I. ALKYLATION OF PHENOL WITH A HOMOALLYLIC CHLORIDE II. CONFORMATIONS OF SOME SEVEN-MEMBERED RINGS Thesis for Ike Degree of DH. D. MICHIGAN STATE UNIVERSITY James L. Corbin 1962. THESIS EAéI gamma, IanILihN ABSTRACT I.. ALKYLATION OF PHENOL WITH A HOMOALLYLIC CHLORIDE II. CONFORMATIONS OF SOME SEVEN-MEMBERED RINGS by James L. Corbin Wagner (1) found that Z-methyl-5-chloro-2-pentene reacted with phenol at 1500 to yield two isomeric products "A" and "B. " He showed that ”A" was 5, 5-dimethylhomochroman. The identification of. the phenolic product "B" was the first purpose of this thesis. The infra-red, NMR, ultraviolet, and mass spectra of "B" indicated that it was probablyr 1, l-dimethyl-S-tetralol. This was confirmed by synthesizing this tetralol as shown. OCH3 OH NI—I as 27: V During the investigation of "A" (1), and also from- the reaction of 2, 2-dimethyltetrahydrofuran with phenol (2), a second‘isomeric phenol "C" was isolated. Another purpose of this thesis was to identify "C. " It was shown to be an approximately equimolar mixture of "B" and 1, 1-dimethy1-7-tetralol, an authentic sample of the latter being synthe- sized as shown. The unusual orientation (the tertiary group was meta, CH3O o CH3O 0 HO 1. NaO-iPr 1. W.K. t ——-—’ ‘ 2. CH3I 2. HI James L. Corbin not para) of the phenolic products of the alkylation reaction, with the homoallylic chloride suggested that the reaction may proceed through one of the ions shown. The ion formed could attack phenol at the + + CH m CH I CH CH HTH (s 3 ,v' z 3 x”' z 3V>C = CH CH; - C1—-> >C75-3-‘CH l or >C73-7-‘C‘ H CH3 \ CH3 \CH2 CH3 “CH; CH; unsymmetrical symmetrical oxygen atom, or ortho, or para positions, and then cyclize to give only the three products actually isolated. To determine which ion was inn volved, 2-methyl-5-chloro-2-pentene-5-dz was used in the reaction. The labelled compound was synthesized by the route shown. This was allowed to react with phenol, and the products examined by NMR, which BTCHZ-‘CHZ-COZCzHS did-$9 BrCHz-sz‘CDzOH E9 BrCHz"CHz“CD3C: 3 K CH3 H M I CN )C(0H)CHz-CHz-CD,C1 3 CszozC-CHz-CHz-CDZCI CH3 C H OH - - .- Cl -H,o HC1 NC CH2 CH2 CD, CH3\ v /C = CH-CHz-CDZCI CH3 showed that the two methylene groups of the homoallylic chloride had become equivalent during the reaction. Thus the symmetrical ion was involved. James L. Corbin The ultraviolet spectrum of "A" suggested that there was a possible barrier to the seven-membered ring interconverting between the two chair forms (3). Examination of its NMR spectrum indicated that the ring was "flipping" at room temperature and also at -1000. An analogous hydrocarbon in which the barrier to interconversion of O \ \ -CH(CH,), . 1. NaNHz ; 1. W.K. \ 2. isoprene " 2. BF3 7 hydrobromide I the 7-membered ring was increased was synthesized as shown. Examination of its NMR spectrum showed the ring to be flipping at room temperature, but in the region of --30 to -600, this flipping appeared» to c eas e. REFERENCES l. C. R. Wagner, Ph. D. Thesis, Michigan State University, 1955. 2.. H. Hart, unpublished work. 3. H. Hart and C..R. Wagner, Proc. Chem. Soc., 284 (1958). I. ALKYLATION OF PHENOL WITH A HOMOALLYLIC CHLORIDE . II. CONFORMATIONS OF SOME SEVEN-MEMBERED RINGS BY James L. Corbin A THESIS f Submitted to AMiChigan State University in partial fulfillment of the requirements for the degree of DOCT OR OF PHILOSOPHY Department of Chemistry 1962 ACKNOWLEDGMENT _ The author wishes to express his sincere appreciation to Professor Harold Hart for his encouragement and guidance throughout the course of this investigation. Appreciation is also extended to the Petroleum Research Fund of the American Chemical Society whose fellowship provided personal financial assistance from September 1960 through December, 1961. ii TABLE OF CONTENTS Page INTRODUCTION...... ........ 1 RESULTS AND DISCUSSION ............... . . . . 5 Part I A. Structure and synthesis of "B" . . . . ....... . 6 B. Structure and synthesis of "C" .......... . 13 C- Mechanism of the reaction of phenol with 2- -methyl- 5- Chloro- Z— -pentene ................. . . 30 Part II A.. Stereochemistry of 5, 5-dimethy1homochroman. . . . 58 B. Synthesis and stereochemistry of l, 1,4,4-tetra- methylbenzocycloheptene ........ . . . . . . . 67 EXPERIMENTAL......... ....... 93 Gas chromatography, Spectra, Microanalyses, Melting points.... ............ 94 Part I A. Synthesis of 1,1-dimethy1-5-tetralol . ........ 95 Preparation of 1, l-dimethyl-S-methoxy-Z- tetralone ................. . . . . 95 Preparation of 1, l-dimethyl-S-tetralol ...... 96 B. Synthesis of 1,1-dimethy1-6-tetralol . . . . ..... 98 Preparation of isopropylidenesuccinic acid . . . . 98 Preparation of terebic acid . . . . . . ....... 99 Preparation of 4-methy1-3-pentenoic acid ..... 99 Preparation of 4- methyl-4—(p-methoxyphenyl) pentanoic acid. . . . . . . .......... 99 Preparation of4, 4- -dimethyl- --7 methoxy-l- tetralone .................... 100 Preparation of 1,1-dimethy1-6-tetralol. . . .' . .. 101 C. Synthesis of 1, 1- -dimethy1- -7- tetralol ......... 102 Preparation of 2, 7—dimethoxynaphtha1ene ..... 102 iii TABLE OF CONTENTS - Continued Page Preparation of 7-methoxy-2-tetralone ....... 103 Preparation of 1,1-dimethyl-7-methoxy-2- tetralone .................... 103 Preparation of1,1-dimethy1-7-tetralol ...... 104 D..Mixture of 1,1-dimethy1- 5- and 7- tetralols ...... 106 Separation of "C" .................. 106 - Preparation of synthetic "C" ......... . . . 106 Reaction of 2-methy1-5-chloro-2-pentene with phenol. . . . . ................. 106 E. Attempted synthesis of 2- --methy1 5- chloro- 2- -pentene- 5-.dz...... .. . ..... . ..... 108 Preparation of 2, 2- -dimethyltetrahydrofuran- --5 dz . 108 Reaction of 2, 2- dimethyltetrahydrofuran-5- d; with Lucas reagent ............ . . . 110 F. Synthesis of 2- -methyl- 5- chloro- 2- --pentene 5- d2 . . . 111 Preparation of ethyl 3- -bromopropionate ...... 111 Preparation of 3-bromopropanol-l-dz ....... 111 Preparation of 1-bromo-3-chloropropane-3-dz . . 112 Preparation of 4- -chlorobutyronitrile- --4 dz. . . . . 113 ' Preparation of ethyl 4- Hchlorobutyrate -4- dz . . . 114 Preparation of 2--methy1 5- Chloro- 2- pentanol- 5- -dz 114 Preparation of 2- m-e-thyl 5- chloro- 2- --pentene --5 dz. 115 Reaction of 2-methy1-5-chloro-2-pentene-5-dz with phenol .................... 1 15 Part II A.. Synthesis and attempted resolution of 5, 5-dimethy1- homochroman-7-carboxylic acid. ..... . . .. 118 Preparation of 5, 5-dimethy1homochroman . .. . . . 118 -Preparation of 5, 5-dimethy1-7-acetylhomochro- man ...... . . ...... 118 Preparation of 5, 5- -dimethy1homochroman carboxylic acid ............. . . . 119 Preparation of 5, 5- -dimethy1homochroman- 7- carboxamide ........ .. . . 120 Preparation of methyl 5, 5-dimethy1homochroman- 7- -carboxylate ....... . .......... 121 Conversion of 5, 5- dimethylhomochroman-7- , carboxamide to 5,5- -dimethylhomochroman. . . 121 Attempted resolution of 5, 5-dimethy1homochroman- 7-carboxylic acid ...... . . . . . . . . . . 122 iv TABLE OF CONTENTS - Continued - Page B. Attempted synthesis of 1, 1, 4, 4- -tetramethy1benzo- cycloheptene ............... . . 122 Preparation of 3- -methyl- 3- -phenylbutyric acid. . 122 Preparation of 3- --methyl 3- --phenyl 1- butanol. . . 123 ~Preparation of the tosylate of 3-methy1- 3-pheny1- 1- butanol ............... . . . 123 Preparation of 5- -methyl- -5- -pheny1hexanoic acid. 124 Preparation of 5-methy1-5-pheny1hexanoy1 chloride ............. . . . . . 125 Preparation of 5, 5- dimethylbenzocycloheptene- 1- -one ............. . ....... . 125 Preparation of 2, 2, 5, 5-tetramethy1benzocyclo- heptene- 1-one .................. 127 Attempted catalytic reduction of 2, 2, 5, 5-tetra1-. methylbenzocycloheptene- l-one ........ 128 Preparation of 2,2, 5, 5-tetramethy1benzocyclo- heptene- 1- ol ................. 129 Attempted reduction of 2, 2, 5, 5- -tetramethy1benzo- cycloheptene- 1- ol ................ 130 Preparation of 1-chloro-2, 2, 5, 5-tetramethyl- ‘benzocycloheptene ......... . . . . ...._.... 130 Reduction of 1- chloro-Z, 2, 5, 5-tetramethylbenzo- cycloheptene . . ....... . . . 131 C. Synthesis of 1, 1, 4, 4- -tetramethy1benzocycloheptene. 131 Preparation of 1- bromo- 3- -me-thyl 2- butene . . . 131 -Preparation of l- --pheny1 2, 2, 5- trimethy1-4- hexene-1-—one ............. .. 132 Preparation of 2, 4, 4- -trimethy1- --6 --phenyl 2- hexene ............. . . ..... 133 Ozonolysis of 2, 4, 4- -trimethy1- --6 -pheny1- 2- hexene ........ . . . . 134 Reaction of 2, 4, 4- -trimethy1- 6- pheny1—2- hexene with aluminum chloride . . . . . . . . . . . 134 Reaction of 2, 4, 4- -trimethy1- -6- phenyl- 2- hexene with hydrogen chloride and aluminum chloride 135 . Preparation of 1, 1, 4, 4-tetramethylbenzocyclo- heptene . . . ............. 136 Reaction of 2, 4, 4- -trimethy1- 6- -pheny1- -2- hexene with ferric chloride ......... . . . . . 137 TABLE OF CONTENTS - Continued Page Reaction of 2, 4, 4-trimethy1-6-pheny1-2-hexene with stannic chloride ..... . ......... 137 SUMMARY ..... . ...................... 138 LITERATURE CITED ........... . ....... 141 vi TABLE 1. LIST OF TABLES Ultraviolet Spectra of Several Phenols ....... . . Ultraviolet Spectra of Several Phenols ......... Calculated and Experimental NMR Peak Areas for Deuterated 5, 5-Dimethy1homochroman ...... . . . Calculated and Experimental NMR Peak Areas for the Deuterated Phenolic Products . . . . . . ....... Properties of 5-, 6- and 7-Membered Cyclic Aromatic Ethers . . . . . ........ . ............ Ultraviolet Spectra of Some Aromatic Acids . ..... Ultraviolet Absorbtion of Some Aromatic Hydrocarbons vii Page 11 18 54 55 58 87 LIST OF FIGURES FIGURE 1. Ultraviolet Spectrum of "B" in Cyclohexane . . . . . 2. Infra-red Spectrum of "B" in Carbon Disulfide 3. . NMR Spectrum of "B" in Carbon Tetrachloride . . . . 4. Infra-red Spectrum of "C" in Carbon Disulfide 5. Ultraviolet Spectrum of "C" in Cyclohexane ...... 6.’ NMR Spectrum of "C" in Carbon Tetrachloride . . . . 7. Infra-red Spectrum of 4, 4-Dimethy1-7-inethoxy- 1- tetralone(Neat) ............. . . . . ...... 8. . Infra-red Spectrum of 1, 1-Dimethy1-6-tetralol in Carbon Disulfide .......... . . . . . . . . . 9. Ultraviolet Spectrum of 1, 1-Dimethyl-6-tetralol in Cyclohexane ............. . . . . . . . . . . 10. Infra-red Spectrum of 1, l-Dimethy1-7-methoxy-2- tetralone in Carbon Disulfide ....... . . ..... ll. . Infra-red Spectrum of l, l-Dimethy1-7-tetralol in Carbon Disulfide . ........ . .. . . . . . . . 12. Ultraviolet Spectrum of 1, 1-Dimethyl-7-tetralol in Cyclohexane ....... . . . . . . .......... l3. - Infra-redSpectrum of 2, 2-Dimethyltetrahydrofuran-5- dz in Carbon Tetrachloride ........ . . . . . . . 14. NMR Spectrum of (a) 2, 2-Dimethyltetrahydrofuran-S- dz, (b) Unlabelled Material, in Carbon Tetrachloride . viii Page 14 15 17 20 21 22 24 26 27 33 34 LIST OF FIGURES - Continued FIGURE 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. Infra-red Spectrum of 2-Methy1-5-chloro-2-pentene-4- and 5-dz in Carbon Tetrachloride ....... . . . . . NMR Spectrum of (a) 2—Methy1-5-chloro-2-pentene-4- and 5-dz, (b) Unlabelled Material, in Carbon Tetra- chloride........., ............ Infra-red Spectrum of 3-Bromo-1-propanol-1-dz in Carbon Tetrachloride ...... . . . ......... NMR Spectrum of 3-Bromo-1—propanol-1-dz in Carbon TetraChloride O O O O O O O O O O O O O O O O O ..... Infra-red Spectrum of 1-Bromo-3-Chloropropane-3-dz inCarbonTetrachloride . . . . . . . . . . . . . . . . NMR Spectrum of (a) 1-Bromo-3-chloropropane-3-dz, (b) Unlabelled Material, in Carbon Tetrachloride . . . Infra-red Spectrum of 4-Chlorobutyronitrile-4-dz in Carbon Tetrachloride . . . . . . . ........... NMR Spectrum of 4-Chlorobutyronitrile-4-dz in Carbon Tetrachloride ........ . . . . . . . . . . Infra-red Spectrum of Ethyl 4-Chlorobutyrate-4-dz in Carbon Tetrachloride ....... . . . . . . . . . NMR Spectrum of Ethyl 4-Chlorobutyrate-4-dz in Carbon Tetrachloride ....... . . ........ Infra-red Spectrum of 2-Methy1-5-ch10’rp-2-pentanol- 5-dz in Carbon Tetrachloride . . . . . . . . . . . . . . . NMR Spectrum of 2-Methyl-5-chloro-2-pentene-5-dz .‘in Carbon Tetrachloride ...... . . . . ..... Infra-red Spectrum of Z-Methyl-5-chloro-2-pentene-5- d; in Carbon Tetrachloride . . . . ........... ix Page 36 37 39 40 41 42 43 44 46 47 48 49 51 LIST OF FIGURES - Continued FIGURE . Page 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. Infra-red Spectrum of Deuterated 5, 5-Dimethy1homo-~ chroman in Carbon Tetrachloride ........ . . . . 52 NMR Spectrum of Deuterated 5, S-Dimethylhomochroman in Carbon Tetrachloride ............ . . . . . 53 NMR Spectrum of Deuterated 1, 1-Dimethy1-5- and 7- tetralols in Carbon Tetrachloride ......... . . . 56 Ultraviolet Spectrum of 5, 5-Dimethy1homochroman in Cyclohexane . . . . ..................... 59 Infra-red Spectrum of 5, 5-Dimethy1-7-ac etyl-homo- chroman in Carbon Disulfide .......... . . . . . 62 Infra-red Spectrum of 5, 5-Dimethy1homochroman-7- _ carboxylic Acid in Carbon Disulfide ........ . . . . 64 Ultraviolet Spectrum of 5, 5-Dimethylhomochroman-7- carboxylic Acid in Cyclohexane ............ . 65 . NMR Spectrum of 5, 5- -Dimethy1homochroman in Carbon 'Tetrachloride ............. . . . . . .. . . . . 68 Infra-red Spectrum of 3- --Methyl 3- -pheny1- l- butanol (Neat) .......... .................71 . Infra-red Spectrum of 5-Methy1-5-pheny1hexanoic Acid in Carbon Tetrachloride . . ..... . . . ....... 72 . Infra-red Spectrum of 5, 5-Dimethylbenzocycloheptene- l-one in Carbon Tetrachloride. . . . . . . . . . . . . . 7B Infra-red Spectrum of 2, 2, 5, 5-Tetrarnethylbenzocyclo- heptene-l-one in Carbon Disulfide . . . ., ..... . . . 74 . Ultraviolet Spectrum of 2, 2, 5, 5-Tetramethy1benzocyclo- heptene-il-o‘ne in Cyclohexane. ....... . . ..... 75 . Infra-red Spectrum of 2, 2, 5, S-Tetramethylbenzocyclo- heptene- 1-01 in Carbon Disulfide ............. 77 LIST OF FIGURES - Continued FIGURE 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. . Infra-red Spectrum of l-Chloro-Z, 2, 5, 5-tetra1nethyl- benzocycloheptene in Carbon Disulfide ...... . . . Infra-red Spectrum of the ReductionProduct in Carbon Disulfide ......................... Infra-red Spectrum of 1-Pheny1-2, 2, 5-trimethy1-4- hexene- 1-one in Carbon Disulfide ............ , Infra-red Spectrum of 2, 4, 4-Trimethy1-6-pheny1-2- hexene in Carbon Tetrachloride ..... . ....... . Infra-red Spectrum of 1-Isopropy1-3, 3-dimethy1tetra- lin(?) in Carbon Disulfide . . . . . ........... Ultraviolet Spectrum of l-Isopropy1-3, 3-dimethy1- tetralin(?) in Cyclohexane. . . . ..... . ...... . NMR Spectrum of 1-Isopropy1-3, 3-dimethy1tetralin(?) in CarbonTetrachloride. . . . . . . . . . . . . . . . . Infra-red Spectrum of 1, 1, 4, 4-Tetramethy1benzocyclo- heptene in Carbon Disulfide . . . . . . . . ....... Page 78 80 81 82 84 85 86 88 Ultraviolet Spectrum of 1, 1, 4, 4-Tetramethy1benzocyclo- heptene in Cyclohexane . . ........ . ...... NMR Spectrum of 1, l, 4, 4-Tetramethy1benzocyclo- heptene in Carbon Tetrachloride . . . . . . . . . . . . Comparison of the Infra-red Spectra of "B" and l, l-Dimethyl-S-tetralol (in Carbon Disulfide). . . . Comparison of the Infra-red Spectra of "C" and a Mixture of 1, l-Dimethyl-S- and 7-tetralols (in Carbon Disulfide)... ....... ........ ...... Comparison of the Infra-red Spectra of 1, 1-Dimethy1- 7-tetralol and Those of the Component of "C" and of the Alkylation Reaction Product (in Carbon Disulfide). . . xi 89 90 97 107 109 LIST OF FIGURES — Continued FIGURE Page 55. Infra-red Spectrum of the Deuterated Phenolic Products in. Carbon Tetrachloride . . ,, . . . . . . . . 117 56. . Infra-red Spectrum of 5-Methy1-5-phenylhexanoy1 Chloride (Neat) .............. . . . . . . . 126 xii ‘ INTRODUCTION INTRODUCTION Reactive alkyl halides such as those with-a tertiary, allylic, or benzylic alkyl group, can alkylate an activated aromatic nucleus (1. e. . phenol) without the use of a' Friedel-Crafts catalyst (144). A Thus t-Qbutyl chloride (2), a-phenethyl chloride (3), or triphenylmethyl chloride (4) react with phenol to evolve hydrogen chloride and yield nuclearly alkylated products. The possibility that the homoallylic chloride 2-methy1-5-chloro-2- pentene (1), although a primary halide, might be sufficiently active to react similarly was suggested by the work of Favorskaya and Fridman (5). They found that this halide gave dimethylcyclopropylcarbinol (11) upon, treatment withaqueous potassium carbonate. The reverse reaction was ac complished with aqueous CH3 CH2 CH3\ KCO /C=CH-CHz-CHz-C1 -£—?—-> HO—(IS—CH CH, ‘ 439-1— ‘CH, CH; 1 II hydrochloric acid. This suggested that the homoallylic chloride'I was a possible source of the dimethylcycloprepyl cation ‘which might alkylate phenol. - On this premise Wagner' (6) attempted the reaction and demonstrated that I did indeed alkylate phenol without the aid of, a catalyst at 150°. . He isolated two isomeric, colorless, crystalline products, one neutral and one phenolic. . The neutral material was identified as 5, 5-dimethy1homochroman' (111).i 'The structure of ‘ III the phenolic material was not elucidated, but it was definitely not the expected para-(dimethylcyclopropylcarbinyl)-phenol (IV). Following the designation used by Wagner (6), the phenolic material will be CH, CH, \ CH - C OH / CH, H, IV referred to as "B" in this thesis. Prolonged treatment of the homo- chroman III with acid yielded a second isomeric phenol which also was not identified. ‘This'will be referred to as "C. " Both III and "C" were also produced by the reaction of 2, 2-dimethyltetrahydrofuran with phenol in the presence of hydrogen Chloride (7). The purpose of this thesis, in part, was to determine the structures of phenols "B" and "C. " The mechanism of the reaction between phenol and 2-methy1-5-chloro-2-pentene was also to be investigated. Examination of molecular models of 5, 5-dimethy1homochroman showed that the most likely conformations for the molecule were two interconvertible enantiomorphic chair forms (8). If the barrier to inter- conversion of Va to .Vb were sufficient Va , Vb then the homochroman would be resolvable. The resolution of the Cyclo- heptadiene derivative VI, which is a related system, has been accomplished (9). Since the homochroman III was a product of the reaction with which VI VI this thesis deals, its stereochemistry was also investigated, as was that of a related hydrocarbon system. RESULTS AND DISCUSSION RESULTS AND DISCUSSION Part I A. Structure and synthesis of "B. " The phenol referred to as "B, " melting point 113-1140, was obtained by Wagner (6) in 29% yield from the uncatalyzed reaction of 2-methyl-5-chloro-2-pentene with a five-fold excess of phenol at 1500 for 12 hours. "B" had the molecular formula CleléO and formed typical phenol derivatives. . Its ultraviolet absorption spectrum (Figure l) with bands at 278. 5 mp. ( e: 1340) and 271. 5 m11( e: 1290), was typical of an ortho or meta but not a para substituted phenol, since it lacked the characteristic absorption of the latter in the 287-290 mp. region (10, 11). The infra-red spectrum (Figure 2) indicated the presence of a hydroxyl group (2.80 11),. a gem-dimethyl group (7. 22 and 7. 30 11), and 1, 2, 3-trisubstitution (12.8 p). The pattern in the 5-6 11 region was also consistent with 1, 2, 3-trisubstitution (12). The NMR * spectrum of "E" (Figure 3) supported trisubstitution by showing three aromatic protons (ca. 3.4 7" ). The protonat 5. 017' was assigned to the hydroxyl group, the two at 7. 43 7' to benzylic protons, and the four at 8. 33 7" to methylene protons. The single peak at 8. 80 7" corresponding to six protons was consistent with a gem- dimethyl group in which the central carbon was quaternary. The number of protons assigned to each peak was obtained from the relative peak areas (13a). :1: nuclear magnet1c resonance 1.73 x 10“ moles/liter I J 1 I 250 260 270 280 290 Wavelength (mu) Figure 1. Ultraviolet Spectrum of "B" in Cyclohexane. .1 Absorbance MA .eefisflm e330 5 :m: we Eseeoeem 8733 A3 newcofiocwd? 2 o e _ e .N 0.33me « a _ 1K4. 3.21 3.57 Figure 3. Amplitude inc reas ed 3X 1 \ ~N ’ UK. . 1 1 II .10 (7 7.43 (57 8.33 NMR Spectrum of "B" in Carbon Tetrachloride. Epiq—- — —' "'""'" —‘—“"""" -""‘L"" 10 In. mass spectra of alkylated aromatic compounds a major course of fragmentation is the loss of an alkyl group on a benzylic carbon atom (14), as shown in V11. Compound "B" showed a major fragmentation route to be 176 —-> 161 + 15 whichindicated- that the two methyl groups were on a benzylic carbon atom. l C5H6 - C-:I-CH3 I V11 There were several structures which fit the requirements of "B" in having a gem-dimethyl group and 1, 2, 3-trisubstitution. These were the 1, 1-dimethy1-5-and 8-tetralels (VIII, IX) and 1, 1, 3-trimethy1-4-and 7-indanols (X, XI). X and XI were the leastelikely since they contained three'methyl groups OH OH i I I OH i OH i VIII ‘ IX x XI and would not fit the NMR spectrum. Comparison of the ultraviolet spectrum of "B" with those of 5-tetralol (15a) and 4-indanol (15b) showed that "B" closely resembled 5-tetralol in bothband positionand'molar extinction coefficient (Table 1). The NMR spectrum of "B" also fitted the tetralol structures VIII and IX in having two benzylic methylene protons (7.43 7" ) and four other methylene protons (8. 33 7' ). Intetralinitself . the benzylic proton peak is at 7. 307' (16a) and the B-methylene proton peakxis at 8.22 7'- (16b). 11 Table l. - Ultraviolet Spectra of Several Phenols xmax. (mu) 6 Reference "B" 278, 272.5 1340, 1290 Figurel 5-tetralol 279, 272 1350, 1300 15a -4-indano1 276, 272 800, 630 15b The difference between structures VIII and IX was a matter of steric consideration. The hydroxyl in IX being adjacent to the methyl groups, would be sterically hindered compared to the hydroxyl in VIII. Coggeshall (17) has shown that hindered or partially hindered phenols require 0. 2 - 0. 5 N ethanolic sodium hydroxide to give a complete bathochromic shift of the maximum in the ultraviolet. Unhindered phenols give this shift in only 0. 1 N base. . "B" showed the complete shift in only 0. 1 N base (6) and therefore was unhindered. Thus the best structure for "B" was VIII. Since this was not a known compound, it was synthe- sized by the route shown in the scheme. CH,o O l. NELC zHiOH ‘ 1.! NaOi—Pr‘ 2.1-T+ , ’ 2. CH,I ’ OCH3 oCH3 O .7 .CH3 ’ W.K. 4.__...__.. OH oCH3 VIII 12 Both 5-methoxy-2-tetralone and 1, 1-dimethyl-5-m.ethoxy-2-tetra- lone had been previously described (18)., (and. the literature procedure was followed. 1, 6-Dimethoxynaphthalene was reduced with sodium metal in refluxing ethanol to give a 43% yield of the methoxytetralone. This was dimethylated in one step using two moles of sodium isopropoxide and excessmethyl iodide in isopropyl alcohol to give the dimethylmethoxy- tetralone in. 75% yield. This underwent a Wolff-Kishner reduction to give a 77% yield of 1, 1-dimethy1-S-methoxytetralin, which was not purified but was cleaved directly to 1, l-dimethyl-S-tetralol(VIII) using hydriodic acid. The melting point of this phenol was 112. 5 - 113. 5° and showed no depression whenmixed with. "B. " The infra-red spectra of the two were identical. - It ‘was concluded that compound "B" was 1, 1-dimethyl-5- tetralol. The formation of a bicyclic system from the alkylation. of phenol with. 5-chloro-2-methyl-2-pentene was not altogether unexpected. Bruson. and Kroeger (19) had tobtaineda Cyclialkylated products from the Friedel-Crafts catalyzed alkylation of phenol with 2, 5-dichloro-2, 5- dimethylhexane. Although their evidence CI-\I, CH3 H _ C1— . \ / cle, AlCl, ‘ Hz C1—C5 W37 CH3 _ or - BE: C6H50H + / CH, was not rigorous, the structures assigned to the products were quite reasonable. Our alkylating agent differed only in thatone reactive center was tertiary and. one was primary, whereas theirs had two tertiary 13 centers. The orientationin. "B, " with the tertiary carbon meta» and the primary carbon ortho to the hydroxyl group, is remarkable. In Bruson and Kroeger's products the orientation was always meta and para. A later section on the mechanism of the reaction attempts to explain this . B. Structure and synthesis of "C. " The isomer referred to as "C, " melting point 83. 3 - 83. 90, was isolated by Wagner (6) in 30% yield from the prolonged treatinent of 5, 5-dimethy1homochroman with hydrobromic acid in glacial acetic acid. It was also formed in 15% yield in the hydrogen chloride catalyzed alkylation of phenol by 2, 2-dimethyltetrahydrofuran .(7). "C" was phenolic although it was difficultly soluble in aqueous base. It could be extracted into Claisen's solution and formed typical phenol derivatives. The phenoxyacetic acid melted at 179- 179. 5 o(softened at 170°) and the methyl ether (6) was a liquid which. showed only? one peak on vapor phase chromatography (10% silicone column). Theinfra-red spectrumeof "C" (Figure 4) was very similar to that of "B, " the main differences being in. intensities. The gem-dimethyl group was present (7. 23 and 7. 30 p.) and several new bands were present in the 11. 3 - 12. 6 11 region. . Examination of the 5-6 p. region showed very broad bands around 5. 3, 5.5,. 5.8, and; 6.0 11.. They were not as intense or sharp as those in "B" which was known to be 1, 2, 3-trisubsti- tuted. The presence of the bands in the 11. 3-12. 6 11 region could mean 1, 2, 4-substitution. ‘ The ultraviolet‘spectrum of "C" (Figure 5) had bands at 287 my. (6 = 1033), 279 mp ( e = 1714),, and 273 mp. ( e = 1336). The band at 287 mp. suggested para substitution (10, 11). . The spectrum showed the characteristic shift to higher wavelengths: in only 0. l N ethanolic sodium hydroxide which indicated that "C" was not a hindered phenol (17), despite its being difficultly soluble in aqueous base. 14 MA .2 62:85 nephew 5 so: no 5:38am 8.795: A3 fiwcgocfim? o N. _o m .w. oudwfih a _ . 15 1.19 x 10'4 moles/liter . . broad max1mum In 0.1 or 0. 5 N ethanolic NaOI—I " 0. 2 Iii”, ‘\\ a" \\ II \‘ 'l 1‘ / ~, I, ‘\ .\ .—4 0. ___l__ I I r I 260 270 280 290 300 Wavelength (mu) Figure 5. Ultraviolet Spectrum of "C" in Cyclohexane. Abs orbanc e 16 The NMR spectrum of "C" (Figure 6) was much like that of "B" suggesting a similar skeleton. . The tentative assignments were aromatic, hydroxyl, benzylic methylene, methylene, and methyl protons as in "B. " The significant difference with» respect to the spectrum of "B" was the doublet instead of a. single line for the six methyl protons. This could result «from several factors. Either the two methyl groups were on a quaternary carbon atom and were not magnetically equivalent, or they were present as an. isgpropyl group and» a-spin. interaction arose (13b). This question. could be resolved by running the NMR at two different frequencies. Alternatively the question. could be answered by amass spectrum. If the two methyl groups were on a benzylic carbon atom a major course of fragmentation would be 176 --> 161 + 15, but if an. iso- propyl group were there, it would be 176 -—-) 133 + 43' (14).. The mass ‘ spectrum of "C" was very similar to that of "B" andshowed 176 "3“) 161 but no. 176 —-> 133 fragmentation, indicating loss of methyl rather than isopropyl. The similarity of the infra-red, NMR,, and mass spectra of ”B" and "C" would suggest similar structures, possibly positional isomers. . Since the infra-red spectrum of "C" could support 1, 2, 4-trisubstitution, and the ultraviolet region indicated para substitution, a 6-tetralol skeleton-(X11) would be plausible. HO XII Table 2 compares the ultraviolet spectrum of "C" with those of 6-tetralol and several other phenols with para substituents. » The striking feature was the very low extinction coefficient of the high wavelength band- in "C" compared to that in 6-tetralol itself. In general, para substitutiongives rise to approximately equal extinctions for the two high wavelength bands 17 (6) Amplitude increas ed 3X / ‘ (4) (1) (3) 1 1 i (Z) V \, I__ l I I I I J;L > 3.22 3.55 ”5.391I 7.39 If 8.42 8.798.88 7" Figure 6. NMR Spectrum of "C" in Carbon Tetrachloride. 18 Table 2. . Ultraviolet Spectraof Several Phenols A Compound I'Lmax. (mu) 7 5 Reference "C" " "287, 279, 272 1033, 1714‘, 1336 Figure 5 6-tetralol 288, 282, 279 2600, 2550, 2550 (15c) p-cresol 284, 277, 274 1950, 2150, 1850 1(15d) 3,4-xy1enol 286, 279, 271 2340, 2450, 1400 (15c) 5-indanol 290, 283, 275 2700,. 3000, 2350 (15f) as shown. However in "C" the second band was almost twice as intense as the first. On the supposition that "C" was a '6-tetra1ol (with perhaps some unusual structural feature to account for the discrepancy in the ultraviolet extinction coefficient) the two isomers XIII and XIV were possible, in analogy to "B. " Neither 1, l-dimethy1-6-tetralol (XIII) nor 1, 1-dimethy1-7-tetralol (XIV) had been reported in the literature, so both had to be synthesized. The first one synthesized was XIII since it had the tertiary group in the para position. This seemed most likely since "C" was formed in an alkylation reaction. H HO XIII XIV The route used in the synthesis of 1, 1-dimethyl-6-tetralol is shown in the scheme. .19 (CH3)2C = CH'CHz‘COZH Céiilocst > p‘CH3O"C6E'C(CH3)Z"CH2‘CH2’“C02H ' 3 1. PCI, 2. SnCl‘ HO CH ’ 0. CH3 HI W.K. . XIII The method of Linstead (20) was used to prepare 4-methy1-3-pentenoic acid from isopropylidenesuccinic acid.. This was used to alkylate anisole in the presence of aluminum chloride to give 4-methy1-4-(p-methoxyphenyl) -pentanoic acid in 72% yield. The neutralization equivalent was correct and the melting point agreed with the literature value (21). The acid was converted into the acid chloride with phosphorus pentachloride and then acyclized with stannic chloride to give an. 83% yield of liquid 4, 4-‘dimethyl-7- methoxy-l-tetralone. This ketone was not a known compound. An oxime (m.p. 97-980) was prepared and gave a correct carbon, hydrogen, nitrogen analysis. The infra-red Spectrum.(Figure 7) of the ketone sup- ported the assumed structure. The ketone was reduced by the Wolff- Kishner method to 1, 1-dimethy1-6-methoxytetralin in 80% yield. This liquid methyl ether was not purified but was cleaved to 1, 1-dimethy1-6- tetralol.(XIII) with :hydi-iodic acid in glacial acetic acid (65% yield). 1t crystallized as colorless needles melting at 93-940 and gave the correct elemental analysis. Its infra-red spectrum.(Figure 8) was consistent with the structure given. A phenylurethane derivative (m.p. 136-137°) was prepared,; and this also analyzed correctly. The ultraviolet spectrum (Figure 9) of XIII had bands at 286 mp. (6 = 2140), 279. 5 mp ( 6: 2010), and 278 mp. (6 = 2020). Again the two high wavelength bands were of about equal intensity, unlike "C. " "C" was definitely not 1, l-dimethy1-6-tetralol. 20 .Aumozv 983.33.; u>xo£uogunntm£uoaw01¢ .w mo 6550on pouumufifi .w oufimfim 1: “image’s? 2 3 o N. m m fl _ — A I—r _J 21 .opfifismwfl conflumU CM Hofimuuouuotaafiofiwflta J Ho 5550on pouumuwcH .w oudmfim HA A3 Auwcoaoxfim? o Ni _ o m m A _ E 4 22 2. 94 x 10" moles/liter 1 I I 1 260 270 280 i 290. 300 250 Wavelength (mp) Figure 9. Ultraviolet Spectrum of l, l-Dimethy1-6-tetralol in Cyclohexane. Absorbance 23 The other choice for "C" was 1, 1-dimethyl-7-tetralol (XIV), and this was synthesized, as shown in the scheme. 2, 7-Dimethoxynaphthalene O 1. Na, CszOli W 1. NaOi-PrA 2.1477 2.CH,I ’ 0. ° ”.K. XIV was prepared from the corresponding dial with methyl sulfate in 97.. 5% yield. The diether was reduced with sddium metal in refluxing ethanol to give 7-methoxy-2-tetralone.- An oxime of this ketone was prepared and its melting point agreed with the value given by Zaugg (22). The ketone itself melted at 30. 5-31. 70. Zaugg gave the melting point as room temperature. . The ketone was dialkylated in one step using .methyl iodide and sodium isopropoxide to give 1, 1-dimethy1-7-methoxy-2- tetralone in 76% yield. This Compound was not reported in the literature. It melted at 53.5-54. 30 and gave a correct carbon, hydrogen, and methoxyl analysis. . The methoxyl analysis was necessary to exclude the 'monomethyl compound. . The infra-red spectrum(Figure 10) was con- sistent with the structure assigned. An oxime was prepared (m.p. 148- 1490) and analyzed correctly. - Wolff-Kishner reduction of the ketone . gave a 73% yield of crude 1, 1-dimethyl-7-methoxytetralin. . This was not purified, but was cleaved directly to 1, l-dimethy1-7-tetra1ol (XIV) in .opafismflfl Conflfimo a: ocofimH—ounm1>xo£uoauwufi>£aoaafltH .H mo 85.3.0on Contoumam .oH ondmfm A3 Aumnoaourm? 2 u. S a a: m m _ 24 a 11 _ _ 25 81% yield with thdr'iodiic acid inglacial acetic acid. . This'tetralol melted at 105-31060 and gave a. correct elemental analysis. The infra- red spectrum (Figure 11) was inagreement withithe proposed structure. The ultraviolet spectrum (Figure 12) had bands at 287. 5 mp.( e = 2140), 279. 5 mp. ( 6: 2270), and 272 mp. (6 = 1590). Again unlike "C, " but as in other para substituted phenols the two high wavelength bands were of about equal intensities. Thus "C" was not 1, l-dimethy1-7-tetralol. - It would be possible to explain the anomalous ultraviolet spectrum of "C" if it were a mixture of the 5- and 6- or 7-tetralols. . This would result in a decrease in the intensity of the high wavelength band. . This mixture would also explain the difficulty in interpreting the infra-red . spectrum with respect to substitution. The easiest way to check this possibility was by gas chromatography. The methyl ether of "C" had shown only one component by this method, and "C" itself had a sharp melt- ing point characteristic of a single compound. . .But gaichromatography , of a sample of the phenols (in methylene chloride solution) shogved two . components in about equal amounts. 'The peaks overlapped considerably. The retentiontime of the first peak was identical with that of "B. " The second peak had the same retention. time as 1, 1-dimethy1-7-tetralol, but 1, 1-dimethy1-6-tetralol was very close. 1 To differentiate between ‘ the two, a portion of the second component was trapped. Its infra-red spectrum showed that it was 1, 1-dimethyl-7-tetralol even though it could not be entirely separated from "B. " As a further test, pure 1, l-dimethyl-S- and 7- tetralols were mixed in equal portions and recrystallized from hexane. The product melted at 83-840 and showed no depression upon admixture with "C" and the infra-red spectrum .was identical to that of "C" in every respect. ’Thus "C" was an approximately equimolar‘mixture of theitwo dimethyltetralols and not a single substance as originally thought. Again. the fact that both. components had the tertialry groupain the meta position was surprising. 26 MA .3235 .8980 5 Hoassoosélafiofiao; .z 1: Auwcofioc/m? 3 a mo £330on postman: .2 ensure. _ q :o .m _ ‘ _ 27 1.91 x 10" moles/liter J w L 1 L 1 250 Figure 12. 260 270 280 290 300 Wavelength (mp) Ultraviolet Spectrum of l, l-Dimethy1-7-tetralol in Cyclohexane. O . 3 N .1 Absorbance 28 This then explains the difficulty in. interpreting the various spectra of "C. " The infra-red Spectrum in the 5-6 p region showed very broad indistinct peaks which would arise from a mixture of 1, 2, 3 and l, 2, 4-tri- substitution. .The same would apply to the long wavelength. region. The ultraviolet spectrum would be merely a combination of the spectra of the two components and give rise to a weaker band at 287 mp. since only one of the components absorbs there. The 280 mp region would not be altered as much since both components absorb there. The double peak for the-methyl groups in the NMR could be explained by the gem-dimethyl groups in each component having slightly different magnetic environments. . Apparently the reason the methyl ether of "C" showed only one peak was that the two components had essentially the same retention times under the conditions employed. The phenoxyacetic acid derivative of "C" melted at 179-179. 50 with softening at 170°. This must have been the derivative of the l, l-dimethy1-7-tetralol portion of "C" sincethe same derivative of the other component‘("B") melted at 141-142.40. As a check, the phenoxyacetic acid derivative of pure 1, l-dimethy1-7-tetralol-was prepared and it melted at 18071820. Probably the derivative obtained from "C" still contained some of the derivative of the other component ("B") which slightly depressed the melting point. This impurity would not affect the carbon hydrogen analysis since it would be isomeric. With the elucidation of the structures of "B" and "C, " the alkylation reaction (the reaction of Z-methyl-5-chloro-2-pentene with phenol) could be shown. OH CH ' o O 1 3\C '3 CH-(CHz)z-C1+ C6H50H —5—O') + / CH, The cleavage of 5, 5-dimethy1homochroman proceeded as shown. 0 OH HO 0 aqueous HBr g + reflux 7 29 ' It was then decided to rerun the reaction of the chloro-olefin with phenol to determine whether any 1,1-dimethy1-7-tetralol was formed in addition to the other two isolated products. This might be expected since it was obtained along with- "B" in the treatment of the homochroman with acid. , The alkylation reaction was run. on a small scale and the crude reaction. mixture was analyzed by gas chromatography. Five components were found in addition to recovered phenol. . Besides the homochroman and "B, " there was a» component which had almost the same retention time as "B" and was present in about the same amount. The two peaks overlapped as they did for "C" and it appeared that this was the expected 7-tetralol. As in the case of "C, " trapping of! this component in a pure state was virtually impossible due to the closeness of the two peaks. However it could be obtained sufficiently pure to show that it was definitely 1, l-dimethyl-7-tetralol by its infra-red spectrum. Thus this same compound was formed in both the alkylation and the ether cleavage along with "B. " He CH, O >C = CH-CHz-CHz-Cl %%%21——> . + CH, 0H + From their‘long retention times the other two components from the alkylation were apparently very highboiling (probably higher alkylation products) and were not investigated since they appeared to be present in only small amounts. It was striking that both. of the phenols produced in the alkylation had the tertiary group in the meta position. This t0pic will be further 30 discussed in. the next section of the thesis which deals with the mechanism of the reaction. C. Mechanism of the reaction of phenol with 2-methyl-5- chloro-Z-pentene In proposing a mechanism for the reaction of 2-methy1-5-chloro- 2-pentene with an excess of phenol at 150°, several considerations must be made. The olefinic double bond must be involved in the initial reaction of the chloride, since saturated primary halides are not reactive. For example n-hexyl chloride reacts only very slightly or not at all under the same conditions (6). The meta orientation of the tertiary groups in, the phenolic products, 1, l-dimethyl-S-and 7-tetralols, must also be accounted for. Rearrangement of the 5, 5-dimethy1homochroman, once formed, to give the phenols cannot be a major course of reaction, since it can be recovered in almost 80% yield when subjected to the reaction conditions (6). Two quite similar reaction mechanisms were considered which differed only in the ion (XV or XVI) formed initially from the homo- allylic chloride. + + :——L-—1 _ r—‘—1 CH .CH 0. N a” 2 x 7- ~ (CH3)2C£}{\ /CHZ"C1 —9(CH3)zc::CHl 01' (CH3)3C"—;C\HE CH, \ H, ‘CH, 8 XV ' XVI Both types of ions have been discussed as intermediates in the solvolysis of homoallylic derivatives (23), and thus also seem applicable here since the reaction Z-methyl-5-chloro-2-pentene with phenol could be considered as solvolytic. Reaction of any of the electron rich centers of the phenol molecule (oxygen atom, or ortho or para ring positions) with ion XV or XVI at the methylene carbon would yield the three olefins XVII-XDC and hydrogen chloride (after loss of a proton). These three olefins 31 0H 0 +0 XV O phenol O °r 1500 a . . + )CVI XVII XVIII iXIX 1 0H1 O llllllllllllll :Hci ’111 17111 3:117 couldthen cyclize in the presence of the acid formed to give the homo - chroman and the two tetralols. These are the only products predicted by this mechanism, and were in fact the only products found. A method to distinguish between the two possible intermediate ions was to use deuterium-labelled Z-rmethyl-5-chloro-2-pentene (XX) in the reaction. .If it proceeded via the unsymmetrical ion‘ (route 1)., 'reaction with. phenol would-lead to the homochroman XXI with all of the deuterium adjacent to the oxygen atom. - If, however, the symmetrical ion (route 11) were the intermediate, phenol could react with either position 0. or (3 of , this ion to give an equimolar mixture of homochromans XXI and XXII which differ only in the position of the deuterium. CH3 0 DD t o \C = CH-CH -CD Cl rou e I unsymmetncal / 2 2 - Ion CH3 \ . XXI route XX .11 . symmetrical ion —-> + XXI CorreSpondingly, the phenolic products would have all of the deuterium adjacent to the benzene ring (XXIII) by route I, or would consist of an equimolar mixture of XXIII and'XXIV by route 11. 32 OH DD _ o D XXIII XXIV Thus by employing labelled 2- -methyl- 5- chloro- 2- -pentene as a reactant and isolating the products, the course of the reaction could be shown. The position of the deuterium in the products could easily be shown by their NMR spectra, since deuterium does not give a signal at the proton frequency. The required labelled chloride was not a known compound and‘a method of preparation had to be devised. The first approach and the results obtained are shown below. 4 ,4-Dimethyl-y-butyrolactone C1 Q0 2. -H,o fl?) H—-C-—-1 (CH3)zC-CH,-CH,-CD,C1 distil 7.72 6.65 (CH3)2C =CH‘CHz-CD2C1 ‘I' (CH3)ZC=CH'CDZ"CH2C1 XXV 'XX + 931, = C(CH,)-CH,—_c_:§_,-CD,C1 XXVI- was reduced with lithium aluminum deuteride and dehydrated to give. 2, 2-dimethyltetrahydrofuran-5-d2 in good yield. . A sample was purified by gas chromatography (20% silicone) and gave infra-red (Figure 13) and NMR (Figure 14) Spectra consistent with this structure. The lack of a peak at 6. 26 Tin the NMR spectrum of theclabelled compound showed that the 5-position was essentially all deuterium. . This peak occurs as a triplet in the unlabelled compound (Figure 14b) due to splitting by the adjacent methylene group. This tetrahydrofuran was then warmed with 33 .oEuoEooSoH C0930 E upumnCdHDHOHptwflouuouinflpoaMQ1N .N mo Eduuoomm ponuonCH A3 nuwcofiocfim? ma 3 o p m .2 Sosa m J - . fl _ _ 6030300508 £00230 5 .Hoioonz cozonflco 3v sols.-eonoeo8323333854 .N 3 no Eonsooam fizz .3; use 2; measure . on .o _ 3V nm0 .o O \r m emu awé (a new a a £0 35 Lucas' reagent to give presumably the dichloride. . This was not purified, but was heated to give a mixture of chloro-olefins (XX,.. XXV and XXVI). The‘presence of Z-methyl-S-chloro-1-pentene-5-dz was indicated (ca. 20% of the products) by peaks in the infra-red spectrum at 6. 06pand 11. 22p' characteristic of terminal olefins. This could not be separated completely from the other products by gas chromatography (20% silicone). The infra- red spectrurn (Figure 15) of the deuterated 2-methy1-5-chloro-2-pentene obtained from gas chromatography still showed these peaks, but with. less intensity. TheNMR spectrum (Figure 16) of this product also indicated the presence of this other isomer (ca. 15-16% by integration); it also showed that the deuterium in the 2-olefin was distributed equally between the 4- and 5-positions. The small peaks at 5. 29 7’ and ca. 87‘ corres- ponded to the vinyl and methylene protons (underlined) in XXVI. The fact that the 2-methyl-5-chloro-2-pentene was an equimolar mixture of the 4- and 5-dideuterocompounds (XX and XXV) was shown by the areas of the 6.65 7’ and 7. 72 7’ peaks, since they appeared in a 1:1 ratio. . If the product were all XX, then no protons would be found at 6. 65 7" ' and two at 7. 72 T . . The opposite would-apply if it were all XXV. Thus this material was not suitable for use in the mechanism study. 7 The NMR . spectrum of unlabelled 2-methy1-5-chloro-2-pentene is shown. in Figure 16b for comparison. .A second route to the desired labelled compoundris shown below. 6.49 7.96 6.46 7.74 Br-CHz-CHz-COzCsz $72i-9 Br-CHz‘fCHz'CDzOH SEE.) Br-CHz‘VCHz‘CDzCI 3 ‘ XXVH XXVHI KCN 7.57 7.95 7.48 7.90 C,H,0,C—CH,-CH,-CD,C1 «(5%? NC—CH,-CH,-CD,C1 XXX XXIX _H o 8.31 14.88 7.58 (CH3)ZC(OH)-CHz-CH,-CDZC1 —§-——> (CH,),C=CH—CH,-CD,C1 _-XXX1 .XX (The NMR 7"-va1ues are given above the appropriate protons.) 36 ME .opfinofiaomflofi C0930 Cw Nptm pom notoCouComumnouofioimufwnuogiw mo 55.30on pouumnwcH 0‘ 11V aumcoaocyo? N. .2 ensure _ : "1m 37 .oEuoEomSoB connoU GM .Howuouoz oozooonoo a: and one -e-ocoocoo-~-ono3o-m-Efiozé 3 no 6:388 ~32 of use .2: ensure. k0 red in NoN moo : . one I J_ .3III/\/\./n\/i\ 34 «x A3): 3v 3 5 Pt m... Amy 30 No 3; 38 $4. moo sun oar. / n_\1 6.46 7.74 7" 0 B (3 7 «1 Br-CHZ'CHz-CHZ‘CI 'Y (\b)AZLZIIIIK/\'\ AJIW >- 6.33 6.46 7.75 *7" Figure 20. NMR Spectrum of (a) l-Bromo-3-chlor0propane-3-dz, (b) Unlabelled Material, in Carbon Tetrachloride. 43 MA .opCOHAUMSoH co9~o0 Cw ~01wuofinficouuwuaaaouodanw Ho 8550on postman; HA 1: Sewage/m? N. .3. ensure _ j ‘ 44 8 o. C1-CDz-CHz-CHz-CN 7.48 7.90 7“ Figure 22. . NMR Spectrum of 4-Chlorobutyronitrile-4-d, in Carbon Tetrachloride. 45 This type of spectrum(mirror image multiplets) also occurs in. l-chloro- 2-bromoethane, whichhas been discussed in the literature (25). No attempt was made to interpret the spectrum of XXIX due to-its com- plexity. -Apparently the multiplet at 7.487‘ was due essentially to the protons of the -CHZCN group, while the one at 7. 90T' (broadened) was due to the methylene group adjacent to the deuterium. The chloronitrile XXIX was refluxed with ethanol in the presence of hydrogen chloride to give ethyl 4-chlorobutyrate-4-dz (XXX) in 83. 5% yield. The infra-red spectrum (Figure 23) was consistent with the structure. . The NMR spectrum(Figure 24) was essentially the. same asthatof themitril’e, with two multiplets..(mirror images) at 7. 57 T'Iand 7. 95 T" (broad). The inter- pretation of this was the same as for the nitrile. There was also a triplet for themethyl protons (8. 74 1- ) and arquartet for the methylene protons (5. 90 T") of the ethyl group in the ester. The ester XXX was treated with excess methyl magnesium iodide in ether to yield the alcohol XXXI. This was not purified, but an infra-red spectrum (Figure 25) of the crude material did support the structure. The alcohol was dehydrated by distillation .fromfreshly fused potassium bisulfate. The product of this was mainly the expected 2-methyl-5-chloro-2-pentene-5-dz (XX), along with. some Z-methyl-S-chloro-l-pentene-S-dz (XXVI), and labelled 2, 2-dimethyltetrahydrofuran. The main product was separated by gas chromatography (20% silicone), although it could not be completely freed (of the l-isomer. This difficulty was also encountered in the earlier attempt to prepare XX. This time the NMR spectrum (Figure 26) indicated that the deuterium was present in the 5-position and not distributed between the 4 and 5-positions as before. This fact could be shown from the peak areas. The triplet at 4. 88 7' arose from the one vinyl proton being split by the two allylic methylene protons. . This was further Split by the two methyl groups. The doublet at 7. 58 7" represented two allylic methylene protons being split by the vinyl proton and broadened by the adjacent 46 .opflogoouuofi C0980 Ca mpuvuoumCaudnonoHAUIw 35.0 0 Eduuoomm pontoCmCH 1.: CamCoHoZmNS : o N. m .2 3:ow _ J C . .mguozumuumh. Gen—HMO 5 NauvuoHMu>usnouoEUuw TFBH mo E53025 ”322 .vm vudmmh 33m mag. >m.> cod 47 § W «mokmukookmokmokavo , 4 d a r 4 a 48 .oEuoEumuuoH conumU 5 NvumufiocmfikmmumuOHOH£oumuanumgnm mo 93.30QO “Enigma: .mN muswwh A3 suwcofimfim? ma 3 o N. m m . _ _ _ i J 49 .mEpoEomuumB conumU 5 mvumumcmucmaumuouoioumugfimgam mo 9330QO fizz . .oN madmfm \n fid 5d 3.5 mmg $3 qurllaw— 14 u a was $4 w a N... “‘F‘ I V NM 60 mam .H 93 mafia—“393% _-_F— Updokmoio u ognmuv ,V a d k 50 deuterium atoms. (Note that this value is somewhat different fromthat determined on XX prepared via the Lucas reagent. The former was examined on. the HR-6O and the latter on the A-60. The latter is more accurate for assigning T—values.) A The doublet at 8. 31 7" was for the six methyl protons. The small peak at 5. 28 7" was due to the terminal methylene protons of the isomeric impurity XXVI (ca. ‘ 13%). A The weak peak at 6. 48 7" indicated the presence of about 8-10% of material un- labelled in the 5-position (possibly a small amount of XXV). The infra- red spectrum-(Figure 27) was consistent with the structure XX, and the weak bands at 6. 06p. and 11. 22p. were characteristic of the terminal olefinic impurity XXVI. . r This Z-methyl-S-chloro-2-pentene-5-d7_ was next heated with an excess of phenol for 8. 5 hours at 150°. The neutral material was separated in the usual manner and then purified by gas chromatography (20% silicone) to obtain. labelled 5, S-dimethylhomochroman in a 15% yield as colorless plates. The infra-red» spectrum of this ether is shown . inFigure 28. r The NMR spectrum(Figure 29) indicated four aromatic protons (ca. 3 ‘7’) as a complex multiplet, a single peak at 6. 15 7'" representing (in area) only one methylene proton adjacent to the oxygen atom,. a complex multiplet (ca. 8. 3 7")corresponding in. area to three other methylene protons; -and a sharp peak at 8. 66 T for six methyl protons. The NMR spectrum of the unlabelled homochroman is shown in‘Figure 35 for comparison. . This spectrum. showed an identical aromatic region (four protons), a triplet at 6. 14 7" for the two protons adjacent to the oxygen atom (split by the adjacent methylene group),. a complex multiplet for four methylene protons (ca. 8. 2 7"), and a sharp peak at 8. 677' for six. methyl protons. From these data it was concluded that the deuterium was equally distributed between the two positions as shown inXXI and XXII. 51 .opcozomhoh. conumU 5 Np..mnocoacomumnonogouméwfluogum Ho 5530on pouumuwsH 1.: Auwcofiocfiw? 2 S a a . m .3 033m _ a . _ . 52 .0 Co v. Eomnuofi conumU 5 cmaougoogoaafioafiflnm .m veganousofl mo £3.30on popuofiwfl t: “images/m? 2 S o N. m .3 0.33m _ - a q q _ ‘ ‘b 53 Al‘ .oEnoEomfioB connmU cw cmEOHAUOanaguocflfiQum .m woumngdofl Ho 8550on M342 cod 1f g g _ r. < omd mic 5 A: V Kb .00 moo n on“ 332%"... _ _ _ _ _ _ _ _ _ _ _ _ _ T _ LB .7 mo.m 2; .mm 39mg 54 D XXI . XXII Had all of the deuterium been present as in XXI, then no peak at 6. 14 7" , would have been seen, and four methylene protons would have been . found around 8. 2 7" instead of three. The presence of a single peak at 6. l4 7' (one proton) showed that half of the material was structure XXI since these -CHZO- protons were not split by'any adjacent hydrogen ‘ atoms. The data are summarized in Table 3. Table 3. Calculated and Experimental NMR- Peak Areas for Deuterated 5, S-Dimethylhomochroman \ Aromatic ~CHZO- -(CH2)- (033); Protons Protons Protons Protons Unlabelled material 4 2 4 6 Calcd. from. XXI 4 0 4 6 Calcd. for equimolar mixture of XXI + XXII 4 1 3 6 Found 4 l 3 6 The labelled phenolic product was isolated and most of the unre- acted phenol was removed by distillation. The remaining mixture of phenols was purified by gas chromatography (20% silicone) and collected as armixture of the deuterated 5- and 7-dimethy1tetralols, since they were difficultly separable as mentioned before. The NMR spectrum of this mixture (predominately the 5-isomer) could be interpreted easily 55 since the corresponding peaks of each isomer occurred at almost identical T—values, as was seen in the spectrum of compound "C. " The necessary data could be obtained by comparing the areas of the benzylic methylene and the other methylene protons of this mixture. The spectrum (Figure 30) showed a multiplet (ca. 3.47) for the three aromatic protons, a single peak at 5. 28 7' for the hydroxyl proton, a double peak at 7. 387" and 7.437' for the benzylic methylene protons of each isomer (the one at 7.437’ was larger due to the excess of the 5-isomer), which together corresponded to one proton in area, a multiplet around 8. 4 7" which had a total area corresponding to three methylene protons, and a double peak at 8: 777' and 8. 847' for the methyl protons of each isomer. The data are summarized in Table 4. OH D D OH D XXIII XX IV From this it was seen that the phenolic products were an equimolar mixture of structures XXIII and XXIV. The spectrum of an equimolar mixture of unlabelled l, l-dimethyl-S- and 7-tetralols is shown in Figure 6 for comparison (compound "C"). Table 4. . Calculated and Experimental NMR Peak Areas for the Deuterated Phenolic Products m M Unlabelled Structure Equimolar mixture Phenols XXIII of XXIII and XXIV Found Benzylic Protons Z 0 1 1 '(CH2)2- Protons 4 4 3 3 56 (1) (3) I Amplitude I increased 4X I I < l (3) II I I I I I (1) ' I I N I I l I I I l I l J , , I l I > 3.41 3.53 7.387.43 I ' 8.42 8.77 8.84 Figure 30. . NMR Spectrum of Deuterated 1, l-Dimethyl-S— and 7-tetralols in Carbon Tetrachloride. 57 Since the deuterium label was found to be distributed equally between two adjacent positions in all three of the reaction products, it was concluded that the mechanism must involve an intermediate in which the two methylenes of the original chloro-olefin become equivalent. Presumably this is the symmetrical ion XVI. The recent work of XVI Whitham (26) indicates this same type ohsymmetrical ion is an inter- mediate in the solvolysis of cholesteryl derivatives. R0 58 Part II A. Stereochemistry of 5, 5-dimethy1homochroman The ultraviolet spectrum of 5, 5-dimethylhomochroman (Figure 31) suggested some unusual conformational properties (6). .The low absorption intensity at 271. 5 mp. ( e = 725) could be interpreted in terms of the -CHz-O- bond being twisted out of the plane of the benzene ring, a condition not favorable to maximum resonance interaction between the unshared electrons on the oxygen atom and the Tr-electrons of the aromatic ring. Baddeley (27) had previously demonstrated this phenomenon by means of ultraviolet spectra and reaction rates. Table 5 summarizes these data. . In coumaran the -CHz-O- (bond is held in the plane of the Table 5. Properties of 5-, 6- and 7-Membered Cyclic Aromatic Ethers (27) kr'el} krel} (solvolysis xmax. (mil) 6 (bromination) of p-chloromethyl , derivative) coumaran 289, 283 3055, 3099 76 ‘ 363 chroman 284, 276 2520, 2125 28 106 homochroman 272, 267 694, 678 1.4 1.86 benzene ring and maximum resonance interaction can occur. In chroman and homochroman the -CHz-O- bond is progressively twisted out of the plane described by the aromatic ring and thus less resonance interaction can occur. This is reflected in the ultraviolet spectra by a decrease in intensity and a shift of the maxima to shorter wavelengths. This effect is also shown in the decreasing rates of aromatic bromination, and of solvolysis of the para- substituted chloromethyl derivatives. Both reactions are favored by groups which increase the electron density on 59 4. 5 x 10" moles/liter L 1 l 256 262 268 274 280 . 2 .1 Absorbance Wavelength: (m p) Ultraviolet Spectrum of 5, 5-Dimethy1homochroman .in Cyclohexane. Figure 31. 60 the ring. As the interaction of the oxygen atom with the ring decreases, the rates fall off sharply. A sample of 5, 5-dimethylhomochroman was sent to Professor E.. M. 'Arnett (28) who was studying quantitatively the basicity of various ethers toward the proton. It was found to be about 104 times more basic than anisole. This shows that the unshared electrons on oxygen in the homochroman are less involved in resonance interactionwith the n-electrons of the aromatic ring than in anisole, and are therefore more available for accepting a proton. Two enantiomorphic chair forms of the homochroman are possible '(IIIa, b). IIIa The -CHz-O- bond was found to be 75-800 out of the plane of the ring from measurements of Dreiding models (29). This) was in agreement with the observed babicity and spectral data. Further examination of the, models showed that it was impossible to go from conformation Illa to -IIIb 'without first going through the boat forms IIIc andIIId. And in the conversion of IIIc file to IIId an intermediate conformation IIIe arose'in which there was a serious. non-k'bonded interaction betweena C5 methyl group and a C2 hydrogen atom. This interaction is shown below. Without a‘ngle deform- ‘ o . . ation, this overlap amounts to about 0. 3 A (as measured from models).- 61 If this barrier to interconversion were sufficiently high, then it .might be possible to resolve the two-enantiomorphs. The firstmethod tried was to place a carboxyl group on the aromatic ring and then attempt to resolve the acid via its brucine or quinine salt, The 5,5-dimethyl- homochroman used in this work was prepared by the reaction of phenol with 2, 2-dimethyltetrahydrofuran and hydrogen “chloride; The crude ketone XXXII was not usually purified but used directly in the haloform reaction. The yield of the crude ketone was on the order of 90%. O CH3COC1‘ NaOBr A1C313 7 ‘ NaOH A CH, CH3 XXXII 3 H3 XXXIII In one case it was distilled (110-1120 at 0.091mm.) yielding a colorless viscous oil which would not solidify. However a solid dinitrophenyl- hydrazone derivative was prepared (m.p. 199. 5-201. 5°) which gave a correct analysis. The infra-red spectrum of the ketone (Figure 32) was consistent with the indicated structure. The haloform reaction on the crude ketone yielded 5,.5-dimethylhomochromancarboxylic acid (XXXIII) in nearly 70% yield (m.p. 174-1750). It was not a known 62 .obflfidmwfl 280230 5 coaougooaoauifioomahuaifioawflam .m mo Eduuoomm pounmnwcm 1: 5.983335 2 3 o N. _ o m .3 833m a _ n J. _ 63 compound but gave a correct analysis and neutralization equivalent. The infra-red spectrum (Figure 33) was also consistent with the structure. The 5-6 p region of the infra-red spectrum fitted the 1, 2,4-substitution pattern (12) with bands at 5. 26, 5. 55, and 5.81 p. The carbonyl absorption would obscure the higher wavelength band (ca. 6 p). The amide and methyl ester which gave correct analyses’were also prepared. To demonstrate that the seven-membered had remained intact throughout the synthesis, the acid was degraded to the homochroman by the route shown. The acid was converted to the amide in essentially a 1. 3,0012 Naoqj , '1.HNO§ 2. mon Ar‘CONHZ NaOH Ar‘NHZ 2.H3Po, HOZC XXXIII 0 III quantitative yield using thionyl chloride followed by ammonium hydroxide. The amide underwent the Hofmann rearrangement with sodium hypochlorite to give the amine, which“was not isolated but was directly diazotized with nitrous acid and the resulting diazonium compound reduced with hypophosphorous acid to give the homochroman. III which was identified by its melting point and infra-red spectrum. The yield based on the amide was 42%. The position of the carboxyl group in 5, 5-dimethylhomochroman- carboxylic acid was not unequivocally determined, but it was in either the 7- or 8-position since the infra-red spectrum showed 1, 2,4-substitution. The 7-position is most likely since the acid was prepared via an electro- philic acylation which would presumably put the entering group para to the oxygen. A comparison of the ultraviolet spectrum of the acid (Figure 34) with those of several other aromatic acids also supports this conclusion and is shown in Table 6. 64 633520 coornmu cw 30¢. SaxonumouwucmEougooanTffioEflQum .m mo 53.30on pouanwcm 3 £82383 Ms 2 a ~_8 m .2 3&3 ‘DH _ — _ 65 6.70 x 10-5 moles/liter I I I I 240 250 260 270 Wavelength. (mp) Figure 34. Ultraviolet Spectrum of 5, 5-Dimethylhomochroman-7-carboxylic Acid in Cyclohexane. 66 Table 6. . Ultraviolet Spectra of Some Aromatic Acids V —r . ,_fi )‘max. (mu) 6 Reference L n meta-anisic acid 234*, 292 7800, 3100 ‘ 30 para-anisic acid 250 17400 30 benzoic acid 231, 273. 5 10,000, 1000 31 5, 5-dimethylhomo- 253. 5 10,900 ' Figure 34 chromancarboxylic acid The homochroman acid resembles para-anisic acid with respect to the position of the maximum and the molar extinction coefficient. The extinction coefficient of the homochroman acid would be expected to be lower than that of the corresponding anisic acid (but not as low'as benzoic acid) due to the :CHz-O: bond being twisted out of the plane of the ring. Thus the best structure for the acid appears to be 5, 5-dimethyl- homochroman-7-carboxylic acid. Resolutiionlof the acid was attempted using brucine in acetone. Only a small amount of low-melting solid was obtained even after removal of part of the solvent. The use of quinine in acetone or 10% acetone-ethyl acetate also gave the same results. It was decided to examine the NMR spectrum of 5, 5-dimethylhomochroman itself at various temperatures. If the seven-membered ring were rapidly interconverting from one enantiESmorphic chair form to the other, then the two methyl groups should give a single sharp line (being magnetically equivalent). Should the molecule, however, be locked in a chair conformation, then one of the methyl groups would lie in the plane of the benzene ring and. the other would be almost perpendicular to it. Thus they would not be-magnetically equivalent and should give rise to two peaks Or one broad peak in the NMR. 67 At room temperature the NMR spectrum (Figure 35) showed only a single peak for themethyl groups. Upon cooling to -1000 no change in the spectrum was noted.- It was concluded'that the seven-membered ring was still interconverting and that the barrier was insufficient for r e s olution. B. Synthesis and stereochemistry of 1, 1,4, 4-tetramethyl- benzocycloheptene Since the barrier to interconversion of the two enantiomorphic forms of 5, 5-dimethylhomochroman appeared too small to permit resolu- tion, it was decided to synthesize a similar molecule in which the inter- action between the 2- and 5-positions would. be increased. The compound chosen was 1, 1, 4, 4-tetramethylbenzocycloheptene (XXXIV). . In this case the interaction would occur between two methyl groups in the XXX IV transition between the two enantiomorphic chair forms. In the case of 68 AL (x \O ('1) .opwpogumuuoh £03.30 5 :mEOHAUOEOQTFSOCGMQIm .m Ho 8550on #322 on .m _ ._________ __ ______,__h v KN. be new u 05 35384 : ~\ wad : E : d wo.m .3 38E 69 the homochroman studied this interaction occurred between a methyl group and a hydrogen atom. The reason for using a hydrocarbon rather than the analogous homochroman (XXXV) was a matter of stability. . O XXXV Tertiary aromatic ethers such as XXXV are not easily handled due to their sensitivity to acids (32). The NMR spectrum of the hydrocarbon XXXIV could be examined to see if it were interconverting from one chair form to the other, as was done with the homochroman. . If it were locked in one conformation the NMR peaks for each set of gem-dimethyl groups would be broadened or possibly occur at different fields due to their non-equivalence. The peak corresponding to the benzylic methylpne protons might also dis- tinguish between them, since one proton would be quasi-axial and the other quasi-equatorial. The desired hydrocarbon was not a known compound and had to be synthesized. . The first route takenis shown, in the scheme. C6H5-C(CH3)z-CHz-COZH ‘Iil—AlI-ili» C6H5-C(CH3)z-(CHz)z-OH \I’bsCl 1. malonic estej 1' CbHs’CICHah-(CHzlscozH 7. OH“, -co,, C6H5'C(CH3)2*(CHz)z-OTS 1.socn 2- A1c1, 1. NaNH W. K. ('2) CH312. ”L XXXVI XXXVII XXXIV 70 This route had been used by Julia (33) to prepareXXXVI. . 3-Methyl- 3-pheny1butyric acid was prepared by thevmethod of Dippy and Young (34) in 94% yield from condensation of 3, 3-dimethylacrylic acid. and benzene with. aluminum chloride. This acid was reduced to the corresponding alcohol in 85% yield by lithium aluminum hydride. The infra-red. spectrum . of 3-methy1-3-phenyl-l-butanol is shown in Figure 36. This alcohol was converted to its para-toluenesulfonate in a quantitative yield. . The tosylate was'used to alkylate malonic ester and yielded 5-methyl-5-phenylhexanoic acid after saponification and decarboxylation. The yield was 54% (infra- red spectrurn, Figure 37). r This acid was converted to the corresponding _ acid chloride (95% yield) which was then cyclized with aluminum chloride to give 5, 5-dimethylbenzocyclohepten-l-one XXXVI in 78% yield. - A dinitrophenylhydrazone derivative was prepared and melted at 160-1620. . Julia (33) reported a melting point of 1620. The infra-red spectrum of this ketone is shown in. Figure 38. The ketone was dimethylated in two steps using methyl iodide and sodium amide to give 2, 2, 5, 5-‘tetramethyl- benzocycloheptene—Lone XXXVII in 64%) yield. The infra-red spectrum (Figure 39) was consistent with the structure. The ultraviolet spectrum (Figure 40) had a peak at 240 mp. ( 6 = 5840), and a shoulder at 272 mp. ( 6 = 960). The low extinction coefficient of the main band was consistent with the carbonyl group being twisted out of the plane of the benzene ring. For example acet0phenone, where the carbonyl function can easily be coplanar with the ring (for maximum resonance interaction) has a high extinction ( 6 = 13,000) at 240 mp (35). No carbonyl derivatives of this tetramethyl-ketone could be prepared. This was not too surprising since the methyl groups tend to hinder any approach to this center. The ketone did give a correct elemental analysis. 'It had been, previously thought that the carbonyl group of the tetra- methyl ketone could be reduced by a Wolff-Kishner reaction to give the desired hydrocarbon. . However, since the ketone would not form a 71 48.8073 Hosanna;nacoamumnrfifioznm mo €5.30on penumumcm . .om onswfim A3 gumsofiouwo? o N. m d u d 72 MA .opfiuogumuuob GonzamU a“ p34 ofiocoxozTncoamumquEuozam Ho 8530on penumumcH 3 £82263 .2 o . n m .8 28E m a _ a _ 73 .opfihofifloouuofi conumU am 6:04 Iodoumoaofiou»00583133anum .m mo ash—06mm popumenH .wm 0.3%er 1: sewage/m? 2 3 me n m m _ - . _ . . 74 .opgsmwfl conflumU cm 0G0;umeoumvoHoenuoncoflnffie58.30,Hum .m .N .N mo 8336on pohimnwdH .om ouswwh A3 Aumcofieurm? 2 3 me N. — o m m n. . .I d .oGMXofloHuunU a“ one;nocoumoaofiuuwooNcoflgfluoamflm,H...m .m .N .N mo 85.8.0on uofiogwufib .1 EV ”imaged/m? .3. ensure com 0?... CNN oom omm owm omm _ n _ _. d a .— om r or I 06 00 IT...- 1 w. n m S m. om Io u 00 I cm I uofioouco£>£uoc§muuouumum .N .NIOHOEUIH mo asbuomm postman: .ch ouswwh A3 sumcmfiouwd? 2 Z a n 3 m. m _ _ _ fl _ _ 79 temperature. . The product was a complex mixture (via gas chromato- graphy on 20% silicone) and the component present in the greatest amount was separated. The infra-red spectrum of this material is shown in. Figure 43. This may have been the desired hydrocarbnn, but it was not investigated further since it was obtained in only a very small amount. . Instead a better synthetic route Was sought because of the length. of the previous one and also the unreactivity of the intermediate tetramethyl-ketone . The scheme finally arrived at is that shown. This involved the p o c,,Ii,,-§-.c1i(cn,)2 + (CH3),c = CH-CHZBr 31% 5 Catalyst XL XXXIV ' XLI alkylation of isobutyrophenone with 1-bromo-3-methyl-2-butene using sodium amide. This reaction afforded an 87. 5% yield of l-phenyl-Z, 2, 5- triniethyl-4-hexen-l-one XL as a colorless liquid. This was not a known compound. . Its infra-red spectrum is shown in Figure 44 and is con- sistent with-the structure. Titration of the double bond (bromate-bromide) showed 1.01 olefinic double bonds per'molecule. A 2, 4-dinitropheny1- hydrazone was prepared (m.p. 86. 8-87. 2°) and gave a correct elemental analysis. This ketone (XL) was reduced by the Wolff-Kishnermethod to give a 62% yield of 2, 4, 4-trimethyl-6-phenyl-2-hexene (XLI). A sample of the liquid olefin was purified by gas chromatography (20% silicone) and gave correct analytical data. The infra-red spectrum of this olefin is shown in Figure 45. Titration of the double bond content gave 0. 99 double 80 033520 conumu 5 nonconnm :oflospom on» mo £3.30on postman: A3 “imaged/m? 2 a e _ o m .3 $an ‘ d d _. 81 «L .opgdmwfl conumU cm ocotalocoXoatwnififioECutm .N .Ntfincoamna mo 55.30on pontoumcH A43 guwcoaouwmg NH om w o .3. enema _ q _ 82 .opCoEomuuoB conHmU cw oCoXosumtacoamuougfioemnhtv .v .N mo Sunbeam postman; .mv osswflh A3 Jawsofioxemg 3 NH 2 m 0 av d _ . . a . . 83 bonds per molecule. . To prove that the double bond was in the position indicated inXLI, a sample was ozonized and the acetone produced was isolated as iodoform by the method Of Bailey (37). . The yield was 76% which. indicated that at least 76% of the double bond was in the position desired. - Attempted cyclization. of this olefin with aluminum chloride at room temperature to give 1, 1, 4, 4-tetramethylbenzocycloheptene yielded a mixture consisting of eight components (by gas chromatography). This was not investigated further. Cyclization at 00 with aluminum chloride and hydrogen chloride gave a single product. The infra-red spectrum(Figure 46) showed 1, 2-disubstitution. The ultraviolet Spectrum (Figure 47) had bands at 274 mp. ( 6 = 935), 267 mp. ( 631020), and 260 mp ( 6 = 930). A mass spectrum showed a major fragmentation route to be 202 —9 159 + 43, which meant that a three carbon fragment was easily lost. A sample was purified by gas chromatography (20% silicone) and analyzed for Clstz. . This material was probably the previously unknown 1-isopropyl-3, 3-dimethyltetralin (XLII), resulting from closure to give the less strained six-membered ring. . The NMR Spectrum (Figure 48) of the product was complex. X LII It was decided to employ the milder catalyst, boron trifluoride- etherate, for the cyclization of the olefin. The reaction was carried out in methylene chloride for 21 hours, and gas chromatography indicated two products, in a ratio of 2:1. The major component was identical in retention time and infra-red spectrum to the supposed isopropyl com- pound XLII formed when aluminum chloride and hydrogen chloride were used in the cyclization. The minor component, although. having the same 84 .ogfidmwfl consmU cw Atcfimtouaguegfivum .muamoumoflaa mo 8530on contour: A3 sewage/m? 2 2 o a _ o m .3 93E "WM f u q 0 — _ 85 1. 98 x 10-3 moles/liter l l I 250 270 290 Wavelength (mp) Ultraviolet Spectrum of 1—Isopropyl-3, 3-dimethyltetralin( ?) in Cyclohexane . Figure 47. Absorbance 86 .mpfiuoEomsuoB ~50.st g :vcflmhuofilfifioaflpum .Mugmoumoflua mo Esau—00mm .522 .wv oufimfm “1x 25, was 3.x ng ‘ L. 35 - 5:. z _ j d _ m : } ‘4 .fi _ Km .N commence“ _ 332mg}. 87 retention time as the starting olefin, was 1, 2-disubstituted from its infra-red spectrum (Figure 49), and showed no unsaturation (bromine in carbon tetrachloride). It analyzed for Clstz- The ultraviolet spectrum (Figure 50) showed a very low extinction coefficient ( e = 232 at 262 mp) characteristic of a benzocycloheptene. . For example, Arnold (38) has shown that the extinction coefficient drops off as the size of the ring changes from five to seven (Table 7). . Further support that this Table 7. Ultraviolet Absorbtion of Some Aromatic Hydrocarbons (38) N xmax. (mu) 6= 5 273 1450 6 274 627 (CHZ)N-2 7 271 292 product was indeed the desired 1, 1, 4, 4-tetramethylbenzocycloheptene (XXXIV) came from the mass spectrum. This showed a major frag- mentation route to be 202 ——> 187 + 15, which indicated a‘methyl group was easily lost (benzylic). The NMR spectrum (Figure 51) also supported 7.34 9.14 ca. 3.4 / 8. 68 XXXIV the structure. This showed four aromatic protons (ca. 3.91"). a single peak for two benzylic methylene protons (7. 34 7"), a multiplet for four 88 633530 £03.80 GM ocoumoflofiotwoonsmflgauoflfimuum,an .v J J m0.§9uuoemm wouumecH .ov 0.2;me 1.: AuwceHQ/m? 2 3 o. h T. m m w a d _ q — 89 a ou'eqzo s qV J.o N O M O m.o .mGdXoaoHo>U Ga ocoummgofiofwoonconaguofimnuo81v JV J J mo 5550on uofiogmsfib .om 6.2..th oom own A165 fiumcoaokrd? com owu b ‘p u H0u3\moHoE nnoJ x wH .N 90 (6) (6) Amplitude increased 3X (2) < (4) (4) ——-—- ———— ———-——-b—_ ‘ l 1 I -I'L jnkll _ 1 [IMMLV/WA ‘ N 7.347F 2.85 3.11 8.44 8.68 * 9.14 ‘7' Figure 51. NMR Spectrum of l, 1, 4, 4- -Tetramethylbenzocycloheptene in Carbon Tetrachloride. 91 other methylene protons (8.44 7'), and sharp peaks at 8.68 7“ and 9. 14 7" , each representing six methyl protons for the gem-dimethyl groups on the l-and 4-positions respectively. The fact that the peaks representing the gem-dimethyl groups were sharp indicated that they were equivalent and that the two enantiomorphic chair forms were rapidly interconverting at room temperature. The sharpness of the peak for the benzylic methylene protons also supported this conclusion. This peak would be' broadaor possibly separated into two peaks if one of these protons were locked in a quasi-axial and the other in a quasi-equatorial conformation. ‘-._EI‘he_acyclizationr.reaction was repeated using boron trifluoride etherate and following the reaction by gas chromatography. . It appeared that the reaction was complete after 2 hours, since the peak ratios became constant at 2:1 after that time. This would indicate that the formation of the two products was kinetically controlled and that the seven-membered ring, once formed, was not isomerized under the reaction conditions . CbHs-CHz-C(CH3)z-CHz-CH = C(CH3)2 -'B—Fi<: Ferric chloride and stannic chloride were alsotried as catalysts. . Ferric chloride gave the two products in the same ratio (2:1)'and even more rapidly than with boron trifluoride etherate. . No reaction occurred with stannic chloride. The product obtained in the reduction of 2, 2, 5, S-tetramethylfil- chlorobenzocycloheptene withpalladium appeared to have the same 92 infra-red spectrum as the one obtained from the boron trifluoride- etherate cyclization. . No further work was done on this since the desired product was easily prepared by the route just described. At this writing the NMR spectrum of 1, 1, 4, 4-tetramethylbenzo- cycloheptene is being examined at low temperatures by Dr. Ernest Grunwald at the Bell Telephone Laboratories. . Preliminary data indicate that the interconversion process is drastically slowed down at -30°. . At -600 both peaks for the methyl groups are further resolved into doublets. ' EXPERHVIENTAL . 93 EXPERIMENTAL Gas chromatography Gas chromatography was done on. either a BeckrnanMegachrom or a Perkin-Elmer Model 154. The columns were packed with. silicone on Chromasorb W, usually 20% by weight. Spectra The infra-red spectra were obtained on a Perkin-Elmer Model 21 instrument, using sodium chloride cells. The ultraviolet spectra were obtained on a Beckman DK-2 and a Beckman-DU. The curves were first obtained on the DK-Z and the extinction coefficients then measured accurately on the DU instrument. . The NMR spectra were obtained on either Varian Model A-60 or a Varian Model HR-60 instrument. .All spectra were obtained at 60' Mc. using tetramethylsilane as aninternal standard. The band positions were recorded in Tunits as prescribed by Tiers (39). The relative peak areas were obtained by electronic integration. The mass spectra were run at the American Oil Company through the courtesy of Mr. Seymour Meyerson. . Mic r oanalys e s All of the microanalytical data were obtained from the Spang ’Microanalytical Laboratory, Ann Arbor,. Michigan. Melting points All melting points are uncorrected. 94 95 Part I A. Synthesis of 1, l-dimethyl-S-tetralol (VIII) Preparation of 1, 1-dimethyl-5-methoxy-2-tetralone The procedure of Cornforth (18) was followed. . Sodium metal (5. 2 g.) was added over a 10 minute period to a refluxing solution of 5. 0 g. (0. 0266' moles) of 1, 6-dimethoxynaphthalene in 60 ml- of ethanol. The solution was refluxed until no more sodiumremained (about 45 minutes). Water (57 ml.) was added followed by 57 m1. of hydrochloric acid (as quickly as possible). The mixture was refluxed for 15 minutes to cleave the enol ether. After cooling, it was extracted three» times with ether. The combined extracts were washed successively with water, 10% sodium bicarbonate, water, and dried over potassium carbonate. The solvent was removed on a rotary evaporator, and the resulting oil was shaken with 60 ml. of saturated sodium bisulfite solution for 10 minutes. After standing overnight in the cold, the crystalline adduct was filtered off and washed with ether. 7 The white solid was stirredwith 60 ml. of water (partially dissolved) and 50 m1. of 20% sodium carbonate was added. Stirring was continued for 2 hours at 20°. The organic material was taken up in ether (three extractions), washed with water, and dried over potassium carbonate. 7 Removal of the solvent on a rotary evaporator yielded. 2.0 g- (43%) of the ketone as a yellow oil. . This was not purified further. . The crude 5-methoxy- 2v-tetralone (2.0 g. , 0.011 moles) was added to sodium isopropoxide (from 0. 53 g. of sodium metal and 20 m1. of Z-propanol). The solution was cooled and4. 5 g. of methyl . iodide in 10 m1. of Z-propanol was added slowly. The reaction mixture was refluxed for 3 hours, cooled, and 5 m1. of sulfuric acid in 20 'ml. of water was added. , The organic material was extracted into ether (two extractions),the combined extracts washed with Water, and the solvent 96 removed on a rotary evaporator. The semi-solid which remained was shaken with saturated sodium bisulfite solution to remove any starting material. . The mixture was extracted with ether and the solvent removed. . The yellow crystals remaining were recrystallized from hexane twice to give a pale yellow product (1, 1-dimethyl-5-methoxy-2-tetralone) melting at 81-820 (1.7 g., 75%). The semicarbazone melted at 189-1900 after one recrystallization from dilute ethanol. Cornforth reported melting points «of 83.-84° for‘the'lvetonenand;19.2r1.9.491for'thetfsemicarbazone (18). Preparation of 1, l-dimethyl-S-tetralol (VIII) A mixture of 1.4 g. (0.0069 moles) of 1, l-dimethyl-S-methoxy- Z-tetralone, 2.0 m1. of 85% hydrazine, 2.4 g. of potassium hydroxide, and 18 m1. of ethylene glycol was refluxed for 3 hours. The solution was distilled until the pot temperature reached 1900 and then refluxed for 2. 5 hours. The cooled reaction mixture was extracted three times with ether and the combined extracts washed twice with water. -After drying over sodium sulfate and removal of the solvent, 1.0 g. (77%) of the crude 1, l-dimethyl-5-methoxytetralin remained. The infra-red spectrum, (Figure 2) showed no carbonyl band. Without further purification, the crude ether was cleaved by refluxing for 2 hours with 4 m1- of hydriOdidu acid in 4 ml. of glacial acetic acid. . Upon the addition of 20 ml. of water and chilling, tan needles separated. . These were filtered and recrystallized from hexane yielding 0. 5 g- of colorless needles of VIII melting at 110-1120. A second recrystallization raised the melting point to 112. 5-113. 5°. This showed no depressionin a mixed melting point with. "B" and also had an. identical infra-red spectrum (Figure 52). 63:35 E530 5 Ti; 33:... 88-1.3885- . . H [II J 98 A V :9. mo .930on posuoumsH o3» mo acmwnoaa . 2 2 A3 numcofiocflmg 00 mm ousmwh q _ a J a no m m 4 . 97 _I "-’_ ~ I\ 98 B- Synthesis of 1, l-dimethyl-6-tetralol (XIII) ' Preparation of isopropylidenesuccinic acid The general procedure of Overberger (40) was followed. . The reaction was carried out in a 2-1. three-necked flask equipped witha mechanical stirrer, dropping funnel, and a reflux condenser fitted with a calcium chloride drying tube. Potassium t-butoxide was prepared by dissolving 44. 5 g. (1. 14 moles) of potassium metal in. 900 m1. of t-butyl alcohol. A mixture of 215 g. (1. 23 moles) of diethyl succinate and 58 g. (1.0 moles) of acetone was added over a twenty minute period to the refluxing basic solution. Reflux was continued for forty minutes. . Most of the solvent was distilled under reduced pressure. The mixture was acidified with 1:1 hydrochloric acid, and the remainder of the solvent was removed. The residue was extracted three times with ether and the combined extracts were washed twice with water. , The aqueous washes were extracted twice with ether and these extracts were added to the original ether solution. The ether solution was then extracted twice with saturated sodiumbicarbonate solution. Upon acidification of the combined basic extracts with 1:1 hydrochloric acid a pinkish oil separated. The aqueous layer was extracted twice with ether, these extracts combined with the oil, and dried over sodium sulfate. The solvent was removed on a rotary evaporator leaving the viscous half- ester (165 g.) which had a neutralization equivalent of 192 (theory, 186). The crude half-ester was hydrolyzed by refluxing for two hours with 200 ml. of 20% potassium hydroxide. The mixture was distilled until the temperature reached 960,. the residue was cooled and acidified with 1:1 hydrochloric. acid. The white solid was filtered and dried, yielding 113 g. of iksopropylidenesuccinic acid melting at 158-1590. Ether extraction of the filtrate yielded an additional 20 g. (84. 3%). . Overberger . (40) gives a melting point of 161.5-1620. 99 Preparation of terebic acid (B-carboxy-‘qmmethyl-'y-valerolactone) The method of Linstead (20) was used. . A mixture of 133 g. (0. 843 moles) of isopropylidenesuccinic acid and 250 m1. of concentrated hydrochloric acid was refluxed for fourteen hours. No solid remained and the solution was red-brown in color. Water (250 ml.) was added and the mixture was cooled to 00. Off-white crystals formed and were filtered. The filtrate was concentrated to 1/3 of its original volume, cooled, and made almost neutral with sodium bicarbonate. . More solid formed which was filtered. The total yield was 95. 5 g- (72%). Preparation of 4-methyl-3-pentenoic acid The procedure of Linstead (20) was followed. Terebic acid (95. 5 g. , 0.605 moles) in ca. 20 g. portions was distilled (no column) witha free flame until the pot temperature reached 2409.: . The distillate was dissolved in saturated sodium bicarbonate solution and extracted once with ether to remove any lactone. . The aqueous layer was made acidic with 1:1 hydrochloric acid and extracted four times with ether. The combined ether extracts were dried over sodium sulfate and the solvent was removed on a rotary evaporator. . Distillation of the residual oil through. a spiral column yielded 24. 7 g. (36%) of a colorless liquid, boiling at 87.-91° (2 mm. ). . Linstead (20) reported a boiling point of 99° at 10 mm. In another run, the yield was 20% and much 4, 4-dimethyl-7-butyrol- actone was formed (70-720/3.6 mm. , 117-3 1.4314-1.4316). Preparation of 4-methyl-4—(p-methoxyphenyl)pentanoic acid The reaction was carried out in a 300-ml. three-necked flask equipped with a mechanical stirrer, dropping funnel, and. a reflux con- denser fitted with a calcium chloride drying tube. A solution of 5. 0 g. (0. 038 moles) of aluminum chloride in 95 m1. of anisole was cooled in an 100 ice bath and 10 g. (0.088 moles) of 4-methyl-3-pentenoic acid was added dropwise with stirring. Another 5.0 g. of aluminum chloride was added and the dark solution was stirred for four hours at 00 and then 5. 5 hours at room temperature. Hydrochloric acid (1:1) was added until all of the precipitated hydroxide dissolved. , The organic layer was washed with dilute acid, water, and sodium hydroxide. The basic extract was washed once with ether, acidified with hydrochloric acid, and extracted twice with ether. The combined ether extracts of the acidic solution were dried over calcium chloride and the solvent removed on a rotary evaporator. This yielded 15.4 g. of crude semi-solid acid which upon recrystallization from petroleum ether gave 12.0 g. (72%) of the desired acid melting at 65-660. The literature value (21) is 66. 5-670. - A neutralization equivalent gave an equivalent weight of 220' (theory, 222). Preparation of 4, 4-dimethyl-7-methoxy- l -tetralone The general cyclization method in Organic Reactions (41) employing stannic chloride was used. The reaction was carried out in a 300-ml. three-necked flask equipped with a mechanical stirrer, dropping funnel, and a reflux condenser fitted with a calcium chloride drying tube. A solu- tion of 5.0 g. (0. 0225 mole. ) of 4-methyl-4-(p-methoxyphenyl)pentanoic acid in 20 ml. of benzene was treated with 5. 5 g. (0. 025 mole) of phOSphorus pentachloride in small portions with ice bath cooling. The mixture was stirred forty-five minutes at room temperature, refluxed for five minutes, and then cooled to 00 with an ice bath. . A solution of 6 ml. of stannic chloride in 8 ml. of benzene was added rapidly and the reaction mixture immediately became maroon colored. . The mixture was stirred one hour at 00 and 15 g. of ice was added, followed by 25 ml. of concentrated hydrochloric acid. The organic layer was separated and washed successively with 20% hydrochloric acid, water, saturated sodium bicarbonate (twice), and again with water. The solution was dried over "Drierite" and the 101 solvent removed on a rotary vacuum evaporator. . The crude ketone (3.7 g., 83%) was a yellow oil. Distillation- through a small Vigreux column gave a pale yellow liquid boiling 107-1080 at o. 1 mm. (n3 1. 5565). The infra-red spectrum (Figure 7) showed an aromatic carbonyl (5. 92 (l) and l, 2, 4-trisubstitution (11.45 and 12.05 11). An oxime of 4, 4-dimethyl-7-methoxy-l-tetralone was prepared by the method described in Shriner and Fuson (42). . It melted at 97-98° after two recrystallizations from hexane. 4152131. Calcd. for C13H17NOZ: C, 71.20; H, 7.81; N, 6.39. Found: C, 70.96; H, 7.80; N, 6.33 Preparation of l, l-dimethyl-6-tetralol (XIII) A mixture of 2 g. (0.01 mole) of l, l-dimethyl-6-methoxy-2- tetralone, 2. 5 ml. of 85% hydrazine hydrate, 3. 5 g. of potassium hydroxide, and 25 ml. of ethylene glycol was refluxed for two hours. The condenser was removed and the reaction mixture distilled until the pot temperature reached 1900. Reflux was then continued for three hours. After cooling, the mixture was extracted twice with ether, the combined ether extracts washed three times with water and then dried over calcium chloride. Removal of the solvent yielded 1.6 g. (80%) of yellow oil with no carbonyl band in the infra-red. . This crude l, 1-dimethyl-6-methoxy- tetralin was not purified further, but was cleaved to the phenol. . The crude ether (1.0 g. , 0.0052 mole) was refluxed two hours with 4 ml. of hydriOdi‘d acid and 4 m1. of glacial acetic acid. Upon cooling, 10 m1. of water was added and the mixture was extracted twice with toluene. The combined extracts were washed with 15% sodium hydroxide. The combined sodium hydroxide extracts were acidified with hydrochloric acid and the crude tetralol was extracted into ether. .After drying over calcium chloride, the ether was removed on a- steam bath. The residue (0. 6 g.) crystallized upon standing. One recrystallization from petroleum 102 ether and two from hexane gave colorless needles of XIII melting at ' 93-940. The infra-red spectrum (Figure 8) showed bands at 2.80 1‘ (hydroxyl), 7.22 and 7. 32 p. (gem-dimethyl group), 11.45 and 12.52 (.1 (1, 2, 4-trisubstitution). . The ultraviolet spectrum(Figure 9) had bands at 286 mu ( e = 2140), 279.5 mu( 6: 2010), 278 mu (' e: 2020) and 272mg (6 = 1490). firil. Calcd. for C12H1602 C, 81.77; H, 9.15. Found: C, 81.72; ‘H. 9.22. A phenylurethane of 1, 1-dimethy1-6-tetralol was prepared by the usual procedure (42a) and melted at 136-1370 after one recrystallization from hexane. A231. Calcd. for ClgHZINOZ: C, 77.25; H, 7.17; N, 4.74. Found: C, 76.93; H, 7.16; N, 4.98. C. Synthesis of 1, l-dimethyl-7-tetralol (XIV) Preparation of 2, 7-dimethoxynaphthalene The reaction was carried out in a 500-ml. three-necked flask equipped with a mechanical stirrer, addition funnel, and a reflux condenser. . The solution of 40. 0 g. (0. 250 mole) of 2, 7-dihydroxynaphthalene (Aldrich Chemical Co.) in 260 ml.. of water containing 20. 8 g. (0. 520 mole) of sodium hydroxide was cooled in an ice bath and 38 ml.. of dimethyl sulfate was added slowly (stirring). . The mixture was stirred and heated on a steam bath for 60 minutes. . After cooling in ice, the solid product was filtered, washed twice with 10% sodium hydroxide, once with water, and dried under vacuum. The yield was 45. 3 g- (97. 5%). A small portion was recrystallized from alcohol and melted at 136-70. The literature value is 138° (43). 103 Preparation of 7-methoxy-2-tetralone Sodium metal (21 g.) was added in small pieces (over a 35-minute period) to a refluxing solution of 22. 7 g. (0. 121 mole) of 2, 7-dimethoxy- naphthalene in 260 ml. of absolute alcohol. The mixture was stirred and refluxed until no more sodium remained (45 minutes). Water (200 ml) was added followed by 110 m1. of hydrochloric acid as quickly as possible. The mixture was thenstirred and heated on a steam bath for 15 minutes to cleave the enol ether. Water (300 ml.) was added and the solution was extracted three times with ether. The combined extracts were washed with water and dried over sodium sulfate. The solvent was removed on a rotary vacuum evaporator, and the residual oil distilled. The fraction boiling at 105-70 (0. 2 mm.) amounted to 13. 9 g. (65%); it solidified and melted at 30. 5-31. 70 after recrystallization from hexane-ether. - An oxime was prepared and melted out 127. 5-128o after one recrystallization from alcohol. The ketone is reported to melt at room temperature, boil. at 123-125° (0.4 mm. ), and a melting point of 126-127. 5° is given for the oxime (22). Preparation of 1, 1-dimethyl-7-methoxy-2-tetralone The reaction was carried out in a 500-ml. three-necked flask equipped with a mechanical stirrer, addition funnel, and a reflux . condenser. -A solution of sodium isopropoxide was prepared by dissolving 5. 1 g. (0. 22 mole) of sodium metal in 100 ml. of dry isopropyl alcohol ar reflux temperature, 18.4 g. (0.104 mole) of 7-methoxy-2-tetralone in 30 m1. of the same solvent was added to the cooled basic solution (stirring). . Methyl iodide (45 g. , 0. 32 mole) was added slowly to the orange solution. The solution became colorless and precipitate formed (sodium iodide). The solution was refluxed two hours, cooled, and 5 ml. of sulfuric acid in 300 m1. of water was added. . The organic layer was 104 separated and the aqueous phase extracted once with ether. . The combined organic layers were washed with water and dried over sodium sulfate. The solvent was removed on a rotary evaporator and the residual oil dis- tilled through a vacuum-jacketed tantalum spiral column. . This ylelded 16 g. (76%) of a colorless liquid boiling at 99-101.50 (0.4 mm.) which solidified upon standing. . Recrystallization from dilute methanol and then petroleum ether gave colorless needles of 1, 1-dimethyl-7-methoxy-2- tetralone melting at 53. 5-54. 30. The infra-red spectrum(Figure 10) showed bands for a gem-dimethyl group (7. 22, 7. 33 (.1). -Anal. Calcd. for C13H1602: C, 76.44; H, 7.90; methoxyl, 15.18. Found: C, 76.32; H, 7.73; methoxyl, 15.04. 1, 1~Dimethyl-7-methoxy-2-tetralone (1. 0 g.), 1.0 g. of hydroxyl- amine hydrochloride, 4 g. of potassium hydroxide, and 25 m1. of ethanol were refluxed for two hours. The reaction mixture was poured; into 150 ml. of water and the colorless needles filtered off. One recrystallization from dilute methanol and two from ligroin raised the melting point of the oxime from, 137-.142>°‘,to 148- 149°. Anal- Calcd. for C13H17NOZ: C, 71.20; H, 7.81; N, 6.39. Found: C, 71.34; H, 7.80; N, 6.45. . Preparation of 1, l-dimethyl-7-tetralol (XIV) A mixture of 4.4 g. (0.0216 mole) of 1, 1-dimethyl—7-methoxy-2- tetralone, 5. 5 m1. of 85% hydrazine, 7. 5 g. of potassium hydroxide, and 55 ml. of ethylene glycol was refluxed for thirty minutes. Water and hydrazine were distilled off until the pot temperature reached 1860 and refluxing was continued for two hours. , After cooling, 60 ml. of water was added and the solution extracted twice with, ether. . The combined ether extracts were washed with water and dried over sodium sulfate. The solvent was removed on a rotary evaporator. . This yielded 3.0 g. (73%) of crude l, l-dimethyl-7-methoxytetralin. The infra-red spectrum 105 showed no carbonyl present. . The crude methyl ether (2. 0 g. , 0.0105 mole) was cleaved by refluxing with 10 ml.. of 47% hydriodie acid and 25 m1. of glacial acetic acid for eight hours. Ice (20 g.) was added to the cooled reaction mixture and crystals separated after stirring for several minutes. The addition of 50 m1. of water yieldedimore product. The solid was filtered, dried in air, and recrystallized from ligroin using Norite. This yielded 1. 5 g. (81%) of colorless needles melting at 105-106°. . The infra-red spectrum showed a hydroxyl band at 2. 80 p (Figure 11). 7 The ultraviolet spectrum (Figure 12) showed bands at 287.5 mp. (e = 2140), 279.5 mp. ( 6 = 2270), and 272.5 mp ( 6 =1590). v-A—n__a.1., Calcd. for C12H160: C, 81.77; H, 9.15. Found: C, 81.63; H, 9.11. XIV (0. 10 g.) was dissolved in 0. 5 ml. of 33% potassium hydroxide solution. and 2 ml. of water. . Chloroacetic acid (0.15 g.) was added, the mixture shaken, and then heated one hour on a steam bath. Water (5 ml.) was added and the solution was acidified with dilute hydrochloric acid. The organic material was extracted into 8 ml. of ether and this washed once with 4 ml. of water. The ether layer was extracted with two 8-m1. portions of 5% potassium carbonate solution. Acidification of the basic extracts yielded the crude derivative as a white solid. This was filtered and washed with 2 ml. of water. - After recrystallization from dilute ethanol the material melted at 180- 182°. . The phenoxyacetic acid derivative of "C" was prepared in an identical manner and melted at ‘179-179. 5°(softened at 170°) after three recrystallizations from dilute ethanol. -_A_n§;l_. Calcd. for C(4H1803: C, 71.77; H, 7.74. Found: C, 71.62; H, 7.76. 106 D- Mixture of l, l-dimethyl-S-and 7-tetralols Separation of "C" Compound "C" was examined by gas chromatography on 10% silicone at 181°. Two components (retention times of 50 and 56 minutes) were found. The two peaks overlapped considerably. The first com- ponent had a retention tilne identical to that of 1, l-dimethyl-S-tetralol ("B"), and the second component had a retention time which was the same as the 7-isomer. However, the 6-isomer had a very similar retention time (57 minutes). The second component was separated by this method, and its infra-red spectrum (Figure 54) showed that it was definitely the 7-isomer (although it could not be completely freed of the 5-isomer). Preparation of synthetic "C" A mixture of 50 mg. of 1, 1-dirnethyl-5-tetralol ("B") and 50 mg. of 1, 1-dimethyl-7-tetralol was recrystallized from 2 m1. of hexane. The crystalline product (ca. 90 mg.) melted at 83-840, and this showed no depression upon admixture with authentic "C. " Their infra-red spectra were identical (Figure 53). Reaction of 2-methyl-5-chloro-2-pentene with phenol A mixture of 15. 7 g. (0. 133 mole) of Z-methyl-S-chloro-2-pentene and 64 g. (0. 68 mole) of phenol was heated at 1500 for 14 hours and hydrogen chloride was slowly evolved. - After cooling, any residual hydrogen chloride was removed by pumping on the dark reaction mixture (water aspirator) for several minutes. Gas chromatography of the crude material (10% silicone, 181°) showed three main products (in addition to unreacted phenol), and two minor ones. The two minor components were apparently very high boiling polyalkylation products, judging by their long 107 was umtfffioaflfltfi J «0 935832 a pan Alv :0: m0 .930on @9753de 0:» mo GOmComEoU Ma .3385 sohsao 8.. 7:; Eosnsoooé A..: numsoaourm? : o h—o .3 ensure _ _ _-: :3 Q‘- _.--- --?- ---- c--\-——I-- . fl -“ ’l (\ T. o‘-'=. I) ‘J’ -8 I ‘s ’ \..,. ‘ \‘ (-‘\ ’ t\ ‘ \ ’--- J ook; o\\\l h—-- ‘ ~---“— 3 108 retention times (98, 137 minutes). Two of the other three components had retention times identical to 5, S-dimethylhomochroman (15 minutes) - and 1, 1-dimethy145-tetralol (50 minutes). The third (56 minutes), which was not completely resolved from the other tetralol (and was present in about the same amount) had a retention time identical to that of 1, l-dimethyl-7-tetralol (56 minutes) and was quite close to that of the 6-isomer (57 minutes). The reaction mixture was dissolved in 200 m1. of methylene chloride and extracted once with 250 ml.. of 20% potassium hydroxide, followed by three extractions with Claisen's solution. , The combined basic extracts were acidified with 1:1 hydro- chloric acid and the organic material taken up in methylene chloride. After drying over anhydrous sodium sulfate the solvent was removed and the unreacted phenol distilled through a Vigreux column under reduced pressure. The residual dark oil was distilled through a small Vigreux column and two fractions were collected. The first amounted to 5. 0 g. (85-950, 0.08 mm.) and solidified. Gas chromatography (same conditions as above) showed this to be mainly 1, l-dimethyl-S-tetralol (50 minutes). The second fraction (95-1150, 0.08 mm.) amounted to 4. 0 g. and partially solidified. Attempts to recrystallize it from hexane failed. Gas chromatography was used to partially purify this fraction, and although the main component (56 minutes) could not be completely separated from the 5-isomer present, it was definitely 1, 1-dimethy1-7- tetralol by virtue of its infra-red spectrum (Figure 54). E. Attempted Synthesis of Z-methyl-S-chloro-2-pentene-5-dz (XX) Preparation of 2, 2-dimethyltetrahydrofuran-5-dz The reaction was carried out in a 300-ml. three-necked flask equipped with a dropping funnel, mechanical stirrer, and a reflux condenser fitted with a calcium chloride drying tube. A solution of 109 L k --_ - ‘- - ‘QOhlooa-cgg «Io-c...“ ’ ’ “---.r-——-—- I a \ ~ z. I. \‘Is..-- -‘~—‘-’ \~ " '3‘ :T:E..— - — c . .. . I. L. .--‘..b-.-- —-— -0. O D 6‘;.=‘- L. \-_- -- - .7‘.‘:-.=_-.-_—— .- . — -00-.----- - fl -Qi- “’ c.-_. 1:; .d. ‘m “to -& was?" "'" u‘ .7 :5 J' (’ I. , | . \ l ' .l" l , l , I . I 0 I l I I ‘ o O 0 1 4 I ' \ o \ I \ ..".‘\ ' o . ."".-8£ C *5- — _-— ‘----*—-----ag= ~:.0.:: . —=--_ -- ' ' ' ' ~53..- ."'J_; 13 11 6'7 Wavelength (p) ) and Those of the Component of "C" (--) and of the Alkylation Product (- ° ° -) in Carbon Disulfide. Comparison of the Infra-red Spectra of 1, l-Dimethyl-7-tetralol ( Figure 54. 110 9. 2 g. (0. 081 mole) of 4, 4-dimethyl-7-butyrolactone (obtained in the preparation of 4-methyl-3-pentenoic acid) in 25 m1. of ether was added dropwise to a well stirred slurry of 1. 9 g. (0. 045 mole) of lithium aluminum deuteride in 50ml. of ether. . Much. solid material was formed. The mixture was stirred thirty minutes at room temperature and then refluxed for thirty minutes. Water (2 ml.) was added cautiously, followed by 45 ml. of 4 N hydrocthric acid. . The mixture was saturated with calcium chloride and. stirred two hours.. The aqueous phase was separated and extracted with five 25-m1. portions of ether. After standing overnight the aqueous phase was extracted with three more portions of ether. The combined organic material was dried over calcium chloride and the solvent removed through a Vigreux column (to 500). . The residue was distilled with 0. 5 ml. of 20% sulfuric acid through a short Vigreux column (so-95°). This yielded 8. 5 g. (ca. 90%) of pale yellow liquid. Gas chromatography on 20% silicone at 1150 showed only the dimethyl- tetrahydrofuran (appearance time, 10 min.) with a small amount of ether present. . The infra-red spectrum (Figure 13) showed. carbon-deuterium bonds at 4. 55 p. and 4. 79 p. The NMR spectrum is shown in Figure 13. Reaction of 2, 2-dimethyltetrahydrofuran-S-deithiLucas' reigent 2, 2-Dimethyltetrahydrofuran-S-d,_ (6.0 g. ,7 0. 059 mole) and 18 ml. of Lucas' reagent were heated for ninety minutes on a steam bath. , The solution became dark and at secondtphase slowly formed. . The upper layer (4. 5 g.) was separated and distilled without the use of a column. . This yielded 2. 5 g. of pale yellow liquid boiling at 80-1450. Vapor phase chromatography on 20% silicone at 1150 showed it to be mainly Z—methyl- 5-chloro-2-pentene (appearance time, 24 min.) with about 20% of a slightly lower boiling material (probably Z—methyl-S-chloro-1-pentene). The product was purified by gas chromatography. The infra-red Spectrum 111 (Figure 15) Showed some terminal methylene (6.07 p, 11. 22 p.) present. This isomer could not be completely separated by gas chromatography on 20% silicone. The NMR spectrum is shown in'Fi'gure 16. F. Synthesis of 2-methyl-5-chloro-2-pentene-5-dz (XX) All reactions were run first with unlabelled compounds. Preparation of ethyl 3-bromopropionate The procedure in "Organic Syntheses ‘ (44) was followed. 7 The re- action was carried out in a 500-ml. three-necked flask equipped with a gas inlet tube and reflux condenser fitted with a calciumchloride drying tube. Anhydrous hydrogen bromide (53 g. , 0. 65 mole) was passed into a solution of 59. 5 g. (0. 595 mole) of freshly distilled ethyl acrylate and 0. 5 g. of hydroquinone in 110 ml. of ether. . The mixture was stirred magnetically and cooled in ice during the addition. - After standing at room temperature for twenty hours, the solvent was distilled until the pot temperature reached 850. The residual liquid was distilled througha Vigreux column giving 102 g. (94. 5%) of the colorless bromoester boiling at 65. 5-660, 8-9 mm.; n3 1.4514 for four fractions taken. . Preparation of 3-bromopropanol-1-dz (XXVII) The procedure of Nystrom (24) was followed. 7 The reaction was carried out in a 1-1. three-necked flask equipped witha reflux. condenser, mechanical stirrer, dropping funnel, and calcium chloride drying tube. A solution of 9. 8 g. (0. 074 mole) of aluminum chloride in 100 ml.. of ether was added rapidly to a stirred Slurry of 3. 1 g. (0.074 mole) of lithium aluminum deuteride in 100 ml. of ether while cooling in an. ice bath. The mixture was stirred five minutes and then cooled in a Dry Ice- acetone bath. -A solution of 23 g. (0. 127 mole) of ethyl 3-bromopropionate 112 in 100 m1. of ether was added over thirty-five minutes. ~ After stirring thirty-five minutes, the bath was removed and the mixture allowed to warm to room temperature. . Stirring was continued one hour and a small sample was removed and hydrolyzed- The infra-red spectrum showed some carbonyl present, so 0. 2 g. of lithium aluminum deuteride in 40 ml.. of ether was added to the reaction mixture and stirring was continued thirty minutes. A small sample showed no carbonyl present, and the reaction was then hydrolyzed by cautiously adding 60 m1. of water followed by 55 ml. of 6 N sulfuric acid. . The aqueous phase was separated and extracted with four 30-ml. portions of ether. The combined organic material was dried over anhydrous magnesium sulfate- The sol- vent was removed through a Vigreux column and the residue distilled (no column). . This yielded 14. 8 g. (78.4%) of the bromoalcohol (XXVII) boiling at 71.. 5-740, 9. 5-10 mm. , 113‘ 1.4840. . Nystrom (24) reported 70-720 at 10 mm. for the boiling point of the unlabelled alcohol. The infra-red spectrum showed carbon-deuterium bonds at 4. 52 and 4. 74 p (Figure 17). The NMR Spectrum is shown in Figure 18. Preparation of 1-bromo-3-chloropropane-3-dz (XXVIII) The reaction was carried out in a 50-ml. three-necked flask equipped with. a dropping funnel and reflux cofndenser fitted with a calcium chloride drying tube.. Thionyl chloride (11:41.2 ml. , 0. 20 mole) was added over forty- five minutes to 14.:6 g. (0.0969 mole) of 3-bromo-1-propanol-1-dz (containing five drops of pyridine) at ice bath temperature (magnetic stirring). The mixture was stirred at room temperature for eight hours, then refluxed for two hours. The infra-red spectrum of a sample after 1. 5 hours of reflux showed only a slight amount of hydroxyl present. The reaction mixture was poured Slowly on ice (ca. 30 g.) and allowed to stand ten minutes with occasional swirling. . The organic phase was separated and washed twice with water. . The aqueous phase was extracted 113 once with pentane (10 ml.) and the combined organic phases dried over anhydrous sodium sulfate. . The solvent was removed through a Vigreux column and the residue distilled- This yielded 13.6 g. (83.5%) of color- less liquid boiling at 51.-55O (31-32 mm. ). Gas chromatography on 20% silicone at 1300 showed it to be about 85% of the desired product (appearance time, 30 min. ), the remainder being a somewhat lower boil- ing component (probably the dichloride). . The infra-red and NMR Spectra are shown inFigures l9 and 20. . Preparation of 4-chlorobutyronitrile-4-dz (XXIX) The reaction was carried out in a 300-ml. three-necked flask equipped with a mechanical stirrer, dropping funnel, and reflux condenser. . The procedureE used was essentially that inO rganic Syntheses (45). .. 3 A solution of 5. 2 g. (0.030 mole) of potassium cyanide in 9 ml. of water was added over thirty minutes to a stirred, refluxing solution of 12.8 g! (0.0758 mole) of 1- bromo- 3- chloropropane- 3- d; in 26 ml. of ethanol. The mixture was stirred and refluxed an additional ninety minutes. After cooling in. ice, 40 ml.. of water were added, followed by 10 ml- of chloroform. . The mixture was stirred five minutes and the organic phase separated. The aqueous layer was extracted withr6 ml.. of chloroform and then the combined organic layers were washed successively with 8 ml. of calcium chloride solution (one part saturated calcium chloride and one part water) and 5 m1. of water. The material was dried over calcium chloride and the solvent removed through a Vigreux. column. . The residual oil was distilled through a small Vigreux column giving 4.4 g. (55. 5% yield) of colorless nitrile (XXIX) boiling at 72-750 (12’mm. , n3 1,4433). The pot residue amounted to 0. 5 g. Bruylants (46) reported a boiling point of 75-77° at 13 mm. and n3 1.4446 for the unlabelled nitrile. The infra- red Spectrum (Figure 21) showed a nitrile group present (4.45 p). The NMR spectrum is Shown in Figure 22. 114 ' Preparation of ethyl 4-chlorobutyrate-4-dz (XXX) The procedure used was that of Fehnel (47). The reaction was carried out in a 25-ml. flask equipped with a reflux condenser and calcium chloride drying tube. A solution of 4. 3 g. (0. 041 mole) of 4-chlorobutyronitrile-4—dz in 11 ml. of absolute ethanol was cooled in an ice bath and saturated with dry hydrogen chloride. The mixture was allowed to stand two hours at room temperature, and then refluxed for two hours. After; several minutes of refluxing, a white solid began to appear. The contents of the flask were almost all solid after thirty to forty minutes. . After cooling in ice, 25 ml. of water was added and the lower (organic) layer was separated.. The aqueous phase was extracted with four 8-ml. portions of ether and the combined organic phases were washed, successively with 5 m1. of water, 5 ml. of 5% sodium bicarbonate, and 3 ml.. of water. The solvent was removed (Vigreux column)::after drying over anhydrous magnesium sulfate and the residue distilled through a Vigreux column. . This yielded 5.2 g.’ (83. 5%) of colorless ester boiling at 69-700/9‘mm. , 1112; 1.4320. . Fehnel (47) reported a boiling point of 81-84°/2o mm., and Blomquist (48) gave 72.5—74°/9.5-lo mm., n3 1.4285, for the unlabelled ester. The infra-red spectrum (Figure 23) showed an ester carbonyl at 5. 76 p. The NMR Spectrumis shown in .Figure 24. Preparation of 2-methyl-5-chloro-2-pentanol-5-d2 (XXXI) The reaction was carried out in a 300-ml. three-necked flask equipped with a dropping funnel, mechanical stirrer, and reflux con- denser fitted with a calcium chloride drying tube. A solution of 5. 0 g. (0. 033 mole) of ethyl 4-chlorobutyrate-4-dz in 30 ml. of ether was added Slowly to a stirred solution of methylmagnesium iodide (from 2. 1 g. , 0. 086 mole of magnesium and 13 g. of methyl iodide) in 50 ml. of ether 115 at ice bath temperature. The addition required thirty minutes. a After an additional thirty minutes at this temperature, the mixture was stirred two hours at room temperature, then hydrolyzed by the addition of 55 ml. of saturated ammonium chloride. . The aqueous phase was separated. and extracted with 20 ml. of ether. The combined ether layers were dried over anhydrous magnesium sulfate. The solvent was removed by distillation until the pot temperature reached 500. The product remained as a yellow oil (5.4 g.) and still contained some solvent. The infra-red Spectrum of this crude alcohol is shown in Figure 25. The unlabelled material had been prepared previously by Henry (49) and Campbell (50). Preparation of 2-methyl-5-chloro-2-pentene-5-dz (XX) The crude Z-methyl-5-chloro-2-pentanol-5-dz (4.0 g.) was dis- tilled from 0. 5 g. of freshly fused potassium bisulfate without the use of a column (50-1400) to give 2. 7 g. of a pale yellow liquid consisting of some ether, 2,2-dimethyltetrahydrofuran, Z-methyl-S-chloro-l-pentene- 5-d2, and mainly the desired Z-methyl-S-chloro-2-pentene-5-dz, as shown by gas chromatography(retention time of 25 minutes at 1150 on 20% silicone). The product was purified by this method and amounted to 0.8 g. (22%). It still contained some of the 1-isomer (13%) as evidenced by the infra-red (Figure 27) and NMR (Figure 26) Spectra. Reaction of 2-methyl-5-chloro-2-pentene-5-dfizf with phenol Phenol (4.0 g.) and Z-methyl-S-chloro-2-pentene-5-dz (0. 50 g.) were heated at 1500 for 8. 5 hours, and hydrogen chloride was evolved. The dark mixture was cooled and taken up in 15 ml. of methylene chloride, extracted successively with 10 ml. of 20_ potassium hydroxide, two 4-m1. portions of Claisen's solution (5.1), washed once with water, and dried over anhydrous sodium sulfate. The solvent was removed on a steam bath, and the residual oil was purified by gas chromatography (at 1750 on 116 20% silicone) to give 110 mg. (15% yield) of deuterated 5, 5-dimethyl- homochromanas colorless plates. The basic extracts were acidified with hydrochloric acid and the phenolic material was taken up in methylene chloride. This solution was washed once withwater and then distilled through a small Vigreux column at 20 mm. to remove most of the unreacted phenol (to 85°). . The residue was purified by gas chromatography on 20% silicone at 1810 to give a mixture of deuterated 1, l-dimethyl-S- and 7-tetralols (30-50 mg. ). The mixture was mainly the 5-isomer since the 7-isomer came off of the column second and was not completely trapped. This was due to the very high boiling points of the two products and the difficulty of main- taining the outlet line of the gas chromatograph at a sufficiently high temperature. The low yield was also due to this. The NMR (Figure 30) and infra-red (Figure 55) Spectra were taken on the material washed out of the trap with carbon tetrachloride. .ovwnozoouuofi 2.83.8.0 cm muosposm oSocoanH pououooBoQ on» mo 9530on poutouwcH .‘ C: sewage/84$ h 4 .mm enema 117 mJ 118 Part II A. Synthesis and attempted resolution of 5, 5-dimethylhomochroman- 7-carboxy1ic acid (XXXIII) Preparation of 5, 5-dimethylhomochroman (III) The procedure of Hart (7) was followed. The reaction was carried out in a 500-ml- three-necked flask equipped with a gas inlet tube, reflux condenser, and thermometer. The flask was charged with 150 g. (1.6 moles) of phenol and 32 g. (0. 32 mole) of 2, Z-dimethyltetrahydro- furan and the mixture heated to reflux (152°). . Dry hydrogen chloride was then passed in. slowly. A "crackling" noise began and within 2 hours the reflux temperature had dropped to 110°. . Reflux was continued for a total of 7 hours, and the reaction mixture was cooled. Benzene (250 ml.) was added and the solution extracted with two 250-ml. portions of 20% potassium hydroxide. The organic layer was then extracted with two 150-ml. portions of Claisen's solution (51). The benzene layer was dried over anhydrous potassium carbonate and the solvent removed. The dark residue was distilled using no column and the portion boiling between 90 and 1600 at 45 mm. was redistilled through a small packed column. The colorless liquid boiling at 77--79o (0. 7 mm.) solidified upon standing and yielded 8.4 g. (15%) of colorless 5, 5-dimethylhomochroman melting at 47-480 after recrystallization from hexane. Preparation of 5, 5-dirnethyl-7-acetylhomochroman '(XXXII) The reaction was carried out in a 1-1. three-necked (flask equipped with a mechanical stirrer, addition funnel, and a reflux condenser. In a typical run, the flask was charged with 400 ml. of methylene chloride, 11.6 g. (0.085 mole) of aluminum chloride, and 9.0 g. (0.115 mole) of freshly distilled acetyl chloride. The mixture was stirred in an. ice bath 119 until all of the aluminum chloride had dissolved. A solution of 12. 0 g. (0. 682 mole) of 5, 5-dimethylhomochroman in 100 ml. of methylene chloride was added dropwise with-cooling and stirring. The stirring was continued for 24 hours at room temperature and the dark mixture hydrolyzed with 120 ml. of 1:1 hydrochloric acid. The organic layer was washed successively with dilute hydrochloric acid, water, potassium carbonate solution, water again, and dried over anhydrous sodium sulfate. The solvent was removed through a Vigreux column. . The residual crude ketone weighed 13.4 g. (90%). .It was usually not purified further; however, in one case it was distilled through a small Vigreux column. . This gave 5, 5-dimethyl-7-acetylhomochroman as a very viscous liquid boiling at 110-1120 (0. 09 mm.) with n3 1. 5497. It could not be induced to crystallize. The infra-red spectrum (Figure 32) exhibited bands at 5. 95 p (aromatic carbonyl), 8.0 p (aromatic ether), 10. 98 and 11. 92 pt (1, 2,4-trisubsti- tution). The 2, 4-dinitrophenylhydrazone of 5, 5-di1nethyl-7-acetylhomo- chroman was obtained in almost quantitative yield using the sulfuric acid method (42b). After two recrystallizations from chloroform the red- orange needles melted at 199. 5-200.5". Anal. Calcd. for Con22N4Os= C, 60.29; H, 5.57; N, 14.06. Found: C, 60.21; H, 5.68; N, 13.92. . Preparation of 5, 5-dimethy1homochroman-7-carboxylic acid (XXXIII) The procedure for the haloform reaction using a dioxane-water solvent system (52) was employed. . The reaction was carried-out in a 1-1. three-necked flask equipped witha mechanical stirrer, addition funnel, and reflux condenser. A thermometer suspended through the condenser dipped into the reaction mixture. Sodium hypobromite was prepared 33231 by Slowly adding 13. 3 ml. (0. 244 mole) of bromine to 37 g. (0. 92 moles) of sodium hydroxide in 180 ml. of water with cooling. 120 This solution was then warmed to 400 by a warm water bath. and a solu- tion of 14. 0 g. (0.0643 mole) of crude, 5, 5-dimethyl-7-acetylhomo- chroman in 140 ml. of dioxane was added over a 60-minute period with stirring. The heat of the reaction was sufficient to maintain the temperature near 400. . The mixture was stirred an additional 3. 5 hours and a second. dark phase slowly separated (bromoform). Water (300 ml.) was added. and the dark lower layer was discarded., The remaining aqueous phase was extracted with 100 ml. of hexane. Acidification of the water layer with hydrochloric acid and chilling yielded an off-white solid. This was air-dried and recrystallized from benzene to give the acidas colorless prisms of XXXIII melting at 174-1750. . The yield was 9. 5 g. (67%). In similar runs the yield was about 65%. The product had a neutralization. equivalent of 221 (theory, 220). The infra-red spectrum (Figure 33) showed a band at 5. 93 p. (aromatic carboxylic acid) and the 5-6 p. region indicated 1, 2, 4-trisubstitution (5. 26,, 5. 55, and 5. 81 u). The ultraviolet spectrum (Figure 34) showed a maximum at 253. 5 mp ( e = 10, 900). Anal.. Calcd. for C13H16O3: C, 70.88; H, 7.32. Found: C, 70.77; H, 7.33. . Preparation of 5, 5-dimethylhomochroman-7-carboxamide A solution of 2.0g. (0.0091 mole) of 5, 5-dimethylhomochroman- 7-carboxylic acid in 7. 0 m1. of thionyl chloride was refluxed for 20 minutes, then poured into 80 m1. of concentrated ammonium hydroxide and ice. The semi-solid that formed was extracted into ether and dried over anhydrous sodium sulfate. ~ Most of the ether was removed by distillation and hexane was added. The flocculent white amide that separated was filtered and dried in air. . The yield was 1. 9 g. (96%) and the product melted at 153-1540. »- Anal. Calcd. for C13H17NOZ: C, 71.20;. H, 7.81; N, 6.39. Found: C, 71.18; H, 7.86; N, 6.24. 121 Preparation of the methyl 5, 5-dimethylhomochroman-7-carboxy1ate A solution of 4. 0 g. (0. 0182 mole) of 5, 5-dimethy1homochroman-7- carboxylic acid and 1 m1. of sulfuric acid in 100 ml.. of absolute methanol was refluxed for 6 hours. Most of the alcohol was removed by distilla- tion, the oily residue was taken up in ether and washed twice with 5% potassium carbonate solution, once with water, and dried over anhydrous sodium sulfate. Slow evaporation of the solvent yielded the ester as colorless plates melting at 60.7-61. 20 (3.7 g. , 87%). .1521. Calcd. for C14H1803: C, 71.77; H, 7.75. Found: C, 71.77; H, 7.77. Conversion of 5, 5-dimethylhomochroman-7-carboxamide to 5 , 5-dim ethylhomoc hr om an 5, 5-Dimethylhomochroman-7-carboxamide (0. 90 g. , 0. 0041 mole) was added to 15 ml. of freshly prepared 0. 5 N sodium hypochlorite solu- tion and 0. 5 g. of sodium hydroxide. Sufficient dioxane (ca. 15 ml.) was added to effect solution. Two phases formed. The reaction mixture was heated to 450 by a water bath and stirred magnetically for 75 minutes. The system was now dark red and consisted of only one phase. This was saturated with sodium chloride and extracted four times with ether. The combined ether extracts were then extracted with two IO-ml. portions of 10% hydrochloric acid to get the crude amine as the hydro- chloride. This solution was cooled to 00 and diazotized by the addition of 0. 2 g. of sodium nitrite in small portions. 7 The solution containing the diazonium salt was poured into 20 ml. of ice-cold 30% hypophosphorous acid and allowed to stand overnight at 0°. . The solution (containing some red oil) was extracted twice with ether. . The ether was extracted twice with 10% sodium hydroxide (to remove any phenolic material formed by hydrolysis of the diazonium compound) and dried over‘magnesium sulfate. The pale yellow solution was treated with Norite and the solvent allowed 122 to evaporate. . This yielded 0. 3 g. (42%) of 5, 5-dimethylhomochroman as pale yellow plates melting at 45-460. 7 The infra-red spectrum was identical with that of authentic material. r Attempted resolution of 5, 5-dimethy1homochroman-7-carboxylic acid A solution of 5. 83 g. of brucine and 2. 75 g. of 5, 5-dimethylhomo- chroman-7-carboxylic acid in 75 ml. of warm acetone was filtered and allowed to cool. No solid was formed, even after evaporation of half of the solvent. Evaporation of more solvent yielded a small amount of oil. . The resolution was attempted using 2. 95 g- of quinine and 2. 00 g. of the acid in 10% acetone-ethyl acetate as a solvent. A very small amount of low melting solid (30-500) was formed. . The same results were obtained using quinine in acetone. The solid material was not investigated. B. Attempted synthesis of l, l, 4, 4-tetramethylbenzocyclo— heptene (XXXIV) Preparation of 3-methyl-3-phenylbutyric acid The reaction was carried out in a 2-1. three-necked flask equipped with a mechanical stirrer. A solution of 60 g. (0.60 mole) of 3, 3-di- methylacrylic acid (Aldrich Chemical Co.) in 1200 ml. of benzene was cooled to 50 by means of an ice bath. ~Alurninum chloride (150 g. , l. 12 moles) was added in small portions with stirring and cooling so that the temperature did not rise above 100. The yellow solution was stirred for fifteen. hours at room temperature and then allowed to stand for two and a half days. The mixture was poured on ice and hydrochloric acid, the organic phase separated, and the aqueous phase extracted twice with 150-m1. portions of benzene. The combined organic layers were washed twice with water and the solvent removed by steam distillation. The residual oil was taken upin hot sodium carbonate solution. and filtered. 123 The filtrate was treated with Norite and acidified with 1:1 hydrochloric acid. - After cooling in an ice bath, the colorless crystals were filtered and dried under vacuum. The yield was 100 g. (94%), and the product melted at 52-540 without recrystallization. .Dippy and Young (34) reported a melting point of 57-580 using the same procedure. Preparation of 3 —methyl- 3 -phenyl- 1 -butanol The reaction was carried out in a 2-1. three-necked flask fitted with a reflux condenser, addition funnel, and a mechanical stirrer. -A solution of 135 g. (0. 760 mole) of 3-methyl-3-phenylbutyric acid in. 300 m1. of tetrahydrofuran was added slowly with stirring to a slurry of 38 g. (0. 975 mole) of lithium aluminum hydride in 250 m1. of tetrahydrofuran. The addition was at such a rate as to maintain gentle reflux. . The reaction mixture was stirred for one hour and then 40 ml. of water and 25 ml. of tetrahydrofuran were added cautiously. . This was followed by 40 g. of sodium hydroxide in 300 ml. of water. After standing overnight, the liquid phase was decanted and the residual white solid washed by decan- tation with ether. The organic phases were combined and most of the solvent was removed through a Vigreux column. Water (100 ml.) was addedto the residual oil and the organic layer was separated. The aqueous layer was extracted with ether and the combined organic layers dried over anhydrous sodium sulfate. Removal of the solvent and dis- tillation of the residual oil through a vacuum- jacketed spiral column gave 105 g. (85%) of 3-methyl-3-phenyl-l-butanol boiling at 83-850 (0.1 mm.) with nz°°8 1. 5228. Bogert (53) reported a boiling point of 137-133°at D 16 mm. . Julia (33) reported a boiling point of 1350 at 14 mm. Preparation of the tosylate of 3emethyl-3-phenyl- l-butanol The procedure of Julia (33) was followed. . A solution of 23 g. (0. 122 mole) of p-toluenesulfonyl chloride in 60 ml.. of pyridine was cooled 124 in an ice bath and 20 g. (0. 122 mole) of 3-methyl-3-pheny1-l-butanol was added. . The mixture remained in. the ice bath one hour and then stood at room temperature for twenty-one hours. Much solid pyridine hydrochloride was present. The reaction-mixture was poured into ice and water and this extracted three times with benzene. The combined extracts were dried over anhydrous sodium sulfate. . The solvent was removed on a rotary vacuum evaporator, keeping the temperature below 50°. The viscous residue amounted to 39 g. (quantitative yield) which had no hydroxyl band in the infra-red. The tosylate was used without further purification. Preparation of 5-methyl-5—phenjlhexanoic acid The reaction was carried out in a 1-1. ,. three-necked flask equipped with a reflux condenser, addition funnel, and mechanical stirrer. Ethyl malonate (70. 5 g. ,. 0.44 mole) in 100 ml. of dioxane was added slowly to a well stirred suspension of 10. 5 g. (0.436 mole) of sodium hydride in 200 ml. of dioxane. The mixture was refluxed for thirty minutes. A solution of 118 g. (0. 366 mole) of crude 3-methyl-3-phenyl-l-butyl tosylate in 200 ml. of dioxane was added dropwise with stirring. . The mixture was then stirred and refluxed for twenty hours. Approximately 450 ml. of water was added. . The organic layer was separated. and the aqueous phase was extracted three times with ether after acidification with hydrochloric acid. . The organic layers were combined and most of the solvent was removed at atmospheric pressure. . The crude product was saponified by heating with 56 g. (l. 0 mole) of potassium hydroxide and 200 ml. of ethylene glycol on a steam bath for four hours. Water ' (100 ml.) was added and the mixture was acidified with 1:1 hydrochloric acid followed by three extractions with ether. The combined ether extracts were washed once with water and dried over anhydrous sodium sulfate. The solvent was distilled at atmospheric pressure and the dark residue 125 was heated at 165° for two hours to complete the decarboxylation. . The residue was distilled without a column yielding 40 g. (54%) of the viscous acid boiling at 138° (0.2 mm.). . Julia (33) reported 145-1480 at 0. 2 mm. for the boiling point of the acid prepared by the same method. The infra-red spectrum is shown in. Figure 37. . Preparation of 5-methy1-5-phenylhexanoyl chloride In a typical preparation 60. 0 g. (0. 291 mole) of 5-methyl-5-phenyl- hexanoic acid and 58 g. (0.49 mole) of freshly distilled thionyl chloride were placedvin a 250-ml. flask fitted with a condenser and calcium chloride drying tube. The mixture was then refluxed for l. 5 hours. The excess thionyl chloride was distilled under reduced pressure (no column), and the yellow residue distilled at 0. 1 mm. The colorless acid chloride boiled at 96-980/mm.‘:ian-d amounted to 6‘37g. (96%) yields)’.: Julia (33) reported a boiling point of 150-1600 at 2 mm. The infra-red spectrum showed a single carbonyl band at 5. 59 p. (Figure 56). Preparation of 5, 5-dimethylbenzocycloheptene-l-one (XXXVI) The reaction was carried out in a 1-1. three-necked flask equipped with a reflux condenser, mechanical stirrer, calcium chloride drying tube, and a Hershberg (54) funnel. A In a typical run, 30.0 g. (0. 134 mole) of 5-methyl-5-phenylhexanoyl chloride in 400 ml. of carbon disulfide was added over twenty hours to a well- stirred solution of 36 g. (0. 270 mole) of aluminum chloride in 200 m1. of the same solvent. The dark solution was stirred an additional ten hours and allowed to stand overnight. . Most of the solvent was distilled and replaced with 100 ml. of ether. The reaction mixture was hydrolyzed by: adding (withcooling in ice) sufficient 1: 1 hydrochloric acid to dissolve the aluminum hydroxide formed. . The organic layer was successively washed with 25 m1. of 1:1 hydrochloric acid, 25 m1. of water, 25 ml. of 5% potassium hydroxide, 126- s: .Gmozv 03.830 Twocmxofiswsoamumtaapogtm mo 5.9.50on postman”; A3 numsofioefim? NH 2 w e .om oudmfim CI! _ q 4 127 and 25 m1. of water. The material was dried-over anhydrous potassium carbonate and the solvent removed on a rotary vacuum evaporator. The residual oil was distilled through a small Vigreux column at 0. 06 mm. and yielded 19.4 g. (78%) of the pale yellow ketone boiling at 82-850/0.07 mm. Julia (33) reported a boiling point of 94-960 at 0. 3 mm. A dinitrophenylhydrazone derivative was prepared and melted at 160-1620 after one recrystallization from ethanol (orange plates). Julia reported a melting point of 1620 (orange plates). . In one case the reaction was carried out in refluxing methylene chloride instead of carbon disulfide and the yield was only 60%. The ketone could be obtained as a colorless liquid by steam distillation of the crude reaction product; however, this was arvery slow process since the steam volatility was low. The infra-red spectrum is shown in Figure 38. . Preparation of 2, 2, 5, 5-tetramethylbenzocycloheptene-l-one (XXXVII) The reaction was carried out in a 50-ml. three-neckedflask equipped with a mechanical stirrer, dropping funnel, and a reflux condenser fitted with a calcium chloride drying tube. A solution of 10. 0 g. (0.053 mole) of 5, 5-dirnethylbenzocycloheptene-l-one (XXXVI) in 10 ml. of toluene Was added over 20 minutes to a warm, stirred suspension of 2. 26 g. (0. 058 mole) of sodium amide in 10 ml. of the same solvent. The deep red solution was refluxed for 25 minutes, and then.8 ml. of methyl iodide was added slowly. The mixture was refluxed an additional 2 hours and then cooled. After washing with two lO-ml. portions of water, it was dried over anhydrous sodiumsulfate. The solvent was removed under reduced pressure (ca. 100 mm.) and the dark residue was distilled (no column) to separate the product from any tars. This yielded 8. 7 g. of a pale yellow liquid boiling at 78-900/0. 08 mm. The alkylation reaction was repeated on this crude product using the same procedure but only 1. 7 g. of sodium amide. Distillation through a Vigreux column gave 7. 3 g. ‘128 (64%) yield) of the tetramethyl-ketone (XXXViII) boiling at 79-830/0. 075 mm., n25 1.5280-1. 5278 as a pale yellow liquid. The infra-red. and D ultraviolet spectra are shown in Figures 39‘and 40. In several other runs the yield was 50-60%. ' -Anal. Calcd. for ClstoO: C, 83.28; H, 9.32. Found: C, 83.36; H, 9.24 No carbonyl derivatives of this hindered ketone could be prepared. Those attempted were the oxime, dinitrophenylhydrazone, and semi- carbazone. Only unreacted ketone was recovered. Attempted catalytic reduction of 2, 2, 5, 5-tetramethylbenzo- cycloheptene- 1-one A solution of 10. 0 g. of 2, 2, 5, 5-tetramethy1benzocycloheptene-1- one in 50 m1. of ethanol was treated with hydrogen at 200° and 2200 p. s. i. for 18 hours in a-Magne-Dash reaction bomb using 1. 5 g. of copper chromite catalyst. The catalyst was filtered and the solvent removed at atmOSpheric pressure. The infra-red spectrum of the residue showed a strong carbonyl band. No uptake of hydrogen was noted during the reaction. The reduction was next attempted using Raney nickel as the catalyst. The ketone (2.0 g.) in 20 ml. of ethanol was treated with hydrogen at 1800 and 2100 p. s.i. for 10 hours in the presence of 0. 5 g. of Raney nickel. Some hydrogen uptake was noted and after working up as before,the:crude product showed only a slight carbonyl band in the infra-red. Gas chroma- tography of the material showed six components (no major one) and the experiment was concluded here. The last catalyst employed was 5% palladium on charcoal. The reduction was carried out in a low pressure hydrogenation} apparatus using 0. 5 g. of the ketone, 2 drops of perchloric acid, and 15 ml. of ethanol at 500 and 58 p. s. i. No hydrogen was taken up and only starting material was recovered. 129 Preparation of 2, 2, 5, 5-tetramethylbenzocyclo- heptene-l-ol (XXXVIII) The reaction was carried out in a 50—ml. three-necked flask equipped with a mechanical stirrer, dropping funnel, and a reflux‘condenser fitted with a calcium chloride drying tube. A solution of 5. 8 g. (0. 0268 mole) of 2, 2, 5, 5-tetramethylbenzocycloheptene-1-one in 10 ml. of tetrahydro- furan was added slowly to a stirred suspension of 1. 0 g. (0. 0264 mole) of lithium aluminum hydride in 15 m1. of the same solvent. The solution refluxed from the heat of the reaction. Reflux was continued an additional 2 hours. Sodium hydroxide (10% solution) was added until no more pre- cipitate formed, and the mixture was allowed to stand overnight. The clear organic layer was decanted and the residue was washed with two 20-m1. portions of solvent. The organic layers were combined and the solvent was removed on a steam bath. The colorless residue was dis- tilled through a small Vigreux column to give 5. 2 g. (89% yield) of the viscous alcohol boiling at 108-111°/0.09 mm., rig 1.5338. The infra-red spectrum (Figure 41) showed no carbonyl band. . In several larger runs the yield was essentially the same. A solution of 0.4 g. of 2, 2, 5, 5-tetramethylbenzocycloheptene-l-ol, 0. 5 g. of 3, 5-dinitrobenzoyl chloride, and l. 5 ml. of pyridine was heated on a steam bath for 8 hours. The dark solution was poured on ice and extracted with three 7-m1. portions of ether. The combined ether ex- tracts were washed successively with dilute sodium carbonate, dilute hydrochloric acid, and water. After drying over anhydrous sodium sulfate, the solvent was removed on a steam bath. . This yielded a very viscous yellow oil which would not crystallize. Chromatography on alumina with 20% ether-hexane still did not give a solid derivative. Attempted crystallizations from ethanol, hexane, and ethyl acetate also failed. A solid a-naphthalylamine adduct (36) was finally obtained by adding a solution of 0. 3 g. of the amine in 2 m1. of ether to 0. 3 g. of the viscous 130 ester in 2 ml. of the same solvent. . An oil separated which crystallized after three days to give reds-orange needles. One recrystallization from ethanol gave 0. 3 g. which melted at 94-950. . A second recrystallization from the same solvent raised the melting point to 94. 5-95. 50. £1131. Calcd. for C3ZH33N3O6: C, 69.17; H, 5.99; N, 7.56. Found: C, 69.21; H, 5.99; N, 7.56. Attempted reduction of 2, 2, 5, 5-tetramethylbenzocycloheptene-l-ol A solution of 4. 0 g. (0. 018 mole) of 2, 2, 5, 5-tetramethylbenzo- cycloheptene- l-ol in 20 m1. of ethanol was treated with hydrogen at 2200 and 2350 p. s.i. for 10 hours using 1.0 g. of c0pper chromite catalyst. No hydrogen was taken up. The crude product, after filtering and re- moval of the solvent still showed a strong hydroxyl band in the infra-red. Preparation of l-chloro-Z, 2, 5, 5-tetramethylbenzocyclo- heptene (XXXIX) A mixture of 5. 0 g. (0. 023 mole) of 2, 2, 5, 5-tetramethylbenzocyclo- heptene-l-ol, 3. 0 m1. of pyridine, 2. 5 ml. of thionyl chloride, and 20 ml. of carbon tetrachloride was stirredmagnetically for 12 hours; at room temperature and then refluxed for 2 hours. The reaction mixture was poured on ice and the organic layer was separated, washed. once with water, and dried over anhydrous sodium sulfate. The solvent was removed by distillation at atmospheric pressure, and the residual oil was distilled through a small Vigreux column. This yielded 4.4 g. of a pale yellow liquid boiling at 85--94o (0. 3 mm. ). . Redistillation gave 3. 9 g. (71% yield) of the chloride boiling at 89-91. 5°/0. 2 mm., 113 1.5380. A middle fraction was used for analysis and to obtain an infra-red spectrum (Figure 42). - In another run the yield was 56% with no reflux period. .Anal. .Calcd. for ClstlCl: C, 76.09; H, 8.94; C1, 14.97. Found: C, 75.94; H, 8.85; Cl, 14.82. 131 Reduction of 1-chloro-2, 2, 5, 5-tetramethylbenzocycloheptene A solution of 7. 3 g. (0.031 mole) of 2, 2, 5, 5-tetramethylbenzo- cycloheptene in 30 ml. of ethanol was hydrogenated in the presence of 5% palladium on charcoal. Hydrogen was taken. up slowly (room temperature) for 20 hours. , The solution was filtered, the solvent removed through a small Vigreux column, and the residual oil amounted to 5. 74 g. Gas chromatography indicated six components with one in about two-fold predominance over the others. . This was separated and purified by this method (2050 on 20% silicone). Only a very small amount of material (ca. 0. 2 g.) was obtained (retention time, 29 minutes) after several passes through the column (r123 1. 5279). The infra-red spectrum D is shown in Figure 43. C- Synthesis of 1. 1.4.4-tetrarnethylbenzocycloheptene (XXXIV) Preparation of l-bromo- 3 -methyl- 2-butene The procedure of Staudinger (55) was used. The reaction was carried out in a 300-ml. three-necked flask equipped with a mechanical stirrer, reflux condenser, and dropping funnel. ~ An ice-cold solution of 30. 5 g. (0. 373 mole) of hydrogen bromide in 40 ml. of glacial acetic acid (prepared by passing anhydrous hydrogen bromide into glacial acetic acid at 5°) was added slowly to a well stirred solution of 25.0 g. (0. 358 mole) of freshly distilled isoprene in 30 m1. of glacial acetic acid. at 3-60. After the addition was complete the mixture was stirred two hours at 50, the flask stoppered, and. allowed to stand 46 hours at 20... The yellow solution was poured into 200 ml. of ice water, the oily layer separated, washed with. 20 ml. of water, and dried over anhydrous sodium sulfate. Distillation through a Vigreux column yielded. 33. 5 g. (62%) of the colorless bromide boiling at 61-630/61 mm. , 2° 1.4880-l.4900. - Young and Linden, (56) reported n20 1.4900. . In another run a yield of 64.4% D was obtained (b.p. 54-55.5°/45 mm., n3 1.4858-1.4865). 132 Preparation of l-phegrl-Z, 2, 5-trimethyl-4-hexene-1-one (XL) The reaction was carried out in a 300-ml. three-necked flask equipped with a mechanical stirrer, dropping funnel, and reflux condenser fitted with a calcium chloride drying tube. A solution of isobutyrophenone (32. 8 g. , 0. 222 moles) in 25 ml. of dry benzene was added dropwise to a warm (50°) suspension of 9.44 g. (0. 242 mole) of sodium amide in. 50 ml. of the same solvent with stirring. The dark solution was stirred and refluxed for forty minutes. A solution of 33. 5 g. (0. 222 mole) of 1-bromo- 3—methyl-2-butene in 25 m1. of dry benzene was added over thirty minutes to this hot solution. After stirring an additional two hours, the mixture was refluxed for twelve hours. Water (100 ml.) was added to dissolve the precipitated sodium bromide and the organic layer separated. . After washing with. three 25-ml. portions of water it was dried over-anhydrous sodium sulfate. The solvent was removed and the residual oil was distilled through a Vigreux column. 7 This yielded 42g. (87. 5%) of the colorless ketone boiling at 81-83°/0.07 mm., n19 1. 5182-1. 5186. . Redistillation D gave 35.7 g. boiling at 75-760/0.06 mm., n20 1.5180. The infra-red spectrum (Figure 44) showed a strong carboIliyl band at 5. 95 (1 which masked the double bond absorption. . In another run the yield was 72. 7% of ketone which boiled at 88-890/0. 09 mm. , n3 1. 5181 .after two distillations. A procedure similar to that described by Smith and Shriner (57) was used to determine the double-bond content of the ketone. An 0. 3020 g. (1.40 mmoles) of XL was dissolved in 50 m1. of glacial acetic acid, and 32. 00 ml. of 0. 1054 N potassium bromide-bromate solution added followed by 5 ml. of 6 N hydrochloric acid. The mixture was allowed to stand in the dark for five minutes and then 10 ml. of 20% potassium iodide was added. . The iodine liberated was titrated with 0.0585 N (9.60 ml.) sodium thiosulfate using a starchindicator. . The millimoles of bromine taken up by the olefin was then 1/2 [(32.00) (0.1054) - (0.60) (0.0585)] = 1.41. 133 Thus 1. 01 moles (l.41/1.40) of bromine was taken up by one mole of the unsaturated ketone indicating one olefinic double bond. A duplicate determination gave identical results. The ketone XL (0. 30 g. , 0. 0014 moles) in 10 m1. of ethanol was mixed with 6. 0 ml. of 0. 25 N 2, 4-dinitrophenylhydrazine reagent (58). The solution became cloudy immediately, but no solid separated for two hours. The yellow derivative was filtered off and recrystallized-twice from dilute ethanol. This yielded 0. 36 g. (65%) of yellow needles melt- ing at 86. 8-87. 20 (softened at 820) after drying under vacuum at room temperature. final—l. Calcd. for CZIHMN404: C, 63.62; H, 6.10; N, 14.14 Found: C, 63.52; H, 5.95; N, 14.21. Preparation of 2, 4, 4-trimethyl-6-phenyl-2-hexene (XLI) A mixture of 24. 9 g. (0. 115 mole) of 1-pheny1-3, 3, 5-trimethyl- 4-hexene-1-one, 30 g. of potassium hydroxide, 30 ml. of 99% hydrazine, and 200 m1. of ethylene glycol was refluxed for three hours in a 500-ml. flask fitted with a distilling head and a thermometer extending into the liquid. After distilling the reaction mixture until the pot temperature reached 1880, the reflux was continued for six hours. . Water (250 ml..) was added to the cooled mixture and the oily product extracted into ether. The aqueous phase was extracted with two 50-ml. portions of ether. The combined extracts were washed twice with water and dried over anhydrous sodium sulfate. The solvent was removed through a Vigreux column and the residual oil distilled (same column) under reduced pressure. The colorless olefin boiled at 670/0.07 mm. , n3 1. 5063 and amounted to 14.4 g. (62%). The infra-red spectrum (Figure 45) showed a tri-substituted double bond (5. 99 u) and no carbonyl band. The pale yellow pot residue amounted to 8.4 g. Titration of the olefin 134 by the potassium bromate-bromide method (previously described) showed 0. 99 mole of bromine taken up per mole of Clstz- » In another run using 10 g. of ketone and distilling until the pot temperature reached 1950 gave 79% of the olefin (b.p. 72-740 at 0.08 mm). An analytical sample was purified by gas chromatography (20% silicone) and had n3 1. 5060. ~A_na_l_:_Calcd. for C15H223 C, 89.04; H, 10.96. Found: C, 89.13; h, 11.09. Ozonolysis of 2, 4,4-trimethyl—6-phenyl-2-hexene (XLI) A procedure similar to that of Bailey (37) was employed. .An 0. 108 g. (0. 000536 mole) sample of the above olefin in 40 ml. of methanol was treated with ozone (Welsbach Ozonator) at -500 until the solution became pale blue (ca. 45 minutes). After warming to room temperature, 0.4 g. of solid potassium iodide was added followed by 0. 9 g. of sodium hydroxide in 8 m1. of water. Then 0. 7 g. of iodine in 50 ml. of water containing 1. 0 g. of potassium iodide was added, and the yellow precipitate of iodoform separated immediately. lAfter standing overnight, the solid was filtered off, washed with 10 ml. of cold water, and dried (thirty minutes at 85° and two hours under vacuum at room temperature). The iodoform amounted to 0. 160 g. (76% yield) and melted 120-1220. Reaction of 2, 4, 4-trimethyl-6-phenyl-2-hexene with aluminum chloride The reaction was carried out in a 100-ml. of flask equipped with a dropping funnel, magnetic stirrer, and reflux condenser fitted with a calcium chloride drying tube. A solution of 1.0 g. (0. 0050 mole) of 2, 4, 4-trimethy1-6-phenyl-2-hexene in 15 m1. of methylene chloride was added over 6. 5 hours to a well stirred mixture of 0. 73 g. (0. 0055 mole) of aluminum chloride in 33 ml. of the same solvent. The dark solution 135 was stirred an additional eight hours and then hydrolyzed bythe addition of 30 ml. of 1:1 hydrochloric acid. The organic layer was separated and washed once with 10% hydrochloric acid and twice with water- After drying over anhydrous potassium carbonate, the solvent was removed and the residue distilled (no column) to give 0.65 g. of pale yellow liquid which decolorized bromine in carbon. tetrachloride. Gas chromatography (20% silicone, 2100) showedveight components (no major one). The mixture was not investigated further. Reaction of 2, 4, 4-trimethyl-6-phenyl-2—hexene with hydrogen chloride and aluminum chloride The reaction was carried out in a 250-1111. flask fitted with a dropping funnel and stirred magnetically. A solution of 1.40 g. (0.00694 mole) of 2, 5, 5-tri’methyl-6-phenyl—2-hexene in30 ml. of methylene chloride was added solution of 0. 30 g. (0. 0022 mole) of aluminum chloride in 40 m1. of the same solvent at -20(ice-salt bath) over five hours; hydrogen chloride had been passed for several minutes through the solvent used for the catalyst. . The reaction mixture was allowed‘to stand for twenty hours at 20 and then poured on ice. The organic layer was separated, washed once with dilute hydrochloric acid, twice with water, and dried overanhydrous potassium carbonate- The solvent was removed through. as Vigreux column and the residue distilled (no column) to give 1. 13 g. (80. 7% yield) of a pale yellow liquid. This was purified by gas chromatography (2100 on 20% silicone) which showed essentially one component (retention time, 28 minutes). . The product was distilled to remove any silicone and had 2° 1. 5136. . The infra-red spectrum (Figure 46) had bands at 7. 25 (.1 and 7. 34 p. (gem-dimethyl group), and at 13. 61 p. (1, 2-disubstitution). . The ultraviolet spectrum (Figure 47) showed peaks at 274 mu (6 = 935), 267‘ mu ( 6 = 1030), and1260O mu (6 = 930). .A mass spectrum indicated a major fragmentation course 136 to be 202 -—>159‘ + 43 and none by 202 -—-> 187 + 15. The product was probably 1-isopropyl-3, 3-dimethyltetralin (XLII). .Anal. Calcd. for C15H22= C, 89.04; H, 10.96. Found: C, 89.09; H, 10.93. Preparation of l, 1, 4, 4-tetramethylbenzocycloheptene (XXXIV). The reaction was carried out in a 150-ml. two-necked flask equipped with a magnetic stirrer, dropping funnel, and a reflux condenser fitted with a calcium chloride drying tube. Boron trifluoride etherate was prepared _i_n_s_i_t_u_ by pas sing boron trifluoride gas into 1 m1. of ether, with cooling in ice. . Methylene chloride (25 ml.) was added, and a solution of 1.0 g. (0.0049 mole) of 2, 5, 5-trimethyl-6-phenyl-2-hexene(XLl)in 25 ml. of the same solvent was allowed to drop in over 2 hours withstirring (ice bath). The bath was removed and stirring continued for 19 hours. Water (25 ml.) and concentrated ammonium hydroxide (5 ml.) were added, the organic layer separated, washed once with water, and dried over anhydrous sodium sulfate. .The solvent was removed through a Vigreux column, leaving a yellow liquid (ca. 1 g.). Gas chromatography (2100 on 20% silicone) showed two components, one with the same retention time as the starting olefin (31. 5 minutes), and the other the same retention time (28 minutes) as the product obtained when the reaction was carried out using aluminum chloride and hydrogen chloride. The latter (XLII) was the predominant product by a ratio of 2:1 (by peak areas). Both components were purified by gas chromatography. The major component had an infra-red spectrum identical to that obtained from the aluminum chloride-hydrogen chloride catalyzed cyclization. The minor component (11% 1. 5192) was not starting material, since it showed no unsaturation (bromine in carbon tetrachloride), and its infra- red spectrum indicated 1, 2-disubstitution (Figure 49). . The ultraviolet and NMR spectra (Figures 50 and 51) indicated that this was the desired 137 1, 1, 4, 4-tetramethylbenzocycloheptene (XXXIV). A mass spectrum showed a major fragmentation route to be 202 ——> 187 + 15. A second run was made to obtain more of the material. . In this run, small samples were removed periodically, washed with water andanalyzed by gas chromatography. The peak ratio became constant at 2:1 after 2 hours. £11131. Calcd. for Clstz: C, 89.04; H, 10.96. Found: C, 88.87; H, 11.12. Reaction of 2, 4, 4—trimethyl-6-phenyl-2-hexene with ferric chloride The reaction was carried out in a 50-ml. three-necked flask equipped with a magnetic stirrer, dropping funnel, and a reflux condenser fitted with a calcium chloride drying tube. . A solution of 1. 0 m1. of 2, 4, 4-trimethyl-6-phenyl-2-hexene in 20 m1. of methylene chloride was added over a 15 minute period to a stirred solution of 1.0 g. of ferric chloride in 10 ml. of the same solvent. Small samples (ca. 1 ml.) were removed periodically, washed with water, and analyzed by gas chroma- tography on 20% silicone at 1700. The same two products were obtained as with boron trifluoride etherate. The ratio (area) of the peaks became constant at 2:1 (as before) after less than 30 minutes after the addition of the olefin. The peak ratio was still the same after 3 hours. Reaction of 2, 4, 4-trimethyl-6—éphenyl-2-hexene with stannic chloride The procedure was the same as used with ferric chloride, except that 1 g. of stannic chloride was substituted for therferric chloride. Gas chromatography was also carried out in the same manner. Only the olefin was present even after 12 hours at room temperature. SUMMAR Y Wagner found that 2-methyl-5-chloro-2-pentene reacted with phenol at 150° to yield two isomeric products "A" and "B. " He showed that "A" was 5, 5-dimethylhomochroman. The identification of the phenolic product "B" was the first purpose of this thesis. The infra-red, NMR, ultraviolet, and mass spectra of "B" indicated that it was probably 1, l-dimethyl-S-tetralol. . This was confirmed by synthesizing this tetralol as shown. OCH3 OH During the investigation of "A, " and also from the reaction of 2, 2-dimethyltetrahydrofuran with phenol, a second isomeric phenol "C" was isolated. Another purpose of this thesis was to identify "C. " It was shown to be an. approximately equimolar mixture of "B" and 1, 1- dimethyl-7-tetralol, an authentic sample of the latter being synthesized as shown. The unusual orientation (the tertiary group of meta, not para) CH3O o CH3O 0 HO m l. Nao-iP; . 1..W. K. g 2. CH3I 2.. HI 138 139 of the phenolic products of the alkylation reaction with the homoallylic chloride suggested that the reaction may proceed through one of the ions shown. The ion formed could attack phenol at the oxygen atom, or + + F—A—W ,CH2 CH3\ CH2 CH3 N CH3 r’. , >C —_Q}'c1i2 - Cl —-5 >C '-=:-'CH | or /C‘«'-=-‘C<§| CH, / CH, \CHZ CH, CH, CHz unsymmetrical symmetrical ortho, or para positions, and then cyclize to give only the three products actually isolated. . To determine which ion was involved, 2-methyl-5- chloro-Z-pentene-S-dz was used in the reaction. The labelled compound was synthesized by the route shown. This was allowed to react with BrCHZ-CHz-COZCZHS fi‘gD—b BrCHz-CHZ-CDZOHflé-rBrCHz-CHZ-CDZCl 3 KCN CH3 ~' CH M I /‘C(0H)CH,-CH,-CD,c1 «LS— CZHSOzC-CHz-CHZ-CDZCI CH, W -H,o \—i—i-—C (H OH NC-CHz-CHz-CDZCI HCl CH 3\ /C = CH- CHz-CDzCI CH3 phenol, and the products examined by NMR, which showed that the two - methylene groups of the homallylic chloride had become equivalent during the reaction. . Thus the symmetrical ion was involved. . The ultraviolet spectrum of "A" suggested that there was a possible barrier to the seven-membered ring interconverting between the two chair forms. Examination of its NMR spectrum indicated that the ring 140 was "flipping" at room temperature and also at -100°. , An analogous hydrocarbonin whichthe barrier tointerconversion of the 7--m,embered O\ \ C- CH(CH3)Z 1. NaNHz ‘ x 2. isoprene 2. BF, 7 hydrobromide ring was increasedwas synthesized as shown. Examination of its NMR spectrum showed the ring to be "flipping ‘ at room temperature, but in the region of -30>to -60°, this “flipping" appeared to cease. 10 11. 12.- 13.. 14. 15. 16. . LITERATURE CITED 1A..Baeyer and V. Villiger, Ber. .35, 3013 (1902). M . H- Hart. and J. H. Simons, J..Am.. Chem. Soc., 71, 345 (1949). . H. Hart, W. L- Spliethoff, and H.. S- Eleuterio, J. Am. Chem. Soc. , 3.9: 4547 (1954). .. H. Hart and F. A. Cassis, J..Am. Chem. Soc., Z53, 1634 (1954). . T. A. Favorskaya and Sh-A. Fridman, J- Gen. Chem. (USSR), $5: 421 (1945). .. C. R. Wagner, Ph- D. Thesis, Michigan State University, 1955. . H. Hart, unpublished work. .. H. Hart and C..R. Wagner, Proc. Chem. Soc., 284 (1958). . D. C. Iffland and H. Siegel, J. Am. Chem. Soc., ’89,, 1947 (1958). .. H. Hart, Anal. Chem., 24', 1500 (1952). H. Hart and E-A. Haglund,sJ. Org. Chem., ’13, 396 (1950). C. W. Young, R- B. DuVall, and N. Wright,.Anal-Chem., “2'3, 709 (1951). J. D. Roberts, Nuclear Magnetic Resonance, .McGraw-Hill, N. Y. , 1955,. p. (a) 30,, and (b) Chapter 3. F. W- McLafferty, Anal.. Chem. , fir 306 (1956). R. A. Friedel and M. Orchin,. Ultraviolet Spectra of Aromatic - Compounds, John Wiley, N. Y. , 1951,. plates (a) 49, (b) 51,. (c) 52, (d) 34,, (e) 42,. (f) 50. G. V. D. Tiers, Characteristic NMR Shielding Values for Hydrogin in Organic Structures, Part I, (a) p. 19, (b) p. 25 (1958). 141 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29.. 30. 31. 32. 33. 34. 35. 142 N. D. Coggeshall and A. S. Glessner," J- Am. Chem. Soc. , 11’, 3150 (1949). ' J. 'W. Cornforth,. R- H. Cornforth, and R. Robinson, J. Chem. Soc., 689 (1942). ‘ H. A. Bruson and J. W. Kroeger, J. Am. Chem. Soc., '63., 36 (1940). A. G. Goldberg and R. P.. Linstead, J. Chem. Soc., 2343 (1928) I. N. Nazarov and A. 1.. Kuznetsova, Bull. acad. sci- URSS, Classe jsci.chim., 431, (1941). r C. A., 2‘9. 1297.309421- B. W..Horrom and H. E. Zaugg,» J. Am..Chem.. Soc., 72, 721 (1950). W S. Winstein andrE- M. Kosower, J. Am..Chem. Soc., 81, 4399 (1959). R. F- Nystrom, J. Am..Chem. Soc., 81’, 610 (1959). J- A. Pople, 'W. G. Schneider, and H. J. Bernstein, .Can- J. Chem., 22, 1060 (1957). G. H. Whitham, Proc.. Chem. Soc.,. 422 (1961). G. Baddeley and J. Vickars, J. Chem. Soc., 2455 (1956). E. M. Arnett and C. Y. Wu, J. Am. Chem. Soc., “83‘, 5660 (1960). A. Dreiding, Helv. Chim. Acta, $2“, 1339‘ (1959). .H. E.Ungnade,, E- E- Pickett, L. Rubin, and E. Youse, J. Org. Chem. , ’16, 1318 (1951). H. E.Ungnade and R. W. Lamb, J..Am..Chem. Soc., 14‘, 3789 (1952). R. A. Smith, J. Am... Chem. Soc., 32,. 3718 (1933). S- Julia, M. Julia, and B. Bemont, Bull- soc. Chim.. France, 1449 .(1959). J- F. T- Dippy and J- T. Young, J. Chem. Soc. ,. 3919 (1955). A. E. Gillam and E. S. Stern, Electronic Absorption Spectroscopy, Edward Arnold-Ltd.,. London, 1957, p. 141. 143 36. O. T. Benfey, J. R. Stanmyer, B..Mi11igan, and E.M. Westhead, . J. Org. Chem., ’29, 1777 (1955). 37. P. S. Bailey, 5- 5. Bath, W. F. Thomson, H. C. Nelson, and E. E. .Kawas, J. Org. Chem., $1,197 (1956). 38. W. R- Moore, E- Marcus, S. E. Fenton, and R. T. Arnold, .Tetrahedron,~§\, 171 (1959). 39. G. V. D. Tiers, J. Phys. Chem.,gi, 1151 (1958). 40. C. G. Overberger and C. W- Roberts, J- Am.. Chem. Soc., 71, 3618 (1949). 41. R. Adams, ed., -Org. Reactions, 2, 136 (1944). 42. R. L- Shriner and R. C. Fuson, Identification of OrganicCompounds, John Wiley, N..Y., 1948, (a) p. 163, (b) p. 171. 43. O. Fischer and W. Kern,-J. prakt. Chem., 24", 34 (1916). 44. E. C- Kendall and B. McKenzie in Org. Synth. , C011- Vol-I, 246 (1941). "' "" 45. C. F. H. Allen in. Org. Synth., Coll. Vol-I, 156 (1932). W 46. A. Bruylants,. M. Tits, C. Dieu, and R. Gauthier, Bull. soc. Chim. Belg., 61, 366 (1952). 47.. E. A. Fehnel, J. Am. Chem. Soc., '14., 1569(1952). 48. A. T- Blomquist and C. J. Buck, J. Am. Chem. Soc., ’81, 672 (1959). 49.. L. Henry, Compt. rend., 142, 133 (1905). M 50. B- K. Campbell and K..N. Campbell, J. Am.. Chem. Soc., 379’ 1372 (1938). 51. G. H- Stillson, D. W. Sawyer, and C. K. Hunt, J. Am. Chem. Soc. 61, 303 (1945). 52. Houben—Weyl, Methoden der Ozganischen Chemie, Georg Thieme, Stuttgart, 1952, Vol. 8, p. 415. 53 54. 55.. 56. 57. 58. 144 .r M. T. Bogert and D- Davidson, J. Am.. Chem. Soc., 129’ 182 (1934). E. B. Hershberg in Organic Synth. ,. Coll. .Vol. 2, 129 (1943). H. Staudinger and K. Schilt, Helv. Chim.~ Acta, A5‘, 746 (1922). W. G. Young and S. L. Linden, J. Am. Chem. Soc.,.6~2, 2912 (1947). W . T. Smith and R- L. Shriner, The Examination of New Org-Elie -Compounds, John Wiley, N. Y., 1956, p. 124. G. D. Johnson, J. Am.. Chem. Soc., 73, 5888 (1951). NM CHEMISTRY ”BRA!“ 1?? 9:732‘ 001-8... I31