" "';_I 4‘ £7.33)... | "51,27,756" Q33 fifi“$.$de<-u.‘ E¥1f *.: 3“». p. I'M «'ch ' ‘f,. . “A?!“ 7‘9” Imp-wk“ ‘ II. I'I Vega“ III ‘IJ-I Wv'.—’ :A‘ if, A -'>“' V . H" ‘ 5 fi“' from: '3*: ludfikf “J Ci: : I.) | “If“... ("I WI: 21.». .-. I‘I'IEI‘n'. "5“ on:| :hul . , , l 1%; } ”'.KI§ y; ' . m“. (P»‘ j #1, J'I, , .I'” &.'¥({’~’. .‘ .. .1 I a. ‘fifll. "' I'JJ- ‘ l' I . I S 1 :9 {‘1 16”»! f:;:.l|;1‘:l I. “V. . I ~ - '. -. P. x» mm n. ‘- '- (nu‘ljp Q‘IJ *Kfl‘l 'dz- , '1‘l .HIU' j ‘1?“ I“. “H _ H” '-. A I In...a. ~ UL.- .v&.m.x'-...u ' '4" . ’.'. ”N. ‘.I I In. . II; ,' u. . . ”‘ "M" Wfit’k‘fiznrfi. “3’3", .,.u:;.. “ J “ pf)” "3:; -.‘- |-. W“, l IINIIQHI‘ I 17 K ‘ ,I 'Lf'. é 4.’ . I ‘ - ‘. . n..r‘ ‘35 '. "3 '- »-‘..,,:_':' " .. ,.H,\“"(‘\I" s,,._ ‘33: f‘“ W 31 I I. , HE»; gs,“ “W"I“'IL.‘ ",2. , Jib“ 3; v \ “3.“: 4’ Arlyn? Ir f (I? “H “-l: :I'.I‘}‘l . 1 $11)? “‘1‘ \“g ' a 1”“, 'K...' r {J'NN " “ARM. '.. :‘: .})-}VL¥ A": 1"» fi 13?} IN . 4l"“_‘h,v (I I I m I.» .. III»... ». , .. H . .f‘WII-‘I‘I III-II M ’ .‘ . f I I“1J‘I“." _. 'H. {IVY ,. ‘ ‘II'I. I.“ " 1"]:IFI'3' gh'fl‘.‘ 'I‘ I" .' If.” .' K, ._ 0 . rl' «I '. WIWII‘ ‘1 ‘1‘: '.‘ ‘W. ”am? E ‘In. - .3. f) NJ -" .I Ill-“.Il 3?“, “2"“ .‘*¥ ' I. ' . n J. n~ “ :33.‘ .q‘ 3:.“ ‘ 31"} , .} mfia' " . . . . I . K. v Mi. '- s wflrfl ..; ”1". ‘ a" J 1- “105 "n"! “2' i K .I‘ ‘1 :4 :°u':': fl§w :tilllfi'll Ii‘g‘l 59“.". ugl‘4:\ ‘ ' uI ' I" ‘I " ' . .,‘. , , . ., .“ 1".“ 1, ‘4le“? W ‘ ‘ "L .1". . ‘ 'l'l“ . .~ I ' I ' 1 l , ‘ "I '> I' ll .LI‘} I“ ‘Fi " .1” NH; I. .‘ "i’:!:“l‘r“. V .1. j" 8 'z'lrllut'.’ ' I . '. .. .‘ . .: .‘III‘ ‘ . I- . I '9' 'I.'.,: . . I',‘ 3' In. _ -.. '.I1\' I (3:. ' . : ;.,',‘.II..';.u,';H;I u'. t. -, 1., - "I . I. ‘.- . Hz‘ufir- ;.,.I".‘.- .. 'L'I-fK'Iv‘.;"’v:‘. “3-. ; ww'j, ' '...'.v, 5:1,.- .. . H' ',-..9' .“-wt I'- z'. w 'u' 'IHI;MW.7::Ia ” k 1%. --:\'"~.¢-a .m%=¢wL .. ..,I WI H‘IH‘ ‘ I .l 1" .." ' "I I. . i. .I' “ VI." .. .‘ 14"» . v .' I h. .. .I .1. " and-1.? . ' p. .,. ' ‘ . '-; . -. ‘ . , ' ‘l o." II I. . ‘ K " V " "H" I". } I" ’I J ’. : ." ‘ l ‘ I H'v' .! .II‘. I‘ H W‘ 7‘6. . '. . . .I )I "I1 I I “ K 1‘ ' '~ ‘1 It‘i' . I: .'I' I , ' ' I. ’.l" I ‘ .' I I.. la." ' z: ' I' r' ‘l ' ' I n‘l , I" I'M-l...» . . .. .. .. .. . . - . ”l “I 1".”1. ' ‘ I l |'l‘g. ."J. .H '1‘, :IL. . I I' 'I I,“ “1" IIII‘,'|I 3).); H .. "l'..".,..". 3",] ‘ '.' .J‘ l‘ 5:. ‘r: I'.' I" ‘:| :“I‘I ‘u‘lll f" “‘6‘ INS” I ‘0‘!“ '5' M“I".I I‘. I" .- HI 5" . . v’..- '. ‘Iylm. .. .K' IF: 1 I" KI. ‘“ J. ”1% .‘I If|,ll',w..,‘ I .HIII‘,.'I|I".-V “if“: I'. ".!\.. . II. I, .v" II.) I ""1 ' In '3' H 'Il‘“. "-' M ' . ‘ ‘I'.’ "(K "NM ' . " 4.6:! ' 'f‘ m“ I W "I" h ‘. v,“ ' 1| .. " ‘I . ‘ 't ‘ ' , I, ‘ ,f' ”I'l‘} ,jIISE'. . I..." '-"’ ...". Wm "*"”;* ,.' R ” .:;wpri‘:fi M‘. ..d“LuM... 9'» w” $2 ' x " "K ‘ ’- H I .'. \K ‘I 'H' N: Kl “"“H‘J I "u ""0... ' ‘l " ‘ I V‘ .".""‘M #9“ ‘1' "l" I I’ " " ‘ I l. . ' .w‘. '. * . . . -;:‘."‘....,- ‘ ‘ \w .. a‘nI-"lrvz‘ "2|, :10'“I'\'s-:V‘I-;f.'4,!"".f L'jfi .n‘ ._..'J I" 'i $.39 33:4 IJIIII ’5'": I LI??QARY Micr- ‘-- 1 state l Um v ersity ¥ J This is to certify that the thesis entitled Part I. Applications of Bis-Aryne Equivalents to Organic Synthesis Part II. Attempted Synthesis of Thiophene or Furan Fused Radialene Analogues presented by Chung-Yin Lai has been accepted towards fulfillment of the requirements for Mdegree in Chemistry W 3+”? Major professor DateMMOL “A 'figl 0-7639 __ -———-‘___._‘ _ 4 V W, {[flflNx L ~.:I‘ in; x \‘W OVERDUE FINES: 25¢ per day per in. RETURNING LIQRARY MATERIALS: Place in book return to name charge from circulation records PART I APPLICATIONS OF BIS-ARYNE EQUIVALENTS TO ORGANIC SYNTHESIS PART II ATTEMPTED SYNTHESIS OF THIOPHENE OR FURAN FUSED RADIALENE ANALOGUES By Chung Yin Lai A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1981 ABSTRACT PART I APPLICATIONS OF BIS—ARYNE EQUIVALENTS TO ORGANIC SYNTHESIS PART II ATTEMPTED SYNTHESIS OF THIOPHENE OR FURAN FUSED RADIALENE ANALOGUES By Chung Yin Lai In Part I of this thesis, an efficient synthesis for per-substituted acenes was developed via bis-aryne equiva- lents (derived from the reaction of n-butvllithium with a tetrahaloarene such as tetrabromo-p-xylene, 3%). These underwent addition to various pyrroles, and subsequent N-bridge extrusion gave the desired acenes. A series of N and/or C substituted pyrroles were selected as dienes for the bis-annelation reaction. Bis- adduct formation was affected primarilytmrthe electron density of the diene system but not by the bulkiness of the substituents on C-2 or C-5 of the nyrrole. Thus, adducts Chung Yin Lai having bulky substituents, as for example bis(N-methyl)— 2,3,6,7,9,lO-hexamethyl-l,A,5,8-tetraphenyl-l,U,5,8-tetra- hydroanthracen-1,A; 5,8-bis-imine, were prepared without difficulty. Similarly, no change was found in the yield of bis—adducts with respect to the size of the substituents on the nitrogen of the pyrroles. Both N-iso-propyl-tetra- methylpyrrole and N-phenyl—tetramethylpyrrole reacted with bis-aryne equivalents to give a good yield of bis-adducts. Besides symmetric bis—adducts unsymmetric bis—adducts could also be prepared, by a stepwise process. When 5% was treated with one equivalent of n-BuLi and N-methylocta- hydrocarbazole a mono—adduct was isolated which was subse- quently reacted with another mole of n—BuLi and pentamethyl- pyrrole. An unsymmetric bis-adduct was thus prepared. With the use of an N-(dimethylamino)-pyrrole as the diene, the bis-adduct obtained became an excellent pre- cursor for per-substituted acenes. For instance in the presence of N-(dimethylamino)tetramethylpyrrole (éé) the reaction of neBuLi with p-dimethoxytetrabromobenzene or 2,3,6,7-tetrabromo—l,A,5,8-tetramethylnaphthalene provided bis-adducts which subsequently could be converted quantita- tively by pyrolysis to 9,lO-dimethoxyoctamethylanthracene or dodecamethylnaphthacene respectively. Octamethyl- naphthalene and decamethylanthracene were similarly prepared. Due to various limitations which plagued known methods for aromatizing 7-aza—bicyclo[2.2.ljheptadienes, this two step Chung Yin Lai process for peracene synthesis represents the most suc- cessful preparation developed to date. Cycloaddition of N-(dimethylamino)-octahydrocarbazole llg with a bis—aryne equivalent provided a short synthesis of 2,3-difunctionalized triphenylenes. Reaction of %%Q and 3% with one equivalent of n-BuLi gave a mono-adduct which was converted to l,A-dimethyl-Z,3—dibromotriphenylene by pyrolysis and dehydrogenation. The preparation of an unsymmetric peracene via this sequence was also studied. Thus bis(tetrahydrobenzo- [l,2;3,u])-5,6,7,8,9,lO-hexamethylanthracene was prepared in a stepwise manner from pyrrole 83 and octahydrocarbazole mm 110. mmm In Part II of this thesis, the synthesis of cyclo- buta[1,2-c;3,A—c'] dithiophene IQQ and its oxygen analog lgfl was attempted. The first approach explored was the coupling of thiophene moieties by the Ullmann reaction. However, neither the direct coupling of 3,A-dibromothiophene nor the closure of 3,3'-dibromo-A,A'-bithienyl appeared suitable for this purpose. Therefore another approach, which started from an intermediate which already contained the four- membered ring, was planned. Thus cyclobuta[l,2-c;3,A-c']- octahydrodithiophene 3%2 and cyclobuta[1,2-c;3,A-c']-octa- hydrodifuran Eli or similar intermediates but with the sulfur oxidized to a sulfone group (e.g., tricyclo[5.3.0.02’6]- 3,9—dithiadecane-3,3-dioxide Egg and tricyclo[5.3.0.02’6]-3- thia-9-oxodecane 223 became the intermediate synthetic targets. ’L’Vb Chung Yin Lai Preparation of all the above mentioned intermediates has been accomplished. Preliminary results revealed that sulfone gag and sulfide glg could not be readily dehydrogen- ated, nor could they be chlorinated by N-chlorosuccinimide. Although the ultimate synthetic targets have not yet been achieved, it is hoped that other manipulations of the inter- mediates glz, gig, éég’ or ééé may provide a route to difuran léfl and dithiophene lég. ACKNOWLEDGMENTS I wish to express my sincere appreciation.to Professor Harold Hart for his enthusiastic assistance, encourage- ment and guidance throughout the course of this study. Appreciation is extended to Michigan State University, National Science Foundation, and National Institutes of Health for financial support in the form of teaching and research assistantships. Many thanks go to my parents, my wife Shiow and my sisters for their support and constant encouragement during these years. ii Chapter TABLE OF CONTENTS LIST OF TABLES. LIST OF FIGURES PART I — APPLICATIONS OF BIS-ARYNE EQUIVALENTS TO ORGANIC SYNTHESIS. . . INTRODUCTION. RESULTS AND DISCUSSION. 1. The Preparation of Unsymmetric Bis- adduct. 2. The Introduction of Substituents other than Methyl Groups. . . . . 3. The Effect of the Size of Groups Attached to the Pyrrole Nitrogen Atom. . . . . A. Aromatization 5. The Applications of N-(Dimethylamino)— pyrrole to the Preparation of Polyacenes. 6. The Applications of N-(Dimethylamino)- octahydrocarbazole to Arene Synthesis 7. Attempts to Prepare Bis-adduct with a N-N Bridge. EXPERIMENTAL. I. General Procedure 2. Tetrabromo—p-xylene (gl). iii Page xi xii 1A 19 20 2A 27 35 A7 53 63 63 63 Chapter Page 3. Bis(N—methyl)-l,2,3,A,5,6,7,8,9,10- decamethy1—1,A,5,8-tetrahydroanthracen- l,A;5,8—bis-imine (2A). . . . 6A A. N- -Methyl- -bis(tetrahydrobenzo[l, 2 ,3 A])- 6, 7- dibromo— 5, 8— —dimethylnaphthalen— l, A— imine (37). . . . . . . . . 65 5. Bis(N—methyl)- bis(tetrahydrobenzo— [l 2 ,3 A])— 5, 6, 7, 8 ,9 10- -hexamethyl- ,A5 ,8- -tetrahydroanthracen- -l ,A;5,8- bis- imine (38). . . . . . . . . . . . . . . 66 6. 2 ,3 Dimethyl- l, A— —diphenyl- -l, A- butanedione (A3).. . . . . . . . . 67 '\J 7. 2,5-Diphenyl-l,3,A,-trimethylpyrrole (AZ). . . . . . . . . . . . . . . . . . . . 67 8. N- -Methyl- —l, 2 ,3,A, 5, 6 3,7 8— —octahydro- carbazole (A9). . . . . . 68 9. N- -Methyl- 2 3, A ,5- tetraphenylpyrrole one 25 8.2.7. “1:22: ,- VA L a; h ., Jh~ pvtgmr ‘1..v.u\ a 13 One possible reason for this unsatisfactory result may be the instability of the product toward acid. Although attempts were made to increase the yield by changing the reaction conditions, such as the amount of buffer reagents or solvent, and the reaction temperature, no improvement was observed. Therefore, it seemed worthwhile to search for a different bridge in the bis—adduct, one which could be eliminated without using acidic or basic reaction conditions. We have successfully used N-(dimethylamino)pyrroles as the diene component in bis-aryne equivalent cycloadditions, and found that the bridge can then be removed quantitatively in most of the cases by pyrolysis. It is the purpose of this part of the thesis to describe these studies and the applications of this approach to the preparation of highly substituted polyarenes. RESULTS AND DISCUSSION Many of the preceding methods for preparing highly methylated polyarenes begin by fusing a new benzene ring onto the starting hydrocarbon (as in Scheme 2).7’10 Ideally, if the starting hydrocarbon were to possess two sets of functionalities capable of being manipulated to re- act with two building synthons, the resulting bis-adduct 2% might then be converted to a polyarene such as 22 by con- ventional methods (Scheme 5) X \/ ' e’ Paco—woo l X z 2U ,0, >0©0+>©©0 -2_9 .32. Scheme 5 l& rfi, V. on «A B! 15 This synthetic plan was realized as follows. Tetra- bromo-p-xylene 22, prepared in 90% yield by the bromina- tion of p-xylene in carbon tetrachloride, reacted with two equivalents of n—butyllithium in the presence of a substi- tuted pyrrole at -78°C in THF to give a mixture of syn and anti bis-annelated adducts 2% in good yield. 2"» R=R =Me __R_19.R Scheme 2 This bis-annelation reaction via a bis-aryne equivalent could provide a successful strategy for the synthesis of highly methylated polyarenes if a subsequent bridge-remov- ing process for converting 22 to 29 were available. This is an attractive strategy for the following reasons (see Scheme Q). The synthesis is designed to proceed with bis— adduct formation and aromatization as two separate stages. In the formation of bis-adduct, the arrangement of the sub- stituents (R, R') is such that the steric interaction between them is minimized (Figure l). l6 Figure 1. Structures of the syn and anti bis-adduct. In other words, the crowdedness due to the substituents can be accommodated at this stage because of molecular distortion by the bridges. Furthermore, the reaction in- volved in forming the bis—adduct is highly exothermic, since it proceeds via a very reactive intermediate (i;§;, aryne). The peri interaction between R' and R is then introduced into the molecule when the bridges are removed. Neverthe— less, this unfavorable strain energy is more than compen- sated by the energy obtained from aromatization. Therefore, this synthetic sequence enables the easy intro- duction of the substituents via the bis-adduct, and provides a driving force for the conversion of the bis-adduct to an arene. One similar bis-annelation was previously reported by Wittig and his co-worker.lLl 2,6-Difluoro-3,S-dibromo-p— xylene ii, when reacted with magnesium in the presence of furan, gave a low yield (5%) of bis-adduct 3%: which was 17 ’of afi+¢ .91. —— We “2 comm-fit- 33 then reduced with palladium on charcoal and dehydrated to give 9,10-dimethylanthracene %% in unspecified yield (the overall yield must be less than 5%). In contrast to the unsatisfactory results of Wittig, we were able to optimize the reaction for bis-annelation and found that the follow- ing procedure provides a good yield of bis-adduct. A mixture of 10 mmole of bis-aryne equivalent and 20 mmole of pyrrole in 200 mL of anhydrous solvent is cooled to -78°C under argon atmosphere as 25 mmole of n—butyllithium (diluted 5-fold with hexane from the commercially available 2.“ M reagent) is added dropwise over a period of 2 hours. The mixture is allowed to warm slowly to room temperature and 18 is then quenched with methanol. The reaction was usually worked up by extraction of the adduct into methylene chloride and purification by chromatography and/or re- crystallization. O O The mechanism of this reaction, as demonstrated by Hart and co-workers,15 is a stepwise process involving the initial formationcfifmono-adduct §g via dibromodimethylbenZyne. Sub— sequent lithiation of the mono—adduct gives intermediate éé, which undergoes another Diels-Alder cycloaddition via an aryne, to give the bis-adduct @é. 1 ! a. B: l9 1. The Preparation of Unsymmetric Bis-adduct. As indicated at the beginning of this section, sym- metric bis-adducts were derived from the bis-aryne equiva- lent and 2 moles of pyrrole. This bis-annelation process can also be extended to the synthesis of a non-symmetric bis-adduct by taking advantage of the stepwise nature of the process. A representative example is described as follows. Tetrabromo-p-xylene was allowed to react with one equivalent I Br Br N + we n_-_B___>uli Itfifir 49 , _3.__7 \ “I n-Buli 6+ $0 39 Scheme 1 20 of n—butyllithium in the presence of N-methyloctahydrocar— bazole in toluene at -78°C. The mono-adduct %1 was isolated in 73% yield. It was subsequently reacted with another equivalent of n—butyllithium and pentamethylpyrrole to give the bis-adduct éé in 58% overall yield from gl. This example typifies the applicability of the process to the preparation of unsymmetric bis-adducts, and such adducts should even— tually lead to a non—symmetric polyarenes with substituents at the desired positions, as illustrated in Scheme 1. 2. The Introduction of Substituents other than Methyl Groups. In order to fully take advantage of the bis-aryne equivalent cycloaddition reaction, it was of interest to investigate the possibility of introducing different sub- stituents other than methyl groups. Therefore, a series of pyrroles QQ—QQ was prepared using either the Paal—Knorr 1 n 2 n1 n2 " MCNHZ / c u 9 N-‘Me 6 6 \ a / a 1 R R1 39 R =Me, R2=H QQ R1=Me, R2=H 3% R1=R2=Me 5Q R1=R2=Me 1 2 1 2 5g R =Ph, R =H 5g R =Ph, R =H mg R1=Ph, R2=Me ”Z R1=Ph, R2=Me "\J ’b 21 g NNHM 19 2 “2 92 / ‘" H 32 mama nccn R2 m 9 ""M" 9 cc 2 Manager R \ R1 1_ 2= 3% R —Me, R Ph Mg R 1R 2=(CH 2)“ pyrrole synthesis16 (azeotropic distillation of the cor- responding l,U-diketone QQ-Qé with methylamine), or the procedure developed by Posvic (QQ, Q2).l7 Tetraphenyl-N-methylpyrrole éQ was prepared in an overall yield of “0% by the dehydrocyclization of benzoin with zinc and ammonium acetate,18 followed by methylation of the pyrrole anion with dimethyl sulfate. Me OOH 1,, /® 1. N all _ II é— A 503 Hi When the bis—annelation reaction was carried out with these pyrroles, the results showed that the bulkiness of the substituents on the pyrrole did not change the yield I IA. ‘ ”V‘s. . i mun . . “Hf‘fi vv ;..;Y il _. V‘ 6 v.. r-. 'o-h ‘ . "1‘, J , J 22 of bis-adduct dramatically (see Table l). The results do suggest that the electronic factor may be much more important than the steric factor in these cycloadditions. For example, pyrroles Q1, 3% were used because they have the same substituents, ELgL, two methyl and two phenyl groups, but the substituents are arranged such that the pyrroles differ sterically at C-2 and C-5. Presumably, if the steric effect of the substituents at the point of bond formation (C-2, C-5) were predominant in cycloaddition, the yield of bis-adduct QQ should be ap- preciably less than that of bis-adduct éé due to the tremendous change in steric hindrance at C-2 and C-5. How- ever, the results show that the yields are about the same. On the other hand, in comparing Qé, fig and fig in which the substituents on 0-2 and C-5 of the pyrrole were the same, the cycloaddition was found to be unsuccessful if the methyl substituents on C-3 and C-4 were replaced by hydrogen atoms or phenyl groups. The failure of these cycloadditions might be ascribed to insufficient electron density in the diene system because of the inductive effect of the hydrogen and phenyl groups, which are electron-withdrawing relative to the methyl group. Unfortunately, the orientation of the phenyl groups in pyrroles gé, Q1 and fig is not known. Therefore this conclusion is based on the assumption that the phenyl groups are similarly oriented in all three pyr- roles. 23 Table l. The Cycloaddition of a Bis-aryne Equivalent with Substituted Pyrroles at -78°C.a I R' “' Br+ \N II-BIIlI 7 R @.@ R ' R R I . . N R R \ / 44 51 73% H I N M 45 24 79% I N \ / 59 52 76% I N Ph Ph l N PII \ 47 54 57% I Ph II Ph M 46 0% fl HI 'I‘ ' \ / so 0% Ph Ph aThe yields reported are the total isolated yield of syn and anti isomers. 'R‘; van- Q no, n-fi ...,. a- .‘ J 2A 3. The Effect of the Size of Group Attached to the Pyr- role Nitrogen Atom. In the previous section, attention was focussed on the effect of the C-substituents on the pyrrole. Another question which can be raised is whether the yield of bis— adduct would be effected if the pyrrole had an N-substituent other than a methyl group. It might be suspected that the more bulky the N—substituent on the pyrrole the more dif- ficult it would be to have effective orbital overlap for bis—adduct formation (the orbitals considered here are the active orbitals on the bis-aryne equivalent and those at the 2- and 5- position of the pyrrole). Therefore a study was planned to address this question. A series of 19’20 with different bulky groups on the N—atom pyrroles were prepared and reacted with a bis-aryne equivalent as described in the Table 2. The results indicate that the bulkiness of the sub- stituents on the N-atom of the pyrrole does not effect drastically the yield of bis—adduct. These observations are consistent with data obtained by Wolthuis and Boeer who found that when benzyne was reacted with N—substituted pyrroles iz-Qg, the yield of 1:1 adduct was not affected greatly by the bulkiness of the N-substituents. Apparently the formation of bis-adduct proceeds through a transition state according to the Alder rule, which 25 Table 2. The Cycloaddition of a Bis-aryne Equivalent with Different N—substituted Pyrroles at -78°c.a R I Z, Mir ——> @0® I :"I e 4. m mfiu I ioPro ‘ , W __._._ a lcuzph We. ._._ Ph I yr 4.. a aThe yields reported are the total isolated yield of syn and anti isomers. .1 (I) I y‘ 26 «I + \N/ —"-‘—> O(€I gs R=Bu 63% 51 R=CH2Ph 50% QQ R=Ph 51% 53 R=cyclo-C6Hll 59% requires maximum w to n orbital overlap between the diene and dienOphile. The overlap is achieved by approach of the benzyne and pyrrole moieties in two parallel planes rather than in a head to head fashion (as indicated by Figure 2). Figure 2. The transition state of cycloaddition with respect to basis set orbitals. 27 Consequently steric hindrance by the N—substituent is not important in the cycloaddition. The results of the foregoing studies can be summarized as follows: I) As described in Scheme 1’ a particular bis-adduct having substituents at desired positions usually can be synthesized in a stepwise process with different pre- designed pyrroles. II) The annelation of pyrroles to benzynes is pri- marily controlled by the inductive effects of the substitu- ents rather than by steric factors. The bis-adduct can usually be prepared from an electronically efficient pyrrole (2.2;: one which contains electron-donating sub- stituents) regardless of the size of the substituents. III) The applicability of pyrroles as dienes is also not limited by the bulkiness of the substituent on the N- atom. Having arrived at these conclusions concerning factors which affect the bis-annelation, we now turn to the question of converting the bis-adducts to aromatic compounds. A. Aromatization. A variety of dienes have been employed in benzene ring construction based on the cycloaddition-extrusion se- quence.22’23 Some extrusions occur with only a little 28 effort. For example, the cycloaddition of dimethyl acetylene— dicarboxylate to tetracyclone gave the adduct QQ which spon- taneously extruded carbon monoxide to give the tetraphenyl I II P R R - 60 ———> Ph Ph R R PI: R: 602MC 64 diester quantitatively.22 Similarly, benzene derivatives have been prepared from 2—pyrone and acetylenes.23 This is a typical Diels-Alder reaction which is followed by ' 3 (I) (I) 29 loss of carbon dioxide from the adduct éé to provide the aromatic product. But the presence of a highly nucleophilic reagent which is also a strong base, iLEL, n-butyllithium in our bis—adduct forming reaction does not permit the use of these carbonyl-containing dienes. Although the utilization of pyrroles for adduct forma- tion has been reported elsewhere, subsequent aromatization usually cannot be performed with ease, as exemplified by previous examples (this is especially true when trying to prepare highly strained arene58’13). . C] , 9 OO O BIT” / f‘ \' 22L 7‘ 01“; 30 The formation of a small amount of naphthalene as a by-product in the reaction of benzyne with N—methyl-pyrrole was first noted by Wittig and Behnisch.2u N-methyl-lO,ll- dihydro-l,2-benzocarbazole éz, the major product, presumably arose from the reaction of adduct éé with another mole of benzyne to form an intermediate zwitterion éz, which then underwent ring closure to Qg. The mechanism for obtaining the complementary product naphthalene was not explained. The l,u-dihydronaphthalene-l,u—imine-2,3-diesters 1Q and Z1, on heating with excess dimethyl acetylenedicar— boxylate, also lead to the deaminated naphthalene diester 7T) m=m=Me 72 71 m=Me th 73 Wfl©©+ 75 76 31 25’26 Analogously compound 1% Z2 and 1%, respectively. on heating at 180°C with acetylenic ester, gave some of the anthracene Zé. But in none of these cases was the deamination product the major product. Furthermore, Diels- Alder reaction of the expected arene with excess dimethyl acetylenediester leads to a secondary product, such as 1Q, adding another disadvantage to this process. A successful method for aromatizing 1,4-naphthaleneimines was reported by Gribble and his co-workers.12 The observa— tion was made that peroxide or peracid oxidation of the l,U- imine rapidly generated an N—oxide 11 and subsequently produced the aromatic hydrocarbon 1% with the elimination of nitrosoalkane. This method does afford a useful I F O m-CPBA -MeN0 f r synthetic approach to simple polyarenes. One competing reaction is oxidation of the carbon-carbon double bond. Also, the oxidation was carried out with a peracid which on reduction gives a carboxylic acid that may cause 32 isomerization of peri—substituted polyarenes. As a result of this anticipated disadvantage, we modified the reaction conditions for our systems by adding a weak base, such as Na2CO3 or NaHCO3, along with m-CPBA during the oxidation. It was intended that the acid generated during the oxidation would be neutralized by the pre-added bases and reduce the chance of isomerizing the product. The modification did give a respectable improvement in yield for the preparation of decamethylanthracene,8 but the yield was poor for dodecam- ethylnaphthacene,l3 and the reaction failed completely for the highly rigid anthracene derivative 12. @o(a "2:22 o O. Several other modifications were tried, such as changing the solvent, reaction time, or using pyridine as a base. All the results turned out to be essentially the same, so an alternative method was sought. Schultz27 reported very recently that the reaction of N—carbomethoxyaminopyrrole fig with excess dimethyl 33 acetylenedicarboxylate in refluxing toluene (U8 h) gave the tetralin 22% in 55% isolated yield. This simple method m ”MA” 6 0 II’: 002Me c 2n _8__q_ \" / DMAD R, . II’ ‘J!1_59:NH2 _§_g_ II‘: IIIIIIIe _8_lI1: NMez for benzene ring construction also worked for the reaction of pyrroles Ql - g; with DMAD on an NMR tube experimental scale. The thermal decomposition of adduct fig to arenes seemed attractive for our purposes because no other reagents are involved in the transformation. N-(dimethylamino)-2,3,u,5-tetramethylpyrrole §§ was easily obtained in 90% yield by azeotropic distillation from a mixture of 3,A-dimethyl-2,5-hexanedione and 1,1- dimethyl-hydrazine. The cycloaddition of pyrrole Q% was 3II first carried out at dry ice temperature in THF with dibromo- prehnitene 85, and gave an 82% yield of N-(dimethylamino)- l,2,3,4,5,6,7,8-octamethyl-l,U-dihydronaphthalen-l,U-imine 8Q. The mass spectrum of adduct 85 shows a very weak parent at m/g 298 but a base peak at m/e 2&0, which corresponds to the fragment without the nitrogen bridge. The 1H NMR spectrum of 8Q in CDC13 consisted of five sharp equal inten— sity singlets at 51.60, 1.93, 2.08, 2.21, 2.30. MezIIIIIIz \LI‘Mez r \ / ' Z. ‘1 -A—> 85 86 18 Upon pyrolysis at 200°C, adduct 8g was converted to octamethylnaphthalene 18 quantitatively in 30 min. At- tempts were made to investigate the volatile compounds from the pyrolysis, but the results from GC-Mass spectrometric analysis showed that the composition of the pyrolysate was not reproducible. The mixture was too complex to lead to 35 any conclusions. The pyrolysis may involve a nitrene ex— 27 Alternatively, trusion as suggested by Schultz and Shen. it might proceed via the elimination of nitrogen and ethane. The mechanism of the bridge removal is not clear at this point and requires further study. 5. The Applications of N-(Dimethylamino)pvrrole to the Preparation of Polyacenes. N-(Dimethylamino)-tetramethy1pyrrole 8% was used in a bis-cycloaddition with the bis—aryne equivalent derived from tetrabromo-p-xylene and n-butyllithium. A mixture of syn and anti bis-adducts 81 was obtained in 64% overall yield. The mixture could be resolved by column chroma- tography. However, no attempt was made to distinguish which isomer was which. In contrast to the smooth con- version of the naphthalene imine 8% to naphthalene 18, the pyrolysis of 81 was complex and the product composition depended on the pyrolysis time. When bis-adduct 81 was heated at 180°C for 30 min, decamethylanthracene IQ and its isomerization product 82 were isolated in “5% and 36% yield respectively. Prolonged pyrolysis only increased the amount of 82 at the expense of IQ. Compound 82 could be formed in several ways. Possibly the isomerization is induced by an active species such as a nitrene generated during pyrolysis. It might also be formed by a thermally allowed 1,5-H migration process on 36 the central ring of the anthracene, to relieve the peri interactions of the methyl groups at the 1,M,5,8,9, and 10 positions. However, the latter explanation was ruled out by the complete recovery of decamethylanthracene after heat- ing it under the same conditions used for the conversion of 3000+000 22 .. 2,2- :— eooc __A_ The double bridge removal occurs stepwise. Thus, by interrupting the pyrolysis at an earlier stage, some N- (dimethylamino)-l,2,3,U,5,6,7,8,9,lO-decamethyl-l,u-dihydro- anthracen—l,H-imine 88 was isolated, although most of the starting bis-adduct was recovered. However, stepwise bridge removal was not beneficial in increasing the yield of decamethylanthracene, because when the mono-imine 88 was 37 pyrolyzed at 200°C for 30 min, compound 88 and 8% were isolated in the same ratio as in the pyrolysis of bis— adduct 81. Surprisingly, when bis—adduct 81 was treated one equivalent of m-CPBA in the presence of Na2CO3 in reflux- ing CHC13, decamethylanthracene was isolated in 72% yield along with 12% of its tautomer 8%. This improvement over the result with 88 is reasonably explained because the reactiOn involves extrusion of nitrosodimethylamine, a thermodynamically more stable species than nitrosomethane. Oxidation of the bis-adduct with lead tetraacetate also gave 18 and 8%, but in trace amounts. Whether the reac- tion went through a nitrene, as in the conversion of 28 to 3%328 is not known. EtOfl 38 To avoid the isomerization which occurs when the substituents at C-9 and C-10 are methyl groups, another bis-aryne equivalent with two blocking groups at the para positions was used. Tetrabromo-p-dimethoxybenzene 2% was prepared in 75% yield by the bromination of p-dimethoxy- benzene. After carrying out the bis-annelation reaction, a mixture of syn and anti bis-adduct bis-N-(dimethylamino)- 9,10-dimethoxy-l,2,3,u,5,6,7,8-octamethylanthracen-1,A; 5,8-bis-imine 2% was obtained in 69% yield. Me Me r Br o 2.77 " o , one Br OMe r i We 93 M" 23" 95 39 Without resolution of the two isomers, the mixture was pyrolyzed at 150°C for 60 min, and gave 9,10—dimethoxy- 1,2,3,u,5,6,7,8-octamethylanthracene 28 quantitatively as shiny yellow crystals. The 1H NMR(CDC13) spectrum showed only three singlets at 52.38(12 H), 2.76(l2 H), and 3.33 (6 H). The ultraviolet spectrum of 28 is similar to that of decamethylanthracene (Figure 3) but with an appreciable log c L 280 lnm 380 400 Figure 3. Ultraviolet absorption spectra of $8 and 28. HO hypsochromic shift for all the principle absorption maxima. The effect of steric strain on the ultraviolet absorp- tion spectra of polycyclic aromatic molecules has been well documented. Data for polymethylated naphthalenes and 29,30 anthracenes were collected. In both series of com- pounds the band shifts could be correlated with the location Table 3. Absorption Spectral Data of Substituted Anthra- cenes.a Compound nm nm nm nm Anthracene (A) 253 339 356 375 2-methyl A 255 3ND 358 377 2,3-dimethyl A 257 391 358 378 2,3,6-trimethyl A A 259 3U2 360 379 2,3,5,6-tetramethy1 A 261 393 359 379 2,3,9-trimethyl A 261 350 368 388 2,3,9,lO-tetramethyl A 265 359 379 MOO 2,3,6,7,9,lO—hexamethyl A 296 360 380 uo2 9,10-dimethoxy—l,2,3,M,5,6,7,8- octamethyl A 280 378 399 U18 1,2,3,U,5,6,7,8,9,10- decamethyl A 288 --- “03 U27 aSolvent: cyclohexane. U1 of the methyl groups. A nearly 2 nm bathochromic shift per "beta" methyl group on the shortest wavelength intense maximum of anthracene (253 nm) was reported. However, methyl substitution at the "peri" positions causes a more pronounced shift of about A nm. (See Table 3.) Although this is an empirical conclusion, there is no doubt about a qualitative correspondence between the batho- chromic shifts due to methyl substitution and the change of molecular structure caused by the overcrowding. The data clearly imply that the more crowded the substitution on the anthracene the larger the bathochromic shift which will be found. No adequate theoretical explanation of the observed shift is available, but it is generally assumed that this phenomenon is associated with the alignment of the substituent with respect to certain axes of the molecule. All the principal absorption maxima of 9,10-dimethoxy- octamethylanthracene 28 have a bathochromic shift relative to 2,3,6,7,9,lO-hexamethylanthracene, but have a hypso- chromic shift relative to decamethylanthracene. Therefore, the UV data support the conclusion that 9,10—dimethoxy substitution experiences less of a peri effect than 9,10- dimethyl substitution on the anthracene skeleton. In other words, 9,l0-dimethoxy-octamethylanthracene might have a less distorted geometry than decamethylanthracene. The new aromatization technique has been combined with bis-annelation to improve the preparation of dodecamethyl- naphthacene 28. Tetrabromo-l,“,5,8-tetramethylnaphthalene H2 38 was—preparedl3 and reacted with N-(dimethylamino)-tetra- methylpyrrole 83 and n-butyllithium at -78°C, and gave a 62% total yield of two isomeric bis-adducts 28. The mix- ture of stereoisomeric bis-adduct 28 was pyrolyzed at 180°C, and the red crystalline dodecamethylnaphthacene 88 was. obtained quantitatively. Compared to the previously re- ported work (8% overall yield from é§),13 the utilization of N-(dimethylamino)-tetramethylpyrrole as a building block represents a dramatic success in this synthesis. *2 96 ' 28 In general, the reactivity of acenes (naphthalene, anthracene, naphthacene, pentacene, hexacene) increases with increasing molecular weight. It was reported that heptacene is so unstable that all attempts to prepare the compound in a pure state failed.31 A similar difficulty “3 was encountered in the preparation of tetradecamethylpent- acene 88% by our methodology. Two possible synthetic routes to pentacenes (102 or 108) ’b’b’b ’b’b are outlined in Schemes 8 and 9. The key feature for dif- ferentiating the better approach must reflect the nature of the peri-substituted arene. In Scheme 8, the difficulty lies in the preparation of 2,3,6,7-tetrabromo-1,U,5,8,9,10- hexamethylanthracene 888. It was hoped that 888 could be Br n \§Li27/ .92. l .22 came —— :2 3: 00000 ' Scheme 8 ’b MU prepared by the bromination of 1,4,5,8,9,10—hexamethyl- anthracene 22. However, the preparation of 22 according to current method was not satiSfactory. Treatment of tetrabromo-p-xylene 3% with n-butyllithium and N-(dimethyl- amino)-2,5-dimethylpyrrole 21 only provides 7% of the bis- adduct 28. Also anthracene 3%, with quadruple peri-inter- actions, is predicted to be unstable under the acidic condition which prevail during bromination. Therefore, neither the pyrolysis of bis—adduct 28 to 82 nor the bromination of anthracene 22 was tried. Alternatively, the synthesis shown in Scheme 2 was pursued. The required N-(dimethylamino)-l,3,4,7-tetra- methylisoindole 88% was prepared using the procedure _developed by Bonnett.32 Reaction of 2,5—hexanedione and pyrrole 31 in refluxing benzene gave a U5% yield of iso- indole 88% as an extremely unstable air-sensitive compound. Therefore, instead of isolating 88% pure, crude material was used directly in the bis-aryne equivalent cycloaddition. Bis-adduct 888 was obtained in 56% yield, and it gave a satisfactory elementary analysis. However, the 1H NMR IIMno4 db MezNNIIZ AGE NOAc .193. M on heating to 1,2,3,A,5,6,7,8-octahydro-9,10,ll,l2-tetra- methyltriphenylene 888. Dehydrogenation of 888 with di- chloro-dicyanoquinone gave 1,2,3,U-tetramethy1triphenylene 888. Compound 88% could also be obtained by the meCPBA #8 . l ox1dation Of %%Z. The H NMR(CDC13) spectrum of klz shows peaks at 52.03(S, 6 H), 2.78(S, 6 H), 7.30(m, U H), 8.03 (m, 2 H), 8.30(m, 2 H). “9 Conventional methods for preparing triphenylenes usually give symmetric derivatives, either by trimerization of a 33 or by building up three new benzene rings 3“ benzene moiety from a trifunctionalized benzene. Therefore, the method described here is particularly useful, since it constitutes a new synthesis for unsymmetric triphenylenes. 2,3—Difunctionalized triphenylenes can also be prepared by the same sequence. Thus llg was reacted with the benzyne derived from gl and one equivalent of n-butyllithium in toluene. The adduct llg was obtained in 88% yield. 0 0 00:00 o ,, .112 n-Buli 00 "9 Similarly, l,U—dimethyl-2,3-dibromo-5,6,7,8,9,lO,ll,l2- octahydrotriphenylene llg and l,u—dimethyl—Z,3—dibromotri- phenylene llé can be subsequently obtained by the same 50 pyrolysis and dehydrogenation process. Attempts were made to couple two triphenylenes by reacting %%é with n-butyl- lithium. However, only a trace of the expected bis-tri- phenylene llg was detected by GC-Mass spectral analysis, along with the butylated triphenylene llg. Attempts to isolate llg or llg were not successful. As suggested previously (Scheme Z, page 19), unsym- metric anthracenes can be prepared by using bis—aryne equivalents in a stepwise process. The following example explores this possibility. Treatment of %l with one equiva- lent of n—BuLi in the presence of one equivalent of N-(di- methylamino)-tetramethylpyrrole Qé, gave N-(dimethylamino)- 6,7-dibromo-l,2,3,U,5,8-hexamethyl-l,u-dihydronaphthalen— l,u-imine lgg in 62% yield. This mono-adduct was then converted to 2,3-dibromohexamethylnaphthalene g9 quantita- tively by pyrolysis at 200°C. Compound fig has been pre- pared before from dimethylisatin in six steps with an overall yield of 39%.8 The current synthesis, which in- volves only two steps with an overall yield of 62%, repre- sents a significant improvement. Reaction of £9 with n- butyllithium and pyrrole llg gave N-(dimethylamino)-bis- (tetrahydrobenzo[1,2;3,A])-5,6,7,8,9,lO—hexamethylanthracen- l,U—imine lgg, which was aromatized by heating at 150°C. This gave anthracene lg; quantitatively, as yellow crys- tals. Anthracene lgé was not stable in solution, and isomerized to %3% rapidly at room temperature. This re- arrangement can be catalyzed by a trace of CF3COOH, with 51 H 124 which the transformation of lgé to lgg was complete within 5 min. Presumably the facile isomerization of lgé is caused by the rigidity of the anthracene skeleton as a consequence of the fused ring substitution. Distortion of the molecular geometry required for accommodating the peri-interaction, is no longer as easily achieved as in decamethylanthracene. Using this argument, an anthracene derivative such as lgé should be even more rigid and thus more labile toward 52 isomerization. Indeed, pyrolysis of bis-adduct leg, which was obtained from the reaction of bis-aryne equivalent gl with llQ, gave what is presumably zg as a yellow powder. Purification of this powder could only be accomplished by 'rapid recrystallization from CHCl3/MeOH solution, and al- ways gave a substantial loss of 12 by decomposition. The loss of 12 is not a clean isomerization reaction as in the case of lgg. Several intractable products were found in the mixture, and they were not separable. An even more difficult situation was encountered in the preparation of naphthacene derivative lgg, obtained as a red powder from the pyrolysis of lgz. A correct molecular 53 ion peak was found at m/g 500 in the mass spectrum of this 1H NMR spectrum could be powder. However, no satisfactory obtained, probably for two reasons, the insolubility of the red powder in organic solvents and the rapid decolorization of the red solution during the period of running the spec— trum. It is believed that naphthacene derivative %%§ was generated by this reaction, but conclusive evidence could not be obtained. "0 EHM> 00 0000 7. Attempts to Prepare Bis-adduct with a N-N Bridge. Compounds with distorted arene rings, such as the cyclo- phanes, in which more than two atoms of an arene ring are 5H incorporated into a large ring system, have been of in- terest for the study of molecular strain.35 Most investi- gations have dealt mainly with the geometry of the molecule itself. For instance, how much distortion from planarity can an arene ring accommodate? How will an arene ring . behave when the ring and bridge are compressed face to face? The questions concern how the physical properties and reactivities will deviate from those of the normal molecule. It seemed of interest to prepare the bridged compounds lag - l§g to study the possibility that the planarity of the benzene or naphthalene rings could be altered by chang- ing the number of methylene groups in the bridge. It seemed feasible to prepare these compounds either by the reaction on a bis-aryne equivalent with a bridged bis- pyrrole such as léé — lgg in a one to one fashion, or by APE “=6 Aéé n=3 we n=7 m =u tit “=8 l§g n=5 55 the reaction of a bis-aryne equivalent with functionalized pyrroles $£§ (Scheme lg, page 57) which can subsequently undergo intramolecular cyclization to form a ring. The bridged bis-pyrroles liQ - lag were prepared in excellent yield from the corresponding diamines and 2,5- hexanedione, using the same method used for the preparation of pyrrole Q§. + H,N-(CH2)n-NH2 fie : —(cu2),.—<: 6 fli3y4fil6,7Jl lag—m The reaction of bis-pyrrole lég - lag with one equivalent of bis-aryne equivalent éé was tried in the following ways: high dilution technique, simultaneous addition, inverse 56 addition. However, no adduct of type iii could be obtained; only polymeric residues were formed. It was found by Hart and co-workers11 that a mono- adduct of the type lg§ can be obtained in fair yield by reacting one equivalent of n—butyllithium and tetrabromo- p-xylene with one equivalent of a bis-pyrrole in toluene at -78°C. Nevertheless the subsequent intramolecular benzyne addition did not lead to the expected bridge com- pound despite all efforts. Presumably, the geometry of the mono-adduct %Q% is such that it can only undergo an inter— molecular Diels-Alder reaction rather than the intra- molecular process. Consequently, attempts to prepare and cyclize these mono-adducts were abandoned. / u-—-(c"2)n fiz’mflz)" 1. Br 3, w &' Br 11?— La Instead, the second approach (Scheme lg) was tried. Pyrrole lgg was prepared by the reaction of diketone Ql with 2-aminoethanol. The O—anion of lfifi was then methylated 57 \/ucu oum‘La/ucco / I \ 20H2 zMezso4 \ “2 "2 MC 144- 145 _4L \ / > \ NCHZCONMez with dimethyl sulfate to give N-(2-methoxyethyl)-pyrrole lUQ. One bis-adduct isomer predominated (lulleB = 1:10) ’b’b ’b’b ’b’b’b g“, i f6 °"’tl=uz N N 145 EBOMe $Mww10 58 from the reaction of 2% with %N2 was isolated in 77% yield. Its 1H NMR(CDC13) spectrum was consistent with the struc— ture, having peaks at 61.58(s, 12 H), l.65(s, 12 H), 2.20 (s, 6 H), 2.28(t, u H, 1:8 Hz), 3.23(s, 6 H) and 3.33(t, A H, i=8 Hz). An E-ray crystallographic analysis showed that the bis-adduct had the anti configuration (Figure A). Therefore subsequent ring cyclization to the bridged com- pound l%% was clearly impossible. However, it might be pos- sible to prepare a doubly bridged compound as indicated on structure 1&8. ’b’b’b Figure 4. Stereo drawing of bis—adduct lag (Z = OCH3). Attention was then focussed on manipulation of the methoxy group to a useful function for later conversions. Acidic deprotection of the methyl ether with concentrated hydrogen iodide was not plausible because it is well known that exposure of a naphthalen-l,A-imine such as %32 to acid will lead to an aromatized d-naphthylammonium salt l§9.36 Therefore, this method was not pursued. Accord- ingly, we applied the procedure developed by Jung.37 59 H + H / “\+/ N: “ -f / —-> Q ——> C 149 150 However, difficulties in removing the methyl group were encountered with this method also. More than 95% of mono- adduct lgl or dimethoxy bis-adduct IQQ was recovered when the adducts were refluxed with trimethylsilyl iodide, even after 96 hours at reflux. W W38" 0 . o 15] 153 Since pyrrole lag, with the unprotected hydroxy group, did not provide any bis-adduct lég in a bis—aryne 6O cycloaddition, various protecting procedures for the OH group were tried. These are listed in Scheme ll. Un- fortunately none of them were successful due to the in- stability of pyrrole kNN in acid. / z \ Se :9 3‘ 52 w / {:2 \ 144, C-c-CI SOCI 9,29 \/ Scheme ll The unsatisfactory attempts at protecting the hydroxyl group or deprotecting the methoxyl group prompted us to search for another functionality. The use of N-butyllithium for generating the bis-aryne equivalent does not leave many 38 on the lithiation of function~ choices. A recent study alized benzenes implies the compatibility of the N,N- dialkylamide group with n-butyllithium at low temperature. This suggested that an amide group ontflmapyrrole might solve this problem. The desired pyrrole could be derived from 61 the reaction of a diketone with the dimethylamide of glycine. Glycine N,N-dimethylamide l2; was prepared from the reaction of glycine anhydride téé with dimethylamine,39 and provide pyrrole lfié (page 57) quantitatively by the previously mentioned procedure. 9 9 soc: 9 ’9 Ph-c-o-c-N-c-c-on ——2—> Ph-c-o-c-u-c-cm 155 l (5 9 N MezN-c-c-NNZ V" MEZNN 04;):0 157 _156 0 0 1 J “2N“ [tau-é-n-c-E-ou \ 159 9. \ ° EtZN-O-C-N / f EtZN-E-c-Nllz 16‘ £9. However, treatment of 1&6 with bis-aryne equivalent 21 ’b’b’b ’Vh only gave a 6% yield of bis-adduct (stereochemistry not known). Possibly the reaction was complicated by competi— tion between the lithium-halogen exchange of 2% by n— butyllithium and nucleophilic addition to the amide group '62 14_6 X 0 o 21 n-Buli I MEZNG $.w CNMEZ 158 by n-butyllithium. It was thought that the N,N-diethyl- amide pyrrole lél would have a better chance for cyclo- addition, due to a decreased tendency for l,2—addition of n-butyllithium to the carbonyl group. Unfortunately, the reaction of diethylamine with lQé did not give the ex- pected glycine N,N-diethylamide lég but the acid derivative $52- All of these difficulties discouraged us from further investigating this aspect of the project. Two possibilities for further efforts might be to search for a proper func- tional group which is compatible with n-butyllithium, or to adopt methods that do not require n-butyllithium to generate the bis-benzyne equivalents. EXPERIMENTAL 1. General Procedures 1 H NMR spectra were measured in CDCl or CClu solution 3 on a Varian T-6O or on a Bruker WM-250 spectrometer with chemical shifts reported in 6-units from tetramethylsilane as the internal standard. 13C NMR spectra were determined on a Varian OFT-2O spectrometer. UV spectra were determined on a Unicam SP-BOO or Cary-1756 spectrometer. Mass spectra were obtained with Hitachi Perkin—Elmer EMU-6 and a Pinni- gan H000 spectrometer. High resolution mass spectra were obtained with a Varian CH5 spectrometer. Elemental analyses were performed by Spang Microanalytical Laboratories, Eagle Harbor, Michigan. The melting points were determined on a Thomas Hoover Unimelt apparatus and are uncorrected. 2. Tetrabromo-p—xylene(2l) To a solution of p-xylene (2% g, 0.23 mol) in 100 mL Of CClu were added dropwise six equivalents of Br2 at room temperature with stirring. The mixture was refluxed over- night and cooled. A saturated aqueous NaHSOu solution (about 700 mL) was then introduced until the bromine color disappeared. The precipitate was collected by filtration, 63 6A and recrystallized from chloroform/methanol to give needle- like crystals of 21 (79.5 g, 9U%); mp 2A9-251°C (lit.uO 251-252°c); 1 H NMR(CDC13) 62.78(s, 6 H); mass spectrum, m/g (relative intensity) U22(100), 3Ul(5u), 262(25), 182 (21), 102(67). 3. Bis(N-methyl)—l,2,3,U,5,6,7,8,9,lO-decamethyl-l,u,5,8- tetrahydroanthracen-1,U;5,8-bis-imine_(gg) A solution of tetrabromo-p-xylene (21, “.22 g, 10 mmol) and pentamethylpyrrole (2.79 g, 20 mmol) in anhydrous tol- uene (200 mL) was cooled to -78°C under argon (most of the g1 precipitates out). To this suspension n-butyllithium (22 mmol in 30 mL of hexane) was added dropwise with constant magnetic stirring. After addition (2 h), the mixture was kept at -78°C for three hours, then warmed slowly to room temperature and left for one more hour. Water (20 mL) was added, the layers were separated, and the aqueous layer was extracted with methylene chloride. The combined organic layers were dried over magnesium sul— fate and evaporated under vacuum, and the residue was re- crystallized from CHCl /hexane to give 2.97 g (79%) of a 3 mixture of the syn and anti isomers of 2%. One isomer was separated from the mixture by washing with ether (1.72 g, u5%); mp 256-258°c; lH NMR(CDC13) 61.63(s, 12 H), 1.66 (s, 12 H), l.93(s, 6 H), 2.23(s, 6 H); mass spectrum, m/g (relative intensity) 376(8), 361(1), 322(6), 266(27), 65 l3“(7); high resolution mass spectrum, calculated for C26H36N2, 376.28786; found, 376.29012. Attempts to purify the minor isomer were not successful; it remained a mixture of the two isomers. “. N-Methyl-bis-(tetrahydrobenzofl,2;3,“j)-6,7-dibromo-5,8- dimethylnaphthalen-l,“-imine (31) A suspension of tetrabromide 21 (“.2 g, 10 mmol) and octahydrocarbazolel7 £2 (1.89 g, 10 mmol) in 200 mL of anhydrous toluene was cooled to -78°C, and neBuLi (11 mmol, diluted S-fold with hexane from the commercially avail- able 2.“ M reagent) was added using a constant rate addi— tion funnel over 2 h, under argon atmosphere. The mixture was stirred for 3 h, warmed to room temperature for l h, water (20 mL) was added, the layers were separated, and the aqueous layer was extracted with methylene chloride. The combined organic layers were dried (MgSOu) and con- centrated under reduced pressure to give crude product, which was recrystallized from methanol/chloroform to provide 3.2“ g (72%) of pure adduct 37; mp l68-169°C; lH NMR(CDC13) 51.52 (m, 12 H), l.85(s, 3 H), 2.20-2.65 (m, “ H), 2.50(s, 6 H); mass spectrum, m/g_(relative intensity) “53(u5), “51(76), uu9(u0), A36(13), A23(25), u08(32), 39“(13), 370(5), 290(5), 188(100). 66 5. Bis(N—methyl)-bis(tetrahydrobenzo[1,2;3,“J)-5,6,7,8,- 9,10-hexamethyl-l,“,5,8-tetrahydroanthracen-l,“;5,8- bis-imine (3%) A solution of mono-adduct 37 (2.25 g, 5 mmol) and N- methyl-tetramethylpyrrole “5 (1.5 g, 11 mmol) in anhydrous THF (200 mL) was cooled to -78°C (most of the 37 precipitated out). To this suspension kept under argon n—BuLi (8 mmol in 30 mL of hexane) was added dropwise with constant mag- netic stirring. After addition (2 h), the mixture was kept at —78°C for three additional hours, then warmed slowly to room temperature and left for one hour. Water (20 mL) and methylene chloride (50 mL) were added, the layers were separated, and the aqueous layer was extracted with CH2C12. The combined organic layers were dried with magnesium sulfate and concentrated under vacuum. The resi— due was triturated with acetone (20 mL) and filtered to give 1.71 g (80%) of pure bis-adduct 38 (a mixture of two stereo- isomers). The mixture was chromatographed on alumina with hexane/ether as eluent. The first fraction (0.12 g) was a mixture of two isomers, the second fraction gave a pure isomer (1.59 g, 75%); for the major isomer; mp 265-266°C; lH NMR(CDCl3) 61.60—2.0l(m, 12 H), l.63(s, 6 H), l.66(s, 6 H), 1.86(s, 3 H), l.96(s, 3 H), 2.20-2.56(m, “ H), 2.30 (s, 6 H); mass spectrum, m/g (relative intensity) “28(29), “13(3), “00(2), 37“(27), 359(6), 188(8), 56(100); high resolution mass spectrum calculated for C30HUON2: “28.31916; 67 found, “28.322“26. 6. 2,3-Dimethyl-l,“-diphenyl—l,“:butanedione (“3) To a solution of propiophenone (67 g, 0.5 mol) in “0 mL of glacial acetic acid, was added potassium permangan- ate (30 g 0.2 mol) in small portions with stirring over 1 h (exothermic reaction), while the temperature of the mixture was kept at 90-95°C by a cooling bath. After the solution became brown (about 2 h), the mixture was refluxed for another 3 h, then cooled to room temperature, treated with water and extracted with ether. The extract was washed with 10% aqueous Na2C03 solution, then water and dried (MgSOu). After removal of the ether, the resulting oil was purified by chromatography on alumina with hexane as eluent, to give 13.“ g of the white l,“—diketone “3; mp “1 no mp reported); 1 202°C (lit. H NMR(CDC13) 51.23(d, 6 H, i=8 Hz), 3.83(m, 2 H), 7.23(m, 6 H), 7.80(m, “ H); mass spectrum, m/e (relative intensity) 266(2“), 161(17), l“7(8), 105(100), 77(20). 7. 2,5-Diphenyl-1,3,“-trimethy1pyrrole (“7) A mixture of l,“-diketone “3 (“.8 g, 18 mmol), methyl- amine (“0 mL, “0% aqueous solution) and benzene (200 mL) was heated under reflux for 10 h, the water being removed continuously with a trap. The solution was dried with MgSOu and evaporated under reduced pressure. The residue 68 was washed with 95% EtOH, and gave essentially pure pyr- role “7 (“.35 g, 88%) after filtration; mp 107-109°C (lit“2 1 101-102°C); H NMR(CDC13) 6 2.08(s, 6 H), 3.3l(s, 3 H), 7.26(S, 10 H). 8. N—Methy;71,2,3,“,5,6,7,8-octahydrocarbazole (“9) A mixture of l,“-diketone 109u3 (13.5 g, 0.07 mol) and “0% aqueous methylamine solution (300 mL) in 200 mL of benzene was refluxed for 8 h. Additional (200 mL) “0% methylamine solution was added and the resulting mixture was then refluxed overnight. The aqueous layer was separated and extracted with 200 mL of diethyl ether. The combined organic layers were washed with water several times, dried over anhydrous MgSOu and concentrated to give l“.8 g of crude white crystals which were recrystallized from petrol— eum ether to give 13 g (100%) of pyrrole “9; mp 9“-96°C (litl7 9u00); lH NMR(CClu) 61.5-1.8(m, 8 H), 2.5-2.7(m, 8 H), 3.2'5(s, 3 H). 9. N—Methyl-2,3,“,5-tetraphenylpyrrole (58) To a solution of tetraphenylpyrrole 50a18 (3 g, 8 mmol) in 150 mL of anhydrous THF was added NaH (0.29 g, 12 mmol). The resulting suspension was refluxed for 2 h. Dimethyl sulfate (10 mL) was introduced dropwise with cooling. The solution was stirred for another 10 h, then 69 quenched first with MeOH and then with water (5 mL each), and extracted with CH2C12. The organic layer was dried OVGP MgSOu and concentrated under vacuum. The residue was recrystallized from hexane to give 2.07 g (67%) of 50 ““ as white crystals; mp 209-21000 (lit 209-211°C); 1H NMR(CClu) 53.36(s, 3 H), 6.76, 7.1(two broad peaks, 20 H); mass spectrum, m/e (relative intensity) 385(7“), 77(100). 10. Bis(N—methy1)-l,“,5,8,9,10—hexamethyl-l,“,5,8-tetra- hydroanthracen-l,“;5,8-bis-imine (51) Using the same procedure as for the preparation of bis- adduct 2“, bis-adduct 51 was prepared by the reaction of tetrabromide 21 (“.2 g, 10 mmol) and N—methyl-2,5-dimethy1- pyrrole (2.8 g, 20 mmol) with n—BuLi (22 mmol). The reac- tion gave 2.39 g of a white powder which was recrystallized from chloroform/hexane to provide 2.33 g (73%) of 51 as a mixture of the syn and anti isomers, ratio 27/73 (determined by integrating the peaks at 51.71 and 1.98). The mixture was chromatographed on alumina with hexane/ether as the eluent. The first fraction was a mixture of two isomers; the second fraction gave a pure isomer (1.58 g). For the major isomer; mp 2“2-2““°C; lH NMR(CDCl3) 61.71(s, 12 H), l.98(s, 6 H), 2.25(s, 6 H), 6.53(bs, “ H); mass spectrum, m/g (relative intensity) 32o 20(1), 2793 observed data. The data were reduced [Wei K.-T. and Ward, D. L. (1976). Acta Crystallographica, B32, 2768-2773] and the structures were solved by direct methods [Main, P. (1978). "MULTAN78. A System of Computer Programs for the Automatic Solution of Crystal Structures from X—ray Diffraction Data." Univ. York, England.]; and the refinement was by full-matrix least squares techniques; Zalkin A. (197“), private communication.]. The final R value was 0.055. The final difference Fourier map showed densities ranging from +.38 to -.31 with no indication of missing or incorrectly placed atoms. 60. Reaction of Trimethylsilyl Iodide with Bis-imine 152 (or Mono-imine 151) A mixture of hexamethyldisilane (0.6 g, “ mmol) and iodine (l g, “ mmol) was heated at 65°C in a 20 mL flask 103 fitted with a reflux condenser. An exothermic reaction took place and a homogeneous solution was formed. The solution was refluxed for 2 h. Bis-imine 15% (0.“7 g, 1 mmol) in CClu was then added slowly at room temperature. The solution was refluxed for 72 h. Water (20 mL) and chloroform (50 mL) were added, the aqueous layer was separated and extracted with chloroform (30 mL). The com- bined organic solutions were dried (NaZSOM) and concentrated. Bis—imine 151 (“27 mg, 91%) was recovered. Starting with mono-imine 151 and using the same pro- cedure described above, no demethylation could be observed. Only the starting imine could be isolated from the re- action mixture. 61. Reaction of Tetrabromo-p-xyleneg1 with n—BuLi in the Presence of Pyrrole 111 Tetrabromo—p-xylene 11 (“.2 g, 10 mmol) was reacted with neBuLi (25 mmol) in the presence of hydroxypyrrole 111 (“17 mg, 25 mmol) using the same procedure as for the preparation of bis-adduct 11. After worked up, only a poly- meric powder was obtained and most of the starting pyrrole (“03 mg) was recovered. lOU 62. Reaction of Pyrroleglgg with Methanesulfonvl Chloride To a mixture of MsCl and pyridine (10 mL each) at 0°C, was added slowly a solution of pyrrole $53 (5 g, 30 mmol) in 50 mL of pyridine. The solution turned purple im- mediately with precipitation. The suspension was stirred for 3 h. Ether (100 mL) was added. The ether solution was then washed with water three times and dried (Nazsou), The solvent was removed under vacuum to give a gummy black residue. No pure compounds could be isolated from this black tar. The same result was found when pyrrole lQQ was treated with thionyl chloride using the procedure described above. 63. Reaction of Dimethylamine (or Diethylaminelgwith N-Carboxyglycine Anhydride léé A mixture of carbobenzoxyglycine léé (35 g, 0.16 mol, Aldrich Chemical Co.) and 100 mL of freshly distilled thionyl chloride was warmed on the water bath at UO°C for 50 min. The precipitate which formed was collected by filtration and air dried to give N-carboxyglycine anhydride gag (15.u g, 86%); mp l60-l62°C (lit.39 100°C). The crude léé was then stirred overnight with excess anhydrous di- methylamine (MO mL) at 0°C. After evaporation of the ex— cess amine, the residual oil was distilled under reduced pressure to give glycine N,N-dimethylamide 157 (10.1 g, 62%); ’L’L’L lO5 bp 110-11600/5 torr (iit.39 60°C/O.8 torr). l H NMR(CClu) 51.63(s, 2 H), 2.83(s, 3 H), 2.86(s, 3 H), 3.23(s, 2 H). Starting with diethylamine and using the same method as described above, acid derivative 15% was obtained (16.5 g, 63%); rather than the expected glycine N,N-diethyl- amide iég. For $33: mp 95-9700; 1 3 H, i=7 Hz), l.28(t, 3 H, i=7 Hz), 2.90(q, 2 H, i=7 Hz), H NMR(CDC13) 61.11(t, 3.20(q, 2 H, £fi7 Hz), 3.66(d, 2 H, i?“ Hz), 5.26(bs, l H, N-H), 9.63(bs, l H, CO2H); mass spectrum, m/g (relative intensity) 17U(10), 159(5), 130(18), 115(7), 100(17), 72(28), 58(100); high resolution mass spectrum calculated for C7H1AN2O3’ 174.100AA, found, 17u.101uo. 6U. N-(2-[N,N—dimethylacetamido])—2,3,U,5—tetramethyl- pyrrole (136) A mixture of 3,A—dimethyl-2,5—hexanedione 5% (6.9 g, U9 mmol) and amide $51 (5 g, U9 mmol) in 100 mL of benzene was refluxed for 6 h. Removal of the benzene under reduced pressure gave yellowish crystals which were washed with hexane to give amide pyrrole 1&6 (8.5 g, 8A%). Amide pyr— role lfié oxidizes rapidly and turns brown when exposed to air; mp iuu—1u60c; 1H NMR(CDC13) 61.86(s, 6 H), 2.03(s, 6 H), 2.90(s, 3 H), 2.96(s, 3 H), u.38(s, 2 H); IR(CClu) 1675(s), 1660(8), lUU5(w), lUOO(s), 1370(m) cm-l; mass spectrum, m/§_(relative intensity) 208(91), 193(6), 136(lOO), l22(27); high resolution mass spectrum calculated for C12H2ON20, 208.15757, found, 208.15631. 106 65. Bis(N—[2-(N,N-dimethylacetamido)l)-l,2,3,“,5,6,7,— 8,2,10-decamethyl—l,“,5,8-tetrahydroanthracen-l,“;- 5,8—bis-imine(15§l Bis-imine 155 was prepared by the reaction of tetra- bromide 2% (“.2 g, 10 mmol) and amide—pyrrole 135 (“.2 g, 20 mmol) with neBuLi (25 mmol) in anhydrous THF following the preparation of bis-adduct 2“. The residue was chroma- tographed on alumina with ethyl acetate as eluent to give only one isomeric bis-adduct 155 (62“ mg, 12%); mp 257— 259°C; 1 H NMR(CDC13) 61.63(bs, 2“ H), 2.23(s, 6 H), 2.86 (s, u H), 3.06(m, 12 H); 13c NMR(CDC13) 6172.19(CO), 1“7.68, 1“5.37, 126.18, 77.57, 49-37, 37.72, 35.99, 16.5“, l“.70: 11.“2; IR(KBr) l6“5(s), l“70(m), l“OO(m), 1275(w) cm‘l; UV(MeOH) Ama 260 nm (loge 3 52), 253(3.53), 2“7(3.68), x 23“(“.“2); mass spectrum, m/e_(relative intensity) 518(13), 503(2), 464(8), ““6(22), 392(1“), 337(1“), 127(100), 86(35); high resolution mass spectrum calculated for C32HM6N“O2’ 518.36209; found, 518.36100. PART II ATTEMPTED SYNTHESIS OF THIOPHENE OR FURAN FUSED RADIALENE ANALOGUES 107 INTRODUCTION Radialenes, a class of "exocyclic" polyenes in which the number of exocyclic double bonds is the same as the number of ring carbons, have attracted considerable in- terest both with regard to synthesis“6 and theory.£17 Aside from (5)-radialene 112, each of the lower members of this series 11g - 11% has been prepared either as a very I73 I74 “8 reactive hydrocarbon or as a reactive intermediate.)49 Consequently, studies have been limited to spectral properties at low temperature and to comparatively few 108 109 chemical reactions because of their high tendency to polymerize. fl 6 .6 u l \xlllllz’ /’ A ”it” I \ 2! .0' I75 I76 177 In contrast to the instability of the parent radialenes 3 derivatives having substituents on the terminal carbons, such as hexamethyltrimethylenecyclopropane 175,50a hepta- phenyltetramethylenecyclobutane 115,5Ob 50c and hexaethyl- idenecyclohexane 177, are reported to be very stable hydrocarbons. Stabilization of radialenes can also be achieved by incorporating a hetero-atom into the cross-conjugated systems. Thus, in 1972 naphthotetrathiophene 178, a naphthoradialene, was synthesized by Wynberg and Heers.51 (As shown on page 110). Naphthotetrathiophene 178 was reported to greatly resembly dibenzo(g,p)chrysene 179 in physical properties. Therefore, it is reasonable to view a radialene such as 178 as a naphthalene which is converted in a formal sense from having ten overlapping 110 p orbitals associated with endocyclic "double bonds" to having those orbitals associated with exocyclic "double bonds". In other words, we might accept the concept of a 111 O 179 17.93. 52 nonclassical condensed thiophene by using the sulfur d orbitals as in structure 178%, thus restoring the endo- cyclic aromatic naphthalene moiety. Such a resonance hybrid may contribute to the overall stability of struc- ture 178. Using the same argument, it is not surprising that the fused hexaradialenes should be stable compounds be- cause of their perfect aromatic sextet arrangement. Thus trioxo-hexaradialene 188,53 trithia-hexaradialene 1875“ and triaza-hexaradialene 78%5u were prepared recently and described as having very similar properties to those of triphenylene 183. 112 18 In light of the above known examples, fused tetra- radialenes such as 888, 785 and 188 became very attractive targets. The main interest in synthesizing compounds $88 - 188 arose from two sources. The first is the concept of anti-aromaticity. In contrast to 778 - 183 which have “n+2w electrons, compounds 78% - 788 have a total of “nn electrons. Such systems are not only non-stabilized,56 but are actually destabilized by the conjugation. The second reason for interest in those compounds is that they could provide model compounds for the study of induced 113 0 0 S S 0 S VFW ’ W / \ 184 185 El; paramagnetic ring currents in a paratropic system. Each of them potentially contains a cyclobutadiene ring in the central part of the molecule. U N H U—w H \ K! 18 <—> /:. \lt—\|<_>\‘L—l/ 187 Biphenylene £87 was the first stable compound of this type to be prepared, and it indeed shows the predicted rather small diamagnetic current in the six-membered ring 57 and large paramagnetic current in the four-membered ring. Its preparation was easily accomplished by two classical . 58 . methods as indicated in Scheme 8%. However, this 11“ \r—W/ ”an/im— X X Scheme %% approach cannot be applied to the preparation of a com- pound such as 888. A bis-Wittig reaction was therefore selected as the key step in the synthesis, the “n+2n ring being constructed in the final step. Thus, 2-thianorbi- phenylene 882 was prepared from benzocyclobutadienequinone O /' “guns, @F—(ouoz man /o \I —kouo;——) \o ' 1.3!; EaPa s 33H 7 :57\ @153 <5’\ ¢5’\ O @"zé—OZg/Wfl g) ‘9‘ BE L32. 188 in l“o yield. 59 Subsequently 882 was oxidized to ’b’b’b sulfoxide 150 and to the sulfone 191. "b ’b 'L’L’b A comparison of the 1H NMR absorptions of $82 - 88% 115 Table u. A Comparison of the 1H NMR Chemical Shifts of Compounds 789 — 757 with Those of Biphenylene. CDCI3 100 MHZ H ,H H H “2 "3 1 2 2 t/ 6.86 (bs) 6.149 (s) s \ V 189 H l"2 H3 1 . 6.6 I \Ffi/so 7 51 (m) 5 (S) "2 "3 "1 V I9] “2 "$2 / O " \ 6.70 (H) 6.60 (H) 18 1 2 116 with that of biphenylene is given in Table “. In 88% both the benzene ring protons and thiophene protons are at higher field absorption than is normal in such systems. This result was ascribed to the paratropic contribution of the four—membered ring. Oxidation of 189 to 188 lead to a downfield shift of the benzene type protons (0.65 ppm), and further oxidation to 897 resulted in a further small downfield shift. These shifts were considered to be the result of removing the paratropic component from the four- membered ring. A different approach, construction of the five-membered heterocycles from a corresponding l,“—diketone, was ex— amined by Garratt.6O The reaction of tetraketone l 8 with phosphorus pentasulfide in pyridine gave a 5% yield of the thiophene fused cyclobutadiene 82%. Because the yield of 89% was not reproducible, no study has been done so far on this paratropic system. Therefore, whether the stability of 825 arises from the four phenyl substituents Ph Ph Ph Ph 0\ __J /o , s / \K Ph m: Ph Ph 192 I93 117 or from the stability of the parent structure itself is still not known. Although the synthesis of the parent dithiophene 885 61 has been attempted previously without success, we pursued this synthesis using our own approaches. The preparation of the target compounds 18“ - 186 has not yet been achieved. ’b’b’b ’VVL However, we have prepared several potentially useful pre— cursors. It is the purpose of this part of the thesis to illustrate the approaches we used and report the synthesis of the precursors we have prepared. RESULTS AND DISCUSSION 1. The Coupling of Thiophene Moieties by the Ullmann Reaction The Ullmann biaryl synthesis has been successfully applied to the coupling of five-membered heterocycles.62 For example, an Ullmann reaction was the key step in the synthesis of cyclopentadithiophene 827, a compound having two thiophenes fused to a cyclopentane ring. Compound 797 differs from the desired $85 by only one methylene group. 3-Bromo-“—thienyllithium reacted with 3-bromo- thiophene-“-carboxaldehyde to give bis(3~bromo—“-thienyl) carbinol 728, which was oxidized to ketone 825. On Br Br Br Br 1. mmui ‘\. /" 003 '\\ .z’ _) S Rag s ’ \ s "We s/ \ B\ /c OH 194 0 I95 NH2NH2 NaOH 6 Br Br 3\ /sé'_sfls\ / sEln-Blllis\ /S \ / __ _ WSZCuC|2 / \ 00 H 2 .128. I9_7 ms. 118 119 Wolff-Kishner reduction of the carbonyl group, dibromide 828 was obtained. Treatment of 858 with n-butyllithium followed by anhydrous cupric chloride provided cyclopenta— 20 ° . - dithiophene 787 in a 7 total yield Compound 827 under went autooxidation rapidly with oxygen in the presence of \ ..—— Scheme 13 co,“ 63 potassium t-butoxide. An ionic mechanism similar to that of the Haller—Bauer reaction was proposed for the ring cleavage (Scheme $5). Although many factors are involved in the conversion 120 of £57 to $58, the existance of severe strain in the central five-membered ring was assumed to be the main driving force for the facile ring Cleavage. This assumption implies that the closure of any bithienyls 892 (n=0 or 1) is likely to be difficult, especially if a four—membered ring is to be constructed (Scheme 88). Scheme 88 Nevertheless biphenylene, the benzenoid analog of %85, was successfully prepared by this approach, via either benzyne dimerization or intramolecular cyclization of 2,2'-dihalo- biphenyls (Scheme %§) 58 Therefore, it seemed worthwhile first to try this straightforward approach. In several instances,58 biphenyl- enes cannot be obtained through the original Ullmann coupl— ing of an o—dihalobenzene. It was therefore not surpris- ing that when 3,“-dibromothiophene was heated at 250°C with active copper powder none of the desired compound 785 was found; only tarry products were formed. Modifica- tions have been developed for broader application of this 121 Br “"2 isoamyl-ouo j cozn . X X W “”2” o OTO X3 sz,l2 Scheme 85 6“ coupling reaction via lithium-copper transmetallation. Thus 3,“—dibromothiophene 888 was chosen as the starting material with the hope that under high dilution conditions €88 might undergo head to head coupling. However, when 888 reacted with excess n-butyllithium followed by an- hydrous CuC12, only 3,3'-dibromobithienyl 887 could be isolated from the reaction mixture. Analogously, no coupl- ing product of 3,“-diiodothiophene was found when its solution in cyclohexane was irradiated at 253 nm with a mercury lamp. 122 lies/Eu; CMNZ S \\ //S 533/ It seemed possible that further lithiation of 881 followed by metal exchange with cupric chloride might result in an intramolecular cyclization to 185. However 8 \ / S\\ / S n_-Bui_> S\ ’8 + trimer (mag /’ ‘\. Br Br / \ 20] S 223 203 -B l' /\ n U! A, §a§:--T/' CuCl2 W 123 the literatureél4 shows that 3,3'-dibromobithiophene only underwent dimerization to 888 and trimerization to trimer £81, with none of the expected 185 being formed. The ring strain in 185 could be the reason for lack of success in this cyclization. Therefore, 3,3'-dibromo-2,2'5,5'-tetra- methylbithienyl 885 was selected as a starting material because it was reasoned that steric hindrance between two methyl groups would change the preferred conformation of 185 in favor of bond formation between 3— and 3'- positions (see structure 885). However, on treatment of 888 with n—butyllithium followed by anhydrous cupric chloride, B s ' s S m» w 0 - m ' Br Br Br Br Br U m : 1. CUCIZ A , f 2. H20 :\:E/ _____ -s / s ‘ S ”. Ltif (F “~ ” ‘Efidz-j 12“ reduced product 39% was obtained in 75% yield instead of the required 59%. Lack of coupling was presumably due to the pronounced "ortho effect" of the methyl and bromine substituents. From these results and literature precedents,61 we felt that this approach had to be abandoned. Preparation of the anthranilic acid analog $92 was also tried, with the hope that a 3,U-thiophyne could be 11:2?" M lM—> "”—>"°2 M ":0 | N02 circus { 200' : 3 ‘V S S \ / e — ----- \ / cozn NH2 coal! «02 209 prepared and dimerized. The first two steps have been carried out (to make 38%), but in disappointingly low yield. Therefore, further work using this approach was dropped. 125 2. Construction of the Intermediate Synthetic Target %%2 Since methods which are successful for the preparation of biphenylene proved unrewarding for synthesizing léé, attention was shifted to a different precursor which already contains the required four-membered ring. One dithiophene derivative lgg has already been pre- pared using this approach,60 although the preparation of other heterocycles (gig, fill) by the analogous reaction 65,66 were not successful. The synthesis still seemed a reasonable strategy for preparing the parent dithiophene 95% ° mm F’J\ m P“ / "95;“ 210 Ph Ph \ ("“4)2003 k / /¢’%> H" N" 3% \ 21L H1 Ph The photodimerization of dimethyl fumarate in the solid state provides only one stereoisomeric cyclobutane dimer été’ with the cis-trans-cis configuration (due to the fixed 126 67). crystal lattice structure of the monomer The technique used in the dimerization consisted of irradiating a thin layer of monomer glg on the surface of a glass plate, using 300 nm light in a Bayonet reactor. Under these conditions monomer aka underwent 60% conversion to cis, trans, cis- l,2,3,U—tetracarbomethoxycyclobutane %l% in 2“ h. The tetra- ester %l% was hydrolyzed quantitatively to %l% with concen— trated hydrochloric acid, and the acid was converted to its salt glg with four equivalents of aqueous sodium hydroxide. However use of the Volhard-Erdman thiophene synthesis pro- 68 cedure, i.e., pyrolysis of the salt glg with excess of phosphorus pentasulfide, did not provide any of the desired dithiophene 18%; all of the starting salt 21% was recovered. ’VL 'Vb “We COOMe We 00" coon hv S_ I/C£§::%E;Jl HCI 5_ 000m 000m oOOMe ooou coon 212 213 N 214 00Na cooua -‘. 15:3\ @255 S L“/' 67’ ,r ‘- ' COONa 00Na 2fl£5 127 Consequently, another synthetic plan was devised start- ing from glé. The tetraester was reduced with lithium aluminum hydride. Difficulties have been reported69 in the isolation of the tetraol glé, due to complex formation between the polyol and aluminum salts from which the product [All \/ 213 could not be separated conveniently. However, we found that if the reduction was carried out in anhydrous dioxane followed by a very careful quenching process (the amount 128 of water added is critical), tetraol 8&8 can be separated in a pure state after passing the oily product through a cationic exchange resin column. The tetraol was charac- terized as its benzoate by treatment of 3&8 with benzoyl chloride. Tetraol 3&6 is very susceptible to intramolecular ether formation. For example, when glé was reacted with phosphorus tribromide in pyridine the major product isolated from the reaction mixture was identified as all. The mass spectrum showed a parent peak at m/g 140 and the 1H NMR (CDC13) spectrum had peaks at 52.ou(d, U H lbc=u Hz), c’ 3 38(dd, U Hb, gab=9 Hz, gbc=u Hz), and 3.98(d, u Ha, £ab=9 Hz). Decoupling results at 180 MHZ were also con- sistent with this assignment. Upon irradiating the peak at 52.uu, the spectrum became two equal intensity doublets at 53°38(iab=9 Hz) and 3.98(gab=9 Hz). Similarily, when ir- radiation was done at 63.38, the spectrum consisted of only two singlets, at 62.MU and 3.38. Compound gig was also obtained in a 65% yield by the treatment of %%Q with p-toluenesulfonic acid in dioxane. Surprisingly, the conversion of 6&8 to its tetra- bromide gl8 could be accomplished by reacting PBr3 with g%é which was coated on sea sand. The reaction gave glé in 52% yield. Subsequently, tetrabromide gl8 was treated with sodium disulfide in 95% ethanol to give bis-thioether glg quantitatively. 129 Conversion of 219 to the target compound 185 was tried ’b’b’b ’L’b’b by reacting 3%2 with BBQ in refluxing solvent, or by at- tempting to chlorinate and dehydrochlorinate the bis-sul— fide 212. However 21% could not be readily dehydrogenated ’VL ’VL nor could it be chlorinated with N—chlorosuccinimide, even under forced conditions. DDQ , A > N R S / toluene ' ' 25 CCI4 ' Therefore, it seemed desirable to prepare a similar syn- thetic target, such as Egg, with a more reactive functional group which might enhance the reactivity of the compound for later manipulation. 3. Construction of the Intermediate Synthetic Target £39. The stabilization of a carbanionic center by an ad- Jacent sulfone group has been widely used for many valuable transformations in organic synthesis.70 Although they are not many, some methods for the reduction of the sulfone group to a sulfide are available.71 Consequently, another 130 approach to léé was planned as outlined in Scheme l6. E E /\r—1/\ 1. Base A (a) A K/L—JtfiF’ S 9%: 5 L; 2.22. ‘ ‘ -HE ,’ S'~\ ”'S ‘<:::::: §:::[::1:::§<$:fl / \ \ Scheme 16 ”MN One possible way to prepare gag would be the partial oxidation of a%% with a stoichiometric amount of oxidizing agent, but the time consuming process for the dimerization of dimethyl fumarate and the modest yield (52%) of 3%8 from 8&8 discouraged us from repeating the whole sequence. The preparation of %%Q therefore was carried out as follows: photochemical cycloaddition of 2,5-dihydrothiophene-l,l- dioxide and maleic anhydride in acetone (quartz reactor, high pressure mercury-vapor lamp), gave the adduct 33% in 1 h with a yield of 60%.72 The reduction of %%% with lithium aluminum hydride gave ggg, which was treated with 131 0" 0H SW2 255% PBr3 sand TsOH Br Br S < Nazs 532 532 sq: 220 224 223 DIBAl gig \V PBr3 by the same technique used for the preparation of €18, giving dibromide ggg in 82% yield. Reaction of gag with sodium disulfide afforded the sulfone-sulfide ggg in good yield. Sulfone 220 could be reduced to sulfide 212 by ’VL’L ’V'b 132 diisobutylaluminum hydride with a 97% yield (20% conversion). Dehydrogenation of sulfone-sulfide gag was tried by the following processes: DDQ oxidation, catalytic dehydrogena- tion with Pd/C and S or Se dehydrogenation. Compound 22g remained inert, although some decomposition occurred with prolonged reaction times. 000 . A \ / toluene 7' N'R' I M ’A% N.R. EtOH S , 4» 2E \ > N.R. More surprisingly, abstraction of the a-H adjacent to sulfone group was also not successful. All efforts to generate the carbanion by a variety of bases, or trap the carbanion (assuming it was generated) with different brominating agents, were in vain. Investigations have been conducted on the acidity of cyclic sulfones.73 The acidity heavily depends on being able to achieve the required pyramidal geometry of the carbanion. As the ring size becomes smaller, the shorter chain enforces a more rigid conformation for the pyramidal 133 1. Base 5 502 + S.NL 2. E base: n-BuLi, s-BuLi, t-BuLi, LDA. \/ + E : Br I2, BrCN, bromo-Meldrum's acid. 2) carbanion and results in an increasing repulsion between the charge on the carbanion and the partial negative charge on the sulfone oxygen atom. Therefore, the carbanion is _ Nfié} l UL—w s l /\s | m K/ 219 — l /\ /\ 8:29 /\ l V 135 so: — of | \/ 1 2&3 l R—@ s o H 403) [e 13H destabilized, with a resulting decrease in the acidity of the sulfone. This might be the reason for the unexpected difficulty in abstracting the a-H from sulfone-sulfide 22g. In summary, we reported the preparation of disulfide 212 and sulfone 220 along with their oxygen containing ’b’b ’V’b’b analogues 217 and 223. Although conversion to the target ’VVL ’VL compounds encountered more difficulties than was anticipated, the preparation of 18“ — 186 via this approach seems still ’L’b’b ’D’L’b plausible. Methods such as those outlined in Schemes Ix and 18 have not yet been tried. They avoid the difficulties raised during the manipulation of the a position of the sul- fone, because functionalization would start at an earlier stage of the synthetic sequence. E E V/\ - . ./’\ 4HE {Q§:-1[:;} W wL—\ E E Scheme17 L g l.-2HE /\ % -2uzo 4x 8 U 05 so =>s EU 1(0) \‘;Q-—~\g/ \gfi’ "‘ E Scheme 18 EXPERIMENTAL l. U,u'—Dibromo-3,3'-bithienyl (@011 To a precooled (—78°C) solution of 3,U-dibromothio- phene76 (l g, 4.2 mmol) in 150 mL of anhydrous ether under nitrogen, was added dropwise with stirring a solution of n-butyllithium (90 mmol in 50 mL of hexane). The result- ing mixture was stirred for 1.5 h. Anhydrous cupric chloride (1.u g, 10 mmol) was then introduced all at once (as crys- tals) while keeping the reaction temperature at —78°C. The solution turned brown slowly and was stirred for another 2 h. After removing the dry ice bath, the reaction mixture was stirred overnight and worked up by washing with u N HCl, and then water. The organic layer was dried over magnesium sulfate and concentrated under reduced pressure. Recrystal- lization of the residue from ligroin gave product 201 (0.42 g, 63%) as white crystals; mp 122-12U0C (lit.7u 127-129°C); lH NMR