tº . ~~
sº tº *
A
4.A. : ;
§3*
jä，3 × s-º
¿№ „
* * ·，≤º
. +→.≡ »
§§§
ſae！
<aºx ***-->
· §*ş ", . . ^（.*¿.*
&& zºº|-----
× × ×… №º_° ******
！

' ' ' t t , , , , ,
1 * : * * * * * ,
* } • ,
, , , ,
tº:
}} | {
* * r * ,
* * * * * * *
. . . . . . . . . . . . .
* * * * ,
! { * *
* * * * *
t * * * * ,
g
#:
g
4

On Some
Para-Hydroxytriphenylmethane
Derivatives
A Contribution to the Chemistry of
- Free Radicals
BY
R. L. JICKLING
I9I4.
On Some
Para-Hydroxytriphenylmethane
Derivatives
A Contribution to the Chemistry of
- Free Radicals
A THESIS -
SUBMITTED IN PARTIAL FULFILMENT OF THE REQUIREMENTS FOR
THE DEGREE OF DOCTOR OF PHILOSOPHY IN. -
THE UNIVERSITY OF MICHIGAN
BY
R. L. JICKLING
I9I4.
I beg to acknowledge the painstaking care with which Professor
M. Gomberg has directed me in my work and I wish to express to him
my great appreciation for all that he has helped me. to accomplish. -
- - ROBERT LEE-JICKLING.
ANN ARBOR, Mich., - ~
November 13, 1914. . .
CONTENTS.
I. Introduction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
II. The reaction between benzophenone chloride and phenol. . . . . . . . . . . . . . . . . . .... 6
The preparation of benzophenone chloride. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
The mechanism of the reaction. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
Diphenoxydiphenylmethane. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
p-Hydroxytriphenylcarbinol . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . IO
Di-p-hydroxytetraphenylmethane. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . II
Mono-p-hydroxytetraphenylmethane. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I2
III. Concerning the free radical, p-hydroxytriphenylmethyl. . . . . . . . . . . . . . . . . . . . . . I2
p-Hydroxytriphenylcarbinol chloride. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . - - - - - - . . I3
p-Carboethoxytriphenylcarbino1. . . . . . * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * * I4
p-Carboethoxytriphenylcarbinol chloride... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I4.
The action of metals. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I4.
p-Carboethoxytriphenylmethyl peroxide. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I4.
Isomerization of the free radical. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I5
p-Benzoxytriphenylcarbinol. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I6
p-Benzoxytriphenylcarbinol chloride. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I7
p-Benzoxytriphenylmethyl peroxide. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I7
p-Acetoxytriphenylcarbinol. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
p-Acetoxytriphenylcarbinol chloride... . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I7
p-Acetoxytriphenylmethyl peroxide. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I8
IV. Concerning p-hydroxytriphenylmethyl ether. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I8
p-Carboethoxytriphenylmethyl ether. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I8
Hydrolysis. . . . . . . . . . . . . . . . . . . . . . . • - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - - I9
V. Summary. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I9
VI. Bibliography. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2O
282416
I. Introduction.
It was suggested by Kehrmann' and proved by Gomberg” through
experimental evidence that the triarylmethyl halides exist in two forms,
the benzoid as colorless and the quinoid as colored, and that there is
a tautomeric equilibrium between the two:
sºmºsºmeº -: H
*-*-C = *-e-Cº.
C1 I. II.
In solution the presence of these two forms is evident, but in the solid
state only the one, the benzoid, has been isolated. In the cases of the
triarylmethyl sulfates” (III), perchlorates,” etc., the other desmotropic
form constitutes the solid phase, i. e., these derivatives are obtained
Only in the colored, quinoid state. The halides” of the xanthenols
and thioxanthenols, which after all are but triarylmethyl derivatives,
exist in some instances as colorless solids (IV), like simple triphenyl-
chlormethane, and in others as colored, resembling the triarylmethyl
sulfates. The latter case is illustrated by the chloride of phenyl-3-
hydroxyxanthenol" (V).
C6H5 C6H5
lyci |
2CK /°s/s
-: H
(C.H.).c =&TX OH
N=/ YSOH /
No.2% No/N/\ci
III. IV. V.
The same explanation of color formation has been applied to the free
radicals themselves." Accordingly triphenylmethyl would exist in
solution as the two forms of the free radical, the benzoid and the quinoid,
in tautomeric equilibrium:
== H
(C6H5)3C– — — Tº: (C6H5)2C ={ ×
X.
From any solvent triphenylmethyl crystallizes in the colorless form,
but there are, however, certain analogs of triphenylmethyl, which
exist as colored solids.” !
In each of the three series mentioned—the halides, the sulfates, and
the free radicals—one and only one of the two forms has been isolated
as a solid, in some cases the benzoid and in others the quinoid. Lately
it was shown that a hydroxy group in the para position to the central
carbon atom increases greatly the tendency toward tautomerization
to the quinoid state.” Triphenylcarbinol is known only in the ben-
zoid form; its p-hydroxy derivative, however, has been obtained by
Gomberg” in two distinct forms, the colorless and the colored, separable
from each other and either readily changeable into the other merely
6
by choice of suitable solvent. The significance of this latest fact be-
comes evident at once. The conception of the tautomeric existence
of all other triarylmethyl derivatives in these two forms becomes more
certain than before as the result of the conclusive proof that this one
series, the carbinols, actually exists in the two desmotropic forms.
In view of these facts, the questions naturally arise: What influence
would a p-hydroxyl group in triphenylmethyl exert as regards the
tendency of the free radical toward tautomerization? Would p-hy-
droxytriphenylmethyl, like the simple triphenylmethyl, exist as two
forms in solution only, and as a solid, colorless? Would the proportion
of the quinoid tautomer in solution be greater than in the case of tri-
phenylmethyl? Would this p-hydroxylated free radical exist even
as a solid in the two desmotropic forms, as is the case with the corre-
sponding carbino1? Or might the quinoid form of the p-hydroxy-
triphenylmethyl be the only one obtainable in the solid state?
As a related problem, we have also attempted to prepare p-hydroxy-
triphenylmethyl ether, the existence of which has been the subject of
considerable discussion in recent literature. s g
For the preparation of the free radical and of the ether, p-hydroxy-
triphenylcarbinol must serve as the starting point. The methods
previously used to prepare this carbinol are both tedious and extended.
A far simpler synthesis has been devised during the course of this work,
and since this method has proved to be efficacious in the preparation
of many homologous hydroxy derivatives, a detailed study of the same
has been undertaken.
II. The Reaction between Benzophenone Chloride and Phenol.
In an unsuccessful attempt to prepare diphenoxydiphenylmethane
(I) by the reaction of sodium phenoxide on benzophenone chloride,
Mackenzie” obtained instead di-p-hydroxytetraphenylmethane (II).
He also noticed that the same product resulted from the direct action
of phenol itself on benzophenone chloride. Through use of an analogous
reaction, Smedley” prepared di-p-hydroxydiphenyldiphenylenemethane
(III), from fluorenone chloride and phenol. Zincke” made use of the
reaction mentioned by Mackenzie, and obtained the same di-p-hydroxy-
tetraphenylmethane, using a large excess of phenol and heating on the
water bath for three days. Sachs and Thonet” have condensed ben-
zophenone chloride with catechol by means of absolute sulfuric acid,
and claim to have obtained 3,4-dihydroxytriphenylcarbinol (IV).
OC5H5 º C6H4OH CGH4 C6H4OH: 2C6H3(OH)2
cºcº. (choic. Jº º (C.H.).có
I. II. III. - * IV.
A careful study of the reaction between benzophenone chloride and
phenols has been undertaken and it has been found that the condensation
7
takes place in several successive steps. By observing the proper con-
ditions, certain intermediary products of this reaction have been isolated,
and in this way there has been devised a valuable and advantageous
method of preparing various hydroxytriphenylcarbinols.
The Preparation of Benzophenone Chloride.—In studying the Friedel
and Crafts’ synthesis, Boeseken” found that the reaction between carbon
tetrachloride and benzene under the influence of the catalyst, aluminium
chloride, proceeds in at least two steps:
CC14 + C6H6 —- C6H5CC13 -H HCl
C6H5CCl3 + C6H5 —- (C6H5)3CCl2 + HC1
(C6H5)2CCl3+ C6H6 —- (C6H5)3CC1 + HC1
He was unable to isolate benzotrichloride, but by using an excess of
carbon tetrachloride he obtained benzophenone chloride in good yield.
We desire to emphasize the value of this reaction as an excellent means
for the preparation of benzophenone chloride in large quantities.
We have obtained benzophenone chloride in 90 per cent. yields by
observing the following procedure: In a wide-mouth, two-liter bottle,
135 grams (I mol.) finely-divided aluminium chloride are suspended in
300 cc. carbon tetrachloride. To this are added through the course of
an hour or more, 156 grams (2 mol.) benzene mixed with about an equal
volume of carbon tetrachloride. By proper cooling and shaking the
reaction mixture should be kept below 30°. The following morning the
granular mass, consisting of the double salt of the keto-chloride and
aluminium chloride, is decomposed in the same bottle, preferably glass-
stoppered, by adding a considerable quantity of ice and shaking vigor-
ously. The temperature of the mixture is lowered to such an extent by
the ice and the hydrochloric acid set free that hydrolysis of the ben-
Zophenone chloride is reduced to a minimum. The carbon tetrachloride
solution, after drying over calcium chloride, is concentrated and the
residue distilled in vacuum. The resulting product contains, as a rule,
90 to 95 per cent. benzophenone chloride, the balance being benzo-
phenone. Redistillation in vacuum after addition of the calculated
amount of phosphorus pentachloride gives pure benzophenone chloride.
Mechanism of the Reaction.—The reaction between benzophenone
chloride and phenol proceeds, as previously mentioned, in several suc-
cessive steps. The first stage in all cases consists most probably in the
formation of the diphenyl ether of benzophenone:
OC6H5
I. (C.H.).cc. 4-2C.H.OH → (CH),C( + 2 HC1
OC6H5
Secondly, under the influence of the hydrogen chloride present, this
diphenoxy compound suffers an intramolecular rearrangement with
respect to but one of the phenoxy groups:
8
C6H4OH
OC6H5
The resulting phenyl ether, like all similar compounds, must be easily
decomposed by hydrogen chloride with the formation of the chloride of
p-hydroxytriphenylcarbinol, which, as Gomberg has shown, exists only
in the quinoid state. º
C6H4OH C6H4OH OH
och. Tº (chocó. -* (CH3)2C = ché.
And thirdly, at elevated temperatures this fuchsone hydrochloride
reacts readily with more phenol—the final step of the reaction—with
the production of di-p-hydroxytetraphenylmethane:
OC6H5
II. (choic. —- (C6H5)2CK
6.D.J.5
III. (C.H.).có
OH C6H4OH
IV. (C.H.).c = CHK. H. C.H.OH → (C.H.).c{ /
C1 Y C6H4OH
The reasons which force us to interpret the entire reaction according
to this above scheme are as follows: (1) It has been found possible
to adjust the conditions of the experiment—presence of solvent—so
as to obtain only diphenoxydiphenylmethane as the final product
(Equation I). (2) Either by starting with this diphenoxy compound
and exposing the same to hydrogen chloride (Equation II), or by carry-
ing out the condensation of benzophenone chloride and phenol in the
absence of solvent and at a sufficiently low temperature, we have ob-
tained as the main product p-hydroxytriphenylcarbinolehloride, isolated
as its hydrolyzed product, the carbinol:
C6H4OH
C1
(3) Furthermore, we have found that the di-p-hydroxytetraphenyl-
methane is formed not only by heating benzophenone chloride with phenol
as Mackenzie observed, but also by the transformation of diphenoxy-
diphenylmethane or by condensation of p-hydroxytriphenylcarbinol-
chloride with phenol (Equation IV). -
Thus it can be seen, that this reaction lends itself not only for the
preparation of the tetraphenyl compound, as Mackenzie found, but
also for preparing diphenoxy derivatives and p-hydroxytriphenyl- .
carbinols. The usefulness of this reaction becomes apparent when we
take into consideration its wide applicability, its simplicity, and the
yields of the products obtained. And in all these we feel that it is
superior to the previously described syntheses.
Diphenoxydiphenylmethane.—Mackenzie” in his attempts to prepare
diphenoxydiphenylmethane did not follow the reaction between ben-
zophenone chloride and phenol in any other solvent than an excess of
phenol and for this simple reason failed. By carrying out the above
reaction in benzene we have prepared the diphenoxy compound. Molten
V. (C6H5)2CCl2 + C6H5OH —2- (C.H.).có
9
phenol was added slowly to an equal weight of benzophenone chloride
mixed with a considerable volume of benzene, the reaction mixture
being kept at about 50°. A stream of dry air was drawn, under slight
vacuum, through the mass to remove the hydrogen chloride as formed,
and thus prevent the decomposition as well as the rearrangement of the
diphenoxy derivative. Concentration of the benzene solution and
addition of petroleum ether gave the diphenoxydiphenylmethane in
clusters of white needles melting at 132°, after recrystallization from
alcohol. The yield was about 85 per cent of the calculated amount.
Wieland" described as diphenoxydiphenylmethane a compound
which he had obtained by heating the diphenyl ether of benzpinacone
to 280°, resulting assumably as follows: sº
(C6H5)2C – C(C6H5)2 (C6H5)2C — OC5H5 -
2 | —- | + (C6H5)2C = C(C6H5)2
C6H5—O O — C6H5 C5H5—O
A sample prepared according to Wieland’s method was found to be identi-
cal with the product from the reaction of benzophenone chloride and
phenol in benzene solution. The melting point of a mixture of the two
corresponds to that of either product, 132°. Thus our results corroborate
Wieland’s conclusion as regards the constitution of the diphenoxy
derivative. Consequently they also lend support to his conception of
the constitution of the above mentioned diphenyl ether of benzpinacone
and that of its dissociation product, the corresponding free radical,
diphenylphenoxymethyl:”
(C6H5)2 = C – C = (C6H5)2 (C6H5)2 = C
—- 2
(C.H.) — () () — (C.H.) T (C.H.) – 6
Diphenoxydiphenylmethane is far more stable than dimethoxydi-
phenylmethane and its homologs, which as described' by Mackenzie
suffer decomposition even upon exposure to air. Neither boiling water
nor normal alkali has any effect upon the diphenoxy derivative. On
the other hand, like its analogs, diphenoxydiphenylmethane is hydrolyzed
by dilute acids, even by acetic acid, into benzophenone and the corre-
sponding alcohol, i. e., in this case, phenol. When exposed to hydrogen
chloride, as has been previously mentioned, it undergoes intramolecular
rearrangement, forming either triphenyl- or tetraphenyl derivatives
according to the conditions. In a quantitative experiment I.45 grams
of substance gave, when exposed to an atmosphere of dry hydrogen
chloride at room temperature, O.67 gram p-hydroxytriphenylcarbinol
but no di-p-hydroxytetraphenylmethane. A 3 gram sample, when
treated similarly but at 50°, gave I.3 grams of the carbinol and O.5.
gram of the tetra compound together with some unchanged diphenoxy
derivative. The significance of this latter result will be brought out
below under a discussion of the formation of the tetraphenyl derivative.
IO
The Preparation of p-Hydroxytriphenylcarbinol
Bistrzycki and Herbst” first prepared p-hydroxytriphenylcarbinol
through elimination of carbon monoxide from p-hydroxytriphenyl-
acetic acid by means of concentrated sulfuric acid. This step is accom-
plished with satisfactory yield although the preparation of the acid by
condensation of mandelic acid with phenol involves a rather laborious
process. Baeyer and Williger” obtained the identical product by
demethylating p-methoxytriphenylcarbinol by boiling the latter in a
mixture of acetic and sulfuric acids for twelve hours, the methoxy
carbinol having been prepared by Grignard's synthesis from anisic
ester and phenyl bromide. Gomberg” demethylated the methoxy
carbinol by means of aluminium chloride in benzene solution, a method
which gives at the same time more or less diphenylguinomethane and
other by-products. A highly satisfactory method of preparing this
carbinol is offered by the reaction of benzophenone chloride and phenol
under the following conditions.
Molten phenol (4 mol.) is chilled in a flask in such a way as to be
evenly distributed on the interior surface. Benzophenone chloride
(I mol.) is then added, the mouth of the flask being protected by a calcium
chloride tube. The reaction begins at once with evolution of hydrogen
chloride, and an occasional rotation of the flask brings fresh portions of
the phenol into reaction. After ten hours or more at a temperature
between 20 and 25°, the mass is subjected to steam distillation in the
same flask to remove the excess of phenol. The residue is digested with
five per cent. alkali and the alkaline solution extracted” with ether
in order to remove any benzophenone which may be present. After
filtration the dissolved ether is removed from the alkaline solution by
a stream of air. Addition of ammonium chloride or treatment with car-
bon dioxide liberates the carbinol and the dihydroxy-tetraphenyl com-
pound in the form of a paste, which changes on standing to a granular
mass. As a means of separation of the carbinol from the tetraphenyl
derivative either alcohol or 95 per cent. acetic acid may be used, as
the latter product is soluble in either solvent to an extent of not more than
o, I gram in 100 ce. Using about 6 cc. alcohol to each gram of the mixed
products dissolves the carbino1 and leaves a fine suspension of the tetra
compound. The filtration, often quite troublesome, may be facilitated
* This extraction is necessary at this point not only to remove the benzophenone
in suspension but also the benzophenone held in solution by the sodium salt of the
carbinol. The solubility of benzophenone under these conditions has been verified
by a blank experiment in which it was found that benzophenone, insoluble itself in
alkali, was jointly soluble with the carbinol in n-NaOH. In the extraction a large
amount of ether is to be avoided since the carbinol through hydrolysis of its salt is also
removed to some extent by this solvent.
II
by the addition of a few ce. Of a concentrated aqueous solution of sugar.
Addition of water and a few drops of ammonium hydroxide to the alco-
Holic filtrate precipitates the carbinol in a beautiful crystalline form.
Though hardly essential the carbinol may be recrystallized from benzene
in order to remove traces of the tetraphenyl derivative.
Following the above method we have obtained with samples of 12
to 18 grams of benzophenone chloride yields of 90 to 98 per cent. of the
calculated amount of the carbinol, with but a few hundred milligrams
of the by-product, di-p-hydroxytetraphenylmethane. The difficulty
of controlling the temperature of the viscous reacting mixture may cause
at times the formation of a gram or more of the tetra compound and a
corresponding decrease in the yield of carbinol. Adoption of the above
method was the result of a series of experiments in which the temperature
and the time of reaction, as well as the relative amounts of benzophenone
chloride and phenol, were varied. A few typical examples, tabulated
|below, will make evident the reasons for the selection of the described
procedure as the one most probable to give a maximum yield of the
carbinol from a given quantity of benzophenone chloride.
Chloride. Phenol. Temperature. Time. Carbinol. “Tetra.”
I2 g. 16 g. 90° 2 hrs. O g. I6 g.
I 2 3.6 90 4. 3.6 O
I 2 IO 4O-50 I 8.7 O .. 2
I 2 2O O-2O I 8. O O. 3
I 2 I7 25–35 I IO. 8 I . 8
I 2 2O 2O-25 I5 I 3. 4. I . 2
Di-p-hydroxytetraphenylmethane.—As has been described in part,
di-p-hydroxytetraphenylmethane results (1) from the action of dry
Bydrogen chloride on diphenoxydiphenylmethane, (2) from the conden-
sation of p-hydroxytriphenylcarbinol and phenol, using hydrochloric
acid as a catalyst, or (3) from the reaction of benzophenone chloride and
phenol at high temperatures. In the first case we may account for its
formation by assuming a double intramolecular rearrangement.
5 HC1 6H4OH Hol &º
OC6H C
(CH),C( —2- (C.H.).có ~ (C6H5)2C
5 C6H5 C.H.OH
Against this interpretation stands the objection that we have been
unable to isolate the intermediary mono-phenyl ether (II), but have
found that instead of this product, p-hydroxytriphenylcarbinol chloride
and phenol are produced. Consequently the tetraphenyl derivative
must result from the condensation of these two products upon heating.
That such may be the case has been actually verified by the following
experiment: Hydrochloric acid gas was passed into a mixture of 4
grams p-hydroxytriphenylcarbinol and Io grams phenol at 50° and after
I2
a short time 3.5 grams of the tetra compound were isolated. In acetic
acid solution, 3 grams carbinol and 3 grams phenol gave, after passing
in hydrogen chloride, 2.8 grams of the tetraphenyl derivative. As
regards the third method, Zincke heated three parts of benzophenone
chloride with four of phenol on the steam bath for three days. According
to our experience, this reaction is practically completed at the end of
but a few hours with a 90 per cent. yield of the tetraphenyl compound.
Di-p-hydroxytetraphenylmethane crystallizes from acetic acid in
glistening, white flakes or needles, melting at 286° without decompo-
sition. It forms a di-acetyl derivative by boiling in three times its weight
of acetic anhydride together with a small amount of sodium acetate.
Mono-p-hydroxytetraphenylmethane.—Finding the reaction between
p-hydroxytriphenylcarbinol and phenol to proceed so readily, we con-
cluded that we were dealing not with a case of anhydrolysis but with.
one of condensation under the influence of a catalyst. The correctness
of this inference was then tested in the case of mono-p-hydroxytetra-
phenylmethane which Baeyer and Villiger” had prepared by treating
triphenylcarbinol and phenol in acetic acid solution with a large quantity
of concentrated sulfuric acid. Our experience has shown that a dehy-
drating agent is not essential in order to bring about this condensation.
Upon addition of 1 cc. of concentrated hydrochloric acid to 2 grams
triphenylcarbinol and 3 grams phenol, heated on the water bath, the
tetraphenyl derivative began to separate in a short time, and at the end
of 15 minutes 2.2 grams of this compound had formed. It gave a melt-
ing point of 28.2°, as Baeyer and Villiger found. Similarly, one drop of
sulfuric acid gave even better yields, while in benzene solution the
carbinol and phenol condensed quantitatively in a short time under the
influence of but one drop of sulfuric acid.
HC1 {
(C6H5)3COH + C6H5OH —2- (C6H5)3C * C6H4OH + H2O.
The ease with which the p-hydroxyphenyl group is introduced stands
in sharp contrast with the difficulties encountered in the preparation
of the hydrocarbon tetraphenylmethane with its four phenyl groups.
This curious fact is as yet unaccounted for, the theory of steric hindrance
being inadequate to explain this difference.
III. Concerning the Free Radical, p-Hydroxytriphenylmethyl.
The general method for the preparation of free radicals in the tri-
arylmethyl series consists in subjecting the triarylmethylcarbinol halide
to the action of metals. A
R3CX + M —- R3C + MX.
Often the free radical thus produced exists wholly in the monomolecular
form, as indicated by the above equation. In other instances there is,
I3
without doubt, an equilibrium between the monomolecular and the
dimolecular forms:*
* 2 R3C Tº: [R3Cl2
This equilibrium is influenced in each particular case by the temperature,
by the nature of the solvent, as well as by the constitution of the radical
itself. In this paper when speaking of free radicals, the existence of
such an equilibrium is to be understood.
Unfortunately p-hydroxytriphenylcarbinol chloride” presents an
exception to the general reaction employed in preparing free radicals.
This had also been found to be the case with the chlorides of p-hydroxy-
benzo-Y-pyranols” and of p-hydroxyxanthenols.” The cause for this
exceptional behavior we believe to be the same in all cases, namely,
the chlorine and the hydroxyl group are linked to one and the same
carbon atom in these carbinol chlorides. The formation of the above
p-hydroxy chloride takes place by exposing in the solid form either
fuchsone or one of the two isomers of p-hydroxytriphenylcarbinol to
gaseous hydrochloric acid. The resulting products are identical hence
we must assume the usual equilibrium between two tautomeric isomers.
-- - OH
– = – /
(C6H5)2C K X O —- (C6H5)2C K DX.
C6H4OH C6H4OH
(C.H.).có —2- (C.H.).có
- OH C1
On treatment with molecular silver a molecule of hydrogen chloride
is split off instead of the usual single chlorine atom, giving consequently
fuchsone and not the free radical, p-hydroxytriphenylmethyl.
From these results it became evident that the only procedure for
the study of the free radical would come through protection or alteration
of the p-hydroxyl group. It had been previously found that, while
the hydroxyl group in the p-position tended to force the tautomeric
equilibrium toward the quinoid form, the methoxy, the acetoxy, and the
benzoxy group in the same position, on the other hand, cause their
derivatives to exist only in the benzoid form. Consequently it was
considered entirely feasible to prepare the free radical from the corre-
sponding stable chlorides, and the free hydroxy-radical according to the
following scheme:
C1
/ Ag
(C6H5)2 = C — CºHAOR —- (C5H5)2 = C – CºHAOR
H2O -
(C5H5)2 C — C6H4OR –2- (C6H5)2 = C — C6H4OH
in which R = CH3, COCH3, or COC5H5.
* As regards the constitution of the dimolecular modification of the free radical,
some assume it to be a molecularly associated complex of the free radical, while others
consider it as a hexarylethane, readily suffering dissociation into its halves.
-:
I4.
Realizing, however, the difficulties usually encountered in the removal
of any of these protecting groups, we turned first to the carboalkoxy
groups, used so extensively and successfully by Emil Fischer” as a tem-
porary protection of hydroxyl groups, in his recent studies on tannins.
A carboalkoxy group affects tautomerization advantageously by forcing
its derivatives into the benzoid form, and is readily removed through
bydrolysis, thus restoring the hydroxy compound.
p-Carboethoxytriphenylcarbinol.–14 grams p-hydroxytriphenylcarbinol
are dissolved in 55 cc. n-NaOH and a considerable amount of ice added
to the solution. To the suspension of the sodium salt of the carbinol
are added with vigorous stirring 6 grams (I. I mol.) chlorcarbonic ester.
The resulting stiff paste upon becoming granular is filtered out. After
drying partially, the product is recrystallized from alcohol by addition
of water. The yield is quantitative. -
O O
| |
C2H5O — C – C1 + NaOC6H4. (C6H5)2C — OH —- C2H5O—C—OC5H4. (C6H5)2COH.
p-Carboethoxytriphenylcarbinol is readily soluble in benzene, ether,
etc., crystallizing from these solvents upon addition of petroleum ether
as white needles, melting at I 19°. - •
p-Carboethoxytriphenylcarbinol Chloride.—According to the usual
method of preparing triarycarbinol chlorides, this chloride is formed by
passing gaseous hydrochloric acid into a benzene solution of the carbinol
in the presence of calcium chloride. Addition of petroleum ether to
the concentrated benzene solution gives white crystals of the chloride,
which after washing with a mixture of ether and petroleum ether, melt
at 98°. -
Calculated for C22H18O3Cl. : C1, 9.67. Found: C1, 9.62.
. The Action of Metals.—As has been mentioned, the usual result of
the action of molecular silver or other metals on the chlorides of the
triarylcarbinols in solution is the formation of the free radical with carbon
in the trivalent state. The unsaturated nature of this class of compounds
is illustrated, among others by the avidity with which they unite with
oxygen to form peroxides.
2 R3C + O2 —- R3C — O – O — CR.3
So characteristic is this reaction, that the isolation of the peroxide is
now generally accepted as sufficient evidence of the existence of the
corresponding free radical, though even temporary. Tested in this
manner the carboethoxy-chloride gives rise unquestionably to its free
radical. 3. -
p-Carboethoxytriphenylmethyl Peroxide.—When treated with molecular
silver out of contact with air a benzene solution of p-carboethoxytri-
phenylcarbinol chloride assumes after several hours a dark, cherry-red
I5
color. On exposure to air this color disappears in a way analogous to
that in the case of triphenylmethyl. Evaporation of the filtered ben-
zene solution and washing the residue with ether to remove any un-
changed chloride gives a relatively small amount of the peroxide, which
when recrystallized from benzene melts at 171 °. Much better yields
of the peroxide result from boiling the chloride and silver in benzene
with a stream of dry air passing through the solution. Like other
triary1 peroxides, p-carboethoxy-peroxide is but slightly soluble in ben-
zene and nearly insoluble in ether.
Calculated for C44H880s: C, 76.05; H, 5.52; Mol. Wt., 694.
Found: - C, 75.73; H, 5.56; Mol. Wt., 701.
All attempts to obtain the dihydroxy-peroxide by removing the
carboethoxy groups alone in this peroxide resulted in the splitting of
the product either to p-carboethoxytriphenylcarbinol or to p-hydroxy-
triphenylcarbinol. Warming the solution of the peroxide in a mixture
of acetic and sulfuric acids (4: I) gives the former carbinol in good yield.
The peroxide is not affected by boiling n-NaOH, but sodium ethoxide
or a solution of potassium or barium hydroxides in methyl alcohol de-
composes it to the hydroxy-carbino1.
Isomerization of the Free Radical. — The formation of the above-de-
scribed peroxide proves definitely that the free radical, p-carboethoxy-
triphenylmethyl, is formed as the initial step of the reaction between
the chloride and molecular silver. We next attempted to isolate the
unsaturated free radical in the solid state by concentration of the colored
solution of the same in an atmosphere of carbon dioxide. But the
residue thus obtained proved to be colorless, was found to be entirely
devoid of unsaturated properties, and would no longer yield a peroxide
on exposure to air, either as the solid or in solution. Consequently
we may assume that the free radical exists only temporarily, and must
have been affected in some way during the subsequent stages of its iso-
lation. In fact, even in the preparation of the peroxide through simul-
taneous action of silver and air upon the chloride, there is formed as a
coating on the silver more or less of this colorless, Saturated compound
in addition to the peroxide. More of this substance, and relatively
less of the peroxide, is obtained if the solution of the free radical is kept
for some time previous to oxidation; still more, if the solution is heated.
It is thus not surprising that we find upon boiling the benzene solution
of the chloride with silver for several hours, that practically none of
the peroxide can then be obtained. For under such treatment, the
reactive free radical is transformed completely into this colorless, inert
substance. - . " -
This, new product is a white, amorphous powder insoluble in ether
and but slightly soluble in boiling benzene or acetic acid. It melts
I6
at about 280° without decomposition. A sample from the last-men-
tioned solvent gave an analysis corresponding closely to that calculated
for the free radical. A molecular weight determined according to the
method of Menzies” proved, however, to be approximately twice that
of the free radical.
Calculated for C22H18O3: C, 79.72; H, 5.78; Mol. Wt., 331.
Found: C, 79.36; H, 6.04; Mol. Wt, 680, 664.
The above results indicate that this unsaturated compound is either
produced from the monomolecular free radical through polymerization,
or more likely, is a metamer resulting from, transformation of the di-
molecular modification of the free radical. A change of this kind has
been proved to take place in the case of the simplest free radical in this
series, namely, triphenylmethyl. It was shown by Gomberg” that this
dimolecular free radical is changed in benzene under the influence of
hydrochloric acid, as a catalyst, to a stable isomer, the constitution
of which has been definitely determined by Chichibabin” as p-benzhy-
dryltetraphenylmethane.
(HCl) h
2 (C6H5)3C —- (C6H5)2 = C – CºHA.C(C6H5)3.
Later Schlenk” found that this same transformation is brought about
by the action of metallic sodium on the free radical in ether solution.
The question arises: May not a similar metamerization take place
spontaneously in the case of certain other radicals? In other words,
may not the change of the free radical to its stable isomer proceed even
without the presence of a catalyst? We consider this entirely possible,
and, at the present time, this is the most plausible explanation that
we can offer for the relatively impermanent existence of the free radical,
p-carboethoxytriphenylmethyl. We have modified the methods in
attempting to prepare this radical by various means—by employing
different solvents, by using mercury or copper instead of silver, by con-
centrating the solutions at very low pressure and temperature—but
the result has invariably been the formation of the same stable metamer
of the free radical.
Acetoxy- and benzoxytriphenylmethyl chlorides behave in this respect
entirely analogously to the carboethoxy-chloride. They, too, give
free radicals which isomerize to stable metamers. The occurrence of
such spontaneous transformation we believe to be far more frequent than
has been assumed in the past. The apparent non-existence of the
free radicals, tri-p-tolylmethyl, di-p-tolylphenylmethyl, p-tolyldiphenyl-
methyl,” tri-p-anisylmethyl,” etc., may be explained as due to a simila
cause, namely, to isomerization. - w *
p-Benzoxytriphenylcarbinol.—Following the Schotten-Baumann re-
action, I4 grams p-hydroxytriphenylcarbinol are dissolved in an excess
\
17
of n-NaOH, 14 grams benzoyl chloride added and the mixture thoroughly
shaken for some time. The resulting oily precipitate solidifies on
standing. Its benzene solution is extracted with Io per cent. alkali
and dried over calcium chloride. Addition of petroleum ether gives
the benzoxy-carbinol in the form of small crystals, melting at 132°
after recrystallization from glacial acetic acid. We have nothing to
add to the properties as given by Bistrzycki and Herbst.” " -
p-Benzoxytriphenylcarbinol Chloride.—This chloride is prepared by
Saturating a benzene solution of the carbinol with hydrogen chloride in
the presence of calcium chloride. Addition of petroleum ether to the
concentrated benzene solution gives white crystals of the chloride in
nearly quantitative yield, melting point Iosº.
Calculated for C26H13O2C1 : C1, 8.90. Found: C1, 8.74.
p-Benzoxytriphenylmethyl Peroxide.—With molecular silver, a solution
of the benzoxy-chloride gives the usual deep color of a free radical.
On evaporation of this solution, however, a mixture of two substances
is obtained, the one the inert isomer of the free radical and the other the
peroxide. For identification the latter was prepared by simultaneous
action of silver and oxygen on the chloride: The peroxide is slightly
soluble in boiling benzene, nearly insoluble in ether, and melts at 167°
with decomposition. This peroxide is also readily hydrolyzed to its
corresponding carbinol by warming in a mixture of acetic and sulfuric
acids.
Calculated for C52H33O8: C, 82,29; H, 5.05. Found: C, 82.4o; H, 5.40.
As in the case of p-carboethoxytriphenylmethyl, we were unable to
isolate the free radical, p-benzoxytriphenylmethyl. On boiling the
chloride and silver in benzene, the resulting highly-colored solution of
the free radical loses its color, even during concentration, with the
formation of the colorless isomer. A sample of the latter, recrystallized
from acetic acid, melted at 266–9°, and gave an analysis and molecular
weight corresponding closely to that calculated for the dimolecular
free radical. l
Calculated for C52H38O4: C, 85.92; H, 5.27; Mol. Wt., 726.
- Found: C, 85.7O; H. 5.55; Mol. Wt., 7 Io.
p-Acetoxytriphenylcarbinol.—The following procedure has been adopted
for the preparation of this carbino1: 2O grams of p-hydroxytriphenyl-
carbinol and 3 grams of anhydrous sodium acetate are boiled for two
hours in 50 grams acetic anhydride. The product is precipitated by
pouring the above solution into IOO ce. water and, if necessary, may be
recrystallized from acetic acid. It possesses the properties described
by Bistrzycki” for this carbino1. * -
p-Acetoxytriphenylcarbinol Chloride.—Gomberg” has prepared this
chloride by the usual method, using calcium chloride as a dehydrating
I8
agent. We find it advisable to avoid the use of hydrogen chloride. The
carbinol is suspended in an excess of acetyl chloride (2: 3 mol.) and the
mixture heated until all goes into solution. The excess of solvent is
removed by careful heating. Upon the gradual addition of petroleum
ether and allowing to stand, the acetoxy-chloride separates out slowly
as white needles, melting at 88°. It is very soluble in benzene or ether
but only slightly so in petroleum ether. - ^
Calculated for Co. HiſO2C1: C1, Io.53. Found: C1, Io.52. -
p-Acetoxytriphenylmethyl Peroxide.—This peroxide must be prepared
by the simultaneous action of silver and oxygen on the chloride in solu-
tion. The resulting product, after recrystallization from benzene and
ether, melts at 172°. It is readily hydrolyzed to the corresponding
carbinol.
Calculated for C40H840s: C, 79.46; H, 5.40.
Found: C, 79.67; H, 5.8O.
In an analogous way, the acetoxy polymer was prepared as a white,
amorphous powder, the purification of which gave a product melting
with decomposition at 255–270°. In its low solubility, its non-con-
version through hydrolysis into the corresponding carbinol, and its
lack of unsaturated properties, this polymer resembles the other analogs.
as described. -
Iv. Concerning p-Hydroxytriphenylmethyl Ether.
A number of compounds, presumably triarylmethyl ethers, have been
described in the literature, but it was recently shown” that these products
could not possess the constitutions attributed to them. On certain
negative results, Schlenk” concluded that the simplest member of this
series, triphenylmethyl ether, is apparently incapable of existence.
Gomberg,” on the other hand, has worked, out a method by means of
which this substance, as well as its analogs, can be prepared, and has
found that these ethers or oxides possess a fair degree of stability.
A product described as p-hydroxytriphenylmethyl ether” appeared
especially doubtful on the evidence presented. On account of the ease
of hydrolysis of the carboethoxy group, the preparation of the corre-
sponding ether seemed to offer a ready means of obtaining this p-hydroxy-
ether. - - -
p-Carboethoxytriphenylmethyl Ether.—Io grams p-carboethoxytri-
phenylcarbinol chloride, 30 grams mercuric oxide and 50 cc. benzene are
sealed and shaken for a month at room temperature. The resulting
solution, after warming, is filtered from the mercuric oxide and chloride
and concentrated. Upon addition of petroleum ether the product is
precipitated as a yellow crystalline mass, which after recrystallization.
from benzene and absolute ether is white and melts at 219°.
19
Calculated for C4H8807: C, 77.84; H, 5.65; Mo1. Wt., 678.
- Found: C, 77.80; H, 5.74; Mol. Wt., 694.
The p-carboethoxy ether resembles other triaryl ethers, being very
readily hydrolyzed by acids. Thus it is converted into the carbinol
by boiling in acetic acid or in alcohol containing a few drops of sulfuric
acid. - -
. As a rule, the carboethoxy group is easily removed by cold dilute
alkali or aqueous ammonia, acetone or pyridine being added if necessary
to increase the solubility of the products.” Treated in this manner,
p-carboethoxytriphenylmethyl ether does lose its carboethoxy groups,
but unfortunately suffers further hydrolysis to p-hydroxytriphenyl-
carbinol. Various conditions and hydrolytic agents of alkaline nature
have been employed but as yet it has been found impossible to remove
only the carboethoxy groups and thus obtain the p-hydroxy-ether.
It is reasonable to suppose that the first action of the hydrolysis does
consist in the destruction of the carboethoxy groups, and the hydroxy-
ether so produced must become further hydrolyzed in its turn. In
other words this hypothetical ether, unlike other triarylmethyl ethers,
is hydrolyzed to its carbinol not only in the presence of acid but also in
the presence of alkali. It will be recalled that a similar behavior was
met with in the case of the corresponding p-carboethoxytriphenylmethyl
peroxide.
W. Summary.
1. A detailed study of the reaction between benzophenone chloride
and phenol has been carried out. Contrary to previous reports in the
literature, the reaction was found to proceed in several successive stages,
and the mechanism of the reaction has been studied.
2. A simple and excellent synthetic method for the preparation of
p-hydroxytriphenylcarbino1 has been devised, a method which promises
to become a valuable means for obtaining analogous compounds.
3. p-Hydroxytriphenylcarbinol chloride on treatment with metals
gives up hydrochloric acid and not chlorine, thus making it impossible
to prepare the corresponding free radical according to the usual method.
4. p-Carboethoxy-, p-benzoxy-, p-acetoxycarbinol chlorides do give
the respective free radicals, but these were found to be too unstable
to be isolated as such.
5. From the experimental facts obtained, an explanation is derived
to account for the impermanent character of these free radicals, which
may also serve to explain similar results previously reported by others.
6. It was found impossible to obtain p-hydroxytriphenylmethyl ether,
by partial hydrolysis of its carboethoxy derivative, since at the same
time the latter is hydrolyzed completely to p-hydroxytriphenylcarbino1.
:
VI. Bibliography.
. Ber., 34, 3818 (1901). - -
. Ber., 40, 1860 (1907); Ber., 42, 406 (1909).
. Ber, 40, 1847 (1907). -
. Gomberg and Cone, Ann., 370, 193 (1909). -
. Gömberg and Cone, Ann., 370, 150 (1909); Ann., 376, 184 (1910).
. Gomberg and West, J. A. C. S., 34, 1562 (1912).
7.
Schmidlin, “Das Triphenylmethyl,” p. 206; Schlenk, Ann., 372, 16 (1909);
Jacobson, Ber, 38, 196 (1905); Gomberg and Cone, Ber, 39, 3274 (1906).
me
8.
9.
IO
II
I 2
I3
I4.
I5
I6
I7
I8
I9
2O.
2 I
22
23
24
25
26
27
28
thyl,’’ p. 24.
29
3O.
3I
32
33
34.
35
36
37
Schlenk, Ann., 372, 16 (1909).
Gomberg and West, J. A. C. S., 34, 1529 (1912).
Gomberg, J. A. C. S., 35, ro35 (1913).
Mackenzie, J. Chem. Soc., 79, 1209 (1901).
Smedley, J. Chem. Soc., 87, 1252 (1905).
Zincke, Ann., 363, 279 (1908).
Sachs and Thonet, Ber., 37, 3329 (1904).
Boeseken, Rev. d. trav. chem., 24, I (1905).
Wieland, Ber., 44, 2554 (1911).
Ber., 44, 2551 (1911).
Ber., 34, 3073 (1901); Ber. 35, 3133 (1902).
Ber., 36, 2791 (1903).
J. A. C. S., 35, 209 (1913).
Ber., 35, 3018 (1902).
Gomberg, J. A. C. S., 35, IO4I (1913).
Gomberg and Cone, Ann., 370, 203 (1909).
Gomberg and West, J. A. C. S., 34, 1562 (1912).
Ber., 46, 3253 (1913); J. A. C. S., 36, 1170 (1914).
J. A. C. S., 32, 1615 (1910).
Ber., 36, 378 (1903). . .
Ber., 37, 4708 (1904); Ber., 41, 2422 (1908); Schmidlin, “Das Triphenyl-
Ber., 47, 1665 (1914). A
Gomberg, Ber., 37, 1627 (1904).
Schmidlin, “Das Triphenylmethyl,” p. 153.
Ber., 34, 3077 (1901).
Ber., 34, 3076 (190},... . . .
Gomberg, J. A. C. S., 35,306 (1913).
Ann., 394, 178 (1912). -
Bistrzycki and Herbst, Ber., 34, 3075 (1901).
Fischer, Ber., 46, 3257 (1913).
Jul 3 1915
a . g . . 4
3.
::::::::::
* * * ..}{{...º.
''': lºt is, ***, * 1:
{{: ; ; ; ; ; ; , , , ; !;
* * ::::::::::::::
*
* * . ;
tº
: ; , ;
" * * * * * 1: , ; , ;
. . . . * * * * * *
** : * * * *
* * * * * *
*
i
|
* *
*** * * * *
*... : * * *
# 8 º
}
it; . . . . . .
; :::::::::::::::,
*::: ; ; ; ; ; ; ;
* * * * $ 1. w &
* * * * *
* * * : * *
* : * r * ,
it is...}{i};
: ; ; ; ; ; ; ; ; ;:
* * * * * *
* *
* ,
t
# * \, ,
": { * * : i , ,
* * * : . . .
4
* * * i
# * *
a
* * * * * * * *
* : * * * , , ,
* * * * *
* * * * * * * .
tº 8 # * * *
* *
* tº li, , ! # tº
* * * * * *; , , * * * * ,
* * * * * * * * * * ,
* {
UNIVERSITY OF
||||
O15 O770
|||||

tº:
§§
}*, *, *.
º
... r. º §
- º; x *
º
a - - #
sº sº.
.*.*.*.*.*, ****
f
Żºł
*
3!º