(4,5-c) Furotropylidene--a ten-pi-electron carbene


Material Information

(4,5-c) Furotropylidene--a ten-pi-electron carbene
Furotropylidene-a ten-pi-electron carbene
Physical Description:
vi, 73 leaves. : illus. ; 28 cm.
Ledford, Thomas Howard, 1942-
Publication Date:


Subjects / Keywords:
Carbenes (Methylene compounds)   ( lcsh )
Chemistry thesis Ph. D
Dissertations, Academic -- Chemistry -- UF
bibliography   ( marcgt )
non-fiction   ( marcgt )


Thesis -- University of Florida.
Bibliography: leaves 70-72.
General Note:
General Note:

Record Information

Source Institution:
University of Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 000580753
oclc - 14081119
notis - ADA8858
sobekcm - AA00004930_00001
System ID:

Full Text






This work is dedicated to the past for my parents, to the

present for my wife, and to the future for my children.


I would like to express my sincere gratitude to Professor W. M.

Jones and all the members of my supervisory committee for their

scholarly guidance during the preparation of this work. A special

debt is owed to Dr. R. W. King of the University of Florida who

good-naturedly suffered all of my questions about molecular spec-

troscopy. It is such men as these who keep teaching in its honored

place among the professions.

The financial aid of the Woodrow Wilson National Fellowship

Foundation, the Graduate School of the University of Florida, and

the National Science Foundation made this work possible.


"Tom he said ... the trouble about arguments
is, they ain't nothin'i, but theories, after all, and
theories don't prove nothing, they only give you a
place to rest on, a spell, when you are tuckered out
butting around and around trying to find out something
there ain't no way to find out. "

Huckleberry Finn in
"Tom Sawyer Abroad"
by Mark Twain


ACKNOWLEDGMENT ................... ... iii

PREFACE ... ......... ..... .......... iv

ABSTRACT ............................ vi

INTRODUCTION .......................... 1

RESULTS .. .... ... .............. .... 11

DISCUSSION ...... ..... .. ..... ......... 45

EXPERIMENTAL ................. ........ 57

LIST OF REFERENCES ................... 70

BIOGRAPHICAL SKETCH. ............ ......... 73

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



Thomas Howard Ledford

August, 1973

Chairman: Professor William M. Jones, Department of Chemistry

(4, 5-c)Furotropylidene has the required structure and the required

number of pi-electrons to belong to the class of aromatic carbenes, a

group of carbenes that show predominantly singlet behavior and react

preferentially with electron-poor olefins. Contrary to this expectation,

(4, 5-c) furotropylidene appears to behave as a triplet above about

400 C. and adds to ordinary olefins. It also has the unusual property

of undergoing the first step of carbene-carbene rearrangement at low

temperatures. The second step of rearrangement, opening of the

intermediate cyclopropene to the rearranged carbene, is not detected.

Possible rearrangement, though not disproved, is shown not to be the

predominant reaction path such as is seen in certain slightly destabi-

lized aromatic carbenes.


It has been established that incorporating a vacant orbital of a

carbene into a ring containing conjugated double bonds can, when the

resulting system obeys the Huckel "4n + 2" rule, result in establish-

ment of so-called "aromatic" carbene systems that have unusual
1, 2, 3, 4
reactivity patterns. These conditions are satisfied when the

number of double bonds conjugated with the vacant orbital of the

carbene is an odd number. (See Figure 1.)


(Figure 1)

Examples of such aromatic carbenes include the 2-pi-electron
system diphenylcyclopropenylidene (1), 4the 6-pi-electron system,
1, 2
cycloheptatrienylidene (2), 12 and the 10-pi-electron carbene derived

from the 1, 6-methano(ll) annulene ring system (3). 5(See Figure 2.)

(1) (2) (3)

(Figure 2)

Aromatic carbenes display behavior patterns that are signifi-

cantly different from those shown by other carbenes. The delocali-

zation of charge density from the conjugated double bond system into

the vacant orbital of an aromatic carbene can be expected to increase

the nucleophilicity of the carbene. Also the carbene orbitals can be

expected to split into two different energy levels, affording the possi-

bility of a stabilized singlet state. (See Figure 3.)


(Figure 3)

Perhaps the most well known of these aromatic carbenes is

cycloheptatrienylidene (2). 1,' This carbene shows the properties one

might expect of a stabilized singlet with increased nucleophilic

character. It prefers to react with electron-deficient, rather than

electron-rich, olefins. For example, in a Hammett study with sub-

stituted styrenes, cycloheptatrienylidene showed a reaction rate con-

stant of +1.05. This compares with reaction rate constants of -0. 619
7, 8
for dichlorocarbene and -0. 38 for carbethoxycarbene. Not only is

the sign of the reaction rate constant significant, but there is also

significance in its larger absolute value, an indication that cyclohepta-

trienylidene is more discriminating than other carbenes; i. e., more

stable. Consistent with the hypothesized stabilization of its singlet

state, cycloheptatrienylidene reacts with acceptor olefins to form

cyclopropanes in which the olefin stereochemistry is preserved.
(Figure 4. )



(Figure 4)

Among the substituted cycloheptatrienylidenes that have been

prepared and studied, the following annelated compounds have shown

some interesting new departures in cycloheptatrienylidene chemistry.

(Figure 5.)

Carbenes (4) and (5) have been generated under conditions that

allow observation of their low-temperature esr spectra. Both have

triplet ground states and react with electron-rich olefins such as 2-

butene to form cyclopropane adducts. 10, 11, IZ Evidently both (4) and

(5) behave much like diphenylcarbene. Annelation has, in these two

cases, changed the cycloheptatrienylidene so significantly that a

singlet ground state is impossible. Carbenes (6), (7), and (8) show

even more dramatic effects of annelation upon cycloheptatrienylidene.

All three of these undergo carbene-carbene rearrangement at low to

moderate temperatures according to the following equations. 13, 14, 15

(Figure 6.)

The ground states for these carbenes are unknown, but it is

assumed that the rearrangements, at least, proceed through a singlet
16, 17
state. 16, 17 The nature of the intermediate or transition state leading

to this kind of carbene-carbene rearrangement has been somewhat

controversial. There is recent convincing evidence that such re-

arrangements proceed via a cyclopropene intermediate such as shown

(2) (4)

(5) (6)



(Figure 5)





4.. C:


(Figure 6)

in Figure 7. 1In this example, the intermediate cyclopropene (9)

appears to have been trapped by a Diels-Alder reaction with each of

several dienes. The intermediate (9) has also been approached from

another source as shown in Figure 7. 18



(Figure 7)

Monoannelation is said to substantially decrease the stability of

the tropyl cation. 19 Since carbene stability is thought to parallel

cation stability, and since monoannelation should have little effect

upon the stability of the intermediate cyclopropene, monoannelation is

thought to cause a destabilization of the carbene relative to the cyclo-

propene intermediate, thus increasing the probability of rearrange-
ment. Following this line of reasoning further, one could expect to

anticipate rearrangements in other carbenes by an analysis of aroma-

ticity and cation stability relative to the tropyl system.

The subject of this study is the carbene (4, 5-c)furotropylidene

(11) shown in Figure 8. This carbene should be expected to show at

least some aromatic character, since it does satisfy the Huckel 4n + 2


(Figure 8)

rule (n = 2), having 10 pi-electrons. The structure is not actually a

simple annelated cycloheptatrienylidene in one sense, because it lacks

a double bond analogous to the one between positions 4 and 5 in cyclo-


The question of whether aromatic character can be expected in

carbene (11) cannot be answered a priori. In fact, the whole concept

of aromaticity in troponoid ketones has been attacked by Bertelli;21' 22

but the concept seems so useful in explaining the behavior of troponoid-

derived carbenes that its continued use seems justified for the time

being. The following analysis, though it is mitigated by Bertelli's

argument, has been used to arrive at estimates of relative aroma-
ticities in the following series of ketones.3

As delocalization of electrons increases in the ring systems, the

bond order of the exocyclic C=O groups will decrease. This will


(13) (14)

(Figure 9)

parallel the increasing contribution of the dipolar form of the
+ -
carbonyl group C-O, thus paralleling the ability of the ring system to

stabilize positive charge at the carbonyl carbon atom. This trend

should be in the same direction as cation stability, thus in the same

direction as carbene stability.

A measure of C=O bond order can be obtained by a study of the

carbonyl absorption positions in the infrared spectra of the ketones in

this series. It has been shown that, for geometrically similarly

disposed C=O groups, there is good correlation between the frequency

of absorption and the calculated bond order; i. e. as bond order of the

carbonyl group increases, the higher will be the absorption frequency.

As the aromaticity in the series of ketones increases, the bond order

of the C=O groups should decrease, showing a lower infrared absorp-

tion frequency. The ketone (4, 5-c)furotropone (14), having a C=O

absorption at 1599 cm., is therefore less aromatic than benztropone
(13), having its C=O absorption at 1590 cm. 1 In turn, benztropone

(13) is less aromatic than tropone, since the carbonyl frequency of

tropone (12) is 1582 cm. By this criterion furotropone (14) has
more delocalization than a cross-conjugated cycloheptadienone because

all its bands appear at lower frequencies than the carbonyl band of
1 23
the dienone (15) shown in Figure 10 (ca. 1635 cm. -1).


(Figure 10)

The inference that furotropylidene (11) should have less
aromatic character than cycloheptatrienylidene (2) suggests that

furotropylidene, like other slightly destabilized aromatic carbenes (6),

(7), and (8), might be expected to undergo carbene-carbene rearrange-
ment. The examination of a hypothetical reaction pathway of a hypo-

thetical reaction pathway (Figure 11) suggests a possible complication.

Although the carbene (11) should be destabilized relative to the



(11) (16) (17)
(Figure 11)

cyclopropene (16), the product (17) from opening the cyclopropene
should give some caution. The rearranged carbene (17) has a carbene

center attached to an isoben -ran skeleton. Although isobenzofuran
is a 10-pi system, it apparel does not show the stabilization


associated with other aromatic systems. The parent hydrocarbon,

isobenzofuran, has been prepared for the first time only recently.24

It polymerizes readily in solution at moderate temperatures. The

reaction pathway in Figure 11 suggests that, for the first time, one of

the destabilized, partially aromatic carbenes might be headed into a

rearrangement pathway in which product stability is quite low. This

is in contrast to the highly stabilized aromatic products of carbene

rearrangements presented on page 4. The effect of this point will

become apparent as the results of this study are presented.


The carbene (4, 5-c)furotropylidene (11) was prepared in all
cases either by pyrolysis or photolysis of the sodium salt of the
tosylhydrazone of (4, 5-c)furotropone (18). The synthetic scheme for


+ + N2

0 o CH
(18) (11) H3
(Figure 12)
producing the required ketone was originally developed by Cook and
Forbes. Some modifications of their procedures were used in pre-
paring the ketone for this study. For example, commercially avail-
able furan-3, 4-dicarboxylic acid w a s converted to its diacyl

0 ->0 --0O
S O02H o- COCI oo2cH3
(19) (20)
(Figure 13)
chloride (19) by the action of thionyl chloride in the presence of a
catalytic amount of N, N-dimethylformamide. The acid chloride (19)
was never purified and characterized. Its presence was inferred from

its reaction with methanol to afford a quantitative yield of the known
dimethyl ester (20), previously characterized and reported by Cook
and Forbes. Using the procedures of Cook and Forbes, the reduc-

tion of the dimethyl ester was carried out using lithium aluminum

hydride, but the 76 percent yields reportedly attainable did not result.

Direct reduction of the crude diacyl chloride (19) did afford the di-

alcohol (21) in yields of about 70 percent. The di-alcohol was treated


0 0J-Ia.0->
(19) (21)
(Figure 14)
with activated managanese dioxide to effect oxidation of one of the

alcohol groups to the aldehyde stage. Again the yields reported in
the literature did not result. The reaction usually produced only

about 50 percent of the maximum amount of 3-hydroxymethyl-furan-

4-carboxaldehyde (22), accompanied by about half of the unreacted

di-alcohol (determined by proton resonance spectroscopy). This

situation was made usable by the fact that lead tetra-acetate oxidation
of this crude mixture of di-alcohol and mono-aldehyde afforded the

C/" CH \H
(21) + MnOZ --- O 1dc
(22) (23) i|
(Figure 15)

di-aldehyde (23) in yields of about 24 percent based on di-alcohol.

The 3, 4-furandicarboxaldehyde (23) was condensed with acetone

using the procedure of Cook and Forbes to give exactly the reported
yield of 38 percent.

C*C\ H CH3
o C- ac=o OH ,O>O

11 CH3
(23) (14)

(Figure 16)

Conversion of furotropone into its tosylhydrazone was best

carried out by treatment of the ketone (14) with tosylhydrazine in

tetrahydrofuran containing a trace of anhydrous phosphoric acid. The

reaction worked best when the reactants were merely allowed to stand

together at room temperature for five to seven days. This procedure

gave the tosylhydrazone (24) in 65-70 percent conversion.

J o-1"-=N-NH--T

(14) (24)

(Figure 17)

A solution of the tosylhydrazone in tetrahydrofuran was treated

with sodium hydride to produce the sodium salt of the tosylhydrazone

(18). The weight of the sodium salt produced suggests from the

stoichiometry of the reaction that one mole of tetrahydrofuran is

included in the salt as bound solvent. All yields in reactions of this

sodium salt have been adjusted to reflect this effect.

/N- N-Ar

/ H2 +

(24) (Figure 18) (18)

Thermal decomposition of the sodium salt of (4, 5-c)furotropone

tosylhydrazone (18) in the presence of benzene at 1880 C. led to

formation of the formal C-H insertion product (25) in 43 percent

isolated yield. The structure of (25) was assigned primarily on the

basis of its spectral properties. At tau 2. 71 and 2. 72 there were two

singlets, assigned to the furan hydrogens and to the benzene hydro-

gens, respectively. The total of both peaks was seven hydrogens.

The vinyl hydrogens (Ha) appeared at 3. 5-3. 9 tau as a doublet, split

by 11.5 Hz. through coupling to (H1b). Each peak of this doublet

showed a slight splitting (ca. 1 Hz.) attributable to allylic coupling

to the tertiary hydrogen (Hc). The vinyl hydrogens (Hb) appeared as

a doublet of doublets at tau 4. 33-4. 6. In this pattern the coupling

(11. 5 Hz. ) between vinylic protons and the coupling (5-5. 3 Hz. )

between (Hb) and (Hc) were both easily discernible. The tertiary

hydrogen (Hc) appeared at tau 5. 64. It was primarily a triplet pattern

showing some superimposed allylic splitting. The infrared spectrum

indicated the monosubstituted benzene structure by its absorptions at

762 cm. -1 and 700 cm. -1

,-, ,J 1' 10* O

L> __ ___j ,I

(18) +



(Figure 19)

The proton magnetic resonance spectrum of the crude reaction

mixture showed only the benzene C-H insertion product (25). No

evidence of a cycloheptatriene structure was present. Careful

examination of the reaction mixture by analytical thin-layer chroma-

tography failed to show any biphenyl in the sample.

In a competition reaction allowing the carbene equal access to

benzene and d6-benzene, essentially equal amounts of deuterated and

non-deuterated products were produced as determined by both mass

spectroscopy and by 100 MHz. proton magnetic resonance.

An effort to prepare the product of C-H insertion into cyclo-

hexane failed because of low yields. The carbene (11) undergoes

reaction with olefins in dioxane solution without t.i:.ir-, dioxane into the

reaction mixture. These observations suggest that the benzene C-H

insertion reaction probably does not result from direct insertion, but

through an intermediate that will be discussed in a later section of

this report.

Attempted addition of the carbene (11) to the double bond in

cyclohexene resulted in a mixture that could not be separated cleanly

enough to allow characterization of any of the products. The proton

magnetic resonance spectrum of the crude product did suggest that

some addition to the double bond had occurred., The presence of other

products in the reaction mixture suggests that (4, 5-c)furotropylidene

is not incapable of C-H insertion, but one is left to speculate about

whether the products arise by direct reaction of the carbene or by

secondary processes.

Decomposition of the tosylhydrazone salt in a refluxing solution

of styrene in dioxane (b. p. 1010C.) was successful in producing a

phenylcyclopropane (26) that could be separated and characterized.

Yields as high as 50 percent were produced in solutions that were

quite dilute (1.5-3 percent styrene). Little, if any, dioxane was

attacked by the carbene. The major side-reaction was production of

considerable amounts of polystyrene. The 100 MHz. magnetic

resonance spectrum of the phenylcyclopropane (26) showed a sharp

singlet at tau 2. 86 with a correct integral for the five phenyl hydrogen.

There were two small singlets (total ZH for both) representing the

furan hydrogens, nonequivalent in this molecule, at tau 2. 96 and

3. 01. The vinylic hydrogens (Ha) and (Ha,) appeared as two doublets

in the region from 4. 0 to 4. 33 tau. Each of the doublets was split by

12 Hz. through coupling to the hydrogens (Hb) and (Hbi). The value of

this coupling between cis olefinic hydrogens suggests that they are

connected to a seven-membered ring. 25 The hydrogens (Hb) and

(Hb') appear as two doublets at 5. 25 to 5. 95 tau, again spaced by

about 12 Hz. ; but each peak in these doublets is split very slightly

again (about 2 Hz.), suggesting coupling across the ring between (Hb)

and (Hb,). This coupling is to be expected because these hydrogens

(non-equivalent because of the phenyl group) are situated for W-form

coupling. Although the furan hydrogens are also situated for W-form

coupling, and their chemical shift difference is ca. 12 Hz., the cou-

pling between them is only barely discernible. It is interesting to

note that the facing pairs of hydrogens on the seven-membered ring

and in the furan system show decreasing chemical shift differences

with increasing distance from the symmetry-disturbing phenyl group.

The cyclopropyl hydrogens in the phenylcyclopropane (26) present the

expected ABX pattern. The (Hc) hydrogen (X) appears in the 7. 53-7. 85



H\ /H
(18) + C=C ->

\dH (26)
HH' \C6H5

Hf' /Hb, CH

(Figure 20)



11 .Itill

(Figure 21)

~------- --r --

I _I -1 r r ----~ IT T_ -I_ -~I~-, ~1 ~:--17 T~-.=j ;._ -I-

tau area as two doublets that show overlap between the two central

peaks. The geminal cyclopropyl hydrogens (AB pair) appear as the

expected pair of overlapping quartets in the region 8.7-9. 0 tau. The

spacing between the midpoints of the two quartets (1/2 abs. value of

JAX + JBX) allows easy calculation of the predicted spacirin between
lines 9 and 12 in the X portion of the spectrum. The predicted

spacing between lines 9 and 12 (15. 6 Hz.) was observed and permitted

assignment of lines 9 and 12 as the two outside lines in the X portion

of the spectrum. 2The value of JAB = 5. 4 Hz. was directly

measurable from the spectrum.

The infrared spectrum of (26) shows absorptions near 700 cm.-1
and 750 cm. consistent with the mono-substituted benzene
-1 -1
structure. Absorptions at 860 cm. -1 and 1028 cm. offer confirma-

tory evidence of the cyclopropane ring indicated by the absorption at
1 27
3060 cm.

Pyrolytic decomposition of the tosylhydrazone salt (17) in the

presence of 1-butene gave the expected ethylcyclopropane (27) in

about 50 percent isolated yield. The cyclopropane was accompanied

by three minor by-products that were never identified. The proton

magnetic resonance spectrum of (27) showed a two-hydrogen singlet at

2. 95 tau for the furan hydrogens. The vinylic protons (Ha) and (HaI)

that showed a pair of doublets in the phenylcyclopropane (26) appeared

in this ethylcyclopropane as an overlapped pair of doublets split by

11. 5 Hz. at tau 3. 87-4. 3. The other pair of vinylic hydrogens (Hb)

and (Hb') appeared as a pair of separated doublets at 5. 25-5. 8 tau

showing the same W-form coupling observed in the phenylcyclopropane

(26). It is interesting to observe the smaller symmetry disturbance




C= --

"-!- I -H (27)

(Figure 22)



produced by the ethyl group in (27) compared with the larger effect

of the phenyl group in (26). Whereas the furan hydrogens were re-

solvable in (26), they were not resolvable in (27). Further evidence

of lower disturbance of symmetry is provided by the fact that the pair

of doublets representing the vinylic hydrogens (Ha) and (Ha,) are over-

lapped in (27), but well separated in (26). The remainder of the

spectrum of the ethylcyclopropane (27) was a complex eight-hydrogen

signal in the region of 8. 3-9. 5 tau that included the cyclopropyl

hydrogens and the hydrogens on the ethyl group.

Thermal decomposition of the tosylhydrazone salt (17) in the

presence of isobutene gave a remarkably clean reaction producing the

dimethylcyclopropane (28) in 28 percent yield. The structure was

assigned primarily on the basis of the proton magnetic resonance

spectrum. This molecule provides an excellent example of the pro-

found effects of molecular symmetry on nuclear magnetic resonance.

A plane of symmetry can be drawn through the dimethylcyclopropane

(28). This plane includes the plane of the cyclopropyl ring and bisects

the plane of the fused furotropyl ring system. This symmetry results

in magnetic equivalence of the furan hydrogens and both sets of vinyl

hydrogens in the seven-membered ring. This results in a simplified

spectrum for the compound. A two-hydrogen singlet for the furan

hydrogens appeared at 2. 83 tau. Instead of the more complex vinyl

absorptions observed in the styrene adduct (26) and in the 1-butene

adduct (27) a simple AB pattern appeared. A doublet centered at 3.81

tau showed a two-hydrogen signal for the (Ha) pair. Another doublet

centered at 5. 09 tau was presented by the (Hb) pair of hydrogens. The

coupling between (Ha) and (Hb) was 11. 5 Hz., about the same value


(1 7) + C-




(Figure 23)

P~ ts e~'3~ CO II
,P~ .,-17_ .l.P_ '---~-`
----- -------~-- -- -------------~---------~

observed in other compounds in this series. The two methyl groups

produced the expected six-hydrogen singlet at 8. 9 tau, accompanied

by a nearby singlet for the two equivalent cyclopropyl hydrogens at

9. 2 tau.

The profound effects of changes in symmetry in spirocyclo-

propanes such as (26), (27), and (28) provide an excellent basis for

assignment of stereochemical configurations in cis- and trans-1, 2-

disubstituted spirocyclopropanes by nuclear magnetic resonance.

Trans 2-disubstituted spirocyclopropanes (see Figure 24) can be

expected to have equivalent sets of furan hydrogens and vinylic

hydrogens facing each other across the ring. This is because rotation


(Figure 24)

about the twofold axis of symmetry shown in the drawing makes

these sets of hydrogens equivalent. On the other hand, cis-1,2-

disubstituted spirocyclopropanes can be expected to show the same

kind of complex pattern observed for the vinyl hydrogens as was seen

in the monosubstituted spirocyclopropanes (26) and (27), resulting

from the non-equivalency of facing pairs of hydrogens on the opposite

sides of the seven-membered ring. A model for cis-disubstituted

spirocyclopropanes of this type has been prepared by Krajca from the
reaction of 4, 5-benzotropylidene with cyclohexene. 1This compound

(29) shows a nuclear magnetic resonance pattern in the vinyl region

that is essentially identical to the pattern shown by the phenylcyclo-

propane (26). A similar vinylic absorption pattern has been used to

(29) H

(Figure 25)

assign stereochemical configurations in a series of spirocyclopro-

panes derived from the reactions of 4, 4-dimethylcyclohexadienylidene

with various olefins. 27 This is shown in Figure 26. Both cis- and

trans-1, 2-disubstituted spirocyclopropanes of this type were prepared.


ss ,,- R ~ s- H
R "H

(Figure 26)

Both isomers showed the expected effect of symmetry differences upon

the nuclear magnetic resonance spectra in the vinylic region.

With the above-described basis for making stereochemical

assignments in 1, 2-disubstituted spirocyclopropanes, it is possible to

study the stereospecificity of the reaction of furotropylidene (11) with

olefins. The stereospecificity test is widely used as a chemical
test for distinguishing between singlet and triplet states in carbenes.

Stereospecific addition; i. e. addition to olefins to produce cyclopro-

panes in which olefin stereochemistry is preserved, is characteristic

of singlet carbenes. Non-stereospecific addition, in which olefin

stereochemistry is not preserved, is characteristic of triplet


A stereochemical study was undertaken using cis- and trans-2-

butenes as acceptor olefins for the carbene (11). Thermal decompo-

sition of the tosylhydrazone salt in the presence of cis-2-butene at

118-1200C. and in the presence of trans-2-butene at the same tem-

perature produced two crude reaction mixtures that were virtually

identical in their proton magnetic resonance spectra. Gas chromato-

graphic examination of the crude reaction mixtures using a 100-foot

capillary column coated with Ucon LB-550 showed at least 11 com-

ponents in the reaction mixtures. Most of the chromatographic peaks

were in the same quantitative relation to each other in both mixtures.

Separation of the main peak on a preparative gas chromatographic

instrument, though it gave a less-perfect separation than the capillary

instrument, did allow some narrowing in the choice of the significant

peaks in the chromatograms prepared on the capillary instrument,

since this fraction was shown by nuclear magnetic resonance to con-

tain the major components present in the crude product. The signifi-

cant area turned out to be a group of two smaller peaks and one major

peak that were not even well separated on the capillary instrument.

Quantitative differences were seen in the relation of the two smaller

peaks when comparing samples prepared by thermal reaction with the

cis and trans olefins, but the significance of this difference between

these two smaller peaks may be trivial because of the following obser-

vations: 1. The proton magnetic resonance spectra videe infra) of

both reaction mixtures were identical. Both crude reaction mixtures

appeared to be predominantly the trans-spirocyclopropane videe infra).

2. There was no indication of the presence of the cis-spirocyclopro-

pane in either sample to the limit of detection by the proton magnetic

resonance spectra. 3. Thermal reaction of the tosylhydrazone salt

with trans-2-butene is most likely to produce the more stable trans-

spirocyclopropane if product isomerization is taking place. A photo-

chemical decomposition of the salt in the presence of cis-2-butene is

most likely to produce the cis-spirocyclopropane, because of expected

lower probability of thermal cis-trans isomerization at the milder

temperatures, ca. 500C., used. A comparison of the capillary

chromatograms of these two reactions showed the same quantitative

relation among the three peaks in this significant area. Apparently

the major peak is the trans-spirocyclopropane videe infra). The two

minor peaks were never identifiable for the reasons of small sample

size and difficulty of purification. The proton magnetic resonance

spectra suggest that these are probably mainly C-H insertion products.

The attainment of the same product mixture from carbene reactions

with a pair of isomeric cis-and trans-olefins is the criterion for com-

plete loss of stereospecificity in the reaction.

The failure to isolate any of the cis-spirocyclopropane from the

reactions with the 2-butenes and to demonstrate the stability of the cis-

isomer to reaction conditions does leave the experiment open to the

criticism that the cis-isomer is possibly being formed, then is decom-

posing to either the trans.isomer or to some other product. This

possibility is impossible to exclude rigorously in the present case,

but some inferences for the stability of the cis-isomer can be drawn

from a study of known model compounds. Cyclopropanes of the type

(30) are subject to a cleavage of the cyclopropyl ring followed by


30: a. R = COZCH3
b. R = H
(30 a, b)

(Figure 27)

isomerization to an indane derivative. The substituted cyclopropane

(30a) undergoes isomerization at 1300C., but the unsubstituted cyclo-

propane (30b) is stable at 1500C. Similarly, one should expect an

enhanced rate of isomerization in the vinyl-substituted spirocyclo-

propane (31) (Figure 28) because of stabilization of radical inter-

mediate (32). The isomerization is slow at 1000C., since the cyclo-

propane can be isolated as the main product from reaction mixtures

exposed to that temperature for 0. 5 hr. videe infra). This su'-pests

that cis -dialkyl-spirocyclopropanes would require substantially

higher temperatures before isomerization to the trans isomer would

occur at a significant rate.

(31) (32) (33)
(31) (32) (33)

(Figure 28)

Isolation and characterization of the trans-spirocyclopropane

produced from the reaction of carbene (11) with the cis- and trans-2-

butenes proved to be as difficult as the foregoing discussion would

suggest. Reaction of trans-2-butene by decomposition of the sodium

salt at 1150C. produced a crude reaction mixture, the proton mag-

netic resonance spectrum of which suggested that the main component


(18) + =C -

H CH3 (CH3
3 (34)

(Figure 29)

was the trans-1, 2-dimethylcyclopropane (34) contaminated with C-H

insertion products. Preparative layer chromatography on silica gel

plates did not improve the appearance of the spectrum very much

until the main band was collected and re-chromatographed on silica

gel plates using very low sample loading. This allowed separation

into three bands, the major one of which gave a spectrum suggesting

a fairly pure sample of the trans-adduct (34). Because of small

sample size, neither of the two minor components was identified. The

yield of the trans adduct (34) appears to be in the neighborhood of

25 percent, but extensive handling of small samples makes this

number unreliable. The assignment of the structure (34) rests pri-

marily on the proton magnetic resonance spectrum. There is the

usual sharp singlet at 2. 9 tau for the furan hydrogens. From the

discussion on pages 17, 18, and 19 one would expect the AB pattern

that is observed in the vinylic region, produced by the hydrogens on

S--. I .. .. 3 ,

\ / 3
(18) + C= ---

CH 3


9 ,H (34)

(Figure 30)

the seven-membered ring. One of the AB doublets is centered at 3. 8

tau, the other at about 5. 3 tau, with a coupling of 11. 5 Hz. At 8.65

to 8.9 tau there are two peaks whose relative intensity su2,r.sts they

are coupled to the cyclopropyl hydrogens that appear sliilitly upfield.

There is a third sharp peak just downfield of these cyclopropyl

hydrogens at about 8. 93 tau, the shape and intensity of which leave

its interpretation open to question. It is probably a spurious peak

due to the presence of some impurity, but it could also be the result

of so-called "virtual coupling" through the cyclopropyl hydrogens.

The integration curve is not much help in deciding, since the effect

of this peak on the total is not very great. The best integral does

result from consid, rint it to be a spurious peak, though. The spacing

of this suspicious peak from the closest of the other two is, whether

fortuitous or not, equal to the spacing between the other two and equal

to one of the spacing patterns seen in the signal for the cyclopropyl

hydrogens whose multiple appears at 9. 05 to 9. 5 tau. The integration

curve for the cyclopropyl hydrogens appears to fall just a little bit

short of the required amount, but some of this signal may be buried

under the "suspicious" peak already discussed. To judge from this

spectrum, there is very little, if any, of the cis-spirocyclopropane

present in the sample.

An overview of the results of the stereochemical study with the

2-butenes suggests that a study with another olefin, one that would

hopefully give a cleaner reaction, would reinforce the argument for

the loss of stereospecificity in addition reactions of this carbene.

Accordingly a study was carried out using trans-deuteriostyrene as an

acceptor for the carbene. From the experience gained with the non-

deuterated styrene reaction it was known that this reaction (shown in

Figure 20) can be used to produce rather pure samples of the phenyl-


The required deuterated styrene was prepared by the stereo-

specific addition of dicyclohexylborane to phenylacetylene followed by
hydrolysis with deuterioacetic acid to free the styrene. Formal

addition of the carbene to this olefin was carried out by pyrolysis

of the tosylhydrazone salt in a dilute solution of the olefin in boiling

dioxane (b. p. 1010 C.). The resulting phenylcyclopropane was

separated by preparative layer chromatography. Use of d6-benzene

as a solvent allowed observation of the geminal cyclopropyl hydrogens

by 100 MHz. proton magnetic resonance spectroscopy as two doublets

appearing at 8.7-9. 2 tau. One of the doublets was split by 8. 5 Hz.;

the other, by 7. 0 Hz. By double irradiation to decouple the neighboring

cyclopropyl hydrogen (Hc) from the geminal pair, the four signals

were caused to collapse to two signals having a separation of about

12 Hz. Integration of the four signals (before decoupling) and the two

signals (after decoupling) showed the presence of an equal mixture

of the two possible isomers.

Though the formation of an equal mixture of the two possible

deuterio phenylcyclopropanes in this study suggests non-stereospecific

addition of the carbene to the olefin, the result is not conclusive unless

the possibility of olefin isomerization before reaction and the possi-

bility of product isomerization after reaction are excluded. The

olefin was determined to be stereochemically stable under the reaction

conditions by a control experiment. The stability of the product is not

so easily proved. Separation of the two stereoisomeric products is not

^ A







C6 H5


(Figure 31)

-I_ 11 ~1 -r -I I ~1 ~1 _1 I r r -r -1 _I -r .r r --1





possible, so a direct test for isomerization under reaction conditions

is not possible. The best rim-,airriinr, option is to conduct the reaction

at a temperature at which product isomerization is highly unlikely.

One can also draw inferences about the thermal stability and the

photochemical stability of the phenylcyclopropane adduct by examina-

tion of model compounds videe infra).

The reaction with trans-deuteriostyrene was repeated by decom-

posing the sodium salt of the tosylhydrazone photolytically at about

450C. This procedure also produced an equal amount of the two

possible stereoisomers. Though it might have been desirable to

have carried out the photolysis at even lower temperatures, the

properties of this carbene are such that it does not add readily to

olefins at low temperatures. This point will be discussed further in

connection with reactions of this carbene with butadiene. The styrene

did not isomerize under photolysis.

The photolytic and thermal stability of the phenylcyclopropane

(26a) can be inferred from the following data: 1. The vinylcyclopro-

pane (31) (see Figure 28) requires temperatures greater than 1000 C.

for an appreciable rate of ring-opening, followed by closure to the

cyclopentene (33). 2. The same vinylcyclopropane was determined

videe infra) to be photolytically stable under reaction conditions. 3.

The somewhat similar l-phenylspiro(2. 6)nona-4, 6, 8-triene (35)

shown in Figure 32 requires temperatures greater than 750C. for

isomerization to the 8-phenylbicyclo(5. 2. 0)nona- 1, 3, 5-triene (36), 31

but its isomerization is aided by the formation of a new stable com-

pound of a type that cannot be formed from (26). 4. The vinylcyclo-
propane 37) rearranges to 38 at 50-7 31 On the other hand, the
propane (37) rearranges to (38) at 50-750 C. On the other hand, the

butadiene adduct (39) is stable enough to be isolated by preparative

gas chromatography. 27








(Figure 32)

Pyrolysis of the tosylhydrazone salt (18) in the presence of 1, 3-

butadiene at 1180 C. produces almost exclusively the 1, 4-addition

product (33) (55 percent yield) shown in Figure 33. It has been hypoth-

esized that triplet carbenes might react with 1, 3-dienes in the 1, 4-

addition mode. Few carbenes, if any, actually do add in this manner




by direct reaction. 3Most adducts arising from a formal 1, 4-

addition are products from the thermal isomerization of initially

formed 1, 2-addition products such as (31). That proved to be true

in this case also. Thermal decomposition of the tosylhydrazone salt

in 1, 3-butadiene at 1000 C. for short reaction times (0.5 hr. or less)

produced the 1,2 adduct (31). Heating of the vinylcyclopropane (31)

(18) -t O 180)

(31) (33)

(Figure 33)

at 1200 C. for 0. 5 hr. caused complete conversion to the isomeric

cyclopentene (33).

Structural assignment of the 1, 4-addition product (33) was based

on the following spectral data. In the proton magnetic resonance

spectrum there is the expected two-hydrogen singlet at 2. 82 tau for

the furan hydrogens. Since this molecule has the same kind of

symmetry as the dimethylcyclopropane (28) shown in Figure 22, one

can predict the same kind of AB pattern for the vinyl hydrogens (Ha)

and (Hb) in the seven-membered ring. This expected four-line AB

pattern is observed in (33). One of the doublets in the AB pattern is

centered at 3. 9 tau and is split by 11. 5- 12 Hz. The other doublet is

centered at 4. 57 tau (representing the (Hb) hydrogens), but the left

half of the doublet has a partially superimposed peak from the vinylic

hydrogens in the cyclopentene ring (Hc). The integral for the lower-

,-^ -------

to 70 64 20 t n, 4 to T

H Ha Hb

S1 (33)

(Figure 34)

field doublet is two hydrogens. The integral for the upper-field

doublet containing the signal for the cyclopentene olefinic hydrogens

indicates a total of four hydrogens. The remainder of the spectrum

is a sharp singlet at 7. 49 tau with a correct integral for the four

allylic hydrogens (Hd). The lack of discernible splitting of the

allylic hydrogen is consistent with the reported 0. 5 Hz. allylic
splitting in cyclopentene itself. The high symmetry of the 1,4-

adduct (33) gives rise to some doubt as to whether the C=C bond in the

cyclopentene ring should even be infrared active at all. Nevertheless,
there is a weak absorption at 1618 cm. that does fit the known

pattern for C=C stretch in five-mrembered ri,'rs (cyclobutene, 1566
-1 -1 -1 34
cm. ; cyclopentene, 1611 cm. ; cyclohexene, 1649 cm.

Structural assignment of the 1, 2-addition product with butadiene

was based on the following information. The furan hydrogens appeared

as a two-hydrogen singlet at 2. 9 tau. The vinyl region showed

clearly the results of the symmetry-disturbing exocyclic vinyl group.

The pattern for the (H a) and (Fa,) hydrogens was a partially over-

lapping pair of doublets showing the same 11. 5 Hz. coupling between

the AB pair in the seven-membered ring that has been observed in

(26) and (27). This four-line signal for the (Ha) and (Ha,) hydrogens

was about 3. 9-4. 2 tau. Another four-line signal for the (Hb) and (Hb,)

hydrogens appeared at about 5. 1-5.65 tau. Once again, since these

two hydrogens are nonequivalent, the W-form coupling of ca. 2 Hz.

was observed in addition to the coupling with the (Ha) and (Hai)

hydrogens. The vinylic hydrogens belonging to the exocyclic vinyl

group appeared between the two sets of signals for the AB pair in the

seven-membered ring. The (HI) signal appeared from about 4.2 to

J f-ujI



Hf Ha Hb

-, H1 (31)

H- "'H2

(Figure 35)

r r I -r I -~I _I 1 _:1 -~II II 1 7~-~7=



(Figure 36)

r =s r

4. 45 tau with primarily a four-line pattern. The (Hz) and (H3)

hydrogens were at 4. 75-5. 0 tau presenting a complex pattern that had

so much fine structure that direct measurement of the coupling

constants was not possible. Use of a 100--.TTz. spectrometer made

it possible to resolve each set of vinylic hydrogens, both the AB

pair and the exocyclic vinyl hydrogens, sufficiently to allow an accu-

rate integration for each signal. All of the integrals were satisfac-

tory. The remainder of the spectrum presented the expected ABX

pattern for the cyclopropyl hydrogens. Direct measurement of JAB

was 5. 0 Hz. The AB portion of the spectrum (for the vicinal cyclo-

propyl hydrogens) was at 8. 6-9. 0 tau. The X portion was at about

8. 15-8. 5 tau. The AB signal allowed easy recognition of the expected

pair of overlapping quartets. The X signal gave an integral that was

slightly lower than the correct value because some of the lines were

buried in instrument noise. Four of the lines were visible, but only

two of them were very strong. The spectrum also showed a sharp

singlet at about 8. 6 tau from a contaminating inhibitor (2, 6-di-tert,

butyl-4-methyl phenol) picked up during exposure of the sample to a

commercial grade of tetrahydrofuran. Elemental analysis was made

impossible because of the presence of the inhibitor, since it was

difficult to separate from the sample. The problem was sur-

mountable by the ready conversion of the 1, 2-adduct to the 1, 4-

adduct (33), which was easy to separate from the inhibitor and to

provide in pure form for elemental analysis. The exocyclic vinyl

group in (31) was shown by infrared. 35

Decomposition of the tosylhydrazone salt by photolysis at low

temperatures in the presence of 1, 3-butadiene caused a remarkable

change in the character of reaction with this olefin. At temperatures

of -600 to -30 C. it produces the product (40), shown in Figure 37,

in about 40 percent yield as the only hydrocarbon product identifiable.

The structure of (40) was identified by the striking similarities in its

spectra with the spectra of a number of similar compounds recently

prepared and elucidated in detail by Coburn. 1In the proton mag-

netic resonance spectrum the furan hydrogens produced a two-hydrogen

singlet at 2. 83 tau. The vinylic hydrogens (Ha) produced an AB

pattern centered near 4. 0 tau split by 10 Hz. The other vinylic

hydrogen (Hb) produce a poorly resolved peak at 4. 47 tau. The

allylic hydrogens appear as a broadened peak at 7. 52 tau. The

cyclopropyl hydrogen (He) appears as a doublet at about 7. 85 tau,

coupled by about 5 Hz. to the other cyclopropyl hydrogen (Hi) which

appears upfield as a multiple at 9. 15-9. 45 tau.

Photolysis of the tosylhydrazone salt at intermediate tempera-

tures (ca. 400C.) produced a mixture of (40) and the 1,2-addition

product (31) from reaction with butadiene. The ratio was about

45:55. None of the 1, 4-adduct (33) was produced.

In control experiments the 1,2-adduct (31) was shown to be

stable to photolysis; therefore, it is not the source of the product (40).

The product (40) was shown to be thermally stable for at least 20

minutes at 140 C., since it could be purified by preparative gas


To see if normal carbene behavior could be elicited at low

temperatures, an effort was made to add the carbene to trans-2-

butene by photolysis of the tosylhydrazone salt at -500 C. Normal

carbene addition to the olefin did not occur, as shown by the proton




o0 70 o60 5 0 M 4 4.0 0 20 lA

(18) +


Ha H

/ Hb (40

(Figure 37)



__ i ,T 4.-

magnetic resonance spectrum of the crude product. The friable

appearance of the product suggested that it was at least partly


In experiments designed to allow equal amounts of olefin

acceptors to compete for the carbene (11), the following relative rate

data were obtained:


1-butene 0.8
isobutene 1. 0
1, 3-butadiene 9.0

(Table 1)

One experiment was done to attempt to observe a signal in the

proton magnetic resonance spectrum indicating the operation of the

chemically induced dynamic nuclear polarization (CIDNP) phenom-
enon. Such an observation would be indicative of the presence of a

triplet carbene. Thermal decomposition of the tosylhydrazone salt in

solution in an nmr sample tube containing a mixture of approximately

20 percent cyclohexene in d6-dimethyl sulfoxide failed to show the

CIDNP phenomenon. This could be attributable to the low solubility

of the tosylhydrazone salt in this medium, indicated by the failure to

observe the presence of the salt in the spectrum of the solution.


Thermolysis or photolysis of tosylhydrazone salts of tropone

and substituted tropones in solution have been found to give at least

five different kinds of reactive species (Figure 38). Unsubstituted

tropone (12) shows chemistry of only the singlet carbene (I). 9

Mono-annelated tropones (13) and (41) show some chemistry expected

of the singlet carbene (I), but in general, their chemical behavior

is dominated by the bicycloheptatriene (III) and the rearranged

singlet and triplet aryl carbenes (IV) and (V). 14, 0 The di-annelated

tropone (42) shows only the chemistry of the bicycloheptatriene (III)

and the aryl carbene, presumably singlet (IV) and triplet (V). 15 The

di-annelated and tri-annelated tropones (43) and (44) show typical

diaryl carbene chemistry. They have been shown to have triplet

ground states, but their chemistry is dominated by the singlet. 10, 11, 12

The reasons for these differences can be qualitatively rationalized

in terms of the expected relative energies of the different intermediates.

Carbene stabilities are thought to run parallel to cation stabili-

ties. Mono-annelation, known to de-stabilize the tropyl cation, should

not be expected to have significant effect on the stability of the inter-

mediate cyclopropene (III). Mono-annelation should then decrease the

stability of the carbene relative to the cyclopropene intermediate,

making the rearrangement easier.20 The di-annelated species (43) and

the tri-annelated species (44), by incorporating into the fused benzene

systems the double bond that must suffer attack in order for





lA A'

8 B B

D D1

A 8 D


(12): A, B, D = H
(13): A, D = H; B = fused benzene ring
(41): A = fused benzene ring; B, D = H
(42): A, B = fused benzene ring; D = H
(43): A, D = fused benzene ring; B = H
(44): A, B, D = fused benzene ring

(Figure 38)

rearrangement to occur, reduce the probability of rea rr an cement by

increasing the relative energy of the intermediate cyclopropene

because of loss of benzenoid aromaticity. The di-annelated species

(42) does not require as much loss of aromaticity to form the cyclo-

propene (III), so it undergoes rearrangement easily. Many carbenes

that are formed in their singlet states react in their singlet states,

because the singlet is so reactive that reaction occurs before colli-

sional deactivation to triplet, if that is the ground state, can occur.

Equilibration between a reactive singlet and a relatively unreactive

triplet can also cause the same effect.

The present carbene (11) fits this overall scheme, but as a

result of its unusual structure, it seems to have a unique place in the

scheme. In the first place, unlike any of the other carbenes studied,

under certain conditions (above about 400 C.) its chemistry is appar-

ently dominated by the triplet.

The complete loss of stereospecificity in reactions of (4,5-c)furo-

tropylidene is consistent with triplet behavior. The nonstereospecific

addition of a carbene in solution is now well established as a criterion

for interpreting the reaction in terms of a two-step reaction; i. e. via

triplet. The present stereospecificity studies must be taken with the
caveat of Gaspar and Hammond 3in mind that "Nonstereospecific

addition cannot be taken as a proof that an attacking species is a

triplet unless it has also been shown that under some other conditions

a species of the same composition can give stereospecific addition. "

Closs, in a more recent view, asserts that nonstereospecific reac-

tions can always be interpreted as proceeding via the two-step mech-

anism; i. e. via triplet. 29

The relative reactivities of (4, 5-c)furotropylidene in reactions

with olefins also fit the triplet pattern. It is well accepted that con-

jugated dienes, such as 1, 3-butadiene, show a high relative rate of

reaction with triplets because of allylic stabilization of the di-radical

intermediate (Figure 39) in the two-step reaction. The common use

of butadiene as a "triplet scavenger" to improve stereospecificity of

r +C_ --- 2

(FiLire 39)

carbene reactions by selectively draining off triplet illustrates this
principle. 3The relative rates found in this present study also fit

the relative rate pattern for the rate of radical addition vs. abstraction

with the same olefins. 39

Interprcting the reaction of (4, 5-c)furotropylidene with benzene

in terms of triplet chemistry is aided by consideration of some related

reports in the literature. Bis(carbomethoxy)carbene has been gener-

ated by photolysis of the corresponding diazo compound under two
sets of conditions. Direct photolysis produces a carbene that reacts

in the singlet state as shown by the stereospecificity of its reactions

with olefins. Photosensitized decomposition produces a carbene that

reacts in the triplet state as shown by the loss of stereospecificity in

its reactions with olefins. The same carbene, prepared by each of

the two methods, was allowed to react with benzene. Direct photolysis

of methyldiazomalonate in benzene gave the cycloheptatriene (45) and

the C-H insertion product (46) in a ratio of 2. 7 to 1.0. The photo-

sensitized reaction gave the same two compounds in a ratio of 1. 6 to



(Figure 40)

1. 0. The increased amount of the phenylmalonate (49) when the

carbene is prepared in the triplet state is consistent with the inter-

mediacy of the di-radical, which can either close to the norcaradiene

related to the cycloheptatriene, or undergo hydrogen shift to form the

phenylmalonate. Increased triplet character in the attackinoh carbene

increases the amount of the C-H insertion product. If the slow step

of the reaction is attack of triplet carbene upon a benzene double

bond, the absence of a deuterium isotope effect is to be expected.

This was demonstrated in the present study with (4, 5-c)furotropylidene.

Still, there are hazards in interpreting the insertion of furo-

tropylidene into the C-H bonds of benzene as necessarily a triplet

behavior. A di-radical intermediate such as that shown in Figure 40

could arise from another path. Consider, for example, the six-

membered carbocyclic carbene, 4, 4-dimethylcyclohexadienylidene.

It apparently reacts with olefins in the singlet state in solution. It

reacts with benzene to produce a spironorcaradiene (47) shown in

Figure 41. This spironorcaradiene isomerizes at 1000C. to produce

the intermediate (48) that is very much like the di-radical intermediate

that could arise from triplet attack upon the benzene double bond.

Here is apparently a singlet pathway to the di-radical intermediate.

\ N (48)



(Figure 41)

None of the analogous norcaradiene was detected in the furotropylidene

case, even when the reaction was carried out by photolysis at room

temperature; but the possibility of that intermediate is very real be-

cause of the complexity of the mixture that was produced in the reac-

tion. The absence of a deuterium isotope effect would also be

expected from the singlet pathway.

Although no one piece of evidence in this report can be said to

rigorously prove that (4, 5-c)furotropylidene is behaving as a triplet

at temperatures above 400C., certainly the mass of evidence taken

as a whole looks fairly convincing. One thing is certain--the cyclo-

propene intermediate (type III, Figure 38) dominates at lower tempera-

tures. This is shown by the trapping of the cyclopropene intermediate

(16) (see Figure 11) by the Diels-Alder reaction with butadiene to form

the adduct (40) shown in Figure 37. The cyclopropene seems likely

to have formed from the singlet state of the carbene, since the car-

bene is almost certainly initially formed in the singlet state, and since

intramolecular reactions seem to be favored for carbenes in the

singlet state. 17' 16 For example, direct irradiation of aliphatic

alpha-diazoketones produces a predominance of the photochemical

Wolff rearrangement; but photosensitized irradiation, which should

increase triplet formation, produces an increased amount of cyclo-

propanes, suggesting normal intermolecular carbene reactions.

Formation of the cyclopropene intermediate (16) is a particu-

larly surprising result, since the (4, 5-c)furotropylidene (11) has not

shown any evidence of rearrangement to the isobenzofuran skeleton

as might have been expected (Figure 11). This rearrangement, if it

does occur, might be impossible to detect with certainty because of

the high reactivity of the isobenzofuran molecule. It polymerizes
rapidly in solution. 4While one cannot say with certainty that none

of the cyclopropene opens to the isobenzofuranyl carbene, the fact

that yields of up to 50 percent of formal furotropylidene addition

products are formed does allow one to say that the rearrangement is

not the overwhelmingly predominant process such as is observed

with the annelated cycloheptatrienylidenes (13), (41), and (42) in

Figure 38. Perhaps the ring-opening of the cyclopropene intermediate

(16) to the isobenzofuranyl carbene is precluded because not enough

aromaticity is gained in that direction.

Why does triplet chemistry predominate in the reactions of

furotropylidene at moderate to higher temperatures? The apparent

ease of crossing from singlet to triplet suggests that these two elec-

tronic states are at very similar energy levels in this carbene. The

effect of temperature in changing the character of the reactions of this

carbene has a few interesting parallels in the literature.

Closs has reported a case in which there may be a temperature
effect upon a singlet-triplet equilibrium. Diphenyl carbene, known

to have a triplet ground state, was produced by irradiation of di-

phenyldiazomethane in the presence of olefins. In reactions with cis-

and trans-2-butenes, cyclopropanes account for no more than 10 per-

cent of the hydrocarbon products. Hydrogen abstraction was the main

reaction pathway. At -100C. the cis-and trans- 2-dimethyl-3, 3-

diphenylcyclopropanes were formed in a ratio of 3. Z from the cis-2-

butene. The corresponding ratio from the trans olefin was 0. 04.

Lower temperatures caused increased stereospecificity. At -660C.

the product ratio from the cis -2-butene was 9. 0. At a given tem-

perature the product ratio was found to be independent of the butene

concentration over a range of 150-fold dilution with cyclohexane. The

presence of oxygen failed to change the isomer ratio of products.

Closs postulated the following scheme (Figure 42) as a possible ex-

planation of his observations. He suggested that intersystem crossing

is much faster than any other reaction in the system and that the

reverse crossing is also very fast so that both singlet and triplet are

effectively in equilibrium. The relative rates of the singlet (kas) and

the triplet (kat) addition steps and the position of the singlet-triplet

equilibrium both determine the fraction of stereospecific singlet-state

addition. Since diphenylmethylene is known to have a triplet ground

state, the rate of crossing to the triplet (ki) must be greater than the

rate of triplet crossing to the singlet (k_i); therefore, in view of the

observed product ratios, the rate of singlet addition (kas) must be

much greater than the rate of triplet addition (kat). If the difference

in the free energies of activation for the two addition reactions is

larger than the free energy difference between the two electronic

states, the temperature difference could be explained on this basis

alone. It is not possible to determine whether a temperature effect

upon the position of singlet-triplet equilibrium is being observed, but

this is a possibility.

k A r-C-Ar k
A Ar A Ar

Ar C --Ar

k \k
at / at

Ar Ar Ar AP

(Figure 42)

Thermal effects upon the population of electronic states are

known in certain photochemically produced noncarbene species. An
example is a study of pyrene-dl0 in a polymethylmethacrylate matrix.

The triplet yield plus the fluorescence yield was near unity at -1960C.

As the temperature was raised, two effects were observed. First,

the triplet yield increased with increasing temperature, suggesting a

temperature-dependent process that produces increasing intersystem

crossing from vibrationally excited singlet to second triplet state

(TZ). The second effect observed was a falling off of the sum of

triplet yield and fluorescence yield from the expected value of unity

as temperature increased. This suggested a thermally dependent

radiationless transition from the first singlet state to the ground

state. The energy of activation for the temperature-dependent com-

ponent of the intersystem-crossing process was determined to be

about 2. 6 kcal, per mole. The energy of activation for the radiation-

less transition from singlet to ground state was about 0. 9 kcal per


A somewhat similar study of 1, 12 benzperylene has shown that,

since the second excited singlet of this molecule lies only about
1300 cm. 1(3, 7 kcal. per mole) above the first excited singlet, there

is significant thermal population of the second excited singlet state

at 23 C.42

Whether furotropylidene is showing a similar thermal effect

upon population of electronic states is not possible to determine so

long as the electronic states themselves cannot be observed except

through their chemistry. This is because the relative rates of re-

action of the electronic species with their trapping agents are


It is possible to draw several speculative schemes that could fit

the presently known facts about (4, 5-c)furotropylidene. Some of

these are shown in abbreviated form in Figure 43. It seems reason-

able to assume that the cyclopropene is lower in energy than the

initially formed singlet carbene. The relative energies of the triplet

and singlet states shown in Figure 43 can only be the subject of




(Figure 43)



speculation from the present data. It is interesting to consider the

question as to whether equilibria exist between the species in Figure

39, but there is no experimental basis for a determination of this

question. A hypothetical experiment can be devised to answer this

question. If one can show that there is X percent formation of cyclo-

propene under a given set of conditions and that there is more than

(100-X) percent of carbene addition observed under the same condi-

tions in the absence of a cyclopropene trap, one could reasonably

conclude that an equilibrium between cyclopropene and singlet carbene

does exist. Such an experiment seems to call for extraordinarily

high yields in these carbene reactions that are unlikely to be attain-

able. In all of these schemes it seems reasonable that singlet

chemistry is not observed via intermolecular olefin trapping, since

the intramolecular reaction to form the cyclopropene would be ex-

pected to be much faster than the intermolecular reaction.

It is interesting to speculate that perhaps triplet chemistry

predominates at higher temperatures because the singlet, through

its aromatic character, is relatively less reactive than triplet, and

therefore has a sufficiently long lifetime to allow intersystem crossing

to occur before singlet reaction occurs. An equilibration between

singlet and triplet, with the triplet the more reactive of the pair,

would also fit the data.

Perhaps the study of minor reaction products of (4, 5-c)furotro-

pylidene would shed additional light on these matters, but the separa-

tion and purification of such large molecules formed in such low yields

presents formidable experimental difficulties.


General. Melting points were taken in a Thomas-Hoover

Unimelt apparatus and are uncorrected. Elemental analyses were

performed by Atlantic Microlab, Incorporated, Atlanta, Georgia.

Accurate mass measurements were provided by Dr. R. W. King,

using the MS-30 high-resolution mass spectrometer equipped with

automatic data system, at the University of Florida. Infrared spectra

were recorded on a Beckman IR-10 spectrophotometer. In all cases

where the liquid film technique was not used, the KBr pellet technique

was used. Nuclear magnetic resonance spectra were determined on

a Varian A-60A high-resolution spectrometer, or in some cases, a

Varian XL-100 instrument. Chemical shifts are reported in tau

values from internal tetramethylsilane standard. Low resolution

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


Analytical thin-layer chromatography was done on 2 in. x 8 in.

plates coated in these laboratories with 0. 25 mm. layers of E. Merck

HF-254 silica gel; preparative work was conducted on 8 in. x 8 in.

plates coated with 1. 0 to 1. 5 mm. layers of HP-254 silica gel. Com-

ponents were visualized by their quenching of fluorescence under

ultraviolet light. Analytical gas chromatography was accomplished

with a Varian Aerograph Series 1200 flame ionization instrument using

a 100-ft. capillary column coated with Ucon LB-550. Analytical

results were obtained by planimetric measurement and by peak height

times peak-width-at-half-height measurement.
r 7

All chemicals are reagent grade used as supplied unless other-

wise stated. The furan-3, 4-dicarboxylic acid was used as supplied

by Aldrich Chemical Company, Milwaukee, Wisconsin. Solvents

were dried by passage through a column of either freshly re-activated

Linde Molecular Sieve (4A) or Woelm basic alumina, activity grade 0,

followed by storage over calcium hydride under a nitrogen atmosphere.

3, 4-Di(hydroxymethyl)furan. This compound has been reported

as the product of the reduction of dimethyl-3, 4-furandicarboxylate. 23

The reported yield of 76 pc recent did not result from use of the pub-

lished procedure. The following procedure gave 72 percent conver-

sion based on the diacid. A mixture of 31.2 g. (0. 2 moles) 3, 4-

furandicarboxylic acid, 47.2 g. (0.4 moles) thionyl chloride, 200 ml.

benzene, and 1 ml. N, N-dimethylformamide was heated at reflux for

1 hr. The reaction is essentially complete when all of the solid has

dissolved. The benzene and excess thionyl chloride were removed in

vacuum by rotary evaporator. The crude diacyl chloride, formed

in essentially quantitative yield, was reduced directly without purifi-

cation using the following procedure. The crude acid chloride was

dissolved in ca. 300 ml. tetrahydrofuran. This solution was dripped

into a stirred suspension of 30 g. lithium aluminum hydride in 800 ml.

dry tetrahydrofuran. The mixture was stirred at room temperature

overnight, then refluxed 8 hr. The reaction mixture was cooled. The

excess hydride was destroyed by addition of about 300 ml. of 5 percent

sodium hydroxide solution that had been saturated with sodium

chloride. The ether layer was separated by decanting from the white

granular slurry. This white residue was washed several times with

diethyl ether. The washings were combined with the first (THF)

extract, washed with brine, dried with anhydrous M.ISO4 and filtered.

Removal of the solvent on a rotary evaporator using aspirator vacuum

gave 26. 4 g. of crude 3, 4-dihydroxymethylfuran. The product was

identified by the correspondence of its spectral properties with the

values reported in the literature.

3, 4-Furandicarboxaldehyde. This compound was prepared from

3, 4-di(hydroxymethyl)furan in two steps by the procedure of Cook and
Forbes. The first step, partial oxidation of the di-alcohol with

activated manganese dioxide, gave yields of about 50 percent instead

of the reported 80 percent. The best yields of dialdehyde were ob-

tained by lead tetra-acetate oxidation of the crude 3-hydroxymethyl-

furan-4-carboxaldehyde containing about 50 percent of unreacted

glycol, rather than by separation and purification of the mono-

aldehyde. This procedure allowed the lead tetraacetate to oxidize,

not only the mono-aldehyde in the mixture, but also the glycol that

had not been oxidized by the manganese dioxide. This required use

of about 50 percent more lead tetraacetate than would have been

required for oxidation of an equal weight of 3-hydroxymethylfuran-4-

carboxaldehyde to the dialdehyde. This procedure gave about 25 per-

cent conversion of the glycol to 3, 4-furandicarboxaldehyde. The

produce was identified by the correspondence of its spectral properties

and melting point with the values reported in the literature by Cook

and Forbes. 23

(4, 5-c)Furotropone. This compound was prepared by condensa-

tion of 3, 4-furandicarboxaldehyde with acetone using the procedure of
Cook and Forbes. The yield and the physical and spectral properties

of the product were exactly as reported.

(4, 5-c)Furotropone tosylhydrazone. A solution of Z. 0 g.

(0. 014 moles) p-toluenesulfonylhydrazine and a trace of phosphoric

acid in 20 ml. of dry tetrahydrofuran was allowed to stand in a

stoppered flask for three to seven days at room temperature. The

solution was diluted with one volume of chloroform and allowed to

stand in a refrigerator cabinet (ca. 5-70C.) for 0. 5 to 1 hr. The

resulting slurry of crystals was poured onto a Buchner filter. The

collected yellow crystals were washed with fresh chloroform on the

filter. The combined wash solvent and mother liquor were eluted

from a column of silica gel (4. 5 x 15 cm. ) using methylene chloride.

The first (yellow) fraction was collected and evaporated to dryness.

The residue was washed with chloroform and filtered. The resulting

second crop of yellow crystals when combined with the first crop on

the filter gave a total of 2. 9 g. (66 percent conversion) of the ketone

tosylhydrazone, m.p. 214-2150C. w. decomposition.

Anal. Calcd for C16H14N203S: C, 61. 13; H, 4.49; N, 8.91.

Found: C, 60.97; H, 4. 54; N, 8.85.
The spectral data were: ir (KBr, cm. ) 3190, 1640, 1595,

1395, 1325, 1162, 1052, 930, 885, 830, 762, 680. nmr (d6-DMSO)

2. 1 to 4. 13 (complex pattern, total 10H), 7. 62 (singlet, 3H).

(4, 5-c)furotropone tosylhydrazone, sodium salt. A solution of

3. 9 g. furotropone tosylhydrazone in 100 ml. dry tetrahydrofuran was

stirred under dry nitrogen while 0. 5 g. sodium hydride (washed with

pentane) was added. After 0. 5 to 1. 0 hr. at room temperature, 50-75

ml. pentane was added to the reaction mixture. The resulting slurry

of yellow solid was filtered in a dry nitrogen atmosphere (dry box) to

recover 5. 2 g. of the sodium salt.

Decomposition of tosylhydrazone salt in presence of benzene.

(4, 5-c)Furotroponetosylhydrazone sodium salt (0.3 g., 0.7 mmole)

was stirred with 50 ml. benzene in a sealed Fischer-Porter Aerosol

Compatibility Test Tube (containing an atmosphere of dry nitrogen)

and heated in an oil bath kept at 1180C. After 1 hr. the tube was

cooled and opened. The dark brown slurry was taken from the tube

and filtered through a sintered glass funnel. The solid filter cake

weighed 0. 24 g. The filtrate, upon evaporation of the benzene, left

a residue of 0. 14 g. This crude residue was chromatographed on

preparative silica gel plates developed with hexane containing 5-10

percent benzene. The leading band of the chromatogram was col-

lected, stripped from the adsorbent with ethanol, and recovered by

evaporating the filtered solution. This resulted in collection of 0. 063

g. of the benzene insertion product (25), m. p. 75-770 C.

Anal. Calcd for C15H12O: C, 86. 49; H, 5.82. Found: C, 86. 38;

H, 5.85.
The spectral data were: ir (KBr, cm ) 1595, 1490, 1450,

1123, 1040, 872, 852, 800, 797, 762, 700; nmr (CDC13) 2.71 and 2. 72

(two singlets, total 7H), 3.7 (complex, 2H), 4.5 (complex, 2H), 5.64

(complex, 1H); mass spectrum (70 eV) 208 (molecular ion), 131, 77.

Decomposition of tosylhydrazone salt in equimolar benzene-d6-

benzene. A repeat of the above preparation in the presence of an

equimolar mixture of benzene and hexadeuterated benzene produced

a 50:50 mixture of the benzene insertion product and the deuterated

benzene insertion product as determined by nmr (100 MHz.) and by

mass spectroscopy.

Decomposition of tosylhydrazone salt in presence of styrene. A

solution of 0. 42 g. (4 mmoles) styrene in 15 ml. dry dioxane was

heated to 1000C. in a flask equipped with thermometer, stirring bar,

and an inlet for dry nitrogen. Dry solid tosylhydrazone salt (0. 33 g.,

0. 8 mmoles) was added to the solution all at once. After 0. 3 hr. the

reaction mixture was quickly cooled in an ice bath as stirring was

continued. The crude brown slurry in the flask was removed and

filtered, then treated on a rotary evaporator to remove the dioxane

and as much styrene as possible. The resulting residue was dis-

solved in chloroform and streaked on a preparative silica gel plate.

Development of the plate in a mixture of hexane and chloroform gave

0. 07 g. of somewhat impure spiro adduct in the major band. This

material was purified by repetition of the silica gel chromatography

using hexane as the solvent for development of the plate. This gave

0. 06 g. of the oily liquid phenylspirocyclopropane (26), conversion

32 percent.

Anal. Calcd for C17H140: C, 87. 13; iH, 6.03. Found: C,

86.81; H, 6. 01.

The spectral data were: ir (film, cm.-) 3130, 3080, 3060,

3020, 2995, 1662, 1600, 1540, 1495, 1450, 1410, 1210, 1125, 1047,

980, 875, 855, 815, 790, 698; nmr (CC14) 2.86 (singlet, 5H), 2.96

and 3.01 (two singlets, total 2H), 7.55-7.82 (complex, 1H), 8.5-8.76

(complex, 2H); mass spectrum (70eV) 234 (molecular ion), 216, 205,

191, 130, 128.

Thermal decomposition of tosylhydrazone salt in presence of

trans-deuteriostyrene. The above preparation was repeated using

trans-deuteriostyrene in place of styrene. Examination of the nmr

spectrum showed that the product consisted of equal parts of the cis

and trans cyclopropanes. The spectrum showed a simplified ABX

pattern as described in the text of this report.

Photolytic decomposition of tosylhydrazone salt in presence of

trans-deuteriostyrene. A solution prepared as in the experiment

above was irradiated in a sealed tube (magnetically stirred) with two

Sears-Roebuck sunlamps at a distance of approximately 8-10 inches.

During the reaction and the workup the product was not exposed to

temperatures exceeding 500 C. The resulting phenylcyclopropane

consisted of equal parts of the cis and trans products as shown by


Test of the thermal and photolytic stability of trans-deuterio-

styrene. A small sample of trans -deuteriostyrene in an nmr sample

tube was heated in a steam cone for 0. 5 hr. The nmr spectrum was

unchanged by the heating. The sample was also unchanged after it

was irradiated by two Sears Roebuck sunlamps for 0. 75 hr.

Decomposition of tosylhydrazone salt in presence of 1-butene.

The salt (0.2 g. 0.48 mmoles) was heated with 5 g. 1-butene (liquid)

that had been distilled into a Fisher-Porter Aerosol Compatibility

Test Tube. The tube was kept in an oil bath at 1100C. for 1 hr. The

excess 1-butene was then released. The crude residue was slurried

with benzene and filtered through a sintered glass filter. The solid

filter cake weighed 0. 14 g. The crude filtrate left a residue of 0. 05 g.

after evaporation of the benzene. This residue (about 90 percent

pure) afforded the ethyl spirocyclopropane (27) after purification by

preparative vapor phase chromatography on an 8 ft. x 1/4 in. column

packed with 60/80 mesh Anakrom ABS coated with 20 percent w/w


Anal. High resolution mass spectroscopy (70 eV): Calcd for

C13H140: 186. 1044. Found: 186. 1036.

The spectral data were: ir (liquid film, cm. ), 3145, 3070,

3035, 3000, 2975, 2940, 2880, 1670, 1540, 1470, 1460, 1132, 1050,

980, 880, 850, 810; nmr (CC14) s. 95 (singlet, ZH), 3. 87-4. 3 (over-

lapping doublets, total 2H), 5.25-5.8 (complex, ZH), 8. 3-9. 5 (com-

plex, 8H); mass spectrum (70 eV) 186 (molecular ion), 171 (C12H110),

158.07 (CI1H100), 157.06 (C11H90), 144. 05 (C10H80), 130.04


Decomposition of tosylhydrazone salt in presence of isobutene.

The salt (0. 3 g., 0. 7 mmoles) was heated with ca. 4 g. liquid iso-

butene in a sealed Fisher-Porter Aerosol Compatibility Test Tube in

an oil bath at 1120C. for 1-1.5 hr. The excess isobutene was then

released to cool the contents of the tube. The crude residue that

remained was slurried in benzene and filtered. The solid filter cake

weighed 0. 2 g. The filtrate, after evaporation of the benzene,

weighed 0. 038 g. Purification of this residue by taking the leading

band on a thin-layer plate (silica gel) developed in hexane gave

0. 017 g. of the purified dimethyl spirocyclopropane (28). The high

purity of the crude product, as shown by its nmr spectrum, suggests

that a large loss of material occurred during handling that was not

attributable merely to purification.

Anal. High resolution mass spectroscopy (70 eV): Calcd for

C13H140: 186. 1044. Found: 186. 1060.

The spectral data were: ir (liquid film, cm. -) 3055, 3030,

2980, 1770, 1725, 1540, 1440, 1365, 1130, 1110, 1050, 880, 825;

nmr (CC14) 2.83 (singlet, 2H), 3.7-3.9 doublett, 2H), 4.9-5.2

doublett, ZH), 8. 9 (singlet, 6H), 9. 2 (singlet, 2HO; mass spectrum

(70 eV) 186. 10 (molecular ion), 185. 10 (C13H130), 172. 08

(C12H110), 158.07 (C11H100), 157. 06 (C11H90), 144. 05 (C10H80).

Thermal decomposition of tosylhydrazone salt in presence of

cis- and trans-2-butencs. The same pyrolysis technique described

above was used to decompose samples of the tosylhydrazone salt in

the presence of cis- and trans-2-butenes. The resulting crude re-

action mixtures had essentially identical nmr spectra and gas chro-

matograms (capillary column, Ucon LB-550). Pyrolysis of a 0. 3-g.

sample (0. 72 mmoles) of the salt with 15 ml. liquid trans-2-butene

at 1180C. produced a crude product weighing 0. 09 g. Careful

preparative layer chromatography (silica gel adsorbent, hexane

solvent) of this material at low plate loadings gave 0. 03 g. of trans-

dimethylspirocyclopropane (34), 23 percent conversion.

Anal. High resolution mass spectroscopy (70 eV): Calcd for

C13H140: 186. 1044. Found: 186. 1052.
The spectral data were: ir (liquid film, cm. -) 3030, 3000,

2960, 2935, 2855, 1665, 1540, 1455, 1387, 1130, 1088, 1050, 880,

810; nmr (CC14) 2.88 (singlet, ZH), 3.9 doublett, 2H), 5.3 doublett,

2H), 8. 75-9. 0 (three sharp peaks, total 6H), 9. 0-9. 5 (complex, 2H);

mass spectrum (70 eV) 186. 10 (molecular ion), 171.08 (C12H110),

158. 07 (C11H100), 157. 06 (C11H90), 144. 06 (C 1H80), 128.06

(C 10H8).

Photolysis of tosylhydrazone salt in presence of cis-2-butene.

The photolytic decomposition of 0, 33 g. tosylhydrazone salt with 12 g.

cis-2-butene was carried out by irradiating the stirred slurry in a

sealed tube for 1 hr. using two Sears-Roebuck sunlamps at a distance

of about 10-12 inches. This procedure produced a crude product

mixture that gave an nmr spectrum and gas chromatogram that were

essentially identical to those produced by the thermal decomposition

of the salt in the presence of cis- and trans-2-butenes described


Photolysis of tosylhydrazone salt in presence of trans-2-butene

at low temperature. Photolytic decomposition of 0. 4 g. tosylhydra-

zone salt by irradiation for 1 hr. with a Hanovia 550-watt mercury

lamp at a temperature of -300 C. produced a crude reaction mixture

that contained no cyclopropane (34) as determined by nmr.

Determination of relative rates of reaction with various olefins.

Relative rates of reaction with various olefins were determined using

the pyrolysis method in a sealed tube as previously described. The

temperature of the oil bath was kept at 1180C. for all runs. In each

run a comparison of product formation from each of two olefins was

done. Each olefin was present in equimolar amounts, measured by

condensing equal volumes of the gaseous olefins into the reaction

tube by use of a mercury-filled gas buret. The product ratios were

determined by capillary column gas chromatography as described

under the General heading of this section. The results are presented

in Table I, page 31.

Pyrolysis with 1, 3-butadiene at 1100 C. Furotropone tosylhy-

drazone salt (0. 25 g., 0. 6 mmoles) was heated with ca. 20 ml. liquid

1, 3-butadiene in a sealed Fisher-Porter Aerosol Compatibility Test

Tube in an oil bath kept at 1100 C. for 4 hr. Excess butadiene was

vented to the atmosphere after the tube was removed from the bath

and opened. The residue that remained in the tube was slurried in

benzene and filtered through sintered glass. The clear amber ben-

zene solution was streaked on a preparative layer plate (silica gel)

that was developed with hexane. The leading band of material gave

0. 06 g. of the 1, 4-adduct of butadiene (33), 55 percent yield. A

small band of material following the 1, 4-adduct was too small for


Anal. Calcd for C13H120: C, 84. 75; H, 6.57. Found: C,

84. 49; H, 6. 65.
The spectral data were: ir (liquid film, cm. -) 3060, 3020,

2930, 2850, 1618, 1540, 1440, 1340, 1132, 1052, 948, 882, 848,

800, 670; nmr (CC14) 2. 82 (singlet, ZHO, 3. 8-4. 7 (two doublets with

overlapping signal, total 6H), 7. 5 (singlet, 4H); mass spectrum

(70 eV) 184 (molecular ion), 169, 155, 130, 129, 128, 54.

Pyrolysis with 1, 3-butadiene at 1000C. The tosylhydrazone

salt (0. 36 g. 0. 86 mmoles) was pyrolyzed with 1, 3-butadiene by the

above-described method using an oil bath temperature of 1000C. and

reaction time of 0. 5 hr. Similar workup and chromatography showed

only a very weak leading band corresponding to the 1, 4-adduct (33)

(0. 01 g.), followed by a second band that afforded 0. 025 g. of the 1, 2-

adduct, the vinyl cyclopropane (31). A yield figure is not given in

this reaction because the short reaction time and low temperature

probably did not decompose all of the sodium salt.

Anal. Analysis was done by thermal isomerization of the 1, 2-

adduct to the known 1, 4-adduct by heating it at 1300 C. for 0. 5 hr.
The spectral data were: ir (liquid film, cm. ) 3140, 3090,

3010, 1542, 1217, 1132, 1052, 1000, 910, 880, 813, 790; nmr (CC14)

3.0 (singlet, sH), 4-5. 7 (complex, 7H), 8. 2-9. 1 (ABX pattern, 3H).

An impurity gave a singlet at 8. 64.

Photolysis of tosylhydrazone salt with 1, 3-butadiene at low

temperature. The tosylhydrazone salt (0. 4 g., 0. 98 mmoles) was

photolyzed in a stirred reactor at -400 to -500C. using the Hanovia

550-watt lamp for 2 hr. Workup, including thin-layer chromatog-

raphy, as described before afforded 0. 07 g. of the Diels-Alder adduct

(40), 39 percent conversion.

Anal. High resolution mass spectroscopy (70 eV): Calcd for

C13H120: 184.0887. Found: 184.0880.

The spectral data were: ir (liquid film, cm. -), 3040, 2960,

2880, 2840, 1630, 1430, 1280, 1220, 1050, 1110, 1025, 892, 860,

780, 740; nmr (CC14) 2. 83 (singlet, 2H), 4. 0 doublett, 2H), 4. 47

(broad, 211), 7. 52 (broad, 4H), 7. 85 doublett, 1H), 9. 15-9.45

multiplee, 1H); mass spectrum (70 eV) 184. 0887 (molecular ion),

182. 073 (C13H100), 169. 065 (C12H90), 168.057 (C12H80), 165.070

(C 13H90).

Photolytic stability of vinylcyclopropane (31). A sample of the

vinylcyclopropane (31) was irradiated with two Sears-Roebuck sun-

lamps for 0. 5-0. 75 hr. It was unchanged after irradiation.

Photolysis of tosylhydrazone salt with 1, 3-butadiene at 400 C. A

small-scale photolysis (ca. 25 mg. salt) was run in the presence of

1, 3-butadiene at 400 C. The product ratio was determined by a gas

chromatographic analysis (see General heading) of the crude product,

followed by another similar analysis after removal of all 1, 2-adduct

by thin-layer chromatography. The result showed that the Diels-Alder

adduct (40) and the vinylcyclopropane (31) were present in a 45:55

ratio with none of the 1, 4-adduct (33) present.


CIDNP Experiment. A saturated solution of the tosylhydrazone

salt in an nmr tube containing a solution of ca. 20 percent cyclohexene

in d6-DMSO was heated in the variable temperature probe of the

Varian A-60A at 1200 C. for 10 min. No change in the spectrum

was detected before, during, and after heating.


1. C. L. Ennis, Ph.D. Dissertation, University of Florida,
March, 1968.

2. W. M. Jones and C. Lawrence Ennis, J. Am. Chem. Soc.,
91, 6391 (1969).

3. R. Gleiter and R. Hoffmann, J. Am. Chem. Soc., 90,
5457 (1968).

4. W. M. Jones, M. E. Stowe, E. E. Wells, Jr., and
E. W. Lester, J. Am. Chem. Soc., 90, 1849 (1968).

5. P. H. Gebert, Ph.D. Dissertation, University of Florida,
March, 1972.

6. L. W. Christensen, E. E. Waali, and W. M. Jones, J.
Am. Chem. Soc., 94, 2118 (1972).

7. D. Seyferth, J. Y. -P. Mui, and R. Damrauer, J. Am.
Chem. Soc., 90, 6182 (1968).

8. J. E. Baldwin and R. A. Smith, J. Am. Chem. Soc., 89,
1886 (1967).

9. W. M. Jones, Burrell N. Hamon, Robert C. Joines, and
C. L. Ennis, Tetrahedron Lett., 3909 (1969).

10. I. Moritani et al., Tetrahedron Lett., 373 (1966).

11. I. Moritani et al., J. Am. Chem. Soc., 89, 1259 (1967).

12. S. I. Murahashi, I. Moritani, and M. Nishino, J. Am.
Chem. Soc., 89, 1257 (1967).

13. K. E. Krajca, Tsutomu Mitsuhasni, and W. M. Jones,
J. Am. Chem. Soc., 94, 3661 (1972).

14. K. E. Krajca, Ph. D. Dissertation, University of Florida,
August, 1972.

15. Thomas Coburn, Ph.D. Dissertation, University of Florida,
August, 1973.

16. P. S. Skell, Accounts of Chemical Research, 6, 97 (1973).

17. Maitland Jones, Jr. and Wataru Ando, J. Am. Chem. Soc.,
90, 2200 (1968).

18. W. M. Jones and J. P. Mykytka, unpublished results, 1973.

19. A. Streitwieser, Jr., "Molecular Orbital Theory for Organic
Chemists, John Wiley and Sons, New York, N. Y., 1961,p.357.

20. W. M. Jones et al., J. Am. Chem. Soc. 95, 826 (1973).

21. Domenick J. Bertelli and Thomas G. Andrews, Jr., J. Am.
Chem. Soc., 91, 5280 (1969).

22. Domenick J. Bertelli, Thomas G. Andrews, Jr. and Phillip O.
Crews, J. Am. Chem. Soc., 91, 5286 (1969).

23. M. J. Cook and E. J. Forbes, Tetrahedron, 24, 4501 (1968)
and references cited therein.

24. Ronald N. Warrener, J. Am. Chem. Soc., 93, 2346 (1971).

25. 0. L. Chapman, J. Am. Chem. Soc., 85, 2014 (1963).

26. C. N. Banwell, in "Nuclear Magnetic Resonance for Organic
Chemists," D. W. Mathieson, Ed., Academic Press, New York,
N. Y., 1967, p. 85.

27. Maitland Jones, Jr., Arnold M. Harrison, and Kenneth R.
Rettig, J. Am. Chem. Soc., 91, 7462 (1969).

28. P. S. Skell and R. C. Woodworth, J. Am. Chem. Soc., 78,
4496 (1956).

29. G. L. Closs, in "Topics in Stereochemistry, Vol. 3, E. L.
Eliel and N. L. Allinger, Eds., Interscience Publishers, New
York, N. Y., 1968, p. 226.

30. George Zweifel, G. M. Clark, and N. L. Polston, J. Am. Chem.
Soc., 93, 3395 (1971).

31. E. E. Waali and W. M. Jones, J. Am. Chem. Soc., in press.

32. Maitland Jones, Jr., Wataru Ando, Michael E. Hendrick,
Anthony Kulczychi, Jr., Peter M. Howley, Karl F. Hummel,
and Donald S. Malament, J. Am. Chem. Soc., 94, 7469 (1972).

33. K. B. Wiberg and B. J. Nist, J. Am. Chem. Soc., 83, 1226

34. L. J. Bellamy, "Advances in Infrared Group Frequencies,"
Meuthen and Co., Ltd., London, 1968, p. 24.

35. L. J. Bellamy, "The I. R. Spectra of Complex Molecules,"
John Wiley and Sons, New York, N. Y., 1954, p. 31.

36. Harold R. Ward, in "Free Radicals," J. K. Kochi, Ed., John
Wiley and Sons, New York, N. Y., 1973, p. 239.

37. Peter P. Gasper and George S. Hammond, in "Carbene
Chemistry," Wolfgang Kirmse, Ed., Academic Press, New
York, N. Y., 1964, p. 270.

38. W. J. Baron et al., in "Carbenes, Vol. 1, Maitland Jones,
Jr. and Robert A. Moss, Eds. John Wiley and Sons, New
York, N. Y., 1973, p. 81.

39. K. U. Ingold, in "Free Radicals, J. K. Kochi, Ed., John
Wiley and Sons, New York, N. Y., 1973, p. 92.

40. G. L. Closs, in "Topics in Stereochemistry, Vol. 3, E. L.
Eliel and N. L. Allinger, Eds., Interscience Publishers, New
York, N. Y., 1968, p. 224.

41. J. L. Kropp, W. R. Dawson, and M. W. Windsor, J. Phys.
Chem., 73, 1752 (1969).

42. W. R. Dawson and J. L. Kropp, J. Phys. Chem., 73, 1752


Thomas Howard Ledford was born August 24, 1942, in Macon,

Georgia, to Mr. and Mrs. Howard William Ledford. He was gradu-

ated from Swainsboro High School, Swainsboro, Georgia, in 1960

and entered the University of Georgia as a four-year General Motors

Scholar that September. While there he was elected to Phi Beta

Kappa and received the Merck Award and the American Institute of

Chemists Award. He obtained the degree of Bachelor of Science in

Chemistry in June, 1964. The period 1965-1968 was spent in indus-

trial research in organic chemistry with Tennessee Eastman

Company, Kingsport, Tennessee. In 1968 he enrolled in the Graduate

School of the University of Florida with a Woodrow Wilson National

Fellowship. He was also a Graduate School Fellow during his

graduate study. He is a member of the American Chemical Society

and Phi Beta Kappa.

Mr. Ledford is married to the former Joan McDaniel of Oneonta,

Alabama. He will be working for the Esso Research Laboratories

of Exxon, U.S.A., in Baton Rouge, Louisiana.

I certify that I have read this study and that in my opinion it con-
forms to acceptable standards of scholarly presentation and is fully ade-
quate, in scope and quality, as a dissertation for the degree of Doctor
of Philosophy.

William M. Jones
Professor of Chemistry

I certify that I have read this study and that in my opinion it con-
forms to acceptable standards of scholarly presentation and is fully ade-
quate, 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 con-
forms to acceptable standards of scholarly presentation and is fully ade-
quate, in scope and quality, as a dissertation for the degree of Doctor
of Philosophy.

George B. Butler
Professor of Chemistry

I certify that I have read this study and that in my opinion it con-
forms to acceptable standards of scholarly presentation and is fully ade-
quate, in scope and quality, as a dissertation for the degree of Doctor
of Philosophy.

RogefjG. Bates
Professor of Chemistry

I certify that I have read this study and that in my opinion it con-
forms to acceptable standards of scholarly presentation and is fully ade-
quate, in scope and quality, as a dissertation for the degree of Doctor
of Philosophy.

,.- L .

Richard H. Hammer
Associate Professor of Pharmaceu-
tical Chemistry

This dissertation was submitted to the Department of Chemistry in the
College of Arts and Sciences and to the Graduate Council, and was ac-
cepted as partial fulfillment of the requirements for the degree of Doc-
tor of Philosophy.

August, 1973

Dean, Graduate School


3 1262 08556 7443