Title: Part I. Thermal "ene" reactions of cyclopropenes with classical eneophiles
CITATION PDF VIEWER THUMBNAILS PAGE IMAGE ZOOMABLE
Full Citation
STANDARD VIEW MARC VIEW
Permanent Link: http://ufdc.ufl.edu/UF00098930/00001
 Material Information
Title: Part I. Thermal "ene" reactions of cyclopropenes with classical eneophiles Part II. Generation and rearrangement of cyclopropenyl carbinyl anions
Alternate Title: Thermal "ene" reactions of cyclopropenes with classical eneophiles
Generation and rearrangement of cyclopropenyl carbinyl anions
Physical Description: xi, 280 leaves : ill. ; 28cm.
Language: English
Creator: Galley, Richard Allen, 1947-
Publication Date: 1975
Copyright Date: 1975
 Subjects
Subject: Cyclopropenes   ( lcsh )
Chemistry thesis Ph. D
Dissertations, Academic -- Chemistry -- UF
Genre: bibliography   ( marcgt )
non-fiction   ( marcgt )
 Notes
Thesis: Thesis--University of Florida.
Bibliography: Bibliography: leaves 272-279.
Statement of Responsibility: by Richard Allen Galley.
General Note: Typescript.
General Note: Vita.
 Record Information
Bibliographic ID: UF00098930
Volume ID: VID00001
Source Institution: University of Florida
Holding Location: University of Florida
Rights Management: All rights reserved by the source institution and holding location.
Resource Identifier: alephbibnum - 000161255
oclc - 02666852
notis - AAS7595

Downloads

This item has the following downloads:

partithermalener00gall ( PDF )


Full Text













PART I. THERMAL "ENE" REACTIONS OF CYCLOPROPENES WITH
CLASSICAL ENEOPHILES
PART II. GENERATION AND REARRANGEMENT OF CYCLOPROPENYL
CARBINYL ANIONS












By

RICHARD ALLEN GALLEY


A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF
THE UNIVERSITY OF FLORIDA
r IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE
DEGREE OF DOCTOR OF PHILOSOPHY










UNIVERSITY OF FLORIDA

1975




































TO BARBARA
















ACKNOWLEDGMENTS


The author wishes to acknowledge the financial assistance

of the Graduate School of the University of Florida (1969-

1972) and the National Science Foundation (1972-1975).

To Dr. Merle A. Battiste, his research director, go the

author's sincere gratitude for his constant suggestions, en-

couragement and friendship. The helpful assistance of the

other members of his supervisory committee, Dr. E. J. Gabbay,

Dr. P. Tarrant, Dr. R. C. Stoufer, and Dr. C. Allen, is also

greatly appreciated.

Dr. R. W. King deserves a special thanks for his com-

puter program and his assistance in interpreting mass

spectral data.

Darlene Brandner is acknowledged for her excellent

typing.



















TABLE OF CONTENTS


DEDICATION ............................. ............

ACKNOWLEDGMENTS .... .......... ..... ................

LIST OF TABLES.......................................

LIST OF FIGURES .. ...................................

ABSTRACT. ............... ............................

PART I

Chapter

I. INTRODUCTION. ...............................

II. STRUCTURE AND STEREOCHEMISTRY OF SEVERAL
CYCLOPROPENE "ENE" ADDUCTS.................

III. KINETIC AND MECHANISTIC INVESTIGATION OF
THE "ENE" REACTIONS OF CYCLOPROPENES.......

IV. EXPERIMENTAL ...............................

PART II

Chapter

I. INTRODUCTION ..............................

II. SYNTHESIS OF CYCLOPROPENYLCARBINYL
PHOSPHONIUM SALTS ...........................

III. GENERATION AND REARRANGEMENT OF
CYCLOPROPENYLCARBINYL ANIONS...............

IV. PHOTOCHEMICAL SYNTHESIS OF FLUORANTHRENES..

V. EXPERIMENTAL ..............................

BIBLIOGRAPHY...... ..................................

BIOGRAPHICAL SKETCH .................................


Page

ii

iii

v

vii

ix






2


21


44

68






118


133


144

193

204

272

280
















LIST OF TABLES


Table Page

1.1 Proton Coupling Constants of "Ene" Adducts
(41) and (44).................................. 28

1.2 Ultraviolet Spectra Maxima for Triphenylcyclo-
propene (33) and "Ene" Adducts (34), (41),
(44), (45), (46), and (48)..................... 38

1.3 Rates of Formation of "Ene" Adducts (34)
and (35)........................................ 50

1.4 Example of NMR Integration for Dimerization
of (33) ........................................ .. 98

1.5 Reproduction of Computer Printouts for the
"Self-Ene" Reaction of (33) at 99.8 + 0.10C... 99

1.6 Reproduction of Computer Printouts for the
"Self-Ene" Reaction of (33) at 120.4 + 0.1C.. 100

1.7 Reproduction of Computer Printout Data for the
"Ene" Reaction of (33) with (20) at 100.1
+ 0.1 ... ..................................... 101

1.8 Reproduction of Computer Printout Data for the
Reaction of (33) with (20) at 120.4 + 0.10C... 102

1.9 NMR Integration Employed in Determining the
Relative Reactivities of Cyclopropenes (33)
and (56) Toward Dimethyl Acetylene Dicar-
boxylate...................................... 111

1.10 NMR Integrations Employed in Determining the
Kinetic Isotope Effect for the "Ene" Reaction
of (33) with (20).............................. 114

1.11 Effect of Solvents on the Rates of the
Reaction of (33) with (20).................... 116

2.1 Ultraviolet Spectra of Cyclopropenylcarbinyl
Phosphonium Salts in Methylenechloride......... 136










Table Page

2.2 Proton Magnetic Resonance Data for Cyclo-
propenylcarbinyl Phosphonium Salts............... 140

2.3 Proton Magnetic Resonance Data of Isomeric
4,5-Diphenyl-2,4-Pentadiene Esters.............. 159

2.4 Proton Magnetic Resonance Spectral Data of
1.4,5-Triphenyl-2,4-Pentadienones (83), (84)
and (85)............ ............................ 166

2.5 Proton Magnetic Resonance Spectral Data of
1.2-Diphenyl-1,3-Butadienes (94), (95), (96)
and (97)........................................ 178

2.6 Ultraviolet Spectra of Fluorenylidene
Derivatives (49), (107), (108) and (109) ........ 201
















LIST OF FIGURES


Figure Page

1.1. Proton magnetic resonance spectrum of
1-(1', 2', 3-triphenylcyclopropenyl)-
succinic anhydride (41) in deuterchloro-
form at ambient temperature................. 29

1.2. 100 Iiz proton magnetic resonance spectrum
of the ABX portion of 1-(1',2', 3'-triphenyl-
cyclopropenyl)-succinic anhydride (41) in
deuterochloroform at ambient temperature.... 30

1.3 Proton magnetic resonance spectrum of cis-
1-(1',2',3'-triphenylcyclopropenyl)-2-
deutero-succinic anhydride (44) in deutero-
chloroform at ambient temperature............ 31

1.4. Proton magnetic resonance spectrum of 1-
(1', 2', 3'-triphenylcyclopropenyl)-
dimethyl maloate (45) in deuterochloro-
form at ambient temperature................. 33

1.5. Proton magnetic resonance spectrum of
diethyl 1-(1', 2', 3'-triphenylcyclopro-
penyl)-hydraazodicarobxylate (46) in deuter-
iochloroform at ambient temperature.......... 36

1.6. Graphic presentation of kinetic run number
1 at 99.8 + 0.10 for the "self ene" reaction
of (33)...................................... 103

1.7. Graphic presentation of kinetic run number
2 at 99.8 + 0.10 for the "self ene" reaction
of (33) .. .................................. 104

1.8. Graphic presentation of kinetic run number
1 at 120.4 + 0.1' for the "self ene" reaction
of (33) ...... ............................... 105











Figure Page

1.9. Graphic presentation of kinetic run number
2 at 120.4 + 0.1 for the "self ene" reaction
of (33)..................................... 106

1.10. Graphic presentation of kinetic run number
1 at 100.1 + 0.10 for the "ene" reaction
of (33) with (20)........................... 107

1.11. Graphic presentation of kinetic run number
2 at 100.1 + 0.10 for the "ene" of (33)
with (20) ................................... 108

1.12. Graphic presentation of kinetic run number
1 at 120.4 + 0.10 for the "ene" reaction
of (33) with (20) .......................... 109

1.13. Graphic presentation of kinetic run number
2 at 120.4 + 0.10 for the "ene" reaction
of (33) with (20)........................... 110

2.1. Line reproduction of the expanded nuclear
magnetic resonance absorptions of H1 and
112 in Z-1,2-diphenyl-1,3-butadiene (95) ... 177


viii

















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



PART I. THERMAL "ENE" REACTIONS OF CYCLOPROPENES WITH
CLASSICAL ENEOPHILES
PART II. GENERATION AND REARRANGEMENT OF CYCLOPROPENYL
CARBINYL ANIONS



By

Richard Allen Galley

August, 1975



Chairman: Dr. Merle A. Battiste

Major Department: Chemistry

This work consists of two main areas, the mechanistic

investigation of the thermal "ene" reactions of cyclopro-

penes with some classical eneophiles and the synthesis and

base catalyzed hydrolysis of several (2,3-diphenyl-2-cyclo-

propen-l-yl)carbinyltriphenylphosphonium salts.

In the "ene" reaction investigation, triphenylcyclo-

propene was reacted with the classical eneophiles maleic

anhydride, dimethyl acetylenedicarboxylate and dimethyl

azodicarboxylate. In each case, one to one adducts were

isolated and characterized. Likewise, the unsymmetrically

substituted cyclopropene 1,2-diphenyl-3-(9'-fluorenyl)-cyclo-










propene yielded a single adduct upon addition to dimethyl

acetylenedicarboxylate. 3-d-Triphenylcyclopropene was

shown to add in a stereospecific cis manner to the double

bond of maleic anhydride.

The entropy of activation, AS was determined for

the dimerization of triphenylcyclopropene and for the

addition of triphenylcyclopropene to dimethyl acetylene-

dicarboxylate at 120.40. In the former case, AS was found

to be -37 cal/mole and in the latter case, -26 cal/mole.

The latter reaction was also employed to probe the inter-

molecular isotope effect (kH/kD), solvent effects and

substituents effects. The isotope effect was found to be

1.25 at 1000. The reaction was insensitive to changes in

solvent and the p-value was determined to be -0.31. These

results are discussed in terms of zwitterionic, hydride

transfer and concerted cyclic mechanisms and the conclusion

is drawn that the concerted process best explains the ex-

perimental results.

Lastly, triphenylcyclopropene was found to add to

dimethyl acetylenedicarboxylate at least 3000 times faster

than the acyclic olefin, 1,2,3-triphenylpropene. This rate

enhancement is discussed in terms of both ground state and

transition state effects.

In the second investigation, a series of resonance

stabilized phosphoranes afforded cyclopropenylcarbinyl-

phosphonium salts when treated with diphenylcyclopropenium










perchlorate. Subsequent base (OH ) catalyzed decomposition

of these salts in aqueous dimethyl sulfoxide resulted in

the formation of triphenylphosphonium oxide and products

resulting from solvent capture of ring opened butadienyl

anions. No products resulting from capture of an inter-

mediate cyclopropenyl carbinyl anion were detected. The

stereochemistry of the phenyl substituents in the butadiene

products varied from salt to salt and steric effects appear

to govern the product distribution. Several products were

synthesized authentically via the Wittig reaction of the

phosphoranes with E- and Z-a-phenylcinnamaldehyde.

1,2-Diphenylpropenylidenefluorene, obtained from

hydrolysis of 9-(2,3-diphenyl-2-cyclopropen-l-yl)-fluorenyl-

triphenylphosphonium perchlorate, was found to yield 1,2-

diphenylfluoranthrene upon photolysis. This reaction was

extended to the synthesis of the previously unknown ring

system, phenanthro[9,10-b] fluoranthrene.
















PART I



THERMAL "ENE" REACTIONS OF CYCLOPROPENES WITH
CLASSICAL ENEOPHILES















CHAPTER I


INTRODUCTION



The "ene" reaction, or indirect substitutive addition,

has marked similarities to the well-known Diels-Alder re-

action. The latter involves the cycloaddition of a diene

and a dieneophile, while the former reaction results from

interaction of a molecule containing at least one allylic

hydrogen (the "ene" component) and an eneophile. These re-

actions are illustrated for their simplest compounds in

Scheme I.


C+ 4s + 2 O





2s + 2 + 2
H
H
Scheme I


In both cases elevated temperature and/or pressure are re-

quired to promote reaction. It should be noted that the

newly formed double bond occupies the same relative position

in both products.









In the Diels-Alder reaction two new sigma bonds are

formed and two double bonds are consumed, while in the

"ene" reaction one new sigma bond is formed at the expense

of one double bond. Also, both reactions are thermally

allowed, concerted, suprafacial transformations under the

orbital symmetry rules developed by Woodward and Hoffmann.1

Accordingly, the Diels-Alder reaction is a 4s + ,2s cyclo-

addition and the "ene" reaction a ,2 s 2 +s +2s allylic

hydrogen transfer.2 In spite of these similarities, the

Diels-Alder reaction has received far greater attention

regarding its scope, stereochemical requirements, and mech-

anism than the "ene" reaction.

The "ene" reaction was first studied by Alder in 1943.3

He observed that propene reacted with maleic anhydride (1)

at elevated temperature to give allyl succinic anhydride (2)

as shown below. Although no mechanistic work was done,

Alder envisioned the product as arising from attack by (1)

at the terminal carbon of the carbon-carbon double bond of

propene, migration of the double bond, and transfer of an

allylic hydrogen to the anhydride.




0




CH3
O O


~









4,5
Many other olefins, including 3-pinene,5 methylene-

cyclohexane,6 3-phenyl-l-butene,7 cis- and trans-2-butene,8

cyclopentene,8 and the natural product caryophyllene, ex-

hibit similar behavior toward maleic anhydride.

Two studies have correlated the structure of olefins

with reactivity toward eneophiles.10'11 The results were

ascribed to steric effects, a 1,1-disubstituted olefin be-

ing preferred over a 1,1,2-trisubstituted olefin which is

in turn preferred over a tetrasubstituted olefin. Within

a given substitution pattern, for example monosubstituted

olefins, the molecule containing the greater number of

allylic hydrogens appears to show the greatest reactivity

propenee > 3-methyl-l-butene).

Generally, 1,3-dienes do not undergo the "ene" re-

action when treated with eneophiles, but ratheryield

Diels-Alder adducts. However, cases are known in which a

1,3-diene does react preferentially in the "ene" reaction

mode. One example is the reaction of 1,3-cyclohexadiene

with diethyl azodicarboxylate (3), which gives the "ene"

product (4) and the Diels-Alder product (5) in 80% and 10%

yields respectively.12


R


OR 1 NR

R R

R = CO C H (3) (4) (5)
225









Preference for formation of the "ene" product in this ex-

ample can be attributed to the fact that (3) normally

possesses the more stable trans-configuration, which causes

severe interactions in the transition state leading to the

Diels-Alder product. The same reaction, run under ultra-

violet light, yields only the Diels-Alder adduct (5).13

Irradiation of the trans-diethyl azodicarboxylate produces

some of the cis-isomer in a photo-stationary equilibrium

and this cis-isomer is more reactive in Diels-Alder cyclo-

additions.14

Acetylenes and allenes also participate in the "ene"

reaction but only if very powerful eneophiles such as

benzyne (6), perfluor-2-butyne (7), or perfluorocyclo-

butanone (8) are employed. Allene reacts with perfluoro-

cyclobutanone, yielding the acetylene (9) by attack of the

eneophile at one of the terminal carbons.15





S2 2OH /



CH2
(8) (9)

Substituted allenes such as tetramethylallene react

differently; attack of the eneophile occurring at the central

carbon.6 Acetylenes are generally less reactive than

olefins. For example, 1-hexyne gives only a 4% yield of

the "ene" product when treated with benzyne.









Although most systems employed in the "ene" reaction

contain carbon-carbon multiple bonds, strained sigma bonds

have also shown reactivity toward eneophiles. Bicyclo[l.1.0]

butanes and bicyclo[2.1.0]pentanes, in which the zero bridge

exhibits considerable unsaturation, react with a number of

eneophiles. For example, l-cyano-3-methylbicyclo[l.l.0]

butane (10) reacts with hexafluoracetone (11) and dicyano-

acetylene (12) in refluxing ether to yield the cyclobutenes

(13) and (14) respectively.17,18

CH3
3 (13)




CN CF3 CF3

SCN
NCC--CCN CN

(10) (12) 1^ CN

CI3 CN (14)

It is clear from the examples above that a wide var-

iety of unsaturated and even some saturated compounds can

be employed as the "ene" component in the "ene" reaction.

The same versatility has been demonstrated for the eneophile.

Since Alder's initial study using maleic anhydride, a number

of olefins, alkynes, azo compounds and carbonyl compounds

have been shown to have eneophilic character. The unifying

feature of all eneophiles is the presence of one or more

electron withdrawing groups bonded directly to the multiple










bond participating in the reaction. In this regard, eneo-

philes and dieneophiles are similar in that each contain an

electron poor multiple bond.

Maleic anhydride has been the most widely used eneo-

phile of the alkene variety. Generally, reaction times of

twenty hours and temperatures of 150-200 C are required for

reaction to occur. A typical example is the reaction of cis-

2-butene with maleic anhydride at 2250 for eight hours, which

results in 30% conversion to the "ene" adduct 3-(l-butenyl)-

succinic anhydride. Numerous other electron deficient ole-

fins have also been studied. The natural product ergosteryl

acetate reacts with acrylonitrile, yielding three "ene"
19
adducts.19 Acrylonitrile also reacts with a number of sim-

ple olefins.0 Methyl acrylate and acrolein react with 3-

pinene (15) at 1350 to give the corresponding adducts (16)

and (17) in 20% and 30% yields respectively.20



0





+ --- -

R
(15) (16), R = OCII

(17), R = H



In general, olefins activated by only one electron withdraw-

ing group are poorer eneophiles than maleic anhydride,









requiring higher temperatures and longer reaction times

to produce "ene" adducts.

Acetylenes show enhanced eneophilic character over

their olefinic counterparts. This enhanced reactivity

may be attributed to the reduced pi-bond energy of the

triple bond and to the greater stability of the newly

formed vinylic carbon-hydrogen bond. Acetylene itself

reacts with a variety of olefins, forming 1,4-dienes, but

only under extreme conditions. Thus cis-2-butene reacts

with acetylene at 3500 and 2500 psi to give 3-methyl-l,4-

pentadiene.21

Acetylenes containing electron withdrawing groups

show far greater reactivity. The "ene" adducts (18) and

(19) are formed in the reactions of isobutene with (7) and

dimethyl acetylenedicarboxylate (20) at 1450 and atmos-

pheric pressure.22 These eneophiles also react with allenes

under mild conditions, affording good yields of the "ene"
16
adducts.





H
R

C CH HI R
H + 111 -- I 1

H R
R

(7), R = CF (18), R = CF3

(20), R = CO2CHJ (19), R = CO2CH3











Benzyne (6) reacts readily with simple olefins such

as 1-octene, cyclohexene, and isobutene.2324 In fact, (6)

is one of the most powerful eneophiles known and when em-

ployed with open chained dienes such as 2-methyl-l,3-buta-

diene, the "ene" reaction competes favorably with the Diels-

Alder reaction.24 In an extreme case, 2,5-dimethyl-2,4-

hexadiene reacts with (6) to yield the "ene" adduct ex-

clusively.25 The absence of any Diels-Alder adduct was

attributed to a trans-configuration of the isobutylene units

in the diene.

Diethyl azodicarboxylate (3) has received the most

attention among azo compounds found to have eneophilic

character. The reaction of 1,3-cyclohexadiene with (3) has

already been discussed.12 The reaction of 1,4-cyclohexa-

diene with (3) also results in formation of an "ene" adduct.26

Iuisgen has studied the reaction of acyclic olefins with

this eneophile and has found exculsive formation of the "ene"

adducts.27

Some carbonyl compounds also show eneophilic character.

The product that results from the "ene" reaction of an ole-

fin with a carbonyl compound is an alcohol, arising from

transfer of hydrogen from the "ene" component to the oxygen

of the carbonyl compound. Arnold first observed this re-

action when he treated methylenecyclohexane with formalde-

hyde, isolating as the sole product the unsaturated alcohol

(21).6









CH-

0 O il
+ II 0t

O H


(21)

Products resulting from reverse addition, that is vinyl

ethers, have not been observed in this reaction. Pre-

sumably, this preference results from the greater gain

in bond energy (D(O-H) + D(C-C) = 190 kcal./mole) in the

observed reaction than in ether formation (D(C-O) + D(C-H) =

177 kcal./mole).28

Although acetone shows no tendency to act as an

eneophile, its perfluoro derivative, hexafluoroacetone (11),

has marked eneophilic character, yielding the "ene" adduct

(22) upon reaction with isobutylene at 250.29




CH CH3

A2+ 1
CH3 CI3 CF3 CF
CF3 CF3

(11) (22)

As shown previously, (11) also reacts with the strained

sigma bond in bicyclo[l.l.0] butanes.17

Another perfluoroketone, hexafluorocyclobutanone (8),

is perhaps the most reactive eneophile known. Its tendency









to undergo the "ene" reaction with allenes when most other

eneophiles fail has been well documented.15'29

Keto-esters should also act as eneophiles. This was

demonstrated by Arnold when he isolated the "ene" adduct

(23) in 55% yield from the reaction of methyl pyruvate (24)

and 3-pinene (15).30'31

OH
CH3

CH2, CO2Et
0 0
+ CH C--COEt -


(15) (24) (23)

From the above discussion, it is obvious that a great

deal of work has been done to expand the scope of the "ene"

reaction. In addition, considerable attention has been fo-

cused on elucidating the mechanism of this reaction.

In general, three reaction mechanisms can be envisioned,

shown for the simplest case in Scheme II.


a



H H
H b

+ 0






Scheme II










In mechanism (a) a zwitterionic intermediate is formed

followed by intramolecular proton transfer. Mechanism (b)

involves formation of a diradical intermediate followed by

intramolecular hydrogen transfer. In mechanism (c), a con-

certed process involving a six-centered cyclic transition

state is proposed.

An ionic mechanism was first proposed by Rondestvedt

to account for the products observed in the reaction of

olefins and a-substituted maleic anhydrides.32 He envisioned

polarization of the reactants, with the negative dipole of

the olefin attacking the positive dipole of the eneophile.

The ionic intermediate formed would then yield the product

by intramolecular proton abstraction. This mechanism ac-

counted for the products obtained with a-substituted maleic

anhydrides on grounds of steric and inductive effects but

failed when the stereochemical orientation of the products

was considered.









o o
0 0 // 0




H H

0 0 0



Arnold first proposed a concerted mechanism to account

for the stereochemistry of the "ene" reaction. He observed

that the reaction of B-pinene (15) with maleic anhydride (1)









yielded a mixture of two optically active products, one

of which predominated.4 This concept was given support

by Hill.7 He reacted the optically active olefin 3-phenyl-

1-butene (25) with (1) and observed that an optically active

product (26) was obtained. The transfer of optical activity

from olefin to product is consistent with a concerted pro-

cess. In addition, had an ionic intermediate been involved,

it would be reasonable to assume that rearrangement to the

more stable benzylic cation would occur. No products

corresponding to such a rearrangement were observed.


0 0


+ ---

Ph II Ph CH \
CH3 O 0

(25) (1) (26)

A concerted process would also predict cis-addition

across the double bond of the eneophile. This was shown

to be the case by Friedrich.33 The deuterated olefin (27)

was treated with (1) and the product analysed by nmr spectro-

scopy. Comparison of the coupling constants in the product

(28) with those obtained from the undeuterated product con-

firmed the cis-addition of deuterium and the alkyl component.

However, because of partial deuterium scrambling, the limit

of detectability for trans-deuterated (28) was approximately

30%. Therefore, the "ene" reaction was at least 70% stereo-

selective.











CH
3
D D // C3 2

CHI 2
C3 CD CIO D

CH3 CH3 0
H

(27) (1) (28)

This result is consistent with a concerted mechan-

ism, but a biradical or zwitterionic intermediate in which

cis hydrogen transfer is preferred is also comparable.

The question of orientation in the transition state

for a concerted mechanism has been studied by several work-

ers. In their investigation of the reactions of optically

active olefins with (1), Hill and Rabinovitz demonstrated

that, not only was optical activity transferred to the pro-

duct, but the bulky phenyl substituent was oriented away

from the eneophile. This is to be expected on the basis

of simple steric considerations. The preference for endo

addition was demonstrated by Berson from a study of the

reaction of both cis- and trans-2-butene with (1). Thus,

cis-2-butene yielded the threo isomer (29) as the major pro-

duct (80%) while trans-2-butene gave the erythro isomer (30)

as the major product (57%).









CHI
3 CH3


H 0 H
0 IH



0 H

CH 3 H


H CH3
0 3


H

0


Additional evidence supporting the concept of endo addition

has been supplied by Hill et al. from a study of the reaction

of cis- and trans-3-d-2-pinene with (1).5 This group de-

termined that the configuration of the adduct in each case

is constant with endoid addition, and further that a per-

pendicular orientation of the allylic carbon-hydrogen bond

and the carbon-carbon double bond of the "ene" component is

preferred over a cisoid orientation.


7 7
/ I / /
/ /

I A / I
2 IIi L -- --------
x
S H


X = H; Y = D

X = D; Y = H


Cisoid


Perpendicular


CH
I 3









Thus, the reaction of trans-3-d-3-pinene with (1) results

in exclusive deuterium abstraction while cis-3-d-B-pinnen

yields exclusive hydrogen abstraction. Simultaneous to

this work, Arnold arrived at the same conclusions through

the study of cis- and trans-3-d-8-pinene with benzyne and

methyl-phenylglyoxylate.31

Despite the evidence presented, Berson has been care-

ful to point out that the stereochemical results do not

prove a concerted mechanism because "the preservation of

assymmetry in the product of the "ene" syntheses with op-

tically active olefins are necessary but insufficient con-

ditions, since a stepwise mechanism in which the carbon-

carbon bond is formed first is also compatible with them."

Other evidence for a concerted mechanism has been

provided by Franzus.26 In his study of the reactions of

1,3-cyclohexadiene and 1,4-cyclohexadiene with diethyl

azodicarboxylate (3), he found the entropy of activation,

AS to be -31.8 eu in the former case and -40.7 eu in the

latter. These large negative values are consistent with a

concerted mechanism involving a highly ordered transition

state. Huisgen studied the reactions of l-p-tolyl-3-phenyl-
27
propene and l-phenyl-3-p--tolylpropene with (3). Varying

the solvent from cyclohexane to nitrobenzene resulted in

only a four-fold rate acceleration. This minimal effect

on the rates of these reactions suggests a transition state

involving little if any charge development.










A diradical mechanism does not appear likely in the

"ene" reaction. The photochemical reaction of ketones and

olefins to form oxetanes is known to proceed via a dirad-

ical triplet; however, in these reactions, "ene" adducts

are absent.34 On the other hand, the thermal "ene" reactions

of olefins and ketones or aldehydes fail to produce any

oxetane products. Nevertheless, a diradical intermediate

appears necessary to account for the products obtained by

Taylor from the "ene" reaction of tetramethylallene and per-

fluoro-2-butyne (7) or dimethyl acetylenedicarboxylate (20).16



CH CH
3R 3 CII 3
/ H



CH l R
3 R R

CH3 CH3 R



Dolbier has shown that the related reaction of allene with
29,35
perfluorocyclobutanone proceeds via a concerted pathway.2

Thus, the result of Taylor appears to be unique for a spec-

ific example and not a general mechanism for the "ene"

reaction of allenes.

Olefins showing the greatest reactivity in the "ene"

reaction are those containing strained double bonds. The

natural product, caryophyllene, which contains a strained

trans-double bond and an exocyclic double bond, reacts

smoothyl with maleic anhydride (1) in refluxing benzene to

produce the adduct (31).











O
S1CH3 -I
/ + 0.
+3 C



(31)


Although the exocyclic double bond is more accessible to

the eneophile, it is not involved in the reaction. An

even more spectacular example is the below room temperature

dimerization via an intermolecular "ene" reaction of the

bridged cis-trans-1,5-cyclooctadiene (32).

CH



CI
3


(32)

In both of these examples, relief of strain in the "ene"

products provides the driving force for the facile re-

actions observed.

One system which contains a strained double bond and

yet has received very little attention regarding its re-

activity toward eneophiles is the cyclopropene series.

This is surprising in light of the enormous interest in the

thermally allowed Diels-Alder reactions of cyclopropene and

its derivatives. In these reactions the cyclopropene func-

tions as the dieneophile and the initial product of its











reaction with a diene contains a cyclopropane ring. The

decrease in strain energy (27 kcal./mole) in going from

cyclopropene to cyclopropane offers an attractive ex-

planation for the facility of these Diels-Alder reactions.3

Two recent but familiar examples of "ene" reactions

involving cyclopropenes only are the dimerizations of tri-

phenylcyclopropene (33) and cyclopropene.38,39 Triphenyl-

cyclopropene dimerizes in 20% yield in refluxing toluene

to give the cyclopropylcyclopropene (34), resulting from

transfer of hydrogen from one molecule to the least hin-

dered face of the other.

Ph
h Ph Ph
Ph
Ph P H

Ph H IPh
H \ (34)

Ph Ph / Ph
(33) Ph Ph

A concerted process was proposed for the reaction since rad-

ical inhibitors failed to effect the rate of dimerization.

Even more facile is the dimerization of the parent cyclopro-
39
pene at -25.39 This is probably not a radical process

either, since neither nitrobenzene nor benzoquinone effect

the rate of dimerization.

In these cases, the cyclopropene moiety acts as both

the "ene" component and the eneophile, and again strain re-

lief would appear to provide the driving force for these











reasonably facile reactions. This would not be true if a

different eneophile was employed. Nevertheless, Razin and

Gupalo have reported that the cyclopropenes (35) and (36)
40
react with benzyne to form the 'ene" adducts (37) and (38).


CO CH




Ph R Ph
Ph
(35); R = Ph (37); R = Ph

(36); R = CH3 (38); R = CH3



Both "ene" adducts result from transfer of the cyclopropene

hydrogen to the eneophile and migration of the cyclopropene

double bond. Iere, the extreme reactivity of benzyne may

be responsible for the observed reaction.

The reactions of cyclopropenes with less reactive

eneophiles remains an area of interest. This study was thus

undertaken to investigate the reactivity of cyclopropenes

with classic eneophiles and if appropriate cases could be

found, to investigate the mechanism of the "one" reactions

through the use of product studies, kinetic results, deut-

erium isotope effects and competition reactions.
















CHAPTER II


STRUCTURE AND STEREOCHEMISTRY OF SEVERAL
CYCLOPROPENE "ENE" ADDUCTS



Introduction
39
The dimerizations of cyclopropene at -25 and of
38
1,2,3-triphenylcyclopropene (33) at 11038 via the "ene"

reaction are facile processes in which the relief of strain

(27 kcal/mole) achieved in transforming a cyclopropene into

a cyclopropane may provide the required driving force.

However, strain relief is not required for the "ene" re-

actions of cyclopropenes to occur, as demonstrated by Razin
40
and Gupalo.40 These workers reacted 1,2,3-trisubstituted

cyclopropenes with benzyne and obtained tetrasubstituted

cyclopropenes as products. In addition, the substituent

bound to the saturated carbon in the starting cyclopropene

was found to reside on the double bond in the product cyclo-

propene, demonstrating that these reactions proceeded with

migration of the cyclopropenyl double bond. Because of the

extreme reactivity of benzyne, however, very little informa-

tion can be gained regarding the reactivity of the cyclo-

propene system relative to acyclic allylic systems. One

objective of this investigation is to determine whether the








bonding and geometry of the cyclopropene ring system in-

creases its reactivity toward classical eneophiles. Com-

parison of the reactivities of suitably substituted cyclo-

propene and a similarly substituted acyclic model toward

an appropriate eneophile should provide an answer to this

question.

Secondly, there exists the possibility of a unique

mechanism in these "ene" reactions; that of hydride trans-

fer from the cyclopropene to the eneophile. In such a

mechanism, an aromatic 2n cyclopropenium cation would be

formed. Combination of the cation and anion would then lead

to the "ene" adduct.





H H H H

R3 R1 R R


I H3
0 + +
R1 R2 R2 R3 R1 R 3 Rp1 R2





This mechanism is not unlikely in view of the fact

that hydride ion abstraction from 1,2,3-trialkyl- and 1,2,3-

triarylcyclopropenes is easily accomplished with triphenyl-

methyl perchlorate at room temperature.41 The hydride ion

abstraction mechanism should be easily distinguished from a

concerted process, in which little if any charge is created

along the reaction pathway leading to products, by invest-









igating solvent effects on the rate of reaction. In addition,

cyclopropenes bearing substituents (R = R2 R3) should,

in the hydride abstraction mechanism, yield two products

resulting from attack of the anion at C3 or at either C1

or C2 of the cyclopropenium cation. In the concerted mech-

anism, only one product should result from such cyclopro-

penes.

Several criteria must be met by the cyclopropenes

selected for this investigation. First, one of the sub-

stituents at C must be hydrogen to fulfill the requirement

that the "ene" component have an allylic hydrogen. Second,

the cyclopropene should be relatively stable to the elevated

temperatures required for the "ene" reaction. The parent

cyclopropene (R1 = R2 = II), which dimerizes at -25", would

not be a suitable choice since the eneophiles chosen for

this investigation could not compete favorably with this

dimerization. Finally, the cyclopropene chosen for the

initial investigation should have, for simplicity of pro-

duct separation and identification, the same substituents

at C1, C2, and C In this case, regardless of the mech-

anism operating, only one "ene" adduct should result.

One cyclopropene that meets all these requirements is

1,2,3-triphenylcyclopropene (33). This compound is sym-

metrically substituted, contains a hydrogen at C3, and has

the desired thermal stability, dimerizing in 20% yield after

heating in refluxing toluene (1100) for thirty-nine hours.38

The classical eneophiles reviewed previously should compete

favorably with this dimerization.









Preparation of 1,2,3-Triphenvlcyclo-

propene (33) and its Thermal Reaction

with Maleic Anhydride (1)

1,2,3-Triphenylcyclopropene (33) was prepared from

trans-stilbene by the method of Battiste.42 Bromine add-

ition followed by dehydrobromination yielded diphenylacetyl-
43
ene (39). Reaction of (39) with benzal chloride and po-

tassium-t-butoxide followed by treatment with hydrogen bro-

mide gas gave the cyclopropenium salt (40), which upon sod-

ium borohydride reduction afforded (33). This reaction

sequence is shown below.


Ph

Ph H 1 ) PhCHCl2/ P Ph H

2) KOH" 2) HBr Ph' Ph APh Ph
Ph







The first eneophile investigated was maleic anhydride
, chosen because of its extensive use with various other







"ene" reactants. After failure to obtain the desired "ene"

adduct at room temperature or in refluxing benzene, the re-

action was conducted in refluxing chlorobenzene (1300). In

addition to the "ene" adduct l-(l',2',3'-triphenylcyclopro-
penyl)-succinic anhydride (41), the triphenycyclopropene
i. 2) KOH 2) HBr Ph Ph ph ph


Ph

(39) (40) (33)








dimer (34) and 1,2-dieneophile indene wervestigated was maleic anhydride

charact, chosen because of its extensive use with various othe
"ene" reactants. After failure to obtain the desired "ene"

adduct at room temperature or in refluxing benzene, the re-

action was conducted in refluxing chlorobenzene (130"). In

addition to the "ene" adduct l-(1',2',3'-triphenylcyclopro-

penyl)-succinic anhydride (41), the triphenycyclopropene

dimer (34) and 1,2-diphenylindene were also isolated and

characterized. Formation of (41) and (34) arise from the










"ene" reactions of (33) with maleic anhydride and with a

second molecule of (33), respectively. Indene (42) pre-

sumably arises by the known acid-catalyzed rearrangement

of triphenylcyclopropene via a ring opened allylic cation.44

Thermal rearrangement of (33) under these conditions does

not appear likely in light of the fact that Breslow reports

a greater than 99% conversion of 3-d-1,2,3-triphenylcyclo-

propene (43) to its dimer in refluxing xylene (1450).



Ph Ph



(33) + O 0 + Ph
130 H

O 0 H


(1) (41) (42)



+ (34)


The assignment of structure (41) to the "ene" adduct

of (33) and maleic anhydride is based on its spectral data.

The 60 MHz nmr spectrum of (41) is given in Figure 1.1.

Four of the aromatic protons are shifted slightly to lower

field. These protons can be assigned to the ortho-positions

of the phenyl substituents bound to the double bond of the

cyclopropene ring. This ortho-deshielding effect is common

for other 1,2-diarylcyclopropenes such as (33) and (34).

The multiplcts at _4.30, 3.07, and 2.92 can be assigned to









the three protons bound to the succinic anhydride portion of

the molecule. These signals,are, in reality, all part of an

ABX system, as revealed by the 100 MHz nmr spectrum given in

Figure 1.2. Analysis of this system yielded the coupling

constants shown in Table 1.1. The assignments are based

upon comparison with a number of other substituted succinic

anhydrides, which show that in every case, the cis-proton

coupling constant is larger by 2-5 Hz than the correspond-

ing trans-coupling constant.33

Further support for these assignments was obtained

from the nmr spectrum, given in Figure 1.3, of the "ene"

adduct (44), isolated from the reaction of 3-d-l,2,3-tri-

phenylcyclopropene (43), with (1). The ABX pattern observed

in (41) collapsed into two doublets in (44). Good agreement

is observed between the coupling constant (J = 10.0 Hz) in

(44) and the cis-coupling constant (J = 9.1 Hz) in (41),

indicating that in each of these adducts the elements of

hydrogen or deuterium and triphenylcyclopropene have added

in a cis-mode across the double bond of maleic anhydride.

The mechanistic implications of this stereospecific addition

will be discussed later. The line broadening of the up-

field doublet in (44) is indicative of hydrogen-deuterium

coupling and thus this signal can be assigned to HA. The

downfield doublet must then be due to H
X

Additional evidence confirming the proposed structure

of (41) is obtained from its infrared, ultraviolet, and mass

spectral data. Carbon-oxygen stretching bands observed at









-l
1865 (m) and 1780 (s) cm- are indicative of a cyclic anhyd-

ride. The ultraviolet spectrum showed maxima at 330 (e =

19,800) and 314 nm (E = 24,700), comparing favorably to the

maxima observed for (33) and (34). This data is shown in

Table 1.2. The mass spectrum of (42) showed a molecular

ion at m/e 366.1263 (calcd. 366.1255) and a P+l ion at m/e

367.1284 (calcd. 367.1288), confirming that (41) is a one

to one adduct of (33) and maleic anhydride. In addition, a

major fragment at m/e 267 indicates the presence of the tri-

phenylcyclopropene ring.



The Thermal "Ene" Reaction of (33)

with Dimethyl Acetylenedicarboxylate (20)

Since activated acetylenes, in general, show greater

eneophilic character than similarly substituted olefins,

the successful "ene" reaction of (33) with maleic anhydride

presaged success with dimethyl acetylenedicarboxylate (20).

Treatment of (33) with a five-fold excess of (20) at

1000 in chlorobenzene resulted in the formation of a single

product, the expected "ene" adduct (45), in 57% yield. Un-

like the previous example with maleic anhydride, no 1,2-di-

phenylindene (42) or dimer (34) were detected in this re-

action.
















TABLE 1.1

PROTON COUPLING CONSTANTS OF "ENE" ADDUCTS (41) and (44)

Compound JABa JAX JBX


18.5 Hz 9.1 Hz 6.9 Hz


(41)


--- 10.0 Hz ---


(44)


aAt ambient temperature in CDC13.





















0


0






-I-- -- 1
O 4
N-J













0
r1






CO



0 -H .

1 -
O -H
0) -
U00






-4 -.4
00
0 oU
a >,4





C 0U
m -i-i
oa uu







0 0 0


L-





















44C)
O


0 .1


r..
S- -H




--- __ 04 .,.- J








o_ 04Jl
4r-

0 4





Cd o
rcl a







0-



>1-
u,
*H 1
-- Oi C-



0- 0
".'I 0*



-0 4J
10




S) -I
4-. c'



0--I
4- 1
,
^ ~ 0 'U
-;- _
5. 0, !-
~f ~ ~ ~ ~ rl+
_r^ I (
^ --- ->_- o i




______^ -^ -:

rl -:










_
1l












0 O









u 0-
3to
N-O

-'



OO -


H 0




16 1
S0 rl 4m




00




0-H
0 a) 0)



f2 0 4T









oO
010











O i
400
oo
GOO
00








i '


An


r----- -


~~~_~ ~

---~CP-,
-~--
'~--~--










CO2Me Ph Ph


PhCH 2C h 02M


H- CO2Me

CO2 Me


(20) (45)



Assignment of structure (45) to the "ene" adduct is

based on its spectral data. The nmr spectrum of (45), given

in Figure 1.4, is consistent with the assigned structure.

As in the previous "ene" adduct (41), four of the aromatic

protons are shifted to lower field by the ortho-deshielding

effect common to 1,2-diarylcyclopropenes. The downfield

singlet at 5.93 integrates for one proton and its position

is in agreement with a vinyl proton bonded to the B-position

of an a,B-unsaturated ester. The two upfield singlets at

3.65 and 3.50 each integrate for three protons and corres-

pond to the methyl ester protons of (45).

Supporting evidence for structure (45) is provided by

infrared, ultraviolet and mass spectral data. The cyclopro-
-l
pene double bond stretching frequency at 1840 (w) cm is

in agreement with the value observed for a number of other

1,2-diarylcyclopropenes.3845 In addition, intense carbon-
-i
oxygen double bond stretching at 1730 (s) and 1710 (s) cm

as well as strong carbon-carbon double bond stretching at

1630 (s) cm- lend support to the proposed structure. The























I 40




-0

-,-O
O-
-4 o



-O









0-i
rl 0








o
0





Q)r
4-4











OH
0r-




CI





4-J r-i- 0
0) >1 C0

S 0o

14


fa0
* 000
00,




..



Hi- c









ultraviolet spectrum, given in Table 1.2, agrees with other

1,2-diarylcyclopropenes. The mass spectrum shows a parent

ion at m/e 410, confirming that (45) is a one to one adduct

of (33) and (20). A minor fragment ion at m/e 379 (P-OCH ),

is consistent with the methyl ester formulation while frag-

ment ions at m/e 351 and m/e 292 correspond to sequential

loss of two CO CH3 radicals from the parent ion. The frag-

ment that results from these losses is formally an acetyl-

enic cyclopropenyl radical cation. An obscure and deep

seated rearrangement with sequential loss of hydrogen rad-

icals would then account for the intense fragment ions ob-

served at m/e 291, 290, and 288 (base peak). Such a re-

arrangement would also explain why the fragment at m/e 267,

associated with the triphenylcyclopropenium cation is only

8% of the base peak in this molecule.



The Thermal "Ene" Reaction of (33)

with Diethyl Azodicarboxylate (3)

The previous examples demonstrated the ability of tri-

phenylcyclopropene (33) to undergo the "ene" reaction with

the electron deficient olefin (1) and acetylene (20). A

third class of eneophiles, those containing an electron

deficient nitrogen-nitrogen double bond, was investigated

using diethyl azodicarboxylate (3). This eneophile was

synthesised according to the procedure of Rabjohn.46

The reaction of (33) with (3) was accomplished in re-

fluxing toluene and the "ene" adduct (46) was isolated in










71% yield as the sole product. Again, as in the previous

example, none of dimer (34) or indene (43) was detected.





Ph Ph

SCO.Et
N/ C2Et C02Et
(33) + I t PhCH2C 1 Ph
N 800
CO2Et H CO2Et



(3) (46)


The spectral data of (46) confirm the assigned

structure. The nmr spectrum is shown in Figure 1.5. The

aromatic portion again reveals the characteristic ortho-

deshielding effect common to 1,2-diarycyclopropenes. The

singlet at k6.78 is assigned to the amide nitrogen proton.

The methylene protons of the ester groups appear as two

overlapping quartets at 4.11 and 4.01 while the methyl pro-

tons appear as two triplets at 1.08 and 1.03.

The infrared, ultraviolet and mass spectral data

further support the assigned structure. Nitrogen-hydro-

gen stretching, nitrogen-hydrogen bending and carbonyl
-l
stretching bands at 3230 (m), 1510 (m), and 1695 (s) cm1,

respectively, are indicative of a secondary amide. In

addition, the characteristic cyclopropene double bond
-1
stretching band is observed at 1820 (w) cm The ultra-

violet spectrum of (46) exhibits maxima at 318 (c = 20,300)

and 303 nm (E = 24,000). These values represent a slight













I--


36












0


-0




o 1
14-,-



OdI
.0 0



tO~






0 (d
>1
F: 4 4-)









0> 0


0,a
r-iH
uOE
i1 >10
o SO



0 -H (
1 P
P4 4J
i- o if
ri
U &4-









hypsochromic shift relative to the maxima of (41) and (45)

but agree favorably with the values observed for other 1,2-

diarylcyclopropenes containing an electron withdrawing sub-

stituent at C3.45 47 The mass spectrum shows a parent ion

at m/e 442, confirming that (46) is a one to one adduct of

(33) and (3). In addition, the triphenylcyclopropenium

cation at m/e 267 is the base peak in this compound.



The Thermal "Ene" Reaction of 1,2-Diphenyl-

3-(9'-fluorenyl)-cyclopropene

The unsymmetrically substituted cyclopropene (47) was

prepared from fluorene and 1,2-diphenylcyclopropenium per-

chlorate by the method of Jones.48 Treatment of (47) with

one equivalent of dimethyl acetylenedicarboxylate (20)

yielded a single "ene" adduct (48).


Ph Ph EtO2C H

Ph
H\ H EtO2C
(20) 2>
S Ph
PhCH2Br


(47) (48)


The spectral data of (48) confirm the assigned

structure. The nmr spectrum showed a broad aromatic multi-

plet at J7.8-6.9 which integrated for eighteen protons. The

olefinic proton appeared as a sharp singlet at 5.80 and the

9-fluorenyl proton as a broad singlet at 5.33, which is






































1--











H
H
H




H


r r
x -

E> H
r< H












H

















0



u


n co

(N -
CO
r'I
H





C0











O G
H 0


CN







0G


0O 0
N k.J
N






m
N




rn










O
LnO
N
(N
















0U



no


H
M -











m -


o

Ln












r-
N














OO
ro-






0 G
o r~











I
I
D
C)
GD
GD G









considerably deshielded with respect to (47). The methyl

ester protons appeared as sharp singlets at 3.60 and 3.55.

Supporting evidence for structure (48) is provided

by infrared, ultraviolet and mass spectral data. The cyclo-

propene double bond stretching at 1845 (w) cm-, carbonyl

stretching at 1730 (s) and 1705 (s) cm-1, and olefinic

stretching at 1635 (m) cm-1 all lend support to the pro-

posed structure. The ultraviolet spectrum, given in Table

1.2, agrees with other cyclopropenes containing one aryl

substituent bound to the cyclopropene double bond. The

mass spectrum showed a parent ion at m/e 498, corresponding

to a one to one adduct of (47) and (20).



Attempted Thermal "Ene" Reactions of (33)

with Various Other Eneophiles

Attempts to extend the scope of the "ene" reaction of

(33) to a variety of other potential eneophiles met with

little success. For example, treatment of (33) with 1,4-

benzoquinone and 1,4-napthaquinone resulted only in the re-

covery of starting material and dimer (34). The same result

was obtained with diethyl maleate and diethyl fumarate.

When bromomaleic anhydride and chloromaleic anhydride were

employed as eneophiles, 1,2-diphenylindene (42) was isolated

as the only product. None of the desired "ene" adduct could

be detected in any of these reactions.

The reactions of (33) with 4--phenyl- and 4-methyl-l,2,

4-triazoline-2,5-dione, (49) and (50) respectively, gave










questionable results. The addition of a methylene chloride

solution of (49), which has a deep purple color, to a solu-

tion of (33) in the same solvent at 00 resulted in complete

discharge of color within thirty minutes. A white powder

was isolated from the reaction and the nmr spectrum (aro-

matic protons and a sharp singlet at 6.43), mass spectrum

(parent ion at m/e 443) and infrared spectrum (strong car-

bonyl stretching) suggested that the desired adduct was

obtained. However, lack of any significant absorption in

the 300 to 330 nm region of the ultraviolet spectrum and

the inability to obtain an adequate elemental analysis makes

assignment of structure (51) to this material rather dub-

ious.

Likewise, the reaction of (50) with triphenylcyclo-

propene (33) resulted in immediate loss of color and iso-

lation of a white powder. Again, the spectral data gave

strong indication that the desired "ene" adduct (52) was

formed, but all attempts to obtain an adequate elemental

analysis on this compound also failed.


Ph Ph
0- 0
Ph H CH2Cl2

R N Ph R
Ph Ph N- H

0 0
(49); R = Ph (51); R = Ph

(50); R = Me (52); R = Me










Decomposition of the "ene" adducts (51) and (52),

coupled with their isolation as an admixture with the cor-

responding urazil compounds might explain the inability to

obtain adequate analysis. When either (51) or (52) were

spotted on thin layer plates and eluted, considerable

streaking occurred. Some support for this explanation was

obtained by dividing the streak into two portions and

analyzing each by ultraviolet absorption. The top portion

showed considerable enhancement of the typical cyclopropene

absorptions in the 330-300 nm region whereas no absorptions

in this region were observed for the botton portion.



Comparison of the Reactivities of Tri-

phenylcyclopropene and 1,2,3-Triphenyl-

propene Toward Dimethyl Acetylenedicarboxylate (20)

To determine whether the reaction of triphenylcyclo-

propene (33) with eneophiles is accelerated with respect

to the reaction of a similarly substituted acyclic olefin,

1,2,3-triphenylpropene (53) was synthesized. This olefin

was chosen because it, like (33), contains a diphenyl sub-

stituted double bond and a phenyl substituent at the allylic

position, and was synthesized from phenylacetic acid as

shown.

1) PCl
PhCH2CO2H 2) AlCl3/ PhCH2-C -Ph
2) AIC1 /
(54)
Benzene











1) Mg; PhCH2Cl OH (CH3CO)2
(54) + PhCH --C-CH2Ph -
2) H302 Ph CH 3CO2H

(55)


H CH2Ph



Ph Ph

(53)

Phenylacetic acid was converted into its acid chloride

followed by Friedel-Crafts substitution on benzene, yielding

phenylbenzyl ketone (54).49 Treatment of (54) with benzyl

Grignard gave the tertiary alcohol (55), which upon dehyd-

ration afforded (53), shown by nmr spectroscopy to be a 4.25

to 1.00 mixture of the trans and cis isomers.50 This mixture

was employed without separation in the subsequent rate com-

parisons.

Dimethyl acetylenedicarboxylate (20) was chosen as the

eneophile for this investigation because its reactivity is

not so great as to preclude selectivity between the olefins

and has already been shown to react cleanly with (33).

One to one solutions of each "ene" component and (20)

were prepared in chlorobenzene with an internal standard

(tetrachloroethane) added. The reactions were carried out

at 119.90 and monitored by nmr spectroscopy, following the

disappearance of the "ene" component. The reaction of (33)

with (20) proceeded with a half-life of 0.8 hours, but









1,2,3-triphenylpropene (53) failed to show any measurable

reaction after 125 hours.

The initial concentration of each reactant in these

reactions was 0.5 M. If a minimum limit of detectability

by nmr analysis of 5% is assumed, a lower limit for the

half-life of the reaction of 1,2,3-triphenylpropene with

dimethyl acetylenedicarboxylate (20) may be estimated to

be 2400 hours. This treatment results in an estimated

rate acceleration for the reaction of triphenylcyclopro-

pene with (20) of 3000. The factors possibly responsible

for the increased reactivity of (33) with respect to (53)

toward the eneophile (20) will be discussed in Chapter III.
















CHAPTER III


KINETIC AND MECHANISTIC INVESTIGATION OF
THE "ENE" REACTIONS OF CYCLOPROPENES

Introduction

The reactivity of triphenylcyclopropene (33) with the

classic eneophiles maleic anhydride (1), diethyl azodicar-

boxylate (3), and dimethyl acetylenedicarboxylate (20)

prompted an investigation of the mechanism of these "ene"

reactions with the objective of determining whether a con-

certed or stepwise process is operating. Several approaches

were used to probe the mechanism of these reactions, in-

cluding product studies, determination of the kinetic acti-

vation parameters (E and AS ), competition reactions, sol-
a
vent effects and deuterium isotope effects.



Results

In each of the thermal reactions of triphenylcyclo-

propene (33) with eneophiles (1), (3), and (20), discussed

in Chapter II, only one "ene" adduct was isolated, although

a substantial amount of the "self ene" dimer (34) was also

formed during the reaction of (33) with (1). In this latter

case, dimerization of (33) became competitive with adduct

formation because of the substantially longer reaction time








and higher reaction temperature. The formation of a single

adduct in these reactions is consistent with either a con-

certed or a step wise hydride-transfer mechanism, since in

the latter case, a symmetrical cyclopropenium cation is in-

volved as the product determining intermediate. However,

the reaction of 1,2-diphenyl-3-(9'-fluorenyl)-cyclopropene

(47) with (20) also led to the isolation of a single product,

(48). The mechanistic implication of this result as well as

the previously stated observation that the "ene" reaction of

3-d-triphenylcyclopropene (43) with maleic anhydride gave

stereospecific cis-addition across the double bond of the

eneophile will be explored later.

Two "ene" reactions, the dimerization of (33) and the

reaction of (33) with (20) were subjected to kinetic invest-

igation. The dimerization reaction was carried out on 0.5

molar solutions of (33) in bromobenzene and the reaction of

(33) with (20) was conducted on bromobenzene solutions that

were 0.5 molar in each reactant. Both reactions were car-

ried out in sealed nmr tubes and the disappearance of (33)

relative to an internal standard, toluene, was monitored.

In this way, the concentration of (33) at a given time could

be calculated and in turn, the rate constants determined

using equation 1, the integrated rate expression for a sec-

ond order reaction in which the initial concentrations of

both reactants are equal, x being the concentration of (33)

at time t and ao being the initial concentration of (33).

k =- x_ Equation 1
a (a -x)t
0 O










From the rate constants for these reactions at two

temperatures, the energies of activation, Ea were cal-

culated using equation 2, in which kl is the rate constant

at absolute temperature TI, k2 is the rate constant at

absolute temperature T2, and R is the gas constant.



2.303 log k2 = Ea (T2-T1) Equation 2

kl R (T2T1)



Employing the energies of activation calculated from

equation 2, the entropies of activation, AS were calcu-

lated from equation 3, where AH = E -RT.
a


2.303 R log k, 47.198 = -AH + AS Equation 3

T T1
1



Using the entropy of activation, AS the pre-expon-

ential factor of the Arrhenius equation was calculated

from equation 4.


T
AS = 4.576 (logA-13.23) Equation 4



The rate constants for the dimerization of (33) and

the reaction of (33) with (20) and the accompanying E
a
AS and log A values are given in Table 1.3.










The reaction of (33) with (20) was also used to probe

the effect of solvent on the "ene" reactions of cyclopro-

penes. Thus, 0.5 molar solutions of each reactant in car-

bontetrachloride, bromobenzene, and nitrobenzene were pre-

pared and heated at 100.0. As before, the progress of the

reaction was monitored by following the disappearance of

(33) with respect to the internal standard, toluene. The

rate of reaction was found to be insensitive to changes in

solvent polarity, increasing by less than a factor of two

as the solvent was changed from carbontetrachloride to nitro-

benzene.

To determine what effect structural changes have on

this "ene" reaction, a competitive study was conducted using

the eneophile (20) and the "ene" components (33) and 1,2-

diphenyl-3-p-tolycyclopropene (56). A bromobenzene solution

containing equal molar amounts of (33) and (56) and approx-

imately 0.1 mole equivalents of (20) was heated in a sealed

tube and the relative rates of formation of the "ene" adducts

(45) and (57) were determined simply by isolating the adduct

mixture and comparing the nmr integral heights of the allylic,

methyl ester and tolyl-methyl protons.



CO Me
H R Ph CO Me

+ (20)
Ph Ph Ph RH
(33); R = Ph (45); R = Ph
(56); R = p-tolyl (57); R = p-tolyl









The ratio of tolyl-methyl protons to olefinic protons

was 1.60 and the ratio of methyl ester protons to tolyl-

methyl protons was 3.78. In a product mixture containing

equal amounts of (45) and (57), these ratios would be 1.50

and 4.00 respectively. Therefore, the adduct mixture was

found to contain 53% (57) and 47% (45). To check the ac-

curacy of this method, the ratio of methyl ester protons

to olefinic protons was measured and determined to be 6.03.

Regardless of the composition of the product mixture, the

theoretical value of this ratio is 6.00.

The rate ratio for this reaction was calculated em-

ploying equation 5, where A stands for "ene" adduct and C

stands for starting cyclopropene.



k(56) (57) C(33)
x Equation 5
k A C
(33) (45) (56)



From equation 5 and the percent composition of the

"ene" adduct mixture, k(56)/k(33) was found to be 1.13. A

one point p value was calculated for the "ene" reaction

using the Hammett equation,51 equation 6, where k = k(56)'

k = k(33), and a = -0.17 for a para-methyl substituent.52



log = op Equation 6
0


This treatment yields a value of p = -0.31. It may be more

appropriate to employ a for the reaction at hand,










since the reacting center at C3 of the cyclopropene ring

can interact directly with the C3-phenyl substituent.53

For a para-methyl substituent, + has a value of -0.3150

and in this case, p = -0.17.

The intermolecular isotope effect was determined by

reacting an equalmolar mixture of triphenylcyclopropene (33)

and 3-d-triphenylcyclopropene (43) with approximately 0.1

mole equivalents of dimethyl acetylenedicarboxylate (20).

After isolating the "ene" adduct, the composition of the

product mixture was determined by nmr integration of the

vinyl and methyl ester protons. The starting cyclopropenes

were also isolated and the isotopic composition determined

by comparing the nmr integration of the cyclopropenyl C3-

hydrogen with that of an accurately measured amount of

toluene. This mixture was then used to obtain a duplicate

value of the intermolecular isotope effect. The rate ratio

is given by equation 7, where A stands for the "ene" adduct

and C stands for the starting cyclopropene.



k A C
k (H) A(H) C(43)
k(H) A x () 3 Equation 7
(D) "(D) C(33)





The intermolecular isotope effect k(H)/k( was found to be
1.25 + 0.03.
1.25 0.03.





50






t i * -H

0 m


O 0 04

Sdo
(1) 4






O
+ m + 0



o o 0 0 0
u 0oH o0



0 -
co o . J
S 0 0 0



O 4 + 0 + o
.q N +N













O I
0 H OH H
S o0 H co0 -N
0 >I
< LN o- 0
in U O' m OO co D V


Hl C-C- 10 0O












S4
O n
Hl CC' H h- u in r-' co rmH
U I * * * * ,
0 0 H in C -N Nn o A-
o j +1+1 +1+1










Q 0 L +1+1 1+10 0
S N 00 C N M
H A C. * in .0 0r-i -1 C- 0-H










C-)
HL 0




o >1'0







0lrl u
I-I0










S4-1 0
0 OH

0 0










Discussion

The results obtained from the investigation of the

mechanism of the "ene" reactions of cyclopropenes point

toward a concerted rather than a stepwise process. The for-

mation of a single "ene" adduct from the reactions of (33)

with the eneophiles (1), (3), and (20) is inconclusive,

but the isolation of a single "ene" adduct from the re-

action of (47) with (20) does appear to support a concerted

process. If this reaction were preceding by complete hyd-

ride transfer to the eneophile, the unsymmetrically sub-

stituted cyclopropenium cation (58) would be formed as the

product determining intermediate. Subsequent combination

of (58) with the corresponding anion would be expected to

yield two products, (48) and (59).


Ph Ph

H CO2Me H CO2Me
SH CO2Me C02

CO 2 Me CO Me
H CO2Me 2
Ph R
(48); R = Ph
R1 = 9-fluorenyl

(59); R = 9-fluorenyl
R1 = Ph


Charge distribution and steric requirements in (58)

might favor combination at one of the phenyl-substituted

carbons to give (48), but complete absence of the thermody-










namically more stable product (59) appears unlikely in

light of the fact that unsymmetrically substituted cyclo-

propenium salts containing three substituents other than

hydrogen yield mixtures of cyclopropene products when

treated with sodium borohydride, methyl lithium, or methyl

magnesium iodide.54 For example, diphenyl-methylcyclopro-

penium perchlorate, upon treatment with methyl lithium at

-70, gives approximately equal amounts of 1,2-diphenyl-

3,3-dimethylcyclopropene and 1,3-diphenyl-2,3-dimethylcyc-

lopropene, while the same salt with methyl magnesium iodide

gives the same products in a seven to one ratio.54 Sodium

borohydride reduction of diphenyl-p-tolycyclopropenium per-

chlorate yields a mixture of 1,3-diphenyl-2-p-tolycyclopro-

pene and 1,2-diphenyl-3-p-tolycyclopropene in which the

former predominates.54

The stereospecific cis-addition of 3-d-triphenylcyclo-

propene (43) to maleic anhydride also supports a concerted

mechanism. A stepwise process, involving hydride transfer,

might be expected to yield a mixture of products resulting

from cis- or trans- addition to maleic anhydride, represented

by paths a and b respectively.


Ph Ph

Ph 0 -
a b H
(44) 0- H 0

Ph h H
O









Although stereospecific addition is necessary for a

concerted process, it is not sufficient evidence to prove

such a mechanism since carbon-carbon bond formation yield-

ing a zwitterionic intermediate followed by a kinetically

preferred cis-hydrogen transfer would also account for the

observed cis-addition. However, such an intermediate (60),

would contain a cyclopropyl cation and the propensity of

such systems to undergo disrotatory ring opening to give

allylic cations is well documented.44'55 Rapid cis trans-

fer of deuterium in (60) would account for the observed pro-

ducts but one might question why products such as (62) aris-

ing from collapse of a zwitterionic intermediate of the type

(61) are not observed if this mechanism is operating. Of

course, stabilization of the positive charge by the phenyl

substituent at the carbonium ion site in (60) could suf-

ficiently slow ring opening to (61), and if hydrogen trans-

fer were extremely rapid, this could account for the lack

of other products.




Ph
Ph
+ D

D Ph +D 0 Ph
Ph Ph


Ph Ph O


(60)


(61)


(62)








Likewise, hydride transfer from the cyclopropene to

the eneophile, resulting in formation of a tight ion pair

which could then collapse both regiospecifically and stereo-

specifically, is also consistent with formation of a single

product from the unsymmetrical cyclopropene (47) and the

stereospecific cis-addition of (33) and (43) across the

double bond of maleic anhydride.

Even though the zwitterionic mechanism or tight ion

pair mechanism could explain the formation of a single pro-

duct in the reactions of unsymmetrical cyclopropenes with

eneophiles, the negligable solvent effects observed for the

reaction of (33) with dimethyl acetylenedicarboxylate (20)

argues strongly against mechanisms involving significant

charge generation. For reactions proceeding from neutral

reactants to charged intermediates, it may be said that the

more polar the solvent, the faster the reaction will be,

since the intermediate ion or ion pairs will be better sol-

vated by more polar solvents. The same is true if there is

considerably more charge generated in the transition state

than in the reactants; enhanced solvation of the transition

state leading to a lowering of the energy of activation,

which is reflected in an enhance rate. Since no significant

rate increase was observed in the reaction of (33) with (20)

when the solvent was changed from carbontetrachloride to

bromobenzene to nitrobenzene, the dielectric constants of

which are 2.22, 5.40, and 34.8,56 respectively, a reaction

pathway involving generation of a cyclopropenium cation or









zwitterionic intermediate is very unlikely. However, a

concerted process involving some degree of charge separ-

ation in the transition state is not incompatible with this

result since any positive charge generated in the cyclopro-

pene moiety of the transition state would be dispersed

throughout the ring, thus decreasing the demand for ex-

ternal solvation by polar solvents. Huisgen and Pohl have

also investigated the effect of solvent on the "ene" re-

action of 1,3-diarylpropenes with diethyl azodicarboxylate

(3)27 and observed only a four-fold rate enhancement on

changing the solvent from cyclohexane to nitrobenzene.

This result, coupled with product studies, led to the con-

clusion that a concerted process was operating.

The results from the competitive study involving the

"enes" (33) and 1,2-diphenyl-3-p-tolylcyclopropene (56) and

the eneophile (20) are also supportive of a concerted mech-

anism. The observed rate ratio, k(56)/k(33 = 1.13, indi-

cates that this reaction is only slightly sensitive to sub-

stituent effects. A single point Hammett p value of -0.31

is determined for the "ene" reaction of 1,2-diphenyl-3-

arylcyclopropenes with dimethylacetylene dicarboxylate (20)

This value is consistent in both magnitude and sign with

reactions involving cyclic transition states. For example,

the rearrangement of p-substituted cinnamyl ethers at 1800

exhibits a p value of -0.4057 and the Diels-Alder reaction

of 1-aryl-substituted butadienes with maleic anhydride at

450 displays a p value of -0.61.58









The small p value observed for the "ene" reaction leads

to the conclusion that there is little charge development

in the transition state. This conclusion may be biased,

however, because any charge generated at C3 of the cyclo-

propene in the transition state will not be localized at

this carbon, but will be dispersed throughout the cyclo-

propene ring, thus decreasing the effect of substituents.

Although additional substituents should be examined to ob-

tain an accurate p value, there is no reason to believe

that the one point p value determined for this "ene" re-

action is not of the proper magnitude and sign.

The large negative entropies of activation observed

for the dimerization of (33) and the reaction of (33) with

(20), -36.7 eu and -25.6 eu, respectively, are also con-

sistent with a concerted process. These values are in good

agreement with the AS value of -36 eu observed for the

Diels-Alder reaction of cyclopentadiene with maleic anhyd-

ride,59 and the AS values of -40.7 and -31.8 eu obtained

by Franzus for the "ene" reactions of 1,3- and 1,4-cyclo-

hexadienes respectively with diethyl azodicarboxylate.26

Although concerted, bimolecular processes are accom-

panied by large, negative entropies of activation, these

values are, of themselves, insufficient to prove such a

mechanism since reactions proceeding from neutral reactants

to dipolar or ionic intermediates also exhibit similar en-

tropy effects due to increased orientation of the surround-

ing solvent molecules. Of course, reactions proceeding









through a dipolar or ion pair intermediate would be sensi-

tive to changes in solvent and that is not the case in

these "ene" reactions.

Although the isotope effect, k(H)/k(D) = 1.27, ob-

served for the reaction of (33) with (20), is relatively

small in comparison to the intramolecular value of 2.8 re-

ported by Huisgen27 for the reaction of 1,4-dideuterio-

1,4-dihydronaphthalene with diethyl azodicarboxylate or

the intermolecular value of 1.99 reported by Dai and Dolbier29

for the reaction of allene with perfluorocyclobutanone,

the most satisfactory rationalization seems to require a

concerted mechanism.

The primary isotope effect arises largely from the

difference in zero-point energy between a bond to hydrogen

and a corresponding bond to deuterium. The major portion

of the kinetic isotope effect results from changes in the

zero-point energy which occurs when reactants are converted

into an activated complex. The difference in zero-point

energy for C-H and C-D bonds is estimated from their corres-

ponding zero-point vibration energy to be roughly 1.15

kcal/mole. If this stretching vibration is lost in the

transition state, then the difference in Ea for the C-H

bond and C-D bond will be 1.15 kcal/mole, which at 2980C

corresponds to a rate factor of about seven. Although this

is a somewhat simplified treatment, experimental data con-

firm that, for reactions involving a linear, symmetrical

transition state, isotope effects of five to seven are

generally observed.









Obviously, the primary isotope effect of 1.25 observed

for the "ene" reaction of (33) with (20) is smaller than the

maximum effect, so that the reaction cannot proceed through

a linear symmetrical transition state. Examples of small

primary isotope effects are numerous and rationalization of

these low values generally consider three possibilities.

First, the maximum isotope effect calculated from the

zero-point energies of the C-H and C-D bonds assumes a sym-

metrical transition state. If the C-H bond is only slightly

perturbed at the transition state and the stretching vibra-

tion is still strong, then the effect of deuterium substi-

tution on the rate will be much smaller. The effect would

be similar in the event of a very late, or product-like

transition state in which the new bond to hydrogen is almost

completely formed. This dependence of isotope effect on

the position of the transition state is nicely demonstrated

by the variable isotope effects observed for base-catalyzed

proton removal from carbon, the variations rationalized in

terms of the differences in basicity between the proton don-

ors and proton acceptors.60,61

Secondly, a significant primary isotope effect will

be observed only if a C-H bond is broken in the rate-deter-

mining step.62 Slow formation of an intermediate followed

by fast removal of a proton in the product-determining step

usually results in a negligible intermolecular effect.

Finally, a broad group of reactions in which primary

isotope effects are small are those proceeding via non-









linear transition states.63 Examples of this class include

low isotope effects observed in 1,2-hydrogen shifts,64

E2 eliminations,65 and other hydrogen transfer processes

proceeding via three,66 four, five, and six membered cyclic

transition states. Theoretical explanations for and cal-

culations of these low isotope effects observed for re-

actions proceeding through non-linear transition states

have been discussed by O'Ferrall.63 A qualitative agree-

ment was obtained by correlating the isotope effect with

the 6-angles (63) of hydrogen migration in the transition

state, the magnitude of the effect increasing from 0.9 to

3.0 as the B-angle increased from 60 to 1200. These ef-

fects were calculated assuming equality of bond breaking

and bond formation in the transition state. Since the con-

certed "ene" mechanism proceeds through a six-membered cyc-

lic transition state, small intermolecular isotope effects

should be observed.




H

----C C --


(63)



Considering the first possibility, a linear, unsym-

metrical abstraction of hydrogen from (33) by dimethyl acety-

lenedicarboxylate (20) appears unlikely. Abstraction of a

hydrogen from (33) should result in the formation of two









radical species (64) and (65). Although combination of

these radicals could account for the observed product, one

might question why only one product is observed when unsym-

metrical cyclopropenes are employed. Also, (64) is much

less stable than the corresponding cyclopropenium cation,

owing to the presence of the odd electron in an anti-bond-

ing molecular orbital,67 and is known to undergo fast di-

merization to bis-triphenylcyclopropene.68 None of this

product or the subsequent rearrangement product, hexa-

phenylbenzene was detected in this reaction. Further,

Breslow has previously demonstrated that the dimerization

of (33) via the "ene" reaction did not proceed by a radical

mechanism.38





Ph H CO2Me

+ (45)
Ph Ph CO2Me

(64) (65)



A stepwise mechanism is more attractive than the rad-

ical mechanism, but as previously discussed, still cannot

account for all of the experimental observations, especially

the negligible solvent effects.

The mechanism most compatible with all the experimen-

tal results is a concerted one, involving a cyclic transition





61





state similar to that first described by Arnold6 to account

for the condensation reactions of maleic anhydride with

olefins. This process should not be interpreted as being

completely synchronous; indeed, some charge development in

the transition state might be anticipated and as previously

discussed, would not be incompatible with the negligible

solvent effects or small p-value observed in these "ene"

reactions. Furthermore, the intermolecular kinetic iso-

tope effect of 1.25 anticipates, by the treatment of

O'Ferrall,63 a S-angle of approximately 80. This angle

seems to be rather small for these reactions but to the ex-

tent that hydride character is developed in the transition

state, that is bond breaking and bond formation become un-

equal, a larger and more reasonable B-angle would be pre-

dicted. The transition state (66), proposed for these re-

actions, is given below.




Ph

p


Ph
0 0 H


0.7,










One might predict, from the mechanistic picture pro-

posed here, that the ability of an electrophilic reagent

to stabilize any negative charge developed in the trans-

ition state would be a factor in determining its ability

to participate in the "ene" reaction involving the cyclic

mechanism depicted on the proceeding page. Indeed, this

appears to be the case; maleic anhydride requiring higher

reaction temperatures and longer reaction times than di-

methyl acetylenedicarboxylate, which in turn is less re-

active than diethyl azodicarboxylate.

Having concluded that the thermal "ene" reactions of

the cyclopropenes (33) and (47) proceed via a concerted

mechanism, the enhanced reactivity of triphenylcyclopropene

(33) with respect to 1,2,3-triphenylpropene (53) toward the

eneophile dimethyl acetylenedicarboxylate (20) must be

rationalized.

One explanation for the disparity in rates may re-

sult from the differences in the orientation of the re-

acting centers in these molecules. Both Hill and Arnold

have recently demonstrated that the preferred orientation

of the "ene" component is the one in which the allylic car-

bon-hydrogen bond is perpendicular to the carbon-carbon

double bond.531 In (33), this geometry is inherent, a

consequence of the rigidity of the three membered ring sys-

tem. In (53), however, models indicate that a perpendicular

arrangement of either of the allylic hydrogens with the

double bond results in non-bonded interactions between the










phenyl substituents and that these interactions are at a

minimum when the allylic phenyl substituent is perpendicular

to the double bond.

A second factor which may increase the reactivity of

(33) with respect to (53) is the disposition of the phenyl

substituents bound to the carbon-carbon double bond. In-

ternal bond compression in the cyclopropene ring results

in an external 0 (C=C-Ph) angle of about 1500, compared

to the estimated 0 (C=C-Ph) angle of 1200 for (53). This

increased angle results in a greater intramolecular distance

between the phenyl substituents, allowing them to assume a

more planar alignment with the double bond in (33) than is

possible in (53). Thus, in (53), where the phenyl substi-

tuents are more skewed, steric inhibition to approach by

the eneophile could contribute to the decreased reactivity

of this olefin.

An alternate explanation may be found in the electronic

nature of the double bonds of the "ene" components. Elec-

tronic effects on the rates of the related Diels-Alder re-

action have been extensively investigated. The general

rule, originally proposed by Alder,70 that electron with-

drawing substituents at the double bond of the dieneophile

enhances the rate of the Diels-Alder reaction, has firm

experimental support. Sauer, for example, has found that

the rate of reaction of cyclopentadiene with dieneophiles

increases by a rate factor of greater than 107 in going

from acrylonitrile to tetracyanoethylene.71










Recently a second class of Diels-Alder reactions, in

which the electron demand is reversed, have been investi-

gated. In these reactions, electron donating substituents

at the double bond of the dieneophile enhance its reactivity

toward electron deficient dienes. The reactivity of the

electron-poor diene hexachlorocyclopentadiene with various

dieneophiles demonstrates this type of electron demand.

Thus, Sauer has found that the electron-rich system, cyclo-

pentadiene, acting as a dieneophile, reacts by a factor of

about 103 times faster with this diene than does the electron-

poor dieneophile maleic anhydride.72 In addition, changes

in the electronic nature of the diene has a pronounced ef-

fect on the rate of Diels-Alder reactions with a given di-

eneophile, as shown by the fact that maleic anhydride re-

acts 5 x 104 times faster with 9,10-dimethylanthracene than

with hexachloropentadiene.72

A qualitative explanation for these two types of Diels-
73
Alder reactions has been offered by Sustmann. In the

normal Diels-Alder reaction, the prominent bonding inter-

action is between the highest occupied molecular orbital

(HOMO) of the diene and the lowest unoccupied molecular orb-

ital (LUMO) of the dieneophile. By introducing electron

withdrawing substituents in the dieneophile, and thus low-

ering its orbital energies, this interaction is enhanced.

Likewise, electron donating groups in the diene will raise

its orbital energies and again produce an increased inter-

action between its (HOMO) and the dieneophile (LUMO). This










interaction is increased more rapidly than the concomitant

decrease in the interaction of the diene (LUMO) and diene-

ophile (HOMO), resulting in a net energy gain. In the

inverse electron demand case, where the diene is now electron

deficient and the dieneophile electron rich, just the op-

posite interactions predominate. Now, the major bonding

interaction involves the diene (LUMO) and the dieneophile

(HOMO). Electron donating groups in the dieneophile increase

this interaction as do electron withdrawing groups in the

diene.

One class of dieneophiles that react readily with

electron-poor dienes is the cyclopropene ring system.

Battiste and co-workers74 have shown that triphenylcyclo-

propene (33) reacts smoothly with the electron deficient

dienes tetraphenylcyclopentadienone and 3,6-disubstituted

tetrazines. Although strain relief undoubtably plays an

important part in these facile transformations, it cannot

explain the differences in reactivities of cyclopropenes with

these dienes. Whereas (33) reacts conveniently with 3,6-di-

phenyltetrazine in refluxing benzene in 80% yield, 3-carbo-

methoxy-1,2-diphenylcyclopropene gives only 32% reaction

with the same diene after refluxing in benzene for fourteen

days. Replacing the C3-phenyl substituent in (33) with the

electron withdrawing carbomethoxy substituent greatly re-

duces the reactivity of the cyclopropenyl double bond, dem-

onstrating that the electron-rich nature of this double bond

plays an important factor in its ability to interact with

electron-poor dienes.










Photoelectron spectroscopy provides a quantitative

measure of the orbital energies of the highest occupied

molecular orbitals of unsaturated bonds. The ionization

potentials for the compounds 1,2-diphenylcyclopropene,

trans-stilbene, and cis-stilbene are 7.57,75 .00,76 and

8.2076 ev respectively. The (HOMO) of 1,2-diphenylcyclo-

propene is, therefore, about 0.4 ev higher in energy than

that of trans-stilbene and about 0.6 ev higher than that

for cis-stilbene, and as a result, is better able to mix

with the (LUMO) of electron-poor dienes.

The "ene" reaction is also subject to the same elec-

tronic influences found operating in the Diels-Alder reac-

tion. Eneophiles bearing two electron withdrawing groups

are more reactive than those containing only one such group

and electron-rich "ene" components show greater reactivity

than their electron-poor counterparts. In their "ene" re-

actions cyclopropenes participate as the "ene" component

and therefore, the electron-rich nature of the cyclopro-

penyl double bond relative to the double bond of propenes,

should facilitate reaction with electron-poor eneophiles

just as electron-rich dienes are facilitated in their re-

actions with electron-poor dieneophiles.

The effects described above may be considered to be

primarily identified as structural effects. The operation

of specific stabilizing effects in the transition state (66)

for reaction of (33) with (20) may also be considered.

Relative to the transition state (67), involved in the










reaction of (53) with (20), (66) contains an extra bonding

interaction in the "ene" component of the transition state

that should provide added stabilization. Also, to the ex-

tent that charge is developed in the two transition states,

the cyclopropenium character in (66) will allow greater

stabilization than in (67). In the extreme case of com-

plete hydride transfer from the "ene" component to the

eneophile, this added stabilization would be the differ-

ence in delocalization energy between a triphenylcyclo-

propenium cation and a 1,2,3-triphenylallylic cation.


Ph


PhPh
Ph---

'H






(67)


Although it is difficult to access which, if any, of

the aforementioned steric and electronic factors is most

responsible for the enhanced reactivity of triphenylcyclo-

propene (33) toward dimethyl acetylenedicarboxylate (20),

their cumulative effect at least offers a reasonable ex-

planation for the observed results.















CHAPTER IV


EXPERIMENTAL



Physical Measurements

General

Melting points were taken on a Thomas-Hoover capillary

melting point apparatus. All melting and boiling points

are reported uncorrected.

Elemental analyses were performed by Atlantic Microlab,

Incorporated, Atlanta, Georgia.



Spectra

Infrared spectra were recorded on a Perkin-Elmer 137

Infrared spectrophotometer as potassium bromide pellets or

liquid films on sodium chloride or silver chloride plates.
-i
Infrared absorptions are given in cm and are described as:

w = weak, m = medium, s = strong, vs = very strong, sh =

sharp, and b = broad.

Ultraviolet spectra were recorded on a Cary Model 15

recording spectrophotometer, using Beckman 1 cm cells. Re-

ported maxima are given in nanometers and are followed by

the extinction coefficient, (e).

All nuclear magnetic resonance spectra were recorded in









deuterochloroform (CDCl3) with tetramethylsilane (TMS) as

the internal standard unless otherwise stated. All spectra

were obtained using a Varian A-60A or Varian XL-100 spect-

rometer.

lass spectra were obtained on a Hitachi Perkin-Elmer

RMU-6E instrument or on an AEI MS-30 high resolution instru-

ment at an ionizing voltage of 70 electron volts.



Reactions

Diphenylacetylene (39)43

A solution of trans-stilbene (45.0 g, 0.25 moles) in

750 ml of ether was prepared in a 1-liter three-necked

round-bottomed flask equipped with a mechanical stirrer,

reflux condenser, and dropping funnel. Bromine (14 ml, 43 g,

0.27 moles) was added dropwise over fifteen minutes to the

stirred trans-stilbene solution. Stirring was continued

for an additional hour, and the precipitate collected on a

Buchner funnel and washed with ether until the washing were

colorless. After air drying, 1,2-dibromo-l,2-diphenylethane

(69.0 g, 80%) was obtained as a white solid, mp 234-2360

(Lit.43 mp 235-237').

A solution of potassium hydroxide (90 g, 0.62 moles)

in 150 ml of absolute ethanol was prepared in a 500-ml

round-bottomed flask equipped with a reflux condenser and

heated with an oil bath maintained at 130-1400. The di-

bromostilbene prepared above was added in small portions,

replacing the reflux condenser after each addition. The

reaction solution was heated at 130-1400 for 24 hours and










then poured into 750 ml of cold water. The product was

collected on a filter, washed with water, and dried over

calcium chloride in a vacuum desiccator. Crude diphenyl-

acetylene was obtained as a brown solid (37.8 g, 83%)

and was dissolved in a minimum amount of hexane and chrom-

atographed over a column of florex eluting with hexane.

After removing the elutant under reduced pressure the

residue was recrystallized from absolute ethanol, yielding

diphenylacetylene (39) (31.7 g, 70%) as white prisms,

mp 60-62 (Lit.43 mp 60-61).



1,2,3-Triphenylcyclopropenium

bromide (40 )42

A solution of diphenylacetylene (39) (23.0 g, 0.129

moles) and potassium tert-butoxide (70.1 g, 0.620 moles)

in 1500 ml of freshly distilled benzene (from P205) was

prepared in a 3-liter three-necked round-bottomed flask

equipped with a mechanical stirrer, dropping funnel, reflux

condenser and nitrogen purge. Freshly distilled benzal

chloride (48.4 g, 0.305 moles) in 150 ml of dry benzene was

added dropwise over a period of 1 hour. The reaction mix-

ture was refluxed for five hours, cooled to room temper-

ature and quenched with 300 ml of water. The benzene

layer was separated and dried over magnesium sulfate and

the aqueous layer was extracted with ether and the extracts

dried over magnesium sulfate. The benzene solution was

filtered and concentrated to 200 ml, the ether extracts









filtered into the concentrated benzene solution and the

resulting mixture diluted to 1500 ml with ether. The

ether-benzene mixture was stirred and saturated with a

stream of hydrogen bromide gas. The precipitate that

formed was collected on a filter, washed with ether and

dried under vacuum, yielding triphenylcyclopropenium bro-

mide (40) (24.2 g, 54%) as a yellow-brown powder, mp 253-

256 .



1,2,3-Triphenylcyclopropene (33)42

1,2,3-Triphenylcyclopropene was prepared from triphenyl-
42
cyclopropenium bromide by the method of Battiste. A

suspension of triphenylcyclopropenium bromide (38.3 g, 0.110

moles) in absolute ethanol (200 ml) was prepared in a 1-

liter three-necked round bottomed flask equipped with a

dropping funnel, mechanical stirrer and nitrogen purge. A

suspension of sodium borohydride (5.5 g, 0.15 moles) in

100 ml of absolute ethanol was added dropwise over a period

of 1 hour. The reaction mixture was stirred for an addition-

al three hours, cooled to 0, and quenched with 250 ml of

water. The product was collected on a filter, washed twice

with cold 95% ethanol and air dried. The crude (33) thus

obtained was placed on a column of alumina and eluted with

hexane. After removing the elutant under reduced pressure,

(33) (28.5 g, 96%) was obtained as a white solid, mp 107-

1090. Recrystallization from ethanol yielded (33) (26.9 g,

91%) as white plates, mp 110-1110 (Lit.38 mp 110-112).








3-(1',2',3'-Triphenylcyclo-

propyl)-1,2,3-triphenylcyclopropene (34)3

A solution of triphenylcyclopropene (33) (0.268 g,

1.00 mmoles) in chlorobenzene (2 ml) was heated at reflux

with magnetic.stirring for 44 hours. After cooling the

reaction mixture to room temperature, the solvent was re-

moved under a stream of nitrogen and the oily residue that

remained was dissolved in a 1:1 benzene-hexane solution

and cooled. The precipitate that formed was collected on

a filter, washed with hexane and recrystallized from the

same solvent system. The precipitate was again collected

on a filter, washed with hexane and dried under vacuum,

yielding (34) (0.092 g, 34%) as colorless prisms, mp 179-

1800 (Lit.38 mp 178.5-181).



l-(l',2',3'-Triphenylcyclopropenyl)-

succinic anhydride (41)

Triphenylcyclopropene (33) (0.268 g, 1.00 mmoles),

freshly sublimed maleic anhydride (0.098 g, 1.00 mmoles)

and bromobenzene (5 ml) were stirred at 1300 for 40 hours.

After cooling to room temperature, the reaction mixture

was examined by thin layer chromatography. Several com-

ponents were detected and separation by column chromato-

graphy was attempted. The solvent was evaporated under a

stream of nitrogen and the reaction residue was placed on

a column of silica gel and eluted with hexane, 4:1 hexane-

benzene, benzene, and 7:3 benzene-chloroform. The collected









fractions (30 ml) were combined on the basis of their thin

layer behavior. The first three components to elute were,

in order of elution, (33) (0.043 g, 12%), 1,2-diphenyl-

indene (42) (0.082 g, 22%), and dimer (34) (0.060g, 16%),

identified by their nmr spectra and melting points. The

final component to elute was identified as the "ene"

adduct (41). Recrystallization from benzene-petroleum

ether yielded colorless plates (0.147 g, 40%), mp 171-1730.

The nmr spectrum (CDC13) showed an aromatic multiple

at S8.00-7.30(15H) and an ABX multiple characterized by

the chemical shifts 4.30 (1H), 3.07 (1H), and 2.97 (1H),

and the coupling constants J = 18.5, J = 6.9, and
gem cis
Jrans = 9.1 Hz.

The infrared spectrum (KBr) showed absorptions for

carbon-hydrogen stretch at 3080 (w), 3060 (w), and 3030 (w),

carbon-oxygen double bond stretch at 1860 (s) and 1760 (vs),

along with other absorptions at 1600 (m), 1492 (s), 1445 (s),

1415 (w), 1220 (s), 1072 (s), 1050 (s), 900 (m), 758 (s),

733 (s), 690 (s), and 630 (w).

The mass spectrum showed a molecular ion at m/e

366.1263 (21%),(calcd., 366.1255), a P+l ion at m/e

367.1284 (5%), (calcd., 367.1288), and other abundant

fragments at m/e 325 (base peak), 267 (80%), and 105 (50%).

The ultraviolet spectrum (95% ethanol) showed maxima

at 330 (19,800), 314 (24,700), 295 (18,800) along with

strong end absorption.










cis-l-(1',2',3'-Triphenylcyclopro-

penyl)-2-d-succinic anhydride (44).

3-d-l,2,3-Triphenylcyclopropene (43) (0.269 g, 1.00

mmoles), freshly sublimed maleic anhydride (2.94 g, 30.0

mmoles), and chlorobenzene (10 ml) were heated at reflux

for 44 hours. The reaction mixture was cooled to room

temperature and examined by thin layer chromatography.

Several components were detected and separation of the

"ene" adduct from the hydrocarbon products by column

chromatography was attempted. The solvent was removed

under a stream of nitrogen and the reaction residue placed

on a column of silica gel. The hydrocarbon products were

eluted with 9:1 hexane-benzene and when thin layer exam-

ination revealed that no more material was being eluted,

the elutant was changed to 6:4 hexane-chloroform. The

collected fractions (30 ml) were combined on the basis of

their thin layer behavior. The solvent was removed under

reduced pressure and the residue was recrystallized from

benzene-petroleum ether, yielding (44) (0.161 g, 43%) as

colorless plates, mp 174-175.

The nmr spectrum (CDCl3) showed an aromatic multi-

plet at 7.90-7.15 (1511), a sharp doublet at 4.28 (1H,

J=10.0 Hz), and a broad doublet 3.05 (1H, J=10.0Hz).

The infrared spectrum (KBr) showed carbon-hydrogen

stretch at 3050 (w), carbon-oxygen stretch at 1860 (s)

and 1780 (vs) and other absorptions at 1490 (m), 1445 (m),









1220 (m) 1090 (m) 1075 (m), 930 (s), 915 (s) 880 (m),

792 (m), 775 (m), 760 (s), 733 (m), and 690 (s) cm- .

The ultraviolet spectrum (95% ethanol) showed maxima

at 330 (19,800), 314 (23,600), and 295 nm (18,600) along

with a strong end absorption.



Dimethyl-l-(l',2',3'-triphenylcyclo-

propenyl)-maleate (45)

Triphenylcyclopropene (0.268 g, 1.00 mmoles) and

freshly distilled dimethyl acetylenedicarboxylate (20)

(0.710 g, 5.00 mmoles) were dissolved in chlorobenzene

(15 ml) and stirred at 1000 for 42 hours. The reaction

mixture was cooled to room temperature and examined by

thin layer chromatography. In addition to starting material,

a new component was detected and separation by column

chromatography was attempted. The solvent was removed

under a stream of nitrogen and the reaction residue was

placed on a column of silica gel and eluted with hexane.

This failed to elute any material and the elutant was

changed to benzene. The collected fractions (30 ml) were

combined on the basis of their thin layer behavior. Re-

moval of the solvent resulted in an oil which was taken

up in benzene-petroleum ether and cooled. The precipitate

was collected on a filter, washed with cold petroleum ether

and air dried, yielding the adduct (45) (0.233 g, 57%) as

colorless prisms, mp 121-122.

Anal. Calcd. for C27H2204: C, 79.00: H, 5.40.

Found: C, 78.99: H, 5.44.









The nmr spectrum (CDC13) showed aromatic multiplets

at 17.95-7.65 (4H) and 7.65-7.11 (1111), and sharp singlets

at 5.93 (1H), 3.65 (31H), and 3.50 (3H).

The infrared spectrum (KBr) showed carbon-hydrogen

stretch at 3030 (w), 2990 (w), and 2930 (w), cyclopro-

penyl carbon-carbon double bond stretch at 1840 (w), car-

bon-oxygen double bond stretch at 1730 (s) and 1710 (s),

and carbon-carbon double bond stretch at 1630 (m), along

with other absorptions at 1495 (m), 1440 (s), 1430 (s),

1345 (s), 1275 (s), 1240 (m), 1195 (s), 1170 (s), 1110 (s),

1035 (m), 976 (m), 879 (m), 866 (m), 791 (m), 767 (m),

753 (s), 734 (m), 706 (s), and 683 (s) cm-1.

The mass spectrum showed a molecular ion at m/e

410 (7%), along with other abundant fragments at m/e 319

(15%), 318 (25%), 292 (26%), 291 (73%), 290 (38%), 288

(base peak), 276 (18%), 267 (8%), 215 (29%), and 77 (17%).

The ultraviolet spectrum (95% ethanol) showed maxima

at 330 (16,000), 314 (17,600), 285 (13,600), 235 (23,100)

and 228 nm (23,600).



Ethyl hydraazodicarboxylate46

Hydrazine hydrate (20 g, 0.42 moles) and 95% ethanol

(250 ml) were placed in a 1-liter three-necked round-

bottomed flask equipped with two dropping funnels and a

thermometer and cooled to 100 with an ice water bath. One

dropping funnel was charged with ethyl chloroformate (87 g,

0.80 moles) and the other with potassium carbonate (55 g,









0.43 mmoles) and water (250 ml). Ethyl chloroformate was

added dropwise with stirring while maintaining the temper-

ature of the reaction solution between 15-200. After ap-

proximately one-half of the ethyl chloroformate had been

added, simultaneous dropwise addition of the potassium

carbonate solution was begun, always maintaining an excess

of ethyl chloroformate and a temperature below 200. During

the addition, a white solid formed. After all the reactants

had been added, the precipitate on the upper walls of the

flask was washed down with water and the reaction mixture

was stirred for an additional hour. The solid was then

collected on a filter, washed with cold water and dried

in an oven at 800, yielding ethyl hydraazodicarboxylate

(43.1 g, 61%) as a white solid, mp 130-132 (Lit.46 mp

131-1330)



Ethyl azodicarboxylate (3)46

Ethyl hydraazodicarboxylate (43 g, 0.24 moles) and

70% nitric acid (100 ml) were placed in a 500-ml three-

necked round-bottomed flask equipped with a thermometer,

mechanical stirrer and gas outlet tube and cooled to 50

with an ice-water bath. Cold, fuming nitric acid (100 ml)

was added and the reaction mixture stirred at 50 for 2

hours. The dark orange reaction mixture was then cautiously

stirred into a mixture of ice (250 g), water (250 ml) and

methylene chloride (250 ml). The organic layer was separ-

ated and the acid layer was extracted with three 100 ml









portions of cold water, stirred for ten minutes with cold

10% potassium carbonate solution (300 ml), washed twice more

with cold water and dried over magnesium sulfate. After

filtering, the solvent was removed under reduced pressure

and the reaction residue was vacuum distilled. Ethyl azodi-

carboxylate (24.1 g, 58%) was collected as an orange oil at

106-111/15 mm.



Diethyl-l-(1'2'3'-triphenylcyclopro-

peneyl)-hydraazodicarboxylate (46)

Triphenylcycloproprene (0.268 g, 1.00 mmoles) and

diethyl azodicarboxylate (0.174 g, 1.00 mmoles) were dis-

solved in toluene (20 ml) and heated at reflux for 14 hours.

The reaction mixture was cooled to room temperature and

examined by thin layer chromatography. Two components were

detected, the faster moving of which was identified as (33).

Separation was attempted by fractional crystallization. The

solvent was removed under a stream of nitrogen and the res-

idue was taken up in methylene chloride-petroleum ether and

cooled. The precipitate was collected on a filter, washed

with cold petroleum ether and air dried, yielding (46)

(0.314 g, 71%) as a white powder, mp 138.5-141. Recrystal-

lization from ethanol-petroleum ether gave (46) (0.298 g,

67%) as small white prisms, mp 140-141.

Anal. Calcd. for C27H26N204 : C, 73.28: H, 5.92:

N, 6.30.

Found: C, 73.24: H, 5.93: N, 6.41.










The nmr spectrum (CDC13) showed aromatic multiplets

at 7.90-7.60 (411), and 7.60-7.00 (11H), a sharp singlet

at 6.78 (1H), overlapping quartets at 4.11 (2H, J = 7.0 Hz),

and 4.01 (2H, J = 7.0 Hz), and two triplets at 1.08 (31,

J = 7.0 Hz) and 1.03 (311, J = 7.0 Hz).

The infrared spectrum (KBr) showed absorptions for

nitrogen-hydrogen stretch at 3230 (s), carbon-hydrogen

stretch at 2910 (m), cyclopropenyl carbon-carbon double

bond stretch at 1820 (w), carbon-oxygen double bond stretch

at 1695 (s), along with other absorptions at 1510 (s),

1470 (m), 1425 (m), 1375 (m), 1350 (m), 1250 (s), 1155 (m),

1053 (s), 768 (m), 771 (s), 761 (s), 741 (m), 710 (s),

and 693 (s) cm-1.

The mass spectrum showed a molecular ion at m/e

442.5 (0.2%) along with other abundant fragments at m/e

325 (16%), 297 (16%), 280 (22%), 268 (24%), 267 (base peak),

77 (16%), and 29 (47%).

The ultraviolet spectrum (95% ethanol) showed maxima

at 318 (20,350), 303 (24,000) and 290 nm (19,100) along

with strong end absorption.



1,2-Diphenylcyclopropene-3-

carboxylic acid77

Freshly distilled ethyl diazoacetate (14.9 ml, 16 g,

0.14 moles) was added dropwise over a 5 hour period to a

mixture of diphenylacetylene (50 g, 0.28 moles),cyclohexane

(30 ml) and anhydrous copper sulfate (0.07 g) prepared in










a 500-ml three-necked round-bottomed flask equipped with a

mechanical stirrer, dropping funnel, reflux condenser and

nitrogen purge and heated with an oil bath maintained at

1250. When the addition was complete, the reaction mix-

ture was cooled to room temperature and 150 ml of a 10%

methanolic potassium hydroxide solution (11 g of potassium

hydroxide in 100 ml of methanol) was added dropwise. The

resulting dark red solution was heated at reflux for 4

hours, cooled to room temperature and diluted with 1 liter

of water. The unreacted diphenylacetylene was removed by

extraction with hexane. The aqueous solution was cooled

to 50 with an ice-water bath and acidified to a pH of less

than four by the dropwise addition of concentrated hydro-

chloric acid. The crude cyclopropene acid was collected

on a filter, washed with water and air dried. The crude

acid was placed on a column of florex and eluted with chloro-

form. The solvent was removed and the residue recrystallized

from acetone, yielding 1,2-diphenylcyclopropene-3-carboxylic

acid (9.8 g, 38% based on diphenylacetylene consumed) as
77
white prisms, mp 209-2110 (Lit. mp 209-211).



Diphenylcyclopropenium perchlorate47

A slurry of 1,2-diphenylcyclopropene-3-carboxylic

acid (6.0 g, 25 mmoles) in acetic anhydride (20 ml) was

added in one portion to an ice cold solution of perchloric

acid in acetic anhydride, prepared by adding 6.6 ml of










cold 70% perchloric acid to 40 ml of ice cold acetic an-

hydride, in a 250-mi three-necked round-bottomed flask

equipped with a mechanical stirrer and nitrogen purge.

The resulting dark brown suspension was stirred at 00 for

2.5 hours. Ether was then added and the precipitate

collected on a filter, washed extensively with ether and

air dried to give diphenylcyclopropenium perchlorate

(4.6 g, 61%) as a light tan solid, mp 148-150 (Lit.47

mp 148.5-150.50)



1,2-Diphenyl-3-(9'fluorenyl)-
A 48
cyclopropene48

A solution of fluorene (1.664 g, 10.00 mmoles) in

dry ether (15 ml) was prepared in a 100-ml three-necked

round-bottomed flask equipped with a reflux condenser,

syringe cap and nitrogen purge. n-Butyl lithium (7.4 ml,

10 mmoles) was added via syringe over a period of 30 min-

utes. The orange reaction mixture was then stirred at

reflux for 1.5 hours, cooled to room temperature and added

via syringe to a stirred suspension of diphenylcyclopro-

penium perchlorate (1.162 g, 4.000 mmoles) in ether (25 ml)

at -78' prepared in a 50 ml round-bottomed flask equipped

with a syringe cap, until the orange color remained. The

reaction mixture was stirred at -780 for 30 minutes, warmed

to room temperature, extracted with dilute hydrochloric

acid, brine, and water, dried over anhydrous sodium sulfate,

filtered and the solvent removed under reduced pressure.









The yellow residue was placed on a column of acidic alumina

(activity 1) and eluted with hexane. The collected

fractions (30 ml) were combined on the basis of their thin

layer behavior. The solvent was evaporated and the residue

recrystallized from benzene-petroleum ether. The product

was collected on a filter, washed with petroleum ether and

air dried to yield 1,2-diphenyl-3-(9'-fluorenyl)-cyclo-

propene (0.320 g, 22%) as white needles, mp 169-1700

(Lit.48mp 170-171.5)



Dimethyl-l-[2'-(9''-fluorenyl)-1',3'-

diphenylcyclopropenyl]-maleate (48)

A solution of 1,2-diphenyl-3-(9'-fluorenyl)-cyclo-

propene (0.064 g, 0.18 mmoles) and dimethyl acetylenedi-

carboxylate (0.026 g, 0.18 mmoles) in bromobenzene (3 ml)

was heated at 1000 for 18 hours. The reaction mixture was

cooled to room temperature and examined by thin layer

chromatography. In addition to a small amount of starting

material, one major component was observed and separation

by column chromatography was attempted. The solvent was

removed under a stream of nitrogen and the reaction residue

was placed on a column of silica gel and eluted with hexane

to remove the unreacted (47). The elutant was then changed

to 4:1 hexane-chloroform and 20 ml fractions were collected

and combined on the basis of their thin layer behavior.

The solvent was removed under reduced pressure, yielding (48)








as a waxy orange material (0.065 g, 72%). Attempts to

crystallize this product from hexane, hexane-benzene,

chloroform-hexane, and methanol were all unsuccessful.

The nmr spectrum (CDC13) showed an aromatic multi-

plet at L 7.8-6.9 (18H), a singlet at 5.80 (1H), a broad

singlet at 5.33 (1H), a singlet at 3.60 (3H), and a sing-

let at 3.55 (3H).

The infrared spectrum (film on AgC1) showed carbon-

hydrogen stretch at 2950 (m), cyclopropene carbon-carbon

double bond stretch at 1840 (w), carbon-oxygen double bond

stretch at 1760 (s) and 1720 (s), and carbon-carbon double

bond stretch at 1650 (m), along with other absorptions at

1490 (m), 1475 (m), 1450 (s), 1355 (m), 1035 (m), 978 (m),

887 (m), 762 (s), 745 (s), and 698 (s) cm-1.

The mass spectrum showed a molecular ion at m/e

498.1830 (67%), (calcd., 498.1838) along with other abund-

ant fragments at m/e 466 (22%), 439 (29%), 437 (63%), 407

(32%), 379 (45%), 376 (50%), 351 (25%), 302 (25%), 265

(20%), 189 (19%), 188 (22%), 180 (23%), 166 (31%), 165 (base

peak), and 163 (12%).

The ultraviolet spectrum (95% ethanol) showed maxima

at 300 (8,700), 287 (10,500) and 255 nm (23,400) along with

strong end absorption.










Attempted reaction of Triphenyl-

cyclopropene (33) with 1,4-benzoquinone

Triphenylcyclopropene (0.268 g, 1.00 mmoles) and

freshly sublimed 1,4-Denzoquinone (0.108 g, 1.00 mmoles)

were dissolved in chlorobenzene (2 ml). A 1-ml aliquot

of the resulting solution was sealed in an nmr tube and

heated at 1300 for 19 hours. The reaction mixture was

cooled to room temperature and examined by nmr spectro-

scopy. In addition to starting material, the only new

product was shown by comparison with an authentic sample

to be the dimer (34). None of the desired "ene" adduct

was detected and further investigation of this eneophile

was discontinued.



Attempted Reaction of Triphenyl-

cyclopropene (33) with 1,4-Naphthaquinone

Triphenylcyclopropene (0.178 g, 0.664 mmoles) and

freshly sublimed 1,4-naphthaquinone (0.105 g, 0.662 mmoles)

were dissolved in benzene (2 ml) and heated at reflux for

24 hours. The reaction mixture was cooled to room temper-

ature and examined by nmr spectroscopy. In addition to

starting material, the only new product was a small amount

of dimer (34), identified by comparison with an authentic

sample. None of the desired "ene" adduct was detected and

this reaction was not investigated further.










Attempted reaction of Tripheny-

cyclopropene (33) with Diethyl maleate

Triphenycyclopropene (0.108 g, 0.403 mmoles) and

diethyl maleate (0.138 g, 0.802 mmoles) were dissolved

in chlorobenzene (2 ml) and a 1 ml aliquot of the re-

sulting solution was sealed in an nmr tube and heated

at 1050 for 140 hours. The reaction mixture was cooled

to room temperature and examined by nmr spectroscopy.

The concentration of diethyl maleate was unchanged and

the only new signal corresponded to the dimer (34). No

"ene" adduct from the reaction of (33) with dimethyl maleate

was detected and this reaction was not investigated further.



Attempted Reaction of Triphenyl-

cyclopropene (33) with Diethyl fumarate

Triphenylcyclopropene (0.108 g, 0.403 mmoles) and

diethyl fumarate (0.138 g, 0.802 mmoles) were dissolved

in chlorobenzene (2 ml) and a 1 ml aliquot of the result-

ing solution was sealed in an nmr tube and heated at 1050

for 48 hours. The reaction mixture was cooled to room

temperature and examined by nmr spectroscopy. The con-

centration of diethyl fumarate was unchanged and the only

new signal corresponded to the diner (34), and this re-

action was not investigated further.










Attempted Reaction of Triphenyl-

cyclopropene with Bromomaleic anhydride

Triphenycyclopropene (0.068 g, 0.25 mmoles) and

bromomaleic anhydride (0.177 g, 1.00 mmoles), as sup-

plied by Aldrich Chemicals, were dissolved in chloro-

benzene (1 ml) and the resulting solution was sealed in

an nmr tube and heated at 130. After two hours, the

reaction solution was cooled to room temperature and

examined by nmr spectroscopy. The signal corresponding

to the C3 hydrogen of (33) was absent and a new signal

corresponding to 1,2-diphenyindene was observed. The

solvent was removed under a stream of nitrogen and the

reaction residue was placed on a column of silica gel

and eluted with hexane. After the solvent was removed,

the residue was taken up in petroleum ether and cooled.

The precipitate was collected on a filter and washed with

cold petroleum ether and air dried, yielding 1,2-diphenyl-

indene (0.059 g, 87%) as white needles, mp 177-178.



Attempted Reaction of Triphenyl-

cyclopropene (33) with Chloromaleic anhydride

Triphenylcyclopropene (0.030 g, 0.11 mmoles) and

chloromaleic anhydride (0.25 g, 0.53 mmoles), supplied as

a 50% mixture with maleic anhydride by Aldrich Chemicals,

were dissolved in chlorobenzene (1 ml) and the resulting

solution was sealed in an nmr tube and heated at 130 for

5 hours. The reaction mixture was cooled to room temperature









and examined by nmr spectroscopy. The signal correspond-

ing to the C3 hydrogen of (33) was completely consumed and

a new signal corresponding to 1,2-diphenylindene was ob-

served. The solvent was removed under a stream of nitrogen

and the residue dissolved in petroleum ether and cooled.

The precipitate was collected on a filter, washed with

cold petroleum ether and air dried, yielding 1,2,-diphenyl-

indene (0.023 g, 77%) as white needles, mp 176-178.



Attempted Reaction of Triphenyl-

cyclopropene (33) with 4-Phenyl-

1,2,4-Triazoline-3,5-dione (49)

A solution of 4-phenyl-l,2,4-triazoline-3,5-dione

(0.525 g, 3.00 mmoles) in methylene chloride (25 ml) was

added dropwise to a solution of triphenycyclopropene

(0.804 g, 3.00 mmoles) in methylene chloride (100 ml) pre-

pared in a 250 ml three-necked round-bottomed flask equipped

with a dropping funnel and nitrogen purge and cooled to

00 with an ice-water bath. The reaction mixture was stirred

for 1 hour, after which time the dark red color of (49)

was discharged and a pale yellow solution remained. The

solvent was removed under reduced pressure and the residue

dissolved in ether. Hexane was added and the precipitate

that resulted was collected on a filter, washed with cold

hexane and air dried to yield a white fluffy solid (1.157 g,

87%), mp 264.5-265.5.










The nmr spectrum (CDC13) showed an aromatic multiple

at _7.85-7.00 (20H), and a singlet at 6.43 (1H).

The infrared spectrum (KBr) showed nitrogen-hydrogen

stretch at 3450 (m), carbon-oxygen stretch at 1785 (s) and

1730 (vs) and other absorptions at 1600 (w), 1505 (m),

1405 (s), 1290 (m), 1140 (w), 1080 (w), 1025 (w), 763 (s),

750 (m), and 697 (s) cm-1.

The mass spectrum showed a parent ion at m/e 443 (20%),

along with other abundant fragments at m/e 366 (24%), 322

(82%), 267 (24%), and 119 (base peak).

The ultraviolet spectrum (95% ethanol) showed a

broad maxima at 305 nm.



Attempted Reaction of Triphenyl-

cyclopropene (33) with 4-Methyl-

1,2,4-triazoline-3,5-dione (50)

A solution of 4-methyl-l,2,4-triazoline-3,5-dione

(0.034 g, 0.30 mmoles) in methylene chloride (5 ml) was

added dropwise to a solution of triphenylcyclopropene

(0.081 g, 0.30 mmoles) in methylene chloride (5 ml) pre-

pared in a 25-ml three-necked round-bottomed flask equipped

with a dropping funnel and nitrogen purge and cooled to 00

with an ice-water bath. The reaction mixture was stirred

at 00 for 45 minutes, during which time the dark red color

of (50) was discharged and a pale yellow solution remained.

The solvent was removed under reduced pressure and the


1









residue dissolved in ether. IIexane was added and the re-

sulting precipitate was collected on a filter and washed

with hexane, yielding a white fluffy solid (0.067 g, 58%),

mp 238-243.

The nmr spectrum (CDC1 ) showed an aromatic multiple
3
at Z7.80-7.10 (15H), a singlet at 6.30 (Hi) and a singlet

at 3.0 (3H).

The infrared spectrum (KBr) showed nitrogen-hydrogen

stretch at 3400 (m), carbon-oxygen double bond stretch at

1735 (s) and other absorptions at 1600 (m), 1545 (m), 1497

(m), 1445 (m), 1235 (w), 1075 (w), 1030 (w), 975 (w), 770

(m), 760 (m), 733 (w), and 697 (s).

The mass spectrum showed a molecular ion at m/e 381

(base peak) along with other abundant fragments at m/e

351 (71%), 304 (94%), 295 (79%), 267 (32%) and 105 (24%).

The ultraviolet spectrum (95% ethanol) showed maxima

at 317 (8,660), 303 (10,500) and 287 nm (9890).



Phenylbenzyl ketone (54)49

Phosphorous trichloride (56 g, 0.40 moles) was added

to phenylacetic acid (54.4 g, 0.400 moles) in a 1 liter

three-necked round-bottomed flask equipped with a reflux

condenser and gas outlet tube leading to a potassium hydrox-

ide trap. The reaction mixture was heated at 90-100' until

gas evolution ceased, cooled to room temperature and ex-

amined by nmr spectroscopy. No phenyacetic acid was de-

tected and the phenylacetic acid chloride was used in the


i




University of Florida Home Page
© 2004 - 2010 University of Florida George A. Smathers Libraries.
All rights reserved.

Acceptable Use, Copyright, and Disclaimer Statement
Last updated October 10, 2010 - - mvs