77-2525

VOLZ, William Edward, 1947- STRAINED ALICYCLIC HYDROCARBONS: SYNTHESIS, CHLOROSULFONYL ISOCYANATE ADDITIONS AND ELECTRONIC PROPERTIES.

The Ohio State University, Ph.D., 1976 Chemistry, organic

Xerox University Microfilms,Ann Arbor, Michigan 48106 STRAINED ALICYCLIC HYDROCARBONS: SYNTHESIS, CHLOROSULFONYL

ISOCYANATE ADDITIONS AND ELECTRONIC PROPERTIES

DISSERTATION

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

by

William Edward Volz, B.S.

*****

The Ohio State University

1976

... Reading Committee: Approved by

Dr. Leo A. Paquette Dr. Gideon Fraenkel Dr. John S. Swenton

Adviser Department of Chemistry DEDICATION

Monica, Tracy, and ACKNOWLEDGMENTS

I am truly grateful to Professor Leo A. Paquette for his direction, encouragement and enthusiasm during this -work. The stimulating atmo­ sphere provided by the research group he has attracted is something that

I am very proud to have experienced.

I wish to thank my friend Mike Geekle for all of the nmr work he did for me. I especially value the companionship he gave me and my family.

The photoelectron spectra and related molecular orbital calculations are courtesy of Professor Rolf Gleiter and Dr. Peter Bischof at the Tech- nischen Hochschule Damstadt, West Germany.

I sincerely appreciate the sacrifices Monica, Tracy and Eric made to my education, and the confidence Monica, my parents and my parents-in-law had in me.

I would also like to acknowledge The Ohio State University and the people of Ohio for their financial support. VITA

October 15, 1947 ...... B o m - Des Moines, Iowa

1970 ...... B. S., The Iowa State University, Ames, Iowa

1970-1974...... Teaching Assistant, The Ohio State University, Columbus, Ohio

1974-1975 ...... Research Assistant, The Ohio State University, Columbus, Ohio

PUBLICATIONS

T. J. Barton, W. E. Volz, and J. L. Johnson, J. Org. Chem., 56? 5565 (1971).

W. E. Volz, L. A. Paquette, R. J. Rogido, and T. J. Barton, Chem. Ind. (London), 771 (1974).

L. A. Paquette, W. E. Volz, M. A. Beno, and G. G. Christoph, J. Am. Chem. Soc., 97, 2562 (1975).

W. E. Volz and L. A. Paquette, J. Org. Chem., 4l, 57 (1976).

L. A. Paquette and W. E. Volz, J. Am. Chem. Soc., 9 8, 2910 (1976 ).

R. Gleiter, P. Bischof, W. E. Volz, and L. A. Paquette, J. Am. Chem. Soc., in press.

iv TABLE OF CONTENTS

Page

DEDICATION...... ii

ACKNOWLEDGMENTS ...... iii

VITA ...... iv

LIST OF FIGURES...... *...... vii

LIST OF TABLES...... viii

SECTION I - Reactions of Chlorosulfonyl Isocyanate with Homo- and Benzhomobarrelenes ...... 1

INTRODUCTION...... 2

RESULTS...... 20

DISCUSSION...... k-3

SECTION II - Reaction of Chlorosulfonyl Isocyanate with Barrelene, Benzobarrelene, and Dibenzobarrelene .... 54

INTRODUCTION ...... 55

RESULTS...... 59

DISCUSSION...... 72

SECTION III - Synthesis of 1-Cyanosemibullvalenes via Chlorocyanation of Barrelene, Benzobarrelene, and Dibenzobarrelene...... 83

INTRODUCTION...... 84

RESULTS...... 90

DISCUSSION...... 102

v Page

SECTION IV - Reactions of Chlorosulfonyl Isocyanate with Bicyclopentane and 1,3 -Dimethylbicyclobutane...... 113

INTRODUCTION...... Il4

RESULTS...... 118

DISCUSSION...... 129

SECTION V - Synthesis and Electronic Properties of Bicyclo- [4.1.1]octane, Bicyclo[4.1.l]oct-3-ene, and Bicyclo[i)-. 1. l]octa-2,^-diene...... 135

INTRODUCTION...... 136

RESULTS...... 153

DISCUSSION...... 185

EXPERIMENTAL...... 203

REFERENCES...... 275

vi LIST OF FIGURES

Figure Page

1 A lateral display of bond lengths and bond angles of 1-cyanosemibullvalene (190a) ...... 93

2 A ventral display of bond angles of 1-cyanosemi- bullvalene (igOa) ...... 94

3 Hie 60 MHz XH nmr spectrum (upfield region) of 2-aza- 1 .4-dimethyl-3 -oxo-bicyclo[2.2.0]hexane (229) ...... 128

4 A diagram of the symmetry elements for bicyclo- [4.1.1]octa-2,4-diene (255) ...... 149

5 The 60 MHz % nmr spectrum of bicyclo[4.1. l]octa- 2.4-diene (233) in CDC13 ...... 167

6 Hie ultraviolet spectrum of bicyclo[4.1.l]octa- 2.4-diene (233) in hexane...... VJO

7 The photoelectron spectrum of bicyclo[4.1.l]octa- 2.4-diene (255.) ...... 174

8 The photoelectron spectrum of bicyclo[4.1.1]oct-3- ene (254) ...... 175

9 The photoelectron spectrum of bicyclo[4.1.1]octane (253) ...... 175

10 Orbital interaction diagram for bicyclo[4.1. l]octa- 2.4-diene (255) based upon perturbation theory ...... l8l

11 An illustration of the Fletcher-Powell computer optimized structures for 253 ? 254, and 255 ...... 183

vii LIST OF TABLES

Table Page

I Relative Rates for the Cyeloaddition of Chloro­ sulfonyl Isocyanate to Various Olefins in Dichloro- methane Solution corrected to 25° ...... 5

II Relative Rates for the Reaction of Chlorosulfonyl Isocyanate with 2-Ethylhexene-l in Solvents of Varying Polarity at 2 5 ° ...... 5

III TO eV Mass Spectroscopic Fragmentation Patterns of Tricyclo[3.2.2.02’4 ]nona-6,8-diene (45) and anti-6,7- Dideuteriotricyclo[3.2. 2.02’ 4 ]nona-67B~-diene (50) ...... 23

IV Product Distribution After Chromatographic Separation of the CSI-Homobarrelene Adducts Formed at Various Reaction Times in Dichloromethane Solution (25°) ...... 24

V Eu(dpm) 3 Induced Chemical Shifts in the NMR Spectra of exo,anti-3-Aza-4-oxoquadricyclo[4.3« 2.02j5.07’9]- undec-10-ene (52) ...... 26

VI Eu(fod) 3 Induced Chemical Shifts in the NMR Spectra of endo,anti-3-Aza-4-oxoquadricyclo[4.3.2.02?4.07j9]- undec-10-ene (53 ) ...... 29

VII Eu(fod) 3 Induced Chemical Shifts in the NMR Spectra of syn(n)-4-Aza-5-oxopentacyclo[8.1.0.02}s.O3 ’8.07>9]- undecane (54) ...... 31

VIII Eu(fod) 3 Induced Chemical Shifts in the NMR Spectra of exo,anti-3-Aza-10,ll-benzo-4-oxoquadricyclo- [4.3.2.02’5.07 >9]undec-10-ene (58) ...... 37

XX Eu(fod) 3 Induced Chemical Shifts in the NMR Spectra of endo,anti-3-Aza-10,ll-benzo-4-oxoquadricyclo- [4.3.2.02’5.07’9]undec-10-ene (59) ...... 40

X Eu(fod)3 Induced Chemical Shifts in the NMR Spectra of exo-3-Aza-T,8-benzo-4-oxotricyclo[4. 2.2.02’5]deca- 7,9-diene (ill) ...... 64

viii Table Page

XI 100 MHz N M R Data from Double Resonance Experiments for 8-Aza-2,3-benzo-9-oxotricyclo[5.3* 0 . 10]deca-2, 5- diene (112) ...... 67

XII 70 eV Mass Spectroscopy Fragmentation Pattern of 7,8- Benzo-anti-2-carbamyl-6-hydroxytricyclo[2.2.2.03 ’ 5 ] - oct-7-ene (114) ...... 88

XIII Variable Temperature 1H NMR Data (100 MHz) for 1(5)- Cyanos emibullvalene (190a 2 190b) in CD2C12-CF2C12 ..... 98

XIV Variable Temperature 1H NMR Data (100 MHz) for 1(5)- Cyanosemibullvalene (190a £ 190b) in C12C=CC12 . 98

XV -^H NMR Coupling Constant Data for Chloro Nitriles 191, 192, and 193,...... 98

XVI 13C NMR Chemical Shifts for Cyanosemibullvalenes llg, 190, and 19ft ...... lo2<-

XVII Bond Lengths in Cyclopropane Derivatives ...... 105

XVIII Estimated Equilibrium constants and Free Energy Differences for 1 (5 )-Substituted Semibullvalenes (174 a 5* 174 b) Based on Chemical Shifts of H4, H6 and C4, C6 Relative to 4H and 1SNMR Data for Semi- bullvalene (174, R=H) at -l60°...... 110

XIX Eu(fod) 3 Induced Chemical Shifts in the NMR Spectra of 6-Aza-7-oxobicyclo[3.2.0]heptane (225) Resulting from the Addition of Chlorosulfonyl Isocyanate to Bicyclopentane (224) ...... 120

XX Eu(fod) 3 Induced Chemical Shifts in the NMR Spectra of 6-Aza-7-oxobicyclo[3*2.0]heptane (225) Resulting from the Addition of Chlorosulfonyl Isocyanate to Cyclopentene (227) 122

XXI Eu(fod) 3 Induced Chemical Shifts in the XH NMR Spectra of 2-Aza-l,4-dimethyl-3-oxobicyclo[2. 2.03- hexane (235 ) ...... 126

XXII Data from the 13C NMR Spectra of Bicyclo[4.1.1]- oetane (253.), Bicyclo[4.1. l]oct-3-ene (254), and Bicyclo[4.1.l]octa-2,4-diene (255) ...... 185

ix Table Page

XXIII Chemical Shifts and Coupling Constants from the 1H NMR Spectra and Known or Estimated Dihedral Angles Between the Planes of the Double Bonds for some Cyclic, Bicyclic, and Tricyclic 1,3-Dienes...... 168

XXIV Data from the Ultraviolet Spectra of some Cyclic, Bicyclic, Tricyclic, and Spirocyclic Conjugated Dienes...... 171

XXV Comparisons of the Vertical Ionization Potentials and Calculated Orbital Energies of 2.55, 25fe, and 253 • • • 176

XXVI Comparison of the Carbon-Carbon Intemuclear Distances for Planar 25^ and 255 Derived from the Data in Figure 11 and Table XXVII...... 188

XXVII Comparison of the Bond and Cyclobutyl Dihedral Angles of the Carbon Skeletons of Planar 25jj- and 255 Derived from the Data in Figure 11 and Table XX V I 189

x SECTION I

Reactions of Chlorosulfonyl Isocyanate with

Homo- and Benzhomobarrelenes

1 INTRODUCTION

Since the synthesis of chlorosulfonyl isocyanate (CSI, l) by i Graf in 1956, its powerful electrophilic reactivity and remarkable propensity for cycloaddition reactions has seen frequent application.

0 II Cl — S-N=C=0 II 0

2 Graf suggested for these reactions a two-step mechanism involving the formation of a 1,4—dipolar intermediate (2) which could lead to either type of observed product via ring closure to give N-chlorosulfonyl 3-

CH3 I CSI + H2C=C(CH3 )2 * ch3 -c-ch2 + I / C = 0 -N I S02C1

Type I / ~ 's^fy-pe II CHa / > CH3 I I ch2=c-ch2 -c=o ch3 -c-ch2 I I I HNS02C1 N-C=0 I 3 S02C1

2 lactam (^) or proton transfer to give f3,Y-unsaturated N-chlorosulfonyl amide (3 ). The ratio of the Type I to Type II products for simple olefins does not change with solvent or reaction temperature but does vary with substrate. The g-lactam or Type II product usually pre­ dominates.

The high electrophilicity of the carbon atom of CSI is the result of the strong electron withdrawing ability of the chlorosulfonyl group attached to the cumulative double bond. This polar functionality can also stabilize the developing negative charge at nitrogen in the transition state (5) of a ''near concerted'' mechanism as well as in

8 + 1 1 — (J — 0 — I i I I 8 . ' ' N— C. C1S02X 0

5.

dipolar intermediate 2. The valence bond resonance structures of CSI point up this electrophilic character.

0 0 II + - II + 1 C1-S~N = C -0 -f »- Cl-S —N = C — 0 II II 0 0

la lb

3-14 The Markovnikov nature of CSI additions to olefins and conju­ gated dienes, the ordering of relative rates (Table i), and the en- 4 hancement of rate with increasing solvent polarity (Table II) combine

to suggest development of positive charge on the olefinic carbon and/or

at least some dipolar character in the transition state. Additions 9 ,16-18 of CSI to conjugated dienes support the proposition of dipolar

charge development which in such systems would involve stabilized

allylie carbonium ions. For example, the Markovnikov-oriented 1,2 7 addition product (6 ) of CSI and isoprene at -10° in ether is unstable

to purification but can be isolated at -65° in 8c$> yield. Warming 6

CH3 ch2=ch~

.N. ^ R' CISOs 0

6 Ta, R = KS02C1 8a, R'= S02C1 Tb, R = 0 8b, R'= H

to 40° in ether for 1 hour gave Ta_ which on hydrolysis gave the stable

lactone Tb. Prolonged heating gave the a,3-double bond isomer of 7a.

Hydrolysis of this product mixture gave the latter lactone and 8b.

Moriconi and Meyer suggested a dipolar intermediate 9 and/or a near-

9 10 5

Table I, Relative Rates for the Cycloaddition of Chlorosulfonyl Iso­ cyanate to Various Olefins in Dichloromethane Solution corrected to 2 5 ° .3

Olefin (R = alkyl) kg rel.

rhc=ch2 1

RHC=CHR (cis) 1

RHC=CHR (trans) 0.4

R2C=CH2 30,000 R2C=CHR 20,000 r2c=cr2 15,000 Styrene TOO or-Me thyls tyrene 70,000 Butadiene 20 Isoprene 7,000

Table II. Relative Rates for the Reaction of Chlorosulfonyl Isocyanate " with 2-Ethylhexene-l in Solvents of Varying Polarity at 25° . 3

Solvent n-Hexane Ether Dichlorome thane Nitromethane

Dielectric 1.90 4.22 8.93 3 8 .6 constant k2 rel. 1 30 H O 0 20,000 concerted polarized transition state (10) where bonding occurs at different rates to explain their results.

Interestingly, P-lactones have not been isolated from the! hydro­ lyzed products of CSI additions to olefins and there has been no mention of 2-chlorosulfonylixnino oxetanes (11) in the literature of

CSI chemistry. Although C-0 bond energy is greater than C-N bond energy, 1,4 closure on oxygen from intermediates such as 2 or £ to give 11 is not seen perhaps because of higher strain energy relative to the N-chlorosulfonyl P-lactarns, less conjugative stabilization of

' s Ai cjk-* n m

11 the oxygen with the imino function, and/or rapid reversibility of carbon-oxygen bond formation. The possibility of fast hydrolysis of

11 or the resulting P-lactone to give a P-hydroxy acid seems doubtful because none have been isolated, although the possibility cf ready loss of carbon dioxide to regenerate olefin does exist (however, this 19 reaction generally requires elevated temperatures ).

Notwithstanding, closure on oxygen to form cyclic lactones and

N-chlorosulfonyl imino ethers in five membered rings and larger have been demonstrated. In this connection, CSI additions to cyclohexa- 11 11 10,18 1,3 -diene, bicyclo[4.2.0]octa-?,4-diene, cycloheptatriene, and 7 io cycloocta-l,3>5-triene show parallel behavior. For example, the addition of CSI to cyclohexa-l,3-diene in dichloromethane at room temperature initially gives 13_after 5 minutes. After 30 hours only

1^_ is present. And finally if the mixture is refluxed in chloroform for 17 hours only 15_ results. Intermediate 12_ is an obvious choice to explain these findings. All four of these processes are formally

.1 ,2 1,4 and 1 ,2 -* 1 ,6 rearrangements involving the stabilizing

CSI +

0 ' I > N-S02C1

12

,S02C1

15 Ik

effects of an allylic carbonium ion and relief of strain in the N- chlorosulfonyl p-lactam.

Solvent polarity does appear to control the proportion of closure on oxygen to give an imino lactone and on nitrogen to give a lactam.

For example the addition of CSI to camphene at -60° initially gives spiro (3-lactam 17_ as the kinetic product, which on warming rearranges via 16 to 18 to give the thermodynamic products 19 and 20. The ratio

-6o CSI +

* n -so 2ci 16 17

V S02C1 S02C1 18 19 20 of 1^ to 20_ in hexane is 3 :1, in dichloromethane the proportions are essentially equal (l:l), while in nitromethane solution 20^is seen to dominate by a factor of 1:6.7* The increased amount of lactone may result from preferential solvation at nitrogen where the negative charge density is likely highest with resultant retardation to ring closure on nitrogen. 20,21 The special case of CSI addition to cyclooctatetraene is of interest because of the generation of a homotropylium cation 21 and its subsequent collapse to 22. No N-chlorosulfonyl-(3-lactam formation

_ »0 C102s-N C102S

CSI +

21 22 was observed under the reaction conditions (50°, no solvent). Another example where no P-lactam has been observed is in the 1 ,6 addition 22 of CSI to 23 which gave after hydrolysis 2b. The structure of 2b

CSI

23 2b

was determined by X-ray crystallography. This is one of the first examples where addition across an unconjugated diene system has been observed.

The high stereospecificity of the (2+2) cycloaddition of CSI to 4 ,5 ,12,15 olefins and conjugated dienes has alternatively been inter­ preted as a concerted or ''near concerted’’ ( t^s + n^a) thermally 23 allowed process, with the cumulative carbon-nitrogen n-bond acting as the antarafacial component. A ( j^ s + j^a) process is not thermally 10 allowed. Barton and Rogido have cited two examples as evidence for a concerted cycloaddition involving CSI. First, the addition of CSI 24 to 2-cyclopropylpropene in ether at -60° followed by reductive hydrolysis at -30° gave 26, while this addition at 0° gave Type I

V V CSI + \ 0 S02C1 £ H

25 26

N-chlorosulfonyl amide products 28_ and cis and trans 29; The authors argued for a concerted addition followed by heterolytic ring opening to intermediate 27 to explain their results.

25 N-S02C1 'NH I S02C1

27 28 &

In a second example, Barton and Rogido showed that the first pro- 25 duct in the addition of CSI to diphenylmethylenecyclopropane was 30 which further rearranged to 32. Dipolar species ;5JL is considered an intermediate in the reaction leading from 30 to 32 but not in the 11

0 Ph CSI ^[X ^ ------* iXWi ^ Ph Ph ^ ^ Ph

22.

Ph-^2-- * • £\ i N-S0 2C1 Ph

32 31

initial formation of 350, which was thought to be a concerted (i^s + jt2a ) cycloaddition.

If carbon-nitrogen bond formation falls behind carbon-carbon bond formation in a '’near-concerted’’ process as Moriconi and coworkers have suggested, a polarized intermediate, e.g. 5_ or 10, suitably- aligned for bonding could well be involved. If this is so, then

Woodward-Hoffmann rules do not apply. In the stereospecific cycloaddi- 4 tion of CSI to cis- and trans-(3-methylstyrene, Moriconi and Kelly believe the reaction to likely be preceded by it-complex formation.

Other mechanistic possibilities also suggested were the quick collapse of an internal ion pair to P-lactam product before bond rotation, a ''pseudo-concerted'' mechanism, and a modification of the Woodward-

Hoffmann rules for the cumulenic system because of d-orbital partici­ pation of sulfur and/or nitrogen lone pair interaction. 7 Later, in the case of (2+2) cycloaddition of CSI to dienes,

Moriconi and Meyer proposed that the electrophilic reactivity of CSI and its capacity to act as a -n^a component is decreased by molecular orbital interaction (33 ) of the olefin n system and the low-lying carbonyl it orbital of valence bond structure la, thus making a ''near-

11 concerted'' (jt2a + j^s) thermal process allowed. The authors also suggested that CSI could not participate as a jt2s component in a sym­ metry allowed ( tt4s + jt2s ) thermally allowed cycloaddition (which was not observed) because the carbonyl it system would be orthogonal in the transition state and thus be unavailable for secondary overlap. The seeming inability of CSI to enter into Diels-Alder reactions as a n2s 3 component is also exemplified in the addition to cyclopentadiene, a better ir4s diene than acyclic dienes. The only products isolated (-50°, ether) were 1,2 rather than 1,4 adducts resulting from attack of CSI at the terminal carbon of the diene unit. It appears at this point that the distinction between what is called a near-concerted or dipolar transition state and a true 1,4 dipole that quickly closes to a kinetically favored {3-lactam may be a matter of semantics. Dipolar addition with varying amounts of charge development depending upon the stability of the resulting carbonium ion followed by fast closure is all that is necessary to explain the products. To be sure, the ability of N-chlorosulfonyl P-lactams to

* * undergo heterolytic ring opening can be predicted on the basis of carbonium ion stabilities. But if this cleavage can occur after P- lactam formation, it must also be possible for these dipolar interme­ diates to occur before P-lactam formation (principle of microscopic reversibility).

The one best example of a concerted process involving P-lactams 6 is their thermal cleavage at 600° in a (o^a + c^s) process. Under these conditions cis- and trans-3 ,4-dimethyl and diethyl 2-azetidinones

(34) gave better than 9Cf?o yields of olefins (3 6 ) with a remarkable 99$

0 H — N=C=0

+

H & 22. Ri R4 stereospecificity. Here the N-chlorosulfonyl group is not present to stabilize a possible 1,4 dipolar, two-step cleavage. That no bond rotation occurs at 600° can be taken as strong evidence for a twist transition state (35 ) leading to concerted cleavage.

In addition to the high stereospecficity, the CSI reagent also exhibits high stereoselectivity in many instances. Notable examples of this feature are the addition of CSI to norbornene, norbomadiene 26 and dicyclopentadiene. In each case, the lone product (a P-lactam) results from exo attack at the more strained double bond. Moriconi and

Crawford suggested an initial electrophilic attack of CSI on the exo face to give it complex 37 which rearranges to generate 1,4 dipole 38

CSI +

37 15 and its subsequent closure to 39* An exception to the rule may be 8 the addition of CSI to 5-ethylidenebicyclo[2.2.l]hept-2-ene which affords ^0. Here the less strained but more sterically accessible and more electron rich double bond is attacked exclusively. Homo- allylic participation of the developing carbonium ion center might also be invoked here as with norbomadiene.

CSI S02C1

Another general area where the stereoselective nature of CSI is 9 , 12- 14 evident is in its addition to isoprenoid bicyclic monoterpenes.

In every instance CSI approach occurred from the less hindered side of the double bond. In all of these studies, Wagner-Meerwein rearrange­ ments played an increasingly important part at higher temperature.

A previously cited example is the addition of CSI to camphene leading to lg_ and 20. 15 As a uniparticulate electrophilic species, CSI can generate and quickly trap carbonium ions in alicyclic systems where competition between Wagner-Meerwein rearrangement and cyclization is possible. It has been observed that in rearrangement prone norbomyl systems, CSI effectively traps the developing carbonium ion without skeletal 16 26 rearrangement. In a more typical example of bimolecular electrophilic 27 addition, bromine adds to norbomadiene with rearrangement. The normal trans addition of bromine to simple olefins implies a bridged intermediate or bromonium ion analogous to it complex 3J_ but in the fast ring closure step the amidate anion of CSI must approach in a cis fashion, an approach which might otherwise indicate a totally free carbonium ion.

Because of thoir unique symmetry, the ease of detecting skeletal rearrangement, and their lack of ring strain, bicyclo[2.2.2]octyl systems sometimes serve as subtle tests for the stereoselectivity of electrophilic addition. As concerns bicyclo[2. 2.2]octatriene

(barrelene, 4l) and its analogs, the opportunity to assess anchimeric assistance and delocalization of positive charge presents itself.

The rigid geometry and special three-dimensional it-electron character of barrelene has prompted experimental scrutiny of its capability 23-30 to enter into molecular rearrangements. High levels of inter­ action between the adjacent bridges are seen. The interpolated heats 31 of hydrogenation show that the interaction of all three it systems is destabilizing by approximately 10 kcal/mole. Hydrogenation of k-2 and 43 gives values consistent with unstrained double bonds, so it must be the electronic effects of interacting the third double bond with the first two that are energetically unfavorable. Photoelectron spectroscopy and MINDO/2 calculations of 4l-44 support the hypothesis 32 of strain instability of 4l. 17

AHhyd. * - 2 7 . 9 a

^ y d = -2S.3

The symmetry properties of barrelene are such that stereochemical tests cannot be applied. To gain such information, some structural perturbation becomes necessary. Electrophilic addition to tricyclo-

[3 .2.2.02 ,4 ]nona-6,8-diene (homobarrelene, 45) and its isomeric benzo derivatives 46 and 47 was therefore examined for the first time and 18 forms the subject of the present discussion. Molecules other than 45 can in principle be selected to address the stereochemical issue, but none of these share with homobarrelene the unique features imparted by the cyclopropane ring. Brown's examination of the steric effect caused by 7,7“dimethyl substitution at the bridge carbon in norbomene has revealed a reversal in the customary exo direction of reaction 33 in some, but not all, cases. A comparable assessment of the influence of a cyclopropane ring laterally fused to a homologous bicyclo[2.2.2]octatriene frame has not been made. In benzologs 46_ and

47 the available site for reaction is predetermined on structural grounds and only in one of these (46) is possible 02,4~% ,9 inter­ action attainable. The prevailing steric situations in 46_and 47_are obviously quite different.

In addition to the purely steric factor, preferential interaction of the internal cyclopropane c orbital with the C6>7 rc-bond can become important in 45. The possibility of selective stabilization of this 34 type gains support from the photoelectron spectral behavior of 45.

Bruckman and Kelssinger compared the EE spectra of 42, 4g, and 45 to determine net stabilization of the it systems. It was found that small decreases in the it MO energy levels were present. Moreover, strong interaction between the proper Walsh orbital of the cyclopropane ring with the double bonds of 45 introduced relatively large stabilization.

In marked contrast, replacement of the cyclopropane ring of 45_ with a double bond to give barrelene imparts destabilization to the system. Because of the high reactivity, stereospecificity, stereoselecti­

vity, and steric requirements of uniparticulate electrophilic reagents,

the comparative behavior of b5j and kj_ toward CSI was studied.

Furthermore, the N-chlorosulfonyl 0-lactams which are formed initially

under conditions of kinetic control are frequently rearrangement-prone

and sometimes experience ring opening-isomerization processes of

mechanistic importance. For these reasons, CSI additions to homo- barrelene and syn and anti benzhomobarrelenes were undertaken. RESULTS

The preparation of 45_ was achieved by electrolytic decarboxyla- 35 tion of the diacid obtained by aqueous hydrolysis of the cyclohepta- 36 37,38 triene-maleic anhydride adduct. In agreement with earlier reports,

45 was found to exhibit a 1H nmr spectrum, showing nearly degenerate

secondary cyclopropyl protons (6 0.60, m, 2 ), tertiary protons of the cyclopropane (l.10, m, 2) and bridgehead variety (3 .60, m, 2), and

two distinctive sets of olefinic proton pairs (5-90 and 6.45, q, 2 each). When 45_was treated with an equimolar quantity of CSI in deuteriochloroform solution and the progress of reaction monitored with nmr spectroscopy at ordinary probe temperatures, the 0 6.45 signal

disappears within 15 minutes. The second olefinic quartet was somewhat

reduced in area and some minor alteration in chemical shift was apparent

as reaction progressed. A companion infrared study revealed the rapid

appearance of a carbonyl band at 1825 cm" 1 indicative of N- (chloro-

sulfonyl) p-lactam formation. With the passage of time, this peak was gradually (although never completely) supplanted by an absorption

at 1T90 cm"1. 39 Using diamagnetic anisotropy arguments, Rhodes and his co- 38 1 workers attributed the high field olefinic quartet of 45__ to Ha,Hg.

Although support for the assumed selective shielding was gained by

direct spectral comparisons with syn- and anti-tricyclo[3 .2.2.02’4]-

20 non-6-ene, the difficulties sometimes encountered in determining the 40 magnitude of long-range cyclopropane shielding effects and possible anisotropic contributions of one double bond upon the other prompted us to establish the accuracy of these assignments in an unequivocal fashion. To this end, the synthesis of homobarrelene specifically labeled -with deuterium atoms at positions 6 and 7 was accomplished 41 (Scheme i). Modified Lindler reduction of dimethyl acetylenedicar- boxylate with deuterium gas gave dimethyl maleate-2,3-d2 (|+8 ), the 1H 42 nmr spectrum of which consists of a sharp singlet at 6 3»75* This diester was saponified and the resulting dideuterio-maleic acid was

Scheme I

D2 D D % Fd-BaS04 CH3OOC - C = C - COOCH3 quinoline-S CH3 OOC COOCH3 CH3 OH

1. Na2CC>3 , H2O 2. HaO+ A, -H20 D 3* Electrolysis EtsN, py, 22

condensed directly with cycloheptatriene in refluxing xylene from which water was azeotropically removed. Subsequent hydrolysis and

electrolytic decarboxylation of resulting anhydride 49_ afforded 50_

in low yield. The 1H nmr and mass spectra of this hydrocarbon con­

firmed greater than 9&?° incorporation of two deuterium atoms. More

specifically, the labeled diene almost completely lacked the downfield

quartet centered at 5 6.45. With the knowledge that cycloheptatriene

enters into Diels-Alder cycloadditions to give adducts having an anti 43 cyclopropane orientation relative to the entering dienophile, J50

is necessarily the 6,7-d2 derivative and Rhodes' original assignments

are fully confirmed. Additional evidence for 50_ comes from comparative

mass spectral data of 45 and 50 (Table III).

These preliminary findings suggested that electrophilic attack

on 45_ occurs with total or near exclusiveiy at the anti double bond.

Because the nmr spectra of the unpurified reaction mixtures were com­

plex (several adducts were clearly in hand), product characterization was necessarily preceded by reductive dechlorosulfonylation and column

chromatography. The latter separation was unavoidably deleterious to

the resulting lactone and lactams with the result that high overall

yields of pure materials were not realizable. Conversions to unpuri­

fied product mixtures were invariably near-quantitative. The addition

proceeds to give four adducts (Scheme II), the relative proportions

of which change with the duration of the experiment (Table IV). The

endo,exo product ratio for the five-hour reaction is 4:1. 23

Table III. 70 eV Mass Spectroscopic Fragmentation Patterns of Tri- cyclop. 2. 2.02? 4 ]nona-6,8-diene (4^) and anti-6,7-Dideuterio- tricyclo[3 . 2.2.0^’4 ]nona-6,8-diene (50 ).

relative relative m/e abundance m/e abundance

120 .44 121 3.8 119 5.5 120 41.8 118 58.9 119 100.0 117 100.0 118 34.3(m) 116 2.7 (m) U 7 19.7 J-J-Jlir 1. 1 ll6 12.2 103 .b6 115 2.4 92 .0 2 105 4.5 91 1.3 10i|. 4.3 90 .09 9b 1.6 89 .15 93 18.8 78 .k2 92 12.2 77 .23 91 5.6 65 .29 90 1.6 63 .21 80 7.5 51 .2 8 T9 5.6 39 .3 6 78 3-9 77 1.7 67 2.3 66 3.1 65 2.2 6i«. 1.9 63 2.2 51 3.2 50 2.8 40 3.6 39 6.3 24

Scheme II

xo

0

»M>||

Table IV. Product Distribution After Chromatographic Separation of the CSI-Homobarrelene Adducts Formed at Various Reaction Times in Dichloromethane Solution (25°).

Yield, fo Reaction time, hr 51 53 54

0.5 a a 29-9b 9.4b 0.5 1.5 1.5 9.1 1.0 5 2.5 5-2 8.3 2 .6 24 a a 9-0 28.0 aPercent yield not determined. Crude yields. 25

The availability of both f3-lactam isomers facilitated the stereo­ chemical assignments. The less dominant of these products, exo,anti-

3-azo-4-oxoquadricyclo[4.3. 2.02’ 5.07} 9]undec-10-ene ( 52 , 1747 rvr,w in ex cm”1) was shown by XH nmr analysis to have retained the high field olefinic multiplet and to be of unrearranged structure (see Experi­ mental Section). Appropriate Eu(dpm)3-induced chemical shift studies established the proximal orientation of the lactam functionality to the tertiary cyclopropyl hydrogens. For example, the relevant A E u 4 4 values for H7 and Hg of 52, were determined to be -11.8 and -3.1, respectively (Table V). If the carbonyl oxygen is the preferred site of lanthanide complexation, it follows that H3 (-11.7), H5 (-15.6), and

H© (-9*5) should be affected to a greater extent than H2 (-6.5 ) and

Hi (-2.8), and that the remaining olefinic protons should also be mini­ mally but differently perturbed (Hio> -2.7; Hu, -3 *8)« In contrast, the tertiary cyclopropyl hydrogens in endo,anti-3-aza-4-oxoquadricyclo-

[4.3* 2.02>5. 07>9]undec-10-ene (53 5 1754 cm”1), are characterized max by small AEu shifts (Hy,He, -1.8), as seen in Table VI.

That anti(c=0)-4-aza-5-oxopentacyclo[8.1.0.02’6.03 ,s.07,9]undecane

(5j0 is a Y-lactam follows from its intense carbonyl stretching frequency of 1700 cm-1. The 1H nmr spectrum is characterized by the absence of olefinic signals and the appearance of three new cyclopropyl protons.

The presence of a total of seven such hydrogens is revealed by the array of multiplets appearing at 1.45-1.83 (IH), 1.18-1.43 (2H),

0.42-1.18 (2H), 0.02-0.38 (1H), and -0.08 to -0.37 (1H). The relative orientation of the lactam and cyclopropane groups was established by 26

Table V. Eu(dpm)3 Induced Chemical Shifts in the NMR Spectra of exo,- anti-3-Aza-4-oxoquadricyclo[4.3.2.0s’5.07>9]undec-10-ene (52).

Mole ,DCC13 £Eu Slope Proton Percent °TMS Eu(dpm) 3 (ppm) (ppm)

H i 0 3.07 - 2.8 0.028 5 3 .1 6 10 3 .2 2 15 3.45 20 3-57 30 3.90 40 4.17 50 4 .4 l

h2 0 3 .60 -6 .5 0.065 5 3.77 10 4.08 15 4.45 20 4.87 30 5.48 4o 6.13 50 6.70

h3 0 6.33 -1 1.7 0.117 5 6.67 10 7.40 15 8.07 20 8.70 30 9.93 4o 11.06 50 11.97

h5 0 3.07 - 15.6 0.156 5 3.55 10 4.31 15 5.01 20 6.01 30 7.63 40 9.23 50 10.57 27

Table V. (Continued)

Mole .DCCI3 _ 1 Proton Percent TMS u °^)e Eu(dpm)3 (ppm) (ppm) (A6/Am-7o)

He 0 3 .0 7 -9.5 0.094 5 3.38 10 3.80 15 4 .31 20 4 .8 1 30 5.80 40 6.80 50 7.60

K7 0 1.13 - 11.8 0.118 5 1.67 10 2.23 15 2 .8 7 20 3.^3 30 4.70 40 5.83 50 6.93

Hsa 0 0.10 -2 .7 0.027 5 0 .17 10 0.30 15 0.42 20 0 .62 30 0.88 4o 1.17 50 1.40

Hsb 0 0.10 -1 .4 0.014 5 0 .17 10 O.30 15 0.42 20 0.44 30 0.44 40 0.73 50 0.83

He 0 1.00 -3.7 0.037 5 1.20 10 1.37 15 1.57 20 1.77 30 2.17 4o 2.50 50 2.80 28

Table V. (Continued)

Mole .DCCI3 Proton Percent TMS U ope Eu(dpm)3 (ppm) (ppm) (A6/Am-$)

H10 0 5-93 -2.7 0.027 5 5.9^ 10 6.10 15 6.50 20 6 .4 7 30 6.64 40 6.93 50 7.27

Hxx 0 5.93 5 5.9^ 10 6.10 15 6.30 20 6.60 30 7.00 40 7.40 50' 7.63 29

Table VI. Eu(fod) 3 Induced Chemical Shifts in the NMR Spectra of endo, anti-3-Aza~4-oxoquadricyclo[4.3.2.02>4 .07 59"lundec-10-ene (53).

Mole •DCC13 £Eu Slope Proton Percent TMS Eu(fod)3 (ppm) (ppm) (hb/m-%)

Hx 0 3.07 -1.5 0.015 5 3.12 10 3 .2 0 15 3-30 20 3.41 30 3.53 40 3.63 50 3 .80

h2 0 3.63 -2.7 0.027 5 3.77 10 3-93 15 4.08 20 4.25 30 4.50 40 4.70 50 4.98

h3 0 6.37 -6.4 0.063 5 6.70 10 7.08 15 7.45 20 7.81 30 8.47 4o 8.93 50 9.53

Hs 0 3.35 -7-7 0.076 5 3.77 10 4.18 15 4.61 20 5.01 30 5-77 40 6.37 50 7.17 30 Table VI. (Continued)

Mole .DCCI3 Proton Percent 6 TMS Slope Eu(fod)3 (ppm) (ppm) (A6 / Am-^

H6 0 3 .07 -5 .2 0.051 5 3.37 10 3.67 15 3.93 20 4.21 30 4.70 4o 5.13 50 5.63

Hy,9 0 0.93 - 1.8 0.018 5 1.01 10 1.13 15 1.21 20 1.35 30 1.53 4o 1.63 50 1.80

Hea,8b 0 0.37 -1.3 0.013 5 0.45 10 0.50 15 0.55 20 0 .6 7 30 0.80 4o 0.90 50 1.03

H 10 0 5.67 -1 .4 0.014 5 5.67 10 5.75 15 5.83 20 5.92 30 6.06 40 6 .17 50 6.31

Hxi 0 5.67 -3 .2 0.031 5 5.93 10 6.11 15 6.28 20 6.45 30 6.75 4o 7.00 50 7.30 31

Table VII. Eu(fod)3 Induced Chemical Shifts in the NMR Spectra of syn(n )-4-Aza-5-oxopentacyclo[8.1.0.02j s.03 »8.07 » 9]undecane W)-

,DCC13 £Eu Slope Proton Percent TMS Eu(fod)3 (ppm) (ppm) (A6/&m-%)

HijHxo 0 0.97 -2 .1 0.0207 5 1.12 10 1.18 15 1.30 20 l.4 o 30 1.60 4o 1.80 50 2.03

h2 0 2.35 -4 .7 0.0465 5 2.60 10 2.83 15 3.02 20 3-27 30 3-77 4o 4.20 50 4.68

h3 0 3.48 -3 .0 0.0294 5 3.65 10 3.80 15 3.95 20 4.10 30 4.40 40 4.68 50 4.95

H4 0 6.10 -11.1 0.1102 5 6.72 10 7.28 15 7.92 20 8 .5 2 30 9-70 40 10.67 50 11.53 32 Table VII. Continued)

Mole .DCC13 £Eu Slope Proton Percent TMS Eu(fod)3 (ppm) (ppm) (A6 / tja-%)

He 0 2.65 -10.8 0.1074 5 3 .2 2 10 3.72 15 4.27 20 4.83 30 5-95 4o 6.97 50 8.00 HT 0 1.57 -5.1 0.0516 5 1.70 10 1.93 15 2.18 20 2.53 30 3.05 40 3.53 50 4.07 0 1.67 -1.9 0.0188 5 1.80 10 1.93 15 2.03 20 2.07 30 2.28 4o 2.47 50 2.63 anti 0 O.38 -1.3 0.0122 5 0.45 10 0.50 15 0.58 20 0.77 30 O.83 40 0.87 50 O.98 0 -0.08 -1 .6 0.0160 "•laJt 3 “ 5 0 .02 10 0.08 15 0.17 20 0.25 30 0.47 4o 0.55 50 0.T2 33 lanthanide shifting (Table VIl). The effect of added Eu(fod)3 was most pronounced at H6 (-10.8)., Hy (-5-l)> and H2 (-4.7) as well as at the >N-H site (-11.l).

The most rapidly eluted adduct, anti(C=0)-4-oxa-5-oxopentacyclo-

[8.1.0.02,e.03 ’8.07’,9]undecane (51) ? was a Y-lactone (VCHC13 Qjycj m&x cm"1), the ^ nmr spectrum of which was very similar to that of 54

(see Experimental Section). As the direct result of substitution of lactone for lactam, downfield shifting of the three low-field multi- plets 4.55-4.75 (M), 2.85-2.98 (in), and 2.52-2.79 (is) is observed.

Both the basic structure and stereochemical disposition of the lactone unit were clear from this spectrum. 45-47 Reaction of benzobarrelene (55) with the Simmons-Smith 48 reagent gave a four-component mixture consisting of pairs of mono- and bis-adducts. Increased reliability and somewhat improved yields 49 resulted from the use of diethyl zinc and this method was therefore preferred (Scheme III). Chromatography on silica gel-silver nitrate permitted isolation of the individual hydrocarbons in pure form. A distinction" between syn and anti isomers in the monocyclopropanated

products is reliably founded on 1H nmr spectral data. The through-

space shielding effect of the three-membered ring in 4)7 results in appearance of its olefinic protons at substantially higher field (6 5*97-

6.30 ) than those for 46 (6.68-6.90). Also, the secondary cyclopropyl

protons in 46_ (0.18-0.58; -0.26 to -0.53) appear upfield relative to

those in 47. (1.03-1.40; O.38 -O.9 5)»as a consequence of the diamagnetic

anisotropy of the proximal benzene ring in the syn isomer. In evalua- 34

Scheme III

55 46 47

+ Yi

56 57

tion of the bis-adducts, the symmetry in 57.follows convincingly from

the pairing of its proton signals. Definitive evidence for its syn,syn

stereochemistry includes aromatic shielding of the individually resol­

vable secondary cyclopropyl protons (0.13 and -0.77)• The aryl hydro­

gens of 57_ are at approximately the same chemical shift as those of 56.

but are split into a well resolved AA'BB1 multiplet. The aryl protons

of 56, appear as a narrow, unresolved multiplet. Hydrocarbon 56. is less

symmetric and therefore uniquely reconcilable with prevailing syn,anti

stereochemistry because of the four different secondary cyclopropyl resonances. Molecular models indicate that formation of an anti,anti isomer of 56_ and 57_ is not sterically feasible.

Chlorosulfonyl isocyanate reacted with at room temperature to give three lactams which could again be separated by silica gel chromatography, but with appreciable loss of product. The major com­ ponent (l6 f o ) to be isolated from this mixture was exo,anti-3-aza-10,11- benzo-4-oxoquadricyclo[4.3 . 2.02,5.07,8]imdec-10-ene (58), the exo configuration of which was deduced by pseudo contact shifting of its

XH nmr spectrum with Eu(fod)3 (Table VIIl). Its carbonyl stretching frequency in chloroform solution (17^8 cm-1) differed insignificantly from that of endo isomer 59 (1752 cm”1) which was next eluted ( k . Q P j o

Scheme IV isolated). Qhe major distinguishing characteristic of these (2+2) adducts is the chemical shift difference of tertiary cyclopropyl protons B7 and Hg which appear further downfield in 58^ (1.67-1.83 ) than in 59, (1.00-1.37). The &Eu values of these protons in 58, (-6.6,

-2.4) conform expectedly to closer proximity to the lactam functionality than is possible in 59, (-2. 5 5 -1«9).

The minor (1.3%) product, endo-9-aza-2,3-benzo-10-oxoquadricyclo-

[6.3 .O.04 ,lo.05,7]undec-2-ene, which accompanies the two P-lactams is assigned structure 6o_ on the basis of its intense infrared absorp­ tion at 1700 cm-1 and key 1H nmr signals at 6 4.15-4.33 (2H) due to the protons flanking the lactam functionality, 3.55-3*93 (2H) arising from the benzylic bridgehead protons, and -0 .1 to 0 .9 (series of four one-proton multiplets) attributable to the individually distinctive cyclopropyl hydrogens. The precise alignment of the amide unit rela­ tive to the remainder of the molecular framework cannot be conclusively established from these data because the two adjoining methine protons

(4.15-4.33) appear at chemical shifts sufficiently similar to preclude revealing double resonance studies; rather, this detail follows from mechanistic reasoning (vide infra).

Attempts to add chlorosulfonyl isocyanate to 47, under conditions where other analogous compounds reacted quite rapidly proved uni­ formly unsuccessful. 37

Table VIII. Eu(fod)3 Induced Chemical Shifts in the NMR Spectra of exo,- anti-3-Aza-10,ll-benzo-4-oxoquadricyclo[4.5.. 2.02} 5.O7’9]- undec-10-ene (58).

Mole -DCClg £Eu Slope Proton Percent TMS Eu(fod)3 . (ppm) (ppm) (Afi/Am-%)

Hi 0 3.53 -2.0 0.020 5 3-78 10 3.78 15 3.85 20 4.00 30 4.20 40 4.38 50 4.57

H2 0 3.53 -4.0 0.040 5 3.78 10 4.00 1 15 4. <20 20 4 .47 30 4.82 4o 5.18 50 5-53

Ha 0 6.33 -8.0 0.078 5 6.93 10 7.25 15 7.75 20 8.20 30 8.95 40 9.67 50 10.27

h 5 0 3.13 -9.9 O.O98 5 3.78 10 4.25 15 4.73 20 5.30 30 6.2 7 40 7.20 50 8.07 38

Table VIII. (Continued)

Mole sdcci3 £Eu Slope Proton Percent TMS Eu(fod)3 (ppm) (ppm) (Afi/Am-%)

He 0 5.53 -6 .2 0.062 5 3-78 10 4.16 15 4.52 20 4.8? 30 5.47 40 6.03 50 6.57

Hr 0 l.4l -6 .6 0.065 5 1 .8 7 10 2.23 15 2.53 20 2.93 30 3.57 4o 4.05 50 4.77

Hsa 0 0.10 -0 .9 0.009 5 0 .2 2 10 0 .2 7 15 0.30 20 0 .32 30 0.38 40 0.48 50 0.60

0 -0.80 -0.2 0.018 5 -0.73 10 -0 .6 2 15 -0.53 20 -0.43 30 -O.25 40 -0 .0 7 50 +0.10 39

Table VIII. (Continued)

Mole .DCC13 £Eu Slope Proton Percent TMS Eu(fod)3 (ppm) (ppm) (A6/Am-%)

Ife 0 1.4l -2.4 0.024 5 1.43 10 1.58 15 1.67 20 1.85 30 2.10 40 2 .32 50 2.55

0 7.07 -2.2 0.022 5 6.90 10 7.32 15 7.32 20 7.53 30 7.70 40 7.90 50 8.08

Hl3-15 0 7.07 -1.5 0.015 5 6.90 10 7.32 15 7.32 20 7.45 30 7.55 4o 7.67 50 7.75 40

Table IX. Eu(fod)3 Induced Chemical Shifts in the NMR Spectra of endo,- . anti-3..Aza-10, H-benzo-4-oxoquadricyclo[4.3. 2. O2’5.O7’9^ undec-10-ene (59).

Mole ,DCC13 Slope Proton Percent TMS £Eu Eu(fod)3 . (ppm) (ppm) (A6/&m-$)

H i 0 3 .6 2 -1.9 0.019 5 3.65 10 3-TO 15 3 .8 2 20 4.12 30 4.15 4o 4.35 50 4.57

h2 0 3.97 -2.5 0.026 5 3.97 10 4.17 15 4.27 20 4.4o 30 4.67 40 4 .9 2 50 5 .20

h 3 0 5.48 -6.0 0.060 5 5.78 10 6.08 15 6.33 20 6.65 30 7.27 40 7.85 50 8.47

h 5 0 3 .6 2 -7 .5 0.075 5 3.97 10 4 .3 0 15 4.63 20 5 .0 2 30 5 .82 4o 6.57 50 7.37 41

Table IX. (Continued)

Mole «DCC13 Proton Percent TMS °^8 Eu(fod)3 (ppm) (ppm)

H6 0 3 .6 2 -5.3 0.053 5 3-97 10 4 .1 7 15 4.38 20 4.65 30 5.22 40 5.75 50 6.30

H7 0 1.18 -2.5 0.025 5 1.30 10 1.38 15 1.45 20 1.55 30 2.00 40 2.13 50 2.42

Hsa 0 0.28 -1.4 0.014 5 O.35 10 o.4o 15 0.45 20 0.53 30 0.68 4o 0.80 50 0 .9 7

Hsb 0 -O.58 -1 .9 0.020 5 -O.58 10 -0 .4 2 15 -O.33 20 - 0.23 30 - 0.03 4o 0 .1 7 50 0 .3 7 Table IX. (Continued)

Mole AEu Slope Proton Percent aTUVB S S 13 Eu(fod)3 (ppm) (ppm) (A6/&m-?o)

Hg 0 1.18 -1.9 0.020 5 1.23 10 1.30 15 1.38 20 1.48 30 1.72 40 I.83 50 2.17

Hz2 0 6.98 -3.5 0.035 5 7.17 10 7.37 15 7.57 20 7-72 30 8.13 40 8.43 50 8 .7 1

Hl3-15 0 6.98 -1.3 0.012 5 7-17 10 7.25 15 7.28 20 7.35 30 7.45 4o 7.50 50 7.65 DISCUSSION

Fusion of a cyclopropane ring to barrelene obviously exerts

large effects upon the direction and stereochemistry of chlorosulfonyl.

isocyanate addition. The presence of the bulky three-membered ring modifies the steric environment such that electrophilic attack at

the syn double bond lacks kinetic importance. The present results

reveal further that a directive effect operates at the anti olefinic

site. Differentiation arises in favor of endo attack, i.e., preferred

addition from that direction proximal to the second double bond. This

preference is not maintained during cuprous chloride-promoted diazo-

*vir\*f*Tn r\v"> 4*-? 4- )l C *.tV« -*»/-» o *v>o r>r\-v*4- prl 4- r \ " K p 4liv WW r W w J V * W^OVii. wvU WV ' * V i l U W V * 50 in yields of 46, 25, and 21%, respectively. Reaction of 45 with

6l 62 63

51 diiodomethane and diethyl zinc leads in 85% yield to relative pro­

portions of 62, 12, and 26%, respectively. Hydrocarbon 6l_ does not

undergo further addition. Since only 62_leads to 63_, and if 62_results

from exo attack on the anti double bond of 45, then the smaller steric

43 requirements of these reagents, coupled with its probable impingement upon the center rather than termini of the n bond may allow for a

lesser degree of selectivity. On the other hand, 62.might also

arise via endo attack of the syn double bond of 45_ and therefore it

cannot be said that only the anti double bond of is reactive

toward these reagents. 52 Schueler and Rhodes synthesized syn-tricyclo[3.2.2.02,4]non-6-

ene (6^_), anti-tricyclo[5. 2. 2. 02’ 4]non-6-ene (§5), syn-tricyclo-

[3*2.2.02’4]nonan-6-one (66), and anti-tricyclo[3.2.2.02’4]nonan-6-

one (67), and made stereochemical assignments of the alcohols resulting

from hydroboration-oxidation of the olefins and lithium aluminum hydride reduction of the ketones. Olefin 6k gave only 68a while 65 gave a 3:1 mixture of 70a and 71a, respectively. In contrast, 66 gave

only 69a while 67 gave a 2:1 mixture of 7_0a and 71a, respectively.

It is obvious that the cyclopropane ring effectively precludes attack

on the syn side of the double bond of 6b_ and syn side of the carbonyl of 66.

The steric effects of the tertiary cyclopropyl protons of 65_ and 67. are more subtle. No obvious steric bias exists, but models

show that 65_ may give a slight advantage to attack on the side of the

double bond syn to the cyclopropane bridge. Use of either diborane or thexylborane gave an almost identical ratio of 70a to 71a. Schueler and Rhodes therefore suggested the preferences for attack might be

of electronic origin or that the product ratio is determined by product control, rather than approach of the reagent.

When 67. was reduced with lithium aluminum tri-tert-butoxyhydride,

the sensitivity of this reaction to steric effects resulted in the

reversal of the ratio relative to lithium aluminum hydride so that 71a

predominated over 70a by a factor of 2:1. Thus it was determined that

only this last reaction showed the expected steric bias. However, it

may be that the lithium aluminum hydride reagents are not a reliable

test of stereoselectivity because they may approach the carbonyl of 67_

from the open side of the molecule away from the tertiary cyclopropyl

protons.

Once N-(chlorosulfonyl) P-lactam formation occurs from the exo

direction of ^ further skeletal rearrangement is precluded and 52.is

the unique end product isolated from this reaction pathway. In 46 contrast, the (2 +2 ) adduct arising from endo capture (7 2 ) can experience ring opening with formation of zwitterion 73_• Cyclopropyl participa­ tion to give 74.may also operate. Unfortunately, the disposition of the amidate anion of 74_ precludes the trapping of such a carbonium ion at the cyclopropyl carbon. The interaction of this laterally fused

-SOgCl - M d02C1

NS02C1 72 w 75 \

74 76 cyclopropane ring with the anti double bond may indeed be the source of the endo selectivity of 4+>_ toward CSI. Discounting the electronic effects of the syn double bond of 45j, a larger charge development in the electrophilic addition of CSI relative to diborane may explain the apparent reversal of exo selectivity for diborane addition to 65. 47

A preliminary report of the solvolysis of 69-TL (R = OBs) has 53 recently appeared. Brosylate 68b solvolyzes with cyclopropyl migra­ tion. The solvolyses of 69b and 70b proceed with anchimeric assistance to give cyclopropylcarbinyl cations. Only 71b showed direct long- range cyclopropyl participation, but products and intermediates have 54 not been published. Kirmse and Wahl did publish the results of the solvolysis of 71b in acetic acid and found that this brosylate gave 77 as the only isolable product (93%)- Further, deamination of 71c gave

AcO

OH

77 78

95-9% of 71a» Deamination of 70c gave 6 . 6 f o of 70a and 84.7$ of 78 via Wagner-Meerwein rearrangement. It appears that the cyclopropane ring of 71^ will participate as a homocyclopropyl carbinyl cation but will not migrate to form a secondary carbonium ion. Epimer 70_ experiences charge development away from the cyclopropane thus exhibits anchimeric participation with the appropriate a-bond. While deamination of 72. afforded 9 8 . 6 jo of 80_ and 1.4$ of 8l

79 81 like treatment of 82 produced 76$ of 83 via cyclopropyl migration,

OH + 80 WH2

flo 84 of 80_via homoallylic participation, and 3 $ of unrearranged 84.

These results were interpreted in terms of intermediates 85 and 86 which lose nitrogen. Diazonium ion 85_ exhibits homoallylic participa­

tion to give 8]7 and no vinyl migration, whereas 86_ has the option of

forming homoallylic carbonium ion 88_which can participate to give

8 7, or of forming 89^ via cyclopropyl migration.

These observations suggest that Jb may be an important interme­

diate for mechanistic consideration, although none of the isolated

products (51-54) requires this intermediate. If bonding occurs in

stepwise fashion, 73 and/or 74 would be formed initially. These

centrally important dipolar species can, of course, cyclize to 72.

With involvement of the neighboring double bond, the possibility for more remote C-N and C-0 closure with formation of 75_ and 76 ,is also

possible and does operate. Cyclopropyl migration in 73 would lead to

90. Such a product was not observed.

90

It is interesting to note that at longer reaction times (Table

III) 75_ and j6_ are not formed at the expense of 72. If conditions of

thermodynamic equilibrium exist, the isolation of a (3-lactam is in 4 ,T ,9-14 contrast with the observations of others concerning the

initial kinetic formation of p-lactams (e.g., 6> 13 j 17> and 3 0 )

which undergo Wagner-Meerwein and other carbonium ion rearrangements

under similar conditions to form larger ring lactams and lactones 50 irreversibly. The formation of the second cyclopropane ring of 75_ and

76 must weigh against the relief of strain in going to the Y-lactam and Y-iminolactone. The N-chlorosulfonyl lactams could not he iso­ lated (because of decomposition) to test this hypothesis.

Again, the low material balance in the addition of CSI to 45 makes a full assessment of the stereoselectivity of electrophilic attack difficult. A zwitterion analogous to 73, resulting from exo attack on the anti double bond of 45_ may very well have formed since syn double bond participation should be facile. A 1,2 vinyl shift would thus result in the formation of 9 1* If instead of vinyl migra-

S02C1

r-S02Cl

91 tion cyclopropyl participation had occurred, 92_would be formed. Such rearrangements must either be unfavorable, i.e., the zwitterion is irreversibly trapped to give 52. after hydrolysis, or and 9 2, must be consumed by further reaction with CSI or simply decomposed in workup.

The same can be said of the fate of 90. 26 Moriconi and Crawford did not detect rearrangement via 1,2- vinyl migration in the addition of CSI to norboradiene. The only product in 95% yield was the exo (3-lactam. Two other relevant examples are the addition of CSI to norbomene to give (3-lactam 59 in 86% yield 51 and an analogous p-lactam from the addition of CSI to 43_ in 6yj> yield.

In neither case was Wagner-Meerwein rearrangement reported.

After workup, a product such as 93_probably could be isolated since it is similar to in structure. That 95.is not found is negative evidence for the fact that CSI does not attack the steri-

■0

/ so2c i cally accessible endo face of the syn double bond of 1 5. P-Lactam might add a second molecule of CSI under the reaction conditions and therefore not be isolable after column chromatography.

The added benzene ring in k6_ removes some possibilities for pro­ duct development and alters steric environment of the endo side of the molecule. The isolated product ratio implies a predominance for exo attack by CSI as a result of the increased steric bulk on the endo surface. The isolation of 5%. shows that endo addition of CSI is not totally impeded. Adduct 60_ can be seen as arising via a phenyl migration from intermediate 95, to give cyclopropylcarbinyl cation which is trapped to give 97. 1,2-Phenyl migration in 95_ leads to a minor product but was detectable, while a similar 1,2-vinyl migration in the CSI addition to to give 92, was not observed. The possibility exists that 95 is a more stable zwitterion than its non- 52

0

N-S02C1

95 9 6 97

benzo analog which is quickly trapped before vinyl migration; 95_ is

sufficiently long lived to allow phenyl migration.

The total lack of reactivity of VT under comparable conditions

serves to emphasize the adverse effects engendered by the combination

of the fused benzo substituent and a syn-oriented cyclopropane ring.

Neither homocyclopropylcarbinyl nor homobenzyl participation appear

operational in the hypothetical developing zwitterion 98.

O'^N-SOaCl

98

In summary, the predominance (greater than h: l) of endo attack

of CSI on the anti double bond of homobarrelene (h5) cannot be explained

by steric effects alone. The influence of the lateral cyclopropane

ring of during the development of positive charge is considered an

important factor also. This conclusion is in accord with the unreactivity of anti-benzhomobarrelene (4j) toward this reagent. The combined steric and electronic effects of the fused benzene ring in syn-benzhomobarrelene (46_) induces an opposite product ratio (exo/endo SECTION II

Reactions of Chlorosulfonyl Isocyanate with Barrelene,

Benzobarrelene, and Dibenzobarrelene. INTRODUCTION

In order to investigate further the steric and electronic effects of electrophilic addition of CSI to bicyclo[2.2.2]octyl systems, barrelene (4l), benzobarrelene (55), and 2,3 3 5,6-dibenzobicyclo[2.2.2]- 55 octatriene (dibenzobarrelene, 99) were prepared. Because of the

55 99

difficulty of synthesis of 41, the chemical reactivity of this triene has 28,29 been little studied. The 13C nmr spectrum is now available and exhibits two absorptions at 140.55 ppm and 48.29 ppm in deuteriochloro- form using TM3 as an internal standard. The J13n values for the C-H olefinic and bridgehead carbons are 175*3 Hz and 143.4 Hz, respectively.

From 13C satellite measurements, the published values for the neat 29 liquid are 176 + 1 Hz and 140 + 1 IIz.

The bromination of 4l_ gives rise to 100 and 1 0 1 in a ratio of 29 3:1» respectively. Analogous behavior has been observed in the addi- 27 56 tion of bromine to both norbornadiene and bicyclo[2.2.2]octadiene

(42). No other electrophilic additions to barrelene have been reported.

55 56

Br Br Br

NC 100 101 102

It is interesting to note that similar products have also resulted 28 from the addition of dienophiles to 4l. For example, dicyanoacety- lene reacts in a (2 + 2 + 2) fashion to give 102_at room temperature in 95$ yield. There is the possibility that this cycloaddition is

stepwise and dipolar rather than a concerted ( it2 s + n2s + i ^ s ) thermal

process.

While there have been no studies of electrophilic addition to

55 (recently, the addition of nitrosyl chloride in chloroform at -30 °

to 55 has been recorded, but the dimeric product(s) was not fully 57 characterized ), there have been reported a rich variety of electro- 58 59 philic additions to 99. Addition of chlorine and iodine in carbon 60 tetrachloride, iodobenzene dichloride in chloroform (in the dark), 61 58 bromine in acetic acid, iodine and silver acetate in benzene and 61 62 acetic acid, performic acid in ether, lead tetraacetate and tri- 63 methylsilyl azide in dichloromethane, iodine isocyanate in ether and 64 benzenesulfenyl chloride in acetic acid to 99_all proceed with

skeletal rearrangement to give syn, exo, and endo products 105, where

Y is predominantly exo. All these reactions are throught to go via 57

X

99 + XY

103

an ionic mechanism. Indeed 99^ has become a test system for ionic mechanisms.

The addition of benzenesulfenyl chloride to 99, in carbon tetra- 65 66 chloride or ethyl acetate unexpectedly gives unrearranged trans

products of type 104. This result was attributed to fast trapping of

99 + XY

104

the carbonium ion before rearrangement could occur. Similarly, mer­ er curie acetate in acetic acid adds to give the unrearranged cis adduct.

In aqueous acetone this reagent gives 30°/> 104 (Y = OH, X = HgOAc) and

70% of the cis isomer, again without skeletal rearrangement. Under

conditions of ultraviolet irradiation, 99_ undergoes free radical chlori- 68 nation with iodobenzene dichloride in chloroform to give 104. Irra-

5! diation of 99. in the presence of iodine at -22° in carbon tetrachloride

also gives 104. Where X = Y = I, syn,exo-103 slowly isomerizes at room 58 temperature in the dark in carbon tetrachloride or deuteriochloroform to a mixture of iodine, 99, the thermodynamically more stable anti,exo isomer of 103 , and surprisingly 104. RESULTS

The addition of CSI to barrelene (4l) at 0° in dichloromethane with gradual warming to room temperature followed by dechlorosulfonyla- tion after 1 hour and chromatography on Florisil led with substantial loss of material to 105, 1 06, and 10Tb_ (Scheme V). y-Lactone 105, the

Scheme V

J+l 105

+ ^ X l 10

106 107a, R = S02C1 107b, R = H first compound eluted (3* lf> yield), was characterized by the absence of

N-H stretching in the ir but the presence of a strong carbonyl at 1778 6o cm”1. The nmr spectrum showed two olefinic protons at 6 6.37 (m, J7 ,s>

= 6 Hz, Js ,9 = 8 Hz) and 6 5*70 (t of m, J9jio = 7 Hz) for Hs and He, respectively. Three tertiary cyclopropyl protons were present, one downfield (6 2.33) and allylically coupled to Hs was assigned as Hy, and two different upfield signals at 2 .0 1 and 1.84 were assigned to

Hg and H5 , respectively. All three exhibited 6 Hz coupling to each other. Comparisons of chemical shifts of 105 and 51 show obvious analogies.

Lactams 106 and 107b were isolated in 4.97° yield as a mixture which predominated in 107b. Unfortunately, 1071) was destroyed while attempting its sublimation. The crude nmr spectrum showed a broad N-H absorption at 6 5 *7 5? two vinyl protons at 6.47 and 5*8 9, and tertiary cyclopropyls between 1.6 and 2.5 ppm. Hi appears at 6 3*59 as a multi- plet and H4 was observed as a multiplet overlapping with Hio on the downfield side of the range 6 2.90 to 2 .6 3 . Sublimation did give 106 in 0.7% yield, this product being characterized by a broad N-H absorp­ tion and a P-lactam carbonyl stretch at 1751 cm-1 in the infrared.

Its nmr spectrum showed four olefinic protons in a narrow (2+2) pattern from fl 6.52 to 6.12 and a broad singlet at 5 .8 0 ppm. No upfield tertiary cyclopropyl protons were evident. Comparisons of the nmr spectrum of 106 with 52^ and that of lOJb with 5^ are in accord with the assigned structures.

Mixing of equimolar quantities of barrelene (4l) and CSI in dichloromethane solution at -78° followed by gradual warming to room temperature without hydrolysis led in 74% yield to an oily product consisting chiefly of N-(chlorosulfonyl) Y-lactam 107a. This structural CHCl assignment follows mainly from infrared (v^^. 3 1722 cm-1) and H nmr spectral data. In particular, only two well-separated olefinic C-H signals are present at 6 6.5^ (mj J7 ,s = 6 Hz, Js,9 = 8 Hz, Hs) and

5.80 (t of d, Jsjio = 8 Hz, J7 j9 = 2 Hz Hg). Additionally, those protons flanking the lactam functionality are distinctive [4.j8 (m,

Ji,6 = Ji, xo = 2 Hz, Hx) and 3-03 (br s, H4)], while the three cyclo­ propyl hydrogens appear at 2.53 (br of d, Js,7 = J6 ,7 - 5-5 Hz, H7 ) and I.85-2 .3 8 (overlapping t, H5 and H6). Since 107a dominates in the crude product mixture, it must be assumed that this compound is destroyed to a greater extent during reductive dechlorosulfonylation and chromatography than the other products.

As a minor variation in synthetic method, benzobarrelenone (108) was converted into its tosyl hydrazone (109) which in turn was treated 67 with methyllithium to give 55_ in 52% yield (Scheme VI). Kitahonoki

Scheme VI

0 N-NHTs 62 46 and Takano's method gave significantly less than bCffo yield via oxime 4T reduction, methylation, and Cope elimination. Zimmerman, et.al., produced 2 b% of 55_ from 108 by sodium borohydride reduction followed by xanthate ester formation and pyrolysis.

The reaction of one equivalent of CSI and benzobarrelene (55) at room temperature in dichloromethane solution proceeded to completion in 30 hours. 'Che mixture was worked up and passed through Florisil to give a kk°jo yield of an unseparated mixture of five products (Scheme VIl).

Scheme VII

10

CSI + 10

55 111 110

lOi NH

OH 1 12

11? Chromatography on silica gel of a similar mixture gave in order of elution (4.356), 110 (2.7$), 111 (7.51°), 112 (6.056), and a mixture of

113 (1 .8%) and 114 (3* 756), which was separated by manual partitioning of the different crystalline forms.

8,9-Benzo-2-oxa-3-oxoquadricyclo[4.4.0.04 ’10.05 ’T]dec-8-ene (IIP) was assigned its structure based on the absence of N-H in either the nmr or ir spectrum and the presence of a strong carbonyl stretch at

1780 cm-1. No olefinic protons were evident by nmr, but a series of upfield tertiary cyclopropyls at 6 2.79 (d of d, Js5y = 6.4 Hz, J6,t =

7.0 Hz) due to Hy and two signals at 6 1.80 and 2-37 as overlapping multiplets due to H5 and H6 were apparent. Also a multiplet at 6 4.47 could be assigned to methine proton Hi.

The assignment of the exo configuration for 3-aza-7,8-benzo-4- oxotricyclo[4. 2.2. 02 >5]deca-7,9-diene (ill) was more difficult. By ir spectroscopy it was obviously a P-lactam because of an N-H stretch and a carbonyl absorption (1755 cm-1). An unresolved two-proton olefinic multiplet at 8 6.18-6 .7 2 was removed by hydrogenation and replaced by a series of upfield methylenes as the result of prevailing shielding TO effects. Better evidence was gained from the lack of change in the chemical shifts of H2 (8 3.48-3.71, m) in 111 to 3.53-3-76 in 113 and of H5 (3.16-3.40, m) in 111 to 2.98-3.42 in 115. Further proof came’ from Eu(fod)3 pseudocontact shifting (see Table x) where Hiqj the ole­ finic proton syn to the carbonyl, had a &Eu value of -4.3 while the ortho aromatic proton syn to the carbonyl had a £Eu value of -2.2. 64

Table X. Eu(fod)3 Induced Chemical Shifts in the NMR Spectra of exo- 3-Aza-7>8-benzo-4-oxotricyclo[4. 2.2.02’ 5]deca-7,9-diene (111) •

Mole- -CDC13 AEu Slope Proton Percent TMS Eu(fod) 3 (ppm) (ppm) ( A5 / Am-$)

Hi 0 4.08 - 2 .0 0.0196 5 4.22 10 4.32 15 4.43 20 4.53 30 4.70 4o 4.90 50 5.08

h2 0 3.65 -3 .6 0.0360 5 3.88 10 4.03 15 4.23 20 4.40 30 4.80 40 5.13 50 5.45

Ha 0 6.33 -7 .6 0.0747 5 6 .7 1 10 7.20 15 7.6 7 20 8.05 30 8 .8 2 40 9.45 50 10.02

Hs 0 3-33 -9 .9 0.0986 5 3.85 10 4.38 15 4.90 20 5.38 30 6.38 40 7.30 50 8.28 65

Table X. (Continued)

Mole- .CDC13 AEu Slope Proton Percent TM3 Eu(fod) 3 (ppm) (ppm) (A6/Am-1o)

H6 0 4.18 -6 .7 0.0667 5 4.55 10 4 .8 7 15 5.20 20 5.52 30 6.1 6 40 6.75 50 7.62

h9 0 6 ,58 -1 .8 O.OI87 5 6.58 10 6.70 15 6.80 20 6.90 30 7.10 4o 7.28 50 7.45

H i o 0 6.57 -4 .3 Oo 0423 5 6.90 10 7.12 15 7.33 20 7.53 50 7.97 4o 8.58 50 8 .7 2 Three anti 0 7.23 -1 .2 0.0121 aromatics 5 7-27 10 7.33 15 7-40 20 7.45 30 7.57 4o 7.70 50 7.83 Aromatic 0 7.23 -2 .2 0.0210 syn to 5 7.38 carbonyl 10 7.50 15 7.6 7 20 7.77 30 7.93 40 8.18 50 8 .2 7 Double resonance studies were necessary to determine the structure of 8-aza-2,3-benzo-9-oxotricyclo[5.3*0.04>lo]deca-2,5-diene (112, see 7 1—73 Table Xl). All values are typical of a [3 .2. l]bicyclic structure; however, the lactam linkage causes JljY to be a little larger than 73 normal (1.8 + 0.5 Hz) and Jx5io to be smaller than expected (4.2 +

1.0 Hz) for dibenzo systems. The y-lactam functionality is in evidence from an ir carbonyl stretch at 1707 cm-1.

2-Aza-8,9-benzo-3-oxoquadricyclo[4.4.0.O4’10.05 >7]dec-8-ene (113) has spectral properties similar to 110. Again there is an N-H absorp­ tion and an unstrained carbonyl ir stretch (1669 cm”1) for a Y-lactam.

No vinyl protons were detected in the nmr spectrum but a benzylic tertiary cyclopropyls (1.78-2.35? H5 and H6) did appear upfield.

The last product isolated was 7,8-benzo-anti-2-carbamyl-6- hydroxytricyclo[2.2 .2.03 ,5]oct-:7-ene (Il4) a direct hydrolysis product of 113. The nmr spectrum of 114 shows a broad two-proton singlet at 6 6.13 and a broad one-proton singlet at 5-40. Amide I and II bands appear in the ir spectrum. Further evidence for a hydroxy amide 67

Table XI. 100 MHz NMR Data from Double Resonance Experiments for 8-Aza-2,3-benzo-9-oxotricyclo[5.3- 0.04’ 10]deca-2,5-diene (112) in CDC13 -TM3.

Chemical Coupling Proton Shift Constant Irradiated (ppm S) (Hz)

Hx 3-40 Jl,5 = 0*8 Jl, 6 = ^ J x,r = 3*0 Jx, 10 = 2.0

h 4 3.50 J4,5 = 6.0 J4,6 =1.4 J4,10 = 4.5

h 5 6.19 J5 ,6 - 9* 2 J5,7 = J5 , 10 =0.7

He 5-71 J6,7 = 4. 5

Hr 3.79 Jv, 10 = 0 H10 4.05

(rather than an amino acid)comes from mass spectroscopy (see Table

XIl), which shows loss of H2NC0 (or C02) at m/e = 171? but also shows a H2NC0 fragment (m/e =44).

Addition of CSI to dibenzobarrelene (99) (Scheme VIII) in refluxing dichloromethane during 72 hours without effecting hydrolysis gave 2 . Tk 68

Table XII. TO eV Mass Spectroscopy Fragmentation Pattern of 7,8-BenzO' anti-2-carbamyl-6-hydroxytricyclo[2. 2.2.03>5]oct-7-ene TOO.

Relative m/e Abundance Remarks

215 68.5 parent peak 198 42.6 loss of «0H or NH3 197 18.5 loss of HOH 196 1 1 .1 182.3 metastable loss of • OH 175 20.4 17 2 1 2 .0 + 171 6 0 .2 loss of H2N=C=0 or C02 170 31.5 169 15.7 168 1 0 .2 + 154 20.4 loss of •OH and H2N=C=0 153 82.4 o - [ > - C 6H4 — <^j 15 2 24.0 151.8 metastable m/e: 198 -* 173 144 8.7 143 37.0 142 18.5 l4l 63 .O 139.2 metastable m/e: 171 - 154 139 1 2 .6 138.7 metastable m/e: 171 - 153 136.9 metastable m/e: 171 - 152 130 7.8 129 46.3 128 1 0 0 .0 naphthalene 127 20.4 116 13.7 115 44.4 114.6 metastable m/e: 173 -♦ l4l *7*7 \ 1 1 0 .0 76 8 .9 63 14.6 59 1 0 .2 51 1 2 .0 + 44 1 8 .3 h 2n =c =o 43 10.5 doublet 39 13.9 cyclopropenium ion 69

Scheme VIII

NC

CSI + 99

116 117a, R = S02C1 iilk R = H

‘N-R

Il8a, R = S02C1 ll8b, R = H

of Il6, an estimated 2.2% of 117a, and 19.1% of ll8a after chromato­

graphy on silica gel. Adduet 117a was not amenable to purification.

More importantly, 117a does not isomerize to ll8a under the reaction

conditions. When a mixture of 117a and ll8a (2:3) was heated in CH2C12

at the reflux temperature for lUO hr, no increase in the amount of

118a could be noted although some decomposition of ll_7_a was evident.

Refluxing pure ll8a for 72 hours gave no detectable amount of 117a by

infrared analysis. This intrinsic unreactivity of ll8a is without

question the factor which permits its isolation. Reductive dechloro- TO T4 sulfonylation of this pair of adducts was effected to give 117b and

13.8b, the spectral properties of which were also consistent with the respective formulations.

initial structural assignment of 116 was based on its infrared spectrum which showed no carbonyl absorption but did exhibit a promi- CHCX l nent nitrile stretch (v 3 22^6 cm” ), and its H nmr spectrum max which suggests a rearranged [3 .2.l]bicyclic skeleton based upon coupling T3 constants and chemical shifts of other similar systems (e.g., 1 03 ).

The exo-4,anti-8 stereochemistry of II6 follows from low order spin- spin interactions of H4 and H5 (2.2 Hz) as well as between Hq and Hx, and Hs (~ 1 Hz). Further proof of structure comes from the base pro­ moted cyclization of 116 to 13-9 (Scheme IX) with potassium tert- T5 butoxide in DMSO. An authentic sample of 119 was also prepared by re dehydration of the known carboxamide IgO.

Scheme IX

116

CN C0NH2

119 120 71

Little evidence for 117a is available per se but it was detected as a component of several chromatography fractions by ir spectroscopy

1815 cm-1). Reductive dechlorosulfonylation of 117a as a mixture with ll8a did give 117b which was characterized as a (3-lactam by ir ( v ^ P -3 3^20 and 1758 cm”1) and XH nmr spectroscopy. The latter shows two nearly degenerate benzylic bridgeheads (6 ^.3 3 > t, Jx,2 = 77 J5 j6 =3-5 Hz, Hi and H6) expected for a bicyclo[2.2.2]octyl frame­ work. The two upfield multiplets flanking the lactam functionality do not show the high spin-spin interaction of cis vicinal protons

(8 .8 +0.8 Hz) but are totally consistent with the data for 106, 111, and 113 as well as 5 2, 53 _, 58, and 59»

N-Chlorosulfonyl lactam Il8a was uniquely able to withstand chromatography on silica gel and isolation techniques. The infrared spectrum of ll8a 1772, 1^06, 1186, 1156, and 114^ cm”1) nicely o . / - . r ^ r u 3713.X indicates the functionalities present. The ^-H nmr spectrum exhibits the strong electron-withdrawing effects of the N-chlorosulfonyl group on H7 (8 5-37, dd, Jlj7 = 5 Hz, Jr ,io = 3 Hz, >CH-N<), and the syn 73 orientation of the carbonyl by the high spin-spin interaction of H i0

(8 3*78, t of d, Ji,10 = J4,10 = 5 Hz, >CH-C=o) and the bridgehead

T4 PHP1 1 protons. Reduction of ll8a-rr— gave ll8b —"— 1*»■« (v__„ max 3 3^-05, 1710 cm” ) where the electron-withdrawing effect on H7 has been reduced (6 4.18) while coupling constants remain similar. Comparison of 112, Il8a, and ll8b is instructive. DISCUSSION

It is important to note the general order of reactivity of CSI toward (in decreasing order) 4lj 55, and 99. Barrelene (4l) is the fastest to react not only for steric reasons hut also because of elec­ tronic factors. Stereoselectivity of CSI addition to 4l_ cannot be reflected in the products, which can be explained by homoallylic parti­ cipation of 121. Intermediate 121 can reorganize bonding to give cyclopropylcarbinyl cation 122 or quickly collapse to give P-lactam

125. 1,3-Closure of 121 or 122 through nitrogen gives 107a while closure on oxygen gives 124. In addition there is the possibility of a concerted (itas + tt^s + jt^s) cycloaddition of CSI with formation of 107a and 124 as Zimmerman and coworkers have found in the addition 28 of dienophiles to 4l_ (e.g. , 102).

Vinyl migration in 121 to give a rearranged [3 .2. l]bicyclic ene-allylic carbonium ion (1 25) which would produce the non benzo analog of 112 was not observed. Ring opening of 122 to give 126 en route to 107a and 124 cannot be excluded. Both 125 and 126 would be antiaromatic if fully delocalized.

72 N-SOaCl

N-S02C1 74 These results correlate well with the findings of Grob and co- 56 workers who found that the electrophilic addition of peracid to 42_ gave only 1,2 addition to the electron-rich endo side with formation of

127. Homo 1,4 addition of bromine to 42_ to give 128 (30$) was appa­ rently less favored than trans 1,2 addition to give 129 (70$).

Br +

Br Br 127 42 128 129

The ability of homoallylic participation to win out over vinyl migration in bicyclo[2. 2. 2Joctyl systems is further indicated from the

OAc OTs OAc

OTs

130 121 133

OAc OAc + 132 'OAc

134 135 136 75 78 results of LeBel and Huber, who solvolyzed 130 and 151 in acetone containing tetraethylammonium acetate at 56°. Tosylate 151 gave 22fo of 152 and 78$ of 155? showing the propensity of the ethylene bridge for anchimeric assistance and intervention of an allylic carbonium ion.

Tosylate 150 gave k-6% of 134 and 3 2 of 155? both the result of homo- allylic participation, 12fo of 156, the product of vinyl migration, and lOfo of 152. Clearly, vinyl migration in bicyclo[2.2.2]octyl systems is unfavorable. 79 Acetolysis of benzobicyclo[2.2.2]octadienyl brosylates at 25° reveals that solvolysis of 137 is 3 *8 times faster than 142 and proceeds

OAc OAc

OBs 138 139

.OAc Ac

140 l4l

mostly with vinyl migration (138 and 159) although homoallylic partici pation does occur (140 and l4l). The relative proportions are 72, 11, 76

10, and 7%, respectively. Solvolysis of 1^2 gave 83 % of lkj> and 17% of

lMf, resulting entirely from phenyl migration.

OBs OAc

'OAc

1 H

The addition of CSI to 55, proceeds -without any detectable vinyl migration products, but the results do show that phenyl migration

operates in the production of 112. Homoallylic participation, analogous to that found in the addition of elecurophiles and dienophiies to ^ and b-2, also occurs to give 110 and 113. One initial intermediate (1A5 ) can be seen to give rise to 1h6 by quickly trapping the carbonium ion, or rearrange to give allylic carbonium ion 1kj which would close to

148. Bond reorganization of ±b^ to give iMj) from which closure on oxygen gives 150 and closure on nitrogen gives 151 may operate. Formation of

150 and 151 could also be explained as a concerted (n2s + n2s + rt2s) cycloaddition. The reversibility of these pathways was not investigated.

Significantly, the steric effect of the fused benzene ring slows the rate of addition of CSI to 55_ relative to 41, and allows the reagent to approach only from the anti side of each double bond. Lithium 80 aluminum hydride reduction of 108^ showed only a 3:2 preference for exo versus endo attack (similar reduction of 6? showed 2:1 selectivety S02C1 N-S02C1

+ / +

147 145

ir 149 / / ,S02C1

x S02C1 0 N

148 151 150 78 for endo attack), but if the hydride reagent complexes with the carbonyl oxygen of 108 first, then the steric bulk of the benzene ring is appreciably avoided and product development may become more important.

No products resulting from endo attack and/or vinyl migration (such as 132 or 153) were observed in the addition of CSI to 55. Lactam 132 would be easily identified by nmr spectroscopy if it were present

'N-H

l

in detectable amounts. 3-Lactam 153, might have been more elusive due to its similarity to 111. Vinyl migration in 145 to give 154 would close to give 150 and 151, but there is no necessity to invoke this many rearrangements.

The products from the addition of CSI to 99, also be envisioned as arising from one intermediate (155) which can quickly close to give

155 157 4 4 4 118a 117a 116 79

117a or rearrange via a phenyl shift to give 156 and 157. Zwitterion

156 can close to form Il8a, but 157 cannot cyclize and must decompose to chloride ion which combines with the benzyl carbonium ion and S03 which leaves behind a chloronitrile (II6 ). These pathways are irreversi­ ble.

It is important to note the effects of the two fused benzene rings on the rate of addition of CSI to 99. which is slower than to 55*

Stereoselectivity is again lost in §9.. The obtention of 117a is of interest because ionic additions to 99, generally proceed with carbon skeleton rearrangement to 4,8-disubstituted dibenzobicyclo[3. 2. l]octa- dienes in both nonpolar and polar solvents. The possibility of a concerted [:t2 s + it2a] cycloaddition mechanism exists in the formation o f 117.

The dibenzobicyclo[3. 2. l] and [2.2. 2]octadienyl systems have been 59 found to undergo equilibration under neutral conditions. Both endo and exo 105 (X = H, Y = OH) have been converted to 104 (X = H, Y = OAc) 81 in acetic acid-sulfuric acid. Similar results were observed by treat­ ing the same two alcohols with p-toluenesulfonic acid in benzene to give 82 10^ (X = H, Y = OTs). Acetolysis of this tosylate in turn gave both 82 endo and exo acetates (103, X = H, Y = QAc). Nenitzescu and coworkers thus suggested an equilibrium between phenonium ion 158 and benzyl carbonium ion 159 to explain these results.

It appears that polar additions to 99, occur under conditions of kinetic control to give dibenzobicyclo[3 .2 . l]octadienyl products which can equilibrate to more stable thermodynamic products (dibenzobicyclo-

[2.2.2]octadienes). However, the thermodynamic equilibration of 117a and 118a could not be accomplished in refluxing dichloromethane. The product ratios of the addition of CSI to 99, may reflect kinetic control, but a definitive statement cannot be made because of the low isolated yield, especially since 117a is unstable.

A comment on the serendipitous formation of ll6 is warranted.

Mechanistically, migration of the ''wrong'' phenyl ring (syn to the N- chlorosulfonyl amidate function) of 155 would lead to 157» This process isolates the amide function from the benzyl carbcnium ion and capture cannot occur. Loss of chloride ion which quenches the cation would leave

160, or alternatively, oxygen may displace chloride to give l6 l. 81

Transfer of oxygen to sulfur in l60 or heterolytic C-0 bond cleavage in l6l could result in 162 which will lose S03 to form 116.

Precedence for nitrile formation under neutral and nonnucleophilic 83 conditions does not exist. Vorbruggen decomposed a series of N- chlorosulfonyl amides (163 ) with triethylamine in acetonitrile or dichloromethane to give nitriles (1 6 6) and proposed intermediates 164

0 0 11 11 1 R-C-NH-S02C1 ------R-C-N-S02C1 ►> R-C=N 3

163 164 / O-SO2CI

R-CN ----- R-C. ^ N © l66 165

and 163 to explain the products. Heating N-chlorosulfonyl P-lactarns in 8 DMF to give a , P-unsaturated nitriles is well documented. Even THF has 84 been observed to displace chloride in 167 to give 168.

CH3 02C-N-S02C1 Na® + THF >- CH3 02 C-N=S02* 0 I + NaCl © O

167 168 In conclusion, CSI is seen to exhibit steric preferences during addition reactions to unsaturated bicyclo[2.2.2]octyl systems such as

46 and 5£. Superimposed upon this steric control is an electronic effect caused by the laterally fused cyclopropane ring present in

46, and 47. The influence of a laterally fused benzene ring is reflected in the reactivity of 46_, 55j and 22.* Generation of carbonium ions on the bicyclo[2.2. 2]octyl skeleton has resulted in phenyl migra­ tion while vinyl migration in similar systems (4l_ and 45) has not been detected. SECTION III

Synthesis of 1-Cyanosemibullvalenes via Chlorocyanation

of Barrelene, Benzobarrelene, and Dibenzobarrelene.

8? INTRODUCTION

Substituent effects on the fluxional character of unsymmetrical homotropilidenes (169) have been shown to affect the equilibrium compo­ sition. Alteration of the equilibrium has been accomplished by incor­ poration of the controlling substituent between Cj. and C5 or by external attachment to the carbocyclic frame. These substituents have generally preferred bonding to the cyclopropane ring to varying degrees. A polar substituent effect ought to be better satisfied by connection to an sp2 center where conjugation is possible (e.g., C4, C5, or C6)- 85 Schroder's investigation of monosubstituted bullvalenes (170) has

171

X 0R ^C-NH b x C=Cx b

H v COOMe 0 CHq b c -c x h R ✓ N ¥ H S A .OCHa RO c— c V C=N b ✓ V & ✓ \ 122.

c -ch 2 C— NH b ✓ \ b I I H-C-C-H / \ 84 indicated that such weakly accepting and donating groups as methyl, chloro, bromo, and iodo show little discrimination between the various available sites. In fluorobullvalene, however, that isomer predominates

(80-85%) where the electronegative fluorine is situated at the only sp3- hybridized aliphatic carbon (C5). With alkoxy and carbomethoxy sub- 85, ae stituents, the preferred orientation is that illustrated in 171a-171c.

When an electron-withdrawing carbonyl group is introduced as in bull- 8T 20 88 valone (lTld) or the lactams 171e and 171f, an opposite directing 89 effect results. 11118 trend is maintained in the azabullvalenes lTlg 90 and (3-lactam 171h. 85 Goldstein's more recent finding that homobullvalenone 172 is characterized by preferred bonding of the carbonyl terminus to C5 conflicts with those trends found in the lower homologs and is not easily reconciled with available theoretical assessments of electronic 91,92 effects in Cope equilibria. Because complications from larger 93 longicyclic frameworks can arise, the true electronic perturbational effects in 172 are not evident. Hiis is not so in the two-fold degenerate barbaralone series (173) where methyl is recognized to prefer Ci (K =

R

172 174b

R = ch3, c 6h5, ch2och3, ch2oh 86

173 175, R = CH3, CH2OCH3, F

94 3.28) and deuterium C5 (K = 0.80). In those 1(5)- and 2(4)-substituted 95,96 semibullvalenes (174 and 1 7 5 , respectively) examined to date, the 91,92 predicted equilibrium values agree with experimental data. With the exception of 175-F, the equilibria clearly illustrate preferential attachment to olefinio > cyclopropyl > aliphatic regardless of sub­ stituent character. However, the effect of an efficient ^-electron acceptor such as cyano on the facile Cope rearrangement of semibull- 97 valene had eluded assessment.

Thermochemical studies have indicated the effects of nitrile groups 98 on strained ring stability to be minor. Such combustion measurements deal, however, with total stabilizing effects and do not consider specific bond weakening and strengthening influences which may be 94 operating within the molecule. That individual cyclopropane ring bonds are affected by such forces is suggested by the appreciable shift in the cycloheptatriene-norcaradiene equilibrium witnessed in the 99,100 101,102 case of the 7-cyano and 7,7-dicyano derivatives. Unlike the semibullvalene valence isomerization process, this last equilibrium 87

DMF CSI + 'C»c^ ' c = c ' ▼ H * CN CIO^ 177 176

I I DMF - c —C-H © I C — C — C-H ,0 C102S-Nq ^ 0 I I + Cl' .N — C . / 0 178 I 1 H-C 'N(CH3 )2 I I — C —C-H © © I © + Cl 182

x . 0 0 I I - C — C- H © 179 © I + Cl'

I I I I I OsS chn(ch3 )2 0 ■ C — C— H © © I + Cl C 183 III N © S03'©

180

I I -C— C-H r © I + Cl- + S03 177 + HC1 C=N

181 88

is not degenerate, lies too far to one side in the parent system, and 101 consequently is difficult to assess quantitatively. (Hydrogenation of cycloheptatriene-norcaradiene over rhodium gives less than 0 .0 0 5$ norcarane.) 95,96,103-106 The currently available access routes to semibullvalenes are not conducive to introduction of a cyano substituent. The formation of 1JL6_ from the addition of CSI to 99, and its subsequent conversion to

1 19 has lead to the development of procedure for the synthesis of 1- cyanosemibullvalenes. As a consequence of the pioneering work of Graf 107 and his coworkers, it is now feasible to convert a simple alkene to an a,P-unsaturated nitrile (ITT) through solvolysis in dimethylforma- mide solution of its chlorosulfonyl isocyanate (CSl) adduct (lj6). Moriconi and Jalandoni have investigated the mechanism of these N-(chlorosul­ fonyl) P-lactam (1 7 6 ) ring opening reactions and shown the process to 8 be general. They hypothesized both thermal (1 7 8-1 8 1 ) and nucleophilic

(1 8 2-1 8 5) mechanisms to arrive at a common intermediate l8l which can lose sulfur trioxide and a proton to give 177. Yields of 177 were

poor without an excess of DMF at 70-80°.

Where a comparably N-substituted Y-lactam is concerned, the action of dimethylformamide at elevated temperatures results instead in conversion to a Y-chloro nitrile. The transformation of 185 into I8 6 108 is illustrative. The addition of CSI to benzvalene (184), followed by treatment with DMF for 48 hours at room temperature gave 1 8 5. Other

Y-lactams were isolated with 1 8 5. Treatment of 186 with a variety of bases did not lead to the desired ring closure (cyanoprismane). Lactams 89

Cl

CSI DMF H

CN C102S 184 186 185

of general structure 185 are known to result on occasion from CSI additions to polyolefinic or strained hydrocarbons prone to structural isomerization under electrophilic conditions. For example, the addi­ tion of CSI to a-pinene gives a (3-lactam which rearranges at room 13 temperature to give 187. However, heating 187 in DMF gave only 188.

DMF

S02C1 CHO

187 188

Barrelene (4l) contains the requisite eight carbon atoms and five degrees of unsaturation. Anticipation of a Y-chloronitrile resembling 116 and l86 could result by heating 107a in DMF. If 1,3- elimination of HC1 could be accomplished, cyclopropyl ring formation would then lead to 1-cyanosemibullvalene. RESULTS

Mixing of equimolar quantities of barrelene (4l) and CSI in dichloromethane solution at -78° followed by gradual wanning to room temperature led in 74$ yield to an oily product consisting chiefly of N-(chlorosulfonyl) Y-lactam 107a (Scheme X). This structural assignment follows mainly from infrared (v^ 3 1722 cm-1) and ^ nmr spectral data (CDC13 solution). In particular, only two well-separated olefinic C-H signals are present at 6 6.54 (m, J7,a = 6 Hz, Ja,g = 8 Hz,

He) and 5-80 (t of d, J9 ,io = 8 Hz, J7j9 = 2 Hz, He). Additionally, those

Scheme X

CSI > DMF

41 S02C1 189

107a

protons flanking the lactam functionality are distinctive, 4.78 (m,

Ji,6 = Ji,10 = 2 Hz, Hi) and 3«°3 (br s, H4 ), while the three cyclo-

propyl hydrogens appear at 2.53 (tr of d, J5 )7 = J6 ,7 =5.5 Hz, H7) and

I.85-2.38 (overlapping t, H5 and H6). Heating of this unpurified

90 91 material in dimethylformamide at 75-95° for 40 hr and isolation of the resulting product by preparative vpc afforded exo-4-chloro-anti-8- cyanobicyclo[3-2. l]octa-2,6-diene (182.) as a colorless solid in 31 % yield. The definitive spectral data for 189 include an intense infrared band (in CH2C12) at 22^5 cm”1 and a XH nmr spectrum which indicates not only that four olefinic protons are present [6 6.6k (q, Ji,r = 3 Hz,

J6,7 = 5*5 Hz, H7 ), 6.26 (m, Jx,2 = ^ Hz, J2,3 —9*5 Hz, J2, 4 = 2 Hz, H2),

5.95 (q, J5 ,6 = 3 Hz, He), and 5 .k-3 (m, J3, 4 = 2 Hz, J3f5 = 1 Hz, H3)] 71-73 but also uniquely defines the stereochemical features at H4 and Ife. 73 As with many dibenzo[3.2.l]octadienes, the low level of spin inter­ action between H4 (4.4l, m, J4j5 = 3 Hz) and H5 (3*32, m) as well as between the pair of bridgehead protons (Hx, H5) and Ha (< 1 Hz) is reconcilable only with an exo, anti arrangement of the two substituents.

Treatment of 189 with potassium tert-butoxide in DMSO-THF gave

190 in 5^% isolated (vpc) yield (Scheme Xl). The semibullvalene was obtained as indefinitely stable colorless needles which proved quite

Scheme XI

CN CN

I89

190a 190b amenable to three-dimensional X-ray crystal structure analysis. Beno 30,109 and Cristoph have subjected 190_ to three-dimensional X-ray cry-

stallographic analysis at -45°. Under these conditions, the crystals

were found to consist wholly of tautomer 190a. Two illustrations of

the structure appear (Figures 1 and 2). The substance crystallizes

with the orthorhombic space group P2 12 12 i, with four molecules per

cell. The cell constants are: a = 6 .6 9 3(2 ), b = 8 .3 0 1 (3 ), and c =

12.393(4) ft. The phase problem was solved by direct methods using 110 the program MULTAN and the structure refined by the usual methods.

After refinement of all atomic parameters including those of the

hydrogen atoms, the R factor is O.O5 6 for 1192 independent reflec- 30,109 tions.

The molecule, with bond distances and angles of interest, is

illustrated in Figure 1. As shown in Figure 2, the structure possesses

nearly exact noncrystallographic mirror symmetry, with the mirror con­

taining C9, Ci, and C5 and bisecting the cyclopropane ring. Bonds C3 -C4

and C0-C7 are highly localized and characteristically olefinic. The

distances between C4 and C6 [2.352(3) ft] and C3 and C7 [3 .0 7 2(3 ) ft] are

too long for bonding interaction and quite possibly can be no shorter

because of repulsive interactions between the opposed n orbitals. The

two five-membered rings are significantly nonplanar, with C3 -C4 and

C6—Cy being forced outward from the planes of the other three atoms in

each ring. The angle between the two planes formed by C2-C3 -C4 -C5 and

Cs-Ce-Cy-Ce is 98.1°, placing the p-it lobes of C4 and C6 nearly per­

pendicular. Figure 1. A lateral display of bond lengths and bond angles of 1-cyanosemibullvalene (lgOa). Figure 2. A ventral display of bond angles 1-cyanosemibullvalene (lgOa). 95

Tables XIII and XIV contain typical low- and high-temperature

100 MHz ^ nmr data for the reciprocation of 190a and 190b in two different solvent systems. In CD2C12-CF2C12, the small amount of CDHC12 present served as an internal standard for monitoring changes in chemical shift as a function of AT. Two sets of downfield signals are seen, each with a relative area of 2, and these exhibit negligible variable-temperature dependence. Since the H3,H/ and bridgehead (H5) protons are likewise not appreciably affected either in the low temperature range or between 0 and 100°, these data do not permit determination of K as has been possible with other semibull- —eq 95-97,111,112 valenes, although clear indication is provided that 190a is heavily dominant. The chemical shift of H4 ,H6 is outside the limits for such a calculation.

An estimate of is made possible by 13C nmr spectroscopy. The

13C spectrum of semibullvalene at room temperature is characterized by three resonances at 50.0 (Ci,C5), 86.5 (C2 ,C4 ,C6,C8), and 120.4 ppm 113 (C3,C7 ). At -l6o°, where the Cope rearrangement becomes ''frozen out,'' five signals appear at 42.2, 48.0, 53*1? 121.7, and 1 3 1 .8 ppm 97 assignable to C2 and Ca, Ci, C5, C3 and Cr , and C4 and C6, respectively.

The results for IgO determined at room temperature in the present study indicate a comparable five-line series of resonances at 43*3 (C^),

49.9 (C2>Ca), 56.0 (C5), 119.2 (C3,C7), and 130.0 ppm (C4,C6).

Expectedly, the inductive influence of the nitrile substituent exerts a shielding effect at Cx while deshielding C2,Ca. Because of their proximity to CN, these carbons are not directly relatable to those of 96

Table XIII. Variable Temperature XH NMR Data (100 MHz) for l(5)-Cyano- semibullvalene (190a ^ 190b) in CD2CI2-CF2CI2 (ppm 6).

Temp, °C CDHC12 HsjHs H3,H7 h 4,h 6 Hs

-1 5-30 3-^5 5.21 5.75 3.51 -IT 5.31 3.46 5.21 5.77 3.53 -31 5.32 3.48 5.22 5.78 3.56 -50 5-34 3-47 5.23 5.81 3.56 -75 5.36 3.48 5.24 5.83 3.58 -95 5.38 3.50 5.26 5.85 3.60 -115 5.40 3-53 5.27 5.87 3.64

Table XIV. Variable Temperature 1H NMR Data (100 MHz) for l(5)-Cyano- semibullvalene (19Qa ^-9Ob) in CI2CCCI 2 (ppm 6).

Temp, °C H25Ha h 3 ,h7 h 4,h 6 h 5

-19 3.38 5.15 5.69 3.46 -4 3.38 5-15 5.67 3.45 19 3.39 5-15 5.64 3.44 32 3-41 5.14 5.63 3.43 49 3-41 5.14 5.60 3.43 60 3.42 5.14 5.58 3.41 72 3.43 5.14 5.57 3-42 84 3.43 5.14 5.55 3-41 99 3.46 5.14 5.53 3.41 97 the parent hydrocarbon. However, the permanently olefinic carbons

(C3,C7) show a quite close correspondence in the two systems (121.7 vs

119.2 ), presumably as a consequence of their more remote position

relative to the cyano group. Given the assumption that C4,C6 would

also be little effected, the chemical shift of these carbons in

''frozen'' semibullvalene (h2.2 and 1 3 1 .8 ppm) should be directly

relatable to that of the similar carbon in 190a (130.0 ppm). Using 33 _2 these values, a K of 2.05 x 10“ is calculated, which translates ’ —eq ’ into an energy difference of 2 .3 6 kcal/mol or 0.10 eV favoring 190a. 91 Hoffmann and Stohrer previously calculated the energy difference between lgOa_and l^Ob. to be 8 .5 kcal/mol or 0.37 eV in favor of the

1-isomer. The equilibrium constant demanded by these calculations is

5- 8 x 10-7 at 25°.

Under similar conditions, benzobarrelene (55) was stirred with

one equivalent of CSI at room temperature for 18 hours. Direct treat­ ment of the cycloadduct mixture with warm dimethylformamide followed by careful silica gel chromatography led to isolation of three chloro- nitriles 191-195 i*1 yield of 23 .5j 23 .8 , and 6.5%, respectively (Scheme

XII). A synopsis of the proton coupling constants exhibited by these

closely related benzobicyclo[3.2.l]octadienes is given in Table XV.

Those coupling patterns which are important for stereochemical

characterization of the chloro and cyano groups of this triad are J4js,

Jl,8> and J5 q. The chemical shifts and coupling constants of the ole­

finic protons are useful in determining the positions of the fused benzene ring. 98

Scheme XII

Table XV. XH NMR Coupling Constant Data for Chloro Nitriles 191, 192, and 195 (CDC13 solution, 60 MHz).

Spin Interaction Coupling Constant, Hz J 191 192 195

1,2 6.4 — — 1,7 --- 2.9 2.8 1,8 4.0 < 1 < 1 2,5 9-5 ---- — 2,4 1.2 ----- — 2,8 1.0 -- - — --- — 3,h ' 3-3 5,5 1.7 — ^,5 1.2 2.9 5.1 5,6 --- 2.9 2.5 5,8 4.0 < 1 < 1 6,7 5.5 5.6 99 As a direct consequence of the rigid geometries and the recog- 1X4 nized reliability of the Karplus rule in bicyclo[3.2.1]octane 71-73 systems, the dihedral angle relationships of the C5-H bond to an endo- (~ 60°) or exo-4-proton (~ 40°) and to an anti- (~ 40°) or

syn-8-proton (~ 80°) are sufficiently distinctive to be revealed by the magnitude of the spin-spin interactions. Thus, the low order of J4js

in 191 and 192 (l.2-2.9 Hz) is accommodated by an endo-4-protcn, whereas the larger coupling constant in 193 (5*1 Hz) connotes an exo

orientation for H4. That 192 and 193 contain a syn-8-proton is indi­

cated by the weak coupling (< 1 Hz) to Hj. and/or H5, a phenomenon not

shared by 191 (4.0 Hz) whose 8-proton must therefore be anti. The higher vinyl coupling in 191 (^2,3 ~ 9*5 Hz) suggests a 6,7 fused benzene ring, while the smaller values for 192 and 193 (J6,7 = 5*5 and

5.6 Hz, respectively) indicates a 2,3 ring fusion for the latter two

structures. The coupling constants for 19JL were consistent with 79 those of 1^3 (J2 ,4 = 1.0, J3j4 = 9*5 Hz). Likewise 192 and lg3 corre­

lated with 138 and 139, respectively. Compounds 138 and 139 gave J6 ,7 =

5*8 Hz and Jxt7 = Js,6 = 2.9 Hz.

Upon individual brief treatment of 191-193 with potassium tert-

butoxide in DMSO-THF solvent at room temperature, dehydrohalogenative

closure occurred with formation of 1-cyanobenzosemibullvalene (194) in

each instance (Scheme XIIl). Bender and Shugarman have recently

established that 194 can also be obtained exclusively by subjecting 2- 115 cyanobenzobarrelene (195} to triplet-sensitized irradiation. A

sample of 194 was provided by these investigators for a mixed melting

point. 100

Scheme XIII

KOtertBu hv 191, 192, 193 DM30-THF R 1COCH3

195 194

The XH nmr spectrum of 194 exhibits 4 aromatic protons, two ole- finic protons (Je,7 =5*3 Hz) at 8 5-67 (H6) and 5*14 (H7), an allylic- benzylic proton at 4.16 (H5), and two tertiary cyclopropyl protons

(J2 ,8 = 7*3 Hz) at 3 .6 5 and 3 .2 8 ppm. Table XVI is a comparative listing of the 13C nmr data for 119, 190? and 194.

When equimolar amounts of dibenzobarrelene (99) and CSI were heated at the reflux temperature of dichloromethane for 70 hr (note diminished reaction rate) and the resulting product mixture treated directly with dimethylformamide, chloronitrile 3.1.6 could be isolated in low yield (lOfo) after silica gel chromatography. As previously mentioned the exo-4 anti-8 stereochemistry of ll6 follows primarily from the low order of spin-spin interaction between H4 and H5 (2.2 Hz) 73 as well as between Ha and the two bridgehead protons (~ 1 Hz). The presence of a rearranged [3 .2 .1] bicyc3i.c skeleton in 116 is fully 75 consistent with its facile base-promoted cyclization to dibenzocyano- semibullvalene 119, an authentic sample of which was prepared from 101

Table XVI. 13C NMR Chemical Shifts for Cyanosemibullvalenes 119, 190 and 194 (CDCI3 solution, TMS internal standard, 33C^).

Chemical Shift, ppm 6 Carbon 190 194 119

-CN 1 1 9 .0 Ci 43.3 43.1 44.4 C2 49.9 45.9 45.2 C3 119. 2 134.0 133.9

C4 1 3 0 .0 148.0 149-0 C5 56.0 56.6 57.1 C6 1 3 0 .0 1 3 6 .4 1 4 9 .0 Cy 1 1 9 .2 1 1 8 .9 133.9 Ce 49.9 44.2 45.2 aromatic -- 1 2 7 .0 127.7 1 2 6 .8 127.3 125.4 125.2 1 2 1 .1 1 2 1 .1

76 known carboxamide 120 by conventional methods. The independent syn­ thesis of 119 is worthwhile because this particular cyanosemibullvalene lacks all but three of the ring protons which so uniquely characterize the nmr spectra of the less substituted analogs I90 and 194. The spectrum of H 9 consists of eight aromatic protons and two singlets at

6 4.72 (H5 ) and 3*67 (H2 and Ha). Data from the 13C nmr spectrum are given in Table XVI. DISCUSSION

The mechanism of the transformation of 107a (or the analogous N- chlorosulfonyl imino ether 12k) in DMF may result from a thermal lactam ring opening to give cyclopropylcarbinyl cation 122 which could open

N(CIfe)2 0 0

Cl DMF 107a 122 126 >

196

DMF

196 189

so2 I 0-CH=N(CH3)2 197 ©

to allylic cation 126 (see page 73)* DMF assisted displacement of chloride at some stage of this sequence could result in neutralization

1 0 2 103 of charge to give 196 and loss of S03 to give 189. If DMF-assisted displacement of chloride occurs initially to give 197, subsequent nucleo- philic attack of chloride at cyclopropyl would produce 196. If the exo configuration of the chlorine in 189 is not the result of thermo­ dynamic equilibration, a solvent cage effect might be the reason for attack from the syn (n) side of 197.

Base induced 1,3 elimination of HC1 probably results in extraction of Hq in 189 giving 198 followed by S ^ ’ displacement of chloride to produce 190. The exo stereochemical disposition of the chloride is 1X5 compatible with this mechanism.

N = C

189 > // / ► 190

198

Direct comparison of the bond lengths in 190a with the structural features of semibullvalene as determined by electron diffraction in 117 the gas phase (see 199) and of the 7,7-dicyanonorcaradiene 200 102 (X-ray methods) is particularly informative. For example, while

the Ci-Cs bridge in 199 is a short single bond at 1.485 X, the 1-cyano

substituent has the effect of returning this bond to the more normal

length of 1.544 5L Wang and Bauer recognized the labile C2-C8 bond in

semibullvalene to be of abnormally large dimension (l. 600 %). Because

the cyano group can constantly maintain a it-interactive conformation relative to the cyclopropane ring in 190 c, the expectation follows that strengthening of the C2-C8 linkage should result. In actuality, a reduction in length of approximately 0.03 X is seen, hut the value of 1.572 A remains appreciably larger than that of 1. 501 £ seen for

200. Although 200 is geminally substituted by two nitrile groups and their combined influence will serve to further shorten the opposite cyclopropane a bond, the C2-C8 distance in 190c lies well outside the expected range for substituted cyclopropanes (Table XVIl). m fact, the length is closely akin to that determined for the highly strained central bond in the small-ring propellane, 8,8-dichlorotricyclo-

[3.2.1.01’5 ] octane (Table XVIl).

The C2-C8 bond length variation of semibullvalenes, which is typical of this skeleton, is sensitive to electronic influences. Thus, the effect of a lone cyano substituent at C* is to shift the Cope equilibrium heavily to that direction in which cyclopropyl bonding pre­ vails. By comparison, one 7-cyano group does not suffice to move the cycloheptatriene equilibrium in the norcaradiene direction far enough Table XVII. Bond Lengths in Cyclopropane Derivatives.

Remote or Proximal or Compound Method Internal Bond (X) Edge bond (a ) Ref

Cyclopropane M 1.510 b

Chlorocyclopropane M 1.515 1.513 c

1,1-Dichlorocyclopropane M 1.534 1.532 d

Cyclopropane 1,1-dicarboxylic acid X 1.462 1.534 e

Bicyclopropyl X(-100°) 1.48? 1.501 f

Cyclopropanecarboxamide X 1.481 1.507 g

Cyclopropanecarbohydrazide X 1.48 1.50 h cis-l,2,3-Tricyanocyclopropane X 1.518 i

Bicyclo[l. 1.0]butane M 1.497 1.498 a

Bicyclo[2.1.0]pentane E 1.439 1.521 k exo,anti-Tricyclo[3 • 1.1.02>4]heptan-6-yl X 1.54 1.48 1 p-nitrobenzoate

8,8-Dichlorotricyclo[3.2.1.01’5]octane X(-40°) 1.572 1.458 m Table XVII. (Continued)

„ ,a Remote or _ Proximal or _ _ ComPcuna Methcd Internal Bond (X) Edge bond (8 ) Ref

2.5-Dimethyl-Tj 7-dicyanonorcaradiene X 1.501 1*556 n

Semibullvalene E 1.600 1.550 o syn-8,8-Dichloro-4-phenyl-5,5-dioxabicyclo- X 1.53 1.48 p [5.1.0]octane

Axivalin hydrate X 1.51 1.49 q.

6.6-Diphenyl-3,3-diethyl-3-azabicyclo- X 1.525 1.517 r [3.1.0]hexane anti-8-Tricyclo[5.2.1.0g? 4]octyl X 1.54 i 1*51 s p-bromobenzene sulfonate Benzocyclopropapyran X 1.51 1*52 t trans-Bicyclo[5.1.0]octane-4-carboxylic acid X 1.31 1.49 u trans-Eicyclo[5•1.0]octane-4-methanol X 1.20 1.4l u p-bromobenzenesulfonate Potassium trans-bicyclo[5.1.0]octane-4- X 1.464 1.509 v carboxylate cis-Bicyclo[5.1.0]octane-4-exo-p-bromo- X 1.499 1.500 u benzenesulfonate References for Table XVII.

o "h .M = microwave spectroscopy; X = X-ray diffraction; E = electron diffraction. 0. Bastiansen, F. N. Fritsch, and K. Hedberg, Acta Cryst., IT, 538 (1964). CR. H. Schwendeman, G. D. Jacobs, and T. M. Kriggs, J. Chem. Riys., 40, 1022 (1964). ^W. H. Flygare, A. Narath, and W. D. Gwinn, J. Chem. Phys., 36, 200 (1962). eM. A. M. Meester, H. Schenk, and C. H. Mac Gillavry, Acta Cryst., 827, 630 (1971). J. Eraker and C. Romming, Acta Chem. Scand., 21, 2721 (1967); see also 0. Bastiansen and A. de Meijere, ibid., 20, 516 (1966). %. E. Long, H. Maddox, and K. N. Trueblood, Acta Cryst. , B25, 2083 (1969). hD. B. Chestnut and R. E. Marsh, Acta Cryst., IX, 413 (1958). 1A. Hartman and F. L. Hirshfield, Acta Cryst., 20, 80 (1966). JK. W. Cox and M. D. Harmony, J. Chem. Riys., 50, 1976 (1969). ^ K. Bohn and Y. H. Tai, J. Amer. Chem. Soc., 92, 6447 (1970). ^S. Masamune, R. Vukov, M. J. Bennett, and J. T. Purdham, J. Amer. Chem. Soc. , 94, 8239 (1972). InK. B. Wiberg, G. J. Burg- maier, K. Shen, S. J. LaFlaca, W. C. Hamilton, and M. D. Newton, J. Amer. Chem. Soc., 94, 7402 (1972). nRef 102. °Ref 117. %. R. Clark and G. J. Palenik, JCS Perkin II, 194 (1974). qG. D. Anderson, R. S. McEwen, and W. Hertz, Acta Cryst., B29, 2783 (1973)* rF. R. Ahmed and E. R. Gabe, Acta Cryst., 17, 603 (1964). SA. C. MacDonald and J. Trotter, Acta Cryst., 18, 243 (1965). ^L. G. Guggenberger and R. A. Jacobson, Acta Cryst., B25, 888 (1969). A. Kershaw, Hi.D. Thesis, The Ohio State Univer­ sity, 1974. VR. A. Kershaw, M. S. Thesis, The Ohio State University, 1972. 108 99 for detection. Two such groups are necessary to achieve complete 102 dominance of the norcaradiene valence isomer. With 7-cyano-T- trifluoromethyl substitution, about 1 Cffo of the equilibrium composition 1 0 0 is in the norcaradiene form.

The two five-membered rings in 190 were found to be significantly nonplanar with C4 and C6 being forced in an outwardly direction. The net effect of this distortion is to cause the two pit orbitals to become orthogonal. This introduces some anti bonding character.

The resulting C4-C6 distance (2.352 $) is clearly too large for a it- bonding interaction, although it is understandably shorter than that between the more remote C3 -C7 pair (3-072 &).

The activation energy for the Cope rearrangement of semibullvalene 92 has been calculated to be 2.3 and later 3-3 kcal/mole. Anet and coworkers 97 used variable temperature i nmr spectroscopy and line shape analysis to determine that the free energy of activation and the enthalpy of activation to be 5-5 and 4.8 kcal/mole, respectively. The 91 entropy of activation was found to be -5.4 eu. Hoffmann and Stohrer

postulated a possible negative activation energy for the Cope rearrange­ ment by studying substituent effects. That is, the transition state would be of lower energy than either valence bond isomer of the semi­ bullvalene. Such effects would lead to a neutral bishomobenzene species.

Among the effects the latter authors discussed was the observation

that jt-electron donors interact with the Walsh orbitals of cyclopropane

to weaken the 2-3 bond while ^-electron acceptors strengthen the 2-3 bond. Instead, it appears that all substituents in the 1(5) position 109 show preference for a cyclopropyl environment over aliphatic. The effects of a strong n-electron donor at Cj has not been studied. Of the substituents studied to date (Table XVIIl), the effect of in­ creasing electron withdrawing character of the substituent stands in accord with the existing predictions of Hoffmann and Stohrer in trend, but not in magnitude.

Ry ignoring temperature solvent and inductive effects in the 1H nmr spectra of 174 and the 13C spectrum of 190\, it is possible to esti­ mate the equilibrium parameters given in Table XVIII using a simple 3 7 mole fraction technique and Anet's ’'frozen out' 1 values of H2 and

H4 (8 2. 79 and 5*59j respectively) for semibullvalene or the corres­ ponding 13C data. The high temperature 1H nmr data of 190 (Table XIV) do show a perceptible thermal chemical shift dependence for H4 ,HS.

The change in chemical shift over the temperature range -19° to 99° is 96 comparable to the change in 174 (R = C6H5) from -126° to +33 °.

Qualitatively, the equilibrium constant for 190 (2.05 x 10~2 at 33°) can be assumed to be less than the phenyl compound, but still 105 times larger than Hoffmann and Stohrer predicted.

Mechanistically, it is possible that 191? 192, and 193 arise by initial CSI addition to the sides of the double bonds anti to the benzene ring in 55_ to give 146, 148, 150, and 151 (see p 77). Thermal ring opening of 146 to 145 with a 1 ,2 phenyl shift would produce 148.

The action of DMF on l4j3 could result in 191, possibly by displacement of chloride ion, which in turn on S„T2' attack would lead to 201. Attack N -— - Table XVIII* Estimated Equilibrium Constants and Free Energy Differences for l(5)-Substituted Semibullvalenes (174a £ IT^-b) Based on Chemical Shifts of H4 , H6 and C4 ,C6 Relative to 1H and 13NMR Data for Semibullvalene (174, R = H) at -l60° . 97

9 7 96 96 96 R H ch3 CH20CH3 c 6h5 CN

Solvent CC12F2 c s2 cd2c i 2 cd2c i 2 CDCI3

Nucleus XH, 13C XE XE XH 13C

t ( ° c ) — 36 35 33 33

K 1.00 0.407 0.202 0.021 eq 0.057 AG° (kcal/mole) 0.00 0.552 0.980 1.74 2 .3 6

% 174a 50% 71% 83% 9% 9Q%

I of displaced chloride on 202 would lead to 192 and 193 after loss of

DMF and S03. The action of DMF on 203 to give 20_4 is an alternative

possibility. However 132 was not isolated as a product of endo addi-

0 .CN

.Cl

203

tion of CSI to 55. In addition the approach of chloride to form 204 would most likely be endo for steric reasons and syn,endo 204 was not

isolated as a product.

All three chloronitriles (191-193) undergo base-initiated cycliza

tion in high yield. In view of the recognized ease of reversible 75 abstraction of C4 benzylic protons under such alkaline conditions,

it is likely that the 4-chloro substituent in 192 is epimerized prior

to being displaced. Such a process may also be operative in 191. 1 1 2

However, the stereochemical requirements are not as rigid here, since an Sn 2' mechanism having precisely the opposite stereochemical require- 116 ment to the direct S^2 pathway can gain importance. This feature is also relevant to the cyclization of 189. It is possible that intramolecular epimerization could occur from 205 to 206 followed by S^2 displacement to give 194. Intermediate 205 is bishomoaromatic.

NC

Cl 194

206

The low yield of II6 may be due to side reactions of 117a and 118a 13 to give N-nitroso lactams from DMF. Silica gel chromatography of the crude product did give a large amount of high retention amorphous solid and oil that was not characterized. The non-lability of the C-N bond of both 117a and ll8a shown in equilibrium studies (see p 69) can be used to rationalize the poor yield of II6. The formation 6f 119 has been previously discussed"(see p 70).

To summarize, the reaction series of CSI addition to barrelenes ijjL,

55, and 99; conversion of the resulting N-chlorosulfonyl lactams to 4- chloro-8-cyanobicyclo[5.2.l]octa-2,6-dienes ll6, I89, 191, 192, and 195 with warm DMF, and finally treatment of the chloronitriles with strong base, has been shown to lead to 1-cyanosemibullvalenes 119, 190, and 194.

Most importantly, this technique has provided a route to the elusive parent compound, 1-cyanosemibullvalene. SECTION IV

Reactions of Chlorosulfonyl Isocyanate with

Bicyclopentane and 1,3-DimethyTbicyclobutane

113 INTRODUCTION

Studies of CSI addition to p-rich a bonds of strained hydrocarbons 1 5 ,1 1 8 ,1 1 9 have been reported. While the reaction of CSI with 1,1,2,2-

tetramethylcyclopropane has been shown to proceed by acid-catalyzed 120 rearrangement to 2,3,3-trimethyl-l-butene prior to cycloaddition,

the electrophilic capture of this reagent by bicyclo[l.1.0]butanes and bicyclo[2.1. 0]pentane have not been comparably interpreted. A cyclo-

propylcarbinyl cation intermediate has often been invoked in the addi- 22,24,25,108 tion of CSI to strained hydrocarbons containing double bonds.

In the case of benzvalene (18^), CSI was seen to first attack the double 115

Relief of strain in 207 is marginally accomplished in a second tran­ sient intermediate (208), which likely reorganizes quickly to 209 and

210. Structure 208 is a cyclopropylcarbinyl cation but must be highly strained since it possesses a bicyclopentyl framework as well.

Zwitterion 209 closes to form 185 while 210 serves as progenitor to 211.

Though the preference of CSI for ir bonds persists in the presence of strained sigma bonds, banana bond breakage need not operate. For 15 example, addition of CSI to 212 leads to 213, in the expected stereo-

c s i + \ \ 7 x ° H

212 213

selective, Markovnikov manner without rupture of the cyclopropane ring.

Having only strained a bonds, 214 nevertheless reacts with CSI to 15 give 31% of 213 and 9% of 216. P-Lactam 216 is considered to be the result of initial rearrangement of 214 to bicyclo[4.1.0]hept-2-ene before reaction with CSI in a fashion analogous to the conversion of

212 to 213. The mechanism for the formation of 213 has been interpreted as the result of backside S„2 attack at Cx with heterolysis of the most ii* strained 1,7 bond to give 217, or similar attack at C^ with 1,2 cleavage

to give 218. If 217 were formed, it might be expected to rearrange to

218. 116

N-H CSI +

214 215 216

C1-02S-N

+ 217 218

When 219 was treated with CSI in dichloromethane at -78°, the pro­ ducts observed after hydrolysis and silica gel chromatography were 220

CSI + H-N

0 219 220 221

-S02C1

0

223 (43%) and 221 (4%). Center bond cleavage would lead to intermediate

222 which is capable of rearrangement to 223 ; edge-bond cleavage alter­ natively can generate 223 directly. Ring closure of 223 through nitrogen and oxygen leads to 220 and 221, respectively. The correspon­ ding addition of CSI to l,2,2,3-tetramethylbicyclo[l.1.0]butane gave 15 an entirely similar y-lactam. RESULTS

121 Re-examination of the addition of CSI to bicyclopentane (22b) at room temperature in dichloromethane (Scheme XIV), followed by reductive

Scheme XIV ,H + CSI ‘0 o

224 225 227

.H

0 226

74 dechlorosulfonylation with alkaline sodium sulfite and silica gel chromatography afforded a lactam product assigned structure 225 rather than 226. (Cyclopentane-5-carboxamide and -3-carboxylic acid have also 1 5 ,1 1 8 been isolated from the product mixture. ) Sublimation and recry­ stallization of 225 gave white needles, mp 47-49° 1757 cm”1). max Moriconi and Dutta reported mp 32-3^° after distillation of 225 and an 1X8 ir stretch at 5-72 p,. 1748 cm"J‘) for the same product. Paquette.

118 _ U s s t 1 15 Allen, and Broadhurst found mp 32-33 and v 1745 cm" . Both ’ e max groups Initially assigned structure 226 to this product. 5 The addition of CSI to cyclopentene (227) has been repeated, and

after workup, sublimation and recrystallization afforded white needles,

mp 47-49°. Bestian and coworkers recorded mp 50-51° and a carbonyl

stretch of 1748 cm-1 for 225. A mixed melting point for the lactam

products of the two reactions was 47-49°. Further support for structure

225 was derived from Eu(fod) 3 induced chemical shift studies of the

nmr spectra of 225 from 224 (Table XIX) and from 227 (Table XX). 122 Recently Jagt and van Leusen synthesized 22'6 unambiguously and found

it to be a white solid, mp 106-109° (1690 cm 1).

In another experiment, addition of CSI to 1,3-dimethylbicyclobutane 123 (228) at -78° in diehloromethane (Scheme XV) and subsequent hydrolysis led

to the previously described lactam (VCHC13 1745 cm-1, mp 74.5-75°) max in 4.3$ yield. Elution of 229 from a silica gel column was preceded

£chemeJ£Y CSI + d t f + CSI 120

Table XIX. Eu(fod)3 Induced Chemical Shifts in the NMR Spectra of 6-Aza-7-oxobicyelo[3.2.0]heptane (225) Resulting from the Addition of Chlorosulfonyl Isocyanate to Bicyclopentane {22k).

Mole ,dcci3 AEu Slope Proton Percent TMS Eu(fod)3 (ppm) (ppm) (AS/Am-fo)

Hx 0 3-^5 -8.7 O.O87 5 3-93 10 4.37 15 4.78 20 5.22 30 6.13 to 6.97 50 7.77

^2endo 0 1.80 -6.5 O.O63 5 2.37 10 2.67 15 2.97 20 3.30 30 3.93 40 4.50 50 5.03

H2exo 0 1.80 -3.6 O.O36 5 1.87 10 2.15 15 2.35 20 2.60 30 2.90 40 3.23 50 3.57

Ha,H3 ' 0 1.80 -2 .1 0.021 H4,H4' 5 1 .87 10 1 .8 7 15 1.93 20 2.08 30 2.43 4o 2.57 50 2.80 1 21

Table XIX. (Continued)

Mole .DCCI3 Proton Percent TMS U °^e Eu(fod)3 (ppm) (ppm) (A6/&n-f°!

H5 0 4.02 -3 .3 O.O33 5 4.18 10 4.33 15 4.50 20' 4.67 30 5.05 40 5-35 50 5.62

Hs 0 6 .0 2 -6 .6 0.065 5 6.50 10 6.92 15 7.18 20 7.52 30 8.23 40 8.80 50 9.27 12 2

Table XX. Eu(fod)3 Induced Chemical Shifts in the RMR Spectra of 6-Aza-7-oxobicyclo[3. 2 . 0]heptane (2 2 $ ) Resulting from the Addition of Chlorosulfonyl Isocyanate to Cyclopentene (227).

Mole ,dcci3 AEu Slope Proton Percent TM3 Eu(fod)3 (ppm) (ppm) (A6/&n-$)

Hi 0 3.45 -7.9 0.082 5 3.42 10 3.88 15 4.33 20 4.80 30 5-70 1*0 6.50

H2endo 0 1 .67 -6 .0 0.061 5 1.82 10 2.17 15 2.50 20 2.82 30 3.45 4o 4.00

H2exo 0 1 .6 7 -2.8 0.030 5 1.33 10 1.80 15 1.93 20 2.02 30 2.38 40 2 .6 7

Ha^Ha' 0 1. 67 - 1.0 0.013 5 1.33 10 1.37 15 1.43 20 1.53 30 I.83 4o 1.97 123

Table XX. (Continued)

Mole ,DCC13 £Eu Slope Proton Percent TMS Eu(fod)3 (ppm) (ppm) (AS / Am-%)

Hs 0 4.03 -2 .1 0.024 5 3.67 1 0 5.85 15 4.02 20 4.20 30 4.52 4o 4.75

He 0 6 .52 -6 .0 0.063 5 6.42 1 0 6 .82 15 7. 22 20 7-57 30 8.27 Q rrO 4o C . | 0 124 by polymeric material. A rerun of the infrared spectrum of 229 ex- CH Cl hibited a carbonyl stretch of negligible difference (v 2 2 1751 cm-1). max Comparison of 229 (from 228) with the lactam product of the addition 124 of CSI to 1,2-dimethylcyclobut.ene (230), which also gave 229. in 197° yield (mp 74-74.5°, vC^ 4 1767 and 1745 cm-1), proved the two products max to be identical. Recently Aue and coworkers also reported the trans- 125 formation of 230 to 229. Their physical data, bp 70° at 0.1 mm; mp 70-72°; v ^ 4 1750 and 1760 cm-1; 6 ^ 4 7.6 (br s, m), 1.6-2.4 (m,

4h ), 1.29 (s, 3H), and 1.16 (s, 3H), correspond closely to those previously detailed.

Die certainty of structure 229 is increased by Eu(fod)3 pseudo contact shift experiments (Table XXl). Although 2^1; the original 15 structure assignment, was not available for direct examination, the syn (or anti) protons at C5 and C6 of 23 1 are chemically equivalent and would be expected to move downfield together with increasing Eu(fod)3 concentrations. The methyl protons of 229 and 2gl do not share the same electronic environment, so without 231 for comparison, conclusions based on the magnitude of £Eu values for the methyl groups must be vague. The actual result of the Eu(fod)3 studies of 229 is that the endo C5 proton (AEu = -6.3 ) moves downfield faster than the other three methylene protons (AE = -3 • 3 ) •

The unsymmetrical ABCD ^ nmr pattern (Figure 3) of the lactam product from 228 or 230 is in accord with structure 229• The methylene pattern of 231 could be expected to be more symmetrical than that ob­ served for 229. The infrared carbonyl stretch is in agreement with a 125

P-lactam. A Y-lactam derivative such as 251 would be expected to dis­ play a carbonyl band at lower frequency. 1 26

Table XXI. Eu(fod)3 Induced Chemical Shifts in the ^ NMR Spectra of 2-Aza-l,4-dimethyl-3-oxobicyclo[2.'2.0]hexane (233).

Mole -DCC13 £Eu Slope Proton Percent TMS Eu(fod)3 (ppm) (ppm) (AS/hm-fo)

1CH3 0 1.28 -2.3 0.023 5 1.40 10 1.52 15 1.63 20 1.75 30 2.00 4o 2.22 50 2.38

h2 0 6.37 -8 .1 0.081 5 6.87 10 7.03 15 7.73 20 8.02 30 8.73 ho 9 .0 0 50 10.37

4CHa 0 1.20 -5.9 0.059 5 1.45 10 1.73 15 2.02 20 2.30 30 2.93 ho 3 .8 0 50 3.97

^sendo 0 2.07 -6.3 0.065 5 1.95 10 2.47 15 3.07 20 3.33 30 3.93 40 4.45 50 5.07 127

Table XXI. (Continued)

Mole .DCCI3 Proton Percent TM3 °^e Eu(fod)3 (ppm) (ppm) (A6/Am-%)

Hsexo 0 2.2.07 °7 -3.3 0.033 1.95 H6, He' 5 10 2.47 15 2.53 20 2. 67 30 3.00 40 3.37 50 3 .6 0 128

CH: h 3c

ll \

3.0 2.0 1.0 0.0 S

Figure 3. The 60 MHz XH nmr spectrum (upfield region) of 2-aza-l,4- dimethyl-3-oxo-bicyclo[2.2.0]hexane (229) in CDC13-TMS. DISCUSSION

In support of mechanism 214 -* 217 -* 2l8 -» 215, the acetolysis and ethanolysis of 232 was found to give 233 , 234, and 233 in the ratio

Q> • Q> • Q OTs OR OR 0R

233 234 235

15:4:1. The predominance of 233 and 234 without internal return and the large rate enhancement relative to the exo isomer of 232 or cyclo- hexyl tosylate indicates the predilection for a 1,2 bond shift in 232 .

This migration is well reconciled with the rearrangement of 217 to 218.

The cation component of intermediate 217 is thermodynamic ally less stable than cyclopropylcarbinyl cation 218, but may be the kinetic result of backside attack at Cx or Cy since the back lobes of the central bond are larger, more accessible, and richer in p character.

Ultimately, rearrangement of transient 217 (if it exists) to 218 must occur to explain the products. Backside attack with homolytic center bond cleavage has been observed in the addition of benzyne to bicyclo- 1 2 7 , 128 butane, 214. and 228 as well as bicyclopentane.

129 130

The addition of deuterioacetic acid to 219 gives 236 and 237 with 129 retention of configuration and high stereoselectivity. Similarly, 130 the reaction of 228 with acetic acid to give 238 can be interpreted

AcO 219 + DOAc

236 237

as a backside electrophilic attack at Cj.. The tertiary cyclopropylcar- binyl cation responsible for 236 and 237 must be stable enough to pre­ clude further rearrangement before quenching or deprotonation. Inter­ mediate 239 is a primary carbocation; its further ring cleavage and

-H

240 OAc 228 HOAc 11 238

rearrangement results in nucleophilic capture by acetate ion at a secon­ dary site. Intermediate 240 may also play a role in the reaction.

The previous discussion provides the basis for explaining the origin of 229 arising from CSI addition to bicyclopentane (228). The 151

CSI + 224 >- 225

0 N-SO2CI -S02C1

241

uni particulate electrophilic backside attack at Cj. of 224 resulting in center bond heterolysis is a process which allows maximum relief of strain, but leaves the N-chlorosulfonyl amidate function remote from the carbonium ion center. If hydride migration occurs before the five- membered ring flattens or flips, 242 is produced. Ring closure and hydrolysis leads to 225. The additions of bromine and chloride to 224 131 is an important precedent. Bromination at -30 to -25° in chloroform in the dark yields 85 to S*Ofo of trans-l,2-dibromocyclopentane. Brid­ ging 1,3 and 1,2 bromonium ion intermediates in the mechanism were postulated by LaLonde.

Ihe possibility of isomerization of 224 to 227 before addition of 15 CSI was considered in the Paquette group's initial study. The reaction was monitored with nmr spectroscopy and no olefinic proton absorptions were ever detected. Very small concentrations of 227 cannot be totally discounted however.

Contrasting observations that electrophiles can add 1,3 without rearrangement (hydride migration) are also well documented. The addi- 132 132 133 tion of mercuric acetate, thallium triacetate, lead tetraace- 1 3 3 ,1 3 4 tate and acetic acid in the presence of sulfuric or p-toluene- 135 sulfonic acid to 22 k all proceed with central bond cleavage to give

1,3 adducts and substituted cyclopentenes. A wide variety of addi­ tions to 22k involving free radical mechanisms also occur via backside 136 attack at Ci and center bond homolytic cleavage.

There are three apparent mechanistic possibilities for appraisal in the addition of CSI to 228. First, 228 could be catalytically converted to 230 prior to addition of the reagent. However no such 137 isomerization has been observed during silver catalyzed rearrangement, 130 130 130 pyrolysis, acetolysis, or additions of iodine, carbenes and/or 130 128 138 carbenoids, benzyne, maleic anhydride, and hexafluoroacetone.

The other possibilities arise from the outcome of electrophilic attack at Ci* If as previously discussed (e.g., CSI + 219 and HCAc + 228) central bond cleavage occurs, 2kp results, while edge-bond cleavage gives 2kk. Both pathways are plausible if 2^3 and 2kk can equilibrate.

N-SO2CI < 0 + ^ N-SOsCl 0 > N-SO2CI 2k3 0 2kk 2k5 133

Intermediate 244 can further rearrange to 245, the immediate precursor of 229. Two related mechanistic processes may he operating in the very fast reaction of acetic acid with bicyclobutane (246) to give a 1:1 139 mixture of 247 and 248.

+ HOAc CH2OAc P OAc

246 24-7 2 4 8

140 The recent work of Saunders and Rosenfeld is relevant to this issue. Entry into the manifold 249 £ 230 has been accomplished from

c r 1 — c f — X — IX ,.

251 249 250 252

either direction in superacid (1H nmr studies). The equilibrium con­ stant was found to be 50 at room temperature. At low temperature, reaction of 251 or 252 with antimony pentafluoride gave 24-9 which iso- merizes on warming to give predominantly 250. Since the addition of CSI to 228 is performed at -jQ°t 229 is a reasonable product to expect since intermediates 243 and 245 are probably favored at low temperatures. One could therefore expect 229 and/or 231 as products of the addition of 13b the addition of CSI to 228.

Whether central or edge bond cleavage occurs initially is a moot question. Nevertheless, edge bond cleavage is the simpler process to explain all the observations of CSI additions to bicyclobutanes. Nmr studies have shown that edge and corner protonation of cyclopropanes are 141 competitive equilibria. In addition, Wiberg and coworkers observed a marked decreased in electrophilic reactivity when a cyclopropane ring is replaced with a cyclobutane ring, even though the strain 139 , 142 energies are quite similar. The additions of CSI to 22b and 228 are entirely consistent with these earlier findings. The reaction of

22b was slower since it contains only one cyclopropane ring while 228 contains two. Furthermore, both reactions occur with an initial cyclo­ propane bond cleavage. SECTION V

The Synthesis and Electronic Properties of Bicyclo[4.1.1]octane,

Bicyclo[4.1. l]oct-3-ene, and Bicyclo[4.1. l]octa-2,4-diene

135 INTRODUCTION

Our final endeavor has been the synthesis of bicyclo[4.1.1]octane

(253), bicyclo[4.1.1]oct-3-ene (254) and ultimately bicyclo[4.1.1]- octa-2,4-diene (255,), in order to record their spectral properties especially those from photoelectron spectroscopy. The interactions

253 255 of the Walsh orbitals of the 1,3 cis fused cyclobutane ring with the diene moiety of 255 relative to 253 and 254 is of special interest.

Several 1,3 cis-fused cyclobutane rings in alicyclic systems are known. 143 144 145 Meinwald and coworkers have synthesized 256, 257? 258, and 145,146 145 146 14V 259. Meinwald and Zimmerman found that 258 and 259

256 257 258

136 thermally rearrange to dihydrosemibullvalene and semibullvalene (l,jk,

R=H) respectively. Sigma MO's of the cyclobutane rings of 258 and 259 must certainly interact with the n systems during these rearrangements.

Both norpinane (260) and 2-norpinene (261) have been prepared from 14a 262 and 263 . The interaction of the cyclobutane sigma bonds with

260 261 262 263

the carbene or carbonium ion generated by treating the tosylhydrazone of 262 with various bases leads mostly to rearrangement products involving participation and migration of the adjacent cyclobutane bonds rather than elimination product 261.

A few substituted bicyclo[4.1.l]octane systems are known. Dibromo- carbene has been added to a - p i n e n e and the resulting 2\6k thermally rearranges with elimination of HBr to give a 4:1 mixture of 265 and 266, 138

149 respectively. The chlorine analogs of 264-266 have also been 149 , 150 150 reported. Joulain and Rouessac have prepared 267-269. In addition, 270 was synthesized by an acyloin condensation, the only method that proved successful in the synthesis of 253-255*

267 268 269 270

Photoelectron spectroscopy (EES) has become a very useful tool for determining the energy levels of the atomic and/or molecular 151 orbitals (MO's) of relatively small organic molecules. This technique has provided such experimental data by the direct measurement of the energy required to remove an electron from the valence shells of atoms and molecules. The results are in remarkable agreement with

MO energy levels derived from theoretical calculations. In return, such calculations have been used to make energy band assignments and interpret the PE spectra of more complex organic molecules. (Some authors have used PES to suggest that Spectroscopic Potentials-adjusted

Intermediate Neglect of Differential Overlap (SPINDO).calculations are preferred for interpreting and predicting the observed MO energy levels 152 of through-bond and through-space interactions. ; The MO interactions of different segments of an organic molecule can be observed by the shifts in energy levels in a series of related systems. Symmetry control, energy matching, and orbital overlap have important effects on 139 these interactions.

According to the Einstein photoelectric law (E = hv-l), the energy, E, of an electron ejected from a molecule is equal to energy of the photon used minus the ionization potential, I, the energy necessary to remove an electron from a given orbital. The observed ionization energies in a EE spectrum correspond to the differences in energy between the ground state of a given MO and the radical cation states that result when an electron is removed from that orbital.

If the geometry of the ground state and the excited state radical cation remain unchanged during ionization (Born-Oppenheimer Approxi­ mation), the magnitude of the ionization energy is equal to the energy of the atomic or molecular orbital in the absence of electronic reorganization in the radical cation (Koopman's Theorem).

When the photoionization of a neutral molecule occurs, the result­ ing EE spectrum may also indicate a Jahn-Teller distortion in cases where an electron is ejected from one of a set of degenerate MO's.

The photoelectron spectra of allene, which is isoelectronic with ketene, exhibits this phenomenon as a double maximum contour in the 154 first ionization band (it region). Ketene does not show Jahn-Teller effect because its two it bonds are quite different.

Since a number of excited vibrational states are populated in

the radical cation, vibrational fine structure is often observed in

EES. The EE spectra of unsaturated organic molecules, especially

those with rigid skeletons, often express good resolution of vibrational

fine structure (a Franck-Condon profile) in the various energy bands. iko

Tn larger, more conformationally flexible and less symmetrical mole­ cules, the large number of valence electrons give rise to a large set of excited radical cations whose ionization potentials overlap.

This causes loss of resolution of fine structure. The jt bands are usually resolvable until the onset of the myriad a bands.

The conjugative (jt-tt) and hypereonjugative (n-ff) interactions of the MO's of organic molecules have been shown to change drastically with variations in geometry. The energy levels of the two it MO's in 3»6-bridged 1,^-cyclohexadienes 271 “* 2.jk reverse as the dihedral 155 angle between the two ethylene units increases. The interplay of symmetry controlled through-bond (hyperconjugative) and through-space 155,156 (homoconjugative) interactions has been discussed by Heilbronner.

As the dihedral angle and the distance between the p-rt orbitals 141 increase, the through-bond interactions increase at the expense of through-space interaction. The interactions of the strained cyclo- butane a bond Walsh orbitals with the jt MO's of Dewar benzene (275) 157 make it impossible to correlate it with the others. Other examples of 32 such interactions are the barrelene series (41-44, see page 16), 34 the homobarrelene series (see page 18), where cyclopropane Walsh orbitals play an important role, and the conjugative interaction of the it and Walsh cyclopropane orbitals of bullvalene (170, R=H), 156 dihydrobullvalene, tetrahydrobullvalene, and hexahydrobullvalene.

The study of conjugative interactions between a cyclobutane ring and an adjacent vinyl group or sp2-hybridized center has attracted 34, 158-163 increasing interest recently. A suitable method for the investigation of such interactions is provided by photoelectron spectroscopy. The observed band positions in various EE spectra have been satisfactorily reproduced within the framework of a Zero Differ­ ential Overlap (ZDO) model by adopting a resonance integral for the

P = < P 1tlH|pw> = -1.9 eV (1)

interaction between a 2p atomic orbital on the double bond (prt) and a

2p atomic orbital on the cyclobutane ring (pw) separated by one carbon- carbon single bond.

The common component of 253 , 254, and 255 is the 1,3 cis-fused disubstituted cyclobutane ring. There are no photoelectron spectra available on a species containing this moiety, but the EES spectrum of 142

164 cyclobutane Itself reveals two high-lying Walsh type orbitals with vertical ionization potentials of 10.7 and 11.3 eV. These energy values are the result of a Jahn-Teller distortion of two degenerate orbitals and jzf2 of orbital symmetry 3eu if cyclobutane is assumed 159,165,166 to be square-planar (D4ft) or of orbital symmetry 4e if (as 167, 168 experiment indicate ) a puckered (D2d) symmetry is assumed for 167,170 cyclobutane. The next highest carbon-carbon g-bonding orbital

(^3) that could interact with a it MO is of symmetry group lbx (or lblg

if the cyclobutane ring is planar), and was found to have an energy

of -12.5 eV. An even lower energy (-15.9 eV) MO, (4, is of little

interest here.

5*1 (Was) ^2 (Wsa) lkj>

Calculations for a set of degenerate (and unoccupied) orbitals and ^2*) at 15.2 eV and another ($3*) at 17.0 eV for 1 6 3 , 169 planar cyclobutane, are not of low enough energy to interact with the occupied it MO's of 254 or 255.

A study of the effects of cyclobutane substitution in a cis 1,2 34 fashion in tricyclo[4. 2.2.02,5]decyl systems showed the expected shift of the eu Walsh orbital energy levels to 1.7 eV higher energy because of inductive effects. Bruckmann and KLessinger noted that the interaction of the cis fused cyclobutane ring of 276? 277? and 278 were similar in nature but lower in magnitude than for quadricyclo-

[3 .2.2.02j4]nonane (66) and homobarrelene (45 ) presumably because of

276 277 278 less p character in the cyclobutane Walsh a orbitals relative to cyclopropane. A second representation of the degenerate cyclobutane Walsh orbitals involved are fa and fa. In another 1,2 cis-fused

1TX cyclobutane series, Heilbronner and coworkers showed that in both

syn-and anti-tricyclo[^. 2.0.02’5]octyl systems, e.g., 279-281? the

interactions of the it systems are through-bond, namely the cyclobutane

279 280 281

a bonds, and to a lesser degree through-space. Again in the series

279 -» 281, the high-lying Walsh orbitals exhibit considerable involve­

ment in lowering the energy of fa and fa which are mathematically 159 equivalent to fa and fa. In the series 276-278, the highest Walsh

orbital, es, dropped from -9*^5 to —9-55 to -9 .8 eV while the ea

orbital went from -9*9 to -9*95 to -10.3 eV. Similarly in the 279-281

series, the au c orbital energy levels were -9*23, -9-86, and -10.13 eV

while the lower bu q values were -9.66, -10.06, and -10.57 eV,

respectively. The stronger interactions of the latter system are

evident from the larger energy drops. 145

The ultraviolet spectra of 258 and 259 were early examples of measurement of the interaction of a cis 1,3 disubstituted cyclobutane 160 ring with a x bond. All three high-lying cyclobutyl MO's

^2, and jz43) interacted significantly with the re MO's of 259. In 258 , only one of the two degenerate Walsh orbitals ( ^ and $2 ) can interact in a symmetry allowed fashion with the double bond. The calculated and observed x -* n* and Walsh cr -* jt* transitions for both compounds exhibit high levels of interaction. The photoelectron spectra of 257 162 and 258 also indicate the strong interactions of the Walsh and x orbitals in 258. The first two ionization potentials of 257 are 9* 78

(e) and 10.45 (e) eV, and were assigned to Walsh orbitals (x and f2.

For 258, the values were 8.63 (b2, x-$2), 10.59 (hi, ^1), 11.54 (b2, x + (2), and 12.10 (a2, $3 ) eV. The energy difference due to Jahn-

Teller distortion of the Walsh orbitals of 257 was 0.6 7 eV, very close 164 to the 0.55 eV difference for cyclobutane. Ample precedents exist for considerable interaction of the cyclobutane ring with the cr and x systems of the four carbon bridges of 253 , 254, and 255»

In general, photoelectron spectra of saturated cyclic and bicyclic 172 hydrocarbons like 253 give no striking features because of the overlap of ct bands. For cyclohexane and norbomane the onset of these 173 bands occurs at approximately 10 eV. For cyclooctene and benzo- 173, 174 cyclooctene the a bands begin at 10.0 and 10.54 eV, respectively.

The PE spectrum of cyclopentene shows two x band peaks of nearly equal intensity at 9*01 eV (adiabatic) and 9*18 eV (vertical) due to a large Frank-Condon factor (an ionization resulting in two vibrationally 146

173 different radical cation excited states.) A similar phenomenon is possible for 254 because of the rigidity of its structure. Cyclo- hexene exists in a more flexible half-chair conformation and exhibits 1 7 2 ,1 7 3 only the vertical transition at 9*12 eV. Cycloheptene is similar with a it band at 9.04 eV. Norbomene and bicyclo[2. 2.2]octene 172 (4^) have it bands at 8.97 and 9*05 eV, respectively. These cis disubstituted cyclic and bicyclic olefins give an indication of what is expected for 254.

There is a possibility of symmetry-allowed interaction of a Walsh orbital (fa) with the occupied it orbital in 254. An energy difference of ~ 1.7 eV, assuming the it MO energy level to be -9.0 eV, is rather wide for a strong interaction. The most probable interaction of fa and it will be hyperconJugative (through-bond) involving itCH2 MO's of the sp3 hybridized centers at C2 and C5, since the relevant Walsh and it orbitals are held too far apart for strong through-space interaction 155 as was the case for 1,4-cyclohexadiene (274). Models predict the dihedral angle to be 135° for olefin 254. and ~ 145° for 274. 175 155 The experimental values for 274 are 159° and 1^3 » suggesting steric and electronic repulsion of the it clouds. The syn C7 and Ca protons of 254 should exhibit even greater steric repulsion, from the it bond forcing the molecule into a nearly planar conformation. Such a conformation would also minimize steric repulsion of the hydrogens on

Ci through C6. The best through-bond interactions in 254 would occur if Ci, C2, C3, C4, C5, and C6 were all in the same plane as shown in

282 and suitable orbitals were available. The a-bond energies 14-7 should be about 0.4 eV lower upon interaction with a double 15 6 ,1 7 2 ,1 7 6 bond. Puckered conformation 2o$ suggests a predominant through-space interaction, but ignores steric repulsion.

0

282 283

The expected splitting of energy levels for the two it MO's of a conjugated diene results from a linear combination of atomic or bond­ ing orbitals. 'When two double bonds are brought into conjugation the average cr center of gravity of the mixed it bonds (7^ and it2) is close to that of an isolated double bond. The PE spectra of 151 151,177 ethylene (jt = -10.51 eV) and butadiene (^1 = -9*07 eV and it2

= -11.46 eV) illustrate this phenomenon.

* 1 * 2

Cisoid conjugated dienes have a slightly greater first ionization 178 potential than transoid dienes, but the magnitude is overpowered by the inductive effects that result from attempts to hold the diene cisoid. Methyl-, phenyl-, and methoxy- substitution on butadienes 148

179 raises the it MO energy levels relative to butadiene. Bringing a third ethylene unit into conjugation increases the first it band ioni­ zation potential to ~ 8.3 eV in the case of cis and trans-1,3 ,5- 180 hexatrienes.

Cyclopentadiene is an atypical cyclic, cisoid conjugated diene, 180 presumably due to hyperconjugation through the ftcH2 orbital at C5.

The jti and it2 vertical ionization potentials for cyclopentadiene,

1 ,3 -cyclohexadiene, and 1 ,3 -cycloheptadiene are 8.58 and 10.74, 8.25 and 18 1 , 182 10.75, and 8 .3 1 and 10.63 eV, respectively. Assuming that the largest difference in tcx and it2 energies signifies the greatest conjugative interaction and thus the most planar diene, then 1,3 - cyclohexadiene is the best model for comparison with 255. Actually, electron diffraction studies have shown that the double bonds of 1,3 - 183 cycloheptadiene are coplanar, while the double bonds of 1,3 -cyclo- 1 7 5 , 184 hexadiene are canted by 17 to 18.3 . Itofortunately, most examples of bicyclic conjugated dienes, for which photoelectron 1 8 2 , 185 spectra have been reported, also contain other it interactions 18 6 or have exocyclic double bonds. Recently, the EE spectrum of 187 quadricyclo[4.3*!•O1’6]deca-2,4-diene was recorded. The first three ionization potentials are 8.12, 9* 42, and 10.3 eV.

If one assumes that Cx, C2, C3, C4, C5, and C6 of 255 all lie in the Px plane, the molecule must contain two planes of symmetry and be in symmetry group C2V (see Figure 4). The other symmetry plane, P2, bisects the C3 -C4 bond and passes through Cr and C8. The symmetry

groups and symmetries (S = symmetrical, A = anti symmetrical) with 149

Figure 4. A diagram of the symmetry elements for t>icyclo[4.1. l]octa- 2,4-diene (255).

respect to Pi and P2 are for b^ and AS; for jr2, a2 and AA; for

bj_ and AS; for j^2, b x and AS; and for f3i a2 and AA. The symmetry-allowed interactions of 255 are (x with rtj. and jt2 with /3.

The ^2 MO is of the wrong symmetry to interact with any of the avail­ able MO’s. Structures 284 and 285 illustrate the symmetry-allowed 150 interactions possible in 255. Note there are no nodes for 284 and

two for 285. Because the expected energy difference is smallest between itx and this interaction should be the strongest.

Interactions of a diene and a cyclobutane ring different from 160 those in 255 have been studied in the case of 286, 28J, and 288.

In the case of 287 and 288, analysis of the interaction between the

it and Walsh orbitals was complicated because two Walsh orbitals

and $3 ) had to be considered. Furthermore, in the region below 10 eV, interpretation of the EE spectrumwas difficult due to strongly

overlapping er-bands. The interactions of the Walsh orbitals of the

u v/\ rC ^ \ O/ \

286 287 288

cyclobutane ring and the diene in 288 were found to be comparable to 1 6 3 , 188 those of spiro[4.4]nona-2,4-diene (289), but less than those 189 observed for spiro[2.4]hepta-4,5-diene (homofulvene, 290) or 190 fulvene (291), because of the greater p character of cyclopropane

00 cx o-

289 290 291 151 bands and it bonds of the latter two examples. Nevertheless, this is good EE spectroscopic evidence for direct conjugation of the Walsh orbitals in a cyclobutane ring with the it-orbitals of adjacent double bonds.

The symmetry correlations of 288 are different from those of

255. There was no symmetry-allowed interaction between jt2 of 288 and any of its high-lying cyclobutane orbitals. There were strong interactions between itx and (1 and fi3, resulting in vertical ioniza­ tion potentials of 8.38 (a2, jt2), 10.12 (bx, itx), 10.80 (a*, $2)9

11.22 (bx, ^x)j 331(1 12.58 eV (bx, ^3 ). For both 255 and 288 the degeneracy of and 3 is removed, but for reasons of symmetry it should be possible to study separately the interaction between itx and ^x an<1 between it2 and 3 in 255.

Hie conformational flipping of the cyclobutane ring of 288 should have little effect on energy levels of the mixed MO's, since both conformers are equivalent. The same holds for the conformations of the cyclopentadiene ring of 288. The cyclobutane ring conformation of 255 9 25^, or 255 will be frozen in only one puckered non-planar conformation. The deviation from planarity of the cyclobutane ring of these hydrocarbons may not be identical, but the variations should be insignificant and allow comparisons in their EE spectra. Experi- 191 mentally, cyclobutane was found to be puckered by 20 to 37°* 1 9 1 ,1 9 2 Calculations predict a deviation from planarity of 7 to 32 .

Calculated carbon-carbon hybridizations for cyclobutane, bicyclo[l. 1.1]- pentane (292), and the cyclobutyl bonds of 257 are sp3*38, sp3*43 to 152 1 9 3 , 194 sp3*59, and sp3,46, respectively, an increase in p character with increasing strain.

292 293 294

The cyclobutane ring puckering of 292 is necessarily 60° from 194 planarity. The divergences from planarity found for bicyclo- 195 196 197 [2.1.1]hexene (256) were 56.5°, 53*3°, and 50.5°. For 293 195 the value was 55»0°. The electron diffraction study of bieyclo- 198 [3 .1.1]heptane (294), which exists in a fast-flipping boat confor- X4S mation, showed that the Ci-Cy bond length was 1.553 + 0.009 ft

(the reported carbon-carbon bond length of cyclobutane was 1.548 + o 167 0.003 A.) and the cyclobutane ring of 294 is 43 from planarity, not very far from the 35° average value for cyclobutane. Therefore, the

puckering of the cyclobutane ring in 253 , 254, and 255, is expected

to be about 40°. The effect on Walsh MO fz (b2) will be small relative to inductive effects but a significant effect on overlap of

the p orbitals of ^ (bx) may result. RESULTS

The synthetic approach to 253, 254, and 255 starting from penta- 199 erythritol (295) made use of the work of Allinger and Tushaus 200 (Scheme XVl) and Russell, Whittle, and Keske (Scheme XVIl). The

Scheme XVI

(CH2Br)2 Q C(CH2OH)4 ----->- C (BrCH 2)2 ^ qH ^ (ch 2o h )2 295 297 296

(H°2C)2 O C o> - ^ > (i-C5H11-02C)2 < ^ X ^ ° ) - < g )

299 298

< (H02C)2 o (co2h ). % 300 301

153 154 documented overall yield was 11.4$ from 295 to 301* Repetition of the sequence produced an 8.8$ yield of 501 (see Experimental). Hie 200 reported yield of $02 from 501 was 6l$, much less than the 92. 6$

Scheme XVII

301 >-0Si(CH3)

0Si(CH 3)3

302 303

found herein. Diester 502 was converted to 503 via an acyloin con­ densation, and the product was readily purified by Florisil chromato­ graphy. Thus the overall yield from 295 to 303 was 6.8$. The crude

Scheme XVIII

0 301 3QT 155 byproduct (3 Ob) from the thermal bis-decarboxylation of 300 ? could be esterified (305, R = -CH2CH3 ), epimerized in methanol and sodium methoxide (305 and 306, R =-CH3), saponified (3 Oft- and 307) and dehydrated to give 301 and 30^ (Scheme XVIIl). Synthetic intermediates 199 ,2 0 0 30b through 307 (Rs = -CH3 and -CH2CH3) are known compounds.

All attempts to synthesize 253, 25^, and 255 were derived from

303 or its hydrolyzed form 308 (Scheme XIX). The conversion of

Scheme XIX

503 508 (R = OH) 310 309 (R = OAc) acyloin 308 to ketone 310 involved the initial formation of the a- 201 acetoxyketone (309), followed by reduction with calcium in liquid 202 ammonia. Reduction of 309, with activated zinc dust and mercuric 203 chloride in acetic acid did not work. Unfortunately, ketone 310 could not be used to obtain significant amounts of olefin 2pb or bicyclo[4.1.l]oct-2-ene via enol phosphate formation and reduction 2 0 4 ,2 0 5 with lithium in liquid ammonia or tosylhydrazone formation and 69 treatment with methyllithium. In fruitless attempts to use the modi- 206 fied Perkow reaction to form enol phosphates, the mesylate of 308 207 was prepared and heated in triethyl phosphite, and treated with 156 208 209 trimethyl phosphite in methanol and acetic acid.

An original synthetic approach as well as the shortest and fastest route to 2pk (Scheme XX) involved treatment of 505 in THF with

Scheme XX

— k / \ 0Si(CH3)3 1 / \ OP(OCH2CH3)2 0Si(CH3)3 0P(0CH2CH3)

505 511

greater than two equivalents of methyllithium and quenching the 2X0 resulting dianion with excess diethyl chlorophosphate to give a

78$ yield of 511. Characterization of 511 was incomplete but the vague nmr and ir evidence (see Experimental) suggests structure 511 because of nmr integrations and multiplets typical of ethyl phosphate esters and the absence of a carbonyl band. This material could not be distilled or chromatographed. The reduction of crude, dried 511 with 204,205 lithium and t-butyl alcohol in ammonia at -78° gave 8.1$ of

2 after vapor phase chromatography.

All other synthetic approaches to 2pk (Scheme XXl) first required reduction of 505 or 50^ with sodium borohydride in ethanol, resulting in a good yield of cis-bicyclof^. 1. l]octane-5;^-diol (512)» The assignment of cis stereochemistry to 512 is required by the W C nmr 157

Scheme XXI

OH OH

315

spectrum (including single frequency off-resonance decoupling). There

are three secondary carbon resonances, (TMS =0.0 ppm) a two-carbon

signal at 35*6 ppm (C2 and C5) and two one-carbon signals at 31*1 (Cq )

and 29.9 (C7) which are respectively anti and syn to the hydroxyl

groups based on shielding effects. If 312 was a trans diol, the C7

and Cb chemical shifts would be equivalent. In either stereochemical 158 case Ci and C6 (32. k ppm), C2 and C5 (35*6) and C3 and C4 (7k. 2) would have been equivalent. After examination of models, it was evident that intramolecular hydrogen bonding in the infrared spectrum of either the cis or trans diol is possible. Further nmr, ir, and mass spectral data can be found in the Experimental Section. Chemical evidence, i.e. cyclic phosphate and thioearbonate formation, may also be ambiguous.

Low yields of 2.5k were obtained from two reactions using di­ mesylate 313 , which was prepared from 312 at 0° using methanesulfonyl 211 chloride in dichloromethane and pyridine. Reduction of 315 with 212 sodium anthracenide in THF at 0°, aqueous workup, pentane extraction, fractional distillation of solvent, vacuum transfer, and vapor phase chromatography gave k% of 25k contaminated with a double elimination product, possibly 255. Only enough material for a mass spectrum of the mixture was obtained. Alternatively, treatment of 513 with sodium 213 iodide and activated zinc in HMPA at 100 for 2k hours followed by aqueous workup, pentane extraction, percolation through alumina, fractional distillation of the pentane and gas chromatography gave 10$ of 25k. 214 The recently developed method of Marshall and Lewellyn was applied in the conversion of 312 to 25k. Treatment of the diol with 215 N,N-dimethyl dichlorophosphoramide in pyridine and THF produced a

97$ yield of crude 31 k, which could be neither distilled nor chromatographed. The XH nmr spectrum of the oil (see Experimental) exhibited a series of overlapping multiplets suggesting a mixture of 159

syn and anti isomers (at phosphorus) of 314. Reduction of this cyclic phosphoric amide with lithium in anhydrous liquid ammonia gave 14$ of 2^4 after quenching, pentane extraction, aqueous workup,

fractional distillation of solvent, percolation through alumina, and vapor phase chromatography.

The last route to 2^4 from 312 made use of the Corey olefin 2ie synthesis. Thioearbonate 3JL5 was obtained in 93% yield from 312 and heated in distilled triethyl phosphite at 150° for 90 hours to give 18% of 234 after a typical workup and gas chromatography which was necessary to separate 254 from the excess triethyl phosphite before further use. Column chromatography on alumina, silica gel

or Florisil was not capable of removing the solvent. Hie destruction

of a large portion of the available 254 during gas chromatography was

intolerable.

Hie overall yields for the five approaches to 254 were 6.1% via

311 , 4.5% via 313 using sodium anthracenide reduction, 5. &!<> via 315 using zinc-sodium iodide reduction, 10. & j o via J l h and 9*0$ via 315.

The lower isolated vpc yields were avoidable in all but the final

instance. Marshall's method was utilized for the most part because

it was the least discouraging pathway to the diene (255). The best

overall yield of 254 from pentaerythitol (2 9 5 ) was 0.7$,

The 1H nmr spectrum of 254 is ambiguous because of the overlap

of complex multiplets in the up-field region. There is a narrow

down-field multiplet for the olefinic protons (6 5* 57? J < 2 Hz) and

more complex absorptions due to the bridgehead (2.48) and allylic l6o

(2.26) hydrogens. "The non-symmetrical distribution of the remaining four cyclobutyl protons [~2. 48 (1H) and 1.03-1.57 (5H)] suggests a boat conformation for 254 which results from the deshielding of a bridge proton. The variable temperature features of this spectrum were not examined. The 13C nmr data for 254 will be presented together with the data for 253 and 255.

Bicyclo[4.1. l]octane (253) was prepared by hydrogenation of 254 and by Clemmensen reduction of 508. (Scheme XXIl). Atmospheric

Scheme XKII

OH 254 253 308

hydrogenation using ten percent palladium on carbon in ether or methanol at room temperature was slow but effective and gave a 75% yield of 253 after vapor phase chromatography. The reduction of 308 gave a 17% yield of 253., however; the crude product after pentane extraction, aqueous workup, percolation through alumina, and concentration by

fractional distillation unexpectedly contained three parts 254 to seven

parts 253 by nmr and vpc analysis. Hydrogenation of the mixture as

above gave 25£ only as a low melting white solid. The best route to 253

from 295 was via the Clemmensen reduction (1.2% overall yield). l6l

The ^ nmr spectrum is definitely not first order and is little more than a mass of overlapping multiplets (see Experimental). The infrared spectrum of 253 is consistent with a saturated bicyclic hydrocarbon without primary hydrogens. An accurate mass measurement of 253 (m/e = 110.10976) agreed with the calculated value (110.10954) for a CsHi4 bicyclic hydrocarbon.

Realization of the elusive bicyclo[4.1. l]octadiene (255) was a torment. Double elimination of the cis diol (312) using Burgess' 217 reagent failed. Attempts at a double elimination of dimesylate 313 218 219 using potassium t-butoxide in DMSO, or DBU in THF, or activated 205, 220 alumina in dichloromethane were all futile. Yet another 221 approach combined the work of Sealey and McElwee and Sharpless and 222 Lauer and involved conversion of the diol (312) to its benzaldehyde

acetal (316) which on treatment with NBS in carbon tetrachloride gave

*- OH OH

316 31?

> ° H * * - h < > -

320 319 318 162 bromoester 517? which was transformed into epoxide 318. It was hoped that attack of sodium phenylselenide on 3j-8 would lead to allylic alcohol 519? which in turn could be oxidized to ^20. Ketone 520 would 204,205 be a suitable precursor for the diene (255) via the enol phosphate 69, 223 or the tosylhydrazone. Unfortunately and inexplicably, the organoselenium reagent reverted 518 to 512.

Another approach resorted to was bromination of the olefin (254) 224 with pyridinium hydrobromide perbromide, followed by elimination.

The trans dibromide was readily obtained and in a vain effort treated 219 with DBU in warm THF. The dibromide was recovered by gas chromato- 225 graphy and treated with potassium t-butoxide in DMSO and THF to

yield not 255, tut 4.2% of 254 after vpc'.

The one successful route to 255 came by free radical allylic 226 bromination of 254- with KBS in refluxing carbon tetrachloride. The

reaction could not be stopped at mono bromination to give 321 , but

continued to give dibromide 322 (Scheme XXIIl). The use of one equiva-

Scheme XXIII

Br 254 > Br

321 322 255 lent of WBS gave a mixture of 25k and 322. Use of more than two equivalents of NBS was not detrimental. The crude product (322) was dissolved in pentane and percolated through silica gel eluting with pentane to give 8k% of 322 as a colorless oil. This oil may consist of a mixture of cis and trans dibromides. The 1H nmr spectrum of the dibromide(s) exhibited two olefinic protons in a narrow multiplet at

6 5.73, and two allylic protons in an unresolved multiplet at 4.93.

A 13C nmr spectrum of 322 was not obtained, so the stereochemistry must remain ambiguous. An accurate mass of 322 could not be measured in the mass spectrum but the accurate mass for loss of bromide ion to leave carbonium ion [CsHio 79Br]+ was observed at m/e = 184.99696

(calculated to be 184.99663). Additional evidence for loss of one bro­ mide ion was a second peak at m/e 187 of equal intensity for [C8Hxo 81Br]

Before finding a successful way of converting 322 to 255, 322 227 was treated with potassium iodide in acetone, activated zinc in 228 229 refluxing THF, and activated zinc dust in DMF and ether. A trace of 255 was obtained using sodium carbonate, sodium iodide and activated 230 zinc in DMF at 100° or using potassium iodide in HMPA at 100°. Only when a concoction of zinc-copper couple, potassium iodide, iodine and DMF was used was 255 isolated in reasonable yield (18$) after gas chromatography. The best overall yield of 255 was 1.6$, starting with 505 a11'1 0.1$ starting from pentaerythritol (295). The actual yield is probably higher when 2pk is not subjected to gas chromatography.

There are at least eleven steps in the synthetic series involving the intermediacy of 314. 164

As shown in Table XXII, 253, 254, and 255 must possess some sym­ metry since only four signals of approximately equal intensity were observed for each compound. This trait was especially important in making the structure assignment of 254, since the 2,3 double bond isomer of 254 is less symmetrical than 254. Because only four chemical shifts are present in the fully decoupled 13C spectra of 233-255., their conformational geometries demand that on the average C2,

C3, C4 , C5, and C6 lie in one plane or contain C2 symmetry, otherwise one would not expect C7 and C8 to be equivalent. A single frequency off-resonance decoupling (SFOBD) experiment was used to make the assignments of chemical shift for the olefin (254). The olefinic carbons were very similar to the olefinic carbons of 1,4-cyclohexa- 23 1 231,232 diene (6 124.5)5 cyclohexene (127.2), and cyclopentene 232 (130.6). By comparing the chemical shifts of cis diol 312 with 25j5. and 254, and matching resonances for C* and C6 (6 32.4, 31*6, and 31*8, respectively), C2 and C5 (35-6, 3^*0, and 35*6, respectively), and

Cy and C8 (29.9 and 31*1? 29.9, and 30.3> respectively), reliable assignments were possible. Die values for C3 and C4 of 312 (6 74.2),

253, and 254 were quite divergent and easily assigned. The 13C chemical shifts of cyclopentane, cyclohexane, and cycloheptane are 233 25.6, 26.9j and 28.5 ppm 6, respectively, and are similar to the value of C3 and C4 in 253.

The chemical shifts of 255 were easy to assign, but unusual.

Ci and C6 are allylic and appear somewhat downfield. The difference in chemical shift (14.9 ppm) of the two types of sp2 hybridized carbon in 165

Table XXII. Data from the 13C NMR Spectra of Bicyclo[4.1.1]octane (255), Bicyclo[4.1.l]oct-3-ene (25^0, and Bicyclo[^. 1.1]- octa-2,4-diene (255) in Deuteriochloroform Solution (TMS =0.0 ppm).

Chemical Shift Carbon Center S3. 254 255

Ci,C6 3 1 .6 3 1 .8 3^.8

C2jC5 3^.0 35.6 138.6

C3 JC4 25-5 125.8 123. T

Cy,CS 29.9 30.3 21, 255 is very large and may indicate unusual electron density distri­ butions. Hie corresponding olefinic carbon chemical shifts of some other planar or nearly planar dienes are for cyclopentadiene 8 132.8 234 233 and 132.2, for naphthalene 128.1 (Ci) and 125.9 (C2), and 132.1 233 (C2) and 122.1 (C3) for quadricyclo[4.3.l.O1’6]deca-2,4-diene (326).

Hie chemical shift of Cy and C8 in 255 is very close to that of cyclo- 233 butane (6 22.4).

The XH nmr spectrum of 255 was non-first order and consisted of a 4-proton M'BB' olefinic multiplet at 8 5.67-6.53 , and 3 two-proton multiplets at 2.73 to 3 *28, 2.12 to 2.73 ? and 1.08 to 1.63 assigned to the bridgeheads, the anti methylenes and the syn methylenes, respectively (see Figure 5). With irradiation of Hj. and H6, the AA'BB' pattern of H2, H3, H4 , and H5 was recorded at 100 MHz. The olefinic chemical shifts and coupling constants for 255 were solved with an 235 iterative least squares computer program, LA.OCOON III. The rms error was +0.1 Hz. The results are listed in Table XXIII along with data for 1,3 -cycloheptadiene (323), 7,7-dicyanonorcaradiene (324)? 323?

326 , 327 , 328 , 329, 330 , 1,3 -cyclohexadiene (331)? 1,3 -cyclooctadiene

(332 ), and cyclopentadiene (333 )» 6.0 3 . 0 2.0 1.0 0.0 8

Figure 5* The 6o XH nmr spectrum of bicyclo[4.1. l]octa-2,4-diene (255) in CDC13 -TMS. H CF\ Table XXIII. Chemical Shifts (ppm 8) and Coupling Constants (Hz) from the NMR Spectra and Known or Estimated Dihedral Angles (0) Between the Planes of the Double Bonds for some Cyclic, Bicyclic, and Tricyclic 1,3-Dienes.

Chemical Shift Coupling Constant Dihedral Compound Angle Ref. h 1,h 4 h2,h 3 J l , 2 J l, 3 J l , 4 J 2 ,3 (9)

251 6.22 5.88 11.4 1.0 0.7 6.9 0° This Work 323 5.6 to 5.75 11.6 0.7 0.7 6.9 0° 183,236 324 A6 0.25 9.4 0.8 1.1 6.2 0° 237 323 6.36 6.16 9.3 0.7 1 .2 6 .0 - 238 326 6.07 5.71 9.3 0 .6 1.3 5.9 0° 238,239 3,27 5.63 6.01 9.5 0.8 1.0 5.9 - 239 328 5.27 5.70 9.7 0.7 1.1 5.5 10° 238,239 329 A6 = 0.21 9-7 1.0 1.0 5.5 0° 240 330 5.31 5.72 9.6 0.9 0.9 5.4 - 238,239 331 5.68 5.79 9.7 1.0 1.1 5.1 17° 184,238,241 332 5.59 5.79 11.3 -0 .6 0.5 4.1 40-65° 241,242 333 6.28 6.43 5.1 1 .1 1.9 1.9 0° 241,243

H o\ CO For a series of conjugated cyclohexadienes, Bothner-By and Moser have suggested that when 3 JS (or J2 3) exceeded 5-88 Hz, the two double bonds of the diene were probably coplanar (0 =0°). This idea 241 has been reiterated by others. In a series of conjugated mono- 243 cyclic dienes, cyclopentadiene (333 ) although planar, gives an anomalously low J2 j3 °f only 1.9 Hz because the strained bond angles of the five membered ring require different coefficients in the modi- 245 fled Karplus equation, Js = A + B cos 9 + C cos 2 0, where 0 is the dihedral angle between the double bonds or between H2 and H3.

1,3-Cyclohexadiene (331) has been found to have a dihedral angle of approximately 17°, and thus a J^,3 value well below 5*88 Hz at 5.1. 183 The diene unit of 1,3-cycloheptadiene is planar and shows coupling constants remarkably similar to 255. This is a strong indication that the diene unit of 255 is planar as a model shows. 1,3-Cyclooctadiene 242 (33£) is doubtlessly nonplanar with a small J2,3 of 4.1 Hz. It can also be inferred that the diene unit of 324 is planar, since 102 Fritchie found the diene of 2,5-dimethyl 324 (200) to be planar.

Other estimates of 0 in Table XXIII are postulated. 1 200 250 300 350 nm

Figure 6. The ultraviolet spectrum of bicyclo[4.1. l]octa-2,h- diene (255) in hexane. 171

Table XXIV. Data from the Ultraviolet Spectra of some Cyclic, Bicyclic, Tricyclic, and Spirocyclic Conjugated Dienes.

Compound X (nm) Extinction Solvent Dihedral Ref. max Coefficient Angle (e) (e)

255. 200-213 1600 hexane - 0 ° This 258 (shoulder) - work 266 2800 277 3800 288 3750 301 2000

288 257 >1500 pentane 0° 246 261 1950 ethanol 247

289 254 2200 ethanol 0° 248,249

290 257 2000 ethanol 0° 248

525 248 7400 isooctane 0° 250

324 271 2900 cyclohexane 0° 101

525. 258 4900 isooctane - 251

326 249 2800 0° 252 255 2800 273 2950

327 243 7900 - - 253

328 235 4700 cyclohexane 254 285 2600

331 256 8000 hexane 17° 255 257 44oo cyclohexane 256

332 228 5600 cyclohexane - 257

333 200 10000 hexane 0° 258 239 3400 234 2800 ethanol 248 172

Table XXIV (Continued)

Compound \ (nm) Extinction Solvent Dihedral Ref. max Coefficient Angle (e) (e)

212 hexane 0° 188 269

218 5350 ethanol 249 276 1100

555, 184 60000 hexane 0° 259 20k 7900 25 6 200 (230-270)

234 ethanol 260 238 239 243 249 255 261 268

336 274 3300 cyclohexane 261 173

A comparison of ultraviolet spectroscopy data (Table XXIV) for cyclopentadiene (333 , 200 and 239 nm)? 1,3 -cyclohexadiene (331 , 256 nm), 1,3 -cycloheptadiene (323 , 248 nm), and the diene (233, 27T nm)

shows a strong red shift similar to that in going from 333 to spiro-

[4.3]octa-5,7-diene (288, 257 nm). Based on ring strain in the mole­

cular models the jt -* rf** transitions of 255 without interactions with

the cyclobutane ring might be expected between the values given for

331 and 323 . The surprising vibrational fine structure observed for

255 (see Figure 6 ) is a strong indication of structural rigidity and

symmetry as observed for 526, spiro[4.4]nonatetraene (334), which also

exhibits strong diene-diene interaction, and benzene (335 )» which

exhibits a highly resolved band in the 230-270 nm wavelength range

strikingly similar to the 240-310 nm band of 255 (see Table XXIV).

Cd Go O Ch

337 338 339

The only other bicyclooctyl hydrocarbons that might conceivably

fit all the spectroscopic data for 253? 254, and 255 are 336? 337? 338 ,

and 332.. The vinyl protons of 338 in the 40 MHz XH nmr spectrum 262 resonate as a multiplet at 8 5.82, while the multiplet of 255 at

100 MHz is centered at 6.10 (Figure 5). However, 33jS exists in thermal 261 equilibrium with l,3 ?5-cyclooctatriene and cannot be isolated 174 discretely. The bicyclo[4.2.0]octa-2,4-diene-iron tricarbonyl coin- 263 plex has been prepared and 336 has been isolated and irradiated in o 264 an argon matrix at 20 K. Bicyclo[4.1.l]octa-2,4-diene (255) is thermally stable. Olefin 254 possesses a vinyl 1H resonance at 8 5.57 265 •which is nearly identical to the 5>58 ppm signal for trans-357.

However, the infrared data for 254 and 557 are different. The vinyl 266 protons of 538 (6 5*96) are also at variance. Hydrogenation of

254 gave 255« The assignment of structure to 253 depends on the structure of 254. There was not enough diene 255 to hydrogenate.

The EE spectra of 253-255 a-re shown in Figures 7-9 and the relevant EE data are listed in Table XXV together with results of molecular orbital calculations, all thanks to Rolf Gleiter and Eeter

Bischof at the Technischen Hochschule Darmstadt in West Germany.

®@© ©©

111 cc< 1- z 3 o o

6 7 8 9 10 11 12 1314 15 16 17 18 19 20 I.P. (eV)

Figure 7- The photoelectron spectrum of bicyclo[4.1.l]octa-2,4- diene (255). 1T5

Ul £ a. H z 3 O O

19 20 I P. (eV)

Figure 8. The photoelectron spectrum of bicyclo[4.1.l]oct-3-ene (25^).

T6 0 1 0 11 12 13 14 15 15 IT 18 19 20 I . P . ( e V )

Figure 9. Hie photoelectron spectrum of bicyclo[i|-. 1.1] octane (255)• 176

Table XXV. Comparisons of the Vertical Ionization Potentials (i^ ) and Calculated Orbital Energies of 255, 254, and 253.

— Orbital energies-- s Corapd Band Assignment ZDO MINDO/3 EH ■STjJ

I8,11? ® 18 .29( 2sL2(n-a) - 8 .1 0 - 8.54 -12.05 (8.48J

© 10.15 5b2 (it-w) -10.14 -10.06 -12.99 255 © 10.72 5b2(w) -10.6 -10.08 -12.95 © 11.13 4b^( Jt+w) -II.56 -11.65 -15.97 © 12.1 15ax(CT) -10.41 -13.40 © 12.5 lagCii+a) -12.6 -12.61 -14.30

8.90 9.08 -12.54 Q < > 6b1( jt) - 9-42 - 9.24

V. 9.40'' J 254 - -12.88 © 10.15 9b2 (w) -10.07 - -13.08 © 10.75 5b1(w) -10.37

- -13.40 © 11.15 15a1(a) -10.61

© 10.0 l6b (w) -10.08 - -12.86 255 © 10.5 15b (w) -10.19 - -12.91 (o>) © 11.0 15&(CT) -10.55 - -15.50 1 7 7 153 In this study, use was made of Koopmans' theorem. In this approxi­ mation the orbital energy gj is set equal to minus the measured

e = -I (2) J V,J vertical ionization potential IV, j. This allows correlation of the bands of the EE spectra with orbital energies obtained from model calculations. A Zero Differential Overlap (ZDO) model and semiempirical 267 methods of the Extended Huckel (EH) and Modified Intermediate 268 Neglect of Differential Overlap (MINDO/3) type have been used for the interpretation of the EE spectra. For the ZDO model, the wave functions and basic orbital energies of the butadiene and cyclobutane 162,163 fragments are needed. These have been given in earlier publications, 166,269 and are briefly repeated here. The relevant ir and Walsh orbitals

(iti, tt2, ^2? and were illustrated above (see pages 14-2 and 147).

The corresponding wave functions are:

+ P j) + 0*602 (p . + p ) ♦1 = 0.5T2 (Pxa *xd vtxb xc Bi (3)

- P „) + 0.372 (p. - p ) a 2 ♦2 - °-6 0 2 (P*a Fxd vtxb *xc' (*)

01 = 0 .5 (

■ 0 .5 (d - Ai (6) 02 vtye ^yf + pyg " ^yh^

= 0.5 (p + a 2 (7) 03 ye Pxf “ Pyg " Pxh^ 178

In these equations the are pure p^ atomic orbitals, while the

are sp” hybrid atomic orbitals parallel to the |j,-axis. The

irreducible representations according to which these wave functions

transform in the point group C2v are listed following these equations.

Only the occupied it and Walsh orbitals are being considered. The 159,166,269 lowest occupied Walsh cyclobutane orbital (aig in D^h) is

omitted since, for reasons of symmetry, it does not interact with

either of the occupied it orbitals of the butadiene moiety.

To construct the corresponding wave functions of 255 the atomic

orbitals at centers a and d of the butadiene moiety have to be com­ bined with those at centers f and h of the cyclobutane ring. The

choice of a unique wave function in the cases of j^x to is difficult.

Semiempirical calculations, however, suggest that the orbital density

is equally distributed over all four centers. This implies that

alkyl substitution should only slightly remove the degeneracy of fix

and If it is assumed that the first two ionization potentials in

25^ are due to the removal of an electron from the Walsh orbitals

corresponding to fix and this hypothesis is experimentally verified.

The split (AI = 0.6 eV) is in line with that found in cyclobutane 164 162 (0.55), methylcyclobutane (0.57), and tricyclo[3.5-0.02’6]- 162 octane (257, 0.67). The basis orbital energies for cis-butadiene 162,163 have been discussed previously by Bischof and Gleiter.

g (tt2) = -8.40 eV (8)

s(jtx) = -10.90 eV (9) 179

Assuming as in the case of 288 that the inductive shift of the Walsh orbitals realized upon introduction of a butadiene unit amounts to 0 .2 163 eV, the following basis orbital energies are obtained.

e(ji) = e(rfz) = -10.8 eV (10)

eOfe) = -12.3 eV (ll)

To set up the interaction matrix, values for the resonance integrals

Hjj must be estimated. Equations (3 ) and (5) give

Hjtw ~ < *l|Ht ^ = 2’°5 ' 0*372 (12)

Eq. (4) and (6) yield

= < % |h| 5^3 > = 2-05 • 0.602 (13 )

while substitution of Eq (l) into (12) and (13 ) provides

H = -0.707 eV (14) itw

H = -1.144 eV (15) IlCT

This treatment leads to the following secular equations for the inter­ acting basis orbitals belonging to the irreducible representations

Bi and A2. i8o

-8.4-e -1.144 -10.9-e -0.707 = 0 and -1.144 -12.3-e -o. 707 -10.8-e

By solving these secular determinants, the following eigenvalues are obtained:

ei(a2 ) = -8.1 eV

62(^2 ) = -12.6 eV 1 11 6 H 63 O^i) • eV 1 II H ON 64 (^1) H • eV

As can be seen from Table XXV, these values are very comparable

to those measured for bands (T), (2), (4) , and (&). The chosen basis

orbital energy for the highest occupied Walsh orbital of the cyclobutane ring (e = -10.6 eV) corresponds to the value measured for band (5) •

The vibrational fine structure encountered in band (l) (see Figure 7 and . 173 Table XXV) is very similar to that seen in 1,3-cycloheptadiene. In

the latter compound, the vibrational spacing is reported to be 0.19 eV;

in 255 a value of 0.18 eV is observed.

The only large difference between the ZDO results and the experi­

mental values is found for e^bi), which differs by 0.43 eV from -1^. ^

(see Table XXV). This discrepancy could be due to the interaction of

this orbital with a lower lying C-H a orbital of Bi symmetry which leads

to the observed destabilization. The results of the ZDO calculations

are summarized in Figure 10. In this figure an interaction diagram

has been constructed between the occupied it orbitals of butadiene on l8l

t j M

- 8 a a2

- 9

-1 0 b 1

b2 e f

12

a 2

-13

a

Figure 10. Orbital interaction diagram for bicyelo[4.1.l]octa-2,4- diene (255) based upon perturbation theory. 182 the left and the three highest occupied Walsh orbitals of cyclobutane on the right, The orbital energies are taken from the ZDO calculations.

The wave functions have been drawn schematically. This diagram nicely illustrates that 255 is an ideal molecule for the study of interactions between jt systems and highest occupied Walsh orbitals of cyclobutane.

Jn hydrocarbons 25k and 2J55, no conjugative interaction with n orbitals is possible and thus direct interpretation of the first bands of the EE spectra is uncomplicated. The assignment given for bands (T) and

(2) in the EE spectrum of 255 agrees very nicely with data from a 162 series of other alkyl substituted cyclobutane derivatives.

The assignment of the first three bands of 25k on the basis of 153 Koopmans' assumption is also straightforward. The value of its first ionization potential (8.90 eV) is close to the first ionization 173 potential of cycloheptene {9.0k eV) and cyclooctene (8.98 eV). The vibrational spacing of 0.16 eV observed in its first band is also identical to that observed for cycloheptene and cyclooctene.

The ionization potentials for bands (2) and (3 ) of 25k, which are due to the ejection of an electron out of the highest occupied Walsh orbitals, are lowered by about 0.2 eV compared with the Walsh orbitals of 253. This is due to a dominance of the inductive effect of the double bond over the hyperconjugative effect through the C2 and C5 CH2 groups. A similar lowering (-0.18 eV) is observed by comparing the center of gravities of the rt-orbital energies for cyclohexene (9-12 eV) _ 172 with 1,4-cyclohexadiene (lv = 9*3 eV). The Walsh orbitals for 233 and 162,164 25^ were split by 0. 6 and 0. 5 eV, respectively, which is typical. 183

To test the analysis given above, a comparison of the experimental 267 268 results with EH and MINDO/3 calculations was made. Since the detailed structures of 233 to 255 are not known, their geometries have 268 been optimized within the MINDO/3 scheme. For this optimization a 270 modified Fletcher-Powell search procedure was used. The heats of

formation so calculated are: 253 , +6.06 kcal/mol; 254, -18.74 kcal/mol;

255, -49.09 kcal/mol. Figure 11 indicates the geometrical parameters obtained for the carbon skeletons. The cyclobutyl bond lengths obtained 198 for 2j>3-2%? are the same as those found for bicyclo[3.1. l]heptane (294), but the cyclobutane ring of 294 is more puckered (137°). The average 168 non-planarity for cyclobutane is 145°. For 253, 254, and 255, C2,

Cg, and C2v symmetry, respectively, was assumed throughout the geometry

optimization procedure in order to reduce the large number of variables.

U5J

1.91 147 139 1.99 1.93 1.53 1.9!

25? (Ca) Sit Cc2v) 255 (Cgv)

270 Figure 11. An illustration of the Fletcher-Powell computer optimized structures for 253., 254, and 255. 184

Molecular model studies indicate that these results are reasonable.

Although 234 was assumed to have only one plane of symmetry, the structure converged to the higher symmetry C^y. This result was quite unexpected. However, it is in line with the known tendency of MINDO/3 271 to underestimate puckering amplitudes of hydrocarbon rings. In the case of 2^3, C3 and C4 are situated 0.23 2. above and below the plane defined by C1? C2, C5, and C6.

The structural parameters realized for the cyclobutane segments in 253 to 233 closely parallel those derived from electron diffraction 168,272 and X-ray experiments on cyclobutane, bicyclo[2.1. l]hexane 195,197 196 (293), bicyclo[2.1.l]hexene (236 ), bicyclo[3.1*l]heptane 198 273 (294), and tricyclo[3 .3 *0.02,6]octane (237). DISCUSSION

The aim of this study has been to elucidate the orbital inter­ actions talcing place between the butadiene and cyclobutane units in

255. A measure of the magnitude of this interaction is given by the atomic resonance integral inEq. (l). Interestingly, the value 162,163 found for this parameter not only confirms our previous results, but is seen to be identical to the p value found for the interaction 34, 156, 162, 163, 3,90,274 between a cyclopropane ring and a double bond. The interaction integrals, H , between a double bond and a cyclopropane or a cyclobutane ring, however, are quite different due to the wide numerical divergence in the coefficients of the atomic orbitals in the corresponding wave functions. This gap in the interaction integrals

2 H (cyclopropane) = P*c *c = -1.9 *0.707 = *<1.1 eV irw it w /IP

H (cyclobutane) = p.c -c = -1.9*0.5*0.707 = -0.67 eV JTW It W is a measure of the difference in the ability of the particular ring to stabilize a it system, and is largely the result of the greater p character of cyclopropane bonds. The numbers are entirely consistent with the outcome of semiempirical calculations on the cyclobutyl- 159 carbinyl system.

185 186

Hoffmann and Davidson's EH calculations for a cis-1,3 - cyc lobutyl dicarbinyldication predict a bisected-bisected geometry, analogous to that found in 255 (/]_ and itx), to be the most favored conforma- 159 7 / tion. In this instance a linear combination of px and P3 was

possible. The symmetry elements of 255 preclude such mixing. Using 275 the Complete Neglect of Differential Overlap (CNDO) method, Wiberg

also found the bisected or out-of-plane conformation of the cyclobutyl 276 cation to be the more stable conformation. Hehre has examined

the rotational energies of vinylcyclobutane by ST0-3G ab initio MO

calculations and found that s-trans conformation 340 (a bisected

interaction) to be 2 kcal/mol lower in energy than gauche rotamer 3/t-l which was 0.3 kcal/mol lower than s-cis conformation 342. The energy

340 341 342

difference between 340 and g4l may possibly be less than 2 kcal/mol

since the n bond of 341 can interact with nearly as well as 277 interacts with the it bond in 340 and ^42. Shanshal has also dis­

cussed the cyclobutylcarbinyl cation and vinyl cyclobutane using a

simple perturbational molecular orbital method not involving computer

calculations. 278 The chemical evidence for the ability of the cyclobutane ring

to stabilize an ar-carbinyl cation is not as striking as the effect of a 187 cyclopropane ring partly "because of the smaller coefficients of the p orhitals of and ^2 relative to the Walsh orbitals of cyclopropane.

Tn addition, x and f2 offer two rotational conformations 90° apart for overlap and thus exhibit less conspicuous conformational preference relative to cyclopropane.

The high 13C chemical shift (6 138.6 ppm) at C2 and C5 of 233 relative to other cyclic dienes may be an indication of unusually low electron density at these sites. The electron density drawn into $x should increase the antibonding repulsion of the C7 and C8 p-type orbitals leading to a distortion of the cyclobutane ring. At the same time the bonding interactions in fx would increase, moving Cx and

Cg closer together. The optimized structure of g55 has the smallest

^CsCjCg (108°) versus 1110 for 25^ and 112° for 25^ and the least puckered cyclobutane ring (see Table XXVIl). These angle changes seem to counterbalance each other. Using the bond angle and bond length data from Figure 11 and simple trigonometry, it is possible to show that Cj. and C6 are about 0.05 ft closer in 255 (2.12 ft) than in

25ft- (2.17 ft) assuming Cx through C6 lie in one plane for 25ft- and 255.

The Cy-Cg internuclear distance is 0.03 ft larger in 255 than in 25ft-

(see Table XXVl). The angles in the cyclobutane rings conform to these distortions. This distortion of the cyclobutane ring would destablize j b y pushing its antibonding orbitals on Cj and C6 closer together. The pinching effect of the sp2 hybridized carbons of the diene also forces C* and C6 together. However, the angles at C3 and

C4 of planar 25ft- are strained more (10°) than in 2^5 (6°). The angle 188

Table XXVI. A Comparison of the Carbon-Carbon Internuclear Distances for Planar 25b and 2.55 Derived from the Data in Figure 11 and Table XXVII.

Distance ($) Nuclei , Difference 25b 255

CiC2 1.53 1.51 0.02

C1C3 2. 6l 2.55 0.06

C1C4 3.12 3.10 0.02

C1C5 3.07 2.96 0.11

CiC6 2.17 2.12 0.05

C^Cy 1.55 1.55 0.00

C2C3 1.50 1.35 0.15

C2C4 2.58 2.51 0 .0 7

C2C5 3 .2 7 3 .0 6 0.21

C2Cy 2.61 2.58 0.03

C3C4 1.3^ 1.47 -0.13

C3C7 3.15 3.16 -0.01

CyCg 2.12 2.15 -0.03 189

Table XXVII. A Comparison of the Bond and Cyclobutyl Dihedral (9) Angles of the Carbon Skeletons of Planar 254 and 255 Taken or Derived from the Data in Figure 11 and Table XXVI.

Angle Nuclei Difference 254 255,

CXC2C3 119° 126° -7°

CXC6C5 1110 108° 5°

C xCyC6 89.0° 86.5° 2.5°

C2C3C4 130° 126° 4°

C2C XCy 116° 115° 1°

C7CXCS 86.3° 87.7° -1.4°

0 147° 144° 3° 190 strain at C2 and- C5 of 234 may increase the p character of the orbitals and thus increase the overlap of with the it and MO's of 25b.

The interactions of f$x ^1 (284-), and and jt2 (285) as also illustrated in Figure 10 are somewhat like a dihomobenzene where the four p-type orbitals of lx and a-type orbitals of ^3 do the work of two p-jr orbitals. There are unfortunately eight electrons involved in mixing four MO's rather than the six or ten necessary for aromati- city. The carbon-carbon bond distance of benzene is 1.39 X. The average bond distances of Cx through C5 for 25 5r and 25b are 1.44 and

1.48 ft, respectively.

Some other interesting molecules to study by photoelectron spectroscopy for the interactions of cyclobutane and cyclopropane

Walsh orbitals with it systems are bicyclo[2.1. l]hexane (2.93.) Edd­ ies bicyclo[2.1.l]hexene (256), although the results are predictable.

The results of similar interactions in [4. 2.2]propella-2,4-diene (^4^),

343

[4.2.l]propella-2,4-diene (344), and [4.1. l]propella-2,4-diene (343) are a few possibilities that might not be so readily foretold.

[4.3. l]Propella-2,4-diene (326 ) has received some attention recently. 191

Tricyclo[4. 2.2.01’s]decane has been synthesized and is fairly 142,280 stable. Bicyclo[4. 2.0]octadiene (3pp) as well as 337 and 35^, would make an interesting study except that 336 exists mostly as its 261 valence tautomer 1,3,5-cyclooctatriene. Since neither 326 nor 281 bicyclo[4.1.0]hepta-2,4-diene-l,6-dicarboxylic anhydride (346)

0

3 46 347 348

rearrange, tautomerizations of 34^, 344, or 343 are not expected. As

a means of preparing 345, ^47 might serve as an intermediate in a

double elimination of 2,5-dibromobicyclo[4.1.l]oct-3-ene (322). The

replacement of the cyclobutane ring of 255 with a bicyclobutane ring 282 results in 345, or tricyclo[5.1.0.02,8]octa-3,5-diene (348). Meinwald 283 was unable to convert 349 to 348. However, Prinzbach has thermo-

X 350, X = 0 352, X = 0 351, X = C(CN)2 353, X = C(CN)2 192 lyzed quadricyclanone (350 ,) and quadricyclene 351, to give 30 to 35 percent of 252 and 55 to 65 percent of 353? respectively. These recent results present several synthetic possibilities for the pre­ paration of 3 ^-8.

Some bicyclobutane orbitals that would be of proper energy to 165 interact with a diene are, in order of increasing energy: pa

^CH^’ (ctCH£’ acc^ ^aCHJ ^ 8 ^gcc^J ^acc» ^cc^’ 811(1

(crcc#> nc *). The similarities in symmetry between ^, ( y , and

and $2, $3, and 5^1**, respectively (see page 1^2 ), are obvious, although the relative energies for bicyclobutane > y) are opposite

* e

h K 193 of that for cyclobutane (j^x > ^3 )* For 3^5 a «x would interact some­ what with fig and strongly with ^ and n2 would overlap well with ^5.

For 5j+8, jtx would mix well with and possibly with j^y*. The puckering of the bicyclobutane ring would keep the antibonding orbitals at Cy and Cq of ^y* remote from the p orbitals of C2 and C5 of rtx*

This interaction would be net-stabilizing as an overlap of occupied / * and unoccupied orbitals but the energy separation of and py in

3^8 is unfortunately very large. The same is true for a HOMO-LUMO interaction of ^3 or with Tt3* of 3jj-5 or 3J+8, respectively. For of ^k-8, has the right symmetry and energy for considerable inter­ action. There is no symmetry allowed interactions for fQ in or 348. 284 The difference in Diels-Alder reactivities of and 333 is chemical evidence of the strong interactions of the endo,endo-2 ,k-

354 355

bicyclobutane and 1,3 -cis-cyclobutane moieties with cis-butadiene and cis-2-butene. The higher reactivity of 35jj; with tetracyanoethylene

(TCNE) is a result of the greater stabilizing effect of the Diels-Alder adduct of 35^« One of the highest occupied molecular orbitals (HOMO,

) and the lowest unoccupied molecular orbital (LUMO, ^y*) of bicyclo­ butane are of the proper symmetry and energy for strong stabilizing interactions with the it bond LUMO and HOMO of a 2,3 -bonded cis-2-butene unit. The symmetry allowed overlap of it2 and in is a HOMO-HOMO interaction and is net-destabilizing (overlap repulsion).

For 3J?3, the overlap of the LUMO of cis-butadiene (rt3'K’) and the

HOMO of the cyclobutane ring (j^) is stabilizing relative to the dominant interaction of the cis-2-butene moiety (jt) with which is 161 an overlap of two filled orbitals (overlap repulsion). The energy difference between fix and ft* or fix and it is too great for much of a stabilizing interaction. There is a stabilizing interaction of fi3 and it* in the Diels-Alder adduct of 335 and TONE but it is relatively small. An analogous HOMO (^3.)-LUMO (:%*) stabilizing interaction in

255 was not observed in the EE spectrum. There was, in fact, a large destabilization observed for the fourth ionization band of relative to the calculated energy of the itx + fix MO (see Table XXV and Figure 10).

356 357 256

EH calculations on the isodesmic reactions of 336 "with ethylene

to give benzvalene (l8j+) and butadiene and of 331 with ethylene to

give 236 and butadiene showed the former reaction to be exothermic by 161, 284 17 kcal/mole and the latter by o kcal/mole. These data indi­

cate that both cyclobutane and bicyclobutane overlap better with the cis-2-butene moiety than with the cis-butadiene unit. The advantage in the overlap with the bicyclobutane orbitals is the smaller energy differences between the interacting orbitals. For gV? and 8_ the perturbation of the it bond of the diene moiety increases the energy gap between and it2 relative to a simple double bond. The same is true for itx and and itx and ^>y in g^t-8. Therefore less interaction should be observed for and 3,^8.

The reaction of bicycloC^. 1. l]octa-2,4-diene (25J.) with dieno- philes presents some engaging possibilities. For instance the Diels-

Alder reaction of 25j> with maleic anhydride, dimethyl acetylene- 285 dicarboxylate and a-chloroacrylonitrile would yield ££8, 359? and

^ -^\_^C02CH3

c o 2c e 3 err 0 359 p60. Basic hydrolysis of 3^0 or treatment with sodium sulfide in hot

ethanol gives 361 . hydrolysis of 358 and bis-decarboxylation leads to V V V 562. The concerted (it2s + ct2 s + e^s) thermal decomposition of

361 , or 362 would produce dimethyl phthalate, phenol or benzene, respectively, and possibly bicyclobutane. The formation of these aromatic compounds could provide the driving force necessary to over- 286 come most of the strain energy in bicyclobutane. The difference in heats of formation between bicyclobutane and cyclobutane is

^5.5 kcal/mol. The difference in strain energies is 39*6 kcal/mol.

The reaction of 2£5, with dimethyl acetylenedicarboxylate to form ^ 9 is especially interesting since this reaction lends itself to a one-step or flow system process without isolation of 35^ but rather may lead to dimethyl phthalate and bicyclobutane directly.

Upon examination of molecular models, the steric congestion around

C2 and C5 of 25£ is obvious and may prohibit the approach of the dieno- phile. The calculated C2-C7 distance (2.58 %) in Table XXVI is only slightly less than the syn H7-C2 intemuclear distance observed in

the model. The shielding of the syn C7 and C8 protons in the 1H nmr

spectrum of 2£j? (see Figure 5) also indicates the close proximity of

the syn C7 and C8 protons to the it clouds. The steric repulsions of

the cyclobutyl methylenes and the ethano and etheno bridges in £58j

35^, and g60 would be even greater and mitigates the Diels-Alder

reactivity of 2 5 If these intermediates could be prepared, the

steric effects would provide further impetus for the thermal decompo­

sitions of 559, g6l, and 362 .

Strong steric influences are observed in the sodium borohydride

reduction of acyloins 3P3, ^ ^08 give cis diol 312 (see Scheme XXl). 197 284 288 Relatively unhindered ^6j3. and cyclic 1,2-diketones respond to sodium borohydride to give mostly trans 1,2-diols. Because of the 289 290 greater steric crowding around the carbonyl of and 36 ^, the

OH OH

363 364 363 sodium borohydride reduction gives predominantly cis diols. The for­ mation of only cis diol ^12 on reduction of 308 _ is an indication of the hindrance encountered by reagents that approach C3 or C4 of

3,4 disubstituted bicyclo[4.1. ljoctyl systems. The most successful reactions of this system have involved attack at the substituent rather than at C3 or C4 (see Schemes XX and XXl).

Since both ^£3 and ^08 give only cis diol, it is likely that in sodium borohydride and ethanol 30 ^ is first converted to 308 _by nucleophilic attack of ethanol or ethoxide on silicon to give enol

366 and/or a-trimethyIsiloxy ketone ^67. Ethanolysis of ^67 leads to

^ >-0Si(CH3)3 — > -OSi(CH3)3 > -OSi(CHe)3

OSi(CH3)3 qu

303 366 367 367

OH

.OH OH OH 312

acyloin 308_. Of the two possible configurations (308 a 308b), the axial hydroxyl in 308 a must experience more steric repulsion from the syn C7 or Ca bridge hydrogen than from the equatorial hydroxyl group in 308 b. Thus conformer 3i2§Jl ^svored. Approach of boro­ hydride on the least hindered side of the carboxyl in 308b, anti to the hydroxyl group, leads to cis diol 312 .

The treatment of enediol (bis)-trimethylsilyl ether 303 . with 210 two equivalents of methyllithium (Scheme XX) to form dianion 368 is an example of a chemical reaction at the periphery of the molecule.

Attack of methyllithium at silicon liberates tetramethylsilane leaving 291 the lithium enolate anion. Quenching at oxygen with the first 204,205 equivalent of diethyl chlorophosphate would give 3^2.* The second equivalent to add would give 311 . Both processes are the result of attack at the least hindered site. Proton nmr integration ruled out displacement of ethoxide at phosphorous to give 3,70 or 372 via resonance isomer 371. There was also no carbonyl absorption band in the infrared spectrum of crude 311 . 199

*- iLi

OLi

368

q _ p — och2ch3

3TO

311

371

372

The reduction of 311 with lithium and t-butanol in ammonia also does not require the solvated electron to approach the highly hindered interior of the molecule. If intermediate anion 37^ could not he readily protonated to give 273, may decompose to which is doubtlessly unstable. Enol phosphate 374 can undergo a second normal reduction to give the desired olefin 254. The low yield of this process can be rationalized by steric hindrance and the possibility of 200

0 - v M 311 V - op(och2ch3)2 op(och2ch3 ); Li

3T3 37*1-

375

the transient formation of Further examples of successful reactions that occur away from C3 and C4 and result in the formation of a C3 -C4 double bond are shown in Scheme XXI.

Die reason for the failure of cis diol 512 to undergo a double 217 dehydration with Burgess' reagent or dimesylate to undergo a 205,218-220 double elimination is probably conformational. Molecular models predict 3,^-disubstituted bicyclo[4.1. l]octyl systems to adopt

376a 376b 201

a "boat conformation rather than a chair conformation (jj6a).

The repulsion of an axial substituent at C3 by CT and the syn hydrogen

on C7 will be severe. The dihedral angle of an axial substituent at

C3 in 376 a and the trans hydrogen and C2 is an ideal 180°. The

dihedral angles of the equatorial substituent at C4 of 276a are about

60° as are all the C2-C3 and C4-C5 dihedral angles in 2T6b.

Finally, one recent alternative for synthetic access into the 242 bicyclo[b. 1.1]octane ring system has appeared. The acetolysis

of endo,endo-tricyclo[3 . 2.1.02’4 ]octan-6-yl brosylate (^Tj) at 85°

+ 5 other products

BsO 377

gave 31 percent of 378 as the major product. 'There were unfortunately 293 five other acetates isolated from the reaction. Synthesis of 37J

begins with the Diels-Alder reaction of cyclopentadiene with cyclo-

H0

HO 380 propene to form 5J2, and its exo isomer. Hydroboration to give exo alcohol ^80_, oxidation of J>80_ with chromium trioxide in pyridine to give ketone ^§1, and reduction of %8l with lithium aluminum hydride to give 582 led to 577.

Acetate pyrolysis of should result in bicyclo[4.1. l]octa-2,4- diene formation. Treatment of 3JQ with strong base would probably abstract the allylic proton at C4 and cause loss of acetate to yield

252,. Saponification of 578 would lead to homoallylic alcohol 58^.

OH

Oxidation of 585 could produce either 584 or 58^. Both ketones and

385 would also be suitable precursors of 255. EXPERIMENTAL

All melting points were determined in open capillaries. Melting

points and boiling points are uncorrected. The microanalyses were per­

formed by the Scandinavian Microanalytical Laboratory, Herlev, Denmark.

The proton magnetic resonance spectra were acquired on Varian A-60A,

Varian HA-100, Jeolco MH-100 and Bruker HX-90 Spectrometers. 13C

Nuclear magnetic resonance data were procurred on the Bruker HX-90.

Infrared spectra were recorded on Perkin-Elmer Model 137 and k67 spectro­ meters. Mass spectra were obtained on an Associated Electrical Indus­

tries, Ltd. Model MS-9 spectrometer at an ionizing potential of 70 eV.

Ultraviolet spectra were taken in ethanol on Cary Models 14 and 15. Gas-

liquid phase chromatographic separations were accomplished on an Aero­

graph Model A-700 instrument.

Benzene was stored over sodium wire. Diethyl ether was dried by

distillation from benzophenone-sodium ketyl. Pentane and hexane were

redistilled. Tetrahydrofuran (THF) and dioxane were distilled from

lithium aluminum hydride prior to use. Dimethyl sulfoxide (DMSO),

dimethyl formamide (DMF), and hexamethyl phosphoramide (HMPA) were

routinely distilled from calcium hydride. Dichloromethane was filtered

through neutral alumina (Activity i) and stored over activated ^ £

molecular sieves. Chlorosulfonyl isocyanate (CSl) was distilled under

nitrogen and stored only for short times in polyethylene bottles before

203 204

use. Organic solutions were dried over anhydrous, powdered magnesium

sulfate unless otherwise noted.

anti-Tricyclo[3- 2. 2.02’ 4 ]non-8-ene-endo-6,7-dicarboxylic Anhydride 36 (386 ). A mixture of 40.0 g (0.43 mol) of distilled cycloheptatriene,

43.2 g (0.44 mol) of maleic anhy­

dride, and 100 ml of xylene was

I I refluxed for 5 hr, cooled and

/ V y n diluted with pentane (200 ml) to

0 give 67.6 g (81.9$) of crude 386 . 386 Recrystallization from carbon

tetrachloride afforded 56.4 g (68.37°) of yellow crystals, mp 101-103 °

( t ■! 4- — n ■> n o n nllP ^ • BCDCI3 c 0.7 f O T T n 1 „ . p . ! v, z. 1,0 O T T 1Mb bridgehead), 3*28 (m, 2H, >CH-C0-), 1.17 (m, 2H, tertiary cyclopropyl),

and O.33 (m, 2H, secondary cyclopropyl); 1820 and 1227 cm"1. max

anti-Tricyclo[3.2.2. Q2’4 ]non-8-ene-endo,endo-6,7-dicarboxylic Acid 36 (387)- A solution of 30.0 g (0.158 mol) of 386 and 17.0 g (0.16 mol)

of sodium carbonate in water

(300 ml) was heated at reflux for

30 min and allowed to cool.

Hydrochloric acid (100 ml, 25%)

jgy was added to crystallize 387 .

The white flaky crystals were washed with 200 ml of cold water and recrystallized from 800 ml of water to 205 294 give 25.0 g (70.0%) of 386, mp 168-171° (gas evolution) (lit. mp 17O-

1T4° ); ^ D3)2C0 5,83 2B’ olefinic), 3.13 (m, 2H, bridgehead),

2.00 (m, 2H, >CH-C02-), 1.01 (m, 2H, tertiary cyclopropyl), and 0.13

(m, 2H, secondary cyclopropyl); vKBr 2940 (broad), 2700, 2625, and max 1724 cm”1.

35,295 Tricyclo[3.2.2. 02j4]nonadiene ( .45 ). A. By Electrolysis.

A well-stirred solution of 1.04 g

(5 .0 mmol) of 387 and 1.25 ml of

triethylamine in 100 ml of 10%

e aqueous pyridine was electrolyzed

between platinum wire and cylin­

drical gauze electrodes at 100 V

(dc) for 3 hrs during which time the current dropped from 0.5 to 0 .1 amps. The temperature was maintained at 5-20° and a dark brown color developed as gas was evolved. Hie solution was poured into water

(100 ml). This solution was combined with that of a second similar run and extracted with pentane (5 x 25 ml). The combined extracts were washed with 50 ml of 5% hydrochloric acid and dried. The pentane was removed by fractional distillation and the residue was vacuum distilled 36 to give 0.24 g (20%) of yellow liquid, bp 80° (40 mm) (lit. bp 83 -86° at 40 mm); 8.50 (m, 2H, anti olefinic), 5*91 (m, 2H syn olefinic),

3 .58 (m, 2H, bridgehead), 1.08 (m, 2H, tertiary cyclopropyl), and 0.57

(m, 2H, secondary cyclopropyl); 3030 , 2940, 1353? 1038, 900, 813 , IuctX and 699 cm-1. 206 46,296 B. By Lead. Tetraacetate Decarboxylation. After oxygen was

bubbled through 500 ml of pyridine for 15 min, 106.4 g (0.240 mol) of

dry lead tetraacetate and 25-0 g (0.120 mol) of 287 were introduced and

the suspension was heated to 70°. The material dissolved, the solution

turned dark brown, and carbon dioxide was evolved rapidly. The oil bath was immediately removed and after 10 min,gas evolution subsided. The

product mixture was cooled, poured into 1 liter of 15$ nitric acid,

and filtered. The filtrate was extracted with ether (4 x 200 ml) and

pentane (5 x 200 ml). The combined extracts were washed with water

and dried. Fractional distillation gave 0.40 g (2.8$) of 45, bp 47°

(20 mm).

41 . . . 29T Maieie Acid-2,5-da (48). The sulfur-Quinoline poison was prepared by

heating 1.0 g of sulfur in 6.0 g

of pure synthetic quinoline at

D x D the reflux temperature for 5 br, 0=0 H00C C00H dilution with 50 ml of xylene, and

filtration to remove insolubles.

A mixture of 4.00 g (28.2 mmol) of

dimethyl acetylenedicarboxylate, 100 ml of absolute methanol, and 0.40

g of 5$> palladium on barium sulfate was treated with 8 drops of the

poison and subjected to an atmosphere of deuterium gas at atmospheric

pressure. After 4.5 hr of vigorous magnetic stirring, the mixture was

rid of catalyst by filtration and the filtrate was evaporated under

reduced pressure. The resulting yellow oil was dissolved in diethyl 207 ether and remaining coagulated material was separated by filtration.

The organic phase was extracted with 5% hydrochloric acid solution until yellow color was no longer removed and dried. Vacuum distillation gave 2.70 g of colorless liquid, bp 85-95° (2.5 Him)? the 1H nmr spectrum of which showed the material to consist of maleate (65%) and succinate (35%)• The ratio was confirmed by vpc methods and comparison with undeuterated authentic samples.

A 2.38 g sample of this mixture was heated at reflux with 40 ml of 10^ aqueous potassium hydroxide for 4 hr. The pH of the cooled solution was adjusted to 5 with concentrated hydrochloric acid. Con­ tinuous ether extraction (14 hr) furnished 30 mg of succinic acid, mp 182-186°. The pH was now adjusted to 3 and continuous ether extrac­ tion was resumed (2 days). From the dried organic extracts, there was obtained 800 mg of pale yellow solid, mp 120-180°. Adjustment to pH 1 and final extraction afforded 1.0 g (805& based upon estimate of 48 originally present) of maleic acid-2,3-d2, mp 132-134.5°. Recrystalli­ zation from ether with charcoal decolorization gave pure product, mp

137-1380; * 2 * 3^00 (broad), 2890, 1700, 1630 , 1565, 1447, 1275, 917 TTlgtX (broad), and 868 cm-1.

35,295 Tricyclo[3.2.2.02 4]nona-6,8-diene-6,7-d2 (50). A mixture of

maleic acid-2,3-d2 (640 mg, 5.42

mmol), 10.0 g (0.108 mol) of dis­

tilled cycloheptatriene, and 10 ml

D of xylene was carefully heated to D reflux (foamingi) in a 100 ml 50 208 round-bottomed flask fitted with a Dean-Stark trap. After 1 day at

IPO0, the solvent was evaporated under reduced pressure. The residual anhydride 49_ and 2 .0 g of sodium carbonate in 50 ml of water was heated to reflux for 1 hr, cooled, adjusted to pH 6, and evaporated in vacuo.

Of the 4.57 g of residue obtained in this manner, 4.2 g was dissolved in a small amount of hydrochloric acid (pH l) and the volume reduced under pressure until crystals began to form. The entire mixture was then extracted with ether (2 x 200 ml) to give 5^0 mg (49%) of crude diacid, mp 159-187° (with gas evolution).

This diacid (560 mg, 2.66 mmol) was dissolved in 100 ml of 10% aqueous pyridine containing 1.25 nil of triethylamine and the solution was electrolyzed at 100 V and 7-19° for 5 hr. The resulting brown solution was poured into 200 ml of water and extracted with pentane

(5 x 100 ml). The combined pentane extracts were washed with 51° hydro­ chloric acid (4 x 50 ml) and saturated sodium carbonate solution (100 ml), dried, and reduced in volume to 2 ml by careful fractional distill­ ation. The hydrocarbon product was isolated by preparative vpc on a

6 ft x 0.25 in. 1 0 FFAP column (105°). There was obtained 30 mg (9-4fo) of 50) 2932, 2305 (very weak), 2268 (weak), 1430 , 1340, 1240,

1090, IO38 , 892, and 695 (strong) cm"1; 6 ^ 4 6.50-6.6 7 (m, 0.10 H, residual H6,HT), 5-82-6.12 (m, 2H, Hq,H9), 3-45-3-75 (m, 2H, Hx,Hs),

0.98-1.32 (m, 2H, H2,H4), and 0.28-0.78 (m, 2, H3 ); m/e 121 (3-8), 120

(41.8), 119 (100.0), 118 (34.3), and 117 (19-7)- 209

Addition of Chlorosulfonyl Isocyanate to 45_with Stirring for 5 Hr.

A solution of 450 mg (3-81 (mmol) of 45_ in 10 ml of dry dichloromethane was stirred at 20° under nitrogen while 5.40 mg (3*81 mmol) of chlorosulfonyl isocyanate in 1 ml of the same solvent was introduced via syringe. After 5 hr, 50 ml of anhydrous ether was added and the T4 solution maintained at 5° while 10 ml of 25% aqueous sodium sulfite and 1 ml of 10% potassium hydroxide solution were slowly added. After

1 hr at room temperature, the organic phase was separated and combined with an ether extract of the residual aqueous solution. After being washed successively with water (50 ml), 2% hydrochloric acid (50 ml), saturated sodium bicarbonate solution (50 ml), brine (50 ml), and water, the organic layer was dried and evaporated to leave 160 mg of pale yellow oil. Chromatographic separation was achieved on silica gel

(Baker) using incrementally higher percentages of ethyl acetate in hexane as eluent.

The first substance to elute proved to be lactone 51_ (14 mg, 2.3%) which was obtained as white crystals, mp 88.5-89.5°, after sublimation

(50°, 0.02 mm) and recrystalliza­

tion from ether-pentane; ^ * max 3035, 3010 , 2945, 1775 (strong),

1360 , 1317 , 1149, 1100, 1080,

1004, 978, 947, 9H, 888, 830 , and

819 cm"1; 6 ^ 13 4.55-4.75 (m, 1H,

>CH-0-), 2.85-2.98 (m, 1H, XJH-C0-), 2.52-2.79 (m, 1H, bridgehead), 1.60-

2.20 (m, 3H, tertiary cyclopropyls), 0.77-1. 52 (m, 2H, tertiary cyclo- 210 propyls), 0.46-0.72 (m, 1H, anti secondary cyclopropyl), and 0.12 to

-0.17 (m, 1H, syn secondary cyclopropyl).

Anal. Calcd for CioHio02: C, 7^*06; H, 6.22.

Found: C, 75*88; H, 6.23 .

The second substance isolated was P-lactam 52_ (20 mg,3 .2$>), white crystals, mp 150-1510, after sublimation (80°,0.02 mm) and

recrystallization from ether;

^ 3^15 , 3250 (broad), 3000 ,

2955, 2850 (weak), 1747 (strong), 0 1352, 1334 , 11T6, 1093, and 849

cm-i; 6 ^ 13 6.33 (br s, 1H, >WH), 52 5.77-6.05 (m, 2H, olefinic),

3 .44 -3 .6 8 (m, 1H, >CH-N<), 2.85-3.23 (m, 3H, >CH-C0- and bridgeheads),

0.77-1.42 (m, 2H, tertiary cyclopropyls), and 0.02-0.33 (m, 2H, secondary cyclopropyls); m/e (calcd) l6l.0841, (obs) 161.0843.

Anal. Calcd for CioHn.N0: C, 74.51; H, 6.88.

Found: C, 74.44; H, 6.89.

The third component was identified as P-lactam 53_ (51 nig, 8.3 %), white crystals, mp 111.5-112.5°? from ether-pentane; VCHC^3 3360 , 3205 ID3X (broad), 2915, 1754 (strong), 1315,

1172, 1043, 1012, 965, and 850 cm-1;

6TMSl3 6,45 (br S’ 5*62 (m, 2H, olefinic), 3*40-3.65 (m, 1H,

>CH-N<), 3.12-3.33 (m, 1H, XJH-C0-),

53 2.73-3.10 (m, 2H, bridgehead), 211

0.60-1.00 (m, 2H, tertiary cyclopropyl), and 0.05-0.33 (m, 2H, secondary cyclopropyl).

Anal. Calcd for C10HxxNO: C, 74.51; H, 6.88; N, 8.69.

Found: C, 74.70; H, 6.85; N, 8.53-

The last product eluted was characterized as v-lactam £4_ (16 mg,

2.6 /0 ), colorless crystals, mp 134-135°j from ether; vCHCls 3390, 3185 msx (broad), 2950, 1700 (strong), 1380 ,

XI 1305 , 1105, 1082, 1022, 1008, 10 990, 830 , and 820 cm"1; 6^ ls

7.00 (br s, 1H, >MH), 3*28 (br s,

an, XJH-NC), 2.40 (br s, 1H, X3H-C0-),

2.13 (m, 1H, bridgehead), 1.45-1.83

(m, 1H, tertiary cyclopropyl), 1.18-1.43 (m, 2E, tertiary cyclopropyl),

0.42-1.18 (m, 2H, tertiary cyclopropyl), 0.02-0.38 (m, 1H, secondary cyclopropyl), and -0.08 to -O.3 7 (m, 1H, secondary cyclopropyl).

Anal. Calcd for CxOHxxN0: C, 74.51; H, 6.88; N, 8.69.

Found: C, 74.53; H, 6.85; N, 8.69.

2-Gxo-7,8-benzobicyclo[2. 2.21oct-7-ene-endo-5,6-dicarboxylic Anhydride

46 (388 ). To 529 g (3*66 mol) of molten 2-naphthol was added 360 g

(3 .6 6 mol) of molten maleic anhy­

dride. The solution was heated to

240° for 1 hr, cooled to 150°, and

diluted with the careful addition

388 of 2 liters of ethyl acetate. The 212 solution was transferred at 60° and allowed to stand. Further crystalli­ zation occurred with the addition of ether (500 ml). Filtration gave

**•53 g (51$) of crude 5 8 8, which was dissolved in dichloromethane, treated with charcoal, and recrystallized to give 361 g of pale yellow 298 _ 388 , mp 192-196° (lit. mp 194-195°); vKBr 1866, 1776, 1730, 1224, max 1073, 947, 913 (broad), 773, 749, and 719 cm-1.

2-0xo-7,8-benzobicyclo[2.2.2]oct-7-ene-endo,-endo-5,6-dicarboxylic Acid

(389). A mixture of 75.0 g (0 .3 1 mol) of 388 and 1 liter of water

0 was heated at reflux for 2 hrs and

rotary evaporated to give 79*0 g

(98$) of white crystalline 589,

mp 164-186° (gas); 5135,

389 1748, 1718, 1300 , 1193, 1170,

1136 , 1107, 1013, 997, 869, 792,

751, and 724 cm"1.

96 2-Qxo-5,6-benzobicyclo[2.2.2]octa-5,7-diene (108). A solution of

50.0 g (0.192 mol) of 389 and

133*0 g (0.30 mol) of lead tetra­

acetate in 700 ml of pyridine was

stirred at room temperature. The

108 orange-brown solution turned

bright yellow and the temperature rose to 50°. Gas was evolved and the solution turned dark brown. 213

When gas evolution ceased the solution was poured into a mixture of

150 g of concentrated nitric acid in 2 liters of ice. This solution was extracted with ether (5 x 500 ml). The extracts were combined, washed with 5% hydrochloric acid and saturated sodium bicarbonate, and

dried. Evaporation of the solvent left 6 .6 g of brown oil which was

chromatographed on silica gel. Elution with ether (0-10Jo) in

petroleum ether (bp 30 -60°) gave 4.1 g (l2fo) of yellow oil, which 46 crystallized on seeding to give 108, mp 53 -56° (lit. aip 56.5-58.O0);

yHCCls 2874 , 1736 (strong), I608, 1466, 1406, 1332 , 1299, 1227 (broad), ID33C ll6l, 1145, 1122, 1080, 1022, 1000, 962, 942, 892, and 684 cm"1;

7.05-7*47 (m, 4h, aromatic), 6.47 -6.95 (m, 2H, olefinic), 4.07-4.57

(m, 2H, bridgehead), and 1.70-2.45 (m, 2H, methylenes).

69 2-0xo-5,6-benzobicyclo[2.2.2]octa-5,7-diene Tosylhydrazone (109).

A mixture of 5.0 g (29.4 mmol)

of 108, 5.5 g (29.5 mmol) of £- 2 7 9 toluenesulfonyl hydrazine and

• 5 drops of concentrated hydro—

chloric acid in 50 ml of THF was 109 refluxed for 4 hrs. Evaporation

of the solvent gave a yellow oil which was shaken in 100 ml of

dichloromethane to give 11.71 g (9^%) of a white 1:1 complex of 109

and dichloromethane, mp 118-122° (gas evolution). Recrystallizations

from ether gave pure 109, mp 162-164° (dec. 156°); vHCC‘1'3 3021 , 2950,

1626, 1577, 1458, 1326 , 1140 (strong), 1106, 1019, 976, and 901 cm”1; 6^?i(l3 7.72-7*97 (m, 2H, aromatic ortho to -SO-?-), 7-62 (br s, 1H, iiVlD ” -NH-), 7.02-7.42 (m, 6h, aromatic), 6.50-6.72 (m, 2H, olefinic),

4.45-4.65 (m, 1H, Cx bridgehead), 4.00-4.25 (m, 1H, C4 bridgehead),

2.38 (s, 3H, methyl), and I.9O-2.36 (m, 2H, methylenes).

Anal. Calcd for CX9HX8N20S: C, 67.43; H, 5-56; N, 8.28.

Found: C, 67.31; H, 5*41; N, 8.31 .

69 2,3-Benzobicyclo[2.2.2]octatriene ( 55.)» To a solution of 10.0 g

(23.6 mmol) of the dichloromethane-

IO9 complex and 100 ml of dry THF

under nitrogen was added 55 ml

(88 mmol) of 51> methyllithium

in ether over 30 min. The solu­

tion turned yellow and finally brown during the addition, while the temperature rose to 40°. After

10 hrs the solution was cooled (0°) and 60 ml of water was added.

This mixture was extracted with ether (4 x 200 ml). The combined ex­

tracts were washed with 5% hydrochloric acid and saturated sodium bicarbonate solutions and dried. Solvent removal gave 3*9° g of brown

011 which was chromatographed on silica gel (hexane elution) to give 45 2.00 g (55-0$>) of crystalline 55j> mp 56-61° (lit. mp 65-65.5°);

v^ 13 5185, 3003, 1653, 1603 , 1466, 1456, 1389, 1335 (strong), 1164 IHcLX 1092, 1018 , 930, 917 (strong), 835, and 8ll cm-1; 6 ^ 13 6.70-7.28 (m,

8h, aromatic and olefinic) and 4.65-5*07 (m, 2H, bridgehead). 215 49,300 Cyclopropanation of Benzobarrelene (5j5)* A solution of benzobarrelene

(J55j 3.47 g, 22.5 mmol) in 50 ml of anhydrous ether containing 7 .0 ml

(70 mmol) of diethylzinc was stirred mechanically under a dry nitrogen

atmosphere at room temperature while 6.20 g (25.0 mmol) of diiodo- methane in 10 ml of dry benzene was introduced dropwise. Upon comple­

tion of the addition, the mixture was heated at reflux for 10 hr.

An additional 4.0 g (15.0 mmol) of diidomethane in 5 ml of benzene was again added and heating was continued for 24 hr. Hydrochloric acid

(5$, 50 ml) was slowly introduced with ice cooling and the hydrolyzed

reaction mixture was poured into 300 ml of ether. The organic phase was washed with 5$ hydrochloric acid and these washings were extracted with ether. The combined organic layers were shaken with saturated

sodium carbonate (300 ml) and sodium chloride solutions (300 ml),

dried, and evaporated to leave 4.15 g of yellow oil.

This oil was chromatographed on silica gel-silver nitrate (10$);

elution was performed with increasing amounts of ether in hexane. The

first component to elute was iden­

tified as syn,syn-9,10-benzoquadri-

cyclo[3 ,3 » 2.02>4 .06>8]dec-9-ene

( 57, 130 mg, 3.2$). Sublimation

(70°, 50 mm) afforded colorless

needles, mp 129-130 °; vmax?Cl4 3020 , 3005 , 2938, 1585, 1464, 1436, 1379, 1506, 1182, 1117, 1087, and 1039

cm-1; 8.75-7.25 (AA'BB', 4h, aromatic), 3*47 (5-line m, 2H, 216 bridgehead), 1.08-1.53 (m, ^H, tertiary cyclopropyl), 0.13 (d of t,

^gem = Hz’ ^cis = ^Z’ anti secondary cyclopropyl), and

-0.77 (d of t, J^-rans = 3*8 Hz, 2H, syn secondary cyclopropyl); m/e

(calcd) 182.1095, (obs) 182.1098.

Anal. Calcd for C14 H14: C, 92.26; H, 7-74.

Found: C, 92.02; H, 7*81.

The next hydrocarbon proved to be bis-adduct 56, rhombic crystals

(from hexane) which were isolated in 6.1$ yield (250 mg). Sublimation

(70°, 50 mm) gave material of mp

91-92°; v^ 4 3075 , 3010 , 2930, max 1478, 1459, 1445, 1318 (doublet),

1170, 1089, 1034, 1020. and 905

cm”1; 7.04 (narrow m, 4h , tjg TMS aromatic), 3-37 (5-line m, 2H, bridgehead), 1.79 (6-line m, 1H, secondary cyclopropyl), 0.53-1-27 (m,

5H, one secondary and four tertiary cyclopropyls), -0.27 (d of t, gam = 5-5 Hz, J . = 7-2 Hz, 1H, secondary cyclopropyl), and -1.05 (d of t, OX s Jtrans = 5-8 Hz, 1H, secondary cyclopropyl); m/e (calcd) 182.1095,

(obs) 182.1098.

Anal. Calcd for C14 H i4: C, 92.26; H, 7-74.

Found: C, 92.23; H, 7-76.

The third component was a colorless oil (1.12 g, 29- 6f>) identified CDC1 as anti-6 ,7-benzotricyclo[3 -2. 2.02’4 ]nona-6 ,8-diene (47 ); v 3 2976, ■■ "■ _ *wnij max 2915, 1460 , 1430 , 1545 , 1263 , 1233, 1155, 1088, 1040, 1020, 1013 , 840, and 800 cm"1; 6 ^ 13 6.87-7-38 (m,

4H, aromatic), 5-97-6.30 (m, 2

olefinic), 3-80-4.12 (m, 211, bridge­

head), 1.03-1-40 (m, 2H, tertiary

cyclopropyl), and O.38 -O. 95 (m, 2H,

^7 . secondary cyclopropyl).

Anal. Calcd for C13 H12: C, 92.81; H, 7.19*

Found: C, 92.45; H, 7-20.

The last product was the corresponding syn isomer ( 46 ; 270 mg,

7.1$) which likewise was a colorless oil; 29^5, 2925, 1458, 1325 , ITlaX 1265, 1235 , 1210, 1165, 1108, 1090,

1037 , 1022, 845, 821, and 783

cm”1; 6.92-7.08 (narrow m,

4h,aromatic), 6.68-6.90 (m, 2H,

olefinic), 3-82-4.20 (m, 2H, bridge­ 46 head), 1.23-1.60 (m, 2H, tertiary cyclopropyl), 0.I8-O.58 (m, J . = 7 Hz, 1H, secondary cyclopropyl), and CIS -0.26 to -0.53 (d of t, Jgem = 5-8 Hz, J|-rans =3-8 Hz, XH, secondary cyclopropyl).

Anal. Calcd for Ci3 Hi2: C, 92.8l; H, 7.19*

Found: C, 93-06; H, 7.21.

Finally, 1.45 g (4l. 8$>) of benzobarrelene was recovered unchanged. 218

Addition of Chlorosulfonyl Isocyanate to 46. To a solution of 46

(300 mg, I.78 mmol) in 6 .0 ml of dry dichloromethane was added with stirring under nitrogen 257 mg (1.82 mmol) of chlorosulfonyl isocyanate via syringe. The mixture was maintained at 25° for 72 hr before dilu- 74 tion with ether (20 ml) and introduction at 0° of 25% sodium sulfite solution (5 ml) and 10% potassium hydroxide (0.5 ml) followed by 10 ml of water. After 50 min at room temperature, the contents were poured into ether (100 ml) and the organic phase was washed with 5% hydrochloric acid (25 ml), saturated sodium carbonate (50 ml), and saturated brine solutions (50 ml). The organic layer was dried and evaporated to leave a yellow oil (409 ing) which was chromatographed on silica gel (elution with increasing amounts of ether in hexane).

Initially, there was recovered 60 mg (2CP/o) of 46 . This was followed by P-lactam 58 (60 mg, 16.0°/>) which was obtained as small

white needles, mp 238 -239° (from

ether-hexane); 34l8, 3260 9 max ’ (broad), 3075 (weak), 3002 , 2950, rNH 1748 (strong), 1475, 1464, 1357,

1338, 1314, 1287, 1270, 1171, 58 1039, and 994 cm"1; fl® ? 13 6.87- IflqjC 7.30 (m, 4H, aromatic), 6.37 (br s, 1H, >N-H), 3*57-5.90 (m, 3H, >CH-M< and bridgeheads), 2.98-3.28 (m, 1H, >CH-C0-), I.67-I.83 (m, 2H, tertiary cyclopropyl), -0 .1 2 to 0.28 (m, 1H, secondary cyclopropyl), and -0.68 to -0.93 (m, 1H, secondary cyclopropyl). 219 Anal. Calcd for C14 Hi3 WO: C, 79-59; H, 6.20; N, 6.63 .

Found: C, 79-22; H, 6.37; N, 6.71.

The second product to elute was found to be p-lactam 59 (15 mg, k.Cffo), off-white crystals, mp 164. 5-166° (from ether); v^ cla 3412, max 3270 (broad), 3075 (weak), 3003 ,

2950, 1752 (strong), 1484, 1455,

1379, 1354, 1308 (doublet), 1283 ,

0 1220 (broad), 1171, 1041, 1012, 982, and 950 cm”1; 5^ 13 6.80-7.38

(m, 4, aromatic), 5-43 (br s, 1H,

>NH), 3-77-4.06 (m, 1H, >CH-N<), 3.38-3.72 (m,3H, >CH-C0- and bridge­ heads), 1.00-1.37 (m, 2H, tertiary cyclopropyl), 0.07-0.47 (m, IH, secondary cyclopropyl), and -0.45 to -0.70 (m, 1H, secondary cyclo­ propyl); m/e (calcd) 211.0997, (obs) 211.1000.

Lastly, there was isolated 5 mg (1-3$) of _6(^ as white needle CKCl clusters, mp 170.5-171' (from ether); max 3 3^18? 3035? 2995? 2965, 2928, 1700 (strong), i486, 1453,

NH 1427 , 1384 , 1347 , 1294, 986, 893,

and 670 cm"1; 6.80-7-32

(m, 4, aromatic), 6.22 (br s, IH,

>MH), 4.15-4.33 (m, 2H, >CH-C0- 60 and >CK-N<), 3-55-3-93 (m, 2K, bridgehead), 0.8-0.9 (m, IH, tertiary cyclopropyl), 0.4-0. 7 (m, 1H, tertiary cyclopropyl), -0.1 to 0.1 (m, IH, secondary cyclopropyl), and 220

-0.5 to -0.7 (m, IH, secondary cyclopropyl); ra/e_ (calcd) 211.09973

(obs) 211.1000.

301 Coumalic Acid (390). To a mixture of 300 g (2.24 mol) of dl-

malic acid and 260 ml of concen­

trated sulfuric acid were added H°2C ' ® three 75-nil portions of 65% fuming

sulfuric acid at 45-min intervals.

The mixture was then heated on a 390 steam bath for 2 hr, allowed to cool and diluted with ice to a volume of 2 liters. The resulting yellow precipitate was filtered, washed with water and dried in the open 301 to give 3l. 5 g (525S) of yellow, powered solid, mp 205-209° (lit. mp 206-209°).

29 of-Pyrone (391). a mixture of 50.0 g (0.214 mol) of 390 and 15 g

of sand was thoroughly ground in

a 250 ml round-bottomed flask, r ^ o attached to a heated right-angle

adaptor, and heated to 180° at

391 0.3 mm. The adaptor lead to a

horizontal 1 x 20 in. Vycor pyrolysis tube packed with coiled copper screen that was heated to 650°.

At the other end of this tube was a right-angle vacuum adaptor and a

Dry Ice-isopropyl alcohol cooled receiver. The pressure rose to 0.8 mm while the coumalic acid was subliming into the pyrolysis tube and 221 then dropped back to the original pressure. After 2b hrs the crude product, 15.1 g (73 $) was removed from the apparatus and distilled to give 391 as a pale yellow liquid, bp 70° (l. 5 mm), which solidified in 20 the freezer (lit. bp 83 ° at 2 .6 mm).

29 3,7-endo,endo-Dicarbomethoxybicyclo[2.2. 2]oct-2-ene (39g). A solu­

tion of 23.56 g (0.2^5 mol) of

391, ^5-3 g (0.539 mol) of methyl

acrylate and 0 .5 g of hydroquinone

was magnetically stirred at the CH3 02C reflux temperature under nitrogen

392 for 103 hrs. A gradual increase in temperature from 91 to 188° was noted. The pale yellow liquid was then distilled at 0.04 to 0.05 mm to give the following fractions:

The fraction preceding this was seeded and all crystals were filtered, washed with pentane, and recrystallized from pentane to give 13.0 g 29 (23.5ft) of 392, mp 90-92° (lit. mp 93-5-9^.0°).

Epimerization of Dicarbomethoxybicyclo[2. 2. 2]oct-2-ene Isomers.

Absolute methanol (l. 5 A) was stirred while 5»0 g (0.217 mol) of sodium shot was added. To this solution was added 300 g (1-33 mol) of various non-crystalline, distilled oils from the above fractional distillation. After stirring for 5 hrs at 25°, 100 g of ice, 200 ml of 222 saturated ammonium chloride solution, and 150 ml of 5$ hydrochloric acid were added (pH 6 ). The resulting solution was diluted with 1 liter of saturated brine and extracted with ether (3 x 500 ml). The combined ether layers were washed with saturated sodium bicarbonate solution and dried. The solvent was evaporated to leave 260 g of dark orange oil. Distillation at 0.05 mm gave the following fractions: bp

45 -80°, 1 .7 g; bp 80-91°, 4.9 g (1.6$); bp 91-92°, 51.9 g (3-9.7$); bp

92-9*)-°, 92.4 g (38 .0$); bp 94-95.5°, 52.8 g (17.6$); and bp 95.5-100°,

46,8 g (15.6$). Only the last fraction crystallized spontaneously in the refrigerator, but some crystals formed upon seeding the fraction imme­ diately preceding it. Further epimerizations were unsuccessful.

29 Bxc.vcIolS. 2. 2]oct-2-ene-5 ,7-sndo,endo-dlcarDohydrazide (393). A sut>-

, 0 ■ pension of 3 0.0 g (0.134 mol) of nh2-nh-c* 392 in 200 ml of absolute ethanol

was magnetically stirred while 30

ml (0.62 mol) of 99$ hydrazine

MH2-NH 0 hydrate was added. Upon heating 393 „ -— to 50 the material dissolved.

The solution was refluxed for 10 hours, the solvent evaporated, and the pale yellow solid was recrystallized from ethanol-water to give 25.05 g 28 (86.8$) of 393, mp 224-227° (lit. mp 229.5-230.5°).

29 5,7-endo,endo-Diaminobicyclo[2. 2. 2]oct-2-ene Dihydrochloride (394)*

A solution of 25.0 g (0.111 mol) of 393 in 600 ml of water was cooled to 0°. To this was added 700 ml of ether and 20 ml of concentrated 223

hydrochloric acid. A solution of ,KH3C1 15.4 g (0.223 mol) of sodium

nitrite in 100 ml of water and 50

g of ice was added quickly to the EH3 CI above mixture with vigorous stir- 394 ring while the temperature rose to 2°. After 5 minutes the layers were separated and the aqueous layer was extracted with 3°0 ml of ether. The combined ether layers were washed with 300 ml of saturated sodium bicarbonate solution and dried.

The dried solution was fractionally distilled over 50 g of acti- vated molecular sieves (4 &). When most of the ether had distilled,

600 ml of benzene was added. When the head temperature rose to 75°? the mixture was refluxed for 20 hours under nitrogen.

The mixture was cooled to 0° and filtered into 600 ml of hydro­ chloric acid to which 600 ml of water was then added. Distillation of this solution (45°, 36 left 11.44 g (48.8%) of 394 as a yellow oil.

This material was dissolved in methanol, filtered, and crystallized from isopropanol-methanol to give white solid 394; mp > 400°.

5,T-endo,endo-Bis(trimethylammonium)bicyclo[2.2. 2]oct-2-ene Diiodide 2g (395)* A solution of 5 .0 g (0.024 mol) of 394 in 30 ml of water was

stirred while 4.6 g (0.115 mol)

of sodium hydroxide was added.

This solution was filtered and

diluted with 100 ml of methanol

395 degassed with nitrogen, and refluxed under nitrogen. To this solution was added dropwise over 2.5 hours,

93 g (0.66 mol) of methyl iodide. After 11 hours of reflux, 93 g more methyl iodide was added dropwise and the mixture was heated for a fur­

ther 3 hours. The solvent was evaporated and the remaining oil was

diluted with methanol and poured onto a 2 x 50 cm column of silica gel

packed with 50% benzene-methanol. The column was eluted with 1 liter

of benzene-methanol, followed by 1.5 liter of methanol. Methanol-

soluble residues from all fractions were combined to give ^ .1 g (36 %)

of crude 393.

29 Bicyclo[2. 2. 2]octatriene ( 4l). A solution of 2.75 g (68.8 mmol)

of sodium hydroxide in 50 ml of

/I water was slowly poured into a

rapidly stirred solution of 10.5

g (6l. 8 mmol) of silver nitrate in

100 ml of water. The resulting

brown, muddy silver oxide was

filtered, slurried in 50 ml of water, filtered, and washed again.

The silver oxide was added to a solution of 6.0 g (12.5 mmol) of

595 in 40 ml of water "under a sodium hydroxide , carbon dioxide

exclusion tube, causing the mixture to turn silver gray. The mixture

was gradually warmed to 50° and held there for 50 minutes before being

cooled and filtered. The colorless filtrate was distilled at 50° at

0.1 mm to remove most of the water. The residue was pyrolyzed at 80-

115° and 0.1 mm. The volatile products were collected in Dry Ice 225 traps, and dissolved in ether. The organic phase was extracted with 50 ml of saturated sodium chloride and dried. The resulting solution was carefully concentrated in preparation for vapor phase chromatography.

The crude product was chromatographed on a | in x 6 ft column of

10$ SE-30 on 60-80 mesh Chromosorb W at 70° with a flow rate of

60 ml/min. The retention times were 1.7 minutes for benzene and 9*0 minutes for 4l^. The yield of 4l which was obtained as a colorless liquid that solidified in the freezer was 0.6 7 g (51.2$)* 13C nmr:

140.55 (doublet of multiplets, Ji3 ^ „ = 173*5 Hz, 6 C, vinyl) and L-ri 48.29 (doublet of multiplets, Ji3 „ „ = 143.4 Hz, 2 C, bridgehead). C-H

Addition of CSI to 4l and Subsequent Hydrolysis. A 800 mg sample

(7*7 mmol) of 4l in 50 ml of dichloromethane was treated with 1 .15 g

(8 .0 mmol) of chlorosulfonyl isocyanate in 10 ml of dichlorome thane at 0° as above. After 1 hr at 20°, the reaction mixture was cooled to

0°, diluted with ether (25 ml), and treated with freshly prepared T4 saturated sodium sulfite solution (50 ml) and ether (25 ml). After a pH adjustment to 8, the emulsion was stirred for 1 hr at 0° and the layers separated. The aqueous layer was extracted with dichloro- methane (2 x 100 ml) and the combined organic layers were washed with water (100 ml) and saturated ammonium chloride solution (50 ml).

Drying and evaporation left a yellow solid which was subjected to chromatography on Florisil.

Elution with ether-pentane (4:1) afforded 37 nig (3*1$) of 2-oxa-

3“Oxoquadricyclo[4.4.0.04 ’lo.05 ’7]dec-8-ene (ip5) which was sublimed (60° and 50 nun) and recrystallized

from hexane. The fine white needles

had mp 61-62°; v ^ 3 2958 , 2932,

2858, 1778 (strong), 1353, 1317, 0 1504, 1158, 1136 , 993, 982, 956,

and 901 cm-1; 6 ^ 13 8.57 (m, J7j8and J7j8 =

6 Hz, Js,9 = 8 Hz, IH, Ha ), 5-70 (t of m, J9,10 = 7 Hz, IH, H9 ), 4.23

(t, J1?6 = Jijl0 =2.5 Hz, IH, Hx), 3.06 (d of m, IH, H10), 2.33 (m, lH, H7), 2.23 (m, IH, H10), 2.01 (t of m, Js}6 ^ ^6,7 = ^ Hz, 4H, H8), and 1.84 (t of m, J5j7 = 6 Hz, IH, H5); m/e (calcd) 148.0526, (found)

148.0528.

Anal. Calcd for C g H ^ : C, 72.96; H, 5.44.

Found: C, 73-19; H, 5.65.

Elution with increasing proportions of ethyl acetate in ether gave

60 mg (4 .9fo) of a mixture of 107b and 106 rich in the first lactam:

W l3 3k00, 5205 (broad)> 2985 ’ 8 2915, 1705 (strong), 1375, 1205,

1105, 1025, 992,and 822 cm"1;

c cl3 6 A i m > Hs)’ 5,89 (a IH, H9 ), 5.75 (br s, IH, >HH),

^ 5 . 3 .5 9 (m> >CH-H<), 2.63-2.90

(m, 2H, >CH-C0- and H10), and 1.6-2.5 (m, 3H, cyclopropyls).

Sublimation of the lactam mixture (70° and 30 mm) caused selective

destruction of 107b and furnished 8 mg (0.Tjo) of 3-aza-4-oxotricyclo- 227

[4.2.2.02,5]deca-7,9-

as off-white prisms, mp 152.5-

15^° (from ether-hexane); v^ 3

3^14, 3360 (broad), 2926, 2857,

106 1751 (strong), 1352, 1326, 1316,

1276, 1225, 1130 , 1071, 99^, 959, and 897 cm”1; 6^P£jl3 6.12-6.52 (m, *f-H, olefinic), 5*80 (br s, 1, >NH), 1Mb

3.88-^. 22 (m, 2H, bridgeheads), 3 .8 0 (t of d, Jx,2 = J2,5 = ^*5 Hz,

J2jX = 1 Hz, IH, >CH-N<), and 3-5^ (t of d, J5,6 = *•••5 Hz, J5jy =1.5

Hz, IH, X3H-C0-); m/e (calcd) l b j . 0 6 8 b , (found) l k j . 0 6 8 7 .

N-(Chlorosulfonyl)-2-aza-3-oxoquadricyclo[4.k. 0.04>10.05’73dec-8-ene

(107a). A magnetically stirred solution of 6l0 rag (5.86 rmnol) of

barrelene (^1) in 15 ml of dry

dichloromethane cooled to -78° 5 was treated dropwise with a solu­ 10 tion of chlorosulfonyl isocyanate C1-S0. (830 mg, 5.86 mmol) in 10 ml of

the same solvent. After 7 hr at

-78°, the temperature was allowed to return slowly to 20° overnight.

Upon cooling to -10°, the mixture was diluted with ether (10 ml) followed by 25 ml of saturated sodium sulfite and 2 ml of 10% potassium 74 hydroxide solutions (rapid stirring). The pH was adjusted to 8 with 5% hydrochloric acid and stirring was continued at room temperature for

10 min. The layers were separated, the aqueous phase was extracted with 228 ether (50 ml), and the combined organic solutions were washed with 50 ml portions of 51° HC1, saturated NaHC08, and brine solutions prior to drying and evaporation. The residual pale yellow oil (995 mg, 74%) PTTPT was now comprised chiefly of 107a; v 3 3005 , 2955, 1722 (strong), m_r».ji- li- lj f f i c l X

1437, 1375 (strong), 1258, ll80 (strong), 1108, 1075, 1005 (strong), and 915 cm’1; 6 ^ 13 6.54 (m, J7j8 = 6 Hz, J8,9 = 8 Hz, IH, Hs), 5.80

(t of d, J9jlo = 8 Hz, J7j9 = 2 Hz, IH, H9), 4.78 (m, J1?e = J1>10 = 2

Hz, IH, Hi), 3-22 (br d, IH, H10), 3-03 (br s, IH, H4), 2.53 (t of d,

J5,t = ^6,7 =5*5 Hz, IH, H7), and 1.85-2.38 (overlapping t, 2H, H5 and H6).

Addition of CSI to 55^ and Subsequent Hydrolysis. A magnetically stirred solution of 55_ (l. 00 g, 6.48 mmol) in 10 ml of dry dichloro- methane cooled to 0° under nitrogen was treated dropwise during 30 min with a solution of chlorosulfonyl isocyanate (920 mg, 6.50 mmol) in

10 ml of the same solvent. Reaction was allowed to proceed for 24 hr at room temperature before removal of solvent in vacuo. The residual yellow oil was dissolved in ether (15 ml) and added to a rapidly stirred mixture of ether (10 ml) and 25% aqueous sodium sulfite solution (15 7 4 ml) at 0°. During this addition, the pH was maintained at 7-8 with small amounts of 10% potassium hydroxide. This mixture was maintained at 0° for 30 min and at room temperature for an equal length of time.

The ether layer was separated and the aqueous phase extracted with ether (2 x 50 ml). The combined organic layers were dried and evaporated to give a pale yellow solid, the analysis of which showed it to be a fairly complex mixture. 229

A 1.57 g sample of such a mixture was carefully chromatographed on silica gel. Elution with hexane-ethyl acetate (4:l) afforded lactone

110 (130 mg), mp 146.5-147° (from

ether-hexane);’ vCHCls max 3033? 5 3017, 2960 (weak), 1780 (strong), 1491,

1^78, 1365 , 1315 (doublet), 1158,

1138 , 1114, 995, 979, 965, 944,

935, 925, 895, 868, and 852 cm"1-,

6^ ls 7.03-7.50 (m, 4h, aryl), 4.47 (m, 1, H x), 3.49 (m, IH, H10),

2.79 (d of d, J5jY =6.4 Hz, Je,7 =7.0 Hz, IH, Hr ), 2.43 (m, IH, H4), and 1.80-2.37 (m, 2H, H5 and He).

Anal. Calcd for C^HnC^: C, 78.77; H, 5*09.

Found: C, 78.53; H, 5-11.

Gradual increase in solvent polarity to 80

(from dichloromethane-ether);

vm S l3 3kl6, 3250 (broad)’ 307°, 2995, 2965, 1755 (strong), 1470,

1353, 1319, 1282, 1271, 1172, 1155,

1124, 1019, 992, 971, and 675

(broad) cm"1*, 6 ^ 13 6.91-7-33

(m, 4h, aryl), 6.18-6.72 (m, 2H, olefinic), 3.90-4.24 (m, 2H, benzylic),

3.48-3-71 (m, IH, >CH-IK), and 3.16-3-40 (m, IH, X3H-C0-).

Anal. Calcd for C13 HnN0: C, 79*17; H, 5-62*, N, 7.10.

Found: C, 78.86; H, 5-61; N, 7.13- 230

With ethyl acetate as eluent, there was obtained a mixture of the above P-lactam (360 mg) and Y-lactam 112 (297 mg, percentage composition

0 derived from *11 nmr integration). The latter component was isolated in

pure form by repeated recrystalli­

zation from dichloromethane-ether

as colorless crystals, mp 191-191-5°;

v- 13 5410 ’ 5250 (broad)> 3000,

1707 (strong), 1468, 1459, 1380 , 1323 , 1237, 991, 854, 640, 633 , 593, and 557 cm"1; 6^ ls 6.93-7-43 (m, 4h, aromatic), 6.70 (br s, IH, >NH),

5.98-6.33 (m, IH, olefinic), 5-50-5-81 (q of m, IH, olefinic), 3-92-

4.20 (m, IH, >CH-C0-), 3-66-3.92 (m, IH, >CH-N<), and 3-21-3.60 (m, 2H,

benzylic).

Anal. Calcd for C13 HnW0: C, 79-17; H, 5-62; N, 7-10.

Found: C, 78.88; H, 5-59; N, 7-20.

Two additional products were admixed in the fractions immediately

following (dichloromethane-ethyl acetate elution). Recrystallization

from dichloromethane-ether gave

pure hydroxy amide 114 as colorless

needles, mp 149-149-5°; v^ 3

3532, 3496, 3410, 3300 (broad), CONHp 3190 (broad), 2998, 2912, 1669

(strong), 1591, 1484, 1465, 1409,

1293 (broad), 1166, 1118, 1096, 1055, 998, 866, and 834 cm”1; 6 ^ 13 231

6.86-7.28 (m, 4h, aryl), 6.13 (br s, 2H, -NH2), 5-39 (br s, IH, -OH),

3 .5 6 (m, IH, >CH-0), 3.11 (m, IH, bridgehead), and 1.68-2.45 (m, >CH-CO- and cyclopropyl).

Anal. Calcd for C13 Hi3 N02: C, 72.54; H, 6.09; N, 6.51.

Found: C, 72.15; H, 6.12; N, 6.50.

Hie mother liquors from above were evaporated and the residue crystallized from ether. Physical separation of the solid rosettes

(due to lactam 113 ) from the needle

clusters of 114 and reerystalliza-

tion of the first material from

dichloromethane provided pure

113, mp 210-210.5°; vCHCl3 3442, ? * ’ max s 3220 (broad), 3016 , 2998, 1669

(strong), 1492, 1468, 1380 , 1355,

1312 , 1290, 1074, 992, 976, 910, 857, and 623 cm-1; 7-00-7.48

(m, 4h, aryl), 6.82 (br s, IH, >KH), 3.20-3.43 (m, 2H, >CH-N< and

bridgehead), 2.77 (apparent t, J =6 .7 Hz, IH, benzylic cyclopropyl),

2.18-2.33 (m, IH, XJH-C0-), and 1.78-2.35 (m, 2H, cyclopropyl).

Anal. Calcd for C13 Hi;iN0: C, 79* IT; H, 5-6l; N, 7-10.

Found: C, 78.80; H, 5-75; N, 7.05.

Hydrogenation of exo-3-Aza-7,8-benzo-4-oxotricyclo[4. 2.2.02’ 5]deca-

7,9-diene (111). A solution of 25 mg (0.127 mmol) of 111 in 10 ml of

ethyl acetate containing 10 mg of 10ft palladium on carbon was hydro­

genated for 12 hours at atmospheric pressure. This mixture was filtered 232

through Celite, rinsing with

methylene chloride and dried. The

solvents were evaporated to give

25 mg (99$) of 113. This material

113 was sublimed at 80° at 0 .0 2 mm

then recrystallized from ether to give white crystals of 115. mp 179-180°; v ^ ls 3420, 3225 (broad), 3005, r^*rrjr' nicw£ 2955, 2900, 1753 (strong), 1469, l46l, 1359, 1342, 1314, 1265, 1172,

1100, 1066, and 970 cm-1; 7.03 -7.31 (apparent s, 4h, aromatic),

6.49 (br s, IH, >NH), 3.53-3°76 (m, IH, -CH-Nc), 2.98-3-42 (m, 3H,

>CH-C0- and two bridgeheads), 2.02-2.38 (m, 2H, methylenes syn to the benzene ring), and 1.13-1.53 (m, 2H, methylenes syn to the lactam ring).

Anal. Calcd for C13 H 13 NO: C, 78.36 ; H, 6.58; N, 7.03 .

Found: C, 77-99; H, 6.51; N, 7-01.

55 trans-11,12-Dichloro-9,10-dihydro-9,10-ethanoanthracene (396). A

mixture of 10.0 g (5 6 .1 mmol) of

anthracene and 50-0 g (0.50 mol)

of trans-1,2-dichloroethylene

was sealed in a glass tube. The

mixture was heated at 200-208°

for 30 hr, removed and dissolved in 200 ml of hexane and poured onto a column of 300 g of neutral alumina.

The column was eluted with 1 liter of hexane. When white solid began to appear, the solvent was changed to carbon tetrachloride (l. 5 liters). 233 The combined carbon tetrachloride fractions gave 9*33 g (60.4$>) of pale 55 green solid 396, mp 105-111° (lit. mp 113.5-114°).

55 Dibenzobarrelene (99_). A solution of 9«0 g (32.7 mmol) of 396 in

250 ml of n-amyl alcohol was

refluxed (120°) while 3*76 g

(0.164 mol) of sodium was added

slowly. After heating for 1 hour

99 an additional 3*76 g of sodium

was added. The temperature of the solution at that time was 130°. Again after 1 hour of reflux,

3 .7 6 g of sodium was added and the temperature of the reaction mixture rose to 140°. A final hour of reflux resulted iu a tempera lure of 150° being reached. The solution was then allowed to cool gradually and 200 ml of water was added.

This mixture was poured into 1 liter each of hexane and water.

The organic layer was washed with water (3x1 liter) and the combined water washes were extracted with hexane (2 x 400 ml). The organic phases were combined and washed with 500 ml each of 5% hydrochloric acid, saturated sodium bicarbonate solution and brine before drying.

The solvent was evaporated and the residual amyl alcohol was removed at 80° and 20 mm. This crude product was dissolved in 95% ethanol, treated with activated charcoal, and recrystallized from to give 4.65 g

(69.^) of 99 as pale yellow plates.

11 234

Chromatography on neutral alumina packed in hexane (elution with hexane and 2-107 ethyl acetate in hexane) gave 3 * TO g (55* 47) of 9 9.

Recrystallization from 957° ethanol gave 3*22 g of 99, mp 118-120.5° 55 (lit. mp 119.5-122°).

Addition of CSI to 99 _. A solution of 2.247 g (11.0 mmol) of 99 and I.5 6 g (ll.O mmol) of CSI in 25 ml of dry dichloromethane was heated at reflux under argon for 72 hr while magnetically stirred.

After cooling, the solvent was evaporated and the residue was chromato­ graphed on silica gel.

Elution with hexane-ethyl acetate (4:1) gave 80 mg (2.77) ^§3 mp 185.5-186.5° (from dichloromethane-ether); v 3 3072 -3012 , 2956, ITIciX 224b, 1485, 1471, l4bO, 1322 , 1267, NC Cl i244, 1176, 1153, 1139s 1108, 886, H 687, 643, 631 , 616, 580, and 571

cm-1; 6 ^ 13 6.87-7.46 (m, 8, aryl).

116 5.12 (d, J4j5 = 2.2 Hz, IH, H4),

4.22 (narrow m, IH, Hi), 4.01 (m,

IH, H5), and 3*82 (narrow m, IH, Hs)j m/e (calcd) 265.0658, (found)

265.0663 .

Anal. Calcd for C1THi2CM: C, 76.84; H, 4-55; N, 5-27.

Found: C, 76.6l; H, 4.75*, N, 5.27.

Enhancement in solvent polarity to 50$ ethyl acetate furnished

730 mg (19.17) of ll8a dense white prisms, mp 210-211° (from dichloro­ methane-ether); vCHC13 3070 , 3026 , 2960, 1772 (strong), 1483, 1470, max 235

1459, 1406 (strong), 1334, 1289, 0 1277, 1186 (strong), 1156 (strong), -S02C1 1144 (strong), 1120, 1101, 1091,

1053, 1047, 1027, 973, 896, 863 ,

ll 8a 608 (strong), 573 , and 562 cm-1;

6 tms1 3 6 ’8 2 ~ ^ k 2 (m > 8 h >

5.37 (dd, Jlj7 = 5, Jt, 10 =1.5 Hz, IH, >CH-N<), 4.27 (m, 2H, benzylic bridgeheads), and 3*78 (t of d, Jx,xo = J4 , xo = 5 Hz, IH, >CH-C0-).

Anal. Calcd for CX7HX2CIUO3 S; C, 59-03; H, 3-50; N, 4.05.

Found: C, 58.98; H, 3-83; N, 4.06.

Several early fractions in the latter series contained small amounts of 117a (v^^Cl2 1815 cm"1, estimated 2.2$ yield) which resisted max purification by fractional cry­

stallization or preparative tic.

'Ihe derived P-lactam could in

contrast be purified without diffi-

117a culty ( see below).

74 Dechlorosulfonylation of ll8a. A solution of 110 mg (0.318 mmol) of

118a in 5 ini of dichlorome thane and

10 ml of ether was stirred magne­ KH tically while 10 ml of 25$ sodium

sulfite solution and 10 drops of

118b 10$ K0H were added. After 4 hr,

dichloromethane was added to dissolve the solid which formed and stirring was continued for an additional hr.

After addition of ether (50 ml) and 5% HC1 (100 ml), the organic layer was separated, washed with saturated NaHC03 and NaCl solutions, dried, and evaporated. Lactam ll8b was obtained in quantitative yield, mp

256.5-25T.50 (from dichloromethane-ether); 3405, 3200 (broad), nicix 3000 , 2965, 1710 (strong), 1482, 1472, l46l, 1384, 1209, 1167, 1001,

638 , and 592 cm"1; 6^ ls 6.73-7-32 (m, 8h, aryl), 5-93 (br s, IH,

>KH), 4.33 (m, IB, H4 benzylic bridgehead), 4.1 8 (m, 2H, >CH-N< and benzylic bridgehead), and 3*54 (t of d, IH, >CH-C0-); m/e (calcd)

247.0997, (found) 247.1001.

T4 Dechlorosulfonylation of 117a. A 273 mg sample of a mixture of 117a

and llOa was treated with sodium ---- sulfite solution as above to give

200 mg (80$>) of a mixture of 117b

(3 parts) and ll8b (5 parts).

Repeated recrystallization from

dichloromethane-ether removed ll8b CHC1 and provided ultimately 26 mg of pure p-lactam, mp 246-247° dec; v 3 ItlcLX 3420, 3320 (broad), 2998, 2968, 1758 (strong), l46o, 1354, 1342, 1310 ,

1262, 1165, 1100 (broad), 1009, 810 (broad), and 556 cm"1; 6^ ls 6. 83 -

7.83 (m, 8h, aryl), 5-58 (br s, IH, >mh), 4.33 (t, J1>2 = J5,e =3-5 Hz,

2H, benzylic bridgeheads), 3*75 (m, IH, >CH-IJ <), and 3*42 (t of m, IH,

>CHC0-); m/e (calcd) 247.0997, (found) 247.1001. l-Cyanodibenzoseroibullvalene (119), 1,5-Elimination from 115. A

solution of 116 (40 mg, 0.15 mmol)

in dry tetrahydrofuran (l ml) and

anhydrous dimethylsulfoxide (l ml)

was stirred magnetically under

CN argon while 35 Mg (0-51 mmol) of freshly sublimed potassium tert- butoxide was introduced. The light brown reaction mixture was stirred for 10 min at room temperature and worked up as above to give 31 Mg

(90$) of 119 as white needles, mp 211-212.5° (from ether-hexane); vCHC13 3011(.5 2970, 2238 , 1472, 1462, 1174, 1018, 960, 818, 691, HldX and 628 cm-1; 6.89-7.38 (m, 8h, aryl), 4.72 (s, IH, H5), and

3 .6 7 (s, 2H, cyclopropyl); 13C nmr (CDC13) 149-0 (2C), 133*9 (2c),

127.7 (2C), 127*3 (2C), 125.2 (2C), 121.1 (2C), 57.1, 45.2 (20), and

44.4 ppm. (see Table XVI, page 101).

Anal. Calcd for C1THi2N: C, 88.67; H, 5.25.

Found: C, 88.93; H, 4.85.

re 1-Carboxamidodibenzosemibullvalene (120). A suspension of 500 mg

(2 .0 1 mmol) of dibenzosemibull-

valene-l-carboxylic acid in 5 nil

of benzene was stirred at room

temperature under nitrogen while

0 .2 ml of pyridine and 0.30 ml CONH2 (0.50 g, 4.2 mmol) of thionyl 120 chloride was added. The solution turned brown during the 2 hours it was maintained at reflux. After cooling, the solution was poured onto

200 ml ammonia-saturated ether (a light brown precipitate formed); 150 ml of water was then added. The layers were separated and the ether layer was washed with 100 ml of water and dried. Removal of the

solvent left 5^0 mg of light brown solid which was recrystallized from acetone-benzene to give 555 nig (71.5°/°) of off-white crystals of 120, 76 mp 206.5-207.5° (lit. mp 208-209°).

Dehydration of 120. A suspension of 150 mg (0.6l mmol) of 120 in 15 ml of sodium-dried benzene and 180 p,l of anhydrous triethylamine was heated with magnetic stirring until dissolution. The heating bath was removed and l80 mg (1.5 mmol) of thicnyl chloride was introduced.

After 5 hr at the reflux temperature, the solution was cooled, diluted with ether (50 ml) and dichloromethane (50 ml), and poured into ice water. The aqueous phase was separated and extracted with ether (50 ml). The combined organic layers were washed with 50 ml aliquots of

5$ hydrochloric acid, saturated sodium bicarbonate, and brine solutions

prior to drying and evaporation. The crude product (170 mg) was puri­

fied by preparative tic (silica gel; benzene elution) and recrystalliza­

tion from ether-hexane. There was isolated 55 mg (39/o) of 119 as white

needles, mp 211-212.5°, which proved identical to the material above.

exo-^-Chloro-anti-8-cyanobicyclo[5.2.11octa-2,6-diene (189)• A solu­

tion of 4l (30 mg, 0.29 mmol) in 2 ml of dry dichloromethane cooled to

-78° was treated with ^1 mg (0.29 mmol) of CSI while stirred magnetically. The temperature of the solution

was allowed to warm gradually to

room temperature and maintained

there for 1 hr. Three ml of

189 dimethyl formamide was introduced and after 3 hr the dichloromethane was removed by fractional distillation (pot temp 80°). The clear brown solution was heated at 75-95° for 40 hr and poured into 10 ml of cold water. The aqueous layer was diluted with 5 ml of saturated NaHC03 solution and extracted with ether (20 ml). The residue obtained after drying and evaporation of the combined organic layers was puri­ fied by preparative vpc on 10$ SE-30 at 130° (0,25 in. x 6 ft column).

There was collected 15 mg (31$) of 189 as a colorless oil which cry­ stallized upon sublimation at 50° and 25 mm, mp 70-71°; 8923, fllctX 2857, 2245 (strong), 1371 (strong), 1340, 1317 (doublet), 1198 (strong),

1184, 1106, 1066, 1015 (doublet), 973 , 964, 955, 932 (doublet), 879,

846 (strong), 809, and 6ll cm"1-, 6 ^ ^ 13 6.64 (q, JljT = 3 Hz, J6jT = 5*5

Hz, IH, Hy), 6.26 (m, Jx,2 = 6 Hz, J^,3 = 9-5 Hz, J2j 4 “ 2 Hz, IH, H2), 5

(q, Js,6 = 3 Hz, IH, Hs), 5.43 (m, J3}4 = 2 Hz, J3 ,s = 1 Hz, IH, H3),

4.4l (m, J4>5 = 3 Hz, IH, H4),3 .38 (s, J < 1 Hz, IH, Hs), 3 .3 2 (m,

IH, H5), and 3.17 (m, IH, Hx).

Anal. Calcd for C9HsC1N; C, 65.27; H, 4.87.

Found: C, 64.91; H, 4.99. 240

1(5)-Cyanosemibullvalene (190). A magnetically stirred solution of

freshly purified

189 (50 mg, 0.30 CN mmol) in 5 ini of 0 anhydrous tetra- hydrofuran cooled

190a 190 b to "5° under nitrogen was treated dropwise with a solution of freshly sublimed potassium tert-butoxide in 3 .5 ml of dry dimethyl sulfoxide. After completion of the addi­ tion, the solution was allowed to warm to 20° where it was stirred for 20 min before being poured into pentane (50 ml) and water (20 ml).

The layers were separated and the aqueous phase was reextracted with pentane (2 x 30 ml). The combined pentane fractions were washed with water (20 ml), dried, and concentrated by fractional distillation.

The concentrate was subjected to preparative vpc on a silanized 5$

SE-30 column (1/4 in. x 2 ft) through which triethylamine had been passed earlier (70°). The only component was isolated to give 22 mg

(55$) of 190: v ^ e Cl2 3065 , 2965, 2233 (strong), 1575, 1334 (strong), ,xr,iV max 1172, 1139, 1101, 1046, 963, 919, 862 (strong), 801 (strong), 637, and 592 cm"1; 6^ (CD2C12-CF2C12) 5-75 (d of m, J3j4 = Js,7 = 5-3 Hz,

J4,5 = Js,e =2.2 Hz, J2j4 = J6j8 =1.2 Hz, 2H, H4 and H6), 5.21 (d of

J2 ,3 = J7,a =1.4 Hz, J3 ,5 = Js,7 =0.4 Hz, 2H, H3 and H7), 3*51

(br t, 1H, Hs), and 3-45 (m, 2H, H2 and Hq); 6 ^ 5.65 (m, 2H, H4 and He),

5.17 (m, 2H, H3 and H7), 3.35-3.48 (m, 3H, H2, H5, and Ha); 13C nmr 241

(CDC13) 130.0, 119.2, 56.0, 49.9j and 43*3 PP“i; \c2HsOH -^nj-ense msix absorption only; m/e (calcd) 129.0578, (found) 129.0580.

Chlorocyanation of Benzobarrelene ( 55 ). Chlorosulfonyl isocyanate

(2.83 g, 20.0 mmol) dissolved in 10 ml of dry dichloromethane was added dropwise during 20 min to a magnetically stirred solution of 55^

(3 .00 g, 19.5 mmol) in 45 ml of the same solvent at room temperature voider argon. After 18 hr, 1 ml of dimethylformamide was added and the dichloromethane removed by distillation while an additional 30 ml of dimethylformamide was introduced. When the internal temperature reached 8o°, distillation was discontinued. The reaction mixture was heated at 80° for 50 hr, cooled, and poured into 300 ml of ice water.

Workup in the predescribed fashion afforded 3*21 g (74.4$) of a mix­ ture of chloro nitriles.

This oil was chromatographed on silica gel (150 g) using hexane elution containing 1-20$ ether. The first component isolated was

identified as 192 (1.00 g, 23 -8$),

thick needles, mp 147.5-148.5°

(from ether-hexane); v^ ^ ’3 2950,

2250 (strong), 1966, 1932, 1826,

192 1815, 1657, 1605, 1478 , 1453,

1346, 1320 (strong), 1267-1158

(broad), 1115, 970, 905, 86l, and 571 cm"1; 6 ^ 13 6.83 -7.50 (m, 4h, aryl), 6.54 (q, J6j7 =5*5 Hz, Jlj7 = 2.9 Hz, 1H, H7 ), 5 .87 (q, J5,e =

2.9 Hz, 1H, Hs), 4.99 (d, J4>5 = 2.9 Hz, 1H, H4), 3-71 (d, 1H, Hi), 242

3.54 (s, 1H, >CHCN), and 3-49 (t of m, 1H, Hs); m/e (calcd) 215.0502,

(found) 215.0504.

Anal. Calcd for CiaHxoClN: C, 72.39; H, 4.67; N, 6.49*

Found: C, 72.25; H, 4.75; N, 6.18.

The second component (O.988 g, 23.51°) proved to he 193. A pure

sample of this chloro nitrile was obtained by recrystallization from

ether-hexane and sublimation (90°

and 0 .0 2 mm) as a colorless cry­

stalline solid, mp 140.5-l4l.5°;

'Smax?’3 ^ 5 , 2250 (strong), 1966, 193 1934, 1605, 1477, 1454, 1352 ,

1337, 1309 (strong), 1270-1180

(broad), 1160, 1114, 986, 948, 900, 857, and 576 cm"1; 6 ^ 13 6.83 -7.65

(m, 4h , aryl), 6.48 (dd, Js,v = 5*6 Hz, Jx,t = 2.8 Hz, 1H, Hy), 5*99

(dd, J5,s = 2.5 Hz, IH, Hs), 5-24 (d, J4, 5 = 5-1 Hz, 1H, H4), 3-53-5.80

(m, 2H, Ha and K5), and 3*05 (s, 1H, Hq); m/e (calcd) 215.0502, (found)

215.0504.

Anal. Calcd for CxsHxoClH: C, 72.39; H, 4.67; N, 6.49.

Found: C, 72.43; H, 4.69; n, 6.25.

Lastly, 274 mg (6.%) of 191 was isolated, white needle clusters,

mp 109-111° (from ether-hexane); 2925, 2855, 2247 (strong), 1961, max 1920, 1806, 1626, 1608, 1591, l46l (strong), 1375 (strong), 1316

(strong), 1288, 1275-1170 (broad), 1152, 1114, IO96, 1071, 965 (strong),

891 (strong), 870, and 597 cm-1; 6^ 3 7.05-7.53 (m, 4h, aryl), 6.35 243

(q of t, J2>3 -9*5 Hz, Ji}2 ~

6.4 Hz, J2 j4 = 1.2 Hz, J2j8 =1.0

Hz, IH, H2), 5-58 (d of q, J3 j4 =

3*3 Hz, J3 j5 =1.7 Hz, 1H, H3),

4.47 (m, J4)5 = 1.2 Hz, IH, H4),

3.62-4.00 (m, 2H, Hx and H5), and 3.46 (t of d, Jx,s = Js,s = Hz, IH, Ha); m/e (calcd) 215.0502,

(found) 215.0504.

Anal. Calcd for Cx3 Hi0C1N: C, 72.39; H, 4.67; N, 6.49.

Found: C, 72.6l; H, 4.52; N, 6.67.

1-Cyanobenzosemibullvalene 194. A. Cyclization of 192. A solution

of 200 mg (I.78 mmol) of sublimed

NC potassium tert-butoxide in 2 ml

of dry dimethyl sulfoxide was

added in one portion to a magne­

tically stirred solution of 192

194 (200 mg, O.927 mmol) in 3 ml of

dry dimethyl sulfoxide and 3 ml of anhydrous tetrahydrofuran under argon at room temperature. After 10 min, the solution was poured into a mixture of ether (50 ml) and ice water (100 ml). The layers were separated and the aqueous phase was extracted with ether (50 ml) and dichloromethane (50 ml). The combined organic layers were washed with water (2 x 30 ml), dried, and evaporated to leave a pale yellow oil which was subjected to silica gel chromato­ 244 graphy. Elution with % ether in hexane furnished 147 mg (88/0) of 194? mp 9lj._95° (from ether-hexane) (lit. mp 95-98°); v^ c^-3 2232 (strong), D18X 1587, 1459, 1348, 1312 , 1301 , 1153, 1130, 954, 939, and 895 cm"1;

C Cl3 6*95-7.45 (m, 4h, aryl), 5 .6 7 (q of m, J5 ,6 =2.4 Hz, J6, 7 =

5.3 Hz, IH, Hs), 5-14 (q, J7, 8 = 2.4 Hz, 1H, H7 ), 4.16 (d, IH, H5),

3.85 (d, J2 , 8 =7.3 Hz, IH, H2), and 3 .2 8 (q of d, JXj8 =0.6 Hz, IH,

Hq); 13C nmr (CDC13) 148.0, 134.0, 136.4, 127.0, 126.8, 125.4, 121.1,

119.0, 118.9, 58.6, 45.9, 44.2, and 43-1 ppm; m/e (calcd) 179-0735,

(found) 179.0739.

B. Cyclization of 193- A 200 mg (0.927 mmol) sample of 193 was treated with potassium tert-butoxide (200 mg, 1.78 mmol) as described above. Chromatography of the crude product on silica gel afforded

145 mg (871o) of 194, mp 94-95°.

C. Cyclization of 191 From 27 mg of 191 and 27 mg of KOtertBu, there was obtained after chromatography 19 mg (85%) of 194, mp 94-95°.

2,3,6,7-Dibenzo-exo-4-chloro-anti-8-cyanobicyclo[3 - 2. l]octa-2,6- diene (Il6). A solution of 400 mg (1.95 mmol) of 99 in 5 ml of dry

dichloromethane was treated with NC 310 mg (2.19 mmol) of CSI while Cl L H stirred magnetically under argon.

After 70 hr at the reflux

116 temperature, the yellow solution

was diluted with 5 ml of dimethyl- 245 formamide and the majority of the dichloromethane was removed by dis­ tillation. Upon heating the residual liquid at 80° for 50 hr under argon, a clear brown solution was obtained. This was poured into ice water and the product was extracted with ether (2 x 50 ml) and dichloromethane (50 ml). The combined organic phases were washed with water and saturated sodium bicarbonate solution before drying and evaporation. There remained 530 mg of a yellowish semisolid which was chromatographed on silica gel (35 g)* Elution with 71° ether in hexane afforded 45 mg (10$ based on recovered 99) of 116, pale yellow cry­ stals, mp 185.5-186. 5°.

121 Diethyl 2,3-Diazabicyclo[2.2.l]hept-5-ene-2,3-dicarboxylate (397).

A solution of 30.0 g (0.172 mol)

of diethyl azodicarboxylate in 30 ml ^N-CO Et / 2 of ether was stirred at 0° under a 'N-C02Et calcium sulfate drying tube while

5^7 12.0 g (0.182 mol) of cyclopenta-

diene in 20 ml of ether was added dropwise. The initial orange color lightened after one hour at 5° and the ice bath was removed. After stirring for 3 hours at 28°, the solu­ tion became colorless. At that time the ether was evaporated and the 121 product was distilled, bp 157-183° (l mm), (lit. bp 119-120° at 0.4 mm) to give 32.45 g (79$) of 397 as a viscous oil.

121 Diethyl 2,3-Diazabicyclo[2.2.l]heptane-2,3-diearboxylate (398). A solution of 32.45 g (0.135 mol) of 397 in 50 ml of absolute ethanol containing 0.1 g of 5$ palladium on carbon was placed in a Paar hydro- genator at 50 psi of hydrogen.

After 2 hours the solution was re­ N-C02Et N-C02Et moved and filtered and the solvent was evaporated to furnish 32.41 g 398 (99-1%) of 398.

121 2,3-Diazabicyclo[2. 2. l]hept-2-ene (399). Nitrogen was bubbled

through 150 ml of ethylene glycol

with concomitant mechanical stir­

ring and heating for 25 minutes

before 35 g (0.536 mol) of potassium 399 hydroxide was added. When the temperature of this solution had reached 125°, 32 .^.' g (0.13 )! mcl) cf 398 was added dropwise under dry nitrogen at a rate that maintained the reaction temperature at 124-125°. After 1 hour the reaction mixture was allowed to cool and was poured onto a mixture of 60 ml of concentrated hydrochloric acid and 150 g of ice. The pH was adjusted from 1 to T with concentrated ammonium hydroxide. The addition of 20 ml of 2 M cupric chloride solution in 1 ml portions caused the yellow solution to turn red. The pH was adjusted from 3 to 6 with 2 ml of concentrated ammonium hydroxide solution. This process was repeated until 75 ml of the cupric chloride solution had been added. Filtration of this mixture gave the bright red solid cuprous chloride hemihydrate complex of 399-

The complex was rinsed with water into 100 ml of saturated ammonium chloride solution. The resulting slurry was filtered. The red cake was washed with 95% ethanol (2 x 100 ml) and cold water (2 x 100 ml), then 247 partially dried by drawing air through it for 30 min to give 38.5 g of complex, which was suspended in a solution of 15 g of sodium hydroxide in 200 ml of water. The resulting mixture was continuously extracted with pentane for 52 hours. The pentane extract was dried at 0° and the pentane was fractionally distilled to leave 14.3 g (79%) of 399«

121 Bicyclo[2.1.0]pentane (224). A 14.3 g (0.149 mol) sample of 3ftft was

heated to 60° to remove residual

pentane through a 30 cm Vigreux

column. The receiver was cooled to

-78° while the flask was heated to 224 170°. After about 3 hours the

column began to cool and the flask was gradually heated to 200°. The resulting 10 ml of distillate was dried and distilled to give 3*93 g (39%) of bicyclopentane (224), bp 121 39-44 .5° (lit. bp 45.5°).

15,IIS Addition of CSI to Bicyclopentane (224). A solution of 3*00 g

(44.0 mmol) of 2g4 in 25 ml of

H dichloromethane was stirred under

nitrogen at -5 to 0° while a solution

of 6.24 g (44.1 mmol) of CSI in 10

ml of dichloromethane was added

dxopwise. The reaction mixture was allowed to warm to room temperature. Fourteen hours later the reaction mixture was refluxed for 4 hours. 248

The contents were cooled to 0° and 50 ml of ether was added, fol- T 4 lowed by saturated sodium sulfite (50 ml) and 15$ sodium hydroxide solutions (10 ml). The temperature quickly rose to 20° and after cool­ ing back to 5°) 50 ml of additional sodium sulfite solution was added.

Three ml more 15$ sodium hydroxide solution was required to adjust the pH to 7. The ice bath was removed and stirring was continued while the mixture warmed to room temperature. The layers were separated and the aqueous phase was extracted with 50 ml of dichloromethane. The organic layers were combined and washed with 50 ml each of 5$ hydrochloric acid and saturated sodium bicarbonate solutions and brine, dried, and evaporated to leave 800 mg (16.4$) of yellow oil which appeared to be nearly pure

6-aza-7-oxobicyclo[3 .2.0]octane (225) by nmr.

The neat liquid was loaded onto a 1 x 15 inch column of silica gel packed in 30-60° petroleum ether. The column was eluted with increasing percentages of ether in petroleum ether, and finally with dichloromethane to give 500 mg (10.2$) of 225 as a yellow, viscous oil. Molecular distillation at 75° and 2. 5 mm afforded a lighter colored oil. Sublima­ tion (80°, 0.05 mm) and recrystallization (hygroscopic) from hexane at 15,118 -78° gave small white needles of 225? mp 47-49 (lit. mp 48.5-50.5°).

A mixed melting point of this material with product of the addition of rT)rl CSI to cyclopentene was also 47-49°; 100 MHz XH nmr: 6^ ^ 3 6.50 (br s,

IH, >NH), 4.05 (t, Jj.,5 =4.0 Hz, J5jX = 4.4 Hz, IH, >CH-N<), 3-41-3.57

(m, IH, >CH-C0-), 1.63-2.15 (m, 4h, Cg, and C4 methylenes), and 1.19-1.63

(m, 2H, C3 methylenes); v ^ Cla 3412 , 2930, 2870, 1757 (strong), 1355, IUaX 1335, 1181, IO69, 1037, 1021, 959, 923, and 576 cm"1. 249

5 Addition of CSI to Cyclopentene (227). A solution of 3*26 g (47.8 mmol) of distilled 227 in 25 ml of dichloromethane was stirred under nitrogen at 3-5° while a solution of 6.80 g (48.0 mmol) of CSI in 5 ml of dichloromethane was added dropwise during 15 minutes. After 18 hours at room temperature, the reaction mixture was refluxed for 12 hours to complete the reaction. This mixture was added to a rapidly stirred mix- 7 4 ture of 75 ml of saturated sodium sulfite solution and 50 ml of ether at 5-10°. The pH was maintained at 7 to 9 hy adding 10$ potassium hydroxide solution. After 15 minutes, with warming to room temperature,

the layers were separated and the aqueous layer was extracted with 50 ml

of dichloromethane. The combined organic layers were washed with 50 ml each of 5$ hydrochloric acid and saturated sodium bicarbonate solutions,

dried, and evaporated to give 2.72 g (51.2$) of 225 as a viscous yellow

oil.

This material was dissolved in ether and crystallized from hexane

at -78° (hygroscopic) to give 225_ as white needles, mp 46-47.5°. Sublima- 5 tion (40°, 0.05 mm) and recrystallization gave a best mp 47-49° (lit.

mp 50-51°.

302 l,3-Dibromo-l,3-dimethylcyclobutane (400). A 5 x 40 cm tube equipped

with a Dry Ice condenser was cooled CH3 with liquid nitrogen under nitrogen. Br f— At least 113 ml (2.0 mol) of methyl- Br acetylene was condensed into the CH3 2^00 tube. On top of this was condensed

at least 45 ml (l. 5 mol) of hydrogen 250 bromide. The liquid nitrogen bath was replaced by a Dry Ice-isopropyl alcohol bath. After 8 hours the two layers had melted and diffused together. After 6j hours at -78° the solution was allowed to warm gradually. At -60° the reaction mixture was refluxing and at -34° the temperature stabilized. After 10 hours, the temperature was 0° and the bath was removed. The solution began to bump vigorously.

The residual pale yellow liquid was fractionally distilled up to

57° and 90 mm. The residue which became brown and then deep blue-green, was filtered through Pyrex wool, diluted 1:1 with pentane, and cooled to -78°. The resulting gray-green solid was filtered and sublimed at 40°

(0 .1 mm) to give 9*52 g (0.03 %) of 400_ as a white solid, mp 50-54°.

123 1,3-Dimethylbicyclobutane (228). A mixture of 80 g (0. 40 mol) of

mercury and 100 ml of mineral oil

were heated under a stream of

argon to l80° and then 1.4 g (0.20

mol) of lithium wire in x/& inch

pieces was added and pushed one at

a time into the mercury with a glass 303 rod. After the mixture had cooled the 1.7% lithium amalgam lumps were transferred into hexane and crushed with wire cutters. The resulting gravel was washed with hexane and dioxane.

To a solution of 9-50 g (59-5 mmol) of 400 in 60 ml of dioxane, was added under nitrogen 80 g of such lithium amalgam. This mixture was stirred at ambient temperature for 50 hours under an ice cooled condenser.

The product mixture was centrifuged, the clear solution was poured off, 251 an equal volume of dioxane was added to extract the residue, and the process was repeated. The dioxane solutions were combined and fraction-

1 3 0 , 138 ally distilled to give 2.0 g ( 60f o ) of 228 , bp 49-55° (lit. bp

54.3-54.5°).

15, 119 Addition of CSI to 1,3-Dimethylbicyclobutane (228). A solution of

0.90 g (10.9 mmol) of 228_ in 20 ml

of dichloromethane was stirred

under nitrogen and cooled to -78°.

To this solution was added dropwise

229 during 1 hour a solution of 1.58 g

(11. 2 mmol) of CSI in 30 ml of dichloromethane. After 6 hours at -7&°, the reaction mixture was stirred at -10 to 15° for 14 hours.

The contents were cooled to -5° and stirred rapidly while 25 ml of T4 saturated sodium sulfite solution was added dropwise using 15% sodium hydroxide to adjust the pH to 8. To this emulsion was added 25 ml of ether with continued stirring for 1 hour while warming to room temperature.

The aqueous layer was separated and extracted with 50 ml of dichloro­ methane. The organic layers were combined and washed with 2 x 100 ml of water, diluted with 200 ml of ether, and washed with 50 ml of 5°k hydro­ chloric acid and 50 ml of saturated sodium bicarbonate solution. The

organic phase was dried and the solvent was evaporated to leave 815 mg

of pale yellow oil.

This material was applied to a column of 75 g of silica gel packed

in 30-60° petroleum ether. The column was eluted with slowly increasing 252 percentages of ethyl acetate in petroleum ether. Elution of the column with pure ethyl acetate gave 59 mg (4.3$) of crude 2-aza-l, 4-dime thy 1-

5~oxobicyclo[2.2.0]hexane (229) as a yellow oil.

Attempts to crystallize 229 at -78° from ether-hexane failed, so

the oil was sublimed onto a Dry Ice cold finger at 90° and 0.5 mm to

15 PT)P1 give 31 mg of pale yellow oil (lit. mp 74.5-75 ); 6^ 3 6.37 (br s,

IH, >NH), 1.77-2.43 (m, 4h, methylenes), 1.28 (s, 3H, Cj. methyl), and 1.20

TT p p l (s, 3H, C4 methyl); v"2"' 2 3390, 2960, 2862, 1751 (strong), 1450, 1381 , HlcU£ 1340, 1316 , 1244, 1170, 1140, 1018, and 904 cm-1.

199 2,2-Bis(bromomethyl)-l,3-Pi*opanediol (2g6>). A mixture of 204 g (l.50

mol) of pentaerythritol (295) i*1 1250

ml of glacial acetic acid and 25 ml

(BrCH2 )2C(CH20H) 2 (0 .2 2 mol) of 48% hydrobromic acid

was heated to dissolve 295 before

296 680 ml (6.05 mol) of 48$ hydro­

bromic acid was added dropwise

during 1 hour. After refluxing for 18 hours, 1.5 & of solvent was

removed by distillation at reduced pressure. One liter of absolute

ethanol was added, and the solution was stirred overnight at ambient

temperature. Distillation of more than 1 liter of ethyl acetate followed

by evacuation of remaining solvents on a steam bath left a heavy,

orange-brown oil.

The oil was mixed with 250 ml of toluene, the solvent was removed

by distillation, and the residue was kept in vacuo at room temperature 253 overnight. The oil was crystallized under benzene in the refrigerator.

Recrystallization from water gave 160 g (4o.7j6) of a beige solid 296,

1 9 9 nr\ni mp 107-112° (lit. mp 109-110°); 3 3-76 (s, 4h, -CH2Br), 3 .5 2

(s, 4h, -CH20H), and 2.10 (br s, 2H, -OH).

109 2-Phenyl-5,5-dibromomethyl-l,3 -dioxacyclohexane (297)« A mixture

of 81.5 g (0.311 mol) of 296 in

200 ml of benzene was heated at -0 (BrCH2)2(^ reflux while 35*0 g (0.53 mol) of

benzaldehyde and 0.3 g of ja-toluene-

29J sulfonic acid was added. Water

was removed with a Dean-Stark trap.

Ihe cooled solution was washed with 100 ml of saturated sodium carbonate solution and dried over anhydrous potassium carbonate. Rotary evapora­ tion left a yellow oil which was crystallized from 50 ml of methanol in the freezer to give 105.5 g (96*9$) °f pale yellow solid, mp 68.5-70°

(lit. mp 68.0-68.3°); 7*18-7-65 (m, 5H, aromatics), 5*36

(s, IH, benzyl), 3*62-4.42 (m, 4h, C4 and C6 methylenes), 3*94 (s, 2H,

-CH2Br), and 3.24 (s, 2H, -CH2Br),

199 Isoamyl 7-Phenyl-6,8-dioxaspiro[3,5]nonane-2,2-dicarboxylate (298).

In 500 ml of isoamyl alcohol was

dissolved 12.7 g (0 .5 5 2 mol) of (i-C5Hn-02C ) 2 Q C 0> - ^ P ) sodium. Diethyl malonate (120 g,

0.75 mol) was added, and 200 ml of 298 ethanol and isoamyl alcohol was removed by distillation through a twelve inch Vigreux column until the boiling point reached 127°. Refluxing was discontinued while 88.0 g

(0.2*+8 mol) of 297 was added. This solution was refluxed for 20 hours and distilled again to remove an additional 200 ml of solvent.

The cooled liquid was partitioned between 300 ml each of ether and water. The aqueous layer was extracted with 2 x 200 ml of ether. All ether layers were combined, washed with water, and filtered through potassium carbonate. The solution was evaporated and the residue distilled to give 74.14 g (69*1$) of solid, bp 210-225° at 0.05 mm 1 9 9 pT)Pl (lit. bp 190-208° at 0.1+ mm); 6^ 3 7.22-7*57 (m, 5H, aromatics),

5.38 (s, IH, benzyl), 4.l8 (t, J = 6.5 Hz, 4h, -C02-CH2-), 4.15 (d,

J = 11 Hz, 2H, C5 and C9 methylenes), 3 .70 (d, J = 11 Hz, 2H, C5 and C9 methylenes), 2.78 (s, 2H, cyclobutyls), 2.21 (s, 2H, cyclobutyls), and

0.78-1.95 (m, 18H, isobutyl groups).

189 7-Ehenyl-6,8-dioxaspiro[3,5]nonane-2,2-dicarboxylic Acid (299)*

A solution of 74.14 g (0.171 mol)

of 298 in 150 ml of absolute

ethanol was poured into a hot

solution of 33 g (0.59 mol) of 299 potassium hydroxide in 800 ml of

absolute ethanol. Heating on a steam bath was continued for 30 minutes while the product precipitated.

The mixture was cooled to 0° and filtered. The white solid was slurried in absolute ethanol, filtered, and washed with ether. Air drying over­ 255 night gave 69.5 g of white powder, which was dissolved in 400 ml of water and stirred while 150 ml of 3 N hydrochloric acid was added (pH 3)*

The precipitate was collected and recrystallized from benzene. The yield of the resulting white solid, mp 188.5-I89.5° (gas evolution),

(lit. mp 185-186° dec.) was 34.1 g (68.2$); 11.88 (br s, 2H,

-CO2H), 7.15-7.85 (m, 5H, aromatics), 5-61 (s, IH, benzyl), 4.49 (d,

J = 11 Hz, 2H, C5 and C9 methylenes), 3*88 (d, J = 11 Hz, 2H, C5 and Cg methylenes), 3.28 (s, 2H, cyclobutyls), and 2.63 (s, 2H, cyclobutyls).

1 9 9 1,1,3,3-Cyclobutanetetracarboxylic Acid (300). A mixture of 57*3 g

(0.196 mol) of 299 in 45 ml of

2 N nitric acid and 10 ml of water

(H02c)2 (0 0 ^ ) 2 was refluxed for 15 min. The

two phase system was cooled and

500, partitioned between ether and water.

The aqueous layer was washed with ether and the ether layers were combined and extracted with 20 ml of water. The combined aqueous layers were heated on a steam bath for 1 hour before adding a mixture of 200 ml of 70$ nitric acid and 5 ini of

90$ (fuming) nitric acid at 70-80° during 2 hours. The evolving nitrogen dioxide was vented through a reflux condenser. The resulting yellow solution was refluxed for 3 hours and allowed to cool overnight.

This colorless solution was distilled (45 mm) at up to 100° until only white solid remained. The solid was cooled to 0° and 15 ml of 97$ formic acid was added slowly. This mixture was allowed to 256 warm to room temperature. The formic acid addition process was repeated twice followed by JO ml of formic acid, added at room temperature.

The remaining yellow solution was reduced by vacuum distillation to leave a tacky solid which was stirred in 800 ml of ether until the product was a crystalline powder. Filtration and vacuum drying gave 18.6 g of yellow solid. The filtrate was decolorized with char­ coal, diluted with 100 ml of benzene, and concentrated to 100 ml to permit crystallization of 25.4 g of JOO. The total product, 44.0 g

(96.8$), was recrystallized from ether-benzene to give white solid, 1 9 9 T)OT) mp 209-211° (gas evolution) (lit. mp 205° dec); 6DSg J.10 (s, 4h, cyclobutyls).

199 cis-1,5-Cyclobutanedicarboxyllc Anhydride (J0l). A flask containing

40.0 g (0.172 mol) of J00 was

heated to 220° for 15 min while

gas evolved. After cooling, 60 ml

of acetyl chloride was added

through the condenser. The mixture

boiled spontaneously and was refluxed for 2 hr. Distillation of the excess acetyl chloride and acetic anhydride left a tacky, dark brown solid, which was distilled from 110° and 2 mm to 130 ° and 0 .0 2 mm to give 10.57 g (48.7$) of white

3 OI. Recrystallization of this material from benzene gave large color- 199 less needles, mp 150-152° (lit. mp 131 -132 °); mass spec: m/e

(relative intensity): 127 = P+1 (0*4), 126 (0 ), 82 (66), 54 (100), 257 CDCI3 3.30 (t of m, J = 6.0 Hz, 2H, bridge­ 39 (79), and 27 (65); 6TMS heads), 2.53-2.91 (m, 2H, anti-cyclobutyls), and 2.37 (t of m, J = 8.0

Hz, 2H, syn-cyclobutyls); v^2<”^ 2 1820, 1800, 1768 (strong), 1306, max 1296, 1169, 1066, 999, 960 (very strong), and 927 cm"1.

2 0 0 Dimethyl cis-l,3-Cyclobutanediacetate (302). A mixture of 2.00 g

(15.9 mmol) of 301 and 2 ml of 6 N

CH2 -CO2-CH3 hydrochloric acid was heated to

reflux for 15 min and allowed to

CH2-C02-CH3 cool. The crystalline diacid was

extracted with ether (2 x 100 ml). 302 The ether layers were combined,

d which was mixed with 40 ml of dry benzene, 7 .0 ml (97 mmol) of thionyl chloride, and

5 drops of pyridine, and stirred for 2.5 hours with protection from

atmospheric moisture. The solution was refluxed for 1 hr, and the

solvent and excess thionyl chloride were removed in vacuo. The residual

oil was twice dissolved in 40 ml of dry benzene and evaporated to leave

3 .0 6 g of brown oil.

The diacid chloride was dissolved in dry benzene (2 x 20 ml) and

syringed slowly into a solution of at least 100 mmol of diazomethane

in 400 ml of ether at -10°. The diazomethane solution was prepared by

distilling a precooled mixture of 200 ml of ether, 6 .0 g (150 mmol) of

sodium hydroxide in 50 ml of water, 50 ml of triethylene glycol and

25.0 g (70 mmol) of 70$ EXR-101. The distillate was collected at -78° 258 over potassium hydroxide, allowed to warm briefly to above 0° to melt any ice crystals and cooled to -78° again, under a drying tube. This

solution was decanted into a reaction flask containing 200 ml of THE.

After stirring for 20 hours, the yellow solution was distilled until the distillate was colorless. The yellow product solution was

evaporated to give 5*81 g of bright yellow solid, which was dissolved

in 50 ml of absolute methanol and cooled to 0° while a slurry of 0.5 g

(3 mmol) of silver acetate in 7 ml of triethylamine was added 1 ml at

a time. The reaction mixture darkened as nitrogen evolved. The

solution was refluxed for 30 min, decolorized with charcoal, and

filtered through Celite. The brown solution was evaporated and the

residue was dissolved in 50 ml of ether. The ether solution was washed with 50 ml each of 5$ hydrochloric acid and saturated sodium bicarbonate

solutions and dried. Rotary evaporation gave 2.94 g (92.6$) of 302

as a yellow oil. Molecular distillation at 40-95° and 0.05 mm afforded

2.1 g (66$) of pale yellow oil. A 5*20 g crude sample of 302 was

purified by application to 60 g of silica gel in hexane. Elution with

300 ml of hexane left the product to be eluted with 10 to 20$ ether

in hexane. The yield was 4.07 g (89$) of 302 as a colorless oil;

^TMS13 (s’ 6h’ -CO2CH3 ) and 2.20-2.55 (m, 10H, cyclobutyls and

-CH2CO2-); 2955 (strong), 2858, 1738 (strong), 1460, 1435, l4l8, IUctX 1358 , 1314 , 1257, 1237, 1194, 1170, 1097, 1045, 995, 878, 841, 8o4,

and 700 cm”1. 259 200,304 3,4-Bis(trimethylsiloxy)bicyclo[4.1.l]oct-3-ene (303). A solu-

tion of 2.00 g (9«99 mmol) of 302 ,

6.5 ml (51 mmol) of chlorotri- 0-Si(CHa) 3 methylsilane and 200 ml of dry

^ 3 3 ether was added dropwise during

30 min to a mixture of 3 .0 g (100 305 mmol) of 1 :1 sodium-potassium alloy in 100 ml of dry benzene at room temperature under nitrogen.

The mixture was refluxed for 20 hours and allowed to cool. The solu­ tion was filtered through glass wool and evaporated to give 2.37 g

(83«5%) of 303 as a yellow oil. This material can be molecularly distilled at 70° (0 .0 1 mm) or gas chromatographed on a 6 ft x J in. column of 10% QF-1 on 60-80 mesh Chromosorb G at 150°.

Chromatography on 50 g of Florisil of a similar crude product

(l g) gave 780 mg (75%) of 303 , when eluted with 10% ether in hexane;

2.1-2.8 (m, 10H, cyclobutyls and allylics) and 0.18 (s, 18h, IWd methyls), v™ 3 2965> ^35, 1709, 1685, 1371, 1317, 1253 (strong), ITlcLX 1188 (strong), 1168, 1114, 1068 (strong), 897 (strong), 874 (strong), and 844 (strong) cm-1.

304 Bicyclo[4.1.1]octan-3-on^4-ol (308 ). A solution of I.35 g (6.74

mmol) of 302 and 4.5 ml (35*5

mmol) of chlorotrimethylsilane

in 100 ml of ether was added to

1*6 g (55*2 mmol) of Na-K alloy

308 in 50 ml of benzene and refluxed 260 overnight as described in the preceding paragraph. The resulting 2.76 g of crude, yellow 303_ was dissolved in 15 ml of THF and stirred while nitrogen was bubbled into the solution for 10 minutes. Five ml of 5$ hydrochloric acid was added and bubbling was continued while the solu­ tion was refluxed for 1 hour. The acid was neutralized by adding 1 g of CaC03 and stirring for 1 hour.

The solvent was evaporated to give 2.3 g of oil which was par­ titioned between 50 ml of water and 200 ml of ether. The layers were separated and the aqueous layer was extracted with ether (2 x 100 ml).

All ether layers were combined, dried, and evaporated to give 0.80 g

(84.6$) of 508. Some of this oil was gas chromatographed at l40° on a

6 ft x 4 in. column packed with 10$ QF-1 on 60-80 mesh Chromosorb G.

The crude acyloin can be chromatographed on a Florisil column packed in 30-60° petroleum ether containing 10$ ether. Elution with

40 and 50 percent ether-petroleum ether gave 308 as a colorless oil;

6TM3l3 4*T° (q’ Ja = 10,5 Hz’ Jb = T*5 Hz, IH, >CH-0-), variable 3.5-4.45 (br s, IH, -OH), and 0.08-3.07 (m, 10H, methylenes and cyclo­ butyls); y^ ; 13 3495 (broad), 2937, 2870, 1704 (strong), 1440, 1408,

1348, 1253 , 1242-1190 (broad), 1108, 1067 (strong), 1039, and 1006 cm-1.

Anal. Calcd for C8H1202: C, 68.55; H, 8.63 .

Found: C, 68.96; H, 8.86.

2 0 x 4-Acetoxybicyclo[4.1. l]octan-3-one (309). A solution of 425 mg

(3 .0 mmol) of 308 _in 3 ml of pyridine was stirred under nitrogen at 0° while 4 ml of acetic anhydride was added dropwise. The solution was stirred for 30 minutes at 25° and 1.5 hours at 50°. The yellow product 261

solution was poured into 50 ml

each of ether and ice water with ,0 0 vigorous stirring. The layers ll C CHs were separated and the aqueous

209 phase was extracted with ether (2 x

50 ml). The combined ether layers were washed with 5°b hydrochloric acid, saturated sodium bicarbonate solution, and brine before drying. Chromatography on silica gel packed in hexane gave 400 mg (73$) of & pale yellow oil when eluted with 50$ ether-hexane.

An analytical sample of 309 was obtained by vapor phase chroma­ tography on a 6 ft x ^ in. column of 10$ QF-1 on 60-80 mesh Chromosorb

G at 175°. Recrystallization from ether-hexane gave needles, mp

82-82.5°; 6 ^ 13 5-53 (q, Ja = H Hz, Jb = 7 Hz, IH, >CH-0-), 1.1-3.1

(m, 10H, bridgeheads, cyclobutyls, and methylenes), and 2.13 (s, 3H,

-O2C-CH3 ); vlCS l3 2935, 2870, 1736 (strong), 1722 (strong), 1368 , 1316 ,

1170-1265 (broad), 1125, 1108, 1021-1062 (broad), 966, and 904 cm”1; mass spec, m/e (calcd) = 182.0943, (obs) = 182.0947.

202 Bicyclo[4.1. l]octan-3-one (310). A solution of 850 mg (4.66 mmol)

of 309 20 ml of THF was added

q dropwise over 10 minutes to a

solution (deep blue) of 5*0 g

(125 mmol) of calcium powder in

310 250 ml of liquid ammonia (distilled

from lithium) under argon. After 262

10 minutes, the still deep blue solution was cooled to -78° and quenched with 2 ml of bromobenzene (blue-gray suspension) and 25 ml of water. The emulsion was poured into 500 ml of ice water and 500 ml of ether. This mixture was filtered through Celite, the layers separated, and the aqueous phase extracted with ether (200 ml) and dichloromethane

(2 x 150 ml). All the organic layers were combined and washed with water, 5$ hydrochloric acid, and saturated sodium bicarbonate solution prior to filtration through potassium carbonate and drying.

The product solution was filtered, evaporated (30° at 35 mm), and applied to a 30 g column of Florisil packed in pentane. Elution with pentane gave bromobenzene, benzene, and other hydrocarbons.

Elution with 20-50$ ether in pentane gave 550 ^ (92$) of 310 as a yellow oil. Vapor phase chromatography at 145° on a 6 ft x i in. column of 10$ QE-1 on 60-80 mesh Chromosorb G with a flow rate of 100 ml/min afforded pure 310 (t = 5 .1 minutes) as a colorless oil;

^TMS^3 “5 *^ (*, with peaks at 2.80, 2.69, and 2.58, 8h, bridgeheads and methylenes) and 1.3 -2.3 (m, with peaks at 1.9 1? 1.80, and 1.59?

4h, cyclobutyl methylenes); 2930, 2862, 1697 (strong), 1414-1447

(broad), 1351, 1315, 1127, 1104, 985, 942, and 906 cm-1.

Anal. Calcd for C8H 120: C, 77-38; H, 9*74.

Found: C, 77.06; H, 9-82.

305 3,4-Bis(diethylphosphonyl)bicyclo[4.1. l]oct-3-ene (511). To a solution of 1.00 g (3.51 mmol) of 505 in 10 ml of THF was added 5 .6 ml

(8.4 mmol) of 1.5 M methyllithium in ether at ambient temperature.

The resulting bright yellow solution was stirred for 1 hr and 1.51 g (8.74 mmol) of diethyl chloro- 0 II phosphate dissolved in 10 ml of o p (o c h 2c h 3 )2 THF was added. The solution which o p (o c h 2c h 3)2 II turned clear was stirred overnight. 0 Evaporation left an oily solid that 311 was partitioned "between 100 ml of ether and 50 ml of water. The aqueous layer was extracted with chloroform (2 x 50 ml). The combined organic layers were dried, filtered through Celite, and evaporated to give 1.09 g (75%) of 511 CDC1 as a pale yellow oil, that was not further purified; 6^ , 3 5*7-4-. 4-

(five-line multiplet, 8h, -P0CH2), 1.0-2.8 (overlapping multiplets,

10H, allylies, "bridgeheads and methylenes), and 1.55 (three-line multiplet, 12H, methyls); vneat 2910, 1725, 1265, II65, IO55, 980, ITlStX and 84-0 cm-1. An accurate mass of the parent ion could not "be obtained. cis-Bicyclo!-^. 1.1]octane-5,4—diol (512). A. Reduction of 505* A mix­

ture of 530 mg (1.86 mmol) of 505,

10 ml of absolute ethanol, and

150 mg (4.0 mmol) of sodium

borohydride was stirred at room

temperature under a condenser

and drying tube for 2 hours and then refluxed for 2 hours. The solution was evaporated and the remain­ ing white solid was dissolved in 40 ml of 5% hydrochloric acid and 50 ml of ether. The aqueous layer was extracted with ether (2 x 100 ml) and dichloromethane (100 ml). All organic phases were combined, dried, and evaporated to give 460 mg of off-white solid. This material was dissolved in dichloromethane and crystallized from hexane to yield 206 mg (77.8$) of 312. The product was sublimed (70°, 0.02 mm) and recrystallized from ether-hexane to give colorless platelets, mp 129*5-

130 .5°; mass spec, m/e (calcd) = 142.09963, (obs) = 142.09960; m/e

(relative intensity), 142 (2.1 ), 124 (6.5 ), 101 (ll), 83 (23 ), 80 (17),

71 (21), 70 (100), 67 (17), 57 (22), 55 (50), 43 (19), 41 (23), and

39 (19); 6^ ls 4.26 (t of m, J = 5-5 Hz, 2H, X3H-0-), variable 2.4-3 .6

(br s, 2H, -OH), 2.38 (br s, 2H, cyclobutyls), 2.02-2.62 (m, 3H, cyclo­ butyls), 1.82-2.02 (q of m, J = 6.0 Hz, J = 5*5 Hz, 4h, methylenes), and 1.23-1.55 (q of m, J7>7 =9*5 Hz, Jlj7 = 5*5 Hz, IH, cis, syn cyclobutyls); 13C nmr (FT mode, 5 KHz sweep width, 8192 data points,

(SF0KD); 6 ^ 13 74.2 (d, 2C, C3 and C4), 35*6 (t, 2C, C2 and C5),

32.4 (d, 2C, Cx and C6), 31.1 (t, 1C, anti C8), and 29.9 (t, 1C, syn

C7); ^ 3590, 3425 (broad), 2997, 2932, 2875, 1433, 1323, 1237-1190 ulcLX, (broad), 1158, 1102, 1053, 1026, 996, 964, 924, 888, and 832 cm"1.

Anal. Calcd for c8H1402: c, 67.57; H, 9*92.

Found: C, 67.45; H, 9*74.

287,288 B. Reduction of 3?8. A solution of 500 mg (3 .6 mmol) of

508 in 3 ml of absolute ethanol was added dropwise to a mixture of 135 mg (3. 6 mmol) of sodium borohydride in 5 ml of absolute ethanol. The mixture was refluxed 2 hours and evaporated. The residue was dissolved 265 in 25 ml each of ether and water and the aqueous layer was extracted with ether (2 x 100 ml). All ether solutions were combined, washed with brine, and dried. Evaporation of the solution left TOO mg of crude oil.

The oil was chromatographed on Florisil packed in 30 -60° petroleum ether. Elution with 10$ methanol in dichloromethane gave

160 mg of crude diol, which was identical to 312 , described in the preceding experimental.

a n cis-3)^-Pimesyloxybicyclo[4.1. l]octane (313). A solution of 300 mg

(2.11 mmol) of 312 in 10 ml of

dichloromethane was added drop-

wise to a solution of 490 fil j/ \ OSO2CH3 0S02CH3 (6*33 mmol) of methanesulfonyl

chloride in 5 ml each of dichloro- 313 methane and pyridine at 0°. The solution was stirred for 2 hours at 0° and for 30 minutes at room temperature before pouring into 100 ml of ice water. The layers were separated and the aqueous layer was extracted with ether and dichloro- methane. £he organic layers were combined and washed with ice water, cold 5$ hydrochloric acid, and cold sodium bicarbonate solution before drying. Evaporation gave 650 mg of pale orange oil which was crystallized from dichlorome thane-hexane to give 450 mg (71*5$) of 323, mp 133.5-13^5°; 6 ^ 13 5.38 (t, J = 6 Hz, 2H, >CH-0-), 3-14

(s, 6h, -SO2-CH3 ), 1.80-2.78 (m, 9H, cyclobutyls and methylenes), and 1.47 (q, 1H, Ja = 11 Hz, Jt, = 6 Hz, 1H, syn cyclobutyl); VKCC13 3025 , IflclX 2942, 1355 (broad and strong), 1174 (strong), 993, 970 (strong), 94l

(strong), 909 (strong), 900 (strong), 876, 854 , and 845 cm"1; mass spec: m/e (relative intensity) 298 (0.05), 257 ( 0.17), 244 (0.25),

219 (0.75), 204 (1.3 ), 203 (8.1 ), 202 (6.3 ), 161 (58), 148 (3 0 ), 123

(25), 107 (28), 106 (88), 95 (24), 91 (3 2 ), 83 (100), 80 (3 2 ), 79 - doublet (69), 78 (44), 67 (38 ), 55 - doublet (47), 4l (82), and 39 (29).

Cyclic N,N-Dimethylphosphoramide of cis-Bicyclo[4.1.1]octane-3,4- diol (314) . 214 A solution of 350 mg (2.4 mmol) of 512, 486 mg (3 .0

mmol) of N,N-dimethyl dichloro- 215 phosphoramide and 590 mg (7*5 J r r ° K ° ‘^'^^"'^0' \l(CH3); miuul) of pyridine in 3 ulL of THF

was refluxed for 18 hours. 314 During this time the solution

darkened and a beige solid formed.

The mixture was partitioned between 150 ml of chloroform and 50 ml of water. The chloroform layer was washed with 5% hydrochloric acid and saturated sodium bicarbonate solutions before drying. The solvent was evaporated and the residue was dissolved in ether and filtered to give, on evaporation of the filtrate, 550 mg (97%) of a yellow oil, CDC1 which decomposed on silica gel; 6, ^ 3 3*8-5.8 (series of m, 2H,

>CH-0), 2.6-3.0 (4s, 6h, exo and endo -11(01^)2), and 0.7-2.5 (series of m, 10H, bridgeheads and methylenes). 267 2 1 6 4,6-Dioxa-5-thiotricyclo[7*1.1*03’~]undecane (315)* To a solution

of 710 mg (5.00 mmol) of cis-

bicyclo[4.1.l]octane-3,4-diol

(312 ) in 60 ml of xylene was added

1.25 g (7 .0 mmol) thiocarbonyl-

diimidazole. This solution was

distilled under nitrogen to remove

5 ml of cloudy distillate and refluxed for 8 hours. The volume was reduced to 20 ml by rotary evaporation. The solution was diluted with 10 ml of dichlorome thane before applying to a column of Florisil packed with 10% dichloromethane in 30 -60° petroleum ether.

After eluting the xylene with 10 and 2Cff> H2CC12 in petroleum ether, the product was eluted with40-80$> H2CC12 as a sweet-smelling pale yellow oil that slowly solidified. The yield of 315 was 853 mg

( 9 2 Recrystallization from ether-hexane gave white needles, mp 73.8-7^. 5°; -5.28 (t of m, J = k.2 Hz, 2H, >CH-0-), 2.53 (m,

3H, bridgeheads and methylenes), 2.28 (m, 5H, methylenes), and 1.03-1.73

ttppi (two m, 2H, syn cyclobutyls); v^ 3 2995 , 2970, 2$b2, 2873 , 1382, 1 iilcUi 1337 (strong), 1321 (strong), 1292 (strong), 1187, 1152, IO98, IO33 ,

997, 973, and 932 cm”1.

Anal. Calcd for CaH1202S: C, 58.67; H, 6.56.

Found: C, 58.82; H, 6.53 .

216 Bicyclo[4.1. l]oct-3-ene (25^). A. Elimination of 315. A solution of 450 mg (2 .kh mmol) of 315 in 5 nil (2 k mmol) of distilled triethyl phosphite was heated to 150° under nitrogen and a condenser for 90 268

hours. The solution was cooled

and diluted with pentane and

applied to a 50 g column of

neutral, activity I alumina packed

in pentane. The column was eluted

with 200 ml of pentane. The

pentane eluate was reduced in volume by careful fractional distilla­

tion and the higher boiling residue was gas chromatographed at 80°

on a 6 ft x I in. column of l O f o QF-1 on 60-80 mesh Chromosorb G. The PDPT yield of 254, a colorless liquid, was 48 mg 6^ 3 5-57 (narrow m, J < 2 Hz, 2H, olefinics), 2.48 (m, 3H, bridgeheads and cyclobutyls),

2.26 (m, 4h, allylics), and 1.03-1.57 (m, 3H, cyclobutyls); 13 C nmr

(FT mode, 5 KHz sweep width, 16,384 data points, (SFOKD);

125.8 (d, 20, C3 and C4), 35*6 (t, 20, C2 and C5), 31*8 (d, 20, Cx

and Ce), and 30.3 (t, 2C, Cr and Ca); vmax**2CC12 3000, 2955 (strong), 2928 (strong), 2823, 1420, 1 1 9 6, 1067, 1027 (strong), 971 (strong), 9^3,

876, 807, and 651 cm-1.

Anal. Calcd for C8Hx2: C, 88.82; H, 11.18.

Found: C, 88.64; H, 11.6 3 .

214 . B. 'Lithium in Ammonia Reduction of 314. A solution of 3*0 g

(13.0 mmol) of 31 ^ in 20 ml of ether was added dropwise to a solution

of 2 .0 g (290 mmol) of lithium in 100 ml of liquid ammonia (distilled

from lithium). After 1 hour at the reflux temperature, the mixture

was cooled to -78°, diluted with 100 ml of pentane, quenched with 269 water, and poured onto 300 g of ice. The aqueous layer was extracted with pentane (3 x 150 ml). The pentane layers were combined and washed with 5$ hydrochloric acid and saturated sodium bicarbonate solutions before drying and fractional distillation. The remaining 10 ml of residue was diluted with 10 ml of pentane and passed through a column containing 10 g of alumina (pentane elution). A second fractional distillation left 200 mg (lk%) of 25^-, as verified by vpc and nmr.

212 C. Sodium Anthracenide Reduction of 313« Nitrogen was bubbled through 350 ml of THF, h. 80 g (26.9 mmol) of anthracene was added under nitrogen, and after dissolution, 0. 7 g (30 mmol) of sodium was added. Stirring was continued for 2 hours while the deep blue reagent formed. This solution was concentrated to 0.25 M by distilling out 250 ml of THF. The sodium anthracenide solution was added drop- wise under nitrogen to a degassed solution of 200 mg (0 .6 7 mmol) of 313 in 15 ml of THF at 0° until the reaction mixture remained deep blue.

The solution was stirred for 12 hours, cooled to 0°, and diluted with

100 ml of pentane. The reaction mixture was quenched with 50 ml of saturated ammonium chloride solution, water was added, and the layers were separated. The aqueous layer was extracted with 30-60° petroleum ether. The organic extracts were combined, washed with water, filtered through K2C03, and concentrated to 60 ml. This remaining solution was vacuum transferred (30° at 0.05 mm) to a Dry Ice trap, dried, and fractionally distilled to obtain 2 ml of residue which was gas chromatographed at 70° on a 6 ft x 4 in. column packed with 10/o 270

SE-30 on 60-80 mesh Chromosorb G. There resulted 3 mg (*$) of 2^k.

The mass spectrum of this product showed a significant m/e = 106, which corresponds to a double elimination product as an impurity in the desired olefin (m/e = 108).

213 / N D. Zinc Reduction of 315. A mixture of 30 mg (0.10 mmol; of

313, 150 mg (1.0 mmol) of sodium iodide, and 65 mg (l. 0 mmol) of activated zinc in 2 ml of HMPA was heated at 100° for 2k hr, cooled, and filtered. The residue was rinsed with ether (2 ml) and pentane

(30 ml) and the combined filtrates were washed with water (3 x 50 ml) and 10% sodium thiosulfate solution (50 ml) before drying. The filtrate was passed through a column containing neutral activity I alumina (5 g) using pentane elution to give 80 ml of solution. The solvent was removed by fractional distillation and the residue was purified by gas chromatography on a 6 ft x J in. column packed with

20% Carbowax 20 M on 60-80 mesh Chromosorb P at 110° with a flow rate of 80 ml/min. Collection of the only major peak (t ^ = k.$ min) gave

1.1 mg (10%) of 25k, as verified by retention time and 1H nmr spectro­ scopy.

204,205 E. Reduction of 3H» A solution of 800 mg (1.9^ mmol) of

311 and 288 mg (3 .8 9 mmol) of t-butyl alcohol in 20 ml of ether was added to a solution of 135 mg (1 9-5 mmol) of lithium in 100 ml of ammonia (distilled from lithium) at -78°. The solution was allowed to reflux for 2 hours and recooled to -78°. Pentane (25 ml) and water

(75 ml) was added carefully. The mixture was poured onto 400 g ice and 271 extracted with pentane (2 x 80 ml) and dichloromethane (2 x 50 ml).

The organic layers were combined and washed with water (2 x 100 ml),

5$ hydrochloric acid (2 x 100 ml), and saturated sodium bicarbonate solution (100 ml) before drying and fractional distillation to remove the solvents. Vapor phase chromatography on a 6 ft x i in. column of

10$ QF-1 on 6O-8 0 mesh Chromosorb G at 70° with a flow rate of 60 ml/min gave 17 mg (8.1$) of 254, which was verified by retention time (5*7 min) and nrar spectroscopy.

Bicyclo[4.1.l]octane (253). A. Hydrogenation of 254. A mixture of

16 mg (0 .1 5 mmol) of 254, 20 ml

of ether, and 100 mg of 10$ palla­

dium on carbon was hydrogenated

at room temperature and atmos­

pheric pressure for 3 hours.

(The same reaction is complete in

2 hours in methanol.) The solution was filtered through 5 g of silica gel (pentane elution). The solvent was removed by fractional distilla­ tion and the residue was purified by gas chromatography on a 6 ft x 4 in. column of 5$ 0V-11 on 60-80 mesh Chromosorb G at 80° with a flow rate of 65ml/min. Collection of the material at 4.5 min gave 12 mg (75$)

PDP1 of 253_as colorless low melting solid; 3 1.95-2 .8 0 (m, 4h, bridge­ heads and anti Hy and H s ) and 1.00-1.95 (m, 10H, methylenes); 13C nmr

(FT mode, 5KHz sweep width, 1 6 ,3 8 4 data points); 34.0 (2C, C2 and

C 5 ), 3 1 .6 (2C, Cj. and C6), 2 9 .9 (2C, Cy and Ca ), and 25-5 (2C, C3 and 272

C4); v^ C l2 ^ (strong), 2855, 2840, 1603 (weak), 1458, 1447, 1232 ITIq A (weak), 986 (weak), and 858 (weak) cm”1; mass spec: m/e (calcd) =

110.10954, (obs) = 110.10976.

306 B. Clemmensen Reduction of 508. To a mixture of 700 mg (5.0 mmol) of 5 0 8, 2 ml of water, and 3 “ 1 of concentrated hydrochloric acid was added 5.0 g (l40 mmol) of activated amalgamated zinc. The mixture was heated at 100° for 5 hr with addition of 1 ml of concen­ trated hydrochloric acid every hour. The mixture was cooled, diluted with 50 ml of ether and 20 ml of water, and filtered. The aqueous phase was diluted with brine (20 ml) and extracted with ether (25 ml).

The organic layers were combined and washed with water and saturated sodium bicarbonate solution before drying. The solvent was removed by fractional distillation and the residue was dissolved in pentane and percolated through 10 g of alumina (neutral, activity I, pentane elution). Vapor phase chromatographic analysis (100°, JO ml/min, 6 ft x 4 in. 20% Carbowax) and nmr spectroscopy showed the product (253) to be contaminated with approximately 30% 254. Hydrogenation as above left only 253. Isolation by vapor phase chromatography on the above

Carbowax column gave 95 mg (17%) of 253 as a low melting white solid.

The retention times of 253 and 254 were 4.8 and 5-1 minutes,

respectively.

226 2,5-Dibromobicyclo[4.1.l]oct-3-ene (3 2 2 ). A solution of 4l mg

(O.3 8 mmol) of 254 in 4 ml of carbon tetrachloride containing 75 mg

(0.42 mmol) of N-bromosuccinimide and 5 mg of azobisisobutyronitrile 273

was refluxed for 2 hours. Vapor

Br phase chromatographic analysis at

this time showed olefin still

Br present. More KBS (100 mg, 0.56

^ 2 mmol) and AIBN was added and the

mixture was refluxed for an additional 2 hours. There remained only a trace of olefin so the mixture wascooled, filtered, and evaporated. The resulting dark yellow oil was applied in pentane to a 15 g column of silica gel and eluted only with pentane (15 ml fractions). Fractions 8 through 12 CDCl were combined to give 85 mg (84fo) of 3 2 2 , as a colorless oil; &TMS 3

5.73 (narrow m, 2H, olefinics), 4.93 (m, 2H, allylics), 2.67-3.20

(m, 2H, bridgeheads), and 1.18-2.67 (series of m, 4h, Cr and C8

methylenes); vCH2Cl2max 2995 (strong), 2855 (strong), 1725, 159^, 13^1, 1095-1005 (broad and strong), 94-5, 8 35-660 (broad and strong), and 599 cm”1; mass spec: 08H loBr+; m/e (calcd) = 184.99663, (obs) = 184.99696,

(m/e + 2 )'(1.0 3 ) = m/e.

229 Bicyclo[4.1. l]octa-2,4-diene (255)• A mixture of 150 mg (0.5 6 mmol)

of 5 2 2, 300 mg (4.6 mmol) of zinc-

copper couple, 100 mg (0 .6 0 mmol) 227 of potassium iodide, 50 mg

(0 .2 0 mmol) of sublimed iodine 255 and 2 ml of DMF was stirred under

nitrogen for 12 hr. Ten ml each of pentane and ether were added and the suspension was filtered through

10 g of alumina (neutral, activity i). The column was eluted with 50 ml of pentane and rinsed with 10 ml of ether. The eluent was washed with water (3 x 100 ml) and brine (100 ml) before drying. Removal of the solvent by fractional distillation and gas chromatographic isolation (110°, 65 ml/min, 6 ft x ^ in. 20$ Carbowax of 60-80 mesh

Chromosorb p) of the major component at 8.1 min afforded 18 mg (l8fo) p-npn of 255^ as a colorless oil; 6 ^ 3 5 .67-6 .5 5 (AA'BB' m, 4h, olefinics),

2. 73-3«28 (m, 2H, bridgeheads), 2.12-2.73 (“9 2H, anti CY and C8 methylenes), and 1.08-1.6 3 (m, 2H, syn C7 and C8 methylenes); 13C nmr

(FT mode, 5 KHz sweep width, 8192 data points); 5ODCI3 1 5 8 .6 (2C, C2 max and C5), 123.7 (2C, C3 and C4), 34.8 (2C, Cx and C6), and 21.4 (2C,

CY and C8 ); v ^ f f 12 3025 , 2940 (strong), 2858, 1601, 1384 , 1333, UlclX 1209, 1149, 1052, 984, 9 50, 924, 8 3 6 , and 802 cm-1; Xhexane 200-213 max (e 1600), 258 (shoulder), 266 (2800), 277 (3 8 0 0 ), 288 (3 7 5 0 ), and 301

(2000) nm; mass spec: m/e (calcd) = 106.07824, (obs) = 106.07842. REFERENCES

1. R. Graf, Chem. Ber., 8 9, 1071 (1956).

2. R. Graf, Angew. Chem., 80, 179 (1968); Angew. Chem. Intern. EcL. Engl., 7, 172 (1968).

5 . H. Bestian, Pare Appl. Chem. , 27, 6ll (1971)-

4. E. J. Moriconi and J. F. Kelly, Tetrahedron Lett., 1435 (1968).

5. H. Bestian, H. Biener, K. Clauss, and H. Heyn, Justus Lietigs Ann. Chem., 718, 94 (1968).

6. L. A. Paquette, M. J. Wyvratt, and G. R. Allen, Jr., J. Am. Chem. Soc., 92, 1763 (1970). I 7. E. J. Moriconi and W. C. Meyer, J. Org. Chem., 3 6 , 2841 (1971).

8. E. J. Moriconi and C. C. Jalandoni, J. Org. Chem., 35, 2073 (1970).

9. J. R. Malpass and N. J. Tweddle, Chem. Commun., 1244 (1972).

10. J. R. Malpass, Chem. Commun., 1246 (1972).

11. J. R. Malpass and N. J. Tweddle, Chem. Commun., 1247 (1972).

12. G. T. Furst, M. A. Wachsman, J. Pieroni, J. G. White, and E. J. Moriconi, Tetrahedron, 29, 1675 (1973)*

13. T. Sasaki, S. Eguchi, and H. Yamada, J. Org. Chem., 3 8 , 679 (1975)*

14. J. R. Malpass, Tetrahedron Lett. , 4951 (1972).

15. L. A. Paquette, G. R. Allen, Jr., and M. J. Broadhurst, J. Am. Chem. Soc., 93, 4503 (1971).

16. J. Haug, F. Lohse, K. Metzger, and H. Batzer, Helv. Chim. Acta, 51., 2069 (1968).

17. P. Goebel and K. Clauss, Justus Liebigs Ann. Chem., 722, 122 (1969).

18. E. J. Moriconi, C. F. Hummel, and J. F. Kelly, Tetrahedron Lett., 5325 (1969).

275 276

19. W. Adam, J, Baeza, and J. Liu, J. Am. Chem. Soc., 94, 2000 (1972); M. U. S. Sultanbawa, Tetrahedron Lett., 4569 (l968~)T and references contained therein.

20. L. A. Paquette and T. J. Barton, J. Am. Chem. Soc., 8 9, 5480 (1967).

21. P. Wegener, Tetrahedron Lett., 4585 (1967).

22. A. Fischli, H. Mayer, and W. Oberhansli, Helv. Chim. Acta, 57, 1477 (1974).

25. R. B. Woodward and R. Hoffmann, Angew. Chem., 8l, 797 (1969); Angew. Chem, Intern. Ed. Engl., 8, 781 (1969).

24. T. J. Barton and R. J. Rogido, Chem. Commun. , 878 (1972).

25. T. J. Barton and R. J. Rogido, Tetrahedron Lett., 3901 (1972).

26. E. J. Moriconi and W. C. Crawford, J. Org. Chem., 33, 370 (1968).

27. S. Winstein, J. Am. Chem. Soc., 8 5, 1516 (1 961), and references contained therein.

28. H. E. Zimmerman and R. M. Paufler, J. Am. Chem. Soc. , 8 2, 1514 (i9 6 0); H. E. Zimmerman and G. L. Grunewald, ibid., 867~1435 (1964); H. E. Zimmerman, R. J. Boettcher, N. E. Buehler, and G. E. Keck, ibid., 97, 5635 (1975).

29. H. E. Zimmerman, G. L. Grunewald, R. M. Paufler, and M. A. Sherwin, J. Am. Chem. Soc., 91, 2330 (1969).

30. L. A. Paquette, W. E. Volz, M. A. Beno, and G. G. Christoph, J. Am. Chem. Soc., 97, 2562 (1975); L. A. Paquette and W. E. Volz, ibid., 98, 2910 (1976).

31. T. G. Traylor, Accounts Chem. Res. , 2, 152 (1969); R. B. Turner, J. Am. Chem. Soc. , 8 6, 3586 (1964).

32. E. Haselbach, E. Heilbronner, and G. Schroder, Helv. Chim. Acta, 54, 153 (1971).

33* H. C. Brown, J. H. Kawakami, and K.-T. Liu, J. Am. Chem. Soc. , 95 , 2209 (1973); H. C. Brown and K.-T. Liu, ibid., 97, 2469 (1975); H. C. Brown and J. H. Kawakami, ibid., 97, 5521~(l975) and other papers in this series.

34. P. Bruckmann and M. Klessinger, Angew. Chem., 84, 453 (1972); Angew. Chem. Intern. Ed. Engl., 11, 524 (1972). 277

35* P» Radlick, R. Klem, S. Spurlock, J. J. Sims, E. E. van Tamelen, and T. 'Whitesides, Tetrahedron Lett. , 5117 (1968); H. H. Westberg and H. J. Dauben, Jr., ibid., 5123 (1968).

3 6 . K. Alder and G. Jacobs, Chem. Eer. , 8 6, 1528 (1953).

37- J- Daub and P. v. R. Schleyer, Angew. Chem. , 80, 446 (1968); Angew. Chem. Intern. Ed. Engl., 7, 468 (1968).

3 8 . Y. E. Rhodes, P. E. Schueler , and V. G. DiFate, Tetrahedron Lett. , 2073 (1970).

39» E. J. Snyder and B. Franzus, J. Am. Chem. Soc., 8 6; 1166 (1964); P. Laszlo and P. v. R. Schleyer, ibid. , 86", 1171 (1964); B. Franzus, W. C. Baird, N. F. Chamberlain, T. Hines, and E. J. Snyder, ibid., 90; 3721 (1968); A. P. Marchand and J. E. Rose, ibid. , 90, 3724~Q_968); M. C. Thorpe and W. C. Cobum, J. Org. Chem. , 3?, 2576 (1969).

40. K. Tori and K. Kitahonoki, J. Am. Chem. Soc., 8 7, 386 (1965) and references contained therein.

41. D. J. Cram and N. L. Allinger, J. Am. Chem. Soc. * j8, 2518 (19 56).

42. The spectrum of dimethyl maleate consists of two singlets at 6 3*78 (6h) and 6.24 (2H): ''Varian Spectra Catalog,'' Varian Associates, Palo Alto, Calif., 1962, spectrum 213.

43. M. Goldstein and A. Gervitz, Tetrahedron Lett., 4417 (1965) and relevant references cited therein.

44. P. Demarco, T. K. Elzey, R. Lewis, and E. Wenkert, J. Am. Chem. Soc., 92; 5734 (1970).

4 5 . R. G. Miller and M. Stiles, J. Am. Chem. Soc., 8 5, 1798 (1963).

46. K. Kitahonoki and Y. Takano, Tetrahedron Lett., 1597 (1963)*

47. H. E. Zimmerman, R. S. Givens, and R. M. Pagni, J. Am. Chem. Soc., 9 0, 6096 (1968).

48. H. E. Simmons, T. L. Cairns, S. A. Vladuchick, and C. M. Hoiness, Org. Reactions, 20, 1 (1973); R* Shank and H. Shechter, J. Org. Chem., 24, 1825 Tl959).

49. J. Furukawa, N. Kawabata, and J. Nishimura, Tetrahedron, 24, 53 (1968).

50. A. de Meijere and C. Weitemeyer, Angew. Chem. , 82, 392 (1970); Angew. Chem. Intern. Ed. Engl., 95 376 (1970). 278

51. A. de Meijere, 0. Schallner, and C. Weitemeyer, Tetrahedron Lett., 5483 (1973).

52. P. E. Schueler and Y. E. Rhodes, J. Org. Chem,, 39, 2063 (1974).

53* P* E. Schueler and Y. E. Rhodes, Abstracts, l64th National Meeting of the American Chemical Society, New York, N. Y. , August, 1972, No. ORGN-129: J. Haywood-Farmer, Chem. Rev., 74? 315 (1974).

54. W. Kirmse and K. H. Wahl, Chem. Ber., 107, 2768 (1974).

55* S. J. Cristol and N. L. Hause, J. Am. Chem. Soc., 74, 2193 (1952); S. J. Cristol, R. P. Arganbright, G. D. Brindell, and R. M. Heitz, ibid., 79, 6035 (1957).

56. C. A. Grob and J. Hostynek, Helv. Chim. Acta, 46, 1676 (1963); C. A. Grob and A. Gagneux, ibid., 42, 2006^~1753 (1959)*

57. R. S. Givens and L. Strekowski, J. Am. Chem. Soc., 97; 5867 (1975).

58. S. J. Cristol, R. P. Arganbright, and D. D. Tanner, J. Org. Chem. , 28, 1574 (1963).

59* D- Tanner and B. G. Brownlee, J. Am. Chem. Soc., 8 8, 771 (1966).

60. S. Masson and A. Thuiller, Bull. Soc. Chim. Fr., 4368 (1969).

61. S. J. Cristol, F. P. Parungo, and D. E. Plorde, J. Am. Chem. Soc., 87; 2870 (1965).

62. S. J. Cristol and R. K. Bly, J. Am. Chem. Soc. , 82, 6155 (i9 6 0).

63 . E. Zbiral and A. Stutz, Tetrahedron, 27, 4953 (1971).

64. A. Hassner, R. P. Hoblitt, C. Heathcock, J. E. Kropp, and M. Lorber, J. Am. Chem. Soc. , §2, 1326 (1970).

65. S. J. Cristol, R. Caple, R. M. Sequeira, and L. 0. Smith, Jr., J. Am. Chem. Soc., 87, 5679 (1965).

66. S. J. Cristol and B. B. Jarvis, J. Am. Chem. Soc., 8 8, 3091 (1966).

67. V. I. Sokolov, Bull. Acad. Sci. USSR Div. Chem. Sci., 1213 (1968); F. R. Jensen, J. J. Miller, S. J. Cristol, and R. S. Beckley, J. Org. Chem., 37, 4341 (1972). 279

6 8. S. Masson and A. Thuiller, C. R. Acad. Sci. Paris, Ser. C, 266, 987 (1968).

69. Method of R. H. Shapiro and J. H. Duncan, Org. Syn., 51, 66 (1971).

70. K. Tori, Y. Takano, and K. Kitahonoki, Chem. Ber., 97, 2798 (1964).

71. G. W. Klumpp, J. W. F. K. Bamick, A. H. Veefkind, and F. Bickelhaupt, Recueil des Travaux Chimiques des Pays-3as, 8 8, 766 (1969). S. Winstein, M. Ogliaruso, M. Sakai, and J. M. Nicholson, J. Am. Chem. Soc., 89^ 3656 (1967).

72. C. W. Jefford, S. N. Mahajan, and J. Gunsher, Tetrahedron, 24, 2927 (1968).

73. S. J. Cristol, J. R. Mohrig, and D. E. Plorde, J. Org. Chem. , 30, 1956 (1965); A. R. Katritzky and B. Wallis, Chem. Ind. ^London), 2025 (1964).

74. T. Durst and M. O'Sullivan, J. Org. Chem., 35, 2043 (1970).

75. 3.

7 6. S. J. Cristol, G. C. Schloemer, D. R. James, and L. A. Paquette, J. Org. Chem., 37, 3852 (1972).

77. S. J. Cristol, T. W. Russell, J. R. Mohrig, and D. E. Plorde, J. Org. Chem. , 31, 58l (1 966).

78. N. A. LeBel and J. E. Huber, J. Am. Chem. Soc., 8 5, 3193 (1963).

79* H. Tanida, K. Tori, and K. Kitahonoki, J. Am. Chem. Soc., 8 9, 3212 (1967).

80. H. Hart and G. M. Love, J. Am. Chem. Soc., 95, 4592 (1973)*

81. F. Pwu, Ph.D. Thesis, University of Colorado, 1961*, Piss. Abstr., 22, 2996 (1962); S. J. Cristol, F. P. Parungo, D. E. Plorde, and K. Schwarzenbach, J. Am. Chem. Soc., 8 7, 2879 (1965)-

82. E. Cioranescu, A. Mihai, G. Mihai, E. Elian, and C. D. Nenitzescu, Rev. Roumaine Chem., 10, 175 (1965)-

8 3 . H. Vorbruggen, Tetrahedron Lett., 1631 (19 68).

84. E. M. Burgess and W. M. Williams, J. Am. Chem. Soc., 94, 4386 (1972). 280

85. G. Schroder and J. F. M. 0th, Angew. Chem. , 79, 458 (1967); Angew. Chem. Intern. Ed. Engl. , 6 , 4l4 (1967); H.-P. Loffler and G. Schroder, Angew. Chem. , 80, 758 (1968); Angew. Chem. Intern. Ed. Engl., 7, 736 (1968); J. F. M. 0th, R. Merenyi, H. Rottele, and G. Schroder, Tetrahedron Lett., 39*+1 (1968); J. F. M. 0th, E. Machens, H. Rottele, and G. Schroder, Justus Liebigs Ann. Chem., 745, 112 (1971).

86. M. J. Goldstein, R. C. Krauss, and S.-H. Dai, J. Am. Chem. Soc., 94, 680 (1972).

87. W. von E. Doering, B. M. Ferrier, E. T. Fossell, J. H. Harten- stein, M. Jones, Jr., G. Klumpp, R. M. Rubin, and M. Saunders, Tetrahedron, 25, 39*+3 (1967).

88. G. R. Krow and K. C. Ramey, Tetrahedron Lett., 3141 (1971).

89. L. A. Paquette, T. J. Barton, and E. B. Whipple, J. Am. Chem. Soc. , 8 9, 5481 (1967); L. A. Paquette, J. R. Malpass, G. R. Krow, and T. J. Barton, ibid., 9 1, 5296 (1969).

90. L. A. Paquette, S. Kirschner, and J. R. Malpass, J. Am. Chera. Soc., 91? 3970 (1969).

91. R. Hoffmann and W.-D. Stohrer, J. Am. Chem. Soc., 93? 6941 (1971).

92. M. J. S. Dewar and W. W. Schoeller, J. Am. Chem. Soc. , 93, l48l (1971); M. J. S. Dewar and D. H. Lo, ibid., 93, 7201 (1971).

93. M. J. Goldstein and R. Hoffmann, J. Am. Chem. Soc., 93, 6193 (1971). ~~

9**. J. C. Barb or ak, S. Chari, and P. v. R. Schleyer, J. Am. Chem. Soc., 93, 5275 (1971).

95. L. A. Paquette, D. R. James, and G. H. Bimberg, Chem. Commun., 722 (1974).

96. D. R. James, G. H. Birnberg, and L. A. Paquette, J. Am. Chem. Soc., 9 6, 7^65 (1974).

97. A. K. Cheng, F. A. L. Anet, J. Mioduski, and J. Meinwald, J. Am. Chem. Soc., §£, 2887 (197*0.

98. H. K. Hall, Jr., and J. H. Baldt, J. Am. Chem. Soc., 93, 140 (1971).

99. C. H. Bushweller, M. Sharpe, and S. J. Weininger, Tetrahedron Lett. , *<•53 (1970).

100. E. Ciganek, J. Am. Chem. Soc., 8 7, 1149 (1965). 281

101. E. Ciganek, J. Am. Chem. Soc., 8 9, 1454 (1967)*

102. C. J. Fritchie, Jr., Acta Crystallogr. , 20, 27 (1966).

103. H. E. Zimmerman and G. L. Grunewald, J. Am. Chem. Soc., 8 8, 183 (19 66); H. E. Zimmerman, R. ¥. Binckley, R. S. Givens, G. L. Grunewald, and M. A. Sherwin, ibid., 91? 3316 (1969); H. E. Zimmerman and H. Iwamura, ibid., 9 0, 47*63 (1968).

104. J. Meinwald and D. Schmidt, J. Am. Chem. Soc., 91, 5877 (19^9); H. E. Zimmerman, J. D. Robbins, and J. Schantl, ibid., 91? 5878 (1969).

105. L. A. Paquette, J. Am. Chem. Soc., 92, 5766 (1970); R. Askani, Tetrahedron Lett., 3349 (1970); 447~Cl97l); R. M. Moriarty, C. L. Yeh, and N. Ishibe, J. Am. Chem. Soc., 93? 3085 (1971); R. Askani, I. Gurang, and W. Schwertfeger, Tetrahedron Lett., 1315 (1975).

106. R. Malherbe, Helv. Chim. Acta, 5 6, 2845 (1974).

107. R. Graf, DBP 1,218,448, Farbwerke Hoechst AG (1963); K. Matterstoek and G. Lohaus, DAS 1,253?704, Farbwerke Hoechst AG (1964).

108. T. J. Katz and K. C. Nicolaou, J. Am. Chem. Soc., 9 6, 1948 (1974).

109. M. A. Beno and G. G. Christoph, manuscript in preparation; full experimental details and description of the structure will be reported elsewhere.

110. G. Germain, P. Main, and M. M. Woolfson, Acta Crystallogr., 27, 368 (1971).

111. L. A. Paquette, R. E. Wingard, Jr., and R. K. Russell, J. Am. Chem. Soc., 94, 4739 (1972); R. E. Wingard, Jr., R. K. Russell, and L. A. Paquette, ibid., 9 6, 7474 (1974).

112. L. A. Paquette, R. K. Russell, and R. L. Burson, J. Am. Chem. Soc., 97, 6124 (1975).

113. E. Wenkert, E. W. Hagaman, L. A. Paquette, R. E. Wingard, Jr., and R. K. Russell, Chem. Commun. , 135 (1973)*

114. M. Karplus, J. Chem. Fhys., 30, H (1959); J. Am. Chem. Soc., 8 5, 2870 (1963).

115. C. 0. Bender and S. S. Shugarman, Chem. Commun., 934 (1974). 282

116. See, for example: G. Stork and W. N. White, J. Am. Chem. Soc., 7 8, 4609 (1956); C. W. Jefford, S. N. Mahajan, and J. Gunsher, Tetrahedron, 24, 2921 (1968).

117. Y. C. Wang and S. H. Bauer, J. Am. Chem. Soc. , 94, 5651 (1972).

118. E. J. Moriconi and C. P. Dutta, J. Org. Chem., 35 , 2443 (1970).

119. W. E. Volz, L. A. Paquette, R. J. Rogido, and T. J. Barton, Chem. Ind. (London), 771 (1974).

120. E. J. Moriconi, J. F. Kelly, and R. A. Salomone, J. Org. Chem. , 33 3 3448 (1968).

121. P. G. Gassman and K. T. Mansfield, Org. Syn., 49, 1 (1969).

122. J. C. Jagt and A. M. van Leusen, J. Org. Chem., 35 , 2043 (1970).

123. K. Griesbaum and P. E. Butler, Angew. Chem., 79, 467 (1967); Angew. Chem. Intern. Ed. Engl., 6 , 444 (1967).

124. R. J. Rogido, Ph.D. Thesis, Iowa State University, 1975*

125. D. H. Aue, H. Iwahashi, and D. F. Shellhammer, Tetrahedron Lett. , 3719 (1973).

126. K. B. Wiberg and B. A. Hess, Jr., J. Am. Chem. Soc. , 8 9, 3015 (1967).

127. P. G. Gassman and G. D. Richmond, Chem. Commun., 1631 (1968); J. Am. Chem. Soc., 9 0, 5637 (1968); ibid., 92, 2090 (1970).

128. M. Pomerantz, G. W. Gruber, and R. N. Wilke, J. Am. Chem. Soc., 9 0, 5040 (19 68).

129 • L. Skatteb^l, Tetrahedron Lett. , 2361 (1970); W. R. Moore, K. G. Taylor, P. Muller, S. S. Hall, and Z. L. F. Gaibel, ibid., 2365 (1970).

130. W. v. E. Doering and J. F. Coburn, Jr., Tetrahedron Lett., 991 (1965).

131. R. T. LaLonde, J. Am. Chem. Soc. , 8 7, 4217 (1965).

132. R. Y. Levina, V. N. Kostin, and T. K. Ustynyuk, Zh. Obsch. Khim., 30 , 359 (I9 6 0).

135* R. J. Ouellette, A. South, Jr., and D. L. Shaw, J. Am. Chem. Soc., 8 7, 2602 (1965). 283

134. R.Criegee and A.Rimmelin, Chem. Ber., 90, 4l4 (1957).

135* R« T. LaLonde and L. S. Forney, J. Am. Chem. Soc. , 8 5, 3767 (1963).

136. For a review: P. G. Gassman, Accounts Chem. Res., 4, 128 (1971)*

137. L. A. Paquette, R. P. Henzel, and S. E. Wilson, J. Am. Chem. Soc., 9k, 7780 (1972).

138 . M. R. Rifi, J. Am. Chem. Soc. , 8 9, 4442 (1967).

139* K. B. Wiherg, K. C. Bishop, III, and R. B. Davidson, Tetrahedron Lett., 3169 (1973).

140. M. Saunders and J. Rosenfeld, J. Am. Chem. Soc., 92, 2548 (1970).

141. M. Saunders, P. Vogel, E. L. Hagen, and J. Rosenfeld, Accounts Chem. Res., 6 , 53 (1973)*

142. K. B. Wiherg and G. J. Burgmaier, J. Am. Chem. Soc. , 94, 7396 (1972).

143. J. Meinwaid and F. Uno, J. Am. Chem. Soc., 90, 800 (1968).

144. J. Meinwaid and H. Tsuruta, J. Am. Chem. Soc., 8 9, 2611 (1967).

145. J. Meinwaid and D. Schmidt, J. Am. Chem. Soc., 91, 5877 (19^9)•

146. J. Meinwaid and H. Tsuruta, J. Am. Chem. Soc. , 91, 5877 (1969)*

147. H. E. Zimmerman, J. D. Robbins, and J. Schantl, J. Am. Chem. Soc., 91, 5878 (1969).

148. K. Grychtol, H. Musso, and J. F. M. 0th, Chem. Ber. , 105, 1798 (1972).

149. J. Hatem and S. Waegell, Tetrahedron Lett. , 2069 (1971); A. N. Vereshehagin, S. G. Vul'fson, N. I. Gubkina, and B. A. Arbuzov, Iz. Akad. Hauk. SSSR Ser. Khim. , 2467 (1970).

150. D. Joulain and F. Rouessac, Compt. rend., 273, C 561 (1971).

151. For a review: E. Heilbronner and J. P. Maier, ''Electron Spectroscopy: Theory, Techniques, and Applications,'' in press, Academic Press, London; H. Bock and P. 0. Mollere, J. Chem. Ed., 51, 506 (1974); H. Bock and B. G. Ramsey, Angew. Chem., 85, 773 Xi§73); Angew. Chem. Intern. Ed. Engl., 12, 75^- (1975); S. D. Worley, Chem. Rev. , 71, 295 (1971); A. D. .Baker, Acc. Chem. Res. , 28b

IT (1970)', D. W. Turner, C. Baker, A. D. Baker, and C. R. Brundle, '’Molecular Photoelectron Spectroscopy,'’ Wiley- Interscience, London (1970).

152. E. Heilbronner and A. Schmelzer, Helv. Chim. Acta, 5 8, 936 (1975).

155. T. Koopmans, Physika, 1, lOit- (193^); W. G. Richards, J. Mass. Spectrom. Ion Phys., 2, ^19 (1969).

15^. C. Baker and D. W. Turner, Chem. Commun., 480 (1969).

155. E. Heilbronner, Isr. J. Chem., 10, 1^3 (1972); M. J. Goldstein, S. Natowsky, E. Heilbronner, and V. Homung, Helv. Chim. Acta, 56, 29^ (1973).

156. P. Bischof, R. Gleiter, E. Heilbronner, V. Hornung, and G. Schroder, Helv. Chim. Acta, 53, 16^5 (1970).

157. G. Bieri, M. J. Goldstein, and E. Heilbronner, Z. Physik. Chemie (Heue Folge), in press (1976).

158. K. B. Wiberg, Tetrahedron, 2k, IO83 (1968); L. Radom, J. A. Pople, V. Buss, and F. v. R. Schleyer, J. Am. Chem. Soc., 92, 638 O (1970). -

159. R. Hoffman and R. B. Davidson, J. Am. Chem. Soc., 93, 5699 (1971).

160. R. Gleiter and T. Kobayashi, Helv. Chim. Acta, 5_b, 108l (1971).

161. W. L. Jorgensen and W. T. Borden, J. Am. Chem. Soc., 95, 66^9 (1973); W. L. Jorgensen, ibid., 97, 3082 (1975).

162. P. Bischof, R. Gleiter, M. J. Kukla, and L. A. Paquette, J. Electron Spectroscopy, b, 177 (197^)•

163 . P. Bischof, R. Gleiter, A. de Meijere, and L.-U. Meyer, Helv. Chim. Acta, 57, 1519 (197^).

164. P. Bischof, E. Haselbach, and E. Heilbronner, Angew. Chem. , 82, 952 (1970); Angew. Chem. Intern. Ed. Engl., 9, 953 (1970).

165. ¥. L. Jorgensen and L. Salem, ’’The Organic Chemist's Book of Orbitals,'' Academic Press, New York (1973).

166. L. Salem and J. S. Wright, J. Am. Chem. Soc., 9 1, 59^7 (1969); J. Chem. Soc. D, 1370 (1969); L. Salem, Chem. Brit. , 5, ^ 9 (1969). 285 l6j. A. Almenningen, 0. Bastiansen, and P. N. Skancke, Acta Chem. Scand,, 15, Til (1 9 6 1).

168. L. C. Snyder and S. Meiboom, J. Chem, Phys., 52? 3857 (1970); J. D. Dunitz and V. Schomaker, ibid., 20, 1703 (1952).

169. J. S. Wright and L. Salem, J. Am. Chem. Soc. , 94? 322 (1972).

170. W. J. Hehre and J. A. Pople, J. Am. Chem. Soc. , 97? 69k! (1975).

171. E. Gleiter, E. Heilbronner, M. Hekman, and H.-D. Martin, Chem. Ber., 106, 28 (1975).

172. P. Bischof, J. A. Hashmall, E. Heilbronner, and V. Hornung, Helv. Chim. Acta, 52, 1745 (1969)*

173. P. Bischof and E. Heilbronner, Helv. Chim. Acta, 53? 1677 (1970).

174. F. Brogli, E. Giovannini, E. Heilbronner, and R. Schurter, Chem. Ber., 106 , 961 (1973).

175* H. Oberhammer and S. H. Bauer, J. Am. Chem. Soc., 91? 10 (19^9).

176. R. Gleiter, E. Heilbronner, M. Kehman, and H. -B. Martin, Chem. Ber. , 106 , 28. (1973).

177* C. R. Brundle and M. B. Robin, J. Am. Chem. Soc. , 92? 5550 (1970); C. R. Brundle, M. B. Robin, N. A. Kuebler, and M. Basch, ibid., 9k, 1451 (1972).

178. M. J. S. Dewar and S. D. Worley, J. Chem. Phys., 49, 2454 (1968).

179* R* Sustmann and R. Schubert, Tetrahedron Lett., 2739 (1972).

180. M. Beez, G. Bieri, H. Bock, and E. Heilbronner, Helv. Chim. Acta, 5 6, 1028 (1973).

181. P. Bischof and E. Heilbronner, Helv. Chim. Acta, 55? 1677 (1970)*

182. P. Bischof, R. Gleiter, and E. Heilbronner, Helv. Chim. Acta, 53, 1425 (1970).

183 . J. F. Chiang and S. H. Bauer, J. Am. Chem. Soc., 8 8, 420 (1966).

184. S. S. Butcher, J. Chem. Phys., 42, I83 O (1965). 286

185. E. Heilbronner, Pure Appl. Chem., 7, 9 (1971); R« Gleiter, E. Heilbronner, L. A. Paquette, G. L. Thompson, and R. E. Wingard, Jr., Tetrahedron, 29? 565 (1973); M. T. Reetz, R. W. Hoffman, W. Schafer, and A. Schweig, Angew. Chem. , 8 5, 45 (1973); Angew. Chem. Intern. Ed. Engl. , 12, 81 (1973); W. Schafer, H. Schmidt, A. Schweig, R. W. Hoffmann, and H. Kurz, Tetrahedron Lett., 1953 (1974); P. Bischof, D. Bosse, R. Gleiter, M. J. Kukla, A. de Meijere, and L. A. Paquette, Chem. Ber., IO8 , 1218 (1975).

186. P. Asmus and M. KLessinger, Tetrahedron, 30, 24'77 (1974).

187. L. A. Paquette, D. C. Liotta, C. C. Liao, T. G. Wallis, N. Eickman, J. Clardy, and R. Gleiter, J. Am. Chem. Soc., in press.

188. C. Batich, E. Heilbronner, E. Rommel, M. F. Semmelhack, and J. S. Foos, J. Am. Chem. Soc., 9 6, 7662 (1974).

189. R. Gleiter, E. Heilbronner, and A. deMeijere, Helv. Chim. Acta, 54, 1029 (1971).

190. E. Heilbronner, R. Gleiter, H. Hopf, V. Hornung, and A. de Meijere, Helv. Chim. Acta, 54, 783 (1971)*

191. Leading references: W. J. Hehre and J. A. Pople, J. Am. Chem. Soc. , 97, 6941 (1975); J- S. Wright and L. Salem, ibid., 322 (1972). ro • to J. S. Wright and L. Salem, Chem. Commun., 1370 (1969); J. L. Nelson and A. A. Frost, J. Am. Chem. Soc., 94, 3727 (1972); A. Y. Meyer and R. Pasternak, Theoret. Chim. Acta, 33 , 215 (1974).

193. K. Kovacevic and Z. B. Maksic, J. Org. Chem. , 39, 540 (1974).

194. D. R. Whitman and J. F. Chiang, J. Am. Chem. Soc., 94, 1126 (1972).

195. J. F. Chiang, J. Am. Chem. Soc., 93, 5044 (1971).

196. C. S. Wang and M. D. Harmony, J. Am. Chem. Soc., 9 8, 1108 (1976).

197. G. Dallinga and L. H. Toneman, Rec. Trav. Chim., 8 6, 171 (19^7)•

198. G. Dallinga and L. H. Toneman, Rec. Trav. Chim., 8 8, 185 (1969)*

199. N. L. Allinger and L. A. Tushaus, J. Org. Chem., 30, 1945 (1965)-

200. G. A. Russell, P. R. Whittle, and R. G. Keske, J. Am. Chem. Soc. , 93, 146? (1971).

201. L. A. Paquette, J. Ward, R. Boggs, and W. B. Farnham, J. Am. Chem. Soc. , 97, H O I (1975). 202. J. H. Chapman, J. Elks, G. H. Phillips, and L. J. Wyman, J. Chem. Soc., 4344 (1956). 287

205. R. L. Shriner and F. W. Newmann, ''Organic Synthesis,'' Coll. Vol. Ill, Wiley, New York, N.Y. (I955),p 75*

204. R. E. Ireland, D. C. Muchmore, and U. Hengartner, J. Am. Chem. Soc., 94, 5098 (1972).

205. L. A. Paquette, W. B. Farnham, and S. V. Ley, J. Am. Chem, Soc., 97? 7275 (1975).

206. R. K. Crossland and K. L. Servis, J. Org. Chem. , 55, 5195 (1970).

207. D. B. Denny, N. Gershman, and J. Giacin, J. Org. Chem., 51, 2855 (1966); T. Mukaiyama, T. Kumamoto, and T. Nagaoka, Tetrahedron Lett., 5563 (1966).

208. P. A. Chopard, V. M. Clark, R. F. Hudson, and A. J. Kirby, Tetrahedron, 21, 1961 (1965).

209. I. J. Borowitz, M. Anschel, and S. Firstenberg, J. Org. Chem., 5 2, 1725 (1967).

210. T. Wakamatsu, K. Akasaka, and Y. Ban, Tetrahedron Lett., 3879 (1974).

211. L. A. Paquette, R. H. Meisinger, and R. E. Wingard, Jr., J. Am. Chem. Soc., 95,, 2250 (1975).

212. J. C. Carnahan, Jr., and W. D. Closson,Tetrahedron Lett. , 3447 (1972).

215. J. Cleophax, J. Hildesheim, and S. D. Gero, Bull. Soc. Chim. Fr., 4111 (1967).

214. J. A. Marshall and M. E. Lewellyn, Syn. Commun., 5, 293 (1975).

215. E. N. Walsh and A. 0. F. Toy, Inorg. Syn., 7, 69 (1963).

216. E. J. Corey and R. A. E. Winter, J. Am. Chem. Soc., 8 5, 2677 (1963).

217. E. M. Burgess, H. R. Penton, Jr., and E. A. Taylor, J. Org. Chem. , 5 8, 26 (1975); L. A. Paquette, I. Itoh, and W. B. Farnham, J. Am. Chem. Soc., 97? 7280 (1975).

218. C. A. Cupas and W. S. Roach, J. Org. Chem. , 3 4 , 742 (1969); W. Reusch and D. W. Frey, Tetrahedron Lett., 5193 (1967); c* Chang and N. F. Wood, Steroids, 4, 55 (1964); C. H. Snyder and A. R. Soto, J. Org. Chem. , 29, 7^2 (1964). 288

219. H. Oediger and Fr. Moller, Angew. Chem. , 79, 53 (1967); Angew. Chem. Intern. Ed. Engl., 6 , 76 (1967).

220. C. Mercier, P. Soucy, W. Rosen, and P. Deslongchamps, Syn. Commun., 3» l6l (1973); 0. H. Posner, R. J. Johnson, and M. J. Whalen, Chem. Commun., 281 (1972).

221. D. A. Seeley and J. McElwee, J. Org. Chem., 3 8 , l68l (1973).

222. K. B. Sharpless and R. F. Lauer, J. Am. Chem. Soc., 955 2697 (1973).

225. W. G. Dauben, M. E. Lorber, N. D. Vietmeyer, R. H. Shapiro, J. H. Duncan, and K. Tomer, J. Am. Chem. Soc., 90, kj62 (1968).

224. L. F. Fieser, J. Chem. Ed. , 51? 291 (1954).

225. L. A. Paquette and J. H. Barrett, Org. Syn., 49, 62 (1969).

226. L. Velluz, Subst. nat. syn., 7, 31 (1953).

227. E. Vogel, ¥. A. Boll, and E. Lohmar, Angew. Chem., 8 3 , 403 (1971); Angew. Chem. Intern. Ed. Engl., 10, 399 (1971).

228. E. Vogel and F. Hogrefe, Angew. Chem., 8 6, 779 (1974); Angew. Chem. Intern. Ed. Engl., 15, 735 (1974).

229. J. A. Elix, M. V. Sargent, and F. Sondheimer, Chem. Commun., 508 (1966).

230. F. N. Stepanov and V. D. Sukhoverkhov, Angew. Chem., 79, 860 (1967); Angew. Chem. Intern. Ed. Engl., 6 , 864 (1967).

231. L. F. Johnson and ¥. C. Jankowski, " Carbon-13 NMR Spectra, ' 1 John ¥iley and Sons, Inc., New York, N.Y. (1972).

232. R. G. Parker and J. D. Roberts, J. Am, Chem. Soc. , 92, 743 (1970).

233. G. C. Levy and G. L. Nelson, ’'Carbon-13 Nuclear Magnetic Resonance for Organic Chemists,'' John Wiley and Sons, Inc., New York, N.Y. (1972).

234. J. B. Stothers, ''Carbon-13 NMR Spectroscopy,'' Academic Press, New York, N.Y. (1972).

235. S. Castellano and A. A. Bothner-By, J. Chem. Phys., 4l, 3863 (1964).

236. P. Crews, Chem. Commun. , 583 (1971). 289

237* C. Ganter and J. D.Roberts, J. Am. Chem. Soc., 88, 741 (1966).

238 . H. Gunther and H. H. Hinrichs, Justus Liebigs Ann, Chem., 706, 1 (1967).

239. H. Gunther and H. H. Hinrichs, Tetrahedron Lett., 787 (1966).

240. P. Crews, J. Am. Chem. Soc., 95, 636 (1973).

241. M. A. Cooper, D. D. Elleman, C. D. Pearce, and S. C. Manatt, J. Chem. Phys., 53., 2343 (1970).

242. E. A. Braude, Chem. Ind. , 1557 (195*0; C.-Y. Chen, R. J. W. LeFevre, and K. M. S. .Sundarem, J. Chem. Soc. , 555 (1965 H G. Favini, F. Zuccarello, and G. Buemi, J. Mol. Struct., 3> 585 (1969).

243. V. W. Laurie, J. Chem. Phys., 24, 635 (1956); L. H. Scharpen and V. W. Laurie, ibid., 437 2?6* (1965).

244. A. Ac Bothner-By and E. Moser, J. Am. Chem. Soc., 90, 2347 (1968).

245. M. Karplus, J. Am. Chem. Soc., 8 5, 2870 (1963).

246. A. deMeijere and L.-U. Meyer, Angew. Chem., 8 5, 908 (1973)» Angew. Chem. Intern. Ed. Engl., 12, 858 ~(l975T-

247. R. D. Miller, M. Schneider, and D. L. Dolce, J. Am. Chem. Soc., 95., 8468 (1973).

248. C. F. Wilcox, Jr., and R. R. Craig, J. Am. Chem. Soc., 83 , 4258 (1961).

249. M. F. Semmelhack, J. S. Foos, and S. Katz, J. Am. Chem. Soc., 94, 8637 (1972).

250. E. Pesch and S. L. Friess, J. Am. Chem. Soc., 72, 5756 (1950).

251. E. Vogel, W. A. Boll, and H. Gunther, Tetrahedron Lett., 609 (1965).

252. E. Vogel, W. Wiedemann, H. Kiefer, and W. F. Harrison, Tetrahedron Lett., 673 (1963).

253* E. Vogel, W. Meckel, and W. Grimme, Angew. Chem., 7 6, 786 (1964); Angew. Chem. Intern. Ed. Engl., 3., 643 (1964).

254. E. Vogel, W. Maier, and J. Eimer, Tetrahedron Lett., 655 (1966). 290

255- H. Booker, L. K. Evans, and A. E. Gillam, J. Chem. Soc., 1453 (1940); V. Henri and L. W. Pickett, J. Chem. Phys., 7, 439 (1939)*

256. W. J. Bailey and W. B. Lawson, J. Am. Chem. Soc., 79, 1444 (1957).

257. A. C. Cope and L. L. Estes, J. Am. Chem. Soc. , 72, 1128 (1950).

258. G. Scheibe, Chem. Ber., 59, 1521 (1926).

259. R. M. Silverstein and G. G. Bassler, ''Spectrome trie Identifica­ tion of Organic Compounds,'' John Wiley and Sons, Inc., New York, N.Y. (1967), P 163 . 260. J. R. Dyer, '’Applications of Absorption Spectroscopy of Organic Compounds,’’ Prentice-Hall, Inc., Englewood Cliffs, N.J. (1965), P 17.

261. K. Alder and H, A. Dortmann, Chem. Ber., 8 7, 1492 (195*0; A. C. Cope, A. C. Haven, Jr., F. L. Ramp, and E. R. Trumbull, J. Am. Chem. Soc., 74, 4867 (1952).

262. A. C. Cope, P. T. Moore, and W. R. Moore, J. Am. Chem. Soc., 80, 5505 (1958).

265. R. Burton, L. Pratt, and G. Wilkinson, J. Chem. Soc., 59*1- (l96l).

264. P. Dutta, T. D. Goldfarb, and R. S. Boikess, J. Am. Chem. Soc., 91, 5429 (1969).

265. M. Maynadier, A.. Casadevall, and E. Casadevall, C. R. Acad. Sci. Paris, Ser. C, 269, 506 (1969); C. Largeau and E. Casadevall, ibid., 267, 1755 (19687.

266. W. G. Dauben and R. L. Cargill, J. Org. Chem., 27, 1910 (1962).

267. R. Hoffmann, J. Chem. Phys., 59j> 1597 (1 965); R. Hoffmann and W. N. Lipscomb, ibid., 36,~2179, 5489 (1962); 37, 2872 (1962).

268. R. C. Bingham, M. J. S. Dewar, and D. H. Lo, J. Am. Chem. Soc., 9T, 1285 (1975).

269. J. S. Wright and L. Salem, Chem. Commun., 1370 (1969); C. A. Coulson and W. E. Moffitt, Phil. Mag., 40, 1 (1949); D. Peters, Tetrahedron, 39,, 1539 (1963); L. Klasinc, Z. Maksic, and M. Randic, J. Chem. Soc. A, 755 (1966).

270. R. Fletcher and M. J. D. Powell, Computer Journ., 6 , 163 (1963). 291

271. R. C. Bingham, M. J. S. Dewar, and D. H. Lo, J. Am. Chem. Soc., §7? 129^- (1975).

272. D. A. Dows and N. Rich, J. Chem. Phys., kj, 333 (1967).

273. F. Chiang and S. H. Bauer, Trans. Faraday Soc., 6 b, 22bj (1968).

27b. R. Askani, R. Gleiter, E. Heilhronner, V. Homung, and H. Musso, Tetrahedron Lett. , l*l*6l (1971).

275* K. B. Wiberg, Tetrahedron, 2k, IO83 (1968).

276. W. J. Hehre, J. Am. Chem. Soc., 9k, 6592 (1972).

277* M. Shanshal, Z. Naturforsch. , 27a, 1665 (1972).

278. R. Breslow, in ''Molecular Rearrangements,'' P. de Mayo. Ed., Interscience, New York, N.Y. (1 963), p 233 and references therein; C. F. Wilcox and M. E. Mesirov, J. Am. Chem. Soc., 8b, 2757 (1962); R. C. Hahn, T. F. Carbin, and H. Shechter, ibid., 9 0, 31+04 (19 68); W. G. Dauben, J. L. Chitwood, and K. V. Scherer, Jr., ibid. , 90? 10ll* (1968); J. J. Gajewski, R. L. Lyle, and R. P. Gajewski, Tetrahedron Lett. , II89 (1970); P. v. R. Schleyer, private communication.

279* P. E. Eaton and K. Nyi, J. Am. Chem. Soc., 93? 2786 (1971).

280. K. B. Wiberg, K. C. Bishop, III, and R. B. Davidson, Tetrahedron Lett., 3169 (1973).

281. R. Darms, T. Threlfall, M. Pesano, and A. Eschenmoser, Helv. Chim. Acta, b6, 2893 (1983).

282. G. E. Gream, L. R. Smith, and J. Meinwald, J. Org. Chem., 39? 3lf6l (1974).

283 . H. Prinzbach, H. Babsch, and H. Fritz, Tetrahedron Lett., 2129 (1976); H. Babsch, H. Fritz, and H. Prinzbach, ibid., 4677 (1975).

28b. P. Th. van Duijnen, P. van der Ploeg, H. Hogeveen, and W. F. J. Huudeman, Tetrahedron Lett. , 573 (1975)? W. J. Jorgensen and W. T. Borden, ibid., 223 (1975)•

285. E. J. Corey, N. M. Weinshenker, T. K. Schaaf, and W. Huber, J. Am. Chem. Soc., 91? 5^75 (1969)? D* A* Evans, W. L. Scott, and L. K. Truesdale, Tetrahedron Lett., 121 (1972); J. Damiano, S. Geribaldi, G. Torri, and M. Azzaro, ibid., 2301 (1973); R* P» Gragson and R. N. Mirrington, Chem. Commun., 598 (1973)- 292

286. S. Kaarsemaker and J. Coops, Rec. Trav. Chim., 71, 26l (1952); N. L. Allinger, J. A. Hirsch, M. A. Miller, I. J. Tyminski and F. A. Van-Catledge, J. Am. Chem. Soc. , 9 0, 1199 (1968); S. W. Benson, F. R. Cruickshank, D. M. Golden, G. R. Haugen, H. E. O'Weal, A. S. Rogers, R. Shaw, and R. Walsh, Chem. Rev., 69, 279 (1969); P. v. R. Schleyer, J. E. Williams, and K. R. Blanchard, J. Am. Chem. Soc., 92, 2377 (1970); and E. M. Engler, J. D. Andose, and P. v. R. Schleyer, ibid., 95, 8005 (1973).

287. H. Kwart and J. Ford, Jr., J. Org. Chem., 24, 2060 (1959).

288. J. Dale, J. Chem. Soc., 910 (1961).

289. D. E. Applequist, P. A. Gebauer, D. E. Gwynn, and L. H. O'Connor, J. Am. Chem. Soc., 94, 4272 (1972).

290. P. Y. Johnson, D. Jacobs, and D. J. Kerkman, J. Org. Chem. , 40, 2710 (1975).

291. G. Stork and P. F. Hudrlik, J. Am. Chem. Soc., 90, 4462, 4464 (1968).

292. G. R. Wenzinger and P. Beng, Tetrahedron Lett., 727 (1976).

293- K. B. Wiberg and G. R. Wenzinger, J. Org. Chem., 30, 2278 (1965).

294. E. P. Kohler, J. Am. Chem. Soc., 6l, 1057 (1939).

295. N. B. Chapman, S. Sotheeswaren and K. J. Toyne, Chem. Commun., 214 (1965).

296. C. D. Nenitzeseu, M. Avram, J. J. Pogany, G. P. Mateescu, and M. Farcasiu, Acad. Rep. Populare Romine, Studii Cereetari Chem., 11; 7 (1963).

297. K. W. Rosenmund and F. Zetzsche, Chem. Ber. , 54, ^-36 (1921). 298. K. Takeda, S. Nagakura, and K. Kitahonoki, Pharm. Bull. (Japan), I , 135 (1953).

299- L. Friedman, R. L. Little, and W. R. Reichle, Org. Syn., 40, 93 (I9 6 0).

300. A second preparation of 46, 47, 5 6, and 57 has appeared subse­ quently: K. Kitahonoki, K. Sakurawi, K. Tori, and M. Ueyama, Tetrahedron Lett. , 263 (1976).

301. R. H. Wiley and N. R. Smith, Org. Syn., Coll. Vol. IV, 201 (1963). 293

302. K. Griesbaum, W. Naegele, and G. G. Wanless, J. Am. Chem. Soc., 87; 3151 (1965).

303 . J- Alexander and G. S. K. Rao, J. Chem. Ed., 47, 277 (1970).

304. U. Schrapler and K. Ruhlman, Chem. Ber., 9 7, 1383 (1964).

305 . T. Wakamatsu, K. Akasaka, and Y. Ban, Tetrahedron Lett., 3^79 (1974).

306. E. L. Martin, Organic Reactions, I, 155? John Wiley and Sons, Inc., New York, N.Y. (1957). ~

307. E. LeGoff, J. Org. Chem., 29, 2048 (196k).