Synthesis of some iodinated benzoylbenzoic acids and anthraquinone derivatives

Northwestern
University Library
Manuscript
T h es e s
U n p u b l i shed th eses s u bm i tt e d for the M a s t e r ’s and
D o c t o r ’s d eg r ee s and d e p o s i t e d in the N o r t h w e s t e r n U n i v e r s i t y
L i b r a r y are o p e n for inspec tio n, but are to be use d only with
Bibliographical
due r e g a r d to the ri gh ts of the authors.
r e f e r e n c e s m a y be noted, but p a s s a g e s m a y be c op ie d only with
the p e r m i s s i o n of the authors, and p r o pe r credit m u s t be gi ve n
E x t e n s i v e c opy in g
in s u b s eq u en t wri tten or p u b l i s h e d work.
or p u b l i c a t i o n of the thesis in w h o le or in part r e qu ir es
also the c onsent of the D ea n of the G r a d u a t e S c h oo l of
Northwestern Univ e r w i t y .
T hi s thesi s by
has b ee n used by the fo l lo wi n g persons, whos e si|
a tt es t the ir a c c e p t a n c e of the above r es tr i c t i o n s
ures
A l i b r a r y w h ic h b or r ow s this thesis for use by
its p a t r o n s is e xp e c t e d to secure the s ig na tu re of each user.
NAME AND ADDRESS
D A TE
NORTHWESTERN UNIVERSITY
SYNTHESIS OF SOME IODINATED
BENZOYLBENZOIC ACIDS AND ANTHRAQUINONE DERIVATIVES
A DISSERTATION
SUBMITTED TO THE GRADUATE SCHOOL
IN PARTIAL FULFILLMENT OF THE REQUIREMENTS
for the degree
DOCTOR OF PHILOSOPHY
DEPARTMENT OF CHEMISTRY
BY
ROBERT WILEY HIGGINS
EVANSTON, ILLINOIS
JUNE, 1940
ProQuest Number: 10101530
All rights reserved
INFORMATION TO ALL USERS
The quality of this reproduction is depen dent upon the quality of the copy submitted.
In the unlikely event that the author did not send a com plete manuscript
and there are missing pages, these will be noted. Also, if material had to be removed,
a note will indicate the deletion.
ProQuest 10101530
Published by ProQuest LLC (2016). Copyright of the Dissertation is held by the Author.
All rights reserved.
This work is protected against unauthorized copying under Title 17, United States C ode
Microform Edition © ProQuest LLC.
ProQuest LLC.
789 East Eisenhower Parkway
P.O. Box 1346
Ann Arbor, Ml 48106 - 1346
ACKNOWLEDGMENT
The author wishes to express his appreciation
for the many suggestions and endless encouragement
given by Dr* C* M* Suter during the course of this
investigation*
The author is also indebted to Dr* Smith
Freeman for pointing out the possible utility of
iodinat ed-hydroxyanthraquinones•
The author is grateful to Northwestern Univer­
sity whose grant of a University Fellowship made a
portion of this work possible*
TABLE OF CONTENTS
page
INTRODUCTION
1
HISTORICAL
I*
II*
III*
Preparation of
anthraquinones
Preparation of
Preparation of
anthraquinones
Chloro- and Bromo• • • • • • • • • « • • • •
6
Iodoanthraquinones • • • • 18
Halogenated Hydroxy• • • • • • * • • • • • • •
19
DISCUSSION
I*
II*
III*
IV*
V*
VI*
VII*
Chemical Structure and Staining Action • • 24
Condensations of Tetraiodophthalie
A n h y d r i d e ..................
26
Preparation of Iodinated Phthalic
Anhydrides • • • » • • • • • • • • • • • •
50
Condensation Reactions of the Partially
Iodinated Phthalic Anhydrides • • • • • • •
35
Search for the Proper Solvent for the
Friedel-Crafts Reaction * • • • • • • • • •
44
Conversion of the Benzoylhenzoic Acids
into Anthraquinone Derivatives • • • • • •
48
Miscellaneous Reactions • • • • • • • « • •
55
EXPERIMENTAL
I*
VII*
VIII*
IX*
Preparation of the Iodinated Phthalic
Anhydrides • • • • • • • • • • • • • • • •
59
Condensation of Benzene with the
Iodinated Phthalic Anhydrides * ........... 66
Preparation of Some Simple Iodinated
Anthraquinones * o * * * « * * * * * * * *
70
Condensation of Anisole with Some
Iodinated Phthalic Anhydrides* • » • • • •
73
Various Other Condensation Reactions
of the Iodinated Phthalic Anhydrides • • • 76
Methylation and Acetylation of
2-(2-hydroxy-5-chlorohenzoyl)-3,4(?)diiodobenzoic acid * • * • • • • • • • • •
84
Preparation of the Diiodophthalimides . * 87
Miscellaneous Reactions « • • • • • • • •
88
Methods of Analysis • • • • • • • • • • »
92
SUMMARY
95
BIBLIOGRAPHY
96
VITA
98
II*
III*
TV*
V*
VI*
INTRODUCTION
It has been known for several years that the bones of
animals are stained red if the animals are fed on a diet con­
taining the root of the madder plant, Rubia tinctorium#
Duhamel and Hunter have found that the growing parts of the
bones of young animals are stained much more deeply than the
fully formed bones of the old animals#
This selectivity of
the madder stain in differentiating between growing bone and
fully formed bone has become of considerable value in study­
ing bone growth#
Many difficulties are encountered, however,
chief of which are that the amount of staining material in
madder is variable, and certain experimental animals, such as
cats, cannot be made to eat the quantity of madder necessary
for the effective staining*
It is known that the madder plant contains a number of
coloring matters related to alizarin but the actual component
responsible for staining the bone seems to be in dispute#
Since the isolation of alizarin from madder has been accom­
plished, a number of investigators have tried replacing madder
by alizarin#
Gottlieb1 was able to vitally stain the bones of
test animals by injecting sodium allzarinsulfonate*
He also
found that the bones of rats were stained when the animals
were fed on a diet containing alizarin.
He observed, however,
that the stain did not show the same depth of color as that pro—
Cl) Gottlieb, Anat. Anz. 46, 179 (1914)*
2
duced by the madder itself*
Some time previous to this,
Schreiber2 found it impossible to obtain any staining of the
bones of frogs or pigeons with alizarin and he concluded that
ruberythric acid, a soluble glycoside of alizarin which is
also present in madder, might be responsible for the staining.
A reinvestigation of the coloring matters in madder has
shown that in addition to alizarin, ruberythric acid, and
purpurin, madder contained a considerable amount of purpurin3-carboxylic acid and its glycoside*
Richter3 has isolated
this substance and its glycoside in a pure crystalline condi­
tion and he found that the purpurin-3-carboxylic acid was the
most active component in madder responsible for staining bones
red*
0
OH
0
OH
0
OH
COOH
alizarin
ruberythric
acid
purpurin-3-carboxylic
acid
Various organic and inorganic substances have been used
by rSntgenologists in studying different parts of the body
of both man and animals.
The organic substances containing
iodine seem to be of special value in this respect.
In men­
tioning (just a few of the organs made rBntgenologically visi-
(2) Schreiber, Arb* path* Anat* Bakt. 4, 257 (1904).
(3) Richter, Biochem. J. 31, 591 (1937J*
3
ble by the use of iodine-containing compounds, one thinks of
the gall bladder, the kidney, and the spinal fluid#
There are at present two different iodine-containing
organic substances used in studying the gall bladder by means
of X-rays*
On the one hand it was found that the monoalkali
metal salts of tetraiodophenolphthalein when prepared in the
colloidal form are of extreme value in rendering the gall
bladder visible4; on the other hand diiodoatophan, 2-p-iodophenyl-6-iodo-4-quinoline-carboxylic acid, C16H90sNIs, serves
admirably for the rBntgenological display of this same organ6.
In attempting to perfect a compound which would be of
use in the rBntgenological display of the kidney, Binz6 and
co-workers investigated some seventy-four different compounds,
finally observing that uroselectan, the sodium salt of 1,2 dihydro-5-iodo-2-keto-l-pyridineacetic acid, seemed to be the
best.
This compound acts so rapidly that within 10 to 15
minutes after intravenous injection the rBntgen picture can be
taken.
In regards to its toxicity, a 60 kg. man can withstand
180 g. of this iodopyridone without harm.
Finally, iodine-containing compounds have been found to
be of use in the X-ray examination of the spinal cord and its
membranes.
Smitt and Bok7 have reported using SJlipoidoln, a
compound of iodine and papaver oil which does not irritate
the tissues, and which when injected intraspinally can be of
(4) C. A. 23, 4777 (192.9) Brit. pat. 304589
(5) Pribam, Deut. Med. Wochschr. 52, 1291-4 (1926)
(6) Binz, R&th, and Lichtenberg, Z. angew. Ohern. 43, 452-5
(1930)
(7) Smitt and Bok, Nederland. Tijdschr. Seneeskunde. 68,
2213-26 (1924)
4
service in' testing by means of X-rays whether the normal
permeability inside the spinal cord is still preserved*
Apparently the above mentioned compounds are specific
for those particular parts of the body that they rendered
X-ray opaque.
Obviously, their opaqueness to X-rays was a
direct result of the iodine content, although it has been re­
ported that the rBntgenological visibility of the kidney and
urinary passages does not depend entirely upon the iodine or
bromine content of the injected compound8. From previous
work it has been found that the hydroxy anthraquinones of the
alizarin type exhibited a specificity in staining the bones.
Thus, it seemed reasonable to suppose that if one could in­
sert several iodine atoms into the alizarin nucleus or possibly
into the nucleus of other anthraquinones having a hydroxyl
group in the alpha position, one might obtain an iodinated or­
ganic molecule which would deposit a stain in the bones that
would be detectable by X-rays.
The synthesis of such compounds
was precisely the aim of this investigation.
It is believed that sodium alizarinsulfonate precipi­
tates out in the bones because active calcification is taking
place there.
If so, this compound would be expected to pre­
cipitate out at other foci in the body where calcification is
occurring.
Thus, when a person is recovering from tuberculosis,
(8) Binz and Maier-Bode, Biochem. Z. 252, 16-21 (1932)
calcium is being deposited about the lesions in the lungs.
If an iodinated alizarin molecule, which was opaque to
X-rays, could be injected into the body of that person, it
is possible that the progress of his recovery could be active­
ly followed through X-ray photographs*
There are also many
other possible physiological applications such as, detection
of gall stones and hardening of the arteries, both of which
are characterized by some deposition of calcium*
Thus, such
a compound would be of considerable medicinal value*
Of course it was realized that such an iodinated mole­
cule might prove to be toxic to the organism and in that event
It would be of little or no physiological importance*
How­
ever, this fact did not have any inhibitory effect on the in­
terest with which this problem was attacked, for there also
existed a possibility that these Iddinated hydroxy-anthraquinones might be of some value as dyes for inanimate ma­
terials*
Considerable difficulty was experienced at the outset
in trying to prepare any Iodinated hydroxy anthraquinones
that had at least one hydroxy group ortho to a carbonyl*
This need for an alpha hydroxyl group will be explained more
fully later*
Thus, as a preliminary to the synthesis of the
desired compounds, it was decided to prepare some simple
Iodinated anthraquinones since they too were practically un­
known*
It was believed that the knowledge gained through
their synthesis might be of some aid in preparing the more
desired substances*
6
HISTORICAL
The synthesis of alizarin from anthracene by Graebe
and Liebermann9 in 1868 and the subsequent establishment of
the relationship between anthracene, anthraquinone, and
alizarin by these same investigators marked the beginning
of a new era in the s tudy of anthraquinone and its deriva­
tives*
It is true that anthraquinone itself was known some­
time previous to this, having first been prepared by Laurent10
in 1840 by nitric acid oxidation of anthracene*
However,
following the announcement of this epoch making discovery of
Graebe and Liebermann, which soon led to the technical pro­
duction of synthetic alizarin, such an interest was stimu­
lated in research in this field that there has accumulated
so much knowledge about anthraquinone and its derivatives,
that today one may speak of an “anthraquinone chemistry1’•11
Since the present interest is focused on one phase of
this vast field of anthraquinone chemistry, namely the syn­
thesis of iodinated and iodinated-hydroxy derivatives, the
historical facts have been confined to a discussion of the
preparation of the known halogenated and some of the known
halogenated-hydroxy compounds*
I*
Preparation of Chloro- and Bromoanthraquinones
The chloro.anthraquinones have been well characterized
and a large number of all the possible isomers have been pre­
pared*
The two theoretically possible monochloroanthraqui-
(9) Graebe and Liebermann, Ber* 1, 49 (1868)
(10) Laurent, Ann* 34, 287 (1840)
(11) Phillips, Chem. Rev* 6, 157 (1929)
7
nones have been synthesized in several ways.
12
Schilling'
prepared the 1-chloro derivative from sodium anthraquinone1-sulfonate, while Bayer13 and the Badische Aniline and SodaFabrik14 obtained the identical compound from the correspond­
ing amino and nitroanthraquinone respectively.
The 2-chloro
derivative has likewise been prepared from the corresponding
sodium anthraquinone-2-sulfonate15,16,17 and from the 2-amino
derivative.18
Furthermore, the 2-chloroanthraquinone has been
obtained by a nuclear synthesis, both from 4-chlorophthalic
anhydride and benzene,l9,s° and from phthalic anhydride and
chlorobenzene.531
It has also been stated that a chloroan-
thraquinone may be obtained by direct halogenation with sulfuryl chloride in the presence of a trace of iodine.
No
statement was made as to which isomer was obtained.28
Theory predicts ten possible isomeric dichloroanthraquinones.
All of these have been synthesized.
The various
methods utilized in their preparation are summarized in the
accompanying table*
(12
(13
(14
(15
(16
(17
(18
(19
(20
(21
(22
Schilling, Ber. 46, 1066 (1913)
Bayer, German Patent 131, 538, Frdl. (3, 311
Badische Aniline and Soda-Fabrik, German Patent 252,
578 Frdl. 11, 545
Ullmann and Knecht, Ber. 44, 3128 (1911)
Coppens, Rec. trav. Chim. 44, 907 (1925)
Bayer, German Patent 205, 195, Frdl. £, 673
Kaufler, Ber. 37, 63 (1904)
Egerer and Meyer, Monatsh. 34, 69 (1913)
Ree, Ann. 255, 240 (1886)
HSchst, German Patent 75,288, Frdl. j5, 260
Houben, Anthracene and Anthraquinone; George Thieme,
1929; 274
8
TABLE I
Isomeric DIchloroanthraquinones
Prep, by
Prep* by replacement of
Dichloro- Haloge- Prep, by oxida­
anthranation nuclear
tion of
sulfo­ amino nitro hydroxy]
quinone
synthesis halogenic group group group
nated an­ acid
thracene
group
1.2
1,3
23*. 24
.
31
1.4
19.
26
25
23
27
28
15.17
U5_._
13
30
!».«
14
29
30,28
i.v
12
.
1*8
. 8.3
„TT
2 ,6
.
....
33
23,33,35
13
34
35
32
32
* Numbers indicate references*
This and several other derivatives have also been prepared
by a rearrangement of the halogen*
(25)
(24)
(25)
(26)
(28)
(29)
(30)
(31)
(32)
(33)
(34)
(35)
Ullmann and Billig, Ann* 581, 11 (1911)
Fierz-David, J* Am. Chem. Soc. 49, 2534 (1927)
Junghans, Ann* 599, 316 (1913)
Hammerschlag, Ber* Ij?, 1109 (1886)
Fierz-David, Helv* chim. Acta 10, 197 (1927)
Badische Anilin und Soda-Fabrik, German Patent 254,450,
Frdl* 11, 546
Goldberg, J* Chem* Soc* 1771 (1931)
Goldberg, ibid. 2829 (1931)
Badische Anilin und Soda-Fabrik, German Patent 197,554
Frdl. 9, 765
C.A. 27, 4687 (1933)
Kircher, Ann* 238, 344 (1887)
Tanaka, Proc* Imp* Acad., Tokio 3, 82 (1927)
9
Of the fourteen isomers of triehloroanthraquinone pre­
dicted by theory nine have been prepared*
Ullmann and Con-
g7
setti
obtained from phthalic anhydride and 2,4-dichloro-
phenol a dichlorohydroxybenzoylbenzoic acid which after
closure of the anthraquinone ring and replacement of the
hydroxyl group by chlorine gave a trichloro derivative*
They
also obtained the same substance by chlorination of 1-chloro—
4—hydroxyanthraquinone and the subsequent replacement of the
hydroxyl group by use of phosphorus pentachloride*
This com­
pound was of course the 1,2,4-trichloro derivative.
Further­
more, this isomer has been obtained from 3,4,6-trichlorophthalic anhydride and benzene, after ring closure of the
intermediate benzoyltrichlorobenzoic acid*3®
Two other isomeric trichloro derivatives were known be­
fore the work of Goldberg30 in 1931*
The authors of a French
patent37 state that chlorine at high temperatures reacts with
both anthraquinone alpha and beta monosulfonate to give di—
chloroanthraquinone alpha and beta monosulfonate*
These in­
vestigators then fused one of the compounds obtained with
boric acid and isolated a quinizarin-monosulfonic acid*
How­
ever, no mention was made as to which monosulfonic acid of
(27) Ullmann and Conzetti, Ber. _53, 826 (1920)
(36) Graebe and Rostowzew, Ber. 34, 2113 (1901)
(37) C.A. 2 , 3410 (1908) French Patent 384,471
10
quinizarin they had obtained*
Somewhat later the authors of
a German patent38 replaced the sulpho groups in the two origi­
nal dichloroanthraquinonesulfonic acids with chlorine and ob­
tained two trichloroanthraquinones*
No mention was made of
the positions of the chlorine atoms but Houben22 (page 278)
refers to them and apparently correctly30 as the 1,4,5- and
1.4.6-trichloroanthraquinones*
A compound having the same
melting point as the supposedly 1,4,6 isomer, prepared above,
was obtained from the condensation of 4-chlorophthalic anhy­
dride with p-dichlorobenzene followed by dehydration of the
intermediate benzoylbenzoic acid to what must be the 1,4,6trichloroanthraquinone*19
Six more of the isomeric trichloro derivatives were pre­
pared by Goldberg,30 they being the 1,2,3-, 1,2,5-, 1,2,6-,
1.2.7-, 1,3,6-, and 1,3,7- derivatives#
In addition he also
prepared the 1,4,5- and 1,4,6- derivatives by his method which,
with the exception of the 1,2,3- derivative, consisted essenti­
ally in sulfonating alpha-chloroanthraquinone and selected dichloroanthraquinones, separating the isomeric chloroanthraquinonedisulfonates and the dichloroanthraquinonemonosulfonates, and replacing the sulfo groups by chlorine*
It was
stated that this method of preparation of the trichloroanthra­
quinones was superior to the benzoylbenzoic acid synthesis be­
cause of limited applicability of the latter and of the ten­
dency of halogens to migrate under the influence of sulfuric
(38) Badische Anilin und Soda-Fabrik, German Patent 214,714,
Frdl. 9, 678
11
acid at high temperature*29
The 1,2,3-trichloroanthraquinone
was prepared through diazotisation of 1,3-dichloro-2-aminoanthraquinone•
Anthraquinone itself is attacked by halogens only with
considerable difficulty*
Nevertheless many chloro derivatives,
especially the poly substituted ones have been prepared in
this manner*
In carrying out the chlorination it is necessary
to. use either halogen carriers or certain solvents such as
fuming sulfuric acid*22 By use of free chlorine anthraquinone
can be chlorinated only In oleum and if one calculates the
amount of chlorine added, the halogen can be successively In­
troduced*
The alpha position is attacked most readily and
the 1,4,5,8-tetrachloroanthraquinone ultimately results.40
This same tetrachloro derivative was also prepared by Schilling
by replacement of the sulfonic acid groups, and by Hofmann41
through a nuclear synthesis*
An isomeric tetrachloro deri­
vative, the 1,2,3,4-tetrachloroanthraquinone, has been pre­
pared from tetrachlorophthalic anhydride
and benzene*34
Antimony pentachloride is of particular value in pre­
paring the poly substituted chloro derivatives.
It does not
react with anthraquinone at low temperatures, but on warming
there results, according to the quantity of the pentachloride
used and the reaction temperature, different degrees of chlor­
ination up to and including, the 1,2,3,4,5,6,7-heptachloro
-C.39) Heller, Ber. 45, 792 (1912)
(40) Bayer, German Patent 228,901, Frdl* _10, 578
(41) Hofmann, Monatsh. 36, 805 (1915)
2Lg
12
4; 2
derivative.
This heptachloro derivative on further treat­
ment with antimony pentachloride, is broken up forming perchlorobenzoylbenzoic acid, hexachlorobenzene, and tetrachloro­
phthalic acid.
However, it is possible to prepare the per-
^
ei
COOH
Cl
Cl/ \ c O O H
Cll
JcOOH
Cl
chloroanthraquinone if the anthraquinone is boiled with antimony
pentachloride in the presence of some iodine.
Besides the octa-
chloro derivative, there also results some 1,4,5,8-tetrachloroanthraquinone and some 1,2,3,4,5,6,8—heptachloroanthraquinone*
The 1,2,3,4,5,6,8—heptachloro derivative has likewise been pre­
pared by a nuclear synthesis from tetrachlorophthalic anhy­
dride and 1,2,4-trichlorobenzenef2
Some other interesting polychloro substituted anthraquinones, namely a penta- and a hexachloro derivative, have been
prepared from tetrachlorophthalic anhydride, chlorobenzene
and dichlorobenzene respectively^1
(42) Eckert and Steiner, Monatsh. 56, 269 (1915)
(42a) Eckert, J. prakt. Chem.
2 102, 561 (1921)
13
The mono— and dibromo substituted anthraquinones have
also been well investigated, both mono derivatives and all
ten of the dibromo derivatives having been prepared*
Again
the methods used in preparing the various isomers have been
quite varied*
Both the 1- and 2-monobromo derivatives have
been prepared by a nuclear synthesis; the 1—bromo derivative
from 3-bromophthalic anhydride and benzene,43 and the 2-bromo
derivative from the 4-bromophthalic anhydride and benzene.44
Other methods of preparing the 1-bromoanthraquinone are by re­
placement of the sulfonic acid or by the replacement of the
amm o group*
The 2—bromo derivative has also been prepared
from the corresponding sulfonic acid and by oxidation of
2,9,10-tribromoanthracene*45
The classical synthesis of alizarin by Graebe and Lieber­
mann from anthracene, through anthraquinone, was brought about
by brominating anthraquinone in a sealed tube at 100°G, and
fusing the resultant dibromo derivative with alkali.
The
actual structure of the bromoanthraquinone obtained by them
seems to be in dispute.
It was stated in one paper that the
derivative obtained was mainly the 2,3-isomeride which suffer­
ed rearrangement to the 1,2 derivative on treatment with alkali
However, according to Houben®2 the dibromo derivative actually
obtained was the 2,7-isomeride.
(43)
(44)
(45)
(46)
At any rate the 1,2-dihydroxy-
V. Pechmann, Ber. _12, 2127 (1879)
Waldmann, J. prakt. Chem. 126, 65 (1930)
Graebe and Liebermann, Ann. Spl. 7, 288 (1870)
Cain and Thorpe, Synthetic Dyestuffs and Intermediate
Products; Charles Griffin and Co., 1933; 212,
14
anthraquinone was obtained on alkaline fusion.
and Liebermann
Later Graebe
found it more satisfactory to first bromi-
nate anthracene and then oxidize the resulting compound to
the corresponding dibromoanthraquinone which w a s subsequently
converted into alizarin.
The methods of preparation of the ten isomeric dibromo
derivatives are summarized in Table II.
The higher bromo derivatives have not been well invest­
igated.
No reference to any tribromoanthraquinones of known
structure was found.
DiehlJ5 however, prepared some poly
substituted bromo derivatives by the further bromination of
a dibromoanthraquinone prepared according to the directions
of Graebe and Liebermann.
In addition to a tetra- and a
pentabromo derivative he isolated a tribromo substituted
anthraquinone but no statement was made as to the positions
taken by. the bromine atoms.
He also stated that all attempts
at preparing substituted derivatives higher than the pentawere unsuccessful.
However, somewhat later according to work
at the Elberfelder56 dye works, bromination of the anthraqui­
none nucleus, carried out in oleum, leads to the formation of
a tetra- and a heptabromoanthraquinone.
Eckert and Steiner42
on reinvestigating this patent were able to find only the
1,2,3,4,5,6,8-heptabromo derivative.
These same investigators
attempted, too, to brominate anthraquinone by using a mixture
(55) Diehl, Ber. 11, 181 (1878)
(56) Bayer, German Patent 107,721, Frdl. 5
,302
15
TABLE II
Isomeric Dibromoanthraquinones
/
Prep, by
Prep, by replacement of
Dibromo*-* Haloge-i Prep, by oxida­
anthra­ nation nuclear
tion of
sulfo­ amino nitro hydroxyl
quinone
synthesis halogenic group group group
nated an­ acid
thracene group
47
48,52
_1,8
to
1
-1
49
49
47
48,52
U4._.
_
53
1.5
47,52
53
1,6
17,47
47
1,7
47
1.8
47,52
26,50
2,3
2,6
54
44
51
47
47
~T? '
48,52
54
53
47,50,55
(47) Battegay and Claudin, Bull* soc* chim* 4 29, 1017 (1922)
(48) Grandmougin, Compt. rend. 175, 839 (1921)
(49) Ullmann and Eiser, Ber* 49, 2154 (1916)
(50) Perkin, J. Chem. Soc* 37, 555 (1880)
(51) Kaufler and Imhoff, Ber. 37, 4707 (1904)
(52) Battegay and Claudin, Bull. soc. ind. Mulhouse 86, 632 (1920)
(53) Dhar, J. Chem. Soc. 117, 993 (1920)
(54) Scholl, Eberle, and Fritsch, Monatsh. 32, 1043 (1911)
16
of antimony tribromide and bromine*
However, the attempt
was unsuccessful as the free bromine cleaved the anthraquinone
in a manner exactly analogous to the splitting with antimony
pentachloride•
Some poly substituted anthraquinones have been obtained
through a nuclear synthesis*
Thus, the 1,2,3,4-tetrabromo-
anthraquinone has been prepared from tetrabromophthalic anhy­
dride and benzenei1 In dehydrating the intermediate benzoyltetrabromobenzoic acid to the anthraquinone 13$ oleum was
used* •Hofmann
41
also condensed tetrabromophthalic anhydride
with bromobenzene and converted the benzoylbenzoic acid ob­
tained into a pentabromoanthraquinone*
The yield of the penta
bromo derivative was said to be quite low*
Furthermore, this
same investigator has condensed perbromophthalic anhydride
with a dibromobenzene, but no mention was made of the yield
obtained in transforming this intermediate product Into the
hexa substituted anthraquinone*
Briefly, it has been seen that several methods have been
used in preparing the chloro- and bromoanthraquinone derlvativesf in fact, there were seven distinct methods used, vary­
ing considerably In their degree of applicability*
They are*
direct halogenation, replacement of the sulfonic acid group,
replacement of the nitro group, replacement of the amino group
replacement of the hydroxyl group, oxidation of halogenated
anthracene derivatives, and by nuclear synthesis*
Halogen atoms in the alpha position in the anthraquinone
17
nucleus are much more labile than those in the beta position.
The alpha halogens are particularly easy to remove by reduc­
tion.
Thus, Kircher34 was able to obtain from 1,2,3,4-tetra-
chloroanthraquinone, 2,3-dichloroanthracene.
was brought about by zinc dust and ammonia.
The reduction
It was also found
that the alpha halogens could be removed by heating with po­
tassium acetate and a trace of copper in nitrobenzene.7 Thus,
the difficultly attainable 1—methylanthraquinone Is prepared
in this manner from the l-methyl-4-chloroanthraquinone*
If
the halogenated anthraquinone is heated with more copper and
without potassium acetate the halogen atoms are similarly re­
moved but the two anthraquinone residues unite.7958
On the other hand, the halogens in the beta position can
also become extraordinarily reactive through an alpha placed
cyano group.
At least this is the case with l-cyano-2-bromo-
anthraquinone, which reacts smoothly with amines and with
aminoanthraquinone with the splitting off of HBrf9 Likewise
the reactivity of the alpha placed halogens is enhanced by a
neighboring nitro group.
Further indication of the mobility of the alpha placed
halogens is given by the rearrangements that such substituted
anthraquinones undergo.
Thus 1-chloroanthraquinone on several
hours heating at 200°C. with concentrated sulfuric acid re­
(57) Ullmann and Minajeff, Ber. 45, 687 (1912)
(58) Scholl, Ber. 40, 1696 (1907*T"
(59) Schaarscbmidt, Ann. 405, 95 (1914)
18
arranges into the isomeric derivative*
the monobroxno derivative.
The same is true of
Likewise, the 1,5-dichloroanthra-
quinone rearranges into the 2,6 isomer and the 1,6-dichloro
derivative into the 2,7 compound*
Eckert and Steiner42 also
observed such a rearrangement on heating with antimony penta­
chloride*
Thus, they converted 1,2,3,4,5,6,8-heptachloro-
anthraquinone into the 1,2,3,4,5,6,7 derivative.
IX*
Preparation of Iodoanthraquinones
Even though many halogenated anthraquinones have been
prepared, the known iodine derivatives are limited in number.
The 1- and the 2-iodoanthraquinones have been obtained from
the amines
*61 and the 1,5-diiodoanthraquinone has been pre­
pared from the corresponding diaminef* An attempt to prepare
higher substituted lodo derivatives, was carried out by Varma
63
and Subramanyam*
These men investigated the action of iodine
and sodium nitrite in fuming sulfuric acid upon anthracene*
They obtained a mixture of mono- and diiodoanthraquinones of
unknown structure, the physical properties of which were not
even recorded*
A few years previous to this it was reported64
that the action of iodine on anthraquinone in
180°yields a complex mixture of products
70% oleum at
from which a tri-,
a tetra-, and a pentaiodoanthraquinone were isolated*
The
tetra compound was characterized by its high (500°) melting
point*
(60)
(61)
(62)
(63)
(64)
Laube, Ber. 40, 3566 (1907)
Kaufler, Ber. 37, 60 (1904)
Scholl, Haas, Meyer, and Seer, Ber.
62B, 107 (1929)
C.A., 25, 2994 (1931)
--Eckert and Klinger, J, prakt. Chem. 2 121, 281 (1929)
19
The only attempt made so far to prepare an iodinated
anthraquinone by nuclear synthesis is also due to Eckert and
Klinger®4 They found that the benzoyltetraiodobenzoic acid
obtained from tetraiodophthalic anhydride and benzene when
heated with concentrated sulfuric acid at 190-200° lost
iodine but nevertheless there resulted a 20# yield of a
tetraiodoanthraquinone, melting at 476° with decomposition,
which they believed was not the 1,2,3,4 compound because of
its high melting point*
Finally, an attempt to bring about
ring closure by the use of aluminum chloride resulted in the
evolution of iodine and hydrogen chloride with the formation
of a diiodoanthraquinone, melting at 290-292°, of unknown
structure.
One other condensation involving tetraiodophthalic
anhydride has been reported.
Laurance85 succeeded In condens­
ing it with toluene but no mention was made of any attempt to
convert the p-toloyltetraiodobenzoic acid into the anthra­
quinone derivative.
III. Preparation of Halogenated Hydroxyanthraquinones
A survey of the literature indicated that the chloro
and bromoanthraquinones have been well characterized but
that little was known relative to the iodoanthraquinones.
An analogous situation exists in regard to the corresponding
halogenated hydroxy derivatives, many chloro- and bromohydroxyanthraquinones having been prepared while the known iodohydroxy derivatives are very limited in number.
(65) Laurance, J. Am. Chem. Soc., 43, 2577 (1921)
No attempt
20
is being made here to summarize the preparation of all the
known chloro- and bromohydroxyanthraquinones*
Only those
that are pertinent to this investigation, such as the chloroand bromo- alizarin and quinizarin derivatives, have been
included*
On the other hand all the known iodohydroxy de­
rivatives are discussed*
The three isomeric monochloroquinizarins are known*
66
5-
The
67
and 6-chloroquinizarins have been prepared by a nuclear
synthesis, having been synthesized from the corresponding 3and 4-chlorophthalic anhydride and hydroquinone by fusion in
an anhydrous aluminum chloride and sodium chloride melt*
The
2-chloroquinizarin was prepared by direct chlorination in
g7
glacial acetic acid* Furthermore, several of the isomeric
dichloroquinizarins have been synthesized; these are the
5,6-, 5,7-, 5,8- and 6,7- dichloro derivatives*
The 5,6-,
5,8- and 6,7- dichloroquinizarins were first prepared by
68
Frey* He obtained these compounds by heating respectively
the 3,4-, 3,6- and 4,5-dichlorophthalic anhydrides and hydro­
quinone with boric acid for two and a half hours at 190°C*,
followed by heating the powdered intermediate product with
94$ sulfuric acid at 160-165°C.
The yields of the desired
products were low*
. Somewhat later these isomeric dichloroquinizarins^S*67
(66) Waldmann, J* prakt* Ghem* 150, 92 (1931)
(67) Waldmann and Mathiowetz, ibid- 126, 250 (1930)
(68) Frey, Ber. 45, 1361 (1912)
21
and in addition the 5,7-dichloro derivative, were prepared
by Waldmann and coworkers.
These investigators used the
same starting products, namely the isomeric dichlorophthalic
anhydrides (the 5,7—dichloroquinizarin being prepared from
the 3,5-dichlorophthalic anhydride) and hydroquinone#
How­
ever, as a condensing agent they used fused anhydrous alum­
inum chloride and sodium chloride*
reported#
Excellent yields were
Of the higher substituted chloroquinizarins the
5,6,7,8-tetrachloro69 derivative seems to be the only one
prepared.
It was synthesized by heating tetrachlorophthalic
anhydride, hydroquinone and boric acid, followed by heating
the intermediate in concentrated sulfuric acid.
low yields were obtained.
Only very
It was found that the two alpha
chlorine atoms could be readily replaced by hydroxyl groups
on heating with copper, water, and calcium oxide at 250° for
twenty-five hours#
The bromoquinizarins are not as well known as the chloro
compounds#
Nevertheless, the three isomeric monobromoquiniz-
arins have been prepared.
The 5-bromoquinizarin has been
obtained In two ways, first from 5-aminoquinizarin through
diazotisation and secondly, by a nuclear synthesis from 3bromophthalic anhydride and hydroquinone.
Likewise, 6-bromo-
quinizarin has been prepared by the condensation of 4-bromophthalic anhydride with hydroquinone in a manner already men­
tioned#
The third monobromo derivative, namely 2-bromoquiniz-
(69) IlBvermann, Ber. 47, 1210 (1914)
22
arin, was prepared by direct halogenationl0971 Only one di-
bromoquinizarin has been prepared, that being the 2,3-dibromo
derivative*
It was also prepared by direct brominationl0
The halogenated alizarin derivatives are not very well
known*
The 3-chloro compound has been prepared by halogena-
7S
tion,
and also by replacement of the sulfonic acid group
•4.
73
with halogen.
The corresponding bromo derivative has been
prepared in the same manner*
It has also been reported that
the gentle action of alkali oh 1,3—dibromo—2—hydroxyanthra—
quinone leads to the formation of 3-bromoalizarinI4 Some high­
er substituted halogenated alizarin derivatives were prepared
75
by Diehl*
He obtained a mono-, a di-, and a tetrachloro-
and a mono-, a di-, and a tetrabromo&lizarin by direct halogenation.
stated.
The position taken by the halogen atoms was not
A dibromoalizarin of known structure has been report­
ed by Dimroth and coworkers.
It was the 3,4-dibromo deriva­
tive and was obtained by bromination.
Briefly it has been seen that the chloro— and bromohy—
droxyanthraquinones have been prepared In a variety of ways
chief of which are, through a nuclear synthesis, direct halogenation in various media, replacement of the sulfonic acid
group in hydroxyanthraquinonesulfonates, and by replacement
(70) Liebermann and Riiber, Ber. 33, 1658 (1900)
(71) Dimroth, Schultze, and Heinze, ibid. 54, 3035 (1921)
(72) Wed., German Patent 189937, Frdl* 9, 685
(73) Hochst, German Patent 77179 Frdl. 4, 330
(74) Hardacre and Perkin, J. Chem. Soc. 180 (1929)
(75) Diehl, Ber. 11, 190 (1878)
23
of the amino group.
The known iodohydroxy derivatives have "been prepared by
76
a direct iodination in pyridine.
It was found possible to
iodinate 2-hydroxyanthraquinone and 2,7-dihydroxyanthraquinone.
Both compounds yielded only the 3-iodo derivative*
The 3-iodo-2-hydroxyanthraquinone was also prepared by the
iodination of l-bromo-2-hydroxyanthraquinone.
It was stated
that the elimination of the bromine atom might be ascribed
to the reducing action of the hydriodic acid which was simult­
aneously produced.
An attempted iodination of alizarin, an-
thrapurpurin, and flavopurpurin gave only negative results*
Later investigation, however, showed that if the 1-hydroxyl
group In alizarin were methylated the compound could be
iodinated in pyridine solution quantitatively.
Demethyla-
tion with halogen acid at 140°C. gave 3-iodoalizarin.
Even
though this method of iodination is effective for alizarin
1-methyl ether, it does not appear to have general applica­
tion because from the analogous anthrapurpurin 1,7 dimethyl
ether, an iodo compound could not be obtained!7
(76) Perkin and Story, J. Ghem. Soc. 2620 (1931)
(77) Perkin and Story, J. Ghem. Soc. 339 (1928)
24
DISCUSSION
I# Chemical Structure and. Staining Action
In organic compounds containing a hydroxyl group ortho
to a carbonyl there is apparently an interaction between the
H-atom ol the hydroxyl group and the carbonyl oxygen; that
Is, the hydrogen atom is more or less strongly attached to
the second oxygen by a hydrogen bond*
known as chelation.
This phenomenon is
The characteristic lake-forming pro­
perties of mordant dyes is a chelation phenomenon where
according to Werner78 the colored metallic lakes formed are
internal complex salts in which the metal is doubly linked
to the organic component.
However, the exact chemical nature
of the alizarin-lakes seems to be in dispute.
and Smith79 indicated that m
Work by Morgan
the case of the cobaltic alizarin—
lakes definite compounds were formed.
On the other hand more
recent work80 seems to indicate that the lake formation is
merely an adsorption phenomenon.
Regardless of whether or not definite compounds are form­
ed the phenomenon of chelation does play an important part in
the use of hydroxyanthraquinones as dyes. rRichter3 has stated
that in the vital staining of the bones by purpurincarboxylic
acid chelate rings were probably formed.
Therefore in attempt­
ing to prepare some iodinated hydroxyanthraquinones, which
would be of value as dyes, interest was confined to those that
t
(78) Karrer, Organic Chemistry, Nordemann Publishing Co.,
New York (1938)
(79) Morgan and Smith, J. Chem. Soc. 121, 160 (1922)
(80) Weiser and Porter, J. Phys. Chem. 31, 1824 (1937)
25
had a hydroxyl group ortho to the carbonyl.
The initial efforts were directed toward the preparation
of an iodinated alizarin derivative.
It was decided to
attempt the preparation of an iodinated alizarin molecule
"because alizarin itself was one of the hydroxyanthraquinones
known to "be effective in staining the bone.
A tetraiodo de­
rivative was desirable because, obviously the higher the per­
centage of iodine in the molecule the greater the possibility
of it being opaque to X-rays.
On the other hand it also
seemed that the iodinated quinizarin derivatives might be
equally as valuable because, by the same token that alizarin
forms an insoluble colored lake, quinizarin also should ex­
hibit the same property.
Finally, in order to markedly in­
crease the solubility of these compounds they were to be
converted into the sodium sulfonate derivatives.
The struc­
ture of the two most highly desired compounds is indicated.
I
0
I
0
OH
sodium
tetraiodoalizarinsulfonate
I
0
OH
I
0
OH
sodium
tetraiodoquinizarinsulfonate
In the vital staining of bones by compounds of this class
chelation probably has an important bearing on the formation
of the insoluble, highly colored calcium lake; the following
reaction may occur.
26
OH
S03 Ga
2
II.
Condensations of Tetraiodophthalic Anhydride
The first attempt at preparing a tetraiodoalizarin de­
rivative, carried out in this laboratory by other investigators, involved the condensation of tetraiodophthalic anhydride and 4-chloroveratrole.
However, it was not found possi­
ble to condense these tY/o components to the intermediate benzoylbenzoic acid.
Another approach to the synthesis of a
tetraiodoalizarin derivative was undertaken in the present
investigation.
An obvious method seemed to be the condensa­
tion of tetraiodophthalic anhydride and veratrole®
I
0
I
0
OCH3
I
The
0
00 -O"—coo
COOIi
I
benzoybenzoic acid thus obtained was to be chlorinated or
brominated, followed by ring closure to the anthraquinone and
subsequent demethylation.
27
0
I
OCHa
I
OCH.
Br«
I
OCHa
I
OCH.
k Br
GOOH
COOH
I
I
i
0
fI3 S 0 4
OCHf
I
or
HBr
OCH;
I
I
0
Br
It would have been necessary to brominte the benzoylbenzoic
acid obtained in the first reaction, otherwise on ring closure
hystazarin dimethyl ether would have been the predominating
product.
1
However, this method also failed, it being found
0
I
I
1
OCH.
I/'
OCH
1
0
XT' V>CH3
J0CHa
COOH
I
impossible to condense the two starting materials.
Anhy­
drous aluminum chloride was used as the condensing agent and
excess veratrole served as the solvent.
The reaction was run
at 120° for six hours.
Attention was next focused on the possible preparation
of tetraiodoquinizarin.
As quinizarin itself may be readily
prepared by condensing phthalic anhydride and p-chlorophenol
in a sulfuric acid medium containing boric acid^1 a possible
(81) Organic Syntheses
Vol. 6, p. 78
28
method of preparation seemed to be the condensation of tetra­
iodophthalic anhydride and p-chlorophenol carried out in an
analogous manner*
1
0
Again success was not realized, the reOH
CO 1
0
Cl
I
0
OH
I
0
OH
action mixture being decomposed into a tarry mass*
Since tetraiodoquinizarin could not be prepared in this
one step reaction, it was decided to attempt the preparation
of an intermediate product, a benzoylbenzoic acid, from
tetraiodophthalic anhydride and p-chlorophenol.
It was ex­
pected that on dehydration of this intermediate substance
tetraiodoquinizarin would result*
The procedure followed
was identical to that used by Graves and Adams8® in the con­
densation of 3,6-dimethoxyphthalic anhydride and p-cresol*
After running the reaction for a few hours the mixture became
too viscous to stir, finally caking and thus any reaction
that might have been taking place was stopped.
tive results were obtained.
Again nega­
However, it seemed that if the
condensation were run in some inert solvent the reaction
might be induced to proceed.
Ullmann and Schmidt83 were quite successful In preparing
benzoylbenzoic acids, through the Friedel-Crafts reaction,
(82) Graves and Adams, J. Am# Ghem. Soc. 45, 2439 (1923)
(83) Ullmann and Schmidt, Ber. 52, 2098 (1919)
29
using tetrachloroethane as a solvent for the mixture and re­
ported yields as high as 90# of the theoretical.
A reaction
was attempted under the identical conditions which proved to
be so fruitful for the original investigators.
However, in
the first place tetraiodophthalic anhydride did not appear
very soluble in this solvent and secondly considerable de­
composition ensued.
The black tarry mass left after removal
of the solvent gave no condensation product.
A survey of the literature indicated that tetrachlorophthalic anhydride had been condensed with p-cresol.83 There­
fore, it appeared that it might be possible to condense tetra­
iodophthalic anhydride and p-cresol under analogous conditions.
However, it was not found possible to do so.
It was also
found impossible to condense the tetraiodo anhydride with
m-cresol and 3-chlorocatechol.
Most of these reactions that
proved unsuccessful were tried several times under a variety
of conditions but to no avail.
It thus seemed as though it were impossible to condense
tetraiodophthalic anhydride with any component at all.
How­
ever, a search of the literature showed that it had been con­
densed with benzene64 and with toluenef5 Both reactions were
repeated but the yields obtained were much lower than those
reported by the original investigators.
Furthermore, in the
course of this research it has been found possible to con­
dense this anhydride with anisole and with o- and m-cresyl
methyl ethers.
not identified.
In the latter case the products obtained were
In the condensation with o-cresyl methyl
30
ether, the reaction being run for six hours at about 150°
with excess ether as the solvent and anhydrous aluminum
chloride as the condensing agent, two products were obtain­
ed*
The one was soluble in sodium carbonate solution and
the other, insoluble in sodium carbonate solution but soluble
in a solution of sodium hydroxide*
Both solutions on acidi­
fying with hydrochloric acid yielded precipitates, which
after crystallizing from dilute alcohol melted at 208-210°
£
and 251-253° respectively. Only the product insoluble in
sodium carbonate solution was obtained, in a pure state, from
the condensation with m-cresyl methyl ether.
The condensa­
tion was run in the same manner as in the case of the iso­
meric cresyl ether.
The product melted at 240-242°.
These
products were not further investigated because they were ob­
tained only in small amounts.
III.
Preparation of Iodinated Phthalic Anhydrides
Since a study of the literature revealed that the pre­
paration of 1,2,3,4-tetraiodoanthraquinone, by dehydrating
2-benzoyl-3,4,5,6-tetraiodobenzoic acid, had not been accom­
plished in spite of numerous investigations it became
apparent that it would be difficult to prepare tetraiodoali­
zarin or quinizarin by a nuclear synthesis even if the proper
conditions for the preparation of the intermediate benzoyl­
benzoic acids were found.
Purthermore, since Pratt and Per­
kins84 have shown that the mixture of di- and triiodophthalic
(84) Pratt and Perkins, J. Am. Ghem. Soc. 40, 219 (1918)
4s Only those melting points recorded in the experimental are
corrected.
31
anhydrides obtained by partial iodination of phthalic anhydride
in oleum can be separated, attention was directed toward the
possible preparation of the Isomeric diiodoalizarin and quin—
izarin derivatives*
The preparation of the isomeric diiodophthalic anhydrides
was carried out as indicated by Pratt and Perkins.
However,
the ease with which the various reaction products were sepa­
rated and obtained in a pure state was much less than the
original article indicated*
The reaction was reported as
being carried out in two different ways.
On the one hand the
iodination was effected in a short period of time at a high
temperature and on the other hand at a low temperature over
a long period of time.
The iodination was run a number of
times, making use of both procedures*
It was found that the
low temperature iodination gave the more satisfactory results*
The experimental procedure is given in detail In another
section.
Mention is being made here only of the modifica­
tions necessary to give more satisfactory results*
According
to the original article the neutral solution of the sodium
di- and triiodophthalates was fractionally precipitated by
successive additions of acid, warming after each addition
and stirring for two hours with cooling*
In a typical experi­
ment the addition of 15 cc. of acetic acid was said to give
a precipitate of 17 g* of nearly pure monosodium 4,5-diiodophthalate.
A second fraction of 31 g., which was chiefly
monosodium 3,4,6-triiodophthalate, was then obtained on the
addition of 10 cc* of acetic acid*
Finally, two other fractions
32
were obtained by use of hydrochloric acid; 20 cc, of acid
giving 91.5 g. of chiefly monosodium 3,4-diiodophthalate
and following its removal, the addition of a large excess
of acid, to the filtrate, gave 47 g* of nearly pure 3,6diiodophthalic anhydride.
The concentration of the acids used by Pratt and Perkins
m
effecting the fractional precipitations was not specified.
However, it was assumed that glacial acetic acid and the cus­
tomary 36% hydrochloric acid were used.
A few runs were
made using acids of these concentrations.
Not in any case
could the results of Pratt and Perkins be duplicated.
By
using glacial acetic acid the first fraction was always much
more than 17 g.
acid
The possibility then arose that the ”acetic
specified in the original work may have referred to 6 N
acetic acid and not to glacial acetic acid.
This was found
to be the case and in all the subsequent runs 6 N acetic acid
was used.
On the other hand the strength of the hydrochloric
acid used was the customary 36% acid.
The addition of 15 cc. of 6 N acetic acid to the warm
neutral solution of iodinated phthalates (see experimental
part) gave a precipitate weighing 17 g.
This, however, was
found not to be pure monosodium 4,5-diiodophthalate but was
contaminated chiefly with monosodium 3,4,6-triiodophthalate.
After its removal the addition of 27 cc. of 6 N acetic acid
gave a precipitate of 42 g.
This, too, was much more of a
mixture than was originally stated.
The remaining two frac­
tions of 110 g. and 41 g. were obtained by using 15 cc# and
33
20 cc, of concentrated hydrochloric acid, respectively.
The third fraction was supposedly nearly pure monosodium
3,4-diiodophthalate but actually proved to be contaminated
with many other products.
Finally, the fourth fraction was
likewise found to be a mixture of products and not the nearly
pure 3,6-diiodophthalic anhydride as stated by Pratt and Per­
kins*
Nevertheless it was found that a tedious purification
of the precipitates finally gave iodinated phthalic anhydrides
satisfactory for further work*
IV* Condensation Reactions of Partially Iodinated
Phthalic Anhydrides
An extensive series of experiments was carried out in
an effort to find satisfactory conditions for condensing the
di- and triiodophthalic anhydrides with phenolic compounds*
Many of the experiments gave negative results and in others
the yields were small.
One of the first reactions attempted
was the condensation of 4,5-diiodophthalic anhydride and
4-chlorove.patrole, using excess of the latter as a solvent
for the reaction and anhydrous aluminum chloride as the con­
densing agent*
This reaction was expected to yield a benzoyl­
benzoic acid as indicated in the following equation*
0
ALCL.
COOH
This
34
benzoylbenzoic acid on ring closure and demethylation was
then expected to yield 4-chloro-6,7-diiodoalizarin.
0
01
0
01
OCHg
0
0CHa
COOH
0
Cl
0-
OH
The initial reaction step was run Tor four hours at 130°•
However, the only product that could be obtained from the
reaction mixture was 4,5—diiodophthalic acid which resulted
from the hydrolysis of the 4,5-diiodo anhydride.
As it was found impossible to condense either the tetra~
,iodo~ or 4,5-diiodophthalic anhydrides with 4-chloroveratrole,
it became interesting to know whether the latter mentioned
substance could be made to condense with phthalic anhydride
itself.
At first, excess of the chloroveratrole was used
as a solvent for the reaction, which in one instance was
run for four hours at 130°•
Anhydrous aluminum chloride was
again used as the condensing agent.
In working up the re­
action mixture considerable difficulty was encountered in
the removal of the excess chloroveratrole by steam distilla­
tion.
In order to remove the last traces of the solvent,
an alkaline solution of the residue was extracted with ether.
35
A product was then obtained from the alkaline solution
which after one crystallization from acetic acid melted
at 189-193°.
The reaction was repeated using tetrachloro-
ethane as a solvent but instead of obtaining much of the de­
sired compound a considerable amount of tar was formed.
It
was apparent that a reaction occurred between these two
components but the reaction product has not been identified.
Approaching the preparation of the diiodoalizarin and
quinizarin derivatives in the same, general manner followed
in attempting to prepare the tetraiodo derivatives the next
step involved an attempted condensation of 4,5-diiodophthalic
anhydride and veratrole.
A method has been found whereby
these two components could be condensed.
However, the re­
action did not give good yields, in addition to being some­
what erratic.
A condensation of veratrole with the two
other isomeric diiodo anhydrides was therefore not attempted.
Likewise the other steps, indicated In the attempted prepara­
tion of the tetraiodoalizarin derivative, were not tried.
In spite of the fact that tetraiodophthalic anhydride
could not be condensed with p-chlorophenol in a sulfuric
acid medium containing boric acid, an attempt to condense
4,5-diiodophthalic anhydride with p-chlorophenol under the
same conditions seemed to be worthwhile.
Since both iodine
atoms in this anhydride were in the beta positions to the
carbonyls it seemed that they would be much less labile than
those in the alpha position and thus would not be readily
lost.
A mixture of 12.8 g. (0.032 mol) of 4,5-diiodophthalic
anhydride, 5.2 g« (0.040 mol) of p-chlorophenol, 2.5 g. of
boric acid and 50 g. of 95$ sulfuric acid was slowly heated
up to 200° over a period of one hour.
It was then kept at
that temperature for three and one half hours.
course of the reaction iodine was evolved.
During the
Likewise sulfur
dioxide was detected ensuing from the reaction mixture.
The
reaction mixture was poured on Ice and the precipitate ob­
tained coagulated by heating.
The solution was filtered and
the residue extracted with potassium hydroxide solution, a
large insoluble residue remaining.
On acidifying the alkaline
solution, a precipitate was formed which completely redissolved In sodium carbonate solution.
The product obtained from
it was then crystallized from hot water.
It melted at 215-*
218° and was obviously 4-,5-diiodophthalic acid.
In all 4 g.
of it was recovered.
Meeting with no success in this reaction attention was
then directed toward the possible synthesis of diiodoquinizarin derivatives through the intermediate benzoylbenzoic
acids.
Several attempts have been made to condense the iso­
meric diiodophthalic anhydrides with p-chlorophenol.
In the
case of the 3,4-diiodo anhydride the proper conditions for
obtaining the benzoylbenzoic acid have been found.
The
yields obtained seemed to vary considerably, fluctuating be­
tween 50 and 80$ under apparently identical experimental con­
ditions.
This reaction is discussed in detail in the experi-
37
mental section.
On the other hand the condensations of the 4,5- and
3,6-diiodophthalic anhydrides offered more difficulty.
After several attempts to condense the 4,5-diiodo anhydride
with p-chlorophenol, both, in excess of the chlorophenol as
the solvent and in various inert solvents, a small amount of
what was presumably the benzoylbenzoic acid was isolated.
It
has not been definitely identified because it was obtained In
such a small quantity.
Only 0.15 g. of pure product melting
at 280-282° was obtained from 3*1 g. (0.020 mol) of the 4,5diiodo anhydride, 35 g. (0.27 mol) of p-chlorophenol and 10
g. (0.075 mol)
of anhydrous aluminumchloride.
The reaction
was rym. at 75° for twenty-two hours, after which time 10 cc.
of O-dichlorobenzene was added.
tinued for another hour.
The reaction was then con­
On working up the reaction mixture
the condensation product was first isolated as the crystalline
sodium salt.
It was then converted into the free organic
acid and twice crystallized from a dioxane-alcohol-water
mixture, after which it melted as Indicated.
In addition
to the reaction product 6.5 g. of pure 4,5-diiodophthalic
acid was recovered.
The condensation of 3,6-diiodophthalic anhydride with
p-chlorophenol gave a mixture of substances from which no
pure product could be isolated*
This condensation was
brought about using 50 g. (0.39 mol) of p-chlorophenol, 15
g. (0.037 mol)
of 3,6-diiodophthalic anhydride, and 18 g.
(0.135 mol) of anhydrous aluminum chloride.
As was custo-
38
mary in all the- Friedel-Crafts reactions, the aluminum
chloride was added in small portions*
In this particular
reaction the temperature was gradually raised up to 65°
and run at that temperature for three hours, after which
20 cc# of Q-dichlorobenzene was added as a diluent#
The
reaction was then continued for fourteen more hours at 75°•
After the reaction mixture was acidified and the ex­
cess solvent steam distilled, an oily residue remained which
partially solidified on cooling#
This water-insoluble por­
tion was dissolved in hot sodium carbonate solution#
On
cooling there was no deposition of sodium benzoylbenzoate#
The alkaline solution was then acidified whereupon the con­
densation product was again obtained as an oil but on stand­
ing it resolidified#
An effort was first made to crystallize
it from an alcohol-water mixture#
Not meeting with any
success it was decided to try anhydrous solvents#
The pro­
duct was first crystallized from benzene and then from
toluene#
In all 10 g# of product was obtained which melted
over a wide range from 165 to 180°•
At first it was believed
that the compound might have been contaminated with the sol­
vent from which it was crystallized#
Therefore a sample was
dried in the Abderhalden at 15 mm# and 100° for five hours#
The product turned yellow and lost about 0#7 g# but it still
melted over the same range#
was therefore discontinued#
Further work on this reaction
39
Another possible approach to the preparation of a
diiodoalizarin derivative has also been investigated*
Al­
though tetraiodophthalic anhydride did not undergo conden­
sation with 3-ehloroeateehol, an effort to condense 4,5diiodophthalic anhydride with it seemed worthwhile*
The 3-
chlorocatechol was obtained by chlorinating catechol accord­
ing to the method used by Willst&tter and M&ller.06
In
this condensation the reaction mixture consisted of 2*7 g*
(0.0067 mol) of 4,5-diiodophthalic anhydride, 3*8 g* (0*028
mol) of anhydrous aluminum chloride, and 1 g* (0*007 mol);
of 3-chlorocatechol*
The reaction was run in boiling car­
bon disulfide for twenty hours.
It was expected that the
following reaction would take place.
It was believed that
the benzoybenzoie acid thus obtained could be readily cycliz*0
OH
q
off
A1C1,
COOH
ed into 3*chloro-8,7-diiodoalizarin*
However, it was not
found possible to bring about the initial condensation under
the experimental conditions employed*
Graves and Adams82 have shown that when 3,6- and 3,5**
dimethoxyphthalic anhydrides were condensed with o- and m-
f85) Willst&tter and M&ller, Ber. 44, 2182 (1911)
40
cresol by means of aluminum chloride the cresol nucleus was
substituted in the ortho-position in respect to the hydroxyl.
It thus seemed probable that in a condensation involving
phenol the ortho-position, in the phenol nucleus, would
likewise be substituted.
Therefore, it seemed that a like­
ly method of preparing an iodinated monohydroxyanthraqulnone,
having the hydroxyl group ortho to the carbonyl, would
be to condense the diiodophthalie anhydrides with phenol
and cyclize the benzoylbenzoic acids thus formed with sulfuric acid#
The practicability of this synthesis was investigated,
an effort being made to condense 4,5-diiodophthalic anhy­
dride with phenol.
The reaction mixture consisted of 5 g.
(0.012 mol) of the diiodo anhydride, 2.7 g. (0.087 mol) of
phenol, 7.7 g. (0.040 mol) of anhydrous aluminum chloride
and 50 cc. of carbon disulfide.
The reaction was run for
twenty-five hours at the boiling point of the solvent.
A
reaction product was isolated which, after one crystalliza­
tion, melted over a 28° range.
It was at first thought
that a mixture of isomers had been obtained.
As the pro­
duct did not crystallize very well, it was decided to
methylate the hydroxyl group and attempt crystallization of
the methyl ethers.
If the phenol nucleus had been substitut­
ed para to the hydroxyl the methylation process would re­
sult in the formation of 2-(p-anisoyl)-4,5-diiodobenzoie
an±dr— This acid had been previously prepared by condensing
41
4,5-diiodophthalic anhydride and anisole and was very easily
purified hy crystallization*
The total reaction product of 4 g*, melting at 215245°, was dissolved in sodium hydroxide*
Methyl sulfate was
then added in small portions until a precipitate was formed
and the yellow color of the alkaline solution had disappear­
ed*
The precipitate was filtered off and dissolved in hot
potassium carbonate solution*
After filtering and cooling
the alkaline solution a crystalline potassium salt was ob­
tained*
It was removed from the mother liquor, redissolv­
ed in hot water and acidified*
acidified*
The mother liquor was also
Both precipitates were then crystallized*
The
product that had first been crystallized as the potassium
salt melted at 224°-225° (cor*), after two crystallizations
from dilute alcohol*
This corresponds to the melting point
of the product obtained from anisole and 4,5-diiodophthalic
anhydride*
A mixed melting point of the two substances
showed no depression*
Thus, it is evident that the conden­
sation, at least In part, went para to the hydroxyl group*
On crystallization of the product that had not been
first crystallized as the potassium salt, a product melting
at 202°-204° was obtained*
At first it was believed to have
been the isomeric acid, that is 2-(0-anisoyl)-4,5-diiodobenzoic acid*
However, a neutral equivalent Indicated that
this was not the case*
The value of the neutral equivalent
42:
was found to be 235 and 236; had the compound been the
isomeric benzoylbenzoic acid, the value should have been
508#
Despite the fairly sharp melting point (202-204°),
this other product was apparently impure 4,5-diiodophthalic
acid, since it Itself has a neutral equivalent of 209#
Therefore, It may be said that the condensation went para
to the hydroxyl in the phenol nucleus and not ortho as
had been anticipated*
This is In agreement with some work
done by Bentley86 and others, who have found that in con­
densing phthalic anhydride with the cresols the condensa­
tion occurred ortho to the hydroxyl only when boric acid
was used as the condensing agent*
When aluminum chloride
was used the cresol nucleus was substituted In the para
position to the hydroxyl group*
Both the 4,5-diiodo- and 3,4,6-triiodophthalie anhy­
drides have been condensed with m-eresyl methyl ether*
The
product isolated from the condensation of the former anhy­
dride has been identified and is described In the experi­
mental section*
On the other hand the condensation reac­
tion of the 3,4,6-triiodo anhydride has not been thoroughly
Investigated*
However, it was found that these two com­
ponents could be made to condense*
The reaction was carried
out in boiling carbon disulfide for twenty-two hours*
Aluminum chloride was used as the condensing agent; 10$ in
Bentley, Gardner, and Weizmann, J* Chem* Soc* 91, 1626
(1907)
43
excess of two moles was used to every mole of the iodinated
phthalic anhydride.
On working up the reaction mixture in
the customary manner, the crystalline potassium salt of the
condensation product was deposited, in part at least, on
cooling the potassium carbonate extract#
The organic acid
obtained from it melted from 226-236° after one crystalliza­
tion from alcohol#
It seems probable that the large range
in melting point is do to an isomeric mixture#
On recrys-
tallization of the product from y ,y diehlorodiethyl ether
a small amount of a white crystalline product melting at
258-259° was obtained, not enough, however, to permit fur­
ther characterization#
Probably this reaction could have been developed but
work was not continued for two reasons#
In the first place
a mixture of isomers was obtained and secondly, several
additional steps would have been necessary in order to ob­
tain an anthraqulnone derivative having an alpha hydroxyl
group#
Several other condensation reactions involving 4,5diiodophthalic anhydride have been attempted such as the
condensation with benzene, anisole, and resorcinol monoand dimethyl ethers#
The products of these condensations
are reported in the experimental section as are condensa­
tions of the other iodinated phthalic anhydrides#
The benzoylbenzoic acids obtained, quite frequently,
on attempted crystallization came down as oils#
This was
44
particularly true If too much water was added to the sol­
vent being used in the crystallization*
In some cases it
was found that these products could be crystallized better
from anhydrous solvents*
However, dilute alcohol, dilute
dioxane, on a mixture of both was most frequently used,
with care being taken not to add too much water*
V. Search for the Proper Solvent for the Friedel-Crafts
Reaction
Various inert liquids have been used as solvents for
the Friedel-Crafts reaction*
Probably the one most common­
ly used is carbon disulfide*
This solvent was used in many
of the present condensations with some success*
However,
since the iodinated phthalic anhydrides were not appreciably
soluble in carbon disulfide, it was thought that better
yields could be obtained if a better solvent were avail­
able*
As previously mentioned Ullmann and Schmidt83 found
that tetraehloroethane was an excellent solvent for the
Friedel-Crafts reactions*
Nevertheless,,tetraehloroethane
did not prove to be a satisfactory solvent for use in this
investigation*
In the first place the iodinated phthalic
anhydrides were not appreciably soluble and secondly there
was always a considerable amount of decomposition*
Craves
and Adams8® also observed that in using tetraehloroethane
as a solvent, in the attempted condensation of 3,6-dimethoxyphthalic anhydride with the cresols, there was always
an appreciable amount of decomposition*
45
Another objection to using carbon disulfide as a sol­
vent was that the reaction could not be carried out at a
high enough temperature*
Since it was desired to run the
condensations at about 70-80°, it was decided to prepare
some olefin free, high boiling, petroleum ether and test
its applicability as a solvent for these reactions*
The
olefin free ether was prepared by extracting the crude
ether with 15$ oleum, followed by two extractions with 95$
sulfuric acid*
During the acid extractions, the ether-
acid mixture was stirred for about two hours*
After wash­
ing with water, drying over sodium carbonate and distill­
ing, the fraction boiling at 67-87° was available for use
as a solvent*
A mixture of 5 g* (0*012 mol) of 3,4-dilodophthalic
anhydride, 2*6 g* (0*020 mol) of p-chlorophenol, 6*6 g*
(0*040 mol) of anhydrous aluminum chloride, and 45 cc* of
olefin free petroleum ether was heated at the about 70®
for twenty-four hours*
The petroleum ether was not a very
good solvent for the reaction and the only product that
could be isolated from the reaction mixture was 3,4-di­
lodophthalic acid*
The possibility of using benzene as a solvent in the
condensation of these iodinated phthalie anhydrides with
various phenolic compounds was then experimentally tested*
Even though benzene itself was found to condense readily
with the iodinated phthalie anhydrides, it seemed probable
46
that if a phenolic compound were present, condensation
would take place with it at a faster rate than with ben­
zene and thus, the latter substance could be used as a sol­
vent*
Two condensations were run to test this possibility*
In both cases the reaction mixture consisted of 4,5-diiodophthalic anhydride, p—chlorophenol, anhydrous aluminum
chloride and benzene*
One mole of aluminum chloride was
used for every mole of the phenol derivative and two moles
for every mole of the diiodo anhydride*
The reactions were
run at the boiling point of benzene for ten hours*
In the
one case an 80$ yield of 2-benzoyl-4,5-diiodobenzoic acid
was obtained, while in the other a 70$ yield of the same
benzoybenzoic acid resulted*
It seems likely that one ex­
planation for obtaining no condensation with p-chlorophenol,
may have been that the formation of the aluminum chloride
salt, with the phenolie hydroxyl, cuts down its reactivity
considerably*
Furthermore, since the 4,5-diiodophthalic
anhydride could not be condensed with p-chlorophenol under
other experimental conditions it seems that this was not a
fair test for determining the use of benzene as a solvent*
Possibly had 5,4-diiodophthalic anhydride been used there
might have been a different result*
Attention was next turned to the possible use of o-dichlorobenzene as a solvent for these reactions*
It was first
47
decided to see If the o-dichlorobenzene itself would con­
dense with the iodinated anhydrides*
This was found to be
the ease, the 4,5- and 3,4-diiodophthalic anhydrides both
condensing with it*
The experimental details are given in
another section of this dissertation*
A reaction was then
run to test the possibility of condensing 4,5-diiodophthalic
anhydride with anisole using the dichlorobenzene as a sol­
vent*
The reaction was run for six hours at 80°•
A 90#
yield of 2-(p-anisoyl)-4,5-diiodobenzoic acid was obtained*
Thus, the reaction with anisole occurs at a much faster
rate than that with o-dichlorobenzene*
Nevertheless, its
applicability as a solvent seemed to be somewhat limited*
Thus, in an attempted condensation of tetraiodophthalic
anhydride and anisole, using o-dichlorobenzene as a solvent,
an appreciable amount of decomposition occurred*
This, how­
ever, may have been do to the high temperature (100°) at
which the reaction was run*
Furthermore, in condensations
using 1,2,4-triehlorobenzene as the solvent decomposition
likewise occurred*
In those condensations involving p-chlorophenol and
p**cresol it was found that excess of these reagents proved
to be the best solvent*
Only in one other case, when tetra-
chloroethane was used as the solvent, could any indication
of a condensation between the iodinated anhydrides and pchlorophenol be detected*
A small quantity of product melt­
48
ing at 250-267°, being extracted from the tarry residue
remaining after the removal of the tetrachloroethane, in
a reaction involving the 3,4-diiodo anhydride and p-chlorophenol*
VI*
Conversion of the Benzoylbenzoic Acids into
Anthraquinone Derivatives
In the cyclization of 2-(2-hydroxy-5-chlorobenzoyl)—
benzoic acid the chlorine atom is eliminated and quinizarin results rather than l-hydroxy-4—chloroanthraquinone,
if 95$ sulfurie acid is used*
However, if 100$ sulfuric
acid is employed the chlorine is not effected and the
ehlorohydroxyanthraquinone is readily obtained*
Further­
more, since iodine was evolved in the attempted cycliza­
tion of 2-benzoyl-3,4,5,6-tetraiodobenzoic acid using 95$
sulfuric acid, it was at first believed that in the dehy­
dration of the isomeric benzoyl-diiodobenzoic acids to the
anthraquinone derivatives, it would be necessary to employ
100$ sulfuric acid*
However, it was later found that this
was not the case as the 1,2— and 1,4—diiodoanthraquinones
were readily prepared from the 2-benzoyl-3,4 (?)-diiodobenzoic and 2-benzoyl-3,6-diiodobenzoic acids, respectively,
on treatment with 95$ sulfuric acid*
Apparently the ex­
planation for the loss of the chlorine atom in the first
instance is that it has been activated by the ortho car­
bonyl and the para hydroxyl.
A similar case has already
been mentioned in the preparation of 3-bromoalizarin*
Here
the alpha bromine, in l,3-dibromo-2-hydroxyanthraquinone,
49
is probably activated by the ortho carbonyl and hydroxyl
groups, for on mild treatment with alkali it is readily
eliminated*
Ring closure of 2-benzoyl-4,5-diiodobenzoic acid readily
gave the 2,3-dliodoanthraquinone*
This compound had the
same melting point as a diiodoanthraquinone, of unknown
structure, previously prepared by Eckert and KlingerJ4 It
had been prepared by them through the action of aluminum
chloride upon benzoyl-tetraiodobenzoic acid*
The identity
of these anthraquinones is reasonable as the Iodine atoms
in the 1 ,4-positions are eliminated most readily*
Some difficulty was experienced in the cyclization of
the two isomeric benzoyl triiodobenzoic acids*
As In the
other cyclizations 100$ sulfuric acid was first employed to
effect ring closure*
In one instance In an effort to pre­
pare the 1 ,2 ,4-triiodoanthraquinone in a considerable quan­
tity 11 g* of the high melting benzoylbenzoic acid (the
condensation of 3,4 ,6-triiodophthalic anhydride with benzene
gives two isomers) was added to 300 g* of 100$ sulfuric
acid*
The reaction was carried out at 100-110° for four
hours*
stirred*
During this time the mixture was continuously
The sulfuric acid used contained a small amount
of sulfur trioxide in order to take care of the water form­
ed in the reaction*
After permitting the reaction mixture to stand over­
night, the flask was placed in an ice bath and at the same
50
time small pieces of ice were dropped into it*
observed that iodine was being evolved*
It was soon
This never occurr­
ed in dehydrating the benzoyl-diiodobenzoic acids when the
reaction mixture was worked up in this manner*
A much better
method of isolating the desired product, however, was to
pour the reaction mixture directly into ice*
Since iodine was so easily evolved from the reaction
mixture, involving the cyclization of the benzoyltriiodobenzolc acid, it seemed certain that 100$ sulfuric acid
was necessary in the preparation of 1,2,4-triiodoanthraquinone*
Nevertheless, a dehydration reaction of both iso*
meric acids was tried using 95$ sulfuric acid*
The reac­
tions were run in the same manner and under the same temperature
conditions as before*
Iodine was not evolved but on the other
hand, on working up the reaction mixtures only a small trace
of substance insoluble in sodium hydroxide, was produced
in each case*
Neither substance was identified because not
much was obtained*
An effort has also been made to prepare 1,2,3,4-tetraiodoanthraquinone by the cyclization of 2-benzoyl-3,4,5,6tetraiodobenzoic acid*
It was found that at 145°, even
though 100$ sulfuric acid was used iodine was evolved*
This
loss of Iodine on attempted cyclization of the benzoyltetraiodobenzoic acid confirms a previous report*64
An
effort to effect the dehydration at a lower temperature was
also unsuccessful*
Iodine was not evolved, but neither had
51
ring closure been effected.
It thus seems that tempera­
ture is an important factor in the determination of whether
or not iodine is evolved.
In the attempted cyclizations of the more complex io­
dinated benzoylbenzoic acids* such as 2-(Ssk-hydroxy-5-chlorobenzoyl)-3,4(?)-diiodobenzoic acid, much more difficulty
has been experienced.
This is not suprising, however, as
many investigators have observed this difficulty in those
benzoylbenzoic acids which contain an ortho or para direct­
ing group meta to the position at which the condensation
must take place#.8®
This difficulty is especially pronounc­
ed In the case of phenol derivatives because of the ease
with which sulfonation takes place#
In an effort to prepare 5,6-diiodoquinizarin from the
benzoylbenzoic acid mentioned above, the cyclization was to
be effected by use of 95% sulfuric acid.
been tried twice.
The reaction has
On each occasion it was carried out at
110° for three hours.
Practically all the product obtain­
ed from the reaction mixtures was soluble in water but was
readily salted out with sodium chloride.
However, a trace
of product insoluble in water and in potassium carbonate
solution was isolated; in such a small quantity, however,
that nothing could be done with It#
A possibility of cutting down the tendency to sulfo­
nate seemed to be the use of some reagent as a diluent for
the_gnlfuric acid.
This scheme has been tested and it was
52
found that the tendency to sulfonate was practically
eliminated but also the cyclization to the anthraquinone
did not occur in appreciable amounts#
The dehydrating
mixture employed consisted of 25 cc# of syrupy phosphoric
acid, 50 cc# of concentrated sulfuric acid and a trace of
boric acid#
About 2 g# of 2-(2-hydroxy-5-chlorobenzoyl)-
3,4(?J-diiodobenzoic acid was added to this mixture and
heated at 140° for eight hours#
A small amount of alkali
insoluble material was isolated from the reaction mixture#
After one crystallization from glacial acetic acid it melt­
ed at 265-275°#
Not enough was obtained to permit further
work#
As it was realized that benzylbenzoic acids were more
easily cyclised than the corresponding benzoylbenzoic acids,
it was decided to attempt a reduction of the carbonyl to a
methylene group, and thus convert the iodinated benzoyl­
benzoic acids into the corresponding benzylbenzoic acids
before attempting cyclization#
The practicability of the
scheme was first tested on 2-benzoyl-4,5-diiodobenzoic acid#
It seemed probable that the desired transformation could be
brought about by a mild alkaline reduction, such as report­
ed by Fieser#7
The reduction was brought about by adding
r-therTodinated benzoylbenzoic acid to a mixture of 2 g# of
~f87) Fieser and Jones, J# Am# Chem# Soc# 60 1940 (1938)
53
zinc dust and 150 cc* of 2 N sodium hydroxide*
was run for forty-one hours at 100°*
The reaction
The product isolated
from the mixture melted at 113—114°, after three crystalli—
zations from dilute methyl alcohol*
This melting point
corresponds to that of o-benzylbenzoic acid*
Therefore,
not only was the carbonyl reduced to a methylene group but
both iodine atoms were removed as well*
0
Zn
NaOH
COOH
2-benzoyl-4, 5<
diiodobenzoic
acid
2-benzylben<
zoic acid
In the attempted cyclization of 2-(2-hydroxy-5-chlorobenzoyl)-3,4(?)-diiodobenzoie acid it was evident that the
rate of sulfonation was much greater than the rate of de­
hydration and as a result the anthraquinone derivative was
not obtained*
However, it was anticipitated that if the
rate of sulfonation could be cut down there would be a
greater possibility of obtaining the desired compound*
First of all it seemed likely that if the hydroxyl group
were methylated or acetylated the tendency for sulfonation
would be greatly reduced and thus the possibility of cycli­
zation enhanced*
tion*
The first approach was through methyla-
Considerable difficulty was encountered, in fact
it was not found possible to methylate the hydroxyl In an
54
aqueous medium*
However, the methyl ether has been prepar­
ed by another procedure, the details of which are given in
the experimental section*
It is probable that the explana­
tion for this difficulty in the methylation of the hydroxyl
group is chelation, for when the hydroxyl group occupied
the para position to the carbonyl group no difficulty was
experienced*
Because of this difficulty in methylation this
approach was abandoned*
The acetylation of the hydroxyl
gave much better results and a considerable quantity of the
acetylated product was prepared#
It is believed that a more feasible approach to the
successful t ermination of the ring closure reaction would
be a method, whereby the predominating sulfonation reaction
would be prevented entirely*
Therefore, it was decided to
attempt bromination of this benzoylbenzoic acid by a method
e o
similar to that used by Jacobson and Adams*
Since the
bromine atom could take up only one position, that of ortho
to the hydroxyl which was the only one available for sulfo­
nation, the benzoylbenzoic acid could not then be readily
sulfonated in the subsequent dehydration reaction*
Another
desirability of. having a bromine atom in the position ortho
to the hydroxyl, is that the position where the ring closure
must take place would then be activated by the para direct­
ing influence of the bromine atom and the ortho directing
^power of the chlorine*
In the cyclization reaction 95%
sulfuric acid should be used*
Thus, it would be antioif*:.
"(88) Jacobson and Adams, J* Am* Chem* Soc* 46, 1312 (1924)
55
pated that the quinizarin derivative should result*
Schematically the method involved is as indicated*
I
0
OH
This
T
OCX) ^ coo
Cl
Cl
i
o
OH
0
OH
series of reactions will be investigated later*
VII
Miscellaneous Reactions
Theoretically there are four possible homonuclear
diiodoanthraquinones*
Three of these, namely the 1,2**,
1,4-, and 2,3-diiodo derivatives were readily prepared by
the respective condensation of £,4~, 3,6-, and 4,5-diiodophthalie anhydrides with benzene, followed by the subse­
quent dehydration of the intermediate benzoylbenzoic acids*
However, since the 3,5-diiodophthalic anhydride was not ob­
tained in the partial iodination of phthalie anhydride, and
as it had not been previously synthesized, it was not avail­
able for the synthesis of 1,3-diiodanthraquinone*
Therefore,
another approach to the preparation of this remaining iso­
meric homonuclear diiodoanthraquinone was investigated*
The
method employed involved the condensation of phthalie anhy­
56
dride with m-diiodobenzene, followed by cyclization to the
anthraquinone derivative*
\
/
G-oldberg33, had previously 33m —
y
I
j
HgSO^
COOH
0
thesized 1,3-dichloroanthraquinone in this manner and thus
the possibility of success seemed rather good*
Three attempts were made to effect the first condensa­
tion, that is to synthesize the 2,4-diiodobenzoylbenzoic
acid*
In one instance carbon disulfide served as the sol­
vent for the reaction, the reaction being run for twentynine hours at the boiling point of this liquid*
As the
original products were recovered on working up the reaction
mixture, it seemed that possibly had a higher temperature
been employed, the condensation might have been effective*
Therefore, in the second attempt a higher boiling liquid,
tetrachloroethane, served as the solvent*
A mixture of 2*2
g. (0*015 mol) of phthalie anhydride, 4.4 g* (0*013 mol) of
m-diiodobenzene and 50 cc* of tetrachloroe thane was heated
up to 120°, at which temperature 4*3 g* (0*032 mol) of an-
57
hydrous aluminum chloride was slowly added#
Iodine was
evolved as soon as the condensing agent was added#
Never­
theless, the reaction was continued for six hours at 145°*
Prom the reaction mixture, an alkali insoluble product was
obtained#
It was not further investigated as it was known
not to be the desired compound*
Therefore, since it was
impossible to effect condensation at 45° and as decomposi­
tion of one of the components set in at 120°, an effort was
made to choose some intermediate temperature#
In this
attempt nitrobenzene was used as the solvent and the reaction
was run for three hours at 85° •
Again no condensation pro­
duct was obtained#
During the course of this investigation the question
of other possible uses for these partially iodinated phthalie
anhydrides arose#
Time, however, did not permit a thorough
Investigation of their many possibilities#
The one possi­
bility that seemed to be of some physiological Importance
and thus which attracted the most attention was the pre­
paration of some iodinated aspirin derivatives*
Their syn­
thesis was to be effected by the conversion of the iodinated
phthalie anhydrides into the corresponding phthalimides,
followed by the transformation of the latter, through the
Hofmann reaction, Into the iodinated anthranilic acids,
which on diazotisation and acetylation would yield the de­
sired aspirin derivatives#
In this present work the 3,4-
and 3,6-diiodophthalic anhydrides have been converted into
58
a
I
\
COOH
‘]
0
/
NHS:
0 0
COOH
COOH
OH
0-C-CHfi
0
the corresponding phthalimides in excellent yield*
Fur­
thermore, the 3,6-diiodophthalimide has been further trans«
formed into the 3,6-diiodoanthranilic acid*
continued on this phase of the investigation*
Work is being
59
EXPERIMENTAL
I.
Preparation of Iodinated Phthalie Anhydrides
These anhydrides were prepared by direct iodination by
a modification of the procedure of Pratt and Perkins*4 A
typical run is described*
A mixture of 76 g* (0*51 mol) of
phthalie anhydride, 127 g* (1 mol) of iodine, and 270 g* of
50$ oleum was heated In a 5 1* round-bottom flask for six
days in an oil bath, the temperature of which fluctuated be­
tween 70-80°•
At the end of this heating period the reaction
mixture was poured into a large evaporating dish and exposed
to the air overnight*
About 250 g* of ice was then added*
After it had melted and the mixture had been thoroughly
stirred the iodinated phthalie anhydrides, which precipi­
tated, were filtered off*
It was found necessary to wash
the precipitate with a sodium sulfite solution In order to
remove the free iodine; a yellow solid then remained*
This yellow residue was treated with 250 cc* of a warm
20$ sodium hydroxide solution*
It did not dissolve readily
but excessive heating had to be avoided because hot alkali
decomposes the partially iodinated phthalie anhydrides*
If
the residue remaining after a short time appeared to be
white, the addition of more water dissolved it; this was
probably a mixture of the sodium salts of the desired com­
pounds*
On the other hand when the residue was yellow a
little more 20$ sodium solution had to be added*
60
In this case, all the precipitate dissolved after more water
had been added*
After filtration, in order to remove dirt,
ferric oxide, etc*, the solution was made neutral to litmus
with acetic acid*
The total volume then measured about
450 cc*
The neutral solution of the sodium salts of the iodi­
nated phthalie acids, thus obtained, was fractionally pre­
cipitated by successive additions of acid into four fract­
ions.
The first fraction (1) was obtained by adding 15 cc*
of 6 N acetic acid to the previously heated neutral solution,
whereupon after stirring and gradually cooling over a two
hour period the monosodium 4,5-diiodophthalate, contaminat­
ed with the monosodium salt of 3,4,6-triiodophthalic acid,
crystallized from the solution.
After removal of the cry­
stalline precipitate the filtrate was reheated, treated with
18 cc* of 6 M acetic acid, and on stirring and cooling over
a two hour period the second crystalline fraction settled
out*
This fraction, i*e. fraction 2, also proved to be
monosodium 4,5-diiodophthalate, contaminated with monosodium
3,4,6-trIiodophthalate.
The filtrate from the second frac­
tion was again reheated, this time being treated with 20 cc.
of commercial hydrochloric acid, after which it yielded on
stirring and gradual cooling over the customary two hour
period another fraction.
This fraction, i.e. fraction 3,
consisted essentially of the monosodium salt of 3,4-diiodophthalic acid having present as impurities the monosodium
j3 ,4 ,6 -triiodophthalate and 3,6-diiodophthalate respectively.
61
Upon removal of the third fraction the filtrate was heated
on the steam hath for ten hours with 80-100 cc. of commer­
cial hydrochloric acid*
It was necessary to heat the solu­
tion for a long time in order to completely precipitate the
3,6-diiodo derivative as the anhydride.
Fraction 4 being
largely 3,6-diiodophthalic anhydride contaminated with
3.4-diiodophthalic acid.
The first two fractions, 1 and 2, were treated separate­
ly with 50 cc. of commercial hydrochloric acid and heated on
the steam bath for from 5 to 6 hours.
The monosodium salt
of 4,5-diiodophthalate present in both fractions was thus
converted into 4,5-diiodophthalic acid.
Likewise the mono­
sodium salt of 3,4,6-triiodophthalate, also present in both
fractions, was converted into 3,4,6-triiodophthalic anhy­
dride.
These two fractions were then combined and the total
precipitate after drying and powdering was extracted with
benzene.
The benzene extraction removed the 3,4,6-triiodo-
phthalic anhydride (extract 1) but did not dissolve the
4.5-diiodophthalic acid (residue 1).
Fraction 3 was redissolved in enough 20% sodium
hydroxide to effect solution, after which it was filtered
by suction, made neutral to litmus with acetic acid and
heated to about 80°.
The addition of 30 cc. of glacial
acetic acid failed to precipitate,after stirring and cool­
ing for two hours, any of the 3,4,6-triiodo derivative known
to be present in fraction 3 as an impurity.
After reheating the solution to about 80° it was treated
6a
with 15 cc• of commercial hydrochloric acid*
Upon stirring
0-hd. gradual cooling for two hours the precipitate (A) which
formed was filtered off*
It was largely mono sodium 3,4-di-
iodophthalate contaminated with the 3,4,6-triiodo derivative
and possibly small amounts of the 3,6-diiodo derivative*
The
filtrate from precipitate A was reheated to 80° and treated
with another 15 cc* portion of commercial hydrochloric acid,
this being followed by a two hour stirring and cooling period*
The precipitate (B) which formed was also the monosodium salt
of 3,4-diiodophthalic acid contaminated with the 3,6-diiodo
derivative and possibly some 3,4,6-triiodo derivative*
Both of these precipitates, A and B, were separately
treated with 150 cc* of commercial hydrochloric acid and
heated on the steam bath for ten hours*
The precipitates
were stirred frequently and care was taken to break up all
the lumps of product*
Here again this treatment was carried
out in order to liberate the 3,4-diiodophthalic acid and to
convert any 3,4,6-triiodo and, or, any 3,6-diiodo deriva­
tives present into their respective anhydrides.
During
this treatment both precipitates, A and B, turned yellow
indicating the presence of the 3,4,6-triiodo and 3,6-diiodo
derivatives*
The precipitates, A and B, after the above treatment,
and fraction 4 were dried and powdered*
All three were
separately extracted with benzene, using a Soxhlet extractor*
In extracting precipitate A the residue left, i*e* residue 2,
was 3,4-diiodophthalic acid, while the extract, i*e* extract
2, was largely 3,4,6-triiodophthalic anhydride contaminated
with, some 3,6-diiodophthalic anhydride*
Extraction of pre­
cipitate B with benzene also left a residue, i*e* residue
2, which was 3,4-diiodophthalic acid, while the extracted
product, i*e* extract 3, was 3,6-diiodophthalic anhydride
with the 3,4,6-triiodo derivative present as a possible im­
purity*
Furthermore, some 3,4-diiodophthalic acid remained
after extracting fraction 4 with benzene.
The benzene ex­
tract, i.e. extract 4, was 3,6-diiodophthalic anhydride.
The benzene extracts were then worked up in the follow­
ing manner.
Extracts 1 and 2 were combined and the benzene
was removed by distillation.
Likewise the benzene was
distilled from extracts 3 and 4.
The solid product isolated
from extracts 1 and 2 was dissolved in just enough 20$
sodium hydroxide solution to dissolve it, after which it
was filtered with suction and made neutral to litmus with
acetic acid.
The neutral solution was then heated to about
80° and treated with 15-20 cc. of glacial acetic acid.
On
stirring and gradual cooling for two hours a crystalline
precipitate settled out which was monosodium 3,4,6-triiodo­
phthalate, precipitate C.
The residues remaining after removal of the benzene
from extracts 3 and 4 were similarly treated.- That is,
both were separately dissolved in as small a volume of 20$
sodium hydroxide as possible, filtered with suction, heated
to 80° and made neutral to litmus with acetic acid.
A 25-30
__o_c. portion of glacial acetic acid was added to each fraction
64
In neither case did a precipitate form on stirring and cool­
ing for two hours.
This apparently indicated the absence of
any 3,4,6-triiodo derivative at least in any quantity, for
had it been present the acetic acid would probably have pre­
cipitated it as the monosodium salt.
Both solutions were then treated with 30 to 50 cc. of
commercial hydrochloric acid and heated on the steam bath
for ten hours.
The two precipitates obtained were combined.
This precipitate, D, was dried, powdered and extracted with
benzene.
A considerable residue remained.
It was 3,4-diio­
dophthalic acid and was thus added to residue 2.
On concentrating the benzene extract from precipitate
D, the 3,6-diiodophthalic anhydride was obtained.
different fractions were isolated.
24 g. melted at 229-231°•
Three
The first fraction of
The other two fractions had large
melting ranges which could not be much improved by recrystalli­
zation.
In this particular run 24 g. of pure 3,6-diiodophtha-
lic anhydride and 25 g. of impure product was obtained, as
compared to a yield of 56.5 g. of product reported by the
original investigators.
Precipitate G, which was obtained from extracts 1 and 2,
was treated with 20 cc. of commercial hydrochloric acid and
heated on the steam bath for ten hours.
This treatment con­
verted the monosodium 3,4,6-triiodophthalate into the corre­
sponding anhydride, E.
of it dissolved.
On extraction of E with benzene all
Upon concentration of the benzene solution
the 3,4,6-triiodophthalic anhydride crystallized.
Only 9.7 g.
65
of product melting at 227-228° was obtained as compared to
20 go isolated by Pratt and Perkins.
Residue 1 was crystallized from hot water.
Acetic acid
could have been used and it is to be preferred to water be­
cause of the low solubility of the 4,5-diiodophthalic acid
in the latter.
Approximately 17 g. of pure 4,5-diiodophtha­
lic acid was obtained.
originally.
This checks the value reported
The acid was converted into the corresponding
anhydride by refluxing in dry benzene and a slight excess
of the calculated amount of acetic anhydride.
About 14 g.
of the 4,5-diiodophthalic anhydride was isolated.
It melt­
ed at 214-215°.
Residue 2 was crystallized from dilute acetic acid.
A
yield of 42 g. of product melting at 210-211° was obtained*
The original investigators reported a yield of 52.5 g.
This diiodo acid was converted into the corresponding anhy­
dride in a manner identical with that followed for the
preparation of 4,5-diiodophthalic anhydride.
In this run
34 g. of the pure 3,4-diiodophthalic anhydride melting at
195-196° was obtained.
This preparation has also been run on a large scale*
Prom 780 g* (5.2 mol) of phthalie anhydride, 1270 g. (10
mol) of iodine, and 2500 g. of bQ/o oleum after heating for
six days at 70-80° were obtained 100 g. of 4,5-diiodophthalic anhydride, M.P. 215-216°; 147 g. of 3,4,6-triiodophthalic
anhydride, M.P. 220-223° (further crystallization raised this
'to 226-227°); 325 g. of 3,4 diiodophthalic anhydride, M.P.
66
194 196 , and 395 g* of 3,6-diiodophthalic anhydride, M.P.
228-231°.
II.
Condensation of Benzene with the Iodinated Phthalie
Anhydrides
fj~_Benzoyl-4,5-diiodobenzoic acid.
A mixture of 5 g. (0.0125 mol) of 4,5-diiodophthalic
anhydride, 3.65 g. (0.0274 mol) of anhydrous aluminum
chloride (which was added in small portions), and 40 cc.
of benzene was refluxed for twenty-four hours.
The reaction
mixture was stirred continuously during this time.
It was
then acidified with commercial hydrochloric acid and the ex­
cess benzene, steam distilled.
The water-insoluble residue
was removed by filtration and treated with hot potassium
carbonate solution which left only a trace of undissolved
material.
After filtering through a hot water funnel and
then cooling the alkaline solution, the potassium salt of
the benzoylbenzoic acid crystallized.
redissolved in hot water and acidified.
It was filtered off,
On crystallizing
the precipitate from dilute alcohol 4.8 g. of product melt—
X
Ing at 244-245° was obtained. This corresponds to an 80/b
yield.
Neutral Equivalentt
Subs., 0.1275 g.; required 6.85 cc. of 0.0387 N NaOH soln.
0.1063 g.; required 5.70 cc. of 0.0387 N NaOH soln.
X
_ All melting points reported in the experimental are corrected.
67
Calcd. for C14H803IS: neut. equiv., 478
Pound:
neut. equiv., 481, 481
2-Benzoyl-3,4(?)-diiodobenzoic acid*
The procedure was identical to that followed in the
previous condensation.
From 5 g. (0.0125 mol) of 3,4-di­
iodophthalic anhydride, 5.8 g. (0.028 mol) of anhydrous
aluminum chloride and 130 cc. of benzene there was obtained
4.5 g. of pure product melting at 223-224°.
ponds to a yield of QQ%%
This corres­
This product was purified initial­
ly by crystallizing its potassium salt and finally by cry­
stallizing the free acid from dilute alcohol.
Theoretically
it was possible to obtain two isomers in this reaction but
only one was obtained.
This condensation was also run for
a period of nine hours instead of twenty-four.
A yield of
10% of the condensation product resulted.
Neutral Equivalent:
Subs., 0.1128 g.; required 2.81 cc. of 0.0842 N NaOH soln.
0.1564 g.; required 3.88 cc. of 0.0842 N NaOH soln.
Calcd. for C14H803I3 : neut. equiv., 478
Found:
neut. equiv., 476, 478
2-Benzoyl-5,6»«diiodobenzoic acid.
Again the procedure was the same as that followed in the
previous condensations.
From a mixture of 5 g. (0.0125 mol)
of 3 ,6-diiodophthalic anhydride, 3.8 g. (0.028 mol) of an­
hydrous aluminum chloride, and 130 ml. of benzene, 4 g. of
product melting at 218-220° v/as obtained.
In this in­
stance it was not found possible to effect purification by
68
crystallizing the potassium salt of the "benzoylbenzoic acid®
The free organic acid was purified by crystallizing from
dilute alcohol®
A yield of 70% was obtained®
Neutral Equivalent:
Subs®, 0.1433 g®; required 3.53 cc. of 0.0842 N NaOH soln.
0.1183 g®; required 2.95 cc. of 0.0842 N NaOH soln.
Calcd. for Cx4He03I3: neut. equiv., 478
Found:
neut® equiv®, 482, 476
2-Benzoyl-5,4,6- and 5,5,6-triiodobenzoic acids.
In condensing 3,4,6-triiodophthalic anhydride with
benzene 18*1 g® (0®136 mol) of anhydrous aluminum chloride
was added in small portions to a mixture of 32.6 g. (0.061
mol) of the triiodo anhydride and 200 cc® of benzene.
The
reaction mixture was stirred continuously during the twentyone hour period of refluxing, followed by acidification and
steam distillation.
The water-insoluble material remaining
was extracted with hot potassium carbonate solution®
After
filtering, the solution was cooled but the potassium salts
of the condensation products failed to crystallize.
On
acidifying with hydrochloric acid they were precipitated.
The precipitate obtained was heated on the steam bath
for fourteen hours in concentrated hydrochloric acid.
This
was done in order to convert any 3,4,6-triiodophthalic acid
present into the corresponding anhydride which could then be
removed by benzene extraction.
The precipitate after this
treatment still appeared white and hence it was taken to
indicate that little, if any, 3,4,6-triiodophthalic anhy­
69
dride could be present.
was omitted.
Therefore, the benzene extraction
The total weight of the dried product at this
point was about 37 g.
Refluxing this product for forty-five
minutes in 400 ml. of 95% alcohol left an insoluble resi­
due of 20 g., which melted at 252-262°C.
This residue was
extracted with hot dioxane leaving a residue of 2.4 g. of
inorganic material.
On concentrating and cooling the dioxane
solution it gave 13.5 g. of product melting at 257-258°; and
2.8 g. melting at 254-257°•
On concentrating and cooling,
the alcohol extract gave 10.7 g. of product melting at 224—
227°, and further smaller fractions which were practically
pure.
A total of 16.3 g* of the high melting isomer and
14.4 g. of the low melting product was obtained.
The purest
samples of these two acids melted at 257-258° and 225-227°,
respectively.
The yield based on impure product was almost
quantitative but based on the quantity of pure products ob­
tained it corresponds to 82^.
Neutral Equivalent:
Subs.,
0.1043 g.;
required 4.44 cc. of 0.0387 N NaOHsoln.
0.1178 g.; required 5.06 cc. of 0.0387 N NaOH soln.
Calcd. for C14H703I3 (M.P. 257-258°)r neut. equiv. 604
Found:
neut. equiv. 607, 602
Subs.,
0.1106 g.j
required 2.19 cc. of 0.0842 N NaOH soln.
0.1473 g.;
required 2.85 cc. of 0*0842 N NaOH soln.
Calcd. for C14H703I3 (M.P. 225-227°):
Found:
neut. equiv. 604
neut. equiv. 600, 613
''Subs.. , 0.1698 g.; required 7.19 cc. of 0.0387 N NaOH soln.
Calcd. for Ci 4H7Os I8 (M.P. 245-252°)r neut. equiv. 604
Found:
III*
neut* equiv., 610
Preparation of Some Simple lodinated Anthraquinones
2_,5-Diiodoanthraquinone*
To 40 g. of 100$ sulfuric acid
at 140° was added gradual­
ly over a thirty minute period with stirring, 2 g. of 2benzoyl-4,5-diiodobenzoic acid.
70-80
The mixture was cooled to
and diluted slowly by adding small pieces of ice*
The
precipitate formed was coagulated by heating on the steam
bath*
The mixture was then filtered and the precipitate
warmed with aqueous potassium carbonate solution to remove
any unreacted acid*
The insoluble portion was crystallized
from a benzene-acetic acid mixture.
ed at 291-292°*
The yellow crystals melt­
The yield was1.6 g. or 80$ of the theoreti­
cal amount.
Analysis:
Subs., 0.0722 g.; Agl 0.0735 g.
0.0910 g.j Agl 0.0931 g.
Calcd. for C14H60SIS; I, 55.21
Found:
I, 55.0, 55.3
1,2-Di iodoanthraqulnone *
This was prepared in a manner similar to that used in
preparing the 2,3-isomer.
It crystallized In orange-red
plates from a benzene-toluene mixture and melted at 236-237°.
It was also found possible to convert the 2-benzoyl-3,4(?)diiodobenzoic acid into 1,2-diiodoanthraquinone by using 95$
sulfuric acid.
To 50 cc. of concentrated sulfuric acid was
71
added 1 g# of the benzoylbenzoic acid.
It was added slowly
and the sulfuric acid solution was stirred continuously*
The reaction was run at 110-115° for approximately two
hours*
After permitting to cool the sulfuric acid mixture
was poured into ice*
The diluted reaction mixture was then
heated on the steam bath for some time in order to coagulate
the precipitate*
It was removed by filtration, extracted
with dilute sodium hydroxide solution, dried, and crystalliz­
ed as already stated.
Only about a 30$ yield was obtained
in this second method*
Analysis:
Subs*, 0*0614 g.j Agl 0*0632 g*
0.1022 g.; Agl 0.1052 g*
Calcd. for C14H60SIS :
Pound:
I, 55.21
I, 55.6, 55.63
1,4-Diiodoanthraquinone.
It was found possible to convert 2-benzoyl-3,6-diiodobenzoic acid Into this anthraquinone derivative by dehydra­
tion either with 100$ or with 95$ sulfuric acid.
Prom
heating 2 g. of 2-benzoyl-3,6-diiodobenzoic acid with 75 g.
of 100$ sulfuric acid at 100° for two hours there was ob­
tained 1.1 g. of the purified 1,4-diiodoanthraquinone.
It
was purified by first heating the crude product In potassium
carbonate solution and then crystallizing the residue from
alcohol.
The product meltedat 218-219°.
Likewise itwas
found that on heating 1,5 g. of the samebenzoylbenzoic
acid
72
in 50 cc. of 95$ sulfuric acid for two hours there resulted,
after dilution, coagulation and purification a 34$ yield of
the 1,4-diiodoanthraquinone,
In the first reaction a 60$
yield was obtained#
Analysis:
Subs#, 0,0675 g,; Agl 0,0692 g,
0.0610 g.; Agl 0.0625 g.
Calcd. for Cl4H60sIs; I, 55.21
Pound:
I, 55.4, 55.3
1,2,4-Triiodoanthraquinone,
To 60 g. of 100$ sulfuric acid was added 2 g. of the
high-melting benzoyltriiodobenzoic acid.
The sulfuric acid
was maintained at a temperature of 105°, while the organic
acid was added slowly over a thirty minute period.
During
this time the sulfuric acid was continuously stirred.
all the reaction was run at 105° for one hour.
mixture was then cooled and poured on ice.
In
The reaction
The resulting
colloidal suspension was coagulated by heating on the steam
bath.
The precipitate obtained was next treated with dilute
sodium hydroxide solution in order to remove the unconverted
benzoylbenzoic acid.
The red residue which was left upon
crystallization from acetic acid yielded 0.5 g. of product
melting at 202-204°•
This is about 25$ of the theoretical
amount*
The low-melting benzoyltriiodobenzoic acid was treated
In an analogous manner.
From 2 g* of the organic acid 0.4
_g_.-of the triiodoanthraquinone was obtained.
Apparently
73
both, acids are converted into the anthraquinone derivative
with the same degree of difficulty.
After two crystalli­
zations from acetic acid the M.P. was 202-203.5°.
Analysis?
Subs., 0.0780 g.j
Agl 0.0937 g.
0.1093 g.;
Agl 0.1313 g.
Calcd. for C14Hs02I3 :
Pound:
I, 64.84
I, 65.0, 64.9
IV#
Condensation of Anisole with the Iodinated
Phthalic Anhydrides
2-p«*Anisoyl-4.5-diiodobenzoic acid
A mixture of 3 g# (0.0075 mol) of4,5—diiodophthalic
anhydride, 1.9 g.
(0.018 mol) ofanisole,
45 cc. of carbon
disulfide, and 2.2 g. (0.0165 mol) of anhydrous aluminum
chloride (added in small portions) was stirred continuously
and heated to the boiling point of the carbon disulfide for
twenty hours#
It was then acidified and warmed on the steam
bath to remove the carbon disulfide#
The water insoluble
product remaining was dissolved in hot potassium carbonate
solution.
The hot alkaline solution, after filtering through
a hot water funnel, on cooling deposited the crystalline
potassium anisoylbenzoate.
It was separated from the mother
liquor, redissolved in hot water and acidified.
The preci­
pitate obtained, after one crystallization from dilute alco­
hol melted at 224-225.5°•
duct.
There was isolated 2.8 g. of pro­
This corresponds to a yield of 73$.
As already stated
74
3?98-c‘kIon has been brought about also by using o-dichlorobenzene as the solvent and In that case a 90$ yield of the
same product was obtained#
Neutral Equivalents
Subs#,
0*1339 g#^required 3#12 cc# of
0*0842N
0*1361 g#;required 3*23 cc. of 0.0842
Calcd#
for CigHioO^Ijgt
Pound:
neut. equiv#, 505, 500#
N NaOH
NaOHsoln#
soln.
neut. equiv., 508
2-p-Anisoyl«*5,4(?)-diiodobenzolc acid
Prom a mixture consisting of 5 g. (0.0125 mol) of 3,4diiodophthalic anhydride, 1.5 g. (0.014 mol) of anisole,
3.8 g# (0.028 mol) of anhydrous aluminum chloride and 75 cc#
of o-dichlorobenzene, after running the reaction at 70° for
five hours, there was obtained 3.1 g. of product melting at
247-249°•
The condensation product was extracted from the
residue, remaining after steam distillation, with hot sodium
carbonate solution.
The solution was then filtered through
a hot water funnel and on cooling yielded the crystalline
sodium anisoylbenzoate#
The free organic acid obtained from
the crystalline sodium salt was crystallized from dilute
alcohol and melted as stated.
A yield of 50$ was obtained*
Neutral Equivalent:
Subs., 0*1878 g.; required 6.07 cc# of 0.0614 N NaOH soln#
0.2105 g.; required 6.83 cc. of 0.0614 N NaOH soln#
Calcd# for CiBH lo04Is :
Poundt
neut. equiv., 508
neut. equiv., 504, 502.
75
2*p~AnisoyI«»3,6-aiioaobenzoio acid
A mixture of 5 g. (0.0125 mol) of 3,6~diiodophthalic
anhydride, 3 cc. of anisole, 110 cc. of carbon disulfide,
and 3.8 g. (0.028 mol) of anhydrous aluminum chloride gave
1.7 g. of condensation product melting at 171-173°, after
one crystallization from dilute methyl alcohol.
The reaction
was run in an identical manner to those previously reported.
In this particular case the reaction time was sixteen hours.
In the purification of the condensation product it was not
found possible to first crystallize it as the sodium or
potassium salt.
of the
In addition to the reaction product, 3 g.
diiodo anhydride was recovered.The reaction
also been carried out using o—dichlorobenzene
has
as the solvent.
In this case the reaction was run at 75° for nine hours.
Only 0.6 g. of product was isolated.
This corresponds to
a 10$ yield.
Neutral Equivalent?
Subs.,
0.1310
g.j
required 4.18 cc. of0.0614N NaOH soln.
0.1965 g.; required 6.23 cc. of
Calcd. for Cl5Hlo04l2 :
Founds
0.0614 N NaOH soln.
neut. equiv., 508.
neut. equiv., 510, 513.
2-p-Anisoyl-5,4»5j6-tetraiodobenzoic acid
The reaction mixture, from which this condensation pro­
duct was extracted, consisted of 10 g. (0.0153 mol) of tetraiodophthalic anhydride, 1.6 g. (0.0148 mol) of anisole, 4 g.
(0.030 mol) of anhydrous aluminum chloride, and 50 cc. of
76
carbon disulfide.
The reaction was run at the boiling
point of the solvent for sixty hours.
The product remain­
ing after acidification and distillation proved to be only
sparingly soluble in hot potassium carbonate solution.
In
order to insure the removal of any unreacted anhydride the
reaction mixture was warmed on the steam bath in a dilute
alkaline solution for over an hour.
The insoluble product
was filtered from the solution and treated with hydrochloric
acid in order to eonvert the potassium salt of the condensa­
tion product into the free benzoylbenzoic acid.
The product
was then crystallized from a dioxane-water mixture.
compound melted at 249-851°*
isolated in the pure state.
The
Only 4.6 g. of product was
This corresponds to a yield of
40$.
Neutral Equivalents
Subs., 0.1831 g.j required 2.93 cc. of 0.0842 N NaOH soln.
0.1241 g.j required 1*98 cc* of 0.0842 N NaOH soln.
Calcd. for CleH804I4 f
Pounds
neut. equiv., 760
neut. equiv., 742, 744*
V.
Other Condensation Reactions of the Iodinated
Phthalic Anhydrides
&-(5»4-dichlorobenzoyl)-4,5-diiodobenzoic acid
A mixture of 75 cc. of o-dichlorobenzene, 3 g. (0.0075
mol) of 4,5-diiodophthalic anhydride, and 2 g. (0.015 mol)
i
of anhydrous aluminum chloride
was heated at 100-105° for
fehe aluminum chloride was always added gradually in small
portions unless otherwise stated*
77
seven hours.
The residue remaining after acidification
and steam distillation, was dissolved in hot sodium car­
bonate solution.
After first filtering through a hot water
funnel, the filtrate on cooling deposited the sodium benzoylbenzoate.
The dichlorobenzoyldiiodobenzoic acid obtained
from this pure sodium salt was crystallized from dilute
alcohol.
The product obtained then melted at 230-231°.
was also recrystallized from dilute acetic acid.
It
However,
this second crystallization of the free organic acid had no
effect on the melting point.
A total of 1*2 g. of conden­
sation product was isolated which corresponds to a 30$ yield.
Neutral Equivalents
Subs., 0.1081 g.; required 3.30 cc. of 0.0614 N NaOH soln.
0.0895 g.; required 2.73 cc. of 0.0614 N NaOH soln.
Calcd. for C14H60aClsIs :
Pounds
neut. equiv., 547
neut. equiv., 534, 534
2-(3j4-dichlorobenzoyl)-5 34 (?)-diiodobenzoic acid
Prom a mixture of 75 cc. of o-dichlorobenzene, 5 g.
(0.0125 mol) of 3,4-diiodophthalic anhydride, and 3.8 g.
(0.028 mol) of anhydrous aluminum chloride, after running
the reaction in the usual manner at 90°-95° for ten hours,
there was obtained 1 g. of pure condensation product.
percentage yield was low being only about 15$.
The
In this re­
action no attempt was made to crystallize the condensation
product as the sodium or potassium salt.
The water insolu-
78
ble residue left after steam distilling the excess o-dichlorobenzene was extracted with a hot solution of potassium
carbonate*
The precipitate obtained on acidification of the
alkaline solution was first crystallized from dilute alcohol,
then from dilute acetic acid.
It melted at 288-290°•
2- (5*4-dimethoxytoenzoyl)-4,5“diiodobenzoio acid
In the condensation of veratrole with 4,5-diiodophthalic
anhydride the following quantities of reagents were used;
50 cc. o-dichlorobenzene, 3 cc. veratrole, 6 g. (0.015 mol)
of the diiodophthalic anhydride, and 4.2 g. (0.031 mol) of
anhydrous aluminum chloride.
The reaction was run for five
hours at 70°, during the course of which it was continuously
stirred.
After acidification and steam distillation the
residue remaining was dissolved in a hot sodium carbonate
solution.
However, on cooling, the sodium salt of the con­
densation product did not crystallize.
The precipitate ob­
tained on acidifying with hydrochloric acid was crystallized
from dilute alcohol.
One fraction of 2.2 g. melting at 239-
240° was Isolated and on further concentration of the
solution 1 g. of product melting at 235-238° was obtained.
79
Based on the total amount of product obtained this corres­
ponds tp a 40# yield*
Neutral Equivalent?
Subs., 0.1519 g.j required 4.62 cc. of 0.0614 N NaOH
0.1754 g.j required 5.27 cc. of 0.0614 N NaOH
Calcd. for Cie^isOglg:
Found:
neut. equiv., 538
neut. equiv. 535, 536.
2-(2-Methyl-4-methoxybenzoyl)-4,5-dilodobenzoic acid
The Friedel-Crafts reaction, whereby this product has
been prepared, has been run both in carbon disulfide and in
o-dichlorobenzene•
In the one instance the reaction was run
for twelve hours at the boiling point of the carbon disul­
fide, while in the other the reaction was run for five hours
at 80°•
The same experimental procedure was followed in
both cases, it being analogous to that already described.
As in most of the condensations of 4,5-diiodophthalic anhy­
dride the reaction product was readily purified through its
sodium salt, followed by the crystallization of the free
benzoylbenzoic abid from dilute alcohol.
From a mixture of
3.2 g. (0.008 mol) of 4,5-diiodophthalic anhydride, 1 g.
(0.0082 mol) of m-cresyl methyl ether, 2.1 g. (0.016 mol)
of anhydrous aluminum chloride and 50 cc. of carbon disul­
fide, 2 g. of the pure condensation product was isolated.
Thus, the compound melting at 230-231° was obtained in a
50# yield*
In the other reaction the following quantities of re­
agents were used; 75 cc. of o-dichlorobenzene, 5 g. (0.0125
80
.
mol) of the 4,5-diiod© anhydride, 3 cc. of m-cresyl methyl
ether, and 3.8 g. (0.028 mol) of anhydrous aluminum chloride#
About 4 g. of product melting from 222-228° was isolated after
two crystallizations, first as the sodium salt and then as
the free aeid.
obtained.
On further crystallization two fractions were
The one melted at 230-231°, the other at 214-220° •
At first it seemed that two isomers had been obtained but it
now seems more likely that some condensation with o-dichloro­
benzene had occurred.
The yield of the desired compound in
this instance was 35# of the theoretical#
Neutral Equivalentt
Subs., 0.0426 g. j required 1.01 cc. of 0.0778 N NaOH soln#
0.0378 g.j required
Calcd. for Cx6HiSQ*Ia:
Found?
0.93 cc. of 0.0778 N NaOH soln#
neut. equiv., 522
neut. equiv., 530, 523.
Condensation of 4,5-diiodophthalic Anhydride
with Besorcinol Dimethyl Ether#
The condensation was brought about to give a 65# yield
of
a benzoylbenzoic acid. The reaction was run in o-dichloro­
benzene for four hours at 90°•
The quantities of reagents
used were; 75 cc. of o-dichlorobenzene, 5 g. (0.0125 mol) of
the diiodo anhydride, 3 cc. of resorcinol dimethyl ether, and
3.8 g. (0.028 mol) of anhydrous aluminum chloride.
The reac­
tion product was first crystallized as the sodium salt, follow­
ed by the subsequent crystallization of the free acid from an
alcohol-water mixture.
The product isolated melted at 235-236°•
When the reaction was run in carbon disulfide only a 22# yield
of product was obtained.
In determining the neutral equiva­
lent of this product the compound itself served as the indi­
.
81
cator, the color change being from white to yellow.
How­
ever the results obtained were misleading since they indi­
cated the compound to be 2-(2,4-dimethoxybenzoyl)-4,5-diiodobenzoic acid.
obtained.
It was expected that this compound would be
However, an iodine analysis and a sodium analysis
of the salt indicated that one methyl group had been cleaved
and thus the product was the same as that isolated in the
condensation of this same diiodo anhydride with resorcinol
monomethyl ether.
Analysis?
Subs., 0.1148 g.$ Agl 0.1032 g.
Calcd. for CX6H1o0bI8 5 I, 48.47
Found:
I, 48.58
Subs., 0.1512
N&gSO^ 0.0196 g.
0.1229 g.j Na^SO* 0.0161 g.
Calcd. for Ci6H©0eIaNaj Na, 4.21
Found:
4.19,
4.24
2- (2-Hydroxy-4-methoxybenzoyl) -4«5-diiodobenzoic acid
The condensation of resorcinol monomethyl ether and 4,5diiodophthalic anhydride resulted in a 53# yield of a product
melting at 238-240°♦
A somewhat different procedure was
followed in carrying out this reaction.
A 5.5 g. (0.041 mol)
portion of anhydrous aluminum, chloride was added t© 100 cc. of
carbon disulfide, to which 5 g. (0.0125 mol) of the diiodo
anhydride had previously been added. Then 2 g.(0.015
©f resorcinol monomethyl ether, which wasdissolved
j©f carbon disulfide, was
added to the
mol)
In 75 cc.
82B
aluminum chloride complex drop by drop.
run for twelve hours at the boiling point
sulfide.
The reaction was
of the carbon di­
As the resorcinol monomethyl ether
was added a
precipitate formed, on the wall of the reaction vessel, at
the point where it was dropping.
The reaction product was
purified first as the sodium salt.
Then the free acid was
crystallized from dilute methyl alcohol.
In determining its
neutral equivalent phenol red was used as the indicator.
Neutral Equivalents
Subs., 0.1978 g.j required 4.24 cc. of 0.0885 N NaOH soln.
0.1829 g.j required 3.94 cc. of 0.0885 N NaOH soln.
Calcd. for Ci5Hxo0eIa t
Pounds
neut. equiv., 524
neut* equiv., 527, 524.
Analysis t
Subs., 0.1614
0.1108
g.j NaaSO* 0.0209 g.
g.j NaaS04 0.0142 g.
Calcd. for Cx8H 906
Pounds
4.19,
IsNaj Na, 4.21
4*15
2-(2-Hydroxy-5-methylben2oyl)-5,4(?)-dliodobenzoic acid
The quantities of reagents used in the preparation of
this acid were; 50 g. (0.463 mol) of p-cresol, 15 g. (0.037
mol) of 3,4-diiodophthalic anhydride, and 17 g. (0.128 mol)
of anhydrous aluminum chloride.
The temperature at which
the reaction was run was gradually raised to 75° over a three
hour period.
The reaction was then continued at that tempera­
ture for eighteen hours more.
The reaction mixture became
83
very viscous but, nevertheless, it was possible to stir it
continuously*
After acidifying and steam distilling, the
reaction mixture was immediately filtered.
The purpose of
filtering the residue while hot was that some of the un­
reacted 3,4-diiodo anhydride, which was now present as the
acid, would be in solution and could be readily removed and
recovered*
The residue remaining after this filtration was then ex­
tracted with hot sodium carbonate solution*
amount of residue remained*
An appreciable
It was filtered off and found
to be soluble in alcoholie potassium hydroxide, which indi­
cated that it was a phthalide derivative*
On cooling the
sodium carbonate solution, the sodium salt of the benzoylbenzoic acid was deposited*
The organic acid obtained from
this sodium salt, after one crystallization from an alcoholwater mixture, melted at 273-275°*
In the best run 6 g* of
condensation product was obtained*
This corresponds to a
yield of 31$.
Analysis?
Subs*, 0*1229 g*j Na2S04
0*0165 g*
0*1775 g.; UasSQ*
0.0239 g*
Calcd* for C15H 904I2Na ; Na, 4*34#
Found?
4*35, 4*36
2- (2-Hydroxy-5-ohlorobenzoyl)-3,4 (?) -diiodobenzoic acid
The condensation of 3,4-diiodophthalie anhydride and
p-chlorophenol has been accomplished in good yield*
The
84
reaction was run in a similar manner to the condensation in**
volving p-eresol*
However, in this condensation the reaction
mixture quite frequently got too viscous to stir*
In those
cases 10-20 cc* of o-dichlorobenzene was added as a diluent*
A typical run Is described*
The reaction mixture consisted
of 60 g* (0*469 mol) of p-chlorophenol, 16 g* (0*037 mol) of
the diiodo anhydride, and 18 g* (0*135 mol) of anhydrous
aluminum chloride*
The temperature at which the reaction
was run varied between 85-90°•
After running for sixteen
hours at that temperature 10 cc* of o-dichlorobenzene was
added*
The reaction was continued for five more hours*
The
reaction mixture was then worked up in the usual manner*
The steps in the purification of the reaction product In­
volved the crystallization of its sodium salt, followed by
a crystallization of the free acid from a dioxane-alcoholwater mixture*
The compound melted at 266-268°♦
particular case an 80$ yield was obtained*
In this
There was no
indication of a phthalide by-product in this condensation*
Analysis t
Subs*, 0*1260 g*j NasS04
0*0172 g*
0*1479 g.j NagSO*
0.0194 g.
Calcd* for C^Hel-sClNa;
Found:
VI*
4*42,
Ha, 4*18
4.25
Methylation and Acetylation of 2-(2-Hydroxy-5chlorobenzoyl)-3,4(?)-diiodobenzoic Acid
2-(2-Methoxy-5-ohlorobenzoyl)-5J4(?)diiodobenzoic methyl ester
Since methylation of the acid could not be brought about
85
in an aqueous medium, even after repeated attempts on the
same sample, it was decided to use a modified method, using
an anhydrous solvent for the reaction mediumt7 A portion of
the benzoylbenzoic acid was dissolved in hot potassium hydr­
oxide solution.
On cooling the potassium salt was deposited.
It was believed to be the dipotassium salt of the acid.
After
filtering off and drying, a 4 g. (0.0066 mol) sample was then
suspended in 70 cc. of toluene and 2.5 g. (0.02 mol) of di­
methyl sulfate added drop by drop.
refluxed for two hours.
steam distilled.
solidified.
The reaction mixture was
Water was then added and the toluene
An oily residue was left but on cooling it
This residue was then extracted with alcoholic
potassium hydroxide.
The product remaining, after one
crystallization from a dioxane—alcohol—water mixture, melted
at 143-145®*
About 2 g. of product was obtained which is
about a 55$ yield*
In an attempt to further purify the pro­
duct by recrystallization it melted over a larger range,
146-150°*
(87) Hickinbottom, Reactions of Organic Compounds, Longmans,
Green and Co* (1936)
86
^ ig-Methoxy-5~chlorobenzoyl)-5,4(?)-diiodobenzoic acid
A portion of the dimethylated product, prepared as
just indicated, was dissolved in a solution containing 40
cc. of dioxane, 40 cc. of water, and 2 cc. of concentrated
sulfuric acid and refluxed for four hours•
The product was
then precipitated upon removal of the dioxane and addition
of water.
After extraction with a dioxane-potassium hydr­
oxide solution, filtering, acidification, and crystallization
from a dioxane-aleohol-water mixture, a product was isolated
melting at 256-259°.
Analysis;
Subs., 0.1236 g.j Na^SO*
0.0157 g.
0.1095 g.j NasS04
0.0140 g.
Calcd. for CiBH804IaClNaf
Found:
4.11,
Na
4.07
4.14
2-(2-Acetyl-5-chlorobenz oy1)-3,4(?)-dilodobenzoic acid
From a reaction mixture consisting of 3 g. of the corre­
sponding hydroxy compound, 40 cc. of acetic anhydride and
three drops of sulfuric acid, 1.7 g. of the acetylated pro­
duct melting at 218-220° was isolated.
The reaction mix­
ture was heating on the steam bath for three hours, then
poured into water and kept in the ice box for several hours.
The product was crystallized from a dioxane-alcohol mixture.
The acetylation reaction has also been carried out in
Q7
pyridine, using acetyl chloride as the acetylating agent.
From a mixture of 8 g. (0.0151 mol) of the hydroxybenzoylbenzoic acid, 4.7 g. (0.0603 mol) of acetyl chloride, and
87
50 g* of pyridine, 6*5 g* of product was isolated, melting
at 213-216°•
This corresponds to a 75$ yield*
As in the
previous case a dioxane-aleohol-water mixture served as a
crystallizing medium*
711*
Preparation of the Diiodophthalimides
5,4-Diiodophthalimide
The procedure followed was analogous to that used by
Pratt and Perkins for the conversion of tetraiodophthalic
3Q
anhydride into the corresponding phthalimideo
In this re­
action 30 g* (0*075 mol) of the 3,4-diiodo anhydride, 47 g*
(1*04 mol) of formamide and 100 cc* of nitrobenzene were
the reagents used*
The anhydride was first dissolved in the
nitrobenzene and the formamide added to the solution which
was at a temperature of 150°•
The temperature was then
raised to 195°-200° and maintained at that range for a half
hour*
On cooling the reaction mixture the desired product
crystallized from the nitrobenzene*
About 26 g* of pro­
duct was obtained which corresponds to a 90$ yield*
small portion was crystallized from xylene*
A
It melted at
305-306°, while originally it had melted at 30 0-301°*
Analysis?
(Kjeldahl method)
£ Analyses by Mr* Washburn
(88) Pratt and Perkins, J* Am. Chem. Soc. 40, 198 (1918)
88
Subs. , 0.1234 g.j
required 5.16
cc. of
0.0584 N EaS04 soln.
0.1009 g.j
required 4.21
cc. of
0.0584 N H2S04 soln.
Calcd. for C8Hs02.HI2 :
Pounds
N 3.50
N, 3.53, 3.41.
3,6-Diiodophthalimide
This compound was prepared in the same manner as the
previous isomeric
phthalimide.
Furthermore, the same
quantities of reagents were used.
crude product was obtained.
yield.
In this case 23 g. of
This corresponds to an 80$
A small portion was recrystallized from xylene and
found to melt at 301-303°.
Analysis:
(Kjeldahl method)
Subs., 0.1009 g.j
required 4.18
cc. of
0.0584 N H2S04 soln.
0.1034 g.j
required 4.26
cc. of
0.0584 N H2S04 soln.
Calcd. for C8H302NI2 :
Pound:
N 3.50
N, 3.39, 3.37
VIII.
Miscellaneous Reactions
Fhenol-4.5-diiodophthalein dimethyl ether
Priedel and Crafts in the preparation of p-toylbenzoic acid, from phthalic anhydride and toluene, using
one and one half mols of aluminum chloride to every mol of
anhydride reported the formation of a small quantity of
resinous by-product.
Von Peckmann on repeating the work
found that this by-product was ditolylphthalide.
Analyses by Mr. Washburn
Further-
89
more, Bubidge and <S|aa89found a phthalide was also formed in
the reaction between benzene and phthalic anhydride*
These
investigators stated that the formation of the phthalide
was due to the action of an excess of the anhydride on the
intermediate addition product of aluminum chloride and the
benzoylbenzoic acid*
These results have been confirmed in
the condensation of 4,5-diiodophthalic anhydride and anisole,
a 60$ yield of the phthalide being isolated*
From a mixture of 10 g* (0*025 mol) of 4,5-diiodophtha­
lic anhydride, 5*1 g. (0*039 mol) of anhydrous aluminum
chloride, and 40 cc* of anisole there was obtained 9 g* of
reaction product*
three hours*
The reaction was run at 85° for twenty-
The product remaining after steam distillation
was extracted with aqueous potassium carbonate*
the product seemed to dissolve*
Not much of
However, it was readily
soluble in alcoholic potassium hydroxide solution*
On cool­
ing the latter solution a crystalline product was deposited*
This salt was then converted into the free organic substance
and after crystallization from an alcohol-water mixture melted at 149-150°♦
Analysis:
Subs*, 0*4413 g*; Agl 0.3453 g*
0*2699 g*; Agl 0.2115 g*
Calcd* for Cjga-HigO^Ig!
Found:
C89)
I 42*49
I 42*41, 42*33
Klipstein
Ind. Eng. Chem* 18, 1327 (1926)
90
5,6-:Dilodoanthranilio acid
The procedure followed In the conversion of 3,6-diiodophthalimide Into the corresponding anthranilic acid was
analogous to the method used by Spath and Holler90 In a
similar conversion of an imide to an anthranilic acid de­
rivative*
A 10 g. (0*0251 mol) sample of the phthalimide
was dissolved in 50 cc* of 10$ NaOH solution which was then
diluted with 50 cc* of water and mixed with a bromine solu­
tion made from 125 cc* of water, 10 g* (0*063 mol) of
bromine and 5 g. of sodium hydroxide*
The mixture was
then heated on the steam bath for one and one half hours*
In working up the reaction mixture it was decided to acidi­
fy it with hydrochloric acid rather than neutralize the alka­
li with sulfur dioxide as was done by these investigators*
On adding the hydrochloric acid iodine was evolved.
In order to precipitate the reaction product copper
sulfate solution was added to the filtered reaction mix­
ture, which had been made acid to litmus with acetic acid*
A green precipitate was formed which was allowed to settle
and then filtered off*
It was then suspended in water and
hydrogen sulfide was bubbled in for three hours*
The in­
soluble copper sulfide was filtered off and the organic sub­
stance crystallized from the hot water solution*
It was
recrystallized from hot water and then melted at 151-153°•
(90) Spath and Holler, Ber* 56, 2454 (1923)
91
Only 2 g. of pure product was Isolated*
4j
Analysis?
(KJeldahl method)
Subs*, 0.1014 g.j required 4.54 cc. of 0.0584 N HsS04 soln*
Galcd. for C7H B0sNIs;;:
Pound:
N 3.59
N 3.66
£ Analysis by Mr. Washburn
92
IX.
Methods of Analysis
Considerable difficulty was encountered in the analysis
of the iodinated anthraquinones for iodine.
The usual pro­
cedure with the Parr bomb gave only partial decomposition
of the compounds.
The Stepanow01 Method using sodium and
absolute ethyl alcohol gave erratic results, which were
high, until it was found that digestion of the silver iodide
precipitate for several hours with dilute nitric acid gave
consistent results on both known compounds and the iodinat­
ed anthraquinones.
Despite the fact that this method of iodine analysis
gave consistent and correct results a speedier method was
desired.
Baubigney
Therefore, the oxidation method of Chavanne and
was investigated.
results at first.
This method also gave high
It was later found that this was due to
the formation of chromic oxide, formed by the decomposition
of ammonium dichrornate.
In the original analytical scheme the oxidizing mixture
consisted of 40 cc. of concentrated sulfuric acid and 4 to
8 g. of potassium dichromate.
Chavanne reported that quite
frequently on diluting and cooling the reaction mixture,
after the reaction was completed, silver dichromate was preci­
pitated from the solution.
It would not be filtered off be­
cause it was mixed with silver iodate.
However, they pro
-(91) Cook and Cook, J. Ind. and Eng. Chem* Anal. Ed. 5, 186
(1933)
(92) Chavanne, Compt. rend. 136 1197 (1903)
93
ceeded to remove it by adding ammonium salts*
The silver
dichromate was thus converted into the more soluble ammonium
dichromate*
The difficulty encountered here was that the
ammonium dichromate was easily decomposed and the chromic
oxide formed always made the analysis high*
It seemed, however, that if the amount of potassium di­
chromate were limited this difficulty would be overcome*
Pour analyses have been run In .which the amount of potassium
dichromate has been limited to 25% in excess of the theore­
tical amount*
It was found that consistent and accurate
results could then always be obtained*
With one other ex­
ception the method was the same as that reported in the
literature*
Instead of carrying out the reduction of the
iodate with a solution of sulfur dioxide, sodium sulfite
solution was added to the iodate solution*
Too much sodium
sulfite had to be avoided, otherwise free silver would be
obtained*
It was found that 2 g* in excess of the theoreti­
cal amount of sodium sulfite was sufficient to bring about
the complete reduction of the iodate to the iodide*
The reactions involved in this analytical procedure are
as indicated*
3CM H leG4I*. 4 53KsCrs07 4 212H2S04 --- »
66C0S 4 53KsS04
* 53Crs (S04 )3 4 233HS0 4 6HI03
6AgN03 4 6HI03
6AgI03 4 18NaffS03
» 6AgI03 4 6HN03
--- >
6Agl 4 18NasS04
The analyses were carried out in the following manner*
^The oxidizing medium was prepared by dissolving 1 to 1*5 g*
of silver nitrate in 40 cc* of concentrated sulfuric acid*
The solution of the silver nitrate was hastened by heating
and agitation*
Then
in excess of the theoretical amount
of potassium dichromate was added*
Again the solution was
warmed and agitated in order to effect more rapid solution*
This oxidizing mixture was then cooled and the vial contain­
ing the sample to be analyzed was dropped into the 300 cc*
Erlenmeyer flask containing the oxidizing solution*
After
several minutes of agitation the temperature was gradually
raised to 150-160°•
After cooling in an ice bath 140-150
cc* of water was added*
The iodate was then reduced by
adding 2 g* in excess of the calculated amount of sodium
sulfite*
The solid material could be added directly but
there was less splattering when a solution was added*
After
10-15 minutes digestion of the precipitate, followed by a
cooling period, the silver iodide was filtered off, dried
and weighed*
This method Is also applicable in determining iodine
in the presence of chlorine and bromine*
The chlorine and
bromine are evolved in the free state but the iodine being
more readily oxidized is retained as the iodate*
As yet
this phase of the analytical procedure has not been per­
fected*
The intermediate benzoylbenzoic acids were titrated In
aqueous alcohol solution with standard alkali, usually using
either phenolphthalein or thymol blue as the indicator.
95
SUMMARY
1*
The partially iodinated phthalic anhydrides have been
condensed with benzene and the resulting benzoylbenzoic
acids converted into the anthraquinone derivatives*
Three
of* the four theoretically possible homonuclear diiodo—
anthraquinones have been prepared*
2*
Three isomeric diiodophthalic anhydrides and tetraiodo-
phthalic anhydride have been condensed with anisole*
No
attempt, however, was made to convert the anisoylbenzoic
acids into the anthraquinone derivatives*
3*
The intermediate compound involved in the nuclear syn­
thesis of an iodinated quinizarin derivative has been pre­
pared but the proper conditions for ring closure were not
found*
4*
Several other iodinated hydroxybenzoylbenzoic acids
have been prepared and their conversion into the correspond­
ing anthraquinones has been investigated*
5*
Miscellaneous reactions of the partially iodinated
phthalic anhydrides have been investigated*
6*
Two methods for the quantitative estimation of iodine
have been improved*
96
BIBLIOGRAPHY
1.
go
3.
4*
dm
6*
7.
8.
9#
10o
11m
12.
13.
14*
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25o
26.
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
41.
42.
2a.
43i
44*
45.
Gottlieb,'Anat. Anz. 46, 179 (1914)
Schreiber, Arb. path.~5nat. Bakt* 4, 257 (1904)
Richter, Biochem. J. 3 1 , 591 (1 9 3 7 T
C.A. 2 3 , 4 7 77 (1 9 2 9 ) Brit. Patent 304589
Pribam, Dent# Med. Wochschr. 5 2 , 1291 (1 9 2 6 )
BinZ, R&th, and Lichtenberg, Z. angew. Chem. 43, 452 (1930)
Smitt and Bok, Hederland. Tj-dschr. Geneeskunde 68,
2213 (1926)
—
Binz and Maier-Bode, Biochem. Z. 252, 16 (1932)
Graebe and Lieberman, Ber., 1, 49 (1868)
Laurent, Ann. 54, 287 (1840)
Phillips, Chem. Rev. 6, 157 (1929)
Schilling, Ber., 46, 1066 (1913)
Bayer, German Patent 131538, Frdl. 6, 311
Badische Anilin und Soda-Fabrik, German Patent 252,578
Prdl. 11, 545
Ullmann and Krecht, Ber., 4 4 , 3128 (1911)
Coppers, Rec. trav. Chim. 44, 907 (1925)
Bayer, German Patent 205,195, Prdl* 9, 673
Kaufler, Ber*, 57, 63 (1904)
Egerer and Meyer, Monatsh. 34^ 69 (1913)
Ree, Ann. 255, 240 (1886)
HSchst, German Patent 75,288, Prdl. 2>, 260
Houben, Anthracene and Anthraquinone; George Thieme,(1929)
Ullmann and Billig, Ann. 581, 11 (1911)
Fierz-David, J* Am. Chem. Soc. 49, 2334 (1927)
Junghans, Ann., 399, 316 (1913)
Hammerschlog, Ber., 3L9, 1109 (1886)
Ullmann and Conzetti, Ber., 53, 826 (1920)
Fierz-David, Helv. chim. Acta 10, 197 (1927)
Badische Anilin und Soda-Pabrik, German Patent 254,450,
Prdl. 11, 546
Goldberg, J* Chem. Soc. 1771 (1931)
Goldberg, ibid. 2829 (1931)
Badische Anilin und Soda-Pabrik, German Patent 197,554
Prdl. 9, 765
C.A. 27, 4687 (1933>
Kircher, Ann., 258, 344 (1887)
Tanaka, Proc. Imp. Acad., Tokio 3, 82 (1927)
Graebe and Rostowzew, Ber., 54, 2113 (1901)
C.A. 2, 3410 (1908); French Patent 384,471
Badische Anilin und Soda-Pabrik, German Patent 214714,
Prdl. 9, 678
Heller, Ber. 45, 792 (1912)
Bayer, German Patent 228,901, Prdl. LO, 578
Hofmann, Monatsh. 56, 805 (1915)
Eckert and Steiner7~Mona.tsh. 36, 269 (1915)
Eckert, J. prakt. Chem. 2 IQS’, 361 (1921)
V. Pechmann, Ber., 12, 2127 (1879)
Waldmann, J* prakt. Chem. 126, 65 (1930)
Graebe and Liebermann, Ann. Ski., jf, 288 (1870)
97
46.
47.
48.
49i
50.
51.
52.
53.
54.
55.
56.
57.
58.
59.
60.
61.
62*
63.
64.
65.
66.
67.
68.
69.
70.
71.
72.
73.
74.
75.
76.
77*
78.
79.
80.
81.
82.
83.
84.
85.
86.
87*
88.
89.
90.
91.
92.
Cain and Thorpe, Synthetic Dyes tuffs and Intermediate
Products; Charles Griffin and Co. (1933)
Battegay and Claudin, Bull. soc. chim. 4 29, 1017 (1922)
Grandmougln, Compt. rend. 175, 859 (1921)
Ullmann and Eiser, Ber. 49, 2154 (1916)
Perkin, J. Chem. Soc. 37, 555 (1880)
Kaufler and Imhoff, Ber. 37, 4707 (1904)
Battegay and Claudin, Bull. soc. ind. Mulhouse,86, 632
(1920)
— '
Dhar, J. Chem. Soc. 117, 993 (1920)
Scholl, Eberle, and Fritach, Monatsh. 32, 1043 (1911)
Diehl, Ber. 11, 181 (1878)
““
Bayer, German Patent 107,721, Frdl. 5, 302
Ullmann and Minajeff, Ber. 45, 687 (1912)
Scholl, Ber. 40, 1696 (19071“
Schaarschmidt, Ann. 405, 95 (1914)
Lauhe, Ber. 40, 5566 (1907)
Kaufler, Ber. 37, 60 (1904)
Scholl, Haas, Meyer, and Seer, Ber. 62B, 107 (1929)
C.A. 25, 2994 (1931)
Eckert and Klinger, J. prakt. Chem. 2
121,281
(1929)
Laurance, J. Am. Chem. Soc. 43, 2577 (19^X7
Waldmann, J. prakt. Chem. 150, 92 (1931)
Waldmann and Mathiowetz, ihid. 126, 250 (1930)
Frey, Ber. 45, 1361 (1912)
HBvermann, Ber. 47, 1210 (1914)
Liebermann and RiTber, Ber. 33, 1658 (1900)
Dimroth, Schultze, and Heinze, ibid. 54, 3035(1921)
Wed., German Patent 189937, Frdl. £ , 6 8 5
Hfichst, German Patent 77179, Frdl. 4,330
Hardacre and Perkin, J* Chem. Soc. 180 (1929)
Diehl, Ber. 11, 190 (1878)
Perkin and Story, J. Chem. Soc. 2620 (1931)
Perkin and Story, ibid. 229 (1928)
Karrer, Organic Chemistry, Nordemann Publishing Co.,
New York (1938)
Morgan and Smith, J. Chem. Soc. 121, 160 (1920)
Weiser and Porter, J. Phys. Chem. 31, 1824 (1937)
Organic Syntheses, Vol. 6, p. 78
Graves and Adams, J. Am. Chem. Soc. 45, 2439 (1923)
Ullmann and Schmidt, Ber. 52, 2098 (1919)
Pratt and Perkins, J. Am. Chem. Soc. 40, 219 (1918)
Willst&tter and Muller, Ber. 44, 2182*^"l911)
Bentley, Gardner, and Weizmann, J. Chem. Soc. 91, 1626
(1907)
Hioklnbottom, Reactions of Organic Compounds, Longmans,
Green and Co., (1936)
Pratt and Perkins, J. Am. Chem. Soc. 40, 198 (1918)
Klipstein
,Ind. Eng. Chem. 18, 1327 (1926)
Spath and Koller, Ber. J56, 2454 (1923)
Cook and'Cook, J. Ind. and Eng. Chem. Anal. Ed. _5, 186 (193
Chavanne, Compt. rend. 136, 1197 (1903)
98
~VTTA
ROBERT WILEY HIGGINS
Born:
April 27, 1913 at New Hope, Pennsylvania
Education:
New Hope Public Schools, New Hope, Pennsylvania
1920-1931
B# S#: Franklin and Marshall College, Lancaster,
Pennsylvania, 1932-1936
Positions held:
Assistant in Chemistry, Franklin and Marshall College,
1935-1936
Graduate Assistant in Chemistry, University College,
Northwestern University, 1936-1937
Graduate Assistant in Chemistry, Northwestern
University, 1937-1938; 1939-1940
Graduate Assistant in Chemistry, Northwestern
University, Summer Session, 1938
University Fellow in Chemistry, Northwestern
University, 1938-1939
Affiliations
Phi Beta Kappa
Phi Lambda Upsilon
Sigma Xi
Publications:
Synthesis of Iodinated Benzoylbenzoic Acids and
Anthraquinone Derivatives#
Journal American Chemical Society 61. 2662
(with C. M. Suter)#