69-4845 BEGLAND, Robert Walter, 1941

This dissertation has been microfilmed exactly as received 69-4845

BEGLAND, Robert Walter, 1941- PARTICIPATION AND STERIC EFFECTS OF NEIGHBORING DIVALENT OXYGEN.

The Ohio State University, Ph.D., 1968 Chemistry, organic

University Microfilms, Inc., Ann Arbor, Michigan PARTICIPATION AND STEPJ.C EFFECTS OF

NEIGHBORING DIVALENT OXYGEN

DISSERTATION

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

Robert Walter Beglnnd, B.S., M.S.

* it * * it it *

The Ohio State University 1968

Approved by

Adviser \J Departmeirt of Chemistry DEDICATION

This dissertation is dedicated to my w ife June, ny con Michael Brian and my son Douglas Samuel.

i i ACKNOWLEDGMENT

The author wishes to express his appreciation to Dr. L. A. Paquette for his guidance, encouragement and considerable time spent in the course of this research.

The author also wishes to express his appreciation to his parents, Mr. and Mrs.

Walter C. Bcgland, for their assistance and encouragement throughout his college training, and to lir. and Mrs. Lester Berndt.

i i i VITA

July 23, 1941 Born - Oak Park, Illinois

1963 B.S. in Education, The Ohio State University Columbus, Ohio

1965 M.S., The Ohio State University Columbus, Ohio

1968 Ph.D., The Ohio State University Columbus, Ohio

PUBLICATIONS

1. Transannular Participation of Ether Oxygen in the Hydrolysis of a Mesocyclic Dienamine, J. Am. Chen. Soc. . 87, 3784 (1965).

2. Stabilized Derivatives of cis,cis,cis-1,3.5-Cyclodecatricne Keto-Enol Tautomerism in 2,3-DicarboiEeth.oxy-cis,c is -3,5~ cyclodecadienones and c.is-3-Cycloalkenones. J. Am. Chem. Soc. . 88, 4685 (1966).

3. Oxocane. Synthesis and Conformational Isomerization. J. Org. Chem. . 32, 2723 (1967).

4. The Interception of Homoallylic Cations by Neighboring Hydroxyl. J. An. Chem. Soc. . 90, (1968).

5. Acid-Promoted Rearrangements Involving Transannular Ether Oxygen Participation. J. Am. Chem. Soc., 90. (1968).

FIELDS OF STUDY

Major Field: Organic Chemistry

iv TABLE OF CONTENTS

Page

D e d i c a t i o n ...... 11

Acknowledgment ...... I l l

Vita ...... iv

List of Tables ...... vi

List of Illustrations ...... v ii

Introduction ...... 1

Part I: Acid-Promoted Rearrangements Involving Transannular Ether Oxygen Participation . . . 4

Part II: Oxocane: Synthesis and Conformational Isomerization ...... 15

Part III: Acetolycis of Oxocan-3-yl £ 1^1 3 ,4 , 7 , 8 - Tetrahydro“2H-oxocin-3-yl Brosylates ...... 18

Part IV: The Interception of Homoallylic Cations by Neighboring Hydroxyl ...... 42

Part V: Stabilized Derivatives of cls.cis.cis-1,3,5- Cyclodecatriene. Keto-Enol Tautomerism in 2.3-Dicarbomethoxy-cis,cis-3,5-Cyclodecadienones and cis-3-Cycloalkenones ...... 57

Part VI: Novel Products from the Reaction of 1-Diraet.hylamino-l,3-Cyclooctadiene with Methanesulfonyl Chloride ...... 84

Experimental ...... 95

R eferences ...... 144

v LIST OF TABLES

T able Page

1. Acetolysis Rate Data for 27. and 4 3 ...... 24

2. Relative Capabilities of Several Heteroatomic Groups for Anchimeric Assistance ...... 56

3. Ultraviolet Absorption Data for Soiae Cyclic Polyolefins ...... 66

4. Infrared Absorptions of Some 2,3-Dicarboraethoxy- cis, cis-3,5-cyclodecadicnones and cis-3" cyclonlkenones ...... 71

5. Ultraviolet Absorption Data of Some 2,3- Dicarboraethoxy-cis,cis-3,5-cyclodecadienones and cis-3-cycloalkenoncs ...... 74

6. Chemical Shift Values (6 units) and Enol Contents of Some 2,3-Dicarborr.ethoxy-cis, cis- 3,5-cyclodecadienones and cis-3-cycloalkcnones . . 76

7. Sulfones Formed Under Various Reaction C onditions ...... 86

8„ Example of a Solvolytic Rate Calculation ...... 120

9. Solvolytic R.r.te Data Experimentally Determined for 27 and 4S ••••••••••»••••«••• 121

v i LIST OF ILLUSTRATIONS

Figure Page

1. Partial NMR Spectra of 2,3-Dicerbonethoxy- 3- c is -cy c1ono neno ne ( 1 3 3 ) ...... 80

v i i INTRODUCTION

Leonard, in his studies of the transannular reactions

of medium-sized heterocyclic molecules, was able to detect

transa\uiular N-C^q and interactions in the spectra of 1 2 ketones such as 1. and 2 In addition, such ketones could

be converted to their respective bicyclic salts upon treatment with strong acids. By contr*ast, l-oxacyclooctan-5-one <3) CH

HCIO, -v

l OH

h c ip. CIO/ © 2 exhibited no such transannular effects in its spectra, a result

which was attributed to the relatively low order of electron-

donating ability of oxygen.^

Despite the marginal nucleophilicity of ether oxygen, many

facts are available to suggest that it should be entirely capable

of transannular bonding in carbontum ion reactions. With particular

1 regard to aliphatic and alicyclic ethers, the anomalously rapid rates of acetolysis of 4-methoxybutyl and 5-methoxype-ntyl brosylates have suggested that these reactions proceed through cyclic oxonium ions.

The enhanced solvolytic behavior of UJ-(2-tetrahydrofuranyl)-3-propyl brosylate also pointed to anchimeric assistance by ether oxygen and formation of a bicyclic oxoniun ion intermediate/* Additional evidence

5 of such neighboring group participation has recently been reviewed.

The work described herein was undertaken to study both the transannular reactions and neighboring group participation of divalent oxygen. Also considered are the steric effects brought about by the replacement of a methylene group by an oxygen atom in medium-sized rings.

Because the present investigation has dealt with a number of different problems relating to the effects of divalent oxygen, the various aspects of this work are divided into the following sections:

Part I . Acid-promoted rearrangements involving transannular ether oxygen participation are presented. The attempted synthesis of 3,4,

7.8-tetrahydro-211-oxocin and the problems which arose from this work are discussed.

Part I I . The low temperature nmr spectra of oxocane are compared to those of cyclooctane in terms of the strain minimization brought about by replacement of a methylene group by an oxygen atom.

Part III. The rates and products of acetolysis of oxocan-3-yl and

3.4.7.8-tetrahydro-2H-oxocin-3-yl brosylates are presented. These com pounds are compared with their carbocyclic analogs and the dramatic effects of the ether oxygen upon the products obtained are discussed.

Part XV. The silver ion-induced cyclization of selected cis- and trnns-iodohydrlns in which the iodine substituent occupies a homoallylic position has been investigated. Evidence is presented which supports the fact that ring closure proceeds by way of intermediate homoallylic cations which are subject to intramolecular nucleophilic attack by the neighboring hydroxyl group. The ambident electrophilic nature of homoallylic cations is considered.

Part V. Stabilized derivatives of cis. cis. c.is-1.3,5-cyclodecatriene were prepared and the keto-enol-fcau-tomerism in the 2,3-dicarbomethoxy- cis, cis-3.5-cyclodccadienones and cis-3-cycloalkcnoncs was studied.

^ne cnol/keto ratio (K£) is considered in terms of probable stereo chemical and steric factors.

Part VI.o The reaction of l-dimcthylamino-l ,3-cyclooctcdiene with methanesulfonyl chloride gave seven products. The structures of these products and the mechanisms of their formation are discussed. PART I

ACIB-PR01DTED REARRANGEMENTS INVOLVING

TRANSANNULAR ETHER, OXYGEN PARTICIPATION

Introduction

The phenomenon of transannular interaction is perhaps the most remarkable of a number of unique features exhibited by medium-sized rings. The origin of this characteristic is found in the existence of certain conformations wherein opposite sides of the ring are brought into close proximity. The pioneering work in this area, due to Prelog^ end Cope^, has centered predominantly upon transannular hydride shifts.

These studies have shown that when substantial carboniurn ion character develops in reaction intermediates involving mesocycles, transannular rearrangements are very likely to be of great importance. For example, the action of phosphoric acid on 4. results in the formation of 6-methyl- g cyclodecanone by a 1,6-hydride shift.

CH CH.

4 5

A further distinguishing feature of medium-sized rings is their tendency to convert to bicyclic structures to relieve existing non bonded interactions. To illustrate, solvolysis of 4-cycloocten-l-yl brosylate (6)in acetic acid yields acetates of 7 and 8 in addition

4 5

to 4-cyclooctenyl-l-acetate

\X)B:is

■-H -■OH 6 7 8 1 2 Leonard, * in a broadening of such concepts to heterocyclic molecules was able to detect transannular N-C interactions

as discussed earlier. This effective electron-donating ability was

found, hovjever, to be restricted only to those irolecules in which

the interacting groups are diametrically opposed. Thus, in contrast

to the behavior of 1_, treatment of 9_ with acid, does not lead to the

formation of a product with a transannular bond across the ring.^-

Interestingly, l~oxacyclooctan~5~one (2 ), a molecule with diametrically

opposed functional group

exhibited no transannular effects in its spectra, a result which was attributed to ether oxygen's relatively low order of electron

donating ability.^ Ketone 3. would also not form a salt; significantly,

however, reaction with aqueous hydrochloric acid gave 1,7-dichloro-4~

heptanone in a process which may have passed through a bicyclic oxoniura

ion.

Although no participation of the ether oxygen in ketone 3, was observed, there are many examples of neighboring group participation 3~*> by aliphatic and alicyclic inethoxyl groups. Participation by ether oxygen has been shown to increase the rate of solvolysis whan

a five- or six- membered oxonium ion intermediate can be formed;

however, when a three- or four- membered oxonium ion intermediate

3 must intervene, rate enhancement is not generally noted. Methoxyl

group participation in MeO-3^ and MeO-4^ carbonium ion reactions

has been implicated by virtue of products isolated.

Our interest in transannular reactions of ether oxygen arose

from the observation that 5-(3-chloropropyl) dihydro-3(211) -furanone

(13a) was obtained upon hydrochloric acid hydrolysis of 7,8-dihydro- 12 N,N-diraethyl-2U-oxocin~3-amine (12). The formation of 13a can be

accounted for only by a transannular reaction and provides evidence that

ether oxygen, when suitably disposed in a mssocycle, v.’il l profoundly

affect the course of reactions proceeding through carbonium ions.

Results and Discussion

Hofmann elimination of methiodidc H) under carefully controlled

conditions resulted in the formation of 5~dimethylamino-7-oxa-l,3-cyclo-

octadiene (1_1) . Heating of this aroinodiene to 140-150° for a short

period of time resulted in facile intramolecular dienyl 1,5-hydrogen 1 ^ transfer*'' and smooth conversion to dienamine 12. Alternatively, 12

could be prepared directly from JLO in 78% yield. Hydrochloric acid

hydrolysis of 12 gave rise in good yield to 13a. Upon addition of

47% hydrobromic acid to 1J2, there was obtained a 61.57. yield of the

analogous bromide 13b. 7

Scheme I

N(CH

10 11

HX h2o ~

12 13a, X ** Cl

H-

w 14 15

In contrast, unrearranged P,y -unsaturated ketone 14 was obtained in quantitative yield when the hydrolysis was performed in refluxlng

15% fluoboric acid for a short period of time. To remove from consider ation the possibility that such transformations were occuring because of the la b ility of cy-alkoxyketones in strong acidic solutions, saturated ketone _15 was prepared and proved to be totally inert to the acidic reaction conditions utilized above.

In view of these results, it would appear that the observed rearrangements can be explained in terms of the mechanism shown in

Scheme II. 8

Scheme IX

In the presence of strong acid, isomerization of 1_4 to the r/, P -isomer

16 can be expected to occurtrnnsanmilar nucleophilic attack by the remotely located divalent oxygen atom in a manner reminiscent of acid- 15 catalysed Michael additions leads to intermediate 17_. This reactive oxonium salt (17) suffers nucleophilic at tack by halide ion and ring cleavage in the indicated manner; ketonization eventuates in the forma tion of 13.

Further insight into this mechanism has been obtained in our laboratory by P. C. Storm. ^ 7,8-Dihydro-5-phenyl-2H-oxocin-3(4H)-one

(18) was prepared and heated with hydrochloric acid. The rearrangement observed earlier was virtually duplicated and furanone 19_was isolated.

However^ whereas the rearrangement of v?as complete in less than

10 minutes, the half-life of 1J3, under virtually identical conditions, was seen to be approximately 15 minutes. The diminished rate of re arrangement of 18, is likely due to a decrease in the electrophilicity of C-5, the seat of oxygen attack, brought about by delocalization of 17 the positive charge developing at this site. 9

HCl

The mechanism shown in Scheme II may be u tiliz ed for the transformation of _18 to 19. However, a mechanism which does not require such isomerization prior to bonding can be written for ketone

18 (Sch erne III) in view of the stabilization provided the carbonium ion by the phenyl group. To provide validity for this proposal, 5-phenyl-3,

4,7,8-tetrahydro-2iI-oxocin (20)was prepared.16

Scheme III

When olefin 20. was treated with aqueous hydrochloric acid as above, tetrahydrofuran _22 was formed. This successful conversion demonstrates that the transannular bonding process may proceed without the assistance of a carbonyl function and that a mechanistic alternative to Scheme II can in certain cases be operative (Scheme IV). 10

Scheme IV

> < ( © 20 21

( ~ V 6H5 N d ^ICI^^CI

An examination of the reversibility of the conversion of 20 to

2.2 was undertaken. In principle, development of the incipient carbonium ion character at the terminal carbon of the chloropropyl sidechain in 22 should be anchinerically assisted by the tetrahydrofuranyl oxygen atom,

The consequence of C-0 bond formation would be generation of bicyclic oxonium ion .21_ which in the absence of efficient external nucleophiles could suffer concomitant proton loss and ring expansion to the eight- membered heterocycle 2£. Experimental support for this hypothesis was available from the reaction of .22 with an aqueous etlianolic solution of silver n itra te which gave rise to 20 in 487. yield. Thus, the microscopic reversal of the transannular reaction was realized. In this connection, it appeared that reactions of the 2 2 - ^ 2 0 variety could provide easy synthetic entry to otherwise difficultly accessible medium-sized heterocycles. However, various similar reactions with 18 13b and 23 were observed not to afford products of ring expansion; rather, solvolysis of the halogen substituent v?as encountered.

The above results demonstrate unequivocally that the formation of transannular products can result from participation of ether oxygen.

What is equally significant about the experimental findings is the apparent dependence of the propensity for rearrangeraent upon the intensity of electron deficiency at the developing carbonium ion center.

The question of whether the double bond in ^4 must migrate into conjugation with the ketone prior to transannular attack of ether oxygen was s t i l l unanswered. To this end, the synthesis of ketone 24 was undertaken.

14 -C H a l- KOi-Bu

Dimethylation of 14 was achieved with excess potassium t>butoxide 19 and iodoraethane according to established procedures. Obviously, in ketone 24 double bond migration into conjugation with the carbonyl group is not possible unless a methyl substituent migrates. Therefore, should a furanone result, the conclusion could be reached that such a non*»conjugated double bond is capable of transannular attack by oxygen. When 2A was refluxcd with dilute hydrochloric acid for periods of time ranging from 15 minutes to 12 hours, no furanone \7as detected; only starting ketone and some polymer (in the prolonged runs) were obtained.

The fact that no transannular reaction occurred upon treatment of

24 with hydrochloric acid does.not definitely prove that the

P -isomer of _14 is the reacting species. Steric retardation, caused by the gem-dinsthyl group in 24 may prevent transannular reaction and a b etter model compound would be 3,4,7,8-Tetrahydro“2lI-oxocin

(25). It is

25 conceivable that 25_ would be subject to acid~promoted rearrangements despite the absence of a stabilizing group at C-5.

Preliminary attempts to reduce the carbonyl group of .14 by means of the Huang-Kinlon modification of90 the Wolff-Kishner procedure, Raney nickel deculfurization of the corresponding ethylene 21 dithioketal, sodium borohydride reduction of the tosylhydrasone 23 and lithium aluminum hydride-aluminum chloride failed to produce 25

With no success in attempts to directly remove the carbonyl function of

14. attention was directed toward alcohol 26.

Reaction of 14 with lithium aluminum hydride gave .26 in good yield.

The brosylate of .26 (27)was prepared and treated with lithium aluminum hydride in the expectation of obtaining olefin 25, however,

a 26 27 28 the product isolated displayed no vinyl protons in its nmr spectrum but did exhibit a nultiplet of four protons in the 0 - 1§ region.

On the basis of spectral data and elemental analysis (see Experimental) structure 28 was assigned to the product isolated. Either hydride ion attack on the double bond which in turn displaces the brosyl group in an

Sjt2* reaction or hydride ion attack on the homoallylic carbonium ion formed by loss of the brosyl group could account for the formation of

28.

To further investigate the transannular interactions of both the double bond and the ether oxygen in 27, a solvolysis study of

27 and its saturated analog was carried out. The results of this study are reported in Part XXI.

Meinwald^ has shown that 2 ,3 ,6 ,7 tctrahydrooxepin (30) can be prepared by the reaction of 29_ v;ith silver oxide in ether. A similar r reaction on an 14 iodohydrin containing one more carbon atom (ID could lead to the eight-membered analog 25. The preparation of 31. (see Part IV) was accomplished by a reaction sequence similar to that of Mcinwald,

However,

H H i(c h 2) 2C-e(CH2) 2-OH ->

3.1 32

the formation of 25. w&s not realized and 2-cyclopropyltetrahydrofuran

(32) was the only low boiling product isolated. Attack of the homoallylic carbonium ion v/hich is generated upon loss of iodide ion, by the alcohol function gives 3.2. This interception of hornoallylic carbonium ions by neighboring' hydroxyl was studied in detail and is presented in Part IV. PART I I

OXOCANE: SYNTHESIS AND OINTOFKATIONAL ISOMERIZATION

Introduction

For some years there has been keen interest in the conformational analysis of medium-sized rings. 25 Recently, attempts to analyze by means of temperature-dependent nmr spectroscopy the conformations and conformational processes occurring in cyclooctane and certain of its derivatives have been reported. 26 The conclusion has been reached ' that the evidence to date can best be interpreted in terms of the preferred existence of a boat-chair (33) and/or a twist- boc'.t-chair (34) conformation(s) for cyclooctane.

33 34

At the same time, the interesting proposal has been advanced that the conformation of eight-mcmbercd rings can be expected to remain essentially unchanged when a methylene group is replaced by a heteroatora such as oxygen.27,28 However, recent studies in this lab- 29 oratory have indicated that strain minimization accompanies the replace ment of a CH 2 group by oxygen in medium-sized rings. Further demonstra tion of the smaller steric requirements of ether oxygen relative to a methylene group has been found in the preferred axial orienta-

15 16 tion of the Jt-butyl group in cis-2~alkyl-5-t-butyl-4.3rdioxg.ttcs.3Q

In view of these divergent considerations, we have deemed i t of interest to prepare oxocane, the oxygen heterocycle related to cyclooctane, and to examine its temperature-dependent nmr behavior.

Synthesis. Catalytic hydrogenation of the previously described

5-dimcthylamino-7-oxa“l,3~cyclooctadicr>e (11) led to 3-diraethylaminooxo- c^ne (35) , which V7as in turn subjected to Hofmann elimination. The exclusive product of this reaction was the er.ol ether 36_ thus indicating that the?-> -hydrogens a to the ether oxygen atom were preferentially Ol removed in the transition state for elimination. Hydrogenation of 36. over Adams' catalyst gave oxocane (37) in good yield.

38 37

A lternatively, arainodiene 14 was converted to it s ruethiodide 38; when th is substance was exposed to hydrogen over Adame' catalyst, 17

hydrogenolysis accompanied the expected hydrogenation and 37. was

isolated in fair yield.

Temperature Dependent Near Analysis. In general, cyclooctane and

i t s derivatives undergo at least two changes which are visible on the

nmr time scale, one near -30° to -90°, the other at about -150°.^

Above these temperatures, spectral variations are not observed.

Therefore, low temperature nmr spectra of oxocane have been recorded.32

At -100°, there was no apparent alteration in the absorption peaks

when compared to the spectrum recorded at room tem perature. 33 uor were there any changes"’encountered at -128° and -146°. At -160°,

soma slight broadening of the two peaks was encountered, and no

sp littin g of signals was obvious. Such minimal results are most

consistent with a highly flexible conformational model for oxocane

(37), one in which methylene wagging, pseudorotation, and other

ring inversion processes have been facilitated by the introduction

of the oxygen atom. We therefore conclude that, at least for

the eight-membered ring case, the space requirements of an oxygen

atom are, in fact, significantly less than the steric demands of

a methylene group with the reTsult that one or more very rapid

averaging process are available to 37. even at -160°.

These resu lts are interesting in view of the quite exact

parallelism which exists between tetrahydropyran and cyclohexane

in their nmr properties,3^ and in light of the examples which suggest

that the introduction of multiple hetero atoms into a ring system generally

effects an increase in the barrier to ring i n v e r s i o n .35 PART I I I

ACE TOLY SIS OF OXOCAN-3-YL AND 3,4,7 ,8-1'E TRAHYDRO - 2H-

OXOCIM-3--YL BRQSYLATES

Introduction

As discussed earlier, studies of the relative rates of solvolysis ofOJ~nethoxyalkyl p-bromobenzenesulfonates, for example

38 and 40, have indicated that the driving force for nucleophilic participation of the methoxyl group is very high only in those cases where five*- or six-membered cyclic oxonium ions such as 39_ and 4_1 3 are possible intermediates,, It is interesting that in this series,

HOAc > 08 s

38 39

18 the primary methoxyl groups of 2-methoxyethyl and 3-methoxypropyl p-bromoben::enesulfonates provided no evidence for anchimeric assistance, although the absence of MeO-3 and lieO-4 participation after the rate determining ionization step cannot be discounted.

I t must be remembered, however, that the inductive effect of the oxygen in the Ke0~3 case would significantly decrease the rate of * ionization and could overshadow any small neighboring group participation effects of the ether oxygen. Secondary and tertiary methoxyl groups, on the other hand, definitely exhibit MtO-3 p a rtic i- 36 37 pation. * This participation was observed by product studies, not rate studies. For example, the rnethoxybromide 42_ with silver acetate in acetic acid affords only the methoxy acetate of retained configuration, presumably because of the intermediacy of oxoniun ion 43. One case of

42 43

MeO-4 participation has recently been reported.^ To account for the products formed upon solvolysis of 44 the oxonium ion intermediate 20

HOAc ^ 'CH3 »Ts

44 45

In contrast to the considerable work published on aliphatic and alicyclic ether participation, few examples of neighboring group participation attributable to ether oxygen when present in a cyclic system have been observed or studied,

A further consideration becomes important when dealing with certain cyclic ethers. Thus, an oxygen atom, when present in the

(3-position to an incipient carboniun ion, has an inductive effect 38 which reduces an Ionization rate by a factor of approximately 100,

However, with favorable geometry, this same oxygen atom can stabilize a neighboring carbonium ion by a resonance effect. The results of the solvolysis of the exo-isomer of 7-oxanorbornyl chloride (46) suggest that this molecule displays these two effects one after the other as the geometry of the ring is altered during the progress of the reaction. ^9 Because of the failure to observe any major rate differences in the exo/endo reactivity ratios relative to the norbornyl system,

Martin and Bartlett have suggested that anchimeric assistance of the type shov;n in 47 is not very important in the rate "determining step.

However, strain factors generated in the incipient formation of 47 may be the reason possible neighboring group participation is inhibited. ~

or 47

The purpose of the following work was to determine to what extent ether oxygen, when situated beta to an incipient carbonium ion in a medium-sized ring, would affect the rates and products of solvolysis.

To this end, brosylates and 48, were prepared and solvolyzed. 22

27 48

The solvolysis of the carbocyclic analogs, (49) and (50), of these heterocyclic brosylates have been studied by Cope and 9 Petersoiip The results of their work serves as a model to which the solvolysis of _27_ and 4j3 may be compared. I t is interesting to note that two of the products formed from 49 are due to participation of the double bond.^ The solvolysis of Z7 gives an opportunity to compare the amount of participation of ether oxygen with that of a double bond which is capable of forming a homoallylic cation.

49 487. 397. 137.

50 23

Results and Discussion

Synthesis., Unsaturated brosylate 27. was prepared from alcohol J26

as described in Part I„ Brosylate AS was obtained by reaction of

ketone 15. with lithium aluminum hydride to give 51 which was subsequently

converted to A8 by treatment with p-broraobenzenesulfonyl chloride in pyridinec

Q \ LAH BsCl ; U8 ( - > 15 51

Rate measurements0 Brosylates 27 and 4.8 v?ere solvolyzed in acetic acid containing a slight excess of sodium acetate. The procedure used (see Experimental Section) was the same as that 9 used by Cope in the acetolysis of A9 and 50. The results are

summarized, in Table 1. TABLE 1

Acetolysis Rate Data for 27^ and 48

Relative Rates at 70? Cyclo- Temp. Rate kx 10 sec"* hexyl Brosylate H * S + Compound (°C) kxlO5 sec"1 70.0°(calc.) «_1,0______Qccal./mole (e.u.)

O B s 39.8° 2.20 ± .03

50.7° 8.13 ± .03

w 63.9° 34.1 ± 0.2 21 70.0° 67.3 8.4 23.3 ± 0.2 -5.81+0.77 (calc.) OBSO B s 39.8° 0.224 ±.004

51.0° 1.01 ± .01

63.9° 4.55 ± .06

70.0° 8.77 1.1 25.5 ± 0.3 -2.97 ± 1.02 (calc.) TABLE 1 (Continued)

Acetolysis Rate Data for 27 and 48

Relative Rates at 709 Cyclo Temp. Rate k xlO^ sec”^ hexyl Brosylate H * S * Compound (°C) kx 105 sec“^ 70.0°(calc.) “ 1.0 .(kcal. /mole)(e.u.)

3-cycloocten-l-yl brosylate3 (49) 2970 370 19.0 -13.4

cyclooctyl brosylatea (50) 14S0 185 21.0 -3.9

4-cycloocten-l- yl broxylatea (52) 23.2 2.4 26.2 2.0

Cyclohexyl Brosylate3 8.02 1.0

These values were taken from reference 9. 9 An examination of the data obtained for the carbocyclic brosylates shows that the solvolysis rate of cyclooctyl brosylate

(50) is strongly enhanced over that of cyclohexyl brosylate, the

latter being a suitable reference standard whose rate of solvolysis is comparable to that of acyclic brosylates. This enhancement is due

to relief of steric strain in the transition state for formation of a 9 planar carbonium ion. The solvolysis rate of 4~cycloocten-l~yl brosylate (52) is much slower. tha*a that of 50. Thus i t appears that introduction of one double bond into a cyclooctane derivative is sufficient to relieve much of the strain in the molecule (part of which arises from several eclipsed hydrogen, interactions). Additional evidence for this relief of strain is provided by the fact that

4-cycloocten-l-yl bromide can be prepared in good yield by addition of hydrogen bromide to 1,5-cyclooetadiene,^ indicating that addition of a second mole of hydrogen bromide to the 5-bromocyclooctene is somewhat slower, presumably because of the strain attending the formation 9 of the completely saturated compound.

From the data in Table 1, 3-cycloocten-l-yl brosylate (49) is seen to solvolyze 130 times more rapidly than the 4-cycloocten-l-yl derivative at 70°, The rate of 49. unassisted by double bond participation may be considered to be approximately the same as for 52. The amount of double bond participation, therefore, is then indicated by a rate enhancement of approximately 130 fold.

The inductive effect of ether oxygen when situated p to an incipient 27 carbonium ion would be expected to cause the rate of solvolysis of

48 to be approximately 100 times slower than that of 50. Of course, the provision must be made that there is no neighboring group participation by the oxygen in the rate determining ionization step.

In fact, as seen in Table 1, 48. solvolyzed 170 times slower than 50.

This large decrease in rate cannot be totally accounted for by the in ductive effect of the P oxygen atom and must be p artially due to the r e lie f of strain which accompanies the replacement of a methylene group by an oxygen atom in an erght-meinbered ring.

The rate of solvolysis of brosylate 27_ would be expected to be

170 times slower than that of 3-cycioocten-l-yl brosylate (49) i f the ether oxygen was exerting the same steric and electronic effects which i t did in the solvolysis of 48. From the data in Table 1, i t is seen that brosylate 27. solvolyzed only 44 times slower than its carbocyclic analog 49. Apparently the relief of strain which is caused by the ether oxygen in J27_ facilitates participation of the double bond and formation of the homoallylic cation. A comparison of the solvolysis products of 2J_ and 49 lends support to the proposal that there is more double bond participation in .27. than in 49. Thus, the cyclopropyl acetates from the solvolysis of 49. corresponded to only 52% of the total products while the cyclopropyl acetates from the solvolysis of 2J_ comprised 91% of the total.

A final analysis of the rate data show’s that both inductive and steric effects caused by the ether oxygen in 27. and 4jl were quite definitely involved in the rate determining step of solvolysis. However, no indications of rate enhancement due to neighboring group participation 23 by the ether oxygen were obtained.

Solvolysi s Products of Oxocan-3-yl Brosylate

and 3 ,A.7 . 8 - T c t rahyd r o - ?.I1 - 0 xo c 1 n - 3 y 1

The crystalline brosylates were solvolyzed at 70° for 10 half- lives in acetic acid containing a slight excess of sodium acetate.

After workup (see Experimental), the products were distilled and analyzed by vapor phase chromatography (vpc). Pure samples of each component V7ere obtained by preparative vpc.

Distillation of the acetolysis products of oxocan-3-yl brosylate

gave a lov? boiling fraction, a high boiling fraction and a solid residue which remained in the distillation pot. The overall yield of recovered products was 93.8%. Scheme V lists the products and their respective percentages.

Scheme V

55 ( 25 . 2%) 5 5 (2 1 -5 %) Ch^OAc H2OBs (93.8%) HOAc >

48 52 (32.3%) 5 1 (8. 4 %)

+ AcO-Chyg-tH £6 ( 12. 6%) 29

The low boiling fraction displayed one peak on the v p c ^ a and was assigned structure 5^ on the basis of the following data. Hydrogenation of 53 gave the known oxocane (37). The infrared spectrum showed no peak in the enol ether region (6.1|j) as was observed for 360 The nmr spectrum (CCl^-TMS) displayed a multiplet at 5.42 6 for two vinyl protons, a narrow m ultiplet at 4.04 6 for two protons which are a to oxygen and allylic, a multiplet at 3.60 £ for two protons cc to oxygen, two a lly lic protons as a m ultiplet from 2,25-2.65 6 and four protons as a multiplet from 1.45-1.75 6.

Vpc analysis of the higher boiling fraction showed two peaks 44a which were separated by preparative vpc. The higher retention time product was assigned structure 5,6 on the basis of its III spectrum which exhibited peaks at 2760 cm ^ for an aldehyde proton and 1750 cm"^- for an ester and aldehyde carbonyl, its nmr spectrum which exhibited a triplet (J K 1.7 Hz) at 9.97 for an aldehyde proton and a singlet at

2.07 6 for the methyl group on an acetate function, and the elemental analysis of its 2,4-dinitrophenylhydrazone.

The nor spectrum of the lower retention time product displayed two sharp singlets in the acetate region and it was decided that two components of identical retention time were present. Lithium aluminum hydride reduction of this acetate mixture gave two alcohols which were readily separated by preparative vpc.^a The two alcohols were assigned structures 51 and 53. A sample of 51 was identical in all respects to the alcohol from which brosylate 48 was prepared. The proposed structure of 5J3 was confirmed by conversion to its brosylate (57)which upon treatment with lithium aluminum hydride gave 2-methyloxepane (59). An independent synthesis of 59 was

BsCl LAH > >

57 59 accomplished by hydrogenation of 2-nethyloxepine.^

The assignment of structure J57 to the solid which was recovered from the solvolysis of 48 was based on the fact that its melting point was identical to that of the brosylate of alcohol. 58, Further structure proof was obtained by the cleavage of brosylate 57_ with sodium naphthalene anion radical^ to give alcohol 58.

The isolation of brosylate f>7 from the solvolysis of 4j8 was riot totally unexpected. When the rate of solvolysis of 48 was determined 31 i t was found that a good first-o rd er rate plot could be obtained only if points taken during the first 0.3 half-life were used.

After this time a decrease in the rate of disappearance of brosylate was observed. To account for this rate decrease, the internal return of brosylate ion to give a more stable primary brosylate was proposed.

The five products obtained from the solvolysis of 4J5 can be derived mechanistically from the oxonium ion intermediate 60.

Thus, loss of a proton from in structure (50 v?ould lead to olefin

53. Attack of acetate ion at C 2 would give 5,4 while attack at would give J>5. Internal return of brosylate ion at would eventuate in the formation of 57. Aldehyde 5J5 could be formed by loss of a proton fron to give olefin 6_1 which would be expected to undergo protonation in acetic acid with the formation of carbonium ion 62.

Attack of acetate ion in the manner -shown would then lead to 56,

\ 60 - > 56

AcO 61 62 A further transannular phenomenon known to occur in eight-membered rings containing a carbonium ion is the 1,5-hydride sh ift.^ I f such a process were occurring in the solvolysis of it would not be observed in the products formed because carbonium ions 63_ and 64 are identical. To determine to what extent, if any,a 1,5-hydride shift was occurring, brosylate 65. was prepared and its

63 solvolysis products were studied. Ketone 15. was reduced with lithium aluminum deuteride to give the dcuterated alcohol 66; reaction of 66 with brosyl chloride in pyridine gave 65.

LAD >

15 66 65

Brosylate 65. was solvolyzed as previously described, the crude acetate mixture was reduced with lithium aluminum hydride and the alcohols thus obtained, 66. and 67., were separated by preparative vpc.^a Integration of the nmr spectra of £6 and 67. showed only 33

five protons at low field indicating that no 1,5-hydride shift had occurred. Further evidence that the deuterium v/as still in its in itia l environment v/as obtained by converting solvolysis product 66

to its acetate <58,, The nmr spectrum of acetate 54 displayed a one proton multiplet at 4.85 6 for the hydrogen a to the acetate group.

Therefore, when the nmr spectrum of 68 exhibited no absorption in the

4.856 region, it was clear that very little if any 1,5-hydride shift had occurred. Mass spectra of the deliberated (67_ and 68) and non-deuterated

(56 and 54) compounds were cor/pared and showed that very l i t t l e if any of the deuterium initially present in brosylate 65. had been lost during the reaction sequence.

Mechanistically, products 53, 54 and 5_6 could be formed without invoking the participation of ether oxygen. However, the fact that no 1,5-hydride shift was observed in the solvolysis of 48 indicates that the oxonium ion 60 is either being formed before other reactions occur or that steric minimization due to the presence of the ring oxygen discourages transannular involvement. This point remains to be tested.

Although no rate enhancement due to neighboring group participation of ether oxygen was observed in the solvolysis of brosylate 48,

the products isolated from the solvolysis establish the fact that

ether oxygen can indeed profoundly affect the course of a reaction in an oxygen-containing mesocycle. Additional evidence of ether oxygen participation v/as obtained when a small sample of brosylate 37 was

solvolyzed in acetic acid. The acetate mixture thus obtained v/as 34

treated with lithium aluminum hydride and the alcohol mixture v/as

1) HOAc, NaOAc 2) LAH

57 (567.) (217.) 58

subjected to vpc analysis . ^ la In addition to a 567, yield of starting

alcohol 5_8 and 217, of an unknown compound, a 237, yield of 51 was formed.

Oxonium ion intermediate 6Cl must again be proposed to account for the

formation of 51.

Nov/, the possibility that bond migration, rather than oxonium ion

formation, may be accounting for the ring expansion of 57. to 51_ should

be removed from consideration. I t has been shown^ that the acetolysis

of cyclohcptylcarbinyl brosylate, the carbocyclic analog of 57_, yields

only cycloheptylcarbinyl acetate and no cyclooctyl acetate. The

inductive effect of the ether oxygen in 57_ would be expected to further

decrease the possibility of bond migration and e_gain gives evidence to

support the intermediacy of 60_ in the solvolysis of 48.

Brosylate 27 was solvolyzed and the reaction mixture worked up

in the usual manner to give a 99.57, overall yield of products isolated.

No low boiling olefin fraction v/as obtained upon distillation of the

crude acetates and no residue remained in the distillation pot.

Scheme VI lists the products obtained and their respective percentages

as determined by vpCo^fa»^

I 35

Scheme VI

? 4- (1.57.) ( 2 . 2%) (4.1%) HcAc (99.5%)

-4- 'OAc

(68.9%) (22.3%) a . 0.7.) 21 72

Pure samples of 69-72 were d iffic u lt to obtain ov.’ing to the similarities of their vpc retention times.^a”c Acetates 7j0 and

71 displayed overlaping vpc peaks and the peal; corresponding to

72 followed very close after these. Conversion of the acetate mixture to alcohols made it possible to separate 70a; however, 71a and 72a now had the same vpc retention times. - -

"OH

26 70a 71a 72a 36

A sample of 6£ was found to have identical vpc retention times and infrared spectrum to those of an authentic sample. Reduction of the crude acetates from the solvolysis of 27. to their alcohols gave 26. which was also identical to a previously prepared sample.

The structure of 70a v/as determined from the following data.

Hydrogenation of 70a gave the known alcohol 51. thereby establishing the position of the alcohol group. The nmr spectrum of 70a displayed six protons in the 3.5-4o3Sregion, two vinyl protons, too allylic protons and too higher field protons. Only too structures f i t this data; 70a and 73. An independent synthesis of 73. v/as carried out. OH

Q V n

Upon refluxing a solution of 14 and p-tolueneculfonic acid in benzene*~fa for 15 min. an equilibrium mixture containing 85% of 14 and 15% of 16. v/as achieved. (This is probably not the true equilibrium ratio since considerable decomposition accompanied this reaction.)

An 80% pure sample of 16. v/as obtedncd by preparative vpc^fC

0 0 X TsOH LAH 15 ■* i. V

14 1 6

(apparently some isomerization back to 14 occurred on the vpc column.) The structure of 16. was confirmed by hydrogenation to the kiiown JL5 and by comparison of it s u ltrav io let spectrum EtOH ECOII 229o 5 (8000)] to that of 2-cyclooctcnone 229.5 (8100) max max Reaction of 1_6 with lithium aluminum hydride at 0° for 5 minutes gave

73. The solvolysis product under study exhibited spectral properties

different from those of J3_ and was therefore assigned structure 70a.

Additional evidence in support of structure 70a is found in the great

similarity of its nmr spectrum to that of olefin 53.

The structures of the cndo and exc> cyclopropyl alcohols 71a

and 72a were assigned on the basis of elemental analysis and spectral

data (see Experimental). To show that 71a and 72a were indeed

epimers, a pure cample of 71a was oxidized with Jones reagent

to give ketone 74. This ketone has an unusual double carbonyl

peak V7ith maxima at 17 20 and 1695 cm.“^, which was also observed

+ 72a

(597.)

in its carbocyclic analog^ and explained in terms of two different

conformations, only one of which possesses the steric features

necessary for lowering the carbonyl frequency by conjugation with the cyclopropane ring. Reduction of ketone 74, with lithium aluminum hydride gave a mixture of the two isomeric alcohols in the ratio of

41% 71a to 59% 72a.

The remaining problem was the assignment of exo and endo

stereochemistry to the two cyclopropyl acetates obtained. In the

solvolysis of the cnrbocyclic. analog of 27 the major cyclopropyl compound was shown to be the endo isom er.^

A mechanism can be written for the formation of the cyclopropyl acetates, in %?hich a bridged carboniun ion 7,5 is formed. The endo

75 7,1

isomer would be predicted to be the major product from mechanistic considerations (assuming that the solvolysis was a concerted process or proceeded via a horaoallylic carbonium ion intermediate)

I t has been r e p o r t e d ^ * t h a t the hindered nature of an axial hydroxyl situated on a cyclohexane ring as compared with an equatorial substituent causes an axial alcohol to be more weakly adsorbed on alumina than its equatorial epimer. Studies of models of 7J;_ and 7_2 showed that the acetate group of the endo isomer is much more hindered than that of tlie exo isomer. Therefore, the major acetate, which had the lower vpc retention time, would be 39 predicted to be the endo isomer 71.

The reduction of ketone 74. with lithium aluminum hydride was found to give 41% of the endo isomer and 59% of the exo isomer.

Thus, the major product (72a) arose by attack of hydride ion from that side of the ring on v/hich the cyclopropyl group is situated.

Other examples where the exo isomer is the major product of hydride reduction have been reported and are illu strated below.

_■> endo + exo (ref. 40) 30% 70%

CH. LAH ■> endo + exo (ref. 51) 33% 67%

CO CO Due to the paramagnetic current associated with an alcohol » or an acetate group, the chemical shift of protons in the immediate vicinity ofthe hydroxyl group is shifted downfield. Thus, when a hydroxyl or acetate group is endo to a cyclopropyl group, the cyclopropyl protons are strongly deshielded.51>54 ^ examination of the nmr spectra of the two cyclopropyl acetates obtained from the solvolysis of 27_ shows that the cyclopropyl protons of the 68.9% product (0.52”1<,43 6 ) are shifted downfield when compared to the cyclopropyl protons of the 22.37. product (0.28-1.32 ® ). Fro;n this, the major product is again predicted to be the endo isomer 71.

Although the data presented thus far strongly suggests that the major product is the endo isomer, an anomalous chemical shift in the runr spectrum of the exo-isomer 72 must be explained. The nmr spectra of the carbocyclic analogs of 71a and 7_2a display

Ct -carbinol hydrogen peaks at 4.2 ® for the endo compound and 3.3 6 for the exo compound.It has been shown*^ that the peak due to an axial hydrogen when (y. bo an oxygen atom in a cyclohexane ring is observed at higher field in the nmr. This argument was used by Cope**"* to explain the high field shift of the Gt -carbinol hydrogen in the exo isomer. Similar high field shifts of protons which are cis to a cyclopropyl function have been attributed to a shielding effect of the cyclopropane ring.*^"’^

From the above results, the endo cyclopropyl acetate 7_1 would be expected to display an a -proton in the nmr at lower field than the

q, -proton of its exo isomer 72. The opposite result was observed

and might be construed as negative evidence for the endo and exo structure assignments. However, an examination of models of 7_2 41

demonstrates that although the a -hydrogen is in a position v;here

it v.’ould be shielded by the cyclopropane ring; it is also held

in close proximity to one of the non-bonded pairs of electrons

on the ether oxygen situated across the ring. The deshielding

of the &-proton of 7J2, owing to the paramagnetic current associated v?ith the proximate ether oxygen, apparently is overriding the

shielding caused by the cyclopropane ring. Therefore, the a-proton

of 72. appears at lower field in the nmr than expected.

OAc

72 PART IV

THE INTERCEPTION OF 1IOKOALLYLIC CATIONS BY NEIGHBORING HYDROXYL

— Introduction

Of the various effects which can be exerted by a substituent on the reactions of an organic molecule, that which involves direct or partial bonding of the substituent to the developing or completely unfolded reaction site continues to enjoy widespread interest.”*

The propensity of such groups as amino,thioethcr,**^ carboxyl,^ and alkoxyl^ for neighboring group participation in certain simple systems has been recognized for some years. In contrast, the neighboring hydroxyl group as such has been l i t t l e s tu d ie d ,^ although alkoxide ions produced from such alcohols in alkaline solution have received considerable attention.** The great majority of the cases examined have been acyclic and cyclic compounds where the substituent had but one reactive site with which to interact intramolecularly.

I t was our intent in this work to investigate systems in which a hydroxyl substituent could avail itself of two widely differing sites in the course of intramolecular cyclization. Such a study forms part of 63 a continuing program designed to examine new aspects of oxygen and sulfur**^ neighboring group effects.

The investigation of such systems was deemed of special interest because of the possible, relevance.of the results to several important theoretical points among which may be cited the development of our

42 43

knowledge concerning the phenomenon of neighboring group participation

and the factors which control ring closure. For reasons of synthetic

accessibility, cis- and trans-iodohydrins of types 76. and 7_7 were utilized as substrates. Reaction of 7J3 and 77 with sources of

(CH?) OH

' -C/ \ i (ch2 (CH ) OH I(CH, H 76 77 monovalent silver and characterization of the products which result in the individual cases have been accomplished and the results are here presented and discussed.

Results

The 6-Hydroxyl Substitu e n t. The synthesis of the isomeric cis- and trans-8-iodo-5-octen-l-ols (85C and 85T) began in both instances by the conversion of tetrahydropyran-2-methyl chloride (78) to 5-hexyn-l-ol

(79) with sodium amide in liquid ammonia solution (Scheme VI). Blocking of the hydroxyl group by treatment of 79. with dihydropyran led to the formation of ^0. Careful addition of excess ethylene oxide to the anion of 80, prepared in situ by means of lithium amide in liquid ammonia, afforded £51 in good yield. Partial catalytic hydrogenation of 131. gave

the desired cis olefin 82s which was subsequently converted into its p-toluenesulfonate ester (83C) in 83% yield. Removal of the tetrahydro- pyra.nyl group resulted in formation of cis-hydroxytosylate 871. Without purification, 84C was transformed into the corresponding iodohydrin 85C 44

Scheme VI_

NaNH -----> HCuC(CH2)4OH H Cl. NH, HC1 2 3

78 79

1) LiNlU, NH3 HCeC(CII2)40-^ ^ ------> K0(CH2)2CkC(CH2)40 2) 80 \ / 81

H \ J 1 2 ^ / C=C\ T£C1 Pd-C 110(CH2) 2/ \CII ) 0 ( ) ------> \ — / py 82c,

(h2S04) -/H \y - c Te0(CH2) 2 X(CH2)4-0-^ 'j aq. CH-jOH

83C

H H Nal

TsO(CH2) 2/ n (ch2)4oh acetone I(CH2) 2 N c H ^ O lI

84C 85C

Ag,0 Li

ether

87 45 by treatment with sodium iodide in acetone. Slov? addition of an ethereal solution of 85C to a rapidly stirred suspension of powdered silver oxide in ether at room temperature resulted £h quantTItative formation of silver iodide. Careful workup of the organic solution led to the isolation of 2~cyclopropyltetrahydropyran (86) . bp 76-77° (55 mm), in

26% yield and^>S 8% purity, together with somewhat lesser quantities of a mixture of three higher boiling components, bp 105-125° (3.5-0.6 mm), which were not further characterized. The structure of 86 followed from its spectral parameters (see Experimental Section) and was confirmed by independent synthesis from cyclopropyllithium and 2-bromotetra- hydropyran (87).

The successful synthesis of 85T, summarized in Scheme VII, was achieved in a scries of four steps starting from 81. The assignment

Sch erne VIT.

Na TsCl 81 NH K0(CH2) ^ \ l py 3 82T

TsO(CH9) / nh TsO(CH2) / nh

84 T83T

Nal A£20 \ =/ CH2>40H } 86 acetone i (ch2) / h ether

85 T 46 of trans-stereochemistry to the olefinic linkage introduced during the sodium-liquid ammonia reduction of the triple bond is consistent with the well-recognized stereoselectivity consistently observed in such transformations.^ When 85T was gradually added to silver oxide in ether in the predescribed maimer, 8 6 ^ was again isolated

(39% yield,>98% purity). Exhaustive distillation likewise afforded an approximately equal amount of significantly higher boiling materials which were not examined.

The 7-Hydroxy 1 Substituent. 4-Pentyn-l-ol was converted into its tetrahydropyranyl ether and hydroxyethylated using ethylene oxide and lithiuia amide in liquid ammonia to give 881 . By procedures similar to those employed with 81, this alcohol ( 8 8 ) was transformed into the isomeric cis- and trans-7-iodo-4-hepten-l-ols, 31C and 31T, respectively (see Scheme VIII). When 31C was added under conditions

Scheme VIII

H0(CH2 ) 2 C2C(CH2 -> H 0(Cll2)2CH=CH(CH2)3'

88 89C, cis double bond

89T. trans double bond

TsO (CH2) 2 CH=CH(CH2) 31 -> Ts0(CH2 ) 2 CH=CH(CH2 )30H —

90n and 90 T 91C and 91T

i(ch 2)2 cii=ch(ch2) 3oh

31C and 31T 32 47

of moderately high dilution to a rapidly stirred slurry of pov?dered

silver oxide in ether, 2-cyclopropyltetrahydrofurnn (3?;) was produced

in 47% yield. This volatile liquid gave a single peak when subjected

to gas chromatography on several columns thus establishing that this material v;as the sole cyclization product. The formulation of 32^

as 2-cyclopropyltctrahydrofuran v.'as supported by its elemental analysis

and by the great similarity of its infrared and nmr spectra with those of 86. In particular, the nmr spectrum of 32 (in ODClg) displayed a

5-proton multiplet in the 6 0.1-1.1 region (cyclopropyl ring), a

4-proton multiplet at 1,5-2.1 (5-ring methylene groups), and a

3-proton multiples at 3.1-3.8 ( ^£U-0-).

When trans-iodohydrin 31T was subjected to similar silver oxide

treatment, 3j2 was again isolated (57% yield) after careful workup. As in the previous example, 3_2 was found by vpc to be uncontaminated with other lov.'-boiling products.

The /3"Hydroxyl Substituent. The reaction of cis-6-iodo~3-hexen-3r ol

(29c) with silver oxide in ether has been studied previously by Meinwald *) / and Nozaki f who found that 2,3,6,7-tetrahydrooxepin (30) v;as produced

\ /H Ag20 — r : 30 29C in 25% yield and a high state of purity. As in the above examples, quantities of high-boiling m aterials (apparently a 3-component mixture) 24 were formed along with 30. At the outset of the synthesis of trails isomer 29T, sodium in

liquid ammonia reduction of acetylenic alcohol 92 was observed to

yield two products, the major component (34.9%) proved to be the

known trans-3-hoxen-l-ol (93T) . The alcohol formed in lesser quantitie

(24.8%) was shown to be the desired trans olefin 94T (see Scheme IX).

The unprecedented production of 93T can best be rationalized in terms

of displacement of the tctrahydropyranyl group from intermediate

radical anion 9J3. Thus, one-electron reduction of the acetyl enic bond

Scheme IX

II0(CH2)O-C(CH2)20 + 92 93®

KO(CII2) 2 TsO(CH2) 2

94 T 95T

(h2so 4) H\ / ( clI2>2OH — > aq. CH3OH tso (ch2) 2/ H 49 in £2 gives rise to radical anion 98^5 which clearly can cyclize with

Nal polymer

■ - - > acetone I(CK2/ 2 ether

9 7

concomitant 1,3-displacement of the oxygenated function. Rapid reduction of the resulting methylene cyclopropane radical (99) results in conversion to anion 100 which can

9 2 Nc'i* . HO (CII2) 2C^CCiI^CH2^ ^ > ^

98 ^ h4 i2

H0(CH2)2C=<^] Na‘ . H0(CH2)2ci<^j ------

99 100

further H0(CH2) CECCH C H j ------> 93 T reduction 101 be expected to rearrange with rupture of the three-membercd ring and generation of acetylenic alcohol 101. Reduction of 3.01 to 93t needs no additional mechanistic clarification. Although nucleophilic displacement of tetrahydropyranyloxy groups has nr , been observed frequently, it is well recognized that 1,3-intrarr.olecular displacements are especially facile;^ further, intramolecular alkylation reactions of intermediate radical anions derived from carbonyl, ^ o'>P-unsaturated 50 68 6 Q carbonyl, and acetylenic functions have recently been shown to be

especially useful synthetic reactions.

trans-Alcohol 94 t was converted to trans-iodohydrin 29T according

to the established procedure. TIhen an ethereal solution of 29T was

added slowly to a rapidly stirred slurry of powdered silver oxide

in anhydrous ether at room temperature, only polymer formation was observed. No low-boiling products could be detected.

Discussion

It is now well established that a rr bond disposed in p, y-fashion to a developing carbonium ion significantly enhances the rate of ionization, dramatically affects product stereochemistry in rigid systems, and frequently promotes the formation of rearrangement products. 7 0 The magnitude of this effect has been found to vary substantially as a function of solvent (enhanced in less nucleophilic media) and substitution about the double bond.^1 These phenomena have, been ascribed to significant delocalization of the neighboring

•n" electron pair, such an interaction leading to the generation of a stabilized cation termed a homoallylic cation.^ Conceptually, the homoallylic cation may be considered to resu lt from mixed T f, o' overlap of the vacant p orbital with the most proximate it lobe of the olefinic 72 linkage, v iz .. 102. The strain introduced by pronounced compression of the 1,3-distance in 102 is apparently more than offset by an increase in resonance energy estimated (LC/D-MQ methods) to be in the 72^73 order of 2.8*>6 kcal/mole, t*’ In valence-bond terms, this homo- 51

c h ^ c h c h 2c h 2x

conjugative stabilisation is seen to involve electron delocalisation across intervening carbon atoms as in 103 (—) 104.

103 104

The homoallylic cation, therefore, can be considered an ambident electrophile, irrespective of whether it is termed a "classical" or

"non-classical" entity.^ In actual fact, the capability of a homoallylic cation to offer two alternative sites to an approaching nucleophilic species was first discovered in the early work within v. the cholesteryl-i-cholesteryl series.^ Further, the conformational rigidity conferred upon the cholesteryl cation (105) by the steroidal framework clearly testified to the propensity of N: to approach 105 exclusively along paths a and b and become bonded exclusively to the sides of C-3 and C-6, respectively. 52 N

105

The various cyclic ethers produced upon treatment of homoallylic iodides 85c, 85T, 31C, 31T. and _29_C with silver oxide may arise from two distinctly different mechanistic routes. The first of these path ways involves attack by elcctrophilic silver ion at the site of the iodine substituent to produce a homoallylic cation such as 103 which reacts intrar.-olecularly with the neighboring hydroxyl group to give product. A lternatively, i t is possible that the hydroxyl group firs t releases its acidic proton to the somewhat alkaline silver oxide reagent and that cyclization actually results from intramolecular nucleophilic displacement of iodide by neighboring alkoxi.de ion. To resolve this dichotomy,iodohydrins 31 C and 31 T were subjected in one instance to the action of aqueous silver n itrate at room temperature, and, 53

in a second set of experiments, to aqueous sodium hydroxide

(one equivalent) under identical conditions. Whereas mild

treatment with silver nitrate led again to the isolation of 2-

cyclopropyltetrahydrofuraii (32) in yields roughly comparable to

those previously obtained, the experiments in alkaline solution

were found to give 60-70% recovery of the iodohydrins together with

smaller quantities of the related diolsc The presence of

32 could not be detected in the la tte r experiments. These results

clearly establish the fact that the cyclisations in question proceed

by way of attack of the hydroxyl substituents upon intermediate

homoallylic cations.

An analysis of the above data in terms of the structural factors which 7r affect 110-4, -5, -6, and -7 involvement and which control the ultimate

site of bonding of the hydroxyl group to the electron-deficient homo

allylic moiety is now in order. Firstly, the interception of homo

allylic cations by neighboring hydroxyl groups is obviously not dependent upon the geometry of the double bond of the iodohydrin precursor if

intramolecular nucleophilic attack is to occur at the more highly substi

tuted internal carbon atom of the homoallylic cation (compare 85C and

85 T. 31C’and 31T). The steric situation present in the pair of

geometric isomers is illustrated in structures 106 and 107. It is quite

evident in these formulas that the carbon center which is to be the ultimate seat of nucleophilic attack is equidistant from the neighboring

hydroxyl group in both examples. 54

(CH2)n-O H

106 107

In contrast, a total dependence on geometry obtains in those cases where attack at the less highly substituted terminus of the homoallylic

cation is demanded by the structure (compare 2 Sq_ and ?.9T) . The reason

for this is quite apparent in structures 108 and 109. Thus, whereas

the hydroxyl group of the ci_s isomer (108) is ideally oriented for

H O -(C H 2)2

108 109 approach to the lower lobe (sec 105) of the electron-deficient C-6

carbon atom, the same group in 109 finds its e lf in a very unfavorable orientation for participation. The important point established by

this comparison is that a very large difference in reactivity cannot be reliably predicted for a given cis:trans isomer pair unless allowance

is made for the fact that intramolecular neighboring group cyclization 55

involving homoallylic cations can take place at two sites v?hich differ

considerably in their relationship to the functioruilited sidechain.

It is clearly seen from the foregoing account that the formation

of fivc-(3_2) and six-membercd (86) oxygen-containing rings is

favored and that no detectable quantities of eight- and nine- membered cycles, respectively, were produced. The larger rings could

theoretically arise fx*om intramolecular nucleophilic attack at the

alternate terminus of the homoallylic cation (HO-8 raid HO-9 partici

pation). It is also significant that seven-ring formation (i.e., 30)

is observed, but that the oxetane derivative 97. was not obtained.

This behavior is not unexpected and, in fact, finds a great deal in

common with the known enchimeric assistance capabilities of neighboring 61 58 59 60 methoxyl, amino, sulfide, and carboxylate groups (Table 2).

A particularly interesting aspect of the present study is found in

the observations that the neighboring hydroxyl groups in each instance

participate by interaction exclusively with only one of the two possible

sites on the homoallylic cation. T able 2

Relative Capabilities of Several Heteroatomic Groups

for Anchineric Assistance

Requisite Size Solvolysis of Cyclization of Cyclization of of Intermediate ch3o (ck 2) xobs Br(CH2)xNH2 C1(CH0) SR in 507. aq. Keterocycle in EtOH, 75° lc (min "^•>25°)b acetone, k (nin ) c (relative rates)®

3-ring 0.25 0.036 -----

4-ring 0.67 0.0005 ----

5-ring 20.3 30 too rapid for measurement (R = CH3, 80°)

6-ring 2.84 0.5 0.016 (R=CH3,80°) 0.053 (R=C2H5,100°)

7-ring 1.19 0.001 0 . 0 0 0 7 (r=c2h5,ioo°) . ■ ■ i■■■ ■■ - ■ ■■■—■ i b c Reference 61* Reference 58. Reference 59a and 59b.

Ln O' PART V

STABILIZED DERIVATIVES OF ClS.ClS,CIS- l .3,5-CYCLODECATRIENE

KETO-ENOL TAUT015KRISM IN 2.3-DICARBOMSTKOXY - ClS,ClS

3, 5-CYCLODKCADIENONES AND ClS-3-CYCI.OALKENONF.S

Introduction

Of the many cyclic polyolefins known to date, the conjugated

trienes of the medium-sized rings are certainly among the most in triguing. The recent interest to be accorded such molecules is due in part to the unusual highly-stereospecific nature of their valence- bond isomerization during which the triene with its six ^ -electrons is converted to a cyclohexadiene with four % - and two a -electrons.

Such rearrangements, although purely thermal in nature, occur under very mild conditions (usually at appreciable rates in the temperature range 0-50°). For example, c ls , c ls ,c is -1,3,5-cyclononatriene (110) rearranges smoothly to c is -bicyclo [4.3.0] nona-2,4-diene (111) at

25o0° with a first-order rate constant of (7.62 ± 0,13) x 10^ sec."*.

( A II* s 23.0 kcal/mole; A -4.7 e.u.).*^ The transvcis,cis-isomer of

110 (112) has been prepared recently by the ultraviolet irradiation of

111 this triene (112) likewise rearranges by first-order kinetics,^ but gives rise exclusively to trans-bicyclo [4.3.0] nona-2,4-diene (113).

110 111

57 58

111 hv

112 113

Essentially the same observations have been noted v;i.th 1,3,

5-cyclodecatrieue derivatives. Thus, Corey and Hor tanarm® ^ have

shovjn that triene 115, generated in a photoequilibration of 114,

25°

114 115 116

undergoes complete thermal conversion to 116 within 6 hours at 25°

in methanol solution. An identical mechanistic course has been 81 postulated for the photoconversion of isodehydrocholesterol (117) 82 to its 5-epimer (11.8) . in both instances, the high reactivity

of the intermediate conjugated cyclodecatrienes precluded their isolation. 59

H H

117 118

The dramatic stereospecificity of the above electrocyclic reactions has been interpreted recently by I'oodward and Hoffmann on the basis of orbital symmetry control daring the cyclization process. Such considerations predict, on the basis of extended

Huckel theory, that the energetically preferred mode in proceeding from triene to diene will be disrotatory as is observed.

The present investigation was initiated to explore the possi bility of preparing derivatives of 1,3,5-cyclodec.atriene which would be stable to rearrangement at ambient temperatures and above, by virtue of the presence of selected stabilizing substituents. A second goal which developed during the above study was the determina tion of the extent to which c is ,c is-3 , 5-cyclodccadienones undergo enolization and an elucidation of the probable stereochemical and steric factors which might control this equilibrium process.

Synthesis and Discussion

Consideration of possible substituents necessary to achieve stabilization of the 1,3,5-cyclodecatriene system led to the selection 60 of a dialkylamino group and a carboraethoxy group because of the appreciable delocalizntion of charge known to exist in structures 83 such as A. Hie selection further viac dictated in part by the

©0 © ^COCH 2 \ c=c/ 3 R2 \ . r.-^C0CH3 > \ very limited number of available synthetic approaches. For example, in view of the novel method available for the expansion of a carbocyclic ring by two carbon atoms under mild conditions, ' 8 4 it - 8 6was hoped that eight-membered dienemines^ likewise might condense with dimethyl- acetylencdicarboxylate to afford 1,3,5-fcyclodecatriene derivatives t>y a sequence involving the intermediate formation of a cyclobutene adduct. /ilthough reports of the reactions of dienamines derived from cyclic ketones with dicnophiles are generally lacking, it has been established that the cyclic dienamine 119 condenses vith dimethyl- acetylencdicarboxylate in Diels-Alder (4+2 x) fashion.®? Apparently,

co2ch 3

CH c c h S ^ 3 III c

co2ch3

119 however, this la tte r example is an exception to the normal enamine reaction pathway, for when l-dimethylamino-l,3~cyclooctadiene (120) 61 was treated with the same dier.ophile in the present study, the

1,3,5-cyclodecatviene derivative 121 was obtained rc-aclily. In

:o 2c h 3 n (c h 3'2 'N!CH3)2 -c o 2c h 3 + 25° 'C 0 2CH3 \ = / \ = H 120 co2c h3 B

N(CH N(CHo)

P t 0 2 ■COoCH

121 122

/ ~ v O r H! [ + II!

C02CH3{ 123

agreement with structure 121. the product exhibited pronounced ultraviolet absorption in ethanol and the requisite infrared and nmr absorptions (see Experimental). Further corroborative evidence for structure 121 was provided by its selective hydrogenation with pi atinura in ether solution to yield a dihydro derivative (1.22) similar in spectral properties to 123 which was prepared in an unequivocal manner from 1-pyrrolidinocyclooctene.

In similar fashion, treatment of 12 with dimethylacetylenedicarboxylate afforded the heterocyclic 1,3,5-cyclodecatriene 124. 88

This structure was established by the usual analytical and spectral data (see Experimental), and by partial hydrogenation to 125. The

N(CH COpCHo N(CH3 2 i A rC °2 CH3 + in c - > 1 0 -2 5 "C°2CH3 co2ch3 H

(ch3)2 (CH3)2 C02CH3 Y^-C02C h3 h2 Pt02 1\ ^ \ ^ ' C 0 2ch3 ^ N 3 o 2ch3

124 125 variety of accumulated data, when taken collectively, indicated that

124 was the desired medium-ring triene.

A remaining problem was the assignment of the proper geometric configurations to the olefinic bonds in the trienes 121 and 124.

Woodward and Hoffmann^ and Longuet-Higgins^ have discussed the two possible inodes of thermal ring opening of cyclobutene systems, namely conrotatory and disrotatory. In the conrotatory process (see D), 63

CH. CH-

n - c o 2c 2 h5

the substituents at Ci and C move in the same sense with respect x 4 to the ring to give a cis,trans-dicne as product. In the disrotatory example (E), the substituents move in opposite directions to afford cis.cis-diene. The thermal isomerization of cyclobutenes such as

D is usually conrotatory.However, steric factors such as the

five-membered ring in E render the conrotatory process highly un favorable and disrotatory opening is observed.^ The influence of a fused cis-cyclooctene ring^ on rotational thermal reorganization of intermediates such as B and C was unknown. Should the fused eight- membered ring exert little strain upon the transition state of the ring opening, then a conrotatory process would be anticipated; however, models indicate a somewhat-distorted extended-tub conformation

for the cis-cyclooctene portion of B and C which imparts significant rigidity to the overall molecular structure and suggested the

\ 64

* CH3I

N(CH3)2

B, X=CH2 121 » X=ch 2 C, X=0 124 , X=0 likelihood that steric control might promote disrotatory ring opening. Furthermore, Wraiding models of the three possible resulting triencs demonstrated the presence of significant transannular nonbonded interactions in the cis,tm ns,cis- and trails, c is , c.1 s -trie n e s; on the other hand, these effects are minimized in a relative sense in the cis, cis. cis-tri cite model. Tliis diminution of steric strain in the latter structure a priori might be expected to be of sufficient importance to promote a disrotatory ring opening.

Indeed, this latter mechanistic pathway was followed and the support ing experimental facts follow.

The ultraviolet spectral data of Table 3 clearly demonstrate that the extinction coefficients of the all cis-isomers do not parallel in magnitude those of their isomeric cis,trans-counterparts, lbst significantly, the emax values of 122.123, and 125 are roughly comparable to those of 126-128, the implication being that they must a ll be of the same geometry about the chroraophore. Given the fact that 126-128 most certainly resu lt from disrotatory ring opening of intermediate F,93 then it follows that they are of the cis,cis- 65

configuration. Consequently, 122. 123, and 125 must contain a similar cis.cis-diena grouping, and since the remaining olefinic linkage in 118 and 121_ must be cis,^ * ^ then the presence of an all-ci_s

configuration for these molecules is established.

Ri NR: c o 2c h 3 ✓^o 2ch3 + c 126 -128 + II! c (CH2)r b:o2CH3 ^02CH3 H

n = 0 to 2 T able 3

Ultraviolet Absorption Data for Some Cyclic Polyolefins

Compound A max, mu c. Solvent Reference

110 296 4,010 Cyclohexane a

112 290 2,050 Cyclohexane b cis,cis-l,3-cyclooctadiene 223 5,600 Cyclohexane c cis,trans-1,3-cyclooctadiene 230.5 2,600 Cyclohexane c cis,cis-1,3-cyclononadiene none>210 4,000(220 mu )Ether d cis,trans-1,3-cyclononadiene 213-220 7,300 Ether d ci3,cis-l,3,-cyclodecadiene none>210 3,900(220 mjj) Ether d cis,trans-1,3-cyclodecadiene 218-220 7,200 Ether d

122 319 13,400 Ethanol e

123 315 15,200 Ethanol e

125 329 12,040 Ethanol e

oO' T able 3 (C ontinued)

Ultraviolet Absorption Data for Some Cyclic Polyolefins

Compound Amax. tt/j Solvent Reference

334 9,760 Methanol

126

327 10,940 Methanol / ■ :o 2c h 3

^ c o 2c h 3

127

323 13,600 Ethanol

O' 128 aSee ref. 77. See ref. 78. CA. C. Cope and C.L. Bumgardner, J. Aa. Chen. Soc.. 78, 2812 (1956). ^G. J.. Fonken, private communication. eThis work. ?iee ref. 84. 68

With the elucidation of the complete structural formulas for 121 and

124, brief mention of their stability and reactivity is warranted.

Both substances are remarkably stable yellow solids v?hich do not decompose at their melting points. They have exhibited no propensity for valence-bond isomerization under the usual laboratory conditions; preliminary attempts at photoisomerization have proven inconclusive.

In methanolic hydrochloric acid, these cyclodecatrienes were readily hydrolyzed to the corresponding cis,cis~3,5-cyclodecadienones (129 and 130) which, hov;ever, exist predominantly as their enol (tricne) tautomers (see foJ.lov7ing section).

129 , X= CH2 130 , X=0

Studies of Keto-Enol Tautoner1am: Spectra

The cistcir,-3,5-cyclodecadienones 129 and 130, obtained from

121 and 124, respectively, in the manner described above, were found, upon cursory examination, to exist almost totally in the enolic form.

In view of the difficulty encountered by earlier workers in their

attempts to prepare similar s y s t e m s , these observations appeared 69

Interesting end v7ere investigated in greater detail.

For purposes oi: comparison, several 2.3-dicarboaethoxy-cis-

3-cycloalkenoncs of medium ring size, namely 131-135, structures in which the additional olefinic bond of 129 and 130 was absent, were desired. These P -ketoesters were prepared conveniently by the mild acid hydrolysis of the corresponding enamine derivatives and were obtained crystalline except in the case of 134. Whereas compounds 129-132. and 135 crystallized in the enol form, 133 solidified with some reluctance and then as the koto tautcmer.

On the other hand, 134 was obtained only as a viscous colorless oil which consisted of an equilibrium mixture of the tv?o tautomeric forms. These conclusions were derived from spectral evidence (see below) .

The characteristic infrared absorption bands found for this group of compounds are summarised in Table 4. The band assignments are patterned after, and confirm, those derived from a number of earlier studies.^ Specifically, the 1590-1610 cm."* absorption has been assigned to the diene or triene stretching vibrations, the 1655-

1660 cm.“* band to the enol chelate, and the absorption in the

1720-1725 cm.”* range to the carbonyl stretching vibrations of the unsaturated ester carbonyl groups. In those examples which displayed a significant concentration of keto tsutomer, an additional band was observed at 1765 cm.”*; this high frequency absorption has been attributed to the saturated ester carbonyl group present in the keto fora (G). The ketone carbonyl absorption which is likewise 70

•C-OCH- x ;02CH3 > B

present in such tautom.crs is believed to fa ll under the absorption

attributed earlier to the unsaturatcd carbornethoxy functions (1720*

1725 cra."^). This assignment appears reasonable in view of the

recorded carbonyl-stretching frequency for mesocyclic ketones^® when allowance is made for the expected ketone frequency shifts

in the medium ring -keto diester (Av~0-9 cmT"1) ^ The general

change in contour of this band in such cases also supports this

conclusion. T able 4

Infrared Absorptions of some 2.3-Dicarbonethoxy-cis.cis-3.5-cyclodecadienones

and cis-3-cycloalkenones

u Ketone and ^ Enol Unsaturated v Saturated VDiene, Chelate, Ester Carbo- Ester Carbo Compound nyls,i cm. -1 nyl , cm. cm„ cm, -1

1606(m) 1660(s) 17 25(s)

0 H

-C 0 2CH3 1610(m) 1655(s) 1725(s)

_ /^ S : o 2ch3 Table 4 (Continued)

Infrared Absorptions of some 2,3-Sicarbomethoxy-cis,cis -3,5-cyclodec* :dienones

and cis-3-cycloalkenones

^Ketone and Enol Unsaturated V Saturated vDiene, Chelate, Ester Carbo Ester Carbo- Compound cm. cm. -1 nyls, cm."*■ n y l. cm.

o 2c h 3 1725(s) 1765(m)

1606(w)a 1655(w) 17 20 (s) 1765(m) ^ u 02CH3 o 2c h3

'%5^ ° 2 Cli3i606(m) 1655(m) 17 25(s) 1765(m)

°2CH3

134 1610(m) 1665(s) 1725(s) COoCH

135 ro Table 4 (Continued)

Infrared Absorptions of some 2,3-Dicarbomethoxy-c is.cis-3.5-cyclodecadienones

and cis_-3-cycloalkencr.es

VKetone and v Enol Unsaturated 17 Saturated V Diene, Chelate, Ester Carbo- Ester Carbo- -1 Compound cm. cm. nyls. en.~^ nyl. cm.~1

1590(s) 1655(s) 1725(s)

.. T C0 2CH3 /cOoCH \ ^ y k o 2c H3

1660 ( s ) 17 25(s)

b m

This set of values was obtained after an equilibration period of 4 days at room temperature. Table 5

Ultraviolet Absorption Data of Some 2.3-Dicarbomethoxv-c ir.cis-3.5-cvclode-

cadienones and cls-3-cycloalkenones

Compound Solvent A max, mu

131 Cyclohexane 261 9,100 Ethanol 259 8,720 23.0 210 10,150 Acetonitrile 262 6,550 2 .6

132 Cyclohexane 256 10,520 Ethanol 257 10,380 70.5 213 11,300 Acetonitrile 258 7,720 2.7

133h Cyclohexane 254c 6,500 Ethanol 255 3,500 1 . 2 ca, 230 5,900 Acetonitrile 255 2,770 0.74

134 Cyclohexane 255° 5,330 Ethanol 256 5,130 26.0 ca. 225 7,500 Acetonitrile 263 2,480 0.87 245 1,550

-^i Table 5 (Continued)

Ultraviolet Absorption Data of Some 2.3-Dicarbomethoxv-cis,cis-3.5-cyclode

eadienones and cis-3 -cycloalkenonc s

Compound Solvent . mu e1 a

135 Cyclohexane 253 9,700 Ethanol 23S 9,700 Acetonitrile 255 6 , 0 0 0 1 . 6

129 Cyclohexane 11,900 254d Ethanol 26 5d 9,370 3.7 Acetonitrile 264c 6,950 1.4

130 Cyclohexane 243° 15,500 ---- Ethanol 248f 14,000 9.4 Acetonitrile 2483 12,400 4.0

a ^ enol/ e' 2 apparent extinction coefficient. K 2 apparent keto ratio Considerable fine structure was in evidence ill these spectra. dAlso a shoulder at 290 mju (5350). eAlso a shoulder at 290 m.fx (3500). ^Also a shoulder at 272 m(j. (10,300). "Also a shoulder at 263 (9300). ^After 12 hr. at room temperature 133 (still predominantly as 133a) displayed the following spectral properties: Scyclohexane 239 ny. (5640); \ EtOH 231 (6120) and 253 sh max Table 6

Chemical Shift Values ($ units) and Enol Contents of Some 2,3-Dicarboraethoxy-

cis.cis-3.5-cyclodecadienones and cis-3 cycloallcenones

Compound Solvent Proton Aa Bb C D % En

131 cci4 ----- 3.64 and 3.68 6.95 12.51 100 CSC13 ----- 3.72 7.08 12.60 100 Acetone-d_ 4.71 3.65 and 3.69 7.02 12.55 95 8 Benzene 3.47 and 3.57 7.02 12.98 100 DM£0 -dg 4.72 3.69 7.04 12.49 90

132 cci4 ----- 3.64 and 3.68 6 . SO 12.57 100

CDC13 ----- 3.70 6.93 1 2 .6 8 100

Acetone-dg 3.66 and 3.70 6.90 1 2 .6 6 100 Benzene 3.46 and 3.57 6.91 13.04 100 DMSO-d_5 3.69 6.90 12.60 100

133 CC14 4.64 3.69 and 3.79 6 . 2 0 12.48 46 CDC13 4.77 3.73 and 3.82 6.28 12.52 31 Acetone-d 4.73 3.70 and 3.80 6.30 30 _ s 12.61 Benzene 4.86 3.54 6 . 1 0 12.91 42 DMSO-dg 4.74 3.69 and 3.78 6.28 12.52 22

134 cci4 4.59 3.67 end 3.78 6.18 12.33 57 CDC13 4.73 3.74 and 3.S3 6.30 12.45 41 Acetone-ds 4.69 3.68 and 3.79 6.28 12.46 39 1 Benzene 4.81 3.51 and 3.53 6.08 12.85 57 DMS0-d 6 4.75 3.70 and 3.80 6.33 12.43 36 O' Table 6 (Continued)

Chemical Shift Values (

~~cis,cis-3.5-cyclodecadienones and cis-3-cycloalkenones~~

~~Compound Solvent Proton A 3° C 2 7. Enol~~

~~135 cci4 3.66 and 3.72 6 . 2 2 12.32 100 CDC13 4.55 3.74 and 3.75 6.30 12.40 92~~

~~Acetone-d 6 4.58 3.67 and 3.73 6.31 12.47 85 Benzene 4.58 3.44 and 3.51 6.18 12.82 95~~

~~129 cci4 ----- 3.66 and 3.71 d 12.81 100~~

~~C-CI3 3.73 d 12.87 100~~

~~Acetone-d 6 4.67 3.67 and 3.74 d 12.91 66 Benzene 3.42 and 3.46 d 13.24 100 DMSO-d 4.72 d G 3.69 12.53 43~~

~~130 CCI4 ---- 3.63 and 3.72 d 12.29 100~~

~~CBC13 ----- 3.73 and 3.77 d 12.40 100 Acetone-ds 3.66 and 3.76 d 12.45 100 Benzene ---- 3.46 and 3,51 d 12.76 100 SMSO-d 4.79 3.63 and 3.72 d 12.23 94~~

a b For proton designations, see structures G and K. Where only one value is given, the resonance peaks of the two methyl groups overlap. cBerchtold and U hlig^ have claimed that 131 exists in the enol form to the extent of approximately 60-707.; this value was obtained by integration of absorption peaks at & 12.54 and £ 5.73 (solvent unspecified). He have concluded that the absorption at 6 5.73 was perhaps an impurity in the sample of these. workers. aThe complexity of the resonance signals did not allow a specific assignment for in the presence of the two additional vinyl protons. 78

Table 5 summarizes the ultraviolet spectra of the keto diesters which were measured in cyclohexane, ethanol, and aceto n itrile. The region of maximum absorption is quite, solvent sensitive and is de creased markedly with increasing solvent polarity. Such evidence is in accord with the expected high enol content in hydrocarbon solvents with increasing ketonization as the polarity of the medium is enhanced. Because the absorption characteristics of enols are known to be almost totally independent of the particular 99 medium employed, and since the assumption can be made that the apparent extinction coefficients in cyclohexane are representative of the extinction coefficient of the enol tautomer, equilibrium constants (k-e) for the more polar solvents may be calculated and these values a.re shown in Table 5.

The nmr spectra of the tautomeric pairs were examined in sol vents of widely differing dielectric constant*^ in order to maximize alterations in the equilibrium induced by such polarity changes or by preferential hydrogen bonding between solvent molecules and one or the other of the tautomers (the internally hydrogen-bonded enol molecule is less polar than the keto species). The results are listed in Table 6 . Of the spectroscopic data presented heretofore, the nmr data are the most informative because a quantitative rather

~~than a qualitative measure of the keto-enol equilibrium is p r o v i d e d .1 0 2~~

All solutions were 2.5 M in substrate and were allowed to equilibrate at room temperature for four days.103 q>he signal at & 4.6-4 . 8 was assigned to the keto a -proton (absent in those systems which were 79

~~totally enollc) and the absorption in the 6 12.4-13.0 region, which~~

~~showed considerable variation in chemical sh ift, to the enol OH~~

~~proton. The per cent enol content was determined by integration~~

~~of the areas of these two absorptions. A representative set of~~

~~spectra is shown in Figure 1; the cis-3-cyclononenone derivative~~

~~133 was selected as the example because it displayed the greatest~~

~~change with time of a ll the systems studied. The absorptions with~~

~~chemical sh ifts of 6 3.7-3 . 8 can be assigned readily to the protons~~

~~of the carbomethoxy group. The remaining 6 6 .1-7.0 signal which~~

~~varied greatly in splitting and general line shape was attributed~~

~~to the vinyl proton (H^,); in addition to the fact that this proton~~

~~would be expected to resonate at lower magnetic field, the spin-~~

~~spin coupling of the Hq absorption should reflect certain first-~~

~~order and long-range coupling phenomena which would be dependent upon the keto-enol ratio, and such is observed.~~

~~Chemical shifts of the various proton groups in benzene solution were generally at higher magnetic field, except for the hydroxyl proton~~

~~of the enol tautomer which was shifted to lower field. This effect,~~

~~perhaps attributable to ring current effects arising from solute-~~

~~solvent complexes, has been observed previously for a number of~~

~~solvents.*04 AFTER DISSOLUTION~~

~~AFTER TWO DAYS AT R.T.~~

~~AFTER FOUR DAYS AT R T.~~

~~Figure 1. Partial nrar spectra of 2,3-Dicarbomethoxy-3-ci cyclononenone (133) at 60 Me. in 2.5 M carbon tetrachloride solution after various periods o equilibration at room temperature. 81~~

~~Conclusions~~

The data of Tables 5 and 6 clearly show that the kel'o-enol equilibria vary markedly with structure, Whereas the seven- and eight-membered cis-3-cycloalkenonos (131 and 132, respectively) are almost totally enollc throughout a broad spectrum of solvent polarity, the nine- (133) and teu-membered (134) congeners display marked decreases in the degree of enolization. A further interesting change occurs when an additional olefinic bond is placed in the ten-membored ring as seen by the increased enol content of 129 when compared to 134. An additional now factor is observed when one of the ring methylene groups is replaced by an ether oxygen atom; that such a modification results in an enhanced enol content is evident when 130 and 133 are compared with their carbocyclic analogs.

An understanding of the fundamental origin of these equilibrium differences demands that the conformations of the molecules under investigation be known with some degree of accuracy. This requirement is most formidable where medium-sized rings are concerned. Cyclic systems of eight to ten members are known to be quite flexible and to be subject to rapid low-energy conformational interconversions;^^ the problem is complicated further by the fact that the introduction of double bonds into such rings drastically changes their geometry and implants some elements of rigidity.In addition, substitution of an oxygen atom for a tetrahedral carbon atom has been claimed to 82

~~exert a negligible conformational influence on the rin g ,^^’ although~~

~~the. removal of some non-bonded interactions which accompanies such a~~

~~change might be expected to relieve I~strain^^ and transannular~~

~~steric compression. This latter effect is reflected quantitatively~~

~~in the high enol contents of 130 and 13^5.~~

~~As a guide to the probable conformational composition of the~~

~~tautomers in the above series, recourse was made to Dreiding models.~~

~~However, variation of the percentage composition could not be foretold~~

~~from a careful study of these models, and indeed an explanation of~~

~~the trend appears to be subtle. As the ring size increases, non-~~

~~bonded interactions and torsional strains are seen to increase~~

~~regularly and such effects probably are reflected in the observed~~

~~decrease in eriol content'*^ (reversed by the introduction of an~~

~~ether oxygen atom).~~

~~It is quite likely that the conjugated double bonds in these molecules cannot attain coplanarity for reasons of steric compression.~~

~~The increase in „ (see Table 5) when going from the seven- (131) QclX — to the eight-membered (132)rings, both of which are 100/C enolic~~

~~in non-polar solvents, suggests that in the somewhat more flexible~~

~~larger ring a greater degree of coplanarity can be achieved.~~

~~Unfortunately, this extrapolation cannot be carried to the next~~

~~larger rings because of the high keto content in these systems even~~

~~in cyclohexane solution. However, significant e increases are J max observed when 129 and 134 are compared to their oxygen analogs 130~~

~~and 135, respectively. Such enhanced absorption can be rationalized 83 on the basis of strain minimization accompanying the replacement of a~~

~~Cll^ group by 0 which permits a higher (but s t i l l incomplete) degree of coplanarity to the conjugated system.~~

~~Finally, the sizeable increase of enol tautomer when an additional double bond is introduced into the ten-membcred cycle ( i.e ., 134 as~~

compared to 129) may be explained by a substantial change in conformational composition during which a number of steric compressions, which formerly were prohibitive to enol formation, are relieved. PART VI

~~THE REACTION OF l-DIMETHYLAMINa-1,3-~~

~~CYCLOOCTADIEKE UJ.TH HE1'HANESULF01\TYL CHLORIDE~~

~~Introduction~~

The pronounced ab ility of sulfenes to add to electron rich olefins such as cnaraines by means of a cycloaddition pathway is now a well-known phenomenon. 109,110 j^ g ^ V7}ic;n an alkancsulfonyl chloride is treated with an enamine in the presence of triethylamine the formation of thietane 1,1-dioxides generally results. Several

~~NR- NRo CH3 SQ2Ct E t3 N~~

~~examples have been reported, however, v;herc straight chain sulfones , . , . 109i,1,111 are the products of this reaction.~~

~~From mechanistic considerations the reaction of 1-divnethylamino-~~

1.3-cyclooctadiene ( 1 2 0 ) with rr.othanesulfonyl chloride might be expected 11? to yield products of 1,4- * and 1,2-cycloaddition (136 and 13V) as well , 111,112 as a bis-adduct (138). The possibility of observing a

~~1.4-cycloaddition product and the fact that 137 would be a possible~~

~~84 85~~

~~120 136 137 138~~

~~precursor to the cyclooctatetraene derivative 139 stimulated an investigation of this reaction.~~

~~139.~~

~~Results~~

~~A solution of 120 and 1-2 equivalents of triethylamine in tetrahydrofuran at -10° was treated dropwise with a corresponding equivalence of methanesulfonyl chloride. Reaction workup (see~~

~~Experimental Section) and chromatography on neutral alumina,~~

F lo ris il, or silic a gel gave seven crystalline sulfone-containing products. The respective yields of these compounds were dependent upon the reaction temperature, the rate of addition of methanesulfonyl chloride, the quantity of methanesulfonyl chloride, the rate of elution through the chromatography column and the adsorbent used.

~~Table 7 shows the products obtained under various conditions. Table 7~~

~~Sulfones Formed Under Various Reaction Conditions~~

~~Special eqv. Compound number and yield Adsorbent Conditions ch3so ?ci 140 137 * 141 142 143 144 145~~

~~Alumina 1 3.37, 17 . 27. - - 0.87. 3.1%~~

~~Alumina 2 0.77. 10.27. 9.47. - 0.8% 8.3%~~

~~Alumina heat on steam bath with alumina before chrom 1 41.S% - - 0.7% 4.9%~~

~~Alumina 2 35.6% - - 0.5% 13.5%~~

~~Alumina Add CH3S02 Cl over 3 hr. 15.87. 14.77. -~~

~~Alumina Add CH3S02C1 at room temp. 1 2.4% - 5.6%~~

~~S ilica Gel 2 8.9% - 347. 3.3%~~

~~F lo risil 2 337. 7.5%~~

~~F lo risil elute immediately with ether 2 5.2% 35.8% - 5.37.~~

O'00 Structure Assignments. Sulfone 140 was known to be a mono- addition product frora its elemental analysis. The infrared spectrum of 140 exhibited peaks at 6.12 and 6.18p indicating an ertamine function and the ultraviolet spectrum f *sooct. 2 0 4 (9900) and L v max 286 mu (1350) showed extended conjugation. The nmr spectrum

~~(CCl^-TMS) displayed a ono-proton doublet of doublets (J ° 9.3 and 7.3~~

Hz) at 7.00 6 , a one-proton triplet**^ (J ~ 7.9 Hz) at 4.90 6 f a three-proton singlet at 2,78 6 (-SC^-CH^) , a six-proton singlet at 2c58 & corresponding to a dimethylenamino group and eight protons from 1.0-2.5 6 . This combination of data suggested structure

~~140. However, to eliminate from consideration similar structures~~

^ ~ so2ch3 in which the two vinyl protons would be adjacent, the nmr spectrum of 140 was decoupled. Irradiation of Ha (7.00 6 ) brought about no change in Hb (4.90 6 ), and irradiation of Hb did not change the appearance of Ha. Upon irradiation of the protons in the 2.0-2.16 region of the nmr spectrum both 11a and Hb were observed as broad singlets.

Sulfone 137 is apparently a precursor to several of the other reaction products and was obtained only upon rapid elution from the chromatography column. I t s structure assignment is based mainly on i t s nmr spectrum (CDCl^) which displayed a two~proton m ultiplet at

5.83^ for the two vinyl protons, a broad one-proton peak at 5.17 6 88 for He, a one-proton doublet of doublets (J ■ 14 and 3 Hz) at 4.32 6 for Ha, a one-proton doublet of doublets (J » 14 and 1 Hz) at 3.42 6 for Hb, a six-proton singlet at 2.426 for the dimethylamino group and eight protons as a multiplet from 1.1-2.36 . Structure 137 is

~~137 also indicated by the fact that it is converted to sulfone 142 upon standing in chloroform solution for too days.~~

The structure of sulfone 141 was revealed from its elemental analysis which indicated addition of and loss of its ultraviolet spectrum 246 rap (5860)J which showed max extended conjugation, and its nmr spectrum (CC)l^--TMS) which exhibited a two proton multiplet at 5.96 6 for two vinyl protons, a two-proton singlet at 4.20 6 for the protons a to the sulfone group, a four- proton multiplet from 2.2-2.7 6 for the allylic protons, and a four- proton multiplet from 1.4-2.16 .

~~141~~

~~Sulfone 142 proved to be a mono-adduct from i t s elemental analysis.~~

~~An eriamine function was indicated from the infrared spectrum ( \ CCl^ 89~~

6 .lip ) find the ultraviolet spectrum ^®tOh ^37 (5 7 7 0 ) L m a x The nmr spectrum (CDClg-TMS) consisted of a multiplet at 5.65 6 for two vinyl protons, a m ultiplet from -3.7~4.5 6 for four protons, a one proton peak at 2.92 6 , a six proton singlet at 2.716 for a dimethyl- enamine group and a six proton multiplet from 1.4-2.36 • Hydrolysis of 142 with hydrochloric acid gave sulfone 143 and, as previously mentioned, sulfone 137 was readily converted to sulfone 142.

~~This combination of data suggests the ten-membered heterocycle* shown below as the structure of 142. An examination of models of~~

~~NM&~~

~~142~~

~~142 shows that Ha is in the shielding cone of the double bond.~~

~~Therefore, the 2.926 peak in the nmr spectrum may be due to this proton.~~

Sulfone 143 was obtained only when the crude reaction mixture v;as chromatographed on F lo risil or silic a gel. Hydrolysis of 142 was also found to yield 143. A molecular formula of CglljyOjS was obtained from elemental analysis and a carbonyl function was indicated by the infrared spectrum (XHCCI 3 5.80(j). The 2,4-dinitrophenylhydrazone derivative of Ikj (rnp 219- 20°) was prepared as proof of a carbonyl function. The nmr spectrum (CDCI 3 -TMS) exhibited a two-proton H H multiplet at 5.75 6 (-C-C-), a two proton singlet at 4.216 90 0 , a two proton doublet (J * 8 Hs) at 3.806

~~(-S0 2 ”CH2 “^cC~) , a tv/o proton m l tipi et from 2.40-2.65 6 for the~~

~~allylic protons, and a six proton multiplet from 1.60-2.35 6 . Structure~~

~~143 was arrived at from this data.~~

~~Q 02~~

~~143~~

~~Sulfone 144, v;hich was obtained in very low yield, proved to be~~

~~a mono-adduct from i t s elemental analysis. The infrared spectrum~~

~~(HCClg) showed no characteristic absorptions for an enamine, carbonyl~~

~~or conjugation and the ultraviolet spectrum displayed only end~~

~~absorption. The nmr spectrum (CDClg, THS) exhibited a one-proton multiplet at 5.52 ^, a two-proton multiplet from 3.34-3.70 $, a~~

~~one-proton singlet at 3.20 5 , a one-proton doublet (J « 4 Ha) at~~

~~2.78 6 , a one-proton multiplet at 2.50 5 , a six-proton singlet at~~

~~2.20 6 and a seven-proton multiplet from 1,2-2,4 6 . Reaction of~~

~~144 with methyliodi.de in refluxing ethanol gave only starting material.~~

~~This compound is still under investigation and decoupling of the nmr~~

~~spectrum should be very helpful in assigning a structure for 144.~~

~~Sulfone 145 was obtained as bright yellow crystals; its~~

~~infrared spectrum (HCGl^) displayed a strong absorption at 6.48(J.~~

~~and its ultravi spectrum f } Et0H 218 (6580), 263 (6580) and L A iaax 352mp (13,200) conjugated1 chromophore. The~~

~~elemental analysis shewed that 145 \ias a bis-adduct. The nmr spectrum 91~~

~~(CDClyTi'S) exhibited a one-proton singlet at 7.74 6 , a six-proton~~

~~singlet at 3.30 6, a three-proton singlet at 3.02 6, a three-proton~~

~~singlet at 2.94 6 and eight protons at higher field. Sulfone 143 v/as hydrolysed with hydrochloric acid to give 146 which was found~~

~~NMe~~

~~SOoCH- 145~~

~~to be a mixture of the ketone and enol ferns. The infrared spectrum~~

~~(HCCl-,) of .146 exhibited a small peek at 5.89/j for the keto-form and~~

~~J> •S CH / -so2ch3 H~~

~~146 so2ch3 707. 307. a strong peak at 6.24 ;J for the enol-forrn. The nmr spectrum~~

~~(CDCl^-TMS) displayed a 0.7H peak at 10.61 6 for Ha and a 0.3H peak at 5.00 6 for Hb. The predominance of the enol form of 146 is not unexpected by analogy to the eight-membered ring carbomethoxy~~

~~compound studied in I'art V.~~

Mechanistic Conoiderations. It is obvious that ketone 143 arises by hydrolysis of 142 on the silica gel or Florisil column. Several mechanisms, however, can be postulated for the formation of 141 and 92

~~and 142, Sulfone 137 is apparently the precursor of both 141 and 142.~~

~~Loss of a proton from along with loss of the **N(CIIg ) 2 group would~~

~~result in the formation of 141. It is also possible to lose a proton~~

~~1^1~~

~~from C2 or Cg resulting in the formation of 147 or 1.48; migration of the double bond in either of these structures could again give~~

~~147 148~~

~~141. The formation of 142 con be visualized as a loss of a proton~~

~~from Cg together with opening of the thietane dioxide ring. In~~

~~either a concerted or step-wise process the carbanion thus formed~~

~~(149) could abstract a proton from the solvent. Another possible and more attractive mechanism involves participation of the lone-pair~~

~~149 93~~

~~of electrons on nitrogen in opening the thietane dioxide ring. The~~

~~a-sulfonyl carbmiion thus formed is situated in such a position~~

~~(see 150) that it could abstract a proton from across the ring~~

~~eventuating in the formation of 142.~~

~~142~~

~~150~~

~~Sulfone JL40 may be accounted for as shown below. The carbanion~~

~~140~~

151 152 formed upon attach of sulfene by the enatainc is situated (see 151) in a position where i t can abstract a proton ck to the dimethylira- oniura group through a six-membered transition state. Isomerization of the double bond in the enamine thus formed (152) would lead to

~~140.~~

The formation of sulfone 145 can be explained through abstraction of a proton from by the carbgnion as shown in 153. A repetition of S4 this process at of diencjnine 154 v?ould eventuate in the formation of 145.

~~©NM92 NMe-~~

~~145~~

~~X ^ S °2 CH3~~

~~153 154 EXPEPJJffiNX&L SECTION~~

~~Melting points are corrected and boiling points art uncorrected.~~

~~The microanalysis were performed by the Scandinavian Microanalytical~~

~~Laboratory, Herlev, Denmark: Galbraith Laboratories, Inc., Knoxville,~~

~~Tennessee, and by M-H-IJ Laboratories, Garden City, Michigan. Infrared~~

~~spectra v?ere determined v?ith a Perkin-Elmer Model 237 spectrometer.~~

~~Ultraviolet spectra v;ere recorded with a Cary Model 14 spectrometer.~~

~~The war spectra were determined v?ith a Varian A-60 spectrometer.~~

~~The mass spectra v?ere measured with an AE1 l'S-9 mass spectrometer at~~

~~an ionizing energy of 70 cv. The vapor phase chromatographic separations~~

~~and analysis v?ere carried out with an Aerograph A-700 gas chromatographic u nit.~~

~~95 P a rt I~~

~~ACID-PllOI'DTRD REARRANGEMENTS INVOLVING~~

~~TRANSANNULAR ETHER OXYGEN PARTICIPATION~~

~~5"Diniethylfira:Lno-7“OXci-l ,3-cyclooctadiene (11) . A solution of~~

~~18.0 g (0 .064 mole) of _10 in water was eluted through a column of~~

~~Amberlite IRA-400 in its basic form. The alkaline eluate was~~

~~evaporated in vacuo below 45°. The residual quaternary methohydroxide v/as decomposed by heating at 45-50° under a nitrogen atmosphere at~~

~~6 mm for 2 hr. The liquid vjhich formed was taken up in ether, dried~~

~~over anhydrous magnesium sulfate, and evaporated to give 7.2 g (73.2%) of yellow liquid, [X Et0H 227 mp(5330)}, max 7 ,8-Dihydro-N, N-dimethyl-2H-o:~~

~~Amberlite IRA-400 (hydroxide form). The eluate v/as collected until~~

~~it was no longer alkaline, at which point the major portion of the water v/as evaporated under reduced pressure. The residue v/as rapidly dis~~

~~tilled at approximately 15-60 nn and 90-150°. The organic product v/as rapidly separated from the liberated water by extraction with~~

~~ether and the ether layer v/as dried. The recovered oil was redis~~

~~tilled at 123° (20 mm) to give 12.8 g (787.) of dienamine I2t~~

1.5240; v ^C*4 1 5 9 0 cm"* (dienamine); \ *sooctane 277.5 (8,600); max max CC] 6 '4 ,ca.5.7 (multiplet, 2H, vinyl protons), 4.78 (doublet,J « 4.5 Hz, TMS vinyl proton at C-4), ca. 3.4 (multiplet, 2H, protons at C-8), and

~~ca. 2.2 (multiplet, 2H, allylic protons).~~

~~96 97~~

~~Anal. Calcd for C^H^NO: C,70.55; H, 9.87; N, 9.14. Found:~~

~~C, 70.44; H, 9.87; N, 9.16.~~

~~5-(3~Chloropropyl)dihydro-3(2H) -furanone (13a) . A. Hydrolysi s of 12.~~

~~To a warm solution of 1.5 g of 1_2 In 6 ml of concentrated hydrochloric~~

~~acid was added 6 ml of water. The resulting solution was heated on a~~

steam bath for 15 min, cooled, and extracted with ether. The combined organic layers were dried and evaporated and the residue was distilled to give 0.71 g (457.) of 13a as a colorless liquid, bp 100-101° (2.2 mm), n 25 1.4722; v CC14 1760 cm"1 ( ^G-'O) ; u max 6 ^ * 4 ca. 4.2 (complex m ultiplet, 1H, proton at C-5) , 3.80 and tms 3.87 (singlets, 1H each, protons at C-2), ca. 3.6 (multiplet, 2H,

~~-Cl^Cl) > 2.0-2,7 (series of six peaks, 2H, protons at C-4) , and~~

~~ca_._ 1.9 (multiplet, 4H, remaining methylene protons).~~

~~The 2,4-dinitrophonylhydrazone of 13a, prepared in the customary~~

~~fashion, was obtained as orange crystals from ethanol, mp 94-95°;~~

~~XEt0H 228 (15,200), 255 sh (11,000), and 356 m p (20,800). max Anal. Calcd for C^H^dN^O^: 45.55; H, 4.41; Cl, 10.35;~~

~~N, 16.35. Found: C,45.77; H, 4.69; Cl, 10.41; N, 16.33.~~

B. Rearrangement of 14. To a warm solution of 0.418 g (3.3 mmoles) of 14 in 2 ml of concentrated hydrochloric acid was added 2 ml of water. The resulting solution was heated on a steam bath for 10 min, cooled, and extracted with ether. The ether was removed and

the product was collected by preparative scale gas chromatography.^**

~~The clear liquid thus obtained (0.226 g) afforded an infrared spectrum and vpc retention time superimposable upon those of the sample obtained 98~~

~~in A.~~

romopropyl ) dihydro-3(2H) -furanone (13b) . A 2.0 g (13.0 mmoles) sample of _1_2 was hydrolyzed with 477. hydrobromic acid in the manner described above. There v/as obtained 1.66 g (61.5%) of 13b, bp 89-90° (0.18 mm); n£)29 1.4944; v CC14 1765 cm"1 (^C>0); max The nrar spectrum v/as very similar to that of 13a. The 2,4-dinitrophenylhydrazone melted at 107-108°.

~~Anal. Calcd for C7Kn Br02: C, 40.43; H, 3,90; N, 14.47;~~

~~Br, 20.64. Found: C, 40.28; 11, 4.02; N, 14.24; Br, 20.50.~~

~~7,8-Dihydro-2H-o:~~

~~Hydrolysis. A solution of 3.1 g (0.020 mole) of 1_2 in ml of~~

107. aqueous acetic acid v/as refluxed for 2.5 hr. The solution was cooled and extracted three times with ether. The ether solution was dried and evaporated, and the remaining oil v/as distilled to give 1.55 g (60.9%) of colorless lit]uid consisting of T4 (937.) 115 and JL_2 (77^), Preparative scale gas chromatography at 165° gave pure 14; v CC14 1720 cm"1 C=C-0) ; X isooctane 29Q (146^ max max 298 (185), 308 (161), and 318 mu (89); 6 CGl4 5.73 (complex TMS multiplet, 2H, vinyl protons), 3.92 (singlet, 2H, protons at C-2),

~~3.28 (doublet, J - 6.5 Hz, 2H, protons at C-4), and ca. 2.3 (multiplet,~~

~~2H, allylic protons).~~

~~Anal. Calcd for C, 66.64; H, 7.99. Found: C, 66.34;~~

~~II, 8.01. The 2,4-dinitrophenylhydrazone of 14^ prepared in the customary fashion, v/as obtained as orange crystals from ethanol, 99~~

~~mp 114-116°; X Et0H 228 (14,600), 255 sh (11,400), and 356 mu max ^ (21,600).~~

~~Anal. Calcd for C^H N C>5: C, 50.98; H, 4.61; N, 18.29.~~

~~Found: C, 51.06; H, 4.77; N, 18.50.~~

~~B. Fluoboric Acid Hydrolysis. A solution of 3.0 g (19.6~~

~~mmoles) of 1_2, 6 ml of 507. fluoboric acid and 15 ml of vrater was~~

~~refluxcd for 15 min, cooled, neutralized with sodium bicarbonate~~

~~and extracted with three 100 ml portions of ether. The ether layer~~

~~was dried, the ether removed and the remaining solution was distilled~~

~~to give 2.46 g (99%) of 1_4, as a colorless liquid, bp 85-87°~~

~~(15 mm), nD25 1.4838.~~

~~Oxocan-3-one (15). A solution of 2.0 g (15.9 mmoles) of 1_4~~

~~in 50 ml of diethyl ether was hydrogenated over palladium on~~

~~charcoal catalyst for 1.5 hr at 45 p.s.i. The catalyst was re moved by filtration, and the filtrate was carefully distilled to~~

~~give 1.95 g (96%) of _15 as a colorless liquid, bp 77-78° (11 mm) ,~~

~~nQ2^ 1.4588; l/*^4 1720 era * ( ^C-0) . The ketone was characterized max as its 2,4-dinitrophenylhydrazone, mp 174-175°, from ethanol.~~

~~Anal. Calcd for C, 50.64; H, 5.23; N, 18.18.~~

~~Found: C, 50.34; H, 5.34; N, 18.10.~~

~~7 .8-?)ihydro-4,4-dime thyl-2H-oxocin-3-one (24) . To a solution~~

~~of 1.51 g (0.012 inole) of 14 in 40 ml of absolute J:-butyl alcohol,~~

~~to which 1.40 g (0.036 g~atom) of potassium had been added, was~~

~~slowly added 10.2 g (0.072 mole) of iodomethane. The resulting~~

~~solution was stirred overnight, poured into 100 ml of water, and extracted with ether. The ether layer was dried and evaporated,~~

~~and the remaining liquid was distilled to give 1.35 g (73%) of~~

~~clear liquid, bp 89-90° (11 mm), n^25 1.4755; v ^ 4 1710 cm"1;~~

~~6 ^^4 ca. 5.4 (multiplet, 2H, vinyl protons), 4.06 (singlet, TMS 0 2H, -OCH^C-) , ca. 3.6 (m ultiplet, 211, -OCIL,-) > ca. 2.1 (m ultiplet,~~

~~2H, allylic protons), and 1.22 (singlet, 611, methyl groups).~~

~~The 2,4-dinitrophenylhydrazone of 24 was obtained as orange~~

~~crystals from 957. ethanol, mp 152-153°.~~

~~Anal. Calcd for C, 53.89; H, 5.43; N, 16.76.~~

~~Found: C, 53.80, H, 5.48; N, 16.65.~~

~~3.4.7 .8-Tctrahydro-2H-oxocin-3--ol (26) . To a slurry of 2.0 g~~

~~(0.053 mole) of lithium aluminum hydride in 50 ml of dry ether~~

~~was added dropwise a solution of 3.4 g (0.02.7 mole) of 14_ in 20 ml~~

~~of ether. The mixture v/as stirred for 4 hr and cooled in an ice~~

~~bath. Hydrolysis was achieved by the slow addition of 2.0 ml of~~

~~water, 2.0 ml of 30% sodium hydroxide solution, and 6.0 ml of water~~

~~in that order. The slurry v/as filtered and the residual solid was washed thoroughly with ether. The combined filtrates were evuporat~~

~~and the remaining liquid v/as distilled to give 3.4 g (987.) of 26,~~

~~bp 94-96° (18 mm); 6 ca. 5.8 (multiplet,.2H, vinyl protons),~~

~~4.29 (singlet, 1H, OH), ca. 3.6 (multiplet, 511), ca. 2.3 (multiplet~~

~~4H, allylic protons).~~

~~The a. -naphthylurethan of 26 was obtained as white crystals~~

~~from carbon tetrachloride-ether, mp 143.0-144.5°,~~

~~Anal. Calcd for CjgH^hT^: C, 72.70; H, 6.44; N, 4.71. 101~~

~~Found: C, 72.60; H, 6.45; N, 4.76.~~

~~3.4,7,8-Tetrahvdro-2H~oxocin-3-yl Erosylate (27). A solution~~

~~of 3.0 g (0.023 mole) of 26_ in 20 ml of pyridine was cooled in an~~

~~ice bath and added to a solution prepared by dissolving 1 2 . 0 g~~

(0.047 mole) of p-bromobenzcnesulfonyl chloride in 40 ml of cold pyridine. After 19 hr at 5° the excess brosyl chloride was hydrolyzed by addition of ice. Finally 100 ml of water was added and the mixture was extracted with two 100 ml portions of ether. The

~~ether layer was washed with 50 ml portions of iced IN hydrochloric acid u ntil the washings remained acidic. The ether layer was then washed with 50 ml of 5% sodium carbonate, dried over magnesium~~

~~sulfate and the ether was removed to give 7.57 g (93%) of white~~

~~solid, mp 74-77°. Recrystallization from ether-ligroin gave white crystals, rap 79-80°.~~

~~Anal. Calcd for ^BrO^S: C, 44.96; H, 4.35; S, 9.23.~~

~~Found: C, 44.99; H, 4.31; S, 9.19.~~

~~3-Oxabicyclo f5.1.01 octane (28). To a solution of 0.38 g~~

~~( 1 0 mmoles) of lithiina aluminum hydride in 15 ml of dry ether was added dropwise a solution of 1.45 g (4.2 mmoles) of brosylate _2£.~~

The resulting mixture was stirred under reflux for 14 hr and worked up in the previously described manner to give 0.40 g (85%) of clear liquid bp 75-76° (60 nan); 5 ^ ^ 4 2.9-4.2 (m ultiplet, 411, -CHo-O-CHo-) , TMS 1.3-2,4 (multiplet, 4K), 0.1-1.3 (multiplet, 4H, cyclopropyl protons).

~~Anal. Calcd for CyH^O: C, 74.95; H, 10.78. Found: C, 74.78;~~

~~H, 1 0 .7 7 . F a r t IX~~

~~OXQCANS: SYNTHESIS AND CORit)RKAXION A'L ISOMERIZATION~~

~~3-Dimethyleminooxocane (35) . A solution of 400 rag (2C6 mmoles)~~

~~of 12. in 40 ml of e t h y l acetate was hydrogenated at atmospheric~~

~~pressure in ethyl acetate solution over Adams' catalyst. Hydrogen~~

~~uptake ceased after the uptake of two moles. The solution was~~

~~filtered and evaporated and the residual liquid was distilled to~~

~~give 35 as a colorless liquid, bp 97-98° (16 ami), n ^ “* 1.4653.~~

~~The corresponding methiodide was prepared in the usual manner~~

~~and was obtained as white crystals from ethanol-ether, mp 147-143°.~~

~~Anal. Calcd for C 1 0 H2 2 IKO: C, 40.14; II, 7.41; N, 4.68.~~

~~Found: C, 40.19; H, 7*18; R, 4.55.~~

~~3,4,5,6-Tetrahydro-2H-oxocin (36). An aqueous solution of~~

~~15.0 g (0.050 mole) of the. above methiodide was eluted through a~~

~~column of Amberlite IItA-400 in its basic form. The alkaline eluate was concentrated under reduced pressure and the residue was distilled~~

~~(with elimination beginning at c.a. 150°) to give 1.87 g (33.47.) of~~

~~36 as a colorless liquid, bp 75-80° (15 mm) , n ^ 1,4612; v ^ ^ 4 max~~

~~1650 cm”'*' (vinyl ether); X end absorption; 6 ^f.^4 1.70 max TRS (broad absorption, 611, H-3,4,5), ca. 2.16 (raultiplet, 2H, H- 6 ) ,~~

~~ca. 380 (m ultiplet, 211, H-2), ca ^ 4.88 (multiplet, 111, H-7), and~~

~~6.00 (doublet "of doublets, J s 6 and 1 Hz, 1H, H- 8 ).~~

~~A sample purified by preparative gas chromatography (10' x V~~

~~stainless steel column packed with 157. Carbovax 20il on Chromosorb W,~~

~~60/80 mesh) at 118° and microredistillation was submitted for analysis.~~

~~102 103~~

~~Anal. Calcd for Cyll^O: C, 74.95; H, 10.78. Found: C, 75.03;~~

~~H, 1 0 .7 6 .~~

Oxocane (37). A. Catalytic Hydrogenation of 36. A solution of 850 mg (7.6 mmoles) of 3_6 in 100 ml of anhydrous ether, was hydrogenated at atmospheric pressure over Adams 1 catalyst. One molar equivalent of hydrogen was absorbed. The solution v/as filtered, and the ether v/as carefully removed. The remaining liquid v/as purified by preparative gas chromatography (same column as used for 36) and microredistillation to give 380 mg (447.) of oxocane, 1.4486;

~~5 CC1A i o62 (broad absorption, 1011, methylene protons) and 3.56 TMS (broad absorption, 4H, -CH^OCIL,-) «~~

~~Anal. Calcd for C?H 0: C, 73.63; H, 12.36. Found, C, 73.43;~~

~~H, 12.29.~~

~~B. Catalytic Hydrogenation of 38. The methiodide 38 vas prepared in the usual manner. From 6.0 g (0,039 mole) of 11^, there was obtained~~

~~9.8 g (90.5?;) of 38, mp 170-171°; \ Et0H 2 2 0 m ^ (17,100). max Anal. Calcd for C^H IKO: C, 40.69; II, 6.15. Found: C, 40.38;~~

~~H, 6.45.~~

~~A solution of 885 mg (3.0 mmoles) of 3_8 in 30 ml of methanol was hydrogenated at atmospheric pressure over Adams’ catalyst.~~

~~Approximately three molar equivalents of hydrogen were absorbed.~~

The solution was filtered and the methanol was carefully evaporated under reduced pressure. Ether was added to the residual oil and 330 mg of trimethylammonium iodide, mp•262-263°, v/as precipitated. The filtra te 104 was evaporated to give 150 mg of oxocane (37) which proved to display an infrared spectrum and vpc retention times identical V7ith those derived from, the sample of part A.

~~Low-temperature Nmr Spectra. These spectra were recorded at~~

Northwestern University with a Varian HU-60 instrument equipped with a low temperature probe. The oxocane spectra were obtained on approximately 107. solutions of 37_ in vinyl chloride with TMS as the internal standard. P a r t I I I

~~ACETOLYSIS OP OXOCAN"3-YL AND 3 ,4 ,7 ,8-TETRAHYBRO-2H-~~

~~0X0CIN-3-YL BRQSYLATES~~

Oxocan-3-ol (51). To a solution of 1,15 g (0.030 mole) of lithium aluminum hydride in 70 ml of dry ether v/as added dropv/ise a solution of 1.85 g (0,015 mole) of 1_5 in 30 ml of ether. The resulting mixture was stirred for 4 hr and worked up in the pre viously described manner to give 1.67 g (89%) of 5_1 as a clear liquid, bp 110-112° (19 mm), r ^ 29 1.4750; 6 ^ 4 4.43 (singlet,

~~1H,-OK) , cn^ 3.7 (m ultiplet, 511, £cH-DH) , 1.67 (broad peak, 8 H, ring methylene protons).~~

~~The Qj-naphthylurethari of 51_ was obtained as white crystals from ether, mp 1.12,5”113,5°.~~

~~• Anal. Calcd for Cl 8 H2 1 N03: C, 7 2.21; H, 7.07; N, 4.68.~~

~~Found: C, 72.32; H, 7.03; N, 4.77.~~

0xocan"3~yl Brosylate (48). A solution of 1.67 g (0.013 mole) of 51_ in 10 ml of pyridine was cooled in an ice bath and added to a cooled solution of 6.60 g (0.026 mole) of p-bromobensensulfonyl chloride in 20 ml of pyridine. After standing for 19 hr at 5° the reaction mixture was v/orked up in the usual manner to give 4.32 g

~~(96%) of white solid, mp 56-58°. Recrystallization from ether- ligroin gave v/hite crystals, mp 58.0-59.0°.~~

~~Preparative Scale Solvolysis of Oxocan-3-yl Brosylate (48).~~

~~A solution of 4.45 g (0.0127 mole) of 48, 55 ml of glacial acetic acid and 0.80 g (0.0085 mole) of sodium carbonate v/as stirred for~~

~~105 106~~

22 hr (10 half-lives) at 70°. The solution was cooled, 150 ml of ice water was added and the resulting mixture was extracted with three 100 ml portions of ether. The combined ether extracts were washed with saturated sodium bicarbonate solution u n til neutral and dried, and the ether was carefully removed. Distillation of the remaining liquid gave 0o34 g (23.67.) of 5J3, bp 52-53° (30 mm); HH 6 CCl/f 5.43 (multiplet, 2H, -C”C~) , 4.05 (narrow multiple!:, 2H, XUS

-0 -CH2 -CII») , 3.62 (m ultiplet, 2H, -0~CH2 -CH2-) , 2.45 (m ultiplet, 2H, allylic protons), ca. 1.62 (multiplet, 4H); 1.34 g (62.37.) of a mix ture of acetates, bp 66-67° (0.10 mm), and a distillation pot residue consisting of 0.35 g (7.97.) of brosylate 57, mp 78.0-79.0° (from ether-ligroin). The overall yield of isolated material v?as 93.87..

The acetate fraction v;as subjected to vpc analysis 44 c a and two peaks in the ratio of 81:19 were displayed. Preparative vpc separated these two components. The 817. peak had a vpc re tention time identical to that of an independently synthesized sample of 54. However, its nmr spectrum in CCl^ displayed two acetate methyl peaks at 2.02 and 2.05 6 indicating that two compounds of identi cal retention time were present. The 197, component was shown to be pure 56, v CC14 2760 and 1750 cm**1; 6 CC14 9.97 (triplet, J = 1.7 max TMS Hz, 1H, aldehyde proton), 4.14 (triplet, J = 6.5 Hz, 211, AcO~-CH2) ,

~~2.50 (triplet of doublets, J - 7.0 and 1.7 Hz, 2H, -C1I 2 -CH2 -C0H) ,~~

~~2.07 (singlet, SlIjCH^COg") and 1.20-190 (complex absorbtion, 8 H).~~

~~Anal. Calcd for CgHj^O^: C, 62.76; H, 9.36. Found: C, 62.48;~~

~~H, 9 .3 3 . 107~~

~~The 2,4~dinitrophenylhydrazone of 56_ was obtained as orange~~

~~crystals from ethanol-water, mp 77.0-77.5°.~~

~~Anal. Calcd for C, 51.13; H, 5.72; N, 15.90.~~

~~Found: C, 51.21; II, 5.83; K, 15.95.~~

~~Hydrogenation of 5. 6 ,7 ,8-Tetrahydro-2Ii-&?:ocin (53). A small~~

~~sample of ,53 in ether was hydrogenated over 107. palladium on~~

~~carbon catalyst at atmospheric pressure for 2 hr. The solution was filtered and the ether was carefully removed to give a compound Q with identical vpc retention time, * infrared and nrnr spectra to~~

~~those of an authentic sample of oxocane (37).~~

Oxocnn-3-yl Acetate (54). To a solution of 0.16 g (1.6 mmoles) of acetic anhydride end 0.13 g (1.6 mmoles) of pyridine cooled in an ice bath v/as added 0.15 g (1.15 mmoles) of alcohol 51. The resulting

~~solution was allowed to stand at room temperature for 18 hr; water and~~

ether were added and the ether layer was separated. The ether layer v/as v/ashed with dilute iced hydrochloric acid until neutral and once with water. The ether layer was dried over magnesium sulfate and the CC1 ether was removed to give 0.18 g. (927.) of 54, u 4 1730 cm”*; — max 6 TMS4 4‘85 (KultiPle t> 1H,-CH(0Ac)-), 3.62 (multiplet, 4H,

~~-CH2 -0-CH2-), 2.02 (singlet, 311, CH 3 ~C02») , 1.50-1.90 (multiplet,~~

~~811) . A pure sample v/as obtained by preparative vpc^a and submitted~~

~~for analysis.~~

~~Anal. Calcd for C^H^O^: C, 62.76; H, 9.36. Found: C, 62.75;~~

~~H, 9 .3 1 .~~

~~S 108~~

~~Reduction of Acetates 54 and 55 to Alcohols 51 and 58. To a~~

~~slurry of 1.13 g (0. 030 mole) of lithium aluminum hydride in 20 ml of dry ether was added dropwise a solution of 0.58 g (3.4 mmoles) of a mixture of J>4 and J55. The resulting mixture was stirred for~~

~~14 hr and worked up in the usual manner to give 0.43 g (97%) of a mixture of alcohols 51_ and 58. The two alcohols were separated by preparative vpc^!a and found to be present in the ratio of~~

~~40% (51) to 60%. (58). The 40% fraction was identical in all respects to a previously prepared sample of 51,. The 60% peak CC1 was assigned structure 58. 6 4 3.20-4.15 (complex m ultiplet,~~

~~6H,-CH2 -0-CH2 -CH0H"), ceu 1.61 (broad peak, 8 H).~~

~~Anal. of 58. Calcd for Cyll]^ 0 2 : C, 64.58; H, 10.84.~~

~~Found: C, 64.45; H, 10.92.~~

2-Oxepanylmethyl Brosylate (57). From 0.10 g (0.77 mmole) of 5j3 and 0.41 g (1.6 mmoles) of p-bronobenzencsulfonyl chloride in 3 ml of pyridine there v/as obtained 0.26 g (977.) of 57, as white crystals from ether-ligroin, mp 78.0-79.0°. The infrared spectrum of this compound was superimposible upon that of the brosylate isolated from the solvolysis of 48.

~~Cleavage of Brosylate 57 to Alcohol 53. A solution of 0.12 g~~

(0.34 mmole) of 57,, which was isolated from the solvolysis of oxocan-3-yl brosylate, in 2~inl of tetrahydrofuran under nitrogen was treated with 3 ml ( 6 eqv.) of a green solution of sodium naphthalene [prepared^ from 0.31 g (0.014 mole) of sodium and

1.80 g (0.015 mole) of naphthalene in 30 ml of tetrahydrofuran] 109 added by means of a syringe. A few drops of water and a spatula of magnesium sulfate were added, the. solution was filtered the solvent was removed and the resulting alcohol was collected by preparative vpc.^a The alcohol thus obtained was identical to

~~58.~~

2-Hethyloxepane (59) from 58. To a solution of 1.04 g (4.0 mmoles) of p-bromobcnzencsulfonyl chloride in 2 ml of pyridine at 0 ° was added 0,25 g (1.9 mmoles) of a mixture of alcohols from the solvolysis of brosylate 48. The resulting solution was allowed to stand overnight at 5° and worked up in the usual manner to give

0.48 g of a mixture of brosylat.es. The crude brosylate mixture was treated with 0.38 g (10 mmoles) of lithium aluminum hydride in ether and stirred for 1 2 hr. V/orkup in the usual manner and re moval of solvent gave 0.13 g (607.) of a mixture of reduction products.

~~Two major peaks were observed upon vpc. analysis ^ * 3 of this mixture.~~

~~The high retention time peak was found to be oxocane QD and the low retention time peak was found to be 2-methyloxepanc (59) ;~~

~~6 CCl4 3.30-3.80 (complex m ultiplet, 3H, -CVU-O-CU-), cfu 1.60 IMS (broad peak, 8 H, ring methylene protons), 1.10 (doublet, J n. 6.5 Hz,~~

~~3H, H-C-CH3).~~

~~Anal. Calcd for C 7 H1 4 0: C, 73.63; H, 12.36. Found: C, 73.33;~~

~~H, 12.30. 110~~

Preparation and Hydrogenation of 2-Methyloxepinc .^ To a suspension of 1.50 g (0,16 mole) of 2,5-dihydrotoluene, 75 ml of methylene chloride and 17.2 g (0.21 mole) and anhydrous sodium acetate vas slowly added (one hr) 27.6 g (0.15 mole) of 407. peracetic acid. The solution v/as stirred for an additional 3 hr, cooled in an ice bath to precipitate the salt formed and filtered.

The organic layer was washed twice with 107c sodium hydroxide solution, once with water, dried over magnesium sulfate and the methylene chloride was removed. Distillation of the remaining oil through a 1 ft. packed column gave 10.5 g (607>) of 1,2-epoxy-

~~1-methylcyclohex-4~enc, bp 60-04° (27 mm) . Redistillation gave~~

~~99-t-T, pure compound, bp 54° (17 mm), 1.4642. [ L i t . ^ bp 42-43°~~

~~(12 mm), n - ,2 0 1.4668] .~~

~~To a solution of 5.8 g (0.053 mole) of 1,2-opoxy-l-methylcyclohex-4-ene in 100 ml of methylene chloride at -78° vas added dropwise 8.5 g (0.053 mole) of bromine in 30 ml of methylene chloride.~~

~~The methylene chloride was removed and the remaining oil was kept at 0.10 mm at room temperature for 2 hr to give 14.0 g (987=.) of~~

~~1,2-epoxy-l-methyl-4,5-dibromocyclohexane. The crude dibromide~~

~~(6 . 0 g, 0 . 0 0 2 mole) was then added to a solution of sodium methoxide~~

(from 1.60 g, 0.070 mole of sodium) in 30 ml of ether and refluxed for 3 hr. The resulting solution was cooled, filtered and 100 ml of ice water was added. The layers were separated, the aqueous layer was washed with pentane and the combined pentane-ether solution was washed with water and dried over magnesium sulfate. The organic I l l

~~solution was then placed in a Parr Hydrogenator along with 2 ml of~~

~~triethylamine and 0.10 g of platinum oxide. After 11 hr the~~

~~catalyst was removed by filtration, the solvent was carefully~~

~~removed and the remaining liquid was subjected to preparative vpc^^ which gave 0,33 g (137.) of 2-methyloxepane, identical in all res~~

~~pects to the previously prepared 59.~~

~~Solvolysis of 2-0xepanylmcthyl Brosylate (57). A solution~~

~~of 0.21 g (0.60 mmole) of brosylate 57_t 0.043 g (0.40 mmoles) of~~

~~sodium carbonate and 3 ml of glacial acetic acid was heated at~~

~~110° for 65 hr in a sealed tube. The reaction mixture was worked up in the previously described manner to give 0.09 g (887.) of an~~

~~acetate mixture. Since the expected acetates’"54 and 55_ were known~~

~~to have the same vpc retention time; the acetate mixture was~~

~~treated with lithium aluminum hydride to give the corresponding~~

~~alcohols. Vpc analysis of the alcohol mixture showed 567. J38,~~

~~237. 51_ and 2171 of an unknown alcohol. Samples of 51^ and 58 were obtained by preparative vpc^fa and their infrared spectra were~~

~~identical to previously prepared samples.~~

~~Oxocan-3-ol-3-d ( 6 6 ) . To a slurry of 0.36 g (9.0 mmoles) of~~

~~lithium aluminum deuteride in 30 ml of dry ether was added 2.0 g~~

~~(15.6 mmoles) of ketone 1^5 in 20 ml of ether. After stirring for 5~~

~~hr the reaction mixture was v;orked up in the previously described manner to give 2.0 g (997.) of 6 6 , n 2 5 1.4755; 5 CC14 3.96 D TMS (singlet, 1H, alcohol proton), ca. 3.5 (multiplet, 4H, ,~~

~~1.61 (broad peak, 8 H). - 112~~

~~Oxocan-3“Vl-3-d Brosylate (65). The brosylate of 6 6 _ (65) was prepared in the usual manner. From 1,95 g (14.9 mmoles) of 6 6 and~~

7.65 g (30 nmoles) of p-bromobenzcnesulfonyl chloride there v;ac obtained 5.15 g (997a) of 6^5 as white crystals from ether-ligroin, mp 57.5-58.5°; 6 ® ^3 7.71 (sharp peak, 4H, aromatic protons), 3.60 TMS (sharp peak, 4H,-C^-O-CH^**), ca. 1 . 6 (multiplet, 8 H)C

Solvolysis of Brosylate 65. A 4.0 g sample of 65 was solvolyzed in the same manner as its non-deuterated analog (48). After reaction workup the acetates were converted to their corresponding alcohols with lithium aluminum hydride and separated l y preparative vpc/^a

~~A sample of 6 6 from the solvolysis had identical spectral properties~~

~~to an authentic sample. This alcohol was converted to its corresponding acetate 6_8 with acetic anhydride av*d pyridine in~~

~~847, yield. The nrnr spectrum of 6j3 displayed no peak in the 4.8 6 region.~~

~~Preparative Scale Colvolysis of 3,4,7 ,8-Tetrahydro-2II-oxocin-~~

~~3-yl Brosylate (27).~~

~~A solution of 4.75 g (0.013 mole) of 2J_, 60 ml of glacial acetic acid and 0.95 g (0.00S0 mole) of sodium carbonate was stirred for~~

2 hr 50min (10 half-lives) at 70°. The solution was cooled, 150 ml of ice water was added and the resulting mixture v/as extracted with three 100 ml protions of ether. The combined ether extracts were washed with saturated sodium bicarbonate solution until neutral, dried over magnesium sulfate and the ether was carefully removed.

~~Distillation of the remaining liquid gave 2.32 g (99.5%) of a mixture of acetates, bp 82-84° (0.8 mm). Analysis by vpc^a showed~~

three peaks in the ratio of 74.5/2.2/23.3. Preparative vpc^**

separated these three peaks and the 2.27. component was found to have identical infrared spectrum and vpc retexition times^a_c to those of an autheivtic sample of 69. The 23.37. component (an impurity consisting of approximately 1% yield based on starting brosylate 2 1 was present in this vpc^ cut) was found to be the exo-cyclopropyl CCX CC1 acetate 7_2, 1740 cm"*; 6 5.28 (m ultiplet, 1H, CH-OAc) ,

~~3.4-4.1 (multiplet, 411, ~ C112 - 0 - CH. ^ - ) , 2.01 (singlet, 3H, CH. 3 C02“) , ca. 1.7 (m ultiplet, 211), 0.28-1.32 (approximately 20 sharp peaks,~~

4H, cyclopropyl protons). The 74.57. component ( two sr.’.al 1 er com pounds ware found to be present in this fraction upon treatment with lithium aluminum hydride) v/as assigned structure 71_, the endo- cyclopropyl acetate, 1740 cm 6 £^4 4.95 (multiplet,

~~1H, CH-OAc), 3.3-4.2 (nrultiplet, 411, -CH2 -0"CH2~) , 2.06 (singlet, 311,~~

~~CH3 C0 2 “) , 1.4-2.0 (m ultiplet, 211), 0.52-1.43 (complex m ultiplet, 4H, cyclopropyl protons).~~

3,4.7,8-Tetrahydro-2H-oxocln-3"yl Acetate (69). To a solution of 0.30 g (3.0 mmoles) of acetic anhydride and 0.2.4 g (3.0 mmoles) of pyridine cooled in an ice bath was added 0.30 g (2.3 mmoles) of alcohol _2(S. The resulting solution was allowed to stand at room temperature for 18 hr and worked up in the previously described manner to give 0.38 g (96%) of 69, p CC*4 1730 cm"*; 6 ca. max iiiS 5.8 (multiplet, 2H, vinyl protons), 4.87 (multipletj 1H, CH-OAc),

~~3.2-4.1 (complex m ultiplet, AHj-C^-O-CIL,-) , 2.1-2.9 (m ultiplet, 4H, 114 a lly lic ), and 2.03 (singlet, 3H, CH-jCQ^*).~~

~~Anal. Calcd for CqH^O^: C, 63.51; H, 8.29. Found: C, 63.42;~~

~~H, 8.35.~~

~~Conversion of Acetates 70 and 71 to Alcohols 70a and 71a.~~

~~To a mixture of 1.S0 g (50 mmoles) of lithium aluminum hydride in~~

30 ml of dry ether v/as added dropv/ise a solution of 0.9S g (5.8 mmoles) of the 74.57. product from the solvolysis of 27 in 10 ml of ether. The resulting mixture v?as stirred for 3 hr and worked up in the usual manner to give 0.73 g (99%) of a mixture of alcohols. Analysis of the mixture by vpc^k showed three peaks which when based on starting brosylate 27 amounted to 1.5%, 4.17. and

~~68.97.. The 1.5% product could not be isolated in quantities sufficient for characterizations. The 4.17. product was assigned structure 70a;~~

~~6 5.59 (narrow multiplet, 2H, vinyl protons), 3.4-4.3~~

(multiplet, 6 H,-CII^-O-CI^-CHOH-) , ca. 2.6 (m ultiplet, 2H, a lly lic protons), ca. 1.8 (multiplet, 2H). A small sample of this 4.1% product v/as hydrogenated over platinum oxide and gave a compound with very similar infrared and nrnr spectra to those of 51.

~~The 68.9% component was assigned structure 71a; 6 ^*4 2.95-4.6 TMS~~

~~(complex m ultiplet, 6 H), 1.55-2.10 (multiplet, 2H), 0.33-1.40~~

~~(multiplet, 4H, cyclopropyl protons). A sample of 71a v/as converted back to acetate 2 JL previously described manner and submitted to analysis.~~

~~Anal. Calcd for : C, 63,28; H, 8.32. Found: C, 63.51;~~

~~H, 8 .2 9 . 115~~

~~Oxidation of a Mixture of Bxo and Endo 3-Oxabicyclo [5.1.0]~~

~~octan-3-ol (71a and 72a) to Ketone 74. To a solution of 1.50 g~~

~~(11.7 mmol es) of a wdxture of 71a and 72a in 1 0 0 ml of acetone~~

~~AO was added dropwise with stirrin g a solution of Jones Reagent~~

~~(prepared from 26.7 g of chromium trioxide in 23 ml of concentrated~~

~~sulfuric acid and diluted to 1 0 0 ml with water) until a yellow~~

~~color persisted. The resulting mixture was stirred for 5 min and~~

~~isopropyl alcohol was added until the yellov? color had disappeared.~~

~~The mixture was filtered, the filtrate was washed with acetone,~~

~~the organic layers were combined and the acetone was removed. Ether~~

~~( 1 0 0 nl) was added and the resulting solution was washed with~~

~~saturated sodium bicarbonate solution, dried over magnesium sulfate CC1 and the ether removed to give 1.10 g (757.) of 74, v 4 1720 and~~

1695 cm 6 approximately 40 sharp peaks from 0.9-4.6. Xi'iO Oxidation of 71a to 74. A vpc/:^a pure sample of 71a was oxidized in the manner described above and gave a product with identical spectral properties to those of the previously prepared

~~74.~~

~~The 2,4-dinitrophenylhydrazone of 74. was obtained as red-orange~~

~~crystals from 95% ethyl alcohol, mp 189-191°.~~

~~Anal. Calcd for C, 50.98; H, 4.61; N, 18.29.~~

~~Found: C, 50.65; H, 4.71; N, 18.32.~~

~~Reduction of 74 to 71a and 72a. To a solution of 0.61 g (16 mmoles) of lithium aluminum hydride in 50 ml of dry ether was added dropwise a solution of 1.10 g (8,75 mmoles) of ketone 74. After 116~~

~~stirring for 3 hr the reaction mixture was v/orked up in the usual~~

~~manner to give 1.11 g (99%) of a mixture of alcohols. Since alcohols~~

~~71a and 72a had identical vpc retention tim e s^ a"c th is mixture of~~

~~71a and 72a vzas converted to acetates 7_1 and 75! with acetic anhydride~~

~~and pyridine in 8S.4% yield. Vpc analysis^3 of this mixture showed~~

~~41% of 7J, and 59% of 72..~~

~~7.8”Sihydro-2H--o::oclri:~3(6iO -one ( 16) . A solution of 2.0 g~~

~~(15.6 mmoles) of fi, y -ketone 14., 1.72 g (10 mmoles) of anhydrous~~

~~p-tolueueculfonic acid and 40 ml of dry benzene v/as heated at reflux~~

~~for 15 min, cooled and poured into 100 ml of saturated Ki-^COj solution.~~

~~The layers were separated, the organic layer was dried over magnesium~~

~~sulfate and the benzene v/as removed. Distillation, gave 1.1 g of clear~~

~~liquid, bp 82-£4° (11 ran). This liquid v/as subjected to the vpc^'c analysis which showed an 84.8% peak corresponding to 14. and a 15.2% peak for 16. Preparative vpc gave an 80%. pure sample of 16,~~

~~V 1680 cm"*; \ 229,5 1/1^,(8000). Thee max v/as corrected~~

~~for purity.~~

~~Hydrogenation of 16 to 15. A solution of 0.034 g (0.26 nmoles) of 16., 0.01 g of 10% palladium on carbon and 30 ml of ether v/as~~

~~shaken in a Parr Hydrogenator for 5 hr under 40 p .s .i. of hydrogen.~~

~~The catalyst v/as removed by filtration and the ether was carefully evaporated to give 0.030 g (88%.)of oxoc.an-3-one (15) .~~

~~3,6,7,8-Tetrahydro~2H-oxocin-3-ol (73). To a slurry of 0.32 g~~

~~(8 .45 nmoles) of lithium aluminum hydride in 40 ml of dry ether cooled in an ice bath was added dropwise a solution of 1.05 g (8.20 mmoles) of an equilibrium mixture of ^4 and 1_6. After stirring~~

~~for 5 min at 0° the reaction mixture v/as v/orked up to give 1.00 g~~

~~(93.57=) of a mixture (85.2% of _26 and 14.87. of 73) of alcohols. A~~

~~pure sample of 7_3 was obtained by preparative vpc^a; 6 5.62 ii’lb (multiplet, 2H, vinyl protons), ca. 4.5 (multiplet, 1H,-CH(OII)-),~~

~~4.33 (singlet, 111, alcohol proton), 2.9-4.0 (multiplet,~~

~~CH2-) , 1.3-2.5 (complex m ultiplet, 411, a lly lic and ring methylene protons) .~~

~~Anal . Calcd for C, 65.59; 11,9.44 „ Found: C, 65.30;~~

~~H, 9.54 .~~

~~Acetolysis Kinetics of 27 and 48~~

~~Reagents. Anhydrous acetic acid v/as prepared by refluxing a~~

~~solution of acetic anhydride in glacial acetic acid for 24 hrs and~~

subsequent fractional distillation in a dry atmosphere. Perchloric acid (ca. 707.) was standardised against anhydrous sodium carbonate which had been heated over an open flame and allowed to cool in a desiccator. Brorn cresol green v/as used as an indicator. Standard perchloric acid in acetic acid (ca. 0.015 M) used in titrating acetolysis aliquots was prepared by the addition of an accurately weighed amount of standard 707. perchloric acid to a known volume of anhydrous acetic acid. Standard sodium acetate in acetic acid

(ca. 0.03 1J) v/as prepared by the addition of anhydrous acetic acid to anhydrous sodium carbonate; the water of neutralization was not re moved, The sodium acetate in acetic acid v/as standardized against perchloric acid in acetic acid using bromophenol blue as the indicator 118

~~the color change is from yellow to colorless. The automatic pipets were calibrated with glacial acetic acid; the density of acetic acid~~

~~is obtained from the following formula: d s 1.0724 [ -1.1229 X T°C ] .~~

~~Kinetics procedure. A ca. 0.02 M solution of brosylate~~

(accurately weighed) in acetic acid-sodium acetate was prepared in a 25 ml volumetric flask. Aliquots of this solution (ca. 2.2 ml) were removed and sealed in glass ampoules. The ampoules were placed in a constant temperature bath and after 10 min the first ampoule was removed and quenched in ice water. At this point an accurate timer was started. The ampoule was then placed in a beaker of water at room temperature for ca. 4 min, whereupon exactly 1.925 ml. of

solution vms removed from the ampoule, by means of an automatic pipette and immediately titrated with standard perchloric acid in acetic acid. Two drops of saturated bromophenol blue in acetic acid was used as the indicator and the end point was considered to be reached when the yellow solution turned clear. The remaining ampoules were removed at appropriately timed intervals, immediately quenched in ice water and titrated as previously described.

~~The rate of solvolysis of brosylate 27 was determined utilizing an infinity titer. The infinity titer was taken after~~

20 half-lives. The rate of solvolysis of brosylate 48 was determined using a theoretical infinity titer. The theoretical infinity point was necessary because it was found that the rate of solvolysis of 4j5 decreased during the course of the reaction. This fact was due to internal return of the brosylate anion with formation of a more stable 119

~~brosylate.~~

~~Due to the decreasing rate of solvolysis of 48 it v/as necessary~~

~~to use only those points taken, during the first 0.3 half-life of the~~

~~reaction for calculations of the f ir s t order rate constant. During~~

~~this time good first order plots were obtained.~~

~~Determination of a solvolytic rate constant. The determination~~

~~of a solvolytic rate constant is accomplished by the completion of~~

~~Table 8. The table shovm was used for brosylate _27_ at 50°. The~~

~~titrsnt was 0.0154-5 M perchloric acid in acetic acid and 1.925 ml~~

~~aliquots were titrated. Column (1) lists the time at which the ampoules were placed in the ice bath. Column (2) l i s t s the time~~

~~in seconds. Column (3) l i s t s the volume of titra n t. Column (4)~~

~~lists the molarity of sodium acetate in each aliquot. Column (5)~~

~~lists the acid generated since the first aliquot, determined by~~

~~substracting each entry in column (4) from the first entry. Column~~

~~(6) lists the molarity of brosylate in each aliquot, determined by~~

~~subtracting each entry in column (5) from the last entry in column~~

~~(5). Column (7) lists the logarithm of each entry in column (6).~~

~~The infinity titer, the last entry in column (3), is taken after at least 20 half-lives. The half-life is determined from the relationship~~

~~ty. - (0.693/k).~~

~~When an infinity titer cannot be obtained a theoretical~~

~~end point must be used. The theoretical infinity concentration of~~

~~sodium acetate is obtained by subtracting the original molarity of brosylate frora the original molarity of sodium acetate. 120~~

~~Table 8. Examp1 c of a Solvolytic Rate Calculation~~

~~(1) (2) (3) (4) (5) (6) (7) sec. HC10/ Acid Time x id3 (ml) NNaOAc gen. log M~~

~~1:40 0 3.525 .02829 0 .02233 1.3489~~

~~2:00 1.20 3.265 .02620 .00209 .02024 1.3062~~

~~2:20 2.40 3.034 .02435 .00394 .03 813 1.2646~~

~~2:45 3.90 2.769 .02222 .00607 .01626 1.2111~~

~~3:15 5.70 2.497 .02004 .00825 .01408 1.1486~~

~~3:45 7.50 2.255 .01810 .01019 .01214 1.0342~~

~~4:05 8.70 2.115 .01697 .01132 .01101 Ti-0418~~

~~4:25 9.90 1.992 .01599 .01230 .01003 1.0013~~

~~4:45 11.10 1.876 .01506 .01323 .00910 0.9590~~

~~Infin. 0.742 .00596 .02233~~

~~The specific rate constant for a first-order reaction is experimentally determined by the equation~~

log M = - (k/2o303)t + const. (1) where M is the molarity of brosylate at time t and k is the. specific rate constant. A plot of log M vs. t should be linear with a slope of -lc/2.303. The values in columns (2) and (7) are used for this plot.

~~The method of least squares can be used to obtain the best slope to fit a linear plot. According to this method the slope, m, of an equation of the form~~

~~y = mx -<• c (2) where y and x are variables and c is a constant may be determined by~~

~~the equation~~

~~m £ xy - n_____ £ ( x 2) - — ^ n In this equation x and y are specific values of various sets and~~

~~n is the number of such sets. A comparison of equations (1) and~~

~~(2) shows that~~

~~y c log II x « t; m ~ -k/2.303~~

~~From the procedure previously described, and using the method of~~

~~least squares to determine the slopes, the rate constants of brosylates~~

~~27 and A8 were obtained at various temperatures. The error included~~

~~v/ith the rate constants is an average deviation for two runs.~~

~~Table 9. Solvolytic Rate Data Experimentally Determined for 27 and 48.~~

~~Temp.itu up. i.at-eRate 1 H il S"' Compound °C. J . S O C . ) (kca l/molc)____(e.u.)~~

~~39.80° (2.20 ±.03)xl0“5 50.75° (8.13 ±.03)>:10-5 27 63.90° (3.41± .02)xl0-A 23.2±.2 -5.61± .77 70.00°(calc) 6,73 x 10"4~~

~~)Bs 39.80° (2.24±.04)xl0“6 51.00° (1.01±.01)xl0"5 h8 63.90° (4.55±.06)xl0“5 25.5±.3 -2.97±1.0 70.00°(calc) 8.77 x 10“5~~

~~Activation parameters. Tire values of the activation parameters~~

~~AH* and are calculated by means of the equation~~

~~kr = (kT/h) e /& e" (3) 122~~

~~where k r specific rate constant~~

~~~ k “ Boltzmann constant (1.380 x 10“^ )~~

~~h » Planck's constant (6,62.4 x 10"'^)~~

~~R " gas constant (1.987)~~

~~T B temperature (°K)~~

~~Dividing (3) by T and taking the logarithm gives~~

~~log (kr /T) s log (lt/h) + a S*/2.303I1 - a K^/2.303RT (4)~~

~~- 10.31876 + A S*/4.576-Alf*74.576T (5)~~

~~I t is seen iron (5) that a plot of log (k /T) vs. (1/T) should b e linear and have a slope of -A ll /4.576. Using the least squares~~

~~treatment the best slope to fit this linear plot can be determined.~~

~~From equation (3) it follov/s that~~

~~y ~ log (kr /T)~~

~~x = 1/T i m * -All74.576 i The values of All are calculated accordingly and are listed in~~

~~Table 9. Since the maximum fractional error in the specific rate constant (p.) is very small, the following relationship as described 1 1 f\ 4- by Uiberg11D affords the error in ill ( 6).~~

~~6 b 2R ■— ^_____ g (0) T* T . Furthermore, rearranging (5) gives an equation for AS :~~

~~AS^ = [log (kr /T)- 10.31876] 4.576 + a Ii^/T (7) 123 i + To obtain a value for a S , the calculated value of A II r is substitute*:~~

~~into equation (7) at the three different temperatures. The error in~~

~~A (a) is linearly related to that of a H^ and thus is calculated~~

~~by equation (8) .~~

~~a z 6 (1/T) 'I- aR i/T + I!-"-! ( 8) 2TT i An average of the three values of a S for each brosylate is also~~

~~listed in Table 9.~~

~~The following rearrangement of equation (5) permits the~~

~~calculation of the specific rate constant at any new temperature~~

~~from the activation parameters of the particular reaction.~~

~~i i A ll- T a ST (9) log kr = 10.31876 + log T 4 .576T~~

~~The reaction rates for acetolycis of 27 and 48, arc calculated~~

~~by equation (9) are liste d in Table 9. Part IV~~

~~THE INTERCEPTION OF 1IOHOALLYLIC CATIONS~~

~~BY NEIGHBORING HYDP jOXYL~~

~~5-Kexyn-l-ol (79). This alcohol was prepared by the method of~~

Englington, Jones, and Uhiting. From 110 g (0.82 mole) of tetrahydropyran-2-methyl chlorj.de (78) ^ 7 and sodium amide (prepared from 56.5 g of sodium metal) in 1509 ml of liquid ammonia, there was obtained 53 g (667c) of 7j9, bp 79-80° (13 mm), 1.4490

~~[ l i t . 117 bp 75° (16 m:n), x^16 1.4510 1~~

~~6-(2“Tctrabydropyranylo:;y) -1 •-hc-xyne (80) . A w ell-stirred mixture of~~

53 g (0.54 mole) of 79 and 44 g (0.54 mole) of dihydropyran was cooled to -30° and treated dropwise with 1 ml of concentrated hydrochloric acid. When the exothermic reaction had subsided, the reaction mixture was treated with a small, quantity of powdered potassium carbonate and d is tille d directly. There was obtaixxed 92 g (93.5%) of 8^0, bp 77-78° 22 (0.55 mm), 1.4588, which was used without further purification.

8-(2"Totrahydropyran.yloxy) -3-octyn-l-ol (81) . To a stirred mixture of lithium amide (from 3.40 g of lithium wire and ferric nitrate catalyst) in 1000 ml of liquid ammonia was added dropwise during 30 min a solutioir of 69 g (0.38 mole) of 8C) in 100 ml of ether. Stirring was continued for 30 min followed by addition of

50 g (1.13 moles) of ethylene oxide. The resulting solution was stirred for 5 hr, the ammonia was allowed to evaporate, 1000 ml of ether was added, and 75 ml of saturated ammonium chloride solution together with 300 ml of water were slowly added in order. The

~~124 125 combined organic layers were dried, the ether was evaporated, and the residue was d is tille d to give 67 g (78%) of 81, bp 115-117° (0,04 ra) , np23 1.4820.~~

~~Anal. Calcd for ^13^22^3* 68,99; H, 9.80. Found: C, 68.38;~~

~~H, 9.75.~~

~~cis-8-(2-Tetrahydropyranyloxy) -3-octen-l-0l (82£) • A solution of 33.6 g (0.15 mole) of 81_ in 150 ml of ethyl acetate containing~~

~~0.5 g of 5% palladium on cabron was hydrogenated in Brown Hydrogenator.~~

~~After the uptake of 1 vrole of hydrogen, the catalyst was separated, the filtrate evaporated, and the residue distilled to give 31.9 g~~

~~(94.57.) of 82G, bp 116-117° (0.05 mm) , nD24 1.4739.~~

~~Anal. Calcd for C, 68.38; II, 10.60. Found: C, 68.08;~~

~~II, 10.66.~~

~~c is-8-(2- Tetrahydropyrany 1 oxy) -3-octenyl jv-Toluenesulfonnte (83C) .~~

A solution of 28.5 g (0.125 mole) of 82C and 26.8 g (0.14 mole) of ju-tolucnesulfonyl chloride in 100 ml of pyridine was stored at 0° for 24 hr. Hater and ether were added and the ether layer was washed with iced 107. sulfuric acid, water, and sodium bicarbonate solution.

~~The ether layer was dried and the ether was evaporated to give 40.0 g~~

~~(837.) of a clear oil, n ^ 1.5050, which was used directly.~~

~~c.is~S-Iodo-5~octen-l-ol (85C)« To a stirred mixture of 40.0 g~~

(0.104 mole) of crude 83C. 150 ml of methanol, and 100 ml of water was addexl dropwise at 0° 11 ml of concentrated sulfuric acid. Stirring was continued for 18 hr and the mixture was neutralized with solid 126 sodium carbonate. Water was added to dissolve the solid, methanol was removed under reduced pressure, and the residue was extracted with methylene chloride. The organic layer was washed with sodium bicarbonate solution, dried, filtered and evaporated to give 30.4 g

~~(93%) of 84C.~~

~~The crude 84C thus prepared (39.0 g, 0.097 mole) and 29.2 g~~

~~(0.19 5 mole) of sodium iodide in 600 tul of anhydrous acetone was stirred at room temperature for 1 hr and at reflux for 18 hr.~~

The precipitated solid was filtered, the solvent was evaporated, water and methylene chloride were added, and the organic layer was washed with sodium thiosulfate solution and water, dried, and evaporated. Distillation of the residue gave 21.1 g (86%) of

~~85C, bp 112° (0.09 rra), n^22 1.5345.~~

~~Anal. Calcd for CgH^IO: C, 37.81; H, 5.95. Found: C, 37.27;~~

~~H, 5.82.~~

Cyclization of 85C. A solution of 18.6 g (0.073 mole) of 85C in 20 ml of anhydrous ether was added dropwioe over 4 hr to a rapidly stirred mixture of 35 g (0.15 mole) of freshly prepared silver oxide in 400 ml of anhydrous ether. The resulting slurry was stirred for

5 hr and filtered to remove the salts. The filtrate was carefully distilled to give 2.40 g (26%) of 2-cyclopropyltetrahydropyran (86), bp 76-77° (55 no), n 21 1.4525; 6 CDC13 0.1-1.1 (multiplet, 5H, U TMs cyclopropyl protons), 1.2-2.0 (multiplet, 6H, ring methylene protons), and 2.35-4.1 (three multiplets, 1H each, -CH-O-). 127

~~Anal. Calcd for CgH^O: C, 76.14; H, 11.18. Found: C, 75.87;~~

~~II, 11.20.~~

~~More exhaustive distillation afforded 1.72 g of a mixture of three higher boiling products, bp 105-125° (3.5-0.6 mm).~~

trans-8-(2-Tetrahydropyranyloxy) -3-octen-l-ol (82.7') . a solution of 9.2 g (0.40 g-atom) of sodiura in 1000 ml of liquid ammonia was added dropwise over 30 min a solution of 30.0 g (0.13 mole) of J31 in 50 ml of ether. The solution was stirred for 2 hr~ 600 ml of ether was added, and solid ammonium chloride was added in portions until the blue color no longer persisted. Hater was added and the ammonia was allowed to evaporate. The organic layer was separated, washed with water, dried, and distilled to give 7.6,3 g (87.4%) of

~~82T, bp 116° (0.05 run), n 24 1.4753.~~

~~Anal. Calcd for 0, 68,38; H, 10,60. Found: C, 68.92;~~

~~H, 10.47.~~

~~trans-8-Iodo-5-octon-l-ol (85T) . A 24.5 g (0.11 mole) sample of~~

82T was treated as above with 23.0 g (0.12 mole) of £ - toluene so. If onyl chloride in 100 ml of pyridine. The usual workup afforded 34.5 g (87%) of J33T, njj^ 1.5080, vhich was used without further purification.

~~Removal of the dihydropyran blocking group gave 26.0 g (96%) of 84T. nD25 1.5143.~~

~~From 24.5 g (0.082 mole) of 84T and 24.8 g (0.165 mole) of sodium iodide in 600 ml of anhydrous acetone, there was obtained~~

~~18.4 g (88%) of 85T,bp 110-111° (0.10 mm), nD22 1.5369. 128~~

~~Anal. Calcd for CgH^IO: C, 37.81; H, 5.95. Found: C, 37.77;~~

~~H, 5 .9 3 .~~

~~Cyclization of 851’. Dropwise addition of a solution of 13.5 g~~

~~(0.053 mole) of 85T in 200 ml of anhydrous ether to a rapidly stirred mixture of 28 g ( 0 . 1 2 mole) of silver oxide in 400 ml of ether over a~~

~~4 hr period and subsequent stirring for 5 hr gave 2,03 g (39%) of 8 6 , bp 82° (57 mm), n ^ * '1.4530. Infrared and nmr spectra and vpc~~

~~retention times V7ere identical to those of authentic 8 6 .~~

~~2-Cyclopropyltetrahydropyrr.n (85) . To a solution of cyclopropyl-~~

~~lithium in 75 ml of ether prepared from 6.0 g (0.049 mole) of~~

~~cyclopropyl bromide and 9.76 g (0.11 g-atom) of lithium wire according~~

~~■ tin to the procedure of Scyferth and Cohen'' 1 was added dropwise during~~

~~30 min to a solution of 2-bromotetrahydropyrau (87) [prepared .from 119 4.30 g (0.051 mole) of dihydropyran according to Paul ‘jin 50 ml of.~~

~~the same solvent. The mixture v.Tas stirred at room temperature" f"or 5 hr~~

~~and filtered to remove the inorganic salts. The filtrate was washed with water, dried, and carefully evaporated. Distillation of the~~

~~residue afforded 2.10 g (33.8%) of 86 _ (>957. purity). The nmr and~~

~~infrared spectra, as well as the vpc retention times on a variety 25 of columns, of purified quantity, 1.4524, were identical to~~

~~those of the previous samples.~~

~~7-(2-Tetrahydropyranyloxy)-3-heptyn-l-ol ( 8 8 ). Treatment of a~~

~~mixture of 116 g (1.38 moles) of 4 -pentyn-l-oll^O an(j g (1.38 moles)~~

~~of dihydropyran with 1 ml of concentrated hydrochloric acid at -30°~~

~~afforded 206 g (88.5%) of 5-(2-tetrahydropyranyloxy)-1-pentyne, bp 73-75° 129 30 (1.7 mm) n^j 1.4532, which was used without further purification.~~

~~Condensation of the lithium salt of this acetylene with ethylene~~

~~oxide in liquid ammonia solution provided 88 _ in 77% yield, bp 112-114° 30 (0.04 irin), 1.4791.~~

~~Anal. Calcd for C, 67.89; H, 9.50. Found: C, 67.69;~~

~~H, 9.35.~~

~~cis-7-(2-Tetrahydropyranyloxy)-3-hepten-l-ol (89G). Partial~~

~~hydrogenation of 123 g (0.58 moles) of 8j3 yielded 117 g (94%) of~~

~~89C, bP 108-110° (0.03 nn) r 24 1.4741.~~

~~Anal. Calcd for C1ZH 203: C, 67.25; H, 10.35. Found: C, 67.46;~~

~~H, 10.60.~~

~~cis-7-Iodo-4-hepten-l-ol (31C) . Fro::: 25.0 g (0.116 mole) of 89C~~

and 24.8 g(0.13 molc-s) of £-toluenesolfonyl chloride in 75 ml of pyridine, there was obtained 36.5 g (85%) of 90C as a clear oil, n ^ 2 * 1.5062. Treatment of this o il with sulfuric acid in aqueous methanol gave 27.0 g (96%) of 91C as a viscous oil which v?as subjected

~~directly to sodium iodide in acetone solution. Distillation of this~~

~~reaction product gave 18.6 g (84%) of yellow liquid, bp 97-99°~~

~~(0.13 men), n D28 1.5399.~~

~~Anal. Calcd for Cyll^IO: C, 35.02; II, 5.46. Found: C, 34.95;~~

~~H, 5.54.~~

~~Cyclization of 31C. Dropwise addition of a solution of 29.3 g~~

~~(0.15 mole) of 31C in 500 ml of anhydrous ether to a rapidly stirred mixture of 75 g (0.32 xnolc) of silver oxide in 750 ml of the same~~

solvent gave upon careful workup 6.56 g (477.) of 2-cyclopropyltctrahydro- 130 fur on (32), bp 71-72° (70 r-sn) , n 25 1.4430; 6 €?X!13 0.1-1.0 — D IMS (broad ir.ultiplet, 5H, cyclopropyl protons), 1.5-2.1 (multi.plet,

~~4H, ring methylene protons), and 3.1-3 . 8 (m ultiple!, 311, -CH-0-).~~

~~Anal. Calcd for C 7 H120: C, 74.95; II, 10.78. Found: C, 74.66;~~

~~H, 10.77.~~

~~trans-7-(2-Tctrahydropyrnnyloxy) -3-hep tcn-1-ol (897.) . Sodium~~

~~(13.8 g, 0.6 g-atorn) in liquid ammonia (700 ml) reduction of 88~~

~~(42.5 g, 0.2 mole) gave rise to 37.4 g (877.) of 89Y, bp 105-6°~~

~~(0.06 i-u), n^26 1.47 23.~~

~~Anal. Calcd for C H 0o: C, 67.25; 1!, 10.35. Found: C, 67.05; 12 2 2 3 H, 10.38.~~

~~trans-7-Iodo-4~heptc-n-l-ol (31T) . From 20,0 g (0.09 mole) of 89T and 19.1 g (0.10 mole) of jo-tolucnesulfcnyl chloride in~~

60 ml of pyridine, there was obtained 23.1 g (677) of 90T as a clear oil, 1.4995. Partial hydrolysis of this oil with sulfuric acid in aqueous methanol gave 16.1 g (90.67.) of 9173, 2(3 1.5091, which was treated directly with sodium iodide in acetone solution to afford 9.5 g (757.) of 31T as a yellow liquid, bp 88-89° (0.05 nm), n 25 1.5381. D Anal. Calcd for C^H^IO: C, 35.02; H, 5.46. Found: C, 35.08;

~~H, 5.50.~~

~~Cyclization of 31T. A 7.1 g (0.030 mole) sample of 31T was treated with powdered silver oxi.de (13.9 g, 0.06 mole) in the usual manner..~~

~~Careful distillation of the reaction product led to the isolation of 1.90 g (57%) of 32, bp 69-70° (65 mm), n p 24 1.4430, identicalin all respects to the material obtained from 17C.~~

~~trans“6"(2“Tetrahydropyranyloxy)-3-hexen-l-ol (94T). To a solution of 24.2 g (1.05 g-atoras) of sodium in 1500 ml of liquid ammonia was added dropwise a solution of 70 g (0o35 mole) of~~

~~24 121 92. * in 100 ml of ether. The resulting solution turned pi.nk aiid an additional 14 g (0.61 g-ntom) of sodium was added.~~

~~The blue solution was worked up as described above to give 12.3 g~~

~~(35%) of tryns-3-hexc-n-l-ol (93T), bp 48-49° (2.4 mi), n 23 - ■ i )~~

~~1.4390, jj-nitrophenylurethan, mp 84.0-84.5° [ lit*2^ bp 51-53° 20 (9 mm), n^ 1.4374; £~nitrophenylurethan, mp 84-85° 3 and 17.5 g~~

~~(25%) of 94T, bp ] 02-103° (0.03 m n) , n D23 1.4748.~~

~~Anal. Calcd for C^H Q03: C, 65.97; H, 10.07. Found: C, 65.79~~

~~H, 10.09.~~

~~tra ns- 6 -Iodo-3-hexen-l«ol (29T). From 15.5 g (0.077 mole) of 94T and 16.3 g (0.085 mole) of £-toluenesulfonyl chloride in~~

~~80 ml of pyridine, there was obtained 22.5 g (82%) of 95T as a clear oil. Selective hydrolysis of this oil gave 16.0 g (947.) of~~

~~96T which was converted directly to 29T in the usual manner, 6.5 g~~

~~(497.), bp 94-95° (0.05 mm), nD23 1.5418.~~

~~Attempted Cyclization of 29T. Reaction of 5.3 g (0.023 ruole) of~~

~~29T with 12.7 g (0.055 mole) of powdered silver oxide in the customary fashion produced only a viscous high boiling liquid.~~

~~Cyclization of 31C with Aqueous Silver Nitrate. A solution of 132~~

~~3.0 g (12.5 nmoles) of 31C in 25 ml of fcetrahydrofuran was added~~

• *» dropwise during 30 min to a stirred solution of 2.53 g (15 mmoles) of silver nitrate in 40 ml of water. A yellow precipitate of silver iodide formed rapidly. The mixture was stirred overnight, filtered, and treated v/ith 100 ml of vmter. The filtrate was extracted with ether and the combined organic layers were dried and carefully evaporated. Distillation of the residue gave 0.43 g (317.) of

~~32, bp 60-61° (45 mm).~~

~~Cyclization of J3JX with Aqueous Silver Nitrate* Treatment of~~

~~5.0 g (0.02.1 mole) of 31T with 4.2 g (0,025 mole) of silver nitrate in the manner described above gave 0.68 (297.) of 3J?, bp 66-67°~~

~~(50 mm). P a r t V~~

~~STABILIZED DERIVATIVES OF CIS. CIS.CIS-1,3 ,5~CYCI/)BZCATRIEN£.~~

~~KETO-ENOL 3aUTOMEAIS4 IN 2,3~DICARBOKETHOXY-CIB,CIS~~~

~~3 ,5-CYCIX>i'ECADIEKONES AND ClS-3-CYCLOALKEKONES~~

~~i 23 1 -Di ruethylamino-1,3-cyclooc.t.adiene (1 20) „ ' A solution of~~

~~1 0 / 40.0 g (0.143 mole) of N-methylgranatenine nethiodide * in water was passed through a column of Ambe.clite IRA-400 (in it s basic form).~~

~~The alkaline eluate was collected and the water was removed under reduced pressure. The remaining oil was d istille d at 100-120°~~

~~(30 orn). The distillate was dissolved in ether and dried over magnesium sulfate to remove the remaining water. The ether was evaporated and the product redistilled to give 19.1 g ( 8 o%) of~~

~~a. -des-dimethylgranateninc, bp 80-85° (15-20 ran) 1.4580~~

~~[ l i t . 1 2 4 bp 80° (12 ran), n£)25 1.4583].~~

~~This diene was pyrolyzcd at 160-170° (140 mm) to induce a~~

1,5-dienyl hydrogen shift and give rice to l-di.methylaraino-1,3- 175 cyclooctadiene ( 120) . ' This dienomine was obtained in 507. yield 27 and was purified by re d is tilla tio n , bp 100-101° (16 mm), 1.5260, neat V 1604 cm .

1-Dime thy lamino-2,3-dicarhomethoxy -ci s, ci s, c.i s-1,3 ,5-cyclodecatriene (121). A solution of 7.2 g (0,048 mole) of dienamine 120 in 150 ml of anhydrous benzene was cooled to a thick slurry in a dry ice-acetone bath. To this frozen mixture was added a cold solution of

~~6 . 8 g (0.048 mole) of dimethylacetylenedicarboxylate i n 50 ml, of the same solvent. This mixture was allowed to warm to room temperature~~

~~133 134 i or and to stand for 45 hr. The ben?;one was evaporated under reduced pressure and the remaining brown solid was washed with ether to afford~~

~~6.9 g (49.3%) of yellow powder. Recrystallization of this solid from acetone gave pure 1_21 as yellow crystals, mp 130-132°. v N ujo1 1620(w),~~

~~1670(g), and 1735(s) cm"1; \ Et0U 250 (11,600), 267 (10,700), 320 max (10,900), and 356 mp (8,750); its nrar spectrum (in DHSO-ck,) displayed peaks at 6 2.97 (s>N(CH3 ) 2 ), 3.49 and 3.66 (s,C00Cfbj),~~

~~5.5-6 . 8 (311, vinyl protons) and 8 protons in a broad upfield absorption band (allylic and methylene protons).~~

~~Anal. Calcd for C, 65.51; II, 7.90; N, 4.78.~~

~~Found: C, 65.43; H, Co02; N, 4.62.~~

Selective Reduction of 1,21. l-DiinethylDinino-2,3~dic«rbcmethoxy~ cis,cis-l,3-cyclooctadiene (122). A solution of 0.60 g (2.0 mmoles) of 121 in 20 ml of ether was hydrogenated at atmospheric pressure over Adams' catalyst. After the consumption of one mole of hydrogen, the hydrogenation was stopped. The catalyst was removed by filtration and the filtrate v;as evaporated under reduced pressure to give 280 mg

(497.) of a white solid and a. yellow viscous oil (non-reduced and over-reduced materials). Recrystallization of the solid from ether afforded pure white crystals of 122, mp 106.5-1080. u ^^14 1540(s),

~~1695(s), and 1740(s) cm”1; its nmr spectrum (in CCl^) displayed peaks at 6 2.97 (s, N(CH^)^). 3.42 and 3.60 (s, COOCH^) , and 5.63~~

~~(q, J-12.5 and 4.5 Hz, vinyl proton).~~

~~Anal. Calcd for ^5.05; H, 8.53; N, 4.74. Found:~~

~~C, 6 5 .2 1 ; H, 8 .4 3 ; N, 4 .6 7 . 135~~

~~1 -Dimethyl amino-2,3 ~dicarbomethoxy~9 -oxa-ci s»c is, ci s»l ,3,5-~~

~~cyclodecatriene ( 1_24) . To a solution of 3,2 g (0.021 mole) of~~

~~l-dimetliylaroino"7-oxa-l,3"cyclooctadiene (12) in 50 ml of anhydrous~~

~~ether was added 2.96 g ( 0 .0 2 1 mole) of dimethylacetylene-~~

~~dicarboxylate over 10 min keeping the temperature below 10°. The~~

~~solution was allowed to stand for 30 min at room temperature and then was cooled in an ice-bath. The crystals thus obtained were recrystallised~~

~~from acetone to give 3.15 g (51%) of yellow crystals, mp 145-146°.~~

~~Further reery stallisatio n of this solid from acetone, or ethanol "hexane~~

~~afforded bright yellow crystals of 124, mp 147-148°. Nujol 1620(w),~~

~~1680(s), and 1735(a) cm"1; \ Et0H 257 (11,200), 270 (11,900), max 326 (10,200), and 356 n H (9,300); its nmr spectrum (in DM£9-d ) ©~~

~~displayed peaks at ca, 6 2.25 (complex rrultiplet, allyl protons),~~

~~3.00 (s, H(CH3)2), H l 3-3° (“» H-8 ), 3.48 and 3.64 (s, CCOCH 3 ) , ca_._~~

~~4.42 (trt, H-10) , end 5.8-7.1 (in, vinyl protons).~~

~~Anal. Calcd for C15N C, 61.00; K, 7.17; N, 4.74.~~

~~Found: C, 60.89; H, 7.22; N, 4.76.~~

~~Selective Reduction of 124. l~'.)imethylamino"2,3-dicarbomethoxy~~~

~~9-oxa-cis,cis-l,3-cyclooctadiene (125). A solution of 3.15 g~~

~~( 0 . 0 1 1 mole) of 124 in 60 ml of ethyl acetate was hydrogenated over~~

~~Adams' catalyst at atmospheric pressure. The consumption of one * equivalent of hydrogen required 30 min, at which point the hydrogen~~

~~ation was stopped. The catalyst was removed by filtration, the~~

~~filtrate was evaporated under reduced pressure, and the residual~~

~~viscous oil was crystallized from acetone. There was isolated 1.94 g 136~~

~~(61%) of white crystals, mp 127.5-128.5°. Further recrystallization of this material from acetone gave pure 125, mp 128-129°.~~

~~1610(e), 1675(s), and 1710(s) cra”^; in the nmr (CCl^ solution), the compound gave evidence of absorption at ca.. 6 1.50 (broad lines,~~

~~11-6, and II-7) , ca. 2.50 (broad lines, allyl protons), 3.60~~

~~(overlapping) and 3.70 (s, COOCHg) , 3.95 (s, H-10), and ca, 5.96~~

~~(quadruplet, J - 11,5 and 5.0 Hz, vinyl proton).~~

~~Anal. Calcd for C, 60.59; II, 7.80; N, 4.71.~~

~~Found: C, 60,52; II, 7.79; N, 4.67.~~

2.3-Oicarboraethoxy-ci.s,cis-3,5“cyclodccadienone (129) . To a solution of 2.35 g (8.0 ir.roles) of 121 in 12 ml of methanol was added 3 ml of concentrated hydrochloric acid and the solution was heated to reflux on a steam bath. Uatcr (1 2 nl) was added and the solution was heated on a steam bath for 10 min. The solution v:as allowed to cool and crystallization v?as induced by stretching. The crystalline product ( 1 . 0 g, 477.) was collected and recrystallized from ether to give pure white crystals of 1 2 9 , 1TIP 90-92.5°.

~~Anal. Calcd for C, 63.14; K, 6.81. Found: C, 63.22;~~

~~H, 6.80.~~

2.3-Dicarbor.:ethoxy"9-oza-cls,cis-3,5-cyclodecadienone(130) „ A solution of 0.35 g (1.2 mmoles) of 124 in 3 ml of methanol aud 1 ml of concentrated hydrochloric acid was hydrolyzed as above. The resulting oil was recrystallized from methanol to give 250 rag

~~(78.5%) of 130 as white crystals, mp 70-71°.~~

~~Anal. Calcd for c, 58.20; H, 6.01, Found: C, 58.28;~~

~~H, 6 .0 6 . 137~~

~~2 . 3 -Bicnrbom 2 thoxy~9 -oxa~icin>- 3 -cyclodecenone (135) . The hydrolysis of 1.00 g (3.7 mmoles) of 1213 by the method described above yielded~~

~~780 mg (787.) of 1.35, mp 90-92° (from ether).~~

~~Anal. Calcd for C., 57.77; II, 6.71. Found: C, 57.73;~~

~~II, 6.75.~~

2.3-Bicarbom.ethoxy-3"C;is-cyclohc'ptci'ione ( 131) . Thisproduct was gA prepared in two steps by the method of Berchtold and Uhlig, and was obtained as white crystals from aqueous methanol, mp61-62° (lit. mp 63..5-54.0°,84 55~57°85).

2.3-Dicarboraethoxy“3-jEls-»cyclooctenone (j_32) . This product was 84 prepared in tv;o steps by the method of Berchtold and Uhlig, andwas obtained as white crystals from aqueous, methanol, mp 74-75°(lit, mp 75.4-76.3°,84 74-75°85).

~~2.3-Picarbonethoxy3-cis-cyc 1 onor.c po ne (133). Hydrolysis of~~

~~8 . 0 g (26.0 mmoles) of 1 -(H-pyrrolidino)-2.3-dicarbomethoxy-cis,cis-~~

1,3-cyclononadiene, prepared in 76,5% yield by the method of Brannock, et a l . , 8"’ mp 139-141° ( l i t . 8j mp 109o5-110.5°), according to the above procedure afforded 2.4 g (35%.) of 133 as white crystals from ether, mp 76-77.5°.

~~Analo Calcd for C-j^H^O,.: C, 61.40; H, 7.14. Found: C, 61,47;~~

~~H, 7.14.~~

~~l-(N-?y rrolidino)-2,3-dicarbomethoxy-cis,cis-1,3-cyclodecadiene~~

(123). This substance was prepared in the same manner as its lower 127 homologs. From 29 g (0o16 mole) of 1-pyrrolidinocyclooctene, there was obtained 27.5 g (53%) of white crystals, mp 104-105° (from CC1 i ether) o u 1540(s), 1695(e), and 1740(e) cm’ ; it s nmr spectrum

~~(in CCI4 ) displayed peaks at 6 3.42 and 3.60 (GQOCH^) and 5.68~~

~~(q, J = 12,5 and 4.5 Hz, vinyl proton).~~

~~Anal. Calcd for 67.26; H, 8.47; N, 4.36.~~

~~Found: C, 67.20; H, 8.46; N, 4.30.~~

~~2,3-'Dicarbomethoxy“3-£is-cyclodecenone (134) . The hydrolysis of~~

~~4.0 g (12.4 mmoles) of 1P._3 by the method described above yielded~~

~~2.3 g (70%) of a clear oil which could not be made to crystallize, but which was submitted to two molecular distillations prior to analysis (see text for further discussion).~~

~~Anal. Calcd for C^H Q05: C, 62.67 ; Ii, 7.51. Found: C, 63.25;~~

~~II, 7.60.~~

~~In frared Spectra. /-II spectra were determined on freshly prepared~~

~~(unless otherwise noted) 10% solutions in carbon tetrachloride with a~~

~~Perkin-Elmer model 237 infrared spectrometer using sodium chloride~~

~~solution cells.~~

Ultra v io le t Spectra. Absorptivity measurements were made in matched 1-cm quartz cells with a Cary model 14 recording spectrophoto meter in which the cell compartment was maintained at 27*2°. The

~~spectra were obtained on dilute solutions which had been equilibrated~~

~~at room temperature for 3 days before measurement,~~

~~Nmr Spectra. Proton magnetic resonance spectra were obtained~~

~~on 2.5 H solutions which had been allowed to equilibrate at room~~

~~temperature for periods of 2 and 4 days and were recorded with a 139~~

Varian A-60 spectrometer. Chemical shift and equilibrium constant measurements have been made at 35-2°. Chemical sh ift values are reported in 6 units from internal tetramethylisilane ( 6 “ 0 ) + to within - 0.02 p.p.m. Equilibrium constants were calculated

~~from the integrated intensities of keto and enol proton signals; the areas of the resonance peaks were hand-integrated repeatedly~~

~~enol_ s keto with a planinoter, and the preccntagc-s are accurate to within-!■ 2%. ParL VI~~

~~NOVEL PRODUCTS FROM THE REACTION OF 1 -DIMETRYLAMINO -1,~~

~~3-CYCL00CTADIENE KITH KETHANESULFONYL CHLORIDE~~

~~Reaction of l-Dmethylnmino-l ,3-cyclooctadien (120) with~~

~~Methanesulfonyl Chloride. To a stirred solution of one equivalent of 120 and 1 - 2 equivalents of triethylamine in tetr&hydrofuran~~

(i.e. 5.0 g of 1.20 in 75 ial of TilF) at -10° and under a nitrogen atmosphere v?as added dropv.’ise over 0.5 hr a corresponding equivalence of me than esul forty 1 chloride in an equal amount of tetrahyc’.ro furan.

The resulting solution was stirred for 1 hr at -10° and for an additional hour at room temperature. The solution was filtered to remove the precipitated triethylamine hydrochloride and the solvent was removed at room temp crate re. The resulting o il v;ao chromato graphed.

~~Under various reaction and workup conditions (see Table 7) seven products were obtained.~~

~~Sulfone 140 was recrystallizcd from ether-ligroin to give white crystals, rap 85-86°; X 6 .12.(in) , 6.18(m), 7.62(s), and~~

8.80 p ( s ) ; XlGOOCt- 204 (9900) and 236 (1350); 6 ^ 4 7.00 max ^LllJ (doublet of doublets, J •» 7.5 and 9.5 Ha, 1H,-CH-C-S0 2CII3) , 4.90 H (triplet, J - 8.0 Hz, 1H, -CK-C-NMe2), 2.78 (singlet, 311, -£0 2CH3),

~~2.58 (singlet, 6 H, -N(CH ) 2) , 1.0-2 . 8 (roultiplet, 8 H).~~

~~Anal. Calcd for Cn H 1 9N02 S: C, 57.62; H, 8.35; N, 611; S, 13.96.~~

~~Found: C, 57.56; H, 8.39; N, 6.02; S, 1394.~~

~~Sulfone 137 v?as obtained as white crystals from ether-ligroixr,~~

~~140 141~~

~~max~~

6 CDClg 5,83 (imltiplet, 2H, vinyl protons), 5.17 (multiplet, TMS He), 4.32 (doublet of doublets, J = 14 and 3 Hz, Ha), 3.42 (doublet of doublets, J c 14 and 1 Hz, Hb) , 2.42 (singlet, 611), and 1.1-2.3

~~(miltipiet, 811).~~

~~Anal. Calcd for C^H^KO S : C, 57.62; H, 8.35; H, 6.11;~~

~~S, 13.96. Found: C, 57.56; H, 8.38; K, 5.85; S, 13.97.~~

~~Sulfone 141 was obtained as white crystals from ether, mp 94-95°;~~

~~X CC14 7.61(g), 8.200.0,8.42(s) , 8,54(s) and 8.9000; ? Rt0U max m 246 mp(5860); 6 „ 4 5.96 (mu1 tiplot, 2H, vinyl protons), 4.20 HIS~~

~~(singlet, 211,-S 0 2 "CH2 ~) , 2,2-2.7 (multiplct, 4-H, allylic protons),~~

~~1.4-2 .1 (multiplct, 411).~~

~~Anal . Calcd for C, 58.66; II, 6.59; S, 17.40.~~

~~Found: C, 58.66; H, 6,60; S, 17.41.~~

~~Sulfone 142 was obtained as white crystals from etjhcr-llgroin, mp 109-110°; X ^ 4 6.11(m), 7.57(e), 8 . 8 8 ( e ) ; \ 237 mp,(5770);~~

~~6 CDClg 5 ^ 5 (I'u.ltiplct, 211, vinyl protons), 3.7-4.5 (multiplet, 4H), TI'IS 2.92 (one proton peak, Ha), 2.71 (singlet, and 1.4-2.3~~

~~(mult ip le t, 611).~~

~~Anal. Calcd for C^lIj^NOgS: C, 57.62; II, 8.35; N, 6.11; S, 13.96.~~

~~Found: C, 57.47; II, 8.28; N, 5.94; S, 13.96.~~

Sulfone 143 was recrystallized from chloroform-ligroin to give FCC1 white crystals, mp 144-145°; X ' 3 5.80(m), 7.52(s), and 8.82|j,(s); max x EtOH 286 mp (23); 6 5.75 (multiplet, 211, vinyl protons), max

~~g le t, 2H, -E-CH 2 -£02"), 3.80 (doublet, J = 8 Hz, 2H.-S0 - - - 142~~

~~CH^-CH-), 2.40-2.65 (multiplet, 211, allylic protons), 1.60-2.35~~

~~(multiplet, 211, allylic protons), 1.60-2.35 (multiplet, 611).~~

~~Anal. Calcd for Cgllj^G: C, 53.46; 11, 6.98; S, 15.82.~~

~~Found: C, 53.43; H, 7.01; S, 15.74.~~

~~The 2,4-dinitrophenylhydrazone was obtained as orange crystals~~

~~from 95% ethyl alcohol, mp 219-220°.~~

Sulfone 144 was recrystallized from ether to give white crystals, mp 169.0-170.5°; \ HCC13 7.61 ( b) and 8.80 u(s); 6C°C13 5.52 max TMS (multiplet, 1M), 3.34-3.70 (multiplet, 211) , 3.20 (singlet, 1H),

~~2.78 (doublet, J = 4 Hz, 1H) , 2.50 (multiplet, 111), 2.20 (singlet,~~

~~6 H,-NIie2) 1.2-2.4 (m ultiplet, 711).~~

~~Anal. Calcd for Cj^H^NO^: C, 57.62; H, 8.35; N, 6.11;~~

~~S, 13.96. Found: C, 57.39; II, 8.30; K, 5.75; S, 14.11.~~

~~Sulfone 14_5 was obtained as yellow crystals from methanol, mp 192-193°; \ HCC13 6.48(c), 7.73(b), 8.83(s) and 8.97 u (s); max r~~

~~XEt0R Sh 218 (6580), 263 (6580) and 352 m,L (13,200); 6 CDC13 7.74 max TMS (singlet, 1H, vinyl proton), 3.30 (singlet, 61I,-NMe2), 3.02 (singlet,~~

~~3H,-S02 CH3), 2.94 (singlet, 3H,-S0 2CH3), and 8 H at higher field.~~

~~Anal. Calcd for C, 46.90; H, 6.89; N, 4.56;~~

~~S, 20.82. Found: C, 46.63; H, 6 . 8 8 : N, 4.41; S, 20.78.~~

~~Hydrolysis of 145. A solution of 1.30 g (4.2 mmoles) of 145,~~

~~4 ml of concentrated hydrochloric acid and 10 ml of water was heated~~

at reflux for 10 min. The resulting solution was cooled, extracted with methylene chloride, the organic layer was dried ancT the solvent was removed to give 1.01 g ( 86 %) of solid material. Recrystallization from chloroform-ligroin gave white crystals, mp 133.0-134.5°;

~~\ HCC13 5.89 (w) , 6.24(b), 7.59(e) rnd 8.81 n(s); 10*61 max r i-Mo (broad peak, 0.7H, enol proton), 7.28 (singlet, 0.711, “CH-A"),~~

~~7.12 (doublet, J = 5.5 Ha, 0.3H, = C H ~ A h - £ 0 2 - ) , 5.00 (doublet,~~

~~J = 5.5 Hz, 0.3K, =CH-4h-S02), 3.15, 3.07 and 3.04 (three sharp peaks, 611, --020112), 2.2-2 ,8 (multi.plet, 4H, allylic. protons) and~~

~~1.5-2.1 (mu 1 tip le t , 411).~~

~~Anal. Calcd for C, 42.84; H, 5.75; S, 22.87.~~

~~Found: C, 42.70; H, 5.74; S, 22.73.~~

~~Hydrolysis of .142.* From 0.17 g (0.74 nmole) of 142, 0.5 ml of concentrated hydrochloric acid and 2 ml of water there was obtained REFERENCES~~

~~lo N. J. Leonard, Rec. Chera. Progr., L7, 243 (1956).~~

~~2« N. J. Leonard, T. 17. Milligan and T. L. Broun, J. /.m, Chen„ Soe. . 82, 4075 (1960).~~

~~3, S. Winstein, E, Allred, R, Heck and R. Glick, Tetrahedron, 3, 1 (1958).~~

~~4. Go T 0 Kwiatkowski, S. J. Kav&rnos and II. D„ Closson, J. Heterocyclic Chem. , 2> 11 (1965)0~~

~~5o B. Capon, Quart. R.ev. , 18, 45 (1964).~~

~~6 . V. Prelog and J. G. Traynham in P. de Mayo (ed.) "Molecular Rearrangements," Inter science Pub].., Nev; York, 1963, Vol. I, p. 593c~~

~~7. A. C. C.ope, Mo M. Martin, and M. A. McKervey, Quart. Revs. , 20, 119 (1966).~~

~~8 . V. Prelog and 11.Kung, Helv. Ch.cm. Acta, 39_, 1394 (1956).~~

~~9c A. Co Cope and P. E. Peterson, J. Am, Chem. Soc ., 81, 1643 (1959),~~

~~10o S, Winstein and R. B, Henderson, J, Am. Chem. Soc. , 65, 2196 (1943).~~

~~11. P. G. Gassman and J, L. Marshall, Tetrahedron Letters, 2429 (1968).~~

~~12. (a) R. !?, Begland, M.Sc. Thesis, The Ohio State. University, 1965, (b) Lo A. Paquette and R. 17. Begland, J. Am. Chem. Soc. , 87_, 3784 (1965).~~

~~13. This thermally induced sigmatropic change of order [l,5] is presumably concerted and suprafacial R. B. Woodward and R. Hoffmann, J. Am. Chem. Soc. . 87, 395, 2511 (1965).~~

14. (a) A mixture of 2- and 3~cyclooctenones is known to achieve equilibrium under acidic conditions with a composition of 207. of the g :,P -isomer; No Heap and G. H, Whitham. J. Chem. Soc., ,B, 164 (1966). (b) We have found that 14 reaches an equilibrium containing 157. of its o!,P “isomer under similar conditions. See Part III, p 3 6 .

~~15. R.elated acid-catalyzed transannular reactions have recently been described: L. A, Paquette and L, D, Wise, J. Am. Chem, Soc,, 87,~~

~~144 145~~

~~1561 (1965); L. A. Paquette and M. K. Scott, J. Org. Chem. , 33, 2379 (1968).~~

~~16. P. C. Storm, M.Sc. Thesis, The Ohio State University, 1S67.~~

~~17o L. A. Paquette, R. W. Begland and P. C. Storm, J„ Am. Chem. Soc. , 90, (1968).~~

~~18. H. Gilman and A. P. Hewlett, Rec. Trey. Chi in. , 51.93 (1932).~~

~~19. R0 B. Woodward, A. A, Patchett, P. II, R. Barton, D. A. J. Ives, and R. B, Kelley, J . An. Chem. Soc. . 76, 2852 (1954).~~

~~20. Huang-Mini on. J. Am. Chem, Soc... 6 8 , 2.487 (1946).~~

~~21. I.. F. F iescr, ib id ., 76, 1945 (1954).~~

~~22. A. C. Cope, D. L. Nealy, P. Scheincr, and G. Hood, ib id ., 87_, 3130 (1965).~~

~~23. A. C. Cope and B. A. Paw son, J. An. Chem, Soc., 90. 636 (1968).~~

~~24. J. Koinwald and II. Hozaki, J.Am, Chrra Soc. , 80. 3132 (1958).~~

25. For summaries of past work onthis subject, see (a) F.. L. E liel, N. L. Allingcr, S. J. Angyal, and G. A. Morrison, ’'Conformational Analysis," Interscience Publishers, Inc., New York, N.Y., 1965.

26. (a) J. D. Roberts, Abstr. 19th National Organic Symposium, Tempo, Arizona, June, 1965; (b) F. A. L. A.net and M. St. Jacques, J. Am.Chem. Soc. , 8 8 , 2585 (1966); (c) F. A. Anet and M. St. Jacques, ibid. . 8 8 , 2586 (1966); (d) M. St. Jacques, M. A. Brown, and F.A.L. Anet, Tetrahedron L etters. 5947 (1966).

~~27. Reference 25a, p. 210.~~

~~28. J. Dale, J. Chem. Soc. . 93 (1963);J. Dale, Angcw. Cheia. In tern . Ed., 5, 1000 (1966)'.~~

~~29. L. A. Paquette and R. Vh Begland, J. Am. Chem. Foe.. 8 8 , 4685 (1966).~~

~~30. E. L. Eliel and Sr. M. C. Knoeber, ib id ., 8 8 , 5347 (1966).~~

~~31. After completion of this work, the synthesis of by a totally different route appeared: F„ Nerdel, J. Buddrus, W. Brodowski, and P. Weyerstahl, Tetrahedron Let t e r s . 5385 (1966).~~

~~32. The author wishes to express his heartfelt appreciation to Prof. Joseph B. I.ambert of Northwestern University for his invaluable assistance in obtaining these spectra. 146~~

~~33. In actuality, temperature increments of 30° were recorded between +30° and -100°. No indication of spectral modification was noted in any of the scans.~~

~~34. J. B. Lambert, R. G. Keske, and 11. K„ Ueary, J. An. Chem. Soc. , 89, 5921 (1967).~~

~~35. J. M. Lehn and F. G0 Riddell, Chem. Comm., 803 (1966); J. E. Anderson, Quart. Revs.. 19, 426 (1965).~~

~~36. S. Uinstein and L. L. Ingraham, J. Am. Chem, Soc., 74, 1160 (1952).~~

~~37. S. Uinstein, C. R 0 Lindegren, and L. L. Ingraham, ib id . . 75, 155 (1953).~~

~~38. S. Uinstein, E. Greenwald, and L. L. Ingraham, J . Am. Chem. Soc., 70, 821 (1948).~~

~~39. J. C. Martin and P. D. Bartlett, J. Am. Chem. Soc.. 79, 2533 (1957).~~

~~40. A. C. Cope, S. Moon and P. E. Peterson, J. Am. Chem. Soc.-, 84, 1935 (1962).~~

~~41. K. Ziegler and H, Ullms, Ann., 567, 1 (1950),~~

~~42. L, A. Paquette and R. U„ Begland, J. Org. Chem,, 32, 2723 (1967).~~

~~43. See Reference 29, and Parts II and V of this thesis.~~

44. (a) A 5 ft x 0.25 in, aluminum column packed with 207. Carbowax 20H on Chromosorb P was employed in conjunction with an Aerograph A-700 gas chromatrographic unit, (b) 12 ft x 0.25 in. column packed with 107, Carbowax 6000 on Chromosorb G. (c) 5 f t x 0.25in. column packed with 10% 8F-96 on Chromosorb P.

~~45. E0 Vogel and H 0 Gunther, Angew. Chem.. 79. 429 (1967).~~

~~46. H. D. Closson, P. Uriede, and S. Bank, J. Am. Chem. Soc., 8 8 , 1581 (1966).~~

~~47. J. U. Hilt and D. D. Roberts, J. Qrg. Chem., 27,3434 (1962).~~

~~48. K. Bowden, I. M. Heilbron, E.R..H. Jones and B. C. L. Ueedon, J. Chem. Soc.. 39 (1946).~~

~~49. E. M. Kosower and S. Uinstein, J. Am. Chem. Soc.. 78, 4347 (1956). 147~~

~~50. (a) S. Uinstein and H. J . Holness, ib id . . 77, 5562 (1955). (b) D. 11. R„ Barton, J . Chem. Soc. , 1027 (1953).~~

~~51. S. P. Acharya and H. C. Brown, J. Ava. Chem. Soc.. , 89, 1925 (1967).~~

~~52. Ho C. Brov;n and A. Suzuki, ib id, 89. 1933 (1967).~~

~~53o H. C. Brown and G. Zwcifel, ib id .. 86,393 (1964).~~

~~54. (a) VJ, G. Dauben and W. T. Uipke, J. Org . Chem,. 32, 2976 (1967). (b) L. Birladeanu, T. Kanafusa, and S. iiinstein, J. An. Chca. Soc., 8 8 , 2315 (1966).~~

~~55. A. C. Cope, S. Moon, and C. H. Park, J. An. Chem. Soc., 84, 4843 (1962).~~

~~56. P.. U. Lcr.n.eux, P.. K. Kullnig, K. J. Bernstein, and 17. G. Schneider, ib id ., 80, 6098 (1958).~~

~~57. H. Prinabach and E, Lruckrcy, Tetrahedron Letters, 34_, 2959 (1965).~~

~~58. (a) H. Freundlich and 1C. Kroepelin, 7,- physic. Chem., 122, 39 (1926); (b) G. Salomon, Helv. Chim. Acta. , 16, i36l (1933) .~~

59. (a) G. Ik Bennett, F. Heathcoat, and A. N. Mosses, J. Chc-ra. Soc. , 2567 (1929); (b) G. M. Bennett and E. G. Turner, ib id . , ~813~ (1938); (c) see also 1J. E. Truce, 17. 17. Bannister, and P,, II, Knopee, J. Org. Chem, 27, 2821 (1962).

~~60. J. F. Lane and H. 17. Heine, J. An. Chon. Soc, , 73, 1348 (1951); II. 17. Heine, E, Becker, and j. F„ Lane, ibid. , 75, 4514 (1953),~~

~~61. (a)Reference 3; (b) E. L. Allred and S. Uinstein, J. Am. Chem. Soc,, 89, 3991, 3998, 4003, 4012 (1967); (c) E. R. Novak and” D.~ S. Tarbell, ibid. , '89. 73, 3086 (1967).~~

~~62. H. 17. Heine, A. D. K iller, 17. H, Barton, and R. W. Greiner, ibid. . 75, 4778 (1953).~~

63. (a) L. A. Paquette, Tetrahedron Letters. 1291 (1965); (b) L. A. Paquette and II. Stucki, J. Org. Chem. , 31, 1232 (1966); (c) L. A. Paquette and R. 17. Begland, ref. 12; (d) L. A. Paquette, R. 17. Begland and P. C. Stom , J. Am. Cheap Soc. , 90, (1968).

~~64. (a) Lo A. Paquette and L. D. Y7ise, J. Am. Chem. Soc., 89, 6659 (1967); (b) L. D. Wise, Ph.D. Thesis, The Ohio Stat'e University, 1967. 148~~

~~65. H. 0. House, "Modern Synthetic Reactions," 17. A. Benjamin, Inc., Nc-w York, N.Y., 1965, pp 71-72.~~

~~6 6 . (a) D. J. Craia, "Fundamentals of Carbanion Chemistry," Academic Press, Inc., New.York, N. Y., 1965 pp 243-256; (b) A. Nickon and H. H. Uerstiuk, J. An. Chan. Soc., 89. 3014 (1967).~~

~~67. H. 0. House, J. -J. Riehl, and C. G. P itt, J . Or?;. Chem., 16, 650 (1965).~~

~~6 8 . G. Stork, P. Rosen, N. Goldman, R. V. Coombs, and J. Tsuji, J. Am. Chem. Soc., 87, 275 (1965).~~

~~69. G. Stork, S, Ualhotra, H. Thompson, and II. Uchibayashi, ibid. , 87, 1148 (1965).~~

70. For recent reviews of this subject, see (a) M. Hanack and H. J. Schneider, Angew. Chem. Intern. Ed. (English), 6_, 6 6 6 (1967); (b) L, H. Ferguson and J. C. Nnadi, J. Chem. Ed., 42. 529 (1965); (c.) P.. Breslow, in "Molecular Rearrangements," P. de Mayo, Ed., Interscience Publishers, New York, N. Y., 1963, Chapter 4; (d) J. A. Berson, ibid. , Chapter 3.

~~71. (a) M. Hanack, S. Kang, J. Haffnur, and K. Gorier, Ann., 690, 98 (1965); (b) K. L, Sc-rvis and J. D. Roberts, J. Am. Chem. Soc., 87, 1331 (1965).~~

~~72. (a) M. Simonetta and S. Uinstein, ibid., 76, 18 (1954): (b) S. Uinstein and E. K. Kosower, ibid. , 81, 4399 (1959).~~

~~73. H. E. H. Howden and J. D. Roberts, Tetrahedron. 19, Suppl, 2, 403 (1963).~~

~~74. P. D. B artlett, "Nonclassical Ions," U. A. Benjamin, Inc., New York, N. Y., 1965.~~

~~75. C. N. Shoppee, J. Chem. Soc., 1147,(1946); S. Uinstein and R. Adams, J. Am. Chem. Soc.. 70, 838 (1948); S. Uinstein and A. H. Schlesinger, ibid. , 70, 3528 (1948).~~

76. The symbolism HO-n is herein employed to denote the size of the heterocyclic ring (n members) ultimately produced (or at least considered producable £ priori) in the cyclization, as suggested earlier for neighboring methoxyl participation.

~~77. D. S. Glass, J. V7. H. Uatthey, and S. Uinstein, Tetrahedron Letters. 377 (1965); J. U. H. Uatthey and S. Uinstein, J. Am. Chem. Soc.. 85, 3715 (1963). 149~~

~~78. E. Vogel, 17. Grircne, and E. Dirtue, Tetrahedron L etters. 391 (1965).~~

79. Relative to 110. 112 was observed to rearrange at a slightly faster rate with a lower A (20.1 - 1.0 kcal/raole). The rate process for the conversion of 112 - 113 was followed by nmr spectroscopy in the temperature range 28-40°. The entropy of activation was calculated to be - 1 2 . 8 e,u,

~~80. E. J. Corey and A. G. Hortrc'.ann, J. AmMChem. Soc., 87, 5736 (1965).~~

~~81. P. de Mayo and G. T. Reid, Quart. Rev. (’London), JL5, 393 (1961).~~

~~82. A. Windaus and G. Zuhlsdorf, Ann. 536, 204 (1938).~~

83. A similar "push-pull" concept for purposes of stabilization recently has been applied in an. independent study to the cyclo- butadiene system: R. Breslow, D, Kivelevich, M. J. M itchell, W0 Fabian, and K. .7 end ell, J. Am. Chen. Soc,, 87, 5132 (1965).

~~84. G. A. Berchtold and G. F. Uhlig, J. Org. Chon. , 28, 1459 (1963).~~

~~85. K. C, Brannock, R. P. B urpitt, V. V J . Goodlett, and J. G. Thweatt, ib id . . 28, 1464 (1963).~~

~~8 6 . A. K. P.ose, G. Mina, K. S. Mannas, and E. Rzucidlo, Tetrahedron Letters, 1467 (1963).~~

~~87. For a brief discussion of nomenclature pertaining to such enar.iines, refer to L. A. Paquette, J. Am. Chem. Soc. , 86, 4092 (1964).~~

~~8 8 . Brief mention of this result has been made: L. A. Paquette, Trans. N. Y. Acad. Sci.. 2.8, No. 3, 387 (1966).~~

~~89. H. C„ Longuet-Higgins and E. 17, Abrnharason, J. Am. Chem. Soc. , 87, 2045 (1965).~~

~~90. For examples, cf. E. Vogel, Ann. , 615, 14 (1958); R. Criegee and K. Noll, ibid. . 627, 1 (1959).~~

~~91. For examples, cf. L. A. Paquette and J. H. Barrett, J. An. Chen. Soc. , 8 8 . 1718 (1966), and pertinent references cited in footnote 13 of this paper.~~

~~92. The double bond at C 5 “C^ is known to be cis because of the un equivocal synthetic route employed in the preparation of dienamines 120 a.nd 1 2 . 150~~

93<, This conclusion has been reached by making recourse to results obtained with the parent hydrocarbon systems (vide infra); inherent in this extrapolation is the inherently reasonable assumption that the substituents in F will not seriously affect the mode of thermal organization which, in all of these examples, occurs in a disrotatory fashion primarily because of steric factors.

~~4 5 0 - 500 °~~

~~Oo L. Chapman, D. J. Pasto, G. W. Borden, and A. A. Griswold, J. Chem. Soc.. 84, 1220 (1962).~~

~~So F. Chappell, I I I , and R. F. Clark, Chem- I.vid 0 (London), 1198 (1962).~~

~~o N. an J. Fonken, J. An. 87, 3996 (1965).~~

~~940 It is not expected that isomerization of this bond would occur in the cycloaddition reaction.~~

95o The numerous varied approaches to cyclodecapentaene are cases in point. Of especial pertinence may be cited Johnson, Bass, and Williamson's unsuccessful attempts [ Tetrahedron, 19., 861 (1963)] to induce i to enolize. 151

96. The sesquiterpene germacrone (ii) , which represents an interesting substrate for keto-enol tautomcrism, has been investigated to a considerable extent but unfortunately no comments regarding its enol content are available. However, a small amount of enol

~~i i~~

~~tautomcr perhaps is suggested by its published infrared spectrum: G. Ohloff, H. Farnow, W, Fhillinp, and G. Sell ad e, Ann., 625, 206 (1959) , and references cited therein.~~

When attempts were made to vacuum d is til germacrone above 165°, the indicated rearrangement was elucidated (r.ot observed with 21 or 22): I. Ogujanov, I). Ivanov, V. Herout, M„ llorak, J. Pliva, and F.~Form, Coll. Cucch. Chem, Comm., 2.3, 2033 (1958); V. Herout and M. Suchy, ibid. , 23, 2169 (1958): M. Suchy and F, Sorm, ibid, 23, 2175”'(1958)7"

~~97. S. J. Rhoads, J. C. G ilbert, A. W. Decora, T. R. Garland, R, J. Spangler, and 11. J. Urbigke.it, Tetrahedron, 19, 1625 (1963), and references cited therein.~~

~~98. C. N. R. Rao, "Ch emical Applications of Infrared Spectroscopy," Academic Press, New York, N. Y,, 1963, p 214.~~

~~99. G. S. Hammond, W. G. Borduin, and G„ A. Guter, J. Am. Chen. Soc, , 81, 4682 (1959). •~~

100. Such a conclusion, however, does not seem ju stified in the case of 133 and 134 because their combined spectral data (Tables 2-4) suggest that a high concentration of enol is never attained with these compounds even in non-polar media. Although the derived K values are at best qualitative in these examples, they are usefuf to indicate the relative increase in keto tautomcr (or, conversely, the relative decrease in enol content) when passing to solvents of increasing polarity.

~~101. For some values, see J. Kine, "Physical-Organic Chemistry," Second Ed., McGraw-Hill Book Co., New York, N. Y.,1962, p 39. 152~~

102o In recent years, direct measurement of the degree of: enolization of a number of keto-enol pairs has been made; cf._, for example, (a) J 0 Ao Pople, Wo Go Schneider, and H„ S 0 Bernstein, "High Resolution Nuclear Magnetic Resonance," McGraw-Hill Book Co0, New York, NoY«, 1959, Chapter 17, avid references cited therein; (b) E0 Wo Garbisch, J 0 Am. Chem0 £oc0, 85, 1696 (1963); 87, 505 (1965); (c) Jo Lo Burdett and M 0 T0 Rogers, ibid. , 8 6 , 2105 (1964); (d) S„ Forsen, F, Merenyi,” and M 0 Nilsson, Acta Chem. Scand., 18, 1208 (1964); (e) I'U T. Rogers and J c L 0 Burdett, Can. J. Chem,, 43, 1516 (1S65; (f) S» Jo Rhoads, J Q Orp,.Chem. . 31, 171 (1966); (g) G. Allen and R„ A« Dwok. J. Chem. focM Phys. Org. , 161 (1966).

103o Random samples were examined again after standing for an additional week, but no further changes in Kc were noted. Similarly, after selected samples were heated on a steam bath for varying periods of time, no further changes in enol content was noted.

104. (a) G. V. Hatton and R,. E. Richards, Hoi. Rhys., 3, 253 (1960); (b) U. G. Schneider, J. Fhys. Chem. , (56, 26iTi (1962) ; (c) L. LaPlanche and M. T 0 Rogers, J . Am. Chem. Soc., 85, 3728 (1963); 86 _, 337 (1964) o

~~105. (a) J. B. Hendrickson, ib id , 86 _, 4854 (1964); (b) K0 B. Uiberg, ibido, 87, 1070 (1965),~~

~~106o (a) Jo Sicher, F rog . S te rcochemo, 3, 202 (1962); (b) J„Dale, Jo Chem. Soc,, 93 (1963)T~~

~~107o Ho C. Brown and M» Gerstein, J, Am, Chem, Eoc, , J_2, 2926 (1950); H. C. Brown, J 0 Chcrn. Soc. , 1248 (1956) 0~~

108o This statement implies that the rigidity imposed by the cls.cis- dienoid or els, cis,cis-tricnoid components of the enols causes such molecules to be more sensitive to non-bonded stcric in te r action? than their keto tautomers which are more flexible (because the sp -hybridized keto carbon spreads one of the C-C-C bond angles etc.). Dreiding models bear out the validity of this conclusion.

109. Enamines: (a) G. Stork and I. J. Borowitz, ib id . . 84, 313 (1962); (b) Go Opitz and H. Adloph, Angew. Chem,, 74, 77 (1962); (c) Wo Eo Truce, J. R., Morel 1, J. E. Richman, and J„ P. Walsh, Tetrahedron Letters, 1677 (1963);(d) G» Opitz and K, Fisher, Zo Naturforschp 18b, 775 (1963); (e) L, Ao Paquette, J. Org.Chemo, 29, 2851, 2854 (1964); (f) Do C„ Dittmer and F 0 A0 Davis, ibid, 29, 3131 (1964); (g) L. A, Paquette, i bido . 30, 629 (1965); (h) Go Opitz and K 0 Rieth, Tetrahedron Letters,3977 (1.965); (i) Go Opitz, H. Schempp, and H. Adolph, Ann. , £34., 92 (1965); (j) L, A. Paquette and M. Rosen, Tetrahedron Letters, 311 (1966); 153

J. N. Wells and F. S. Abbott, J. Med. Chem.. 9, 489 (1966); (1) Jo Jo Looker, J„ Or go Chem. , 31, 2973 (1966); (in) G. Opitz and Do Bucher, Tetrahedron Letters, 5263 (1966); (n) L. A. Paquette, H. Stucki, and M. Rosen, J. Org. Chem,, 33, (1968).

110. Ketene acetals and aminals: (a) W. E„ Truce, J. J„ Breiter, Do J. Abraham, and J, R. Norell, J . Ara. Chemo Soc. , 84, 3030 (1962); (b) II. E. Truce and J. R.'~Norell, ibid.', 85,“ 3231 (1963); (c) R. Ii. Rnsek, P 0 G. Gott, R. H 0 Mcen, and J 0 C„ Martin, J. Org. Chem. , .28, 2496 (1963); (d) W. E„ Truce and P 0 N. Son, ibidp, 30, 71 (1965); (e) R. H. Hasek, R 0 H. Meen and J 0 C. Martin, ibid. , 30, 1495 (1965); (f) G. Opitz and II. Schcnpp, Ann,, .684, 103 (1965); (g) G. Opitz, K, Rieth, and G. Walz, Tetrahedron Letters, 5269 (1965).

~~111. Lo A. Paquette and M. Rosen, J e Am. _Chrra.__SoCo, 89., 4102 (1967).~~

~~112. G. Opitz and F. Schwcincbert, Angcw. Chem. , 77, 811 (1965),~~

113. The vinyl proton of l-cyano-2,3-diethyl-l,3-cyclooctadiene was observed as a triplet (J = 7.3 Hz) at 5.606 . J. G. Atkinson, Do Eo Ayer, G. Buchi, and E. II. Robb, J. Am. Chem. Soc., 8 J>, 2257 (1963).

~~114. Compare N. S. Bhacca, I). P. H ollis, L. F 0 Johnson, and E. A 0 Pier, "NI'l Spectra Catalog," Vol. 2, Varian A.scociates, Palo Alto, California, 1963, Spectrum No. 439.~~

~~115. A 20 ft x 3/8 in stainless steel column packed viith 20% SE-30 on Chromosorb U was employed.~~

~~116. K. B. Wiberg, Physical Organic Chemistry, John Wiley 4 Sons, New York, 1964, pp 378-9.~~

~~117. Go Englington, E. R. H. Jones, and M. C. Whiting, J. Chem. So c ., 2873 (1952).~~

~~118. Do Seyferth and H„ M. Cohen, Inorg. Chem. , 2, 625 (1963); J. Organoraetallic Chem. ,1, 15 (1963).~~

~~119. R. Paul, Bull. Soc. Chira.France, I_, 1397 (1934).~~

~~120. E, R. Ho Jones, G. Englington, and M. C„ Whiting, Ore. Syn., Coll. Vol. XV. 755 (1963).~~

~~121. R0 A. Raphael and C. M. Roxburgh, J. Chem. Soc.. 3875 (1952).~~

~~122. Lo Crombie and S„ H. Harper, J. Chem. Soc . . 873 (1950). 154~~

123. G. Herling, Chen. Ber.. 24, 3123 <1891); R. W illstatter and Ko tracer .ibid. , 44, 3423 (1911); R. Tvillstatter and M. Heidelberger, ibid. 46. 517 (1913); for a brief discussion of this early work, see R0 Fo Mancke and H. L. Holmes, ”The Alkaloids,” Vol. I, Academic Press, New York, N.Y0, p 285 (1949)»

~~124. A. C0 Cope and C 0 G0 Overberger, J 0 Arn. Chcia. Soc. , 70, 1433 (1948).~~

~~1250 For a related r ear rang eraent, refer to A. C. Cope and A. A 0 D’Addieco, J. Am, Chem. Soc.. 73, 3419 (1951).~~

~~126. We have experienced that failure to mix these two reagents when very cold resulted in a vigorously exothermic reaction which afforded, upon work-up, a non-descript black tar.~~

~~127. H. E. Kuehne, J. An. Citcm. Soc. , 81, 5400 (1959).~~