ANALYST. APRIL 1989, VOL. 113 505 Computerised Compensation Method for the Spectrophotometric Determination of a Single Substance in the Presence of Interferences Abdel-Aziz M. Wahbi, Hoda Mahgoub and Magda Barary Pharmaceutical Analytical Chemistry Department, Faculty of Pharmacy, Alexandria University, Alexandria 21521, Egypt The difference curves obtained using the compensation method for the spectrophotometric determination of a single substance in the presence of interferences were computed using a BASIC computer program. The exact balance-point (end-point) was defined on a plotter. When the absorption characteristics of the substance to be determined disappear, the resulting curve corresponds to the interferences, and the concentration of the substance in the sample solution is equal to that in the reference solution.The applicability of the method was illustrated by the determination of atropine sulphate in an injection solution and ephedrine hydrochloride and nadolol in tablets. The results showed good agreement with those obtained using the experimental compensation method and a second-derivative ultraviolet spectrophotometric method. Keywords: Computerised compensation spectrophotometry; interferences; pharmaceutical analysis; BASK program The compensation method’ is a non-mathematical method for the detection and elimination of unwanted absorption during spectrophotometric analysis. The unwanted absorption curve is assumed to possess the simplest possible shape and none of the characteristics of the pure compound. The method may therefore be regarded as an analogue of the more advanced types of mathematical correction which involve similar assumptions.The compensation method involves a compari- son of several difference spectra (sample, s, - reference, r) using different concentrations of a reference solution (c,) in the reference cell. Hence, if ASi and Arj refer to the absorbances of the relevant cells against air at a wavelength, i , then AAi = ASj - Arj. The characteristic peak of the pure compound, which may be observed in the difference curve, gradually decreases as c, increases and finally disappears at the balance-point (end- point), for which c, = c,. A further increase in c, then leads to an over-compensated difference curve which exhibits an inversion of the pure compound’s characteristic peak (Figs.The difference curve at the balance-point coincides with the unwanted absorption, if present. The accuracy of the compen- sation method for the determination of a single substance and for the analysis of multi-component mixtures depends on correct evaluation of the balance-point.*-4 Apart from its ability to compensate for unwanted absorp- tion, the method also possesses the usual advantage of a “null method” in that it cancels the non-linearities of both the instrument and the chemical system. Hence, when the balance-point is associated with a negligible difference curve, the spectrophotometer is little more than a “balance detector” and all the usual errors of measurement5 (excluding those due to cell path length) are eliminated.With regard to the chemical system, Beer’s law deviations,h which may be accentuated by constituent interactions, disappear at the final stage, when both “sample” and “reference” mixtures have been equalised. Further, although the interactions between the constituent(s) and the unwanted absorption remain, they usually represent only a minor problem. The work described in this paper involves replacing the practical compensation steps by a suitable BASIC program that can carry out the entire operation and define the balance- point. Only two solutions are required, a reference solution of suitable, known concentration and a sample solution. The absorbances of the two solutions are measured over a wide wavelength range in order to cover all the absorption 2-4).characteristics. The proposed method was applied to the determination of atropine sulphate in an injection solution (1 mg ml-I), ephedrine hydrochloride in tablets (30 mg per tablet) and nadolol in tablets (80 mg per tablet). Experimental Reagents Atropine sulphate crystals Merck, Darmstadt, FRG. Ephedrine hydrochloride crystals. Knoll, Chemische Fab- Nadolol. Squibb Laboratories, Princeton, NJ, USA. All other reagents and solvents were of analytical-reagent riken, Ludwigshafen am Rhein, FRG. grade. Instrumentation A Perkin-Elmer Model 550s UV - visible spectrophotometer with 1-cm quartz cuvettes and a Hitachi Model 561 recorder. an Olivetti M24 personal computer, 128 k byte, and an Epson HI-80 plotter were used. Reference Solutions Atropine sulphate. 0.08% mlV in 0.1 M hydrochloric acid.Ephedrine hydrochloride, 0.06% mlV in 0.05 M sulphuric Nadolol, 0.025% mlV in 0.1 M hydrochloric acid. acid. Pharmaceutical Preparations and Sample Solutions Atropine sulphate injection USP X X . Misr, 1 mg ml-1, used as received. Ephedrine hydrochloride tablets. Nile, labelled to contain 30 mg of ephedrine hydrochloride per tablet; average mass, 0.1345 g. Ephedrine hydrochloride sample solution. Weigh and powder 20 tablets. Transfer an amount of the powder equivalent to about 0.090 g of ephedrine hydrochloride into a 100-ml calibrated flask. Add about 70 ml of 0.05 M sulphuric acid, shake mechanically for 30 min, and then make up to the mark with 0.05 M sulphuric acid. Mix and filter, discarding the first few millilitres. Nadolol tablets (Corgard tablets).Squibb, Egypt, labelled to contain 80 mg of nadolol per tablet; average mass, 0.22597 g.ANALYST. APRIL 1989. VOL. 114 Start c2 Read data for sample, Read data for reference wavelength ( i ) and A,; lnitialise the plotter and draw x and y axes t I t -1 Enter the variable; K 1 Compute the data to be plotted AA; = [A,; - (KA,;)] 1 I Plot AA, against i I K7 Exit Fig. 1. Flow chart for the BASIC computer program Nadolol sample solutiorz. Weigh and powder 20 tablets. Transfer an amount of the powder equivalent to about 0.100 g of nadolol into a 100-ml calibrated flask. Add about 70 ml of 0.1 M hydrochloric acid, shake mechanically for 30 min and make up to the mark with 0 . 1 ~ hydrochloric acid. Mix and filter. discarding the first few millitres.Dilute 20 ml of the filtrate to 100 ml with 0.1 M hydrochloric acid. Second-derivative Spectra7.8 These were recorded between 290 and 240 nm at a scan speed of 60 nm min-1 and a chart speed of 120 nm min-1 with the response set at 6 and the ordinate minimum and maximum settings of -0.05 and 0.05, respectively. The amplitude was measured in millimetres for both the standard and sample solutions. Procedures Compiiterised compensation method Measure the absorbances of the reference, A,,, and sample, A,,, solutions against the solvent from 230 to 290 nm at 2-nm intervals. Using the BASIC program (Fig. I ) , plot A,, against wavelength, i. Note the absorption characteristics of the sample solution. Select the variable, K , so that KA,, is about 0.5A,, at the wavelength of maximum absorbance.Plot AA, (= A,, - KA,,) against the wavelength. Continue changing K and plotting the difference curves until the characteristics of the peak for the substance to be determined disappear completely and the balance-point (end-point) is reached at K = K, . The calculations are performed as follows: atropine sulphate in injection solution (mg ml-1) = ephedrine hydrochloride (mg per tablet) = KCP x 0.080 x (10001100) K, X 0.060 X (average masdmass taken) X 1000 0.6 0.4 a, U C ru e 0.2 n Q cn 0 - _ _ _ _ - - - - - _ -0.2 , j 240 250 260 270 280 Wavelengthinm Fig. 2. (A) Absorption curve of atropine sulphatc injection solution; (B and C) the computed difference curves obtained by compensation; (Z) the balance-point; and (D and E) over-compensated difference curves nadolol (mg per tablet) = K, x 0.025 x (average massimass taken) x (100120) x 1000 Experimental compensation method Record the absorption curve of the sample solution against the solvent. Prepare a series of reference solutions (by appro- priate dilution of a suitable reference solution) starting with concentration differences of 10, 5 and 2 mg per cent.for atropine sulphate, ephedrine hydrochloride and nadolol. respectively, followed by a second series with concentration differences of 1, 1 and 0.5 mg per cent., respectively, near the balance-point. The concentration increments are selected to give effective difference curves when using the compensation technique. Reducing the increments does not lead to notice- able differences. Keeping the sample solution in the sample cell, fill the solvent cell with the reference solutions in succession starting with the most dilute reference solution and recording the difference absorption curve in each instance under the sample absorption curve.Follow the disappearance of the absorption characteristics of the substance to be determined by increasing the concentration of the reference solution, c,. Determine the exact balance-point at which the concentration of the substance in the sample solution is equal to that in the reference solution. Results and Discussion The compensated difference curves, AA,, can be regarded as being equal to A,, - KA,,, where A, and A, represent the absorbance of the sample and reference solutions, respect- ively, at a wavelength, i ; the value of K can be varied to reach the balance-point in the computerised compensation method.If the sample is as pure as the reference, then A,, - KA,, = 0 , i.e., the balance-point coincides with the wavelength scale. Fig. 1 shows a flow chart for the BASIC program used. Fig. 2 shows the difference curves obtained with the computerised compensation method for atropine sulphate in an injection solution. The peak characteristics decrease as K increases and finally disappear at the balance-point (end-point) where K = K , p . The curve obtained at the balance-point represents the absorption curve of the interferences present in the injection solution. This is not surprising because the atropine sulphate injection solution9 is a sterile solution of atropine sulphate in water (0.1%); the acidity of the solution is adjusted to pH 3. Owing to the relatively low molar absorptivity of atropine sulphate, the injection solution is measured without any further dilution.The additives used to adjust the pH of the solution are the source of the absorbing interferences.ANALYST. APRIL 1989, VOL. 114 507 Table 1. Spectrophotometric determination of atropine sulphate in an injection solution using the computerised compensation, experimental compensation and second-derivative UV methods Q) m 0.4 +? 8 2 0.2 0 -0.2 1 I I 230 240 250 260 270 Wavelengthhm Fig. 3. (A) Absorption curve of ephedrine hydrochloride \ample solution in 0.05 M sulphuric acid; (B and C) the computed difference curves obtained by compensation; ( Z ) the balance-point; and (D and E ) over-compensated difference curves 8 0.2 c m 11 L g o a 11 -0.2 -0.4 250 260 270 280 290 Wavelengthinm Fig.4. (A) Absorption curve of nadolol sample solution in 0.1 M hydrochloric acid; (B and C) the computed difference curves obtained by compensation; and (D and E) over-compensated difference curves. Note that the balance-point coincides with the wavelength axis A further increase in K so that it exceeds Kc.p. leads to an inversion of the pure compound's characteristic peak and results in over-compensated difference curves (Fig. 2). This last step is sometimes necessary to locate the balance-point accurately. Figs. 3 and 4 show the difference curves obtained with the computerised compensation method for ephedrine hydro- chloride and nadolol sample solutions, respectively.The curve obtained at the balance-point shown in Fig. 3 represents the unwanted absorption caused by tablet fillers and excipients. In the example shown in Fig. 4, the balance-point was found to coincide with the wavelength axis, indicating that the interfer- ence from tablet fillers and excipients was negligible owing to excessive dilution during the preparation of the final sample solution. In these instances, the absorption curves of the interferences are linear. Accordingly, they can be eliminated by recording derivative absorption curves7.10.11 and, in par- ticular, second-derivative spectra.7.8,' The computerised compensation, experimental compensa- tion and second-derivative spectrophotometric methods were applied to the determination of atropine sulphate in an injection solution and ephedrine hydrochloride and nadolol in tablets.The results obtained are shown in Tables 1, 2 and 3, respectively. In view of the fact that location of the balance- point is subjective and in order to eliminate any personal bias, Method Computerised Experimental co m pe n s a t i o n Experi- compensation ment No. K , * Found, Yo f c, , Yo $ Found. (Yo j- 1 1.24 99.2 0.100 100.0 2 1.25 100.0 0.101 101.0 3 1.25 100.0 0.100 100.0 4 1.25 100.0 0.101 101.0 K , = K at the balance-point. + Percentage of label claim (1 mg ml-1) found. ri: c, = reference concentration at the balance-point. Second derivative Found. '%t 100.4 100.6 100.1 - Table 2. Spectrophotometric determination of ephedrine hydrochloride in tablets using the computerised compensation.experimental compensation and second-derivative UV methods Method Computerised Experimental Second Experi- compensation cornpensation derivative ment No. K , Found. % ' c,/mg o/o Found, Yo* Found, % * 1 1.40 93.3 85 94.4 94.3 2 1.41 94.0 85 93.4 93.9 3 1.40 93.3 84 93.3 93.4 4 1.41 94.4 85 94.4 93.6 * Percentage of label claim (30 mg per tablet) found. Table 3. Spectrophotometric determination of nadolol in tablets using the computerised compensation. experimental compensation and second-derivative UV methods Method Computerised Experimental Second Experi- compensation compensation derivative ment No. K , , Found, cJmg Yo Found, o/o * Found. "/C, * 1 0.78 97.5 19.5 97.5 97.4 2 0.77 96.2 19.5 97.5 97.6 3 0.77 96.6 19.5 97.5 97.3 4 0.78 97.5 19.5 97.5 97.2 * Percentage of label claim (80 mg per tablet) found.four analysts determined the exact balance-point indepen- dently using both methods (i.e., computerised and experimen- tal compensation methods ) for the three compounds investi- gated. Each assay result was obtained by one analyst. With regard to precision, all the results were within ~ 1 % of each other and of their mean. Regarding the accuracy of the computerised and experimental compensation methods, the average percentage value found deviated from that obtained using the second-derivative method by less than 1%. This indicates that both compensation methods give precise and accurate results. However, the computerised method needs only two solutions, a reference and a sample solution.The final result is obtained on the plotter by applying the BASIC program. The repeatability of the computerised and experimental compensation methods cannot be evaluated from the results obtained by one analyst due to bias. Further comparative studies between the computerised and experimental compensation methods were carried out. Hence, the difference curves obtained mid-way through the compensation technique and at the balance-point for the three compounds were compared over the wavelength range used.508 ANALYST. APRIL 1989. VOL. 114 The differences between the experimental and computed difference curves were found to be negligible and only affected the third decimal place of the values obtained. This is further confirmation that the computerised version of the experimen- tal Compensation technique can be applied successfully to the determination of a single substance in the presence of unknown interferences.The proposed method is vercatile, fast and easily automated for routine analysis. Further, with the advent of modern spectrophotometers, which can be interfaced to a computer, the entire operation can be carried out with the minimum of time and effort. The procedure can be concidered to be a specfrophotometric titration technique for the determination of a single substance in the presence of interferences. Work is continuing to extend the method to two- and multi-component analyses. It should be noted that for successful application of the computerised method, the substance should obey Beer‘s law over the concentration range involved in the compensation process and it should not interact with the interferences present or with the solvent used.Adherence to Beer’s law can be tested by conventional means. Possible interaction between compounds can be checked by testing for the additive properties” of their absorbances. I . 2. 3 . 4. 3. 6. 7 . 8. 9. 10. 1 1 . 12. 13. References Jones. J . 1 1 . . Clark. G. R . , and Harrow, L. S . , J . A.ssoc. Ojy. Agric. (‘hem.. 1951, 34, 135. Schiaffino, S . S . , Loy, H. W., Kline, 0. I-., and Harroiv, L. S., J . Assoc. Off. Agric. Chcrn., 1956. 39. 180. , J . Ptiarm. Sci., 1961. 50, 693. . F.. Anal. Chrm., 1961, 33, 927. Goldring, L. S., Hawes, R. C., Hare, G. H . . Bcckman, A . 0.. anti Stickncy, M. E., Anal. Chrwz., 1953. 2 5 . 869. Loth i a t i , G . F . , “Absorption Spec t rop h o t o m e t ry . ’’ S e co ti d Edition, Hilger and Watts, London, 1958, p. 75. Davidson, A . G., and Elsheikh, H . , Analysi. 1982. 107, 879. Davidson, A . G . , and Hassan, S. M.. .I. Pharm. Sci., 1984. 73, 413. L‘United States Pharmacopeia XXl.” Twinbrook Parkway. Rockville, MD, 1985, p. 80. Wahbi, A. M.. and Ebel, S., Anal. Chirri. Actu. 1974, 70. 57. Fell, A . F.. Proc. Anal. Div. C‘hem. SOC.. 1978, 15, 260. Davidson, A . G . , and Hassan, S. M . , J . Pharrri. Phurmucol.. 1984, 36. 7 . Wahbi, A. M., and Barary. M., Anal. Lett., 1983. 16(A20), 1617. Puper 8101 H50F Received May I I th, I988 Accepted Decmzber 7th, I088