Xerox University Microfilms , Ann Arbor, Michigan 48106

75-3012

BIRNBERG, Gary Harold, 1948- I SYNTHESIS AND REACTIONS OF SELECTED MONOSUBSTITUTED DIBENZOSEMIBULLVALENES, SEMIBULLVALENES, AND BULLVALENES. j

The Ohio State University, Ph.D., 1974 | Chemistry, organic ( ! | 1

THIS DISSERTATION HAS BEEN MICROFILMED EXACTLY AS RECEIVED. SYNTHESIS AND REACTIONS OF SELECTED MONOSUBSTITUTED

DIBENZOSEMIBULLVALENES, SEMIBULLVALENES, AND BULLVALENES

DISSERTATION

Presented in Partial Fulfillment of the Requirements for the Degree Doctor of Philosophy in the Graduate School of The Ohio State University

Gary Harold Bimberg, B. Sc. *****

The Ohio State University

197^

Reading Committee: Approved By

Dr. Leo A. Paquette Dr. Robert J. Ouellette Dr. John A. Secrist, III

Department of Chemistry DEDICATION

To my parents and siblings

ii 1 'Would you tell me, please, which way I ought to go from here?1'

''That depends a good deal on where you want to get to,'' said the Cat.

''I don't much care where said Alice.

''Then it doesn't matter which way you go,'' said the Cat.

''... so long as I get somewhere,1' Alice added as an explanation.

''Oh, you're sure to do that,’1 said the Cat, ''if you only walk long enough. ''

Lewis Carroll

iii ACKNOWLEDGMENTS

The author wishes to express a profound indebtedness to Professor

Leo A. Paquette for the opportunity to work both with and for him and to the personnel of Armand’s Army ttho made my stay in Cowtown enjoyable.

Special thanks to Mr. J. M. Geckle without whose dedication to someone else's research this thesis would not have been possible at this time.

iv VITA

March 6, 19bQ ...... Born - Brooklyn, New York

1967...... A. A., Ml ami-Dade Jr. College, Miami, Florida

1969...... B. Sc. , with honors, University of Florida, Gainesville, Florida

1969-1972 ...... Teaching Assistant, The Ohio State University, Columbus, Ohio

1972-1973 ...... Goodyear Tire and Rubber Fellow, The Ohio State University, Columbus, Ohio

1973-197^ ...... Research Assistant, The Ohio State University, Columbus, hio

197^...... Hi. , The Ohio State University, Columbus, Ohio

PUBLICATIONS

Leo A. Paquette and Gary H. Bimberg, ' '5-Dibenzosemibullvalenylcar- binyl Carbene and Carbonium Ion. Some Observations on the Fate of Reactive Groups Attached to the Bridgehead Carbon of a Non- fluxtional Semibullvalene,' 1 J. Amer. Chem. Soc., 9j+, 16k (1972).

Leo A. Paquette, Gary H. Birnberg, Jon Clardy, and Bruce Parkinson, ''Stereoselective 1,4-Bromination of Semibullvalene and Tri-n- butyltin Hydride Reduction of the Dibromide,'' JCS Chem. Commun., 129 (1973)-

FIELDS OF STUDY

Major Field: Organic Chemistry

v TABLE OF CONTENTS

Page

DEDICATION...... ii

ACKNOWLEDGMENTS...... iv

VITA...... v

LIST OF TABLES...... vii

LIST OF FIGURES...... viii

- INTRODUCTION...... 1

RESULTS AND DISCUSSION

Part I. 5-Dibenzosemibullvalenylcarbinyl Carbene and Carbonium I o n ...... l4

Part Ila. Bromination of Semibullvalene...... 28

Part lib. Synthesis of Selected Monosubstituted Semibullvalene s 3 2

Part lie. Reactions and Properties of Selected Mono substituted Semibullvalenes..... *...... 4l

Part III. Synthesis and Reactions of Selected Mono substituted Bullvalenes ...... 55

EXPERIMENTAL...... 66

REFERENCES...... 123

vi LIST OF TABLES

Table Page

I Decomposition of Dibenzosemibullvalene-5-Carboxaldehyde Tosylhydrazone (51) ...... 15

II Results of Direct and Indirect Addition of PTAD to Various Monosubstituted Cyclooctatetraenes ...... 55

III Summary of Conditions for Conversion of Tetracyclic Dienes to 9jlO-Diazabasketanes ...... 57

IV Summary of Conditions for Transformation of 9 >10- Diazabasketanes to 9>10-Di&z&snautanes ...... 38

V Variable Temperature Rnr Shift Data for 106d 5=^ 106'd. (6, 60 MHz, TMS) ...... 42

VI Variable Temperature Rnr Shift Data for 107d ^ 107.' d (6, 60 MHz, TMS) ...... 44

VII LIS Rnr Data for 11£ (fi, 60 MHz, TMS) ...... 49

VIII LIS Rnr Data for 114 (6, 60 MHz, TMS)...... 51

IX LIS Rnr Data for 14^ (6, 60 MHz, TMS) ...... 62

X Decomposition of Bullvalenylcarboxaldehyde Tosyl hydrazone 115

vii LIST OF FIGURES

Figure Page

I Newman projection along the 6-7 bond of 6 ^ ...... 25

II Newman projection along the 5-6 bond of 7 1 ..... 25

III Plot of chemical shift vs mole $ of Eu(fod)3 for 11 ^ .... 50

IV Plot of chemical shift vs mole $ of Eu(fod)3 for lift- 52

V Plot of chemical shift vs mole $ of Eu(fod)3 for 1^ 5 .... 65

viii INTRODUCTION

1 Subsequent to the prediction by Doering that bullvalene (l) should be a molecule capable of degenerate Cope rearrangement to the extent that it would be totally fluxional (1,209,600 possible valence isomers), a great deal of attention has been given to molecules that are 2 inherently capable of comparable valence isomerization. For example,

etc.

semibullvalene (2), a close structural relative of 1, has attracted much 3a interest since first synthesized and initially reported to have a pmr

' 1 spectrum which is temperature independent from -110° to +117°. Re- 4 cently, Anet reported the temperature dependent prar spectra (-110° -♦

-165°) for 2, observed coalescence at -143°, and calculated the energy of activation for the degenerate Cope rearrangement of 2 to be 5*1 kcal/mole, the lowest yet found. This is slightly higher than the 2.3- 5 3*3 kcal/mole value MIHDO/2 calculations had predicted.

The synthesis of bullvalene was first accomplished in 80$ yield by the irradiation of that cyclooctatetraene dimer which melts at 76° 6,7 (j5). Later a ' 'rational'' synthesis of 1 appeared based upon the

1 preparation of barbaralone (4) from benzene in five steps, and ring expansion of this ketone with diazomethane to give bullvalone (5) 8 together with.epoxide 6, Subsequent borohydride reduction of 5_, acetyl- ation of the alcohol, and pyrolysis of the acetate at 345° gave .1 admixed with cis-9,10-dihydronaphthalene (j). This particular route suffers from a large number of steps and the several tedious separations involved. I

c h 2n 2

1) NaBH4 2) Ac20 3) 3^5°

A rather novel pathway to 1^ in three steps from methane(tri-a- 9 diazoacetone) (8) has appeared recently. The treatment of 13 in refluxing xylene with anhydrous copper sulfate in thioanisole under high dilution conditions afforded in 2$ yield the '’bullvalenetrione’' 9.

Subsequent reaction of the tris-tosylhydrazone of 9_with excess methyl- lithiura gave 1 in approximately 20$ yield.

0SMe 1) TsNHNH2 HC(CH2C0CHN2)3 CuS04 2) MeLi 8

9 1 0 The interconversions of various (c h )io hydrocarbons also provide pathways for the synthesis of 1. For example, irradiation of bicyclo ll, 12 [4.2. 2]deca-2,4,7,9-tetraene (10) affords 1 in 64$ yield, while the photoisomerization of cis-9,10-dihydronaphthalene (T) gives 1 in 13,14,15 addition to 10, iumibullvalene (11), and naphthalene (12). The formation of 1 from 7 has been suggested to arise from the conversion

hv 1 + 10

11 12

of 7 to 10 in the primary photoprocess, followed by further excited 14 e, state reaction of 10 to form L, Of the above routes to 1, Schroder’s is at present the most convenient.

Access to substituted bullvalenes has evolved principally from 2a the chemistry of 1 and Methylbullvalene (lj?) and phenylbullvalene (lA), for example, have been prepared by the action of methyllithium

and phenylmagnesium bromide, respectively, on 5_ followed by dehydra- 8 tion. Bromobullvalene (15), available from the bromination-dehydro-

1) MeLi l) 0MgBr A

16, 17 halogenation of JL, has been a key compound in providing convenient 16-20 routes to a number of other substituted bullvalenes. Chlorobull- 18 valene (l6) has been obtained in a similar manner, while the remaining

halobullvalenes have been derived from 1£_. Thus, when bullvalenyl-

magnesium bromide (IT) was allowed to react with iodine, iodobullvalene

(18) resulted, while the action of silver fluoride on 15, or 18 furnished 18 fluorobullvalene (19) • Conversion of 15 to the various alkyl ethers

l) Br2 1) S0C12 •Br 2) KOtBu 2) KOtBu

15 16

20 takes place upon treatment of 15 with the corresponding potassium Mg 15 MgBr

AgF

18 2SL

16, IT alkoxide in dimethylsulfoxide. Carbonylation of Grignard reagent

MeMgBr 15 15 >. — DM30 OR

20 CuSR 17 in the usual manner gives the carboxylic acid -21, esterification of which with diazomethane leads to carbomethoxybullvalene (22). Reaction of 15^and 17 in the presence of cobalt(ll) chloride gave bibullvalene

(2p). Similarly, methylmagnesium bromide was coupled to 15_ to give Ip.

The thioethers 24 were prepared by .the action of copper(i) alkylsul- 20 fides on 15.

CH2N2 CO2CH3

Syntheses of substituted bullvalenes based on the interconversion of (CH) 10 hydrocarbons are limited due to the relative inaccessibility of the necessary functionalized (CH)^ precursors. Three known examples of this route are the photoisomerizations of 7,8-diphenylbicyclo[4.2. 2]- 21 deca-2,4,7,9-tetraene (25), cis-9,10-dihydro-9,10-dicarbomethoxy- 8 22 naphthalene (26), and isomers of chlorofluorotricyclo[4.2.2]deca- 23 2,4,7,9-tetraene (27a,b) which furnish the corresponding disubstituted bullvalenes 28, 2g_, and £0.

CO2CH3 hv

CO2CH3

26 29

hv hv

Cl 27b 27a Interconversions of various (CH)8 hydrocarbons have formed the 3,24-27” basis for the initial syntheses of semibullvalene (£). Zimmerman 24 and coworkers have transformed barrelene (3JL) and cyclooctatetraene 25 (3£) into 2 by suitable acetone sensitized photorearrangement, in the latter case at low temperatures (-40° to -60°). The first route is badly hampered by the requisite laborious synthesis of 31, while the latter is troubled by the difficult separation of £ from large amounts of unreacted g2.

hv hv > 2 <------0 ~ 0 A , A . -k o ° 31 5£

VJhen tricyclo[3.3.0.02>6]octa-3,7-d.iene was synthesized, it was 26,27 found to undergo facile isomerization to £ at room temperature.

But this method also suffers from the tedious synthesis of starting material.

Gram quantities of £ became readily available via the indirect 2 8 structural isomerization of 32. All three routes rely upon oxidation of hydrazo compound to the intermediate azo compound ^ which spon taneously loses nitrogen to give 2.

[0] -N2 H N N I N H Sit 25 29 A recent synthesis of 2 by Malherbe originates with the diazo- ketone 36. Acetolysis of 36_ to the acetoxyketone 37, followed by reduc tion with lithium aluminum hydride to give the diol 38, conversion to the dimesylate 39> and double elimination furnishes in 20% overall yield.

CH=N; AcOH 0 40°

OAc

21 21 38

1) n-BuLi KOtBu OMs 2) MsCl DMF Substituted semibullvalenes have.been prepared by the thermal iso

merization of some cyclooctatetraenes, notably the 1,3,55 7-tetramethyl 25^30 (40) and octamethyl derivatives (42). This is somewhat unusual

CH. CH3 240° CH.

40 4l

CH. 240

CH3

CH.

42 52.

25 in that the isomerization of £2 to £ occurs photochemically as well.

Askani has developed a synthesis of substituted semibullvalenes 31 from l,5-dimethylbicyclo[3-3*0]octa-3,7-dione (44). By judicious manipulation of the stoichiometry and sequence of addition of methylmagnesium iodide, phenyllithium, or lithium aluminum hydride and subsequent 12 dehydration of the resultant alcohols, he obtains • mixtures of the desired dienes 45 and 46. Treatment of these dienes -with N-bromosucci-

CH3 CH3 several 0 R R R2 steps

44 45 46

Ri r2 ch3 a HH l) UBS b ch3 H 45 + 46 R R2 c c i4 c CH3 CH3 2) Li(Hg) d i> H e Et20 i 0 f P CHs 4Ta-f

nimide followed by lithium amalgam gives rise to the desired substituted semibullvalenes 47a-f. Of necessity these semibullvalenes carry

1,5-dimethyl substitution.

An interesting question arises when one considers the possibility of neighboring group participation by JL and 2 toward various reactive intermediates such as -CH2 , -CH: , or -CH2N2 . Would the system behave as if the reactive site were attached to an sp3-(bridgehead), sp2*25- 13

(cyclopropyl) or sp2-(olefinic) hybridized position or some combination of these. In 1, the reactive site has the potential of adjoining any one of the possible hybridized positions, -whereas in 2, there are three distinct possibilities. Position a interconverts between sp2 ,25 and

sp3 hybridization and position b between the sp2* 25 and sp2 states, while position c is permanently sp2.

a a

c ~ c

In order to lay the foundation for the possible types of reaction pathways that might be expected in the parent semibullvalenyl system, various reactive intermediates were first generated adjacent to dibenzosemibullvalene. Syntheses of the necessary monosubstituted bullvalenes and semibullvalenes for the generation of the desired reactive interme diates were then undertaken. Finally, some observations on the fate of these reactive intermediates are described. RESULTS AM) DISCUSSION

Part I

5-Dibenzosemibullvalenylcarbinyl Carbene and Carbonium Ion

The synthesis of the desired derivatives of the 5-substituted

dibenzosemibullvalene system began with 5-carbomethoxydibenzosemibull- 32 valene (48). The conversion of ester 4J3 to tosylhydrazone 51 is

- depicted in Scheme I. Lithium aluminum hydride reduction of 48 gave

Scheme I

LIAIH4 DMSO

CO2CH3 CH2OH

48 49

TsNHNH2

CHO CH=NNHTs

14 15

the alcohol 49 "which when oxidized under Rfitzner-Moffatt conditions

afforded the aldehyde in 66$ overall yield. The transformation to

£ 1 was cleanly accomplished by separately dissolving equimolar quanti

ties of 50 and tosylhydrazine in minimal amounts of hot ethanol, com

bining these, and adding two drops of concentrated hydrochloric acid.

Under these conditions 5JL precipitates quickly and is isolated in 90$

yield.

Reaction of jjl with varying amounts of sodium methoxide in anhy

drous diglyme at 125° resulted in the formation of benzo[b]fluorene

(52) in yields of 31 to 50$ (Table i). Substitution of the alkoxide

Table I. Decomposition of Dibenzosemibullvalene 5-Carboxaldehyde Tosyl- r~r*~'~r~ hydrazone (51)

Run Reactant, Base, $ Composition No. mmoles Base mmoles Solvent 53

1 0 .50 Na0CH3 0 . 75 Diglyme 31 2 2 .5 0 Na0CH3 3-75 Diglyme 50 — 3 1.00 n-BuLi 3 .0 0 Diglyme 91 — 4 1.00 n-BuLi 1 .8 0 Ethylene glycol 12 a 5 4.00 n-BuLi 12.00 Ethylene glycol 12 88 6 1.00 n-BuLi 3 .0 0 Ethylene glycol-d2 30b a 7 1. 62 n-BuLi 2.43 Ethylene glycol-d2 4ob 45c

®No attempt was made to isolate 5j5_ in these runs. In these cases, 52-d3 was the product. Obtained as 53-da* values cited are actual yieldTs determined after initial chromatography on silica gel. base by a threefold molar excess of n-butyl lithium, the same product was obtained in 91$ yield. The structure of 5£ was confirmed by corn- 34 35 parison of infrared and pmr spectra with those reported for authen tic material. When the decomposition of 51 was effected in protic ethylene glycol with the corresponding lithium alkoxide as the base,

52 was obtained as a minor product (12$). The major product was iden tified as alcohol 55 > which arises from skeletal rearrangement and solvent capture. The infrared spectrum of shows a band at 3^50 cm"1

hochsCHsoh

a which is indicative of the hydroxyl group. Particularly diagnostic of the ring-expanded structure of 5^5. the characteristic AB pattern (J^g

16 Hz; Avjyj « 39 Hz) in the pmr spectrum, which is unambiguously assign able to the ring methylene group. The rather sizable difference in the chemical shifts of protons A and B, which is also seen in the related structures 59 and 60, arises from the rigidity of the carbon framework resulting in the positioning of and Hg in substantially different chemical environments. This is in contrast to derivatives of the type

49 and where the methylene group is attached to a symmetrical frame work and likely to be almost freely rotating so that absorptions for and Ht. are equivalent. 15 Decomposition of 51 in ethylene glycol-d2 to which greater than

1 equiv of n-butyllithium had been added, afforded 52-d3 and 53-d2.

10 8 ft. 7 •9

0 52-d3 HOCH2CH2 53-la Isotopic exchange of the benzofluorenyl methylene protons was expected under these conditions. The pmr spectrum (in CDC13) of 52-d3 indicated clearly by the absence of the absorption at 6 4.02 that two deuterium atoms were attached to Cjv Also the lack of the characteristic H5 absorption which appears at 6 8. IT, conveniently separated from the 35 other aromatic protons by virtue of its unique environment, specified that a third deuterium was attached to C5. The remaining aryl protons fall into two groups: one multiplet at ca 6 7.88 comprising H-4, -6, -9, and -10, while the other at ca 5 7*53 is due to H-l, -2, -3, -7, and -8.

The absence of H5 in the pmr spectrum in addition to the proper relative integration of the two remaining aryl proton multiplets, definitively fixes C5 as the site of the third. Incorporation of the approximately three deuterium atoms was confirmed by combustion analysis (see Experi mental Section).

lyrolysis of the dry sodium salt of 5JL at I3O0 (0 .5 mm) in an apparatus designed to remove all volatile material from the heating zone as it was produced resulted in the immediate appearance of a white cry stalline solid in the cool portion of the apparatus which proved to he exclusively £2 (90f> yield). No evidence was found in this experiment for any intermediate in the conversion of ^1 to 52.

The 5-dibenzosemibullvalenyl cation was approached from two direc tions: acetolysis of tosylate 5j+ and deamination of £8. The synthetic route to and is outlined in Scheme II. Conversion of to jk

Scheme II

TsCl AcCl

CH2OTs CH2OH CH2OAc St

1 ) NaOH 1) S0C12 k8 >. 2) Hsd^ 2) NH3/Et20 @ X ® c o 2h conh2

56 51

^ © C p ® ch2nh2 19 was carried out in the usual manner with £-toluenesulfonyl chloride in pyridine. Amine 58_ was prepared by saponification of 4j3 to 5j5 followed by transformation to the acid chloride and reaction with ethereal ammonia to form 57» Reduction of amide 57 proceeded efficiently to 36 58 with diborane in tetrahydrofuran solution. 37 Acetolysis of 54_ proved, not unexpectedly, to be appreciably retarded. Heating a solution of in acetic acid buffered with sodium acetate at 118° for k j2 hr succeeded in returning 26$ of unchanged tosylate. The only other material isolated was the tertiary acetate 59*

This new acetate and the corresponding alcohol 60 derived from the reduction of 5j9 with lithium aluminum hydride were shown not to be identical to £5, and 49_, respectively. The ring expanded structure of the acetate is recognizable from the appearance of an AB pattern (J^g =

14.5 Hz and A v ^ = 48 Hz) for the methylene protons. This is consistent

HOAc LiAlR* NaOAc 118°

59 60

with structure 5g. tut not with 6l, which contains a plane of symmetry.

From the above, it follows that phenyl migration proceeds to the exclu sion of cyclopropyl migration. 20

OAc

6l '

Despite the strain which must be encountered in the skeletal rear

rangement of this bridgehead neopentyl system, no unrearranged products were detected, in agreement with earlier studies on cations of similar 38,39 type. This observation is indicative of the instability of the

primary carbonium ion 62. There is no evidence on whether 62 is a

discrete intermediate or is bypassed to give 6j5_ directly via a concerted

1,2-shift.

When the acetolysis of was carried out in a sealed tube at 220°

for 2h hr, the only product obtained was jj2. Under the same conditions

acetate 59_ was found to give 52 in 56$ yield with recovery of 59.

These data show that at higher temperatures carbonium ion 63 prefers to

aromatize with loss of a proton. 21

HQAc ^ HOAc 54. ► 52---- <------59 *— NaQAc '— NaQAc 220 ° 220°

Deamination of the perchlorate salt of 5^ in acetic acid containing 40 sodium nitrite afforded ^ in 58$ yield, Aprotic diazotization of

58 with isoamylnitrite in diglyme containing 1 equiv of acetic acid gave a product mixture consisting of (2.4$) and 59 (4-1$). In the presence of 10 equiv of acetic acid the product mixture contained 52_

(1.7$) and 5^ (4-5.5$). As was noted in the acetolysis of 54, the deami nation of 58_ proceeds through phenyl migration. This possibly occurs in concert with the elimination of nitrogen to preclude other possible reactions of the diazonium ion or its derived cation, if produced.

Great care must be taken to avoid the onset of cationic mechanisms 41 in the alkaline decomposition of tosylhydrazones in aprotic solvents.

Nonetheless, the base-assisted decomposition of 51, probably proceeds through the diazo compound 64 to the carbene 65_ (Scheme III). This inference is based mainly on the observation that the vacuum pyrolysis of the dry sodium salt of £1 resulted only in the efficient production of £2. The transformation of 6j5_ to 52 can proceed by 1,2-rearrangement 42 of one of the phenyl rings to the electron deficient carbenoid center.

The resulting highly strained bridgehead olefin 66 would, like other 43 bridgehead olefins, be quite reactive. Subsequent rearrangement to

67, followed by a [1,5] H-shift would result in relief of strain and aromatization to give 52. The product (68) which would be expected from 22

Scheme III

HC:

64 65

52 H-shift

66 67

1,2-insertion into the Cx-Cs bond does not appear to have any reasonable pathway to 5 2. In addition, migration of the cyclopropyl group is known 44, 45 to be kinetically slower than that of phenyl. 23

The reaction path was dramatically altered when the solvent for

the decomposition of £1 was changed to ethylene glycol. Again 52 was

formed, hut the major product was now £3• This is in contrast to the

results for 1-dihenzosemibullvalenylcarboxaldehyde tosylhydrazone where

the product composition is essentially the same in protic and aprotic 46 solvents.

Formation of 55, can be rationalized as occurring through solvent 47 capture of the rearranged cationic species 63 > or trapping by solvent

of 66 after its formation from 65. Although the use of ethylene

glycol-d2 does reveal deuterium incorporation into the aldehyde proton

and that this carbon atom must become C5 in 52, it does not provide a

clear distinction between carbene or cationic mechanisms. The incorpora

tion of deuterium quite likely does not involve either 51, or its anion

since tosylhydrazones are normally completely converted to their anions under such conditions and exchange at the sp2-hybridized carbon under 48 these conditions is known to be slow. Rather, exchange via the diazo

compound probably occurs by deuteration at carbon to give the corres- 49 ponding diazonium ion 69 in reversible fashion. The similarity in the

product distribution observed in the independent generation of 69 is

a 2k perhaps the strongest indication of the intervention of the cationic mechanism in the formation of 53*

Due to their relative orientation in the structurally rigid di- benzosemihullvalenyl system, there is no conjugative interaction possible between the developing carbonium ion site and the cyclopropyl or phenyl rings. Geometrically, the phenyl migration must proceed through a cr- bonded species, 70, rather than a phenonium ion. Therefore, the additive

70 inductive effects of the neighboring electron withdrawing rings in Jjh. and 69 will exert a substantial strengthening of the bond between the 3 7 leaving group and carbinyl carbon. This would be primarily reflected in the enthalpy of activation and in the rate retardation in the solvolysis of 5^_ which has already been noted.

In this case there are two possible cations available by rearrange ment, but they are not equally likely. To begin with, phenyl assistance in neopentyl solvolyses is approximately five times faster than cyclo propyl neighboring group participation, at least under normal circum- 50 ^ stances. Secondly, inspection of molecular models indicates that 63_ is capable of greater stabilization due to better overlap of the carbonium 25

P .M'

I I 1I Figure I. Newman projection Figure II. Newman projection along the 6-7 bond along the 5-6 bond of 6 3. of 71.

ion center and the phenyl ring (Figure i) than in 71 (Figure II). Fur ther, 71 is inductively destabilized by the two flanking phenyl rings.

On this basis it can be concluded that 6^_ is a more likely intermediate than 71.

The lack of interconversion between 6^ and 72 by a 1,2-hydride shift is not unexpected. The carbon-hydrogen bond that would be involved in the migration is out of alignment with the p-orbital of the cationic 26

center "by b5°. Thus, the energy barrier to migration in this rigid system is certainly higher than those involved in capture by a nucleo- 51 phile.

At elevated temperatures (220°) 6^ did undergo bond reorganization.

Cleavage of the cyclopropane ring apparently occurs (Scheme IV) to give

Scheme IV

[1,5] shift 27

73, a [1,5] hydrogen shift which provides 7^_. Aromatization of lb hy loss of a proton or its capture by solvent followed by aromatiza tion would lead to 52.

Thus the reaction pathway preferred in the 5-dibenzosemibullvalenyl carbonium ion and carbene is ring expansion via phenyl migration. Part Ila

Bromination of Semibullvalene

Synthesis of monosubstituted semibullvalenes was initially attempted in a manner analogous to the preparation of monosubstituted bullvalenes. Bromination of bullvalene (l) at -78° in dichloromethane proceeds by trans 1,4-addition to afford dibromide 75,. Its treatment with potassium t-butoxide results in elimination of the elements of hydrogen bromide and reformation of the bullvalene system to give 1 6 , IV bromobullvalene (1 5).

KOtBu Br Br

1 75 15

When semibullvalene (2) is brominated under the same conditions as for 1, a dibromide (76) is obtained. The relative stereochemistry of the bromine atoms in 76 could not be assigned by pmr spectroscopy, but 52 was found to be cis and exo by X-ray crystal structure analysis. When

76 was exposed to potassium t-butoxide, only dark tarry material was

28 29

Br Br2 CH2C12 - HC] , 0 -78° Br

76 formed. No evidence was seen for the formation of 2(4)-bromosemi- bullvalene (7 7) nor for the intermediacy of pentalene (7 8)-

e .

78 77

Treatment of 76 with trl-n-butyltin hydride in benzene at 70°, analo- 53 gous to the procedure for bullvalene dibromide, resulted in the pro duction of a mixture of dienes 79 and 80 in a ratio of 3:1* Isolation

^ (n-Bu)3SnH 7b > — I k CD * CD AIBN, 70° 12 80 was achieved in 7&f> yield by vapor phase chromatography using a 6 ft x 0.25 in. 5$ SF-96 on 6 0 /8 0 mesh Chromosorb G column, while separation was effected using a 2 ft x 0 .2 5 in. 10$ silver nitrate on 6 0 /8 0 mesh firebrick column at 60°. The pure dienes Tg, (^re-t = ^1*2 min) and 80

(t = 1 6 .2 min) had ir and pmr spectral features in accord with r6t 54 literature data.

The stereochemical control observed in the bromination of 2 can be nicely rationalized by initial sterically controlled exo attack of bromine to give intermediate 8la or its bromonium ion equivalent 8lb.

Subsequent attack of bromide ion at position a, again sterically con trolled, yields dibromide 76. The absence of 82, which would have

8la 8lb been expected from attack at position b in 8la or 8jLb, could be explained by poor stereoelectronic control along this reaction pathway. Application of the above rationale to the bromination of 1 would predict the formation of cis dibromide, mainly on the basis of compar able stereoelectronic control since steric differences are minimal. IT However, the less stable trans isomer is produced. The reasons for 55 this apparently kinetically controlled capture of bromide remains unclear.

The lack of formation of 77 when 76 is treated with base was mainly attributed to its apparent preference for 1,2 or 1,1* elimination. It was thought that under these conditions the intermediates that would have resulted would rapidly polymerize. Support for this idea came when Askani subsequently reported the preparation of 84_ from 83 by 56 treatment with potassium t-butoxide in dimethylsulfoxide. Blockage of the 1 ,5 positions of the bicyclo[3 -5 *0]octa-2 ,6 -diene system with methyl groups now prevents any 1 ,2 or 1,4 eliminations.

KOt-Bu DMSO

83 84

Reduction of 76 with tri-n-butyltin hydride resulted in straight forward production of 7J? and §0 with no evidence for any products of skeletal rearrangement. Part lib

Synthesis of Selected Monosubstituted Semibullvalenes

Adaptation of the procedure for the transformation of cycloocta-

tetraene (32) to semibullvalene (2) to the synthesis of its monosubsti

tuted' derivatives requires access to variously functionalized tetra

cyclic dienes of type 8 5. The two methods available for the production

of 85_from 32 were readily extended to structurally modified cycloocta-

- tetraenes. The process of direct dienophile addition occurs through

Scheme V

hv © sens

85 86

l) hydrolysis Y 2) oxidation

V"* 0 87 2

32 53

Scheme VI

0 a 9^ j]?TAD |PTAD R R

C b

89 90

88 R

92 91

| PTAD jPTAD

N_ 0 a, R = -CO2CH3 R T b, R = - c h 2 o h T r ’i c", R = -CH2OAc d, R = - c h 2och3 2i e, R = -CH2OMs 95 7, R = -CHO 34 a valence isomeric bicyclo[4. 2 .0 ]octatriene form obtained by prior disrotatory ring closure. In the case of a monosubstituted cyclooctatetraene (8 8), the possible involvement of four bicyclic forms (8g, gO, gl^ and g2) arises. Consequently, four tetracyclic dienes (93?

94, g5 , and g6 ) can in theory be produced when the very reactive dienophile 4-phenyl-l,2,4-triazoline-3»5-

Vi). Unlike most other dienophiles, PTAD demonstrates a tendency for competitive direct 1,4 cycloaddition to cyclooctatetraenes giving the tri 57-59 cyclic adducts g7 as well. Elimination of this undesired competing

f 0

97 reaction may be achieved by use of the alternative method of indirect 60 addition (Scheme VII). Often the relative proportions of the dienes

Scheme VII

V H 0 98 0 99 35

Table II. Results of Direct and Indirect Addition of PTAD to Various Monosubstituted Cyclooctatetraenes

Adducts, $ Mode of R Addition 93 2 1 25 96 2 1

-CeHs• Direct8, 84 ~ 1 — Indirect 6 9

-ch3 Direct8, 11 1 — u b

— — O

Indirect KjJ 1 0 0 -CH2QAc Direct 32 H 1 1 Indirect 10 35 10 ,

-ch2och3 Direct 42 i4 7 .5 — 17 Indirect 23 32 -COOCH3 Direct 8 27 2 10d Indirect 31 -CN Direct8 26 24 31 2-4 Indirect8 12 9 49 -F Direct8 — 95 Indirect8 90 ••a. mm wm

-Cl Direct6 27 30 ------0 -Br Direct 12 25 io d Indirect mmmm 28 -I Direct‘d 13 — - — 8.5b a Td c T>. R. James, personal communication. Two isomers. Three isomers. cL 6 f A single isomer detected. See footnote 63. See footnote 64.

9 ^ -9 6 in the product mixture are quite different when compared to the

results of addition (Table II). In some cases these two methods comple

ment each other such that all four possible dienes become available. 36

Acetone sensitized i^s + jr2s photocyclization of 85 resvilted in 61 smooth conversion to 8(5 in 84$ yield. Similarly when each of the various monosubstituted tetracyclic dienes was irradiated through Vycor or Corex under nitrogen in acetone or acetone-benzene solvent systems, closure resulted to give the corresponding 9jl0 -dia28>ka’Sketanes 1 0 0 , 62 101, and 102. Dienes gji a*11* 2^. 8X6 seen to yield the same photo closure product (101). A summary of the photoclosure data is compiled in Table III.

‘N ^ 0

0 1 0 0 101 102

Chemical conversion of the 9»10-diazabasketane derivatives into their 9 51 0-diazasnoutane counterparts was achieved through the agency of Ag(l) catalysis, which has been used widely in recent years for 65 the rearrangement of cubyl systems. Two sets of reaction conditions were used for the conversions: (a) silver nitrate in aqueous methanol, and (b) silver perchlorate in benzene. The latter reaction medium was used in cases where deleterious side reactions of the R group in the former medium made it necessary. Table IV summarizes the data relating to these rearrangements. 37

Table III. Summary of Conditions for Conversion of Tetracyclic Dienes

to 9>10-Diazabasketanes.

Starting 9»10-Diaza- diene (g) Solvent (ml) Filter Time, hr basketane Yield,

93c Acetone- Corex 100c 8 2 benzene, 1 :1 (0.50) (350)

94 c:9 6c Acetone- Corex 101c 72 (f-y1 :1 n , 0 .5cr>\ 0 ) benzene, (30Q)’ 1 :1

Acetone- Corex 102c 70 benzene, 1 :1 (1 .1 7) (350)

Acetone Vycor 24 lOOd 24 (5 0 0) (1.11)

94d Acetone Vycor lOld 34 (350) (0 . 71)

93d Acetone Vycor 12 102d 64 (3 0 0 ) (0.90)

Acetone Vycor 48 102a 35 (3 0 0 ) (1. 00) 358

Table TV. Summary of Conditions for Transformation of 9jlO-Diazabaske

tanes to 9>1 0 -Diazasnoutanes.

9,10-Diaza- Catalyst Solvent basketane (mmol) systems Diaza- (immol) (ml) Time, days snoutane Yield, ^

86 AgBF4 CHC13 7 87 81 (34) (lT)a (1 0 0 )

100c AgC104 CeHs 7 105 c 95 (4.0) (30) (150)

101c AgC104 c6h6 7 104c 75 (0.65) (5) (25)

102c AgC104 c6h6 8 10.5 c 91 (2.3) (1 0 ) (5 0)

lOOd AgN03 ch3oh-h2o 7 103d 40 (1.4) (59) (4:1, 75)

lOld AgNQa ch3oh-h2o 8 104d 82 (2 .0 ) (8 8) (4:1, 7 5)

10 2d AgNQa CH30H-H20 10 105 d 67 (1 .8) (70) (4:1, 50)

102a AgC104 0 ^ 6 8 105 a 88 (3-T) (2 0 )(1 0 0 )

aSee footnote 28a. 39

100

V"* o 105

0 103

Alkaline hydrolysis of 87, under forcing conditions (potassium hydroxide and aqueous ethylene glycol at ~ 100°) afforded the intermediate hydrazo compound 3j+_ which was subsequently air oxidized to the very labile azo compound 35. Spontaneous cheletropic elimination of nitrogen 28a from ^2. led "to £• Under less strenuous hydrolytic conditions the 66 semicarbazide is formed by rupture of the urazole ring. Oxidation of the semicarbazide to ^ using manganese dioxide is an adaptation of the observations, of Kelley who found the manganese dioxide oxidizes 6 7 phenylhydrazides smoothly in aqueous acetic acid at room temperature and of Pratt who uncovered the efficient oxidation of hydrazobenzene to 60 azobenzene with this reagent.

Application of the latter two-step process to the synthesis of mono substituted semibullvalenes met with varied success. Under the above conditions the conversion of 10^d, 104a, and 105d to 106d, 107d, and

108d, respectively, proceeded well. However, no substituted semibullvalene could be obtained from 105a, 10^c, lOjt-c, and 105c. When the oxidation step was carried out with air rather than Mn02} 103c and 105c afforded the corresponding semibullvalenyl carbinols 106b and 108b, respectively. However, lOjia and 104c under the same conditions did not yield the analogous semibullvalenes. Attempts to isolate the copper(l) complexes of the intermediate azo compounds also failed.

1 0 6 ' 106

e . - & e Part lie

Reactions and Properties of Some Substituted Semibullvalenes

Variable temperature pmr analysis of 1.06d ^ 106'd (Table V) indi cated that the methoxymethyl substituent prefers attachment to the cyclopropyl ring. Confirmation of this preference was obtained by double resonance techniques at -4l°, which showed H5 to be coupled to only H4 and H6 (J4,s = Js,s = 2.5 Hz). The accuracy of the chemical

CH3 CH3 OCH.

H3 1 0 6 'd 106d shift and temperature data (2 -3 significant figures) and the small ob servable changes in chemical shifts did not permit use of the method of 69 TO Wood, Fickett, and Kirkwood as later improved by Mislow to calculate the equilibrium composition.

However, the equilibrium constant can be, in principle calculated from the equations: «„ = p«v + Ci - p)»c

P - - »cV ( 6v - »c)

s * - p/(l - p) 4l /

Table V. Variable Temperature Emr Shift Data for 106d ^ 1061 d (6, 60 MHz, TM3).

Temp, °C Solv Methyl -0CHa- H2,He H3 ,H7 Hi, He Hs

+35 CD2C12 3 .20, s 3-35, s 3-37, m 5.12, m 5.1 2 , m 3.04, ta

-4l CD2Cl2-Freon 11 (1:1) 3.24, s 3-35, s 3.23, m 5*14, m 5.26, m 3 .0 8, ta

-6 9 CD2C12-Freon 11 (1:1) 3.24, s 3-35, s 3 .18, m 5*14, m 5.3 0 , m 3.09, ta b -107 CD2Cl2-Freon 11 (l:l) 3.27, s 3.36, s 3.14, m 5.16, m 5-42, m 3-14, m b -119 CDaCl2-Freon 11 (l:l) 3 -26, s 3-35, s 3 .10, m 5-16, m 5.45, m 3 .1 0 , m

aJ4 ,5 = J5,s = 2.5 Hz.

^The signals due to H2, Hs, and H5 overlap substantially at these temperatures.

4=- ro where 6^ is the observed chemical shift at a temperature m of a proton undergoing rapid exchange, &v and 6c are the respective chemical shifts of the vinyl and cyclopropyl protons in the absence of rapid exchange, 71,T2 and p is the mole fraction of one of the isomers. Values used 4 for 6 and 6 are those obtained by Anet for the 1 ’frozen'' form of v c the parent hydrocarbon. In the calculations, protons H4 and He are used since they are the least likely to be perturbed by the substituent.

On this basis, form 106d is seen to be favored to the extent of 83$ at

+35° with an increase to 95$ at -119°. These calculations are only approximations since they ignore possible perturbations by the substi tuent and any solvent-induced chemical shift changes as the temperature is lowered.

Examination of the variable temperature pmr data for 107d•^==>'10Td clearly showed the substituent to be at an olefinic position, that is, favoring 107d. Again this was confirmed through double resonance experi ments.

CH2OCH3

107d 107’d

Analogous results were seen for l(5)-methyl-, l(5)phenyl-, 2(4)- 7 3 methyl-, and 2(4)fluorosemibullvalenes. The limited number of mono- Table VI. Variable Temperature Par Shift Data for 107d S ^ 107'd (8, 60 MHz, TM3).

Temp, °C Solv Methyl -0CHa- H7 Ha He Ha H2 Hi H S

+60 CDCla 3.33, a 3.9k, ABa 5.22, ddb 5.16, br d° k.71, a 3.76, a 3.72, a 3.11, a 3.06, a

*50 CDCI3 3.3^. a 3.9k, ABa 5.28, ddb 5.12, br dc k.73, a 3.76, a 3-72, a 3.13, a 3.0T, a

-1 CDCI3 3-35, a 3.95, ABa 5.29, ddb 5.lh, br dc ^.7k, m 3* 77, a 3.73, a 3*16, a 3.08, a

-32 CDC13 3-37, s 3.96; ABa 5.32, aab 5.22, br dc k.73, a 3.82, a 3.78, a 3.18, a 3.11, a

CDaClg-Preon 11 (1:1) 3.31. a 3-91, ABa 5.27, ddb 5-1's br dc k.77, a 3.67, a 3*62, a 3 .1 b , a 3.05, a

-81 CDaCl2-Freon 11 (l: X) 3.32, a 3-93, ABa 5.2^, ddb 5.16, br de k.79, a 3.69, a 3 > & , a 3.17, a 3.07, a

-101 CD2Cl2-Freon 11 (l:l) 3.3k. a 3.93, ABa 5.30, ddb 5.18, br d° k.79, a 3.71, m 3.69, a 3-19, a 3.08, a

aJ^g ■ 11.5 Hz with dovnfleld portion further split by H3 (J ■ 1.0 Hz). \rTf8 - k.o Hs*, Jo, 7 " 5.0 Hz.

°J2>3 ■ 3.0 Hz. All apparent splittings are rounded to the nearest half Hz.

■p- ■p- substituted semibullvalenes thus far investigated indicates preferential

attachment to olefinic > cyclopropyl > aliphatic positions regardless

of the location or nature of the substituent. The fluoro case is the 74 only one that appears to contradict the calculations of Hoffmann,

who predicted that 109 should be favored over 110. At this point in-

109 110

sufficient variation in the types and position of substituents prevents

extensive testing of Hoffmann's predictions.

As expected, 3-methoxymethylsemibullvalene (108d) had a temperature

independent pmr spectrum over a similar temperature range.

All attempts to isolate the aldehyde 108f, necessary for the pre

paration of the corresponding tosylhydrazone, from oxidation of 108b have failed. A variety of mild reagents have been examined: dimethyl- 7 5 sulfoxide and dicyclohexylcarbodiimide, ruthenium tetroxide in carbon 76 tetrachloride, pyridine-sulfur trioxide and triethylamine in dimethyl- 7 7 78 formamide, manganese dioxide, and N-chlorosuccinimide-dimethyl- 7 9 sulfoxide with triethylamine. The sensitivity of the desired aldehyde

appears to be the reason behind our inability to isolate this substance.

This view is partially supported by observations from attempts to 80 oxidize 108b with silver carbonate on Celite. Progress of the oxidation c h 2oh CHO

108b 108f

was monitored by pmr analysis during which a singlet absorption was

seen to appear at 6 9* 76. Decomposition of the sample was evident

even in the short time necessary to run the pmr spectrum. An insoluble

white solid was generated and even storage at -30° under argon did not

prevent a decrease in intensity of the new signal.

Likewise all attempts to isolate the mesylate 106>e ^=i-106,e failed, however, hydrolysis products were found that could have been formed

through the intermediacy of 106e ^ .1 0 6 1e (Scheme VIII).

Scheme VIII

MsCl 106b 1 0 6'b [106e 106' e] Et3N Et20

CH2 CH2 II II

+ 115 e OH OH 47

Reaction of 106b^ l06tb with sulfene, generated in situ by combination 81 of methanesulfonyl chloride and triethylamine, gave upon aqueous workup the indicated mixture of alcohols, 11^, 114, and 115. An attempt to directly observe the mesylate by pmr resulted in the forma tion of a dark insoluble material being generated in the pmr tube.

The mixture of alcohols was inseparable by tic, but was resolved into two components by vpc (6 ft x 0.25 in. column packed with 51° XF-

1150 on 6 0 /8 0 mesh Chromosorb G at 120°): A, t ^ = 5*1 min and B, t £ =8.0 min (48:52). Both had identical accurate mass determinations for C9H3.0 O, nearly identical mass spectral splitting patterns, and were individually subjected to a lanthanide induced shift (LIS) study using Eu(fod)3. From the LIS study it became clear that A was a mixture of two compounds and that B was a single isomer. Component A was separated into its constituents by preparative vpc using a 12 ft x 0.25 in. column packed with FFAP on 60/80 mesh Chromosorb W at 120°: A3., tyet = 26.1 min and Az , t ^ = 27-9 min (61:39)* Quantities of Ax and

As isolated were sufficient for characterization by FTpmr but not for further investigation by LIS or combustion analysis.

All three alcohols had pmr absorptions characteristic of exo methylene protons; in addition Ai and B had the same type of olefinic 82 pattern as observed for other bicyclo[3.2. l]octadienes. 3he stereo chemistry of the hydroxyl function was assigned based upon the relative shifting of the exo-methylene protons in the LIS study. In the exo alcohol the exo-methylene protons are expected to move more rapidly I

kQ than the endo. Thus 113 was assigned to B and 11^ to A x (see Tables

VII, Figure III and Table VIII, Figure IV, respectively). These data combined with the presence of hydroxyl absorptions at 3350 cm-1 in the ir spectra strongly support the structures 113 and ll4.

The structure of 113 is less certain. Two possibilities ja and la are based on solvent trapping of intermediates depicted in Scheme IX.

Scheme IX

CH2 .PH2

116 117 118

H20 HpO

c h 2 Hcxy

Rnr spectral data for ll^ show the presence of four olefinic protons

(6 6.16-6.72, m), an exo-methylene group (6 ^*95 and ^.9^3 s), a three Table VII. LIS Bnr Data for (6, 60 MHz, TMS).

Chemical shift of Mole $ h 2 OH Eu(fod)3 Hi H3 h 4 H5 He H7 Hg Hg’

0 3*63 4 .3 2 5*25 6.4o 2.9^ 6.73 5*95 4.57 4.49 1.80

__a __a a _a _ a 6.3 4.6 7 3.36 4.78 4 .70 10.5 a a a 11*9 5*67 5*03 7*^5 3*78 5.09 4*93 » _b 1T.0 6.63 10.20 9*32 7.90 4.17 8.20 8 .81 > * 5*15 _b 22.8 7*69 12.30 10.80 8.42 4.69 8.70 9.85 5*39 5*35 a a _b 33*0 9 A 9 15*55 12.80 5.24 11.40 5*73c _b 42.8 10.57 18.10 14.32 9*75 5*69 10.00 12.41 5*94c

^ o t discernable due to overlap. ^Not observed. COverlap.

•t=- vo Figure III. Plot of chemical shift vs mole $ of Eu(fod)3 for 115. for shift of PlotEu(fod)3of chemical vsmole $ III.Figure • Chemical Shift (S) 0 2 0 OH 10 113 OH H OH oe% Eu(fod)3 Mole% 20

30 40 50 J~— 1 rj1j Table VIII. LIS Pmr Data for Il4 (8, 60 MHz, TMS).

Chemical shift of Mole $ Eu(fod)3 Hi H2 h3 H4 !H5 Hs Hr He He'

0 3-03a 4.03 5.30 6.37 3.03 6.68 6.10 4.62b 4.65b c c c 4.7 4.32 5.97 3.48 6.38 5.37 5.20 c c c c 9-1 5-55 3.94 6.60 6.02 5-64

13-4 6.67 9-40 8.77 7.85 4.33 7.50 6.93 6.63 6.05

18.7 7.97 11.45 10.10 8.42 4.82 7.81 7.24 7.32 6.57

29.0 10.3 2 . 15.10 12.30 9.40 5.66 8.38 7.84 8.72 7.42

39.3 12.10 18.15 13.90 10.06 6.23 8.80 8.25 9.34 7.97

49-5 13.10 20.10 14.70 10.43 6.52 9.06 8.53 9.69 8.21

aOverlap with H5. bCould be interchanged. cNot discemable due to overlap.

vn H

53 proton multiplet (5 3 * 7^--^. 16 ) and the hydroxyl proton (6 1.35 j d,

J = 9.0 Hz). These data support b over ja as the structure of 11J5.

For proposed future work 113 and llA- could be prepared independently 83 (Scheme X). The known dienone ll^, prepared in five steps from nor-

Scheme X

2) NaBH4 , OH 119 120 114

L1AIH4

bornadiene, could be converted to the triene 120 through the agency of methylenetriphenylphosphorane. Subsequent photooxygenation would yield the exo-hydroxytriene 114 as the major product. Oxidation of 114 with manganese dioxide, followed by reduction of the resulting trienone 121 with lithium aluminum hydride would give predominantly 113.

Similarly, confirmation of the bicyclo[2. 2. 2]octane ring system in

113 could be shown by hydrogenation of 113 to 122 folllowed by Jones 84 oxidation to the known ketone 123 (Scheme XI).

Scheme XI

115

122 123

From the results given above we can conclude that reactions of the mesylate occur as if it were a cyclopropylcarbinyl system, that is, it reacts through valence tautomer 106e. Bart III

Synthesis and Reactions of Selected Monosubstituted Bullvalenes

For the purpose of this study it was desired to have available a single monosubstituted bullvalene that would allow in a minimum of syn thetic manipulations for the preparation of precursors to the reac tive intermediates. Cyanobullvalene (124) appeared to fulfill these requirements based on the possible reaction sequences shown in Scheme XII.

Scheme XII

•CN CHO CH2OH

124 125 126

1 I

ch2nh2 CH=NNHTs c h 2or

127 128 129

carbenic deamination decomposition solvolysis V V

55 56 8 5 Reaction of 1^ with sodium dicyanocuprate in dimethylformamide at reflux afforded 12k in Qdf> yield- Subsequent reduction of 12k with

1 equiv of dii sobutylaluminum hydride (DIBAL-H) gave the aldehyde

125- The desired alcohol was generated by treatment with sodium borohydride in ethanol during 2k hr at room temperature (56$ from 12k).

DIBAL-H NaBH, 12k 125 126 0H EtOH

Attempts to reduce 124 or 12j> as well as the oxime derived from

125 with lithium aluminum hydride resulted in complex product mixtures

(greater than four components by tic). Examination of the pmr spectra showed clearly that the bullvalene system was no longer intact. Defini tive identification of the products was not pursued.

Aldehyde 125 was seen to be very prone to acid catalyzed rearrange ment. Exposure of 125 to a solution of j>-toluene sulfonic acid in benzene at room temperature for 12 hrresulted in conversion to a single new aldehyde identified as 150 by comparison to an authentic sample. At

!H0

130 S i

shorter exposure times (l hr), a mixture of I30 and 1^1 (72:28) was ob tained, which proved to be inseparable by standard chromatographic 57 techniques. The product ratio was determined by relative integration of the aldehydic proton absorptions in the pmr spectrum. The 2,4- dinitrophenylhydrazone (2,4-DNP) derivatives of the mixture of 130 and

1^1 did prove to be separable by fractional crystallization. The 2,4-

DNP derivative of I3I was shown to.be identical to one prepared from 86 the aldehyde generated from ester 152. Reduction of lg2 with 2 equiv

2 equiv DM30 -> 151 DIBAL-H DCC py* TFA

132 133

of DIBAL-H followed by oxidation under Moffatt conditions gave 131, spectrally identical to that prepared by acid catalyzed rearrangement of 123. Direct reduction of 152^with one equivalent of DIBAL-H at low temperature gave only a mixture of 133 and 152. Aldehyde lgl also rearranged when exposed to manganese dioxide. Thus oxidation of 155. with Mn02 gave only 130.

m . m°s > raa —

An interesting feature of this acid catalyzed rearrangement is that only one each of the possible bicyclo[4. 2. 2]decatetraene carbox-

aldehydes and naphthaldehydes are produced exclusively. This can be 58 concisely rationalized by protonation of 125a (Scheme XIII) to give the 87 bishomotropylium ion 1^5, further bond reorganization of -which affords 131. Isomerization of 1^1 to 9jl0-dihydro-2-naphthaldehyde

Scheme XIII

CHOH II

CHO CHOH 321 125 a 3&L

CHOH CHOH CHOH

-H > -> 131

322. 137 followed by oxidation with adventitious oxygen would give 1^0. Were protonation of 125b or 125c to operate and subsequent bond reorganization to follow, exclusive access to either 130 or 131 would appear very unlikely. The synthesis of 128 was completed by the heating of an ethanol

solution containing equimolar quantities of 125 and tosylhydrazine on

a steam bath for 20 min. Tosylhydrazone 128 crystallizes upon cooling

and is isolated in kk^p yield. Decomposition of 128 in glyme or ethylene

glycol with n-butyllithium or a sodium alkoxide respectively as the

base, or pyrolysis of 1^8_ results in the formation of a single compound

(139) in J>h-h2$> yield (see Table X, Experimental Section). The mass

spectrum shows a parent ion at m/e 170, which corresponds to a molecular

formula of ChHioN2- In addition, the pmr spectrum clearly shows the

128 0CHMTs

88 presence of a homotropilidine structure, and the ir spectrum shows

the presence of an >N-H absorption. In light of these data, the product was concluded to be a pyrazole. Ample precedent exists for the con- 89 version vinyl diazo compounds to pyrazoles. For example, vinyl diazomethane (lft-O) is converted to pyrazole (l4l) within a few minutes in

H H

H-Shift 60 8 9 refluxing isopropyl ether solution. The intermediate pyrazolenine can "be isolated where the [1,5] H-shift is prevented by gem-dimethyl 89 b,c substitution. A similar mechanism can be envisioned for the con version of the intermediate bullvalenyl diazomethane to the pyrazole

■>

l42a 142b m 90 (l3^). Similar results have been obtained for the pyrolysis of 1Up.

143 144

Combustion analysis of l$g indicates that it contains water of crystallization to the extent of one molecule of water per three mole cules of 139. This was confirmed by FTpmr analysis at 90 MHz of a portion of the analytical sample. At the low concentration level of the sample the one N-H and water of crystallization appeared in the same region as protons Ha, Ha ’, H5, and H^' resulting in a total area of 7.47 relative to protons Hc (area 2.00). This is a surplus of 0.47 6 l proton, attributable to the water of crystallization (calcd 0.6j proton).

When deuterium oxide was added the relative areas changed to 5•TO:2.00.

Attempts to prepare the tosylate of 126 by the stepwise addition of n-butyllithium and p-toluenesulfonyl chloride resulted instead in the isolation of a new alcohol. This was shown to be isomeric with 126 by its mass spectrum. Rnr analysis indicated the presence of six olefinic protons (m, centered at 6 6.00), an exo methylene (d,‘ 6 5*00, J =

2.0 Hz, d, 6 4.78, J = 2.0 Hz), a proton 01 to the hydroxyl group (m,

6 4.1l), two bridgehead protons (m, 6 3 *3 2 ) and the hydroxyl proton (bs,

6 I.9 8). An LIS study of the new alcohol showed clearly that the exo methylene is distant relative to the hydroxyl group (see Table IX, and * Figure V). This eliminates l46 (R=H, X C H ^ as a possibility. In

l) n-BuLi 2) TsCl CH2

addition, little similarity is seen between this LIS study and the one 91 reported for 146 (R=CH3 , X“H2) by Schroder.

1^6 \

Table IX- LIS Rnr Data for 14^ (6, 60 MHz, IMS).

Mole $ Hi h2 Ha h4 ,ht H8,H10a He Eu(fod)3 He He’ h5

_b _b _b _b _b _b 0 4.12 4.80 5.00 __b __ b b _b _b _b 9-9 5.13 5.38 4.12 b 19.4 10.00 9.52 7.58 6.78 5.38 5.70 4.70 8.63

28.9 9-07 12.73 11.27 8.47 7. 22 5.68 6.05 5.32 9.83 _b 39-2 15-25 12.85 9.08 7.58 5.95 6.35 5.87 10.8Q

48.5 11.90 17.10 13.92 9-95 7.83 6.13 6.55 6.23 11.62

59-6 12-90 18.85 14.75 9.90 8.07 6.27 6.72 6.50 12.15

aThese four protons fall into two groups of two and could not be individually assigned.

Not discemable due to previous overlap. 20 H9 H|q H2 18

H, 14 *H|

12 Hc

10 H4»Hjo

8 H 7, H8 7 H6

4 '

0 1 () 10 20 30 40 50 60 Mole % Eu(fod)3

Figu re V. Plot of chemical shift vs mole $ of Eu(fod)3 for 1^5. The structure 145 is consistent with these data and is easily ration

alized mechanistically. Nucleophilic attack hy water as depicted in

Scheme XIV during ionization of 12gb (R=Ts) or 129c (R=Ts), probably

Scheme XIV

-ROH 145 -ROH

129c

‘under control by steric and stereoelectronic factors, affords 145

directly. Hydrolysis of the £-anisoate 147 in a sealed tube at 125° in

70:30 (v/v) acetone: water for 24 hr also gave 145 (85$ yield based on

recovered l4j?) as the only product.

. p-AnCOCl 125 126 — =?---- ; > 0 ______145 2,6-Lutidine II H20/ acetone CH20C-£-An It would seem therefore that rearrangement of potential carbonium ion intermediates (such as 148 and ljt-9) can be initiated from different

148

valence isomers and may depend upon the method of generation and the timing of the transition state along the reaction coordinate. EXPERIMENTAL

Melting points are corrected and boiling points are uncorrected.

Proton magnetic resonance spectra were obtained on Varian A60-A, Varian

HA-100, and Jeolco MH-100 spectrometers; apparent splittings are given in all cases. Infrared spectra were determined on Perkin Elmer model

137 and 467 instruments. Mass spectra were recorded on an AEI-MS9 spectrometer at an ionization potential of 70 eV. Elemental analyses were performed by the Scandinavian Microanalytical Laboratory, Herlev,

Denmark. Microanalyses were not acquired for the various semibullvalenes due to their air sensitivity and thermal instability at room temperature.

The following vpc columns were used: A, 6 ft x 0.25 in 5$ SE-30 on

6 0/80 mesh Chromosorb G; B, 2 ft x 0. 25 in 10$ SF-9 6 on 6 0/80 mesh

Chromosorb G; C, 6 ft x 0. 25 in 5$ SE-30 and 3$ K0H on 60/80 mesh

Chromosorb G; D, 6 ft x 0. 25 in 5$ XF-1150 on 60/80 mesh Chromosorb G;

E, 12 ft x 0. 25 in 15$ FFAP on 60/80 mesh Chromosorb W; F, 6 ft x 0. 25 in 5$ SF-96 on 6 0/80 mesh Chromosorb G; G, 2 ft x 0.25 in 10$ AgN03 on 60/80 mesh firebrick; H, 6 ft x 0.25 in 10$ SF-9 6 on 60/80 mesh

Chromosorb G.

32 92 5-Carbomethoxydibenzosemibullvalene (48). * A mixture of 50.0 g

(0 .2 8 1 mol) of anthracene, 150 ml of nitrobenzene, and 30 ml (44.6 g,

0.351 mol) of oxalyl chloride was placed in a 500-ml flask and refluxed

66 67

for 7 hr. After this time, the

excess oxalyl chloride was removed

in vacuo and the mixture was then

transferred to a steam distillation C02CH3 • apparatus and the nitrobenzene

removed. To the remaining mixture

was added 100 ml of 10 N sodium hydroxide solution and the entire was refluxed for 1 hr. The cooled solution was filtered to remove suspended solids and the filtrate was neutralized with 10$ sulfuric acid. The precipitated solid was collected by suction filtration, washed with water, and dried in vacuo at 150°.

This solid was dissolved in 1 & of acetone, decolorized with charcoal, concentrated to 75 ml, and cooled. There was obtained 40 g (6if$) of 9- 93 anthroic -acid as yellow needles, mp 215-216° dec (lit mp 208-212°).

Treatment of 22. 2 g (0.10 mol) of 9-anthroic acid with ethereal diazomethane afforded, after recrystallization of the crude solid from methanol, 25 g (97-5$) of 9-methyl anthroate as yellow prisms, mp 111-

94 112° (lit mp 114°, 115°).

l-Carbomethoxy-7,8-dichlorodibenzobicyclo[2. 2. 2]octadiene was ob tained by heating 20.0 g (O.985 mol) of the above ester and 70 g of cis-1,2-dichloroethylene in a thick-walled sealed glass tube at 195°

(total immersion) for 28 hr. After cooling, the tube was opened, the solution was removed, the tube was rinsed with methanol (10 ml), and the combined solutions were evaporated to dryness. Recrystallization of the 6 8 solids from isopropyl alcohol yielded l4 g (49*5$) of white needles,

9 5 mp 156-157° (lit mp 159-l60o ). The combined filtrates from these recrystallizations generally contained 9-methyl anthroate and adduct in a ratio of 7:5 (pmr analysis). This mixture was suitable for recycling.

A mixture of 16.8 g (0.05 mol) of the adduct, 30 6 of Zn-Cu 96 couple, and 250 ml of absolute ethanol was refluxed for o hr. The hot solution was filtered and the solids washed twice with 10-ml por tions of hot ethanol. The filtrate was evaporated and the residue was recrystallized from isopropyl alcohol. There was obtained 12.5 g (95%) of l-carbomethoxydibenzobicyclo[2.2.2]octatriene as white needles, mp

137-1380 (lit95 mp 139-1^0°).

A solution of 10 g (3*8 mmol) of this ester in 500 ml of acetone was irradiated through Ityrex with a 450-W Hanovia lamp for 10 hr. The solvent was removed on a rotary evaporator and the resulting yellow oil was chromatographed on Florisil (500 g). The elution solvent was

9$ benzene in hexane and 100 ml fractions were taken. Fractions 1-150 were blank; fractions 151-350 contained 3*91 g of **8; fractions 351-441 contained 2.11 g of the 2-isomer. Recrystallization of both samples of benzene-hexane gave 40$ of 48, mp 128-129° (lit ’ mp 129-130°) and 32,95 21$ of the 2-isomer, mp 99-102° (lit mp 101-102°).

5-Dibenzosemibullvalenylmethanol (49). A slurry of 0.95 g (0.025 mol) of lithium aluminum hydride in 35 ml of dry tetrahydrofuran cooled to

0° was treated dropwise with a solution of 1.31 g (0.005 mol) of 4!3 in

10 ml of the same solvent. After 1.5 hr at this temperature, the mixture was stirred at room temperature for

1 hr. Hydrolysis was achieved by

dropvd.se addition of 0.95 ml of

water, 0.95 ml of 30$ sodium hydroxide CH20H .solution, and 2.85 ml of water at 0°.

The customary processing gave 0.80 g (69$) of 49 as white crystals, mp 136-137° (from aqueous methanol);

3500 cm_1’ 5TM3l3 6.8o-7.30 (m, 8, aryl), 4.28 (s, 2, -CH20-),

3*63 (t, J = 6.5 Hz, 1, cyclopropyl), 3*02 (d, J = 6.5 Hz, 2, cyclopropyl), and 2.07 (s, 1, -OH).

Anal. Calcd for CxtH^O: C, 87.15* H, 6.02.

Found: C, 87.16; H, 6.06.

3-Dibenzosemibullvalenylcarboxaldehyde (50 ). To a solution of 2.15 g

(9 .2 mmol) of 4 ^ in 20 ml of

dimethyl sulfoxide and 20 ml of

benzene was added 5.68 g (27-8 mmol)

of dicyclohexylcarbodiimide, 0.73 CHO ml of pyridine, and 0.37 ml of tri-

fluoroacetic acid. After standing overnight at room temperature, the mixture was treated with ether (100 ml) and stirred for 1 hr at which point 3*50 g ofoxalic acid was added.

After 30 min, the precipitate was filtered and washed withether (2 x

50 ml). The combined ether layers were washed with saturated sodium bicarbonate solution (2 x 100 ml) and water (100 ml), dried, and eva- 70 porated. Chromatography of the resulting oil on Florisil (elution with hexane) gave 2.04 g (95$) of as white needles, rap 124-125°, after recrystallization from benzene-, 1720 cm-1; 1 0 .2 0 (s, 1 ,

-CH0), 6.90-7.50 (m, 8, aryl), 3*95 (t, J = 6.5 Hz, 1, cyclopropyl), and 3.22 (d, J = 6.5 Hz, 2, cyclopropyl).

Anal. Calcd for C 3.7H 12O: C, 87.9°; H, 5*21*

Found: C, 87.74; H, 5 .23.

5-Dibenzosemibullvalenylcarboxaldehyde Tosylhydrazone (51.). To a hot

solution of 2.00 g (8.6 mmol) of 5(3

in.30 ml of 95$ ethanol was added

a hot solution of 1 .6 0 g (8.6 mmol)

of tosylhydrazine in 15 ml of 95$ CH=NNHTs ethanol. Upon addition of 2 drops

of concentrated hydrochloric acid, a precipitate developed immediately. The mixture was cooled and the solid was collected and washed with ethanol (3 x 10 ml).There was ob tained 3 -0 9 g (90$) of £1, mp 230-2310 dec; 6 ^ ls 6.80-8 .0 0 (m, l4, aryl, -CH=N< and>N-H), 3-78 (t, J = 6.0 Hz, 1, cyclopropyl), 3 .18 (d,

J =6.0 Hz, 2, cyclopropyl), and 2 .38 (s, 3> methyl).

Anal. Calcd for C24 H 20N2O2S: C, 71-98; H, 5-03; N, 7.00.

Found: C, 71-60; H, 4.97; N, 7-00.

Base-Induced Decomposition of 5JL with Sodium Methoxide. Dry, powdered tosylhydrazone £1 (l. 00 g, 2.5 mmol) was added in one portion to a suspension of 203 mg (3*75 mmol) of sodium methoxide in 20 ml of dry 71 diglyme. The flask was immersed in a' preheated oil bath and the tem perature was raised rapidly. At ca 90°, vigorous gas evolution com menced*, by 115°, gas evolution was essentially complete and the reaction mixture became red. The temperature was then raised to 125° and main tained there for 15 min. Hie cooled solution was added to ice water

(100 ml) and extracted with ether (4 x 80 ml). The combined pink ethereal layers were washed with water, dried, and evaporated. Chroma tography of the red oil on silica gel and elution with pentane yielded 35 250 mg (50$) of 52, white plates, mp 207-209° (from hexane) (lit mp

208°). The infrared and nmr spectra of 52_ were identical with those 3 4 , 3 5 reported for benzo[b]fluorene.

Base-Induced Decomposition of 5JL with n-Butyllithium. To a stirred slurry of 1. 60 g (4.0 mmol) of 5JL in 20 ml of dry ethylene glycol was added dropwise during 5 min a 5-ml sample of 1.21 M n-butyllithium in hexane (6.0 mmol). The reaction flask was immersed in a preheated oil bath, the temperature was raised rapidly to 125°, and maintained there for 15 min. The mixture was worked up as above to give after silica gel chromatography: (a) elution with pentane - 91 mg (11$) of 5j*>

(b) elution with ether-pentane (1:1) - 700 mg (87-5$) of 55? mp 96-97°; vma£ 5450 cm"1; 6™ s l3 6- 76"T- 25 (m’ 8’ 3 .50-3 .83 (m, 4, -CH20-), 3 .5 0 and 2 .9 0 (AB pattern, = 14. 5 Hz, Av = 37 Hz, 2, methylene), and 2.50-3 .3 0 (m, 3 , cyclopropyl).

Anal. Calcd for CigHisOs: C, 81.98; H, 6.52.

Found: C, 81.89; H, 6.42. 72

Base-Induced Decomposition of 51 in Ethylene Glycol-d2. To a slurry of

650 mg (1 .6 2 mmol) of £ 1 in 8 ml of ethylene glycol-d2 was added under nitrogen 2.10 ml of 1.16 M n-butyllithium (2. kb mmol) in hexane via

a syringe. As above, the mixture was heated rapidly to 130°, maintained

at that temperature for 15 min, and quenched by the addition of water

(20 ml). The milky suppension was worked up as described earlier to afford upon elution with hexane 120 mg (40$) of 52-d3 and upon elution with ether-hexane (l:l) and ether 202 mg (45$) of 53-dg*

For 52-d3: combustion analysis (falling drop method - Josef

Nemeth, Urbana, Illinois): 2.89D per molecule; —/e: 219; ^ M S ^3

8.05 (m, 4, H-4, 6 , 9, 10) and 7.18-7.70 (m, 5, H-l, 2, 3 , 7, 8).

For 52-d2: 6 ^ 1;3 6.75-7-25 (m, 8, aryl), 3.50-3-83 (m, k,

-CH20-), and 2.5O-3 .3O (m, 3» cyclopropyl).

Vacuum pyrolysis of the Sodium Salt of 51.. To a suspension of 800 mg

(2.0 mmol) of £1 in 10 ml of dichloromethane was added 84 mg of 57$ sodium hydride-mineral oil suspension (2.0 mmol) in one portion.

Hydrogen was evolved gradually while the mixture was stirred at room temperature for 6 hr. The precipitated solid was filtered but not dried completely and transferred while moist to a beaker where it was kept in vacuo overnight. There was obtained 680 mg (8l$) of white solid.

This solid (90 mg, 0.21 mmol) was introduced into a pyrolysis flask to which was applied a vacuum of 0.5 mm. Upon lowering the apparatus into a preheated (.130°) oil bath, a white solid appeared I

73 immediately in the cool zone of the vessel. Aftei; 30 min, these white plates were collected (43 mg, 9C/$>) and examined by pmr spectroscopy.

The spectrum was identical with that of benzo[b]fluorene.

5-Tbsyloxymethyldibenzosemibullvalene (^4). A solution of 234 mg (1.0

'mmol) of 4£ and 3^1 mg (2.0 mmol)

of freshly recrystallized jd-toluene-

sulfonyl chloride in 6 ml of pyri

dine was maintained at 0° for 45 hr.

CHaOTs reaction was quenched by addi

tion of 10 g of ice water. Extrac tion with ether (3 x 25 ml) was followed by washing of the combined ether layers with 6 N hydrochloric acid (2 x 25 ml) and water (2 x 30 ml). The ether layer was dried and evaporated and the resulting oil was crystallized from hexane (15 ml). The yield of dried crystals was

328 mg (85* )i W 139.5-141°*, 1360, 1193, and 1 1 8 0 cm"1; fl ^ 3

6.70-7.90 (m, 12, aryl), 4.75 (s, 2, -CH20-), 3*58 (t, J = 6 .5 Hz, 1, cyclopropyl), 3*08 (d, J = 6.5 Hz, 2, cyclopropyl), and 2.42 (s, 3, methyl).

Anal. Calcd for C24 H2 0 O3S: C, 74.20; H, 5-19; S, 8.25*

Found: C, 74.45; H, 5*31; S, 8.14.

5-Acetoxymethyldibenzosemibullvalene (5J>) ■ A solution of 234 mg (l. 0 mmol) of 4§. 71*5 m (1*0 mmol) of acetyl chloride in 10 ml of pyri dine was stirred at room temperature for 2 hr and 50 ml of water was then added. The product was extracted with ether and the ether layers were dried and evaporated. Recry

stallization of the residue from

hexane-benzene afforded 200 mg (70$)

of 55 as colorless prisms, mp 155- CH20Ac — 3-56°; v - 1727 cm"1; 6 ^ 13 6.90-7-40

(m, 8, aryl), 4.91 (s, 2, -CH20-),

3*62 (t, J = 6.0 Hz, 1,cyclopropyl), 3*10 (d, J = 6.0 Hz, 2, cyclo propyl), and I.98 (s, 3> methyl).

Anal. Calcd for C^HisOs: C, 82.58*, H, 5*84.

Found: C, 82.35; H, 5-91.

5-Dibenzosemibullvalenecarboxylic Acid (56). A solution of 1.00 g (3-9

mmol) of 48 in. 5 nil of methanol

and 30 ml of 20$ aqueous sodium

hydroxide was refluxed overnight.

Neutralization was effected with 10$ C02H sulfuric acid and the resulting milk-

like suspension was digested on a steam bath for 30 min. The white plates so produced were collected, washed with water, and dried to give 880 mg (91$) of £6. Recrystalli zation from hexane-benzene yielded fluffy white needles, mp 202-203°;

VS £ 1698 Cm‘1; 6TM3l3 6m9° ‘7 ’60 (m’ 83 aiyl)’ ^ 10 (t> J = 6'5 HZ’ 1> cyclopropyl), and 3>18 (d, J = 6 .5 Hz, 2, cyclopropyl).

Anal. Calcd for C xtHx202: C, 82.24; H, 4.87.

Found: C, 81.95; H, 4.80. 5-Dibenzosemibullvalenecarboxamide (57). To a suspension of 2.08 g

(8.4 mmol) of £6 in 7 ml of benzene

was added 2.00 g (l6.8 mmol) of

thionyl chloride and 600 mg (8.4

mmol) of pyridine. After being

C0NH2 stirred for 20 hr at room tempera

ture, the reaction mixture was freed of solvent and excess thionyl chloride in vacuo. The residual tan solid was dissolved in ether and added in one portion to 200 ml of ether previously saturated with ammonia. After 30 min, water (150 ml) was added, and the ether layer was separated, washed with water (100 ml) and saturated sodium chloride solution (100 ml), and dried. After evaporation of the ether, the crude solid was recrystallized from benzene-hexane to afford 1.98 g (9&fc) of ^7 as white needles, mp l8l-

182°; 1625-1695 cm"1 (broad)-, 6 ^ 13 6.90-7*50 (m, 8, aryl), 6.45

(br, 2, -KH2), 3.78 (t, J = 6.0 Hz, 1, cyclopropyl), and .3 .21 (d, J =

6.0 Hz, 2, cyclopropyl).

Anal. Calcd for CXTHi3N0: C, 82.57; H, 5-50; N, 5-66.

Found: C, 82.78; H, 5*43; N, 5*31.

5-Dibenzosemibullvalenylmethylamine (^8)* To a solution of 1.00 g (4.05

mmol) of 57_ in 15 ml of anhydrous

tetrahydrofuran was added dropwise

32 ml (3 2 .0 mmol) of diborane in the

same solvent (ca 1 M) during 15 min

CH2NH2 under nitrogen. The mixture was

subsequently refluxed for 8 hr, cooled, and treated carefully with 16. ml of 6 N hydrochloric acid. The tetrahydrofuran was evaporated and the resulting paste was brought to pH 10 with 5 N sodium hydroxide. Extraction with ether (3 x 50 ml) was followed by washing, drying, and evaporation of the organic layer.

The resulting oil was dissolved in ether and treated with 70$ perchloric acid-ethanol (l: l). There was obtained 850 mg (63$) of perchlorate salt, mp 246-248° dec.

Anal. Calcd for CxyHisdNO*: C, 6l. 18; H, 4.83; N, 4.20.

Found: C, 60.92; H, 4.83; N, 4.08.

Alternatively, the free amine can be obtained as white needles, mp 117-118°, by direct crystallization of the oil initially produced from hexane; 6.90-7*40 (m, 8, aryl), 3*60 (t, J = 6.0 Hz, 1, cyclopropyl), 3*55 (br s, 2, -CH2N), 3*1° (d, J =* 6.0 Hz, 2, cyclopropyl), and 1.35 (br, 2, -NH2).

Acetolysis of 54. A^, A solution of 60 mg (0.154 mmol) of 54^ and 14 mg

(0 .1 3 2 mmol) of anhydrous sodium carbonate in 5 “1 of dry acetic acid was refluxed (ll8°) for 432 hr. Aliquots (l ml) were removed after 24- and 2l6-hr periods; the first of these showed only unreacted 54, while the latter indicated but minimal (~ 15$) conversion to a single product.

The acetic acid was removed in vacuo and the residual solid was shown by pmr to consist only of £4 and £2.. Preparative tic of this solid on silica gel G afforded 9 mg (26$) of recovered 54, and 11 mg (43$) of 59

(elution with 10$ ether in hexane - R_ = 0.71; R_ = 0.80). This f j n * 15 acetate was obtained as white prisms, mp 117-118°, from hexane; vmax 77

1748 cm"1*, 6.82-7*30 (m, 8, aryl), 3.82 and 3.02 (AB pattern,

= l4.5 Hz, Av = 48 Hz, 2, methylene), 2.55-3*08 (m, 3, cyclopropyl), and 2.06 (s, 3, methyl).

Anal. Calcd for C 19H 16O2: C, 8 2.58; H, 5*84.

Found: C, 82.42; H, 5*85*

B. A mixture of 49 mg (0.126 mmol) of 54> 12 mg (0.10 mmol) of anhydrous sodium carbonate, and 2 ml of dry acetic acid was sealed in a tube and kept at 220° (total immersion) for 24 hr. The tube was cooled and the contents were treated with 20 ml of water. The milky suspension was extracted with ether (2 x 25 ml) and the combined ether layers were washed with water (2 x 30 ml) and saturated sodium bicarbonate solution

(2 x 30 ml), and dried. The yellow oil remaining on evaporation was subjected to tic purification as above. The only identifiableproduct was benzo[b1fluorene (52, 14 mg, 50$). The remainder consisted of a viscous yellow-brown gum.

A control experiment in which 27.6 mg (0.10 mmol) of 59, was substi tuted for 54 , led under the same reaction conditions and workup to a

60:40 mixture of £2 and j?9.

Reduction of To a stirred suspension of 63 mg (l. 65 mmol) of lithium aluminum hydride in 4 ml of anhydrous tetrahydrofuran was added dropwise at 0-5° a solution of (91 mg, 0.33 mmol) in 2.5 ml of the same solvent. After 1 hr at room temperature, the reaction mixture was quenched with water and processed as predescribed to yield 42 mg (55$) KBr of 60 as white crystals, mp 174-176°, from benzene-hexane; v_D„ 3510-3125 max t

78 cm-H 6 ^ 13 6.80-7.30 (m, 8, aryl), 3-45 and 2 .89 (AB pattern, = l4. 5 Hz, Av = 33*5 Hz, 2, methylene), and 2.40-2.90 (m, 3> cyclopropyl).

Deamination of 5j3 Perchlorate. To a suspension of 300 mg (0.90 mmol) of the perchlorate salt of 58 in 10 ml of dry acetic acid was added 120 mg (1.74 mmol) of sodium nitrite in small portions. The solid gradually dissolved, nitrogen was liberated, and the solution became yellowish- orange. Benzene (30 ml) and water (30 ml) were added and the water layer was extracted with additional benzene (2 x 30 ml). The combined organic layers were washed in turn with saturated sodium bicarbonate (2 x 80 ml) and sodium chloride solutions (2 x 80 ml) and dried. The solvent was evaporated and the resulting oil was chromatographed on silica gel. The yellow oil which eluted with 20$ ether in hexane was further purified by tic (silica gel G) to afford a faint yellow solid.

Two recrystallizations from hexane furnished 54 mg (58$) of 52, as white prisms, mp 117-118°.

Aprotic Deamination of 58. A solution of 270 mg (l. 16 mmol) of 58 in 20 ml of diglyme was treated with 60 m-1 of dry acetic acid and 171 M-l of freshly distilled i^-amyl nitrite and then heated rapidly to 130°. When gas evolution ceased, the flask was cooled and 50 ml of water was added.

The products were extracted with ether (50 ml) and this organic layer was washed with water (50 ml), dried, and evaporated. Ihe resulting yellow oil was chromatographed on a preparative tic plate (silica gel G - elution with 20$ ether in pentane). The higher band yielded 6 rag 79

(2.4$) of 52 and the lower R^. hand furnished 130 nig (4l$) of 59; The remaining material which did not migrate from the origin was an ill- defined yellow gum.

■When a ten-fold increase in the amount of acetic acid was utilized, there was obtained 4 mg (l.7$) of 5£ and 138 mg (43*5$) of 59.

97 Carbomethoxycyclooctatetraene (88a). In a quartz vessel 2.5 I of

benzene and 45 ml (45.6 g, 0.543

mol) of methyl propiolate were co .C02CH3 irradiated for three 12-hr periods.

Between each period the vessel was

cleaned of polymer by rinsing with

25# hydrofluoric acid-75$ nitric acid followed by water, 3 N ammonium hydroxide solution, water, and acetone. After the irradiation was completed, the excess benzene and methyl propiolate were removed by distillation in vacuo. The red gummy residue was distilled to give 9*12 g (11$) of 88a as an orange liquid, bp 90-92° (0.5 mm); 5*85 (s, 7H, olefinic) and 3*70 (s, 3H, methyl).

98 Cyclooctatetraenylmethanol (88b). To a 2-X Morton flask fitted with

an overhead stirrer, Friedrich con

denser, and a 250-ml dropping funnel CH20H was added 300 ml of dry ether and

13*7 g (360 mmol) of lithium aluminum

hydride. The slurry was cooled in an ice bath and 21. 45 g (132 mmol) of 88a in 300 ml of dry ether was added dropwise over 45 min. The ice bath was removed and the orange slurry was refluxed for 2 hr, cooled in an ice bath, and quenched with

125 ml of saturated aqueous ammonium chloride. After an additional

0.5 br, the salts were filtered through a Celite pad and the filtrate was dried over sodium sulfate and evaporated in vacuo. The residue was distilled to yield l4.64 g (83$) of 88b as a yellow liquid, bp 95-98°

(0 .5 mm); 5*86 (s, 7H, olefinic), 4.05 (s, 2H, methylene), and

1.99 (bs, 1H, hydroxyl).

99 Acetoxymethylcyclooctatetraene (88c). A solution of 7.00 g (52.2

mmol) of 88b and 25 ml of acetic

anhydride was heated on a steam bath CH20Ac for 1 hr and poured into 600 ml of

water and 200 ml of pentane. The

organic layer was washed with water

(3 x 600 ml) and saturated brine, dried over sodium sulfate, and evaporated in vacuo. The residue was distilled to give 8 .9 1 g (97$) of 88c, bp 71-74° (0 .5 mm), as a sweet smelling bright yellow liquid; 5*82 (s, 7H, olefinic), 4.52 (s,

2H, methylene), and 2.07 (s, 3H, methyl).

Methoxymethylcyclooctatetraene (88d). A 4.0 g sample of 57$ sodium hydride oil suspension (93*0 mmol of Nail) was washed with pentane under nitrogen (3 x 50 ml). To the dry sodium hydride was added 100 ml of freshly distilled dry dimethylformamide. A solution of 5*55 g (4l.4 81

mmol) of 88b in 25 ml of dry dimethyl-

formamide was added dropwise over .c h 2oc h3 15 min. As gas evolution proceeded

the solution darkened in color to

an orange-brown. After addition

was completed, the mixture was brought to 40° for 10 min, then cooled in an ice bath while 28.il- g (193*0 mmol) of methyl iodide was added in one portion. After stirring for 5 min the reaction mixture was again yellow and contained a large amount of suspended solid. Water (10 ml) was carefully added to quench the excess sodium hydride. The mixture was then added to 200 ml of water and kO ml of pentane. The aqueous layer was separated and washed again with 40 ml of pentane. The combined pentane layers were washed with water (3 x 200 ml) and saturated brine (50 ml), dried over sodium sul fate, and evaporated in vacuo. Distillation of the residue gave 5*03 g

(82$) of 88d, bp 115° (air bath temperature) at 0 .5 mm, as a bright yellow liquid. The analytical sample was obtained by preparative gas chromatography (column A, 100°, 60 ml/min, t = 1 3 .8 min); 5*72

(s, 7H, olefinic), 3 .8 0 (s, 2H, methylene), and 3 .2 8 (s, 3H, methyl).

Anal. Calcd for CiOHx20* C, 8l. 04; H, 8.l6.

Found: C, 81.00; H, 8.34.

Synthesis of Substituted 7,8-Diazatricyclo[4. 2.2.02 >5]deca-3>9-diene-

7,8-dicarboximides. Direct Addition. Equimolar quantities of the substituted cyclooctatetraene and 4-phenyl-l,2,4-triazoline-3,5-d.ione

(PTAD) were refluxed in ethyl acetate (100 ml per gram of the COT) until the color of the PTAD was discharged. The solvent was removed in

vacuo and the residue was chromatographed.

Indirect Addition. To a Dry Ice-isopropyl alcohol cooled solution of

the substituted cyclooctatetraene in dichloromethane (75 nil per gram

of the COT) was added dropwise one equivalent of bromine dissolved in

one-half the above volume of dichloromethane during 30 min. The solu

tion at this point was usually a slight orange due to a slight excess

of bromine. Stirring was maintained at this temperature for an addi

tional 15 min before the addition of one equivalent of PTAD. The reac

tion mixture was heated to reflux, whereupon the color of the PTAD was

almost totally discharged within J>0 min. The solvent was removed in vacuo and replaced with dry glyme (75 nil per gram of the COT); zinc-

copper couple (10 g per gram of the COT) was added and the mixture was

refluxed for 9 hr. After cooling, the reaction mixture was filtered,

the filtrate evaporated in vacuo, and the residue chromatographed.

Indirect Addition to 88a. Using the procedure for indirect addition,

1 .0 0 g (6 .1 7 mmol) of 88a was

reacted sequentially with 1 .0 0 g

(6 .1 7 mmol) of bromine, 1 .10 g (6 .1 7

mmol) of PTAD and Zn/Cu couple. Cry

stallization of the crude reaction

product from ethanol afforded 640 mg

*At this point the residue could be crystallized from methanol. Usually the 1-substituted isomer crystallized first. \

(31$) of 95a, recrystallization of which from benzene-hexane gave white prisms, mp 176-177°, 7*^° (m> 5H, aromatic), 6.75 (m, 1H, olefinic), 6.10 (m, 3H, olefinic), 5*°5 (m, 1H, bridgehead), 3*92 (s,

3H, methyl), 3 .65 (d, J = 4 Hz, 1H, H2), and 3*45 ( " t " , J = 4 Hz, 2H,

H 3).

Anal. Calcd for C 18H 3.5N3O4 : C, 64.09; H, 4.48; N, 12.46.

Found: C, 64.36; H, 4.58; N, 12.37.

Direct Addition to 88a. A mixture of 4 .3 8 g (26.4 mmol) of 88a and 4.73

g (27.0 mmol) of PTAD was refluxed

for 6 hr in ethyl acetate. Very

careful chromatography on Florisil

(400 g) (elution with 25$ ether-

hexane) gave in order of elution:

1 .0 8 g of 88a, 1 .8 8 g (28$) of

as a white solid, and 0.14 g (2$) pTV"*! of slightly impure 9,6a; 3 7*41

(m, 5H, aromatic), 7.06 (dd, J = 2.0,

6.0 Hz, 1H, olefinic a to ester),

6.00 (s, 2H, cyclobutene olefinic),

5.59 ( " q " , J = 2.0, 1H, bridgehead), 96a 5.18 (m, 1H, bridgehead), 3*59 (s,

3H, methyl), and 3-43 (m, 2H, cyclobutene aliphatic). Also isolated was

0.55 g (8$) of 93a as white needles, mp l8l-l82° [6 ^ 13 7-42 (m, 5H, aromatic), 6.63 (s, 1H, cyclobutene olefinic), 6.20 (m, 2H, olefinic),

5.16 (m, 2H, bridgehead), 3-72 (s, 3H, methyl), 3*53 (t, J = 4.0 Hz, 1H, 84 cyclobutene aliphatic), and 3*30 (t, 'J = 4 . 0 Hz, 1H, cyclobutene ali phatic)] and O.6 9 g (10$) of an open isomer, which was not investigated further.

Anal. Calcd for C 1SH 15^304 : C, 64.09; H, 4.48; N, 12.46.

Found: C, 63.96; H, 4.58; N, 12.30.

Indirect Addition to 88c. As described for the indirect addition proce

dure 3 .0 0 g (1 7 .0 mmol) of 88c was

treated sequentially with 2.73 g CHpOAc (1 7 .0 mmol) of bromine, 2.93 g (17-0

mmol) of PTAD and Zn/Cu couple. The

crude reaction product was chromato

graphed on a short column of Florisil

(elution with chloroform) to give a , c h 2oac mixture of the three dienes 9£c, §4c,

96c as an oil. Crystallization

from ethanol gave 2 .1 0 g (35$) of

95c as white fluffy needles, mp

155-156°; 6^ ls 7.45 (m, 5H, aro

matic), 6.19 (m, 2H, olefinic), 5-92

(s, 2H, cyclobutene olefinic), 5*°5

(m, 1H, bridgehead), 4.92 (s, 2H,

methylene), 3*35 (®, 2H, cyclobutene

aliphatic), and 2.12 (s, 3H, methyl). CHgQAc

& Anal. Calcd for CigH^^O.*: C, -64.95; H, 4.88; N, 11.96.

Found: C, 64.66; H, 4.79; N, 12.09.

The filtrate contained 1.23 g (20.6$) of 94c and 96c in roughly equal amounts as determined by inspection of the methyl signals in the pmr spectrum. The structural assignments were confirmed by the fact that this mixture of dienes yielded the same cage compound (101c ) when irra diated. Calcd ra/e 351.1219, found 351-1224.

Direct Addition to 88c. As described previously for direct addition,

CH20Ac 2 .7 0 g (15*3 mmol) of 88c and 2.68 g

(15.3 mmol) of PTAD were refluxed

for 4.5 hrs in ethyl acetate. Chroma

tography on Florisil (elution with W 0 dichloromethane) yielded 1 .7 0 g 22« (32$) of an oil which crystallized from ethanol to give 9pc_ as a white powder, mp 108-108.5°; 7- 42

(m, 5H, aromatic), 6.14 (m, 2H, olefinic), 5*83 (bs, 1H, cyclobutene olefinic), 5*00 (m, 2H, bridgehead), 4.4l (bs, 2H, methylene), 3*20 (m,

2H, cyclobutene aliphatic), and 2.02 (s, 3H, methyl).

Anal. Calcd for C19H XTN304: C, 64.95; H, 4.88; N, 11.96.

Found: C, 64.64; H, 4 .98; N, 11.99-

Continued elution with dichloromethane yielded a mixture of three open isomers which were not investigated further. Indirect Addition to 88d. Using the previously described procedure for

indirect addition, 1 .0 0 g (6.75 mmol)

of 88d was reacted sequentially with

1.14 g (6.75 mmol) of bromine, 1.25

g (6. 75 mmol) of PTAD and Zn/Cu

couple. Chromatography on silica 0 gel (elution with 25$ ether-hexane)

gave in order of elution: 3^9 mg

(11.5$) of an unknown substance

whose mass spectrum indicates that

it has the formula: C18Hi7N303Br2,

calcd m/e (for ^Br) 480 .9638, found

V ’V 480.9645 [when reexposed to Zn/Cu 0 sis. couple, the compound was recovered unchanged]*, 693 mg (32$) of 95d, recrystallized ftom ethanol to give flat white needles, rap l46-l46.5°;

7*42 (m, 5H, aromatic), 6.05 (m, 4H, olefinic), 5*05 (m, 1H, TMb bridgehead), 4.18 (AB pattern, = 11.0 Hz, A v ^ = 10.0 Hz, 2H, methylene), 3 .45 (s, 3H, methyl), and 3 .3 3 (s, 2H, aliphatic cyclobutene)

511 mg (23$) of 94 a, recrystallized from ethanol to give white plates mp 201-202°; 7-48 (m, 5H, aromatic), 6.15 (m, 4h, olefinic),

5.05 (m, 2H, bridgehead), 3*75 (AB pattern, J ^ = 9-0 Hz, - 19.0

Hz, 2H, methylene), 3*43 (s, 3H, methyl), and 2.95 (d, J = 5*0 Hz, 1H, aliphatic cyclobutene). Anal. Calcd for C 18H irN303: C, .66.86; H, 5-30; N, 13.00.

Found (for 93d): c, 66.74; H, 5.39; N, 13-22.

Found (for 94d): C, 66.56; H, 5-27; N, 12.99*

Direct Addition to 88d. As described for the direct addition, 2.00 g

(13.5 mmol) of 88d and 2 .8 6 g (13.5

mmol) of PTAD was refluxed for 6 hr

in ethyl acetate. Chromatography on

silica gel gave, in order of elution

with chloroform, 750 mg of recovered

88d, 203 mg (7-5$) of 95d, 324 mg

(l4$) of 94d> 1 .1 1 g (4l$) of 9^d, and 460 mg (17$) of open isomers 9Jd.

Eecrystallization from ethanol gave ggd as white prisms, mp 114-115°;

6TMSl3 ^ 0 (”» 5H’ aromatic)» 6- 18 (m> SH* olefinic), 5-85 (s, 1H, cyclobutene olefin), 5.05 (m, 2H, bridgehead), 3-78 (s, 2H, methylene),

3-3° (s, 3H, methyl), and 3-25 (m, 2H, cyclobutene aliphatic).

Anal. Calcd for C 18H 17N3O3: c, 66.86; H, 5*30; N, 13.00.

Found: C, 66.53; H, 5.33; N, 12.92.

The open isomers were not investigated further.

Conversion of Cyclooctatetraene (3 2) to N-Phenyl-9,10-diazapentacyclo-

[4.4.0.02’ 4 .03 ,s.05*T]decane-9,10-dicarboximide (8 7). A solution of

30.0 g (288 mmol) of 32 in 300 ml of dichloromethane was cooled in a Dry

Ice-isopropyl alcohol bath while 46.1 g (288 mmol) of bromine in 3°0 ml of dichloromethane was added dropwise during one hour. Fifteen minutes after addition was completed, 50.4 g (288 mmol) of PTAD was added and the reaction mixture allowed to reach- room temperature. When ambient temperature was almost attained, the reaction became exothermic and refluxed for five minutes as the color of the PTAD discharged and a white precipitate was deposited. The solid was isolated by filtration, washed with acetone, and air dried to give 85.4 g of §2. (R=H) as a white powder. The filtrate and washes were combined and the solvent was removed in vacuo. The orange residue partially crystallized from acetone to give an additional 24. 6 g of £9 (R=H) as a white powder, for a total yield of 110.0 g (87$); ®^]VE^3 7*23-7.57 (m, 5H, aromatic),

6.1*0-7.11*- (m, 2H, olefinic), 5.01-5-52 (m, 2H, bridgehead), 4.62-4.99

(m, IK, >CHBr), 4.12-4.23 (m, 1H, >CHBr), and 3- 51-5- 77 (m, 2H, ali phatic). This solid was treated with 30 g of Zn/Cu couple and 1.0 I of DMF at the reflux temperature for 9 hr, filtered while hot, and freed of dimethylformamide at 5 mm on a rotary evaporator. The resulting solid was isolated by filtration using 2 I of 3 N sulfuric acid as the rinse followed by washing with water and air drying to give 69.5 g (86$ from 3 2 ) of 8£; 6 ^ 13 7-42 (m, 5H, aromatic), 6.19 (t, J = 4.0 Hz, 2H, olefinic), 6.00 (s, 2H, cyclobutene olefinic), 5-02 (m, 2H, bridgehead), and 3-33 (m> 2H, cyclobutene aliphatic). Irradiation of 85_in 5-0-g portions in 500 ml of acetone was effected using a Hanovia 200 W medium- pressure mercury vapor lamp through Vycor optics for 24 hr. The acetone was removed in vacuo to give a yellow residue. The combined residues of the fourteen runs were chromatographed on a short alumina (Act i) column (elution with chloroform) to afford 52.0 g (75$) of 86 after CDC1 recrystallization from ethanol; 6^ 3 7-46 (m, 5H, aromatic), 5.01 89

(in, 2H, bridgehead), 3 .6 0 (m, 4h, equatorial), and 3 .15 (m, 2H, axial).

The silver(l) rearrangement of 86 was carried out in three runs of

IT. 3 g (l6l mmol) each, by refluxing with 600 ml of 4:1 methanol-water and 130 g (765 mmol) of silver nitrate in the dark for 3 days. After cooling, the reaction mixture was added to 1 .5 $> of distilled water and extracted with chloroform (3 x 200 ml). The combined chloroform layers were washed with water and brine, dried over magnesium sulfate and evaporated in vacuo to yield the crude 8? as an off-white solid. The three runs were combined and recrystallized from ethanol to give 4l. 6 g (80%) of 87 as white needles; 6 ^ 13 7-46 (m, 5H, aromatic), 5.07

(m, 2H, bridgehead), and 2.00 (m, 6h, cyclopropyl).

Irradiation of Substituted 7>8-Diazatricyclo[4 .2. 2.02»5]deca-3J9"diene-

7^^dica£boximides. A solution of the diene in acetone was irradiated

through Vycor or Corex optics after deaeration with oxygen-free nitrogen

for 20 min. The reactions were followed by thin layer chromatography.

The closure product invariably had a lower R^ than starting material.

Once starting material has disappeared the solvent was removed in vacuo

and the crude reaction product was either chromatographed or directly

crystallized.

Yields can usually be increased by changing the solvent to a 5®: 50

mixture of acetone and benzene, lowering the concentration of the diene,

and by using Corex optics. 90 l-Carbomethoxy-N-phenyl-9,10-diazopentacyclo[4.0 .02 >5.03 »8.04» 7]decane-

9.10-dicarboximide (l02a). A solution of 1.00 g (2.97 mmol) of 95a in

350 ml of acetone was irradiated

through Vycor optics for 48 hr.

Chromatography of the crude reaction

product on alumina (Act I, Woelm,

elution with dichloromethane) yielded

350 mg (35$) of 102a as a white solid,

which was recrystallized from benzene- hexane to give a white powder, mp 202.5-203.5°; 7-42 (m, 5H, aromatic), 5*05 (m, 1H, bridgehead), 3*89 (s, 3H, methyl), 3*82 (m, 4h, equatorial), and 3*23 (m, 2H, axial).

Anal. Calcd forC 18Hi5N304: C, 64.09; H, 4.48; N, 12.46.

Found: C, 64.36; H, 4.76; N, 12.40.

4-Acetoxymethyl-N-phenyl-9,10-diazapentacyclo[4.4.0.02’5.03 ’8.04>7] decane-

9.10-dicarboximide (lOOc). A 0.50 g sample (1.42 mmol) of 93c was dis

solved in 350 ml of- 50: 50 acetone-

CH20Ac benzene and irradiated for 6 hr

through Corex optics. Chromatography

of the crude product on silica gel

(elution with chloroform) gave 410

mg (82$) of 100c as white needles

from ethanol, mp 159-l60o; 6^ j l3

7.49 (m, 5H,aromatic), 5-03 (m, 2H, bridgehead), 4.12 (s, 2H, methylene), 91

3*54 (m, 4h, equatorial), 3*12 (m, 1H, axial), and 2.07 (s, 3H, methyl).

Anal. Calcd for C 19H 1TN304: C, 64.95; H, 4.88.

Found: C, 64.81; H, 4.84.

2-Acetoxymethyl-N-phenyl-9,10-diazapentacyclo[4.4.0 .02* 5.03,s. 04’7]- ** ^ —i ^ ~ ^ ^ ^ ^ M--ri — 1 li—1 _r_ ■—11—11 <— r - 1^ ~ c _t~ j ~ i ci_— — rj~.n r r ■— ^ ^ ^ ~ decane-9? 10-dicarboximide (l01c_). Irradiation of 0. 50 g (1.42 mmol) of

a mixture of 94c and g6c_ in 300 ml

,CH20Ac of 50:50 acetone-benzene through

Corex optics for 3 hr gave, after

chromatography on Florisil (elution % with ether) and crystallization from 0 ethanol, the desired 101c

(72$) as a white powder, mp 94-95°;

^TMs"^3 (m» 5H, aromatic), 5-02 (m, 2H, bridgehead), 4.23 (AB pattern, = 12.0 Hz, central portion of pattern not resolved, 2H, methylene), 3«59 (m, 3H, equatorial), 3*1° (ms 2H, axial), and 2.00 (s,

3H, methyl).

Anal. Calcd for C19H 17N304: C, 64.95; H, 4.88; N, 11.96.

Found: C, 64.97; H, 5.26; N, 12.09. l-Acetoxymethyl-N-phenyl-9j10-diazapentacyclo[4.4.0 .02’5.03 ’8.0 4’7]- decane-9,10-dicarboximide (102c). Irradiation of 1.17 B (5-?4 mmol) of

95c in two portions through Corex

H20Ac optics in 350 “1 of 50:50 acetone-

benzene each for 7 hr, yielded an

orange residue which was chromato- 92

graphed onFlorisil (elution with chloroform) to afford 0.822 g (70$) of 102c as a white solid which was recrystallized from ethanol, mp

170-171°; 7.1+8 (m, 5H, aromatic), 5*08 (m, 1H, bridgehead), 4.73

(s, 2H, methylene), 3.62 (m, 4h, equatorial), 3.22 (m, 2H, axial), and

2 .1 0 (s, 3H, methyl).

Anal. Calcd for C19H1TN304: C, 64-95; H, 4.88; N, 11.96.

Found: C, 64.84; H, 4.78; N, 12.14.

4-Methoxymethyl-N-phenyl-9,10-diazapentacyclo[4.4.0.0gj 5.03 ?a.04?7]- decane-9,10-dicarboximide (lOOd). A 1.11 g (3*43 mmol) sample of 93j-

was dissolved in 500 ml of acetone CH20CH3 and irradiated through Vycor optics

for 24 hr. Crystallization of the

crude reaction product from ethanol

gave 268 mg (24$) of lOOd as white

prisms. Recrystallization from

ethanol again gave white prisms, mp

156-156.5°; 6^ la 7.45 (m, 5H, aromatic), 5-05 (m, 2H, bridgehead),

3,55 (m, 4h, equatorial), 3.42 (s, 2H, methylene), 3*32 (s, 3H, methyl), and 3 .2 0 (m, 1H, axial).

Anal. Calcd for CieHi7N303: C, 66.86; H, 5-30; N, 13.00.

Found: C, 66.42; H, 5*33; N, 12.86.

2-Methoxyme thyl-N-phenyl-9,10-diazapentacyclor4.4. 0.0 2*5.03 ’8.04’7]- decane-9,10-dicarboximide (lOld). Irradiation of a solution of 711 mg

(2.20 mmol) of 94d in 350 ml of acetone for 3 hr through Vycor optics gave a yellow residue after solvent

removal in vacuo. Crystallization

from isopropyl alcohol gave 240 mg

(34#) of an off-white powder. Re

0 crystallization from ethanol afforded

lOld as a white powder, mp j 6. 5-77.5°

®TMS^3 (m> 5H, aromatic), 5*03 (m, 2H, bridgehead), 3*64 (m, 3H, equatorial), 3*52 (s, 2H, methylene), 3*37 (s, 3H, methyl), and3*l8

(m, 2H, axial).

Anal. Calcd for C18H 1TN3O3 : C, 66.86; H, 5*30; N, 13.OO.

Found: C, 66.79? H, 5*40; N, 13.05.

1-Methoxymethyl-N-phenyl-9,10-diazapentacyclo[4.4.0.02>5.03 >8.04>T]- decane-9,10-dicarboximide (l02d). A solution of 905 mg (2.80 mmol) of

93d in 300 ml of acetone was irra

diated through Vycor optics for 12

,CH20CH3 hr. Chromatography on silica gel

/ ^ 0 (elution with 75$ ether-hexane) N I gave 580 mg (64$) of 102d as a white

^ powder, mp 145-146° after crystalli- PDCl zation from ethanol; &TMS 3 7*50

(m, 5H, aromatic), 5*09 (m, 1H, bridgehead), 4.08 (s, 2H, methylene),

3 .6 3 (m, 4h, equatorial), 3*50 (s, 3H, methyl), and 3.20 (m, 2H, axial).

Anal. Calcd for CisHrrNgCfe: C, 66.86; H, 5-30; N, 13.00.

Found: C, 66.73; H, 5*36; N, 12.97. Silver Catalyzed Rearrangement of Substituted 9>10-di&zapentacyclo-

Ik .k .0.02>5.03 ’8.04>7]decane-9,10-dicarboximides. Method A. A solution of the cube compound and silver nitrate in 4:1 methanol-water was re fluxed 3-10 days in the dark. The product was isolated by adding the cooled reaction mixture to a large quantity of water followed by ex traction with chloroform. The combined chloroform layers were washed with water and brine, dried over sodium sulfate, and evaporated in vacuo to give the crude reaction product.

Method B. A solution of the cube compound in saturated (0.2 N) silver perchlorate in benzene was refluxed 7-10 days in the dark under anhy drous conditions. It is absolutely necessary that the conditions be anhydrous, and the silver perchlorate be stored under high-vacuum (0.01-

0.05 mm Hg) for 2-3 days before the saturated benzene solution is pre pared. For best results, freshly prepared silver perchlorate in benzene must be used and any last traces of water removed using a Dean-Stark trap at the outset of the reaction. The workup is identical to that employed in method A. l-Carbomethoxy-N-phenyl-9,10-diazapentacyclo[4.4. 0.02»4.03>8.05»7]decane-

9,10-dicarboximide (lOjija). Method B: 1.26 g (3*7^ mmol) of 102a and

100 ml of silver perchlorate-benzene,

. 196 hr reflux period. Crystalliza- (C I LCOoCHo IILL 7— I -< / V tion of the crude product from I benzene-hexane gave 1.11 g (88$) of

^ 105_a as a white solid, mp I58-I590; 6TM3l3 7mk5 (m> 5H’ ajrormt^ > 5.11 (m, W, bridgehead), 3-91 (s, 3H,

-CH3 ), and 2.12 (m, 6h, cyclopropyl).

Anal. Calcd for Ci8H 15N304 : C, 64.09; H, 4.48; N, 12.46.

Found: C, 64.06; H, 4.58; N, 12.19.

4-Acetoxymethy1-N- phenyl-9,10-diazapentacyclo[4.4.0.02’ 4.03,e. 05,7]~ decane-9,10-dicarboximide (l03c_). Method B: 1.4l g (4.02 mmol) of

100c together with 150 ml of silver

CH20Ac perchlorate-benzene solution; reflux

y>yrj period of 7 days. The crude reac

tion product was crystallized from

ethanol to give 1.35 g (95$) of

0 105c as white needles, mp 147-148°; PDC1 6TMS 3 ^*1*0 aromatic)» 5.07 (m,*2H, bridgehead), 4.17 (s, 2H, methylene), 2.03 (s, 3H, methyl), and 2.06 (m, 5H, cyclopropyl).

Anal. Ceiled for C19H 17N3O4 : C, 64.95; H, 4.88; N, 11.96.

Found: C, 64.64; H, 4.88; N, 12.05.

2-Acetoxymethyl-N-phenyl-9,10-diazapentacyclo[4.4.0.0 2j 4 .03 38.05»7]- decane-9,10-dicarboximide (l04c). Method B: 229 mg (0.653 mmol) of

101c and 25 ml of silver perchlorate- CHgOAc benzene solution; reflux period of 7

days. Chromatography on a short Flori

sil column yielded 172 mg (75$) of

104c, an oil which did not crystallize ^TMES^3 7.46 (m, 5H, aromatic), 5«12 (m, 2H, bridgehead), 4.03 (AB

pattern, J = 12.5 Hz, Avat, =7*2 Hz, 2H, methylene), 2.12 (m, 5H, A n AB cyclopropyl), and 1.95 (s, 3H, methyl); parent ion in the mass spectrum

was too weak for an accurate mass measurement.

l-Acetoxymethyl-N-phenyl-9,10-diazapentacyclo[4.4.0.02’4.03’8.05»7]-

decane-9,10-dicarboximide (105c). Method B: 822 mg (2.34 mmol) of

102c and 50 ml of silver perchlorate-

benzene solution; reflux period of 8

days. Crystallization of the crude 0 product from ether gave 752 mg (91$)

of 105c. Recrystallization gave 0 ---- flat needles, mp 153-154°; 6 ^ 13

7.40 (m, 5H, aromatic), 5-H (m,

1H, bridgehead), 4.88 (s, 2H, methylene), 2.11 (s, 3H, methyl), and

2 .0 5 (m, 6h, cyclopropyl).

Anal. Calcd for C 19H 17N3O4 : c, 64.95; H, 4.88; N, 11.96.

Found: C, 64.72; H, 5.01; N, 12.18.

4-Methoxymethyl-N-phenyl-9,10-diazapentacyclo[4.4.0.02»4 .03 >8.05? -

decane-9,10-dicarboximide (103d). Method A: 470 mg (l.40 mmol) of

lOOd, 10.0 g (59*0 mmol) of silver

nitrate, and 75 ml 4:1 methanol-

water; reflux period of 7 days.

Crystallization of the crude reac

tion product from ethanol gave 194 97 mg (hOjo) of 103d, mp 112-113°;o. 6.CDC13 7* 4l (m, 5H, aromatic), 5*09 ’ TMS (m, 2H, bridgehead), 3*^9 (s, 2H, methylene), 3*32 (s, 3H, methyl), and

2.05 (m, 5H, cyclopropyl).

Anal. Calcd for CaeH17N303: C, 66.86; H, 5*30; N, 13.00.

Found: C, 66.73; H, 5*33; N, 13*05*

2-Methoxymethyl-N-phenyl-9,10-diazapentacyclo[4.4.0. 02’4* 03>s. 05>7]- decane-9,10-dicarboximide (104d). Method A: 650 mg (2.01 mmol) of

lOld, 15*0 g (88.3 mmol) of silver

CH20CH3 nitrate, and 75 ml 4:1 methanol-

water; reflux period of 8 days.

Chromatography of the. crude reaction

product on silica gel (short column)

viscous oil, which could not be crystallized; 7*43 (m, 5H, aromatic), 5*08 (m, 2H, bridgehead),

3*34 (AB pattern, “ 10.5 Hz, AvAfi = 5 .8 Hz, 2H, methylene), 3.25

(s, 3H, methyl), and 2.04 (m, 5H, cyclopropyl); calcd m/e 323* 1270, found 323.1276. l-Methoxymethyl-N-phenyl-9,10-diazapentacyclo[4. 4.0.02>4 .03 ’8,05’ 7] - decane-9,10-dicarboximide (I05d). Method A: 580 mg (l. 80 mmol) of 102

1 2 .0 g (70.5 mmol) of silver nitrate,

and 50 ml 4:1 methanol-water; reflux

period of 10 days. Crystallization

of the crude product from ethanol

yielded 390 mg (67$) of 105d, mp 0 I

97-98°*, 6 ^ 13 7-60 (m, 5H, aromatic), 5-08 (m, 1H, bridgehead), lf.l6 1Mb (s, 2H, methylene), 3.^9 (s, 3H, methyl), and I.9 8 (m, 6h, cyclopropyl).

Anal. Calcd for CiaHxTN303: C, 66.86; H, 5*30; N, 13*00.

Found: C, 66.71; H, 5-58; N, 12.72.

Preparation of Semibullvalene (2 ) by the Hydrolysis and Manganese

Dioxide Oxidation of 87. A 100 ml three-necked round-bottom flask

equipped with condenser, serum cap, magnetic stirrer, and stopper was

charged with 500 mg (1-79 mmol) of 87 and 1.0 g (17*8 mmol) of potassium

hydroxide. The flask was flushed with nitrogen and 20 ml of isopropyl

alcohol was introduced through the serum cap. The mixture was brought

to reflux with stirring and held at this temperature for ^5 min before

cooling in ice. Hydrochloric acid (3 N) was added until pH ~ 2 and the

mixture was stirred for 5 min. Ammonium hydroxide (3 N) was added drop-

wise to give pH ~ 8, at which point 6 ml of pentane and 1.5 6 g of

manganese dioxide were introduced sequentially in single portions.

Die resulting black suspension was stirred for 1 hr at 0° and 30 min

at room temperature. Decantation of the solution and rinsing of the

remaining sticky solids with small amounts of pentane was followed by

washing of the combined organic layers with water (2 x 50 ml) and brine

(50 ml). The aqueous layers were reextracted with pentane and the

pentane layers were dried. Filtration gave a clear solution which

when distilled gave a residue which was isolated by preparative vpc on

column A at 65°. There was isolated 85 mg (46$) of semibullvalene. 99 Preparation and Decomposition of the Copper(i) Chloride Complex of

9,10-diazapentacyclo[4.4.0.02>4 .03 >8.05>T]dec-9-ene (;5J5;CuCl). A mix

ture of O.5 0 g (1 .8 0 mmol) of 87

and 0 .3 0 g (7*5 mmol) of sodium

hydroxide in 30 ml of isopropyl N alcohol was refluxed for 1 hr under ■ K N CuCl nitrogen. The reaction mixture

was then cooled in an ice bath and

3 N hydrochloric acid was added to pH 2-3, at which point the preci pitated solid dissolved; gas evolution was noted. After basification with 3 N ammonium hydroxide solution, the isopropyl alcohol was removed in vacuo leaving an aqueous phase with a solid which was partitioned between 50 ml of water and 25 ml of CHC13 with gentle swirling (not shaken to avoid air oxidation as much as possible). The organic layer was dried over sodium sulfate and the solvent was removed in vacuo to give a yellowish solid which was taken up in 5 ml of methanol and to which a solution of O.3 6 g (2 .1 1 mmol) of copper(il) chloride dihydrate in 5 ml of water was added until no additional complex formed (approxi mately 2/ 3 of the solution was introduced). The complex was collected by suction filtration to give 56 mg (13.5$) of a brick-red solid, 3jj»CuCl, 28C _ mp 9^-95°, dec (lit 95-96 , dec). Decomposition of 29 mg (0.125 mmol) of 35*CuCl in 2 ml of ether with 3 N ammonium hydroxide solution was vigorous. The ether layer was dried over sodium sulfate and the product isolated by preparative gas chromatography (column A) was shown to be semibullvalene, 4 mg (31$), by pmr analysis. No trace of cyclooctatetraene was in evidence by pmr. 100

Bromination of Semibullvalene (2). To a solution of 896 mg (8.60 mmol)

of 2 in 50 ml of dichloromethane Br cooled with a Dry Ice-isopropyl

alcohol bath was added dropwise

over 30 min a solution of 1 .38 g Br (8 .6 0 mmol) of bromine in 50 ml of 76 dichloromethane. The solution was stirred for an additional 15 min at -78°, then allowed to reach room temperature and evaporated in vacuo (bath temp never above 25°C). The resulting dark oil was chromatographed quickly on silica gel (2 x 20 cm column, elution with 20$ dichloromethane-hexane). There was ob tained a yellowish oil which was taken up in 10 ml of pentane and placed in the freezer under nitrogen. The white needles which formed were collected by filtration to give 0.30 g of 76. The filtrate was concentrated and distilled at 90° (bath temperature) and 0.5 mm Hg using a molecular still to give 1 .2 2 g of 76 as a pale yellow liquid which crystallized from hexane in the freezer. Total yield 1.52 g (67$) of 76, mp 71-71- 5°; 5.80 (nm, 4H, olefinic), 4.73 (nm, 2H, H ot to Br), and 3*99 (nm, 2H, bridgehead). The dibromide 18^ was somewhat unstable, it slowly decomposed, and for this reason was not submitted for combustion analysis.

Reaction of 76 with Tri-n-butyltinhydride. A mixture of 132 mg (0. 50 mmol) of 76, 302 mg (l. 05 mmol) of tri-n-butyltinhydride and 10 mg of azobisisobutyronitrile in 0. 50 ml of benzene was sealed in a thick wall ioi tube (25 cm long, 2 .0 mm wall, 3/a’' o.d.) and heated using an air bath at J0° for 10 hr. The tube was opened and the reaction mixture was injected directly onto a vpc column heated to 75° (Column F, flow rate 60 ml/min). Only one peak was observed in addition to benzene.

Isolation of this peak gave 4l.4 of a mixture of 79 and 80.

The ir spectrum of this mixture wa

(flow rate 60 ml/min), t ^(79_) = 1 6 .2 min (25$), tfc^(8o ) = 4l. 2 min 54 (75$)* The pmr spectra of 79 and 8(3 are in accord with the literature;

6lSs4 6.34-6.6 8 (m, 4h, olefinic), 3-17-3-57 (m, 2H, bridgehead), and 1.88-2 .8 2 (m, 4h, vinylic); (T9) 6.43-6.78 (m, 4h, olefinic), I Mo — 3 .56-3*87 (m, 1H, bis vinylic), and 1.85-3 .12 (m, 5H, vinylic and bridgehead). Repetition of this reduction at higher dilution (same quantities except that 25 ml of benzene was used) gave similar results.

Reaction of 76 with Potassium t-Butoxide. To a solution of 102 mg (O.39 mmol) of 7*5 in 2 ml of dry benzene was added 300 mg (2 .7 0 mmol) of potassium t-butoxide in 2 ml of benzene. The mixture immediately turned black. This was quenched with 10 ml of water and extracted with

10 ml of pentane. The organic layer was washed twice with water (15 ml each) and brine, and dried over sodium sulfate. Analysis by tic showed only a small amount of 76_ to be present. Identical results were realized at 0°.

3-Methoxymethylsemibullvalene (l08d). Treatment of 300 mg (0.93 nmol) of 105d with 200 mg (5 .0 mmol) of sodium hydroxide in 10 ml of 2-propanol 102

at reflux under nitrogen for 1. 5 hr and subsequent processing and oxi

dation (2 .0 g of manganese dioxide) as described before for the prepara

tion of 2_ gave after vpc isolation on column C at 80°, 62 mg (45$) of

108d as a clear, colorless liquid*, 4.93 (t, J = 3.5 Hz, l), 3*92

(m, 4), 3*55 (s, 2 ), 2 .9 8 (s, 3 ), and 2.85 (m, 2 )-temperature invariant.

Calcd for C 3.0H 12O: m/e 148.0888; found 148.0890.

1(5)-Methoxymethylsemibullvalene (I06d ^ 106' d). In an analogous manner,

diazasnoutane 103d (250 mg, 1.24 mmol) was partially hydrolyzed with

250 mg (6.2 mmol) of sodium hydroxide in 15 ml of 2-propanol. Subse

quent oxidation with manganese dioxide (2 .0 g) and preparative vpc purification on column C at 80° yielded J2 mg (39$) of 106d

Calcd for C 10H 12O: m/e 148.0888; found 148.0890.

2(4)-Methoxymethylsemibullvalene ( lO^Td)'» In a similar manner, diazasnoutane 104d (517 mg, 1. 60 mmol) was partially hydrolyzed with

320 mg (8 mmol) of sodium hydroxide in 20 ml of 2-propanol. Subsequent oxidation with manganese dioxide (2 .0 g) and preparative vpc purification on column C at 80° yielded 99 mg (42$) of 10Td ^ 107’d. Calcd for

CiOH 120: m/e 148.0888; found 148.0890.

Hydrolysis-Oxidation of 105a. A 50-ml three-necked flask fitted with two septa and a condenser was charged with 337 mg (l. 00 mmol) of 105a and 560 mg (1 0 .0 mmol) of potassium hydroxide then flushed with nitrogen.

To this was added 10 ml of isopropyl alcohol via syringe and the mixture was brought to reflux by means of an oil bath at 90°. As the temperature rose, the starting material dissolved and the solution became clear.

Soon after a fine -white precipitate formed. After 45 min at reflux the reaction mixture was cooled in an ice bath and 3 N hydrochloric acid was added until attainment of pH 2 where decarboxylation took place.

After 5 min 3 N ammonium hydroxide was added to return the pH to 6.

Pentane (3 ml) and dichloromethane (3 ml) were added followed by 870 mg of manganese dioxide, at which point immediate evolution of gas was noted. Stirring was maintained for one hour until gas evolution sub sided. The mixture was filtered and treated with diazomethane in ether (10 equivalents); no gas evolution was seen. The organic layer was washed with water and brine and dried over sodium sulfate. The sol vent was reduced to a volume of 2 ml in vacuo. Analysis by gas chroma tography on column F at 100° (flow rate 100 ml/min) showed only one peak identified as aniline by isolation and comparison of pmr spectra.

Repetition keeping the pH at 2 gave similar results.

Use of copper(il) chloride as the oxidizing agent and trap for the intermediate azo compound, in a manner analogous to that of the parent system, gave no complex.

3-Semibullvalenylcarbinol (108b). A mixture of 200 mg (0.57 mmol) of

105c, 150 mg (3.85 mmol) of sodium hydroxide, and 10 ml of 2-propanol was refluxed under nitrogen for 1 hr. The cloudy reaction mixture was cooled in ice while 3 H hydrochloric acid was added to bring the pH to ca 2. The solid that was present dissolved and the solution became purple in color. After being stirred for 5 min, this solution was 104 returned to pH 8 with 3 N ammonium hydroxide, treated with 300 ml of water, and extracted with chloroform (6 x 15 ml). The combined organic layers were washed with water and brine, dried, and evaporated in vacuo.

The resulting oil was chromatographed on a short Florisil column (1.5 x

10 cm). Elution with ether-pentane (1:1 gave a white solid, recry stallization of which from pentane-ether furnished 43 mg (57$) of 108b

TCRt* as white needles, mp 71-71.5°; v"~ 5520, 1350, 1007, 819, 759, 751, HlCwC and 703 cm-1; 5*21 (t, J = 4.0 Hz, l), 4.14 (m, 4), and 4.01 (s,

2) - temperature invariant. Calcd for CgHioO m/e 134.0732; found

13b-0755. l(5)-Sexnibullvalenylcarbinol (106b 106 *b). A mixture of 300 mg (0.85 mmol) of 103d, 150 mg (3*75 mmol) of sodium hydroxide, and 15 ml of 2- propanol was brought to reflux under nitrogen while being stirred mag netically. After 2 hr, the solution was cooled in ice and 3 N hydro chloric acid added. Decarboxylation occurred as the pH was adjusted to

2. Aqueous ammonia (3 N) was introduced to return the pH to ca 8 and this mixture was added to 100 ml of water and extracted with methylene chloride (6 x 10 ml). The combined organic layers were washed with water, 0.3 N hydrochloric acid, saturated sodium bicarbonate solution, and brine. Drying and solvent evaporation was followed by preparative tic on silica gel (elution with 20# ether in carbon tetrachloride) to give 62 mg (54#) of the carbinol as a pale yellow oil; 4.87-5*22

(m, 4), 3 .5 0 (s) overlapping 3.42-3-60 (m, 4, total), 3*27 (s, 1, -OH), and 3.02 (t, J = 3.0 Hz, l). Calcd for C9H lo0 m/e 134.0732; found 134.0735- 105

Preparation and Reaction of l(5)-Mesyloxymethylsemibullvalene (l06e ^

106’e). A mixture of 142 mg (l. 06 mmol) of 106b ^ 106 'b, 0.23 nil (l. 65 mmol) of triethylamine and 5 .0 ml of dry ether tinder argon was cooled using a Dry Ice-carbon tetrachloride bath (-25°). To this was added dropwise 90 M>1 (l- 20 mmol) of me thane sulfonyl chloride over 2 min; a precipitate started forming immediately. After addition was completed the cooling bath was allowed to warm gradually to room temperature over approximately 20 min. The reaction mixture which now contained a large amount of precipitate was added to a separatory funnel containing

50 ml of water and 25 ml of ether. The ether layer was washed with saturated sodium bicarbonate solution and brine, and dried over sodium sulfate. Thin layer chromatography on silica gel showed only one spot (R^ = O.3 8, 2056 ether/carbon tetrachloride). Gas chromatographic analysis on column D at 120° (flow rate 120 ml/min) showed two peaks: t .j. =5*1 and 8.0 min, relative areas 48:52. These two components were separated on the above column to give 16 mg of the 5 *1 min component, a solid, and 19 mg of 113 the 8.0 min component, a liquid. Both of these were identified as C9H 3.0O isomers by their mass spectra, calcd m/e

134.0732, found 134.0733* Their pmr spectra and lanthanide shift studies

[Eu(fod)3] indicated that the 5*1 min component was a mixture of two compounds and that the 8.0 min component was a single substance. Resolu tion of the 5 -1 min mixture into its components was accomplished on column E at 120° (flow rate 120 ml/min): t £ a 26.1 and 27*9 min, relative areas 61:39- Isolation of these two components gave 2.3 mg of the 2 6 .1 min component and 1.4 mg of the 27-9 min component, 114 and 106

115, respectively. Isolated yields are low, since a large amount of material was lost in search of a column to effect their separation;

6S b 13 (-11?* 90 MHz) 6.65 (dd, J6 j7 = 6.0 Hz, J5>6 = 3 . 0 Hz, 1H, He),

6.3k (ddd, J3 j4 = 9 .5 Hz, J4 j5 =6.0 Hz, J2>4 - 1.7 Hz, 1H, H4), 5.89

(dd, Je,7 = 6.0 Hz, Jx>7 = 3-0 Hz, 1H, H7), 5*22 (dm, J3>4 =9-5 Hz, 1H,

H3 ), 4.67 (s, IH, H9 or H9'), 4.48 (s, 1H, Hq or H9 ’), 4.32 (m, 1H,

H2), 3.54 (m, 1H, H x), 2.93 (dd, J4, 5 = 6 . 0 Hz, J5 , 6 =3-0 Hz, 1H,

H5), and 2.02 (m, 1H, hydroxyl); 6 ^ 13 (ll4, 90 MHz) 6.65 (dd, J6>7 =

5.6 Hz, J5>6 = 3-0 Hz, 1H, Hs), 6.35 (dd, J3>4 = 9*4 Hz, J4j5 = 6 .6 Hz,

1H, Hi), 6.06 (dd, J6>7 = 5-6 Hz, Jx>7 = 3.0 Hz, 1H, H7 ), 5*30 (ddd,

J3 ,4 = 9-4 Hz, J2, 3 = 3-4 Hz, Ji, 3 = 1.9 Hz, IH, Ha), 4.70 (s, 1H, H9 orHe'), 4.6 7 (s, IH, H9 or Ife'), 3*94-4.20 (m, 1H, H2), 2.98-3.14 (m,

2H, Hx and H5), 1.72 (bd, 1H, J ^ = 9-1 Hz); (115, 90 MHz) 6.16-

6.72 (m, 4h, Hx, H2, H3, and H4), 4.94 (s, IH, H9 or H9 '), 4.93 (s, IH,

H9 or Hs'), 3-74-4.16 (m, 3H, H5, H s, and H7 ), and 1.35 (bd, J = 9-0 6, On Hz).

Anal. Calcd for C9H XoO: C, 80.56; H, 7*51.

Pound (for. 113): C, 80.03; H, 7.48.

There were insufficient quantities of 114 and 115 for combustion analysis.

Attempted Oxidation of 108b^with Manganese Dioxide. A solution of 28 mg

(0 . 21 mmol) of 108b in 0. 50 ml of carbon tetrachloride was stirred magnetically under nitrogen with 50 mg (0 .5 8 mmol) of manganese dioxide.

The mixture was filtered directly into a pmr tube. The spectrum showed 107

that no aldehyde signal was present and the absorptions due to 108b

were absent. No other peaks were in evidence.

77 Attempted Oxidation of 108b by Doering's Method. To a solution of 20

mg (0 .1 5 mmol) of 108b in 0 .5 0 ml of dry dimethylsulfoxide and 125 Hi

(90 mg, 0 .9 0 mmol) of triethylamine at room temperature under nitrogen

was added 72 mg (0.^5 mmol) of pyridine-sulfur trioxide complex in 0 .5 0

ml of dry dimethylsulfoxide. This mixture was stirred magnetically

for 2 hr. Addition of the reaction mixture to 30 ml of water and

- extraction with hexane gave no material after solvent removal. Ex

traction with ether also gave nothing. When the aqueous phase was 77 acidified, as specified in the workup by Doering, and extracted with

ether no material was found.

79 Attempted Oxidation of 108b by Corey’s Method. A solution of 35 mg

(0.26 mmol) of freshly recrystallized N-chlorosuccinimide in 1.0 ml of

dichloromethane under argon was cooled in an ice bath and 26 p.1 (0 .3 6

mmol) of dimethylsulfide was introduced. A precipitate immediately

formed. The mixture was cooled in a Dry Ice-carbon tetrachloride bath,

23 mg (0 .1 7 mmol) of 108b in 0 .5 0 ml of dichloromethane was added and

the entire mixture was stirred at this temperature for 2 hr. Triethyl

amine (36 p,l, 0. 26 mmol) was added followed by 15 ml of ether. Water

(15 ml) was added and the ether layer was washed with water and dried

over sodium sulfate. After solvent removal in vacuo, pmr analysis

showed no aldehyde absorption; nor were any other signals evident. io8

Attempted Oxidation of 108b with Ruthenium Tetraoxide. The RuO^ was prepared by vigorously stirring 17 mg (0.11 mmol) of Ru02 hydrate

(Englehardt) with ^ ml of carbon tetrachloride and 6 ml of 10$ aqueous

sodium periodate at 0° until the black Ru02 dissolved (ca 15 min). The yellow carbon tetrachloride layer was pipetted out and added to an ice

cooled solution of 15 mg (0 .1 1 mmol) of 108b in 1 .0 ml of carbon

tetrachloride. The color of the RUO4 was immediately discharged and a brown-black precipitate was formed. The mixture was stirred for 5 min

and filtered through a small Celite pad. The clear filtrate was con

centrated and pmr analysis showed no aldehydic absorptions and only a

small amount of unreacted 108b.

75 Attempted Oxidation of 108b by Moffat's Method. In 1.0 ml of dry

dimethylsulfoxide and 2 .0 ml of benzene was dissolved 35 mg (0 .2 6 mmol)

of 108b and 162 mg (0.7 8 mmol) of dicyclohexylcarbodiimide. To this magnetically stirred solution at room temperature was added 1 0 .8 |il

(16.8 mg, 0.13 mmol) of dichloroacetic acid. Stirring was continued

for 6 min whereupon 10 ml of ether was added followed by 157 mg (l- 7^ mmol) of oxalic acid. After gas evolution ceased, the reaction mixture was added to 30 ml of water, stirred for 5 min and filtered to remove

the dicyclohexylurea. The ether layer was separated, washed with

saturated sodium bicarbonate solution, water and brine, dried over

sodium sulfate, and evaporated in vacuo. The resulting oil did not

contain any aldehyde or any semibullvalenyl absorptions (pmr analysis). 109

80 Attempted Oxidation of lO&b with Silver Carbonate on Celite. To a

solution of 25 mg (0 .1 9 mmol) of 108b in 2 .0 ml of dichloromethane was added 0.57 g (l.0 mmol of Ag2C03 ) of silver carbonate on Celite. The reaction mixture was purged with argon and refluxed for 8 hr. A new component appeared by tic (R^ = 0.60, 20$ ether-carbon tetrachloride).

No change in the relative amounts of the new component and 108b occurred even after 2k hr of reflux. The reaction mixture was cooled and

filtered and the ether was removed in vacuo, to give 12 mg of yellow

oil. Rnr analysis showed an aldehyde absorption at 9* 76 6. Storage of this solution in the freezer (-30°) under argon overnight resulted in the generation of a solid in the pmr tube (pmr shows less of the peak at 9 .7 6 6). Preparative thick layer chromatography was tried to

separate the new component (Rf 13 0.60) but isolation of the aldehyde from

the preparative tic plate yielded 1. 5 mg of a clear oil which when taken up in carbon tetrachloride for pmr analysis rapidly deposited

insoluble material. Only a very low intensity aldehyde absorption was

seen.

Cyanobullvalene ( l2k). A mixture of k.36 g (20.5 mmol) of bromobull-

valene, 1 .0 1 g (20.5 mmol) of

sodium cyanide, 1. 8k g (20.5 mmol)

of cuprous cyanide, and 200 ml of

freshly distilled dry dimethylforma-

mide was refluxed under nitrogen for

8 hr. The reaction mixture gradually 1 1 0 turned brown then black. After cooling to room temperature, the black solution was transferred to a separatory funnel with 500 ml of 10 N sodium cyanide solution and 500 ml of ether. The ether layer was washed with water (2 x 1000 ml) and brine (l x 1000 ml), dried over magnesium sul fate, filtered, and evaporated in vacuo. The resulting yellow oil was purified by column chromatography on silica gel (elution with 25$ ether-hexane). The colorless oil obtained was crystallized from 4 ml of ethanol in the freezer overnight as needles. These were collected by suction filtration, washed with cold ethanol (-30°) and air dried to yield 2 .6 2 g (82$) of 124, mp 60-60.5°; 2190 cm'1; 6 ^ 13 (-35°)

6 .7 6 (d, IH, olefinic P to -CN), 5*97 (m, 4h, olefinic), and 2 .5 6 (m,

4h, cyclopropyl and aliphatic); 6 ^ 2 13 (108°) b. 45 (s, half-width 7*5 IMo Hz).

Anal. Calcd for C lxIfeN: C, 85.13; H, 5.85.

Found: C, 85.27; H, 6.16.

Reduction of 124 with Lithium Aluminum Hydride. To a slurry of 53 n>6

(l. 40 mmol) of lithium aluminum hydride in 4.1 ml of dry ether was added a solution of 179 “g (1*15 mmol) of 124 in 4.0 ml of ether at such a rate that gentle reflux was maintained. Approximately 4min was required. The now white slurry was refluxed for 40 min, cooled in ice, and quenched with 1 .0 ml of water followed by 2 .0 ml of 3°$ sodium hydroxide solution. The ether layer was separated, dried over sodium sulfate, and evaporated to give an orange oil. Pmr analysis showed clearly that the bullvalene skeleton was no longer present. Tic analysis Ill revealed a mixture of at least four components; eluent 75$ ether-hexane

(developed twice), = 0.28, 0.52, 0.67, 0.91* This reaction was not investigated further.

Bullvalenylc arb oxaldehyde (125) - A solution of 1.00 g (6.44 mmol) of

124 in 50 nil of dry benzene was

stirred magnetically at room tempera

CHO ture and 5 *6 ml of 26$ diisobutyl-

aluminum hydride by weight in hexane

(1 .1 equiv) was added via syringe.

The temperature of the reaction mix ture was raised to *+u" using a warm water bath and stirred for 0. 5 hr.

The solution was then cooled in ice and residual active hydride was quenched by careful addition of 2. 0 ml of methanol followed by 2 .0 ml of water.’ After being stirred for 1 hr, the mixture was filtered through a pad of Celite and the filtrate dried over magnesium sulfate, filtered, and evaporated in vacuo to yield an orange oil. This material crystallized slowly from benzene/hexane to give 0.678 g (67$) of 125 as a yellowish solid, mp 153-159°; 2830, 1629, l6l0 cm-1;

9.05 (s, IH, -CH0) and 4.50 (vb, 9H, bullvalenyl).

Acid-Catalyzed Rearrangement of 125. To 6.3 ml of a solution of £- toluenesulfonic acid in benzene (1 ,0 mmol/ml) was added 50 mg (0.63 mmol) of 125 and the mixture was stirred for 12 hr. Addition of the solution to water (30 ml) and extraction with ether (10 ml), followed 1 1 2 by washing of the organic layer with water and saturated sodium bicar bonate solution, drying over sodium sulfate, and removal of solvent in vacuo gave 29 mg (58$) of 2-naphthaldehyde (130). This was identified by pmr comparison with a known sample, and by means of its 2,V-DNP 1 0 0 derivative, mp 269-270.5° (lit mp 270°).

When the duration of reaction was limited to 1 hr, a mixture of two aldehydes was detected by pmr after the above work-up; these were determined to be 151 and 130 in a ratio of 72: 28 as determined by relative areas of their aldehyde absorptions in the pmr spectrum; 31 mg

(62*).

Use of j>-toluene sulfonic acid-di gave identical results and shoved no deuterium incorporation. These two aldehydes could not be separated by tic and although separation could be achieved on a gas chromatograph using column H at 155° (flow rate 100 ml/min), the collected material that would correspond to 151 (‘t|re^. = 9*3 min) was found on a preparative scale to still be a mixture of 131 and 150 ('tre^ = 13*5 min). Rearrangement was apparently taking place in the detector.

Bullvalenylcarboxaldehyde Tosylhydrazone (128 ). A mixture of 316 mg

(2 .0 0 mmol) of 125 and 372 mg (2.00

mmol) of tosylhydrazine in 10 ml of

ethanol was heated on a steam bath IHKNHTs for 20 min. The hot solution was

filtered and allowed to cool to

room temperature. The tosylhydrazone 113

128 crystallized as pale yellow needles; after further cooling in the freezer the needles were collected to give 290 mg (44$) of 128, mp 151° dec; 1355, 1328, 1300, and 1160 cm-1; 7*52 (m, 6h, aromatic, majc iiyo imine and >NH), 4.50 (vb, 9H, bullvalenyl), and 2.40 (s, 3H, methyl).

Anal. Calcd for CiaHiaNpOaS: C, 66.23; H, 5*5&; N, 8.58.

Found: C, 66.6l; H, 5*6l; N, 8.47. pyrolysis of the Sodium Salt of Bullvalenylcarboxaldehyde Tosylhydrazone

(138). A stirred solution of 163 mg (0. 50 mmol) of 128 in 2 ml of

dichloromethane (which had been NH stored over sodium hydroxide pellets)

was treated with 21 mg (l mol equiv)

of 47$ sodium hydride suspension.

Gas evolution started immediately

m and stopped after 3 min. The solvent was removed in vacuo and the pyrolysis apparatus assembled. The system consisted of the reaction flask, a bent adaptor leading into a straight vacuum adaptor, and a receiver. When a vacuum of 0.01 mm Hg had been attained, the apparatus was lowered such that the reaction flask became immersed in a salt bath preheated to 200° and the receiver in a Dry Ice- isopropyl alcohol bath. Within a few seconds bubbling was noted in the reaction flask and deposition of a yellow oil was seen in the cooler portions of the bent adaptor. No change in the pumping speed was noted, indicating that no nitrogen was being evolved. After approximately two minutes no activity was evident in the reaction flask, only a gray solid 114 remained. The yellow oil crystallized upon scratching. After prepara tive thick layer chromatography, there was isolated 34 mg of the pyrazole

1^8 (Rf =* 0 .25, 75$ ether-hexane) as a white solid, mp 83-87°; vBjax 3360 and 3180 cm“H 6^ 13 7*35 (s, IH, imine), 5-80 (m, 2H, olefinic),

4.10 (m, 4h, olefinic ^ cyclopropyl), 3-27 (t, J = 8. 5 cps, IH, cyclo propyl ^ aliphatic), and 3*05 (t, J = 8. 5 cps, IH, cyclopropyl^ ali phatic); calcd for ClltH 10N2 m/e 170.0844, found 170.0847.

Anal. Calcd for (CnHxoNaJs'HaO: C, 74.97; H, 6.10.

Found: C, 74.65; H, 6.27.

Base Induced Decomposition of 128. To a stirred solution of 3 equiv of the base in 5 ml of dry solvent under nitrogen was added one equivalent of 128 in one portion. This mixture was then heated in a 125° oil hath for 10 min. The only observable change was a darkening of the reaction mixture. This was then cooled in ice and rinsed into a separatory funnel with 80 ml of water and 25 ml of ether. The ether layer was washed with water, dried over sodium sulfate, filtered, and evaporated in vacuo. The resulting oil was chromatographed on a shortcolumn of silica gel (elution with 75$ ether/hexane). The only isolated product was the pyrazole 139 • 115

Table X . Decomposition of Bullvalenylcarboxaldehyde Tosylhydrazone.

mmol of 128 Solvent Base $ yield of 139^

0 .8 0 diglyme n-BuLi 37$ 0. k6 ethylene glycol n-BuLi 33$ 0 .8 6 diglyme NaOMe k2$

Treatment of 139 with Mercuric Bromide. A solution of 25 mg (0.15 mmol) of 15g> and 13 mg (O.O36 mmol) of mercuric bromide in 5*0 ml of methanol was stirred at room temperature for 3 days. Solvent was removed in vacuo and the clear residue was chromatographed on a preparative thick layer plate (elution with 75$ ether-petroleum ether) to give a single component of the same as 139. Pmr also showed no other material to be present.

Treatment of 139 with £-Toluenesulfonic Acid. To a solution of 25 mg

(0.15 mmol) of 139 in 5*0 ml of benzene was added 10 mg (0 .053 mmol) of

£-toluenesulfonic acid monohydrate. Water of hydration was removed as the benzene azeotrope and the resulting solution was refluxed for

2k hr, cooled, and added to 20 ml of water and 5 ml of benzene. The organic layer was washed with saturated sodium bicarbonate solution (20 ml) and brine (20 ml), dried over sodium sulfate, and evaporated in vacuo to leave an off-white solid which by pmr was uniquely 139- 116

7-'Carbomethoxybicyclo[4. 2. 2]deca-2,4,.7s9-tetraene (1^2). Hydrolysis of

2. 5 g of the filtrates from the preparation of 124 was accomplished with sodium hydroxide (2 g, 4 mmol) in refluxing methanol-water (1:4) for 16 hr. Acidification of the cooled reaction mixture gave a white precipitate which was collected by suction filtration and air dried to give 1 .3 0 g of acid, esterification of which with ethereal diazomethane in the usual manner gave a clear oil. Treatment of this oil with 0.525 g (l. 5 mmol) of mercuric bromide in 25 ml of methanol for 40 hr at room temperature gave after chromatography on silica gel (elution with 25$ ether-hexane) 450 mg of 1^2 as a clear oil, whose pmr spectrum corres- © © ponded to that reported for 13_2; ^TM3*^3 MHz) 6.63 (d, J = 6.0 Hz,

IH, olefinic a to ester), 5-36-6.31 (m, 6h, olefinic), 3*83 (m, 1H, bridgehead), 3*65 (s, 3H, methyl), and 3-26 (m, IH, bridgehead).

Iteduction and Oxidation of 132^ To an ice-cold solution of 100 mg (0.53 mmol) of 132 in 5 nil of dry ether under argon was added via syringe

O.9 2 ml (1.17 mmol) of di i sobutylalumlnum hydride (26$ by weight, 1.27 mmol/ml) in hexane. This clear solution was stirred at 0° for 1 hr, then treated with 0. 5 ml of methanol and 0.5 ml of water. After 1 hr, the aluminum salts were removed by filtration through Celite and the gel washed with three 20-ml portions of ether. The combined filtrates were dried over sodium sulfate and evaporated in vacuo. The residue was chromatographed on a short Florisil column (1-5 x 8 cm); elution with 20$ ether-carbon tetrachloride gave 133 as a clear oil, 94 mg

(probably containing trapped carbon tetrachloride); 6 ^ ^ 13 5.25-6 .2 0 117

(m, 7H, olefinic), 3*91 (s, 2H, methylene), 2.85-3*^3 (m> 2H, "bridgehead), and 2.88 (bs, 1H, hydroxyl). This oil was transferred to a 10 ml flask and 1 ml of dry dime thylsulfoxide, 1 ml of dry benzene, 358 mg (1.74 mmol) of dicyclohexylcarbodiimide, 60 p,l (0. jk mmol) of pyridine and

30 nl (0.40 mmol) of trifluoroacetic acid were added. This solution was stirred magnetically for 24 hr at room temperature, at which point it was poured into 15 ml of ether to which was added 240 mg (2.67 mmol) of oxalic acid in 6 ml of methanol. After an additional 30 min, water

(10 ml) was introduced, the organic layer was separated and washed sequentially with water and saturated sodium carbonate solution, dried over sodium sulfate and evaporated in vacuo. Thin layer chromatography showed that incomplete oxidation had taken place. Elution with 25$ ether-hexane showed a minor component R^ = 0.63, 131, and a major one

R^ = 0.22, 133. Isolation of l^l by preparative tic gave 9 Mg (11$ from

132) which was spectroscopicly identical with 131 obtained from acid- catalyzed rearrangement of 125. The 2,4-dinitrophenylhydrazine deri vative of 151 prepared by the two routes were also identical by infrared and tic.

Oxidation of 133 with manganese dioxide gave 2-naphthaldehyde.

Bullvalenylcarboxaldehyde Qxime (150). To 0.50 g (7.2 mmol) of hydroxyl- amine hydrochloride dissolved in 3 ml of water was added 2 ml of 10$ sodium hydroxide solution followed by 100 mg (0.63 mmol) of 125 and 5 ml of ethanol. The mixture was heated on a steam bath for 15 min. The solid which was initially present quickly dissolved upon warming. 118

Subsequent cooling deposited a pale

green solid -which was collected by-

.CH=NOH suction filtration, taken up in hot

ethanol, and treated with Norit. To

.the filtrate was added 5 ml of water

and the total volume was reduced to

10 ml on a steam bath. Slow cooling yielded 98 mg (85$) of 150 as white needles. The analytical sample was obtained by two recrystallizations from ethanol, mp 184-185°:, v1®17 3230, 1620, 1290, 972, 953 , 938, and max 824 cm-1-, 5^ 3 ^ 0 7.41 (s, 1H, imine), 4.38 (vb, 9H, bullvalenyl), and

2 .7 8 (bs, 1H, hydroxyl).

Anal. Calcd for C XiHlaN0: C, 76.27; H, 6.40; N, 8.09.

Found: C, 76.02; H, 6.44-, N, 7-82.

Reduction of 150 with Lithium Aluminum Hydride. To an ice-cold slurry of 30 mg (0 . 80 mmol, 4.1 equiv) of lithium aluminum hydride in 5 ml of dry tetrahydrofuran under nitrogen was introduced via syringe 68 mg (0.39 mmol) of 150 in 2 ml of dry tetrahydrofuran dropwise over 2 min. Gas was evolved after each drop. After addition was completed, the now yellowish solution was stirred at ice-bath temperature for 15 min. The slightly cloudy reaction mixture was then heated to reflux; more preci pitate formed and the solution became yellow. After a few minutes at reflux the precipitate dissolved and another reappeared. The cooled reaction mixture was quenched after 2 hr at reflux with 0 .5 ml of 10$ sodium hydroxide solution and 0. 5 ml water. The THF solution was separated from the aluminum salts, dried over magnesium sulfate, and evaporated in vacuo to yield a brown oil which by tic was a continuum

(no distinct spots). The pmr showed that the bullvalene system was no longer present. The same result was obtained by using 4.4 equivalents of diisobutylaluminum hydride at 50° for 12 hr. These reactions were not investigated further.

Bullvalenylcarbinol (126). A mixture of 510 mg (3*22 mmol) of 125 , 200

mg (5-0 mmol) of sodium hydroxide,

and 1.46 g (38*8 mmol) of sodium

borohydride in 50 ml of ethanol c h 2oh was stirred for 24 hr at room tempera

ture and then added to 500 ml of

water and 100 ml of ether. The resulting emulsion was broken using salt flakes, whereupon the ether layer was separated, washed with water and brine, and dried over magne sium sulfate. After removal of the ether in vacuo, the resulting yellow oil was chromatographed on silica gel (l.5 x 10 cm) by elution with 50$ ether-hexane to give 428 mg (83$) of 126 as a pale yellow oil. The analytical sample was obtained by preparative vpc on column F at 170°

(flow rate 150 ml/min, tyet = 9-5 min, only peak); ^*55 (vbs, 9H, bullvalenyl), 4.02 (s, 2H, methylene), and 1.88 (ba, IH, hydroxyl); v1®1* 532O and 1015 cuT1. max Anal. Calcd for CnHisO: C, 82.46; H, 7*55*

Found: C, 82.32; H, 7*73* 1 2 0

Bullvalenylmethyl j>-Anisoate (147). To a solution of 158 mg (0.99 mmol)

of 126 in 3 ml of 2,6-lutidine was

added 171 mg (l.00 mmol) of £-

anisoyl chloride in 2 ml of 2,6-

lutidine. A precipitate formed

immediately and the mixture was

placed in the refrigerator for 18 hr.

The orange reaction mixture was added to 40 g of ice and water and the precipitate was collected by suction filtration, washed with JO ml of cold water, and dried in air. Preparative tic (elution with 75$ ether- petroleum ether) yielded (R^. = 0.66) 151 mg (52$) of 147 as an off-white solid. The other component (R^. = 0.50) was identified as p-anisoyl anhydride. Recrystallization from hexane gave pure 147 as white prisms, CDC1; 1720, 1610, 1275, 1260, 1710, and 1110 cm-1; 6 max TMS and 6.83 (M'BB'j 4h, aromatic), 4.59 (s, 2H, methylene), 4.30 (v'b, 9H, bullvalenyl), and 3 .8 0 (s, 3H, methyl).

Anal. Calcd for C 19H 18O3: C, 77-53» H, 6.16.

Found: C, 77-45; H, 6.23.

Hydrolysis of 147. A sealed tube was prepared using glass tubing 3/ b "

0. D., 2.0 mm walls, and 30 cm long.

The tubing was sealed at one end, HO ", charged with 107 mg (0 .3 8 mmol) of

V l4j and 1 .0 ml of 70:30 (v/v) acetone:

water and then sealed 20 cm from the 145 bottom under vacuum while cooled with 1 2 1

a Dry Ice-isopropyl alcohol hath. The sealed tube was allowed to warm

to room temperature then totally immersed in an air bath at 125°.

After 10 min the tube was removed and inverted a few times to dissolve

the molten ester at the bottom. The tube was then replaced and heated

for 2k hr. After cooling in a Dry Ice-isopropyl alcohol bath the tube was opened and the yellowish reaction mixture was added to 40 ml of water and extracted with ether (2 x 20 ml). The combined ether layers were washed with water (l x 30 nil) and saturated sodium carbonate solu

tion (l x 30 ml), dried over sodium sulfate, and evaporated in vacuo to yield a yellow oil which after preparative tic was separated into its two components: R^. =* 0. 73 > 24 mg, recovered 147; R^ * 1.40, 38 mgj rear ranged alcohol 145, 85$ based on recovered 147, white needles recry

stallized from pentane, mp 77-77* 5°; 3380 and 1015 cm-1; IT1£LX ^TMR13 ^ •00 (m> olefinic), 5*00 (d, J = 2.0 Hz, 1H, exo methylene),

4.78 (d, J = 2.0 Hz, 1H, exo methylene), 4.11 (m, 1H, >CH-0-), 3*32 (m,

2H, bridgehead), and 1.9 8 (bs, 1H, hydroxyl); calcd m/e l60.0888; found,

I6O.O890.

Attempted Preparation of 129. (R=Ts). To a magnetically stirred solution of 212 mg (1 .3 2 mmol) of 126 in 5*0 ml of dry tetrahydrofuran under nitrogen at 0° was added 0. 63 ml of 2.10 M n-butyllithium (l equiv).

The clear solution yellowed upon addition of the n-butyllithium; no other change was noted after 0 .5 hr. To this was added 250 mg (1.32 mmol) of jd-toluenesulfonyl chloride. After 0.5 hr, solvent was removed in vacuo, 2 ml of dichloromethane was added, and the resulting suspension filtered. Only one spot was seen by tic (R^. = 0..81, 50$ ether-hexane).

Isolation by preparative tic gave lft-5 (52 mg, 2^$). REFERENCES

1. (a) W. von E. Doering, Zh. Vses. Khim. Obshchest., 7, 308 (1962); (b) W. von E. Doering and W. R. Roth, Tetrahedron, 1^, 715 (1963).

2. (a) G. Schroder and J. F.M. Oth, Angew. Chem. Inter. Ed. Engl., 6, 4l4 (1967); (b) G. Schroder, J. F. M. Oth, and R. Mer^nyi, ibid., 4, 752 (1965); (c) L. A. Paquette, ibid., 10, 11 (1971).

3 . (a) H. E. Zimmerman and G. L. Grunewald, J. Amer. Chem. Soc. , 88, 183 (1966); (b) H. E. Zimmerman, R. W. Binkley, R. S. Givens, G. L. Grunewald, and M. A. Sherwin, ibid. , gl, 331^ (1969)*

4. A. K. Cheng, F. A. L. Anet, J. Mioduski, and J. Meinwald, J. Amer. Chem. Soc. , §6, 2887 (1974).

5* (a) M. J. S. Dewar and W. W. Schoeller, ibid., 93, 1481 (1971); (b) M. J. S. Dewar and D. H. Lo, ibid. , 23, 7201 (1971).

6. G. Schroder, Angew. Chem., J5j 722 (1963).

7. G. Schroder, Chem. Ber., 97, 3140 (1964).

8. W. von E. Doering, B. M. Ferrier, E. T. Fossel, J. H. Hartenstein, M. Jones, Jr., G. Klumpp, R. M. Rubin, and M. Saunders, Tetrahedron, 22, 3943 (1967).

9. J. Font, F. Lopez, and F. Serratosa, Tetrahedron Lett., 2589 (1972).

10. L. T. Scott and Maitland Jones, Jr., Chem. Rev., 72, l8l (1972).

11. M. Jones, Jr. , and L. T. Scott, J. Amer. Chem. Soc. , 89, 150 (1967).

12. M. Jones, Jr., S. D. Reich, and L. T. Scott, ibid., 9 2, 3118 (1970).

13* W. von E. Doering and J. W. Rosenthal, ibid., 88, 2078 (1966).

14. Idem., Tetrahedron Lett., 349 (1967).

15. M. Jones, Jr., J. Amer. Chem. Soc. , 89, 4236 (1967)*

16. G. Schroder, R. Merenyi, and J. F. M. Oth, Tetrahedron Lett., 773 (1964).

123 124

17- J. F. M. Oth, R. Merenyi, J. Nielsen, and G. Schroder, Chem, Ber., §8, 3585 (1965).

18. J. F. M. Oth, R. Merenyi, H. Rottele, and G. Schroder, Tetrahedron Lett., 3941 (1968).

19. J. F. M. Oth, E. Machens, H. Rottele, and G. Schroder, Justus Leibigs Ann. Chem. , 743, 112 (1971).

20. G. Hoogzand, J. Nielsen, and J. F. M. Oth, Tetrahedron Lett., 2287 (1970).

21. U. Kruerke, Angew. Chem. Inter. Ed. Eng. , 6, 79 (1967).

22. E. Vogel, W. Grimme, W. Meckel, H. J. Riebel, and J. F. M. Oth, ibid., 8, 590 (1966).

23. H. Rottele, P. Nikoloff, J. F. M. Oth, and G. Schroder, Chem. Ber. , 102, 3367 (1969).

24. H. E. Zimmerman, R. W. Binkley, R. S. Givens, andM. S. Sherwin, J. Amer. Chem. Soc., 8 9, 3932 (1967).

25. (a) H. E. Zimmerman and H. Iwamura, ibid., 90, 4763 (1968); (b) Idem. , ibid. , 92, 2015 (1970).

26. (a) J. Meinwald and D. Schmidt, ibid., gL, 5877 (1969); (b) J. Meinwald and H. Tsuruta, ibid., ffl, 5877 (1969).

27. H. E. Zimmerman, J. D. Roberts, and J. Schantl, ibid. , 91, 5878 (1969). “

28. (a) L. A. Paquette, ibid., 92, 5765 (1970); (b) R. Askani, Tetrahedron Lett. , 3349 (1970); (c) R. M. Moriarty, C.-L. Yeh, and N. Ishibi, J. Amer. Chem. Soc., 93, 3085 (1971).

29. R. Malherbe, Helv. Chim. Acta, 56, 2845 (1973).

30. R. Criegee and R. Askani, Angew. Chem. Inter. Ed. Eng., 7, 537 (1968).

31. R. Askani, Tetrahedron Lett., 447 (1971).

32. E. Ciganek, J. Amer. Chem. Soc. , 88, 2882 (1966).

33. K. E. Pfitzner and G. G. Moffatt, ibid., 8 7, 5670 (1965).

34. Sadtler infrared spectrum no. 1191- 125

35* B. Deubzer, E. 0. Fischer, H. P. Fritz, C. G. Kreiter, N. Krie- bitzsch, H. D. Simmons, Jr., and B. R. Willeford, Jr., Chem. Ber., loo, 308U (1967).

3 6. H. C. Brown and P. Heim, J. Amer. Chem. Soc., 86, 3566 (196^).

37. (a) B. R. Ree and J. C. Martin, ibid., 91, 5882 (1969); 92, 1660 (1970); (b) P. von R. Schleyer and V. Buss, ibid., 9,1, 588*0 (1969); (c) R. S. Bly and R. T. Swindell, J. Org. Chem., 3 0 , 10 (1965), and references cited therein.

38. (a) R. L. Bixler and C. Niemann, J. Org. Chem. , 23, 7^2 (1958); (b) C. A. Grob, M. Ohta, E. Renk, and A. Weiss, Helv. Chim. Acta, 4l, 1191 (1958)', (c) C. A. Grob, R. M. Hoegerle, and M. Ohta, ibid. , b_5, 1823 (1962)-, (d) W. G. Dauben and J. B. Rogan, J. Amer. Chem. Soc., 79, 5002 (1957); (e) S. J. Cristol and D. K. Pennelle, J. Org. Chem., £5 , 2357 (1970).

39* (a) In the acetolysis of bicyclo[2.1.l]hexane-1-methyl tosylate, 2$ of unrearranged acetate was observed: K. 3. Wiberg and B. R. Lowry, J. Amer. Chem. Soc., 85, 3188 (1963); (b) A true carbonium ion equilibrium has been demonstrated in the adamantylcarbinyl- homoadamantyl interconversion: J. E. Nordlander, S. P. Jindal, P. von R. Schleyer, R. C. Fort, Jr., J. J. Harper, and R. D. Nicholas, ibid., 88, kbT5 (1966)*, (c) The hydrolysis of norbornyl- 1-carbinyl and norbornenyl-l-carbinyl tosylates in acetone-water (60:40) lacking collidine leads exclusively to ring expanded pro ducts. In contrast, when collidine is present, 8.5$ and k2$ un rearranged alcohols, respectively, are produced; J. W. Wilt, C. T. Parsons, C. A. Schneider, D. G. Schultenover, and W. J. Wagner, J. Org. Chem., 33_> ^9^ (1968). ho. L. Friedman and J. H. Bayless, J. Amer. Chem. Soc., 91, 1790 (1969) •

4l. J. A. Smith, H. Shechter, J. Bayless, and L. Friedman, ibid., 8 7, 659 (1965). h2. 1,2-Alkyl migrations of a related type have been noted previously: R. A. Moss and J. R. Whittle, Chem. Commun. , 3^1 (1969), M. H. Fisch and H. D. Pierce, Jr., ibid., 505 (1970); R* D« Allan and R. J. Wells, Austr. J. Chem., 2£, 1625 (1970).

43. See, for example: J. R. Wiseman, J. Amer. Chem. Soc., 89, 5966 (1967); J> R. Wiseman, H.-F. Chan, and C. J. Ahola, ibid., 91, 2812 (1969); J. A. Marshall and H. Faubl, ibid., 8 9, 5965 (19^7)i 92, 9^8 (1970); J. R. Wiseman and W. A. Pletcher, ibid., 92, 956 Tl970); G. Kobrich and H. Heinemann, Chem. Commun., h93 (I969X” 126

44. A. R. Kraska, L. I. Cherney, C. G. Moseley, G. M. Kaufman, and H. Shechter, private communication.

45. In the present instance, a two-fold statistical advantage for phenyl migration needs also to be considered.

46. L. A. Paquette and G. V. Meehan, J. Amer. Chem. Soc. , 92, 3039 (1970).

k j. J. W. Powell and M. C. Whiting, Tetrahedron, 7, 305 (1959)» J» W. Wilt, C. A. Schneider, H. A. Dabek, Jr. , J. FT Kraemer, and W. J. Wagner, J. Org. Chem., 31, 1543 (1966).

48. K. B. Wiberg and J. M. Lavanish, J. Amer. Chem. Soc. , 8 8, 5272 (1966); F. Cook, H. Shechter, J. Bayless, L. Friedman, R. L. Foltz, and R. Randall, ibid., 8 8, 3870 (1966).

- 49. H. Zollinger, ’'Azo and Diazo Chemistry, ' 1 Interscience Publishers, New York, N.Y. , 1961, p. 45-

50. Y. E. Rhodes and T. Takino, J. Amer. Chem. Soc. , 9,2, 5270 (1970).

51. Similar rationale has been employed to explain the lack of rear rangement of: (a).the 2-adamantyl cation to the more stable 1- adamantyl cation on acetolysis [P. von R. Schleyer and R. D. Nicholas, J. Amer. Chem. Soc., 83, 182, 2700 (196I)]; (b) the behavior of l-arainodibenzobicycXo[2. 2.2]octadiene on deamination [W. R. Benson, Ph.D. Thesis, University of Colorado, 1958]9 (c) the like behavior of 1-aminomethyltriptycene. 38e

52. Determined by Professor Jon Clardy, Iowa State University, Ames, Iowa.

53. H.-P. Loffler, Chem. Ber. , 104, 1981 (1971).

54. (a) P. Yates, E. S. Hand, and G. B. French, J. Amer. Chem. Soc., 82 , 6357 (I960)-, (b) J. D. Roberts and W. F. Gorham, ibid. , 2278 (1952)* (c) W. von E. Doering and W. R. Roth, Tetrahedron, 1£, 715 (I963)i (d) J. E. Baldwin and M. S. Kaplan, J. Amer. Chem. Soc. , 93, 3969 (1971).

55. H.-P. Loffler and G. Schroder, Chem. Ber., 103, 2105 (1970).

56. R. Askani and H. Sonmez, Tetrahedron Lett. , 1751 (1973)-

57. A. B. Evin, R. D. Miller, and G. R. Evanega, Tetrahedron Lett., 5863 (1968). 127

58. R. Huisgen, W. E. Konz, and G. E. Green, J. Amer. Chem. Soc., 92, 4105 (1970).

59* R> Huisgen, W. E. Konz, and U. Schnegg, Angew. Chem. Int. Ed. Engl., 11, 715 (1972).

60. (a) W. Rappe, 0. Schlichting, K. Klager, and T. Toepel, Ann., 560, 1 (1908); (b) M. Avram, E. Sliam, and C. D. Nenitzescu, ibid., 184 (i960); (c) D. G. F a m u m and J. P. Snyder, Tetrahedron Lett., 5861 (1965).

61. J. P. Snyder, Hi. D. Thesis, Cornell University, Ithaca, N. Y., 1965; Dissert. Abstr. , 26, 5728 (1966). We thank Professor Farnum for making available to us the details of their synthesis of 86 in advance of publication.

62. In the later stages of thie study, the finding was made that a change in our earlier conditions of pure acetone solvent and Vycor optics to acetone-benzene and a Corex filter system was particularly advantageous in the acetoxymethyl series. We have not reinvesti gated the other derivatives described herein, but point out to those interested in preparing such compounds in the future that yields can probably be increased by making this changeover and also by decreasing still further the concentration levels of diene utilized by us.

65. R. Huisgen, W. E. Konz, and G. E. Gream, J. Amer. Chem. Soc., 92, 4105 (1970).

64. R. Huisgen, W. E. Konz, and U. Schnegg, Angew. Chem. Int. Ed. Engl. , 11, 715 (1972).

65. For a review see L. A. Paquette, Accounts of Chem. Res., 4, 280 (1971).

66. M. Heyman, V. T. Bandurco, and J. P. Snyder, JCS Chem. Commun. , 297 (1971).

67. R. E. Kelley, J. Org. Chem. , 28, 453 (1965)* R- B. Kelley, G. R. Umbreit, and W. R. Liggett, ibid., 29, 1275 (1964).

68. E. F. Pratt and T. P. McGovern, ibid., 29, 1540 (1964).

69. W. W. Wood, W. Fickett, and J. G. Kirkwood, J. Chem. Riys. , 20, 561 (1952).

70. H. Joshua, R. Gans, and K. Mislow, J. Amer. Chem. Soc. , 90, 4884 (1968). We especially thank Professor Mislow for a copy of the computer program. 128

71- L. A. Paquette, S. Kirschner, and J. R. Malpass, ibid. , 91, 3970 (19 69); 92, 4330 (1970).

72. For application of the same principle to conformational analysis, see (a) E. L. Eliel, Chem. Ind. (London), 78 (1959); (6) N. S. Bhacca and D. H. Williams, ''Applications of Nmr Spectroscopy in Organic Chemistry,'1 Holden-Day, Inc., San Francisco, Calif., 1964, pp 153-154; see also (c) J. A. Pople, W. G. Schneider, and H. J. Bernstein, ''High-Resolution Nuclear Magnetic Resonance,'' McGraw-Hill Book Co., Inc. New York, N. Y., 1959, Chapter 10.

73* D. R. James, personal communication.

74. R. Hoffmann and W.-D. Stohrer, J. Amer. Chem. Soc., 93 5 6941 (1971).

75- K. E. Pfitzner and J. G. Moffatt, ibid., 87, 5661 (1965).

76. (a) L. A. Paquette, K. H. Fuhr, S. Porter, and J. Clardy, J. Org. Chem. , 39 , 467 (1974)-, (b) H. Gopal, T. Adams, and R. M. Moriarty, Tetrahedron, 28, 4259 (1972)', (c) R. M. Moriarty, H. Gopal, and T. Adams, Tetrahedron Lett. , 4003 (1970); (d) H. Nakata, Tetrahe dron, 1£, 1959 (1963).

77* J. R. Parikh and W. von E. Doering, J. Amer. Chem. Soc., 8 9, 5505 (1967).

78. For a summary of various modifications, consult L. F. Fieser and M. Fieser, ''Reagents for Organic Synthesis,'' Vol. I, p 636, J. Wiley and Sons, Inc., New York, N. Y., 1967*

79* E. J. Corey and C. U. Kim, J. Amer. Chem.Soc., 94, 7587 (1972).

80. M. Fetizon and M. Golfier, Compt. Rend., 267, 900 (1968).

81. R. K. Crossland and K. L. Servis, J. Org. Chem. , 35, 3195 (1970).

82. (a) S. Winstein, M. Oharurso, M. Sakai, and J. H. Nicholson, J. Amer. Chem. Soc., 89, 3656 (1967); (b) W. R. Moore, W. R. Moser, and J. E. LaProde, J. Org. Chem., 28, 2200 (1963).

8 3. G. W. Klumpp, G. Ellen, and F. Bickelhaupt, Rec. Trav. Chim. Pays-Bas, 88, 474 (1969) and references cited therein.

84. A. Orahovats, M. Tichy, and J. Sicher, Collect. Czech. Chem. Comm., 35, 838 (1970).

8 5. H. 0. House and W. F. Fischer, Jr., J. Org. Chem., 34, 3626 (1967). 129

86. H.-P. Loffler and G. Schroder, Angew. Chem. Int. Ed. Eng., 7, 736 (1968).

87. See for example, G. Schroder, U. Prange, B. Putze, J. Ohio, and J. F. M. Oth, Chem. Ber. , 10j+, 3406 (1971).

88. R. Merenyi, J. F. M. Oth, and G. Schroder, Chem. Ber. , 97, 3150 (196*0 .

89. (a) D. W. Adamson and J. Kenner, J. Chem. Soc. , 286 (1935); (*>) G. L. Closs and W. A. Boll, Angew. Chem. Int. Ed. Eng., 2, 399 (1963); (c) G. L. Closs, L. E. Closs, and W. A. Boll, J.~*Amer. Chem. Soc. , 8 5, 3796 (1963); (d) R. K. Bertlett and T. S. Stevens, J. Chem. Soc. C , 1964 (1967); (e) C. D. Hurd and S. C. Lui, J. Amer. Chem. Soc. , 5J7, 2656 (1935); (f) D. Y. Curtin and S. M. Gerber, ibid. , 7^7 4052 (1952).

90. H. Shechter and D. Sanders, private communication.

91- M. R. Willcot, J. F. M. Oth, J. Ihio, G. Plinke, and G. SchrSder, Tetrahedron Lett. , 1579 (1971).

92. The Chemical Abstracts name for 413 is methyl 5-dibenzotricyclo- [3 .3 .0.02,8]octadienylcarboxylate. Throughout, the dibenzosemibullvalenyl nomenclature is employed for convenience.

95- H. G. Lanthan, Jr. , E. L. May, and E. Mosettig, J. Amer. Chem. Soc., 70, 1079 (1948).

94. R. C. Parish and L. M. Stock, J. Org. Chem. , 30, 927 (1965); R- 0. C. Norman and P. D. Ralph, J. Chem. Soc. , 2S21 (1961).

95- We thank Dr. Ciganek for providing us with details of his procedure in advance of publication.

96. Prepared by treating 38 g of zinc powder with three 100-ml portions of 256 aqueous copper sulfate solution, and washing with 5 0 0 ml of ethanol.

97. L. A. Bock, Ph.D. Thesis, University of California at Los Angeles, 1969.

98. A. C. Cope and D. F. Rugen, J. Amer. Chem. Soc. , 75, 3215 (l953)>

99. K. A. Henzel, personal communication. 100. R. L. Shriner, R. C. Fuson, and R. Y. Curtin, ''The Systematic Identification of Organic Compounds, ' ’ 5th ed., p 321, John Wiley and Sons, Inc., New York, N. Y. (1967)-