simultaneous determination of chlordiazepoxide and clidinium bromide in pharmaceutical formulations...

International Journal of Pharmaceutics 189 (1999) 67–74

Simultaneous determination of chlordiazepoxide andclidinium bromide in pharmaceutical formulations by

derivative spectrophotometry

M. Ines Toral *, Pablo Richter, Nelson Lara, Pablo Jaque, Cesar Soto,Marcela Saavedra

Department of Chemistry, Faculty of Sciences, Uni6ersity of Chile, P.O. Box 653, Santiago, Chile

Received 4 January 1999; received in revised form 9 July 1999; accepted 9 July 1999

Abstract

A direct and simple first derivative spectrophotometric method has been developed for the simultaneous determina-tion of clidinium bromide and chlordiazepoxide in pharmaceutical formulations. Acetonitrile was used as solvent forextracting the drugs from the formulations and subsequently the samples were evaluated directly by derivativespectrophotometry. Simultaneous determination of the drugs can be carried out using the zero-crossing method forclidinium bromide at 220.8 nm and the graphical method for chlordiazepoxide at 283.6 nm. The calibration graphswere linear in the ranges from 0.983 to 21.62 mg/l of clidinium bromide and from 0.740 to 12.0 mg/l ofchlordiazepoxide. The ingredients commonly found in commercial pharmaceutical formulations do not interfere. Theproposed method was applied to the determination of these drugs in tablets. © 1999 Elsevier Science B.V. All rightsreserved.

Keywords: Chlordiazepoxide; Clidinium bromide; Simultaneous determination; First derivative spectrophotometry; Pharmaceuticalformulations

www.elsevier.com/locate/ijpharm

1. Introduction

Chlordiazepoxide (7-chloro-N-methyl-5-phenyl-3H-1,4-benzodiazepin-2-amina-4-oxide) is a seda-tive-hypnotic drug widely employed as atranquilizer and antidepressant (Gasparic and Zi-mak, 1983) and clidinium bromide (3-[(hydroxy-diphenylacetyl) - oxy] - 1 - methyl - 1 - azoniabicyclo-

[2.2.2] octane bromide) is effective for anxiety-re-lated conditions including spastic colon (Rudyand Senkowski, 1973).

Tensoliv and Libraxim are pharmaceutical for-mulations which contain both drugs. Some chro-matographic techniques, including liquidchromatography (Yuen and Lehr, 1991) and areversed-phase high performance liquid chro-matography (Jalal et al., 1987) have been reportedfor their simultaneous determination. The pro-* Corresponding author. Fax: +56-2-2713888.

0378-5173/99/$ - see front matter © 1999 Elsevier Science B.V. All rights reserved.

PII: S0 378 -5173 (99 )00238 -0

M.I. Toral et al. / International Journal of Pharmaceutics 189 (1999) 67–7468

posed method by US Pharmacopoeia 23 (Anon.,1995) is a liquid-chromatographic method, involv-ing an UV detector and a 4-mm×25-cm columnthat contains packing L1. This simultaneous de-termination by HPLC requires careful prepara-tion of the column and it is expensive. On theother hand the derivative spectrophotometry ismore simple, inexpensive and has been used insimultaneous determination of organic com-pounds (Toral et al., 1996a,b; Wrobel-Zasada etal., 1996) and inorganic compounds (Toral et al.,1993, 1997; El-Sayed and Khalil, 1996).

In this work a simple, accurate, precise, rapidand inexpensive method by derivative spectropho-tometry is proposed. Simultaneous determinationof both compounds was carried out using acetoni-trile as solvent for extracting the drugs from theformulations and subsequently the samples wereevaluated directly by first order derivativespectrophotometry.

A study of the spectral behavior of these com-pounds in different solvents, the optimization ofthe spectral variables, the determination of pKavalues and the application of the proposedmethod to the determination of these drugs intablets are included in this work.

2. Materials and methods

2.1. Instruments

A Shimadzu UV-160 spectrophotometer with10-mm cells was used for measurements of theabsorbance and derivative absorption spectra. Forall solutions, the first derivative spectra wererecorded, over a range of 400–200 nm againstsolvent, at a scan speed of 480 nm/min with aDl=3.2 nm. The spectra derivatives were ob-tained digitally by software incorporated in theShimadzu UV-160 spectrophotometer.

2.2. Reagents

All reagents were of analytical reagent grade.Clidinium bromide (I) and chlordiazepoxide (II)were kindly provided by Laboratorio Chile, San-tiago, Chile.

Stock solutions of clidinium bromide and chlor-diazepoxide were prepared, by dissolving 21.6290.01 and 14.9890.01 mg of each compound inacetonitrile in order to give 1×10−3 M solution,corresponding to 432.4- and 299.8-mg/l solutionsof each specie. Other ranges of concentrationswere prepared by appropriate dilution using thesame solvent. The tablets containing clidiniumbromide and chlordiazepoxide were also dissolvedwith the same solvent and assayed.

2.3. Calibration procedure for determination of Iand II in mixtures

Aliquots of the stock solution of I and II weresimultaneously diluted in acetonitrile over theconcentration range 0.9–22.0 mg/l. The calibra-tion procedure was carried out for each com-pound in the presence of 8.6 mg/l clidiniumbromide solution and 6.0 mg/l chlordiazepoxidesolution, respectively. In all cases the correspond-ing absolute values of the first derivative spectraat 220.8 and 283.6 nm for clidinium bromide andchlordiazepoxide, respectively, were obtained andwere plotted against the correspondingconcentrations.

2.4. Procedure for determination of I and II intablets

A total of 20 tablets of each formulation wereweighed and powdered. A quantity of powderequivalent to 10–35 mg of tablet containing I andII was accurately weighed and dissolved in ace-tonitrile, transferred into separate 100 ml cali-brated flasks and diluted to mark. The contents ofthe flasks were shaken for 20 min and then thefirst derivative spectra were obtained.

3. Results and discussion

The structures of chlordiazepoxide and cli-dinium bromide are shown in Fig. 1. As can besee the structures of these drugs are quite differ-ent. Because of this, the spectral behavior of bothcompounds can be expected to be quite different.The spectral behavior of chlordiazepoxide in dif-

M.I. Toral et al. / International Journal of Pharmaceutics 189 (1999) 67–74 69

ferent solvents such as methanol, ethanol,dimethylsulfoxide, dimethylformamide, acetoni-trile and formamide, is similar. In all cases themolar absorptivities are near to 4000 l(mol×cm)−1. In contrast, clidinium bromide shows adifferent spectral behavior because the molar ab-sorptivities are considerably lower, with valuesbetween 100 and 200 l(mol×cm)−1. Only whenacetonitrile is used as solvent does the molarabsorptivity of this compound increase to 4000l(mol×cm)−1.

According to results obtained from this spectralstudy, acetonitrile was selected, because only inthis solvent is the ratio of the molar absorptivitiesfor both compounds near to one. Under thiscondition, the simultaneous determination is moreeffective and the sensitivity for the determinationof clidinium bromide is improved. Further, in thissolvent both drugs are stable for at least 24 h.

3.1. pKa Determination

The determination of pKa of chlordiazepoxidein Britton Robinson Buffer ranging between pH0.0 and 11.0 was carried out by the first derivativetechnique, plotting the derivative unit at l=260.6

Fig. 2. Determination of pKa of chlordiazepoxide and cli-dinium bromide in Britton Robinson Buffer. Chlordiazepox-ide: (A) derivative unit at 260.6 nm versus pH; (B) derivativeunit at 289.4 nm versus pH. Clidinium bromide: (C) derivativeunit at 213.0 nm versus pH; (D) derivative unit at 227.6 nmversus pH.

and l=289.4 nm versus pH. These derivativeunits correspond to the absorption of the acid andthe basic forms, respectively (Fig. 2). As can besee in Fig. 2A,B, a pKa of 4.9 was obtained: thisvalue is consistent with the pKa=4.8 reported inthe literature. In the case of clidinium bromide,the pKa determination was also carried out inBritton Robinson Buffer, ranging between pH 0.0and 13.0. The first derivative technique was used,plotting the derivative unit at l=213.0 nm (acidform) and l=227.6 nm (basic form) versus pH.Fig. 2C,D shows that the pKa of this compoundis near 11. This value is also in accordance withthe low characteristic acidity of clidinium bro-mide. For the pKa determination of both com-pounds, it was necessary to use the derivativespectrophotometry technique because, in bothcases, the bands of the acid forms and the basicforms were very overlapped.

3.2. Spectral features

The zero-order spectra of both analytes (Fig.3), using acetonitrile as solvent, show that onlychlordiazepoxide could be determined directly be-tween 240 and 350 nm. However, it is not possibleto determine clidinium bromide directly in mix-tures, because its spectrum is totally overlappedby that of chlordiazepoxide (Fig. 3). In order tocarry out simultaneous determinations of multi-

Fig. 1. The structure of chlordiazepoxide and clidinium bro-mide. (a) Chlordiazepoxide (7-chloro-N-methyl-5-phenyl-3H-1,4-benzodiazepin–2-amina-4-oxide); (b) clidinium bromide(3 - [(hydroxydiphenylacetyl) - oxy] - 1 - methyl - 1 - azoniabicyclo-[2.2.2] octane bromide).


Fig. 3. Spectra of chlordiazepoxide and clidinium bromidedissolved in acetonitrile measured against acetonitrile. (A)Clidinium bromide, 8.6 mg/l; (B) chlordiazepoxide, 6.0 mg/l.

ples of these types of approaches. The latter casehas been proposed in the resolution of mixtures ofPAHs (Rossi and Pardue, 1985), which presentsome problems when separated bychromatography.

In this work, we adopted derivative spectropho-tometry for resolution of analyte bands, becausethis approach is simple and it does not requiremany data mathematical processes.

3.3. Selection of spectral 6ariables

3.3.1. Deri6ati6e orderTo choose the optimum derivative order, the

first, second, third and fourth derivative spectra ofthe solution containing separately the respectiveanalytes were recorded (Fig. 4a–d).

As can be seen in Fig. 4a–d, the second, thirdand fourth derivatives are more resolved, butthese also have more noise than the first deriva-

components when the spectra overlap, there aresome chemiometric approaches which have beensuccessfully applied to spectrophotometric signals.Derivative spectrophotometry and multiwave-length evaluation methods are well-know exam-

Fig. 4. (a) First derivative spectra of chlordiazepoxide and clidinium bromide dissolved in acetonitrile measured against acetonitrile.(A) Clidinium bromide, 8.6 mg/l; (B) chlordiazepoxide, 6.0 mg/l. (b) Second derivative spectra of chlordiazepoxide and clidiniumbromide dissolved in acetonitrile measured against acetonitrile. (A) Clidinium bromide, 8.6 mg/l; (B) chlordiazepoxide, 6.0 mg/l. (c)Third derivative spectra of chlordiazepoxide and clidinium bromide dissolved in acetonitrile measured against acetonitrile. (A)Clidinium bromide, 8.6 mg/l; (B) chlordiazepoxide, 6.0 mg/l. (d) Fourth derivative spectra of chlordiazepoxide and clidiniumbromide dissolved in acetonitrile measured against acetonitrile. (A) Clidinium bromide, 8.6 mg/l; (B) chlordiazepoxide, 6.0 mg/l.


Fig. 5. (a) Effect of chlordiazepoxide over the first derivativespectra: (A) 1.0×10−5 mg/l; (B) 2.0×10−5 mg/l; (C) 3.0×10−5 mg/l. (b) Effect of the clidinium bromide concentrationover the first derivative spectra: (A) 1.0×10−5 mg/l; (B)2.0×10−5 mg/l; (C) 3.0×10−5.

5a shows that the graphical method (O’Haver,1982) at 283.6 nm can be used for determinationof chlordiazepoxide and Fig. 5b shows that thezero crossing method (O’Haver, 1982) can be usedfor the determination of clidinium bromide at220.8 nm. The selection of the analytical wave-lengths was confirmed by recording a series of thefirst derivative spectra of acetonitrile solution con-taining chlordiazepoxide and clidinium bromideseparately, between a concentration range of1.0×10−5 and 3.0×10−5 M of each analyte.Fig. 4a shows that at 286.3 nm, the distance, hB isonly proportional to the concentration of chlor-diazepoxide. In contrast the clidinium bromidedoes not absorb in this wavelength (Fig. 5b).Similarly, the measurement of the derivative spec-trum at an abscissa value of 220.8 nm (hA), corre-sponding to the zero-crossing point of thederivative spectrum of chlordiazepoxide, can besatisfactory to determine clidinium bromide (Fig.4a) without interference of chlordiazepoxide.

3.3.3. Selection of Dl 6alueIn order to select the Dl value for differentia-

tion, a series of first derivative spectra of a mix-ture of 6.0 mg/l chlordiazepoxide with increasingconcentration of clidinium bromide ranging from2.16 to 15.13 mg/l, were evaluated under theselected conditions, at different Dl values. In allcases good linearly was obtained (Table 1).

The sensitivity increase was proportional to theincrease in Dl, between 1.6 and 6.4 nm. Howeverin the selection of Dl it is also necessary to studythe effect of the presence the clidinium bromideon the chlordiazepoxide signal (Fig. 6a). Simi-

tives. In this context, the first derivative was se-lected in the simultaneous determination of thesecompounds, because the ratio signal/noise ishigher.

3.3.2. Selection of the analytical wa6elengthsThe selection of the analytical wavelengths was

carried out taking into account the first derivativespectra of both analytes separately (Fig. 4a). Fig.

Table 1Effect of Dl over the sensitivities

SensitivityAnalyte Intercept R Dl/nm

0.0007Chlordiazepoxide 0.999 1.60.0050.010 3.20.0014 0.999

0.999 4.80.015 0.00120.999 6.40.020 0.0017

1.60.9970.002Clidinium bromide 0.0040.997 3.20.008 0.003

0.012 4.80.9960.0046.40.9970.0030.016


Fig. 6. (a) Effect of concentrations of clidinium bromide overthe signals to analytical wavelength of 6.0 mg/l chlordiazepox-ide at different Dl values. (A) Dl=1.6 nm; (B) Dl=3.2 nm;(C) Dl=4.8 nm; (D) Dl=6.4 nm. (b) Effect of concentrationsof chlordiazepoxide over the signals to analytical wavelengthof 8.6 mg/l clidinium bromide at different Dl values. (A)Dl=1.6 nm; (B) Dl=3.2 nm; (C) Dl=4.8 nm; (D) Dl=6.4nm.

selected, the sensitivity increased but the accuracyand the precision were not affected, because underthese conditions there was no interference be-tween the analytes.

3.4. Analytical features

Under the working conditions stated, a linearrelation was observed between correspondingderivative units and the respective clidinium bro-mide and chlordiazepoxide concentrations. Thecalibration graphs (n=8) were obtained by plot-ting the first-derivative value hA for clidiniumbromide at 220.8 nm and hB for chlordiazepoxideat 283.6 nm using a Dl=3.2 nm value for differ-entiation, versus the analyte concentrations. Theequations of the regression line obtained were:

clidinium bromide: hA=0.008×C(mg/l)

+0.003 r=0.997 (1)

chlordiazepoxide: hB=0.010×C(mg/l)

+0.0014 r=0.999 (2)

where h is in derivative units and C correspondsto analyte concentration in mg/l.

The determination ranges were found to be0.983–21.62 mg/l for clidinium bromide and0.740–12.0 mg/l for chlordiazepoxide. The detec-tion limits (calculated by using the 3s recommen-dation) were found to be 0.295 mg/l for clidiniumbromide and 0.222 mg/l for chlordiazepoxide.

The repeatability of the determination ex-pressed as the relative standard deviation of 11replicates of solutions containing 8.6 mg/l cli-dinium bromide and 6.0 mg/l chlordiazepoxide,and the relative standard deviations were 2.4 and1.3%, respectively.

In order to establish the ratios at which onesubstance can be accurately measured one in pres-ence of the other, a study of the recoveries ofcompound was carried out in standard mixturesof clidinium bromide and chlordiazepoxide at dif-ferent ratios. The results (Table 2) show that thecontent of each compound can be reliably deter-mined in mixtures at different molar ratios, usingthis first derivative spectrophotometric method.

According to the results obtained, the simulta-neous determination of both compounds in

larly, a series of first derivative spectra of 8.65mg/l clidinium bromide with increasing concentra-tions of chlordiazepoxide ranging from 1.5 to10.49 mg/l were evaluated under the selected con-ditions, at different Dl values.

The sensitivities at different Dl are show inTable 1 and the effect of chlordiazepoxide concen-tration over signals of clidinium bromide at differ-ent Dl values are show in Fig. 6b. For clidiniumbromide, the sensitivity increase is also propor-tional to the increase in Dl, in same range men-tioned before.

According to Fig. 6a,b when Dl is equal to 1.6and 3.2 nm, in both cases the signals of theanalytes that remain constant are not affectedwhen the concentration of the variable analyteincreases. When Dl value equal to 3.2 nm was


Table 2Determination of clidinium bromide and chlordiazepoxide in different standard mixtures

Ratio I:II Found concentration per mg/laStated concentration per mg/l

III I II

2.003:1 6.1090.26.00 2.0690.34.001:2 2.0090.22.00 3.9790.28.00 3.9690.34.00 7.9590.41:28.00 1.9790.4 7.9390.31:4 2.00

10.00 1.9590.3 9.8690.42.001:5

a Mean of ten determinations.

4. Conclusion

The structures of chlordiazepoxide and cli-dinium bromide are quite different. Because ofthis, the spectral behavior of both compounds canbe expected to be quite different. In this work,after a study of the spectral behavior of chlor-diazepoxide and clidinium bromide it was possibleto select a solvent in which the ratio of the molarabsorptivities for both compounds was near toone. In these conditions the simultaneous determi-nation is more effective and it is possible to obtainaccurate results.

In addition, an accurate, precise, direct, andinexpensive method by first derivative spectropho-tometry has been developed for the simultaneousdetermination of clidinium bromide and chlor-diazepoxide in pharmaceutical formulations. Theingredients commonly found in commercial phar-maceutical formulations do not interfere. The pro-posed method is a good alternative for thesimultaneous determination of both drugs in phar-maceutical formulations.

Acknowledgements

The authors are grateful to the National Fundfor Development of Sciences and Technology(FONDECYT), project 1961024, for financial sup-port and Laboratory Chile for supply of the drugs.

References

Anon., 1995. The United States Pharmacopoeia 23. The Na-tional Formulary, 18, 334.

Table 3Simultaneous determination of chlordiazepoxide and clidiniumbromide by first derivative spectrophotometry

Amount found per mgDosage form

Chlordiazepoxide Clidiniumbromide

Tensoliv (Laboratorio 2.4690.054.9090.09Chile)

Libraxim (Roche) 4.9390.10 2.4290.04

tablets was possible, taking into account that theusual ratio of clidinium bromide and chlor-diazepoxide in commercial formulations is 1:2.

3.5. Application

The accuracy of the method was tested for bothdrugs by analysis of synthetic formulations. Sam-ples of 5.0 mg of chlordiazepoxide and 2.5 mg ofclidinium bromide were accurately weighed andmixed with 96.1 mg of excipients. The excipients inthe synthetic mixtures approximately containedmagnesium stearate+gelatine 3–5% and lactose-starch 95%. The recoveries were found to be97.890.5% and 101.090.5% for chlordiazepox-ide and clidinium bromide, indicating that thecommon excipients do not interfere. The methodwas applied to different tablet samples (Tensoliv,Libraxin, Lerogin) in which the nominal contentsof all the dosage forms are 5 mg chlordiazepoxideand 2.5 mg of clidinium bromide. The amountsfound for these different samples are shown inTable 3.


El-Sayed, A.I., Khalil, M.M.H., 1996. Simultaneous first-derivative spectrophotometric determination of iron(III)and molybdenum(VI) in cobalt-chromium and nickel-chromium alloys. Talanta 43 (4), 583–588.

Gasparic, J., Zimak, J., 1983. Analysis of the 1,4-benzodi-azepines by methods based on hydrolysis. J. Pharm.Biomed. Anal. 1 (3), 259–279.

Jalal, I.M., Sa’sa, S.I., Hussein, A., Khalil, M., 1987. HSreversed-phase high-performance liquid-chromatographicdetermination of clidinium bromide and chlordiazepoxidein tablet formulations. Anal. Lett. 20 (4), 635–655.

O’Haver, T.C., 1982. Derivative spectroscopy: theoretical as-pects. Anal. Proc. 19 (1), 22–28.

Rossi, D., Pardue, H., 1985. Effects of wavelength range onthe simultaneous quantitation of polynuclear aromatic hy-drocarbons with absorption spectra. Anal. Chim. Acta 175,153–161.

Rudy, B.C., Senkowski, B.Z., 1973. Comprehensive descrip-tion. In: Florey, K. (Ed.), Analytical Profiles of DrugSubstances, vol. 2. Academic Press, New York, pp. 145–161.

Toral, M.I., Richter, P., Silva, L., Salinas, A., 1993. Determi-nation of mixtures of cobalt and iron by first-derivative

spectrophotometry. Microchem. J. 48 (2), 221–228.Toral, M.I., Richter, P., Araya, E., Fuentes, M., 1996a. Deter-

mination of mefenamic acid and paracetamol by first-derivative spectrophotometry. Anal. Lett. 29 (15),2679–2689.

Toral, M.I., Richter, P., Martinez, O., 1996b. Determinacionsimultanea de paracetamol y codeina por espectrofotome-trıa derivada de primer orden. Bol. Soc. Chil. Quim. 41,283–289.

Toral, M.I., Richter, P., Rodrıguez, C., 1997. Simultaneousdetermination of copper and iron by second derivativespectrophotometry using mixtures of ligands. Talanta 45,147–153.

Wrobel-Zasada, K., Wrobel-Kaczmarczyk, K., Lopez-de-Alba, P.L., Lopez-Martinez, L., Garcia-Lopez, M.A.,1996. Application of internal standard for derivative spec-trophotometric determination of azinphos-methyl in com-mercial formulations. Talanta 43 (7), 1055–1060.

Yuen, S.M., Lehr, G., 1991. Liquid-chromatographic determi-nation of clidinium bromide and clidinium bromide-chlor-diazepoxide hydrochloride combinations in capsules. J.Assoc. Off. Anal. Chem. 74 (3), 461–464.

.

simultaneous determination of chlordiazepoxide and clidinium bromide in pharmaceutical formulations...

Documents