Aug.

20, 1953 CONVERSION

OF BENZONITRILE
TO BENZAMIDE 3961

[CONTRIBUTION FROM THE DEPARTMENT

OF CHEMISTRY AND CHEMICAL ENGINEERING,
UNIVERSITY OF WASHINGTON 1

The Mechanisms of Hydrogen Peroxide Reactions. I. The Conversion of Benzonitrile

to Benzamide
BY KENNETH
B. WIBERG
RECEIVED
APRIL6, 1953

The rate of reaction between hydrogen peroxide and benzonitrile shows a first-order dependence on the concentration of
the nitrile, hydrogen peroxide and hydroxyl ion. Benzonitrile oxide has been shown not to be an intermediate in the reaction,
and the oxygen introduced into the nitrile has been demonstrated to arise from hydrogen peroxide and not from water or
hydroxyl ion. The effect of substituents on the rate of the reaction parallels that observed in nucleophilic attacks on the
carbon of the nitrile. These data suggest that the first step is a reaction between the nitrile and peroxide anion, folIowed by
u fast reaction with a molecule of hydrogen peroxide. The peroxide anion appears to be a particularly effective nucleophilic
agent in its reaction with benzonitrile.

Introduction was inert to neutral potassium permanganate. Hydrogen

peroxide was a C.P. grade of 3% strength, the same lot
Hydrogen peroxide is an unusual reagent in that being used for all the experiments.
it is known to function both as an oxidizing agent The boric acid solution (0.82 M) was prepared by cooling
and as a reducing agent. Although many examples a saturated aqueous solution to 20' for two hours, followed
of both types of reactions are known, little has been by filtration. The sodium hydroxide solution was 0.100 N ,
prepared carbonate free. The sodium thiosulfate solution
done in an effort to elucidate the mechanisms of (0.1 N ) was used within a week after its preparation.
these transformations. A particularly interesting Analytical Method.-The rate of reaction was followed
example is the reaction of an aromatic nitrile with by measuring the rate of disappearance of hydrogen per-
hydrogen peroxide in weakly basic so1ution.l oxide. Ten milliliter aliquots were removed from the re-
action flask and added to a solution of two grams of potas-
The stoichiometry of the reaction suggests that it sium iodide and 10 ml. of 18 N sulfuric acid in 125 ml. of
effectively involves an oxidation-reduction of water. The liberated iodine was titrated with 0.1 N sodium
hydrogen peroxide with simultaneous hydration of thiosulfate solution. Under these conditions, neither ace-
the nitrile. tone nor oxygen interfered with the analysis.
The pH of the solutions was determined using a Beckman
0 model G $H meter. It was noted that an 0.001 N solution
I/ of sodium hydroxide in 25% acetone gave an observed PH
CsHSCX + 2HzO2 +CsHs-C-NHn + Ot + H20 of 11.4. The PH values determined in this solvent are
therefore approximately 0.4 pH unit high.
It was believed that an examination of the kinet- Kinetic Method .-One hundred milliliters of an acetone
ics and other pertinent data concerning this re- solution containing four times the desired quantity of the
action would lead to the formulation of a mecha- nitrile was prepared. Ten milliliters of 0.82 M boric acid
nism for the reaction, and also might give one an solution, the necessary amount (5, 10 or 20 ml.) of 0.1 N
sodium hydroxide solution for the desired PH, 25 ml. of the
insight into the nature of the other reactions of nitrile solution and 20 ml. of distilled water were mixed in
hydrogen peroxide. a 100-ml. volumetric flask and allowed to warm in a thermo-
stat maintained at 40.0'. The desired amount of hydrogen
Experimental peroxide solution (10 or 20 ml.) which had been warmed to
The example chosen for study was the reaction of benzo- 40' was added at zero time, and sufficient 40' water was
nitrile with hydrogen peroxide in buffered aqueous acetone added to make 100 ml. The reaction flasks were wrapped
solution at 40'. This temperature was chosen because the in aluminum foil in order to minimize photochemical de-
decomposition of hydrogen peroxide is fairly slow under composition of the peroxide. Most of the runs were made
these conditions.2 In order to obtain a sufficiently high in duplicate, the results thus obtained being in very good
concentration of the nitrile, it was found necessary to add agreement. The reactions were carried to 2040% com-
an organic solvent. Acetone, ethanol, dioxane and tetra- pletion.
hydrofuran were considered a s possible solvents. Acetone Blank runs were made simultaneously with each rate de-
was chosen since hydrogen peroxide showed a lower rate of termination in order to estimate the extent of self decompo-
decomposition in 25% acetone than in mixtures of the other sition of the hydrogen peroxide. This reaction was about
solvents with water. Acetone-water has been used with one-seventh as fast as the nitrile reaction, and was corrected
success as a solvent for this reaction.' for in calculating the rate constants. The rate constants
Since it is known that the reaction is base catalyzed,' it given are in effect based on the rate of appearance of product
was necessary to use a buffer in order to control the pH. It and were calculated using the natural logarithm. The
was observed that the addition of hydrogen peroxide to the probable error in the rate constants was estimated to be
borate buffer used, caused a decrease in PH in excess of that about 10%.
due t o the sulfuric acid present a s a stabilizer in the hydro- It was not possible to use 25% acetone as the solvent for
gen peroxide. It is known3 that boric acid reacts with hy- the reaction of the substituted benzonitriles because of their
drogen peroxide to form perboric acid which is a stronger limited solubility. Therefore, 50% acetone was employed.
acid than boric acid. However, the equilibrium constant The method used was the same a s above, except that 50 ml.
for the reaction is very small, indicating the presence of only of an acetone solution and correspondingly less water was
a small quantity of perboric acid in the reaction mixtures. used.
The amount of hydrogen peroxide removed from solution by The Reaction of Benzonitriie with Hydrogen Peroxide.-
this reaction was estimated to be less than 5% for most of Benzonitrile oxide was prepared by the method of Wieland4
the runs. A similar effect was noted when a phosphate from benzhydroxamyl chloride and dilute base. It was
buffer was used,although it was of a smaller magnitude. used immediately in order to avoid polymerization. To a
Reagents.-Benzonitrile was a purified Eastman Kodak mixture of 1.5 ml. of 20% hydrogen peroxide and 5 ml. of
Co. white label grade. The other nitriles were obtained water was added 0.8 g. of benzonitrile oxide and sufficient
from the same source. Acetone was a C.P. grade which acetone to effect solution. After the addition of 3 ml. of
10% sodium carbonate solution, the mixture was allowed to
(1) B.Radziszewski, Be?., 17, 1289 (1884); L. McMaster and C. R. stand overnight. The acetone and water were removed by
Noller, J . l a d . Chcm. Soc., 12, 652 (1935): J. V. Murray and J. B. evaporation on a steam-bath. The infrared absorption
Cloke, THISJOURNAL, 56, 2749 (1934). spectrum of the product (in acetonitrile solution) was com-
(2) Fr. BUrki and Fr. Schaft, Hclv. Chim. A&, 4, 418 (1921).
(3) H.Menzel. Z. p h y s i k . Chem., 105,402 (1923). (4) H. Wielsnd, Bev., 40, 1670 (1907).
3962 H WIBERG
K E ~ T B. Vol. 75
pared with that of benzamide, and indicated the absence of TABLE I1
a significant amount of the latter. The product was found Solvent, 25% acetone; temp., 40.0'.
not to melt under 250",and on heating on a spatula, melted,
charred and emitted a lachrymatory vapor. Under these 0.1
Na-N 0.82.44 k X k101d
X
conditions, using twice the quantity of hydrogen peroxide, Run 0 H . b HaBOs,b 2nd 3rd
benzonitrile gave a quantitative yield of benzamide. no. [HxO;] [CsHsCN] pH" ml. ml. order order
The Reaction of Benzonitrile with Hydrogen Peroxide 1 0.180 0.090 7.58 10 10 3.9 5.1
and H~0l~.-To a mixture of 1 ml. of 90% hydrogen per-
oxide and 5 ml. of water containing 1.4% HnOUwas added 1 2 .180 .090 7.58 20 20 3.7 4.9
g. of benzonitrile and sufficient acetone to effect solution.6 3 .E42 .090 7.58 5 10 3.8 5.0
After the addition of 3 ml. of 10% sodium carbonate solu- 4 .184 .045 7.58 10 10 4.0 5.3
tion, a n exothermic reaction occurred and oxygen was 5 .181 .090 8.07 20 10 12.3 5.2
evolved. After standing overnight, the solvents were re- 6 .183 .090 7.19 5 10 1.47 4.8
moved by evaporation, and the product was washed with
cold water and dried. 7 .181 .090 8.06 11.2 4.8
The product was analyzed for its 0" content by passing So .I82 .090 8.08 10 10 16.3 . ..
its vapor, mixed with nitrogen, through a tube containing " T h e observed PH values in this solvent are approxi-
granular quartz heated to 425°.8 The benzonitrile and mately 0.4 pH unit high. The observed values are listed
water formed were collected in a small trap and equilibrated in this table. Volume per 100 ml. of reaction solution.
with carbon dioxide overnight. Analysis of the carbon Observed rate constant: -d[H~Ol]/dt = 2&[H202][RCN];
dioxide using a mass spectrometer indicated that the amide units, 1 mole%ec.-'. Calculated rate constant: -d-
contained 0.030% 0l8over normal abundance.'
[H,Oz]/dt = 2&(K./K,)[HaOz][RCN][HO-];Ks = 2 X
lO-'a. e 25% ethanol was employed for this run. f 10 ml.
Results of 1 .O N NazHPOd and 2 ml. of 0.5 N HCl were used per 100
The results of a typical kinetic run are shown in ml. of reaction mixture.
Table I. These results clearly indicate the reaction
which a phosphate buffer was used, gave essentially
TABLE
I the same observed rate constant as run 5, indicating
Solvent, 25% acetone (by volume); temp., 40.0'; [HtOz] again that the buffer did not enter into the re-
= 0.180; [CsHsCN] = 0.090; PH 7.58. action. Finally, run 8 was carried out in %yo
Time
Reaction
of HzOr, %
Reaction
of C E H ~ C N%,~
k X 10'
2nd order
ethanol, by volume, instead of 25% acetone, and
gave qualitatively the same rate as run 5 which
30 1.4 1.2 3.8
60 2.8
had the same pH. This indicated that the solvent
2.4 3.8 acetone did not enter into the reaction. The third
120 5.2 4.3 3.5
240
order rate constants listed in the last column are
10.9 9.5 4.0
300 13.0 11.2 4.0
based on the observed pH value which is about 0.4
PH unit high. Thus, these values of K should be
450 19.7 17.0 4.3
multiplied by a factor of 2.5 to obtain the actual
1380 43.8 39.5 4.5 rate constants.
Calculated from the rate of disappearance of hydrogen The results of an investigation of the effect of
peroxide from the reaction mixture, and a n independent
blank. substituents on the rate of the reaction are shown
in Table 111. The introduction of a chloro group
to be of the second order. The increase in k with
time can be attributed to the small increase in pH TABLE I11
as the hydrogen peroxide concentration decreased. Solvent, 50% acetone; [HzOZ]= 0.180; temp., 40.0'
A more significant conclusion may be reached on Run k X 10'b
no. Compound Concn. pHa 2ndorder
the basis of the data in Table 11. In this case the 9 Benzonitrile 0.090 8.60 3.6
rate constants were derived from the initial rate 10 p-Chlorobenzonitrile .090 8.60 15
of the reaction in order to avoid errors caused by a 11 p-Nitrobenzonitrile ,050 8.60 85
change of @H during the reaction. In all cases,
a Swain and Scott, THISJOURNAL,
the rate constants were calculated for a considerable indicated that the observed PH values 75, 141 (1953),have
in 50% acetone
proportion of the reaction, with results similar to are high by about 1.2 pH units. The table gives observed
those shown in Table I. pH values. See note c , above.
A comparison of runs 1 and 2 indicates that the
borate buffer did not enter into the reaction, since increased the rate, and the effect of a nitro group
the rate was essentially the same when the buffer was observed to be considerably greater. These
concentration was doubled. Runs 1, 3 and 4 reactions were carried out in 50% acetone rather
indicate the reaction to be first order in both than 25y0 acetone because of the limited solubility
hydrogen peroxide and in benzonitrile, since halving of the substituted compounds. It may be noted
the concentration of either compound resulted in no that the rate constant in 50% acetone was 3.6
significant change in the second-order rate con- X 1. mole-' set.-' a t a corrected PH of 7.4,
stant. Runs 1, 5 and 6 showed the rate of reaction and in 25y0 acetone it was 3.9 X at a corrected
to be proportional to the concentration of hy- PH of 7.2. The effect of changing dielectric con-
droxyl ion within experimental error. Run 7, in stant is therefore small.
In order to obtain some information concerning
(5) This is essentially the procedure suggested by Murray and Cloke, this reaction, other than rate data, the conversion
ref. 1.
(6) R. S. Boehner and A. L. Ward, THIS JOURNAL, 88, 2506 (1910), of benzonitrile to benzamide was carried out in the
found that benzamide was almost quantitatively converted to benzo- presence of H20'8.8 Pyrolysis of the benzamide
nitrile and water under these conditions. The products obtained in the followed by analysis of the water gave a value of
first run were discarded in order t o prevent dilution of the product 0.030% excess 01*over normal abundance. A
water with residual water in the cracking catalysts.
(7) The 0 1 3 content was kindly determined by Mr. R. Wade using a (8) Obtained from the Stuart Oxygen Co., under allocation from the
Consolidated Engineering Corp. Model 21-201 mass spectrometer. Atomic Energy Commission.
Aug. 20, 1953 CONVERSION
OF BENZONITRILE
TO BENZAMIDE 3963

value of 0.6-0.7ojO would be expected if the oxygen If the proposed mechanism may be considered
arose from water or hydroxyl ion instead of hy- to be established, a comparison of the rate of this
drogen peroxide. reaction with other nucleophilic attacks on the
In order to determine if benzonitrile oxide could carbon of the nitrile would be of interest. The
be an intermediate in this reaction, the oxide was available data are summarized in Table IV. It is
treated with hydrogen peroxide under conditions apparent that peroxide anion is a particularly
which led to a quantitative conversion of benzo- effective nucleophilic agent in its reaction with
nitrile to benzamide. The product of the reaction benzonitrile.
was a high melting, unidentified solid which ap- TABLEIV
peared to contain little if any benzamide. Temp.,
Anion Solvent "C. k, 1. mole-' sec-1 Ref.
Discussion H02- 25% acetone 40 1
Since the rate of the reaction showed a first-order HS- 100% ethanol 69 2-8 X lo-* 10
dependence on the concentration of hydrogen EtO- 100% ethanol 25 z x 10-4 11
peroxide, the slow step must involve only one HO- 25% acetone 40 -1 X lo-' 12
molecule of the peroxide, and a successive fast
reaction must be responsible for the second hy- It is well known that sulfur containing anions
drogen peroxide molecule required by the stoichio- have high nucleophilic reactivity, as compared to
metry. The rate expression suggests that either hydroxyl ion, in the S N displacement ~ reaction.13
hydroxyl ion and hydrogen peroxide react indepen- This reactivity is presumed to be a consequence of
dently with the nitrile, or peroxide anion alone the higher polarizability of the sulfur, which will
reacts with the nitrile in the slow step. The first accommodate some of the extra electron density
possibility is very unlikely since the attack of a t the carbon in the transition state. It is also
hydroxyl ion on the nitrile would be expected to known that hydroxyl ion is particularly efTective
lead to simple hydrolysis without the participation in its reaction with P-propi~lactone,~~ this being due,
of hydrogen peroxide. a t least in part, to the possibility of having the
If peroxide anion is assumed to be the reacting carbonyl oxygen bear the negative charge in the
species, it can react with the nitrile either a t the transition state.la Although the present reaction
carbon (A) or a t the nitrogen (B). The inter- would appear to resemble the ester reaction more
mediate I could then react with a molecule of than the S N displacement,
~ a comparison with p-
hydrogen peroxide (C) to give the amide. It is not propiolactone is not fruitful since hydroxyl ion be-
equally apparent how the nitrile oxide might give haves differently in this reaction than do other ions.
the amide. The fact that benzamide was not Unfortunately, the reactions of ordinary esters
obtained from benzonitrile oxide makes reaction B or other acid derivatives with nucleophilic agents
appear unlikely. Reactions A and C then remain have been little studied. Although it is known
as a reasonable interpretation of the course of the that some esters and acid chlorides react with
reaction. sulfhydryl ion to give thiol acids,16the rate of this
Hz0 reaction as compared to that of hydroxyl ion does
CeHiCEN + H0z8 slow
+ CeHi-C=NB
I
+
fast
not appear to have been measured. If sulfhydryl
0-OH
ion were more reactive than hydroxyl ion in this
e reaction, then the reactivity of peroxide anion
CeHi-c=NH + OH (A) might be explained by an argument similar to that
I for the sulfur containing anions. Although per-
0-OH
e oxide anions does not have the benefit of higher
CsHsC=N + H0z8 +C&-C=N + 0 + OH (B) polarizability a t the attacking atom, it does have
n n the possibility of transferring part of the negative
CeHb-C=NH H-0-0-H + charge to the other oxygen. It is interesting to
Id note that even though peroxide anion has a nega-
OzOH
tive charge, it in effect transfers a neutral oxygen.
CeHs-C-NHz + 02 + H e + H 0 8 (C) If, on the other hand, hydroxyl ion is more &ec-
I/ tive than sulfhydryl ion in its reaction with esters
0
or acid chlorides, it would then be necessary to
If the proposed mechanism is correct, the intro-
(10) K. Kindler, ref. 9, gave a value of 9 X 10-8 I. mole-1 min.-l
duction of substituents into benzonitrile should as the first-order rate constant for the reaction between benzonitrile
result in a change of rate which would correspond and a constant concentration of sulhydryl ion. He unfortunately
to that observed in a nucleophilic attack on benzo- neglected to mention the concentration of sulfhydryl ion. From the
nitrile. Kindler;s in his study of the reaction of experimental details, one may calculate that the HS- concentration
substituted benzonitriles with sulfhydryl ion, has must have been between 0.01-0.002 M ,depending upon the size of the
volumetric flask he employed (50-250 d.).This reaction will be rein-
shown that chloro and nitro groups accelerate the vestigated.
reaction in that order. The same order has been (11) E. K. Marshall, Jr., J. P. Harrison and S. F. Acree, Am. Chcm.
observed in the present case, lending further sup- J . , 49,400 (1913), have determined the rate of the reverse reaction and
also the equilibrium constant. From these data, the rate of the for-
port to the proposed mechanism. Additional ward reaction was calculated.
evidence may be found in the observation that the (12) Benzonitrile (0.1 M ) and sodium hydroxide (0.1 M ) in 25%
oxygen introduced into the nitrile during the acetone at 40' gave about 45% reaction in 22 hr. as determined by
reaction arose from hydrogen peroxide rather than the ultraviolet absorption spectrum.
from water or hydroxyl ion. (13) Cf.C. G. Swain and C. B. Scott, THISJOURNAL, 76, 141 (1953).
(14) P. D. Bartlett and G. Small, ibid., 7'2, 4867 (1950).
(9) K. Kindler, Ann., 460, 1 (1926). (15) A. Kekule, 2. Chem.,106 (1867).
3964 JACK HINEAND W. H. BRADBR,JR. VOI. 75
examine the reactions of nitriles in order to find an directed toward a more detailed investigation of
explanation for the high reactivity of both sulf- these points.
hydryl ion and peroxide anion. Current work is SEATTLE 5, WASHINGTON

[CONTRIBUTION FROM THE SCHOOL OF CHEMISTRY OF THE GEORIGA

INSTITUTE OF TECHNOLOGY]

The Effect of Halogen Atoms on the Reactivity of Other Halogen Atoms in the Same
Molecule. 111. The SN2Reactivity of Ethylene Halides’
BY JACK HINEAND W. H. BRADER,
JR.
MARCH
RECEIVED 25, 1953

The interpretations of earlier data as showing that the replacement of ,&hydrogen atoms in alkyl halides by halogen atoms

- - -
causes increased sN2 reactivity are shown t o be unreliable. The kinetics of the bimolecular substitution reaction of sodium
thiophenolate with ethyl bromide and five of its B-substituted derivatives have been studied. For the various &substituents,
the rates vary thus: H > CtHb CHa > C1 Br F. The heats of activation are found t o increase with electron-with-
drawing power of the substituents studied, while the entropies of activation decrease with increasing size.

The substitution of a p-hydrogen atom in an alkyl have now found that the reaction of ethylene iodide
group by a halogen atom is commonly said2 to in- with sodium thiosulfate yields largely ethylene, it is
crease the reactivity by the s N 2 or one-step bimo- clear that these data may not be used as the basis
lecular displacement mechanism. The evidence for generalizations concerning displacement on car-
quoted includes the data of Tronov and Gershe- bon.
vichI4who reported rate constants for the reaction It should also be noted that although Slator ob-
of sodium methoxide with n-propyl bromide, iso- tained reasonably steady first-order rate constants
propyl bromide, propylene bromide, trimethylene for the reaction of ethylene chlorobromide and chlo-
bromide, 1,%,&tribromopropane and other halides roiodide with excess sodium thiosulfate, these are
of propane. From what is known of the chemistry not the rate constants for first-order solvolysis, as
of polyhalides of this type it might be expected has been suggested.’ This follows from the fact
that a major reaction with the strongly basic rea- that the rate constants for the solvolysis of
gent sodium methoxide would be dehydrohalogen- these compounds are smaller by factors of more
ation.6 Thus, the fractions of the rate constants than one thousand than the constants reported
obtained by these workers which are due to sub- by Slator,* as well as from inspection of some
stitution reactions are not definitely known. The of Slator’s data.g Although a second-order reac-
data of Slator, who studied the reactions of sodium tion in which rate “constants11climb (due to some
thiosulfate with a number of halogen compounds,6 replacement of the second halogen atom) could give
have also been quoted in this connection. Slator reasonable first-order rate constants, we cannot be
indeed found that the second-order rate constant sure that this is being observed in the present case,
for the reaction with ethylene iodide is considerably since another side reaction, hydrolysis of the Bunte
greater than that for reaction with ethyl iodide, salt to a mercaptan and acid, was shown by Slator
but no attention appears to have been called to to be present and other reactions such as dehalogen-
his observation that ethylene bromide was consid- ation may also have been occurring.
erably less reactive than ethyl bromide. Since we From the data of Taylor’O the rate of the second-
(1) Papers I and I1 of this series: J. Hine and D. E. Lee, THIS
order substitution reaction of ethanolic alkali with
JOURNAL, 73,22 (1951); 74, 3182 (1952). ethylene bromide may be calculated to be two or
(2) E. R. Alexander, “Principles of Ionic Organic Reactions,” three times that for reaction with ethyl bromide,
John Wiley and Sons, Inc., New York, N. Y.,1950, p. 89; M. J. S. but this calculation depends on a difficult and not
Dewar, “ T h e Electronic Theory of Organic Chemistry,” Oxford altogether satisfying analysis of the products of
University Press, London, 1949, p. 71; A. E. Remick, “Electronic
Interpretations of Organic Chemistry,” 2nd Ed., John Wiley and the former reaction.”
Sons, Inc., New York, N. Y.,1949, p. 394. E. D. Hughes and C. K. We have not been able to devise a method for
Ingold [ J . Chem. Soc., 244 (1935)1, have also made a statement of determining accurately the effect of &iodine atoms
this type, but in an article which has come to our attention since this on s N 2 reactivity, since the nucleophilic reagents
investigation began [E. D. Hughes, Quart. Reus., (London), 6, 264
(1951) 1, it is recognized t h a t none of the existent data unambiguously which have been studied attack the @-iodineatom
describe the rate of a substitution (in contrast to elimination) process (7) L. P. Hammett, “Physical-Organic Chemistry,” McGraw-Hill
and stated “. . . investigations of suitable systems are being under- Book Co., Inc., New York. N. Y., 1940, p. 208. See also E. A. Moel-
taken.” wyn-Hughes, A’dIiw~. 133, 294 (1934).
(3) For thr significance of the term 9 ~ 2 see . 1. Dustruvsky, I? I). (8) J , Hine and W H. Brdder, unpublished work.
Hughes and C. K. Ingold, J . Chem. S O L , 173 (19461, and earlier (9) The rate constants calculated as first order in ethylene chloro-
papers. bromide increase with increasing thiosulfate concentration. T h e order
(4) B. V. Tronov and A. I. Gershevich, J . Russ. P h y s . Chon. Soc.. of a reaction in which an excess of ethylene chloroiodide was used was
69, 727 (1927); c. A , , aa, 3380* (1928). reported t o be about 1.85.
( 5 ) Sodium hydroxide and 1,2,3-tribromopropane give 2,3-dibromo- (10) W. Taylor, J. Chem. SOL.,1514 (1935).
propene in more than 80% yield (R. Lespieau and M. Bourguel, “Or- (11) The work of B. V. Tronov and L. V. Laduigina [ J . Russ.
ganic Syntheses,” Coll. Vol. I, 2nd Ed., John Wiley and Sons, Inc., Phys. Chem. Soc., 69,2105 (1930); C. A., 26,3957 (1931) 1, P. Petrenko-
New York, h-.Y., 1941, p. 209). Ethylene bromide with sodium hy- Kritchenko and V. Opotsky [Bcr., 69B,2131 (1926)l and many other
droxide in homogeneous aqueous dioxane solution gives more than 90% data on reactions of nucleophilic reagents with 1,2-dihslides, also prob-
elimination (J. Hine and W. H. Brader, unpublished work). ably include large amounts of elimination as well as substitution reac-
(a) A. Slator, J. Chcm. SOC.,86. 1286 (1904). tiond.