Citation
Radical cyclizations involved in cyclopolymerizations

Material Information

Title:
Radical cyclizations involved in cyclopolymerizations
Creator:
Smith, Thomas William, 1944- ( Dissertant )
Butler, George B. ( Thesis advisor )
Battiste, Merle A. ( Reviewer )
Brown, Henry C. ( Reviewer )
Deyrup, James A. ( Reviewer )
Vala, Martin H. ( Reviewer )
Place of Publication:
Gainesville, Fla.
Publisher:
University of Florida
Publication Date:
Copyright Date:
1972
Language:
English
Physical Description:
viii, 104 leaves. : ill. ; 28 cm.

Subjects

Subjects / Keywords:
Absorption spectra ( jstor )
Chlorides ( jstor )
Ethers ( jstor )
Flasks ( jstor )
Hydrides ( jstor )
Infrared spectrum ( jstor )
Liquids ( jstor )
Reactive oxygen species ( jstor )
Sodium ( jstor )
Vinyl radical ( jstor )
Chemistry thesis Ph. D
Dissertations, Academic -- Chemistry -- UF
Polymers and polymerization ( lcsh )
Ring formation (Chemistry) ( lcsh )
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Abstract:
The objective of this research project was to investigate the mechanism involved in the cyclization reactions of radicals such as the initial radicals involved in the cyclopolymerization of 1,6-dienes. The approach taken was to observe the products obtained from the reactions of tributyltin hydride with 5 ,6-unsaturated hexyl radicals substituted at the 5-position. By generating the radicals in the presence of different tributyltin hydride concentrations and running the reactions at different temperatures, the kinetics of the competing radical reactions were studied. The product distributions resulting from the reactions of the 2- methallyloxyethyl radical and the 2-(2-phenylallyloxy) ethyl radical, generated from the corresponding bromides, with tributyltin hydride were determined. The products obtained were synthesized independently so that the product concentrations in the reaction mixtures could be determined by gas chromatography by comparison with standard solutions of products. The 2-allyloxyethyl radical reacts with tributyltin hydride to give 3-methyltetrahydrofuran as the only cyclic product. The 2- raethallyloxyethyl radical cyclized to give up to 3 percent 3-methyltetrahydropyran in addition to the 3 5 3-dimethyltetrahydrofuran and ethyl raethallyl ether. 3-Phenyltetrahydropyran was the predominant cyclic product from the 2-(2-phenylallyloxy)ethyl radical. These observations support a mechanism involving classical radical cyclizations rather than the cyclic products arising from a concerted attack by the hydrogen donor on the uncyclized radical, perhaps in the form of an internal radical-TT complex. The kinetic treatment for a reaction scheme involving a competition between irreversible cyclization reactions and abstraction of hydrogen by the initially formed 2-methallyloxyethyl or 2-(2-phenylallyloxy)ethyl radical satisfactorily fit the data. The reversibility of the cyclization reactions was checked by synthesizing 3-phenyl-3-bromomethyltetrahydrofuran and generating the corresponding radical in the presence of tributyltin hydride. 3-Phenyl-3-methy] tetrahydrofuran was the only product, indicating that the radical cyclizations of the 2-(2-phenylallyloxy) ethyl radical and probably also the 2-methallyloxyethyl radical are irreversible. Thus the kinetic treatment is also in agreement with a mechanism involving a simple competition between irreversible cyclization reactions and abstraction of hydrogen from tributyltin hydride by the acyclic radical. The data do not, however, eliminate the possibility of some type of complexed species being formed at some point during the course of the reaction. Additional information about the rate constants and activation energies for the competing reactions was obtained from the kinetic data. It appears the driving force for cyclopolymerization is simply a competition between the reaction of the initial radical with the other double bond of the same molecule or with another molecule of monomer, which strongly favors formation of the cyclic radical. The structures of the cyclic repeating units depend on steric and electronic factors and, in cases in which the initial radical is quite stable and reversibility of the radical cyclizations is possible, on the relative stabilities of the cyclic radicals.
Thesis:
Thesis -- University of Florida.
Bibliography:
Bibliography: leaves 100-103.
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
022665034 ( AlephBibNum )
13978010 ( OCLC )
ADA5106 ( NOTIS )

Downloads

This item has the following downloads:


Full Text

PAGE 1

Radical Cyclizations Involved in Cyclopolymerizations By THOMAS WILLL\:-I SMITH A DISSERTATION PRESENTED TO THE GR^^DUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REOUIRBfFNTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1972

PAGE 2

iiil

PAGE 3

ACKNOWLEDGEMENTS The help of Dr. G. B. Butler, whose guidance, patience, and encouragement were invaluable during the execution of this research project, is acknowledged with deep appreciation and gratitude. A special thanks is also extended to the other members of my supervisory committee for giving their valuable time and helpful advice. I also wish to thank my fellow workers in the laboratory for their friendship and their helpful sharing of ideas. The financial assistance received from a National Aeronautics and Space Administration Traineeship and a University of Florida Graduate School Fellowship is gratefully acknowledged.. A special expression of gratitude goes to my wife, Linda, whose love and understanding were of great assistance to the completion of this project. Also in line for special thanks are my family and Linda's family who provided a great deal of support and encouragement.

PAGE 4

TABLE OF CONTENTS Page Acknowledgements ii List of Tables iv List of Figures v Abstract vi Chapter 1 I INTRODUCTION 1 A. History of Cyclopol5mierizatlon 1 B. Structures of Cyclopol^Tners 3 C. Radical Cyclization Reactions 5 D. Mechanism of Cyclopoly:7ierization 7 E. Statement of the Problem 14 II RESULTS AND DISCUSSION 17 A. Synthesis of Reactants and Products .... 17 B. Generation and Reactions of Radicals .... 29 C. Kinetics of the Radical Reactions 36 D. Discussion • • 51 III EXPERBIENTAL 58 A. Equipment and Data 58 B. Syntheses 58 C. Preparation and Analysis of Reaction Solutions . 96 Bibliography 100 Biographical Sketch 10'^

PAGE 5

LIST OF TABLES Table Page I Reaction Solutions for Generation and Reactions of 2-Methallyloxyethyl Radical 32 II Product Concentrations from Reactions of the 2Methallyloxyethyl Radical 32 III Product Percentages from Reactions of the 2-^ethallyloxyethyl Radical 33 IV Ratios of 3,3-Diinethyltetrahydrofuran to 3-Methyltetrahydropyran Produced 33 V Reaction Solutions for Generation and Reactions of the 2-(2-Phenylallyloxy)ethyl Radical 34 VI Product Concentrations from Reactions of the 2(2-Phenylallyloxy) ethyl Radical 34 VII Product Percentages from Reactions of the 2-(2-Phenylallyloxy) ethyl Radical 35 VIII Ratios of 3-Phenyltetrahydropyran to 3-Phenyl-3-methyltetrahydrofuran Produced 35 IX Values of k /k or k /k„ lA^hich Best Fit the Data According to Equation 4 40 X Composite of Rate Constant Ratios 42 XI Activation Energy Differences and Frequency Factor Ratios from Figures 1-6 49

PAGE 6

LIST OF FIGURES Figure ZSS^. 1 Ln k-/k Versus 1/T for the 2-Methallylox5'ethyl Radical 43 2 Ln k^/k» Versus 1/T for the 2-Methallyloxyethyl Radical 44 3 Ln k /k Versus 1/T for the 2-Methallyloxyethyl Radical 45 4 Ln k /k Versus 1/T for the 2-(2-Phenylallyloxy)'ethyl Radical 46 5 Ln k /k^ Versus 1/T for the 2-(2-Phenylallyloxy)ethyl Radical 47 6 Ln k /k Versus 1/T for the 2-(2-Phenylallyloxy)ethyl Radical 48

PAGE 7

Abstract of Dissertation Presented to the Graduate Council of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy RADICAL CYCLIZATIONS INVOLVED IN CYCLOPOLY^IERIZATIONS By Thomas William Smith June, 1972 Chairman: Dr. G. B. Butler Major Department: Chemistry The objective of this research project was to investigate the mechanism involved in the cyclization reactions of radicals such as the initial radicals involved in the cyclopolymerization of 1,6-dienes. The approach taken was to observe the products obtained from the reactions of tributyltin hydride with 5 ,6-unsaturated hexyl radicals substituted at the 5-position. By generating the radicals in the presence of different tributyltin hydride concentrations and running the reactions at different temperatures, the kinetics of the competing radical reactions were studied. The product distributions resulting from the reactions of the 2methallyloxyethyl radical and the 2-(2-phenylallyloxy) ethyl radical, generated from the corresponding bromides, with tributyltin hydride were determined. Tlie products obtained were synthesized independently so that the product concentrations in the reaction mixtures could be determined by gas chromatography by comparison with standard solutions of products. The 2-allyloxyethyl radical reacts with tributyltin hydride

PAGE 8

to give 3-methyltetrahydrofuran as the only cyclic product. The 2raethallyloxyethyl radical cyclized to give up to 3 percent 3-methyltetrahydropyran in addition to the 3 5 3-dimethyltetrahydrofuran and ethyl raethallyl ether. 3-Phenyltetrahydropyran was the predominant cyclic product from the 2-(2-phenylallyloxy)ethyl radical. These observations support a mechanism involving classical radical cycli?,ations rather than the cyclic products arising from a concerted attack by the hydrogen donor on the uncyclized radical, perhaps in the form of an internal radical-TT complex. The kinetic treatment for a reaction scheme involving a competition between irreversible cyclization reactions and abstraction of hydrogen by the initially formed 2-methallyloxyethyl or 2-(2-phenylallyloxy)ethyl radical satisfactorily fit the data. The reversibility of the cyclization reactions was checked by synthesizing 3-phenyl-3-bromomethyltetrahydrofuran and generating the corresponding radical in the presence of tributyltin hydride. 3-Phenyl-3-methy] tetrahydrofuran was the only product, indicating that the radical cyclizations of the 2-(2-phenylallyloxy) ethyl radical and probably also the 2-methallyloxyethyl radical are irreversible. Thus the kinetic treatment is also in agreement with a mechanism involving a simple competition between irreversible cyclization reactions and abstraction of hydrogen from tributyltin hydride by tlie acyclic radical. The data do not, however, eliminate the possibility of some type of complexed species being formed at seme point during the course of the reaction. Additional information about the rate constants and activation energies for the competing reactions was obtained from the kinetic data. It appears the driving force for cyclopolymerization is simply

PAGE 9

a competition between the reaction of the initial radical with the other double bond of the same molecule or with another molecule of monomer, which strongly favors formation of the cyclic radical. The structures of the cyclic repeating units depend on steric and electronic factors and, in cases in which the initial radical is quite stable and reversibility of the radical cyclizations is possible, on the relative stabilities of the cyclic radicals.

PAGE 10

Chapter I INTRODUCT ION A. Hist ory of Cyclopolymerization The discovery by Butler and Ingley (1) that certain nonconjugated dienes polymerize to give saturated, soluble polymers was in direct contrast to the earlier belief that polymerization of such niononiers must lead to cross-linked polymers. They found that certain diallylammonium salts polymerized in aqueous solution using t-butylhydroperoxide as catalyst to give water-soluble polymers. Under the same conditions triand tetraallylammonium salts yielded cross-linked polymers while the monoallylammonium salts failed to polymerize. To explain the unusual polymers obtained from the diallylaramonium salts, Butler and Angelo (2) proposed an alternating intramolecularintermolecular chain propagation mechanism to form plpcridine rings in the chain. This type of process is now commonly termed cyclopol\nnerization. (I) (I) (a 4 1)

PAGE 11

The initiating radical attacks one double bond of the diene. The radical thus produced attacks the other double bond of the molecule, forming a cyclic radical (I) . This radical attacks another molecule of the diene, and the process repeats itself to form a saturated polymer containing cyclic repeating units. The fact that cyclic structures were indeed present in these polymers was substantiated by the work of Butler, Crawshaw, and Miller (3). They converted poly(diallylammonium bromide) (II) to the benzamide (III) , which was oxidized with potassium permanganate. Tlie structure of the resulting polymer (IV) was confirmed by elemental analysis, titration, and the infrared spectrum. PoljTner (IV) decomposed upon heating to give benzoic acid and a cross-linked poljTiier.. NaOH C,H^C0C1 6 5 Br" (II) (HI) (IV) These workers also degraded poly (diallyldimethylammonium bromide) (V) Dy means of Hofmann exhaustive methylation to obtain trimethylamine and a cross-linked pol^Tuer. These degradative studies conclusively demonstrated the presence of cyclic structures in the pol^Tner chains, although they did not unequivocably establish the ring size in the repeating units. The postulated piperidine rings or pyrrolidine rings or a mixture of the two could have given the observed results.

PAGE 12

CH. CH, J Br" 2) heat 1) CH^I 2) to hydroxide 3) heat N^ NCCH^)^ cross-linked polymer Marvel and Vest (4) found evidence for the presence of six-membered rings in the polymers formed from the free-radical polymerization of l,6~heptadiene-2,6-dicarboxylic acid and its ethyl ester. I-Jhen these two polymers xrere partially dchydrogenated with potassium perchlorate, the infrared and ultraviolet spectra indicated the presence of aromatic rings in the dehydrogenated material. Since the initial reports on cyclopolymerization a large number of nonconjugated dienes have been cyclopolymerized using all of the known types of initiators. Several good review articles (5,6,7) cover the scope of the work that has been done. B. Structures of Cyclopolymers A number of methods have been employed to determine the structures of cyclopol^miers . The techniques which can be utilized depend on the particular monomers used. The degradative methods of Butler and coworkers mentioned above demonstrated the presence of either piperidine

PAGE 13

or pyrrolidine rings in the cyclopolymers from diallylammonium salts. The presence of aromatic rings in the dehydrogenated cyclopolymers of Marvel and Vest specifically indicated the presence of some sixmembered rings in the cyclopol^aners of l,6-heptadiene-2 ,6-dicarboxylic acid and its ethyl ester. Several workers have studied the poljnners obtained from N-substituted dimethacrylamides. Goetzen and Schroeder (8) isolated ethylene and trim-ethylmaleimide from the pyrolysis of poly (N-methyldimethacrylamide) . Koton and coworkers (9) have shown that N-ethyl , propyl, and phenyl derivatives give similar products. A-o ~^ 0>N»A< heat HC^H^ Sokolova and Rudkovskaya (10) polymerized the same four N-substituted dimethacrylamides radically and also found that the cyclic units were composed of succinimide rings. They determined the structures of the cyclic units in their polymers by comparing the carbonyl absorptions in the infrared spectra of the polymers with those of similar buccinimides and glutarimides and also with those of polymers which contain only glutarimide rings, obtained by partial deamination of poly (methacrylamides). They found that poly (dimethacrylamide) is composed of both fiveand six-membered rings, with the latter predominating. Butler and Meyers (11) used both infrared and Nlfil spectroscopy to analyze the polymers obtained from dimethacrylamide and its N-methyl and N-phenyl derivatives. Their results agreed with the findings of

PAGE 14

the other authors, with the exception that the NMR spectrum of the Nmethyl protons of poly (n-methyldimethacrylamide) showed that a small percentage of six-membered rings were present in addition to the fivemembered rings, Sultanov and Arbuzova (12) found that both fiveand six-membered rings are formed in the polymerization of divinyl acetal. They hydrolyzed the polymer and analyzed the resulting 1,2and 1,3-glycol units derived from the fiveand six-membered rings respectively. The vicinal hydroxyl groups were oxidized with periodic acid, and the ratio of sixto f ive-membered rings was found to be 77:23. C. Radical Cyclization Reactions The amount of information which can be established concerning the microstructures of these cyclopolymers is usually quite limited and depends to a large extent upon the nature of the monomer. Most of the cyclopolymers which have been prepared do not offer such distinguishing features to serve as a probe of the cyclic structures as those discussed in the preceding section. Another approach to the study of the cyclic units formed in the radical cyclization reactions of cyclopolymerizations is through the study of the radical cyclization reactions of selected model compounds. In 1958 Friedlander (13) reported that free-radical catalyzed addition of various active chain transfer agents to 1 ,6-diolef ins provides a new synthetic tool for preparing six-membered heterocyclic and carbocyclic compounds. This was in agreement with the earlier polymerization studies of Butler (2,3) and Marvel (.4) where the polymers were postulated to have six-membered rings as the cyclic repeating units.

PAGE 15

Lamb, Ayers, and Toney (14) in 1963 made the surprising discovery that 6-heptenoyl peroxide decomposes to give methylcyclopentane as the major product. Only a very small amount of cyclohexane was formed. This result is hard to justify. Cyclohexane has been shown to be more stable than methylcyclopentane at the reaction temperature (15). Cyclohexaneformyl peroxide decomposes 34 times as fast as cyclopentaneacetyl peroxide (16), indicating that the cyclohexyl radical is more stable than the cyclopentylmethyl radical. Lamb, Ayers, and Toney (14) also decomposed these tv70 peroxides to determine that formation of the cyclic radicals from the 5-hexenyl radical are both irreversible processes. The irreversibility of the conversion of 5-hexenyl radical to cyclopentylmethyl radical was also demonstrated recently by the ESR studies of Kochi and Krusic (17). y -15° ^ < •^ This unexpected discovery by Lamb was just the first of many such radical cyclization reactions leading to f ive-membered ring products or mixtures of fiveand six-membered rings. Walling and Pearson (18) observed almost exclusive f ive-membered ring formation in the cyclization of 5-hexenyl mercaptan with triethyl phosphite. Increasing the reaction temperature increased the relative amount of cyclic products observed as well as the proportion of six-membered rings. SH + P(OEt). 3 60° 0% 120° 3%

PAGE 16

Brace (19) has reported the products obtained from the addition of active chain transfer agents to 1,6-dienes. In each case he reported the exclusive fornation of five-member ed ring cyclic products. His results contradicted the earlier results of Friedlander (13) and also cast a doubt concerning the reported cyclic repeating units of poly(diallylcyanamide) (20). Aso (21) did a similar study of the cyclization of diallyl ether and divinyl formal and observed f ive-membered rings also. He used infrared and IIMR spectroscopy to identify the products. A number of other radical cyclization reactions have been reported. Some of them will be mentioned later in connection with the mechanism of cyclopolymerization. D. Mechanism of Cyclopolymerization A satisfactory explanation for the fact that cyclization is favored over intermolecular propagation to the extent that many nonconjugated 1,6-dienes give totally saturated cyclopol>Tne.rs has not been agreed upon. Three mechanistic explanations have been proposed — two emphasizing electronic interactions and the other based on steric effects. Butler (22) proposed that an electronic interaction might exist between the double bonds of the 1,6-dienes which might provide an energetically favorable pathway for cyclization. He and his coworkers (23,24) have observed bathochromic shifts in the ultraviolet absorptions for some nonconjugated double-bond systems, which can be attributed to interspacial interactions of the type proposed to lead to cyclization.

PAGE 17

Further evidence for an electronic interaction was provided by the discovery by Butler and Meyers (11) that solid state polymerization of dimethacrylamide and the N-methyl and N-phenyl derivatives gave completely cyclized poljTuers. This was interpreted to mean that the 1,6-dienes were oriented in the crystals in a conformation favorable for cyclopolymerization. Baucom (25) recently in an attempt to correlate the extent of electronic interaction between the nonconjugated double bonds of some di-unsaturated esters and ethers and the tendency of these dienes to undergo cyclopolymerization was unable to observe any evidence for an electronic interaction in the ultraviolet, NMR, or infrared spectra of the dienes. A strong argument against an electronic interaction of this type leading to exclusive cyclic polymer lies in the fact that the activation energy for the cyclization step has been found to be always greater than or equal to the activation energy for the intermolecular propagation leading to formation of pendant double bonds (5). Butler (22) and Marvel and Stille (26) suggested that an interaction might exist between the initially formed radical and the other double bond of the molecule.

PAGE 18

This type of interaction has also been suggested by Lamb and coworkers (lA) to explain the formation of methylcyclopentane from the 5-hexenyl radical, explaining that it arises from attack of the hydrogen donor on the more easily approached position of the complex, Walling and coworkers (27) investigated the reaction of 6-bromo1-hexene with tributyltin hydride to generate the 5-hexenyl radical. Br + Bu SnH iBu Snll AIBN V Bu^Snlllk^ k^^^n^SnH Jb SnH tBu SnH They did a study of the reaction kiaetics, varying the tributyltin hydride concentration and running the reactions at 40° and 130°. One percent of cyclohexane was formed at 130° and a trace at 40°, thus k^ was not considered in the kinetics. Their data at 40° fit equation (1) with

PAGE 19

10 d[Bu Snll] k^ + k^lBu^Snll] d[Hexene] k [Bu„SnH] d[Bu SnH] k^ + (k^ + k JlBu^SnH] d[Hexene] k2[Bu2SnH] (1) (2) k /k^ = 10. The data at 130°, however, would not give a good fit for equation (1). Equation (2) would give a good fit for both sets of data. It includes k , , the rate constant for formation of methylcyclopentane by attack of tributyltin hydride on the uncyclized radical, perhaps in some form of complex — a "concerted cyclization" process. Struble, Beckwith, and Gream (28) studied the reactions of the 4-(l-cyclohexenyl)butyl radical, generated by reacting the bromide v/ith tributyltin hydride in the presence of a radical initiator. They found a constant ratio of spiro compound to decalins as well as a constant + Bu SnH J > i Bu SnH '4"; k^Pu SnH Bu SnH cis & trans

PAGE 20

11 cis-: trans-decalin ratio over a seven-fold change in hydride concentration. Both observations were suggested as evidence against a concerted cyclization and hydrogen transfer. The cis: trans-decalin ratio was similar to that previously obtained by Bartlett and covrorkers (29) from decalyl radical generated from the decomposition of both cisand trans-9-decalyl peresters. A concerted cyclization and hydrogen transfer, capable of leading to both spiro product and decalins, should lead preferentially to cisdecalin. Nozakura and coworkers (30,31) polymerized divinyloxydimethylsilane using both cationic and radical initiation. Radical polymerization gave polymers containing f ive-membered rings, six-membered rings, and pendant vinyl groups, whereas cationic polymerization yielded little or no cyclization. If a tricentric complexed species were important in leading to cyclopolymer, one would expect cationic initiation to yield a higher degree of cyclopolymerization than radical initiation. Field (32) polymerized 2 ,6-diphenyl-l,6-heptadiene by free radical, cationic, anionic, and Ziegler-type initiation. The polymers obtained in all cases were soluble and contained little or no residual unsaturation according to their infrared spectra. He found that the best method for obtaining both high molecular weight and high conversion was by anionic initiation. This observation causes doubt as to the necessity of a complexed species leading to cyclopolymers , since a tricentric species having four electrons is highly unfavorable. Once again the fact that the activation energy for cyclization is greater than or equal to the activation energy for intermolecular propagation leading to pendant double bonds contributes an argument against the complex as a mechanistic explanation for cyclopolymerization.

PAGE 21

12 The third mechanistic approach is a proposal by Gibbs and Barton (6) that steric factors alone bring about cyclopolymerization. They point out that bulky groups on the radical slow down interraolecular propagation greatly, as indicated by a comparison of the propagation rate constants for polymerization of methyl acrylate and butyl acrylate. The k for methyl acrylate is over 50 tiuies greater than the k for P P butyl acrylate. The fact that a double bond on the pendant group can react with the radical to form a conveniently sized cyclic structure, while interraolecular propagation is inhibited by the bulky pendant group, could explain the exclusive formation of cyclopolymers . This would predict a value of k /k^ greater than one but within a moderate range, since the intermolecular propagation is simply' competing unfavorably with the intramolecular reaction. The experimental values generally range from about 5 to 20 moles/liter. The formation of methylcyclopentane from the reaction of 5-hexenyl mercaptan with triethyl phosphite was explained by Walling and Pearson (18) as arising from the attack of the radical at the more accessible end of the double bond, the process being irreversible. Butler and Raymond (33) noted that for approach of the radical to the double bond with the p-orbitals in a common plane formation of fivemembered rings would be less sterically hindered. One of the terminal hydrogens lies in the nodal plane directly between the radical carbon and the carbon on the terminal end of the double bond, thus hindering six-membered ring formation. No such steric interference exists for approach of the radical to the other end of the double bond, leading to f ive-membered rings. Struble, Beckwith, and Gream (28) also offered a steric explanation

PAGE 22

13 for forraation of f ive-membered ring products, saying the general preference, for f ive-membered ring formation results from attack of the radical at the position on the double bond more readily approached vertically. Julia and coworkers (34,35,36) have also offered a steric explanation for the products they have obtained from cyclization of a number of substituted 5-hexenyl radicals. They found that placing stabilizing groups on the initial radical Increased the proportion of six-^membered rings formed. Placing a methyl group on either position of the double bond further increased the tendency of the stabilized radicals to yield six-membered ring products. A studj'^ of the product distribution from the cyclization of a stabilized radical at different temperatures showed that the proportion of six-membered ring products increased with increasing temperature. This was in agreement with the finding of Walling and Pearson (18), and indicates a higher activation energy for the formation of the cyclohexyl derivatives relative to the cyclopentyl derivatives. 65° 74% 26% 35° 61% 39% 22° 50% 50% -70° 20% 80% Julia (36) decided to check the reversibility of the radical cyclizations of this stabilized radical. He synthesized the two peresters (VI) and (VII) , which were decomposed to give the desired cyclic radicals. They found that the product ratios were similar from both peresters and agreed closely with the 16:84 ratio of products obtained from the open-chain radical. CN CO^Et

PAGE 23

14 ,C0 tBu 20% 80% \ — (-<;o Et (J Q CN .Et CO^Et 15% 85% Julia interpreted his data to mean that f ive-membered rings are the kinetically favored products, whereas six-membered rings can form in some cases when equilibration of the radicals is possible. E. Statement of the Problem The major objective of this research project was to synthesize and study the reactions of a series of compounds from which we could generate 5-substituted 5-hexenyl radicals. Xv X = CH^, C^H^, etc. J o 5

PAGE 24

15 We wanted to find out what effects the substituents at the 5-position would have on the ring size of the cyclic products from the reactions of the radicals. This information should help to define the mechanism and driving force for cyclization of these radicals and of analogous substituted radicals in cyclopolymerizations. If formation of f ive-membered rings arises from attack by the hydrogen donor on a complex at the more sterically favorable position, as suggested by Lamb, Ayers , and Toney (14), the 5-substituted 5-hexenyl radicals should show an even greater tendency to form f ive-membered ring products. If f ive-membered ring formation arises from attack of the radical at the more available end of the double bond sterically, the 5-substituent should cause a decrease in f ive-membered ring formation relative to six-membered ring products because the steric advantage for cyclopentane formation is decreased by the substituent at the 5position. Another consideration is that if cyclization involves classical radicals, placing a substituent at the 5-position means formation of a primary radical versus a tertiary or benzylic radical in going to cyclopentyl and cyclohexyl products respectively. Even if the cyclizations are irreversible processes, the lower energy of the substituted cyclohexyl radicals could cause a lowering of the activation energy for cyclohexyl product formation relative to the unsubstituted case and enable the cyclohexyl product formation to compete with formation of cyclopentyl derivatives. The initial work was aimed toward the synthesis of 6-substituted 6-heptenoyl peroxides, which could be decomposed to yield the 5-substituted 5-hexenyl radicals. UTicn this system presented a number of

PAGE 25

16 synthetic problems, we changed to the 2-substituted 2-allyloxyethyl radical system. By this time it had been shovTi by Walling and coworkers (27) and by Carlsson and Ingold (37) that generation of such radicals by reaction of the bromide with tributyltin hydride in the presence of a radical initiator provides a good method for studying the kinetics of the reactions of these radicals. Thus we set out to synthesize methallyl 2-bromoethyl ether and 2-phenylallyl 2-bromoethyl ether and the products which could arise from the generation of the corresponding radicals in the presence of tributyltin hydride. In this way we could study the kinetics of the reactions of the radicals by determining the product distributions at different tributyltin hydride concentrations and at different temperatures.

PAGE 26

Chapter II RESULTS AND DISCUSSION A. Synthesis of R.eactants and Products Attempted synthesis of 6-methyl--6-heptenoyl peroxide The preparation of 6-methyl-5-heptenoyl peroxide was to involve a series of reactions beginning with methallyl chloride. The envisaged reaction scheme is depicted below. CH3 CH^C-CH^Cl + Mg CH ether ' > CH^C-CH^MgCl .,A 2) NH.Cl > Xj 1) OH' 2) H+ 3) A ico^^^Vi <: NaOC^H^/C^H^OH PBr, C3H3N V CO^H SOCl, -> Xf !f2^ °2>2 The Grignard reagent of methallyl chloride was formed and reacted with ethylene oxide to yield 4-methyl-4-penten-l-ol. Many attempts V7ere made to convert this alcohol to the corresponding bromide or iodide. It is likely that a great many of the problems encountered 17

PAGE 27

18 could have been overcome by exploring more variations in reaction conditions. Reaction of 4-methyl-4-penten-l-ol with phosphorus tribroraide and pyridine, either neat or in benzene solution, yielded in every case a mixture of 5-bromo-2-methyl-l-pentene and 5-bromo-2-methyl-2-pentene, favoring the latter undesirable isomer. .OH + PBr, C3H3N ^Ve^ Br Reaction of the alcohol with half a mole of thionyl chloride in the presence of pyridine yielded the expected sulfite. \<7hen this sulfite was subjected to a mole of freshly prepared thionyl bromide plus pyridine, a black tar was formed. Repeating the reaction without the pyridine led to a mixture of 5-bromo-2-methyl-l-pentene and 5bromo-2-methyl-2-pentene, with the latter predominating once again. .OH C^H^N + 0.5 SOCl^ ^ ^ > 0-^2^0 SOBr^ C3H3N > tai SOBr, N^ Br Br The procedure of Landauer and Rydon (38) for converting alcohols to iodides was applied to 4-methyl-4-penten-l-ol. It was heated with equimolar portions of triphenyl phosphite and methyl iodide, yielding a mixture of 5-iodo-2-methyl-l-pentene and 5-iodo-2-methyl-2-pentene in a 1:5 ratio according to the integrated areas of the NIIR absorptions of the vinyl hydrogens.

PAGE 28

19 OH + P(0C,HJ, + CH_I D D J J A final attempt to prepare 5-lodo-2-inethyl-l-perLtene employed the procedure of Ansell (39) by way of the tosylate. 4-Methyl-4-pentenl-ol was added to a cooled slurry of p-toluenesulfonyl chloride in pyridine to produce 4-methyl-4-pentenyl p-toluenesulfonate. The tosylate was added to a solution of sodium iodide in acetone, and the liquid was refluxed several hours to yield once again a mixture of 5-iodo-2methyl-1-pentene and 5-iodo-2-methyl-2-pentene. Ansell (39) reported the preparation of 5-iodo-2-methyl-l-pentene by this procedure. He OH + TsCl C3H3N OTs Nal ine^ used elemental analysis and the infrared spectrum to establish the product's structure. The most obvious evidence, by far, for the presence of a mixture of the two isomers appears in the NMR absorptions of the vinyl hydrogens. The two vinyl protons of 5-iodo-2-methyl-lpcntene appeared as a singlet with hyper fine splitting at 5.28 T. The one vinyl proton of 5-iodo-2-methyl-2-pentene appeared as a triplet with hyperfine splitting centered at 4.89 T. The next step in the synthesis involved the reaction of 4-methyl-

PAGE 29

20 4-pentenyl p-toluenesulfonate with diethyl malonate and sodium ethoxide in ethanol. The product of this reaction, diethyl A-methyl4-pentenylmalonate, was obtained in good yield. OTs + CH^ (00202115)2 NaOC H ^^S"5>2 The diester was hydrolyzed by refluxing a mixture of the diester and a concentrated aqueous solution of potassium hydroxide for several hours until the mixture was homogenous. The big problem arose in attempting to neutralize and decarboxylate the diacid. Acidifying the solution to a pH of three to four with dilute hydrochloric acid and refluxing the solution resulted in a mixture of 6-methyl-6-heptenoic acid and 6-methyl-5-heptenoic acid. Neutralizing the solution with ammonium chloride to a pH of six and refluxing for a week produced a few drops of pure 6-methyl-6-heptenoic acid. ^°2S^)2 KDH H2O HCl. O2H A / NH CI CO2H O^H

PAGE 30

21 The conditions were adjusted a number of ways to try to get the diacid to decarboxylate in good yield without isomerizing, but all attempts were unsuccessful. We decided to abandon this system and attempt to prepare 3-methallyloxypropionyl peroxide, which upon decomposition would give the 2-methallyloxyethyl radical. Although allyl ethers are known to isomerize to vinyl ethers (40), the conditions are such that they could be easily avoided. The 2-methallyloxyethyl radical system Our first plan for generating the 2-methallyloxyethyl radical was by decomposing 3-methallyloxypropionyl peroxide. C02>2 A This peroxide was to be prepared by the following reaction sequence: CH2=C-CH20H + CH2=CH-CN ^J^^^ HCl °2^2 Na202 OCl SOCl ^ 3— Methallyloxypropionitrile was prepared, in good yield from the reaction of acrylonitrile with methallyl alcohol and a small amount of concentrated sodium ethoxide. Attempts to hydrolyze this nitrile to

PAGE 31

22 3-methallyloxypropionic acid by the procedure of Christian and llixon (Al), using hydrochloric acid, failed. 3-Methallyloxypropionic acid was readily prepared, however, by the reaction of methallyl alcohol with B-propiolactone. CH„ I 3 CH =C-CH20H + ^ By that time it had been established from the work of Walling and coworkers (27) and Carlsson and Ingold (37) that generating radicals of this type by the reaction of the corresponding bromides with tributyltin hydride in the presence of a radical initiator provides a much better way of studying the competing reactions which the radicals undergo. + Bu^SnH a™ > + Bu„SnBr i This method offers a number of advantages over generating tha radical from the peroxide. The coupling products and products containing a C0„ group are eliminated. The number of products which are formed when such a radical is generated is only two or three. The kinetics of the competing radical reactions can be easily studied in the case of the radicals generated from the bromides by reaction with the tributyltin radical.

PAGE 32

23 Tributyltin hydride was prepared according to the procedure of Kuivila and Buemel (42) by reaction of tributyltin chloride with lithium aluminum hydride. It was stored under nitrogen in a sealed container. Bu.SnCl + LiAlH, 3 4 -> Bu SnH Preparation of 2-brom.oethyl methallyl ether and the products expected from generation of the 2-methallyloxyethyl radical — ethyl methallyl ether, 3,3-dimethyltetrahydrofuran, and 3-methyltetrahydropyran — provided the next synthetic objectives. Synthesizing 2-bromoethyl methallyl ether involved a relatively straightforward three-step synthesis. Reacting the sodium alkoxide of ethylene glycol with methallyl chloride produced 2-methallyloxyethanol in good yield. HOCH CH ONa + CH -C-CH CI ' ^ ^'^° > This alcohol was then converted to the tosylate as before using a slurry of p-toluenesulfonyl chloride in pyridine. The unpurified tosylate derivative was then converted to the bromide by refluxing in a solution of lithium bromide in acetone. Purification of the bromide was accomplished by column chromatography followed by a flash distillation. LiBr + TsCl C II N

PAGE 33

24 The alcohol was also converted directly to the bromide in low yiel 1 by reaction with phosphorus tribromide and pyridine. The yield from this route was only 11 percent, whereas the alcohol was converted to the bromide by way of the tosylate in 50 percent yield. Ethyl methallyl ether was prepared by reacting sodium ethoxide with methallyl chloride in ethanol. CH2=C-CH2C1 + C2H30Na -^^^ 3,3-Dimethyltetrahydrofuran was prepared from 2 ,2-dimethylsuccinic acid by a three-step synthesis. The diacid was converted to the diethyl ester by the acid-catalyzed reaction between 2 ,2-dimethylsuccinic acid and an excess of ethanol in toluene. Diethyl 2 ,2-diraethylsuccinate was reduced to 2 ,2-dimethyl-l,4-butanediol by reaction with lithium aluminum hydride in ether. The 2 ,2-dim£thyl-l,4-butanediol was then converted to 3 ,3-dimethyltetrahydrof uran by reaction with 60 percent sulfuric acid at 100°. CO„H CO„C„H, CH„OH I 2 I 2 2 5 I 2 l"2 C„H,OH
PAGE 34

25 3-Methyltetrahydropyran was prepared from a similar series of reactions beginning with 2-methylglutaric acid. By completely CO^H I ^ CH I CH„ I ' CIl-CH, CO^H C2H OH toluene CO2C2H3 CH„ I ^ CH„ I ^ CH-CH I CO2C2H3 LA H V ether CH-OH I ^ CH_ I ^ CHI ^ CH-CH, CH OH analogous reactions, the diethyl ester was formed and converted to 2-methyl-l,5-pentanediol. The diol was again cyclized by heating with 60 percent sulfuric acid, but this time a single cyclic product was CH-OH I ^ ™2 CHI ' CH-CH„ CH OH H2O CH2+ CH„ CH. CH-CH, CH OH CH OH CH, CH, CH-CH, CH2+ CH-OH 1 2

PAGE 35

26 not obtained. The main product, 3-methyltetrahydropyran, was contaminated with a second cyclic product, which was not purified, but was identified by its NMR spectrum to be 2 ,2 -dimethyltetrahydrofuran. Formation of this product could arise simply from a 1,2 hydride shift on the carbonium ion before cyclization occurs. A careful distillation through a spinning-band column was required to separate the 3-methyltetrahydropyran from the 2 ,2-dimethyltetrahydrof uran. The 2-(2-phenylallyloxy)ethyl radical system 2-Bromoethyl-2-phenylallyl ether was prepared in a three-step synthesis analogous to the one used for the methallyl system. Reaction between the sodium alkoxide of ethylene glycol and 2-phenylallyl chloride produced 2-(2-phenylallyloxy)ethanol, which was converted by way of the tosylate to the bromide in 78 percent yield. CH^=C-CH^C1 + HOCH^CH^ONa 2)2 22 glycol ^6»5 C.H. ^ / „ C^H, 1^ acetone l^ The three products expected from the reaction of the 2-(2-phenylallyloxy) ethyl radical with tributyltin hydride were ethyl 2-phenylallyl ether, 3-phenyltetrahydropyran, and 3-methyl-3-phenyltetrahydrofuran. Bu-SnH ^/ o D 3 >

PAGE 36

27 Ethyl 2-phenylallyl ether was prepared from the reaction of sodium ethoxide with 2-phenylallyl chloride in ethanol. CH„=C-CH„C1 + C^H^ONa 2 I 2 2 5 C2H^0H The other tvro expected products were obtained by reacting equimolar portions of 2-bromoethyl 2-phenylallyl ether in benzene solution in an autoclave at 130°. At this temperature the main products were 3-phenyltetrahydropyran and 3-methyl-3-phenyltetrahydrofuran. These tv70 compounds were distilled away from the tributyltin bromide and were separated and purified by preparative scale gas chromatography. The 3-(3-phenyltetrahydrofuranyl)methyl radical A five-step reaction sequence, which was reported by Colonge and Inf arnet (43) , was employed for the synthesis of 3-phenyl-3-bromomethyltetrahydrofuran. C^H^-CHCCO^C^H^)^ + BrCH^CO^C^H^ TsCl V C„H N NaOC^H^ C2H^0H 85% H„PO, 3 4 100° ^6^5-2 CH2CO2.C2H3 LAH ether C^H3-C(Cll20H)2 CH2CH2OH LiBr methyl isobutyl ketone / V~CH2Br

PAGE 37

28 Reaction of diethyl phenylmalonate in a solution of sodium ethoxide and ethanol with ethyl bromoacctate gave a 59 percent yield of triethyl 2-phenyl-2-carboxyinethylsuccinate and a 33 percent yield of ethyl ethoxyacetate . C,H -CH(CO„C-H_)C,H -C(CO„C-H,)„ ^ ^ 2 2 5 2 NaOC^H^ ^ ^ \ 2 2 5 2 + BrCH^CO^C^H^ C^H OH ^ ^"2^°2SS The triethyl 2~phenyl-2-carboxymethylsuccinate was reduced with lithium aluminum hydride to 2-phenyl-2-hydroxymethyl-l,4-butanediol in 88 percent yield. This was then reacted with 85 percent phosphoric acid to give a 37 percent yield of 3-phenyl-3-hydroxymethyltetrahydrofuran. The literature (43) reported a 33 percent yield of this product based on the amount of triethyl 2-phenyl-2-carboxymethylsuccinate used. Tliis is exactly what was observed in our case. C,H -C(CO„C H ) C,H -C(CH OH) ^ ^ I 2 2 5 2 LAH , ^ ^ | ^ ^ h+ CH^CO^C^H^ ^ CH2CH2OH 33% The tosylate was prepared by reacting the alcohol with ptoluenesulfonyl chloride and pyridine. A suitable solvent had to be sought for the conversion of the tosylate to 3-phenyl-3-bromomethyltetrahydrofuran. Acetone was not suitable because its reflux temperature was not hot enough to bring about the reaction. Methyl isobutyl ketone, boiling at 117°, proved to be a good solvent for the reaction. The bromide was obtained in 80 percent yield relative to the alcohol.

PAGE 38

29 "CH„OH / 1 — > 2 ^ T.n . / \ ^ _LiBx. + TsCl "cThTn \^ "methyl J J *^ isobutyl ketone B. Generation and Reactions of Radicals The 2-methallyloxyethyl radical The techniques involved in generating the radicals and analyzing the resulting products were modeled after the procedures of Walling and coworkers (27) and Wilt and coworkers (44). In all cases the molar ratio of the bromide to tributyltin hydride in the reaction mixtures was about three to one. Quantitative determinations of the concentration of products resulting from the reactions of the radicals with tributyltin hydride were made by gas chromatography, using a Hy-Fi Aerograph with Disc Integrator. A nine-foot b ,b '-oxydiproplonitrile column at 53-55^ gave the best separation of the products from each other and from the solvent benzene. A number of standard solutions were prepared containing the products in the range of concentrations in which they were prepared during the radical reactions. A standard calibration curve was prepared for the peak area versus concentration for each of the

PAGE 39

30 products. The accuracy of the determinations of product concentrations in the reaction mixtures was checked often by preparing solutions containing the concentrations of the three products in the reaction mixture according to the calibration curves. In this manner slight errors could be corrected and the calibration curves could be checked often. Duplicate reactions were occasionally run at a particular concentration of tributyltin hydride, and in each case the reaction mixtures showed the same product concentrations. The first attempt to study the reactions of the 2-methallyloxyethyl radical with tributyltin hydride demonstrated the importance of having very pure solvent and reagents and of quickly and carefully preparing the reaction mixtures. The solvent was benzene which had been distilled from sodium. The 2-bromoethyl methallyl ether was pure according to gas chromatographic analysis. The AIBN was recrystallized from methanol. The tributyltin hydride was prepared a few days before it was used, but it was not stored in an airtight container and there was a cloudy white material in the bottom of the bottle. The tributyltin hydride was weighed in a flask, followed by the bromide. Then the contents were diluted to the mark with benzene and the AIBN was added. The solutions were transferred to the reaction tubes. They were degassed by three freeze-thaw cycles, sealed, and heated at the desired temperature until the reaction was complete— until all the tributyltin hydride was consumed. The lack of any definite trend in the product distributions from the reaction mixtures heated at 90°, along with the fact that the yields were somewhat lower than anticipated, led to suspicions about the presence of impurities in the reaction solvents and about the

PAGE 40

31 method in which the solutions were prepared. The speed at which the reactions reached completion, even at 40°, also caused some concern about the care necessary to prevent reaction from occurring before the solutions were put in the constant temperature baths. Thus it was decided that the reactions should be repeated, taking greater care about the purity of tie reagents and the contact of the reactants prior to their being heated at a constant temperature. The solvent used this time was spectrograde benzene which had been dried over activated 3A molecular sieves. Tributyltin hydride was freshly prepared and stored in an airtight container under nitrogen. The amount of AIBN used was reduced from 3 mole percent to 1.5 mole percent based on hydride concentration. The solutions were prepared rapidly and were promptly added to the tubes and frozen. The AIBN was weighed out first and set aside. The bromide was weighed next and set aside. The tributyltin hydride v/as weighed in a volumetric flask. The bromide was washed in with solvent and the flask was filled with solvent. Then the AIBN was added, the mixture was shaken, and the solution was put in the tubes and frozen. The solutions prepared in this manner are described in Table I. The product concentrations resulting from these reaction solutions are given in Table II. The percent yields of the products based on the initial concentrations of the tributyltin hydride are listed in Table III. The ratio of 3 ,3-dimethyltetrahydrofuran to 3-methyltetrahydropyran formed at each temperature was approximately the same for all three concentrations of reactants. The ratios observed at the three reaction temperatures are listed in Table IV.

PAGE 41

32 TABLE I. REACTION SOLUTIONS FOR GENERATION AND REACTIONS OF 2-METHALLYLOXYETHYL RADICAL

PAGE 42

33 TABLE III. PRODUCT PERCENTAGES FROM REACTIONS OF THE 2-METHALLYLOXYETHYL RADICAL

PAGE 43

34 and 3-phenyltetrahydropyran, were best separated on a five-foot 30 percent SE-30 column heated at 190°, with the injection port at 290°. Table V shows the compositions of the solutions prepared to study the reactions of the 2-(2-phenylallyloxy)ethyl radical with tributyltin hydride. TABLE V. REACTION SOLUTIONS FOR GENERATION AND REACTIONS OF THE 2-(2-PHENYl.ALL^i-LOXY)ETHYL RADICAL Solution Temperature :Bu SnH] Kr [AIBN] 401

PAGE 44

35 The percent 3'ields of the products based on the initial concentrations of the tributyltin hydride are listed in Table VII. TABLE VII. PRODUCT PERCENTAGES FROM REACTIONS OF THE 2-(2-PHENYLALLYLOXY)ETHYL RADICAL

PAGE 45

36 The 3-(3-phenyltetrahydrofuranyl)methyl radical The 3-(3-phenyltetrahydrofuranyl)methyl radical V7as generated from 3-phenyl-3-broniomethyltetrahydrofuran under the usual reaction conditions at 90° to check the reversibility of its formation from the 2-(2-phenylallyloxy)ethyl radical. It was reasoned that if either 6 SN-^^'^ • ? of the cyclization reactions of the 2-(2-phenylallyloxy)ethyl radical was reversible, the one sho\\Ti above should be since the other radical leading to 3-phenyltetrahydropyran should be a more stable benzylic radical. A solution was prepared, as in the previous cyclization experiments, which was 0.1000 molar in 3-phenyl-3-bromomethyltetrahydrofuran, 0.1000 molar in tributyltin hydride, and 0.0017 molar in AIBN. It was heated at 90° then analyzed by gas chromatography. 3-Phenyl-3-methyltetrahydrofuran was the only main product of the reaction. There was no ethyl 2-phenylallyl ether or 3-phenyltetrahydropyran formed. Thus it appears the radical cyclization reactions of the 2-(2phenylallyloxy) ethyl radical are irreversible processes, just as they were sho^im to be for the cyclization reactions of the 5-hexenyl radical (14). C. Kinetics of the Radical Reactions The groundwork for studying the kinetics of the reactions of radical systems of the type studied in this work was established by Carlsson and Ingold (37) and by Walling and cov/orkers (27). Carlsson

PAGE 46

37 and Ingold established the fact that for the reaction of an alkyl bromide with tributyltin hydride in the presence of a radical initiator, the rate-controlling step is abstraction of hydrogen from the tributyltin hydride by the alkyl radical. Thus the competing reactions of the alkyl radical can be studied kinetically by this method. Walling and coworkers (27) reported a kinetic treatment for the competing reactions of the 5-hexenyl radical in the presence of a limited amount of tributyltin hydride. They used an excess of 6-bromo1— hexene and allowed the reactions to proceed until the tributyltin hydride was completely consumed. A diagram of the reaction scheme was presented on page 9 of the introduction. They ran the reactions at 40° and 130° using different tributyltin hydride concentrations. Their data at 40° could be accounted for by simple classical reactions as well as by the more complex treatment involving a dependence on the tributyltin hydride concentration for the formation of methylcyclopentane, which was required to account for the product distributions observed when the reactions were rva^ at 130°. The two kinetic equations which they used are shown on page 10. They omitted k„ because they observed only a small amount of cyclohexane in the reaction mixtures. Equation 1 is applicable for the simple case, while Equation 2 takes into account the dependence of methylcyclopentane formation on the tributyltin hydride concentration — the formation of methylcyclopentane by direct attack of tributyltin hydride on the uncyclized radical, perhaps in some form of radical-double bond complex. The simple equation with k„ included. Equation 3, quite satisfactorily fit the data obtained for the reaction products obtained from the 2-methallyloxyethyl radical and the 2-(2-phenylallyloxy)ethyl

PAGE 47

38 radical. The equation was integrated between the initial and final reaction conditions to yield Equation 4. d[Bu SnH] dxk^ + k2 + k^LBu^SnH] k^LBu^SnH] (3) L_ oJ k + k k + k + k [Bu SnH] The scheme for the reactions of the two radical systems studied is shown on the following page. It is just like the reaction scheme described for the generation and reactions of the 4-(l-cyclohexenyl)butyl radical by Struble, Beckwith, and Gream (28). Just as they observed a constant ratio of the spiro compound to the decalin formed + Bu SnH AIBN X.^. Bu SnH Bu SnH Bu SnH X

PAGE 48

39 over a seven-fold change in tributyltin hydride concentration, we observed a constant ratio of tetrahydrofuran derivative to tetrahydropyran derivative over a four -fold change in tributyltin hydride concentration at each temperature for both of the radical s^'stems studied. Since the cyclization of the 2-(2-phenylallyloxy)ethyl radical to the 3-(3-phenyltetrahydrofuranyl)methyl radical was shown to be an irreversible process, it would seem reasonable to assume that all four cyclic radicals formed in the two radical systems studied were C.H, X formed irreversibly. The methyl analog of the five-member ed cyclic radical should be no more likely to undergo the reverse reaction than the phenyl system, while the two six-membered cyclic radicals, one being tertiary and the other benzylic, should be less likely to undergo the reverse reaction. The lack of reversibility of the radical cyclizations coupled with the constant ratios of cyclic products at each temperature supports the proposed reaction scheme, whereby the cyclic products arise from irreversible radical cyclizations whose rates are independent of the tributyltin hydride concentration. Thus the ratio of k to k^ is just the constant ratio of tetrahydrofuran derivative to tetrahydropyran derivative formed at that temperature. (5) a"

PAGE 49

40 With this relationship hetween k and k established, it was. possible to solve Equation 4 by trial and error for a value of k /k or k„/k. which would fit each set of data. The values of k /k or k„/k„ which gave the best fit for the products obtained from the 2-mcthallyloxyeth3'l radical and the 2-(2-phenylallyloxy) ethyl radical respectively are listed in Table IX. Also listed in Table IX Xv are the values of 'Xj , the calculated values of the calc. concentration of uncycllzed product, which are obtained when the k„/k or k„/k„ which best fits the data is. plugged into Equation 4. From Table IX it can be seen that a single value of k„/k^ or k„/k„ gives a good fit of the data to Equation 4 at each temperature. The particular value of k./k or k, /k. which best fits the results from a single reaction solution to Equation 4 is no more than 3 percent away from the value which gives the best overall fit for the data at that temperature. The values of k /k„ , k„/k , and k /k„ which were obtained for the two radical systems at each of the three reaction temperatures are collected in Table X.

PAGE 50

TABLE IX . VALUES OF k^/k^ or k /k WHICH BEST F IT THE. DATA ACCOED ING TO E.Q5kT ION 4 SYSTEM 41 Temperature Xj obs. k3/k^ calc. 41

PAGE 51

42 TABLE X. COMPOSITE OF RATE CONSTANT RATIOS Temper, , , , Radical ature \'^2 S'\ ^^'^l ^eS 40

PAGE 52

43 3.02.82.6 r2.4 2.2 2.0(1/T) X lO' Figure 1. Ln k /k versus 1/T for the 2-methallyloxyethyl radical.

PAGE 53

44 4.03.83.63.43.23.0 2.4 (1/T) X 10 Figure 2. Ln k /k versus 1/T for the 2-methallyloxyethyl radical.

PAGE 54

45 7.06.6 6.2 5.85.45.0 2.4 2.6 2.8 3.0 3.2 (1/T) X 10" Figure 3. Ln k /k versus 1/T for the 2-methallyloxyethyl radical,

PAGE 55

46 2.42.2 2.01.81.6 2.4 (1/T) X 10 Figure 4. Ln k /k versus 1/T for the 2-(2-phenylallyloxy)ethyl radical,

PAGE 56

47 0.700.680,660.640.620.600.580.56 2.4 2.6 (1/T) X 10' Figure 5. Ln k /k versus 1/T for the 2-(2-phenylallyloxy)ethyl radical.

PAGE 57

48 3.23.02.62.42.2 2.4 2.6 2.8 3.0 3.2 (1/T) X 10Figure 6. Ln k /k versus 1/T for the 2-(2-phenylallyloxy)ethyl radical.

PAGE 58

49 Equation 4, considering the product concentrations in the reaction mixtures were within about 2 percent of the measured values. TABLE XI. ACTIVATION ENERGY DIFFERENCES AND FREQUENCY FACTOR RATIOS FROM FIGURES 1-6 ^6«5 Property [\ \.J ^A ~ ^A (Kcal/mole) 1.7 ± 0.15 0.26 ± 0.05 2 "1 ^1 " \ ^2 ^3 E E (Kcal/mole) 2.2 ± 0.15 1.3 ± 0.25 ^1 ^2 ^A ~ \ (Kcal/mole) 3.9 ± 0.3 1.6 ± 0.25 A^/A^ 0.35 ±0.08 2.8 ±0.2 A /A^ (liter/mole) 0.44 ± 0.07 2.3 ± 0.8 A^/A^ (liter/mole) 1.3 ± 0.55 0.85 ± 0.25 Although the product distributions obtained from the reactions of the two radical systems studied could be satisfactorily explained by the scheme on page 38 involving simple irreversible radical cyclizations, the possibility of the presence of a radical-it complex at some point in the reaction cannot be ruled out. In the following scheme a complex is included. It is possible that a complex is formed reversibly from the uncyclized radical and then collapses irreversibly to the cyclic radicals. The fact that 3-phenyl-3-methyltetrahydropyran was the only product obtained from the 3-(3-phenyltetrahydrofuranyl)methyl radical shows that such a transformation from a complex to the cyclic radical, if it is involved, is irreversible. A direct route from complex to cyclic products could be possible if the rate-determining step is formation of the

PAGE 59

50 V^ 'Br + Bu SnH AIBN < Bu SnH ^2 "-^ \ s t) 3u SnH V \^ SnH X -Bu^^H u SnH complex rather than attack of the complex by the tributyltin hydride. The steric factors, if important enough in view of the electronic factors involved, would discount the last possibility since the products obtained were not those expected from attack on a complex by the hydride at the more accessible position. Some of these problems might be answered by a study of a similar system with a bulky alkyl group on the double bond.

PAGE 60

51 D. D iscussion As was set forth in the "Statement of the Problem," the major objective of this research was to determine the mechanism involved in the cyclization reactions of 5,6-unsaturated radicals. The approach taken was to observe the products obtained from radicals of this type with a substituent at the 5-position. If formation of f ive-membered cyclic products arises totally or in part from attack by the hydrogen donor on the uncj'clized radical in some type of a radical-ir complex at the more sterically favorable position, as suggested by Lamb, Ayers, and Toney (14) and by Walling and coworkers (27), the 5substituted radicals should have shown an even greater tendency to form f ive-membered cyclic products. The data obtained in this research project show that this is not what was observed. The product distributions from the 2-methallyloxyethyl radical (Table III) shov? that up to 3 percent 3-methyltetrahydropyran was formed in £iddition to the two major products. No tetrahydropyran formation was reported by Lamb, Pacifici, and Ayers (45) or by Walling and coworkers (27) from the reactions of the 2-allyloxyethyl radical. In the case of the 2-(2-phenylallyloxy)ethyl radical the 3-phenyltetrahydropyran was the predominant cyclic product. Thus the data suggest that the cyclic products are not formed by way of a concerted attack by the hydrogen donor on the uncyclized radical in the form of a complex. Julia (46) recently reported that the 5-methyl-5-hexenyl radical gave a ratio of methylcyclohexane to 1 ,1-dimethylcyclopentane of about 1.7. This also lends support against the necessity of a complex leading to cyclic products.

PAGE 61

52 The kinetic results show that the reaction scheme shown on page 33 quite adequately describes the reactions of the two radicals studied in this research. Equation 4, which was derived from this reaction scheme, gave a good fit to the data for both radical systems at all three reaction te'^-peratures. The formation of the cyclic products can be accounted for by irreversible cyclizations of the initially formed radicals — processes whose rates are independent of the tributyltin hydride concentration. The irreversibility of the cyclizations was checked by generating the 3-(3-phenyltetrahydrofuranyl)methyl radical from the corresponding bromide. 3-Phenyl-3methyltetrahydrofuran was the only product formed. From the kinetic results it was possible to get an idea of the activation energies for the cyclization reactions relative to the activation energy for the abstraction of hydrogen from tributyltin hydride by a primary alkyl radical. Wilt and coworkers (44) have calculated the activation energy for this hydrogen abstraction process to be between 6.8 and 8.2 Kcal./mole using rate constants reported by Carlsson and Ingold (37). The comparisons of the rate constants for the cyclization reactions to that for the hydrogen abstraction from tributyltin hydride by the initially formed primary alkyl radicals were presented in Table XI. The rate constants for hydrogen abstraction by the 2-methallyloxyethyl and the 2-(2-phenylallyloxy)ethyl radical should be very nearly the same. Thus it is possible to compare the activation energies for the cyclization reactions leading to the four cyclic products in the two radical systems. The order of activation energies for the four cyclizations are presented in relation to X, the activation energy for the hydrogen

PAGE 62

53 abstraction by the initially formed primary radicals. Product A X (Kcal./mole) 3-Phenyl-3-methyltetrahydrofuran 1.3 ± 0.25 3-Phenyltetrahydropyran 1.6 ± 0.25 3,3-Dimethyltetrahydrofuran 2,2 ± 0.15 3-Methyltetrahydropyran 3.9 ± 0.3 These values show that the activation energy for the cyclization of the 2-(2-phenylallyloxy)ethyl radical to form 3-phenyltetrahydropyran is about 2.3 Kcal./mole less than the activation energy for the cyclization of the 2-methallyloxyethyl radical to form 3-methyltetrahydropyran. This helps to explain why formation of the sixmembered cyclic product competes so much more favorably in the phenyl system than in the methyl system. The relationships between the three rate constants in both of the radical systems v^-ere established. Approximate values of the rate constants for the cyclization reactions of the two radicals were obtained from the rate constant relationships and the reported value (37) of the rate constant for abstraction of hydrogen from tributyltin hydride by the 1-hexyl radical (k = 1.0 x 10 M sec. at 25°). Comparing this value to the ratios reported in Table X gives the following values for the rate constants of the four cyclization reactions at 40°: Product c (sec. ) at 40° 4 3-Phenyltetrahydropyran 9.6 x 10 3,3-Dimethyltetrahydrofuran 6.1 x 10 3-l'henyl-3-methyltetrahydrofuran 5.3 x 10^ 3-Methyltetrahydropyran 1.4 x 10

PAGE 63

54 The total rate constant for cj'clization in either of the tv/o radical systems would be simply k^ + k„ , the sum of the two rate constants for the cyclization reactions. The approximate values 4 for the methyl and phenyl systems studied in this work are 6.2 x 10 and 1.5 x 10 sec. respectively. The fact that this total rate constant for cyclization is more than twice as great for the phenyl system as for the methyl system appears to arise from the differences in rate constants for cyclization to the tetrahydropyran derivatives. The rate constants for formation of the tetrahydrofuran derivatives are nearly the same. This big difference is likely due to the greater stability of the benzylic radical leading to 3-phenyltetrahydropyran compared to the tertiary radical leading to 3-methyltetrahydropyran. Carlsson and Ingold (37) determined the rate constants for cyclization at 40° for the 1-hexenyl radical studied by Walling and coworkers (27) and for the 4-(l-cyclohexenyl)butyl radical reported by Struble, Beckwith, and Gream (28) by comparing the reported rate constant ratios with the rate constant for hydrogen abstraction from tributyltin hydride by the 1-hexyl radical at 25°. The values of k obtained 5-1 4-1 were 1 x 10 sec. for the 5-hexenyl radical and 4 x 10 sec. for the 4-(l-cyclohexenyl)butyl radical. They compared these constants with the rate constant for addition of an ethyl radical to 1-heptene 3 -1 -1 at 40°, 1 X 10 M sec. , and stated that the "effective double bond concentration" for the intramolecular cyclization of such radicals is about 40 to 300 M. Such a comparison with the cyclization rate constant for the 2-metballyloxyethyl radical yields a value of 62 M. For the 2-(2-phenylallyloxy)ethyl radical the value would be 150 M

PAGE 64

55 but a more accurate basis for comparison would be with the rate constant for addition of a primary radical to an a-alkyl styrene, 3 —1 -1 which should have a value greater than 1 x 10 M sec. at 40°. This enhanced "effective double bond concentration" for reaction of the initial radical with the double bond to give a cyclic radical, coupled with the increased steric hindrance to approach of another monomer molecule when the initial radical is part of a growing polymer chain, should be sufficient to explain why so many 1,6-dienes undergo free-radical polymerization to yield polymers composed entirely of cyclic repeating units. Propagation by reaction of the initial radical with another molecule of monomer simply cannot compete with the cyclization processes in most cases. The products obtained from the reactions of the two radicals studied in this work were consistent with a reaction scheme whereby the cyclic products arise from irreversible radical cyclizations. The kinetic treatment further supports this explanation. Equation 4 when applied to the data of Struble, Beckwith, and Gream (28) for the products from reactions of the 4-(l-cyclohexrnyl)butyl radical with tributyltin hydride at 40° gave a good fit for k^/(k^ + k2) = 40. Equation 4 also gave a good fit to the data reported by Beckwith and Gara (47) for the reactions of the 2-(3-butenyl)phenyl radical with tributyltin hydride with k„/k^ =2.0. The data of Walling and coworkers (27) for the reactions of the 5-hexenyl radical with tributyltin hydride at 40° were satisfied by Equation 4 with k^/k^ = 10. Only Walling's data at 130°, which showed no distinct trend in product distribution, failed to satisfactorily fit Equation 4.

PAGE 65

56 From our results and the results of other studies on similar radical systems, it would appear that a reaction path involving a complex between the radical and the double bond is not necessary to explain the formation of the cyclic products. A schem.e involving a competition between the simple irreversible radical reactions leading to the cyclic and nonc^'clic products satisfactorily explains the observed product distributions. It is possible, however, that some type of a nonclassical radical species is involved in the reaction at some point. The pofcdibilities were discussed at the end of the previous section. The fact that the cyclic radicals were found to form irreversibly discounts the possibility of the cyclized radicals beiiig in equilibrium with a complex. If such a complex were low in energy relative to the cyclic radicals, one might expect the cyclic radicals to revert back to the complex form and partition themselves between the two cyclic products and perhaps also the uncyclized product. The results of Kochi and Krusic (17) also cast some doubt as to the formation of a nonclassical radical species in radical cyclizations of this type. By ESR they were able to see the 5-hexenyl radical and the cyclopentylmethyl radical but no other in between. The proportions of f ive-membered and six-membered cyclic products in the two radical systems studied appeared to be dependent upon the activation energies for the competing cyclization reactions and also upon steric factors concerning the ease of approach of the radical to the two ends of the double bond. These effects would also be important in determining the ring-size of the repeating units of cyclopolj'T^ers. IJhen the initial radicals are stable enough that

PAGE 66

57 the radical cyclization reactions are partially or totally reversible, as in the case of some of the radical systems studied by Julia (36) , the stabilities of the cyclized radicals become an additional factor in determining the proportions of f ive-membered and six-membered cyclic repeating units in cyclopolymers.

PAGE 67

Chapter III EXPERIMENTAL A. Equipment and Data All temperatures are reported uncorrected in degrees ceiitigrade. Nuclear magnetic resonance (NMR) spectra were obtained on a Varian A-60A Analytical NMR Spectrometer. The chemical shifts were measured in carbon tetrachloride and deuterium oxide, using tetramethylsilane as reference. Infrared spectra were obtained with a Beckman IR-8 Infrared Spectrophotometer . Refractive indices were obtained with a Bausch and Lomb Abbe 34 Ref ractometer equipped with an anchromatic compensating prism. Elemental analyses were performed by Atlantic Microlab, Inc., Atlanta, Georgia, and by PCR, Inc., Gainesville, Florida. Gas chromatographic analyses were done on a Hy-Fi Aerograph Model 600-D Gas Chromatograph. Preparative scale gas chromatography was done on an F & M Model 775 Prepmaster Gas Chromatograph. The reactions were run in a constant temperature bath controlled by a Sargent Therraonitor to within ± O.Ol^C. B. Syntheses Preparation of A-methyl-4-penten-l-ol 4-Methyl-4-penten-l-ol was prepared according to the procedure of Johnson and Ou^yang (48). The reaction was carried out in a hood. 58

PAGE 68

59 A 2000-ml. three-necked, round-bottomed flask was equipped with mechanical stirrer, dropping funnel, dry-ice condenser, and nitrogen inlet. With nitrogen flowing through the system, 48.6 g. (2.0 moles) of magnesium turnings, 100 ml. of anhydrous ether, and a few iodine crystals were put in the flask. A solution of 90.6 g. (1.0 mole) of methallyl chloride in 900 ml. of anhydrous ether was added at such a rate as to maintain gentle refluxing. The mixture was refluxed for one hour on a steam bath, then 44 g. (1.0 mole) of ethylene oxide was added. The addition was accomplished by allo\;ing the ethylene oxide to evaporate into a stream of nitrogen, which bubbled in just below the surface of the Grignard mixture. The mixture was allowed to stir for one hour at room temperature after addition was completed. The reaction mixture was poured over 300 ml. of ice water. The excess magnesium was removed by pouring the mixture through a strainer. This was followed by hydrolysis vrith saturated ammonium chloride solution until the mixture was almost neutral. The aqueous solution was extracted with ether 24 hours in a continuous extractor. The ether solution was dried over anhydrous magnesium sulfate and filtered. The ether was removed at reduced pressure. The remaining crude liquid was distilled through a 16-cm. Vigreux column to give 75 g. (75 percent) 23 of a clear, colorless liquid b.p. 68-70/20mm. , n^ 1.4403 (lit. (49) 20 n^^ 1.4372). Gas chromatographic analysis using a 9-ft. 30 percent Carbowax 20M column at 120° C with the injector at 200° C indicated the product to be about 95 percent pure. This accounts for the fact that the refractive index does not agree too closely with the literature value.

PAGE 69

60 The infrared spectrum (neat) of the product showed absorption bands at 3350 (broad, s) , 3090 (m) , 2980 (s) , 2940 (s), 2880 (s) , 1650 (m), lAAO (m), 1375 (ir.) , 1060 (s) , 1020 (w) , 995 (w) , 880 (s) cm."-"-. The NMR spectrum (CCl.) showed a multiplet centered at 8.35 T overlapping a singlet with hyperfine splitting centered at 8.28 T, a broadened triplet centered at 7.93 T, a singlet at 6.96 T, a triplet centered at 6.48 T, and a singlet with hyperfine splitting centered at 5.33 T with area ratios of 2:3:2:1:2:2 respectively. Attempted preparation of 5-bromo-2-methyl-l-pentene This synthesis was attempted by the adaptation of a procedure by Gaubert, Linstead, and Rydon for the preparation of 5-bromo-lpentene (50). The reaction was carried out in a hood. A 200-ml. three-necked, round-bottomed flask, fitted with mechanical stirrer, dropping funnel, thermometer, and exhaust hose to a trap of aqueous potassium hydroxide, was charged with 54.3 g. (0.20 mole) of freshly distilled phosphorus tribromide. The flask was maintained below 0° C while a mixture of 47.5 g. (0.47 mole) of 4-methyl-4-penten-l-ol and 15 g. of pyridine was added dropwise over a period of three hours. The reaction mixture was left at room temperature for 48 hours, then it was distilled at a pressure of 26 mm. Caution ! An oil bath was used and its temperature was not allov^7ed to exceed 140°C. Higher temperatures can result in an explosion. The liquid thus obtained was distilled through a 16-cm. Vigreux column to give 42.3 g. (55 percent) of a clear, colorless liquid b.p. 56-58/24.5 mm. The NMR spectrum (CCl.) of the liquid indicated the presence of a mixture of 5-bromo-2-methyl-l-pentene and 5-bromo-2-methyl-2-pentene

PAGE 70

61 in an approximate ratio of 1:2 respectively. The ratio was determined from the integration of vinyl proton peaks in the NMR spectrum. The two vinyl protons of 5-bromo-2-methyl-l-pentene appeared as a singlet with hyperfine splitting at 5.28 T. The one vinyl proton of 5-bromo2-raethyl-2-pentene appeared as a triplet with hyperfine splitting centered at 4.89 T. The desired compound probably could have been obtained if greater care had been taken in the workup procedure. The isomerization probably occurred during the distillation. Preparation of thionyl bromide Thionyl bromide was prepared according to the method in Inorganic Syntheses (51). The reaction was carried out in a hood. In a 200-ml. trap was placed 100 ml. (162.8 g.) of thionyl chloride. Hydrogen bromide gas was bubbled through very slowly for 16 hours while the trap was maintained at 0°C. The acidic fumes were collected in a trap containing concentrated potassium hydroxide solution. Distillation yielded 111 g. of a pale red liquid b.p. 43-44/19 mm. The yield was greatly reduced by a number of mishaps caused mainly by a bad valve on the hydrogen bromide cylinder. Attempted preparation of 5-bromo-2-methyl-l-pentene A procedure reported by Frazer and coworkers (52) for converting alcohols to bromides was employed. The reaction was carried out in a hood. An ice-methanol bath was used to cool a 300-ml. three-necked, round-bottomed flask, equipped with mechanical stirrer and dropping funnel, and containing 20 g. (0.20 mole) of 4-met}iyl-4-penten-l-ol ,

PAGE 71

62 40 g. of pyridine, and 100 ml. of ether to -10°C. Dropwise addition of 11.9 g. (0.10 mole) of thionyl chloride was followed by a brief stirring period and then filtration of the precipitate. The N14R spectrum of the remaining liquid was that expected for di-4-methyl-4pentenyl sulfite: a singlet with hyperfine splitting at 8.27 T overlapping a multiplet centered at 8.15 T, a multiplet centered at 7.95 T, a multiplet centered at 6.06 T, a singlet with hyperfine splitting at 5.28 T with area ratios of 3:2:2:2 respectively. Attempts to distill this high-boiling liquid, even at very low pressures, resulted in some isomerization of the double bond, thus the crude sulfite was used in the second step of the reaction. The sulfite and 20 g. of pyridine were placed in a 200-ml. threenecked, round-bottomed flask, equipped with mechanical stirrer, condenser, and dropping funnel, and 20.8 g. (0.10 mole) of thionyl bromide was added slowly. The mixture was heated to 70°C and stirred at that temperature for three hours. The resulting reaction mixture was a messy black tar. Attempts to isolate any product by extraction with various organic solvents failed. When this second step was repeated omitting the pyridine, the bromide was formed, but the double bond underwent isomerization to a large extent — about 50 percent. Attempted preparation of 5-iodo-2-methyl-l-pentene The reaction was carried out according to the procedure of Landauer and Rydon (38) for converting alcohols to iodides. The reaction was carried out in a hood. In a 500-ml. two-necked, roundbottomed flask, equipped with mechanical stirrer, reflux condenser.

PAGE 72

63 and drying tube, were put 155.0 g. (0.50 mole) of triphenyl phosphite, 71.0 g. (0.50 mole) of methyl iodide, and 50.0 g. (0.50 mole) of 4-methyl-4-penten-l-ol. The mixture was heated at 80-90 °C for 60 hours, cooled, and poured into 300 ml. of ether. The ether solution was washed with two 200-ml. portions of 2N potassium hydroxide and then with 150 ml. of water, and dried over anhydrous magnesium sulfate. The ether was removed at reduced pressure and a sma.1 '' portion of the product was distilled b.p. 60-62/15 mm. The NMR spectrum (CCl, ) of the product indicated the double bond had isomerized a great deal, v-zith the internal double bond of 5-iodo2-methyl-2-pentene predominating over the terminal double bond of 5-iodo-2-methyl-l-pentene at a ratio of 5:1, according to the relative areas of the vinyl proton peaks in the N'-IR spectrum. Preparation of A-methyl-A-penten^'l p-toluenesulf onate 4-Methyl-4-pentenyl p-toluenesulf onate was prepared according to the procedure of Ansell (39). The reaction was carried out in a hood. In a 200-ral. three-necked, round-bottomed flask, equipped with mechanical stirrer, dropping funnel, and drying tube, were put 18 ml. of pyridine and 34.6 g. (0.18 mole) of p-toluenesulf onyl chloride. The slurry was stirred briefly, then 16.5 g. (0.16 mole) of 4-methyl-4penten-1-ol was added dropwise while the temperature was maintained at 20°C. The mixture was stirred one hour at room temperature then it was cooled to 0°C, followed by addition of 12 ml. of water. After brief stirring, the mixture was poured over 70 ml. of ice water. The organic portion was taken up in 50 ml. of ether. The aqueous layer was washed with two 30-ml. portions of ether. The ether portions

PAGE 73

64 were combined and vcashed twice each with 30 ml. of saturated ammonium chloride solution, 30 ml. of saturated sodium bicarbonate solution, and 30 ml. of water. It was dried over anhydrous potassium carbonate, and the ether was removed on a rotary evaporator. The remaining liquid could not be distilled. The crude product weighed 32.0 g. (79 percent). The infrared spectrum (neat) of the product showed absorption bands at 3090 (w) , 2990 (m) , 2940 (m) , 1650 (w) , 1600 (m) , 1440 (w) , 1355 (s), 1295 (w), 1190 (s) , 1175 (s) , 1100 (m) , 960 (m) , 920 (m) , 810 (m), 740 (w), and 660 (m) cm." . The NMR spectrum (CCl.) showed a singlet with hyperfine splitting at 8.33 T overlapping a multiplet centered at 8.21 T, a multiplet centered at 7.98 T, a singlet at 7.56 T, a triplet centered at 6.03 T, a singlet with hyperfine splitting centered at 5.38 T, a doublet centered at 2.73 T, and a doublet centered at 2.30 T with area ratios of 3:2:2:3:2:2:2:2 respectively. Attempted preparation of 5-iodo-2-m&thyl-l-pentene 5-Iodo-2-methyl-l-pentene was synthesized according to the procedure of Ansell (39). A 300-ml. round-bottomed flask was equipped with a magnetic stirring bar, condenser, and drying tube. To a solution of 11.0 g. (0.073 mole) of sodium iodide in 65 ml. of acetone was added 16.0 g. (0.06 mole) of 4-methyl-4-pentenyl p-toluenesulfonate from the previous experiment. The mixture was refluxed l-;r hours, then it was cooled to 0°C and filtered. The acetone was removed on a rotary evaporator. The residue was washed with 30 ml. of 10 percent sodium thiosulfate solution and 30 ml. of water and dried over anhydrous magnesium sulfate. Filtration and distillation yielded 4.0 g.

PAGE 74

65 of a light pink liquid whose composition was determined from the NMR spectrum (CCl.) to be a 3:1 mixture of 5-iodo-2-methyl-l-pentene and 5-iodo-2-methyl-2-pentene respectively. The second distillation fraction was 1.1 g. of a 1:3 mixture of the products. It was shown by heating the mixtures that the proportion of 5-iodo-2-methyl-2pentene increases with heating. The presence of 5-iodo-2-methyl-2pentene in the first fraction is not easily detected in the infrared spectrum, but it is easily seen in the vinyl proton region of the NMR spectrum. Preparation of dieth^'^l methallylmalonate The reaction was carried out according to a procedure in Organic Syntheses (53). The reaction was carried out in a hood. A 1000-ml. three-necked, round-bottomed flask, equipped with a mechanical stirrer, condenser and drying tube, dropping funnel, and nigrogen inlet, was purged with nitrogen, followed by addition of 300 ml. of dry ethanol and 13.8 g. (0.6 mole) of sodium chunks. When all of the sodium had reacted, 96.0 g. (0.6 mole) of diethyl malonate was added slowly. The stirring had to be very slow during the addition, or foaming resulted. To the clear solution was slowly added 54.3 g. (0.6 mole) of methallyl chloride. The reaction mixture was refluxed for several hours, but it remained very basic. Some potassium iodide was added and the mixture was refluxed again for several hours , at which time the mixture was less basic. As much ethanol as possible was distilled off with a steam bath. The residue was shaken with 250 ml. of very dilute hydrochloric acid. Some sodium chloride had to be added to facilitate separation of the diester. The crude product was dried over anhydrous

PAGE 75

66 magnesium sulfate. Distillation through a 16-cm. Vigreux column yielded 41.6 g. of 90 percent pure diethyl methallylmalonate, 42.8 g. of 75 percent pure, and 8,0 g. of 60 percent pure product. The approximate purity of the different fractions was determined by gas chromatography. A small portion of the 75 percent pure fraction was redistilled carefully through a 16-cm. Vigreux column to yield 3 g. of 98.5 percent 35 pure product b.p. 122-3/15 mm. (lit. (54) b.p. 94-6/5 mm.) and n^ 1.4300 (lit. (54) n^° 1.4340). The bulk of the product was not purified further, since the purpose of this experiment and the following one was just to test the procedures with an inexpensive halide similar to the one that was to used in later experiments. The infrared spectrum (neat) of the product showed absorption bands at 3100 (m), 3000 (s), 2950 (s) , 2920 (m) , 1740 (broad, vs) , 1650 (s) , 1450 (s), 1370 (s), 1330 (s) , 1310-1130 (broad, s) , 1095 (s), 1045 (s) , 1030 (s), 890 (s) , 860 (m) , and 790 (w) cm."-^. The N>IR spectrum (CCl.) showed a triplet centered at 8.76 T, a singlet with hyperfine splitting at 8.26 T, a doublet centered at 7.47 T, a triplet centered at 6.60 T, a quartet centered at 5.86 T, and a singlet with hyperfine splitting centered at 5.29 T with area ratios of 6:3:2:1:4:2 respectively. Preparation of d.ihydro-5 ,5-dimethyl-2 (3H)-furanone A modification of a procedure in Vogel's Textbook of Practical Organic Chemistry (55) for the hydrolysis and decarboxylation of malonic esters was used. The reaction was carried out in a hood. A

PAGE 76

67 solution of 41.5 g. of potassium hydroxide in 41.5 ml. of water was prepared in a 500-ml. round-bottomed flask. To the hot solution v;as added 41.5 g. of 90 percent diethyl methallylmalonate. The mixture was refluxed until the liquid was completely soluble in water. Then 45 ml. of water was added and 48 ml. of liquid was distilled off to insure the complete removal of ethanol. Addition of 175 ml. of 3N hydrochloric acid dropped the pH to 4. The solution was refluxed 6 hours, and then the product was removed by continuous extraction with ether for two days. The solution was dried over anhydrous magnesium sulfate, filtered, and stripped of ether by rotary evaporation. The remaining liquid was distilled through a 16-cm. Vigreux column, yielding 11.2 g. of a clear, 22 colorless liquid product b.p. 79-81/10 mm., n 1.4340. The product was shown to be a single component by gas chromatography. The structure of the product was shown by NI4R and infrared spectra to be not the expected 4-methyl-4-pentenoic acid, but dihydro-5 ,5-dimethyl2(3H)-furanone. The boiling point and refractive index agree closely with the literature values (56) for dihydro-5 ,5-dimethyl-2 (3H)-furanone: b.p. 79.5/10 mm., nj^ 1.4313. The infrared spectrum (neat) of the product showed absorption bands at 2990 (s), 2950 (m) , 2890 (m) , 1840 (broad, vs), 1460 (m) , 1420 (w) , 1390 (s), 1375 (s), 1320 (w) , 1270 (s) , 1250 (w) , 1165 (s), 1130 (s) , 1110 (s), 1030 (w), 1005 (w), 955 (s) , 930 (s) , 890 (m) , 800 (m) , 770 (w) , and 640 (m) cm. The NMR spectrum (CCl.) showed a singlet at 8.61 T and an ^^2 pattern centered at 7.76 T with area ratios of 3:2 respectively.

PAGE 77

68 Pre paration of diethyl 4-methyl-4-pentenylmalonate A modification of a procedure in Vogel's Textbook of Practical Organic Chemistry (57) for preparing alkyl-substituted malonic esters was used. The reaction was carried out in a hood. To 2.7 g. (0.115 mole) of sodium chunks in a dry 300-ml. three-necked, round -bottc-.ed flask, equipped with mechanical stirrer, condenser and drying tube, dropping funnel, and nitrogen inlet, was added 85 ml. of ethanol at such a rate as to maintain gentle refluxing. Then 19.5 g. (0.12 mole) of diethyl malonate was added slowly, followed by addition of 29.7 g. (0.115 mole) of crude 4-methyl-4-pentenyl p-toluenesulfonate. A slow nitrogen purge was maintained throughout the reaction. The mixture was refluxed until it was almost neutral. Enough water was added to dissolve the precipitated sodium p-toluenesulfonate, and the aqueous layer was extracted overnight with ether. The ether solution was dried over anhydrous magnesium sulfate and filtered. Removal of the ether on a rotary evaporator left 31.0 g. of liquid xjhose MR spectrum (CCl^) indicated the mixture to contain mainly diethyl 4-methyl-4-pentenylmalonate with a little bit of diethyl malonate and 4-methyl-4-pentenyl p-toluenesulfonate. The infrared spectrum (neat) of the product showed absorption bands at 3100 (w) , 2990 (s), 2940 (s) , 2890 (w) , 1740 (broad, vs) , 1650 (w), 1450 (s), 1370 (s), 1330-1140 (broad, s) , 1100 (m) , 1030 (s) , 890 (m) , 860 (w) , 810 (w) , and 660 (w) cm." . The NMR spectrum (CCl.) showed a triplet centered at 8.75 T, a singlet with hyperfine splitting at 8.33 T overlapping a multiplet centered at 8.25 T, a triplet with hyperfine splitting centered at 8.00 T, a triplet centered at 6.72 T, a quartet centered at 5.90 T, a singlet with hyperfine splitting centered at 5.40 T with area ratios

PAGE 78

69 of 6:3:2:2:1:4:2 respectively. This previously unreported compound was identified by its infrared and NMR spectra. It was not purified, thus no elemental analysis, refractive index, nor boiling point was reported. Preparation of 6-methyl-6-heptenoic acid The procedure used previously for the hydrolysis and decarboxylation of diethyl methallylmalonate was employed (55). A great deal of trouble was encountered in the decarboxylation step. The problem arose because acidifying the solution sufficiently to bring about decarboxylation with heating resulted in partial isomerization of the carbon-carbon double bond. A small amount of 6-methyl-6-heptenoic acid was obtained by neutralizing the basic solution with solid ammonium chloride to pH of 6 and refluxing the mixture for one week. The product was extracted from the reaction mixture with ether. The ether solution was dried over anhydrous magnesium sulfate and the ether was removed by rotary evaporation. Only a few drops of product remained. The NMR spectrum (CCl,) showed a multiplet centered at 8.41 T overlapping a singlet at 8.30 T, a triplet with hyperfine splitting centered at 7.97 T, a triplet with hyperfine splitting centered at 7.66 T, a singlet with hyperfine splitting at 5.33 T, and a singlet at -0.85 T with area ratios of 4:3:2:2:2:1 respectively. The synthesis of 6-methyl-6-heptenoic acid was reported by Julia and Gueremy (58). It was prepared in very low yield from a series of three reactions starting with cyclopentanone and isopropenyl Grignard. The 6-methyl-6-heptenoic acid was distilled from acidic solution, and its identification was based on elemental analysis and the infrared

PAGE 79

70 spectrum. They reported the presence of absorption at 890 cm. , characteristic of a gem -disubstituted alkene. The I'fl'IR spectrum was the only basis for identification of the 6methyl-6-heptenoic acid prepared in this experiment. Preparation of 3-methallyloxypropionitrile This synthesis was modeled after a procedure reported by Koelsch (59). The reaction was carried out in a hood. A dry 500-ml. threenecked, round-bo :tomed flask was fitted with mechanical stirrer, condenser, and drying tube. A mixture of 53.0 g. (1.0 mole) of acrylonitrile and 85.7 g. (1.22 mole) of methallyl alcohol in the flask was treated with 25 drops of concentrated sodium ethoxide. The temperature was maintained around 30-35° for 4 hours. The mixture was neutralized with glacial acetic acid. Distillation through a 16-cm. Vigreux column yielded 105.3 g. (84 percent) of a clear, colorless liquid b.p. 10:-( '21 mm. (lit. (60) b.p. 209-212/760 mm.), n^^ 1.4361. The product was shown to have no detectable impurities by gas chromatography . The infrared spectrum (neat) of the product showed absorption bands at 3090 (m), 2990 (s) , 2930 (s) , 2880 (s) , 2250 (m) , 1655 (m) , 1450 (s) , 1410 (m), 1375 (ra) , 1350 (m) , 1330 (w) , 1250 (w) , 1220 (w) , 1110 (s) , 1000 (m), 900 (s) cm."""-. The N14R spectrum (CCl.) showed a singlet with hyperfine splitting at 8.27 T, a triplet centered at 7.49 T, a triplet centered at 6.43 T, a singlet at 6.10 T, and a singlet with hyperfine splitting centered at 5.10 T with area ratios of 3:2:2:2:2 respectively. This compound, which was previously reported in the literature (60) ,

PAGE 80

71 was identified by its infrared and NMR spectra. Attempted preparation of 3-giethallyloxypropionic acid A procedure by Christian and Hixon (41) for the hydrolysis of nitriles to acids was followed. The reaction was carried out in a hood. A mixture of 25.2 g. (0.21 mole) of 3-methallyloxypropionitrile and 32.7 ml. of concentrated hydrochloric acid was stirred in a 250-ml. Erlenmeyer flask at 60-70° for 4 hours. The mixture was allowed to cool, and enough water was added to dissolve the precipitate. The organic layer was separated and the aqueous layer was extracted overnight with ether. The combined organic layers were dried over anhydrous magnesium sulfate. An attempted distillation proved to be messy and the product decomposed and polymerized. The vinyl protons had all disappeared according to the infrared and NtlR spectra. Preparation of 3-methallyloxypropionic acid A procedure reported by Lamb and coworkers (45) was followed. The reaction was carried out in a hood. A 300-ml. round-bottomed flask, equipped with magnetic stirrer, condenser, and drying tube, was charged with 36.0 g. (0.5 mole) of 3-propiolactone and 50.4 g. (0.7 mole) of methallyl alcohol. This was heated at 68-70° for 26 hours. The mixture was distilled through a 16-cm. Vigreux column, yielding 33.0 g. (46 percent) of a clear, colorless liquid b.p. 79-80/0.09 mm., n^ 1.4436. The infrared spectrum (neat) of the product showed absorption bands at 3200-3000 (broad, m) , 3100 (w) , 2990 (s) , 2930 (s) , 2680 (broad, m) , 1730 (vs), 1655 (m), 1430 (m) , 1375 (w) , 1350 (w) , 1260 (broad, m) , 1185 (s), 1100 (s), 1075 (w), 1050 (w) , 1010 (w) , and 900 (s) cm. .

PAGE 81

72 The NMR spectrum (CCl.) showed a singlet with hyperfine splitting at 8.30 T, a triplet centered at 7.43 T, a triplet centered at 6.38 T, a singlet at 6.17 T, a singlet with hyperfine splitting centered at 5.14 T, and a singlet at -1.17 T with area ratios of 3:2:2:2:2:1 respectively. This compound was reported previously in a patent (61). It was identified by its infrared and NMR spectra. Preparation of 2-methallyloxyethanol A procedure reported by Hurd and Pollack (62) was followed. The reaction was car^ried out in a hood. A 300-ml. three-necked, roundbottomed flask, equipped with mechanical stirrer, dropping funnel, condenser, and drying tube, was charged with 79 ml. of ethylene glycol, and 11.7 g. (0.5 mole) of sodium (small chunks) was added slowly. The solution was heated with stirring on a steam bath while 46.4 g. (0.51 mole) of methallyl chloride was slowly added. The mixture was heated an additional two hours, then allowed to cool. The sodium chloride was filtered off. The liquid was distilled through a 60-cm. spinningband column with Teflon band to yield 45.5 g. (82 percent) of 2-methallyloxyethanol b.p. 163/760 mm. (lit. (63) b.p. 172/760 mm.), n^^ 1.4386 (lit. (63) n^^ 1.4372). The infrared spectrum (neat) of the product showed absorption bands at 3420 (broad, s) , 3090 (w) , 2990 (m) , 2930 (s) , 2870 (s) , 1655 (m), 1450 (m), 1375 (w) , 1350 (w) , 1110 (s) , 1050 (s), 1010 (w) , and 890 (s) cm.""'". The NMR spectrum (CCl, ) showed a singlet with hyperfine splitting at 8.28 T, a singlet at 7.03 T, an A-B^ system centered at 6.49 T, a

PAGE 82

73 singlet at 6.13 T, and a broad singlet with hyperfine splitting centered at 5.12 T with area ratios of 3:1:4:2:2 respectively. Preparation of 2-bromoethyl methallyl ether A procedure reported by Hurd and Pollack (62) i^as followed. The reaction was carried out in a hood. A 200-nil. three-necked, roundbottomed flask, fitted with mechanical stirrer, thermometer, dropping funnel, and drying tube, was charged with 27.1 g. (O.IC mole) of phosphorus tribromide and cooled to -10° in an ice-methane 1 bath. A mixture of 29.0 g. (0.25 mole) of 2-methallyloxyethanol and 5.0 g. (0.06 mole) of pyridine was added dropwise while the reaction mixture was stirred at -10°. Addition took slightly more than one hour. The mixture was stirred an additional hour while returning to room temperature. The next day the reaction mixture was distilled at 25 mm. , and the oil bath temperature was not allowed to exceed 143°. The solid residue was discarded. The liquid was washed twice with 20 ml. of dilute sodium hydroxide, twice with 20 ml. of dilute sulfuric acid, and once with 20 ml. of water. It was then dried over anhydrous sodium sulfate and filtered. Distillation and redistillation through a 16-cm. Vigreux column yielded 4.8 g. (11 percent) of a clear, colorless liquid b.p. 75-6/25 mm., n^^ 1.4655. The infrared spectrum (neat) showed absorption bands at 3100 (m) , 2990 (s), 2930 (s) , 2870 (s) , 2760 (w) , 1660 (m) , 1450 (s) , 1425 (m) , 1375 (m), 1350 (m) , 1280 (s) , 1260 (w) , 1220 (w) , 1190 (w) , 1110 (s) , 1040 (m), 990 (m) , 900 (s), 820 (w) , and 670 (w) cm." . The NMR spectrum (CCl.) shovjed a singlet with hyperfine splitting at 8.28 T, an A B^ pattern centered at 6.50 T, a singlet at 6.13 T, and

PAGE 83

74 a singlet with hyperfine splitting centered at 5.12 T x^rith area ratios of 3:4:2:2 respectively. Anal. Calcd. for C,lL,BrO: C, 40.24; H, 6.19; Br, 44.63. Found: C, 40.36; K, 6.08; Br, 44.99. Preparation of 2-methallyloxyethyl p-toluenesulfonate The procedure of Ansell (39), which was previously used to prepare 4-inethyl-4-pentenyl p-toluenesulfonate, was used once again. A yield of 39.0 g. (74 percent) of a viscous liquid was obtained. The infrared spectrum (neat) of the product shovjed absorption bands at 3090 (w) , 2990 (m) , 2930 (m) , 2870 (m) , 1660 (w) , 1600 (m) , 1450 (m), 1360 (s), 1290 (w) , 1250 (w) , 1190 (s) , 1175 (s) , 1100 (s) , 1020 (s), 915 (s), 815 (s) , 775 (s) , and 660 (s) cm."-*-. The NMR spectrum (CCl, showed a singlet with hyperfine splitting at 8.35 T, a singlet at 7.58 T, a triplet centered at 6.50 T, a singlet at 6.23 T, a triplet centered at 5.93 T, a singlet with hyperfine splitting centered at 5.20 T, a doublet centered at 2.75 T, and a doublet centered at 2.29 T with area ratios of 3:3:2:2:2:2:2 respectively. This previously unreported compound was identified by its infrared and NMR spectra. Preparation of 2-bromoethyl methallyl ether A modification of a procedure reported by Wilt and coworkers (44) was employed. A solution of 36.0 g. (0.41 mole) of lithium bromide in 1100 ml. of acetone was prepared in a 3-1. round-bottomed flask, equipped with magnetic stirrer, condenser, and drying tube. The 39.0 g. (0.14 mole) of crude 2-methallyloxyethyl p-toluenesulfonate was added, and

PAGE 84

75 the solution was stirred and refluxed 17 hours. The flocculent precipitate was filtered off, and the acetone was removed on a rotary evaporator using lukewarm water. Attempts to take up the product in 100 ml. of ether resulted in two layers. l-Jhen 50 ml. of water was added, it was taken up in the bottom laj'er; thus it was extracted with ether over the weekend. The combined organic phases were dried over anhydrous potassium carbonate. The ether was removed on a rotary evaporator. Distillation through a 16-cm. Vigreux column gave four fractions, none of which was pure. Purification was accomplished by column chromatography, using a 60-cm. column containing 75 g. of silica gel. The solvent system used was composed of 90 percent petroleum ether and 10 percent ethyl ether. The chromatographed product was distilled, yielding 17.1 g. (68 percent) of a clear, colorless liquid b.p. 70-1/20 mm., n^ 1.4651. The product was shown to be pure by gas chromatography. The refractive index, infrared spectrum, and Nl'IR spectrum were identical to those for the 2-bromoethyl methallyl ether previously prepared. Preparation of ethyl methallyl ether Ethyl methallyl ether was prepared according to the procedure of Tamele and coworkers (64). A mixture of 90.6 g. (1.0 mole) of methallyl chloride, 92.0 g. (2.0 moles) of ethanol, and 96.0 g. (1.2 moles) of 50 percent aqueous sodium hydroxide in a 1000-ml. round-bottomed flask was refluxed eight hours, then it was allowed to stand overnight. Addition of 100 ml. of water dissolved the precipitate, and the organic phase was separated and washed with 50-ml. portions of water until all the ethanol was removed — five washings were required. The ether was dried over calcium chloride and filtered. The product was distilled through

PAGE 85

76 a 60-cin. spinning-band column with Teflon band to yield 55.0 g. (55 percent) of a clear, colorless liquid b.p. 86.5/760 mm. (lit. (64) b.p. 25 20 84.8-86.8/760 ram.), n^ 1.3970 (lit. (64) n^ 1.4067). Gas chromatographic analysis using several different columns showed the product to be composed of a single component. The infrared spectrum (neat) of the product showed absorption bands at 3C (m) , 2990 (s) , 2940 (m) , 2880 (s) , 1660 (m) , 1450 (m) , 1380 (m), 1350 (m) , 1260 (w) , 1170 (w) , 1120 (s) , 1100 (s) , 1010 (w) , and 895 (s) cm.""^. The MR spectrum (CCl.) showed a triplet centered at 8.84 T, a singlet with hyperfine splitting at 8.30 T, a quartet centered at 6.62 T, a singlet at 6.23 T, and a broad singlet with hyperfine splitting centered at 5.18 T with area ratios of 3:3:2:2:2 respectively. Preparation of diethyl 2-methylglutarate A procedure in Organic Syntheses (65) for the esterif ication of dicarboxylic acids was used. The reaction was carried out in a hood. Into a 500-ml. round-bottomed flask were put 48.7 g. (0.33 mole) of 2-methylglutaric acid, 120 ml. (2 mole) of absolute ethanol, 60 ml. of toluene, and 0.5 g. of concentrated sulfuric acid. The flask was set up for downward distillation and heated with an oil bath. An azeotropic mixture of alcohol, toluene, and water began to distill over at 76°. Distillation was continued until the temperature reached 78°. The distillate was collected in a 250-ml. Erlenmeyer flask containing 50 g. of anhydrous potassium carbonate. It was shaken, filtered, and returned to the reaction mixture. The flask was heated again, and 115 ml. of liquid, again coming over at 76°, was collected and discarded. The

PAGE 86

77 remaining liquid was distilled, yielding 64.1 g. (96 percent) of a clear, colorless, volatile product, b..p. 128/20 mm. (lit. (66) b.p. 125/20ram.), n^^ 1.4230 (lit. (66) n^° 1.4265). The infrared spectrum (neat) of the product showed absorption bands at 2990 (s), 2950 (s) , 2920 (m) , 2890 (m) , 1730 (vs), 1460 (s) , 1420 (w) , 1375 (s), 1320 (w), 1300 (w) , 1250 (s) , 1180 (broad, s), 1120 (s) , 1095 (s), 1065 (m), 1025 (s) , 930 (w) , 850 (m) , 780 (w) , and 755 (w) cm."-*-. The NMR spectrum (CCl.) showed a doublet centered at 8.87 T overlapping a triplet centered at 8.77 x, a multiplet centered at 8.20 T, a triplet centered at 7.75 T overlapping a multiplet centered at 7.65 T, and a quartet centered at 5.92 T with area ratios of 3:6:2:2:1:4 respectively. Preparation of diethyl 2 ,2-dimethylsuccinate The procedure used was the same as for the preparation of diethyl 2-methylglutarate, the same quantities being used because the two diacids have the same molecular weight. Distillation yielded 57.0 g. of a clear, colorless liquid, b.p. 115-118/20 mm. (lit. (67) b.p. 101/15 mm.), n^^ 1.4244 (lit. (67) n^° 1.4233). The properties of the product were not exactly those expected for diethyl 2 ,2-dimethylsuccinate. The boiling point and refractive index differ from the reported values. The carbonyl region of the infrared spectrum showed an absorption at 1735 cm. as expected for the ester carbonyl and two other absorptions at 1790 and 1860 cm. , characteristic of an anhydride. The NMR spectrum (CCl.) also indicated the presence of a mixture of diethyl 2 ,2-dimethylsuccinate and 2 ,2-dimethylsuccinic anhydride in a 2:1 ratio. The absorptions for the anhydride

PAGE 87

appear as a singlet at 8.58 T and a singlet at 7.23 x with area ratios of 3:1 respectively. The mixture of diester and anhydride was combined with 120 ml. of ethanol, 60 ml. of toluene, and 0.5 g. of concentrated sulfuric acid in a 500-ml. round-bottomed flask, set up for downward distillation. The mixture was heated and 160 ml. of a mixture of ethanol and toluene was collected. The remaining liquid was washed with 50 ml. of saturated sodium bicarbonate and dried over anhydrous magnesium sulfate. Filtration through a 16-cm. Vigreux column yielded 54.0 g. (81 percent) of a clear, colorless liquid, b.p. 107/15 mm. (lit. (67) b.p. 101/15 mm.), nj^ 1.4201 (lit. (67) nj° 1.4233). The infrared spectrum (neat) of the product showed absorption bands at 2990 (s), 2950 (s) , 2920 (m) , 2890 (m) , 1735 (vs) , 1475 (s) , 1450 (m) , 1410 (w), 1390 (m), 1370 (s) , 1345 (s) , 1305 (s) , 1255 (s) , 1210 (s) , 1180 (s), 1130 (s), 1100 (m), 1030 (s) , 950 (w) , 860 (m) , and 770 (m) -1 cm. The NMR spectrum (CCl.) showed a singlet at 8.78 T overlapping a triplet centered at 8.77 T, a singlet at 7.53 T, and a quartet split into doublets centered at 5.92 x with area ratios of 3:3:1:2 respectively. Preparation of 2-methyl-l ,5-pentanediol A procedure in Vogel's Textbook of Practical Organic Chemistry (68) for reducing esters to alcohols was followed. The reaction was carried out in a hood. Into a 2000-ml. three-necked, round-bottomed flask, fitted with mechanical stirrer, dropping funnel, condenser, and drying tube, was put 10.5 g. (0.276 mole) of powdered lithium aluminum hydride. The lithium aluminum hydride was washed into the flask wi a 300 ml. of

PAGE 88

79 anhydrous ether. After 10 minutes of stirring, a solution of 50 g. (0.25 mole) of diethyl 2-methylglutarate in 150 ml. of ether was added at such a rate as to maintain gentle refluxing of the ether. The slurry was stirred an additional 15 minutes, followed by cautious addition of 50 ml. of water to decompose the excess lithium aluminum hydride, the slurry turning from gray to white. The suspension was filtered, and the ether phase was dried over anhydrous magnesium sulfate. The white sludge remaining in the filter was dissolved in 400 ml. of 20 percent sulfuric acid, and this solution was extracted continuously with ether overnight. The resulting ether solution was dried over anhydrous magnesium sulfate. The ether solutions were filtered, combined, and stripped of ether on a rotary evaporator. The infrared spectrum of the crude product showed absorptions in both the hydroxyl and carbonyl regions. Since it appeared conversion to the diol was incomplete, the crude product was recycled through the procedure with the addition of a one-hour reflux before decomposing the excess lithium aluminum hydride. The continuous extraction was allowed to run for three days, because all of the diol was not removed overnight. The infrared spectrum of the crude product showed no absorption in the carbonyl region. Distillation through a 16-cm. Vigreux column yielded 19.9 g. (68 percent) of a clear, colorless 21 liquid, b.p. 135-6/6 mm. (lit. (66) b.p. 140/20 mm.) n,^^ 1.4524 (lit. (66) 20 n^^ 1.4545). The infrared spectrum (neat) of the product showed absorptions at 3650-3050 (broad, vs) , 2990 (m) , 2940 (broad, s) , 2880 (s) , 1460 (s) , 1380 (m), 1180 (w), 1100 Cm), 1040 (broad, s), 935 (w) , and 895 (m) cm." . The NMR spectrum (DO) sho^^^ed a doublet centered at 9.07 T, a multiplet centered at 8.48 T, a doublet centered at 6.52 T overlapping a

PAGE 89

80 triplet centered at 6.38 x with area ratios of 3:5:2:2 respectively. Preparation of 2,2-dimethyl-l,4-butanediol The procedure for preparing 2-methyl-l,5-pentanediol (68) was used with the following modifications: The amount of lithium aluminum hydride was increased to 14.0 g. (0.37 mole), and 50 g. (0.25 mole) of diethyl 2 ,2-dimethylsuccinate was used instead of the diethyl 2-methylglutarate in the previous experiment. Three 50-ml. portions of ether were added when the mixture became viscous. The reaction mixture was stirred two hours at room temperature and one hour under reflux. The continuous extraction was complete after one day. The infrared spectrum of the crude product shovjed a small amount of carbonyl absorption. Distillation through a 60-cm. spinning-band column with Teflon band yielded 3.0 g. of starting material and 12.3 g. (42 percent) of a clear, colorless liquid, b.p. 163/15 mm. (lit. (67) 25 20 b.p. 117/8 mm.), n^ 1.4436 (lit. (67) n^ 1.4513). The infrared spectrum (neat) of the product showed absorption bands at 3650-3050 (broad, vs) , 2990 (m) , 2960 (s) , 2880 (s) , 1465 (s) , 1390 (m), 1360 (m), 1150 (w) , 1040 (broad, s) , 980 (m) , and 900 (w) cm."""". The NI'R spectrum (DO) showed a singlet at 9.08 T, a triplet centered at 8.44 T, a singlet at 6.66 T, and a triplet centered at 6.28 T with area ratios of 3:1:1:1 respectively. Preparation of 3-methyltetrahydropyran Into a pressure bottle were put 30 g. of 60 percent sulfuric acid and 10.0 g. (0.085 mole) of 2-methyl-l ,5-pentanediol. The liquids were heated to 100° for one hour. Then they were cooled overnight. The

PAGE 90

81 bottle was opened and the contents were diluted with 30 ml. of water and distilled. The product and water azeotroped over, and 7.5 g. of crude product was obtained. The product was dried over 3A molecular sieves. Preliminary infrared, NMR, and gas chromatographic analyses on the crude product showed it to be a binary mixture which was not easily separated by distillation. A careful distillation through a 50-cm. spinning-band column with teflon band afforded some separation, yielding 3.8 g. (45 percent) of a clear, colorless, volatile liquid, b.p. 108/760 mm. (lit. (66) b.p. 109/733 mm.), nj^ 1.4194 (lit. (66) n^° 1.4210). The identity of the impurity was indicated by the NMR spectra of two product mixtures in different proportions. The positions of absorptions of the NMR (CCl.) of the impurity were a singlet at 8.85 T, a multiplet centered at 8.25 T, and a triplet centered at 6.2 T with area ratios of 3:2:1 respectively. This is the spectrum expected for 2,2-dimethyltetrahydrofuran, which could have resulted from a 1,2-hydride shift to form a tertiary carbonium ion prior to cyclization. Gas chromatographic analysis showed the 3-methyltetrahydropyran to be greater than 98 percent pure, and a small portion was purified essentially free of impurity by another spinning-band distillation for later use in quantitative analyses. The infrared spectrum (neat) of 3-methyltetrahydropyran showed absorption peaks at 2960 (s) , 2940 (s) , 2880 (w) , 2850 (s) , 2760 (w) , 2740 (w), 1460 (m), 1390 (w) , 1300 (w) , 1275 (w) , 1225 (m) , 1185 (m) , 1095 (s), 1035 (m), 990 (w) , 970 (m) , 890 (m) , and 860 (m) cm.""*-. The NMR spectrum (CCl,) showed a doublet centered at 9.20 T, a broad multiplet centered at 8.5 x, a multiplet centered at 7.0 x, and a multiplet centered at 6.3 x with area ratios of 3:5:2:2 respectively.

PAGE 91

82 Preparation of 3,3-dimethyltetrahydrofuran The same general technique as for the preceding synthesis was used. From 6.0 g. (0.051 mole) of 2 ,2-dimethyl-l,4-butanediol was obtained 4.8 g. (94 percent) of a clear, colorless, volarile liquid, b.p. 99/760 9S 20 mm. (lit. (67) b.p. 98-9/750 mm.), n"^ 1.4102 (lit. (67) n^ 1.4121). Gas chromatographic analysis showed a single peak with no impurity. The infrared spectrum (neat) of the product showed absorption peaks at 2970 (s), 2940 (m) , 2880 (s) , 1460 (m) , 1390 (w) , 1370 (m) , 1165 (w) , 1065 (s), 975 (w), 920 (w) , and 895 (m) cm.""*". The NMR spectrum (CCl,) showed a singlet at 8.93 T, a triplet centered at 8.38 T, a singlet at 6.70 T, and a triplet centered at 6.25 T with area ratios of 3:1:1:1 respectively. Preparation of tributyltin hydride Tributyltin hydride was prepared according to the procedure of Kuivila and Buemel.(42). The compound was obtained in 75 percent yield from the reaction of tributyltin chloride with lithium aluminum hydride (0.41 mole). The liquid was distilled at pressures of 0.33 to 0.5 mm. and boiled around 75-80°. The product was refrigerated in a flask purged with nitrogen and sealed with a septum cap. Preparation of 2-(2-phenylallyloxy)ethanol The procedure of Hurd and Pollack (62) used previously for the preparation of 2-methallyloxyethanol was employed. The reaction was carried out on a half molar scale, using 11.7 g. of sodium, 76.3 g. of 2-phenylallyl chloride, and 79 ml. of ethylene glycol. After the sodium chloride was filtered off, the reaction mixture was diluted with three

PAGE 92

83 parts of water and extracted with ether for four days. The ether solution was dried over anhydrous sodium sulfate and filtered. The ether was removed on a rotary evaporator. The remaining liquid was distilled through a 16-cm. Vigreux column to yield 59.3 g. (67 percent) of a 27 clear, colorless liquid, b.p. 109-111/0.5 ram., n 1.5456. The infrared spectrum (neat) of the product showed absorption bands at 3430 (broad, s) , 3100 (m) , 3070 (m) , 3040 (m) , 2940 (s) , 2870 (s) , 1630 (m), 1600 (w) , 1575 (m) , 1495 (s) , 1445 (m) , 1410 (v/) , 1360 (m) , 1310 (w), 1120 (s), 1095 (s), 1065 (s) , 1030 (s) , 905 (s) , 830 (w) , 780 (s), and 710 (s) cm."-^. The NMR spectrum (CCl.) showed a singlet at 7.80 T, a multiplet centered at 6.48 T, a singlet at 5.67 T, a doublet centered at 4.73 T, a doublet centered at 4.57 T, and a multiplet centered at 2.71 T with area ratios of 1:4:2:1:1:5 respectively. Anal . Calcd. for C^^H^^02: C, 74.13; H, 7.92. Found: C, 73.96; H, 8.12. Preparation of 2-(2-phenylallyloxy)ethyl p-toluenesulfonate The procedure of Ansell (39) used previously for the preparation of 2-methallyloxyethyl p-toluenesulfonate was used with modifications. The reaction mixture was left at room temperature overnight before being worked up. The product obtained was shown by thin layer chromatography to consist of two major components. They were separated by column chromatography using silica gel and a solvent system of 85 percent petroleum ether (65-110) and 15 percent ethyl ether. The two separated components were identified as the expected p-toluenesulfonate and 2phenylallyl 2-chloroethyl ether. Formation of the chloride is not

PAGE 93

84 surprising. It has been stated in Fieser and Fieser (69) that in the presence of pyridinium chloride at room temperature primary tosylates are converted to clilorides. Thus the reaction mixture should not be allowed to exceed 20°. The chloride was distilled through a 16-cm. Vigreux column, yield25 ing 5.2 g. of a clear, colorless liquid b.p. 88/0.25 mm., n 1.5428. The infrared spectrum (neat) showed absorption bands at 3100 (m) , 3070 (m), 3040 (m) , 2970 (m) , 2870 (s) , 1630 (m) , 1600 (w) , 1575 (m) , 1495 (s), 1445 (m), 1430 (m) , 1370 (w) , 1360 (w) , 1300 (s) , 1250 (w) , 1200 (m), 1120 (s), 1100 (s) , 1055 (s) , 1030 (s) , 990 (w) , 910 (s) , 845 (w), 780 (s), 750 (m) , 705 (s) , and 665 (s) cm." . The NMR spectrum (CCl.) showed a multiplet centered at 6.24 T, a singlet at 5.65 T, a doublet centered at 4.73 T, a doublet centered at 4.55 T, and a multiplet centered at 2.70 T with area ratios of 4:2:1:1:5 respectively. Ana l. Calcd. for C^^H CIO: C, 67.17; H, 6.66; CI, 18.03. Found: C, 67.23; H, 6.69; CI, 17.86. The crude p-toluenesulfonate was obtained in a yield of 6.3 g. (19 percent). Another procedure (70) provided a good yield of the desired 2-(2phenylallyloxy) ethyl p-toluenesulfonate from the alcohol. The reaction was carried out in a hood. A 500-ml. three-necked, round-bottomed flask, equipped with a mechanical stirrer and a thermometer, was charged with 17.8 g. (0.10 mole) of 2-(2-phenylallyloxy)ethanol and 31.6 g. (0.40 mole) of pyridine. The liquid was cooled to 10° followed by addition of 21.5 g. (0.11 mole) of p-toluenesulfonyl chloride at such a rate as to maintain the temperature below 20°. The mixture was stirred three and

PAGE 94

85 a half hours with the temperature kept below 20". The contents were cooled with an ice-water bath, and a solution of 50 ml. of concentrated hydrocliloric acid and 175 ml» of water was added slowly. The tosylate was taken up in 200 ml. of ether, and the aqueous phase was washed with three 50-ml. portions of ether. The ether phases were combined and about half of the ether was evaporated off. The solution was washed with 50 ml. of water, twice with 50 ml. of saturated sodium bicarbonate solution, and again vrith 50 ml. of water. It Vvras dried over anhydrous magnesium sulfate, filtered, and the ether removed by rotary evaporation. The crude tosylate was obtained in 89 percent yield (29.7 g.). The infrared spectrum (neat) of the tosylate showed absorption bands at 3100 (m) , 3070 (m) , 3040 (m) , 2970 (m) , 2940 (m) , 2880 (s) , 1630 (m), 1600 (s) , 1575 (m) , 1495 (s) , 1445 (s) , 1400 (m) , 1360 (s) , 1310 (m), 1290 (m), 1240 (w) , 1210 (w) , 1190 (s) , 1175 (s) , 1120 (s), 1090 (s), 1020 (s), 910 (s), 810 (s), 775 (s), 710 (s) , 690 (w) , and 660 (s) cm. The NMR spectrum (CCl,) showed a singlet at 7.60 T, an A-B system centered at 6.17 T, a singlet at 5.73 T, a doublet centered at 4.80 T, a doublet centered at 4.59 T, a doublet centered at 2.79 T overlapping a multiplet centered at 2.72 x, and a doublet centered at 2.31 T with area ratios of 3:4:2:1:1:2:5:2 respectively. Preparation of 2-bromoethyl 2-phenylallyl ether A modification of the procedure of Wilt and coworkers (44) used previously for preparing 2-bromoethyl methallyl ether from the tosylate was again employed. A solution of 29.5 g. (0.089 mole) of 2-(2-phenylallyloxy) ethyl p-toluenesulfonate and 25 g. (0.29) of lithium bromide

PAGE 95

86 in 500 ml. of acetone was refluxed for three hours in a 1000-ml. roundbottomed flask, equipped with magnetic stirrer and condenser. The lithium tosylate was filtered from the cool solution, and the acetone was removed on a rotary evaporator. As before, the ether failed to dissolve the remaining liquid, so it was diluted with 50 ml. of water and extracted 36 hours with ether. The ether solution was washed twice with 50-ml. portions of water and twice with 50 ml. of saturated sodium bicarbonate solution and dried over anhydrous magnesium sulfate. Filtration and removal of the ether left an oil which was distilled through a 16-cm. Vigreux column, yielding 18.9 g. (88 percent) of a clear, colorless liquid, b.p. 98/0.2 mm., n^^ 1.5593. The infrared spectrum (neat) of the product showed absorption bands at 3100 (m), 3070 (m) , 3040 (m) , 2980 (m) , 2870 (s) , 2800 (w) , 2760 (w) , 1950 (w), 1885 (w), 1825 (w) , 1750 (w) , 1630 (m) , 1600 (m) , 1575 (m) , 1495 (s), 1460 (m), 1445 (s) , 1420 (w) , 1390 (m) , 1370 (m) , 1360 (m) , 1305 (m), 1280 (s) , 1230 (m) , 1220 (m) , 1185 (m) , 1125 (s) , 1095 (s) , 1040 (s), 1030 (s), 1010 (m), 980 (m) , 945 (m) , 910 (s) , 825 (w) , 805 (w), 780 (s), 745 (w), and 710 (s) cm."-*-. The NMR spectrum (CCl.) showed an A„B„ system centered at 6.47 T, a singlet at 5.65 T, a doublet centered at 4.73 T, a doublet centered at 4.55 T, and a multiplet centered at 2.70 T with area ratios of 4:2:1:1:5 respectively. Anal. Calcd. for C^^H^^BrO: C, 54.79; H, 5.43; Br, 33.14. Found: C, 54.85; H, 5.51; Br, 32.97. Preparation of ethyl 2-phenylallyl ether A procedure by Baucom (25) was used. The reaction was carried out

PAGE 96

»7 in a hoodA 500-ml. three-necked, round-bottomed flask was equipped with a drying tube, magnetic stirrer, dropping funnel, and nitrogen inlet. With nitrogen at a high flow rate the system was flamed out thoroughly. VJhen it had cooled, 10.6 g. (0.22 mole) of 50 percent sodium hydride was added. The sodium hydride was washed twice with 50-ml. portions of petroleum ether (20-40). After each washing the sodium hydride was allowed to settle to the bottom and the ether was decanted while the nitrogen flow rate was high. Cautiously 250 ml. of dry dimethyl sulfoxide was added. Then 9.2 g. (0.20 mole) of ethanol was added dropwise. A slow nitrogen flow was maintained while the reaction was stirred overnight. Then 30.5 g. (0.20 mole) of 2-phenylallyl chloride was added dropwise to the alkoxide. The reaction was allowed to stir one day after all of the chloride had been added. The contents were then poured into an equal volume of water and extracted with four 150-ml. portions of pentane. The pentane extracts were combined and washed three times with 85 ml. of water. The pentane solution was dried over anhydrous magnesium sulfate and filtered. The pentane was removed by distillation at atmospheric pressure. Analysis of the remaining liquid by gas chromatography indicated the presence of a good bit of the starting chloride, thus another 0.11 mole of the alkoxide was prepared and the liquid was added and heated one hour at 90°. This time the conversion was complete, but purification proved to be quite a problem. Column chromatography was used with silica gel as the solid phase and benzene as the solvent. The chromatographed product was distilled through a 16-cm. Vigreux column, yielding 14.7 g. (45 percent) of a clear, colorless liquid, b.p. 99.5/10 mm. (lit, (71) b.p. 96/10 mm.). n^^ 1.5215 (lit. (71) nj^ 1.5202)

PAGE 97

Gas chromatographic analysis showed the product to be completely free of impurities. The infrared spectrum (neat) of the product showed absorption bands at 3100 (m), 3070 (m) , 30A0 (m) , 2980 (s), 2940 (m) , 2870 (s), 2810 (w) , 1950 (w), 1885 (w), 1825 (w) , 1690 (w) , 1630 (m) , 1600 (m) , 1575 (m) , 1495 (s), 1410 (ra), 1390 (m) , 1375 (m) , 1355 (m) , 1305 (w) , 1280 (w) , 1265 (w), 1175 (m), 1160 (m) , 1125 (s) , 1090 (s) , 1025 (m) , 1005 (w) , 905 (s), 845 (w), 775 (s), 745 (w) , and 705 (s) cm.""'-. The NMR spectrum (CCl,) showed a triplet centered at 8.84 T, a quartet centered at 6.54 T, a singlet with hyperfine splitting at 5.77 x, a doublet centered at 4.77 T, a doublet centered at 4.62 T, and a multiplet centered at 2.75 T with area ratios of 3:2:2:1:1:5 respectively. Preparation of 3-phenyltetrahydropyran and 3-methyl-3-phenyltetrahydrofuran A solution of 6.03 g. (0.025 mole) of 2-bromoethyl 2-phenylallyl ether, 7.55 g. (0.026 mole) of tributyltin hydride, and 0.062 g. (0.0004 mole) of AIBN in 500 ml. of spectrograde benzene was sealed in an autoclave (one-liter capacity) . The contents were run through three freezethaw cycles using a vacuum system (0.02 mm.) and a dry-ice-isopropanol bath. Then it was placed in a preheated cavity and heated at 130° for three hours. The contents were allowed to cool, and most of the benzene was removed on a rotary evaporator. The remaining benzene was removed by distillation at atmospheric pressure. The residue was distilled at a pressure of 7.5 mm. to separate the more volatile products from the tributyltin bromide. The two major products, 3-phenyltetrahydropyran and 3-methyl-3-phenyltetrahydrofuran, v/ere isolated by means of

PAGE 98

89 preparative gas chromatography, using an F & M model seven-seventy-five prepmaster gas chromatograph with an 8-foot 3/4 -inch column of 20 percent SE-30. Pure 3-methyl-3-phenyltetrahydrofuran was isolated in 17y percent yield (0.7 gram). It is a clear, colorless liquid, b.p. 101/4 mm., 25 n^ 1.5271. The infrared spectrum (neat) showed absorption bands at 3100 (m) , 3070 (m), 3040 (m) , 2980 (s), 2940 (s) , 2880 (s) , 1950 (w) , 1880 (w) , 1810 (w), 1740 (w), 1600 (m) , 1580 (w) , 1495 (s) , 1445 (s) , 1380 (m) , 1360 (w), 1340 (w), 1290 (w) , 1260 (w) , 1150 (m) , 1090 (m) , 1075 (m) , 1055 (s), 1030 (s), 975 (m) , 940 (w) , 900 (s) , 760 (s) , 695 (s) , and 670 (w) cm."""". The NMR spectrum (CCl.) showed a singlet at 8.61 T, a multiplet centered at 7.92 x, a singlet at 6.22 T overlapping a triplet with hyperfine splitting centered at 6.09 x, and a singlet at 2.81 X with area ratios of 3:2:2:2:5 respectively. Anal . Calcd. for C K ,0: C, 81.44; H, 8.70. Found: C, 81.57; H, 8.70. Pure 3-phenyltetrahydropyran was isolated in 35 percent yield 25 (1.4 grams). It is a clear colorless liquid, b.p. 104/4 mm., n 1.5295 (lit. (72) n^^ 1.5267). The infrared spectrum (neat) showed absorption bands at 3100 (w) , 3070 (m), 3040 (m) , 2940 (s) , 2850 (s), 1950 (w) , 1880 (w) , 1800 (w) , 1600 (m), 1580 (w) , 1490 (m) , 1465 (m) , 1450 (m) , 1380 (w) , 1350 (w) , 1275 (m), 1210 (m) , 1175 (m) , 1140 (m) , 1100 (s), 1085 (s) , 1025 (s) , 990 (m), 960 (m) , 915 (m) , 900 (m) , 855 (m) , 845 (m) , 750 (s) , 695 (s) , and 630 (ra) cm.

PAGE 99

90 The NMR spectrum (CCl.) showed a multiplet between 7 . 87 T and 8. 48 T, a broad multiplet between 6.95 and 7.53 T , a triplet centered at 6.72 T, a broad doublet centered at 6.10 T, and a singlet at 2.84 T with area ratios of 4:1:2:2:5 respectively. This spectrum agrees well with the one reported in the literature (70) . Anal . Calcd. for C,^H^^O: C, 81.44; H, 8.70. Found: C, 81.54; H, 8.69. Preparation of triethyl 2-phenyl-2-carboxymethylsuccinate A 1-liter three-necked, round-bottomed flask, fitted with mechanical stirrer, dropping funnel, and condenser and drying tube, was charged with 300 ml. of absolute ethanol. To this was added 14.5 g. (0.63 mole) of sodium chunks at such a rate that the ethanol did not reflux too violently. When the sodium had completely reacted, 149 g. (0.63 mole) of diethyl phenylmalonate was added. Then 102.8 g (0.61 mole) of ethyl bromoacetate was added at such a rate that a gentle reflux was maintained. When the reaction was complete, the pH of the solution was about 7. The condenser was set for downward distillation and 250 ml. of ethanol x-zas distilled off. Just enough water to dissolve all of the sodium bromide (125 ml.) was added. The aqueous layer was separated and washed with three 15-ml. portions of ether. The combined ether solution was washed with 50 ml. of brine and dried over anhydrous magnesium sulfate. The solution was filtered and the ether was removed by rotary evaporation. The remaining liquid was distilled through a 40-cm. heated column packed with saddles, yielding 116.5 g. (59 percent) of a clear, colorless, viscous liquid, b.p. 164-5/1.2 mm., n 1.4904. The infrared spectrum (neat) of the product showed absorptions at

PAGE 100

91 3490 (s), 3030 (m) , 3000 (s) , 2950 (s) , 2920 (s), 2890 (m) , 1740 (broad, vs), 1600 (m), 1585 (v;) , 1500 (s) , 1470 (s) , 1450 (s) , 1410 (m) , 1390 (s), 1370 (s), 1345 (s), 1300-1150 (broad, vs) , 1095 (s) , 1060 (s) , 1030 (s), 915 (w), 860 (s), 805 (w) , 760 (m) , 730 (m) , 695 (s) , and 630 (m) cm. The NMR spectrum (CCl.) showed overlapping triplets centered at 8.78 and 8.80 T, a singlet at 6.82 T, overlapping quartets centered at 5.93 and 5.81 T, and a singlet at 2.75 T with area ratios of 9:2:6:5 respectively. Anal . Calcd. for C^^E^^O^: C, 63.34; H, 6.88. Found: C, 63.45; H, 6.96. The byproduct, ethyl ethoxyacetate, was obtained aa a clear, color25 less liquid, b.p. 34/2.2 mm. (lit. (73) b.p. 55/11 mm.), n^ 1.4031 25 1 (lit. (73) n 1.4019). This product was obtained in 332" percent yield (27 g.). The infrared spectrum (neat) of the product showed absorption bands at 2990 (s), 2950 (m) , 2920 (m) , 2890 (m) , 1755 (vs) , 1480 (m) , 1450 Cm), 1430 (m), 1400 (m) , 1380 (m) , 1370 (m) , 1350 (w) , 1275 (s) , 1200 (s) , 1180 (m), 1130 (vs), 1030 (s), 920 (w) , 850 (m) , 800 Cw) , and 710 (w) -1 cm. The NMR spectrum (CCl, ) showed overlapping triplets centered at 8.80 T and 8.72 T, a quartet centered at 6.48 T, a singlet at 6.08 T, and a quartet centered at 5.84 T with area ratios of 3:1:1:1 respectively. Preparation of 2-phenyl-2-hydroxymethyl-l,4-butanedlol A 2-liter three-necked, round-bottomed flask was equipped with mechanical stirrer, dropping funnel, and dry-ice condenser and drying

PAGE 101

92 tube. The system was flame dried and purged with nitrogen, followed by addition of 19.0 g. (0.5 mole) of lithium aluminum hydride (pulverized), which was washed in with 300 ml. of sodium-dried ether. The mixture was stirred half an hour, followed by addition of a solution of 89.0 g. (0.276 mole) of triethyl 2-phenyl— 2-carbox^^nethylsuccinate in 200 ml. of dry ether at such a rate as to maintain gentle refluxing. Addition took two hours, after which the suspension was stirred an hour and a half at room temperature then refluxed half an hour. Dropwise addition of 75 ml. of water consumed the excess hydride. The mixture was allowed to stand for two days, then it was filtered.. The ether solution was dried over anhydrous magnesium sulfate. The sludge was dissolved in 500 ml. of 20 percent sulfuric acid (cool). To the organic layer on top was added 75 ml. of ether and the aqueous layer was separated and extracted continuously with ether for two days . The three separate ether solutions were dried over anhydrous magnesium sulfate, filtered, and stripped of ether. The infrared spectra of all three crude products showed carbonyl absorptions. Thus they x^ere recycled with 12.0 g. (0.32 mole) of pulverized lithium aluminum hydride in 150 ml. of anhydrous ether in a 1-liter flask. The mixture was refluxed with warm water for an hour after addition was complete. Then it was stirred at room temperature one hour and allowed to sit overnight. The excess lithium aluminum hydride was destroyed by dropwise addition of 38 ml. of water. The sludge was filtered through a sintered glass funnel. The ether solution was dried over anhydrous magnesium sulfate. The sludge was dissolved in 200 ml. of 20 percent sulfuric acid. The organic phase was separated and the aqueous layer was extracted with six 25-ml. portions of ether. The combined ether solution was dried over anhydrous

PAGE 102

93 magnesium sulfate. Both ether solutions were filtered and stripped of ether. The residues showed very little carbonyl absorption in the infrared spectra. Further purification was not attempted because the literature reference claims the triol is not stable to distillation 0^31.. Thje yield of crude product was 47.1 g. (88 percent). The infrared spectrum (neat) of the product showed absorption bands at 3370 (broad, vs) , 3100 (m) , 3070 (m) , 3040 (m) , 2980 (s), 2940 (s) , 2900 (s), 1600 (m), 1580 (w) , 1495 (m) , 1445 (m) , 1380 (m) , 1270 (m) , 1045 (s), 1025 (s), 880 (m) , 760 (s), and 695 (s) cm.""''. The NMR spectrum (DO) showed a triplet centered at 8.1 x, a triplet centered at 6.5 T, a singlet at 6.1 T, and a singlet at 2.7 T with area ratios of 2:2:4:5 respectively. Preparation of 3-phenyl— 3-liydroxymethyltetrahydrofuraiv A 100-ml. three-necked, round-bottomed flask, equipped with mechanical stirrer, thermometer, and condenser, was charged with 26.0 g. (0.13 mole) of 2-phenyl-2-hydroxymethyl-l,4-butanediol. To this was added 67 g. (0.58 mole) of 85 percent phosphoric acid. The mixture was stirred and heated at 100° for 1.5 hours. Then it was allowed to cool and added along with 50 ml. of water to a 250-ml. separatory funnel. The organic product was taken up in 75 ml. of ether, and the aqueous portion was neutralized with sodium hydroxide then washed with six 25-ml. portions of ether. The ether layers were combined and dried over anhydrous magnesium sulfate, filtered, and stripped of solvent. Distillation of the crude liquid yielded 12.1 g. of a viscous liquid which was sho\im by infrared, NMR, and gas chromatography to be a mixture containing about 75 percent of the desired product. The product was redistilled

PAGE 103

94 through a 40-cm. heated column packed with saddles, yielding three fractions containing about 80 percent, 88 percent and 92 percent product according to the gas chromatographs. The yields of these three fractions were 6.0, 3.5, and 0.6 grams respectively. This calculates to be about 37 percent of the theoretical yield. Further purification was not attempted since it was planned to make the tosylate, vrhich could be purified and identified. The purest fraction was a clear, colorless, viscous liquid, b.p. 93-4/0.05 mm. (lit. (43) b.p. 119-20/0.6 im.) , n^^ 1.5496 (lit. (43) n^^ 1.5471). The infrared spectrum (neat) of the product showed absorption bands at 3600-3140 (broad, s) , 3100 (w) , 3070 (ra) , 3040 (m) , 2940 (s), 2880 (s), 1950 (w), 1880 (w), 1720 (w) , 1600 (m) , 1580 (w) , 1495 (s) , 1445 (s) , 1375 (w), 1340 (w), 1295 (w) , 1260 (w) , 1210 (w) , 1130 (s) , 1070 (s) , 1050 (s), 1035 (s), 1020 (s) , 975 (m) , 900 (s) , 760 (s) , 700 (s) , and 650 (w) cm. The NMR spectrum (CCl.) showed a singlet at 7.90 T overlapping a multiplet centered at 7.84 T, a singlet at 6.49 T, a multiplet centered at 6.13 T, and a multiplet centered at 2.80 T with area ratios of 1:2:2:4:5 respectively. Preparation of 3-(3-phenyltetrahydrofuranyl)methyl p-toluenesulfonate The same procedure as for the preparation of 2-(2-phenylallyloxy)— ethyl p-toluenesulfonate was followed (70). A 250-ml. three-necked flask containing 6.65 g. (0.033 mole) of 88 percent 3-phenyl— 3-hydroxymethyltetrahydrofuran and 10.5 g. CO. 13 mole) of pyridine was cooled to 10° while the contents were stirred, and 8.4 g. (0.044 molel of p-toluene— sulfonyl chloride was added at such a rate as to maintain the temperature

PAGE 104

95 well below 20°. The slurry was maintained around 15° while it vjas stirred an additional four hours, then a solution of 20 ml. of hydrochloric acid in 70 ml. of ice water was added slowly. The resulting suspension was filtered through a cold Euchner funnel. The solid product was recrystallized from methanol to give a crystalline white solid, m.p. 115° (lit. (43) m.p. 104°). The infrared spectrum (KBr) of the product showed absorption bands at 3100 (w), 3070 (m) , 3040 (m) , 2980 (m) , 2950 (m) , 2920 (m) , 2880 Cs) , 1930 (w), 1880 (w), 1800 (w) , 1600 Cs) , 1495 (m) , 1450 (m) , 1400 Cw) , 1350 (s), 1310 (m), 1290 (m) , 1210 (w) , 1190 (s) , 1175 (s>, 1135 (m) , 1120 (w), 1095 (s), 1070 (s) , 1045 (s) , 1020 (m) , 975 (s) , 900 (m) , 845 (s), 810 (s), 770 (m) , 750 (m) , 695 (s), and 660 (s) cm."-"-. The NMR spectrum (CCl.) showed a multiplet centered at 7.77 T overlapping a singlet at 7.60 x, a multiplet centered at 6.16 T overlapping a singlet at 5.99 T, a multiplet centered at 2.85 T overlapping a doublet centered at 2.79 T, and a doublet centered at 2.50 T with area ratios of 2:3:4:2:7:2 respectively. Anal . Calcd. for C^gH2QS0^: C, 65.04; H, 6.06; S, 9.65. Found: C, 64.88; H, 6.19; S, 9.50. Preparation of 3-phenyl-3-bromomethyltetrahydrofuran A 300-ml. round-bottomed flask was charged with a solution of 8.3 g. (0.025 mole) of the tosylate from the preceding experiment and 7.0 g. (0.080 mole) of anhydrous lithium bromide in 150 ml. of methyl isobutyl ketone. The solution was refluxed for 2.5 hours, then allowed to cool to room temperature. The precipitated lithium tosylate was filtered off and the methyl isobutyl ketone was removed by rotary evaporation.

PAGE 105

96 The remaining liquid was taken up in 60 ml. of ether, while 25 ml. of water was added to dissolve the white solid (LiBr). The aqueous layer was separated and washed with 15 ml. of ether.. The combined ether layers were washed with 20 ml. of water and dried over anhydrous magnesium sulfate. Filtration and removal of the ether left 6.3 g. of a viscous liquid, which was distilled to give 5.2 g. (87 percent) of a clear, 25 colorless, viscous liquid, b.p. 92-3/0.07 mm., n 1.5676. The infrared spectrum (neat) of the product showed absorption bands at 3100 (w), 3070 (m) , 3040 (m) , 2970 (s) , 2930 (m) , 2880 (s) , 1600 (m) , 1580 (w), 1495 (s), 1445 (s) , 1425 (m) , 1365 C\-}) , 1340 (w) , 1290 (w) , 1260 (m), 1240 (s) , 1210 (w) , 1160 (w) , 1125 (w) , 1070 (s) , 1030 (w) , 1005 (w), 980 (m), 910 (m) , 840 (w) , 760 (s), 695 (s) , and 620 (s) cm.""*". The NMR spectrum (CCl.) showed a multiplet centered at 7.70 T, a singlet at 6.40 T, a multiplet centered at 6.03 x, and a multiplet centered at 2.81 T with area ratios of 2:2:4:5 respectively. Anal . Calcd. for C H BrO: C, 54.79; H, 5.43; Br, 33.14. Found: C, 54.85; H, 5.43; Br, 33.24. C. Preparation and Analysis of Reaction Solutions Reactions of 2-bromoethyl methallyl ether and tributyltin hydride The solutions for the initial attempt to study the reactions of 2-bromoethyl methallyl ether with tributyltin hydride uere prepared in the following manner: The AIBN, which had been recrystallized from methanol, was weighed out first and set aside. Tributyltin hydride, which had been prepared two or three days and stored in a screw-cap bottle, was weighed in a volumetric flask. Then the 2-bromoethyl methallyl ether, which was free of impurities according to gas

PAGE 106

97 chromatograpliy , was weighed in the volumetric flask and the flask was filled to the mark with the solvent, reagent grade benzene which had been distilled from sodium. The AIBN was added and the solution was shaken and transferred to reaction tubes. The tubes were degassed by three freeze-that^ cycles, sealed, and heated at the desired reaction temperature until the reactions had reached completion. The product distributions resulting from the various reactions were analyzed by gas chromatography on a Hy-Fi Aerograph, using a nine-foot b ,b'-oxydipropionitrile column at 50-55° with the injection temperature at 185°. Calibration curves for concentration versus peak area were prepared for the reaction products using standard solutions of mixtures of the products in benzene. The accuracy of these determinations was checked often by preparing solutions containing the concentrations of the three products in the reaction mixture according to the calibration curve. In this manner slight errors could be corrected and the calibration could be checked often. A method using an internal standard to calibrate the product concentrations was initially tried, but this method proved to be very unsatisfactory. Due to the lack of a distinct trend in product distributions and the speed at which the reactions reached completion, it was decided that the reactions should be repeated, using greater care to exclude impurities from the solutions and prevent reaction from occurring prior to heating at the desired temperature. The following changes were made in the reaction system: The solvent used was spectrograde benzene which had been dried over activated 3A molecular sieves.. Tributyltin hydride was freshly prepared and stored in an airtight container under nitrogen.. The amount of AIBN used was

PAGE 107

98 reduced from 3 mole percent to 1.5 mole percent based on hydride concentration. The solutions were prepared rapidly and were promptly added to the tubes and frozen. The AIBN was weighed out first and set aside. The tributyltin hydride was weighed in a volumetric flask and diluted with solvent. The bromide was washed in with solvent and the flask was filled to the mark. The AIBN was added, the mixture was shaken, and the solution was put in the tubes and frozen. The solutions prepared in this manner are shown in Table I. The product distributions were determined in the same way. Duplicate reactions were run to check the accuracy, and in each case both reaction mixtures showed the same product concentrations. The product concentrations and percent yields from the reaction solutions listed in Table I are given in Tables II and III. Reactions of 2-bromoethyl 2-phenylallyl ether and tributyltin hydride The improved procedure from the preceding section for preparing the reaction solutions listed in Table I for the 2-methallyloxyethyl radical was again employed to prepare the reaction solutions for the 2-(2-phenylallyloxy) ethyl radical, which are described in Table V. The product analyses were carried out as before on a Hy-Fi Aerograph, using standard solutions of the products to calibrate a five-foot 30 percent SE-30 Column at 190° with the injection port at 290°. The product concentrations and percent yields determined for the reaction solutions listed in Table V are given in Tables VI and VII. Reaction of 3-phenyl-3-bromomethyltetrahydrofuran with tributyltin hydride Again the same procedures for preparing and analyzing the samples

PAGE 108

99 employed in the preceding experiments were used. A solution was prepared 0.1 molar in tributyltin hydride, 0.1 molar in 3-phenyl-3-bromomethyltetrahydrofuran, and 0.0017 molar in AlBN. The degassed reaction tube was heated at 90° for four hours, then analyzed. The only product of the reaction was 3-phenyl-3-methyltetrahydrofuran. No ethyl 2-phenylallyl ether or 3-phenyltetrahydropyran was observed.

PAGE 109

BIBLIOGRAPHY 1. G. B. Butler and F. L. Ihgley, J. Am. Chem. Soc , 73_, 895 (1951). 2. G. B. Butler and R. J. Angelo, J. Am. Chem. Soc . , 79, 3128 (1957). 3. G. B. Butler, A. Crawshaw, and W. L. Miller, J. Am. Chem. Soc . , 80 , 3615 (1958). A. C. S. Marvel and R. D. Vest, J. Am. Chem. Soc , 79^, 5771 (1957). 5. G. B. Butler, Encycl. Polymer Sci. Tech ., 4^, 588 (1966). 6. W. E. Gibbs and J. M. Barton, Ch. 2 in "Kinetics and Mechanism of Pol}TTierization," Vol. 1, Part 1, Vinyl Polymerization , G. E. Ham ed., Dekker, New York, 1967, p. 59. 7. G. B. Butler and G. C. Corfield, Cyclopolymerization , in press. 8. F. Goetzen and G. Schroeder, Makromol. Chem ., 88, 133 (1965). c 9. M. M. Koton et al . , Chem. Abstr ., 64_, 3693/? (1966). 10. T. A. Sokolova and G. D. Rudkovskaya, J. Poly. Sci. C , 16 , 1157 (1967) 11. G. B. Butler and G. R. Meyers, J. Macromol. Sci.— Chem , A5 , 135 (1971) 12. KSultanov and I. A. Arbuzova, Chem. Abstr . , 64, 12799h (1966). 13. W. S. Friedlander, A. C. S. Meeting Abstracts , 133 , 18N (1958). 14. R. C. Lamb, P. W. Ayers, and M. KToney, J. Am. Chem. Soc . , 85 , 3483 (1963). 15. A. L. Glasebrook and W. G. Tonell, J. Am. Ch em. Soc. , 61 , 1717 (1939). 16. H. Hart and P. D, Wyman , J. Am. Chem. Soc . , 8^1, 4891 (1959). 17. J. K. tochi and P. J. Krusic, J. Am. Chem. Soc . , 91, 3940 (1969). 18. C. V:alling and M. S. Pearson, J. Am. Chem. So c. , 86, 2262 (1964). 19. N. 0. Brace, J. Am. Chem. Soc , 86, 523 (1964). J. Org. Chem ., 31, 2879 (1966). Ibid . , 32^, 2711 (1967). Ibid., 34^, 2441 (1969). J. Poly. Sci. A-1 , 8^, 2091 (1970).

PAGE 110

101 20. S. G. Matsoyan e_t al . , Vysokomdl. Soedin , 5^, 854 (1963). 21. C. Aso , F. Kunitake, and J. Tautsumi, Kbgyo Kagaku Zasshi , 70, 2043 (1967). 22. G. B. Butler, J. Poly. Sci ., 48^, 279 (1960). 23. G. B. Butler and M. A. Raymond, J. Org. Chen ., 30^, 2410 (1965). 24. B. lachia, Ph.D. Dissertation, University of Florida, August, 1966. 25. K. B. Baucom, Ph.D. Dissertation, University of Florida, June, 1971. 26. C. S. Marvel and J. K. Stille, J. Am. Chem. Soc . , 80, 1740 (1958). 27. C. Walling et al . , J. Am. Chem. Soc . , 88, 5361 (1966). 28. D. L. Struble, A. L. J. Beckwith, and D. E. Gream, Tetrahedron Letters , 3701 (1968). 29. P. D. Bartlett et al . , J. Am. Chem. Soc . , 87, 2590 (1965). 30. M. Furue, S. Nozakura, and S. Murahashi, Kbbunshi Kagaku , 24 , 522 (1967). 31. M. Sumi, S. Nozakura, and S. Murahashi, ibid . , 24 , 512 (1967). 32. N. D. Field, J. Org. Chem ., 25, 1006 (1960). 33. G. B. Butler and M. A. Raymond, J. Poly. Sci. A , 3^, 3413 (1965). 34. M. Julia, Pure and Appl. Chem ., 15, 167 (1967). 35. M. Julia and M. Marray, Bull. Soc. Chim. Fr . , 2641 (1967). 36. M. Julia, M. Maumy, and L. Mion, Bull. Soc. Chim. Fr . , 2641 (1967). 37. D. J. Carlsson and K. U. Ingold, J. Am. Chem. Soc . , 90, 7047 (1968) 38. S. R. Landauer and H. N. Rydon, J. Chem Soc . , 2224 (1953). 39. M. F. Ansell, J. Chem. Soc . , 539 (1961). 40. T. J. Prosser, U.S. 3,168,575 (February 2, 1965); Chem. Abstr . , 62^, Pl6059g (1965). 41. R. V. Christian, Jr. and R. M. Hixon, J. Am. Chem. Soc . , 70^, 1333 (1948). 42. H. G. Kaivila and 0. F. Beumel, Jr., J. Am. Chem. Soc . , 83 , 1246 (1961). 43. J. Colonge and Y. Infarnet, C. R. Acad. Sci. Paris (C) , 264, 894 (1967).

PAGE 111

102 44. J. W. Wilt ot al. , J. Org. Chein >, 35, 2803 (1970). 45. R. C. Lamb, J. G. Pacific!, and P. W. Ayers, J. Org. Chem . , 30 , 3099 (1963). 46. M. Julia, Accounts Chem. Res . , ^, 386 (1971). 47. A, L. J. Beckwith and W. B. Gara, J. Am. Chem. Soc . , 9j^, 5691 (1969) 48. W. S. Johnson and R. Ox/yang, J. Am. Chem. Soc . , 86, 5593 (1964). 49. M. F. Ansell and D . Z. Thomas, J. Chem. Soc , 1163 (1958). 50. P. Gaubert, R. P. Linstead, and H. N. Rydon, J. Chem. Soc , 1972 (1937). 51. Inorganic Syntheses , 1, 113 (1939). 52. M. J. Frazer et a]^. , Chem. Ind . , 931 (1954). 53. Organic Syntheses Coll. Vol. I , 250 (1941). 54. CM. Stevens and D . S. Tarbell, J. Org. Chem . , 19, 1996 (1954). 55. A. I. Vogel, Textbook of Practical Organic Chemistry , 3rd ed . , 486 (1959). 56. B. J. Clarke and R. P. Hildebrand, J. Inst. Brew ., 73^, 60 (1967); Chem. Abstr ., 67_, 32303a (1967). 57. A. I. Vogel, Textbook of Practical Organic Chemistry , 3rd ed., 485 (1959). 58. S. Julia and C. Gueremy, Bull. Soc. Chim. France , 2994 (1965). 59. C. F. Kbelsch, J. Am. Chem. Soc . , 65^, 437 (1943). 60. H. A. Bruson, U.S. 2,280,780; Chem. Abstr ., 36^, P55889 (1942). 61. W. D. Stewart, U.S. 2,535,875 (December 26, 1950); Chem. Abstr . , 45, P3986e (1951). 62. C. D. Kurd and M. A. Pollack, J. Am. Chem. Soc , 60, 1905 (1938). 63. R. I. Meltzer, A. D. Lewis, and A. Fischman, J. Org. Chem . , 24 , 1763 (1959). 64. M. Taraele et al . , Ind. Eng. Chem ., 33 , 115 (1941). 65. Organic Syntheses Coll. Vol. II , 264 (1943). 66. E. Hanschke, Chem. Ber., 88, 1048 (1955).

PAGE 112

103 67. Y. K. Yur'ev, G. Y. fondrat'eva, and N. K. Sadovaya, J. Gen. Chem . USSR , 23, 883 (1953); Chem. Abstr . , 48^, 39551 (1954). 68. A. I. Vogel, Textbook of Practical Organic Chemistry , 3rd ed. , 879 (1959). 69. L. F. Fieser and M. Fieser, Reagents for Organic Synthesis , 965 (1958). 70. Organic Syntheses Coll. Vol. Ill , 366 (1955). 71. W. C. feith, U.S. 3,230,205; Chem. Abstr . , 64, 8497c (1966). 72. G. Descotes and A. Laily, Bull. Soc. Chim. Fr . , 2989 (1967). 73. A. Schoenberg and K. Praefcke, Chem. Ber ., 99 , 196 (1966).

PAGE 113

BIOGRAPHICAL SKETCH Thomas William Smith was born July 19, 1944, in Baton Rouge, Louisiana. He graduated from Port Allen High School in Port Allen, Louisiana, in June 1962. He received his Bachelor of Science degree in chemistry, cum laude, from Louisiana State University in Baton Rouge, Louisiana, in May 1966. As an undergraduate he was a member of Acacia Fraternity, was named the Phi Kappa Phi outstanding freshmen and the outstanding senior in chemistry, and held memberships in Phi Eta Sigma, Phi Kappa Phi, Pi Mu Epsilon, and Phi Lambda Upsilon honorary fraternities. In September 1966 he enrolled in the Graduate School of the University of Florida and has pursued the degree of Doctor of Philosophy. He is a member of the American Chemical Society and has been active in a number of outside activities in music during his years in college. On June 13, 1970, he married the former Linda Kathleen Tarrant of Gainesville, Florida. 104

PAGE 114

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. George B. Butler, Chairman Professor of Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. 7 -'M.A/fAJ^^ M Merle A. Battiste Professor of Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Henry C. Brown Professor of Chemistry

PAGE 115

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Ijiame-s A. Deyrup Associate Professor of Chemistry I certify that 1 have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Martin Vala, Jr. Assistant Professor of Chemistry This dissertation was submitted to the Department of Chemistry in the College of Arts and Sciences and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. June, 1972 Dean, Graduate School

PAGE 116

7i SM5 7 17. 1-1 T i»l