Migratory Aptitudes in the Elimination- Rearrangement
Reactions of 1, 1-Diaryl-2-Bromoethylenes
RALPH ANTHONY DAMICO, JR.
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
The author wishes to express his gratitude to the faculty
members of the Chemistry Department, especially those persons who
served on his graduate committee, and to his student associates
for their many helpful suggestions and advice.
He is especially indebted to Dr. W. M. Jones, Chairman of
the author's supervisory committee, for his advice, assistance and
The author also wishes to thank Miss Jan Brockett for typing
and drawing the figures of this dissertation, and Texaco, Inc., for
their financial assistance.
TABLE OF CONTENTS
ACKNOWLEDGEMENTS . .
LIST OF TABLES . .
LIST OF FIGURES . .
INTRODUCTION . . . .
EXPERIMENTAL . . . .
A. Syntheses and Reactions
B. Spectra . . . .
C. Kinetic Methods . .
D. Dipole Moments . . .
RESULTS . . . . .
DISCUSSION OF RESULTS . .
SU Y . . . . .
REFERENCES . . . . .
. . . . . . . 1
. . . . . . . 4
. . . . . . . 13
. . . . . . . 23
. . . . . . . 27
. . . . . . . 31
. . . . . . . 43
. . . . . . 62
. . . . . . . 63
. . . . . . . . . . . 66
LIST OF TABLES
1. Data for the calculation of cis-l-p-chlorophenyl-1-
phenyl-2-brmnoethylene in trans-l-p-chlorophenyl-1
phenyl-2-brmnoethylene . . . . . . . . 9
2. Absorption Maxima of Compounds Used in Kinetic Measure-
ments . . . . . . . . . . . . 19
3. Extinction Coefficient Determination for 1-p-Methoxy-
phenyl-2-phenylacetylene . . . . . . . . 20
4. Extinction Coefficient Determination for trans-l-p-
Methoxyphenyl-l-phenyl-2-broaoethylene . . . .. 20
5. Extinction Coefficient Determination for cis-l-p-
Methoxyphenyl-l-phenyl-2-bromoethylene . . . . 21
6. Extinction Coefficient Determination for 1-p-Chloro-
phenyl-2-phenylacetylene . . . . . . . . 21
7. Extinction Coefficient Determination for Diphenylacety-
lene . . . . . . . . . . . . . 22
8. Quantitative Analysis of 1,1-Diphenyl-2-brcnoethylene
and Diphenylacetylene Mixtures by Vapor-phase Chraooto-
graphy . . . . . . . . . . . . 26
9. Data for the Determination of Dipole Moments of cis-
and trans-l-p-Methoxyphenyl-l-phenyl-2-bromoethylene . 32
10. Kinetic Data for the Reaction of trans-l-p-Methoxy-
phenyl-l-phenyl-2-bromoethylene with 0.20 M Potassium-
t-butoxide at 950 . . . . . . . . . 37
11. Summary of Kinetic Data for the Reactions of 1,1-
Diaryl-2-bromoethylenes (EH) with Potassium-t-butoxide
(KB) at 950 . . . . . . . . . . 38
12. Average Rate Constants of the Reactions of 1-Phenyl-
1-p-substituted Phenyl-2-bramoethylenes with 0.20 M
Potassium-t-butoxide . . . . . . . . . 52
LIST OF FIGURES
1. Determination of o0 for cis- and trans-1-p-
Methoxyphenyl-1-phenyl-2-bromoethylene . . . 33
2. Determination of 6/ for cis- and trans-i-p-
Methoxyphenyl-1-phenyl-2-bromoethylene . . . 34
3. Reaction of trans- and cis-1-j-Methoxyphenyl
1-phenyl-2-bromoethylene with 0.197 M Potas-
sium-t-butoxide at 950 . . . . . . .. 40
4. Comparison of Rate Constants of 1,1-Diaryl-2-
bromoethylenes at 0.20 M Potassium-t-butoxide . 41
The elimination and rearrangement reactions of 1,1-diaryl-2-
haloethylenes with strong bases to yield diarylacetylenes (tolans)
was first reported by Fritsch [l1, Buttenberg [2 and Wiechell [3 .
C = C + B~ -- AR -C=C- AR + BH + X-
They used ethanolic solutions of sodium ethoxide at 180-200o
to promote reactions. More recently, potassium amide in liquid am-
monia [4,5 butyl- and phenyl-lithium in ether 6,7 and potassium-
t-butoxide in t-butyl alcohol [8,9] have been used to produce high
yields of tolans. Due to the general base initiation, the first step
in the reaction is believed to be the abstraction of a proton to form
the diaryl vinyl anion.
AR H AR
\ / \
C =C + B -*- C ==C- + BH
/ \ / \
AR X AR X
Curtin and Flynn  have proposed this to be the rate deter-
mining step in their system in which they used butyl-lithium as the
base, while Pritchard and Bothner-By  have shown it to be a rapid
equilibrium reaction under their conditions with potassium-t-butoxide.
The rearrangement step has been shown to be stereospecific.
In the reaction of cis- and trans-l-p-bromophenyl-l-phenyl-2-bromoethy-
lene with potassium-t-butoxide, the group trans to the leaving bromide
ion migrates preferentially . Cis- and trans-l-p-chlorophenyl-1-
phenyl-2-bromnoethylene rearranges similarly when promoted by butyl-
lithium [10|. These results exclude the possibility of a carbene
being involved in the major reaction and strongly suggest a geometri-
cally stable vinyl carbanion as an intermediate in the reaction.
c =-C: C =C-
/ / \
Ar AR Br
The effect of the leaving halogen on the rate of the reaction
is known . The purpose of our work was to make a kinetic study
of the migratory aptitudes of phenyls containing various substituents.
For this purpose the cis and trans isomers of 1-2-methoxyphenyl-l-
phenyl-2-bromoethylene (cis- and trans-I), 1-p-chlorophenyl-l-phenyl-
2-bromoethylene (cis- and trans-II) and 1,1-diphenyl-2-bromoethylene
(III) were prepared. Dipole moment studies were also undertaken to
determine the configurations of the high and low melting isomers of
Potassium-t-butoxide-t-butyl alcohol was chosen as our reac-
tion medium for several reasons. Proton transfer is known to be a
rapid step under these conditions [91 leading to the presumption that
substituted phenyl groups will change the rate of the reaction as
compared to phenyl groups. Few, if any, side reactions occur with
this base and relatively simple analysis of products can be obtained.
A. Syntheses and Reactions
This compound was prepared by adapting the method used by
Nielson and McEwen 1n] in the preparation of a similar compound.
Twenty-four grams (1 mole) of magnesium was placed in a three-necked
one-liter flask equipped with a reflux condenser and a stirrer.
Seventy-eight grams (0.5 mole) of dry, freshly distilled bromoben-
zene in 200 ml. of dry ether was added dropwise over a two-hour
period. The reaction started with stirring after a few milliliters
of bromobenzene had been added. The reaction mixture was then re-
fluxed for one hour, cooled, and 50 g. (0.32 mole) of p-methoxy-
acetophenone in 50 ml. of dry ether added over one hour. After re-
fluxing for one hour the reaction mixture was hydrolyzed with 30 g.
(0.565 mole) of ammonium chloride in 60 ml. of water. The organic
and aqueous layers were separated and the aqueous layer extracted with
ether. The organic layers were combined. The ether was removed by
distillation and 100 ml. of 20 percent sulfuric acid was added to the
cooled mixture. After refluxing for one hour the organic layer was
separated. Yellow crystals appeared upon cooling the organic layer.
Recrystallization from 95 percent ethanol* gave 40.8 g. (0.178 mole,
*Ninety-five percent ethanol will be called ethanol in this text.
55.5 percent) of white crystals, melting point* 74-76o (lit.  750).
Repeating this experiment under essentially the same conditions af-
forded a yield of 53 percent.
cis- and trans-l-p-Methoxyphenyl-l-phenyl-2-bromoethylene (cis- and
Thirteen grams (0.064 mole) of 1-p-methoxyphenyl-l-phenyl-
ethylene was brominated with 10.5 g. (0.065 mole) of I-ronine in 50 ml.
of carbon disulfide. The addition took forty-five minutes and was
allowed to reflux with no aid. Hydrogen bromide was liberated after
about 20 ml. of the bromine solution had been added. The mixture
was then refluxed overnight. Carbon disulfide was removed by distil-
lation and the residual oil dissolved in ethanol. White crystalline
needles were obtained upon cooling. These were removed and the solu-
tion set aside for one week, during which time more crystals came out
of solution. A total of 4.45 g. (0.015 mole, 23.4 percent) of the
trans** isomer was obtained. These were recrystallized from ethanol,
m.p. 81-82o (lit  82.50)
The resultant filtrate, from which the trans isomer had been
removed, was passed through a 1 x 30 cm. column of alumina eluting
with ethanol. Twenty milliliter samples were collected and allowed
to evaporate at room temperature. Yellow oil and white crystalline
flakes came out of solution and were separated by filtering.
*All melting points and boiling points are uncorrected.
**The trans configuration was assigned by Stoermer and Simon 12
and confirmed by dipole moment measurement (see discussion) by the author.
Recrystallizing the white crystals from ethanol gave 1.05 g. (0.0036
mole, 5.5 percent) of cis isomer, m.p. 55-560 (lit.  520).
1,1-diphenylethylene was prepared according to the method of
Allen and Converse .
The bromination of 1,1-diphenylethylene was performed using
the procedure of E. Hepp [14 .
The product was recrystallized from methanol to yield very
white crystalline flakes, m.p. 42-430 (lit.  400).
This compound was prepared using a modification of Curtin's
 procedure. Seventy-two grams (3 moles) of magnesium was placed
in a dry two-liter three-necked flask. Approximately 143 g. (1.5
moles) of methyl bromide in 500 ml. of dry ether was added over three
hours. The reaction was allowed to reflux without external heating.
At the end of this time an ether suspension of 150 g. (0.695 mole)
of p-chlorobenzophenone was added in small portions. The reaction
was refluxed for two hours and then hydrolyzed with 53.5 g. (1 mole)
of ainonium chloride in 200 ml. of water. After separating the
layers and washing the organic layer with water, the ether was re-
moved by distillation and the resultant carbinol was dehydrated with
200 ml. of 20 percent sulfuric acid. After separating the organic
and aqueous layers the organic layer was dried and the remaining oil
fractionated under reduced pressure through a 6-in. column packed with
glass helicies. One hundred grams (0.45 mole, 66.5 percent) of 1-p-
chlorophenyl-l-phenylethylene, b.p. 151-158o (5-8mm.) lit. [10j b.p.
153-158o (10 mm.), was obtained.
cis- and trans-l-r-Chlorophenyl-l-phenyl-2-bromoethylene (cis- and
Fifty grams (0.232 mole) of 1-r-chlorophenyl-l-phenyl-ethylene
was brominated with 44.2 g. (0.245 mole, 6 percent excess) of bromine
in 200 ml. of carbon tetrachloride at room temperature. The mixture
was refluxed for eighteen hours to eliminate hydrogen bromide. After
stripping off the carbon tetrachloride more hydrogen bromide was
evolved during the distillation of the mixture. The resultant oil
was then refluxed with 14 g. (0.25 mole) of potassium hydroxide
dissolved in 100 ml. of absolute alcohol. After refluxing for one
hour, water was added and two layers separated. The water layer was
extracted with ether and the organic layers combined and dried with
anhydrous sodium sulfate. The ether was removed on a steam-bath
and the remaining oil fractionated through a 6-in. column packed
with glass helices. The middle fraction was collected, b.p. 149-155o
(1-1.3 mn.), to yield a total of 49 g. (0.167 mole, 72 percent) of
colorless liquid. On standing at room temperature white crystals
formed. These were removed and the filtrate allowed to remain one
week in the refrigerator. The crystalline product was recrystallized
from ethanol, m.p. 88-89O (lit.  88-890), and has the cis con-
figuration as assigned by Bergmann 15].
The remaining filtrate was then redistilled and a slightly
yellow oil collected, b.p. 155-157o (1-1.3 mm). Infrared spectra in-
dicated this product had not more than 30 percent of cis-II as an im-
purity (Table 1). This was made on the basis of the absorption at
723 ca71 where the cis isomer has a strong absorption. Attempts to
crystallize the oil from pentane both before and after passage through
an alumina column proved fruitless (lit. [15 m.p. 43-44o).
Analysis: Calculated for C14H10BrCl: C,57.27; H, 3.43.
Found: C, 57.45; H, 3.29
Diphenylacetylene was prepared according to the method de-
scribed in Organic Syntheses .
This compound was obtained impure using a modification of the
method of Orekhoff and Tiffeneau 17 .
To an ether solution of benzyl magnesium chloride, prepared
from 22.4 g. (0.18 mole) of benzyl chloride and 3.4 g. (0.14 mole)
of magnesium, 14 g. (0.10 mole) of p-methoxybenzaldehyde was added
over thirty minutes. The solution was refluxed for one hour, cooled,
and hydrolyzed with 40 ml. of saturated ammonium chloride solution.
The ether and water layers were separated, the water layer extracted
with ether and the organic layers combined. The ether was allowed
to evaporate and the residual oil was then dehydrated with 50 ml. of
DATA FOR THE CALCULATION OF cis-l-p-CHLOROPHENYL-1-
PHENYL-2-BROMOETHYLENE IN trans- l-P-CfHLOR0PI1MNYL-1-
1.14 x 10-1
9.00 x 10-2
7.89 x 10-2
1.85 x 10-la
a impure trans-II isomer.
Extinction Coefficient = EAvg.= 44.1
Maximum Amount of cis-II in trans-II
+ 2.0 liters/mole-cm.
= 0.120/(42.1)(0.0518) = 0.0551
Maximum Percent of cis-II in trans-II = (0.0551/0.185)(100) = 29.8
30 percent sulfuric acid. After refluxing for fifteen hours the
mixture was poured into ice water and pale yellow crystals separated.
These were recrystalized from ethanol, m.p. 125-130o (lit. 17]
137-1380). This product was assumed to be 1-k-methoxyphenyl-2-
Six grams (0.028 mole) of this product was brominated with
4.5 g. (0.028 mole) of bromine in 100 ml. of chloroform for forty-
five minutes at room temperature. After the addition was completed
the chloroform was removed on a steam-bath and a red solid resulted.
A total of 7 g. (0.19 mole, 67.7 percent) of 1-p-methoxyphenyl-2-
phenyl-1, 2 dibromoethane was obtained, m.p. 170-1750 (lit. 17]
In order to effect the elimination of two moles of hydrogen
bromide from this compound a variety of bases were used. Two and
one-half grams (6.7 moles) of the dibromo compound was refluxed
for five hours with 1.85 g. (13.4 mmoles) of potassium carbonate
in 75 ml. of water. The reaction mixture was then cooled and a reddish-
ycllow oil separated. The oil was dissolved in ether. After drying
with anhydrous sodium sulfate the ether was removed and ethanol added
to crystallize the oil. A small amount of yellow-red solid separated,
m.p. ) 1800 (lit.  89-90o).
Potassium hydroxide in absolute ethanol, sodium hydroxide
melt, and potassium hydroxide melt were tried to obtain the 89-90o
product. In all cases a reddish colored oil was recovered. Adding
ethanol resulted in two solids, a yellow solid soluble in hot ethanol
and a red solid insoluble in hot ethanol. The yellow solid had a
melting range from 60 to 1000, with most of the solid melting about
60. The infrared spectrum of this compound was identical to the
spectrum of the product (m.p. 57-58o) obtained from the reaction of
trans-l-2-methoxyphenyl-l-phenyl-2-bromoethylene (trans-I) with po-
tassium-t-butoxide (described later) except for the absorptions at
965, 895, and 859 c-1. Absorptions 965 and 859 cm.-1 correspond to
two peaks observed in the infrared spectrum of l-k-methoxyphenyl-2-
phenylethylene. Vapor-phase chromatography, using a Perkin-Elmer
Model 154B, showed two peaks for the yellow solid. These peaks had
the identical retention times as the peak observed from the product
of the potassium-t-butoxide with trans-I reaction and the peak of
1-p-methoxyphenyl-2-phenyethylene. Analysis of the product obtained
from the potassium-t-butoxide reaction agreed with the formula of
1-k-methoxyphenyl-2-phenylacetylene. From this it is concluded that
the yellow solid is mainly a mixture of 1-L-methoxyphenyl-2-phenylacety-
lene and 1-k-methoxyphenyl-l-phonylethylene. Attempts to separate
these components were unsuccessful.
The red solid melted above 1800. No further identification
mgo- and dJ-Stilbene Dibromide
These compounds were prepared from trans-and cis-stilbene and
pyridinium bromiOd perbromide as described by Fieser [18 .
Three and six-tenths grams (0.105 mole) of meso-stilbene di-
bromide was added to 0.59 g. (0.0105 mole) of potassium hydroxide
dissolved in 10 ml. of absolute ethanol. After refluxing for one
hour the mixture was cooled, excess water added, and then extracted
with two 25 ml. portions of ether. The ether solution was dried over
anhydrous sodium sulfate and evaporated under dry nitrogen. A red
oil resulted which solidified at 160 (lit.  190) in an ice-bath.
trans-Monobromost i lbene
Eight-tenths of a gram (2.3 moles) of dl-stilbene dibromide
was added to 0.13 g. (2.3 mmoles) of potassium hydroxide in 5 ml.
of absolute ethanol. The reaction was treated in the same way as in
the preparation of cis-monobromostilbene. A light yellow oil was
recovered and solidified by dissolving in ethanol and cooling in a
dry ice-acetone bath. White crystals, m.p. 29-30o (lit. [i9] 320),
t-Butyl alcohol was purified by refluxing with a small amount
of sodium for one hour and then distilling through a 3-ft. column
packed with stainless steel sponge. The fraction, with a boiling
range of 82-82.50, was collected for use.
Potassium-t-butoxide in t-Butyl Alcohol
Potassium metal, cut under white mineral oil and dipped once
in dry pentane, was added to t-butyl alcohol and allowed to dissolve
under dry nitrogen. Concentrations of the solutions were measured
by titrating aliquots added to water against standard hydrochloric
acid solution using phenolphthalein as the indicator.
Reaction of 1,1-Diphenyl-2-bromoethylene (III) with Potassium-
One gram (3.9 =moles) of 1,1-diphenyl-2-bromoethylene (III)
was dissolved in 50 ml. of dry t-butyl alcohol and 240 mg. (6.1
mmoles) of potassium was added. The reaction was refluxed for
seventy-two hours and then quenched by adding nitric acid until the
mixture was neutral. Water was added until an oil separated. The
oil was dissolved in ethanol to yield 440 mg. (2.5 mmoles, 64 percent)
of diphenylacetylene, m.p. 57-58o (lit.  60-610). An authentic
sample of diphenylacetylene mixed with this compound gave no depres-
sion in melting point. The infrared spectra of both samples were
Reaction of 1,1-Diphenyl-2-branoethylene (III) with Potassium-t-
This reaction was carried out in exactly the same manner as
reaction I except that the reaction time was twenty-four hours. Ap-
proximately 650 ag. of oil were recovered. Vapor-phase chromatography
using a 6-ft. Tide column at a temperature of 1900 and a pressure of
10 lbs./sq.in. showed two peaks, retention times 2.7 and 4.7 minutes.
Under identical conditions a mixture of III and diphenylacetylene
also exhibited two peaks, retention times 2.7 and 4.8 minutes. Pre-
pared trans-monobromostilbene indicated one peak with a retention
time of 7.9 minutes. Cis-monobromostilbene also showed one peak with
a retention time of 3.8 minutes. A mixture of cis-monobromostilbene
and III gave only one broad peak. A broad peak was also observed with
the reaction mixture and the cis-isomer at approximately the same re-
tention time. Attempts to resolve these broad peaks were unsuccessful.
The infrared spectrum of the reaction mixture was identical to an
infrared obtained from a mixture of III and diphenylacetylene. Cis-
monobromostilbene had several peaks (2900, 1360, 1190, 1095, 890, and
792 cm-1) which did not appear in the infrared spectrum of the reaction
Reaction of 1-p-Methoxyphenyl-l-phenyl-2-bromoethylene (I) and Potas-
One gram (3.5 mmoles) of trans-I was dissolved in 25 ml. of a
0.40 M (10 mmoles) solution of potassium-t-butoxide in t-butyl alcohol.
The reaction mixture was refluxed for six days. After neutralizing
with hydrochloric acid and extracting with ether, 500 mg. (2.4 mmoles,
69 percent) of yellow-white crystals (m.p. 53-580) resulted on evapora-
tion of the ether extract. These were recrystalized from ethanol to
give a white crystalline product, m.p. 57-58o (lit.  89-90o).
Similar results were obtained with the cis-I isomer.
Analysis: Calculated for C154120: C, 86.54; H, 5.75
Found: C, 86.56; H, 5.82.
Reaction of 1-p-Chlorophenyl-l-phenyl-2-bromoethylene (II) with
Fifty-four hundredths of a gram (1.84 mmoles) of cis-l-2-
chlorophenyl-l-phenyl-2-bromoethylene (cis-II) was refluxed for
sixty-nine hours with a large excess of potassium-t-butoxide in
20 ml. of t-butyl alcohol. Following this, water was added to the
reaction mixture until it became cloudy. After placing the mixture
in a refrigerator for two days, 240 mg. (1.1 moles, 61.5 percent)
of 1-p-chlorophenyl-2-phenylacetylene was recovered. Recrystalliza-
tion from ethanol gave a pure product, m.p. 81-830 (lit. 10 82-830),
as evidenced by one peak recorded by vapor-phase chromatography.
Reaction of 1-y-Methoxyphenyl-l-phenyl-2-bramoethylene (I) with
Two hundred and thirty milligrams (0.79 mmole) of trans-I was
refluxed for forth-eight hours with 100 mg. (4.35 mmoles) of sodium
in 15 ml. of absolute ethanol. Water was added to quench the reac-
tion and crystals separated which were identical with the starting
material as evidenced by the melting point, 79-810, and a mixed
melting point. Cis-I gave similar results.
Reaction of trans-l-p-Methoxyphenyl-l-phenyl-2-bromoethylene (trans-I)
Two hundred milligrams (0.69 nanole) of trans-I was treated with
150 mg. (3.94 mmoles) of potassium in 20 ml. of t-amyl alcohol for
sixty-one hours at reflux temperature. Water was added to the reaction
mixture and white crystals separated, m.p. 45-90o. These were not
Reaction of trans-l-p-Methoxyphenyl-l-phenyl-2-bromoethylene (trans-I)
with Potassium-t-butoxide in Diglyme
Five hundred milligrams (1.73 =moles) of trans-I was treated
with 300 mg. (2.68 mmoles) of potassium-t-butoxide in diglyme solvent
for ninety-six hours at reflux temperature. The addition of water
resulted in an oil which was separated from the other liquid and dis-
solved in ethanol. White crystals separated, m.p. 45-900, and were
not characterized further.
Reaction of trans-l-p-Methoxyphenyl-l-phenyl-2-bromoethylene (trans-I)
with Potassium Amide in Liquid Ammonia
One and one-tenth grams (3.8 mmoles) of trans-I was dissvoled
in 40 ml. of dry ether and added dropwise to 1.0 g. (18.2 mmoles) of
potassium amide in 150 ml. of liquid ammonia. After the addition, the
solution was allowed to evaporate overnight. Water and ether were
added to the residue and a reddish-white solid was obtained from the
ether layer. Recrystallization from ethanol yielded a white solid,
m.p. 48-500. The infrared spectrum showed the presence of C C --
at 2225 cm.-1. Vapor-phase chromatography showed one peak at the same
retention time as an authentic sample of 1-2-methoxyphenyl-2-phenylacety-
lene run under identical conditions.
Reaction of cis-l-p-Methoxyphenyl-l-phenyl-2-bromoethylene with
Phenyl-lithium in Ether
Two hundred milligrams (0.69 mole) of cis-I was treated with
10 ml. of 0.23 N (2.3 moles) of phenyl-lithium in dry ether for two
hours at reflux temperature. After the removal of ether, water and
hydrochloric acid were added to the residual oil. Yellow-white
crystals separated, m.p. 500. Vapor-phase chromatography indicated
this was starting material.
Attempted Rearrangement of trans-l-p-Methoxyphenyl-l-phenyl-2-bromo-
ethylene (trans-I) to cis-l-p-Methoxyphenyl-l-phenyl-2-bromoethylene
with Ultraviolet Light
Three hundred milligrams (1.04 mmoles) of trans-I was dissolved
in 20 ml. of ethanol and placed under an ultraviolet lamp for eight
hours. Water was added and yellow-white crystals separated, m.p.
76-78o. The crystals were redissolved in ethanol and placed under
the lamp for an additional fifteen hours. The solution was treated
as before. A solid, m.p. 75-780, and oil were recovered. Some impure
crystals were obtained from the oil, m.p, 70-80o. To the mother li-
quor, after the separation of solid and oil, ether was added, separated
from the water layer, dried with anhydrous sodium sulfate and evaporated
almost to dryness. Ethanol was added and some crystals were recovered,
1. Ultraviolet Absorption Spectra
Ultraviolet absorption spectra were employed to determine the
concentrations of starting materials and products used in the kinetic
measurements. Ninety-five percent ethanol was used as our solvent.
A scan of the spectra from 340 mp to 220 mp by the use of the Beckman
DK-2 Spectrophotometer indicated the maximum absorptions for our
materials. The extinction coefficients for each of the maxima needed
for our concentration calculations were determined utilizing the Beck-
man Model DU. Table 2 summarizes the maxima observed for each of our
compounds used in kinetic measurements. In cases where there are
several maxima the largest two are given.
Data for the calculation of the extinction coefficients of
tolans formed from the reactions as well as some of the starting
materials are summarized in Tables 3, 4, 5, 6, and 7.
2. Visible Absorption Spectra
It was desirable to follow the formation of bromide ion in
the kinetic measurements. Since the concentration of starting mater-
ial was approximately 10-3 M an accurate method of determining bro-
mide ion was desirable. Accordingly, the method used by Goldman and
Byles was tried . The principle of the method was the oxidation
of bromide ion with chloramine T reagent and the bromination of phen-
ol red in a buffer solution. A color comparison of brominated com-
pound versus distilled water was made with a Beckman DU Spectrophoto-
meter at 590 mp. The brominated compound should appear reddish to
ABSORPTION MAXIMA OF COMPOUNDS USED IN KINETIC MEASUREMENTS
due to impurities
EXTINCTION COEFFICIENT DETERMINATION FOR
moles/liter x 105
1.73 0.351 2.02
3.46 0.698 2.02
4.34 0.865 1.98
i.00 0.20b 2.06
0.10 0.021 2.01
2.00 0.404 2.02
E = Extinction coefficientAvg. = 2.02 + 0.007 x 104 (306 mp)
EXTINCTION COEFFICIENT DETERMINATION FOR
moles/liter x 104
Extinction CoefficientAvg. = 1.98 +
0.015 x 103 (306 mp)
EXTINCTION COEFFICIENT DETERMINATION FOR
Concentration in Optical Density Extinction Coefficient
moles/liter x 104 (306 mp) x 10-3
1.22 0.139 1.14
2.44 0.267 1.09
3.63 0.422 1.16
2.61 0.296 1.14
1.31 0.154 1.16
E = Extinction CoefficientAvg. = 1.14 + 0.01 x 103 (306 mp)
EXTINCTION COEFFICIENT DETERMINATION FOR
Concentration in Optical Density Extinction Coefficient
moles/liter x 105 (302 my) x 10-4
0.764 0.190 2.48
1.53 0.409 2.68
0.580 0.153 2.64
0.725 0.182 2.52
E = Extinction CoefficientAvg. = 2.58 + 0.03 x 104 (302 mp)
EXTINCTION COEFFICIENT DETERMINATION
moles/liter x 105
E = Extinction CoefficientA = 2.48 + 0.015 x 104 (296.5 mp)
violet, depending upon its concentration. By constructing a calibra-
tion curve of bromide ion concentration from 0.1 to 1.0 mg./1. versus
optical density at 590 mp a sharp differentiation within + 1 percent
can be made between varying quantities of bromide.
C. Kinetic Methods
1. Ultraviolet Absorption Method
Essentially the kinetic method described by Pritchard and
Bothner-By  was used. The reactions of 1,1-diaryl-2-bromoethy-
lenes with potassium-t-butoxide in t-butyl alcohol were carried out
Standard base was prepared as previously described. Glassware
used was washed twice with acetone, once with tap water, then with
distilled water, and then dried in an oven at 1350.
Weighed amounts of 1,1-diaryl-2-bromoethylenes were accurately
diluted to a standard volume with t-butyl alcohol. Twenty-five milli-
liter samples were added to a 50 ml. volumetric flask and standard
base was added until the total volume was 50 ml. The contents of the
flask were thoroughly mixed and added to a 50 ml. buret. Five milli-
liter samples were then added to 10 ml. ampoules, previously flushed
with dry nitrogen, using a needle on the tip of the buret in order to
obtain a fine stream of liquid. The ampoules were capped with medi-
cine dropper tops, after flushing again with dry nitrogen, and placed
in an ice-bath. The tubes were then sealed, preheated in a water bath
for 20 seconds at ca. 950 and immediately transferred to a constant
temperature mineral oil bath, which was maintained at a temperature
of 950 + 0.020. The purpose of preheating the frozen tubes was to
eliminate the large drop in temperature in the mineral oil bath which
would result from adding the tubes directly. Zero time was taken
when the tubes were placed in the bath. Ampoules were withdrawn at
intervals and quenched by placing in an ice-bath. While frozen the
tubes were opened and distilled water added to convert any potassium-
t-butoxide remaining to potassium hydroxide. The tubes were then
allowed to warm to room temperature and the solution transferred to
a volumetric flask. The tubes were rinsed once with ethanol and once
with distilled water. After accurately diluting with ethanol to a
standard volume at room temperature the ultraviolet spectra were run
versus ethanol as a blank.
Optical densities were measured at wave lengths corresponding
to maxima of tolans. These were selected because starting olefins
either absorbed very little or did not absorb at all at these wave
Concentrations were calculated from Beer's Law, assuming the
formation of product is equal to the disappearance of starting material.
The general equation A = abc, was used, where
A = absorbance or optical density of the solution,
a = molar extinction coefficient of the solute,
b = thickness of the cell containing the solution, and
c = concentration of the solution in moles/liter.
For the reactions of trans- and cis-l-p-methoxyphenyl-l-phenyl-2-
bremoethylene (I), which absorbed slightly at the maximum of
1-p-methoxyphenyl-2-phenylacetylene, the following modified equation
ATotal = aTbcT + aEHbcEH,
where Total refers to l-p-methoxyphenyl-2-phenylacetylene (tolan)
and cis- or trans-I. T represents p-methoxytolan and EH refers to
cis- or trans-I.
Knowing the concentration of starting material the relation
cT + cEH = constant was used to substitute in the equation and de-
termine the concentration of product or starting material during the
A plot of -log cEg versus time gave a straight line, indica-
tive of a first-order reaction. Values of k were calculated by the
method of least squares [21 .
2. Vapor-phase Chromatography
Quantitative analyses of mixtures of 1,1-diphenyl-2-broBoethy-
lene and diphenylacetylene were made in the anticipation of following
the rate of the reactions by the use of vapor-phase chromatography.
Samples were accurately weighed, dissolved in chloroform and injected
into a vapor-phase fractometer, Perkin-Elmer Model 154. A 6-ft. Tide
column at temperature ca. 1800 and a pressure of 25 lbs./sq.in. (helium)
was used to separate the mixtures. The areas under, the curves for
each component were measured by constructing triangles over the curves
and calculating the areas of the triangles. Table 8 sunmarizes the
results obtained. This method of analyses was not used in the kinetic
QUANTITATIVE ANALYSIS OF 1,1-DIPHENYL-2-BROMOETHYLENE
AND DIPHENYLACETYLENE MIXTURES BY VAPOR-PHASE CHROMATOGRAPHY
Mole percent of Mole percent of
1, 1-Diphenyl-2-bromoethy lene Diphenylacetylene
Calculated 77.4 60.6 41.9 22.6 39.4 58.1
Measured 78.5 59.2 44.1 21.5 40.8 55.9
D. Dipole Moments
Dipole moments were obtained by measuring the dielectric con-
stant of dilute solutions in benzene, a nonpolar solvent. The dielec-
tric constant D was determined by measuring the ratio of the capacity
of a condenser with the substance between the plates of a cell compared
to the capacity of the condenser with a vacuum between the plates of
Capacities of the cell were measured by balancing a General
Radio Company type 716-C capacitance bridge containing the measuring
cell (unknown capacitor) and a precision capacitor (balancing capa-
citor). These measurements were made at frequencies of 90 and 70 kilo-
cycles obtained by a General Radio Company oscillator type 1302-A.
Null or balance points were found by observing the smallest sine wave
on a Tektronix type 541-A oscilloscope. A J.C. Balsbaugh 2TN50 cell
was used for measurements. The volume of liquid used in this cell
was ca.15 ml. The cell was immersed in a water bath held at 250 +
0.020. Since the temperature of the room was ca. 240 little heating
was necessary to maintain the temperature. It was found that shield-
ing the thermoregulator, stirring motor and heater from the cell with
a 10" x 10" x 1/8" piece of copper metal removed any interference ob-
served on the oscilliscope.
Readings of capacitance were made after the solutions had
10 minutes to reach equilibrium. Five readings were taken at each
Refractive indices of the solutions were measured with an
Abbe refractometer held at 250 + 0.020.
Densities of the solutions were determined with a 5 ml. pycno-
meter calibrated with distilled water. The pycnometer was filled at
ca. 200 and immersed in a constant temperature bath, 250 + 0.020, for
ten minutes. After removing, it was carefully dried and weighed.
Densities were not corrected for the buoyancy of air.
Solutions were made up by weighing both solute and benzene
for the first sample and then accurately diluting this sample with
volumetric pipets in a room at ca. 200 to obtain the other solutions.
Five solutions were prepared for each compound. These ranged from
0.015 to 0.002 mole fraction of solute (x2).
Pure benzene previously prepared for dielectric studies was
used, n250 = 1.4982, d250 = 0.8736.
Spectral grade carbon tetrachloride was used.
The purest available samples of cis- and trans-l-p-methoxyphenyl-
1-phenyl-2-bromoethylene were used.
a. Molar Refraction
Molar refractions of the solutions were calculated according
to the equation
n2 1 XIMI + x2M2
R12 =n- + 2 d12 '
where R12 is the molar refraction of the mixture, n is the refractive
index, x1 and x2 are the mole fractions of solvent and solute respec-
tively, and M1 and M2 are the molecular weights of solvent and solute.
d12 is the density of the solution.
The values of R12 were determined from the equation
R12 = x1R1 + x2R2,
where R1 and R2 are the molar refractions of solvent and solute, re-
b. Dielectric Constant
The total capacitance of a cell is made up of the capacitance
of the cell leads CL and the capacitance of the cell when filled
sith a substance DCo. When the capacitance of the cell is measured
with air (D = 1.000), and with benzene of a known dielectric constant
(D = 2.273 at 25)  the cell constants may be determined and used
in the measurement of the dielectric constant of an unknown material.
The following standardization was made:
Capacity with air in cell = 143.35 mmf. = Co + CL
Capacity with benzene in cell 177.6 mmf. = 2.273Co + CL
(mmf. = micromicrofarads).
From these two equations,
1.273Co = 34.25
Co = 26.90 Baf.
CL = 116.45 mmf.
To check the standardization pure carbon tetrachloride was
placed in the cell and its dielectric constant calculated.
Capacity with CC14 in cell = 176.4 mmf.
176.4 = [(D)(26.90)] + 116.45
DCCI4 = 2.228 (lit.  = 2.227)
c. Dipole Moment
The dipole moment p was calculated using
p = 0.01281 x 10-18 (P2oo R2)T 1/2
where R2 is the molar refraction of the solute,
temperature and P2oo is the polarization of the
dilution. P2oo was determined by the method of
using the equation
T is the absolute
solute at infinite
(DI 1)(M2 N4i) + 3M1D1
P2oo = (DI + 2)dI dl(DI + 2)2
symbols have the usual significance. 8 and ao are the
the lines obtained from the following equations:
D12 = D1(1 ax2)
dl2 = dl(1 + x2)
D12 and dl2 are the dielectric constants and densities of the solu-
tions. These slopes were calculated by the method of least squares.
The dipole moments of the cis and trans isomers of 1-p-
methoxyphenyl-l-phenyl-2-bromoethylene were measured. The trans
configuration was assigned to the higher melting isomer and the cis
configuration to the lower melting isomer on the basis of the ob-
served moments as compared to the calculated moments. These results
are summarized in Table 9. Figures 1 and 2 were used to determine
ca and 0 used in the calculation of the dipole moments. 0c and
are the slopes of the lines in Figures 1 and 2, as determined by
the method of least squares .
Five compounds were prepared for kinetic measurements. These
were the cis and trans isomers of 1-p-methoxyphenyl-l-phenyl-2-bromo-
ethylene (cis- and trans-I), the cis and trans isomers of 1-p-chloro-
phenyl-l-phenyl-2-bromoethylene (cis- and trans-II) and 1,1-diphenyl-
2-bromoethylene (III). All were obtained pure, except trans-II which
was contaminated with a maximum of 30 percent of cis-11 as calculated
quantitatively by infrared spectra.
Cis- and trans-monobromostilbene, products of possible in-
termediates in the reaction of 1,1-diphenyl-2-bromoethylene, were
m 9w (7 0
a) to 0n
co CO t- t-
N NH C
0 (0 r-
e r-4 0 0
yr-4 in tl-c
co 0o 0)
* 0 .
r-4 NO CV) qwP
\ 8 a
\- CD )
\ I \ C
\u;So \ J;o s
x = cis
. = trans
2 .4 .6 .8 1.0 1.2 /4 /.6 /. 8
Mole Fraction of Solute x 102
Figure 2. Determination of 0 for cis- and trans-I-p-methoxyphenyl-l-phenyl-2-bromoothylene.
Products from the reactions of 1,1-diphonyl olefin (III) and
cis and trans-p-methoxy compounds (I) with potassium-t-butoxide were
synthesized by an alternate route. They were diphenylacetylene and
1-p-methoxyphenyl-2-phenylacetylene (obtained impure).
All results indicate that the reactions of the bromoethylenes
with potassium-t-butoxide give mainly diphenyl and substituted diphenyl-
acetylenes (tolans). Thus, spectroscopic analysis indicated at least
85 percent of the reaction yielded tolans for all of the starting bromo-
ethylenes (e.g. see Table 10). Furthermore, the reaction of 1,1-diphenyl-
olefin (III) with excess base gave a 64 percent yield of diphenylacety-
lene, after recrystallization. Trans- and cis-k-methoxy compounds (I),
when reacted with'potassium-t-butoxide, yielded 69 and 62 percent re-
spectively, of essentially pure tolans.
A careful analysis of the crude reaction mixture, resulting
from the incomplete reaction of 1,1-diphenyl olefin (III) with potas-
sium-t-butoxide, indicated it consisted only of starting material and
diphenylacetylene. This mixture was analyzed by:
1. Comparing the vapor-phase chromatograms of the reaction
mixture and a mixture of 1,1-diphenyl olefin (III) and diphenylacety-
2. Preparing cis- and trans-monobromostilbene, two possible
products from intermediates in the reaction of 1,1-diphenyl olefin
(III) with base, and comparing their vapor-phase chromatograms and
infrared spectra to those of the reaction mixture.
3. Comparing the infrared spectrum of the reaction mixture
to the spectra of the starting material and diphenylacetylene.
Since spectroscopic evidence indicates that all of the olefins
gave over 85 percent of the tolan, the reaction may reasonably be
djt -dt kobs.CEH*
where CT and CEH are molar concentrations of tolan and bromoethylene
derivatives, respectively, and kobs. is the observed rate constant for
the reaction. A plot of -log CEH versus t (time) gave a straight line
representative of a first-order reaction in all cases. Values of k
were calculated from the slope of the lines by the method of least
squares . Table 10 indicates the typical data used in the kinetic
The kinetic results of the five compounds studied are summarized
in Table 11. The same base concentration (ca. 0.20 M) was used in all
cases to eliminate any differences in the rate constants resulting from
the base term inherent in these constants and to minimize differences
in salt effects between compounds. Three runs were made for each com-
pound except 1,1-diphenyl-2-bromoethylene, whose rate constant was
previously reported under very similar conditions . A concentra-
tion of ca. 1.0 x 10-3 M of starting material was used. In cases
where this varied the pseudo-first-order rate constant remained essen-
tially the same.
KINETIC DATA FOR THE REACTION OF trans-l-p-METHOXYPHENYL-1-
PHENYL-2-BROMOETHYLENE (trans-I) WITH 0.20 M POTASSIUM-t-
BUTOXIDE AT 950
Concentration of tr -I
in moles/liter x 103
a 1 ml. of sample to 9 ml. of solvent. All
of sample to 49 ml. of solvent.
others were diluted 1 ml.
b Measured initial concentrations were within 5 percent of those cal-
culated by weighing of samples and diluting to standard volume.
SUMMARY OF KINETIC DATA FOR THE REACTIONS OF 1,1-
DIARYL-2-BROMOETHYLENES (EH) WITH POTASSIUM-t-
BUTOXIDE (KB) AT 950
of KB in
of EH in
a tra-1-P-chlorophenyl-l-phexyl-2-bromoethylene was contaminated
with a maximum of 30 percent of cis-l-p-chlorophenyl-l-phenyl-2-
bA kobs. of 0.85 x 10-3 (min.-1) at 0.212 M KB agrees w11 with a
reported kobs. of 0.73 x 10-3 (min.-1) at 0.215 M KB l91.
A characteristic plot of -log CEH versus time (minutes) is shown
in Figure 3.
It is desirable to compare the rate constants of the bromoethy-
lenes with each other. For that purpose Figure 4 was drawn. Since the
rate constants are independent of the concentration of bromoethylenes,
the average rate constants of each compound are used to determine the
average slopes of the lines in Figure 4. The base concentration of
0.20 M was used. The calculated rate constants of 1.62, 1.06, 0.45,
1.04, 0.82 x 10-3 for trans- and cis-p-methoxy compounds (I), trans-
and cis-E-chloro compounds (II) and 1,1-diphenyl olefin (III), respec-
tively, were used. A distinct difference of slopes, except between cis-
k-methoxy-I and cis-j-chloro-II, can be observed in each case.
Reactions with Other Bases
A few bases, other than potassium-t-butoxide in t-butyl alcohol
were tried to effect the rearrangement. Treatment of cis- or trans-
1-p-methoxy compounds (I) with sodium ethoxide in ethanol and phenyl-
lithium in ether gave no 4-methoxytolan. Potassium amide, in liquid
ammonia, was successful while potassium-t-amylate in t-amyl alcohol
and potassium-t-butoxide in diglyme (diethylene glycol dimethyl ether)
gave dubious results.
Attempted rearrangement of trans---methoxy-I to cis-p-methoxy-I
under ultraviolet radiation was unsuccessful.
x a cis
* = trans
80 JZO O
240 280 320
Reaction of trans- and cis-1-p-methoxyphenyl-l-phenyl-2-branoethylene (I) with 0.197 M
Potassium t-butoxide at 950.
120 18o 240 300
Comparison of rate constants of 1,1-diaryl-2-bramoethylenes at 0.20M potassium-t-butoxide.
A quantitative method of calculating the percent of 1,1-diphenyl-
2-bracoethylene and diphenylacetylene by vapor-phase chromatography
was developed and found to be accurate within + 1 mole percent.
DISCUSSION OF RESULTS
In order to interpret the relative reaction rate constants
and properly discuss reaction mechanisms the proper geometric con-
figuration of the isomers of 1-p-methoxyphenyl-l-phenyl-2-bromoethy-
lene (UI2- and trans-I) and of the isomers of 1-r-chlorophenyl-l-phenyl-
2-bromoethylene (gcs- and tras-II) must be known. Usually the best
way to assign these configurations is by dipole moment measurements.
This has been done for cis- and trans-II and the higher melting isomer
was found to have the cis configuration (bromine and 2-chlorophenyl
Stoermer and Simon , who first prepared the cis- and trans-
p-methoxy compounds (I), assigned the trans configuration to the higher
melting isomer. Their assignment was based on two assumptions:
1. The one isomer of 1-phenyl-l-o-hydroxyphenyl-2-bromoethy-
lene, which reacts with base to yield 2-phenylbenzofuran, necessarily
has the cis configuration.
O /H /0
2. Since this isomer was lower melting, the trans configura-
tion could be assigned to all the higher melting isomers of 1,1-diaryl-
2-bromoethylenes with a substituent in one of the benzene rings. These
assumptions are not valid in all cases, as evidenced by the cis-p-
chloro compound (II) being the higher melting isomer of the isomers
of cis- and trans-p-chloro compounds. Thus, a determination of the
dipole moments of the cis and trans-p-methoxy compounds (I) was under-
Theoretical dipole moments were calculated upon the following
1. The carbon-bromine bond moment is 1.38 debye units, the
value assigned by Smyth .
2. The bond moment of the vinyl hydrogen is taken as zero
because Gent  has stated the carbon-hydrogen bond moment is ca.
zero in ethylene.
3. The moment of the p-methoxyphenyl group is 1.25 debye
units, the same as that reported for anisole . The moment
angle of the mothoxy group is 780 from the line of attachment to
the benzene nucleus and at a direction into the benzene ring [27 .
A further assumption made here is the free rotation of the methoxy-
methyl group around the carbon-oxygen bond resulting in the rotation
of the methoxy-moment around the line of attachment to the benzene
4. All other moments in the molecule are zero.
5. The bond angles are not distorted from their normal values.
Using these assumptions the dipole moments of the cis- and
trans-p-methoxy compounds (I) were calculated by the law of cosines.
In the case of trans-I the moment of the methoxy group is always
780 from the carbon-bromine moment. In the cis isomer the angle be-
tween the methoxy moment and the carbon-bromine moment varies from a
minimum of 420 to a maximum of 1620 when the methoxy moment rotates.
t'he calculated value of cis-1 is taken as the average of the maximum
and minimum values. The calculated dipole moment of the trans-p-
methoxy compound (I) is 2.05 debye units, while the calculated value
of the cis-p-methoxy compound (I) is 1.45 debye units, an average of
a maximum value of 2.46 and a minimum value of 0.43 debye units.
The observed dipole moments of the trans- and cis-p-methoxy compounds
(I) are 2.65 (820 isomer) and 2.18 (530 isomer) debye units, respec-
tively. The observed values are approximately 0.6-0.7 debye units
higher than the calculated values, but the difference between the
moments observed of the cis- and trans-p-methoxy compounds (I) is
essentially that calculated. On this basis the isomer with the high-
est dipole moment is assigned the trans configuration. This is the
same as that originally given by Stoermer and Simon .
Migratory Aptitudes and Reaction Mechanisms
The tendency for various groups to rearrange or migrate in
the elimination rearrangement reaction of 1,1-diaryl-2-bromoethylenes
is called the migratory aptitude of the groups. One reasonable approach
to measuring this tendency is to examine the rates of the elimination-
rearrangement reaction of various substituted phenyl groups on the vinyl
bromide. However, inherent in such a study is the problem of the effect
of the substituent group on the acidity of the vinyl hydrogen. In view
of the known stereospecificity of this reaction, it was thought that
this problem could be obviated by examining the relative rates of re-
arrangements of the cis and trans isomers in which only one of the phenyl
rings contained the substituent under consideration. Thus, the cis and
trans isomers of 1-L-methoxyphenyl-l-phenyl-2-bromoethylene (cis- and
trans-I), 1-p-chlorophenyl-l-phenyl-2-bromoethylene (cis- and trans-II)
and 1,1-diphenyl-2-bromoethylene (III) were prepared. In these five
different compounds we have an electron-donating group (-OCH3) on the
migrating phenyl, an electron withdrawing group (-C1) on the migrating
phenyl and what we could call our neutral or reference group (-H). In
a pair of isomers, if the trans group is migrating with its electrons
the electron-donating group would be expected to increase the rate of
the reaction. If, on the other hand, the phenyl is migrating without
its electrons, the electron-donating group would be expected to retard
the reaction whereas the electron withdrawing group would be expected
to increase the rate. This study of migratory aptitudes can be com-
pared to the known migratory aptitudes in the pinacol rearrangements.
R' -C--C --R
In the pinacol rearrangement it was found that the migratory apti-
tudes for a phenyl migrating are as follows :
p OCH3 > p alkyl >p H >p Cl
The general reaction mechanism (8,9 for the reaction of
1,1-diaryl-2-bromoethylenes with potassium-t-butoxide may be written
C = C
+ B- C = C"
AR--C C--AR + Br-
AR = aryl group
Because of the wide variety of strong bases 1-10] which can
initiate the rearrangement a reasonable first step in the reaction is
the abstraction of the vinyl proton to form anion V. By the use of
deuterated solvent Pritchard and Bothner-By (9] have exchanged the
vinyl hydrogen of IV with a deuterium atom. They have calculated
the exchange reaction to be at least a hundred times faster than the
conversion to diphenylacetylene at the same temperature. Moreover,
the rates for the rearrangement reactions are considerably faster in
deuterated solvent than they are in t-butyl alcohol. From this in-
formation it is plausible that the reaction mechanism involves rever-
sible carbanion (V) formation. If the formation of carbanion V were
irreversible, so that its formation would be the rate-determining
step, the reaction of deuterated substrate would proceed more slowly
than that of undeuterated substrate (IV). Further evidence that the
abstraction of a proton is not the rate-determining step, using po-
tassium-t-butoxide as the base, is the difference in rate constants
of the cis and trans isomers observed by the author. In my opinion,
the cis- and trans-p-methoxy compounds (I), for example, should be
identical as far as the abstraction of the vinyl proton is concerned,
yet they have rate constants of 1.06 and 1.62 x 10-3 (min.-1), re-
spectively. The most reasonable explanation of these rate constants
is that the migration is involved in the rate-determining step, rather
than occurring after it. On the other hand, Curtin and Flynn (7) have
demonstrated that in the formation of diphenylacetylene, 1,1-diphenyl-
2-bromoethylene-2-d reacts at an appreciably slower rate with butyl-
lithium than does 1,1-diphenyl-2-brmaoethylene. This would indicate
that the breaking of the carbon-hydrogen bond is the rate-determining
step when butyl-lithium is used as a base. However, under these con-
ditions, the first step of the reaction is most probably not reversible.
The second step, or rate-determining step of the reaction, us-
ing potassium-t-butoxide as the base, is the migration of the group
trans to the bromine atom. Implicit in this system is the configura-
tional stability of anion V. The fact that the trans group migrates
and anion V was stable was demonstrated by Bothner-By [81 and by Cur-
tin, al. [101. Bothner-By prepared the cis and trans isomers of
l-p-bromophenyl-l-phenyl-2-bromioothylene-1-14C and converted them to
labeled 4-bromotolans using potassium-t-butoxide in refluxing t-
butyl alcohol (ca. 330). By oxidizing the tolans, radioactive benzoic
and p-bromobenzoic acids were obtained. From this he found the cis
isomer to rearrange with 88 percent migration of the phenyl group
while the trans isomer rearranged with 92 percent migration of the
p-bromophenyl group. Curtin, et al., found similar results with cis
and trans p-chloro compounds using butyl-lithium to initiate the re-
action at -350.
It might be well to point out now the similarity between this
rearrangement and the Beckmann rearrangement . The Beckmann re-
arrangement involves migration between doubly-bonded carbon and
nitrogen with the migration of the group trans to the leaving -OH or
R OX 0 X
\ / II I
C = N acid or heat R-- C --N --R'
X = H or R group,
R', R = alkyl or aryl
Moreover, the nitrogen atom is isoelectronic with carbanion V. There
are, of course, differences in the reactions, the most obvious being
the initial reaction with strong base in our rearrangement while the
Beckmann is initiated with acids or heat.
There remains unanswered the important questions of migratory
aptitudes and of the detailed reaction mechanism of the second step
of our reaction. The three most reasonable mechanisms for the migra-
tion step are:
1. The slow migration of the trans aryl group without its
electrons followed by a fast loss of bromide ion:
C =-C slow fCa=C fast AR- C =-C -AR' + Br-
/ \ / \
AR Br AR Br
2. The same as mechanism 1 except that the first step is fast
and the second slow.
It was important to consider these two mechanisms because the
rearrangement of phenyl groups without their electrons to carbanions
has been postulated by Zimmerman and co-workers [30 .
3. The slow rearrangement of the trans group with its electron
accompanied by the displacement of bromide ion.
C= C- slow AR C C-AR' + Br-
The values of the average rate constants, a measure of migra-
tory aptitudes in our reactions, are summarized in Table 12.
It can be seen, from a comparison of the rate constants of the
cis- and trans- p-methoxy compounds (I), that the migrating group
donating electrons into our system speeds up the reaction over the
migrating neutral or reference group. Conversely, the cis- and
trans-2-chloro compounds (II) indicate that the migrating group with-
drawing electrons slows down the reaction compared to the migrating
reference group. Normally, this type of information could be taken
as evidence that a group is migrating with its electrons [28,29 .
Thus, mechanism 1 would appear unlikely because the slow migration
without electrons would predict the opposite orders. For example,
Zimmerman and Zweig  have shown that a p-tolyl group migrates
slower than a phenyl group in the rearrangement reactions of 1-chloro-
2-phenyl-2-p-tolylpropane with lithium. This is consistent with an
electron-donating group (-CH3) retarding the rate of rearrangement
compared to a phenyl group when the groups are rearranging without
their electrons. Any effect of the various substituted phenyl
groups on anion VI would not affect the rate of the reaction since
this anion is formed after the rate-determining step. However, in
mechanism 2 the stability or instability of anion VI would certainly
be expected to affect the rate of the reaction since this anion is
formed before the slow step of the reaction. For example, anion
VI-a of the cis-p-chloro compound (II) would be expected to be more
AVERAGE RATE CONSTANTS FOR THE REACTIONS OF 1-PHENYL-1-p-
SUBSTITUTED PHENYM-2-BROMOETHYLENES WITH 0.20M POTASSIUM-
Compounds k x 103(min.-1)
trans-l-p-methoxyphenyl- l-phenyl 1-2-
stable than anion V1-b of the trans-p-chloro compound (II) due to
the electron-withdrawing effect of the p-chloro group.
/ \ / \
y = -Cl or -OCH3
Conversely, anion VI-a of the cis-p-methoxy compound (I)
would be less stable than anion VI-b of the trans-p-methoxy com-
pound (I) due to the destabilization of the cis-I anion by the
p-methoxy group. Thus, from an examination of the relative rates
of reaction of the cis-trans pairs of isomers it is seen that the
results can be made equally consistent with either mechanism 2 (rapid
equilibration of the two anions followed by slow loss of bromide) or
mechanism 3 (migration of the phenyl with its electrons). Further-
more, a more careful comparison, not only of the relative rates of
the cis-trans pairs, but also of the relative rates of various com-
pounds, leads to the same conclusions; although, admittedly, some of
the assumptions required to rationalize the results with mechanism 2
are rather unlikely. However, in the discussion to follow, it is seen
that mechanism 2 becomes a very unlikely possibility.
As was pointed out earlier, Bothner-By [91 succeeded in trap-
ping anion V with deuterium.
+ t BuOD ',
C = C
+ t BuO-
It is certainly reasonable that anion VI, in which the carbanion
electrons have the possibility of being conjugated with a phenyl
or substituted phenyl group, would be at least as stable, and probab-
ly more stable, than anion V. Therefore, if present, anion VI or
VI' should react with solvent by an analogous reaction to that of
anion V to give VII, VII', or both
[- = = -
+ t BuOH C = C
C = C
C = -C
A reaction of 1,1-diphenyl-2-bromoethylene (III) with potassium-
t-butoxide was, therefore, allowed to proceed to partial completion
and the crude mixture, after neutralization, was carefully analyzed
for trans- and cis-monobromostilbene, the products of anions VI and
VI'. By the use of vapor-phase chromatography, infrared spectra
and a comparison with authentic samples it was shown that the re-
action mixture resulting from the partial reaction of the 1,1-diphenyl
olefin (III) with base consisted only of starting material and diphenyl-
acetylene. There was no detectable trace of the bromostilbenes. These
results tend to exclude mechanism 2 as being operative.
The third mechanism involves the slow rearrangement of the
trans group with its electrons. The rate constants for the cis and
trans isomers are consistent with this mechanism. The trans-p-methoxy
compound (I) would be expected to react faster than the cis-p-methoxy
compound (I) because the migrating group is p-methoxyphenyl in trans-I
and phenyl in cis-I. According to mechanism 3, the p-methoxy group,
being an electron-donating group, increases the rate of the rearrange-
ment compared to the rate of rearrangement of a phenyl group. Con
versely, the cis-p-chloro compound (II) reacts faster than the trans-
2-chloro compound (II) because a p-chlorophenyl group is migrating
in trans-It, retarding the rate of the reaction.
Phenyl is the migrating group in the cis-p-methoxy compound (I)
and in the cis-p-chloro compound (II). The differences in rate con-
stants between the cis isomers and 1,1-diphenyl olefin (III) is rea-
sonably due to the effects of the stationary substituted phenyl groups.
Probably the most important of these effects is the acidity of the
vinyl hydrogen abstracted in the first step of the reaction.
Roberts and co-workers  have measured the rates of deuterium-
proton exchange of o-, m-, and p-deuterated fluorobenzene, benzo-
trifluoride and anisole. In such reactions the combined inductive
and field effects of the substituents on the benzene ring appear
to be of major importance and the electromeric effects of minor im-
portance. They have shown that fluorine, although four carbon atoms
from the deuterium atom, increases the rate of the removal of deuterium
two hundred times over that of the unsubstituted deuterobenzene. The
methoxy group had little effect on the acidity of the deuterium atom.
In view of this paper we can be fairly certain that the chlorine atom
of the cis-p-chloro compound (II), having the same type of effect as
the fluorine atom, would increase, at least to some degree, the acid-
ity of the vinyl hydrogen over that of the acidity of the vinyl hy-
drogen of 1,1-diphenyl olefin (III). It is reasonable, therefore, to
say that the cis-p-chloro compound (1) reacts faster than the 1,1-di-
phenyl olefin (III) due to the greater acidity of the vinyl hydrogen.
The cis-p-methoxy compound (I) also reacts faster than 1,1-
diphenyl olefin (III). Since Roberts, et al. ,31], have shown that
the methoxy group has little effect on the acidity of the deuterium
atom it is difficult to rationalize cis-I reacting faster than III.
One possible explanation of this increase is the possible stabiliza-
tion of the incipient carbon-carbon triple bond by the p-methoxyphenyl
group. Probably occurring simultaneously with the migration of the
trans aryl group is the formation of the carbon-carbon triple bond.
Since the p-methoxyphenyl group is conjugated with the vinyl group
it probably has some effect on the formation of the triple bond.
Depuy and Leary  have shown that in the pyrolytic elimination
reactions of some acetates of substituted 1, 3-diphenyl-2-propanols
the reaction is controlled by the stability of the olefin formed.
In these cases, the olefin formed conjugated with a p-methoxy phenyl
group predominated by a factor of ca. three over that of an olefin
formed conjugated to a phenyl group. These results would indicate
that the p-methoxyphenyl group stabilizes the incipient double bond
compared to a phenyl group. The fact that the cis-p-methoxy compound
(I) reacts faster than 1,1-diphenyl olefin (III) could be explained,
therefore, by the greater stability of the triple bond being formed
in the reaction of the cis-p-methoxy compound (I). The trans-p-
methoxy compound (I) reacts faster than 1,1-diphenyl olefin (III) due
to the p-methoxyphenyl group aiding in the rearrangement with elec-
trons. Conversely, the trans-p-chloro compound (II) retards the rate
of the reaction compared to 1,1-diphenyl olefin (III) because the
p-chlorophenyl group destabilizes the rearrangement of the trans group
with its electrons. Comparing the rate constants of the cis and trans
isomers to each other is consistent with mechanism 3. In view of
this and the reasonable rationalization of the relative rate constants
of the cis and trans isomers to 1,1-diphenyl-2-bromoethylene, mechanism
3 is certainly the most probable.
There remains the question of just what is the driving force
of the rearrangement, or why do groups rearrange with their electrons
to an electron-rich site? In the Beckmann rearrangement the polarity
of the N-OX bond is thought to influence the rearrangement . The
more powerfully the OX group attracts electrons the faster the rear-
rangement. This could conceivably give rise to a partial breaking of
the N-OX bond, leaving a vacant orbital to which the trans group could
migrate. There is no evidence that there is any actual dissociation
into free ions. This would result in a nitrogen sextet intermediate
sometimes postulated for the Beckmann rearrangement .
This intermediate is unreasonable due to the stereospecificity of
It is then conceivable that the partial breaking of the
carbon-bromine bond and the probable rehybridization to give a vacant
p orbital, into which the trans group could migrate, would be the
driving force of the rearrangement. There are many cases of ioniza-
tion of a halide resulting from the formation of a carbanion. The
most common of these are the well known carbene reactions [34-36.
In particular, the reaction of chloroform with sodium hydroxide
involves an initial rapid proton transfer and then a slow loss of
chloride ion from the resulting carbanion [34 .
CHC13 + OH" "-- ~CC13 + H20
CC13 low :CC12 + Cl"
The fact that the chloride ion does not ionize until the carbanion
is formed might explain why the rearrangement is initiated by the
removal of the vinyl proton. Once the vinyl carbanion is formed the
bromide atom can ionize, at least to some degree, causing the forma-
tion of the vacant p orbital necessary for the rearrangement. An ex-
treme case of this might be depicted as such:
/C = C-
This poses the problem of why, in the Beckmann rearrangement,
the rate constant of the p-methoxyphenyl group migrating is so much
greater than that of the phenyl group migrating, while in our rearrange-
ment it is not? For instance, R. Huisgen and co-workers  have found
trans-2-methoxyacetophenone oxime to react ca. 150 times faster than
acetophenone oxine, yet trans-l-methoxy compound I reacts only 1.6 times
as fast as cis-p-methoxy compound I. Even though the nitrogen atom in
the Beckmann rearrangement is isoelectronic with the carbanion in
our rearrangement the fact remains that the nitrogen atom is neutral
and the carbon has a negative charge. If the driving force of the
reaction is the partial breaking of the C-Br bond, this would still
leave some negative charge on the carbon atom while the partial break-
ing of the N-OX bond would leave the nitrogen atom partially positive.
This difference in charge on the migrating terminus affecting the
rates of reaction is reflected in other rearrangements where the group
is migrating with its electrons. For instance, the rate constant for
the migrating p-methoxyphenyl in the symmetrical pinacol rearrangement
is five hundred times greater than the migrating phenyl group .
In this rearrangement the group is believed to migrate to a positively
charged terminus. In the Lossen and Hofmann rearrangements the phenyl
or substituted phenyl groups are migrating to a negatively charged ni-
trogen atam. Hauser, et al. , have indicated the migrating p-me-
thoxyphenyl group to be ca. six times as fast as the migrating phenyl
group in the Lossen rearrangement while in the Hofmann rearrangement
they found it to be approximately ten times as fast. This comparison
of charge on migrating terminus would tend to predict a small differ-
ence in rate between p-methoxyphenyl and phenyl due to a group migrat-
ing with its electrons to an electron-rich site.
The p-chlorophenyl group, on the other hand, does not affect
the rates of rearrangement to a radical degree [28,29,39 Chapman
and Fidler  have shown that the picryl ether of p-chlorobenzophenone-
-oxime reacts 1/5 as rapidly as the picryl ether of benzophenone
under the same conditions. The fact that the trans-2-chloro compound
(II) reacts 1/2 as rapidly as 1,1-diphenyl olefin (III) is not sur-
prising from this comparison.
A kinetic study of the migratory aptitude, or tendency
of a group to rearrange, was made on the cis and trans isomers of
a series of 1-phenyl-l-2-substituted phenyl-2-broaoethylenes. The
kinetics of the reactions of the bromoethylenes with base was fol-
lowed by measuring the ultraviolet absorption spectra of the pro-
ducts obtained. Rate constants for the rearrangement of the bromo-
ethylenes showed that an electron-donating group on the migrating
phenyl increases the rate of the reaction compared to a phenyl
group. Conversely, an electron withdrawing group on a migrating
phenyl retards the rate of the reaction. Kinetic results also in-
dicated that the cis-bromoethylenes reacted faster than 1,1-diphenyl-
Three mechanisms have been considered for the reactions and
the most reasonable consists of a rapid equilibrium involving the re-
action of the bromoethylene with base followed by a slow rearrange-
ment of the phenyl or substituted phenyl group with its electrons.
A comparison of the elimination-rearrangement reaction to
the Beckmann, Pinacol, Lossen and Hofmann rearrangements is made,
and the differences between the elimination-rearrangement reaction
and the others is rationalized.
LIST OF REFERENCES
1. P. Fritsch, Ann., 279, 319 (1894).
2. W. P. Buttenberg, ibid., 279, 327 (1894).
3. H. Wiechell, ibid., 279, 337 (1894).
4. G. H. Coleman and R. D. Maxwell, J. Am. Chem. Soc., 56, 132 (1934).
5. G. H. Coleman, W. H. Holst and R. D. Maxwell, ibid., 58, 2310
6. D. Y. Curtin and J. W. Crump, ibid., 80, 1922 (1958).
7. D. Y. Curtin and E. W. Flynn, ibid., 81, 4714 (1959).
8. A. A. Bothner-Dy, ibid., 77, 3293 (1955).
9. J. G. Pritchard and A. A. Bothner-By, J. Phys. Chem., 64, 1271
10. D. Y. Curtin, E. W. Flynn and R. F. Nystrom, J, Am. Chem. Soc.,
80, 4599 (1958).
11. D. R. Nielson and William E. McEwen, ibid., 76, 4042 (1956).
12. R. Stoermer and M. Simon, Ber., 37, 4167 (1904).
13. C. F. H. Allen and S. Converse, "Organic Syntheses, Collective
Volume I," John Wiley and Sons, Inc., New York, 1951, p. 226.
14. E. Hepp, Ber., 7, 1410 (1874).
15. E. Bergmann, J. Chem. Soc., 402 (1936).
16. L. I. Smith and U. M. Falkof, "Organic Syntheses, Collective
Volume III," John Wiley and Sons, Inc., New York, 1955, p. 350.
17. A. Orekhoff and M. Tifteneau, Bul. Soc. Chim., 37, 1414 (1925).
18. L. F. Fieser, "Experiments in Organic Chemistry," D. C. Heath and
Company, Boston, 1955, pp. 180, 185.
10. P. Rumpf and M. Gillois, Bul. Soc. Chim., 1348 (1955).
20. E. Goldman and D. Byles, Jour. AWWA, 55, 1051 (1959).
21. W. J. Youden, "Statistical Methods for Chemists," John Wiley
and Sons, Inc., New York, 1951, p. 42.
22. J. Timmermans, "Physico-Chemical Constants of Pure Organic
Compounds," Elsevier Publishing Co., Inc., New York, 1950.
23. G. Indastand, Z. Physik. Chem., B2, 428 (1929).
24. C. P. Smyth, "Dielectric Behaviou and Structure," McGraw-Hill
Book Company, Inc., Now York, 1955, p. 244.
25. W. L. G. Gent, Quart. Rev. Chem. Soc., II, 383 (1948).
26. C. P. Smyth, "Dielectric Behavior and Structure," McGraw-Hill
Book Company, Inc., Now York, 1955, p. 253.
27. L. F. Sutton, Proc. Roy. Soc. (London), 133A, 668 (1931).
28. (a) W. E. Bachmann and J. W. Ferguson, J. Am. Chem. Soc., 56,
(b) W. E. Bachmann and 11. R. Steinberger, ibid., 56, 170 (1934).
29. B. Jones, Chem. Revs., 35, 335 (1944).
30. (a) H. E. Zimmerman and A. Zweig, J. Am. Chem. Soc., 83, 1196
(b) H. E. Zimmerman and F. J. Smentowski, ibid,, 79, 5455 (1957).
31. G. E. Hall, R. Piccolini and J. D. Roberts, ibid., 77, 4540 (1955).
32. C. IH. Depuy and R. E. Leary, ibid., 79, 3705 (1957).
33. G. H. Wheland, "Advanced Organic Chemistry," 2nd Ed., John Wiley
and Sons, Inc., New York, 1949, p. 483.
34. (a) J. Hine, J. Am. Chem. Soc., 72, 2438 (1950).
(b) J. Hine, ibid., 76, 2688 (1954).
35. P. S. Skell and A. Y. Garner, ibid., 78, 5430 (1956).
36. D. Y. Curtin and W. HI. Richardson, ibid., 81, 4719 (1959).
37. R. Huisgen, J. Witte, H. Walz and W. Jira, Ann., 604, 191 (1957).
38. (a) W. B. Renfrow, Jr., and C. R. Hauser, J. Am. Chem. Soc., 59,
(b) W. B. Renfrow, Jr., and C. R. Hauser, J. Am. Chem. Soc., 59,
39. A. W. Chapman and F. A. Fidler, J. Chem. Soc., 448 (1936).
Ralph Anthony Damico, Jr., was born in Chester, Pennsylvania,
on July 2, 1935. In June, 1957, he received the degree of Bachelor
of Science in Chemistry from Saint Joseph's College, Philadelphia,
Pennsylvania, and enrolled in the graduate school of the University
of Florida. During his graduate study at the University of Florida
he has held a graduate assistantship, a teaching assistantship and
a Texaco research fellowship in the Chemistry Department.
Mr. Damico is married to the former Darrina Dee Turner. He
is a member of the American Chemical Society and of Gamma Sigma
This dissertation was prepared under the direction of the
chairman of the candidate's supervisory committee and has been ap-
proved by all members of that committee. It was submitted to the
Dean of the College of Arts and Sciences and to the Graduate Council,
and was approved as partial fulfillment of the requirements for the
degree of Doctor of Philosophy.
February 3, 1962
Dean, College of Arts an
Dean, Graduate School