Citation
Carbene-carbene rearrangements

Material Information

Title:
Carbene-carbene rearrangements a study in the benzocycloheptatrienylidene-naphthylcarbene system
Added title page title:
Benzocycloheptatrienylidene-naphthycarbene system
Creator:
Krajca, Kenneth Edward, 1944-
Publication Date:
Copyright Date:
1972
Language:
English
Physical Description:
xii, 92 leaves. : illus. ; 28 cm.

Subjects

Subjects / Keywords:
Absorption spectra ( jstor )
Carbenes ( jstor )
Cyclohexenes ( jstor )
Mass spectroscopy ( jstor )
Photolysis ( jstor )
Protons ( jstor )
Pyrolysis ( jstor )
Reaction mechanisms ( jstor )
Sodium ( jstor )
Spectral index ( jstor )
Carbenes (Methylene compounds) ( lcsh )
Chemistry thesis Ph. D
Dissertations, Academic -- Chemistry -- UF
Genre:
bibliography ( marcgt )
non-fiction ( marcgt )

Notes

Thesis:
Thesis -- University of Florida.
Bibliography:
Bibliography: leaves 87-91.
Additional Physical Form:
Also available on World Wide Web
General Note:
Typescript.
General Note:
Vita.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
022663170 ( AlephBibNum )
13958996 ( OCLC )
ADA5054 ( NOTIS )

Downloads

This item has the following downloads:


Full Text

















CARBENE-CARBENE REARRANGEMENTS: A STUDY IN THE
BENZOCYCLOHEPTATRIENYLIDENE-NAPHTHYLCARBENE SYSTEM








By





KENNETH EDWARD KRAJCA


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


UNIVERSITY OF FLORIDA
1972










DEDICATION

This dissertation is dedicated to the author's wife,

Sherry Lee, whose work, understanding, and love made this

work possible and to his daughter, Laura, who provided him

with much happiness.











ACKNOWLEDGMENTS

The author is indebted to Dr. W. M. Jones for his

professional guidance, patience, and personal friendship

during the course of this work. His advice and encourage-

ment will be of lasting benefit. The counsel and discussion

of the author's fellow associates are highly regarded.

The author wishes to thank his parents for their

guidance and help.

Finally, the author wishes to thank the Petroleum

Research Fund and the National Science Foundation for

providing financial assistance during this period of study.











TABLE OF CONTENTS


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

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

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

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

CHAPTER

I. INTRODUCTION AND BACKGROUND....................

II. RESULTS ........................................

III. DISCUSSION......................................

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

General. .....................................

Pyrolysis of the Sodium Salt of Benzaldehyde

Tosylhydrazone in Diglyme.............

Pyrolysis of the Sodium Salt of Tropone

Tosylhhyrazone in Diglyme ............

4,5-Benzotropone ...........................

4,5-Benzotropone Tosylhydrazone ...........

B-Naphthaldehyde Tosylhydrazone...........

General Preparation of the Sodium Salt of

Tosylhydrazones ......................

Pyrolysis of the Sodium Salt of 4,5-Benzo-

tropone Tosylhydrazone in Diglyme.....









Photolysis of the Sodium Salt of 4,5-Benzo-

tropone Tosylhydrazone in Diglyme..... 62

Pyrolysis of the Sodium Salt of 4,5-Benzo-

tropone Tosylhydrazone in Benzene..... 63

Photolysis of the Sodium Salt of 4,5-Benzo-

tropone Tosylhydrazone in Benzene..... 64

Preparation of 2-naphthaldazine ............ 64

Pyrolysis of the Sodium Salt of 8-Naphth-

aldehyde Tosylhydrazone in Cyclohexene. 64

Pyrolysis of the Sodium Salt of 4,5-Benzo-

tropone Tosylhydrazone in Cyclohexene

at 83 ..................................... 66

Pyrolysis of the Sodium Salt of 4,5-Benzo-

tropone Tosylhydrazone in Cyclohexene

at 140 ................................. 67

Pyrolysis of the Sodium Salt of 4,5-Benzo-

tropone Tosylhydrazone in Cyclohexene--

Diglyme at 820........................ 67

Photolysis of 4,5-Benzotropone Tosylhydrazone

Sodium Salt in Cyclohexene............ 68

Pyrolysis of the Sodium Salt of 4,5-Benzo-

tropone Tosylhydrazone in Cyclohexene. 69

Reaction of the Sodium Salt of 4,5-Benzo-

tropone Tosylhydrazone with Dimethyl

Fumarate at 1050...................... 69






Page..

Reaction of the Sodium Salt of B-Naphth-

aldehyde Tosylhydrazone with Dimethyl

Fumarate at 1050..................... 70

Reaction of S-Naphthaldehyde Tosylhydrazone

Sodium Salt with Dimethyl Fumarate

at 160 ............................... 71

Reaction of the Sodium Salt of 4,5-Benzo-

tropone Tosylhydrazone with Dimethyl

Fumarate at 1600..................... 72

General Procedure for the Photolysis of

Tosylhydrazone Sodium Salts in the

2-Butenes. ............................ 72

Photolysis of the Sodium Salt of 6-Naphth-

aldehyde Tosylhydrazone in trans-2-

Butene ..................... .......... 73

Photolysis of Benzaldehyde Tosylhydrazone

Sodium Salt in trans-2-Butene........ 74

Photolysis of the Sodium Salt of B-Naphth-

aldehyde Tosylhydrazone in cis-2-

Butene................................ 75

Pyrolysis of B-Naphthaldehyde Tosylhydrazone

Sodium Salt in cis-2-Butene.......... 76

Photolysis of 4,5-Benzotropone Tosylhydrazone

Sodium Salt in cis-2-Butene.......... 77

Photolysis of 4,5-Benzotropone Tosylhydrazone

Sodium Salt in trans-2-Butene......... 77







Pyrolysis of the Sodium Salt of 4,5-Benzo-

tropone Tosylhydrazone in cis-2-Butene. 77

Pyrolysis of the Sodium Salt of 4,5-Benzo-

tropone Tosylhydrazone in cis-2-

Butene and Benzene..................... 78

Pyrolysis of the Sodium Salt of 4,5-Benzo-

tropone Tosylhydrazone in cis-2-Butene

and Perfluorocyclobutane.............. 79

Pyrolysis of the Sodium Salt of 4,5-Benzo-

tropone Tosylhydrazone in cis-2-Butene

and Oxygen ..................... ...... SO

Pyrolysis of the Sodium Salt of 4,5-Benzo-

tropone Tosylhydrazone in cis-2-Butene

and trans,trans-2,4-hexadiene.......... 80

Sensitized Photolysis of the Sodium Salt of

4,5-Benzotropone Tosylhydrazone in

Benzene................................ 81

Thermal Stability of the syn- and anti-1-

(2-naphthyl)-cis-2,3-dimethylcyclo-

propanes ................ .............. 81

Stability of cis-2-Butene under Reaction

Conditions ......................... 82

2-Methyl-4,5-Benzotropone .................. 82

2-Methyl-4,5-Benzotropone Tosylhydrazone... 82

Pyrolysis of the Sodium Salt of 2-Methyl-

4,5-Benzotropone Tosylhydrazone in

Benzene................................ 83






Page

2,3-Benzotropone ............................ 84

2,3-Bentotropone Tosylhydrazone............. 84

Pyrolysis of the Sodium Salt of 2,3-Benzo-

tropone Tosylhydrazone in Benzene....... 85

LIST OF REFERENCES .................................... 87

BIOGRAPHICAL SKETCH ................................... 92











LIST OF TABLES

Table Page

1. Stereochemical Results: The Percent "Wrong
Isomer" from the Decompositions in cis-2-
Butene of the Carbene Precursors both Thermally
at 1050 and Photolytically at -25 ............. 27











LIST OF FIGURES

Lgure Page

1. Vpc trace: Pyrolysis of 26 in Cyclohexene.... 16

2. 100 MHz Nmr of 33... ............ ............. 18

3. Decoupled Nmr of 33. ........................... 19

4. Closure of vinylcarbene to cyclopropene....... 48

5. Boat Conformation of Benzocycloheptatrienyli-
dene.............. .......................... 50

6. Planar Conformation of Benzocycloheptatrienyli-
dene. ....................................... 51











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


CARBENE-CARBENE REARRANGEMENTS: A STUDY IN THE
BENZOCYCLOHEPTATRIENYLIDENE-NAPHTHYLCARBENE SYSTEM


By


Kenneth Edward Krajca

August, 1972


Chairman: W. M. Jones
Major Department: Chemistry


The rearrangement of benzocycloheptatrienylidenes to

naphthylcarbenes in solution thermally at temperatures in the

range of 1000 and photolytically at temperatures as low as

-350 is reported. The rearranged naphthylcarbenes were

detected by their reaction products with solvent. Decomposi-

tion in benzene gave the corresponding naphthylcyclohepta-

trienes in good yield, and decompositions in cyclohexane and

diglyme yielded the corresponding insertion products. In

cyclohexene, the formation of naphthylnorcaranes and naphthyl-

carbene insertion products into cyclohexene was temperature

and concentration dependent. This is consistent with a

bimolecular reaction of an intermediate with cyclohexene in

competition with a unimolecular rearrangement that gives

0-naphthylcarbene.






The rearrangement of a carbene precursor was excluded

and possible mechanisms were explored. Stereochemical studies

indicate that the singlet state of the carbene is responsible

for the rearrangement and that there is a small amount of

nonstereospecific addition which is unaffected by dilution

but is affected by the addition of triplet scavengers. This

is compatible with a singlet-triplet equilibrium wherein the

rate of triplet-singlet intersystem crossing is greater than

the rates of addition of the triplet carbene to olefin, but

the rate of triplet carbene addition to dienes (and oxygen)

must be greater than the rate of triplet-singlet intersystem

crossing. Methyl substitution has an effect upon the rearrange-

ment which may occur via a fused, twisted cyclopropene inter-

mediate (benzobicycloheptatriene) from the boat conformation

of the carbene, the planar conformation, or via a concerted

pathway.










CHAPTER I

INTRODUCTION AND BACKGROUND


Carbene-carbene rearrangement has been established as a

firm, although relatively rare, chapter in the chemistry of

divalent intermediates. In 1967, Skattebol' reported several

examples of "non-trivial" carbene-carbene rearrangements, e.g.,

that of vinylcyclopropylidene (1) to cyclopentenylidene (2).



B F> ->
Br2
1 2

In this type of rearrangement, however, the carbene carbon

retains its integrity and, since the carbene was generated by

a-elimination, a free carbene may not have been obtained. In

fact, the stereoselectivity in product formation in one case

suggested a carbenoid species.

Cais et al.2 reported a carbene-carbene rearrangement in

the deprotonation of ferrocenyltropylium fluoborate (3) with

the admonition that the carbene intermediacy was not established.





F BF4 0
3 F







Along a slightly different line, Strausz3 and coworkers

investigated the rearrangement of the carbonylcarbene derived

from the photolysis of 3-diazobutanone-2 and homologous com-

pounds in both the gas and condensed phases. Russell and

Rowland" also investigated similar systems but came to slightly

different conclusions.

In 1965, Shechter and Vander Stouw5 reported the rearrange-

ment of various methyl- and dimethylphenylcarbenes and eluci-

dated this work in a more recent report.6 In their work,

(2-methylphenyl)diazomethane (4), for example, was pyrolyzed

to yield, among other products, styrene (5), and benzocyclo-



CHN2

CH3 0,

4 5 6
butene (6). The above products and a labeling study indicated

the process was occurring via a multiple carbene-carbene

rearrangement.

Jones et al.7 later elaborated on the phenylcarbene-cyclo-

heptatrienylidene rearrangement when heptafulvalene (7) was

isolated from the gas phase pyrolysis at 2500 of the sodium

salt of benzaldehyde tosylhydrazone as well as from pyrolysis

of the sodium salt of tropone tosylhydrazone.8 Pyrolysis of



H- 'C'



7





a mixture of the sodium salts of benzaldehyde tosylhydrazone
and p-tolualdehyde tosylhydrazone resulted in the formation of
crossed dimer which proved that the rearrangement occurred in
the gas phase.
Multiple carbene-carbene rearrangements of diphenyl-
methylene (8) were also studied by Jones.9 Pyrolysis at 3500
of diphenyldiazomethane yielded fluorene (9), tetraphenyl-
ethylene (10), triphenylheptafulvene (11), diphenylheptafulva-
lene (12), and the corresponding azine. Note that these


C:


a


4-


0 0

00


13



1l


S-


2L



01
O iN


results require that the diarylcarbene expand to form the
arylcycloheptatrienylidene 13 and the arylcycloheptatrienyli-
dene contract to form a biphenylmethylene which is the
immediate precursor of the fluorene.







Wentrup and Wilczek0l also reported the reversibility of

the phenylcarbene-cycloheptatrienylidene rearrangement when

they isolated stilbene and heptafulvalene from the indepen-

dent gas phase pyrolysis at 3000 of the sodium salts of both

benzaldehyde tosylhydrazone and tropone tosylhydrazone.

Hedaya and coworkers" studied the ring contraction and

expansion of phenylcarbene in the gas phase by pyrolysis of

benzaldehyde tosylhydrazone sodium salt. They also observed

the formation of heptafulvalene as one of the reaction products

and by incorporating a labeling study on tolylcarbene'2 they

decided that very small energy barriers separate phenylcarbene,

cycloheptatrienylidene, and possible bicycloheptatriene inter-

mediates.

Lastly, M. Jones, Jr.13 investigated the interconversion

of o-, m-, and p-tolylcarbene by observing the ratios of the

styrene (5) and benzocyclobutene (6) produced in the pyrolysis

of the corresponding diazo compounds.
CHN2

CH3






CHN2
CH3

CHN2 6




CH3







It is very interesting to note that most of the

researchers on the phenylcarbene-cycloheptatrienylidene

interconversion believe that the rearrangement occurs via a
5-7 9-13
bicycloheptatriene intermediate (14). Jones and









14


Mitsuhashi" elucidated this mechanism with their study on

the importance of bond order in the gas phase carbene-carbene

rearrangement of a series of substituted naphthylcarbenes.

It was found that scrambling of the carbene occurs only

between positions having the highest bond order which augurs

well for the bicyclic intermediate as opposed to a Wolff-type

rearrangement (shown below).





C '*-





To date, the latest work in carbocyclic carbene-carbene

rearrangements was done by Jones and Gebertis on the ring

expansion of the ten w-electron 1,6-methano[10]annulenyl-2-

carbene (15). The products of the reaction are apparently







rearranged dimers and the mechanism is assumed to be similar

to that mentioned above.





^C:

15

Nitrogen analogues to the phenylcarbene-cyclohepta-

trienylidene rearrangement have been known for a long time.

The ring expansion of phenylnitrene to yield substituted

azepines was studied as early as 195816,17 and has recently

received thorough investigation.18-24 Pyrolysis16 and

photolysis24 of phenyl azide (16) in the presence of amines

led to 2-amino-3H-azepine derivatives (17) as final products.


EtNEt2




16 17


The thermal1 and photolytic21,22 deoxygenation of nitro-

benzene, as well as the thermal deoxygenation of nitroso-

benzene,18 in the presence of amines also led to 2-amino-3H-

azepines. As in the analogous carbene-carbene rearrangements,

these reactions can be envisioned as proceeding through a

bicyclic intermediate which can yield an azacycloheptatrienyli-

dene. From the experimental results, however, workers have

not been able to determine whether or not the azacyclohepta-






trienylidene is an intermediate. In fact, another mechanism,
addition of the amine to the bicyclic azirine (18), has
received much more attention.25


18
6 -0 -


Et2NH

NH
-NEt 2
0- >


S- Et NEt2






NEt2


Crow and Wentrup2 reported that the thermal generation
of 2-pyridylphenylcarbene (18a)led to carbazole (19) products
which were believed to arise via a substituted azacyclohepta-


S 0


I
19 H







trienylidene (20) intermediate or transition state. Crow and

Wentrup have extended their work on these systems27 as well

as investigated the pyridylnitrene-diazacycloheptatrienylidene

rearrangement.2e Perhaps the most interesting was the work

on 2-quinolylnitrene (21), an annelated nitrene which is








=N 2.1 2







N-"
N N:





believed to rearrange via an annelated diazacycloheptatrienyli-

dene (22).29 However, naphthyl azides have not been known to

give rise to detectable azepine.3"

The work described in this dissertation was done to

investigate the effects of annelation on the phenylcarbene-

cycloheptatrienylidene rearrangement. It was hoped that the

effects of annelation would give insight on the nucleophilicity

of the aromatic carbene, on the relative rate of rearrangement

and on the mechanism of the rearrangement. The approach was

very feasible in that the precursors for both the annelated

arylcarbene and the annelated aromatic carbenes were avail-





9


able. The systems chosen for study were 4,5-benzocyclohepta-

trienylidene (23), 2-methyl-4,5-benzocycloheptatrienylidene

(24), 2,3-benzocycloheptatrienylidene (25), and the related


S** CH3 **c a






23 24 25
naphthylcarbenes. While other benzocycloheptatrienylidenes

have been prepared and studied, they did not rearrange and

many of their reactions were typical of diarylcarbenes."3












CHAPTER II

RESULTS


The first results needed were for the model system of

phenylcarbene and cycloheptatrienylidene. Thus, the pyrolysis

of the sodium salt of benzaldehyde tosylhydrazone was effected

in diglyme at 1650 to produce what is believed to be the

phenylcarbene-diglyme insertion products. This residue, a

light yellow oil, had an nmr spectrum very similar to that of

the naphthylcarbene-diglyme insertion products videe infra);

there was also benzaldazine produced in the reaction. The

nmr spectrum of the reaction material indicated no heptaful-

valene was present, and comparative tic of the material and

a known sample of heptafulvalene was carried out using silica

gel eluting with pentane. This confirmed the absence of

heptafulvalene in the reaction material.

Tropone tosylhydrazone sodium salt was also decomposed

in diglyme at a temperature of 2100. The nmr spectrum of the

residue, a dark solid, was identical to that of heptafulvalene

with an additional small amount of aromatic resonance at 2.8 T

but none at 6.2-7.0 T (the position expected for the resonances

of any diglyme insertion products).

The aromatic carbene, 4,5-benzocycloheptatrienylidene

(23), was generated by the thermal and photolytic decomposi-







tion of the sodium salt of 4,5-benzotropone tosylhydrazone

(26) in an appropriate solvent. Pyrolysis or photolysis of

the sodium salts of tosylhydrazones form the corresponding

diazo compounds in situ.32 The diazo compounds then decompose

under aprotic reaction conditions to yield the carbenes.

The synthetic route to the sodium salt of 4,5-benzo-

tropone tosylhydrazone is shown in the scheme below. The


SCO2Et

C=O
c+ o 0

*COzEt IZ





N-N-Ts --- N-Ts
Na+ H

26


ketone, 4,5-benzotropone (27), was prepared by the method of

Thiele and Weitz33 and the tosylhydrazone by a modification

of the procedure of Closs.34 The conversion of the tosyl-

hydrazone to its sodium salt was done with sodium hydride in

tetrahydrofuran.35 The last step was carried out in a dry

box; however, contrary to some reports the sodium salts did

not appear to be hygroscopic in most cases.

In a similar manner, commercially available B-naphth-

aldehyde was converted to its tosylhydrazone36 and then to its







tosylhydrazone sodium salt. This salt was the immediate

precursor of B-naphthylcarbene.

In most cases, the naphthaldehyde tosylhydrazone sodium

salt was decomposed under the same or very similar conditions

as those of the 4,5-benzotropone tosylhydrazone sodium salt

so that the chemistry of the two carbenes could be compared

and, where possible, as a check on the authenticity of the

naphthyl products formed from the rearranged carbene.

Thus 4,5-benzotropone tosylhydrazone sodium salt was

decomposed in diglyme at 1000 to give a 29% yield of the

isomeric B-naphthylcarbene-diglyme insertion products (28).



HaCN- 0 O/ CH





280

While these isomeric compounds could not be readily separated,

there is little doubt as to their identity. The nmr spectrum

of these products shows aromatic proton absorptions at 2.00-

2.80 T and methyl and methylene protons adjacent to oxygen at

6.10-7.20 T (there are more signals than necessary in the

region of 6.60 T for methyl protons of a single product).

Other spectral properties and the elemental analysis of the

mixture are in complete agreement with the assigned structures,

and regardless of the carbene precursor videe infra), the

products formed in diglyme are the same.







The decomposition of naphthaldehyde tosylhydrazone

sodium salt in diglyme at 1000 produced a 13% yield of the

same B-naphthylcarbene-diglyme insertion products produced

in the above decomposition. The product yield from the

decomposition of the naphthaldehyde tosylhydrazone salt was

typically lower because there was competition of aldazine

formation.

Next, the photolytic decomposition of the salt of 4,5-

benzotropone tosylhydrazone in diglyme was carried out at 300

and resulted in the formation of the familiar naphthylcarbene-

diglyme insertion products.

When 4,5-benzotropone tosylhydrazone sodium salt was

heated in benzene at 800 there was formed, in 96% yield, a

product identified as 2-(2,4,6-cycloheptatrien-l-yl)naphthalene

(29). Although this compound has been prepared before, there









29

is no complete report of its physical and spectral properties.37

The existence of the cycloheptatrienyl moiety was proven by

comparison of the above compound's nmr spectrum with that of

l-phenyl-2,4,6-cycloheptatriene38 (the spectra were identical

except for the aromatic region). That it was the B-naphthyl

function (as opposed to a-naphthyl) was shown by careful







examination of the compound's ir spectrum; the absorption for

the out-of-plane bending at 861 cm-1 indicated an isolated H

on the naphthyl ring system, that at 821 cmI indicated two

adjacent H, and that at 745 cm1 was indicative of four adja-

cent H39 (the nmr of the product showed a ratio of 4:3 for

the a to 8 protons, respectively, in the aromatic region).

Photolysis at 300 of the sodium salt of 4,5-benzotropone

tosylhydrazone also produced the adduct 29 in 59% yield.

Photolysis (300) and thermolysis (1250) of the sodium

salt of S-naphthaldehyde tosylhydrazone in benzene gave, in

addition to the aldazine, 29 in 67% and 78%, respectively.

The existence of 8-naphthaldazine in the reactions

involving the B-naphthaldehyde tosylhydrazone sodium salt was

proven by comparison of the reaction product with an authentic

sample of the azine"0 which was independently synthesized."1

While decompositions of both carbene precursors in

benzene gave the same hydrocarbon products, the decompositions

of 4,5-benzotropone tosylhydrazone sodium salt in cyclohexene

were unique in that there were additional products formed

when compared to the 8-naphthaldehyde tosylhydrazone sodium

salt decompositions.

First, the pyrolysis of 8-naphthaldehyde tosylhydrazone

sodium salt was carried out in cyclohexene at 135-1450. The

relative yields of the three volatile components are as

follows: the naphthylcarbene-cyclohexene insertion product

30 (34%), the syn-norcarane 31 (26%), and the anti-norcarane







32 (40%). The aldazine was formed in 10% yield, but it is

not in the above data as they only include volatile hydro-

carbon products.













30 32 31

The insertion product 30 was readily identified by nmr;

the two vinylic and two "benzylic" protons were readily dis-

cernible. The syn- and anti-norcaranes required slightly more

attention to effect differentiation, however. The corresponding

phenylnorcarane models"' and the guidelines put forth for their

identification by Closs and coworkers3'," aided the charac-

terization of the naphthyl moieties. Thus, by analogy, the

syn-7-(2-naphthyl)norcarane 31 should have the naphthyl ring

system facing the cyclopropane ring and this conformation

would then have two separate effects on the nmr spectrum of

the compound: first, several of the cyclohexyl ring protons

would be held in the shielding cone of the naphthalene ring

system and be consequently shifted upfield as far as 0.7 ppm

compared with the anti-epimer; second, the aromatic protons

would not interact with the cyclopropyl ring and their signals

would be relatively narrower than the other compound. The

anti-epimer 32 should exist in the conformation with the







naphthyl ring system bisecting the cyclopropyl ring"4 which

would cause one of the o-protons to lie in the shielding zone

of the cyclopropane ring and to be shifted upfield approxi-

mately 0.2 ppm relative to the syn-product (of course the

cyclohexyl ring protons would be slightly downfield and not

so broadened when compared with the former).

Second, when the sodium salt of 4,5-benzotropone tosyl-

hydrazone was decomposed in cyclohexene heated to reflux (830),

five products were formed. These products were detected and

isolated by vpc. The vpc trace of the decomposition of 4,5-

benzotropone tosylhydrazone sodium salt in cyclohexene at

830 is shown in Figure 1. Two products accounted for 66% of


32
3031
30


Figure 1. Vpc trace: Pyrolysis of 26 in Cyclohexene.







the material: one is unidentified, but the other has the

properties of the spiro-adduct of benzocycloheptatrienylidene

to cyclohexene (33). The 100 MHz nmr of 33 is shown in








33

Figure 2 and the decoupled spectrum in Figure 3. Perhaps the

cyclohexyl moiety is introducing the dissymmetry present in

the spectra, but it is difficult to rationalize the large

differential chemical shift of the two different cyclohepta-

trienyl protons adjacent to the benzene ring system. The uv

spectrum, Xmax = 240 mp (log e = 4.2) and 265 mi (log e = 3.5),

compares well with that of 3,4-benzocycloheptatriene, Xmax

230 mi (log E = 4.7) and 265 my (log e = 3.7).~' The mass

spectrum of this compound indicates a molecular ion at m/e

222. The ir contributed little toward the identification
-1
except that the strong absorption at 750 cm is indicative

of an o-disubstituted benzene.'* Also, other absorption are

present in the ir spectra of both 3,4-benzocyclohepta-1,3,6-

triene'5 and the above unknown compound with approximately
-1
the same intensities (1620, 1480, 1440, 795, 745 cm and

several weak bands). The last three products and their rela-

tive yields were, respectively, (cyclophexene-3-yl)-2-naphthyl-

methane, the naphthylcarbene-cyclohexene insertion product,

(7%), syn-7-(2-naphthyl)norcarane (12%), and anti-7-(2-naphthyl)-
---- I fln /1c \




























h
























4-,








4-
0
























44,
o












oo











A
*i'


pri
~a~p-




r


3

m


--Y-


7-

zl
0


C)






a)

F-.
0
CDl
Pt


3.--- ------ ---






Pyrolysis of the sodium salt of 4,5-benzotropone tosyl-
hydrazone in cyclohexene at 1400 led to the same five products
as the lower temperature one.
The sodium salt of 4,5-benzotropone tosylhydrazone was
thermally decomposed at 820 in a 2:1 molar mixture of diglyme--
cyclohexene. Analysis of the hydrocarbon products indicated
the presence of the same five products as the other cyclohexene
decompositions with the addition of the naphthylcarbene-diglyme
insertion products. The diglyme products amounted to 27% of
the total products.
Finally, the photolysis of 4,5-benzotropone tosylhydrazone
sodium salt in cyclohexene at 300 produced the same products
as the pyrolyses plus a new unidentified one. In the
photolysis, the benzocycloheptatrienylidene addition product
33 is a major component; in the pyrolyses, the opposite is
true. In the cyclohexene--diglyme reaction, interfering
materials made accurate analysis impossible.
When cyclohexane was used for the substrate, the product
isolated in 51% yield from the decomposition of 4,5-benzo-
tropone tosylhydrazone sodium salt was (B-naphthyl)cyclohexyl-
methane (34). While the yield was not exceptional, the nmr
spectrum of the crude reaction material indicated almost
exclusively the insertion product and cyclohexane.



cc m
Z111 I'l







In an effort to trap the diazo compound of 4,5-benzo-

tropone, the tosylhydrazone sodium salt was decomposed in benzene

at 105-1080 in the presence of an equimolar amount of dimethyl

fumarate. There was considerable nitrogen evolution, and the

product isolated from the reaction was the naphthylcarbene-

benzene adduct 29 in 63% yield. The nmr spectrum of the

crude reaction material indicated only the adduct and dimethyl

fumarate. Comparative tlc, on silica gel eluted with 50:50

ethyl ether--pentane, of the known pyrazoline videe infra) and

this reaction mixture indicated no pyrazoline was present in

the latter.

Naphthyldiazomethane was treated in a similar manner.

The sodium salt of B-naphthaldehyde tosylhydrazone was decom-

posed in benzene at 105-1080 in the presence of an equimolar

amount of dimethyl fumarate. The product, an almost colorless,

very viscous oil, had all of the properties of trans-3,4-

dicarbomethoxy-5-(2-naphthyl)-2-pyrazoline (35). The oil was



MeO 2 C CO2Me



H

35

not stable to ordinary distillation conditions and even

though column chromatography on silica gel effected some

purification, it was not sufficient to provide a sample which

gave a good elemental analysis (e.g., calcd. N, 8.97; found:

N, 8.31). The ir spectrum of this compound had a strong







absorption at 3440 cm-1 indicative of N-H stretching, one at

1730 for the unconjugated ester carbonyl stretching, one at

1700 for the conjugated ester carbonyl, and one at 1555 indi-

cating a C=N bond. The nmr spectrum contained absorptions at

2.10-2.60 T indicating the naphthyl protons, a broad singlet

at 3.24 T for the proton bonded to the nitrogen, a broad

doublet (J = 9.4 Hz) for the "benzylic" proton at 4.60 -, a

double (J = 9.4 Hz) for the proton adjacent to the carbo-

methoxy group at 5.92 T, a singlet for the conjugated ester

methyl at 6.18 T, and a singlet for the unconjugated ester

methyl at 6.24 T (in addition, there were other resonances

which changed in intensity depending upon work up and purifi-

cation steps and were therefore not believed to originate

from the pyrazoline). The trans-configuration was evident

when compared with nmr data of trans-3,4-dicarbomethoxy-5-

phenyl-2-pyrazoline47 which had J = 10 Hz for the protons on

C-4 and C-5 and a chemical shift of 4.87 T for the benzylic

proton on C-5. It is also interesting to note that the above

phenyl pyrazoline was a viscous oil which resisted all attempts

to induce crystallization.' The mass spectrum had no signal

for the molecular ion, but there was one at 284 (the molecular

ion with the loss of N2). The max in ethanol were 224 my

(log E = 4.53), 275 (3.75), and 287 (3.79). Using 2-methyl-

naphthalene9 and the known 3-carbomethoxy-4-phenyl-2-pyrazo-

line50 as models, the following absorptions would be expected

in the uv: 223 my (log e = 4.22), 274.5 (3.73), and 296 (3.98).







The decomposition of B-naphthaldehyde tosylhydrazone

sodium salt in benzene at 1600 in the presence of an equimolar

amount of dimethyl fumarate gave rise to a different product

than the low temperature one. The product was identified as

trans-l,2-dicarbomethoxy-3-(2-naphthyl)cyclopropane (36) and

was isolated in 67% yield. The structure was easily desig-


O2Me



Me02C
H
36

nated trans because of the two different methyl proton absorp-

tions (either cis-cyclopropane would have had only one signal

for the two methyl groups as they would be magnetically

equivalent) and because the thermal decomposition of trans-

3,4-dicarbomethoxy-5-phenyl-2-pyrazolines is known to give

the trans-cyclopropanes .

The decomposition of the sodium salt of 4,5-benzotropone

tosylhydrazone in benzene at 1600 in the presence of dimethyl

fumarate led to the formation of 2-(2,4,6-cycloheptatrien-l-yl)-

naphthalene (29) and what is believed to be 2-(2,4,6-cyclo-

heptatrien-4-yl)naphthalene (37) in a ratio of 1.0:1.6 as

determined by nmr. The latter product could arise via a [1,51

hydrogen shift of the former; the hydrogen shift product was

not isolated but its presence was inferred from the nmr

spectrum and from work done in the phenylcycloheptatriene

series.sl

















29 37

8-naphthaldehyde tosylhydrazone sodium salt was decomposed

by photolysis in the presence of trans-2-butene to yield

1-(2-naphthyl)-trans-2,3-dimethylcyclopropane (38) along with

a small amount of cis- and unknown materials. While other


H


H3 CH



38 H

properties were consistent with this structure, the identifi-

cation of this compound was hampered by the unusual nmr

spectrum. The two signals arising from the methyl group

protons were essentially singlets (they would be expected to

be doublets). Two things were tried in order to solve this

dilemma: first, the nmr spectrum was taken in a different

solvent, benzene, which caused one of the singlets to split

by about 1.5 Hz, and,second, the corresponding phenyl cyclo-

propane was prepared and its nmr spectrum examined and found

to be very similar to the naphthylcyclopropane. Thus, the

structure is in accord with the observed spectral properties.







The photolysis of the sodium salt of B-naphthaldehyde

tosylhydrazone in cis-2-butene gave the two expected epimeric

cyclopropanes along with a small amount of trans-cyclopropane

and unknown materials. Syn-l-(2-naphthyl)-cis-2,3-dimethyl-



H3

H3 H
HH CH3

H H CH3
H

39 40

cyclopropane (39) and anti-l-(2-naphthyl)-cis-2,3-dimethyl-

cyclopropane (40) were differentiated in a manner similar to

that of the epimeric naphthylnorcaranes videe supra). In the

syn-cyclopropane 39 the methyl protons appear 0.2 ppm upfield

relative to the anti-compound, and the naphthyl resonances are

relatively less broad. Naturally, the methyl resonances were

at lower field and one of the o-protons on the naphthyl system

appeared at higher field (causing a broadening of the aromatic

resonances) in the anti relative to the syn.

Pyrolysis of the naphthaldehyde tosylhydrazone sodium

salt in cis-2-butene gave, in addition to the epimeric cyclo-

propanes, an unknown product formed in considerable amounts.

When 4,5-benzotropone tosylhydrazone sodium salt was

decomposed by photolysis in the presence of cis-2-butene, it

gave the expected epimeric cyclopropanes, a small amount of

trans-cyclopropane, and a considerable amount of unknown







material. The photolysis in trans-2-butene gave as a major

product the expected trans-cyclopropane (in addition to small

amounts of cis-cyclopropane and unknown material).

The pyrolysis of 4,5-benzotropone tosylhydrazone in cis-

2-butene yielded large amounts of the epimeric cyclopropanes,

small amounts of trans-cyclopropane, and some extraneous

materials.

The stereochemistry of the addition of the carbenes to

olefins was then determined by analysis of the amount of

trans-cyclopropane formed in the decomposition of the two

carbene precursors in research grade cis-2-butene. The

decompositions were carried out under a variety of conditions.

They were done using the neat olefin as solvent, they were

done with the olefin diluted with benzene (1:3 molar ratio,

respectively) and perfluorocyclobutane (two runs, 1:2 molar

ratio and 1:6 molar ratio, respectively), and they were done

with olefin in the presence of oxygen and trans,trans-2,4-

hexadiene (two runs, 1.0:1.5 molar ratio and 1.0:2.1 molar

ratio, respectively). Dilution had no effect upon the ratio

of cis- and trans-cyclopropanes formed; however, the presence

of oxygen and the diene significantly altered the above ratio,

producing relatively more cis-cyclopropane. When the mole

percent of trans,trans-2,4-hexadiene was greater than 66%,

there was too little cis-2-butene adduct formed to determine

the stereochemistry. The results of the stereochemical

investigation are shown in Table 1.











TABLE 1

Stereochemical Results: The Percent "Wrong Isomer"a from the
Decompositions in cis-2-Buteneb of the Carbene Precursors
both Thermally at 1050 and Photolytically at -250.

Tosylhydrazone Sodium Salt % "Wrong Isomer"c
Solvent System (molar ratio) A hv

Naphthaldehyde d
cis-2-butene 10 3

Benzotropone
cis-2-butene 7 4

Benzotropone
cis-2-butene/benzene (1:3) 6

Benzotropone
cis-2-butene/perfluorocyclobutane (1:2) 6

Benzotropone
cis-2-butene/perfluorocyclobutane (1:6) 6

Benzotropone
cis-2-butene/oxygen (saturated) 4

Benzotropone
cis-2-butene/2,4-hexadiene (1.0:1.5) 1


a 1-(2-naphthyl)-trans-2,3-dimethylcyclopropane
b research grade 99.94 mole %

c standard deviations were typically 1

ddatum only approximate due to interfering material







To be sure that the products were stable to vpc, the cis-

cyclopropane was preparatively chromatographed and then

analyzed for trans-material--none was found. Also, the olefin

was analyzed after the reaction and it was found that the olefin

had isomerized less than 1 percent during the course of the

reaction.

In hope of sensitizing the rearrangement, the sodium

salt of benzotropone tosylhydrazone was decomposed photolyt-

ically in benzene using xanthen-9-one (ET = 74 Kcal/mole) as

a sensitizer. Unfortunately, no hydrocarbon products were

isolated from the reaction.

The substituted tropone, 2-methyl-4,5-benzotropone was

made according to the method of Hielbronner52 (the conver-

sions necessary to obtain the sodium salt of the corresponding

tosylhydrazone were the same as those used for 4,5-benzotro-

pone itself).

The sodium salt of 2-methyl-4,5-benzotropone tosylhydrazone

was thermally decomposed in benzene at 1300 to yield two

products: 2-vinylnaphthalene (41) and 2-(2,4,6-cyclohepta-

trien-l-yl)-3-methylnaphthalene (42). The authenticity of the




-4Y CH3






2-vinylnaphthalene from this reaction was verified by comparison

of its spectral properties with those of commercially avail-

able material. The substituted cycloheptatrienylnaphthalene

was identified by its spectral properties and by comparison

with material obtained from the thermal decomposition of the

sodium salt of 3-methyl-2-naphthaldehyde tosylhydrazone in

benzene videe infra).

The sodium salt of 3-methyl-2-naphthaldehyde tosylhydra-
zone was thermally decomposed in benzene at 1450 and yielded

the above cycloheptatrienylnaphthalene 42 (39%) and aldazine

(42%).
The products from the decomposition of the benzo precursor

were analyzed by vpc and it was found that the relative ratios

of the products were 2-vinylnaphthalene 72%, unknown material
6%, and 2-(2,4,6-cycloheptatrien-l-yl)-3-methylnaphthalene
22%.

Another annelated tropone, 2,3-benzotropone was prepared

as shown in the scheme below53 (again the conversions to the

tosylhydrazone sodium salt were the same as those necessary

for 4,5-benzotropone).
PPA

CH2-(CH2)3-CO2H


BrZ/CCl1

SiC/DMFBr
LiCl/DMF
[ii4 ^^







The decomposition of the sodium salt of 2,3-benzotropone

tosylhydrazone in benzene at 1200 gave 1-(2,4,6-cyclohepta-

trien-l-yl)naphthalene (43). The compound was identified by









43

its spectral properties as follows: the nmr has resonances

at 2.20 T three a-protons of the naphthyl system; 2.55 r four

B-protons of the naphthyl system; 3.30 T two protons on the

4 and 5 positions of the cycloheptatriene ring; 3.70 T two

protons on the 3 and 6 positions of the cycloheptatriene

ring; 4.50 T two protons on the 2 and 7 positions of the

cycloheptatriene ring; and 6.60 T for the one proton on the

1 position of the cycloheptatriene ring. The mass spectrum

gave peaks at m/e 218 for the molecular ion, 127 for the

naphthyl ring system, and 91 for the tropylium fragment.

That it was the a-substituted naphthalene was shown by the

ratio of the a- to B-protons in the nmr spectrum and by close

examination of the ir spectrum (Cf. 2-(2,4,6-cycloheptatrien-

l-yl)naphthalene); the lack of any absorptions at 861 cm-

and 821 cm1 and the appearance of a band at 800 cm-1 (for

three adjacent H) and the mutual band at 745 cm1 was in

agreement with this structure.












CHAPTER III

DISCUSSION

To date, the interconversion of aryl and aromatic

carbenes2,6,7,9,10,12,16,18 (and nitrenes26,28) --such as the

conversion of phenylcarbene to cycloheptatrienylidene --has

been limited to the gas phase and rather high temperatures

(the exceptions to this have been the ring expansion of 1,6-

methano(10)annulenyl-2-carbene'5 and the small amount of

rearrangement observed in the photolysis at room temperature

of o-, m-, and p-methylphenylcarbeness) Also, with the

exceptions of aromatic carbene dimerizations and of the

reaction of the expanded 1,6-methano(10)annulenyl-2-carbene

with 1,2-cyclononadiene,* in no case has it been possible to

examine intermolecular reactions of any of the rearranged

carbenes. At this time, the rearrangement of 4,5-benzocyclo-

heptatrienylidene to B-naphthylcarbene and that of 2,3-benzo-

cycloheptatrienylidene to a-naphthylcarbene in solution at

temperatures as low as -350 is reported. These reactions

constitute examples of carbene-carbene rearrangements which

are apparently not subject to either of the previous restric-

tions.


W. M. Jones and P. H. Gebert, unpublished results.







For this study, the phenylcarbene-cycloheptatrienyli-

dene reaction was taken as the prototype. Phenylcarbene (44)

was generated in diglyme under conditions where the sodium

salt of tropone tosylhydrazone gives heptafulvalene and the

reaction mixture was carefully examined for cycloheptatrienyli-

dene dimer. No trace was found. Cycloheptatrienylidene

(45) was then generated under conditions where phenylcarbene

gives clean reaction with solvent and the reaction mixture

carefully analyzed for phenylcarbene insertion products. No

trace was found.

The failure of phenylcarbene to expand in solution is

probably due to a favorable competitive intermolecular

reaction with solvent. Cycloheptatrienylidene, on the other


H\C



Insertion < DivIyme g -Diglyme i Dimer


44 45

hand, is very unreactive with most solvents, preferring to

simply dimerize. Thus, its failure to contract reflects an

activation energy high enough to allow dimerization to dominate.

As one possible way to reduce the activation energy for

contraction while hopefully retaining enough of the low

reactivity of the aromatic carbene to prevent extensive

reaction with solvent, benzocycloheptatrienylidene was

studied. This system was chosen because monoannelation is







known to substantially decrease the stability of the tropylium
ion5s but it should have little effect upon the stability of
the intermediate cyclopropene. Thus, if the stability of
the carbene parallels the carbonium ion and if the rate of
contraction is reflected in the relative stability of the
intermediate, then benzocycloheptatrienylidene should contract
faster than cycloheptatrienylidene.
The carbenes were generated from the sodium salts of the
corresponding tosylhydrazones and the rearranged naphthyl-
carbenes were detected by their reaction products with solvent.
Thus, the formation of the naphthylcycloheptatriene 29 suggests
a carbene-carbene rearrangement followed by reaction of the
rearranged carbene 46 with solvent. Formation of the rearranged

H


26





10
46










29







carbene was further supported by the formation of 8-naphthyl-

cyclohexylmethane (34) upon decomposition of the salt in

cyclohexane. Considering decompositions in cyclohexene, it

is interesting to note that the relative yield of the benzo-

cycloheptatrienylidene adduct (33) appears to decrease with

increasing temperature.

The presence of 33 and its apparent decrease with

increased temperature is consistent with a bimolecular

reaction of the unrearranged carbene with cyclohexene (to

give 33) in competition with a unimolecular rearrangement

that gives 8-naphthylcarbene. Lastly, formation of the

various insertion products can be taken as strong evidence5s

for formation of the rearranged carbene.

The mechanism of the carbene-carbene rearrangement, how-

ever, is quite a detailed problem. Initially, one must con-

sider the rearrangement of a carbene precursor. Thus, the

decompositions of B-naphthaldehyde tosylhydrazone sodium salt

and an equivalent amount of 4,5-benzotropone tosylhydrazone

sodium salt were carried out under identical conditions in

the presence of an equimolar amount of dimethyl fumarate.

As expected, the $-naphthyldiazomethane reaction gave as the

primary product the pyrazoline. In contrast, the benzotropone

tosylhydrazone salt reaction gave the now familiar benzene

addition product of 8-naphthylcarbene 29 with no detectable

trace of the pyrazoline 35. When the reactions were carried

out at higher temperatures, the former yielded the 1,2-

dicarbomethoxy-3-naphthylcyclopropane 36 and the latter gave,







in addition to the usual f-naphthylcarbene-benzene adduct, a

new product 37. The nmr of the unisolated material suggests

that it is the thermal (1,5) hydrogen shift product of the

original adduct. All of the above data are consistent with

the rearrangement occurring via a carbene, not a carbene

precursor.

Once it was determined that a carbene was probably the

rearranging species, it became necessary to differentiate

between rearrangement occurring from the singlet state and

that from the triplet state and knowledge of the multiplicity

of the initial carbene was imperative. In principle, this

facet of the problem may be explored by investigating the

stereochemistry of the addition of rearranged 8-naphthyl-

carbene to olefins. Stereospecific addition is taken as

strong evidence for the reaction of the singlet state, whereas

the triplet state appears to add nonstereospecifically.75,s8

Furthermore, the ground state of B-naphthylcarbene has been

shown to be triplet.59 Thus, a high degree of stereospecific

addition would point to the singlet as the reacting species

(unless the singlet and triplet are in equilibrium --this

will be discussed below).

The rearrangement via the triplet state might occur as

shown in the scheme below (this is simply a diradical form

of the cyclopropene mechanism, vide infra).



















Photolytically, in the pure olefin, B-naphthylcarbene

from both precursors gave 96-97% stereospecificity which is

in good agreement with the work of Closs"3 who found 95-97%

stereospecificity with phenylcarbene in pure olefins.

Thermally, the stereospecificity was less than the cor-

responding photochemical reactions (ca. 93%); this may be

because of a thermal dependence of intersystem crossing.57

It is in agreement with the amount of nonstereospecificity

observed in the addition of thermally generated dicyano-

carbene to cis-2-butene (8% in neat olefin), however, singlet-

to-triplet transition in dicyanocarbene is competitive with

addition, and thus the stereospecificity is sensitive to

dilution effects.60

The dilution effect is caused by the different concen-

tration dependence of intersystem crossing and addition to

the olefin; the former is unimolecular and the latter, of

course, is dependent upon olefin concentration." When

applied to the naphthylcarbene system, dilution either with

benzene or octafluorocyclobutane had no effect upon the

stereospecificity.







Another experimental test for multiplicity is scavenging

of paramagnetic species by oxygen or dienes. As Closs

states,57 the stereospecificity should be increased in the

presence of a diene (if the triplet is not in rapid equili-

brium with the singlet) because the scavenging of the para-

magnetic species "results in a correspondingly greater pro-

portion of the singlet-state reaction." Thus, when the B-

naphthylcarbene is generated from the salt of 4,5-benzotropone

tosylhydrazone in the presence of olefin and oxygen, the

addition is somewhat more stereospecific. By making some

approximations from existing data,61 namely that the solubility

of oxygen in benzene is similar to that of butene and that the

solubility is a linear function of temperature, a saturated

solution of 120 millimoles of butene should contain ca. 0.2

millimoles of oxygen under reaction conditions. This concen-

tration was apparently sufficient to increase the stereo-

specificity by 2% (which was also the amount it was increased

when applied to the fluorenylidene system by M. Jones, Jr.).62

Furthermore, more effective scavenging was accomplished

by utilizing trans,trans-2,4-hexadiene. In this case the

effect was pronounced in that the stereospecificity was

increased to 99%. In the hexadiene experiments, a large

part (ca. 90%) of volatile product was apparently from

reaction with the diene (relative to the amount of reaction

with butene as determined by vpc). This is quite in line

with the data of Moss'5 who, using similar conditions,

reported that approximately 70% of the reaction product







appeared to be derived from the reaction of phenylcarbene

with the diene.

In summary, the (-naphthylcarbene added to olefins with

a high degree of stereospecificity which was unchanged by

dilution but was increased by the addition of triplet

scavengers. These data are comparable with a scheme in which

there is a thermally driven singlet-triplet equilibrium

wherein the stereospecificity is independent or only slightly

dependent upon olefin concentration. If there is a rapid

singlet-triplet equilibrium, one can no longer be assured

that the singlet is the rearranging species. The rearrange-

ment could be occurring via the triplet state, but since

there is an equilibrium it might not be reflected in the

stereochemistry.

These data appear inconsistent with a scheme in which

intersystem crossing is in effective competition with addition

of the singlet to olefins as in the case of fluorenylidene.62

The stereospecificity of addition of fluorenylidene to olefins

is dependent upon both the concentration of olefin and the

addition of triplet scavengers. They are also incompatible

with a scheme where there is a rapid singlet-triplet equili-

brium in which the stereospecificity is unaffected by the

addition of triplet scavengers (as in the case of diphenyl-

carbene).57

Initially it seems that a dicotomy exists, but one can

envision a system whereby all of the above facts can be

readily explained and which leaves little doubt but that the

initially formed B-naohthvlcarbene is a sinplet.







The following scheme indicates the effects of changing

the concentrations of diene and olefin upon cis- and trans-

cyclopropane ratio; the explanation of symbols is as follows:

B = benzocycloheptatrienylidene

S = singlet naphthylcarbene

T = triplet naphthylcarbene

D = diene (2,4-hexadiene)

0 = cis-2-butene

P = diene-carbene products


B k--- S




ki [01






cis-


d is= k [0] [S] + k7[O] [T]


ks [D] > p


k7 (0]


k, [O]


trans-


d trans
dt = k4l [O] (T]


d cis
i -t ki[0] [S] + k,[0] [T]
d trans k4 [0] [T]



Integrate over all time and assume the right-hand side to be

constant.







cis k [O] [S] + k7 [0] [T] ki [S] k7
trans k4 [0] [T] I TIV T (1)

Assume a steady-state concentration for [T] which is "particu-

larly good when the intermediates are very reactive and,

therefore, present at very small concentrations.""6

d[T]
= k [S] k [T] kk[T [- ] [T] k [0] [T) k5 [D] [T] = 0


Sk2[S]
[T] = k + k[0O] + k7 [0 + ksl[D (2)


Substitute for [T] from eq. (2) into eq. (1) and simplify

cis k1k3 + kIk4 [O] + kik7 [O] + klks[D] + k2k7
trans k2k4

If there is a rapid equilibrium, k3 >> k [O], k[O]0, ks[D]

and k2 >> ki[0].

Then cis klk3
trans R214i (3)

and, the ratio is independent of olefin and diene concentra-

tions.


If the reaction is irreversible, k [0] >> k3, k [0] >> k3.

Disregard the diene for first case.

Then cis ki(k + k7) [0] + k2k7
trans k2Ik

and the ratio is dependent upon olefin concentration.


Now include the diene in the scheme k [0] >> k3, k7 [0]>> ks,

and ks [0] >> k.

cis kiks[D] + ki(k4 + ky) [0] + k2k7
trans k2k4 (4)

and the ratio is dependent upon diene and olefin concentra-







If, however, k3 >> k4 [0], k3 >> k7 [0], but ks [D] >> k3 then

cis kiks [D] + k2k7 (
trans k2k--- (

and the ratio is dependent only upon the diene concentration.

One must consider, however, that the rearrangement might

be going via the triplet benzocycloheptatrienylidene (one

need not concern oneself with triplet benzocycloheptatrienyli-

dene rearranging to singlet naphthylcarbene). Thus, the

scheme below indicates entry into the triplet rearranged

carbene.
B

k D
p, keD] S k kD [D] p
k3


ki[0] k [0]

k7[0]

cis trans


d is = k [0] [S] + k7 [0] d[T k4 [0] [T]


Integrate as in the first case

cis k [O] [S] + k7 [0] [T ki[S] k (6)
trans k [0] [T] kTTT E (6)

Assume a steady-state concentration of [S].

d ] = k [T] k [S] k [O] [S]- ka[D] [S] = 0


k3 [T]
[S] = k, + k [O] + k [D]







Substitute [S] from eq. (7) into eq. (6) and simplify.


cis klk3 k7
trans k k2 + kO + ka [D] ) (8)

If ka[D] >> k2, but k2 >> ki[0] then


cis = klk3 k7
trans +kk [D] T (9)

and the ratio is inversely proportional to the diene concen-

tration.

Finally, one might envision a scheme wherein the benzo-

cycloheptatrienylidene is the intermediate which undergoes

the singlet-triplet intersystem crossing. Each state of the

unrearranged carbene then yields the corresponding B-naphthyl-

carbene as shown in the scheme below; the explanation of

symbols is as follows:

Bs = singlet benzocycloheptatrienylidene

Bt = triplet benzocycloheptatrienylidene

Ns = singlet naphthylcarbene

Nt = triplet naphthylcarbene

D = diene

O = cis-2-butene

P = diene-carbene products


Bs Bt D p



Ns Nt



cis- trans-







There would be no dilution effect because intersystem

crossing and rearrangement are both unimolecular; the

scavenging effect would arise from reaction of diene with

the unrearranged carbene. In agreement with the arguments

above, the singlet still must be the state by which most of

the rearrangement occurs in order to explain the high degree

of stereospecificity observed.

Excluding the benzocycloheptatrienylidene intersystem

crossing, what is required for the explanation of the facts

concerning the stereospecificity of B-naphthylcarbene addi-

tion is that the rate of triplet-singlet intersystem crossing

proceeds at a rate considerably greater than those of the

additions of the triplet carbene to the mono-olefin. Also,

the rate of triplet addition to the diene (and oxygen) must

be greater than the rate of triplet-singlet intersystem

crossing. Accordingly, it has been reported that diphenyl-

carbene reacts with 1,3-butadiene more than one hundred times

faster than unconjugated olefins.6 The above kinetic scheme

depicts the desired situation in eq. (5) and is also compatible

with the earlier situations of fluorenylidene as shown in

eq. (4) and diphenylcarbene as shown in eq. (3). It also

vitiates the mechanism whereby the triplet is the rearranging

species because the addition of diene would be required to

decrease the cis-trans ratio as shown in eq. (9).

With the triplet pathway ruled out, at least four

mechanisms have been considered which arise from the singlet

state of the unrearranged carbene. The singlet mechanisms







are a carbocyclic version of the Wolff rearrangement, a

Skattebol-type rearrangement, that going via a fused cyclo-

propene intermediate, and a concerted rearrangement.

The Wolff-type rearrangement (or perhaps more accurately

the "retro-Wolff" rearrangement) is shown below. This

..7H
C
.. + H








mechanism is not very favorable for two reasons. First,

annelation should retard it because of loss of aromaticity

in the benzene ring. Second, Jones and Mitsuhashi have postu-

lated that the Wolff and the cyclopropene mechanisms would

differ experimentally in that the latter mechanism should be

favored by high double bond character whereas the Wolff

mechanism should be retarded." Their work on the importance

of bond order in the interconversion of aryl and aromatic

carbenes by studying the gas-phase, carbene-carbene rearrange-

ment of a series of substituted naphthylcarbenes,showed that

net cleavage of bonds of high bond order is observed; net

cleavage of bonds of low bond order is not. Thus, the

evidence mitigates against a Wolff mechanism.

Another mechanism which can be written for the contrac-

tion of cycloheptatrienylidene (and also for benzocyclohepta-

trienylidene) is essentially the Skattebol' vinylcyclopropyli-






dene-cyclopentenylidene rearrangement with cleavage of
different bonds. However, the effect of annelation argues
against this possibility since it should retard norcaradiene
formation (to prevent loss of aromaticity) and hence the rate
of contraction.
H

> /
.C






The differentiation of the mechanisms involving the fused
cyclopropene (route a) and the concerted pathway (route b)
present a subtle problem. In fact, unless the benzobicyclo-
heptatriene intermediate can be trapped or observed (spectro-
scopically, etc.) there can be no real distinction between
the two. If the reaction is not concerted, the activation



H\ H\
c:


/-b9-9>-
8 89







energy for ring opening must be quite low* as evidenced by

the fact that benzobicycloheptatrienylidene contracts to

B-naphthylcarbene at temperatures as low as room temperature.

The low activation energy is not unreasonable since the

energy lost in cleaving the single bond of the twisted cyclo-

propene (the benzobicycloheptatrienylidene) might well be of

the same order of magnitude as the naphthyl resonance energy

gained. In the model system it has been suggested that

"relatively small energy barriers separate phenylcarbene,

cycloheptatrienylidene, and bicycloheptatrienylidene isomers."12

For the discussion of a detailed mechanism of the

rearrangement it will be assumed that the process occurs via

the fused cyclopropene intermediate since this facilitates

the dissection of the problem and because it seems intuitively

likely that the cyclopropene is a distinct intermediate.

Accordingly, Bergman6 recently reported that the thermal

isomerization of a substituted cyclopropene very likely went

through a vinylcarbene intermediate. He gave no suggestions

as to the pathway of ring opening and closure except to rule

out a "diradical" mechanism.


Assuming that the intermediate has a half-life of 0.1 sec

and that the pre-exponential factor (A) is similar to that for

cyclopropene opening,6s a value of ca. 15 Kcal/mole is found

for the energy of activation at 270.







Thus while the preferred conformation for closure of

vinylcarbenes to cyclopropenes is unknown, it may occur

from a structure in which the filled orbital of the carbene

is orthogonal to the a-system which was suggested by Closs67

and by calculations which imply that this should be the

conformation of the ground-state singlet of vinylcarbene.68

Using Zimmerman's69 MO Following approach to correlation

diagrams, one can arrive at the preferred conformation of the

rearranging (or rearranged) carbene. This approach is applied

to cases where there is no element of symmetry and which con-

sist of a linear array of orbitals. Considering benzocyclo-

heptatrienylidene and treating the system as a vinylcarbene-

cyclopropene addition, the remainder of the benzenoid system

is disregarded (the subsequent ring opening is just the

reverse of the addition). Application of Zimmerman's "four

basic rules" yields the correlation diagram shown in Figure 4.

It is immediately obvious that p, P4 is the occupied orbital

(in the singlet state) and, since it accepts the electron

pair, it will have added s character. The final result is

that of the benzocycloheptatrienylidene with the vacant

orbital vertically oriented (pl + P4) and conjugated with

the w-system, and the orbital containing the non-bonded

electrons, s + (pl P4) weighted, horizontally oriented in

a plane containing the sigma framework.












1 4



2 _


(/23) 12
01 2


(1,3)

S," ( 3 )

+1-2+3-4


S (2.,3) *
3 4


(2)

pi+ps .1.. +-+1-2-3+4


PI-P, --
(1)
+1+2-3-4\


(o)
T (0 +1+2+3+4
S23


(0.1 4


12Figure 4. Closure of Vinylcarbene to Cyclopopene.

Figure 4. Closure of Vinylcarbene to Cyclopropene.








In hope of elucidating the mechanism somewhat, the

effect of methyl substitution upon the rate of carbene-

carbene rearrangement was cursorily examined. From the

preferred formation of 41 versus that of 42 in the decompo-

sition of the sodium salt of 2-methyl-4,5-benzotropone tosyl-

hydrazone in benzene, it appears that alkyl substitution

favors rearrangement. The validity of this conclusion


H CH3
C CHa CH3






48 47






CH3

42 41
rests on the fact that neither 3-methyl-2-naphthylcarbene (48)

nor methyl-2-naphthylcarbene (47) ring expands under the

reaction conditions and thus indicates that both reactions

are irreversible.* It is recognized that this datum will

not enable one to pinpoint a single mechanism, but in order


* From the product distribution, the free energy differ-

ence between the two product transition states is on the

order of 1 Kcal/mole at 1300.








to explain it, attention must be focused on some interesting

concepts about this general type of rearrangement. In dis-

secting this problem, two types of interactions or conforma-

tions must be considered: a boat-shaped system wherein the

ring containing the carbene center is puckered and delocali-

zation of n-electrons into the carbene p orbital is minimal,

and a planar system with delocalization of r-electrons into

the vacant p orbital of the carbene (similar to cyclohepta-

trienylidene8 which calculations imply has a planar struc-

ture).* In each of these two conformations, one must examine

three effects: steric, geometric, and electronic.

Models suggest that steric effects are unimportant

since in neither conformation is the methyl group able to

significantly aid in the formation of the transition state by

lessening any non-bonded interactions.

The transformation from the boat conformation (Figure 5)

could be envisioned as occurring by interaction of the empty











CH3
Figure 5. Boat Conformation of Benzocycloheptatrienylidene.


J. Sabin, Physics Department, University of Florida,

personal communication.







p orbital of the carbene with the rotating p orbital of the

y-carbon to form the new sigma bond and by interaction of

the sp2 orbital of the carbene with the p orbital of the

B-carbon to yield the twisted cyclopropene double bond.

From an examination of models, it appears that in the

boat conformation there is very good orbital orientation for

efficient overlap. Thus, it seems readily adapted to the

formation of a fused, twisted cyclopropene (i.e., the ground

state geometry of the boat form closely resembles that of a

possible transition state). It is quite possible that the

boat form is the lower energy conformation in this system in

spite of the assumption that the benzotropylium cation has a

planar structure.70 However, coplanarity is not requisite

as the activation energy for the rearrangement may include

the energy necessary for the planar system to become puckered.

In the planar case (Figure 6), the cyclopropene sigma

bond would be formed by the interaction of the back lobe of

the inverting sp2 orbital on the carbene carbon with the

rotating p orbital on the y-carbon. Localization of a pair


Figure 6. Planar Conformation of Benzocycloheptatrienylidene.








of n-electrons in the still conjugated p orbitals of the a-

and B-carbons would then form the new double bond (in a

manner similar to the nucleophilic attack of carbenes upon

electron-deficient double bonds).8

The planar form has to rehybridize to accomplish the

inversion at the carbene center (this is not required of

vinylcarbene cyclization). From examination of models, this

rehybridization introduces strain into the ring before there

is what appears to be good orbital overlap; this could increase

the energy of activation for closure. That is, in contrast

to the boat form, in this case there seems to be less initial

interaction leading to the rehybridization in that initially

the two potentially interacting orbitals are orthogonal.

The published electron effects are obscure and, conse-

quently, lend very little aid to differentiation of the two

modes of attack. Closs attributes alkyl substitution effects

to the possibility of both steric and electronic components

in the formation of cyclopropenes from the ring closure.of

vinylcarbenes;67 it appeared, though, that the steric portion

was dominant. Closs also observed that an alkyl group in the

olefinic position of the cyclopropene was considerably less

effective in its ability to stabilize the cyclopropene than

alkyl groups at the saturated center.71

From an examination of possible electronic effects in

the boat conformation, it appears that in the formation of

the transition state of the rearrangement the B-carbon

acquires a partial positive charge. This charge deficiency







could arise from loss of conjugation with the p orbital of

the y-carbon due to the latter's rotation toward the carbene

center. What the electron deficiency requires is a greater

degree of bond breaking between B- and y-carbons than bond

formation between the a- and B-carbons (shown below) which













CH3



is not unlike the unequal bond formation predicted for the

electrophilic addition of carbenes to double bonds.72 It is

known that alkyl substitution facilitates this type of addi-

tion. 3

Unfortunately, one cannot make conclusive predictions

about the electronic effects of the methyl group in the

planar conformation. There is the possibility, though, that

the methyl group would have little or no effect upon the rate

of the rearrangement. Thus, concerning the closure of vinyl-

carbenes to cyclopropenes, Closs67 stated that with progres-

sive rotation of the y-carbon there was the possibility of a

decrease of r-electron density at that site due to the forma-

tion of the double bond between the a- and B-carbons








(implying little change in electron density at the B-carbon).

Using the same reasoning, it would appear that there would

be little or no charge deficiency on the methyl-substituted

3-carbon of the planar system resulting in little or no

effect by the methyl group on the rearrangement.

It should now be pointed out that while geometric and

electronic effects do not form any real basis for prediction

of a preferred pathway, they are certainly consistent with

the "electrophilic-like" type of interaction discussed

earlier. Thus, the prediction of Closs and the calculations

of Zimmerman need not be incorrect as vinylcarbene may

rearrange in the manner described by them. But, carbocyclic,

aromatic carbenes may well rearrange by the other pathway

due to electronic effects brought about by geometric constraints.

For example, the apparent similarity of the boat form to a

possible transition state, and, in contrast to the assumed

geometry for the closure of vinylcarbenes,67 the geometry of

cycloheptatrienylidenes is such that the sp2 orbital of the

carbene is trans- to the double bond.

A clue to the reason for the rather dramatic effect of

annelation on the ease of ring contraction may be found in

the effect of annelation on the pKR+ values of model cations.5

Carbonium ions should be fair models for these carbenes since,

in aromatic carbenes, the vacant orbital is probably conju-

gated with the 7-system with the non-bonded pair of electrons

in the plane of the sigma framework.7 The tropyl cation has

a value of 4.7 and the benzotropyl cation has a value of 1.7.







Annelation would not be expected to substantially change

the stability of the intermediate, but annelation of the

tropyl cation causes a marked decrease in the pKR+ signaling

a substantial destabilization. Thus, the energy of activa-

tion for ring contraction should be less in the annelated

system than in the nonannelated one.

von R. Schleyer* has done "back of the envelope" calcu-

lations on the "strain energy" on phenylcarbene, cyclohepta-

trienylidene, and the bicycloheptatriene and come to the

conclusion that they are all of similar energy. Thus, it

appears reasonable that even modest annelation effects could

be enough to cause the energetic of the annelated system to

favor ring contraction over the ring expansion exhibited in

the cycloheptatrienylidene series.

Lastly, in order to test the generality of the benzo-

cycloheptatrienylidene-naphthylcarbene rearrangement, the

sodium salt of 2,3-benzotropone tosylhydrazone was decomposed

in benzene to yield the a-naphthylcycloheptatriene. The

smooth rearrangement fits perfectly into the overall picture

of the rearrangement.

In summary, benzocycloheptatrienylidenes rearrange to

naphthylcarbenes in solution thermally at temperatures in

the range of 1000 and photolytically at temperatures as low

as -350 with the rearranged carbenes being detected by their

reaction products with solvent. The rearrangement may occur


P. von R. Schleyer, Princeton University, personal com-

munication.





56

via a fused cyclopropene intermediate from the boat form of

the singlet unrearranged carbene, from the planar form of

the singlet, or via a concerted pathway.












CHAPTER IV

EXPERIMENTAL

General.--Melting points were taken in a Thomas-Hoover

Unimelt apparatus and are uncorrected. Elemental analyses

were performed by Atlantic Microlab, Inc., Atlanta, Georgia.

Ultraviolet spectra (uv) were recorded on a Cary 15 double-

beam spectrophotometer using 1 cm silica cells. Infrared

spectra (ir) were recorded with a Beckman IR10 with absorp-

tions reported in reciprocal centimeters. In all cases where

the KBr pellet technique was not used, sodium chloride plates

were substituted. Nuclear magnetic resonance spectra (nmr)

were determined on a Varian A-60A high resolution spectrometer

or a Varian XL 100 spectrometer. Chemical shifts are reported

in tau (T) values from internal tetramethylsilane standard.

Mass spectra were determined on a Hitachi model RMU-6E mass

spectrometer.

Analytical thin-layer chromatography (tlc) was accom-

plished on 2 in x 8 in plates coated in these laboratories

with 0.25 mm layers of E. Merck HF-254 silica gel. Components

were visualized by their quenching of fluorescence under uv

light. Vapor phase chromatography (vpc) was carried out

using a Varian Aerograph 90-P thermal conductivity instrument

using column A (18 ft x 0.25 in in 20% SE-30 on Chromosorb W

at 2100) or column B (3 ft x 0.25 in 15% SE-30 on Chromosorb

W at 1900) and a Varian Aerograph Series 1200 flame ioniza-







tion instrument using column C (9 ft x 0.125 in 5% SE-30 on

Chromosorb W at 1550). The analytical technique for deter-

mining product ratios from vpc data was that of cutting and

weighing. All chemicals were reagent grade unless otherwise

stated. Diglyme (diethylene glycol dimethyl ether) and

tetrahydrofuran were dried by passage over activity grade I

Woelm alumina and subsequent storage over CaH2. The analyzed

research grade cis-2-butene was from Matheson Gas Products,

Inc. (cis-2-butene, 99.94%; trans-2-butene, 0.04%; and buta-

diene, 0.02%).


Pyrolysis of the Sodium Salt of Benzaldehyde Tosylhydra-

zone in Diglyme.--To 30 ml of dry diglyme in a 3 oz Fischer-

Porter Aerosol Compatibility Tube was added 0.10 g (0.34

mmol) of the sodium salt of benzaldehyde tosylhydrazone.38

The sealed tube was placed in an oil bath preheated to 1650

for 30 min. At the end of this time all trace of color of

the diazo was gone, leaving the solution colorless. The

mixture was allowed to cool, poured into water (200 ml), and

extracted with pentane (three 30 ml portions). The combined

pentane extracts were washed with water (three 50 ml portions),

dried (MgSO4), and concentrated on a rotary evaporator under

reduced pressure. The nmr spectrum of the residue, a light

yellow oil, indicated no heptafulvalene was present, only

what is believed to be the phenylcarbene-diglyme adducts.

The nmr (CDCls) showed r 2.80 (phenyl protons) and 6.5

(diglyme protons). In addition, there was aldazine produced

in the reaction. Comparative tic of the residue with a






known sample of heptafulvalene was carried out using silica

gel eluting with pentane. This confirmed the absence of

heptafulvalene in the reaction products.


Pyrolysis of the Sodium Salt of Tropone Tosylhydrazone

in Diglyme.--Into a Carius tube was placed 80 mg (0.27 mmol)

of the sodium salt of tropone tosylhydrazone8 and 10 ml of

dry diglyme. The sealed tube was immersed in an oil bath

preheated to 2100 and allowed to remain for 30 min. Upon

cooling, the tube was opened and the contents poured into

water (120 ml) and extracted with pentane (three 33 ml por-

tions). The combined pentane extracts were washed with

water (three 50 ml portions) and dried (MgSO). The dry

pentane solution was concentrated on a rotary evaporator

under reduced pressure to yield a very dark solid. The nmr

spectrum of the residue was identical to that of heptaful-

valene with a small amount of aromatic resonance at 7.2 T,

but none at 6.2-7.0 .(the position expected for any diglyme

insertion products).


4,5-Benzotropone.--The 4,5-benzotropone used in this

research was prepared by the method of Thiele and Weitz3

with the exception that a steel high pressure reactor was

used in place of a sealed glass tube.


4,5-Benzotropone Tosylhydrazone.--In a typical run,

2.00 g (0.013 mol) of 4,5-benzotropone, 2.38 g (0.013 mol)

of p-toluenesulfonylhydrazide, and 2 drops of cone sulfuric

acid were placed in 55 ml of 95% ethyl alcohol and heated at







reflux for 1 hr. Upon cooling, 4,5-benzotropone tosylhydra-

zone precipitated from the solution and was isolated by fil-

tration. Recrystallization from absolute ethyl alcohol

yielded 2.5 g (51%) of the product. There were two crystal

forms of the tosylhydrazone: red plates (mp 186-1890) and

yellow needles (mp 180-1850). The other properties of the

tosylhydrazone were as follows: ir (KBr) 3200, 1638, 1325,

1165, 575, 555 cm-l; nmr (acetone-d6) T 2.00-2.80 (A2B2, 4H,

J=8, toluyl aromatic), 2.62 (s, 4H, benzo), 2.90-3.80 (m, 4H,

cycloheptatrienyl), 7.60 (s, 3H, toluyl methyl); mass spectrum

324 (M+).

Anal. Calcd. for C18Hi6N202S: C, 66.64; H, 4.97; N,

8.64; S, 9.89. Found: C, 66.52; H, 5.07; N, 8.57; S, 9.97.


B-Naphthaldehyde Tosylhydrazone.--In a typical run,

B-naphthaldehyde (Aldrich Chemical Co.), an equivalent amount

of p-toluenesulfonylhydrazide, and a few drops of conc HC1

were placed in 95% ethyl alcohol and allowed to stir overnight

at room temperature and the resulting crystals filtered and

recrystallized from ethyl alcohol to give the tosylhydrazone.

The product had a mp 1750, lit.36 1740.

General Preparation of the Sodium Salts of Tosylhydra-

zones.--The sodium salts were prepared by dissolving the

tosylhydrazone in dry THF and adding one equivalent of sodium

hydride (57% in mineral oil Alfa Inorganics) slowly with

stirring. The mixture was allowed to stir for 1 hr. At the

end of this period, pentane was added to precipitate the







sodium salt (this was not always necessary). The salt was

filtered, washed with additional pentane, and dried. The

entire operation was carried out in a dry box, although some

of the salts seemed to be stable to laboratory conditions.


Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosyl-

hydrazone in Diglyme.--To 13 ml of dry diglyme in a 3 oz

Fischer-Porter Aerosol Compatibility Tube was added 0.45 g

(1.3 mmol) of the sodium salt of 4,5-benzotropone tosylhydra-

zone. The sealed tube was placed for 20 min in an oil bath

which had been preheated to 1000. At the end of this time

78% of the theoretical amount of nitrogen had been evolved.

The reaction mixture was allowed to cool, poured into water

(200 ml), and extracted with pentane (three 30 ml portions).

The combined pentane extracts were washed with water (three

50 ml portions), dried (MgSO4), and concentrated on a rotary

evaporator under reduced pressure. The remaining yellow oil

was chromatographed on alumina and eluted with ethyl ether to

yield 103 mg (29%) of the naphthylcarbene-diglyme insertion

products. Although the yield was only 29%, the nmr spectrum

of the crude reaction mixture showed mainly diglyme and the

adducts. The adducts had the following properties: ir (film)

3050, 2920, 2880, 1600, 1450, 1100, 860, 820, and 750 cm-1;

nmr (CDCls) T 2.00-2.80 (m, 7H, naphthyl), 6.10-7.20 (m, 15H,

diglyme and benzylic); mass spectrum 274 (M+).

Anal. Calcd. for C17H220s : C, 74.42; H, 8.08; 0, 17.49.

Found: C, 74.59; H, 8.05; 0, 17.36.







The sodium salt of B-naphthaldehyde tosylhydrazone was

thermally decomposed and worked up in a manner analogous to

the above pyrolysis. The naphthylcarbene-diglyme adduct was

isolated in 13% yield; the major product was naphthaldazine.

The adduct's nmr spectrum was identical to that of the

products from the pyrolysis of the sodium salt of 4,5-benzo-

tropone tosylhydrazone videe supra).


Photolysis of the Sodium Salt of 4,5-Benzotropone Tosyl-

hydrazone in Diglyme.--To 15 ml of dry diglyme in a photolysis

vessel which consisted of a Pyrex tube with a side arm and

equipped with a small mechanical stirrer, was added 250 mg

(0.72 mmol) of the sodium salt of 4,5-benzotropone tosylhydra-

zone. The mixture was rapidly stirred and the photolysis

carried out by mounting the apparatus vertically alongside

and as close as possible to a 550 watt Hanovia medium pressure

mercury lamp in a Pyrex tube. Both the vessel and the tube

containing the lamp were immersed in a circulating water

bath which kept the temperature at 300. The salt was photolyzed

for 2 hr with 100% of the theoretical amount of nitrogen

being evolved as monitored by a gas buret fitted to the side

arm of the photolysis vessel. The reaction mixture was

worked up as in the above pyrolysis to yield 51 mg (25%) of

the naphthylcarbene-diglyme insertion products. It should

be noted that while the yield was low, the nmr spectrum of

the crude reaction mixture indicated that the diglyme adducts

were the major products.







Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosyl-

hydrazone in Benzene.--To a 3 oz Fischer-Porter Aerosol

Compatibility Tube containing 35 ml of benzene was added

0.20 g (0.58 mmol) of the sodium salt of 4,5-benzotropone

tosylhydrazone. The sealed tube was placed in an oil bath

preheated to 1200 and allowed to remain for 2.5 hr. The

color of the reaction mixture changed from orange to light

yellow during the course of the reaction. Upon cooling, the

nitrogen evolution was monitored and 81% of the theoretical

amount had evolved. The reaction mixture was filtered and

the residue was washed with ethyl ether. The total filtrate

was concentrated on a rotary evaporator under reduced pressure

to give 122 mg (96%) of crude 2-(2,4,6-cycloheptatrien-l-yl)-

naphthalene. Recrystallization from abs methyl alcohol

yielded 68 mg (54%) of yellow crystals mp 85-860 (sublimation

at 900 and 0.1 mm Hg gave white plates with same mp): ir

(KBr) 3060, 3020, 1600, 1500, 1280, 1260, 950, 900, 860, 820,

745, 700, 480 cm-1; nmr (CDC13) T 2.10-2.70 (m, 7H, naphthyl),

3.26 (broad t, 2H, 4 and 5 cycloheptatrienyl), 3.55-3.90

(complex m, 2H, 3 and 6 cycloheptatrienyl), 4.33-4.70 (m, 2H,

2 and 7 cycloheptatrienyl), 7.10 (m, 1H, methine); mass

spectrum 218 (M+).

Anal. Calcd. for C17H14: C, 93.54; H, 6.46. Found:

C, 93.39; H, 6.48.

The sodium salt of F-naphthaldehyde tosylhydrazone was

pyrolyzed and worked up in a manner analogous to the one

above to yield 78% 2-(2,4,6-cycloheptatrien-l-yl)naphthalene

and 6% of the aldazine.







Photolysis of the Sodium Salt of 4,5-Benzotropone Tosyl-

hydrazone in Benzene.--To the photolysis vessel, previously

described in the photolysis of 4,5-benzotropone tosylhydrazone

sodium salt in diglyme, containing 60 ml of benzene was added

0.10 g (0.29 mmol) of the sodium salt of 4,5-benzotropone

tosylhydrazone. The mixture was stirred and irradiated with

a 550 watt Hanovia medium pressure mercury lamp for 30 min

at 300. The nitrogen evolved was 84% of the theoretical

amount. The reaction mixture was filtered and the residue

was washed with ethyl ether. The total filtrate was concen-

trated on a rotary evaporator under reduced pressure to

yield 37 mg (59%) of 2-(2,4,6-cycloheptatrien-l-yl)naphtha-

lene which had all the properties of the pyrolysis product.

The sodium salt of 0-naphthaldehyde was photolyzed in a

manner analogous to the above salt to yield 67% of 2-(2,4,6-

cycloheptatrien-1-yl)naphthalene.


Preparation of 2-naphthaldazine.--The compound was

prepared by the general procedure given in "Organic Synthesis"''"

and it had mp 235-2360. Lit. 2310.0


Pyrolysis of the Sodium Salt of B-Naphthaldehyde Tosyl-

hydrazone in Cyclohexene.--To a 3 oz Fischer-Porter Aerosol

Compatibility Tube containing 15 ml of distilled cyclohexene

was added 0.40 g (1.15 mmol) of the sodium salt of B-naphth-

aldehyde tosylhydrazone. The sealed tube was placed in an

oil bath preheated to 1350 and allowed to remain for 70 min

while the temperature was raised to 1450. Upon cooling, the







mixture was filtered and the residue was washed with ethyl

ether. The total filtrate was concentrated on a rotary

evaporator to yield 212 mg of an oil. Analysis by vpc indi-

cated there were three volatile components which were sepa-

rated by preparative vpc (column A): (cyclohexene-3-yl)-2-

naphthylmethane (34%) retention time 30 min, syn-2-(7-nor-

caryl)naphthalene- (26%) retention time 32 min, and anti-2-

(7-norcaryl)naphthalenea (40%) retention time 38 min. There

was about 10% aldazine formed in the reaction but it did not

appear in the vpc chromatogram, thus the above data only

refer to volatile hydrocarbon products.

(cyclohexene-3-yl)-2-naphthylmethane showed: ir (film)

3050, 3020, 2920, 2850, 1600, 1500, 1445, 855, 820, 750 cm-1;

nmr (CDClI) T 2.00-2.90 (m, 7H, naphthyl), 4.40 (s, 2H, vinyl),

7.35 (s, 2H, methylene), 7.80-9.20 (m, 7H, cyclohexyl); mass

spectrum 222 (M+).

Anal. Calcd. for C17H16: C, 91.84; H, 8.16. Found:

C, 91.64; H, 8.20.

syn-2-(7-norcaryl)naphthalene showed: mp 68-700;

ir (melt) 3050, 3000, 2930, 2860, 1600, 1500, 1445, 830, 750,
-i
725 cm 1; nmr (CDCI3) T 2.00-2.80 (m, 7H, naphthyl), 6.60-

9.50 (m, 11H, norcaryl); mass spectrum 222 (M+).

Anal. Calcd. for C17Hia: C, 91.84; H, 8.16. Found:

C, 91.83; H, 8.20.

anti-2-(7-norcaryl)naphthalene showed: ir (film) 3050,

3005, 2925, 2850, 1625, 1598, 1505, 1442, 809, 778, 740 cm"-1

nmr (CDC13) T 2.20-3.10 (m, 7H, naphthyl), 7.90-9.30 (m, 11H,

norcaryl); mass spectrum 222 (M+).








Anal. Calcd. for C17Hi8: C, 91.84; H, 8.16. Found:

C, 91.56; H, 8.09.


Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosyl-

hydrazone in Cyclohexene at 830.--Into a 50 ml three-neck,

round-bottom flask fitted with a thermometer, condenser with

a drying tube, and a solid-addition funnel, was placed 25 ml

of distilled cyclohexene. The cyclohexene was heated to

reflux and 0.10g (0.29 mmol) of the sodium salt of 4,5-

benzotropone tosylhydrazone was added by means of the solid-

addition funnel. The mixture was heated at reflux for 4 hr

and then stopped even though the reaction was incomplete as

indicated by the presence of an orange color in the reaction

flask. The reaction mixture was filtered and the residue

was washed with ethyl ether. The total filtrate was concen-

trated on a rotary evaporator under reduced pressure. The

remaining oil was subjected to vpc analysis (column A) which

indicated five volatile components. The ratio of products

was two unknown products (66%), the naphthylcarbene-cyclo-

hexene insertion product (7%), syn-norcarane (12%), and anti-

norcarane (15%). While neither of the unknown products has

been completely characterized, one appears to be the adduct

that would result from the addition of benzocycloheptatrien-

ylidene to cyclohexene. The compound showed ir (film)
-l
3010, 2920, 2580, 1585, 1480, 1440, 795, 745, 700 cm-1; nmr

(CDC13) T 2.75 (d, 4H), 3.25 (d, IH), 3.60 (d,lH), 4.50 (d,

1H), 4.66 (d, 1H), 8.70 (broad m, 10H); mass spectrum 222 (M+);







uv (95% C2HsOH) 240 mi (log E = 4.2), 265 (3.5).

Anal. No satisfactory analysis was obtained (probably

due to insufficient sample size).


Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosyl-

hydrazone in Cyclohexene at 1400.--Into a 3 oz Fischer-Porter

Aerosol Compatibility Tube containing 15 ml of distilled

cyclohexene was introduced 0.40 g (1.15 mmol) of the sodium

salt of 4,5-benzotropone tosylhydrazone. The tube was

immersed in an oil bath preheated to 1350 and allowed to

remain for 70 min while the temperature was increased to 1450.

The mixture was allowed to cool and was filtered. The residue

was washed with ethyl ether and the total filtrate concentrated

on a rotary evaporator under reduced pressure to yield 150 mg

(58%) of products. Analysis by vpc (column A) indicated the

same five products as the 830 pyrolysis: the two "unknown"

products, the insertion product, the syn-norcarane, and the

anti-norcarane. The amount of unrearranged-carbene products

relative to the naphthyl products was approximately the same

as in the 830 pyrolysis.


Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosyl-

hydrazone in Cyclohexene--Diglyme at 820.--Into a 3 oz

Fischer-Porter Aerosol Compatibility Tube containing 17.5 g

(0.13 mol) of dry diglyme and 4.95 g (0.06 mol) of distilled

cyclohexene was introduced 0.10 g (0.29 mmol) of the sodium

salt of 4,5-benzotropone tosylhydrazone. The sealed tube

was placed in an oil bath preheated to 820 and allowed to








stir for 11 hr. The reaction mixture was allowed to cool and

poured into water (100 ml), extracted with pentane (three 35

ml portions), and the combined pentane extracts washed with

water (three 75 ml portions). The pentane solution was dried

(MgS04) and concentrated on a rotary evaporator under reduced

pressure to yield 105 mg of products. Analysis of volatile

hydrocarbon products by vpc (column A) indicated the "un-

known" products, the insertion product, the syn-norcarane,

and the anti-norcarane. There were also diglyme insertion

products amounting to 27% of the total products. Interfering

materials prevented accurate analysis of the mixture.


Photolysis of 4,5-Benzotropone Tosylhydrazone Sodium

Salt in Cyclohexene.--To the photolysis vessel, previously

described in the photolysis of 4,5-benzotropone tosylhydrazone

sodium salt in diglyme, containing 60 ml of distilled cyclo-

hexene was added 200 mg (0.58 mmol) of the sodium salt of 4,5-

benzotropone tosylhydrazone. The mixture was stirred and

irradiated with a 550 watt Hanovia medium pressure mercury

lamp for 1.25 hr. The photosylate was filtered, and the fil-

trate was concentrated on a rotary evaporator at reduced pres-

sure to yield 67 mg of product. Analysis by vpc (column A)

of the volatile hydrocarbon products indicated the "unknown"

products, a new, additional unknown product, the insertion

product, the syn-norcarane, and the anti-norcarane. The

amount of unrearranged-carbene products relative to the

naphthyl products appeared to be greater than in the pyrolyses.







Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosyl-

hydrazone in Cyclohexane.--In a 3 oz Fischer-Porter Aerosol

Compatibility Tube containing 20 ml of cyclohexane was placed

0.40 g (1.1 mmol) of the sodium salt of 4,5-benzotropone

tosylhydrazone. The sealed tube was placed in an oil bath

which was preheated to 1150 and allowed to remain for 1.25 hr.

Nitrogen evolution was 100% of the theoretical amount. The

reaction mixture was allowed to cool and then filtered. The

residue was washed with ethyl ether and the total filtrate

concentrated on a rotary evaporator under reduced pressure.

The remaining yellow oil was chromatographed on alumina,

eluted with pentane, yielding 131 mg (51%) of (B-naphthyl)

cyclohexylmethane. The nmr spectrum of the crude material

indicated almost exclusively adduct and cyclohexane. An

analytical sample of the compound was purified by preparative

vpc (column A at 2220) retention time 32 min and had the

following properties: ir (film) 3042, 3010, 2920, 2845, 1601,
-l
1445, 815, 795 cm1 ; nmr (CDCIa) r 2.10-2.90 (m, 7H, naphthyl),

7.35 (d, 2H, J=7, benzylic), 8.10-9.20 (m, 11H, cyclohexyl);

mass spectrum 224 (M+), 141 (C11H9), 83 (C6H11).

Anal. Calcd. for C17H20: C, .91.01; H, 8.99. Found:

C, 90.84; H, 8.94.


Reaction of the Sodium Salt of 4,5-Benzotropone Tosyl-

hydrazone with Dimethyl Fumarate at 105S.--Into a 3 oz

Fischer-Porter Aerosol Compatibility Tube containing 20 ml

of benzene were placed 0.20 g (0.58 mmol) of the sodium salt

of 4,5-benzotropone tosylhydrazone and 83 mg (0.58 mmol) of







dimethyl fumarate (previously washed with saturated aqueous

sodium bicarbonate and then recrystallized from chloroform).

The sealed tube was placed in an oil bath preheated to 1050

and allowed to remain for 5 hr. Upon cooling, the mixture

was filtered. The residue was washed with ethyl ether and

the total filtrate concentrated on a rotary evaporator under

reduced pressure to yield 76 mg (63%) of 2-(2,4,6-cyclohepta-

trien-l-yl)naphthalene. The nmr spectrum of the crude reac-

tion material indicated the above mentioned adduct and a con-

siderable amount of unreacted dimethyl fumarate. Comparative

tic on silica gel eluted with 50/50 ethyl ether--pentane of

the known pyrazoline videe infra) and this reaction mixture

indicated no pyrazoline was present in the latter.


Reaction of the Sodium Salt of B-Naphthaldehyde Tosyl-

hydrazone with Dimethyl Fumarate at 1050.--To a 3 oz Fischer-

Porter Aerosol Compatibility Tube containing 20 ml of benzene

was added 0.20 g (0.58 mmol) of the sodium salt of B-naphth-

aldehyde tosylhydrazone and 83 mg (0.58 mmol) of dimethyl

fumarate (previously washed with saturated aqueous sodium

bicarbonate and recrystallized from chloroform). The sealed

tube was placed in an oil bath preheated to 1050 and was

allowed to remain there for 5 hr. Upon cooling, the reaction

mixture was filtered. The residue was washed with ethyl

ether and the total filtrate was concentrated on a rotary

evaporator yielding 90 mg (50%) of an almost colorless, very

viscous oil. The oil was not stable to ordinary distillation

conditions and even though column chromatography on silica gel







effected some purification, it was not sufficient to provide

a sample which gave a good elemental analysis (e.g. Calcd.

N, 8.97; found: N, 8.31). The other physical data were all

consistent with the 3,4-dicarbomethoxy-5-naphthyl-2-pyrazoline

structure: ir (film) 3440 (N-H), 3050, 2945, 1730 (unconju-

gated ester carbonyl), 1700 (conjugated ester carbonyl), 1555

(C=N), 1440, 1200, 1015, 855, 815, and 750 cm-1; nmr (CDC13)

T 2.10-2.60 (m, 7H, naphthyl), 3.24 (broad s, IH, nitrogen),

4.60 (broad d, 1H, J=9, benzylic), 5.92 (d, 1H, J=9, a-carbo-

methoxy), 6.18 (s, 3H, conjugated ester methyl), 6.24 (s, 3H,

unconjugated ester methyl)*; mass spectrum no (M+), 284 (M -

28); uv (95% C2HsOH) 224 mp (log c = 4.53), 275 (3.75), 287

(3.79).


Reaction of g-Naphthaldehyde Tosylhydrazone Sodium Salt

with Dimethyl Fumarate at 1600.--This decomposition was

carried out in the same manner as the low temperature one

with the exception that the sealed tube was placed in an oil

bath preheated to 1050 and allowed to remain for 2 hr while

the temperature was raised to 1600. Upon cooling, the mixture

was filtered. The filter paper and the residue were washed

with ethyl ether and the total filtrate concentrated on a

rotary evaporator under reduced pressure to yield 110 mg (67%)


In addition there were other resonances which changed in

intensity depending upon work up and purification steps

and are, therefore, not believed to originate from the

pyrazoline.








of crude trans-1,2-dicarbomethoxy-3-(2-naphthyl)cyclopropane.

The solid was recrystallized from abs methyl alcohol to yield

35 mg (21%) of pink needles mp 91-930: ir (CC14) 3028, 2975,

1730, 1435, 1300, 1170, and 909 cm-1; nmr (CDC13) T 2.00-2.90

(m, 7H, naphthyl), 6.23 (s, 3H, methyl), 6.58 (s, 3H, methyl),

6.70-7.50 (m, 3H, cyclopropane ring); mass spectrum 284 (M+).

Anal. Calcd. for C17H1i6O : C, 71.82; H, 5.67; 0, 22.51.

Found: C, 71.93; H, 5.73; 0, 22.34.


Reaction of the Sodium Salt of 4,5-Benzotropone Tosyl-

hydrazone with Dimethyl Fumarate at 1600.--This decomposition

was carried out in the same manner as the low temperature one

with the exception that the sealed tube was placed in an oil

bath preheated to 1050 and allowed to remain for 2 hr while

the temperature was raised to 1600. Upon cooling, the mixture

was filtered. The residue was washed with ethyl ether and

the total filtrate concentrated on a rotary evaporator under

reduced pressure. The products were not isolated, only cur-

sorily examined by nmr spectroscopy. The materials, 2-(2,4,6-

cycloheptatrien-l-yl)naphthalene and 2-(2,4,6-cycloheptatrien-

4-yl)naphthalene, were readily identified by their character-

istic spectra and determined to be present in a ratio of

1.0:1.6, respectively, as shown by integration.


General Procedure for the Photolysis of Tosylhydrazone

Sodium Salts in the 2-Butenes.--The photolyses were carried

out in an apparatus which consisted of a Pyrex tube (20 cm x

2.5 cm) with two side arms and equipped with a small mechanical







stirrer. One of the side arms was connected to a Dry Ice

condenser with a drying tube and the other to a stopcock.

The desired amount of tosylhydrazone salt was introduced into

the apparatus which was consequently flushed with dry nitro-

gen. The apparatus was then placed in a small Dry Ice-acetone

bath and the 2-butene condensed into the vessel. The vessel

was then mounted vertically alongside, and as close as pos-

sible to,a 550 watt Hanovia medium pressure mercury lamp in

a Pyrex cooling jacket. Both the apparatus and the lamp

system were placed in a large Dewar flask containing methyl

alcohol, with the temperature of the entire system being

regulated by the cooling jacket of the lamp which in turn was

cooled by a Dry Ice-acetone heat exchanger. The mixture was

then rapidly stirred and irradiated for the desired amount of

time at temperatures in the range of -15 to -30.


Photolysis of the Sodium Salt of B-Naphthaldehyde Tosyl-

hydrazone in trans-2-Butene.--Into the previously described

photolysis apparatus was placed 0.20 g (0.58 mmol) of the

sodium salt of 8-naphthaldehyde tosylhydrazone and 30 ml of

trans-2-butene (C.P. Grade, typical analysis 99.3 mole %

trans-2-butene, 0.2 mole % butene, and 0.5 mole % cis-2-

butene). The reaction mixture was irradiated for 2.0 hr at

temperatures between -15 to -200. At the end of this time,

the butene was allowed to evaporate and the residue was

dissolved in ethyl ether, filtered, and concentrated on a

rotary evaporator under reduced pressure to yield 88 mg of a








light yellow oil. In addition to a small amount of unknown

products, the material contained three products, l-(2-naphthyl)-

trans-2,3-dimethylcyclopropane, and syn- and anti-l-(2-

naphthyl)-cis-2,3-dimethylcyclopropane, as shown by vpc

analysis. The trans-cyclopropane was analyzed (column A) and

purified by prep vpc (column A).

1-(2-naphthyl)-trans-2,3-dimethylcyclopropane showed:

ir (film) 3050, 3000, 2940, 2920, 2860, 1628, 1600, 1500,
-l
1450, 1058, 850, 815, 745 cm-1; nmr (CCl4) T 2.10-2.95

multiplee, 7H, naphthyl protons), 8.15 multiplee, 1H, proton

on C-l), 8.35-9.65 multiplee, 8H, protons on C-2 and C-3

and methyl protons); mass spectrum m/e 196 (M+).

Anal. Calcd. for C15H16: C, 91.78; H, 8.22. Found:

C, 91.72; H, 8.31.


Photolysis of Benzaldehyde Tosylhydrazone Sodium Salt in

trans-2-Butene.--In the described photolysis apparatus was

placed 0.40 g (1.35 mmol) of the sodium salt of benzaldehyde

tosylhydrazone,38 then 25 ml of trans-2-butene (C.P. Grade)

was condensed into the apparatus. The reaction mixture was

irradiated for 2.5 hr at temp between -150 and -200. At the

end of this time, the butene was allowed to evaporate and the

residue was dissolved in ethyl ether, filtered, and concen-

trated on a rotary evaporator under reduced pressure to yield

0.09 g (45%) of crude 1-phenyl-trans-2,3-dimethylcyclopropane.

The cyclopropane was purified by preparative vpc (column A).

The nmr (CDClI) T 2.93 (broad singlet, 5H, phenyl protons),







8.30 multiplee, 1H, proton on C-1), 8.68-9.36 multiplee,

8H, protons on C-2 and C-3 and methyl protons) agreed with

the data of Closs.43


Photolysis of the Sodium Salt of 6-Naphthaldehyde Tosyl-

hydrazone in cis-2-Butene.--The photolysis of the sodium salt

of 8-naphthaldehyde tosylhydrazone in cis-2-butene was carried

out in a manner similar to that of trans-2-butene. In this

case, 0.29 g (0.84 mmol) of the sodium salt was decomposed in

30 ml cis-2-butene (C.P. Grade, typical analysis 99.5 mole %

cis- and 0.5 mole % trans-2-butene) and the photolysis was

done over a period of 2.75 hr at a temp between -200 and -300.

There was produced 0.126 g of a light yellow oil. Again the

products were analyzed by vpc and the same three products

were present (relative yields): l-(2-naphthyl)-trans-2,3-

dimethylcyclopropane (3%), syn- and anti-l-(2-naphthyl)-cis-

2,3-dimethylcyclopropane (97%). The syn- and anti-cyclopro-

panes were purified by preparative vpc (column A).

syn-l-(2-naphthyl)-cis-2,3-dimethylcyclopropane showed:

ir (film) 3050, 3000, 2940, 2870, 1630, 1602, 1500, 1385,

1160, 890, 860, 818, 748 cm-1; nmr (CC1I) r 2.20-3.00 (multi-

plet, 7H, naphthyl protons), 7.90 multiplee, 1H, proton on

C-l), 8.50-9.50 multiplee, 8H, protons on C-2 and C-3 and

methyl protons); mass spectrum m/e 196 (M+).

Anal. Calcd. for C1sH16: C, 91.78; H, 8.22. Found:

C, 91.55; H, 8.15.







anti-l-(2-naphthyl)-cis-2,3-dimethylcyclopropane showed:

ir (film) 3050, 3000, 2940, 2860, 1628, 1600, 1500, 1082,

850, 815, 782, 745 cm-1; nmr (CC14) T 2.20-3.10 multiplee,

7H, naphthyl protons), 8.50-9.15 multiplee, 9H, protons on

C-1, 2, and 3 and methyl protons); mass spectrum m/e 196 (M+).

Anal. Calcd. for C1sH16: C, 91.78; H, 8.22. Found:

C, 91.63; H, 8.29.


Pyrolysis of B-Naphthaldehyde Tosylhydrazone Sodium Salt

in cis-2-Butene.--In a 0.5 oz reusable Fischer-Porter Carius

Tube was placed 0.10 g (0.29 mmol) of the sodium salt of

B-naphthaldehyde tosylhydrazone. The tube was fitted with a

Dry Ice condenser by means of a rubber stopper, flushed with

dry nitrogen, and placed in a Dry Ice-acetone bath. Then 10

ml of trans-2-butene was condensed into the Carius Tube; the

condenser was quickly removed and the tube sealed. The tube

was then immersed in an oil bath (preheated to 1050) for 1 hr.

At the end of this time the tube was allowed to cool to room

temperature and then cooled further in a Dry Ice-acetone

bath. The tube was opened, the bath removed, and the butene

allowed to evaporate. The residue was dissolved in ethyl

ether, concentrated on a rotary evaporator at reduced pres-

sure, and analyzed by vpc (column C) to yield the following

data (relative yields): l-(2-naphthyl)-trans-2,3-dimethyl-

cyclopropane (-10%),* and syn- and anti-l-(2-naphthyl)-cis-

2,3-dimethylcyclopropane (-90%).


A large amount of interfering material makes this data

only approximate.







Photolysis of 4,5-Benzotropone Tosylhydrazone Sodium

Salt in cis-2-Butene.--The photolysis of the sodium salt of

4,5-benzotropone tosylhydrazone was effected using 0.29 g

(0.84 mmol) of the salt in 30 ml of cis-2-butene (C.P. Grade)

and by irradiating for 2.3 hr at -250. Work up yielded 0.061

g of product and vpc analysis (column C) gave the following

data: 1-(2-naphthyl)-trans-2,3-dimethylcyclopropane was a

minor product (4%); the major products were syn-l-(2-naphthyl)

cis-2,3-dimethylcyclopropane and anti-l-(2-naphthyl)-cis-2,3-

dimethylcyclopropane (96%).


Photolysis of 4,5-Benzotropone Tosylhydrazone Sodium

Salt in trans-2-Butene.--The photolysis of the sodium salt of

4,5-benzotropone tosylhydrazone was effected as previously

described using 0.20 g (0.58 mmol) of the salt in 30 ml of

trans-2-butene (C.P. Grade) and by irradiating for 2.3 hr at

-250. Work up yielded 0.046 g of a yellow oil. Vpc analysis

(column C) gave the following data: 1-(2-naphthyl)-trans-

2,3-dimethylcyclopropane was the major product along with a

small amount of the epimeric l-(2-naphthyl)-cis-2,3-dimethyl-

cyclopropanes.


Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosyl-

hydrazone in cis-2-Butene.--In a 0.5 oz reusable Fischer-

Porter Carius Tube was placed 0.10 g (0.29 mmol) of the

sodium salt of 4,S-benzotropone tosylhydrazone. The tube was

fitted with a Dry Ice condenser by means of a rubber stopper,

flushed with dry nitrogen, and placed in a Dry Ice-acetone







bath. Then, 10 ml of cis-2-butene (research grade) was con-

densed into the Carius Tube; the condenser was quickly removed

and the tube sealed. The tube was allowed to warm to room

temperature and then immersed in an oil bath (preheated to

1050) for 3.5 hr. At the end of this time the tube was

allowed to cool to room temperature and then cooled further

in a Dry Ice-acetone bath. The tube was opened, the bath

removed, and the butene allowed to evaporate. The residue

was dissolved in ethyl ether, concentrated on a rotary evapo-

rator under reduced pressure to yield 7 mg of a yellow oil

which was analyzed by vpc (column C). The reaction material

contained the syn- and anti-l-(2-naphthyl)-cis-2,3-dimethyl-

cyclopropanes (93%), l-(2-naphthyl)-trans-2,3-dimethylcyclo-

propane (7%), and some extraneous material. In all cases

following the percentages given in parentheses do not imply

total percent yield but relative yields of the cis- and trans-

cyclopropanes and were done on column C unless otherwise

stated.


Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosyl-

hydrazone in cis-2-Butene and Benzene.--The reaction was

carried out as in the above experiment except that 0.08 g

(0.23 mmol) of the sodium salt, 7.0 ml (79 mmol) of benzene,

and 2.5 ml (28 mmol) of research grade cis-2-butene were

heated at 1050 for 3 hr. The work up was the same yielding

10 mg of a yellow oil which contained the cyclopropanes in

the following relative yield: syn- and anti-l-(2-naphthyl)-

cis-2,3-dimethylcyclopropanes (94%) and 1-(2-naphthyl)-trans-







2,3-dimethylcyclopropane (6%). The 2-naphthylcarbene-benzene

adduct was also produced as evidenced by vpc.


Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosyl-

hydrazone in cis-2-Butene and Perfluorocyclobutane.--Again,

the reaction was carried out in a manner analogous to that of

the cis-2-butene reaction. In this series two reactions were

carried out: the first with cis-2-butene and perfluorocyclo-

butane in a 1:2 molar ratio, respectively, and the second

with a 1:6 molar ratio, respectively.

In the first reaction, 0.10 g (0.29 mmol) of the sodium

salt of benzotropone tosylhydrazone, 8.0 ml (60 mmol) of per-

fluorocyclobutane, and 2.5 ml (30 mmol) of research grade cis-

2-butene were heated at 1050 for 5 hr. The work up yielded

7.5 mg of a yellow oil which, when analyzed by vpc, was shown

to contain the syn- and anti-l-(2-naphthyl)-cis-2,3-dimethyl-

cyclopropanes (94%) and 1-(2-naphthyl)-trans-2,3-dimethyl-

cyclopropane (6%).

In the second case, 0.10 g (0.29 mmol) of the sodium

salt of benzotropone tosylhydrazone, 9.0 ml (67 mmol) of

perfluorocyclobutane, and 1.0 ml (11 mmol) of research grade

cis-2-butene were heated at 1080 for 4 hr. The work up

yielded 7.0 mg of a yellow oil which gave the same cis to

trans cyclopropane ratio as the first experiment (namely,

cis 94% and trans 6%).







Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosyl-

hydrazone in cis-2-Butene and Oxygen.--This reaction, too,

was carried out like the cis-2-butene reaction with the fol-

lowing exceptions: 0.10 g (0.29 mmol) of the sodium salt of

benzotropone tosylhydrazone and 12 ml of research grade cis-

2-butene were placed in the Carius Tube and then oxygen was

bubbled through the mixture for 1 hr. The reaction mixture

was heated at 1080 for 3.6 hr and the work up yielded 21 mg

of a yellow oil which contained the two epimeric cis-cyclo-

propanes (96%) and the trans-cyclopropane (4%).

A control experiment, done exactly the same as the above

experiment only bubbling nitrogen through the mixture, gave

the typical 94% cis- material and 6% trans-.


Pyrolysis of the Sodium Salt of 4,S-Benzotropone Tosyl-

hydrazone in cis-2-Butene and trans,trans-2,4-hexadiene.--

Again, the reaction was carried out in a manner analogous to

that of the cis-2-butene reaction. In this series two sepa-

rate reactions were carried out: the first with cis-2-butene

and trans,trans-2,4-hexadiene in a 1.0:1.5 molar ratio, respec-

tively, and the second with a 1.0:2.1 molar ratio, respectively.

In the first reaction, 0.09 g (0.26 mmol) of the sodium

salt of benzotropone tosylhydrazone, 7.0 ml (60 mmol) of

trans,trans-2,4-hexadiene, and 3.6 ml (40 mmol) of research

grade cis-2-butene were heated at 1050 for 4.5 hr. Work up

yielded 24 mg of product (again, a yellow oil) which had the

cis-cyclopropanes produced in 99% yield relative to the trans-







cyclopropane in 1%. However, what is assumed to be the

products from the diene account for ca. 90% of the total

products.

In the second case, 0.10 g (0.29 mmol) of the sodium

salt of benzotropone tosylhydrazone, 7.3 ml (63 mmol) of

trans,trans-2,4-hexadiene, and 2.7 ml (30 mmol) of research

grade cis-2-butene were heated at 1050 for 5 hr. Work up

yielded 18 mg of a yellow oil which proved to be almost

exclusively the assumed diene products (there was too little

of the butene adducts to provide reliable results).


Sensitized Photolysis of the Sodium Salt of 4,5-Benzo-

tropone Tosylhydrazone in Benzene.--In the photolysis vessel,

described previously, was placed 1.0 g (5.1 mmol) xanthen-9-

one, 0.10 g (0.29 mmol) of the sodium salt of benzotropone

tosylhydrazone, and 40 ml of benzene. The stirred mixture

was photolyzed at 300 for I hr. The solvent was evaporated

and the residue was chromatographed on grade III Woelm basic

alumina. There were no hydrocarbon products isolated from

the reaction mixture.


Thermal Stability of the syn- and anti-l-(2-naphthyl)-

cis-2,3-dimethylcyclopropanes.--A sample of the cis-cyclopro-

panes was preparatively chromatographed (column A at 2300)

and then analyzed for any isomerization during chromatography.

There was no trace of trans-cyclopropane as evidenced by vpc

analysis (column B).







Stability of cis-2-Butene under Reaction Conditions.--As

in a typical run, 0.10 g (0.29 mmol) of the sodium salt of

benzotropone tosylhydrazone and 9.0 ml cis-2-butene (research

grade) were heated in a Carius Tube at 1050 for 4.25 hr.

Analysis of the cis-2-butene (column C) after the usual

reaction period indicated that less than 1 percent isomeri-

zation had occurred during the reaction.


2-Methyl-4,5-Benzotropone.--The ketone was prepared by

the method of Heilbronner and had a mp 68-690. Lit. 700.52


2-Methyl-4,5-Benzotropone Tosylhydrazone.--In a typical

run, 2.00 g (0.012 mol) of 2-methyl-4,5-benzotropone, 2.18 g

(0.012 mol) of p-toluenesulfonylhydrazide, and 5 drops of

coned sulfuric acid were placed in 40 ml of abs ethyl alcohol

and heated at reflux for 1 hr. Upon cooling, 2-methyl-4,5-

benzotropone tosylhydrazone precipitated from solution and

was isolated by filtration. Recrystallization from abs ethyl

alcohol yielded 2.4 g (60%) of the product as yellow-gold

plates mp 182-1840: ir (KBr) 3180, 1630, 1590, 1555, 1420,

1385, 1330, 1165, 1022, 915, 805, 760, 680, and 580 cm-1;

nmr (CDC13) T 2.00-3.00 (A2B2, 4H, J=8, aromatic), 2.78 (s,

4H, benzo), 3.05-3.70 (m, 3H, cycloheptatrienyl), 7.62 (s,

3H, p-methyl), 7.88 (d, 3H, cycloheptatrienyl methyl); mass

spectrum 339 (M+).

Anal. Calcd. for C19iHiN202S: C, 67.43; H, 5.36; N,

8.28. Found: C, 67.39; H, 5.33; N, 8.40.







Pyrolysis of the Sodium Salt of 2-Methyl-4,5-Benzotropone

Tosylhydrazone in Benzene.--To a Fischer-Porter Aerosol Com-

patibility Tube containing 35 ml of benzene was added 216 mg

(0.60 mmol) of the sodium salt of 2-methyl-4,5-benzotropone

tosylhydrazone. The sealed tube was placed in an oil bath

preheated to 1300 for 20 min. By the end of this time the

yellow mixture had changed to a very light yellow-green one.

The mixture was allowed to cool and was filtered. The residue

was washed with ethyl ether and the total filtrate concen-

trated on a rotary evaporator under reduced pressure to yield

80 mg of products. Analysis of the product mixture by nmr

indicated the presence of 2-vinylnaphthalene and 2-(2,4,6-

cycloheptatrien-l-yl)-3-methylnaphthalene (comparisons were

made using commercially available 2-vinylnaphthalene from

Aldrich Chemical Co. and 2-(2,4,6-cycloheptatrien-l-yl)-3-

methylnaphthalene from the pyrolysis of the sodium salt of

the tosylhydrazone of 3-methyl-2-naphthaldehyde in benzene).

Analysis by vpc (column B) indicated three products in the

reaction mixture (the known ones were established by compari-

son with known samples): 1) retention time 1 min, 2-vinyl-

naphthalene (72%); 2) retention time 10 min, unknown compound

(6%); 3) retention time 14 min, 2-(2,4,6-cycloheptatrien-l-

yl)-3-methylnaphthalene (22%). Products 1 and 3 were isolated

in 44% and 20% overall yield, respectively. 2-(2,4,6-cyclo-

heptatrien-l-yl)-3-methylnaphthalene sublimedd at 900 and

0.1 mm Hg) had the following properties: mp 65-680; ir (KBr)

3055, 3010, 1600, 1500, 1440, 1220, 1010, 890, 745, 700, and








480 cm-1; nmr (CDC13) T 2.00-2.90 (m, 6H, naphthyl), 3.25

(broad t, 2H, 4 and S cycloheptatrienyl), 3.50-3.90 (complex

pattern, 2H, 3 and 6 cycloheptatrienyl), 4.35-4.76 (complex

pattern, 2H, 2 and 7 cycloheptatrienyl), 6.95 (broad t, IH,

methine), 7.65 (s, 3H, methyl); mass spectrum 232 (M+).

Anal. Calcd. for Cs1H16: C, 93.06; H, 6.94. Found:

C, 92.85; H, 7.00.

The sodium salt of 3-methyl-2-naphthaldehyde tosylhydra-

zone was pyrolyzed in the same manner as the above salt

except that 40 ml of benzene was used and the temperature

was 1450. The products were aldazine (42%) and 2-(2,4,6-

cycloheptatrien-l-yl)-3-methylnaphthalene (39%).


2,3-Benzotropone.--2,3-Benzotropone was prepared by the

method of Collington and Jones,53 and the ir spectrum of the

product was in complete agreement with that reported in the

literature.


2,3-Benzotropone Tosylhydrazone.--To 1.0 g (0.6 mol) of

2,3-benzotropone in 10 ml of 95% ethyl alcohol was added

1.1 g (0.6 mol) of g-toluenesulfonylhydrazide in 25 ml of 95%

ethyl alcohol and one drop of acetic acid. The solution was

heated at reflux for 75 min, and then the reaction mixture

was allowed to cool. The solvent was removed on a rotary

evaporator under reduced pressure and the resinous residue

was chromatographed on silica gel eluting with methylene

chloride. The thus purified tosylhydrazone, still a viscous

resin, was obtained 54% yield and had the following properties:







ir (film) 3210, 1582, 1400, 1385, 1340, 1320, 1160, 1080,

1020, 810, 800, 760, 700, 680, 655 cm-1; nmr (CDCl3) T 1.60

(broad s, 1H, proton on nitrogen), 2.00-4.00 multiplee, 12H,

protons on benzo ring, cycloheptatriene ring, and the aromatic

toluyl protons whose A2B2 quarter could be picked out of the

multiple, 7.70 (s, 3H, methyl).

The tosylhydrazone was never obtained in a state of

purity which gave a good analysis (e.g. Calcd. C, 65.38;

H, 5.16. Found: C, 65.10; H, 5.84). Mass spectrum had no

parent peak, (none of any significance above m/e 156 even at

10 eV) 156 (100), 141 (9), 91 (17), 64 (28).

The resinous tosylhydrazone was converted to its sodium

salt cleanly by the outlined procedure (this was the only

water-sensitive salt made).


Pyrolysis of the Sodium Salt of 2,3-Benzotropone Tosyl-

hydrazone in Benzene.--In a 3 oz Fischer-Porter Aerosol

Compatibility Tube containing 35 ml of benzene (deoxygenated

by bubbling in nitrogen for several min) was placed 0.28 g

(0.81 mmol) of the sodium salt of 2,3-benzotropone tosylhydra-

zone. The tube was then placed in an oil bath which was pre-

heated to 1200 and allowed to remain there for 35 min. The

solution was allowed to cool, filtered, and concentrated to

a yellow oil on a rotary evaporator under reduced pressure.

Chromatography of the residue on activity grade III basic

Woelm alumina yielded 0.045 g (26%) of a colorless oil which

was identified as l-(2,4,6-cycloheptatrien-l-yl)naphthalene.




86


The properties of the compound are as follows: ir (film)
-l
3050, 3020, 1600, 1510, 1395, 800, 780, 745, 705 cm ; nmr

(CDCla) x 2.20 multiplee, 3H, a-protons of naphthyl system),

2.55 multiplee, 4H, B-protons of naphthyl system), 3.30

(triplet, 2H, J=3 Hz, 4 and 5 protons on cycloheptatriene),

3.70 multiplee, 2H, 3 and 6 protons on cycloheptatriene),

4.50 (quartet, 2H, J=5.5 Hz, 2 and 7 protons on cyclohepta-

triene), and 6.60 multiplee, 1H, 7 proton on cyclohepta-

triene); mass spectrum m/e 218 (M+), 127 (CloH07), 91 (C7H7).

Anal. Calcd. for C17Hil: C, 93.54; H, 6.46. Found:

C, 93.32; H, 6.61.












LIST OF REFERENCES


1. L. Skattebol, Tetrahedron, 23, 1107 (1967).

2. P. Ashkenazi, S. Lupan, A. Schwarz, and M. Cais, Tetra-
hedron Lett., 817 (1969).

3. G. Frater and 0. P. Strausz, J. Amer. Chem. Soc., 92,
6654 (1970); D. E. Thornton, R. K. Gosavi, and 0. P.
Strausz, ibid., 92, 1768 (1970); I. G. Csizmadia,
J. Font, and 0. P. Strausz, ibid., 90, 7360 (1968).

4. R. L. Russell and F. S. Rowland, ibid., 92, 7508 (1970).

5. G. G. Vander Stouw, Dies. Abst., 25, 6974 (1965) and
Chem. Abet., 63, 13126b (1965) undeer the direction of
H. Shechter.

6. G. G. Vander Stouw, A. R. Kraska, and H. Shechter, J.
Amer. Chem. Soc., 94, 1655 (1972).

7. R. C. Joines, A. B. Turner, and W. M. Jones, ibid., 91,
7754 (1969).

8. W. M. Jones and C. L. Ennis, ibid., 91, 6391 (1969).

9. J. A. Myers, R. C. Joines, and W. M. Jones, ibid., 92,
4740 (1970).

10. C. Wentrup and K. Wilczek, Helv. Chim. Acta, 53, 1459
(1970).

11. P. O. Schissel, M. E. Kent, D. J. McAdoo, and E. Hedaya,
J. Amer. Chem. Soc., 92, 2147 (1970).

12. E. Hedaya and M. E. Kent, ibid., 93, 3285 (1971).

13. W. J. Baron, M. Jones, Jr., and P. P. Gaspar, ibid., 92,
4739 (1970).

14. T. Mitsuhashi and W. M. Jones, ibid., 94, 677 (1972).

15. P. H. Gebert, Ph.D. Dissertation, University of Florida,
Gainesville, Florida, 1972, under the direction of
W. M. Jones.




Full Text

PAGE 1

CARBENE-CARBENE REARRANGEMENTS: A STUDY IN THE BENZOCYCLOHEPTATRIENYLIDENE-NAPHTHYLCARBENE SYSTEM By KENNETH EDWARD KRAJCA A DISSERTATION PRESENTED TO THE GRADUATE COUNCIL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF DOCTOR OF PHILOSOPHY UNIVERSITY OF FLORIDA 1972

PAGE 2

3 1262 08552 4709

PAGE 3

DEDICATION This dissertation is dedicated to the author's wife. Sherry Lee, vsfhose work, understanding, and love made this work possible and to his daughter, Laura, who provided him with much happiness.

PAGE 4

ACKNOWLEDGMENTS The author is indebted to Dr. W. M. Jones for his professional guidance, patience, and personal friendship during the course of this work. His advice and encouragement will be of lasting benefit. The counsel and discussion of the author's fellow associates are highly regarded. The author wishes to thank his parents for their guidance and help. Finally, the author wishes to thank the Petroleum Research Fund and the National Science Foundation for providing financial assistance during this period of study.

PAGE 5

TABLE OF CONTENTS Page ACKNOWLEDGMENTS iii LIST OF TABLES ix LIST OF FIGURES x ABSTRACT xi CHAPTER I . INTRODUCTION AND BACKGROUND 1 II. RESULTS 10 III. DISCUSSION 31 IV . EXPERIMENTAL 5 7 General 5 7 Pyrolysis of the Sodium Salt of Benzaldehyde Tosylliydrazone in Diglyme 58 Pyrolysis of tlie Sodium Salt of Tropone Tosylhhyrazone in Diglyme 59 4 ,5-Benzotropone 59 4, 5-Benzotropone Tosylhydrazone 59 B-Naphthaldehyde Tosylhydrazone 60 General Preparation of the Sodium Salt of Tosylhydrazones 60 Pyrolysis of the Sodium Salt of 4, 5-Benzotropone Tosylhydrazone in Diglyme 61 iv

PAGE 6

Page Photolysis of the Sodium Salt o£ 4,5-Benzotropone Tosylhydrazone in Diglyme 62 Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosylhydrazone in Benzene 63 Photolysis of the Sodium Salt of 4,5-Benzotropone Tosylhydrazone in Benzene 64 Preparation of 2-naphthaldazine 64 Pyrolysis of the Sodium Salt of 3-Naphthaldehyde Tosylhydrazone in Cyclohexene. 64 Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosylhydrazone in Cyclohexene at 83° 66 Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosylhydrazone in Cyclohexene at 140° 67 Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosylhydrazone in CyclohexeneDiglyme at 82° 67 Photolysis of 4 ,5-Benzotropone Tosylhydrazone Sodium Salt in Cyclohexene 68 Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosylhydrazone in Cyclohexene. 69 Reaction of the Sodium Salt of 4,5-Benzotropone Tosylhydrazone with Dimethyl Fumarate at 105° 69 V

PAGE 7

Page Reaction of the Sodium Salt of 3-Naphthaldehyde Tosylhydrazone with Dimethyl Fumarate at 105° 70 Reaction of 6-Naphthaldehyde Tosylhydrazone Sodium Salt with Dimethyl Fumarate at 160° 71 Reaction of the Sodium Salt of 4,5-Benzotropone Tosylhydrazone with Dimethyl Fumarate at 160° 72 General Procedure for the Photolysis of Tosylhydrazone Sodium Salts in the 2-Butenes 72 Photolysis of the Sodium Salt of 3-Naphthaldehyde Tosylhydrazone in trans -2Butene 73 Photolysis of Benzaldehyde Tosylhydrazone Sodium Salt in trans 2-Butene 74 Photolysis of the Sodium Salt of 6-Naphthaldehyde Tosylhydrazone in cis -2But ene 75 Pyrolysis of 3-Naphthaldehyde Tosylhydrazone Sodium Salt in cis 2-Butene 76 Photolysis of 4 , 5-Benzotropone Tosylhydrazone Sodium Salt in cis 2-Butene 77 Photolysis of 4 , 5-Benzotropone Tosylhydrazone Sodium Salt in trans 2-Butene 77 vi

PAGE 8

Page Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosylhydrazone in cis 2-Eutene . 77 Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosylhydrazone in cis -2Butene and Benzene 78 Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosylhydrazone in cis -2-Butene and Perfluorocyclobutane 79 Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosylhydrazone in cis -2-Butene and Oxygen SO Pyrolysis of the Sodium Salt of 4,5-Benzotropone Tosylhydrazone in cis 2-Butene and trans , trans -2 ,4-hexadiene 80 Sensitized Photolysis of the Sodium Salt of 4,5-Benzotropone Tosylhydrazone in Benzene 81 Thermal Stability of the syn and anti -1(2-naphthyl) cis 2 , 5-dimethylcyclopropanes 81 Stability of cis -2-Butene under Reaction Conditions 82 2-Methyl-4 , 5-Benzotropone 82 2-Methyl-4,5-Benzotropone Tosylhydrazone... 82 Pyrolysis of the Sodium Salt of 2-Methyl4 , 5-Benzotropone Tosylhydrazone in Benzene 83 vii

PAGE 9

Page 2,3-Benzotropone 84 2 , 3-Bentotroponc Tosylhydrazone 84 Pyrolysis of the Sodium Salt of 2,3-Benzotropone Tosylhydrazone in Benzene 85 LIST OF REFERENCES 87 BIOGRAPHICAL SKETCH 92

PAGE 10

LIST OF TABLES Table Page 1. Stereochemical Results: The Percent "Wrong Isomer" from the Decompositions in cis -2Butene of the Carbene Precursors both Thermally at 1050 and Photolytically at -25° 27

PAGE 11

LIST OF FIGURES Figure Page 1. Vpc trace: Pyrolysis of 2_6 in Cyclohexene. . . . 16 2. 100 MHz Nmr of 33 . . 18 3. Decoupled Nmr of 33^ .19 4. Closure of vinylcarbene to cyclopropene 48 5. Boat Conformation of Benzocycloheptatrienylidene 50 6. Planar Conformation of Benzocycloheptatrienylidene 51

PAGE 12

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 CARBENE-CARBENE REARRANGEMENTS: A STUDY IN THE BENZOCYCLOHEPTATRIENYLIDENE-NAPHTHYLCARBENE SYSTEM By Kenneth Edward Krajca August, 1972 Chairman: W. M. Jones Major Department: Chemistry The rearrangement of benzocycloheptatrienylidenes to naphthylcarbenes in solution thermally at temperatures in the range of 100 and photolytically at temperatures as low as -35 is reported. The rearranged naphthylcarbenes were detected by their reaction products with solvent. Decomposition in benzene gave the corresponding naphthylcycloheptatrienes in good yield, and decompositions in cyclohexane and diglyme yielded the corresponding insertion products. In cyclohexene, the formation of naphthylnorcaranes and naphthylcarbene insertion products into cyclohexene was temperature and concentration dependent. This is consistent with a bimolecular reaction of an intermediate with cyclohexene in competition with a unimolecular rearrangement that gives 3-naphthylcarbene .

PAGE 13

The rearrangement of a carbene precursor was excluded and possible mechanisms were explored. Stereochemical studies indicate that the singlet state of the carbene is responsible for the rearrangement and that tlierc is a small amount of nonstereospecif ic addition which is unaffected by dilution but is affected by the addition of triplet scavengers. This is compatible with a singlettriple t equilibrium wherein the rate of triplet-singlet intersystem crossing is greater than the rates of addition of the triplet carbene to olefin, but the rate of triplet carbene addition to dienes (and oxygen) must be greater than the rate of tripletsinglet intersystem crossing. Methyl substitution has an effect upon the rearrangement which may occur via a fused, twisted cyclopropene intermediate (benzobicycloheptatriene) from the boat conformation of the carbene, the planar conformation, or via a concerted pathway.

PAGE 14

CHAPTER I INTRODUCTION AND BACKGROUND Carbenecarbene rearrangement has been established as a firm, although relatively rare, chapter in the chemistry of divalent intermediates. In 1967, Skattebol^ reported several examples of "non-trivial" carbene-carbene rearrangements, e.g that of vinylcyclopropylidene (1) to cyclopentenylidene (2_) . ^ ^ ^ > 1 2 In this type of rearrangement, however, the carbene carbon retains its integrity and, since the carbene was generated by a-elimination, a free carbene may not have been obtained. In fact, the stereoselectivity in product formation in one case suggested a carbenoid species. Cais et_ aJ • ^ reported a carbene-carbene rearrangement in the deprotonation of f errocenyltropylium fluoborate (3^) with the admonition that the carbene intermediacy was not established, a Fe ^^^''" @

PAGE 15

Along a slightly different line, Strausz^ and coworkers investigated the rearrangement of the carbonylcarbene derived from the photolysis of 5-diazobutanone2 and homologous compounds in both the gas and condensed phases. Russell and Rowland"* also investigated similar systems but came to slightly different conclusions. In 1965, Shecliter and Vander Stouw^ reported the rearrangement of various methyland dimethylphenylcarbenes and elucidated this work in a more recent report.^ In their work, (2-methylphenyl) diazome thane (4^), for example, was pyrolyzed to yield, among other products, styrene (5), and benzocycloCHN; CHa -:^ Qp 4 '5 6 butene (6^). The above products and a labeling study indicated the process was occurring via a multiple carbene-carbene rearrangement. Jones et al . ^ later elaborated on the phenylcarbene-cycloheptatrienylidene rearrangement when heptafulvalene (7^) was isolated from the gas phase pyrolysis at 250° of the sodium salt of benzaldehyde tosylhydrazone as well as from pyrolysis of the sodium salt of tropone tosylhydrazone.® Pyrolysis of J ->

PAGE 16

a mixture of the sodium salts of benzaldehyde tosy Ihydrazone and 2." tolualdehyde tosylhydrazone resulted in the formation of crossed dimer which proved that the rearrangement occurred in the gas phase. Multiple carbene-carbene rearrangements of diphenylmethylene (8^) were also studied by Jones. ^ Pyrolysis at 350 of diphenyldiazomethane yielded fluorene (9^) , tetraphenylethylene (10^) , triphenylheptafulvene (Ij^) , diphenylheptafulvalene (12), and the corresponding azine. Note that these 10 11 results require that the diarylcarbene expand to form the arylcycloheptatrienylidene 1_3 and the arylcycloheptatrienyli dene contract to form a biphenylmethylene which is the immediate precursor of the fluorene.

PAGE 17

Kcntrup and Kilczek^° also reported the reversibility of the phen)'lcarbene-cycloheptatrienylidene rearrangement when they isolated stilbene and heptaf ulvalene from the independent gas phase pyrolysis at 300° of the sodium salts of both benzaldehyde tosylhydrazone and tropone tosy Ihydrazone . Hedaya and coworkers ^^ studied the ring contraction and expansion of phenylcarbene in the gas phase by pyrolysis of benzaldehyde tosylhydrazone sodium salt. They also observed the formation of heptaf ulvalene as one of the reaction products and by incorporating a labeling study on tolylcarbene ^ ^ they decided that very small energy barriers separate phenylcarbene, cycloheptatrienylidene , and possible bicycloheptatriene intermediates. Lastly, M. Jones, Jr.^^ investigated the interconversion of o, m, and £tolylcarbene by observing the ratios of the styrene (5_) and benzocyclobutene (6^) produced in the pyrolysis of the corresponding diazo compounds. CHN2 CH3 CHN; CH3 0^-^

PAGE 18

It is very interesting to note that most of the researchers on the phenylcarbene-cycloheptatrienylidene interconversion believe that the rearrangement occurs via a 5-79-13 bicycloheptatriene intermediate (14) . Jones and <> <> (i ^ Mitsuhashi^ * elucidated this mechanism with their study on the importance o£ bond order in the gas phase carbene-carbene rearrangement of a series of substituted naphthylcarbenes . It was found that scrambling of the carbene occurs only between positions having the highest bond order which augurs well for the bicyclic intermediate as opposed to a Wolff-type rearrangement (shown below) . FI\, ^ ^
PAGE 19

rearranged dimers and the mechanism is assumed to be similar to that mentioned above. -> Nitrogen analogues to the phenylcarbene -cycloheptatrienylidene rearrangement have been known for a long time. The ring expansion of pheny Initrene to yield substituted azepines was studied as early as 1958^^>^'' and has recently received thorough investigation.^®'^'* Pyrolysis^^ and photolysis^"* of phenyl azide (1_6) in the presence of amines led to 2-amino-3H-azepine derivatives (1_7) as final products EtzNH -> a 17 The thermal^^ and photolytic^ ^ ' ^ ^ deoxygenation of nitrobenzene, as well as the thermal deoxygenation of nitrosobenzene,^® in the presence of amines also led to 2-amino-3Hazepines. As in the analogous carbene-carbene rearrangements, these reactions can be envisioned as proceeding through a bicyclic intermediate which can yield an azacycloheptatrienyli dene. From the experimental results, however, v/orkers have not been able to determine whether or not the azacyclohepta-

PAGE 20

trienylidene is an intermediate. In fact, another mechanism, addition of the amine to the bicyclic azirine (1_8) , has received much more attention. ^^ Crow and Wentrup^^ reported that the thermal generation of 2-pyridylphenylcarbene ( 18a ) led to carbazole (19) products which were believed to arise via a substituted azacyclohepta-

PAGE 21

trienylidcne (20) intermediate or transition state. Crow and Kentrup have extended their vv'ork on these systems^ ^ as well as investigated the pyridylnitrene-diazacycloheptatrienylidene rearrangement.^^ Perhaps the most interesting was the work on 2-quinolylnitrene (21) , an annelated nitrene which is believed to rearrange via an annelated diazacycloheptatrienylidene (2^).^^ However, naphthyl azides have not been known to give rise to detectable azepine.^° The work described in this dissertation was done to investigate the effects of annelation on the phenylcarbenecycloheptatrienylidene rearrangement. It was hoped that the effects of annelation would give insight on the nucleophilicity of the aromatic carbene, on the relative rate of rearrangement and on the mechanism of the rearrangement. The approach was very feasible in that the precursors for both the annelated arylcarbene and the annelated aromatic carbenes were avail-

PAGE 22

able. The systems chosen for study were 4 , 5-benzocycloheptc trienylidene (25) , 2-methyl-4 , 5-benzocycloheptatrienylidene (24), 2 ,3-benzocycloheptatrienylidene (25^), and the related 2^ 24 25_ naphthylcarbenes . While other benzocycloheptatrienylidenes have been prepared and studied, they did not rearrange and many o£ their reactions were typical of diary Icarbenes . ^ ^

PAGE 23

CHAPTER II RESULTS The first results needed were for the model system of pheiiylcarbene and cycloheptatrienylidene . Thus, the pyrolysis of the sodium salt of benzaldehyde tosylhydrazone was effected in diglyme at 165 to produce what is believed to be the phenylcarbene-diglyme insertion products. This residue, a light yellow oil, had an nmr spectrum very similar to that of the naphtliylcarbene-diglyme insertion products ( vide infra ) ; there was also benzaldazine produced in the reaction. The nmr spectrum of the reaction material indicated no heptafulvalene was present, and comparative tic of the material and a known sample of heptafulvalene was carried out using silica gel eluting with pentane. This confirmed the absence of heptafulvalene in the reaction material. Tropone tosylhydrazone sodium salt was also decomposed in diglyme at a temperature of 210°. The nmr spectrum of the residue, a dark solid, vvras identical to that of heptafulvalene with an additional small amount of aromatic resonance at 2.8 x but none at 6.2-7.0 t (the position expected for the resonances of any diglyme insertion products) . The aromatic carbene, 4 , 5-benzocycloheptatrienylidene (23) , was generated by the thermal and photolytic decomposi10

PAGE 24

11 tion o£ the sodium salt of 4 , 5-benzotropone tosylhydrazone (26) in an appropriate solvent. Pyrolysis or photolysis of the sodium salts o£ tosylhydrazones form the corresponding diazo compounds in situ . ^ ^ The diazo compounds then decompose under aprotic reaction conditions to yield the carbenes. The synthetic route to the sodium salt of 4, 5-benzotropone tosylhydrazone is shown in the scheme below. The COzEt ( .^=^>r-^^ 26 ketone, 4 , 5-benzotropone (27) , was prepared by the method of Thiele and Weitz^^ and the tosylhydrazone by a modification of the procedure of Gloss. ^'* The conversion of the tosylhydrazone to its sodium salt was done with sodium hydride in tetrahydrofuran. ^ ^ The last step was carried out in a dry box; however, contrary to some reports the sodium salts did not appear to be hygroscopic in most cases. In a similar manner, commercially available g-naphthaldehyde was converted to its tosylhydrazone^^ and then to its

PAGE 25

12 tosylhydrazone sodium salt. This salt was the immediate precursor of 6-naphthylcarbene . In most cases, the naphthaldehyde tosylhydrazone sodium salt was decomposed under the same or very similar conditions as those of the 4 ,5-benzotropone tosylhydrazone sodium salt so tliat tlie chemistry of the two carbenes could be compared and, wliere possible, as a check on the authenticity of the naplithyl products formed from the rearranged carbene. Thus 4 , 5-benzotropone tosylhydrazone sodium salt was decomposed in diglyme at 100 to give a 29% yield of the isomeric 3-naphthylcarbene-diglyme insertion products (28). H3CN While these isomeric compounds could not be readily separated, there is little doubt as to their identity. The nmr spectrum of these products shows aromatic proton absorptions at 2.002.80 T and methyl and methylene protons adjacent to oxygen at 6.10-7.20 T (there are more signals than necessary in the region of 6.60 t for methyl protons of a single product). Other spectral properties and the elemental analysis of the mixture are in complete agreement with the assigned structures and regardless of the carbene precursor ( vide infra ) , the products formed in diglyme are the same.

PAGE 26

13 The decomposition o£ naphthaldehyde tosylhydrazone sodium salt in diglyme at 100 produced a 13% yield of the same 3-naphthylcarbene-diglyme insertion products produced in the above decomposition. The product yield from the decomposition of the naphthaldehyde tosylhydrazone salt was typically lower because there was competition of aldazine formation. Next, the photolytic decomposition of the salt of 4,5benzotropone tosylhydrazone in diglyme was carried out at 30 and resulted in the formation of the familiar naphthylcarbenediglyme insertion products. When 4 , 5-benzotropone tosylhydrazone sodium salt was heated in benzene at 80° there was formed, in 96% yield, a product identified as 2(2 ,4 ,6-cycloheptatrien-l-yl)naphthalene (29) . Although this compound has been prepared before, there 29 is no complete report of its physical and spectral properties The existence of the cycloheptatrienyl moiety was proven by comparison of the above compound's nmr spectrum with that of l-phenyl-2,4 ,6-cycloheptatriene^^ (the spectra were identical except for the aromatic region). That it was the g-naphthyl function (as opposed to a-naphthyl) was shown by careful

PAGE 27

14 examination of the compound's ir spectrum; the absorption for the out-of-plane bending at 861 cm indicated an isolated H on the naphthyl ring system, that at 821 cm indicated two adjacent H, and that at 745 cm was indicative of four adjacent H^^ (the nmr of the product showed a ratio of 4:3 for the a to B protons, respectively, in the aromatic region). Photolysis at 30 of the sodium salt of 4 , 5 -benzotropone tosy Ihydrazone also produced the adduct 29_ in 59-0 yield. Photolysis (30 ) and thermolysis (125°) of the sodium salt of B-naphthaldehyde tosylhydrazone in benzene gave, in addition to the aldazine, 2_9 in 67% and 78°6, respectively. The existence of g-naphthaldazine in the reactions involving the 6-naphthaldehyde tosylhydrazone sodium salt was proven by comparison of the reaction product with an authentic sample of the azine**" which was independently synthesized."*^ While decompositions of both carbene precursors in benzene gave the same hydrocarbon products, the decompositions of 4 , 5-benzotropone tosylhydrazone sodium salt in cyclohexene were unique in that there were additional products formed when compared to the 3-naphthaldehyde tosylhydrazone sodium salt decompositions. First, the pyrolysis of g-naphthaldehyde tosylhydrazone sodium salt was carried out in cyclohexene at 135-145°. The relative yields of the three volatile components are as follows: the naphthylcarbene-cyclohexene insertion product 30 (34%), the syn -norcarane 51 (26%), and the anti -norcarane

PAGE 28

15 32 (401). The aldazine was formed in 101 yield, but it is not in the above data as they only include volatile hydrocarbon products . 30 32 31 The insertion product 30_ was readily identified by nmr ; the two vinylic and two "benzylic" protons were readily discernible. The syn and anti -norcaranes required slightly more attention to effect differentiation, however. The corresponding phenylnorcarane models'*^ and the guidelines put forth for their identification by Closs and coworkers'*^''*'* aided the characterization of the naphthyl moieties. Thus, by analogy, the syn -7(2-naphthyl)norcarane 31 should have the naphthyl ring system facing the cyclopropane ring and this conformation would then have two separate effects on the nmr spectrum of the compound: first, several of the cyclohexyl ring protons would be held in the shielding cone of the naphthalene ring system and be consequently shifted upfield as far as 0.7 ppm compared with the anti -epimer ; second, the aromatic protons would not interact with the cyclopropyl ring and their signals would be relatively narrower than the other compound. The anti -epimer 32 should exist in the conformation with the

PAGE 29

16 naphthyl ring system bisecting the cyclopropyl ring"*** which would cause one of the o-protons to lie in the shielding zone of the cyclopropane ring and to be shifted upfield approximately 0.2 ppm relative to the syn -product (of course the cyclohexyl ring protons would be slightly downfield and not so broadened when compared with the former) . Second, when the sodium salt of 4 , 5 -benzotropone tosylhydrazone was decomposed in cyclohexene heated to reflux (83°) five products were formed. These products were detected and isolated by vpc. The vpc trace of the decomposition of 4,5benzotropone tosylhydrazone sodium salt in cyclohexene at 83 is shown in Figure 1. Two products accounted for 66% of Figure 1. Vpc trace: Pyrolysis of 26_ in Cyclohexene

PAGE 30

17 the material: one is unidentified, but the other has the properties of the spiro-adduct of benzocycloheptatrienylidene to cyclohexene (35) . The 100 MHz nmr of 33 is shown in 33 Figure 2 and the decoupled spectrum in Figure 3. Perhaps the cyclohexyl moiety is introducing the dissymmetry present in the spectra, but it is difficult to rationalize the large differential chemical shift of the two different cycloheptatrienyl protons adjacent to the benzene ring system. The uv spectrum, X = 240 my (log e = 4.2) and 265 my (log e = 3.5), compares well with that of 3 ,4-benzocycloheptatriene , X 230 my (log e = 4.7) and 265 my (log e = 3.7).'*5 The mass spectrum of this compound indicates a molecular ion at m/e 222. The ir contributed little toward the identification except that the strong absorption at 750 cm is indicative of an o-disubstituted benzene."*^ Also, other absorptions are present in the ir spectra of both 3,4-benzocyclohepta-l , 3,6triene**^ and the above unknown compound with approximately the same intensities (1620, 1480, 1440, 795, 745 cm" and several weak bands). The last three products and their relative yields were, respectively, (cyclophexene3-yl) 2-naphthylmethane, the naphthylcarbene-cyclohexene insertion product, (7%), syn -7(2-naphthyl) norcarane (121), and anti 7(2-naphthyl)

PAGE 31

18 1 ..^i-l-M-^-?-

PAGE 32

19 ,.4-l-i

PAGE 33

20 Pyrolysis of the sodium salt of 4 , 5-benzotropone tosylhydrazone in cyclohexene at 140 led to the same five products as the lower temperature one. The sodium salt of 4 , 5-benzotropone tosylhydrazone was thermally decomposed at 82 in a 2:1 molar mixture of diglyme-cyclohexene. Analysis of the hydrocarbon products indicated the presence of the same five products as the other cyclohexene decompositions with the addition of the naphthylcarbene -diglyme insertion products. The diglyme products amounted to 271 of the total products. Finally, the photolysis of 4 , 5-benzotropone tosylhydrazone sodium salt in cyclohexene at 30 produced the same products as the pyrolyses plus a new unidentified one. In the photolysis, the benzocycloheptatrienylidene addition product 33^ is a major component; in the pyrolyses, the opposite is true. In the cyclohexene-diglyme reaction, interfering materials made accurate analysis impossible. When cyclohexane was used for the substrate, the product isolated in 51% yield from the decomposition of 4, 5-benzotropone tosylhydrazone sodium salt was (3-naphthyl) cyclohexylmethane (34) . While the yield was not exceptional, the nmr spectrum of the crude reaction material indicated almost exclusively the insertion product and cyclohexane. ^ 34

PAGE 34

21 In an effort to trap the diazo compound of 4,5-benzotropone, the tosylhydrazone sodium salt was decomposed in benzene at 105-108° in the presence of an equimolar amount of dimethyl fumarate. There was considerable nitrogen evolution, and the product isolated from the reaction was the naphthylcarbenebenzene adduct 29_ in 65% yield. The nmr spectrum of the crude reaction material indicated only the adduct and dimethyl fumarate. Comparative tic, on silica gel eluted with 50:50 ethyl ether-pentane , of the known pyrazoline ( vide infra ) and this reaction mixture indicated no pyrazoline was present in the latter. Naphthyldiazomethane was treated in a similar manner. The sodium salt of g-naphthaldehyde tosylhydrazone was decomposed in benzene at 105-108° in the presence of an equimolar amount of dimethyl fumarate. The product, an almost colorless, very viscous oil, had all of the properties of trans -3,4dicarbomethoxy-5(2-naphthyl) 2-pyrazoline (3_5) . The oil was 3_5 not stable to ordinary distillation conditions and even though column chromatography on silica gel effected some purification, it was not sufficient to provide a sample which gave a good elemental analysis (e.g., calcd. N, 8.97; found: N, 8.31). The ir spectrum of this compound had a strong

PAGE 35

22 absorption at 3440 cm iiulLcative of N-H stretching, one at 1730 for the unconjugated ester carbonyl stretching, one at 1700 for the conjugated ester carbonyl, and one at 1555 indicating a C=N bond. The nmr spectrum contained absorptions at 2.10-2.60 T indicating the naphthyl protons, a broad singlet at 3.24 T for the proton bonded to the nitrogen, a broad doublet (J = 9.4 Hz) for the "benzylic" proton at 4.60 t, a doublet (J = 9.4 Hz) for the proton adjacent to the carbomethoxy group at 5.92 x, a singlet for the conjugated ester methyl at 6.18 x, and a singlet for the unconjugated ester methyl at 6.24 x (in addition, there were other resonances which changed in intensity depending upon work up and purification steps and were therefore not believed to originate from the pyrazoline). The trans configuration was evident when compared with nmr data of trans 3 , 4-dicarbomethoxy5phenyl2-pyrazoline'* ^ which had J = 10 Hz for the protons on C-4 and C-5 and a chemical shift of 4.87 x for the benzylic proton on C-5. It is also interesting to note that the above phenyl pyrazoline was a viscous oil which resisted all attempts to induce crystallization.'*^ The mass spectrum had no signal for the molecular ion, but there was one at 284 (the molecular ion with the loss of Na). The X^^^ in ethanol were 224 my (log £ = 4.53), 275 (3.75), and 287 (3.79). Using 2-methylnaphthalene** ^ and the known 3-carbomethoxy-4-phenyl2-pyrazoline^° as models, the following absorptions would be expected in the uv: 223 my (log e = 4.22), 274.5 (3.73), and 296 (3.98).

PAGE 36

23 The decomposition o£ 3-naphthaldehyde tosylhydrazone sodium salt in benzene at 160 in the presence of an equimolar amount o£ dimethyl fumarate gave rise to a different product than the low temperature one. The product was identified as trans 1 , 2-dicarbomethoxy-3(2-naphthyl) cyclopropane (56) and was isolated in 67% yield. The structure was easily desigH 36 nated trans because of the two different methyl proton absorptions (either cis cyclopropane would have had only one signal for the two methyl groups as they would be magnetically equivalent) and because the thermal decomposition of trans 3 ,4-dicarbomethoxy-5-phenyl-2-pyrazolines is known to give the trans -cyclopropanes . '*° The decomposition of the sodium salt of 4 ,5-benzotropone tosylhydrazone in benzene at 160 in the presence of dimethyl fumarate led to the formation of 2(2 ,4 ,6-cycloheptatrien-l-yl) naphthalene (29) and what is believed to be 2(2 ,4 ,6-cycloheptatrien-4-yl)naphthalene (32.) in a ratio of 1.0:1.6 as determined by nmr. The latter product could arise via a [1,5] hydrogen shift of the former; the hydrogen shift product was not isolated but its presence was inferred from the nmr spectrum and from work done in the phenylcycloheptatriene series . ^ ^

PAGE 37

24 3-naphthaldehyde tosylhydrazone sodium salt was decomposed by photolysis in the presence of trans 2-butene to yield 1(2-naphthyl) trans 2 , 3dimethyl cyclopropane (38) along with a small amount of cisand unknown materials. While other properties were consistent with this structure, the identification of this compound was hampered by the unusual nmr spectrum. The two signals arising from the methyl group protons were essentially singlets (they would be expected to be doublets) . Two things were tried in order to solve this dilemma: first, the nmr spectrum was taken in a different solvent, benzene, which caused one of the singlets to split by about 1.5 Hz, and, second, the corresponding phenyl cyclopropane was prepared and its nmr spectrum examined and found to be very similar to the naphthylcyclopropane . Thus, the structure is in accord with the observed spectral properties.

PAGE 38

25 The photolysis o£ the sodium salt of 3 -naphthaldehyde tosylhydrazone in cis 2-butene gave the two expected epimeric cyclopropanes along with a small amount of trans -cyclopropane and unknown materials. Syn -1(2-naphthyl) -cj^2 .3dimethyl 39 12. cyclopropane (39^) and anti -1(2-naphthyl) cis -2 , 3-dimethylcyclopropane (40) were differentiated in a manner similar to that of the epimeric naphthylnorcaranes ( vide supra ) . In the syn cyclopropane 39 the methyl protons appear 0.2 ppm upfield relative to the anti -compound, and the naphthyl resonances are relatively less broad. Naturally, the methyl resonances were at lower field and one of the o-protons on the naphthyl system appeared at higher field (causing a broadening of the aromatic resonances) in the anti relative to the syn . Pyrolysis of the naphthaldehyde tosylhydrazone sodium salt in cis -2-butene gave, in addition to the epimeric cyclopropanes, an unknown product formed in considerable amounts. When 4 ,5-benzotropone tosylhydrazone sodium salt was decomposed by photolysis in the presence of cis 2-butene , it gave the expected epimeric cyclopropanes, a small amount of trans -cyclopropane , and a considerable amount of unknown

PAGE 39

26 material. The pliotolysis in trans 2-butene gave as a major product the expected trans -cyclopropane (in addition to small ainonnts of c is cyclopropane and unknown material). The pyrolysis of 4 , 5 -benzotropone tosylhydrazone in cis 2-butenc yielded large amounts of the epimeric cyclopropanes , small amounts of trans -cyclopropane , and some extraneous materials . The stereochemistry of the addition of the carbenes to olefins was then determined by analysis of the amount of trans -cyclopropane formed in the decomposition of the two carbene precursors in research grade cis 2-butene . The decompositions were carried out under a variety of conditions. They were done using the neat olefin as solvent, they were done with the olefin diluted with benzene (1:3 molar ratio, respectively) and perf luorocyclobutane (two runs, 1:2 molar ratio and 1:6 molar ratio, respectively), and they were done with olefin in the presence of oxygen and trans , trans -2 ,4hexadiene (two runs, 1.0:1.5 molar ratio and 1.0:2.1 molar ratio, respectively). Dilution had no effect upon the ratio of cis and trans -cyclopropanes formed; however, the presence of oxygen and the diene significantly altered the above ratio, producing relatively more cis -cyclopropane . \Vhen the mole percent of trans , trans 2 ,4-hexadiene was greater than 661, there was too little cis_2-butene adduct formed to determine the stereochemistry. The results of the stereochemical investigation are shown in Table 1.

PAGE 40

27 TABLE 1 Stereochemical Results: The Percent "Wrong Isomer"^ from the Decompositions in cis 2-Butene of the Carbene Precursors both Thermally at 105° and Photolytically at -25°. Tosylhydrazone Sodium Salt I "Wrong Isomer" Solvent System (molar ratio) A hv Naphthaldehyde d cis -2-butene 10 3 Benzotropone cis 2-butene 7 4 Benzotropone cis 2-butene /benzene (1:3) 6 Benzotropone cis 2 -butane/ perfluorocyclobutane (1:2) 6 Benzotropone cis 2-butene/perf luorocyclobutane (1:6) 6 Benzotropone cis 2-butene/ oxygen (saturated) 4 Benzotropone cis -2-butene/ 2 ,4-hexadiene (1.0:1.5) 1 1(2-naphthyl) trans -2 , 3-dimethyl cyclopropane research grade 99.94 mole % standard deviations were typically ±1 datum only approximate due to interfering material

PAGE 41

28 To be sure tliat the products were stable to vpc, the cis cyclopropane was prcparatively chromatographed and then analyzed for trans -material --none was found. Also, the olefin was analyzed after the reaction and it was found that the olefin had isomerized less than 1 percent during the course of the reaction . In hope of sensitizing the rearrangement, the sodium salt of benzotropone tosylhydrazone was decomposed photolytically in benzene using xanthen-9-one (E„ = 74 Kcal/mole) as a sensitizer. Unfortunately, no hydrocarbon products were isolated from the reaction. The substituted tropone, 2-methyl-4 ,5 -benzotropone was made according to the method of Hielbronner^^ (the conversions necessary to obtain the sodium salt of the corresponding tosylhydrazone were the same as those used for 4 ,5-benzotropone itself) . The sodium salt of 2-methyl-4 ,5-benzotropone tosylhydrazone was thermally decomposed in benzene at 130° to yield two products: 2viny Inaphthalene (41^) and 2 (2 ,4 , 6-cycloheptatrien1-yl) -3-methylnaphthalene (42_) . The authenticity of the r:^ .^ 41 42

PAGE 42

29 2-vinylnaphthalene from this reaction was verified by comparison of its spectral properties with those of commercially available material. The substituted cycloheptatrienylnaphthalene was identified by its spectral properties and by comparison with material obtained from the thermal decomposition of the sodium salt of 3-methyl2-naphthaldehyde tosylhydrazone in benzene (vide infra) . The sodium salt of 3-methyl2-naphthaldehyde tosylhydrazone was thermally decomposed in benzene at 145° and yielded the above cycloheptatrienylnaphthalene 42_ (39%) and aldazine (42%). The products from the decomposition of the benzo precursor were analyzed by vpc and it was found that the relative ratios of the products were 2-vinylnaphthalene 72%, unknown material 61, and 2(2 ,4 ,6-cycloheptatrien-l-yl) -3-methylnaphthalene 22%. Another annelated tropone, 2 ,3-benzotropone was prepared as shown in the scheme below^^ (again the conversions to the tosylhydrazone sodium salt were the same as those necessary for 4 ,5-benzotropone) . PPA a > I CH2(CH2)3-C02H ^ Brz/CCl^ < LiCl/DMF o^ Br Br

PAGE 43

30 The decomposition of the sodium salt of 2 , 3-benzotropone tosylhydrazonc in benzene at 120° gave 1 (2 ,4 , 6-cycloheptatrien1-yl) naphthalene (43^). The compound was identified by 43 its spectral properties as follows: the nmr has resonances at 2.20 T three a-protons of the naphthyl system; 2.55 t four B-protons of the naphthyl system; 3.30 x two protons on the 4 and 5 positions of the cycloheptatriene ring; 3.70 t two protons on the 3 and 6 positions of the cycloheptatriene ring; 4.50 t two protons on the 2 and 7 positions of the cycloheptatriene ring; and 6.60 t for the one proton on the 1 position of the cycloheptatriene ring. The mass spectrum gave peaks at m/e 218 for the molecular ion, 127 for the naphthyl ring system, and 91 for the tropylium fragment. That it was the a-substituted naphthalene was shown by the ratio of the ato g-protons in the nmr spectrum and by close examination of the ir spectrum (Cf. 2(2 ,4 ,6-cycloheptatrienl-yl)naphthalene) ; the lack of any absorptions at 861 cm and 821 cm and the appearance of a band at 800 cm (for three adjacent H) and the mutual band at 745 cm \ias in agreement with this structure.

PAGE 44

CHAPTER III DISCUSSION To date, the interconversion of aryl and aromatic carbenes^.S^.^.i°>i2,i6,i8 (and nitrenes^^ . ^ «) --such as the conversion o£ phenylcarbene to cycloheptatrienylidene --has been limited to the gas phase and rather high temperatures (the exceptions to this have been the ring expansion o£ 1,6methano(10)annulenyl-2-carbene^^ and the small amount of rearrangement observed in the photolysis at room temperature of o, m, and £-methylphenylcarbenes ^ "*) . Also, with the exceptions of aromatic carbene dimerizations and of the reaction of the expanded 1 , 6-methano (10) annulenyl2-carbene with 1 ,2-cyclononadiene ,* in no case has it been possible to examine intermolecular reactions of any of the rearranged carbenes. At this time, the rearrangement of 4 ,5-benzo cycloheptatrienylidene to B-naphthylcarbene and that of 2,3-benzocycloheptatrienylidene to a-naphthylcarbene in solution at temperatures as low as -35° is reported. These reactions constitute examples of carbene-carbene rearrangements which are apparently not subject to either of the previous restrictions . W. M. Jones and P. H. Gebert, unpublished results. 31

PAGE 45

32 For this study, the phcnylcarbcnecycloheptatrieny li dene reaction was taken as the prototype. Pheny Icarbene (44) was generated in diglyme under conditions where the sodium salt of tropone tosylhydrazone gives heptafulvalene and the reaction mixture was carefully examined for cycloheptatrienylidene dimer. No trace was found. Cycloheptatrienylidene (45) was tlien generated under conditions where phenylcarbene gives clean reaction with solvent and the reaction mixture carefully analyzed for phenylcarbene insertion products. No trace was found. The failure of phenylcarbene to expand in solution is probably due to a favorable competitive intermolecular reaction with solvent. Cycloheptatrienylidene, on the other 7 \\ "^^^>^"^^ > Dimer hand, is very unreactive with most solvents, preferring to simply dimerize. Thus, its failure to contract reflects an activation energy high enough to allow dimerization to dominate, As one possible way to reduce the activation energy for contraction while hopefully retaining enough of the low reactivity of the aromatic carbene to prevent extensive reaction with solvent, benzocycloheptatrienylidene was studied. This system was chosen because monoannelation is

PAGE 46

33 known to substantially decrease the stability of the tropylium ion^^ but it should have little effect upon the stability of the intermediate cyclopropene . Thus, if the stability of the carbene parallels the carbonium ion and if the rate of contraction is reflected in the relative stability of the intermediate, then benzocycloheptatrienylidene should contract faster than cycloheptatrienylidene. The carbenes were generated from the sodium salts of the corresponding tosylhydrazones and the rearranged naphthylcarbenes were detected by their reaction products with solvent. Thus, the formation of the naphthylcycloheptatriene 29^ suggests a carbene-carbene rearrangement followed by reaction of the rearranged carbene 4^ with solvent. Formation of the rearranged 46 29

PAGE 47

34 carbene was further supported by the formation of 3-naphthylcyclohexylmethane (34_) upon decomposition of the salt in cyclohexane. Considering decompositions in cyclohexene, it is interesting to note that the relative yield of the benzocycloheptatrienylidene adduct (3_3) appears to decrease with increasing temperature. The presence of 3^ and its apparent decrease with increased temperature is consistent with a bimolecular reaction of the unrearranged carbene with cyclohexene (to give 5^) in competition with a unimolecular rearrangement that gives 6-naphthylcarbene . Lastly, formation of the various insertion products can be taken as strong evidence^^ for formation of the rearranged carbene. The mechanism of the carbene-carbene rearrangement, however, is quite a detailed problem. Initially, one must consider the rearrangement of a carbene precursor. Thus, the decompositions of g-naphthaldehyde tosy Ihydrazone sodium salt and an equivalent amount of 4 ,5-benzotropone tosylhydrazone sodium salt were carried out under identical conditions in the presence of an equimolar amount of dimethyl fumarate. As expected, the 3-naphthyldiazomethane reaction gave as the primary product the pyrazoline. In contrast, the benzotropone tosylhydrazone salt reaction gave the now familiar benzene addition product of g-naphthylcarbene 29^ with no detectable trace of the pyrazoline 3_5. When the reactions v^?ere carried out at higher temperatures, the former yielded the 1,2dicarbomethoxy-3-naphthylcyclopropane 36^ and the latter gave,

PAGE 48

35 in addition to the usual 3-naphthylcarbene-benzene adduct, a new product 3_7. The nmr of the unisolated material suggests that it is the thermal (1,5) hydrogen shift product of the original adduct. All of the above data are consistent with the rearrangement occurring via a carbene, not a carbene precursor. Once it was determined that a carbene was probably the rearranging species, it became necessary to differentiate between rearrangement occurring from the singlet state and that from the triplet state and kno^vledge of the multiplicity of the initial carbene was imperative. In principle, this facet of the problem may be explored by investigating the stereochemistry of the addition of rearranged B-naphthylcarbene to olefins. Stereospecif ic addition is taken as strong evidence for the reaction of the singlet state, whereas the triplet state appears to add nonstereospecif ically . ^ ^ ' ^ ® Furthermore, the ground state of B-naphthylcarbene has been shown to be triplet. ^^ Thus, a high degree of stereospecif ic addition would point to the singlet as the reacting species (unless the singlet and triplet are in equilibrium --this will be discussed belov\f) . The rearrangement via the triplet state might occur as shown in the scheme below (this is simply a diradical form of the cyclopropene mechanism, vide infra).

PAGE 49

36 Photolytically , in the pure olefin, g-naphthylcarbene from both precursors gave 96-971 stereospecif icity which is in good agreement with the work of Gloss"* ^ who found 95-97% stereospecif icity with phenylcarbene in pure olefins. Thermally, the stereospecif icity was less than the corresponding pliotochemical reactions (ca. 93^) ; this may be because of a thermal dependence of intersystem crossing.^' It is in agreement with the amount of nonstereospecif icity observed in the addition of thermally generated dicyanocarbene to cis 2-butene (8% in neat olefin), however, singletto-triplet transition in dicyanocarbene is competitive with addition, and thus the stereospecif icity is sensitive to dilution effects.^" The dilution effect is caused by the different concentration dependence of intersystem crossing and addition to the olefin; the former is unimolecular and the latter, of course, is dependent upon olefin concentration.^' When applied to the naphthylcarbene system, dilution either with benzene or octaf luorocyclobutane had no effect upon the stereospecificity.

PAGE 50

37 Another experimental test for multiplicity is scavenging o£ paramagnetic species by oxygen or dienes. As Gloss states, ^^ the stereospecificity should be increased in the presence o£ a diene (if the triplet is not in rapid equilibrium with the singlet) because the scavenging of the paramagnetic species "results in a correspondingly greater proportion of the singlet-state reaction." Thus, when the 3naphthylcarbene is generated from the salt of 4 , 5-benzotropone tosylhydrazone in the presence of olefin and oxygen, the addition is somewhat more stereospecif ic. By making some approximations from existing data,^^ namely that the solubility of oxygen in benzene is similar to that of butene and that the solubility is a linear function of temperature, a saturated solution of 120 millimoles of butene should contain ca. 0.2 millimoles of oxygen under reaction conditions. This concentration was apparently sufficient to increase the stereospecificity by 2% (which was also the amount it was increased when applied to the f luorenylidene system by M. Jones, Jr.).^^ Furthermore, more effective scavenging was accomplished by utilizing trans , trans -2 ,4-hexadiene . In this case the effect was pronounced in that the stereospecificity was increased to 99%. In the hexadiene experiments, a large part (ca. 901) of volatile product was apparently from reaction with the diene (relative to the amount of reaction with butene as determined by vpc) . This is quite in line with the data of Moss^® who, using similar conditions, reported that approximately 70% of the reaction product

PAGE 51

38 appeared to be derived from the reaction of phenylcarbene with the diene. In summary, the 3-naphthylcarbene added to olefins with a high degree of stereospecif icity which was unchanged by dilution but was increased by the addition of triplet scavengers. These data are compatable with a scheme in wliich there is a thermally driven singlettriplet equilibrium wherein the stereospecif icity is independent or only slightly dependent upon olefin concentration. If there is a rapid singlettriplet equilibrium, one can no longer be assured that the singlet is the rearranging species. The rearrangement could be occurring via the triplet state, but since there is an equilibrium it might not be reflected in the stereochemistry. These data appear inconsistent with a scheme in which intersystem crossing is in effective competition with addition of the singlet to olefins as in the case of f luorenylidene . ^ ^ The stereospecif icity of addition of f luorenylidene to olefins is dependent upon both the concentration of olefin and the addition of triplet scavengers. They are also incompatible with a scheme where there is a rapid singlettriplet equilibrium in which the stereospecif icity is unaffected by the addition of triplet scavengers (as in the case of diphenylcarbene) . ^ ^ Initially it seems that a dicotomy exists, but one can envision a system whereby all of the above facts can be readily explained and which leaves little doubt but that the initially formed S-nanhthvlcarbene is a singlet.

PAGE 52

39 The following scheme indicates the effects of changing the concentrations of diene and olefin upon cis and trans cyclopropane ratio; the explanation of symbols is as follows B = benzocycloheptatrienylidene S = singlet naphthylcarbene T = triplet naphthylcarbene D = diene (2 ,4-hexadiene) = cis -2-butene P = diene-carbene products >P ^^|i^ = ki[0] [S] + ky[0] [T] d trans dt k,[0] [T] d cis "^t ^ k,[0] [S] + ky[0] [Tl d trans kTToTm ~^[t Integrate over all time and assume the right-hand side to be constant.

PAGE 53

40 cis _ ki [0] [S] ^ k7 [0] [T ]_ ki [Sj_ k7 r.. FriTHi' knorm i^n ^ ^^ Assume a steady-state concentration for [T] which is "particularly good when the intermediates are very reactive and, therefore, present at very small concentrations."^^ ^^ = k2 [S] k3 [T] k^ [0] [T] k7 [0] [T] ks [D] [T] = PPT kzLS] ^ ^^^ "TTT^k.LO] + k7[0] + ksLD] (2) Substitute for [T] from eq. (2) into eq. (1) and simplify cis ^ kika + kik^[0] + kik7[0] -+ kiksED] + kgky trans IcTk^ If there is a rapid equilibrium, ks >> k^ [0] , k7 [0] , ks [D] and kz >> ki [0] . Then cis _ kik; trans ~ kzki* (3) and, the ratio is independent of olefin and diene concentrations . If the reaction is irreversible, k^ [0] >> ks, k? [0] >> ks. Disregard the diene for first case. Then '^^^ = ^1 (^'^ -^ Y''} ^^^ '^ ^^^^ trans FITii and the ratio is dependent upon olefin concentration. Now include the diene in the scheme k^ [0] >> ks, k7[0]->> ks and ks [0] >> ks . cis ^ kik5[D] + kiCk^ + k7) [0] + k2k7 ,.^ trans kaki^ '^ -' and the ratio is dependent upon diene and olefin concentra-

PAGE 54

41 If, however, kj >> k^ [0] , ka >> ky [0] , but ks [D] >> ka then cis kiks [D] + k2k7 kakii (5) trans and the ratio is dependent only upon the diene concentration. One must consider, however, that the rearrangement might be going via the triplet benzocycloheptatrienylidene (one need not concern oneself with triplet benzocycloheptatrienylidene rearranging to singlet naphthylcarbene) . Thus, the scheme below indicates entry into the triplet rearranged carbene . >P d cis = ki[0] [S] kyLO] [T] Integrate as in the first case cis ^ ki [0] [S] + k7 [0] [T] ^ trans k,* [0] [T] d trans ""cTt = k^ [0] [T] ki [S] k, [T] k„ • Assume a steady-state concentration of [S] . ki [0] [S]ksED] [S] = C6) d [S] dt

PAGE 55

Substitute [S] from eq. (7) into cq, (6) and simplify. 42 CIS kikj + ^7 k^(k2 + ki [0] + ka [D]~T 1^ trans If ka [D] >> kz, but ka >> ki[0] then (8) cis _ kika ^ k? trans k^ka [D] F^ (9) and the ratio is inversely proportional to the diene concentration. Finally, one might envision a scheme wherein the benzocycloheptatrienylidene is the intermediate which undergoes the singlet-triplet intersystem crossing. Each state of the unrearranged carbene then yields the corresponding 3-naphthyl carbene as shown in the scheme below; the explanation of symbols is as follows: B = singlet benzocycloheptatrienylidene B = triplet benzocycloheptatrienylidene N = singlet naphthylcarbene N = triplet naphthylcarbene D = diene = cis 2-butene P = diene-carbene products

PAGE 56

43 There would be no dilution effect because intersystem crossing and rearrangement are both unimolecular ; the scavenging effect would arise from reaction of diene with the unrearranged carbene. In agreement with the arguments above, the singlet still must be the state by A^hich most of the rearrangement occurs in order to explain the high degree of stereospecif icity observed. Excluding the benzocycloheptatrienylidene intersystem crossing, what is required for the explanation of the facts concerning the stereospecif icity of g-naphthylcarbene addition is that the rate of triplet-singlet intersystem crossing proceeds at a rate considerably greater than those of the additions of the triplet carbene to the mono-olefin. Also, the rate of triplet addition to the diene (and oxygen) must be greater than the rate of triplet-singlet intersystem crossing. Accordingly, it has been reported tliat diphenylcarbene reacts with 1 , 3-butadiene more than one hundred times faster than unconjugated olefins.^** The above kinetic scheme depicts the desired situation in eq. (5) and is also compatible with the earlier situations of f luorenylidene as shown in eq. (4) and diphenylcarbene as shown in eq. (3). It also vitiates the mechanism whereby the triplet is the rearranging species because the addition of diene would be required to decrease the cis trans ratio as sliown in eq. (9). With the triplet pathway ruled out, at least four mechanisms have been considered which arise from the singlet state of the unrearranged carbene. The singlet mechanisms

PAGE 57

44 are a carbocyclic version of the Wolff rearrangement, a Skatteboltype rearrangement, that going via a fused cyclopropene intermediate, and a concerted rearrangement. The Wolff-type rearrangement (or perliaps more accurately the "retro-Wolff" rearrangement) is shown below. This > mechanism is not very favorable for two reasons. First, annelation should retard it because of loss of aromaticity in the benzene ring. Second, Jones and Mitsuhashi have postulated that the Wolff and the cyclopropane mechanisms would differ experimentally in that the latter mechanism should be favored by high double bond character whereas the Wolff mechanism should be retarded.^"* Their work on the importance of bond order in the interconversion of aryl and aromatic carbenes by studying the gas-phase, carbene-carbene rearrangement of a series of substituted naphthylcarbenes, showed that net cleavage of bonds of high bond order is observed; net cleavage of bonds of low bond order is not. Thus, the evidence mitigates against a Wolff mechanism. Another mechanism which can be written for the contraction of cycloheptatrienylidene (and also for benzocycloheptatrienylidene) is essentially the Skattebol^ vinylcyclopropyli-

PAGE 58

45 dene-cyclopentenylidene rearrangement with cleavage o£ different bonds. However, the effect of annelation argues against this possibility since it should retard norcaradiene formation (to prevent loss of aromaticity) and hence the rate of contraction. The differentiation of the mechanisms involving the fused cyclopropene (route a) and the concerted pathway (route b) present a subtle problem. In fact, unless the benzobicycloheptatriene intermediate can be trapped or observed (spectroscopically, etc.) there can be no real distinction between the two. If the reaction is not concerted, the activation

PAGE 59

46 energy for ring opening must be quite low* as evidenced by the fact that benzob icycloheptatrieny lidene contracts to 3-napmhylcarbene at temperatures as low as room temperature. The low activation energy is not unreasonable since the energy lost in cleaving tlie single bond o£ the twisted cyclopropene (the bcnzobicycloheptatrienylidene) might well be of the same order of magnitude as the naphthyl resonance energy gained. In the model system it has been suggested that "relatively small energy barriers separate phenylcarbene , cycloheptatrienylidcne , and bicycloheptatrienylidene isomers. "^^ For tlie discussion of a detailed mechanism of the rearrangement it will be assumed that the process occurs via the fused cyclopropene intermediate since this facilitates the dissection of tlie problem and because it seems intuitively likely that the cyclopropene is a distinct intermediate. Accordingly, Bergman^ ^ recently reported that the thermal isomerization of a substituted cyclopropene very likely went through a vinylcarbene intermediate. He gave no suggestions as to the pathway of ring opening and closure except to rule out a "diradical" mechanism. Assuming that the intermediate has a half -life of 0.1 sec and that the pre-exponential factor (A) is similar to that for cyclopropene opening/^ a value of ca. 15 Kcal/mole is found for the energy of activation at 27°.

PAGE 60

47 Thus while the preferred conformation for closure of vinylcarbenes to cyclopropenes is unknown, it may occur from a structure in which the filled orbital of the carbene is orthogonal to the ir-system which was suggested by Closs^' and by calculations which imply that this should be the conformation of the ground-state singlet of vinylcarbene . ^ ^ Using Zimmerman's^^ MO Following approach to correlation diagrams, one can arrive at the preferred conformation of the rearranging (or rearranged) carbene. This approach is applied to cases where there is no element of symmetry and which consist of a linear array of orbitals. Considering benzocycloheptatrienylidene and treating the system as a vinylcarbenecyclopropene addition, the remainder of the benzenoid system is disregarded (the subsequent ring opening is just the reverse of the addition). Application of Zimmerman's "four basic rules" yields the correlation diagram shown in Figure 4. It is immediately obvious that p. p. is the occupied orbital (in the singlet state) and, since it accepts the electron pair, it will have added s character. The final result is that of the benzocycloheptatrienylidene with the vacant orbital vertically oriented (p.. + p.) and conjugated with the TT-system, and the orbital containing the non-bonded electrons, s + (p-, p.) weighted, horizontally oriented in a plane containing the sigma framework.

PAGE 61

48 ILAI "3 • (3) +1-2+3-4 Pl+P"*

PAGE 62

49 In hope of elucidating the mechanism somewhat, the effect of methyl substitution upon the rate of carbenecarbene rearrangement was cursorily examined. From the preferred formation of 4_1 versus that of 4_2 in the decomposition of the sodium salt of 2-methyl-4 ,5-benzotropone tosyl hydrazone in benzene, it appears that alkyl substitution favors rearrangement. The validity of this conclusion 42 £1 rests on the fact that neither 3-methyl-2-naphthylcarbene (48) nor methyl2-naphthylcarbene (47) ring expands under the reaction conditions and thus indicates that both reactions are irreversible.* It is recognized that this datum will not enable one to pinpoint a single mechanism, but in order * From the product distribution, the free energy difference between the two product transition states is on the order of 1 Kcal/mole at 130°.

PAGE 63

50 to explain it, attention must be focused on some interesting concepts about this general type of rearrangement. In dissecting this problem, two types of interactions or conformations must be considered: a boat-shaped system wherein the ring containing the carbene center is puckered and delocalization of -rr-electrons into the carbene p orbital is minimal, and a planar system with delocalization of ir-electrons into the vacant p orbital of the carbene (similar to cycloheptatrienylidene® which calculations imply has a planar structure).* In each of these two conformations, one must examine three effects: steric, geometric, and electronic. Models suggest that steric effects are unimportant since in neither conformation is the methyl group able to significantly aid in the formation of the transition state by lessening any non-bonded interactions. The transformation from the boat conformation (Figure 5) could be envisioned as occurring by interaction of the empty Figure 5. Boat Conformation of Benzocycloheptatrienylidene . J. Sabin, Physics Department, University of Florida, personal communication.

PAGE 64

51 p orbital of the carbene with the rotating p orbital of the Y" carbon to form the new sigma bond and by interaction of the sp^ orbital of the carbene with the p orbital of the 3-carbon to yield the twisted cyclopropene double bond. From an examination of models, it appears that in the boat conformation there is very good orbital orientation for efficient overlap. Thus, it seems readily adapted to the formation of a fused, twisted cyclopropene (i.e., the ground state geometry of the boat form closely resembles that of a possible transition state). It is quite possible that the boat form is the lower energy conformation in this system in spite of the assumption that the benzotropylium cation has a planar structure. '° However, coplanarity is not requisite as the activation energy for the rearrangement may include the energy necessary for the planar system to become puckered, In the planar case (Figure 6) , the cyclopropene sigma bond would be formed by the interaction of the back lobe of the inverting sp^ orbital on the carbene carbon with the rotating p orbital on the y-carbon. Localization of a pair Figure 6. Planar Conformation of Benzocycloheptatrienylidene,

PAGE 65

52 of ir-electrons in the still conjugated p orbitals of the aanJ Bcarbons would tlien form the new double bond (in a manner similar to the nucleophilic attack of carbenes upon electron-deficient double bonds) . ^ The planar form has to rehybridize to accomplish the inversion at the carbene center (this is not required of vinylcarbene cyclization) . From examination of models, this reliybridization introduces strain into the ring before there is what appears to be good orbital overlap; this could increase the energy of activation for closure. That is, in contrast to the boat form, in this case there seems to be less initial interaction leading to the rehybridization in that initially the two potentially interacting orbitals are orthogonal. The published electron effects are obscure and, consequently, lend very little aid to differentiation of the two modes of attack. Closs attributes alkyl substitution effects to the possibility of both steric and electronic components in the formation of cyclopropenes from the ring closure. of vinylcarbenes ; ^ ^ it appeared, though, that the steric portion was dominant. Closs also observed that an alkyl group in the olefinic position of the cyclopropene was considerably less effective in its ability to stabilize the cyclopropene than alkyl groups at the saturated center.''^ From an examination of possible electronic effects in the boat conformation, it appears that in the formation of the transition state of the rearrangement the 3-carbon acquires a partial positive charge. This charge deficiency

PAGE 66

53 could arise from loss of conjugation with the p orbital of the y-carbon due to the latter' s rotation toward the carbene center. What the electron deficiency requires is a greater degree of bond breaking between $and ycarbons than bond formation between the aand 3carbons (shown below) which is not unlike the unequal bond formation predicted for the electrophilic addition of carbenes to double bonds. ''^ It is known that alkyl substitution facilitates this type of addition.'^ Unfortunately, one cannot make conclusive predictions about the electronic effects of the methyl group in the planar conformation. There is the possibility, though, that the methyl group would have little or no effect upon the rate of the rearrangement. Thus, concerning the closure of vinylcarbenes to cyclopropenes , Closs^' stated that with progressive rotation of the y-carbon there was the possibility of a decrease of ir-electron density at that site due to the formation of the double bond between the aand 6-carbons

PAGE 67

54 (implying little change in electron density at the 3-carbon). Using the same reasoning, it would appear that there would be little or no charge deficiency on the methylsubstituted 3-carbon of the planar system resulting in little or no effect by the methyl group on the rearrangement. It should now be pointed out that while geometric and electronic effects do not form any real basis for prediction of a preferred pathway, they are certainly consistent with the "electrophiliclike" type of interaction discussed earlier. Thus, the prediction of Closs and the calculations of Zimmerman need not be incorrect as vinylcarbene may rearrange in the manner described by them. But, carbocyclic, aromatic carbenes may well rearrange by the other pathway due to electronic effects brought about by geometric constraints. For example, the apparent similarity of the boat form to a possible transition state, and, in contrast to the assumed geometry for the closure of vinylcarbenes , ^ ' the geometry of cycloheptatrienylidenes is such that the sp^ orbital of the carbene is trans to the double bond. A clue to the reason for the rather dramatic effect of annelation on the ease of ring contraction may be found in the effect of annelation on the V^-d+ values of model cations. ^^ Carbonium ions should be fair models for these carbenes since, in aromatic carbenes, the vacant orbital is probably conjugated with the TT-system with the non-bonded pair of electrons in the plane of the sigma framework.^"* The tropyl cation has a value of 4.7 and the benzotropyl cation has a value of 1.7.

PAGE 68

55 Annelation would not be expected to substantially change the stability of the intermediate, but annelation of the tropyl cation causes a marked decrease in the pK^^ signaling a substantial destabilization. Thus, the energy of activation for ring contraction should be less in the annelated system than in the nonannelated one. von R. Schleyer* has done "back of the envelope" calculations on the "strain energy" on phenylcarbene , cycloheptatrienylidene , and the bicycloheptatriene and come to the conclusion that they are all of similar energy. Thus, it appears reasonable that even modest annelation effects could be enough to cause the energetics of the annelated system to favor ring contraction over the ring expansion exhibited in the cycloheptatrienylidene series. Lastly, in order to test the generality of the benzocycloheptatrienylidene-naphthylcarbene rearrangement, the sodium salt of 2 , 3-benzotropone tosylhydrazone was decomposed in benzene to yield the a-naphthylcycloheptatriene. The smooth rearrangement fits perfectly into the overall picture of the rearrangement. In summary, benzocycloheptatrienylidenes rearrange to naphthylcarbenes in solution thermally at temperatures in the range of 100 and photolytically at temperatures as low as -35 with the rearranged carbenes being detected by their reaction products with solvent. The rearrangement may occur P. von R. Schleyer, Princeton University, personal communication.

PAGE 69

56 via a fused cyclopropene intermediate from the boat form of the singlet unrearranged carbene, from the planar form of the singlet, or via a concerted pathway.

PAGE 70

CHAPTER IV EXPERIMENTAL General . -Melting points were taken in a Thomas-Hoover Unimelt apparatus and are uncorrected. Elemental analyses were performed by Atlantic Microlab, Inc., Atlanta, Georgia. Ultraviolet spectra (uv) were recorded on a Gary 15 doublebeam spectropliotometer using 1 cm silica cells. Infrared spectra (ir) were recorded with a Beckman IRIO with absorptions reported in reciprocal centimeters. In all cases where the KBr pellet technique was not used, sodium chloride plates were substituted. Nuclear magnetic resonance spectra (nmr) were determined on a Varian A-60A high resolution spectrometer or a Varian XL 100 spectrometer. Chemical shifts are reported in tau (t) values from internal tetramethylsilane standard. Mass spectra were determined on a Hitachi model RMU-6E mass spectrometer. Analytical thinlayer chromatography (tic) was accomplished on 2 in X 8 in plates coated in these laboratories with 0.25 mm layers of E. Merck HF-254 silica gel. Components were visualized by their quenching of fluorescence under uv light. Vapor phase chromatography (vpc) was carried out using a Varian Aerograph 90-P thermal conductivity instrument using column A (18 ft x 0.25 in in 20% SE-30 on Chromosorb W at 210°) or column B (3 ft x 0.25 in 151 SE-30 on Chromosorb W at 190°) and a Varian Aerograph Series 1200 flame ionizaS7

PAGE 71

tion instrument using column C (9 ft x 0.125 in 5°o Sli-30 on Chromosorb \\ at 155°). The analytical technique for determining product ratios from vpc data was that of cutting and weighing. All chemicals were reagent grade unless otherwise stated. Diglyme (diethylene glycol dimethyl ether) and tetrahydrofuran were dried by passage over activity grade I Woelm alumina and subsequent storage over CaH2 . The analyzed research grade cis 2-butene vv'as from Matheson Gas Products, Inc. ( cis -2-butene , 99.941; trans 2-butene , 0.04^; and butadiene, 0.02"OPyrolysis of the Sodium Salt of Benzaldehyde Tosylhydrazone in Diglyme . --To 30 ml of dry diglyme in a 3 oz FischerPorter Aerosol Compatibility Tube was added 0.10 g (0.34 mmol) of the sodium salt of benzaldehyde tosylhydrazone . ^ ^ The sealed tube was placed in an oil bath preheated to 165° for 30 min. At the end of this time all trace of color of the diazo was gone, leaving the solution colorless. The mixture was allowed to cool, poured into water (200 ml), and extracted with pentane (three 30 ml portions) . The combined pentane extracts were washed with water (three 50 ml portions) , dried (MgSOi,), and concentrated on a rotary evaporator under reduced pressure. The nmr spectrum of the residue, a light yellow oil, indicated no heptafulvalene was present, only what is believed to be the phenylcarbene-diglyme adducts. The nmr (CDCI3) showed t 2.80 (phenyl protons) and 6.5 (diglyme protons) . In addition, there was aldazine produced in the reaction. Comparative tic of the residue with a

PAGE 72

59 known sample of heptafulvalene was carried out using silica gel eluting \\rith pentane. This confirmed the absence of heptafulvalene in the reaction products. Pyrolysis of the Sodium Salt of Tropone Tosylhydrazone in Diglyme . Into a Carius tube was placed 80 mg (0.27 mmol) of the sodium salt of tropone tosylhydrazone^ and 10 ml of dry diglyme. The sealed tube was immersed in an oil bath preheated to 210° and allowed to remain for 30 min. Upon cooling, the tube was opened and the contents poured into water (120 ml) and extracted with pentane (three 33 ml portions). The combined pentane extracts were washed with water (three 50 ml portions) and dried (MgSOit). The dry pentane solution vv'as concentrated on a rotary evaporator under reduced pressure to yield a very dark solid. The nmr spectrum of the residue was identical to that of heptafulvalene with a small amount of aromatic resonance at 7.2 t, but none at 6.2-7.0 x (the position expected for any diglyme insertion products) . 4 , 5 -Benzo tropone . -The 4 , 5 -benzotropone used in this research was prepared by the method of Thiele and Keitz with the exception that a steel high pressure reactor was used in place of a sealed glass tube. 4 , 5 -Benzo tropone Tosylhydrazone . -In a typical run, 2.00 g (0.013 mol) of 4 , 5-benzotropone , 2.38 g (0.013 mol) of £toluenesulfonylhydrazide , and 2 drops of cone sulfuric acid were placed in 55 ml of 951 ethyl alcohol and heated at

PAGE 73

60 reflux for 1 hr. Upon cooling, 4 , 5-benzotropone tosylhydrazone precipitated from the solution and was isolated by filtration, Recrystallization from absolute ethyl alcohol yielded 2.5 g (51%) of the product. There were two crystal forms of the tosylhydrazone : red plates (mp 186-189 ) and yellow needles (mp 180-185 ). The other properties of the tosylhydrazone were as follows: ir (KBr) 3200, 1638, 1325, 1165, 575, 555 cm'-*-; nmr (acetone-de) t 2.00-2.80 (A2B2, 4H, J=8, toluyl aromatic), 2.62 (s , 4H, benzo) , 2.90-3.80 (m, 4H, cycloheptatrienyl) , 7.60 (s, 3H, toluyl methyl); mass spectrum 324 (M+). Anal . Calcd. for CisHieNzOzS: C, 66.64; H, 4.97; N, 8.64; S, 9.89. Found: C, 66.52; H, 5.07; N, 8.57; S, 9.97. 3-Naphthaldehyde Tosylhydrazone . In a typical run, 3-naphthaldehyde (Aldrich Chemical Co.), an equivalent amount of £toluenesulfonylhydrazide , and a few drops of cone HCl were placed in 95% ethyl alcohol and allowed to stir overnight at room temperature and the resulting crystals filtered and recrystallized from ethyl alcohol to give the tosylhydrazone. The product had a mp 175°, lit.^^ 174°. General Preparation of the Sodium Salts of Tosylhydrazones . -The sodium salts were prepared by dissolving the tosylhydrazone in dry THF and adding one equivalent of sodium hydride (57% in mineral oil Alfa Inorganics) slowly with stirring. The mixture was allov^?ed to stir for 1 hr. At the end of this period, pentane was added to precipitate the

PAGE 74

61 sodium salt (this was not always necessary) . The salt was filtered, washed with additional pentane, and dried. The entire operation was carried out in a dry box, although some of the salts seemed to be stable to laboratory conditions. Pyrolysis of the Sodium Salt of 4 , 5-Benzotropone Tosylhydrazone in Diglyme .--To 13 ml of dry diglyme in a 3 oz Fischer-Porter Aerosol Compatibility Tube ^vas added 0.45 g (1.3 mmol) of the sodium salt of 4 ,5-benzotropone tosylhydrazone. The sealed tube was placed for 20 min in an oil bath which had been preheated to 100 . At the end of this time 781 of the theoretical amount of nitrogen had been evolved. The reaction mixture was allowed to cool, poured into water (200 ml), and extracted with pentane (three 30 ml portions). The combined pentane extracts were washed with water (three 50 ml portions), dried (MgSO^), and concentrated on a rotary evaporator under reduced pressure. The remaining yellow oil was chromatographed on alumina and eluted with ethyl ether to yield 103 mg (29%) of the naphthylcarbene-diglyme insertion products. Although the yield was only 291, the nmr spectrum of the crude reaction mixture showed mainly diglyme and the adducts. The adducts had the following properties: ir (film) 3050, 2920, 2880, 1600, 1450, 1100, 860, 820, and 750 cm""*"; nmr (CDCI3) x 2.00-2.80 (m, 7H, naphthyl) , 6.10-7.20 (m , 15H, diglyme and benzylic) ; mass spectrum 274 (M+) . Anal . Calcd. for C17H22O3: C, 74.42; H, 8.08; 0, 17.49. Found: C, 74.59; H, 8.05; 0, 17.36.

PAGE 75

62 The sodium salt of 3-naphthaldehyde tosylhydrazono was thermally decomposed and worked up in a manner analogous to the above pyrolysis. The naphthylcarbene -diglyme adduct was isolated in 151 yield; the major product was naphthaldazine . The adduct 's nmr spectrum was identical to tliat of the products from the pyrolysis of the sodium salt of 4,5-bcnzotropone tosylhydrazone ( vide supra ) . Photolysis of the Sodium Salt of 4 ,5-Benzotropone Tosylhydrazone in Diglyme . --To 15 ml of dry diglyme in a photolysis vessel which consisted of a Pyrex tube with a side arm and equipped with a small mechanical stirrer, vv^as added 250 mg (0.72 mmol) of the sodium salt of 4 , 5-benzotropone tosylhydrazone. The mixture was rapidly stirred and the photolysis carried out by mounting the apparatus vertically alongside and as close as possible to a 550 watt Hanovia medium pressure mercury lamp in a Pyrex tube. Both the vessel and the tube containing the lamp were immersed in a circulating water bath which kept the temperature at 30°. The salt was photolyzed for 2 hr with 100% of the theoretical amount of nitrogen being evolved as monitored by a gas buret fitted to the side arm of the photolysis vessel. The reaction mixture was worked up as in the above pyrolysis to yield 51 mg (251) of the naphthylcarbene-diglyme insertion products. It should be noted that while the yield was low, the nmr spectrum of the crude reaction mixture indicated that the diglyme adducts were the major products.

PAGE 76

63 Pyrolysis of the Sodium Salt of 4 , 5-Benzotropone Tosylhydrazone in Benzene . --To a 3 oz Fischer-Porter Aerosol Compatibility Tube containing 35 ml of benzene was added 0.20 g (0.58 mmol) of the sodium salt of 4 , 5-benzotropone tosylhydrazone . The sealed tube was placed in an oil bath preheated to 120° and allowed to remain for 2.5 hr . The color of the reaction mixture changed from orange to light yellow during the course of the reaction. Upon cooling, the nitrogen evolution was monitored and 81% of the theoretical amount had evolved. The reaction mixture was filtered and the residue was washed with ethyl ether. The total filtrate was concentrated on a rotary evaporator under reduced pressure to give 122 mg (96%) of crude 2(2 ,4 ,6-cycloheptatrien-l-yl) naphthalene. Recrystallization from abs methyl alcohol yielded 68 mg (54%) of yellow crystals mp 85-86 (sublimation at 90 and 0.1 mm Hg gave white plates with same mp) : ir (KBr) 3060, 3020, 1600, 1500, 1280, 1260, 950, 900, 860, 820, 745, 700, 480 cm'"""; nmr (CDCI3) t 2.10-2.70 (m, 7H, naphthyl) , 3.26 (broad t, 2H, 4 and 5 cycloheptatrienyl) , 3.55-3.90 (complex m, 2H, 3 and 6 cycloheptatrienyl), 4.33-4.70 (m, 2H, 2 and 7 cycloheptatrienyl), 7.10 (m, IH, methine) ; mass spectrum 218 (M+) , Anal . Calcd. for CiyHji^: C, 93.54; H, 6.46. Found: C, 93.39; H, 6.48. The sodium salt of 3-naphthaldehyde tosylhydrazone was pyrolyzed and worked up in a manner analogous to the one above to yield 78% 2(2 ,4 ,6-cycloheptatrienl-yl)naphthalene and 6% of the aldazine.

PAGE 77

64 Photolysis of the Sodium Salt of 4 , 5-Benzotropone Tosylhydrazone in Benzene . --To the pliotolysis vessel, previously described in the photolysis of 4 , 5 -benzotropone tosylhydrazone sodium salt in diglyme, containing 60 ml of benzene was added 0.10 g (0.29 mmol) of the sodium salt of 4 , 5-benzotropone tosylhydrazone. The mixture was stirred and irradiated with a 550 watt Hanovia medium pressure mercury lamp for 30 min at 30°. The nitrogen evolved was 84% of the theoretical amount. The reaction mixture was filtered and the residue was washed with ethyl ether. The total filtrate was concentrated on a rotary evaporator under reduced pressure to yield 37 mg (59%) of 2(2 ,4 ,6-cycloheptatrienl-yl)naphthalene which had all the properties of the pyrolysis product. The sodium salt of 6-naphthaldehyde was photolyzed in a manner analogous to the above salt to yield 67% of 2(2,4, 6cycloheptatrien1-yl) naphthalene . Preparation of 2-naphthaldazine . -The compound was prepared by the general procedure given in "Organic Synthesis"' and it had mp 235-236°. Lit. 231°.'° Pyrolysis of the Sodium Salt .of g-Naphthaldehyde Tosylhydrazone in Cyclohexene . -To a 3 oz Fischer-Porter Aerosol Compatibility Tube containing 15 ml of distilled cyclohexene was added 0.40 g (I.IS mmol) of the sodium salt of 3-naphthaldehyde tosylhydrazone. The sealed tube was placed in an oil bath preheated to 135 and allowed to remain for 70 min while the temperature was raised to 145°. Upon cooling, the

PAGE 78

65 mixture was filtered and the residue was washed with ethyl ether. The total filtrate was concentrated on a rotary evaporator to yield 212 mg of an oil. Analysis by vpc indicated there were three volatile components which were separated by preparative vpc (column A): (cyclohexene-3-yl) -2naphthylmethane (341) retention time 30 min, syn 2(7-norcaryl)naphthalene(26%) retention time 32 min, and anti 2(7-norcaryl)naphthalene-' (40%) retention time 38 min. There was about 101 aldazine formed in the reaction but it did not appear in the vpc chromatogram, thus the above data only refer to volatile hydrocarbon products. (cyclohexene-3-yl) 2-naphthylmethane showed: ir (film) 3050, 3020, 2920, 2850, 1600, 1500, 1445, 855, 820, 750 cm'"^; nmr (CDCI3) t 2.00-2.90 (m, 7H , naphthyl) , 4.40 (s , 2H, vinyl), 7.35 (s, 2H, methylene), 7.80-9.20 (m, 7H, cyclohexyl) ; mass spectrum 222 (M+). Anal . Calcd. for Ci7Hia: C, 91.84; H, 8.16. Found: C, 91.64; H, 8.20. syn -2(7-norcaryl)naphthalene showed: mp 68-70°; ir (melt) 3050, 3000, 2930, 2860, 1600, 1500, 1445, 830, 750, 725 cm"-*-; nmr (CDCI3) x 2.00-2.80 (m , 7H, naphthyl), 6.609.50 (m, IIH, norcaryl) ; mass spectrum 222 (M+) . Anal . Calcd. for C17H18: C, 91.84; H, 8.16. Found: C, 91.83; H, 8.20. anti -2(7-norcaryl)naphthalene showed: ir (film) 3050, 3005, 2925, 2850, 1625, 1598, 1505, 1442, 809, 778, 740 cm'-*"; nmr (CDCI3) t 2.20-3.10 (m, 7H, naphthyl), 7.90-9.30 (m, IIH, norcaryl) ; mass spectrum 222 (M+) .

PAGE 79

66 Anal. Calcd. for CiyHis: C, 91.84; H, 8.16. Found: C, 91.56; H, 8.09. Pyrolysis of the Sodium Salt of 4 ,5-Benzotropone Tosylhydrazone in Cyclohexene at 85 .--Into a 50 ml three-neck, round-bottom flask fitted with a thermometer, condenser with a drying tube, and a solid-addition funnel, was placed 25 ml of distilled cyclohexene. The cyclohexene was heated to reflux and O.lOg (0.29 mmol) of the sodium salt of 4,5benzotropone tosylhydrazone was added by means of the solidaddition funnel. The mixture was heated at reflux for 4 hr and then stopped even though the reaction was incomplete as indicated by the presence of an orange color in the reaction flask. The reaction mixture was filtered and the residue was washed with ethyl ether. The total filtrate was concentrated on a rotary evaporator under reduced pressure. The remaining oil was subjected to vpc analysis (column A) which indicated five volatile components. The ratio of products was two unknown products (661) , the naphthylcarbene-cyclohexene insertion product (7%) , syn -norcarane (121) , and anti norcarane (15%) . While neither of the unknown products has been completely characterized, one appears to be the adduct that v^7ould result from the addition of benzocycloheptatrienylidene to cyclohexene. The compound showed ir (film) 3010, 2920, 2580, 1585, 1480, 1440, 795, 745, 700 cm'-*-; nmr (CDCI3) T 2.75 (d, 4H) , 3.25 (d, IH) , 3.60 (d,lH), 4.50 (d, IH) , 4.66 (d, IH) , 8.70 (broad m, lOH) ; mass spectrum 222 (M+) ;

PAGE 80

67 uv (951 C2H5OH) 240 my (log e = 4.2), 265 (3.5). Anal . No satisfactory analysis was obtained (probably due to insufficient sample size). Pyrolysis of the Sodium Salt of 4 , 5-Benzotropone Tosylhydrazone in Cyclohexene at 140° .-Into a 3 oz Fischer-Porter Aerosol Compatibility Tube containing 15 ml of distilled cyclohexene was introduced 0.40 g (1.15 mmol) of the sodium salt of 4 ,5-benzotropone tosylhydrazone. The tube was immersed in an oil bath preheated to 135° and allowed to remain for 70 min while the temperature was increased to 145°. The mixture was allowed to cool and was filtered. The residue was washed with ethyl ether and the total filtrate concentrated on a rotary evaporator under reduced pressure to yield 150 mg (58%) of products. Analysis by vpc (column A) indicated the same five products as the 83 pyrolysis: the two "unknown" products, the insertion product, the syn -norcarane , and the anti -norcarane . The amount of unrearranged-carbene products relative to the naphthyl products was approximately the same as in the 83 pyrolysis. Pyrolysis of the Sodium Salt of 4 , 5-Benzotropone Tosylhydrazone in Cyclohexene-Diglyme at 82 .--Into a 3 oz Fischer-Porter Aerosol Compatibility Tube containing 17.5 g (0.13 mol) of dry diglyme and 4.95 g (0.06 mol) of distilled cyclohexene was introduced 0.10 g (0.29 mmol) of the sodium salt of 4 ,5-benzotropone tosylhydrazone. The sealed tube was placed in an oil bath preheated to 82 and allowed to

PAGE 81

68 stir for 11 hr. The reaction mixture was allowed to cool and poured into water (100 ml) , extracted witli pentane (three 35 ml portions), and the combined pentane extracts washed with water (three 75 ml portions) . The pentane solution was dried (MgSO^) and concentrated on a rotary evaporator under reduced pressure to yield 105 mg of products. Analysis of volatile hydrocarbon products by vpc (column A) indicated the "unknown" products, the insertion product, the syn -norcarane , and the anti -norcarane . There were also diglyme insertion products amounting to 11% of the total products. Interfering materials prevented accurate analysis of the mixture. Photolysis of 4 , 5Benzotropone Tosylhydrazone Sodium Salt in Cyclohexene . -To the photolysis vessel, previously described in the photolysis of 4 , 5-benzotropone tosylhydrazone sodium salt in diglyme, containing 60 ml of distilled cyclohexene was added 200 mg (0.58 mmol) of the sodium salt of 4,5benzotropone tosylhydrazone. The mixture was stirred and irradiated with a 550 watt Hanovia medium pressure mercury lamp for 1.25 hr. The photosylate was filtered, and the filtrate was concentrated on a rotary evaporator at reduced pressure to yield 67 mg of product. Analysis by vpc (column A) of the volatile hydrocarbon products indicated the "unknown" products, a new, additional unknown product, the insertion product, the syn -norcarane , and the anti -norcarane . The amount of unrearranged-carbene products relative to the naphthyl products appeared to be greater than in the pyrolyses,

PAGE 82

69 Pyrolysis of the Sodium Salt of 4 , 5-Benzotropone Tosylhydrazone in Cyclohexane . In a 3 oz Fischer-Porter Aerosol Compatibility Tube containing 20 ml of cyclohexane was placed 0.40 g (1.1 mmol) of the sodium salt of 4 , 5-benzotropone tosylhydrazone . The sealed tube was placed in an oil bath which was preheated to 115 and allowed to remain for 1.25 hr. Nitrogen evolution was 1001 of the theoretical amount. The reaction mixture was allowed to cool and then filtered. The residue was washed with ethyl ether and the total filtrate concentrated on a rotary evaporator under reduced pressure. The remaining yellow oil was chromatographed on alumina, eluted with pentane, yielding 131 mg (51%) of (3-naphthyl) cyclohexylmethane . The nmr spectrum of the crude material indicated almost exclusively adduct and cyclohexane. An analytical sample of the compound was purified by preparative vpc (column A at 222 ) retention time 32 min and had the following properties: ir (film) 3042, 3010, 2920, 2845, 1601, 1445, 815, 795 cm'"'-; nmr (CDCI3) t 2.10-2.90 (m, 7H, naphthyl) 7.35 (d, 2H, J=7, benzylic), 8.10-9.20 (m, IIH, cyclohexyl) ; mass spectrum 224 (M+) , 141 (C11H9), 83 (CeHu). Anal. Calcd. for C17H20: C,.91.01; H, 8.99. Found: C, 90.84; H, 8.94. Reaction of the Sodium Salt of 4 , 5-Benzotropone Tosylhydrazone with Dimethyl Fumarate at 105° . --Into a 3 oz Fischer-Porter Aerosol Compatibility Tube containing 20 ml of benzene were placed 0.20 g (0.58 mmol) of the sodium salt of 4 , 5-benzotropone tosylhydrazone and 83 mg (0.58 mmol) of

PAGE 83

70 dimethyl fumarate (previously waslied with saturated aqueous sodium bicarbonate and then rccrystallized from chloroform). The sealed tube was placed in an oil bath preheated to 105 and allowed to remain for 5 hr. Upon cooling, the mixture was filtered. The residue was waslied with ethyl ether and the total filtrate concentrated on a rotary evaporator under reduced pressure to yield 76 mg (63%) of 2(2 , 4 ,6-cycloheptatrienl-yl)naphthalene . The nmr spectrum of the crude reaction material indicated the above mentioned adduct and a considerable amount of unreacted dimethyl fumarate. Comparative tic on silica gel eluted with 50/50 ethyl ether-pentane of the known pyrazoline ( vide infra ) and this reaction mixture indicated no pyrazoline was present in the latter. Reaction of the Sodium Salt of g-Naphthaldehyde Tosylhydrazone with Dimethyl Fumarate at 105 .--To a 3 oz FischerPorter Aerosol Compatibility Tube containing 20 ml of benzene was added 0.20 g (0.58 mmol) of the sodium salt of 3-naphthaldehyde tosylhydrazone and 83 mg (0.58 mmol) of dimethyl fumarate (previously washed with saturated aqueous sodium bicarbonate and recrystallized from chloroform) . The sealed tube was placed in an oil bath preheated to 105° and was allowed to remain there for 5 hr. Upon cooling, the reaction mixture was filtered. The residue was washed with ethyl ether and the total filtrate was concentrated on a rotary evaporator yielding 90 mg (50%) of an almost colorless, very viscous oil. The oil was not stable to ordinary distillation conditions and even though column chromatography on silica gel

PAGE 84

71 effected some purification, it was not sufficient to provide a sample which gave a good elemental analysis [e.g. Calcd. N, 8.97; found: N, 8.31). The other physical data were all consistent with the 3 ,4-dicarbomethoxy5-naphthyl-2-pyrazoline structure: ir (film) 3440 (N-H) , 3050, 2945, 1730 (unconjugated ester carbonyl) , 1700 (conjugated ester carbonyl) , 1555 (C=N), 1440, 1200, 1015, 855, 815, and 750 cm""^; nmr (CDCI3) T 2.10-2.60 (m, 7H, naphthyl) , 3.24 (broad s, IH , nitrogen), 4.60 (broad d, IH, J=9, benzylic) , 5.92 (d, IH, J=9, a-carbomethoxy) , 6.18 (s , 3H, conjugated ester methyl), 6.24 (s , 3H, unconjugated ester methyl)*; mass spectrum no (M+) , 284 (M 28); uv (951 C2H5OH) 224 my (log e = 4.53), 275 (3.75), 287 (3.79). Reaction of g-Naphthaldehyde Tosylhydrazone Sodium Salt with Dimethyl Fumarate at 160° . --This decomposition was carried out in the same manner as the low temperature one with the exception that the sealed tube was placed in an oil bath preheated to 105 and allowed to remain for 2 hr while the temperature was raised to 160 . Upon cooling, the mixture was filtered. The filter paper and the residue were washed with ethyl ether and the total filtrate concentrated on a rotary evaporator under reduced pressure to yield 110 mg (67%) In addition there were other resonances which changed in intensity depending upon work up and purification steps and are, therefore, not believed to originate from the pyrazoline.

PAGE 85

72 of crude trans 1 ,2-dicarbomethoxy5(2-naphtliyl) cyclopropane . The solid was recrystalli zed from abs methyl alcohol to yield 35 mg (21'b) of pink needles mp 91-93°: ir (CCIO 3028, 2975, 1750, 1435, 1300, 1170, and 909 cm""*-; nmr (CDCI3) x 2.00-2.90 (m, 7H, naphthyl) , 6.23 (s, 3H, methyl), 6.58 (s, 3H, methyl), 6.70-7.50 (m, 3H , cyclopropane ring); mass spectrum 284 (M+). Anal . Calcd. for dyHieO^: C, 71.82; H, 5.67; 0, 22.51. Found: C, 71.93; H, 5.73; 0, 22.34. Reaction of the Sodium Salt of 4 , 5 Benzotropone Tosylhydrazone with Dimethyl Fumarate at 160° . --This decomposition was carried out in the same manner as the low temperature one with the exception that the sealed tube was placed in an oil bath preheated to 105 and allowed to remain for 2 hr while the temperature was raised to 160°. Upon cooling, the mixture was filtered. The residue was washed with ethyl ether and the total filtrate concentrated on a rotary evaporator under reduced pressure. The products were not isolated, only cursorily examined by nmr spectroscopy. The materials, 2(2,4, 6cycloheptatrien-1-yl) naphthalene and 2(2,4 ,6cycloheptatrien4-yl)naphthalene , were readily identified by their characteristic spectra and determined to be present in a ratio of 1.0:1.6, respectively, as shown by integration. General Procedure for the Photolysis of Tosylhydrazone Sodium Salts in the 2-Butenes . -The photolyses were carried out in an apparatus which consisted of a Pyrex tube (20 cm x 2.5 cm) with two side arms and equipped with a small mechanical

PAGE 86

73 stirrer. One of the side arms was connected to a Dry Ice condenser with a drying tube and the other to a stopcock. The desired amount of tosylhydrazone salt was introduced into the apparatus which was consequently flushed with dry nitrogen. The apparatus was then placed in a small Dry Ice-acetone bath and the 2-butene condensed into the vessel. The vessel was then mounted vertically alongside, and as close as possible to, a 550 watt Hanovia medium pressure mercury lamp in a Pyrex cooling jacket. Both the apparatus and the lamp system were placed in a large Dewar flask containing methyl alcohol, with the temperature of the entire system being regulated by the cooling jacket of the lamp which in turn was cooled by a Dry Ice-acetone heat exchanger. The mixture was then rapidly stirred and irradiated for the desired amount of time at temperatures in the range of -15 to -30°. Photolysis of the Sodium Salt of g-Naphthaldehyde Tosylhydrazone in trans 2-Eutene . -Into the previously described photolysis apparatus was placed 0.20 g (0.58 mmol) of the sodium salt of 6-naphthaldehyde tosylhydrazone and 30 ml of trans -2-butene (C.P. Grade, typical analysis 99.3 mole I trans2-butene, 0.2 mole % butene, and 0.5 mole % cis -2butene) . The reaction mixture was irradiated for 2.0 hr at temperatures between -15 to -20°. At the end of this time, the butene was allowed to evaporate and the residue was dissolved in ethyl ether, filtered, and concentrated on a rotary evaporator under reduced pressure to yield 88 mg of a

PAGE 87

74 light yellow oil. In addition to a small amount of unknown products, the material contained three products, 1(2-naphthy 1) trans 2 , 5dimethylcyclopropane , and syn and anti -1(2naphthyl) cis 2 , 5 -dimethyl cyclopropane , as sliown by vpc analysis. The trans cyclopropane was analyzed (column A) and purified by prep vpc (column A). 1(2-naphthyl) trans 2 , 3dimethyl cyclopropane showed: ir (film) 3050, 3000, 2940, 2920, 2860, 1628, 1600, 1500, 1450, 1058, 850, 815, 745 cm'""-; nmr (CCIO x 2.10-2.95 (multiplet, 7H, naphthyl protons), 8.15 (multiplet, IH, proton on C-1), 8.35-9.65 (multiplet, 8H, protons on C-2 and C-3 and methyl protons); mass spectrum m/e 196 (M+) . Anal. Calcd. for CisHig: C, 91.78; H, 8.22. Found: C, 91.72; H, 8.31. Photolysis of Benzaldehyde Tosylhydrazone Sodium Salt in t_rans2-Butene. In the described photolysis apparatus was placed 0.40 g (1.35 mmol) of the sodium salt of benzaldehyde tosylhydrazone,^^ then 25 ml of trans 2-butene (C.P. Grade) was condensed into the apparatus. The reaction mixture was irradiated for 2.5 hr at temp between -15 and -20 . At the end of this time, the butene was allowed to evaporate and the residue was dissolved in ethyl ether, filtered, and concentrated on a rotary evaporator under reduced pressure to yield 0.09 g (45%) of crude 1 -phenyltrans 2 ,3dimethylcyclopropane . The cyclopropane was purified by preparative vpc (column A) . The nmr (CDCI3) t 2.93 (broad singlet, 5H, phenyl protons).

PAGE 88

75 8.30 (multiplet, IH, proton on C-1), 8.68-9.36 (multiplet, 8H, protons on C-2 and C-3 and methyl protons) agreed with the data of Closs."*^ Photolysis of the Sodium Salt of g-Naphthaldehyde Tosylhydrazone in cis -2-Eutene.-The photolysis of the sodium salt of g-naphthaldehyde tosylhydrazone in cis -2-butene was carried out in a manner similar to that of trans -2-butene . In this case, 0.29 g (0.84 mmol) of the sodium salt was decomposed in 30 ml cis 2-butene (C.P. Grade, typical analysis 99.5 mole % cis and 0.5 mole I trans 2-butene) and the photolysis was done over a period of 2.75 hr at a temp between -20° and -30°. There was produced 0.126 g of a light yellow oil. Again the products were analyzed by vpc and the same three products were present (relative yields): 1(2-naphthyl) trans -2 ,3dimethylcyclopropane (31), syn and anti -1(2-naphthyl) cis 2 ,3-dimethylcyclopropane (97%). The syn and anti -cyclopropanes were purified by preparative vpc (column A) . syn -1(2-naphthyl) cis -2 ,3-dimethylcyclopropane showed: ir (film) 3050, 3000, 2940, 2870, 1630, 1602, 1500, 1385, 1160, 890, 860, 818, 748 cm"-*-; nmr (CCIO x 2.20-3.00 (multiplet, 7H, naphthyl protons), 7.90 (multiplet, IH, proton on C-1), 8.50-9.50 (multiplet, 8H, protons on C-2 and C-3 and methyl protons); mass spectrum m/e 196 (M+). Anal. Calcd. for CisHig: C, 91.78; H, 8.22. Found: C, 91.55; H, 8.15.

PAGE 89

76 anti 1(2-naphthyl) cis 2 , 3dimethyl cyclopropane showed: ir (film) 3050, 3000, 2940, 2860, 1628, 1600, 1500, 1082, 850, 815, 782, 745 cm'"^; nmr (CCIO t 2.20-3.10 (multiplet, 7H, naphthyl protons), 8.50-9.15 (multiplet, 9H, protons on C-1, 2, and 3 and methyl protons); mass spectrum m/e 196 (M+) . Anal . Calcd. for CisHie: C, 91.78; H, 8.22. Found: C, 91.63; H, 8.29. Pyrolysis of P-Naphthaldehyde Tosylhydrazone Sodium Salt in cis2-Butene . In a 0.5 oz reusable Fischer-Porter Carius Tube was placed 0.10 g (0.29 mmol) of the sodium salt of 6-naphthaldehyde tosylhydrazone. The tube was fitted with a Dry Ice condenser by means of a rubber stopper, flushed with dry nitrogen, and placed in a Dry Ice-acetone bath. Then 10 ml of trans 2-butene was condensed into the Carius Tube; the condenser was quickly removed and the tube sealed. The tube was then immersed in an oil bath (preheated to 105°) for 1 hr, At the end of this time the tube was allowed to cool to room temperature and then cooled further in a Dry Ice-acetone bath. The tube was opened, the bath removed, and the butene allowed to evaporate. The residue was dissolved in ethyl ether, concentrated on a rotary evaporator at reduced pressure, and analyzed by vpc (column C) to yield the following data (relative yields): 1(2-naphthyl) trans -2 ,3 -dimethyl cyclopropane (-10%),* and syn and anti -1(2-naphthyl) cis 2 , 3-dimethylcyclopropane (~90l). A large amount of interfering material makes this data only approximate.

PAGE 90

77 Photolysis of 4 , 5 -Benzotropone Tosylhydrazone Sodium Salt in cis2-Butene . -The photolysis o£ the sodium salt of 4 , 5-benzotropone tosylhydrazone was effected using 0.29 g (0.84 mmol) of the salt in 30 ml of cis_2-butene (C.P. Grade) and by irradiating for 2.3 hr at -25°. Work up yielded 0.061 g of product and vpc analysis (column C) gave the following data: 1(2-naphthyl) trans 2 ,3-dimethylcyclopropane was a minor product (4^); the major products were syn -1(2-naphthyl) cis 2 ,5dimethyl cyclopropane and anti-1(2-naphthyl) -cis_2 ,3dimethylcyclopropane (96%). Photolysis of 4 ,5-Benzotropone Tosylhydrazone Sodium Salt in trans -2But ene. --The photolysis of the sodium salt of 4 , 5-benzotropone tosylhydrazone was effected as previously described using 0.20 g (0.58 mmol) of the salt in 30 ml of trans -2-butene (C.P. Grade) and by irradiating for 2.3 hr at -25 . Work up yielded 0.046 g of a yellow oil. Vpc analysis (column C) gave the following data: 1(2-naphthyl) trans 2 ,3-dimethylcyclopropane was the major product along with a small amount of the epimeric 1(2-naphthyl) cis 2 ,3-dimethylcyclopropanes . Pyrolysis of the Sodium Salt of 4 , 5-Benzotropone Tosylhydrazone in cis -2-Butene . In a 0.5 oz reusable FischerPorter Carius Tube was placed 0.10 g (0.29 mmol) of the sodium salt of 4 , 5-benzotropone tosylhydrazone. The tube was fitted with a Dry Ice condenser by means of a rubber stopper, flushed with dry nitrogen, and placed in a Dry Ice-acetone

PAGE 91

78 bath. Then, 10 ml of cis 2-butcne (research grade) was condensed into the Carius Tube; the condenser was quickly removed and the tube sealed. The tube was allowed to warm to room temperature and then immersed in an oil bath (preheated to 105°) for 3.5 hr. At the end of this time the tube was allowed to cool to room temperature and then cooled further in a Dry Ice-acetone bath. The tube was opened, the bath removed, and the butene allowed to evaporate. The residue was dissolved in ethyl ether, concentrated on a rotary evaporator under reduced pressure to yield 7 mg of a yellow oil which was analyzed by vpc (column C) . The reaction material contained the syn and anti 1(2-naphthyl) cis 2 , 3 -dime thy 1cyclopropanes (93%), 1(2-naphthyl) trans -2 ,3-dimethylcyclopropane (7%), and some extraneous material. In all cases following the percentages given in parentheses do not imply total percent yield but relative yields of the cis and transcyclopropanes and were done on column C unless otherwise stated. Pyrolysis of the Sodium Salt of 4 , 5-Benzotropone Tosylhydrazone in cis 2-Butene and Benzene .-The reaction was carried out as in the above experiment except that 0.08 g (0.23 mmol) of the sodium salt, 7.0 ml (79 mmol) of benzene, and 2.5 ml (28 mmol) of research grade cis 2-butene were heated at 105 for 3 hr. The work up was the same yielding 10 mg of a yellow oil which contained the cyclopropanes in the following relative yield: syn and anti -1(2-naphthyl) cis -2,3-dimethylcyclopropanes (94%) and 1(2-naphthyl) trans -

PAGE 92

79 2 ,3-dimethylcyclopropane (6%). The 2-naphthylcarbene-benzene adduct was also produced as evidenced by vpc. Pyrolysis of the Sodium Salt of 4 , S-Benzotropone Tosylhydrazone in cis -2-Butene and Perf luorocyclobutane .-Again, the reaction was carried out in a manner analogous to that of the cis 2-butene reaction. In this series two reactions were carried out: the first with cis -2-butene and perf luorocyclobutane in a 1:2 molar ratio, respectively, and the second with a 1:6 molar ratio, respectively. In the first reaction, 0.10 g (0.29 mmol) of the sodium salt of benzotropone tosylhydrazone , 8 . ml (60 mmol) of perf luorocyclobutane , and 2.5 ml (30 mmol) of research grade cis 2-butene were heated at 105° for 5 hr. The work up yielded 7.5 mg of a yellow oil which, when analyzed by vpc, was shown to contain the syn and anti 1(2-naphthyl) cis 2 , 3dimethyl cyclopropanes (94%) and 1(2-naphthyl) trans -2, 3 -dime thy 1cyclopropane (6%). In the second case, 0.10 g (0.29 mmol) of the sodium salt of benzotropone tosylhydrazone, 9.0 ml (67 mmol) of perf luorocyclobutane , and 1.0 ml (11 mmol) of research grade cis 2-butene were heated at 108 for 4 hr . The work up yielded 7.0 mg of a yellow oil which gave the same cis to trans cyclopropane ratio as the first experiment (namely, cis 941 and trans 6%) .

PAGE 93

80 Pyrolysis of the Sodium Salt of 4 , 5Bcnzotropone TosylIn'drazone in cis 2-Butene and Oxygen .-This reaction, too, was carried out like the cis 2-butcne reaction with the following exceptions: 0.10 g (0.29 mmol) of the sodium salt of bcnzotropone tosy Ihydrazone and 12 ml of research grade cis 2-butene were placed in the Carius Tube and then oxygen was bubbled through the mixture for 1 hr . The reaction mixture was heated at 108 for 3.6 hr and the work up yielded 21 mg of a yellow oil which contained the two epimeric cis -cyclopropanes (96°6) and the trans -cyclopropane (41). A control experiment, done exactly the same as the above experiment only bubbling nitrogen through the mixture, gave the typical 941 cis material and 6% trans . Pyrolysis of the Sodium Salt of 4 , 5 Benzotropone Tosylhydrazone in cis 2-Butene and trans , trans 2 ,4-hexadiene. -Again, the reaction was carried out in a manner analogous to that of the cis 2-butene reaction. In this series two separate reactions were carried out: the first with cis 2-butene and trans , trans 2 ,4-hexadiene in a 1.0:1.5 molar ratio, respectively, and the second with a 1.0:2.1 molar ratio, respectively, In the first reaction, 0.09 g (0.26 mmol) of the sodium salt of benzotropone tosylhydrazone , 7.0 ml (60 mmol) of trans , trans 2 ,4-hexadiene , and 3.6 ml (40 mmol) of research grade cis 2-butene were heated at 105° for 4.5 hr. Work up yielded 24 mg of product (again, a yellow oil) which had the cis -cyclopropanes produced in 991 yield relative to the trans-

PAGE 94

81 cyclopropane in 1%. However, what is assumed to be the products from the diene account for £a. 901 of the total products . In the second case, 0.10 g (0.29 mmol) of the sodium salt of benzotropone tosylhydrazone , 7.3 ml (63 mmol) of trans , trans 2 , 4-hexadiene , and 2.7 ml (30 mmol) of research grade cis 2-butene were heated at 105° for 5 hr. Work up yielded 18 mg of a yellow oil which proved to be almost exclusively the assumed diene products (there was too little of the butene adducts to provide reliable results). Sensitized Photolysis of the Sodium Salt of 4,5-Benzotropone Tosylliydrazone in Benzene . --In the photolysis vessel, described previously, was placed 1.0 g (5.1 mmol) xanthen-9one , 0.10 g (0.29 mmol) of the sodium salt of benzotropone tosylhydrazone, and 40 ml of benzene. The stirred mixture was photolyzed at 30 for 1 hr. The solvent was evaporated and the residue was chromatographed on grade III U'oelm basic alumina. There were no hydrocarbon products isolated from the reaction mixture. Thermal Stability of the syn and anti -1(2-naphthyl) cis 2 , 3-dimethylcyclopropanes . --A sample of the cis -cyclopropanes was preparatively chromatographed (column A at 230 ) and then analyzed for any isomeri zation during chromatography. There was no trace of trans -cyclopropane as evidenced by vpc analysis (column B) .

PAGE 95

82 Stability of cis 2Butene under Reaction Conditions .-As in a typical run, 0.10 g (0.29 nunol) of the sodium salt of benzotropone tosy lliydrazone and 9.0 ml cis 2-butenc (research grade) were heated in a Carius Tube at 105 for 4.25 hr. Analysis of the cis 2-butene (column C) after the usual reaction period indicated that less than 1 percent isomerization liad occurred during the reaction. 2 -Methyl -4 ,5-Benzotropone . -The ketone was prepared by the method of Heilbronner and had a mp 68-69 . Lit. 70°.^^ 2-Methyl-4 , 5-Benzotropone Tosylhydrazone . --In a typical run, 2.00 g (0.012 mol) of 2-methyl -4 , 5-benzotropone , 2.18 g (0.012 mol) of £^toluenesulfonylhydrazide , and 5 drops of coned sulfuric acid were placed in 40 ml of abs ethyl alcohol and heated at reflux for 1 hr. Upon cooling, 2-methyl-4 , 5benzotropone tosylhydrazone precipitated from solution and was isolated by filtration. Recrystallization from abs ethyl alcohol yielded 2.4 g (601) of the product as yellow-gold plates mp 182-184°: ir (KBr) 3180, 1630, 1590, 1555, 1420, 1385, 1330, 1165, 1022, 915, 805, 760, 680, and 580 cm'-""; nmr (CDCI3) t 2.00-3.00 (A2B2, 4H, J=8, aromatic), 2.78 (s, 4H, benzo) , 3.05-3.70 (m, 3H, cycloheptatrienyl) , 7.62 (s , 3H, £-methyl), 7.88 (d, 3H , cycloheptatrienyl methyl); mass spectrum 339 (M+) . Anal . Calcd. for C19H18N2O2S: C, 67.43; H, 5.36; N, 8.28. Found: C, 67.39; H, 5.33; N, 8.40.

PAGE 96

83 Pyrolysis of the Sodium Salt of 2 -Methyl -4 , 5-Benzotropone Tosylhydrazone in Benzene . --To a Fischer-Porter Aerosol Compatibility Tube containing 35 ml of benzene was added 216 mg (0.60 mmol) of the sodium salt of 2-methyl4 , 5-benzotropone tosylhydrazone. The sealed tube was placed in an oil bath preheated to 130° for 20 min. By the end of this time the yellow mixture had changed to a very light yellow-green one. The mixture was allowed to cool and was filtered. The residue was washed with ethyl ether and the total filtrate concentrated on a rotary evaporator under reduced pressure to yield 80 mg of products. Analysis of the product mixture by nmr indicated the presence of 2-vinylnaphthalene and 2(2,4,6cycloheptatrien1-yl) -3-methylnaphthalene (comparisons were made using commercially available 2-vinylnaphthalene from Aldrich Chemical Co. and 2(2 ,4 ,6-cycloheptatrien-l-yl) 3methylnaphthalene from the pyrolysis of the sodium salt of the tosylhydrazone of 3-methyl2-naphthaldehyde in benzene). Analysis by vpc (column B) indicated three products in the reaction mixture (the known ones were established by comparison with kno^>m samples): 1) retention time 1 min, 2-vinylnaphthalene (72%); 2) retention time 10 min, unknown compound (6%); 3) retention time 14 min, 2(2 ,4 , 6-cycloheptatrien1yl) -3-methylnaphthalene (221). Products 1^ and 3^ were isolated in 44% and 20% overall yield, respectively. 2(2 ,4 ,6-cycloheptatrien-1-yl) -3-methylnaphthalene (sublimed at 90 and 0.1 mm Hg) had the following properties: mp 65-68°; ir (KBr) 3055, 3010, 1600, 1500, 1440, 1220, 1010, 890, 745, 700, and

PAGE 97

480 cm'^; nmr (CDCI3) x 2.00-2.90 (m, 6H, naphthyl), 3.25 (broad t, 2H, 4 and 5 cycloheptatrienyl) , 3.50-3.90 (complex pattern, 2H, 3 and 6 cycloheptatrienyl), 4.35-4.76 (complex pattern, 2H , 2 and 7 cycloheptatrienyl), 6.95 (broad t, IH, methine) , 7.65 (s, 3H , methyl); mass spectrum 232 (M+) . Anal . Calcd. for CisHie: C, 93.06; H, 6.94. Found: C, 92.85; H, 7.00. The sodium salt of 3-methyl2-naphthaldehyde tosylhydrazone was pyrolyzed in the same manner as the above salt except that 40 ml of benzene was used and the temperature was 145°. The products were aldazine (421) and 2(2,4, 6cycloheptatrien1-yl) 3-methylnaphthalene (39%) . 2,5-Benzotropone .--2 , 3-Benzotropone was prepared by the method of Collington and Jones, ^^ and the ir spectrum of the product was in complete agreement with that reported in the literature. 2 , 3-Benzotropone Tosylhydrazone . --To 1.0 g (0.6 mol) of 2 ,3-benzotropone in 10 ml of 95% ethyl alcohol was added 1.1 g (0.6 mol) of £toluenesulf onylhydrazide in 25 ml of 95% ethyl alcohol and one drop of acetic acid. The solution was heated at reflux for 75 min, and then the reaction mixture was allowed to cool. The solvent was removed on a rotary evaporator under reduced pressure and the resinous residue was chromatographed on silica gel eluting with methylene chloride. The thus purified tosylhydrazone, still a viscous resin, was obtained 54% yield and had the following properties:

PAGE 98

85 ir (film) 3210, 1582, 1400, 1385, 1340, 1320, 1160, 1080, 1020, 810, 800, 760, 700, 680, 655 cm'-'-; nmr (CDCI3) x 1.60 (broad s, IH, proton on nitrogen), 2.00-4.00 (multiplet, 12H, protons on benzo ring, cycloheptatriene ring, and the aromatic toluyl protons whose A2B2 quarter could be picked out of the multiplet, 7.70 (s , 3H, methyl). The tosylhydrazone was never obtained in a state of purity which gave a good analysis (e.g. Calcd. C, 65.38; H, 5.16. Found: C, 65.10; H, 5.84). Mass spectrum had no parent peak, (none of any significance above m/e 156 even at 10 eV) 156 (100), 141 (9), 91 (17), 64 (28). The resinous tosylhydrazone was converted to its sodium salt cleanly by the outlined procedure (this was the only water-sensitive salt made). Pyrolysis of the Sodium Salt of 2 ,3-Benzotropone Tosylhydrazone in Benzene . --In a 3 oz Fischer-Porter Aerosol Compatibility Tube containing 35 ml of benzene (deoxygenated by bubbling in nitrogen for several min) was placed 0.28 g (0.81 mmol) of the sodium salt of 2 ,3-benzotropone tosylhydrazone. The tube was then placed in an oil bath which was preheated to 120° and allowed to remain there for 35 min. The solution was allowed to cool, filtered, and concentrated to a yellow oil on a rotary evaporator under reduced pressure. Chromatography of the residue on activity grade III basic Woelm alumina yielded 0.045 g (26%) of a colorless oil which was identified as 1(2 ,4 , 6-cycloheptatrien-l-yl)naphthalene.

PAGE 99

86 The properties of the compound are as follows: ir (film) 3050, 3020, 1600, 1510, 1395, 800, 780, 745, 705 cm""^; nmr (CDCI3) T 2.20 (multiplet, 3H , a-protons of naphthyl system) 2.55 (multiplet, 4H, 6-protons of naphthyl system), 3.30 (triplet, 2H, J = 3 Hz, 4 and 5 protons on cycloheptatriene) , 3.70 (multiplet, 2H , 3 and 6 protons on cycloheptatriene), 4.50 (quartet, 2H , J=5.5 Hz, 2 and 7 protons on cycloheptatriene), and 6.60 (multiplet, IH, 7 proton on cycloheptatriene); mass spectrum m/e 218 (M+) , 127 (C10H7), 91 (C7H7). Anal . Calcd. for CnHmi C, 93.54; H, 6.46. Found: C, 93.32; H, 6.61.

PAGE 100

LIST OF REFERENCES 1. L. Skattebol, Tetrahedron, 23^, 1107 (1967). . 2. P. Ashkenazi , S. Lupan, A, Schwarz, and M. Cais, Tetrahedron Lett., 817 (1969). 3. G. Frater and 0. P. Strausz, J. Amer. Chem. Soo., 9^, 6654 (1970); D. E. Thornton, R. K. Gosavi, and 0. P. Strausz, ibid., 92_, 1768 (1970); I. G. Csizmadia, J. Font, and 0. P. Strausz, ibid., £Cl_, 7360 (1968). 4. R. L. Russell and F. S. Rowland, ibid., 9^, 7508 (1970). 5. G. G. Vander Stouw, Diss. Abst., 25, 6974 (1965) and Chem. Abst., 6_3, 13126b (1965) un3e"r the direction of H. Shechter. 6. G. G. Vander Stouw, A. R. Kraska, and H. Shechter, J. Amer. Chem. Soc, 94, 1655 (1972). 7. R. C. Joines, A. B. Turner, and W. M. Jones, ibid., 91 , 7754 (1969). 8. W. M. Jones and C. L. Ennis, ibid., 91_, 6391 (1969). 9. J. A. Myers, R. C. Joines, and W. M. Jones, ibid., 92 , 4740 (1970). 10. C. Wentrup and K. Wilczek, Belv. Chim. Acta, 5^, 1459 (1970). 11. P. 0. Schissel, M. E. Kent, D. J. McAdoo, and E. Hedaya, J. Amer. Chem. Soa., 92_, 2147 (1970). 12. E. Hedaya and M. E. Kent, ibid., £3, 3285 (1971). 13. W. J. Baron, M. Jones, Jr., and P. P. Caspar, ibid., 92 , 4739 (1970). 14. T. Mitsuhashi and W. M. Jones, ibid., 94, 677 (1972). 15. P. H. Gebert, Ph.D. Dissertation, University of Florida, Gainesville, Florida, 1972, under the direction of W. M. Jones. 87

PAGE 101

R. Huisgen, D. Vossins, and M. Appl, Chem. Ber.^ 91_, 1 (1958); R. Huisgen and M. Appl, ibid., 12 (1958). R. A. Abramovitch and B. A. Davis, Chem. Rev., 6_4, 149 (1964) and references therein. R. A. Odum and M. Brenner, J. Amer. Chem. Soc, 88 , 2074 (1966). J. I. G. Cadogan, Quart. Rev. (London), !!_, Ill (1968) and references therein. R. J. Sundberg and S. R. Suter, J. Org. Chem., 35, 827 (1970). ~~ R. J. Sundberg, W. G. Adams, R. H. Smith, and D. E. Blackburn, Tetrahedron Lett,, 111 (1968). R. J. Sundberg, B. P. Das, and R. H. Smith, Jr., J. Amer. Chem. Soc, 9^, 658 (1969). J. S. Splitter and M. Calvin, Tetrahedron Lett., 1445 (1968). W. von E. Doering and R. A. Odum, Tetrahedron, 11, 81 (1966). R. J. Sundberg, S. R. Suter, and M. Brenner, J. Amer. Chem. Soc, 9_4, 513 (1972) and references therein. W. D. Crow and C. Wentrup, Tetrahedron Lett., 6149 (1968) C. Wentrup, Tetrahedron, 26_, 4969 (1970). W. D. Crow and C. Wentrup, Chem. Commun., 1387 (1969). C. Wentrup, Tetrahedron, 11_, 367 (1971) and references therein. P. A. S. Smith in "Nitrenes," W. Lwowski , Ed., Interscience Publishers, New York, N.Y., 1970, Chapter 4. S. Murahashi, I. Moritani, and M. Nishino, Tetrahedron, 11_, 5131 (1971); I. Moritani, S. Murahashi, M. Nishino, Y. Yamamoto, K. Itoh, and N. Mataya, J. Amer. Chem. Soa., 8£, 1259 (1967); S. Murahashi, I. Moritani, and M. Nishino, ibid., 89_, 1257 (1967). W. Kirmse, B, G. von Bulow, and H. Schepp, Justus Liebegs Ann. Chem., 6_91_, 41 (1966). J. Thiele and E. Weitz, ibid., 1>11 , 1 (1910).

PAGE 102

89 34. G. L. Gloss, J. Amer. Chem. Soo., 8^, 3796 (1963). 35. G. M. Kaufman, J. A. Smith, G. G. Vander Stouw, and H. Shechter, ibid., 8^, 935 (1965). 36. L. Cagliote and M. Magi, Tetrahedron, 19^, 1127 (1963). 37. C. Jutz and F. Voithenleitner , Chem. Ber., 9J_, 29 (1964), 38. H. Nozaki, R. Noyori, and K. Sisido, Tetrahedron, 20, 1125 (1964). 39. R. M. Silverstein and G. C. Bassler, "Spectrometric Identification of Organic Compounds," 2nd ed. , John Wiley and Sons, Inc., New York, N.Y., 1967, p. 83. 40. S. Pietra and C. Trinchera, Gazz. Chim. Ital., 8_5_, 1705 (1905); Chem. Abstr., 5_0, 10038b (1956). 41. A. H. Blatt, Ed., "Organic Synthesis," Collective Vol. 2 John Wiley and Sons, Inc., New York, N.Y., 1943, p. 395. 42. R. Jensen and D. B. Patterson, Tetrahedron Lett., 3837 (1966). 43. G. L. Gloss and R. A. Moss, J. Amer. Chem. Soc, 86^, 4042 (1964). 44. G. L. Gloss and H. B. Klinger, ibid., 8^, 3265 (1965). 45. Von G. Wittig, H. Eggers , and P. Duffner, Justus Liebigs Ann. Chem., 619_, 10 (1958). 46. J. R. Dyer, "Applications of Absorption Spectroscopy to Organic Compounds," Prentice-Hall, Inc., Englewood Cliffs, N. J., 1965, Chapter 3. 47. A. Hassner and M. J. Michelson, J. Org. Chem., 21_, 3974 (1962). 48. W. M. Jones, J. Amer. Chem. Soc, 8J^, 5153 (1959). 49. Sadtler Ultra Violet Spectra, Vol. 52, Sadtler Research Laboratories, Philadelphia, Pa., no. 12432. 50. J. A. Moore, J. Org. Chem., 20_, 1607 (1955). 51. A. P. Ter Borg and H. Kloosterziel , Reo. Trav. Chim. Pays-Bas, 8_2, 741 (1968). 52. D. Meuche , H. Strauss, and E. Heilbronner, Helv. Chim. Acta, Al, 2220 (1958).

PAGE 103

90 E. W. Collington and G. Jones, J. Chem. Soq., C, 2656 (1969) . W. M. Jones, R. C. Joines, J. A. Myers, T. Mitsuhashi, K. E. Krajca, E. E. Waali, T. L. Davis, and A. B. Turner, submitted for publication in J. Amer. Chem. Soc. A. Streitwieser, Jr., "Molecular Orbital Theory for Organic Chemists," John Wiley and Sons, Inc., New York, N.Y. , 1961. W. Kirmse, "Carbene Chemistry," Academic Press, New York, N.Y. , 1964; J. Hine , "Divalent Carbon," The Ronald Press Co. , New York, N.Y. , 1964. G. L. Gloss in "Topics in Stereochemistry," Vol. 3, E. L. Elicl, Ed., Interscience Publishers, New York, N.Y. , 1968, pp. 193-235. R. A. Moss and U. H. Dolling, J. Amer. Chem. Soc, 9_3, 954 (1971). A. M. Trozzolo, E. Wasserman, and W. A. Yager, ibid., 87, 129 (1965). E. Ciganek, ibid., 8^, 1980 (1966). W. P. Linke, "Solubilities of Inorganic and Metal Compounds," Vol. 2, 4th ed., Ajnerican Chemical Society, Washington, D.C., 1965, pp. 1228-1239. M. Jones, Jr., and K. R. Rettig, J. Amer. Chem. Soo., 87, 4013 (1965); ibid., 8^, 4015 (1965). A. F. Frost and R. G. Pearson, "Kinetics and Mechanism," 2nd ed. , John Wiley and Sons, Inc., New York, N.Y. , 1965, p. 172. R. M. Etter, H. S. Skovronek, and P. S. Skell, J. Amer. Chem. Soc, 8j^, 1008 (1959). R. Srinivasan, ibid., 9_1, 6250 (1969). E. J. York, W. Dittmar, J. R. Stevenson, and R. G. Bergman, ibid., 94, 2882 (1972). G. L. Gloss and L. E. Gloss, ibid., 85^, 99 (1963). R. Hoffmann, G. D. Ziess, and G. W. Van Dine, ibid., 90, 1485 (1968). H. E. Zimmerman and L. R. Sousa, ibid., 94, 834 (1972).

PAGE 104

91 70. G. Naville, H. Strauss, and E. Heilbronner, Helv. Chem. Aata, 43, 1221 (1960). 71. G. L. Gloss, L. E. Gloss, and W. A, Boll, J. Amer. Chem. Soo., 85_, 3796 (1963). 72. R. Hoffmann, ibid., 90, 1475 (1968). 73. W. Kirmse, "Carbene Ghemistry," 2nd ed., Academic Press, New York, N.Y., 1971, Ghapter 8. 74. R. Gleiter and R. Hoffmann, J. Amer. Chem. Soc, 90, 5457 (1968). —

PAGE 105

BIOGIUPIIICAL SKETCH Kenneth Edward Krajca was born April 1, 1944, at Wichita Palls, Texas. In June, 1962, he was graduated from Wichita Palls Senior High School. In June, 1967, he received the degree of Bachelor of Science with a major in Chemistry from Midwestern University. In 1967 he enrolled in the Graduate School of the University of Florida. He worked as a teaching assistant in the Department of Chemistry until June, 1969, and as a research assistant until the present time while he has pursued his work toward the degree of Doctor of Philosophy. Kenneth Edward Krajca is married to the former Sherry Lee Walters and is the father of one child, Laura. He is a member of the American Chemical Society. 92

PAGE 106

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. \ ^ -^William M. Jones , Chairman Professor of Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Merle A. Battiste Professor of Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. William R. Dolbier, 3x.j' Associate Professor of Chemistry I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Willis B. Person Professor of Chemistry

PAGE 107

I certify that I have read this study and that in ray opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. ilip Bifcon PTT Associate Professor of Mathematics This dissertation was submitted to the Department of Chemistry in the College of Arts and Sciences and to the Craduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. August, 1972 Dean , Graduate School

PAGE 108

n% "'/ DM18. 17 897.