Full text of "Organic seminar abstracts"

See other formats

'LIBRA RY

OF THE

U N I VERS ITY

Of 1LLI NOIS

G.547

U<6s
l94l/42

REFERENCE

Return this book on or before the
Latest Date stamped below.

University of Illinois Library

Ll<. 1—1141

Digitized by the Internet Archive

in 2012 with funding from

University of Illinois Urbana-Champaign

http://archive.org/details/organicsemi4142univ

SEMINAR REPORTS

I Semester 1941-42

Pae;e
Alkaloids of Crotalaria, Senecio, Heliotropium Trichodesma and Erechtites 1

Roger Adams

Progesterone and Related ConiDOunds 6

R. L. Frank

Koelsch's Work on Cyclic Diketones 12

Stanley Wawzonek

The Rearrangement of Allyl G-roups in Three-Carbon Systems 18

Clay Weaver

Syntheses of Aldehydes through Grignard Reagents 23

R. B. Carlin

Synthetic Estrogenic Compounds 27

C. F. Jelinek

Derivatives of ^%thallyl Chloride 33

R. S. Voris

Pinacol Rearrangements 38

J. 0. Corner

New Syntheses of ^exatrienes and Squalene 44

P. F. ^ar field

Attempts to Determine the Structure of Phthioic Acid 48

R. G. Chase

The Optical Isomers of Cis-9-Methyl-l-Dec alone 52

J. D. Garber

Structures of Pyrethrins I and II 56

D. ff. He in

The Michael Condensation: Some Recent Investigations 60

R. E. Foster

Reaction of Alkyl Benzoates with Sodium Alkoxides 64

G. L. Schertz

The Structure of Lignin 68

B. C. McKusick

Syntheses in the Triphenylene Series 72

P. L. Southwick

iq/| /a. 7

Sigh Octane Aviation Fuels

Q. F. Soper

Page
78

The Mechanism for the Coupling of Diazonium Salts with Aromatic 82
Amines and Phenols

W. E. Blackburn

A Review of the Organic Chemistry of Arsenic 90

C. W. Theobald

Addition Products between Ketenes and Unsaturated Hydrocarbons 99

W. H. Kaplan

Solvolytic Reaction Mechanism 104

0. D. Jones

Aromatization of Aliphatic HydrocarDons 108

S. P. Rowland

Thiazoles: Some Syntheses and Reactions 112

G. W. Cannon

High Pressure Hydrogenations over Nickel and Copper Chromite 118

S. M. Himel and R. C, Gunther

The Chemistry of Organoboron Oomp0unds 130

R. M. Roberts

Cardiac Aglycones of the Strophanthidin Group 135

R. F. Phillips

Preparation of Nitriles by the se of Cuprous Cyanide 139

A, V. Mcintosh, Jr.

The ^eduction of Multiple Carbon-Carbon Bonds 142

J. C, ^obinson, Jr.

Polyenes and Cumulenes 145

S. B. Speck

Reactions By Pyridine 149

F. J. ffolf

ALKALOIDS OF CROTALARIA, SENECIO,
HELIOTROPIUM TRICHODESMA, AND ERECHTITES

Crotalaria is a genus of leguminous plants, many species of
which are commonly used in the southern part of the United States
as soil-enriching legumes. All contain smaller or larger quantities
of alkaloids. At present Crotalaria spectabilis is the most
widely planted. Its alkaloid, extracted by alcohol, has the for-
mula C16Hs306N. Through its chemical reaction it has teen shown
to be closely related to alkaloids extracted from various species
of Senecio, Heliotropium, Trichodesma, and Erechtites. In view of
the fact that Senecio alone has over 1200 species, the alkaloids of
this general type promise to be the largest single class known. Of
the 20-30 thus far characterized, not a single one has had its
structure clarified. These alkaloids are esters, made up of an acid
containing eight or ten carbons (typical formulas are C8H1205,
C8H1004, C10H14t04, C10H1606) and a bicyclic nitrogen base con-
taining two hydroxyls, at least one of which is esterified with a
carboxyl of the acid. All the bases appear to be derivatives of
methyl pyrrolizidine.

This report will be confined to the experiments on mono-
crotaline, the alkaloid of Crotalaria spectabilis. The two most
important reactions which cleave the molecule into parts more
suitable for structural study are (l) hydrolysis with alkali;
(2) hydrogenolysis with platinum of nickel catalyst and hydrogen.

Monocrot aline

2Ha C16H33p6N AgNaOH

(CTH1i03)C00H + C8H150N (CgHnOjCOOH + C8H1302N

monocrotalic acid retronecanol monocrotic retronecine

acid

I NaOH ^

Monocrotalic acid (optically active) loses carbon dioxide upon
treatment with alkali to give monocrotic acid (i). Its properties
follow: (l) optically inactive; (2) monomethyl ester with diazome thane;
(3) forms a dinitrophenylhydrazone; (4) gives iodoform with iodine
and alkali; (5) oxidizes with sodium hypobromite to a mixture of
meso and racemic a,a'-dimethylsuccinic acids; (6) on heating loses
water to give an unsaturated neutral substance reconverted to the
acid by hydrolysis. Monocrotic acid thus appears to be a, ^-dimethyl
levulinic acid which was established by synthesis.

JuAtv.

2 -

CH3-CH-COCH3

I
CH3-CH-COOCH3

Monocrotic Acid

CK3-CH-COCH3

I
CH3-CH-COOH

CK3-C=C-CH3

CK3-CH-CO
II

CH3-CH-COOH
CK3-CH-COOK

CH3-CH-CH-CH3

1 >

CH3-CH-CO

Konocrotalic acid is obviously closely related to Jimettoyl
levulinic acid. Monocrotalic acid was shown to have the x clicking
properties? (l) back titration from alkali indicates a earboxyl
ilT^llone linkage; (2) monomethyl ester with_ thiazomethane which
shows one active hydrogen; (5) heat decomposition gives a, P> Y-
"^methylangelicalactone (II ; (4) heat decomposition of methyl
monocrotalate gives an unsaturated ester by loss of a molecule of
?r the unsaturated ester can be reduced to a saturated ester
which'is hydrolyzed to a stable, crystalline ^tonic acid; the
unsaturated ester can be hydrolyzed with acia to the -actone oh
tained by heat decomposition of monocrotalic acid. These facte may
be explained by any of three formulas which decarboxylase in

OH
I
CH3-CH-C-CH3

xo
/

CH3-C— CO
CO OH

III

COOH/OK
CK3-C— C^-CHa
0

Cn3-CH-CH

CH,-CH-C-CHaC00H

CK3-CH-CO

-lkpline solution to the tautomeric form of dimethyl levulinic
acid and whose esters dehydrate to unsaturated lactone. They are
the lactone forms of ^-keto acids.

The observation that monocrotalic acid esterifies with great
difficulty with acid and methanol led to the elimination, perhaps
uncorrectly, of V. The necessity of high pressure reduction made
unlikely the unsaturated ester corresponding to IV. Orxgin^liy
III was accepted as the most likely. However, the results of tne
study"of compounds VI and VII led to different conclusions.

COOR
CH0-CH-COCH3 I

3 l 3 CH3-C-COCH3

CH3-C-C00R I

3 | CH3-CH-COQR

COOR

VII

■* ,~1

f. • - '-V • »

•;;•

Acid hydrolysis of compound VI resulted in formation of di-
methyl levulinic acid (I) which indicates this configuration is un-
stable to acid while monocrotalic acid is stable to acid. Alkali
saponification gave a dibasic acid and no decarboxylation took
place. Acid and alkaline hydrolysis of VII also gave dimethyl
levulinic acid. Compound V now seems the most likely formula but
no proof is yet available.

The basic part of the molecule obtained by saponification is
retronecine which has been shown by previous investigators to be a
di-hydroxy methyl pyrrolizidine containing one double bond. By
hydrogenolysis of retronecanol a oonohydroxy methyl pyrrolizidine
results*

Menshikov dehydrated retronecanol to the corresponding un-
saturated compound and reduced this product to the saturated base.
His proof that this substance, called heliotridane, C8H15N, is 1-
methyl pyrrolizidine was as follows: it had the following
properties — (l) a tertiary base and no N-alkyl groups, thus it
must be a bicyclic molecule with a nitrogen atomj common zo both
li.igSs Exhaustive methylation foil*
pyrollidine derivative shown by its
pyrrol. Subsequent reduction gave l_
dine, a synthesis for which by Menshikov failed. He prepared all of

smooth dehydrogenation to a
in optically inactive pyrolli—
' tiled. He prepared

except the desired one and drew his conclusions by elimination.
Thus the three possible methyl pyrrolizidines would yield an ex-
haustive methylation depending on the point of attack on the
molecule the following compounds.

ssible pyrrolidines which could be fori.ed theoretic

CH3

•CH2CH2Cri2CH3

VIII

•N"
CH3

■CH2CH2CR3

-CH-

"1 _ ,

L v— CH3CH(CH3 ) 3

in*

CH,

V ) — CH2CH2CH3
N

CH,

CH.,

CH3

CKCH2CH3

XII

XIII

_ 4 -

All the pyrrolidines except XIII were synthesized and the
solid picrates compared in m.p. with the picrate of the optically
inactive product obtained by degradation of heliotridane. None were
identical. Menshlkov degraded further the natural pyrrolidine base
and obtained an open chain compound which gave a picrate identical
with that obtained by a similar reaction with XII.

\/.,jr CH2CH3CH3

* CH3CH3-CH — CH~CH3CH2CH3
fea N(CH3)3

CHr

XIII

CH3

•CH3

■ CrICn3uR3

XII

This reaction establishes the structure except for one point. Con-
pound XII can exist in two stereochemical forms and Menshlkov may
have isolated the one not identical with the compound from the
natural oroduct.

In this laboratory compound XIII was synthesized and shewn to
give an identical picrate (m.p. 116°).

Oil 3
C6H5OCH2-CH2-C-CN

COOC^H

CK3

CcHcOCHpCHp— Cn— l»w

3ll-5

CH3
C6H5OCH2CH2-CH-COCH2CK2CH:

HBr

CK3
3rCH3CH2-CK-CH-C3H7

NHCH3

PtO.

GH3NH3

^6^5^'-'

■CH3

NaOH
*

CHj

C3H7

CH3
Gg n5OCri2 Cn 3 — C n— C nC 3 rl 7

NHCH3

Use of copper chromite, dry methyl amine and hydrogen gave a dia-
stereoiscmeric pyrrolidine picrate (m,p* 126°).

Placement of the hydroxyls and double bond has not yet been
completed. On the basis of the exhaustive methylation of retronec=
nol, Menshikov claimed he obtained a tertiary alcohol. Accepting
this as correct, retronecanol would have the structure

s/V

,-•

- 5

the

A ■

other

The following experimental facts were used to place
hydroxyl and double bond in retronecine. (l) One hydroxy! is
readily replaced by hydrogen; (2) after reducing the double bond
both hydroxyls are stable to reduction; (3) hydroxyl s in the
saturated molecule esterify at different rates; (4; no enol group;
(5) saturated dihydroxy compound readily forms an internal ether.
The only formula satisfactorily fulfilling these conditions is
shown in XIII, assuring retronecanol to be XIV.

CK3
— K^,

■w

CHa

"Tl

— CH3OH

XIII

XIV

Compound XV represents another structure on the assumption of
no tertiary hydroxyl. It satisfies all the other facts and also
the stability which might not be true of a vinyl amine type, (XIII )

Bibliography

Neal, Rusoff,

Ahmann , J .

Adams and

Rogers,

ibid.

Adams, Rogers and Sprules
Adams, Rogers and Long,
Adams and
Adams and

Long, ibid. . 62
Rogers, ibid.

Am. Chem. Soc. 57,
61, 2815 (1939).

ibid., 61, 2819 (1939),
ibid.. 61. 2822 (1939).

2560, (1935).

2289 U940).
63, 228 (1941).

Adams and Rogers, ibid., 63, 537 (1941).

Reported b} Roger Adams
September 24, 1941

PROGESTERONE AND RELATED COMPOUNDS

Progesterone is the female sex hormone produced in the corpus
luteum, a small yellow body in the ovary. Its function is to
condition the uterus for fertilization.

The isolation of pure crystalline progesterone from sows1
ovaries was first announced by Butenandt and his coworkers at the
Kaiser-Wilhelm Institute in Berlin in 1934 (only shortly before
similar announcements by Slotta, Allen and Wintersteiner, and Hartmann
and Wettstein). Its structure was first suggested by Slotta but was
confirmed largely through the synthetic work of Butenandt and his
collaborators, who synthesized the hormone from pregnanediol and from
stigmasterol.

31CK3

sol

Progesterone

yn3

Progesterone is the most highly specific of the sex hormones.
Very few of the compounds related to it show any of its biological
activity. Some of those which do are the 3-enol acetate of
progesterone (100$) , testosterone (low), 17-methyltestosterone (low),
17-ethinyltestosterone (good), 6-dehydroprogesterone (b0%) t 6-
hydroxyprogesterone acetate, 21~hydroxyprogesterone, 21-methyl-
progesterone, and 20-norpregnanolcne,

Cole has very recently announced the preparation of a number of
compounds of the type II:

CH3

R = Me, Et, or phenyl

These were prepared by the addition of dialkylcadmium or zinc com-
pounds to the acid chloride of A5-3-acetoxybisncrcholenic acid (from
stigmasterol),. The pure products were inactive, but the crude
materials showed progestational activity.

Considerable research has been carried out since 1934 in an
effort to find practical syntheses for progesterone from the cheaper
sterols such as cholesterol, stigmasterol, the bile acids, and the

- .-

\ \

J - . (K-

- 2 -

saponins. At attempt has been made to summarize this synthetic
work in the following pages.

Syntheses of Progesterone from Stigmasterol:

HC02H

ozone

AlA/ —
,kXJ m

acetate di

bromide

Stigmasterol

3-Hydroxybis-
norcholenic acid

COCH3

cyclohexanone
al-isopropcxid

Br2 in AcOH

Pyridinium salt

<A>-1, 2-isomer

1 .1

'. .. Ui t.: - , •» -

-3~

Syntheses of Progesterone from Cholesterol

/KV"

CrO,

KOH

VIII acetate
difcromi

Cholestero

CK2CH3

L-OH

Acetate

t, Jij. . -nr\m NaOH

Pyridine + P0C13 MeQK
1 CHCH3

CH,

^ C-C02Et

0s04

4 CKOKGH3
LoH

CH.

Ac,0

pyridine

CHOCOCH3
,4- OH

COCHgf

cyclohexanone
al-isopropoxide

■ /

17-isoprogesterone

HC1

'NaOH

hA/V

/ Dehydroandrosterone
chlero- \
e (/KCCH KCN\AcOH
H ^1 CN

OH - i_OH

x— — j

Kg acetamide

pyridine +
4/ POCI3
CN

Free acid
-C02

cyclohexanone
al-ieopropoxide

Progesterone

-4-

Spielman and Meyer have developed a method of oxidizing di-
bromochole sterol directly to progesterone by means of acid per-
manganate, The yield is 0„2$ and the product is not isolated in a
pure state, but the concentrates have been found to be satisfactory
in biolog ).c al w ox k ,

Convex 'sion p f 3h £ ogenin s to Pregnane Derivatives

! - / C^3 A 0

CH--C > CH-CR3 -a-*

X 6 CH2-CH2 200

HO
Sarsasapogenin

CK3

CHoOH

Pr egnane di one-3 , 20

CH=C»CHSCHSCH

! \

\^°

CH,

Pseudo sarsasapogenin

KOAc

* Pregnane-

EtCK diol-5,20

*«. Tiogenin (differing from sarsasapogenin in having an allo-H
at C5) can be similarly transformed.

Marker has recently applied this series of reactions to
diosgenin, which has a double bond at C5 6, and obtained progesterone
in good yields (See Organic Seminar by EiH. Riddle, April 30, 1941),

Synthesis of Progesterone from Bile Acids

XII

C02H

Barbier-
Wieland

NiArV

COpH

Barbier-
Wieland

Hyodesoxycholic acid

-5-

C03H

/•

Barbier-
Wieland

COCH3

KHSO4 ^

-K20
(3-acetate )

ISO1

Progesterone

Hoehn and Mason have prepared lithocholic acid from desoxy-
cholic acid (from bile) in 50 per cent yield by preferential'- _
esterif ication of the 3-OK and oxidation of the hydroxy! at
position 12 to a keto group. This was then removed by Wolff-
Kishner reduction. Lithocholic acid can be converted to progesterone
by known steps.

/\jL/X~\AcOaH

/M^V

* *

Desoxycholic Acid

CO^H

Lithocholic Acid

1367

04tu ^-LbKtu;.
m.t IX, 331 (1940); X, 303 (1S41).
)hal, and Hohlweg, Zeit. Physiol£hem. ,

Bibliography

G-ilaan, Organic Chemistry, p

Ann. Repts., 37, 540 (1940).

Ann. Rev. Block e

Bute.nandt, Westph„.

Slotta, Ruschig, and Pels, Ber. , 67, 1624 (1934).

Allen and Winter steiner, Science, 80, 190 (1934).

Kartmann and Wettstein, Helv. Chim. Acta, 17, 878

Bufcenandt and Schmidt, Ber., 67, 1893, 1901

Butenandt, V/estphal, and Cooler, Ber., 67.

Kahlbaum a~G. Fr. 822, 551 (Jan. 4 1938).
rfestphal, Naturwiss., 24, 696 (1936); Ruzicka, Hcfmann,

227, 84 (1934)

1365 (1934 ).
(1934).
1611 (193 4)^- Schering-

Eelv. Chim.

388, 1367

318 (1940

Soc. Exp
tfamoli, Ber., 71.
Ehrhart, Ruschig,

nd Meldahl,
Acta, 21, 371 (1938); Wett stein* et al. . 'ibid. . 25.

7 1371 (1940); Ehrenstein and Stevens, J. Org. Chem., 5,
0); Van Keuverswyn, Collins, Williams, and Gardner, Proc.
. Biol, Med., 41, 552 (1939).

2701 (1938).

and Aumuller,

Z. Angew. Chem., 52, 365 (1939),

Butenandt and Marnoli, Ber., 67, 1897 (1934); 68, 1847, 1850 (1935).

. _

« -. t

t--« ■ <

r4s

t ^

1 1

-6-

Marker, Wittle, and Plambeck, J. Am..Chem. Soc, £1, 1332 (1939).

Butenandt, Dannenbaum, Hanisch, and Kudszus, Zeit. Physiol Chem.,

257, 57 (1935).
Butenandt, Schmidt-Thome, and Paul, Ber., 72,1112 (1939).
Miescher and Kagi, Helv. Chim. Acta, 22, 184 (1959).
Goldberg and Aeschbacher, ibid. . 22, 1185 (1939) „
Butenandt and Schmidt-Thome, Ber., 71, 1487 (1938),
Spielman and Meyer, J. Am. Chem. Soc, 61, 893 (1939).
Marker and Rohrmann, ibid. r 62, 518, 898 (1940).
Marker and Krueger, ibid., 62, 3349 (1940),
Marker and Krueger, ibid. . 62, 79 (1940).
Hoehn and Mason, ibid. , 62, 569 (1940).

Reported by Robert L. Frank
October 1, 1941

KOELSCH'S WORK ON CYCLIC DIKETONES

Koelsch's work on diketones derived from cyclopentane may be
divided into two parts: (X) Enolizability; (2) Reactivity of the
carbonyl groups when adjacent to each other B

A survey of the literature indicated that the following
generalities could be made concerning the enolization of poly-
ketocyclopentanes ;

(1) Cyclopent&nes containing more than one ketone group and
having the ability to enolize twice, exist as mono enols.

(2) If the ketone is a derivative of cyclopentene, enolization
is hindered.

Under the influence of reagents which produce enolization
such as alkali, the above statements do no hold. Treatment with
base, depending upon the amount used, forms a yellow or red sodium
salt of the compound if only one double bond is present in the ring,
or a purple one if a second double bond is introduced. Necessary
conditions for the fulfillment of these effects are the presence
of the following chromophonic groupings:

One example of this behavior in the literature is oxalyldibenzyl-
ketone.

0 0

* ».

NaO
0

— 0

NaO
0

ONa
0

0
purple

yellow yellow

Exceptions to these generalities were the iollowing:

0 0

,\ — C6H5

H
=0

(A)

(B)

(C)

St-X V v>

- 2 -

l,2-Diketo-3-phenylhydrindene(A) is completely enolic while
l,3-diketo~3-phenylhydrindene (B) is predominantly enolic in polar
solvents, l,3-Diketo-p_-iodophenylhydrindene (C) gave two crystalline
forms, depending upon the solvents used. From acetic acid red
violet prisms were obtained, while from non-polar solvents colorless
needles were obtained.

The colors of the salts of the compounds in this series were
likewise peculiar, Compounds of type (D) are red and dissolve
easily in alkali to give deep blue or green salts. l,5-Diketo-2-
phenylhydrindene (B) dissolves in aqueous alkali to give a deep red

H
^0

JsrO

>v c

c=o

(D)

(E)

i
OM

(F)

(G)

sodium salt the structure of which is written as E. von Braun, who
worked with compounds of type D primarily on the basis of the
difference in color, assigned the ortho-ouinoid structure (F) to
these compounds. Koelsch, using 1, 2-diketo-3-phenylhydrindene as a
representative of group D, investigated the enolate structure of this
compound. He found that the deep blue alkali salt when methylated
with (CK3)sS04 gave a red 0-methyl derivative which had an absorption
spectra similar to the enol itself and to 2,3-diphenyrindone* The
compound added one mole of C6H5MgBr to give G-, since on oxidation
O-dibenzoylbenzene was obtained. This behavior indicated that the
ortho-ouinoid structure was not present,

.Turning now to the enolization of the diketones themselves,
two explanations are possible for their abnormal behavior.

1. The double bond common to both rings can migrate totally
or martially into the benzene ring, depending upon the environment.

-^s

-OH
0

2. Since 1, 2-diketohydrindene and 1,3-diketohydrindene are
both ke tonic, the grouping H influences the enolization in

such systems.

— '■?«-

. ~<s

-V

•

, • A ' A •

• . y ■

V ' V

A /\

A* )»

_,•

..—*■

. • -

- 3 -

To distinguish between these two possibilities, l,"2-diketo-
3,4,5-triphenylcyclopentene-3, 1, 2-diketo-3,4-diphenylcyclopentene-3
and lr3-diket0"2,4,5~triphenylcyclopentene-4, analogs to the above
in which the double bond is fixed, were synthesized and their
enolizations compared with corresponding hydrindenes.

Methods of synthesis used were as follows:

0-C=O

GH,

ctCr

c=o

0_C=CR
\

C=0

0C-?

OH H3

>-0

NOH

0ChsCOOH

— >

0-C
0-C

' \
0

0
0

NaOCH.

-0

0CN /

HNO,

A comparison of reactions is given in Table I (page 4).

Results point to a ketonic character for the cyclopentene
analogs and, therefore, indicate that the first explanation was th<
correct one. Reduction oroductfc ~in this series are enolic.

In the second part of the work the purpose was to find out
which of the groups in the cyclic a-dike tones was more reactive and
how this reactivity compared with that of carbonyl groups in an
acyclic cc-diketone.

Activation of carbonyl groups is usually characterized by a
tendency to add alcohols or water. Acyclic a-dike tones are
apparently not active enough to do this but a-ke toaldehydes and
triketones (2, 3, 4-triketopentane ) apparently are.

To study this activation cyclic dike tones were chosen which
showed no tendency to enolize. A summary of the observations made :
given in Sable II (page 5),

■J ..

■ ■

': V

■

•

.: .•

• ■

Reagent

0 H

_4_

TABLE I

0-C-C

0-^V

H 0

C=0

Coffipounds

/V'\ J

/XH

0
0

color

color in

alkali

Br:

0COC1 in
alkali

/C0C1 in
pyridine

0MgBr

red

blue-green

in cold
CCI4

1 mole
addition

orange
yellow

white

blue-green red
in alco-
holic alka-
li

Reacts in
hot HOac

0-benzoate

2 moles
addition

in cold alco-
hol but not
in ether or
CHC13

0-benzoate

1 mole
addition

yellow

blue in alco-
holic alkali

Reacts in hot

HOAc

C-benzoate

C-benzoate +
0-benzoate

Reduction

- .- .

■ -. 1 r ..

Compound
/V"\=0

-W-

-U

-5-

TABLE II
RCH 1 eq= NHSCH

UnsteMe J car-boj't.yi-2. oximated
aeetal

1 eq. 0MgBr

0 OCH3

/\X=0

carbonyl-2-
gives aeetal

carbonyl-2
gives aeetal

no rx

carbonyl-2
gives aeetal

carbonyl-2
oximated

carbonyl-2

oximated

carbonyl-1
reacts

=0
0

no rx

carbonyl-1

oximated

■

• •

»•'' " V

--■ U'i

■

. ^.-A

- 6 -

From the table one can conclude that in compounds bearing no
phenyl group on carbon— 3, carbonyl-2 shows Tore activation than a
earbon yl group in acyclic co-diketone. One phenyJ group diminishes
this acti ration while two removes it entirel^c

One explanation for the greater activity cf the 2-carbonyl
group is that 'Che double bond conjugated with the 1-carbonyl tends
to lessen the polarize t ion of this group. Therefore in the
rev or £.icle reactions of oxime or hemiacetal formation, any reagent
attacking the more exposed. l-oarbon-yl is eventually given up to
the L-earbonyl. Phenylmagnesiun bromide in its addition being
irreversible adds to the l-carbonyl •

The effect of the phenyl groups is to deactivate the 2-carbonyl
through electron donation. If the effect were one of steric
hindrance alone only the rate of reaction would be diminished.

Bibliography

Eoelsch, J. Am, Chem. Soc. . 53, 1321 (1936).
Keel 3 en' and Hochman, J. Org- ~Chem. , 3. 503 (1938).
Koelech and G-eissman_. ibid. , 3, 480 TiS38-V.
G-e i s sman and Koe 1 s ch '. i oi d „', 3 , 489 ( 1938 ) «
Koelsch and LeClaire"- ~ibTd*. , .6, 516 (1941).
Wawzonek, Thesis, University of Minnesota, 1939.

Reported by S. Wawzonek
October 1, 1941

THE REARRANGEMENT? OF ALLYL GROUPS IN
THREE-CaRBON SYSTEMS
Cope, et al

while carrying out an investigation of the alkylation of
cyanoacetic ester, Cope and Hardy found that when the sodium
derivative of ethyl (1-methylpropylidene )-cyanoacetate was
treated with aliyl halides the. expected product I was obtained.

CH3 CK3

CH3CH2C=C(CN)COOC2H5 NaQC3H5 , [CH3CH=C-C(CN )COOC2H5]~Na+

CH2=CHCH2X

CH3
CH3 CH=C- C ( CN } COOC 2H5

CH2 CR=l>R2
I

Although the ester I was sufficiently stable to permit purification
by means of distillation under reduced pressure, repeated dis-
tillations caused isomerization with the formation of a new com-
pound having a higher boiling point and a higher index of refrac-
tion. A good yield of the isomer could be obtained by refluxing I
at atmospheric pressure. That the isomerization involved a shift
of the allyl group from the aloha- to the yam ma —position was shown
by establishing the structure of the product as ethyl (1,2-
dimethyl-4-pentenylidene)-cyanoacetate (ll}»

CH3

pAfj0 I

I > CH3CH-C=C(CN)COOC2Hs

i «

C3H5

The structure of II was established by cleavage with aqueous
ammonia to unsym-methyl allyl acetone and cyanoacetamide., and
verified by synthesis from the same ketone and ethyl cyanoacetate.

CH3

II ^H4OH > CH3CH-C=0 + CH2(CN)CONH2

C3H5

CH3CR C=C + CH2(CN)COOC2H5 -* II

C3K5 CK3

- 2 -

Recently the investigation has been extended to include
several additional cyanoacetic esters, as well as malonic esters
and malononit riles. All three classes of compounds were found to
undergo the rearrangement involving a shift of the allyl group .
from the a- to the ^-carbon atom with an accompanying shift of the
double bond from the p, ^f- to the a, ^-position.

i ( i

-C=C-C-Y
i

C3H5

X
i ; i
-C-C=C-Y
i

C3H5

X and I = CN or COOCPH

2115

The isomerizations were brought about by refluxing the pure
compounds in a partial vacuum, t]

produce the desired boiling temp
all the cases investigated the r
ment product was higher than tha
pletion of the reaction was evid
steady increase. The structure
either by cleavage with ammonia
or by reduction followed" by cond
with urea to give a derivative o

;he pressure being regulated to
erature of the liquid. Since in
efractive index of the rearrange-
t of the starting material, com-
enced by a constant value after a
of the products was established ,
to ketones, as illustrated for II,
ensation of the reduced compound
f barbituric acid.

R3C-CH=C(COCC3H5) -£->• R3C-CH2CH(COOC3K5)2

urea

<5»H

3r-5

O^jtl

NaOCpH

3ju7

2iA5

0=C-

R3G-
1

CaH

•CH3-CK
l
0=G-

-NH

C=0

■NH

3ri7
v.

In some instances the structure was verified by means of an
independent', synthesis.

Cyanoacetic Ester Series

Six cyanoacetic esters, disubstituted by an alkyl vinyl and an
allyl group, were rearranged and the products identified by
cleavage with ammonium hydroxide.

C(CN)COOC3H

3A16

CaH

3^5

III

CH3 CH3

1 \

C3H5CH=C~C(CN)CCCC3K5 n-C4H9CK=C-C(CN)COOCsI%

CaH

3aB

C3H

3ix5

CH«

CRH

6ix5

C3H5

i-C3H„CH=C-C ( CN) COCC 3H5 CH3=C-C ( CN ) COOC3K5 CH3CH«C-C ( CN ) CCCC3H5

C3n5 CnHi

C3H5
VI

VII

'3n5

VIII

- 3 -

Mslononitrile Series

Two repr*eentative£ of this class, l-cyclohexenyl-allylmalono-
nitrile (IX) and 1-ethylpropenyl-allylmalononitrile (X) were found
to isomerize very readily. The structures of the rearrangement
products were established by cleavage with aqueous ammonia

■C(CN)a

C3H5

C3H5
CH3CH=C-C(CN)2

C3H5^
X

Malonic Ester Series

"GWO

Of the four disubstituted malonic esters investigated, only
ethyl propenylallylmalonate (XI) s.nd ethyl (1-butenyl)-
allylmalonate (XII), could be rearranged with the formation of
pure products. Side reactions occurring during the rearrangement
of the higher molecular weight esters (XIII and XIV) produced
mixtures of indefinite boiling points*

CH3CH=CH-C(COOC2H5)

C-aHt

C3H5CH=CH-C ( CCOC2H5 ) a
C3H5

XII

( CH3 ) 3CHCH=GH-C ( COCC2H5)

C3H1

XIII

-C(C00C2K5)

1 rii

XIV

The structures of the isomers produced by the rearrangement of XI
and XII were established by reduction followed by condensation
with urea. The barbituric acid derivative thus obtained was com-
pared with an authentic sample.

The rates of the rearrangements were studied for the ten com-
pounds above which gave clean reactions. The progress of the re-
actions was followed by means of the increasing refractive index
and rate constants were calculated from the following equations:

k, =

t 2 - tl

1 - xx *i and x2 = fraction rearranged at
2,305 log - times t1 and t3

1 - x

per cent rearranged at time t =

"t - n1
nt - ni

x 100 of starting mat-
erial

nt = refractive index
of product.

- f ,

■

•■

- fc

- 4 ~

Good first order rate constants were obtained in all cases except
the malonic ester XII, the rearrangement of which was accompanied
by side reactions affecting the refractive index. Correspondence
of the rates to first order kinetics was interpreted as strong
evidence that the rearrangement is intra- rather than inter-
molecular. This view receivea support when it was found that there
was no interchange of groups when mixtures were rearranged. For
example, rearrangement of a mixture of ethyl isopropenyl
crotylmalonate (XV) and ethyl (l-methyl 1-hexenyl )--allylcyano-
acetate (V) yielded no crotylcyanoacetate or allylmalonaie
derivatives.

QH3

JHs

XV CH2=C-C(COOC2H5)2 + V A ■ > CH2-C=C(C00C2H5) 2 +
CH3CH=CHCH3 'v CH3CHCH=CH2

pH3
C4H9CK-C=C(CN)COOC2H5

A great difference in ease of isomerization of the three
classes of compounds was found. The most striking difference in
rate exists between the malononitriles, which rearrange much
faster than the cyanoacetic esters, which in turn isomerize
faster than the malohic esters*

In order to determine whether the allyl group undergoes in-
version during the rearrangement, the crotyl derivatives XV and

XVI were isomerized*

CH3 CH3

CH3CH=C-C(CN)COOCaHs ■ ■" ■> CH3CH-C=C(CN)COOC2H5

CH3CH=CHCH2 CH3CHCH=CH2

XVI

In both cases, the only product was the one resulting from an in-
version of the crotyl group.

The authors conclude that the rearrangement of allyl groups
in acyclic or alicyclic three-carbon systems is a general reaction
when the a-carbon atom is attached to two nitrile or carbethoxyl
groups, or to one of each of these groups. The isomerization is
similar in type to the Claisen rearrangement of allyl ethers of
phenols and enols in which the allyl group shifts from the electron
attracting oxygen atom to a carbon atom.

_C=c-0-C3H5 -+* ► C3H5-C-C=0

In the isomerization involving a three-carbon system, the allyl
group becomes detached from the a-carbon atom, electron attracting
because of the two negative groups attached to it, and recombines

OC 22

- 5 -

with the 0- carbon atom which is lees electron attracting. It is
generally accepted that one motivating force responsible for the
Claisen rearrangement is the unequal sharing of the electron
pair binding the allyl group to the a-atonic The fact that the
nitrile group is more electron attracting than the carbethoxyl
group offers a ready explanation for the greater ease with which
a malononitrile isomerizes as compared with the cyanoacetic and
malonic esters.

Bibliography

Ccpe and Hardy, J. Am.' Chem. Soc, 62, 4<±1 (1940).
Cope. Hoyle and Heyl, ibid. . 65, 1843 (1941),
Gope^ Hofmann and Hardy, ibid., 65 1852 (1941)..
Tarbell, Chem. Rev.", 27," 495 (1940).

Reported by Clay leaver
October 8, 1941

SYNTHESES OF ALDEHYDES THROUGH GRIGNARD REAGENTS

Formic Esters

The first preparation of aldehydes by means of Grignard re-
agents was reported by Gattermann and Maffezzoli, who treated
various Grignard reagents with a three-molar excess of ethyl formate
at -50°C> Aldehydes were formed in yields varying between ten and
fifty per cent.

/)MgX F+
RMgX + HCOOEt -» RCH — » RCHO + Mg(OEt)X

OEt

Cther esters of formic acid, namely methyl and amyl formates, have
also been used in this reaction. At the low temperature prevailing
throughout, secondary alcohol formation was minimized.

Ethyl Orthoformate

That Grignard reagents could be caused to react with ethyl
orthoformate to yield aldehyde acetais was simultaneously recorded
by Tschitschibabin and by Bodroux.

RMgX + HC(OEt)3 -* RCH(OEt)2 + Mg(OEt)X

Tschitschibabin added the orthoformate to an ether solution of
the Grignard reagent, refiuxed, and then evaporated most of the
ether. After most of the ether had been removed, a very vigorous
reaction occurred and considerable heat was evolved. The pasty re-
action mixture was acidified and the acetal isolated by distillation.
The yields ranged between fifteen and eighty per cent*

Tschitschibabin and his coworkers were able to show that large
excesses of Grignard reagent and increasingly strenuous conditions
resulted in two, or even three, of the ortho ester ethoxyl groups
being replaced by the organic portion of the Grignard reagent .

In general, it has been shown, aliphatic and arylaliphatic
Grignard reagents give better yields of aldehydes than do the aromatic

compounds.

Formic Acid

Zelinsky found that aldehydes were formed when Grignard re-
agents were treated with formic acid.

H +
KCOOH + RMgX -> HCR(OMgX)a > RCHO + Mg(OH)X

Houben investigated this reaction and reported that the yields
were never greater than thirty per cent. Use of copper and other
salts of formic acid in place of the acid itself failed to improve
the reaction.

CI*

- 2 -

Disubstitut ed Forme.mid.es

~jouveau.it showed that disubstituted formamidee reacted with
Grignarc reagents to give products which formed aldehydes on
hydrolysis,,

M2 K+

IICONHg + R'MgX -> RCR\ ► R;CHO + R2NH + Mg(OH)X

OMgX

He reported no yields, although he pointed out that a secondary re-
action id prominent in some cases.

HCON(CH3)2 + gCBHnMgCi — > HC(CsKi J aN(CHsJ a + MgO + MgCl2

A number of G-rignard reagents were used, and a variety of secondary
am?i,es w&.'s- studied^ In addition to the simultaneous reactions in-
dicated above, ether side reactions occurred in special cases, and
tne reaction was generally undefendable. Yields of aldehyde ranged
from twenty to fifty per cent, In all cases twenty to forty per
cent of the substituted amide was recovered unchanged.

I soo.vanid.es

Sachs and Loevy reported that benzaldehyde was formed when
phenylmagnesium bromide was treated with methyl isocyanide.

C£H5
CPHBMgX + CH3N=C -» CH,K=C( -> CH3N=CHC6HB -> CsHBCHO + CH3NHa

MgX

Gilman and Heckert later investigated this reaction and dis-
covered uhat the particular case studied by Sachs and Loevy was the
only one of the general type which could be made to occur. Even this
reaction produced benzaldehyde in negligibly small yield.

Et ho xy methyl ene An 11 in e

Monier-Williams showed that aldehydes are produced when G-rig-
nard reagents are treated with ethoxymethylene aniline.-

RMgX + CsH-EN-CHOEt -* EtOMgX + RCH=NC6HB -> RCHO + C6HBNHa

Ke reported yields of thirty to sixty per cent.

G arbodit hio Acids

Wuyts prepared aldehydes from Grignard reagents by means of a
series of reactions, with the carbodithio acid as the predominant

intermediate.

RMsX -— > RCS3H

'O'

ROSSH + K^KCOKKNHg * RCSNHCQNHNH2 + H2S

RCSNHCONHNH2 > RCK=NCONHNH2 + S

RC=NCONHNHa ^ —■ ► RCHO

- 3 -

Wuyts and his coworkers found that semicarbazide was superior to
other well-known nitrogenous aldehyde-ketone reagents in this re-
action. Side reactions are prominent, however, in this preparation.
Most troublesome of these is nitrile formation. Aromatic aldehydes
are formed in better yield than aliphatic aldehydes, since reactions
leading to the latter frequently stop after the second step*

Chi-;, al

Savariau reported yields of aldehydes of about fifty per cent
from the reaction of Grignard reagents with chloral

RMgX + C13CCH0 -> C13CCH(0H)R NaQH ■ > RCHCOOH *?°Pa» > RCHO

l
OH

Aldehyde Syntheses in which the Group *r.MgX is replaced by a Formyl
homologue

Spath prepared substituted acetaldehydes by means of the re-
action between Grignard reagents and ethoxyacetal.

EtOCH3CH(OEt)3 RMfiX> EtOMgX + RCHCH2OEt HsS°4 i RCH3CKO

OEt

* EtOMgX + EtOH +RCH=CHOEt 1 H23°4

It was demonstrated that aldehyde was formed from both intermediates.

A clever preparation of substituted acetaldehydes from Grignard
reagents was devised by Herschberg, who coupled Grignard reagents
with allyl bromide.

RMgX + BrCH3CHCH3 -> MgXBr + RCH3CH=CK3 — ► Dibromide

4/
RCH3CHO 4 Glycol 4r Diacetate

Either ethoxyacetic ester or its phenoxy analogue was used as
a starting material for a Grignard synthesis of disubstituted
acetaldehydes by Behal and Sommelet.

ROCH3COOEt + R'MgX -* R0CH3C{0H)R3, -» R3 » C=CHOR -> R3 ' CHCHO
EtOCH3CQR + R»MgX ■-* RRf C(0H)CHs0Et >RR'CHCHO

Comparative Studies by Smith and Nichols

Of the Grignard reactions leading to aldehydes in which the
-MgX group is replaced by the formyl group, Smith and Nichols
selected those involving ethyl orthof ormate, ethoxymetbylene aniline,
and carbcdithio acids as the ones showing the greatest promise.
Grignard reagents were prepared from a series of bromomethyl- and
bromopolymethylbenzenes which ranged from p_- and £-bromotoluene to
bromopentamethylbenzene and compared the yields of aldehydes obtained
when each Grignard reagent was converted to the corresponding

- 4 - 26

aldehyde by each of the three different methods. Best of the three
methods proved to be that involving ethoxymethylene aniline, by means
of which the aldehydes were produced in yields of sixty to eighty
per cent. The yields obtained in this synthesis were four to seven-
teen per cent better than those obtained by means of ethyl ortho-
formate, which ranged between forty-five per cent and seventy-five
per cent. Only the Grignard reagent from p_-bro mo toluene produced
tA respectable yield of aldehyde by means of the carbodithio acid
"vnthesis, p_-tolualdehyde being formed in sixty per cent yield. In
'-'~.l other cases the reaction was useless a-s a preparative method.

An outgrowth of this work which is worthy of mention is the
- ; ., . v- of the formation of bisulfite addition products of these
lr-r,'-?ere3. aldehydes. It has long been recognized that benzaldehydes
he • -.'.v:g at least one ortho group unsubstituteu formed bisulfite
addition products rapidly and in good yield. Smith and Nichols were
able to show that aldehydes having both ortho positions occupied
by hindering groups but its para position unsubstituted formed
bisulfite addition products in excellent yield, although slowly.
If, however, both ortho positions and the para position were sub-
stituted, only very poor yields of bisulfite product could be
obtained, even after a long time.

Bibliography

Smith and Bayliss, J, Org. Chem-.., 6, 437 (1941)..
Smith and Nichols, ibid., 6, 489 Tl94l).

All previous references are cited in the above two papers

Reported by R. B. Carlin
October 8, 1941

SYNTHETIC ESTROGENIC COMPOUNDS

Dodds, Lawson, et al, Courtauld Institute of
Biochemistry, London

Robinson, et al, Oxford University

Estrone and estradiol, the female sex hormones secreted by the
Graafian follicle, possess the physiological function of preparing the
vagina and uterus for fertilization.

These hormones are very expensive because (l ) they do not occur
naturally in great quantities, (2) they cannot be prepared cheaply
from related natural compounds such as cholesterol, sligmasterol, or
equilenin, and (3) great difficulty has been encountered in synthe-
sizing them from cheap starting materials.. Consequently, much work
has been done in recent years to synthesize more accessible com-
pounds which would possess estrogenic activity.

Estradiol

Dodds and Lawson in 1934 started a study to determine the
molecular structure essential for a compound to exhibit estroegnic
activity. It was found that a cyclo-pentano-phenanthrene structure
was not essential when they discovered' that 1-k.eto-l :2:3 :4-tetrahydro-
phenanthrene (III) displayed marked estrogenic activity. They syn-
thesized this by the method used by R.D. Haworth for its preparation;

y\?

CH,

CH.

-C

,4>

A-i-C 1 ^

nitrobenzene

Chg— CHg— CH3— C02H

8b% H2S04

100'

C-CH2-CH3-C-0H

Clemmensen

III

They next found that a compound need not contain a phenanthrene-
type structure in order to be estrogenically active. As an example,

. ;

./,;

- 2 - 28

diphenyl-a-naphthyl~carbinol i\?as moderately active. However, the
corresponding £>-naphthyl-carbinol was not effective. Of much
greater importance was the fact that while 4:4?-dihydroxybiphenyl
was moderately active, 4;4*-dihydroxv stilbene displayed a much
greater activity.

On observing these facts, Robinson noted that both hydroxyl
groups might well be phenolic, although only one is phenolic in
estradiol, and that a stilbene type of compound seemed to have an
enhanced activity. Consequently, he synthesized a series of C-
alkylated derivatives of 4:4,-dihydroxystilbene by the following
method which used anisole as the starting material:

i£°L „ „ /-=\ 8 S /=v OOH Sn-H3-

CK3Q-< ><H — CH3Q-< >C^-<"

KC1

CHaO^ >~CH2-C~<, />-OCH

C3H5ONa

S— -N, ? 9 /^=r*\ RMgBr

CH30-< > C-C-<v ,,VoCH3 —

X , , £ N. V

(90$ yield)

PBr.

\ xA-0CH3

R OH
(90$ yield)

CHC1.

^N ? ^ N K°K, C2H50H

ch3o-<; ^>-?"c"< />-och3 : >

X / h >s V 200°

(83$ yield)

H0-<^ ^>-C=C-^ ^Xqh

R
(85$ yield)

Biological tests showed that a remarkable peak in estrogenic
activity was found when R=Ethyl, and that this compound is two to

■•■'■
■.■-•■

.. I

•

•.'

- 3 r-

three times as active as estrone itself, The reason for this
activity probably is that it is closely related stereochemical^
to estradiol;

CH,

/%*/

P
JS

.CH2

CH,

CHa

CH8
C=

Trans-Die thylst lib estrol
"Stilbestrol"

cia-Di ethyls tilbestrol

One difficulty connected with this synthesis is that in addition
to the highly active diethylstilbestrol (m.p,, 171°), a less active
geometrical isomer (mfp., 141°) is obtained. By stereochemical
analogy, it is apparent that the trans-isomer is more closely related
to estradiols and,, therefore, that the trans-isomer should be the
more active isomer. - Wessely showed that this is true by hydrogenating
the two isomers of dime thy Is tilbestrol, the structures of which are
known,, Hydrogenation of els-dimethyl stilbestrol gives the me so-
compound, while hydrogenation of trans-dimethyls tilbestrol gives a
racemic compound 9 Since hydrogenation of the more active diethyl-
Sbilbestrol gives a racemic compound^ it must have the trans-
configuration0 CI s~di ethyl stilbestrol can be converted completely
into trnns-diethyls tilbestrol by exposure to sunlight, so that none
of the less active cis-isomer need be obtained*

Another difficulty connected with this synthesis is that in the
dehydration of the carbinol intermediate by means of phosphorus
tribromide, the dehydration can give an unsaturated side~chain as

we-Lx

suit
this
that
them

as the stilbene compound, so that two racemic mixtures could re-
in addition to diethylstilbestrol. Wessely showed that
really happens by isolating the two racemic mixtures. He found
these could be converted into diethyl stilbestrol by warming
in a chloroform solution of iodine.

In the course of their work, Dodds and Robinson hydrogenated
stilbestrol and obtained a compound which is even slightly more
active than stilbestrol. This compound is called "hexestrol, "

Diethyl stilbestrol

/CH3 OH

CHS f/S/

Hexestrol

.I-. .. • *.',

•

•f <■ \

-•

' : :

"2 } ■ - • - . . ■

* -i. ' . ' ,' ■"," ■ ■ ■ j .{ ; :
* - "i i • '■ o i

;.u:

: •' .' . ;

!••:,/

'r. v'^

\ '/

- 4-

cis-3tilbestrol on hydrogenation in the presence of platinum oxide
yields a compound melting at 184°, while the trans-isom-aTon hydrogena-
tion in the presence of a palladium catalyst- yields a compound
melting at 128°. The estrogenic activity of the lower-melting form
is much less than that of the higher-melting form. As shown by
Figure IV, the compound analagous stereochemically to estradiol is
the meso compound, so it was postulated that the active, higher-
melting compound was the meso compound, and that the less active,
lower-melting form was a racemic mixture, Wessely has shown this to
be true by resolving the lower-melting product by means of its a-
bromo-TT - camphor- sulfonat e.

Dodds and Robinson also prepared hexestrol by the following method:

0"
H

CpH,

Al.Hg

moist ether

^V"

ACpO

Di-
acetate

OAc

Hexestrol

btlll another method of preparation of hexestrol is that
developed simultaneously by Dodds and by Bret Schneider whereby ethyl
feiagnesiun bromide is added, to anisaldazine with simultaneous loss of
nitrogen:

J
OH,

NH2-NH:

/Vc=N-N=C
CH3

■OCH.

p8HRMgBr
-Nft

OCH,

KOH

Hexestrol

200

It was possible to isolate both the racemic dimethoxyhexestrol and the
meso-dimethoxyhexestrcl, and it was found that the racemic mixture
could be converted into the highly active me so-compound by heating in
the presence of a palladium- charcoal catalyst.

a.-ril

■■■.,'•■ » ■ ■

' i

.-»••

. '. ■•

'■ '' ■ .... (--•■'

r ;'■;: H£'t-

-5-

Since the over-all yield in the latter synthesis employing the
aldazine intermediate is low, and since the hydrogenation of stil-
bestrol to hexestrol is very difficult, various investigators have
been investigating possible new methods of synthesis of hexestrol in
order to develop a cheap way of preparing this compound.

Two syntheses which are promising necessitate the preparation of
the Intermediate, a-p^-metnoxyphenyla^-bromopane,

C— CHp^CH;

The method employed by Bernstein and Wallis is represented below by
Method I, while that of Docken and Spielmann is represented by

Method II.

Method I

0
II
C-C2H5

NaOH

(CH3)2SC4

0
CH3

C-C2H5
Na

— ■ — >

CpHk0H

dry HBr
0°

C-C2H5
H Na

CaH

2xi5

CH-CH

11 ^A

-*• Hexestrol

0CH?

Overall yield, based on ,p_-hydroxypropiophenone, about 7%

Method II

CH=CH-CH.

CH3

Ane thole

HBr

V V-

OHs

CH-C3H5
Br

a-p_~methoxyphenyl-a-bromopropane

Docken and Spielmann obtained hexestrol from the a-jD-methoxyphenyl-a<-
bromopropane by coupling the propane derivative with magnesium,
followed by demethylation at high temperatures by means of potassium
hydroxide. The over-all yield, based on anethole, is 10-15$.

I I

I :

1 ' :

> ■

- 6 - 32

Bibliography:

Cook, Dodds, Hewett, Law son, Proo. Roy. Soc, 114B. 272 (1934).

Haworth, J. Chem. Soc, 1932, 1125.

Dodds and Lawson, Proo. Roy, Soc, 125B. 222 (1938); Nature, 159,

627, (l&r.?Y-; ibid., 159, 1068 (193777 ibid. , 140, 772 (1937).
Dodds, Go:.'';crg, Lawson and Robinson, ibid. , 141, 247 (1938); ibid. ,

.V;,2. 34 v-"3); ibid., 142, 211 (193877 Proc, Roy. Soc, 127B.

140' Ur33.}6
Rename and iloblnson, J. Chem. Soc, 1955 , 607.
Wease'iy ar.c. ivVilebsr, Naturw, , 28, 780 (1940),
jifes36ly, Kerschbaum, Kleedorfer, Prillinger, and Zajic, Monatsh.,

i o , -!.;, . v ■'■-' ±^' J «

Oaiv.pbell, Socles, and Lawson, Proo. Roy. Soc, 128B. 255 (1940).
BretsnhneJldsr, dsJonge-Bretschneider, and Ajtai,, Ber.# 74, 571 (1940).
flerr.steir: ana \vallis, J. Am. Chem. Soc, 62_, 2871 (194077
Doclr.en and Gpielmann, ibid. t 62 , 2163 (1940).

Reported by C. F. Jelinek
October 15, 1941

« % 't

/ J- «. .

• 1 • '

•

» .

...... * §

/ ■

DERIVATIVES OF METHALLYL CHLORIDE
Shell Development Company, Emeryville, California

As a result of the recent advances In the chlorine substitution
of isobutylene _ on a commercial scale methallyl chloride (^-methyl-
allyl chloride) has assumed importance as a synthetic intermediatec
The presence of a reactive chlorine atom and an olefin linkage makes
possible a large number of useful reactions,,

Commercial methallyl chloride containing 4$ of £, p-dimethylvinyl
chloride is prepared by allowing a mixture of chlorine and liquid iso-
tu'jylene to react for 0,006 second at such a temperature (C-1500 0)
thai; a liquid film on the walls will catalyze substitution, and then
passing the reaction products into water tc reduce secondary re-
actions of HC1 on the olefins present, The impurity of vinylic
chloride, which is extremely inert, is ol little importance in re-
placement reactions; this is not true for reactions involving the
clouble bond.

Metathesis Reactions of Methallyl Chloride

The reactivity of the chlorine atom in this molecule has been
measured by means of the reaction velocity with potassium iodide;
U.ns a comparison shows that allyl < methallyl <crotyI <M-chloro-2-
ine ieh.yl- 2-but ene .

CH3=CHCH3C1 <;'CH2=CCHSC1 <T CH3CH=CHCHSC1 < CH3CH=CCH3C1

CHa CH3

The chloride is readily hydrolyzed by alkali, the product de-
pending upon the temperature, alkalinity, and agitation. Inefficient
mixing may cause local points of acidity and rearrangement to iso-
butyraldehyde; high alkalinity favors formation of dimethallyl
tther from the methallyl alcohol and unreacted methallyl chloride.
The necessity for a large excess of water and careful pH control
indicates that allylic chlorides are capable of reacting in two
ways; i.e* by a unimolecular reaction with the solvent (alcohol or
water) and by a bimclecular replacement of the chlorine atom by a
hydroxyl or alkoxyl group. It is interesting to note that methallyl
chloride is structurally incapable of eliminating HC1. Careful
control will give the alcohol in 90$ yield.

Methallyl chloride is so reactive that many of its ethers can be
made by heating with the alcohol and concentrated aqueous sodium
hydroxide* Thus, formation of dime tidily 1 ether in "90$ yield is
maintained by the exothermic reaction cf the chloride and alcohol in
50% alkali. The ethers from the more highly dissociated alcohols -
methyl, ethyl, isopropyl, and phenol - which furnish a more favorable
ratio of alkoxyl to hydroxyl ions give good yields of the correspond-
ing ether even in the presence of water. Acid-catalyzed dehydration
of the alcohol fails because methallyl alcohol rearranges to iso-
butyraldehyde in acid medium. The isobutyl group can be introduced
into phenols without formation of tertiary phenols by rearranging
methallyl phenyl ether to isobutyl phenol (Ciaisen Rearrangement),
followed by hydrogenaticn.

' i - ■

\i i.. &

. ;■< : •:••::;;■ • ■■'.'

■. ti •»■■■;'■■•' ;> '' ': .
.!.■;■ '.;•; ■ /; I .:•* :■■■/'■
''■■'. i- •

;■;•:■ ! ...:■' ■ ■
. . !: f .'..-.. . ; ..
■ •: i ■ .. :.. 5 '■•
f : ; ■ f ■.: ;

Treatment of met allyl chloride with aqueous ammonia in an auto-
clave for two minutes yields 56$ primary, 26$ secondary, 8$ tertiary,
and 5$ quaternary amines; the high percentage of secondary amine
shows that primary methallyl amine is more reactive than ammonia.
By including ammonium chloride in the reaction mixture the amount
of secondary amine is reduced, since the primary amine forms a hydro-
chloride salt as produced*

Methallyl chloride reacts with many metal salts, such as KBr, KI,
Na3S, NaSH, NaSCN, to form the bromide, iodide, sulfide, mercaptan,
and thiocyanate respectively. Cuprous cyanide must he used in pre-
paring the cyanide since the alkali cyanides cause a shift of the
double bond. Substituted barbituric acids can be prepared from meth-
allyl chloride, a monosubstituted barbituric acid, and caustic or by
the chloride with malonic ester and sodium, followed by urea.

Magnesium gives first the Grignard reagent, which immediately
couples to form dimethallyl. However, 90$ G-rignard reagent may be
obtained by using a large excess of magnesium and ether. The Barbier
reaction has been applied to utilize the Grignard as formed - for
example, acetone yields 2,4-dimethyl-4-penten-2-ol.

Reactions Involving the Double Bond

o
Stirring methallyl chloride with 80$ sulfuric acid at 10 , then
decomposing the sulfuric ester with ice yields 65$ of isobutylene
chlorohydrin (chloro-tert. butyl alcohol), the remainder of the
methallyl chloride being rearranged to £, p-dimethylvinyl chloride.
At 40° the sulfuric acid treatment causes 85$ rearrangement; the re-
moval of unreacted methallyl chloride from the vinylic chloride is
effected by saponifying with KOH.

Concentrated HC1 adds to the double bond| it has been reported
that HF catalyzes the addition of methallyl chloride to benzene to
form (chloro-tert. butyl) benzene,

CH3=C-CH3C1 + C6H6. ijfffru C6H5-C(CH3)3CH3C1

CHo
Chlorination of liquid methallyl chloride or j3, p-dimethylvinyl
chloride results in ?0$ of the unsaturated dichlorides; bromine gives
93$ addition* A solution of chlorine in xvater gives chiefly dichloro-
tert. butyl alcohol, along with some dichloro-isobutylenes, tri-
chlor-alcohols and trichloropropane. A reasonable mechanism postu-
lates that chlorine substitutes to form a positive charge on the
tertiary carbon atom, to which water or chloride ion may add or from
which a proton may be expelled to cause unsaturation at different
points in the chain,

CH3=CCH3C1 fil*-* CH3C1-C+-CH3C1 "■2-° > CK3C1-C~CH3C1 + H+
CH3 CH3 ^

C1-
+C1- > CH3C1-CC1-CH3C1

CH2C1-C-CH3C1 or CH3C1-C=CHCL
U H-

CH2 h+ CH-

■•: : : -;. ■.:'.

■(•

. y

f .-,••« <«

.... - . ......V -mi

' ..-_..:

» 81*

In comparing isobutylene and methallyl chloride, it is seen
that the former is more reactive, as evidenc-ed in polymerization,
addition of HC1, and chlorine substitution.

Reactions of Methallyl Alcohol

Methallyl alcohol is rearranged by 12$ sulfuric acid in almost
quantitative yield to isobutyraldehyde; any impurity of dimethallyl
ether is also hydrolyzed and rearranged.

CH8=CCH8OH -> CH3GCHO

CH3 CH3

This aldehyde can be oxidized to isobutyric acid; the overall
conversion from isobutylene to isobutyric acid is 75$.

Refluxing methallyl alcohol and isobutyraldehyde in 2b% sulfuric
acid gives isobutyleneglycol-J-spbutyracetal^ which is decomposed
to the glycol and aldehyde on heating with dilute acid.

9 ^H* . 9H

CH3CHCH-0~(jjCH3 -* CH3CHCHO + CH3CCH3OH

Esterification proceeds by distilling the organic acid with
methallyl alcohol. Again, use of mineral acid must be avoided be-
cause of rearrangement to the aldehyde* Methallyl alcohol, its
esters, and ethers may be hydrogenated below 200° over nickel to the
corresponding isobutyl compound.

Methacroiein is prepared either by dehydrogenation or oxidation
of methallyl alcohol; the oxidation process is employed because of
less rearrangement to isobutyraldehyde. The process consists of
passing alcohol and air over silver gauze at 500°. Unless inhibited,
methacroiein polymerizes on standing at room temperature to a white
granular solid.

Methallyl xanthate, ueed in ore-flotation, can be made from
methallyl alcohol, alkali, and carbon disulfide.

6-methvl glycerol and its Derivatives

Methallyl chloride has a series of derivatives similar to those
of allyl chloride, hence the "£-methylg3.ycerol" homologs*

CI OH CI , . 0

CH3=CCH3C1 3°£k^ 6H8dCHa J£i2££8 CH3-C-CH3

CH3 CH3 01 6h3

dichlorc-tert .butyl p-methylepichlorohydrin
alcohol

H30
,H3S04

'. .1

C l '.

on j

lo U .-v i ; ,.; '• .

. .1

1 s,

«'■

rt .2 1 u {

: ; . '. ' .•-..•

* '

x u

• -I

£*<*« •-'• i: <■ -n*

l >, T f

Ka0

OH OH OH

chsc — ch2 4-

fy H2S04

p-methylglycerol

- 4 m
NaOH

CI OH
CH3-C—
CH3

OH
I

OH-

p-methylglycidol

£~methylglycerol mono-
chlorohydrin

The advantage of this series of reactions over the direct hydrolysis
of di chl or o-tert .butyl alcohol or jis-methylglycerol monochlorohydrin
is that the salt need not be removed from the glycerol produced.

Derivatives of Methallyl Chloride

isgbiitY]*-] Isobut^racetalHoSQ Isobutylme
aiaehyQ.3 iof iSobutylene Glycol

RQH

NaOH
Ks0

Methallyl
Ethers

OH-

Methallyl
Alcohol

H8S0j Glycol

o'xidi. £ e

Isobutyr-
jjwgange with. ikgpj aldehyde

Alcohol
NaOH

Metal

Methallyl
Chloride

MRthally.l ,

Dimethallyl

Ether

dehyi-r

Salts

JSSi

Methallyl Bromide,
Iodide j Sulfide
|Thiocyanate,
jMercaptan, etc*

dil.

"Sethacrofein

oxidize

RCOMMethallyl
I Esters ,

Isobutyric

Acid

^sfl so butyl j
S ilAlcohol j

Grignard
Metais

Methallyl
Amine

Methallyl

Chloride

Dimethallyl
Amine

etc*

Methallyl Subst.
Hydrocarbons.
[Dimethallyl

H0C1

I Dichlor o-tert*
(butyl alcohol

distil

£-methyl-

epichloro-

hydrin

Ha0

H2S04

KoSO,

HaS04

NH* j OH-

£-methyl
Glycerol
Monochlor-
hydrin

Rearrange; p-dimethylvinyl

chloride

H20
JQLl

Isobutylene
Chlorohydrin

1,3-diamino-

2-methyl-

propen-2-ol

NapH jp-methyl |
glycidol — i

ilstil

Isobutylene
Oxide

H2C
H2£04

^-methyl
glycerol

Dichloro-
Isobutylenes

p-methyl glycerol differs from glycerol in its easy conversion
to the unsaturated aldehyde by atmospheric distillation from 12$
sulfuric acid. These same conditions give methacrolein from any
member of the above series* Trichloro-tert .butyl alcohol, a minor

•

. ■

•

! — >,L. .

'■

—

■

i ' ' ■-'

,.- ivvr.J

Ji.

•-f-i-v ■

• - '..

■ «■..-.... ..

"~*

<5 :

" t ,

...

jr.

-■• -

-> • ■*.- *

- 5 -

by-product from the chlorohydrination, , gives an analagous series
containing a chloromethyl group in. place of the methyl.

Compounds Containing Conjugate Unsaturatlon from Methallvl Chloride.,

Methallyl chloride affords a feasible approach to the synthesis
of substituted isoprenedicarboxylic acids through ^-methyl glut a conic
ester.

CH3-C

CH3

W H0C1

CHa-Cl

CH2C1
I
CH3-C-.0H
/
CH3C1

KCN

CHoCN

CHa-C«-OH

EtOH

I dry HC1

GH*GN k°So4

CH3
I
0-CH=C~C=CH-COOH

COOH

a-benzal-p-methylglutaconic acid

/ uGl

. 0CHO
<:

KOH

CH2COOEt
i

CH3-C-OH

CH2C00Et

distil..

CH-COOEt
/I

CH3-C
I

CH2COOEt

p-methylglutaconic
ester

By use of £>-cyclocitral in place of benzaldehyde, this gives a new
route to the carat enoids and Vitamin A..

Bibliography

Burgin, Engs, Groll,tand Hearne, Ind. Eng* Chem., 31, 1413 (1939).-

Tamele, Ott, Marple, ' and Kearne/ ibid., 33, 115 tl94l),

Burgin, Hearne, and Rust, ibid..f 33 '. 385"Tl94l)«

Hearne,. Tamele, and Converse, ibid. f 35r, 805 (1941),

Hearne and DeJong, ibid,, 33, 941 (1941 5.

Hurd and Abernathy, J, Am* Chem. Soc, 63, 976 (1941).,

Reported by R. S. Voris
October 15, 1941

li, -aoGlq ft; i* :

Li !

•

■ . - - ■ t .

• ; &:•:. < »• » »...-,. ........

I •-• • •

,'.,;;:.•■ i .. • -" - ' . •

• i - • i ' %-U.J. t • ' > -■ ; * ■ ' • ■ \ • ■■> '■ ■

i .

PINACOL REARRANGEMENTS

The rearrangements of tri~ and tetrasubstituted ethylene glycols
and a, £-amino alcohols are summarized in this report.

I. Plnacol-oinacolone rearrangement

When tetrasubstituted ethylene glycols (pinacols) are de-
hydrated, they rearrange to form pinacolones*

R R R3£C-c£-R

;>c~c< _

R 0 0 R

H H

Sulfuric acid of various concentrations, hydrochloric acid, acetyl
chloride, and acetic acid and iodine are the most common reagents used
to bring about the transformation* The mechanism usually given is
the following, using pinacol itself as an example.

CH3 2

CH3 :

C : 0
* • • •

C : 0
• »

H
H

CH;

CH,

CH3 S p

ch3 :

l3 * H

ch3 ; u

» a

• •

: H

0 9

• »

\ H

• *

CH3 :
CH3 :

3 •

C : 0: H

ran

3 Li.

CH3
CH3

(CH3)3 >C~C~CH3 + HX + H.O

There are four types of pinacols, which can undergo the types
of reaction shown.

xc— c^

R^O 0XR
H H

Type I

HO— C-—R
VR

(CH2)

c— -c

H H

0NR

Type II

(OH.)

\R °

/ //

C— C-R or

(CH.)

R
R

i'l'. . [fiA3/; .::••.•.::

■

.1 1.

I . ■ i •

■

• .'

\ i_

-,.;-*.

•'•..

•</■■

» «

i . :

t- ■

:l ', ■■

r.—~ 'I. ■-■■{- \

t »•

» *

« . ■» ♦

.A-

**-*-*4^« .

- 2

r —

(CH3)n

\ /

c— c

/O 0\

Type III

<?H2)m

(CH2)

, C

(CHa)

(CH2)

(CH3)n

nS f n

(CHS)

R
.C-OH

(CHa)

.C-OH
R

C-R

.0=0

(CK2)

0=0

Type IV

The course of the rearrangement depends, first, on which carbon
atom loses an hydroxyl group, and, secondly, on which radical mi-
grates to stabilize the molecule. In type I, if the radicals were
all different, four different products could result.

Ro~ ■— C— 0— R^

#°/R»

R i — C— C~ — R3

Ra^ 0

XR3
<

0 RA

Rs — C— C--R3
R$

R2> C£C— 0f R3
NR*

Since the hydroxyl group takes with it a pair of electrons, it will
be lost from the carbon atom having the greater capacity for electron
release* The second consideration is the migratory ability of the
groups on the other carbon atom. The migrating group carries a pair
of electrons so that its electron-attracting power, of affinity cap-
acity, is important. In the aliphatic series there is an inverse
relationship between migratory power and affinity capacity while in
the aromatic series the two go hand-in-hand, showing that in one
case or the other there is another factor under observation.

The pinacol rearrangement has been used to compare the affinity
capacities and migratory power of various groups. Bachmann has
studied the rearrangement of symmetrical pinacols of the type
ArAr' C ( OH)C ( OH ) Ar Ar ' , The question of which hydroxyl group is lost
is eliminated and the product depends solely on the relative mi-
gratory tendencies of Ar and Ar' . In all cases both products were

( :.

. .;

i • :■

*'■»' 1.

n-nm

i : - *

- 3 -

obtained, but the amounts of each gave an indication of the migratory
power of the groups. If the relative migratory power of A and B and
that of «A and C were known, it was found that the behavior of B and
C together could be predicted. Thus the tendency to migrate was shown
to be a property of the group and not of the molecule. The following
series gives the relative migratory power of various aryl radicals
(phenyl = 1; anisyl, 500; phenetyl, 500; p_-tolyl, 15,7; £-biphenyl,
11.5 ; p_-isopropylphenyl, 9; p_~ethylphenyl, 5; m-tolyl, 2;
m-metnoxyphenyl, 1.6; phenyl and p_-iodophenyl, 1; p_-bromophenyl, 7;
p_-chlorophenyl, .66; o-methoxyphenyl; .3; m-bromophenyl, o- and &-
chlorophenyl, nearly zero. Except for steric factors this series is
comparable to the series for affinity capacities of these groups.

Bachmann has also studied the rearrangement of unsymmetrical
pinacols of the type ArArC(CH)C(OH)ArfArf . The migrating tendencies
found here were greatly different from those found in the symmetrical
type. They bore no simple relationship to the other values and could
not be predicted. This is because the ease of loss of hydroxyl
group rather than the migratory power was the factor under observa-
tion. Apparent migratory capacities in this series are as follows:
p-biphenyl> phenyl JN m-tolyl^ p_-tolyl\ p_- chlorophenyl J> phenetyl
anisyl,)' p_~fluorophenyle

The rearrangement of pinacols of types II, III, and
governed not only by the two factors mentioned above, bu
size of the ring. Thus, in type II, if n is four, a rin
will take place, as is shown by the following example:

C=0

IV is
also

; enla

by the

rgement

If n is five, a mixture of the possible products is form
of the type represented below rearrange to both fluorene
Dhenanthrenes.

and

ed, Pinacols
s and

Bachmann investigated pinacols of type IV
Only one reaction was found to take place.

.n the phenanthrene series.

■' ■

. , ,■ . ... , . j ■» - 1 • .

■■-■■■ r, , ' ■ '•!'««

■

1 . :. J

! \ *

- .,

- 4

Ar
-OH

Ar
OH

— -*

Ar 9 anisyl, p_-tolyl,

m-tolyl, p_-chloro-
phenyl, p_-fluoro-
phenyl, jo-biphenyl,
and a-naphthyl

In the allcyclic series, type IV gives a ring contraction in one case.

CH3 yt^ GH3

-OK

CH3
OH

Recently it has been suggested that the pinacol rearrangement in-
volves a Walden invej&sion. Bartlett has studied the rearrangement of
the els- and trans- forms of 7,8-diphenylacenaphthenediol-7,8.

0_C C~0

CUt-t-C — 0

Both geometrical isomers gave the same product, but the kinetics of
l-he reaction showed that the rearrangement of the trans- form was
much slower, corresponding to conversion to the ci&r "form before re-
arrangement. Some e^s~ pinacol was actually isolated from a partially
rearranged portion of _t:j..ans'- pinacol. Since the oi.s- form Is the one
which can rearrange by ftalden inversion, that procedure seems most
likely in the light of these facts.

Semipinacol Rearrangement

The dehydration of trieubstituted ethylene glycols can take
p. lace in a number of ways.

1.. vinyl dehydration
H R« r ft:

R-C— C-R;i i R-C^C-R"

Q 0
K H

R-i?"

R?
1

C-R"

I'..

. ir.

■ ■

. '

'■>■■

' '"• . '■ .

! H

; II

■

■ >

■ !

i ,-■■.' -::

^....." . ., r -

^.•'

..'*

r ...

,/!,

■ • • • . ' •■.•'. i "- •; :lv.r^: .J-

•

-«. ■'* '- - .■--—' —

- 5 -

H R»

i I

r_C— C-R"

6 6

H H

2. Semihydrobenzoin rearrangement

rpNr '

|r| _C— -C-R"
6

H-C-C-R'

11 ^R»

3. Semipinacol rearrangement

H R'
i I

r_C— .C-R"

I I

0 0

H H

-*->

H ri

R-C— C-R"

I I

R'-C-C-R"
R/ (,

riCc-C-R'

In the first two reactions, the tertiary hydroxyl group is eliminated,
but in the last one the secondary hydroxyl group is eliminated. The
removal of one or the other is controlled by the nature of the
groups attached to the respective carbon atoms as well as by the
nature of the reagent. Concentrated acid promotes reaction 3, while
milder conditions favor reaction 2, Dehydration of aryl dialkyl
glycols, ArCH(OK)C(OH)RR, with oxalic acid or dilute sulfuric acid
gives aldehydes (reaction 2) and with concentrated sulfuric acid gives
ketones (reaction 3). As the radicals become larger the tertiary
hydroxyl group is stabilized and the semipinacol rearrangement takes
place regardless of the reagent used.

Alkyl hydrobenzoins C6H5CK(0H)C(0H)RC6K5, give desoxybenzoins by
loss of tertiary hydroxyl when R is methyl, isobutyl, or phenyl. When
R is ethyl, propyl, isopropyl, butyl, Isoamyl, and cyclohexyl, of
desoxybenzoins and benzhydryl alkyl ketones, (CSH5)3CH-C=0 are

R
produced. Evidently the size of the radical is not the determining
factor in the rearrangement of alkyl hydrobenzoins*

It has been shown that the presence of an aryl group on the
carbon atom carrying the secondary hydroxyl group is necessary before
a semipinacol rearrangement can occur.

III. Retrooinacol r ear t a n gement

The dehydration of alcohols derived from pinacolones is
accompanied by the reverse pinacol rearrangement.

R3^C-C-R

fi-

RS>C=C <R.

Mixtures result when the radicals are different, Whitmore studied the
dehydration of two isomeric pinacolyl alcohols and found that they
gave the same three products.

■

.;■ O:,, . :,. |

•

■

■■ ■: .

'4 .:

f '

* - ■

. ;

■» -

■ ■ / . • ■ .

'. ' ■ •

x : ' ' •

.' . .

■

....

! 0 ' .;

"' I ■ , '■ • ' ';..-•

w 6 *•

C4H9 H / 9Ha 9H3
C4H9*(k~- C-CH3 ( CaH»-CH=C- C-C4H

CHa 0 / H

H _ — M / CH3 CH3

CH3 JJ ^ C4H-9— C ■ '■G—C4H9

J I

C4H9O — C— C4H9

ch3 6 I f3 F*H*

jj \ CH3-C=== C--C4H9

IV, Deamination reactions

Pinacol and semipinacol rearrangements may take place if one
hydro xyl group of a pinacol or trisubstituted ethylene glycol is re'
placed by an amino group. In this case there is no question of the
course of the reaction. The amino group is always eliminated and
migration takes place away from the carbon atom carrying the
hydroxyl group.

1. Pinacolic deamination

C6H5x /CgHs hono jP /G6H5

>C— C ► C6H5-C~ C-Me

C6H5X^ I XMe XC6H5

2, Semipinacol ic deamination

CsH5x?_C^e _HONO^ C6H5_cl0c^e

C6H5/ 0. NH3 NSeH5

The relative ease of migration of aryl groups may be determined by
making the groups different.

Whitmore has shown that semipina colic deamination involves a
Walden inversion. The optically active amino alcohol was used in
reaction 2 above and an optically active ketone resulted. By re-
lating the configuration of starting material and product to com-
pounds of known configuration, it was proved that a Walden inversion
has occurred.

Bibliography

Ann. Reports, 25, 134 (1928); 27, 116 (1930)* 30, 181 (1933),
Complete references given.

Bachmann et al«, J.Am, Chem. Soci, 54, 1124, 1969, 2112 (1932); 55,

3819 (1933); 56, 170, 2081 (1934); 57, 1095 (1935); 58,~Tll8
(1936), ' —

Whitmore, ibid., 55, 1528 (1933); 61, 1324 (1939).

Bartlett, jjbid., 6g, 2927 (1940).

Tiffeneau, "Glycols", Masson and Co,, Paris (1940),

Reported by J. 0. Corner, October 22, 1941

': - . -.,

•' £.M - . .,

' ■'■'■'■■ '■'■■'■ ♦ ;

1 ••..-• , J-1,

> ■ ~

i ' :

* ■•■•■■ '"

NEW SYNTHESES OF HEXATRIENES AND SQUALENE

The basis of these syntheses is the use of the Grignard compound
of 1,4-dibromobutane in reactions with carbonyl compounds to form
1,6-glycols by regular (1,2) addition, subsequent dehydration and
dehydrogenation giving the desired hexatriene.

The 1,4-dibromobutane is obtained by treating N-benzoyl
pyrrolidine with PBr5 according to the von Braun method?

NCO0

PBr,

•7"

Br
C0

— CHgBr
\N=CBr0

BrCH3CH3CH3CH3Br + 0CN

When
the G-rigna
mine, and
been shown
It is the
C6H5MgBr r
dissolved
almost qua
the purifi
with Se03,
lization o
pale green

BrMgCH2CH3

an ether solution of ben
rd compound of 1,4-dibro
the reaction product sep

that this product is 1,
same product that Bouvet
eact with ethyl adipate,
in water, a spontaneous
ntitative yields of 1,1,
ed hexadiene is dissolve

one obtains 1,1,6,6-te
f the product from aceti

plates - the color aris

CH3CH3MgBr + C6H54-C6H5

zophenone is added to a solution of
mobutane, the solution turns car-
arates as a viscous mass. It has
1,6, 6~tetraphenyl-l,6~hexanediol.
obtained when he let excess
When this impure alcohol is
dehydration takes place giving
6, 6-tetraphenyl-l, 5-hexadiene. If
d in glacial acetic acid and heated
traphenyl hexatriene upon recrystal-
c anhydride. It is in the form of
ing from the conjugation.

OH OH

} i

— > CgHs— C— CH3CH3CH3CH 3— C—CgHs — >
I 1

CbHk CeHc

CgHs— C— CH— Cri3GH3CH— C— Cg H5
I I

CeH5 C6H5

C6H5-C=CH-CH=CH-CH=C-C6H5
C6H5 C6H5

The hexadiene may also be converted to the hexatriene by heating
it with jp_-benzoquinone at 170-180°.

Fluorenone was also made to react with the Grignard solution of
1,4-dibrornobutane to form l,6-difluorenyl-l,6~hexanediol. This
turned out to be so very insoluble that ordinary recrystallization
was quite difficult. Hence the impure alcohol was heated with
benzene sulfonic acid in acetic anhydride, which eliminated the water
to give the hexadiene. This was converted to the hexatriene by
means of Se03 in glacial acetic acid as above. The result was 1,6-
dipiphenylene hexatriene.

iV,Xz

. ■

■

i '■

I ,

, ■■•-■yl\

t
i ,

•'• a»l*-,^:

ft*-

" t. : ■■•*"' .■■■"■>'■■ ■"" "' k.

!>v

-. -,:v ...; .-;■ ■ ... .:t.- /<-:,

ff«t/i

I;.

\ x . •

■ y •/'•- . i!vri:o ,f; r :r ..... ;.> ; .

.-.-: . ' ,u/-

• ■

- 2

CO + BrMgCH2CH3CH3CH3MgBr

*7%>.,

X ^

NC-CKSCHSCHSCH3-C

C=CH-CH=CH-CH=C/'

Acetophenone and benzaldehyde were also used in this reaction;
good yields of the hexatriene were obtained in each case. It is
interesting to note that the l,6-diphenyl*-l,6-dimethyl-l,6~
hexanediol from the acetophenone gives rise to two forms — one a
racemic modification and- the other a meso form. This results from
the two similar asymmetric carbon atoms which the molecule possesses.

OH ?H

C6H5-C*~CH2-CH3-CH2-CH2~C*-C6H5
l I

CH3 CH3

There are few classes of natural products that are harder to syn-
thesize than the carotenoids.

This is mainly due to the long branched carbon chain and the ex-
tensive system of conjugated unsaturations which are characteristic
of this class of compounds.

One of the most important of the carotenoids is ^-carotene which

can be converted by the body into Vitamin A, The new synthesis of

hexatriene s gives us a good means of producing 0-carotene.j
0Ha CH3

■ CH:=CH-C=CH-CH=CH~

CH3
CHa

R— C— Cn3CH2CH2CH3— C—R

CH:

CH«

R-C=0 + BrMgCH2CH2CH2CH2MgBr
CH3

R-OCHCH3CH3CH=C~R

CH,

R-C=CH~CH=CH~CHs=C-R
. i i

CH3 CH3

|3-carotene

However it should be noted that the ketone with which this syn-
thesis starts is not know but it has been w ell established that it
could be produced by condensing £-ionylidene acetaldehyde with
acetone;

» _ r.,

:. - ■ '

•

' j'l &

f-JV.f ' '..'

■ i

«,1 ',

;-*0- ■ ■

•. . '

. s-iv

» .r - > f

; - '"

gfc-. J W1'

' <'

,vi.'

,. i

-/,-.•:

■ .. i ■ . -

;;,; '• .. ;, '■ ■• ' . ;
':. ■■■:■ I: ■'J

CH3 CH3

CH=CH-C=0
\
CH9

CH3 CK3

,CH=CH-C=CH-CHO

CK3CHO

CHS

CH.

{5-ionylidine
acetaldehyde

CH3 C--CH3

CH3 CH3

CH=CH~C=CH-CH=CK-C=0
CH3 CH3

^CH3

It would seem simpler to use l,4-dibromo-2-butene and thereby
omit the dehydration, but this is not possible - only the pinacol
is isolated.

Squalene is an unsaturated hydrocarbon which occurs in the liver
oil of fish of the shark family and also in yeast. It is made up of
six isoprene units and gives crystalline compounds with HC1 and
HBr that are suitable for its characterization.

It is interesting to note that the Barbier reaction is employed
instead of the G-rignard reaction in this synthesis of squalene. In
the Barbier reaction the 1,4-dibromobutane is added to the geranyl
acetone in an absolute ether solution of Mg with a crystal of iodine
as catalyst, and the "nascent" G-rignard reagent reacts in a normal
manner before complicating side reactions can set in. A vigorous
reaction takes place and the diol is formed. This is converted into
squalene by distillation under diminished pressure.

The geranyl acetone is first formed from geraniol by an aceto-
acetic ester synthesis: q 0

CH3c£cH2-C^OC2H5
CH3C*CHCH3CHaC=CHCH2OH = R'CH R'CH PBra , R*Br — = —

CK3 CH3

CaHKOH

genanlol

•c

R' 1

i^°

CH,c£cH + (£oC,Hs Ba(°H)3

co:

H30
Ketonic cleavage

CH3C=CHCH3CH2C=CHCH2CHsC=0
Cn3 CH3 CH3

geranyl acetone

Then the Barbier reaction is carried out, using 1,4-
dibromobutane with the above product:

•

.1* k

. . . » V.

■'•• •

[ \

b. 4 -

OH OH

R'CH2C=0 + BrCH3CH3CH3CH3Br — ^-— ► R1 CH3C~CH2CH2CH3CH3-CCH3Rf

CH3 j CH3 CH3

■ — > CH3C— CHCH3CH3C— CHCH3CH3C— CRCH3GH2CH=C— CH2Cn3CH=CCH2CH3CH^.
CH3 CK3 CK3 CH3 CH3 Qjfa

CH3
Squalene

The advantages of this squalene synthesis are:

(1) Cheaper and more easily obtainable starting materials

(2) Squalene is produced in better yields and in purer crystal-
line form.

The old preparation started with farnesol — > farnesyl bromide

MgBr

► squalene. See reference to Karrer.

Bibliography

Josef Schmitt, Ann., 547, 103 (1941).

Jose;? Schmitt, Ann., 547. 115 (1941)

M. T. Bogert, G-ilman - Organic Chemistry, 1158.

Karrer, Organic Chemistry, 53,

All other references are given in Ann.

Reported by P. F-» Warfield
October 22, 1941

' ■ n J ■-

■: ■ -. it S ;

■ .

- <-. ■ ' -

'*. .1

s .1 '

ATTEMPTS TO DETERMINE THE STRUCTURE OF PHTHIOIC ACID
From the 1940 Presidential Address by Robert Robinson

Isolation

Phthoic acid was isolated by R. J. Anderson. He extracted
moist Hue tubercule bacilli with an equal mixture of alcohol and
water. The residue was further extracted with chloroform. The
chloroform fraction yielded a wax. The alcohol-ether extract was
treated with acetone which caused a precipitation of phosphatides
and left in solution various glycerides.

The phosphatide precipitate was decomposed by hydrolysis,
yielding 33% of water soluble constituents, consisting of carbo-
hydrates and glycerophosphoric acid, and 67% of a mixture of fatty
acids. This mixture of acids was converted into a corresponding
mixture of lead soaps. This mixture was extracted with ether,
leaving a residue, which yielded pulmitic acid on decomposition with
acid. The ether extract was hydrogenated catalytically and the re-
sulting mixture of acids was converted to the lead salts and extracted
with ether. A residue was left which, when decomposed, yielded
stearic acid. The stearic acid resulted from the hydrogenation of
oleic acid present Juthe original mixture of fatty acids* The ether
extract was decomposed yielding a yellow oil which solidified on
cooling. The melting point of the resulting solid was 21°, the
rotatory power La.]so£ was +11.96° and analysis showed an empirical
formula C26H530a. Anderson named this product phthioic acid.

residue

CHC13

residue
wax extract

tubercule
bacilli

alcohol

ether

acetone
extract > extract containing glycerides

a precipitate

hydrolyzed

33% water soluble gly-
cerides and glycero-
phosphoric acids

67% fatty acids

fatty acid
mixture

Pb pulmitate
Pb oleate
Pb phthioite

ether

extract

H+

Pb pulmitate
oleic acid
phthioic acid

cat.

stearic acid
phthioic acid

ether
►

extract

Pb stear-
ate

phthioic
acid

V-

J >■ v i ■._

....

•

'•

- 2 -

A mixture of fatty acids was also obtained from the wax frac-
tion. On distillation at low pressure two fractions were isolated;
a low "boiling one, which proved to be a single compound, was called
tuberculostearic acid, the high boiling fraction was shown to be
identical with phthioic acid*

Spielman oxidized tuberculostearic acid with chromic acid
and obtained n-octyl .ketone, azelaic and j^-octoic acids. He,
therefore, showed tuberculostearic acid to be —

H
CH3(CH3)7C-(CHs)8COOH

CH3 Q

H v nv,n CK3 (GH2) 7C-CH3

CH3(CH3)7C-(CH3)8C00H K^r°7

I H3S04 CH3(CH3)6COOH

H03C(CH3)7COOH

Attempt to determine the structure

Phthioic acid, when oxidized with chromic acid, gave a compound
with the formula C1:LH3303| which was claimed to be different from
n-undecoic acid because two derivatives melted 20° too low. This
"evidence is not thought to be conclusive in view of the small quanti-
ties used and the probability of mixtures.

Charguff synthesized a number of C36 acids and, from the fact
that the melting point of the n-hexacosanic acid is lowered 20-30°
from 88° by the introduction of one side chain, concludes that
phthioic acid must have at least three hydrocarbon chains.

Wagner-Jaureyg found by the Kuhn-Roth method of estimating
side chain methyl that phthioic acid gave 2.4 moles of C3H403 per
mole while 'tuberculostearic acid gave 1.4 moles of C3H4,0S per mole*
This is suggested as evidence for three carbon chains in the molecule,

Anderson believes that there is a methyl group in the alpha
position and another in the neighborhood of the eleventh carbon atom.

X-Ray reflections from multilayer films of the barium salt
showed that the length of the molecule was that of a chain of twelve
to fourteen carbon atoms.

Phthioic acid differs from known fatty acids in that it is in
thin films. On water it forme a very compressed unimolecular layer
collapsing at an area of ZQk0* per molecule. On the other hand, n-
decyl-n-dodecylacetic acid forms a much more expanded film collapsing
at about 6OA0 , The surface dipole moment- of this film was much
smaller than that formed by phthioic acid. Stenhagen suggests the
presence of a small alkyl group in the a-position to the carboxyl
because this might account for the observed close packing of the
chains. He suggested the formula —

&SftJ

0 i.

► i i : :■. .

I *

■

-. i ti .

■ : i.

. g . .. ; •

*. K

■

:.". ■' '

I ':'■ '

\ ■■:< U '■

• f.u

.■:

■.: v' '

•. - i - ;:

;.;Ci ■:•

3 >., v

J. v

■ :) ;•

- 3 -

CH3 (CHe) x
CH3(CH3)y __^CC03H

CH3(CH8)2^

in which x and y are about 10-12 and different and z = 0 or 1,
the most probable formula thought to be the one in which z = 1,
x = 9, y = 11.

Synthesis of Substances of the phthioic acid type

With the ultimate purpose of synthesizing phthioic acid, Birch
synthesized several compounds of the phthioic acid type. Methyl
di-n-octyl acetic acid was produced by application of the method
devised by Reichstein.

CeHi

C8H17-CC1 +

KMnO

v! >-C03Me
N0

Aici3 , S8^72c m '

CHa^ ^0^~C°3Me

CqEx 7 ^
CgHi? — ' — CCOgH
CH3 ^

This acid formed the compressed films of the phthioic acid type.

a,a-Dimethyl n-decyl acetic acid was synthesized according to
the directions of Haller.

NaNOs CH3 \
— 2* CH3~C-C03H

\\ /CH3

CH3N //0

C€H5C-C-CH3 + NH3Na -

-> CHa-C-C

C10H8i

C10HS1/ N

NHNa H3S04 C10H3"^
This acid also produced films of the type shown by phthioic acid.

••• i •

*- .-, ■ —

X
V i

■

'7 V

After several attempts £^-~^i-£^-octylbutyric acid was

synthesized.

0 CH3-COsH ethyl sodio

C8H17iS-CH3 GUareSCH CH3-9-.CH17 - ester - ester

chloride a-acetyi

CH2C0sH n-heptoate

Clemmensen

0 CH

hydrolysis , j 3 Clemmensen

reduction

CfiH

8n*7

CH3 (CH2) 7 \

CH3(CH2)7 CCH3C03H

This acid also gives films of the phthioic acid type.

CQn.cXu£..ion

., Tuberculosis and leprosy are caused by infection with
bacilli characterized by the possession of a fatty or waxy envelope.
In the case of leprosy, chaulmoogric acid and its derivatives prove
to be bactericidal. It has been shown that the bactericidal action
is not characteristic of any one group of atoms but rather due to the
physical character of the molecule. The bacilli probably suffer an
impairment of the fatty envelopes in the presence of chaulmoogric acid
and its derivatives. It has also been shown that the best results
were obtained with molecules containing 16 or 17 carbon atoms*
Molecules of the phthioic acid type synthesized by Robinson and
Birch, especially £, p-di-n-octyl butyric acid, are found to be
bactericidal to tubercule bacilli. From the above analogy it is
expected that the compounds of this type containing 16 or 17 carbon-- (
atoms may be even more active when synthesized.

Bibliography:

Anderson, J. Biol. Chem., 74, 525 (1927); ibid., 83, 169 (1929),

Anderson and Charguff, ibid,, 85, 77 (1929-30); ibid.. 106, 87 (1934).

Vl^Quff, Ber., 65, 745" (1932).

Wagner- Jaureyg, Z. Physiol. Chem., 247, 135 (1937).

Reichstein, Helv. Chim. Acta, 18, 271 (1938)*

Haller, Ann. Chim,, 1, 15 (19147.

Shriner and Adams, J. Am. Chem. Soc, 42, 2727 (1925).

Adams and Stanley, J. of Pharmacology and Exp. Therap., 45, 121-162

Robinson, J. Chem. Soc, 505-j(l940),

(1932)

Reported by R. G. Chase
October 29 f 1941

• -

. .

■

•£ >•;.:

^ « .......

t ■*' .

••• ! .

• ' -' .

" % ■- .'

. * ... ■ ' i. ■ . ,

THE OPTICAL ISOMERS OF CIST9-METHYL-1-DECAL0NE

Plentl, et al

In recent years the synthesis of steroids has been of special
interest, but the question of their optical isomerism has not been so
fully studied.

The ultimate aim of all of these syntheses was a comparison of
the synthetic steroid with its corresponding natural compound. Since
the latter, in all cases, are optically active the synthetic steroid
must be a pure substance and not contain stereoisomers impurities.

cis-9-Methyl-l-decalone 8 first reported by Chuang, was selected
as a raw material for study to obtain an isomer of definite composi-
tion. This compound has the advantage of possessing an angular group
at C9 which gives dissymmetry as well as being a general characteristic
of all steroids. From this compound, methods are available for the
preparation of both perhydrophenanthrenes and cyclopentenophenanthrenes.

Methylcyclohexenylbutyric Acid

The only methods reported for the preparation of 9-methyl-l-
decalone are dependent on the synthesis of methylcyclohexenylbutyric
acid; consequently the method previously worked out by Elliot and
Linstead was tried. In this a Grignard reaction is run with 5-
bromo-1-pentene and methylcyclohexanone with permanganate oxidation
of the tertiary alcohol, followed by dehydration of the hydroxy acid.

CH3=CHCH2CH2CH2Br -> CH2=CHCH2CH2CH2MgBr -+

CH3 OH

CH2CH2CH2CH:=Cri2 — ►

However, appreciable amounts of methylcyclohexanol were found in the
Grignard reaction, which is in agreement with the observation of
Butenandt that aliphatic magnesium bromides with moderately long side
chains have reducing properties. The dehydration of the hydroxy acid
also forms a spirolactone. This* need not be discarded but is simply
converted to ethyl methylcyclohexenylbutyrate in very good yields by
boiling with thionyl chloride in benzene solution and pouring the mix-
ture into absolute alcohol. ^"Lactones are usually split in this way
and apparently the procedure is equally as good with C -lactones.

Simultaneously with the above study a new synthesis of this acid
was attempted by Plentl and Bogert which proved superior in many ways.
This is shown in Flow Sheet A. Essentially the method is the elonga-
tion of the side chain of methylcyclohexenylacetic acid by two
successive Arndt-Eistert rearrangements. The substituted acetic acid
had previously been prepared by Chuang as follows:

(' I

1 -

. lii r

^ >

,-)

• ■ ..■ -.;/

OC 53

CKa 0 CH3 cHsC02Et ?Hs

BrCH2C02Et
Zn

cis-9-Methyl-l-decalone

- 2 -

Z_\ / ^SrCHBCOaEt

KHSO4

This preparation is also outlined on Flow Sheet A, and is
Essentially a cyclization of the acid by the Darzens reaction, as
modified by Cook and Lawerence, to the corresponding chloroketone
which after removal of the HC1 yields the unsaturated ketones.

Resolution

Only the purest fraction of cis.-9-methyl-l-decalylamine was used
for the resolution. Although Hueckel aid Kuehn were able to resolve
a-decalylamine using camphorsulfonic acid, this reagent was unsatis-
factory here because of the excessive solubility of its salts.
However, bromocaraphorsulfonic acid formed salts which were sufficiently
difficultly soluble in ethanol to be separated.

The amines were regenerated from these salts and allowed to re-
act with nitrous acid to give a mixture of a hydrocarbon and a partially
inverted alcohol. The hydrocarbon was probably Z^1'2 -9-methyloctalin
and was not further studied. The conversion to the alcohols seemed to
involve partial inversion at Cx, since the mixtures obtained from the
d- and 1-amines could not be brought to equal and opposite rotation.
However since oxidation of the mixture gave quite pure d- and 1-9-
methyl decalone, this indicated that the Walden inversion was confirmed
to Ox,

These facts are in agreement with Hueckel 's observations that in
compounds of this type partial inversion occurs, although he was
unsuccessful in isolating the alcohol. Further evidence is found
here that the compound is eis and not trans .

since in trans compounds of this type Hueckel showed that no inversion
occurs.

Bibliography

Plentl and Eogert, J. Org. Chem., 6, 669 (1941).

Elliot and Linstead, J. Chem. Soc.7 1958 f 660.

Butenandt, Cobler, and Schmidt, Ber. , 69, 448 (1936).

Cook and Lawerence, J. Chem. Soc, 1955 1637.

Hueckel and Kuehn, Ber., 70 2479 (1957).

Hueckel, Ann,, 53J3, 1 (1958),

Chuang, Tien, and Ma, Ber., 69, 1494 (1956).

Reported by J. D. G-arber
October 29, 1941

Uli-J

* ♦

t-...-^....'.<jt. ( ' ' • ■ *;.! ' »,«*«■ « '-- r .

.- ', ■ t *r • J • ' ■- »

-'-■'-■■'■ *

FLOW SHEET A
CH3

CH3

CH.

COCHN2
I
■CH2

CONH:

CH3
0

/n/s

. t '

•

'•'■■

. '

J.vj

FLOW SHEET B

HON
(Han

NOH

NHS

CH3

■»

Co*

NH:

CH3J

Bromocamphor sulfonate
[aD]HP0 + 75°

Bromocamphor sulfonate
[aD3H 0 + 59*2°

(~)cis-9-Methyl-l-decalyl-

amine hydrochloride

o
[a ] - 6.9

Stereoisomeric decalols

(-)ci8-9-Methyl-l~decalone

Bromocamphor sulfonate
[aD]H,0 + 68-8°

( +)jgis-9-Methyl-l-deca-
lylamine hydrochloride

£aD^H20 + 7-°

Stereoisomeric decalols

( + ) cis-9-Methyl-l«
decalone

CaD]Et0H " 3«9

^aD^Et0H + 4*2

:\ *.; . ;%>.:

r :

V' /

'"' ■'''■'>'

STRUCTURES OF PYRETKRINS I AND II

Pyrethrins I end II ere the active insecticidal constituents
of Pyrethrum flowers. It is important to determine their structures
so that attempts may be made to produce them synthetically.

Staudinger and Ruzicka did the first work on the structure of
the pyrethrins and arrived at formulas I and II for the respective
pyrethrins:

H ?Hs „ CH3

\i H H3 H H * 4 L

H2-C' V-C— C=C=C-CH3 Ha-C/f'V— C-C=C=C-CH.

II ■ II

H-C — C=0 K-C— C=0

0 0

c=o c=o

AH /GH3 (CH3)eC<C-Sg-C=C/ *"

(GH3)3C^0-C=CN _ H NC0CH:

H ^CH3 b

I II

Structure of the Acid Component s> — Staudinger and Ruzicka on
saponification of pyrethrin extracts were able to isolate but one
alcohol which they called pyrethrolone and two acids which they
named chrysanthemum monocerboxylic acid and chrysanthemum dicar-
boxylic acid corresponding to the acid portions of I and II. Some
chrysanthemum dicarboxylic acid monomethyl ester was also separated,
which indicates which -C00H group is attached to the pyrethrolone
nucleus* Ozonolysis of chrysanthemum monocerboxylic acid yielded
acetone and 1-trans-ceronic acid; ozonolysis of chrysanthemum di- .
carboxylic acid yielded 1- trans.- car onic acid and pyruvic acid,
thus establishing the structures of the acids as shown in I and II.
LaForge and Kaller discovered a third acid constituent whose formula
is C16H3002 but have not identified it*

Structure of the Pentenolone Nucleus 4 — The results of Staudinger
and Ruzicka indicated the structure shown in I and II for the cyclo-
pentanolone nucleus. LaForge and Heller showed that the analyses of
pyrethrolone, tetrahydropyrethrolone, and their semicarbazones
indicated two less H atoms than represented by their empirical for-
mulas. They reduced pyrethrolone (III) to tetrahydropyrethrone (V):

CH3 CH3 CH3

H2-C "C *" C5H7 p rr Hg— C C— C5H11 Hg— C C— C5H11

HOC C=0 HOC — C=0 C1C C=0

8 H H

III IV

tr. i

' 'i •

, ■ , ' ■; ■

j > ... ; ; f.j

1 •

:f . ••;

■i ;

' R

■ -

! •

I -

i ■ Si y ■
) -

• i > ■' ..•-*

CH3

/£%

Kg-^C C-CsHix

I / V

Hg-C — C=0

Mixed melting points showed the semicerbazone of tetrahydropyre-
throne (V) to be identical to thgt of dihydrojasrnone of known struc-
ture V.

structure of the SldecbeiEU — Studies of the structure of the
sidechein have attracted considerable attention, for if it has the
structure indicated by VI, it will be the first natural product known
to contain a cumulated system of double bonds.

Staudinger and Ruzicka postulated structure VI in the side-
chein from their studies. Lfter Ruzicka and Pfieffer abandoned VI
in favor of VII .

CH3

fT-T

H2HH pHHHH

H3-C XC-C-C=C=C-CH3 H3-C '0-0=0-0=0-0^

[12346 j j 1 3 3 4 S

HOG 0=0 HOC 0=0

I i

H H

VI VII

LaForge and Haller's early work was evidence against a. con-
jugated system but did not allow a definite choice,

Ozonolysis yielded acetaldehyde, establishing the position of
the 3,4-double bond. Further oxidation with hydrogen peroxide
yielded malonic acid which is evidence for the cumulated system.
Only in one instance was an acid of probable structure VIII isolated
in small yield from the ozonolysis.

CH3

H2-C C-CK2C00H

| I VIII

AcOC — 0=0

Treatment of pyrethrolone (VI ) with aluminum amalgam should
have given 1,4-addition if a conjugated system were present. Sub-
sequent reduction of the -OH group would have resulted in jasmone.
Reduction with aluminum amalgam yielded pyre throne (IX) however.
Pyrethrone on catalytic hydrogene tion furnished a tetrehydro deriv-
ative identical with dihydro jasmone. (Nuclear double bond is very
resistant to hydrogenation, )

•

• ■ . -t

■ •

,)■•
i. •

.-■•■•■• ■ .
-

•■:

; ,.

. . ..

..-";

■

■■

■>

^y 58

CH3 fH3

Hs-C C-C5H7 A1 tj H3-C C-C5H7

I I . A1 rtg > ! I

HOC C=0 H3~C — C=0

VI IX

Pyrethrolone and pyrethrone do not form characteristic products
with raaleic anhydride or ^naphthoquinone,

Pyrethrone absorbed one mole bromine readily in an indifferent
orgrnic solvent. If 1,4-addition had occurred, the product should
have yield_ed jasmone when reduced with zinc in acetic acid, but the
original pyrethrone resulted. Evidently bromine added on adjacent
cprbon atoms and no choice could be made.

Addition of one mole of bromine in ethanol solution yielded
p. monobromo compound plus nearly one mole of hydrogen bromide, Con-
sequently the bromine reaction was considered as one of substitution.
Addition of two moles of bromine gave a dibromo compound plus nearly
two moles of hydrogen bromide. Both bromo derivatives gave pyrethrone
when reduced with zinc. Only the sidechains are involved in that
tetrahyclropyrethrolone and tetrahydropyrethrone do not decolorize
bromine.

Pyrethrone was heated with sodium in a sealed tube and the
product was treated with carbon dioxide resulting in the formation
of an acid. This is typical of ^C=CxCHCH3.

No conclusion can be reached from the above results except that
no such behavior has been noted with a conjugated system. The liter-
eture furnishes no clues of the behavior of cumulated systems toward
halogens except that a lien e is stated to add four atoms of bromine
to give an unstable tetrabromide. Therefore, Acree and LaForge
preppred l-phenyl-l,2-butadiene, l-cyclohexyl-2,3-pentadiene, and
2,3-pentadiene to compare their reactions with halogens with those
of pyrethrone and pyrethrolone.

The above allenes did not react with maleic anhydride or
a-naphthoquinone (cf, pyrethrone above).

All three of the allenes mentioned gave dibromo and dichloro
pddition products with bromine and chlorine in indifferent organic
solvents in the cold as did both pyrethrolone and pyrethrone.

Conant and Jackson and later Jackson and coworkers have re-
ported that certain compounds with ethylenic linkage yield methoxy-
bromo derivatives (a) as well as the normal dibromo derivatives (b)
when treated with bromine in methyl alcohol*

(a) ^C=CC + ROH + Br s -» ^C— CC + HBr

6 Br
R

(b) ^C-£^ -h Br, -* ^C-CC

Br 3r

« •

4,-

; '

-4- q 59

Both reactions proceed at a. slow rate*

Both the dibromo and the methoxybromo products were formed when
each of the above mentioned alienee was treated with bromine in
methrnol. Liberated hydrogen bromide amounted to 60-70$ equivalent,
showing that reaction (a) predominates* In contrast to ethylenic
compounds, the allenes reacted instantly in the cold. Pyrethrone
reacted anrlogously in methanol to give a methoxybromo and a dibromo
derivative. Titration of liberated hydrogen bromide (60-70^ equiv-
alent) and methoxyl and bromine determine tions showed that reaction
(a) predominated. Reduction of the mixture of methoxybromo and
dibromo products with zinc regenerated pyrethrone in excess of the
amount expected from the dibromo derivative present. Evidently
zinc gives an analogous reaction with the alkoxybromo compound.
Analogy for this reaction is found in an article by Dykstra, Lewis
and Boord who reported the t a, p-alkoxybromo compounds are readily
dehalogenated by zinc to form a double bond between the carbon atoms
that carried the substituents*

All the above reactions of allenes and pyrethrone are com-
patible and the cumulated system of double bonds seems the most
likely arrangement in the sidechain.

LaForge and Acree have reported the reactions of allenes with
lead tetraacetate but have not reported comparable reactions for
pyrethrone yet.

Bibliography

Staudinger and Ruzicka, Helv. Chim, Acta, 7, 177 (i), 201 (II),

212 (III), 236 (IV), 245 (V), (1924).
LaForge and Ha Her, J. Org. Chem . , 2, 56 (1937); J.Am«Chem»Soc. , 58,

1061 (1936).
Ruzicka and Pfeiffer, Helv, Chim. Acta, 16, 1208 (1933).
LaForge and Haller, J. Org, Chem,, 2, 546 (1938).
Acree, F., Jr., and LaForge, J. Org. Chem., 5, 430 (1940); 6, 208(1941)

Reported, by D. W, Hein
November 5, 1941

* ■> I .'•.. ■■: '

■

,.f.

'. v

. ■ .y '!

'•:':'.> '■:/

'Iv -

;

i ♦

■t". :i''

■ ' ■'" <£ r '• ' ' •' i. ) ■'■

■ ... -'- '.'■

THE MICHAEL CONDENSATION: SOME RECENT INVESTIGATIONS
Connor, et al. University of Pennsylvania

The Michael condensation may be generally represented by
the equation

i i
-C^C-Li

+
L3CHL3

Piperidine
or NaCR

-C-C-Li
i

Lft—C—Lo

in which Llt L3, and L3 are labilizing groups. Examples have been
reported in which Lx is -COOR, -COR, -CN, -CONK3, ~N03, -S03R> and
in which L3, L3, or both, are -COOR, -COR, -CN, -CONH3, -N03,
-S03R, -CHO. The acceptor may be acetylenic rather than olefinic,
or it may be a quinone , Either the acceptor or addendum may be
vinylogs of these structures.

Influence of Experimental Conditions and the Structure of the
Acceptor on the Condensation

Connor's work has shown that secondary amines, for instance,
piperidine, are the safest catalysts; they seldom cause any reaction
other than normal condensation. However, amines often fail to •
bring about reactions that occur in the presence of NaOR, and the
rate is so slow, even in favorable cases, that a long reflux is
necessary. One-sixth to one-third of an equivalent of NaOR may
bring about condensation when amines do not; the use of one equi-
valent of NaOR is most likely to cause condensation, as well as
side reactions. With sodium alkoxides as catalysts, the best re-
sults are obtained by permitting the reaction mixture to stand at
room temperature for twenty to one hundred and fifty hours. Higher
temperatures may give lower yields, probably because side reactions
and retrogression are favored.

Nature of Lx x

Arrangement of groups in order of activation of double bond
is not generally possible but unsaturated ketone^ corresponding-
ester > nitrile. Examples in Table I.

TABLE I

Acceptor

Addendum

% Yield Conditions

1. 0CH=CHCO0

2. 0CH=CHCOOEt

3. 0CH=CHCOOEt

4. 0CH=CHCN

5. p_-O3NC6H4CH=CHCO0

0CH3COOEt
0CH3COOEt
0CH3COOEt
0CHaCOOEt
CH3(COOEt )s

6. p_-03NC6H4CH=CHCOOEt CH3(COOEt)

.,■ -

-v jz

■

; ■■

>'')

; ■■ ■'■ ft

■

•'...V

•

; ui

1 j , v

- 2 -

Conditions

A. Piperidine catalyst, long reflux

B. Small amount of sodium alkoxide, stand at room temperature

C. Equivalent amount of sodium alkoxide, hot.

Substitution on a and g atoms

(1) Reactivity of acceptor decreases as hydrogens are replaced*

(2) Reactivity decreased if substituent is alkyl, aryl,
carbethoxyl or acyl.

Typical results are shown in Table II.

TABLE II

Acceptor

Addendum

% Yield Conditions

1. 0CH=CHCO0

2. 0CH=C(COOEt)CO0

3. 0CH=CHCOOEt

4. 0CH=C(0)COOEt

5. CH3CH=CHC00Et

6. CH3CH=C(CH3)C00Et

7. (CH3)2C=CHC00Et

Remote Substitution

0CH2CCOEt

0CH2COOEt

0CH3COOEt

0CH2COOEt

0 CHgCOOEt

0CH3COOEt

Groups not directly attached to the double bond of acceptor
may have greater effect than would be expected (Cf. Table III).

TABLE III

Acceptor

Addendum

% Yield Conditions

o-03NC6H4CH=CHC00Me
m-02NC6 H4 CH=CHC00Me
p_-02NC6H4CH=CHC00Me

0CH=CHCO Mes

CH2(C00Me)2
CHa(C00Me)a
CH2(C00Me)3
CH2(C00Me)2

Instability of some Addition Products

In some cases, there is an unusual difference in reactivity
between a substituted active methylene compound and the next higher
homolog.

I : i !'

■

'. ,-i

■ . ■ : , ..i

".'J i

U • :

*K\ .. :■ .

"■. r

.1 ■ " s ' . ."si

.iV: '

- • '

■' : 1 ! » ".:

-3-

0CH=CHCO0

+
RCH(CO0Et)g

N fcOEt

0CHCH3CO0
R-C(COOSt)3

I R = CH3- ou/o
II R = C3H5 0%

If the difference is due to spacial interference, II, once
prepared, would be expected to be stable. An addition product of
a Michael condensation was alkylated, as a possible synthesis of IL

0CHCH3CO0
CH(COOEt)3

III
A

0CH=CHCO0

+
CK3(COOEt )a

0CRCH2CO0
C3H5C(COOEt};

/|\

ir i

0CH=CHCO0

+
C3K5CH(C0OEt)3

None of compound II was i sole ted even at -78 . Steric hin-
drance is improbable when one considers the results of Connor and
Andrews obtained by the reaction of the sodium derivative of ethyl
ethylmalonste with benzalacetophenone to give ethyl cc-ethylcinnama te
and ethyl benzoylacetate which must be the results of a Michael
condensation.

Thus the data indicate ttu t the expected product is so
readily cleaved by NaOR that isolation is impossible and steric
hindrance does not prevent reaction.

Activation of the Methylene Group by Carbon -Carbon Unsaturation

The possibility of L2 and L3 both being aromatic or olefinic
was investigated by Connor. The reactivities of fluorene, cyclo-
pentadiene and 1, 4-pentadiene were studied.

Fluorene reacted with benzalacetophenone, benzal-p_-bromo-
acetophenone, and benzalacetone in the presence of one equivalent
of sodium ethoxide (yields 2-27%). No reaction occurred with
a, p-unsaturated esters or m- or p_-nitrobenzalacetophenone.

0CH=CHCO0

CH3

NaOEt>

0CHCH3CO0
CH .

Cyclopentadiene reacted with a, p-unsaturated ketones when
piperidine was used as a catalyst. This indicates that it is a
highly reactive compound. The reaction ves carried out under
pressure to prevent the loss of the hydrocarbon (yields 25-30%).

■y

-4~ 63

0CH=CHCOC6H4Br(p_) 0CHCH3-COC6H4Br (p_)

. + I

^CH^. — ■ > /GHX

CH CH (CH3)5NH CH CH

I. II |! !l

CH CH CH— CH V

Pentediene-1,4 reacted with benzal. p.~bromoacetophenone in
the presence of en equivalent amount of sodium ethoxide. The same
compound was obtained by using the sodium derivative of the diene.

0CH=CHCOC6H4Br(p_) Ich^CH^CH^ > 0CHCH3COC6K4Br (pj

CH3=CH-CH-CH=CH3
(CH3=CH)3CHNa , vi VI

Bibliography

Andrews and Connor, J. Am, Chem . Soc, , 57, 895 (1935); 5_6, 2713 (1934)

Connor and McClellen, J. Org. Chem,, 3, 570 (1939).

deBenneville, Clagett rnd Connor, ibid"., 6, 690 (1941).

Ingold, Perren and Thorpe, J. Chem, Soc.,^119, 1976 (1921)*

Ingold and Powell, ibid,, 121, 1771 (1921).

Taylor and Connor, J. Org/ Chem., 6, 696 (1941).

Reported by R. E. Foster
November 5, 1941

■?

REACTION OF ALKYL BENZOATES WITH SODIUM aLKOXIDES
McElvain, et al, University of Wisconsin

A study of the products formed when alkyl benzoates are heated
with the corresponding sodium alkoxide led McElvain at 'Wisconsin to
suggest that a reverse Tischtschenko occurred. The "benzaldehyde and
aliphatic aldehyde or ketone formed may then take part in one or
several of the following reactions:

1. Forward Tischtschenko

2. Mixed aldol condensations followed by loss of CO to give
alkyl phenyl carbinols which in turn give ketones.

3. Further condensation of the ketone, as acetophenone, with un-
reacted ester to give 1,3-diketones,

4. Saponification of the ester by water formed in various conden-
sations.

5. Condensation of the alcohol or its ester with sodium alkoxide
through the Guerbet reaction.

6. Acyl exchanges between esters and 1,3-diketones.

Adickes had treated ethyl benzoate with sodium ethylate and ob-
tained not only more sodium benzoate than expected but also a 20 per
cent yield of dibenzoylmethane. McElvain found that heating 4 moles
of the ester with one mole of the alkoxide at 175-180° for two hours
gave a maximum yield, 40$, of dibenzoylmethane from ethyl benzoates.
These conditions were then employed for treating the methyl, ethyl, '
propyl, icopropyl, butyl, isobutyl, and neopentyl esters. Table I
is a summary of most of the results.

The formation of benzylbenzoate as a common product is cited as
evidence that although the reactions are apparently quite different
they may follow a common initial course, a reverse Tischtschenko, and
the benzaldehyde formed could then give' benzylbenzoate ■ :

by a forward Tischtschenko. Reverse Tischtschenkos with neopentyl
benzoate and isobutyl benzoate would give trimethylacetaldehyde and
isobutyraldehyde. These aliphatic aldehydes have either none or only
one alpha hydrogen and might give esters hy forward Tischtschenko
reactions. Neopentyl trimethylacetate was found in good yield and
enough isobutyl isobutyrate was formed to indicate that isobutyralde-
hyde might have functioned in a forward Tischtschenko.

If ethyl and propyl benzoates took part in a reverse Tischt-
schenko reaction the resulting aliphatic aldehydes would be capable
of entering into aldol condensations with the benzaldehyde. The
following mechanism is proposed to explain the origin of a 40 per cent
yield of dibenzoylmethane from ethyl benzoates.

C6H5C00C2H5 -> C6HBCK0 + CH3CH0

C6HsCH0 + CH3CH0 -* C6HSCH0HCH2CH0 ~C0 > C6HsCH0HCH3

C6H5CH0HCH3 -^22 ► C6H5C0CH3

In a similar manner propiophenone would be formed from propyl
benzoate.

;r .

■i ■

■

■ .

■■ 'J ,' ;; '«

''ri "<5i*5fli.4fi

5/Vi

* 2L-

Dibenzoylmethane, produced by the ethyl ester, could have been
formed by a Claisen condensation,,

C6H5COCH3 + C6H5COOC3H5 Na0CsH5 > (C6H5CO)2CH2

It is recognized that the loss of carbon monoxide by an alpha
hydroxyaldehyde is without precedent in the literature. Those esters
giving high yields of the carbinol also gave high yields of carbon
monoxide. Trimethylacetaldehyde formed in the neopentyl benzoate re-
action could not give an aldol condensation to form an alpha hydroxy-
aldehyde and this reaction did not liberate carbon monoxide.

The results from isopropyl benzoate are of particular interest be-
cause the acyl exchange reaction suggested proved to be of considerable
importance when studied further.

C6H5COOCH(CH3)3 NaOCH(CH3)2 ? ^^0 + (CK3)2C0

C6HBC00CK(CH3)2 Claisen ) C6h5cooCH3COCH3

C6H5C0CH8C0CH3 Acyl jlxchanSe , (C6H5C0)3CH2 + CH3C00CH(CH3 ) 2

The presence of 2-methyl-l-pentanol among the reaction products
of the n-propyl ester and of 2-ethyl-l-hexanol from the n-butyl ester
can be explained as being formed by the G-uerbet reaction.

Acyl Exchanges Between Esters with 1.5-Diketones and Esters with g-
Keto Esters

The acyl exchange reaction mentioned above was investigated in
more detail. If the reaction proposed is correct it should be possible
to prepare 1, 3-diketones and {3-keto esters by such reactions as follows

(1) C6H5C0CH2C0CH3 + C6H5C00C2K5 -> C6H5C00CH2C0C6H5 + CH3C00C2H5

(2) CH3C0CH3C0CC3H5 + C6H5C0CC3H5 -* C6HeC0CH2C00C3H5 + CH3C00C3H5

o
Attempts to obtain the exchange products by heating at 130-160
with sodium ethoxide caused the ester, e»g« ethyl benzoate, to take
part in the reverse Tischtschenko and subsequent reactions to such an
extent that the acyl exchange was obscured. When the less basic
sodium enolates of the 1,3-diketone or j3~keto ester were employed good
results were obtained. The sodium enolates were heated with a
sufficient excess of the ethyl ester to form a homogeneous reaction
mixture. The temperature employed was sufficiently high to permit the
ethyl acetate formed to be removed by distillation. Results of the
exchange are shown in Tables II and III.

In Table II it is to be noted that there are good yields and
correlations between yields in runs 1,4, and 5 where the acetyl group
is replaced by benzoyl and a-furyl. Since js-chlorobenzoyl did not
replace benzoyl in run 7 it is believed that the d-j>-chlorobenzoyl-
methane formed in run 2 (see footnote a) was formed by the reverse
Tischtschenko and subsequent reactions.

The yield obtained in run 1 of Table III suggests a useful method
of preparing acylacetic esters provided the esters employed have a

•,

, • . ' -.. -....> I,

r r

! . >•:-''!

~ 3 -

boiling point sufficiently high to permit the ethyl acetate formed
to distil from the reaction mixture*

A mechanism is proposed for this exchange which accounts not
only for the acyl exchange but also explains the origin of considerable
alcohol which always distilled along with the volatile ester*
A0 Na+ ^O-Na

CSH5CT + :CHCOC6H5 ► C6H5C-~

■OEt

COCK.

CH3COOEt + C6H5C(ONa)=CHCOC6H

OEt

•CHCOC6H5
COCH3

^ COCH3
C6H5C(ONa)=C-COC6H5 + EtOH

Intermediate I formed through a carbanion mechanism could decompose
by path (a) to give the acyl exchange product II or by path (b) to
give ethyl alcohol. Evidence is lacking for the formation of a
triacyl methane, II,. but it may have been hydrolyzed into a diacyl
compound and an acid. Acids roughly equivalent to the alcohol dis-
tilled were separated from the mixture.

Under somewhat different conditions ethyl isobutyrate and ethyl
benzoyldimethylacetate form ethyl benzoate and ethyl isobutyryliso-
butyrate. The net result is an acyl exchange. This reaction pro-
ceeds in the presence of sodium ethoxide and triphenylme thane at
ordinary temperature and is explained by Hauser as a series of forward
and reverse acetoacetic ester condensations.

TABLE I

Ratio of Moles Product to Moles Unrecovered Ester for the Reaction
of Four Moles of C6H5C03R with One Mole of RONa at 175-180°

Iso-

Ethyl

Prolyl

Propyl

n- Butyl

Isobutyl

Neopentyl

Unrecovered

Ester, Moles

2,10

2.60

1.75

3.08

3.00

2,10

Benzoic Acid

.26

.35

.34

.87

.34

.41

ROH

.08

.60

1.00

,41

.47

.40

1,3-Diketone

.19

. » • .

»09a

.05^

%■'■•••

...

• * »

Esters from

Tischtschenko

,07°

Trace

.06°

.013°

.05°

.07°

Reaction

.10d

.06|
.46*

RCH0HC6H5

• • •

.08

• • •

.11

.06

.12§

.138

..09S

• • *

.27

.41

• • •

.34

• • •

Guerbet

Products

• • •

.015
.09S

t f f

.03
.08&

• • •

Cl ■

'■• L -

3 r

-j.'.(

j - »

r-'ifl

■'>••' ;

' »

'••"•■■■

I '; • , '

.->

Table I (Conf)
aC6H5COCH2COC6K5
^CexHBCOCHsCOCHa
cBenzyl benzoate

- 4 -

Isobutyl isobutyrate
eBenzyl trimethylacetate
^Neopentyl trimethylacetate
^Benzoic ester of

Run

TABLE II

Acyl Exchange Between the Sodium Enolate of 1,3-
Diketones and Esters

RCOCHNaCOR' + R^COgEt -

R»

C6H5
p_-C6H4Cl

CsHgCHa

C4H30 [a-
C4H30
CH3
jp.-C6H*Cl

(a) a 4:0% yield of ethyl benzoate was also obtained

(b) a mixture of CsH5COCHaCOC6H4Cl and C1-C6H4C0CH2C0C6H4-C1
was obtained. Yield not given.

C6H5

CH3

Q«HB

CH3

C6HS

CK3

C6H5

CH3

C6H5

C6K5

RCOCHNaCOR"

+ R

fC03Et

% yt<

3 Id 1

R'C02Et

RC0CH2C0R"

49*

'uryl)51

32°

TABLE III
Acyl Exchange between the Sodium Enolate of p-Keto-Esters and Esters,

RCOCR'NaC03Et + R"C02Et

R"C0CR'NaC02Et + RC02Et
% yield of

.un

R" R!C02Et

R"C0CHK,C

02Et

CH3

C6H5 56

49*

2
3
4
5

CH3

(CH3^CH
CH3
CH3
C2H50

C2H5
C2H5
H

H
H

C6H5 10a

C6H5 72

C4H3O (a-Furyl) 66

C8H4N (|5-Pyridyl)73

C6H5 10

61b

(a) yield of \0% ethyl acetate and 60* ethyl butyrate

(b) This product was ethyl benzoylacet&te
.(c) Main product was a non-distillable tar.

Bibliography :

Magnani and McElvain, J, Am, Chem. Soc, £0, 813 (1938).
McElvain and Weber, ibid, t 63, 2192 (1941)7
Adickes, Mullenheim, and Simeon, Ber., 66B, 1904 (1933).
Hauser and Hudson, J. Am, Chem. Soc, 62, 62 (1940),

Reported by G, L* Schertz
November 12. 1941

- 8

•••••' .•;

*£:

■' i.. •

■'■ Jb *-. >'- ■•-; - .,- ■•

• - .

THE STRUCTURE OF LIGNIN

Hibbert - McGill
Freudenberg - Heidelberg

Brauns- Inst, of Paper

Chemistry
Adkins - Wisconsin

Lignin is a constituent of the woody portion of plants. Besides
b0-5b% of cellulose, wood contains 2£-3C$ of lignin, 15-20$ of
sugars and other low molecular weight carbohydrates, and a few per
cent of resins, fats, and proteins. ; Since wood contains so muqh lig-
nin the structure and properties of lignin have been studied exten-
sively by the paper industry, both with a view to finding more
efficient ways of separating it from cellulose and to discover uses
for it after it is removed.

Lignin is not readily
being chemically combined w
or glycoside linkages, and
cellulose. Rather drastic
heating a number of hours w
chloride or various acids,
doubt but that the amorphou
differs from the lignin as
the nature of lignin varies
and its method of isolation
nin is a highly complex sub
and controversial one.

separated from the other wood constituents,
ith the lower carbohydrates through ether
the whole being intimately mixed with the
chemical methods are necessary, such as
ith sodium bisulfite, alcoholic hydrogen

Whatever method is used, there is little
s brown solid which is finally obtained
it originally occurs in wood. Moreover,

quite a bit depending both on its source
For these reasons and the fact that lig~

tance, the lignin problem is a difficult

In spite of this, most authorities seem agreed as to the essentia}?
structure of lignin, differ though they may in some details. Just as
a protein may be regarded as a condensation product of amino acids
and a polysaccharide as a condensation product of monoees, so lignin is
regarded as a condensation product of a number of closely related
aromatic compounds. The units are believed to be derivatives of
phenylpropane. Typical examples are:

I RCOCHOHCH, II RC0C0CH3

R is guaiacyl (IV) or syringyl (V)

CH,0

III RCH=CHCH2OH

OH
CH— C-CH:

CHOH-CO-CH3

These units are thought to be joined
together as in VI,. Most of the evidence
indicates that this explanation is funda-
mentally correct and that lignin is comprised
of structures such as VI and polymers of VI.
Unpolymerized compounds like I, II, and III may
also be present.

The proof of the above hypothesis rests upon analytical data and
the general reactions of lignin as well as upon the products obtained
by the degradation of lignin by caustic fusion, oxidation,

o -f-

1 «

. t

- 2 -

hydrogenation, sulfonation, and alcoholysis. Studies of known sub-
stances with structures similar to that assumed for lignin have also
proved helpful.

Analytical Data and General Reactions

Lignin has a methoxyl content of around 16$* It has an hydroxyl
content of about 10$, Color reactions and a certain acidity indicate
the presence of phenol hydroxyl groups, but while dimethyl sulfate
almost doubles the methoxyl content of the lignin, diazomethane in-
creases it by only about 3$. This is taken to mean that most of the
phenol hydroxyls of the guaiacyl and syringyl radicals are tied up in
ether linkages as in VI.

Lignin can be chlorinated, bromlnated, nitrated, sulfonated, and
mercurated in a way reminiscent of sensitive aromatic compounds. The
absorption spectra of lignin sulfonic acid indicate an aromatic struc-
ture. A positive haloform reaction shows the presence of CH3-C0-
or CH3-CHOH-.

Brauns has isolated a lignin by merely soaking wood in alcohol
for several days. On the basis of analytical data, he assigns it the
formula C42H3306 (0CH3)4{0H)4 (CO) and says that it contains one phenol
hydroxyl and an enolizable carbonyl group. This he believes is the
fundamental building stone of lignin, an opinion not shared by Hibbert.

Degradation by Oxidation

Methylation of phenol groups with diazomethane, treatment with
alkali to break ether linkages, another methylation, and finally oxi-
dation with permanganate, gives veratric acid ,VII* / isohemipinic
acid VIII, and trimethylgallic acid IX. The yields are very low,
which is attributed to the fact that the acids themselves are not very
resistant to permanganate. The isolation of isohemipinic acid is
good evidence that the side chain of one phenylpropane unit has con-
densed with the benzene ring of another.

COOH

0CH3
VII

HOOC

0CH3
VIII

0CH:

All three of these acids are obtained from hard wood lignin, but
only veratric acid and isohemipinic acid are obtained from soft wood
lignin. This is in line with other facts indicating that the guaiacyl
radical IV is the only important radical in soft wood lignin, while
both the guaiacyl radical and the syringyl radical V are important in
hard wood lignin.

I '•'

.-vj •;- .1

•;

( i .

■

~ 3 ~

Degradation by Hydro aenat ion

Adkins recently reduced lignin catalytically. His lignin was ex-
tracted by the relatively mild reagent methanol-dry hydrogen chloride.
The lignin took up more hydrogen than a corresponding weight of ben-
zene, and besides a large amount of methanol, a 40$ yield of X, XjC,
and XI J, all cyciohexylpropane derivatives, was obtained. Unidentified
compounds of higher molecular weight were also found*

CH3CH3CH3 CHSCH3CH3 ^CHsCHgCHsOH

OH
X

OH
XII

When he hydrogenated lignin obtained industrially by the rather
severe soda process, he isolated small amounts of X and XII plus
some cyclohexanol, 4-methylcyclohexanol, and 4-ethylcyclohexanol.
However, most of the products were alcohols and glycols related to
polycyclic hydrocarbons with 20 to 70 or more carbon atoms in the
molecule. None was identified. Ke concludes that lignin isolated by
mild means is made up of phenylpropane units joined in chains, and
that the soda process causes considerable cyclization, giving mole-
cules very stable to hydrogenolysis. This is good evidence that lig-
nin has a skeleton of many carbon atoms.

Degradation by Bisulfite and Alkali

Lignin sulfonic acid from the sulfite liquor of paper manufacture
when treated with alkali in the presence of m-nitrobenzenesulf onic
acid, gives as high as a Ab% yield of vanillin, XIII, and syringic
aldehyde, XIV. This process is actually being used commercially to
prepare vanillin.

Small amounts of acetovanillone, XV, and acetosyringone, XVI,
were also found.

CHO

x J-0CH3 CH30

OH
XIII

CO-CH3

OCH3 CH3O

OCH,

Degradation by alcoholysis

Refluxing wood with et handle hydrogen chloride, a relatively
mild method of isolating lignin, gives a quantity of water soluble
products, some of which have been identified recently by Hibbert as
vanilloyl methyl ketone, XVII, syringoyl methyl ketone, XVIII, a-
ethoxypropiovanillone, XIX, and a-ethoxypropiosyringone, XX.

»/•;•

■

. ■ •■. ■ • '• • ; - lOV:

; • -

»-* ' t • ■ -

-f ti ' .'

. . .<

.1 ,. .1

■

■ . ■■ ■■'■;%■

■* ' ' ...

#■,».' . . I, , ' * J,

'-";• •• : , , J.j. .

■ /.:..■

* « «.

-C0-CH3

CO-

-CO-CH3

,CO-CHOEt-

1 I

-CH

3 CO-

-CHOEt
CH3

1
T

^OCHs

CH30/^fNOCH3
OH

kJ-OCH3.
OH

c*°

W^OCHg

xvi :

XVIII

XIX

This is further proof that lignin is built up of phenylpropane
units such as I, II, and III (this was originally postulated on some-
what theoretical grounds) and is the first good indication of the
structure of the side chain of the phenylpropane units.

The ethoxy group in XIX and XX probably came from the ethyl
alcohol with which the wood was refluxed.'

Studies of Lignin Models

ri'he acid XXI has a structure very similar to that postulated for
lignin. When heated with alkali, methylated, and then oxidized with
permanganate, it gives the same products as does lignin under the
same conditions (veratric acid, VII, and isohemipinic acid, VIII).

Coniferyl alcohol, XXII, which occurs in plants as the glucoside
feritl, is readily condensed by a trace of acid to an amorphous
iubstancs similar to lignin. Thus this substance gives the same oxida-
tion products, reacts with bisulfite to give similar sulfonic acids,
and has about the same composition. There seems to be a definite
relationship between coniferin and lignin, but so little is known about
the life processes of plants that it is hard to say just what it is.
Likeivise, although there is no lack of theories regarding the forma-
tion of lignin and its possible function in' the plant, not much can
be said with any degree of certainty.

coni^fc

CH,0

m3o<f >-CH CH~CH3

-n

CH30

CH=CHCH3OH

■COOH

XXI

OH
XXII

-OCH3

Bibliography:
Phillips, Chem.

Revs., 14, 103 (1934).

Harris, D'J^nni, and Adkias, J. Am. Chem. Soc, 60, 146? (1938).

Freudenberg, {Inn. Rev. Biochem., 8, 88 (1939).

, 61, 2120 (1939).

J. . Am. Chem. Soc. ,

Hibbert, Paper Trade Journal, July 24, 1941. ■

Brauns
Adkins

Am* Chem. Soc,
Frank, and Bloom,

63, 549 (1941).

Reported by B. McKusick
November 12, 1941

, •

*;• ■

••,■••

- i
■■*■".

♦'

- .ti?

: j

......

.. ■','>'■ ,l\

SYNTHESES IN THE TRIPHENYLENE SERIES

Until very
phenylene series,
pared. However,
ties of derivativ
prepare compounds
Several useful me
have been issued
for the preparati
dye intermediates
the chemistry of

recently little attention had been paid the tri-

and few derivatives of triphenylene had been pre-
following the discovery of the carcinogenic proper-
es of 1,2-benzanthracene , an effort was made to

in the triphenylene series isomeric with these,
thods of synthesis have resulted. Moreover, patents

within the past few years for several processes
on of triphenylene derivatives suitable for use as

Thus, there is evidence of increasing interest in
the triphenylene derivatives.

Triphenylene, the parent compound of the series, occurs to the
extent of one to three per cent in the chrysene fraction of coal tar.
The first useful synthesis of triphenylene was that developed by
Mannich in 1907, although the formation of small amounts had previous-
ly been reported in the pyrolysis of benzene, and in treatment of
bromobenzene .with sodium. Mannich prepared it by heating cyclo-
hexanone with a 30$ solution of sulfuric acid in methanol. The re-
action is apparently exactly analogous to that in which mesitylene is
formed from acetone:

H3SO4

CH3OH

^ f

HsS04
CHaOH

450 - 50(

^ .0 <

/»

The product of the condensation, 1, 2, 3,4, 5, 6, 7, 8, 9, 10, II, 12-dodeca-
hydrotriphenylene (I), is obtained in a yield of about Sf. By passing
the vapors of this product in an atmosphere of carbon dioxide over a
copper catalyst at 450-500C, Mannich dehydrogenated it almost quanti-
tatively to triphenylene (II). Mannich assigned these compounds their
correct structure on the basis of their oxidation to mellitic acid
by fuming nitric acid in a sealed tube.

A modification of this method of preparation has recently been
patented. Cyclohexanone or 1-cyclohexylidenecycylohexanone is con-
densed to the same dodecahydrotriphenylene by heating under pressure
in the presence of a dehydration catalyst and a rare earth oxide.

Most of the recent work on the triphenylene series concerns
the synthesis of derivatives by ring closure methods which could
leave no doubt about the position of substituent groups. Com-pounds
to be tested for carcinogenic activity must, of course, be of known

_ o _

structure. Moreover, it was considered desirable to obtain reference
compounds which could be used to determine the structures of new com-
pounds in this series.

A few general methods have been developed for syntheses of
this type. The first of these is that introduced by Bergmann and
Blum- Bergmann, and subsequently extended and improved by Fieser
and Joshel. Triphenylene , the 1- and 2-methyltriphenylenes, and 1,2-
climethyltriphenylene were prepared by the series of reactions shown
in flow sheet A, starting with 9-bromophenanthrene. The percentages
given under the arrows indicate the yields obtained.

Bachmann and Struve have prepared 1-methyltriphenylene by a
similar series of reactions. Their synthesis, however, began with
the preparation of £-[9~(l, 2,3,4-tetrahydrophenanthroyl )] propionic
acid by a Friedel-Crafts reaction between l,2,354--tetrahydrophenan~
threne and succinic anhyaride. Phenanthrene itself could not be
need, since, in the Friedel-Crafts it gives substitution mainly in
the 3-position and to a lesser extent in the 2-position, not in the
'^••position as desired here.

To prepare 1,4-dimethyltriphenylene, Fieser and Joshel were
forced to adopt a different scheme after attempts to add the methyl
Grignard reagent to the carbonyl group of |3-(9-phenanthroyl)
propionic acid and its methyl ester had failed. The method finally
employed is shown in flow sheet B. It involves an unusual aldehyde
synthesis.

A novel method of forming the triphenylene skeleton is that
investigated by Bergmann and Bergmann. When various 9-vinylphenan-
threnes were treated with maleic anhydride, a Diels-Alder reaction
occurred. The necessary 1,3-diene system was furnished by the vinyl
group of the side chain and the 9,10-bond of the phenanthrene
nucleus. Examples are shown in the following equations:

CH— C

II ^0

boil

toluene
\^ {U% yield)

- 3 -

„0

CH-— C '
+ H ^

CH-— C'

boil

toluene
\ (58$ yield)

The most recent contribution to the field is the method worked
cut by Rapscn. As in Mannich's original synthesis of triphenylene ,
the starting material was cyclohexanone . This was condensed to 1-
c .■/,;:1.chexylidenecyclohexanone in almost quantitative yield when
saturated with dry hydrogen chloride and allowed to stand. When this
v :'->ciuct is treated with arylmagnesiumbromides, carbinols result
Vuich can be cyclized to triphenylene derivatives. The equations in
tlow sheet C show the syntheses accomplished by Rapson using this
procedure. Yields in the cyclization step seem to have been low in
general, although they were not reported in all cases. This scheme,
however, involves fewer steps than the other general methods, and
gives more convenient access to some derivatives.

Within the last six years a number of processes for making
derivatives in this series have appeared in patents, 2-Triphenylene-
eulfonic acid is made by the sulfonation of triphenylene in nitro-
benzene solution. 2-Hydroxy-3-triphenylenecarboxylic acid is made
by treating the alkali salt of 2-hydroxy triphenylene with carbon
rdoxide at high temperature and pressure. Various anilides of this
acid are made by condensing it with substituted anilines such as the
toluidines, the anisidines, and the chloroanilines. Triphenylene-
carboxylic amides can be made directly by a Friedel-Crafts reaction
between triphenylene and two moles of carbamic chloride. This
method can be applied to other compounds having three or more con-
densed aromatic rings, or two benzene rings condensed with a hetero-
cyclic ring. All of these products have been patented as dye inter-
mediate s.

Bibliography:

W.nich, Ber., 40, 159 (1907).

Bergmann and Blum Bergmann, J. Am. Chem. Soc., 59, 1441 (1937).

Fieser and Joshel, ibid. . 61, 2958 (1939).

Bergmann and Bergmann, ibid., 59, 1443 (1937),

Rapson, <T. 'Chem. Soc, 1941. 15.

lei--. Farbenind. A.G,, French Pat. 790,565 (1935); Chem. Abs. 50.

2994 2 (1936).
I.G, Farbenind, A.G., German Pat. 654,283 (1937); Chem. Abs., 32,

23722 (1938).
1,0-. Farbenind.A.G., German Pat. 654,715 (1937); Chem. Abs., 32,

36301 (1938).
I.G. Farbenind. A.G. , French Pat. 799,598 (1956); Chem. Abs., 50,

7585* (1936).
I.G. Farbenind, A.G,, French Pat. 797,072 (1956); Chem. Abs., 30,

6589 7 (1936).

Reported by P. L. Southwick, November 19, 1941.

■

_ 4 -
FLOW SHEET A

/XCH9 zn;Hg \

KC1

•d

■H

•H

ULI

cti

^CHs (79jg)
C02H

/CH2

^CHS

dry HF
0°

VII

/CHs (87#)
^ C02H

Zn-Hg
HC1 .

III

s
p,r?o°

la'

2200

CH3 MgCl

320

-H20

(42$ yield in I
last three steps) CH3

./^

-H'

- 5 -
FLOW SHEET B

•V

MgBr CH3OCH3CN
(66#)

CH<
I
, CH

CHO

III

0
G

k G
yy\/ XCH2-OCH;

/NaHSOa
<\

KHSO*
^180°
\

(58Jg)

CHgMgCl

(95%)
^CH3

^CHsOCHg

CH3'C03H)3
Pip .ridine

(eof)

CH;

CH
/
C03H

y <

CH3

VII

Adams ' catalyst

S
250°

-Hi

Uslm hf _^

(B2%)

(. 2202

' -Hs0

(35$ over-all yield
in last three steps)

- 6 -

FLOW SHEET C

^~^s

CS2

AICI3

0° (19$

VIII

V\A

0°

/,-

\\R

III

cs2
A1C13

( 29 fo when
v R = CH3O)

Pd + C
300o

(quant. )

Pd + C
300°
(quant. )

(R = CH3 or CH30)

HIGH OCTAINE AVIATION FUELS

It has been known for many years that hydrocarbons of
different structures have varying efficiencies as fuel for the internal
combustion engine. Thus, gasoline obtained by direct distillation
of petroleum, because of' its tendency to knock, can be used only in
engines of low compression ratio. The power output and efficiency
of the engine, however, improve as the compression ratio is increased.
In :i s'cudy of the antiknock qualities of pure hydrocarbons of the
gasoline range, it has been found that highly branded hydrocarbons
tire superior fuels. Among the best is isooctane, while n-hepfcane
is one of the poorest. In standardizing gasolines the fuel is compared
30 a mixture of isooctane and n-heptane.

The cracking process has brought a great improvement in the
auality of gasoline. The unsaturated hydrocarbons produced have a
I. rod antiknock rating, and the cracking is accompanied by isomeriza-
ticn to more highly branched compounds. New processes for the
production of airplane gasolines of very high quality involve the
i".1i Liza lion of gases from the cracking units. The first of these
\.ec: developed for the production of isooctane. When isobutylene is
r.c-Gsed over sulfuric or phosphoric acid supported on inert material
it .reacts with itself to give a mixture of isooctenes. These are hy-
drogenated in the presence of a catalyst and isooctane is obtained.

."uw Temperature Sulfuric Acid Alkylatlon Process

A valuable modification of this process consists in a similar
ureatment of the fraction containing all the four-carbon hydrocarbons.
Isobutylene reacts preferentially with the butylenes. At the same
time isobutane adds to the butylenes yielding saturated hydrocarbons.

H SO
Isobutylene + butylene — ~ — $~* ootenes

Isobutane + butylene HasQ4 > octanes

The last reaction is accompanied bv the formation of higher and lower
hydrocarbons (pentanes and decanes). Since n-butane does not enter
into the desired reactions, it is converted to isobutane and
butylenes by the following processes.

n-butane ^^ ► butylenes + H2

oxide catalysts

n-butane -►- isobutane

A1C13

The commercial development of the sulfuric acid process has
gone forward very rapidly and it has been estimated that by the fall
of 1940 plants had been installed capable of producing over 7,200,000
barrels per year of 92-94 octane fuel (unleaded) from C4 olefins and
isobutane alone.

The entire process may be summarized as follows: With pure
hydrocarbons it has been shown that the reaction in the presence of
sulfuric acid may be generally applied to the lower olefins except
ethylene. Thus propylene, isobutene, 1- and 2-butene , trimethyl-
: ethylene, together with lower polymerides of isobutene, di- and

- 2 -

triisobutene, and the butene-isobut.ene copolymer have all been shown
to react with isobutane to give good yields of saturated products
possessing octane numbers of 90+. Of the paraffins investigated,
isobutane, isopentane, and isohexane reacted smoothly although the
octane number of the saturated product falls off rapidly with in-
creasing length of the carbon chain- used.

The olefin is passed into a well-stirred mixture of the iso-
paraffin and 96-97$ sulfuric acid kept at temperatures varying from
-10 to 30°C. The overall yield of gasoline to the point where
conversion was no longer considered economical in experiments with a
! 4 cut containing 56$ of unsaturated compounds, was 310$ by weight
cf the acid used.

High Pressure Hydro ge nation Process

The production of hydrogenated fuels to augment the natural
supply has been made possible by the development of new catalysts
tor* the high pressure hydrogenation operations. These catalysts
possess the property of converting petroleum oils boiling outside
the naptha range into lower boiling cyclic and branched chain com-
pounds imparting desirable high octane values to the product. The
fuels obtained by this method generally possess antiknock qualities
superior to those of straight run naphthas.

A typical run gives conversions cf 50-75$ per pass and yields
80-95$ of 75-78 octane gasoline as compared to 74 for natural
aviation naphtha. The desirability of hydrogenated fuel is increased
by its high octane number response on additions of lead tetraethyl.

In addition to the production of aviation naphthas high pressure
hydrogenation is currently employed on a commercial scale for the
preparation of blending agents. One of the principal blending agents
thus made results from the hydrogenation of octenes available from
copolymerization of iso- and n-butylenes. The hydrogenated codimer
is not an aviation fuel in itself since it does not possess the
boiling point range available for standard engine use. It is,
therefore, principally employed as a blending agent to increase the
octane number level of natural or hydrogenated aviation naphtha.
In this case the increase in octane number is from 84 for the ccdimer
to 100 for the hydrogenated polymers.

Mechanisms of Low Temperature Alkylations

McAllister and his coworkers have postulated the following
reaction mechanisms for these alkylations:

1. Simple addition of the isoparaffin to the double bond of
the olefin. Most of the products obtained can not be explained by
this mechanism.

2. Polymerization and depolymerization. The fact that a large
proportion of octanes are obtained by the alkylation of isobutane with
butylene trimers is evidence for this reaction.

3. Rearrangement of the primary products. This reaction un-
doubtedly occurs but not to a large extent under the given conditions.

_ 5 -

4. Carbon-carbon cleavage and addition of the fragments to the
double bond. Pauling has calculated that it takes less energy to
cleave a C-C bond than it does to cleave a C-H linkage, hence this
reaction is more plausible than dehydrogenation. The reaction of iso-
fcutylene and isobutane can be postulated as follows:

5 ?

C-C-C -+ C-Cr- + c-

c-c-c-C-C

C c

l i
C-C-c-c-C

isooctane

Caesar and Francis propose a method whereby the formation of the
observed compounds could be easily explained. They state that the
olefin is able to wedge itself in between a methyl group and the
rest of the isoparaffin so that the methyl group adds to one side of
the double bond and the rest of the isoparaffin to the other. The
methyl group farthest from the tertiary carbon is the one split off in
the case of isopentane. Because it is necessary to use an isoparaffin
they suggest as the first step that there is a bonding between the
tertiary carbon, or its lone hydrogen atom, and the catalyst.

The following table gives the paraffins found and the manner in
which they are explained by this method.

1. Isobutane + ethylene

C~C-c-c-C

2. Isobutane^* isobutene
C c

■C-c-c-<

C-C-c-c

C c
C-C-c-c-C

Isopentane + propylene
C

C-C- C-c-c-c

C-i-C-'c-c-C

Consideration of the thermodynamic equilibria of isomeric
paraffinic hydrocarbons show the following interesting point: "In
any group of isomeric paraffins formed by this process," the relative
amounts of the isomers agree closely with those computed by thermo-
dynamic equilibria when those isomers are excluded which are not
permitted by this mechanism." Thus in the case of the hexanes:

n-hexane
2-methylpentane
3-methylpentane
2, 2- -dime thy lbutane
2,3-aimetnylbutane

AF (n =

- 4 -

Eqi
P

jiilib
sr Ce

rium
nt

Per

nt
at

in Alk-.l-
;e

Calc

Found

-558

10-25

-558

-1341

-1165

75-90

bibliography

Miirphree, G-ohr, and Brown, Ind, Eng. Chem., 31, 1083 (1939).
li-ch, et al., ibid. , 31, 884, 1081 (1939).
SSbAllister, et al., J. Org. Chem., 6, 647 (1941).
-oar and Francis, Ind. Eng. Chem,, 33, 1426 (1941).

Reported by Q,. Soper
November 19, 1941

THE MECHANISE FOR THE COUPLING OF DIAZONIUM SALTS
WITH AROMATIC AMINES AND PHENOLS

Bartlett and Wistar Hauser and Breslow

The diazo compounds have been a most interesting subject
of investigation since their discovery by Griess in 1858. They
have been proven to be a most important and useful group of com-
pounds.

The structure of these compounds has been a very cont rovers if.:,
matter. Griess thought that both nitrogen atoms were attached di-
rectly to the benzene ring, but this idea was soon disproven by
Kekule, who postulated the structures CeH5N=NX for the salt, and
CsH5N=NOH for the free base. Shortly, afterward, Blomstrand'
Erlanger, and Strecker independently suggested the identical struc-
ture, C6H5N=X, but for different reasons.

For a number of years, little attention was devoted to the
structure of these compounds, the major interest being in their uses
However, the issue was raised again in 1892 by von Pechmann, and
for many years a tumultous controversy raged between Hants oh,
Bamberger, Angeli, and many others over various suggested struc-
tures. The conflict between Hantsch and Bamberger was perhaps the
most acrimonious in the history of Organic Chemistry. The matter
has not been entirely clarified even at the present time.

The subject is complicated by the several isomeric compounds
which are known, either directly or through their derivatives.
Although many others have been suggested, supporting evidence
exists for the following structures:

(1) Kekule' s dizao formula, ArN=NOH

(2) Blomstrand' s diazonium formula, ArN=N

' X

(3) Hantsch' s stereoisomers modification of Kekule1 s
diazo formula

(4) von Pechmann' s nitrosamine formula, ArNHNO

Probably all these formulas are correct under certain conditions.
Taylor and Baker summarize the relationship in the following manner:

Mineral Acid

S \ &

x Ar OM Ar Ar^

Diazonium , &~Diazotate Isodiazotate Isodiazo- Nitrosa-
Salt hydrate mine

The limited scope of this report precludes a discussion of the
evidence for and against the various structures which have been
suggested from time to time. For further discussion of this point
your attention is directed to the General References listed with '
this report.

'• •.

- .... .t^

■-.-• \ • (T '3

•. •

,, ,-; o. uwuao *>v~ 5*? --•^— *

••■:'',.

K* OKA 3311M* 0i:-MO«i :al»

I- •

.. I -.. d k-i ■» • v."-

',:■-:. 1 ' i

i •

. IS \ O
li J. .-I

■

• &»•'•' •

• . . r ; .

. . I . .. •■ -

JO I .:. .. It.i

- ■

... I? • '

f ' .V-

. , . .■ i

"■ I.

!-!,'/s

i ■

, J.*?;, j.

• ■■•'

■:., '{-Mr. i'- ;-y- ; '

* C u •*

'.'.:■. "■ J . ' ''

r - .

" ; :. ■ j , . ..

.jjil'.tfcj -; '"koaA

J If- J .

'1 'V.

■I :i:'

■in

m*t+ i&

. - 1; ■

- 2 -

The primary interest of this discussion is not the possible
structure of the diazo compounds, but the mechanism by which one
of their most useful reactions occurs. It was discovered in 1870
by Kekule and Hidegh that diazo compounds would unite with sub-
stances such as phenols and amines to form stable azo compounds.
Many hundreds of these • substances have been and are now used as
dyestuff s.

This coupling reaction occurs in weakly acidic or weakly
basic solutions. The mechanism by which it occurs has been a source.
of uncertainty. It was generally supposed that the coupling took
place between undissociated diazohydroxide and the phenol or amin^
The reaction has been usually represented as follows:

ArN=N0H + ArOH -* ArN=NArOH + HOH
or ArN=N0H + ArNR3 -► ArN=NArNR3 + HOH

Since the reaction did not go well in strongly acid solutioni it
was presumed that the addition of alkali to reduce the acidity
simply converted the diazonium salt into the diazohydroxide* and
thus promoted the coupling reaction

This point of view does not appear consistent with the modern
electronic conceptions of aromatic substitution, which require that
the substituting reagent be an electron acceptor attacking a region
of high electron availability. It seems more logical that the di-
azonium ion, which can readily serve as an electron acceptor is the
2!lw?«-rea?en/ lnJhe paction* Tt -1* known that electron attracting
S^inUS?if (SuCh *! the Altro ST™?)* whi<* would increase the g
electrophilic character* activate the diazonium component. It is
also known that the same substituents, which would reduce the elec-
tron donor tendency, deactivate the amine or phenolic component.

The diazonium ion may be represented in two resonance forms,

..+

ArjNj-N f ~> Ar:N::N

<*j (B)

tuS°ofh^eS?nanCLf0rm £ Probably contributes largely to the struc-
le of the ion it may be assumed that at the approach of an elec-
tron-donating molecule, resonance form B becomes the major struc-
ture.. Similarly, the activation of the ion by the presence of
electron attracting substituents in the aromatic ring can be explained
by such a structure. The electron attracting substituents would

bvUILwl6nTah'e f°r? B t0 contribUte ffiore to the ionic structure
by drawing the electrons closer to the aromatic nucleus.

It also appears that the free amine rather than the substituted
Stv°^ or the phenoxide ion rather than the free phlnol

be mor^TfkPif ?at6r tendenc? for electron donation, and hence vould
be more likely to serve as the other active componeAt in the

^onicnf r^ctio4i- Similarly, the presence of electron attracting
groups in the aromatic nucleus would obviously reduce the tendency
of this component to act as an electron donor*

- 3 -

The most obvious mechanism for the coupling reaction w ould
appear to be

<CIJ>^r +H£2^N'(CH3)s "* <C>N=N^^3>=^+(CH3)3

__^>N=N<^3>N(CH3)2 < "H+

Hauser and Br e slow have shown that the diazonium ion is
capable of entering directly into the coupling reaction. They usea
anhydrous media (pyridine), and observed that phenyldiazonium
chloride couples readily with either p-naphthol or sodium ^j-naphtho^
ide to form l-benzeneazonaDhthol-2.

The argument might be made that a molecular oxy-azo compound
(analogous to the diazohydroxide in aqueous solution) between the
diazonium ion and the (3-naphthol might be formed first, followed
hy a reaction between the oxy-azo compound and unchanged |5-
naphthol. ' In answer to this possible objection, investigation of
the reaction of the nitrogen-azo compound benzenediazopiperidide
(analogous to the hypothetical oxy-azo compound postulated ab'ove)
with p-naphthol and its anion was made. Coupling did not occur.
The nitrogen-azo compound was used because of the difficulty in
obtaining the oxy-azo compounds in a pure state.

These results become even more convincing when we consider
that addition of pyridinium chloride to the solution of benzene-
diazopiperidide and (3~naphthol (or sodium p-naphthoxide) brings about
some coupling almost immediately. The pyridinium ion effects the
decomposition of the benzenediazopiperidide to form piperidine and
benzenediazonium ion, which is then free to couple with the p-
naphthol or sodium 6-naphthoxide.

C6K5N=N-NC5H10 CsH5NH > C6H5N=N-NC5H10 -> C6H5NS+ + C5H10NH

The above results seem to show conclusively that the diazonium
ion is capable of serving as the active reagent in the coupling re-
action, at least in anhydrous media.

Further evidence in favor of this mechanism has been obtained
by Bartlett and Wistar, making use of an entirely different method
of approach, i.e., from a kinetic study of the reaction. The fact
that diazonium coupling with phenols and amines takes place only
when the acidity of the solution was kept below a certain value
suggested to Conant and Peterson that the rate of coupling is a
function of the H+ activity of the solution,, They investigated
this point with different diazonium salts and phenols. Heavily
buffered dilute solutions of constant ionic concentration were used,
and the progress of the reactions was followed colorimetrically.
It was found that the reaction is strictly bimolecular and free
from complications. Over the pH range investigated, when log k was
plotted against pH, curves were obtained which had a slope of 1.

• .

4 -

(1) Diazotized o-anisldine + disociium-
2-naphthol-3-6~di sulfonate

(2) Diazotized o_-anisidine + sodium- 1-
naphthol-4-sulfonate

13) Diazotized sulfanilic acid +
di sodium 2-naphthol-3-6-
disulfonate

(4) Diazotized sulfanilic acid +
sodium l-naphthol-4-sulfonate

A probable mechanism was suggested -

=* ArN=iMOH + X"

Figure I

(1) ArN3X + 0H~ *

(Fast and reversible)

(2) ArN=NOH + P qp#t ArN=NP + Hs0

(Relatively slow and irreversible)

The equilibrium in the first
step would be a function of the (OH""")
activity, which is a direct function of 1.0
pH. The rate controlling reaction (2)
would depend directly on the concentration
of ArN=N0H, and thus the rate would be a direct
function of the pH value of the solution.

It is to be noted that both of the react ants are capable
of acid-base equilibria.

p o

log k

3.0

ArNs

and

+ OH
ArOH

ArHs0H

ArO~ +

However, the phenol equilibrium does not reach appreciable dissocia-
tion in the pH region in which coupling occurs, and hence is of
little importance in the actual kinetic study.

Recently Bartlett and Wistar have investigated the coupling
reaction using amines instead of phenols. Unlike the phenols,
the acid-base equilibria of amines reaches dOfo in the pH range
where the coupling reaction occurs.

ArNH.

+ H"

a ArNH3+

Hence, this equilibrium must be of practical importance in the re-
action, and the curve, log k vs pH, cannot be a straight line, as
in the case of phenols.

There are four possibilities for actual reacting components
in diazonium salt-amine coupling.

(1) Diazohydroxide couples with the substituted ammonium ion

(2) Diazonium ion couples with the free amine

- 5 -

(3) Piazonium ion couples with the substituted ammonium ion

(4) Diazohydr oxide couples with the free amine

In interpreting the results obtained, it is advantageous to
consider the general shape of the curves (log K vs pH) resulting if
the various possible combinations above are the active reactants.
For simplicity, a model case with the following dissociation con-
stants can be set up.

For the substituted ammonium ion K = 1 x 10~4

For the diazohydroxide » .& = 1 x 10"3

For the phenol K = 1 x 10~8

7/e can calculate the fractions of the total amine existing as free
base and as. substituted ammonium ion in buffer solutions of
different pH values* Similarly we can calculate the fraction of
the diazonium compound existing at; the diazonium ion and as
undissociated diazohydroxide in these same buffers.. Likewise the
fractions of phenol and phenoxide ion can be calculated, "Vie re-
sults of these model case calculations are tabulated below „

Effect of pH on fraction
of each component in acid
and basic form

(B)

(C)

(D)

(E)
(F)

ArNH2

ArNHa+ + ArNH.

ArNK

?!±~3L

ArNH3 '" ■*■ ArNH 3
ArO~

ArOH + ArO
ArOH
ArOH + ArO""

ArggOH

ArNaOH + ArN3+

ArN:

ArNoOH + ArN=+

Referring back to the possible combinations of diazonium com-
pounds and amines, if Case (l) represents the facts, then the
velocity constant for the rate of coupling v/ill change with pH in
proportion to the change in product

(ArN3OH) (ArNH3+)

(ArN3OH) + (ArNp )

(ArNH3+) + (ArNHp)

Corresponding expressions will be obtained for the other three
cases. Plots of these products, therefore, indicate the general
slope of the Rate-pH curve which should result from each of the
eligible mechanisms. The theoretical curves are listed below.

•

- 6

(1) ArN3OH + ArNH3

(2) ArN2+ + ArNHs

(3) ArNs + + ArNH,

(4) ArNsOH + ArNH;

Figure 3

log k values plotted are actually
the fraction product illustrated
above.

The general shapes of these
curves would be the same even if
the ionization constants were not
the same as in the assumed case,
only the points of inflection being
shifted. Actually the assumed
values are near the constants for
the amines and salts actually usei(.
To be perfectly accurate, it is
necessary to note that these curves are not strictly correct, since
the acidic character of the diazohydroxide is neglected. Only the
alkaline portion of the curves would be affected, however. Experi-
mentally this is of no importance as the pH regions investigated
included only those regions where this effect is negligible, i.e.,
pH range 2.04-6.25. It is to be noted that the curves for
mechanisms (l) and (2) are identical with each other in general
shape, but are quite different from those of mechanisms (3) and (4).

The experimental work consisted in a determination of the
velocity constants at various pH values for the coupling of

(1) 1-Faphthyl amine 4~sulfonic acid with diazotized sulfanilic
acid.

(2) 1-Naphthyl amine 8-sulfonic acid with diazotized aniline

Determinations, were made in heavily buffered solutions of constant
ionic concentration and temperature, duplicating the conditions
used by Conant and Peterson exactly. Progress of the reaction was
followed colorimetrically. Results were obtained as shown by the
following curves.

Figure 4

_^*- — •£

1 -

' ^

0 -

pH
1 1 1 h-

-4—

Figure 5

12 3

. ;

■

- 7 -

Figure 4 (Cont)

1-Naphthyl Amine~4**Sulfonic Acid
+ Diazotized Sulfanilic Acid

Figure 5 (Cont)

1-Naphthyl Amine-8-sulf onic
Acid + Diazotized Aniline

(Solid circles are experimental
points. Open circles are
theoretical superposed values)

Only a portion of the curve is shown in each case as it is obviously
impossible to obtain experimental results over the entire pH range.
However, this portion is sufficient to establish the general shape.

Of the theoretically eligible mechanisms, it is seen that
the curves for mechanisms (l) and (2) are similar to the experimental
curves. In fact, by making the slight necessary change in assumed K,
the exact coincidence between the theoretical and experimental
curves is obtained. This has been done in the graphs above.
Mechanisms (o) and (4,) are definitely eliminated from consideration,
although (4) has been usually considered as the most probable.

In deciding between mechanisms (l) (Diazohydroxide coupling
with substituted , ammonium ion) and mechanism (2) (Diazonium ion
coupling with free amine), there appears to be little doubt that
mechanism (2) is the correct one. In those cases where aromatic
substitutions have been carried out on anilinium ions, these ions
have proven both unreactive and powerfully meta directing. It
appears most unlikely that such a substance could act as an active
intermediate in a reaction which fails to occur except with highly
activated benzene derivatives, and which always results in para or
ortho substitution. On the other hand, mechanism (2) is perfectly
in accord with the modern electronic theories of substitution.

The phenol coupling reactions studied by Conant and Peterson
are subject to the same analysis. Theoretical curves are obtained
as follows:

Figure 6

-4

M -8

ho
o

-12 --

-16 --

(5) ArN3 + ArOH

(6) ArN2+ + ArO~

(7)ArN20H + ArOH

(8) ArNaOH + ArO~

log k is actually the fraction
product of the reacting pairs

Although for obvious reasons, complete experimental data
is lacking, it is seen that only curves (6) and (7) have any por-
tion that is linear with a slope of 1 as observed. The choice here
is between the reactive pairs, diazohydroxide and phenol, or diazoni-
um ion and phenoxide ion. On theoretical grounds, the second mech-
anism is to be preferred here also. Certainly it is auite definitely
shown that the results of Conant and Peterson are not incompatible
with the postulated mechanism.

Since two entirely independent and different lines of
approach to the problem have led to the same conclusion, and since
this conclusion is thoroughly in accord with modern electronic
theory, it appears to be definitely established that the dlazonium
ion and the free amine or phenoxide ion are the active agents in
the dlazonium coupling reaction.

Bibliography:

Hauser and Br e slow, J. Am. Chem. Soc, 63, 418 (1941 ),
Wistar and Bartlett, ibid. , 63, 413 (i94l'jr
Conant and Peterson, ibid. , 52, 1220 (193 o) .
Hauser and Ereslow, ibid., 62, 2389 {l940),
Hammett, "Physical Organic Chemistry", p. 314

General References:

Saunders, "The Aromatic Diazb Compounds"

Taylor and Baker, "Sidgwick's Organic Chemistry of Nitrogen",
Chapter XIII

Reported by W. E. Blackburn
November 26, 1941

A REVIEW OF THE ORGANIC CHEMISTRY OF ARSENIC

Since there are no naturally occurring organic arsenic com-
pounds which might serve as starting material in a synthesis, one
is immediately faced with the problem of preparing a compound with
the desired 3~As linkage. With this end in view a large number
of arsonation reactions have been attempted with varying success.
Of these many reactions only a few with general applicability can
be picked out,.

I. Methods of Arsonation.

A0 Aliphatic

Cahours and Richie in 1854 found that by heating an alkyl
halide, usually the iodide, with sodium arsenide they obtained a
mixture of arsinesc

RI + Na3As — > R2As-AsR3 + R3As 4 R4As I"

This reaction was important in pointing out the relationship
between the various types of arsines then known, but is of little
preparative value because of the mixture obtained. In 1859 Cahours
announced an improvement of the synthesis.

RI + Zn3As3 -» R4As+IcZnIa •—y™^ R4As+I~ ^|^» R3As + RI
The cadmium salts could also be used,

Michaelis and Reese in 1882 advanced another more versatile
method for preparing these arsines „

3RX + bNa + AsX3 -» R3As + 6NaX

Using the following types of arsine halides, which are
available from other reactions, RAsX3 and RR;AeX, one can obtain
tertiary arsines of the type RAsR2 and R'R* rR: f "As respectively.

The most widely used synthetic method which is also capable
of industrial application is the Meyer reaction.

As-ONa + RI -> Nal + RAS-ONa
^ONa ^ONa

Meyer carried the reaction out in an aqueous medium usually
under pressure to avoid the loss of the alkyl iodide being used.
Dehn extended this reaction and found that he could improve the
yields by using dilute alcohol solutions and potassium rather than
sodium arsenite. He thus obtained an homogenous reaction medium.

The procedure as modified by Dehn still left much to be de-
sired, because the higher homologs were obtained in increasingly
small yields and because a considerable amount of alkyl iodide was
lost due to ether formation with the solvent alcohol. Valeur and
Delaby found, in contrast to Dehn, that they obtained better re-
sults with aqueous solutions which allowed better agitation and re-
frigeration. Adams and Quick selected this method for preparing
higher aliphatic arsonic acid and found that better yields could be
obtained by the use of alkyl bromides or chlorides. The sodium
salts of the arsonic acids produced are soluble in the reaction

- 2 - SI

mixture and difficult to isolate. When the alkyl chloride or brom-
ide was used these sodium salts could be converted to tie free
acids which are much less soluble. When the iodides were used the
hydrogen iodide immediately reduced the ar sonic acids.

The reaction can be extended to further alkylation as follows:

RAs-ONa + R'l -* RR'As-ONa + Nal
xONa

RR'As-ONa + R"I -> RR'RnAs=0 + Nal

Still another method of importance because of its application
especially to the production of lethal compounds of military value
is the addition of arsenic halides to acetylenes.

R-C=C-RT + AsCl3 -> R-C«

CI .

*C-R!

ASCI;

The reaction has also been carried out on ethylene

B.. Aromatic.

The oldest method of aronation in the aromatic series is
that of Bechamp.

^>NH3+As04Hs- 19°-S°°° ■ NHs<^/>As03Hs

The reaction can be carried out on nuclear substituted amines
with an open £ or o position. If both the c_ and the jo positions
to the amine group are filled the reaction fails.

By far the most important method of ar.omatic arsonation is the
Bart reaction or one of its various modifications.

<^ ^-N-NOH + K3As03 Ka° i <^ "\as03K3 + N2 f + KOH

According to Bart's original procedure a solution of the di-

azonium salt was prepared and then made alkaline and heated to a

rather high temperature, then with continued heating the arsenite

solution was poured into the diazo solution. He later suggested

the use of copper, nickel, silver, and cobalt and their salts as
catalysts.

Since Bart's original procedure appeared it has undergone
cany modifications in the hands of other workers. Schmidt suggests
the use of a neutral or even faintly acid medium with no catalyst,
Mouneyrat uses alkaline, neutral, or acid reaction medium with a
dual catalyst consisting of a metallic salt and reducing agent
chosen with reference to the pH of the solution.

No blanket statement can be made in regard to the best con-
ditions for carrying out the Bart reaction. The choice in any
given case must be made after experimentation with the particular

- 3 -

amine under consideration. Even then the successful completion of
the reaction is dependent on the technique employed so that two
operators have difficulty in reproducing yields.

As in the case of the Meyer reaction the higher arylated
arsinic acids and arsine oxides can be prepared by proper choice of
starting material*

^> N=NOK

J— *s

N=NOK

K2O3As0

KO-As-02

02As-OK + N8

03As=O + N2

Chatt and Mann recently developed

iodides of the type As (abed)
stereochemically.

AICI3 -*

These

1 +

AsCl3 +

AsCl2 +

RR'AsCl

a method for making arsonium
compounds are of interest

+ T~

04AsTI

KI + _
+ AICI3 -» — * 03RAs I

+ AlClo ~*

+T-

* RR'02As I

R,R,,RMAs +

+ A1C1:

RR"R"0As+I~

Many attempts at direct arsonation of aromatic compounds by
the Meyer reaction have been made. Rosenmund was successful in
isolating benzene arsonic acid from bromobenzene and aqueous po-
tassium arsenite. Hamilton and Leudeman confirmed this report and
attempted to extend the reaction. They found that c_-chlorobenzene-
arsonic acid could be arsonated in aqueous medium to arsonophthalic
acid in good yields but that other compounds with equally active
halogen gave poor yields or none at all. It thus appears that
direct arsonation is not a general reaction in the aromatic series.

II. Transformations in the Arsenic Series

The presence of arsenic in an aromatic nucleus does not
materially alter the aromaticity of the ring. The arsonic acid
group has about the same m orienting effect as the sulfonic acid

- 4 - 93

group and activates the ortho and to a lesser extent the para
positions. p_-Chlorobenzenearsonic acid is sufficiently reactive
to react with aromatic and aliphatic amines and with alcohols i The
amino arsonic acids are of especial synthetic value since they
undergo all of the characteristic aromatic amine reactions. The
presence of an amine group or hydroxy group in the para position
seems to weaken the C-As linkage and in this case the use of halogen
containing reagents must be avoided since by some of them (i.e.,
PC13, SOsClg, S0C12, CI, Br) the arsonic acid group is cleaved from
the ring being replaced by halogen.

There are numerous derivatives of arsenic known due to its
existence in multivalent states. The arsonic acids are the most
important derivatives since they are most readily prepared. Since
interconversion in the series is quite easy, all of the other
derivatives may be prepared from the arsonic acids. There are note-
worthy differences between the aliphatic and aromatic arsenicals
of the same class, but they are differences, as a rule, of stability
and degree rather than of kind.

The types of arsenic compounds, their nomenclature, and their
interconversion are summarized in Tables I and II.

III. A Comparison of Nitrogen and Arsenic Chemistry

As can be seen from Tables I and II, there are several types
of compounds similar in structure to well known nitrogen compounds.
There are also many arsenic compounds whose nitrogen analog is not
known and vice versa.

Arsenic compounds react very readily v/ith halogens to give
haloarsenic compounds; a reaction not observed in the nitrogen
series. This is to be expected since inorganic arsenic halides
are more stable and easier to prepare than nitrogen chlorides. The
ability of arsenic compounds to become oxidized to a higher valence
state is also more pronounced than in the nitrogen compounds. In-
deed, with certain arsenic compounds their air oxidation occurs
so frapidly that they are spontaneously combustible.

A study of the reactions and properties show that compounds of
similar structural formula often show fundamental differences in
their chemistry.

The arsonic acids most closely resemble the nitro compounds.
The reactions due to the acidic hydrogens of arsonic acids, of
course, have no parallel in the nitro compounds. The electronic
arrangement of nitro compounds is suitable for the formation of a
dibasic acid and the inability to do so has been attributed to the
small atomic size of the nitrogen atom which will not permit it to
coordinate the required number of groups. Arsonic acids can exis'-j
in the anhydride form RAs02, but unlike the nitro compounds, these
are hydrolyzed by water to the arsonic acid. The arsonic acid group
has no chromophonic properties, all the arsonic acids being white
solids. Like the nitro group, the arsono group exerts a m directing
effect on orientation and like the nitro compounds the arsonic acids
can be reduced to any desired lower valence state by the choice of
suitable reagents.

c
u

CD
Eh

P
C

CD
rH
Cd
>
•H

p
G
(D

rH
Cd
>

cd
P

CD
PM

rH
O
CO
<

•H

rH <&

>> <-

is! CD CD

•H vH
CO

co
m

rH-H

cd co
•H ^
fH cx3

<
n

isi CD

H CO

cd >H Tj

•H CO -H

Sh rH X

P cd O

>i-h w

-5-.P O

•H

«3

CD
P

cd
a?

•H
C

J-3

co
H

cd
o

•H

CD
CO

u
<

W
PQ o

•H

cd
Xi
ft
•H

P
CJ
CD

cd
>

•H
Sh
Eh

73777* /y/./aT"

•H
!h

+^
CJ
0)

Co
>

cd
p
C5
CD

<«J

S-(CD

cdT3

•H
rHX

rH < O

rHCD W
Cd£
w-»-H
«C>CQ

7^777"

CD
H TJ

>s CD CD >H

- o .yc-dc

rH «H »H cd

Cd tO rH >/

•H Si «i o

ft; 'O ajx!

pc ^ a?

CO rH £

<t; cd -h

I cd co

co U U

< p aj

0> CD -H

ft! Eh ^d

V
CO

pc;

>3 CD
rH -H

cd to

•H fs

•a cd

H
O

CD I CD

C cdrd

•H Sh-H
co+3rH
£h CDcd

cd pjcJ o

■m

, 1

U
PQ

^ 1

~6-

/*"

\ !

Cti
>
•H

iary
sine

- <

,«

CO C ^

pc{ 4^> cd

— X

>s \ \,

u \ <$

§ \ ,

v\-r

o V+«

•H

cd co

■ CO

r$ <C

" ?-i

C3? 'f

oc;

re r-

— ^ O

>s'd CD

rH tf

•p

cd CD-H

•H CjrH

Sh-h cd

-•4

•H
Jh

4J crj£j <;

oc;

4^>

•H

•^

4-3

CO
SH CD

1 rH

C c»

c rt

CTJ<d

a tn

u o

OJrH

NJ Ph

•H

cd u:

rH <!

rH X

CO rH

o hx;

"-*-'

a5 O

>>o

< >>>>

re rn

X CD K

^H CD

.")

5 <c

OS cd

o c: o

cd a

<rj

. J

•H

c-h ,w

■H ^H

• — ■■

En 0d t3

TJ "d CO \

HO*

rj CO

13^

»"3

•H

Secon
ent

rH CD

cr.

M«tj

' 1

Ch-H

crj

4-3

cdrH

°S

o — —

rH rH£i
O >y,H

^1 < -

CO N

rHrH

«3

re .
Ph r

cdn-3

tto if n pi * q

1/ c\. ^~

""*"

/DM

J.' t/

CD
rH ^

13
1 -H

U X

C, CO

\ 1 1 o

"* ?H -H rO

cd o

W K

cd ;3

re c

T5 rd -H W

O CJ

■

> c

O O

M z rH

.<3

CJ CD

re p^

' ^)

>' !

•H 1

rH CD

N 1

rH -Hnd

1 , O rH CD O ^

CD Nl

-T

^r- i

C c

t» c

>> CJ-H

-> CO J

>)drl

)

CO CJ

ry-

l_ i

Eh <
P

in -H
cd CO

C, CD.O

cd co cd

/ r

>/ pq cd co o .*

-j

?H CD

cd,o

.-s ^

ri i

Sr)0

"*"^

■—

*-.-— . -

t s^

BoV>*

c(7 J

r?7

~j ,

^-/

-•T-i

cd
6

•H

M
V

-r

J
■0 ..r

*^> —

Ph
4->

' /

r
0
V /

r -a
r p

•J

, i

CD 1

•v

a 5UCD

t£-\. '

o i

•h or5

(

CO HH

> i

cd o

0 '

O
\

- o

^1 /

rH Orrj

^ !

~ X

=<.

>.-^>

>5 C0.H

r-< cd

cd 4J

u u o .

cd cd cd

The arsines bear little chemical resemblance to the amines.
They are nearly devoid of basic properties. Methylarsine and the
tertiary arsines are basic enough to form salts with halogen acids,
but these salts are completely decomposed by water. The tendency
to oxidize increases as the number of organic groups is increased.
Arsine itself is stable to air oxidation, the mono organic arsines
are fuming liquids which must be protected from air to preserve
them, and the di-organo arsines are spontaneously combustible. The
tendency to air oxidation is less in the aromatic series.

On the other hand, the arsonium compounds bear a striking
resemblance to the ammonium compounds. The salts are highly
ionized, and the arsonium hydroxides are strong bases absorbing
water and carbon dioxide from the air. The arsonium iodides are
not decomposed by aqueous alkali, but must be distilled from
potassium hydroxide to decompose them.

The areeno compounds are quite dissimilar to the azo compounds.
The arseno group is not chromophoric although some arseno compounds
are colored. Oxidation of the arseno compounds occurs under very
mild conditions, silver oxide being reduced by an arseno compound,,
The comparative instability of t^e arseno group is shown by the
fact that halogens, sulfur and hydrogen cleave the As=As linkage and
by the fact that heating a solution of two different arseno compounds
causes an exchange reaction.

RAs=AsR + R'As=AsR' v=^ 2RAs=AsRf

Likewise, the arsine oxides bear only structural resemblances
to the nitroso compounds. The arsine oxides are amphoteric, giving
with halogen acids the dihalogen arsines, and with bases the salts
of hydroxy arsines R2As-ONa or arsenous acids RAs(ONa)s. In the
aliphatic series the acids corresponding to these salts cannot be
isolated, but esters can be prepared with an alkyl halide under
suitable conditions. In the aromatic series the arsenious acids.
and hydroxy arsines are not stable unless nuclear nitro or carboxyl
groups are present. In the latter event the arsine oxides may be
hard to obtain.

Compounds having a formal resemblance to the dia 20, diazonium,
hydrazo and azcxy groups and to the hydrazines are not known in
the arsenic series.

In general it may be said that analogous reasoning from the
nitrogen compound to its arsenic analog is not justified, although
significant resemblances can be found.

IV. The Stereochemistry of Trivalent Arsenic

To date the best physical evidence indicates that trivalent
arsenic has a rather flattened pyramidal configuration with the
arsenic atom at the apex. The bond angles are calculated to be
100° + 3°. Recent chemical evidence seems to support this view.

Lesslie and Turner have resolved compounds of type I and found .
them to be very stable to racemization.

- 8 -

CaHBI

\Aas\^C°*

1 CH3 also C3H5

The optical activity
(a) The three rings
to the asymmetry of
behind the plane.)*
the rings would have
asymmetry of the mol
activity. (c) The
"asymmetric arsenic
in the absence of an

r^V V%

II s N

CH3 C2H5

-COaH

of these compounds could be due to three causes:
could be coplanar with the optical activity due
the molecule (i.e., methyl group in front or

(b) If arsenic has a pyramidal configuration

to be folded along the 0 to As axis. The
ecule would still be the cause of optical

optical activity could be due to an
atom" which would impart optical activity even

asymmetric molecule.

Kamai attempted the resolution of compounds of the type
R-A»-Rf-C03H but failed. This negative evidence, of course, does

not rule out the third possibility. The evidence in favor of (b) is
rather meager. Lesslie and Turner found that the optical isomers
of I were very stable to racemization, being only slowly racemizel
in boiling benzyl alcohol. However, in
presence of an alkyl iodide

ethyl alcohol in the
racemization was rapid. Treatment

with ethyl iodide would produce II in which the arsenic atom, being
no longer trivalent, would probably not have the pyramidal configura-
tion. If II has a tetrahedral arsenic atom, then it should be
resolvable, but all attempts at resolution have met with failure c

Chatt and Mann have advanced interesting evidence in favor o f
the pyramidal arsinic atom. They synthesized 9, 10 di-j>-
tolylarsanthrene (III, IV, or V) and found it to consist of a mix-
ture of two forms which they separated by fractional crystallization,
Assuming the pyramidal construction the following three compounds
are indicated.

f I *

The left hand ring is projecting from the plane of the paper
toward the observer and a dotted line to the ]>-tolyl group indi-
cates that it extends behind the plane of the paper and toward the
center of the molecule. From an examination of the model for IV
it is seen that the two p,-tolyl groups are almost coincident in
space. They offer this as an explanation of their failure to obtav
a third form.

- 9 -

It thus appears that trivalent arsenic does have the pyramidal
form. However, this pyramidal form might be stabilized by the
fused ring systems present in I, III, and IV. The evidence on
hand points to the fact that I is resolvable due to the asymmetry of
the molecule but does not yet prove or disprove the existence of an
"asymmetric arsenic atom."

Bibliography

For coverage of literature to 1923 see -
Raiziss and Gavron, J. Am, Chem. Soc. Monograph. .
Lesslie and Turner, J. Chem. Soc, 1954^ 1170; 1955. 1051;

1956. 750,
Chatt and Mann, J. Chem. Soc, 1940. 1192.
Hamilton and Leudeman, J. Am. Chem. Soc, 52, 5284 (1930).
Kamai, Ber., 68B, 960 (1935); 68B, 1893 (1935).
Tatum and Cooper, J. Pharmacol., 50, 198 (1934).
Eagle, ibid.. 66, 423 (1939).

Reported by C. W. Theobald
November 26, 1941

ADDITION PRODUCTS BETWEEN KETENES AND UNSATURATED HYDROCARBONS

Staudinger prepared several addition products between ketenesr
and unsaturated hydrocarbons and suggested 'chat they were substi-
tuted cyclobutanones. In the past few years the structure of the c,e
addition products has been definitely established.

Dipheny.lketene and Cyclopentadiene. — Staudinger condensed

and cyclopentadiene and obtained a dicyclic unsat-

i&lble st.

dime thy Ike tene

urrted ketone, C9H130. He suggested I and II as two poo

tures, where R=CH3. After unsuccessfully attempting to character

ize the degradation products of Staudinger' s compound, Simonsen

studied the compound in which R=C6H5 jrnd proved the structure to

be I.

' V.

RoC-CH CH

0=C-CH-CH2

permono-

phtna
acid

lie 0
CH""/

CH2

/ \

R3C-CH CH

i ! I

0=C-CH-CH
II

CHOH
02CH-CH CHOH

glycol

alkali

H > \
02C-CH CH

III

0=C-CH-CH2

fusion

^ H03C-CH-CH2

a and p

III

CH3C00H>dlecetfte
H3S04 03CH-CHCOOH

a and p
forms CH2COOH

CHCOOH

:2«

KMnO,

lead

tetra

acetate

The a and p forms of IV were synthesized by the hydrolysis
of the condensation product of bromodiphenylmethane and methyl
sodio-propane-a,a, P,X -tetracarboxylete, one of the carboxyl groups
being eliminated.

CfiH

6rl5

^6^5

CH Br

COOCH.

Na I -C
i

-CH-CH2COOCH.

COOCH3 COOCH3

HC1

■CO.

03CH-CHCOOH
CHCOOH

CH2COOH

Dlpheny Ike tene and Styrene. — Structures A and B were suggested
by Staudinger for the condensation product of diphenylketene and
styrene. Hydrolytic fission (NaOH, H20) of the adduct gave an
acid, C22H2002, which he suggested was a, a,^ -triphenylbutyric acid
(I) or a, a, p-triphenylbutyric acid (II).

03C- C=0

I I

> 0CH-CH2

02C-CO2H
I

H2C-CH3

i 11

02C-

-C=0

CH2-CH0

02C-CO2H

HC-CH20

-2-

100

Both of these acids were synthesized by Bergmann and found to
differ from Staudinger's acid.

Addition of the ketene might involve reaction at the nuclear
double bond (as with benzophenonephenyllmj.de end phenylnagne&iwp
bromide, and ethyl azodicarboxylate and styrene derivative s) to
give compounds like:

0CH

'c=o

CH;

CH0

Bergmann prepared a series of compounds which would result
hydrolytic fission of the above rnd analagously formed addition
products; none Was identical with Staudinger's acid.

Simonsen in his work on the adduct with cyclopentadiene had
found that hydrolytic fission occurred between the diphenyl carbon
and the carbonyl group, A similar fission of A would give the
acid III. This acid was synthesized by Bergmann and found to be
identical with the acid derived from the addition product of di-
phenylketene and styrene, thus establishing A as the structure o.p
the addition product.

0SCHOCH3 + Na

08CHNa
A
0CK=CHCO3CHS

02CH C02H
0-Ch— CH2

III

Structure B was also ruled out by the consideration that it
would not account for the isomer! zation of the original adduct to
C6H5CH=CHCCH(CeH5)3.
0

Using a different method of degradation Smith confirmed
Simonsen' s structure for the adduct with cyclopentadiene. Smith
also found that diphenylketene reacted with cyclohexadiene and
with 2, 3-dimethylbutadiene. However, 1,4-dimethylbutadiene did
not react and neither did cyclopentadiene react with ketene.

Farmer and Farooq found that diphenylketene reacted with
cyclohexene, cyclopentadiene, cyclohexadiene, dimethylbutadiene
and pioerylene. However tetramethylethylene would not react,

Diphenylketene and Phenylacetylene. — In an attempt to obtain
further evidence thr t styrene and diphenylketene reacted to give
cyclobutanone derivatives Agre and Smith tried the reaction be-
tween diphenylketene and phenylacetylene. The compound obtained
was soluble in methanolic potassium hydroxide and gave a positive
phenol test (Folin). The G-rignsrd machine showed one active
hydrogen and no addition of reagent. The principal reactions

-3-

101

are below:
OH

N=NC6H4S03Na

0 KaCa<807/ \0

f ^Pb_(oAc)4 o J&

COOH

"+ 0COOK acetrte and /^
^ >COC6H5 meth^ ether j l|

NrOAc
(AcO)20

acetate

and
methyl
ether

OAc

OA:
0

All the reactions were adequately explained by structure I,
but the reactions would be just as consistent with the l^-naphthol

Following the lead of earlier work by Franssen, Smith and
Hoehn synthesized I by the addition of 1, 4-naphthoquinone to an
excess of 5 moles of phenyl magnesium bromide.

/VS

0MgBr

oxime

COOK

+ 0COOH
COC6Hs

Because of the unexpected n.-ture of the addition product an
attempt was made to determine the mechanism of the reaction. The
first step was to determine the location of the aryl group from
the acetylene. The reaction between p_-tolyl acetylene and di;Dhenyl-

-4-

10;

ketene indicated that the aryl group from the acetylene goes to
the 3-position in the naphthalene.

alkaline

Jc6h5ch3(d)

-^.

KMnO,

COOK

ICrHi

x' 0

COOK

Diphenylketene reacts with many compounds with active hydrogen
to give compounds containing the diphenylacetyl group. This might
be the first step in the reaction. The compound which would be
formed if this were the first step, 02CKCC=C0, was prepared by the

reaction between 0CsCMgBr and 02C=C=O, Attempts to cyclize the
phenyldiphenylacetylacetylene with fused ZnCl2 in acetic acid
failed. This eliminated this compound as a possible intermediate o

The fact that the dieubstituted acetylene reacted in tne same
manner as the monosubstituted acetylene indicated that the active
hydrogen of the phenyl acetylene was not necessary to the reaction.

00^30 + 02C=C=O

vV0

| 02C=C=O

The addition between diphenylketene and phenylacetylene re-
sembles the Diels-Alder reaction in that no catalyst is required
gnd that elevated temperatures are not needed. The diphenylketene
supplies a 1,4 conjugated system, if a double bond of a ring is
included. But the diene synthesis could only lead to p-naphthols.

^\ *CH

-Vv^0

C6HB

-5-

103

The work of SUauci.ing.er, Simons en, Farooq, end Bergmann had
demonstrated that ketenes react with ethylenic compounds to give
substituted cyclobutanone s. It is therefore quite probable that
a cyclobutanone is an intermedir te of the reaction with acetylenes.

+ CH

'CH

CC6H5

'or

CH
II

°*L^

.*/

The cyclobutanone is the first and most important inter-
mediate. It explains the formation of an cc-naphthol, why the
rcetylenic aryl group goes to the 3 position, why both mono- and
diary lacetylenes give analagous products. The bond which is bro]
in subsequent steps is the same one which breads upon hydro lytic
fission of the adducts of diphenylketene and ethylenes.

Attempts to obtain direct evidence for the existence of the
cyclobutanone intermediate have been unsuccessful.

.en

Bibliography

Lewis, Raraage, Simonsen and Wainwright, J. Chem . Soc . , 1837 (1937).

Farmer and Farooq, Chem. pnv Ind. , 1079 (1937 ).

Bergmann anc"! Bergmann, J. Chem. Soc., 727 (19^8).

Smith, Agre, Leekley anc1 Pri chare1, J. Am. Chem. Soc, 61, 7 (1939).

Smith anc1 Hoehn, ibir1 . , 61, 2619, (1939); 63, 11*75 (1941).

Reported by W, H, Kaplan
December 3, 1941

104

SOLVOLYTIC REACTION MECHANISM

Hammett-Columbiaj Ingold, Hughes-University College, London;

Bartlett-Harvard

Second-order replacement reactions by (a) anions or (b)
electron donating neutral reagents are quite well established as
occurring by a bimolecular inversion mechanism* This Lewis-London-
Polanyi-Olson mechanism explains the Walden inversion.

yH3 CH3

(a) I* + C6H13-CHI ^zr±. C6H13CHI* + i"

(b) NH3 + RC1 -> RNH3+ + Cl~ (Menschutkln )

It is almost equally well established that not all anion re-
placement reactions occur by this mechanism. For example: cc-
phenylethyl chloride is hydrolyzed to the alcohol at a rate inde-
pendent of the hydroxyl ion concentration, converted to the ethyl
ether by alcoholysis at a rate independent of the ethoxyl ion con-
centration, and to the acetate in glacial acetic acid at a rate
independent of the acetate ion concentration. ■

Such first-order replacements occur in solvents such as water,
alcohols, phenols, carboxylic acids, and sulfur dioxide and the
reaction rate is influenced by the solvent concentration when a
diluent is used.

To explain this first-order (in the presence of large excess
solvent) replacement, Ward in 1937 proposed a solvolytic ioniza-
tion mechanism, which Hammett and Ingold have accepted with
modifications. It is unlike the bimolecul:-r mechanism where the
energy to eliminate the departing anion is partly furnished from
the energy of the bond forming with the simultaneously entering
anion.

In the solvolytic mechanism the rate-determining and energy-
providing step is the solvation with insipient ionization of the
anion-formlng part of the molecule. The electron donating or

R R+

Step 1. R-CX ^± RC X" (solvated) slow

R R

Step 2. Y" + RC — X -* YCR + x" rapid

R R

"nucleophllic" reagent Y" is not involved in the rate-determining
step 1.

' !

■ . . f j| a

;•• ;•'.

-2-

105

The hypothesis that step 2 occurs with the anion being removed
not over 1 A away explains why inversion of configuration occurs
with large but incomplete racemization when optically active com-
pounds are being substituted on the asymmetric carbon.

The ionization mechanism will be favored relatively tft the
bimolecular mechanism by large electron-release from R, strong
electron-affinity in X, low nucleophilic activity (basicity) and
low concentration of Y, and high ionizing capacity of the solvent*
Such a mechanism is not possible when the group is held very firmly
as in aryl hs.lides and applies essentially to substitution at a
saturated carbon atom and in solution.

Primary halides tend to react by the bimolecular mechanism;
secondary and tertiary, by the ionization mechanism. The "mecha-
nistic critical point" is illustrated in work of Hughes and Ingold
on the decomposition of sulfonium hydroxides and salts in water at
100°.

.+,

OH" + RS R3»
(CH3)3S+ -JL

ROH + RS'S
CH30H second-order

( C2H5 )3S

-^ C3H50H

CH3S(C3H,-i)8 _E3^C3H70H
(CH3)2£c4H9-t -^±_>C4H90H

second-order

first-order

ki = 9 ks

k4 = 2600 k3

Trimethyl sulfonium salts were decomposed in alcohol

basicity

anions: .OH > 0C6H5

decreasing rate, second -order

first-order and equal rate

In addition to the three characteristics illustrated,
(l) an ionization favoring solvent, (2) a first-order reaction
rate independent of the concentration of the replacing ion, (3)
en inductive effect in reverse to that of bimolecular replacement,
the solvolytic mechanism can be diagnosed stereochemically.

d-sec-octyl bromide IN NaOH 80

60f EtOH
acid

\ j±.

\!/

^-alcohol 65^ racemized

1-alcohol lesa than

racemized
(probably due to secondary

reaction )

-3-

ICS

Some of the oth^r evidence for the mechanism is as follows:

1. In a mixed solvolysls experiment a linear rate dependency on
water was observed in the alcoholysis of benzhydryl chloride;
however; it was not a case of two superimposed bimolecular
reactions because less than 1/4 the theoretical proportion
(calculated from the relative rates) of benzhydrol to benz-
hydrylethyl ether wps formed. The water facilitated ioniza-
tion but entered step P. to a minor extent.

?.. In the solvolytic hydrolysis of p_,p_f-dimethylbenzhydryl chlo-
ride in aqueous acetone the two steps do not have widely dif-
ferent rrtee; therefore, a mass effect can be observed. Due
to building up chloride concentration there is a progressive
decrease in the overall reaction rate; furthermore a 0.05M
addition of chloride ion depresses the reaction rate one-third.

3. Only those halides which cause Friedel-Craf t alkylation were
effective in racemlzing a-phenylethyl chloride and the order
of racemizing effectiveness coincided with alkylrting effec-
tiveness (SbCl5> BC13 > SnCl4 > ZnCl3? HgCl). The solvolytic
ionization is analogous to that caused by these strongly elec-
trophilic reagents; it is presumably less in degree and due

to hydrogen bonding.

Bartlett has found that phenols exhibit hydrogen bonding with
hydrogen chloride.

4. Because the rate of hydrolysis of benzhydryl chloride in ace-
tone increases nearly thirtyfold for a tenfold increase in
water concentration, and other similar results, Hammett refers
to "polymolecular solvolysis." Dime.ric water must be more
effective in solvolysis than monomeric.

An unusual case is the alcoholysis of d-bromopropionic acid
to methoxypropionic acid. The reaction is mainly second-order at
sodium methoxide concentrations from 0.5 to 1 M, but mainly first-
order from 0,03 to 0.06 M. The second-order reaction inverts the
configuration but the first-order reaction does not and is accom-
panied by no racemization* The interpretation of the first-order
reaction involves cyclization and double inversion; the reaction
may or may not be solvolytic.

Br ,p H 0 /P

d-CH3~C — C -» 1-CH3-C-— JC=0 -* d-CK3-C-C

n 0- ti 0-

The hydrolysis of ethyl benzene sulfonate shows the charac-
teristics of solvolytic replacement.

-4-

10'

First-order elimination reaction? have been formulated by
Hauser as occurring by tne solvolytic initial step.

As a result of this work, to quote Hammett, "It seems very
probable that every displacement on an asymmetric carbon the rate
of which is proportional to the concentration of the nucleophilic
displacing ion or molecule involves an inversion of configuration."

Another step was taken toward the chemical explanation of
homogeneous catalysis when it was shown that the solvent is
kinetically important in first-order replacements.

Bibliography

Hammett, Physical Organic Chemistry , McGraw-Hill Company, New York

(1940).

Watson, Ann. Reports, |?, 236 (1940);. 35, 208 (1938).

Hughes, Trans. Faraday Soc, 34, 185, 202 (1938).

Hammett and McCleary, J. Am. Chem . Soc, 63, 2254 (1941 ).

Farinacci, ibid. . 63, 1799 (1941 ).

Bartlett and Dauben, ibid. , 62, 1339 (1940).

Reported by G. D. Jones
December 3, 1941

103

AROMATIZATION OF ALIPHATIC HYDROCARBONS

In recent years steps toward more efficient utilization of
natural hydrocarbon resources have led to development of such
processes as cracking, nitration, chlorination, and automatiza-
tion of gaseous hydrocarbons. The conversion of aliphatic hydro-
carbons to aromatic hydrocarbons is by no means a recent discovery
but the practical solution of the problem is the latest success
of the petroleum industries.

A bulk production of aromatics from open chain compounds in
industry obviously must make use of suitable naphtha cuts from
petroleum sources, light oils arising from low temperature dis-
tillation of coal, and the fractions of light spirit derived from
either high pressure hydrogenation of coal or from synthetic
processes for hydrocarbon production starting with water gas
(Fischer-Tropsch synthesis).

Aromatization is best approached by consideration of thermal
decomposition of paraffin hydrocarbons. This might proceed in
the following manner in the case of n-hexane:

1. 2C6H14-> C13H36 + Ha, 6. C6H14~> C6H10(cyclic ) + 2 ESiJ

2. C6Hj.4 — ► C6H6 + 4H3, 7. C6Hj.4 — > C5H10 + CH4,
?. C6Hi4 — * C6H12 + H3, 8f 2C6H14 — + C5H13 + C7H1S,

4. C6H14 -> C6H12(cyclic) f H3, 9. C6H14-^ 6C + 7H3.

5. C6H14 -+C5H9CH3 (cyclic) + H3,

The maze of conceivable reactions may be simplified to a certain
extent by a thermodynamic consideration of the possibility of
each transformation at different temperature intervals. By means
of linear equations expressing the free energies of formation of
the hydrocarbons from graphite and hydrogen in the range 300° to
1000° A, the standard free energies accompanying the various
transformations ca.n be determined. The conclusions to be drawn
for a thermal decomposition at atmospheric pressure are: a re-
forming reaction (l) cannot proceed since conditions favor the
reverse reaction; cyclization and dehydrogena tion to benzene (2)
begin about 500° A and the equilibrium is well on the aromatic-
side at 600°A; dehydrogena tion to hexene (3) or cyclohexene (4'
is not appreciable below 800 A; cyclization to methylcyclopen-
tane (5) sets in about 350° A; cracking (?), disproportionation
(Q), and decomposition to carbon and hydrogen (9) are appreci-
able above 400° A.

-2-

IC9

Reactions 3,4,5 serve to produce substances which can form
benzene by further dehydrogenation so that successful solution
of the problem of aromatization of n-hexane hinges on the sup-
pression of reactions 7,8,9 which involve fission of the G-C
bond. The partial or complete elimination of these reactions
will depend upon the choice of catalyst. The catalyst, moreover,
must be specifically a dehydrogenation catalyst since the ener-
gies of the C-C and the C-H bonds are, respectively, 58600 and
87300 cal./mole. Such selective action has been realized in
mixtures of oxides of elements of groups VI, V, and IV (in order
of their effectiveness) of the periodic table. The most common
catalyst is amorphous Cr203 suspended on an inert carrier such
as alumina, magnesia, or silica gel but mixtures containing
molybdenum, vanadium, and sometijr.es titanium and cerium in ad-
dition to the chromium give the best results. The method of
preparation of these catalysts as well as the ratio of the dif-
ferent oxides is very important to the activity of the catalyst.

Because investigators have employed different catalysts,
reaction temperatures, and rates of flow, their results are not
usually directly comparable. Nevertheless, the following table
summarizes yields obtained in a few dehydrocyclizations for a
single passage of the hydrocarbon over the catalyst.

Hydrocarbon

2-Methylpentane
n-Hexane

2-Methylhexane
2,5-Dimethyl-
hexene

n-Heptane

Composition of liquid product in weight
per cent of starting material

Aromatic s

4.3
17.0 (benzene)

28.0 (toluene )

46.0

(80^ p_- xylene)

Naph-

Ole-

Parst-

thenes

fins

2.2

11.5

75.5

30.0 (toluene)
26
*12.1

2,6-Dimethylheptsne 82.0 (m-methylisopropylbenzene )

n-Octane 38.0 (55^ p_-xylene, 35^ c_-xylene, 5% m-xy-

lene, b% ethylbenzene )

n-Nonane 52.0 (greater than 90^ methylethylbenzene )

Butylbenzene 12.0 (naphthalene)

What would seem to be low yields of aromatics in the table
are in most cases satisfactory since the olefins and paraffins
appearing in the product are converted to aromptics on recycli^e 0
Thus, the latest experimental data on the dehydrocyclization or
n-heptane (starred drta in table), show that there was obtained
98.1 weight per cent of liquid products consisting of 12.1 weight
per cent of toluene. The calculated recycle yield (yield ob-
tained by repassing olefins and paraffins over the catalyst) of
toluene is 89 weight per cent or 97 per cent of the theoretical..

-o-

All paraffin hydrocarbons with the exception of methane
have "been converted to aromatice with the expected variations
in yield. Those hydrocarbons whose structure permits formation
of a six-membered ring are aromatized to a marked extent.
Branched chain paraffins having less than six carbons in a
straight chain give low yields presumably due to the diffi-
culty of the necessary isomer! zation to a six-membered chain.
Extent of aromrtization increases with the number of carbons
and generally the more ways a cyclization can take place the
greater the extent to which it occurs. Nevertheless, all
possibilities of ring formation are not realized to an equal
extent. It seems that the shortest sidechains (on aromatic
rings) are formed preferentially*

The nature of the predominating aromatic products is
governed by the reaction temperature. High temperatures fevoi?
formation of naphthalene and anthracene. However, the pre sen-:
value of the conversion of aliphatic s to pure aroma tics lies
in the production of toluene and benzene.

The arometization of paraffins may be conceived as taking
place in one of various ways. Thermodynamically, all are

Paraffin

\S i \

Naphthene <■ h Olefin

Aromatic

possible. Evidence indicates that the olefin is an inter-
mediate in the reaction and not merely a by-product. The next
step, cyclization, is considered to involve the double bond
in a 5 or 6 position. This is in harmony with the observation
that those olefins that have not undergone reaction and are
isolated in the product contain the double bond in a 2 or 3
position. These remain only because a shift of the double bond
must constitute a step preliminary to cyclization.

In no case have appreciable quantities of naphthenes been
isolated from the reaction product. This does not eliminate'
the naphthene structure as an intermediate in the proces*s
since it would be dehydrogena ted more easily than the other
reactants (probably before it could leave the surface of the
catalyst). Under a given set of conditions the extent and
ease of aroma tization from the following starting materials
increases in the order: paraffin< corresponding olefin <
six-membered naphthene < corresponding cycloolefin.

Aromatization of relatively pure hydrocarbons proceeds
smoothly without formation of large amounts of by-products.
On the other hand, the economic cracking of heavy petroleum
residues to aroma tics depends on the disposal of considerable
quantities of g°s, tar, "and carbon.

I I !

•4-

The latest variation of the dehydrocyclization reaction is
the hydro forming: process, a dehydrogenation carried out in the
presence of hydrogen. By this process, utilizing low octane
heavy naphtha, there may be obtained either a. product consist-
ing of 40 to 50 per cent aromatics (high grade gasoline) or
one having upwards of 80 per cent aromatics (toluene production)

Bibliography

GoXdwaeser and Taylor, J. Am. Chem. Soc. , 61, 1766 (1939).
Hoog, Verheus, and Zeiderweg, Trans. Far. Soc, 35, 993 (1939).
Taylor and Turkevich, ibid. , 35, 921.(1939).
Grosse, Morrel, and Mattox, Ind. Eng. Chem., 32, 528 (1940).
Kazan sky, Losek, Zelinsky, Compt. rend. acad. sci. URSS.,
27, 565 (1941); C.A., 35, 305 (1941). .

Reported by S. P. Rowland
December 10, 1941

112

TKIAZOLES: SOKE SYNTHESES AND REACTIONS

M. T. Bogert - Columbia University
E. Ochiai - University of Tokyo
T. B. Johnson - Yale University

Thiazoles contain a f ive-membered ring with two hetero-
atoms, nitrogen and sulfur, separated by a carbon atom. In num-
bering the ring, one begins with the sulfur atom and proceeds in
a manner such that nitrogen is the number three atom.

Recently there has been a great amount of interest in the
pharmacological properties of thiazoles. Johnson has synthe-
sized many thiazoles with a B, 4-dihyd poxy phenyl group in the
2-position. They possess definite analgesic and anesthetic ac-
tion and are equally effective when administered intravenously;
subcutaneously or orally. Three of the most recent chemothera-
peutic agents are thiazoles — 2-sulfanilamidothiazole (sulfa-
thia zole ), 2~sulfanilamido-4~methylthiazole ( sulfamethylthia-
zole ) and ?.-sulfanilamido-4-phenylthiazole ( sulfaphenylthiazole ).
They are superior to sulfanilamide and sulfapyridine in their
bacteriostatic action on pneumococci types I, II and III, on
p-hemolytic streptococcus A and on gonococcus. The 4-methyl
derivative is even more effective then sulfathiazole itself,
4-Methyl-5~( p-hydroxy ethyl )-thiazole is the sulfur-containing
portion of vitamin Bx.

The thiazole nucleus is important in the dye industry. A
yellow dye, primuline, war the first thiazole dye. There is
still a demand for Thioflavine T discovered by Green in 1889
because it is one of the few dyes giving yellow shades with a
prized greenish tone,

In 192.1 it was reported that mercaptobenzothiezole was a
very powerful vulcanization accelerator. At present the organic
accelerator most widely used in industry is mercaptobenzo thia-
zole, known to the trade as "Captax." Other popular accelerators
are its zinc salt (Zenite) and its benzoyl derivative (Ureka C)0
These accelerators have nearly flat vulcanization curves making
the vulcanization of thick rubber articles practical.

Syntheses of Mononuclear Thiazoles. — (l) The condensation
of a-halogen aldehydes and ketones with thioacid amides.

This, one of the most important methods, was worked out by
Hantzsch, The reaction is vigorous and is controlled by the
use of a diluent such as alcohol, water or ice.

2-Methyl-4-phenylthiazole is obtained from cc-bromoaceto-
phenone and thioacetamide.

-2-

I 13

HCBr

C6H5COK

HS,

HC— S

HN^

^C-CH.

C„HfiC— N^

,CCH3'HBr + H20

'6115'

Thiazoles in which the 2-position is unsubstituted ere
obtrined when thiof or .tip mid e are used.

HCC1

HCOH

HS,

*CH

HN'

HC-S.

HC -N'

The thiazole portion of vitemin Bx has been synthesized
by three groups of workers, Buchman^ synthesis, reported to be
the best, yields 8 - 12 grams of the thiazole from 100 grams of
ethyl acetoacetate.,

CH3C0CH2C00C2K5

H0CH2CH2C-Sv
CH3C-N/

0
NgOCaHs CH3COCHC^

/ >

CH2 CH2

S02C12

Cn2— CH->
0

0
CH3C0CC1C^

1 >

CH3 Ch2

HC:

*NH.

0 CI
CH3C~CCH2CH2OH
H

Severpl recent patents appear to be based on this synthesiso

Johnson has made 4~(a-chloromethyl )- thiazoles by using
symmetrical dichloroacetone and thiobenzarnide or its derivatives,.

HCCL HS

+ "CC6HS

HC — S.

ClCHoCOH

C1CH2C— N

CCSH5

(2) Condensation of cc~helogen aldehydes and ketones with
thioureas.

This synthesis wgs first employed by Traumann in the
preparation of 2-aminothiazole s. It is illustrated by the
synthesis of 2~amino~4-methylthiazole given in Organic Syntheses.

HCC1
CH3C0H

)PNH2

HC— Sv

CH„C— N^

CNH-

In a search for nevj sulfanilamide derivatives, Ziegler Iia s
used it recently for the preparation of 2-amino~4~?lkylthia zolt r

-5-

I 14

CE3COCERCOOC2E5

Br:

CS.

4 BrCH2C0GHRC00C2H5

NHsGSNKj

EC=CCERCOOC2E5
I I

NEa

NaOE

EC = CCERCOONe

EC1

RCHaC— S>

S N
\*

NH2

50-600

CNH-

EC— N^

(?) Treatment of a-acylated amino aldehydes and amino-
ketones with phosphorus penta sulfide,

This reaction was first used by G-abriel in the synthesis o:
2-methyl*-5-phenylthia zole from acetaminoacetophenone.

C«H=COCEPNHCOCE.

P2S

3^5

C«E*C

6lA5

170°

EC— N^

CCE-

Similarly:

RC0NECH2C0CH3

P2S

CE,C— S

S'-'S

EC — N^

Syntheses of Benzothia zoles. — (l) Condensation of p_-amino-
phenylmercaptans with acids, acid anhydrides, acid chlorides,
aldehydes and ae tones.

Eofmann developed this reaction for the preparation of ben-
zothiazoles. Condensing acetaldehyde with p_-aminophenylmercap~
tan results in the formation of 2-methylbenzothia.zoles.

+ CEoCEO

r V \

CCE, + EP + EP0

However, these mercaptans are frequently unavailable since
they ere usually prepared from the benzothiazoles and are often
unstable. With some modification, this method has been especi-
ally fruitful in the hands of Soger t and his coworkers.

N02C6E4SSC6E4N02

EOAc

■$»

-■

I 15

-4-

RCHO

or RCOC1

(2) Heating aroma tic amines and their derivatives with
sulfur.

Bogert and Abrahamson report that the fusion of benzanilic.e
with sulfur is the best method for the synthesis of 2-phenyl-
benzothiazole .

c6h5c;

NHCfiH

6Xi5

CCRH

6Xi5

Similarly, benzotniazole itself would be obtained from
f ormanilide.

The principal product from the fusion of p_-toluidine with
sulfur for the manufacture of thiazole dye intermediates is
dehydrothio-p_~toluidine .

CH3

£— > CH3C6H^ lCC6H4NH2(p_)

NH-

A larger molecule is obtained upon further heating with
sulfur » The sodium salt of the sulfonic acid obtained by the
action of fuming sulfur id acid on this molecule is primuline0
An entire series of dyes is obtained by diazotizing and coupling
primuline with amines and phenols.

(3) Rearrangement of compounds with the group CH-N-C-S-

1 A

Jacobson found that thioanilides and phenylthiourethans are
oxidized by alkaline potassium ferricyanide at room temperature
to benzothiazoles. Bogert and Mey?r used this method in tht
synthesis of 2-(p_-tolyl )-benzo thiazole.

C-NHC6H4CH3

^\/S

K3Fe(CN)

■■A

CC6H4CH3

-5-

CH3C-NHC6H6

K3Fe(CN),

CCH3

Properties and Reactions of Thiazoles. — The benzothiazoles
resemble the quinolines. They are stable to acids but yield an
c_-aminophenylmercapte.n end an acid when boiled with alcoholic
alkali or fusing with potassium hydroxide.

The thiazoles resemble the pyridines. Their aqueous solu-
tions react neutral and they form stable salts with acids which
have an acid reaction. They are oxidized by potassium perman-
ganate. They are unaffected by most reducing agents but sodium
amalgam and alcohol cleave the ring with the formation of an
amine and a mercaptan. 2-Aminothiazole s can be diazotized and
then treated as any primary aromatic amine. 2~Amino~4-elkyl~
thiazoles couple in the 5-position with diazotized amines to
form azo dyes. Two molecules of the 2-amino -4-alkylthiazole
will also condense with an aromatic aldehyde to form a compound
of the type

ArCHf-C — S

RC— N^

"CNH:

The chlorine atom in the 2-phenyl-4-(cc~chloromethyl )~thiazoles
is comparable to that in benzyl chloride. Ammonia will replace
it with an amino group at room temperature.

Most of the work on the substitution
zoles has been done in the last three year
coworkers. Bromine ti on s were conducted by
zole in ice cold dilute sulfuric acid with
it to stand overnight and by allowing the
overnight with chloroform and bromine. Tii
methyl thiazoles were unaffected while the
aminothiazoles gave the 5-bromo derivative
are obtained when thiazoles are bromine ted

at 250° and 400u
ben zothie zole is

The 2-bromo
brominated a t

derivative
450°.

reactions of thia-
s by Ochiai end his

treating the thie-

bromine and allowing
thiazole to stand
e 4-methyl and 5-

2-hydroxy and 2~
2-* Br omo thiazoles

in the vapor phase
is also obtained when

4-Methylthie zole cannot be nitrated, 2-Hydroxy-4-methyl-
thiazole gives the 2-hydroxy-4-methyl~5~nitrothie zole . The
analogous 2-a.minothiazole gives the corresponding 4-methyl-2-
nitremino-5-nitrothie zole. When a phenyl group occupies the
4- or the 5~position, a p_-nitrophenyl derivative is obtained.

4~Methylthiezole~5-sulfonic acid is obtained when 4-
methylthiazole is sulfonated. The 2-amino derivative yields
2~sulfemino^4-methylthiezole when sulfonated in the cold.

I 17

-6-

Further treatment produces the corresponding 2-sulf amino -4-
mrthylthiazole-5~sulfonic pcid.

A thiazole unsubstitut ed in the 2-position is obtained
when 2~mercaptothiazoles are oxidized by acid or neutral hydro-
gen peroxide. Oxidation with alkaline hydrogen peroxide pro-
duces the 2- sulfonic acid,

2-Hydroxy-4-methylthiazole undergoes the Friedel-^Crafts
reaction with acetylchloride to give 2-hydroxy~4-methyl-5-acetyl~
thiazole; with benzoyl chloride to give 2-hydroxy~4~methyl-5~
benzoylthiazole. 4-Methylthiazole would not undergo the
Gattermenn reaction but 2 ~hydroxy-4-m ethyl thiazole did, forming
the 5-eldehyde.

An amino group is introduced in the 2-position both of
thiazoles and benzothia zoles when they are treated with
sodamide at 140-150°.

Fox and Bogert have observed that 6~alkoxy-7-nitroben-
zothiazoles, in the presence of alcohols and small amounts of
dilute caustic alkali, exchange their alkyls for those of the
alcohol used.

Bibliography

Bogert and Meyer, J. Am. Ghem. Soc, 44, 1568 (1922).

Bogert and Abrahamson, ibid., 44, 826 (1922),

Fox and Bogert, ibid . , 63, 2996 (1941).

Olin and Johnson, ibid., 63, 1475 (1931 ).

Ochiai, Kakude, Nekayama and Masuda, J. Pharm. £>oc . , Japan,

59, 462 (1939); Chem. Abs., 34, 101 (1940).
Buchmrn, J. Am. Chem. Soc, 58. 1803 '1936),
Ziegler, ibid. , 63, 2946 (1941).

Lawrence, Proc. Soc. Exptl. Biol. Med. . 43_ 92 (1940),
Cadwell and Temple, A. C. S. Monograph, 74., Chaot, VIII.
Meyer and Jacobson, "Lehrbuch der Organisohen Chemie," Waltei

de Gruyter and Company, Berlin and Leipzig (1920), Vol. Z.

Part 3, p. 535.

Reported by G, W. Cannon
December 10, 1941

I 18

HIGH PRESSURE HYDROGENhTIONS OVER NICKEL AND
COPPER CHROMITE

Adkins et. al. University of Wisconsin

The use of hydrogen at high pressures and temperatures in
laboratory syntheses has been made practical by the work of
Adkins and his students-. For the most part, this seminar is con-
cerned with the results which they have obtained. In a previous
seminar, the reactions of organic molecules over platinum,
palladium, and other catalysts at low pressures were covered.
These data will not be included in this report.

Hydrogen reacts with organic molecules either by addition to
a multiple linkage (h.ydrogenat ion) or by cleavage of the molecule
(hydro genolysis) . For convenience, the reaction of hydrogen
with common functional groups will be considered before discussion
of the selective hydro genat ion of polyfunctional molecules.

The most widely used catalysts for these hydro genat ions are
copper chromite and nickel. Nickel catalyst is prepared either
on a support such as kieselguhr or as "Raney" nickel. It is a
highly active, versatile catalyst even at room temperatures and
at low hydrogen pressures. In many partial hydrogenations it
is difficult to stop the reaction when the calculated amount of
hydrogen has been taken up. In cases like these the "limited
(16) hydrogenation" procedure of Adkins and Durland can be used. This
procedure is carried out by filling the bomb with nitrogen at
100 atmospheres and then adding the calculated amount of hydrogen a
The nickel functions as a catalyst until the partial pressure of
hydrogen approaches zero.

Copper chromite (Adkins5 catalyst) 3s a highly useful,
stable catalyst prepared by the decomposition of copper ammonium
chromate. It is active at temperatures above 100° and at hydro-
gen pressures greater than 20 atmospheres, and is particularly
useful for the hydrogenation of esters to alcohols and amides to
amines. It is notably inactive for the hydrogenation of the
benzenoid nucleus and, therefore, may be used for the hydrogena-
tion of aryl compounds without much danger of saturating the ben-
zene ring. As an example, naphthalene can be hydrogenated over
copper chromite to tetralin, at which point the reaction stops.
With nickel as catalyst tetralin can only be obtained by in-
terrupting the reaction when the theoretical amount of hydrogen
has been absorbed. A review of the preparation and properties
of these catalysts is given in recent publications. (1,2)

-2-

! (9

Hydrogenation of the alkene linkage: In general the alkene
linkage is easily and quantitatively reduced. A selected list
of examples is given below. Of all the hydro carbons, the only
carbon-carbon bond which is easily cleaved is the ethane bond in
phenylated ethanes. For example, tetraphenylethane undergoes
hydrogenolysis (cleavage) of the carbon-carbon bond. The corres-
ponding cyclohexyl derivatives have a stable ethane bond*

Hvdrogenation of Alkenes

Products**

ethyl benzene
ethyl benzene
1,1,2,2,-tetra-
phenyl ethane

1,1,2,2-tetra-
phenyl ethane
cyclohexane

*Ni(k) - nickel on kieselguhr **The yields averaged 96-100$

Compound

°c

Time(Min. )

Catalyst*

styrene

NiU)

styrene

125

CuCrO

1, 1,2,2, -tetra-

100

130

Ni(k)

phenyl ethylene

1,1,2, 2-tetra-

150

CuCrO

phenyl ethylene

cyclohexene

165

CuCrO

(3)
(3)

(3)

(4)

Hydrogenation of carbonyl compounds: The carbonyl group in
aldehydes, ketones, esters, and lactones is of varying reactivity.
Hydrogenation of the carbonyl group in amides will be discussed
separately.

(a) Aldehydes and ketones generally react with hydrogen
under conditions as mild as those required for the reduction of
the alkene link. The yield of alcohol is practically quantita-
tive even with aldo or keto esters and alcohols. Carbonyl com-
pounds which are reduced to benzyl alcohols or to 1-3 or 1-4
glycols can undergo hydrogenolysis of the C-OH bond. This side
reaction is a function of the conditions; the higher the tempera-
ture and the longer the duration of the reaction, the greater will
be the amount of hydrogenolysis. For example, benzflldehyde is
reduced in 92$ yield to benzyl alcohol; however, reduction of
ethyl benzoate requires high temperatures and gives a quantitative
yield of toluene rather than benzyl alcohol.

o (b) Esters are hydrogenated over copper chromite at 200-
250 to give almost quantitative yields of the corresponding al-
cohols or glycols. Esters which are reduced to unstable alcohols
or glycols give lower yields. lb reduce diethyl succinate to
tetramethylene glycol (74$ yield) the reaction is carried out as
rapidly as possible in order to cut down the cleavage of the 1-4
glycol.

(c) 1-3 Diketones can be reduced either to the keto al-
cohol, or to the 1-3 glycol as well as undergoing hydrogenolysis
of the carbon-carbon bond. Keto alcohols have been isolated in
the reduction of a series of 1-3 diketones in 35-68$ yield. The
corresponding 1-3 glycols could be prepared in 50-90$ yield. If
the methylene carbon of the dlketone is substituted, hydrogenolysis

120

- 3 -

takes place quite readily. The labile carbon-carbon bonds in 1-3
diketones are shown by dotted lines in the formula below —

|l H

R— C . . . C .
H

50-90^

<JH H OH

* R-C-— £^C— R1
H H h

0 H OK

58~68^ -> R-C-C-C— R'

H H

(11)
(12)

Compound
acetone

cyclopentanone
acetophenone

aaetophenone

H.vdrogenation of Carbonyl Compounds
T°C Time Catalyst Yield

Product

125
100
150

13 m
6 hr
30 m

Ni(k)
Ni(k)
CuCrO

110 lOmn Ni(R)

100

propanol-2 (3)

cyclopentanol (6)

phenylmethylcarbi-
nol (7)

diphenylcarbinol (6

1,2-diphenylethyl-
ene glycol (b)

benzyl alcohol C4)
toluene (5)
toluene (8)
ethylbenzene
diphenylme.thane (4
hexanol-1 (8)
octadecanol-1 (9)

tetramethylene gly-
col (10)
butyrolactone

hexanediol-1-6 (10

2-isopropylbutane-
diol,l-4 (10)

2-phenylbutanediol?

1-4
3-phenylbutanol-l
(10)

The functional group in cyanides, imines, hydro xylamines,
oximes and nitro compounds is readily reduced under mild condi-
tions to give good yields of amines. Formation of secondary
amines is a side reaction in the reduction due to the reaction of
an imine with the primary amine. This secondary amine formation
can be reduced to a large extent by carrying out the reduction

benzophenone

160

1 hr

Ni(k)

benzil

125

1 hr

Ni(k)

benzaldehyde

180

1 m

CuCrO

benzylalcohol

125

10 m

Ni(k)

ethyl benzoate

250

—

CuCrO

100

acetophenone

175

6 m

Ni(k)

benzophenone

175

1 hr

CuCrO

ethyl caproate

250

5 hr

CuCrO

ethyl stearate

250

3 hr

CuCrO

diethyl succinate

250

1-1/2

CuCrO

74
18

diethyl adipate

250

—

CuCrO

diethyl a- methyl
succinate

250

3 hr

CuCrO

diethyl a-phenyl
succinate

250

6 hr

CuCrO

12
67

\2\

_4_

rapidly, and by using methanol- ammonia as solvent. By this
procedure alkyl cyanides can be reduced to primary amines in
yield as high as 95$.

R-C=N

K
R-C=NH

R-CHa-NH:

H H ^

R-C-N-CHa-R
I

NH3

H
R-CHa-N-CHs-R

n-butyl cyanide 125

n-butyl cyanide
benzal aniline
benzal aniline

methanol-ammonia

65 5m Ni(k)
175 25 m CuCrO

30 m Ni(R) 67$ n-pentyl amine (8)

16 di-n-pentyl amine

95 n-pentyl amine (13)

9? phenyl benzyl amine

97 phenyl benzyl amine .

(15)

Reduction of amides to amines: Although this hydrogenation
requires drastic conditions (temperature 250-265°, pressure 200-
300 atmospheres) the yield of amines is very good. With substi-
tuted amides, the yields range from 80-95$ while with simple
amides the yield is lowered due to formation of secondary amines
at the high temperature necessary for the reduction. In addition
to reduction of the carbonyl, the linkages shown by dotted lines
in the formula below are subject to hydrogenolysis. For the most
part side reactions due to cleavage of these bonds are small* (l)

0=C,..N...R*

i H

Reduction of Amides (CuCrO catalyst)

Lauramide
Heptamide
Kept amide

250
250

250

Salicylamide 250
Lauroylpiperadine 250

N-diethylheptamide 250

42 m

4 hr

2 hr

1 m

1 hr

n-duodecylamine (10)
di-n-duodecylamine
n-heptylamine (10)
di-n-heptylamine
q-cresol (10)
n-duodecyl alcohol (10)
N-n-duodecyl piperadine
n-heptylethylamine (10)
n-heptyldiethylamine
dl-n.-heptylamine

Reaction of glycols with amines: Secondary and tertiary
amines can be prepared in 30 to 70$ yields by reaction of gly-
cols with amines in the presence of copper chromite. When 1-4,
1-5, or 1-6 glycols are used the corresponding pyrrolidine,
piperidine, and hexahydroazepine is produced. The yield of the
seven member ring compound (hexahydroazepine) is low (17$), but
that of pyrrolidines and piperidines averages 50-75$.

■

••■

•

■ .

' -

.. ..

■':■

^B^fmm

122

CH3v.

R-NH3 + HO-CH3-(CHs)n-CH20H -> R-N ' /(°Hs)n

ur«3

Pentane 1-4 diol 250 CuCrO 60 l-n-amyl-2-methylpyrroli«-:.
n-amylamine dine (17)

hexane 1-5 diol 250 CuCrO 75 l-n-amyl-2-methylpiperi-
n-amylamine dine (17)

Hydrogenation of N-substituted amides of dicarboxylic acids :
When the amides are prepared from 4,5,6-carbon chain dicarboxylic
acids, hydrogenation gives excellent yields of pyrrolidines,
piperidines, and hexahydroazepines. By this method £ and fr
substituted piperidines can be prepared by starting with the
appropriate substituted glutaramide.

0 0 '

H M i , , h H ^CH2-C

R-N-C-C-C-C-C-N-R -» R-H l-C"

I I I VCH3-S'

R amyl, benzyl, phenylethyl

Adipamides cyclize in this reaction to give hexahydroazepines in
35^ yield

In the case where the piperidine amide of these acids is
reduced, ring closure is impossible and straight chained substi-
tuted cadaverines are formed,

CH3-CH3 || jj CH3-CH3v
CH3 >-C-(CH3) -C-Nsx CH3 -* py-CH3-(GH3)n-CH3-py
^CR3-CK3 CH3-CH3

N-n-amylsuccimide 200 30 m CuCrO 79 N-n-amylpyrroli-

dine (10
glutaramide 250 2 h CuCrO 70 piperadine (14

di-N-n-amyl- ^ , £ dimethyl 250 4-1/2 CuCrO 69 l-n-amyl-4 , 4 » -
glutaramide h dimethyl piperi-

dine (14)
£-methylglutarimide 250 1-1/2 CuCrO 50 4-methyl piperi-
dine (14)

Hydrogenation of pyrrolldones and pioeridonas: (14)

CHa-C . CH3-CH3

f ^N-R -* I ) N-R

CH3-CH3 CH3-CH3

Cri3— C * CH3— CH3

CH3 N-R -* CKa SN-R

NCH3-CK3 NCHS-CHSX

It is not impossible that these cyclic imides would be in-
termediates in the hydrogenation of amides of dicarboxylic acids
to cyclic amines. However, the yield of cyclic amines by

■

-6- 123

reduction of these compounds is less than the overall yield from
the straight chained amides. The piperidones and pyrrolidones
are prepared in excellent yield by partial reduction of the
appropriate cyclic imides. Nickel is used as a catalyst in
this reduction,

0 .0

s CHa-C* Ni „ CH3-C*
CHSn 7J-R r-* CH3^

CH2-C; 220° NCH2-CH3
0

Preparation of amines from aldehydes and ketones: The re-
duction of aldehydes and ketones in the presence of ammonia or
amines (Mignonac's method) is a very satisfactory preparative
method for secondary and tertiary amines.

butyraldehyde 125 2 hr Nl(k) 70 dicyclohexylamine (15)

piperidine 93 N-n-butylpiperadine (15)

cyclohexanone 125 2 hr Ni(k) 70 dicyclohexylamine (15)
cyclohexylamine

In hydrogenation reactions today the critical problem is not
better, but more selective hydrogenation. By selective hydro-
genation is meant the use of the proper catalyst and conditions
so as to bring about the preferential hydrogenation of the
important functional groups in the presence of each other.

Control of reaction: Success in selective hydrogenation de-
pends primarily upon the selection of the catalyst, temperature,
and duration of reaction, and secondarily upon the medium of
reaction.

(a) Catalyst: Proper choice of a catalyst is perhaps
the one most important factor to be recognized in choosing con-
ditions for a selective hydrogenation. For an example, copper
chromite is relatively inactive toward benzenoid nuclei, hence
aldehydes, ketones, esters, and amides containing an aryl

group may be hydrogenated to the corresponding alcohols or amines
containing the benzenoid ring. Nickel, on the other hand, is in-
active toward oxygen-containing groups as the amido and
carbalkoxyl groups, so compounds of these types containing aryl
groups may be converted to the corresponding amides or esters
containing cyclohexyl groups. Other selective catalysts will
be encountered later.

(b) Temperature: In the discussion of hydrogenation re-
actions it was noted in many instances that one functional group
reacts at a temperature sufficiently below that required for
another functional group to make selective hydrogenation easily
attainable. An example of this is the selective hydrogenation

of the alkene linkage in preference to other unsaturated carbon-
carbon linkages as in benzene, etc.

124

~7-

phenylethyl alcohol ie obtained only by interrupting the hydro-
genation before all the ester has undergone the first step in
the following reaction: (29)

250°

(1) C6H5CH3C00C2H5 + 2H3(CuCrO) > C6H5CH3CH30H + C2H50H

(2) C6H5CH3CH30H + H3(CuCrO) 2-50° > C6H5CH3CH3 + HOH

(d) Modification of reaction medium: If diethyl furfural
acetal is hydrogenated two principal products are obtained:

(1)

>-CH(0CaH5)3 + 2K:

I > CH(0C3HB)3

(2)

CH(0C3H5)3+ 3Hj

CH30C3H5 + C3H50H

Over a nickel catalyst reaction (2) takes place to give the
ether in 95$ yields. However, if a little sodium carbonate,
sodium ethoxide, or any one of several amines is added, reaction

(l) takes place and the saturated
yields. (5)

.cetal is obtained in 30-80

Alkynes

Alkenes

Smides

10.

Oximino

11.

Nitro-nitroso

12.

Cyanides

13.

Aldehydes

14.

Structure and ease of hydrogenat ion : Attempts to summarize
the relative ease of hydrogenat ion of a series of monofunctional
compounds will embody contradictory facts which have no
immediate explanation. However, on the basis of present data,
the following qualitative list may be proposed:

Ketones

Furanoid

Pyridinoid (subst.)

Benzenoid

Pyrroioid (N-subst.)

Esters

Amides

The relative ease of reaction of two compounds with different
functional groups taken separately is not a safe basis of pre-
diction as to the relative rates of hydrogenation of the two com-
pounds in a mixture or of the two groups when both are in the
same molecule. d-Alpha pinene is hydrogenated much more rapidly
than cinnamic acid, yet in a mixture of the two, cinnamic acid is
completely hydrogenated to the exclusion of the pinene. Likewise
in a single molecule the functional group which taken alone is
most reactive is not always the most active when the molecule
contains a second functional group.

Selective hydrogenations:

1. Hydrogenation of unsaturated esters: Despite the
fact that the carbon to carbon double bond in alkenes is

hydrogenated under as mild conditions as is any other functi
group, and despite the fact that the carbalkoxyl group requi
a temperature of over 200° and pressures exceeding 100
atmospheres, certain unsaturated esters have been hydrogenat
to the corresponding unsaturated alcohols using zinc chromit
as a catalyst. (18)

o
Butyl oleate 300 11 h ZnCrO 65
Butyl erucate 300o 11 h ZnCrO 68

125

onal
res

ed
e

octadecenol

ducosenol

2. Hydrogenat ion of benzenoid type corn-pounds : The
benzenoid nucleus requires more drast j c conditions for hydro-
genation than most other functional groups.

1 h Ni(R) 100 eye lone xane (30)
5 m Ni(R) 100 me thylcyclohexane (30)
4 h Ni(R) 92 1, 3 , 5-trime thylcyclo-
hexane (30)

Benzene 150
Toluene 175°
Mesitylene 200°

9-10 dihydro-
phenanthrene
90$

*VScP"

sym-octahydro-

phenanthrene

85$ jJHiiRL.

t e trade cahydro-
phenanthrene
85-90$

(a) Selective Hydrogenatioh of Phenanthrene (16)

' (19)
dodecf hydro-

>i phenanthrene

a s ymm- oc t ahydr o-

s^s phenanthrene

y Ni(R) v 30$
r3TJo->

1,2,5, 4-tetra-
hydrophenanthrene
40$

5. Hydrogenat ion of furanoid
genation of the furanoid nucleus occurs
conditions than those used for the benz
differs in that the furan ring is a eye
ject to hydrogenolysis.

o
2-Kethyl furane 150

o
Furfuryl alcohol 125

o
Furoin (in EtOH) 150

type compounds: Hydro-
under somewhat milder
3noid ring and also
lie ether and hence sub-

1 h

Diethyl furfuryl
acetal (l g.
amyl amine )

175

Ni(k)

2.5h Ni(k)
0„5h ^7i(k)

83
85
93

5 h Ni(k) 76

2-methyl tetrahydro-

furane (20)
t e tr ahy dro furfuryl

alcohol (21)
1; 2-dihydroxy-l, 2-
di t e tr ahydr o furfuryl
ethane (21)
diethyl tetrahydro-
I'urfural acetal (21)

-9-

126

4. Hydrogenation of gyridinoid type compounds: Pyri-
dine requires a somewhat higher temperature for conversion to
piperidine than does the transformation of benzene to cyclo-
hexane , However, derivatives of pyridine are generally more
readily hydrogenated than are derivatives of benzene. This is
evidenced by the preferential hydrogenation of the pyridine ring
in compounds such as quinoline and phenyl pyridine.

Pyridine compounds with substituents
benzyl, etc., in the 2 or 2,6-positions a
at lower temperatures than is pyridine it

due to the effect of
of the nitrogen to

the

"poison",

sub st it ue
i.e..

nt a
to

pyridine
3-acetyl

pyridine

2-methyl pyridine
2-benzyl pyridine
2, 6-dicarbethoxy
pyridine

o
200
145°

200
100(
137 '

7 h
4.5h

0.6h
3 h
O.Olh

Ni(R)
Ni(R)

Ni(R)
Ni(R)
Ni(R)

com

83
28
61

90
85
66

, viz., carbethoxy,
re hydrogenated rapidly
self. This i s no doubt
lowering the tendency
bine with, the catalyst,

Refs. (22)(4)
piperidine

3-ethyl piperidine
3-piperidyl methyl ke-
tone
2-methyl piperidine
2-benzyl piperidine
2, 6-dicarbethoxy
piperidine

Copper chromite is also active for hydrogenation of the
pyridine ring in compounds such as quinoline.

5. Hydrogenation of i&yrroloid type compounds: Pyrroles
are far less reactive toward hydrogen than are the derivatives
of benzene, pyridine, or furane. The 2,3,4 and (or) 5-alkyl
pyrroles react with hydrogen over nickel or copper chromite at
200-250°. However, the 2,3,4 and (or) 5-carbethoxy pyrroles re-
quire such drastic conditions th:. t hydrogenation of the nucleus
does not occur without simultaneous hydrogenation and hydro-
genolysis of the carbethoxyl group. On the other hand, pyrroles
bearing a carbethoxy or aryl group on the nitrogen atom readily
react with hydrogen from room temperatures upward.

Because of the resistance of the nucleus toward hydrogena-
tion, acyl pyrroles have been converted to alkyl pyrryoles in e
cellent yields over copper chromite. Similarly carbethoxy
pyrroles have been converted to methyl pyrroles although careful
control of the reaction is necessary so as to avoid excessive
formation of pyrrolidines.

c.
d.

h
h

Pyrrole 180 1

1-Phenyl 135° 1

pyrrole 0

1-Carbethoxy 70 0.3h

pyrrole 0

2,4-diacetyl- 160 0.3h
3,5-dimethyl

pyrrole 0

Ni(R)
Ni(R)

Ni(R)

CuCrO

47
63

2,4-dicarbeth-
oxy-3,5-di-
methyl pyrrole

220 0.3h CuCrO 53

pyrrolidine (25)
1-phenylpyrrolidine (23)

1-carbethoxypyrrolidine

(24)
2,4-diethyl-3,5-
dimethyl pyrrole (23)

3-carbethoxy-2,4,5 tri-
methyl pyrrole (23)

2,3,4,5-tetramethyl
pyrrolidine

-10-

127

f. 1, 2-dicarbeth-
oxy pyrrole

200

1 h

CuCrO 42 2-methyl pyrrole (25)

. <_ a &« — Hydrogenation of substituted amides; As previously
oon P«AohydJ0ge£ under 200-300 atmospheres reacts with amides at

°I °ner the i?flue2ce of copper chromite to give amines.
A study of tne reaction of various substituted pyrrolidones and
piperidones and open chain amido esters with hydrogen offers an
excellent example of selective hydro genat ion. (25)

o=c.

CuCrO

N /~C00C3H5 210-220 30 m.

o=c

IT
H

93^

o=c

H
CONC5H

1 1

/ Np-C00C3H5

■CH,

C5Hn

CuCrO

210°

35 m.

o=c

H
sN/*-CKaN-C6Hlx

CuCrO

240-250" 20 m 0=C

H
r—C0NCBHxl

0-C

CuCrO

235-240° 60 m

C5H1X

CHpOH

F 7 — VI

\E,

NG5H.L1

'— CH,0H

C 5 H ! x

C8H50C0CH3CH3C-^

x —

CuCrO 57/^ KOCH3CH3CH2CK2-N
230 35m 20^ H0CH3CH3CH3CH30H

^ti/fnmth«%ab0Vf^ata.il: iS observed ^at a carfcethoxy or
group in the 5-position in a pyrrolidone-2 is more reactive J
hydrogen than is the lactam group of the ring. In the"r>1-er:
rtp£wVhl8 taT relationship holds with respect to the r^rhl

ttlol^ZLi* Tth the araid° ^ivatives, preferential h/£
ation of the lactam group in the ring occurs. In the oren >i

cm^?t!nJerS\bf h the 6Ste^ gT0Up a*d amid0 group undergo
was fouSf fL h^^gen?1!i0S iand h^ogenolysit) since no evic
was found for either the hydroxy amide or the amino ester.

am i do
toward
1. c? i n 9
sthoay

■O.c p;>-

•c.ln
lence

123

(26)

•11-

D'l^nni and Adkins/have shown that under the influence of
either fUney nickel or copper chromite various N-pentamethylene
amides having a hydroxyl group in the a, ft %X t oppositions are
converted to the corresponding amino alcohols in yields of 51-
79$. In the case of the ^-hydroxy amide the hydroxyl group is
eliminated and the chief product is the alkyl piperidine.

7. Hydro e; enat Ion of P.vrones (27)

.A/V

•CpH

3iA5

-^ O V

/y vcaH,
w92,

2-ethyl chromone

\X/-CH3(CH2)3

CH;

Flavone (2-phenylchromone ) can be hydrogenated to the
corresponding series of compounds.

(28)
8. Hydrogenation of 6-ketoni trilea: Wiley and Adkins/
have found that the products obtained from the easily reduced
ketonitriles depend upon the temperature used in the hydrogena-
tion process, At 35-40° over nickel the ketoamines were iso-
lated in 10-60$ yields, but hydrogenations carried out rapidly
at 150-250° gave 30-60% yields of the aminoalcohols. Hydro-
genolysis also occurred to some extent at 150-200°.

Bibliography;

Adkins "Reaction of Hydrogen" University of Wisconsin Prsss.

Madison, Wisconsin, 1937. (.I.-*

Adkins and Shriner, Gilman chapter (in press). (°.)
Adkins and Zartman, J. Am. Chem. Soc, 54, 1668 (1932) (;>'

Adkins and Connor, ibid., 53, 1091 ( 19317. (•*)

Adkins, Covert, and Connor, Ibid, 54, 1651 (1932). ,(5j

Adkins and Cramer, ibid., 52, 4349 (1930). /6)

Covert MS Thesis, University of Wisconsin v7)

-12-

Burdick, MS Thesis, University of Wisconsin
Adkins ana Folkers, J. Am. Chem. Soc, 54, 1145
Wojcik, ibid. , 55, ■ 4349 (1933).
and Sprague, ibid., 56, 2669 (1934).
and Stutsman, ibid. . 61, 3303 (1939).
and Schwoegler, ibjdH 61, 3499 (1934).
and Paden, ibid., 5J., 2487 (1936).
and Winans, ibid. , 55 , 2051 (1934).
Durland, ibid, 60, 1501 (1938)*
Hill, ibid., 60, 1033 (1938).
Sauer, ibid., 59, 1 (1937).
Durland, ibid., 59, .135 (1938).

129

Adkins,

Adkins

Adkins,

Adkins

Adkins,

and

and
et

and

dnd

and

Wojcik and Covert,
Zartman and Cramer,

Connor, ibid. . 54, 4678 (1932,
Eurdick, ibid., 56. 438 (193'
al., ibid., 56, 2425 (1934).
Signaigo, ibid., 58, 709 (1936).
Rainey, ibid., 61, 1104 (1939).

Sauer, ibid., 60, 402 (1938).
D'lanni, ibid., 61, 1675 (1939).

Mozingo, ibid. , 60
Wiley, ibid. , 60

669 (1938).
914 (1938).
ibid., 55,
ibid., 53

1669
1425

(1933).
(1931).-

(8)
(1932). (9
(10

(11
(12
(13
(14
(15
(16
(17
(18
(19
(20
(21
(22
(23
(24
(25
(26
(27
(28
(29
(30

Reported by C. M. Himel

R. C. G-unther
December 17, 1941

I »

130

THE CHEMISTRY OF ORGANOBORON COMPOUNDS

Organic compounds containing boron comprise a field which
for many years has received relatively little attention, and is
still generally unfamiliar. la recent years interest in these
compounds has growr.'? for several reasons. For one thing boron is
situated next to carbon in the periodic table and hence is very
similar in effective nuclear charge ant atomic radius.; therefore,
one may expect studies of organic boron compounds to be of value
in interpreting the behavior of analogous carbon electronic sys-
tems. Another emphasis is placed on boron compounds because of
the ease of effecting nuclear disintegration of the boron aton,
Kruger has recently shown that when neoplastic tissue which has
been impregnated with boric acid is bombarded with slow neutrons
the tumor cells are killed very effectively by the boron disin-
tegration products. His experiments were carried out in vitro . bu
he concludes that the same results could be obtained in the living
body if the boron could be applied to the diseased tissue. This
suggests the possibility of developing organic compounds contain-
ing boron which may be specifically absorbed by the tissue.

Therefore, it is the purpose of this report to briefly
outline the preparation and properties of the different classes
of organoboron compounds., with reference also to certain mechanism
explanations.

I. Trial^yl and Triaryl Borings (R.^B)

A. Preparation* The first organoboron compounds were pre-
pared by Frankland in 1862: he obtained trimethyl- and triethyi--
borine by the interaction of zinc alkyls and ethyl borate. Ifrai.se
and Nitsche have more recently obtained good yields of the berimes
by the action of Grignard reagents on the etherate of boron
trifluoride. Johnson, Snyder, and Van Carapen have also produced
these compounds from methyl borate and Grignard reagent s; but
find the BF3 method preferable,

3ZnR3 + (EtC)3B ~~^-> R33 ■-'- ?7.nPX'Et

3RMgX + Et20:EF3 .'-a— » R,!E •■? l<EzXJ -'- Et3C (80# y\-ld)

3RMgX + (MeO)3B -l&—> RSB + 3MgX0Me {50% yield)

(R = alkyl or aryl)

B. Properties: The aliphatic borine s have been stucied
more thoroughly recently. Trimethyl borine is a gas - the only
gaseous organometallic compound; the higher members of the
series are colorless liquids.

(l) These compounds are so easily oxidized that they are
spontaneously inflammable in air. (2) They are stable toward
water, in sharp contrast to thy organic derivatives of the

13!

- 2 -

neighboring elements Be and Al. (3) With dry HBr a hydrocarbon
is produced by the removal of one alkyl group:

R3B + HBr -> R3B-Br + R-H

In the presence of water the boron bromide is hydrolyzed to
R2B-OH, but with anhydrous HBr the reaction is clear-cut, and
the dialkyl boron bromide produced is a new type compound which
may be important in the synthesis of other new organoboron com-
pounds, (5) With dry Brs more complex reactions take place,
with some substitution of bromine in the alkyl groups forming;
HBr which may then react with the borine as shown above
however, it is definitely shown that two alkyl groups aia.- s^iii
off. This will be mentioned later. (6) As would be expected.,
NH3 forms an addition complex; a more unusual reaction is the
addition cf the alkali metals - the latter is specific for t;
aromatic borine s, it seems.

'OP

II. Boronic and Borinic Acids, Alkyl Boron Oxides [RB(OH)s,
R5BOH, RBO]

1. Aromatic

A. Preparation: These were first prepared by Michaelis
and Becker in 1880 by use of an organomercurial. Several bettei
methods have been developed since then: Krause obtained the
acids by careful oe-cdation of tria..'"vj. bonines ore-oared frcn
BF3 and iUSgX-

R3B 4^~* R2E~0R --'- RE.(TK)8 •-— ** fiE'-">H;c

Khotinsky and Me lame ci used methyl borate and aryl magnesium
bromide,

RMgBr -: 'LleC}33 — ; RD(CMe)P ~3a£u RB(0Il)a

Konig and Scharrnbeck modified :.".:. ;, procedure by the use of i-.o-
butyl borate, and Bean and Johnson y Lie use of n--butyi borate.
The last method is probably the b.^j • of the lot.

Diaryl borinic acids are f 01 ned as by-products in the
above reactions.

B. Properties: The acids are crystalline white Solaris,
soluble in organic solvents, sparingly soluble in water They
are fairly stable substances, for benzene boronic acid has been
nitrated and the nitro compound reduced to the amine without
cleavage of the 0-3 bond by Johnson and his coworkers. This
bond may be cleaved, however, upon warming with metallic salts
such as HgClg, Zn012, and by aqueous H2Os or bromine water.
They do not undergo atmospheric oxidation. Upon treatment with
ammoniacal Ag20 the hydrocarbon R-H is formed.

•

- 3 -

2. Aliphatic

A. Preparation: Frankland first prepared an aliphatic
boronic acid by the careful oxidation of triethyl borine,
followed by hydrolysis. They are now prepared by the same ,
methods used for the aromatics. Snyder, Kuck, and Johnson have
obtained the best results using the method of Khotinsky and
Melamed, with slight modification, to prepare a series of ali-
phatic boronic acids. The borinic acids may be prepared by the
action of aqueous HBr on trlalkyl borines, but are very easily
dehydrated, giving R3B-0-BR3. Their esters may be obtained by
careful oxidation of the borine, as shown above.

B. Properties: The aliphatic boronic acids are much
weaker acids than their aromatic analogs, as evidenced by the
fact that the latter can be titrated with standard alkali upoi
the addition of mannitol, whereas the former cannot. They are
not cleaved by reagents which cleave the aromatics (HgCl3,
Br3-H30), but do resemble them in being cleaved by H303. They
undergo autooxidation in the air very rapidly.

RB(OH)3 H*°* , ROH * H3B03 Hs°s' * = ^ °r a"yl
air air, R = alkyl only.

They react quite differently with ammoniacal Ag30 also, giving
the R-R hydrocarbon instead of R-H:

2n-C4H9-B(0H)3 + Ag(NH3)3 + 2H30 -> n-CeHie + Ag + 2H3B03 + 2NIL

An interesting change takes place if the aliphatic borer?-;...;
acids are dried over P30s or concentrated H3S04;

RB(OH)3 -HoO, RBO

The produce ere called alkyl boron oxides. Although they b^v?
the comj /vsvticn RBO, molecular weight determinations show ii.am
to be trimeri - This suggests a cyc.lrc ':^t >. .er ara."' o^ous to the
paraldehyde .-. r?hi-.a it is seen that the aliphatic corcnic acids
show a defini'-e similarity to the al-phatio aldehydes^ {x.) Th ;;
combine with mvv.ecular oxygen, (2) the}' .-.-duce aAQ!oaf.ar-?l j*g?0.
(3) they undergo cyclization to a pVienberea ring structure.
This may be attributed to analogous 3?:.ectrt*nio configurations.

To shew ::he distinction between s.rooa.tic and allpr.nu: o
boronic acids 3 Gill more clearly, Johnson Van demp-in, aid
Grummitt pr^pavec' benzyl-, t-butyl •■ ?■ furyl-, and'" 2-tb.i^ny2 -
boronic acids md tested them with Ag(KH, ;."'.' and aii! :>iiat^ci.
The first two a ere shown to react definitely as aliphatic, tbi
last pair as aromatic - confirming the wel- knowr a^.'macio
character of these groups.

III. Mechanism Studies

Mention was made on page 2 of the reaction of dry bromine
with trlalkyl borines, and it was noted that the significant
point of the reaction is that two alkyl groin™ *rP m^vp^ fVnm a

• •

--•

' ■ I

133

- 4 -

This dibromide formed (RB-Br3) must come from the action of
bromine on the monobromide (RaE— Br) since the latter has been
shown to be inert toward KBr.

These reactions may be considered as proof of the theory
that many organic reactions take place through a mechanism in
which one group acts as an electron acceptor, the other as a
donor. Here, the electron shell of the B atom may be completed
by an unshared electron pair of a Br atom. This would cause a
surge of electron density in the direction B — ■> R, and increase
the mobility of the potential alkyl anion; an irreversible
a - y shift within the complex would complete the reaction.

' R 1 R

:Br: -* B:R j -» :Br:3:R f R:Br:

• • • • > •

; Br : R

L " J

Failure of HBr to effect cleavage of more than one group may be
attributed to the diminuished acceptor activity of B after the
compound R3B-Br is formed, due to an internal resonance effect:

• • • • •©

;Er:Br: + B:R
* • « • • • •

R:B:3r: e ^ R:3::8r:
R* •' R*

Br 2 undergoes coordination even with this weaker acceptor
effective enough to produce sufficient mobility of the alkyl
group, but HBr does not.

Investigation was carried further in this field by Johnscrj
and Van Campen, who studied the results of oxidation by aqueous
H303 and autooxid^tion in air:

R3B -^-U R3BOR ■ (°< > RB(OR)3
I %i

They concluded that these reactions also go through a coordina-
tion complex, because they found that the reaction would be
stopped at stage I in the presence of water, and would not go
at all with the ammonate of the borine (R3B £-> NH3 ) .

A survey of the behavior of various alkyl boron compounds
toward HBr, Br3, H303, 0S, etc., indicates a definite gradation
of reactivities; R3B <R3BOH >RB(OH)3 , and this is nicely ex-
plained by reference to resonance possibilities within the
molecule, leading to lessened reactivity.

IV. U se_ of. _D_lazonlum Borofluorides in Synthesis f

Dunker, Starkey, and Jenkins have s.aown that the diazonium
borofluorides may be readily prepared as follows:

Ar_NH3 + NaN03 + 2HBF4 > Ar-N3BF4 + Na3F4 + 2H30 (90-97$

00 yields)

They were interested in these compounds because when they are

■ '

- 5 -

134

obtained. More recently, Starkey has used these diazonium com-
pounds for synthesizing molecules in which the nitro group is
ortho or oara to a nitro, carbonyl, or similar group. For
example, his method is included in Organic Syntheses for the
preparation of p_- and o-dinitro "benzenes in yields of 67-82$
and 33-38$ respectively. His reactions are carried out according
to the following equation:

N0a

NO.

NaNOg -i- Cu pow d e r

efficient
stxrring

N03

+ Na + NaBF4

Bibliography:

Frankland,

Krause and

J. Ohem.
Nitsche .

Soc.

Ber.

15,

363 (1862).
2784 (1921);

55, 1261 (1922).
Johnson, Snyder, and Van Campen, J. Am. Chem. Soc, 60, 115

(1938)
Michaelis and Becker, Ber,, 13, 58 (1880); 15, 180 (1882).

ibid.f 42, 3090 (19095*.

J. prakt, Chem,, 128

Khotinsky and Kelamed,

Konig and Scharr'nbeck,

Bean and Johnson, J. Am, Chem. Soc., 54,

Snyder, Kuck, and Johnson, ibid.

153 (1930).
443.5 (1932).

60, 105 (1941).
G-rummitt, ibid, ,

0 111 (1938).

Johnson, Van Campen, ana

Johnson and Van Campen, ibid.

Dunker, Starkey, and Jenkins,

Starkey, ibid., 59, 1479 (1937*57

Organic Synthesis, Vol. 19, p. 40.

Kruger, Proc. Natl. Acad. Sci,, U. S., 26, 181 (1940).

60, 121 (1938),
ibid., 58, 2308 (1936).

Reported by Royston M. Roberts
January 7, 1942,

CARDIAC AGLYCONES OF THE STROPHANTHIDIN GROUP

135

W. A. Jacobs - Rockefeller Institute for Medical

Research
R.C. Elderfield - Columbia University

Plants of the digitalis-strophanthus group contain
glycosides which have a characteristic and powerful action on
cardiac muscle. The nucleus, which is common to all the agly-
cones of this group, is given below (l):

HSC-C

C=CH

Recent evidence indicates an a,, p unsaturated lactone.

The following table lists the best known aglycones.

Aglycone

Grouos

Ca-OH/R*

Rings A/B*

Plant Source

Digitoxigenin

trans

cis

CH3

Digitalis

Thevetigenin

cis

CH3

Thevetia

Uzarigenin

cis

trans

CH3

Uzara tree

Digoxigenin

11, 14

trans

cis

CH3

Digitalis

Gitoxigen

14, 16

trans

cis

CH3

Digitalis

Periplogenin

5, 14

trans

cis

CH3

Periploca

Sarmentogenin

11, 14

trans

cis

CH3

Strophanthus

Strophanthidin

5, 14

trans

cis

CHO

Strophanthus

Probabl

e structures

In the glycosides the sugar residue is linked through the

C3-0H.

In 1934 Tschesche elucidated the cyclopentanoperhydro-
phenanthrene ring system by selenium dehydrogenation of mono-
anhydrouzarigenin to Diels' hydrocarbon (II) and also by degrada-
tion of the same material to etioallocholanic acid (II!

COpH

III

- 2 -

Correlation of the above results
principally by W. A, Jacobs from
infer the complete structures of

with the extensive work,
1922 on, made it possible to
the more important aglycones.

136

The highly important and characteristic lactone group may
be titrated with alkali at moderately elevated temperatures.
It absorbs one mole of hydrogen to give a saturated lactone.
The 3, Y unsaturation was inferred from the behavior of the agly-
cones on treatment with sodium nitroprusside (Legal* s test) and
ammoniacal silver solution (Tollens* reagent) which resembled

that of ^^,V angelica lactone (V) and not that of ^ a» ^
angelica lactone (IV).

CH=^CH

H3CCH C=0
^ 0'

CH
»
H3CC

so'

CK3
i

c=o

Therefore, the side chain appeared to be the lactone of an
enolized aldehydo acid. Upon saponification and then acidifi-
cation or merely by treatment with alcoholic alkali without
saponification, a characteristic rearrangement of either the
aglvcone or the unhydrolyzed glycoside to a saturated isomer
(VI) takes place. Jacobs inferred that it was preceded by a
stereochemical inversion on C17 by which the lactone and the
C14-OH, originally trans, assumed a cis relationship.

rt, ^ 9 ?H2

c=o

KOH

CH-

.0=0

/wK

CH CH:

0-4-CH c=o

-o"

The structure of the isomer (VI) is well established.

Evidence for the positions of the other functional groups
of strophanthidin follows:

C14-0H. One molecule of water is split out on mild treat-
ment (alcoholic HCl) of strophanthidin. The anhydrogenin cannot
be isomerized. The C14 position alone allows the hydroxyl to
be both tertiary and y or J with respect to the aldehyde group
of the side chain,

C3-0H. The secondary hydroxyl has been located at C3 by
degradations involving the opening of ring A,

. ...

■

137

- O -

C10-CHO, Wolff-Kishner reduction of a strophanthidin
derivative gave a derivative of periplogenin. Conversion of
this to a derivative of digitoxigenin was accomplished as
follows:

.OH

CrO:

y^\ .,

0 OH

Hft0

^N^X/

Digitoxigenin has been degraded to etiocholanic acid (a
stereoisomer of III).

C5-OH. A second tertiary hydroxyl has been demonstrated.
The extremely easy dehydration of the hydroxyketone in the
previous transformation suggested a § hydroxyketone. To be
both tertiary and p to C3 the OH must be attached to C5.

Recently Elderfield has considered the synthesis of cardiac

aglycones by transformations of other steroids. His attempts to

f*> v
prepare model j=> substituted jx* unsaturated lactones have

thrown new light on the structure of the lactone ring in the

aglycones. His more successful syntheses made use of the fact

that half-ethers of primary-tertiary glycols are converted

smoothly into aldehydes when heated with acids;

CsHlxMgBr + CH3OCH3CN -* CeHj. iCOCHsOCHg

CsHnCOCHaOCHa + BrCH3CC2C2H5

KHSOj

SweT

^6^1 i^H

-* C6H11C(OH)CH2C03C3H£

CeHnCH
CHO

VII

a^y-Saponff^- CH30CH3

tion

saponification
then HBr

HOCH CO , .

^0s C6H5-C(OH)-CH C6HnC

VIII 1 t + I

.CH

C=0

CH.

IX \ HBr * X

It was not possible to dehydrate VII to an unsaturated lactone.

The following considerations have led Elderfield to

suggest that the lactone in the aglycones is really /£
unsaturated:

138

- 4

The comparison of the lactones of Keto acids (angelica
lactones) with lactones of aldehydo acids is not justified. X
and strophanthidin gave identical color reactions with a modified

6 Y

Legal test and with ferricyanide while £v ' angelica lactone

differed remarkably. £\a>fc angelica lactone gave only a slow
response to Legal ' s reagent.

Representative aglycones, X, and ethyl crotonate have
similar absorption spectra. Vinyl acetate (comparable to

Z\ lactones) differed greatly.

Hydroge nation of X and the aglycones yielded exclusively
the saturated lactones. When the double bond is at the point of

lactonization (as in the A^' type), varying amounts of
desoxy-acids are obtained.

The action of alkali on X parallels that noted in aglycone
derivatives.

^6^1 i~C

-CH

CH-

KOH in

CH3UH

ao. NaOH

Cs^i iGH

CHO

CeHn-C:

■CHa

COOH

=CH

CHoOH COCNa

The isomerism of the glycosides and aglycones by alkali may be
formulated as follows:

=rCH

CH3 / CO

/CH

' o'

CH3
I

c=o

CH.

CH — CH;

I !
CH CO
v 0'

Here the lactone is assumed cis to the C14 hydroxyl and no in-
version on C17 is necessary. The observed failure of the gly-
coside uzarin and of allostrophanthidin to isomerize is
probably due to a trans configuration. Allostrophanthidin is
produced by the action of an enzyme found in strophanthus
seed upon strophanthidin. Allostrophanthidin has been shown to
be a diastereoisomer of strophanthidin for which C17 or C14 is
responsible.

Bibliography:

G-ilman, Organic Chemistry. Wiley, New York, (1958) pp. 1314-1337,

Elderfield et al., J. Org. Chem. , 6, 260 (1941).

Chen, Robbins, and Worth, J. Am. Pharm. Assoc, 27, 189 (1938),

Reported by R. F. Phillips

• .•

139

PREPARATION OF NITRILES BY THE USE OF CUPROUS CYANIDE
J. F. Koelsch, University of Minnesota

Cuprous cyanide has been used as a reagent to produce nitriles

by the replacement of halogen atoms in aryl halogen compounds,

vinyl halides, allyl halides, and acyl halides. These reactions
will be discussed in order.

The Rosenmund-Von Braun nitrile synthesis,

2 ArX + Cu2(CN)2 -> 2 ArCN + Cu2X3

is best known. Reactions reported in the literature vary widely
in experimental conditions. Koelsch found that iodo- and bromo-
diphenylindones mixed with cuprous cyanide and heated at 240-250°C«
for three hours gave nearly quantitative yields of the nitriles.
Fieser and Seligman found that when pyridine was added to a mix-
ture of 4-chloro-7-methylhydrindene and cuprous cyanide the tem-
perature necessary for reaction was lowered from 265° to 225°C,
the time of heating could be cut from eighteen hours to two hours,
and the yield of nitrile was higher. Newman has given the most
detailed directions for the use of pyridine in the Rosenmund-Von
Braun synthesis. Quinoline has also been used as a solvent to aid
reaction.

Recently Koelsch and Whitney carried out a study to determine
the mechanism of the transformation involved in this synthesis,
and optimum conditions for the reaction. Though the reaction could
be simply written, it was not free of complications. Reaction be-
tween cuprous cyanide and p_-bromo toluene was only 15 per cent com-
pleted after sixty minutes at 250°, but was 75 per cent completed
in the next thirty minutes. This induction period indicated an
auto-catalytic reaction.

The catalyst might be the aromatic nitrile, forming a complex
with cuprous cyanide and bringing the solid into solution. To
test this assumption p_-tolunitrile was added to the reaction mix-
ture. Small amounts increased the rate of reaction; larger amounts
were less advantageous. The induction period was not entirely
eliminated.

The authors believed that peroxides might act as anticatalytic
agents. These if present at the start of the reaction would be
destroyed during its course. Accordingly, small amounts of hydro-
quinone were added. The induction period was almost doubled.

. .

-2-

140

This suggested the use of an oxidizing agent. Cupric sulfate
was found to have a pronounced catalytic effect. The actions of
both cupric sulfate and p_-t©lunitrile on the reaction of p_-
(C6H5 )3CHC6H4Br with two equivalents of cuprous cyanide are shown
in the graphs below. The rate of reaction for this compound
without catalysts is also shown, illustrating the induction period,

10Q

<h
P
O

cti

p pn

c •

CD
Ph

0 I.

100 r

0 90 180 270 *60

Time, minutes

p_-(C6H5)2CHC6H4Br
2 equiv. CuGN

ctf

50J

, ,v

0 0.2 0.4 0.6

C7H7CN or CuS04, grams

0.5 g. p_-(C6H5)3CHC6H4Br
with I. p_-C7H7CN, 45 mins
II. CuS04j 60 minutes

A mechanism was suggested to explain the catalytic effect
of cupric sulfate in the Rosenmund-Von Braun synthesis:

ArX + Cu

(ArX ->Cu)++ + Cu+

(ArX ->Cu)

Ar+ + CN"

(ArX-^ Cu)++

(ArX ->Cu)+ + Cu++
Ar+ + CuX
ArCN

This mechanism rests on assumptions that; (l) The aryl halide
does not form a complex through the halogen atom with cuprous
copper; (2) The cuprous complex can decompose to give an aryl ion

The study led to the formulation of a practical synthetic
procedure. To a mixture of an aryl halide with excess cuprous
cyanide is added a few drops of tolunitrile, and a little cupric
sulfate* The whole is placed in a bath at 250°C. Completion of
the reaction, as indicated by marked dimunition in volume of
solid copper salts and formation of a dark liquid phase follows
rapidly (ten to thirty minutes).

-3-

14 i

The nine compounds used in Koelsch' s experiments were rated
as to relative reactivity with cuprous cyanide by measuring the
time required for fifty per cent of a substance to react with
cuprous cyanide. The order of reactivity was: p_-bromotriphenyl-
methane < m-bromotoluene <p_-bromobenzophenone / p_-bromotoluene <^
bromobenzene <_ bromomesitylene / cc-bromonaphthalene / p_-bromobenzoic
acid.

Many examples in the literature show that bromine or iodine
is much more easily replaced by the nitrile group than is chlorine.

The reaction of vinyl halides with cuprous cyanide should be
similar to the reaction of aromatic halogen compounds. Koelsch
found that triphenylvinylbromide heated two hours with cuprous
cyanide at 240° gave a quantitative yield of triphenylacry lonitrila.
However, cc~p_-bromophenyl-|3, p-diphenylvinylbromide under similar
conditions gave only a black resin and at lower temperatures did
not react.

Allyl cyanide has been made by shaKing allyl chloride for
eight days with a concentrated solution of potassium cyanide. The
yield was ten per cent. Bruylants obtained a 95 per cent yield
by refluxing allyl iodide or bromide one hour with a five per cent
excess of cuprous cyanide,

ct-Ketonitrile s have been prepared from the lower members of
the fatty acid series in sixty to eighty-five per cent yields by
refluxing the acyl bromide and cuprous cyanide, without solvent,
for one and a half to two hours. This is the method of
Tschelinzeff and Schmidt. Previously silver cyanide had been used
with acid chlorides. An autoclave was necessary, and there were
many side reactions.

Bibliography:

Koelsch and Whitney, J. Org. Chem,, 6, 795 (1941).
C. F. Koelsch, J. Am. Chem. S0c . , 58, 1528 (1956).
Fieser and Seligman, ibid. , 58. 2482 (1936).
Newman, ibid, . 59, 2472 (1957).

Bruylants, Bull. soc. Chim. Belg., 51, 175 (1932).
Tschelinzeff and Schmidt, Ber., 62, 2210 (1929).

Reported by A. V. Mcintosh, Jr.
January 14, 1942

•

142

the reduction of multiple
carbon-carbon bonds

Because of the ease of replacement of the hydrogen atoms of
acetylene by sodium or by the MgX group, substituted acetylenes m?.j
be readily prepared in a pure state. These should afford a source
of the corresponding olefins, provided that a satisfactory reagen'
or catalyst for partial reduction can be found.

The ordinary catalysts, nickel, palladium, and platinum, in
their various forms, have been used successfully for this process
Zalkind has reduced a great many substituted acetylenic glycols
using the noble metal catalysts and in all cases has found that
there is a definite break in the curve at half-reduction. Campbeli
and O'Connor have continued some of the earlier work of d'uPont us:;;;

Raney nickel on variously substituted acetylenes. Their results
may be summarized as follows:

Type Half-reduction

R-C=C-H Very slight change in slope

R-C=C-R» Noticeable change, especially if R = R-

Ph-C=C-H and Ph-C=C-Me No change

Ph-C=C-Ph Stops at one mole

Analysis and Raman spectra indicate that there is better than 99.5$
olefin at the half-way point.

Less familiar methods include the use of chromous chloride,
"Raney" iron, and sodium in liquid ammonia. The iron catalyst,
prepared exactly as Raney nickel, was developed by Paul and Hilley
and was found to be almost specific for triple bonds.. It failed in
only a few isolated cases;, for example, tolane is completely re-
duced to diphenyleth-.ne in the presence of this catalyst . Thompson
and Wyatt have extended the use of the catalyst to the reduction of
various alkynes, enynes, and dienynes.

Alkynes

OK OH

' _ / o

CH3-C-C=C-C~CH3 150 C 80$ olefin

CH3 CH3 14°0#

H
i

H-C^C-C-OH 100 C 2nd mole of H enters

1000# one fourth as fast as

, i . first

(CH3)2

EnyneS QH 100°C

CHa=C-C=CH 1000# 50$ isoprene

R 100°C

R-CH=C-C=C-R 1000// Stops at diolefin

- 2 -

Dienynei

r_C=C-C=C-C=C-R

H H K H
(R is GHa or C6H1X)

100 C
1000#

'f ?

R-C=C-C=C=C-C~R

143

Sodium in liquid ammonia is a reducing agent developed by
Lebeau and Picon and has been used to prepare olefins, especially
trans olefins, by Campbell and Eby. Several trans octenes, hexenes,
and decenes as well as butadiene have been prepared from the corres-
ponding acetylenes by this method. They recommend it, incidentally,
as the most satisfactory for obtaining terminal ethylenic compounds,

may be

As regards the stereochemical course of reduction reactions, it
said that in general catalytic methods lead to cis forms
whereas chemical methods produce trans modifications. Bourguel,
G-redy, She rr ill and Zalkind have all used metal catalysts and have
obtained cis compounds in 90-100$ yields. The melting points of
known compounds and the Raman spectra data on the unknowns have
been offered as evidence.

Interesting work in this field has been done by Schroter, Ott,
and Farkas. Shroter found that by using the same batch of catalyst
Over and over in the reduction of ethyl acetylenedicarboxylate, the
ratio of ethyl furuarate to ethyl maleate gradually increased.
Changing the amount of catalyst or the introduction of a catalyst
poison such as carbon monoxide or hydrogen sulfide diminished the
rate of hydrogen absorption but did not change the nature of the
product. This led Ott to postulate in a later paper that the
production of the more labile form is favored by an increase in
the activity of the catalyst. He recognized that the velocity of
the reaction on a catalytic surface is not measurable in terms of
hydrogen absorbed. He assumed that the activity of catalysts is
similar in nature to the reduction potential of the ordinary re-
ducing agents. A metal of low potential should give a stable trans
form, one of higher potential would give more of the cis isomer,
and one high in the electromotive series which gives hydrogen of
extremely high potential would carry the reduction all the way to
the ethane. Propiolic acid was reduced chemically by the following
combinations with the results indicated:

100% trans cinnamic acid

Mn-Alkali 50% cis. 4^ trans, and an oily mixture
Mg, Na, etc. gave mixtures of the saturated hydro-

Cr-HCl

Zn-alkali 90^ trans
50% cis.

cinnamic acid and unreduced propiolic acid.

Thus between zinc and manganese lies the critical area, within which
ordinary catalysts probably belong as far as reduction potential or
activity is concerned.

Farkas and Farkas have also reviewed the literature and
summarized the results rather briefly. Catalytic methods add two
hydrogen atoms simultaneously, thus giving a cig form; nascent

- 3 -

hydrogen on the other hand adds stepwise giving the opportunity for

able

modification to result
form is the

and with most oiefinic

the fum^roid form the

the more

pairs the malenoid form is the more labile,

more stable. This may be carried further into the reduction of

olefins to ethanes and in general the me so form resembles the cis

and the racemic modification the trans. Most reductions conform

to the following scheme:

Modification reduced

C=C

C=C (cis)

C=C (trans)

If the temperature is
catalytic reduction m-
modification.

Product from
Nascent hydrogen Catalytic hydrogen

trans
dl-pa ir
me so

CIS

me so
dl-p^ir

too high or if enolization is possible, the:
y lead partially or totally to the trans

So far, the work on these methods of partial reduction has
been confined to the laboratory only. The industrial research in
this field has been limited almost entirely to attempts to prepare
butadiene from vinylacetylene. Most important patents are to du T ••"»
Pont for the catalytic reduction of vinylacetylene, to Jasco for
reduction by metallic zinc in alkaline solution, and to I. G-. Far-
benindustrie for the reduction by sodium and zinc in the presence
of butyl naphthalene sulfonate, Hurukawa has published results in-
dicating a 60/6 yield of the diene from an electrolytic reduction at
a palladium black cathode* There is also a process for the semi-
reduction of most of the common acetylenes by means of palladium
on clay and there is another patent for the reduction of several
substituted dienynes to mono-olefins in the presence of nickel.

Bibliography

Zalkind, Vishnyakov and Morev, J. Gen. Chem.
Campbell _.nd O'Connor, J. Am. Chem, Soc, 61,
Campbell and Eby, ibid, . 63, 216 (1941).
Thompson and Wy:-.tt, ibid. . 62, 2555 (19*0).

(USSR), 3, 91 (1933).
(1939).

2897

Sherrill and Launspach, ibid.

2562 (1938).
rend.. 157, 137 (1913).

Lebeau and Picon, Cou4pt.
Bourguel, ibid., 180, 1753 (I925~T7
Paul and Hilley, ibid., 206, b08 (1938); Bull. Soc. Chim. [5]
6, 218 (1939).

[5] 2, 1029 (1935).
er.. 60B, 624 (1927).

67B, 1669
Farkas and Farkas, Trans. Far. clay Soc,
Hurukawa, J. Electrochem. Soc, Japan,

G-redy, ibid.

Schroter,

Ott, B^rth and Clems er, ibid. .

USP

(1934).
33, 887 (1937).
7, 346 (1939).
2,207,070 cf. C.A. 34, 7932' (1940}
1,920,242 27, 4818 (1933).

2,167,067 33, 8625 (1939

834,111 33, 3393 (1939

837,196 33, 6872 (1939

Reported by John C. Robinson, Jr.
January 14, 1942

. .

145

POLYENES AND CUMULENES
Richar* Kuhn and Coworkers - Kaiser Wllhelm Institut

REVIEW OF POLYENES

Kuhn defines 'polyene' as referring to those substances th
contain many ethylenic linkages in open chains. It is usually
understood that the double bonds are conjugated.

Methods of Synthesis. —

Type A - C6H5£CH=CH3-nC6H5 (n = 1,2,3,4,5,6,7,8,11,15)

1. Treatment of hydrobenzoins with phosphorous iodide (P2I4):

2 C6H5CH=CH-CHO -^~* C6H5CH=CH-CH-CH-CH=CHC6K5 5 4 >

6h OH

C 6H 5CH=CH-CH=CH -CH=CHC 6H 5

This method is suitable for 1,4- and l,6-glyc«ls also. They are
prepared by the action of BrMgC=CMgBr and BrMgC=C-C=CMgBr on
polyene aldehydes.

2. Condensation of polyene aldehydes with h00C-CH2£CH=CH3ftCKa-
COOH (n = 0,1,2) using lead oxide:

C6H5CH=CH-CHO + CH2— CH2 + OHC-CH=CE-C6H5 ->

COOK COOK

C6Hs-CH=CH-CH=CK-CH=CH-Ch=CH-C6HB

3. Action of benzylmagnesium chlorides on pnlyene aldehydes:

(a) C6HBCH=CH-CHO + CH3-CH=CK-CHO H|||r|dine )

C6HsfCH=CH3.5CH0 and C6H5f CK=CK9-7CH0

(b) C6H5-£CH=CH}nCHO + C6H5CH2MgCl ->
C6K5fCH=CK}-nCHOHCK2C6H5 -> C6HB£ CH=CH}-n+1CeH5

4. Coupling of thio- and seleno-aldehydes:

C6H5£-CH=CH]-CHO HsS(H3Se)> C6K5£-CH^CH1 CHS °U> CaC°3 >

piperidine

CeHsfUH-CH^^j^CeKs

-2- 1 46

In this manner, C6HS£CH=CH3-15C6H5 was prepared. It is the
highest known member of the polyene series. The solid is green-
ish black and its solutions are violet-red.

Type B - CH3£CH=CH}nCH0 (n = 1,2, 3,4, 5, 7, 9) and

CH3-ECH=CH>nCOOH (n = 1,2,3,4,5,6,8)

Aldehydes, chiefly crotonaldehyde are condensed by means
of piperidine acetate to give higher polyene aldehydes,
CK3£CH=CH}nCHO. These aldehydes will condense with malonic
acid:

—CO
CH3fCH=CH}nCHO + CHa(COOH)a -> CH3-£CH=CHi-nCH=C(C00H )a ■ s >

CH3-ECR=CH}n+1C00H

By this method the total synthesis of stearic acid and cetyl
alcohol was effected.

Type C - CH3£CH=CH}nCH3 (n = 1,2,3,4,6)

Polyene aldehydes react normally with alkylmagnesium ha-
lides to form the carbinols. These are dehydrated by treating
with a l-2# solution of p_-toluene sulfonic acid in ether:

-Ha0
CH3£CH=CH3nCH0 + CH3CHaMgBr -► CH3GCH=CH^CH0HCH3CH3 Z >

CH3fCH=CH9n+1CK3 (n = 2, m.p, 52°)

Type D - H03C£CH=ClQnCOOH (n = 1,2,3,4,5,7) and
H03C-CH3-£CH=CH3nC00H (n = 1,2,3,4)

1. Claisen Condensation:
EtOaC-C03Et + CH3tCH=CK^nCOaR

Et02C-C0CHa£CH=CH3nC0aR

pyridine
AcaO

Et02C-C=CH£CH=CH4 COaR A1» H£ >
OAc

Et03C-CHtCH=CH3 CHaC05R R'OH, NaOH >

OAc

R'OaCiCH=-CHan^C03R»

147

- 3 -

The free oxalo-polyene carboxylic acids are obtained by
saponification. Treatment with hydrogen peroxide removes carbon
monoxide :

H02C-COCK2-CH=CH-COCH K3°s » H02C-CH2-CH=CH-C02H

2. Eiological Oxidation:

Kuhn discovered that if polyene mono-carboxy acid amides
were fed to rabbits, the mono-amide of the corresponding dicar-
boxy acid would be obtained from -the rabbit's urine:

CH3£CH=CH33C0NHa -> H02CtCH=CH33C0NH2

Yields of 20-80 per cent have been obtained.

CUMULENES

Cumulenes are compounds containing an uninterrupted series of
double bonds.

Methods of Synthesis: —

The first method used by Kuhn was the treatment of acetylene
or diacetylene glycols with P2I4:

(c6h5)2c-c=c-c=c-c(c6h5)2 PgI*> (C6H5)2C=C=C=C=C=C(C6H5)2

(0.2-0.3^)

The yields were greatly improved by treating (I) in ether solu-
tion with dry HC1 and VC12 or CrCl2. Yields of around 90 per cent
were obtained. In this manner the following cumulenes were pre-
pared:

Ryy

_. C— C— C=?C— C— Cx ■ n

H
R = CI
OH,

=c=c=c=c=c

148

- 4 -

Attempts were made to prepare unsymmetrical cumulenes in
order to test out the van't Hoff theory. Unsymmetrical benzo-
phenones were treated with 3rMgC=CMgBr and the corresponding
acetylene diols obtained. These, however, could not be converted
into the butatrienes.

Kuhn attempted to prepare cumulenes from diphenylketene by
the following reactions:

2(C6H5)2C=C=0 + BrMgC=CMgBr -> (C6H5 ) 3C=C-C=C-q=C (C6H5 )

HG1

CrCl3

( C6H5 ) 3C=C=C=C=C=C ( C6H5 ) 3

Instead of (II), however, he obtained a diol of some other struc-
ture, as yet undetermined*

Properties of Cumulenes ~.na lolyenes :

As compared to the corresponding saturated compounds, polyenes
and cumulenes are more highly colored, have higher melting points
and lower solubility, and have absorption bonds in the higher
wave lengths.

Cumulenes are much more highly colored than polyenes and have
higher melting points. They give a negative Baeyer test, do not
react with maleic anhydride, and give no color reaction with
C(N03)4. They are destroyed by K202 and are readily reduced.

Bibliography:

Richard Kuhn, Angew. Chem., 50, 703 (1937). A review.
Richard Kuhn, J. Chem. Soc., 605 (1938). A review.
Kuhn and Wallenfels, Ber., 71B, 783 and 1510 (1938).
Kuhn and Platzer, Ber., 73B, 1410 (1940).

Reported by Stanley B. Speck
January 21, 1942

149

REACTIONS OF PYRIDINE

Pyridine is a well-known organic base which occurs in small
amounts in coal tar. It is a weaker base than ordinary tertiary
amines but does form soluble, stable salts. It may be considered
more aromatic than benzene, as shown by the oxidation of quinoline,
and may be compared in many of its reactions and properties to
nitrobenzene.

It is, however, more easily reduced than benzene. The
selective reduction of phenyl and benzyl derivatives of pyridine
has been accomplished by sodium and alcohol and by hydrogenation
over Raney nickel. The salts may be reduced by Adams1 platinum
catalyst. The ease of reduction is quite different for isomeric
alkyl or aryl derivatives. Thus it has been reported that 4-
phenylpyridine is not reduced with Adams' catalyst under conditions
which effect reduction of 4-benzylpyridine. With Raney nickel
the following is the order of decreasing ease of reduction of the
pyridine ring: 2-benzyl, 2-methyl, 2-phenyl, pyridine, 4-phenyl.
The reduction and alkylation of pyridine to methylpiperidinium
formate by refluxing a mixture of methanol, formic acid, and
pyridine has been reported.

Pyridine undergoes a number of substitution reactions, few
of which may be carried out except under strenuous conditions.
In general, the product formed by chlorination, bromination,
iodination, sulfonation, or nitration is the three or the three-
five derivative. However, conditions and catalysts have a pro-
nounced effect upon the nature of the product. Under the strenuous
conditions necessary to obtain any substitution at all it is diffi-
cult to prevent the simultaneous formation of some di- and poly-
substituted derivatives. The mechanism of the substitution of
pyridine is not known but sufficient data are at hand to show that
it is not a simple one.

The sulfonation of pyridine is best carried out in the
presence of mercuric salts with fuming sulfuric acid at a tempera-
ture of 270°. The main product is pyridine-3-sulfonic acid, al-
though some of the 2-derivative is reported. No di sulfonic acids
are produced. The compound (I) has been prepared by treatment of

S02
CgHsN' j
(I) ^0

pyridine with (S03)3 and has been proposed as an intermediate in
the sulfonation reaction. A few reactions of (I) which show the
nature of the problem encountered are listed* It is noteworthy
that one of the best methods for the iodination of pyridine in-
volves the use of fuming sulfuric acid as an oxidizing agent.
In this case the predominant product is 3-iodopyridine.

- 2 -

150

90%

NaOK

cold

SO,ONa

C^fUN/

'SllB

'oh

CH=CH-N-S02ONa
CH=CH-CHOH

CH-CHO

H
CH-CH=CHONa

V 79$

H20
^ u-

CH-CH=NS02ONa
CH-CH=CHONa

CH-
VI

80$

-> C5H5N + Na2S04
60$

-CH=N

= CH

41$

If pyridine is chlorinated at 170-220 , 3- and 3,5-deriva-
tives are formed while at 260-420°, 2- and 2, 6-derivatives are
formed. When bromination is carried out in the absence of metallic
salts, some of which tend to lower the temperature necessary for
substitution and affect the course of the reaction, definite
transition ranges are noted. Thus at 300° the 3- and 3,5-deriva-
tives are formed; at 400° a mixture, the 2-, 3-, 2,6-, and 3,5-
derivatives are obtained; while at 500° the product is mainly 2-
and 2,6. It is apparent from these facts that some sort of a
polar mechanism as opposed to a free radical type mechanism must
apply and that possibly the character of the nitrogen atom is
changed when the higher temperatures are reached. The formation
of 2-chloropyridine by treatment of pyridine with sulfuryl chloride
has been advanced as evidence that substitution in this position
takes place on some sort of an N-oxide or peroxide derivative.

Nitration of pyridine produces practically negligible yields
of 3-nitropyridine but if the pyridine ring is substituted it is
possible to obtain the nitro compounds in yields as high as 90 per
cent.

o
When pyridine is heated with sodamide to 160-180 2-amino-
or 2,6-diaminopyridine is obtained depending upon the ratio of
the reactants. This reaction may involve an addition of the
sodamide across the C=N linkage.

Treatment with mercuric acetate produces only the 3-mercuri-
derivative under mild conditions and tars under more strenuous
conditions. An intermediate, probably VII, is first formed.

0C0CH3
CSH5N /
VII \HgOCOCH3

151

- 3 -

In general the orientation effects of groups already present
are the same as the same groups in nitrobenzene and good yields of
the substitution products may be obtained if the substituent is
one which activates the ring.

Cases of 1-2 and 1-4 addition to the bond system of pyridine
are encountered when pyridine is treated with phenylmagnesium
bromide or phenyl lithium.

Since all three derivatives are isolated in most cases, free
radical substitution probably does occur when pyridine is treated
with diazotized aromatic amines in alkaline solution*

The comparison to nitrobenzene may be extended to the
activity of the groups' already present, Thus halogen in the 2-
or 4-position is activated much the same as in p_-chloronitrobenzene,
Methyl or alkyl groups in these positions undergo active methylene
condensations with aldehydes and esters and this method may be
used to separate p-picoline from the mixture which is obtained
from coal tar in the same manner that m-nitrotoluene is separated
from the p_- and p_-isomers.

The treatment of amino pyridines with nitrous acid gives
unstable diazonium salts which behave in replacement reactions
much like the corresponding benzene derivatives. The conversion
of 5-chloro-2-aminopyridine to 5-chloro-2-nitropyridine in 40^
yield by treatment with sulfuric acid and hydrogen peroxide has
been reported.

The other amino pyridines may be obtained from the corres-
ponding halogen compounds by treatment with ammonia in an auto-
clave. 3-Aminopyridine may be obtained by the Hofmann degradation
of nicotinamide and 4-aminopyridine is obtained by treatment of
4-pyridyl-N-pyridinium dichloride with ammonia (VIII).

S0C1

CI
I

2C5H5N bU"15) y9 ^N — ^ \\N-HC1 + SO

VIII 94^
Some reactions of VIII are shown:

•4-aminopyridine + pyridine
41%

VIII — > 4-hydroxypyridine + pyridine

150° 84%

\ CfiH50K

4-phenoxypyridine + pyridine
56%

152

- 4

Pyridine-2-sulfonic acid has been produced by oxidation of
the thiol obtained by treatment of 2-bromopyridine with sodium
hydrosulfide. The sulfonic acids may be fused with alkali, metal
hydroxides, or cyanides to produce the corresponding hydroxide or
cyanide.

The hydroxy derivatives raay be produced by tiia&oti nation of
the amine, by alkali fusion of the "sulfonic, acid 'or by hydrolysis of
compounds of the type VIII.

The preparation of nicotinonitrile by the heating of 3-
bromopyridine with cuprous cyanide has been reported. Nicotinic
acid may also be produced by the treatment of 3-bromopyridine with
butyl lithium followed by carbonation and hydrolysis, a process
which is reported to produce yields as high as 70 per cent.

The treatment of 2-bromopyridine with magnesium apparently
produces a polymer to which structure IX was assigned. It was

''/-Br

MgEr

possible to produce phenyl-2-pyridylcarbinol in 50 per cent yields
by treatment of this polymer with benzaldehyde.

A table showing the yields and isomers obtained by some of
the substitution reactions of pyridine is included.

TABLE

Reaction

Product

% yield

Remarks

Ref

chlorination

2-chloro
2,6-dichloro

46
1

copper tube
270^

bromination

3-bromo

3,5-dibromo
polybromo

39
12

mercury salts

pumice

300°

iodi nation

3-iodo

3,5-diiodo

fuming sulfuric

sulfo nation

3-sulfonic acid

mercuric sulfate
230°

nitration

3-nitro

18# fuming HN03
160-180°

14
i

- 5 -

Table (Cont)
Reaction

Product

% yield

Remarks

153

Ref.

amination

phenylmagne sium
"bromide

ethylmagnesium
bromide

phenyl lithium

sodium benzene
diazotate

benzyl chloride

mer curat ion

2-amino
2,6-diamino

2-phenyl

2-ethyl

2-phenyl

2-phenyl
3-phenyl
4-phehyl

2-benzyl
4-benzyl
2,6-dibenzyl
2,4-dibenzyl

3-mercuri

50
51

40-50

24
10

trace
trace
trace

autoclave

110 -toluene

benzyl chloride
heat and copper
powder

160-180

15
16

19
20

Bibliography

Soc, 56, 2425

1. Adkins, Kuick, Farlow and *tfojick, J. Am. Chem
(1934),

2. McElvain and Bailey, ibid. . 52, 1633 (1930).

3. Adams and Hamilton, ibid. , 50, 2260 (1928).

4. Overhoff and Webarit, Rec. trav. chim. , 50, 957 (1931).

5. Mayo, J. Org. Chem., 1, 496 (1936).

6. Baumgarten, Ber., 59B, 1166 (1926); 72B, 567 (1939).

7. Machek, Monatsh, 72, 77 (1938).

8. van Gastel and Wibaut, Rec. trav. chim., 53, 1031 (1934).

9. Rodewald and Plazek, Ber. , 70B, 1159 (19377.

10. Wibaut and Nicolai, Rec. trav. chim., 51. 709 (1930).

11. Colonna, Chem. Abstracts, 34, 7290 (1940).

12. McElvain and Englert, J. Am. Chem. Soc, 51, 863 (1929).

13. Wibaut and Kertog, Patent, Chem. Abstracts, 27, 4542 (1953).

14. Plazek, Ber., 72B, 577 (1959).

15. Wibaut and Dingernanse, Rec. trav. chim., 42, 240 (1923).

16. Shreve, Reichers, Rubenkoenig and Goodman, Ind. Eng. Chem., 32
173 (1940),

17. Shreve, Skeeters and Swaney, ibid., 32, 360 (1940).

18. Bergstrom and McAllister, J. Am. Chem. Soc, 52, 2845 (1950).

19. Organic Syntheses, 18, 70 (John Wiley and Sons, New York, 1958)

20. Haworth, Heilbron and Hey, J. Chem. Soc, 1940. 349.

21. J. Gen. Chem., 10, 1101 (1940) (USSR).

22. Koenigs and Greiner, Ber., 643, 1049 (1931).

23. McElvain and Goese, J. Am. Chem. Soc, 63, 2283 (l94l).

24. JSilman and Spatz, ibid., 62, 446 (19^0).

25. Overhoff and Proost, Rec. trav. chim., 57, 199 (1938).

26. von Braun and Pinkemelle, Ber., 64B, 187i(l93l).

Reported by F. J. Wolf, January 21, 1942

SEMINAR REPORTS

II Semester 1941-42

Pa pre
Epinephrine and Related Compounds: The Influence of Structure on 1
Physiological Activity

K. E. Hamlin, Jr.

Attack at the Alpha Carbon Atom of Alpha, Beta-Unsaturated Carbonyl 9
Compounds

J. E. Mahan

Hydrogen Bonding by C-H 14

E. Ginsberg

Acidic and Basic Constituents of Petroleum 16

Billie Shive

Cleavage of the Alkyl -Oxygen Bond in the Hydrolysis of Esters 22

W. E. Blackburn

Some Reactions of Hydrocarbons in an Electrical Discharge 28

R. G. Chase

^ar Gases 34

C» F. Jelinek

Developments in Quantitative Organic Analysis 39

G. E« Inskeep

Recent ^thods for Making Acids 44

G. W. Cannon

The Prins Reaction 49

R. E. Foster

Chlorosulfonic Acid as an Organic Reagent 52

S. S. Drake

Some Derivatives of the Polyhydroxybenzenes 58

J. D. Garber

Relation between uptical Rotatory Power and Constitution of the Sterols 63

• H. Kaplan

Reduction of Carbonyl Groups to Methylene Groups 69

B. C. McKusick

The Replacement of Alkyl Groups during Nitration 72

s. I. Meltzer

Page
Ketene Acetals 77

G. L. Schertz

Biphenylene 81

F. W. Spangler

The Constitution of Usnic Acid 85

C. W. Theobald

Alkyl Carbonates in Synthetic Chemistry 92

R. M, Roberts

Glycerol Derivatives 98

J. F. Shekleton

Determination of Branch Chain ^thyl, Citric Acid, and Organic "alogen 102

C. W. Smith

Recent Developments in the Identification of Organic Compounds 106

Q. F. S0per

Sulfadiazines 111

M. Chiddix

The Effect of Catalysts on the Grignard Reaction 116

A. B. Spradii.ng

The Chemistry of Explosives 121

R. S. voris and P. F. Warfield

Recent Developments in the Study of Vitamin E 133

F. W. wyman

The Acylation and Alkylation of the Sodium Enolates of Aliphatic Esters 138

John Whitson

EPINEPHRINE AND RELATED COMPOUNDS
The Influence of Structure on Physiological Activity

Alt h o u gh r ap i d p r o g . ? ess ha s "been

m o rl

e, the relationship

ix'ucr.ure and
it ill not

hiysiolo < leal activity of
under stood. However,

between the chemical
medi c ina 1 pro du 2 1 s i
evidences that there is some such relationship are apparent in
several well-known chemical series, such as the alcohols, amines,
barbiturates, phenols, and "sulfas." The individual importance
and purpose of each intimate group or radical within each series
has yet to be determined- That this relationship is of value to
discover, is essential 00 the chemist synthesizing products which
combat disease. B;y such information, scientific approach can be
made towards specific medicinal s where formerly success was often
only by chance.

The earliest and best known study in attempting to correlate
chemical constitution and physiological behavior was made with
"pressors" or compounds which produce a rise in blood pressure.
Interest in these compounds was first arcu sed when Oliver and
Schaefer in 1894 found that extracts of the suprarenal glands
produced a rise in blood pressure when injected into the blood
vessels of animals. Immediately an intensive study of this gland
in an effort to isolate its active principle was made by chemists,
physiologists, and pharmacologists, credit finally going to Abel
and his coworkers, who first isolated the hormone (as the poly-
benzoyl derivative) in 1897. Once isolated, the formula for
epinephrine was quickly established and this formulation confirmed
by synthesis of the, hormone
is as follows:

carried out by Stolz in 1904, and

1
j/^N.0H ClCHaCOCl

P0C1.

/ns,oh

Q.h\ NH:

/<\Ch

Al . Hg

C-CH3C1
0

y — C R 2 I'm liC II 3

CHOHCHs-
NHCH3

Epinephrine

Since the levo compound is the only one used in medicine,
the racemic mixture obtained is resolved through d-tartaric acid.

Biologically, epinephrine is an extremely active compound.
It has been found to exert an effect on the isolated frog heart
in dilution as low as one part in five billion. Its value as a
drug is based on the effects that it produces. These are
chiefly, (l) its action on smooth muscles to relieve asthma, hay
fever and severe colds; (2) its hyperglycemic effect to convert
muscle and liver glycogen into glucose; and (3) its pressor ac-
tion to cause a high rise in blood pressure or locally to arrest
hemorrhage.

While the interest in epinephrine was increasing, a substance
isolated from putrified meat, acting as a pressor, was identified
by Barger and Dale to be tyr amine . Their investigations on a
group of compounds of putrefactive origin were reported in a
classic which first shows this correlation of physiological
character and structural similarity. Accordingly, when Chen in
1923 brought to light the medicinal virtues of ephedrine, a drug
used by the Chinese for over five thousand years, workers in this
field intensified their efforts to coordinate these physiological-
ly related chemical compounds.

A comparison of the structure of the better known naturally
occurring pressors (which follows) reveals that the p-phenyl-
ethylamine skeleton is corrmon to all.

<^ "^> CHaCHaNHjs HOK^"^ CH3CKaNHa

p-Phenyle thylamine Tyr amine

/> CHOHCK3NHCE3 <f ^> CHOHCHCH3

NKCH3

Epinephrine Ephedrine

As i s noted, however, there are significant structural differences
between each compound. How, then, do these differences affect
their relative physiological behaviors? Such modifications of
structure will next be considered.

It seems appropriate at this point to discuss very briefly
the usual method by which such compounds are compared. Since
pressor action is common to all of these substances, assay is made
by administration to a test animal and recording the resulting rise
or fall in blood pressure. As epinephrine is the animal hormone
and has been most widely studied, it serves as the comparative
standard during the assay. It must be realized that in this, as
in other biological assays, variations occur as a result of many
factors. Hence, in accumulating evidence from many sources, con-
tradictions and inconsistencies are unavoidable. Thus, for
comparison, the different results of a single experimenter are the
most reliable. Also, in interpreting results, the relative
activities ana toxicities must be compared before assigning a
chemical compound a specific therapeutic value.

Considering now variation from the p-phenyle thylamine nucleus,
Barger and Dale investigated a series of compounds in which the
relative positions of the phenyl and amino groups were varied.
Aniline was without effect, benEylamine was slightly active, a-
phenyle thylamine more active, |5-phenyle thylamine had maximum
activity, while Y~phenylpropyIamine was again much less active.
They concluded, therefore, that, in such a series, optimum pressor

activity was shown by compounds having a phenyl group removed
from an amino group "by two carbon atoms. Mere recently, these
observations nave been substantially confirmed by other workers,
thus establishing the ^-phenyl e thy lamine skeleton as essential
for optimum activity.

Naturally occurring compounds of the type under consideration
include tyr amine, epinephrine, and herdenine Qhordenine or
anhaline is an alkaloid isolated from germinating barley having
the formula:

HO<s /> CH3CH3N(CH3 ) 2 ]

primary, secondary, and tertiary amines, respectively
then consider the variation in the amino group.

Let us

Barger and Dale in their early reports stated that for the
non-phenolic compounds, me thylati on of the amino group makes no
appreciable difference in pressor potency. However, a lengthy
survey of reports uy more recent workers indicates that for both
phenolic and non-phenolic substances of this type (with one
exception), the secondary amines are much less active and more
toxic than the corresponding primary amines. Introduction of the
higher alkyl groups in place of methyl is a step in the wrong
direction. Toxicity becomes higher and depressor action results.
Two methyl groups also work to a disadvantage, as in the case of
hordenine which is distinctly of less value than tyr amine. The
effect of alkylating the amino group may best be summarized by
Table I.

TA3LE I
Effect of Substitution in Amino Group

Compound

Pressor Activity Relative Toxicity
Epinephrine = 1

0CHSCH3NH2 (p-phenylethylamine ) 1/350
0CH3CH3NHCH3 1/350

Toxicity doubles on
substitution

0CHCHCH3KH3
0CHOHCH3NHCH3

1/350
1/700

Toxicity about same

0CHOHCHCH3 (Propadrine)

•"'2

./80

§0% more toxic than
primary amine

Table I (Conf )

£ (conf )

0CHOHCKCH3

N xl L* il 3
0OHOHCHCH3

N(CK3)3

0CHCHCHCH3
1

NHC3H5

0CKOHCKCH3
NHC3H7

1/95

1/600

1/150

Depressor

Toxicity progress-
ively increases
with alkylation

D
HO

/-CH2CH2MHS
(Tyr amine)

1/150

H0<^~ ~ />CH3CH3NHCH3
rlO^' ~^>CH3CHSN(CH3)

1/150

1/700

Toxicity increases
with alkylation

H0<. ^> CHsCHgNHCsHs

1/200

.t of the investigations of
it c^n be stated that the

A second principal difference in chemical structure in the
naturally occurring pressors, is the presence of one or more
phenolic hydroxy 1 groups. As a resu!
Barger and Dale, Alles, and Taint er,
intensity of action of phenolic pressor substances is much greater
than that of the corresponding non-phenolic compounds. Although
the ortho-hydroxyl has no apparent pressor influence, it does
serve to increase the toxicity. The met a isomer increases the
pressor activity as much as five times but also increases the tox-
icity. Finally, the, para hydroxyl confers a pressor activity be-
tween the ortho and me t a and, most important, it lowers the
toxicity. Thus, it is not too surprising to find that the 3,4-
dihydroxy compound indicates a maximum activity with a rather low
toxicity. Further phenolic substitution only serves to decrease
activity. With phenolic substitution, however, the action is
more transitory, the duration of response being lowered. Table II
summarizes such comparable compounds.

- 5

Compound

table ii

Effect of Phenolic Hydroxyls

Pressor Activity
Epinephrine = 1

fCH8CHaNHa (I)
<^~^X CE8GHaNH8

~7"

1/550

1/350

Relative Toxicity

More toxic than

H0<^ ' ^- CE8CHaNH8
(Tyr. amine )

« ..!,— .-J.

^>— CH3CH8rJH8

1/150

1/70

Less toxic than I

More toxic than I

L3^'-2-i'±2

H0<C ^>-0H»GHaNHs

1/3!

Toxicity only
slightly more than

H0<^ \^>-CH0H0H3^HCHa (II)

' — ** ( 3y n ep hr i n e ,

1/55

2?~ CEOKCHaNHOHa

( N e o t yn e p hr i n e )

1/1

More toxic than II

H0<x x>-GHOHCHsNH0Ha

T~ (Epinephrine)

More toxic than II

0CHOHCHCH3 (III)

(it oo9 drine )

HG^ >*> - CEOKCHC!

1/80

1/50

1/5 £?.s t jyic as III.

- 6 -
Table II (conf)

C (conf )

<^ ^>- CHOHCHGH3

/ NHa

1/85

3 times as toxic as
III

<x x^~ CHOHCHCH3

NH3
(Cobefrine)

1/4

l/lOO as toxic as
epinephrine

In the use of epinephrine as a therapeutic agent, its chief
handicap is its complete inactivity when administered by mouth.
Since ephedrine, a propane derivative, is

orally active, is the length of the side chain responsible for
this physiological difference? As studied by the Council on
Pharmacy and Chemistry of the American Medical Association,
phenylethanolamine is not active after oral administration,
whereas phenylpropanolamine is. Chen, et al, in their investiga-
tions on the ephedrine s attribute the oral efficacy of ephedrine
to the presence of the third carbon of the side chain. Hartung and
Munch from their results with the phenylpropylamines found that
phenyl-l-amino-2-propane was orally active. Thus, such compounds
as ephedrine, propadrine and benzedrine are all active when ad-
ministered by mouth. However, further extension of the side chain
gives results of a negative character. Thus, phenylbutanolamine
is nearly inactive while phenylpentanolamine is a depressor with a
marked toxicity.

Finally, in a discussion of these pressor substances, it i s
noted that for some, a side chain hydroxyl group is in evidence.
The effect of this group is not too well defined but all evidence
points to an augmentation of the pressor activity and a lowered
toxicity. An examination of Table III illustrates this rather
clearly.

TABLE III
Effect of -alcoholic Hydroxyl

Compound

Pressor activity

Epinephrine = 1 Relative Toxicity

0CHaCH3NHa
0CHOHCHaNKa

1/350
1/350

1/2 as toxic as
simple amine

- 7 -
Table III (Conf )

0CH2CHCH3 (Benzedrine)

NK:

0CKOHCKCK3 (Propadrine)

1/350
1/80

l/o as toxic as
simple amine

<C~>c-

CHaNHCHj

3uil3

. IH

CHOHCHpNHCH.

1/150

1/35

Less toxic than
simple amine

1/35

HO <. /> CHOHCHgNHg

HO vArterenol)

dl-art. = 3
dl-epi. 2

dl-art. — 1

1-epi. 1.

Less toxic than
s:

.e amine
l/3 toxicity of
epinephrine

The above discussions and the resulting conclusions may be
summarized as follows:

1. The optimum pressor activity (production of a rise in blood
pressure) is found in those compounds in which the aromatic
nucleus and the amino group are attached to neighboring or
adjacent carbon atoms, thus Ar-C-C-N.

2. The primary amines are more active and less toxic, generally,
than corresponding methylated, secondary amines.

3. Substitution of hydroxyl groups in the 3, 4-positions of the
phenyl nucleus confers optimum pressor activity.

4. Compounds with three carbon atoms in the side chain are
much more active on the circulation after oral administration
than are the homologs with only two carbon atoms in the side
chain.

5. The alcoholic .hydroxy 1 group increases the activity and
decreases the toxicity or both.

In applying this information, the following two compounds are
ch&sen as representatives of the synthetic field to surplant
perhaps in their individual usefulness, the natural products:

CHOHCHpNH.

HO<T >>• CHOHCHCH3

NH:

Arterenol

Nearly as active as

epinephrine

l/o as toxic as epinephrine

Acts twice as long as

epinephrine

Cobefrine

Orally active

1/100 as toxic as epinephrine

Acts twice as long as

epinephrine

1/4 as active as epinephrine

Bibliography:

Earger and Dale, J. Physiol. 41, 19 (1910).

Chen, K.K., Wu, C. and Henri ks en, E., Journal of Pharmacology and

Experimental Therapeutics, 36. 363 (1929).
Hartung, Iv.K., Chem. Rev., 9, 369 (1931) 251 ref.
Hartung Munch, Miller, Crossley, J. Am. Chem. Soc, 53_, 4149

(1931).
Tainter, Arch. Internat. Fharmac. Ther., 4JL, 365 (1931).
Tainter, ibid., 46, 192 (1933).
Alles, G-ordon a., Journal of Pharmacology and Experimental

Therapeutics, 47, 339 (1933).
Alles, G-ordon a. and Knoefel, Peter K. , U. of Cal. Pub. in

Pharmacol., 1, 101 (1938).
G-unn, J. a., Brit. Med. J., 2, 155, 214 (1939).
Jenkins, G-.L. and Hartung, W.H., The Chemistry of Organic

Merlieinal Products, John S. Swift and Co., 274, (1941).

Reported by K. E. Hamlin, Jr.
February 11, 1942

ATTACK AT THE ALPHh CARBON ATOM OF

alpha, beta-un saturated carbonyl compounds

a, ^-Unsaturated carbonyl compounds are well characterized by
the addition reactions which they undergo with many different
types of reagents. In a large number of cases addition occurs
in such a manner that the elements H and A, of the reagent HA,
finally become attached to carbon atoms 3 and 4, respectively.

4 3 3 1

c=c-c=o

+ HA

-C-CH-C=0

Kohler and his associates have definitely established that
the initial reaction is a 1,4 addition in the case of G-rignard
reagents, and hence, it is generally assumed that most other
HA reagents react by the same mechanism. However, there are a few-
reagents which apparently attack conjugated carbonyl compounds at
the alpha carbon atom, and thus, are in contrast to the general
scheme outlined above.

A. The first of these to be mentioned is the reaction of
aromatic diazo compounds with substances containing a conjugated
carbonyl system. The reaction was discovered in 1899 by Borsche
who found that instead of obtaining an azo dye by the action of
benzene diazonium chloride on p_~nitrosophenol, a carbon-carbon
linkage was established and 2-phenyl-4-nitrosophenol resulted
along with some 2,6-diphenyl-4-nitrosophenol.

0 NgCl

alkaline
solution

/ \

NOH

The nature of this reaction has been further studied and extended
to various benzo- and naphthoquinones by Kvalnes; while Meerwein
and coworkers have established definitely that the attack is at
the a- carbon atom. The reaction is also the subject of a few
patents in connection with the preparation of substituted quinones
to be used as dye intermediates. The work of Meerwein, mentioned
above, was reviewed by Mr. Rabjohn as a seminar topic in the fall
of 1939, so only a few of the more interesting reactions are
listed below:

0 CH-fiH-CHO -(- e~C1.0N2C1

acetone

solvent

0-CH=C-CHO + Ns
0C1

+ KC1 (2)

0-CH=CH-C=N + p_-C10I<2Cl

- 2 -

76^

0-CH=C~G=N + N8 + HC1 (3)

0C1

0-0 „.,w« -. Acidic HC

+ £-C10i\i3Cl ►

Solution

Umbellif erone

Ns + HC1
(4)

Methyl cinnamate and oinnamic acid are also attacked at the
alpha carbon atom, however, in the case of cinnamic acid carbon
dioxide is eliminated and a stilbene results. This reaction has
been used in this Laboratory by Fuson and Cooke to prepare the
methyl and ethyl esters of 4-carboxystilbene.

0-CH=CH-COOH + C1NS

CO OR ^ 0-CH=CH0COaR + HOI +

OO—u dr/o

+ CO,

(5)

It should be pointed out that although aromatic diazo com-
pounds attack the conjugated system at the a-carbon atom, ali-
phatic diazo compounds attack these substances in such a way that
a new carbon link is established at the p-carbon atom. Thus
benzalacetophenone and diazomethane give the two pyrazolines I
and II.

0-CH=CH-C-0
It
0

+ CH2N3 -

■* 0-CH- CH-C-
CH2 N °

-0

0-CH— -C — C-0

! jl 8
CE3 N
\ /

N TI

H ^X

B. In 1927 Dufraisse and Moureu discovered that a, £-
unsaturated ketones could be converted into a-diketones (40-80^
yields) by bromination, treatment with piperidine, and subsequent
hydrolysis. After a careful examination of the reaction the
authors formulated it as follows:

RCE=CH-C0R

-» RCKBrCHBrCOR

-HBr

CBH1XN

RCH=CBr-COR
III

C5HUN

RCH,CBr-COR — 5-

NC5H10
IV

HiiN

R-CH-

■CH-COR + RCH=C-C0R

some
V

NC5H10

mainly
VI

- 3 -

hydrolysis

[R-CH=C-COR]
OH

■* R-CHo-C-C-R

0 0
VII

The French investigators wrote V as the a,a-dipiperidino compound
but the recent work of Cromwell indicates rather definitely
that it is the a, p-isomer. The transformation of III into IV is
of particular interest since it involves an inverse addition to
the conjugated carbonyl system. Compound VI (R=0) is a deep red
crystalline compound and was first prepared by Watson, although
he considered it to be fj-piperidinobenzalacttophenone .

Kohler and Addinall refused to accept either the mechanism
of Dufraisse and Moureu or their structure VI (R=0) for the red
piperidino compound. Instead Kohler and Addinall, who had been

studying the
the reaction

RCK=CBr-COR
III

action of alcoholates on
as follows:

R
RCH=C — C-NC5H10
Br OH
IVa

formulated

C5H1XN

^s-dibro moke tones,

i-» R-CK=C— C-NC5H10

Via

hydrolysis

R
[R-CH=C— C-NC5H10]

OH OH

[RCH-C-COR]
0
H

RCHaC-C-R
0 0

VII..

In order to prove which of these tivo mechanisms was correct
a rigorous proof of structure of the compound represented by VI
or Via was necessary and in so doing the questionable inverse
addition would be clarified. This work was done by Kohler and
Bruce and after an exhaustive study they reluctantly concluded
that the structure represented by VI as assigned by Dufraisse
and Moureu was correct. In brief, their most conclusive proof of
structure was as follows:

0-CH=C-CO0

5MgBr

NC5H10

$,-CH-C=C-CMgX
NC5H10

NK4C1

03-CH-CH

0-0

NCSH10 0

■•X

NC5H10
VIII

0Mg3r

.0:

0-CH=C C-0 2 stePs^ 0-CH2-C-Cx

Villa

- 4 - 12

To decide which of the two formulas, VI II or Villa, was correct
for the carbinol which they obtained was easy; oxidation with
chromic acid gave two moles of benzophenone instead of one as
would be expected from Villa.

More recently Cromwell has further investigated the reactions
of a-bromo-a, ^-unsaturated ketones with amines. He has demon-
strated that morpholine reacts, in cold ether, or petroleum ether,
with a-bromobenzalacetone , IX, and o-bromobenzalacetophenone , X,
to give the a-bromo-a-morpholino compounds XI and XII by inverse
addition.

Br
0-CH=CBrCOR ^4K80 > 0-CHg-C-COR . Na0ii9 0-CH=C-COR

I Ale. (
NC4H80 NC4K80

(IX; R = CK3) (XI; R = CH3 ) (XIII; R = CH3 )

(X; R = 0) (XII; R = 0) (XIV; R = 0)

When XI and XII were treated with sodium acetate in alcohol solu-
tion the cc-morpholino-a, ^-unsaturated ketones XIII and XIV re-
sulted. The latter two compounds, however, could not be induced
to add another molecule of morpholine, nor could this base be
added tot £-morpholinobenz,-.lacetophenone which also contains the
conjugated carbonyl system. Pyrrolidine has been found to add
to a-bromobenzalacetophenone X in the same manner as morpholine,
but the weaker* base, 1,2,5,4-tetrahydroquinoline will not react.

Now when benzalacetone and benzalacetophenone react with
morpholine only saturated p-morpholino ketones result, thus
indicating that amines attack the a-carbon atom of a, ^-unsaturated
ketones only when there is a halogen atom substituted on the
a-carbon.

In studying the reaction of msthoxylamine witn a, ^-unsaturated
ketones, Blatt has shown that this reagent adds smoothly and
reversibly in the absence of a catalyst to produce saturated fj-
methoxy amino ketones.

R-CH=CH-CCR + CH30NHa -» R-CH-CH2-C-R thl00ri3 , R-CE=C-COR

I jl -CH3OH I

NH 0 NH3

0
CH3

^hydrpj^sis^ RCH2-Crg-R

0 0

However, when this reaction is carried out in the presence of a
base or when the £-methoxyamino ketone is treated with sodium
ethylate, methyl alcohol is removed, and an a-amino-a, j3~
unsaturated ketone is formed, which can be further hydrolyzed to
an a-diketone. This shift of an amino group from the p to the a
carbon is easily accomplished in yields which in meny cases
approximate 90^ and thus constitutes a useful method for trans-
forming a, ^-unsaturated ketones into a-dike tones.

- 5 -

C. It is reported in the paper of Meerwein, et al., that
diphenyl ioclinium iodide reacts with a, ^-unsaturated carbonyl
compounds to give substitution at the a-carbon atom just as in
the case of diazo compounds. Thus

0-CK-CH-COOB + &0J3I -> 0GH=CH0 +01 + HI

There was no experimental detail, or reference to previous
work given, merely the statement that the reaction was under in-
vestigation. To date nothing further on this reaction has "been
published.

Bibliography :

W. Bore che, Eer., 32, 2935 (1899); Ann., 312, 211 (1900).

U. S. Patent 1,735,432; C.A., 24, 732 (l9307. German Patent

508,395; C.A., 25, 712 (1931)". English Patent 480,617;

C.a. 32, 6262 (193S).

D. E. Kvalnes, J. Am. Chem. Soc, 56, 2478 (1934).

C. Dufraisse and K. Moureu, Bull. soc. chim. , (4) 41, 457, 850,
1370 (1927).

E. P. Kohler and C.R.Addinall, J. Am. Chem. Soc, 52, 3728 (1930)
E. R. Watson, J. Chem. Soc, 85, 1322 (1904).

C. Dufraisse and R. Netter, Bull. soc. chim., 51, 550 (1932).
S. P. Kohler and "i.F. Bruce, J. Am. Chem. Soc, 53, 1994 (1931).
N. H. Cromwell, ibid., 62, 1672, 2897, 3470 (1940}"; 63, 837,

2984 (1941).
V. E. Stewart and C. 3. Pollard, ibid., 59, 2702 (1937).
A. H. Blatt, ibid., 61, 3494 (1939).
H. Meerwein, E. Buchner and K. van Emster, J. prakt. Chem. 152.

237 (1939).
H. G-. Cooke, Ph.D. Thesis, University of Illinois, 1940.

Reported by John E. Mahan
February 11, 1942

HYDROGEN BONDING BY C-H

Marvel, Copley and Zellhoefer

The concept of association through hydrogen bonding proposed
by Latimer and Rodebush in 1920 to explain the abnormal properties
of certain liquids has been utilized with considerable success
in understanding the behavior of liquids and solutions. In
particular it has proved useful in explaining and predicting ab-
normally high solubility of certain organic substances in various
types of solvents, high heats of mixing, and other abnormal

olution behavior.

The types of hydrogen bonds that have been found are
0H<— 0 or N, XHfr-0 or N, NHe- 0 or N, SH -> 0 or N, and CH<-0 or N.
The present work deals with the latter type.

There are a number of methods of determining the presence
and strength of hydrogen bonds. One method involves shift in the
C-H absorption frequency in the infra-red regions. The methods
used at this school include measurement of the heat of mixing
of liquid mixtures, vapor pressure - composition studies of a
volatile solute in a relatively non-volatile solvent, and solubili-
ty measurements of polymers in various solvents. All of these
methods give relative values of the strength of bond. Qualitatively
they are in excellent agreement. Absolute values for the
strength of a CH(-0 bond, say for chloroform ether, are about
5,000 to 10,000 calories per mole of complex formed.

The CH hydrogen atom is capable of acting as an acceptor
only when it is activated by some negative group such as halogen,
nitro, nitrile, acetylenic linkage, and to a very slight extent
the phenyl group. It has been found that as acceptors,
CHC13 > CHBr3 > CHI3 and CH3C12 > CE3Br3 > CH3I3.
This is quite in keeping with the order of electronegativity of
the halogens: CI > Br>I. It is the presence of the strongly
electron-attracting halogen atoms on the carbon atom which loosen
the hydrogen atom and make it available for coordination to the
donor atom.

One very unusual case of an active CH hydrogen atom is in
benzotrichloride. This compound is a vinylog of chloroform and
most probably owes its bonding activity to the vinylogous action
of the halogen atoms on the para (or ortho) hydrogen atom.

In the case of donor solvent s the order of donor-ability
appears to be, roughly and with a degree of overlapping: alkyl
phosphates > 3C aliphatic amines ^ N,N-dialkyl amides > 1° ali-
phatic amines > ethers, ketones and esters > aromatic amines>
nitriles > alkyl sulfates and sulfonates > nitrates and nitro com-
pounds.

Solvents which form strong intramolecular hydrogen bonds
such as alcohols, unsubstituted amides, and glycols exhibit little
tendency to bond intermolecularly with other acceptors and hence
are poor solvents for such compounds as haloforms. However, they

- 2 - JLO

are good solvents for strong donors such as amines. Dinitriles, -~ _
aliphatic nitro compounds, aromatic amines, and other compounds
which form weak intramolecular hydrogen bonds show less tendency
to form int ermolecular bonds with either donor or acceptor mole-
cules than do unassociated solvents. Similarly, chelation tends
to decrease the solubility of diketones and keto esters in
acceptor solvents.

A steric effect has been noted in the heat of mixing of
chloroform and polyethylene glycol ethers. The curves indicate
that every other oxygen atom bonds to a chloroform. This is
similar to the result obtained with donor solvents and sym-
tetrachloroe thane. In this case with two adjacent active hydrogen
atoms only one appears to bond.

Hydrogen bonding has been of some help in obtaining solvents
for polymeric materials. In general, if the polymer contains a
donor group an acceptor compound will be a good solvent.
However, hydrogen bonding is only a partial explanation for
solubility of polymers.

Hydrogen bonding is a very useful aid to the organic
chemist, but it must be remembered that it is only a part explana-
tion of solubility and not an infallible rule.

Bibliography:

Latimer and Rodebush, J. Am. Chem . Soc, 42, 1419 (1920).

McLeod and Wilson, Trans. Faraday Soc, 51, 596 (1935).

Gordy, J. Am. Chem. Soc, 60, 605 (1938).

Buswell, Rodebush and Roy, ibid, , 60, 2448 (1938).

Zellhoefer, Copley and Marvel, ibid. , 60, 1357 (1938).

Zellhoefer and Copley, ibid., 60, 1343"Tl938).

Copley, Zellhoefer and Marvel, ibid. , 60 2666, 2714 (1938).

Copley and Hoi ley, ibid. , 61, 1599 (1939).

Copley, Marvel and Ginsberg, ibid. , 61, 3161 (1939).

Copley, Zellhoefer and Marvel, ibid., 62, 227 (1940).

Marvel, Dietz and Copley, ibid. , 62, 2273 (1940).

Marvel, Copley and Ginsberg, ibid. , 62, 3109, 3263 (1940).

Copley, Ginsberg, Zellhoefer and Marvel, ibid., 63, 254 (1941).

Marvel, Harkema and Copley, ibid., 63, 1609 (1941

Marvel and Harkema, ibid. , 65, 222lTl941 ) .

Reported by E. Ginsberg
February 18, 1942

ACIDIC AND BASIC CONSTITUENTS OF PETROLEUM

At the University of Texas, under the direction of
Dr* H. L, Lochte and the late Dr. J. R. Bailey, sixty-five indi-
vidual acidic and "basic constituents of petroleum have been
isolated and identified. Several more have been isolated but
not yet identified. Only sixteen of these compounds had been
previously reported, and thirteen of these were phenols and
aliphatic acids. The compounds which have been isolated are
listed in Table I.

TABLE I

Bases

Acids

Quinoline
Quinaldine

8-Dimethylquinoline

3-Dimethylquinoline

•Methyl-8-ethylquinoline

3 , 8-Trimethylquinoline

3-Dimethyl~8-ethylquinoline

3-Dimethyl-8-n-propylquinoline

4-Dimethylquinoline

4, 8-Trimethylctuinoline

4-Dimethyl-8-ethyiauinoline

4-Dimethyl-8-n-propyiquinoline

4-Dimethyl-8-s.-butylquinoline

3,4,8-Tetramethylquinoline

3,4-Trimethyl-8~ethylquinoline

3,4-Trimethyl~8~n~propylquinoline

3,4-Trimethyl-8~l-propylquinoline

3,8-Trimethyl~4-ethylquinoline

3-Dimethyl-4,8-diethylquinoline

3-Dimethyl-4-ethyl-8-n-propylouinoline

Isoquinoline

2-Picoline

4-Picoline

2, 6-Dime thy lpyr idine

2, 5-Dimethylpyr idine

2,4-Dimethylpyridine

3, 5-Dimethylpyridine

2,4 , 5-Trimethylpyridine

2, 4 , 6-Trimethylpyridine

2-s_e_c.-3utyl-4, 5-dimethylpyridine

2, 4-Dimethyl-6- " trans T'-2, 2, 6-
t rime thy lcyclahexvlpyr idine

2,3-Dimethylbenzo (h)ouinoline

2, 4~Dimcthylbcnzo (h)quinoline

Formic

Acetic

Propionic

n- Butyric

iso-Butyric

Valeric

.1 so- Valeric

n-Hexanoic

2-Methylpentanoic

3- Me thylp en tano i c

n-Heptanoic

SrMe t ay liitex&nol c

3-Methylhexanoic

4-Me t hy lhe xano i c

5-Methylhexanoic

n-Octanoic

n-Nonanoic

Dimethylmaleic anhydride

Phenol

o_~Cresol

m-Cresol

p_-Cresol

2,4-Xylenol

2,5-Xylenol

3,5-Xylenol

Cyclopentanecarboxylic

Cyclopentaneacetic

3-Methylcyclopentane-

acetic
2,3-Dimethylcyclopen-

taneacetic
solid 4-methylcyclo-

hexanecarboxylic
"trans "-2, 2, 6-Trimethyl-

cyclohexanecarboxylic
cis-2, 2 , 6-Trime thyl-

cyclohexanecarboxylic

- 2 " 17

One of the interesting problems in connection with this
research was the assignment of structure to a C16HS5N base,
2,4-dimethyl-6~trj^~2,2,6-trimethylcyclohexylpyridine, and the
structural correlation of this base with a C10H1802 acid, trans-
2,2,6-trimethylcycl@hexanecarboxylic acid, also isolated from
petroleum.

The positive contributions toward determining the structure
are diagramed on the Flow Sheet and may be listed as follows:

1. Isolation of pyridine-2,4, 6-tricarboxylic acid from the
oxidation products of the base with dilute nitric acid proved
alkylation at positions 2,4, and 6.

2. Formation of a phthalone established the presence of a
methyl group at position 2 or 6.

3. Condensation of the base with formaldehyde and nitric
acid oxidation of the product yielded a dicarboxylic acid,
G14H19N(COOH)3, which was decarboxylated first to C§ iHi9N(COOH)3,
which was decarboxylated first to Cl4H20NCOOH and then to C14H21N,
indicating the presence of two reactive methyl groups.

4. Formation of a dibenzal derivative also indicated the
presence of two reactive methyl groups.

5. High pressure hydrogenation in the presence of Raney
nickel as a catalyst resulted at 250° in the consumption of six
atoms of hydrogen per molecule of the base. Dehydrogenation
yielded the original base. The base with one mole of methyl iodide
formed a quaternary ammonium salt which was not converted into a
methylated free base by sodium hydroxide. The non-reactivity of
the base with ammonium iodide and hydriodic acid in an atmosphere
of carbon dioxide showed the absence of an N-:.lkyl group. All of
these facts indicated that the base was a substituted pyridine.

One of the' confusing facts about the base was its stability
towards oxidizing agents. Alkaline permanganate as well as chromic
acid has no effect at room temperature. Acid permanganate
attacks the base but no product except carbon dioxide could be
isolated.

Data to this point indicated that the compound contained
a pyridine nucleus with two methyl groups and a C9H17-alicyclic
radical as substituents in positions 2,4, and 6, the exact location
of the individual groups being indefinite.

6. Ring cleavage and removal of the nitrogen atom was
achieved by the method of von Braun, by which benzoylpiperidine
may be converted to pentamethylenedibromide. The dibromide ob-
tained was converted to a diene by alcoholic KOH and ozonized to
obtain a CloH1802 acid, confirming the presence of n C9H17-
substituent on the pyridine nucleus. Direct ozonization of the
base followed by hydrolysis in alkaline peroxide gave the amide
of the acid, establishing the C0H17-group as being .in the °->
position

- 3 - AS

7. The C10H1803 acid obtained from degradation of the
base was found identical with one obtained from petroleum.

8. Conversion to the C9H17NK3 amine was accomplished by
the action of hypobromite on the amide or hydrazoic acid upon

the acid. Treatment with nitrous acid gavn an unsaturated hydro-
carbon, C9H1S, which gave 2-methylcyclopentyl methyl ketone,
leaving only five possible structures (listed in the flow sheet)
for the acid. Since the acid was not esterif iable by methyl
alcohol-hydrogen chloride, the ester prepared from the acid
chloride was not saponified by alcoholic KOH, the e ster was not
hydrogenated under 4QC0 pounds pressure at 250° in the presence
of copper- chromium oxide, and the amide could not be prepared by
ammonolysis of the ester, it appeared at first to be a tertiary
acid. Whitmore's method of distinguishing between secondary
and tertiary acids indicated that this acid was tertiary; so the
structurally possible tertiary acids were synthesized.

After elimination of these tertiary acids, von Braun's
imide chloride method showed the acid to be secondary, whereupon
the two isomers of 2, 2,6-trimethylcyclohexar.ecarboxylic acid
were synthesized. The acid from petroleum and from the degrada-
tion of the base was identical with the higher melting form,
tentatively assigned the trans configuration as the higher
melting form.

Bibliography:

Shive, Horeczy, Wash, and Lochte, J. Am. Chem. Soc, 64, 385 (1942).

Shive, Roberts, Kahan and Bailey, ibid, (in press)

Shive, Crouch, and Lochte, ibid. , 63, 2979 (1941).

Wash, Shive, and Lochte, ibid. , 63, 2975 (1941).

Hancock and Lochte, ibid. , 61, 2448 (1939).

Schutze, Shive and Lochte, Ind. Eng. Chem., Anal. Ed., 12, 262

(1940).
Lake and Bailey, J. Am. Chem. Soc, 55, 4143 (1933).
Armendt and Bailey, ibid,, 55, 4145 TlS33).
Thompson and Bailey, ibid, ,55, 1C02 (1931).
Perrin and Bailey, ibid,, 55.;.' 4136 (1933'
Bratton and Bailey, ibid., 59, 175 (1937
Lackey and Bailey, ibid. . 56, 2741 (1934
Axe and Bailey, ibid., 60, 3028 (1938).
Axe, ibid-, 61, 1017 (1939).
Axe and Bailey, ibid. , 61, 2609 (1939).
G-lenn and Bailey, ibid., 61, 2612 (1939).
Glenn and Eailey, ibid., 65, 637 (1941 J .
Glenn and Bailey, ibid., 63, 639 (1941).
Schenck and Bailey, ibid., 61, 2613 (1939).
Schenck and Bailey, ibid. .. 52, 1967 (1940;.
Schenck and Bailey, ibid.'. 63, 1364 (1941).
Schenck and Bailey ibid., 63, 1365 (1941).

Reported by Billie Shive
February 18, 1942

FLOW SHEET
Degradation
COON

CH3

tSSK^00

+ other oroducts

%/-C00H

N^-G^3

phthalic

anhydride

Phthalone

JH^O

CilgCrigOH

KNO*

C9HX7-^ y-CffaCH8OK

CBH17^N^COOH CbHxtV*^

o
ixi
o

in
W

CD
"O

-^ C9H17\Ny-CH=CK-C6H5

soaa
lime

C9K17-\N^

^16^35^ + 3H3 — » C16H31N
C16H30NCBr3(C6H5) - distillation

CrHcCCCI prv.

-S-S ; C16H30N-C0-C6H5 -i^^^

in vacuum

* C16H30Br;

Me OH

HoO

KOH

C9H17C00H

3^3

,Qa

1#Hj 7-A^ >/*- CNS

* ozonide j?a^H ■> C9H17-C0NH3 + 2CH3C00H +

H3°3

2C0o

' C9H17-CONH3^Jto£.

C9H17-C00H

G9Hx 7NH;

„*^\x6*H&** C9H17i*(CH3)a

CH-

■CH:

CH2 Cri-CH3

CH-C:

CH-

On ■

■CI

ri-

Cri2

H2C CH

CH3HC C(CH3)2

vc6

T COOK

CHS

H2C/\CH(CH3)

I !
H2C G (CH3 ) 2

CO OH

CH2

CHS — CH2

CH CH ( CH3 )

H3C-OH-CH3 COOH
III

CH-CH3
CH-C^CH3

CHg CK2

i !

CH, CH(CH3)

^CH
I
CH3-C-COOHI

CH3
IV

CH2 CH2

1 I

CH2 CH(CH3)

CH
»

CH3-CK-CK2~COOH
V

SYNTHESIS OF I

(CH3)2C=CHCHaCH3~C=CH-C-H ife^L (QH3 )3C=CHCH3CH3-C~CH~COOH f orm^c >

C
HC NCH2

CHa-C C(CHs)a

XCH

COOH

CH3

HoC -CHp

I I

(CH3)HC /C(CH3)2

CH
COOH

ra.p. 74-75°

CH3
Hac, ac20

acid

trace HC1

H2C CH2

(CH.) HC C(CH3)2

COOH

m.p. 82-83'

PROPOSED SYNTHESIS OF BASE

CH3

(CH3)HC ,C(CH3)a

Ui-3

•>- CH

Al.Hg

Uxlg

H2C v CH3

I i

(CH3)HC C(CH3)3

XC^

CH3

«Li £»•— -CH3

CLEAVAGE OF THE ALKLY-OXYGEN BOND IN THE HYDROLYSIS

OF ESTERS

Cohen and Schneider (Harvard )

Although it appears that hydrolysis of sulfonic and other
strong acid esters occurs in the following manner,

-SOa-0-3|+HJ5ff-* t«6020H+R0H

it is generally accepted that hydrolysis of carboxylic esters and
the formation of these esters are reactions which do not involve
rupture of the alkyl carbon-jjiygen bond hut involve an acyl
carbon-oxygen mechanism,

R-C^O-R'+HOH-* RCOOH+R'OH

although the same products could be produced by the alkyl carbon-
oxygen rupture.

The evidence for the acyl carbon-oxygen mechanism can be
summarized as follows:

1. Reid's work on the esterif ication of thio-acids and mercap-
tsns, and the hydrolysis of thio-esters.

(a) Esterif ication of thio-acids produced oxygen esters
and hydrogen sulfide, rather than thio-esters and
water,

*P-

RC-S-H+HpR' -^RCOOR»+H3S

(b) Carboxylic esterif ication of mercaptans produced thio-
esters and water, rather than oxygen esters and hydro-
gen sulfide.

RC-.0-H+HS-R1 -*RC-S-R»+H20

RC-^-R'+HDH -^RCOOH+R'SH

2. Retention.of configuration when carboxylic esters of the

/O H

type R-C-0-£*_rt sre hydro lyzed. It is evident tha t if at
sty time during the hydrolysis, the asymmetric carbon become
free, complete retention of the original configuration is
not possible. Thus, retention of configuration is an indica-
tion that the alkyl-oxygen bond is not broken during the
hydrolysis. There are many cases of such hydrolyses in the
literature, and there is no simple instance of appreciable
inversion of configuration of the alcohol.

-2-

3. Demonstration by Ingold 'and co-workers that single alcoholic
oroducts are obtained when esters of the 0

R~C?-0-Ctt2-C=CR2
R
type are hydrolyzed. If the resonating ion becomes free
at any time during the hydrolysis,
S+ + H

R:C::C:6« >R:C:C::C

R R H R R H
a mixture of alcohols would be obtained.

4. Demonstration by Quayle and Norton that esterif ica tion of
various acids with neopentyl alcohol produced no unsaturated
products, and that hydrolysis of the resulting esters pro-
duced neopentyl alcohol es the only alcohol. "Whitmore and
his collaborators have demonstrated that a positive neo-
pentyl group invariably rearranges to tertiary amyl, the
rearrangement being accompanied by the formation of un-
saturated derivatives.

9H3 + +

CH3C-CH2 > CH3Q-CH3CH3

CH3 CH3

Apparently the oxygen-neopentyl bond is not broken during
esterif ica tion or hydrolysis, even though the a.cid concerned
is as strong as trichloracetic acid.

5. Direct evidence employing isotopically distinguished dxygen.

(a) Alkaline hydrolysis ©f amyl acetate in an aqueous
medium enriched in 018 resulted in the obtaining of
amyl alcohol with the normal isotope ratio, thus
indicating that the oxygen present in the alcohol was
originally attached to the alkyl group in the ester.

(b) Similar results were obtained by Ingold in the acid-
catalyzed hydrolysis of methyl hydrogen succinate.

(c) Urey and Roberts esterified benzoic acid with 018
enriched methyl alcohol, and found that the heavy
isotope was not present in the water produced in more
than the normal rati©, thus indicating that the oxygen
of the water produced was obtained entirely from the
benzoic acid.

There are a few exceptions to this mechanism in the case
of some compounds, which because of special structural features,
show unusual reactivity. These include ^-lactones, the hypo-
thetical cc-lactones, and esters of a secondary allylic alcohol
( CGHs~CH-CH~gH0H~CH3 ) ,

The general conclusion for the mechanism has been based on
a study of esters of primary and secondary alcohols, and
apparently is perfectly valid for such cases. However, there is
one instance reported in the literature in which a very limited
racemization of a secondary optically active alkyl group
occurred during esterif ication, i.e., the esterification ©f

-3-

p-n~octyl alcohol. An alternative mechanism was indicated.
Since the tertiary alkyl-oxygen bond ie more easily broken than
is the secondary , an investigation of tertiary rlkyl esters was
thought to be worth-while.

Cohen and Schneider have made such investigations. Although
the nature of the hydrolysis products would throw no light on
the point of interest (unless optically active alkyl groups or
iso topically distinguished oxygen were employed), the nature
of the products obtained from a primary rlcoholysis of an ester
would provide pertinent information.

The two possibilities are:

(a ) Ester Interchange

s$ *o

RC ~p-R M-HpR" -* RC-O-R" +R T OH
a nd ,••••••

( b ) Ac I-* *n9 Ether .Formation P ,0

RC ~0-jR \ +t Rn0h -> RC-O-H+R * OR"
The occurrence of ester interchange 'would* 'indicate the validity
of the general conclusion as applied to tertiary esters, while
the formation of acid and ether by alcoholysis would indicate
a different mechanism, i.e. Alkyl carbon-oxygen cleavage.

Alcoholysis of i.-butyl benzoate with methyl alcohol in
initially neutral solution produced a good yield of t.~ butyl
methyl ether and appreciable quantities of benzoic acid. Al-
though considerable methyl benzoate was produced, this could
have been formed by esterif ication of the benzoic acid produced
by the acid-ether mechanism. It was also shown that t_-butyl
alcohol and1 methanol do not form J>-butyl methyl ether in the
presence of benzoic acid. No t-butyl alcohol (which would
result from ester interchange) could be detected. Thus it
Bpp-peers that ester interchange does not play a very important
part in this reaction, end that the alkyl carbon-oxygen cleavage
Is probable. Similar results were obtained with t~butyl acetate.

On the other hand, similar alcoholysis in the presence of
alkf li resulted in the normal ester interchange, and no evidence
was found for the formation of t.-butyl methyl ether, thus
indicating the "normpl" cleavage.

steric hindrance

■

~4_

Result? of the hydrolysis of esters of 2,4,6-trimethyl
benzoic acid were as follows:

(1) Acid hydrolysis of JCu-butyl 2, 4, 6-trirnethylbenzoe te
produced a nearly quantitative yield of the acid.

(2) Similar acic1 hydrolysis of the methyl ester was un-
successful.

(3) Comparable basic hydrolysis of the t-butyl ester was
unsuccessful. These results demonstrate that there
is a deep-seated difference between the alkaline and
the acid hydrolyses of the t_-butyl ester, and that an
equally important difference exists between the acid
hydrolyses of the tertiary ester and of the. primary
ester.

When t-butyl 2,4,6-trimethyl benzoate was subjected to
alcoholysis, results pointing to the same conclusion were
observed, namely

(1) Initially neutral alcoholysis gave products obtainable
from the acid-ether mechanism, although the yields
were not nearly fb great.

(2) Base catalyzed alcoholysis did not occur, indicating
that the ester interchange mechanism is not active
in the case of this st erica lly hindered compound.

The base-catalyzed ester interchange and hydrolysis re-
actions of esters probably proceed by a nucleophilic attack
an alkoxide or hydroxide ion on the carbonyl carbon atom of
ester.

the

:0:
R:C:p:R'+:0:Hz

„

• 0:..
R:C:0:H

5 0:R! ^

:0:
tR:cV6:H+ :0:R1

A Stuart model of ^-butyl 2, 4, 6~trimethylbenzoate shows that the
carbonyl carbon atom is strongly hindered, and the reason for
the failure of attempts at alkaline hydrolysis is readily
apparent. The probable mechanism for the "normal" acid-
catalyzed 'hydro ly si s also involves the same difficulty, and the
resistance of primary esters of this acid is understandable.

A possible course for the alcoholysis of tertirry esters to
form the corresponding carboxylic ^cids and ethers ie as follows,
First, a slow reaction occurs in neutral solution involving
hydrogen bonding between en alcohol molecule and the carbonyl
oxygen atom of the ester, followed by ejection of the £~butyl
group and the formation of the products. This uncatalyzed or
solvolytic mechanism accounts for only a very small portion of
the products. However, the acid produced, catalyzes the faster
and more important reaction.

~5- ,.^v^

:0: +:0:K

♦ • • « »

OeH8.;0:0:G(CHs)3+H;A: ^ — > C6H5:C:0:C(CH3 }3+:A:

0-
••■ ■+■

C6H8:C::0:C(CH3)3 ->
< •

•0:

+
C6H5C:0:K+(CH3)Bt3

+ ## +

(CH3)3C +CH3:b:H -* CK3: 0:C(CH3 )3+H

The model of Jj,- butyl 2, 4, 6~trimethylbenzoa te shows that the
carbonyl oxygen atom is not strongly hindered. Since the above
mechanism involves this atom and not the carbonyl carbon, the
above mechanism seems plausible, ?n0 would account for the
acid hydrolysis of this ester ps well as the alcoholysis pro-
ducts obtained from t^butyl benzoate in initially neutral solu-
tion.

There are relevant kinetic data available for hydrolysis
of esters of the type CK3C00R. As the group R is varied from
CH3 to C3K5, i_-C3H7 and J>-C4Hg, the rate of the base-catalyzed
hydrolysis falls off, while the rate of acid hydrolysis falls
through a slight minimum and rises again, that of the. t_~butyl
ester being about 15$ faster than that of the methyl ester.

These facts are consistent with the following interpreta-
tion: in the base-catalyzed ractions, a single mechanism oper-
ates, the attack of the hydroxy 1 ion on the carbonyl carbon
atom followed by the ejection of the alkoxyl group. The rate
of this reaction diminishes as the alkyl group is changed from
primary to tertiary. In the acid-catalyzed reaction, there is
one mechanism available for esters of primary end secondary
alcohols involving a. rupture of the acyl-oxygen bond, and there
is the alternative mechanism available for the tertiary alcohol
ester, involving rupture of the alkyl-oxygen bond and ejection
of the alkyl group.

Further investigation employing optically active tertiary
esters or isotopically distinguished oxygen would provide
further information on this Question.

BIBLIOGRAPHY

Cohen and Schneider, J. Am. Chem. Soc . 63, 3332 (1941)
Annual Reports on the Progress of Chemistry for 1940, London

38, ?29 (1941)
Reid, Am. Chem. J. 43, 489 (1910*

JS7

-6-

Polanyi pnc1 Szabo, Trans. Faraday Soc . 30, 508 (1934)
Roberts and Urey, J. Am. Chem. Soc. 60, 2391 (1938)
Olson and Miller, lipid. 60, 2687 (1938)
Winstein, ibid. 61, 1635 (1939)

Bean, Kenyon and Phillips, J. Chem. Soc. 303 (1936)
Kenyon, Partridge and Phillips, ibid. 85 (1936)
Burton and Ingold, ibid. 904 (1928)
Hughes, Ingold and Masterman, ibid. 840 (1939)
Ingold and Ingold, ibid . 756 (1932)

Datta, Day and Ingold, ibid . 838 (1939)
Quayle and Norton, J. Am. Chem. Soc. 62, 1170 (1940)
Whitmore, ibid. 61, 1586 (1939)

Hammett, "Physical Organic Chemistry", McGraw-Hill (1940),
pp. 211, 213, 356

Reported by ¥. E. Blackburn
February 25, 1942

SOME REACTIONS OF HYDROCARBONS IN AN
ELECTRICAL DISCHARGE

It is possible to excite molecules, causing them to undergo
subsequent chemical reactions, by exposing them to the action
of heat, light, sound waves, a particles, and electrical dis-
charges. Considering only the last method, we find that there are
two distinct types of electrical discharge, the silent or non-
disruptive and the disruptive discharge. The silent discharge
may take the form of the ozonizer, the semi-corona, the corona,
the glow, or the electrodeiess discharge, while the disruptive
discharge may occur as the arc or the spark. The chemical re-
actions which take place in these two types of discharges are
essentially different. In a given reaction, the important
variables to be considered are: (l) the effective discharge in-
tensity; (S) the effective residence time in the discharge.

Reactions of Hydrocarbons in the Silent Discharge-
Ethyl ene is a fairly active hydrocarbon and requires only
mild excitation to cause it to react. With ethylene and higher
olefins reactions occur under mild conditions and are of a com-
paratively simple nature. In a static system, that is a system
in which the ethylene was allowed to stay in the reaction chamber
during the entire course of the discharge, it was found that
when using a high frequency course discharge, the ethylene had
completely reacted after a period cf about ten hours. There
appeared to be an induction period of about two hours which could
be considerably shortened by the introduction of hydrogen into
the reaction chamber. An analysis of the gaseous products showed
67% hydrogen and 20% saturated hydrocarbons. The liquid fraction
consisted of a dark oil with a molecular weight of the order of
five hundred.

The results of an experiment using a dynamic system are more
informative. The ethylene was passed through the discharge in a
stream and the products were condensed at -?0°C. In the
particular experiment cited a high frequency ozonizer discharge
was used. The following products were obtained:

Uncondensed gases (H2, C3H3, C2H6 ) 13%

Butane 45%

1-Butene 15%

a fraction boiling in the range

34-45 °C 4%

a C6 fraction 15%

Higher hydrocarbons 8%

In the C6 fraction 1-hexene and paraffin hydrocarbons were iso-
lated. In other experiments in which somewhat different experi-
mental conditions were used, it was found that the chief product
was acetylene or butadiene. In all cases the energy consumption
was about twenty kilowatt hours per kilogram of ethylene which
reacted. The products identified are thought to be evidence
for the theory that ethylene undergoes two primary reactions:

- 2 -

[1) Dehydrogenation of ethylene to produce acetylene and hydrogen;
.2) Polymerization to give higher molecular, weight hydrocarbons.
Egloff and others suggest that (l) may be accounted for by the
assumption that the energy absorbed from the electrical discharge
disrupts the C-H bond in ethylene leaving the vinyl radical.

II]

H
H

C=(

The vinyl radical then decomposes to give acetylene and more
hydrogen.

H H

•C=CH + H-

It is quite possible, however, that the C-H bond is not actually
broken but merely activated. Under these conditions it is
possible that two molecules of ethylene might react as follows:

H
H
K

vc=c;

c=c-'

C=C
H * ,*H

Nc=cr"

H' H

rigC— Crl— C113CH3

1-butene

from

The butene is isolated in a great many cases. Furthermore, in the
presence of hydrogen, hydrogenation to butane is possible which
would account for the formation of this substance. It is also
possible for the butene to react with another molecule of
excited ethylene forming a hexene and then a hexane. However,
this theory is not sufficient to account for all the observed
products. In many cases the liquid products absorb oxygen f]
the air, a property common to acetylene polymers but not
characteristic of ethylene polymers. This fact suggests the
possibility that acetylene, after being formed, reacts to give
polymers of the butadiene type.

With higher olefins, reactions similar to those of ethylene
take place. In a static experiment using propene, it was found
that 85-90$ of the olefin reacted to form a liquid possessing an
average molecular weight of 232. Isobutene polymerizes readily
to give, as the largest fraction, a mixture of di- and triiso-
butenes.

Acetylene was found to react readily in an electrical dis-
charge. When the reaction mixture was cooled to -60°C, a 70fo
yield of a colorless liquid was obtained. This liquid had a
molecular weight of a triaer and from the reactions of the liquid
it was concluded that it contained 1, 5-hexadiyne, methyl
pentadiyne and unidentified products.

When the re
temperature, both
of the products s
to the benzene ri
action of a silen
If the reaction p
and cooled, relat
allowed to react
compounds or arom
A spectroscopic e
is in the dischar
carbon atom, and

- 3 -

action products were allowed
liquid and solid products w
eemed to have an unsaturated
ng. Thus it seems probable
t discharge on acetylene is
roducts are quickly removed
ively simple polymers may be
further, long-chain, highly
atiea with unsaturated side
xaminati on of the light prod
ge indicates the presence of
the hydrogen atom.

to stand at room
ere formed. Some

side-chain attached
that the primary
to form polymers,
from the system

isolated., . If
unsaturated, aliphatic
chains may be formed.
uced when acetylene

the C+ ion, the

Benzene reacts in the silent electric discharge,
commonly identified product is diphenyl but it is always
accompanied by complex resinous material. In the gaseou_

considerable quantities.

The most
In the gaseous products

hydrogen and acetylene are found in

the reaction products are quickly cooled, dihydrobiphenyl can be

obtained. .

H H

dihydrobiphe nyl

products

net
cooled

Thus benzene aeems capable of undergoing two types of reactions.

> — > polymerization and dehydrogenation (biphenyl)

C6H6 <^ 4Q^

x _Zr — ,. depolymerization (acetylene)

A spectroscopic study of the light emitted during the reaction
shows the presence of C ions, carbon and hydrogen atoms, C2 and
CH molecules. This would indicate the presence of a complex
mixture.

Paraffin hydrocarbons , being less reactive than olefins and
acetylenes, require a more intense die charge to initiate a re-
action. With the more intense discharge a high temperature is
developed and thermal activation is at least partially responsible
for the observed reactions. When methane is subjected to an
ozonizing discharge at room temperature, the reaction is thought
tc take the following general course.

_ 4 -

polymerizes > iiqUid and solid products

CH4 -» H3 + C3H4 +ris > C3H6

"0 h.4

CaH

3^8

G^Kp, + .7

Aiia_i_^3 c4h10

Very little ethylene was actually isolated, but it is not sur-
prising for, if forced, it should immediately react in any dis-
charge which is intense enough to activate methane. In a glow
discharge methane can toe made to yield acetylene but the
products must be quickly removed from the reaction zone and cooled
to prevent the acetylene from undergoing further reaction.

Ethane can undergo both dehydrogenation due to activation
of the C-H bond and demethanation cue to the activation of the
C-C bond.

CsH6 — > Csn^ + H3

2C3H6 ~* OH* + C3K8

The latter reaction is characteristic of the saturated hydro-
carbons in an electrical discharge. This reaction can possibly
be accounted for by assuming the ion-cluster mechanism which
was proposed to account for the reactions of hydrocarbons with
a particles.

„ „ discharge _ ,T + ,„ *

C3n6 1 ^=-> C3H6 + e (1; ionization

0aHe+ + C3H6 -► [C3H:6.C3Hg] {2) clustering

[C3H6.C3K6] ~* LCr^.CgHe] (3) rearrangement of

the valence bonds

[CH4.C3H8] + e — + OH4 + C3HS (4) decomposition

It must be assumed that the ion-cluster is stable for a long
enough period of time to allow the valence bonds to rearrange.
Propane undergoes the same type of reactions.

Reactions of Hydrocarbons in Disruptive Discharges

The disruptive discharges consist of the arc and the spark.
While there may be electrical effects on hydrocarbons in the
spark discharge, there is reasonable doubt that there is any such
effect in the arc. Both discharges cause the same type of re-
action as does high teaperatur e, i.e. above 1500°C. Carbon and
hydrogen are the main products in most cases and there is usually
a considerable amount of acetylene formed. Traces of ethylene,
dipropargyl (HC=CCH3CH2C=OH), benzene, and naphthalene have been
found when using methane, indicating a thorough-going disruption
and recombination. In arc and spark discharges, the electrode

- 5 -

material is of considerable importance and there is seme evidence
to indicate that in the case of graphite electrodes, they may
actually take part in the reaction. It has been noticed that
benzene has a very distinct tendency to form carbon and hydrogen.

Activation of. the _Hc 1 eou.1 e s

Most of the reactions typical of rn electrical discharge
take place in the silent discharge, Taking the gl cw discharge
as an example o^ this type the process of activation will be
considered in a little more detail, The activating electrons are
produced ?t i;he cathode Moving under the force of electro-
static repulsion away fsori. the negative polo. they migrate to
the Grookes dark space Tills area i : the region of greatest
potential drop and the electron experiences .an accelerating force
toward the anode. During this acceleration, the electrons under-
go three types of collisions with molecules,

(1) elastic collisions in which no energy is lost or gained

(2) ionizing collisions in which a secondary electron and a
positive ion are formed

(3) activating collisions in which the quantum state of the
molecule is raised

Each electron leaving the cathode is Known to produce from 50 to
100 secondary electrons by ionizing collisions and since the
energy of ionization is much greater than that required for
activation, it is assumed that many times that number of activat-
ing collisions take place. At all pressures above .1 mm of Hg,
electrons are thought to be the sole factors producing activa-
tion and ionization. The positive ions present are so heavy
that there is not a sufficiently high potential drop to give them
a high velocity, and they are large and, therefore, must hit
other molecules very often, Both of these conditions operate to
prevent the positive ions from ever acquiring sufficient energy
to cause an ionizing or activating collision Elastic
collisions may be neglected as only the ionized and activated
molecules have the energy necessary for chemical reaction. Very
little is known about the reaction of hydrocarbon ions except what
little can be deduced from the nature of the products. The course
of the reaction of tne activated molecules can be followed in
many cases by analogy with reactions which take place- under the
influence of thermal activation.

Commercial Anplications

The action of the silent discharge is being used on a
commercial scale to produce and improve lubricating oils. The
process is known as "Siectricnization" in Belgium and as
"Voltolizati on'-' in Germany. The oil is put in a large mechanical
device which is arranged in such a way as to provide thorough
mixing and a short exposure to the silent discharge. The reaction
is run at about 80°C and the pressure is about 60- -65 mm. The
electrical energy is furnished by a 500 cycle discharge at approxi-
mately 4000 volts. The action of the discharge dehydrogenates
the molecules which then polymerize. Fifteen per cent of the

- 6 - 33

product is blended with untreated oil. The mixture possesses
the following desirable characteristics: (l) the viscosity is
increased, (2) the viscosity temperature coefficient is decreased,
(3) the pour point is lowered, (4) the sludge-forming tendency
of the oil is decreased, (o) the tendency to form emulsions
with water is increased.

The production of acetylene from methane on a commercial
scale can be accomplished at the present time. In the carbide
process we have the reaction:

o
CaO + 5C -» CaC3 + CO ^F29e - 88,400 cal/mol

For the discharge process:

o
2CH4 -* C2H2 + 3Ha AF298 - 75,400 cal/mol

Translated into electrical units this means the carbide process
requires theoretically 5.6 kilowatt hours as against 4.7 for the
discharge process per cubic meter of acetylene. In actual
practise the energy consumption is considerably larger, requir-
ing from 9 to 18 kilowatt hours, depending on the type of furnace.
Various experiments have been carried out with methane. When
the methane is diluted with two volumes of hydrogen to cut down
the carbon forming reaction, it was found that o&% of the methane
which reacted, bl% of Che methane charged was converted to
acetylene giving an energy consumption of 12 kilowatt hours per
cubic meter of acetylene in an arc. It was possible with a glow
discharge tube to obtain yields up to 90$ of methane charged.
Using liquid hydrocarbons, a process has been developed which re-
quires 15.5 kilowatt hours per cubic meter of acetylene but, at
the same time, 1.4 cubic meters of hydrogen are produced as a by-
product.

Several other processes, i.e. the preparation of diphenyl
from benzene, butadiene from fuel oil, and various processes for
the preparation of gasoline hydrocarbons by cracking in an
electrical discharge, have been patented. As far as is known, "
none of these are in actual operation at the present time.

It may be seen that an electrical discharge is a powerful
tool in causing the activation and subsequent reaction of hydro-
carbons. It is possible to expose the reactants to a wide variety
of experimental conditions. The primary reactions seem to be de-
ny drogenation and polymerization. In many cases these are
followed by various secondary reactions which give rise to a large
number of products. One commercial process, i.e. Voltolization,
has been in operation for a number of years and other useful
products may be awaiting only the proper experimental conditions
for large-scale production.

Bibliography:

Harking and Gans, J. Am._Chem. Soc. , 5_2, 2578 (1930).

de Saint-Ceunay, Chi mi e Industrie 29, 1011 (1933).

Brewer, Chem. Rev., £1, 213 (1937).

Thomas, Egloff and Morrell, ibid , , 28, 1 (1941).

Reported by R. G> Chase February 25, 1942

WAR GASES 34

Perhaps the commonest classification of war gases is the
physiopathological classification; that is, a classification
based on the most characteristic action of the war gas on
human beings. By this method, the gases usually are divided
into the following classes:

(1) Lachrymators, or tear gases;

(2) Sternutators, or "sneeze" gases;

(3) Lung injurants;

(4) Toxic gases-those gases having a harmful systemic
effect;

(5) Vesicants, or substances producing blister's on the skin.

This classification is not rigid, since the biological action
of these substances is complex. For instance, chloropicrin is
a lachrymator, a toxic gas, and a lung injurant. , Nevertheless,
this classification is useful in helping to present a clear pic-
ture of this subject.

(A) Lachrymators

The many substances which have been used as lachrymators
include a- end (3-halogenated ethers, ethyl bromoacetate, acro-
lein and its halogenated derivatives, oc-halogenated ketones,
the benzyl and xylyl halides, o_-nitrochlorobenzene, bromobenzyl
cyanide, and phenylcarbylamine chloride.

In general, it has been found that the aromatic compounds
are more satisfactory than the aliphatic compounds. Of the
substances mentioned above, bromsbenzyl cyanide and chloro-
acetophenone are perhaps the most satisfactory lachrymators.

Bromobenzyl cyanide can be prepared in the following way:

C6H5 - CH.ei-gjS-* C6HS - CH2CN -. ,Br« vC^-CH^'
yu ultra-violet 6 5 Vg

Bromobenzyl cyanide is a low-melting solid. One method in
which it may be used is to dissolve it in chloropicrin, which
in itself is a war gas. On heating to 160°, it decomposes to
form dicyanostilbene:

,CN C6H5 - C - CN

2 CeH5 - CH > II + 2 KBr

NBr C6H5 - C - CN

For this reason, its use is limited because of its low sta-
baility t© the explosion of a bursting shell. Another drawback
is that it attacks metals, se that specially-lined containers-

%
-2- ^>0

have to be used.

Chloroacetophenone can be prepared by the fallowing method,

C6HG + CICH3 - C0C1 A1C]:3 .,c6Hs - GO - CH2C1 + KC1.

This substance is also a sclid. One method of using it is to
spray a benzene solution of it into the air. "When the benzene
evaporates, the chloroacetophenone is dispersed in a state of
fine subdivision.

(B) Sternutators

Some of the representative sternutators are phenyl dichloro-
arsine, diphenyl chloroarsine, diphenyl cyanoarsine, pnd phena;:*-
sazine chloride (Adamsite).

( 1 ) Phenyl Dichloroarsine and Diphenyl Chloroarsine

A mixture of the above two compounds was prepared by the
Allies in the last war by the following method:

AsCl3 + 3 C6HBC1 + 6Na- >(C6H5)3 As + 01a CI

(C6H5)3 As + AsGl3 ~^^~* C6H5 ~ AsCl3 + vC6H5)2 -AsCl

60 - 65^ 35 - 40^

This mixture, which was used without further treatment, solid-
ifies to form a low-melting mass. When it was dispersed in the
form of fine particles, it passed through the gas masks then in
use, causing sneezing and vomiting. Since then, layers of felt
have been put into the gee mask to mechanically filter the
sternutator .

( 2 ) Diphenyl Cyanoarsine

This substance can be prepared by the reaction of hydrogen
cyanide with diphenyl arsenious oxide as shown by the following
series of reactions:

2 (C6H5)2 - AsCl + H30 —+ HC1 + [(C6H5)2 As]2 0

[(CSH5)2 As]2 0 + 2 HON * 2 (C6H5)2 AsCN + H20

This also is a low-melting solid, and is utilized in the
same way as diphenyl chloroarsine.

(3) Phenarsa yine Chloride (Adamsite )

This compound was studied in 1918, but wp s not used in the
last war. It wee prepared at Edgewood by the reaction between
diphenylamine and arsenic chloride:

T Yield, Q0%

Later, it was found the t diphenylanine hydrochloride dis-
sociates into the amine and hydrogen chloride at about 100- ,
so the following method was developed: £6kU

/ \

2 (C6HS)S NH.HC1 + Ass 03 >3H30 + 2HN AsCl

^6H4X
95$

This substance melts at 190 , end is also employed in the
form of a fine dust.

( C ) Lung Injurants

Substances which have been employed as lung irritants
include chlorine, thiophoegene , chlcropicrin, phosgene, and tri-
chloromethyl chlorof ormate.

Due to its low persistence, chlorine was not used to a great
extent. In addition to being a lung injurant, phosgene is very
injurious to th- heart."' Phosgene was used extensively, but it
has the disadvantage of hydrolyzing easily.

On the other hand, trichloromethyl chlorof orma te is quite
resistant to hydrolysis at ordinary temperatures. This compound
can be prepared in the following way:

CH30H + C0C12
0

■» HC1 + CI - C

!1:

\)CHo Ultra-violet

CI - c

0-CC1.

This substance is also known as diphosgene, since on heating it
decomposes into two moles of phosgene. It is completely re-
tained by filters of active carbon.

Chlcropicrin was prepared by the action of calcium hypo-
chlorite on picric acid, but now ca-n be prepared by the action
of chlorine on nitrometnane :

CH3 - N03 + Cl3 C? (0C1),5.) Cl:

NO.

Chloropicrin is not as toxic as phosgene or diphosgene, and of
the war gases is one of those most easily held back by active
carbon.

(D) Toxic Gases

Among the toxic gases are cyanogen chloride, dimethyl sul-
fate, carbon monoxide, hydrogen cyanide, and tetrachlor'9 dinitro-
ethane.

3*«

~4~

Carbon monoxide and hydrogen cyanide °re too volatile.
Cyanogen chloride had a limited use Gs a war gas due to its
tendency to trimerize to form cyaruryl chloride.

N *N

3C1CN -> « i

C1~C J3-C1

Dimethyl sulfate is unsuitable because of its low volatility*

Since the last war, it has been found that tetrachloro-
dinitroetha.ne is a very effective toxic gas. It hr s been pre-
pared as follows:

2L -- - „„, N02 C13„C-N03

CI3C-CCI3 > ci2c = cci, -T6^ ci;j_N0;

This substance is six times as toxic and eight times as lach-
rymatory as chlorcpicrin.

(E) Vesiccnt s

The most important vesicant used in the last war was mus-
tard gas, P, |3r~dichlorodiethyl sulphide* In addition to being
a vesicant, mustard gas is five times as toxic as phosgene.

The method of preparation used by the Allies was that
developed by Pope in England and Levinstein in America, Ethylene
is bubbled into sulfur monochloride while the temperature is
kept below 35°, According to Conant; the following reactions
take place:

S3C13 : S + SC13

/C1

CH3 = CH2 + SC12 -* S

NCH* - CHa - CI

CH3 = CH3 + s' ^H3 - CH3 - CI

XCH3 ~ CK3 - CI -* ?"

vH3 — CH3 — CI

The Germans prepared it from thiodiglycol and concentrated
hydrochloric acid by the following series of reactions:

HC1
20H - CH3 - CH3 -CI + Na3S->S(CH3 - CH3OH)3 — -**-»

S(CH3 - CH3C1)

The Allied method was cheaper, but the German method was
more easily controlled. Recently, a cheap method of preparation
of thiodiglycol has been developed:

~5-
2H3S + 1 N)->S(CH3 - CHsOH)s oS

ch/

Consequently, the thiodiglycol method of preparation of mustard
gas may now be the preferable method.

Bleaching powder reacts with mustard gas both as an
oxidizing agent anc* as a chlorinating agent, rendering it harm-
less. For this reason, bleaching power is used to decontaminate
areas infected by mustard gas.

Chloramine-T reacts with mustard gas to form a sulfilimine:

CK3 -C6H.i-30^f/ + S(CH3 - CH3Cl)3^CH3 - C6H4 - S03 -
XC1

N<~S(CH3 - CH3C1)

Thus, chloramine-T can also be used as a decontaminant .

Lewisite is ^-chlorovinyl dichloroarsine, Cl3-As(CH=CHCl)3,
which is prepared from acetylene and arsenic trichloride:

^- „TJ . „, A1C1, fCl3As - CH = CHC1
CH= CH + AsCl3 ±±2±L^ ClAe(CH = CHCl)3

(As(CK = CHC1)3

The chlorovinyl dichlorearsine can be separated from the other
two products by fractional distillation. The secondary and
tertiary arsines can be converted into Lewisite by heating with
arsenic trichloride, so that the over-all yield is satisfactory.

On oxidation, Lewisite yields chlorovinyl arsenic acid,
CI ~ CH = CH ~ As03H3,. which is innocuous.

Lewisite, besides being a vesicant and a toxic gas, is an
irritant to the eyes and lungs, so that it is potentially a
very effective war gas. This gas was developed too late to be
used in the last war, so that it has not been tested in actual
combat.

The vesicants proved to be the most effective gases in the
last war. For instance, in the last year of the war, 77% of
the British gas casualties were caused by mustard gas.

Bibliography

Sartori, _fhe War Gases, D. Van Nostrand Co., (1939).

Prentiss, Chemicals in War, McGraw-Hill Book 0o0, (1937).

B. M. Vanderbilt (to Commercial Solvents Corporation). U. S.

Patent, 2,181,411 (1939). Ch em, Aba., 34, 1993, (1940).

Nenltzescu and Scarla.tescu, Antigaz. 9, 12, (.1935).

Othmer and Kern, Ind . Sng. Chem. 32 ', 160, (1910).

Reported by C. F. Jelinek
March 4, 1942

DEVELOPMENTS IN QUANTITATIVE ORGANIC ANALYSIS

This report is an attempt to note only some of the high
spots in recent developments of quantitative analytical tech-
nique, For clarity, brief descriptions of some methods in
common use will be included.

Signer has introduced a process of molecular weight deter-
mination which, although not new, has received little attention.
Clark describes the method, together with simplified apparatus
for its application. Solutions of the unknown and of a known
compound in the same so]. vent are placed in a sealed, isc thermally
insulated system so that their vapors are in contact. Following
Racult's law. the solvent from one solution will distill into
the other until equilibrium is reached. A measurement ; of the
volume of each solution will then give the molecular weight of
the unknown according to the equation

G-XMV
x ="GV7

where M is molecular weight, G is weight, and V is solution
volume of the standard, and the symbols with subscript x refer
to these values for the unknown. The advantage of the meth©d
is its high accuracy (reported error usually less than 1%,
frequently less than 0.4$); its principal disadvantage lies in
the time consumed in reaching equilibrium (several days).

Most of the recent cha.nges in the standard carbon-hydrogen
combustion determination have been introduced to take care of
the nitrogen oxides from N-containing compounds, because the
lead dioxide fails to do so satisfactorily. Proposed substitutes
include other metal oxides to absorb the nitrogen oxides or
metals to reduce them, or an absorbing agent placed in a U-tube
between the anhydrone and the ascarite absorption tubes. Tests
by Elving and McElroy en a large number of these materials
showed them to be unsatisfactory, either because of incomplete
removal or because of a tendency to absorb C02 or water. They
recommend a strong oxidizing agent (potassium permanganate or
dichromate) in concentrated sulfuric acid solution placed in a
U-tube between the two absorption tubes. Niederl and Whitman
mix the sample with copper oxide and employ a stream of nitrogen;
the oxide oxidizes the sample and the freshly reduced copper
reduces the nitrogea oxides.

Several wet oxidation methods have been developed to simplify
the apparatus required for combustion. Oxidizing agents used
include permanganate, iodate, dichromate, or persulfate, in
solution in concentrated sulfuric or phosphoric acid. The vilume
of evolved C03 may be measured, or it may be determined gravi-
metrlcally or volumetrlcally . Christensen has recently combined
several previous methods in the following microprocedure: the
sample is oxidized with potassium iodate and sulfuric acid; the
carbon dioxide is absorbed in standard barium hydroxide solution
and the excess base is titrated. Since such a procedure gives

~c -

only carbon content, Williams back-titrates the excess oxidizing
agent (iodate or dichromate) and from this value of equivalent
oxygen 0R he calculates hydrogen content..

foil - (#uR 4- loc .- 11/3 '^C)/3 93

This equation is applicable to compounds containing only C; H,
and 0. He reports that the accuracy of the latter value is
greater than that from the usual combustion procedure.. Williams
also uses the oxidation equivalent E and the exao I; molecular
weight M of an unknown compound OHO to obtain its formula*

12x + y + 15 z = M
2x + 0t5y - z = M/E

This oxidation equivalent E is the number of grams of the com-
pound equivalent to one gram-atom of required oxygen. M must
be even, and M/E must be a whole number. The equations are
solved for x and y_ in terms of z and a few trial values of z
readily show which is correct. Compounds containing S and N
can be determined similarly provided the content of these ele-'
ments is known and certain corrections are made. Wet combustions
fail with certain compounds. Acetic acid, phyhalic acid, and
pyridine types are unresponsive to one or more of the oxidants
mentioned. Also, dichromate frequently produces some carbon
monoxide, and the use of phosphoric acid medium leads to oxida-
tion of part of the ammonia from nitrogen compounds.

Dumas nitrogen microanalysis often gives high results,
especially when the sample contains more than 40% N. Fischer
ascribes this ta oxygen from a carbon monoxide- cuprous oxide-
oxygen equilibrium at the high temperature in the combustion
tube and suggests a separate furnace kept at only 200°C for the
farther end of the tube. He reports a result 0,28^ high on a
sample containing 80#N.

In the standard Kjeldahl amino nitrogen procedure the
sample is digested with sulfuric acid and a catalyst t*> pro-
duce ammonium sulfate; the solution is made alkaline with strong
caustic and the ammonia steam-distilled into a standard acid
solution. This method is limited to amines, amides, and amino
acids; nitro groups must be absent, and heterocyclic nitrogen
is in general not attacked. Sometimes the method is used for
total nitrogen by first reducing it, as in the Friedrich pro-
cedure which uses HI and. red phosphorus. The excess HI is
removed with sulfuric acid, and heat. Belcher and Godbert have
applied this method to a wide range of compounds, including
those with hetero nitrogen; they recommend as a catalyst for the
Kjeldahl digestion a mixture of potassium sulfate, mercuric
sulfate, and selenium. The reported results were rather con-
sistently low by about 0.1$ N.

Huffman has perfected a way by which sulfur can be deter-
mined during carbon-hydrogen combustion. The oxides of sulfur
are absorbed on silver pellets to form silver sulfate, and this

is electrolyzed in dilute edueous isopropyl alcohol (so that
an adherent plete may be obtained.) Halogens must be absent.

The standard Zerewitinof f procedure has been modified by
Evans, Davenport, and Revuke s for improving the accuracy of
micro determinations of active hydrogen. They burn the hydro-
carbon evolved and determine C and H in the usual manner.
n-Butyl Grignard is used to obtain more weighable product per
active hydrogen than methyl G-rignard would give.

The classical method for determination of acetyl is by trans-
eeterif ication with "evthenol, distillation of the ethyl acetate
and its saponification in standard alkali. Ma.tchett and Levine
have eliminated the unwieldy procedure of introducing ethanol
vapors during trans-eeterificstion by employing an efficient
fractionating column with total reflux; small fractions of the
distillate at the top are removed at Intervals. This also per-
mits the use of HC1 catalyst in place of the less available
aromatic sulfonic acids.

Methoxyl or ethoxyl content is commonly found by a modifi-
cation of the Z-isel method. The alkoxyl is converted, by HI
to the alkyl iodide, which is distilled into bromine solution.
The excess bromine is destroyed by formic acid, potassium iodide
is added, and the liberated iodine is titrated with thio sulfate.
The re actions are:

RI + Br2 ->RIBr2 -> RBr + IBr

IBr + 2Br2 + 3H20->HI03 + 5HBr

HI03 + 5HI -> ftla + 3K20

Lisle reports * rapid method for approximate determination of
methoxyl. Evolved methyl iodide i's massed over a tegrfc paper
saturated with a palladium chloride-pyridine solution, which
turns brown. After drying it is compered with a set of standards.
The question of whether the methio group will react to the Zeieel
procedure has been investigated. Arndt shows the relationship
of structure to ease of splitting: S-elkyl> S-aryl > S-cerbonyl .
He concludes that the method is generally not applicable, but
that at the same time methio groups will interfere in alkoxyl
determinations .

An illustration of the applicability of absorption spectra
to quantitative organic analysis is given by the work of Gore
and Fatberg on the determination of toluene purity. Most im-
purities are aliphatic hydrocarbons. The aliphatic C-H bond
shows a marked absorption peak in the Infrared region at 3.4y*,
and another smaller peak at 3.49X'. The aromatic C-H bond
exhibits a much lower oeak at 3.3,*, Fig. 1 compares the absorp-
tion curves for cyclohexpne (upper) and toluene. Small quantities
of paraffin present in toluene will show themselves in the peak
at 3.4^ in spite of the toluene methyl. By arbitrarily assuming
a single compound es the impurity (e.g, cyclohexane ) it is

-4-

possible to estimate the degre? fif purity of the toluene from
thp relative absorption at this w?ve length according to Beer*s
lav:

log I0/I = kc

where I, and I reprfsent intensity of radiation through pure
toluene in solution and that through the sample containing im-
purity of concentration c. Results from pure toluene with
added traces of cyclohexane gave fig. 2. The procedure is
rapid and indicates relative purity of the toluene with suffi-
cient accuracy for industrial purposes.

exti'nct.
coeff.

10-

Fi3

wave

Bibliography:
Pregl, Roth, and

P. Blakiston's Son
Niederl and Niederl, "Micromethods

Elementary Analysis," John Wiley

Daw, "Quantitative
and Go . ,

Orgfnic Micro-analysis,"
Phila., 1937, 2nd Eng. Ed.

of Quantitative Organic
and Sons, Inc., New York,

1938

Ed. 2

Shriner, "Quantitative Analysis of Organic Compounds," 1941,
Williams, J. Am. Chem. Soc, 59, 288 (1957); Williams, Rohrman,

and Christensen, ibid., 59, 291 (1937); Christensen, Williams,

and King, ibid., 59, 293 (1937).
Arndt, Loewe, and Ozansoy, Ber., 72B, 1860 (1939).
Lisle, Analyst, 64/ 876 (1939)
Fiech£>2-, Chem. Fsbrik, 1940. 154.

Belcher and Godbert, J.^Soc. Chem. Ind. (Trans.) 60, 196
•<3J^iBterrt?en. Wong, and Facer, Ind. Eng. Chem., Anal. Ed.

364 (1940).
Huffman, ibid. . 12, 53 (1940).
Evans, Davenport, and Revukae, ibid., 12, 301 (1940 )

(1941)
12,

-5-

Matchett and Levine, ibid. . 13, 98 (1941).
Christeneea ?nd Wong, Ibid. , 13, 444 (1941)
Elving pnd McElroy, Ibid. . 15, 660 (1941).
Cxore ?nd Prtberg, ibid., 13. 768 (1941).
Clerk, ibid., 13, 820 (1941).

Reported by G-. E. Inskeep
Merch 4, 1942

RECENT METHODS FOR MAKING -ACIDS
I. Condensations with Oxalyl Chloride

In the presence of light and at room temperature oxalyl
chloride and typical saturated hydrocarbons react with the form-
ation of an acid chloride and the liberation of carbon monoxide
and hydrogen chloride. The reaction appears to be general for
paraffinic end cycloparaff inic hydrocarbons and may be represented^
p s

H2
H3

H H

(C0C1);

H C0C1

+ CO + HC1

2 Ky

The yields are quantitative— that is, a mole of acid
chloride is produced for each mole of oxalyl chloride used.
The conversions, however*, vary from a few per cent for n-heptane
to more than fifty per cent for cyclohexane* The chief cause
of low conversions is the formation of colored matter in the
reaction mixture which shields trie oxalyl chloride from effective
radiation. •

Benzene inhibits the reaction, apparently by the absorp-
tion and degradation of the effective radiation. Toluene and
related compounds Co not react. This is presumed to be due to
the effect of the aromatic ring on the transmission of the
radiation reouired for activation of the oxalyl chloride.

From a consideration of the energy involved the following
mechanism is proposed

Primary process:

(COCl)s -^2-COCl

(C0C1);

I0C0C1 + .CI

Secondary process;

.♦COd'-* CO + CI-
•C0C0C1 -> 2 CO + CI-
RH + CI* -**•»• + HC1
R- + (COCl)a -+RC0C1
•C0C1 -> CO + CI-
CI- + RH -*R« + HC1

-'r

:oci

It was found that the addition of small quantities of
benzoyl peroxide to a reaction mixture of cyclohexane and oxalyl
chloride induced a reaction in which the conversion was practi-
cally complete and the yield was sixty-five per cent.

Neither light nor peroxides h°ve any apparent effect upon
the action of oxalyl chloride- with unsaturated compounds. Gentle
refluxing, however, brings ;bout a reaction which may be

-2-

represented ps

RHC=CH3 + (COCl)s ->RKC=CHC0C1 + CO + HC1

Yields very from over fifty per cent for 1,1-diphenyl-
ethylene to six per cent for 1-methylcyclohexene.

Phenyl acetylene reacts somewhat differently

C6H5C=CH + (COCl)8 ->C6HSC=CC0C1 + CO

CI NH

Most unsaturated compounds do not react under these mild
conditions but the effect of typical catalysts on the conden-
sation was not studied. A highly polar double or triple bond-
is a prereauisite for reaction. The reaction is unaffected
by the usual catalysts and inhibitors for reactions involving
atoms and free radicals. The authors, therefore, believe that
the reaction is strictly of the polar type.

II. Tertiary Cerboxylic Acid Esters

Aston ''n<3 his workers have shown the t sodium alcoholates
in ether act on a-bromo-a, a-oialkyl ketones to yield the ester
of a tertiary acid. The following table gives some esters with
the yield obtained using this method

Isopropyl trimethylacetate 64^

Methyl time thy lac eta te 39^x

Methyl trimethylacetate 61. Z%

Methyl methyl~t-butylacetate 73^

Methyl ethylmethylpropylacetate 75^2

x-the low yield is due to an excess of alcohol used in

preparing the alcoholate.
3-a trace of methoxy ketone formed in the reaction is

included in this yield.

The steps postulated for the reaction are:

R'(R»)CBr-CO-R" ' + NaOR -> R-7 (rM )CBr~C(OR ) (ONe )~RT T T

I
-» R1 (rU )c_C (OR)-R ■ « t _> riri i.R. . iCC00R
\/ II III

Two other reactions are possible:

1. Normal metathesis; e.g. 3-alkoxy-2-butanone from

3~bromo-2-butenone
2.

R*R'»C — -C(OR)-R'»l-+ R,,^OH-^R'R"C(OH)~C(OR)(OR'»").

\) II

-3-

When no alcohol is present (sodium alcoholate suspended
in ether), reaction (2) which requires alcohol is not possible.
The possibility that a change of medium alone is the factor
influencing the rearrangement to the ester is eliminated by
the fact that if any alcohol is present a corresponding amount
of hydroxy ecetal is obtained in ether medium.

The formation of addition product (I) rather than the
product of metathesis is ascribed to steric hindrance. That
the presence of hydrogen on the a-earbon is not the determining
factor is shown by the seventy-three per cent yield of the
pure methyl ester of methyl-t~butylacetic acid obtained from
4, 4-dimethyl-3-bromo-2-pentanone. The rearrangement is retarded
by making the rearranging alkyl or aryl group larger. Thus
cc-bromoisobutyrophenone gave only the a-methoxyisobutyrophenone.

III. Glycolic Acid

Glycolic acid is produced by heating formaldehyde, carbon
monoxide and water at 160-170° for one hour at pressures
between five and fifteen hundred atmospheres. An inorganic
acidic material such as sulfuric acid, dissolved in an organic
acid such as acetic acid, serves as a catalyst. Other suitable
catalysts are boron trif luoridc: , hydrochloric, phosphoric,
formic and glycolic acids.

IV. Substituted Acetic Acids

Substituted acetic acids are formed when carbon monoxide,
formaldehyde or one of its polymers and an inorganic acid react
in the presence of en acidic catalyst according to the equation

n (ECHO) + n (CO) + HnX -> X(CHsCOOH )

When sn organic acid is used, acyloxyacetic acids are
obtained. Thus propionoxyacetic acid would be obtained from
formaldehyde, carbon monoxide and propionic acid.

V. Acetic and Propionic Acids

Acetic acid is produced when methyl alcohol and carbon
monoxide are heated at a temperature of 125-180° and pressures
greater than twenty-five atmospheres, 700-900 atmospheres being
most suitable. Boron trifluoride and water are used es the
catalysts.

By the use of ethyl alcohol and carbon monoxide under
similar conditione, propionic acid is produced. Propionic acid
is also made from a hydra ting agent such as water, carbon
monoxide, and ethylene using boron trifluoride and one to three
moles of water as the catalyst.

In an analogous manner, polycerboxylic acids are obtained
from polyhydroxy alcohols such as ethylene glycol, propylene
glycol and. glycerol.

-4-

The boron trifluoride catalyst is prepared by treating
anhydrous liouid hydrogen fluoride with boric acid, its anhy-
dride or a borate at temperatures below 10°. The reactants
are used in such a proportion that BF3*2H30 or BF3*3H20 will
be formed.

Dreyfus has patented a catalyst consisting of traces of
copoar end ammonium phosphate for the production of acetic
acid by this procedure. The Eastman Kodak Company has a
catalyst containing small amounts of the oxides of zinc and
manganese and larger amounts of the oxides of copper with a
binder such as sodium silicate or cellulose acetate. Other
suitable catalysts are the fluorides of magnesium, calcium and
titanium and the halides of numerous other metals.

VI. Maleic Acid

A process recently patented by the Standard Oil Develop-
ment Company consists in passing a mixture of oxygen, an oxid-
izable unsaturated hydrocarbon containing at least four carbon
atoms and water vapor over a catalyst at a temperature of
250-400°. The oxides or salts of vanadium, nickel, tungsten,
chromium, manganese, molybdenum or mixtures of these are
suitable catalysts.

VII. Fumeric Acid

Fuma.ric acid is produced by the oxidation of furfural.
This is accomplished, by moderate heating in a chlorate solution
using vanadium trioxide and one or more of manganese trioxide,
aluminum oxide or ferric oxide as the catalyst.

VIII. Soap-forming Carboxylic Acids

British and German patents describe a method for the
preparation of soap-forming carboxylic acids or their salts by
the oxidation of technical mixtures of hydrocarbons and fusion
with alkali metal or alkaline earth metal basss. The hydro-
carbon must contain one unsaturated link in en aliphatic chain
and at least one saturated aliphatic chain not less than six
carbon etome in length. The unsaturated tar oils boiling at
200~?50° obtained from the distillation of lignite, cracking-gas
oil fractions boiling at 200~;7>250 (seventy to seventy-five per
cent of which are olefins having twelve to eighteen carbon atoms)
and olefinic products from the hydrogenation of carbon 'monoxide
are suitable starting materials..

Permanganates, oersulf stes, hypochlorites, hydrogen
peroxide, chlorine, oxygen and air are used as oxidizing agents
and in some cases tne hydrocarbons are emulsified before oxid-
ation.;; The use of solvents or diluents such ss mineral oils
or saturated hydrocarbon? which may be present in the initial
hydrocarbon mixture is beneficial.

-5-

In a typical example, a hydrocarbon having a boiling
point above 250° is oxidized at 115° by means of air in the
presence of heavy metal salts of unsaturated acids. The
dried product may then be fused with alkali at 250-280° *
The product consists of salts of stearic, palmitic, myristic",
lauric and capric acids,

IX. a-Chloroacrylic Acid

Paraformaldehyde, trichloroethyiene; 98^ sulfuric acid
and metallic copper (in the ratio 30' 132-84-1 parts) are
heated at 25-30°; the temperature finally being raised to
45°. cc-Chloroacrylic acid is obtained from the reaction mixture

Bibliography

Aston, Clarke, Burgess

64, 300 (1942)
Kharasch, Kane and Brown, i
French Pat. 831,474 (1938);
U.S. Pat. 2,153,064 (1939)

and Greenburg, J. Am. Chem. Soc

bid. ,

Chem.
ibid 3

British Pat. 508,383
British Pat. 527,644
British Pat. 527,645
U. S. Pat. 2,135,453
U. S. Pat. 2,170,325
British Pat. 490,544

1939
1940
1940
1939
1939
1938
1940
1939
1939
1939
1940
1939
( 1940
(1938
U. S. Pat. 2,211,624 (1940
French Pat. 845,230 (1939);

Pat.

217,

650

Pat.

135,

459'

Pat.

135,

454

Pat.

165,

428

Pat.

260,

409

iti

sh Pa

498,

398

Ja]

Danese Pat.

137,755

Briti

sh Pat.

491,

927

ibid
ibid
ibid
ibid
ibid
ibid.
ibid
ibid
ibid
ibid,
ibid,
ibid,
) ibid.
; ibid,
J ibid,
ibid.

64, 333 (1942).

Abs. 33, 2539 (1939)
3, 5006 (1939).

34, 451 (1940).

35, 6981, (1941).
35, 6981 (1941).

33, 995 (1939).

34, 116 (1940).
33, 646 (1939).

35, 1070 (1941).
33, 996 (1939).
33, 1112 (1939).
33, 3214 (1939).

36, 784 (1942).
33, 4272 (1939).
35, 1806 (1941).
33, 1347 (1939).
35, 1068 (1941).
35, 1070 (1941).

Reported by G-, W. Cannon
March 11, 1942.

:•

THE PR INS REACTION

This reaction, discovered by Prins, is the condensation
of formaldehyde with an olefin.

RCH=CHR + CH30 -> RCH CHR H3° » RCH-CHROH

I 1

CHo— 0 CHoOH

SRCH CKR^ /

I \ 7

nu Viij

In some cases the 4-membered ring may rearrange, or the glycol
may dehydrate, to give an unsaturated primary alcohol. . The
condensation is brought about by sulfuric acid in acetic acid,
formic acid, or water. Prins carried out his initial experiments
on styrene, anethole, isosafrol, a-pinene, camphene, and cedrene.

With unsymmetrical olefins, for example styrene, the
addition might proceed in two ways, giving rise to the products:

A. QfCH—CK3 -* 0CH— CH30H

If I

CK3~0 CH20H

B. 0CH— CK2 -> 0CH— CK2CH30H
6.

-CH3 OH

Only one product was obtained; Prins believed it to be of the
structure A, since it gave a diacetate with boiling acetic anhy-
dride, instead of cinnamyl acetate.

Fourneau, Benoi't and Ferminich conclusively proved the
structure to be the unsymmetrical glycol B. They based their
conclusion on the following evidence:

1. Saponification of the diacetate yielded large quantities
of cinnamyl alcohol.

?. The glycol, 1-phenyl-l, ^-propanediol prepared by

Rupe by reducing 0COCH=CHOK gave a dibenzoate of the
same melting point (51°) ^s that from Prins' glycol.

?. The unsymmetrical glycol was prepared from a chloro-
hydrin of known structure.

-2-

C1CH3CH2CH0 + 0MgBr -^ 0CH-CH3CH3C1 ^COgNa> 0CH-CH3GH3OCO0

OH OH I S0G13

0CHCH3CH3OCO0 0 ^3 * ' •0CH-CH3Cti3OCiO0
i(CHa), ci

Prins' glycole -^>Chlorohydrin (CH3)?iNH> JGOGJ,

0CHCH3CH3OCO0

N(CH3)3

The two amino esters were identical.

Prins later offered an explanation as to the mode of
addition of formaldehyde to an un symmetrical olefin derivative.
He assumed that the negative oxygen atom would add to the more
positive carbon atom of the ^C=G<I and he further assumed that
a carbon atom would be the more positive when a hydrogen was
substituted by an atom with a high electron affinity. Thus
the reaction was studied with di-, tri~, and tetrachloroethylene
The neutralization of the cha.rge on the carbon to which the
oxygen became attached, activates the chlorine atom so that
immediate hydrolysis occurs.

The reaction with dichlorcethylene gave only a resin.
Trichloroethylene yielded CH30HCHC1C00H which lost water to
form the ethereal acid 0(CH3CHC1C00H ). The oxygen attacks the
GCl2=group almost exclusively. Tetrachloroethylene yielded
the dichloroacid CHa0HCClPC00H.

Recent Developments and Applications .

Marvel, Harmon, and Riddle found that acetaldehyde con-
denses with vinyl acetate in the presence of sodium to give
a cyclic acetal reminiscent of the classical Prins reaction.

2CH3CK0 + CHa=CH0C0CH3 -> CH3CH-CHaCH-0C0CH3

Or J)

I
CH3

Acid condensing agents had no effect, and no other aldehyde
could be made to undergo the reaction.

Nenitzescu and Przemetzky, in searching for new routes to
8-methyl~l~hydrindanone and similar compounds, discovered that
the best method involved the Prins reaction.

CH3

+ (CHaO).

-3-

CH-

t >CH30H

PBr.

Ma Ionic ^NyCHa

ester

A patent has been issued for the f or ma. t ion o'f a synthetic
plastic from formaldehyde and a cracked petroleum oil fraction.
The reaction is brought about by an acid condensing agent in
the presence of a promoter such as acetic acid.

Cyclic formals can easily be synthesized by reacting one
mole of an aliphatic monoolefin with about two moles of
formaldehyde in the presence of an acid condensing agent.
Similarly, a source of 1, 3-butanediol is the reaction of
formaldehyde hydrate with propylene at super-atmosphere
pressures using HX and ZnCl2.as catalysts.

The important acid, ot-chloroacrylic acid may be manufactured
by heating formaldehyde with trichloroethylene in concentrated
sulfuric acid, and heating the subsequent mixture in the
presence of water.

Bibliography:

Prins, Chem. Weekblad, 16, 1510 (1919)

Prins, Proc. Acad. Sci . Amsterdam, 22, 51 (1919); C. A. 14,

1662 (1920 J
Fourneau, Benoit and Ferminich, Bull. Soc. Chim., 47, 860 (1930 )
Rupe and Miller, Helv. Chim. Acta, 1921, 4, p. 841
Prins, Rec. Trav. Chim., 51, 469 (1932)
Marvel, Harmon and Riddle, J. Org. Chem. 4, 252 (1939)
Nenitzescu and Przemetzky, Ber. 74B, 676 Tl941 )
U. S. Pat. 2,035,123 (1936); C.A. 36,3120 (1936)
British Pat. 507,571 (1939); Ibid, 34, 449 (1940)
U. S. Pat. 2,143,370 (1939); ibid. 33, 2914 (1939)
French Pat. 845,230 (1939): ibid, 35, 1070 (1941)
U. S. Pat. 2,233,835 (1941); ibid, 35, 3651 (1941)

Reported by R. E.Foster
March 11, 1942

Ghloro sulfonic Acid As An ■

Organic Reagent ^>&

I. Preparation of chloro sulfonic acid.

II. Action as a sulfonating and chloro sulfonating agent on

Hydrocarbons and derivatives
Hydroxy compounds
Aldehydes and ketones
Acid derivatives

III. Reactions involving oxidation and chlorination.

IV. Reactions as a qualitative organic reagent.

I. PREPARATION

Numerous examples have been cited in the literature for
the preparation of chloro sulfonic acid, C1S03H, but all are
essentially based upon the production of hydrochloric acid and its
reaction with sulfur trioxide:

HC1 + S03 = CISO3H

For example, dry hydrogen chloride is passed into oleum until no
further absorption occurs, and the chlorosulfonic acid separated
by distillation in a stream of hydrochloric acid. Perhaps a
more widely used method is the addition of solid sodium cnloride
to oleum containing 60-80 mole per cent of sulfur trioxide and
finally separation by distillation.

Chlorosulfonic acid itself is a colorless, fuming liquid
which boils at 158°, melts at -88°, and has a sharp unpleasant
odor. It is very corrosive and therefore must be handled with
care.

II.. SULFONATION AND CHLORO SULFONATE ON

The action of chlorosulfonic acid as a sulfonating reagent,
stressed by Gebauer-Fuelnegg and Haemmerle, is only one of its
extensive and varied reactions.

Starting first with the aromatic hydrocarbons, the initial
step in the reaction is .the elimination of hydrochloric acid
and. the formation of the corresponding sulfonic acid which, in
the presence of excess chlorosulfonic acid, reacts to form the
sulfonyl chloride, thus:

C6H6 + CISO3H = C6H5S03H + HC1, (ben?enesulfonic acid)
C6HBS03H + C1S03H = C6H5S03C1 + H3S04 (benzenesulfonyl

chloride )

The mechanism ie perhaps oversimplified, for the products
depend intimately upon the experimental conditions: the time
of reaction, the temperature, the quantity of reagents, and the
nature and order of additions of the reacting substances. The
type of reaction is determined by the temperature of the
reaction and the amount of chlorosulfonic acid. The results
may be summarized by the following; observations:

A. At low temperature, that is, room temperature although
many reactions are run at 0° and below, and in the
presence of equivalent quantities of the reacting sub-
stances, the sulfonic ^cid is produced.

B. At low temperature and in the presence of an excess of
chlorosulfonic acid, the acid chloride is produced.

C. At higher temperatures, 150°, and in the presence of
equivalent amounts of reacting substances, the sulfone
is produced which may be chlorosulfonated further.

D. At the higher temperatures, and in the presence of
excess chlorosulfonic acid, the product obtained is the
sulfonyl chloride. Under these forced conditions, more
than one sulfonyl group may be introduced.

To illustrate these:

room

CeH6 + C1S03H (excess) — - * C6H5S03C1 (76#)

temp. (benzenesulfonyl chloride)
150°

C6H5 + C1S03H (excess) -» * m-C6H4(S02Cl ) + sulfone.

3 hrs. Tbenzene-m-disulfonyl chloride)

150° excess

C6H5(20 g. ) + C1S03H(15 g. ) — 5> n

10 min. 5 hrs.

(diphenylsulfone-m-di sulfonyl chloride )

Other organic compounds are indicated in the following
table:

-3-

Hydrocarbon

Product s isolated

Toluene

o-xylene
m-xylene
p.- xylene

Naphthalene

Biphenyl

Chlorobenzene
Chlorobenzene and
Bromobenzene
l-Chloro-4-nitro

benzene
Ethyl phenyl sulfone
Aniline
m-phenylenediamine

o- sulfonic acid

p_-sulfonie acid

2, 4-disulfonyl chloride

di (p_~methylphenyl-m~sulfonyl chloride )

sulfone
3, 5-disulfonyl chloride
1, 3-dime thy lphenyl~4- gulf one
1 , 4-dimethylphenyl-2-sulf one
2-eulfonyl chloride
1-sulfonic acid
1-sulfonyl chloride
1, 5-disulfonyl chloride
mono sulfonic acid
4, 4'-disulfonyl chloride
?., 4-disulfonyl chloride
p-mono sulfonic acid
diholosulfobenzide
2-eulfonyl chloride

m- sulfonyl chloride

2,4, 6~tri sulfonyl chloride

4,6 disulfonyl chloride

In its action on aliphatic compounds, chlorosulfonic acid
is similar to sulfuric acid. Branched chain hydrocarbons are
much more reactive, and these offer a possible means of purifi-

cation. Ethylene

adds

C1S03H to form CSH50S03C1

Hydroxy compounds of
very extensively and seem
following mechanism: A mo
followed by the eliminatio
fate which undergoes a Eri
acid, whence the sulfonyl
selected preferentially.

(1) C6HsOH + C1S03H

the aromatic series have been studied
to bring about sulfone t Ion by the
lecular addition compound is formed
n of hydrochloric acid to form & sul-
es rearrangement producing the sulfonic
chloride, the para, position beimg
For example:
OS
nr*CeHBOS03H +j3-K0C6H4S03H

-15

CH.

phenyl hydrogen
sulfate

4-hydroxy benzene
sulfonic acid

(2)

,0H

+ C1S0
(exc

3H -
ess;

AH3

, 2-hydroxytoluene-
2 ^ i d e

+ ClSOoH

110°
i

1 Hr„

QH3
/\s, 0 SO 3 f^s .SO s c 1

JO, _

CK.

ra-cr e so lsulf onylide-di sulfonyl

-4-

The sulfate of equation (l) is formed if the reaction is
carried, out in the presence of pyridine or some other tertiary
base. Other examples are given in the following"

Reactant Prcduote

i \

p_-cresol j)-toluyl hydrogen sulfate

4,4' -dime thyldiphenyl-1 , 2 , 1 ' , 2 ' -

sulf onylide-6, 61 -disulfonyl chloride
2,3-dimethylphenol 4, 6-disulfonyl chloride

2, 3, 2', 3 '-tetrp.me thyldiphenyl-1, 6,1' ,
5,-sulfonylide-4, 4'-disulfonyl

chloride
2,4-dimethylphenol 6-sulfonyi chloride

1,3,1' , 3'-tetramethyldiphenyl-
4, 5 , 4 • 5 » - sulf o ny lide

In the aliphatic series, hydroxy compounds tend to form the
sulfuryl chloride, as ethyl alcohol produces C2H50S02C1. One
interesting application is its use to produce sulfated sugars.
These are converted to the acetates when treated with acetic
anhydride.

Formaldehyde reacts with chloro sulfonic acid to produce
C1CH30S0C1, CH2S04, (C1CH20)3S02, and (C1CH2)30 as if the hypo-
thetical C1CH30H had been formed. Homologues are attached in the
alpha position, producing sulfonic acids; however chloral, being
stable in this position, produces (CCl3CKCl)30 and chloralide.

Acetophenone is of interest because of its unique reaction
iY°-CH* ; 01S0.H „ ffV-CH^O.CX

\s k.^-so3ci

Reactions on acid derivations are indicated in the following
table :

Reactant Products

Sodium acetate acetic anhydride

CHgOCOCl CHgOSOfcCl

ArS03K ArSCUCl

Salicylic acid 3, 5 -disulfonyl chloride

CH3(CONHR)2 di sulfonic acids

The most spectacular reaction of this group is the action
of chloro sulfonic acid on acid chlorides.

nu nr\n% C1S03H 45O poo

CH3C0C1 *--> CH3C03S03C1 -2£-> CH2 (S03H )C0C1 -^—>

CH3S03C1 + C02

03H

-5-

Above 60°, s 3^ yield of 2-methyl-l, 4--pyrone-6~acetic acid
was obtained. Prooionyl chloride reacts similarly to form
CH3CH(S03K)C0C1 and a by-product of ot'-ethyl-p, B •■ -dimethyl-
phorone; butuyryl chloride does not undergo these condensation
reactions, but rather forms dipropyl ketone.

III. OXIDATION AND CKLORINATION

These reactions of oxidation or chlorination are more or
less exceptions to the rule rather than general in application,
however their presence is real and must be considered; besides
they are interesting and unusual.

Res c tent

Products

Hydroquinone

Nitrobenzene

p-phenylenediamine

p-phenylenediamine (ClS03Na)

Pbloroglucinol

Resorcihol

Naphthalene

Cyclohexadiene

Thiophenetole

Thiophenol

Chloranil

Penta chlorophenol

Chloranil

Tetrachlorophenylendiamine

Pentachlorophenol

Hexachlorobenzene

Tetrachlorophthalic anhydride

Benzenesulfonyl chloride

Pe ntachl or 0 thiophenetole

Diphenyldi sulfide

Huntress and co-workers recently applied this versatile
and reactive substance to the identification of organic compound:
aryl halide, alkyl aryl ethers, and alkyl benzenes. The results
have proven very satisfactory, and a procedure has been stan-
dardized for the identification of such compounds. It consists
of adding an excess of chlorosulfonic acid to a cold solution
of the unknown in chloroform solution, separating the sulfonyl
chloride and converting it to the amide by means of solid
ammonium carbonate or concentrated ammonium hydroxide.

C1S03H
» XArSOoCl

Ar X

(NH4)aC0;

4 XA.r S02NH.

He found that under these
possible is invariably formed.

conditions the para isomer where

In general the halid.es of benzene, toluene, and naphthalene
were studied. The iodo compounds all reacted abnormally, thus

(1)

(2) C6H5I

chlorination,
-. ^

(p-IC6H4)sS0;

-6-

Constituents in the ortho and mete positions of iodo com-
pounds reacted according to (l); sulfonea as in (2) were also
produced from f luorobenzene , and from o_-dichloro- and
p_-dibromo benzene.

Some forty-two alkoxybenzenes were studied and found to react
analogously, with only biphenyl compounds behaving pecuriarly.

r> a

In the case of the alkylbenzenes, the para orientation w.
proven by permanganate oxidation to the corresponding sulfamido
benzoic acid where possible. That only one sulf amide group had
been introduced under tne conditions of the reaction was
established by a nitrogen determination.

It was pointed out that even in the cases where abnormal
reactions had taken place, the products served just as well for
the purposes of identification.

Bibliography

Huntress and Autenrieth, J. Amer, Chem. Soc . , 63, 3446 (1941)

Huntress and Carten, ibid. , 62, 511, 603 ( 1940T~

Pollak and Gebauer-Fuelnegg. Monatsh.,47, 109, 537 (1927),

48, 187 (1928)
Lustig and Katscher, ibid. , 48, 87 (1927)

Pollak, Krauss, Katscher, and Lustig, ibid. , 55, 358 (1933)
Burkhardt, J. Chem. Soc, 1933, 337
Young, ibid. . 75, 172 ( 18997"
Chem. Reviews, 25, 67 (1939)

Reported by S. S. Drake
March 18. 1942

SOME DEPIV&TIVES 0r THE POLYHYDROXYBENZENES

Wilson Baker, et al
Oxford University

This paper attempts to cover only a few of the many
possible polyhydroxy benzene derivatives. It is chiefly con-
cerned with certain tetra-, pent a- and hexamethoxybenzenes and
some naturally occurring compounds that may be derived from them

Tetrahydroxybenzene Deri vat iv es

Compounds derived from tetrahydroxybenzene are not often
found in nature, but of these substances the derivatives of
1 :?: 3: 5-tetrahydroxybenzene occur most frequently. Many of
these have been determined, and found to be flavones, flavonols,
and isofiavones.

Derivatives of 1:2: 4:5- tetrahydroxybenzene are contained
in certain lichen coloring matter, and in embeiic acid (2:5-
dihydroxy-;3-lauryl-p_-benzoc'Uinone ) . The latter- has been syn-
thesized by Hasan and Stedman (9) as follows:

GO0lxH2;
^XOH ^

V 0

C0CxlHS3

3n25

fs/ ^v

0*\j?

o^MHMe

C 12^-2 s(n) G13H25

Derivatives of 1:2:3:4 -tetrahydroxybenzene are repres-
ented in nature by parsley apiole (l :4-dimethoxy-2: 3-methylene
dioxyallylbenzene ), fraxetin (7:B-dihydroxy-5-methoxycoumarin ),
and dill apiole ( 3:4-dimethoxy-l :2-methylenedioxy-5-allyl-benzene ).
T^o syntheses of the basic cc-mpound .1: 2: 3:4: tetramethoxybenzene
are Bhwn below v' 1 , 2 ) :

OH CHO

Ar OMe .^

>0H

v/0H

OMe

-2-

OMe
\OMe

OMe

i, }

OMe

OMe
OMe

OMe

A OMe

OMe

!>OMe

The product, (l:2-dihydroxy-3:4~dimethoxybenzene ) ,
resulting from the first series of reactions, was used by B? ker
and Jukes (?) in their synthesis of dill apiole, which occurs
in dill plants, matico oil, and sea-fennel oil.

/Nop

OMe

3Hp=0HCH

OMe

s^ /OMe
V

OMe

CHa=CHCHg

OMe

Dill Apiole

Pentahydroxybenzene Derivatives

Many derivatives of pentahydroxybenzene have been isolated
from plant sources. Some of these include nobiletin (5:6:718:3':
4,-hexamethoxyflavone ) , spinulosin (?:6-dihydroxy-4-methoxy-
2:5-toluquinone ) , erianthin (5:7-dihydroxy-?:6:8:3' :4T~
pentamethoxy flavone), pedicin (2:3~dihydroxy-4:5:6-trimethoxy-
phenyl styryl ketone), and the closely related compound pedicellin
(dimethyl ether of pedicin).

Four separate methods have been recorded in the literature
for the preparation of pentahydroxybenzene derivatives, the
yields varying and in general not good. Baker (5) has devised
two relatively simple procedures which heve been outlined in
Flow Sheet A. Flow Sheet B shows some simple reactions of
these compounds, and in Flow Sheet C are found two syntheses
of pedicellin (5) and also one of funagatin as worked out by
Baker and Raistrick (6). This latter product, 3~hydroxy-4-
methoxy-2: 5-toluquinone, can be isolated as maroon-colored
needles from a mould metabolic product.

Hexahydroxy benzene Derivatives

The only derivative to be considered in this paper is
hexamethoxybenzene . This compound has been prepared by Robinson
f nd Va s ey ( 4 ) , a no by Ba ke r ( 5 ) .

MeO

OMe

-3-

MeO,/
Br

OMe
Br

MeO

OMe
MeCO^S.OMe

MeoL JiOMe

OMe

MeO
MeO

Ac
0

^\.0Me

OMe

0
Ac

OMe
MeOf^\.OMe

MeO

OMe

Bibliography

9.
10.
XI.
12.

Brker and Smith, J. Chem. Soc
Baker, Kir by, end Montgomery,
Baker and Jukes, ibid. \ 1934, 1681.
Robinson and Vesey, ibid., 1941, 660.
Baker, ibid. , 1941, 662.
Baker and Raistrick, ibid. , 1941 ,
Organic Syntheses, 14, 40 (1934)
Einhorn, Cobliner, and Ffeiffer,
Hasan and Stedman, J. Chem. Soc,

, 1951, 2542.
ibid. , 1932, 2876.

670.

Ber., 37, 119 (1904)
1951, 2113.

Chapman, Perkind, and Robinson, ibid. . 1927, 5028.
Graebe and Kess, Ann. 340, 237 (1905).
Dakin, Araer. Chem. J., 42, 477 (1909).

Reported by John D. Gerber
March 18, 1942

Flow Sheet A

OH
HO^\

OMe

MeOi-

OMe
-OMe

OMe

♦ f ij0He

MeO^J>OH
COMe

OMe

MeO^x

Me0\/;

OMe

>OMe

OMe

C03H

COMe

HO^v >OH
OH

COMe

<<v>OMe
OMe

— >

COMe

rAoH

IO^v _/OMe
OMe

■nOMe
4?

H
^V OH

U-OH
OMe MeoLJJoMe

OMe OMe

Flow Sheet B

OMe

M©0^. 1LOH

Moo.

OMe

'OH

OMe
?0\loAc

MeO^^y
OH

OMe
OAc

MeO

OAc

rS
V

OAc

OMe

OAc

OMe

MeO

OMe

V0H

COMe

Flaw Sheet C

HO
HO

OMe
^OMe

-sy

COMe

MeO
MeO

OMe

*w*OMe
OMe'

0 0
8cKaC0

MeOr^ 3 OMe
Meotv >OMe

OMe

MeO-k AOVie
OMe

OH
MeCOr^ j|OMe

MeO-LjoMe
OMe

CCH=CH0
MeOr^r\0Me

MeO ^/ OMe

OMe
Fedicellin

OMe

V/OMe

OMe

COMe
Mer5^ >rOH

OMe

OH
Me^\

OMe

Me^ ^OH
\ J^OMe

Fumagatin

•-

RELATION BETWEEN OPTICAL ROTATORY POWER AND CONSTITUTION

OF THE STEROLS

Evidence for the correlation of optical rotatory power with
constitution of the sterols was presented by Callow and Young
in 1936 after a study of known pairs of compounds between which
the only difference was an inversion of carbon atoms 5, 4, 5
or 17 or the introduction of a double bond. In 15 out of 18
cases there was an increase of dextro-rotatory power by a change
of the 3 hydroxyl from cis to trans relative to the 10 methyl.
Introduction of a double bond at the 4:5 position results in an
increase in dextro-rotation, while introduction at the 5:6
position gives a decrease in dextro-rotation, A double bond
at 8:14 brings a small decrease and at 14:15 a small increase
in d-rotation. Changing the 17 ketone group to a secondary
carbinol brings a decrease in dextro rotation.

However, the results of Callow and Young were merely quali-
tative; the sign of the change but not the magnitude could be
predicted. And exceptions were not rare, even in the limited
data available, although many of these might be ascribed to
impure materials or uncertainty as to structure.

Gorin, Kauzmann and Walter showed the t optical superposition
depended on certain very special conditions to which the carbo-
hydrate molecules readily conform. Steroid molecules in general
cannot conform to these conditions and any method of optical
superposition will not succeed.

Berstein, Kauzmann and Wallis have worked out a method
which in effect divides the steroid molecule into two parts, the
ring system and the side chain, each of which parts functions
as a unit optically and is independent of the other part.

Consider the pairs of compounds, stigmastane and ergostane,
stigma stanone and ergostanone (see next page). In each case
the only difference is in the side chain, one having an ethyl
the other a. methyl on C-24, If the molecule is considered as
being made up of two independent regions of asymmetry, the atoms
in the neighborhood of C-3 and the side chain, 'then the differ-
ence in molecular rotation between the substances in each of the
above pairs should be due only to the difference in effect of
an ethyl and a methyl at C-24, and therefore both differences
should have the same value.

stigmastane +10,470° stimastanone +17,000°

ergostane + 7,670° ergostanone +13,960°

+ 2,800° + 3,040°

Since an error of 1000° in molecular rotation corresponds
to about 2.5° of specific rotation, the agreement of these
values is very good.

-2-

Now compare the pairs ergoetane snd ergostanone, stigmastene
and stigmastanone. In this case the difference in rotation
should be due to the effect of a methylene group and its
environment and a carbonyl group and its environment at C-3.

ergostrnone
ergost?ne

+13,960°
+ 7,670°
+ 5,290°

stigmastanone
stigma stane

+17,000°
+10.470°
+ 6,530°

CH.

ai ^ 23 » /

OHCH3CH2CHCH

CH.

r^^k^

I N

CH.

CHCHpCHoCH

I \

CH.

CH, CH;

Ergostanone

Stigma stane

pH3 CH3

CHCH2CH3CHCH

NCHS

2H5

CH.

CH;

CHCHpCHpCHCH

I CH3
C3H5

Stigmastanone

Whenever the environment of the C<*3 is the some as in
ergostane — that is, with saturated rings and with the 0-5 hydro-
gen trans to the C-10 methyl — the difference in rotation between
the methylene and carbonyl compound should be +6,400°.

cholestanone
cholestane

+15,840°
+ 9,160°
+ 6,680°

Tsitostanone
ysi to stane

+15,760°
+ 8,090°
+ 7,670°

-3-

However, if the C-5 hydrogen is cie to the C-10 methyl, we
would expect the difference between the rotations of the carbonyl
and the methylene compound to have a different value.

coprostanone +14,010°
coorostane ± 9.430°

+ 4,580°

In order to use these relationships more easily, we can set
up a system of notation. Cholestane is taken as the reference
compound and its molecular rotation is denoted by the letter C.
In the change to cholestanone, we can symbolize the change in
rotation as K3t, Thus the rotation of cholestanone is C + K3

The change in rotation in going from cholestane to ergostane
can be called Erg. Thus the rotation of ergostane is C + Erg.
Then knowing the numerical" values for the rotations of these
molecules, we can easily find the value for the individual con-
stants.

cholestane = +9,160° = C
cholestanone = 15,840° = C + K3t

K3t = 15,840° - 9,160°
ergostane = 7,670° = C + Erg
Erg = 7,670° - 9,160°

Using this notation, we can indicate the rotation of ergo-
stanone as C + K3 + Erg + e where e represents the difference
in interaction of the ergostane and cholestane side chains
with the carbonyl and methylene groups at C-3. But since two
centers of asymmetry which are far apart do not influence one
another's contribution to the optical rotation, e should be
negligible. Then since we already know the values of the other
three symbols, we should be able to calculate the molecular
rotation of ergostanone.

C + K3t + Erg = +14,350°
observed value = +13,960°

In this manner a table was set up (Table I) for the values
of a large number of symbols which were assigned to specific
changes on the sterol framework and then these values were used
to calculate the rotation of a number of sterols (Table II).
The calculated and observed values agreed fairly well in most
cases. The authors ascribed large discrepancies to inaccuracies
in obeerved values (difficulties of purification) or to errors
in assigned structures.

In their most recent paper, Berstein, Wilson, and Wallis
adapt a similar method to the calculation of the rotation of
derivatives such as acetates, benzoates and m-dinitrobenzoates
(Tables III and IV) and apply the method to recent experimental
results.

D-

<st<

C5>

•*

<j>

H-

-I-

■■d q

C cd

to -P

. .

rH C

O O

/3 o

• •

• .

9 .

>i<H

if)

Ifl

co o

f — i

CD H1 lO lO
Q ....

n a O) Q) -^ LO

G O rH lO WW

i_iO + I + +

•

—I rH

£ cd

_iO

G CO

q 4J

• •

r-H

<D £

i — 1

•N

•\

•»

•>

EH cu

p) CO

-P

t ^

a co

[

•H C

■J 4-3

O 4-3 «"

If)

> s

i — »

i—J o

> o

O
O

O
P-

i — o

Q C

•>

•»

1 w

•^

-sF

r — 1

r-l

r>-

fio

1 )

^ O

I CD

Cvi

r-l

rH
rH

O
W

H
H

•

1 CD

i i — «•

^ rH

i &

E-t

o
q

CD CD

Cu <»-l
XS 03

o p:;

rH OJ

(0
10

•H

cu 05

i-^

4-3

4-'

— .

*— >*

q cd

* — »

col

^"~N

4-3

CD d

■H

•H

Cfi

•H

e cd

•H

4-3

•H

. .

■f

C 4-3

fc— o

v-»»

x_f O

cti

4-3

O aj

•H

4->

r**

f^\

Px]

q cu

ffi

•H r-l

o UJ

1 — 1

> o

1 1

q ,q

co in

1 — »o

W o

o o

Pr-,

Er-

a>
q

4-3

r-i

c a

•H

M CC

evj

- CO

rO W

cd c

cu o

P o

i rH

4-3

CD ■

bfi

CD '/J

rjj .

..q

•H

o q

q -h

4-3

«P

4^>

cu q

• CO

•H

■P .H

•H

CO Cu

.a p

.,-(

•H

p-.

Pt3

P rQ

H
A-4

£>

• .

en o

"tf

tO CD

00
CD

CO "<^ CO to O)
H CVi rH tO CM
+ + + 1 +

CVi

•N

<+H

+2*

1>S

*=—*

<«;

<ti

o
to

o o

cx> co

rH
I

o
to

CD
■H

CD £

4-5 CO

CO p

O co

r— i £

S o

•H

4^>

•H

>*>

o •

• •

•H

4-3

t£

■P

bfl

•H

■H

fcu

•H

E-t

CO.

GLi.

> 03

•H 3

Sh rH

CD CO

O >

■k

rH
rH

^ £

o a-

<H P

^ P

nd a

■d

£ CO

CD o

crj O

■d

aj o

£>

rH C

O CD

to £

£ fit

0) -H

d o

CJ P

4--> Jh

£ co

CO p

CS rH

CO' .H

P 3

B CS

oj o

1 — i

CD-r-t

P rH

•H rc

3 tf

■^ !!

SDO

03 S

CaIc.[a]D Calc [a]D
using; obs. using Calc'd.

-6-
Table IV

Obs [o].D CM]d of L*JD of

Compound (CHC13) Sterol Sterol

Stigmastanol acetate +15.4 +13.5 +13.4

tf-sitostanol acetate + 9.0 + 7.8 + 8.2

Ergostanol acetate + 6.8 (av.) + 4.8 + 7.5

aj.-sit06tan.ol acetate +39.4 +15.6

ergostanol

jjl-dinitrobenzoate +13.5 (av. ) + 6.6 + 8.6

Stigmasterol acetate -55.6 -51.3

p-sitosterol acetate -41.0 -40.5 -38.2

Brassicasterol acetate -65.0 -64.7 -60.3

(all tables abridged)

Bibliography

Callow and Younr:, Froc . Roy Soc. . (London]) A, 157,194 (1936)
Bernstein and Wallis, J. Am. Chem. Soc, 61,2303 (1939)
Gorin, Kauzmann, and Walter, J. Chem. Phys. , 7., 327 (1939)
Kauzmann, Walter and Eyrinr, Chem. Pie v. , J^,339 (1940)
Bernstein, Kauzmann, and Wallis, J. Orr. Queni. , £..319 (1941 )
Bernstein, Wilson, and Wallis, ibid. , 7,103 T~1942)

Reported by W. H. Kaolan
March 25, 1942

REDUCTION OF CARBONYL GROUPS TO METHYLENE GROUPS , — .

The three chief methods of converting carbonyl groups to
methylene groups are the Wolff-Kishner reduction, the Clemmensen
reduction, and high pressure hydrogenation over copper chromite.
Other methods which have occasional value are reduction with
sodium and alcohol, zinc and alkali, or hydriodic acid end red
phosphorus.

The Wolff-Kishner Reduction (1,2,3,4,5')

The classical method of Wolff is to heat the hydra zone of
an alddhyde or ketone with 5-10^ sodium ethoxide at 180-200°
in a sealed, tube.

>C=0 + N2H4-H20_» >C=NNH3 N^0Et . )CH3 + N3

180-200a 2 3

The semicarbazone can also be used. In this case, the first
step is decomposition to the hydrazone, which then reacts as above
If there is insufficient heating, the reaction may stop at the
hydrazone stage.

)C=NNHC0NH3 -* >C=NNH2 + C02 + NH3

Often it is not necessary to isolate the hydrazone. The

compound can be heated at 180-200° with sodium ethoxide and a.

slight excess of hydrazine hydrate in absolute alcohol, and the
reduction product is obtained, in one step.

Because of the sealed tube feature of the reaction, only
small batches of a compound can be reduced, usually about a gram
at a time. Bigger batches can be reduced if a bomb is available.
Ruzicka recommends the use of sodium benzylate instead of sodium
ethylate. Benzyl alcohol has a high enough boiling point (205°)
so that the reaction can be run in an open vessel. A commoner
means of doing the reaction at ordinary pressure is due to Kish-
ner. He found that heating a hydrazone with powdered potassium
hydroxide accomplishes the desired result. One refluxes the
carbonyl compound with hydrazine hydrate in alcohol, distils off
alcohol, water, and excess hydrazine, and adds powdered potassium
hydroxide to the residual oil. The mixture is then heated until
nitrogen is evolved, usually at 160-180°.

The reaction gives 40-90^ yields with all ordinary types
of aldehydes and ketones, including those which also contain
hydroxyl groups, carboxyl groups, or double bonds. The product
is readily purified. The chief side reaction is azine formation
caused by water being present.

>C=NNH3 + HOH -» N2H4 + >C=0 'C=NNH2> >C=N-N=C<'

The Clemmensen Reduction (5,6,7)

In this reaction, the compound to be reduced is treated with
zinc amelg&m and hydrochloric acid. The oldest method is merely
to reflux the compound six or eight hours. It is often well to

-2-

dissolve the compound in toluene, particularly if it is a com-
pound melting above the boiling point of the hydrochloric acid
solution or if it is very insoluble. This has the advantage
that since the compound is in the toluene layer, well removed
from the zinc, little if any of the resinous by-product often
accompanying the older method forms* Another modification employs
alcohol, dioxan, pcetic acid, or other solvent miscible with
the hydrochloric acid solution in order to form a homogeneous
reaction mixture.

An advantage of the Clemmensen reduction is its extreme
simplicity. The amalgam is preprred by shaking the zinc a few
minutes with a 5-10# mercuric chloride solution containing a
little hydrochloric acid. This polution is poured off, 20-40^
hydrochloric acid and the compound (with or without toluene) are
added, and the whole is refluxed. No stirring is necessary
because the hydrogen evolved provides good agitation. The amounts
of zinc end acid are unimportant as long as there is an excess
of each.

The reaction has not been used with much success on aldehydes
except for a few phenolic ones. Low molecular weight aliphatic
ketones go fairly well, higher ones with difficulty. Many ali-
cyclic compounds, including various sterols, have been reduced
successfully. Aromatic-aliphatic ketones, including the important
0-aroylpropionic acids, are usually reduced in good yield. Aro-
matic ketones are undependable , some reacting smoothly and others
not at all.

Most other functional groups are safe from attack. Excep-
tions are: the hydroxy groups of a-hydroxy acids and benzyl
alcohols; halogens alpha to a carbonyl or carboxyl group; and
the double bonds in pyrroles, isoquinolines, and a~P unsaturated
ketones or acids.

Catalytic Hydrogenation (8,9,10)

This is a relatively recent method which has not had much
practical application as yet, but it seems destined to grow in
importance as time goes on. Its use is limited to carbonyl
groups directly attached to an aromatic ring. The compound is
hydrogenated at 1 50-800° under a pressure of 2000-3000 pounds
over a copper chromite catalyst. Yields are usually 80^ or
better and en easily purified product results. Furan nuclei are
completely reduced. Pyrrole and benjjene rings are not ordinarily
affected.

Fieser found that keto acids were best reduced as an aqueous
solution of the sodium salt. In other cases alcohol or no sol-
vent at all was used.

Low pressure hydrogenation over a platinum catalyst sometimes
reduces a ketone to a hydrocarbon in low yield, the main product
being a carbinol.

-3-

Soci i urn end Alcohol (11,12)

This is an excellent method for reducing diary 1 ketones to
the corresponding diary 1 methanes. Other ketones usually stop
at the carbinol stage, as, indeed, do a few diaryl ketones. The
sodium reduces even hindered ketones not susceptible to the Clem-
mensen or Wolf f -Kishner methods. The yields ere usually good,
but are sometimes low due to cleavage of the ketone.

One merely drops chunks of sodium into a boiling ethyl or
e.myl alcohol solution of the ketone. The hot reaction mixture
is poured into water and the product is filtered or extracted
with ether.

Zinc and_ Alkali ( 6 )

This method often works veil on keto acids and enthrones.
The zinc is activated by shaking it with a copper sulfate solu-
tion. When the blue color disappears, this solution is poured
off and a sodium hydroxide or ammonium hydroxide solution of
the compound is then refluxed 12 to 48 hours. High melting com-
pounds insoluble in alkali are dissolved in toluene.

Hydriodic Acid end Red. Phosphorus

Sometimes ketones which are hard to reduce to the methane
are readily converted to the carbinol. The carbinol can then
be reduced by means of hydriodic acid and red phosphorus. The
carbinol may be changed to the halide first.

A method occasionally used as a last resort is to convert the
carbonyl group to a 1,1 dichloride with phosphorus pentachloride
and then reduce the product with the above reagent. An even more
desperate measure involves the ketone, hydriodic acid, red phos-
phorus, and a sealed tube at 200°.

Bibliography:

1. Suter and Weston, J. Am. Chem. Soc, 61, 232 (1939).

2. Dutcher and Wintersteiner, ibid. , 61, 1992 (1939).

3. Ruzicke and Goldberg, Helv. Chim. Acta, 18, 668 (1935).

4. Schmidt, Hopp, and Schoeller, Ber., 72B, 1893 (1939).

5. Fieser, "Experiments in Organic Chemistry," D. C. Heath and
Co., New York City ( 1941), "Vol. II, p 420.

6. Martin, J. Am. Chem. Soc, 58, 1438 (1936).

7. Mike ska, Smith, end Lieber, J. Org. Chem., 2, 499 (1938).

8. Adkins, "Reactions of Hydrogen," University of Wisconsin
Press, Madison, Wisconsin (1937), pp. 51, 69, 129.

9. Fieser and Heymann, J. Am. Chem. Soc, 63, 2333 (1941 ).

10. Packendorff, Ber., 67, 905 (1934).

11. Klages and Allendorff, ibid. . 31, 993 (1898).

12. Asahina, ibid.., 69B, 1643 (19367.

Reported by B. C. McKusick
March 25, 1942

•••

THE REPLACEMENT OF ALKYL GROUPS DURING NITRATION

Y*it

Halogens, methoxyl, carboxyl, sulfonic acid, and alkyl
groups are all known to migrate or to be replaced by the nitro
group during nitration in certain cases. The removal of an

replacements have been observed.

Although the methyl groups of toluene and of the xylenes
have been replaced by the nitro group in electrolytic nitrations,
such replacement is not the rule when more usual methods are
used.

The trimethylbenzenes have never been known to lose a
methyl group. If 5- or 6-bromopseudocumene is added to fuming
nitric acid and concentrated sulfuric acid and heated, a reaction
proceeds as follows:

CH-

CH3

OsNy/NrCH.

CH3

11no;

0SN<//A\CH3ON03

CH.

NO;

The 3-bromopseudocumene does not give this result.

Continuing on to the tetramethylbenzenes, durene is not
demethylated on nitration, but a nitric acid ester is obtained.
If bromodurene is nitrated with fuming nitric acid or a mixture
of nitric and sulfuric acids, not only cm the nitric acid
ester be formed, but on standing at room temperature in the
presence of the acid, ?~bromo~5, 6-dinitropsuedocumene is
formed.

H,C

CH.

HNO.

h3cY

0°

H*C

NO.

Uroom temp.v |

CH30N03 h3S04 H,C

CH3
N03

In connection with the effect of halide substitution on
the ease of replacement of alkyl groups, Qvist has found that
the more chlorine substituted on the ring of p-cymene, the
easier the replacement of the isopropyl group by tha nitro
group .

Pentamethylbenzene and pentamethylbromobenzene act in
an an analogous manner. The latter can have two methyl groups

■

. !

-2~

replaced by nitration with sulfuric and fuming nitric acids
in chloroform at 0-5°.

In those cases where dinitro compounds were produced by
the elimination of one or two methyl groups, the ortho compound
is formed to the exclusion of the para . In the case of hexa-
and pentaethylbenzene , the para dinitro compound is formed to
the exclusion of the ortho . The best method of preparing the
p_-dinitro, p_-diamino, or p_-quinone of tetraethylbenzene is
through nitration of pentaethylbenzene rather than through
nitration of 1,?,4, 5-~tetraethylbenzene.

The elimination of the isopropyl group was mentioned above.
The latest reaction reported is the f ollowlng :

HN0,-H,S04v

CH,

NO.

82$ +

CH(CH3)

3 /2

CK(CH3)S

CH3-CH0H-CH;
CH,-C0-CH,

Barbier has done work on the elimination of butyl groups.
He found that only groups meta to the first entering nitro
group, were replaced. This is well shown by two examples from
his work.

HaC

H9(iso )
C0CH3

£R3 92# HNO.

0°

C4H9(isp_)

H3cr Noch,

CH3C0»

C4Hg

OaN/^SCOCH.

H,C

N03

CH.

A°8

H3c/^fOCH.

\ywOg

CH^CO

Cihg

HaC

N02

N0S
CH3

C,H9

H3CY >0CH3

\/Jno3

0aN

Amyl and heptyl substituent groups have also been observed
to be affected by nitration of the benzene ring. In the latter
compound, the heptyl group was eliminated but a nitro group did
not take its place

(CoH5 )aG

HNO 3
HaS04

0aN

-3-

MECHANISMS

To date five mechanisms have been advanced to account for
different instances of this type of replacement.

The first mechanism suggested was a Jacob-sen Rearrangement
mechanism.

/4L

H3C
H3C

CH-

HMO 3 -Eg SO/

CK3 0° 30 mm CH
CHC13

CH3

•*>

Ch
CH3

CH.

3 ^8

— >

03N
03N

ch3

CH3
*CH3

The yield in this particular nitration is 70$ of the
theoretical based on p en tame thy 1 benzene. Such a yield would
require that more than the prehnitene itself be nitrated.
Therefore, a mechanism would still be required to explain the
conversion of hexamethyl benzene to dinitroprehnitene. Further-
more, the low temperature and short time required for the reaction
would permit only a negligible Jacobsen Rearrangement to take
place.

Another mechanism suggested is the one by Alfthan proposing
that the alkyl group is first oxidized to a carboxyl and that
this carboxyl is then replaced by the nitro group.

A X3

CH(CH3)3

COCH.

CH.

COOK

•0

N0:

co;

Alfthan was unable, however, to nitrate methyl~p_~tolyl
ketone to the nitrotoluene. It has been shown that in the case
of polymethyl carboxylic acids, the carboxyl group can be
almost quantitatively replaced. The objection to this mechanism
lies in the fact that nitro replacement of a methyl group takes
place in polymethyl compounds even under conditions which do
not favor oxidation of the methyl group.

Barbier has advanced a mechanism which is in agreement
with results of a large number of experiments. From his own
work and from work on the elimination of halogens, carboxyl,
alkoxy, and alkyl groups during nitration, Barbier suggested
that elimination mpj take place when the alkyl group is met a to
a nitro group which has entered the ring. This is in agreement
with such results es the following.

CH3

CH(CH3)

OpN

3 J3

CH(CH3)

3 /S

OH(CHa)

3 / 3

OpM

£H3

NO 3
CH(CH3)3

CH3
rSNOa

NO.

0H3
OgNf V]N03

OH
CH(CH3)3

OaN

N03

CH*

CH3
N0S

In certain cases, however, this mechanism fails to account
for the products of a reaction, e.g,, the nitration of p-cymene
to give p_-nitrotoluene and the nitration of chloro-substituted
p_-cymenes which have no free positions meta to the isopropyl
group. In the case of polymethyl and of polyethylbenzenes,
the mechanism does not explain the facts at all. Pentamethyl-
benzene gives only tne ortho dinitro compound, and pentaethyl
benzene gives only tne para dinitro compound.

From their work with polymethyl and
L. I. Smith and coworkers have also adva
propose that the first step is the nitra
if a position is open. After this nitra
if no position was open), another nitric
to form a benzoyl alcohol ester of nitri
then decomposes to give the nitrobenzene
reaction are illustrated in the nltratio
above., This particular nitration consti
that the reaction c°n take place in tne
Another similar nitration in which the i
were isolated is the nitration of pentae

polyethylbenzenes,
need a mechanism. They
tion of the benzene ring
tion (or the first step

acid molecule reacts
c acid „ This ester
B The step 3 of the
n of bromodurene shown
tutes one of the proofs
steps postulated^
ntermediate products
thylbromobenzene .

-5~

Br
H3Cp |CH

H3C\XCH3

CH3

3 fuming HN03 vCHa
CKC1

alcohols obtained
+ by hydrolysis of
these eaters

inseparable

nitro

compounds

Sn + KCi' GK

CH3
identified

This mechanism obviously does not explain such orientation
of reaction as shown in connection with Barbier's work.

The latest mechanism proposed suggests that the replacement
occurs as a hydrolysis of the a.lkyl group from the benzene ring.
The proponents of this mechanism state as their only support
that p_"Cymene on treatment with nitric and sulfuric acids at zero
degrees gives about Q% of p_-nitrotoluene. Isopropyl alcohol
can be recovered as well as acetone.

None of the above mechanisms by itself can account for all
observed results. One mechanism, however, does not necessarily
exclude all others and. it is possible that a more complete
mechanism may be formulated by incorporating elements from the
different mechanism so far proposed.

Bibliography

Alfthan, Ber . , 55, 78 (1920)

Barbier, Helv. Chim. Acta, 11, 152, 157 (1928)

Doumani and Kobe, Ind. Eng. Chem., 31, 257 (1939); J. Org. Chem.,

7, 1 (1942)
Gottschalk, Ber., 20, 3286 (1887)
Huender, Rec . trav. chim., 3_4, 9 (1915)

Ovist, Acta Acad Aboensis Math, et Phys. 6, sec 11 (1932)
Rinkee, Rec. trav. chim., 57, 1405 (1939), 58, 218, 533 (1939)
Smith, et al . , J. Am. Chem. Soc, 57, 1289 (1935); 59, 1082 (l937)j

57, 1293 (1935); 62, 1349 (19407; 62, 2635 (l9407.

Reported by R.
April 1, 1942

I. Meltzer

■ -

■

KETENE ACETALS

McElvain et al, University of Wisconsin

The most satisfactory method for preparing ketene
diethylacetal is:

, CH33rCH(0Et)3 + KOC(Ch3 )3 ..^^ L±9-°*)°l) CK3=C(0Et )3 + HBr

This ketene adds H30, EtOH, RCOOh, HX, CsHsOH, CH3COGH3COsEt ,
CH3(C03Et)3) NH3; C6H5NH3, CBH10NH, and H3 across the double
bond. In the presence of Cct013 aimers, trimers, and polymers
have been produced and a head- to- tail structure has been indi-
cated. These properties were discussed in this seminar last
year. This report is concerned with the addition of active
halogen compounds, acids, and a, p-unsaturoted carbonyls.

Many of the additive reactions can be interpreted on the
basis of the polarization that is characteristic of the hetero-
enoid structure of ketene diethylacetal:

^ f°

CH2kCV,

xEt

1« Addition of active bromides. --The reaction with active
bromides can be summarized as follows:

R = allyl or benzyl

CH3=C(OEt)3 + RBr -* [RCH3C(OEt )3Br] ->RCH3C03ET + EtBr

RCH=C(OEt)3 + HBr
(not isolated)

\RBr

[CH3CBr(OEt)3] [R3CHC3r (OEt )3]
CH3Cv53Et + HBr R3CHC03Et + EtBr

RCH=C(OEt)3 + EtOH -> RCH3C(OEt)3

Origin of the alcohol used in the last reaction is poly-
merization of some of the ketene.

2. Addition of CHaCOCl. — The products isolated indicate
the following course of the reaction:

CH3=C(OEt)3 * CH3C0C1 -^-4 CH3COCH3C03Et + EtCl

(b)//CH3COCl

CH3C(OCOCH3)=CHC03Et + HCl

-2-

The hydrochloric acid formed by reaction (b) appears to
react in two ways, (c) end (d)n Notice- that reaction (d)
involves s 1,4 addition ecronc two molecules of the acetal.
This unusual reaction is substantiated by oxher additions.

(c )

CH3=C(OEt)3 + hCl -^-MGH3C03St + EtCl

CH3
(EtO)3C **KC1 (a) v ,

CH3 it

^^C(0Et)3

CH3C(OEt)=CHC03Et + EtCl + EtOH

The validity of reaction (b) is indicated by the fact that
acetoacetic ester is acetylated by acetyl chloride with either
pyridine or ketene diethylacstal as reagents for the removal
of KC1.

3. Addition of Acids. — 1,4 addition of acids across two
molecules is further indicated by similar addition to pure
ketene diethylacetal by nine acids as, for example, hydrochloric,
hydrobrornic , trichloroacetic, formic, and acetic in yields of
10-38^, A minimum of acidity seems to be required for this

type of addition since phenol and p_-bromophenol added only as
in (c) above but tribromophenol gave a 26f yield of the 1,4
product.

4, Addition to a, ^-unsaturated carbonyl compounds. — The
ad.dition of maleic anhydride can also be considered as a 1,4
intermolecular addition in which the establishment of the 2,3
double bond of the ordinary type is replaced by a single bond
that joins the two molecules of the ketene involved. The work
with maleic anhydride is summarized on the flow sheet at the
end of the abstract. Because the addition product, II, is
insoluble in ether, only one mole of the anhydride adds in tnis
solvent. In benzene II will add a second mole of maleic anhy-
dride to give III. Chief support for the structure of II is
its dehydrogena tion to the known compound, 3, 5-diethoxyphthalic
anhydride, IV, and hydrolysis of IV to the corresponding
phthalic acid. Heating either the anhydride or the acid with
aniline gave the corresponding N-phenylphthalimide . Since
reaction (b) involves the loss of an a-hydrogen, possible addi-
tion to dimethyl maleic anhydride was studied. No addition
product was obtained.

Addition to quinone and proof of structure was as follows:

'**''•'

•3-

( + 2CHs=C(0Et);
>

pOH3 + CK3C(OEt)3

WCOSt

Colorless

H20
fa

Ester ^MhL. ||

GKpCOpH A

0
K

It is interesting to notice that I, a colorless compound,

can be considered as a vinylog of an ester while quinone,

from which it is derived, is vinylogous to the colored
1, 2-diketones.

The addition to dibenzalacetophenone and proof of struc-
ture was as follows:

C6H5CH— €H~r-C0CK=CEC6H5

II \

CH-a-C(OEt)a \

\HC1
Br2 \Ha0

C6H5CH-CHBrCOCH=CHC6H5 C6H5CH-CH3COCH=CHC6H5
CH3COsEt + EtBr CHsC03K + 2EtOH

Benzalacetophenone added in a similar manner.

Bibliography:

McElvain and Kundiger, J. Am. Chem. Soc, 64, 254 (1942).
McElvain and Cohen, J. Am. Chem. Soc . , 64, 260 (1942).

Reported by G-. L- Schertz
April 1, 1942

OH,

(EtO)aC

-4-

FLOW SHEET

CH— CO

|| p Ether Solution ,

CH-C6 (a)

Benzene

maleic
,,pnnydride
Eenzene

EtOC NCHCO=,H
II i

III 60^

H\yCH2

35(

EtOC7 XHC03H

II I

HC XJHCOgK

(Et0)3C XCH~CO
T)
HaG ^CH-Ctf
XCtOEt)2
I

(tfl

EtOC xCH~Q0

|i j ^0 + 2EtOH

C Jl—CV

xctm
\

715? \

HpO

/CH2
EtOC' ^CKCOoH

II I

HC ^C-C03H

OEt

CH3=C(0Et)2

CH3C(0Et)3

Et0/^\\C0

0
/Et IV

HpO

36H5NH3
Imide

Acid

BIPHENYLENE

Investigation of the interesting hydrocarbon biphenylene
(I) has been carried out because it might contribute to the
theory of aromatic chemistry as a whole, supplying in its central
ring a possible cyclobutadiene system and at the same time
affording an extreme case of the Mills-Nixon effect.

JK.

(I)

The history of its preparation is a series of repeated
failures and of only one isolated e nd iireproducible success.
Hosaeus, in. 189^, carried out a Wurtz reaction with o_-dibromo-
benzene, obtaining biphenyl as the. only hydrocarbone.
Niementowski , in 1901, on treating the diazonium salt of 2,2'-
diaminobiphenyl with copper powder obtained carbazole. An
attempted dehydration of 2~hydroxybiphenyl by Cullinane, Morgan,
and Plummer also failed.

Dobbie, Fox, and Gauge in 1911 rep
of the hydrocarbon in 1007 yield by the
freshly cut sodium on 2,2' -dlbromobiphe

based their proof of

method of synthesis,

and on its analysis.

Indicated a. strained

the ring to give the

derivative of the hydrocarbon. The

the hydrocarbon gave dibenzofuran,

structure of the c
on its oxidation i
Reactions carried
central ring. Tre
starting material

ac
di

derivative, Attempts by other investig
of Dobbie, Fox, and Gauge failed oomple
of a compound containing a cyclobutadie
able.

orted the preparation

prolonged action of
nyl in dry ether. They
ompound C13HQ on the
n part to phthalic acid,

out on the compound
atrnent with bromine opened
as well as a d.ibromo
tion of nitric acid on
nitro, and a tetranitro
a. tors to repea.t the work
tely so that the existence
ne ring became question-

Lothrop has recently investigated biphenylene in connection
with his work on the Mills-Nixon effect in aromatic compounds.
His attempts to eliminate the bromine atoms from 2,2 '-dibromo-
biphenyl by the action of hydrogen, lithium, sodium, and
potassium were all successful, but the product in each case
was biphenyl and a bromine containing oil. Magnesium reacted
with only one of the bromine atoms and calcium, zinc, and pure
copper had no effect. Cuprous oxide, however, g:eve a low yield
of a new hydrocarbon which was volatile in steam and formed a
scarlet picrrte. Pyrolysls of biphenylene iodonium iodide with
cuprous oxide gave a larger yi<=ld of the hydrocarbon with small
amounts of 2, 2'-diiodobiphenyl .

ex.

NH3 HaNV/

Picrate"

M,P. 121~122°C.

M.P. 109~110°C.
(I)

Molecular weight determine tione in camphor and benzene
agreed with the formula for biphenylene rather than for
dimoleculrr coupling possibilities. Analyses of the hydro-
carbon end its picrate agreed with the formula C13H8. Oxidation
with chromic oxide ^pve phthalic acid, thus establishing the
presence of ortho substitution. Reduction with hydrogen over
red hot copper gave biphenyl.

CrO

3,--

(I)

COOH

OCH

2 . 7 -Dim ethyl biphenylene (II). — The 2 , 7 ~di me thy 1 derivative of
biphenylene wee prepared in the same manner In order to see
whether the dehalogena tion reaction could be considered general.
This compound was prepared by two different routes, thus veri-
fying the structural formula of biphenylene. This compound
gave a deep crimson picrate.

. 1

-3-

H3C

Piers te
M.P. 110°C.

f—

3C^\NH8 H3C
NO*

<f-

N02 OaN

\'/

H3C^V/>NH3 H2N^^CH

HgC^X

H*C

M.P. 112-113 C.
(II)

/^SGK

NOo 0aN.

— >

VfCH.

/^CH;

The formulas written for these hydrocarbons pre supported
by the complete agreement found between prediction and experi-
ment, and suggest that the cyclobutadiena ring may not be too
strained to exist if fused with two benzene rings (A), or that
the Mills-Nixon effect may operate to stabilize the molecule in
a less strained cyclobutane form (B), or that the coplaner
molecule has considerable resonance energy.

/V_A

%S — \f>

(A)

(B)

/•

-■'.

-4-

The fused system of three rings would seem to be confirmed
by the formation of such stable and highly colored picrates.

2 , 3 ■ 2 ' . 3 ■ •■ Bin/mhthy len s ( III ) . ---The only other compound reported
in the literature having a structure similar to biphenylene is
2, 3,2 ' , 3' -binaphthylene, prepared bj Rosenhauer, Bre.un, Pummerer,
and Riegelbauer in connection with polymerize bi on studies.
1,4-Nephthoquinone was condensed to form 2, 3, 2* ; 3 ! --binaphthylene-
1, 4,1 ' , 4 ' -diquinone which, on distillation with zinc dust, gave
2*5,2' i 3 l -binaphthylene in 15.5$ yield. This compound formed
a n un s t a bl e r o d p i c r a t e .

CflHsNO.

Py. + HOAc

Distil
Zinc dust

Picrate
M.P, > 260°C

Bibliography :

,/vv

M.P. 365°C
(III)

Hosaeus, Monatsh,, 14, 323 (1B93).

Niementowski, Ber . , 34, 3325 (1901).

Dobbie, Fox, and Gauge, J. Chem. Soc . , 99, 693, 1615 (1911);

105, 36 (1913).
Mills and Nixon, J. Chem. Soc, 2510 (1930).
Schwechten, Ber., 65, 1605 (1932).
Mascarelli and Getti, Gazz. chim. ital,, 63, 654,
Cullinane, Morgan.., and Plummer, Rec . trav, chim.,
Rosenhauer, Braun, Pummerer, and Riegelbauer, Ber
Lothrop, J. Am. Chem. Soc, 63, 1187 '( 1941).
Hammett, "Physical Organic Chemistry", McGraw-Hill Book Co., Inc.,
New York, N. Y., 1940, p. 19

661 (1933).
56, 627 (1937).
, 70, 2281 (1937).

Reported by F. W. Spangler
April 8, 1942

THE CONSTITUTION OF USNIC ACID

Usnic acid is a very widely distributed naturally occurring
compound which has been found in more than ?0 lichens. Usnic
acid was isolated first by Rochleder and Heldt in 1843 and has
been the subject of much investigation since that time. Widman
in 1902 first advanced a formula for usnic acid (I), based
on the observations of several earlier workers and on the re-
sults of his own degradation experiments. In 1927 Shopf, Heuck
and Kraus turned up several new experimental facts which made
it obvious that Widman :s interpretation of his data was erron-
eous.

CO — 0

C03H
CH3COC=C-C=CCH--C8H11

0 C

■CO

C03H
CH3COCH=C~C=CCK-CeH 1 %

0 — to

I II

Kraus demonstrated by a Zerewitinoff determination that
usnic acid had three active hydrogens. Formula I would indicate
only two if we assume complete enolization of the methyl Ketone
group. Moreover, one of the characteristic degradation products,
decarbousnic acid, represented by Widman as II, was demonstrated
by Shopf and Heuck to be an acidic enol, not a carboxylic acid.

The structure of usnic acid is most readily elucidated
from its degradation products which are quite numerous. The
more important are set forth in Chart I for future reference.

The decomposition through usnetic acid to usnetol has
been of greatest aid in elucidating the nuclear structure of
usnic acid and its derivatives. The structure of usnetol
follows from the following degradations and synthesis.

Ci3Hi404 O3
usnetol

~CH3C03H
H30 (B)

>yCnHiaOB + CH3C02H
Alcoholic (Ident. as

^HCl silver salt)

%0 1 C13H1406

4 N'-NaOH

y\°H

CijH.,0, — >di^cetate
Uen&dl

0^ JLsOs

^NaNOa ^

OpN

CH3^H3S04 *(h)

N03

CHa

NO-

v I

■

[ > ;

■

■ ■

- ■

■ '■

<
XX
O

•H

CCS

O rH

■4

•H

> -* O

H +3

JH CD

•

•4

m c:

•

H CO

O. 3

£5

/IS

ffl

o o

m oa

ffi

O Ov

o w

o «

ntrj

to O

w +

w o

O 1

-d

^ O

O co

CD
m o

co a

^ -sh 0.?
I I •
tO W OJ
Oi O CO
rH CV2 <#
+ 1

Ph Ph II

• a

r<-i »2h I — I
I 1

• • «

O -Pi— J
CO o
SU CO'

•H

-V

lO-H

O £

W CO

CO rH

H 2

... •-• O

ffl o

• —

_ j _

~*XX CD

•

ro *h

■-« a

PL,

* t>.

H CO

•

o p

O 3

is*

w
O

p
Pi
o

o
en

o
c
o
o

o
m to

+ CD

•H

iO CO

• 1 o

to to O

r-t O ^

Cv> C\2 CV

•H

C/J

•

•H
3 rH

o
1

O O

•

•H

CD 3

£N

O O

■n Q

h *h

_ o

iX co

•

^ o

•H

ffl

H CO

•

o -d

in 3

o O

C0rQ

Cvi

•d

* fc

K co

•

t^ o

»h a>

P ^

•d

•H

CO o

O co

■"• O

CO.

tX -H

CO rj

■"• O

•

o c

•

*>'

•

• ; :■

■1

-2-

87'

The isolation of the product C13H140€ after decomposition
of the ozonide indicates a double bond in a ring system since
no carbon atoms were lost. The ease with which this compound
is hydrolyzed by alcoholic hydrochloric acid indicates an oxygen
containing five or six memebered ring with one double bond. The
alternative route of degradation (B) indicates one acetyl group,
not present as an acetate. Structures III, IV or V would satisfy
all of the above facts. The isolation of acetic acid and the
compound G11¥{130B by acid hydrolysis of the ozonolysis product
rules out an isomeric structure with an ethyl group and a hydro-
gen at C-2 and C-3 instead of two methyls as shown*

CHCH

COCH.

The correct structure (ill) was established by the follow-
ing synthesis of Curd and Robertson after they had established
the structure of usneol as being III minus the acetyl grouo
at C-7.

CH.

CH,0^

CH3

OH CHsO-^Vq
NCCH3

COCH.

CI9HCH3 HO
OCCH3

KpCO, CH3

COCH3

CH3I

CH.

.3CO3

CH3
*CH3

dimethyl ether of III
(no depression on mixed

M.P.)

The establishment of the orientation of these substituted
phloroglucinols is due to work by Robertson et si.

Usnetic acid must then have the same nuclear structure as
usnetol with the addition of C03 to make the carboxyl group
shown to be prepent. Because of the synthetic difficulties,
usnetic c:ciu itself was not synthesized but the structure of
pyrousruc «cid dimethyl ether (VI} arising from usnetic acid by
loss of the nuclear acetyl group at C-7 has been established
as follows.

ch3o

CH3

r*\

CH3
CO

M^Ls^

0CH3

CH.O

CH.

-3-

OH c^30^^oCH3C03Et

COl3!!;"~>CH3iv

COCH.

OGB

OCH.

CHaO^y^NCHO
CH3^ '

oc:

CH30

^v~^

^K/-

CH30f*

CH.

Gatterman

CH3

^CHO

CH*

OCH;

OCH,

hippuricy
acid

• - 0

CH.

OCH.

10^ ^H oAA

NaOH h3°

■»

CH.

CH3COC03Et
*CH3

CHaOr^
This structure for usnetic acid (VII)
has been confirmed by Asahina by the CH3vv

following degradation.

CH2C02H

KOH y mzo<

CH3 Ha0

3^3

HOpC

CH3C03H
*CH3

CH3N3 CBHxlONO
NaOEt

-4-

C03CH3

CHg

JkO^

acet

HOgC^

!i I'

anhyd, KOECj liCH.

/ \C02H

CH3

The structure of decarbousnic acid, has not been established
by synthesis but there can be little doubt as to its structure .
The following most pertinent facte must be considered in formu-
la ting a structure for decarbousnic acid.

(1)

(2)
(3)

(4)

(5)

Ketonic reagents indicate a 1,3 diketone structure.

a. Hydro xylamlne gives an ieoxazole derivative.

b. Phenylhydrazine gives a pyrazole derivative.
c „ Semicarbazide gives ? pyrazole derivative.
Mono--d.i-tri-tetraacetyl derivatives are known,
Decarbousnic acid is optically inactive.
Decarbousnic acid is a dibasic acid containing no
GC2K group.

Ozonolysis of dlacetyl decarbousnic acid shows that
decarbousnic 'acid has the characteristic substituted
benzofurane nucleus.

^nd

The only formula which seems to fit all
mass of minor experimental facts is VI

these criterea
Experimental

evidence for- the side chain structure will be given
with the ozonolysis of diacetyl usnic acid.

in connection

COCH

•Ha0

cone .

K3S04
60°

! I I

3 X/ CK3 SCX

CK.

TV
1ft

The cyclization of decarbousnic acid VIII to decarbousnol IX
(postulated) was without precedent until Foster and Robertson
(and Kealy) showed that this type of reaction did occur and that
the compounds so obtained were entirely analogous to decarbousnol
and usnolic acid, of similrr structure..

The structure of usnic acid (X) itself is fairly well
established without presupposing the structure of decarbousnic
acid and the evidence thus obtained supports the proposed, formula
for decarbousnic acid. Shopf and Ross performed the following
degradation.

-5-

COCH

l , .CHGOCHg

Oa on

diacetpte

EtOH

CH3C02

CH.

COCHL

A/-

OCOEt
\o 4. P°

„C.H3 CH3COCH3

co3

XII

The isolation of the lpctone XI pnd of acetoneoxalic eeter XII
clearly indicates the presence of the double bond as shown and
consequently of the benzodihydrof uren nucleus, end the 1,3
diketpnic nature of the side chain. The fact that all the
degradation products of usnic acid have been demonstrated to
be substituted ben^ofurans does not constitute a serious draw-
back since the reversion to the more stable conjugated form is
to be logically expected when possible.

v\A

*p°

XIII

The nature of the degradation products
indicates the partial" structure XIII of
usnic acid which must be considered as
3CHCOC_Ha derived from XIII by the replacement of

2H by the -CO- group. One of these H atoms
must be from C6 in XIII in order to stabil-
ize the dihydrocoumarin ring system. There
are then four possibilities left; a) By
enolization and removal of the H at C2 a
sixmembered cyclic lactone results,
process at C4 an 8 membered cyclic lactone
removal of an H at C3 a 6 membered cyclic trike-
tone {X) results, d) by the same process at C5 an 8 membered
cyclic triketone results. Both lactone formulas are readily
ruled out since the 1,3 diketonic system characteristic of usnic
acid as well as decarbousnic acid would not be present. Of the
remaining two the choice of the 6 membered triketone is pre-
ferred on two grounds, first the more ready occurence of 6
membered rings, but this choice is even more strongly indicated
since 1,3, diketones in a ring structure do not show the char-
acteristic behaviour of straight chain 1,3 diketones with
ketonic reagents (i.e. dihydroresorcinol ).

b) by the same
results, c) by
(X)

-6-

The structure advanced has some de
necessary to assume that the linkage be
is rather weak since in all the degrada
linkage is broken. This does not seem
assumption since the linkage under ques
hindered quaternary C atom which is und
of several active ting groups-. The only
out explanation is the easy recemizeabi
again may be considered as a. property o
Cs-Cilj. linkage but no adequate analogy

batable features. It is
tween G3 and Cu in X
tions of usnic acid this
to be an unjustifiable
tion is to a highly
er the combined influence

serious objection with-
lity of usnic acid.. This
f the weakness of the
is known.

Two

other compounds in Chart I which have had only
"nuisance value" in the elucidation of the structure of
acid are usnolic acid XIV and usnonic acid XV a or b
postulated only ) .

usnic
( structures

COCH

COGH

C-COCH,

COCH3

CHCOCH.

0 C

.CH~
CH^CT COCH

XIV

XVa

XVb

Bibliography

Recent literature only. For earlier references see Schopf and
Heuck, Ann. 459, 233 (1927)

Schopf, Clemens Institut fur organische Chemie der Techn.
Hochschule in Darmstadt

Schopf and Heuck (A. Kraus), Ann., 459, 233 (1927)
Schopf and Ross, Natur., 26, 772 (1938)
Schopf and Ross, Ann., 546, 1 (1940-41)

Robertson, U. of Liverpool

Curd and Robertson, J. Chem. Soc . , ,1953, 437, 714,1173

Birch, Flynn and Robertson, ibid. , 1936, 1834

Curd and Robertson, ibid;, ,1937, 894

Birch and Robertson, ibid. , 1958, 306

Foster and Robertson (Healy) ibid. . 1959, 1594

Asahina, Yasuhika; Pharmaceutical Institute, U. of fokio

Asahina, Miyasaka and Sekisawa, Ber., 693, 1643 (1936)

Asahina and Yanagita, ibid. , 70, 66, 1500 (lgs17)

Asahina, Yanagita, and Mayede, ibid. , 70, 2462 (1937)

Asahina and Yanagita, ibid, , 71. 2260 71933.-: 72, 1140 (1959)

Yanagita, ibid., 71, 2269 (1933)

Proc. Imp. Acad. TTokio ) 13, 270 (1937); 15, 311 (1939)

Perkin and Everest, Natural Organic Colouring Matters, (1938) p. 530
Reported by C. W. Theobald, April 8, 1942

•

ALKYL CARBONATES IN SYNTHETIC CHEMISTRY
Wallingford, Homeyer, and Jones — Mallinckrodt Chemical Works

Until very recently the alkyl carbonates have found little
application in synthetic organic chemistry. Within the last
year a series of investigations have been carried out which
have greatly extended the uses of this class of compounds and
have made easily available certain types of compounds which
before were difficult if not impossible to prepare,

!• Synthesis of Malonic Esters.

The first attempt to condense alkyl carbonates with esters
was made by Wielicenus in 1887; he was unsuccessful, and con-
cluded that alkyl carbonates were not suitable for Claisen type
condensations. However, since then, several investigators
(Lux, Nelson and Cretcher, Skinner) have succeeded in carrying
out such condensations by use of sodium or potassium in benzene
or ether solution, producing malonic esters in low yield.
Wallingford and his associates point out that the difficulties
encountered were due to the following factors: (l) The ester
may condense with itself, (2) The metallic sodium and potassium
used readily decompose alkyl carbonates, (?) A metal elcoholate
in alcohol solution may not be used as a condensing agent because
it ha s been shown by Cope and McElvein and by Connor that sub-
stituted malonic esters undergo alcoholysis according to the
equation

RCH(COOEt)2 + EtOH NR^Bt RCH2COO£t + (EtO)2CO

The Mallinckrodt workers have demonstrated that this reaction
may be reversed. By applying the principles of mass action, they
have developed a process which is quite general, and have pre-
pared a large number of malonic esters by condensing alkyl car-
bonates with a variety of esters. The general reaction is

R'CH2COOR + (RO)3CO + MOR -> [Rf C (COOR)3]M + 2R0H

The essentials of the new procedure are (l) the use of a large
excess of alkyl carbonate as a reaction medium or solvent and
(2) removal of alcohol from the reaction mixture by distillation.
The validity of the applied principles of moss action is shown
by the fact tha t an 86% yield of diethyl phenylaceta te was
obtained from ethyl phenylacetate and diethyl carbonate whereas
Nelson and Cretcher have shown that these two will not condense
at all in an alcoholic solution of NaOEt.

It was noted in the carbalkoxyla tion of some of the lower
aliphatic esters that an appreciable amount of alkylation
occurred. For example, from ethyl n-butyrate and diethyl car-
bonate there was obtained diethyl diethylmalonate along with the
expected diethyl ethylmalonetp. This observation led to the
work which will b° mentioned later in this report.

Scope and Limitations: It may be seen from Table I that the
cerbalkoxyl group is introduced most readily into aryl substituted
acetic esters; but the reaction also goes well with alipnatic
esters up to ethyl steerate. The a-carbon must contain two
hydrogen atoms. If the £--carbon of the ester is tertiary the
yield is low. The esterified alcohol of the ester . nd of the
carbonate should be the same.

I I ♦ Synthesis of £-Keto Esters .

It would be expected that if alkyl carbonates will condense
with esters that they should condense with ketones, and such is
the case. The general reaction may oe represented as follows:

R'C^CK3R" + (R0)3C0 + MOR -> [R'C^C-R' ' ]M + 2R0K

GOOR

Welling ford °nd his associates have used the same procedure used
in the cerbelkoxyle tion of esters and have prepared a large
number of p-keto esters from various ketones end alkyl carbonates;
a few of these are shown in Table II. Such a general method
of organic synthesis is particularly important since the products
are one of the most reactive classes of organic compounds.

Limitations: Ketones which are so active that they condense
with themselves or polymerize are not satisfactory. Unreactive
ketones such as camphor give low yields. Forcing by heating
?bove lOOo j_s undesirable because of ether formation by the
reaction between elcoholaae and alkyl carbonate at that temperature
^-Acylation takes place with certain ketones (cyclohexanone ,
for exemole ) giving the carbonic ester of the enol form of the
ketone.

III. Synthesis of a-Cyeno Esters.

Alkyl carbonates may be condensed with nitriles to give
the corresponding cc-cyano esters as follows:

R'CH3CN + (R0)2C0 + MOR

[R'C-CN]M + 2R0H
COOR

Although several authors have reported the condensation of
diethyl carbonate with phenylacetonitrile by various procedures,
none indicated thet this reaction was applicable to other types
of nitriles. The Mallinckrodt workers have employed the tech-
nique -uaed with esters and. ketones for nitriles and have pre-
pared a variety of cc-cyano esters, some of which are shown
Table III.

Alcohol introduced with the metal alcoholete and that pro-
duced in the condensation is removed by distillation. Cope and
Hancock have shown that cyeno esters undergo an alcoholysis

-3-

analogous to that of malonic esters, producing an alkyl car-
bonate and a nitrile. Therefore, here again the distillation
of the alcohol favors completion of the reaction according
to the mass law and prevents cleavage of the reaction product.

Scope end Limitations: Vinylecetonitrile and p_-nitrophenyl-
acetonitrile appear to "be too active and result in tars.
cc-Phenylbutyronitrile failed to react. A general statement
may now be made in regard to the use of alkyl carbonates for
carbalkoxylation:

(1) The method is general for esters, ketones, and
nitriles having two alpha hydrogen atoms.

(2) Na and K alcoholates are equally effective as con-
densing agents; Mg and Al alcoholates are not effective.

(.3) Primary alkyl carbonates react satisfactorily; the
use of secondary carbonates is not practical.

(4) Lower members of each series give poor yields of
cerbelkoxylated products either because the low boiling point
prevents forcing the reaction, or because of polymerization.

IV. Alleviation of Malonic Esters by Alkyl Carbonates.

It was noted in the synthesis of ms Ionic esters with alkyl
carbonates that as a side reaction the sodio derivative was
alkylated; this was found to take place in the following manner:

[R'C(COOR)3]M + (RO)3CO -> R»C(COOR)3 + RMC03

This is a new type of alkylation; organic esters, including
alkyl carbonates, are known to act as alkylating agents in the
Friedel-Crafts reaction, but metal enolates have not been
alkylated by esters of organic acids before.

The alkylation may be carried out on the metal derivative
of a malonic ester, or a mono-substituted acetic ester may be
used as starting material since carbalkoxyla tion gives the
malonic derivative. The reaction is carried out by refluxing
at temperatures of 125-175° with a large excess of the alkyl
carbonate.

Scope ant? Limitations: Ethyl, butyl, isobutyl, ieoamyl,
and benzyl groups have been introduced into a variety of esters
(see Table IV). Secondary carbonates are poor alkylating agents.
Malonic ester itself reacts to form tricarbalkoxymethane . If
the first substituent is a s~aliphatic group, alkylation is
poor. Best results are obtained if the alkyl carbonate, metal
alcoholate, mr1 malonic ester all have the same alkoxy group.
Preliminary experiments with substituted a-cyanoacetic esters
and acetoacetic esters gave negative results.

-4- 95

Vi Alkyl Carbonates as Solvents for Metallation and Alkylation.

In the foregoing synthesis it will be noted that the product
is in the form of the metal derivative. Not only is it easy
to alkylate directly by means of the alkyl carbonate solution
itself, but it has now been shown that alkyl carbonates are
superior solvents for alkylation by means of alkyl halides. •

Previously, esters containing a branched substituent such
as the .g_-butyl group have been very difficult to alkylate further
by the usual procedure using sodium alcoholate in alcohol. This
has been ascribed to the incomplete formation of the sodio
derivative according to the equation

R'CH(COOR)2 + NaOR^=*[R'C(COOR)2]Na + ROH

The alkyl helide when added reacts to a large extent with the
NaOR present. Furthermore, as previously noted, the alkylated
malonic ester formed may undergo alcoholysis. Roth of these
conditions may be avoided by using alkyl carbonates as reaction
media. The above reaction may be forced to completion by removing
the alcohol by distillation; the metal derivative may then be
treated with the alkylating agent in alkyl carbonate solution with
no danger of alcoholysis.

The alkylation of malonic, p-keto, and cc-cyano esters were
carried out in alkyl carbonate solution with r.lkyl halides as
the alkylating agent. Excellent yields were obtained as shown
in Table V. It may be seen that ethyl and allyl groups are
readily introduced into diethyl s-butylmalonate whereas this is
practically impossible in alcoholic solution. The method has
also enabled Wallingford to introduce two s.-alkyl groups as
substituents, which has hitherto been impossible.

As would be expected, slight alkylation may take place by
action of the alkyl carbonate, hence it is desirable to use as
reaction medium the alkyl carbonate corresponding to the group
to be introduced.

Bibliography

Wallingford, Homey er, and Jones, J. Am. Chem. Soc . , 63, 2056 (1941 ).

Wallingford, Homey er, and Jones, ibid. . 65, 2252 (1941).

WiFlicenus, Ber., 20, 2930 (1887).

Lux, ibid. , 62, 1827 (1929).

Nelson and Cretcher, J. Am. Chem. Soc, 50, 2758 (1928).

Skinner, ibid., 59, 322 (1937).

Cope and McElvain, ibid., 54, 4319 (1932).

Connor, ibid. , 55., 4597 (1933).

Wallingford, Jones, and Homey er, ibid. , 64, 576 (1942).

Wallingford and Jones, ibid. , 64, 578 (1942).

Wallingford, Thorpe, and Homeyer, ibid., 64, 580 (1942).

Cope and Hancock, ibid. , 61, 96,776 (193977

Reported by R. M. Roberts, April 15, 1942.

•5-

TABLE I.

Starting Material

Product

% Yield

Ethyl acetate

Ethyl butyrate

Ethyl stearate

Ethyl phenylacetate

Ethyl p_-methylphenylacetate

Diethyl raalonate 25

Tricarbethoxymethane 10

Diethyl ethylmalonate 45

Diethyl di ethylmalonate 10

Diethyl cetylmalonate 50

Diethyl phenylraalonate 86
Diethyl p_-methylphenylmalonate 65

TABLE II.

Starting Material

Pro due t

t Yield

Diethyl ketone

Methyl isobutyl ketone

Di-n -propyl ketone

Acetophenone

Dibenzyl ketone

Propiophenone

Cyclohexsnone

Ethyl p-keto-a-methylvalerate 20

Ethyl p-keto-^-methylcaproate 60

Ethyl (3-keto-a-ethylcaproate 45

Ethyl benzoylacetate 60

Ethyl a, Tf -diphenylacetoacetate 45

Ethyl a-benzoylpropionate 37
Ethyl 1-phenyl-l-propen-l-yl

carbonate 25

Ethyl 1-cyclohexen-l-yl carbonate 20

Starting Material

TABLE HI.
Product

% Yield

Acetonitrile
Butyroriitrile
Capronitrile
Stearonitrile
Phenyl acetonitrile

Ethyl cyanoacetate
Ethyl a-cyanobutyrate
Ethyl a-cyanocaproate
Ethyl a-cyanostearate
Ethyl cc-cyanophenylacetate

10
40
54
75
78

TABLE IV.

Starting^ Material

Ethyl butyrate
Diethyl ethylmalonate
Ethyl ^ -methylceproa.te
Diethyl .s-butylmalonate
Dibutyl ethylmalonate
Di butyl cetylmalonate
Fluorene

Product

% Yield

Diethyl diethylmalonate 36
Diethyl diethylmalonate 54
Diethyl ethylisoamylmalonate 45
Diethyl .s-butylethylmalona te Poor
Dibutyl butylethylrnalona te* 42
Dibutyl butylcetylmalona te* 83
Butyl 9-bu tyl-9-fluor en ecar boxy-
late 45

-6-

TABLE V,

Starting Materiel Alky! Kallde Product % Yield

Diethyl ethylmalonate EtI Diethyl di-:-thylmalonate 83

Diethyl ethylmalonate i-AmBr Diethyl ethylisoa.mylmalonate 75

Diethyl e.-butylmalonate EtBr Diethyl .s-butylethylmalonate 95

Diethyl ^.-butylmalone te AllylBr Diethyl e-butylallylmalonate* 86

Diethyl .s-butylmalona te s-3uBr Diethyl di~s.~butylmalonate* 25

* New compound

GLYCEROL DERIVATIVES

The specificity of triglycerides for various animal species
has been shown. Structurally similar animals frequently eilatorati
similar kinds and proportions of fatty acids combined as tri-
glycerides. LiKewise it has been shown that the more complex
fatty acid mixtures are found in fats of the simplest forms of
plant and animal life, and a. gradual simplification is observed
both in structure and number of component acids on ascending the
evolutionary scale of development. With recent developments in
molecular distillation and fractional crystallization it is
reasonable to assume that more work will be done in the near
future in isolation and purification of the triglycerides, and
their structures will have to be checked by syntheses.

Accordingly, some workers have been engaged in fundamental
research on triglycerides. A brief summary of their successes
in synthetic problems constitutes this seminar report.

Synthesis of Simple Triglycerides

No special procedures need be resorted to in preparing
simple triglycerides. Merely heating three moles of the acid
with one mole of glycerol yields the tri-ester in nearly quanti-
tative amounts. The slight oxidation which occurs at such tem-
peratures may be avoided by use of an inert atmosphere such as

2 )

or SO

2 •

CH3OH

CHOH +

CH3OH

3 RCOOH

2 CO'

6om

CH30C0R

CHOCOR

CH2OCOR

Syntheses of Symme t rical Mixed Triglycerides

The preparation of mixed triglycerides cannot be effected
by the above method, i.e., by heating in either order a mole
of glycerol with a mole of one acid and then with two moles of
another acid. Such a reaction yields a variety of triglycerides,
and the usual methods of fractional crystallization and dis-
tillation can not be applied with any degree of success in
attempting to purify any one product.

The following methods have been used:

CH3OH
CHOH

CH3OH

Ha SO 4

CH30S03H

CHOH

CH3OS03H

2 R'COOH

CH3OCOR»

CHOH

CH3OCORf

RC0C1

-?.-

CK20H

CHOH

C6HsCHO

CKaQ

CHOH piC6H5

CH30'

RCOC1

Pyridine

CH30

choco:

CHoO-

HCpH,

CR3OH

CHOC OR i/R'^Cl

CHoOH

CH2OH

CHOH (C6HjJ3CCl

CH3OH

CH2OC(C6H5)3

HOH
H2OC(C6H5)3

RCOC1

R'COCl

CH2OC(C6H5)3

i
CHOCOR

CH3OC(C6H5)

CH3OH

CHOCOR

CH30H

The final product in each of the above three equations is
the same symmetrical triglyceride. The latter method is of
more recent date and appears to be successful for a wider variety
of glycerides.

Syntheses of Unsymmetrical Mixed Triglycerides

The preparation of unsymmetrical mixed triglycerides
involves first, of course, the formation and isolation in the
pure state of either the a-monoglyceride or a, p-diglyceride.

Among the following syntheses, I is only of historical
interest, II, III, and IV represent fairly satisfactory methods
of syntheses, and V*s authenticity has been questioned because
of its similarity to VI, which yields a product whose formation
seems to involve a so-called acyl shift.

H3C1

H2S04

CH3OS03H

CHOSO3H

CH2C1

RCOOH

CH2OCOR

(JjROCOR

CH3C1

R'COOK

II.

5H3OH

3HOH (C6H5)3CC1

CH3OH

CH2OC(C6H5)3

CHOH 2 RCOC

T -*

CHpOH

1 I

H2OC ( C6H5 );

CHOCOR
CH2OCOR

~3~

CH2OH

CHOCOR

CH3OCOR

R'COCl

III. CH8ONa
I
CHOH

CH2OH

C6H5CH30C0C1

CH2OOCOCHsC6H5

CHOH

GHoOH

CH3OOCOCH3C6H5
CHOCOR "
CH20C0R

2 RCOC1

H>3
Pd

CH2OH

CHOCOR

CHoOCOR

R'COCl

IV.

CH2OH

CHOH

acetone

CHpOH 1^ HC1

CHoQ

I >C(CH3)2

CHO^

CH2OH

R'COCl

cold
quinoline

CH2OH

choh 2 R-aoci

CH2OCOR

f|H2

CH
I

CHpOCOR

HOI

CHgOH

CHI
I

CH2OCOR

RCOOK

CH3OH

CHOCOR

CHpOCOR

R'COCl

VI.

CH2OH
I

CHOH

CH3I

2RC0C1

CH3OCOR

CHOCOR

CH21

AgN03
EtOH

3H3OCOR

taoH

JHaOCOR

The a~g Acyl Shift In Dlglvcerldes

The expected product from VI (above) would be the a,§-
diglyceride, end isolation of the a, a' -diglyceride to the exclusion
of eny of the expected substance can be explained only by spying
that the acyl group first went into the beta position and sub-
sequently shifted to the al-pha position.

-4- 101

H30H

ch3o m CH3OCOR

|"vV -* CHO^ \R -* CHOH
CH3OCOR CH3OCOR CH3OCOR

Fischer postulated the above mechanism of acyl shift
several decades ago. Hibbert and Grieg, in 1931, substantiated
Fischer's hypothesis by obtaining a stable ring structure in
the acylation of ethylene glycol with trichloroacetic acid.

CH3OH CH30 /OK

I + ClgCCOOH -> I NC + H30

CH3OH CH3(/ XCC13

Apparently the polarity occasioned by the presence of the
three chlorine atoms serves to stabilize the oxolone structure.
Hibbert and Grieg presume that in all such acylations of a lpha -
dihydroxy compounds there is an equilibrium between such an
oxolone structure and the normal ester.

This a-p shift, seemingly, is applicable only to aliphatic
acyl groups; several a, (3-aryl diglycerid.es have been found to
be stable. Further, this acyl shift is not encountered during
such catalytic hydrogenations as shown in equations II and III..

Bibliography

Grun, Ber., 40, 792 (1907)

Fischer, Ber., 53, 1621 (1920)

Helferich and Sieber, Z. physiol. Chem., 175, 311 (1928)

Garner, J. Soc . Chem. Ind . , 47, 278, 801 (1928)

Bergmrnn and Carter, Z. physiol. Chem., 191, 211 (1930)

Hibbert and Grieg, Can. J. Research, 4, 254 (1931 )

Jackson and King, J. Am. Chem. Soc, 55., 678 (1933$

Verted e et al, Rec. trav. chim., 54, 716 (1935)

Daubert and King, J. Am. Chem. Soc, 61, 3328 (1939)

Daubert, ibid, 62, 1713 (1940)

Daubert and King, Chem. Revs., 29, 269 (1941 )

Reported by Jos. F. Shekleton
April 15, 1942

DETERMINATION OF BRANCH CHAIN METHYL,
CITRIC ACID, AND ORGANIC HALOGEN

102

Determination of Branch Chain Methyl

Kuhn and L'Orsa developed a method for determining side-
chain methyl groups in natural products (Vitamin A, Carotenoids,
etc*) composed of isoprene units. The method consists in oxidiz-
ing the substance with chromic acid in sulfuric acid, then dis-
tilling and titrating the acetic acid formed. Pregl and Roth
have adapted the process to micro-analytical work. King has
applied the method to volatile materials by carrying out the
oxidation in a micro bomb.

A variety of compounds has been analyzed in this way
(Table I).

Compound
C2H5OH

C2H50C2Hs
C2H5OC- — R

CH3C^— OR

CH3C~CH2R

CH3CH-CH-R

I I

OH 0!

OH
CH3CH=CH-R

CH 3

=CH-C=CH

CH3C

TABLE I.

Moles of Acetic Acid Moles of Acetic Acid

per mole of Compound Compound per mole of compound

^<2>

& N^CH.

.95-1

.85

.95

.85

.90

.10

,12

QK3/CH3

.40

CHo JDH9

lC^C-CH=CH-C

CH.

-CH.

CH^-C

!5h 2.70

CH
I

CH2
CHpOH

C H o C H q

CHa NC~CK=CHC~

L3
I

CHS

CH.

i-CH.

2.00

on3

W N^vNH.

.70

103

It is evident from the figures given that the method is not
as precise as might be desired. It has "been most useful in
choosing one of several possible formulas by comparing the amount
of acetic acid found with the t calculated from the figures given
in Table I. For example, a-ionone would be expected to yield
(0.4 + 0.90 + 0.85) - 2.15 molecules of acetic acid per molecule.
There were found 2.0 molecules.

Determina tion of Citric Acid

Of the various methods used for the determination of citric
acid, the pentabromoacetone method has produced the most satis-
factory results. The details of the procedure are given by
Deysher and Holm in a recent publication. The method consists
of the oxidation of citric acid with potassium permanganate in
the presence of bromine, under controlled conditions. The acid
is converted quantitatively into pentabromoacetone, a white
crystalline solid which can be determined gravimetrically„

CKS-C00H

H0-C-C00H

CHs-COOH

KMn04
KBr

H3S04

CHs-C00H

c=o

CH3-C00H

P
CBr3(f— CBr3H

Kunz was the first to utilize the reaction to determine the
citric acid content of milk, wines, and other food products* The
citric acid content of urine also has been determined by this
method .

Because of the slight solubility and volatility of the penta-
bromoacetone, no absolute method can be prescribed for the com-
plete recovery of citric acid as pentabromoacetone under all
conditions. The method must be standardized with respect to
the conditions and products employed. A complete discussion of
various other methods and modifications is given by Lampitt and
Rooke.

Kometiani determines the pentabromoacetone by treating it
with hydrogen iodide in alcoholic solution to liberate six atoms
of iodine per molecule. The iodine may then be determined
volume trically.

It seemed probable that compounds similar to citric acid
might be determined in a like manner but a study of the litera-
ture disclosed no such examples.

Deter.ii3Jna_ti.on. of Organic Haloget-

The determination of organic halogen by the liquid ammonia-
sodium process is both rapid and accurate. Chablay was the first
to employ this reaction quantitatively. The method of Vaughn
and Nieuwland, an adaptation of the Chablay method, does away

-"&M

104

with all special apparatus, and gives increased rapidity and
accuracy of analysis of insoluble materials. A small amount,
0.1 to 0.4 g. of the halogen-containing organic -material is
introduced into liquid ammonia contained in a beaker. If solu-
tion does not take place upon stirring, some organic solvent (as
ether) inert toward sodium in liquid ammonia is slowly added
until the material is dissolved. One gram of sodium is then
added in small pieces and the covered beaker allowed to stand
until the reaction is complete (usually 30 seconds to two minutes).
The excess sodium is removed by ammonium nitrate, and the solution
allowed to evaporate to dryness. The solids are taken up in
water acidified with nitric acid and the halogen content deter-
mined with silver nitrate either gravimetrically or volumetri-
cally by the method of Fajen or Volhard.

In some cases organic solids are produced during the decom-
position. These must either be filtered off or put in solution
by the addition of aldehyde-free acetone or other suitable
solvent before precipitation of the halide.

Clifford noted that cyanides were formed with certain highly
chlorinated compounds as carbon tetrachloride and hexachloroethane
end caused high halogen values. He devised a method for removal
of the cyanide which consists in boiling a. nitric acid solution.
Dains and Brewster investigated the action of sodium in liquid
ammonia on 12? organic halogen compounds and found cyanide pro-
duced from only eight. In ell these cases where cyanide was
produced two or more halogens were attached to a single carbon
atom. Methyl cyanide, benzyl cyanide, cyanoacetic ester evidently
splits off sodium cyanide directly.

Organic fluorides also may be determined by following this
procedure except that the fluoride is precipitated from the water
solution with calcium nitrate as calcium fluoride and determined
gravimetrically.

Some of the results of Vaughn and Nieuwlend may give an idea
of the reliability of the procedure.

Compound % Halogen Present % Halogen Found

Trichloracetic acid

Chloralhydrate

Hexachloroethane

Chlorobenzene

p_-Dichlorobenzene

Hexa chlorobenzene

Ethyl bromide

3, 5-Dibromobenzoic acid

Styrene dibromide •

n-Butyl iodide

Methylene iodide

Phenyl iodide

Fluorobenzene

1,3-Dime thy 1-5-bromofluoro benzene

o.-Chlorophenylf luorof orm

65.11

65.19 :'

64.33

64.21

89.86

89.91

31 . 53

31 . 42 '

48.28

48.23

74.73

74.80

73.37

73.46

57.15

57.00

60.56

59.95

68.98

68.86

94.77

94.84

62.23

19.79

19.78

9.36

9.30

31.58

31.36

-4- j05

The average per cent difference for all determinations run was
.14 although the large majority ran below 0.1^.

This method ha s been used in this labor? tory with good
results.

Bibliography

Kuhn and L'Orsa, Z. Angew. Chem., 44, 847 (1931)

Pregl and Roth, Quantitative Organic Microanalysis, P. Blakiston's

Son and Co. Inc., Philadelphia, Ed. Ill, 1937, p. 201.
Kuhn and Roth, Ber., 65, 1285 (1932)

Deysher and Holm, Ind. Eng. Chem. Analytical Ed., 14, 4 (1942)
Kunz, Arch. Chem. Mikroskop., 7, 285 (1914)
Lampitt and Rooke , Analyst, 61, 654 (1936)
Kometiani, Z. anal. Chem., 86, 359 (1931 )
Chablay, Ann. Chim. (9) 1, 510 (1914)
Clifford, J. Am. Chem. Soc, 42, 1573-9 (1920)
Vaughn and Nieuwland, Ind. Eng. Chem. Analytical Ed., 3, 274 (1931)

Reported by Curtis W. Smith
April 22, 1942

106

RECENT DEVELOPMENTS IN THE IDENTIFICATION
OF ORGANIC COMPOUNDS

I. Acid s

Dewey
bromid. e as
reagent is
When a hot
acqueous so
the derlvat
The £~ehlor
used "but bo
of the salt

and Shasky suggest the use of p-bromopseudothiouronium
a reagent for the identification of acids. The
prepared from p_-bromobenzyl bromide and thiourea,
alcoholic solution of this bromide is added to an
lution of the sodium or potassium salt of the acid,
ive precipitates out in the pure state on cooling,
opseudothiouronium chloride reagent has also been
th have the disadvantage that the melting points
s formed cover only

a small range (about 140-170 ).

In addition tc It? us-"1 as a oerbonyl reagent, phenyl--
hydrazine has been used to form acid derivatives.. Steffipel and
Schaffel tried this compound on aliphatic carboxy acids when
they observed the success of its use on aliphatic sulfonic acids.
To prepare the phenylhydra zide it is only necessary to boil a
solution of the acid in phenylhydra zine for thirty minutes. The
derivative precipitates on cooling or by the addition of benzene.

II. Alcohols

In order to identify the cellosolves, carbitols, and
related glycols, Seikel and Huntress developed a procedure for
the formation of the solid trityl ethers. The derivatives are
formed by heating the alcohol and triphenylmethyl chloride in
a mixture of pyridine and ether.

By means of a modified Schotten-Baumann reaction Lipscomb
and Baker have discovered a good way to identify alcohols in
an aqueous solution such ag would result from the saponification
of an ester. A solution of 3, 5-dinitrobenzoyl chloride in
benzene is made then d little ligroin is added to prevent freez-

ing of the mixture during the reaction. This
with the aqueous solution of alcohol, sodium

is
ace

shaken at
;ete, and ,

0
■lkali

Alcohols may be identified by the use of saccharin chloride
The reagent is easily prepared from insoluble saccharin and
phosphorus pentachloride. It reacts with alcohols in the
absence of water to give saccharin alkyl ethers in which the
alkyl gr-oup is attached to the carbon through the oxygen bridge.

ROH -* ^\^%^R

+ HC1

-2-

The isomeric nitrogen ethers can be formed by the reaction of
the alkyl helide on sodium seech? rin. Saccharin chloride was
used to identify primary and secondary alcohols.

Ill- Aldehydes and Ketones

By the use of the methyl p.- to luenesulf onate addition
product of nicotinic acid hydra zide it is possible to secure

107

CONKNH.

0 rfli <y bJ ■

solid derivatives of aliphatic aldehydes having a melting point

40 higher than the corresponding 2,4-dinitrophenylhyd.re zones.

It has the advantage in that the aldehyde is very easily recovered

or it can be converted directly into another derivative such

as the 2,4-dinitrophenylhydrazone. It gives oils with unsaturated

ketones and the range of melting points for the C6-C10 aldehyde

derivatives is narrow.

Sah added another carbonyl reagent to his long list when
he described the preoaration and use of p_-iodobenzhydrazide.
This compound is easily prepared from methyl p_-iodobenzoate and
40^ aqueous hydrazine hydrate. Refluxing a mixture of the hydra-
zide, excess aldehyde, and two drops of acetic acid in pure
alcohol gives the desired derivative. This reagent has two
distinct advantages: (a) the yields are so high that it is
easy to operate with 300 mg. of the aldehyde and still get enough
material for a study of physical properties and analysis, and
(b) the melting points of the derivatives of the elkanel series
from C6-C10 are widespread. This is not true of the 2,4-dinitro-
phenylhydrf zones or semicarbe zones for this series.

A new optically active carbonyl reagent of this type has
been described by Woodward, Kohmen, and Harris. This is prepared
from .1-menthol by conversion to l_-menthylhydrazide. It gives
very good derivatives with nice melting points end has the
added advantage of having a characteristic specific rotation.
It is of interest in that it was used in the first successful
resolution of dl-cemphor.

A new type of carbonyl derivative has been developed by
Henze and Speer. In their work with keto ethers, they found
that certain of their compounds did not give solid derivatives
with the ordinary carbonyl reagents but did form solid hydentoins.
The carbonyl compound was warmed in dilute alcoholic potassium
cyanide and ammonium carbonate:

~3~ 108

R'
R-CO-R' + KCN + (NH4)3C03 -> R-C CO

NH-CO-NH

(R1 can be K or R)

In most cases the hydantoin separates In a state of sufficient
purity to eliminate further recrystellization. This is not
useful for formaldehyde, certain unsaturated aldehydes, a few
nitro- and hydroxy-aryl aldehydes and pyruvic acid. However,
levulinic acid and ecetoecetic ester are converted to the
corresponding 5, 5-di substituted hydantoine.

IV. Amines

By the dia zoti zation of ]>-±odobenzhydrazide, one can get
p_~iodobenza zide. When this reagent is refluxed in an anhydrous
medium, usually toluene, the corresponding substituted urea is
formed. Again, as in the case of the hydra zide, only a email
amount (200 mg. ) is necessary for a complete physical examination
pnd analysis.

Billmen et al used 2, 4-dinitrobenzenesulfonyl chloride as
a reagent for the identification of amines. The reagent is
prepared from 2,2 ' ,4, 4'~tetrenitrodiphenyl sulfide and chlorine.

, ^s N03

°3N<C-3>~S~C1 + 2RNHs ~* °^<^l [^,-S-H-R + RNH2.HC1

It can be used even With a 30f aqueous solution of the cmine.

V- Hydrocarbons

Although it is not a new reagent and identification work
with it is not new, trinitrobenzene has been found useful in
identifying aromatic compounds. The derivatives, which are much
like the corresponding picrates, have been listed for certain
aromatic compounds: acids, aldehydes, amines, ethers,
hydrocarbons, ketones, and nitriles; mostly for amines, hydro-
carbons, phenols and phenol ethers. One interesting fact about
this reagent is that the number of trinitrobenzene molecules
used does not vary with the number of functional groups but
with the number of aromatic nuclei!. Certain pyrrole derivatives,
which do not give picrates, do form the trinitrobenzene compound.

The identification of pure paraffinic hydrocarbons has
usually been made only through the use of physical constants of
the compounds. Huntress has developed desoxycholic acid as a
reagent for the preparation of derivatives of these hydrocarbons.
He has applied it to three pentanes, five hexanes, seven heptanes,
eight octanes, three nonanes, eight decanes, and three higher

-4-

paraffins, making a total of thirty-seven hydrocarbons. The
derivative is produced when desoxycholic acid in methanol is
allowed to react with the hydrocarbon.

VI . Phenols

Pnenols can be identified by the use of p_-iodobenzazide
(see first reagent under amines). In this case the result is
a substituted urea.

VII. Sulfonic Acids

Although characterization of carboxylic acids by the
formation of their p_-nitrobenzyl esters has been a standard
procedure, the analogous preparation of the corresponding
sulfonates has hitherto proved impossible. By employing the
silver salt of the sulfonic acid with p_-nitro benzyl chloride in
pyridine, and taking advantage of the fact that the resultant
silver chloride is less soluble in hot than cold pyridine, and
that the solvent combines with the sulfonate to yield a readily
crystallizable pyridium salt with sharp and significant melting
point, Huntress has characterized twenty common aromatic sulfonic
acids.

Chambers and Watt have extended the use of benzyliso-
thiouronium hydrochloride to include 34 sulfonic acids. Impure
sulfonic acids can be used but if there is an amino group
present, the method usually fails.

If there is an amino group present and ordinary identification
methods fail because of inner salt formation or sensitivity of
the amino group, then Allen and Frame suggest the replacement
of the amino group by a chlorine atom througn the Sandmeyer
reaction. The sulfonyl chloride, sulfonamide or sulfonr nilide
can then be made. The yields are good enough so th: t 1.5-2 gms.
are all of the original acid that is necessary.

VIII. The Iodoform Reaction

Rothlin has developed a method whereby it is possible to
distinguish between R-CO-CH3 and R-CHOH-CK3 by means of the
iodoform reaction. Using a reagent mpde of one part potassium
cyanide, four parts iodine, five parts ammonium hydroxide and
fifty parts of water, a positive test can be obtained for the
methyl ketone only. It may also be used quantitatively to
determine the amounts of these two compounds in a mixture. By
the ordinary iodoform reaction, the total amount of the two
types can be determined; then by means of this second reagent
the amount of methyl ketone can be found.

~fi-

110

Bibliography

Dewey and She sky , J. Am. Chem. Soc . , 63, 5526 (1941 ).

Stempel end Schaffel, ibid., 64, 470 (1942).

Seikel and Huntress, ibid. , 63, 593 (1941).

Lipscomb and Baker, ibid'. . 64, 179 (194?,).

Meadow and Reid, Memphis Meeting, American Chemical Society,

April, 1942.
Allen and Gates, J. Org. Chem., 6, 596 (1941).
San and Ksu, Rec. Trav. Chim., 59, 349 (1940).
Henze and Speer, J. Am. Chem. Soc*, 64, 522 (1942).
Woodward, Kohman, and Harris, ibid. , 65, 120 (1941).
Sah, Rec Trav. Chim., 59, 364 (1940).
Sudborough, J. Chem, Soc, 109, 1339 (1916).
Billman et al, J. Am. Chem. Soc, 63, 1920 (1941).
Huntress, Memphis Meeting, American Chemical Society, April, 1942
Sah, Rec. Trav. Chim., 59, 357 (1940).
Chambers and Watt, J. Org. Chem., 6, 376 (1941).
Allen and Frame, J. Org. Chem., 7, 15 (1942).
Rothlin, see C. A., 35, 5091 (1941).

Reported by Q. F. Soper
April 22, 1942

ill

SULFADIAZINES

The antistreptococcic activity of Prontosil, 2,4-diamino-
azobenzene-4' -sulfonamide was discovered in the laboratories
of I. G-. Farbenindustrie by Mietzsch, Klarer, and Domagk in 1935.
That same year Tre'fouel, Trefouel, Nitti, and Bovet at the Pasteur
Institute while synthesizing Prontosil derivatives found that
sulfanilamide was just as active. During the next five years
synthesis of about 1300 new derivatives was published. Among
the most successful of these derivatives were sulfapyradine
(2-sulfanilamidopyridine), sulfathiazole (2- sulfanilamide thiazole ),
sulfaguanidine , and sulfadiazine (2-sulfanilamidopyrimidine ).

After its synthesis by Roblin and co-workers, sulfadiazine
proved to be more effective, more soluble, and less toxic in
general than the other active sulfa-drugs. As a result a great
many pyrimidine and pyrazine derivatives of sulfanilamide were
prepared.

(l) N-CH(6) N=£H HN-g=0 OH

N-CH 6 ) N=CI

hi,, ! '

IC CH 5 HC CI

(2) HC CH(5) HC CH H3N~C CH r=* H3N // N^

, , i L, , n ii it i n*=^

(3) N=CH(4) HC-N

pyrimidine pyrazine isocytosine

Synthesis of Isocytosine.

For the synthesis of isocytosine, an intermediate in the
synthesis of several pyrimidine derivatives, Roblin and co-workers
used a condensation of malic acid with guanidine sulfate in the
presence of fuming sulfuric acid.

Isocytosine
C00HCH3CH0HC00H + H3NC(=NH)NH3 -> [C00HCH3CH0 + HN=C(NH3)3] $

Among the earlier methods reviewed by Johnson and H^hn was
the alkaline condensation of guanidine carbonate with the sodium
salt of ethyl formylacetate.

C8H60C0-CH=CH0Na + HN=C(NH3)3 -> isocytosine

Davidson and Baudish developed a synthesis for uracil from
urea and malic acid.

NHeCONH2 + H00CCHSCH0HC00H f""-HsS04 ) mcll (2,4-di ke topyr imidine )

Using this method Hilbert and Johnson synthesized cytosine
and isocytosine from uracil according to the following reactions:

-2-

±±a

ale. Mixture of the

FU01a > 2,4~dichloropyrim, NH3 2-amino and 4-

amino chloropy-
rimidines

NeOCHa ) separate the H30,HC1 > H0_ / \ and ieocytosine

" methyl ethers XN — 4/

cytosine

Pyrimidlne synthesis

In the preparation of 2-aminopyrimidine Roblin vnd his co-
workers treated isocytosine Kith P0C13 and then reduced off the
chlorine in the (4) position with H2 and Pd(OH)2 on CaC03.

The same series of reactions starting with cytosine would
lead to 4-aminopyrimidine . The reduction can also be carried
out in either case with Zn and H20.

Using a procedure of Hale and Brill to obtain the sodium
salt of 2-hydroxy-5-nitropyrimidine, Roblin worked out a synthesis
of 5-aminopyrimidine.

NH2CONH2 + CHOCH(NOs)CHO ■ Efe0H.> NaO_^ ^-U02 • F0C1* y

2~chloro-5-nitro- Fe.HOAc ; 2~chloro-5~amino ^~> 5-amino-
pyrimidine pyrimidlne ^ pyrimidlne

Nitromalondialdehyd e can be made by the action of sodium nitrite
on mucobromic acid according to Hill who used pyromucic acid and
bromine as a source of the acid. Simoniswas able to make muco-
bromic acid in good yield by the action of bromine on furfural.

furfural + 5 Br3 + 3 H20 — > [pyromucj-c acXdl —> CEOCrzCCOOH

Br Br

Other condensations lead to alkyl and aryl pubstituted
pyrimidines, Gabriel has investigated a number of reactions
between amidines and various diketones.

±13

-3-

CH
NH=CyNH2 + CH3COCH2COCH3 M&L> ^_^f "^ 3 y=GH3 , phenyl, etc.

=^CCH

NH=CyNHa + CH3COCH3COOC2H5 -*• y.

OH
CH,

OH
NH=CyNH3 + NaOCH=CHCOOC2H5 -> y_

^N — ^O

-N

Caldwell and co-workers made 4, 5-dialkylpyrimidines by
refluxing the appropriate hydroxymethylene ketones with guanidine
carbonate in alcohol. The ketones were prepared by a Claisen
condensation of ethyl formate with the proper straight chain or
cyclic ketone.

RCOCH3R» + C2H5OOCH Nf* RCOC(R! )=CHOH KN=C(NHS)^ HgN^^^X^-Rt

** ~^N £r~

Ballard and Johnson synthesized 2-a.mino-5~carboxypyrimidine
as follows:

w ' ... OH
NH=C(SR)NH2 + C3H50CH=C(C00C3H5)3 Na0G*K*> RS-^ \-COOC2H5

% y

2-alkylmercepto- N .

POCla ^ Zn.HaOy 5-carbethoxy- Cla , Cl^T "N^__COOC3H5

pyrimidine ^N— .X^

a 1 o . NH a \ ale .and' 10% KOH ^ 2-amino~5~carboxypyrimidine.

The preparation of ethoxy methylene me Ionic ester was worked out
by Wheeler and Johns.

CH3(C00C3H5)2 + CH(C00C2H5)3 + (Ch3C0)20 + ZnCl2 -> C2H50CH=C(C00Et ),

Previously C2H50CH=C(CM)C00C2H5 and H0CH=C (CKO )C00C2H5 had been
used in the condensation. In the pseucothiouree R may be hydrogen,
ethyl, or benzyl.

-4-

114

Roblin, Winnek and English obtained 5-chloro-2-eminopyrimidine
from the condensation of chloromalondialdehyde with guanidine
carbonate in fuming sulfuric acid. The chloromalondialdehyde
was prepared from tetrachloropropene .

CHCl3C(Cl)=CHCl + 95^ H2S04 -> CE0--CHC1-CH0

CH0CHC1CH0 + NH=C(NH3)3 -* 5-chloro-2-eminopyrimidine .

Pyrazine Synthesis*

Kail and Spoerri prepared aminopyrazine starting with quinoxa
line.

KMnO* s

COOH

sublime v

/ %-COOE

, (70^ yield)

pyrazoic ?cid CKaQH^ ester

NH.

HC1

CH,0H

^ amide , NaOCl

./N^-NHCOaNa

aminopyra zine

The authors say that this is the first time a stable sodium
carbamate has been isolated from the Hofmann hypobromite reaction,

Sulfanilamide derivatives.
N4-substituente:

(4) 5 6

4NV //l

(It

S03NH3

3 2

The synthesis of N -pyrazinoyl sulfanilamide by Daniels and
Iwamoto illustrates one method of preparing N4-deriva tivee.

pyrazinoyl chloride + sulfanilamide dry

pyridine

-5-

4 115

4-chloropyrimidine is used in the synthesis of 2-N -sulfon-

emidopyrimidine according to the method of de Suto-Nagy end Johnson.

4-chloro- + sulfanilamide ale, v N — . NH >^* "%y_S02NH2

pyrimictme >v '

N -substltuents: ■

There are three m^ in methods of preparing N -derivatives,
(l) The use of ecetyleulfanylil chloride on an amine followed by
hydrolysis,.

}l
Hi

N'-<^3> + CH3C0NH_^~X_.S03C1 pyridine^ suifadi8zlRe

(2) The use of p_-nitrobenzene8ulf onyl chloride followed "by reduction

2-amino-5-chloro- , v

pyrimidine + 03N_.< 'X-S02C1 _py_v Fe.HCl , 5'-chloro-

n^ ^ * dil.alc/ sulfadiazine

(3) The action of an acyl chloride on N -acetylsulfaniiamide.

pyrazinoyl , ^

chloride + CH3C0NPL_X V_S03NH3 py . \0% NaOH

N S x ^

N -pyrazinoyl
sulfanilamide
Bibliography

Amundsen, J. Chem. Ed., 19, 167 (1942).

Ballard and. Johnson, J. An. Chem. Soc . , 64, 794 (194-2),

Caldwell, Kornfeld and Donnell, ibid. , 6.3, 2188 (1941 ).

Daniels and Iwamoto, ibid . , 63, 257 (1941).

Davidson and B?udisch, ibid.., 48, 2379 (1926).

de Suto-Nagy and Johnson, ibid. , 65, 3235 (1941).

Gabriel and Coiman, Ber. , 32, 1533 (1899).

Gabriel, ibid. . 37, 3641 (1904).

Hale and Brill, J. An. Chem. Soc, 34, 82 (1912)..

Hall and Spoerri , ibid. , 62, 664 (1940 ).

Hilbert and Johnson, ibid, , 52, 1153 (1930).

Hill, Am. Chem. J., 22, 95 (1899).

Johnson and Hphn, Chen. Rev.. 13, 193 (1933).

Morthey, ibid. , 27, 85 (1940).

Roblln, Willlamr, Winnik, and English, J. An. Chem. Soc, 62,

2002 (1940).
Roblin, Wlr.nik and English, ibid . . 64, 567 (1942).
Roblin and winnik, ibid. , 62, 1999 H940).
Simonis, Ber., 32. 2085 (1899).
Wheeler and Johns, An. Chen. J., 40, 238 (1908).

Reported by M. Chiddix
April 29, 1942

THE EFFECT OF CATALYSTS ON THE GRIGNARD REACTION ±1_6

Kharasch et al, University of Chicago

It is known that several metals and metallic compounds
have a significant effect upon the Grignard. reaction -- the
nature of the products, the rate of reaction, and the yield. In
order to study this effect more closely Kharasch and his co-
workers investigated the catalytic effect of several metallic
halides upon certain Grignard reactions.

The first reaction to be studied was that of isobutyl
magnesium bromide with benzophenone. Ordinarily this reaction
goes as follows:

CH3CHCH2Mg3r + 0CO0 -* 0-OHQH

92% yield

CH3

The reaction is thus a 2-electron Grignard reduction. However,
when the reaction was run in the presence of manganous chloride,
benzopinacol ae well' as benzohydrol was obtained, and the yield
of benzopinacol varied directly with the amount of manganous
chloride present up to 2 mol per cent of the latter. Chromic
chloride and ferric chloride gave similar results but with smaller
yields of benzopinacol. Cuprous chloride shewed no effect upon
the reaction.

The second reaction studied, was that between methyl magnesium
bromide and. benzophenone. The reaction ordinarily yields
diphenylcarbinol, but the presence of small amounts of metallic
halides altered the reaction to yield either benzopinacol or
the carbinol in quantitative amounts.

0CO0 + CH3MgBr, : »# .OH

f XCH3

OH OH

2 mol %

catalyst upoq f. pinacol % carbinol

Mg 0 95

CuCl 0 93

MnCl3 0 93

FeCl3 65 21

CoCl3 93 2

The
me thy Imp g
repgents
amounts o
a yield o
duct was
various o
diagram a

GH,

CH.

~2~

last reaction studied was that of i
nesium bromide. Ordinarily the eddi
to conjugated cyclic unspturated ket
f the 1,2 and 1,4 addition products,
f 71$ of the tertiary alcohol and it
obtained. However, in the presence
ther products were obtained as shown
nd table .

0 + CH3MgBr

ophorone with
tion of Grignard
ones yields varying

a nd i n thi s case
s dehydration pro-
of metallic haiides

in the following

CH-

3<x

Gil 3

, GH

Yields

Mol %

1,2 ad

•

Catalyst

Diene

prbinol

Cpd.V

Pinacol

1,4 addn.

Total

none

48.2

42.6

90.8

1.0 FeCl3

2.2

81.6

9.46

94.0

1.0 CuCl

6.96

82 . 5

89.5

1.0 NiCl3

7.3

22.6

61.1

4.73

96.0

1.0 CoCl3

16.1

78.5

94,6

20 xs Mg

55.5

22.7

1.45

80.0

1.0 AgCl

57.7

35.0

92.7

1.0 MnCls

56.0

28.5

84.5

ordinary

•85 . 5

4 . 35

1.45

91.3

Determination of the structure of compound V proved to be
difficult. On the basis of chemical evidence this compound,
should be either the er.ol form of isophorone or the isomer in
which the double bond is in the (3, /"position. If the latter
structure is correct, then a remarkable rearrangement tak.es
place when isophorone is treated with methyl magnesium bromide
in the presence of 2 mol per cent of ferric chloride:

ilb

Ch^

Ch3
i 3

* < >°

jH3

Following is the evidence for the existence of compound V as
distinct from i gopher one:

1. Carbon end hydrogen pnalysse show an empiric? 1 formula
CgH140, the same ^ s that of isophorone.

2. The physical properties °re different from those of
isophorone .

b,pa °C, b.o>(atm.) nS° d30

compound V 38.0-38*4 (4 mm.) 181-185 1,4620 0.9083

isophorone 69.0-69.3 (5mm.) 210-211 1.4775 0.9215

3. Compound V absorbs oxygen much faster than isophorone.
This property caused much difficulty in securing analyses, and
the compound he d to be distilled and Kept in high vacuum to get
consistent results.

4. Compound V el owl j changes to isophorone at room temp-
erature. This conversion is accelerated by a trace of acetic
acid, and in -the presence of potassium bisulfate at 150°, the
change is complete in one hour.

5. The possibility that compound V might be the alcohol
/\3-3, 5, 5, -trimethyicyclohexanoi f orn ed by the reduction of iso-
phorone by methyl magnesium bromide v° s disproved by comparison
of the physical properties of the two.

B,,p. °C, n^° d30

Compound V 33-38,4 (4mm.) 174820 0.9033

Alcohol 69 (5 mm.) 1.4717 0.9144

6. A comparison of the reactivity of compound V end iso-
phorone toward semi carbamide was carried, out. Compound V gave

a precipitate immediately, out one to three hours were required
for isophorone to form any precipitate. Both s em icarba zones
melted at 186-i87°C, and they did not depress each other's mel-
ting poinrs.

7. Compound V is evidently an isomer of isophorone. The
enol form should give a test with ferric chloride, end compound V
did not, but this test is not conclusive.

3. Each compound reacted with hydro xylamine to give en
oxime melting at 78-79°, cut an equimolar mixture of the two
oximes melted at 00-52?-

-4-

Thus no definite conclusion could be reached as to the
structure of compound V. The best method of deciding on one
of the two proposed structures would seem to be their absorption
spectra since tne infra-red should reveal the presence of a
hydroxy 1 group and the ultra-violet should show the presence of
conjugation.

In attempting to explain the effects of the metallic halides,
certain considerations must be kept in mind. The coupling by
metallic halides has been studied many times, but in every case
at least one mol of the halide was used:

2RMgX + MX3 -> R-R + 2MgXs
2 mols 1 mol 1 mol

However, when one mol of bromobenzene is dropped into a
mixture of phenyl magnesium bromide (one mol) plus 3 mol% of
cobaltous chloride, the bromobenzene acts as an oxidizing agent
in converting the phenyl magnesium bromide to bi phenyl. That
the biphenyl is formed exclusively from the C-rign- rd reagent is
shown by the fact that the bromobenzene may be replaced by
p_-tolyl bromide, ethyl bromide, isopropyl bromide, and others
with exactly the same results. The reaction is applicable to
other biaryls, and n-Q1 bitolyl, SL-SL* bitolyl, 4-4' bianisyl,
2-2' biphenetyl, were obtained in yields better than those
obtained using other coupling agents.

This reaction essentially involves the transfer of an
electron to the halide from the organic radical of the Grignard
reagent. The metallic halide acts as an oxidation-reduction
catalyst since it is reduced by the G-rignard reagent and oxid-
ized back by tne organic halide. The need of only small amounts
of the metallic halide suggests a chain mechanism:

C6K5MgBr + CoCl2 -» C6K5CoCl + MgBrCl

2C6H5CoCl -> C6H5-C6H5 + 2 CoCl-

CoCl' + C6H5Br -> CoOlBr + CSH5«

XC6H5' — > C6E6, C6H5-C6H5, C6H5-C6H4-C6H5, etc.

This mechanism also explains the pinacol rather than the carbinol
in the first reaction studied:

CHgMgBr + CoCl 2 -» CH3CoCl + MgBrCl

2CH3CoCl -* C2H6 + 2C0C1-

2CoCl. + 208CO + MgBrCl -* 2C0C12, + (03C-OKgBr)3

Not all Grignard reagents would react with bensophenone in
the presence of cobaltous chloride to give the pinacol even

_5- 120

though the primary requirement, reaction of the Grignard reagent
Kith the metallic halide, were fulfilled. The seconc" step,
decomposition of the RC0C1 into R-R and CoCl, must be sufficiently
rapid or the normal addition reaction will take place exclusively.
Thus in attempting to alter the normal course of a Grignard
reaction, the rate of normal condensation as well as the stability
of the intermediate must be considered.

Bibliography

Kharascn, Kleiger, Martin, and Mayo, J. Am. Chem. Soc . , 63, 2305

(1941).
Kharascn and Tawney , ibid. . 63, 2308 (1941).
Kha.rasch and Lambert, ibid,, 63, 2315 (1941).
Kharascn and Fields, ibid. , 63, £316 (1941).
Linsteed, J. Chem. Soc, 1603 (1930).

Reported by A. B. Spradling
April 29, 1942

J2i

THE CHEMISTRY OF EXPLOSIVES

Because of the contemporary importance of explosives, their
chemistry, both organic end inorganic, is a topic of extreme
interest. Inasmuch as exact information of modern developments
is unavailable, this report will summarize the practices of the
explosives industry during and following World War I.

History

Historically, tradition records that Berthold Schwarz, a
German monk, discovered black powder around 1250, but various
documents indicate that Roger Bacon in the thirteenth century was
the first European scholar acquainted with the use of saltpeter
in incendiary mixtures. Even earlier, around the eighth century,
the Byzantines and Greeks described various flame-throwers and
pyrotechnics. After guns began to be used in the fourteenth
century, the best proportion of potassium nitrate, charcoal, and
sulfur (6:1:1 ) was found and later developments were chiefly in
the manufacturing methods. As an improvement, the potassium
nitrate end sulfur were replaced by ammonium nitrate to produce
Ammonpulver, a cheap, powerful, flashless, smokeless propellent.

The discovery of smokeless powder by Felouze (1838) and
its development by Abel in the form of guncotton (nitrocellulose)
marked the beginning of modern firearms; Nobel was the first to
gelatinize guncotton end nitroglycerin with acetone.

Properties.

The important physical properties to be considered are:

1. Shattering Power (brisance)

a. Strength of Detonation- depends on volume of gas
evolved and amount of heat liberated. Determined
by measurement of the distension after explosion

in e lead block (Treuzl test), by recoil of a pendulum
in a ballistics gun, or by means of a crusher gauge
in a manometric bomb.

b. Velocity of Detonation- depends on constitution end
packing density. Measured by a rota ting-drum
chronograph or by comparison with standard samples
(Deutriche test).

2. Sensitivity
a . Impa c t

b. Temperature of Ignition

.?. Stability

a. Time end temperature for N02 evolution

b. Hygroscopic nature

\22

•2~

Classification

Explosives may be classified conveniently by reference to
their responses to the stimuli causing the .explosions. Thus,
there ere primary explosives, propellents, and high explosives
(with an intermediate sub-class of boosters).

The behavior of the various classes of explosives is best
illustrated by considering a typical ammunition shell containing
both a. propelling charge and a bursting charge.

Igniter

Propellent

Primer

Driving

Band

Booster
/

u :

Bursting charge

Detonator
Fuze

Thus ,

flame

sets

burni

ga. s e s

collo

inal

acqui

the s

is ne

the p

the fir
, igniti
fire to
ng of th

from th
ided nit
perfora.t
re a rot
oft cop-o
cessary
ro j^ctil

ing pin cau
ng the blac
the igniter
e grains of
e combust! o
rocellulose
ions, cause
a ting mot le-
er driving
to insure a
e.

ses the primer cap to produce a small
k powder in the primer. This in turn
, from which the large, hot flame causes

smokeless powder, the propellent. The
n of the smokeless powder, which is

of cylindrical grains with 1-7 longitud-
s the projectile to move forward and to
n as the rifling of the gun bites into
band. The proper burning of the powder

gradual and powerful acceleration of

If the shell is intended to explode a definite time after
leaving the gun, a firing pin in the fuze undergoes set-back,
striking a primer cap containing a primary explosive which
ignites the slow-burning powder train adjusted by time rings.
When this flame reaches the detonator and bursting charge, the
shell explodes in flight. For detonation on striking the target,
8 percussion element is included. The centrifugal force causes
a firing pin to rise up so that inertia, will cause it to strike
the cap, produce fire in the powder charge, and result in an
explosion. The usual high explosive must be insensitive to
tolerate the shock of the set-back. In general TNT is used for
this purpose along with a booster of tetryl which aids in the
detonation and reduces the quantity of primary explosive necessary
The bursting charge may also contain shrapnel or a war gas.

Naturally, there is some overlapping in the classification
of explosives. For example, nitrocellulose may function as a
propellant if ignited by fire, or as a high explosive if detonated
by a primary explosive such as mercury fulminate. In
adoption of a particular explosive depends to a large
its stability, sensitivity, and reaction to climactic
in addition to its strength.

all cases,
extent upon
conditions

-3- 123

I* The Prime- ry Explosives (initiators, detonators)

These compounds explode or detonate when heated or subjected
to shock, but do not burn. Their uses vary with their brisance;
mercury fulminate, lead azide, ana nitromannite are by far the
most common initiators..

1. Mercury fulminate - Hg(ONC)3 ~ used with potassium
chlorate and sometimes with powdered glass. Prepared

by dissolving mercury in 60% nitric acid, adding alcohol,
and separating the grey oowder. Eleven times as sensitive
as TNT".

2. Lead Azide - Pb(N3 )s - half as sensitive as fulminate.
Prepared by 0

NftNH8 + N80 -+ NaNi Pb^N°*-H Pb(N3)s

3. Nitromannite - the hexe-nitrate of mannitol, which is
made by electrolytic reduction of mannose. Nitrosorbitol
is also used.

Others of lesser importance are lead picrate, trinitroresor-
cinol, m-nitrophenyldia zonlum perchlorate, tetracene, nitrogen
sulfide, copper acetylide, and nitrosoguanidine.

II. The Propellants (low explosives)

These are combustible materials containing within themselves
all the oxygen for their combustion; they burn but do not explode
and function by producing gas which explodes.

1. Black Powder - in the United States, it usually contains
75# sodium nitrate, lb% carbon, and 10$ sulfur. It is
now used chiefly in coal mining, as time fuses, and in
shrapnel shells. Its products of combustion are 56^
finely divided solids.

2. Smokeless Powder - Cellulose in the form of clean cotton
or wood chips is nitrated by a nitric-sulf uric acid
mixture to give a product containing 12.8^ N, which
corresponds to the formula C24H30010(N03 ) 10. This
"nitrocellulose", plasticized by alcohol-ether and
stabilized by diphenylamine, is forced through perforated
dies, cut, and dried to give the desired guncotton. An
improved military rifle powder now being made contains
13.1?? N.

3. Cordite - a double-base powder developed by the British
which contains nitrocellulose and about 30^ of nitro-
glycerin, gelatinized by acetone. Burns more readily
and faster and is well suited for trench mortars.

-4-

4. E. C. powder - semi-gelatinized nitrocellulose with
small amounts of potassium or barium nitrate; used in
hand grenades. NH

5. Nitroguanidine (NH3CNHNOs) made from guanidine nitrate,
which comes from cyanamide by the addition of ammonia,
then subsequent treatment with nitric acid. May be used
with a chlorate or perchlorete.

III. The Hi gh Exp 1 o g 1 v e g ( bur sting cha rge s )

These substances detonate under the influence of a suitable
primary explosive. They do not function by burning and sometimes
are not even combustible. They ere not readily exploded by
heat or shock but are extremely brisant once they are detonated.
The chief members of this class are the dynamite and liquid
oxygen types used for industrial purposes, and the nitro aromatic s
and newly developed nitro aliphatic s employed in warfare.

1. NITROGLYCERIN AND DYNAMITE. -~C3H5 (N03 )3 - The trinitrate
of glycerine is too sensitive; Nobel, however, discovered
dynamite, a solid made by absorbing the liquid in
kleeelguhr. The filler now used in the United States

is wood meal and sodium nitrate and acts as a cushion
for the explosive during handling. A dynamite cartridge
is fired by an electric spark in the sequence of primer
mixture, fulminate, end dynamite.

To eliminate the hazard of frozen nitroglycerin,
ethylene glycol dinltrate or glycerin dini tromono-
chlorohydrin is usually added to lower the freezing point.

2. AMMONITES* — This term induces mixtures of ammonium
nitrate with aromatic nitro compounds or combustibles
such as sawdust and coal.

dynamon - ammonium nitrate with wood chips, widely

used in Russia .
amatol - 20-B0i mixtures of ammonium nitrate and TNT.
ammonal - 15# TNT, 17^ Al, 65% NH4N03, Z% C.
alumatol - same as above except for 3% Al; used in
hand grenades.

3. BLASTINE.— 60# NE4C104, U% TNT, 22# MaN03, 1% wax;
used in mines.

4. LOX. — Liquid oxygen absorbed on carbonaceous material
contained in a canvas wrapper. Retardents such as
diammonium phosphate solution may be used to reduce
handling and storage hazards. Used chiefly in open-air
mining.

5. CARDOX.-- (liquid C03) in steel cylinders is ignited by
a. special electric spark; sometimes used in mining.

~5~

1.25

6. NITROSTARCH. used as filler for hand grenades and
mortar shells.

7. NIB - NITROGLYCERINE.— ( nitroisobutylglycerol trinitrate;
(KOCHg)3CN03) obtained from CH3N03 and CH30 is nitrated.

8. TETRANITROMETHANE, — from acetic anhydride and nitric acid.

9. NITROGUAM1DINE, NITROCELLULOSE. —listed previously as
propellents.

10. TETRACENE (guenyl-nitro saminoguanyl-l-tetracene.) —

This substance ha s a wide range of sensitivity depending
upon its density. It varies from a slow burning com-
pound to a detonator. However neither tetracene nor
its mixtures with potassium chlorate are satisfactory
substitutes for mercury fulminate. Commercial prepar-
ation:

H3N~C-NHNH3

(HNOa)

3 /S

NaNCU v

H30^
HO AC
10°

H3N~C-NH~NH-N=N~C~NH-NH-NO

« ii

NH NH

tetracene

11. HEXOG-EN (trimethylenetrinitramine) (Syn. Cyclonite)
Except for Penthrit this is one of the most brisant
of all explosives. This high brisance makes it a useful
underwater explosive for use in mines and. in depth
bombs. For this purpose it is usually mixed with
30-40% of TNT. It can be made quite cheaply because
the apparatus required is very similar to that used in
the high pressure process for making methanol.
Commercial preparation:

6CH30 + 4NK3 ■?► '^ti "?

'CH3 CH3"

CH3 CH3 J3HZ

Cone .

H exam ethylene tetramine

03N-|}— CHa-N-N08
CH3-N— CH3

tog

Hexogen

12. PENTHRIT fpentr ery thritol tetra nitrate ) This is the
most powerful of the known military explosives. Like
Hexogen, it has a high brisance which makes it a useful

126

underweter explosive, enC also makes it useful as a
bursting charge in artillery shells. It is stable,
easily stored, can be prepared from coal, water, air,
limestone and sulfuric acid, and hence can still be
made by any country even though that country's imports
are completely cut off.
Commericfl preparation:

CH3CH0 + CHaO Cg(0H.^) C(CHgOH)4 -^ * > C(CH3ONOs)4

Hg SO4

13. PENTRI NIT.— This is the name applied to mixtures of

Penthrit and Nitroglycerol in various ra'tios. Pentrinit
60/40 (60°/ Penthrit ?nd. 40% Nitroglycerol) has the
highest rate of detonation of any of the known explosives.

The Nitro Aroma tics

Although several new and very good explosives have been
developed in the aliphatic class, the aromatic nitro compounds
still constitute the largest ahc! most important class of the high
explosives*

1. TNT (2,4.6-trinitrotoluene) (syn. Trotyl, Tolite).--
This is probably the mort widely used of the nilitary
high explosives. It is widely used in demolition
shells," the TNT constituting as much as 60^ of the weight
of the shell.

Commerical preparations A 3-stage direct nitration pro-
cess is generally used in the commercial preparation
of TNT froro toluene, although 1- and 2-stage processes
have been used.

Purification of product: The crude product which
separates from the reaction mixture usually contains
several impurities, the chief of which are: (3~(2, 3, 4-) ,
Y~(3,4,6~), and £~(2,3?6~) TNT, TNB (l, 3, 5-trinitro-
benzene) and DNT (s, 4-dinitro toluene ) . There may also
be impurities from the toluene used. However, this
latter class of impurities is usually removed by an
initial purification of the toluene. In spite of this,
some TNX ( trinitro-m-xylcne ) is usually present in the
final product, but in such a small amount that it does
no harm. , ■ Some of the DNT is removed by washing with
sulfuric acid, and whet little is left causes no trouble.
It is not all removed, because too large an amount of
, TNT would be dissolved by the sulfuric acid. The ^-TNT
is present in too small an amount to cause any trouble;
however, the p- and y-THT must be removed. This is
accomplished by treating the crude product with sodium
"sulfite. The products formed from the p- and Y-TNT
are dissolved in water, and then converted; to m-methyl
tetryl:

-7-

±2

*
t

3H3

P-TNT

Na8S03
warm

CHaNH^

CH3

1 /CH3

N°2 CH3

03N_/V_N03

1 TVH-

02N

CH3

N0S
Y-TNT

.N03 warm

0*N

)

l_S03Ha

CH^NH^

OpN

CHS

^0.

NO.

^CH3

these are H20 sol.

m-methyl tetryl

2. TNB (l, 3, 5-trinitro benzene ) This is the most powerful

explosive among the nitro aromatics, and is less sensitive
to impact than TNT.

Commercial preparation:
CK3

[

COOH

N03 02N
HgSOA

Na2Crs07
40-50°

-CO.

OaN

-/\v-NO

NO,

Although benzene can be nitrated directly to TNB, the
process is too expensive of acid and heat to be used
commercially.

5, TNX ( trinitro-m-xylene ) This compound has a large excess
of carbon in each molecule, so it is used with oxidizing
agents such as NH4N03, It is also used to lower the
m.p. of compounds such as TNT and PA, However, it
attenuates the explosive power of these compounds to some
extent. TNX is., prepared by a 3-stage nitration of m-
xylene.

' .

■'.

»;•

- •

-8-

4. HEXANITROBIPHENYL (2,2,J4,4*,6,6,~) This compound is
said to have explosive propertieB superior to those
of hexil. It is also non-toxic and very stable chemi-
cally. It cannot be made by direct nitration of
biphenyl since only the tetranitro compound is obtained.
Commercial preparation:

Cu -powder

5. PA (Picric Acid) (Syn. Metinite, lyddite, shimose) This
is a very powerful explosive, but has been largely
supplanted by TNT, even though the latter is somewhat
less explosive. This is because PA forms salts with
the common metals (except Sn end Al ) , which are very
sensitive to shock. (in fact lead picrate is an
excellent detonator. ) Because of this reaction with
metals, use of PA in shells necessitates an inside coat-
ing of a varnish.

The straight nitration of phenol is not a satis-
factory method for preparing PA for two reasons:
1. The -OH group weakens the ring and makes it more
sensitive to oxidation, 2. Some p_-nitrophenol is
always lost because of its volatility. Therefore PA
is made commercially by an indirect method:

HNO.

■»

H2S04
40-95°

NsOH

autoclave

N0;

0J-/V

'NO,

NO;,

Picric Acid

' " ■ 1

-9-

129

There is elso a catalytic process which is
commercially feasible. This process gives the same
overall yields as the above method.

HNO.

Hg(N03):

■>

6. AMMONIUM PICRATE, — This explosive is less sensitive to
shock than PA. In fact it is not even detonated by
mercury fulminate, and a booster such as compressed PA
or Tetryl has to be used to explode it. It is very
suitable for use in armor piercing shells, especially
for coast guard guns. A mixture of ammonium picrete and
picric acid is known as "Explosive D" and is quite
widely used in artillery shells in this country.
Commercial preparation: PA is suspended in hot water
and an excess of strong NH4OH is added. The ammonium
picrate separates on cooling..

7. TNA (tetranitroaniline ) This compound ha s a very high
velocity of detonation, but it is too reactive chemically
to be of much use. It does find Borne use as a booster,
however.

Commercial preparation:

N02

NaaS

S°4

NH.

^ HNO a

H3S04'

V 0aN^

N08

N03
_N0S

8. TETRYL (l, 3, 5-trinitrophenylmethylnitramine )
probably the most important of the boosters,
powerful and more brisant than TNT or PA.
Commercial preparation:

This is
It is more

CH,~N-H

CH,NH:

CHa-N-NO

HNO.

0aN

H2SO,

. . ■

:;.i.J'

■

•

■

;

■!

[

■

-10-

180

HEXIL (2,'2l,4r4,,6l6,-he«s.nitrodiphenylaffllne) This is
another important booster like tetryl. It is slightly
less sensitive to shock then tetryl, but it has the
disadvantage of attacking the skin. It is made commer-
cially from aniline and dinitrochlorobenzene:

NO,

NO.

Oaggg

60°

OpN

HNO

NO,

N03 NQa ^

NH_

EXPLOSIVE RIVETS

An important industrial use of explosives that has
recently been patented is the use of explosive rivets.
These are extremely valuable in the manufacture of air-
planes, since they can easily be used when the riveter
can reach only one side of the rivet.

The explosive rivet consists of a rivet with a hollow
shank wherein is placed a very small charge of lead
azide. The rivet is then placed in the hole and a
riveting iron is applied to its head. This iron heats
the rivet, causing the charge to explode within 1.5 to
2.5 seconds. The explosion causes the part of the shank
that has passed through the plates to expand, thus lock-
ing the rivet firmly in place. This expansion of the
rivet can be controlled to within 0.02 inch.

Bibliography

Clark, Ind. Eng. Chem., £5_, 1385 (1933)
Desvergnes, Chimie et induetrie, 2.8, 1038
DeWilde, ibid.,,, 30, 1034 (1933)
Rinkenbecn and Burton, Array Ordnance,
Storm, ibid. . gl, 20 (1941)
Denues and Huff, Ind, Eng. Chem., News Ed.
Lewis, ibid. . 19, 782 (1941 )
Klrkpatrick, Chem. and Met. Eng., 47, 744
Berl, ibid. , 46, 608 (1939)
Woodward, ibid. , 4£, No. 2, 117 (1941 )
Rollend, ibid., 48, No. 6, 96 (l94l)
Be.bie, ibid., 48, No. 10, 76 (1941 )
Hardy, ibid. . 49., No. 4, 76 (1942)
Cullen, J. Soc, Chem. Ind,, 5£, 812 (1939)

(1932)
12, 120 (1931)

, 18, 1114 (1940)
(1940)

i •. i

•

:.■

■

, . ■- -

■■•'...•' 3

: . j ■':"- ... i* fe.

'■':■. . . v -■

. . '■' " ■.'■■'■ • '■

:■'." ■ ■ ' ' '.

/" .■'■■ s

131

-n~

Marshall, Nitroglycerine and Nitroglycerine Explosives, Naoum,

Williams and Wilkine Co. (1928)
Encyclopedia Britannica, Edition 14, Vol. 8 (1929)
Riegel, Industrial Chemistry, Reinhold Publishing Corp., N. Y.

(1937)
Davis, The Chemistry of Powder and Explosives, Vol. I, Wiley and

Sons, N. Y. (1943 )

Reported by R. S. Voris

P,, F. War field
May 6, 1942

^ \ o — ffi — ca

P to O O

«:

o
a

M
Eh
<

4-3 4-3

•H -H

CO O

S to

£} H

o 3

U rH

CO rH
■P CD
CO O
O O
Ih Sh
4-3 +J
•H «H
C C

4j +j

c a

T t

rH rH
O O
4^> 4-3
■H -H

C rO

cd o

Tt T

C/J CD <D
XJ Q «i oq
O 3 O O

h a ^

CU CO o

CO
4-3
C£i o B Cti

K
O

132

' ■■■

■

133

RECENT DEVELOPMENTS IN THE STUDY OF VITAMIN E

Although many p_~hydroxychromans show slight vitamin E
activity, it is not possible to modify the structure of
cc-tocopherol very much and without loss of part or all of the
vitamin E activity . Karrer and coworkers have synthesized
homologs of cc-tocopherol in which R (formula 1^ has been modified
by addition or subtraction of isoprene units. Also they have
synthesized derivatives in which R*,RS and R3 has been varied.
A summary of the homologs they have prepared is given in Table I.

R3k\ A ,?-CH,R

In cc-tocopherol, R?',Ra;R3 = CH3

and R = C15R31 "3 isoprene units"

All of these compounds were obtained by condensing the
proper hydroquinone derivative with the :-llyl bromide or alcohol.

The general methods used for the synthesis of the hydro-
ouinones are as follows:

ilOH

The substituted hydroquinone s were all synthesized by one
of these methods, with only one exception, that being 2-ethyl-
3,6-dlmethyl— ; D-hydroquinone . In this case, the tri-substituted
phenol necessary for the above syntheses could not be obtained
readily by synthesis. For this particular case the following
procedure was employed:

•2-

lol

CH30

OCH.

CH.oOr/

CH^CHOHSN ^OCH

CHoO/^

BrMg

CH3

loc»s

CH30

CH3

OCH3

ZllB

CH3

An application of Fischer's synthesis of pnytol wrs used for
the synthesis of fllyl bromides. For example, l-bromo-3, 7, 11-
trimethyldodecene-2 was preprred in the following manner:

CH3-CH(CH3)3CCH3 + HC=CH 'V2 CH3CH(CH3 )3C-C=CH

CH3 0 CH3 0H3

III

li3^t CHaCH(CH3)3A~CH=CH3 P-^3 CH3Ch(CH* 3C=CH-CK3Br

JH3 CH3
IV

CH3 CH3
V

135

-3-

CpH

a«s

CH3COCHNaCOOEt P NaNH2

— > CH3CH(CH2)3CH(CK2)3CH(CH2)38cH3 J 1 ^

end reduction

CH3 CH3

H2(-Pt

CH3CH(CH2 )3C-C=CH "3*r \

CH3 CH3 VIII

VII

PBr3

, ) CH3CK(CH2)3CH(CH2)3C=CH-CH2Br

CH3 Cn3 CK3

The l~brorno-3,7-dimethylpentene-2 which ws s needed for the
synthesis of the homo log in which R(Formula I) is C5Hn appears
in the above scheme. The higher homolog, i.e., the one in
which R is C20H41, was obtained from phytol by the same method
as above.

One compound containing an unsaturated hetero-cyclic ring
has been prepared by Karrer. The method of synthesis here was
the condensation of trimethyl hydroquinone with the bromide
obtained from the action of PBr3 upon 3,7, 11, 15-tetramethyl-
hexadecyn-l-ol-3. The latter was obtained by treating the ketone,
resulting from the ozonolysis of phytol with acetylene and
sodamide. It also occurs as one of the intermediates in the
synthesis of phytol from compound IX above.

A New Method of Synthesis of Tocopherals

Smith and coworkers have devised another method of synthesis
of 6-hydroxychromens which does not depend on the condensation
of an allyl compound with hydroquinone s. This method is as
follows:

..*'.-

-4-

136

CH30
CH3

CH3

CH3CH3OH

OCH.

-y

ch3o

CH3

CH3
CH3

CK3CH2MgBr CH30

H3 CH3

OCH-

"* CH.

Ch3Ch3C— R

CH3

OCH

HO
CH.

CH3 CH3

This method so far has been applied only to chromrns con-
taining methyl groups in the 2,5,7, and 8 positions. By varying
the nature of the ketone, five 2, 5,7, 8-tetramethyl-2-alkyl-6-
hydroxychromans have been prepared, namely, R=CH3, C2H5,
n_-C3H7, i_sp_-C4H9, end C16H33(4,8, 12-trimethyl decyl).

For the preparation of cc-tocopherol, the ketone was obtained
from phytol by ozonolysis. The carbinol was prepared from
trimethylhydroquinone dimethyl ether by bromination, conversion
to the Grignard reagent and treating with ethylene oxide.

From a synthetic point of view, the proof of the structure
of cc-tocopherol until now has been based upon the assumption
that phytol and its derivatives would behave in the condensation
reaction as the simpler analogs do, i.e. to form a chroman ring.
Since the new method of synthesis leads unquestionably to a
chroman structure, the preparation of cc-tocopherol by this
method affords further proof, by synthesis, that the hetero ring
in a- tocopherol is a chroman.

Bibliography

Smith and Miller, J. Am. Chem. Soc . , 64. 440 (1942).

Smith and Renfrow, ibid. . 64, 445 (1942).

Karrer and Yap, Helv. chim. acta., 23, 531 (1940); 24, 640 (1941 ).

Karrer and Hoffman, ibid. . 23, 1126 (1940 ).

Karrer and Schlapfer, ibid.., 2_4, 298 (1941).

Reported by F.
May 13, 1942

W. Wymen

, i

TABLE I

13?

C 2 0^4 1
C 15^3 x

C 15H3 1
C 15H3 1

C 15^3 1
C15H3 1

Cl5^3 1
C15H3 1

C 15H3 1
C15H3 1
C 15H3 1

C 10^2 1

C5H1 1

on 3

CH3

C2H5

CH3

n u

^2^-5

-,zns

CH3

t~i T T

un3

Complete
Activity

No.
Activity

CH3

30 mg.

mg-.

CH3

3 mg.

mg.

CH3

6 mg .

CH3

5-10 mg.

CH3

CHa

8-10 mg.

CHg

C2H5

16 mg.

C2H5

CH3

riti
on 3

C2H5

CgH5

10 mg.

CH3

P H

10 mg.

C2H5

10 mg.

mg.

CH3

mg.

CH3

rag.

138

THE ACY1ATI0N AND ALKYLATION OF THE SODIUM ENOLATES

OF ALIPHATIC ESTERS

Hauser — Duke University

The scope of reactions of the sodium enolates of esters
was limited until a base was found which would convert esters
largely into their sodium enolates. The Claisen condensation
of simple esters, for instance, required the presence of two
a loha hydrogen atoms, if the condensing agent used was sodium
ethoxide. Two esters qualifying in respect to having two aloha
hydrogen atoms but failing to condense were ethyl isovalerate and
ethyl t-butyl acetate. Also, esters belonging to the class
which has only one alpha hydrogen atom did not condense.

In 1931 Schlenk, Hillemann and Rodlof'f showed that sodium
triphenylmethyl converts methyl diphenylaceta te into its sodium
enolate, consequently the use of sodium triphenylmethyl has been
extended to the condensations, acylations and alkylations of
those aliphatic esters requiring a strong base to form the enolate.
The enolate is formed according to the following equation:

(C6HB)3CNfj + H-C-COOR -> Na[C~COOR] + (C6H5)3CH

Ethyl isobutyrate, which has only one a loha hydrogen a torn,
has been converted to ethyl isobutyrylisobutyrate by sodium tri-
phenylmethyl; and ethyl isovalerate, which was noted above to be
unaffected by treatment with sodium ethoxide, similarly undergoes
self -condensation to give ethyl isovalerylisovalerate :

(CH3)2CHCHsCOOEt +

CHCOOEt

CK(CH3)S

_63^

(CH3 )3CHCH3COCHCOOEt
CH(Ch3)3

+ NaOEt

Spielman and Schmidt have shown that this same self -condensation
also takes place in the presence of mesitylmagnesium bromide,
cl though the yield is lower.

The acylation of the sodium enolate of an ester by a
different ester is satisfactory only when the latter has no
active hydrogen, for the sodium enolate apparently attacks an
available active hydrogen of the second ester more readily than
the carbonyl group to give a mixture of two enolates and two
free esters, from which four beta-keto esters might be formed.

~9_

139

Nitriles having hydrogens alpha to the -C=N group likewise
exhibit this hydrogen interchange:

CH3(CH2)2CH2CN + Na[CHaCOOC(CHB)»] -> Na[CH3 (CH2 )2CHCN] +

n-valeronitrile podium enolate of CK3C0OC(CH3 )3

t-butyl acetate

Na[CH3(CH3)2GHCN] + CH3COOC(CH3 );

0 CH3COCHCN

CH2CH2CH3 + NaOC(CK3)
cc-acetyl-n-valeronitrile

Ethyl oxalate, having no a lpha hydrogens, reacts easily to
give ethoxalyl derivatives, but ethyl formate gives low yields
of formyl esters:

COOEt

COOEt + Na[C(CH3)3COOEt] -}'''

sodium enolate of
ethyl isobutyrate

COC(CH3)2COOEt

GOOEt

+ NaOEt

HCOOEt + Na[C(CH3)2COOEt]

1&%

HCOC(GH3 )2GOOEt + NaOEt

Hauser hap also carried out e series of acylations with
acid chlorides and the sodium enolftes of esters having only one
alpha hydrogen atom. This method, in contrast to that using
esters as the acylating agent, gave satisfactory yields. The
general reaction is as follows, and the results are summarized
in Table I:

RC0C1 + Na[CR2C00Et] -> RCOCR2COOEt + NaCl

The a, a-disubstituted beta-keto esters so produced may be sub-
jected to ketonic hydrolysis to give certain ketones of the type
RCOCHR2. The complete procedure represents an extension of the
acetoacetic ester method of synthesis of these ketones. Table II
lists the experimental results.

Finally, certain other reactions of the sodium enoli tes of
esters might be listed:

110

•3~

1. With phenyl isocyanate the enolate of ethyl isobutyra.te gives
cc,a-dimethylmalonanilide ethyl ester:

C6H5N=C=0 + Na[C(CH3)3COOEt] ^ C6H5NHCOC (CH3 )3COOEt

2. With ethyl benzenesulf onate the enolate of ethyl isovalerate
gives ethyl cx-e thy li so valerate:

C2H5OS02C6H5 + Na

CHCOOEt
CH(CH3)a

33^

C2HBCHCOOEt
C

H(CH3)2 + Na.OS02C6H5

This method represents a gain of 11$ in yield over that obtained
when the same enolate was alkylated. .with ethyl iodide.

3. With ethyl cc-bromoisobutyrate the enolate of ethyl isobuty-
rate gives diethyl tetra.methyleuccinate:

0 CH,

EtOC— C~Br + Na[C(CH3)2COOEt]
CI

IH.

30^

0 CK3 CH3

II I I

EtO-C — -C C C-OEt

I I

CHn CHa

+ NsBr

4. With ethylene oxide the enolate of ethyl ieobutyrate gives
a, a-dimethylbutyrolactone :

CH3

CH2~CH2 + Na[C(CH3)2C00Et] -* NaOCH2CH2C-COOEt

CH,

NaOEt +

CH3

;h3ch2c-c=

CH3

•0'

A previous method due to Blanc for production of this compound
involved the reduction of cc,a-dimethylsuccinic anhydride, but
this procedure also results in the formation of the isomeric
P, P-dimethyllaotone.

-4- 14i

Bibliography

Blanc, Bull. soc. chim., 33, 893 (1905)

Hudson and Hruser, J. Am. Chem. Soc, 63. 3156 (1941)

Hudson and Hauser, ibid., 63, 3163 (1941)

Roberts and McElvain, ibid., 59, 2007 (1937)

Schlenk, Hillemann and Rotfloff, Ann., 497, 135 (1931)

Spielman and Schmidt, J. Am. Chem. Soc, 59, 2009 (1937)

Reported by John Whit son
May 13, 1942

14^

TABLE I

Results of Experiments with Na[CR3COOC2K5] and RC0C1

Acid

-Ester Used, Ethyls (C6H5 )3CNa Chloride Used

Isobutyrate 23

Mole Mole

0.198 0.195 Acetyl

Yield
.of Keto Ester.
B.P,

.190 .185 n-Butyrl

.95 .93 Isobutyryl 102

.875 .85 Benzoyl 123

Ieobutyrate 22

Isobutyrate 110

Isobutyrate 102

Methylethyl- 26 .20

acetate

Methylethyl- 26 .20

acetate

Methylethyl- 25.6 .197 .197 Benzoyl

acetate

Diethyl- 28 .20

acetate

0 Mole G % °C M'n

32 0.408 15.7 51 75-6 15

183-4 760

21.5 .20 20.2 58 109-11 29

.95 128.0 74 92-94 15

373 122.0 65 146-8 15

.20 Propionyl 18.5 .20 19.3 52 97-102 15

.20 Isovaleryl 24.6 .20 21.7 51 116-19 15

27.7 .197 24.0 52 164 18

.195 Benzoyl 28 .20 28.8 59 175-77 20

TABLE II
Ketonic Hydrolysis of Beta-Keto Esters

p-Keto Ester
Used, Ethyl

Ethylbutyryl-

isobutyrate
n-Butyryldi-

methylacetate

Propionyldi-

methylacetate

Isoveleryl-

methylethyl-

acetate

Benzoyldi-

methylaceta te

Benzoylmethyl-

ethylacetate

Benzoyldi-

ethylacetate

, — Hydrolysis Mixture, cc.^

CH3- 50f Time,
G. H3S04 H20 COOH HP Hr. Ketone

Yield 0 B.P.
G % C Mm

14.8 10

21.0 5

14 . 2 4

14.0 8

20.0 10

13.0

15.0

10 30

10 55

4 40

5 38

5 30

4.0 Di-isopropyl 7.0 78 121-25 760

3.5 n-Propyl .. i0.2 79 134-36 760

iso*-propyl
8.0 Ethyl iso- 6.7 78 134-36 760
propyl
14.0 Isobutyl 7.0 75 165-67 760
s- butyl

3.0 Phenyl iso- 13.4 81 102 15

propyl 218 760

75 75 48.0 Phenyl 6.0 69 109 10
s- butyl

75 75 48.0 Diethyl- 3.0 75 117-18 10

acetophenone 247-49 760

Internet Archive Audio

Featured

Top

Images

Featured

Top

Software

Featured

Top

Books

Featured

Top

Video

Featured

Top

Mobile Apps

Browser Extensions

Archive-It Subscription

Save Page Now

Full text of "Organic seminar abstracts"

See other formats