SOME ATTEMPTS TO PREPARE
TRIPHENYLCARBINYL P-TOLUENESULFONATE,
TRITYL TOSYLATE
BY
John Vincent Killheffer, Jr.
Submitted in Partial Fulfillment
of the Requirements for the
Degree of Bachelor of Science
at the
Massachusetts Institute of Technology
1950
Signature of Author
yV
''
Certified by
Thesis Supervisor
Department[ Head
ACKNOWLEDGMENT
The author recognizes that his indebtedness
to Professor C. Gardner Swain for his endless
patience is greater than he can repay,
Table of Contents
Page
Introduction
.
.
.
.
.
.
.
.
.
I
.
.
0
1
.
.
.
1
.
Object of Investigation.
Apocamphyl Tosylate .
.
Past Attempts.
.
.
.
.
Phenylcarbinols
.
.
.
.
.
.
.
.
Trityl Tosylate
.
.
.
.
.
.
.
.
2
Displacement Synthesis .
.
.
0
0
3
Esterification Synthesis
.
.
.
0
0
3
Difficulties of Esterification
.
.
.
3
Side Reactions with Amines.
.
0
0
Separation of Alcohol from Ester.
.
.
.
4
4
Sensitivity to Water.
0
5
.
.
.
.
.
.
0
Other Tosylation Methods
.
.
.
0
0
5
Silver Tosylate-Trityl Chloride Reaction
5
p-Toluenesulfonic Anhydride
6
.
.
.
6
Triphenylmethoxi4e-Sulfonyl Chloride
Reaction .
.
.
.
.
.0.
.
.
.
.
.
.0.
Results and Discussion .
.
.
.
.
The Tipson Procedure.
.
.
.
.
Acid-Alcohol Reaction
Acetyl Tosylate
.
...
.
Ditrityl Ether.
.
.
.
.
.
Tritylpyridinium Chloride .
,
.
p-Toluenesulfonic Anhydride
.
.
Triethylamine .
.
.
.
.
.
.
7
7
8
.
8
8
00
.
.
10
.
0
11
Page
.
.
.
.
12
Silver Tosylate- Trityl Chloride
Reaction
.
12
Water Reaction .
.
.
.
Alcohol-Acid Dehydration.
Experimental Part .
.
Triphenylcarbinol .
Tosyl Chloride .
12
.
.
.
.
.
14
14
.
.
.
.
.
14
.
.
.
.
14
.
.
.
.
.
Pyridine .
.
.
.
.
14
Petroleum Ether.
.
.
14
.
.
.
14
Acetyl Chloride.
.
Benzene
Chloroform
.
.
.
.
.
15
Benzenesulfonyl Chloride.
.
.
.
.
15
Anhydrous Oxalic Acid.
.
.
.
.
15
.
15
.
15
.
15
Triethylamine
.
.
Silver Oxide.
.
.
.
.
.
.
Silver Tosylate.
p-Toluenesulfonic Acid
16
Trityl Chloride.
Ditrityl Ether ....
.
.
.
.
16
.
.
.
16
Trityl Tosylate.
17
Benzenesulfonic Anhydride
.
.
.
.
18
Trityl Tosylate.
.
.
.
.
.
.
.
19
List of References.
.
.
.
.
.
.
.
21
I
Introduction,-
The present project was undertaken with the object
of developing a general method for the synthesis of sulfonic
esters of tertiary alcohols. Only one such ester has been reported
in the literature, viz., 1-apocamphyl p-toluenesulfonatel. This
compound was prepared in 34% yield by heating 1-apocamphanol with
p-toluenesulfonyl (tosyl) chloride in pyridine for twenty hours
on the steam bath. Its recrystallization from 50% ethanol gives
some indication of the abnorm&l properties of this substance: an
open-chain tertiary tosylate would be expected to hydrolyze very
rapidly.
Attempts have been made by several workers to prepare tertiary
sulfonates. Heating t-butyl alcohol with tosyl chloride in pyridine at 1000 yielded only t-butyl chloride.a Gilman and Beaber3
reported "several unsuccessful attempts" to prepare the tosylates
of t-butyl alcohol and triphenylcarbinol. The method employed by
these workers for several other esters, and presumably also for
the tertiary ones, was the addition of powdered potassium hydroxide to a mixture of alcohol and acid chloride in ether solution
at temperatures not allowed to exceed 40*4 These three articles
are the only mentions made of such compounds, except for the assertion of Tipson5 that he was not aware of reports of any sulfonates
of tertiary hydroxy compounds.
Preparation of the sulfonates of even primary and secondary
phenylcarbinols, e. g., benzyl alcohol
nol 0, and ethyl
f-hydroxy-F-phenylpropionate
,
phenylmethylcarbi-
1,
seems to present
especial difficulty, as evidenced by the hardships encountered in
early attempts. The tosylates of the two secondary carbinols are
2
too unstable to be isolated, though their reactions can be studied
by forming them in the reaction mixture by oxidation of the corresponding sulfinates. Benzhydryl tosylate has not been reported,
and experiments by A, J. DiMilo1 2 also indicate that its synthesis
may be very difficult.
Triphenylcarbinyl (trityl) p-toluenesulfonate was chosen as the
model compound to synthesize. The reasons for this choice are
briefly set forth below.
Triphenylcarbinol was selected because of its ready availability, its complete lack of o(-hydrogen atoms, and because trityl
sulfonates were desired for other research. Its structure precludes dehydration, which could become a hazard with t-butyl alcohol, e. g.
A tosylate is, in general, to be preferred over a benzenesulfonate from the preparative viewpoint, since it is more likely
to be higher-melting and easier to crystallize. Some representative melting points are presented foV comparison in Table I.
Table I
M.p,, benzenesulfonate
Liquid 2 6 ,2 7 ,2 8
Liquid 29 ,30 ,3 1
Methanol
M.p., tosylate
28013
Ethanol
32-3*14,15
Phenol
p-Nitrophenol
o-Cresol
94-5*16;95-6018 34-5*; 35*
75033
810 19;81.520
79-85*33; 82034
97-97*5*2l
39-40*35; 35-6*22
54-5*18,22
m-Cresol
48022; 51018
p-Cresol
69 -70*
p-Naphthol
Guaiacol,
125023
85024
105-7*36
80*38
L-menthol
97025
8038
Hydroxy compound
o-Nitrophenol
45022
;67-8*243*'
I
Displacement of a tertiary halide by a metal sulfonate was rejected as a method of synthesis. In general, the polar solvent
necessary to dissolve ionic sulfonates is unsatisfactory for
tertiary halides, they being reactive toward it or insoluble in
it. To this diaadvantage may be added the fact that the Walden
Inversion on a tertiary carbon atom is not easy, and in the case
of sulfonates, the position of the equilibrium might very well
be unfavorable. Thus, it is stated39 that "...alkyl sulfonates
react (with sodium iodide in acetone), producing the corresponding sodium sulfonates as precipitates."
The only remaining satisfactory method for preparing a tertiary sulfonate seemed to be the direct esterification of an
alcohol. The reaction
RSOX + R'OH
-
RS30R' + HX
may be successfully carried out only if the following three conditions are satisfied.
1. The reaction between the hydroxy compound and the ester,
to form an ether, must be slow enough to be unimportant.
2. The hydrogen halide evolved must have little effect on
the hydroxy compound or the ester.
3. The excess hydroxy compound must be easily removed.
Each of these conditions may present difficulty in the present
problem. Tipson's recommendation5 in the case of ether formation
is the use of excess hydroxy compound and low temperatures, the
attendant disadvantages being the slowness of the reaction and
the waste of reagent. The second possibility, that of reaction
of the ester or the alcoholwith the hydrogen halide formed in
the reaction, has been dealt with by earlier workers in a number of ways. Aspiration of air through the reaction mixture was
used40.
In other experiments, bases were provided for the acid
to react with; e. g., sodium hydroxide4
diethylaniline
, and pyridine
,
sodium carbonate
,
.
If amines are employed, new side-reactions must be considered.
Quaternary salts may form. This reaction takes place negligibly
slowly at 0*,
according to Sekera and Marvel45. Another possible
reaction in the presence of tertiary bases is chlorination.
p + C5H5NH1
ROSO 2 0 6 H40H --
--
R0l +C
5 H 5 NH+p-CH
C6H430
This transformation has been encountered in work on phenols4 6 ' 4 7
simple alcohols4 '
,
5 0 51 5 2
and polyhydroxy compounds such as sugars ' ,
A further complication is the difficulty of removing any unreacted hydroxy compound from the ester, In the worst possible
situation, mixed crystals of the two compounds could form, as has
been noted in studies on trityl fluoride.53 Triphenylcarbinol
is a solid and probably trityl tosylate will also be a solid,
and there exist no really good methods of separating two neutral
types like this. Probably the only means available which do not
involve jeopardizing the ester are these:
1. Extraction between some combination of nonpolar solvents.
2. Fractional crystallization from nonpolar solvents, e.g.,
diethyl ether, chloroform, carbon tetrachloride, benzene,
acetonitrile, nitromethane, halobenzenes, benzonitrileetc.
The second of these methods is especially likely to be tedious,
and the determination of the best solvents for a useful extractive separation is not likely to be an easy task. The extreme reactivity of trityl compounds towards water must be kept in mind
during all these considerations. Just this fact, for example,
makes the useful preparation of primary tosylates exemplified
for n-butyl by Roos, Gilman, and Beaber54 , involving aqueous
sodium hydroxide, appear to be impracticable in this synthesis.
A solution of trityl chloride in benzene is quantitatively hydrolyzed by a brief shaking with water in a separatory funnel 5 5
and trityl tosylate should be even more easily hydrolyzed. For
example, ethyl chloride is hydrolyzed by heating with concentrated alkali in an autoclave at 125056. Ethyl tosylate is saponified
in a comparable length of time at room temperature in a 30%
ethanolic solution containing a trace of sodium hydroxide57
Other tosylation methods include the following:
1. Reaction of silver tosylate with trityl chloride.
2. Reaction of p-toluenesulfonic anhydride with triphenylcarbinol.
3. Reaction of sodium triphenylmethoxide with tosyl chloride.
4. Removal of water from an equimolar mixture of p-toluenesulfonic acid and triphenylcarbinol.
Silver salts usually lead to effects different from those of
sodium salts, because of the participation of the heavy metal ion
in the reaction. The silver salt of p-toluenesulfonic acid might
succeed where the sodium salt would not by virtue of a prior
front-side attack by silver ion rather than a Walden Inversion
on the highly hindered trityl group. Trityl perchlorate58 and
ditrityl sulfate59, 6 0 have been prepared by reaction of the silver salts of the appropriate acids with trityl chloride. These
acids are of strength comparable to that of p-toluenesulfonic,
giving promise for the trityl tosylate preparation.
The anhydride of p-toluenesulfonic acid would have an advantage over the chloride in that no chlorination could take place.
A drawback in this approach lies in the difficulty of preparing
the anhydride itself. No really satisfactory synthetic method
has been evolved for this type of compound. Shepherd61 rejected
older methods for preparing this type with the statement that
they gave "variable results." His own procedure could not be
duplicated by Miller
nor in the present investigation, even
though several attempts were made. The reaction used by Meyer
and Schlegl 6 3 for the preparation of p-toluenesulfonic anhydride
(reaction of thionyl chloride and p-toluenesulfonic acid) gave
large proportions of chloride as well as anhydride; moreover,
treatment of methanesulfonic acid with thionyl chloride has been
suggested as one of the better methods of preparing methanesulfonyl chloride 64.
It seems quite possible that one of the original methods may
be the most dependable yet published6 6 ,6 7 ,6 5 . A preparation of
sulfonic anhydrides in consistently good yields would certainly
be of interest in this program.
The triphenylmethoxide-sulfonyl chloride reaction would not
be without analogy, various esters having been prepared by this
means6 8 ,6 9,70,71, though it may be significant that benzyl esters
are not among them. The reaction using potassium benzyloxide and
tosyl chloride yielded no ester
-.
Only half the sulfonyl chlo-
ride was consumed in the reaction and the main product (isolated
in 48% yield) was dibenzyl ether. The authors suggested that the
ester formed, but was as reactive as the starting chloride with
respect to the benzyloxide, displacement of tosylate from carbon
occurring exclusively rather than that of chloride from sulfur.
The dehydration of the acid-alcohol mixture also can be supported by analogous examples, sulfates being commonly prepared
in this fashion, although ditrityl sulfate has been prepared,
apparently, only by the reaction of trityl chloride and silver
sulfate in benzene60 or liquid sulfur dioxide59. Trityl perchlorate has been successfully prepared ,y the method of removing
water by aspirating airthrough a mixture of triphenylcarbinol
and perchloric acid in nitrobenzene, then precipitating the product with benzene7 3 . The synthesis of trityl chloride by treatment
of triphenylcarbinol with dry hydrogen chloride7 4 may also be
cited as an example of this sort of synthesis.
The common preparation of trityl chloride by allowing acetyl
chloride to react with triphenylcarbinol75 leads to speculation
regarding the outcome of interaction between triphenylcarbinol
and acetyl tosylate. Compounds of this type have been reported 7 6
8
Results and Discussion.- The preliminary investigation of this
problem showed that a promising first step was to attempt the synthesis by the procedure outlined by Tipson5 . AccordinglY, triphenylcarbinol was treated with tosyl chloride in pyridine at
about -5*. Dilution with water produced a copious precipitate,
which was worked up. This substance was triphenylcarbinol.
The recovery of triphenylcarbinol could be explained in a number of ways. No reaction at all may have occurred; esterification
could have proceeded, with water hydrolyzing the ester;esterification could have been followed by ether formation, with final
hydrolytic decomposition of this product.
The contingency first considered was that of ether formation.
Since demonstration of the stability of ditrityl ether to pyridinium ion would effectively dispose of this possibility, an attempt was made to synthesize this ether by one of the two published methods, Gomberg's reaction of trityl chloride with mercuric oxide7 7 .
o ether could be obtained by this procedure and
another attempt, by the second method (formation of ditrityl
carbonate by treatment of trityl chloride with silver carbonate,
followed
by thermal decomposition of this ester to the desired
product78 ), was being considered when it was decided that ether
formation was not likely to be the main trouble. The ether preparation was postponed and eventually abandoned.
It seemed likely that there had been simply no reaction under
the mild conditions of Tipson's procedure, and a duplicate experiment was carried out at the temperature of the steam bath, and
__
2.
another at reflux, other conditions being the same as previously.
Some crystalline material was obtained by these high-temperature
procedures which had m.p. 170-1740
(uncorr.), higher than any of
the starting materials. Purification yielded crystals, m.p. 1721750,
qualitative analysis of which showed that it contained no
sulfur, but that it did contain nitrogen and halogen. The elements
present, coupled with the m.p., was good evidence that the product was tritylpyridinium chloride. This compound has been previuosly prepared by two groups of workers7 9 ,80, with whose data
the above m.p. agrees well. The shape of the crystals and the color
of their pyridine solution also compared well with the data of
the German workers.
The conclusion to be drawn from the formation of this compound
is a little more clean cut than in the prior case of triphenylcarbinol recovery. Pyridinium salt formation occurred. Probably
the desired esterification took place, and was followed by an
attack on the ester by either chloride ion or pyridine. In the
first of these possibilities, the trityl chloride formed could
79
react with pyridine, in the fashion observed by Norris and Culver
to give tritylpyridinium chloride, the observed product. The
attack by pyridine would lead to tritylpyridinium tosylate and
precipitation of the chloride would be merely a result of its
lower solubility. An alternative proposal for the mechanism involves the direet production of trityl chloride
from tosyl
chloride and triphenylcarbinol.
Another possible process is the formation of trityl chlbride
10
from triphenylcarbinol and chloride ion, catalyzed by the pyridinium ion formed in the solution. This theory is in qualitative agreement with the fact that pyridinium salt formation is,
in general, an unimportant reaction with primary and secondary
alcohols except those which are easily chlorinated by the Lucas
Reagent, e.g. Especially to be noted in this connection is the
behavior of benzyl alcohol, a primary alcohol noted for both ease
of reaction with the Lucas R-eagent and ease of pyridinium salt
formation in tosylation 5. The apocamphanol case also fits, since
although it is a tertiary alcohol, it is unaffected by the ordinary reagents which convert alcohols to the corresponding halides
Apparently the quaternary salt formation is not limited to
high temperatures. A solution of triphenylcarbinol and tosyl
chloride in pyridine deposited crystals of this substance upon
standing for several weeks.
If the first of the mechanisms is the correct one (i.e.,
at-
tack on trityl tosylate by chloride ion rather than pyridine),
then an attempt to synthesize trityl tosylate by reaction of triphenylcarbinol and p-toluenesulfonic anhydride offers hope,
since no chloride ion need be present in such a reaction mixture.
Attack by tosylate would not affect the product (unless it were
optically active). Accordingly, efforts were directed at the
preparation of this anhydride. The method involving
heating
tosyl chloride with oxalic acid61 was selected, since a large
quantity of tosyl chloride was on hand, and oxalic acid is readily available as the dihydrate. The choice was further influenced
11
by Shepherd's implication that this method was the best available.
In several attempts to repeat the preparation described for
benzenesulfonic anhydride, no crystalline material could be
isolated, but only black semisolid masses which could not be
purified. Reference to the work of Miller 6 2 disclosed that his
efforts to duplicate the benzenesulfonic anhydride preparation
gave results of similar nature. This approach was abandoned
forthwith.
At this point, a reappraisal of the processes of pyridinium salt formation suggested a new approach. It is conceivable
that the chloride ion attacks only the conjugate acid of the
ester or alcohol, in equilibrium with the pyridinium ion in the
solution. If such is the case, the use of a stronger base would
displace this equilibrium in favor of the ammonium ion and decrease the rate of chlorination, Pyridine and diethylaniline,
the only two amines which have been used for this purpose, may be
insufficiently basib, their dissociation constants being 1.4 x 10~9
and 3 x 10-8 (81,82), respectively. Accordingly, some experiments
were carried out using triethylamine, with a constant of 5.65
x 10-4 (83), as the acid acceptor.
The first of these experiments involved the substitution of
triethylamine for pyridine in the procedure of Tipson. The reaction
was carried out at room temperature, and only triphenylcarbinol
was obtained. After heating under reflux, the product was a dark
mass which appeared to react very rapidly with the moisture in
12
the air to give a crumbly brown product from which no crystalline material could be isolated. Additionof water gave triphenylcarbinol.
The addition of a large amount of water to the reaction mixture to stop the reaction and precipitate the ester is an integral part of all the tosylation procedures in the literature,
However, the addition of water to reaction mixtures encountered
in this investigation may have hydrolyzed any trityl compounds
present back to triphenylcarbinol.
The successful preparation of ditrityl sulate by the action
of trityl chloride on silver sulfate59, 6 0 lent support to an
attempt to prepare trityl tosylate in an analogous manner. Silver tosylate and trityl chloride were heated in benzene. Crystalline material separated upon cooling, and this solid was
purified, being recrystallized under anhydrous conditions, It
was apparently unreacted trityl chloride. It seems likely the
the silver tosylate was not finely divided enough and soon became coated with silver chloride, stopping the reaction. Further
study of this reaction seems warranted, with attention to the
facts that the silver tosylate is sensitive to light and that it
should be pulverized for this reaction.
An attempt was made to prepare trityl tosylate by removing
water from a mixture of triphenylcarbinol and p-toluenesulfonic
acid. Gentle heating of a small sample of a mixture of the two
compounds in the absence of a solvent yielded a brilliant red
melt which cooled without decolorizing. Heating to the boiling
point produced a very dark green substance which had the appearance of the melt of triphenylcarbinol and p-toluenesulfonyl
chloride, which do not give the red color upon gentle heating,
however.
Gomberg used nitrobenzene as the solvent in his preparation
of trityl perchlorate
but p-toluenesulfonic acid monohydrate
is not very soluble in nitrobenzene, either alone or when the
solvent contains a large amount of triphenylcarbinol. Dry air
was aspirated through a mixture of alcohol and acid in nitrobenzene. No crystalline material
mixture.
separated from the reaction
14
Experimental Part.- Triphenylcarbinol (student preparation) was
recrystallized from 95% ethanol, yielding practically colorless
crystals, m.p. 158-161*. Complete evaporation of the solvent
left a gummy orange material, probably mostly biphenyl.
Triphenylcarbinol (Eastman white label) was recrystallized from
95% ethanol. Rapid cooling of the solution gave white microcrystalline material, m.p. 160-160.5*.
p-Toluenesulfonyl chloride (Eastman white label) was dissolved
in cold dry benzene, and after filtering to remove insoluble
material, was shaken in a separatory funnel with 5% aqueous
sodium hydroxide. The layers were separated and the benzene solution was dried over anhydrous potassium carbonate. Filtration
and evaporation of the filtrate to a small volume gave large
colorless prisms, m.p.
67
-94, which turned pink on standing for
a few weeks. The mother liquor was used for subsequent recrystallization of other batches.
Benzene (Eastman pure) was dried by distillation until the distillate came over clear. The residue was used.
Pyridine (Eastman pure or Mallinckrodt Analytical Reagent) was
dried over 97% sodium hydroxide pellets for a week, then filtered and distilled. The fraction boiling at 112-5* was used.
Petroleum ether (b.p. 35-60*) was washed with concentrated sulfuric acid until the acid layer was colorless, then with water,
The material was dried over calcium chloride, filtered and distilled, b.p. 45-55*.
Chloroform was purified by washing with concentrated sulfuric
acid, then water. The chloroform was dried over calcium chloride
and distilled. The fraction boiling near 610 was collected.
Acetyl chloride (Eastman practical) was mixed with one-tenth
its volume of dimethylaniline and the resulting blue solution
was slowly fractionated through a six-inch column packed with
glass beads. The fraction boiling at 500 was collected.
Benzenesulfonyl chloride (Eastman practical) was distilled at
the pressure of the water aspirator, b.p. 180-190* under
varying pressure. The distillate was slightly yellow. A dark
high-boiling residue remained.
Anhydrous oxalic acid was prepared by heating the dihydrate
(Mallinckrodt Analytical Reagent) in an oven above 100* for
about forty-eight hours. It was also prepared by drawing dry
air slowly through a flask charged with the dihydrate for fortysix hours. The flask was heated by means of an oil bath at 1300:
Some sublimation was noted.
Triethylamine (Sharples technical) was distilled. The fraction
boiling at 8P-90*
was collected.
Silver oxide was prepared by adding 40 cc. of 10% sodium hydroxide to a solution of 18.2 g. of silver nitrate in distilled
water. The brown solid was filtered off and washed with water.
It was immediately utilized in the preparation of silver p-toluenesulfonAte.
Silver p-toluenesulfonate. The silver oxide from the above
preparation was suspended in distilled water and 18.4 g. of
p-toluenesulfonic acid was added. Warming of the mixture over
16
a flame produced a sudden vigorous reaction. A heavy white
granular precipitate formed, was filtered off and dried. It
quickly darkened upon exposure to light.
p-Toluenesulfonic acid was prepared by boiling tosyl chloride
with water until solition was complete,about three hours. Evaporation of the water and recrystallization from ethanol afforded
the monohydrate, m.p. 1010.
Triphenylmethyl chloride was prepared according to the method
of Organic Syntheses 82. All-glass apparatus was used, and 25 g.
of triphenylcarbinol, with corresponding amounts of other reagents. The weight of the unrecrystallized product was 23 g. (78%)
m.p. 110-2*0.
The synthesis was repeated, using 20 g. of triphenylcarbinol
and proportionate amounts of other reagents. The yield was 16.6
g.,2 (78%),
m*p.
110-2*.
An attempt was made to prepare ditrityl ether by the method of
Gomberg 8 3. Trityl chloride (21 G.) wasdissolved in dry benzene
and 24.5 g. ef mercuric oxide (Mallinckrodt, yellow, C.P.) was
added. The suspension was refluxed for eight and one-half hours,
and allowed to cool. The supernatant liquid was decanted into
a separatory funnel and enough 0.1% hydrochloric acid was added
to give clear phases after shaking. The benzene layer was dried
over calcium chloride, filtered, and concentrated under reduced
pressure. The orange solid mass was recrystallized from benzene
but remained highly colored. A small amount of orange, partially
crystalline material, m.p. 126-136*, resulted after another at-
tempt to recrystallize. This was discarded. The ether washings
of the original crop deposited a small amount of material which
was recrystallized from benzene with the aid of Norite. This
substance proved to be triphenylcarbinol, identified by its m.p.
and mixed m.p. with an authentic sample.
Several attempts were made to prepare trityl tosylate. In following the procedure outlined by Tipson5 , a copious precipitate
formed upon addition of water. This solid was filtered off and
washed with water until free of pyridine. Recrystallization from
ether gave triphenylcarbinol, m.p. 160-160.5*.
Evaporation of
chloroform extracts of the filtrate gave only a small additional
amount of triphenylcarbinol.
A second attempt was made, using the amounts of reagents suggested
by Tipson. The mixture was heated on the steam bath for two hours.
Upon cooling, some crystalline material was deposited. After filtering, the solution was diluted with water, again yielding triphenylcarbinol. The crystals which had been removed melted at 1701740. Recrystallization from the yellow pyridine solution gave
lozenge-shaped plates, m.p. 172-175*, solidifying and remelting
at 172-173*. Qualitative analysis84 showed the absence of sulfur
but the presence of nitrogen and halogen. It was concluded that
the compound was tritylpyridinium chloride 8 5 ,8 6.
An experiment was carried out in which the reaction solution was
refluxed for three hours. Concentration yielded 7.4 g. of crystals
which were combined with those from the steam bath experiment.
An experiment was carried out according to Tipson's alternate
18
procedure in which 5N sulfuric acid was added to the reaction
mixture after fifteen minutes to prevent pyridinium salt formation. The usual precipitate of triphenylcarbinol formed and
was recrystallized from ethyl actate, m.p. 1610.
Triethylamine (41 cc.) was substituted for the usual 100 cc.
of pyridine in one experiment. Addition of water produced a
two-phase mixture, acidification of which gave one liquid phase
and a dirty white precipitate of triphenylcarbinol. Boiling
the precipitated alcohol with water containing sodium hydroxide removed the acid chloride. Recrystallization of the alcohol
from ether gave m.p. 1610.
Refluxing a suspension of tosyl chloride (7.6 g., 0.04 moles),
triphenylcarbinol (10 g.,
0.0385 moles) in 41 cc. of triethyl-
amine (these were the usual amounts used) for three hours followed by overnight cooling gave a brown semisolid mass. When
an attempt was made to filter the ice-cold mixture, it became
very viscous, then crumbly. The whole product was dissolved in
chloroform and attempts were made to obtain crystalline material,
but all such attempts failed.
3enzenesulfonic anhydride was the object of several synthetic
61 , oxalic acid
attempts. Following the procedure of Shepherd
(4.5 g., 0.05 mole) and freshly distilled benzenesulfonyl chloride (17.7 g., 0.10 mole) were heated in a flask to 150-160*.
This temperature was maintained until no more sublimation appeared
to be occurring. The flask was then heated briefly to 200*, attached to the water pump, and kept at 180* until volatile
material stopped distilling. The mixture was by this time quite
dark in color, and at the end of the prescribed heating period
was a dark oil from which no crystals separated upon cooling.
In several successive attempts to carry out this reaction,
only black oils resulted, some of which partially solidified upon
cooling. Attempts to crystallize these mixtures led to isolation only of small amounts of oxalic acid, identified by its
behavior on heating in a melting point capillary.
Similar results were obtained when Shepherd's procedure was
followed using oxalic acid (9.0 g., 0.1 mole), and p-toluenesulfonyl chloride (38.2 g.,
0.2 mole). The reaction flask was
swept with dry air, the effluent fumes being strongly acidic to
moistened pHydrion paper. Attempts to recrystallize the black
solid resulting from this procedure gave only oxalic acid.
An attempt to prepare trityl tosylate was carried out by dissolving trityl chloride (14. Og., 0.05 mole) in 25 cc. of benzene and adding silver tosylate (14.0 g., 0.05 mole). The bright
yellow suspension was refluxed seven and one-half hours and allowed to cool. Some crystals formed around the walls of the flask.
They appeared to be greenish in the solution, which was heated
and filtered through a filter stick constructed according to
FieserF7 . The benzene was removed under reduced pressure, care
being taken to exclude moisture. The brown residue was recrystallized several times from carbon tetrachloride, yielding a substance of m.p. 85-90*, containing so sulfur nor nitrogen, but
containing halogen. This compound was probably impure trityl
20
chloride.
An attempt to carry out the removal of water from a mixture
of triphenylcarbinol and p-toluenesulfonic acid was made. Triphenylcarbinol (2.6 g.,
0.01 mole) and p-toluenesulfonic acid
(monohydrate, Eastman white label; 1.9 g., 0.01 mole) were mixed
in 78 cc. of nitrobenzene. Addition of the acid produced a
bright orange color not observed with triphenylcarbinol or the
acid alone in nitrobenzene. The flask was attached to the vacuum pump as if for distillation, with a fine capillary extending to the bottom of the flask. A calcium chloride tuba was placed
on the inlet of the capillary and another in the line between
the flask and the dry ice-acetone trap protecting the pump. Air
was drawn through the mixture at a pressure of about 3 mm.
for
twelve hours. The weight of the second calcium chloride tube
had not changed appreciably at the end of this period. The reaction mixture had become quite dark, though it remained clear.
Nothing separated from the solution on standing for one week,
nor in two days after dilution with benzene.
21
References
1.Bartlett and Knox, J. Am.
Chem. Soc., 61, 3184 (1939).
2. Bennett and Hafez, J. Chem. Soc., 656 (1941).
3. Gilman and 3eaber, J. Am. Chem. Soc., .41
518 (1925).
4. Hahn and Walter, Ber., 54B, 1531 (1921).
5. Tipson, J. Org. Chem., 2, 235 (1944).
6. Ferns and Lapworth, J. Chem. Soc., 101, 273 (1912).
9.
Medvedev amd Alekseeva, 3er., 653, 131 (1932).
10. Kenyon, Phillips, and Taylor, J. Chem. Soc., 173 (1933).
11. Kenyon, Phillips, and Shutt, J. Chem. Soc., 1663 (1935).
12. A. J. DiMilo, Unpublished work.
13. Ullman and Wenner, Ann., 12Z,
14. Jaworski,
Chem. Zeitung,
121 (1903).
221 (1865).
100 (1867).
15. Otto, Ann.,
142,
16. Otto, Ber.,
12, 1883 (1886).
17. Otto, Ber.,
19,
1832 (1886).
18. Reverdin and Crepieux, Bull. soc. chim., (3)
27, 745 (1902).
19. Ullman and Loewenthal, Ann., 1}2, 63 (1904).
20. Bamberger and Rising, Ber., 24, 241 (1901).
21. Bamberger and Rising, Ber.,
4, 240 (1901).
22. Badische Anilin- und Soda-Fabrik, Ger. Pat. 162,322.
23. Reverdinand Crepieux, Bull. soc. chim., (3)
24. Reverdin and Crepieux, ibid., (3)
5,
1047 (1901).
2L, 1046 (1901).
25. Hilditch, J. Chem. Soc., 22, 238 (1911).
22
26. Wegscheider and Furcht, Monatsh., 2J, 1095 (1902).
27. Krafft and Roos,
Ber.,
?5, 2257 (1892).
28. Praetorius, Monatsh,, 26, 8 (1905).
29. Praetorius, ibid., 2Z, 784 (1906).
30. Billeter, Ber., JZ, 691 (1904).
31. Krafft and Roos, Ber., E,
32.
691 (1904).
Schiaparelli, Gazz. chim. ital., ll, 66 (1881).
33. Georgescu, Buletinul Societatii de Sciinte din Bucuresci,
8, 673; Beilstein, Vol. XI, p. 31.
34. Sciaparelli, Gazz. chim. ital., 11, 70 (1881).
35. Georgescu, Buletinul Societatii de Sciinte din Bucuresci,
8, 668; Beilstein, Vol, XI,.p.
31.
36. Georgescu, Buletinul Societatii de Sciinte din Bucuresci,
3, 417; Beilstein, Vol. XI, p. 31.
37. Georgescu, 3uletinul Societatii de Sciinte din Bucuresci,
8, 669; Beilstein, Vol. XI. p. 31.
38. Patterson and Frew, J. Chem. Soc., 8Q, 333 (1906).
39. Shriner and Fuson, "The Systematic Identification of Organic
Compounds," Wiley, Kew York, 1948, p. 141.
40. Slotta and Franke, Ber., §J, 678 (1930).
41. Georgescu, Ber., 24, 416 (1891).
42. Slotta and Franke, Ber., 0,
678 (1930),
43. Sane, Chakravarty, and Parmanick, J. Indian Chem. Soc., 2,
55 (1933).
44. Reverdin and Crepieux,,Ber.,25, 1439 (1902).
45. Sekera and Marvel, J. Am. Chem. Soc.,
53, 345 (1933).
46. Ullman and Nadai, Ber., 4I, 1870 (1908),
47. Sane and Joshi, J. Chem. Soc., 125, 2481 (1924).
48. Butler, Renfrew, Cretcher, and Souther, J. Am. Chem. Soc.,
a,
227 (1937).
50. Ohle, Euler, and Lichtenstein, Ber., 62, 2885 (1929).
51. Hess and Pfleger, Ann., 507, 48 (1933).
52. Hess and Stenzel, Ber., 68, 981 (1935).
53. Swain, Esteve, and Jones, J. Am.
Chem. Soc., 11, 965 (1949).
54. Org. Syntheses, Coll. Vol. I, 145 (1941).
55. Swain, J. Am. Chem. Soc., 70, 1119 (1948).
56. Izmail'skii and Popov, J. Gen. Chem. (U.S.S.R.), 8, 965 (1938);
C.A. 21,
1263 (1939).
50, 60 (1931).
57. Demeny, Rec. trav. chim.,
58. Gomberg and Cone, Ann.,
170, 194 (1909).
3543 (1904).
59. Gomberg and Cone, Ber., a.
60. Gomberg, Ber., 40, 1849 (1907).
61. Shepherd, J. Org. Chem., 12, 275 (1947).
62. Miller, Ph. D. Thesis, Mass. Inst. of Tech., 1949.
63. Meyer and Schlegl, Monatsh., 34, 561 (1913).
64. Noller and Hearst, J. Am. Chem. Soc., 70, 3955 (1948).
65.
Billeter, Ber., 38, 2016 (1905).
66. Abrahall,
J. Chem.
67. Hubner, Ann.,
Soc.,
49,
692 (1886).
223, 244 (1884).
68. Schieler and Otto, Ber., 2
,
1638 (1876).
24
69.
Otto, Ber., 19, 1832 (1886).
70. Hubner, Ann.,
223, 235 (1884.)
71. Bourcet, Bull. soc. chim., (4) _1l' 363 (1912).
73. Gomberg and Cona, Ann., 2[_0, 194 (19090.
74. Gomberg, 3er., 0,
75.
2401 (1902).
Org. Syntheses,23, 100 (1943).
76. 3aroni, Atti accad. naz. Lincei, (6) 17, 1081 (1933);
Chem.
Zentr.,
77. Gomberg,
I, 1183 (1934).
J. Am. Chem.
78. Halford, J. Am.
Soc.,
2,
Chem. Soc., 31,
79. Norris and Culver, Am.
205 (1913).
2158 (1929).
Chem. J., 29 134 (1903).
80. von Meyer and Fischer, J. prakt. Chem., 82, 523 (1910).
81. Lange, "Handbook of Chemistry," 7th Ed., Handbook Publishing Co.,
Sandusky,
1949,
p. 1410.
82. Hall and Sprinkle, J. Am.
Chem. Soc.,
4, 3469 (1932).
83. Lange,"Handbook", p. 1410.
84. Shriner and Fuson, "Identification of Organic Compounds, " p. 52.
85.
Norris and Culver, Am. Chem. J.,
2,
134 (1903), report
"large pink crystals," m.p. 167-167.5*.
86. von Meyer and Fischer, J. prakt. Chem., 82, 523 (1910),
report colorless prisms, m.p. 171*; the solution of their
compound in pyridine was bright yellow.
87. Fieser,"Experiments in Organic Chemistry", 2nd Ed., Heath,
Boston, 1941, p. 333.