99 "1799:? , , 99'9' 599:9, 939’ 1.9" Vi'h'b ”at, $9 ,9, .1919; T] '99 ‘95,“ ”9““ 9999“,,9,,92" ““9999. ‘ 9’99 ‘ 1,9, g; .‘L& 31'9”“ Si‘; 3! \J‘ "- 9L 1...? ‘R‘Efiw my“, :{L w9;:::9i.9 a“§ \ :§L@ HA9P XV"? -.'.-999 . gf ”$150... :4ng 9 9 . 99 «99mm 9999‘ *999999993 9 '99 9,99 9.999" "”99 N9 ‘1; t ‘ “\6’9‘ 4 9 4:“ . . 9.9,, $3”9,;9;§9g‘9;€9 W99» 4124,;9 ', 9 9 99!}. 39999999 ‘ 1399999191999“: 9‘99 999 999999999 " “9‘19 919 ”‘9' “99,3953! $9919: 9999,99 9‘99 9,9 '99’99‘ 5 999999999 7 9999999999 99919919,; 8‘9; 93‘99 ‘ ” . 59 9 , ~9 , £7?!“ 12539,, 9919 .1999 3. 999,3 99,9 9999999? ,9, 399999.99, 9,, 9,} 1f 1 A, 9 - . fi, is: 9:: 9,9,9;- "9 9,93% 9, 313244?” :‘J Jugg- Egg-1.2.9214.“ A’? l. . Jrrw. 9. _ _ ‘ =_ -. dag, *x-f‘w‘ A J: 2%; r- I: . 9. .. .‘ - . 4b.. “2%; . 9.‘ I ‘ 'Fidipz' - 2‘51“- , .“‘“ , , 1 *3: 44%? ‘9; A Fv» 2.} A - $734 32"“) .1- ,- EZZA$ 42:3' ' 3‘21? ‘ 2,9. 591: :13) (I .l ‘9 9:99 3“), n ‘ , ; ’ d $391329 9 9. , “529' t9 ”’19 5.1 r ‘., 3“ “Eat 33. “791’ " 9.1 99 , 9'9 i9 5 9 v i 9 .‘ “fig-rn‘!‘ \ 3‘, ' '9 ‘4"; ~ ‘1‘. 1' ‘3 ,9, 399 3196",. . ,. ' , R999 :9 11,5 ' r: 9 . , “9‘ ,0, . .33 g ‘t- 9 .9 .1 b '19.‘ :29” E9999» ’9‘ '9'“??? '4‘". :‘g' . ._ 9. k _‘ a 1 5‘ -’«" , r‘ ,-.- . " .. d"? g .:’ H“ “0‘: 3" " ""-‘."’~'4." : “" "‘~ - 3-? s ' I: '2‘ u ’z.‘ s“; ’u'..."j.::‘.: {2‘9 ‘15 8 -' ‘ a 'i‘ "Eviefi’ggfi. - 50" 939 f .: DA'M?ZEP’:W1'9B;.‘ , 'mrrm.mp-nm .....---. i l“ ,’ .. ”a“ 'b y 1." ,. . This is to certify that the dissertation entitled New Methods For Aryl—Carbon Bond Formation presented by Chi—Jen Frank Du has been accepted towards fulfillment of the requirements for Ph. D. degree in Chemistry #M4J Major professor Date 2" 28' 87 MS U is an Affirmative Action/Equal Opportunity Institution 0-12771 MSU LIBRARIES n \— RETURNING MATERIALS: PIace in book drop to remove this checkout from your record. FINES wiII be charged if book is returned after the date stamped beiow. ““7 NEW METHODS FOR ARYL-CARBON BOND FORMATION By ChiaJen Frank Du A DISSERTATION submitted to Michigan State university in partial fulfillment of the requirement for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1987 @1987 CHI —JEN FRANK DU All Rights Reserved ABSTRACT NEW METHODS FOR ARYL-CARBON BOND FORMATION BY Chi-Jen Frank Du Nucleophilic addition of organometallics to an aryne is a potentially useful method for constructing an aryl-carbon bond. The reactions of polyhaloarenes with organomagnesium halide (Grignard reagent), which belong mechanistically in this category, are the subject of this thesis. The mechanisms of these reactions consist of three steps, (1) halogen-metal exchange to form a polyhalogenated aryl Grignard reagent, (2) elimination of magnesium bromide to form an aryne intermediate, and (3) nucleophilic addition of the Grignard reagent to the aryne. For multiple bond formation, the last two steps are repeated. A new one-pot route to p-terphenyls is described in chapter A. Addition of 1,4-dibromo-2,5-diiodobenzene to an excess of an aryl Grignard reagent gives, prior to quenching, a 1,1';4',1"-terphenyl—2',5'-di-Grignard reagent. After aqueous quench, p-terphenyls are isolated in 30% to 50% yield. Chi-Jen Du In chapter B, the reaction of g—dihalobenzenes with aryl Grignard reagents, followed by electrophilic quench, has been developed into a useful synthesis of unsymmetric biaryls. Hexabromobenzene or l,2,4,5-tetrabromo-3,6- dichlorobenzene reacts with an excess of an aryl Grignard reagent in THF to give 1,2,4,5-tetraarylphenyl-3,6-di- Grignard reagent, which can then be quenched with various 0, Br electrophiles (H20, D and I2). Twenty new 1,2,4,5- 2 2 tetraarylbenzenes were prepared in this way. A mechanism that involves a sequence of organometallic aryne intermediates is proposed. The reaction of an aryl Grignard reagent with 2,6- dibromoiodobenzene or other 1,2,3-trihalobenzenes gives 2,6- diarylphenylmagnesium halide. The mechanism involves Grignard exchange at the central halogen, followed by two cycles of magnesium halide loss and regioselective capture of the resulting aryne by the aryl Grignard reagent. Quenching with an electrophile then gives a m-terphenyl in which the outer rings are identical and in which the 2' position in the central ring is substituted with the electrophile. Some preliminary results on the reaction of selected vinyl, alkyl, acetylenic and heteroaryl Grignard reagents with polyhaloarenes are presented in chapter E. Vinyl and alkyl Grignards behave similar to aryl Grignards, and successful examples of p, m and 1,2,4,5 aryl-vinyl or aryl- Chi-Jen Du alkyl bond formation are described. For selected acetylenic and heteroaryl Grignards, the necessary initial halogen- metal exchange step did not take place, preventing the desired reaction. this difficulty was overcome by using one equivalent of alkyl Grignard for the exchange and stoichiometric amounts of acetylenic or heteroaryl Grignards to trap the aryne intermediates. Two unusual examples of the Diels-Alder reaction are described in the final section of this thesis. In both cases, the diene component appeared to be an aryl or heteroaryl Grignard reagent. Reaction of mesitylmagnesium bromide with hexabromobenzene gave after aqueous quench, in addition to 1,2,4,S-tetramesitylbenzene, a mixture of two biscycloadducts in a 20% yield, 9,10-dibromo-1,3,5,7,11,13- hexamethyl-[1,4,5,8]-bisetheno-1,4,5,8-tetrahydroanthracene and 9,10-dibromo-1,3,6,8,11,14-hexamethyl-[l,4,5,8]- bisetheno-1,4,S,8-tetrahydroanthracene. Quenching with deuterium oxide or with methyl iodide showed that these bis- adducts, prior to quench, possess two organometallic functionalities in vinyl positions. In the reaction of 2- thienylmagnesium bromide with 2,6-dichlorophenylmagnesium bromide, there was obtained after aqueous quench, in addition to the expected major product 1,3-bis[2- thienyl]benzene, a low yield of l-chloronaphthalene. The latter product is thought to arise from the cycloaddition of 3-chlorobenzyne to 2-thienylmagnesium bromide, followed by extrusion of sulfur. Quenching with D 0 support this 2 Chi-Jen Du conclusion. These two examples suggest that a carbanion on a diene may enhance its reactivity toward dienophiles in the Diels-Alder reaction. ACKNOWLEDGEMENTS The author wishes to express special gratitude to Professor Harold Hart for his continual guidance and encouragement. Professors W. H. Reusch, G. E. Leroi and D. Nocera are also acknowledged for their serving on my committee. Appreciation is also extended to Mr. K. Harada and Dr. D. Ng for their contribution in this work. Financial support from the Dow chemical company, BASF chemical company, the National Science Foundation, the National Institute of Health and Michigan State University is appreciated Finally, my sincere thanks to my wife, Chiung-Ling, and our families for their unending love, confidence, and support. TABLE OF CONTENTS chapter LIST OF TABLES. . . . . . . . . . . . . . . . . INTRODUCTION . . . . . . . . . . . . . . . . . RESULTS AND DISCUSSION . . . . . . . . . . . . A. A New Synthesis of p-Terphenyls. . . . A.1. Reaction Mechanism. . . . . . . . . A.2. Synthetic Scope. . . . . . . . . . B. The Synthesis of Unsymmetric Biaryls . C. A New Synthesis of 1,2,4,S-Tetraarylbenzenes C.1. Reaction Mechanism. . . . . . . . . C.2. Synthetic Scope. . . . . . . . . . D. A New Synthesis of m-Terphenyls. . . . D.l. Reaction Mechanism. . . . . . . . . D.2. Synthetic Scope. . . . . . . . . . E. Reactions of Polyhalobenzenes with Vinyl, Alkyl, Acetylenic and Heterocyclic Grignard Reagents. . . . . . . . . . . . . . . F. An Unusual Diels-Alder Reaction Between An Anionic Diane and Benzyne. . . . . . . EXPERIMENTAL 1. General procedure. . . . . . . . . . 2. 1,4-Dibromo-2,5-diiodobenzene 1. . . . 3. 1,5-Dibromo-2,4-diiodobenzene 18.. . . 4. General Procedure A for p-Terphenyl Synthesis. page ix 19 19 20 25 30 35 37 43 49 51 54 63 75 86 86 87 87 10. ll. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 2',5'-diiodo-p-terphenyl 4 (Iodine quench) 2,4,6,2",4",6"—Hexamethyl-p-terphenyl 25. . 1,4-[1',1"-Dinaphthyl]benzene 26.. . . . . . Reaction of 18 with phenylmagnesium Bromide. The effect of lithium 2,2,6,6-tetramethyl- piperidide on the reaction of 1 with phenyl- magnesium Bromide. . . . . . . . . . . . . The effect of potassium t-butoxide on the reaction of 1 with phenylmagnesium bromide.. General Procedure B for Biaryl Synthesis. Preparation of 4-Methylbiphenyl 40.. . . . General Procedure C for Biaryl Synthesis. Preparation of 3,4-Dimethoxybiphenyl 45. . . z-Bromobiphenyl 41 (Bromine quench). . . . . 2-Iodo-4,5-dimethoxybiphenyl 46 (Iodine quench). . . . . . . . . . . . . . . . . . 2-Iodo-4,5-dimethylbiphenyl 42. . . . . 2-Iodo-3',4'-dimethylbiphenyl 44.. . . . . . 2-(2'-Iodophenyl)naphthalene 48. . . . 1-(2'-Iodo-4',5'-dimethylphenyl)- naphthalene 49.. . . . . . . . . . . . Hexabromobenzene 55. . . . . . . . . . 1,2,4,5-Tetrabromo-3,6-dichlorobenzene 56. General Procedure D for Tetraarylbenzene Synthesis. preparation of 1,2,4,5-Tetraphenyl- benzene 57. . . . . . 88 89 89 89 9O 9O 91 91 92 92 92 93 93 93 94 94 94 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 1,4-Dibromo-2,3,5,6-tetraphenylbenzene 87 (Bromine quench). . . . .,. . . . . . . . . 1,4-Diiodo-2,3,5,6-tetra-p—tolylbenzene 9O (Iodine quench). . . . . . . . . . . . . . . 1,4-Dideutero-2,3,5,6-tetramesitylbenzene 86 (Deuterium oxide quench).. . . . . . . . . . Reaction of 1,4-dibromo-2,3,5,6-tetrachloro- benzene 68 with Phenylmagnesium Bromide. . . Reaction of bromopentachlorobenzene 70 with phenylmagnesium bromide. . . . . . . . . . . 1,2,4,5-Tetra-p-tolylbenzene 74. . . . . . . 1,2,4,S-Tetra-m-tolylbenzene 75. . . . . . . 1,2,4,5-Tetra-o-tolylbenzene 76. . . . . . . 1,2,4,S-Tetramesitylbenzene 77.. . . . . . . 1,2,4,5-Tetra-1'-naphthylbenzene 78. . . . . 1,2,4,5-Tetra-2'-naphthylbenzene 79. . . . . 1,2,4,S-Tetra-p-biphenylylbenzene 80.. . . 1,2,4,S-Tetra-m-biphenylylbenzene 81.. . . . 1,2,4,5-Tetra(o-ethylphenyl)benzene 84.. . l,2,4,5-Tetra(2,6-dimethylphenyl)benzene 83. 1,2,4,S-Tetra(pentamethylphenyl)benzene 82.. l,4-Dibromo-2,3,5,6-tetra-m-tolylbenzene 88. 1,4-Dibromo-2,3,5,6-tetra—o-tolylbenzene 89. 1,4-Diiodo-2,3,5,6-tetra-o-tolylbenzene 92.. 1,4-Diiodo-2,3,5,6-tetra-m-tolylbenzene 91.. 1,4-Diiodo-2,3,5,6-tetra-(2,6-dimethylphenyl)- O 95 96 96 97 97 98 98 98 99 99 99 99 100 100 100 100 101 101 101 102 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. benzene 93.. . . . 1,2,4,S-Tetra-m-anisylbenzene 85.. 2,6-Dibromoiodobenzene 94. General procedure E for m-Terphenyl Synthesis. Effect of Reactant Ratio on Yield. 2,4,6,2",4",6"-Hexamethyl-1,1':3',1"-terphenyl 104. O O O O O O O 2,2"-Dimethoxy-1,1':3',1"-terphenyl 109. 2,5,2",5"-Tetramethoxy-1,1':3',1"-terphenyl 110 O O O O O O O O 2'-Iodo-1,1':3',1"—terphenyl 96 quench). . . . . . 1,1':3',1"-Terphenyl-2'-carboxanilide 114 From 96 O O O O O O 1,1':3',1"-Terphenyl-2'-carboxanilide 114 From 94 O O O O O O (Iodine 5'-bromo-1,1':3',1"-terphenyl 117. 5'-Chloro-1,1':3',1"-terphenyl 118.. 5"-Bromo-1,1':4',1":3",1"':4"',l""-quinqui- phenyl 119.. . . . 5',5"'-Diphenyl-l,l' :3t,1n:3u’1"0:3ut'1IIII_ quinquiphenyl 113. O 5"-bromo-5',5"'-diphenyl—1,1':3',1":3",1"' :3"',1""-quinquiphenyl 120.. 5',5"-Dipheny1- 1,1':3',1":3",1"'-quaterphenyl 122. . . . . . . . V} 102 102 103 103 104 104 104 105 105 105 106 107 107 107 108 109 109 59. 60. 61. 62. 63. 64. 65. 66. 67. 68. 69. 70. 71. Reaction of o-bromoiodobenzene and styryl- magnesium bromide. . . . . . . . . . . . . Reaction of 1,4-dichloro-2,3,5,6-tetrabromo- benzene and styrylmagnesium bromide. . . . Reaction of 2,6-Dibromoiodobenzene and styryl- magnesium bromide. . . . . . . . . . . . . Reaction of 1,4-Dibromo-2,5-Diiodobenzene and styrylmagnesium bromide. . . . . . . . . Preparation of 1,2-diphenyl-vinylmagnesium bromide 130. . . . . . . . . . . . . . . . Reaction of 2,6-dibromoiodobenzene and 1,2-diphenyl-vinylmagnesium bromide 130. . Reaction of 1,4-dibromo-2,5-diiodobenzene and 1,2-diphenyl-vinylmagnesium bromide 130. . Reaction of o-bromoiodobenzene and phenyl- ethynylmagnesium bromide 136.. . . . . . . Reaction of o-bromoiodobenzene and phenyl- ethynylmagnesium bromide (iodine quench).. Reaction of 2,6-dibromoiodobenzene and phenyl- ethynylmagnesium bromide 136.. . . . . . . Reaction of 1,4-dibromo-2,3,5,6-tetrachloro benzene 68 and phenyl—ethynylmagnesium bromide 136. . . . . . . . . . . . . . . Preparation of 2,6-dichloro-phenylmagnesium bromide 139. . . . . . . . . . . . . . . . Reaction of 2-thienylmagnesium bromide and O 110 110 111 112 112 112 113 114 114 115 115 116 2,6-dichloro-phenylmagnesium bromide.. 72. dichloroiodobenzene 94. 73. 74. 75. Biscycloadduct 160 . BIBLIOGRAPHY. vifi Biscycloadduct 158 . Biscycloadduct 159 . Reaction of ethylmagnesium bromide and 116 117 117 117 118 119 LIST OF TABLES p-Terphenyls from Aryl Grignard Reagents and 1. Biaryls from Aryl Grignard Reagents and o-Dihaloarenes. . . . . . . . . . . . . . . . . Reaction of Hexahaloarenes with PhMgBr at 7°C . Tetraarylarenes from Aryl Grignard Reagents and Hexahalobenzenes. . . . . . . . . . . . . . . . m-Terphenyls from Aryl Grignard Reagents and 1,2,3-Trihaloarenes . . . . . . . . . . . . . . 5-Halo-1,3-diarylbenzenes from Aryl Grignard Reagents and 1,2,3,5-Tetrahalobenzenes. . . . . Mass and NMR Spectral Data of Biscycloadducts . page 26 32 39 44 55 59 77 INTRODUCTION The formation of aryl-aryl bonds is an important mani- pulation in organic synthesis. A number of methods have been developed to effect the homo-condensation of two aromatic rings, including the direct oxidative dehydrodimerization by Pd,1 V2 or T13 (eq. 1) and the reductive coupling with loss of a substituent group as in the classical4a-d and 4e-l Ullmann reaction (eq. 2). Recently, the modified coupling of organometallic compounds catalyzed by transition metals5 has also served as a method of choice (eq. 3). Pd, V, T1 2 ItrII -.Ar*——-Aur (eq- 1) Cu, Ni, Pd 2 1tr-J( D» Aar———.AI' (eq. 2) x = I, Br, C1 Cat. 2 Ar-H —>Ar--Ar (eq. 3) H = T1(orf)2, HgCl, MgBr cat’ = Pd: Rh, T1 However, extention of these methods to the cross-coupling of two unlike aromatic moieties has been limited by their unavoidable homo-coupling side reaction. Therefore, a wide variety of approaches, either modified procedures or new 2 pathways, for the unsymmetric condensation of two different aryl rings either inter- or intramolecularly have been explored to address this problem. Here we have attempted to classify and briefly review these literature methods into several categories based on their reaction features. A. Electrophilic Aromatic Substitutions Reactions that involve the formation of an aryl cation, which then acts as an electrophile to attack another aromatic ring, thus constructing a new aryl-aryl bond, fall into this category. Historically, metal salt induced oxidative dehydrodimerizations of aromatic compounds (eg. 4) have been known for many years and are refered to as the Scholl6 reaction. Such processes have little synthetic xx n ArH + Ar'H >Ar—Ar' (eq. 4) utility, however, primarily because of low reaction yields and complex mixtures of products. They appear to proceed most readily (a) with relatively electron rich aromatic subtrates and (b) when a metal salt is used which can function as a one electron oxidant. For example, electron- rich substrates such as phenols and phenol ethers can be effectively coupled intramolecularly by using vanadium oxytrihalide as an oxidant (eq. 5).7 Recently, several different metal salts including Tl(OTf)38 and Pb(OAc)49 were 3 found to give better yields. Alternatively, controlled 10 of phenolic and phenol potential electrochemical oxidation ether substrates has been developed as a method of choice for the synthesis of certain natural products (eq. 6). one _ OHe H20 \ @/ 0H8 HeO 0H8 HeO l VOF'3 , BF3OEt2 HeO @ 9 @ r . t. 0 (eq . 5) HeO HeO @ OHe OHe e OM HeO I::I| @ - — (eq- 6) -2 e r . Heo -H+ e0 @ N- 6 "‘CH HeO CH3 Another facet of electrophilic substitution reactions II is seen in some recent examples which utilize either the phenoxeniumll or arylnitrenium12 ion as the electrophile to form the corresponding hydroxy- or amino-biphenyl (eq. 7 and 8). Recently, Okamoto13 and coworkers reported a case of H 0-..... W Q i CF3SOBB/CF3COZH 9 ———>— (eq. 7) Benzene OH I I R x R x / R + ?3 ( NH 1 g or so H g 8) 3 3 I @ CF3COZH @ \ I reductive phenylation which involves an iminium-benzenium dication intermediate (eq. 9). The intermediate was generated by the reduction of a nitro aromatic compound with zinc or iron pentacarbonyl in the presence of trifluorome— thansulfonic acid. Reaction with benzene gives mainly 4- amino-biphenyl and the method has general utility for aminobiphenyl synthesis. An intramolecular example of the reductive phenylation is shown in (eq. 10). 1132 .r “2' e @ (37380311 0 Benzene + 2 (eq. 9) Zn or IPC J l::, l[::l major .:\ CF so H @‘W N—CH3 3 3 - N-CH3 (eq. 10) Zn _fl/ ‘° @ t No NH 5 B. Radical-mediated Reactions Radical-mediated reactionsl4 involve the initial formation of aryl radicals, followed by reaction with the other aromatic ring and termination by hydrogen atom abstraction. In the first step, the aryl radicals can be generated by several methods (Scheme 1) including (a) the thermal decomposition of diazonium salts under alkaline conditions (Gomberg-Bachman-Hey reaction) or under acidic conditions (Pschorr reaction), (b) the thermal decomposition of arenecarbonyl peroxides and (c) the photolysis of aryl halides. Next, the aryl radicals attack the aromatic subtrates in much the same way as in the electrophilic substitution, and the subsequent loss of a hydrogen atom results in the formation of unsymmetric biaryls. Scheme 1 acidic or + basic A o _ ° b) Alf-("Z-O-O-C-Ar A]? ——> .-—AI' “H Ar 0 0 Although radical-mediated reactions are widely used for intramolecular cyclizations in natural product synthesis,15 in general, the intermolecular condensation is not as effective. Lately, however, there have been several 6 modifications and new developments. For example, for those reactions involving the aryldiazonium intermediate, higher yields can be obtained either by generating a stable 16 aryldiazonium salt from an aryltriazene in CF3C02H or by introducing the phase transfer reagent 18-crown-6 to catalyze the reaction (eq. 11).17 In the Pschorr cr co‘n N-N N 3 2 )> N co ' \ 2 2CF3 \\:urn . (eq. 11) Arli RDAc, ls-crovn-6 phenanthrene synthesis (eq. 12), the presence of sodium iodide and a phenylsulfonyloxy group on the acceptor ring also improves the reaction yield significantly.18 HeO [:1 ”eo\1E:]| MeO HeO (eq. 12) mzl IO H @ © MeO HeO Yield R = CH C H i-C H CNO l) i-c H cno 5 11 > 45% 2) Hal R = so c H l) i-c H cno 2 6 5 5 11 a 64% 2) Hal 7 C. Nucleophilic Aromatic Substitutions Nucleophilic substitutions rely on electron-withdraw- ing substituents in the substrates to stabilize the electron-rich intermediates. For example, the reaction under alkaline conditions of 2,6-di-t-butylphenol with nitro- benzenes bearing a leaving group leads to the formation of unsymmetric biaryls in good yields via nucleophilic aromatic 9 substitution (eq. 13).1 A. I. Meyers reported an aromatic OH t-Bu t-Bu on N02 VB“ VB“ NaOH, ouso (d * . 0 so c X - 2-halide, -802£h 2, , -802Ph 4 -NO substitution process which involves an oxazoline function as the activating group.20 It activates only organometallics capable of chelation and transfer of the nucleophile from a tight ion pair to the electrophilic site. For example, the reaction of an (o-methoxy)aryloxazoline with aryllithium or aryl Grignard reagent results in methoxy displacement to give o-(aryl)aryloxazolines, which can be rt 4 O/N O/N OMe R R (eq. 14) RM L O 'C.THF (0M8). (OMe). (OMe). 8 further hydrolysed to the corresponding benzoic acid derivatives (eq. 14). The reaction probably proceeds via an addition-elimination sequence (Scheme 2). Scheme 2 ~WQ (eq. 21) Q Q “" Q Q Q @ ..____,. + R R oxidative R conditions maj or ' minor basic conditions minor major E. TransitionéMetal Catalysed Coupling Reactions The catalysis of coupling reactions with transition metals has received a great deal of attention in organic synthesis during last decade. One of the traditional reactions used to effect aryl-aryl bond construction is the Ullmann reaction, in which copper or copper oxide is employed as a catalyst to combine two aryl halides. In the case of cross-couplings, the yield of the Ullmann reaction is determined by the difference in reactivity between the two aryl halides. In general, electronegative groups such as nitro and methoxycarbonyl in the ortho position strongly activate the aryl halides; on the other hand, the reaction is greatly inhibited by substituents such as amino, hydroxyl and free carboxyl group. An optimum yield is usually obtained when one aryl halide is activated and the other is relatively unreactive. For example, the activated 2,4,6- trinitrochlorobenzene is condensed with iodobenzene to give the cross-coupled product in 80% yield (eq. 22). 12 N02 / N02 Cu N02 --C1 + I —>NO2 @ N02 (eq. 22) N02 A recent modification27 proceeding through an arylcopper intermediate requires a two-step sequence (eq. 23). A copper(I) aryl species stabilized intramolecularly by a heteroatom is generated from one aryl halide. It is then reacted with another aryl halide bearing an ortho substituent which may also function as a ligand. This modified pathway was successfully applied to the total synthesis of steganacin (eq. 24).27 @:\Z n-BuLi Z CuX @Z ' Cu R, Bit (1) L1 R1 RI (eq. 23) Z R2 4' + CuX W R. R: ‘ Z 0 F0 X F ’0 can W::1\;2‘ -L NCfl"__. ,_ 1' SJ I CH’O (Z=NorS) x CHD (eq. 24) a,X-Br X' Br c, [’0 R x . NC‘H“; y = scH,CH,0 o ‘R: .‘H R,-OAc,R,-H Steganacin. 13 Another rapidly developing method for aryl-aryl bond formation is the coupling between an arylorganometallic and an aryl derivative catalyzed by a transition metal complex. 8 In 1972, Kumada2 and Corriu29 reported that the cross- coupling of Grignard reagents with aryl and alkenyl halides could be catalyzed by nickel complexes. Since then a great 0 deal of coupling methods3 involving various organometallics and various organoderivatives have been developed, and some of them have been successfully adapted to the construction of aryl-aryl bonds. These include the reaction of lithium 31 diphenylcuprate and aryl halides (eq. 25), the condensa- tion of a higher order mixed cuprate and the aryltriflate 32 derived from phenol (eq. 26), the reaction of aryl Grignard reagents and aryl halides catalyzed by nickel complexes (eq. 27),33 the cross coupling of an arylzinc reagent and an aryl halide in the presence of a palladium or nickel catalyst (eq. 28),34 the reaction of arylboronic acids and aryl halides catalyzed by palladium complexes (eq. 35 29), the condensation of an aryl phosphate with an aryl Grignard reagent in presence of nickel (eq. 30),36 the palladium-catalyzed electro-inductive coupling of aryl 7 halides (eq. 31),3 the nickel-induced reaction between an aryl ether and an arylmagnesium bromide (eq. 32),38 aryl demetallation of aryltin reagents in the presence of palladium complexes (eq. 33),39 the selective arylation of a bis(alkylthio)benzene with an aryl Grignard reagent catalyzed by nickel complexes (eq. 34),40 the reaction of an 14 arylcopper and an aryl halide with a palladium catalyst (eq. 35),41 and the arylation of a triphenylaluminum with an aryl halide using a palladium complex (eq. 36).42 @o + (mm... M Q...__.@ Q @ PdLn NIL H80 @ I + ch1—.———— "one (eq. 23) (eq. 27) g-N(i-Pr)2 (i- Pr) 2N—C=o one —_+@:@Pd1’n OMe Nil.n ‘rrwfib @@ N(CH3)2 N(CH3 )2 Pd 0] 31£®||'}—t- Bu 3 o . @Q ——~ I) I Pd-I 15 @@ + __. (eq. 32> e“? ’ 3 SCHMe2 SCHMe2 o o do Cu + I N02‘—""""’> N02 (@ .1 + flew (eq. 3.) Plausible mechanisms for these coupling reactions (eq. 34) @Q_@ 99? involving transition metals as catalysts include oxidative addition, transmetallation and reductive elimination.43 F. Benzyne-mediated Arylations Another method for aryl-aryl bond formation which has seen only limited synthetic use is the nucleophilic addition of an aryl lithium or other aryl organometallic to an aryne. The reaction was pioneered by Wittig, Huisgen and 16 44,45 coworkers. It was a key reaction in the early 46 but is recognition of benzyne as a reactive intermediate, usually looked upon as an unavoidable side reaction in the preparation, via aryl halides and phenyllithium, of arynes to be used for other purposes. However, the yield of biaryls by this route can sometimes be quite high, as in the following example (eq. 37).47 ocrg In this example, the intermediate aryne is formed by proton removal, but a similar result may be obtained when the aryne is generated from a 1,2-dihaloarene and magnesium metal by halogen-metal exchange (Scheme 3).48 Though metal Scheme 3 17 transfer between a polyhaloarene and an aryl or alkyl Grignard reagent has been observed (eq. 38),49 o-bromo- phenylmagnesium bromide, the precursor of benzyne, has not been prepared by exchange between a dihaloarene and a Grignard reagent. Br . Br Br x’Br PhMgBr or Br /,Br (eq. 38) 4;» Br Br EtMgBr Br Br Br ‘ MgBr In connection with our interest in synthetic application of aryne chemistry, we studied the reaction of aryl Grignard reagents and polyhalogenated aromatics. We find that aryl Grignard reagents can be used to generate and trap arynes resulting in the formation of one or more aryl-aryl bonds in an one pot reaction. The reaction yields are satisfactory and comparable with the known literature methods outlined above. The main reactions that we have studied and that are to be discussed in detail in this thesis are presented below (eq. 39-42). X X AngBr + _.___> “‘1”: (eq. 39) x” x X + —-—> Ar—. (eq. 40) X 18 x x x Ar Ar Ar... . @ __. x’ x Ar Ar x x x©x 'Ar “(eq-42) * “’- X= I, Br, Cl. 19 RESULTS AND DISCUSSION A. A.New Synthesis of p—Terphenyls Because of their interesting structural, electronic and optical properties, p—terphenyls have many practical uses. For example, p-terphenyls, which have a rigid rod-like structure in the solid state, have been widely investigated for their liquid crystal properties.50 The exposure of p- terphenyl to AsF forms an electrically conducting complex 51 5 which is a potential organic conductor. The highly fluorescent p-terphenyls have also been used as liquid scintillators for radiation detection52 3 and as laser dyes to generate short wavelength UV light.5 The whole spectrum of our discoveries originated from a study of the reaction between phenylmagnesium bromide 54 The addition at room and 1,4-dibromo-2,5-diiodobenzene 1. temperature of a solution of 1 in THF to four equivalents or more of phenylmagnesium bromide, also in THF, gave, after an aqueous quench, p-terphenyl 3 (54%) and 1,4-dibromo-2- iodobenzene55 6 (43%) (eq. 43). Quenching with iodine gave instead the diiodoterphenyl 4 and the starting dibromo- diiodobenzene 1. Hence, the actual reaction products are the terphenyl di-Grignard 2 and trihalomono-Grignard 5. This 20 8r 2 z 8: .—-Mch . @ —1flf——~» . (eq. 43) r 2 Br I 8 1 2({=Mox) s(z=MoX) 3(Z=H) 6(2=H) 4(zzl) 1(z=I) finding provides a new general method for the synthesis of p-terphenyls whose two outer rings are identical. This new synthetic method has the following unique features (1) two aryl-aryl bonds are formed in a one-pot reaction, (2) although, regiochemically, the potential for forming m- and p-terphenyls exists, the latter predominate, (3) the reaction proceeds via a novel two aryne sequence, (4) aryl Grignard reagents are used to generate and to trap arynes and (5) the product terphenyl has functionality which allows elaboration of the central ring. A.1. Reaction mechanism. Two plausible mechanisms were considered to account for our reaction. In the first mechanism we assume that Grignard exchange can occur either at iodine or bromine to obtain 5 or 7 respectively (Scheme 4). In this case we are forced to conclude that aryne formation from 5 does not occur at room temperature, so that 5 remains as one of the two observed reaction products, trapped with water or iodine to give 6 or 1. However, elimination must occur from 7 to give aryne 8 which is trapped by the Grignard reagent presumably to give 9 and/or 21 10. Grignard exchange and elimination is then repeated to give aryne 11 which is trapped to give the observed terphenyl di-Grignard product 2. I 8r 8'“ ° : t ‘ PhMgB' —-——> @ + Q I St l Scheme 4 Bng ‘ 5 —M°xz ’ arms 8: Ph Br Br + e w o 1.3. pn I 8880 ' 10 9 8 PhMgBr _ng_l n W > BrMo 3'“ Ph 11 2 An alternate mechanism is shown in Scheme 5. If we assume that Grignard exchange is more facile at iodine than at bromine, we then obtain first mono-Grignard 5, then di- Grignard 12. If this step ( 5712 ) is slow, then some 5 may remain even in the presence of excess phenylmagnesium bromide. Intermediate 5 then leads to the observed products 6 or 1, whereas di-Grignard 12 leads to terphenyls 2 or 4 by reactions analogous to those in Scheme 4. 22 Scheme 5 8M 8 :0? W ”JOE“ B t _ 8! I 3 03' 1 5 12 - l-MQBQ ph 3m Ph ‘ 3M9 ,@ l-Mdflri @ m . I Mr 3' M98: 8' 11 _ 14 13 lama! ”‘63” SIMS Ph 2 Both mechanisms have certain common features. The exclusive formation of p-terphenyls (<1% of m-terphenyl is detectable by gas chromatography when the reaction mixture is quenched with water ) implies that addition of phenyl- magnesium bromide to the organometallic aryne intermediate (1192 in both Schemes, and presumably also 13*14 in Scheme 5) is regiospecific. This specificity probably arises from the importance of having like charges in 2 or 14, and the transition states leading to them, as far apart as possible. Several observations favor Scheme 5 over Scheme 4. For 56 15 is not an effective example, 1,2,4,5-tetrabromobenzene replacement for 1 under the same reaction conditions, although under more severe reaction conditions it does give some p-terphenyl (Scheme 6). For example, phenylmagnesium 23 Scheme 6 +10. \ _ln_._.z_. 15 (1001’ BiJIIIIEf" ft Br Br 15 24 n H 0' 8' ' .‘mfl—filr rt I—L" ‘3 (1“) © 8 (excess) 16 . (60 I) fi'hri‘SLMnQ 3' Q @ 8: 17 (371) bromide and 15 at room temperature gave, after 2 hour, mainly recovered starting material. However after 24 hour, 16% of p-terphenyl and 66% of 1,2,4-tribromobenzene57 were obtained after aqueous quench. After 3 hour at reflux (THF), 12% of p-terphenyl and 57% of 3,4-dibromobiphenyl58 were obtained. Apparently Grignard exchange can occur at bromine, but it is very slow at room temperature. The resulting 2,4,5-tribromophenylmagnesium bromide remains at room temperature (to give 66% of 16: if Scheme 4 is correct, a small amount is converted to di-Grignard and ultimately gives 3). At reflux, however the tribromo-Grignard eliminates to give 4,5-dibromobenzyne, the precursor of biphenyl 17. These results suggest that Grignard exchange with 1 is more likely to occur at iodine than at bromine. This conclusion is reinforced by some low temperature results. Phenylmagnesium bromide (1.8 eq.) was added to one equiva- 24 lent of 1 at -78°C, stirred for 1 hour and quenched, to give 52% of recovered 1 and 43% of 6. No terphenyl and no bromoiodobenzene was observed. Thus exchange appeared to be rapid at iodine even at -78°C, furnishing good evidence for 5 but not for 7. There is also some evidence for the formation of di- Grignard 12. A similar experiment to that just described, but at -22°C gave after aqueous quench 73% of 6, 1.1% of p- dibromobenzene and 6.4% of 1,2,4-tribromobenzene. The presence of p-dibromobenzene suggests that di-Grignard 12 is formed, though not to a great extent at -22°C. The formation of tribromobenzene suggests that the conversion of 5 to 12 is reversible (i.e., 12 can react with PhX to reform 5 when X=I or, if X=Br, its tribromo analog). Thus our mechanistic evidence indicates that Scheme 5 is a permissible mechanism; it also tends to argue against Scheme 4 in that we can find no direct evidence for 7. However Scheme 4 cannot be entirely discounted, since some Grignard exchange does occur at bromine in the case of 15. Still, it would seem that the major Grignard exchange occurs at iodine, favoring Scheme 5. Both schemes postulate aryne intermediates. To check this out we carried out the reaction of phenylmagnesium bromide with 1,5-dibromo-2,4-diiodobenzene 18 in place of 1 (eq. 44). If the reaction follows a path analogous to Scheme 4, aryne 8 should be an intermediate; if it follows Scheme 5, 13 should be an intermediate. In either case we 25 3: 8r 3' 3' @ "raw H p" Ph l(eq. 44) 18 19 20 expect p-terphenyl to be formed even though like halogens in 18 are meta to each other instead of para, as in 1. The situation is somewhat complicated by opposing factors in di- Grignard formation from 18; preferred exchange at iodine would give a meta di-Grignard, whereas to keep like charges as far apart as possible, exchange at one iodine and one bromine might be favored. In fact, the predominant product after aqueous quench was p-terphenyl 3; however some m- terphenyl 19 was also formed. As in the reaction with 1, a dibromoiodobenzene, this time 20,60 was also formed. We think the results with 19 support the proposal of aryne intermediates in these reaction, but the precise origin of the minor amount of 19 from 18 requires further study. A.2. Synthetic scope. A variety of aryl Grignard reagents react with 1 to give p-terphenyls. Table 1 summarizes the results. The reactions were carried out by adding a THF solution of 1 to four equivalents of the Grignard reagent at room temperature over modest addition times. Aqueous quench gave the terphenyl shown, as well as by-product 6. The products were separated and purified by column chromatography and the yields shown (not optimized) are of pure, isolated products. Except for 25 which is new, the terphenyls had melting points which agree with those 26 Table 1. p-Terphenyls from Aryl Grignard Reagents and 1 Grignard reactn yield entry reagent time, h product % m.p. ref. 1 ©Mfl' 2.5 Q Q Q 54 212-213 61 3 2' 1.0 © © @ 53 387-388 62 21 3 var—@u 3.0 cu, @ © Q n, 53 243-249 63 22 . ’5 4 ”{3}... 3.. Q Q Q 50 1.1 .4 23 cu, )5 5 @m 3.0 Q © © 45 145.5-146 55 24 cu, 6 up ugh 6.0 49 181-183 cm 27 Table 1. (Cont'd) 10 0 O O @@@ 3° 26 35 27 @@© ‘3 28 “m“ 3° 29 206-207 259-260 283.5 273-274 66 63 67 68 28 reported in the literature, and spectra consistent with the assigned structures. There are several types of p-terphenyl syntheses in the literature. Our method in general provides a simpler and shorter route and gives as good or better overall yields than all of the literature methods for the known p-terphen- yls in Table 1. One especially noteworthy example is the much simplified synthesis of p-quinquephenyl 21 over the Organic Syntheses procedure.62 The preparation of 24 and 25 in fairly good yield also shows that the reaction is not particularly susceptible to steric factors. It is clear at this stage that although p-terphenyls can easily be prepared in one step by this method, the reaction as so far described has an annoying feature, namely the formation of by-product 6. The products are in fact the di-Grignard 2 and the mono-Grignard 5, so that efforts to elaborate 2 with electrophiles other than the proton would be hampered by the simultaneous elaboration of 5, leading to separation problems. We therefore sought a method to eliminate the undesired by-product. This objective was achieved in several ways. The best to date is to carry out the reaction between 1 and the aryl Grignard reagent in the usual way (1:4 mol ratio), and then to add one equivalent of either lithium tetramethyl- piperidide (LiTMP) or potassium t-butoxide to complete the reaction. In this way, the yields of 3, 22 and 24 were increased to 75—80% and very little or no 6 was present in 29 the final product. Exactly how these reagents function remains to be clarified. It should be mentioned that replacement of the aryl Grignard by an aryllithium does not improve the synthesis. Indeed, using the same reaction conditions as for the experiments in Table 1, the yield of p-terphenyl from 1 and phenyllithium was only 12-23%. The reason why Grignard reagents are superior to aryllithiums is a subject that requires study. 30 B. The Synthesis of Unsymmetric Biaryls Since the extensive studies on the rotational energy barriers and substituent effects on the racemization of chiral biphenyls in the 1930569 until, very recently, with the total synthesis of the potent antileukemic lignan like stegane,7o compounds with a biphenyl skeleton have attracted chemists' attention. All of the methods described in the introduction section can be adapted to the construction of unsymmetric biphenyls, but each method has limitations. The best synthetic route to a particular biphenyl depends on its structure. As mentioned, nucleophilic addition of aryl organometallics to arynes is one method to effect the aryl- aryl bond formation. Its most attractive feature is that the aromatic ring derived from the polyhalobenzene subtrate possesses an organometallic functionality adjacent to the newly formed aryl-aryl bond in the final product. Elaboration of this funtionality with various electrophiles is the unique synthetic characteristic which broadens the scope of our method. The reaction of aryl Grignard reagents with o- dihalobenzenes can be readily extended to the unsymmetric biaryl synthesis. A diluted solution of o-dihalobenzene in THF was added to 2 equivalents (or more) of an arylmagnesium bromide at room temperature. The reaction mixture was 31 stirred for several hours, followed by electrophilic workup. The results are summarized in Table 2. Several features of the results are worthy of mention. Comparison of entry 3 with entries 1 and 2 shows that at least one iodine is needed on the dihaloarene under these mild reaction conditions. This suggests that Grignard exchange occurs mainly at iodine (see Scheme 5). Comparison of entries 5 and 6 or 9 and 10 illustrates how capture of the o-biaryl Grignard product with different electrophiles can be useful. Entries 6 and 7 illustrate how a complementary choice of iodobromoarene and Grignard reagent, followed by quench with an electrophile, can place the latter at the ortho position of either aryl ring (extentions, for example, to the synthesis of specifically labeled, i.e. deuterated, biaryls are obvious). Comparison of entries 8 and 9 illustrates how the yield may be improved, in some instances, by adding one equivalent of LiTMP to the reaction mixture prior to workup. The bromoiodoarenes in Table 2 were selected such that the derived aryne would be symmetric and give a single product on nucleophilic addition of the Grignard reagent. They were also selected, in this initial study, for their ease of synthesis. All of the reactions listed in Table 2 are presumed to proceed via Scheme 7. Following initial Grignard exchange at iodine to give 51, loss of MgXBr to furnish arynes 52 is apparently rapid at room temperature, since little or no by- 32 Table 2. Biaryls from Aryl Grignard reagents and o-Dihaloarenes dihalo- Grignard procedure yield entry arene reagent /quench product % ref. 30 35 40 I 2 @[1 35 B/HZO - 4o ' 69 31 3' b 3 @fi 35 B/HZO ---— 32 4 3o B/Brz' 63 72 36 41 H a: 5 m 36 B/HZO a; © © 66 73 H: 33 42 6 33 36 B/I2 as Q Q 65 H C, 7 .0 mfg}... W H... .4 . I 33 Table 2. (Cont'd) cmo CH : 7. 777720 a..." 77 7. cmo ' _ 45 69 34 34 36 C/HZO 45 c o 34 36 C/I2 who“. 58 74 u 46 Inn: 47 7. “re/.2 7. 12 39 7. me o o .3 'o 49 13 33 In procedure C, LiTMP was added prior to quenching. - No biphenyl ‘ E‘. was isolated and >75% of 32 was recovered. 34 product correponding to the reaction of 51 with the quenching electrophile was observed. This was true when the substituent S was H, Me or MeO, as in all of the entries in Table 2. It is somewhat disturbing, then, to recall that when one S is bromine and the other is iodine (as in 5 or in the Grignard precursor of 20), aryne is apparently not formed from the mono-Grignard under very similar conditions (that is, 6 and 20, and not bromoiodobiphenyls, are the main by-products of the p-terphenyl synthesis. Perhaps the electron-withdrawing halogen substituents in these particular substrates stabilize 51 whereas electron-donors such as Me and MeO favor aryne formation. Other features that require further study are the role of the metal (Mg or Li, for example), of the counterion X in AngX, and the yield-enhancing effect of LiTMP and KO-t-Bu. Scheme 7 Br AMX S a: Mx s f ‘ 9 2 --1-+ - )I S ' S My 5 50 51 52 (s = H. Mo. MeO) AwMgX S A’ + S A! *——-‘ S E S Max 54 53 35 C. A.New Synthesis of 1,2,4,5-tetraarylbenzenes. In continuing our study of the reaction of aryl Grignard reagents with polyhaloarenes, we observed that the reaction of excess aryl Grignard reagents with hexabromobenzene (or with dichlorotetrabromobenzenes) in THF solution at room temperature gave a good yield of 1,2,4,5- tetraarylarenes compounds not readily available by any other method (eq. 45). In a literature survey, it was found that °'@8' mugs: _ @@@ THF rt 7 Br Br ’ (eq. 45) 7 0 © 55 (x: Br) 57 (rt-55%) 56 (X = C!) the origins of this chemistry are quite old. Durand reported that hexabromobenzene 55 reacts with phenylmagmesium bromide to give hexaphenylbenzene,76 but shortly afterwards, Dilthey showed through independent synthesis that the product was in fact 1,2,4,S-tetraphenylbenzene 57.77 Geissman confirmed 78 these results, and showed through obtaining 1,2,4,5- tetraphenylbenzene-3,6-dicarboxylic acid on carbonation that the actual reaction product was the tetraphenyl-1,4- diGrignard (Scheme 8). In these early studies, the solvents 36 were ether, in which 55 is nearly insoluble, or benzene, a poor Grignard solvent. It is not surprising, then, that the product yields were well below 10% and that considerable tars were formed. Scheme 8 Br Br Br PhMgBr “30+ @ © >9 Br Efl'Benzene C) l(:, Br 55 57 C023 .. 7.2 © @ ___>__> @ 58 Thirty years later, Berry and Wakefield treated 55 with magnesium in tetrahydrofuran (THF) using 1,2-dibromoethane as the entrainer and obtained pentabromophenylmagnesium bromide 59 in 32% yield (based on aqueous quench to give Scheme 9 H20 8' Br ———-—5- Br MqBr Br Br Br Br , Mg,THF Br Br Br Br 8" BrCH2CH28r 7 Br Br 60 Br (32%) Br CsHs Br 55 59 ‘3 IiII‘::"Bf . Br L CsHsMgBr, o c. THF f 51 (37) (95%) 37 pentabromobenzene); with ether as the solvent, the yield was under 5%.79 Significant for our work (Scheme 9), they found that if the solution of 59 was heated at reflux with benzene prior to the aqueous quench, a 3% yield of tetrabromobenzobarrelene 61 was obtained. Thus, at reflux 59 eliminated magnesium bromide to give some tetrabromobenzyne. A little later, Tamborski and coworkers reported that the conversion of 55 to 59 could be greatly improved using exchange with phenyl- or ethylmagnesium bromide in place of magnesium“49 The yield was better when the reaction was run in THF than in diethyl ether. In addition, a substantial amount (51%) of 1,2,4,5-tetrabromophenyl-bis-magnesium bromide 62 was formed when 2 equivalents of ethylmagnesium bromide were employed in the exchange process. The synthesis of tetraarylarenes from aryl Grignard reagents and hexahalobenzenes (Durand and Geissman's reaction) had no synthetic importance owing to the low reaction yield. This situation changed when we found that the yield improved dramatically with THF as the solvent. Also, since the actual reaction product is a di-Grignard reagent, elaboration of the final product by treatment with various electrophiles broadens the reaction's synthetic scope. C.1. Mechanism. The mechanism we propose involves as key intermediates the 1,4-di-Grignard reagent 62 and various organometallic arynes as outlined for hexabromobenzene in 38 Mg Br 63 MqBr M93 64 Scheme 10 Br War “93' 8' 8’ NMQBV ¥B'@8r AquBr _Br 6 Br “MOB'L_B’@\ a, Br Br 6. a: a: 9' 8' Br Br war 55 59 62 MqBr MoBr A! A! A! A! Ag M98: / A! 8' F..— f @ «—-—— Br / Br Br Br Br Br MgBr MqBr MqBr MoBr / Ar Br Ar ArMoBr Br.@Ar Ar’@8r V Ar Br \ Br MqBr MqBr AquBr MqBr MoBr / Ar Br Ar Ar M93: B’QM 4 ragga s: Ar 3' / . MqBr MqBr ”93’ MqBr MoBr Ar Ar Ar NWB’ Ar Ar —~— > Br Ar Ar , Ar Ar “98' MqBr 65 66 67 39 Scheme 10. As already mentioned the rapid formation of 59 from 55 and phenylmagnesium bromide is well established. We found that with excess phenylmagnesium bromide and low reaction temperatures, the presence of substantial amounts of the 1,4-di-Grignard also can be demonstrated, as shown by the quenching results in Table 3. When X=Cl, only the bromines exchanged, and the 1,4-di-Grignard was the principal product. The mechanism of such Grignard exchange is not well established, but we believe that other factors being equal, the isomer with the charges as far apart as possible is preferred. The results in Table 3 show that under the same conditions, more di-Grignard is formed when X=Br than when X=Cl. The results in Table 3 establish the plausibility of the first two steps in Scheme 10. Table 3 Reaction of Hexahaloarenes with PhMgBr at 7°Ca substrate products after aqueous quench, % X x X Br Br Br 3' @ Ph:.:”h Br @ Br 3' @ Br 3' Ph Ph x x X X=Br 26 32 22 X=Cl 58 27 8 ‘ Cl - Substrate (2.5 mmol) and phenylmagnesium bromide (15 mmol) in 80 mL of THF were stirred at 7 C for 4 h, then quenched with water and the product mixture analyzed by gas chramato- graphy. 40 The observation that Grignard exchange occurs at bromine but not at chlorine in 56 was used to test the proposal in Scheme 10 that di-Grignard formation is essential to the tetraarylation reaction. At least two bromines must be present on the bromochloroarene. Thus, dibromotetrachlorobenzene 68 reacts with phenylmagnesium bromide to give 5781 (eq. 46); on the other hand, bromopentachlorobenzene 70 gives, under the same conditions, only pentachlorobenzene 71 (eq.47). These observations tend to confirm the proposal in Scheme 10 that the product derives from the di-Grignard 62 and not from the monoGrignard 59. Br Cl Cl Cl C! PhMQB; Hz: :@ + 57 (3676) (eq. 46) Cl Cl ".1671 C! C‘ Br 69(6PI.) 68 Br CI Cl PhMgBr H20 Cl C1 -—————#> ————’- c: o rt.l3h o c: (eq. 47) CI Cl 70 71 The remaining reactions in Scheme 10, once 62 is formed, involve one main assumption, i;§;, that the addition of an arylmagnesium bromide to the intermediate organometallic arynes (i4§;, 63, 66 and so on) occurs regiospecfically always to give a 1,4-(and not a 1,3-) di-Grignard. Thus, 41 after aryne formation from 62 and Grignard addition, we obtain only di-Grignard 64 which may give three possible arynes. Each of these in turn can give but a single diaryl di-Grignard. Each of these di-Grignards can give only one aryne, and addition of aryl Grignard to any of these three arynes will give only 65 which leads to the observed 1,2,4,5-tetraaryl di-Grignard 67. Regiospecificity in the nucleophilic addition of aryl Grignards to the various organometallic arynes in Scheme 10 may be a consequence of two factors. One of these is the maintenance of like charges in the di-Grignards as far apart as possible. The second is the electronic effect of substituents (especially the organometallic substituent) on the charge distribution in the aryne intermediates. If, of the two aryne carbons, the one most remote from the organometallic substituent is the most negative, then a nucleophile will always attack meta to the organometallic substituent, giving the para-di-Grignard. Whatever the reason, the observed specificity is remarkable; only 1,2,4,5-tetraarylbenzenes are formed. Although it is possible to arrive at product 67 from the mono-Grignard 59 by loss of magnesium bromide to give tetrabromobenzyne, this alternative mechanism requires, in its early stages, some arbitrary choices regarding the regiochemistry of nucleophilic addition to neutral arynes. In its later stages, this mechnism requires arbitrary regiochemistry in the Grignard exchange reaction. Finally at 42 the di-Grignard stage, which is eventually necessary, it requires the same regiospecificity assumption as outlined in Scheme 10. Without these multiple assumptions, it would be difficult to rationalize why the two regioisomers of 67 (i4§;, 1,2,3,5- and 1,2,3,4-tetraarylbenzene) as well as pentaarylphenylmagnesium bromide are not also observed as products. We therefore believe that Scheme 10 is essentially correct and that di-Grignard 62 is a necessary intermediate in the reaction mechanism. We cannot tell, however, whether all of the multiple paths between 64 and 65 are essential or whether there is a preference among them. Finally, it should be noted that when 1,2,4,5-tetrabromo-3,6-dichlorobenzene 56 is used in place of hexabromobenzene, Scheme 10 must be modified. The first Grignard exchange presumely occurs at bromine to give 72 and the second exchange would then give 73. One can proceed to 67 by steps analogous to those in Scheme 10, except that aryne can be formed from 73 by elimination of either bromide or chloride, making the mechanism similar to, but formally somewhat more complex than, Scheme 10 (eq. 48). Cl Cl Br MgBr Br MgBr 56 —-> -—-> > -—> 57 (eq. 48) Br Br Bng Br Cl Cl 72 73 43 Since many of the intermediates proposed in Scheme 10 are nucleophiles, as is the final product 67, they could replace the original arylmagnesium bromide reagent in adding to the intermediate arynes, leading to a complex mixture of products. The original reagent is present in large excess, however, so its addition predominates. In this way, the good yields reported in Table 4 are obtained. Some higher molecular weight by-products are formed, however, perhaps in the manner just indicated. They are easily removed from the desired product through chromatography. C.2. Synthetic Scope. Addition of a THF solution of hexabromobenzene to eight equivalents of phenylmagnesium bromide in the same solvent at room temperature, followed by stirring for 12 h, gave, after hydrolysis. A 57% yield of 1,2,4,5-tetraphenylbenzene 57. The less expensive precursor 1,2,4,5-tetrabromo-3,6-dichlorobenzene 56, under the same conditions, gave a 65% of 57. In general, reactions with 56 tended to be cleaner in workup that those with 55. Other aryl Grignard reagents gave analogous products, as summarized in Table 4. Except for the first entry, all of the products in Table 4 are new compounds. Their structures rest on elemental analysis and mass spectra, on analogy with the first entry, and for those products with methyl substituents, on the symmetry required of their proton NMR spectra. For example, 74, 75 and 76 each showed sharp 12- proton singlets for the methyl protons (at 5 2.33, 2.27 and 44 Table 4. Tetraarylarenes from Aryl Grignard Reagents and Hexahalobenzenes hexahalo- Grignard yield entry arene reagent product % m.p. OC 0 9 1 55 C6H5-MgBr @ 57(65) 267-268 9 6 57 O ,9 2 56 p-CH3-C6H4-MgBr @ 73 247-249 6 6 74 3 56 m-CH3-C634-MgBr 71 157-158 ' 0 0 75 4 56 o-CHB-C6H4-MgBr 70 205-206 0 G 76 5 55 2,4,6-( )3- e e a 30 258-259 C6H2-MgBr @ a 6 S6 50 352-355 7 56 57 324-325 45 Table 4. (Cont'd) 10 ll 12 l3 14 56 56 55 55 55 56 55 4-C6H5-C6H4-MgBr 27:75:72.7- o-(CH3CH )- C6H4-MgBr m-CH3O-C6H4 -MgBr @ 0 52 62 57 52 50 57 30 380-381 257-259 376-378 288-290 223-225 199-200 258-259 46 Table 4. (cont'd) 15 16 17 18 19 20 21 56 56 56 56 56 56 56 C -MgBr 6H5 m-CH3-c6H4-Mgsr o-CH3-C6H4-MgBr p-CHB-C634-MgBr m-CH3-C6H4-MgBr o-CHB-C6H4-MgBr 2,6-(CH ) _ c6H3-Masf or: <36 WQW‘SWQ? oc> c>o on on 06) 29? Q ‘0 )0 O <30 oc> -6... _G_ or: c>o «60 oc» —o-: «no 66) -O~s M \O U 50 55 64 48 40 42 38 423-424 327-330 329-330 387-389 327-329 320-322 352-354 47 2.18, respectively) and 77 showed two such singlets, at 5 2.21 and 1.99 in the ratio of 12:24, respectively. These results clearly established the 1,2,4,5-orientation of the aryl groups around the central aromatic ring. The reaction is not particularly susceptible to steric effects. Thus, the compounds 77, 82 and 83 were prepared in comparable yields to less-hindered analogs. The peripheral rings in these compounds are twisted with respect to the central ring and almost certainly do not rotate freely. This conclusion is based on our observation that although the methyl signal in 1,2,4,S-tetra-o-tolylbenzene 76 is a singlet (broad), that signal in the hindered 89 (obtained by quenching the reaction with bromine) shows a complex pattern with five broad peaks at room temperature in CDClB. In the case of the more hindered 92, seven peaks in methyl region were observed under the same conditions. Thus rotation in 89 and 92, each of which has five possible conformers with a total of eight possible methyl peaks, is clearly restricted at room temperature. In 77, rotation is expected to be much more hindered. Indeed, CPK space-filling models of 77 are exceedingly difficult or impossible to construct, even with the mesityl rings at 900 to the central ring. The central ring can be elaborated by treating the reaction mixture with various electrophiles prior to workup. For instances, compounds 87-89 were obtained with a bromine quench, 90-93 with an iodine quench, and 86 with a 030+ 48 quench. The reaction of aryl Grignard reagents with certain hexahalobenzenes provide a unique one-step synthesis of 1,2,4,5-tetraarylbenzenes. Four new carbon-carbon bonds are generated in a one-pot reaction. The products are interesting for their restricted rotation properties. It is possible that they may also be useful intermediates in the synthesis of polynuclear aromatic compounds and for other purposes. 49 D. A.New Synthesis of m-Terphenyls A recent development of host-guest chemistry82 recognizes spherands and hemispherands as two new classes of host molecules that form the complexes with a number of metal ions. Because the synthesis of spherands usually relies on the coupling of m-terphenyl dibromide derivatives (eq. 49), the construction of m-terphenyl moieties is essential for the preparation of both types of hosts. Therefore, efficient methods for the preparation of m- terphenyls is still an area of active exploration. For example, Reinbount recently reported a synthetic route to t) BULB 2) FflAcAch . 3) EDTA ‘CH, 4) NC! 22,. CH’O. a spherand modified m-terphenyls which can be used as precursors for 83 spherand synthesis. The key steps of the preparation are outlined (Scheme 11). The nitro-m-terphenyl is prepared from 50 a highly sophisticated arene precursor in 57% overall yield, and subsequent convertion to the m-terphenyls needed for the coupling reaction takes a few more steps. Scheme 11 OR' R Br.CH2C02R2 Claisen Me Me __> M condensation Gun 0 M200 e Br 8" R1 = Me, R2: El R C0261 Nitromalonaldehyde decarboxylation + sodium salt v R2 ”“0 0' OeOR'O Br W a: R': H, R2: no2 In previous chapters of this thesis, several interesting aspects of the reaction of aryl Grignard reagents with polyhaloarenes were described. They include (1) halogen-metal exchange occurs preferentially at iodine rather than at bromine at ambient temperatures in several polyhaloarenes, (2) the subsequent elimination of magnesium bromide to form an aryne and the nucleophilic addition to the aryne takes place repeatedly when possible, resulting in multiple aryl-aryl bond formation, (3) owing to steric and electronic effects, the reaction usually leads to one predominant product. These considerations prompted us to 51 investigate the reactions of aryl Grignard reagents with 1,2,3-trihalobenzenes. 2,6-Dibromoiodobenzene84 in dry THF was added to three (or more) equivalents of phenylmagnesium bromide at room temperature. After aqueous workup, the reaction gave m- terphenyl in excellent yields (80-90%) (eq. 50). Using an iodine quench under otherwise similar conditions, 2'-iodo-m- terphenyl85 was isolated in 88% yield (eq. 51). Hence, it is clear that the reaction product prior to electrophilic quench possesses one reactive organometallic functionality between the two newly-created aryl-aryl bonds in the central ring. I - O @ BT95” ”29 o $— -—-b- r.t, 4h _ (eq. 50) 94 95 I 9 I O 57— -——>- r.t. 4h (eq. 51) 94 96 0.1. mechanism. We postulate the reaction mechanism shown in Scheme 12. 2,6-dibromoiodobenzene 94 should undergo halogen-metal exchange preferentially at iodine to give Grignard 97. If this mono-Grignard were to eliminate magnesium bromide to form aryne 98, trapping should give 99 rather than its regioisomer 99'. Nucleophilic addition to 3- bromobenzyne is known to occur predominantly in this manner 52 as a consequence of electron-withdrawal by the bromine substituent.121 This addition mode sets the stage for a second aryne raction, and the resulting 3-ary1aryne 100 should again add the nucleophile meta to give 101 instead of its regioisomer 101', probably for both electronic and steric reasons. This reaction mechanism, which was anticipated, leads only to m-terphenyl products. Scheme 12 I MgBr . Br©8r AngBr Br©8r 4493;: Br@ 94 97 98 AngBr MgBr MgBr A' A’ A AngBr ©A’ -M9Br2 Br‘©’~ 101 Ar 10° Ar 99 ”(o)" 8'67“” ’ I 101 99 While studying the mechanism of this reaction, we observed an interesting phenomenon. when only one equivalent of phenylmagnesium bromide was added to a solution of 2,6- dibromoiodobenzene with a reaction time of two hours, aqueous workup gave mainly 1,3-dibromobenzene 102, together with a trace of o-bromoiodobenzene 103. The usual product, m-terphenyl 95, was not observed (eq. 52). However, if two I I Br /Br 1 eq. PhMgBr H3 0+ Br /Br )- —> (eq. 52) r.t. (major) (trace) 94 102 103 53 more equivalents of phenyl Grignard were added before workup, the major product was m-terphenyl 95 (eq. 53). Thus r.t. r.t. ' + 1 eq. PhMgBr 2 eq. PhMgBr H30 @ @ (eq. 53) 94 > > > © 95 (50%) it appears that the elimination of magnesium halide from the intermediate 2,6-dibromopheny1magnesium bromide, to generate 3-bromobenzyne, was facilitated by the presence of excess Grignard reagent. In a relevent example, 2,6- dibromophenylmagnesium bromide (prepared by the exchange reaction between 2,6-dibromoiodobenzene and ethylmagnesium bromide) was mixed with excess pentachlorophenylmagnesium bromide at room temperature. After aqueous workup, we obtained only 1,3-dibromobenzene 102 and pentachlorobenzene 71 as a result of quenching each Grignard reagent (eq. 54). C1 C11 . ”C1 .. ”a.“ 2.: to) KS}: c1 c1 r.t. nun 102 71 This implies that 2,6-dibromophenylmagnesium bromide is quite stable under these reaction conditions. These observations lead to the conclusion that the presence of excess Grignard is necessary for elimination of magnesium halide to form an aryne intermediate. In the first two 54 chapters, we have described a catalytic effect of strong bases, such as LiTMP and t-BuOK, on the p-terphenyl and biaryl syntheses. So far we are not quite sure about the relation between the new observations and the previous data. Further studies on the mechanistic aspects of these reactions should help us to understand the reactions in more detail. D.2. Synthetic scope. 2,6-Dibromoiodobenzene 94 was readily prepared in quantity from 2,6-dibromoaniline88 via diazotization and treatment with potassium iodide. A solution of 94 in THF was added dropwise to somewhat over the theoretically required three equivalents of aryl Grignard in the same solvent, usually at room temperature but in some instances at reflux. After additional stirring for a few hours, the mixture was quenched with dilute aqueous acid. Results are summarized in Table 5. In general, the yields of m-terphenyls are quite good, the only poor example being entry 11. The product structures were clear from spectral data and from comparison of melting points with literature values, except for 104, 109, 110 and 113 which are new compounds. 1,2,3- Tribromobenzene and 2,6-dichloroiodobenzene can be used in place of 94. They gave somewhat lower, but still good, yields of 95 with phenylmagnesium bromide. The value of our method can be seen by comparison with literature syntheses of the known compounds in Table 5. Woods developed what appears to be the best general route to 55 Table 5. m-Terphenyls from Aryl Grignard Reagents and Tribaloarenes trihalo- Grignard yield entry arene reagent conditions product % ref. rt,3h @ : 3 77 81 H W 2‘ W a I O (D a 94 95 Br 2 3, :‘ Br @MqBr rt,15h 95 61 184 3 184 @- MgBr reflux, 3h 95 64 rt,24h 95 53 <92 C? 5 94 @MqBr rt,3h @ a 70 .7 9. WM... 9 g g 9 77 .7 7 94 rt,5h @ 9 Q G 0 62 90 Mg Br 56 Table 5 . (Cont'd) 10 11 12 13 14 94 94 . OCH3 94 @MqBr reflux, 10h 94 94 94 94 reflux,3h MgBr CH30 -@- MgBr rt, 5h OCH, MgBr reflux,10h CH30 MgBr M98! -7. 9 @ reflux,5h @ MgBr @ @ CHSO @ 9 107 © @ 111 85 57 93 91 92 92 57 m-diarylbenzenes, via the enol ether of dihydroresorcinol (eq. 55). Using this three-step sequence, with Ar =Ar' =p- 89 biphenylyl, he obtained a 72% yield of 105; our yield is comparable and requires only one step. In other cases, his method gave significantly lower yields than ours (for 106,90 91 92 17%; for 108, 18%; for 111 and 112, 23%). The advantage 0 0 Ar' Ar' a ArLi A @ Ar'Li 4r @ Pd/C +2 @ (eq. 55) OR - Ar Ar Ar of the Woods synthesis is that the added external aryl rings may be different from one another since they are incorporated in a stepwise sequence. In this sense, the method is more versatile than ours. On the other hand, if the two external rings are identical, our synthesis is shorter and qenerally gives superior yields. Another advantage of our method is that substituents are easily incorporated on the central ring. The actual reaction product is the m-terphenyl Grignard 101 which, in principle, can be treated with various electrophiles other than the proton. In practice, this may lead to difficultly separable mixtures because some excess or unreacted aryl Grignard reagent is present and will also react with the added electrophile. Quenching with a halogen (bromine or iodine), however, works well. For example, 2'-iodo-m- terphenyl 96 was obtained in 88% yield as described in the previous section. This route to 96 is superior to a three- step literature route involving an Ullman cross-coupling.85 58 Of course, 96 can then be reconverted to Grignard reagent 101 and treated with other electrophiles. This two-step sequence is sometimes cleaner and preferable to direct quenching with complex electrophiles. For example, quenching of the reaction of 94 and phenylmagnesium bromide with phenyl isocyanate gave anilide 114 in one step and 63% yield. The same product was obtained in two steps from 94 via iodoarene 96 in higher overall yield (88% x 85% =75% overall ) and with an easier workup (eq. 56). NHc.H, I 2' CGHSN'C'O @ 2. CsHsN-C-O 96 ‘ 2. I2 9“ (eq. 56) 114 Substituents can also be incorporated elsewhere in the central ring by starting with an appropriately substituted derivative of 94 (eq. 57). For example, some m-terphenyls with a halogen (bromine or chlorine) at carbon 5 in the central aromatic ring were prepared in good yields by the reaction of aryl Grignard reagents with 1,3,5- trihaloiodobenzenes (Table 6). I I Ar Ar X z’x reflux B C+ + AngBr ——-> -L-> (eq. 57) THF X X 59 Table 6. S-Halo-l,3-diarylbenzenes from Aryl Grignard Reagents and 1,2,3,S-Tetrahalobenzenes tetrahalo- Grignard . yield arene reagent product % @‘MQB' @ 30 c: | c: g 9 60 C! c; 116 118 g G 42 60 Further exploitation of these halogen substituents by converting the halo-m-terphenyls to the corresponding Grignard reagents and repeating the sequence with trihaloiodobenzene provides an efficient route the oligoarenes. For example, the 5'-bromo-m-terphenyl 117, prepared from the phenylmagnesium bromide and tribromoiodobenzene 115, was converted to the corresponding Grignard reagent and treated with 115 again to give a 42% yield of the 5"-bromo-5',5"'-diphenyl-quinquephenyl 120 which has seven aromatic rings. Thus, it only requires two steps to synthesize 120 starting from materials with a single aromatic ring. This simple synthetic route may be generally useful for preparing polyphenyls, which are widely studied for their electrical conductivities94 and thermal stabilities.95 Scheme 13 We! a) 115 9 @a) Hg 9 @ ___> © © b) "30+ at b) 115 ——>@ G a g 0 0) 330+ 117 120 For the construction of large molecules, synthetic pathways that allow the frequent repetition of similar steps are advantageous. For example, a recent synthesis of arborals96 and starburst dendritic polymers97 clearly demonstrated the synthetic advantages of a cascade-like sequence (Scheme 14). Our synthetic route, shown in Scheme 61 13, follows a similar repeating-step principle and can be readily adapted to the synthesis of new polyphenyls with a particular structure. The main difference in the two approaches is the site where the manipulation take place. It is at the side arms of a "seed" molecule in the cascade-like Scheme 14 _ a: Q ”65"", 9 ”°C‘C°ZE"3 \ ccr4 5' C6H6.DMF :7 - Br C(COZEH3 : :; Hhm(040H) 2 (EtOZC)3C -—— 2 3 \ chos, omso‘7 ' CHX§EH3 ::;3) R A route whereas it is at the tetrahalobenzene, which is eventually transformed into the central ring of the final molecule in our reaction. Both methodologies are useful from a synthetic viewpoint and deserve extensive study. The reaction was extended to molecules with two vicinal trihalo moieties. For example, slow addition of 3,4,5,3',4',5'-hexabromobiphenyl98 121 in THF to a solution of phenylmagnesium bromide in THF at reflux gave, after aqueous quench, the anticipated 5',5"-diphenyl-quaterphenyl 122 in 79% yield (eq. 58). MgBr 122 The reaction of aryl Grignard reagents with vicinal trihalobenzenes, as described here, provides a simple one- step route to m-terphenyls. The method has wide general applicability, the surface of which has just been scratched by the work described in this thesis. 63 E. Reactions of Polyhalobenzenes with Vinyl, Alkyl, Acetylenic and Heterocyclic Grignard Reagents. The reactions of several types of polyhaloarenes with aryl Grignard reagents described in the preceding chapters constitute useful methods for aryl-aryl bond formation. Extension of these reactions to the construction of multiple aryl-carbon bonds between aromatic substrates and other carbon moieties, by using other types of Grignard reagents, was studied in an exploratory way. In this chapter, some preliminary results of the reactions of selected vinyl, alkyl, acetylenic, and heterocyclic Grignard reagents with polyhaloarenes are presented. The mechanism of the reaction of an aryl Grignard reagent with a polyhaloarene follows these important sequential steps: (1) halogen-metal exchange to form a polyhalogenated aryl Grignard reagent, (2) elimination of magnesium halide to form the aryne, and (3) nucleophilic addition of the Grignard reagent to the aryne. In cases of multiple bond formation, the last two steps are repeated. It is evident that halogen-metal exchange plays a crucial role in triggering the reaction. From a thermodynamic vieWpoint, the ability of two reactants to undergo halogen-metal exchange might be determined by two factors. One of these is the inductive 64 effect of substituents on the reactants; that is, electron- withdrawing groups can stabilize the resulting negative charge. The other important factor is the difference in carbon hybridization of the two reactants. In general, the greater the s-character at the carbanionic center, the more stable it will be. The importance of substituent inductive effects is seen in the reactions of unsubstituted aryl Grignard reagents with polyhaloarenes. Metal transfer from the aryl Grignard to the polyhaloarene is thermodynamically favorable, because the adjacent halogen substituents are capable of stabilizing the resulting negative charge. For example, the reactions in equations 59 and 60 proceed in the forward direction because the product Grignard is favored over the reactant Grignard due to the inductive effect of the halogen substituents in the product Grignard. In these examples, the carbon- hybridization is identical in the reactant and product Grignards ( i.e., sp2 ). ‘1: + ._,.gar ‘14:?” 59, 30 123 I l 3981' N Br I 94 124 65 Extension of these reactions to other types of Grignard reagents was initiated with vinyl Grignard reagents because of their carbon hybridization is identical with that of aryl Grignard reagents. We studied the reaction of styrylmagnesium bromide 125 with polyhalobenzenes such as o- bromoiodobenzene 30, 2,6-dibromoiodobenzene 94, 1,4-dibromo- 2,5-diiodobenzene 1 and 3,6-dichloro-1,2,4,5- tetrabromobenzene 56. Each reaction was conducted in dry THF at room temperature, and the results are shown in eq. 61- 64. Distyrylbenzenes 12799 and 128,100 prepared by this method, were isolated as the E,E isomers in 48% and 44% yields respectively. The synthesis of 1,2,4,5- tetrastyrylbenzene 129101 has been reported previously via a multiple step sequence, with an overall yield of only 28%. Our method required only one step and gave the desired product in a 49% yield. rt,2h CH'CHM n @— gBr+ @Br—p THF —,.__}__, @431 CH-@ (c-H,77'/.) (99° 61) l25 30 126 I 4’ 125 + Br\©,Br $1.11,, THF (eq. 62) 9‘ 127 Br I H O" I Br THF 6 (eq- 63) 2>440£ 66 C). l + Br 3: H O CH'CH CH'CH 125 + 1,9?! 3 @- m —@ Br Br HF @‘CH'O' CH'CH‘@ (eq. 64) 4905 129 C1 56 In a similar manner, 1,2-diphenylvinylmagnesium bromide 130 reacted with 94 or 1 in dry THF at reflux to give, after aqueous workup, 131 and 132102 in 64% and 60% yields respectively (eq. 65 and 66). However, the reaction of 130 with 56 failed to produce the highly sterically hindered compound 133 (eq. 67). I /Ph Br 1 /Br reflux “30+ /th ll ~> —> II II MA THF “A Am(eq. 65) 136 131(64%) _ a. \ reflux H 0+ /u\ 130 -———)-—3-—>- Ph- @ l"‘(eq. 66) THF \'/ \ PB 1 132 (54%) (\th Br reflux B 0+ 130 3 )E/Ll';@ :\l/ Pb (eq. 67) want 55 133 It is clear from these preliminary results that the reaction of vinyl Grignard reagents with various polyhalobenzenes can be a useful synthetic method for 67 preparing polyvinylarenes. Extension of this methodology to other vinyl Grignards, including those with substituents that provide latent functionality, should be fruitful. The transposition of negative charge from a less electronegative sp3 carbon to a more electronegative sp2 carbon should be thermodynamically favorable. We may take advantage of this thermodynamic preference. For example, the reaction of one equivalent of 2,6-dibromoiodobenzene 94 with one equivalent of ethylmagnesium bromide 134 gave mainly, after aqueous workup, 1,3-dibromobenzene 102 (eq. 68). This I i + Br /Br 1 eq. EtMgBr a 0 Br\ Br r.t. (eq. 68) 94 102 result indicates that the desired exchange occurs quite readily. With 3 equivalents of ethylmagnesium bromide under the same conditions, a 20% yield of the anticipated 1,3- diethylbenzene 135 was observed (eq. 69). The product was identified by gas chromatographic analysis. Although the reaction yield was not good, this result suggests that further investigations, including a study of reaction conditions, should be fruitful. f .7 Br Br 3 eq- Mar, 3 0 at Rt. 68 For thermodynamic reasons, the transfer of negative charge from a more electronegative sp carbon to a less electronegative sp2 carbon is not favorable. For example, in the reaction of phenylacetylenic Grignard 136 with o- bromoiodobenzene 30, the o-dihalobenzene was recovered unchanged (eq. 70). The inductive effect of the halogen substituent is insufficient to overcome the hybridization change, and the reaction fails. I _ THF @CzC'MQB' 4' @Br no reaction (eq. 70) 136 _ 30 But to construct an aryl-carbon bond between the acetylenic moiety and the aromatic substrate, the problem of the inability of the acetylenic Grignard to initiate metal- halogen exchange needed to be solved. A simple solution for this problem was to use another Grignard reagent to undergo halogen-metal exchange. The best candidate to satisfy this need is an alkyl Grignard reagent (eg. 69). Thus, when ethylmagnesium bromide was added to 2,6-dibromoiodobenzene in the presence of excess phenylacetylenic Grignard at room temperature, aqueous workup gave the expected diphenylacetylene 137 in 61% yield (eq. 71). This procedure also proved to be effective with an iodine quench (eq. 72). 69 (011 EtMgBr ' Ho" Br slovoddmon '9 c..c @ 5'” 3° 3h,rt,THF 137 1: O'CJN @C=C;@ 50% (eq. 72) . 138 This reaction modification demonstrates the use of two different Grignard reagents for different purposes, one to bring about halogen-metal exchange and the other to trap the aryne. Its synthetic merits are two fold. First, it allows Grignard reagents that cannot undergo halogen-metal exchange to be used to form the desired aryl-carbon bond. Second, the quantity of Grignard reagent used to trap the aryne can be reduced to the stoichiometric amount without the loss of one equivalent of reagent for halogen-metal exchange. Following a similar procedure, the reaction of phenylethynylmagnesium bromide with 2,6-dibromoiodobenzene 94 and with dibromotetrachlorobenzene 56 gave 1,3-bis- 103 phenylethynyl-benzene 140 and 1,2,4,5-tetrakis- 101 phenylethynyl-benzene 141 in 38% and 25% yields respectively (eq. 73 and 74). I 136 + Brté/Br EquBr 15h H 0+ @Ct @ .0; -—L.. sknvoddnmm rt C (eq. 73) 3h,rt, THF 94 70 a. C01 19 C\ /C c: Cl ,, ‘c C' 136 + £3): 5 some,but> Cm (eq. 74) c) c) 24h.rt ‘ Cs 8' of Co 156 141 Attempts to optimize the reaction yields resulted in another modification. It was found, in the preparation of 140, that the slow addition of 2,6-dichlorophenylmagnesium bromide 139, (prepared from 2,6-dichloroiodobenzene and ethylmagnesium bromide at 0°C), to a THF solution of 136 gave, after aqueous workup, a higher yield of 140 (52%) (eq. 75) . MgBr , 136 + Cl Cl reflux H20 @C /C@ @ THF ‘c‘kz‘ (eq, 7,5, 139 140 The advantage of this methodology can be seen by comparing the present one-step synthesis of 141 from readily available starting materials with the multiple-step 104 outlined in the following equation. literature method, Tetramethylbenzene 142 was initially brominated, then converted to the corresponding phosphate ester 143. The subsequent reaction of 143 with benzaldehyde gave tetrastyrylbenzene 144. The bromination of 144 and dehydrohalogenation of 145 led to 141 in an overall yield of 13% (eq. 76). 71 O 0 CH CH ’6 - _ 3 3 a) NBS (OE )ZP CH2\ CH2 P(013402 - @ CH / CH b P 0E1: _ 3 3 ) ( )3 (OEt)2g CH2 CH2_g(OEt)2 142 143 Pn-C=C /C=C-Ph 3172 Benaldehyde’ \.\ Ph-0=C / C=C-Ph 144 Ph-(CHBr)2 (CHBr)2-Ph ‘ KO-t-Bu / \ 7’- 141 (eCI- 75) Ph—(CHBr)2 (CHBr)2—Ph 145 Extension of this methodology to heterocyclic Grignard reagents was expected to work, although some complications might be anticipated. For example, extensions to thienyl (or furyl) Grignard might be complicated by the change in ring size, which alters (slightly) the carbon hybridization of the Grignard reagent from that in carbocyclic aryl Grignard. Also, the electronegativity of the heteroatom might affect the metal-halogen exchange step. In fact, a preliminary study of the reaction of o- bromoiodobenzene 30 with 2-thienylmagnesium bromide 146 gave quite a complex product mixture. The products, after aqueous workup, included only a modest yield of the desired 2-phenylthiophene 148. There was a considerable amount of 72 recovered o-bromoiodobenzene 30, as well as low yields of 147 and 149 (eq. 77). The presence of 149 was very /Br t 21'! [10+ /Br r. O, [K + @ -—-—>-9—> S MgBr I . (60%)I 146 30 Br + o—Qo Q43 (4%) ( 15 20%) 16%) 147 ' 148 149 interesting. One rationalization for its formation is shown in Scheme 15. Initial halogen-metal exchange between 30 and Scheme 15 Br \ Br H 0+ \_ + S MgBr S I + I MgBr 150 123 —-HgBr2 “8 $97—53 <———— .I \MgBr 152 151 ! £30 or 150 149 146 would give 150 and 123. Elimination of magnesium halide from 123 gives benzyne 151, which is trapped by the thienyl 73 Grignard to generate the adduct 152. The possible halogen- metal exchange between 152 and 150 or 30 could lead to 149. At this point, it is somewhat premature to draw any conclusions about the reaction mechanism; further study is certainly required. Another example of the reaction between a heterocyclic Grignard reagent and a polyhalobenzene was carried out in a somewhat different manner. The heterocyclic Grignard reagent was used solely to trap the aryne. The aryne precursor was generated in a separate step by metal-halogen exchange with an alkyl Grignard reagent. Thus 2,6-dichlorophenylmagnesium bromide 139 (prepared from 2,6-dichloroiodobenzene and ethylmagnesium bromide at 0°C) was added dropwise to a solution of 2-thienylmagnesium bromide in dry THF at 60°C, with stirring for an additional 1.5 hours at the same temperature. Aqueous workup and column chromatography gave the desired 1,3-dithienylbenzene 155 in 41% yield (eq. 78). The product structure was mainly confirmed by it NMR spectrum, in which the isolated proton at the 2-position of the central ring in 155 coupled with two meta protons in the same ring and showed a triplet signal at 6 7.88. To establish that the immediate precursor of 155 contained an organometallic functionality at the 2 position of the central ring, the reaction was quenched with D 0 (eq. 79). 2 The NMR spectrum of the product 155' now lacked the triplet signal at 5 7.88. 74 HgBI’C \\ +1: reflux H 0+ 6 (EQ- 78) MgBr 139 146 ' 155 Our preliminary studies of the reaction of polyhalobenzenes with vinyl, alkyl, acetylenic and heterocyclic Grignard reagents demonstrate some features that deserve comment. (1) The initial and crucial step for the reaction is halogen-metal exchange, which depends on thermodynamic factors, and in particular on the carbon hybridization at the negative carbon of the Grignard reagent. (2) The exchange can be initiated using an alkyl Grignard reagent. (3) The quantity of Grignard used to trap the aryne can be limited to the stoichiometric amount without loss of one equivalent for exchange. (4) Although several different types of Grignard reagents were employed, the mechanisms and the incorporation of organometallic functionality in the product follow the same pattern as discussed in the first four chapters of this thesis. 75 F. An Unusual Diels-Alder Reaction Between An Anionic Diene and Benzyne During our study of the reactions of polyhaloarenes with Grignard reagents, two instances of unusual by-products were observed. The structures of these by-products suggested that they were formed via [4+2] cycloaddition reactions (Diels-Alder reactions). Furthermore, the diene component in these reactions appeared to be aryl or heteroaryl Grignard reagents. These novel findings are described in this final section of the thesis. F.1. The Reaction Between mesitylmagnesium bromide and Hexabromobenzene. Using the typical procedure, the reaction was carried out by adding hexabromobenzene to an excess of mesitylmagnesium bromide 157 at room temperature. An aqueous workup gave, besides the expected 1,2,4,5- tetramesitylbenzene (30%), a 20% yield of a biscycloadduct assigned the overall structure 158 (eq. 80). _ Br Br /,Br r.t. H+ + -————%>--——¥> Br Br ‘ THF Br MgBr 55 157 76 Br . 1 ‘ I , l ’ 4 ‘ x 0 Br 158a 158 b The product was in fact an inseparable mixture of two regioisomers, 158a (which has a C symmetry axis), and 158b 2 (which has a mirror plane). The structures were assigned on the basis of mass and NMR spectral data. The mass spectrum of the mixture showed the presence of two bromine atoms (m/e 470, 472, 474). The proton NMR spectrum showed peaks corresponding to four vinyl protons (J 6.00), two bridgehead protons (J' 4.98), two bridgehead methyl groups (5 2.12) and four vinyl methyl groups (cf 1.89). The products seem to arise from the [4+2] cycloaddition of 3,6-dibromo-1,4-benzadiyne (or its synthetic equivalent) to two eqivalents of either mesitylene of mesitylmagnesium bromide. To determine whether the hydrocarbon or the Grignard reagent provided the diene component, the reaction was quenched with deuterium oxide or with methyl iodide (eq. 81 and 82). The products were 159 and 160 respectively; their NMR and mass spectral data are summarized in Table 7. The molecular ion of 159 is two mass units higher than that of 158, which clearly shows that two organometallic functionalities are present in the product prior to electrophilic quenching. By comparing the proton NMR spectrum of 159 with that of 158, we can see that two vinyl 77 . Br I Br /Br + r.t. 81) Br Br THF Br MgBr 55 157 159 Br ff}; /« 0‘40 (eq. 82) 160 Table 7 mass and NMR spectral data of biscycloadducts biscycloadduct 158 159 160 molecular ion mass 472 474 500 NMR (in J ) (vinyl) H 6.00 (4H) 6.00 (2H) 5.90 (2H) (bridgehead) H 4.98 (2H) 4.98 (2H) 4.96 (2H) (bridgehead) CH3 2.12 (6H) 2.12 (6H) 2.12 (6H) (vinyl) CH3 1.89 (12H) 1.89 (12H) 1.89 (6H) 1.81 (6H) 1.67 (6H) 78 protons are replaced by deuterium. Thus the organometallic functionality is located at the vinyl positions. The results of methyl iodide quenching confirm this conclusion. Therefore the mesityl Grignard reagent is indeed involved in the Diels-Alder cycloaddition and cycloaddition occur across C2 and C5 of that reagent. An interesting question arises. We can see (Scheme 16) that there are two paths by which an aryne can approach the mesityl Grignard reagent. Path a will generate a product with a organometallic functionality at the bridgehead, whereas path b will place the that functionality at the double bond. The results show that path b is favored. We may now try to rationalize this preference. Scheme 16 a i b ,/ l / path a path b ‘giil’, 0’0 <— —* ‘0 \ - /’ R. N Br I ' M Br R 9 / MgBr 9 b 161 1 162 a The mechanism of the Diels-Alder reaction is still controversial.105 The reaction can proceed in a concerted or a non-concerted fashion, and even if concerted (the usual path), the new bonds can be formed in a synchronous or a nonsynchronous manner. Even though the reactions being 79 discussed may occur in a concerted manner, it will be useful here to consider the possible zwitterionic forms of the transition states. These are shown below, excluding those structures which place a negative charge on the diene, because the diene already carries a substantial negative charge. The transition states for path a are 163 and 164,whereas those for path b are 166 and 167. Transition C”) Q 1...... 164 165 166 167 states 163 and 166 suffer from a steric interaction between the aryl methyl group and the incoming benzyne, and should be disfavored. If transition state 164 were involved, it should collapse to the more stable 165, which is not related to the cycloaddition process. Therefore, 167 would seems to be the most favorable transition state, and would lead to 80 the observed regiochemistry in the product. Also, the methyl groups in 167 are located ortho and para to the new bond and should stabilize the partial positive charge that is present. Finally, this transition state leads to a product with the negative charge at an sp2 carbon, as contrasted with path a transition states, that would give a product with a negative charge at an sp3 (bridgehead) carbon. To the extent that the product energies are reflected in the transition states that lead to them, path b should be favored, as observed. Barry and coworkers49 reported that tetrabromobenzyne reacted with mesitylene to give a cycloadduct 169 in a 10% yield. In the present reaction, the biscycloadducts are obtained in a somewhat better yield (20%). To the extent that any direct comparison is possible, it would seems that the presence of an anion in the diene component may facilitate cycloaddition to some extent. It has been reported theoretically and experimentally that the Diels- Alder reaction can be greatly improved by the introduction of an electron-donating substituent on a diene. Since carbon-magnesium bond is substantially ionic, this moiety acts as an electron-donating group, even though the orbital bearing the negative charge is, to a first approximation, orthogonal to the diene"? orbitals. MgBr © 8 Br Br Br \ \ r a '——-> ____> “a 8" (eq° 83) Br Br Br Br 0 Br Br Br Br (10%) 81 One other factor may be the steric crowdedness of having three adjacent substituents in the mesityl Grignard reagent. Relief of strain on cycloaddition may make a mesityl Grignard reagent more reactive than mesitylene itself as a diene. Clearly there are many unanswered mechanistic questions regarding this unusual cycloaddition reaction. Two reactions were carried out with the hope of extending the reaction scope. The reaction of 2,6- dimethylphenylmagnesium bromide 170 or pentamethylphenyl- magnesium bromide 171 with hexabromobenzene was carried out under the same conditions. Disappointingly, no Diels-Alder biscycloadduct was found in either case. . Br Br /Br r.t. 11+ + ———>-—> NO Cycloadduct (eq. 84) Br Br THF Br MgBr 55 170 r.t. H+ 55 + @ .——>-—-> No cycloadduct (eq 85) THF ° l HgBr F.1. The Reaction of 2,6-Dichlorophenylmagnesium bromide with 2-Thieny1magnesium bromide. The reaction of 2-thienylmagnesium bromide with 2,6- 82 dichlorophenylmagnesium bromide described in chapter five gave, after aqueous workup, a 6-10% yield of 1- chloronaphthalene in addition to the anticipated bisadduct 155 (eq. 86). The presence of l-chloronaphthalene 172 is MgBr c1 I . C1 C1 \ reflux H 0+ + / ——>—3—-> + 155 (eq- 36) S MgBr THF 172 139 146 C]. D C]. .. I32°, + @@ + 155’ (eq. 87) D 173 174 attributed to the occurrence of a [4+2] cycloaddition followed by sulfur extrusion. A plausible mechanism is proposed in Scheme 17. 3-Chlorobenzyne 175 could add to the 2-thieny1 Grignard reagent to give cycloadducts 176 and/or 177. Spontaneous cycloreversion with loss of sulfur would then give naphthalenic Grignard 178 and/or 179. The hydrolysis of 178 and/or 179 would then give the observed l-chloronaphthalene. The involvement of 2-thienylmagnesium bromide in the cycloaddition was confirmed by quenching the reaction with deuterium oxide. The GC-MS analysis of reaction mixtures illustrated the presence of monodeuterated chloronaphthalene, which may be either 173 or 174 or a 83 Scheme 17 Cl “QB: 146 1.39 ———> @| __> MgBr 176 177 L i + . . 1130 1 | 172 <———1 MgBr D20 |il® @ll+é 173 + 174 ‘ MgBr 178 179 mixture of those. Attempts to isolate and identify the particular deuterated chloronaphthalene were not successful (eq. 87). The failure to obtain cycloaddition intermediates 176 and 177, which would have provided direct evidence for cycloaddition, is attributed to the rapid retro- cycloaddition in which a sulfur atom is extruded. This is a general phenomenon for cycloadditions involving thiophene 106 derivatives as the diene. For example, the reaction of tetrafluorobenzyne and thiophene gave tetrafluoronaphthalene 107 But transient NMR 183 instead of cycloadduct 182. absorptions attributable to the bridgehead and vinyl protons of the intermediate 182 were observed. Therefore, we believed that the formation of l-chloronaphthalene is correctly rationalized by Scheme 17 (eq. 88). 84 X X X X X X 010w 00 00s X S X X X X X 180 181 183 X I F 183 X - F Relevant to our finding, anion-assisted pericyclic reactions have attracted a great deal of attention recently. Examples include Ireland's ester enolate-Claisen rearrangement,108 Evans' anionic oxy-Cope rearrangement10 110 9 and the carbanionic Claisen rearrangement. For the Diels- Alder reaction, no such report has yet been published. 0n the other hand, an anionic-accelerated retro-Diels-Alder reaction was reported as early as 1967 by our group.111 A recent quantitative measure of the rate enhancement of retro-Diels-Alder reactions shows that the presence of an alkoxide group can increase the rate by a factor of 106 (Scheme 18).112 Scheme 18 Free Enthalpy of Activation (05') for (402)-Cyctoreversi0n E E .__. + $10k ”and En SE) 0 .. E ’ E E _. + 26.0 kcallmol .Q 0 E 05mm3 OSiMe3 W —. E13 + 0 s20 kcallmol E E 9 e 0 e E=C02Me 85 In summary, two unusual cycloaddition reactions involving Grignard reagents as the diene component have been encountered. The possible role that anionic substituents placed directly on the diene component might play in Diels-Alder cycloadditions is clearly a subject worthy of further study. 86 Experimental 1H NMR spectra were determined 1. General procedures. on a Varian T-60 or Bruker WM-250 spectrometer in CDCl3 solution containing tetramethylsilane as the internal standard. Chemical shifts are reported in J units. Infrared spectra were obtained on a Perkin-Elmer 167 spectrometer, using KBr pellets. Mass spectra were recorded at 70 eV on a Finnigan 4000 spectrometer or, for high resolution, a Varian CHS spectrometer at the Michigan State University Mass Spectrometry Facility supported by a grant (RR-00480) from the Biotechnology Resources Branch, Division of Research Resources, NIH. Melting points were determined on a Thomas- Hoover Unimelt apparatus and are uncorrected. Anhydrous magnesium sulfate was the drying reagent throughout, and the silica gel for chromatography was 230-400 mesh. Analyses are by Guelph Chemical Laboratories, Ltd. and by Spang Microanalytic Laboratory. 2. 1,4-Dibromo-2,5-diiodobenzene 1. A solution of 1,4- dibromobenzene (5.8 g, 24.6 mmol) and iodine (24 g, 94.5 mmol) in 80 mL of concentrated sulfuric acid was vigorously agitated with a magnetic stirrer while the reaction mixture was held at 125-1300C for 6 h. The mixture was poured into 87 ice-water and the precipitated crystalline solid was filtered and washed successively with aqueous sodium bisulfite, sodium bicarbonate and water. Recrystallization from benzene gave 8.4 g (70%) of 1 as white needles, mp 163- 165°C (lit. 161-163°C) 1 H-NMR 6 7.97 (5); mass spectrum, m/g (relative intensity) 490 (28), 488 (54), 486 (M+, 28), 363 (11), 361 (18), 359 (9), 236 (15), 234 (30), 232 (17), 155 (25), 153 (26), 74 (100). 3. 1,5-Dibromo-2,4-diiodobenzene 18. The procedure and workup is essentially the same as for 1. From 5.9 g (25 mmol) of 1,3-dibromobenzene, 12.7 g (50 mmol) of iodine in 80 mL of concentrated sulfuric acid there was obtained 7.9 g 1H-NMR 6‘ of 18 as white needles from benzene, mp 166-167°C 7.78 (s, 1 H), 8.17 (s, 1 H); mass spectrum, m/g (relative intensity) 490 (40), 488 (81), 486 (M+, 41), 363 (14), 361 (24), 359 (11), 236 (18), 234 (34), 232 (18), 74 (100). Anal. Calcd for C6H2Br2 2: C, 14.78; H, 0.41; Br, 32.77; I, 52.04. Found: C, 14.64; H, 0.34; Br, 32.73; I, 51.96. 4. General Procedure A for p—Terphenyl Synthesis. A solution of 1,4-dibromo-2,5-diiodobenzene (2.44 g, 5 mmol) in 40 mL of THF was added slowly over 30 min to a solution of arylmagnesium bromide (prepared from 25 mmol of aryl bromide, 27.5 mmol of magnesium in 60 mL of THF), and the mixture was stirred for an additional 2 h at room temperature. The reaction was quenched with ice and dilute 88 hydrochloric acid and the mixture was extracted with chloroform. The organic layer was dried and the solvent was evaporated under reduced pressure to give a mixture of a solid and an oil, which was washed with hexane and filtered to give the first crop of terphenyl product. Chromatography of the filtrate using hexane as the eluent gave the second crop of terphenyl. The yields reported in Table 1 are based on the sum of two crops of the products collected. 5. Iodine Quench; 2',5'-diiodo—p-terphenyl 4. The general procedure A for p-terphenyl described above was followed, but before quenching with ice, the reaction mixture was cooled to 10°C, 7.6 g (30 mmol) of iodine was added and the mixture was stirred at 10°C for 1 h, then warmed to room temperature. It was successively treated with ice-water, extracted with chloroform and the organic layer was dried and evaporated under reduced pressure. The residue, a solid and oil, was washed with a little benzene and filtered to give mainly diiodo-p-terphenyl. Recrystallization from benzene gave 1.25 g of pure 4 as white needles. Chromatography of the filtrate using hexane as eluent gave 0.7 g (29%) of 1 and an additional 0.1 g of 4 (overall yield 57%); mp 262-2630C, 1 H-NMR 6 7.34- 7.50 (m, 10 H), 7.88 (s, 2 H); mass spectrum, m/g (relative intensity) 483 (10), 482 (M+, 59), 241 (13), 228 (100), 227 (24), 226 (72). Anal. Calcd for C H I 18 12 2: c, 44.84; H, 2.51. Found: C, 44.92; H, 2.60. 89 6. 2,4,6,2",4',6"-Hexamethyl-p-terphenyl 25. The general procedure A was followed. For 25: 1H-NMR52.06 (s, 12 H), 2.34 (s, 6 H), 6.96 (s, 4 H), 7.16 (s, 4 H); mass spectrum, m/e (relative intensity) 315 (23), 314 (M+, 100), 157 (22), 142 (20), 133 (30). Anal. Calcd for C H 24 26‘ c, 91.72; H, 8.28. Found: C, 91.96; H, 8.30. 7. 1,4-[1',1“-Dinaphthyl]benzene 26. The general procedure A was followed. For 26: lH-NMR.J 7.18-8.0 (m); mass spectrum, m/g (relative intensity) 330 (M+, 100), 202 (30), 163 (27); high resolution mass spectrum; Calcd for C H 8. Reaction of 18 with phenylmagnesium.bromide. A solution of 18 (2.44 g, 5 mmol) in 20 mL of THF was added over 15 min to a stirred solution of phenylmagnesium bromide (20 mmol) in 80 mL of THF. The mixture was stirred at room temperature for 2.5 h, then quenched with ice and dilute hydrochloric acid. Extraction with chloroform, drying and evaporation of the solvent gave a solid mixed with an oil. This mixture was washed with a little hexane and filtered to give 0.16 g of nearly pure p-terphenyl. The filtrate was chromatographed using hexane as eluent to give 0.72 g (46%) of 1,3-dibromo-4-iodobenzene 20, mp 45-46OC (lit.6O mp 45-46OC) and 0.32 g of terphenyls. The separation of p- and m-terphenyls by column chromatography was difficult, so the 90 mixture was analyzed by gas chromatography (SE-30 column, 1/4"x6", 180-300°C at 8°C/min) The procedure was standardized with authentic commercially available (Aldrich) P- and m-terphenyls and showed that a total of 0.41 g (37%) of p-terphenyl and 0.07 g (6%) of m-terphenyl was present. 9. The effect of lithium.2,2,6,6-tetramethylpiperidide on the reaction of 1 with phenylmagnesium.bromide. The reaction was carried out according to the general procedure A but prior to the aqueous quench a solution of LiTMP (5 mmol) in 10 mL of hexane was added over 30 min to the reaction mixture. After 1 h of additional stirring at room temperature, the reaction was worked up as usual. There was obtained 0.85 g (74%) of p-terphenyl. A similar procedure increased the yields of 22 and 24 from those shown in Table 1 to 74% and 71% respectively. 10. The effect of potassium t-butoxide on the reaction of 1 with phenylmagnesium bromide. The reaction was carried out according to the general procedure A but prior to the aqueous quench a solution of 5 mmol of potassium t-butoxide in 10 mL of THF was added over 30 min to the reaction mixture. After 1 h of additional stirring at room temperature the reaction was worked up as usual, to give 0.86 g (75%) of p-terphenyl and 0.14 g (8%) of 6. 91 11. General Procedure B for Biaryl Synthesis. Preparation of 4-Methylbiphenyl 40. A solution of o- bromoiodobenzene 30 (2.83 g, 10 mmol) in 20 mL of dry THF was added slowly over 90 min to a freshly prepared solution of p-tolylmagnesium bromide (from 3.42 g 20 mmol of p- bromotoluene and 0.48 g of Mg in 60 mL of THF). The mixture was stirred for an additional 2 h at room temperature, then quenched with ice and dilute HCl. The THF was removed under reduced pressure and the aqueous solution was extracted with chloroform. The organic extract was washed with sodium bicarbonate, water, and dried (MgSO4). Evaporation of the solvent left a brown oil (4.2 g) which was chromatographed on silica gel using hexane as the eluent to give 1.20 g 71 (71%) of 4-methylbiphenyl, mp 48-49OC (lit. value 49~ 50°C). 12. General Procedure C for Biaryl Synthesis. Preparation of 3,4-Dimethoxybiphenyl 45. The reaction was carried out according to procedure B, using 10 mmol of 4- bromo-S-iodo-veratrole and 20 mmol of phenylmagnesium bromide. Prior to aqueous quench, a solution of lithium tetramethylpiperidide (10 mmol) in 10 mL of hexane was added over 30 min and the reaction mixture was stirred at room temperature for an additional 3 h. Then aqueous workup was followed as uaual. The gradient elution of the reaction mixture on silica gel using a hexane and benzene mixture afforded 1.47 g (69%) of 3,4-dimethoxybiphenyl, mp 68-690C 92 (from methanol; lit.74 value 67-680C). 13. Preparation of z-bromobiphenyl 41 (Bromine quench). The reaction between o-bromoiodobenzene (10 mmol) and phenylmagnesium bromide (30 mmol) was carried out using procedure B, but instead of an aqueous quench the reaction mixture was added to 16 g of bromine in 15 mL of carbon tetrachloride cooled in an ice bath. The mixture was then warmed to room temperature, treated with 10% sodium bicar- bonate to destroy the excess bromine, extracted with chloro- form and worked up as usual to give 1.46 g (63%) of 2-bromo- biphenyl and about 3% of 2-iodobiphenyl. 14. Preparation of 2-iodo-4,5-dimethoxybiphenyl 46 (Iodine quench). Procedure C was followed as described above for the preparation of 45, but instead of an aqueous quench, 7.6 g (30 mmol) of iodine was added to the reaction mixture which was then stirred for an additional hour. Excess iodine was destroyed with sodium sulfite and the usual workup gave 1.97 g (58%) of 46 mp (methanol) 107- 108.5°c (lit.74 value 109°C). 15. 2-Iodo-4,5-dimethylbiphenyl 42. Procedure B was followed to afford a 65% yield of 42. B.p. 155-157°C at 1 Torr, 1H-NMR; 6 2.20 (s, 3 H), 2.22 (s, 3 H), 7.05 (s, 1 H), 7.28-7.40 (m, 5 H), 7.69 (s, l H); mass spectrum, m/g (relative intensity) 308 (M+, 100), 166 (53), 165 (72); 93 Anal. Calcd for C14H13I: C, 54.54; H, 4.22. Found: C, 54.72; H, 4.11. 16. 2-Iodo-3',4'-dimethylbiphenyl 44. Procedure B was followed to afford a 64% yield of 44. B.p. 157-158°C at 1 Torr, 1H-NMRHS 2.30 (s, 6 H), 6.94—6.99 (m, 1 H), 7.03- 7.18 (m, 3 H), 7.20-7.37 (m, 2 H), 7.90-7.94 (m, 1 H); mass spectrum, m/g (relative intensity) 308 (M+, 100), 166 (59), 165 (68), 154 (21); Anal. Calcd for C14H13I: C, 54.54: H, 4.22. Found: C, 54.33; H, 4.39. 17. 2-(2'-Iodophenyl)naphthalene 48. Procedure C was followed using hexane as the eluent in column chromatography 1H-NMR to afford a 60% yie1d1of 48. mp 80-82°C from benzene; 5 7.02-7.09 (m, 1 H), 7.38-7.44 (m, 2 H), 7.45-7.78 (m, 3 H), 7.82 (s, 1 H), 7.85-7.93 (m, 3 H), 7.97-8.05 (m, 1 H); mass spectrum, m/g (relative intensity) 330 (M+,39), 204 (100), 203 (46), 202 (72), 101 (26); Anal. Calcd for C16H11I: c, 58.18; H, 3.33. Found: C, 58.23; H, 3.34. 18. 1-(2'-Iodo-4',5'—dimethylphenyl)naphthalene 49. Procedure C was followed using hexane as the eluent in column chromatography to afford a 63% yield of 49. mp 145- lH-NMR: