United States Patent Office 3,235,583

Patented Feb. 15, 1966

1. 2
3,235,583 able on the market, but other like apparatus adapted to
PREPARATION OF ASPRN provide agitation, heat and means for liquid recovery
Robert T. Edmunds, Norwich, N.Y., assignor to The may be employed.
Norwich Pharmacal Company, Norwich, N.Y., a cor In the practice of my invention, the method which I
poration of New York currently prefer consists in mixing acetic anhydride and
No Drawing. Fied July 22, 1964, Ser. No. 384,544 salicylic acid in a glasslined jacketed, conical dryer
3 Clains. (C. 260-480) blender which is revolvable on its central axis. I have
found that an excellent yield of aspirin of exceptional
This invention relates to an improved method of pre quality can be obtained through the practice of my proc
paring aspirin (acetyl salicylic acid) and is a continua ess when a slight molar excess of acetic anhydride is
tion-in-part of my earlier filed pending application, Serial used. The quantity of acetic anhydride exceeds by about
No. 126,440, filed July 14, 1961, and now abandoned, 2% to about 20% (18% being preferred) the stoichio
which, in turn, was a continuation-in-part of my co metric equivalent of salicylic acid that is charged into the
pending application Serial No. 819,236, filed June 10, reaction vessel. An inert, organic diluent such as acetic
1959, and now abandoned. 5 acid, benzene, toluene or xylene may be added advan
For many years the method customarily employed to tageously, but is not essential to the reaction. The re
prepare aspirin of a degree of purity that is acceptable action mixture is heated through the jacket of the dryer
for pharmaceutical purposes, and in a form that lends blender by the introduction of steam, or hot water or
itself to the ready tableting of the aspirin, involved a other suitable heat transfer medium to about 45° to 95
multi-step method which is both cumbersome and ex 20 C. Thereafter the pressure within the reaction vessel
pensive. In that method, salicylic acid and acetic an is reduced by exhaustion to produce a partial vacuum
hydride were reacted at elevated temperature and the while maintaining a jacket temperature of about 45°
product of the reaction was crystallized through con 95 C. The internal pressure is reduced initially to
trolled cooling in such manner that the largest possible about 120-170 mm. of mercury absolute and progressive
amount of the end product recovered thereby would be 25 ly decreased at a rate of about 5 to 50 mm. per hour
in the desired crystalline form most readily adapted to until visible distillation of liquids ceases. For complete
tableting. The end product was then filtered out, washed removal of vapors and drying of the aspirin crystals, the
and dried but it was frequently necessary to resort to re pressure is preferably reduced to somewhere between
crystallization to obtain the desired crystalline form of the minimum obtainable in the system and 30 mm. and
end product and to free it from impurities. The filtrate 30 held at this pressure for several hours. The crystalline
which was produced in this process was distilled to ob product is readily removed from the reactor by dumping.
tain additional quantities of the end product and also While any inert diluent distillable under the above
to recover solvents and unrelated reagents for recycling. conditions may be used advantageously in the above
I have discovered that it is possible to prepare aspirin method, acetic acid is particularly desirable since this
through a vastly simplified method which results in a 35 permits the recovery of both the diluent and by-product
very substantial reduction in the cost-thereof, and it is acetic acid in a single step, in a relatively pure form
a particular feature of my invention that the aspirin suitable for recycling as the diluent for subsequent re
that can be produced thereby is not only substantially actions.
pure but it is also in the form of the desired needle-like The percentage of impurities (normally less than .5%
crystals which most readily lend themselves to easy 40 by U.S.P. test procedures) contained in the aspirin pro
tableting. duced through my new method can be further reduced
My method is based on the discovery that crystalline by introducing into the reaction vessel from 1 to 2% by
aspirin can be produced by heating a mixture of salicylic weight based on the total weight of the reaction mixture,
acid and a fractional molar excess of acetic anhydride of a quench consisting of water or acetic acid or a mix
and directly distilling to dryness at elevated temperatures ture thereof.
and under reduced pressure; the acetic anhydride and The quench is introduced after substantially complete
salicylic acid can be reacted and distilled to dryness in distillation of the reaction mixture and prior to dumping
a single step in one vessel from which nearly theoretical of the crystals, preferably shortly after minimum pres
yields of substantially pure crystalline aspirin are di sure is attained in the reaction vessel. The quench is
rectly obtained. This eliminates filtration, washing and added to prevent trace amounts of acetic anhydride
drying of crystals and workup or refinement of process (which might remain after distillation) from reacting
liquors, operations which require additional equipment with the aspirin crystals to form sodium carbonate in
and which are inherent in past methods. soluble compounds. While these sodium carbonate in
A particularly surprising feature of my method is that soluble compounds, formed in the absence of a quench,
distillation to dryness has produced aspirin of over 99.5% 55 appear to be unstable and disappear upon storage, they
purity, whereas it would be anticipated on the basis of nevertheless prevent the crystals obtained directly from
prior experience that impurities would be formed in ex the reactor from passing the full U.S.P. test procedure.
cess of 0.5%. In addition to the high purity of my Thus, addition of the quench in large scale operations
aspirin, near theoretical yields of aspirin and by-product facilitates the recovery of U.S.P. aspirin crystals directly
acetic acid are obtained by my method. These high 60 from the reaction vessel.
yields are primarily due to the fact that there is no oc In order that my invention may be readily available
casion for loss of these materials through the precipitation, to and understood by those skilled in the art, the follow
filtration, washing and reworking of mother liquors that ing specific examples thereof are given:
have been necessary in the past.
The salicylic acid and acetic anhydride may be re Example I
acted and distillation of the reaction mixture under re
duced pressure to recover liquids and to secure a crystal The following runs were made in a round bottomed
line product may be conveniently and Satisfactorily car Pyrex flask of 1,000 cubic centimeter capacity and the
ried out in apparatus such as stainless steel or glasslined flask was attached to a rotary vacuum connection that
blenders, ball mills and the like which are equipped to 70 would allow the flask to be laid upon its side and rotated
agitate the reaction mixture. I now prefer to use a coni in a water bath at elevated temperatures and reduced
cal, jacketed, dryer-blender of the type presently avail pressures.
 3,235,583
3 4.
(a) 140 grams of salicylic acid and 110 grams of distiliation under vacuum and then the acetic acid and
acetic anhydride (a 6.8% molar excess) were heated at finally the acetic anhydride were introduced by applying
75 C. for 90 minutes and the vacuum then applied at a slight vacuum to the unit. After loading, the blender
about 150 mm. of mercury absolute pressure. The vac was closed and started revolving, while water was circu
uum was gradually lowered to about 30 mm. of mercury 5 lated through the jacket at a temperature of 90° C. When
absolute pressure over a period of about 90 minutes and the temperature within the unit reached about 88 C. as
held at 30 mm. for an additional 30 minutes at said ele- indicated by a dial thermometer installed in one of the
vated temperature. A yield of 179 grams out of a theo- loading ports, vacuum was applied and the pressure re
retical of 181 grams of substantially pure aspirin crystals duced to about 170 mm. of mercury absolute pressure.
was recovered directly from the flask. The U.S.P. test 10 This
for sodium carbonate insoluble materials on the product
pressure was held for one hour after which the ab
solute pressure was lowered approximately 15 mm. of
was negative indicating that no anhydride derivatives of mercury per hour for 4/2 hours, at which point the ab
aspirin had been formed. Impurities of about 1% un- solute pressure was lowered at a rate of 120 mm. of
reacted salicylic acid, however, were found. mercury per hour for the next hour to the minimum ab
(b) 140 grams of salicylic acid, 120 grams of acetic 5 solute pressure attainable in the system; i.e., 20 and 30
anhydride and 142 grams of toluene as a diluent were mm. of mercury absolute. During the controlled de
placed in the flask and heated to 85° C. for about 40 crease in absolute pressure, the liquids in the reactor dis
minutes. The absolute pressure in the fiask was reduced tilled at a steady rate and by the time that minimum ab
to 170 mm. pressure at said temperature and held for 3 solute pressure had been obtained the reactor contained
hours. The absolute pressure was then gradually lowered 20 crystals of aspirin which were relatively dry. In this par
to 2 mm. of mercury over a period of about 30 minutes. ticular series of runs an additional 2 to 3 hours at mini
Aspirin crystals were recovered directly from the flask mum pressure were allowed with the jacket circulating
and washed with carbon tetrachloride and dried in an temperature at 90° C. and then the crystals were dumped
electric oven at 50° C. for 2 hours. 180 grams of aspirin and weighed. Tests of crystals in this series were re
were obtained and 211 grams of solvent (toluene and ex- ? stricted to an assay for percent aspirin and for free sali
cess acetic anhydride) and by-product acetic acid were re- cylic acid as specified in the U.S.P. In Run No. 5, a
covered. The yield of aspirin based on salicylic acid was pite series p U.S.P. tests for aspirin was carried out
99% of theoretical and the recovery of solvent was an t ," E. w satisfactory, A fluctuation In yield
96.5%. Melting point of finished material was 135-138 30 beromnoted.
82% This
of thefluctuation
t eoreticalin toyield
1.04% of theoretical
is due will
to the fact that
C. Free salicylic acid was .01% U.S.P. test and the ma- a certain amount of materal adhered to the walls of the
terial assayed 100.1% aspirin by U.S.P. test. reactor in each run. The reactor was not cleaned between
(c) 140 grams of salicylic acid, 122 grams of acetic an- runs so that after the first run or two almost theoretical
hydride (a 17.5% molar excess) and 95 grams of acetic yields were obtained. It was observed by opening the re
acid were reacted for 16 hours in the apparatus described 35 actor that the material left on the walls from run to run
above in a water bath held at 45 C. The absolute pres- actually dissolved and therefore was eventually recovered.
TABLE
U.S.P. tests
Run. Notes reactor Salicylic acid, Acetic Acetic acid, Water jacket Aspirin yield,
No. Clean | los, anhydride,
lbs. bS. temp., C. percent
theoryof Aspirin assay Free salicylic Insolubility Other tests
percent acid test tests 1

69 56 25 90 82 00.7
69 56 25 90 92 99.78
69 56 25 90 91 99.8
69 56 25 90 00 99.5
69 56 25 90 104 99.5
69 56 25 90 100 99.79

1 Test for materials insoluble sodium carbonate.

sure was reduced to 120 mm. of mercury to produce Example III

distillation and gradually reduced to 20 mm. of mercury Data presented in Table II represent runs made in a
until a dry product was obtained (about 2 hours). This 55 plant or commercial scale 110 cubic foot glassline rotary
product was found to contain 35% unreacted salicylic conical vacuum-dryer made by Glasscote Products, Inc.,
acid with the rest being aspirin. The recovered weight of and similar in design to the 5-cubic foot unit used to col
dry material was 163 grams which checks closely with the lect the data in Table I. A manufacturing change from
theoretical weight that should be obtained for a mixture the small scale work was the installation of a 400-gallon
of 65% aspirin and 35% salicylic acid, resulting from the 60 glassline jacketed scale or weighing tank into which the
starting materials. U.S.P. test for materials insoluble in salicylic acid, the acetic acid, and the acetic anhydride
sodium carbonate solution was negative. were charged. Water was circulated through the jacket of
(d) The above run was repeated without the acetic acid. this tank at 88 C. and the tank agitated until solution
In this case the mixture was held at 45 C. for 20 hours was obtained. By adding warm acetic acid and acetic
before distillation was started, and the resulting dry crys 65 anhydride to this tank, solution was obtained in approxi
tals contained only 20% unreacted salicylic acid. The mately the same elapsed time as in the smaller unit. With
total solids recovered were 170 grams. The above two the material in solution, vacuum was applied to the re
runs indicate that the reaction can be expected to proceed actor (the dryer blender) and the solution in the scale tank
toward completion as the time is increased at these rela was pulled through a stainless steel filter into the reactor.
tively low temperatures. 70 The use of the scale tank allows convenient, accurate
weighing of all ingredients and also permits filtration of
Example II the reaction mixture, a desirable precaution even when us
The runs included in Table I were carried out as foll ing U.S.P. quality raw materials. After the solution was
ows: Salicylic acid was loaded directly into a five cubic in the reactor, the cycle was quite similar to the smaller
foot glasslined, jacketed, conical dryer-blender adapted for runs. The pressure was reduced and held for one hour
 3,235,583
5 6
at 150 mm. mercury absolute, the absolute pressure was of acetic acid was approximately 75% due to the many
further reduced at a rate of 15 mm. per hour for 3% opportunities for losses inherent in the processes.

Water jacket
temp., C. U.S.P. tests
Acetican- Aspirin, Acetic
Run Notes re- Salicylic hydride, Acetic yield per- acid, yield
No. actor clean acid, lbs. lbs. acid, lbs. cent of percent of
Above Below theory theory Aspirin Free Sali, Insolubil- Other
break break assay, per- cylic acid ity tests 2 tests
point point cent test y

1,395 1,119 626 88 () 98.5 97 99.83 OK.------- OK.------. OK.

1,395 1,119 626 88 50 99.2 95 99.99 OK.------- OK. ------ OK.
1,395 1,122 626 88 72 98.9 97.5 99.99 OK.-- OK.------- OK.
1,395 1,122 626 88 72 99.0 96.0 99.523 OK.------ OK.------- OK.
1 Steam off.
2 Test for materials insoluble Sodium carbonate.

hours and at this point the pressure was reduced at the rate What I claim is:
of 120 mm. per hour until a minimum absolute pressure 1. In the method of preparing acetyl salicylic acid by
20
of 20 to 30 mm. of mercury was attained. At the break reacting a mixture of salicylic acid and acetic anhydride
point where the more rapid decrease in absolute pressure at an elevated temperature of 40° C. to 95 C. the im
was begun, various temperatures of the water circulating provement which comprises:
in the reactor jacket were tried as indicated in the table. employing a molar excess of about 2 to about 20% of
In the large scale runs the addition of water or water 25 acetic anhydride;
mixed with acetic acid, provided a practical method of reacting said mixture in a closed vacuum equipped ves
preventing the formation of “materials isoluble in sodium sel at said elevated temperature;
carbonate.” In Run No. 7, 10 lbs. of water in 30 lbs. of maintaining said elevated temperature and reducing the
acetic acid were added when the absolute pressure reached internal pressure to a partial vacuum within the range
its minimum point; i.e., 20-30 mm. In Run No. 8, 20 lbs. 30 of about 120 mm. to 170 mm. of mercury absolute
of water in 20 lbs of acetic acid where added at approxi- and thereafter gradually reducing said pressure to a
mately the same point. In Run No. 9, 20 lbs. of water in range between about 30 mm. and the minimum pres
20 lbs. of acetic acid were added at the point at which visi- ure obtainable in the system;
ble distillation had ceased, approximately 40 minutes af- maintaining said elevated temperature and reduced pres
ter the minimum absolute pressure was reached. In Run 35 Sure at the lower range for about 1 to 3 hours; and
No. 10, 20 lbs. of water in 20 lbs. of acetic acid were thereafter recovering acetyl salicylic acid directly from
added at about 20 minutes after minimum pressure was said vessel in a pure dry crystalline form.
reached. In each case the conditions were maintained 2. The method according to claim , wherein an inert
until all distillation ceased. The material produced was distillable diluent selected from the group consisting of
suitable for use preferably after deodorization and screen 40 acetic acid, benzene, toluene and xylene is added to the re
ing. Removal of residual trace cdors can be accom action mixture prior to distillation.
plished in various ways; (1) by application of an absolute 3. The method according to claim 1, wherein about 1%
pressure in the range of 3 mm. of mercury for a period of to 2% by weight, based on the total weight of the reac
about 40 minutes; (2) by drawing dry air at room tem tion mixture, of a quench selected from the group con
perature through the material for a period of 15 hours; or 45 sisting of water, acetic acid and mixtures thereof, is added
(3) by passing the material once through a standard tubu to said vessel after distilling to substantial dryness and
lar hot air rotary dryer. one-half to two hours prior to recovery of the crystals
The large scale reactor tests showed less of a tendency from said vessel.
for the material to adhere to the walls as can be seen by
the fact that starting with a substantially clean reactor, a 50 References Cited by the Exagainer
yield of 99% of the theoretical based on salicylic acid was UNITED STATES PATENTS
obtained in most cases. This high yield coupled with the 2,423,569 7/1947 Sokol -------------- 260-480
good recovery of by-product acetic acid; i.e., in the neigh- 2731492 1/1956 Kamlet. 260-480
borhood of 97%, represents a substantial Saving over 55 3,061,632 10/1962 Surine et al.
earlier aspirin processes where the over-all yield of aspirin
260-480
was in the range of 91% of theoretical and the recovery LORRAINE A. WEINBERGER, Primary Examiner.