PART 1 SWQIES CONCERNED WETH THE EYfoéxESES CF QCTALEEQE ANfl BICYCLE [@3‘0]BEA-ha,5,7,9nFENTAENVE PART 11 THE REARRANGEMENTS OF MONOSUEfiTWUTED CYCLOOCTATETRAEHEB Thesis 5o;- Nac Dogma of DE. D. MiCHlGAN STATE UNIVERSITY Greg A. Bullock 1968 g is r' ,_ p, Viv-f L i u .- .4 4< 1 ' é ‘lichigan Sm University This is to certify that the thesis entitled Part I: STUDIES CONCERNED WITH THE SYNTHESIS OF OCTALENE .,AND BICYCLO[6.2.0]DECA-l,3,5,7,9-PENTAENE Part II: THE REARRANGEMENTS 0F MDNOSUBSTITUTED CYCLOOCTATETRAENES presented by Greg A. Bullock has been accepted towards fulfillment of the requirements for Ph.D. Chemistry degree in Major pro sor Datuéymililfé? 0-169 -.. .-—*<7-» 4 _.._I-_ ._ .I -a_..-_.-__.IV.._._~_ .I.‘—..._._.,._._-_.—_ ._ .-_ _ ABSTRACT PART I STUDIES CONCERNED WITH THE SYNTHESIS OF OCTALENE AND BICYCLO[6.2.0]DECA-1,3,5,7,9-PENTAENE PART II THE REARRANGEMENTS OF MONOSUBSTITUTED CYCLOCCTATETRAENES by Greg A. Bullock The initial goal of this investigation concerned the synthesis of bicyclo[6.6.0]tetradeca-1,3,5,7,9,11,13-heptaene (octalene). Several synthetic approaches leading to the preparation of octalene were attempted. None of these ap- proaches, however, was successful. During the course of this investigation a new synthetic route to [2.2]paracyclo- phanes was discovered. 4.5.12,13—Tetracarbomethoxy[2.2]para— cyclophane was synthesized from dimethyl 3,6-bis(hydroxy— methyl)-1,4-cyclohexadiene-1,2-dicarboxylate diepftoluene- sulfonate by solvolysis in buffered acetic acid. The bis- (hydroxymethyl)cyclohexadiene was prepared from a Diels- Alder reaction between trans,trans—2,4-hexadiene-1,6-diol and dimethyl acetylenedicarboxylate. The synthesis of bicyclo[6.2.0]deca-1,3,5,7,9-pentaene was also attempted. Synthetic approaches to the bicyclic decapentaene involved the attempted condensation of both dehydrocyclooctatetraene and N{N-diethylaminocyclobctatetra- ene with substituted acetylenes. The condensation of N,N,diethylaminocyclobctatetraene with dimethyl Greg A. Bullock acetylenedicarboxylate yields dimethYLJSZ-naphthalenedicar- boxylate as the only isolable product. A logical proposal for the formation of this product is that the condensation product, dimethyl l-N,N-diethylaminobicyclo[6.2.0]deca— 2,4,6,9—tetraene-9,10-dicarboxylate, ring opens to produce dimethyl l-N,N—diethylaminocyclodecapentaene-2,3-dicarboxyl- ate. The [10]annulene could then cyclize accompanied by the loss of diethyl amine to yield the observed product. The preparation of bicyclo[6.2.0]decapentaene was attempted also from tetramethyltricyclo[6.2.0.03I6]decane-2,7-dione-4,5,9,10- ‘tetracarboxylate without success. The stereochemical con- figuration of tetramethyl bicyclo[6.2.0.0316]decane-2,7- dione-4,5,9,10-tetracarboxylate was established from its nmr spectrum and from the configuration of tetramethyl 11—oxa- tetracyclo[4.4.1.02t5.07:1°]undecane-3,4,8,9-tetracarboxylate, one of its reaction products. The cyclohexadione ring of the tetramethyl dione exists in a boat form with the two cyclobutane rings gig fused to the pseudoequatorial posi- tions. The carbomethoxy groups are believed to be gig to each other and trans to the cyclohexadione—cyclobutane ring junction. Chloro- and bromocyclobctatetraene thermally rearrange to trans—fi-chloro- and Eganng-bromostyrene, respectively. The solvolysis of chlorocyclodctatetraene in methanolic solu— tions containing various nucleophiles is also discussed. Methanolysis of chlorocyclooctatetraene yields a mixture of Greg A. Bullock trans—B-chlorostyrene, trans-B-methoxystyrene, and phenyl- acetaldehyde dimethyl acetal. If methanolfi§_is used as the solvent, the phenyacetaldehyde dimethyl acetal is labeled in the benzylic position. *Methanolysis of chlorocyclobcta- tetraene in the presence of sodium methoxide produces, along with trans-B-chlorostyrene and trans-B-methoxystyrene, meth- oxycyclooctatetraene whereas methanolysis in the presence of lithium bromide yields trans-B-chlorostyrene, trans-B— methoxystyrene, and trans—fi-bromostyrene. These results are compatible with the View that the valence tautomer of chlorocyclooctatetraene, 1-chlorobicyclo[4.2.0]octa—2,4,7- triene, ionizes to the corresponding bicyclic octatrienyl carbonium ion. A mechanism for the formation of the pro- ducts in these reactions is discussed. The rearrangement takes a different pathway when diethylamino- or Efbutoxy- cyclooctatetraene are heated. In both of these cases the a-substituted styrenes are produced. A mechanism for the latter rearrangement is proposed. PART I STUDIES CONCERNED WITH THE SYNTHESIS OF OCTALENE AND BICYCLO[6.2.0]DECA-1,3,5,7,9-PENTAENE PART II THE REARRANGEMENTS OF MONOSUBSTITUTED CYCLOOCTATETRAENES BY ‘09 Greg A. Bullock A THESIS Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Chemistry 1968 ACKNOWLEDGMENT The author wishes to express his sincere gratitude to Professor Eugene LeGoff for encouragement, guidance and en— thusiasm throughout the course of this investigation and also for arranging financial support from September, 1967 to August, 1968. Appreciation is extended to Michigan State University for a teaching assistantship from September, 1965 to August. 1966, t0 Lubrizol Corporation for a research fellowship from September, 1966 to August, 1967 and to Petroleum Research Foundation for a research fellowship from September, 1967 to August, 1968. ii To Lisa iii TABLE OF CONTENTS PART I HISTORICAL AND INTRODUCTION . . . . . . Previous Approaches to Octalene . . . . . . . Previous Approaches to Bicyclo[6.2.0]deca- 1,3,5,7,9-pentaene . . . . . . . . . . . . RESULTS AND DISCUSSION . . . . . . . . . . . . Octalene . . . . . . . . . . . . . . . . . . . Bicyclo[6.2.0]deca-1,3,5,7,9-pentene . . . . EXPERIMENTAL . . . . . . . . . . . . . . . . . . General Procedures . . . . . . . . . . . . Diels—Alder Adduct of 1,4—Dipheny1-2-styryl- 1,3-butadiene and Maleic Anhydride . . . . 1,4,5,8-Tetrahydro-1,4,5,8-naphthalenetetracar- boxylic Acid . . . . . . . . . . . . . . . . Dimethyl 3,6—Bis(hydroxymethyl)-l,4-cyclo- hexadiene—l,2-dicarboxylate . . . . . . . . Dimethyl 3,6—Bis(hydroxymethyl)-1,4—cyclohexa— diene-1,2-dicarboxylate Di-pftoluenesulfonate Attempted Ring Expansion of Dimethyl 3,6-Bis- (hydroxymethyl)-1,4-cyclohexadiene-1,2-di— carboxylate Di-pftoluenesulfonate . . . . 9,10-Diethylbicyclo[6.2.0]deca-l,3,5,7,9- pentaene (Attempted) . . . . . . . . . . . 9-Ethyl-10-methoxybic clo[6.2.0]deca-1,3,5,7,9- pentaene (Attempted ; Preparation of EjAmyl- oxycyclofictatetraene . . . . . . . . . . . . N.N-Diethylaminocycloactatetraene . . . . . . iv Page 10 11 15 15 27 45 45 45 46 47 48 48 50 51 52 TABLE or CONTENTS (cont.) PART I ' Page Dimethyl l-N,N-Diethylaminobicyclo[6.2.0]deca- 2,4,6,9-tetraene—9,lO-dicarboxylate (Attempted); Preparation of Dimethyl 1,2-Naphthalenedi- carboxylate . . . . . . . . . . . . . . . . 53 Tetramethyl Tricycld[6.2.0.03:5]decane-2,7- dione-4,5,9,10-tetracarboxylate . . . . . . 54 Tetramethyl Tricyclo[6.2.0.03I6]decane—2,7-di- one-2,7—bisoxime-4,5,9,10-tetracarboxylate . 54 Tetramethyl Tricyclo[6.2.0.03:6]decane-2,7-dione— 2,7-bistosylhydrazone-4,5,9,10-tetracarboxylate 56 Tetramethyl 2,7-Dihydroxytricyclo[6.2.0.03'6]- decane-4,5,9,10-tetracarboxylate . . . . . . 56 Tetramethyl 2,7-Dibromotricyclo[6.2.0.03:5]- decane-4,5,9,10-tetracarboxylate (Attempted) 57 Tetramethyl 2,7-Dihydroxytricyclo[6.2.O.03'6]- decane-4,5,9,10-tetracarboxylate DifEf toluenesulfonate . . . . . . . . . . . . . . 58 Tetramethyl 11-0xatetracyclo[4.4.1.02[5.07:101- undecane-3,4,8,9-tetracarboxylate . . . . . 59 ll-Oxatetracyclo[4.4.1.02I5.07:1°]undecane- 3,4,8,9-tetracarboxylic Acid Monohydrate . . 60 PART II HISTORICAL AND INTRODUCTION . . . . . . . . . . . 62 RESULTS AND DISCUSSION . . . . . . . . . . . . . 64 EXPERIMENTAL . . . . . . . . . . . . . . . . . . 80 Thermal Rearrangement of Chlorocyclobctatetraene to trans-B—Chlorostyrene . . . . . . . . . . 80 Rearrangement of Chlorocyclobctatetraene in Refluxing Methanol . . . . . . . . . . . . . 80 TABLE OF CONTENTS (cont.) PART II Page Rearrangement of Chlorocyclobctatetraene in Methanolic Sodium Methoxide . . . . . . . . 81 Rearrangement of Chlorocyclobctatetraene in Methanolic Lithium Bromide . . . . . . . . . 82 Tetracyanoethylene Dials-Alder Adduct of Chloro- cyclobctatetraene . . . . . . . . . . . . . 83 Thermal Rearrangement of N,N-Diethylaminocyclo— octatetraene . . . . . . . . . . . . . . . 83 Thermal Rearrangement of EfButoxycyc108cta- tetraene . . . . . . . . . . . . . . . . . . 84 Tetracyanoethylene Dials-Alder Adduct of E- Butoxycyclobctatetraene . . . . . . . . . . 85 The Diels-Alder Reaction of E-Butoxycycloocta- tetraene with Dimethyl Acetylenedicarboxylate 85 LITERATURE CITED . . . . . . . . . . . . . . . . 95 vi SCHEME II. III. IV. V. VI. VII. VIII. IX. XI. XII. LIST OF SCHEMES A Proposed Synthetic Route to Triphenyl- octalene . . . . A Proposed Synthetic Route to Octalene The Synthesis of 3,6-Bis(hydroxymethyl)-1,4- cyclohexadiene-l,2-dicarboxylate Difipf toluenesulfonate The Chemical Reactions of Dehydrocyclobcta- tetraene . . . . The Chemical Reactions of Benzyne A Proposed Synthetic Route to Bicyclo[6.2.0]— deca-1,3,5,7,9-pentaene A Proposed Synthetic Route to Tricyclo- [6.2.0.03I6]decane-2,7—dione-4,5,9,10— tetracarboxylic Acid Proposed Synthetic Routes to Tetramethyl Tri— cyclo[6.2.0.03I6]deca-1,7—diene-4,5,9,10— tetracarboxylate A Mechanism for the Rearrangement of Chloro— cyclooctatetraene in Methanol-d Methoxide . . . . A Mechanism for the cyclobctatetraene Bromide . . . . . A Mechanism for the N.N—Diethylamino- tetraene . . . Rearrangement of Chloro- in Methanolic Lithium Thermal Rearrangement of and EfButoxycyclobcta— vii Mechanism for the Rearrangement of Chloro- cycloocatetraene in Methanolic Sodium Page 16 18 24 29 3O 34 36 38 69 7O 71 77 LIST OF FIGURES FIGURE Page 1. Infrared Spectrum of 1,4,5,8-Tetrahydro- 1,4,5,8-naphthalenetetracarboxylic Acid . . 87 2. Infrared Spectrum of 4,5,12,13-Tetracarbo- methoxy[2.2]paracyclophane . . . . . . . . . 87 3. Infrared.Spectrum of EfAmyloxycyclobctatetraene 87 4. Infrared Spectrum of N,N-Diethylaminocyclo— Octatetraene . . . . . . . . . . . . . . . . 88 5. Infrared Spectrum of Tetramethyl Tricyclo— [6.2.0.03I5]decane-Z,7-dione—2,7—bisoxime- 4,5,9,10-tetracarboxylate . . . . . . . . . 88 6. Infrared Spectrum of the Diacetate of Tetra— methyl Tricyclo[6.2.0.03'6]decane-2,7—dione-2,7- bisoxime-4,5,9,10-tetracarboxylate . . . . . 88 7. Infrared Spectrum of Tetramethyl Tricyclo- [6.2.0.03I6]decane-2,7-dione-2,7-bistosyl— hydrazone-4,5,9,10—tetracarboxylate . . . . 89 8. Infrared Spectrum of Tetramethyl 2,7-Dihydroxy- tricyclo[6.2.0.03t6]decane-4,5,9,10-tetracare boxylate . . . . . . . . . . . . . . . . . . 89 9. Infrared Spectrum of Tetramethyl 11-Oxatetra- cyclo[4.4.1.02v5.07!1°]undecane-3,4,8,9-tetra- carboxylate . . . . . . . . . . . . . . . . 89 10. Infrared Spectrum of 11-Oxatetracyclo- [4.4.1.02I5.07:1°]undecane-3,4,8,9-tetracar- boxylic Acid Monohydrate . . . . . . . . . . 9O 11. Infrared Spectrum of Methoxycyclooctateraene . 90 12. Infrared Spectrum of the Dials—Alder Adduct of EfButoxycyclobctatetraene and Tetracyano- ethylene . . . . . . . . . . . . . . . . . . 9O 13. Nmr Spectrum of 4,5,12,13-Tetracarbomethoxy— [2.2]paracyclophane . . . . . . . . . . . . 91 14. Nmr Spectrum of thmyloxycyclobctatetraene . . 91 viii LIST OF FIGURES (cont.) FIGURE Page 15. Nmr Spectrum of Tetramethyl Tricyclo- [6.2.0.03p5]decane-2,7-dione-2,7—bisoxime— 4,5,9,10-tetracarboxylate . . . . . . . . . 91 16. Nmr Spectrum of the Diacetate of Tetramethyl Tricyclo[6.2.0.03I5]decane-2,7-dione-2,7-bis- oxime-4,5,9,10-tetracarboxylate . . . . . . 92 17. Nmr Spectrum of Tetramethyl Tricyclo- [6.2.0.03I5]decane-2,7-dione-2,7-bistosyl- hydrazone—4,5,9,10-tetracarboxylate . . . . 92 18. Nmr Spectrum of Tetramethyl 2,7-Dihydroxytri- cyclo[6.2.0.03I5]decane—4,5,9,10-tetracar- boxylate . . . . . . . . . . . . . . . . . . 92 19. Nmr Spectrum of Tetramethyl 11-Oxatetracyclo- [4.4.1.02v5.07!1°]undecane—3,4,8,9—tetra— carboxylate . . . . . . . . . . . . . . . . . 93 20. Nmr Spectrum of trans-B-Methoxystyrene . . . . 93 21. Nmr Spectrum of Methoxycyclobctatetraene . . . 93 22. Nmr Spectrum of N,N-Diethylaminocycloocta— tetraene . . . . . . . . . . . . . . . . . . 94 23. Nmr Spectrum of 11-Oxatetracyclo[4.4.1.02'5.O7'1°]- undecane-3,4,8,9-tetracarboxylic Acid Monohydrate . . . . . . . . . . . . . . . . 94 24. Nmr Spectrum of a-N,N-Diethylaminostyrene . . 94 ix PART I STUDIES CONCERNED WITH THE SYNTHESIS OF OCTALENE AND BICYCLO[6.2.0]DECA-1,3,5,7,9-PENTAENE HISTORICAL AND INTRODUCTION Until recently the term "aromatic" as applied to or- ganic compounds was considered synonymous with "benzenoid" (meaning derived from benzene). The ability of these com- pounds to undergo substitution reactions had long been con- sidered as a principle criterion of aromaticity. However, as a result of the investigations into the chemical reac- tions of azulene, chemists recognized that cyclic compounds other than benzene derivatives can undergo substitution re- actions instead of addition reactions. These observations led to such compounds being classified as "non—benzenoid aromatics". This thesis, in part, will concern the attempted syn- thesis of two such non-benzenoid aromatic compounds: octa— lene l and bicyclo[6.2.0]deca-1,3,5,7,9-pentaene g, /\ \/ l 2 The synthesis of either of these two compounds and the study of their physical and chemical properties would serve to support existing theories dealing with the criteria for aromaticity or cause them to be substantially modified. The most straightforward manner of detecting aromaticity in either of these two compounds would be by obtaining their 2 3 nuclear magnetic resonance spectrum. The magnetic field which is applied in obtaining such spectra causes an induced circulation of the delocalized pifelectrons of an aromatic molecule and, therefore, ring currents are established in the molecule. This "ring current effect" is responsible for the deshielding of the protons or any other groups oriented such that they are contained outside of and in the plane of the aromatic ring. Based on this method of experimental in- vestigation, the following definition of an aromatic compound has been suggested: "The essential feature is a ring of atoms so linked that pifelectrons are delocalized right round the ring. We can define an aromatic compound, there— fore, as a compound which will sustain an induced ring current. The magnitude of the ring current will be a func- tion of the delocalization of pirelectrons around the ring and therefore a measure of aromaticity."(1). Although the increased stability of benzene has been recognized for many years, it was not until the 1930's when HUckel, Pauling and others developed the molecular orbital (M0) and valence bond (VB) methods that a theoret- ical basis for this stability was established (2,3). The most important contributions of these theories were that they proved to be well founded from a physical point of view and they allowed the calculation of the thermochemical resonance energy, a measurable quantity. The change in emphasis of aromaticity from chemical behavior to physical properties illustrates the great success of semiquantitative 4 calculations of ground states compared with that of chemical reactivity. The delocalization energy and resonance energy of a molecule are generally used as a criterion for aromaticity. Delocalization energy differs from resonance energy in that the latter has been corrected for strain energy. Thus, de- localization energy is calculated from theoretical treatments and is adjusted by accounting for strain energy to give the resonance energy. Simple Hfickel molecular orbital theory has been used to calculate the delocalization energy of both octalene and bicyclo[6.2.0]decapentaene. In these procedures overlap was neglected, all coulomb integrals were considered equal and all resonance integrals between adjacent carbon atoms were considered equal. The delocalization energy of octalene was calculated (5) to be 4.19B (0.305 per pi-electron), but the delocalization energy is found to be much less (0.285) when a more sophisticated treatment (6) is used. Allinger and Gilardeau (7) have performed calculations that indicate that the non-planar form of octalene is more stable than the planar form by over 50 kcal/mole. They predict that the molecule will exist in a non-planar configuration with alter- nating long and short bonds (similar to the tub configura— tion of two fused cycloBctatetraene molecules). Allinger and Gilardeau (7) have also calculated the expected electronic spectrum for both the planar and non-planar forms of octa- lene. The calculated delocalization energy of 5 bicyclo[6.2.0]decapentaene (8) was 2.845 (0.285 per pi-elec- tron) which is somewhat less than delocalization energies of benzenoid aromatics of comparable size (naphthalene is 0.375 per pifelectron). »Although the pifelectron delocalization energy provides a measure of the overall stability of a mol- ecule, other contributing factors may in some instances lower or cancel its calculated effect. In each of the above individual cases, the possibility must be taken into account of a reduction in stability due to the considerable angle strain which may be present or due to the possible non—co— planarity of eight-membered rings. Bicyclo[6.2.0]decapenta- ene, at least, may be more stable in this respect than octalene since the fused four-membered ring should help to flatten out the eight-membered ring. The fundamental concept underlying current ideas on aromaticity is known as Hfickel's rule (3). In its general form, the rule derived by Hfickel states that "amongst fully conjugated, planan monocyclic polyolefins only those posses- sing (4n + 2)‘pi-electrons, where n is an integer, will have special aromatic stability." In its earliest form, Hfickel's rule was stringently restricted to monocyclic hydrocarbons. Examination of num- erous experimental results has indicated that it is also applicable to bicyclic fused systems. This liberalization of the "4n + 2" rule is readily understood if it is assumed that the bridging bond does not introduce any noticeable change in the aromatic character of the molecule but acts 6 only to preserve the coplanarity of the system. Such a simplification is made more admissible since calculations show that the perturbation caused by the bridging bond in both naphthalene and azulene is small and that it cannot affect the general aromatic nature of the system. If the bridge bond is disregarded in this manner, naphthalene and azulene would be derivable from the ten-membered cyclic hydro— carbon, cyclodecapentaene, which contains ten pifelectrons (i.e. n = 2 in the Hfickel 4n + 2 formula) so these com- pounds should possess aromatic character. In its general form the rule may be stated as "any plane (or nearly plane) fused system containing no atoms common to more than two rings will be aromatic if the number of i—electrons in it is equal to 4n + 2 (where n is a whole number)“ (4). According to the generalized rule, octalene (14 p;- electrons; n = 3) and bicyclo[6.2.0]decapentaene (10 pi- electrons; n = 2) would both be expected to exhibit aromatic character. Another proposal which has been used to distinguish between normal aromatics and pseudoaromatics (polyolefins) is "Craig's rule". This distinction, as developed by Craig (9), is based on symmetry and applies only to molecules having an axis passing through at least two piecenters as illustrated by the following molecules. 0 a» 7 A molecule such as bicyclo[6.2.0]decapentaene that does not fulfill the symmetry requirement cannot be characterized by this method. The rule is applied by first labeling each carbon atom of one of the Kekule-type structures with an alph§_or a beta (spin symbols) alternately as far as possible so that each end of a double bond has opposite spins. The molecule is then rotated 1800 about its axis of symmetry to give another canonical form. 53‘ 65:16 d B B ' 0 and Be: (1 Ct B 6516 ~ I “a l B'B . (1 C1 I BPS a A B C There are two symbols associated with this operation: p and q. P is equal to the number of pifcenters affected by the rotation and q is equal to the number of alphas and bg£2§_which must be interchanged so that the original label- ing scheme is restored. The values of p and q are then applied to the fol- lowing equation: 8 If gpi_is odd, the molecule has a nontotally symmetrical ground state and is a pseudoaromatic (polyolefin). If phi is even, the molecule has a totally symmetrical ground state and is a normal aromatic. A) Benzene: p = 2 because two pifcenters have been interchanged (molecule has been transformed into a Kekulé isomer); q = 0 since no change in the labeling system has occurred; p + q = 2 and therefore benzene is a normal aromatic since phi is even. B) Pentalene: p = 3 since three pifcenters have been affected in the 1800 rotation procedure; q = O for no change in the labeling has occurred; p + q = 3 and ghi_is odd so pentalene is a pseudoaromatic. C) Cyclobutadiene: p = 1 because one carbon atom has been interchanged; q = 0 since no change in the alpha and pg 3 labeling has occurred; p + q = 1 and cyclobutadiene is a pseudoaromatic. Since octalene possesses the necessary symmetry require- ments (an axis passing through two pi-centers), Craig's rule can be applied. For octalene, p = 6 since six pg—centers have been inter- changed in going from one Kekulé-type form to the other and q = 0 because the alpha and beta labeling system has 9 remained unchanged. The value of phi is even and therefore octalene has a symmetrical ground state and should be a nor- mal aromatic. Another classification that has been used to predict aromatic character is the alternant or nonalternant character of a molecule. Coulson and Rushbrooke (10) proposed that al— ternant or starrable compounds would be aromatic whereas non- alternant compounds would be pseudoaromatic. To be an alter- nant molecule it is meant that if alternant carbon atoms around the ring are starred no two starred carbon atoms will be adjacent; otherwise it is a nonalternant molecule. Naphth— alene would be an example of an alternant molecule while azulene is nonalternant. * * * , *- The underlying importance of this process is that in an * alternant molecule the signs of the atomic orbitals will al- ternate about the carbon atoms and there can be complete delocalization over the entire molecule. A nonalternant molecule, however, will have at least one position at which the sign of the atomic orbitals on adjacent carbons will be the same and there should be little, if any,‘pi-electron overlap across this position. Therefore, azulene should show reduced pi-electron overlap across the 9,10-bond and the resonance energy would be expected to be reduced due to this factor. 10 Application of this process to octalene and bicyclo- [6.2.0]decapentaene reveals that both molecules are alternant and should therefore show aromatic character. * * 9(- * * */\ * * *\/ a: Previous Approaches to Octalene To date, there has been only one report in the chemi- cal literature concerning a successful synthesis of an octalene derivative. Breslow and coworkers (11) prepared benzo[c]octalene 5-in a 1—2% yield from the reaction of 1,8-diformylcyclofictatetraene §_with the bisphOSphorane 4, This octalene derivative was not particularly stable, being CHO CH=P <93 CHO CH=P «>3 5.3. i 2. destroyed on exposure to air and, in part, by heat. The octalene protons are unshifted (1 4.30) from normal cyclo- Bctatetraene protons (T 4.31). The ultraviolet spectrum (A max 328 mu) resembles that of benzocycloactatetraene (12) (in which the eight-membered ring is in a tub conformation). This data would seem to indicate that the eight-membered 11 rings exist in an ordinary tub conformation and that benzo— octalene is not aromatic. Related evidence comes from the reaction (11) of 1,8- diformylcycloactatetraene §_with bisphosphorane 6, The pro- duct isolated from this reaction in 10% yield, dihydroBCtal- ene Z, resisted all attempts at dehydrogenation to octalene. This would suggest that octalene does not have any striking stability. (CD3 P=CHCH2 ) 2 g 6 . 474—. .7. Previous Approaches to Bigyclo[6.2.Q]deca-l,3,5,7,9-pentaene A few unsuccessful attempts to prepare bicyclo[6.2.0]- decapentaene have appeared in the literature. Elix, Sargent and Sondheimer (13) have prepared bicyclo[6.2.0]deca-1,3,5,7- tetraene g_by the photochemical cyclization of 7,8-dimethyl- enecycloBcta-1,3,5—triene 8, However, all attempts to con- —“’hv>l l 79’ §. I0 2. vert g'to the bicyclic decapentaene g_were unsuccessful. These same authors (14) attempted to prepare a precursor of 12 g, 9—chlorobicyclo[6.2.0]deca—1,3,5,7-tetraene 11, by irradi- ation of 7-methylene-8-chloromethylenecycloficta-l,3,5-triene ‘10. They found, however, that substitution of a chlorine CH2 -- c1 ééihv > ' I CHCl > a 10 11 atom for a methylene hydrogen in §_deactivates the exocyclic diene system and causes the endocyclic double bonds to be- come more photolabile. Masamune and co-workers (15) synthesized bicyclo[6.2.0]- deca-2,4,6,9-tetraene 1§_by irradiating a solution of bi- cyclo[6.1.0]nona-Z,4,6—triene-Eggpgf9-carboxaldehyde tosyl- hydrazone 14 in tetrahydrofuran containing an equivalent amount of sodium methoxide at 0°C. However, 15 readily isomerized to Eggpgf9,10-dihydronaphthalene 16 at tempera- tures above 0°. H — I NaOCH3 >00 C‘NNHTS w> > CO 00 14 15 16 During the preparation of this thesis, Schroder (16) reported the successful synthesis of two substituted bi- cyclo[6.2.0]deca-1,3,5,7,9-pentaenes. Thus, treatment of the cycloadduct 11 with potassium Efbutoxide resulted in 13 the formation of deep red 8-fluoro-95Efbutoxybicyclo[6.2.0]- decapentaene 19, an extremely air-sensitive but thermally stable compound. In an analogous fashion, B-Efbutoxybicyclo- F2 -— F 17 CH3Li — C12 _'_ > | C1 In -BuOK l EfBuOK ./’ ‘\\ [IF /’ \\::][fi \ / OE-Bu \ / oE-Bu [6.2.0]deca-1,3,5,7,9-pentaene gQ_could be obtained from 18. The nmr spectrum of 29 showed a multiplet around T4.0 for the six olefinic eight-membered ring protons while the cyclobutenyl proton absorbs at surprisingly low field (12.9).1 Schroder has offered the suggestion, based on the nmr data, that 12_and 29 may be best represented by a structure 1The cyclobutenyl protons of bicyclo[6.2.0]deca-1,7,9-tri- ene i. appear (15) at 73.40. \CH // IP- 14 such as 21, He has also pointed out, however, that addi- tional experiments are required to decide if bicyclo[6.2.0]- decapentaene displays delocalization of the‘pi-electrons analogous to the isomeric naphthalene and azulene systems. RESULTS AND DISCUSSION Octalene The initial goal of this investigation concerned the synthesis of bicyclo[6.6.0]tetradeca-1,3,5,7,9,11,13-hepta- ene (octalene). The first approach directed toward the synthesis of octalene involved a double Diels—Alder reaction between 1,4-diphenyl-2-styryl-1,3-butadiene‘22 and gig-3,4- dichlorocyclobutene g§_(Scheme I). Treatment of the product of this reaction 24_with base was expected to result in the elimination of four moles of hydrochloric acid yielding the desired product, triphenyloctalene 25, To ascertain if 1,4-diphenyl-2-styryl-1,3-butadiene‘22 would enter into a double Diels-Alder reaction, the reaction of gg_with maleic anhydride was performed as a model reac- tion. The 1,4-diphenyl-2-styryl-1,3-butadiene was synthe- sized by the method of Bohlman (17). Heating gg_and an ex- cess of maleic anhydride at 120° afforded a 68% yield of the desired double Diels-Alder adduct 26, The Diels-Alder re- action between 22_and gi§f3,4-dichlorocyclobutene was then attempted in refluxing xylene. A brown viscous material was isolated, which would not solidify or distill. Attempted purification by column chromatography produced a yellow glassy material, which again would not crystallize. The Diels- Alder reaction was also tried in refluxing benzene and tolu- ene. In both cases, a yellow, presumably polymeric, glassy material was isolated after column chromatography. A possible 15 Cl 16 Scheme I B: -2HCl (proposed) Cl B: —2HCl ———————9 (proposed) 17 explanation for the failure of this Diels-Alder reaction to occur could be that the rate of thermal isomerization (18) | + o ¢ ¢ Maleic Anhydride of cis—3,4-dichlorocyclobutene to cis, trans-1,4-dichloro- 1,3—butadiene is faster than the rate of the Diels~Alder reaction between 22 and cis-3,4-dichlorocyclobutene. The resultant cis, trans-1,4-dichloro-1,3-butadiene could then copolymerize with 1,4-diphenyl-2-styryl—1,3-butadiene 22, At this point, an alternate pathway for the synthesis of octalene was considered (Scheme II). This reaction sequence would involve the Birch reduction of 1,4,5,8-naphthalenetetracarboxylic acid 22, Lithium alum- inum hydride could then be used to reduce the tetra-acid 22_ to the corresponding tetraol 22, The tetratosylate 29, pre- pared from the tetraol 22, hopefully would ring expand di- rectly to octalene by solvolysis in sodium dihydrogenphosphate- acetic acid solution. 18 Scheme II Hozc COZH Na/EtOH,_ 28 HOzC COZH H02 02H 21 LiAlH4 TsOH2C CHZOTS HOHZC CHZOH TSOH2C CHonS HOH2C CH2 OH s9. 22 HOAc NaHz P04 19 This reaction sequence is similar to that employed by Dauben and Bertelli (19) for the synthesis of dihydroheptalene a- cozn CHZOTs HOAc ‘ —> ——> ? O O NaH2 P04 .0 C02H CHZOTS 31 There were two main uncertainties connected with this proposed path (Scheme II) to octalene. First, the Birch re~ duction of 1,4,5,8-naphthalenetetracarboxylic acid was not a known reaction and may be a difficult step to accomplish; second, the ring expansion of a non-benzenoid fused six-mem- bered ring to an eight-membered ring had not been reported.2 The Birch reduction of 21 to 1,4,5,8-tetrahydro-l,4,5,8- naphthalenetetracarboxylic acid, as previously mentioned, was not a known reaction. There are no reports of a tetracar- boxylic acid successfully undergoing the Birch reduction. 2gig-9,10-p_1_§_(hydroxymethyl)-9,10—dihydroanthracene distosyl- ate 2_has been shown to give mainly the cycloBctatrienol derivative li_on solvolysis with acetic acid buffered with sodium acetate; E. Cioranescu, A. Bucur, M. Banciu and C. D. Nenitzescu, Rey. Roumaine Chim., 29, 141 (1965) (Chem. Abstr., fig, 11456 (1965)). CHZOTS OAc CHonS 20 Sodium in liquid ammonia will reduce aromatic para-dicar- boxylic acids to give dihydro diacids and in all cases the diacids are non-conjugated, being reduced at the 1,4-position with respect to the carboxyl groups. Thus, terephthalic acid 22_and 1,5-naphthalenedicarboxylic acid 22_were reduced to 1,4—dihydroterephthalic acid 24_(20) and 1,4,5,8-tetra~ hydro-1,5-naphthalenedicarboxylic acid 22, (19) respectively. COZH COzH .22 > £2 c02H c02H COZH cozn ,c02H c02H Birch reduction of 1,4,5,8—naphthalenetetracarboxylic acid 2Z_would be expected to yield the intermediate acid 26, which should undergo further reduction to furnish the de- sired product 22, Sodium in liquid ammonia does not reduce isolated double bonds so the 2,3—double bond in 2§_would be expected to be stable under the reaction conditions.(21). 21 Hozc cozn H02C COZH Hozc C02H NaZEtOH > Na/EtOH > Liq NH3 Liq NH3 HOZC COZH Hozc cozn Hozc cozH 36 28 In initial attempts to carry out the above reduction, crude commercially available 1,4,5,8-naphthalenetetracarboxy— lic acid 2Z_was used as the starting material. In all cases, only starting material was recovered. At this point, we con— cluded that possibly the impurities in the crude tetra-acid were preventing the reduction from occurring. To circumvent this problem the tetra—acid was converted to the tetramethyl ester (22). The ester was a tractable compound that could easily be recrystallized from ethanol-water. The pure tetra— acid 21 could then be regenerated almost quantitatively by saponification in a mixture of methanol and water with potas- sium hydroxide. The Birch reduction was attempted on purified 1,4,5,8- naphthalenetetracarboxylic acid 21. The procedure involved the addition of tetra-acid'21 to a solution of six equiva- lents of sodium ethoxide in liquid ammonia at —78°C (this allowed conversion of tetra-acid 21_to the slightly soluble tetra-sodium salt). Ethanol (12 equivalents) was added as a source of protons (23) to react with the proposed intermedi- ate 21 in the sodium-liquid ammonia reduction. Eight equiva- lents of sodium were then added in small pieces over an eight 22 hour period to accomplish the reduction. Since the tetra- sodium salt was only slightly (if at all) soluble in the + reaction medium, the sodium was added over a long period of time so that equilibrium could be established between addi- tions. After all the sodium had reacted, the ammonia was evaporated with a stream of nitrogen. Acidification of the resultant tan solid produced a solid material which could be recrystallized from boiling water. Analysis showed that the composition of the product agreed with the empirical formula for 1,4,5,8—tetrahydronaphthalene—1,4,5,8-tetracarboxylic acid. A neutralization equivalent on this purified acid with standardized sodium hydroxide gave values of 76.7 and 77.3, both in close agreement to the calculated value of 77.1. The data indicated that the product had the desired em- pirical formula and there only remained the problem of estab- lishing its structure. The proof of the structure of the reduced acid was based mainly on the nmr spectrum. The de- sired product would be expected to show two absorptions in the ratio of 1:1. The observed spectrum showed a one proton singlet at 15.32 and a one proton singlet at T7.81 and was consistent with the eXpected spectrum for 1,4,5,8-tetra4 hydronaphthalenetetracarboxylic acid 22, The singlet at 23 15.32 was assigned to the olefinic protons, whereas the singlet at 17.81 would correspond to the allylic protons. -Evidently the dihedral angle between the allylic and olefinic protons is about 90° since coupling was not observed between the two different protons (24). The protons appearing at 77.81 underwent deuterium exchange with basic deuterium ox— ide at 100°, which would be expected for protons glpfig to a carboxylic acid group. From this data, the structure of the reduced acid was assumed to be correct. Later attempts to prepare the reduced acid 22 occasion- ally yielded only starting material. An investigation, therefore, was undertaken into possible modifications of the reduction step. Varying the concentration of all re— agents revealed that the volume of ammonia employed was criti- cal for the successful Birch reduction of the tetra-acid 21, A successful reduction required at least 100 ml of ammonia per gram of tetra-acid 21, Using the modified procedure, 40-50% yields of 1,4,5,8-tetrahydro-1,4,5,8-naphthalenetetra- carboxylic acid 22 could be obtained. Having successfully accomplished the Birch reduction of 1,4,5,8-naphthalenetetracarboxylic acid 21, the next obstacle involved the ring expansion of a six-membered ring to an eight-membered ring. The ring expansion was attempted on the model compound dimethyl 3,6-bis(hydroxymethyl)-1,4-cyclo— hexadiene-l,2-dicarboxy1ate di-pftoluenesulfonate 22_(Scheme III). The acetolysis of 22, however, did not yield dimethyl cycloBctatetraene-l,2-dicarboxylate 22, the compound 24 Scheme III CHZOH cozcn3 CHzOH CHZOTS + fil 1500> COZCHSE‘TSC1> C02CH3 \\ c pyr | COZCH3 COZCHa CHZOH COZCH3 CHZOH CHZOTS 2.2.3. 22 expected from a double ring expansion. Instead, two new products were isolated: 4,5,12,13-tetracarbomethoxy[2.2]- paracyclophane él_and what is believed to be 2,3—dicarbo- methoxy—4—methylbenzyl acetate 32, The latter compound was not characterized but its structure was consistent with its nmr spectrum, which consisted of a 2H AB quartet (J = 8Hz) centered at 12.46 (aromatic protons), a 2H singlet at 15.94 (methylene protons), a 3H singlet at 16.72 (methyl ester protons), a 3H singlet at 16.83 (methyl ester protons), a 3H singlet at 17.82 (methyl group) and a 3H singlet at 18.21 (acetate protons). The structure of the paracyclophane 22_was established from its uv and nmr spectra. The uv spectrum of 22_was very similar to that reported (25) for [2.2]paracyclophane. The nmr spectrum of 42 showed a singlet at 12.93 (4H) for the aromatic ring protons, a singlet at 16.15 (12H) for the carbomethoxy protons and an A2B2 multiplet centered at 16.80 (8H) for the benzylic protons. 25 E pHOAc NaH2P04 7 E Q CHonS E = COZCHs CHonS CHzoAC NaH PO ' 2 4 ‘E E CH3 41 42 The failure of 22 to undergo a double ring expansion can be explained by considering the mechanism that Dauben and Bertelli (19) have proposed for the ring expansion of 1,5-bis(hydroxymethyl)-1,4,5,8-tetrahydronaphthalene difpf toluenesulfonate 22_to dihydroheptalene 22. Because one of the double bonds in 22 is incorporated into an a,5-unsat~ urated diester system, this double bond would not partici— pate in the formation of a cyclopropyl carbonium ion. The lack of availability of one of the two double bonds to stabilize the incipient carbonium ion may slow down the rate of ring expansion enough so that elimination of pftoluene— sulfinic acid to yield dicarbomethoxy-pfxylylene 22 becomes + - CH20TS CH2 ‘OTS ‘HOAc O NaH2P04A m CH20TS CHonS E. H Ii H B B: —> H <— 2 b ’ 4’ HonS CHonS HOAc a NaH2P04 etc. 31 CHonS the major reaction. Elimination of two pftoluenesulfinic acid molecules would certainly be favored in this case due to the acidity of the 3 and 6 ring protons (y-protons in an a,5,A,e-unsaturated system). As has been previously pointed out for a similar reac- tion (26), the dimerization of gg_to 22_must involve a multi- step process since the Woodward-Hoffmann selection rules (27) do not allow a direct thermal 6 + 6 electrocyclic reaction. 27 CHZOTS ' " 02CH3 COZCH3 cozcn3 COZCH3 CHZOTs 44 312 — ‘ This method of preparation of paracyclophanes may have some synthetic utility since the starting bis(hydroxymethyl)— cyclohexadienes can be easily prepared by either of two methods: a) A Diels—Alder reaction between $2222, 2222232,4-hexadiene-1,6-diol and an appropriate acetylene. b) Birch reduction of an appropriately substituted aromatic para—dicarboxylic acid followed by hydride reduction to the corresponding diol. The failure of a six-membered ring to eXpand to an eight-membered ring coupled with Breslow's (11) publication concerning the preparation of benzo[c]octalene induced us to abandon the synthesis of octalene in favor of bicyclo- [6.2.0]deca-1,3,5,7,9-pentaene. Bicyglo[6.2.0]deca-1,3,5,7,9:pentaene Krebs (28) has reported that treatment of bromocyclo- Botatetraene 32 with an ethereal suspension of potassium Efbutoxide yields 1,2-dehydrocycloactatetraene.32 as a 28 transient intermediate. The chemical reactivity of 1,2-de- hydrocyclofictatetraene (Scheme IV) is very similar to that of benzyne gz_(Scheme V). The chemical similarity of 1,2- l- .. B r + jg-BuO- K+ Egg—EL» l 45 46 dehydrocycloéctatetraene and benzyne suggested the direct synthesis of a substituted bicyclo[6.2.0]deca-1,3,5,7,9- pentaene g2_by condensation of 1,2-dehydrocyclo6ctatetraene with a substituted acetylene. This type of reaction has been successfully accomplished with benzyne (29). ¢cscn I I l O 0 ¢ ¢ ¢ \ / 47 1,2—Dehydrocyclooctatetraene was prepared in the pres- ence of a large molar excess of either tolane, 3-hexyne, or l-methoxy-l-butyne. The only products isolated from these reactions were Efbutoxycyclobctatetraene 32 and naphtho- 2,3-cyclobctatetraene 22, These two products result from the reaction of 1,2-dehydrocyclodctatetraene with pfbutyl alcohol (to form 42) and with itself (to form 22). Potas- sium pramylate was also employed as a base to generate 1,2- dehydrocyclooctatetraene. Potassium Efamylate rather than 29 Scheme IV E-Buo t-BuOH IA 0') Phenyl aZide 30 Scheme V / \ ¢ ¢ _ 47 4 90 ¢ Phenyl ¢ -C=0 aZIde ¢ 00 El 31 potassium Efbutoxide would seem to be the preferred base OR" RNOH . j / r’R RCECR' 49. ._t I (______ | _. R —_-Bu \ R I 2;: R u =E-Amyl 48a: R=R'=¢ - fl ‘ 46 48b: R=R'=CH2CH3 _ 50 5.8.9: R=C2H5I R'30CH3 — for this reaction because it is soluble in tetrahydrofuran and can be prepared from stoichiometric amounts of potassium and Efamyl alcohol. Preparation in this manner allows the use of an alcohol-free base. Generation of 1,2-dehydro- cyclo6ctatetraene in the presence of 1-methoxy-1-butyne using potassium.§famylate as the base produced only Efamyl- oxycyclobctatetraene 22_and naphtho-2,3-cyc106ctatetraene 22. The previously reported (30) condensation of N-(l-cyclo- hexenyl)-pyrrolidine 22 with dimethyl acetylenedicarboxylate appeared to be adaptable to the synthesis of 9,10-dicarbo- methoxybicyclo[6.2.0]deca-1,3,5,7,9-pentaene‘22. Condensation 0 o E-C C-E E E = cozcn3 32 of N,N-diethylaminocycloactatetraene 22 with dimethyl acety- enedicarboxylate followed by deamination should yield the desired product 22, § N(C%F5)2 N(C2H5)2 g > ' 'HN(C2H5)2> ’/’ \\ I + ‘3 _ E \ / E 53 54 55 The first step in this synthesis involved the prepara- tion of the previously unknown N,N—diethylaminocycloacta+ tetraene 22, This synthesis was accomplished by the reac- tion of bromocyclooctatetraene with lithium diethylamide in Br , Etzo N(¢2H5)2 + L1N(C2H5)2 > 22_ refluxing diethyl ether. The structure of 22 was established by the nmr spectrum and hydrolysis in 30% acetic acid to cyclobcta-1,3,5-triene-7-one (isolated as the 2,4-dinitro- phenylhydrazone). The condensation of 22 with dimethyl acetylenedicarboxyl— ate yielded, instead of 22_or 22, dimethyl naphthalene—1,2— dicarboxylate 21, A logical proposal for this reaction is that the C1-C8 bond in 21 opens in a conrotatory process to produce the unstable intermediate 22, which cyclizes in a disrotatory process at C1 and C6 accompanied by elimination of diethyl amine to yield the observed product 21. 33 E (Et)2N ._ \‘ Pb(o.Ac)4 > HOZC’/ c02H 34 Scheme VI 0 cn3ozc F,fl\\ C02CH3 CH302C COZCH3 A 22_ B CH302C cnaozc \ / C02CH3 coch3 35 acid 22_has appeared in the literature (32) but all attempts to repeat the reported work failed. Saponification of the tetramethyl ester 22 in either refluxing aqueous potassium hydroxide or dilute hydrochloric acid produced only a glassy polymeric material. The tetramethyl ester 22_and tetracar- boxylic acid 22 contain hydrogens a to a keto group. The acidity of these hydrogens could have caused an aldol conden— sation to occur under the saponification conditions, there- fore producing polymeric material. To circumvent this prob- lem, the reaction sequence in Scheme VII was attempted. The bisoxime derivative 22 of the tetramethyl ester 22_ is easily prepared by reaction of 22 with hydroxylamine hydro— chloride in refluxing ethanol. The bisoxime derivative can be readily converted to the starting dione by oxime exchange in 40% formalin. Difficulty was encountered in the attempted saponification of bisoxime 22, The product eXpected from this reaction would be the bisoxime tetracarboxylic acid 22. The acid 22 could not be isolated from the saponification product because of the apparent high water solubility of the product. Structure determination under these conditions was not possible. Attempts to remove the oxime groups from the crude tetra-acid by the formalin method produced only one isolable compound in a poor yield. The structure of this compound was not determined but is is not the reported tetra-acid 22, At this point, the alternate reaction sequence (Scheme VI; Path B) to tetramethyl tricyclo[6.2.0.03:6]deca-1,7— diene-4,5,9,10-tetracarboxylate 22 was employed. The 36 Scheme VII E O /E + 1' NOH NH3OH c IKKF— pyr E\ ’E 62 E EtOH E/ \ E O NOH \ HCl / as O .KOH A A H H proposed v 0 NOH cozn H /U\ COzH 0 H020/ Y—wozn )1» o 6_3_ 37 attempted preparations of fig are outlined in Scheme VIII. The preparation of the bistosylhydrazone §§_proceeded smoothly. Attempted formation of the tetramethyl diene ester §Q_by reaction of §§_with sodium methoxide in diglyme yielded three products (separated by column chromatography). None of the isolated products showed the presence of olefinic protons in their nmr spectrum so they were not further in- vestigated. Tetramethyl 2,7—dihydroxytricyclo[6.2.O.03:6]decane- 4,5,9,10-tetracarboxylate §g_is obtained in high yield by reduction of the dione §§ with sodium borohydride in tetra- hydrofuran at room temperature. The diol is isolated as a viscous oil that presumably contains the three diastereomeric diols possible from the reduction of the dione. Conversion of the diol to the corresponding dichloride or dibromide was attempted with thionyl chloride, phosphorous tribromide, and phosphorous pentabromide. Hydrolysis of the reaction pro- ducts with dilute hydrochloric acid yielded only the starting diol. Direct dehydration of the diol with a catalytic amount of pyridine absorbed on Woehlm alumina was tried at 220°. This method has been successfully employed (33,34) for the dehydration of terpenes. However, only starting diol was isolated from this reaction. The failure of the latter two reactions to occur may be due to steric factors encountered in the diol. The six-membered ring of the diol §g_is be- lieved to be in the boat conformation (this assignment will be explained later) with the cyclobutane rings cis fused 38 Scheme VIII 0 NNHTs E E E TSNHNHZ i NaOCH 60 THE Diglyme '—— E E H+ E O s 58 NNHT “ g1 NaBH4 THF OH OTs pyr a REEFEE‘19; -§9 E E DMSO Naocn3 6/::[:Or A E CH30H EfTSCl or S:C12, PBr3, PBr5 E \‘ K'E E /, \'E E = C02CH3 X B: proposed 6 39 to the pseudo-equatorial positions. Molecular models strongly suggest that attack at C2 or C7 (carbons at which the hydroxyl groups are attached) will be very unlikely be- cause of shielding of these carbons by either the adjacent hydrogens or the backside of the molecule. The ditosylate §§_could be obtained by reaction of the diol §g_with pftoluenesulfonyl chloride in chloroform. Formation of the tricyclic decadiene §Q_from the ditosylate was attempted using both sodium methoxide in methanol and potassium Efbutoxide in dimethyl sulfoxide. Both reaction products yielded one major product (isolated by column chromatography), which did not contain olefinic protons in the nmr spectrum. The dehydration of the tricyclic decanediol §g_was also attempted with phosphoryl chloride in pyridine. This rea- gent has been found (35,36) to be particularly effective for the dehydration of alcohols. The product isolated from the reaction in 16% yield was tetramethyl 11-oxatetracyclo- [4,4,1,02r5.07:1°]undecane-3,4,8,9-tetracarboxylate 613 A substantial quantity (75%) of unaltered diol was also re— covered. The structure of §Z_was established from the ele- mental analysis and the nmr Spectrum, which showed a singlet for the 1 and 6 hydrogens at T5.65, a singlet for the 12 methyl hydrogens at 16.32 and a symmetrical A232 multiplet centered at 17.0 for the 2,3,4,5,7,8,9, and 10 hydrogens. The A232 multiplet consists of a multiplet for the 2,5,7, 4O OH E \ Poc13 67 —————-———> ’1... L—— pyro “' E Y \E g3, OH POC13 E = C02CH3 pyr- H O E H H / J H02C’ . H 'Hzo E /—V \E Hozc 'EOZH COzH .62. 68 and 10 hydrogens at 16.78 and a multiplet for the protons (3,4,8, and 9) alpha to the carbomethoxy groups at 17.22. The assignment for the A232 multiplet was substantiated by partially exchanging the 3,4,8, and 9 hydrogens of tetra- acid §§_with deuterium using basic deuterium oxide. The nmr spectrum of the partially deuterated tetra-acid‘gg showed a decrease in intensity for the low field A282 ab- sorption as compared to the high field portion of the Asz multiplet. The stereochemistry of §Z_was assigned by the use of the nmr spectrum, molecular models, and the method of forma- tion of tetramethyl tricyclo[6.2.0.03:5]decane-2,7-dione- 4,5,9,10-tetracarboxylate 58, The simplicity of the nmr spectrum of §z_indicated that the molecule was quite sym- metrical. The oxa-bridgehead hydrogens, 1 and 6, are not 41 split signifying that the dihedral angle between these hydro- gens and the 2,5,7, and 10 hydrogens is about 900 (24). To meet this requirement, the cyclobutane rings must be gi§_gng fused. If the cyclobutane rings were cis endo-fused, models show that the angle between H1 and H5, H7 would be ap- proximately 40°. In the latter case, a coupling constant of about 5 Hz would be expected. Steric crowding would also preclude the assignment of two cis endo-fused cyclobutane rings in 61. The appearance of the A2B2 multiplet in the nmr spectrum of §Z_narrows the structure of §Z_to two possible stereochemical configurations, 67a and 67b. 67a 67b Since the cyclohexane ring of 61 is in the boat conform- ation, the cyclohexadione ring of tetramethyl tricyclo- [6.2.0.03I6]decane-2,7-dione-4,5,9,10-tetracarboxylate‘58, the precursor to 61, should also be in the boat conformation assuming that the stereochemistry of the tricyclic dione §§_ has remained unchanged throughout the reaction sequence. The sodium borohydride reduction of the dione §§_should not have altered the stereochemistry of the cyclobutane rings because the low base strength of the borohydride anion usu— ally permits the reduction of a carbonyl function without 42 racemization of an adjacent center of asymmetry (37). Since no protonic solvents were present during the reduction, the enolate anion of the dione should not have been produced. Reaction of the reduction product, the tricyclic diol 64, with phosphoryl chloride in pyridine should also proceed without altering the stereochemistry about the cyclobutane rings. ~When the tricyclic dione 58 was refluxed with pyri- dine for 2 hours, only the unaltered dione 58 was recovered. This indicates that a basic reagent will not change the stereochemistry of the starting material, the tricyclic dione 58. Assuming then that the cyclohexadione ring of the tri- cyclic dione §§_is in the boat conformation and that the stereochemistry has remained unaltered throughout the re- action scheme, the stereochemistry of the cyclobutane rings would follow from the manner of preparation of the tricyclic dione 58, The tricyclic dione §§_was prepared by the photo- chemical dimerization of dimethyl trans, trans—1,4-penta- diene-3-one-1,5-dicarboxylate (32). -Dimerization in a head to head fashion would produce the tricyclic dione §§J which contains the cyclohexadione ring in a boat conformation. The carbomethoxy groups would then be gig to each other and trans to the cyclohexadione-cyclobutane ring junction. The stereochemistry of tetramethyl tricyclo[6.2.0.03'6]decane- 2,7-dione-4,5,9,10-tetracarboxylate §§_would therefore be as shown for ggg, The cyclobutane rings of tetramethyl-11+ oxatetracyclo[4.4.1.0215.07:1°]undecane-3,4,8,9-tetracarboxy1ate 43 O H H E hv a 13') /° x" ———-> I ' I E E 67 should also possess this stereochemistry and structure 6 a would appear to be correct. H O H E’ H E = C02CH3 E ‘E 613_ Corse, Finkle and Lunkin (38) have reported that the configuration of the tricyclic dione §§_consists of the cyclobutane rings being joined by trans fusions to the cyclo- hexadienone ring and the protons alpha to the carbomethoxy groups on the same four-membered ring must be trans to each other and trans to the adjacent protons on the junction car- bon atoms. Their assignment was based solely on the nmr Spectrum (taken in trifluoroacetic acid), which showed a pair of incompletely resolved multiplets centered at 15.30 and 15.80 and a single methoxyl resonance at 16.09. The configuration §§§_would also be in agreement with this nmr spectrum. The configuration §§b_of the tricyclic dione, as pro- posed by Corse and co-workers, contains two cyclobutane rings trans fused to the cyclohexadione ring. It would be expected 44 that under the influence of a mild base (pyridine) the junc- tion carbon atoms adjacent to the carbonyl groups would be E = C02CH3 epimerized to the more stable gg§_fused cyclobutane rings. This type of transformation has been reported by Corey and co-workers (57). Since only the unaltered tricyclic dione was recovered after refluxing in pyridine, the stereochemical assignment 58; rather than §§b_would seem to be preferred. Only one of the three possible diastereomeric diols, .62, can yield the observed tetracyclic ester 61, The other two diastereomeric diols probably do not react under the reaction conditions and are therefore recovered after hydrol- ysis. The mechanism of formation of §Z_from‘§2_most likely occurs through an SN2 displacement of the phosphate group by the alcohol's oxygen. . EXPERIMENTAL General Procedures All infrared spectra were obtained on a Perkin-Elmer Model 237B recording spectrophotometer, using sodium chlor— ide cells. The nmr spectra were obtained with a Varian A-60 or Jeolco C-60H spectrometer. Chemical shifts are reported as 1 values measured from either tetramethylsilane or, when D20 was used as the solvent, sodium 2,2-dimethyl-2-sila— pentane-5-sulfonate. All ultraviolet Spectra were measured with a Unicam Model SP-800 using 1 cm quartz cells. Mass Spectra were obtained by M. Petschel of this department with a consolidated Electrodynamics Corp. Mass Spectrometer Type 21-103C. Analysis by vpc was carried out on an Aerograph A-90—P3 instrument with a thermal conductivity detector us- ing a 15% carbowax 20M column, 5' x if. Melting points were determined on a Thomas Hoover melt- ing point apparatus and are uncorrected. All microanalyses were performed by the Spang Micro- analytical Laboratory, Ann Arbor, Michigan. The molecular models used were Framework Molecular Models by Prentice-Hall, Inc., Englewood Cliffs, New Jersey. Diels-Alder Reaction of 22_with Maleic Anhydride 1,4-Dipheny1-2-styryl-1,3-butadiene, 1 9 (0.00325 mole), and an excess (3 g) of maleic anhydride were dissolved in 46 5 ml of anhydrous toluene. The mixture was refluxed for 48 hr under an atmosphere of nitrogen. The toluene was evap- orated under reduced pressure and the unreacted maleic an- hydride was removed by sublimation at 900 under reduced pressure. The resu1ting brown oil solidified on cooling. The solid material was refluxed with chloroform and the in- soluble white crystalline material was removed by filtration. The DielsSAlder adduct was dissolved in acetone and freed of insoluble material by filtration. Evaporation of the acetone produced 1.109 g (68%) of the dianhydride g§_as a white crystalline solid: m; 305-3080; uv(CH3CN) 230 (e 1650), 253 (910), 259 (870), 265 (695), and 310mu (520). »Ap§;, Calcd for C32H2406: C, 76.18; H, 4.79. Found: C, 76.26; H, 4.90. 1,4,5,8-Tetrahydro-1,4,5,8-naphthalenetetracarboxylic Acid 28 A 1-1., three-necked, round-bottomed flask was equipped with a mechanical stirrer, dry ice condensor and dropping funnel and cooled in a dry ice-acetone bath. Ammonia (700 ml) was distilled into the flask and 23 ml (0.39 mole) of abso- lute ethanol‘ was added dropwise with stirring. Sodium, 3 9 (0.1305 mole), was then added in 0.1 9 pieces. After the sodium had dissolved (30 min), 7 9 (0.0232 mole) of 1,4,5,8-naphthalenetetracarboxylic acid was added and the suspension was stirred for 30 min. Sodium, 4.5 g (0.191 mole), was added in small pieces (0.2 g) over an 8 hr period maintaining the temperature at -80°. After the complete 47 addition and reaction of the sodium, the ammonia was allowed to evaporate. The resulting tan solid was dissolved in 150 ml of water, filtered free of insoluble material, and acidified to pH 1 with 6N hydrochloric acid. The solution was evaporated to dryness under reduced pressure and the tan residue was diluted with 100 ml of cold water. The crude tetra-acid was removed by filtration. Two recrystalliza- tions from boiling water produced 2.8 g (40%) of 1,4,5,8- tetrahydro-1,4,5,8-naphthalenetetracarboxylic acid as color- less plates: mp 228-2290; ir (Nujol) 3300-2500 (S, broad), 1700 (s, broad), 1330 (s), 1273 (m), 1198 (s), 920 (w), and 773 (w) cm-l; nmr (NaOD in D20) 15.32 (s, 4), and 17.81 (S, 4). éflél: Calcd for C14H1208: C, 54.55; H, 3.92. Found: C, 54.53: H, 4.05. Dimethyl 3,6-Bis(hydroxymethyl)-1,4-cyclohexadiene-1,2-di— carboxylate §§_ 2,4-Hexadiene-1,6-diol (39), 0.8 g (0.007 mole), was combined with 2.5 g (0.014 mole) of dimethyl acetylenedicar- boxylate and the resulting semisolid was heated at 150° for 24 hr under an atmosphere of nitrogen. The unreacted di- methyl acetylenedicarboxylate was removed under reduced pressure at 90°. The orange residue was dissolved in a mini- mal amount of boiling ethyl acetate and cooled to 0° pro- ducing 0.338 g (19%) of the diol 38 as a yellow crystalline solid: mp 99.5-102°; ir (Nujol) 3300 (m, broad), 1702 (s), 48 1628 (w), 1270 (s), 1068 (m), and 823 (m) cm-l: nmr (6661,) 14.16 (m, 2), 6.62 (m, 2), and 6.21 (m 12). Dimethyl 3,6—Bis(hydroxymethyl)-1,4-cyclohexadiene-1,2—di- carboxylate Diepftoluenesulfonate‘gg prToluenesulfonylchloride, 2.7 g (0.014 mole), recrystal- lized from 90-120° petroleum ether (mp 64-650), was dissolved in 10 ml of dry pyridine (distilled from potassium hydroxide). This solution was added dropwise to 0.83 9 (0.0065 mole) of dimethyl 3,6-bis(hydroxymethyl)-1,4-cyclohexadiene-1,2-dicar— boxylate §§_dissolved in 5 ml of dry pyridine at a tempera- ture of -6 to 0°. The mixture was kept overnight at 0°. The pyridine was removed under reduced pressure at room temperature and the dark red residue was dissolved in 50 ml of chloroform. The chloroform solution was washed with 50 ml of ice cold 10% hydrochloric acid and 50 ml of 5% sodium bicarbonate. The chloroform extract was separated, dried (M9804), and concentrated producing a dark red oil: ir (Neat) 2945 (m), 1710 (s), 1595 (m), and 1365 (s) cm—l. All attempts to crystallize the ditosylate failed so the crude ditosylate was used forthe following ring expansion. Attempted.Ring Expansion of 3 — 1:— Acetic acid (100 ml) was combined with acetic anhydride (20 ml) and refluxed overnight. The acetic acid used for the following solvolysis was obtained from this mixture by distillation through a 12 inch Vigreaux column (bp 117-118°). 49 Crude dimethyl 3,6-bis(hydroxymethyl)—1,4-cyclohexadiene— 1,2-dicarboxylate di—pftoluenesulfonate'32, 1.65 g (0.00293 mole), was added to a solution of sodium dihydrogenphOSphate monohydrate, 1.8 g (0.013 mole), in dry acetic acid (20 ml) in a 100—ml, three-necked, round-bottomed flask, fitted with a mechanical stirrer, reflux condensor, nitrogen inlet and outlet and thermometer. The mixture, blanketed with dry nitrogen, was heated to 90° with an oil bath for 22 hr. The dark brown solution was transferred to a 400 ml beaker. The glassware was rinsed with water and ether and combined with the original solution. Ether (150 ml) was added and the two-phase system was cooled to 60 in an ice bath. Next, a solution prepared by dissolving 24 g of potassium hydroxide in 400 ml water was added dropwise with stirring at 6-120 until the pH of the aqueous layer was 14. The ethereal ex- tract was separated, washed with water, dried (MgS04) and concentrated producing a yellow semisolid. The crude semi— solid was dissolved in hot methanol and upon cooling white crystals formed. Two recrystallizations from methanol pro- duced 0.134 g (21%) of 4,5,12,13-tetracarbomethoxy-2,2- paracyclophane 41 as colorless needles: mp 203-204.5°; uv (ethanol) 215 (6 2,260), 270 (170) and 302 mp (170); ir (Nujol) 1717 (s), 1188 (s), 1164 (m), 1128 (m), 1104 (m), 1004 (w), 873 (w), and 800 (w) cm-l; nmr (c0013) T2.93 (s, 4), 6.15 (s, 12), and 6.80 (m, 8). Anal, Calcd for C24H2408: C, 65.45; H, 5.49. Found: C. 65.46; H, 5.40. 50 Distillation of the crude methanol solution produced what is believed to be 2,3-dicarbomethoxy—4-methylbenzyl acetate 42: bp 70-740 (3 mm); ir (CHC13) 2946 (m), 1747 (s), 1712 (s), 1600 (w), 1375 (s), 1221 (s), 1177 (m), 1050 (m), 1009 (m), and 922 (m) cm‘l; nmr (cnc13) 12.29 (d, 1, g = 8 Hz), 2.71 (d, 1, g = 8 Hz), 5.94 (s, 2), 6.72 (s, 3), 6.83 (s, 3), 7.82 (s, 3), and 8.21 (s, 3). 9,10-Diethylbicyclo[6.2.0]deca-1,3,5,7,9-pentaene 48b (At- tempted) Bromocycloactatetraene (48), 5 g (0.0273 mole), in 10 ml ether was added dropwise over a period of 1.5 hr to a stirred suspension of 75 ml (54.2 g; 0.66 mole) of 3-hexyne, 75 ml of ether and 4 g of potassium Efbutoxide under nitro— gen. The dark brown suspension was stirred at room tempera- ture for 20 hr. The suspension was hydrolyzed with 100 ml of deoxygenated water. The ethereal layer was separated, dried (Na2804), and concentrated to a volume of 10 ml by evaporation under reduced pressure at room temperature. The residue was column chromatographed under an atomosphere of nitrogen on silicic acid (100 g) eluting with deoxygenated carbon tetrachloride—benzene (15% benzene). Evaporation of the eluents with a stream of nitrogen produced 0.729 g (15%) of tfbutoxycyclobctatetraene: bp 38-390 (0.55 mm) [lit. (28) bp 37-380 (0.05 mm)] and 0.458 g (7%) of naphtho-2,3- cycloactatetraene: mp 109-1120 [lit.(28) mp 113-1140]. 51 9-Ethy1-10-methoxybicyclo[6.2.0]deca-1,3,5,7,9-pentaene 48c (Attempted); Preparation of EfAmyloxycyclobctatetraene‘51 A solution of potassium Efamylate in tetrahydrofuran was prepared by adding 2 g (0.05 mole) of potassium to a solution of 2.64 g (0.03 mole) of tramyl alcohol (distilled from sodium) in 50 ml of dry (distilled from calcium hydride) tetrahydrofuran and refluxing with stirring under nitrogen for 90 hr. The solution was cooled and the unreacted potas- sium was removed by fitration under nitrogen. The potassium 'E-amylate solution was found to be 1.075 molar by titration of a 2-ml aliquot with 0.0983N hydrochloric acid. A solution of potassium Efamylate in tetrahydrofuran, 27 ml (0.029 mole), was added dropwise over a 30 min period to a solution containing 10 ml of dry tetrahydrofuran, 20.4 g (0.25 mole) of l-methoxy-l-butyne (40), and 5 9 (0.0273 mole) of bromocyclooctatetraene with stirring under nitrogen at room temperature. The solution immediately turned brown and warmed slightly. Stirring was continued for 42 hr. The suspension was hydrolyzed with 50 ml of deoxygenated dis- tilled water and 150 ml of ether was added. The ethereal layer was separated, washed with two 50-ml portions of satur- ated sodium chloride and concentrated under reduced pressure at room temperature. The residue was chromatographed under nitrogen on silicic acid (100 g) eluting with carbon tetra- chloride-methylene chloride (23% methylene chloride). *Evap— oration of the eluents with a stream of nitrogen produced 52 0.339 g (6%) of naphtho-2,3-cyclooctatetraene: mp 110-1120 [lit. (28) mp 113-1140] and 1.27 g (21%) of E—amyloxycyclo- Botatetraene: bp 46-480 (0.15 mm); ir (Neat) 3000 (s), 2979 (s), 2940 (m), 2880 (w), 1633 (s), 1460 (m), 1398 (w), 1369 (m), 1224 (m), 1139 (s), 927 (w), 833 (m), 802 (m), and 790 (s) cm—l; nmr (c0c13) 43.72-4.70 (m, 7), 8.39 (q, 2, g_= 7 Hz), 8.77 (s, 6), and 9.12 (t, 3, g.= 7H2); mass spectrum (56 ev) gig (rel intensity) 190 (5), 175 (4), 161 (5), 120 (95), 91 (100), 78 (95), and 70 (90). Stirring 0.40 9 (0.0021 mole) of tramyloxycyclooctatetra- ene in a solution consisting of 23 ml of 95% ethanol, 4 ml of water, and 3 ml of concentrated sulfuric acid for 2 hr under nitrogen at room temperature produced cycloocta-1,3,5- triene-7—one. The cycloBctatrienone was identified as its 2,4-dinitrophenylhydrazone: mp 155.5-157o [lit. (41) mp 158— 159°]. N,N-Diethylaminocycloéctatetraene‘53 Into a dry 250—ml Erlenmeyer flask equipped with a con- densor, dropping funnel, and magnetic stirring bar was placed 4.11 g (0.057 mole) of diethyl amine in 50 ml of dry ether under nitrogen. To this stirred solution was added 34 ml of gfbutyllithium in hexane (1.6M; 0.0546 mole) dropwise over a period of 10 min. The resulting opaque solution of lithium diethylamide was cooled to -4° in an ice-95% ethanol bath and 5 9 (0.0273 mole) of bromocyclodctatetraene in 15 ml of dry ether was added dropwise overa.period of 15 min. The dark 53 red solution was stirred at -4° for 10 min, allowed to warm to room temperature over a period of 20 min, and then re- fluxed for 92 hr. The resulting suspension was hydrolyzed by cooling to -10° in an ice-95% ethanol bath and adding 50 ml of deoxygenated water dropwise over a period of 15 min. The ethereal layer was separated, washed twice with 100-ml portions of cold deoxygenated water and dried (M9304) under nitrogen. Evaporation of the solvent at room temperature afforded 4.2 g (88%) of N,N-diethylaminocyclodctatetraene as a deep red liquid: ir (Neat) 2960 (s), 2935 (m), 2855 (w), 1612 (s), 1379 (m), 1250 (m), 1126 (m), and 672 (m) cm-l; nmr (Neat) 13.9-4.6 (m, 6) 5.52 (d, l, g.= 4 Hz), 6.99 (q. 4, l: 7 Hz), and 9.01 (t, 6, q: 7 Hz). Stirring N,N-diethylaminocyclofictatetraene with 30% acetic acid for 30 min at room temperature and 20 min at 35° furnished cycloocta-1,3,5-triene-7-one. The cycloocta- trienone was identified as its 2,4-dinitrophenylhydrazone derivative: mp 156-157 [lit. (41) mp 158-159°]. Dimethyl l-N,N-Diethylaminobicyclo[6.2.0]deca-2,4,6,9-tetra— ene-9,10-dicarboxylate 54.(Attempted): Preparation of Di- methyl 1,2-Naphthalenedicarboxylate ._5__ Dimethyl acetylenedicarboxylate, 1.20 g (0.00845 mole), in 10 ml of ether was added dropwise over a period of 15 min to a solution of 1.48 g (0.00845 mole) of N,N-diethylamino- cyclodctatetraene in 40 m1 of ether under nitrogen with stirring at 0°. The solution was allowed to warm to room 54 temperature and was then refluxed for 5 hr. The resulting solution was washed twice with water, dried (M9304), and concentrated. The red oily residue was chromatographed on 50 g of silicic acid eluting with chloroform. Evaporation of the eluent and recrystallization from methanol afforded 0.441 g (21%) of dimethyl 1,2-naphthalenedicarboxylate as colorless needles: mp 82-82.5° [lit. (42) mp 82-83°]: ir (c0c13) 2948 (s), 2880 (m), 1725 (s), 1435 (s), 1293 (s), 1270 (s), 1248 (s), 1141 (s), 1043 (m), and 852 (s) cm—17 nmr (CDC13) 11.61-2.30 (m, 6), 5.85 (s, 3), and 5.97 (s, 3). Tetramethyl Trigyclo[6.2.0.03I6ldecane-2,7-dione-4,5,9,10— tetracarboxylate §§_ Dimethyl trans-trans-l,4-Pentadiene-3-one-1,5-dicar- boxylate (32), 4.1 g (0.021 mole), was slurried in methylene chloride (25 ml) and deposited on the walls of a 4-1. beaker. The methylene chloride was allowed to evaporate and the ketone was then irradiated for 8 hr with a Hanovia Type L, 450 watt ultraviolet lamp using a vycor filter. The photolyzed material was washed with hot chloroform and re- moved by filtration yielding 2.8 g (69%) of the tricyclic ester 58: mp 242-243° [lit. (32) mp 242—243°]. Tetramethyl TricycloL6.2.0.03rGldecane-Z,7—dione-2,7-bisoxime- 4,5,9,10-tetracarboxylate‘62 Tetramethyl tricyclo[6.2.0.03v6]decane—2,7-dione—4,5,9,10- tetracarboxylate §§J 1 g (0.00252 mole), was combined with 55 10 ml of ethanol, 5 ml of pyridine and 0.8 g (0.012 mole) of hydroxylamine hydrochloride and refluxed with stirring for 2 hr. The reaction mixture was cooled and concentrated to a yellow oil. The oil was crystallized by addition of 20 ml of water. Two recrystallizations from ethanol—water af— forded 0.763 g (71%) of the bisoxime 62 as colorless plates: mp 223~224°; ir (Nujol) 3400 (s, broad), 1730 (s), 1638 (w), 1275 (m), 1200 (m), 947 (m), and 830 (w) cm-l; nmr (DMSO—da) 15.80—6.78 (m). final. Calcd for C18H22010: C, 50.70: H, 5.20. Found: C, 50.90; H, 5.25. Bisoxime 623 0.22 g (0.516 mmol), was combined with 2 ml of 40% formalin and 0.4 ml of 2N hydrochloric acid and heated on a steam bath for 20 min. The bisoxime immediately dissolved and a white solid started to precipitate out of the solution. The resulting suspension was cooled, diluted with water, and filtered producing 0.2039 g (100%) of the tricyclic dione 58; mp 242-243° [lit. (32) mp 242-243°]. Bisoxime 62, 0.2 g (0.5 mmol), was heated with 1 ml of acetic anhydride on a steam bath for 5 min. The solution was cooled and diluted with 2 ml of water. The resulting white solid was removed by filtration. Two recrystalliza— tions from ethanol produced 0.215 g (85%) of the bisoxime diacetate 623: mp 178.5-179°; ir (Nujol) 1774 (s), 1727 (s), 1640 (w), 1279 (s), 1200 (s, broad), and 887 (m) cm'l: nmr (DMSO-da) T5.70-6.70 (m, 20), and 8.10 (s, 6). 56 Anal. Calcd for C22H26012: C, 51.767 H: 5.13. Found: C, 51.80; H, 5.06. Tetramethyl Tricyclo[6.2.0.03:6]decane-2,7-dione-2,7—bistosyl— hydrazone—4,5,9,10-tetracarboxy1ate 63' Tetramethyl tricyclo[6.2.0.03I6]decane—2,7—dione-4,5,9,10— tetracarboxylate 58” 0.76 g (1.92 mmol), was combined with 0.94 g (5 mmol) of 4-toluenesulfonylhydrazine in 50 ml of tetrahydrofuran containing 1 ml of concentrated hydrochloric acid and stirred for 24 hr under nitrogen at room temperature. The tetrahydrofuran solution was concentrated and the result- ing yellow oil was crystallized by adding 10 ml of ethanol. Two recrystallizations from chloroform-ethanol afforded - 1.207 g (86%) of the bistosylhydrazone §§_as long, colorless needles: mp 208-209.5°; ir (CHC13) 3130 (w, broad), 3010 (w, broad), 2950 (w), 1725 (s), 1600 (w), 1440 (m), 1350 (m), 1273 (m), 1168 (s), 1065 (m), and 960 (w) cm—l; nmr (00013) T1.82 (d, 4, g_= 8 Hz), 2.42 (d, 4, g_- 8 Hz), 5.74-7.0 (m, 20), and 7.61 (s, 6). £231, Calcd for C32H36N401282: C, 52.45; H, 4.95. Found: C, 52.30; H, 5.02. Tetramethyl 2,7-Dihydroxytricycloj6.2.0.03v5]decane-4,5,9,10- tetracarboxylate 64 Tetramethyl tricyclo[6.2.0.03:6]decane-2,7-dione- 4,5,9,10-tetracarboxylate, 3 g (7.56 mmol), was dissolved in 350 ml of dry tetrahydrofuran containing 0.3 g (8 mmol) 57 of sodium borohydride and stirred at room temperature for 48 hr. The yellow solution was diluted with 100 ml of an- hydrous methanol and allowed to stand for 12 hr at room temperature with occasional stirring. The solution was con- centrated, 100 ml of water was added and the solution, con- taining a small quantity of undissolved oil, was acidified to pH 1 with 6N hydrochloric acid. chloroform (100 ml) was added and the resulting two-phase system was allowed to stand for 12 hr with occasional agitation. The chloroform layer was separated and the aqueous extract was saturated with sodium chloride and extracted with three additional 50-ml portions of chloroform. The chloroform extracts were combined, dried (MgSO4), and concentrated affording 2.177 g (72%) of the crude diol as a pale yellow, viscous liquid. Crystallization could be accomplished from chloroform-hexane. Five crystallizations produced 0.6366 g (21%) of the diol 64_ as small colorless needles: mp 239.5-242°; ir (Nujol) 3450 (s, broad), 1725 (s, broad), 1440 (s), 1255 (s), 1200 (m), 1173 (m), 1100 (w), 1056 (w), 948 (m), and 891 (w) cm‘l; nmr (DMSO-ds) 3.93 (m, 2) and 5.91-6.96 (m, 22). Anal. Calcd for C13H24010: C, 54.00: H, 6.04. Found: C, 54.09; H, 5.79. Tetramethyl 2,7-Dibromotrigyclo[6.2.0.03I6]decane-4,5,9,10- tetracarboxylate §§_(Attempted) Bromine, 4.64 g (0.028 mole), dissolved in 6 ml of methylene chloride was slowly added to a cooled (ice bath) 58 solution of 7.59 g (0.029 mole) of phOSphorous tribromide in 8 ml of methylene chloride with stirring. A yellow precipi- tate of phosphorous pentabromide formed. 0161.64, 5.1 g (0.0128 mole), dissolved in 10 ml of methylene chloride was added over a period of 1 hr to the well stirred and cooled mixture. An additional 5 ml of methylene chloride was added and the suspension was stirred for 3 hr. Ice water (25 ml) was added; the methylene chloride layer was separated, washed with water, 10% sodium bicarbonate and saturated sodium chloride. Concentration of the methylene chloride extract afforded only 4.86 g (95%) of the starting diol (by ir). Tetramethyl 2,7-Dihydroxytricyclo[6.2.0.03Isldecane-4,5,9,10— tetracarboxylate Di-pytoluenesulfonate‘gg 0101,64, 14.7 g (0.037 mole), was dissolved in 100 ml of pyridine at 0° and treated with a solution of 21 g (0.111 mole) of pftoluenesulfonylchloride dissolved in 100 ml of chloroform. After standing at 0° for 8 days, the solution was poured into 450 ml of ice cold 2N hydrochloric acid. An additional 100 ml of chloroform was added, the organic phase ‘was separated, washed with 300 ml of 1N hydrochloric acid (0°), 300 ml of water, and 300 ml of saturated sodium chlor— ide solution. Concentration of the chloroform extract under reduced pressure at room temperature afforded 17.2 g (65%) of the crude ditosylate §§_as a pale yellow liquid. All at— tempts to crystallize the ditosylate 66 were unsuccessful so the crude ditosylate was used directly in all reactions: ir 59 (CHC13) 3130 (w), 2995 (m), 2950 (s), 1735 (s, broad), 1600 (m), 1437 (s), 1375 (s), 1280 (s), 1169 (s), 1080 (m), 1008 -1 (m), and 915 (m) cm ; nmr (CDC13) T2.19 (d, 4, g_= 8 Hz), 2.72 (d, 4, g.= 8 Hz), 5.72—7.35 (m, 22) and 7.57 (s, 6). Tetramethyl 11-0xatetracycloL4.4.1.02:5.07r1°1undecane— 3,4,8,9-tetracarboxylate‘61 Phosphoryl chloride, 5 g (0.33 mole), was added drop- wise with stirring to a solution of 3.305 g (8.3 mmol) of diol §4_dissolved in 24 ml of pyridine at 0°. The resulting suspension was stirred at room temperature for 21 hr and at 100° for 1.5 hr. The reaction mixture was cooled in an ice bath and poured over 100 g of ice. The aqueous solution was extracted with three 70-ml portions of chloroform. The chloroform extracts were combined, washed with four 100-ml portions of 2N hydrochloric acid and two 100-ml portions of water. The chloroform extract was dried (M9304) and con- centrated to a brown oil. Addition of 20 ml of methanol caused the tetracyclic ester 61 to crystallize. Two recrys— tallizations from hot methanol produced 0.446 g (16%) of the tetracyclic ester 61 as colorless needles: mp 209.5- 210.5°; ir (Nujol) 1740 (s), 1350 (m), 1300 (m), 1232 (m), 1177 (m), 1154 (m), 1056 (w), 1031 (w), 860 (w), and 826 (w) cm-l; nmr (CDC13) 15.65 (s, 2), 6.32 (s, 12), 6.78 (m, 4), and 7.22 (m, 4). 523;. Calcd for C18H2209: C, 56.54; H, 5.80. Found: C, 56.26; H, 5.65. 60 11-0xatetragycloL4.4.1.02I5.07:1°]undecane-3,4,8,9-tetracar- boxylic Acid Monohydrate 68 Tetracyclic ester 61, 0.030 9 (0.0784 mmol), was com- bined with 15 drops of dioxane, 15 drops of water and 5 drops of 6N hydrochloric acid and heated on a steam bath for 15 hr. The resulting suspension was diluted with water and the tetra—acid was removed by filtration. Two recrystalliza- tions from boiling water followed by drying under reduced pressure at 1000 afforded 0.025 g (92%) of the tetra-acid monohydrate §§_as a colorless crystalline solid: mp 340°; ir (Nujol) 3475 (m, broad), 3300-2500 (m), 1710 (s), 1635 (w), 1350 (w), 1282 (m), 1245 (m), 950 (w), 922 (w), and 742 (w) cm-l; nmr (NaOD in D20) 75.64 (s, 2), 7.09 (m, 4), and 7.40 (m, 4). ‘ Anal. Calcd for C14H16010: C, 48.84; H, 4.68. Found: C, 48.77; H, 4.50. PART II THE REARRANGEMENTS OF MONOSUBSTITUTED CYCLOGCTATETRAENES 61 HISTORICAL AND INTRODUCTION Cycloéctatetraene and its derivatives undergo both acid catalyzed and thermal rearrangements. The acid catalyzed rearrangements have attracted by far the most attention in the literature (43,44,45,46). A mech- anism for this type of rearrangement has been postulated (46). For example, Willstatter and Heidelberger (47) treated cyclo- octatetraene Zl_with hydrogen bromide, Reppe and his coworkers (43) identified the product as a—bromoethylbenzene and Ganellin and Pettit (46) proposed the following mechanism for the re- arrangement: H .fl§£_; H _____> +_1§ CH2 1.1. 22_ H Br H H H Ii HBr __. .— H._ + ‘3' CH2 The rearrangement presumably involves addition of a proton to cyclodctatetraene yielding the cyclooctatrienyl carbonium ion zg_which undergoes a WagnerdMeerwein rearrangement to the carbonium ion 13. Further rearrangement of the valence tautomer of Z§_and the addition of hydrogen bromide yields the observed product. 62 63 Thermal rearrangements of cyclooctatetraenes, on the other hand, have received very little attention to date. In the only reported paper dealing with this subject, COpe and Burg (48) noted that during the preparation of bromo- and chlorocyclooctatetraene, considerable amounts of the cor- responding B-halostyrenes were isolated. Further studies indicated that chlorocyclobctatetraene.15 readily rearranged to ging-chlorostyrene 19 at 200° and bromocyclooctatetraene .11 to trans-B-bromostyrene Z§_at 90°. 0 H H r x 0 Cl 900 H Br — 19 75: X = Cl 30 min — (D H 11; X = Br A mechanism for this type of rearrangement has so far not been proposed. RESULTS AND DISCUSSION Thermal rearrangements of cyclobctatetraenes have com- manded only minor research interest in the past; therefore, this type of rearrangement was studied in order to establish a possible mechanism. As mentioned previously, COpe and Burg (48) reported. that chloro- and bromocycloactatetraene thermally rearranged to ging—chloro- and trans—B—bromostyrene, respectively. Since the work of Cope and Burg on the thermal rearrangement of chloro- and bromocycloactatetraene had been reported be— fore the advent of nuclear magnetic resonance spectroscopy, the products from the thermal rearrangements were reinvesti- gated to determine if the assigned structures were indeed correct. The styrene isolated from the thermal rearrange- ment of bromocycloéctatetraene was shown by its nmr spectrum to be the reported trans-S—bromostyrene 18“ but the product from the thermal rearrangement of chlorocycloBctatetraene was proved by nmr and ir (51) to be trans-B-chlorostyrene §4_instead of the reported gig isomer 16. Both trans—5— halostyrenes showed a typical trans vicinal olefinic coupling (50) of 13.7 Hz. The thermal rearrangement of both chloro- and bromocyclo- 6ctatetraene appeared to be stereospecific. Analysis of the halostyrene from either thermal rearrangement by vapor phase chromatography (vpc) or nmr showed the presence of only the 64 65 trans isomer. There appeared to be no detectable amount of .Eii isomer present. The rate of the rearrangement of chlorocyclobctatetraene and presumably bromocycloooctatetraene noticeably increased in polar solvents. For example, heating chlorocyclooctatetra- ene for 36 hours at 122° in a sealed tube with 2-methyl-2- No reaction /’ 122o c1 36 hr butene produced only unaltered starting material. In re- fluxing acetonitrile (bp 81°), however, chlorocyclobctatetra- ene was completely converted to Egggg—B-chlorostyrene in 24 hours. This suggests that the rearrangement is proceeding through a polar intermediate. Rearrangement of chlorocyclobctatetraene in refluxing methanol produced besides Eggggea-chlorostyrene, two addi- tional products, phenylacetaldehyde dimethyl acetal 81_and EggggrB-methoxystyrene 88, Egggg—B—Chlorostyrene would not convert to either the acetal 81 or the methoxystyrene 88 under the reaction conditions, whereas Eggggffi-methoxystyrene 88_is readily transformed (46) into phenylacetaldehyde di- methyl acetal 81, Phenylacetaldehyde dimethyl acetal 81 66 CH30H , 1 HCl I Cl CH OH H OCH3 ‘_—i7_’ §3, + ¢CH2CH(OCH3)2+ )F:q< 24 hr 9 H 55% 40-45% 0_5% 87 §§. CH3 OH * * HCl and trans-B-methoxystyrene 88 were identified by comparison of their infrared spectra with their reported spectra (52). The rearrangement of chlorocycloéctatetraene was per— formed in methanol-g_to determine the extent of deuterium incorporation. The trans-B-chlorostyrene isolated from the reaction product was found to be free of deuterium (by nmr), but the phenylacetaldehyde dimethyl acetal contained two deuterium atoms, both on the benzylic carbon. The location of the deuteriums at the benzylic position followed from the nmr spectrum, which showed the signal of the methine proton as a singlet at 75.56 rather than the triplet characteristic of the undeuterated compound and the disappearance of the doublet for the benzylic protons. The deuterium content of trans—S-methoxystyrene was unfortunately not determined be- cause the reaction product contained only a small amount of this compound. Since trans-fi-methoxystyrene converts to phenyacetaldehyde dimethyl acetal under the reaction condi- tions, if the methoxystyrene is labeled, the deuterium must be present at the a position. 67 Rearrangement of chlorocyclobctatetraene in refluxing methanol containing an excess of sodium methoxide altered the products and product distribution. Thus, methoxycyclobcta- tetraene 88_was isolated along with EggggrB-chlorostyrene and Egggng-methoxystyrene. -Methoxycyclobctatetraene was found to be stable under the reaction conditions. Phenyl- acetaldehyde dimethyl acetal would not be expected to be pres- ent in the reaction mixture because the addition of methanol to Egggg-B-methoxystyrene is catalyzed only by acid. Meth— oxycyclobctatetraene 88 was identified by its mass spectrum, its rapid hydrolysis in dilute alcoholic sulfuric acid to cyclobctatrienone (characterized as its 2,4-DNP) and the nmr spectrum, which showed a multiplet for six cyclobctatetraene hydrogens at 73.93-4.41, a multiplet for the cyclobctatetra- ene hydrogen a to the methoxy group at 75.05-5.15, and a three proton singlet for the methoxy group at 76.42. The rearrangement of chlorocyclooctatetraene in refluxing meth— anol containing a fifteen mole excess of lithium bromide dramatically illustrates the effect of a strong nucleophile on the reaction. The products obtained from the reaction were Egggng-chlorostyrene (14%), phenylacetaldehyde dimethyl acetal (3%). and trans—B-bromostyrene (83%). The large yield Cl H Cl H Br LiBr >._—._< >_—_< .‘ > + ¢CH2 CH (OCHa ) + 24 hr 0 H . 2 0 H 14% 3% 83% of trans-B—bromostyrene indicates that the nucleophilic 68 bromide ion is successfully competing with chloride ion and methanol in the trapping of the rearrangement intermediate. A possible mechanism for the rearrangement of chloro- cycloéctatetraene in polar solvent, which is consistent with all the acquired data, is proposed in Scheme IX for methanol solution, in Scheme X for methanolic sodium methoxide and in Scheme XI for methanolic lithium bromide. This mechanism proposes initial ionization in the polar methanol solutions of 1-chlorobicyclo[4.2.0]octa-2,4,7-tri- ene 88, one of four possible valence tautomers of chloro- cyclobctatetraene, to the bicyclic octatrienyl carbonium ion 91. Cl - The stereospecific formation of 852-8-substituted bicy- cylo[4.2.0]octa-2,4,6-triene‘81, the precursor to ££22§‘5‘ substituted styrene 88, is rationalized by assuming that the approaching nucleophile only attacks from the 2E2 side of 8; because of the greater steric accessibility of the 252 face and the steric difficulties involved in approaching the gagg_face. A molecular model of the bicyclic octatrienyl carbonium ion 8; supports this assumption by revealing that the Cz‘pi-orbital effectively prevents attack by a nucleo- phile from the endo face by shielding the endo face of the 69 Scheme IX ChlorocycloBctatetraene and Methanol-d oCH3 CH CD 3+ . -- 92 +D+ CH3OD -CH30D -D i OCH3 ‘ 22H con. ring opening OCH3 OCH fi+€§=< ring 3+CH3 on opening’ OCH3 -CH3OD OCH3 70 Scheme X ChlorocycloBctatetraene and Methanolic Sodium Methoxide OCH3 OCH other I 4...— 3 __._s valence ’ ‘ tautomers 94 “- A CH30 con. CE ring 1 opening H 9.2. CH30- V con. H OCH // . 3 — OCH3 ring ; __ \\ E’ opening H 93 71 Scheme XI ChlorocycloBctatetraene and Methanolic Lithium Bromide Br OCH3 ,H Cl 9.5. 94 '__ 84 _H+ +H+ —- con . opening / -- _ —. 42'! Br “Cl \\ ' 3‘ L .1 H ‘ con. OCH3 Ci .r (jaw ._..8—: open1 4’ 3 96 H H con. 88 ._._ ring opening +H+ HH” -H+ OCH3 CH30H OCHa 2.8. .1» 72 / | x- con 0 H X ”r: [ +X- / a X ring _ \\ 48 ‘x I g H 2 H l X- 97 98 vacant‘pirorbital on C8. Once the egng-substituted bicyclo- [;.2.0]octa-2,4,6-triene §Z_has been formed, a thermally al- lowed conrotatory opening (53) of the cyclobutene ring would stereospecifically yield the trans-B-substituted styrene fig. The absence of methoxycyclooctatetraene §2 as a product in Scheme IX is also consistent with the proposed mechanism. Methoxycyclooctatetraene §2_would be in equilibrium with its valence tautomer, 1—methoxybicyclo[4.2.0]octa-2,4,7—triene 22, which would be formed by the addition of methanoljd_to the bicyclic octatrienyl carbonium ion 2;. However, since the addition of methanol-d_to the carbonium ion 2; is a reversible reaction in an acidic medium, any product formed by an irreversible reaction would deplete the reaction mix- ture of methoxycyclooctatetraene. Since both trans—B-chloro- styrene fig and trans-B-methoxystyrene are formed by a con- rotatory ring opening, in this case an essentially irrevers- ible reaction, the presence of methoxycyclobctatetraene‘gg inany detectable amount would not be expected. Subjecting methoxycyclooctatetraene to the reaction conditions (reflux- ing methanol containing a catalytic amount of dry hydrochloric acid) established the reversibility of the reaction sequence gg. i3_gg_;3_g;. The products obtained from this reaction 73 were phenylacetaldehyde dimethyl acetal (97%), EEEEEfB‘ methoxystyrene (3%), and a trace of Eggné—B-chlorostyrene. The presence of unreacted methoxycyclooctatetraene was not detected. The formation of methoxycyclooctatetraene fig” however, would be expected in basic solution (Scheme X), as was ob— served, since the addition of methoxide ion to the bicyclic octatrienyl carbonium ion 2l_would be an irreversible re- action. An alternative mechanism for the thermal rearrangement of n23£_chlorocyclooctatetraene and n§2£_bromocyclooctatetra- ene to the corresponding Eggngfa-halostyrenes is possible. This mechanism involves the stereospecific formation of egg: 8—halobicyclo[4.2.0]octa-2,4,6-triene‘2z_by a 1,3-sigma— tropic suprafacial shift (49) of the halogen atom in the valence tautomer 29, The suprafacial sigmatropic shift is // 1,3-sigma- . con. H X I tropic > “' X ring > ___ \ shift I opening ‘5, H H 90 97 98 allowed because the migrating halogen atom possesses an available p-orbital that can interact with the pi_system in the transition state (49). Conrotatory ring opening of 21 would yield the trans-B-halostyrene fig, Differentiation between the two possible mechanisms (ionization gs, 1,3—sigmatropic shift) for the neat rear- rangements was not attempted. The rearrangement of chloro- cyclooctatetraene in refluxing methanolic solutions most likely does not proceed by the latter mechanism because a 1,3-sigmatropic shift should not be favored by increasing the solvent polarity. An attempt was made to trap 1-chlorobicyclo[4.2.0]octa- 2,4,7-triene 92 by preparing a Dials-Alder adduct of chloro- cyclooctatetraene. The isolated Dials-Alder adduct was found to be derived from valence tautomer 22.rather than 29, C1 C1 C1 99 l ,__. d .— l _ NC CN >=< 3f (TCNE) Imus NC CN c1 / (CN)2 cl / 12}. (CN)2 (CN)2 (CN)2 100 The tub conformation of chlorocyclooctatetraene contains no 1,3-diene system providing the approximately planar 75 configuration essential for a Dials-Alder reaction. On the other hand, the valence tautomers 22, 22 of chlorocycloocta- tetraene do offer a planar diene system, particularly since incorporation of the cyclobutene ring causes further flat- tening of the 1,3-cyclohexadiene system. Neither 1,3-cyclo- octadiene (54) nor 1,3,5-cyclooctatriene reacted with maleic anhydride to form a DielSeAlder adduct, but bicyclo[4.2.0]- octa-2,4-diene $22 (55) was able to do so. These facts sug- gest that a moderate equilibrium concentration of the valence /’ \\ 1 2 tautomers rather than chlorocyclooctatetraene may enter into the Diels-Alder reaction. The failure of valence tautomer 22_to react with tetra- cyanoethylene can be rationalized by assuming that the chlor- ine atom of 22 repels the cyano groups of tetracyanoethylene through electron-electron repulsion. Therefore, the essen— tial pifoverlap required for a Diels-Alder reaction would not be present. A molecular model of the transition state expected from the reaction of 22 and tetracyanoethylene sub- stantiates this assumption. The structure of 22, on the other hand, does not permit this type of electron-electron repulsion and formation of the Dials-Alder transition state can easily occur. 76 The rearrangement takes a different pathway when di- ethylaminocyclooctatetraene‘122 or Efbutoxycyclooctatetra- ene $21 (28) are heated. In both of these cases the a—sub- stituted styrenes, 12§_and 122, are produced cleanly. Chemi- cal evidence for the structures of'122_and lg§_was obtained by their rapid hydrolysis as an enamine and vinyl ether, respectively, by dilute acid forming acetophenone (isolated as the semicarbazide or 2,4-DNP). 1000 (C2H5)2N H mi? — lflé R lgéfi R=N(C2H5)2 ¢ H .121: RénguO 1450 ‘E-Buo H 54 hr > j>===< ,lQQ ¢ H A possible mechanism for the rearrangement of diethyl- amino- and Efbutoxycyclooctatetraene is proposed in Scheme XII. This mechanism involves the formation of 7-substituted bicyclo[4.2.0]octa-2,4,7-triene $22” a valence tautomer of the substituted cyclooctatetraene. The valence tautomer could then rearrange to 7-substituted bicyclo[4.2.0]octa- 2,4,6-triene 122 followed by a conrotatory ring Opening to the a-substituted styrene. An alternate pathway available 'would be the formation of 1-substituted bicyclo[4.2.0]octa- 2,4,7—triene $21, another valence tautomer of the substituted cyclooctatetraene. This valence tautomer, however, would not ionize to the carbonium ion 2l_because of the strong 103. R 104: R / WN(C2H5 _t_-Bu0 H Con. H WN(C2H5 E-Buo opening 77 Scheme XII \\ 108 /’ .__ 109 78 basicity of the leaving group. Preparation of the DielSeAlder adduct $19 of Efbutoxy- cyclobctatetraene established the intermediacy of 773- butoxybicyclo[4.2.0]octa-2,437-trieneIlgg. The structural assignment of llg follows from the nmr spectrum which shows three olefinic protons and four bridgehead protons. The Diels-Alder reaction between tfbutoxycyclooctatetraene and dimethyl acetylenedicarboxylate forms the basis for the as- signment of the Efbutoxy group in llg to the cyclobutene double bond. The latter Diels-Alder reaction yielded not t-Buo / E-Buo , I _TCNE < A \ ¢H A’ / (“)2 (CN)2 108 110 the expected product, but instead its thermal degradation product, dimethyl phthalate. If the tfbutoxy group had been situated on the cyclohexadiene ring of 11; rather than the cyclobutene ring, the product of the reaction would have been a tfbutoxy substituted phthalate rather than the ob- served dimethyl phthalate. This fact coupled with the nmr spectrum's integral ratio suggested that the tfbutoxy group in 110 was located on the double bond of the cyclobutene ring. _t_-Bu0 " t-BuO 79 11 111 fl —-i E = cozcn3 EXPERIMENTAL Thermal Rearrangement of Chlorocyclooctatetraene‘15 to trans- fi-Chlorostyrene Chlorocyclodctatetraene (48), 3 9 (0.0217 mole), was combined with 10 mg of hydroquinone and refluxed at 2000 in a nitrogen atmosphere for 2 hr. Distillation of the brown reaction product yielded 0.873 g (29%) of Ergggffi-chloro- styrene: bp 61-620 (4.5 mm); ir (neat) (51) 1609 (m), 1498 (w), 1448 (m), 1248 (m), 1071 (w), 939 (s), 859 (m), 814 (m), 741 (s), and 695 (s) cm-l; nmr (neat) 12.7-3.1 (m, 5), 3.32 (d, 1, g.= 13.7 Hz), 3.67 (d, 1, g_= 13.7 Hz). Rearrangement of Chlorocyclooctatetraene in Refluxing Methanol Chlorocyclooctatetraene, 0.5 g (2.89 mmol), was com- bined with 2.5 m1 of methanol and refluxed for 24 hr under nitrogen. The reaction mixture was concentrated to a volume of about 0.5 ml. The product was analyzed by vpc, using a 15% Carbowax 20M column (5 ft x i-in) at 100°. There were three components, the ratio from faster to slower was 55:45:5. The materials were separated by vpc. The products were trans-B—chlorostyrene (55%) (retention time, 26 min), identified by the infrared spectrum (51), phenylacetalde- hyde dimethyl acetal (45%) (retention time, 34 min), identi- fied by the infrared spectrum and the nmr spectrum (CDC13) 80 81 12.67-2.95 (m, 5), 5.52 (t, 1, g_= 6 Hz), 6.84 (s, 6) and 7.16 (d, 2, g’3 6 Hz), and transz—methoxystyrene (5%) (re- tention time, 44 min), identified by its infrared spectrum (52) and the nmr spectrum (CDC13) T2.55-2.8 (m, 5), 2.89 (d, 1, g_= 13.7 Hz), 4.16 (d, 1, g_= 13.7 Hz) and 6.33 (s, 3). Rearrangement of chlorocyclooctatetraene in refluxing methanol-deroduced trans-fi-chlorostyrene,‘Egang-B-methoxy- styrene (position of deuterium is uncertain), and 2,2-di- deutero-2-phenylacetaldehyde dimethyl acetal (separated by vpc): nmr (CDC13) 12.70-3.04 (m, 5), 5.56 (s, 1), and 6.82 (s, 6). Rearrangement of Chlorocyclooctatetraene in Methanolic Sodium Methoxide Chlorocyclooctatetraene, 0.320 g (1.85 mmol), was com- bined with sodium methoxide (from 0.12 g (0.52 mmol) of sodium) dissolved in 2.5 ml of methanol and refluxed for 20.5 hr under nitrogen. The reaction product was diluted with 20 ml of water and extracted with two 10-m1 portions of ether. The ether extracts were combined, dried (MgSO4), and concentrated under reduced pressure. Three compounds were isolated by the use of vpc (15% Carbowax 20M, 5 ft x i-in, 106°, retention times, 16, 24, and 37 min). The slower moving fraction (37 min) and the next slowest moving fraction (24 min) were shown to be trgngfa—methoxystyrene (30%) and trans-B-chlorostyrene (31%), respectively, by their infrared spectra (52,51). The fastest moving fraction 82 (16 min) was identified as methoxycyclooctatetraene: ir (CC14) 3005 (s), 2953 (m), 2904 (w), 2840 (w), 1662 (m), 1641 (s), 1465 (m), 1452 (m), 1443 (m), 1403 (m), 1379 (m), 1231 (s), 1202 (s), 1163 (s), 1020 (s), 835 (m), 800 (s), and 756 (s) cm-l; nmr (cnc13) T3.93-4.41 (m, 6), 5.05-5.15 (m, 1), and 6.42 (s, 3); mass spectrum (62.5 eV) m/e 134. Stirring methoxycycloéctatetraene with a solution con- taining 2 ml of ethanol, 1 ml of water, and 0.5 ml of con- centrated sulfuric acid yielded cycloocta-1,3,5-triene-7-one. The cyclooctatrienone was identified as its 2,4-dinitro- phenylhydrazone derivative: mp 153-1560 [lit. (41) mp 158- 159°]. Rearrangement of Chlorocyclooctatetraene in Methanolic Lithium Bromide Chlorocyclooctatetraene, 0.250 g (1.445 mmol), was com; bined with 2 g (20 mmol) of lithium bromide and 5 ml of methanol and refluxed for 24 hr under nitrogen. The pale yellow solution was diluted with 40 ml of water and ex- tracted with 25 ml of ether. The ethereal layer was separ- ated, dried (MgSO4), and concentrated by evaporation under reduced pressure. The reaction product upon vpc, after a small solvent peak, showed three peaks (15% Carbowax 20M, 93°, retention times, 22, 30, and 46 min). The two faster moving peaks were trgngffi-chlorostyrene (14%) and trans—fi- methoxystyrene (3%), respectively (by infrared). The slow- est moving peak was identified as trans-B-bromostyrene (83%) 83 by comparison of its infrared spectrum with that of an authentic sample (48). Tetracyanoethylene Diels—Alder Adduct of ChlorocycloBcta- tetraene Tetracyanoethylene, 0.93 g (7.26 mmol), was combined with 1 g (7.26 mmol) of chlorocyclo6ctatetraene dissolved in 3 ml of benzene and allowed to stand at room temperature for 129 hr under nitrogen. The insoluble material (TCNE) was removed by filtration and washed with 3 ml of benzene. The benzene solution was concentrated to a brown semisolid under reduced pressure. Two recrystallizations from hot methanol yielded 0.057 g (3%) of the Diels-Alder adduct 19; as colorless needles: mp 235-235.5°; ir (KBr) 2950 (w), 2920 (w), 1537 (s), 1270 (m), 1228 (m), 1136 (m), 1090 (m), 920 (m), 798 (m), 745 (s), and 699 (w) cm-l: nmr (DMF) 73.58- 3.71 (m, 2), 4.03-4.21 (m, 1), and 5.81-6.05 (m, 4). fl. Calcd for C14H7C1N4: C, 63.05: H, 2.65. Found: C, 62.75; H, 2.72. Thermal Rearrangement of N,N-Diethylaminocycloactatetraene 103 N,N-DiethylaminocycloBctatetraene (2.0 g) was heated under an atmosphere of nitrogen at 1000 for 0.5 hr. Dis- tillation of the reaction product produced 1.63 g (81%) of a-N,N—diethylaminostyrene 195: bp 47-480 (0.15 mm); nmr (Neat) T2.45-2.90 (m, 5), 5.75 (s, 1), 5.90 (s, 1), 7.06 (q. 4. g,= 7 Hz), and 9.03 (t, 6, g_= 7 Hz). 84 d—N,N-Diethylaminostyrene was hydrolyzed to aceto— phenone by stirring with 30% acetic acid at 90° for 15 min under nitrogen. The acetophenone was isolated as the semi- carbazide derivative: mp 196-198° [lit. (56) mp 198°]. Thermal Rearrangement of EfButoxycycloBctatetraene 104 4“ EfButoxycycloSctatetraene (28), 1.3 g (7.4 mmol), was heated at 144-147° for 45 hr under an atmosphere of' nitrogen. The nmr spectrum of the crude material indicated that approximately 25% of the tfbutoxycyclooctatetraene had rearranged to d-Efbutoxystyrene 126; nmr (Neat) 12.69- 2.81 (m, 5), 6.43 (s, 1), 6.54 (s, 1), and 8.70 (s). The Efbutoxy ether was hydrolyzed by dissolving the crude material in 10 ml of 95% ethanol and then adding 20 ml of 15% sulfuric acid solution. This solution was stir- red under nitrogen for 2 hr at room temperature. Water (45 ml) was added and the solution was neutralized with sodium carbonate. The aqueous solution was extracted with four 10-ml portions of ether. The ether extracts were com- bined, dried (MgSO4), and concentrated under reduced pres- sure. The reaction product was analyzed by vpc, using a 15% Carbowax 20M column (5 ft x i-in) at 121° and there were two components. The ratio of the faster to the slower mov- ing material was approximately 3:1. The material appearing at 14 min was identified as cycloBcta-1,3,5—triene-7—one by its infrared spectrum (41). The minor, slower moving mater- ial (30 min) was acetophenone. The infrared spectrum of 85 the collected acetophenone was identical in all respects to the infrared spectrum of commercially available acetophenone. 1A 2,4-dinitrophenylhydrazone derivative was made of the col- lected acetophenone: mp 247.5-248.5° [lit. (56) mp 250°]. Tetracyanoethylene Diels-Alder Adduct of EfButoxycycloécta- tetraene 104 EfButoxycyclooctatetraene, 1 g (5.7 mmol), was com— bined with 0.726 g (5.7 mmol) of tetracyanoethylene in 3 ml of benzene and refluxed for 19 hr under nitrogen. The ben- zene was removed under reduced pressure producing a dark brown oil, which slowly solidified on standing. Two recrys— tallizations from boiling methanol produced 1.12 g (65%) of the Diels-Alder adduct 110 as a beige crystalline solid: mp 159.5-1610; ir (KBr) 2935 (m), 2940 (w), 1627 (m), 1590 (s), 1367 (m), 1300 (m), 1243 (m), 1200 (m), 349 (m), 723 (m), and 699 (w) cm—l; nmr (cnc13) T3.86-4.52 (m, 3), 6.2-6.54 (m, 4), and 8.53 (s, 9). Anal. Calcd for C18H16N40: C, 71.03; H, 5.30. Found: C, 70.81; H, 5.34. The Diels—Alder Reaction of EfButoxycyclo6ctatetraene with Dimethyl Acetylenedicarboxylate EfButoxycyclooctatetraene, 2.0 g (11.3 mmol), was com- bined with 1.612 g (11.3 mmol) of dimethyl acetylenedicar- boxylate in 4 ml of dry toluene under an atmosphere of nitro- gen and refluxed for 23 hr. The dark brown solution was 86 vacuum distilled through a semimicro column. The distil- lation separated a forerun of 0.32 g of Efbutoxycycloficta- tetraene, bp 47-52° (0.2 mm), from 1.2 g (49%) of dimethyl phthalate: bp 90—94.5° (0.2 mm). The infrared and nmr spectra of the isolated dimethyl phthalate were identical in all respects to those of commercially available dimethyl phthalate. 87 “"30"“ 10.0 ".0 no (1:) m an: .91." -¢-') l ’ moons ' moons ".0 I10 88 .00 -¢--; m 'o-' L- 89 90 . 4.0 moons ‘ . . 3.0 ““00““ 10.0 no no 0 H . ”no 400 150 3m 2:: m Isa no to QC” on t” m 340 :00 an no In a ”S m m m m m u n m T I l I l l’Ile V l l l I I In ' .' LL! _‘ U '1 [Ll '_'J-',l L1,: lu' ' l 71*! ' U '1L":l.?_'_"lr!_'.'_L' '.'_'.IJTI!I.!LJIL_- J..- .._l 1-,_LWL._‘ f Figure "13 ' ' , rJ 11 .L f .. 1 .. H r . 4. . . . . A. ._... . .1 . LAIIIIIIllIlJllllllllt‘lllllljlllIll]llllllJlLLlIIIILIIIAllllLIllllfllJLllll I i I I O 7 I Li T_ [14171411177171 . 50C 1‘?) 2 50 x’) 250 200 ‘50 IO) 50 £75 W no no :00 Id) In I20 00 m I“ M m too 500 200 I” ”S 1 new: m an :00 :50 I 4 «no no no no :00 3° W 5" ‘3" nu: mo :59 a 5 § “ '°° 50° ' :00 zoo m ea: V'|[IU‘1_IIVIIIn“![JiLyjAIIIJIflLLvJ‘IJJllLLLIIIILLLI‘LIIVIJ[LFIIVIfo‘ILLIlIlJlrI—l—rvj—r‘UIIAIIJIILIl 5 , s Figure 15 : ‘ -———- —~.__1.‘ ‘: r l..___ _. 2U) ) Jllllllllll[AIIlllln‘lLLLllllllllIllllAAIllll‘lllllllLALllllllll‘llllllJ 3 4 D 0 I I V LAJJIJAIIIIIII I 92 A am no :50 300 250 no no too so cos a on so :00 an In no 00 005 no eon :00 ”do too to: ”II _,II_II I ,I 1 TI_ I n TI I I I TI. I I I __I_ I .I _I LIL I I I I III _I II I I I I .I_I_I.!.I-I-I I.I-I_.I.I-[_C!_I.l.‘ll_'.lLL!-E‘._LI-'_.II.I.I,I_'.I I1 ..- . ' I E Figure 16 . “-H- o , 1- .1 !.._,fi_._,., I1 ' - 4) ‘ :r—“' H ““ ‘ I M 030 350 300 250 “I no a so M m can no 6m :00 I_l[I III‘IWTTIT_ILLWT[ Trtlvtr [firrlrfT111rtilii F..- .. , .._._. ,_._-_ -__ . .. ___ , -._. .- _ ”Eigure 17 t” ”'1 "_ 3 '3' ‘7' 'i; y--._. ( I o n ‘n. M -a.‘ a. v . o—o .Ao—ow— o... I I 0 - 1":— 'N A. ..~ .. .k 4.444;] ‘ LLL114111119111L11111114L1'11111'111i-l'ill-I'Iwilii-lLILLLL‘I L11111111111‘I111’i11l':1:11.11111'1'114] ‘ o 93 Q some 400 :50 x0 :50 to Iso 100 50 ('58 duo an no 100 :00 240 no no to 005 m In 700 coo 000 200 loo cos rr ' ‘ T' ' r1] ' ‘ '1 rr ' '177 ' ' I VTI ' I 3 tr.“ .T.f.1 I. 11.1.Vilrwrfn'ilf'lfj Tij: LE}; L!_:_'..:.r.'.'-'.3_'.1_ LPK'TELY 1.1,]. -1 l h r '. I 3 o » - 7 D ? 0 " 1 . L I. I ' ! ‘ - ) 1 L A ' A A. L‘III’QTT‘I‘J 1.11 [11.11 11111111111 1111.1 :1 111-1141111111-{1’1le lllLLJlLLLL‘LllgLLllIll-1‘11 111141 3 3 C 3 O 3 ’ U . '- ‘<.. ‘7' ---» ‘. 0 ‘--_ ”.-—-—.— -, .D—u-O"! o--- .. / .."er~..- - T , I, . O LLLLIIIILJILIL'11.111L.J111111111111glLLllllllLllLLllJLJlLLLLlILLlLllLllllllllJLillLLL' 4 E // km . J] ‘ I! J); 0 t TV Vvarv—v vv—vr' v vrvv v 1. 1 , . “m. .H.1 d .. .... . o. . . .. I. 1 . 1 __ h. H M . .1 ., 1 . 1. 1 .1 1. 1 .— —.—-- —-——-— _— ~ 0 .— “of- O l v L I o +-- O 1.111- I 1 , 1r.._..4 I 11 1M ‘L-n-l 0. +07." -.11I,11 ?. 9 - )_. .4: i 3 1 1 fl 1. .1 1 1 1 1 211 H 1 , I10 «.11 1 1.1.51... 1 . _ 4 _ . .1 .. vr .1 m H 1 v. w u . .1 . _ . . 1 - .. . ,. 1 w .toO.1f..4l.£ .Ool. vv,o..9\v ..¢ Jot11101t§nib|vlolz1¢iletlololiwl. .I-.." v...?:'?' .3910? IIIII (01,11 , . . 1 1 1 . . 1 1.1.13.1...1 1,1 1. 1 1 , ”4|; 1.. . . . . ._ u w m l. 21' 1 . _. H v . . 1 H . H w n 1 . 1 _ .. .. r . . 1 fr H ..... ..u [N.... 4 fr .,6... fl. » 1 ilfl-§'\‘o-n A 1 — 1 v . m a v 9110;?! . w m _ . . . .1 9.th . 1 1 1 1 4... .1: . 1 1 1 1 .92.! .1 .111 1 1 . 1 _ n 1 1 M 1. 1 . .1 1 v , . . v 1 .. 1 - s _ vvl-nla’b' H0'0'11IIJ’Io-"o’t\9‘1"0-|lino0..."1H.:- II v' .‘l ' 0.4!.‘II'0159MI0. 6 d .. . , 1 . . . _ 1 1 V 1 w . . M f 1 r 1 u . 1 _ 1 o.) .. . . q 1 1 . 1 1 . . 1 w 1 1 1 . 1 . 1 . . a h . 1 1 1 . 1 10 cf” 1 . . . -h.. 5 u o .. . . . ... .. ................ . fivv 1 w . 1 . , .1” . v 1 I! n H H . £01.53. 1011.».th MJI . fl I. 1 ..... 1 1.1 I 1 11 ..... 1 1 . 1 1 1 1 . 1 , 1 1 . _ wv - . .1— . 1 o r , _ 1 1 1 . 1 . 1 1 , 1 m. 1 . 1 .1 .“TT .9 Val T W o 91H 14 L .\..o ?1ot“c.TV 0., O. I'vw L . - 1». v 1 _ . 1 l- _ . 1 ., . . v 1 . fr .. . ’ . . . fi 1 v u r. . 1. .. . 1 J .1 v 1 . v 1 . _ v 1 .1 wt n w . o .. J . w o v 1 . . v 1 1 TV . e a ....... 1 . .. .WN . r . o 11. 1 «1... 1 1 1 1 . . 1 1.. .14 «1.1. 1 ,1 v 1 1 , . v 1 1 1 . 1 . ,. fl 1 1 1 1 1 1 H 1. . . iv . m 1 . . u , 1.L rfi _ 1 1 . 1 1 , . v 9.? . . ..ol- v.1. «I’ll: 130130. 9+ Ti-‘ll.alhl!lh.‘llnl w .. . 3 . I ..1r.|)l(-,rl.71” 193+ p). f... . . v 1 , 1 . . 2 1 3 1 1 1 . U .1. 1 . 1 y 2 a? 2 _ 1 1 1 . 1.... ”T" W... 1 v .. .. h... . . W . v 1 1 .. . 1 1 ... . 1 . 1 m 1 e 1 , 1 1 um .9 1 . “ . . Iv o r H - .W 0 0| luv)... (1|: L r u 1 . v 1 1 1. v . 1 m. . m. , . 1 1 ., m1 1 . . . . , . v . 1 i _ U fi 1 1 1 1 . 1 . 1 1 _ . I. r . i 1 . 1 1,. , l. r .1 . . . . F . . F _ m . f . . . . H 1 . 1 1 _ 1. . . . 1 1 e_ 161 . 1 1 . . D I - . H . D I . b » > . . "fl ’ b > Jr . I. ' ' 1| 1 ) .J ' ' I 3 1' ‘ 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. LITERATURE CITED J. A. Elvidge and L. M. Jackman, J. Chem. Soc., 859 (1961). E. Huckel, Z. Electrochim., 23, 782 827 (1937). E. Huckel, Z. Physik., 19” 204 (1931). M. E. Vol'pin, Russ. Chem. Rev., 29, 129 (1960). ~A. Streitweiser, "Molecular Orbital Theory for Organic Chemists," J. Wiley and Sons, New York, 1961, p. 298. M. Dewar and G. Gleicher, J. Amer. Chem. Soc., 81, 685 (1965). N. L. Allinger and C. Gilardeau, Tetrahedron, 1569 (1967). V. Boekelheide, ibid., ll” 121 (1961). D. P. Craig, J. Chem. Soc., 3175 (1951). C. A. Coulson and G. Rushbrooke, Proc. Cambridge Phil. Soc., 36, 193 (1940). R. Breslow, W.-Horspool, H. Sugiyama, and W. Vitale, J. Amer. Chem. Soc., 88” 3677 (1966). G. Wittig, H. Eggers, and P. Duffner, Ann. Chem., 619, 10 (1958). J.-E1ix, M. Sargent, and F. Sondheimer, J. Amer. Chem. Soc., 89, 180 (1967). J. Elix, M. Sargent, and F. Sondheimer, ibid., 89, 5081 (1967). S. Masamune, C. Chen, K. Hojo, and R. Seidner, ibid., §2, 4804 (1967). '“'”' G. Schroder and H. Rottele, Chimia, 22” 146 (1968). F. Bohlman, Chem. Ber., §§, 2191 (1956). 4R. Criegee, D. Seebach, R. Winter, B. Borretzen, and H. Brune, ibid., gg, 2339 (1965). .H. J. Dauben and D. Bertelli, J. Amer. Chem. Soc., 83, 4657 (1961). 95 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30. 31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 96 G. W. Brown and F. Sondheimer, ibid., 89, 7116 (1967). -W. Huckel and R. Schwen, Chem. Ber., 89, 481 (1956). W. Treibs and G. Mobius, Ann. Chem., 619, 122 (1958). A. J. Birch, J. Chem. Soc., 1551 (1950). 4J. R. Dyer, "Applications of Absorption Spectroscopy of Organic Compounds," PrenticeéHall, Inc., N.J., 1965, p. 117. H. Winberg, F. Fawcett, W. Mochel, and C. Theobald, J. Amer. Chem. Soc., g2” 1428 (1960). D. T. Longone and M. Reetz, Chem. Commun., 46 (1967). R. Hoffmann and R. B. Woodward, J. Amer. Chem. Soc., 81, 2046 (1965). A. Krebs, Angew. Chem., internat. Ed., 3, 952 (1965). ~M. Stiles, U. Burckhardt, and A. Haag, J. Org. Chem., 21, 4715 (1962). .K. Brannock, R. Burpitt, W. Doodlett, and J. Thweatt, ibid., 25, 1464 (1963). H. Stobbe and E. Farber, Chem. Ber., 58, 1548 (1925). H. Midorikawa, Bull. Chem. Soc. Japan, 26, 302 (1953). E. von Rudloff, Can. J. Chem., 39, 1860 (1961). -E. J. Corey and A. Hortmann, J. Amer. Chem. Soc., 81, 5736 (1965 E. LeGoff, ibid., gg, 1505 (1962). -W. Allen and S. Bernstein, ibid., 11, 1028 (1955). 6N. Gaylord, "Reduction with Complex Metal Hydrides,“ Wiley-Interscience, New York, 1956. J. Corse, B. Finkle, and R. Lunkin, Tetrahedron Lett., 1 (1961). E. Bates, E. Jones, and M. Whiting, J. Chem. Soc., 1854 (1954). J. R. Nooi and J. Arens, Rec. Trav. Chim., 18, 284 (1959). 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 97 A. C. Cope and B. D. Tiffany, J. Amer. Chem. Soc., 18, 4158 (1951). M. Karasch and R. Dannley, J. Org. Chem., 183 406 (1945). ‘W. Reppe, O. Schlichting, K..Klager, and T. Toepel, Ann. Chem., 560, 1 (1948). C. R. Ganellin and R. Pettit, J. Amer. Chem. Soc., 18, 1767 (1957). A. C. Cope, N. Nelson, and D. Smith, ibid., 18, 1100 (1954). c. R. Ganellin and R. Pettit, J. Chem. Soc., 576 (1958). R. Willstatter and M. Heidelberger, Chem. Ber., 18, 517 (1913). ~A. C. Cope and M. Burg, J. Amer. Chem. Soc., 11, 170 (1952). R. B. Woodward and R. Hoffmann, ibid., _1, 2511 (1965). A. A. Bothner-By, "Advances in Magnetic Resonance,“ Vol. 1, Academic Press, New York, 1965, p. 207. M. Schlosser, Chem. Ber., 81, 3219 (1964). S. Miller, J. Amer. Chem. Soc., 18, 6091 (1956). R. B. Woodward and R. Hoffmann, ibid., 81, 395 (1965). K. Ziegler and H. Wilms, Ann. Chem., 567, 26 (1950). A. C. Cope, A. Haven, F. Ramp, and E. Trumbull, J. Amer. Chem. Soc., 13, 4867 (1952). R. L. Shriner, R. C. Fuson, and D. Y. Curtin, "The Systematic Identification of Organic Compounds,“ 4th ed, John Wiley and Sons, Inc., New York, 1960, p. 317. 'E. J. Corey, R. Mitra, and H. Uda, J. Amer. Chem. Soc., fig, 485 (1964). HIHIHIHIH '1... 93 03082 2 NI)” 1) II)I))I))