IV‘ESI_J RETURNING MATERIALS: P1ace in book drop to LJBRAfiJES remove this checkout from —:-—— your record. FINES wiH be charged if book is returned after the date stamped be1ow. PART I THE SCOPE AND MECHANISM OF BISANNELATION OF BISARYNE EQUIVALENTS WITH FURANS PART II DEOXYGENATION OF ARENE-1,4-ENDOXIDES WITH LOW-VALENT TRANSITION METALS By Godson Chukuemeka Nwokogu A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1981 (_- / / ABSTRACT PART I THE SCOPE AND MECHANISM OF BISANNELATION OF BISARYNE EQUIVALENTS WITH FURANS PART II DEOXYGENATION OF ARENE-1,4-ENDOXIDES WITH LOW-VALENT TRANSITION METALS By Godson Chukuemeka Nwokogu In Part I of this thesis, optimal reaction procedures for bisannelation of furans to tetrabromoarenes (bisaryne equivalents) were developed by considering the nature and solubility of the bisaryne equivalents, solvent effects, reaction temperature, concentration of n-butyllithium (g- BuLi) and its rate of addition. The generality of the bisannelation process was demonstrated by its success with various p-disubstituted tetrabromoarenes. Bisaryne equiva- lents with electron-releasing substituents gave good yields of bisadducts while those with electron-withdrawing groups gave poor yields with all furans. Tetrabromoarenes with different 1,4-substituents, e.g. methyl tetrabromo-p-cresolate and tatrabromo-p-chlorotoluene gave poor yields due to a secondary reaction. Godson Chukuemeka Nwokogu The syn/anti ratios of bisadducts showed very little dependence on the bisaryne equivalent structure and on the furan substituents. The isomer ratios for bisadducts of the same bisaryne equivalent with furan itself and methyl- substituted furans were not much different. These results are consistent with the intermediacy of a benzyne. Investigation of the mechanism of bisannelation showed that it could occur through either monolithiation or dilithiation of the tetrabromoarenes. Whereas tetrabromoarenes with electron-releasing groups, e.g., tetrabromo-p-xylene, formed bisadducts by stepwise monolithiation/annelation, those with electron-withdrawing groups like tetrabromo-p- difluorobenzene formed dilithio—arenes, which then underwent aryne formation/annelation. The relative position of the lithium atoms in these dilithio-derivatives was exclusively para. Treatment of most tetrabromoarenes with one equivalent of n-BuLi in the presence of furans resulted in mono-adducts in toluene only. Methyl tetrabromo-p-cresolate and tetrabromo- p-dimethoxybenzene gave monoannelation products also in THF due to their solubility in this solvent. For tetrabromoarenes with different 1,4-substituents, e.g., methyl tetrabromo-p-cresolate g3 and tetrabromo-p- chlorotoluene, treatment with one equivalent of n-BuLi led to two isomeric monolithiation intermediates. One of these monolithiation derivatives formed benzyne only while the other Godson Chukuemeka Nwokogu probably formed benzyne in competition with proton abstrac- tion. For example, treatment of £1 with one equivalent of n-BuLi and excess 2,5-dimethylfuran gave a monoadduct and methyl 3,5,6,-tribromo-p-cresolate. Polylithiation of bisaryne equivalents stOpped at the dilithio-derivative no matter how much excess of n- BuLi was employed. In Part II of this thesis, an efficient new method for the direct deoxygenation of arene endoxides was developed. Low-valent forms of iron, titanium and tungsten were used to aromatize naphthalene-l,4-endoxides and anthracene-l,4:5,8-diendoxides, even when peri-interactions were severe. Hence, 1,4,5,8,9,10-hexamethylanthracene was prepared from l,4,5,8,9,10-hexamethy1-l,4,S,8-tetrahydro- anthracene-l,4:5,8-diendoxide using low-valent titanium. Adducts of furan and 2,5-dimethylfuran were effectively deoxygenated by the three metals. Adducts of tetramethylfuran underwent a 1,3-hydrogen migration under the influence of reduced iron whereas reduced titanium led either to partial deoxygenation or to a mixture of non-oxygen containing hydrocarbons. Reduced tungsten effected deoxygenation of only the endoxide functions of 9,10-dimethoxyanthracene diendoxides. Reduced iron and titanium, however, also effected the replacement of the methoxyl groups with hydrogen. Hence l,4,5,8-tetramethyl-9,lO-dimethoxy-l,4,S,8-tetra- hydroanthracene-l,4:5,8-diendoxide gave 1,4,5,8-tetramethyl—9, Godson Chukuemeka Nwokogu 10-dimethoxyanthracene when treated with reduced tungsten, but led to 1,4,S,8-tetramethylanthracene on reaction with low-valent iron or titanium. ACKNOWLEDGMENTS May I express my deepest gratitude to Professor Harold Hart for his advice and guidance during the course of this study. His constant interest in this project and suggestions were encouraging especially at the difficult spells. Appreciation is extended to Michigan State University, National Science Foundation and National Institutes of Health for financial support in the form of teaching and research assistantships. I am very much indebted to my parents for their unfailing love and support. Many thanks are due to my friends, especially Ms. Jean Ellen Wyche, for their invaluable contributions to my well-being during the course of this study. ii TABLE OF CONTENTS PART I THE SCOPE AND MECHANISM OF BISANNELATION OF BISARYNE EQUIVALENTS WITH FURANS INTRODUCTION . RESULTS AND DISCUSSION . A. B. B(i). B(ii). B(iii). B(iv). Choice and Preparation of Reactants Development of Reaction Conditions for the Use of Tetrabromoarenes as Bisaryne Equivalents with Furans The Effect of Solubility of the Bisaryne Equivalent and Concentration of n-BuLi on Yields of Bisadducts Determination of the Best Solvent for Bisannelation with the Different Tetrabromoarenes Effect of Reaction Temperature on Bisadduct Yield Electronic Influence of Substituents on the Furans Factors Affecting the Distribution of the Syn and Anti Isomers of Bisadducts Mechanism of Bisannelation Attempts at Trilithiation of Tetrabromo- arenes Electronic Effects of the Remaining Substituents on the Metalation of and Aryne Formation from Tetrabromoarenes Some Transformations of the Bisadducts iii .14 .14 -19 .25 .26 . 31 -35 . 38 .43 . 61 . 68 . 76 EXPERIMENTAL . 1. 2. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. General Procedures 9,10-Dimethoxy-l,4,S,8-tetrahydroanthracene-1, 4:5,8-diendoxide (£8) . . . . . . . . 9, 10- -Dimethy1- l, 4, 5, 8- tetrahydroanthracene- l, 4. 5, 8- diendoxide (32) . . . . Methyl Tetrabromo-p-cresolate (£4) Tetrabromo-p-chlorotoluene (gs) Tetrabromo-p-dimethoxybenzene (g9) Tetrabromo-p-difluorobenzene (£8) Tetramethylfuran (49) 1,4,5,8,9,lO-Hexamethy1-l,4,5,8-tetrahydro- anthracene-1,4:5,8-diendoxide (4;) . Decamethyl-1,4,5,8-tetrahydroanthracene-l,4:5, 8-diendoxide (4g) . . . . . . . . . . l,4,5,8-Tetraphenyl-9,lO-dimethyl-l,4,5,8- tetrahydroanthracene-l,4:5,8-diendoxide (4g) . . . 9- -Methyl- 10- -methoxy- 1, 4, 5, 8- tetrahydroanthracene- 1, 4: 5, 8- diendoxide (46) . . . lO-Methoxy-l,4,5,8,9-pentamethyl-1,4,5,8- tetrahydroanthracene-l,4:5,8-diendoxide (41) 9,?,?~Trimethyl-lO-methoxy-l,4,5,8-tetrahydro- anthracene-1,4:5,8-diendoxide (42) . . . l,4,S,8,9-Pentamethyl-10-chloro-l,4,5,8-tetra- hydroanthracene-l,4:5,8-diendoxide (SQ) 9,10-Dimethoxy-l,4,S,8-tetramethy1-l,4,5,8- tetrahydroanthracene-l,4:5,8-diendoxide (£1) 9,10-Dimethoxy-l,2,3,4,5,6,7,8-octamethyl- l 5, 4. 8-tetrahydroanthracene-l,4:5,8-diendoxide (_g) 9, 10- Dichloro- l, 4, L 8- tetrahydroanthracene- L 4: L 8- diendoxide (53) . . . . . . 9,10—Dichloro-l,4,S,8—tetramethyl-l,4,5,8- tetrahydroanthracene-l,4:5,8-diendoxide (£4) iv 80 80 81 82 83 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 9,10-Dif1uoro-l,4,S,8-tetrahydroanthracene- l,4:S,8-diendoxide (EE) . . . . . 5, 8- Difluoro- 6- bromo- L 4- dihydronaphthalene- l, 4- endoxide (57) . . . Attempted Interception of a Dilithio- derivative of Tetrabromo-p-xylene g; Reaction of Tetrabromo-p- xylene 33 with One Equivalent of BuLi in THF. Attempted Preparation of the Monoadduct 62 . . . . . . . . . . . Reaction of Tetrabromo- -p- xylene 33 with Three Equivalents of BuLi in the Presence of Furan at -78° C. The Preparation of the Monoadduct 62 Reaction of 33 with Three Equivalents of BuLi at -78° C in —he Presence of 2, S- dimethyl- furan. l, 4, S, 8- -Tetramethyl- 6, 7- Dibromo- l, 4- dihydro- naphthalene- l, 4- endoxide (61) . . . . L 4, 5- -Trimethy1- 6, 7- dibromo- 8- -methoxy- L 4- -dihydronaphthalene- L 4- endoxide (64) and the Tribromoarene 65 . . . Attempted Deuterium Incorporation in gg l,2,3,4-Tetramethyl-S,8-dimethoxy-6,7-dibromo-l, 4-dihydronaphthalene-1,4-endoxide (99) . l,4-Dimethyl-S,8-dimethoxy-6,7-dibromo-l, 4-dihydronaphthalene-l,4-endoxide (12) 1,4,5-Trimethyl-6,7-dibromo-8-chloro-l,4- dihydronaphthalene-l,4-endoxide (ll); 6i and 21 Interception of the Dilithio- arene 72 in the Presence of Furan. Preparation of 2—5- Dibromo- p-dichlorobenzene 1g . p- Diacetyl Derivative 74 and Monoacetyl Derivative 75 of Tetrabromo- -p- -dichTorobenzene p- Bisphenylthio- 2, L dibromo- 3, 6- dichloro- benzene (76) . . . . Interception of the p-Dilithio Derivative of Tetrabromo-p-dichlorobenzene with Benzaldehyde. Preparation of 11 . . . l,4-Dibromo-2,S-dichloro-3,6-diiodobenzene (Z§) 98 99 .100 .100 .101 .102 .102 .103 .104 .105 .106 .107 .108 .109 .110 .111 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 2,S-Dichloro-3,6-dibromo-p-xy1ene (12) Reaction of Tetrabromo-o-dichlorobenzene with Four Equivalents of BuLT. Preparation of 2,3-Dichloro-S,6-dibromobenzene (8L) 2,3-Dibromo-5,6-dichloro-p-xylene (gg) l,2-Dibromo-3,6-diiodo-4,S-dichlorobenzene (fl) 2,5-Dibromo-p-difluorobenzene (82) Attempted Interception of the Dilithio Derivative of Tetrabromo-p-dichlorobenzene with a-bromoesters Preparation of the Trihalo-pseudocumene 81 p-Dichlorodurene (21) . . . . . . . . . . . . Dichloroprehnitene (2;) Attempted Trimetalation of Hexabromobenzene. Preparation of 2,3,5,6-Tetrabromobenzene (2;) Monolithiation of Tetrabromo-p-chlorotoluene. Preparation of the Tribromoarene 21 Reaction of Tetrabromo-p-chlorotoluene with Two Equivalents of BuLi. Preparation of 97-d1 and 22_ . . . Attempted Oxidative Demethylation of the Bisadduct 28 . . . . . . . . . Endoxide Ring Opening of the Bisadduct 28. Preparation of 103 . . . . . Bisepoxidation of the Bisadduct 32 Preparation of 104 . . Bisepoxidation of the Bisadduct gg. Preparation of 105 . . . vi 112 112 113 114 115 115 115 115 117 117 118 119 119 120 121 121 PART II DEOXYGENATION OF ARENE-1,4-ENDOXIDES WITH LOW-VALENT TRANSITION METALS INTRODUCTION . A. Direct Deoxygenation of Arene-l,4-endoxides B. Deoxygenation of Arene-l,4-endoxides by Indirect Methods . . C. Low-valent Transition Metals in the Deoxygenation of Epoxides RESULTS AND DISCUSSION . EXPERIMENTAL . 1. l,2,3,4,5,6,7,8-Octamethy1anthracene (g) 2. 9,10-Dimethylanthracene (118) 3. L 4, S, 8, 9- -Pentamethyl- 10- -methylene- 9, 10- dihydroanthracene (123) . 4. L 4- -Dimethyl- L 4- -dihydronaphthalene- l, 4- endoxide (129) . . . . 5. l,4-Dimethylnaphthalene (130) 6. Typical Deoxygenation Procedure with FeCl3. Preparation of 1,4, 5, 8-.Tetramethylnaphthalene (L32) . . . . . 7. l, 2, 3, 4- -Tetramethyl— L 4- dihydronaphthalene- 1, 4- endoxide (L33) . . 8. Typical Deoxygenation Procedure with WC16. Preparation of L 2, 3, 4- -Tetramethy1naphthalene (L34) . . . . . . . . . 9. 1, 2, 3, 4, S, 8- -Hexamethyl- l, L dihydronaphthalene- 1, 4- endoxide (L35) . . 10. Attempted Deoxygenation of Endoxide L35 with a l : 2 Ratio of FeCl3 to E” BuLi . . . . . vii 124 125 130 133 138 166 166 167 167 167 168 169 169 170 171 172 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. Typical Deoxygenation Procedure with TiC13. Preparation of 1, 4, S, 8, 9, 10- -Hexamethy1anthracene (L36) . . . . . . . . . . . . . . . . . Decamethyl-l,4-dihydroanthracene-1,4- endoxide (137) . . . . . . . . 9,10-Dimethoxyanthracene (138) Mixture of the Deoxygenated Products of the Bisadduct i; (g, 108 and 150) . . 1,4,5, 8,9,10-Hexamethy1-1,4-dihydroanthracene-1, 4- endoxide (L43) . Deoxygenation of the Bisadduct §§ with FeCl3. Preparation of Anthraquinone (139) Conversion of 9,10-Dimethoxyanthracene to Anthraquinone (139) 1,4,5,8-Tetramethy1-9,IO-dimethoxyanthracene (140) . . . . . . . . . . . . . . . Demethoxylation of anthracene 140. Preparation of L 4,5, 8-Tetramethy1anthraceF (L41) Direct Preparation of Anthracene 141 from Bisadduct El . Attempted Deoxygenation of the Endoxide 133 with Reduced Iron. Preparation of 1, 3, 4- -Trim€—hy1- 2- methylene- 1, 2, 3, 4- tetrahydronaphthalene- 1, 4- endoxide (L44) . Hydrogenation of 133 and 144. Preparation of 1, 2,3, 4-Tetramethy1-1, 2, 3,4-tetrahydronaphtha1ene- L 4- endoxide (L45) . Attempted Deoxygenation of Diendoxide A: with Reduced Iron. Preparation of 146 . . Hydrogenation of 43 and 146. Preparation of Decamethyl- octahyd_oanthracenL L 4: 5, 8- diendoxide (147) . . . . . . . . . . Reaction of the Endoxide 135 with Low-valent Iron. Preparation of 1, 3, 4, S, 8- Pentamethyl- 2- -methy1ene- 1, 2, 3, L tetrahydronaphthalene-. 1, 4- endoxide (L48) . . . . viii .173 .173 .174 .175 .176 .177 .178 .178 '180 .180 '181 '182 ~183 ~184 - 185 26. 27. Hydrogenation of 135 and 148. Preparation of 1, 2, 3, 4, 5, 8- -Hexametfiy1- 1, 2, 3, 4- tetrahydro- naphthalene- 1, 4- endoxide (149) . Deoxygenation of Bisadduct 52 with Low- valent Titanium. The Interception of 152 and 153 ix .186 .186 10. 11. 12. 13. LIST OF TABLES Preparation and Yield of Tetrabromoarenes Bisadducts of Bisaryne Equivalents Solvent Effect on the Yields of Bisadducts Yield of Bisadducts at Different Reaction Temperatures Effect of Solvent and Reaction Temperature on the ratios of Syn and Anti Isomers Syn/Anti Ratios of Bisadducts of the Same Furan with Different Tetrabromoarenes . . . . . Monoadducts of Bisaryne Equivalents Reaction of Dilithio-arenes with Electrophiles Electronic Character of the p-Substituents of Tetrabromoarenes . . . Influence of Solvent Polarity on the Distribution of Monometalation Products of Methyl Tetrabromo- p-cresolate (£4). . . . . . . . Yields of Naphthalene-l,4-endoxides Deoxygenation of Arene-l,4-endoxides with Low- valent Transition Metals Reaction of Adducts of Tetramethylfuran with Reduced Iron at -78°C . . . . . . . . 15 22 3O 33 .39 .41 .51 .57 .69 .73 '139 141 151 PART I THE SCOPE AND MECHANISM OF BISANNELATION OF BISARYNE EQUIVALENTS WITH FURANS INTRODUCTION The methods that have been used for the construction of fused aromatic rings can be classified into the Haworth (Scheme 1)1 and the Dials-Alder (Scheme 2)2 types of anne- lation to an existing benzene or six-membered ring. 1 Scheme 1 /\ coo .. Scheme 2 The Diels-Alder approach is especially suitable when the dienophile is an aryne. If a cyclic diene is used, at least two of the ring carbons of the adduct are sp3 -hybridized through a 1,4- and/or 9,10-bridge. Aromatization, as a last step, involves cleavage of this bridge. This modifi- cation of the cycloaddition pathway to acenes has found effective application in the synthesis of aromatic compounds with severe peri-interactions (Scheme 3).3 The 4a-azonia- anthracene é formed the adduct 4_with 3,6-dimethylbenzyne generated from the appropriate diazonium carboxylate. Reduction of the pyridine bridge to a dihydropyridine allowed a retro Dials-Alder reaction to generate the anthracene é. CCO 1) liAIH4 ‘ ct ‘ 2) A Scheme 3 The introduction of steric strain in the last step is more than compensated for by the stabilization gained from aromatization. This two-step reaction sequence, involving a cyclo- addition of an aryne to a cyclic diene followed by cleavage 3’4 has become the of the bridge from the resulting adduct, most versatile method for the synthesis of simple and strained polynuclear aromatic compounds. Because of the reactivity of arynes, the first step is highly exothermic.5 As an approach to highly substituted acenes, the sp3- hybridization of some of the ring carbons in the adducts allows accommodation of the steric strain due to the substi- tuents at this stage in the assembly of the molecular frame- work. This makes it possible to have all the required substituents present in the diene and dienophile, thus reducing the number of necessary steps for the preparation of a specifically substituted acene. The second step involves as an important driving force, the formation of a new aromatic ring by the elimination of a small fragment X (the bridge). Various types of bridges (X) have been used in the described sequence and their eventual cleavage often is preceded by the reduction or oxidation of the bridging system. The application of a pyridine bridge and its cleavage is illustrated in Scheme 3.3 As shown in Scheme 4 the cyclohexadienone 9 reacted with 3,6-dimethy1benzyne to give the adduct 1 which has a two-carbon bridging system. Reduction of the keto-group to an alkoxide was done to facilitate the retro Diels-Alder cleavage of the bridge in the aromatization step leading to §.6 Oxidation to N-oxides of the imino group of pyrrole 7 adducts allows its loss as a nitroso group. Hart and co- workers have used this bridging system in short syntheses of 1% I . {CHO + as Alum, 2) No“, A 1 Scheme 4 9 octamethylnaphthalene,8 decamethylanthracene g and do- 10 The introduction of the imino-N,N- 11 decamethylnaphthacene. dimethlamino bridge by Lai led to marked improvements in 2|! 0 ,n / / "'34me @ the overall yields of these strained aromatic hydrocarbons. 50$ azzz‘; CO I "-2!: .. mm. A. Oxygen and sulfur have also been used as bridges in the 12 synthesis of simply substituted acenes. One can envision applying this two-step strategy to the synthesis of anthracenes in several ways. The central ring could originate from either the diene component or the aryne precursor. In the former case, the diene can be an 12 isobenzofuran (X = O), isoindole (X = N-R)13 or . R ‘- I + )( —— ‘p— R 32 X.=() 11 Xm=lt-R 1.2. X” 2 For example, the reaction of isobenzothiophene (X = S).1 1,3-dipheny1isobenzofuran 19 (X = O, R = Ph) with benzyne gave the adduct 1§_(X = 0, R 8 Ph) which was deoxygenated either by treatment with zinc in acetic acid or by heating 12.14 The low availability and stability in diglyme at 162°C. of variously substituted isobenzofurans, isoindoles and isobenzothiophenes limits the versatility of this method. If the central ring of the anthracene originates from the aryne component, sequential annelations or Scheme 5 bisannelation might be possible (Scheme 5). A monoaryne derived from IS (A and B are groups capable of generating aryne) can undergo cycloaddition and the resulting monoadduct 11 can then be converted to the anthracene Zl_by one of two sequences (i and ii). The bridge can be eliminated to give the naphthalene lg which is set up, by the presence of groups B, to give another aryne leading to the anthracene 21 via the adduct 12 (sequence i). Alternately, monoadduct 11 can undergo a second aryne cycloaddition to give the bisadduct £9 from which both bridges could be eliminated (sequence ii). If the bridges (X) are different, they have to be cleaved by two different methods. Cleavage can be affected simultaneously, however, if the bridges are identical. Another alternative would be the direct formation of the bisadduct £9 if the groups A and B are such that they could be converted to arynes by the same reagent (sequence iii). In this case, the reactant lg serves as a bisaryne equivalent, although mechanistically, the formation of gg may occur in a stepwise manner, i.e. via 11. Examples of these different synthetic sequences have appeared in the chemical literature. Sequence (1) is exemplified by the synthesis of decamethylanthracene 2 (Scheme 6).9 The initial aryne was generated from the diazonium carboxylate function of g; and trapped with the cyclohexadienone g to form the adduct 2;. Cleavage of the 10 bridge before the next cycloaddition was done to avoid reaction of the carbonyl group of g; with n-BuLi which was used to generate a naphthyne from 34. Trapping of this Scheme 6 aryne with N-butyl~tetramethy1pyrrole resulted in the adduct gg which was converted to 2 using m-chloroperbenzoic acid. The use of two methods of benzyne generation and bridge cleavage have less adverse consequence on the 11 utility of this scheme than the difficulty of preparing the bifunctional benzyne precursor gg. The sequence (ii) was used in the preparation of the furan bisadduct ;§.15 Metal-hydrogen exchange between potassium amide and 2,5-dibromo-p-dimethoxybenzene Z9 in the presence of furan resulted in the monoadduct £1. After isolation, 11 was subjected to a second aryne cycloaddition in the same manner but at a higher temperature to give the bisadduct gg. The special advantage of this sequence is the ease of preparation of the bifunctional aryne synthon g9. Only one mention was made prior to the beginning of this work of the use of tetrahalobenzenes 22 as bisaryne 12 2 X1,2 = Halogen 16 When 2,6-dif1uoro-3,S-dibromo-p-xylene gg equivalents. was refluxed in THF with magnesium in the presence of furan, the bisadduct é; was isolated in 5% yield, the major product being the monoadduct El (34%). Substitution of n-BuLi for L BuLi , Furan magnesium improved the yield of g; to 15% and no monoadduct was isolated, thus giving an example of sequence iii (Scheme 5). It is with the synthetic applications of this latter approach (use of bisaryne equivalents) and the mechanism of 13 bisannelation that the first part of this thesis is concerned. In the second part, a new method for the removal of the bridge when X = 0 will be described. RESULTS AND DISCUSSION A. Choice and Preparation of Reactants: The preparation of aromatic fluorides from the parent arenes involves nitration, reduction to the amine, diazitization and pyrolysis of the diazonium tetrafluorobo- 16 For each additional fluorine substituent, some rate salt. or all of these steps have to be repeated. On the other hand, the polybromination of activated and slightly deactivated arenes is a facile one-pot and often high-yield process. Also, the bond strengths of the C-F and C-Br bonds (117 kcal/mol and 65 kcal/mol respectively) indicate that an g-lithiobromoarene will be more efficient in the generation of benzynes than an g-lithiofluoroarene. Considering therefore the ease of polybromination of most arenes and the ease of loss of LiBr from g-lithio- bromoarenes, it was anticipated that the more accessible tetrabromoarenes would be better bisaryne equivalents than tetrafluoro-, tetrachloro and mixed (especially those containing F and Cl) tetrahaloarenes. A precedent in favor of this expectation was the successful bisannelation of pyrroles to 2,3,6,7-tetrabromo-l,4,5,8-tetramethy1naphtha- 10 lene. Tetraiodobenzenes would be expected to be better 14 15 than other bisaryne equivalents but the substitution of iodine on the ring is a cumbersome and costly process. The tetrabromoarenes used in this work were prepared by standard methods. Activated arenes like p-xylene, p- cresol and p-chlorotoluene were brominated at below -5°C in the presence of catalytic amounts of iron filings and 17 excess bromine without solvent. The less reactive arenes like p- and g-dichlorobenzene, p-difluorobenzene (Table 1) TABLE 1 PREPARATION AND YIELD OF TETRABROMOARENES Entry Precursor Product Yield Reference 73 17 90 19,20 16 TABLE 1 (Cont'd.) Entry Precursor Product Yield Reference 3 Q 72 -- Cl Ci“. . 3' Br 4 . 71 21 Br r ' (”We 3.5 Cl 5 liilll 94 18 Cl ‘ 5 Ir 6 71 24 Br ; 17 TABLE 1 (Cont'd.) Entry Precursor Product Yield Reference (1 Cl c C! O 7 Bf 97 18 3.9 were refluxed in excess bromine with iron filings for about 6 h. Polybromination of p-dimethoxybenzene was accomplished in two steps. gave the 2,5—dibromoarene 29. Initial bromination in glacial acetic acid 21 After isolation, 29 was then treated with excess bromine at room temperature without catalyst and solvent to give gg. The choice of five-membered ring dienes-—in this case furans--is a recognition of the conformational require- ments of the Diels-Alder cycloaddition. In the transition state for the concerted reaction, the diene must have an 18 s-cis conformation. For acyclic dienes, the s-trans is usually the more stable and abundant conformer. Reactions of benzynes with acyclic 1,3-dienes therefore give low yields (often less than 20%) of [4 + 2] adducts. For example, when benzyne reacted with 2,3-dimethy1 butadiene, the major product arose from the ene reaction of the s-trans isomer (Scheme 7).22 With the cyclic dienes--furans and pyrroles--the s-cis conformation is fixed for the diene, thus ensuring the right geometry for the transition state in the cyclo- addition process. Furans were chosen in this work over 0% 6% 40% Scheme 7 pyrroles because they and their adducts are more stable to light and air. 19 All the furans used were commercially available except tetramethylfuran 4923 which was prepared by cyclizing 3,4-dimethy1hexan-2,5-dione at reflux in benzene in the presence of catalytic amounts of p-toluenesulfonic acid 0 /u\/ PbOz #- 0 m Reflux '6 (p-TsOH). The dione was prepared by the radical coupling of Z-butanone with lead dioxide (PbOz). B. Development of Reaction Conditions for the Use of Tetrabromoarenes as Bisaryne Equivalents WithIFurans We have found that tetrabromoarenes with 1,4-substi- tuents function as better bisaryne equivalents than tetrafluoro-, tetrachloro— and mixed (F, Cl, Br) tetrahalo- arenes thus allowing ready conversion of benzenes to acenes in two steps (the second step is the subject of the second part of this thesis). Treatment of the easily available tetrabromoarenes with two or more equivalents of n-BuLi in an anhydrous solvent (tetrahydrofuran, diethyl ether or toluene) at ~78°C or 0°C in the presence of furans gave the bisadducts after quenching either at the reaction temperature or after allowing the mixture to warm to 25°C. 20 The yields of bisadducts depended on the reagents and reaction conditions and in general, represent a marked 16 obtained with improvement over the results which Wittig 2,6-dif1uoro-3,S-dibromo-p-xylene £9. The success of this reaction with different para-disubstituted tetrabromoarenes shows that the method is broad in scope (Table 2). Our success in developing a new and effective method for the direct aromatization of the resulting endoxides, which is described in Part II of this thesis, demonstrates the general utility of this scheme (bisannelation with furans) for the synthesis of substituted, strained polynuclear aromatic compounds. The yields given in Table 2 are for isolated products at optimized reaction conditions; the products consist of a mixture of syn and anti isomers. In the case of Entry 13, about 10% of the 9,10-dibromo analog of the adduct £4 was also present (observed from the mass spectrum of the product mixture). The optimized procedure for bisannelation consisted of dissolution of 5 mmol of the tetrabromoarene in 300 mL 21 of the appropriate and anhydrous solvent containing up to ten molar equivalents of the appropriate furan. The bisaryne equivalents 34, 36 and §§ remained in solution, especially in THF, at this concentration even at -78°C. In the case of 33, 35, 31 and 32, some fraction of these reagents precipitated at the reaction temperature. After the slow addition of 2.5 to 3.0 molar equivalents of n- BuLi diluted with dry hexane under argon atmosphere (2 h), all reaction mixtures were either colored (reddish—brown or yellow) or milky white suspensions. The reaction mixture was stirred for a further 2 to 3 h at -78°C, quenched at the same temperature or allowed to warm to 25°C (6 h) before work up to isolate the product. Purification in most cases was achieved by chromatography and/or recrystalization. 22 TABLE 2 BISADDUCTS OF BISARYNE EQUIVALENTS Entry Bisaryne Equiv. Diene -Bisadduct Yield % B ' ’0 til 1 ‘~ 89 B r 16 33 — £16 23 TABLE 2 (Cont'd.) Entry Bisaryne Equiv. Diene Bisadduct Yield % S 87 6 71 7 73 8 16 24 TABLE 2 (Cont'd.) Entry Bisaryne Equiv. Diene Bisadduct Yield % 10 11 12 _1_6_ g9 41 22 .29. ;6_ 86 73 25 Table 2 (Cont'd.) Entry Bisaryne Equiv. Diene Bisadduct Yield % 13 "' — l4 B(i). The Effect of Solubility of the Bisaryne Equivalent and Concentration of n-BuLi on Yields of Bisadducts The solubility of the tetrabromoarenes and the concentration of n-BuLi have a marked influence on the yield and cleanness of the bisannelation reactions in all salvents. For example, when the volume of solvent and quantity of the bisaryne equivalent were such that the 26 tetrabromoarene precipitated to a large extent on cooling to -78°C, the product often contained some unreacted bisaryne equivalent even when excess n-BuLi was used. This was particularly true when the reaction was quenched at -78°C after 2-3 h. The reaction of a 0.36 M slurry of 33 in dry THF containing three molar equivalents of tetramethylfuran 59 with two equivalents of undiluted n-BuLi at -78°C gave 59% of the bisadduct 33 and about 20% of unreacted 33 was BuLi T HF! 48.0 recovered. When the reaction was repeated using an 0.017 M solution of 33 in dry THF containing excess tetramethylfuran at -78°C with 2.5 equivalents of g-BuLi diluted six times with hexane, the yield of 33_improved to 79%. In this case, the proton NMR spectrum of the crude product showed no trace of starting tetrabromoarene. B(ii). Determination of the Best Solvent for Bisannelation with the Different Tetrabromoarenes Tetrahydrofuran is often used as a solvent of choice 25 for metalation and metal-halogen exchange reactions. While 27 this may be so for this step of the bisannelation sequence, THF is not always the best solvent for the overall bisannela- tion. Reactions with different tetrabromoarenes in different solvents showed that THE worked well for bisannelations with 31 and 33. Even though the yields of bisadducts with 31 and 33 were low (cf. Table 2; Entries 12, 13 and 14), THE was the only solvent from which it was possible to isolate bisadducts of these bisaryne equivalents. Reactions with these tetrabromoarenes in ether and toluene, at various temperatures, gave oils and polymeric materials. The use of THF as reaction solvent is especially adequate if the bis- annelation can be run effectively at -78°C, as is the case with 31 and 33. When reactions in THF have inevitably to be run at 0°C or allowed to warm to room temperature, as is the case with 31 and 33, the yield is often low due to reduction, butylation and polymerization. For example, when the tetrabromoarene 3§_was reacted with n-BuLi in the presence of furan at -78°C and the reaction mixture was warmed to 25°C, the product consisted of the bisadduct 33, a mixture of monoadducts 33-32 (0.25 g) in the ratio 20 : 3 : 2 : 25 (gc determined) and polymeric substances (Scheme 8). 28 r+ Scheme 8 Toluene gave the best results for reactions with the bisaryne equivalent 33. The reaction mixture necessarily had to warm to room temperature for complete reaction but even then, these reactions were often clean. StOpping the reaction at -78°C after about 2 h or stirring led mainly to monoadduct. Even the use of three equivalents of g-BuLi at these conditions gave the monoadducts 30 and 33 as the major products in the attempted bisannelations of furan and 2,5-dimethy1furan to tetrabromo-p-xylene 33 (Scheme 9). 29 ’ 3eq . B'JLi a ’7 \ Toluene l8 4. Scheme 9 Table 3 summarizes the results of a comparative study on the influence of solvents on the yield of products which led to the identification of best solvents for bisannelations with the different tetrabromoarenes. The reaction conditions in each series differ only in the solvent. When a solvent is best for a specific bisaryne equivalent, this conclusion is valid independent of the furan used. This is illustrated by Series 1 and 2 in Table 3. 30 TABLE 3 SOLVENT EFFECT ON THE YIELDS OF BISADDUCTS Series Tetrabromoarene Furan Solvent Bisadduct % I 1 r Toluene 72 13 Ether 35 33 THE 71 I O \ 2 i5; Toluene 76 fl Ether 63 THF 68 Hr f 3 12_ Toluene Oil 8 r Ether Oil !5 , THF 87 B / O 4 r \ Ether 30 E 1.9 THF 78 31 B(iii). Effect of Reaction Temperature on Bisadduct Yield When some bisannelation reaction mixtures (especially the heterogeneous mixtures) were allowed to warm to room temperature before quenching, butylated monoadducts, butylated arenes, reduced monoadducts and reduced haloarenes were found as components of an oily fraction of the product mixture. In some cases, this oil accounted for a major fraction of the total weight of the products. For example, an 0.1 M slurry of 33 in dry ether containing two equivalents of tetramethyfuran, when reacted with three equivalents of undiluted g-BuLi at -78°C followed by allowing to warn to 25°C, 'r + BuLi/Ether ‘ -7a° - 25°C gave only 37% of 53, A large fraction of the product mixture was an oily mass. The GC-MS analysis of the oil from a similar preparation of the bisadduct 33 (cf. Table 2, Entry 9) led to the identification of the following components: 32 This result is similar to that obtained in the reaction of tetrabromo-p-difluorobenzene 33 with furan (cf. Scheme 8) when the reaction mixture was warmed to room temperature before workup. A much better yield and cleaner reaction is achieved, even with heterogeneous bisannelation reaction mixtures, if the reaction can be run effectively at -78°C. As an illustration, the bisannelation of tetrabromo-p-xylene 33 with tetramethylfuran gave a 78% yield of clean bisadduct 33 when the reaction was quenched at -78°C (the yield for the same reaction quenched at 25°C was 37%). Not all bisannela- tion reactions, however, can be run effectively at -78°C as illustrated by monoadduct formation in toluene (cf. Scheme 9) and formation of mainly 31 (cf. Scheme 8) when these reactions were run and quenched at that temperature. Most bisannelations can be effected in THF and diethyl ether at -78°C and Table 4 TABLE 4 33 YIELD OF BISADDUCTS AT DIFFERENT REACTION TEMPERATURES Yield % at Entry Bisadduct Solvent -78°C -78°C to 25°C 1 ”a“ THF 87 71 .33. 2 II I ll Ether 72 63 .12.. 3 Ether 78 4O 4 THF 56 32 Q 5 Ether 30 22 34 illustrates the advantage of quenching the reactions at this temperature in these cases. All other factors except the reaction temperature were the same for the reactions in each entry in Table 4. Some tetrabromoarenes did not give any adducts at all at -78°C. Tetrabromo-p-dimethoxybenzene 33 was totally recovered unreacted in toluene at this temperature. Bisaryne equivalents which contain additional electron- withdrawing groups on the aromatic ring such as 33 and 31 gave better yields (even though still low when compared to the yields obtained with the reactive tetrabromoarenes 33 and 33) when the reactions were run at 0°C. In these cases, the addition of g-BuLi at —78°C, followed by allowing the reaction mixture to warm to 25°C, led to intractable polymeric substances. For example, the tetrabromoarene 33, when reacted with 3-BuLi in THF at -78°C in the presence of 2,5-dimethylfuran, followed by allowing the mixture to warm to room temperature, gave only polymer. The same reaction when run at 0°C and warmed to 25°C, gave the bisadduct 33 (16% isolated yield), a mixture of 33_and the monoadducts 33 and 33 (1 g) in the ratio 2 : 2 : 1 and some polymer (Scheme 10). Presumably, the lithio-derivative of 33 accumulated at -78°C because it could not decompose at this temperature to give benzyne. As the temperature rose to 25°C, benzyne was formed and reacted with the lithioarenes faster than it reacted with the diene. 35 ‘-1a'c to as: r ’ .__, In 6 o («:29 Q \ THF Cl a) (16%) 3' \ Poly“, 4' O a + CI 03 Scheme 10 B(iv). Electronic Influence of Substituents on the Furans Whereas furan and methyl-substituted furans gave good yields with the reactive tetrabromo-p-xylene 33 in toluene, 2,5-diphenylfuran 33 gave less than 5% of the bisadduct 33 and a lot of polymeric material. About 70% of the diene 33 was recovered unchanged. This poor result exemplifies the :15 36 influence of substituents at the 2,5-positions of the furan ring on its reactivity as a diene. The reactive methyl- substituted furans trapped effectively the arynes generated from 33, resulting in the good yields, but with 33 the Ph 'flfluene lhii '75°C to 25’c 4;" (<5%) benzyne underwent other reactions faster than it could react with the diene. There is no doubt that this outcome was due to the reduced reactivity of the furan ring caused by the presence of the phenyl groups. Similar results have been observed when similarly substituted pyrroles and oxazoles were used as dienes in 26:27 In fact, in these cases, no adducts cycloadditions. were formed. Kondrat'eva and Khuan26 found that the introduc- tion of electron-withdrawing groups at the 2- and/or 5- positions of the oxazole ring did reduce the yield of adducts by more than 50% in the cycloaddition reaction of oxazoles with maleic anhydride or maleimide. They also showed that whereas oxazoles with one phenyl group at the 4-position and a methyl substituent at the 2- or S-position gave good yields of adducts, the 2,5-dipheny1 analog with no additional methyl substituents on the ring did not react. A steric 37 explanation for this failure was ruled out by the success of similar cycloadditions with the equally hindered 2-phenyl- S-ethoxyoxazole. C) n + 02" N4 R' / ° 8' / ox C) 1 X1 """ ‘~ ' I» R 1 R \ R N R" N u ' C) X'OQN-R 27 in bisannelations with The results obtained by Lai phenyl-substituted pyrroles are also consistent with the findings of Kondrat'eva and Khuan. For example, he found that both 2,S-dimethyl-3,4-diphenyl-N-methylpyrrole and 2,5- diphenyl-3,4-dimethy1-N-methy1pyrrole gave about the same yield of bisadducts with tetrabromo-p-xylene 33, but that the 2,5-dipheny1 and tetraphenyl derivatives were completely unreactive. R' = Ph; R = Ph, R' = CH3 Yield = 59% R = Ph, R' = H; R = R' = Ph Yield = 0% 38 These results were interpreted to mean that electronic factors are more important than steric factors in cyclo- addition reactions. Phenyl groups are electron-withdrawing and their effect at the 2— and/or 5-positions is to make the heterocycle more aromatic and less of a diene by inducing more delocalization of the lone pair of the hetero atom into the ring system. It appears from the work of Kondrat'eva26 and Lai27 that the presence of electron-releasing groups like methyl (by induction) and methoxyl (by resonance) compensates effectively for the delocalizing influence of an electron-withdrawing group like a phenyl substituent. C. Factors Affecting the Distribution of the Syg and Anti Isomers of Bisadducts The distribution of the syn and anti isomers of the bisadducts depended weakly, if at all, on the tetrabromo- arenes and the furans. Nor did the solvent or reaction temperature affect the syn/anti ratios. These ratios were calculated from the 250 MHz proton NMR spectra of the crude product mixtures by integrating and averaging the ratios for all the clean signals (Table 5). 39 TABLE 5 EFFECT OF SOLVENT AND REACTION TEMPERATURE ON THE RATIosa.b op SYN AND ANTI ISOMERS Series Bisadduct Solvent Temperature Ratio Toluene -78°C to 25°C 50 : 50 ”O THF -78°C to 25°C 50 : so 1 a THF -78°C 47 : 53 g? .Ether 0°C 49 : 51 Toluene -78°C to 25°C 42 : 58 Ether -78°C to 25°C 42 : 58 2 THF -78°C to 25°C 40 : 60 Toluene -78°C 44 : 56 Ether -78°C 41 : 59 12 THF '78°C 42 : 58 a . . React1ons were run at the same reagent concentrations. bIt is not known which is the major isomer. There did not seem to be any difference in the syn/ anti ratios for 2,5-dimethylfuran and tetramethylfuran bisadducts (cf. Table 5, Series 2 and Table 6, Entries 3 and 4). This may be reasonable if steric and electronic effects of only the substituents at the 2,5-positions affect the geometry of the transition state and reactivity of the furan. 40 The ratios also depended more on the furan than on the bisaryne equivalent, because very similar ratios were obtained for the same furan with different tetrabromoarenes (Table 6). The reaction conditions leading to the results in Table 6 were the same for each type of furan. 41 TABLE 6 SYN/ANTI RATIOS OF BISADDUCTS OF THE SAME FURAN WITH DIFFERENT TETRABROMOARENES Entry Tetrabromoarene Puran Isomer Ratio ' C O l 50 50 B r \ 2 l3 2 3' 3' so so 13 Br 3.9 3 36 : 64 3-3 ’ \O 12 e 4 40 : 60 q 19 I: o 42 Newman28 has examined the steric effect on isomer ratios of the adducts formed from Z-substituted furans and 3-methy1benzyne. He found that the ratios varied from 42/58 for 2-methylfuran to 36/64 for 2-(1,3-dioxolan-2-yl)- 02“ ""2 \a: .. d, «A w furan--a very small variation for such a change in steric .ll.~h,'-.Uo CChfifll. ._ “unwosa any mafia owumm No: mumsumnsm asucm wauzoa IHHS mmaaxonzm1v.~-mzmm< no ZOHHxomD NH mam auspopa Hay meHe oHHmm Ho: mumpumpsm Hpusm H.p.usouv NH mHm

Huspopa Haw osHH oHHsa Ho: mumpumpsm spasm Hp.H:ouU HH mHm .6 HHUHH HuHs mpHpHH .p HHuom HHHz mpHmH> .s Ho.NN 62 00’ as .N u o n H NH Ham H H NH HN " o H H 4H HH mm so.HHHH .gxu Hem moo NH HN H o n H OH .¢.. N eHoH> Hoaeoua Haw oeHe oHumm Ho: onHHmnnm Huucm He.ucouv HH mHm

er mole-atom of halogen, led to recovery of the endoxide 3135 at both -78°C, room temperature and at reflux in THF. 111is was not the case, however, with tungsten. Conditions 148 other than total 'replacement' of the halogens worked as well. For example, with the reagent ratios given below, the endoxide 129 was converted to the naphthalene 130 in the yields shown. Ow £33m 3" CD 120 fig; WC16: BuLi Yield % l : 6 88 l : 3 94 The reaction of the bisadduct 11_with tungsten to give hexamethylanthracene 111, its rearranged form 111 and anthracene-l,4-endoxide 111 (cf. p. 146) is an example of the cases where the use of less than the optimal ratio of 1 : 6 : 24 between the substrate, metal halide and BuLi leads to partial deoxygenation. The actual ratio for this reaction was 1 : 2 : 6 respectively. When the ratio was changed to l : 3 : 9, only the rearranged hydrocarbon 191 was isolated, in 49% yield. 149 Of the three metals used, titanium and tungsten gave better results than iron in most cases. This may be either due to reduced iron being a weaker reagent or perhaps the mechanism for deoxygenation is different. An observation that may seem consistent with the latter rationale is the fact that with reduced iron, two reactions other than deoxygenation were encountered with certain polymethylarene endoxides. Although 111 could be converted to 111_in good yield with WCl6 and TiCl3 (cf. Table 12), reduced iron reacted differently at -78°C (if the reaction mixture was allowed to warm to 25°C before the removal of reduced iron species, deoxygenation was affected in 96% yield). For example, the endoxide 111, with a substrate to reagents (FeClS/BuLi) ratio of 1 : 3 : 12 at -78°C, gave only 7% of the desired naphthalene 111 and also 7% of the reduction product 145. The major product resulted from a 1,3-hydrogen migration (144, 59%). With the more substituted substrates (Table 13), reduced iron did not give any deoxygenation product under any of the reaction conditions 150 tried. The endoxides 11_and 111 gave the product of allylic rearrangement with reduced endoxides as minor products. When the reaction of diendoxide 11_with reduced iron was allowed to warm to room temperature before work-up, an equimolar mixture of the rearrangement and reduction products 146 and 147, which could not be separated easily was formed. 151 TABLE 13 REACTION OF ADDUCTS OF TETRAMETHYLFURAN WITH REDUCED IRON AT -78°C Substrate Rearrangement Reduction Deoxygenation (%) (%) (%) 59 (1.41) 7 (.15.) 7 (114.) .12; 70 (fl) so (112) -- i 78 (14_6_) -- -- 152 Separation of the reduced metals from the reaction mixture 00¢ :91. at low temperatures gave madnly 111 and only trace of 111. The use of a mixture of syn and anti isomers of 11 led to a mixture of two isomeric rearrangement products. They were formed in the same ratio as the syn/anti ratio of the starting bisadduct mixture. The use of one isomer of 11 gave a single isomer of rearrangement product. These results indicate that this rearrangement is stereospecific. Physical and chemical evidence were used to assign the structure of the novel rearrangement product 111. The parent peak in the mass spectrum of 111_was the same as the molecular weight of the starting bisadduct (M+, 350) and the proton NMR Spectrum was very different from that of 11. The proton NMR spectrum also revealed an unusual symmetry 153 in the structure of the product. These preliminary observations implicated a molecular rearrangement. The new relationships between the protons in 111_were sorted out through double irradiation experiments performed at 180 MHz. The bridgehead and aromatic methyl groups in 111 (A', A" and B') have almost the same chemical shifts as in the starting bisadduct (A and B, Scheme 24), but the vinyl Scheme 24 methyl groupsCIin 11 had disappeared from their former position giving rise to four new types of protons, HC‘, HD, HE and HF in 111, It follows that it is this part of the molecule that has undergone rearrangement. A doublet appeared at [0.704 and 0.743 (6 H, g = 7 Hz) belonging to the protons of the C' methyl groups. A multiplet for the methine proton HD was centered at 43.556 (2 H). The terminal 154 vinyl protons HB and HF appeared as doublets at Ju.628/4.636 (2 H, 1.8 1.34 Hz) and 4.947/4.960 (2 H, 1_- 2.29 Hz). On double irradiation of HD (J 2.556), the doublet at 50.704/ 0.743 collapsed into a singlet and the doublets of HE and HF also became singlets, showing that the methine proton was situated so that it coupled with the upfield methyl groups (J‘0.704/0.743) and with the vinyl protons. Decoupl- ing of HC, simplified the multiplet at Ab.556 into an unresolved triplet and no other signals were altered by this irradiation. Double irradiation of the vinyl proton at JZ.947/4.960 defined the multiplet for H at 61.556 as a D broad quartet whereas decoupling the proton at 6&4628/4.636 resulted in HD becoming a clean quartet (1_= 7 Hz), with each of the four components split into doublets (1 = 2.3 Hz). This coupling pattern uncovered by these irradiations is consistent for a product of an allylic hydrogen shift in 11 to give 111, whose partial structure is shown in Scheme 24. This proton NMR spectral pattern was similar for the rearrangement products of 111 and 111. The structures of the 1,3-hydrogen shift products were also interrelated to the endoxides by catalytic FeCh/BuLi I - 78°C H2. Pd/C 155 hydrogenation. Thus reduction of the adduct 111 and of its rearranged form 111 gave the same product 119. Similarly, reduction of 11 and its rearranged form 111 gave 111 in each case (see p. 152). The catalytic hydrogenation of 111, 111, 111 and 111 led to one isomer of reduced product in each instance. Single isomers of 11 and its rearranged form 111 gave a single isomer of 111 also. The reduction of the adducts and their rearranged forms is therefore stereo- specific. The methine proton in the rearranged forms (111, 146 and 148) probably has an exo stereochemistry. _ 144 148 The methyl and methine protons at the 2,3-positions of the reduced endoxides 145, 147 and 149 have the same NMR pattern as meso-2,3-dibromobutane - an XSAA'Xg system - which has been discussed in some detail by Anet.67 Detailed analysis of the porton NMR spectra of these compounds would be required in order to determine the coupling constant between the methine protons. This value would, in turn, be used to determine the stereochemistry at the 2,3-positions of the reduced adducts. 156 No rearrangement similar to that described above for the reaction of adducts of tetramethylfuran with reduced iron was observed with reduced titanium and tungsten. For example, bisadduct 11, when stirred with reduced titanium at 25°C for 10 h, gave the partially deoxygenated product 111_( cf. Table 12, Entry 6). Similar reaction with reduced iron gave the product of a 1,3-hydrogen migration 111. At very long reaction times, reduced titanium and tungsten could lead to the reduction and rearrangement of the deoxygenated product. Thus, when the deoxygenation of 11 with reduced titanium was run for 36 h, the desired fiOfl-‘b—r COO product 9 was further reduced to 150 and underwent rearrange— ment to 109. Whereas, therefore, reduced titanium and tungsten do not cause allylic rearrangement and reduction of 157 the endoxides, as reduced iron does, they may cause reduction and rearrangement of the resultant hydrocarbons if such compounds remain in contact for long periods with the low- valent metal species. The presence of traces of reduced arene endoxides in the reactions with reduced iron led us to suspect that they might be intermediates in the deoxygenation process. An attempt to aromatize 142 by this procedure, however, was No Reaction unsuccessful. Surprisingly, aromatization of 9,10-dimethoxyanthra— cene diendoxides was accompanied by the simultaneous loss of the methoxyl groups to give aromatic hydrocarbons. The reduced metals were used in the same Optimal reagent to substrate ratio which was developed for the 9,10-dimethyl analogs. For example, when 11 was stirred with low-valent iron, 1,4,S,8-tetramethylanthracene 1411’66 was obtained. 158 Similar treatment of the bisadduct 11 gave the octamethyl- 2,64 (cf. Table 12). Only FeCl and TiCl 3 3 affected the methoxyl groups in this manner. Reduced anthracene 1 tungsten often gave the 9,10-dimethoxyanthracene as the ultimate product. For example, treatment of the bisadduct 11 with reduced tungsten led to an equimolar mixture of the 159 anthracene 119 and the partially deoxygenated compound 111. It appears that peri-interactions may be responsible for the replacement of the methoxyl groups with hydrogen because, with the furan bisadduct 11, only the quinone 119 could be obtained even after long reaction times (cf. Scheme 23). Attempts were made to determine how and at which stage the methoxyl groups are lost. On page 77 it was reported that it was not possible to prepare the quinone 191 of the bisadduct 11. It follows then that demethoxyla- tion cannot occur before deoxygenation of the endoxide function of the molecule. But 9,10-dimethoxyanthracene 138 was converted in 70% yield to the quinone 139 by 160 FeClS/BuLi, showing that this reagent system can effect demethylation of the 9- and 10-methoxy1 groups. It is therefore probable that for the transformation of 11 to 119, deoxygenation of the endoxide function occurred first to give the anthracene 111 which was then demethylated and oxidized to anthraquinone 119. The conversion of hindered 9,10-dimethoxyanthracene diendoxides to aromatic hydrocarbons with reduced iron and titanium might also involve initial deoxygenation of the endoxide functions to give the hindered 9,10-dimethoxy- anthracenes. These intermediates could then be converted to the hydrocarbons probably through the intermediacy of quinones and other exo-derivatives of the anthracenes. To verify whether hindered 9,lO-dimethoxyanthracenes could be converted to aromatic hydrocarbons by reduced titanium and iron, diether 111, prepared from bisadduct 11 using reduced tungsten, was treated with reduced titanium and actually led to anthracene 141 in very high yield (Scheme 25). 161 Scheme 25 An indication that quinones and other oxo-derivatives of anthracenes might be systematic intermediates in the loss of the methoxyl groups of hindered anthracene diendo- xides and anthracenes (i.e., the 9,10-methoxy1 groups) was provided by the identification of oxo-derivatives 111_and 111 in the reaction of the bisadduct 11 with reduced titanium. From these observations, the conversion of 162 9,10-dimethoxyanthracene diendoxides 11_and 11 to 163 W~~ _§_1_ 14o «— 06 fl" anthracenes 111 and 1 respectively, can be formulated as involving the steps illustrated with 11 above. All the reactions effected with these reduced transition metal Species represent new applications of these reagents. Even though deoxygenation was effected as was surmised by analogy with alkene oxides, there is not enough evidence from our results to confirm that the same mechanism operates here. Further investigation is required to ascertain the nature and oxidation state of the active metal species. In conclusion, low-valent transition metals, especially Ti, W and Fe, are effective reagents for one- 164 step aromatization of arene endoxides even when multiple peri-interactions pose serious problems with the regular methods for aromatizing such compounds or their derivatives. The adducts of 2,5-dimethy1furan and furan itself are deoxygenated in good yields by all three metals. The adducts of tetramethylfuran can be aromatized with reduced titanium and tungsten but the reaction has to be monitored more closely with respect to time to avoid partial deoxygenation or reduction of the resulting aromatic hydrocarbon. Whereas tungsten will give the 9,10-dimethoxy- anthracenes, iron and titanium will simultaneously replace the methoxyl groups with hydrogen in their reactions with the 9,10-dimethoxyanthracene diendoxides. The 1,3-hydrogen shift observed in the reaction of tetramethylfuran adducts with reduced iron is unprecedented and interesting because a more substituted double bond isomerizes to a stable terminal double bond. This process would not be expected purely on thermodynamic considerations from the simple hydrocarbon analogs. The strain between the *\ 165 C1 and C2 methyl groups does not seem to be high enough to account for this unusual rearrangement. It would be instructive to verify whether this process is also possible for the similarly substituted oxanorbornenes, oxanorborne- dienes, norbornenes and norbornadienes. :41 HM‘ EXPERIMENTAL 1. l,2,3,4,5,6,7,8-Octamethy1anthracene (1) To a slurry of 1.96 g (12.1 mmol) of anhydrous sublimed FeCl in 50 mL of anhydrous THF at -78°C under 3 argon was added 32 mL of 1.6 M BuLi. The mixture was stirred at this temperature for 2 h and 0.66 g (1.73 mmol) of the bisadduct 11, suspended in 100 mL of dry THF was run into the reaction mixture. After 90 h of Stirring, during which time the temperature rose to 25°C, the fluorescence of the product reached a maximum (tlc, alumina/hexane). The reaction mixture was filtered through a layer of alumina and the layer was washed many times with THF. The combined filtrates were evaporated to dryness on a rotary evaporator. The yellow solid was triturated with cold hexane to remove oily matter and gave 0.31 g (62%) of anthracene 2: mp 64 297-8°C); 1 290-2°C with decomposition (lit. H NMR (CDClS) 52.45 (s, 12 H), 2.73 (s, 12 H); mass Spectrum, E/E (relative intensity) 291 (M + 1, 31), 290 (M+, 100), 275 (5). 166 167 2. 9,10-Dimethy1anthracene (118) This compound was prepared according to the typical procedure for deoxygenations with FeCl3 (experiment #6). Thus 1.93 g (11.8 mmol) of FeCl 27 mL of 1.7 M BuLi and 3’ 0.47 g (2 mmol) of the anthracene diendoxide 11 gave 0.23 g 16 (56%) of 118: mp 179-180°C (lit. 180-1°C). 3. l,4,5,8,9-Pentamethy1-10-methy1ene-9,lO-dihydroanthra- cene (123) In a Similar procedure as for experiment #15, 2.44 g (6 mmol) of WCl 11 mL of 2.2 M BuLi and 0.57 g (2 mmol) 6’ of the diendoxide 11 gave 0.25 g (49%) of 123: mp 159°C- 161°C; 1 H NMR (0014) 61.27 (d, 3 H, g = 7 Hz), 2.33 (s, 6 H), 2.47 (s, 6 H), 4.23 (q, l H, 1 = 7 Hz), 5.43 (s, 2 H), 6.70 (S, 4 H); mass spectrum, E/E (relative intensity) 262 (M+, 23), 247 (100), 230 (5), 215 (9), 202 (6). Both the precursor to this compound (11) and the aromatic isomer--hexamethy1anthracene 111;-showed correct elemental analyses (see p. 88 and 173). 4. 1,4-Dimethy1-l,4-dihydronaphthalene-1,4-endoxide (129) 23’46 was prepared as in experiment The compound 129 #7 in 66% yield from benzene diazonium carboxylate and 2,5-dimethylfuran. The solvent for the annelation step was THF and prepylene oxide was omitted: bp 97°C (2.5 torr); 168 1H NMR (00013) 51.70 (s, 6 H), 6.40 (s, 2 H), 6.50-6.90 (m, 4 H); 13c NMR (c0013) 619.15, 92.45, 122.14, 128.54, 150.71, 156.79. 5. 1,4-Dimethy1naphthalene (130) To 25 mL of anhydrous THF under argon at -78°C containing 1.62 g (4.1 mmol) of WCl was added 5.6 mL of 6 2.2 M BuLi through a syringe. After 10 min., the reaction flask was removed from the bath and allowed to warm to room temperature in l h. A solution of 0.43 g (2.5 mmol) of the endoxide 119 in 10 mL of anhydrous THF was added to the reaction mixture through a syringe and the stirring was continued for 6 h. Removal of THF on a rotary evaporator left a black residue which was taken up in ether (50 mL). The blue solution was washed several times with aqueous NaOH until the organic layer was colorless. Subsequent washing of the solution with water and drying (NaZSO4) gave 63 0.37 g (94%) of 130 as an oil after evaporation to dryness: 1H NMR (cc14) 42.60 (s, 6 H), 6.90 (s, 2 H), 7.10-7.40 (m, 2 H), 7.60-7.80 (m, 2 H); mass spectrum, E/S (relative intensity) 156 (M+, 100), 141 (93), 128 (13) 115 (18); UV (g-heptane ) Ama 235 (loga = 4.05), 279 (3.8), x 290 (3.89), 300 (3.68); Ir (CC14) 3000, 1660, 1460-1400, 1030 cm’l. 169 6. Typical Deoxygenation Procedure with FeC13. Preparation of 1,4,5,8-Tetramethy1naphthalene (132) Sublimed anhydrous FeCl3 (0.98 g, 6.0 mmol) was added to 25 mL of anhydrous THF under argon at -78°C. g-Butyllithium (12 mL, 2 M) was added and the mixture was stirred for 2 h. A solution of the endoxide 11110 (0.43 g, 2.0 mmol) in 10 mL of anhydrous THF was added (syringe) and the mixture was left to warm to room temperature with stirring during 6 h. After the removal of THF on a rotary evaporator, the black residue was shaken with ether (100 mL) and the suspension was suction filtered. The ether extract was washed with water, dried (NaZSO4) and evaporated to dryness. The solid residue was chromatographed on alumina with chloroform. The product was recrystalized from CHCl3 to give 0.25 g (63%) of 111 as thin needles: mp 129-131°C (lit.10 131-2°C). 7. 1,21114—Tetramethy1-1,4-dihydronaphthalene-l,4-endoxide (133) Anthranilic acid (11.4 g, 8.3 mmol) was suspended in 150 mL of anhydrous THF containingzacatalytic amount (0.18 g) of trichloroacetic acid. The reaction mixture was cooled to 0°C in an ice-salt bath and isoamyl nitrite (19 mL) was added at such a rate that the reaction tempera- ture remained below 18°C. Stirring was continued until the 170 initially formed brick-red precipitate turned tan white. The precipitated diazonium carboxylate was filtered and washed with THF. To a suSpension of the diazonium carboxylate in dichloroethane was added 7.97 g (64 mmol) of tetramethyl- furan, followed by 23 mL of propylene oxide. The mixture was brought to gentle reflux on a water bath. When gas evolution ceased, the reaction mixture was cooled to room temperature, washed many times with aqueous NaOH and finally with water until the washings had a light red color. The solution was dried (NaZSO4) and concentrated on a rotary evaporator to give a red oil. Vacuum distillation (89-95°C, 1.3 Torr) gave 4.5 g (35%) of 133: mp 41-3°C (lit.46 1 45.6-46°C); H NMR (coc13) 61.57 (s, 6 H), 1.60 (s, 6 H), 6.70 (m, 4 H); mass Spectrum, 1]; (relative intensity) 200 + (M , 90), 185 (100), 159 (89), 142 (42), 128 (23), 115 (29). 8. Typical Procedure for Deoxygenation with WC16. Prepara- tion ofil,2,3,4-Tetramethy1naphthalene (134) To a suspension of 2.45 g (6.2 mmol) of WCl6 in 25 mL of anhydrous THF under argon at -78°C was added 10 mL of 2.4 M BuLi. The mixture was stirred for l h and then allowed to warm to 25°C in l h. A solution of the endoxide 111 (0.41 g, 2.0 mmol) in 10 mL of anhydrous THF was added over 10 min. and the mixture was stirred for 5 h. The 171 solvent was removed on a rotary evaporator and the residue was extracted with ether (100 mL). The blue ether solution was washed several times with aqueous NaOH until the blue color discharged, then the organic layer was washed with water and dried (NaZSO4). The solution was concentrated until crystals of 111 deposited; the mixture was cooled and the crystals of 111 were collected by suction filtration as long white needles (0.34 g, 89%; mp 106-7°C, 11t.°4 lO6.S-107.5°C). 9. l,2,3,4,5,8-Hexamethy1-l,4-dihydrongphthalene-l,4- endoxide (135) 3,6-Dimethy1anthrani1ic acid (8.4 g, 5.1 mmol) suspended in 140 mL of absolute ethanol was cooled in an ice-salt bath and 6 mL of concentrated HCl was added. Isoamyl nitrite (14 mL) was added to the reaction mixture at such a rate that the reaction temperature remained below -S°C. After 1 h of stirring, 140 mL of ether was added and stirring was continued below 10°C for another 1 h during which time the diazonium carboxylate dropped from the solution as a yellow solid. The solid was isolated by filtration and suspended in 500 mL of dichloroethane. Tetramethylfuran (6.18 g, 50 mmol) and 21 mL of propylene oxide were added and the mixture was refluxed on a water bath until the evolution of gas subsided greatly. After cooling to 25°C, the dark brown 172 reaction mixture was washed several times with aqueous NaOH and water and then dried (NaZSO4). Evaporation of the solution to dryness on a rotary evaporator left a solid which, after recrystalization from methanol, gave 4.2 g 1 (36%) of 13s as needles: mp 101-3°C; H NMR (0014) 61.50 (S, 6 H), 1.70 (S, 6 H), 2.20 (S, 6 H), 6.30 (s, 2 H); 1; mass Spectrum, Q/g IR (052) 3000, 1380, 1150, 800 cm' (relative intensity) 228 (M+, 6), 185 (100), 174 (42), 155 (19). A derivative of this compound (148, see p.185) showed correct elemental analysis. 10. Attempted Deoxygenation of Endoxide 135 with a l : 2 Ratio of FeCl3 to g-BuLi A slurry of anhydrous sublimed FeCl3 (0.49 g, 3.01 mmol) in 20 mL of anhydrous THF cooled to ~78°C under argon was stirred while 3 mL of 2.1 M BuLi was added through a syringe. After 1 h at this temperature, a solution of the endoxide 111 (0.23 g, 1.01 mmol) in 20 mL of anhydrous THF was added and stirring was continued for 4 h more at -78°C. The solvent was evaporated on a rotary evaporator and the black residue was shaken with 50 mL of ether. The ether solution was washed with water and dried (NaZSO4). Evapora- tion to dryness gave 0.23 g (100%) of the starting endoxide 13 and no other substance was isolated. 173 11. Typical Deoxygenation Procedure with TiClS. Preparation of l,4,5,8,9,TO-Hexamethylanthracene (136) To a stirred suspension of 1.3 g (8.4 mmol) of TiCl3 in 25 mL of anhydrous ether under argon at -78°C was added 16 mL of 2 M BuLi in 10 min. After 1 h at this temperature, the mixture was warmed to room temperature, a suspension of the diendoxide 11 (0.42 g, 1.4 mmol) in 50 mL of anhydrous ether was added in 10 min and the reaction mixture was stirred for 9 h. Filtration of the black mixture through a bed of alumina and concentration of the ether solution gave 0.29 g (79%) of the anthracene 111: mp 196-8°C (p-hexane); 1H NMR (0014) 82.70 (s, 12 H), 2.83 (s, 6 H), 6.87 (s, 4 H); mass spectrum, 1]; (relative intensity) 262 (M+, 100), 247 (88), 232 (19); UV (g-heptane) Amax 425 (logE = 3.71), 404 (3.73), 276 (4.76), 224 (3.90). 1111. Calc'd. for CZOHZZ: C, 91.55; H, 8.45. Found: C, 91.71; H, 8.39. 12. Decamethyl-l,4-dihydroanthracene-1,4-endoxide (137) To a suspension of TiCl3 (1.26 g, 8.2 mmol) in 25 mL of anhydrous ether under argon at -78°C was added 16 mL of 2 M BuLi over 5 min. The mixture was stirred for l h and then allowed to warm to 25°C in l h. A suspension of the adduct 11 (0.48 g, 1.4 mmol) in 75 mL of anhydrous ether was added and the mixture was stirred for 10 h. Filtration 174 of the reaction mixture through a layer of alumina, removal of the solvent on a rotary evaporator and trituration of the residue with hexane gave 0.19 g of a white solid which was identified by its proton NMR spectrum to be unreacted bisadduct 11. The hexane solution was concentrated to give a yellow tinted solid which after recrystalization from hexane, weighed 0.23 g (50% yield of 111): mp 148-150°C; 1H NMR (0014) 51.63 (s, 6 H), 1.84 (s, 6 H), 2.27 (s, 6 H), 2.42 (s, 6 H), 2.45 (s, 6 H); mass spectrum, fl/g (relative intensity) 334 (M+, 11), 291 (100), 276 (11), 261 (11); high resolution mass spectrum calculated for C24H300: 334.22980. Found: 334.22967. 13. 9,10-Dimethoxyanthracene (138) Tungsten hexachloride (1.62 g, 4.1 mmol) was added to 25 mL of anhydrous THF at -78°C under argon. g-Butyl- lithium (6 mL, 2.2 M) was added and the stirred mixture was maintained at this temperature for 2h. The addition of 0.51 g (2.0 mmol) of the bisadduct 11, dissolved in 50 mL of anhydrous THF took 15 min. and the reaction mixture was left to warm to 25°C in 6 h. Filtration through a layer of alumina and evaporation of THF on a rotary evaporator gave a solid which was chromatographed on alumina with hexane. Final elution of the column with THF gave 0.40 g (89%) of 65 138 as yellow granular crystals: mp 195-7°C (lit. 202°C). 17S 14. Mixture of the Deoxygenated Products of the Bisadduct 11 (9, 108 and 150) To a suspension of 2.21 g (14 mmol) of TiCl3 in 25 mL of anhydrous THF at -78°C under argon was added 24 mL of 2.4 M BuLi over 10 min. The reaction mixture was stirred for 1 h at -78°C and then allowed to warm to 25°C in 2 h. A solution of the bisadduct 11 (0.82 g, 2.34 mmol) in 75 mL of anhydrous THF was added in 15 min. and the reaction was followed by tlc (alumina/hexane) until the yellow spot, probably of decamethylanthracene 9, predominated (36 h). The reaction mixture was filtered through a layer of alumina, the greenish-yellow solution was concentrated on a rotary evaporator and the residue was redissolved in 500 mL of hexane. The hexane solution was washed with water and dried (NaZSO4). After removing the solvent, the yellow solid residue was chromatographed on alumina using hexane. Evaporation of the eluate to dryness gave 0.6 g of a yellow solid. The proton NMR spectrum of this solid showed that it was a mixture of 9, 111 and 111. The peaks at 61.60 and 2.67 (ratio 1 : 2) correspond to the 9,10- and 1,4,5,8- methyl protons of decamethylanthracene. The singlet at 6’ 5.17, the quartet centered at £4.30 and the doublet at 6‘ 1.30 are characteristic of the rearranged form 111 of decamethylanthracene. The signals at J1.30 and 1.30 are also characteristic of the 9,10-dihydrodecamethylanthracene 150. On standing in the CC14 solution, the yellow color 176 disappeared totally in l h and the proton NMR Spectrum of this solution showed that the peaks at 61.60 and 2.67 had moved from this position and merged with the methyl peaks originally at 62-2.5. The vinyl peak at J3.17 had intensified as did also the doublet at 61.30. The loss of color and the intensification of the peaks at 65.17 and 1.30 are characteristic of the rearrangement of 9 to 111. Mass spectrum of the mixture, fl/E (relative intensity) 318 (M+, 14), 303 (100), 290 (28), 151 (25), 145 (33), 91 (29), 57 (32); by chemical ionization (M + l) 319. Based on the parent peak in the mass spectrum of the mixture being 318, the total yield of deoxygenation product is 81% (0.6 g). It was not possible to separate the components by chromato- graphy. 15. 1,4,5,8,9,10-Hexamethy1-l,4-dihydroanthracene-l,1- endoxide (143) A suspension of WCl6 (1.60 g, 4 mmol) in 25 mL of dry THF at -78°C under argon was stirred with 5.6 mL of 2.2 M BuLi for 10 min. and the mixture was allowed to warm to 25°C in 1 h. A solution of the bisadduct 11 (0.55 g, 1.9 mmol) in 50 mL of anhydrous THF was added and the reaction mixture was stirred for 3 h more. Evaporation of the solvent left a dark residue which was extracted with ether. The extract was washed successively with aqueous NaOH and water and dried (NazSO4). The dark yellow solid left after the 177 removal of ether on a rotary evaporator was chromatographed on alumina first with hexane to give 0.08 g (16%) of a l : 4 mixture of hexamethylanthracene 111 and its rearranged form 111 respectively. Further elution of the column with THF gave 0.24 g (46%) of 111 as a white solid: mp 155-7°C; 1H NMR (0014) 81.92 (s, 6 H), 2.52 (s, 6 H), 2.63 (s, 6 H), 6.47 (s, 2 H), 6.77 (s, 2 H); mass spectrum, m/g (relative intensity) 278 (M+, 10), 252 (22), 236 (20), 235 (100), 220 (21), 205 (12); high resolution mass spectrum calculated for CZOHZZO: 278.16712. Found: 278.16707. l6. Deoxygenation of Bisadduct 11 with FeCls. Preparation 6f Anthraquinone (139) Anhydrous THF (50 mL) containing 2 g (12 mmol) of FeCl3 was cooled to -78°C under argon and 25 mL of 2 M BuLi was added. The mixture was stirred for 2 h and 0.56 g (2.06 mmol) of the bisadduct 11 dissolved in 100 mL of dry THF was run into the reaction mixture in 15 min. After 48 h, during which time the temperature of the mixture rose to 25°C, the reaction mixture was filtered through a layer of alumina. The filtrate was evaporated to give 0.24 g (55%) of anthraquinone 119 as a yellow powder. The melting point and proton NMR spectrum of this product were identical with those of an authentic sample. 178 When a similar reaction was worked up 3 h after the addition of the substrate 11, 9,10-dimethoxyanthracene 138 was obtained in 20%. 17. Conversion of 9,10-Dimethoxyanthracene to Anthraquinone (129) To a suspension of 1.96 g (12 mmol) of sublimed anhydrous FeCl3 in 50 mL of anhydrous THF at -78°C under argon was added 22 mL of 2.2 M BuLi in 5 min. Stirring was continued for 2 h and the mixture was allowed to warm to 25°C in l h. The anthracene 111 (0.50 g, 2.1 mmol) dissolved in 100 mL of anhydrous THF was added over 20 min and the reaction mixture was gently heated to reflux. The characteristic fluorescence of 111 disappeared after 20 min of reflux but the reaction was continued for 9 h. The black suspension was filtered through a layer of alumina, THF was removed on a rotary evaporator and the residue was redissolved in CH2C12. The solution was washed with water and dried (NaZSO4). Evaporation to dryness and trituration of the residual faint yellow needles with hexane gave 0.31 g (70%) of 139. 18. 1,4,5,8-Tetramethy1-9,10-dimethoxyanthracene (140) A suspension of tungsten hexachloride (1.37 g, 3.5 mmol) in 25 mL of anhydrous THF at -78°C under argon was 179 treated with 6 mL of 2.4 M BuLi; the mixture was stirred at this temperature for l h and then allowed to warm to 25°C in l h. The bisadduct 11 (0.16 g, 0.5 mmol) dissolved in 5 mL of anhydrous THF was added and the mixture was stirred for 24 h more. fHuesolvent was removed on a rotary evaporator and the residue was extracted with ether. The extract was washed successively with 40% aqueous NaOH and water and dried (NaZSO4). Removal of the solvent left a yellow oil (0.149 g) which was chromatographed on alumina with hexane to give 0.069 g (47%) of the anthracene 111 as yellow shiny 1H NMR (00013) 62.80 (s, 12 H), plates: mp 153-5°C (g-hexane); 3.45 (S, 6 H), 6.88 (s, 4 H); mass spectrum, fl/E (relative intensity) 294 (M+, 31), 279 (100), 263 (9), 147 (35). 1111. Calc'd. for CZOHZZOZ: C, 81.60; H, 7.53. Found: C, 81.52; H, 7.70. Further elution of the column with chloroform gave 0.072 g (47%) of 1,4,5,8-tetramethy1-9,lO-dimethoxy-l,4- dihydroanthracene-l,4-endoxide (111): mp 125-7°C (CHC13); 1H NMR (CDC13) 62.05 (s, 2 H), 2.70 (s, 6 H), 3.53 (s, 6 H), 6.63 (s, 2 H), 6.92 (s, 2 H); mass spectrum, m/g (relative intensity) 310 (M*, 31), 267 (100), 237 (22), 165 (20); high resolution mass Spectrum calculated for C20H2203: 310.15708. Found: 310.15690. 180 19. Demethoxylation of anthracene 140. Preparation of 1:1,5,8-Tetramethy1anthracene (141) To a suspension of TiCl3 (0.50 g, 3.2 mmol) in 10 mL of anhydrous THF at -78°C under argon was added 7 mL of 1.9 M BuLi. The mixture was stirred for l h at -78°C, left to warm to 25°C in 2 h and a solution of the anthracene 111 (0.10 g, 0.30 mmol) in 75 mL of anhydrous THF was added. After 16 h, the reaction mixture was filtered over a layer of alumina and the solvent was removed on a rotary evaporator. The residue was chromatographed on alumina (hexane eluent) to give 0.07 g (93%) of the anthracene 111_which was recrystalized from a 1 : 1 mixture of chloroform and 1- heptane: mp 220-2°C (lit.66 221-1.5°C); 1H NMR (0014) J‘ 2.37 (s, 12 H), 6.93 (s, 4 H), 8.30 (s, 2 H); mass spectrum, 1/9 (relative intensity) 234 (M+, 100), 219 (21), 203 (10), 189 (5), 117 (9), 102 (8). 20. Direct Preparation of anthracene 141 from Bisadduct 11 Sublimed anhydrous FeCl3 (4.06 g, 25 mmol) was suspended in 50 mL of anhydrous THF at -78°C under argon. g-Butyllithium (50 mL, 2 M) was added in 20 min. and the mixture was stirred for 2 h. A solution of 1.41 g (4.3 mmol) of bisadduct 11 in 100 mL of anhydrous THF was added in 0.5 h and the mixture was stirred for 18 h during which time the temperature rose to 25°C. Filtration was carried 181 out over a layer of alumina which was washed with THE until the greenish-yellow color of the product was no longer noticeable in the filtrate. Concentration of the combined filtrate left a solid which was redissolved in CHZClz. The solution was washed with water and dried (NaZSO4). A brown solid residue was left on evaporation of the solution to dryness. The residue was shaken with hexane and the extract was passed down a column of alumina. The eluate was evaporated to dryness on a rotary evaporator and the product was recrystalized from hexane to give 0.61 g (60%) of the anthracene 141 as pale yellow crystals: mp ZZZ-3°C. 21. Attempted Deoxygenation of the Endoxide 133 with Reduced IFOn. Preparation of 1,3,44Trimethyl-2Fmefhy1ene-1,2, 3,4-tetrahydronaphthalene-l,4-endox1de (144) To a slurry of 0.90 g (5.6 mmol) of anhydrous sublimed FeCl3 in 25 mL of anhydrous ether at -78°C under argon was added 10 mL of 2.3 M BuLi. After 2 h of stirring at this temperature, a solution of the endoxide 111 (0.38 g, 2.0 mmol) in 10 mL of dry ether was added and the reaction was left for 1 h at -78°C. fUuacold reaction mixture was suction filtered over a layer of alumina and the filtrate was washed with water and dried (NaZSO4). Removal of the solvent on a rotary evaporator left 0.28 g of an oil whose proton NMR spectrum showed the presence of l,2,3,4-tetra- methylnaphthalene 134, the rearranged form of the adduct 182 111 (111) and the reduced derivative of the adduct 111 in the ratio 1 : 8 : 1 respectively. Chromatography of this mixture on alumina with a l : 9 mixture of hexane and chloroform led to the separation of 111 from the other two 64 components: mp of 134; 106-7°C (lit. 106.5-7.5°C). A sample of the pure rearranged adduct 144 was obtained as a 1H NMR liquid by glc at 200°C from its mixture with 111. (c014).Jb.61 (d, 3 H, g = 7 Hz), 1.63 (s, 3 H), 1.70 (s, 3 H), 2.50 (m, 1 H), 4.47 (d, 1 H, 1 = 2.3 Hz), 4.77 (d, l H, 1 = 2.3 Hz), 6.88 (br s, 4 H); mass spectrum, m/g (relative intensity) 200 (M+, 23), 185 (10), 157 (72), 146 (100), 129 (23), 115 (36), 103 (16); high resolution mass spectrum calculated for C14H160: Found: 22. Hydrogenation of 133 and 144. Preparation of l,2,3,4- Tetramethyl-l,2,3,4—tetrahydronaphthaleneJT,4-endoxide (145) A solution of the endoxide 111 (0.35 g, 18 mmol) in 50 mL of absolute ethanol was hydrogenated at 70 psi of hydrogen for 6 h at room temperature over 0.3 g of 10% palladium on charcoal. The catalyst was removed by filtra- tion, the solvent was removed on a rotary evaporator and the residue was purified by chromatography on alumina by first using hexane followed by chloroform as eluents. The 1 liquid product 145 weighed 0.34 g (96%): H NMR (CDC13) d, 183 0.32 (d, 6 H, g = 7 Hz), 1.57 (s, 6 H), 2.07 (m, 2 H), 6.87 (br s, 4 H); mass spectrum, m/g (relative intensity) 146 (100), 131 (38), 115 (17); with chemical ionization, (M + 1)+ = 203. Anal. Calc'd. for C14H180: C, 83.12; H, 8.97. Found: C, 83.01; H, 9.07. The proton NMR spectrum of 14; obtained from the deoxygenation of 133 with reduced iron (vide supra) was identical with that obtained from the hydrogenation of 133. Hydrogenation of a mixture of 133 and 144 (l : 1) obtained in a deoxygenation reaction of 133 gave 145 in 93% yield. 23. Attempted Deoxygenation of Diendoxide 4§_with Reduced Iron. Preparation of 146 To 25 mL of anhydrous THF at -78°C under argon was added 0.93 g (5.7 mmol) of anhydrous sublimed FeCl followed 3 by 10 mL of 2.1 M BuLi. After 2 h of stirring at this temperature, a solution of the bisadduct 43 (0.32 g, 0.9 mmol) in 25 mL of dry THF was added in 15 min. and the reaction mixture was stirred at -78°C for 4 h. Filtration and evaporation of the solvent left a residue which was taken up in CHZClz, washed with water and dried (Na2804). The yellow solid residue obtained after removal of the solvent was chromatographed on alumina and first eluted by hexane to remove the yellow coloring impurity. Further 184 elution with a 9 : 1 mixture of CHZCl2 and hexane gave 0.25 g (78%) of 116: mp 233-s°c (ethanol); 1H NMR (cnc13) J‘ 0.72 (d, 6 H, g? 7 Hz), 1.86 (s, 6 H), 1.89 (s, 6 H), 2.33 (s, 6 H), 2.56 (m, 2 H), 4.63 (d, 2 H, g = 2.3 Hz), 4.95 (d, 2 H, 1 = 2.3 Hz); mass spectrum, mfg (relative intensity) 350 (M+, 63), 306 (48), 296 (77), 253 (100). The hydro- genation derivative of 146, which is described in experiment #24, and the starting bisadduct 43 gave correct elemental analyses. 24. Hydrogenation of 4§_and 146. Preparation of Deca- methyl-octahydroanthracene-l,4:S,8-diendoxide (147) The bisadduct 43 (0.33 g, 0.9 mmol) dissolved in 40 mL of absolute ethanol was hydrogenated at atmOSpheric pressure and room temperature in 12 h over 0.3 g of 10% palladium on charcoal to give 0.31 g (93%) of l,2,3,4,5,8,9, lO-decamethy1-1,2,3,4,5,6,7,8-octahydroanthracene-1,4:5,8- diendoxide 147: mp 251-2°C (ethanol); 1 H NMR (cc14) J‘ 0.43 (d, 12 H, g = 7 Hz), 1.67 (s, 12 H), 1.80-2.30 (m, 4 H), 2.20 (s, 6 H); mass spectrum, m/g (relative intensity) 298 (31), 242 (100), 122 (17); with chemical ionization, (M + 1)+ = 355. Anal. Calc'd. for C24H3402: C, 81.31; H, 9.67. Found: C, 81.45; H, 9.61. Similar hydrogenation of 146 gave the same product (147) in 98% yield. 185 25. Reaction of the Endoxide 135 with Low-Valent Iron Preparation of 1,3,4,5,8-Pentamethy1-Z-methylene— 1,2,3:4-tetrahydronaphthalene-1,4fefidoxide (148) To 25 mL of anhydrous ether at -78°C under argon was added 1.44 g (8.9 mmol) of anhydrous sublimed FeC13, followed by 16 mL of 2.2 M BuLi. After 2 h of stirring at this temperature, a solution of the endoxide 1§§_(0.44 g, 2 mmol) in 20 mL of anhydrous ether was added and the reaction mixture was maintained at -78°C for 4 h. The cold mixture was filtered, the filtrate was washed with water, dried (NaZSO4) and evaporated to dryness on a rotary evaporator. The viscous oil (0.44 g) solidified after standing for several hours. The proton NMR spectrum of the product showed that it consisted of a 7 : 3 mixture of the rearranged adduct 148 and the reduced derivative 149 (i.e., 70% of 148 and 30% of 142). Pure 148 was obtained from this 1 mixture by glc at 210°C: mp ss-7°c; H NMR (cc14) 60.66 (d, 3 H, 1 = 7 Hz), 1.76 (s, 3 H), 1.77 (s, 3 H), 2.27-2.28 (m, 6 H), 2.43-2.49 (m, 1 H), 4.53 (d, 1 H, i = 2.3 Hz), 13 4.89 (d, 1 H, g = 2.3 Hz), 6.67 (s, 2 H); c NMR (CDC1 3) 312.97, 14.87, 16.42, 45.28, 84.42, 84.88, 99.13, 124.22, 124.43, 126.62, 127.03, 127.65, 140.63, 141.80, 155.88; mass spectrum, m/g (relative intensity) 228 (32), 185 (42), 174 (100). Anal. Calc'd. for C16H200: C, 84.16; H, 8.83. Found: C, 84.24; H, 9.00.. 186 26. Hydrogenation of 135 and 148. Preparation of 1,2,3,4, 5,8-Hexamethy1-1,2,3,4-tetrahydronaphthalene-1,4- endoxide (149) A solution of the endoxide 135 (0.21 g, 0.9 mmol) in 50 mL of absolute ethanol was hydrogenated at atmospheric pressure and room temperature over 0.3 g of 10% palladium on charcoal. Filtration of the reaction mixture and evaporation of the solution on a rotary evaporator gave an oil which was chromatographed on alumina (3 : 7 :: hexane chloroform). The yield of 149 was 0.21 g (99%): mp 45-7°C; 1H NMR (cc14) 30.43 (d, 6 H, g = 7 Hz), 1.72 (s, 6 H), 13C NMR 2.06-2.18 (m, 2 H), 2.27 (s, 6 H), 6.57 (s, 2 H); (CDC13) $8.93, 16.62, 17.18, 42.38, 85.72, 126.31, 126.81, 141.71; mass spectrum, m/e (relative intensity) 174 (100), 159 (9); with chemical ionization, (M + 1)+ = 231. Anal. Calc'd. for C16H220: C, 83.43; H, 9.63. Found: C, 83.26; H, 9.55. Similar hydrogenation of 14§_gave a quantitative yield of 149. Also, 142_obtained from the deoxygenation of 135 with low-valent iron (vide supra) had the same proton NMR spectrum as the hydrogenation product. 27. Deoxygenation of Bisadduct 52 with Low-valent Titanium. The Interception of 152 and 153 Using the typical procedure for deoxygenation with TiCl3 (experiment #11), 2.14 g (13.88 mmol) of TiCl 23 mL 3’ 187 of 2.4 M BuLi and 0.83 g (2.17 mmol) of bisadduct 22 were reacted in 72 h to give a mixture of 2, 222 and 22_ (0.3 g). It was difficult to separate the components effectively but small quantities of pure 2 and 222_were obtained by chromatography on alumina using 80% hexane in CHZCIZ. 2: mp 29o-2°c (lit.64 1 297-8°C); H NMR (CDC13) 32.45 (s, 12 H), 2.73 (s, 12 H), 8.53 (s, 2 H); mass spectrum, m/g (relative intensity) 291 (M + 1, 31), 290 (M+, 100), 275 (5). 2 152: This compound was contaminated with 153 and was not obtained in pure form for proton NMR spectrum. It was possible to record its mass spectrum because compound 222 with which it was contaminated did not show any peaks on electron impact mass spectral determination: mass spectrum for 222, m/g (relative intensity) 307 (M + l, 12), 306 + (M , 48), 291 (100), 233 (9), 138 (7). 153: mp — sublimed completely at 280-5°C (lit.2 1 303°C); H NMR (CDC13) 32.23 (s, 12 H), 2.47 (s, 12 H); mass spectrum - there were no peaks observed by electron impact on a pure sample of l 3. Chemical ionization showed (M + l)+ = 321. BIBLIOGRAPHY 10. 11. 12. 13. BIBLIOGRAPHY Caspar, M.L., Stothers, J.B., Wilson, N.K., Can. J. Chem., 22, 1960 (1975). Backer, H.J., Strating, J., Huisman, L.H.H., Rec. Trav. Chim., §§, 761 (1939). Hart, H., Jiang, J.B., Gupta, R.K., Tetrahedron Lett., 4639 (1975). Fields, D.L., J. Org. Chem., 22, 3002 (1971). Hehre, W.J., Pollack, S.K., Tetrahedron Lett., 2483 (1980). Hart, H., Oku, A., J. Org. Chem., 22, 4269 (1972). Gribble, G.W., Allen, R.W., Anderson, P.S., Christy, M.E., Colton, C.D., Tetrahedron Lett., 3673 (1976). Hart, H., Teuerstein, A., Synthesis, 693 (1979). Hart, H., Ruge, B., Tetrahedron Lett., 3143 (1977). Hart, H., Sy, A., J. Org. Chem., 22, 7 (1979). Lai, C-Y., Ph.D. Thesis, Michigan State University, 1981, pp. 34, 36, 38, 42. Wittig, G., Knauss, E., Niethammer, K., Liebig's Ann. Chem., 630, 11 (1960). Gribble, G.W., Allen, R.W., LeHoullier, G.S., Eaton, J.T., Easton, N.R., Slayton, R.I., Sibi, M.P., J. Org. Chem., 22, 1025 (1981). 189 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 190 Beringer, F.M., Huang, S.J., ibid., 22, 447 (1964). Cragg, G.M.L., Giles, R.G.F., Roos, G.H.P., J. Chem. Soc., Perkins I, 1339 (1975). Wittig, G., Harle, H., Liebig's Ann. Chem., 23, 17 (1959). Hennion, G.F., Anderson, J.G., J. Am. Chem. Soc., 68, 424 (1946). Hugel, G., Lerer, M., Fabre, C., Bull. Soc. Chim. France, 173 (1957). Kohn, M., Weisberg, M., Monatsch., 22, 295 (1924); Plant, J., Baddeley, G., J. Chem. Soc., 527 (1943). Deacon, G.B., Farquharson, G.J., Aust. J. Chem., 22, 629 (1976). Kohn, M., Grun, S., Monatsch., 22, 663 (1925). Wittig, G., Durr, H., Liebig's Ann. Chem., 672, 55 (1964). Hart, H., Sasaoka, M., J. Org, Chem., 32, 368 (1979); Wolthuis, E., Bossenbroeck, B., DeWall, G., Geels, E., Leegwater, A., ibid., 22, 148 (1963). Yakobson, G.G., Mironova, N.R., Petrov, A.K., Vorozhtsov, N.N., Jr., Zh. Obschch. Khim., 36 (1), 147, (1966). Whiting, M.C., Ann. Rept. Progr. Chem. (Chem. Soc. London), 22, 189 (1954). 'Kondrat'eva, G. Ya., Chzhi-Khuan, K., Doklady Akad. Nauk. SSSR., 142, 593 (1962). Lai, C-Y. Ph.D. Thesis, Michigan State University, 1981, p. 22. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 191 Newman, M.S., Kannan, R., J. Org. Chem., 22, 3356 (1976). Fields, E.K., Meyerson, 8., Adv. Phys. Org. Chem., 2, 18 (1968). Wittig, G., Merkle, W., Chem. Ber., 22, 1491 (1942). Gilman, H., Gorsich, R.D., J. Am. Chem. Soc., 22, 3919 (1955). Hart, H., Lai, C-Y., Nwokogu, G., Shamouillian, S., Teuerstein, A., Zlotogorski, Ch., J. Am. Chem. Soc., 222, 6649 (1980). Hoffman, R.W., Dehydrobenzene and Cycloaklynes, Academic Press, New York, 1967, pp. 49-53 and references therein. Varma, P.S., Shankarnarayanan, S., J. Ind. Chem. Soc., 22, 31 (1936). Beacall, T., Trans. Faraday Soc., 22, 472 (1945). Finger, G.C., Reed, F.H., Burness, D.M., Fort, D.M., Blough, R.R., J. Am. Chem. Soc., 22, 145 (1951). Gilman, H., Haiduc, 1., J. Organomet. Chem., 22, 394 (1968). Harper, R.J., Jr., Soloski, E.J., Tamborski, C., J. Org. Chem., 22, 2385 (1964). Tamborski, C., Soloski, H.J., J. Organomet. Chem., 22, 245 (1969). Gilman, H., Haiduc, 1., ibid., 22, P4 (1968). Gilman, H., Dua, S.S., ibid., 64, C1 (1974). 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. S4. 55. 56. 57. 192 Gilman, H., Dua, S.S., Howells, R.D., J. Fluorine Chem., 2, 381 (1974). Zinke, Th., Wiederhold, K., Liebigfs Ann. Chem., 320, 206 (1902). Silveira, A., Jr., Bretherick, H.D., Jr., Negishi, E-i., J. Chem. Ed., 22, 560 (1979). Wenkert, E., Stevens, T., J. Am. Chem. Soc., 22, 5627 (1956). Wolthuis, E., J. Org. Chem., 22, 2215 (1961). Caple, R., Chen, G. M-S., Nelson, J.D., J. Org. Chem., 22, 2874 (1971). Polovsky, S.B., Franck, R.W., ibid., 22, 3010 (1974). Kauffman, T., Udluft, K., Angew. Chem. (Int. Ed. Engl.), 2, 45 (1936). Hoffman, R., Woodward, R.B., Acc. Chem. Res., 2, 17 (1968). Ziegler, G.R., Hammond, G.S., J. Am. Chem. Soc., 22, 514 (1968). DeWit, J., Wynberg, H., Rec. Trav. Chim., 22, 281 (1973). Jiang, J.B., Ph. D. Thesis, Michigan State University, 1975, p. 36. Keinan, E., Mazur, Y., J. Org, Chem., 22, 1020 (1978). Cava, M.P., van Meter, J.P., ibid., 22, 538 (1969). Smith, C.A., Grotznec, J.B., ibid., 32, 367 (1976). Olah, G.A., Narang, S.C., Gupta, B.G.B., Malhotra, R., ibid., 22, 1247 (1979). 58. 59. 60. 61. 62. 63. 64. 65. 66. 67. 193 McMurry, J.B., Fleming, M.P., J. Org. Chem., 22, 2555 (1975); McMurry, J.B., Silvestri, M.G., Fleming, M.P., Hoz, T., Grayston, M.W., ibid., 22, 3249 (1978). Kochi, J.K., Singleton, D.M., Andrews, L.J., Tetrahedron, 24, 3503 (1968). Fujisawa, T., Sugimoto, K., Ohta, H., Chem. Lett., 883 (1974). Sharpless, R.B., Umbriet, M.A., Nieh, M.T., Flood, T.C., J. Am. Chem. Soc., 22, 6538 (1972); Sharpless, K.B., Teranishi, A.Y., Backvall, J-E., ibid., 22, 3120 (1977). Fujiwara, Y., Ishikawa, R., Akiyama, F., Teranishi,S., J. Org. Chem., 4;, 2477 (1978). Riemschneider, R., Nolde, K., Henning, K., Monatsch., 222, 987 (1973). Krysin, A.P., Bodoev, N.V., Koptyug, V.A., Zh. Org. 22222, 22, 1290 (1977); Engl. Transl., 22, 1183 (1977). Meyer, K.H., Liebig's Ann. Chem., 379, 37 (1911). Carruthers, W., Stewart, H.N.M., Hansell, P.G., Kelly, K.M., J. Chem. Soc., Perkins I, 2607 (1967). Anet, F.A.L., J. Am. Chem. Soc., 22, 747 (1962).