions used in the oxidation of 42, a sample of pure 46'was not isomerized to dienone 43, 3. The compound with an approximate retention time of 24 minutes, which constituted 24% of the reaction mixture, could not be completely separated from lg'and thus unequivo- cal characterization was not possible. 'However, examination of the photoproducts resulting from a mixture of isomers 46, $1, and Ag'led to theoassignment of compound 4Z'as 2,4,4,54 tetramethylbicyclo[4.4.0]deca-1,5-dien-3-one."Elucidation of the structure was possible since the photochemical paths Of these dienones are known (1,2,4,5). Consequently, the Structures of the reactants may be determined from the pro— duCts. 43 4. The compound with a retention time of 28 minutes, which constituted 20% of the reaction mixture, is considered to be 3,4,5,5-tetramethylbicyclo[4.4.0]deca-1(6),3-dien- 2-one 43, The compound was a white powder, mp 87-890, with a Xfifigg at 250 nm (6 12,000), shoulder at 275 nm (e 4500), ir bands (cc14) at 1660 (shoulder) and 1630 cm’1 (C=C con- jugated), and an nmr spectrum (CC14, Figure 14) with sig- nals at T 8.82 (s, 6H, ggmfdimethyl), 8.25-8.55 (m, 4H, cyclohexyl protons), 8.20 (br s, 3H, g_= 0.9 Hz,allylic methyl), 8.07 (br m, 3H, g_= 0.9 Hz, allylic methyl), and T 7.57-7.88 (m, 4H, cyclohexyl protons). The mass spectrum showed a parent peak at m/e 204. 1533;, Calcd for C14H200: C, 82.30; H, 9.87 Found: C, 82.08; H, 9.96. M. Preparation of 2,2,4,5-Tetramethylbicyclo[4.4.0]deca- 1(6),4-dien-3-one-5-methyl-d3 (18-d§) A 50 mg sample of compound 43 was labeled according to the procedure outlined in Section E. The compound had an nmr spectrum (0014) which had signals at T 8.90 (s, 6H, ggm-diemthyl), 8.37 (m, 4H, cyclohexyl protons), 8.18 (s, 5H, allylic methyl), T 7.47-7.93 (m, 2H, cyclohexyl protons). The mass spectrum of the compound showed a parent peak at m/e 207. 44 N. Preparation of 3,3,4,5-TetramethylbicycloL4.4.0]deca- 1(6),4-dien-2-one-7-d2 (16-d2) A 50 mg sample of compound 43'was labeled according to the procedure outlined in Section E. The compound had an nmr spectrum (CCl4) which had signals at T 8.90 (s, 6H, ggmfdimethyl), 8.37 (m, 4H, cyclohexyl protons, 8.18 (s, 6H, allylic methyl), and T 7.47-7.93 (m, 2H, cyclohexyl protons). The mass spectrum of the compound showed a parent peak at m/e 206. 0. Preparation of 3,4,5,5-Tetramethylbicyclo[4.4.0]deca— 1(6),3-dien-2-one-4-methyl-d3-7—d2 (43—d5) A 50 mg sample of compound 43'was labeled according to the procedure outlined in Section E, except that the exchange required 30 hours for completion. The compound had an nmr spectrum (CC14) which had signals at T 8.82 (s, 6H, ggmrdimethyl), 8.37 (m, 4H, cyclohexyl protons), 8.18 (s, 3H, allylic methyl), and T 7.72 (m, 2.5H, cyclo— hexyl protons). The mass spectrum showed a parent peak at m/e 209. P. Irradiation of 3,4,5,5-Tetramethylbicyclo[4.4.0]deca— 1(6),3-diene-3-one (43) A solution of 100 mg of 43'in 9 ml of methylene chlor— ide was irradiated using a Rayonet photochemical reactor with 2537 X lamps. Complete conversion of the starting 45 material to product occurred in 1 hour and 15 minutes. The photolysis was monitored by vpc (5' x 1/4", SE-30, 190°, 60 ml/min of He). One product was isolated by vpc (above conditions), and was identified as 3,4,4,5-tetramethyltricyclo[4.4.0.03'5] deca-l(6)-ene-2-one 44 by analysis of spectral data. The MeOH- compound had a A max and 320 nm (e 600), ir bands (CC14) at 2950, 1690 (c=o at 237 nm (e 7500), 275 nm (6 2400), conjugated) and 1640 (C-C conjugated) cm.1 and nmr spectrum (CC14, Figure 16) with signals at T 9.03 (s, 3H, cyclo— propylmethyls), 8.90 and 8.87 (s, 6H), 8.80 (s, 3H) and T 7.55-8.55 (br m, 8H, cyclohexyl protons). The mass spectrum showed a parent peak at m/e 204. Anal. Calcd for C14H200: C, 82.30; H, 9.87 Found: C, 82.07; H, 9.66. A second product was isolated by column chromatography using activated alumina and elution with hexane. Final purification was accomplished by vpc (2' x 1/4", 14% OV-25, 90°, 80 ml/min of He). The compound was found to be identi- cal (ir, uv, nmr and mass spectrum) with enone 43 formed from the photolysis of 43, Q. Irradiation of 2,2,4,5-Tetramethylbicyclo[4.4.0]deca- 1(6),4-diene-3-one (13) A solution of 100 mg of 43’in 9 ml of methylene chlor- ide was irradiated using a Rayonet photochemical reactor with 3000 R lamps. The reaction was followed by uv, as the c .'-~I\ hf ins-iv fl".. An .- \ 1...»:o ...p~ % n in.“ 07“,- :‘ no t.- .J ‘p .1 46 product was found to be unstable to vpc conditions. Con- version of the starting material was complete in 28 hours. The product was purified by column chromatography using alumina (J. T. Backer activity 1, 80-200 mesh) and eluted with 400 m1 of hexane. The eluant was collected in 50 ml fractions. Most of the pure product was obtained from fractions 4, 5, and 6. Final purification was effected by vpc (2' x 1/4", 14% OV-25, 90°, 80 ml/min of He). The product, 8,9,10,10-tetra— methyltricyclo[4.3.1.0]deca-8-ene-7-one 43, had a AMigg at 239 nm (e 4800), 275 nm (e 2500) and 320 nm (E 500), ir bands (CC14) at 1690 (C=O conjugated) and 1640 (C=C conju- gated) cm"1 and an nmr spectrum (CC14, Figure 17) with signals at T 9.07 (s, 3H, cyclopropyl methyls), 8.83 (s, 3H, cyclopropyl methyl), 8.45 (br s, g.= 1 Hz, 3H, allylic methyl), 8.10 (br s, g.= 1 Hz, 3H, allylic methyl), and T 7.7-8.7 (br m, 8H, cyclohexyl protons). The mass spectrum showed a parent peak at m e 204. Anal. Calcd for C14H200: C, 82.30; H, 9.87 Found: C, 82.24; H, 9.89. R. Irradiation of 3,3,4,5-Tetramethy1bicyclo[4.4.0]deca— 1(6),4-dien-2-one (16) A A solution of 100 mg of 43 in 9 ml of methylene chlor- ide was irradiated using a Rayonet photochemical reactor With 3000 R lamps. The reaction was monitored by vpc 47 (2' x 1/4", 15% OV-25, 145°, 120 ml/min of He). Conversion of 16 to product was complete in 24 hours. One photoproduct was isolated by vpc (above condi- tions). The compound was found to be identical (ir, nmr) with 44, which was formed as a photoproduct of 43, S. Irradiation of 2,4,4,5-Tetramethylbicyclo[4.4.0]deca- 1,5-dien-3-one (11) A solution containing 300 mg of isomers 43, 11, and £3 in approximately a 1:1:2 ratio in 20 ml of methylene chlor- ide was irradiated using a Rayonet photochemical reactor equipped with 3000 R lamps. The reaction was monitored by ir spectroscopy. Conversion of the starting dienone was complete in 24 hours. Analysis of the photolysate by Vpc (2' x 1/4", 15% OV-25, 140°, 100 ml/min of He) showed the presence of three photoproducts. As expected, two of the compounds had reten- tion times identical with enones 44 and 43, The third com- pound which composed 30% of the mixture was purified by vpc (above conditions) and identified as 2,2,3,5-tetramethyl- tricyclo[4.4.0.01'3]deca-5-ene-4-one 42'by analysis of MeOH spectral data. The compound had a A max at 231 nm (e 6250) and 300 nm (e 1850), ir bands (601,) at 2950, 1690 (”c-0 conjugated) and 1630 (v conjugated) cm"1 and an nmr spec- C=C trmm (CC14, Figure 18) with signals at T 9.10, 9.05 and 8.92 (s, 9H, aliphatic methyl groups), 8.23 (br s, 3H, 48 allylic methyl) and T 7.47-7.90 (br m, 8H, cyclohexyl pro- tons). The mass spectrum showed a parent peak at m/e 204. A233; Calcd for C14H200: C, 82.30; H, 9.87 Found: C, 82.20; H, 9.90. SUMMARY 1. Oxidation of tetramethylbenzocyclobutene with peroxy- trifluoroacetic acid-boron fluoride etherate in methylene chloride resulted in a 36% yield of 2,4,4,5-tetramethyl- bicyclo[4.2.0]octa-1,5—dien-3—one (43). No other volatile components could be detected. 2. Irradiation of a methylene chloride solution of dienone 43'gave no net reaction. Irradiation in methanol resulted in five photoproducts. Nonconjugated ester 34, which com— posed 55% of the mixture, is produced by 1,24addition of methanol to the ketene intermediate. The remaining 45% of the photolysate were conjugated esters 33, 33, 34, and 33, These four geometric isomers result from 1,6-addition to the ketene intermediate. 3. Oxidation of 5,6,7,8-tetramethyl-1,2,3,4-tetrahydro- naphthalene (43) with peroxytrifluoroacetic acid-boron fluoride etherate in methylene chloride afforded four iso- meric dienones in an 84% yield. These 2,4-dienones, 3,3,4,5-tetramethylbicyclo[4.4.0]deca-1(6),4-dien-2-one (1g), 2,4,4,5-tetramethylbicyclo[4.4.0]deca-1,5-dien—3—one (11), and 2,2,4,5-tetramethylbicyclo[4.4.0]deca-1(6),4— dien-3-one (43), and one 2,5-dienone, ' 49 5O 3,4,5,5-tetramethylbicyclo[4.4.0]deca-1(6),3-dien-2-one (43), were isolated in 36, 24, 20, and 20% yield respectively. 4. Irradiation of a methylene chloride solution of 2,5- dienone 43'gave the two expected enones 3,4,4,5-tetramethyl- tricyclo[4.4.0.03'5]deca—1(6)-ene-2-one (44) and 8,9,10,10- tetramethyltricyclo[4.3.1.0]deca-8-ene-7¥one (43). Enones 44 and 43 were also obtained separately upon irradiating 2,4-dienones 43 and 43 respectively. 5. Irradiation of a methylene chloride or methanol solu- tion of 2,4-dienone 4Z'afforded only 2,2,3,5—tetramethyl— tricyclo[4.4.0.01'3]deca-5-ene-4-one (42). In contrast to the results obtained from the photolysis of dienone 43, no ester formation could be detected in the irradiation of 41, The Sigfdiene ketene intermediate is unaffected by the cyclo- hexane ring, since photoisomerization occurs to give enones analogous to those obtained from other hexa-substituted dienones. Part II. Mechanism for Electrophilic Addition to Methylcyclopropane 51 e ‘ r- -‘ ~ u ' I ‘vfi - ‘E "‘u\ A . INTRODUCTION The existence of protonated cyclopropane intermediates has been well established experimentally (14). Both Baird and Aboderin (15,16,17) and Deno (18) demonstrated that the introduction of cyclopropane into a deuterated sulfuric acid solution did not give 3-deuteropropyl hydrogen sulfate which might have been suspected from straightfoward electro— philic addition, but instead resulted in additional hydrogen— deuterium exchange of the cyclopropyl hydrogens. Hydrolysis of the Efpropyl hydrogen sulfate gave 1-propanol which had deuterium at C-1, C—2, and C-3, as shown in the equation. D2804 * * * ,5}: A , CH3 - CH2 — CH2 — OH 46 17 38 %D 22. 2‘1 These were key experiments in establishing protonated cyclo- propanes as important intermediates in electrophilic addi- tions of acids to cyclopropane. The results could not be explained by a classical gfpropyl cation nor were they con— sistent with a 1,3-hydride shift process. Ring closure reactions which occur during the deamina- tion of amines (15,17,19,20,21) and deoxidation of alcohols 52 .71‘ .6“ n... a U «U‘ .‘u. a ( '. I hug-I unva- p '- vu- 0 an 'Ovv u n. . .HQC H .‘v\ “V 53 (22) are also explained using protonated cyclopropane inter- mediates. Baird and Aboderin examined the cyclopropane fraction formed in the nitrous acid deamination of 3,3,3—d3- aminopropane (15) and found 43% cyclopropane-d2 and 57% cyclopropane-d3 to be present. These percentages were in excellent agreement with the statistical value of 3/7 and 4/7 respectively as calculated from the protonated cyclo— propane mechanism. Likewise, examination of the open-chain products from these reactions revealed that deuterium migration to C-2 and C-3 had occurred. Karabatsos (19) found that deamina- tion of 2,2-d2-1-aminopropane perchlorate 33 resulted in alcohols with deuterium at C-1, 33 and 33, This result rules out 1,3- and 1,2- hydride (or deuteride) shifts. A CH3CD2CH2NH3+ ClO4— —-> (C2H3D2)CH20H + C2H5CD20H 5.6. 97.9% 31 1.2% 33 + (C2H4D)CHDOH 0.9% 22, process involving 1,3-shifts cannot result in a deuterium atom at C-1, and the higher proportion of 33’over 33’ [1.2:0.9%] cannot be explained by a series of 1,2-shifts. The mechanism used to explain the addition of acids (15—18,23-25), acid chlorides (26-28), and bromine (29) to cyclopropane also involves the intermediacy of protonated cyclopropane. Upon acylating cyclopropane, Hart and Schlos— berg (27) found products 33 and 33 which could not be 54 explained by mechanisms involving 1,3-addition 143 a classical primary carbonium ion. However, all of the observed products can be nicely accounted for by employ- ing a protonated cyclopropane intermediate. This mechanistic scheme was supported by the demonstra- tion that reverse addition (443, addition of acetyl chlor- ide-aluminum chloride solution to the cyclopropane solution) H H o H H A1013 9 :+\c'5cn3 ——>, H2 3 S? 00’ AcCl °-°CCH3 22. H H L H OH :51, AlCl4- /u\/\/Cl 23% + 4% / 31% resulted in the increased yields of 33 and 33 at the expense of 34'and 33, This result is consistent with a process involving equilibrating ions 34 and 33, since the concentra- tion of nucleophile is less than during normal addition a rla' . nzvn- u‘.~' ,...D u. «I. ‘Y‘w '. ‘6. o V: IN iri- "‘b 55 (4:2, addition of cyclopropane to an acetyl chloride- aluminum chloride solution). Consequently, more time is allowed for the intermediates to equilibrate resulting in higher yields of products from ion 33, Consequently, the evidence indiCates that a protonated cyclopropane intermediate is more stable than a primary carbonium ion, which is the alternative intermediate (18). This result has been verified by molecular orbital calcula— tions, which conclude that an edge-protonated cyclopropane is lower in energy than a primary carbonium ion (30,31,32). Once the preference for protonated cyclopropane over primary alkyl cations was established, the logical extension was to examine alkyl cyclopropanes to compare the relative stabilities of protonated alkylcyclopropanes with secondary alkyl cations. The existence of a protonated methylcyclo- propane has been postulated by Saunders (33) to explain the low temperature nmr spectrum of the Eggfbutyl cation. The cation was produced at —110° by treating 2—fluorobutane with antimony pentafluoride. The solution was observed by nmr spectroscopy at temperatures ranging from -112 to -40°. The spectrum contained two peaks (relative intensity 2:1) at -110° due to the C-1 plus the C—4 protons and the C-2 plus C-3 protons respectively. The C-2 and C-3 protons become magnetically equivalent due to the rapid 1,2-hydride shifts occurring even at -110° in the Eggfbutyl cation. The two peaks observed at -110° coalesce at -55°. The process which coalesces the two peaks in the sec-butyl 56 cation was ascribed to a cyclization to protonated methyl- cyclopropane, followed by proton scrambling and reopening to gggfbutyl cation again. The alternate mechanism for this interconversion involves rearrangement to the primary cation. This route was disfavored by activation energy considerations and by comparing these results with those obtained in the isopropyl cation case. Brouwer has also suggested protonated methylcyclopro- pane as an intermediate in the acid-induced carbon inter- change process observed in gfbutane (34). Treatment of 27 butane-1—13C with HF-SbF5 gave gfbutane with the labeled carbon at C-1 and C-2. The intermediate gggfbutyl cation closes to a protonated methylcyclopropane and reopens to form the rearranged cation, which now has 13C at C-1 and C-2. No isobutane was recovered. The classical mechanism for electrophilic attack on an alkylcyclopropane involves the production of a secondary cation which can then give rise to 1,3-addition products. Hart and Schlosberg obtained two types of products from the acylation of methylcyclopropane with acetyl chloride-aluminum chloride. Ketone 33'is the expected 1,3-addition product. 0 O O —%A1Cl )k/Y + + / H AcCl C1 C1 69 70 71 72 W m rvv W 65% 12% 14% 57 Ketones Z4’and Z3'result from 1,2-addition to 60, enone 72 possibly being fOrmed by elimination of hydrogen chloride from 34, Two possible mechanisms may be proposed to account for 1, 3— —additio::> I/Jl\v/’\\:/’e AlCl4 //:L\/’::\L/’ AlCl ACCl 2% AlCl4 33 isomerizati?n N 9:4 0 / .22 the observed products. The classical mechanism involves normal 1,3-addition ylg_carbonium ion 33'to give the major product 33, However, methylcyclopropane might also isomer- ize to 2-butene ‘14 in the presence of acid. 1,2-Addition of acetyl chloride would then lead to products 34 and 33. An alternate mechanism involves a protonated methyl- cyclopropane intermediate. The major product 33'would be formed from the intermediate 33, Compound Z4'WOuld arise from an attack of chloride ion on intermediate 33, which is in equilibrium with 33, Loss of a proton from ZZ'could give rise to Z3, 58 A1013 AH AcCl 69 C1 C1 71 rvv 72 It was considered important to determine which of thse two mechanisms is operative. It:Esimpossible to discriminate between these two mechanisms with methylcyclopropane itself. However, inclusion of a deuterium at C-1 in Qg'would result in an appropriate system to distinguish between the two mechanisms. The deuterium acts as a label which can be traced in the reaction products. Consequently, a mechanism which involves an equilibrating intermediate such as a pro— tonated cyclopropane, can be detected by a scrambling of the label in the products, whereas a classical process will yield products which have deuterium only at the original carbon. The results obtained from the examination of methyl- cyclopropane-l-d1 zg'with electrophiles are the subject of Part II of this thesis. RESULTS AND DISCUSSION A. Consequences of Deuterium Label It is instructive to consider the possiblities which might result upon reacting deuterated methylcyclopropane Z§’ with electrophiles. In the case of acylation, the deu- terium should be found on different carbons depending upon the mechanism. Figure 4 shows the various positions of the deuterium if equilibrating bridged ions are involved as intermediates. In this scheme only ions with a positive charge adjacent to the methyl group are considered, as the carbonyl group destabilizes an adjacent positive charge. Electrophilic attack on the C1-C2 bond is assumed to be preferable to attack at C2-Ca, since a: methyl group should exert a controlling influence. The major products in this scheme will be compounds §2'and Qgfi since they are derived from the first formed protonated intermediates (12, §lx and §§). Compounds §§’and §§ require equilibration of inter- mediates §§'and §§'prior to reaction with chloride ion to form products. A second possibility is a mechanism which involves two competing processes. The labeled methylcyclopropane may react in the usual manner to form 1,3-addition products. 59 60 .fiolfilmcmmoumoHu>oH%LumE mo coflpmamom mzu as mmocwsvmmcoo mcflaawflmq mw HO O O 3 mm m .ol {.41 A . E. u Hoaa mm Ho O m v HUH< ammo )>\ hm E Q vaoaa ; m mm m mm Hm m mmooo m mooo \\\. ATIIIIII: 0,“; Illllllmv m afloaa .w muzmflm mw O HO O -«Hoaa mm m m 3000 .\ Q WM AH HUU¢ «Hoam Q .u... .v‘.‘ Q '5"‘ a...“ . .. t-"b' ~5'v‘ n: (I) ._.4 :w V‘U I). I r?" Sc? é (D 61 Compound Z§'may also undergo a rapid isomerization to labeled Zébutene-Z-dl 22', which may then react to give 1,2—addition products §§Jand 21, No deuterium scrambling would be expected with this type of mechanism. Since the major product obtained from the acylation of zg'is the 1,3-addition product (27), examination of gg'for deuterium scrambling should give an indication of the' mechanism of its formation. The mechanism of addition of other electrophiles such as hydrogen bromide or hydrogen chloride should likewise be subject to test by analogous labeling experiments. The presence of deuterium at the terminal carbons or hydrogen at the methine position in the product can be accounted for by invoking a protonated intermediate, whereas conventional 1,3-addition of the acid would result in no deuterium scrambling. B. Preparation of Methylcyclopropane-1-d1 (Z§) The desired deuterated methylcyclopropane zg'was syn- thesized using an adaptation of Greenlee and Boord's pro- cedure (35,36) for preparing methylcyclopropane. 62 CH3 CH3 ClCH C-CH fl? ClCH ('3 .52.; 2 2 B2202 2 , CHZBr EtOH D 92 g; 78 W W The necessary precursor, 1-bromo—3—chloro-2—methylpropane— 2-d1 92’ was obtained by adding deuterium bromide to 3- chloro-Z-methylpropene 92p. The deuterium bromide, which was prepared by treating benzoyl bromide with deuterium oxide, was bubbled into an ice-cold solution of gg’contain— ing benzoyl peroxide. Addition of the deuterium bromide occurs in an anti-Markownikoff fashion yielding 82% of the desired product 92; Reflux of the labeled dihalide ggiwith zinc and aqueous ethanol for 24 hours resulted in a 92% yield of methylcyclopropane-l—d1 lg. after purification. Purification was necessary to remove a small amount of olefin formed in a side reaction. This procedure was ac— complished by passing the labeled cyclopropane through a neutral potassium permanganate solution. The compound was identified by comparison of its spectral and physical proper- ties with those of the unlabeled compound. The infrared spectrum showed absorptions at 3070 (VG-H) and 2250 (VC-D) cm—l. The nmr spectrum contained two-proton multiplets at T 9.98 and 9.52 which are due to the cyclopropyl protons and a three-proton singlet, broadened due to coupling with the C-1 deuterium, at T 8.98 ascribed to the methyl protons. The molecular weight was found to be 57 amu by mass spectros— copy. 63 The deuterium content of the labeled cyclopropane was calculated from its mass spectrum. Comparison of the P and P-l peak intensities for labeled and unlabeled methylcyclo- propane showed the deuterium content in the labeled species to be greater than 98%. Although measurement of relative signal intensities in nmr spectroscopy is a less sensitive method than in mass spectroscopy, such measurements also showed that a maximum of 2% hydrogen was present at C—1. C. Reaction of Methylcyclopropane-l-d (Z§) with Electro- philic Reagents 1. Acylation with Acetyl Chloride in the Presence of Aluminum Chloride. The labeled methylcyclOpropane Z§ was acylated ac- cording to the procedure outlined by Hart and Schlosberg (27). An acetyl chloride-aluminum chloride complex was prepared in methylene chloride and filtered through a sintered—glass disk to remove any unreacted aluminum chlor— ide. Compound zg’was slowly bubbled through the ice—cold solution. After addition was complete, the solution was stirred at 50 for one hour. The reaction mixture was hydro- lyzed and the organic layer was isolated. Analysis by Vpc showed that five reaction products were present. Only two components could be isolated in sufficient quantities to be identified. The major component (53%) was found to be 5- chloro-2-hexanone-5-d1 §2 by comparison (ir, nmr) with an 64 unlabeled sample of 82, The compound had an nmr spectrum with a three-proton triplet (g = 0.9 Hz) at T 8.51 (C—6 protons), two sets of two-proton triplets (g_= 6.5 Hz) at T 8.04 and 7.42 (C-4 and C-3 protons) and a sharp three— proton singlet at T 7.92 (C—l protons). No signal was 7.92 O 7.42 8.04 Cl 8.51 so I CH3 "' C -’ CH2 _ CH2 - C " CH3 t D 80 observed at T 6.02 which is the chemical shift observed for the C-5 proton in unlabeled 82, The splitting patterns and relative peak intensities demonstrated that no deuterium scrambling had occurred. The second compound, which comprised 17% of the mix— ture, was found to be a labeled derivative of 2-chlorobutane, isomers 94 and 95; The compound had an nmr spectrum which showed a three-proton multiplet at T 8.99 (C—4 protons), a three—proton doublet (g_= 6.5 Hz) at T 8.54 (C-l protons), a sextet (g_= 6.5 Hz) at T 8.3 (C—3 protons) which had a relative area of 1.5 protons and a pentuplet (g_= 6.5 Hz) at I 6.15 which had a relative area of 0.5 protons. The nmr spectral data can be rationalized by a scheme which produces 2-chlorobutane with 50% deuterium at C-2 and C-3. Presumably the methylcyclopropane-l—d1 is isomerized to 2—butene—2-d1. Hydrogen chloride is present in the reaction mixture in significant amounts as an unavoidable impurity 65 D D D D HCl AlCl3 )\\/ -——9 N + 3) Cl 1 78 90 94 95 C AN NV rw rw in acetyl chloride. Addition of the hydrogen chloride to the labeled 2-butene may occur in two orientations. One results in the chlorine and deuterium atoms being attached to the same carbon as found in 22, while the other orienta— tion produces gg'with the chlorine and deuterium atoms on adjacent carbons. 2. Reaction ofyzglwith Hydrogen Bromide. A solution of methylcyclopropane-l—dl was prepared by passing 2.9 g of gaseous 18 into 30 ml of methylene chloride which was chilled to -5°. Anhydrous hydrogen bromide was slowly passed through the solution for one hour at ice-bath temperature. Analysis of the reaction mixture by vpc showed the presence of only one product, which was isolated and identified as 2-bromobutane-2-d1 (96). The compound had an nmr spectrum which had a three—proton Udplet (g_- 7 Hz) with secondary splitting (g_= 1 Hz) due to the asymmetric center at T 9.0 (C-4 protons) and a trip- let at T 8.37 (C-1 protons), which is coupled to the deu- terium (g.= 1 Hz). A two-proton multiplet at T 8.24 is attributed to the C-3 protons. A very weak signal was 66 observed at T 6.05 with a relative intensity of 0.02 protons 8.37 D 8.24 9.0 ! CH3 " C "' CH2 "‘ CH3 I Br 96 (C-2 proton). This band area was corrected for a maximum unlabeled methylcyclopropane residue of 2% in 28; The small amount of hydrogen at this position can be accounted for by the limited equilibration of a protonated methyl- cyclopropane. Rearrangement of the protonated intermediate D -‘i———> ;::H 78 + D Br 100 gt 2% D Br‘ D N6— /\1/ /\I;/ % 96 98 m to a secondary cation must occur more rapidly than equi— libration. Consequently the major product is 2-bromobutane- 2—d1 (96) formed from the secondary cation 102, while a small amount of scrambled product 92 is formed yia_equi- librating ions 91 and 98, An alternative explanation for the small amount of deuterium scrambling is a 1,2-hydride shift in ion 122, This process would also result in a pro— duct with hydrogen at C-2 (37,38). .vus nr ....v 'I (I) U: "”1 c 8‘. A . vi. ‘7‘? ".h 67 3. Reaction of Z§lwith Hydrogen Chloride. A methylene chloride solution of Z8 was prepared as described in the preceding section. Anhydrous hydrogen chloride was passed through the solution for three hours at 6°. The reaction mixture was allowed to stand for an additional 14 hours at 6°. Analysis of the reaction mixture by vpc showed the presence of one product, which was isolated and identified as 2-chlorobutane-2-d1 101; The nmr spectrum of compound 10; showed a three-proton triplet of triplets (g_= 7 and g.= 1 Hz) at T 9.0 (C-4 protons), a three—proton doublet (g_ 1 Hz) at T 8.56 (C-l protons), a two—proton quartet of triplets at T 8.35 (C-3 protons), and a very weak multip- let with a relative intensity of 0.03 protons at T 6.15 (C-2 proton). This band area was corrected for a 2% residue of unlabeled methylcyclopropane. Again this small amount of product can be accounted for H :+ I 6_ 'l H H H H 7 HCl : ‘:.’D 18 9,. 233. \ ‘ J/Cl- /\f/ <———“- N d? C1 + D 101 100 lgg 68 by the limited equilibration of protonated methylcyclo— propane 21 resulting in the formation of approximately 3% of rearranged product 102. D. Recent Related Studies and Mechanistic Conclusions. During the course of this research, several studies were reported which were also concerned with the protonated methylcyclopropane problem. The approach was different from that employed here, and the results supplement those acquired from the study of deuterated methylcyclopropane. The diazotization of several deuterated isobutyl- and segybutylamines in protic and aprotic solvents was studied by Friedman and Jurewicz (39,40). Methylcyclopropane was found as a minor product (3—15%) in the decomposition of both amines; the major products are isomeric olefins. Di— azotization of isdbutyl—Z-dl-amine 103 proceeded with £313 (EH3 D\ I H ---’CH2 D-C-CHZNHZ ——> /c’—‘=-=—1—CH2 ——> \ +’ \ ' CH3 \ I 103 CH3 T121 -—— CH { D\+ I” 2 _D+ /’ \ ‘79 d1 CHa—t do I H 98 w 100% deuterium retention in the methylcyclopropane fraction. Thus involvement of the methine hydrogen in an equilibrating 69 protonated cyclopropane (91’-¢'9§) is excluded. Limited equilibration between methyl hydrogens is possible. The extent of the equilibration of methyl hydrogens is given by the 3.8% methylcyclopropane-d7 which is formed in the D chD‘3 91377;?) CD3 “é‘CHzNHZ '_'> I: i:\ ——> \\\ I, &D CD3-? CH2 CD3-? CH2 3 D D 122 l 106 19,7; deamination of 2-methyl-d3-nfpropyl—2,3,3,3-d4-amine 105 . The small amount of deuterium lost can be attributed to equilibrating intermediates 101 and 108, A small amount of deuterium scrambling was also found in the protic and aprotic diazotization of sggfbutyl-4,4,4— d3-amine 109 . Approximately 2% of methylcyclopropane—d3 NH CD D --13 CD ---H ' 2 "er-‘3 C\i I \\2+ ’ CD3 -CH2-C-CH3 —-> ,{_ _ _\ CH3 -9 \ ,’ e-——‘—> \ ,’ £1 CH2— c’ Hz— C—H HZ— C-CH3 I | CH 109 H 110 1,13. 3 D112 -D+ -D+ —H+ a .. 9" V‘- a- 6.-. 'uxv“ m.“ xi». . Al ' i- v- ‘.V. s... ‘.| o o- r) I 9) 'v (D 4) fi‘, 70 was formed, illustrating the limited involvement of a partially equilibrating protonated methylcyclopropane (111 ::3_112). The major product, methylcyclopropane-d2, is formed from the methyl-bridged intermediate 119 and need not involve equilibrating ions 111 and 112, The conclusion thus may be drawn that a methyl-bridged species such as 104/ 126/ or 110 is needed to explain the presence of the methylcyclopropane formed in the diazo— tizations. As stated previously, these intermediates must then rearrange to the more stable secondary ion to form olefins faster than they equilibrate. Consequently, only a few per cent of the reaction proceeds via a protonated cyclopropane intermediate. The only cases discussed to this point have involved either methylcyclopropane or an isotopically labeled deriva— tive. In each of these cases, it appears that a secondary alkyl cation is preferred to a protonated methylcyclopropane. A further clarification of this principle is found in a recent report by Deno (41). Substituted cyclopropane- carboxylic acids were treated with deuterated sulfuric acid. In the case of 1—methylcyclopropanecarboxylic acid, the major product is tiglic acid 115. An average of 1.84 deuterium co H + 2 COZH >> C3H5(CH3)C02H ———> *’/Q§§<: 114 115* > /\/ 68% 113 o 32% 1 71 atoms were introduced into the molecule with 1.50 being attached to the methyls and 0.34 to C-3. The fractional incorporation and scrambling of the deuterium are both indicative of equilibrating protonated intermediates. Thus, the destabilizing effect of the electron—withdrawing carboxyl substituent is sufficient to allow the protonated intermediate to successfully compete with the classical cation. Another study reported by Deno (l8) involves a system more closely related to the work in this thesis. The product resulting from the treatment of methylcyclopropane with deuterium chloride was identified as 2—chlorobutane- 4-d1 (102). An equilibrating protonated methylcyclopropane DCl H Cl- H _'>HW -____€>H‘://\\jr’ D 1.11 C1 102 was discarded as an intermediate in favor of the classical 2—butyl cationIQZ, since no deuterium scrambling occurred as determined by nmr band area ratios (3:1:2:2) and split- ting patterns. Unfortunately, this system results in a product which is not easily analyzed for small amounts of deuterium on the carbon atoms, as the measurement of band areas is relatively insensitive to small changes in area ratios. In contrast, the methylcyclopropane-l—d1 system 72 results in products which have mostly deuterium on a methine carbon. Consequently, any signal detected for this carbon can be measured and attributed to an hydrogen migra— tion. Thus, product 122 has nmr band areas ratios which satisfactorily show that the gross process involves a classical 2-butyl cation. However, as found by this re— searcher and by Friedman (39,40), a small amount (2—5%) of the process may occur y13_an equilibrating protonated cyclopropane. Consequently, the reaction of methylcyclopropane with electrophiles proceeds mainly through a classical secondary cation. A very limited amount of the reaction may proceed through equilibrating protonated methylcyclo- propane intermediates. A methyl-bridged species may be involved as an intermediate ion which rearranges to the more stable secondary cation faster than it equilibrates. If the secondary cation is destabilized, a protonated methyl- cyclopropane becomes energetically favorable and acts as an important intermediate in the reaction mechanism. EXPERIMENTAL A. Preparation of Methylcyclopropane-l—d1 (18) 1. Preparation of Deuterium Bromide. The procedure of Brown and Groot (42) for the prep- aration of deuterium chloride was followed with appropriate modifications. To a 500—ml 3-necked flask fitted with a Graham condenser, thermometer, dropping funnel and mag- netic stirrer was added 29.6 g (0.16 mole) of benzoyl bromide. The solution was heated to 70° as 3.2 g (0.16 mole) of deuterium oxide was added dropwise. The gas flow was measured with a manometer. After all of the deuterium oxide was added, the solution was slowly heated to 130°. The generated gas was collected in a Spiral trap cooled to -78°. The system was flushed with nitrogen to remove the last traces of deuterium bromide. A total of 12.5 g (95%) of product was obtained. 2. Preparation of 1-Bromo-3—chloro—Z—methylpropane— 2-d1 (22,). To a 100-ml 3-necked flask fitted with a gas inlet tube, thermometer and reflux condenser protected with a drying tube was added 90.5 g (1 mole) of 3-chloro-2-methylpropene 92’ 73 74 (Aldrich Chemical Company) and 6 g of benzoyl peroxide. The stirred solution was maintained at 00 with an ice—salt bath. Deuterium bromide was slowly bubbled into the solu— tion over an hour. Upon completion of addition, the solu- tion was stirred at room temperature for an additional 30 minutes. The solution was washed(2 x 100 ml)with water, saturated sodium bicarbonate, 15% ferrous sulfate, and water again, dried with magnesium sulfate and distilled. A yield of 140 g (82%) of 1-bromo-3-chloro-2-methylpropane— 2-d1 (99), bp 47° (14 mm) was collected. The nmr spectrum (CC14, Figure 19) of the monodeuterated compound 91 had signals at T 8.90 (t, g_= 0.9 Hz, 3H), 6.55 (t, g_= 0.9 Hz, 2H), and T 6.45 (t, g_= 0.9 Hz, 2H). An extremely weak sig- nal could be observed for the methine proton at T 7.86 at greatly increased amplitude. Due to the low intensity of this signal, no accurate measure of its relative intensity could be obtained. The mass spectrum showed parent peaks at m/e 175, 173, and 171 with the 173 peak being the most intense. 3. Preparation of Methylcyclopropane-1—d1 (29). The procedure of Greenlee and Boord (35,36) was used for this preparation. To a 250-ml 3-necked flask equipped with a mechanical stirrer, dropping funnel, and reflux condenser was added 210 g (3.20 mole) of zinc dust and 300 ml of 95% ethanol. The vessel was heated until gentle reflux was achieved, whereupon the solution was vigorously 75 stirred as 137 g (0.8 mole) of compound 99 was added. Re- flux was continued for 24 hours. The methylcyclopropane— l-d1 was collected yi§_a gas outlet tube attached to the reflux condenser which led to a spiral trap cooled to -78°. After 24 hours the reaction temperature was increased to a vigorous reflux to remove any dissolved product from the solution. The system was swept with a gentle stream of nitrogen to remove the lasttraces of product. Purification was accomplished by bubbling the gas through a neutral solution of potassium permanganate followed by distillation through a 2-ft vacuum-jacketed Vigreaux column fitted with a Dry Ice cold finger. A yield of 43 g (92%) of methyl— cyclopropane-1-d1 (Z9) was obtained after purification. Compound 29 had an ir spectrum (CCl4, Figure 20) with im— -1 portant absorptions at 3070 (vC-H) and 2250 (v ) cm and C-D an nmr spectrum (CC14) with signals at T 9.98 (m, 2H), 9.52 (m, 2H) and T 8.98 (br s, 3H). The mass spectrum contained a parent peak at m/e 57. (Figure 21.) B. Reaction of Methylcyclopropane-l—dl with Hydrogen amiss-2 To a 100-ml 3-necked flask fitted with a cold finger condenser cooled with Dry Ice, thermometer and gas delivery tube was added 30 ml of methylene chloride. The solution was cooled to -5° with a salt-ice bath as 2.9 g (0.051 mole) of Z9 was passed into the solvent. After addition was 76 complete, hydrogen bromide was slowly bubbled through the solution for one hour. The reaction mixture was washed 2 x 30 ml with water and saturated sodium bicarbonate, dried with sodium sulfate and the solvent was removed by fractional distillation. Analysis of the reaction mixture by vpc (10' x 1/4" SE-30, 100°, 60 ml/min of He) showed the presence of only one product. The compound was puri— fied by vpc (above conditions) and identified as 2-bromo- butane-Z—d1 92, The ir spectrum (CCl4) of 99 had bands at 2950 and 2200mm”1 (C-D) and the nmr spectrum (cc14, Fig- ure 22) had signals at T 9.0 (t of t, g_= 7, g_= 1 Hz, 3H), 8.37 (d, g = 1 Hz, 3H), 8.24 (m, 2H), and T 6.05 (m, 0.4H uncorrected). C- Reaction of Methylcyclopropane—1—d1 with Hydrogen Chloride The reaction was performed according to the procedure described above. However in this case, after allowing the hydrogen chloride to bubble through a solution of Zé'for 3 hours, the reaction mixture was allowed to stand at 6° for an additional 14 hours. Analysis of the reaction mix— ture by vpc (10' x 1/4" SE-30, 100°, 60 ml/min of He) showed the presence of one product, which was identified as 2- chlorobutane-Z-d1 (191). Compound 191 had an ir spectrum 1 (cc14) with bands at 2950 and 2200 cm’ (C-D) and an nmr spectrum (CC14, Figure 23) with signals at T 9.0 (t of t, J = 7, g'= 1 Hz, 3H), 8.56 (d, g_= 1 Hz, 3H), 8.35 (q of t, 77 g_= 1 Hz, 2H), and T 6.15 (m, 0.05 H uncorrected). A control experiment was performed using unlabeled methylcyclopropane and employing experimental conditions identical to those used on the labeled compound. Analysis by vpc showed only one product was formed in the reaction. The product had identical ir and nmr spectra and R (10' x t 1/4" SE-30, 100°, 60 ml/min of He) as an authentic sample of 2-chlorobutane. D. Acylation of Methylcyclopropane-l-dl with Acetyl Chloride To a 3—necked 500-ml flask equipped with a sintered glass outlet on the bottom, mechanical stirrer, dropping funnel and drying tube was added 10.1 g (0.08 mole) of aluminum chloride suspended in 30 ml of cold methylene chloride. This suspension was stirred as a solution of 6.5 g (0.083 mole) of acetyl chloride in 20 ml of methylene chloride was added. The mixture was stirred for 30 minutes after addition was complete. The solution was added to a 300-ml 3-necked flask y}g_the sintered glass outlet. The 300-ml flask was equipped with a cold finger condenser cooled with Dry Ice, magnetic stirrer, gas-inlet tube and thermometer. The solution was maintained below 100 with a salt-ice bath. Methylcyclopropane—1—d1 (4.3 g, 0.076 mole) was bubbled into the cold solution over a 35 minute period. The temperature was maintained below 5° for an hour after addition was complete. 78 The reaction mixture was poured into a stirred mixture of 100 g of ice and 40 m1 of hydrochloric acid. After hydrolysis was complete, the organic layer was washed(2 x 50 ml)with water, saturated sodium bicarbonate and again with water, dried with sodium sulfate, and the solvent removed under reduced pressure. The concentrated sample was kept at -78° to prevent decomposition. Analysis of the solution by vpc (10' x 1/4" SE-30, 120°, 60 ml/min of He) showed the presence of five products with R of 2, 4, t 6, 12, and 15 minutes. The compound with the longest Rt’ which was also the major product (58%), was purified by vpc (10' x 3/8" ov-25, 1100, 200 ml/min of He). The compound was identified by analysis of its spectral properties (nmr and ir) as 5—chloro-2-hexanone-5-d1 (99). Compound 99 had identical R as an unlabeled sample. The compound had an t ir spectrum (CCl4) with absorptions at 2950 (VG-H) and 2200 (v ) cm—1 and an nmr spectrum (CCl4, Figure 24) with C-D signals at T 8.51 (t, g-= 0.9 Hz, 3H), 8.04 (t, g_= 6.5 Hz, 2H), 7.92 (s, 3H), and T 7.42 (t, g.= 6.5 Hz, 2H). A second compound with a Rt of 2 minutes was isolated by Vpc (above conditions) and identified as deuterated 2- chlorobutane 94 and 99, The nmr Spectrum (CCl4, Figure 25) contained signals at T 8.99 (m, 3H), 8.54 (d, g_= 6.5 Hz, 3H), 8.3 (sextet, g_= 6.5 Hz, 1.5H), and T 6.15 (pentuplet, J = 6.5 Hz, 0.5H). No additional volatile products could be isolated. The expected products are firchloroketones, 79 which rapidly polymerize at room temperature in concentrated solutions. Thus, attempts to isolate these chloroketones by vpc or by distillation resulted in polymerization of the components. SUMMARY 1. A 92% yield of methylcyclopropane—l-d1 (19) was ob- tained by cyclizing 1-bromo-3—chloro-2-methylpropane-Z- d1 (99) with zinc. Mass spectroscopy showed that (29) contained 98% deuterium. 2. Treatment of Z9 with acetyl chloride-aluminum chloride gave 5-chloro-2-hexanone-5-d1 99’as the major product. A detailed mechanistic consideration of the possible deuter- ated products indicates that the formation of 99 with unscrambled deuterium must occur from a classiCal cationic intermediate. A protonated methylcyclopropane intermediate is not involved. 3. Reaction of Z9'with anhydrous hydrogen bromide gave only one product, 2-bromo-butane—2-d1. Spectroscopic analysis showed that 2% deuterium scrambling occurred. The limited equilibration of a protonated methylcyclopropane intermediate could account for the small amount of scrambled deuterium. The scrambling may also occur as a result of 1,2-hydride (deuteride) shifts in the 2—butyl cation inter- mediate. 4. Similarly, treatment of zg’with anhydrous hydrogen chloride produced only 2—chlorobutane-2—d1. Spectral 80 81 analysis revealed that 3% of the deuterium had migrated from the original carbon. The intermediacy of a protonated methylcyclopropane is not necessary to explain the data. Part III The Photochemistry of Arylnitrile Oxides 82 INTRODUCTION The thermal reactions of nitrile oxides have been well characterized (43) since the first nitrile oxide was pre- pared by Werner in 1894 (44). It was found that they are among the most reactive species in organic chemistry. They readily undergo 1,3-addition reactions with olefins to R R3 R1 R cm 0 + {c C/ R R _ = _H = 2 . \ \ I R R2 R4 7 3 \ R 4 119 120 121 W produce 4,5-dihydro-1,2-oxazolenes 121. In the presence of carbonyl compounds, 119 cleanly produCes 1,2,4—dioxazoles R1 R o \ || R-CEN“>O + O = I O/LRJ, . N\ R2 R2 119 122 LEE 123. Analogous reactions are observed for imines or nitriles N R 0' N -s— + I R C N >0 ? :>' h” ILLx R \o R 119 124 ltzi 83 84 to yield 1,2,4-oxadiazolenes 125. The nitrile oxide func— tion is also susceptible to nucleophilic attack, but weak nucleophiles such as the halogens do not add to the nitrile l/N-OH R—CEN—eo -—3EEL——> R—c’/ ‘\ Nu 119 126 W W Nu: = CH”, R0", Hx, H23, H20, RMgX, HNRZ. oxide even under forced conditions. The nitrile oxide function may be reduced with zinc in acetic acid to give the corresponding nitrile in greater than 90% yield. With few exceptions, nitrile oxides also dimerize I} I? R R f; + {j > IL lml i ( Cr, 0 O O 221 1,132, 1,12, spontaneously to 1,2,5-oxadiazole 1-oxides (furoxans) 191. Simple aliphatic nitrile oxides undergo this reaction ex- tremely fast, whereas aromatic nitrile oxides dimerize within hours or days (45). Aromatic nitrile oxides which have strong electron withdrawing groups such as pr02 or EfCl are considerably more stable than the parent compound, benzonitrile oxide (25 days for pr02 versus 30-60 minutes for phenyl). The dimerization, which occurs by a 1,3-cyclo— addition process, is apparently retarded due to the reduc- tion of electron density at the nitrile oxide moiety in the pr02 and prl cases. 85 An alternate means of stabilizing nitrile oxides is the introduction of sterically hindering groups at the Egg? positions of the aromatic residue. Consequently, 2,4,6-trimethylbenzonitrile oxide 128’may be isolated as a white solid and is stable indefinitely at room temperature CENTeo 128 1M (46). The bulky methyl substituents prevent dimerization to a furoxan, yet allow all other addition reactions to occur. Two examples have also been reported of sterically encumbered nitrile oxides involving substituents other than simple aliphatic groups. Just and Dahl prepared O-methyl- podocarponitrile oxide 129 by oxidation of podocarpinal OCH3 syn-oxime with lead tetraacetate (47). The second example was the bridgehead nitrile oxide 130 synthesized by Ranganathan (48). 86 A I' Br C ll| N 1 1.49, 0 While none of these sterically hindered nitrile ox— ides dimerize to furoxans, a thermal isomerization of the nitrile oxide moiety does occur at elevated temperatures. Refluxing 199 or 199 in xylene results in the complete isomerization of nitrile oxide to isocyanates 191 or 133' 0\ GEN—b0 d;N N=C=° A _____€;’ xylene 3> 128 132 131 A k A A It ———> It ‘. 'r xylene > Br Br ' c N c x’ Ni—‘o " N ‘5 1 130 134 o 133 0 WV NW WV with no furoxan formation. The mechaniSm postulated for these processes involves oxazirine intermediates 132 and 134. The hydrocarbon residue undergoes a 1,2-shift, subse- quent to or concurrent with ring opening to isocyanate. 87 ReCently, Just and Zehetuer (49) found that irradiation of 128 also results in isomerization of the nitrile oxide group to an isocyanate function. Again a mechanism O GEN—>0 H N=C =0 hv [ill] iilll ;> + 128 135 131 W W involving oxazirine 132 was proposed. The formation of lactam 135 supports the mechanism, as the intermediate 132 may undergo an electron rearrangement to yield nitrene 136, 8 I} C—N C‘ : H 132 136 rm which is in excellent position to insert into the neigh- boring alkyl group to form lactam 199, Thus, thereame sufficient chemical data to believe that oxazirine 199 is involved in the thermal and photo- chemical isomerization of nitrile oxides to isocyanates. However, no spectroscopic evidence has been found to substantiate the chemical data. Consequently, Part III of this thesis is primarily concerned with the spectro- scopic detection of intermediates in nitrile oxide photo- chemical reactions. 88 While the primary concern of this study was the detection of an oxazirine intermediate, the examination of nitrile oxides as possible precursors for photolytic oxygen atom transfer was also of interest. Several ex- amples of photolytic oxygen transfers have been observed (50,51,52) upon irradiation of aromatic N-oxides. Jerina has found that oxygen atom transfer results when aromatic- <' w + 7 139 140 Z 030‘— N—oxides are irradiated (50). Two oxidized products 122. and 132 are obtained in low yield upon irradiation of pyridine—N-oxide in the presence of naphthalene. The irradiation of substituted pyridazine N-oxides produces more efficient atomic oxygen transfer. A total yield of 20-30% of oxidized product was obtained from the irradiation of 141 in the presence of an excess of cyclo- O O .-— hV \t + \/ / ——N l 0 O 141 22,2. 22,3; hexene [(51) and references therein]. A yield of 20—40% of deoxygenated pyridazines was obtained in all cases. 89 Oxygen atom transfer should also be observed for nitrile oxides, since the nitrile moiety is more electron withdrawing than a pyridine ring. Consequently, this part of the thesis is also concerned with the photolytic oxygen atom transfer as a result of irradiation of nitrile oxides. RESULTS AND DISCUSSION A. Preparation of pfchlorobenzonitrile Oxide (144) Since only one example of the pmxnchemical isomeriza- tion of nitrile oxide to isocyanate has been reported (49), it was necessary to study several other cases to establish the generality of the conversion. Two nitrile oxides, 3&4 and 145, were synthesized which were less stable than the‘ GEN—>0 GEN—>0 C C1 C1 144 145 literature example 12g; Whereas lgg'cannot dimerize to furoxan, lgg’undergoes this decomposition reaction in 2-3 days and 12§,in several weeks. These compounds were also chosen because of the electron withdrawing nature of the chlorine atoms. It was anticipated that this effect would lower the electron density of the nitrile function, thus facilitating loss of the oxygen atom from the nitrile oxide moiety. The general procedure of Wiley and Wakefield (53) was employed for the synthesis of 144. The oxime of 90 91 pfchlorobenzaldehyde 146 was chlorinated by passing a Hx _N,0H Cl~c= /OH €12 9 0H E C1 C1 C1 1.4.9. La 1,44. moderate stream of chlorine through an anhydrous chloro— form solution of lgg'until the green color persisted. A 70% yield of pfchlorobenzohydroxamic chloride 141 was ob— tained after recrystallization. Compound 122.13 hygroscopic but otherwise stable and may be stored indefinitely in a desiccator. Since nitrile oxide lgg'undergoes noticeable dimerization in one day, it was prepared fresh as needed from precursor 121, A solution of nitrile oxide 144 was easily prepared by treatment of an ice-cold suspension of 122.1“ carbon tetrachloride with 15% sodium hydroxide. The carbon tetrachloride layer showed strong infrared absorp- tions at 2290 (CEN stretching), 1380 (N —> O) and 1100 cm-1 'which are characteristic of pfchlorobenzonitrile oxide (53). It is not possible to isolate this nitrile oxide due to its propensity to dimerize. It can be stored in dilute solution at 00 for several hours without noticeable decom- ‘position. Complete dimerization of 1440to furoxan 14g] results upon refluxing a solution of 144 for 30 minutes. 92 GEN—>0 A Cl 144 148 B. Irradiation of prhlorobenzonitrile oxide (144) A 4% carbon tetrachloride solution of 144 was irradi— ated with an Ultra Violet Products PCQ-XI lamp for four hours at 23°. The solution was maintained at 23° to mini- mize dimerization of lég'to furoxan. The photolysis was followed by infrared spectroscopy. Analysis of the photolysate by vpc revealed the presence of two photoproducts in approximately equal amounts. These compounds were identified as pfchlorophenyl isocyanate 149 and pfchlorobenzonitrile 150. Since both hV a + C1 C1 C1 144 149 150 149 and 150 are known compounds, they were identified by comparison of their physical and spectral data with those reported in the literature. Isocyanate 149 was an oil with .. _1 characteristic strong infrared absorption at 2260 cm . 93 Nitrile lég'was a white solid with characteristic weak infrared absorption at 2240 cm-1. A white precipitate formed upon concentration of the irradiated solution. This nonvolatile solid was identified as 3,4-di(pfchlorophenyl)furoxan lég'by comparison (mp and ir) with an authentic sample. Compound 148 is the expected product from the thermal dimerization of 144; The mechanism for the photochemical pr0cess has not been completely elucidated. Formation of isocyanate 149 /\ hv ¥ \ f ,7 av aryl :> Cl C1 C1 144 151 149 can be explained by postulating a mechanism involving oxazirine 151. Ring opening and migration of the aryl ring to the nitrogen atom produces 151. The formation of nitrile lég'represents the formal loss of atomic oxygen from nitrile oxide 144; The process by which this loss occurs has not been determined (51). The formation of approximately 50% deoxygenated nitrile is in agreement with the results obtained from the irradiation of pyridazine-N-oxides (51). In these cases, 20-40% yields of deoxygenated pyridazine were produced. Consequently, the loss of an oxygen atom from nitrile oxides can be in- duced photochemically. 94 Attempts to intercept the generated oxygen atom were not entirely successful. Irradiation of a carbon tetra- chloride solution of 142.containing equimolar amounts of naphthalene was performed. Analysis of the photolysis solution by vpc revealed the presence of the same compounds that were obtained in the absence of naphthalene. An addi- tional product was observed which had an identical Rt as 2-naphthol. The compound comprised approximately 1% of the total product. Consequently, it was not possible to isolate a sample for positive identification. Other oxygen acceptors such as anthracene and hexamethylbenzene were also employed. No oxidized products could be detected. C. Preparation and Irradiation of 2,4-Dichlorobenzonitrile Oxide (145) The procedure employed for the synthesis of 145 was identical to that outlined for the preparation of 144. c1 \. PLC=N_OH C=N-OH GEN—>0 C C1 Cl _ Cl 2 OH \ CHCI3 ; — HCl 7 C1 C1 C1 1 2 153 145 Treatment of oxime 152 with chlorine produced hydroxamic chloride 153. Baseéinduced elimination of HCl from 153 results in'a nearly quantitative yield of the desired nitrile oxide 145. Compound 145 was found to be a relatively stable 95 nitrile oxide, as it could be isolated as a white solid, mp 74-76°. The infrared spectrum of 145’showed strong bands at 2320 (can), 1380 (N —> o) and 1150 cm-1. Com— plete dimerization of lgé’occurred in three hours at 78°. By comparison, nitrile oxide lég'dimerized in 30 minutes at the same temperature. A 5% carbon tetrachloride solution of 145 was irradi- ated through quartz with an Ultra Violet Products PCQ—XI lamp at 23°. Complete conversion of lgé’occurred in five hours as evidenced by the disappearance of the infrared band at 1380 cm-1. Analysis of the photolysate by vpc showed the existence of two volatile products in a 2:1 ratio. The major component was identified as 2,4-dichloro- phenyl isocyanate 154 by comparison (ir and mp) with the GEN->0 I N=C=O CEN cl ' C c1 hv ;> + C1 C1 C1 145 154 155 literature values. The minor component was established as 2,4—dichlorobenzonitrile 155 by comparison of its physical properties with literature values. 96 D. Preparation and Irradiation of 2,4,6-Trimethylbenzo- nitrile Oxide (128) The procedure of Grundmann (46) was used for the prep- aration of 128. Treatment of a dimethylformamide solution of 156 with sodium methylate and N-bromosuccimide resulted H\C=N_OH CEN‘$O NaOCH3 :> NBS DMF 156 128 in a 90% yield of nitrile oxide 128; The stable white com- pound, mp 111°, showed strong infrared bands at 2310 (CEN), 1360 (N —> O) and 1090 cm-1. As discussed previously, 1§§.is stable indefinitely at room temperature. Recrystal- lization from ligroin is possible without dimerization. Irradiation of a 1% carbon tetrachloride solution of lgfi’through quartz was performed with an Ultra Violet Products PCQ-XI lamp. Complete conversion of the nitrile oxide lfifiito products was complete in five hours as evi— denced by the disappearance of the infrared band at 1360 cm_1. Analysis of the photolysate by vpc showed the presence of two volatile products in a 4:1 ratio. The major component was identified as 2,4,6-trimethylphenyl isocyanate 131 by comparison (mp and ir) with literature values. The minor component was established as 97 0 'HH N=C=O CEN 131 157 135 W W rwv 2,4,6-trimethylbenzonitrile 157 by comparison (mp and ir) with literature values. A third product, isolated from the solution by silica gel chromatography, was identified as 2,4-dimethylphthal— imidine 135, Absorption in the infrared region at 3220 (N-H), 3100 (N-H) and 1685 (c=0, 5-membered conjugated lactam) cm.1 and an nmr spectrum with three-proton reso- nances at T 7.60 and 7.30 (benzylic methyl groups), a two- proton signal at 1 3.00 (phenyl protons) substantiated the structural assignment. Thus, photoproducts lgl’and 122 are identical to those found by Just (49) upon irradiation of 128, The deoxygen— ation product lézihad not been previously reported as a photoproduct of 128, The three nitrilecnddes, 144, 145 and 128, examined in this study produced analogous photopro- ducts, isomerization to an isocyanate or deoxygenation to a nitrile. As previously stated, the formation of iso- cyanate most likely occurs yi2_an oxazirine intermediate. The generality of this rearrangement has now been established. Thermally, only nitrile oxides which are sterically pro— hibited from dimerization isomerize to isocyanate. However, 98 photochemically all nitrile oxides examined rearranged to isocyanate regardless of their thermal stability. Deoxy- genation to yield nitriles is also a general phenomenon observed upon irradiation. Mechanistic studies have not been performed to elucidate the path by which loss of Oxygen OCCUIS . ‘0'. l'; A low temperature study of the photolysis of nitrile L oxide 128 was carried out in an attempt to observe the oxazirine intermediate 132. A 5% solution of 128 in 2- 4 methyltetrahydrofuran was cooled to -170° in a low tempera— - I A ture infrared cell. The apparatus used in this study had the capability of allowing successive irradiation and infra- red scanning. Irradiation of the cooled solution for ten minute intervals with a 450 watt Hanovia lamp was conducted for a total irradiation time of 30 minutes. The photolysis was monitored by infrared spectroscopy immediately after each ten minute interval. No change in the spectrum occur— red in the 1600-2000 cm-1 region. Analysis of the solution after irradiation was complete showed that the reaction had proceeded to 25% conversion. Thus, no infrared absorption which could be assigned to the oxazirine intermediate Egg was detected. The absence of any spectral manifestations of lfig'may result because of the relative instability of 132 even at -170°. Consequently, no appreciable concentra- tion of the intermediate accumulates in the irradiated solu- tion. EXPERIMENTAL A. Preparation of prhlorobenzonitrile Oxide (144) To 7 g of hydroxylamine hydrochloride dissolved in 40 ml of water was added 14 g (0.1 mole) of pfchlorobenzalde— hyde. The slurry was stirred at 0° as 25 g of 40% aqueous sodium hydroxide was added dropwise. The thick slurry was neutralized with concentrated hydrochloric acid (approx. 10 ml). The oxime may be purified by recrystallization from benzene to give pure pfchlorobenzaldoxime, 146, in 90% yield, mp 109-1100 [lit. (54) mp 11001. The crude oxime was dissolved in 400 ml of chloroform which was then cooled to 0°. A moderate stream of chlorine was passed through the solution for approximately 40 minutes or until the green color persisted. A pale yellow solid was collected after the chloroform was removed under reduced pressure. The crude material was recrystallized from ligroin giving pure pfchlorobenzohydroxamic chloride, 141, in 70% yield from 146” mp 86-89° [lit. (53), 82—86°]. To 2.0 g (0.01 mole) of pure 141 suspended in 40 ml of CCl4 at 0° was added 6 ml of 15% sodium hydroxide. The solution was stirred at 0° for 30 minutes. The organic layer was separated and dried over anhydrous sodium sulfate 99 I' "A ' YJTT 111.: all. 100 at 0°. The organic layer showed strong infrared absorp- tions (cc14) at 2290 (CEN aromatic), 1380 (N —> o) and 1100 1 . ‘ . . . . cm (para substituted benzene) which are indicative of pf chlorobenzonitrile oxide 144 (53). B. Preparation of 3,4-Di(pfchlorophenyl)furoxan (148) A CCl4 solution of 144 from 1.0 g of 147 was warmed on a steam bath for 30 minutes. The CCl4 was removed, leaving 0.65 g (81%) of solid 148, mp 141-1430 [lit. (53), 144- 145°]. C. Irradiation of prhlorobenzonitrile Oxide (144) A 4% CCl4 solution of 122,35 prepared in A was irradi— ated in a quartz test tube with an Ultra Violet Products PCQ-XI lamp for four hours at 23°. The solution was cooled by means of a cold finger water condenser to minimize thermal dimerization of the nitrile N—oxide to furoxan. The photo— lysis can be followed by observing the disappearance of the 1380 cm"1 (N —4 0) band and a shift of the 2290 (can) band to 2260 cm-1. The solution is yellow after the photolysis. The reaction mixture was concentrated, at which time 0.5 g of a white solid crystallized. The solid was removed and found to be identical (ir and mp) with an authentic sample of 3,4-di(pfchlorophenyl)furoxan, 148, The remaining solution was analyzed by vpc (10', 3% OV-210, 120°, 100 Inl/min of He) which showed the presence of two other major compounds in approximately equal amounts. 101 The compound with the shorter retention time (6.5 min) was a colorless, pungent oil which was identified as pf chlorophenyl isocyanate, 142; The structure was assigned by comparison with the infrared spectrum as found in Sadtler Standard Spectra, Midget Edition, #22271. Absorp— tions were found at 2260 (very strong, N=C=O) and 1590, ,1 1510 and 1090 cm—1 (benzene ring). I Juli-T" The compound with the longer retention time (12 min) ‘vjj‘l was a white solid, mp 90-90.5°. It was identified as 27 h chlorobenzonitrile, 152 [lit. (54) mp 90°]. It had an ir spectrum (CC14) with bands at 2240 (weak, CEN) and 1600, 1490, and 1100 cm.1 (benzene ring). The mass spectrum had parent peaks at 137 and 139 amu. D. Preparation of 2,4-Dichlorobenzonitrile Oxide (145) Compound 145 was prepared according to the procedure outlined in A, substituting 2,4-dichlorobenzaldehyde for pfchlorobenzaldehyde. The oxime of 2,4-dichlorobenzalde— hyde was obtained as a white solid mp 136° [lit. (54), 136-137°] after recrystallization from ligroin. Chlorin- ation of the oxime produced hydroxamic chloride 153, mp 94-95° after recrystallization from ligroin. The yield of lég’from the aldehyde was 35%. Treatment of a carbon tetra- chloride solution of 153 with 15% sodium hydroxide produced nitrile oxide 122” which could be isolated as a white solid, mp 74-76°. The infrared spectrum (CC14) showed absorptions 102 1 at 2320 (strong, ) 1380 (strong v ) and 1150 cm- . _ VC's-N N->o When 145 was refluxed in carbon tetrachloride for three hours, complete conversion to 3,4-di(2,4-dichlorophenyl)— furoxan resulted. The nitrile oxide could be stored at 0° for two days without noticeable dimerization. E. Irradiation of 2,4-Dichlorobenzonitrile Oxide (145) A 5% carbon tetrachloride solution of 145 as prepared in D was irradiated in a quartz test tube for five hours with an Ultra Violet Products PCQ-XI lamp at 23°. The dis— appearance of the 1380 cm—1 (N-vO) band and the shift of 1 1 the 2320 cm— (CEN) band to 2300 cm- (N=C=O) was observed when the photolysis was monitored by ir spectroscopy. Analysis of the solution by vpc (5' x 1/4" 15% 0v-25, 1400, 100 ml/min of He) indicated the formation of two major volatile products in a 2:1 ratio with retention times of 9 and 13 minutes respectively. The photoproducts were purified by vpc (above conditions). The major component was shown to be 2,4-dichlorophenyl isocyanate. The mp (56-58°) and ir spectrum (CCl4) 2300 cm-1 (N=C=O) correspond to the litera- ture values (57). The minor compound was a white solid, mp 59—610. It was identified as 2,4-dichlorobenzonitrile [lit. (55) mp 61°]. The ir spectrum (CC14) had bands at 2250 (weak ceu) and 1590, 1485, and 1110 om’1 (benzene ring). 103 F. Preparation of 2,4,6-Trimethylbenzonitrile Oxide (128) The nitrile oxide was prepared by the procedure of Grundmann (56). To 25 g (0.167 mole) of 2,4,6-trimethyl- benzaldehyde dissolved in warm methanol was added in one batch a warm solution of 25 g of hydroxylamine hydrochlor— ide and 48 g of sodium acetate in 120 ml of water. The clear solution was warmed on a steam bath for 15 minutes i and poured into 350 ml of ice water. The white solid was : collected and washed twice with ice water. The oxime was i dried by dissolving it in chloroform, treating it with 4* magnesium sulfate and evaporating the solvent under reduced pressure. To 0.815 g (0.005 mole) of 2,4,6-trimethylbenzal- doxime dissolved in 15 ml of N,N-dimethylformamide was added 0.27 g of sodium methylate and the solution was cooled to 10°. A solution of 0.89 g (0.005 mole) of N- bromosuccinimide in 5 ml of N,N-dimethylformamide was added over a 10-minute period. Stirring was continued for 30 minutes after addition was complete. The solution was diluted with ice water until the nitrile oxide started to crystallize, kept at 0° for several hours, filtered and washed with ice water. After recrystallization from ligroin, the stable white compound melted at 110.5-111O [lit. (56) mp 111°] and showed ir bands (CC14) at 2310, 1360, and 1090 —1 cm . 104 G. Irradiation of 2,4,6-Trimethylbenzonitrile Oxide (128) A solution of 600 mg of 128 in 40 ml of carbon tetra— chloride was irradiated for 5 hours with an Ultra Violet Products PCQ—XI lamp. The disappearance of the 1360 (N—eo) and 2310 cm-1 (CEN) bands was observed when the photolysis was monitored by ir spectroscopy. Analysis by vpc (5' x 1/4" 15 ov-25, 150°, 100 ml/min of He) showed the presence of two major volatile products in a 4:1 ratio with retention times of 8 and 11.5 minutes respectively. The photoproducts were purified by vpc (above conditions). The compound with the shorter R comprised the majority t of the mixture and was identified as 2,4,6-trimethylphenyl isocyanate, mp 43—45° [lit. (56) mp 45°]. The ir spectrum (CCl4) contained bands at 2310 (strong, N=C=O), 1520 and 1090 cm.1 (benzene ring). The compound with the longer Rt was a white solid, mp 47-50° [lit. (58) mp 50°], and was identified as 2,4,6—trimethylbenzonitrile. It had ir bands (cc14) at 2950 and 2250 cm“1 (CEN). Silica gel chromatography of the photolysis solution allowed two additional minor products to be isolated. Elution with the following series of solvents, hexane, carbontetrachloride, methylene chloride and chloroform, effected an adequate separation of the two compounds. Further purification was accomplished by vpc (2' x 1/4" 15% OV-25, 200°, 100 ml/min of He). The compound with the shorter Rt (6 min) was 2,4-dimethylphthalimidine. The 105 ir spectrum (CC14) had bands at 3220, 3100 (N-H) and 1685 cm.1 (c=o, 5-membered conjugated lactam) and nmr spectrum (CDCla) contained signals at T 7.60 and 7.30 (br s, 3H, benzylic methyls), 5.67 (br s, 2H, methylene protons) and T 3.00 (br s, 2H, phenyl protons). The mass Spectrum showed a parent peak at m/e 161. The structure of the component with the longer R (10 minutes) was not deter- t mined. The infrared spectrum (CC14) of the sample had absorptions at 2950, 1730, 1470 and 1190 cm-1. H. Irradiation of 2,4,6-Trimethylbenzonitrile Oxide 128 at Liquid Nitrogen Temperature A 5%;solution of 128'in 2-methyltetrahydrofuran was cooled to approximately -170° with liquid nitrogen, forming a transparent glass. The solution was contained in a 0.1 mm KBr infrared cell. The cell was positioned in a metal ring, which was in direct contact with a reservoir of liquid nitrogen. The temperature of the infrared cell was lowered to the desired level within 30 minutes. The KBr cell was protected from atmospheric moisture by an outer cell which contained three ports. Two of the ports were fitted with KBr windows to allow scanning in the infrared region. The third port, at right angles to the other two, contained a quartz (Suprasil 1) window to allow ultraviolet irradiation of the sample cell. The windows of the outer cell were heated to prevent water condensation. 1| .til 5-137;. ‘MrT-Z..—— ‘14 106 Irradiation of the solution of lgg’with a 450—watt Havonia Type L mercury lamp was performed in 10 minute intervals. Immediately after each irradiation interval the infrared Spectrum from 1600 to 2000 cm-1 was recorded A total irradiation time of 30 minutes produced no detect— able change in this region of the Spectrum. The irradiated sample was warmed to room temperature and diluted with carbon tetrachloride. The solution was washed 3 x 100 ml with water to remove the 2—methyltetra- hydrofuran. Examination of the relative intensities of the 2300 and 1360 cm.1 infrared bands Showed that a 25% con- version of the nitrile oxide 128 to product had occurred. w- . ." .' _.¢v-’u 4 o SUMMARY 1. A carbon tetrachloride solution of pfchlorobenzonitrile oxide (144) was prepared by treatment of pfchlorobenzald— oxime with C12. The resulting hydroxomic chloride was dehydro-halogenated with base to yield 144; Irradiation of a carbon tetrachloride solution of 144’resulted in two photoproducts, pfchlorophenyl isocyanate (142) and pfchloro— benzonitrile (152), in approximately equal amounts. The photoisomerization to isocyanate has been postulated to involve an oxazirine intermediate. 2. The preparation of 2,4-dichlorobenzonitrile oxide (145) was accomplished by treating the corresponding hydroxamic chloride with base. Nitrile oxide 145 was found to be rela— tively stable and could be isolated as a white solid. Irrad— iation of a carbon tetrachloride solution of 145 produced two photoproducts, 2,4-dichlorophenyl isocyanate (154) and 2,4- dichlorobenzonitrile (155), in a ratio of 2:1. The photo- isomerization of nitrile oxides to isocyanates and deoxygen— ation to nitrile was established as a general photoreaction. Attempts to effect an oxygen atom transfer were unsuccessful. 3. The preparation of 2,4,6-trimethylbenzonitrile oxide (128) was accomplished by treating 2,4,6-trimethylbenzald— oxime with N-bromosuccinimide and sodium methoxide. 107 108 Irradiation of a carbon tetrachloride solution of 128’pro- duced three photoproducts, the corresponding isocyanate 111 and nitrile 151’as well as 2,4-dimethylphthalimide 135. Compounds 131 and 135’were previously reported as photo- products of 128. 4. A 2-methyltetrahydrofuran solution of 128 was irradiated at -170° for short intervals. Immediately after each ir- radiation, the infrared spectrum of the cooled solution was examined in the 1600-2000 cm"1 region in an attempt to spectroscopically observe the oxazirine intermediate 132; No bands were observed which could be assigned to 132. 17)“??? A I'm..— ' LITERATURE CITED (6) (7) (8) (9) (10) LITERATURE CITED J. Griffiths and H. Hart, J. Amer. Chem. Soc. 90, 3297 (1968). ’”” . M. Collins and H. Hart, J. Chem. Soc., Sect C 895 1967). P ( H. Hart and R. M. Lange, J. Org. Chem. 11, 3776 (1966). D . H. R. Barton and G. Quinkert, J. Chem. Soc. 1 (1960). H. Hart, P. M. Collins, and A. J. Waring, J. Amer. Chem. Soc. 88/ 1005 (1966). H. Hart, J. A. Hartlage, R. W. Fish, and R. R. Rafos, J. Org, Chem. 31, 2244 (1966). D. Hart, in press. Organic Preparations and Procedures. A. C. Gripper Gray and H. Hart, J. Amer. Chem. Soc. 92, 2569 (1068). B. V. Gregorovitch, C. S. Liang, D. M. Auguston, and s. F. MacDonald, Can. J. Chem. gp, 3291 (1968). H. H. Jaffé and M. Orchin, Theory and Applications of Ultraviolet SpectroscoPY. New York: John Wiley and Sons, 1962. H. E. Zimmerman and D. I. Schuster, J. Amer. Chem. Soc. 84, 4527 (1962). H. E. Zimmerman and J. S. Swenton, J. Amer. Chem. Soc. 22: 906 (1967). R. J. Bastiani, D. J. Hart, and H. Hart, Tetrahedron Let., 4841 (1969). C. J. Collins, Chem. Rev., 69, 543 (1969) R. L. Baird and A. Aboderin, Tetrahedron Let., 235 (1963). 109 110 (16) R. L. Baird and A. Aboderin, J. Amer. Chem. Soc., 86, 252 (1964). ’”“ (17) R. L. Baird and A. Aboderin, J. Amer. Chem. Soc., 86, 2300 (1964). ’”” (18) N. C. Deno, D. LaVietes, J. Mockus, and P. C. Scholl, J. Amer. Chem. sog,, 22, 6457 (1968). (19) G. J. Karabatsos, C. E. Orzech, and S. Meyerson, g, Amer. Chem. Soc., 81, 4394 (1965). (20) C. C. Lee, J. E. Kruger, and E. W. C. Wong, J. Amer. Chem. Soc., 81, 3985 (1965). (21) C. C. Lee and J. E. Kruger, J. Amer. Chem. Soc., 81, 3986 (1965). (22) P. S. Skell and I. Starer, J. Amer. Chem. Soc., §fix 3962 (1962). (23) C. C. Lee and L. Gruber, J. Amer. Chem. Soc., 88/ 3775 (1968). (24) C. C. Lee, W. K. Chang, and K. Wan, J. Amer. Chem. Soc., 92, 3778 (1968). (25) C. C. Lee, L. Gruber, and K. Wan, Tetrahedron Let., 2587 (1968). (26) H. Hart and R. Schlosberg, J. Amer. Chem. Soc., 88/ 5030 (1966). (27) H. Hart and R. H. Schlosberg, J. Amer. Chem. Soc., 88/ 5189 (1968). (28) H. Hart and O. E. Curtis, Jr., J. Amer. Chem. Soc., 22, 931 (1957). (239) N. C. Deno and D. N. Lincoln, J. Amer. Chem. Soc., . §§/ 5357 (1966). (30) R. Hoffmann, J. Chem. Phys., 28, 2480 (1964). (31) T. Yoneyawa, K. Shimizu, and H. Kato, Bull. Chem. Soc., . Jap., 19, 1302 (1967). (32) H. Fischer, H. Kollmar, and H. O. Smith, Tetrahedron Let., 5821 (1968). (33) M. Saunders, E. L. Hagen, and J. Rosenfeld, J. Amer. Chem. Soc., 22; 6882 (1968). (49) (51) (52) 111 D. M. Brouwer, Rec. Trav. Chim., 82” 1435 (1968). K. Greenlee and C. Boord, J. Amer. Chem. Soc., 75, 948 (1953). ’V” K. Greenlee and C. Boord, J. Amer. Chem. Soc., 12. 3346 (1953). J. D. Roberts, W. Bennett, R. E. McMahon, and E. W. Holyrod, Jr., J. Amer. Chem. Soc., 12, 4283 (1952). G. J. Karabatsos, R. A. Mount, D. O. Rickter, and S. Meyerson, J. Amer. Chem. Soc., §§/ 5651 (1966). L. Friedman and A. T. Jurewicz, J. Amer. Chem. Soc., 81, 1803 (1969). L. Friedman, A. T. Jurewica, and J. H. Bayless, g:_ Amer. Chem. Soc., 81, 1795 (1969). N. C. Deno, W. E. and S. Schneider, Billups, D. LaVietes, P. C. Scholl, J. Amer. Chem. Soc., 88” 3700 (1970). H. C. Brown and C. Groot, J. Amer. Chem. Soc., 81, 2223 (1952). "Methoden der Organishe 1965. C. Grundmann in Houben Weyl, Chemi,“ George Thieme Verlag, Stuttgart, A Werner and H. Buss, Chem. Ber., 81, 2193 (1894). A. Quilico in "Chemistry of Heterocyclic Compounds", Vol. 17, A. Weissberger, Ed., New York, N.Y., 1962. C. Grundmann and J. M. Dean, J. Org. Chem., 88” 2809 (1965) G. Just and K. Dahl, Tetrahedron Let., 2441 (1966). S. Ranganathan, B. B. Singh, and C. S. Panda, Tetra— hedron Let., 1225 (1970). G. Just and W. Zehetner, Tetrahedron Let., 3389 (1967). D. M. Jerina, D. R. Boyd, and J. W. Daly, Tetrahedron Let., 457 (1970). T. Tsuchiya, H. Arai, and H. Igeta, Tetrahedron Let., 2213 (1970) and reference therein. J. Streith, B. Danner, and C. Sigwalt, Chem. Commun. 979 (1967). 112 R. Wiley and B. Wakefield, J. Org. Chem., 88, 546 (1960). T. M. Heilbron, ed. Dictionary of Organic Compounds, 3rd Edition, Oxford University Press, New York, 1953. M. Gomberg, L. H. Cone, Justus Liebigs Ann. Chem., 370, 182 (1909). C. Grundmann and R. Richter, J. Org. Chem., 88, 476 (1968). N. S. Dokunikhin, L. A. Gaeva, and I. D. Pleteva, Zhur. Obshchei Khim., 81, 174 (1954). C. Grundmann and H. D. Frommeld, J. Org. Chem., 2077 (1964). 9,9,, Jul- '1; 1 11.59!“ ”HEP. . APPENDIX SPECTRA 113 4:3 hulaljfliqm, u.1_L Many. A i ,. . 3 LL... _. mmv osoumusoaoum.a Imuuomo.m.«HoHomoflnawzumEmuumplm.v.v.m mo Mvaouv Eduuowmm HEZ .m musmam L. brh—LFD o ._-.»._..»»M»...—>..(._»».._...»O.N..>. 00 Whgsufla on o.— . r... . i. E 2‘ E N L *“ f 1’: 2'??— 114 . AWMV maou mlmcmt H Imuoono.N.vHoHowoflflahzumeuumulm.v.3.N mo AQHUUV Esuuommm H82 .m musmflm .Hr».)_..r..fi...._.r-.Hr».._-LHT-..L,L,.L(Lli.-.p_.-.-_1...r)..._ 0.. O 0 . 0.x OK 0.0 Thy Sam 06 0.? r 5% am oo— m¢.w.r .4 .- 115 .l U. . L. I. m .1.- _ . . .EE . LED”... 4.... SKIP .- .. of V; .mev aonmnsouo oaoumuosoua ImuooLo. N. vLoHomoaflamzumEMHuwul In v v N no LvHOUv Eduuummm HEZ L .3 .5... L L... L .1 1111.15. ......L L... 3.1 1 111.1141 1111.191: 11.31": ...LL.11.L.1111._L.L 1:21.111“ L .b ousmflm 11 111.1111:1-L.L LI 1.31.1.1"... 111-L. 1 :LL. . r . s u p > > . . . . . . 0.0 1.1.12.3 06. O v 1:;SLLL11sL:xLL 1.1L...L:.L.1a 1.1.. .1._..1.L..1;LL...1 1 116 .mev mumumomthmcmusnoaowouH IamcumEICI.HmcmmoumamsumEHUIN.HVIN Hmsuma mo Avaouv Bauuowmm HEZ .m musmflm . — — ID i P P P P P h P P k), b P P L > r b P _ n P b P h b! r 06 LB.S%. ad ad . .15.. L L L {1. L. filgélgig. 111 L... 117 .Ava mumumommcmusnoaohold.Halahnumald m Ihmcmvaaxmoumaxsqule N HVIN HmnquIAN NV NC A VHUUV Ednuommm HEZ m mu: Hm ? .: :. . _._... .. .. J. :: #_._.._.T:..}r._.= _. \.. ..._._ :i___...._: a? #1:: _........... _._ _ \ .L “-_ _“;‘ ' 118 .Jmm. mumumommcmusnoaumUId.HQIHmsquIC IJmcwwaammonmamnumEHUI N H.IN stumEIJm N. mo JvHUO. Esuuommm HEZ .oH musmflm ._r..... LTPF»L_...>JI»»»..IIIJILI..I...J+..»...p. . . Q. od 5.2.. cm ow aJ r . . ._ .. .,._ :: :. T::% $.33r3 .: J, ... J.J...... rfiflwfié; \_ .JJ.....JJ 14%.}: ..J...»....=..JJ.......J..J: J .JJH}..JIJ_.J..J.J....J’ Emu?:r...._._......J..‘.J..J-W..§..J..J_ $.2JK7JFJJJ 3.5.13.2“ .JJ.....:..JJ.___.. .. .. . J.. . . . .. \IIII\III‘..‘ - \ 119 .wav mumuwumwcmuSQOHOMUId.HQIahnumEId IJmcmvfiawmoumHmsumEHUIm.HJIN HmnumEIJm.mJ mo JvHUUJ Esuuommm HEZ .HH musmflm P r _ P I b .J.r»......fi.........J-I».-. III,4..II.|,II.HJ:IIII.-....J;H.»MII.III .,... r..r. b... O o . 0d OK 0.0 ...b. Sam 0.6. 0.? \. III . . JV}? . J Li J... 3 _._.. 1...... . . .J J: .5. . . j kip... J. :J. J .s JJ . gmfiJqug‘ .{NJLJJ ..J.... \IIrJJJJ. J..... .J J9}... V-fifisfg 5:)Jywr:.r .58"? ‘FJJJJJJJJJm “($33. ngsJJJg JJJ ’34.}! .J (Kai; .... {I _. £43-; 5.31.... L 120 .Ammv mumvwommcmuDQoHumunc.fidlazzumEId IchmwflaxmoumaxsumEHUIN.HyIN HhLumEIAN.mv mo Avaoov Esuuommm HEZ .NH wusmflm — r F b b > M p b g r P L > > H > > P b ¥ L P b H r L > H, u b r r M > b P b r 1* P . b b p > D t b M 0+— » 0.0 _ 0.x _ OK _ 0.0 C: ”2%. Own 0 Q L \1: -- - :\ .1), a... .2. >\.. .. V. r .8:¢.....,.:~ .rifudxni I. :3 7..) La. fir...“ ...s_l?¢i( 2 PE." __ (2,»? Efiygdgléi 3 _. 121 . «my mcoumncmflw -v&mvfi umomufio.v.vaoflowoflnamnumamuumuum.v.m~m mo Avaoov gunpowmm Hez .mH musmflm , —.pLL_ .».~r r_rf»»~r » _ r+fi* . _-+ _+LL_ »» h» J} 9. oo 3 o“ od :5: on o... 122 . Q3 mcoumucmflcég 8 H Imuwwmo.v.wHOHowoflnawzumEmHumulm.msv.m mo Avauuv Esuuommm HEZ .¢H musmflm -fr» H - ++ . -_ . FL - rrfl oo “.32.... o.m o... ON.waP no.m K. \‘._‘ T 123 . aw mao-mucm€u¢.§ H Imumwno.v.vHoHowuflnamzqumuuwulm~v.m~m mo Avaoov Eduuummm HEZ .mH musmflm #»+.,_....—.»L._»p»»~.p»»_.,»»~r+»»p_.»_H»»,.p.r>r_-r.._....h. o.— o.o . o.» ox 0.0 A b V 2.... on... o... ~y’. “o--- 124 .Awwv maoumumcmquvH Imomvfim.no.0.v.vHoHo>oHuuH>£umEmuumUIn.v.v.m mo AvHUUV Esnuommm HEZ .mH musmflm o. i..r :Qo ,- , ca ‘ .,i Lap 4 ; ;.od kflzfif ow, , . oé 125 .Amwv maonhlmcwlwlmowv uHo.H.m.vHoHumoHuuamgumamuumuuoH.oH.m.w mo Avaoov esHuuwmm Hez .hH musmflm 126 .Adwv waonvlmcmlmlmomw IHm.Ho.o.v.vgoHomoHHuamflumEmuumulm~m.N.N mo Avauov Esuuommm HEZ .wH madman Ab .. . .. . p....4LL.. . w. a... . o... as o6 «hit 3 ., oé 127 .«)mxmvlemlmammonmamsumelmlouoHLUIHIOEOHQIH mo 24ro 8350QO HEZ .mH musmfim _ ~ ~ _ f [It 5' Fulfil: Irlur {IL I. Iv. > V P LP + } L tr FL,» x! L p F > » — ._.»|F._.rfpp.»..b.»,, .- : , p, 0.0 . 0.x OK 0.0 .. F“. SEA 9% O V 3.0L. mv. _— .__——-—.— .ANMVaclalmcmmoumoHoxoamsumE mo Avaoov Eduuommm HEZ .om wusmflm 128 tor. .h or _ om of ”heft; ow _ erWL _ F & \Sr§xéJ;3%§}ii1¥53駥}1}}1§2{1xétzt W _ a: o fl n: om _ ‘ 2.3 a. . £5 2.: 'I—‘ fix _ w _ . V f _ é «Sumo 129 mm on S .Awhv lefilmcmmoumoHowoawzumE mo Eduuommm mmmz mv — m\E 0v mm . om . Hm 9:53 mm ___._ _ _ _ _ 0 :0H I O N O 0‘) rOmV 6 L0 qusuequl eAquIau wow .Ob .om OOH 130 \hk .Ammvawlmlmcmusnoaounlm mo AVHUUV Esnuommm HEZ _ _ _ o.o.¢ % ‘_ ‘ ma x. . y. it __w E r‘ _r K? , _ : .Nm mysmflm m a 1:3 1.2.{égs 1411' lllhpllulflvllu'illbi J L4 41‘ 14. ‘{ 4L4 4 ; NI OOfi a 131 .AHQHVlemlmcmusnoonLOIm wo Awaoov Eswuommm HEZ .mm mmsmflm ? L ¥ L— L h b V H D > .Lr b H L! k? LL V’\"l,ll. ‘ Irlrbx‘ls [Inbv VIII}. Irlr» F ‘7 F r r b \b r rrhv} Qo om OK ‘ . Qo Ak.2$. 06 ed 132 .vaw, lemlwcocmxmzlmlouoHLUIm Mo AJUUV Esuuommm HEZ .vm musmflm . . . . . . . r 7!. ,.-r1{tv , iri‘ - .LI. . (rtriivzwlfafPTrL IrLILI-.. rib . . FL , . » F . O— 0.0 0.x OK 0. .h SE; O.m 06 * ‘ N 25);}. .4” u» {xiii .1...“ LL. \§,..{}.}L$..3l!£.§£§%§§i§§§ai¢f 4... .33 Z... ,..T <4 :_ c. i m _ \ 133 .rww¢mcmusnomoH£UIN UmumHmusmU maamfluumm mo AVHUUV Esnuommm HEZ h I. pl. 0‘! . 4. 94:9? I.'| L")! In. . \%§;.§3}5i%<1 5. .mm musmflm [LIL FLL > FL b r b F ob :k,2m¢ ca . ow \l."l.l . 7..5§?f§é§..1§v§§>fi I... . .3, . .4 . c 1%.. w —&..J 4 . .5 3 fir v. .r i \ ; 4 J . .. ._H .n. (