- Email: [email protected]
Decomposition Of Aspirin In The Solid State*
SCIENTIFIC EDITION
May 1958
and reported total alkaloid content as ergotamine in three different samples. The values were reported as 110, 102, and 375 mg. %. No other determinations were made.
SUMMARY AND CONCLUSIONS
Ergot of couch grass was studied with respect to the fat, moisture, and alkaloid contents. It was found that half of the sample was infested with a beetle, and determinations were extended to find out what effect this infestation had on the contents of the drug. It was found that: 1. The moisture content of both infested and noninfested portions was the same. 2. The fat content of infested ergot was lower than that of the noninfested portion. 3. The alkaloid content of the infested ergot was lower than that of the noninfested portion. 4. Both portions assayed above N. F. limits in ergotoxine content, and below N. F. limits in ergonovine content.
329
5 . Evidence was presented to show that ergotamine, ergotaminine, ergosine, ergosinine, ergocornine, ergocorninine, ergocristine, ergocristinine, ergocryptine, ergocryptinine, and lysergic acid were probably present. 6. Ergonovine was positively shown to be present. REFERENCES (1) Youngken, Heber W., “Textbook of Fharmdcognosy,” 6th Ed., Blakiston C o . , New York., 1948, pp. 85-91. (2) Ibid. p. 132-137. (3) “Th; National Formulary,” 10th ed., Mack Publishing C o Easton Pa. 1955, p. 217. (4) ’Silber, A,, aAd Schulze, T . , Pharmasie, 8, 675(1953). (5) “Pharmacopeia of the United States,” 15th rev., Mack Publishing. Co., Easton, Pa.. 1955, p. 1094. ( G ) , Personal contact with C. Pugh, Research Laboratories, Eli Lilly and Co., Indianapolis, Ind. (7) Kolsek, J., Microchim. Acta, 9, 1377(195G). (8) Tyler, V. E., Jr., and Schwarting, A. E., THIS JOURNAL, 41, 345(1952). (9) “The National Formulary,” 10th ed., Mack Publishing Co., Easton, Pa., 1955, p. 218. (10) Ibid p 216. (11) Madski. R. H. F., and Holme?. H. L.. “The Alkaloids, Chemistry and Physiology,” Academic Press, Inc., New York. N. Y. 1952 p. 377. (12j Silbhr, A,, and Dischoff, W.. Arch. Phorm.. 288. 124 (1955).
Decomposition of Aspirin in the Solid State* By LEWIS J. LEESON? and ALBERT M. MATTOCKS The decomposition of aspirin in the solid state has been demonstrated to be dependent on vapor pressure and temperature. A possible mechanism of decomposition is offered, which consists of an initial sorption of a water layer by each particle, diffusion of aspirin into solution, and decomposition taking place by acid catalyzed hydrolysis. A series of equations which describe the mechanism have been derived and the experimental data demonstrated to fit them. Using these equations, the various kinetic constants have been determined.
comprise a large S segment of the drugs marketed today, there is a great need for methods of predicting their INCE SOLID DOSAGE FORMS
stability under various conditions of storage. In order for such a method to be dependable it must be based upon a knowledge of the mechanisms involved and quantitative studies of the variables affecting the reactions. Many pharmaceuticals are known to exhibit decompositionin solid form, and perhaps the most widely used one is aspirin. Aspirin is an especially interesting example of this type of reaction since it is known to be affected not only by temperature and humidity but also by numerous chemical agents with which it is often combined. Decomposition of aspirin in solid form has been noted by Tsakalotos (l), Paolini (2), and
* Received May 3. 1957, from the College of Pharmacy, University of Michigan. Ann Arbor. Presented to the Scientific Section, A. PH. A. New York meeting, April-May, 1957. t Parke. Davis Fellow and American Foundation for Pharmaceutical Education Fellow.
Strathopoulos (3). In addition, Ribeiro, et al. (4), found that certain lubricants, notably stearates, increased the degree of decomposition of aspirin contained in aspirin-phenacetin-caffeine tablets. Ebert (5), and Yamamoto and Takahashi (6, 7) studied the accelerating effects of certain mines, temperature, humidity, pressure, and grinding upon aspirin loss in tablets. These studies were concerned primarily with effects of various chemical agents and conditions and did not attempt to determine reaction rates or mechanisms. There have been a number of kinetic studies on aspirin decomposition in solution (8-11), the most complete investigation being reported by Edwards (12). He agreed with other workers that the decomposition is first order with respect to aspirin concentration, but by studying the reaction at various pH values he demonstrated that the apparent first order rate constant is actually a composite of six rate constants. Each
330
JOURNAL OF THE AMERICANPHARMACEUTICAL ASSOCIATION Vol. XLVII, No. 5
of t h e six constants is associated with one of the hydrolytic reactions depending upon hydrogen ion, hydroxyl ion, or water reacting with either molecular or ionic aspirin. There is no direct method for applying t o solid forms information obtained from studies of decomposition of a drug i n solution, since i t is even questionable whether the same reactions occur. T h u s i t is important t h a t studies be made of reactions of drugs in solid state, and such a study was undertaken in these laboratories using aspirin as the first example.
a t 50°, 60", and 80";and a t varying humidities obtained by use of saturated inorganic salt solutions (14). At various times samples were removed and analyzed for extent of decomposition. Results are shown in Table I. A graph showing the general shape of the curve relating decomposition t o time is shown in Fig. 1.
EXPERIMENTAL Aspirin powder of U. S. P. grade (Mallinckrodt) was used for this investigation. I t was classified into various sizes by means of U. S. Standard sieves. Extent of aspirin decomposition was determined by measurement of the salicylic acid formed; this was accomplished by a modification of the A.O.A.C. procedure (13). The sample was dissolved in 100 ml. absolute alcohol, a suitable aliquot was transferred to a 100-ml. volumetric flask, and sufficient absolute alcohol was added t o make the total alcohol content 50 ml. Two milliliters of ferric ammonium sulfate T . S. was added and the solution was made up t o volume with water. The color intensity was measured on a spectrophotometer at 532 mp. Since it was not known whether aspirin would decompose if no moisture were present, a preliminary experiment was run in the absence of water vapor. One-half gram samples of aspirin of 100/140 screen size were sealed in ampuls and stored a t 35". loo", and 110". Similar samples to 45", 60", &lo, which 2.5% calcium stearate had been added were stored under identical conditions. Calcium stearate was added since it has been reported t o induce extensive decomposition of aspirin (4). Samples were removed at various times over a period of fifty days and assayed. Samples of aspirin alone showed little or n o decomposition at 80" or below. Those containing calcium stearate decomposed in a short time t o the extent of about 1% and then remained constant. In samples stored a t 100" and 110" salicylic acid was found to increase rapidly t o about 2% and then decrease gradually with time. These samples were observed to melt and change color, and it is not known whether the color interfered with the assay or the decrease in salicylic acid might be due to salicylide formation. It is believed that the small amount of decomposition detected in these samples could have been caused by traces of moisture which contaminated the dry aspirin during the process of filling the a m puls. The amount of water necessary to account for the decomposition observed is approximately moles. It was concluded, therefore, that decomposition of aspirin in the absence of moisture is of minor importance. Consequently, attention was directed to the decomposition of aspirin in the presence of water vapor. A series of samples of aspirin of screen size 100/140. weighing approximately 0.25 Gm., were placed in loosely capped, screw-top vials wliich were stored
::::I
30.0
-
,
,
,
0.0 0
80 160 TIME I N DAYS
240
Fig. 1.-Typical decomposition of aspirin in solid state (60", vapor pressure 120.3 mm.). In order t o be certain that the hydrolysis reaction was the only one of major importance and that no salicylic acid was lost by volatilization, an ultraviolet assay for aspirin and salicylic acid (5) was performed on numerous samples. This method consisted of reading solutions of aspirin and salicylic acid in absolute alcohol a t 226 mp and 235 mp, which are the absorbance peaks for aspirin and salicylic acid, respectively. The amount of each present was then determined by the usual two-component system technique. In all cases the original amount of aspirin could be accounted for by a summation of the aspirin and salicylic acid found.
TREATMENT OF THE DATA Examination of the data of Table I by means of typical plots used to determine reaction order made it obvious that the reaction studied did not follow one of the simpler kinetic relationships. The curves obtained, illustrated in Fig. 1, resembled typical autocatalytic plots, but an attempt t o establish autocatalysis by graphical means gave negative results. Therefore, a mechanism was postulated based on the known physical and chemical conditions of the reaction. The proposed mechanism is based on the following tenets: ( a ) Water is rapidly adsorbed onto the surface of the aspirin, the amount of water being a function of vapor pressure. If inultilayer aclsorp-
SCIENTIFIC EDITION
May 1958
331
TABLE I .-DECOMPOSITION OF ASPIRINAT VARIOUS TEMPERATURES AND VAPORPRESSURES -500-
46.02 mim . Time Mole in 4 Days Aspirin
54.40 mm. Time Mole 4 in Days Aspirin
0.0 3.0 24.0 34.1 40.9 47.9 61.9 76.8 103.8 139.8 170.8
99.90 99.87 99.83 99.66 99.53 99.29 99.26 98.73 98.51 97.77 96.86
0.0 3.0 24.0 34.0 40.9 47.9 61.7 76.8 103.8 139.8 170.8
...
...
...
...
...
68.41 mm. Time Mole in % Days Aspirin
0.0 10.8 20.0 27.7 33.8 44.7 48.7 51.8 61.8 83.6 99.1 120.0 141.0
99.90 99.42 99.00 98.27 97.91 96.84 96.35 95.94 94.80 91.38
... ... ..
...
...
...
88.50
85.26 80.74
... ... ...
...
68.30 mm. Time Mole in days Aszrin
0.0 3.0 7.0 21.9 30.0 34.1 40.9 47.9 61.9 76.8 103.8 139.8 170.8
99.90 99.86 99.70 99.43 99.27 98.83 98.98 98.26 97.83 96.91 94.69
74.50 mm. Time Mole in Days Aspirin
--
4
111.9 mm. Time Mole in % Days Aspirin
0.0 2.8 10.8 15.8 23.8 27.7 37.7 41.7 44.7 51.8 61.8 72.8 83.6 99.1 .. 107.7 120.0 141.0
99.90 99.70 99 134 98.71 98.54 97.73 97.00 96.64 95.66 95.08 93.20 91.81 89.97
. ..
..
0.0 2.9 6.8 10.8 15.8 20.7 23.8 27.7 30.8 33.8 37.8 41.7 44.7 48.7 51.8 55.7 61.8 72.8 83.6
...
...
...
~~
~
8.5.66 ._ ..
83.65 79.99 75.95
99.90 99.84 99.60 99.22 99.56 99.05 98.97 98.69 98.10 97.40 95.98 93.82 92.64
99.90 99.60 99.07 99.13 98.20 97.50 97.21 96.21 96.02 94.91 95.15 94.06 93.86 91.11 90.66 89.32 81.46 79.51 73.09
...
80'
181.0 mm. Time Mole in 9 HI. Aspirin
199.5 mm. Time Mole in % Hr. Aspirin
0.0 17.0 40.8 63.3 90.5 112.5 137.0
0.0 17.5 40.3 62.5 90.0 111.8 136.3
ifin ..,.. . i -i 186.5 208.8 232.8 308.8
99.90 99.68 99.69 97.96 95.79 94.45 91.77 88 .- 79 .85.28 83.34 81.96 70.86
ifin n
~._ .
186.0 208.3 232.5 308.3
99.90 99.39 98.16 97.41 95.24 92.39 87.80 85.69 _ _~. . 73.86 78.40 57.31 50.20
tion takes place a film of water will surround each aspirin particle, and the thickness of film will depend o n vapor pressure. If monolayer adsorption takes'place, each aspirin particle will be partially covered by a layer of water, the extent of coverage being dependent on vapor pressure. (b) The water
232.5 mm. Time Mole in % HI. Aspirin
0.0 17.3 40.5 63.3 90.3 112.8 136.8 160.8
1G.j
209.8 232.8
...
74.20 mm. Time Mole % in Days Aspirin
0.0 7.0 16.0 24.0 30.0 34.0 40.9 47.9 61.9 76.8 103.8 139.8 170.8
99.90 99.59 .. .. 99.32 99.65 98.92 98.76 98.72 98.24 96.72 96.16 93.29 89.78 86.75
120.3 mm. Time Mole % in Days Aspirin
0.0 2.9 6.8 10.8 15.8 20.7 23.8 27.7 30.8 33.8 37.8 41.7 44.7 48.7 51.8 55.7 61.8 66.7 72.8
99.90 99.75 99.24 99.22 98.55 95.14 95.24 94.89 91.72 90.80 90.26 86.12 87.42 82.93 83.32 79.04 70.65 64.75 60.45
132.0
35.98
--
99.90 99.25 98.64 97.05 94.40 86.75 84.78 77.44 72.80 67.54 60.76
film is rapidly saturated by solution of a portion of the solid aspirin. Decomposition then occurs in solution, and as a molecule of aspirin is removed via hydrolysis it is instantly replaced by one from the solid. Although oiily a small purtion of the aspirin is
332
JOURNAL OF THE
h m c m PHUACEUTICAL ASSOCIATION Vol. XLVII, No. 5
actually in solution, the reaction may be treated as though all the aspirin is in solution and only a small fraction is in active state. This makes it possible to express the concentration in terms of total amount of solid aspirin. Treatment of this system as a solution makes it possible t o use one of the rate constants derived by Edwards ( 1 2 ) . The constants of Edwards depending on water concentration are so small as t o be negligible, while the Concentration of hydroxyl ion in this system is so low that the extent of reaction dependent on it is negligible. Further, since the ionization constant for aspirin is small and ionization would be depressed by hydrogen ion concentration. it is believed that this system involves primarily molecular aspirin. Thus, only one of the constants described by Edwards, kl, which is associated with hydrogen ion concentration and molecular aspirin, is necessary for the rate expression for this system. Accordingly the rate equation may be written: dC/dt = k l ( A ) ( H + ) ,(Eq. I), in which C i s the concentration of salicylic acid, A is the concentration of aspirin, and Hf is the concentration of hydrogen ion. Using the definition of the ionization constant for acetic acid, K = [(H+)(CHaCOO-)] /(CH&OOH), (Eq. 2), the concentration of salicylic acid, C, may be substituted for that of acetic acid, since one mole of acetic acid is formed for each mole of salicylic acid, and the decrease in concentration of acetic acid due t o ionization is negligible. Also, the concentration of hydrogen ion is equal t o that of acetate ion. Thus, the ionization equation may be C) (Eq. 3 ) . stated: K = In order t o convert concentrations t o the number of moles contained in volume of solution, V , on the surface of the particles, K, which has units of moles/liter, is multiplied by V to give: K V = (Hf)2/(C),(Eq. 4). Solving for hydrogen ion concentration, [ H + ] = ( K V [ C ] ) ' / ' .(Eq. 5 ) ; and substituting Equation 5 in Equation 1 gives dC/dt = k l [ A ] ( K V [ C)'Iz, ] (Eq. 6), which applies to any single vapor pressure, and has units of moles X V - l X t-l. To express the rate in terms of number of moles rather than concentration the equation is multiplied by V, t o give dC/dt = klK'/zV'/2[A][C]'/Z, (Eq. 7). It should be noted that although the units of dC/dt are moles X t-' those of [ A ] and [C] are moles X V - l . Since aspirin samples have different weights they are converted t o a common basis by expressing both [ A ] and [C] as per cent of the total number of moles originally present, giving them the units mole% X V-1. It is possible to express aspirin concentration in terms of salicylic acid concentration, since their sum is always a constant equal t o 100. Thus, [A] IC] = [A01 [CO] = DO,(Eq. 8 ) , and [A] = DO - [ C ] ,(Eq. 9). where A0 and COare initial concentrations of aspirin and salicylic acid. This substitution facilitates integration t o give dC/dt = kiK'/aV'/*(Do - [C]) [C]'/', (Eq. 10). The volume of the solution layer, V,is unknown, but it may be expressed in terms of the Freundlich isotherm equation; thus, for a particular size and shape of particle: V = k'p", (Eq. l l ) , in which p is the vapor pressure, k' is the proportionality constant, and n is the order of the sorption reaction with respect t o p . Substituting this expression for V into Equation 10: dC/dt = kpann ( D O - [C])
+
+
[ C ] ' / ¶(Eq. , 12). where k = k1K1/"''/*. By making the substitution [C]'/' = y the following equation is obtained: Zdy/(Do - y2) = kpanl2 dt, (Eq. 13). which can be readily integrated by standard form, giving:
+
From Equation 9 it follows that (Do'/* [C]'/a) (DO'/' - [ C ] ' / a )= [ A ] ; inserting this into Equation 14:
Setting t equal to zero, the constant of integration,
I,is found t o be:
Rearranging and converting to logarithms to the base 10, the final equation is obtained:
If the mechanism proposed is correct, a plot of the left hand side of Equation 17 versus time should give a straight line with a slope equal to (DO'/'kp3n/2)/2.303. The data of Table I were plotted in this manner and found t o give straight lines. Figure 2 illustrates this type of graph. The values of the slopes of the lines obtained in this manner are shown in Table 11.
r
I
0
120 240 TIME IN HOURS
360
Fig. 2.-Typical plots of decomposition equation Legend: 0-vapor pressure 232.5 (Eq. 17), 80'. mm.; x-vapor pressure 199.5 mm.; .-vapor pressure 181.0 mm.
It can be seen from the equation of the slope that a plot of log (D,'/2kpan/2)/2.303versus log p should give a straight line with a slope of 3 n / 2 and an intercept of log DtIak/2.303. These graphs are shown in Fig. 3, and the values of the slopes and intercepts are presented in Table 111. From chemisorption theory ( 1 5 ) n would be expected to be 1, making 3 n / 2 equal t o 1.5. The results of Table I11 are in agreement with this within experimental error; thus, the intercepts of Table 111were calculated using a value of 1.5for 3n/2. In order to estimate the activation energy for the reaction, k, which is the product of kl, k"/2 and K'/*,
May 1958
SCIENTIFIC EDITION
TABLE II.-SLOPES
333
DECOMPOSITION EQUA- son t o believe that the sorption forces are other than ASPIRIN(EQ. 17) van der Waals, the energy of activation for this reaction can be considered zero (15). Thus the activaVapor tion energy determined from the slope of an ArrhenTemp. Pressure OC. mm. Slope = DO'/2kpn /*/2.303 ius plot is that of the hydrolysis reaction, and was 80 181.0 1.660 X lo-' hours-' calculated t o be 15,065 calories per mole, which is 199.5 1.986 X 10-3 hours-' in good agreement with the value of 15,620 deter232.5 2.475 X 10-3 hours-' mined by Edwards (17). 60 68.41 2.772 X 10-8 days-' 74.50 3.191 X 10-adays-' 111.9 5.218 X days-' 120.3 7.547 X 10-8 days-' DISCUSSION 50 46.02 7.558 X lo-' days-' 54.50 9.980 X lo-' days-' It should be noted that the method used in this 68.30 1.367 X 10-8 days-' work for control of humidity is quite troublesome. 74.20 1.860 X days-' The normal variations in temperature of laboratory ovens is sufficient to cause condensation of water on the inside of the desiccators which frequently causes loss of a whole series of samples. Also, data obtained once the per cent aspirin remaining has 9o1O0 Idropped t o a low value, usually around 2(r30%, 80 is widely scattered. At this point the samples are 7 0 60 1 50 commonly observed t o become sticky and aggregate 40 or even form a semisolid mass. It is believed that 30 a different physical system exists at this point. Data obtained undet such conditions are not reported in this paper and do not follow the mechanism proposed. However, decomposition at this stage is of minor importance. By use of the equations established in this paper, it is possible to predict the stability of aspirin under 5 known conditions of tmperature and humidity. 4 Application of the equation is limited, however, t o 3 the particle size used in this work. Further work is in progress in this laboratory t o determine the relationship between particle size and rate constant. It appears that by knowing the ionization constant 500 1000 10 20 3 0 50 100 200 of a material to be mixed with aspirin, one might LOG VAPOR PRESSURE use these equations t o predict the stability of the Fig. 3-Plots t o determine order of vapor pressure mixture. Other factors such as solubility of the effect. Legend: 0-60"; (3-50'; 0-80'. added substance will have t o be taken into account. Work of this type is planned. The basic assumptions and treatment of data used TABLE I11 in this study show promise of being applicable t o many solid medicinals which decompose as a result Temp. of adsorption of water. "C. 3n/2 Do''Zk/2.303 80 1.5 6.953 X 10-7mm.-'/:hrs.-' 60 1.5 4.976 X 10-6mm.-'I'sdays-1 50 1.7 2.552 X 10-6mm.-'/*days-1 REFERENCES OF THE
TION OF
TABLE IV. DECOMPOSITION RATE CONSTANTS FOR ASPIRINAT VARIOUSTEMPERATURES Temp.
OK
353.1 333.1 323.1
kik"'2
(hlole %) - 1
'2
Liters' '2 Moles-1
mm. -3 I2 Hr. --I
44.58 X 1012.41 X lod 6.06 X lo-'
(1) Tsakalotos, D. E. J . Pharm. Chim.. 14, 174(1916). (2) Paolini. V., Giorn.'chim. ind. cd. appli.. 3, 403(1921); Chem. Absfr., 16,314*(1922). (3) Stathopoulos T. G., Prak. Alkad. AfhZndn, 6, 229 (1931); Chcm. Absfr:, 27, 22506(1933). (4) Ribeiro, D., Stevenson, D., Samyn, J., Milosovich, G., and Mattocks A. M THISJOURNAL 4, 226(1955). (5) Ebert, W.'R., Ph: D. 1 hesis, University of Michigan (1956). ( 6 ) Yamamoto R. and Takahashi, T., Ann. Repfs. Shionoga Research iab.,'3,79(1954). 7) Ibid., 3, 112(1953). 8) Tsakalotos D E and Horsch S . Bull. Sac. Chcm.. 15. 743(1914); 17: i S 6 ( i b i 5 ) . 17, 4 0 i ( i g i 5 ) . (9) Morton, C., Quart. j. Pharm. and Pharmacol., 6,
__,.
4_ 4_ 3_ I\ 1_ 4_ 2 4 ~
is divided by K'A, the value of which is known for each of the temperatures studied (16). This leaves klk"/z the values of which are shown in Table IV. From a plot of log k1k"h versus 1/T the activation energy can be obtained. This value is the sum of two such energies, one associated with the chemical hydrolysis reaction, and the other with the physical sorption reaction. However, since there is no rea-
(10) Ferroni E. and Baistrocchi R. Spcrimenfdc, Sez. chem. b i d . 4 f ( l 9 k 3 ) ' Chcm. Absfr.,' 47,'10974h(1953). (11) St&&ant. . .J: M... J . Am. Chcm. Soc., 64. 723 ( 1950). (12) Edwards L. J. Trans. Faraday Soc. 4 6 723(1950). (13) "Methods of Analysis of the A. 0;A,' C.," 6th ed., A. 0. A. C. Washington D. C. 1945 p. 673. (14) Stoies, R. H., a d Robidson, R. A,, I n d . Eng. Chrm., 41, 2013(1953). (15) Laidler. K. J., "Chemical Kinetics," 1st ed.. McGrawHill, New York, 1950, p. 146. (16) "International Critical Tables," 1st ed., McGrawHili, New York, 1929. p. 263. (17) Edwards, L. J.. Trans. Faraday Soc.. 48, 696(1952).
May 1958
and reported total alkaloid content as ergotamine in three different samples. The values were reported as 110, 102, and 375 mg. %. No other determinations were made.
SUMMARY AND CONCLUSIONS
Ergot of couch grass was studied with respect to the fat, moisture, and alkaloid contents. It was found that half of the sample was infested with a beetle, and determinations were extended to find out what effect this infestation had on the contents of the drug. It was found that: 1. The moisture content of both infested and noninfested portions was the same. 2. The fat content of infested ergot was lower than that of the noninfested portion. 3. The alkaloid content of the infested ergot was lower than that of the noninfested portion. 4. Both portions assayed above N. F. limits in ergotoxine content, and below N. F. limits in ergonovine content.
329
5 . Evidence was presented to show that ergotamine, ergotaminine, ergosine, ergosinine, ergocornine, ergocorninine, ergocristine, ergocristinine, ergocryptine, ergocryptinine, and lysergic acid were probably present. 6. Ergonovine was positively shown to be present. REFERENCES (1) Youngken, Heber W., “Textbook of Fharmdcognosy,” 6th Ed., Blakiston C o . , New York., 1948, pp. 85-91. (2) Ibid. p. 132-137. (3) “Th; National Formulary,” 10th ed., Mack Publishing C o Easton Pa. 1955, p. 217. (4) ’Silber, A,, aAd Schulze, T . , Pharmasie, 8, 675(1953). (5) “Pharmacopeia of the United States,” 15th rev., Mack Publishing. Co., Easton, Pa.. 1955, p. 1094. ( G ) , Personal contact with C. Pugh, Research Laboratories, Eli Lilly and Co., Indianapolis, Ind. (7) Kolsek, J., Microchim. Acta, 9, 1377(195G). (8) Tyler, V. E., Jr., and Schwarting, A. E., THIS JOURNAL, 41, 345(1952). (9) “The National Formulary,” 10th ed., Mack Publishing Co., Easton, Pa., 1955, p. 218. (10) Ibid p 216. (11) Madski. R. H. F., and Holme?. H. L.. “The Alkaloids, Chemistry and Physiology,” Academic Press, Inc., New York. N. Y. 1952 p. 377. (12j Silbhr, A,, and Dischoff, W.. Arch. Phorm.. 288. 124 (1955).
Decomposition of Aspirin in the Solid State* By LEWIS J. LEESON? and ALBERT M. MATTOCKS The decomposition of aspirin in the solid state has been demonstrated to be dependent on vapor pressure and temperature. A possible mechanism of decomposition is offered, which consists of an initial sorption of a water layer by each particle, diffusion of aspirin into solution, and decomposition taking place by acid catalyzed hydrolysis. A series of equations which describe the mechanism have been derived and the experimental data demonstrated to fit them. Using these equations, the various kinetic constants have been determined.
comprise a large S segment of the drugs marketed today, there is a great need for methods of predicting their INCE SOLID DOSAGE FORMS
stability under various conditions of storage. In order for such a method to be dependable it must be based upon a knowledge of the mechanisms involved and quantitative studies of the variables affecting the reactions. Many pharmaceuticals are known to exhibit decompositionin solid form, and perhaps the most widely used one is aspirin. Aspirin is an especially interesting example of this type of reaction since it is known to be affected not only by temperature and humidity but also by numerous chemical agents with which it is often combined. Decomposition of aspirin in solid form has been noted by Tsakalotos (l), Paolini (2), and
* Received May 3. 1957, from the College of Pharmacy, University of Michigan. Ann Arbor. Presented to the Scientific Section, A. PH. A. New York meeting, April-May, 1957. t Parke. Davis Fellow and American Foundation for Pharmaceutical Education Fellow.
Strathopoulos (3). In addition, Ribeiro, et al. (4), found that certain lubricants, notably stearates, increased the degree of decomposition of aspirin contained in aspirin-phenacetin-caffeine tablets. Ebert (5), and Yamamoto and Takahashi (6, 7) studied the accelerating effects of certain mines, temperature, humidity, pressure, and grinding upon aspirin loss in tablets. These studies were concerned primarily with effects of various chemical agents and conditions and did not attempt to determine reaction rates or mechanisms. There have been a number of kinetic studies on aspirin decomposition in solution (8-11), the most complete investigation being reported by Edwards (12). He agreed with other workers that the decomposition is first order with respect to aspirin concentration, but by studying the reaction at various pH values he demonstrated that the apparent first order rate constant is actually a composite of six rate constants. Each
330
JOURNAL OF THE AMERICANPHARMACEUTICAL ASSOCIATION Vol. XLVII, No. 5
of t h e six constants is associated with one of the hydrolytic reactions depending upon hydrogen ion, hydroxyl ion, or water reacting with either molecular or ionic aspirin. There is no direct method for applying t o solid forms information obtained from studies of decomposition of a drug i n solution, since i t is even questionable whether the same reactions occur. T h u s i t is important t h a t studies be made of reactions of drugs in solid state, and such a study was undertaken in these laboratories using aspirin as the first example.
a t 50°, 60", and 80";and a t varying humidities obtained by use of saturated inorganic salt solutions (14). At various times samples were removed and analyzed for extent of decomposition. Results are shown in Table I. A graph showing the general shape of the curve relating decomposition t o time is shown in Fig. 1.
EXPERIMENTAL Aspirin powder of U. S. P. grade (Mallinckrodt) was used for this investigation. I t was classified into various sizes by means of U. S. Standard sieves. Extent of aspirin decomposition was determined by measurement of the salicylic acid formed; this was accomplished by a modification of the A.O.A.C. procedure (13). The sample was dissolved in 100 ml. absolute alcohol, a suitable aliquot was transferred to a 100-ml. volumetric flask, and sufficient absolute alcohol was added t o make the total alcohol content 50 ml. Two milliliters of ferric ammonium sulfate T . S. was added and the solution was made up t o volume with water. The color intensity was measured on a spectrophotometer at 532 mp. Since it was not known whether aspirin would decompose if no moisture were present, a preliminary experiment was run in the absence of water vapor. One-half gram samples of aspirin of 100/140 screen size were sealed in ampuls and stored a t 35". loo", and 110". Similar samples to 45", 60", &lo, which 2.5% calcium stearate had been added were stored under identical conditions. Calcium stearate was added since it has been reported t o induce extensive decomposition of aspirin (4). Samples were removed at various times over a period of fifty days and assayed. Samples of aspirin alone showed little or n o decomposition at 80" or below. Those containing calcium stearate decomposed in a short time t o the extent of about 1% and then remained constant. In samples stored a t 100" and 110" salicylic acid was found to increase rapidly t o about 2% and then decrease gradually with time. These samples were observed to melt and change color, and it is not known whether the color interfered with the assay or the decrease in salicylic acid might be due to salicylide formation. It is believed that the small amount of decomposition detected in these samples could have been caused by traces of moisture which contaminated the dry aspirin during the process of filling the a m puls. The amount of water necessary to account for the decomposition observed is approximately moles. It was concluded, therefore, that decomposition of aspirin in the absence of moisture is of minor importance. Consequently, attention was directed to the decomposition of aspirin in the presence of water vapor. A series of samples of aspirin of screen size 100/140. weighing approximately 0.25 Gm., were placed in loosely capped, screw-top vials wliich were stored
::::I
30.0
-
,
,
,
0.0 0
80 160 TIME I N DAYS
240
Fig. 1.-Typical decomposition of aspirin in solid state (60", vapor pressure 120.3 mm.). In order t o be certain that the hydrolysis reaction was the only one of major importance and that no salicylic acid was lost by volatilization, an ultraviolet assay for aspirin and salicylic acid (5) was performed on numerous samples. This method consisted of reading solutions of aspirin and salicylic acid in absolute alcohol a t 226 mp and 235 mp, which are the absorbance peaks for aspirin and salicylic acid, respectively. The amount of each present was then determined by the usual two-component system technique. In all cases the original amount of aspirin could be accounted for by a summation of the aspirin and salicylic acid found.
TREATMENT OF THE DATA Examination of the data of Table I by means of typical plots used to determine reaction order made it obvious that the reaction studied did not follow one of the simpler kinetic relationships. The curves obtained, illustrated in Fig. 1, resembled typical autocatalytic plots, but an attempt t o establish autocatalysis by graphical means gave negative results. Therefore, a mechanism was postulated based on the known physical and chemical conditions of the reaction. The proposed mechanism is based on the following tenets: ( a ) Water is rapidly adsorbed onto the surface of the aspirin, the amount of water being a function of vapor pressure. If inultilayer aclsorp-
SCIENTIFIC EDITION
May 1958
331
TABLE I .-DECOMPOSITION OF ASPIRINAT VARIOUS TEMPERATURES AND VAPORPRESSURES -500-
46.02 mim . Time Mole in 4 Days Aspirin
54.40 mm. Time Mole 4 in Days Aspirin
0.0 3.0 24.0 34.1 40.9 47.9 61.9 76.8 103.8 139.8 170.8
99.90 99.87 99.83 99.66 99.53 99.29 99.26 98.73 98.51 97.77 96.86
0.0 3.0 24.0 34.0 40.9 47.9 61.7 76.8 103.8 139.8 170.8
...
...
...
...
...
68.41 mm. Time Mole in % Days Aspirin
0.0 10.8 20.0 27.7 33.8 44.7 48.7 51.8 61.8 83.6 99.1 120.0 141.0
99.90 99.42 99.00 98.27 97.91 96.84 96.35 95.94 94.80 91.38
... ... ..
...
...
...
88.50
85.26 80.74
... ... ...
...
68.30 mm. Time Mole in days Aszrin
0.0 3.0 7.0 21.9 30.0 34.1 40.9 47.9 61.9 76.8 103.8 139.8 170.8
99.90 99.86 99.70 99.43 99.27 98.83 98.98 98.26 97.83 96.91 94.69
74.50 mm. Time Mole in Days Aspirin
--
4
111.9 mm. Time Mole in % Days Aspirin
0.0 2.8 10.8 15.8 23.8 27.7 37.7 41.7 44.7 51.8 61.8 72.8 83.6 99.1 .. 107.7 120.0 141.0
99.90 99.70 99 134 98.71 98.54 97.73 97.00 96.64 95.66 95.08 93.20 91.81 89.97
. ..
..
0.0 2.9 6.8 10.8 15.8 20.7 23.8 27.7 30.8 33.8 37.8 41.7 44.7 48.7 51.8 55.7 61.8 72.8 83.6
...
...
...
~~
~
8.5.66 ._ ..
83.65 79.99 75.95
99.90 99.84 99.60 99.22 99.56 99.05 98.97 98.69 98.10 97.40 95.98 93.82 92.64
99.90 99.60 99.07 99.13 98.20 97.50 97.21 96.21 96.02 94.91 95.15 94.06 93.86 91.11 90.66 89.32 81.46 79.51 73.09
...
80'
181.0 mm. Time Mole in 9 HI. Aspirin
199.5 mm. Time Mole in % Hr. Aspirin
0.0 17.0 40.8 63.3 90.5 112.5 137.0
0.0 17.5 40.3 62.5 90.0 111.8 136.3
ifin ..,.. . i -i 186.5 208.8 232.8 308.8
99.90 99.68 99.69 97.96 95.79 94.45 91.77 88 .- 79 .85.28 83.34 81.96 70.86
ifin n
~._ .
186.0 208.3 232.5 308.3
99.90 99.39 98.16 97.41 95.24 92.39 87.80 85.69 _ _~. . 73.86 78.40 57.31 50.20
tion takes place a film of water will surround each aspirin particle, and the thickness of film will depend o n vapor pressure. If monolayer adsorption takes'place, each aspirin particle will be partially covered by a layer of water, the extent of coverage being dependent on vapor pressure. (b) The water
232.5 mm. Time Mole in % HI. Aspirin
0.0 17.3 40.5 63.3 90.3 112.8 136.8 160.8
1G.j
209.8 232.8
...
74.20 mm. Time Mole % in Days Aspirin
0.0 7.0 16.0 24.0 30.0 34.0 40.9 47.9 61.9 76.8 103.8 139.8 170.8
99.90 99.59 .. .. 99.32 99.65 98.92 98.76 98.72 98.24 96.72 96.16 93.29 89.78 86.75
120.3 mm. Time Mole % in Days Aspirin
0.0 2.9 6.8 10.8 15.8 20.7 23.8 27.7 30.8 33.8 37.8 41.7 44.7 48.7 51.8 55.7 61.8 66.7 72.8
99.90 99.75 99.24 99.22 98.55 95.14 95.24 94.89 91.72 90.80 90.26 86.12 87.42 82.93 83.32 79.04 70.65 64.75 60.45
132.0
35.98
--
99.90 99.25 98.64 97.05 94.40 86.75 84.78 77.44 72.80 67.54 60.76
film is rapidly saturated by solution of a portion of the solid aspirin. Decomposition then occurs in solution, and as a molecule of aspirin is removed via hydrolysis it is instantly replaced by one from the solid. Although oiily a small purtion of the aspirin is
332
JOURNAL OF THE
h m c m PHUACEUTICAL ASSOCIATION Vol. XLVII, No. 5
actually in solution, the reaction may be treated as though all the aspirin is in solution and only a small fraction is in active state. This makes it possible to express the concentration in terms of total amount of solid aspirin. Treatment of this system as a solution makes it possible t o use one of the rate constants derived by Edwards ( 1 2 ) . The constants of Edwards depending on water concentration are so small as t o be negligible, while the Concentration of hydroxyl ion in this system is so low that the extent of reaction dependent on it is negligible. Further, since the ionization constant for aspirin is small and ionization would be depressed by hydrogen ion concentration. it is believed that this system involves primarily molecular aspirin. Thus, only one of the constants described by Edwards, kl, which is associated with hydrogen ion concentration and molecular aspirin, is necessary for the rate expression for this system. Accordingly the rate equation may be written: dC/dt = k l ( A ) ( H + ) ,(Eq. I), in which C i s the concentration of salicylic acid, A is the concentration of aspirin, and Hf is the concentration of hydrogen ion. Using the definition of the ionization constant for acetic acid, K = [(H+)(CHaCOO-)] /(CH&OOH), (Eq. 2), the concentration of salicylic acid, C, may be substituted for that of acetic acid, since one mole of acetic acid is formed for each mole of salicylic acid, and the decrease in concentration of acetic acid due t o ionization is negligible. Also, the concentration of hydrogen ion is equal t o that of acetate ion. Thus, the ionization equation may be C) (Eq. 3 ) . stated: K = In order t o convert concentrations t o the number of moles contained in volume of solution, V , on the surface of the particles, K, which has units of moles/liter, is multiplied by V to give: K V = (Hf)2/(C),(Eq. 4). Solving for hydrogen ion concentration, [ H + ] = ( K V [ C ] ) ' / ' .(Eq. 5 ) ; and substituting Equation 5 in Equation 1 gives dC/dt = k l [ A ] ( K V [ C)'Iz, ] (Eq. 6), which applies to any single vapor pressure, and has units of moles X V - l X t-l. To express the rate in terms of number of moles rather than concentration the equation is multiplied by V, t o give dC/dt = klK'/zV'/2[A][C]'/Z, (Eq. 7). It should be noted that although the units of dC/dt are moles X t-' those of [ A ] and [C] are moles X V - l . Since aspirin samples have different weights they are converted t o a common basis by expressing both [ A ] and [C] as per cent of the total number of moles originally present, giving them the units mole% X V-1. It is possible to express aspirin concentration in terms of salicylic acid concentration, since their sum is always a constant equal t o 100. Thus, [A] IC] = [A01 [CO] = DO,(Eq. 8 ) , and [A] = DO - [ C ] ,(Eq. 9). where A0 and COare initial concentrations of aspirin and salicylic acid. This substitution facilitates integration t o give dC/dt = kiK'/aV'/*(Do - [C]) [C]'/', (Eq. 10). The volume of the solution layer, V,is unknown, but it may be expressed in terms of the Freundlich isotherm equation; thus, for a particular size and shape of particle: V = k'p", (Eq. l l ) , in which p is the vapor pressure, k' is the proportionality constant, and n is the order of the sorption reaction with respect t o p . Substituting this expression for V into Equation 10: dC/dt = kpann ( D O - [C])
+
+
[ C ] ' / ¶(Eq. , 12). where k = k1K1/"''/*. By making the substitution [C]'/' = y the following equation is obtained: Zdy/(Do - y2) = kpanl2 dt, (Eq. 13). which can be readily integrated by standard form, giving:
+
From Equation 9 it follows that (Do'/* [C]'/a) (DO'/' - [ C ] ' / a )= [ A ] ; inserting this into Equation 14:
Setting t equal to zero, the constant of integration,
I,is found t o be:
Rearranging and converting to logarithms to the base 10, the final equation is obtained:
If the mechanism proposed is correct, a plot of the left hand side of Equation 17 versus time should give a straight line with a slope equal to (DO'/'kp3n/2)/2.303. The data of Table I were plotted in this manner and found t o give straight lines. Figure 2 illustrates this type of graph. The values of the slopes of the lines obtained in this manner are shown in Table 11.
r
I
0
120 240 TIME IN HOURS
360
Fig. 2.-Typical plots of decomposition equation Legend: 0-vapor pressure 232.5 (Eq. 17), 80'. mm.; x-vapor pressure 199.5 mm.; .-vapor pressure 181.0 mm.
It can be seen from the equation of the slope that a plot of log (D,'/2kpan/2)/2.303versus log p should give a straight line with a slope of 3 n / 2 and an intercept of log DtIak/2.303. These graphs are shown in Fig. 3, and the values of the slopes and intercepts are presented in Table 111. From chemisorption theory ( 1 5 ) n would be expected to be 1, making 3 n / 2 equal t o 1.5. The results of Table I11 are in agreement with this within experimental error; thus, the intercepts of Table 111were calculated using a value of 1.5for 3n/2. In order to estimate the activation energy for the reaction, k, which is the product of kl, k"/2 and K'/*,
May 1958
SCIENTIFIC EDITION
TABLE II.-SLOPES
333
DECOMPOSITION EQUA- son t o believe that the sorption forces are other than ASPIRIN(EQ. 17) van der Waals, the energy of activation for this reaction can be considered zero (15). Thus the activaVapor tion energy determined from the slope of an ArrhenTemp. Pressure OC. mm. Slope = DO'/2kpn /*/2.303 ius plot is that of the hydrolysis reaction, and was 80 181.0 1.660 X lo-' hours-' calculated t o be 15,065 calories per mole, which is 199.5 1.986 X 10-3 hours-' in good agreement with the value of 15,620 deter232.5 2.475 X 10-3 hours-' mined by Edwards (17). 60 68.41 2.772 X 10-8 days-' 74.50 3.191 X 10-adays-' 111.9 5.218 X days-' 120.3 7.547 X 10-8 days-' DISCUSSION 50 46.02 7.558 X lo-' days-' 54.50 9.980 X lo-' days-' It should be noted that the method used in this 68.30 1.367 X 10-8 days-' work for control of humidity is quite troublesome. 74.20 1.860 X days-' The normal variations in temperature of laboratory ovens is sufficient to cause condensation of water on the inside of the desiccators which frequently causes loss of a whole series of samples. Also, data obtained once the per cent aspirin remaining has 9o1O0 Idropped t o a low value, usually around 2(r30%, 80 is widely scattered. At this point the samples are 7 0 60 1 50 commonly observed t o become sticky and aggregate 40 or even form a semisolid mass. It is believed that 30 a different physical system exists at this point. Data obtained undet such conditions are not reported in this paper and do not follow the mechanism proposed. However, decomposition at this stage is of minor importance. By use of the equations established in this paper, it is possible to predict the stability of aspirin under 5 known conditions of tmperature and humidity. 4 Application of the equation is limited, however, t o 3 the particle size used in this work. Further work is in progress in this laboratory t o determine the relationship between particle size and rate constant. It appears that by knowing the ionization constant 500 1000 10 20 3 0 50 100 200 of a material to be mixed with aspirin, one might LOG VAPOR PRESSURE use these equations t o predict the stability of the Fig. 3-Plots t o determine order of vapor pressure mixture. Other factors such as solubility of the effect. Legend: 0-60"; (3-50'; 0-80'. added substance will have t o be taken into account. Work of this type is planned. The basic assumptions and treatment of data used TABLE I11 in this study show promise of being applicable t o many solid medicinals which decompose as a result Temp. of adsorption of water. "C. 3n/2 Do''Zk/2.303 80 1.5 6.953 X 10-7mm.-'/:hrs.-' 60 1.5 4.976 X 10-6mm.-'I'sdays-1 50 1.7 2.552 X 10-6mm.-'/*days-1 REFERENCES OF THE
TION OF
TABLE IV. DECOMPOSITION RATE CONSTANTS FOR ASPIRINAT VARIOUSTEMPERATURES Temp.
OK
353.1 333.1 323.1
kik"'2
(hlole %) - 1
'2
Liters' '2 Moles-1
mm. -3 I2 Hr. --I
44.58 X 1012.41 X lod 6.06 X lo-'
(1) Tsakalotos, D. E. J . Pharm. Chim.. 14, 174(1916). (2) Paolini. V., Giorn.'chim. ind. cd. appli.. 3, 403(1921); Chem. Absfr., 16,314*(1922). (3) Stathopoulos T. G., Prak. Alkad. AfhZndn, 6, 229 (1931); Chcm. Absfr:, 27, 22506(1933). (4) Ribeiro, D., Stevenson, D., Samyn, J., Milosovich, G., and Mattocks A. M THISJOURNAL 4, 226(1955). (5) Ebert, W.'R., Ph: D. 1 hesis, University of Michigan (1956). ( 6 ) Yamamoto R. and Takahashi, T., Ann. Repfs. Shionoga Research iab.,'3,79(1954). 7) Ibid., 3, 112(1953). 8) Tsakalotos D E and Horsch S . Bull. Sac. Chcm.. 15. 743(1914); 17: i S 6 ( i b i 5 ) . 17, 4 0 i ( i g i 5 ) . (9) Morton, C., Quart. j. Pharm. and Pharmacol., 6,
__,.
4_ 4_ 3_ I\ 1_ 4_ 2 4 ~
is divided by K'A, the value of which is known for each of the temperatures studied (16). This leaves klk"/z the values of which are shown in Table IV. From a plot of log k1k"h versus 1/T the activation energy can be obtained. This value is the sum of two such energies, one associated with the chemical hydrolysis reaction, and the other with the physical sorption reaction. However, since there is no rea-
(10) Ferroni E. and Baistrocchi R. Spcrimenfdc, Sez. chem. b i d . 4 f ( l 9 k 3 ) ' Chcm. Absfr.,' 47,'10974h(1953). (11) St&&ant. . .J: M... J . Am. Chcm. Soc., 64. 723 ( 1950). (12) Edwards L. J. Trans. Faraday Soc. 4 6 723(1950). (13) "Methods of Analysis of the A. 0;A,' C.," 6th ed., A. 0. A. C. Washington D. C. 1945 p. 673. (14) Stoies, R. H., a d Robidson, R. A,, I n d . Eng. Chrm., 41, 2013(1953). (15) Laidler. K. J., "Chemical Kinetics," 1st ed.. McGrawHill, New York, 1950, p. 146. (16) "International Critical Tables," 1st ed., McGrawHili, New York, 1929. p. 263. (17) Edwards, L. J.. Trans. Faraday Soc.. 48, 696(1952).