KETENE CAKBODIIMIDE CYCLOADDITIONS

APPROVED:

Graduate Committee:

l/J. Major Professor

Committee Member

Committee Member

TyC^Cs<-

Committee Member

Chairman of the Department of Chemistry

j / c

Dean of tftie Graduate School

KETENE CARBODIIMIDE CYCLOADDITIONS

DISSEHTATION

Presented to the Graduate Council of the

North Texas State University in Partial

Fulfillment of the Requirements

For the Degree of

DOCTOR OF PHILOSOPHY

By

Edwin Darrell Dorsey, B. S.

Denton, Texas

August, 1970

TABLE OF CONTENTS

Page

LIST OF TABLES iv

LJST OF ILLUSTRATIONS

Chapter

I. INTRODUCTION 1

II. EXPERIMENTAL 15

III. RESULTS AND DISCUSSION 30

APPENDIX 42

BIBLIOGRAPHY ^7

iii

LIST OF TABLES

Table Page

I. Preparation of Some Acid Halid.es 18

II. Cycloadducts from Ketenes and Carbodilmldes . . . ^3

III. Second Order Rate Constants for the Cyclo-addltlon of Diphenylketene and Diisopro-pylcarbodiimide at 30° 39

iv

LIST OF ILLUSTRATIONS

Figure Page

1. Diphenylketene Standard Curve **4

2. Diphenylketene Dilsopropylcarbodllmide Cycloaddltion in Isobutyronitrlle at 35 • • ^5

3. Diphenylketene Diisopropylcarbodiimlde Cycloaddltion in Benzene at 35° ^6

CHAPTER I

INTRODUCTION

Ketenes have been known since the early investigations

of Staudinger more than sixty years ago (33)* The ketene

functionality consists of a cumulative system of olefin and

carbonyl groups as illustrated.

Nc=c=o

Ketoketenes are disubstituted and when both substituents are

alkyl or aryl groups, are usually stable and can generally be

Isolated. Aldoketenes, in contrast, are monosubstituted,

unstable, and cannot generally be isolated. Because of their

extreme reactivity, ketenes have become very Important

intermediates in organic synthesis.

Ketenes may be prepared by a variety of different methods.

Pyrolysis is often used for the preparation of a few specific

ketenes (22, 24, 39* 44).

(CH3)2C—C=0 2 (CH3)2C=C=0

O=C—C(CH3)2 A

A (C2H5)2C, ^ (C2H5)2C=C=0 + co2

5 A

9 CH3-C-CH3 ^ CH2=C=0 + CHlj,

A 0

C2H5-CL^ CH. 0 ^ \ ^C=C=0 + CoHc-C-OH

C 2H5-C^ — ^ H 5

0 ^

This pyrolysis method. Is usually limited to ketenes of rela-

tively low molecular weight.

Another method which is more general involves the

dehalogenation of <>£-haloacyl halides. Staudlnger reported

the preparation of diphenylketene by the dechlorination of

diphenylchloroacetyl chloride (3*0 •

(C5H5)2C-C-C1 + Zn ^ (C6H5)2C=C=0 + ZnClg CI

This method has been used for the preparation of a large

variety of ketenes (35# ^0, 4-1).

The most widely used method for the preparation of

ketenes is the dehydrohalogenation of appropriately substituted

acyl halides. Tertiary amines, notably triethylaminei have

been used as dehydrohalogenating agents (38).

-C-C-X + \ ^0=0=0 + (CgHijJ^NjHX H /

Ketenes are very susceptible to dimerlzation. Aldoketenes

dimerlze readily to form ^-lactones (6, 1?, 18, 37) and

ketoketenes generally undergo dlmerizatlon much more slowly

to form 1,3-cyclobutadiones (36, ^-2). Aldoketenes and

halogenated ketenes have a strong tendency to polymerize.

2 RHC=C=0 RHC-C—0

RHC—C=0

R1R2C—C=0 2 R-jRoC^C-O ^ | |

^ 0=C—CR1R2

For this reason many aldoketenes and halogenated ketenes are

prepared In situ with the reacting species without isolating

the ketene.

The two principle types of reactions of ketenes are

nucleophilic addition and cycloaddition. Ketenes undergo

nucleophilic addition reactions readily with alcohols, acids,

amines, and water (29)*

, H O C=C=0 + HNu ^ -6-fi-Nu

X > 1

In recent years the cycloaddition reactions of ketenes

with unsaturated compounds have been renewed with great

interest from a mechanistic point of view. The more common

types of unsaturated compounds include carbonyls, imines,

and olefins (29)• The cycloaddition reactions of ketenes

and olefins have been extensively studied with particular

emphasis on conjugated dlenes (31). enamines (15), and vinyl

ethers (4-3) - The cycloaddition of a ketene and an olefin

produces a cyclobutanone (11, 23, ^3).

I \ \ / -C—C=0 c=c=o + c=c I

/ / \ -C-C-

This cyclobutanone arises from a (2+2) cycloaddition

process and occurs exclusively, even when the olefin is a

conjugated diene. The mechanism of the ketene-olefin cyclo-

addition has been shown to be "near concerted with some

charge separation in the transition state" (1, 2, 3, 1^, 20,

30). Recent theoretical developments proposed by Woodward

and Hoffman are consistent with the (2 + 2) concerted cyclo-

addition process (^5» ^6). These authors have shown that for

the reaction to take place as a concerted process, the if system

of the ketene must act in an antarafacial manner and the

olefin must participate in a suprafaclal manner. This

process (7/ 2S +<?r2&) is allowed by orthogonal approach of

the ketene and olefin.

It has been shown recently that enamines, a particular

type of olefin, may also undergo cycloaddition with ketenes

via a two-step process involving a dipolar intermediate.

In this reaction, under appropriate conditions, it was

shown that the reaction proceeded via two competing pathways,

i.e., concerted and two-step (15)*

(CH3)2C=C=0

(^N-CH=C(CH3)2

V

-c=o <ch3)2c- ^

Q n - c — c ( c h 3 ) 2

H

(CH3 ) 2 q ^ p

( c h 3 ) 2 6 n

V

C=N

A H 0

(CH3)2Q^^P

© /(CH 3) 2

N=CT H

(ch3)2C=C=O

(ch 3 )c V < Y >

(CHolz^CIhJz

H x n

It is Interesting to note that the dipolar intermediate can

cyclize to form the cyclobutanone or react with another

molecule of ketene to form the £ -lactone.

There are many reports in the literature which date back

to the early investigations of Staudinger on the cycloaddition

reactions of ketenes and imines across the C-N bond to form

the /3 -lactam (26).

C—C—0 v ;c=n--c—c=o

I I -C—N-

Pfleger and Jager have extensively investigated the

reaction of diphenylketene and substituted imines. It was

reported that diphenylketene did not react with N-(phenylmethylidaie)-

benzylamlne (C6H5CH2N=CHC6H5) or N-(3-methylbutylidene)-aniline

(lso-Cj[HnCH=N-C£H^) but did undergo cycloaddition with a wide

variety of imines of the type Ph-N=CHAr or Ar-N=CHFh (32).

The ft -lactams that are produced are often stable and can be

decomposed only at elevated temperatures. The mechanism of

this cycloaddition reaction has quite recently received a

considerable amount of attention. Kagan and Luche have

recently proposed a two-step mechanism for this reaction

involving a dipolar intermediate as illustrated (19» 27, 28).

C6H5-CH=N-C6H5 +

(C6H5)2 c=c=o

© © c^-CH^N-ogE^

e (C6H5)2C=C-O

e (C6H5)2C-C=0

C6H5-CH-N-C6H5

(C6H5)2C—C=o

Trapping experiments and solvent and substituent effects

proved to be consistent with the proposed mechanism.

Huisgen and coworkers have also reported evidence for the

formation of a 1,4-zwitterionic intermediate in the cycloaddition

of diphenylketene and benzyl! denemethylamine (13)- It was shown

that in an excess of diphenylketene, the dipolar intermediate

could close to the p -lactam or react with another molecule of

diphenylketene to yield the 1,3-oxazinone derivative.

(C6H5)2C=C=O

+•

H C6H5-C=N-CH3

<C6H5)2C—C=O

CGH5C—N-CH3 H

(o6H5)2c^/p

C6H5—CH-® CH3 <r

(c6H5)2c

(C6E5)2C^t>0

Jt H3C ©XCHG6H5

(C6H5)2C®C«0

'Vv C H3 * ^ C 6 H 5 ) 2

H C 6H 5

Gomes and Joullie reported that in an attempt to react

ketene with benzalaniline in liquid sulfur dioxide as the

solvent, a compound containing sulfur dioxide was Isolated

in good yield (9)« It was proposed that this compound arose

from the addition of a molecule of sulfur dioxide to a

1,4-dipolar intermediate.

C6H5-CH=N-C6H5

+

CH2=C=O

© C6H5-CH-N-C6H5

Ho 0-0=0

SO' ->

C6H5-p -CH—N-C6H5

SOO 0

CH2

However, further investigations led these researchers to believe

that with ketene the reaction does not go as shown above, but

rather involves attack of the reaction product of ketene and

sulfur dioxide on the imine (8).

H2C=C=O +

SOO

?°2

6H2 \ c = o ©

<--> < ^ p = o

so2

Rc

BT -C=N-R,

'CH SO'

2x

R1R2C~

<j:=o

N-R'

8

Ghosez has recently reported on the cycloadditlon of

some halogenated ketenes with aryl and alkyl imines (5).

X I

cixc=c=o + RiR2c=N-R3 > ci-<j:—c=o

R1R2C—N-R^

where X = H, CI

BX - C 6H 5

R2 = C6H5, H

R3 = c 6 H 5 » CH3(CH2)3-, c 6 H N -

Carbodiimides are a class of compounds which contain a

cumulative system of two carbon-nitrogen double bonds as

illustrated.

-N=C=N-

There are a variety of methods which have been employed for

the preparation of carbodiimides. One of the oldest methods

involves the removal of hydrogen sulfide from thioureas.

The desulfurization is effected with yellow mercuric oxide*

lead oxides, and other metallic oxides. These preparations

Include aryl and alkyl carbodiimides (4-7). More recently

carbodiimides have been synthesized from isocyanates. This

method employs 3-methyl-l-ethyl-3-phospholine~l-oxide as a

catalyst (k).

2 R-N=C=0 > R-N=C=N-R + C02

g -CH 3 of N C 2

H 5

The reactions of carbodiimid.es have been studied exten-

sively (21, 25). However, the first report of the cycloaddition

of carbodlimides with ketenes has only recently appeared. In

the first two reports no structures were given and one report

indicated that the adducts were not isolated due to decom-

position (7» 12). Hull has recently reported the cycloaddition

of dichloroketene with two dialkylcarbodiimides to produce

the corresponding imino-ft -lactam (16).

C12C=C=0 Cl-C—C=0 + > | |

R-N=C—N-R R-N=C=N-R

It is interesting to note that when R was cyclohexyl- or

isopropyl-, the reaction proceeded in good yield. However,

when R was o-tolyl-, the cycloaddition did not occur.

With the above considerations in mind, it was proposed

to study the cycloaddition of ketenes and carbodilmides in

some detail. The first objective was to investigate the

general applicability of the reaction as a tool for the

synthetic organic chemist in the preparation of a new olass

of substituted ft -lactams; i.e., imino-^-lactams. The

10

second objective was to determine the mechanism of the

reaction. In studying a mechanism in which a two-step process

is suspected, the best evidence for such a process is direct

observation of the intermediate, either by isolation or

trapping techniques. Other evidence for a two-step process

involving a polar Intermediate is the observation of substituent

and solvent effects (10). It was therefore proposed for this

part of the research problem to look for the intermediate,

either directly or indirectly, by trapping experiments. It

was further proposed to study substituent effects in the

ketene and carbodilmlde and also Investigate the effect of

solvent polarity on the reaction rate. From these data, it

was hoped that the mechanism of the cycloaddition reaction

could be elucidated.

CHAPTER BIBLIOGRAPHY

1. Baldwin, J. E. and. Kapeckl, J. A., "Kinetic Isotope Effect on the (2+2) Cycloaddition of Diphenylketene with °c-and - Deuterostyrene," Journal of the American Chemical Society, 91 (May, 1969) , 3106-31087

2. Brady, W. T. and O'Neal, H. R., "Further Studies on the Mechanism of Diphenylketene Cycloaddition," Journal of

1, 32 (September, 196?), 2704-2707.

"Kinetics and Mechanism of the Cycloaddition of Diphenylketene and Dihydropyran,11

J ournal of Organic Chemistry. 32 (March, 1967), 612-614.

4. Campbell, T. W., Monagle, J. J., and Foldi, V. S., "Car-bodilmides. I. Conversion of Isocyanates to Carbodiimides with Phospholine Oxide Catalyst," Journal of the American Chemical Society. 84 (October, 1962) , 3673-3677.

5. Duran, F. and Ghosez, L., "Cycloaddition of Haloketenes to Imines: A Convenient Synthesis of Functionally Substituted -Lactams and 2-Pyridones," Tetrahedron Letters. 3 (January, 1970), 245-248.

6. Enk, E., and Spes, H., "Trimere Aldoketene," Anprewandte Chemie, (January, 1961), 334-338.

7* Fabrenfabrlken Bayer Akt. Gesellschaft. British Patent. 797.972 (1958); Chemical Abstracts. 53 (1959), 3059.

8. Gomes, A. and Joullie, M. M., "The Chemistry of a Ketene-Sulfur Dioxide Adduct (1)," Journal of Heterocyclic Chemistry, 6 (October, 1969) , 729-73W7

"Cycloaddition of Keten and Imines to Sulphur Dioxide," Chemical Communications. Number 18 (September, 1967) ,"~935-936.

10. Gompper, R., "Cycloaddltions with Polar Intermediates," Angewandte Chemie. International Edition. 8 (May, I 9 6 9 ) , 312-327•

11. Hasek, R. H., Gott, P. G•, and Martin, J. C., "Ketenes III. Cycloaddition of Ketenes to Acetylenic Ethers," Journal of Organic Chemistry. 29 (September, 1964), 2510-2513.

11

12

12. Hoffman, R., Schmidt, E., Wamsler, K., Relchle, A., and Moosmuller, F., German Patent 960, 458 (1957) Chemical Abstracts, 53 (1959)» 1607?.

13. Huisgen, R., Davis, B. A., and Morikawa, M., "Detection of Open-Chain Intermediates in the Formation of -Lactams from Ketenes and Azomethlnes," An#ewandte Chemie, Inter-national Edition, 7 (October 19&8), 8 2 6 - 8 2 7 .

Ik. Huisgen, R., Feiler, L., and Binsch, G., "Stereospecific Addition of Ketenes onto Enol Ethers," Angewandte Chemie, International Edition. 3 (November, 1964), 753-75^*

15* Huisgen, R., and Otto, P., "Demonstration of Two Mech-anistic Path-ways in the Reaction of Dimethylketene with N-Isobutenylpyrrolidine," Journal of the American Chemical Society 91 (October, 1 9 6 9 ) , 5 9 2 2 - 5 9 2 3 .

16. Hull, R., "Some Reactions of Dlhalogenoketens with Carbodi-imides," Journal of the American Chemical Society. C, (196777 115^-1155.

17- Hurd, C. D. and Blanchard, C. A., "The Structure of Diketene and Butylketene Dimer," Journal of the American Chemical Society. 72 (April, 1950JT I46l-l5"63.

18. Hurd, C. D., Cantor, S. M., and Roe, A. S., "Acetylation of Carbohydrates by Ketene," J ournal of the American Chemical Society. 61 (February, 1939)» 426-42$T

19* Kagan, H. B. and Luche, J. L., "Cycloadditions des cetenes sur les imines (II) mise en evidence d'un intermediaire de reaction au cours de l'addition dlphenylcetene-benzalaniline," Tetrahedron Letters. 26 (1968), 3093-3096.

20. Katz, T. J. and Dessau, R., "Hydrogen Isotope Effects and the Mechanism of Cycloaddltlon," Journal of the American Chemical Society. 85 (July, 1 9 6 3 ) , 2172-2173.

21. Khorana, A. G., "The Chemistry of Carbodiimldes," Chemical Reviews. 53 (March, 1953b 1^5-166.

22. Klstiakowsky, G. B. and Mahn, B. H., "The Photolysis of Methylketenes," J ournal of the American Chemical Society. 79 (May, 1957). 2¥l2^2¥l9.

2 3 . Knoche, H., "Cycloaddition of Dlchloroketen an Butin-(2)," Justus Lleblgs Annalen Der Chemie. 722 (Mav. iq^qK 2 3 2 - 2 3 3 .

13

24. Krapcho, A. P. and Lesser, J. H., "Cycloaddition of DimethyIketene to Olefinic Substrates," Journal of Organic Chemistry, 31 (June, 1966), 2030-2032.

25. Kurzer, P. and Douraghi-Zadek, K., "Advances in the Chemistry of Carbodiimides," Chemical Reviews, 67 (April, 1967), 107-152.

26. Lacey, R. N., The Chemistry of Alkenes, edited by S. Patai, New York, Interscience Publishers, Inc., (1964), 1164.

27. Luche, J. L. and Kagan, H. B., "Stereochimie en serie /S -lactame. IV. Cycloaddition d'aldocetknes sur la benzalaniline," Bulletin de la Soclete Chimique de France. 6 (1968), 2450-2451.

28. , "No. 300.-Stereochimie en serie -lactame.^. V. Effect isotopique du deuterium et mechanisme des reactions de Reformatsky ou des cetenes sur les bases de Schiff," Bulletin de la Soclete Chimique de France. 5 (1969)» 1680-I&82.

29. March, J., Advanced Organic Chemistry: Reactions. Mechanisms, and. Structures. New York, McGraw-Hill Book Company, (19^1, 585-591.

30. Martin, J. C.» Goodlett, V. W., and Burpltt, R. D., "The Stereospeclflc Cycloaddition of Dlmethylketene to ols- and trans-Butenyl Ethyl Ethers," Journal of Organic Chemistry. 30 (December, 1965). 4309-4311.

31. Martin, J. C., Gott, G. P., Goodlett, V. M., and Hasek, R. H., "Ketenes. V. Reactions of Ketenes with Dienes and Olefins," Journal of Organic Chemistry. 30 (December. 1965). 4175-41^0:

32. Pfleger, R. and Jager, A., Uber die Darstellung Von -Lactamen Durch Anlagerung Von Ketenen An 2-Thiazoline und Schiffsche Basen," Berlchte der Deutschen Chemischen Gesellschaft. 90 (July, 1957), 25^0-2470.

33. Staudlnger, H., Die Ketene. Stuttgart, Ferdinand Enke, 1912.

34. , "Ketene ein neue Korperklasse." Berlchte der Deutschen Chemischen Gesellschaft. 38 (Arvrn . lorRK 1735-1739.

i *

35- , "Uber Ketene. 3. Mitteilung: Diphenylketen," Berichte der Deutschen Chemischen Gesellschaft. 39 (August, 1906), 3062-7

14

36 . , "liber Ketene. 16. Mittellung: Uber bildung und Spaltung von Vlerrlngen," Berlchte der Deutschen Ghemlschen Gesellschaft. 44 (February, 1911), 521-533*

37. t "Uber Ketene. !?• Mittellung Phenylketen and MethyIketen," Berlchte der Deutschen Chemlschen Gesellschaft, 44 (February, 1911), 533-5^3•

38. • "iiber Ketene, XIX. Uber Bildung und Darstellung des DIphenylketens," Berlchte der Deutschen Ghemlschen Gesellschaft, (May, 1911), 1619-1623•

39* Staudlnger, H.,„Fellz, H. F., and Hardner, H.f "Ketene: L. Mittellung. Uber Additlons-und Polymerizationsreaktlonen des Dlmethylketens," Helvetica Chimica Acta. 8 (1925), 306-313.

% t

40. Staudlnger, H. and Kubinsky, J., "Uber Ketens. 12. Mittellung Zur Darstellung des Ketens," Berlchte der Deutschen Chemlschen Gesellschaft. 42 (October, 1909), 521>¥215.

41. Staudlnger, H. and Ruzlnska, L., "Zur Kenntnis der Ketene. (Vierts Abhandlung) Phenylmethylketene," Annalen der Chemle. 380 (1911). 278-287.

42. Staudlnger, H., Schneider, H., Schotz, P., and Strong, P. M., "Uber Keten: XLIII. Mittellung. Uber Alkyl- und Aryl-substituierte Ketoketene," Helvetica Chlmlca Acta. 6 (1923), 291-303.

43. Staudlnger, H. and Sutter, E., "Ketene XXXII: Cyclobutan-Derivative aus Diphenyl-ketene und Athylen-Verbindungen," Berlchte der Deutschen Chemlschen Gesellschaft. 53 (June. 1920T7T092-1105. ~

44. Williams, J. W. and Hurd, C. D. "An Improved Apparatus for the Laboratory Preparation of Ketene and Butadiene," Journal of Organic Chemistry. 5 (March, 19^0), 122-125.

45. Woodward, R. B. and Hoffman, R., "The Conservation of Orbital symmetry," Anaewandte Chemle. International Edition. 8 (November, 1969), 781-853.

^6. "Selection Rules for Slgmatropio Reactions," Journal of the American Chemical Society. 87 (June, 1965), 25II-2513.

47. Zetzsche, F. and Nerger, W., "Zur Darstellung von Carbodi-imiden aus Thioharstoffen," Berlchte Der Deutschen Chemlschen Gesellschaft. 73 (April, 1940), 467-477.

CHAPTER II

EXPERIMENTAL

Infrared spectra were recorded on Perkln-Elmer Model

237, Perkln-Elmer Model 621, and Beckman Model IR-9 spectro-

photometers. Spectra were obtained using neat samples,

suspensions In potassium bromide discs, or as approximately

ten per cent solutions In carbon tetrachloride, carbon

disulfide, or chloroform. Sodium chloride discs were used

for neat samples and fixed thickness potassium bromide cells

of approximately 0.1 mm thickness were used for the solution

samples.

Nuclear magnetic resonance (nmr) spectra were recorded

on Varlan A-60 and Varian A-60A spectrometers. Spectra were

obtained from solutions in carbon tetrachloride or deuterated

chloroform containing approximately one per cent tetra-

methylsilane as an internal standard (cf = 0.00 ppm).

Mass spectra were obtained using an Hitachi RMU-6E

mass spectrometer.

A Beckman Model DB UV-vlsable spectrophotometer equipped

with a Sargent SRL recorder was used for absorption measure-

ments in the kinetic studies.

A constant temperature bath was used in the kinetic

studies. The bath was heated with an immersion heating

15

16

element coupled to a Fisher proportional temperature controller

that afforded a temperature control of 1 0.02°.

Elemental analyses were performed by the Analytical

Services Section of the Chemistry Department of North Texas

State University, Denton, Texas, and by C. F» Geiger and

Associates of Ontario, California.

Preparation of Reagents

Hexane was refluxed and distilled from n-butyllithium.

Benzene was refluxed and distilled from sodium. Isobutyronitrile

was refluxed over 4A molecular sieves and distilled immediately

prior to use. Liquid sulfur dioxide was obtained by passing

sulfur dioxide gas through a sulfuric acid gas bubbling

tower, and then condensing the gas in a flask immersed in

a dry ice-acetone bath. Trlethylamine was refluxed and

distilled from calcium hydride.

Dlcyclohexyl- and dlisopropylcarbodilmides were obtained

commercially and distilled prior to use. Dlphenylcarbodllmide

was prepared by the desulfurization of N,N'-diphenylthiourea

by the following method.

A 20.0 g (0.1 mol) portion of phosphorous pentachloride

was added to a stirred slurry of 22.8 g (0.1 mol) of N,N'-

diphenylthlourea in 250 ml of carbon tetrachloride. The

mixture was heated at 35° for thirty minutes, then refluxed

for five hours. The solvent was removed under vacuum and

the residue was vacuum distilled to yield 4.2 g (22 per cent

17

yield.) at 138-140° and. 1.0 mm (lit. bp 119-121° and. 0.4 mm)

(1); ir (CCI4) 2140 cm"1 (N=C=N) .

Dime thy Ike tene -was prepared by the pyrolysis of the

commercially available ketene dlmer, tetramethyl-1,3-

cyclobutanedione, and ketene was prepared by the pyrolysis

of acetone (2). Butylethylketene was supplied by Eastman

Chemical Products, Inc., in the form of a twenty per cent

solution in toluene. Dlphenylketene was obtained by the

dehydrochlorination of diphenylacetyl chloride with tri-

ethylamine (3). Phenylethylketene was prepared by the

dehydrochlorination of o<L-phenylbutyryl chloride according

to the following procedure.

A solution of 11.1 g (0.11 mol) of triethylamine in

20 ml of hexane was added dropwise to 18.2 g (0.10 mol) of

<xC -phenylbutyryl chloride in 50 ml of hexane at room tem-

perature. Stirring was continued for several hours after

the addition was complete. The salt was removed by filtration,

the solvent was evaporated, and the phenylethylketene was

distilled at 32-34° at 0.04 mm to yield 7 g (48 per cent

yield) of product; ir 2110 cm"* (C=C=0).

Acid Halldes

All of the acid halides listed in Table I were prepared

from the corresponding acid and an appropriate halogenating

reagent such as thionyl chloride or phosphorous tribromide

according to standard procedures. The reaction was carried

18

TABLE I

PREPARATION OF SOME ACID HALIDES

0 R-J -gh-6-X

R1 R 2 X Halogenating

Agent

c6h5 c6h5 CI S0C12

Br Br CI S0C12

c6h5 c2h5 CI soci2

CH^ CI CI soci2

CI H Br PBr^

c6h5 C! CI pci5

out in a one-neck flask fitted with a reflux condenser and a

calcium chloride drying tube. The halogenating reagent was

added through the reflux condenser and refluxed for an

appropriate period of time. The acid halide was distilled

through a 2^-inch Vigreaux column or, in the case of diphenyl-

acetyl chloride, the product was recrystallized from ether.

Synthesis of Azetldinones (^-lactams)

1-Cyclohexy1-4-cyclohexyllmlno-3.3-dlphenylazetldln-2-one (I)

A 13.2 g (0.068 mol) portion of diphenylketene was added

to a stirred solution containing 1^.0 g (0.068 mol) of

dicyclohexylcarbodiimide in 100 ml of hexane at room

19

temperature. After several hours, the reaction mixture

was filtered to yield 23.6 g (90 per cent yield) of I:

mp I58-1590; ir 1810 (C=0) and 1695 cm'1 (C=N); nmr (CCl^)& 1.57

(multiplet, 20 H), 3.4 (multiplet, 2 H), and 7.12 ppm (multlplet,

10 H).

Analysis: Calculated for C 2^H^2N2 0 : C f ®0,9» H, 8.05»

N, 6.95. Found: C, 80.8; H, 8*35; N, 6.71.

3.3~Dlt>henvl-l-lsoprot>yl-4~lsopropylimlnoazetldln-2-one (II)

A solution containing 6.4 g (0.033 mol) of diphenylketene

and 4.2 g (0.033 mol) of dllsopropylcarbodlimide in 100 ml

of benzene was allowed to stand at room temperature for two

hours. Upon removal of the solvent and recrystallization of

the solid residue from ether, II was obtained in 88 per cent

yield: mp 108.5-109.5°; lr 1810 (C=0) and 1690 cm"! (C=N);

nmr (CCI4) 0.80 (doublet, 6 H), 1.45 (doublet, 6 H), 3.66

(multiplet, 1 H), 4.03 (multiplet, 1 H), and 7-3 PPm (multi-

plet, 10 H).

Analysis: Calculated for C21H24N2O: C, 78.7; H* 7*56;

N, 8.75. Pound: C, 79-0; H, 7.58; N, 8.74.

3-Chloro-l-cyclohexyl"4-cyclohexyllmlno-

A solution of 13.3 g (0.070 mol) of ©C-chloro- -

phenylacetyl chloride in 30 ml of hexane was added dropwise

to a refluxing solution of 14.5 S (0.070 mol) of dicyclo-

hexylcarbodllmlde and 14.2 g (0.141 mol) of triethylamlne in

20

200 ml of dry hexane. After the addition was complete, the

mixture was allowed to continue refluxing for two hours.

The amine salt was removed "by filtration and the hexane was

evaporated to yield 16 g (65 per cent) of III, recrystallized

from methanol: mp 86-88°; ir 1822 (C=0) and 1?00 cm"-*- (C=N);

nmr (CCI4) £ 1.5 (multiplet, 20 H), 3*4 (multiplet, 2 H),

and 7*42 ppm (multiplet, 5 H).

Analysis: Calculated for C21H27CIN2O: C, 70.30;

H, 7-52; CI, 9.90; N, 7*81. Found: C, 70.53; H, 7-83;

CI, 9.82; N, 8.08.

3.3-Dlbromo-l-cyclohexyl-4-cyclohexylimlnoazetidln-2-one (IV)

To a refluxing solution of 12.2 g (0.059 mol) of dicyclo-

hexylcarbodiimlde and 6.55 g (O.O65 mol) of triethylamine in

100 ml of hexane, was added over a period of two hours a

| solution of 16.6 g (0.059 mol) of dibromoacetyl chloride in j

| hexane. After being refluxed for an additional thirty minutes,

the solution was cooled and filtered and the solvent was

I evaporated on a rotatory evaporator. The residue was

j recrystallized from 95 per cent ethanol to yield 14.5 g 1

(59 per cent yield) of IV: mp 121.5-122.0°; ir 1822 (C=0)

and 1716 cm-1 (C=N); nmr ( C C I 4 ) 1 . 7 (multiplet, 20 H), and

3.7 ppm (multiplet, 2 H).

Analysis: Calculated for Ci^aNaOBrg; C, 44.3; H,

5.42; N, 6.90. Found: C, 44.5; H, 5.63; N, 6.73.

21

3-Bth.y 1-1-1 sopropy 1-4-1 sopropyllmlno-~ 3-phenylazetid.In-2-one (V)

A 4.2 g (0.029 mol) portion of phenylethylketene was

added to 20 ml of dry benzene containing 3*62 g (0.02'9 mol)

of dilsopropylcarbodiimide at room temperature. The solution

was allowed to stand at this temperature for forty-eight

hours, after which time the yellow color of the ketene had

disappeared. The solvent was evaporated and the residue was

recrystallized from ether to yield 4.5 g (57 per cent yield)

of V: mp 35-36°; lr 1830 (C=0) and 1710 cm"1 (C=N); nmr

(001^)^0.95 (doublet, 3 H), 1.10 (doublet, 3 H), 1.12

(triplet, 3 H), 1.45 (doublet, 6 H), 2.28 (multiplet, 2 H),

3.42 (heptet, 1 H), 4.05 (heptet, 1 H), and 7.25 ppm (multi-

plet, 5 H).

Analysis: Calculated for C ^ I ^ i ^ O : C, 74.96; H, 8.88;

N, 10.28. Pound: C, 75*25; H, 8.93; N, 10.23.

3.3-Dlmethyl-l-lsopropyl-4-lsopropyllmlnoazetldln-2-one (VI)

To a solution of 40 ml of hexane containing 18.2 g

(0.144 mol) of dilsopropylcarbodiimide, was added 10.1 g (0.144

mol) of dimethylketene. This solution was refluxed for eight

hours, and then the solvent and unreacted carbodiimlde were

removed by vacuum distillation. The residue was recrystallized

from ligroin to yield 9.1 g (32 per cent yield) of VI. The

cycloadduct was further purified by sublimation at room

temperature and 0.01 mm pressure: mp 75-76°; ir 1815 (C-O)

22

and 1695 cm"-1- (C=N); nmr (CCl^) £ 1.10 (doublet, 6 H), 1.35

(singlet, 6 H), and 3.7 ppm (multiplet, 2 H).

Analysis: Calculated for C, 67.3» H, 10.2;

N, 14.3. Found: C, 67.06; H, 9»99; N, 14.45.

3-Chloro-3-methyl-l-cyclohexyl-4-cyclohexyllmlnoazetidin-

Anll.3 S (0.066 mol) portion of -chloropropionyl

•bromide in 50 ml of hexane was added to a stirred refluxing

solution of 13•6 g (0.066 mol) of dicyclohexylcarbodlimlde

and 13*3 g (0.132 mol) of triethylamine in 150 ml of hexane

over a period of two hours. After refluxing for an addi-

tional five hours and cooling, the amine salt was removed.

The solvent was removed on a rotatory evaporator and the

residue was recrystallized from 95 per cent ethanol to yield

4.8 g (25 per cent yield) of VII. Further purification was

obtained by sublimation at 56° In vacuo: mp 61-62°; ir

1825 (C-0) and 1705 cm"1 (C=N); nmr ( C C I 4 ) 1 - 6 (multiplet),

I.85 (singlet), and 3.55 ppm (multiplet). The singlet was

superimposed on the 1.6 multiplet. The areas were in the

ratio of 2:23*

Analysis: Calculated for C ^ H ^ ^ O C l : C, 64.8; H, 8.45;

N, 9-46. Found: C, 64.8; H, 8.77; N, 9.40.

3-Chloro-l-isopropyl»4-lsopropyllminoazetidln-2-one (VIII)

A 10.1 g (O.O65 mol) portion of chloroacetyl bromide in

25 ml of hexane was slowly added to a refluxing solution of

23

8.2 g (O.O65 mol) of dlisopropylcarbodlimide and 6.5 g (0.13

mol) of triethylamine in 100 ml of hexane. The reaction

mixture was refluxed for two hours after the addition was

completed. Upon removal of the salt by filtration and evap-

oration of the solvent, the residue was distilled to yield

2.4- g (20 per cent yield) of impure IX: ir 1820 (C=0) and

1705 cm"1 (C=N); mass spectrum, parent peak 202, P/(P+2) = 3/1,

P - 35 (loss of CI), theory, 202.

3-n-Buty1-3-ethy1-1-1sopropyI-4-1sopropyllmlnoazetldln-

2-one (IxY

To a refluxing solution consisting of 9*95 g (0.079 mol)

of diisopropylcarbodilmide in 50 ml of hexane was slowly

added 9«95 g (0.079 mol) of butylethylketene in 50 ml of

hexane. This solution was refluxed for an additional two

hours. The solvent was evaporated and the residue was

distilled at 80-89° at 0.025 111131 to yield 2.3 g (12 per cent

yield) of Impure IX: ir 1810 (C=0) and 1690 cm""1 (C=N).

1-1sopropyl-4-1sopropyllmlnoazetldln-2-one (X)

An excess of ketene was bubbled into a solution of 8.1 g

(0.07 mol) of dlisopropylcarbodlimide over a period of eight

hours. The solvent was evaporated to yield predominately

unreacted carbodiimlde with only a very small amount (5 per

cent yield) of X: ir 1820 (C=0) and 1700 cm"1 (C=N).

2k

Attempted Reaction of Dlphenylketene and. Diphenylcarbodlimlde

A 2.0 g (O.OO98 mol) portion of diphenylcarbodlimlde

was added to a solution containing 2.0 g (0.0098 mol) of

dlphenylketene in 25 ml of dry benzene. The solution was

stirred overnight at room temperature. Infrared analysis

of the solution revealed only carbodiimide and ketene at

214-0 and 2100 cm~l respectively. Refluxlng the solution for

three hours produced no change. Approximately 10 mg of

aluminum chloride was added and the solution was stirred

and refluxed for three additional hours. Infrared analysis

showed no evidence of reaction. Approximately 100 mg of

cuprous chloride was added and the solution was refluxed

overnight. The solution turned very dark, which indicates

polymerization, with no evidence of azetidlnone being

formed.

Attempted Reaction of Dlphenylketene and Dl-o-tolylcarbodllmlde

A 2.0 g (0.009^ mol) sample of di-o-tolylcarbodlimide

was added to a solution containing 1.8 g (0.0094 mol) of

dlphenylketene in 25 ml of dry benzene. The solution was

stirred and refluxed overnight. Infrared analysis of the

solution revealed only carbodiimide and ketene with no

evidence of the ft -lactam.

25

Trapping Experiments

N-Dlphenylacety 1-N. N' -dllsopropylurea (XI)

A solution containing 6.4 g (0.033 mol) of diphenyl-

ketene in 50 ml of "benzene was added to a solution of 4.2 g

(0.033 mol) of dllsopropylcarbodlimide in 50 ml of "benzene.

This reaction mixture was quenched with 25 ml of water after

four minutes. The solvent and water were evaporated on a

rotatory evaporator and the residue dissolved in chloroform.

This solution was extracted with a dilute sodium hydroxide

solution. Acidification of the basic extract resulted in

the crystallization of diphenylacetic acid (47 per cent yield).

The chloroform containing the cycloadduct was evaporated,

and the residue dissolved in ether. The solution was

fractionally crystallized to yield the ft -lactam (II) in a

40 per cent yield and the substituted urea XI in a 12 per

cent yield: mp 131-1320; ir 3300 (N-H) and 1705 and I760

cm~l (C-0); nmr (CCl^) £ 1.13 doublet, 6 H), 3.8 (multlplet,

1 H), 4.3 (multlplet, 1 H), 5*15 (singlet, 1 H), and 7.2

(multlplet, 11 H).

Analysis: Calculated for C21H26N2O2: C, 74.5; H,

7.75; N, 8.28. Found: C, 74.4; H, 7-82; N, 8.23.

Treatment of Diphenylacetic Acid with Dllsopropylcarbodlimide (Control)

To a solution containing 6.4 g (0.033 mol) of diphenyl-

acetic acid in 100 ml of benzene, was added 4.2 g. (0.033 mol)

26

of dllsopropylcarbodiimide. The solution turned brownish

yellow* with a slight evolution of heat. The solution was

quenched with 20 ml of water after five minutes and the

mixture stirred for thirty minutes. The solvent was evaporated

and the residue dissolved In chloroform. Diphenylacetic acid

was removed from the chloroform solution by extraction with

a dilute sodium hydroxide solution. Evaporation of the

chloroform solution yielded only dilsopropylurea and a small

amount of diphenylacetic anhydride. There was no indication

that N-dlphenylacetyl-N.N'-dilsopropylurea had been produced.

Cycloaddltlon of Dlphenylketene and Dllsopropylcarbodiimide

In Sulfur Dioxide (XII) ~

To a solution of 7*3 S (0.057 mol) of dllsopropyl-

carbodiimide in 125 ml of liquid sulfur dioxide at -78°, was

added 11 g (0.057 mol) of dlphenylketene. The solution was

allowed to warm up to a gentle reflux at about -10°. After

an hour, the solvent was removed with an aspirator to yield

20 g (90 per cent yield) of l,l-dioxo-2(N-isopropyllmlno)-

3-isopropyl-5,5-diphenylthlazolidin-4-one (XII): mp 119-122°

(gas liberated at mp); attempts to recrystallize (XII)

resulted in the loss of sulfur dioxide; ir (CCl^) 1710 (C=0),

1695 (C=N), 1305 cm"1 (S02); nmr (CCl^) 1.1 (complex set

of doublets, Jjjjj = 6 cps, 12 H), 3*7 (multiplet, 1 H), 5.0

(multiplet, 1 H) and 7*3 ppm (multiplet, 10 H); mass spectrum,

parent peak 38*1-, P-64 (loss of S02)» theory, 384.

27

Analysis: Calculated, for C21H24N2SO3: C, 65*6; H, 6.3;

N, 7»3* Found: C, 65*1; H, 7*0; N. 7.1»

Hydrolysis of l.tlrDloxo-2-(N-lsopropyllmlno)-3-lsopropyl-*). 'S-dlphenylthiazolldln-ff-one (XII)

A solution of 2 g of (XII) in 200 ml of ether was treated

with 20 ml of water. The mixture became warm, with the

evolution of sulfur dioxide. The mixture was warmed on a

steam bath until the evolution of sulfur dioxide ceased.

The ether layer was separated and evaporated, yielding

1*7 g (97 per cent yield) of N-dlphenylacetyl-N.N'-diisopropylurea,

(XI). For characterization, see Experimental for (XI) above.

Treatment of 1-Isopropy1-^-1sopropyllmlno-3.3-dlphenylazetldln-2-one. (11). wIth Sulfur Dioxide (Control)

A 0.1 g portion of (II) in 10 ml of liquid sulfur dioxide

was allowed to stand at -78° for twenty-four hours. The

solvent evaporated as the reaction vessel was allowed to

warm to room temperature. An infrared spectrum of the residue

was identical with (II). Also, thin-layer chromatography

revealed only one component, which had the same Rf value as

(II).

Solvent Effects

Preparation of Dlphenvlketene Standard Curve

A stock solution containing 1.11 mg/ml of dlphenylketene

was prepared by diluting 100 jul of freshly distilled dlphenyl-

ketene to 100 ml in dry benzene. This solution was further

28

diluted, to give standards ranging from 0.22 mg/ml to 1.11

mg/ml for the standard curve. Absorption measurements were

made on the standard solutions by scanning through the peak

and measuring the per cent transmission at the wavelength of

maximum absorption, which was approximately *K)0 nm on the

Beckman Model DB. The valiies obtained for per cent trans-

mission were converted to absorbance. Figure 1 shows the

standard curve obtained by plotting absorbance versus concen-

tration. The curve adheres to the Beer-Lambert law in the

concentration range used.

Rate Studies

Cycloadditlons of diphenylketene and dilsopropylcarbodiimide

were run in benzene and also isobutyronitrlle at 30°. Aliquots

of 1.0 ml were removed from the reaction solution at various

intervals, diluted with benzene to 10.0 ml and absorption

measurements determined in the same manner described previously

for the preparation of the standard curve. The concentration

of diphenylketene was obtained from a standard curve. The

solutions were equlmolar with respect to diphenylketene and

dilsopropylcarbodiimide, and the concentrations were 0.050 M.

A diphenylketene dlmerization control revealed that dlmeri-

zatlon under the reaction conditions was negligible.

CHAPTER BIBLIOGRAPHY

1. Campbell, T. W., Monagle, J. J., and Foldi, V. S.. "Carbodiimides. I. Conversion of Isocyanates to Carbodiimides with Phospholine Oxide Catalyst," Journal of the American Chemical Society. 84 (October, 1962),

2. Hanford, W. E. and Sauer, J. C., Organic Reactions. Vol-ume III, ed. by R. Adams, New York, John Wiley and Sons, Inc., 19^6.

3- Staudinger, H., "Uber Ketene. XIX. Uber Bildung und Darstellung des Diphenylketens," Berlchte der Deutschen Chemischen Gesellschaft. 44 (May, 1911), I6I9-I623.

29

CHAPTER III

RESULTS AND DISCUSSION

The reaction of ketenes and carbodilmid.es was found to

be a 1,2-cycloaddition reaction and can be generally repre-

sented as follows. The treatment of the carbodlimide with

an excess of diphenylketene did not produce a detectable

amount of the 2:1 adduct nor was there any evidence of such

an adduct in any of the other systems.

R R~C Q=0 ^C=C=0 + R'-N=C=N-R' > I I

Br R'-N=C—N-R1

The imino-^-lactams that were prepared and the yields of

the preparations are shown in Table II. Dicyclohexyl- and

diisopropylcarbodiimides were chosen because they were

readily available commercially. The types of ketenes

employed in the first phase of the investigation Included

aryl, alkyl, halogenated, and combinations of these as well

as ketene itself.

Since halogenated ketenes are so susceptible to poly-

merication and do not lend themselves to being isolated,

in situ trapping experiments were necessary for this type of

ketene. The halogenated ketenes were generated in situ by

the dehydrohalogenation of the appropriately substituted acid

30

31

halide with triethylamine in the presence of the carbodiimlde.

The ratio of acid halide to carbodiimlde was 1:1 with an

excess of triethylamine. The optimum conditions for the

generation of ketene had previously been shown to be the

dropwise addition of acid halide to triethylamine. If

triethylamine is added to an excess of acid halide, the

reaction of the acid halide with the intermediate enolate

ion formed in the dehydrohalogenation reaction to produce an

^C-halovinyl ester becomes a significant side reaction (1, 2,

9).

© 0 o | o

-6-C-X + (CoHjoN > ^C=C^ aoia N -C-C-0-C=C. h c 5 3 Nx halide £ ±

The optimum conditions for the in situ ketene-carbodiimide

cycloadditions were found to be refluxing hexane. Most of

these cycloadditions were accompanied by polymerization of

the ketene and this undesirable side reaction is, in part,

responsible for the lower yields of the halogenated /-lactams.

This polymerization is apparent by the formation of varying

amounts of undistillable black tar in the reaction mixture.

At lower temperatures, the cycloaddition reaction is retarded

and polymerization of the ketene occurs to the exclusion of

the cycloaddition reaction. For example, the attempted

cycloaddition of methylchloroketene and dicyclohexylcarbodi-

imide at -780 resulted in polymerization with no evidence

32

for the formation of the imino-^ -lactam. Yet in refluxing

hexane a 25 per cent yield of the cycloadduct was obtained.

The aryl- and alkyIketenes were prepared, isolated and

purified prior to each cycloaddition. Diphenylketene reacted

smoothly with the alkylcarbodiimides at room temperature to

produce the imino-y? -lactams in high yields. However,

refluxing hexane was required for the cycloadditions of

phenylethylketene and the dialkylketenes. Even under the

refluxing conditions for two hours, unreacted butylethyl-

ketene and dlisopropylcarbodiimide remained in the reaction

mixture. Difficulty was experienced in the isolation and

purification of the adducts from chloro- and butylethylketenes,

but the spectral data established that these -lactams were

produced. Numerous attempts under varying conditions to

obtain an appreciable amount of the cycloadduct of ketene

and dlisopropylcarbodiimide were unsuccessful. Unlike the

halogenated ketenes, the alkylketenes did not undergo any

appreciable amount of polymerization. Since unreacted

ketene and carbodiimlde were found in several of the prepa-

rations, the low yields must be due to a lack of reactivity

of the ketenes.

The imino-^-lactams were found to be quite stable.

The compounds were stored for periods of up to two years

without any evidence of decomposition. The cycloadduct of

diphenylketene and dlisopropylcarbodiimide was found to be

33

quite stable In a hydrocarbon solvent at a temperature of

200° for several hours. Only a slight coloration appeared,

and infrared spectral evidence indicated that only very little

decomposition had occurred. The aryl and halogenated -

lactams were relatively stable in refluxlng seventy per cent

ethanol. Hydrolysis could be effected by the addition of

considerable quantities of acid or base. The hydrolysis

products were not Isolated and identified since the purpose

of the experiments was to get an indication of the stability

of the new compounds.

The lmino-^ -lactams listed in Table II reveals an

interesting correlation with respect to the reactivity of

the various ketenes as far as the yields are concerned.

Since the yields can be generally correlated to reactivity

of the ketenes, it is interesting to note that, in order

of decreasing yields, the reactivities of the ketenes could

be grouped as follows: diphenyl > phenylhalo, dihalo,

phenylalkyl > dialkyl, alkylhalo, halo > ketene. It should

be emphasized that all of these cycloaddltion reactions

were uncatalyzed and that cycloadditlons effected in the

presence of Lewis acid catalysts would be expected to show

a decrease in the reactivity differences.

The observed substituent effect is the first indication

of the mechanism of the reaction. The large differences in

the ketene reactivities Indicate that the mechanism of the

34

reaction Is not concerted. Moreover, this substituent

effect suggests a two-step process which involves a dipolar

intermediate. That is, the resonance and electronic effects

of the substituents have the capability of stabilizing or

delocalizing the negative charge on an intermediate of the

following type.

? R © R-C — C=0 R-C = C-0

® | ^ ® | R-N=C—N-R' R-N=C—N-R »

This dipolar or zwitterionic intermediate could be initially

formed by nucleophllic attack of the nitrogen on the carbonyl

carbon of the ketene. The aryl substituents would stabilize

the resulting carbanion through resonance derealization.

The halogen substituents would also be expected to stabilize

the carbanion, although to a much lesser degree, through

electron withdrawal because of the high electronegativities.

The alkyl substituents and ketene itself would not be

expected to show a stabilizing effect and indeed, would

probably destabilize the intermediate through electron

releasing effects of the alkyl groups to the carbanion.

Because of such a pronounced substituent effect with

respect to the ketene, a similar study was made with various

carbodllmides. Different types of carbodiimides were

reacted with diphenylketene. This ketene was selected

because of the relative stability and because of the excellent

35

yields of /?-lactams with the dlcyclohexyl- and diisopropyl-

carbodiimides. It was found that diphenyl- and di-o-

tolylcarbodiimide did not undergo cycloaddition with diphenyl-

ketene, even at elevated temperature and in the presence of

Lewis acid catalysts. This result is consistent with the

report of Hull (7) that dichloroketene did not undergo

cycloaddition with di-o-tolylcarbodiimide.

Since the observed substltuent effects suggested a

two-step process involving an intermediate dipolar species*

the next step was to try to observe the Intermediate spectro-

scopially. Cycloadditions of diphenylketene and the alkyl-

carbodiimldes were followed by infrared and ultraviolet

spectroscopy in an attempt to observe evidence for the

intermediate. The disappearance of the ketene and carbodl-

imide absorptions were observed in the Infrared and the

appearance of the C=0 and C=N absorptions of the i m l n o - -

lactams were observed with no evidence for the intermediate.

The disappearance of the ketene chromophore was followed

by ultraviolet spectroscopy with no evidence for the

intermediate.

Since the suspected intermediate could not be observed

directly, an attempt was made to trap the dipolar species.

This was first accomplished by quenching the cycloaddition

reaction with water after a few minutes. This led to the

formation of N-diphenylacetyl-N,N*-diisopropylurea (XI) in

a 12 per cent yield as shown in the following scheme.

36

(G6H5)2c=c=0 + R-N=C=n-R >

(G6H5)2g-c=° <°6H5>2&—

R-N=C—H-E ^ ^ R'-NSC—N-R

4 x m 1 % ° -

. . . 9 9 (e 6H5)2c—C =O (C6H^)2CH-C-g-C-NHR

R-N=C—N-R

II XI

where R = isopropyl

The acetyl substituted urea (XI) results from addition of

water to the dipolar intermediate (XIII) to form the enol

followed by tautomerism to the keto form of the urea. A

control experiment revealed that (XI) could not be produced

by hydrolysis of the ^-lactam (II) under the reaction

conditions. A second control experiment was also necessary

to show that (XI) was not the result of a reaction Involving

the hydrolysis product of diphenylketene (diphenylacetic

acid) and diisopropylcarbodiimide. Thus, the reaction of

diphenylacetic acid and diisopropylcarbodiimide produced

diphenylacetic anhydride and dlisopropylurea as expected

and none of the urea (XI). The Isolation of (XI) is cer-

tainly not consistent with a concerted process. However,

it is very indicative of a two-step mechanism involving a

dipolar intermediate.

Another trapping experiment was performed whereby the

reaction of diphenylketene and diisopropylcarbodiimide was

37

effected in liquid sulfur dioxide as the solvent. A near

quantitative yield of l,l--dioxo-2~(N~lsopropylimlno)-3-

isopropyl-5#5-<iiphenylthiazolidin-4-one (XII) was obtained.

This product could arise from the insertion of sulfur

dioxide into the 1,4-zwitterion as shown below.

(C6H5)2C=C=0 + R-N=C=N-R >

(C6H5]2g—1C=0 (C6H5)2g—C=0

-Kr-n < > R-N=?—N-R S°^ y R-N=C—N-R ^ * R-N=C—N-R XIII

(CfiHcJoC—C=0 6 5'2/ y h20 o 9 02S N-R > (Cc)2CH-C-N-C-NHR ' \ / -S02 B

fi XI N-R

XII

where R = isopropyl

The hydrolysis of (XII) produced a quantitative yield

of (XI) with the loss ofsulfur dioxide. A control exper-

iment revealed that (XII) could not be produced from (II)

under the reaction conditions. This was shown by dissolving

(II) in liquid sulfur dioxide and allowing it to stand at

-78° for twenty-four hours. The solvent was evaporated at

room temperature. Infrared analysis showed that no change

had occurred in (II) and thin-layer chromatography showed

only one spot, which had an Rf value the same as (II).

38

Another Interesting possibility for the formation of

the thiazolidine (XII) is from the reaction of a diphenyl-

ketene-sulfur dioxide adduct and the carbodiimide similar to

that proposed by Gomes and Joullle (4) for ketene and

benzalanlline in sulfur dioxide. However, these researchers

reported that diphenylketene does not form an adduct with

sulfur dioxide since they were unable to obtain a thiazoli-

dine adduct with diphenylketene and benzalanlline.

The isolation of such a large amount of the urea (XI)

(12 per cent) and a near quantitative yield of the thiazoli-

dine (XII) suggest that the dipolar Intermediate is relatively

long-lived. The first step of the cycloaddltion must be

faster than the second step for the accumulation of such a

large amount of intermediate to occur.

A comparison of the reaction rates of the cycloaddltion

of diphenylketene and diisopropylcarbodiimide in benzene

and isobutyronitrile was made. The rate of reaction was

followed by observing the rate of disappearance of diphenyl-

ketene by ultraviolet absorption measurements at about 400 nm.

The rate data shown in Figures 2 and 3 Indicated that the

reaction was second order overall. The rate constants for

the two solvents are shown in Table III. The reaction was

found to proceed about seven times faster in isobutyroni-

trile than in benzene.

39

TABLE III

SECOND ORDER RATE CONSTANTS FOR THE CYCLO-ADDITION OF DIPHENYLKETENE AND DIISOPRO-

PYLCARBODIIMIDE AT 30°

Solvent k x 102, l./mol. mln.

Benzene* . . 6 . 5

Isobutyronitrile 45

The rate of a reaction involving an intermediate ionic

species such as (XIII) is usually expected to be very

dependent upon the polarity of the solvent, e.g., very large

accelerations in the rate might be expected in more polar

solvents. However, this is true only if the formation of

the ionic intermediate is the rate determining step.

Gompper (5) has reviewed reactions of this type in which a

polar intermediate is Involved and yet only a small solvent

effect was observed. In some cases the more polar solvent

caused a small increase in the rate and in some reactions

a small decrease in the rate was observed. This small

solvent effect In conjunction with the trapping experiments

clearly Indicates that the first step (formation of the

dipolar intermediate) is not the rate-determining step in

the cycloaddition.

A mechanism involving formation of a dipolar inter-

mediate would be consistent with the fact that alkylcarbodi-

imides are very reactive toward ketenes and arylcarbodiimides

^0

are not reactive. The alkyl group on the carbodiimide,

through electron releasing effects, makes the carbodiimide

a better nucleophile. The aryl substituents on the carbodi-

imide decrease the nucleophillcity of the nitrogen atom

through resonance derealization of the electron pair

throughout the aromatic ring.

During the course of this Investigation, several

reports have appeared in which a two-step process has been

proposed for the cycloaddltlon of ketenes and imines (3, 8).

These reports, along with the report by Hulsgen (6) that

enamines can undergo cycloaddition with ketenes via a two-

step mechanism, are very consistent with the proposed mech-

anism for the cycloaddition of ketenes and carbodiimides.

The reason for a two-step process being favored over a con-

certed process is obviously the ability of the substituents

on the imines, carbodiimides, and enamines to stabilize the

zwitterionic intermediate, thus lowering the energy require-

ment for the overall cycloaddition.

In conclusion, the scope of the ketene carbodiimide

cycloaddition has been investigated. Various types of

ketenes have been found to react with alkylcarbodiimides.

However, arylcarbodilmides have been found to be unreactive

toward ketenes. The mechanism of this reaction has been

proposed to be a two-step process involving a 1,4-zwitterionic

intermediate. The substituent effects, trapping experiments,

and solvent effect are consistent with the proposed mechanism.

CHAPTER BIBLIOGRAPHY

1. Brady, W. T., Parry, F. H., Roe, R., Hoff, E. F.» and. Smith, L., "Halogenated Ketenes. XII. The Reactions of Some Acid Halides with Triethylamine.^-Halovlnyl Esters," Journal of Organic Chemistry. 35 (May, 19?0), 1515-1517.

2. Geiger, von R., Rey, M. and Dreidlng, A.MS., "Reaktion von cc -Chloroproplonyl-chlorld mit Triathylamin Buildung eines Saurechloride-enolesters," Helvetica Chimica Acta, 51 (1968), 14-66-14-69.

3. Ghosez, L., and Duran, F., "Cycloaddition of Haloketenes to Imines: A Convenient Synthesis of Functionally Substituted Z3 -Lactams and z -Pyrldones," Tetrahedron Letters. 3 (January, 1970), 24-5-248.

4. Gomes, A. and Joullie, M. M«, "The Chemistry of a Ketene-Sulfur Dioxide Adduct (1)," Journal of Heterocyclic Chemistry. 6(0ctober, 1969)* 729-73^.

5. Gompper, R., "Cycloadditions with Polar Intermediates," Angewandte Chemle. International Edition. 8 (May, 1969), 312-327.

6. Huisgen, R. and Otto, P. "Demonstration of Two Mechanistic Pathways in the Reaction of Dimethylketene with N-Isobutenylpyrrolidine," J ournal of the American Chemical Society. 91 (October, 19^9)7 5922-5923.

7. Hull, R., "Some Reactions of Dihalogenoketens with Carbodl-imides," Journal of the American Chemical Society. C, (1967). 115^-1155"

8. Kagan, H. B. and Luche, J. L., "Cycloadditions des cetenes sur les imines (II) mlse en evidence d'un intermedlalre de reaction au cours de l'addition dlphenylcetene-benzalaniline," Tetrahedron Letters. 26 (1968), 3093-3096.

9. Lavanish, J. M., "The Reaction of Dlchloroacetyl Chloride with Triethylamine: Trichlorovinyl Dlchloroacetate," Tetrahedron Letters. 57 (December, 1968), 6OO3-6OO6.

41

APPENDIX

k2

43

TABLE II

CYCLOADDUCTS FROM KETENES AND CABBODIIMIDES

RO

R -C—c=o 1 I !

R_..N=G—N-R-

Compound R1 «2 R 3 % Yield

I C6 H5 C 6H 5 C6H11 90

II C6H5 C6H5 88

III C 6H 5 CI c6h11 65

IV Br Br C6H11 59

V C6H5 C2H5 I-C3H7 57

VI CH3 CH3 I-C3H7 32

VII CH3 CI c6 h11 25

VIII CI H I.-C3H7 20

IX C4H9 C2H 5 I-C3H7 15

X H H 1-C3H7 5

44

1.0

0.8 <1>

O

£o .6 o CO %

0 .4

0.2

0.20 0.40 0.60 0.80 mg/ml

Fig. 1—Dlphenylketene standard curve

"-5

l/c

40 80 120 160

Time in minutes

200

Fig. 2—Diphenylketene diisopropylcarbodiimide cycloaddition in isobutyronitrile at 35°•

46

0.18 -

0.16 -

0.14 -

1/C

0.12

0.10 -

0.080

40 80 120 160 200

Time in minutes

Fig. 3—DIphenylketene diisopropylcarbodlimlde cycloaddition in benzene at 35°.

BIBLIOGRAPHY

Books

Hanford, W« E. and. Sauer, J. C., Organic Reactions. Volume III. edited by R. Adams, New York, John Wiley and Sons, Inc., 19^6.

Lacey, R. N., The Chemistry of Alkenes. edited by S. Patai, New York, Interscience Publishers, Inc., (1964-), 1164.

March, J., Advanced Organic Chemistry; Reactions. Mechanisms. and Structures. New York, McGraw-Hill Book Company,

Staudlnger, H., Die Ketene. Stuttgart, Ferdinand Enke, 1912.

Articles

Baldwin, J. E. and Kapecki, J. A., "Kinetic Isotope Effect on the ( 2 + 2 ) Cycloaddition of Diphenylketene with exp-and jS -Deuterostyrene," Journal of the American Chemical Society. 91 (May, 1969). 3106-310E.

Brady, W. T. and O'Neal, H. R., "Further Studies on the Mechanism of Diphenylketene Cycloaddition," Journal of Organic Chemistry. 32 (September, 1967). 2705^2707.

— , — "Kinetics and Mechanism of the Cycloaddition of Diphenylketene and Dlhydropyran." Journal 21 Organic Chemistry. 32 (March, 1967)» 612-614-.

-Brady, w. T., Parry, F. H., Roe, R., Hoff, E. F., and Smith, L., "Halogenated Ketenes. XII. The Reactions of Some Acid Halldes with Trlethylamine. oC-Halovinyl Esters," Journal QL Organic Chemistry. 35 (May, 1970), 1515-1517.

Campbell, T. W., Monagle, J. J., and Foldi, V. s., "Carbodl-imides. I. Conversion of Isocyanates to Carbodiimides with Phospholine Oxide Catalyst," Journal of the American Chemical Society. 84 (October, 1962), 3673-3677T

Duran, F. and Ghosez, L., "Cycloaddition of Haloketenes to Imines: A Convenient Synthesis of Functionally sub-

fuller wT£tJ^Tli0nea''' T 6 t r a h 6 d r 0 "

^7

48

Erik, E. and. Spes, H.» "Trimere Aldoketene," Angewandte Chemle. (January, 1961), 334-338.

Fabrenfabrlken Bayer Akt. Gesellschaft. British Patent. 797»972 (1958), Chemical Abstracts. 53 (1959), 3059.

Gelger, von R., Key, M. and Dreidlng, A. S., "Reaktion v o n ^ -Chloroproplonyl-chlorld mlt Tri&thylamin Buildung eines Saurechloride-enolesters," Helvetica Chlmlca Acta, 51 (1968), 1466-1469.

Gomes, A. and Joullie, M. M.» "The Chemistry of a Ketene-Sulfur Dioxide Adduct (1)," Journal of Heterocyclic Chemistry. 6 (October, 1969)» 729-73^.

"Cycloadditlon of Keten and Imlnes to Sulphur Dioxide," Chemical Communlcations. Number 18 (September, 1967)» 935-936" ~

Gompper, R., "Cycloadditions with Polar Intermediates," Angewandte Chemle. International Edition. 8 (May, 1969), 312-327.

Hasek, R. H., Gott, P. G., and Martin, J. C., "Ketenes III. Cycloadditlon of Ketenes to Acetylenic Ethers," J ournal of Organic Chemistry. 29 (September, 1964), 2510-2513."

Hoffman, R., Schmidt, E., Wamsler, K., Reichle, A., and Moosmuller, F., German Patent 960, 458 (1957) Chemical Abstracts. 53 (1959). 16077.

Hulsgen, R., Davis, B» A., and Morikawa, M., "Detection of Open-Chain Intermediates in the Formation of / -Lactams from Ketenes and Azomethlnes," Angewandte Chemle. Inter-national Edition. 7 (October, i960), 826-827.

Huisgen, R., Feller, L., and Blnsch, G., "Stereospeclfic Addition of Ketenes onto Enol Ethers," Angewandte Chemle. International Edition. 3 (November, 1964), 753-754.

Huisgen, R. and Otto, P., "Demonstration of Two Mechanistic Pathways in the Reaction of Dimethylketene with N-Isobutenylpyrrolidlne," Journal of the American Chemical Society. 91 (October, 19^917 5922-5923.

Hull, R., "Some Reactions of Dlhalogenoketens with Carbodi-imides," J ournal of the American Chemical Society, c. (1967). 115^-1155. ~

^9

Hurd, C. D. and Blanchard, C. A., "The Structure of Dlketene and Butylketene Dimer," Journal of the American Chemical Society. 72 (April, 19507, i£5i-!lE63.

Hurd, G. D., Cantor, S. M., and Roe, A. S., "Acetylation of Carbohydrates by Ketene," Journal of the American Chemical Society, 61 (February, 1939). 2<3-£28.

Kagan, H. B. and Luche, J. L., "Cycloaddltions des, cetenes sur les imlnes (II) mlse en evidence d'un intermediaire de reaction au cours de l'addltlon diphenylcetene-benzalanlllne," Tetrahedron Letters, 26 (1968) , 3093-3096.

Katz, T. J., and Dessau, R., "Hydrogen Isotope Effects and the Mechanism of Cycloaddltlon," Journal of the American Chemical Society. 85 (July, 1963).2172-2173.

Khorana, A. G., "The Chemistry of Carbodllmides,11 Chemical Reviews, 53 (March, 1953)» 1^5-166. ""

Klstiakowsky, G. B. and Mahn, B. H., "The Thotolysls of Methyl-ketenes," Journal of the American Chemical Society, 79 (May, 19571.2412-2^19.

Knoche, H., "Cycloaddltlon of Dlchloroketen an Butin-(2)," Justus Lleblgs Annalen Per Chemle. 722 (May, 1969) , 232-233.

Krapcho, A. P. and Lesser, J. H., "Cycloaddltlon of Dimethyl-ketene to Oleflnic Substrates," Journal of Organic Chemistry. 31 (June, 1966), 2030-2032.

Kurzer, F. and Douraghi-Zadek, K., "Advances In the Chemistry of Carbodllmides," Chemical Reviews. 67 (April, 1967) , 107-152.

Lavanlsh, J. M., "The Reaction of Dichloroacetyl Chloride with Triethylamlne: Trichlorovlnyl Dlchloroacetate," Tetrahedron Letters. 57 (December, 1968) , 6003-6006.

Luche, J. L«, and Kagan, H. B., "Stereochlmie en serle -lactame. IV. Cycloaddltlon d'aldoc^tenes sur la benzalanl-llne," Bulletin de la Soclete Chlmlque de France. 6 (1968), 2450-2^51.

t "No. 300.-Stereochlmie en s£rie /i -lactame., V. Effect isotopique du deuterium et mechanisme des reactions de Reformatsky ou des cetenes sur les bases de Schiff," Bulletin de la Societe Chimiaue de France. 5 (1969), 1680-1682.

50

Martin, J. C., Goodlett, V. W., and Burpitt, R. D., "The Stereospecific Cycloaddition of Dimethylketene to cls-a n d trans-Butenyl Ethyl Ethers," Journal of Organic Chemistry, 30 (December, 1965)» 4309-4311.

Martin, J. C., Gott, G. P., Goodlett, V. M., and Hasek, R. H., "Ketenes. V. Reactions of Ketenes with Dlenes and Olefins," Journal of Organic Chemistry. 30 (December, 1965). £369-4311.

Pfleger, R. and Jager, A., "Uber die Darstellung Von /S -Lactamen Durch Anlagerung Von Ketenen An 2-Thlazoline und Schlffsche Basen," Berlchte der Deutschen Chemlschen Gesellschaft, 90 (July, 1957)» 2££0-2470.

Staudlnger, H., "Ketene eln neue Korperklasse," Berlchte der Deutschen Chemlschen Gesellschaft, 38 (April, 1905)• 1735-1739.

, "Uber Ketene. 3. Mltteilung: Dlphenylketen," Berlchte der Deutschen Chemlschen Gesellschaft. 39 (August, 1906), 3062-3069*

, "Uber Ketene. 16. Mltteilung: Uber bildung und Spaltung von Vierringen," Berlchte der Deutschen Chemlschen Gesellschaft. 44 (February, 1911) » 521-533•

"Uber Ketene. 17- Mltteilung Phenylketen and Methylketen," Berlchte der Deutschen Chemlschen Gesellschaft. 44 (February, 1911). 533-5^3.

% * it

"Uber Ketene. XIX. Uber Bildung und Darstellung des DiphenyIketens," Berlchte der Deutschen Chemlschen Gesellschaft. 44 (May, 1911), 1619-1623•

Staudlnger, H., Fellz, H. F., and Hardner, H., "Ketene: L. Mltteilung. Uber Additions- und Polymerlzationsreaktionen des Dimethylketens," Helvetica Chlmlca Acta, 8 (1925), 306-313.

Staudlnger, H. and Kubinsky, J., "Uber Ketens. 12. Mltteilung Zur Darstellung des Ketens," Berlchte der Deutschen Chemlschen Gesellschaft. 42 (October, 1909)» 4213-4215#

Staudlnger, H. and Ruzinska, L., "Zur Kenntnls der Ketene. (Vierts Abhandlung) Phenylmethylketene," Annalen der Chemle. 380 (1911). 278-287.

Staudlnger, H., Schneider, H., Schotz, P., and Strong, P. M., "tfber Keten: XLIII. Mltteilung. Ober Alkyl- und Aryl-substltuierte Ketoketene," Helvetica Chimlca Acta. 6 (1923). 291-303.

51

Staudlnger, H. and Sutter, E., "Ketene XXXII: Cyclobutan-Derivatlve aus Diphenyl-ketene und Athylen-Verblndungen," Berlchte der Deutschen Chemlschen Gesellschaft. 53 (June, 1920), 1092-1105.

Williams, J. W. and Hurd, C. D., "An Improved Apparatus for the Laboratory Preparation of Ketene and Butadiene," Journal of Organic Chemistry. 5 (March, 19^0), 122-125*

Woodward, R. B. and Hoffman, R., "The Conservation of Orbital Symmetry," Angewandte Chemle. International Edition. 8 (November, 19°9),781-853.

"Selection Rules for Sigma-tropic Reactions," Journal of the American Chemical Society. 87 (June, 1965). 2511-2513,

Zetzsche, F. and Nerger, W., "Zur Darstellung von Carbodi-imiden aus Thioharstoffen," Berlchte der Deutschen Chemlschen Gesellschaft. 73 (April,19^0), 467-^77.