W T '

~“‘

H!

W
I

III

I
II

III

0—;
04> ;|
N01

 

 

A STUDY OF THE ALKYLATION AND
ARYLATION OF NITRILES WITH HALOGEN
COMPOUNDS

Thesis for the Degree of M. S.
MICHIGAN STATE COLLEGE

Andrew A. HoIzschuh
I955

T H v.81.

.1

1.1: Lo 1—;

Jyvr~y "r“

A STUDY OF THE ALKYLATION AND ARYLATION

CF NITRILES mI'I‘H HAMLEN COMPOUNDS

By

And rev .1 , Holzschuh

A THESIS

Submitted to the School of Graduate Studies of Michigan
State College of Agriculture and Applied Science
in partial fulfillment of the requirements
for the degree of

MASTER CF SCIENCE

Department of Chemistry

1955

9hmw>
0

ACKNOWLEDGMENT

The author wishes to express his
appreciation for the aid and guidance
given by Doctor Gordon L. Goerner dur-
ing the investigation and preparation

, of this thesis.

flm'k
*W

**

ii
352594

VITA

Name: Andrew A. Holzschuh

Date of Birth: August 20, 1913

Place of Birth: ‘Waters, Michigan

Academic Career: High school at Gaylord, Michigan, 1926 to 1930

'Michigan College of Mining & Technology at
Hbughton, Michigan, 1930 to l93h

Degrees Held: Bachelor of Science in Chemical Engineering

Employment: Dow Chemical Company at Midland, Michigan as a
Research Chemist and Supervisor since 1935.

iii

A STUDY OF THE ALKTLATION AND MUTATION

OF NITREES WITH ILKLOGEN CCNPOUNDS

By

Andrew A. Holzschuh

AN ABSTRACT

Submit ted to the School of Graduate Studies of Michigan
State College of Agriculture and Applied Science
in partial fulfillment of the requirements
for the degree of

MASTER OF SCIENCE
Department of Chemistry

Year .1955

Approved glaze“; 56W

ABSTRACT

This investigation was begun as an attempt to phenylate
propionitrile with bromobenzene using sodium hydride as a condensing
agent. Much polymerization of the propionitrile occurred and no
phenylated product was obtained.

Bergstrom and Agostinho (1) reported successful phenylation of
prepionitrile with chlorobenzene when they used potassium amide as a
condensing agent. They obtained L31 of'the mono-substituted product,
hydratroponitrile, and some higher boiling material. This preparation
was repeated as described by these authors. The first fraction,
which would have contained the hydratrOponitrile, boiled over a very
wide range. It was not further investigated. If it had all been
hydratrOponitrile, it would have amounted to 27.7% yield. The second
fraction boiled reasonably constant at 157° at 3 mm. pressure and it
probably was 2-phenylhydratr0ponitrile. It amounted to 33.6% yield
based on the prepionitrile used. It was not further investigated.

One attempt was made to phenylate phenylacetonitrile with bromo-
benzene and sodium hydride. None of the phenylated product was
obtained.

One attempt, also, was made to alkylate propionitrile with nAbutyl
bromide and sodium hydride. The monoalkylated product, 2-methylhexane-
nitrile, CH,-(CH;),JEECBN, was obtained in.23.6% yield based on the
propionitrile used. Only a trace, if any, of the dialkylated product

was obtained.

iv

In order that the work would not get unduly extensive, at this
point the objective was narrowed down to the preparation of the butyl

substituted 2-phenylbutanenitriles,

C Ha-C Ha-Q—CN
R

from 2-phenylbutanenitrile and the butyl chlorides. The effect of the
configuration of the isomer on the yield of product was to be determined.
is a condensing agent both sodamide and sodium hydride were to be used
and comparison of these obtained.

‘With sodamide as a condensing agent, this reaction gave good
yields of alkylation product, when the n-butyl, s-butyl, and iébutyl
chlorides were used as alkylating agents. The yields of 2-butyl-2-
phenylbutanenitriles were: nebutyl, 87.3%; s-butyl, 91.7%; i-butyl,
90.11; t-butyl, approximately, 7.7%.

With sodium hydride as a condensing agent the figures were:
n-butyl, 73.0%; s-butyl, 80.7%; i-butyl, 79.3%. No preparation was
attempted with tAbutyl chloride and sodium hydride.

It was found impossible to hydrolyze these highly hindered nitriles
for preparation of derivatives. However, they could be readily reduced
to amines by the use of lithium aluminum hydride (2). This was done,
and the amines were readily converted to suitable derivatives. The
physical constants of the nitriles and the amines, and the melting

points of the derivatives are shown in the accompanying table.

—

B .

 

Infra-red spectra were obtained on the nitriles and the amines.
The nitriles all showed the characteristic absorption band at h.S
microns. In the spectra of the amines this band was no longer

present.

LITERATURE CITED
(1) F. W. Bergstrom and R. Agostinho, J. Am. Chem. Soc., él, 2152

(19h5).
(2) L. Amundsen and L. Nelson, J. Am. Chem. Soc., 1;, 2L2 (1951).

vi

TABLE OF PROPERTIES

 

 

 

0
Compound B.p., C D25 N25
D
Emil

2-Ethy1-2-pheny1hexanenitrile 126 0.9322 l.h97l
2-Ethyl-3-methyl-
2-phenylpentanenitrile 126 0.9h68 1.5038
2-Ethyl-h-methy1-
2-phenylpentanenitrile 119 0.9327 1.h976
2-Ethyl-3,3-dimethyl-2-phenyl-
butanenitrile a 123-126 0.9571 1.5107
l-Amino-Z-ethyl-2-phenylhexane b 13h 0.9283 1.5125
l—Amino-2~0thyl—3-methyl-
2-phenylpentane c 129.5 0.9h7l 1.5116
l-Amino-Z-ethyl- -methyl-
2-phenylpentane, 120.5 0.9279 1.5130

 

‘ This nitrile was obtained only in very small quantity and a
derivatives was not made. However, it was a solid and melted at

57.2-58.00 after recrystallization.
b M. p. of phenylthiourea, 118.5-119°,
c M. p. of Ci-napthylurea, 176-1770.

d M. p. of phenylthiourea, 1Lo.3-1h1°.

SUGGESTIONS FOR FURTHER‘WORK.....

SELECTED REFERENCES.........

TABLE OF CONTENTS

MROmCTIONOOOOOOOOOOOOOOOO...OOOCOOOOOOOOOOOOOOOOOOO0.0.0....

HISTORICAL....................... ..... .. .......... . ...... ......
EXPERIMENTAL..,.. eeeeeeeeeeeeeeee eoe ..... eeooeeeoeeoeeeoeeoeeee
ChemicaIS....................... ..... . ....... .............

General Information.......................................
Studies of the Arylation of Nitriles......................

A. Attempt to Arylate Prepionitrile in the Presence
of Sodium Hydride.................................

B. Arylation of Prepionitrile in the presence of
Potassium Amide...................................

C. Attempted Phenylation of Phenylacetonitrile in the
Presence of Sodium Hydride........................

Alkylation of Prepionitrile in Presence of Sodium Hydride.

llkylation of 2-Phenylbutanenitrile in Presence of
Sodamide.............................................

Alkylation of 2-Phenylbutanenitrile in Presence of Sodium
Hflri’de.O'COOOOOOOOOOOOOOOOOOOO0....1.0.... 000000 0...

Preparation Of Derivativeseooeeeeeeeeeee eeeee eeeee eeeeee as

A. Attempted Hydrolysis of Nitriles..................
B. Reduction of Nitriles and Preparation of

Derivatives.................... .......... ....
DISCUSSIONeeeoeeo-e eeeee eeeeeeeeeo e 000 0 090.00 0000000000 O
SUMMARYIO0000.00.00.00000000 eeeeee 00000 00 0 00000000 eeeeeee o

OOOOOOOOOOOOOOOOO. ..... 0.0. ........

10
12

13

15

19
21
21
2h
27
32
3b

LIST OF TABLES

TABLE
I Distillation of 2-Ethyl-3-methy1-2-phenylpentanenitrile....
II Yields and Physical Constants for the 2-Butyl-2-phenyl-
butanenitriIGS................................... ..........
III Attempted Alkaline Hydrolysis of 2-Buty1-2-phenylbutane-
nitriles...................................................
IV Attempted Acid Hydrolysis of 2-Butyl-2-phenylbutanenitriles
V Distillation Data for l-Amino-Z-butyl-Z-phenylbutanes......
VI Preperties of the 1-Amino-2-butyl-2-pheny1butanes..........

Page

18

2O

22
23
25
26

INTRODUCTION

This investigation was begun as a follow-up to some work reported
by Bergstrom and Agostinho (1). They found that prOpionitrile could
be aryiated by chlorobenzene in liquid ammonia in the presence of
potassium amide. The products obtained were the mono- and the diphenyl-

substituted propionitriles. The reaction is

H
H30'C’CN
Cl KNH
Q +CH3-CH2-CN -—-’—--> and

Lig . NH3

0

H30‘C‘CN

In this laboratory investigations have shown that the alkylation of
nitriles can be carried out quite satisfactorily and conveniently using
sodium hydride as the condensing agent with a toluene medium at some-
what elevated temperatures. Ring substituted phenylacetonitriles have
.been alkylated under these conditions with benzyl chloride and with
butyl and amyl bromides. Initially the object of the work reported
herein was to see if the use of sodium hydride could be extended to

the phenylation of propionitrile. A number of attempts to do this

resulted in failures. A number of variations of this plan were tried
and it was found that alkylation of propionitrile with n-butyl bromide
and sodium hydride could be accomplished with some success. In order
not to get started on.a more extensive investigation than was practical
the objective was narrowed down to the preparation of the butyl-

sutstituted 2-pheny1butanenitriles,

R - g - C N
a“:

by the use of sodamide as a condensing agent, and to make a minimum
investigation of the feasibility of the use of sodium hydride instead
of the amide for the same reaction. The effect of the configuration
of the butyl chloride on the yield of alkylation product was to be
determined, as well as the effect of the nature of the condensing agent.
In addition, characterization of the reaction products was to be carried

out, since they have not been reported in the literature.

HISTORICAL

The literature on the alkylation of nitriles has been quite
extensively reviewed in a recent thesis by Workman (2). Therefore,
only that work which seems most directly pertinent to this thesis will
be reviewed here. Ramart (3) reported the successful benzylation of
diethylacetonitrile, n-valeronitrile, and n-caprylic nitrile, using
benzyl chloride and sodamide in ether medium. Ziegler and Ohlinger (h)
reported that, after considerable difficulty with polymerization, they
finally succeeded in the alkylation of butyronitrile using n-butyl
chloride and sodamide. They claimed a 68% yield of alkylated product.
Later Ziegler (5) reported the successful alkylation of isobutyro-
nitrile with benzyl chloride and sodium hydride. The work of Bergstrom
and Agostinho (1) on the phenylation of prepionitrile by means of
chlorobensene and potassium amide was cited earlier.

Schuerch and Huntress (6) alkylated acetonitrile with ethyl bromide
and sodamide in ether and obtained substantial yields of the mono- and
diethylated products. Cristol and co-workers (7) investigated the
reaction of sodium hydride with various alkyl halides at usual reaction
conditions and found it to be inert. Workman (2) alkylated hydratropo-
nitrile with the four isomeric butyl halides, using sodamide as the
condensing agent. His yields‘were about 60%, except in the case of

the tébutyl chloride. For the latter the yields were 0512%.

EXPERIMENTAL
Chemicals

Prepionitrile - Eastman practical grade material was redistilled.
A small low'boiling fraction was discarded and the remainder,
which boiled at 95.5-960, was the material used. A bromate-

bromide titration showed that unsaturation was less than 0.01%.
Phenylacetonitrile - Eastman white label material was used as obtained.

2-Phenylhutanenitrile - This compound was prepared by the phOSphorous
oxychloride dehydration of Eastman practical grade <X-pheny1-
butyramide, by the method of Cram and Allinger (8). The product
was fractionated through a column packed with helices. The
fraction saved boiled at llO-lll.5o at 9 mm. pressure and had a

2S

D

refractive index N 1.5060. Reported values by the above

authors are b.p. 122-121:0 at 16 mm. pressure and N35 1.5075.

Bromobenzene, chlorobenzene, and n-butyl bromide - These were Eastman

white label grade materials and were used as obtained.

n-Butyl chloride - Eastman white label grade material was refractionated

and a low'boiling fraction of about L% was discarded. The fraction
used boiled constantly at 77°.

séButyl chloride - Eastman white label grade material was refractionated.

The fraction saved boiled at 66-67°.

t-Butyl chloride - Refractionation of Eastman white label grade

material gave a fraction which boiled at 50.50.

i-Butyl chloride - Eastman white label grade material. 0n refraction-
ation it was found difficult to get a material with a constant

boiling point. The material used boiled at 67-700.

Lithium aluminum hydride and sodium hydride - These compounds were used
as obtained from Metallic Hydrides, Inc. They were repackaged

into small bottles and stored in a desiccator.

Sodamide - This material was obtained from Farchan Laboratories.
It had been in storage for several years in sealed bottles.
It was good in appearance. It, also, was repackaged into small

bottles and stored in a desiccator.

Toluene and absolute ether - These materials were obtained from the

stockroom and were dried over sodium before use.

General Information

In all the experiments, unless otherwise stated, the reactions
were carried out in a 500 cc. three-necked flask, fitted with a stirrer,
thermometer, dropping funnel, and reflux condenser. The reflux con-
denser was fitted with a soda lime drying tube. Fractional distilla-
tions were carried out in either of two columns. One was twelve inches
long and packed with glass helices, the other was a fifteen inch long
vacuum jacketed Vigreux column.

Halide conversion figures were obtained by titration of the aqueous
layer by the Volhard method for halide ion.

All reported melting points were taken in capillary tubes and are
those obtained after the material had been recrystallized repeatedly
until the melting point increased only slightly or not at all. The
final values were determined with standardised thermometers graduated
in one-fifth degree divisions and immersed as marked.

All yield figures were calculated on the basis of the reactant
nitrile used. .

I. STUDIES OF THE ARYLATION OF NITRILES

A. Attempt to Arylate Propionitrile in the Presence of Sodium Hydride

Four different attempts as described below were made in an effort
to arylate prepionitrile with bromobenzene in the presence of sodium
hydride.

To the apparatus described above (see General Information) was
added a mixture of 25.7 g. (0.h67 mole) propionitrile and 38 g. bromo-
bensene. i suspension of 11.2 g. (O.h67 mole) sodium hydride in 38 g.
bromobenzene was placed in the dropping funnel. Enough more bromo-
bensene was used for rinsing to make a total of 150 g. (0.96 mole).
About one-third of the sodium hydride was added at once. A mild gas
evolution was observed. When the mixture was slowly warmed to 1000
it turned brown and gummy. It was cooled to 700 and the rest of the
sodium hydride was added during a few'minutes. The reaction mass,
‘which soon became so sticky that it was difficult to agitate, was heated
at 110-1200 for one hour. It was cooled and treated with a little
alcohol (to react with any remaining sodium hydride) and then with
water. An insoluble slimy impurity was removed by filtration. The
layers were separated and the oil layer was washed once with 50 cc.
water. The water layer was analyzed by the Volhard method and found
to contain 0.185 mole bromide ion. This corresponds to h0.5% conversion
based on the nitrile used. The oil layer was fractionated and approxi-

mately 6 g. low'boiling (up to the boiling point of bromobenzene, lSho)

material were obtained. Next an undetermined amount of bromobenzene
was distilled over. After that only a few drops of material could
be distilled even at 2 mm. pressure. The material remaining in the
distillation flask became very thick and sticky on cooling. It was
discarded.

In the second attempt toluene was used as a diluent and solvent.
The preparation was made with 0.67 mole of each reactant. The sodium
hydride was added with the aid of toluene (100 cc.) because the bromo-
benzene has such a high specific gravity that it floats the hydride
and makes it difficult to feed to the reaction flask. The reaction
mixture was heated to 65-700 and the hydride was added in one hour at
this temperature. The batch was then warmed slowly. At 90° it react-
ed with considerable evolution of heat. This mixture was heated
another half hour at about 1000 and became very sticky and balled up.
The batch was discarded at this point.

Since the nitrile appeared to have been polymerizing, a little
of it was treated with sodium hydride in a test tube. Only moderate
reaction took place at room temperature. when warmed up it reacted
more and more until, at 75-850 it became yellow and sticky. The first
reaction product may have coated the hydride and thereby slowed the
reaction. Heating could then dissolve or otherwise remove this layer
and permit the reaction to continue. A review of the literature
revealed that the formation of dimeric and trimeric condensation products
of acetonitrile and propionitrile can be effected by the alkali metals

(9,10,11,12,l9). The two forms are both soluble in acid. Since the

desired product is acid-insoluble, in the following runs the product
layer was acid washed to remove any of the polymers that might be
formed.

In the third preparation, 3h.h.g. (0.60 mole) propionitrile,
15.0 g. (0.625 mole) sodium hydride, and 120 g. (0.76 mole) bromo-
bensene in SO cc. toluene were permitted to react. The sodium hydride
was added during 15 minutes at room temperature and the mixture stirred
for one-half hour. No noticeable reaction took place. No reaction was
observed as the mixture was heated to 62° during one and one-half hours.
The addition of a few drops of alcohol had no effect. When the mixture
was heated to 70°, reaction occurred and the temperature rose rapidly
to 90°. It became very sticky and the agitator set up. The mixture
was heated to 118°, then cooled and alcohol and water were added.
Considerable reaction took place on addition of the alcohol. The layers
were separated and analysis of the aqueous layer showed 9.2% bromide
conversion. The oil layer was washed once with water, twice with acid,
and again with water. After removal of the excess bromobenzene there
remained about h or S g. of high boiling material which could not be
distilled at reduced pressure without decomposition. This material was
discarded.

In the final experiment in this series, an attempt was made to
avoid a high concentration of nitrile at any time. It was hoped that
in a highly diluted state, with a high concentration of bromobensene

present, the nitrile would be phenylated rather than undergoing

10

polymerization. The sodium hydride (lS.h g., 0.6h3 mole) and bromo-
benzene (188 g., 1.2 moles) were loaded into the reaction flask, and
heated to 90-1000. A solution of 33 g. (0.60 mole) propionitrile in
100 cc. toluene was added over a period of three hours. The mixture
became very sticky as previously observed. It was heated an additional
hour at 120-1250 and then cooled. On addition of alcohol considerable
reaction took place. This may be a decomposition of the sodio deriva-
tive of the polymer. The water layer showed 39% bromide conversion.
The oil layer was acid washed as before. This time the acid wash
separated from the oil layer was made alkaline and about 20 g. of a
sticky oily material separated from it. This amount of polymer cor-
re3ponds to about 60% of the starting nitrile. After removal of the
toluene and bromobenzene from the oil layer, very little material was
left. When it was attempted to distill this material at 2 mm. pressure
it began to decompose and some material froze in the reflux head. The

distillation was discontinued and the material was discarded.

B. Arylation of Prepionitrile in the Presence of Potassium Amide

This reaction was carried out according to the procedure of
Bergstrom and Agostinho (l). The potassium amide was prepared by the
addition of metallic potassium to liquid ammonia in the presence of
ferric oxide catalyst. This was similar to the preparation of sodamide

in liquid ammonia, and in addition all the precautions for the handling

of metallic potassium were observed.

11

Potassium metal (7.8 g., 0.20 mole) was converted to the amide in
about LOO cc. liquid ammonia in a 1-1. three-necked reaction flask.
Another portion of potassium amide was made in a 500 cc. flask from
15.6 g. (0.h0 mole) potassium in about 200 cc. liquid ammonia. To the
former was added 11 g. (0.20 mole) propionitrile, followed by hS g.
(0.h0 mole) chlorobenzene during a few'minutes time. Very little, if
any, visible reaction occurred. The second portion of amide was
pressured slowly over into the reaction flask during a period of 20
minutes. A violent reaction ensued and the mixture turned dark brown.
After LS minutes of stirring, 20 g. ammonium bromide was added to
neutralize any remaining potassium amide. Then 150 cc. water was added
very slowly. The mixture was acidified with dilute hydrochloric acid.
One hundred cubic centimeters toluene was added and the layers were
separated. The oil layer was washed four times with water and frac-
tiohated at 20 mm. pressure. Two fractions were collected. The first
'(7.3 g.) distilled at ll6-lh0o at 20 mm. pressure (not constant any-
where), Ngs 1.512h. If calculated as monoarylated product this
correSponds to 27.7% yield. The second fraction (13.9 g.) distilled
at 1100 at 20 mm. pressure to 1620 at 3 mm. pressure (mostly at 157°
at 3 mm.), Ngs 1.5700. If calculated as the diarylated product this
correSponds to 33.6% yield. Bergstrom and Agostinho (1) report only
that they obtained b3% yield of the monoarylated product (b.p. 10S-112°
at 8 mm.) and some higher boiling material. werkman (2) gives for

hydratroponitrile a boiling point of 82° at 3 mm. pressure, Ngc’l.5120.

12

C. Attempted irylation of Phenylacetonitrile in Presence of Sodium
Hydride

Eighty grams (0.683 mole) phenylacetonitrile and 150 g. (0.95
mole) bromobenzene were loaded into the reaction flask. To this mix-
ture was added l6.h g. (0.683 mole) sodium hydride in 100 cc. toluene
over a period of two hours while the temperature was maintained at
90-1100. The batch was very thick so 50 cc. more toluene was added.
After the mixture had been stirred at 9551050 for 22 hours it was
quite thin and reddish colored. It was treated with alcohol and then
with water (ammonia evolution). It was filtered and the layers were
separated. The water layer showed 22.7%‘hromide conversion.

The oil layer was washed twice with water, twice with acid, and
again twice with water. After the bromobenzene had been removed there
remained about 1h g. material, which could not be distilled at 2 mm.

pressure. It was discarded.

13

II. ALKYLATICN OF PROPIONITRILE IN PRESENCE OF SODIUM HYDRIDE

To a.mixture of 36.6 g. (0.666 mole) prepionitrile and 100 g.
(0.73 mole) n-butyl bromide at 30°, was added 16 g. (0.666 mole)
sodium hydride in 100 cc. toluene during one-half hour. No noticeable
reaction occurred. The mixture was gradually heated to 500 and held
at this temperature for 21 hours. At the end of this time it still was
a thin gray slurry much like it was at the start. Over a period of six
hours it was heated gradually to 1120 and held there for two hours.

The only evidences of reaction were that the reflux temperature had
increased from 980 to 1120 and that the gray color had changed to

almost white. The mixture was treated with alcohol (reaction), followed
with water. After filtration and separation the water layer showed
6h.3% bromide conversion.

The oil layer was acid washed and then water washed. The acid
wash liberated about 18 g. oil when it was made alkaline. This is
evidence of considerable polymerization of the nitrile. After the
toluene and butyl bromide had been removed the oil layer was fraction-
ated and two fractions were collected. The first (17.5 g.) distilled
at 70-930 (mostly 78-800) at 30 mm. pressure, N35 l.h070, D25 0.7985.
This corresponds to a 23.6% yield of.2-methylhexane nitrile. The second
fraction (2.5 g.) distilled at loo-126°, N55 l.t255. 2-Methylhexane-
nitrile has a reported boiling point of b3-SOO at 9 mm., D25 0.797 (20).

One gram of the suspected 2-methylhexanenitrile (first fraction

above) was converted to the corresponding amide by heating with S g. of

1h

75% sulfuric acid on the steam bath for 20 minutes. It was poured
into ice water; the crystals which formed melted at (lb-670.
Recrystallization from benzene raised the m.p. to 68.5-69.00. The
amide of 2-methy1hexanenitrile is reported to melt at 660 (20) and at

70-72.S° (21).

15

III. THE ALKYLATION 0F 2-PHENYLBUTANSNITRILE WITH THE
BUTYL CHLORIDES USING SODAMIDE
The same procedure was used for all the alkylations in this series.
All alkylations were run in duplicate. The reactants were used in the
ratio of one mole sodamide, 0.81 mole nitrile, and 0.9h mole butyl
chloride. Approximately O.h mole sodamide was used in each preparation.
The 2-pheny1butanenitrile, the butyl chloride, and 50 cc. toluene

were charged into the 500 cc. reaction flask. The sodamide in another

 

50 cc. toluene was added from the dropping funnel at a temperature
sufficiently high to make the reaction proceed readily and at a rate
which permitted good control of the temperature. The time of addition
was from one-half to one hour and could have been shortened consider-
ably by use of more cooling. The temperature of addition was 65-800
in the cases of the n-butyl, s-butyl, and i-butyl isomers. For the
t-butyl isomer it was 55-650. The batch was then heated to reflux
temperature which was 118-1200 for all except the t-butyl isomer, for
which it was 112-1160. As will be seen later, in this latter case
the starting nitrile reacted to only a small degree, and, since it has
a lower boiling point than the expected product, it is reasonable that
it should raise the reflux temperature of the toluene a lesser amount.
One further variation was instituted in the case of the t-butyl
chloride runs. It was expected that isobutylene might likely be formed

as a by-product. The tOp of the reflux condenser was connected to a

 

bubbler trap containing 15% sulfuric acid and the outlet from the

trap was connected to a dry ice trap. In both runs a quantity of
material was collected in the dry ice trap, which was sufficient to
account for about 70% yield if it were all isobutylene. The material
was readily flammable. Most of it boiled between -lo° and +100, and
the remainder over the range up to LOO. It appeared to be mostly iso-
butylene, contaminated with a little tebutyl chloride.

After the reaction was completed it was cooled and water and a

:m Vii-Ta-LAP. 71
7%

little filtercel were added. After filtration the layers were

separated and the water layer was analyzed for chloride ion. The oil

 

layer was washed further, in some cases only with water, in others with
salt water, and in still others with acid followed by water. The pur-
pose of the acid wash was to remove polymeric nitriles. Apparently in
this series none of these were formed, as no difference was noted in
the amount of distillation residue as a result of the acid wash. Salt
water in place of plain water served only to make the layers separate
faster. It was not necessary. _

After distillation of the toluene the residual product was frac-
tionated thru the Vigreux or the helices-packed column at 5 mm. pressure.
Fractions were taken and the refractive index observed for each and
recorded as shown in Table I for 2-ethyl-3-methyl-Z-phenylbutanenitrile.
The boiling points and refractive indices followed the same pattern of
variation quite closely for all the successful runs. In all the runs

a negligible amount of distillation residue remained. In the two runs

with t-butyl chloride about 80% of the starting nitrile was recovered.

17

The high boiling fractions from the two runs were combined and re-
fractionated. The boiling range was narrowed down to about 6°, how-
ever, a constant boiling fraction was not obtained. That fraction
which had the least variation in boiling point was used for determin-
ation of physical constants. Three of the fractions solidified. These
were combined, pressed out, and recrystallized from ligroin and again
from methanol, m.p. 57.2-58.00.

The yields, chloride conversions, physical constants, and nitrogen
contents are summarized in Table II. The nitrogen analyses and all
other chemical analyses of the prepared compounds reported herein were
done by the HicroJTech Laboratories, Skokie, Illinois.

In a typical run, h8.1 g. (0.332 mole) 2-phenylbutanenitrile,

35.6 g. (0.386 mole) s-butyl chloride, and 50 cc. toluene were charged '
into the reaction flask. A suspension of 15.9 g. (0.h08 mole).sodamide
in 50 cc. toluene was placed in the drOpping funnel. The flask was
heated and the amide added at such a rate that the temperature was
maintained at 70-850. After the addition of the amide, the mixture was
heated at reflux temperature (118-1200) for one hour. It was cooled
and 100 cc. water and a little filtercel were added. After filtration
and separation of the layers, the water layer was analyzed and found

to contain 0.332 mole cthride ion, which correSponds to 100% chloride
conversion, based on the nitrile used.- The oil layer was washed once
with salt water, twice with acid, and once more with salt water. After
most of the toluene had been distilled at atmOSpheric pressure, the

remaining product was fractionated at 5 mm. pressure. The distillation

18

data and the refractive indices of the fractions taken are shown in

Table I.

TABLE I

DISTIILATION OF 2-mPHL-3-MhTmp2-PHHVYLPENT1NENITRILE

 

 

 

' Fraction B.p., oC. weight, N25
(5 mm.) Grams D
1 82-105 0.8 1.5005
2 105—127 1.6 1.5039
3 127-127.S 8.8 1.5038
h 127.5 9.9 1.5037
5 127.5 1b.5 1.5039
6 127.5 10.0 1.5037
7 127.5-128 12.0 1.5037
8 28-132 5.b 1.50ho

 

The residue and column holdup amounted to about one gram. The density
of fraction five was determined at 250 with a pycnometer and found to
be 0.9h66. The total weight of all the cuts, except the first, was
62.2 g. This corresponds to a yield of 93.3% based on the starting

nitrile.

19

IV. THE ALKYLITION 0F 2-PHENYLBUT1NENITRILE WITH
THE BUTYL CHLORIDES USING SODIUM HYDRIDE

In this series the mole ratios of the reagents were the same as
were used in the sodamide series, namely, one mole sodium hydride,
0.9L mole butyl chloride, and 0.81 mole nitrile. The batch size was
considerably smaller, 0.186 mole sodium hydride. .The procedure used
was substantially the same as that described in Section III. It was
found that the sodium hydride could be added practically all at once
at any temperature below the initial reflux temperature of 105°. In
about one hour the reflux temperature would increase to 113° and remain
there while it was stirred for an additional hour. After the reaction
was complete a little alcohol was added to get rid of any sodium
hydride before the addition of the water. Only the n-butyl, sebutyl
and iébutyl derivatives were prepared and only one run was made of
each. The data is summarized in Table II. The distillation residues
again were negligible. Sufficient low boiling material (mainly the

starting nitrile) was recovered to account for about 10-15% of the
yield. In the case of the i-butyl isomer about h% of high boiling

material was obtained, which was not further investigated.

 

20

.owmn~.~m as vmpflms vHHom one .mnnh mnavmooha 03p
map Eopm unoHpthm mafiafion zmfiz nocaneoo map Mo :ofipmzowpompmou one no woman u
.maapuwcmcméanKszQuw omhm>ooom o
.mcmzpo mag ca vmms mm: mwfiadoow .pCmmm mcfimcovcoo nu coma was mpfihfih: suavom n
.ea.e .z «zeamvao tee eeeefleeaeo a

 

 

 

 

om.o Noam.a Hemm.o omaumma 5.5 .mo eampemup
eo.om m.ee Hapsmup
00.05 N.me Hausmwp
eem4.H o.wfla m.me o.mm Haeemna
eea4.H emmm.o o.mHH H.Hm m.ew paeeem.a
mo.e hem4.H mmme.o o.oNH e.em o.Hm Hanna-“
mmOm.H o.mma n.0a o.~m easesmre
emom.H eeem.o o.m~H m.om m.mm daesmnn
Hm.o wmom.H eese.o m.ema m.ma o.ooa Haeemsm
mems.fl o.ama o.me o.~e namesm-e
Hams.fi mamm.o o.mma o.ew m.mm Hapemue
00.0 Hee4.H ommm.o m.e~H m.em o.ooa easem-e
a . A.sa mv m u :oampm>:oo
am.z mmz mma .oo..a.m .efieaw eeatoaao m
zonmokmohmo

Q

mmfifiHzmz...e§zéa-m-EB.~ a: mom 224.860 .2333 maumqmfi

HH mqmds

21

V. PhaPihnTION 0F DnhIViTIVnS
A. Attempted Hydrolysis of Nitriles

Reports in the literature state quite clearly that the hydrolysis
of highly sterically hindered nitriles is very difficult (2,13,1h,15,15).
In the extreme cases the nitrile is converted to the amide or the
sodium salt of the acid by treatment with alkali in a high boiling sol-
vent such as glycerine or diethylene glycol, or to the amide by treatment
with sulfuric acid. The amide may be converted to the acid by treatment
with butyl nitrite and sulfuric acid.

Both alkaline and acid hydrolyses of these nitriles were attempted.
The product made from the t-butyl chloride was available in only a very
small amount, and no preparation of a derivative of it was attempted.
In Tables III and IV are listed the attempts at hydrolysis by alkali
and by acid, reSpectively. In the alkaline hydrolyses no reaction could
be observed, except, in a few cases, for the evolution of a trace of
ammonia. In all cases an insoluble oil was recovered when the reaction
mixture was poured into water. The water soluble material which was
obtained in the 90% sulfuric acid hydrolyses was assumed to be sulfonated

product.

 

 

 

22

.o>wpstc .anSmom

 

.mcapnfipm aeofindzomz onwnmma o.H Aeomsmv mom mow ampsmuc
moan
m>flpmmmz cam on m.o .Hooham osmHhQOhawpa thsmu:
mom . oqugfiw
m>fipammz cam on m.o .Hoohflm mamahaounwpe ampsmu:
mom
o>figwmmz cam on m.o .Hoohaw mamahaopafiha ahasmu:
esaeamez cam on m.o moan .eeaneeaflc Haeemne
moan .
eenpamez oma m.o Hoeafim eeeaausm Hansm-e
mom
eeneemez OHN-OON m.o .Honoeaa Hansen Haeem.e
mow
.pwzpo athmsocoa
m>fipmmmz oamuoma m.o Hoomam mcmdmaopafipe thsmlc
Hanan-“
.m>aammm: .wcwxmnm mom HApsmIn
esteemed .eanaenae so: oam-oma o.nH .Hoeaaw eeeaanpm Hapsm.e
.m>wpsmo: .mpHfimmm mom
.aeaaanfionae eeoaaeoo 00H 0.: .Hoeaam eceflhesm Haeem.e
nxhdfimh . 00 0.3.90: 1
can mpaaumm .ecspmhmasme .mswa nucmmsmm m

 

 

mmAHmHHzm24asmAszmmlwqueDm1w may mo mHmHuomnwm MZHAdmud nmamzmth

HHH aqmwh

23

 

 

 

w‘ ---«s.::~..1;+;
megapmmoz ow o.m 0:833 THE .930 Hawzmnc
measemen .eepecomaem - ow m~.o . need eatsuadm «mm Hausm.n
e>eeamez oefiuoea m.o pane .eaoa eateaadm amp Hagen]:
e>aeamee .eeeaeeeasm on mN.H ease eateeaem mmm Heeem.n
eeaeamez omaaona m.o ease caeeefidm amp Hagen-»
defined: .83 Eden 0.2-03 od H8 .88 deem.“
e>apdmen .edeeeoaaem cw m.o ease handmasm Rmm Haesm-a
m.o Aeav defines: season
neapdmez omanooa o.H .eaod sandmadm mom Haesmne
mxpweoh .00 926;
psi mpflsnom . .mpdpspoasma .osae mpcmmaom m

 

 

.

mmAHéafisaafimmnméamlm BE. .8 MHnHAOMQHm QHUd 5%.:

PH ”39.9

2h

B. Reduction of Nitriles and Preparation of Derivatives
fibrk going on in this laboratory on another project showed that
nitriles similar to these could be readily reduced to amines by use
of lithium aluminum hydride and that the amines could be converted into
substituted thioureas with isothiocyanates. The work was now directed
towards use of this method for characterization of the new nitriles.
The reduction was carried out according to the procedure of
Amundsen and Nelson (17). Four grams lithium aluminum hydride and
100 cc. dry ether were loaded into the reaction flask (the same appara-

tus as was used for the alkylations). about 10 g. nitrile in another

 

100 cc. dry ether was added during a fewrminutes at room temperature.
After refluxing for about hS minutes the mixture was cooled and h cc.
water was added drapwise. This was followed by 3 cc. 20% sodium
hydroxide and 1h co. more water. Apparently these quantities were so
chosen that the inorganic phase would come out solid and the separation
of gelatinous aluminum hydroxide is avoided. The ether solution was
filtered from the solid and the latter washed twice with fresh ether.
After evaporation of the ether the oil was distilled at 5 mm. pressure
without a fractionating column. The amines were practically water-
white before distillation. The weights of the fractions taken and their
boiling points are listed in Table V.
The second cuts were used for determination of the physical con-
stants and the preparation of derivatives. The substituted urea and
thioureas were prepared by mixing equal parts of the amine and
O<-naphthylisocyanate or phenylisothiocyanate according to the usual

procedure. Reaction took place readily with evolution of heat.

25

TABLE V

DISTILLATION DATA FOR THE l-AMINO-Z-BUTYL-Z-PMYIBUTANES

 

 

 

Fraction n-Butyl s-Butyl i-Butyl
1 133-13h° 126-129.5° 115.5-119.S°
1.3 g. 1.h g. 1.7 g.
2 13h° 129.S° 120.5-121°
S.b g. 5.3 g. 5.2 8.
Residue and 0.8 g. 1.3 g. 1.5 g.
holdup

 

The mixtures were warmed a little and then permitted to cool and
crystallize. The solids were recrystallized from methanol. Phenyl-
isothiocyanate was used for the n-butyl and the i-butyl isomers.
However, it would not give a solid product with the s-butyl isomer so

CK-napthylisocyanats was used to prepare a derivative. The properties

of these amines are listed in Table VI.

26

 

.pmpoohpoo one nvqfion mcapamz
U
.wn.e .z. nonzenzeno nee eoeeaeofiao .eensaanenaez.ee o

.Nn.m .m am~.m .z .mnzenmano new eeeaaeeaao n
.Nm.e .z ”zenmvao to» eeeeadeano a

 

 

. sea

on.a no.m -m.osa no.e omHm.H meme.o m.o~a Haeem-n
. spa
eam.e o-defl No.0 naem.a Hesm.o m.ema Hansm-n
. .mHH .

mm.e so.w -m.mHH am e mmem.fi mamm.o o.an Haeem-e

a .m a .z .o ..e.: A.ee my

D D D o a Q . ta.
nmmpmoazpahamzm as 2 mm: mmn co m m

 

 

«azuumvmohm ohms

©

mm2<eDmghzwmmumlqweDmINIOZHZdIH axe m0 mMHBmmmomm

H> mamaa

27

DISCUSSION

The literature indicates quite definitely that the simple nitriles
(9,10,11,12) polymerize easily with the alkali metals and with strong
bases. The experiments with propionitrile and sodium hydride show
that this polymerization is readily effected by sodium hydride also,
and in fact, to such a degree that alkylation and arylation are
practically impossible.

Sodamide apparently is such a mild reagent that it does not cause
polymerization, but still is active enough to effect condensation with
alkyl halides. However, Ziegler and Ohlinger (h) claimed that they
did get some polymerization of propionitrile when they tried to alkylate
it with sodamide. Bergstrom and Agostinho (1) did their work with
chlorobenzene, potassium amide, and prOpionitrile. in liquid ammonia,
and in addition thet used a second very large portion of potassium amide.
They offered very little explanation of why this was necessary. It was
noticed in this investigation that the addition of this second portion
of amide caused a violent reaction while the addition of the first amide
caused little noticeable reaction. The work of Ziegler and Ohlinger
was done above room temperature.

The experiment with n-butyl bromide, propionitrile, and sodium
hydride showed that with a more active halide some alkylation could be
accomplished. This observation is in general agreement with earlier
work in this laboratory which showed that ring substituted phenyl-

acetonitriles could be alkylated with butyl and amyl bromides and with

benzyl clfloride using sodium hydride. In this latter case the nitrile
being rather highly substituted would be less sensitive to polymeri-
zation.

The preparation from prepionitrile, chlorobenzene, and potassium
amide checked the results of Bergstrom and igostinho (1) only approxi-
mately. They reported a fair yield of monOphenylated product and did -.I
not isolate the diphenylated material. In the present work the lower rflfli‘
boiling material was obtained in only a small quantity and boiled over

such a wide range that it was not further investigated. However, the

 

higher boiling fraction boiled at least reasonably constant and was lh_ w
probably'the diphenylated product. However, this point was not proven.
Only one eXperiment was run here and judging from the published informa—
tion the above authors also ran only one experiment, so the information
must be considered as being very meager.

The results obtained in the alkylation of 2-phenylbutanenitrile
were in good agreement with past experience with nitriles of similar
structure. The 2-phenylbutanenitrile is a quite highly substituted com-
pound and would not be eXpected to be very susceptible to polymerization.
The aliphatic halides are at least moderately active and thereby
alkylation should be favored. It is interesting that the n-butyl
chloride gives a lower yield of alkylated product than do the s-butyl
and the i-butyl chlorides. This observation is in agreement with some
present work in this laboratory with the isomeric amyl halides and also

with the findings of werkman (2) in his work on the alkylation of

hydratrOponitrile with butyl halides and sodamide. He obtained about

 

a 60% yield with the n-butyl isomers and about 70% yields with the
s-butyl and i-butyl isomers. On the other hand Tilford and co-workers
(22) alkylated cyclohexyl cyanide with amyl and.butyl bromides using
sodamide as a condensing agent, and found that in both cases the
n-alkyl bromide gave about 20% more yield than the i-alkyl bromide.
The fact that t-butyl chloride gave almost no yield of alkylated product
was in agreement with the results of both Tilford (22) and werkman (2).
It should be mentioned that with the other factors favorable, sodium
hydride can be used as a condensing agent with fair results. The use
of this condensing agent was investigated to only a minimum degree and
it is very probable that the yields by this method could be increased
with further study of the variables.

It might be worthy of mention here that the attraction for sodium

it could be purchased and stored safely and

hydride was the fact that/the preparation of sodamide could be omitted
from the eXperimental work. This objection to sodamide may no longer
be serious. The material used in this work was purchased in the dry
state, kept in storage for several years, and was found to be in very
good condition. Older experience as reported is that the material
often deteriorated in storage and became hazardous due to the formation
of sodium nitrite due to air oxidation in the presence of moisture.

In connection with the difficulty encountered in the attempts to

hydrolyze these nitriles it is interesting to examine them from the

standpoint of the six-number concept advanced by Newman (23). He made

 

a study of the rates of esterification of a number of acids having vary-

ing degrees of steric hindrance and found that the rate decreases as

30

the six-number increases. His six-number is defined as the number of

 

 

atoms within the molecule which occupy a six position if the carbonyl
oxygen is counted as one. In Figure I are shown the configurations

of the nitriles made by werkman (2) (I through V) and of those made in
this investigation by alkylation of 2-phenylbutanenitrile (VI through
IX). Each sixth atom from the nitrogen atom is starred (*). The

six-numbers are shown for each nitrile in parenthesis. Nitrile I

 

could be hydrolyzed readily by heating it with sulfuric acid (75%) for
fifteen minutes on the steam bath. II and III could not be hydrolyzed
by any of the more moderate methods. However, by heating with potassium
hydroxide and amyl alcohol at reflux for ten hours, hydrolysis was
finally accomplished. These two nitriles have a six-number of 5. IV and

V with six-numbegg of 8 and 11 reapectively could not be hydrolysed in

 

any manner by Werkman (2). The experience with the nitriles VI through
Ix, which could not be hydrolyzed, has already been described in an

earlier Section of this thesis. From the standpoint of their six-numbers

 

this non-reactivity is just what would be expected.
For some reason the steric factors which prevent hydrolysis do not

interfere with the reduction to amines.

31
FIGJRE I

 

Six-numbers of Sterically Hindered Nitriles

 

 

I (2) VI (8)
H’QH“ 3* 3*
HZ-CN ‘
0H3 H30-0*H3-CH*3-0H3- —CN ,
II (5) VII (8) g
H H g
0H,-0""H,-0H§.0H3.c.cn 5
CH3 H301. gin—0H .c-CN If;
H3 0113-0 E:
III (5) VIII (ll)
‘15 9.1-
H H H“ H*
0 ”Ha-OH”£H2-C-CN 0 H 3-CH:-C--O-CN
0H3 tn; CH3 CHg—CH:
IV (8) IX (It)
a v Id
If H H50 H
0* Tia-CH: .0H-4-0N 011:4??4" *
v (11)
H’”\ fill}?
he, {

32

SUMMARY

1. Attempts to arylate propionitrile with bromobenzene using sodium
hydride as a condensing agent were unsuccessful. Much polymerisa-

tion of the nitrile occurred.

2. Propionitrile was arylated in approximately 60% yield by the use of
potassium amide according to the Special procedure of Bergstrom and

Agostinho (l).

3. One attempt to alkylate propionitrile with n-butyl bromide and sodium

hydride resulted in 23.6$ yield of monoalkylated product.

h. One attempt to arylate phenylacetonitrile with bromobenzene and

sodium hydride was unsuccessful.

S. Alkylation of 2-pheny1butanenitrile with the four isomeric butyl
chlorides and sodamide was successful with the nébutyl, s-butyl and
i-butyl chlorides, but unsuccessful in the case of the t-butyl
chloride. In the latter case about 8% yield was obtained. The
n-butyl isomer gave about an 88% yield, while the sAbutyl and the

i-butyl isomers gave yields of about 90-91%.

6. It was found that 2-phenylbutanenitrile could be alkylated with butyl
chloride when sodium hydride was used as a condensing agent. The
yields from the n-butyl, s-butyl, and i-butyl chlorides were about
12% lower than the yields when sodamideewas used and were in the same

order. No run was made with tAbutyl chloride in this series.

 

33

7. Many attempts were made to hydrolyze the 2-butyl-2-phenyl-
butanenitriles for the purpose of characterization. These were all
unsuccessful. It was found that these nitriles could be reduced
quite easily with lithium aluminum hydride to the corresponding
amines. Derivatives were prepared from the amines. Physical con-

stants of the nitriles and the amines were determined.

8. The resistance to hydrolysis of these nitriles and of the 2-butyl-
2-phenylpr0panenitriles prepared by werkman (2) was shown to corre-

late with the six-number concept advanced by Newman (23).

 

10

314

SUGGESTIONS FOR FURTHER WORK

The reaction of propionitrile with chlorobenzene in liquid ammonia
with potassium amide as a condensing agent would be an interesting
subject for further study. It is the only example, as carried out
by Bergstrom and Agostinho (l), of a successful arylation of a
nitrile to be found in the literature. Investigations could be

made to see how generally applicable this technique is, to try to

get more information regarding the intrinsic nature of the reaction
and why the second addition of amide is needed, to see if other media
besides liquid ammonia can be used, and perhaps to determine whether

sodamide will work equally well, since there are definite advantages

in working with sodamide rather than with potassium amide.

In the experiments with sodium hydride, bromobenzene, and propio-
nitrile a substantial degree of bromide conversion was obtained.

It would be interesting to find out what happens to the bromobensene
and what conditions are required to make it happen. Information such
as this would very likely provide some clues as to how to vary the
conditions of the alkylation in such a manner as to improve the

yields.

Obviously the alkylation of 2-phenylbutanenitrile with t-butyl
chloride could stand some further study. Variations in temperature
of reaction, working in a different medium, e.g. liquid ammonia, and
the use of other condensing agents would provide subjects for

firfiwrshfly.

 

10.
ll.
12.
13.

l7.

‘INY

SELECTED RQFERENCES
w. Bergstrom and R. Agostinho, J. Am. Chem. Soc., 61, 2152
(19b5).
R. Workman, M. S. Thesis, Michigan State College, 1950;
G. L. Goerner and W. h. Workman, J. Org. Chem., 19, 37
(l95t).

Ramart, Compt. rend., lfg, 1226 (1926).

Ziegler and H. Ohlinger, Ann., £25, 8h (1932).

Ziegler, b. 5. 1,958,653; C. A., 32, LLBS (193L).
Schuerch and E. Huntress, J. Am. Chem. Soc., 19, 282L (19h5)

. Cristol, J. W. Ragsdale, and J. S. Meek, J. Am. Chem. Soc.,

1;, 19c3 (19t9).

. J. Cram and J. Allinger, J. Am. Chem. Soc., 16, h516 (195h).

Frankland and H. Kolbe, Ann., gg, 269 (18h8).

. Bayer, Ber., a, 319 (1869).

Meyer, J. prakt. Chem., g3, 262 (1880).

. Meyer, J. prakt. Chem., 3§, 336 (1886).

. A. Larsen, A. W. Ruddy, B. Elpern, and M. MacMullin,

J. Am. Chem. Soc., 1;, 532 (19h9).
Zaugg and B. Horrom, J. Am. Chem. Soc., 12, 3006 (1950).

Sperber D. Papa, and F. Schwenck, J. Am. Chem. Soc., 12
3093 (19h8).

9

J. Hickinbottom, "Reactions of Organic Compounds", Longmans,
Green.& Co., New York, 19h8, 2nd Ed., p. 282.

Amundsen and L. Nelson, J. Am. Chem. Soc., 13, 2L2 (1951).

35

 

 

18.

19.

20.
21.

22.

23.

R. L. Shriner and R. C. Fuson, ”Identification of Organic
Compounds", John Wiley and Sons, Inc., New York, 19h0,
2nd Ed., p. 171,

V. Migrdichian, ”The Chemistry of Organic Cyanogen Compounds",
Reinhold Publishing Co., New York, 19L7, p. 3L9,

P. Levene and L. Mikeska, J. Biol. Chem., EEJ 582 (1929).
P. Rasetti, Bull. soc. chim. France, [3], 23, 690 (1905).

C. H. Tilford, L. A. Doerle, M. G. VanCampen, and R. S. Shelton, F—-qh
J- Am. Chem. 500., I}, 1705 (19b9).

M. s, Newman, J. Am. Chem. 800., 1g, L783 (1950).

 

VERSITY LIBRARIES

015 4743

MICHI AN STAT

Hfu MI in

31293

“W
03