Anal. Chem. 1983, 5 5 , 54-56
54
fluorescence file searching routines. Conversely, this absence has diminished the need to develop standard protocols for fluorescence instrumentation. This interface between the requirements of the collection and identification of fluorescence spectra promises to be a challenging area for future investigation. The FORTRAN code for SIPS compiles and runs under the TOPS-10 operating system on the University of Delaware Computation Center DEC-10 system. Source listings are available from the author. Portability is restricted by the IMSL subroutine calls and by use of DELPLOT for hard-copy graphics output.
ACKNOWLEDGMENT The assistance of Cecil Dybowski and the insights he provided regarding the broader aspects of Fourier transforms is gratefully acknowledged. The cooperation and help of Isadore Berlman in making available a computer-readable library file of spectra aided this work immeasurably. Registry No. Anthracene, 120-12-7;3,4-benzophenanthrene, 195-19-7;naphthalene, 91-20-3;deuterated anthracene, 54261-80-2; benzidine, 92-87-5; sodium salicylate, 54-21-7; tetramethyl-p&phenyl, 83399-67-1;tetramethylphenylenediamine,27215-51-6; pyrene, 129-00-0;perylene, 198-55-0; 2-aminoanthracene,613-13-8; rubrene, 517-51-1; chrysene, 218-01-9; 2-phenylphenanthrene, 4325-77-3; 4,4-di(n-butoxy)-l,l-binaphthyl, 4499-67-6; 1,4-diphenylnaphthalene, 796-30-5. LITERATURE CITED (1) Hlrata Y.; Novotny, M.; Peaden, P. A.; Lee, M. L. Anal. Chlm. Acta 1961, 127, 55-61.
(2) Shpol'skll, E. V. Sov. Phys.-Usp.(fngl. Trans/.) 1962, 5 , 522-531. (3) Green, L. G.; O'Haver, T. C. Anal. Chem. 1974, 46, 2191-2196. (4) Warner, I.M.; Christian, G. D.; Davldson, E. R.; Callis, J. 6. Anal. Chem. 1977, 49. 564-573. (5) Miller, T. C.; Faulkner, L. R. Anal. Chem. 1976, 4 8 , 2083-2088. (6) Ylm, K. W. K.; Miller, T. C.; Faulkner, L. R. Anal. Chem. 1977, 49, 2069-2074. (7) Gold, H. S.; Rechsteiner, C. E.; Buck, R. P. Anal. Chim. Acta 1977, 9 5 , 51-58. (8) Stadallus, M. A,; Gold, H. S. Federation of Analyticai Chemistry and Spectroscopy Societies Meeting, Philadelphia, PA, Sept 1981; paper 62. (9) Small, G. W.; Rasmussen, G. T.; Isenhour, T. L. Appl. Specfrosc. 1979, 33, 444-450. (10) Powell, L. A,; Hleftje, G. M. Anal. Chim. Acta 1978, 100, 313-327. (11) Cooley, J. W.; Tukey, J. W. Math. Comput. 1985, 19, 297-309. (12) Horlick, G.; Hieftje, G. M. I n "Contemporary Topics in Analytical and Clinical Chemistry"; Hercules, D. M., Hieftje. G. M., Snyder, L. R., Evenson, M. A,, Eds.; Plenum: New York, 1978; Vol. 111. (13) Horllck, G. Anal. Chem. 1973, 45, 319-324. (14) Lam, R. 8.; Wleboldt, R. C.; Isenhour, T. L. Anal. Chem. 1981, 5 3 , 889A-90 1A. (15) Marshall, A. G. I n "Fourier, Hadamard, and Hilbert Transforms in Chemistry"; Marshall, A. G., Ed.; Plenum: New York, 1982. (16) Berne, B. J.; Pecora, R. "Dynamic Light Scattering"; Wlley: New York, 1976. (17) Berlman, I.B. "Handbook of Fluorescence Spectra of Aromatic Molecules"; Academic Press: New York, 1971. (18) Gold, H. S.; Rechsteiner, C. E.; Buck, R. P. Anal. Chem. 1978, 4 8 , 1540-1 546. (19) Bendat, J. S.; Plersol, A. G. "Random Data: Analysis and Measurement Procedures"; Wiley-Intersclence: New York, 1971. (20) Demas, J. N.; Crosby, G. A. J . Phys. Chem. 1971, 75,991-1024. (21) Yim, K. W. K. Ph.D. Dissertation, University of Illinois at UrbanaChampaign, Urbana, IL, 1978.
RECEIVED for review May 17,1982. Accepted September 24, 1982. Partial support of this research was provided by the American Cancer Society under Grant IN-159.
Determination of Salicylic Acid in Aspirin Powder by Second Derivative Ultraviolet Spectrometry Kelsuke Kltamura" and Ryo MaJima Kyoto College of Pharmacy, 5 Nakauchl-cho, Misasagi, Yamashina-ku, Kyoto 607, Japan
The second derivative spectrum of sallcylic acid showed a trough at 292 nm and a satellite peak at 316 nm. When large amounts of aspirin coexlsted, the trough disappeared but the height of the satelllte peak ( D , ) was not altered even at an aspirin concentration 20 000 times that of salicylic acld (correspondlng to salicylic acld content of 0.005%). A plot of 25 sets of D , values and salicyllc acld concentrations (1.00-10.02 pg/mL) gave a stralght line (correlatlon coefficient = 0.9999) and relatlve standard deviation ( S I X ) for a slope of 1.2%. A typical assay result for commercial aspirin powder was that the content of salicylic acld was 0.0361 f 0.0005% (at the 95% confldence IimR) wlth s/jz of 1.2% for five measurements.
I t has been shown that the application of derivative techniques to spectrophotometry is very useful when there exists signal overlapping or interferences (1-3). Moreover, the experimental procedure is simple and time-saving. Despite these advantages, however, few applications of derivative spectrophotometry have been published (4-7). This paper describes an application of second derivative UV spectrometry to permit a simple and rapid assay of sal-
icylic acid in aspirin powder. As salicylic acid, the major decomposition product of aspirin, irritates the digestive system, the limit of salicylic acid content in aspirin powder is prescribed to be 0.1% by the Pharmacopeas (8,9). But the assay for aspirin powder described in them is qualitative. There have been some reports on the assay for salicylic acid of 0.1% content level in aspirin by gas-liquid chromatography (GLC) (IO),liquid chromatography ( I I ) , and high-performance liquid chromatography (12). Although the retention times of salicylic acid were within 10 min in every one of these chromatographic methods (10-12), time-consuming chromatographic conditioning was essential for all of them and chemical derivatization was necessary for the GLC method (10).
EXPERIMENTAL SECTION Solvent and Chemicals. A 1% chloroacetic acid-ethanol solution was used as solvent. Salicylic acid was twice recrystallized from hexane, and standard salicylic acid solutions of various concentrations were prepared by dilution of a stock solution with the solvent. Salicylic acid free aspirin was obtained by twice recrystallization of aspirin from acetone. Preparation of Assay Solutions. Assay solutions were prepared immediately before measurement. A 150-mg sample of aspirin powder was dissolved in the solvent to bring the mixture t o 25.0 mL.
0003-2700/83/0355-0054$01.50/00 1982 American Chemical Society
ANALYTICAL CHEMISTRY, VOL. 55, NO. 1, JANUARY 1983
I
I
400
500 WAVELENGTHh)
Figure 1. Second derivatlve spectrum of
i3
didymium fllter.
( 1 : 2050)
0.2
0.2
IC ACID
ASPIRIN
300 WAVELENGTH (nrn)
3-ORDER
350
300
3 60
WAVELENGTH (nrn)
Figure 2. Absorption (zero order) and second derivative spectra of salicylic acid (3.0 pg/mL) and aspirin (12.1 pg/mL) (a) and the spectra of their mixture in the c:oncentration ratio of 1:2000 (b).
Second Derivative UV Spectrometry. A differentiator was constructed with electronic derivative (circuits (1, 13). Three circuits having varied differential time constants (2.7, 8.2, and 22.0 s) were incorporated and any one of them could be selected by switching. The differentiator was connected to a double-beam scanning spectrophotometer, Shimazu UV-210A. The derivative spectra were obtained at a slit width of 1 nm, a scanning speed of 120 nm/min, and a time constant of 22.0 s.
RESULTS AND DISClUSSION Second Derivative Spectrum of a Didymium Filter. Figure 1shows a second derivative spectrum of a didymium fiiter measured under the same derivative conditions employed for the assay of salicylic acid (see Experimental Section). The positions of troughs iin the second derivative spectrum were shifted along the scanning direction (toward short wavelength) by about 10 nm relative to corresponding peaks in nonderivative absorption spectrum. The spectrum may serve as a certain index when tlhe derivative conditions are referred to other derivative devices or methods. Absorption (Zero Order) and Second Derivative Spectra of Salicylic Acid, Aspirin, and Their Mixture (1:2000). The absorption spectra of salicylic acid (3.0 pg/mL) and aspirin (12.1 pg/rnL) are depicted ah the bottom of Figure 2a. Since the concentration of aspirin was not much higher than that of salicylic acid, the zero-order spectrum of aspirin showed no absorptioin a t the A,, of salicylic acid (302 nm). Thus the concentration of salicylic acid in a mixture with a comparable amount of aspirin could be determined from the absorbance a t 302 nm without any interference of aspirin signal. However, since the content of salicylic acid in aspirin powder is prescribed not to exceed 0.1 70 (8,9), the concentration of aspirin would be considerably higher than loo0 times that of salicylic acid in practical cases. The zero-order spectrum of a mixture of salicylic acid and a great quantity of aspirin (about 2000 times of salicylic acid) is illustrated in the lower part of Figure 2b. The salicylic acid signal was severely interfered with by a large aspirin band, so that the determination of salicylic acid by conventional absorption spectrometry would not be feasible. The upper spectra in Figure 2a illustrated the second derivative spectra of salicylic acid and aspirin, respectively. The
55
spectrum of salicylic acid showed a trough at 292 nm and a satellite peak at 316 nm, and the spectrum of aspirin showed a trough a t 264 nm and a satellite peak a t 284 nm. I t was apparent from comparing the zero order and second derivative spectra in Figure 2a that the detection power for both salicylic acid and aspirin was enhanced by differentiation. The limit of detection for salicylic acid in the second derivative spectrum was about 0.12 pg/mL, since the height of trough-to-peak was almost equilvalent to twice of peak-to-peak noise a t this concentration. The second derivative spectrum for the mixture of salicylic acid and aspirin (1:2000) is shown in the upper part of Figure 2b. On account of the large satellite peak of aspirin, the trough of salicylic acid at 292 nm disappeared but the satellite peak of salicylic acid a t 316 nm seemed to be reproduced. If the height of the satellite peak measured with respect to the derivative zero, denoted by DL in Figure 2, would not be affected by coexisting aspirin, it could be used for the quantitative determination of salicylic acid in aspirin powder. Effect of Coexisting Aspirin on DL Values. The influence of aspirin signal on DL values could be known by comparing the DL values of a solution of salicylic acid alone with that of a solution having the same salicylic acid concentration but containing large amounts of aspirin. For this purpose salicylic acid free aspirin was needed, since if aspirin added to a salicylic acid solution contained salicylic acid originally, the amount of salicylic acid in the solution would be increased and the results could not be reliable. Salicylic acid in aspirin was removed by recrystallization. When the second derivative spectrum of highly concentrated aspirin solution (6 mg/mL) did not show any signal a t 316 nm (DL = 0), the aspirin was considered salicylic acid free. Among the several recrystallization solvents (benzene, chloroform, dioxane, isopropyl alcohol), only acetone could give salicylic acid free aspirin. One more problem which would make results uncertain was the hydrolysis of aspirin to salicylic acid in the sample solution during the experiment. When a second derivative spectrum of salicylic acid free aspirin in ethanol (6 mg/mL) was measured repeatedly at intervals of a few minutes, the recorded line at 316 nm gradually raised from the base line due to the formation of salicylic acid by the hydrolysis of aspirin. In order to suppress the hydrolysis of aspirin, acid-containing ethanol could be used as solvent (14). Three kinds of acid, phosphoric acid (85%),acetic acid, and chloroacetic acid were tested. A 60-mg sample of salicylic acid free aspirin was dissolved in 10 mL of ethanol which contained one of these acids with 1% concentration and the second derivative spectrum was taken repeatedly for 1 h with some time intervals. The plot of the obtained DL values against the time (0-60 min) yielded a straight line for each acid. The comparison of the suppression effect of each acid was carried out by using slopes of the lines. The results summarized in Table I showed that the hydrolysis of aspirin was most effectively suppressed in 1% chloroacetic acid-ethanol solvent, since the slope of this solvent was about one-seventh that of ethanol. Though the hydrolysis of aspirin could be suppressed to some extent by using this solvent, it was still recommended for higher accuracy that the experimental procedure, from dissolution of aspirin to measurement of spectrum, was to be accomplished within 15 min. Since the time necessary for one scanning was about 1.5 min, measurement of spectrum for a sample could be repeated five times in 15 min. The influence of aspirin signal on DL value was investigated at three different salicylic acid concentrations of 5.0, 1.0, and 0.3 wg/mL and the amount of aspirin incorporated was 0-6 mg/mL. The measurement of spectrum was repeated five times for each samples and the mean value of five DL values
56
ANALYTICAL CHEMISTRY, VOL. 55, NO. 1, JANUARY 1983
Table I. Suppresion Effects of Acids on Aspirin Hydrolysis in Ethanol Solution concn in ethanol, acid added
%
Table 111. Accuracy in Second Derivative UV Spectrometric Determination of Salicylic Acid salicylic acid concn, pg/rnL
slope,' mm/h
2' 1.002 2.003 3.005 5.008 10.015
% accuracy
loo(% - 2)/2
x b
1.004 2.004 2.994 5.024 10.031
0.20 0.05 -0.37 0.32 0.16
none 15.1 phosphoric acid (85%) 1 10.0 acetic acid 1 5.2 chloroacetic acid 1 2.1 'The slope was obtained from the plot of D L values against time. Derivative conditions are given in the Experimental Section.
'Prepared value. Mean value of five x's calculated from D L values using eq 1.
Table 11. Effect of Aspirin Concentration on the Height of Second Derivative Satellite Peak of Salicylic Acid ( D L )
Table IV. Assay Results of Salicylic Acid in Commercial Aspirin Powders
salicylic acid concn, pg/mL 5.014
amt of aspirin added, mg/mL 0.0
6.01 1.003
0.0
6.14 0.301
0.0
6.15 a
fraction of salicylic acid, % 100.0
0.0834 100.0
0.0163 100.0
0.0049
DL,' mm
40.41 40.07 8.28 8.29 2.42 2.46
Mean value of the five duplicate measurements.
was taken. The results are shown in Table 11. At every salicylic acid concentration, there could not be seen any effect of the coexisting aspirin on DL values, even where the amount of aspirin incorporated was more than 20000 times that of salicylic acid (corresponding to salicylic acid fraction of 0.0049 % ). Calibration Curve for Salicylic Acid. T o obtain a calibration curve for salicylic acid, the second derivative spectra of standard salicylic acid solutions were taken for five varied concentrations (1.002-10.015 kg/mL). The spectrum was measured five times for each concentration, so that 25 sets of DL values (y) and salicylic acid concentrations (x)were obtained. The correlation coefficient between y and x was calculated to be 0.9999 and a plot of y against x yielded a good straight line which intercepted the origin. The slope was 8.01 mm/&g/mL) with a relative standard deviation (s/$ of 1.2%. These corresponded to the confidence limits of 8.01 f 0.04 mm/(kg/mL) a t the 95% level. The salicylic acid concentration x pg/mL could be calculated from l/slope X DL,that is x = 0.125 X DL (1) Accuracy of the second derivative method for salicylic acid was calculated by eq 2, using the data of the calibration curve % accuracy = (X - Z)/Z X 100 (%) (2) where It is the prepared concentration of salicylic acid in a standard solution and iis the mean value of the salicylic acid concentrations calculated by eq 1for each five DL values. As the results in Table I11 show, good accuracy was proved for the second derivative method a t every salicylic acid concentration. Assay of Commercial Aspirin Powders. Three kinds of commercially obtained aspirin powders A, B, and C were assayed and the results are listed in Table IV. The salicylic acid content of the aspirin powders tested fell well within 0.1 70. The statistical calculations of the assay results showed satisfactory precision of the derivative method. In conclusion, it was demonstrated that the second derivative UV spectrometry could permit a simple and time-saving assay of salicylic acid in aspirin powder which had sufficiently
salicylic acid aspirin concn, sam- concn, pg/ content, ple mg/mL mL % A
B
C
6.060 5.976 6.092 5.924 6.040
2.16 2.15 2.18 2.15 2.22
0.0356 0.0360 0.0358 0.0363 0.0367
6.248 5.928 6.032 6.100 5.968
2.38 2.24 2.26 2.30 2.20
0.0381 0.0378 0.0375 0.0377 0.0369
5.980 5.992 5.996 6.084 6.154
1.56 1.64 1.59 1.65 1.68
0.0261 0.0274 0.0265 0.0271 0.0273
mean 0.0361
i
0.0005'
i
0.0006'
i
0.0007'
s / F b = 1.2%
mean 0.0376 s / F b = 1.2%
mean 0.0269 s / F b = 2.1%
a Confidence limits at 95% level. deviation.
Relative standard
good accuracy and precision. The method could have a possibility of application to the similar problem of determining very small amounts of impurity in pure chemicals.
ACKNOWLEDGMENT The authors thank M. Takagi for her assistance with the experiments. Registry No. Salicylic acid, 69-72-7; aspirin, 50-78-2. LITERATURE CITED Talsky, G.; Mayring, L.; Kreuzer, H. Angew. Chem., I n t . Ed. fngl. i m . 17. 765-874. O'Haver, T. C. Anal. Chem. 1979, 5 1 , 91A-100A. Fell, A. F. UVG Bull. 1980, 8, 5-31. Fell, A. F. f r o c . Anal. Div. Chem. SOC. 1978, 15, 260-267. Such, V.; Traveset, J.; Gonzalo, R.; Gelpl, E. Anal. Chem. 1980, 5 2 , 412-419. Fell, A. F.; Davidson, A. G. J. Pharm. fharmacol., Suppl. 1980, 32, 97P. Fell, A. F.; Jarvle, D. R.; Stewart, M. J. Clln. Chem. (Winston-Salem, N.C.) 1981, 2 7 , 286-292. "Pharmacopeia of the United States XX"; United States Pharmacopelal Convention, Inc.: Rockvllle, MD, 1960; p 56. "Pharmacopeia Japonica Edltlo Deca"; Ministry of Health and Welfare: Tokyo, Japan, 1980; p 41. Patel, S.;Perrin, J. H.; Wlndheuser, J. J. J . Pharm. Sci. 1972, 6 1 , 1974- 1976. Baum, R . G.; Cantwell, F. F. J . fharm. Sci. 1978, 6 7 , 1066-1069. Gupta, V. D. J . fharm. Scl. 1980, 6 9 , 113-115. Green, G. L.; O'Haver, T. C. Anal. Chem. 1974, 4 6 , 2192-2196. Martln, A. N.; Swarbrlck, J.; Cammarata, A. "Physical Pharmacy", 2nd ed.; Lea & Feblger: Philadelphia, PA, 1969; p 388.
RECEIVED for review June 23,1982. Accepted September 28, 1982. Financial support of this research was provided by Kyoto College of Pharmacy. This research was presented in preliminary form a t the 102th Annual Meeting of Pharmaceutical Society of Japan, Osaka, April 1982.
