Xerox University Microfilms 300 North Zeeb Rood Ann Arbor, Michigan 48106 11

INFORMATION TO USERS

This material was produced from a microfilm copy of the original document. While the most advanced technological means to photograph and reproduce this document have been used, the quality is heavily dependent upon the quality of the original submitted.

The following explanation of techniques is provided to help you understand markings or patterns which may appear on this reproduction.

1. The sign or "target" for pages apparently lacking from the document photographed is "Missing Page(s)". If it was possible to obtain the missing page(s) or section, they are spliced into the film along with adjacent pages. This may have necessitated cutting thru an image and duplicating adjacent pages to insure you complete continuity.

2. When an image^pn the film is obliterated with a large round black mark, it is an indication that the photographer suspected that the copy may have moved during exposure and thus cause a blurred image. You will find a good image of the page in the adjacent frame.

3. When a map, drawing or chart, etc., was part of the material being photographed the photographer followed a definite method in "sectioning" the material. It is customary to begin photoing at the upper left hand corner of a large sheet and to continue photoing from left to right in equal sections with a small overlap. If necessary, sectioning is continued again — beginning below the first row and continuing on until complete.

4. The majority of users indicate that the textual content is of greatest value, however, a somewhat higher quality reproduction could be made from "photographs" if essential to the understanding of the dissertation. Silver prints of "photographs" may be ordered at additional charge by writing the Order Department, giving the catalog number, title, author and specific pages you wish reproduced.

5. PLEASE NOTE: Some pages may have indistinct print. Filmed as received.

Xerox University Microfilms 300 North Zeeb Rood Ann Arbor, Michigan 48106 11

73-18,901

HENZEL, Kay Ann, 1946- PARTICIPATION BY NEIGHBORING CYCLOOCTATETRAENYL GROUPS IN SOLVOLYTIC DISPLACEMENT REACTIONS.

The Ohio State University, Ph.D., 1973 Chemistry, organic

i j University Microfilms, A XERO\Company , Ann Arbor, Michigan j PARTICIPATION BY NEIGHBORING CYCLOOCTATETRAENYL GROUPS IN SOLVOLYTIC DISPLACEMENT REACTIONS.

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

By Kay Ann Henzel, B.S *****

The Ohio State University 1973

Approved by

A U V X O WJL Department of Chemistry DEDICATION

To my unborn child who caused this thesis to be written in great haste, and to the man who made it all possible.

ii ACKNOWLEDGMENT

The author would like to express her gratitude to Dr. Leo A. Paquette for his guidance, patience, and encouragement.

ill VITA

November 19, 19*1-6 ...... Born, Phillipsburg, New Jersey X9^^ B#S«, Bucknell Universi "by, Lewi sburg, Pennsylvania.

1968-I969 ...... Teaching Assistant, Department of Chemistry, The Ohio State University, Columbus, Ohio 1969-1973 ...... Research Assistant, Department of Chemistry, The Ohio State University, Columbus, Ohio

PUBLICATIONS "The Chemistry of Flavandiones. Reaction with Diazomethane, J, Org, Chem.. 3 7 , 277*1- (1972).

FIELDS OF STUDY Major field: Organic Chemistry TABLE OP CONTENTS

Page DEDICATION...... 11

ACKNOWLEDGMENT...... ill

VITA ...... lv

LIST OF TABLES ...... vl

INTRODUCTION...... ll'l 1

RESULTS AND DISCUSSIONS ...... 28

EXPERIMENTAL...... 83

REFERENCES ...... 15^

v LIST OP TABLES

Table Page I. Buffered Hydrolysis Rate Data ...... 32 II. Buffered (NaOAc) Acetolysis Rate Data ... III. Product Variation During Acetolysis of 89-OBs...... 59 IV. Paramagnetic-Induced.Shifts for.Protons of 155-OH ...... 63 V. Solvolysis of 92-C1 in 50# Aqueous Acetone at 30.1°.... .:...... 139 VI. Solvolysis of 92-C1 in 50# Aqueous Acetone at 40.0° ...... 140 VII. Solvolysis of 92-C1 in 50% Aqueous Acetone at 4-9.9° ... .77...... 1^1 VIII. Solvolysis of 88-C1 in 50% Aqueous Acetone at 30.10 ....77...... 1^2 IX. Solvolysis of 88-Cl in 50# Aqueous Acetone at ^o.o° ....r:...... ;...... 1^3 X. Solvolysis of 88-Cl in 50# Aqueous Acetone at 49.9° ___ r:...... ;...... l4if XI. Solvolysis of 1^0-OBs in 0.10065 N NaOAc/ HOAc at 75.00 “ ... ;...... 1^5 XII. Solvolysis of lij-O-OBs In O.IOO65 N NaOAc/ HOAc at 85.0° T T ...... 1 k6

XIII. Solvolysis of l^J-O-OBs in O.IOO65 N NaOAc/ HOAc at 9 5 .00 lTT ...... 7 ...... 1^7

vi LIST OP TABLES (Continued)

Table Page XXV. Solvolysis of 139-OBs in 0.10065 N NaOAc/ HOAc at 45.0° '"77...... 148 XV. Solvolysis of 139-OBs in 0.10065 N NaOAc/ HOAc at 55.0° '"TV...... 149 XVI. Solvolysis of 139-OBs in O.IOO65 N NaOAc/ HOAc at 65.OO ...... 7 ...... 150 XVII. Solvolysis of 89-OBs in 0.10065 N NaOAc/ HOAc at 55.00 ^7...... 7 ...... 151

XVIII. Solvolysis of 89-OBs in O.IOO65 N NaOAc/ HOAc at 6 5 .0 0 ^ ...... 152

XIX. Solvolysis of 89-OBs in O.IOO65 N NaOAc/ HOAc at 7 5 .0 0 153

vii INTRODUCTION

A large part of the current chemical literature Is con cerned with substituents which are capable of Influencing a reaction by becoming bonded or partially bonded with the reaction center In order to stabilize the transition state.’1' This phenomenon is commonly called neighboring group par ticipation and frequently results in an increase in reaction rate due to the enhanced stability of the reaction inter mediate. This type of anchimeric assistance has been recog nized for heteroatom-containing groups such as methoxyl, 2 hydroxyl,^ carbonyl,** amino,thioether,^ imidazole,? etc. Although direct nucleophilic participation by hydrogen in a

Q rate-determining step occurs in very few reactions, an chimeric assistance by carbon is well known. A vast amount of literature is devoted to aryl participation in which the aryl group bridges In the ionization step,7*Q 10 while contra dictory theories have also been presented.1-1' In addition, evidence has appeared to support the theory that neighboring small rings are capable of having a significant influence on *i p reaction rates. c Finally, extensive investigation has shown that neighboring olefinic bonds can function as nucleophilic neighboring groups in intramolecular displacement reactions. This category can be further subdivided into allylic and homoallylic participation. The solvolysis reactions of a vast number of allylic compounds have been studied.^ • Winstein, Grunwald, and JonesA have suggested that the mechanism of these reactions is best described by a single mechanism which is intermediate between Snl and Sn2, rather than by competing unimolecular and bimolecular processes. Support for this hypothesis is found in the effect of solvent properties on rates of solvolysis reactions, which shows a linear relationship between solvolysis rate and solvent ionizing power. If two mechanisms were com peting, the reaction should change from a purely Sn2 process in solvents of very low ionizing power and high nucleophilicity to a purely Snl process in solvents of similar nucleophilicity but much higher ionizing power. A plot of solvolysis rate versus solvent ionizing power would then show a distinct curvature or even a break, designating a region of transition between mechanisms. Much can be learned about the mechanisms of allylic solvolysis reactions by studying the products formed. The solvolysis of most allylic halides yields mixtures of the i "normal" (unrearranged) and "abnormal" (rearranged) subs titution products and this suggests the operation of an Snl mechanism. It is not uncommon to find both members of a pair of isomeric halides giving an Identical mixture of products which is an indication not only that unimolecular substitution is Involved, but that both isomers are reacting via a common intermediate. This could be either an ion pair or a classical, mesomeric carbonium ion. 3

Isomeric allylic derivatives have also been found to give different mixtures of products under identical reaction conditions. The "product spread" simply means that one isomer may favor an Snl or Sn2 mechanism to a greater degree than its counterpart. Solvolysis of crotyl chloride (1) and methy1- vinylcarblnyl chloride (2 ) have been used as models to test the postulate that solvolysis occurs with preliminary ion ization, and each halide was predicted to give the same product mixture. This system was somewhat unsatisfactory as a model

CH3CH=CHCH2C1 CH3CHCHSCH2 Cl 2

ROH, -H+

CH3CHCH3CH2 + CH3CH=CHCH2 0R OR since the solvolysis rates were somewhat dependent on lyate-Ion concentration. However, solvolysis In each case gave a mixture of isomeric products and although the proportions of product isomers were not exactly the same for each halide, the differ ences were In the direction consistent with the postulate that the mechanism is not entirely unimolecular. Many additional examples of allylic rearrangements In simple, straight-chained systems are available in a comprehensive review. ^ Allylic participation has been observed in several simple alicyclic systems. The solvolysis rates of methylcyclobut-2- enyl (3 )* cyclopent-2 -enyl (jO, and cyclohex-2 -enyl (5) bromides

*1 Q have been studied and compared to those of suitable reference i compounds such as oc,Y-dimethylallyl (j§) and ^ iVtY-trimethyl^ allyl (7) bromides. However, direct comparisons vfith aliphatic

,CH' \

CH (c h 3) !H3 CHr B: Br systems may be invalid since conformational and strain effects play a large role in the formation of carbonium ions in ring systems. 2-Cycloheptenyl bromide (8) has been solvolyzed1^ pA as well as the analogous 3-bromocyclooctene (9)* Although

5L no rates were reported in the solvolysis of 8, was found to solvolyze 60 times faster than its saturated analog, cyclo- octyl bromide, which indicates a significant amount of allylic participation in the rate-determining step. Allylic rearrangements have also been found to occur in cycloalkenyl systems such as 10 Involving exo-endo double bond

Isomerization. In the 5-membered ring series, (n“ - 3), 11vv' ^CHgOH (CH2 )„

10Uvv undergoes quantitative conversion to 12 in the presence of hydrochloric acid and aqueous acetone. 21 As with most allylic r

11 12 rearrangements, these interconversions are very dependent upon

substituent effects as evidenced by the fact that 1^ gives

,CH-CH=(2H =c h 2 ih-c h -c h O f >

Wv13 exclusively 1^ upon treatment with acid, while 15 has been found to form only w16 v under similar conditions,21 In the six- OH (pH ^CH-CH=CH-CH'3 /CH=GH-CH-CH3 C f H+ 15 16 membered ring series (n = 4). 17 and 18 are converted to their conjugated isomers upon treatment with dilute acid.22 Alcohol 6 OH 1 c h ~c h =c h 2 H OH A? 19 12 was found to form ljg exclusively, while 1H gave the endo-

H-CH=CH-CH OH-CH-CH-CE 3

18 20 cyclic isomer 20 as the only observable product. In the seven- membered ring series, rearrangements of 2 1 , 2 g, and 2 ^ to their fully conjugated isomers in the presence of dilute acids were effected in nearly quantitative yield.^3 Rearrangement of 21, OH ^3HPh OHPh H

21 OH ,bH-CH-CH

22 .25 OH OH t l

23 26 7 gave 2 *t as expected while acid treatment of 22 gave exclusively

2 5 . In contrast to 2 2 . its higher homolog i/Vv23 underwent rearrangement with migration of the hydroxyl group in the opposite direction to give 26. This isomerization of 23 to 26 corres- ponds with the observations previously made in the cyolopent- enyl2-1, and cyolohexenyl22 series. To complete the study, the cyclooctenyl analogs, 27, 28, u W and 29 were similarly treated with dilute acid.2** Alcohol 27 underwent Incomplete rearrangement to 30, indicating that in OH GHPh CHPh

27 contrast to the smaller rings, this double bond now has con siderable stability in the endocyclic position. Treatment of

£0 with acid gave the same equilibrium mixture as that obtained from its isomer 2£,‘ The diallylic alcohols 28^ and 29, gave 31^ and J2 exclusively, indicating that the direction of rearrange ment is the.same as for analogous compounds with smaller rings.

^H-CH=CH2 r!H-GH-CH2 H+ OH o 28 31 OH OH 8 Homoallylic participation by olefinic double bonds has come under extensive experimental scrutiny since its discovery several decades ago,2^’2^ and it is intimately involved with the topic of classical and nonclassical or bridged carbonium Ions.2'* In any reaction where carbon migrates, one must decide whether ionization to the bridged ion occurs directly, or whether a classical Ion is formed first followed by interconversion to another classical ion, possibly by means of a bridged species. Direct formation of a nonclassical ion means it is of lower free energy than the classical Ion and thus Its formation should be accompanied by an increased reaction rate. One of the fundamental properties of nonclassical ions is intramolecular delocalization of positive charge. However this characteristic is not limited to nonclassical ions since car- nonium ions also possess this capacity and they have seldom, If ever, been classified as "nonclassical'*. The difference lies in the mode of overlap of atomic orbitals which combine to form the molecular orbital which permits the delocalization of positive charge. In tTie benzylic and allylic case, these molecular orbitals are formed exclusively by 7T overlap between the atomic orbitals as in Figure I, In the case of "nonclassical” carbonium ions, the molecular orbitals are formed

Figure I partially by o' overlap between atomic orbitals. This o' over lap Is relatively minor in Winstein's description of the unsymmetrical homoallylic cation which consists of a carbonium ion with a p-olefinic group.Figure II explicitly shows the overlapping p orbitals of the two olefinic atoms (C3 and C/j,) and the cationic carbon atom, C^. Winsteln’s major concept behind the term "homoallyl,, is that the methylene group (C2 ) is a poor insulator of conjugation If the proper rotational

Ci<

Figure II positions about the C^-Cg and C2 -C^ bonds are assumed. Proper positioning of these atoms gives appreciable 1 ,3-overlap of a type intermediate between d and if. Semiempirical molecular orbital calculations suggest substantial stabilization (6 kcal/ mole"*'} from this electron delocalization.2®a A calculation 9 Ah of the homoallyl cation by Roberts and Howden gives a value of only 2 .8 kcal/mole" for simple 1 ,3-overlap, but a value of 11.k kcal/mole"* when a 1 ,^-interaction is also taken into account. Their idea of optimum geometry of the cation is shown In 10 Figure III.

Figure III The concept of homoallylic conjugation first arose in considering the behavior of cholesteryl and i-cholesteryl a Q derivatives in solvolytio reactions. The acetolysis of 3^5- cholesteryl £-* toluene sulfonate (J33&) proceeds at an increased rate compared with that for cyclohexyl £-toluenesulfonate; 27b ' and the product, 3 ^-cholesteryl acetate (35) is formed with

HO AcO a, R = Ts 35 bj R = H retention of configuration. A nonclassical intermediate 2iibwas proposed and described as "homoallyl". Methanolysis of 33a in the presence of potassium acetate yields 90 % of the methyl ether of 3»5-cyclocholestan-6/3-ol 6) and 10# of that of 3§- oholesterol (33b).2^t> The same mixture is obtained from 11

36 37 38

3 »5-cyclocsholestan-6of. and 6{J-yl trichloroacetate and ^8)2^C and a similar mixture of the corresponding alcohols is obtained from the hydrolysis of the £- toluene sulfonate of 3f?- 28« hydroxymethyl-A-norcholest-5-ene (j39) • These four compounds must therefore react by way of a rapidly-equilibrating set of

OTs 39 ions or through a symmetrical ion (3^)« v W Another classic example of homoallylic participation is found in the acetolysis of anti-norborn-2-en-7-yl g-toluenesulfonate {J+0)^ which proceeds at a rate 1013, times faster than the analogous saturated compound, norborn-7-yl E-toluene-

OTs OTs OAc

uw41 42 S3 12 sulfonate (41).^° The electron cloud of the double bond In 40 is particularly well placed-to Interact with the developing carbonium ion at position 7 to yield the nonclassical ion ,4£. In addition to the rate increase, participation by the double bond is indicated by the reaction product, anti-norborn-2-en- 7-yl acetate (43^) obtained with retention of configuration. Norbornadien-7-yl derivatives undergo solvolysis even more readily than those of the anti-norborn-2-en-7-yl series as shown by the hydrolysis of chloride {(4, in aqueous acetone to yield 45 with a relative rate of 750 (norbornadienesnorbornene)

44 45 46

The intermediate carbonium ion, written as 46,, forms a stable fluoroborate salt whose structure has been Investigated by nmr These results support an unsymmetrical, non classical structure with lack of. symmetry resulting possibly from the ion-pair structure. One cannot discuss the subject of homoallylic carbonium ions without mentioning the related topic of blcyclobutonium ions since the latter can also arise from solvolytic reactions of allylcarblnyl compounds. The formation of bicyclobutonium- type Intermediates has been used to explain the rates of cyclopropylcarbinyl, cyclobutyl, and allylcarblnyl compounds^ which 13 ik are solvolyzed 10 to 10 times faster than the open-chain or saturated compounds used for comparison. In addition, solvolysis is often accompanied by complete rearrangement of one system to another and thus these reactions are particularly good examples of neighboring group participation. For example, in the deamination of cyclopropylcarbinyl^ and allylcarblnyl 'ak amines,^ reactions of thionyl chloride with cyclopropyl- carbinol and cyclobutanol, as well as in the reaction of cyclopropylcarblnol with Lucas reagent,33 isotopic labeling experiments have shown that extensive shuffling of methylene carbons occurs. In addition, the reaction products were repeatedly found to occur in the same ratio. These observations led to the assumption that all the isomers gave the same ionic intermediate which was described as three-pyramidal blcyclobutonium ions converting rapidly, but at different rates Into one another. Each ion yields cyclopropylcarbinyl, cyclobutyl, and allylcarblnyl products which differ only in the position of the isotopic label (see Figure IV),Since the reaction of n k . (c£- C)-cyclopropylcarbinyl amine gave a non-uniform distribution

# # CH 2 .. CH2 CH 2 CH2 v CH2 -jCH2

+ ! * A ch ch2 ch -— -dnz d6---- ch2

Figure IV of the isotopic label, the possibility of a fully symmetrical, tricyclobutonium Intermediate (**?) was ruled out.-^ 36* y°H

A CH 2 NH 2 28#--- 36#

C 52 The bicyclobutonium model has been contradicted by a series of further observations. Substitution of methyl groups on cyclopropylcarbinyl £-nitrobenzoates have been found to increase the rate of hydrolysis even when they are attached to the carbon atom that is uncharged in the bicyclobutonium ion picture (Figure IV). ^ This seems to Indicate that the charge must indeed be delocalized over all the carbon atoms. Racemizatlon has also been found to accompany the solvolysis of optically active cyclopropylcarbinyl derivatives.3? The fact that the rates of solvolysis and racemizatlon are the same shows that solvent can attack from both sides of the cyclopropylcarbinyl cation. In the model shown in Figure IV, substitution can take place only from the unbridged side and should therefore be stereospecifio while the rearrangement of cyclic, homoallyl compounds does not yield sterically uniform products.38*39 Hanack and Schneider explained these obser vations by assuming separate but rapidly isomerizing carbonium ions.-"39 A homoallylic ester, for example, would be expected to dissociate relatively slowly to give the homoallylic cation which has little delooalization owing to its high energy- content.It therefore rapidly isomerizes into the cyclo propylcarbinyl and cyclobutyl cations which experience greater nonclassical stabilization. Thus in the solvolysis of homoallylic esters, the product composition depends on the relative stabilities of the isomeric ions as well as the rates of substitution of the ions by solvent and on the ratio of the isomerization energy to the substitution energy.-^® Further evidence shows that homoallyl carbonium salts rearrange immediately, even at low temperature, to give three- and four- membered ring systems. The more stable cyclopropylcarbinyl and cyclobutyl cations, however, isomerize much more slowly,^ Previous Investigation of the solvolysis of allylcarbinyl chloride,^ tienzenesulfonate,**'2 and ^-naphthalene- sulfonate^ have shown no evidence for bicyclobutonium inter mediates. However, allylcarblnyl tosylate has been found to solvolyze 3*7 times faster than n-butyl tosylate In 98# formic acid.**** Product studies showed the formation of cyclopropyl carbinyl, allylcarblnyl, and cyclobutyl products in a ratio of *»■.5*1 which Is In good agreement with the product mix tures usually obtained in carbonium ion reactions of cyclo propylcarbinyl and cyclobutyl compounds. Solvolysis of 1,1- dideuterlo-3-butenyl tosylate showed the isotopic label to be statistically distributed between the methylene groups of both the cyclopropylcarbinyl and cyclobutyl products. The large percentage of allylcarblnyl product formed by Sn2 displacement under the reaction conditions made determination of deuterium scrambling difficult in this product. However, for the most part the deuterium labeling experiment does provide additional support for entrance into the equilibrating bicyclobutonium ion system from the allylcarblnyl side. The dependence of the neighboring group effect on stereo chemistry and strain in a system can be seen in the solvolysis rates of cyclic, homoallylic compounds as well as in the cyclo propylcarbinyl and cyclobutyl products formed. The lowest homolog of the 3-cycloallcene series, is the 3-oyclopentene system, *£8, The planar conformation of the ring prevents double-bond participation as indicated by the low rate of

solvolysis of 48 (X - Br or ONs) in comparison with corres ponding cyclopentyl derivatives.The products obtained from solvolysis of 48, were mainly 2 -cyclopentenyl compounds (50) arising from a hydride shift.^®*^ However, evidence for bicyclopentane (4£) formation has been seen in the reaction of ,4& (X = OTs) with lithium aluminum hydride or sodium borohydride. Rearrangement in the 3-oyclohexenyl system should be facilitated since the carbonium ion can assume a conformation 17 close to the homoallyl model (Figures II and III). Hanack, et. al. found the rate of acetolysis of 51 to be higher than that of cyclohexyl jD-toluenesulfonate and isolated the expected bicyclo [3 «1 *o] compounds on acetolysis and hydrolysis.^

+ Ts OH 51 52 53 7% cisj trans (?0#)

Because of their greater conformational mobility, 3- cycloheptenyl (5 5) and 3-cyclooctenyl (5 6) derivatives can

a 55 56 almost completely attain the geometry required for homoallylic resonance. As shown by Cope, et. al., 24*8 derivatives of these systems solvolyze much more rapidly than the corresponding cyclohexyl systems and the solvolysis is accompanied by exten sive rearrangement into the cis, trans-isomeric bicyclic derivatives. The 2-cycloalkenylmethyl compounds with small rings such as 5 7 ^ and 58^° do not undergo homoallylic rearrangement but 18

Ph CH2OTs 'GH2X Ph- n 58 instead form primarily ring-expanded products. A further increase in ring size in the 2 -cycloalkenylmethyl system allows the homoallylic rearrangement to predominate over ring expansion. 2-Cyclopentenylmethyl derivatives (59* X *= OTs or NH2 ) rearrange in solvolysis or deamination to the stereo-

c h 2x

C r 59 60 isomeric bicyclo [3*1* 0J hexyl compounds (60)* The kinetics also show a neighboring group effect with ££ (X = ONs) sol- volyzing 10 times more rapidly than the saturated cyclopent- ylmethyl ^-naphthylsulfonate.Homoallylic participation is also seen to occur in the methylenecyclohexylcarbinyl j>- toluenesulfonate system (6l) where the steric requirements for

CH, + OH

61 62 W- 63

homoallylic participation are met.-*1 The hydrolysis of 61 leads to W62 V and WA. 63 as well as 64 which is assumed to come from rearrangement of ion 65 which is the ring-expanded intermediate.

An excellent system for the study of the homoallyl re arrangement is the 1-cycloalkenylethyl system (6 6) which is

66. vW67 68 wW only influenced by slight steric effects and is one of only a few cases where cyclopropylcarbinyl and cyclobutyl compounds are formed simultaneously, Hanack and Schneider^ have shown that it is possible to obtain both spirocycloalkane (6£) and condensed cyclobutane (6 8) derivatives in good yield from hydrolysis of 66^ (n = 4,5*6*7) and thus the homoallyl rearrange ment becomes of synthetic importance. Similarly, Closson and Kwiatkowski have found the 2(/£s?--cyclopentenyl)ethyl tosylate

H “ 3 ) * to give bicycloj]3 *2 ,o]heptan-l-ol (69,) and spiro- [4.2 ]heptan-4-ol (?0 ),5 3 Since neither of these studies could determine whether the products were formed from a common nonclassical intermediate or two equilibrating ions, the solvolysis of 2 -(^-cyclobutenyl)ethyl tosylate (6 6* n - 2 ) was studied by Wiberg with this problem in m i n d . Acetolysis of the tosylate of 66 (n = 2 ) gave ?1 and ^2 as the major CHp (20# of D) lb )2 (80# of D)

72 products with no trace of homoallyl-type products (6 7, 6 8). Hydrolysis in aqueous acetone, however, did give a trace amount of VVV67 (n ” = 2) although no 66 was formed. Acetolysis of the 1 ,1-dideuterio derivative of SA/V66, (n = 2 ) gave v\A/71 with deuterium scrambled between the terminal methylene and position 2 of the cyclopentyl ring. This result led to the postulate that this rearrangement proceeds via direct form ation of a spiro[*U2 .3lhexyl-4-cation J ( U\/W7 3)* as well as the bicyclo [2.1. o]pentane- 1-methyl cation (j£j^). Formation of 7^ can lead to V7 /*t W but W7 V^ V must be formed firectlyv as well in order to explain the isotope distribution found in the product. 21 The solvolytic rates and subsequent rearrangement; of allylio and homoallylic systems described previously have provided considerable insight into the mechanism of these displacement reactions as well as a synthetic entry into new systems. An important aspect affecting the facility of i ■ double bond participation is the stability of the inter mediate cation formed in the reaction. That is why many solvolysis reactions involve aromatic and homoaromatic inter mediates. It is for these reasons that the study of the neighboring group effect of cyclooctatetraene would be an interesting one. Not only is the.ability of a double bond of i the cyclooctatetraene system to participate in a solvolytic reaction unknown, but the possibility exists that the kinetics and products of the reaction could provide evidence for the intermediacy of the stable homotropylium cation. Aromaticity has been associated with Increased delocal ization energy of a substrate relative to an acyclic analog as well as the ability of a substance to sustain an induced ring current.-'*' Homoaromaticity has basically the same definition although it is a term generally associated with nonclassical electron delocalization employing, In part, over lap of orbitals which are not entirely TT In nature. In thel case of the homotropylium ion, there is found a six electron homoaromatic system whose preparation and electronic structure have been well established,55 In 1962, Pettit and coworkers^ reported the first 22 generation and direct observation of the monohomotropylium cation which was generated by protonation of cyclooctatetraene in concentrated sulfuric acid-^a* 57 and was isolated as a solid s a l t , -56a CgH^SbCl^ , by treatment of the hydrocarbon with HC1 and SbCltj. The nmr spectrum of the ion (75) showed a value of 8 .5S for protons H2 -H5, a value of 6 .^2 8 for protons and Hr,, a value of 5*10S assigned to Ha , and a value of

-O.67S assigned to H^. Regarding the structure of this CgH^+ ion, the planar cyclooctatrienyl structure (VVW7 7) was excluded Hb Hb Ha Ha

75 v w 11w w since it would be expected to have magnetically equivalent

Ha and H^, methylene protons. The classical structure 76, was also excluded since one would not expect to see a value of 6.42& for protons and on a fully-formed cyclopropane ring by analogy with known cyclopropyl-substituted carbonium ions. Thus the homotropylium ion was assigned 75 and the ex- VW 1 tensive shielding of H-g provided good evidence for the presence of a sizable ring current. Subsequently, a detailed study of the protonation of cyclooctatetraene transit!on-metal carbonyl complexes was made. This proved to be a powerful tool in the study of homoaromaticity since the different electronic re quirements of the transitlon-metal atoms in their complexes 23 exert substantial control on the resulting structure of the

C3H^+ ion. For example, protonation of cyclooctatetraene molybdenum tricarbonyl (7 8) with sulfuric acid yields the homotropylium

(-0.15&) Hb Ha (3.376)

6.O-6 .5 8 1 (V.27 &) o( CO) Mo(CO)3 (co)3 79 80 complex in which all six *Tt electrons of the ion are used in bonding to molybdenum. Because of the 6 Tf electron demand of molybdenum, the structure of the. protonated species is limited to 79 or 80. However, only 79 is able to account for the large observed chemical shift difference between Ha and Hb. This assignment is strengthened by the analogy between

79 and the free ion, 7 5 , An early study of the protonation of the cyclooctatetra ene iron tricarbonyl complex In sulfuric acid at 0° showed the structure of the resulting ion to be best represented by the classical structure, 8 2,^® in which the cyclopropane ring remains closed and only the four IT electrons of the penta- dienyl system are used in bonding to iron. However, a more 24*

(1.53&) Hb Ha (1.35S)

Pe(C0)3 Pe(CO) 3

81 82 recent s t u d y ^ 9 has shown that protonation of 81 at -120 ° yields 8^ which contains a five carbon pentadlenyl system completed to iron as well as a "free” double bond. When the

5.888 h1,5 4*.54*8 H 5 .8 8 8 H3 6 .7 7 8 83 reaction temperature is raised to -60°, 83 undergoes a clean* first-order ring closure (k « 3 x lO"^ sec'^.AG^ ~ 15.7 kcal/mole”*) to the previously-reported^® bicyclo[5.1 .o]octa dienyliron complex, 82. Thus the available data concerning

the Gq H^"** ion favors a nonclassical, homotropylium represent- ation for the free ion and the Mo(C0) 3 complex. This picture involves 1 ,7-overlap of a type intermediate between d* and If and can thus be called "homoaromatic”. The presence of a ring current in these ions can be used to explain the large chemical shift difference between the "inside” and "outside" 25 methylene protons.^® This chemical shift difference has made it possible to measure the rate of ring inversion in the homotropylium ion which has led to a value of 2 2 .3 kcal/mole“^ for the energy difference between the classical cyclooctatrienyl cation (77) and the preferred homoaromatic one (75)*^ Huisgen and coworkers 61 have recently reported a chemical shift difference between "inside" and "outside" protons in homotropylium ions with an 8-chloro substituent which is identical to that for the unsubstituted ion. They also found the rate of equi librium of the epimeric 8-chloro ions to be essentially the same as that of the unsubstituted analog. Further Insight into the electronic structure of the homo tropylium ion has been provided by its uv spectrum^® which shows a closely resembling that of the tropylium ion rather than that expected for a classical, planar cyclooctatrienyl cation (77)* 6? In addition, direct evidence for the aromatic ring current in the homotropylium ion comes from its volume diamagnetic susceptibility measured by Dauben^^ which was found to be slightly larger than that of the tropylium ion. Honohomotropylium ions with 1-methyl, 1-phenyl, and 2 -hydroxy substituents, ^ ^ * 0 as well as the 1-hydroxy derivative^ have been found to be decidedly homoaromatic as confirmed by diamagnetic susceptibility measurements,^ In view of the appreciable stability of the homotropylium cation, its intermediacy has been used to explain the 26 mechani-stic aspects of some electrophllic addition reactions to cyclooctatetraene. Huisgen and coworkers 61x have estab lished that the chlorination of COT occurs via endo-8-chloro- homotropylium cations (8^) which subsequently undergo kinetic- ally-controlled nucleophilic attack by chloride ion from the endo side to give predominantly cls-7»8-dlchlorocycloocta- tetraene (85). Similarly, the addition of chlorosulfonyl-

C1 Clp Cl t \ ^ H Cl 85 isocyanate to COT at 50° ih the absence of solvent has been reported to involve the intermediate dipolar homotropylium cation 86 which then collapses to the observed 1,4-cyclo-

+ ClOoSNCO »

86 addition product, 8 7 . ^ In light of these findings, one lA/v would therefore expect the neighboring group effect of the cyclooctatetraene molecule to be of great interest since the participation of one of its double bonds in anchimerio fashion could lead to the intermediacy of the stable homotropylium cation. It was for this reason that the following study of the solvolytic behavior of cyclooctatetraenylmethyl (8 8) and »a A /3-cyclooctatetraenylethyl (8 9) derivatives was undertaken* 1 v W RESULTS AND DISCUSSION

Evaluation of the neighboring group effect of cyclooctatetraene (COT) began with a study of the behavior of cyclo-

octatetraenylcarbinyl derivatives (8 8) during solvolysls, l A / V

!H2X

Cyclooctatetraenylcarbinol (88-OH) was prepared In the following manner, A solution of methyl propiolate (90) ,in benzene was placed in a quartz vessel and irradiated (2537 & ) for a period of days to give carbomethoxycyclooctatetraene (91). Reduction of 91 In the usual manner using lithium

H-CSC-CO2CH3 I I

WV\

LiAlHjj,

y==rsCH2 0DNB = \ ^ C H 2C1 SOCIp,Py Q r CH20H DNBC1 * c x ^CHCI3 *>y 88-Cl vVw88-OH 88-ODNB

28 29 aluminum hydride gave 8£-0H whose infrared spectrum was identical to that of the cyclooctatetraenylcarbinol prepared by the copolymerization of acetylene with propargyl alcohol. 67 ' Although the yield of the copolymerization process was slightly greater than that in the above preparation, the latter proved to be more convenient with regard to laboratory execution as well as product purification. Treatment of 88-OH with 3,5-dinitrobenzoyl chloride in pyridine gave g8-0DNB In the form of pale, yellow needles. An attempted solvolysis in

aqueous acetone at 80° showed 88-ODNB to be completely unreactive. Subsequent attempts to prepare a more reactive p-toluenesulfonate ester of v88 W were unsuccessful and it was for these reasons that W88-C1 v was chosen as the most suitable derivative for solvolytic study. Its preparation merely involved treatment of 88-OH with thionyl chloride and pyridine in chloroform to give 88-C1 as the sole product. To provide a basis for comparison, the A^-cyclooctenyl- carbinyl system (92 ,) was chosen for a parallel study in order to determine the neighboring group effect of an isolated

W92 ~ v double bond contained in an eight-membered ring. The parent alcohol (92-0H) was prepared by the following sequence. 30 2-Carbomethoxycyclooctanone (9*0 was chosen as the starting material and was prepared according to the procedure of Krapoho, et, al* 68 This involved reaction of cyclooctanone

(J9 3 ) with sodium hydride and diethylcarbonate in benzene* Subsequent hydride reduction 69- of ketoester g£ i gave three products which were identified as the two Isomeric alcohols

93 LiAlH/j,

CH2 0H

OH 92-OH 9 5-OH 96

CH2 0DNB

tm 92-C1 92-ODNB

^AA92 and .95 Vv\ as well as the diol 96, Simple distillation served to separate the alcohols from the diol and subsequent vapor phase chromatography was used to isolate the desired £2-0H in pure form. Treatment of g2-OH with thionyl chloride and pyridine in chloroform gave a 2 :1 mixture of 92 -C1 and 95 -C1 A/V which could only be separated by vpc. Since similar treatment 31 of the distilled mixture (1:4) of 92-OH and 95-OH with thionyl chloride gave the same 2 :1 mixture of the corresponding chlorides, the tedious vpc separation of aloohols 92 and 95 was omitted in subsequent preparations. Treatment of 92 -OH with 3*5-dinitrobenzoyl chloride in pyridine gave S?2-0DNB in good yield. As with 88-ODNB, 92-ODNB proved to be stable * / W wVV upon attempted hydrolysis in 40# aqueous acetone at 80°. In addition, the corresponding p-toluenesulfonate ester of 92 proved to be too reactive to isolate and therefore, as with the cyclooctatetraenylcarbinyl system, the corresponding chloride seemed to be the derivative of choice for solvolytic study. Hydrolysis of 88-C1 and 92-C1 was performed in a medium Wi/ of 50# aqueous ethanol. The rate data and activation para meters are given in Table I. A good model to use as a basis for comparison is l-chloro-2 -methyl-2 -butene (9 7 ) which like v w 92-C1 is a /3- and Y-alkyl-substituted allyl chloride. Although the rate of hydrolysis of this compound has not been determined,

CH3CH=C(CH3 )CH2 C1 CH3CH-CHCH2 CI 97 98 the analogous l-chloro-2 -butene (9 8) can be used as an adequate model since ^-methyl substitution has been found to give only a negligible rate increase.Hydrolysis of S>8 (50# aqueous ethanol, 25 °) was found to proceed at a rate of 32 Table I, Buffered Hydrolysis Rate Data

Chloride Temp. °C k^, sec“* AH^, lccal/mol AS^, eu

2 5 .0 7.77 x 10“6a 3 0 .1 1.44 x 10“5 88-C1 - 20.7 -12.5 40.0 4.24 x 10*°

49.9 1 .2 6 x 10“**

25.0 3.83 x 10“5a 3 0 .1 7.20 x 10~5 92-C1 u 21.2 - 7.80 40.0 2.16 x 10“^ 49.9 6.60 x 10_/* aExtrapolated values based on activation parameters.

1.51 x 10"^ sec~* which indicates that !W92-C1 exhibits solvolytic behavior characteristic of a normal allylic chloride containing a 'Ylalkyl substituent. The fact that §2,-C 1 shows a five-fold rate enhancement over 88-C1 leads one to conclude that the COT ring is unable to provide significant anchimeric assistance. Furthermore, the results show the relative rate retardation of uvv88-C1 to be due to an entropy * w rather than an enthalpy effect. Hydrolysis of 92j-Cl at room temperature in 50# aqueous ethanol buffered with 2,6-lutidine gave the following products: 33

,CH20H n o m 50% aq. CsH^OH . — 2 ,6-lutidine ; a * cxiH 25° 92-OH 95-OH (2?#) (22#)

CH 2 0 CH 2 CH 3 y— y + Gy + N — A ) c h 2 c h 3 9 2 - O C H p C H o 9 5 - O C H o C H o /Vs* « y ^ J (33^) (19 ^)

These products were separated by vpc and the alcohols (92 and

9 5 ) were identified by comparison with authentic samples. Their corresponding ethers were identified by ir, nmr, and elemental analysis. The appearance of roughly 40# rearranged products is not surprising in light of the chlorination re action of g2 -0H which gave 33# rearrangement. Hydrolysis of 88,-Cl under the same conditions as above gave the following:

ch2oh ch2 och2 ch3 CHi 1 CH2 CH0 8 8 -0 1 50#_aq, C2HgOH W /==%/ + ~ 2 ,6-lutidine } 88-OH 88-OCH2 CH3 2% (37#) (17#) (ifO#)

Compound 88-OH was identified by comparison with an authentic sample, whereas 88-0CH2 CH3 was identified by spectral means and by unequivocal synthesis from §8rOH, potassium tert- butoxide, and bromoethane. The synthesis of o-tolylacet- 3 k aldehyde (.92) bas teen reported in the literature,^ however the melting point of its seraicarbazone derivative (196 °) did not compare with that of 99 (176-7°). Therefore it became necessary to prepare an authentic sample of o-tolylacetaldehyde for comparison. Commercially available o-tolylacetic acid (1££) was reduced with lithium aluminum hydride to give the corres ponding ^-(o-tolyl)ethanol (101). Subsequent oxidation with

CH CH !H2 CH2 0H L1A1H

100 101

Collins reagent gave the desired o-tolylacetaldehyde which was found to be identical with the solvolysis product upon comparison of their individual spectra, vpc retention times, and semicarbazone derivative. The alcohol and ether products obtained from the hydrol ysis of 88-C1 presumably arise from simple Sn2 displacement by solvent. However the mode of formation of 99 posed a VV* mechanistic problem and thus a dideuterio chloride (102-C1) was prepared by reduction of carbomethoxycyclooctatetraene

(91)■ A A-l with lithium aluminum deuteride followed by chlorination

91 L1A1D^ , ,d 20H SOClc, Py k A-A* ^ CHCI3 * 102-OH . 102-C1 with thionyl chloride. Hydrolysis of 102-C1 as predescribed gave the corresponding alcohol (102-OH) as the major product (75%) as well as the dideuterioether (lO^-OC^CH^)

50% aq. CpH^OH c h 2c h o o c T” *-idine ' /

102-0H IO2 -OCH2 CH3 lOJ {75%) (1<$) (15#) in 10% yield with only 15% of dideuterioaldehyde (103) formed. The aldehyde, however, was found to have all the deuterium incorporated in the tolyl methyl group with no apparent deuterium loss. The diminished proportion of dideuterioaldehyde 103 in the product mixture compared with that of its unlabeled counterpart 2 % could be due to a secondary deuterium isotope effect which suggests that the ionization of 88-C1 is rate-determining only in the case of the aldehyde formation. Hydrolysis of 88-C1 in a D2 O/C2 H5OD medium gave no deuterium incorporation. In addition, ^ cannot arise from further re arrangement of 88-OH or 83-OCH2 CH3 since both were found to be stable to the reaction conditions• These experiments Indicate that ionization of 88UC1 leads to a cationic intermediate which readily reacts with water to form a benzenoid product in which the methyl and acetaldehyde substitutents are in an ortho arrangement. These requirements are best met by the intermediacy of cations 10^f and/or 105 (Scheme I), whose formation is very reasonable in light of the 36 Scheme I.

CH 104 105

h 2o , -it H20, -H

■3 106 107WvV

•V

CH CH 108 homoaromatic nature of the homocyclopropenium ion.'“72 Such carbocatlons have been found to have an unusually Important 1,3-interaction due to the restraint of the four-membered ring which creates an unusually small 1,3-distance, Capture of 104 and 105 with water (presumably from an exo direction) gives rise to 106 and lO^, respectively. Subsequent conrotatory ring opening gives an ortho-substituted aromatic 37 product (108) which leads to the observed o-tolylacetaldehyde

(99)u w by a tautomeric shift of hydrogen. Evidence for the formation of homocyclopropenlum Ions

of the general type wvV 111 has been found previously in a study of the kinetics and mechanism of the rearrangement of bromo-

cyclooctatetraene (109-Br)1/ W % to trans-bromostyrene (113-Br).^^a w

Scheme II

110

‘■-2

c=c butene derivative (112-Br) which leads to trans-ft-bromostyrene (llJ-Br) upon conrotatory ring opening* The con version of 1,05 — > ll^ was found to be quantitative and highly stereo-selective, with the trans isomer being accompanied by only small amounts of the els (1-3#)* Rearrangement of 109-Br in the presence of lithium iodide yields trans-J?- iodostyrene (11^-1) in addition to lljJ-Br which indicates the presence of a reversible ionization step. Solvent dependence ' t of rate further indicates that this step is rate-determining although the preionization valence tautomerism was found to be rate-determining in solvents of high polarity such as DMSO, acetonitrile, acetic acid, and methanol. Prom the values of Kexp for 109 113 in acetic acid over a range of

50°, a AH* of 2.31 ± .0 5 kcal/mole"1 and aAs+ of -9.5 ~ 1.5 eu were evaluated, corresponding to a barrier of valence isomerization of 109 —* 110* A similar rearrangement was carried out on 1,4-dibromocyclooctatetraene (lli£)'^t) which was found to give a £,p-dibromostyrene (117), presumably by a mechanism similar to the one proposed for the rearrangement of 109* The positions of bromine in 117 substantiated the Br

115 116 39 occurrence of a 1,3-miez‘ation of bromine. Furthermore, 71} Criegee, et. al. ,' have recently studied the thermal Ion ization of substituted benzoblcyclooctatrlenes (118) to vlnylnaphthalenes (120) and have assumed an analogous re action path to explain the observed stereospecific

118 119 120,

isomerizations. The initial valence tautomerism prior to isomerization found in the bromoCOT to bromostyrene rearrangement can be invoked to explain the absence of anchimeric assistance in

the hydrolysis of 88-C1%#w since its reaction rate would depend upon a preequilibrium (K^k^/k^, K'^k^/k^) which, because It does not favor the monocyclic form,?5 would comprise a sig nificant rate-retarding component to the observed rate constant (k0bS=Kk2^* 0ne cannot readily distinguish between the form ation of 121 versus 122 in Scheme III since hydride shifts lead to similar homocyclopropenium ions (10^ and 10£) which have both been shown to lead to o-tolylacetaldehyde in Scheme

I. Ionization of 121,v however, * does lead to a cationic intermediate (12j3) which is capable of stabilization through over lap with the proximate cyclohexadlene system. In contrast, one would not expect significant stabilization in the case

Scheme III

CH2C1 CH2C1

-1 -1

123 D-a I shift

104 105 of 124 where the developing positive charge is not in very good position to be delocalized by the adjacent jf system. On this basis, one might predict 88 —> 104 to be the more favored process, although one cannot eliminate the possib ility of a process involving halide ion departure with con comitant hydride shift to produce 104 and 105 directly. To date, the only evidence for a solvolysls reaction preceded and initiated by valence tautomerism such as that postulated in Scheme III has been reported by Sargent, et. al.for the solvolysis of (7-cycloheptatrienyl)methyl 3,5-dinitrobenzoate (125-ODNB). Solvolysis in the presence of excess base was found to give primarily 125-OH as well as

HOAc Urea ^ 'HoODNB

125 126

•v

128 127

styrene (128). Since 125-OH proved to be stable under solvolysis conditions, styrene must be a primary product of the re action. These results forced the authors to conclude that the solvolysis of 12£-0DNB involved a prior isomerization to its valence tautomer 126 followed by ionization to the cyclopropylcarbinyl carbonium ion (127,) with concomitant fragment ation of the cyclopropyl ring to produce styrene. The possi bility exists therefore that hydrolysis of ,88,-Cl could provide a second example of a reaction, initiated by valence .tautomerism, One cannot overlook the possibility that 88-C1 may 42 isomerize directly to an ionic intermediate such as 129 which V/v%» then could experience valence isomerization to 123 (Scheme IV) and rearrange further by the mechanism shown in Scheme I to give the observed o-tolylacetaldehyde (99). Internal return

Scheme IV

-Cl 88-C1

123 Cl“ i i < — - 104

131 130

104 99

to give 130 is also a possibility since this substituted chloroCOT could now experience valence tautomerism and ion ization similar to that observed by Huisgen?^ for chloroCOT itself to give the homocyclopropenium ion 104 which then gives 99 as in Scheme I* Due to the lack of anchimerio assistance of the COT ring as evidenced by the solvolysis rate of 88-C1, one must eliminate the possibility of 129 ^3 having a flattened, delocalized homotropylium-type structure such as 133. Instead, the tub-shaped nature of 129 not only

133 prevents any delocalization of positive charge throughout the entire Tf system but misaligns the £ orbitals of the carbonlum ion and adjacent double bond In such a manner as to generate inductive destabilization. This misalignment could cause destabilization sufficient to account for the rate-retardation observed in the hydrolysis of 88-C1, v W An example of the pronounced effect this type of in ductive destabilization can have is found In a study of the allylic adamantyl system vW'**13^,^^*^® The structural rigidity Inherent in this molecule holds the adjacent double bond perpendicular to the developing carbonlum Ion center and rate

13^ decelerations on the order of 10^ have “been observed*77*78 In the case of 88-C1, molecular rigidity is not as pronounced and one would expect a significant but smaller rate decele ration. It is difficult to estimate what the degree of rate retardation would be in the case of 88. If one examines a model of 129* it is seen that the developing p orbital of the carbonlum ion resides at an approximate 45° angle with theTt orbitals of the' double bond (assuming a "tub” conformation). In light of this observation, one might expect a larger de celeration effect that the mere five-fold rate decreass found. And Indeed, in the solvolysis of ^-cyclooctatetraenylethyl derivatives (to be discussed subsequently) where a cycloocta- trienylium intermediate similar to 129 is indicated, a larger deceleration effect is observed. In order to further investigate the neighboring group effect of cyclooctatetraene, the solvolytic reactivity of £>- cyclooctatetraenylethyl derivatives (89) was studied. The

Vw parent alcohol, 8

Br Br Br, KOt-Bu

135 136 137

n-BuLi 89-OBs BsCl,Py Li OH 1) . V . 2). H30h DNBC1, py ^ 89-OH 138

as pale, yellow needles. However, LrW 89-ODNB proved to be unreactive when treated with 20JJJ aqueous acetone at 85° nnd thus 89-OBs was prepared, by reacting 89-OH with brosyl chloride « A s ■>A O in pyridine. The analogous ^-(A^-cyclooctenyl)ethyl (13§) and cyclo octylethyl (1^0) derivatives were also prepared in order to provide a basis for comparison in the solvolysis of 8£. Preparation of 139-OH was accomplished by two separate

139 1^0 vw* Q n sequences. The first involved a Reformatsky reaction in which cyclooctanone (1^1) was treated with zinc and ethyl- bromoacetate in benzene to give the corresponding hydroxy- ester (l*f2) in only 30# yield. Treatment of 1^-2 with phosphorus tribromide gave the corresponding bromide 1 which was reacted with ethanolic KOH at room temperature to give .OH

Ibl 1^2

B-r LiAlH c h 2c o 2h k o h ' C2 H5OH 139-OH l4if 1^3

DNBC1 vPy 139-OBa 139-ODNB k ?

A -cyclooctenylacetic acid (l*f*i-).wVwV It was not possible to carry out the dehydrobromination without concomitant sapon ification of the ester with this reagent. However, none of the exocyclic double bond isomer was obtained which was not the case when other bases were used. (For example, treat ment of 1*£2 with triphenylphosphine dibromide in DMF at 100° gave the ethyl ester of l*^ directly, although It was con taminated with approximately 33# of the exo Isomer). Re duction of 1kjj; with lithium aluminum hydride in the usual manner gave 13g-0H, as expected. Alternatively, 1*£2 was pre pared by a similar sequence involving treatment of l*jOL with lithium hexamethyldisllizane®3, and ethylacetate according to the procedure of Rathke.®2 This latter method for preparing

+ LiN[si(CH-j )^] 3 + CH3CO2CH2CH3 CH2 CO2 CH2 CH3

l*fl 1*4-2

{3-hydroxy esters is less tedious and time consuming than the usual Reformatsky procedure®0 and gives much higher yields (an increase from 3 ° % to 95% in the case of lk2). To complete the sequence, 1*4-2 was then treated as before to give 139-OH which was treated with 3 ,5-dinitrobenzoyl chloride in pyridine to give lJg-ODNB. After heating lJg-ODNB at 125° in 20# aqueous acetone for over 2k hrs., no evidence of hydrolysis was found and the starting material was recovered in good yield. There fore the corresponding brosylate (13g-0Bs) was prepared as In 48 the ^-cyclooctenylethyl series, and isolated as a nearly colorless oil which could be purified by low temperature re crystallization. Hydrogenation of 139-OH using a platinum oxide catalyst gave a quantitative yield of 140-OH, Treatment of this O.— ^ 0H Cr >— \ o ^ OBs 139-OH iitO-OH 140-OBs

alcohol with brosyl chloride in pyridine gave 140-OBs as a colorless oil. Like 139-OBs, this derivative was purified by low-teraperature recrystallization. Acetolysis of brosylates 8£, 1^2* and 140 showed good first-order kinetic behavior in all three cases. The rate data and activation parameters are given in Table XI. The relative rate of solvolysis of (2 6 0 ) is compar able to the kunsat^ksat = 350®^ reported for the acetolysis of 4-methyl-3-pentenyl ^-naphthylsulfonate (14^)*

OBs (CH3)2-CH=CH-CH2CH2ONs \

145, 146

The value of ^unsat/^sat ^or acet°lysis l4§r was only 49

Table II, Buffered (NaOAc) Acetolysis Rate Data

Brosy- Rel .1 AS*, late Temp, °C kjL 1 sec" kcal/mol AS*,eu rate, 65*

89-OBs 55.00 5 .02 1.91 X 1 0 -6 65.0 0 ± .03 6.58 X 1 0 "6 25.3 -7.9 5

75.00 ± .03 1 87 X io-5

l0-6a 140-OBs 6 5 .0 1.43 X

75.00 ± .0 3 4.69 X 10“ 6 . 24.9 -11.8 1

8 5 .0 0 ± .03 1 .1 6 X 10"5

95.00 t .03 3.25 X lO" 5

139-OBs 4 5 .0 0 ± .02 3.^3 X 10“ 5

* 55.00 ± .03 1.25 X 10“^ 24.9 -0.78 260 +1 O cn 6 5.00 • 3.74 X 10~^ aExtrapolated value based on activation parameters. 50 40 which reflects the effect strain and ring size can have on homoallyllc solvolysis in these systems. Thus the expected anchimeric assistance seems to be operative in the case of 13^, although the amount of participation occurring in the acetolysis of 82»OBs seems to be negligible by comparison, as the results show its solvolytic reactivity to be more closely allied to that of the fullyw saturated l40-0Bs. v w \ One must also note that the enthalpy value is very similar in all three cases with the activation parameters showing variance in the entropy term alone. Solvolysis of 140-OBs in an acetic acid/sodium acetate medium at 85° for a period of ten half-lives gave 140-OAc as the sole product. (No elimination product(s) could be detected by nmr or vpc analysis). It was identified by its

HOAc/NaOAc OCCH

140-OBs 140-OAc ir and nmr spectral data as well as its elemental analysis. An alternate synthesis of 140-QAc involved treatment of 140-OH 0 OCCH3

140-OH 140-OAc with acetic anhydride in pyridine; the ir and nmr spectra of this acetate were Identical to those of the solvolysis product. 51 Similar solvolysis of 139-OBs at 50° in an acetic acid/ fc»WV sodium acetate medium for a period of ten half-lives using a ten-fold molar excess of NaOAc gave the following products:

> Q , . o f ) OGCHo OCCH' uuuno II J J 0 139-OBs 139-OAc 3>7-OAc 148-OAc ^VW VVV'V* WVV s/v^yV < 3 W (10^) {55%)

These products were separated byvvpc and tentatively identified by ir, nmr, and elemental analysis. The internal return prod uct, 139-OAc, was alternatively prepared by the reaction of 139-OH with acetic anhydride in pyridine. Its isomer, lJ^-OAc, 0 OH + (ch3c-)2o — > r 3

139-OH 139-OAc was prepared in the same fashion from l4£-0H which was pre pared by the following sequence. Cyclooctanone (1^1) was refluxed with cyanoacetic acid and ammonium acetate In benzene with removal of water to give 1^9 as expected.

i 52

CN t /CH-C02H CH2CN NHj^OAo II A (j)H a <-h 2o ) o 149 150

L1A1H,

OH 'H NaNOp Q r ~ O 139-OH 147-OH 151

(CH3§-)20,Py

147-OAc

Pyrolysis of l4

Zn*Cu ^ o « * 95-OH 148-OH 0

lW-OAo anhydride in pyridine gave the desired 148-QAc which proved to he identical to the solvolysis product. As might be expected, the product ratios in the solvol ysis of lj|g-OBs were found to vary with the amount of buffer used. Using only a 10# molar excess of sodium acetate, 148-QAc was found to comprise 59# of the product mixture while 14£-QAc accounted for the remaining 41#. The fact that spiro [2 ,n] and bicycloQn,2.0] compounds are extremely un stable was pointed out by Wiberg^^ who was interested in pre paring 1-hydroxybicyclo[4.2 .0] octane (68-OH, n = 4). Pol- lowing the procedure of Hanack, et. al., 84 he found that he

OTs

CaCO +

66-OTs 67-OH 68-OH (n = 4-) Hanack (33#) (65#) Wiberg (85#) ( 5#) 54 obtained different product ratios. Wiberg claimed that an

equilibrium exists between 6? and 68 in dilute acid with w\A68 predominating. Under his reaction conditions (unstirred), 68-OH Is the major product. Under Hanaclt* s vigorous stirring conditions, however, the buffer is efficient enough to pre vent equilibrium and 6? Is the major product. It was for this reason that in the solvolysis of 139-OBs, a ten-fold molar excess of buffer was used to eliminate the possibility of acid catalyzed rearrangements. As a result, the internal return product 139-OAc was found in significant amount presumably due to an Sn2 process. 1 The product ratios found in the solvolysis of 139-OBs

can be compared with those found by H a n a c k ^ 2 upon the hydrol ysis of 139-OTs in 20# aqueous acetone. The ratios are comparable although in the case of 139-OBs, internal return to give 139-OAc seems to occur at the expense of 148-QAc formation.

OTs — 0H OH 20# aq.acetone + O CaCO~ - > N — ' \—/

139-OTs 139-OH 147-OH 148-OH

< 5* > (15t) (80)6)

Solvolyses of 139-OBs at temperatures higher than 50° were found to give increased amounts of elimination products and even at 50° these hydrocarbons were found to comprise as 55 much as 30# of the mixture. Hpwever, careful examination of the crude solvolysis mixture by nmr showed negligible amounts of elimination products formed and therefore it is reason able to conclude that these hydrocarbons arise primarily in the vpc purification procedure. Partial and sometimes com plete elimination, of acetic acid from the isomeric acetates was found to occur with many vpc columns to give hydrocarbon mixtures. Thus the instability of the splro[7.2 ]decane (157-OAc) and bicyclo j_6,2.0]decane (l*£8-0Ac) products'became evident. The possibility of acetates 139, an<* •?>£§ undergoing solvolysis themselves was considered and the products'were each subjected to the solvolysis conditions at 50°, for a period of ten-half lives of 139-OBs. As might be expected,

139-OAcv w was found to be stable under these conditions'and was shown to be isomerically pure by vpc. However, l**8-0Ac was found to solvolyze to some extent to both 139-OAc and 1^7-OAc. Under the same conditions, 1^7-QAc rearranged to only 139-OAc vVs/

j—s-^OAc /—v/N, HQAc/NaOAc , f f + f JZ1 + f JT c tOAc 50° — ^' —'/^0Ac 148-OAc 139-OAc 14^-QAc 148-OAc <31#) (8#) (61#) with no 148-OAc being detectable by vpc* The possibility that 56

/ \_/V'OAc HQAc/NaOAc ^ C p 50° k J OAc 147/OAc 139-OAc (52 ^) (48#)

139-OAc and 147-OAc are secondary solvolysis products cannot be eliminated, although this possibility seems to be a remote one. Analysis of the solvolysis mixture afterone-third of a half-life showed the presence of all three isomeric acetates in approximately the same ratios that are found after ten half-lives. OAc

q ~ “ \ c b . C X o

139-OBs * 139-OAc 147-OAc 148-OAc (3350 (24#) (43#) The results obtained from the product study of l^g-OBs do not seem to indicate the intermediacy of a bicyclobutonium- type cation. Product studies in the allylcarbinyl chloride system have shown the formation of cyclopropylcarbinyl, allyl carbinyl, and cyclobutyl products to be fofmed in a ratio of 4,5:1:4,5» and this mixture is considered to be a standard one where the equilibrating bicyclobutonium ion system is involved. In the case of lJ^-OBs, this ratio is 5.5'3.4:1 which is very similar to the 7.6:3.8:1 ratio obtained In the solvolysis of 148-OAc. Thus it seems that the mechanism can 57 be more accurately described by assuming separate by rapidly isomerizing carbonlum ions as suggested by Hanack and Schneider.^9 Solvolysis of the homoallylic ester 139-OBs would be expected to involve a relatively slow dissociation

to the homoallyl cation (1 5 2 ) which has generally little delocalization as evidenced by its high energy content.Rapid isomerization would then occur to the cyclopropylcarbinyl

(l/WV1 5 3) and cyclobutyl (15*0 cations which experience greater nonclassical stabilization. The fact that 1^8-OAc is the major product suggests that 1£2[ may be more stable than 15j*» although the product composition also could be influenced by the activation energies Involved in the formation of 148-OAc and 1^7-OAc from 153, a*id respectively. Thus the product composition could be under kinetic control. Due to the

Q p ^HOWNaOA.. Q p + — > Q f t * = * Q d

139-OBs 152 153 15^

I 148-OAci 147-OAc inevitability of Sn2 displacement reactions occurring in this solvolysis as well as the possible occurrence of acid-catalyzed interconversions, one cannot categorically state that the bicyclobutonium ion does not play a role. However, the product ratios do suggest that a rapidly isomerizing carbonlum ion mechanism operates instead. 58 Solvolysis of ^5-cyclooctatetraenylethyl brosylate (89-OBs) In acetic acid/sodium acetate at 75° for a period of ten half-lives gave 89-OAc and 1^-OAc; while 1,2-dihydronaphthalene (jJ56), naphthalene (157)* and tetralin (^58) were formed at low buffer concentrations. As shown in Table III, the product ratios varied greatly with variation in the buffer

Q r HOAo/NaOAc; ^ ^ ^ >PAc +

89-OBs 89-OAo 155-OAc 156

+ 00 157 158 concentration, Solvolysis with one molar equivalent of base I gave dihydronaphthalene (156) as the sole primary product with disproportionation accounting for the minor amounts of naphthalene and tetralin formed. With ten molar excess of buffer at a concentration of 0.2 N, 15£-0Ac was the maJor product with only 12# of 1J6 formed. Using a greater buffer concentration (2.0 N) greatly increased the amount of internal return product (8^-OAc) strongly Indicating an Sn2 mechanism. Increased buffer concentration also caused a decrease in the amount of 156 formed which suggested the possibility that further reaction of 155-OAc could account 59 Table III. Product Variation During Acetolysis of 89-OBs,

Molar proportion, ^ Product composition, %

Hun NaOAc Concen, 89-OAc i55-OAo 156 15§

A 1 0 .1 N — 86 7 7

B 10 0 .2 N 19 69 12 — C 10 2.0 N k9 bQ 3 __

for some of the dihydronaphthalene formation. Indeed, treat ment of 155-OAc with £-toluenesolfonic acid in acetic acid causes almost total conversion to 156. This result suggests

that perhaps 155-OAc and 156 arise from a common reactive intermediate, with formation of 1^-OAc being favored in a medium of high acetate concentration. The minor reaction products J57» 2111,1 were Identified by comparison of their spectral data and vpc retention times with those of authentic samples. The internal \ return product, 89-OAc was alternately prepared by reaction of 89,-OH with acetic anhydride and pyridine to give the

Py 0 — 0H+ (c h 3L 2o - s l , 0 r ° Ao

89-OH 89-OAc 60 corresponding 8g-0Ac which proved to be identical to the sol volysis product. The gross structure of the major product, 155-OAc, was assigned on the basis of its conversion to cis- v w * 1 88 perhydro-l-azulenone seraicarbazone, mp 220-1°). Reduction of 155-OAc with lithium aluminum hydride in the usual manner gave the corresponding alcohol, 155-OH. All

OAc D C C ^ . DMSO 1 H 155-OAc 156

H2 , Pd/C

DCG 0 5 , ;Q DMSO PH 160 ljg H attempts to oxidize ketone gave either dihydro naphthalene (T56) or unrecognizable mixtures. The alcohol was then hydrogenated to 152 an(* subsequent Moffatt oxidation gave l60 which was identified by comparison of speotral data and semicarbazone derivative with the known cls-perhydro-1- azulenone. The major product from the solvolysis of ,8£-OBs was assigned structure 155, although the tautomeric struoture 161 had to be considered also. A spin decoupling experiment on the corresponding alcohol showed no evidence of coupling bet ween protons and which one would expect to find for 61

OH OH

155-oh 161-OH structure appeared to be coupled to only Hg (J s

3Hz) and to only H2 (J = **Hz). However, no evidence of coupling could be seen between Hj and H3 which one would expect for structure 161. If one makes a model of 155* one can measure the dihedral angles between and Hg* and H^o*

H10 and H^, and compute the expected coupling constants from the Karplus equation.The 90° dihedral angle between

approx Karplus observed dihedral angle Jxy Jxy

Hi,H2 120° 3-A cps 4 cps Hi ,h 9 90° 0 cps 0 cps

h 9*h8 55° 2.5-3 cps 3 cps

Hi and H^ satisfactorily explains the absence of an obser vable A. model of l6l shows dihedral angles of about 20° and 95° between H^ and each of the H^ protons. Although a negligible coupling constant would be expected where the angle is 95 °* a dihedral angle of 2 0 ° would exhibit a large J value (ca 7 cps) and this Is definitely not observed. Thus the decoupling data can be interpreted in favor of structure 155. 62 ' In support of this tenuous evidence, a Europium-shift reagent was used in an attempt to verify the position of relative to the hydroxyl function. In the case of 155* one would expect a similarly rapid shift of and reflecting their vicinal relationship. However, in 161 where is further removed from the hydroxyl, one would expect to see Hq shielded to a greater extent than H^. An nmr sample of 155-OH was prepared by dissolving 0.25 mmole of alcohol in 250 ju-1 carbon tetrachloride (with 2#.TMS), Euroshift F was added in increments of 0.05 mole complex/ solute and a spectrum recorded after each subsequent addition. Shifts for all solute protons were to lower field values and increasing the amount of Euroshift F increased the resolution such that the spectrum became first order at a molar ratio of 0 .50. Table IV describes the shift of each proton with each subsequent addition of shift reagent. Earlier findings have shown a linear relationship to exist between chemical shift and metal ion concentration,^1 and thus five points were used to determine the locus for each proton. A least squares analysis was used to determine the slope of the best line in each case. From the slope and intercept of the line, a value for the chemical shift at a molar ratio of one could be determined. Using this value, the term AEu (the para magnetic-induced shift for each solute proton) can be found Table IV, Paramagnetic-Induced Shifts for Protons of 155-OH

Mole complex : solute Hx 0 .05 .10 .15„ .20 /cr50a 1.0a AEu Chemical shift (o)

OH 3.16 7.78 13.29 21.47 — 63.7 124.2 -121 Hi 1,90 2.78 3.66 4.64 5.64 11.3 20.6 - 18.7 h 2 **.77 5.08 5.44 5.80 6.16 8.2? 11.8 - 6.99 H3 5.96 6.22 6.50 6 .7 8 7.10 8.81 11-7 - 5.69 Hi* 6.40 6.54 6.72 6.90 7.12 8.21 10.0 - 3.61

H5 6.22 6.34 6.50 6.64 6.82 7.72 9.21 - 2.99 h 6 5.85 6.00 6.18 6 .3 0 6.48 7.44 9.02 - 3.17 H8sb 2.68 3.28 4.00 4.72 5.**8 9.73 16.8 - 14.1 H8a° 2.30 2.58 3.10 3.5** 4.00 6.66 11.0 - 8.71 H9sb 1.7** 2.58 3.36 4.24 5.1** 10.2 18,6 - 16.9 H9a° 1.66 2.10 2.56 3.00 3.50 6.24 10.8 - 9.15 H10 4.20 5.38 6.50 7.7** 9.08 16.3 28.4 - 24.2 aCalculated bSyn cAnti t 12.0 11.0

10.0 9.0

H 5

«6.0

5.0 8a 4.0 3.0

2.0

0 .25 .50 .75 Mole complex:solute Graph 1. Plot of chemical shift vs. mole complex: solute from Table IV, by computing the difference between the chemical shift prior to complexation and at a molar ratio of one. From a study by Demarco, protons located geminally to an OH function have been found to give rise to

a AE u value of -21,7 to -26.7 ppm. It can be noted from i Table IV that the AEu value for is -2^.2 ppm. Vicinal protons have been found to experience deshielding effects ranging from -8.8 to -19.2 ppm. This effect seems to be dependent upon the magnitudes of the dihedral angle between the hydroxyl and the proton under consideration. The following relationships were found:

e AEu (ppm) 3-5° -1 6 .7 to -19.2

60° -1 3 .6 to -1^.9

122° - 8.8 to - 9.3

180° - 8.2

In 1££* proton H^s has a dihedral angle approximately equal to 30° and the observed AEu is -9.2 ppm. But most import antly, with a dihedral angle of 90° has a AEu of -18,7 ppm which strongly indicates that it is vicinal to the hydroxyl. From the graph of chemical shift vs, metal ion concentration it is obvious that protons % and H^s are moving at a comparable rate which convincingly indicates structure 1££ over 161. It also provides strong evidence for the stereochemistry of the hydroxyl function to he syn to H^. A linear relationship has also been found to exist between AEu and R, the vector distance between the proton In question and the hydroxyl oxygen.^ In structure 161* the R for is about 4.5 corresponding to a AEu of -6 ppm. However in

155» the ™R for proton Hi ■** is 3 A° corresponding to a much larger AEu of -15 ppm. Since the observed AEu for Hj is -18.2 ppm, structure 135, Is again strongly indicated. Having conclusively identified the major product in the solvolysis of ^- O B s as 155,-OAc, the synthesis of deuterium- labeled analogs of 8g-OBs was undertaken in order to study their solvolysis products and attempt to formulate a reason able mechanism for the formation of 155-OAc. The cc»cC-di- deuteriobrosylate (163-OBs) was prepared in the following manner. Oxidation of 89-OH with Jones reagent gave the corresponding cyclooctatetraenylacetic acid (1 6 2 ) which was subsequently reduced with lithium aluminum deuteride in the

OH Jones ox. 2 I'lAlDii j I J ^

89-OH 162 I63-OH

BsCl,Py usual manner to give the dideuterioalcohol (16JI-0H). Treat ment of 16^J-0H with brosyl chloride in pyridine gave l6j-OBs which was solvolyzed in acetic acid/sodium acetate at 75° as predescribed. Analysis of the solvolysis products by nmr determined that the unrearranged acetate (163-OAc) did not undergo deuterium scrambling. However, the isolated

1,2-dihydronaphthalene (1 6 5) vras found to have deuterium

HQAc/NaOAc , [| 1 75° i§3-°Bs 163-OAc 16^-OAc 16§ distributed equally between and C2 . Nmr analysis of the isolated tetrahydroazulene derivative (^L6jJ-0Ac) also showed a scrambling of deuterium with the isotope dis tributed equally between Cq and C9 , Synthesis of the corresponding /3j, /3-dideuterioalcohol (168-OH) was carried out in the following manner. Ester- . ification of l6g with diazomethane (using N-nitroso-N-methyl- urea as precursor) gave the corresponding ester (166) which was treated with sodium methoxide and methanol-d^ to give the dideuterated ester 167, with greater than 95$ isotopic purity. Reduction of 16£ with lithium aluminum hydride gave the ft,^-dideuterioalcohol 168,-OH, which furnished 168,-OBs 68

'T=Ns- CH2 CO2 H CH2 CO2 CH3 QH2N2

162

NaOCH^,DOCH3

% 168-OBs ^BsCl,Py GO2 CH3

168v w -OH 1 6 7. upon treatment with "brosyl chloride and pyridine. Solvolysls of 168-OBs gave the unrearranged acetate (168-OAc) which

D 5 / 0 H 0 HQAc/NaOAc OAc 75° 168-OAc 16^-OAc 165 168-OBs showed no evidence of deuterium scrambling. Also formed were l6f>. and 16£ with deuterium equally distributed. between Cq and

C9 , and and C2, respectively. Thus in the solvolysis of

89-OBs the precursor to the unrearranged acetate 89-OAc and

155 and 156 cannot be identical. Formation of 89,-OAc may ■ simply be due to Sn2 attack by solvent, a conclusion supported also by the data in Table III. The fact that deuterium is equally scrambled between two carbons in both l£j> and 1^6 indicates that their carbonium ion precursor must be symmet rical. That the precursor to 155 (and 156) must be symmetrical needed further substantiation and it became extremely desir able to prepare a suitable derivative of 8j?-OBs which would label a position on the COT ring and further elucidate the mechanism operating in the solvolysis. As a result, the 2-methyl derivative of 89-OBs was prepared by the following route. A solution of ethyltetrolate (169)^2 and benzene in a quartz vessel was irradiated (2537 for a period of four days in a manner similar to that used in the preparation of carbomethoxyCOT (91). The product, 2-methylcarbomethoxy- cyclooctatetraene (170)^3 was subsequently reduced with

.CH

LiAlH CHo / 3 (^PCHOCH., (^- y C H3 MnO?/C CH=CH0CH3 ^— ^CHO

1) Hg(OAo) 2) NaBHi^.OH aluminum hydride in the usual manner to give the corres ponding alcohol, 1 7 1, which was then oxidized with MnOjj/C^ to give 1?2_in good yield. Treatment of 172^ with raethoxy- methylenetriphenylphosphine gave ylide 173. Hydrolysis of 173^ proved to be difficult since the usual protic acids such as perchloric acid would attack the COT ring rather than the ylide side chain. Reaction of 173, with mercuric acetate and sodium borohydride in base gave a mild,*clean hydrolysis to 1£4; and 17£ (2:1). This reaction proved to be an inter esting one in that the major product, 1^ . was a new compound namely 2-methylcycloocta [b]furan. In the cyclooctafuran series, only the cycloocta[cJfuran Is k n o w n , what proved to be even more interesting was the fact that treatment of

ch2oh y = \ y CK0 /=vch-choch3 MnQ?/C \\IJ ^PCHOCH^ y .fi jj

88-OH 176 177

1) Hg(0Ac)2 2) NaBH/j,, OH'

C r " 89-OH the parent cyclooctatetraenylcarboxaldehyde (1 7 6) with meth- oxymethylenetriphenylphosphine followed by subsequent hydro lysis with mercuric acetate gave only the alcohol, 89-OH in 71 nearly quantitative yield. Separation of 1?** and 175-OH was accomplished "by careful column chromatography and 175-OH was subsequently treated with brosyl chloride in pyridine to give the corresponding brosylate. Solvolysis of 175-OBs as pre described gave only the unrearranged acetate. No additional products could be observed by vpc or nmr analysis and thus a

CH ' 3 HQAc/NaOAc v >Ac O ^ x ^ O B s : 1£5,-OBs 175-OAc a methyl label in the 2 position seems to alter the solvol ysis mechanism sufficiently to preclude rearrangement. An attempt was therefore made to prepare a substituted derivative of 89-OBs where the methyl group would be further removed from the solvolysis center. Huisgen, et. al.,'^73b have reported the synthesis of 1,^-dibromoCOT by bromination with subsequent dehydrobromination of bromoCOT. A similar procedure was then used in an attempt to prepare l-bromo-4- methylCOT (17g) from methylCOT (1 7 8). When 179, was treated

178 179 180 with lithium dimethylcopper, the corresponding 1,^-dimethylCOT was formed which appeared to be one component by vpc and identical to an authentic sample.96»97 However, when 1££ 72 was treated with n-butyllithiura and ethylene oxide, vpc analysis showed the presence of a 3 :1 mixture of the 4-methyl and 5-methyl isomers* These compounds were painstakingly

OH 1) n-BuLi

BsCl,Fy BsCl,Py

181-OBs 182-OBs separated by vpc and the pure isomers were treated with brosyl chloride in pyridine to give the corresponding iso meric brosylates 181-OBs and 182-OBs which had significantly different melting points although their nmr spectra were nearly identical. The 1,5-isomer could not be obtained in large enough quantities to permit a solvolytic study and therefore only the behavior of the 1,4-isomer was examined. As a structure proof of 181-OH, a suitable derivative was prepared for X-ray i analysis. Treatment of 181-OH with iron enneacarbonyl in

Fe2 (C0)9 -- > U Ij .. PNBCl,Py ^ 183-OPNB

Pe(CO)3 181-OH 18 3-OH ether gave the corresponding Irontricarbonyl complex I83-OH which was purified by column chromatography and treated with £-nitrobenzoyl chloride in pyridine to give the corres ponding I83-OPNB* Three-dimensional X-ray analysis of 183-' OPNB^? has equivocally established the 1,4-placement of the substituents. Solvolysis of 181-OBs in acetic acid/sodium acetate at* 75° in the predescribed manner gave 20# unrearranged acetate with the remaining 80# of the product mixture composed of the two isomeric tetrahydroazulene derivatives, 184-OAc and 18g- OAc. These isomers could not be separated by vpc or any

OBs

HQAc/NaOAc \ 75° 181-OBs

w n OAc +

H 184-OAc 185-OAc 181-OAc

LiAlHjj, LiAlHjj, 184-OH 185-OH

(CH^)^SiCl (CH-jJ^SiCl \!/ 184-0S1(CH3)3 185rOSi(CH3)3 74 ' other means but were tentatively identified as the 2 -methyl and 4-methyl isomers by careful analysis of the nmr spectrum of the mixture. Reduction of the mixture with lithium aluminum hydride gave the corresponding alcohols which also proved to be inseparable. Treatment of the alcohols with trimethylsilyl chloride gave the corresponding trimethyl silyl ethers which were volatile and could be successfully separated by vpc. The nmr spectra of lSJ-OSMCH^)^ and 185-OSi(CH^)^ were distinctively different with the most outstanding difference appearing in the vinyl region. The nmr of the parent compound (1 5 5 ) shows peaks at 6 . 3 S representing Hj^, and H(j, 5<9$ representing H3 and H^» 4.8S representing Hg.

The spectrum of 184 shows only one proton at 6.38 indic ating a methyl substituent at Hk or H*. In 185* there is no peak at 4 .8 S for Hg and both H9 and have been notice ably deshielded compared with the corresponding value for the parent compound and that observed for ,184. It was on the basis of these nmr data that the positon of the methyl substituents in the solvolysis products was assigned. Once separated, the silyl ethers were hydrolyzed back to their corresponding alcohols by the procedure of Galbraith et. al., 98 and subsequently treated with a trace of p-toluenesulfonic acid in anhydrous acetic acid at 50°. In the case of .184* rearrangement occurred to the correspond ing methyldihydronaphthalene (186) which gave exclusive 75

l84-OSi (H+ )

AcOH, TsOH, 50°

DDQ H3 i§7 186 formation of f?-methylnaphthalene (18£) upon treatment with DDQ, Compound 18^-OH, however, gave a dehydration product upon acid treatment but the subsequent reaction with DDQ gave no volatile components. The failure of 18% to arom atize in a fashion similar to 184, will be explained in the following mechanistic interpretation of these results. The rearrangement of 8^-OBs to l^rOAc is an intriguing one and the data arising from the deuterium and methyl labeling experiments provide valuable Insight into the mechanism. Scheme V outlines a sequence which adequately explains the data and seems to give a reasonable explan ation of the observed facts. Ionization of 8^0Bs is suggested to occur with participation by the double bond of the homoallylic system to give the intermediate spiro cation (188) which contains an element of symmetry in both the ring and side chain and would satisfactorily explain the Scheme V

OBs

89-OBs 188 189

192 190

OAc -H+ OAc* V

156 vC—/ -o 155-QAc cr\ observed deuterium scrambling1 experiments. Interconversion to the bicyclobutyl cation 189 can lead to the strained intermediate v190 w which can readily form the tetrahydroazulene acetate.(15^-OAc) as well as the interconverting cycloheptatrienylcarbinyl-norcaradienylcarbinyl pair, 191 ‘ and 192. The facile transformation of 191 to 192 provides an easy route for the rearrangement of 3^-OAc to 15&* high acetate ion concentration, trapping of 190 (and/or 1^1) occurs faster than the 191 to lg2 transformation and conver sion to 15£-0Ac is the predominant pathway. Low acetate ion concentration allows increased amounts of interconver sion to 192 which can result in exclusive formation of 156/ The effect of a *f—methyl substituent In this sequencer is outlined In Scheme VI, Ionization of 181-OBs gives an Intermediate spiro[7.2 ]cation which now contains symmetry In the side chain but no longer in the ring. Conversion to the cyclobutyl cation then gives two intermediates, lg^f and 1£5, which proceed to give two isomeric tetrahydroazulene acetates (18J and- In the case of conversion to a dihydronaphthalene derivative is facile and proceeds via the norcaradienylcarbinyl intermediate 200 which readily forms 186 with loss of a proton. The position of the methyl group in 1999 however, prohibits such aromatization as one can see by examining the norcaradienylcarbinyl cation 201 . It is not surprising, therefore, that 18§,-0H did not 78

Scheme VI

CH H3G ,OBs "\.t I

181-OBs 193 194 195.

/ i

•Ac 184-OAc 196 197

HoCJ/

198 // h 3 ’OAc CH 199. 185-OAc 200 IT

186 201 79 aromatize upon treatment with acid and DDQ as was the case with 18^-OH. An important question which arises from this mechanism involves the nature of the Intermediate spiro |7.2]cation, 188, formed upon ionization of 181-OBs. 188 has been pic- tured as a classical cyclopropylcarbinyl carbonlum ion in which the COT ring maintains its "tub" conformation thus prohibiting delocalization of the positive charge by the cyclooctatrienyl system. Alternatively, a flattening of the ring would permit charge delocalization to give a homotropylium intermediate (202) which would be homoaromatic in nature. The 52-fold rate deceleration of the p-cycloocta-

202

tetraenylethyl system (,89,) relative to the ^-(/4-cyclooctenyl)ethyl system (132) strongly eliminates the possi bility of a homoaromatic intermediate such as 202 being formed in the rate-determining step. Instead we can assume that double bond participation in the case of 89,-OBs gives an intermediate which exists in the familiar "tub" confor mation as described by 188. Inspection of a molecular model 80 of this intermediate shows that in this conformation, the developing p orbital of the carbonium ion is not parallel to the p orbital of the adjacent double bond. Thus maximum overlap is not feasible and the adjacent double bond instead serves as an electronegative substituent which destabilizes the developing carbonium ion. The effect of cyclopropyl and vinyl groups twisted by 90° from the most favored geometry for conjugation on the stability of an adjacent cationic center has been shown in the solvolysis of spiro [cyclopropane-1,2-adamantyl] (20if) and 2-methyleneadamantyl (205) derivatives.^

203 20£ 205 1 10-2 10 -k

In contrast to the large rate accelerations usually assoc iated with ionizations leading to cyclopropylcarbinyl and allylic cations, compounds 20A and 20fj respectively solvolyze about 10"^ and 10"^ times as fast as the parent compound, 203. A linear correlation of log krei with the inductive substituent constant

2-methyleneadamantyl system (205 ) a rate deceleration of 52 in the case of 89 is not unreasonable. An examination of ' the model also reveals that the cyclopropyl ring and carbon ium Ion cannot attain the most favorable cis-blsected geometry and this too could have a significant but smaller' destabilizing effect, exclusions. The solvolysis of cyclooctatetraenylmethyl (88) and {3-cyclooctatetraenylethyl (89) derivatives have revealed the potential of COT to function as a neigh boring group and have led to intriguing rearrangements. However, from a mechanistic standpoint, the kinetic evidence has failed to indicate the formation of the homotropylium ion in the rate-determining steps of these processes. With both 88 and 89, no significant rate increases were observed (with respect to chosen model systems) to Indicate a decrease in their ionization energy owing to the greatly delocalized charge of the homotropylium intermediate. On the contrary, rate decelerations were observed in both cases. EXPERIMENTAL

General, All melting points were taken In open capillaries and are corrected. Microanalyses were performed by the Scandinavian Microanalytical Laboratory, Herlev, Denmark. Infrared spectra were determined with a Perkin-Elmer Model

237 spectrometer while ultraviolet spectra were recorded on a Cary Model l^J- spectrometer. Nmr spectra were obtained with Varian A-60A, HA-100, or Jeolco MH-100 spectrometers. The mass spectra were measured with an AEI MS-9 mass spectrometer, operated at an ionizing energy of 70 ev.

Carbomethoxycyclooctatetraene (91)* A! solution of 500 ml of benzene and 20 ml of methyl propiolate (£0) was photo- lyzed in a Bayonet reactor with 2537 & light in a quartz tube for a period of ^ days, (The tube needed to be cleaned daily to remove polymetic residue). The benzene was removed

in vacuo and the residue was distilled at 52° and 0.05 mm to give ^.55 6 of yellow liquid; 3.70(s, 3 H, methyl) TMS and 5*85 (s, 7 H, vinyls).

Cyclooctatetraenylcarblnol (88-0Hj. To a suspension of

0.530 g (13.95 mmoles) of lithium aluminum hydride in 50 ml

of ether was slowly added a solution of 2.25 g (13*95 mmoles)

83 84 of carbomethoxyCOT (91) In 100 ml of ether* The reaction mixture was stirred at room temperature for 2 hrs and 0.5 ml of water, 0.5 ml of 10# sodium hydroxide solution, and 1.4 ml of water were added slowly in that order. This slurry was filtered through magnesium sulfate and the sol vent was removed in vacuo to give 1.85 S (99#) of ,88-OH; 2,33(broad s, 1 H, hydroxy), 4.02(s, 2 H, methylene), and 5«83(s, 7 H, vinyls).

an loe-cooled solution

of 200 mg (1.50 mmoles) of WN88-OH in 10 ml of pyridine was added 462 mg (2.00 mmoles) of recrystallized 3*5-8initro- benzoyl chloride. The mixture was stirred for 10 min and left in the refrigerator overnight. The reaction mixture was poured over ice, extracted with ether, and the combined ether extracts washed with water, 10# hydrochloric acid, saturated sodium bicarbonate solution, and dried over mag nesium sulfate. Removal of the ether gave a yellow oil

which was crystallized from absolute ethanol to give 250 mg (51#) of yellow needles, mp. 68-9°; StmI1? 2 H * methylene), 5*89(s, 7 H, vinyls), and 9.13(s, 3 H, aromatics). Anal. Calcd for C16H12N206i C, 58.5^i H, 3.68, Pound: C, 58.46; H, 3.75.

Cyclooctatetraenylcarblnyl chloride (88-Cl). To a solution of 1.00 g (7.^+6 mmoles) of 88-OH and 1.6 ml (20.0 mmoles) of 35 pyridine in $0 ml of chloroform was slowly added a solution of 1.79 g <15*0 mmoles) of thionyl chloride in 5 ml of chloroform. The mixture was stirred at room temperature for 12 hrs, quenched with 25 ml of water, and the chloroform layer was separated. The chloroform solution was washed with 50 ml portions of water, 3 N hydrochloric acid, satur ated sodium bicarbonate solution, and dried ovsr magnesium sulfate. Removal of solvent in vacuo gave a colored oil which was molecularly distilled to give 0.42 g (33^) of yellow liquid; S ^ g 13 3.92

Reduction of 2-Carboetho_xycyclpoctanone (94). To a suspen sion of 2.28 g (60,0 mmoles) of‘lithium aluminum hydride in 50 ml of ether was slowly added a solution of 10.0 g (50.4 mmoles) of 2-carboethoxycyclooctanone.^® The mixture was refluxed for 3 hrs and a basic worlcup was used (2,3 ml of water, 2.3 ml of 10^> sodium hydroxide solution and 7*0 ml of water). Filtration through magnesium sulfate and evaporation of the ether gave 7.17 g of residue which was 1 distilled at 75-6° and 0,2 mm to give 4.55 S of a 1:4 mix ture of 92-OH and 95-OH. Phenyl urethane derivatives were prepared from both isomers (separated on a 6* column of Carbowax-lj£ potassium hydroxide on Chromosorb G at 140°) and were found to match those reported in the literature (mp 60-1°) and 96-8°, respectively).98 Nmr spectral data for g2-0H; 1.52(s, 8 H, aliphatics), 2.20(broad s, 4 H, allylics), 4.06(s, 2 H, methylene), and 5«67(t, J *= 7 Hz, 1 H, vinyl). For 95-OH; 1.33-2.0?(m, 10 H, aliphatics), 2.10-2.33(m, 2 H, allylics), 2.45(s, 1 H, hydroxyl), 4.17(t, J = 6 Hz, 1 H, H alpha to hydroxyl), 4.93(m, 1 H, vinyl), and 5.01(w* 1 H, vinyl).

3^5-Dej^oate^^9 ^_92 -0H . To an ice-cooled solution of 160 mg (1.14 mmoles) of £2,-0H in 10 ml of pyridine was added 346 mg (1.50 mmoles) of recrystallized 3»5-dinitrobenzoyl chloride. The solution was stirred for 10 min and left in the refrigerator overnight. The reaction mixture was poured over ice, extracted with ether, and the combined ether extracts were washed with water, 10^J hydrochloric acid, saturated sodium bicarbonate solution, and dried over magnesium sulfate. Removal of the ether gave a nearly colorless oil which was crystallized from absolute ethanol

to give 240 mg of pale yellow needles, mp 96-7°; S ^ g 13 1.53 (s, 8 H, aliphatics), 2.30(broad s, 4 H, allylics), 3.80(s, 2 H, methylene), and 9.13(s, 3 H, aromatics). Anal. Calcd for Cl6Hl8N206: C, 57.48; H, 5.43; N, 8.35.

Found: C, 57.51; H, 5.55; N, 8.35. 8? Chlorinat1on of J22-OH. To a solution of 340 mg (2.43 mmoles) of 92-OH and 400 mg (5.00 mmoles) of pyridine in 50 ml of chloroform was slowly added a solution of 595 nig (5.00 mmoles) of thionyl chloride in 10 ml of chloroform. The mixture was stirred for 12 h'rs at room temperature, quenched with cold water, and extracted with ether. The combined ether extracts were washed with water, 10# hydro chloric acid, saturated sodium bicarbonate solution, and dried over magnesium sulfate. Removal of the ether gave 375 nig of a dark oil which was chromatographed on silica gel and eluted with pentane. Vpc analysis of the eluent showed a 60:40 mixture of. the isomeric chlorides 92-C1 and 95-C1 which were separated by vpc on a 6» column of 5# V " ' * — i'"* Carbowax-1# potassium hydroxide on Chromosorb G at 95°.

Nmr spectral data for j?2-Cl; 1.50(s , 8 H, aliph atics), 2.20(m. 4 H, allylics), 4.00(s, 2 H, methylene), and 5.72(t, J = 7 Hz, 1 H, vinyl). Anal. Calcd for C^H^Cl: C, 68.13; H, 9.53. Pound: C, 67.86; H, 9.45. Nmr spectral data for (95,-Cl; 1.52(s, 8 H, ring H), 2.20(m, 4 H, ring H), 4.48(m, 1 H,methine), 4.92(s, 1 H, vinyl), and 5*l4(s, 1 H, vinyl). Anal. Calcd for C^H-^Cl: C, 68.13; H, 9.53. Pound: C, 68.10; H, 9.53. 88 ' Kinetics Procedure; A 50# aqueous ethanol solution was pre pared by combining an equal volumetric mixture of absolute ethanol and distilled water both of which had been degassed prior to use, A standard solution of aqueous sodium hydrox ide was prepared and standardized against potassium hydrogen sulfate, A 0,03 M solution of chloride and two equivalents of 1 redistilled 2,6-lutidine in 5°# aqueous ethanol was pre pared, Aliquots of this solution (ca, 1,1 ml) were removed, sealed in glass ampoules and immersed in a constant temper ature bath. After 5 rain the first ampoule was removed, an accurate timer started, and the ampoule quickly cooled in an ice-water bath. The ampoule was then placed in a vessel of water at room temperature. After 5 min exactly 0.923 ml of solution was removed from the ampoule with an automatic pipet, treated with 1 drop of a saturated solution of phenol- phthalein indicator in ethanol, and titrated with standard sodium hydroxide solution, A Fisher Accumet pH meter with microprobe combination electrode was used to determine the end point potentiometrically. The remaining ampoules were removed at appropriately timed intervals, Immediately cooled in ice-water, and titrated as previously described. In each case one ampoule was allowed to remain in the heated bath 1 for a period of at least 10 half-lives. The sample was then titrated as before to give the infinity titer. The rate constants were calculated using a STAT-6 program for the least squares treatment of the data applied to the following equations

[hciL - t>01l Activation parameters and extrapolated rate constants were also calculated* The Wang Electronic Calculator was used for all programming*

Product^Stiidi^s^^^lvol^^ A solution of 325 mg (2.31 mmoles) of g2 -Cl and 2^7 mg (2.31 mmoles) of 2 ,6- lutidine in 25 ml of 50% aqueous ethanol was stirred at room temperature for 5 days. The. reaction mixture was added to 100 ml of water and extracted viith ether. The combined ether portions were washed with water, 3 N hydrochloric acid, saturated sodium bicarbonate solution, and dried over magnesium sulfate. Removal of the ether in vacuo gave 275' mg of colorless liquid which was analyzed by vpc using a 6* column of 5# Carbowax-1# potassium hydroxide on Chromosorb G and shown to contain 92-0H (27/0, 92-0CH~CHo (33/0, 95-OH (22#), and jJ^-OCHgCH^ (19#). Nmr spectral data for g2-0CH2 CH3; 6j“gl3 l.lMt, J = 7 Hz, 3 H, methyl), 1.48(s,

8 H, aliphatics), 2.l6(br s, H, allylics), 3*38(<1, J - 7

Hz, 2 H, -OCH2 CH3 ), 3.80(s , 2 H, -CH2 0-), and 5.56(t, J = 7 Hz, 1 H, vinyl). 90 Anal. Calcd for C ^ H ^ O : C, 78.51; H, 11.98. Pound: C, 78.15; H, 11.88. Nmr spectral data for

Solvol^^s^of^88-^^ A solution of 330 mg (1.95 mmoles) of 88-C1 and 2.4 mg (2.00 mmoles) of 2,6-lutidine in 25 nil of 50# aqueous ethanol was stirred at room temperature for 116.5 hrs. The reaction mixture was added to 100 ml of water and extracted with ether. The combined ether portions were washed with water, 3 N hydrochloric acid, saturated sodium bicarbonate solution, and dried over magnesium sul fate. Removal of the ether in vacuo gave 250 mg of yellow oil which was analyzed by vpc on a 6* column of 10# QF-T on Chromosorb G and shown to contain 88-OH (37#), 88-

OCH2CH3 (l?*), and gg (46*).

Cyclooctatetraenylmethyl ethyl ether (88-OCHpCHo). To a solution of 35° mg (2.61 mmoles) of v"x/v88-OH and 285 mg (2.6l mmoles) of ethyl bromide in 20 ml of tetrahydrofuran was added 292 mg (2.61 mmoles) of potassium t-butoxide. The 91 reaction mixture was stirred at room temperature for 6 hrsf quenched with water, and extracted with ether. The com bined ether portions were washed with 25 ml portions of water, 3 N hydrochloric acid, saturated sodium bicarbonate solution, and dried over magnesium sulfate. Removal of the ether in vacuo gave 300 mg {75%) of j^-OG^CH^ which was purified by vpc on a 6' column of 5% Carbowax-l# potassium hydroxide on Chromosorb G at 135°. The nmr spectral data for 88-OCH2 CH3; S^MS13 1.18(t, J = 7 Hz, 3 H, methyl), 3.48(q, J = 7 Hz, 2 H, ether methylene), 3«86(s, 2 H, allylic methylene), and 5*80{s, 7 H, vinyls). Anal. Calcd for 0, 81.44; H, 8,70.

Found: C, 81.30; H, 8.67. fl-o-Tolyl ethanol (101). To a suspension of 0.635 S (0.017 mmole) of lithium aluminum hydride in 50 ml of ether was slowly added a solution of 250 mg (0.017 mmole) of o- tolyl acetic acid (ljDiD). The reaction was stirred at room temperature for 2 hrs and a basic workup was used (O.65 ml of water, 0.65 ml of 10# sodium hydroxide solution, and 2.0 ml of water). The mixture was filtered through magnesium sulfate and the solvent was evaporated in vacuo to give

2.20 g of o-tolyl ethanol; 1,96(s , 3 H, methyl), 2.52(t, J » 7 Hz, 2 H, benzylic methylene), 3*40(t, J = 7 Hz, 2 H, methylene), 4,05(br s, 1 H, hydroxyl), and 6.78 (s, 5 H, aromatics). 92

• To a solution of 100 mS (0.736 mmoles) of 101 in 50 ml of methylene chloride was added v-v-v 1,16 s (7.36 mmoles) of Collins reasent. The solution was stirred at room temperature for 24 hrs and the solvent was removed in vacuo. The residue was taken up in 100 ml of ether and washed with 50 ml portions of water, 10# hydro chloric acid, saturated sodium bicarbonate solution, and dried over magnesium sulfate. Removal of the solvent in vacuo gave 100 mg of colorless liquid. Vpc analysis (6* column of 5# Carbowax-1# potassium hydroxide on Chromosorb

G at I3O0) showed 58# o-tolylacetaldehyde and 42# o-meth- ylbenzaldehyde. Nmr spectral data for 99; 2.24(s, 1“'V* IMq 3 H, methyl), 2.62(d, J = 2 Hz, 2 H, methylene), 7.12(s, 4 H, aromatics), and 9.60(t, J ^ 2 Hz, 1 H, aldehyde); A semicarbazone of was prepared in the usual manner and recrystallized twice from 95# ethanol to give a crystal line product, mp 176-7°. A mixed melting point proved it to be identical to the derivative prepared from the solvolysis product. Anal. Calcd for C^H-^NO: C, 62.80; H, 6.85; N, 21.98. Found: C, 62.41; H, 6.85; N, 22.04. ac;c(r*Dideuteriocyclooctatetraenylcarbinol (102-OH). To a suspension of 675 mg (16.0 mmoles) of lithium aluminum deuteride in 50 ml of ether was slowly added a solution of 2.60 g (16.0 mmoles) of carbomethoxyCOT (91). The mixture 93 was stirred at room temperature for 2 hrs and processed as with 88-OH to give 1,66 g (76#) of alcohol; 2.52 «-v' IMS (br s, 1 H, hydroxyl), and 5.82(s, 7 H, vinyls), with no proton absorption at 4.026.

for 12 hrs and processed as with 88,-Cl to give 0.660 g (35$) of 102,-Cl, Mass spectral data; calcd m/e 154.0518, found, ' 154.0519. Nmr spectral data; S^g^*3 5.82(s, 7 H, vinyls), with no proton absorption at 3.928 , t

A solution of 1.16 g (6.79 mmoles) of 102-C1 and 0.75 S (7.00 mmoles) of 2,6-lutidine in 75 ml of 50# aqueous ethanol was stirred at room temperature for 7 days. The reaction mixture was processed as with 8£3-Cl and vpc analysis on a 6' column of 5# Carbowax-1# potassium hydroxide on Chromosorb G at 148° showed: 102-OH (75#)* IO2 -OCH2 CH3 (10#), and 103 (15#). Spectral data for 102- OH; calcd m/e 136.0857* found, 136,0859. Spectral data for 1 3 -OCH2 CH3 ; calcd m/e 164.1170, found, 164.1173. £>TMSl3 l.l8(t, J = 7 Hz, 3 H, methyl), 3.46(q, J = 7 Hz, 2 H, ether methylene), and 5.76(s, 7 H, vinyls). Spectral 94 data for .103* calcd m/e 136,0857* found, 136.0859; 2.20(br s, 1 H, -CD2H), 3-64(d, J * 2 Hz, 2 H, methylene), 7.12(m, 4 H, aromatics), and 9.62(t, J « 2 Hz, 1 H, aldehyde).

a s^irred solution of 20.0 g (0,192 mole) of cyclooctatetraene in 80 ml of dry methylene chloride under a nitrogen atmosphere and at a temperature of -65°C was added a solution of 31*36 g (O.196 mole) of bromine in 80 ml of dry methylene chloride over a period of 40 min. The mixture was stirred for an additional 30 min at -65° and 32.70 g (0.292 mole) of powdered potassium t-butoxide was added in 1-2 g portions during 1 hr while stirring vigorously. The bath temper ature was allowed to rise to -10° over a period of 90 min. The mixture was poured into 400 ml of water containing 10 ml of glacial acetic acid. The aqueous layer was satur ated with sodium chloride and left overnight. The emulsion was filtered and the aqueous layer shaken with 200 ml of methylene chloride. The combined organic extracts were washed with 140 ml of sodium bicarbonate solution and 200 ml of cold, saturated sodium chloride solution and dried over magnesium sulfate. The methylene chloride was removed in vacuo and the residue (23.5 s) was distilled at 40° and 0.01 mm to give 21.1 g (60#), 95 Cycj^^tatetraenylli^^ To a solution of 3.85 g (32.0 mmoles) of 1££ in 100 ml of dry ether under a nitro gen atmosphere and at a temperature of -60° was added 26 ml of 1.5 M n-butyllithium (38*0 mmoles) by means of a syringe and rubber septum. The resulting orange solution was stirred at this temperature for 2 hrs. The compound proved unstable and was used immediately.

[9-Cyclooctatetraenylethanol (89-0H). To a fresh solution of 3*52 g (32.0 mmoles) of 1^8 under a nitrogen atmosphere at -60° was added 3*13 6 (71.0 mmoles) of ethylene oxide. The mixture was stirred at this temperature for 30 min and allowed to warm to room temperature. The mixture was washed with two 100-ml portions of 10# hydrochloric acid followed by two 50-ml portions of saturated sodium chloride solution. The ether layer was removed, dried over magnesium sulfate, and concentrated in vacuo. The residue (3.1 g) was distilled at 0.5 mm to yield 2.8 g (60#) of 8^-OH, bp 80-85° and 0.5 mm; ^ £ eat 3310 and 1020-1050 cm"1 (OH);

^TMS13 2 *°9(s, 1 H, hydroxyl), 2.2^(t, J = 6.5 Hz, 2 H, alpha methylene), and 5.70(s, 7 H, vinyls). Anal. Calcd for C10H12O: C, 81.0^; H, 8.16. Found: C, 8 0.65; H, 8.12. 96 ^ a coo^e<^ solution of 3*78 g (25.4 nunoles) of J39-OH in 25 ml of pyridine was added 6.01 g (26.0 mmoles) of recrystallized 3*5-8initrobenzoyl chloride. The solution was stirred for 15 min and left in the refrigerator overnight. The reaction mixture was poured into 50 ml of water and extracted with ether. The combined ether extracts were washed with water, 3 N hydrochloric acid, saturated sodium bicarbonate solution, and dried over magnesium sulfate. Removal of the ether in vacuo gave a yellow oil which was crystallized from absolute ethanol to ‘ give 6.4l g (74#) of yellow needles, rap 69-70°. Nmr spec tral data for 89-ODNB; 2.60(t, J = 6.5 Hz, 2 H, allylic methylene), 4.55(t, J - 6.5 Hz, 2 H, alpha methylene), 5.80(s, ? H, vinyls), and 9*22(s, 3 H, aromatics). Anal. Calcd for C1^H14,N20gs C, 59.65; H, 4.12; N, 8.18. Pound: C, 59.59; H, 4.28; N, 8.15.

Brosylate of 89-OH, To a cooled solution of 1.00 g (6.76 mmoles) of JJ^-OH in 50 ml of pyridine was added 3.60 g (13.5 mmoles) of £-bromobenzenesulfonyl chloride. The solution was stirred under nitrogen for 30 min and placed in the refrigerator overnight. To this solution was added ca 50 g of ice and the resulting aqueous solution was extracted with ether. The combined ether portions were washed with 10# hydrochloric acid, water, saturated sodium bicarbonate 97 solution and dried over magnesium sulfate. Removal of the ether in vacuo gave 2.26 g (91^) of yellow oil which was crystallized from absolute ethanol to give pale, yellow needles, mp *fl-2°. Nmr spectral data for 89-OBs: 2.*l-2(t, J - 7 Hz, 2 H, allylic methylene), *J-.17(t, J = 7 Hz, 2 H, alpha methylene), 5.75(s, 7 H, vinyls), and 7.75(s, *f- H, aromatics).. Anal. Calcd for C-j^H-j^BrO^S: C, 52.31; H, *t-.15» S, 8.82. Pound: C, 52.26; H, if.12; S, 8.73.

Methyl-1-Hydroxycyclooctylacetate (1*1-2). Approximately 5 S (79 mmoles) of mossy zinc was first activated by adding it to a warm solution of 5% sulfuric acid and washing with water, methanol, acetone, and ether. The zinc was heated in a vacuum oven at 110°C for 30 min and placed in a 3- necked flask fitted with overhead stirrer, condenser to which was attached a nitrogen inlet, and dropping funnel. A solution of 10.0 g (79.** mmoles) of cyclooctanone (l^fl) and 13.3 g (79.** mmoles) of ethyl broraoacetate in 100 ml of dry ether was placed in the dropping funnel and 3 ml were added slowly to the zinc. The mixture was heated until the reaction began and the remainder of the solution was added at a rate necessary to maintain reflux. After the addition was complete, the solution was refluxed for *f hrs. After cooling to room temperature, 50 ml of 10^ sulfuric acid was added slowly with vigorous stirring. The layers

1 were separated and the ether layer was washed with 5# sulfuric acid, 10# sodium carbonate solution, and water. The combined aqueous layers were extracted with ether and the combined ether layers washed with saturated sodium chloride solution and dried over magnesium sulfate. The ether was removed in vacuo and the residue distilled at a pressure of 0.02 mm to give 5*32 g (31#) of 95# pure 3^2, bp 93-4°• as well as an additional 3.23 & (19$) of 75# pure 142, bp 81-2° to give a total yield of 45#; ^max^ 3550 cm**' (OH) and 1710 cm”1 (carbonyl); 1.1-2 ..2 (complex m, 17 H, CgHi^OH-CHaCOaCHaCH^), 2.49(s, 2 H, -ch2co2-), 3 .^6 (s, 1 H, hydroxyl), and 4.23(q* J = 10.5 Hz, 2 H, -C02CH2CH3). Q1 Lithium Hexamethyldlsillzane. To a 100 ml' 3-necked flask fitted with condenser, nitrogen inlet tube, and serum stopper, and containing 25 ml of dry ether was added 9.5 ml (45.1 mmoles) of hexamethyldisilizane. To this solution was slowly added 30 ml of n-butyllithium in pentane (1.5 M* 45.1 mmoles). Gas evolution was immediately apparent upon- addition of the n-butyllithium and the gas was assumed to be n-butane. The mixture was refluxed for 30 min, cooled to room temperature, and solvent was removed in vacuo to give a white solid residue which was used quickly and directly in the next reaction. 99 A Iter n a. teSynt hes Is of 142. To the white solid residue of lithium hexamethyldisilizane described above was added 20 ml of tetrahydrofuran and the resulting solution was cooled to -70°. A syringe was used to add 3.3 ml (39.7 mmoles) of ethyl acetate and the reaction mixture was

stirred for ,15 min at -70°, at which time a solution of 5.00 g (39.7 mmoles) of cyclooctanone (1^+DL) in 20 ml of tetrahydrofuran was added. The reaction mixture was stir red for 10 min and acidified with 20 ml of 10# hydrochloric acid. The solution was warmed to room temperature and the organic layer was separated and washed with water, satur ated sodium bicarbonate solution, and dried over magnesium sulfate. Removal of the solvent in vacuo gave 8,10 g (95#) of a colorless liquid,

5*32 g (24.8 mmoles) of 142 at 0° was slowly added 6.9 ml (19.5 S» 72,0 mmoles) of phosphorus tribromide. The solution was stirred at room temperature for 24 hrs, at which time the mixture was added very slowly, with stirring, to 25 ml of ice- water. (Rapid addition resulted in flames). The aqueous solution was extracted with ether which was subsequently washed with a saturated sodium chloride solution and dried over magnesium sulfate. The ether was removed in vacuo to give 6.70 (99#) of crude 143 which was not distilled due to 100 its very high boiling point. Tic analysis showed only one component; ^720 (carbonyl), with absence of -OH absorption,

Ai-Cyclooctenylacetic Acid (144). The crude bromide 143

(6.70 g, 24.2 mmoles) was added to a solution of 8.0 g of potassium hydroxide in 30 ml of ethanol and stirred over night at room temperature. The ethanol was removed in vacuo and 25 ml of ether was added to the residue. The ethereal solution was subsequently washed with water, satur ated sodium carbonate solution and dried over magnesium sulfate. Upon removal of solvent, the residue was dis tilled to give 4.00 g {96%) of cyclooctenylacetic acid, bp 108-110° at 0.1 mm; ^JJax** 1720 cm”1 (carbonyl); 1.50(s, 8 H, aliphatics), 1.9-2.3r s, 4 H, allylics), 3.03

^ slurry °** 0.313 £ (2.20 mmoles) of lithium aluminum hydride and 10 ml of ether was placed in a 100 ml 3-necked flask fitted with reflux condenser, magnetic stirrer, and addition funnel. To this mixture was added a solution of 0.5 £ (2.98 mmoles) of cyclooctenylacetic acid (lJi) 1° ml of> ether. The mixture was refluxed for 24 hrs and the usual workup was used (p.4 ml of water, 0.4 ml of sodium hydroxide sol ution, and 1.2 ml of water). After filtration the ether 101 solution was dried over magnesium sulfate. Upon removal of the solvent, the residue was distilled to give 0.**3 & (98^) of 137-OH, bp 121° at 16 mm; &CDCI3 1 .51(5 8 h, aliphat- TMS ics), 1,9-2.5(complex m, 6 H, allylic ring H and allylic methylene), 2«62(s, 1 H, hydroxyl), 3.72(t, J = 6.5 Hz, 2 H, alpha methylene), and 5.33(t, J = 8*0 Hz, 1 H, vinyl).

3 ,_5-pinltrobenzoate of 139-OH_. To a solution of 3*00 g. (19*3 mmoles) of 1J^-0H and 1.5 ml (2.02 g, 20.0 mmoles) of triethylamlne in 15 ml of hexane was added 4-.62 g (20.0 mmoles) of recrystallized 3,5-dinitrobenzoyl chloride. The solution was warmed slightly and stirred for *1- hrs. A ■ precipitate of triethylamlne hydrochloride appeared almost immediately. The solid was removed by extraction with $0 ml of water and the solution was evaporated to yield 0,66 g (98/6) of the white solid. Upon recrystallization from ethanol, 6.32 g (93^) of 132-ODNB was obtained, mp 55-6°; Srjijjg13 1.51(s, 8 H, aliphatics), 1.9-2,8(complex m, 6 H, allylics), 4.6o(t, J =* 6.5 Hz, 2 H, alpha methylene), 5*58 (t, J = 8.0 Hz, 1 H, vinyl), and 9*21(s, 3 H, aromatics). Anal. Calcd for C17K20N2°6: c » 58.61; H, 5*79; N, 8.0*1-. Pound; C, 58.66; H, 5.92; N, 8.10.

To a solution of 2.00 g (13.0 mmoles) of 135-OH in 100 ml of pyridine was added 5.12 g (20.0 mmoles) of jD-bromobenzenesulfonyl chloride. The solution 102 was stirred at ice bath temperature for 30 min and placed In the refrigerator overnight. Approximately 75 S of ioe was added to the mixture which was subsequently extracted with ether. The combined ether portions were washed with water, 10# hydrochloric acid, saturated sodium bicarbonate solution and dried over magnesium sulfate. Removal of the ether gave 3.23 8 (67#) of colorless oil which was purified by low-temperature recrystallization. Nmr spectral data for 139-OBs; S§§|13 1.42(s, 8 H, aliphatics), 2.05(br s, 4 H, allylic ring H), 2.33(t, J « 7 Hz, 2 H, allylic meth ylene), 4.17(t, J a 7 Hz, 2 H, alpha methylene), 5.3?(t,

J. = 8 Hz, 1 R, vinyl), and ?.73(s, 4 H, aromatics).

$^yclooctyvl^^ To a solution of 950 raS (6.17 mmoles) of 139-OH in 30 ml of methanol was added 3 ml of acetic acid and 100 mg of platinum oxide. Hydrogenation was carried out at room temperature in a Parr hydrogenator (50 psi) for a period of 24 hrs. The reaction mixture was filtered through Celite and 25 ml of water was added to the filtrate. The filtrate was. extracted with ether and the combined ether extracts were washed with water, saturated sodium bicarbonate solution, saturated sodium chloride solution, and dried over magnesium sulfate. Removal of the solvent in vacuo gave 680 mg (71#) of a colorless liquid. Distillation at 0.05 nim gave a forerun (bp 40-1°) which was Identified as ethylcyclooctane^ as well as 577 &8 (60#) of 103 the desired alcohol, bp 101-2°; Srjjjjfl£p3 1.53(b:r s, 17 H, C8H15CH2CH2OH), and 3.63(t, J - 7 Hz, 2 H, -CHgOH).

Brosylate of 1^0-0H. To a solution of 2.20 g (1^.1 mmoles) of lJfO-OH in 75 nil of pyridine at 0° was added 7.67 g (30.0 mmoles) of p-bromobenzenesulfonyl chloride. The solution was stirred at 0° for 15 min and placed in the re frigerator for 36 hrs. To the reaction mixture was added approximately 50 g of ice with stirring. The aqueous sol ution was extracted with ether and the combined ether ext racts washed with water, 3 N hydrochloric acid, saturated sodium bicarbonate solution, and dried over magnesium sul fate. Removal of the ether in vacuo gave 3*72 g (70#) of a colorless oil. Low-temperature crystallization of this oil from absolute ethanol at Dry Ice-isopropyl alcohol' temperatures gave a white solid which, however, melted below room temperature. Chromatography on silica gel re sulted only in decomposition of the brosylate and so the low-temperature recrystallization was used as the method of purification of the ester. Nmr spectral data for l4£-OBs;

STMSl3 1*50(br s, 17 H, C8H15CH2 CH2OBs), 4.12(t, J = ? Hz, 2 H, -CHgOBs), and 7.75(s, 4 H, aromatics),

Kinetic^JProcedui^. Anhydrous acetic acid was prepared by refluxing a solution of acetic anhydride in glacial acetic acid overnight and subsequent fractional distillation in a 104 dry atmosphere. Standard perchloric acid solutions In acetic acid were prepared by dilution of an accurately weighed quantity of standard 70$ perchloric acid with an hydrous acetic acid to a known volume. Sodium carbonate which had been heated over an open flame and cooled in a desiccator was accurately weighed and diluted to a known volume with anhydrous acetic acid to prepare the standard sodium acetate solutions; the water of neutralization was not removed. Standard solutions of brosylate in the sodium acetate solution were prepared. Aliquots of this solution (ca 1.1 ml) were removed, sealed in glass ampoules, and immersed ■ in a constant-temperature bath. After 10 min, the first ampoule was removed, an accurate timer started, and the ampoule quickly cooled in an ice-water bath. The ampoule was then placed in a vessel of water at room temperature. After 5 min, exactly 0.923 ml of solution was removed with an automatic plpet, treated with 1 drop of a saturated * solution of bromophenol blue indicator in acetic acid, and titrated with standard perchloric acid using a Fisher '* Accumet pH meter with microprobe combination electrode to determine the endpoint potentiometrically. The remaining ampoules were removed at appropriately timed intervals, immediately cooled in ice-water, and titrated as previously described. In each case one ampoule was allowed to remain 105 In the heated bath for a period of at least 10 half-lives. The sample was then titrated as above to give the infinity point. The rate constants were calculated using a STAT-6 pro gram for the least squares treatment of the data and activ ation parameters and extrapolated rate constants were also calculated. The Wang Electronic Calculator was used for all programming,

SolTnalysi^ A solution of 0.725 S (1*93 nmoles) of l^O-ODsW v was placed in 8 ml of 0.10065 N sodium acetate/ acetic acid and heated at 85° for a period of 10 half-lives (seven days). The solution was cooled and 10 ml of water added. The resulting solution was extracted with ether and the combined ether portions were washed with water, satur ated sodium bicarbonate solution, saturated sodium chloride solution, and dried over magnesium sulfate. Removal of the ether in vacuo gave 350 mg of yellow residue. Vpc analysis on a 6* column of 5^ Carbowax on Chromosorb G showed cyclo- octylethyl acetate (l40~OAc) as the only product. A sample of this acetate was isolated from this column at 1^0° for spectral comparison with an authentic sample prepared as described below. 106 /^CyclooctyOethyl^J^ A solution of 180 mg (1.15 mmoles) of l^O-OH and 550 mg (5*39 mmoles) of acetic anhydride in 0 .5 ml of pyridine was stirred at room temp erature for 2k hrs. Ice was added to hydrolyze the anhyd ride and the mixture was extracted with ether and the com bined ether portions washed with water, 10# hydrochloric acid, saturated sodium bicarbonate solution, and dried over magnesium sulfate. Removal of the solvent in vacuo gave 205 mg (91#) of colorless liquid. The acetate was purified by vpc using a 5' column of 5% SF-9& on Chromosorb G at 120°. Spectral data for 1^0-OAc; l«53(s, 17 H, c8Sl5cS2CH20Ac)* 2.00(s, 3 H, methyl), and *f.08(t, J = 7 Hz, 2 H, -CHgOAc). Anal. Calcd for 0\2^22p2l G» 72.68; H, 11.18. Found: C, 72.53; H, 11.09.

Solvolysis of 139-OBs. A solution of 1.00 g (2.68 mmoles) of 13^-OBs in 150 ml of 0 .195 ^ N sodium acetate/ acetic acid was heated at 50° for a period of 10 half-lives (30 hrs). The reaction mixture was added to ice and the resulting solution was extracted with ether. The combined ether por tions were washed with water, saturated sodium bicarbonate solution, saturated sodium chloride solution, and dried over magnesium sulfate. Removal of the solvent in vacuo gave O.Jj-3 g of yellow residue. A vpc analysis of the mixture was performed on a 10' column of 10# XF-1150 on Chromosorb W 107 and the following product ratio was found: 139-OAc (34£), 1^7-OAc (10S>), and 148-OAc (56#). These products were iso- W V ^ l / V V \ lated from the same column and their nmr spectra were found to be identical to those of authentic samples*

• A solution of 300 mg (1.95 mmoles) of 139-OH and 550 mg (5*39 mmoles) of acetic anhydride in 0 ,5 ®1 of pyridine was stirred at room temperature overnight. Ice was added to hydrolyze the anhydride and the aqueous mixture was extracted with ether. The combined ether portions were washed with water, 10# hydrochloric acid, saturated sodium bicarbonate solution and dried over magnesium sulfate. Removal of solvent in vacuo gave a quantitative yield (382 mg) of colorless liquid. The acetate was purified by vpc on a 5* column of 5# SF-96 on Chromosorb G at a column temperature of 120°. Nmr spectral data for 139-OAc; 1,48(br s, 8 H, aliphatic ring H),

1.83-2.53(m, 6 H, allylics), 2.00(s, 3 H, methyl), 4.13(t, J a* 7 Hz, 2 H, alpha methylene), and 5*^2(t, J « 8 Hz, 1 H, vinyl). Anal. Calcd for C12H20°2S °» 73-^35 H, 10.27. Founds C, 73.71! H, 10.33. 108 Alter^ ^1-Ac^ o^ybi OAcj. A. ^^-Cyoloootenylaoetonltrlle (15Q),« To a solution of 6.50 S (79.4 mmoles) of cyanoacetic acid and 0.223 S (2 .9 0 mmoles) of ammonium acetate in 25 ml of benzene was added 10.0 g (79.^ mmoles) of cyclooctanone (ljfl). This solution was refluxed for 20 hrs using a Dean-Stark trap. After this time, the theoretical amount of water (approx 1.4- ml) had been collected and the benzene was removed by distillation in vacuo. The remaining yellow solid was de- carboxylated by heating at 120° for 2 hrs. This product was distilled at 55-60° and 0.8 ram to give 9.72 g (82#) of 1£0 ;

^TMS3*3 1 -1|,7(sf 8 H, aliphatic ring H), 1.92-2.^2(m, if H, allylic ring H), 3.03(s, 2 H, methylene), and 5.70(t, J = 7 Hz, 1 H, vinyl).

■1 To a suspension of 0.51 g (13.^ mmoles) of lithium aluminum hydride in 20 ml of ether was slowly added a solution of 2 .0 0 g (13.^ mmoles) of 150. The mixture was refluxed under nitrogen for 12 hrs during which time it turned olive green. The reaction mixture was cooled, and a basic workup was used 0 .5 ml of water, 0 .5 ml of 10# sodium hydroxide solution and l.if ml of water. The mixture was filtered through magnesium sulfate and the ether was removed in vacuo to give 2.01 g (98#) of a yellow liquid. The nmr spectral data for 151; 1.^8(br s, 8 H, aliphatic ring H), 1 .90 -2 .4 8(01, 6 H, allylics), 2 .8 2(t, J s 7 Hz, 2 H, methyl ene alpha to nitrogen), and 5.43(t, J = 7 Hz, 1 H, vinyl).

Ten ^-5% perchloric acid solution was adjusted to a pH of 3.5 by adding 2 N sodium hydroxide solution at 5°. To this was added 2 .0 g (1 3 .0 8 mmoles) of 1 5 1* To this solution was added 0 .2 0 g of sodium nitrite in 10 ml of water, while 1 perchloric acid solution was added simultaneously to main tain a pH of 3.5. After the addition was complete, the reaction mixture was heated at 60° for 3 hrs, cooled, and

20 ml of saturated sodium chloride solution was added. The resulting solution was extracted with ether and the combined ether portions were washed with water, saturated sodium chloride solution, and dried over magnesium sulfate. Removal of the ether gave 140 mg of yellow residue which showed two components by vpc. The mixture was separated on a 6* column of 5% Carbowax-1/6 potassium hydroxide on Chromo sorb G at 150°, The two components were shown to be the desired bicyclic alcohol (14£-0H) as well as l^-OH. Spec tral data for 14£-0H; 1.17-2.17(br s with spikes at 1.47, 1.95). A phenyl urethane derivative of 147-0H was prepared and isolated as a white, crystalline solid, mp 148-9° (lit8** 139-42°). of 50.0 mg (0.325 mmole) of 1^7-OH in 0 .1*J* ml (1.65 mmoles) of pyridine was added 0,16 ml (I.65 mmoles) of acetic anhydride. The solution was stirred at room temperature for Zb hrs and quenched with ice. The resulting aqueous solution was extracted with ether and the combined ether portions were washed with water, 10$ hydrochloric acid, saturated sodium bicarbonate solution, and dried over mag nesium sulfate. Removal of the ether gave 4-5 mg of residue which was purified by vpc using a 10* column of 10$ XF-1150 on Chromosorb W. Nmr data for 1^7-OAc; 1.25-1.92 (m, 10 H, large ring H), 2.00(s, 3 H, methyl), and 2.09-2.Jf2(m, 7 H, ’small ring H and large ring H beta to acetoxyl).

Anal. Calcd for c12 h20 °2 : c» 73.^3; H > 10.2 7 . Found: C, 73.19; H, 10.26.

1-Hydroxyspiro [2 .7] decane (1^8-011). A fresh solution of Zn-Cu couple^ was prepared by adding 7 S of zinc dust to a hot, rapidly-stirred solution containing 0.4 g of cupric acetate monohydrate in 10 ml of glacial acetic acid. After approximately 5 min, the copper was deposited on the zinc and the mixture was shaken for 1 min. The acetic acid was decanted and the Zn-Cu couple was washed with two .10 ml portions of acetic acid and filtered. The couple was washed with acetone, ether, and dried in vacuo and the Simmons- Smith reaction was carried out in the following manner: 1X1 A solution of 110 mg (2.00 mmoles) of Zn-Cu, 40 mg

(1 ,5 0 mmoles) of methylene iodide, and 0 .2 ml (2 mmoles)

of glyme (to complex the zinc iodide) in 5 ml of ether was refluxed for 30 min. To this mixture was added a solution of 100 mg (0.71 mmole) of in 5 ml of ether. The

resulting solution was refluxed for 8 hrs, cooled, and

10 ml of saturated ammonium chloride solution was added. The ether was decanted into a separatory funnel and the aqueous layer was washed with ether. The combined ether layers were washed with saturated sodium chloride solution and dried over magnesium sulfate. Removal of the ether

gave a colorless- residue which was isolated pure from a 6* column of 5% Carbowax-l$ potassium hydroxide on Chromosorb G to give the desired alcohol, 148-0H; o ^ | 13 0.25-0.75(m, 4 H, cyclopropyls), 1.25-1*95(m, 11 H, ring H and hydroxyl), 2.07(br, 2 H, ring H beta to hydroxyl), and 2.95(®» 1 H, ring H alpha to hydroxyl), A phenyl urethane derivative of 148-OH was prepared in the usual manner to give a white, crystalline derivative, mp 67-8°; 0 .17-0 .7 7(m, 4 H, cyclopropyls), 1.22-2.08(m, 12 H, ring H), 4.28(t,

J — 6 Hz, 1 H, ring H alpha to hydroxy]}, 6.67(br s, 1 H, -NH), and 6.92-7*58(m, 4 H, aromatics). Anal. Calcd for C, ?4.69; H, 8.49. Pound: C, 74.56; H, 8.40. 1-Acetoxyspiro [2.71 decane (148-OAc). A solution of 148-OH

(1? mg, 0,11 mmole) in 0,05 ml (0.60 mmole) of acetic anhydride and 0 ,0 6 ml (0,80 mmole) of pyridine was stirred for 24 hrs at room temperature. Ice was added and the mix ture was stirred for 15 min followed by extraction with ether. The combined ether portions were washed with water, saturated sodium chloride solution and dried over magnesium sulfate, Removal of the ether gave 22,4 mg of the desired acetate with the following nmr spectrum; 0 ,25 -0,67 (m, 4 H, cyclopropyls), 1.42-1.83(m, 12 H, ring H), 2,02(s, 3 H methyl), and 4,33(t, J - 6 Hz, 1 H, ring H alpha to acetoxyl),

Solj^jy^nis^ A solution of 53*5 mg (0.27 mmole) of 122-OAc in 1*5 ml of 0.1945 N sodium acetate/acetic acid was heated at 50° for 30 hrs. The reaction mixture was added to 5 ml of water and extracted with ether. The result ing ether solution was washed with water, saturated sodium bicarbonate solution, and dried over magnesium sulfate. Removal of the ether in vacuo gave 50*0 mg of a colorless residue, Vpc analysis (10* column of 10# XF-1150 on Chromo sorb W) showed one peak, subsequently identified as starting material. No other isomers were visible by vpc or nmr. 113 A s°lu-tl°n °f 75 nig (O.38 mmole) of acetate 14^ In *f0 ml of 0.10065 N sodium acetate/acetic acid was heated at 50° for 30 hrs. The solution was poured over ice and subsequently extracted with ether. The com bined ether portions were washed with water, saturated sodium bicarbonate solution, and dried over magnesium sul fate. Removal of the ether in vacuo gave 26.2 mg of residue. Vpc analysis of the mixture (10* column of 10# XF-1150 on Chromosorb W) showed 52# 3A£-0Ac and Jf8# 1^-OAo.

Solv^lyjj^^ A solution of 22 mg (0.11 mmole) of acetate 1^8 in 30 mlof O.IOO65 Nsodium acetate/acetic acid was heated at 50° for 30 hrs. The solution was poured over ice and subsequently extracted with ether. The com bined ether portions were washed with water, saturated sodium bicarbonate solution and dried over magnesium sulfate. Removal of the ether in vacuo gave 30.6 mg of a mixture which was identified by vpc (10* column of 10# XF-1150 on

Chromosorb W) as 6l# l48-0Ac, 31# 139-OAc, and 8# 1^7-OAc,

Solvoly^is^of^8^0^ A solution of 1.00 g (2,73 mmoles) of j^-OBs was placed in 135 ml of 0 .19 *1-5 N sodium acetate/ acetic acid and heated at 75° for a period of 10 half-lives (4 days, 7.5 hrs). The reaction mixture was cooled and added to 100 ml of ice water. This aqueous solution was ex tracted with ether and the combined ether portions were washed with water, saturated sodium bicarbonate solution, 114 and dried over magnesium sulfate. Removal of the ether gave 5°0 ms of dark residue which was first chromatographed on silica gel (pentane elution) to remove gross impurities. This mixture was purified on a 6? column of 5% XF-1150 on Chromosorb G at 140°. The most volatile component (tr = 3

min) was identified as 1,2-dihydronaphthalene (137» 12/S) by ir and nmr by comparison with those of an authentic sample. 81 The second component (tr = 14 min) was similarly identified as 8j?-0Ac (19^) and was found to be spectro scopically identical to an authentic sample. The final

component (t*.* = 20 min) was assigned structure 155-OAc '✓VS/ {69%) on the basis of ir and nmr data. Nmr spectral data for 157; 2.37(br m, 2 H f benzylics), 2.87(m, 2 H, allylics), 6.03(m, 1 H, vinyl), 6.40, 6.58(d of t, J = 1.5, 9.5 Hz, 1 H, vinyl), and 7.08{s, 4 H, aromatics), Nmr spectral data for 155,-OAc; S§§s13 1.90(s, 3 H, methyl), 2.03-2.28(m, 3 H, Hx and H9 ), 2.37-2.70(m, 2 H, H8 ), 5.02 (d of d, J = 4.5, 2.5 Hz, 1 H, H2 ), 5.20-5.45(m, 1 H, H10),

5.87-6.28(m, 2 H, H 3 and H 5), and 6.28-6.57(m, 2 H, % and h 5 ). Anal. Calcd for C^gH^Og: C, 75.76; H, 7.42. Found: C, 75.38; H, 7.60.

Cyclooctatetraenylet hyJL Acetate (89-OAc). To a solution of 500 mg (3.38 mmoles) of 8£~0H in 1.5 ml of pyridine was added 1,6 ml (17.0 mmoles) of acetic anhydride. The solution 115 was stirred at room temperature for 2h hrs. The reaction

mixture was added to 5 ml of water and extracted with 10# hydrochloric acid, saturated sodium bicarbonate solution, and dried over magnesium sulfate. Removal of the ether gave 610 mg (95#) of a colorless liquid. The ester was chromatographed on a 6* column of 5# SF-9 6 on Chromosorb G at 115°. Nmr spectral data for 89-OAc; 2.02(s, 1 H, hydroxyl), 2.37

Anal. Calcd for C12 H ^ 0 2: C, 75.76; H, 7.^2. Foundi C, 75.65; H, 7.^9. *

l-Hydroxy-l,2,3,il'-tetrahydroazulene (155-OH). To a suspension of 27 mg (O.70 mmole) of lithium aluminum hydride in 5 ml of ether was slowly added a solution of 125 mg (0.70 mmole) of 155,-OAc in 10 ml of ether. The mixture was re fluxed for 3 hrs and a standard basic workup was used. The mixture was filtered through magnesium sulfate and the ether was evaporated ^n vacuo to give 100 mg of residue. The alcohol was purified by chromatography on silica gel and

eluted with 10# ether-pentane. Nmr spectral data for 155-OH;v w ^TMS13 1*67-2.20(m, 3 H, Hi and H9 ), 2.42-2.90(m, 3 H, H8 and -OH), 4.39(d of d, J = 3.0, 1.0 Hz, 1 H, Hi0)* **.98(d of d, J = 2 .5, ^.5 Hz, 1 H, H2 ), 5.92-6.33(m, 2 H, H3 and Hg), and 6.33-6.63(m, 2 H, H4 and H^), 116 * A solution of 272 mg (1.8** mmoles) of 155-OH In 5 nil of methanol was hydrogenated at ^rvA/ atmospheric pressure using 25 mg of 10# Pd/C catalyst. The hydrogenation was allowed to proceed for 5 hrs when no further uptake of hydrogen was observed. The reaction mix ture was filtered through Cellte and the methanol was re moved In vacuo to give 111 mg of l£g. Chromatography on silica gel with 10# ether-pentane elution gave relatively pure 159; 8?S§13 0.83(m, 15 H), and 3,70(m, 1 H, H alpha to -OH). f ~ ^e^hydroazulenone _(160). To a solution of 25. mg

(0.162 mmole) of 159 In 15 ml of dry benzene was added 0.5 ml of dimethylsulfoxide, 0.02 ml of pyridine, 0,01 ml of trifluoroacetic acid and 97 mg (0.**86 mmole) of dl- cyclohexylcarbodilmlde. The mixture was stirred at room temperature for 15 hrs and added to 5 ml of water. The aqueous solution was extracted with ether and the ether portion was subsequently washed with water, saturated sodium chloride solution, and dried over magnesium sulfate. Removal of the ether gave **5.1 mg of residue, which con tained some impurities. Ketone 16Q was Isolated pure from a 6* column of 5# SE-30 on Chromosorb G at 125°; 0.92-2.58(br m, spikes at 1.57t 2.10, 2.12). The ir and nmr of this ketone were identical with those of an authentic 11? sample of cis-l-perhydroazulenone. A semlcarbazone derivative of 160 was prepared by- adding 149 mg (0.132 mmole) of semicarbazide hydrochloride and 5^ mg of sodium acetate to a solution of 20 mg of 160 (0.132 mmole) in 5 ml of 95/6 ethanol. The mixture was heated on a steam bath for 10 min and 5 ml of water was added. The reaction was filtered to give 20 mg of white crystals which, upon recrystallization from absolute ethanol, had a melting point of 220-1°, identical with that reported in the literature for the seraicarbazone derivative of cis- OQ 1-perhydroazulenone.

i Cyclooctatetraenylacetic acid (1 6 2 K To a solution of 2.00 g (13*5 mmoles) of 0-cyclooctatetraenylethanol (j3g-OH) in 200 ml of acetone at 5° was slowly added 9,0 ml (1 9 .1 mmoles) of Jones* reagent (2,01 M). After the addition was complete, a red color, persisted and the solution was allowed' to warm to room temperature and neutralized with 10# sodium hydroxide solution. Filtration followed by evaporation of solvent left an oily residue which was dissolved in 200 ml of ether and subsequently extracted with saturated sodium bicarbonate solution. Evaporation of the ether gave 0.80 g of liquid residue which contained some starting material and some COTacetaldehyde, This liquid was reoxidlzed to give 0.40 g of acid (20#). The bicarbonate solution was 118 acidified, extracted with ether and dried over magnesium sulfate to give an additional 0.90 g of acid (6l# combined yield) as a yellow oil. All attempts to induce crystal lization were unsuccessful. Spectral data for 162; 3.07(s, 2 H, methylene), 5.73(s, 7 H, vinyls), and 7.92(s, 1 H, -C00H).

Methyl Cyclooctatetraenylacetate (166), A solution of 17.1 S (79.8 mmoles) of N-nitroso-N-methyl-]D-toluenesulfon- amide in 150 ml of ether was placed in a 250 ml Erlenmeyer flask fitted with a rubber stopper and a piece of bent glass tubing extending into a second flask containing 50 ml of ether and cooled to -5°. Upon addition of 18.0 g (0,32 mole) of potassium hydroxide in 3° ml of water, the diazomethane began to distill over into the cooled receiver. When the evolution of gas ceased, the ethereal diazoraethane solution was carefully added to a solution of ^,35 6 (26.9 mmoles) of 162 in 50 ml of ether. This solution was stirred at -5° for 15 min and allowed to warm to room temp erature where it was stirred for an additional 30 min. The solution was dried over magnesium sulfate and the ether was evaporated to give 3.89 S (81#) of crude ester. The ester was purified for analysis by preparative vpc (6*, 10# SE-30 column, followed by 6* 10# SF-1150). Spectral data for

^TMS13 3.08(s , 2 H, methylene), 3*68(s, 3 H, methyl), and 119

5.83

~~or, of-Dideutero-jg-cyolooctatetraenylethanol (163-OH). To a solution of 0.88 g (21.0 mmoles) of lithium aluminum deut-~~

eride in 25 ml of ether was slowly added 3*70 g (21.0 mmoles) of 162 in $0 ml of ether. The mixture was gently refluxed for 2 hrs and stirred at room temperature for 7 hrs. A basic workup was used (1 ml of water, 1 ml of 15$ sodium hydroxide solution, and 3 ml of water). Filtration through magnesium sulfate and subsequent evaporation of solvent gave 1.92 g (6l$) of 16^-OH. The crude liquid was distil

~~led at 49-5°° and 0.02 mm to give 1 .1 5 g of pure alcohol. The nmr showed 98$ isotopic purity; £^3^3 2.27(s, 2 H, allylic methylene), 3.15(br s, 1 H, hydroxyl), and 5»77(s, 7 H, vinyls).~~

~~Brosylate of 163-OH* To a solution of 770 mg (5#14 mmoles)~~

of I63-OH in 50 ml of pyridine at 0° was added 2,43 g (9*50 mmoles) of £-bromobenzenesulfonyl chloride. The reaction was left to sit in the refrigerator for 12 hrs and was poured over ice. The resulting aqueous solution was extrac ted with ether and the combined ether portions were washed with water, 10$ hydrochloric acid, saturated sodium bicar bonate solution, and dried over magnesium sulfate. Removal of the ether in vacuo gave 1.45 g of yellow oil. All 120 attempts to crystallize this oil were unsuccessful although nmr analysis showed greater than 95% purity* Nmr spectral data for 163-OBs; 2.37(s, 2 H, allylic methylene),

~~5.68(s , 7 H, vinyls), and 7,72(s, k H, aromatics),~~

A solution of 1.10 g (3.00 mmoles) of 163-OBs in 150 ml of 0,19^5 N sodium acetate/acetio acid . ‘- ' W * — was heated at 75° for a period of 10 half-lives (4 days, 7*5 hrs). The mixture was cooled and added to 100 ml of ice-water. The aqueous solution was extracted with ether and the combined ether portions were washed with water, saturated sodium bicarbonate solution, and dried over mag nesium sulfate. Removal of the ether in vacuo gave 550 mg of dark residue which was first chromatographed on silica gel and eluted with pentane. The mixture was purified by isolation from a 6* column of 5% XF-1150 on Chromosorb G at 1^0°. Separation and subsequent nmr analysis showed three components; 16^-OAc (deuterium contained in the alpha methylene only), 16^ (deuterium equally distributed between

~~C0 and C9 ), and 16£ (deuterium equally distributed between Ci and C2 ).~~

~~Methyl {B,/3-Dideuterio-#-cyclooctatetraenylethyl)Acetate (I67J. To i*00 mg (17.0 mmoles) of methanol-d^ was slowly added Jf.O mg of sodium and to this solution was introduced~~

150 mg (O.85 mmole) of 166. The mixture was heated at 50° for 12 hr, cooled to room temperature, and the solvent was removed in vacuo. The residue was taken up in pentane and washed with saturated sodium chloride solution. Removal of solvent gave 89.6 mg of yellow liquid, which was purified by molecular distillation. Nmr spectral data for 167;

~~3 .60(s , 3 H, methyl), and 5.?2(s, 7 H, vinyls).~~

suspension of 0.285 & (7.75 mmoles) of lithium aluminum hydride in 50 ml of dry ether was slowly added a solution of 1.38 6 (7.75 mmoles) of 16^ in 25 ml of ether. The mix ture was refluxed for ^ hrs and a basic workup was used (0.3 ml of water, 0.3 ml of 10^ sodium hydroxide solution, and 0.9 ml of water). The mixture was filtered through magnesium sulfate and the ether removed in vacuo to give 1 .0 5 g of yellow liquid which was purified by molecular distillation. Nmr spectral data for 168-OH; 2.13(s, 1 H, hydroxyl), 3*62(s, 2 H, alpha methylene), and 5«77(s, 7 H, vinyls).

Br0sylate of 168-OH, To a solution of 8*4-0 mg (5*60 mmoles) of 168-OH in 50 ml of pyridine at 0° was added 2.86 g (11,20 mmoles) of ja-bromobenzenesulfonyl chloride. The solution was left to sit in the refrigerator for 12 hrs and was poured over ice. The resulting aqueous solution was extracted with ether and the combined ether portions were washed with water, 10^ hydrochloric acid, saturated sodium bicarbonate solution, and dried over magnesium sulfate. Removal of the ether in vacuo gave 1 .0 0 g of yellow oil which could not be crystallized. Nmr analysis showed greater than 95% purity; *K05(s, 2 H, alpha methylene), 5.72(s, 7 H, vinyls), and 7*75(s, 4 H, aromatlcs),

~~Sol^lysi^^^fig-OBs^. A solution of 1.00 g (2.75 mmoles)~~

of 168-OBs in 30 ml of 1.0 N sodium acetate/aceticacid was heated at 75° for a period of 10 half-lives (*f days, 7 .5 hrs). The mixture was then cooled and added to 100 ml of ice-water. The aqueous solution was extracted with ether and the combined ether portions were washed with water, saturated sodium bicarbonate solution, and dried over mag nesium sulfate. Removal of the ether in vacuo gave **20 mg of.dark residue which was first chromatographed on silica gel and eluted with pentane. The mixture was further purified by vapor .phase chromatography on a 6* column of 5% XF-1150 on Chromosorb G at 1**0°, Separation and subsequent nmr analysis showed three components; 158-OAc (deuterium contained in the beta methylene only), l6J* (deuterium equally

~~distributed between Cq and C9 ), and 16£ (deuterium equally distributed between and Cg).~~

l^Carboethoxy-^-meth^lcjrc^ A solution of 15 ml of ethyl tetrolate (169 ) and 500 ml of benzene was placed in a quartz tube fitted with a glass stopper and irradiated for a total of 32 hrs in a Rayonet reactor using 2537 A6 lamps. The quartz vessel was treated with cleaning 123 solution (potassium dichromate in sulfuric acid) after each 8 hr photolysis period. After a total reaction time of 32 hr the benzene and remaining ethyl tetrolate were removed by distillation (33°/l40 mm) and the residue was distilled (6l°/0.075 mm) to give a 45$ yield of 170. The nmr spectrum was found to be consistent with that reported in the literature^. 1.23(s, 3 H, ester methyl),

~~1.83(s , 3 H, allylic methyl), 4.20(q, 2 H, ester methylene), 5.65-6.12(m, 5 H, vinyls), and 7.00(d, J - 2 Hz, 1 H, vinyl alpha to carboethoxy).~~

slurry of 1.55 g (40.75 mmoles) of lithium aluminum hydride in 75 ml of ether was slowly added a solution of 7.74 g (40.75 mmoles) of 17(} in 50 ml of ether. The mixture was refluxed gently for 5 hr, cooled, and worked up in the customary basic fashion. The mixture was filtered through magnesium sulfate and the solvent was removed in vacuo to give 5*91 g (96#) of yellow liquid. Nmr spectral data for 17J; 1.85(s, 3 H, methyl), 2.47(s, 1 H, hydroxyl), 4.12(s, 2 H, methylene), and 5*S0(s, 6 H, vinyls).

solution of 2.40 g (16.2 mmoles) of 171 in 160 ml of chloroform was added 35 g of Mn02 /c9 \ The slurry was stirred at room temperature for 24 hrs and filtered through Celite to give 2 .2 0 g (93$) of clear, yellow liquid; ^JJax^ 1^95 cm" 1 124

(carbonyl); &Sijjp3 1.83(s , 3 H, methyl), 5.80(m, 2 H, vinyls), 6.05(m, 2 H, vinyls), and 6,82(d, J = 2 Hz, 1 H, vinyl H conjugated with aldehyde group). The 2,4-dinitro- phenylhydrazone of 1£2 was prepared in the following manner. To 300 mg of 1£2, in 1 ml of ethanol was added approximately 3 ml of a 1,5 M solution of 2,4-dinitrophenylhydrazine in aqueous ethanol. Reaction was immediate but was left to sit in the refrigerator overnight. The solid was filtered and recrystallized once from methanol and again from chloroform-pentane to give a dark red, crystalline product, mp 204-5°. Anal. Calcd for : C, 58.89; H, 4.32; N, 17.17. Found: C, 58.62; H, 4.44; N, 16.91.

2^Meth^3^(^metl^x^^T^J^^clo^t^^i«^^^(l^Jij A solu tion of sodium ethoxide in ethanol was prepared by adding 920 mg (40,0 mg-atoms) of sodium to 100 ml of absolute ethanol. After all the sodium had reacted, this solution was slowly added to a solution of 13.7 S (40.0 mmoles) of methoxymethyltriphenylphosphonlum c h l o r i d e ^ in 50 ml of ethanol. The mixture was stirred for 4 hrs and the solvent was removed in vacuo to give a white solid. The crystal line Wittig reagent was taken up in 50 ml of ether and slowly added to a solution of 2.14 g (12.5 mmoles) of 172 in 25 ml ether. The solution was refluxed for 1 hr, cooled, and filtered through alumina. The solvent was removed in vacuo and the residue was triturated twice with pentane and 125 the precipitated triphenylphosphine oxide removed by fil tration. The ethereal solution was concentrated to give 2.14 g (82$) of a yellow liquid which was purified by passage through a 6* column of 5$ Carbowax on Chromosorb G at 135°. The vinyl ether proved to be extremely unstable and polymerized upon standing for more than 24 hrs. Spec

tral data for 173;l#v\/ calcd “m/e “ 174.1045, found, 174.1041; 8TMSl3 1-75(s , 1/3(3 H), methyl), 1.88(s, 2/3(3 H),'methyl), 3.55(s, 3 H, methoxyl), 5.52(d, J « 13 Hz, 1 H, vinyl), 5.70(m, 6 H, ring vinyls), and 6.50(d, J = 13 Hz, 1 H, vinyl adjacent to methoxyl). One must realize that this nmr data is for the major component of the reaction and is not necessarily the sole isomer formed. Also, two methyl groups appear due to the presence of two possible valence tautomers; namely,

~~and CH=CH0CH CH=CH0CH~~

Hydrolysis^of^ To a solution of 3*58 S (12.3 mmoles) of mercuric acetate in 50 nil of 50$ aqueous tetrahydrofuran was added 2.14 g (12.3 mmoles) of 173. The mixture was stirred for 1 min at room temperature and quenched with 20 ml of 15$ sodium hydroxide solution and 467 mg (12.3 mmoles) of sodium bprohydride in 20 ml of 15$ sodium hydroxide solution. The 126 reaction mixture was filtered through Celite and the fil trate was neutralized with 3 N hydrochloric acid* The neutral solution was extracted with ether and the combined ether layers were washed with 5° ml portions of water, saturated sodium bicarbonate solution, and dried over mag nesium sulfate. Removal of the ether gave 1.70 g of residue which was chromatographed on silica gel. Elution with pentane gave 1 .1 0 g ($6%) of .17^ while elution with 20 ^ ether- pentane gave 520 mg of 175-OH. 174. showed an observed accurate mass of 158.0708 (calcd 1 58.0731) and a~Xmax at 235 nm (£ - 31,400) with a shoulder having an £= 4,000. Nmr spectral data for 1^4; 1.87(s, 3 H, methyl), 5*^5 (br s, 3 H, COT vinyls), 6.05(br s, 2 H, COT vinyls), 6.20 (d, J * 2 Hz, 1 H, furan vinyl), and 7.40(d, J = 2 Hz, 1 H, furan vinyl adjacent to oxygen). Anal. Calcd for C11H100: C, 83.51; H, 6.37. Pound: C, 8 3.57; H, 6.44. Nmr spectral data for l£4-OH; 1.68(s, 1/3(3 H), methyl), 1.80(s, 2/3(3 H), methyl), 2.38(t, J = 7 Hz, 2 H, allylic methylene), 3*60(tj J = 7 Hz, 3 H, methylene and hydroxyl), 5«63(s, 2 H, vinyls), and 5«80(s, 4 H, vinyls).

Brosylate of 175-OH. To a solution of 218 mg (1.35 mmoles) of 175-OH in 15 ml of pyridine cooled to 0° was slowly added 344 mg (1 .3 5 mmoles) of ]D-bromobenzenesulfonyl chloride. The reaction mixture was allowed to remain in the 127 refrigerator for 8 hrs before adding it to 50 ml of water. The aqueous solution was washed with ether and the combined ether portions were washed with water, 3 N hydrochloric acid, saturated sodium bicarbonate solution, and dried over magnesium sulfate. The solvent was removed in vacuo to give Jj-50 mg (88^>) of yellow oil which was crystallized from ethanol to give pale yellow needles, mp 74-5°$ 1*63 (s, 1/3(3 H), methyl), 1.75(s, 2/3(3 H), methyl), 2.52(t, J = 7 Hz, 2 H, allylic methylene), 4.15(t, J = 7 Hz, 2 H, methylene), 5-63(s, 2 H, vinyls), 5.78(s, 4 H, vinyls), and 7.82(s, 4 H, aromatics). Anal. Calcd for C ^ H ^ B r O ^ S : C, 53.55; H, 4.49; S, 8.41. Pound: C, 53.46; H, 4.60; S, 8.4 7.

a solution of 710 mg (5*30 mmoles) of 88-OH in 50 ml of chloroform was added 10 g of Mnt^/C^. The mixture was stirred at room temperature for 5 hrs, filtered through Celite and mag nesium sulfate and the solvent was removed in vacuo to give 600 mg (85#) of yellow liquid with the following nmr spec trum; 5.92(m, 7 H, vinyls), and 7.65(s, 1 H, alde hyde ), 128 ^ solution of sodium ethoxide In ethanol was prepared by adding 0.185 g {8,0 mg- atoms) of sodium to 15 ml of absolute ethanol. After all the sodium had reacted, this solution was slowly added to a solution of 2.75 S (8.0 mmoles) of methoxymethyltriphenyl-

phosphonium chloride^ in 20 ml of ethanol. The reaction . was stirred for k hrs and the solvent was removed in vacuo to give a white solid. The crystalline Wittig reagent was taken up in 25 ml of ether and slowly added to a solution of 500 mg (3.79 mmoles) of 176 in 10 ml of ether. The mixture was refluxed for 1 hr, cooled, and filtered through alumina. The solvent was removed in vacuo and the residue was triturated twice with pentane. The precipitated triphenylphosphine oxide was removed by filtration and the ethereal solution was concentrated to give 570 mg of vinyl ether (91*#) which was used immediately in the next reaction.

Hy^olysi^of^l7^. To a solution of 1.03 g (3.25 mmoles) of mercuric acetate in 50 ml of aqueous tetrahydrofuran was added 520 mg (3.25 mmoles) of vinyl ether (177). The reaction was stirred for 1 min at room temperature and quenched with 10 ml of 15% sodium hydroxide and 123 mg (3.25 mmoles) of sodium borohydride in 10 ml of 15# sodium hydroxide solution. The reaction mixture was filtered through Celite and the filtrate was neutralized with 3 N hydrochloric acid. The neutral solution was extracted with 129 ether and the combined ether portions were washed with 50 ml portions of water, saturated sodium bicarbonate solution and dried over magnesium sulfate. Removal of the ether gave 330 mg of 89-OH (70^).

Solvolysis of 175-OBs, A solution of $60 mg (1.47 mmoles) of 175-OBs in 50 ml of a 0.20 N solution of sodium acetate in acetic acid was placed in a preheated oil bath at 75° for 4 days and 8 hrs. At this time the solution was cooled and poured into 100 ml of water. The resulting solution was extracted with ether and the combined ether solutions were washed with water, saturated sodium bicarbonate solution, and dried over magnesium sulfate. The ether was removed in vacuo to give a colored residue which was chromatographed on a short column of silica gel and eluted with pentane to give 270 mg (92#) of a yellow liquid. Nmr and vpc analysis showed the presence of only the unrearranged acetate, l£§-0Ac; S§gg13 1.68(s, 3 H, allyl methyl), 1.85(s, 3 H, ester methyl), 2,32(t, J = 7 Hz, 2 H, alpha methylene), 3.97(t, J - 7 Hz, 2 H, allyl methylene), 5.52 and 5.65(m, 6 H, vinyls). l-Bromo-4-methylcyclooctatetraene and 1-Bromo-5-methyl- cyc^octatetr^^ A solution of 13.7 fi (0.116 mmole) of methylCOT in 200 ml of methylene chloride was cooled to -70° under nitrogen. To this was slowly added a solution of 6.4 ml of bromine (0.116 mmole) in 100 ml of methylene chloride* After the addition was complete, the mixture was stirred at -70° for 1 hr. An Erlenraeyer flask containing

16*8 g (0 .1 5 mmole) of potassium t-butoxide was affixed to the flask by means of a Gooch tube and approximate 1 g portions of base viere added to the mixture at 15 minute intervals. After addition was complete, the mixture was again stirred for 1 hr at -70° and then allowed to slowly rise to room temperature, A solution of 20 ml of acetic acid in 200 ml of saturated sodium chloride solution was added to the reaction and stirred thoroughly. The organic layer was separated, washed with 100 ml portions of water, saturated sodium bicarbonate solution, and dried over mag nesium sulfate. Removal of solvent gave 15.5 S of dark residue which was chromatographed on silica gel and eluted with pentane to give 10,5 6 (50^) of yellow liquid which was used directly in the next reaction. Vpc analysis showed gross impurities and this was supported by the nmr spectrum. l^-^and^jJJ-Diii^jTj!^^^ To a suspen sion of 3.81 g (20.0 mmoles) of cuprous iodide in $0 ml of ether cooled to 0° was slowly added 2 6 .6 ml of a 1,5 N solution (40,0 mmoles) of methyllithium in ether. The mix ture was stirred for 15 min at 0*J cooled to -70° and a solution of 1.00 g (5.08 mmoles) of 172, ether was added. The mixture was stirred at -70° for 1 hr and allowed to rise to room temperature. To the reaction was slowly added 25 ml of 10# aqueous ammonium chloride sol ution. The layers were separated and the ether portion was washed with saturated ammonium chloride solution until the latter was clear (ca 3 washings). The ether portion was washed with water and saturated sodium chloride solution, dried over magnesium sulfate, and the solvent was removed in vacuo to give 0.68 g of yellow liquid. Vpc analysis on a 5* column of 12# OV-11 on Chromosorb G at 77° showed one major peak (67# of volatile components) which appeared to be 1,^-dimethylCOT with a possible contamination of the 1,5-isomer. Ten percent of the reaction mixture was methyl- COT and trimethylCOT (possibly a combination of isomers) accounted for 21# of the mixture. Nmr spectral data; dimethy 1C0T; 1.72(s, 6 H, methyls), 5.53(s, 2 H, vinyls), 5*67 and 5.72(s, 4 H, vinyls); trimethylCOT; l.72(s, 9 H, methyls), 5.**7(s, 3 H, vinyls), and

~~5 .67(s , 2 H, vinyls).~~

182-OHj. A solution of 10,5 S (53*3 mmoles) of l£g in 200 ml of ether was cooled to -70° under nitrogen. To this was slowly added a solution of 26 ml of n-butyllithium in pentane (2.3^ N, 60.0 mmoles). After the addition was complete, a large excess of ethylene oxide (100 g) was added and the mixture was stirred at -70° for an addition 2 hrs before allowing the solution to rise to room temperature. The reaction was acidified with 3 N hydrochloric acid and the 132 ether portion separated, washed with 100 ml portions of water, saturated sodium bicarbonate solution, and dried over magnesium sulfate. Removal of solvent gave a red- orange oil which was distilled at 55-6° and 0.02 mm to give 3*28 g (38#) of a clear, yellow liquid. Vpc analysis on a 6* column of 5$ Carbowax-l$ potassium hydroxide on Chromo- sorb G at 138° showed a 3:1 mixture of the 1,4- and 1,5- isomers (181 and 182). Nmr spectral data: (181-OH);

&TMS13 1 *69 (s , 3 H, methyl), 2.25(t, J = 7 Hz, 3 H, allylic methylene and hydroxyl), 3.53(b** s, 2 H, methylene), 3*82 (s, 2 H, vinyls), and 5*80(s, 4 H, vinyls); (182-OH); 8fflP3 1.6?(s, 3 H, methyl), 2.23(t, vT = 7 Hz, 2 H, allylic methylene), 2.62(s, 1 H, hydroxyl), 3*54(br s, 2 H, methyl ene), 5*52{s, 2 H, vinyls), and 5.70{s, 4 H, vinyls),

8rpsylate^qf^181-OH. To a solution of 82 mg (O.5O6 mmole) of 181-OH in 10 ml of pyridine at 0° was added 205 mg (0,800 mmole) of £-bromobenzenesulfonyl chloride. The mix ture was allowed to remain in the refrigerator overnight before pouring into 50 ml of water. The aqueous solution .* was extracted with two 50 ml portions of ether and the com bined ether portions were washed with 50 ml portions of water, 3 N hydrochloric acid, saturated sodium bicarbonate solution, and dried over magnesium sulfate. Removal of ether in vacuo gave 1.60 g of crude yellow oil. Crystallization from ethanol gave 57.0 mg (31$) of yellow solid, mp 34-5°* 133 Nmr spectral data; 1 *72(s, 3 H, methyl), 2.31(t, J = 7 Hz, 2 H, allylic methylene), 4.02(t, J = 7 Hz, 2 H, methylene), 5.46(s, 2 H, vinyls), 5.62(s, 4 H, vinyls), and 7.66(d, J = 2 Hz, 4 H, aromatics). Anal, Calcd for C-^H^BrO-jS: C, 53-55; H, 4.49; S, 8.41. Found: C, 53.48; H, 4.53; S, 8.50.

Brosylate - OB s,. To a solution of 48 mg (0.286 mmole) of 182-OH in 10 ml of pyridine cooled to 0° was added 103 mg (0.400 mmole) of £-bromobenzenesulfonyl chloride. The subs equent procedure and workup were the same as above. The yield of crystalline brosylate after two recrystallizations from ethanol was 5.1 rag (5$)» mp 47-8°. Nmr spectral data; fiTMSl3 1 '7°(s» 3 H, methyl), 2.35

Iron Tricarbonyl Complex of 181-OH (183-OH). To a solution of 225 rag (1.39 mmoles) of 181-OH in 20 ml of ether was added 1.09 g (3.00 mmoles) of iron enneacarbonyl. The mix ture was refluxed gently for 2 hrs, cooled, and the solvent was removed in vacuo. The residue was chromatographed on silica gel and eluted with 10$ ether-pentane to give 120 mg of a red oil which was used directly in the next reaction. Spectral data; &rj$jjep3 1.86{s, 4 H, methyl and hydroxyl), . . 2.28(t, J = 7 Hz, 2 H, allylic methylene), 3*64(s, 2 H, methylene), and 4.40-5.80(m, 6 H, vinyls). p-Nitrobenzoate of I83-OH, To a solution of 120 mg (0,40 mmole) of 18^-OH In 10 ml of pyridine cooled to 0° was added 110 mg (0 .8 0 mmole) of recrystallized £-nitrobenzoyl chloride. The mixture was stirred at 0° for 1 hr and quenched with 50 ml of water. The aqueous mixture was extracted with ether and the combined ether portions were washed with 50 nil portions of water, 3 N hydrochloric acid, saturated sodium bicarbonate solution, and dried over mag nesium sulfate. Removal of the ether Jja vacuo gave 100 mg of red oil which was crystallized from ethanol to give 81 mg (45#) of red crystalline material, mp 82-3°. Nmr spectral data; 1.90(s, 3 H, methyl), 2.62(t, J = ? Hz, 2 H, allylic methylene), 4.46(t, J » 7 Hz, 2 H, methylene), 4.40-

~~5.80(m, 6 H, vinyls), and 8.20(d, J = 2 Hz, 4 H, aromatics).~~

~~Anal. Calcd for C21 H 17FeN07: C, 55.90; H,3.80; N, 3.10. Found: C, 55.67; H, 3-78; N, 3.21.~~

Sqlyolysls of 181-OBs, A solution of 540 mg of 181-OBs in 75 nil of 0.20 N sodium acetate in acetic acid was placed In a preheated oil bath at 75° for 4 days and 8 hrs. The mix ture was processed as with 17^-OBs to give 300 mg of a slightly colored liquid. Vpc analysis (6* column of 5# SE-30 on Chromosorb G at 125°) showed a 20# conversion to 181-OAc and an 80# mixture of 184-OAc.and 18£-0Ac. All attempts to separate these isomers were unsuccessful. Mass spectral 135 analysis of this mixture showed a parent ion at m/e 204,1148 (calcd 204.1150).

a solution of 50*0 mg (0.309 mmole) of 18£-0H in 0.25 ml (3*10 mmoles) of dry pyridine was slowly added 3 .1 6 mg (3*10 mmoles) of acetic anhydride. The reaction was stirred at room temperature for 5 hrs and quanched with 10 ml of water. The resulting aqueous solution was extracted with ether and the combined ether portions were washed with 10 ml portions of water, 10^ hydrochloric acid, saturated sodium bicarbonate solution, and dried over magnesium sul fate. Hemoval of the ether in vacuo gave 55 m£ of yellow* liquid which was purified by vpc on a 61 column of 5% SF-96 on Chromosorb G at 140°. Nmr spectral data for 181- OAc; o§§g13 l.?4(s, 3 H, allylic methyl), 2.04(s, 3 H, ester methyl), 2.34(t, J = 7 Hz, 2 H, allyl methylene), 4.09(t, J - 7 Hz, 2 H, methylene), and 5*68(m, 6 .H, vinyls). Anal. Calcd for C, 76.05; H, 7*97* Found: C, 76.44; H, 7.90.

of 57 mg (1.50 mmoles) of lithium aluminum hydride in 50 ml of ether was slowly added a solution of 300 mg (1.47 mmoles) of acetate mixture in 50 ml of ether. The mixture was stirred at room temperature for 2 hrs when 0 .0 6 ml of water, 0 .0 6 ml of 10$ sodium hydroxide solution, and 0.20 ml of water were added. The reaction mixture was filtered through magnesium sulfate and the solvent was removed in vacuo to give 200 mg (8*f$) of alcohol mixture 181-OH, 18^-0H, and 185-OH.

Prepara11 on of Tr 1methyjj.si lyl^ Ethers ^of 181-OH , 18^-OH^ and^JL8^-MI. To a solution of 200 mg (1.23 mmoles) of alcohol mixture in 1 ,0 ml of pyridine (12 .^ mmoles) was added a solution of 1.35 S (12.3 mmoles) of trimethylchlor- osilane in 5 ml of ether. This solution was stirred at room temperature for 1 hr, quenched with water, and the ether layer was separated. The ethereal solution was washed with 5 ml portions of water, 2 N hydrochloric acid, satur ated sodium bicarbonate solution, and dried over magnesium sulfate. Removal of ether in vacuo gave 2^2 mg (85$) of product. The ethers were separated on a 6* column of 10$ QF-1 on Chromosorb G at 125° to give isomerically pure 18^ (40$) and 18£ (4-0$), Spectral data for l^-OSMCH^)^; calcd m/e 23^.1440, found, 23^.1*06; 6tmS13 °-28(s , 9 H, ether methyls), 2.10(s, 3 H, methyl), 2.20-3.00(m, 5 Hi Hi, H7, and Hg), 4.^0(br s, 1 H, H^), 5.00(m, 1 H, H2 ), 6.02 (m, 2 H, H3 and Hg), and 6.32(m, 2 H, H*j, and H^), Spectral data for 18J-OS1 (GH3 )yt calcd ra/e 23^. 1^0, found, 23 ^.1^36;

~~^TMeP^ 0*28(s, 9 H, ether methyls), 2.08(s, 3 H, methyl), 2.20-3.00(m, 5 H, Hi, H?, and Hg), ^.80(br s, 1 H, H9 ),~~

6.08(m, 2 H, H3 and Hg), and 6.k0(m, 1 H, H5). 137 To a solution of 30.8 mg (O.I32 mmole) of IR^-OSitCH^)*j in 10 ml of 50^ aqueous tetrahydrofuran was added 30 rag of £-toluenesulfonlc acid. The mixture was stirred at room temperature for 5 hrs and the product was extracted with ether. The combined ether portions were washed with 10 ml portions of water, satur ated sodium bicarbonate solution, and dried over magnesium sulfate. Removal of the ether gave 20 mg of 184-OH.

^-Nethyl-9,10-dihydronaphthalene (186). To a solution of 18£-0H in 2 ml of acetic acid was added 25 mg of E-toluenesulfonic acid. The solution was heated at 50° for a period of 12 hrs, cooled, and diluted with 10 ml of water. This aqueous solution was extracted with ether and the combined ether portions were washed with 10 ml portions of water, saturated sodium bicarbonate solution, and dried over mag nesium sulfate. Removal of the ether gave 15 mg of crude I methyldihydronaphthalene; xrl& 2.5^(s , 5 H, methyl and allylic methylene), 2.96(t, J - 8 Hz, 2 H, benzylic methyl ene), 6.l8(m, 1 H, vinyl), 6.64(m, 1 H, vinyl), and 7,l6(s, b H, aromatics).

~~To a solution of 1 5 .0 mg~~

(0 .10^- mmole) of 186 in 2 ml of benzene was added 82 mg (0.366 mmole) of DDQ. The mixture was stirred at room temp erature for 1 hr and passed through an alumina column, eluting with pentarc. The solvent was carefully removed in vacuo to give a colorless residue which was identified as jB-methylnaphthalene by comparison of its nmr and vpc retention time with that of an authentic sample, Nmr spectral data for 18^; 2*38(s, 3 H, methyl), and 7.06-7*70(m, 7 H, aromatics). 139

~~Table V- Solvolysls of 92-C1 in 5 0 % Aqueous Acetone at 30.1°.~~

~~Concentration of 92-C1: 0.0319 M, Tltrant: NaOH 0.01042 N.~~

~~Aliquot Time (sec) ini (NaOH) 1 0 0.282 2 2000 0.743 3 4*100 0.15 1 4 6000 1.4-50~~

~~5 8000 1.750 6 10100 2 .0 5 1~~

~~7 00 3.400 i4o~~

~~Table^VI^. Solvolysis of Sj^-Cl in $ 0 % Aqueous Acetone at 40,0°.~~

~~Concentration of %2,-Cl: 0,0339 M Titrant: NaOH 0.01042 N~~

~~Aliquot Time(sec) ml(NaOH) 1 0 0.530 2 2000 1.371 3 4000 1 .8 2 0 4 6000 2.142 5 8000 2.341 6 10000 2.481 7 Qo 2 .7 2 0 Table^VII. Solvolysis of 92-C1 in $ 0 % Aqueous Acetone at 49.9°.~~

~~Concentration of 92-C1:w v 0.0294 —M Titrant: NaOH 0.01042 N~~

~~Aliquot Time (sec) ml(NaOH) 1 0 0.570~~

~~2 200 0.821 3 300 0.931 4 400 1.020 5 500 1.163 6 Oo 2.721 Table^VIII. Solvolysis of 88-C1 In 50% Aqueous Acetone at 30.1°.~~

~~Concentration of 88-C1: 0.0270 M — Titrant: NaOH 0.010^2 N~~

~~Aliquot Time(sec) ml(NaOH) ■~~

~~1 ■ 0 0 .3 0 0 2 5000 0.551 3 10000 0 .75 0 15000 0 .9^0 5 20000 1 .1 0 1 6 25000 1.260 7 00 3 .022 Table IX, Solvolysls of JJ8-C1 in 50$ Aqueous Acetone at 40.0°.~~

~~Concentration of 88-C1: 0,0274 M Titrant: NaOH 0.01042 N~~

~~Aliquot Time(sec) ml(NaOH)~~

~~l 0 0.31 2 2 2100 0.571 3 4000 0 .7 8 0 4 6000 0.96 2 5 8000 1.143 6 10000 1 .3 0 0 7 oo 3.121 Table^X. Solvolysis of 88-C1 in 50$ Aqueous Acetone at 49.9°.~~

~~Concentration of UVv88-C1: 0.0241 M — Titrant: NaOH 0.01042 N~~

~~Aliquot Time(sec) ml(NaOH) 1 0 0.550 2 300 0.662 3 600 0.733 4 900 0.801 5 1200 0.872 6 1500 0.930 7 co 2.600 145~~

~~Table^XX, Solvolysis of 140-OBs In 0.10065 K NaOAc/HOAc at 75.0°.~~

~~Concentration of 140-OBs: 0.0600 M Titrant: HClO^/HOAc 0.02043 N~~

~~Aliquot Time (sec) ml(HC10jj,/H0Ac)~~

~~1 0 4 .630 2 2500 ^.531 3 5000 4,442 4 7500 4.411 5 12500 4.630 6 17600 4.352 7 22500 4.303 8 35000 4.171 9 00 2.152 146~~

~~Table XII.* Solvolysls of 140-OBs v^V\ In O.IOO65 N NaOAc/HOAc at 85.0°.~~

~~Concentration of V^i/V140-OBs: 0.0601 M Titrant: HClO^/HOAc 0.02043 N~~

~~Aliquot Time(sec) mKHClOz/HOAc)~~

~~1 0 4.540~~

~~2 2500 4.440~~

~~3 55 00 4.312 4 7600 4.243 5 10000 4.161 6 12500 4.111 7 14700 4,080 8 30000 3.782 9 35000 3.701 10 00 2.170 1J+7~~

TableJKIII< Solvolysis of 140-OBs In 0*10065 N NaOAc/HOAc at 95*0°*

~~Concentration of 3/fO-OBs: 0,0600 M: Titrant: HClO^/HOAc 0.020^3 N~~

~~Aliquot Time(sec) ml(HC10/j,/H0Ac) 1 0 4.5*J-0 2 2500 -^.351 3 5000 J+,190~~

~~4 7500 Jf.032 5 10000 3.893 6 15000 3.601 7 17500 3.5^0 8 co 2.051 148~~

~~TableJQV. Solvolysis of 139-OBs in 0.10065 N NaOAc/HOAc at 45,0°.~~

~~Concentration of 139-OBs: 0.0602 M Titrant: HClO^/HOAc . 0.02043 N~~

~~Aliquot Time(sec) ml(HClO^/HOAc)~~

~~1 0 4.340~~

~~2 1000 4.321 3 2000 4.200 4 4000 4.062 5 6000 3-970 6 8000 3-813 7 10000 3-701 8 12500 3 .6 0 0 9 25000 3-052 10 Oo 2.071 149~~

~~TableJCV. Solvolysis of 139-OBs in 0.10065 N NaOAc/HOAc at 55.0°.~~

~~Concentration of 139-OBs: 0.0602 M~~

~~Titrant: HClO^/HOAc O.OI989 N~~

~~-Aliquot Time(sec) ml (HClO/j/HOAc) 1 0 4.640~~

~~2 . 500 4.501 3 1000 4.303 4 2000 4.100 5 4000 3.671 6 6000 3.281 7 7000 3.130 8 8000 3.042 9 10000 2 .8 6 0 10 Oo 2.140 150~~

~~Table^XVI. Solvolysis of 139^0Bs in 0.10065 N NaOAc/HOAc at 65*0°,~~

~~Concentration of 139,-OBs: 0.0602 M Titrant: HClO^/HOAc 0.02043 N~~

~~Aliquot Time(sec) ml (HClOj^/HOAc) 1 0 4.670 2 250 4.461 3 500 4.242 4 750 4.081 5 1000 3.872 6 1500 3 .6 1 0 7 2000 3.300 8 2500 3 .1 6 1 9 3000 2.992 10 0o 2 .1 9 0 151~~

~~^ble^XVII. Solvolysis of 89-OBs in 0.10065 N NaOAc/HOAc at 55.0°.~~

~~Concentration of 89-OBs: 0.0600 M Titrant: HC10j^/H0Ac 0.01989 N~~

~~Aliquot Time(sec) mKHClOj^/HOAc)~~

~~1 0 4.610 2 10000 4.530 3 26000 4.462 4 50000 4.353 5 90000 4.211 6 135000 4.031 7 180000 3.890 8 190000 3.782 9 200000 3.720 10 OQ 1.961 152~~

~~TableJCVIIl. Solvolysis of 89-OBs in O.IOO65 N~~

~~NaOAc/HOAc at 65.0°.~~

~~Concentration of 89-OBs: 0.0619 M Titrant: HClO^/HOAc 0.01989 N~~

~~Aliquot Time(sec) ml (HClOj^/HOAc)~~

~~1 0 4 ,660 2 5000 ^.592 3 15000 4.362 4 35000 4.091 5 85000 3.590 6 100000 3.411 7 110000 3.362 8 130000 3.213 9 180000 2.860 10 00 2.020 153~~

Table^XIX. Solvolysis of 89-OBs in 0*10065 N NaOAc/HOAc at 75.0°*

~~Concentration of 89-OBs: 010600 M Titrant: HClO^/HOAc 0,01989 N~~

~~Aliquot Time(sec) mlCHClO^/HOAc) 1 0 4.510~~

~~2 2500 4.442 3 5000 4.291 4 10000 4.090 5 15000 3.902 6 25000 3.571 7 30000 3.450 8 75000 2.543 9 200000 1.970 10 00 1.890 REFERENCES~~

(a) B. Capon, Quart, Rev. (London), 18, 45 (1964); (b) A, Streitweiser, Jr., Chem. Rev.fjijS. 675 (1956), (a) S, Winstein, E. Allred, R, Heck, and R, Glick, Tetrahedron, 3. 1 (1958); (b) E. L. Allred and S, Winstein, J, Amer, Chem. Soc., 8 9, 3991. 3998, 4008, 4012 (1967); (c) E. R. Novak and 0. S. Tarbell, Ibid., 8£, 7 3, 3086 (1967). (a) H. W. Heine, A. D. Miller, W. H. Barton, and . R. W. Greiner, ibid., 75, 4778 (1953); (b) H. W. Heine, ibid,. 6268 (1957); (c) L, A, Paquette and R. V/. Begland, ibid., $0,, 5159 (1968). (a) J. F. Lane and H. W. Heine, ibid., 13/^8 (1951); (b) H. W. Heine, E. Becker, and J. F, Lane, ibid., 75, 4514 (1953); (c) B, Capon, M, C. Smith, E. Anderson, R. H. Dahm, and G, H. Sankey, J. Chem. Soc. B , IO38 . (1969 ); (d) B, M. Dunn and T. C. Bruice, J. Amer. Chem. Soc., 92, 6589 (1970). ^ (a) T. C, Bruice and S. J. Henkovic, ibid. , 85, 1 (1963 ); (b) H. Freundlich and K. Kroepelln, &. Physik. Chem., 122,, 39 (1926); (c) G. Salomon, Helv. Chim. Acta., igT 1361 (1933). (a) K. D. Gunderman, Angevf. Chem. Int. Ed., 599 (1963 ); (b) R, E. Ireland and H. A. Smith, Chem. Ind. (London), 1252 (1959); (©) L. A. Paquette, G. V. Meehan, and L. D. Wise, J. Amer. Chem. Soc., gJL 3231

(a) T. C. Bruice and J. M. Sturtevant, ibid., 81, 2860 (1959); (b) M. L. Bender, Chem. Rev.. 6 0 ^ 7 8 (i9 6 0 ); (c) G. L. Schmir and T. C. Bruice, If. Amer. Chem. Soc., 8 5. 1173 (1958); (d) T. C. BruiceT ibid., 8 1, 5 ^ 4 (1959). --- (a) E. D. Hughes, C. K, Ingold, and J. B. Rose, J. Chem. Soc#« 3839 (1953); (b) S. Winstein, et al., J. Amer. Chem. Soc,. Z*t, 1127 (1952); (c) I. L. Reich, A. Diaz, and S. V/instein, ibid., SU>» *>635, 5637* (1969); (d) P. C. Myhro and M. S. Brovm, ibid., gl, 5639 (19o9); (e) P. C. Myhro and E. Evans, ibid., §£. 5641 (1969). (a) D. J. Cram, ibid., 71, 3863, 3875 (1949); (b) D. J, Cram and L. A^Singer, ibid., 85. 1075 (1963); (c) D. J. Cram, ibid., 8j§7376? (196*0; (d) D. J, Cram and J. A, Thompson? ibid., 8 9, 6766 (1967). (a) S. Winstein, Bull. Soc., Chlm. Fr.. 55 (1951); (b) S. Winstein, C. R. Lindegren, H. Marshall, and L. L. Ingraham, J. Amer. Chem. Soo.« 7$, 1*1-7 (1955); (c) S. Winstein and R. Heck, ibid., ££[, 4801 (1956); (d) E. F. Jenny and S. V/instein, Helv. Chlm. Acta., 41,807 (1958); (e) S. Winstein and R. Baker, J. Amer. Chem. Soc., 86, 2071 (1964); (f) L. Eberson, J. P. Petrovich, R?^Balrd, D. Dyches, and S. Winstein, ibid., 87, 3504 (1965); (g) A. Diaz, I. Lazdins, and §. Winstein, ibid., go, 6546 (1968). (a) H. C. Brown, K, J. Morgan, and F. J. Chloripek, ibid.« 82, 2137 (1965); (b) H. C. Brown, R. Bern- heimer, C. J. Kim, and S. E. Scheppale, ibid., 85,, 370 (I9 67); (c) H. C. Brown and C. J. Kim, ibid., 90, 2082 (1968). (a) G. D. Sargent, R. L. Taylor and W. H. Demisch, Tetrahedron Lett.. 2275 (1968); (b) W. G. Dauben and J. L. Chitwood, J. Amer. Chem, Soc., 90, 3835 (1968); (c) R. M. Coates and J. L. Kirkpatrick, ibid., 90, 4162 (1968); (d) H. Tanlda, Accounts Chem. Res., f, 239 (1968); (e) M. J. S. Dewar and J. M. Harresj J. Amer. Chem. Soc.. 90, 4468, (1968); (f) Y. E. Rhodes and T. Takino, ibid.,^90, 4469 (1968); (g) R. Muneyuki, T. Yano, and H. Tanlda?"ibid.. 91, 2408 (1969 ); (h) S. V/instein, Quart. Rev.~(T6ndonT, 23. 141 (19o9); (i) G. N. Flkes, J. Org. Chem., ^4,^513 (1969); (j) J. A. Berson, D. Wegl, G. M. Clarke, and R. G. Bergman, J. Amer. Chem. Soc., £0, 3240 (1968); (k) J. A. Barson, R. G. Bergman? G. M. Clarke, and D. Wege, ibid., 91, 5601 (1969 ); (1) P. K. Freeman and J. N, Blagevich, Chem. Commun., 1357 (1969); (m) J. S, Haywood-Farmer and R. E. Plncock, J. Amer. Chem. Soc.. 91, 3020 (1969 ). 156 13. R. H. DeWolfe and W. C. Young, Chem. Rev,, 56, 788 (1956). ~ 14. S. Winstein, E, Grunwald, and H. W, Jones, J. Amer. Chem. Soc.,' i ' 73. 2700 (1951). 15. A. G. Catchpole and E. D. Hughes, J. Chem. Soc., 4, (1948). 16. W. G. Young and L, J. Andrews, J. Amer. Chem. Soc., 6 6, 421 (1944). 17. J. D, Roberts, W. G. Young, and S. Winstein, ibid,, 64, 2157 (1942). 18. E. P. Kiefer and J. D. Roberts, ibid., 0%, 784 (1962). 19. A. C. Cope, C. H. Park, and D. Scheiner, ibid., 84, 4862 (1962 ). ^ 20. A. C. Cope and P. E. Peterson, ibid., 81, 1643 (1959). v y v 21. E. A. Braude and W. F. Forbes, J. Chem. Soc., 1755 (1951). 22. E. A. Braude and J, A. Coles, ibid., 2014 (1950).

2 3 . E. A. Braude, W. F. Forbes, and E. A. Evans, ibid., 2202 (1953). 24. E. A. Braude, W. F. Forbes, B. P. Gofton, R. P. Houghton, E. S, Waight, ibid., 4?11 (1957). 25. (a) N. J. Demjanov, Chem. Ber., ^Q, 4393, 4961 (1907); (b) P. Bruylants and A. Dewael, Bull. Cl. Sci. Acad. Ray. Belg... 14, 140 (1928); Chem. Zentralbl., 99., "2708 (1928). ~ ' 26. Reviews published in the last decade include R. Brewlow in "Molecular Rearrangements", Part 1, Interscience, New York, N.Y., 1965* pp. 233-294; B. Capon, Quart. Rev. (London). 1§, 45 (1954); G, D. Sargent, ibid., 20, 301 (1966); M. Hanack and H-J Schneider, Angew. gtiem. Int. Ed. Engl., j5, 666 (1 9 6 7); K. B. Wiberg, B. A, Hess, Jr., and A T J. Ashe III in "Carbonium Ions", Vol. Ill, Interscience, New York, N.Y., 1972. 27. P. Bethell and V. Gold, Quart. Rev., (London), 12, 173 (1958). 157 28. (a) M. Simonetta and S. Winstein, J. Amer, Chem. Soc., 76* 18 (1954); (b) S. Winstein, et al., ibid,, 70, 8 3 8, 3528 (1948); (c) S. Winstein, et al,, ibid., 73,9 4347 , 4354 (1956); (d) S. Winstein, et aITT ibid.. ol, 4399 (1959); (e) G. H. Whitham, Proc. cKem. Soc., 422 (1961). 29. (a) S. Winstein, M. Shatavsky, C. Norton, and R. B. Woodward, J. Amer. Chem. Soc., 27* 4183 (1955); (b) H. C. Brovin and H. M. Bell, ibid., 8 3, 2324 (1 963 ); (c) S, Winstein, A. H. Lewin, and R. C, Bande, ibid.. 8g, 2324 (1 96 3 ). 30. (a) S. Winstein and M. Shatavsky, ibid. , £8,, 592 (1956); (b) S, Winstein and E. T. Stafford, ibid., Z9,, 505 (1957). 31. S. Winstein and C. Ordronneau, ibid., 82, 2084 (i9 6 0 ), 32. P. R. Story and M. Saunders, ibid., 84, 4876 (196 2 ); P. R, Story, C. C. Snyder, D. C. Douglass, E. W. Anderson, and R. L. Kornegay, ibid., 8£, 3630 (1 96 3 ). 33. R. H, Mazur,. W. N. White, D. A. Semenow, C. C. Lee, M. S. Silver, and J. D, Roberts, ibid., 81, 4390 (1959). ~ v 34. E. Renlc and J. D. Roberts, ibid., 8J, 878 (1961). 35. M. C. Caserio, W, H. Graham, and J. D, Roberts, Tetrahedron. 13,, 1?1 (i9 6 0 ). 36. P. v. R. Schleyer and G. W. Van Dine, J. Amer. Chem. Soc.. 8g, 2321 (196 6); cf, however, D. D. Roberts, J. Org. Chem., 2£, 294 (1964); 30, 23 (1965); 31, 2000 (1966). ^ 37. M. Vogel and J. D, Roberts, J. Amer. Chem. Soc., 88, 2262 (1966). 38. M. Hanack and H-J. Schneider. Tetrahedron, 20, I863 (1964). ------39. M. Hanack and W. Keberle, Chem. Ber., 86, 2937 (1 9 6 3 ). 40. V/. J, Hehre and P. C. Hiberty, J. Amer. Chem, Soc., 9ft. 5917 (1972). 41. M. Hanack and H-J. Schneider, Angew. Chem., 77* 1023 (1965); Angew. Chem. Int. Ed., 4., 976 (19657/ 158 42. C. Bergstrom and S. Siegel, J. Amer. Chem, Soo,, 7A» 145 (1952). 43. M, Hanack and G, Gbcker, Chem. Ber,, 96, 2121 (1963 ). 44* K. L. Servis and J. D. Roberts, J. Amer, Chem, Soc., VV*86, 3773 (1964). 45. (a) H-J. Schneider, Dissertation, Universitat Lubingen, 1967; (b) P. D, Bartlett and M. C, Rece, J. Org. Chem., 28, 3351 (1963 ). 46. M. Hanack, Angew. Chem., 77, 624 (1965)? Angew. Chem. Int. Ed., &, 603 (1965). 47. M, Hanack and W. Keberle, Chem. Ber., §6, 2937 (1963). 48. A. C. Cope, M. M. Martin, and M. A. McKervey, Quart. 1Rev. " “ (London), A 20, V 119 (1966). 49. R. Breslow, J. Lockhart, and A, Small, J. Amer. Chem. Soc., 84, 2793 (1962). 50. M. Hanack and M. Riedlinger, Chem. Ber., 100, 2107 (1967). 51. H. Schneider-Bernlohr, H-J. Schneider, and M, Hanack, Tetrahedron Lett.. 1425 (196?), 52. M. Hanack and H-J. Schneider, Liebigs Ann. Chem., 686, 8 (1965) 53. w. D. Closson and G. T. Kwlatkowski, Tetrahedron Lett., 3831 (1964); Tetrahedron. 21, 2779 (1965). 54. K. B, Wiberg and J. E. Hiatt, J. Amer. Chem. Soc., 20,v w 6495 (1968). 55. S. V/instein, Quart. Rev. (London), 2^, 141 (1969 )* 56. (a) J. L. Rosengerg, J. E. Mahler, and R. Pettit, J. Amer. Chem. Soc.. 8^, 2842 (1962 ); (b) C. E. Keller and R. Pettit, ibid., 88, 604 (1966): (c) J. D. Holmes and R. Pettit, Ibid., g§, 2531 (1963). 57. S. V/instein, H, D, Kaesz, C. G. Krieter, and E, C, Friedrich, ibid., 8£, 3267 (1965). 58. A. Davison, W, McFarlane, L. Pratt, and G. Wilkinson, J. Chem, Soc.. 4821 (1962 ). 159 59. M. Brookhart and E. R. Davis, J. Araer. Chem, Soc,, 9 2 , ?622 (197 0 ). 60. S. Winstein, C. G. Krieter and J. I. Brauman, ibid., 8 8, 20^7 (1 9 6 6). 61. G. Boche, V/. Hechtl, H. Huber, and R. Huisgen, ibid., §2. 334-4-, 33^5 (1967). 6 2 . N. G. Deno, et al., ibid., 8£, 2153 (1965), 6 3. H. J. Dauben, J. D, Wilson, and J. L. Larty, ibid,, £0 , 811 (1 96 8). 6*f. M. Brookhart, M. Ogliaruso, and S, Winstein, ibid., 8g, 1965 (1967). 6 5. H. J. Dauben, J. Larty, and S. Winstein, unpublished work. 6 6. L. A. Paquette and T. J. Barton, J. Amer. Chem. Soc., 8g f 5480 (1 9 6 7); 2 1 , 4714 (1 9 6 9 )":: 6 7. A. G. Cope and D. F. Rugen, ibid,, 75, 3215 (1953). 68. A. P. Krancho, Org. Syn., £1* 20 (1967), 6 9 . A. S. Dreiding and J, A. Hartman, J. Amer. Chem. Soc., 75, 939 (1955). 70. R. H. DeWolfe and W. G. Young, Chem. Rev., 5 6, 786 (1956). ------~ 71. J. Colonge, G. Descotes, and M. Fournier, Comp. Rend., 258, 4-083 (1964-). 72. S. Winstein, Quart. Rev. (London), 23 ., 1^-7 (1969 ); T.. J. Katz and E. H. Gold, J. Amer. Chem. Soc., 86, 1600 (1964-). • 73. (a) R. Huisgen and V/. E. Kenz, ibid. , 82, 4-102 (1970) (b) R. Huisgen and W. E. Kenz, ibid., 9 2 , 4-104- (1970). 7^. R, Crlegee, C, Schvreickhardt, and H. Knoche, Chem. Ber., 103,v W v 960 (1970). 75. H. Huisgen, F. Mletzsch, G. Boche, and H. Seidl, Chem. Soc., Spec. Publ.. 19, 3 (1965). 7 6. G, D. Sargent, N. Lowry, and S, D. Reich, J. Amer. Chem. Soc., 8 9, 5985 (1 96 7). 160 77. J. C. Martin and B. H. Ree, ibid., 92, 1660 (1969). 78. V. Buss, R. Gleiter, and B. v. R. Schleyer, ibid., 93, 3927 (1971). 79. A. C, Cope, M. Burg, and S. W. Fenton, ibid., 74, 173 (1952). 80. R. L. Shriner, Org. React.. Jl, 20 (1942). 81. E. H. Amonoo-Neizer, et al., J. Chem. Soo., 2997 (1965).

82. M. W. Rathke,9 ■J. Amer. — ■- IChem. ■ Soo.,■■ ' W92, V 3222 (1970). 8 3. M. Hanack and H-J. Schneider, Angew. Chem. Int. Ed., 6, 666 (1967). 84. M. Hanack, et. al., Liebigs Ann. Chem., 717* 4l (1968). ------85. E. LeGoff, J. Org. Chem., 29, 2048 (1964). 1■ ■1 1 86. W. G. Dauben and G, H, Berezin, J. Amer. Chem. Soc.. W85, v 468 (1963), 8 7. J. C, Stowell, unpublished results. 88. C. D. Gutsche, I. Y. C. Tao, and J. Kozma, J. Org, Chem,, 32, 1782 (1967). 89. M. J, Karplus, J. Amer. Chem. Soc,, 88, 2870 (1963 ).

90. C. C. Hinkley, ibid.,———— 91, 5160 (1969 ). 91. P. V. Demarco, et al., ibid., 92, 5734 (1970). 92. G. Markl, Chem. Ber.. §4, 3005 (1961). 93. F. A, L. Anet and L. A. Rock, J. Amer. Chem. Soc., go,, 7130 (1968);. F. A, L. Anet and B. Gregorovich, Tetrahedron Lett.. 5961 (1966).

94. L. A. Carpino, J. Org.. j Chem., 35,vvv 3971 (1970 ), 95. E. LeGoff and R. B. LaCount, Tetrahedron Lett., 2787, (1965). ------.161 96. 1,4-Dimethy1COT was subsequently found to be con taminated with a significant amount of the 1,5- isomer. 97. J. Clardy, personal communication. 98. M. N. Galbraith, et. al., Chem. Commun., 466 (1968). 99. A. C. Cope and P. Burton, J. Amer., Chem. Soc., 82, 5439 (I9 6 0 ). — 100. G.. Wittig and M. Schlosser, Chem, Ber., 94, 1373 (1961), ~