- Email: [email protected]
Sensitive determination of phenothiazines in pharmaceutical preparation and biological fluid by flow injection chemiluminescence method using luminol–KMnO4 system
Talanta 71 (2007) 1124–1129
Sensitive determination of phenothiazines in pharmaceutical preparation and biological fluid by flow injection chemiluminescence method using luminol–KMnO4 system Yinhuan Li a,b , Weifen Niu a , Jiuru Lu a,∗ a
School of Chemistry and Materials Science, Shaanxi Normal University, Xi’an 710062, China b School of Science, Xi’an Jiaotong University, Xi’an 710049, China Received 23 March 2006; received in revised form 2 June 2006; accepted 3 June 2006 Available online 17 July 2006
Abstract A flow injection chemiluminescence method was described for the determination of four phenothiazine drugs, namely, chlorpromazine hydrochloride, perphenazine hydrochloride, fluphenazine hydrochloride and thioridazine hydrochloride. Strong Chemiluminescence (CL) signal was produced when above-mentioned drug was injected into the mixed stream of luminol with KMnO4 . The linear ranges of the method were 0.0020–1.0 g/mL chlorpromazine hydrochloride, 0.0040–3.0 g/mL perphenazine hydrochloride, 0.0020–5.0 g/mL fluphenazine hydrochloride and 0.0050–1.0 g/mL thioridazine hydrochloride. The detection limits were 0.4 ng/mL chlorpromazine hydrochloride, 0.7 ng/mL perphenazine hydrochloride, 2 ng/mL fluphenazine hydrochloride and 0.7 ng/mL thioridazine hydrochloride. The proposed method was applied to the determination of chlorpromazine hydrochloride in injections and in mental patient’s urine samples and the satisfactory results were achieved. The possible CL reaction mechanism was also discussed briefly. © 2006 Elsevier B.V. All rights reserved. Keywords: Chemiluminescence; Flow injection; Phenothiazines
1. Introduction Phenothiazine derivatives are a large group of tricyclic antidepressants which are commonly used for the treatment of psychiatric patients suffering depressions [1]. The monitoring of these compounds is important for quality assurance in pharmaceutical industry and for obtaining optimum therapeutic concentrations in body fluids to minimize the risk of toxicity. Therefore, it is important to develop simple and sensitive methods for the determination of these drugs. A wide variety of analytical techniques are available for the determination of phenothiazine drugs in pharmaceutical preparations and biological samples, such as spectrophotometry [2–4], electrochemical methods [5–7], chromatography [8] and capillary electrophoresis [9].
∗
Corresponding author. Tel.: +86 29 85303911; fax: +86 29 85307774. E-mail address: [email protected] (J. Lu).
0039-9140/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.talanta.2006.06.007
Chemiluminescence (CL) method has been frequently used for the analysis of pharmaceutical compounds in recent years [10–14] because of its promising advantages of low detection limit, wide linear dynamic range and relatively simple and inexpensive instrumentation. Several CL systems have been reported for the determination of phenothiazine drugs [15–23]. They were based on the acidic KMnO4 reaction [15–17], Ce(IV) reaction sensitized by rhodamine B [18], luminol–H2 O2 –Cr3+ reaction [19], luminol–Fe2+ reaction [20], luminol reaction in micelles [21] and electrogenerated Ru(bpy)3 2+ reaction [22,23]. The analytical parameters of the previous reported CL methods for the determination of phenothiazine drugs were summarized in Table 1. We here proposed a simple and sensitive flow injection CL method for the determination of four phenothiazine drugs, including chlorpromazine hydrochloride, perphenazine hydrochloride, fluphenazine hydrochloride and thioridazine hydrochloride. It was based on the fact that strong CL signal could be produced when above-mentioned drug was injected into the reaction mixture of luminol with KMnO4 . The method
Y. Li et al. / Talanta 71 (2007) 1124–1129
1125
Table 1 Analytical parameters of the previous CL methods for the determination of phenothiazine drugs (unit: g/mL) Species
CL reactions
Linear ranges
Chlorpromazine Thioridazine Fluphenazine Chlorpromazine Chlorpromazine Chlorpromazine Chlorpromazine Chlorpromazine
Acidic KMnO4 reaction Acidic KMnO4 reaction Ce(IV) reaction sensitized by rhodamine B Luminol–H2 O2 –Cr3+ reaction Luminol–Fe2+ reaction Luminol reaction in reverse micelles Electrogenerated Ru(bpy)3 2+ reaction Electrogenerated Ru(bpy)3 2+ reaction
1.0–30.0 0.5–90 3.6–355 0.7–7.1 0.05–10 0.3–6.4 0.004–1.1
was successfully applied to the determination of chlorpromazine hydrochloride in injections and in mental patient’s urine samples. 2. Experimental 2.1. Apparatus The schematic diagram of the CL flow system used was shown in Fig. 1. PTFE tubing (0.8 mm i.d.) was used as the connect material in the flow system. A peristaltic pump was used to deliver all solutions; each at a flow rate of 2.1 mL/min. Sample injection (50 L) was automated operated by means of a sixway valve. The CL signal produced in the flow cell was measured using an IFFS-A multifunction CL analyzer (Xi’an Remax Electronic Science-Tech Ltd. Co., China). The CL data acquisition and treatment were performed by using an MCDR-A multifunction data processing system (Xi’an Remax). CL spectra and fluorescence spectra were taken on CRT-970 fluorescence spectrophotometer (Shanghai Third Analytical Instrumental Plant).
Detection limits
References
7.1 0.5 0.01 13
[15] [17] [18] [19] [20] [21] [22] [23]
0.006 0.24 0.001
tions (500.0 g/mL) of each drug were prepared by dissolving 50.0 mg of the drug in 100 mL of water. They were stored in a refrigerator and protected from light. Working solutions were prepared by appropriately diluting the stock solution when used. Stock solution (0.01 mol/L) of luminol was prepared by dissolving 1.771 g luminol in 50 mL of 1 mol/L sodium hydroxide and diluting to 1 L with water. Stock solution (0.02 mol/L) of potassium permanganate was prepared by dissolving 3.16 g of KMnO4 in 1 L of boiled water, filtering through glass wool and protecting from light. 2.3. Procedure As shown in Fig. 1, flow lines were connected with phenothiazine solution, luminol solution and KMnO4 solution, respectively. Luminol solution was firstly mixed with KMnO4 solution via a Y-piece to give a stable baseline. Then 50 L solution of phenothiazine was injected into the merged stream of luminol with KMnO4 by means of a six-way valve to produce CL. The concentration of phenothiazine was quantified by the CL intensity.
2.2. Reagents and solutions 3. Results and discussion All chemicals were of analytical reagent grade except luminol, which was synthesized by the Institute of Analytical Science of Shaanxi Normal University (Xi’an, China). Doubly distilled water was used throughout the experiments. Chlorpromazine hydrochloride, perphenazine hydrochloride, fluphenazine hydrochloride and thioridazine hydrochloride were obtained from the Chinese Pharmaceutical and Biological Test Institute (Beijing, China). Potassium permanganate was purchased from Xi’an Chemical Plant (Xi’an, China). Stock solu-
3.1. Kinetic characteristic of the CL reaction The CL kinetic characteristic of the reaction was investigated using the static system of the IFFS-A multifunction CL analyzer. Fig. 2 shows the typical CL intensity-time profile. A strong CL was recorded when 1.0 mL solution of 20 mol/L KMnO4 was injected into 1.0 mL solution of 10 mol/L luminol (peak a). After CL signal declined about 45s, subsequently injecting 1.0 mL 0.5 g/mL chlorpromazine hydrochloride resulted in a new CL reaction (peak b). About 80s later, the CL reaction terminated and the CL signal declined to baseline. 3.2. Optimization of experimental conditions
Fig. 1. Schematic diagram of CL flow system. (a) Phenothiazines solution; (b) luminol solution; (c) potassium permanganate solution; P, peristaltic pump; V, injection valve; F, flow cell; PMT, photomultiplier tube; HV, high voltage; PC, personal computer; W, waste.
The experimental conditions were optimized using 0.05 g/mL chlorpromazine hydrochloride solution as a model. The optimized parameters were of reaction medium, luminol concentration and KMnO4 concentration. It was observed that higher CL signal and better repeatability could be obtained when sodium hydroxide was added into sample solution than into luminol solution. The effect of sodium
1126
Y. Li et al. / Talanta 71 (2007) 1124–1129
3.4. Interference
Fig. 2. The CL intensity-time profile. (a) injecting 1.0 mL 20 mol/L KMnO4 into 1.0 mL 10 mol/L luminol; (b) injecting 1.0 mL 0.5 g/mL chlorpromazine hydrochloride into above reaction mixture.
hydroxide concentration on the CL reaction was examined in the 0.02–0.1 mol/L. Finally, 0.05 mol/L of sodium hydroxide was employed because it gave larger signal-to-blank ratio and higher CL signal. The concentrations of luminol and KMnO4 were important parameters in the experiments. The effect of different concentrations combination of luminol (1–100 mol/L) and KMnO4 (1–500 mol/L) was studied. The experimental results showed that the maximum CL signal and signal-to-blank ratio was obtained when 10 mol/L of luminol and 20 mol/L of KMnO4 were used. The effect of flow rate on the reaction was examined in the range 0.7–3.5 mL/min. The CL signal increased with increase in flow rate from 0.7 to 2.1 mL/min. When the flow rate was higher than 2.1 mL/min, the CL signal changed slightly. Therefore, 2.1 mL/min of flow rate was employed. 3.3. Analytical performance Under the selected experimental conditions, the calibration graphs for four phenothiazine drugs were prepared. The parameters of calibration graphs and the detection limits (3sb ) were summarized in Table 2. The relative standard deviations for 0.01 and 0.1 g/mL solutions of phenothiazine drugs in 11 repeated measurements were also showed in Table 2. A complete analysis, including sampling and injecting, could be performed in 30 s, giving a sample throughput of 120/h.
In order to assess the possibility of applying the proposed method, the effect of foreign species was examined by analyzing synthetic sample solutions containing 0.050 g/mL of chlorpromazine hydrochloride and various amount of each potential interferent. A foreign species was considered not to interfere if it caused a relative error less than 5% in the determination of chlorpromazine hydrochloride. No interference has been found when including up to a 1000-fold Na+ , K+ , Al3+ , Fe3+ , Fe2+ , NO3 − , SO4 2− , CO3 2− , PO4 3− , starch, glucose, lactose, leucine, 500-fold Ca2+ , tryptophan, 200-fold urea, 100-fold Mg2+ , Cl− , 10-fold Mn2+ , Pb2+ , 1-fold fluphenazine hydrochloride, 0.2-fold ascorbic acid, epinephrine and 0.02-fold uric acid. 3.5. Application 3.5.1. Determination of chlorpromazine hydrochloride in pharmaceutical preparations Ascorbic acid is commonly used as the anti-oxidant for chlorpromazine hydrochloride injection [24]. The experimental results indicated that the effect of ascorbic acid could be eliminated by the addition of Fe3+ . Experimental result indicated that 50-fold amount of ascorbic acid did not interfere with the determination of 0.05 g/mL of chlorpromazine hydrochloride when 50 g/mL Fe3+ was present in sample solution. Five injections were homogenized. A portion of sample was transferred into a 50 mL beaker, to which 2.5 mL of 1 mg/mL Fe3+ solution was added. The solution was stood for 1 min and then 5 mL of 0.5 mol/L NaOH solution were added. The resultant solution was diluted with water to 50 mL and filtered. The filtrate was then analyzed according to the procedure described in Section 2. The results were summarized in Table 3. The t-test assured that there were no significant differences between the results obtained by the proposed method and those obtained by the Chinese Pharmacopoeia method [25] at the confidence level of 95%. 3.5.2. Determination of chlorpromazine hydrochloride in mental patient’s urine In order to apply the proposed method to the determination of chlorpromazine hydrochloride in mental patient’s urine, a chlorpromazine-imprinted polymer was employed to eliminate the interferents existed in urine. The chlorpromazine-imprinted polymer was synthesized by using batch method with chlorpro-
Table 2 Analytical performances Phenothiazines
Chlorpromazine hydrochloride Perphenizine hydrochloride Fluphenazine hydrochloride Thioridazine hydrochloride
Linear range (g/mL)
0.002–1.0 0.004–3.0 0.002–5.0 0.005–1.0
Regression equation (ng/mL)
I = 4.62C + 92.0 I = 2.33C − 38.8 I = 0.76C + 30.0 I = 3.20C − 40.0
Correlation coefficient (r2 )
0.9938 0.9972 0.9972 0.9959
Detection limit (3sb ) (ng/mL)
0.4 0.7 2 0.7
R.S.D. (%) (n = 11) 0.01 g/mL
0.1 g/mL
1.2 1.5 1.6 2.0
0.4 0.3 0.5 0.6
Y. Li et al. / Talanta 71 (2007) 1124–1129
1127
Table 3 Determination of chlorpromazine hydrochloride in injections Sample
Nominal content
Proposed methoda
R.S.D. (%) (n = 5)
Pharmacopoeia methodb
R.S.D. (%) (n = 3)
040912 041110
25 mg/1 mL 50 mg/2 mL
24.9 ± 0.5 (mg/1 mL) 49.9 ± 1.0 (mg/2 mL)
2.1 1.9
24.5 ± 0.2 (mg/1 mL) 49.2 ± 0.4 (mg/2 mL)
0.8 0.9
a b
Average of five determinations ± S.D. Average of three determinations ± S.D.
mazine as the template molecule and the procedure was similar to that reported previously [26]. The procedure of molecular imprinting-chemiluminescence method (MI-CL) consisted of four steps: adsorption of chlorpromazine, removing the other substances except chlorpromazine, CL detection and cleaning the MIP column. The flow system and procedure has been described in detail in elsewhere [27]. The experimental conditions corresponding to the four steps of MI-CL method were examined. The optimum conditions were 180 s of adsorption time, 120 s of washing time, 80 s of detection time and 80 s of cleaning time. When chlorpromazine-imprinted polymer was used, the tolerance limits were 1000-fold urea, 200-fold uric acid, 100-fold ascorbic acid, 50-fold fluphenazine hydrochloride and 2-fold epinephrine. Blank urine samples were collected from the healthy volunteers and the urine samples from the mental patients were collected after taking chlorpromazine hydrochloride for 8 h. After centrifuging at 3000 rpm for 15 min, 1 mL of the supernatant was transferred into a 100 mL volumetric flask, diluted with water to the mark and used as the sample solution. The results for the determination of chlorpromazine hydrochloride in urine samples were summarized in Table 4. Recovery tests were also carried out by adding a known amount of chlorpromazine hydrochloride in samples and the results were also summarized in Table 4. The t-test assumed that there were no significant differences between recoveries and 100% at confidence level of 95%. 3.6. Discussion on the reaction mechanism To elucidate the reaction mechanism, the following experiments were performed using chlorpromazine hydrochloride as a model.
Firstly, weak CL was recorded when alkaline chlorpromazine hydrochloride solution was injected into KMnO4 . Strong CL was detected when chlorpromazine hydrochloride solution was injected into the mixed solution of luminol and KMnO4 . Secondly, the CL spectra of luminol–KMnO4 reaction and luminol–KMnO4 –chlorpromazine hydrochloride reaction were drawn by means of a refitted CRT 970 fluorescence spectrophotometer (Fig. 3). Both spectra had same maximum emission wavelength at 425 nm. This indicated that the CL spectra were independent of chlorpromazine hydrochloride, the emitter of luminol–KMnO4 –chlorpromazine hydrochloride reaction was still the excited state of 3-aminophthalate ions (3-AP*) [28], the oxidation product of luminol [29]. Thirdly, when luminol was mixed with KMnO4 solution, the purple of KMnO4 solution changed gradually into green. This result indicated KMnO4 was firstly reduced to K2 MnO4 [30] and luminol was oxidized to 3-aminophthalate ions (3-AP) [29]. Fourthly, no CL signal was observed when injecting chlorpromazine hydrochloride into the mixed solution of luminol and K2 MnO4 . This experiment indicated that K2 MnO4 was irrelevant to the CL reaction. Fifthly, the alkaline mixed solutions of luminol and KMnO4 with different concentration ratio were prepared remaining luminol concentration at 1.0 × 10−5 mol/L. The concentration ratios of luminol to KMnO4 were 1:0.5 for solution A and 1:5 for solution B. When chlorpromazine hydrochloride was injected into above solution, the CL signal from solution B was much higher than that from solution A. This experiment suggested that KMnO4 was indispensable to the CL reaction. Sixthly, the fluorescent spectra of the reaction mixtures of luminol–KMnO4 (a), KMnO4 –chlorpromazine hydrochloride (b) and luminol–KMnO4 –chlorpromazine hydrochloride
Table 4 Determination of chlorpromazine hydrochloride in urine samples Urine samples
Determineda (g/mL)
Volunteer 1 Volunteer 2
R.S.D. (%, n = 3)
Added Found (g/mL) (g/mL)
Recovery (%)
0
1.00 2.00
0.96 2.04
96 102
0
1.00 2.00
1.03 1.95
103 98
Mental patient 1 0.42 ± 0.01
3.3
1.00 2.00
1.37 2.36
95 97
Mental patient 2 0.78 ± 0.02
3.1
1.00 2.00
1.74 2.71
96 97
a
Average of three determinations ± S.D.
Fig. 3. CL spectra. The CL spectra of the reaction of (a) 10 mol/L luminol + 20 mol/L KMnO4 ; (b) 10 mol/L luminol + 20 mol/L KMnO4 + 2.0 g/mL chlorpromazine hydrochloride.
1128
Y. Li et al. / Talanta 71 (2007) 1124–1129
4. Conclusion
Fig. 4. The fluorescent spectra. The excitation and emission spectra of the reaction mixtures of (a, a ): 10 mol/L luminol + 20 mol/L KMnO4 ; (b, b ): 20 mol/L KMnO4 + 2.0 g/mL chlorpromazine hydrochloride; (c, c ): 10 mol/L luminol + 20 mol/L KMnO4 + 2.0 g/mL chlorpromazine hydrochloride.
(c) were scanned using a CRT 970 fluorescence spectrophotometer (Fig. 4). The excitation spectra and emission spectra were similar for the reaction mixtures (a, c), the excitation wavelength and emission wavelength were within the range of 240–400 nm (λex = 314 nm) and 370–600 nm (λem = 425 nm), respectively. They were identical with the excitation spectrum and emission spectrum of 3-AP [28]. The excitation wavelength and emission wavelength for reaction mixture (b) were 240–370 nm (λex = 306 nm) and 350–470 nm (λem = 386 nm), respectively. It was identical with the spectrum of the oxidation product of chlorpromazine hydrochloride [31], sulphone compound [32]. It was also interestingly found that the excitation spectrum of sulphone compound (b) largely overlapped with that of 3-AP (a, c), which suggested that the energy released by reaction between KMnO4 and chlorpromazine hydrochloride could be absorbed not only by sulphone compound to form excited state of sulphone compound but also by 3-AP to form 3-AP*, respectively. Based on the above discussion, the following possible reaction mechanism was suggested: Chlorpromazine hydrochloride reacted with KMnO4 and the energy was released. The released energy was absorbed by 3-AP to form 3-AP* . The 3-AP* returned to ground state and generated CL. Luminol + KMnO4 → 3-AP + K2 MnO4 + hν
(425 nm)
KMnO4 + chlorpromazinehydrochloride + 3-AP → sulphonecompound + K2 MnO4 +3-AP∗ 3-AP∗ → 3-AP + hν
(425 nm)
The same phenomenon was also found when chlorpromazine hydrochloride was replaced by other three phenothiazine drugs. This may be attributed to their similar molecular structure and fluorescence characteristics [31]. Therefore, the CL reaction mechanism for other three phenothiazine drugs should be identical with that of chlorpromazine hydrochloride.
Four phenothiazine drugs including chlorpromazine hydrochloride, perphenazine hydrochloride, fluphenazine hydrochloride and thioridazine hydrochloride were found to produce strong CL in KMnO4 –luminol reaction. The proposed method has wider linear ranges (The linear range extended 2–3 magnitudes) and lower detection limit (The detection limit was two magnitudes lower than that of most reported CL methods). The proposed method was successfully applied to the determination of chlorpromazine hydrochloride in injections and in mental patient’s urines. The possible CL reaction mechanism was investigated. Acknowledgement The authors gratefully acknowledge the financial support from Shaanxi Normal University for Doctoral Fund. References [1] J. Karpinska, B. Starczweska, H. Puzanowska-tarasiewicz, Anal. Sci. 12 (1996) 161–170. [2] M.A. El-Maaboud Ismail, Talanta 44 (1997) 1173–1182. [3] T. Aman, J. Anwar, A. Ahmad, L. Latif, Anal. Lett. 36 (2003) 2961–2974. [4] K. Basavaiah, J.M. Swamy, Chem. Anal. 47 (2002) 139–146. ˇ anek, J. [5] A. Ferancov´a, E. Korgov´a, T. Buzinkaiov´a, W. Kutner, I. Step´ Labuda, Anal. Chim. Acta 447 (2001) 47–54. ¨ [6] Z. Yang, J. Kauffmann, M.A. Valenzuela, S. Ozkan, Mikrochim. Acta 131 (1999) 85–90. [7] I. Nemcova, K. Nesmerak, D. Rychlovsky, J. Koutnikova, Talanta 65 (2005) 632–637. [8] M.C. Quintana, M.H. Blanco, J. Kacal, L. Hernandez, Talanta 59 (2003) 417–422. [9] F.J. Lara, A.M. Garcia-Campana, F. Ales-Barrero, J.M. Bosque-Sendra, Anal. Chim. Acta. 535 (2005) 101–108. [10] Y. Fuster Mestre, L. Lahuerata Zamora, J. Martinez Calatayud, Luminescence 16 (2001) 213–235. [11] A. Townshend, W. Ruengsitagoon, C. Thongpoon, S. Liawruangrath, Anal. Chim. Acta 541 (2005) 105–111. [12] N. Anastos, N.W. Barnett, S.W. Lewis, N. Gathergood, P.J. Scammells, D.N. Sims, Talanta 67 (2005) 354–359. [13] Y.D. Liang, J.F. Song, T. Tian, Anal. Bioanal. Chem. 380 (2004) 918– 923. [14] X.F. Yang, H. Li, Talanta 64 (2005) 478–483. [15] P.J.L. Lopez, A. Townshend, Anal. Commun. 33 (1996) 31–33. [16] L. Wang, W.R.G. Baeyens, G. Van der Weken, A.M. Garcia-Campana, X.R. Zhang, E. Smet, Luminescence 15 (2000) 128–129. [17] A. Kojlo, J. Michalowski, E. Wolyniec, J. Pharm. Biomed. Anal. 22 (2000) 85–91. [18] F.A. Aly, A.A. Alwarthan, N.A. Alarfaj, Anal. Chim. Acta 358 (1998) 255–262. [19] L. Wang, W.R.G. Baeyens, G. Van der Weken, A.M. Garcia-Campana, X.R. Zhang, E. Smet, Luminescence 15 (2000) 81–83. [20] T. Perez-Ruiz, C. Martinez-Lozano, A. Sanz, M.T.S. Mignel, Lab. Autom. Inf. Manage. 34 (1999) 149–158. [21] W. Shi, J. Yang, Y. Huang, J. Pharm. Biomed. Anal. 36 (2004) 197– 203. [22] G.M. Greenway, S.J.L. Dolman, Analyst 124 (1999) 759–762. [23] G. Xu, S. Dong, Anal. Chem. 72 (2000) 5308–5312. [24] W. Liu, Pharmaceutical Analysis, People’s Medical Publishing House, Beijing, 1999, p. 151. [25] Society of Pharmacopoeia of Hygiene Department. Part II. Chemical Industry Press, Beijing, 2000, p. 574.
Y. Li et al. / Talanta 71 (2007) 1124–1129 [26] F. Nie, J.R. Lu, Y.H. He, J.X. Du, Talanta 66 (2005) 728–733. [27] Y.H. He, J.R. Lu, M. Liu, J.X. Du, Analyst 130 (2005) 1032– 1037. [28] E.H. White, M.M. Bursey, J. Am. Chem. Soc. 86 (1964) 941– 942.
1129
[29] E.H. White, O. Zafiriou, H.H. Kagi, J.H.M. Hill, J. Am. Chem. Soc. 86 (1964) 940–941. [30] J. Du, W. Liu, J. Lu, Acta Chim. Sinica 62 (2004) 1323–1326. [31] T.J. Mellinger, C.E. Keeler, Anal. Chem. 35 (1963) 554–558. [32] J.E. Wallace, J.D. Biggs, J. Pharm. Sci. 60 (1971) 1346–1352.
Sensitive determination of phenothiazines in pharmaceutical preparation and biological fluid by flow injection chemiluminescence method using luminol–KMnO4 system Yinhuan Li a,b , Weifen Niu a , Jiuru Lu a,∗ a
School of Chemistry and Materials Science, Shaanxi Normal University, Xi’an 710062, China b School of Science, Xi’an Jiaotong University, Xi’an 710049, China Received 23 March 2006; received in revised form 2 June 2006; accepted 3 June 2006 Available online 17 July 2006
Abstract A flow injection chemiluminescence method was described for the determination of four phenothiazine drugs, namely, chlorpromazine hydrochloride, perphenazine hydrochloride, fluphenazine hydrochloride and thioridazine hydrochloride. Strong Chemiluminescence (CL) signal was produced when above-mentioned drug was injected into the mixed stream of luminol with KMnO4 . The linear ranges of the method were 0.0020–1.0 g/mL chlorpromazine hydrochloride, 0.0040–3.0 g/mL perphenazine hydrochloride, 0.0020–5.0 g/mL fluphenazine hydrochloride and 0.0050–1.0 g/mL thioridazine hydrochloride. The detection limits were 0.4 ng/mL chlorpromazine hydrochloride, 0.7 ng/mL perphenazine hydrochloride, 2 ng/mL fluphenazine hydrochloride and 0.7 ng/mL thioridazine hydrochloride. The proposed method was applied to the determination of chlorpromazine hydrochloride in injections and in mental patient’s urine samples and the satisfactory results were achieved. The possible CL reaction mechanism was also discussed briefly. © 2006 Elsevier B.V. All rights reserved. Keywords: Chemiluminescence; Flow injection; Phenothiazines
1. Introduction Phenothiazine derivatives are a large group of tricyclic antidepressants which are commonly used for the treatment of psychiatric patients suffering depressions [1]. The monitoring of these compounds is important for quality assurance in pharmaceutical industry and for obtaining optimum therapeutic concentrations in body fluids to minimize the risk of toxicity. Therefore, it is important to develop simple and sensitive methods for the determination of these drugs. A wide variety of analytical techniques are available for the determination of phenothiazine drugs in pharmaceutical preparations and biological samples, such as spectrophotometry [2–4], electrochemical methods [5–7], chromatography [8] and capillary electrophoresis [9].
∗
Corresponding author. Tel.: +86 29 85303911; fax: +86 29 85307774. E-mail address: [email protected] (J. Lu).
0039-9140/$ – see front matter © 2006 Elsevier B.V. All rights reserved. doi:10.1016/j.talanta.2006.06.007
Chemiluminescence (CL) method has been frequently used for the analysis of pharmaceutical compounds in recent years [10–14] because of its promising advantages of low detection limit, wide linear dynamic range and relatively simple and inexpensive instrumentation. Several CL systems have been reported for the determination of phenothiazine drugs [15–23]. They were based on the acidic KMnO4 reaction [15–17], Ce(IV) reaction sensitized by rhodamine B [18], luminol–H2 O2 –Cr3+ reaction [19], luminol–Fe2+ reaction [20], luminol reaction in micelles [21] and electrogenerated Ru(bpy)3 2+ reaction [22,23]. The analytical parameters of the previous reported CL methods for the determination of phenothiazine drugs were summarized in Table 1. We here proposed a simple and sensitive flow injection CL method for the determination of four phenothiazine drugs, including chlorpromazine hydrochloride, perphenazine hydrochloride, fluphenazine hydrochloride and thioridazine hydrochloride. It was based on the fact that strong CL signal could be produced when above-mentioned drug was injected into the reaction mixture of luminol with KMnO4 . The method
Y. Li et al. / Talanta 71 (2007) 1124–1129
1125
Table 1 Analytical parameters of the previous CL methods for the determination of phenothiazine drugs (unit: g/mL) Species
CL reactions
Linear ranges
Chlorpromazine Thioridazine Fluphenazine Chlorpromazine Chlorpromazine Chlorpromazine Chlorpromazine Chlorpromazine
Acidic KMnO4 reaction Acidic KMnO4 reaction Ce(IV) reaction sensitized by rhodamine B Luminol–H2 O2 –Cr3+ reaction Luminol–Fe2+ reaction Luminol reaction in reverse micelles Electrogenerated Ru(bpy)3 2+ reaction Electrogenerated Ru(bpy)3 2+ reaction
1.0–30.0 0.5–90 3.6–355 0.7–7.1 0.05–10 0.3–6.4 0.004–1.1
was successfully applied to the determination of chlorpromazine hydrochloride in injections and in mental patient’s urine samples. 2. Experimental 2.1. Apparatus The schematic diagram of the CL flow system used was shown in Fig. 1. PTFE tubing (0.8 mm i.d.) was used as the connect material in the flow system. A peristaltic pump was used to deliver all solutions; each at a flow rate of 2.1 mL/min. Sample injection (50 L) was automated operated by means of a sixway valve. The CL signal produced in the flow cell was measured using an IFFS-A multifunction CL analyzer (Xi’an Remax Electronic Science-Tech Ltd. Co., China). The CL data acquisition and treatment were performed by using an MCDR-A multifunction data processing system (Xi’an Remax). CL spectra and fluorescence spectra were taken on CRT-970 fluorescence spectrophotometer (Shanghai Third Analytical Instrumental Plant).
Detection limits
References
7.1 0.5 0.01 13
[15] [17] [18] [19] [20] [21] [22] [23]
0.006 0.24 0.001
tions (500.0 g/mL) of each drug were prepared by dissolving 50.0 mg of the drug in 100 mL of water. They were stored in a refrigerator and protected from light. Working solutions were prepared by appropriately diluting the stock solution when used. Stock solution (0.01 mol/L) of luminol was prepared by dissolving 1.771 g luminol in 50 mL of 1 mol/L sodium hydroxide and diluting to 1 L with water. Stock solution (0.02 mol/L) of potassium permanganate was prepared by dissolving 3.16 g of KMnO4 in 1 L of boiled water, filtering through glass wool and protecting from light. 2.3. Procedure As shown in Fig. 1, flow lines were connected with phenothiazine solution, luminol solution and KMnO4 solution, respectively. Luminol solution was firstly mixed with KMnO4 solution via a Y-piece to give a stable baseline. Then 50 L solution of phenothiazine was injected into the merged stream of luminol with KMnO4 by means of a six-way valve to produce CL. The concentration of phenothiazine was quantified by the CL intensity.
2.2. Reagents and solutions 3. Results and discussion All chemicals were of analytical reagent grade except luminol, which was synthesized by the Institute of Analytical Science of Shaanxi Normal University (Xi’an, China). Doubly distilled water was used throughout the experiments. Chlorpromazine hydrochloride, perphenazine hydrochloride, fluphenazine hydrochloride and thioridazine hydrochloride were obtained from the Chinese Pharmaceutical and Biological Test Institute (Beijing, China). Potassium permanganate was purchased from Xi’an Chemical Plant (Xi’an, China). Stock solu-
3.1. Kinetic characteristic of the CL reaction The CL kinetic characteristic of the reaction was investigated using the static system of the IFFS-A multifunction CL analyzer. Fig. 2 shows the typical CL intensity-time profile. A strong CL was recorded when 1.0 mL solution of 20 mol/L KMnO4 was injected into 1.0 mL solution of 10 mol/L luminol (peak a). After CL signal declined about 45s, subsequently injecting 1.0 mL 0.5 g/mL chlorpromazine hydrochloride resulted in a new CL reaction (peak b). About 80s later, the CL reaction terminated and the CL signal declined to baseline. 3.2. Optimization of experimental conditions
Fig. 1. Schematic diagram of CL flow system. (a) Phenothiazines solution; (b) luminol solution; (c) potassium permanganate solution; P, peristaltic pump; V, injection valve; F, flow cell; PMT, photomultiplier tube; HV, high voltage; PC, personal computer; W, waste.
The experimental conditions were optimized using 0.05 g/mL chlorpromazine hydrochloride solution as a model. The optimized parameters were of reaction medium, luminol concentration and KMnO4 concentration. It was observed that higher CL signal and better repeatability could be obtained when sodium hydroxide was added into sample solution than into luminol solution. The effect of sodium
1126
Y. Li et al. / Talanta 71 (2007) 1124–1129
3.4. Interference
Fig. 2. The CL intensity-time profile. (a) injecting 1.0 mL 20 mol/L KMnO4 into 1.0 mL 10 mol/L luminol; (b) injecting 1.0 mL 0.5 g/mL chlorpromazine hydrochloride into above reaction mixture.
hydroxide concentration on the CL reaction was examined in the 0.02–0.1 mol/L. Finally, 0.05 mol/L of sodium hydroxide was employed because it gave larger signal-to-blank ratio and higher CL signal. The concentrations of luminol and KMnO4 were important parameters in the experiments. The effect of different concentrations combination of luminol (1–100 mol/L) and KMnO4 (1–500 mol/L) was studied. The experimental results showed that the maximum CL signal and signal-to-blank ratio was obtained when 10 mol/L of luminol and 20 mol/L of KMnO4 were used. The effect of flow rate on the reaction was examined in the range 0.7–3.5 mL/min. The CL signal increased with increase in flow rate from 0.7 to 2.1 mL/min. When the flow rate was higher than 2.1 mL/min, the CL signal changed slightly. Therefore, 2.1 mL/min of flow rate was employed. 3.3. Analytical performance Under the selected experimental conditions, the calibration graphs for four phenothiazine drugs were prepared. The parameters of calibration graphs and the detection limits (3sb ) were summarized in Table 2. The relative standard deviations for 0.01 and 0.1 g/mL solutions of phenothiazine drugs in 11 repeated measurements were also showed in Table 2. A complete analysis, including sampling and injecting, could be performed in 30 s, giving a sample throughput of 120/h.
In order to assess the possibility of applying the proposed method, the effect of foreign species was examined by analyzing synthetic sample solutions containing 0.050 g/mL of chlorpromazine hydrochloride and various amount of each potential interferent. A foreign species was considered not to interfere if it caused a relative error less than 5% in the determination of chlorpromazine hydrochloride. No interference has been found when including up to a 1000-fold Na+ , K+ , Al3+ , Fe3+ , Fe2+ , NO3 − , SO4 2− , CO3 2− , PO4 3− , starch, glucose, lactose, leucine, 500-fold Ca2+ , tryptophan, 200-fold urea, 100-fold Mg2+ , Cl− , 10-fold Mn2+ , Pb2+ , 1-fold fluphenazine hydrochloride, 0.2-fold ascorbic acid, epinephrine and 0.02-fold uric acid. 3.5. Application 3.5.1. Determination of chlorpromazine hydrochloride in pharmaceutical preparations Ascorbic acid is commonly used as the anti-oxidant for chlorpromazine hydrochloride injection [24]. The experimental results indicated that the effect of ascorbic acid could be eliminated by the addition of Fe3+ . Experimental result indicated that 50-fold amount of ascorbic acid did not interfere with the determination of 0.05 g/mL of chlorpromazine hydrochloride when 50 g/mL Fe3+ was present in sample solution. Five injections were homogenized. A portion of sample was transferred into a 50 mL beaker, to which 2.5 mL of 1 mg/mL Fe3+ solution was added. The solution was stood for 1 min and then 5 mL of 0.5 mol/L NaOH solution were added. The resultant solution was diluted with water to 50 mL and filtered. The filtrate was then analyzed according to the procedure described in Section 2. The results were summarized in Table 3. The t-test assured that there were no significant differences between the results obtained by the proposed method and those obtained by the Chinese Pharmacopoeia method [25] at the confidence level of 95%. 3.5.2. Determination of chlorpromazine hydrochloride in mental patient’s urine In order to apply the proposed method to the determination of chlorpromazine hydrochloride in mental patient’s urine, a chlorpromazine-imprinted polymer was employed to eliminate the interferents existed in urine. The chlorpromazine-imprinted polymer was synthesized by using batch method with chlorpro-
Table 2 Analytical performances Phenothiazines
Chlorpromazine hydrochloride Perphenizine hydrochloride Fluphenazine hydrochloride Thioridazine hydrochloride
Linear range (g/mL)
0.002–1.0 0.004–3.0 0.002–5.0 0.005–1.0
Regression equation (ng/mL)
I = 4.62C + 92.0 I = 2.33C − 38.8 I = 0.76C + 30.0 I = 3.20C − 40.0
Correlation coefficient (r2 )
0.9938 0.9972 0.9972 0.9959
Detection limit (3sb ) (ng/mL)
0.4 0.7 2 0.7
R.S.D. (%) (n = 11) 0.01 g/mL
0.1 g/mL
1.2 1.5 1.6 2.0
0.4 0.3 0.5 0.6
Y. Li et al. / Talanta 71 (2007) 1124–1129
1127
Table 3 Determination of chlorpromazine hydrochloride in injections Sample
Nominal content
Proposed methoda
R.S.D. (%) (n = 5)
Pharmacopoeia methodb
R.S.D. (%) (n = 3)
040912 041110
25 mg/1 mL 50 mg/2 mL
24.9 ± 0.5 (mg/1 mL) 49.9 ± 1.0 (mg/2 mL)
2.1 1.9
24.5 ± 0.2 (mg/1 mL) 49.2 ± 0.4 (mg/2 mL)
0.8 0.9
a b
Average of five determinations ± S.D. Average of three determinations ± S.D.
mazine as the template molecule and the procedure was similar to that reported previously [26]. The procedure of molecular imprinting-chemiluminescence method (MI-CL) consisted of four steps: adsorption of chlorpromazine, removing the other substances except chlorpromazine, CL detection and cleaning the MIP column. The flow system and procedure has been described in detail in elsewhere [27]. The experimental conditions corresponding to the four steps of MI-CL method were examined. The optimum conditions were 180 s of adsorption time, 120 s of washing time, 80 s of detection time and 80 s of cleaning time. When chlorpromazine-imprinted polymer was used, the tolerance limits were 1000-fold urea, 200-fold uric acid, 100-fold ascorbic acid, 50-fold fluphenazine hydrochloride and 2-fold epinephrine. Blank urine samples were collected from the healthy volunteers and the urine samples from the mental patients were collected after taking chlorpromazine hydrochloride for 8 h. After centrifuging at 3000 rpm for 15 min, 1 mL of the supernatant was transferred into a 100 mL volumetric flask, diluted with water to the mark and used as the sample solution. The results for the determination of chlorpromazine hydrochloride in urine samples were summarized in Table 4. Recovery tests were also carried out by adding a known amount of chlorpromazine hydrochloride in samples and the results were also summarized in Table 4. The t-test assumed that there were no significant differences between recoveries and 100% at confidence level of 95%. 3.6. Discussion on the reaction mechanism To elucidate the reaction mechanism, the following experiments were performed using chlorpromazine hydrochloride as a model.
Firstly, weak CL was recorded when alkaline chlorpromazine hydrochloride solution was injected into KMnO4 . Strong CL was detected when chlorpromazine hydrochloride solution was injected into the mixed solution of luminol and KMnO4 . Secondly, the CL spectra of luminol–KMnO4 reaction and luminol–KMnO4 –chlorpromazine hydrochloride reaction were drawn by means of a refitted CRT 970 fluorescence spectrophotometer (Fig. 3). Both spectra had same maximum emission wavelength at 425 nm. This indicated that the CL spectra were independent of chlorpromazine hydrochloride, the emitter of luminol–KMnO4 –chlorpromazine hydrochloride reaction was still the excited state of 3-aminophthalate ions (3-AP*) [28], the oxidation product of luminol [29]. Thirdly, when luminol was mixed with KMnO4 solution, the purple of KMnO4 solution changed gradually into green. This result indicated KMnO4 was firstly reduced to K2 MnO4 [30] and luminol was oxidized to 3-aminophthalate ions (3-AP) [29]. Fourthly, no CL signal was observed when injecting chlorpromazine hydrochloride into the mixed solution of luminol and K2 MnO4 . This experiment indicated that K2 MnO4 was irrelevant to the CL reaction. Fifthly, the alkaline mixed solutions of luminol and KMnO4 with different concentration ratio were prepared remaining luminol concentration at 1.0 × 10−5 mol/L. The concentration ratios of luminol to KMnO4 were 1:0.5 for solution A and 1:5 for solution B. When chlorpromazine hydrochloride was injected into above solution, the CL signal from solution B was much higher than that from solution A. This experiment suggested that KMnO4 was indispensable to the CL reaction. Sixthly, the fluorescent spectra of the reaction mixtures of luminol–KMnO4 (a), KMnO4 –chlorpromazine hydrochloride (b) and luminol–KMnO4 –chlorpromazine hydrochloride
Table 4 Determination of chlorpromazine hydrochloride in urine samples Urine samples
Determineda (g/mL)
Volunteer 1 Volunteer 2
R.S.D. (%, n = 3)
Added Found (g/mL) (g/mL)
Recovery (%)
0
1.00 2.00
0.96 2.04
96 102
0
1.00 2.00
1.03 1.95
103 98
Mental patient 1 0.42 ± 0.01
3.3
1.00 2.00
1.37 2.36
95 97
Mental patient 2 0.78 ± 0.02
3.1
1.00 2.00
1.74 2.71
96 97
a
Average of three determinations ± S.D.
Fig. 3. CL spectra. The CL spectra of the reaction of (a) 10 mol/L luminol + 20 mol/L KMnO4 ; (b) 10 mol/L luminol + 20 mol/L KMnO4 + 2.0 g/mL chlorpromazine hydrochloride.
1128
Y. Li et al. / Talanta 71 (2007) 1124–1129
4. Conclusion
Fig. 4. The fluorescent spectra. The excitation and emission spectra of the reaction mixtures of (a, a ): 10 mol/L luminol + 20 mol/L KMnO4 ; (b, b ): 20 mol/L KMnO4 + 2.0 g/mL chlorpromazine hydrochloride; (c, c ): 10 mol/L luminol + 20 mol/L KMnO4 + 2.0 g/mL chlorpromazine hydrochloride.
(c) were scanned using a CRT 970 fluorescence spectrophotometer (Fig. 4). The excitation spectra and emission spectra were similar for the reaction mixtures (a, c), the excitation wavelength and emission wavelength were within the range of 240–400 nm (λex = 314 nm) and 370–600 nm (λem = 425 nm), respectively. They were identical with the excitation spectrum and emission spectrum of 3-AP [28]. The excitation wavelength and emission wavelength for reaction mixture (b) were 240–370 nm (λex = 306 nm) and 350–470 nm (λem = 386 nm), respectively. It was identical with the spectrum of the oxidation product of chlorpromazine hydrochloride [31], sulphone compound [32]. It was also interestingly found that the excitation spectrum of sulphone compound (b) largely overlapped with that of 3-AP (a, c), which suggested that the energy released by reaction between KMnO4 and chlorpromazine hydrochloride could be absorbed not only by sulphone compound to form excited state of sulphone compound but also by 3-AP to form 3-AP*, respectively. Based on the above discussion, the following possible reaction mechanism was suggested: Chlorpromazine hydrochloride reacted with KMnO4 and the energy was released. The released energy was absorbed by 3-AP to form 3-AP* . The 3-AP* returned to ground state and generated CL. Luminol + KMnO4 → 3-AP + K2 MnO4 + hν
(425 nm)
KMnO4 + chlorpromazinehydrochloride + 3-AP → sulphonecompound + K2 MnO4 +3-AP∗ 3-AP∗ → 3-AP + hν
(425 nm)
The same phenomenon was also found when chlorpromazine hydrochloride was replaced by other three phenothiazine drugs. This may be attributed to their similar molecular structure and fluorescence characteristics [31]. Therefore, the CL reaction mechanism for other three phenothiazine drugs should be identical with that of chlorpromazine hydrochloride.
Four phenothiazine drugs including chlorpromazine hydrochloride, perphenazine hydrochloride, fluphenazine hydrochloride and thioridazine hydrochloride were found to produce strong CL in KMnO4 –luminol reaction. The proposed method has wider linear ranges (The linear range extended 2–3 magnitudes) and lower detection limit (The detection limit was two magnitudes lower than that of most reported CL methods). The proposed method was successfully applied to the determination of chlorpromazine hydrochloride in injections and in mental patient’s urines. The possible CL reaction mechanism was investigated. Acknowledgement The authors gratefully acknowledge the financial support from Shaanxi Normal University for Doctoral Fund. References [1] J. Karpinska, B. Starczweska, H. Puzanowska-tarasiewicz, Anal. Sci. 12 (1996) 161–170. [2] M.A. El-Maaboud Ismail, Talanta 44 (1997) 1173–1182. [3] T. Aman, J. Anwar, A. Ahmad, L. Latif, Anal. Lett. 36 (2003) 2961–2974. [4] K. Basavaiah, J.M. Swamy, Chem. Anal. 47 (2002) 139–146. ˇ anek, J. [5] A. Ferancov´a, E. Korgov´a, T. Buzinkaiov´a, W. Kutner, I. Step´ Labuda, Anal. Chim. Acta 447 (2001) 47–54. ¨ [6] Z. Yang, J. Kauffmann, M.A. Valenzuela, S. Ozkan, Mikrochim. Acta 131 (1999) 85–90. [7] I. Nemcova, K. Nesmerak, D. Rychlovsky, J. Koutnikova, Talanta 65 (2005) 632–637. [8] M.C. Quintana, M.H. Blanco, J. Kacal, L. Hernandez, Talanta 59 (2003) 417–422. [9] F.J. Lara, A.M. Garcia-Campana, F. Ales-Barrero, J.M. Bosque-Sendra, Anal. Chim. Acta. 535 (2005) 101–108. [10] Y. Fuster Mestre, L. Lahuerata Zamora, J. Martinez Calatayud, Luminescence 16 (2001) 213–235. [11] A. Townshend, W. Ruengsitagoon, C. Thongpoon, S. Liawruangrath, Anal. Chim. Acta 541 (2005) 105–111. [12] N. Anastos, N.W. Barnett, S.W. Lewis, N. Gathergood, P.J. Scammells, D.N. Sims, Talanta 67 (2005) 354–359. [13] Y.D. Liang, J.F. Song, T. Tian, Anal. Bioanal. Chem. 380 (2004) 918– 923. [14] X.F. Yang, H. Li, Talanta 64 (2005) 478–483. [15] P.J.L. Lopez, A. Townshend, Anal. Commun. 33 (1996) 31–33. [16] L. Wang, W.R.G. Baeyens, G. Van der Weken, A.M. Garcia-Campana, X.R. Zhang, E. Smet, Luminescence 15 (2000) 128–129. [17] A. Kojlo, J. Michalowski, E. Wolyniec, J. Pharm. Biomed. Anal. 22 (2000) 85–91. [18] F.A. Aly, A.A. Alwarthan, N.A. Alarfaj, Anal. Chim. Acta 358 (1998) 255–262. [19] L. Wang, W.R.G. Baeyens, G. Van der Weken, A.M. Garcia-Campana, X.R. Zhang, E. Smet, Luminescence 15 (2000) 81–83. [20] T. Perez-Ruiz, C. Martinez-Lozano, A. Sanz, M.T.S. Mignel, Lab. Autom. Inf. Manage. 34 (1999) 149–158. [21] W. Shi, J. Yang, Y. Huang, J. Pharm. Biomed. Anal. 36 (2004) 197– 203. [22] G.M. Greenway, S.J.L. Dolman, Analyst 124 (1999) 759–762. [23] G. Xu, S. Dong, Anal. Chem. 72 (2000) 5308–5312. [24] W. Liu, Pharmaceutical Analysis, People’s Medical Publishing House, Beijing, 1999, p. 151. [25] Society of Pharmacopoeia of Hygiene Department. Part II. Chemical Industry Press, Beijing, 2000, p. 574.
Y. Li et al. / Talanta 71 (2007) 1124–1129 [26] F. Nie, J.R. Lu, Y.H. He, J.X. Du, Talanta 66 (2005) 728–733. [27] Y.H. He, J.R. Lu, M. Liu, J.X. Du, Analyst 130 (2005) 1032– 1037. [28] E.H. White, M.M. Bursey, J. Am. Chem. Soc. 86 (1964) 941– 942.
1129
[29] E.H. White, O. Zafiriou, H.H. Kagi, J.H.M. Hill, J. Am. Chem. Soc. 86 (1964) 940–941. [30] J. Du, W. Liu, J. Lu, Acta Chim. Sinica 62 (2004) 1323–1326. [31] T.J. Mellinger, C.E. Keeler, Anal. Chem. 35 (1963) 554–558. [32] J.E. Wallace, J.D. Biggs, J. Pharm. Sci. 60 (1971) 1346–1352.