Solvent Extraction Coupled On-Line to a Reversed Micellar Mediated

A fast and sensitive method for the determination of atropine, an alkaloid closely ... Using the reverse-flow injection and chemical conditions optimi...
0 downloads 0 Views 57KB Size
Download PDF
Anal. Chem. 2000, 72, 1715-1719

Solvent Extraction Coupled On-Line to a Reversed Micellar Mediated Chemiluminescence Detection System for Trace-Level Determination of Atropine Terufumi Fujiwara,*,† Imdad U. Mohammadzai,‡ Katsumi Murayama,† and Takahiro Kumamaru§

Department of Chemistry, Graduate School of Science, Hiroshima University, Kagamiyama, Higashi-Hiroshima 739-8526, Japan, Department of Chemistry, University of Peshawar, Peshawar, Pakistan, and Department of Life Science, Yasuda Women’s Junior College, Yasuhigashi, Asaminami-ku, Hiroshima 731-0153, Japan

A fast and sensitive method for the determination of atropine, an alkaloid closely related to cocaine, is proposed. The principles of on-line ion-pair formation of alkaloid-metal complexes and liquid-liquid extraction are applied to the chemiluminescence determination of atropine. On mixing with a reversed micellar medium of cetyltrimethylammonium chloride in dichloromethanecyclohexane (1:1 v/v)-water (0.3 M Na2CO3) containing luminol, the ion-pair complex of tetrachloroaurate(III) with atropinium produced an analytical chemiluminescence signal when it entered the reversed micellar water pool. Using the reverse-flow injection and chemical conditions optimized for atropine in aqueous samples, a detection limit of 1 ng/mL was achieved and a linear calibration graph was obtained with a wide dynamic range from 10 ng/mL to 100 µg/mL. The proposed method is simple and provides a good precision with a relative standard deviation (n ) 6) of ca. 3% at the atropine concentration of 100 ng/mL. After a preliminary study involving the potential interference from species of organic, inorganic, and metallic nature, the method was applied to the determination of atropine in artificial urine samples and of atropine and scopolamine in pharmaceutical formulations. Alkaloids are naturally occurring amines derived from plant sources. At present, there are some 6000 naturally occurring alkaloids.1 Medicinally, alkaloids have played a key role for millennia, and even today some 25% of commercial drugs are either alkaloids or their structural modifications or analogues.1 Atropine and scopolamine are closely related to cocaine. Besides being used as an antidote for anticholinesterase poisons, atropine is used in several ophthalmic products. Various methods of analyzing tropane alkaloids/metabolites have been developed for forensic applications. The list of reported methods ranges from classical procedures to the most popular chromatographic tech†

Hiroshima University. University of Peshawar. § Yasuda Women’s Junior College. (1) Koskinen, A. Asymmetric Synthesis of Natural Products; JWS: London, 1995; p 196. ‡

10.1021/ac991087d CCC: $19.00 Published on Web 03/03/2000

© 2000 American Chemical Society

niques.2,3 Most of them require a more or less complex treatment of the sample, derivatization, extraction, etc., and an expensive instrumentation if selective quantification is desired. Because of inherent simplicity, speed, and high sensitivity, chemiluminescence (CL) is an established method of detection for routine analysis in several fields of analytical chemistry.4 Trace analysis is among the existing applications where CL-based detection is used, and unparalleled results for analytical measurements at ultralow levels are reported. Also, CL detection in liquid chromatography and flow injection analysis (FIA) has received considerable attention.5,6 The importance of micelles and microemulsions has been widely demonstrated in view of their analytical applications.7,8 In CL measurements, an interesting development of analytical significance is the incorporation of reverse micelles into the detection system. Results from different CL studies indicate many advantages, including sensitivity and improved selectivity in the reversed micellar system.7,9-14 After dispersion in an apolar organic phase, the molecules of surfactant encompass tiny water droplets and are converted into homogeneously distributed micelles referred to as microreactors.15 The signifi(2) Alcada, N. M. P. M.; Lima, L. F. C. J.; Montenegro, B. S. M. C. M. Anal. Sci. 1995, 11, 781-785 (and references therein). (3) Lehr, G. J. J. Assoc. Off. Anal. Chem. 1996, 79, 1288-1293 (and references therein). (4) Gutierrez, F. A.; Pena, d. L. M. In Molecular Luminescence Spectroscopy. Methods and Applications: Part 1; Schulman, S. G., Ed.; Wiley: New York, 1985; pp 463-546 (and references therein). (5) Fujiwara, T.; Kumamaru, T. Spectrochim. Acta Rev. 1990, 13, 399-406. (6) Nieman, T. A. In Chemiluminescence and Photochemical Reaction Detection in Chromatography; Birks, J. W., Ed.; VCH: New York, 1989; pp 99-123. (7) Georges, J. Spectrochim. Acta Rev. 1990, 13, 27-45. (8) Kumamaru, T.; Okamoto, Y.; Yamamoto, M.; Obata, M.; Onizuka, K. Anal. Chim. Acta 1990, 232, 389-391. (9) Fujiwara, T.; Tanimoto, N.; Huang, J. J.; Kumamaru, T. Anal. Chem. 1989, 61, 2800-2803. (10) Hinze, W. L.; Srinivasan, N.; Smith, T. K.; Igarashi, S.; Hoshino, H. In Advances in Multidimensional Luminescence; Warner, I. M., McGown, L. B., Eds.; JAI Press: Greenwich, CT, 1991; Vol. 1, pp 149-206 (and references therein). (11) Fujiwara, T.; Tanimoto, N.; Nakahara, K.; Kumamaru, T. Chem. Lett. 1991, 1137-1140. (12) Imdadullah; Fujiwara, T.; Kumamaru, T. Anal. Chem. 1991, 63, 23482352. (13) Imdadullah; Fujiwara, T.; Kumamaru, T. Anal. Sci. 1991, 7 (Suppl), 13991402. (14) Imdadullah,; Fujiwara, T.; Kumamaru, T. Anal. Chem. 1993, 65, 421-424. (15) See, e.g.: Pileni, M. P. In Structure and Reactivity in Reverse Micelles; Pileni, M. P., Ed.; Elsevier: Amsterdam, 1989 (see also references therein).

Analytical Chemistry, Vol. 72, No. 7, April 1, 2000 1715

cance of reverse micelles in CL analysis is considered to be due to their unique structure (size/shape) and composition. Although not proven, it is believed that reversed micellar mediated CL (RMM-CL) reactions occur at surfactant-water interfaces.10,16,17 With the additional advantage of sensitivity, these microreactors have the capability to transfer species of experimental interest quantitatively into the water pool.9,11,16 Because of the catalytic behavior of different metal ions and the effectiveness of reverse micelles in CL reactions, we used RMM-CL reactions to develop new methods for the off/on-line trace-level quantification of gold(III),12-14,18 rhodium(III),19 iron(III),11,20 iron(II),20 and vanadium(IV).16,17 Most probably via an ion-pair formation in the extraction process, gold(III) was transferred as the tetrachloroaurate ion from aqueous solution into chloroform containing tri-n-octylphosphine oxide.13 The postextraction step was directly coupled to an RMM-CL detection system, and the resulting solvent extraction/RMM-CL hybrid method was then applied to the determination of gold in industrial samples.13,14 Analytical procedures based on the principles of ion-pair formation between a protonated alkaloid and a negatively charged metal-containing species have been developed for the indirect AAS determination of alkaloids.21-23 The feasibility of using AuCl4- for the on-line extraction and CL determination of atropine in aqueous media is examined in this paper. The fundamental theory is closely related to the procedure for gold (III) determination based on CL oxidation of luminol with AuCl4- in reverse micelles as reported previously.12-14,18 The validation of the proposed method is confirmed by applying it to the determination of atropine in synthetic urine samples and of atropine and scopolamine in pharmaceutical preparations. The proposed procedure could be extended to the determination of cocaine. It has the advantages of simplicity, sensitivity, and convenience, since the lengthy processes associated with solvent/back-extraction are avoided. Additionally, minimum sample manipulation is involved before analysis. EXPERIMENTAL SECTION Chemicals. All metal standard solutions (AAS grade), including those of chloroauric acid, hydrochloric acid (electronic grade), urea, uric acid, hippuric acid, glucose, and sodium chloride, were purchased from the Kanto Chemical Co., Inc. (Tokyo). Anhydrous sodium carbonate (99.98%, analytical grade) was purchased from the Asahi Glass Co., Inc. (Tokyo). Luminol was obtained from the Aldrich Chemical Co., Inc. (Milwaukee, WI), and atropine sulfate, scopolamine hydrobromide, harmine hydrochloride, adrenaline, nicotine, and cetyltrimethylammonium chloride (CTAC) were obtained from the Tokyo Kasei Kogyo Co., Ltd. (Tokyo). Standard pharmaceutical samples of atropine sulfate monohydrate and (16) Fujiwara, T.; Theingi-Kyaw; Kumamaru, T. Anal. Sci. 1997, 13 (Suppl.), 59-62. (17) Theingi-Kyaw; Kumooka, S.; Okamoto, Y.; Fujiwara, T.; Kumamaru, T. Anal. Sci. 1999, 15, 293-297. (18) Fujiwara, T.; Murayama, K.; Imdadullah; Kumamaru, T. Microchem. J. 1994, 49, 183-193. (19) Imdadullah; Fujiwara, T.; Kumamaru, T. Anal. Chim. Acta 1994, 292, 151157. (20) Theingi-Kyaw; Fujiwara, T.; Inoue, H.; Okamoto, Y.; Kumamaru, T. Anal. Sci. 1998, 14, 203-207. (21) Nerin, C.; Garnica, A.; Cacho, J. Anal. Chem. 1986, 58, 2617-2621. (22) Travnikoff, B. Anal. Chem. 1983, 55, 795-796. (23) Gallego, M.; Silva, M.; Valcarcel, M. Anal. Chem. 1986, 58, 2265-2269.

1716 Analytical Chemistry, Vol. 72, No. 7, April 1, 2000

Figure 1. Flow diagram for the on-line solvent extraction/RMM-CL determination of atropine: C, dichloromethane; E, dichloromethane; S, sample; L, luminescent reagent; EC, extraction coil; PS, phase separator; D, detector; W, waste; P1, plunger pump; P2, peristaltic pump.

scopolamine hydrobromide trihydrate were obtained from the Tanabe Pharmaceutical Co. (Osaka, Japan) and the Kyorin Pharmaceutical Co. (Tokyo), respectively. Dichloromethane and cyclohexane were purchased from Wako Pure Chemical Industries, Ltd. (Osaka, Japan), potassium phosphate and ammonia were purchased from Katayama Chemical Industries, Ltd. (Osaka, Japan), and calcium chloride was purchased from Yoneyama Chemical Industries, Ltd. (Osaka, Japan). All chemicals were used as received. Deionized water was freshly collected from an Advantec Toyo (Tokyo) model GSU-901 water purification apparatus and utilized in the preparation of all aqueous solutions and for related cleaning purposes. Working solutions were freshly prepared before use. All glassware was soaked in 20% nitric acid and thoroughly cleaned before use. Synthetic urine was prepared as reported earlier,24 stored in clean polythene bottles, and kept refrigerated. Before analysis, urine samples were spiked with the alkaloid and/or interfering species. Apparatus. The multicomponent instrument, used for the reverse FIA system of on-line continuous extraction, phase separation, and CL detection, was the same (Figure 1) as that reported previously.18 The extraction coil (a 3 m PTFE tube of 0.5 mm i.d.) and the homemade phase separator with a Teflon membrane (60 µm thickness and 0.22 µm porosity) were also the same as reported earlier.18,20 PTFE tubing of 0.5 mm i.d. was used throughout the flow system. The CL signals were recorded on an ordinary strip chart recorder. Analytical Procedure. An aqueous sample solution (100 mL) of atropine in 0.1 M HCl containing 10 µg/mL Au(III) as AuCl4(51 µM) was pumped with a flow rate of 5 mL/min using a peristaltic pump. At a flow rate of 2 mL/min, an organic solvent stream of dichloromethane was mixed with the aqueous sample stream; the ratio of the flow rates results in an aqueous-to-organic volume ratio of 2.5. The mixture was passed through the extraction coil where the associated ion pairs were transferred from the aqueous into the organic phase. The two phases were separated using a Teflon membrane. Nearly 85% of the organic phase was membrane-transferred and passed onward in the line. The reversed micellar solution of luminol was prepared as before19 by dispersing a certain volume of the carbonate (0.3 M Na2CO3)buffered luminol stock solution (pH ) 11.5) in a reversed micellar (24) Albero, M. I.; Garcia, M. S.; Pedreno, C. S.; Rodrigues, J. Analyst 1990, 117, 1635-1638.

bulk solvent of dichloromethane-cyclohexane (1:1 v/v) containing CTAC (0.10 M), with a water-to-surfactant molar ratio (R ) [H2O]/ [CTAC]) of 11. The luminol concentration was 4.0 × 10-4 M, calculated on a final total volume basis. Using a rotary injection valve, the luminol solution was sucked into a 100 µL loop which was then inserted into the carrier stream of dichloromethane pumped with a flow rate of 2 mL/min. The luminescent reagent was mixed with the organic phase containing the Au(III)-alkaloid ion pairs in a 70 µL spiral flow cell mounted in front of the phototube of the photometer, and the resultant CL signal was recorded. An aqueous solution of 100 ng/mL atropine was used in optimizing experimental and chemical parameters. RESULTS AND DISCUSSION Ion-Pair Extraction. The applications of ion-pair formation to the liquid-liquid extraction of different species of interest are well established.25 Numerous analytical procedures based on the formation of ion pairs between protonated alkaloids and negatively charged metal complexes extractable into organic solvents have been reported.21,22,26,27 The nitrogen of atropine is first protonated in an acidic medium and is then associated with a negatively charged counterion containing gold(III). As ion pairs of alkaloids are sparingly soluble in water but relatively soluble in slightly polar organic solvents, the hydrated ion-pair associate atropine H+AuCl4-‚nH2O, once formed, is then quickly and efficiently extracted into a slightly polar solvent, e.g. 1,2-dichloroethane or dichloromethane. 1,2-Dichloroethane is more suitable for extracting ion pairs, as reported previously.27 In this work, however, dichloromethane was selected as the extractant, in view the stability of the CTAC reverse micelles and the compatibility of the extractant and the reversed micellar bulk solvent, which are important for obtaining reproducible CL signals. In our previous studies,13,14,18 chloroform was used for extraction, but for the present ion pair of atropine, because of its lower polarity or dielectric constant, the efficiency of chloroform extraction was relatively poor. Effect of the AuCl4- Concentration. The effect of the gold(III) concentration on the formation and extraction of the ionpair complex followed by CL detection was investigated. In the absence of AuCl4-, negligibly small CL signals due to the luminol reagent alone in the reversed micellar solution were observed for aqueous samples of atropine. Also, CL emission resulted when an aqueous gold(III) solution without atropine was used to determine a blank signal, as the gold(III) complex in an acidic medium is also slightly extractable in the form of chloroauric acid. An analytical CL signal was taken as the difference in observed peak heights for the analyte and the blank. As the concentration of AuCl4- in the aqueous medium was increased, the CL intensity increased, indicating the occurrence of ion-pair formation and extraction. However, the blank CL signal also increased, implying that the background problem associated with using higher concentrations of the gold(III) complex for extraction causes difficulties in obtaining reproducible peak heights. For subsequent work, a concentration of 51 µM AuCl4- was thus recommended. (25) Marcus, Y.; Kertes, A. S. Ion Exchange and Solvent Extraction of Metal Complexes; Wiley & Sons: New York, 1969. (26) Garcia, A. M.; Uria, E. S.; Sanz-Medal, L. J. Anal. At. Spectrom. 1996, 11, 561-565. (27) Eisman, M.; Gallego, M.; Valcarcel, M. Anal. Chem. 1992, 64, 1509-1512.

Effect of the HCl Concentration. No CL signals were observed for neutral aqueous solutions, indicating the absence of protonated atropine. With an increase in the concentration of HCl, an increase in the CL intensity was observed, reaching a maximum around 0.1 M. Beyond this value, the CL intensity showed a decline. A similar behavior was reported earlier indicating a decline in the CL intensity when the concentration of HCl was increased beyond a certain level using chloroform as the extraction solvent.18 Although unconfirmed, a possible explanation could be that a small quantity of free HCl, accumulated in dichloromethane during the extraction process, appears to cross the Teflon membrane, probably in the gaseous form, ultimately causing an alteration in the constitution of the luminol buffer in the reversed micellar water pool. A 0.1 M concentration of HCl was, therefore, considered to be the optimum for facilitating both protonation and ion-pair formation. Optimization of the Luminescent Reagent. The variations in the CL intensity with the concentrations of luminol and sodium carbonate were each individually observed for an aqueous solution of 100 ng/mL atropine in 0.1 M HCl containing 51 µM AuCl4-. With an increase in the concentration of luminol in the reversed micellar solution of 0.10 M CTAC at R ) 11, the CL intensity increased and reached a maximum around an optimum value of 4.0 × 10-4 M. Increasing the concentration of sodium carbonate in the buffered luminol solution also caused an increase in the CL intensity and a maximum CL intensity was attained around 0.3 M Na2CO3 (pH ) 11.5), which was chosen as the optimum concentration. Effect of the [H2O]/[CTAC] Molar Ratio. The molar ratio, R, strongly affects the physiochemical properties of the reverse micelles. By changing R, one can modify the size of the reverse micelles.15,28-31 The R value can be varied by changing either the amount of water or the concentration of the surfactant. For an aqueous solution of 100 ng/mL atropine in 0.1 M HCl containing 51 µM AuCl4-, it was observed that, at a constant amount of water buffered with 0.3 M Na2CO3 containing luminol, as the surfactant CTAC concentration in the reversed micellar medium was lowered, a pronounced change in the CL intensity occurred (Figure 2a). The intensity of the CL signal reached a maximum around an R value of 8-12. An optimum R value of 11 was selected and was further investigated by changing the amount of water and keeping the amount of surfactant constant. Figure 2b shows the change in the R value at a constant CTAC concentration of 0.10 M in the micellar medium. Both results for the CL signals in Figure 2 agreed for the corresponding values of R. An increase in the R value (hence more free water was available for CL reaction) caused an increase in the size of the reverse micelles.28,31 It can be presumed that, initially, the surface available for the CL reaction in a reverse micelle is increased, reaching a maximum around 11 in the on-line system. However, with an increase in the R value, the relative amount of CTAC in the reverse micelle is decreased, leading to instability of the reversed micellar system. (28) Day, R. A.; Robinson, B. H.; Clarke, J. H. R.; Doherty, J. V. J. Chem. Soc., Faraday Trans. 1 1979, 75, 132-139. (29) Jada, A.; Lang, J.; Zana, R.; Makhloufi, R.; Hirsch, E.; Candau, S. J. J. Phys. Chem. 1990, 94, 387-395. (30) Nishimoto, J.; Iwamoto, E.; Fujiwara, T.; Kumamaru, T. J. Chem. Soc., Faraday Trans. 1993, 89, 535-538. (31) Lang, J.; Mascolo, G.; Zana, R.; Luisi, P. L. J. Phys. Chem. 1990, 94, 30693074.

Analytical Chemistry, Vol. 72, No. 7, April 1, 2000

1717

Table 1. Metals Tested for Their Possible Interfering Behavior in the Solvent Extraction/CL Determination of Atropine Using Luminol in Reversed Micelles rel CL intensa metal Ca(II) Cr(III) Mn(II) Fe(III) Co(II) Ni(II)

100b

94 102

rel CL intensa

1000b

metal

97 95 97 115 100 123

Cu(II) Zn(II) Pt(II) Pd(II) Ag(I) Pb(II)

100b

1000b 103 148 139 171 107 99

99 100 114

a All CL intensities are relative to the CL signal ()100) for 100 ng/ mL atropine alone. b Metal-to-atropine weight ratio.

Figure 2. Dependence of the CL intensity for 100 ng/mL atropine on the R ()[H2O]/[CTAC]) value at a constant concentration of H2O or CTAC in the reverse micelles: (a) [H2O] ) 1.1 M; (b) [CTAC] ) 0.10 M.

Table 2. Constituent Species of the Synthetic Urine Tested for Their Possible Interfering Behavior in the Solvent Extraction/CL Determination of Atropine Using Luminol in Reversed Micelles species urea

At a constant R value of 11, the CL intensity increased upon an increase in both the CTAC concentration and the amount of water in the reversed micellar solution. Upon changing the amount of water and the surfactant concentration at a fixed R, the sizes of the reverse micelles remain unchanged28 but their population is increased. Therefore, it is presumed that an increase in the surfactant concentration and amount of water in the reverse micelles at a constant value of R leads to the formation of more micelles or microreactors of identical size. CL signals of maximum intensity were observed around a CTAC concentration of 0.10 M, which was selected as the optimum concentration for the surfactant. Analytical Performance. Under the optimized instrumental and chemical conditions, the detection limit (DL) for aqueous samples was 1 ng/mL atropine (3.5 × 10-9 M), where the DL is given as the concentration for which the analytical signal is 3 times higher than the noise level of the baseline. The calibration graph obtained using 51 µM AuCl4- for the extraction is linear in the range from 10 ng/mL to 100 µg/mL atropine in aqueous samples. In our previous work on the Au(III) determination based on the RMM-CL reaction of chloroaurate with luminol,12 a DL of 10 pg/ mL Au(III), 0.05 × 10-9 M, and a linear calibration line up to 1 µg/mL were obtained. These data indicate that only a small amount (ca. 1%) of atropine in the aqueous sample is probably extracted by the present on-line system under the optimized conditions. The proposed CL method is nevertheless more sensitive than other techniques such as potentiometry2 and AAS.27 The relative standard deviation at the atropine concentration of 100 ng/mL was 3% (n ) 6). Interference Effects. Metallic Species. The specificity of the method for atropine in the presence of some common metals that are capable of catalyzing the CL reaction of luminol4,6 and likely to be potential ionic interfering species was examined. The results obtained are summarized in Table 1. None of the species tested interfered with the CL determination of atropine when present at an interference-to-atropine weight ratio of 100. When the ratio was increased to 1000, an enhancement in the CL signal was observed 1718

Analytical Chemistry, Vol. 72, No. 7, April 1, 2000

uric acid hippuric acid glucose lactic acid PO4Na+ K+ Ca2+

concn

rel CL intensa

4.0 × 10-4 Mb 1.0 × 10-4 M 2.0 × 10-4 Mb 8.0 × 10-3 Mb 4.0 × 10-4 Mb 1.5 × 10-4 M 5.0 × 10-4 Mb 5.0 × 102 µg/mLb 4.0 × 103 µg/mLb 2.0 × 103 µg/mL 5.0 × 102 µg/mLb 5.0 × 102 µg/mLb

108 102 100 95 83 96 104 100 360 142 100 104

a All CL intensities are relative to the CL signal ()100) for 100 ng/ mL atropine alone. b The concentrations were reported earlier for synthetic urine.24

only for iron(III), nickel(II), zinc(II), platinum(II), and palladium(II). Species Constituting Synthetic Urine. All species constituting the synthetic urine were tested individually for their possible interfering behavior. The results, summarized in Table 2, indicated that, with the exception of sodium, none of the species interfered with the CL determination of atropine. However, at a Na+-toatropine weight ratio of 2.0 × 104 or 4.0 × 104, significant enhancement in the CL signals occurred. To explain the involvement of Na+ in the process, we propose that probably an ion pair was formed between Na+ and AuCl4-, followed by its extraction into the organic phase and subsequent production of the CL reaction. This interfering behavior of Na+ was further investigated by changing the dichloromethane-to-cyclohexane volume ratio in the organic phase. It was observed (Figure 3) that, at a lower volume ratio, only Na+ was extracted while the atropine-containing ion pair needed a higher volume ratio (6 and above) for extraction. By simply shaking the aqueous sample with a dichloromethanecyclohexane mixture at a volume ratio of 8 prior to analysis with the present CL method, we were able to screen out Na+ from the synthetic urine sample, and hence its interference was eliminated. Alkaloids and Other Organic Species. The effect of some common alkaloids, e.g. harmine and scopolamine, which can form similar ion pairs with the AuCl4- ion, was also examined to check possible interference. An interference from both of the alkaloids

Table 3. Determinations of Atropine and Scopolamine in Pharmaceutical Formulations samplea atropinec scopolamined

found values/µg 10.3 10.2

10.2 9.6

estd valueb/µg 10.0 10.0

a A 20 µL aqueous sample was used. b The values were calculated from nominal values, respectively. c 2(atropine)‚H2SO4‚H2O. d Scopolamine‚HBr‚ 3H O. 2

Figure 3. Effect of the dichloromethane-to-cyclohexane volume ratio on the solvent extraction of the ion-pair associate and its subsequent RMM-CL determination: (O) [atropine] ) 1.0 × 102 ng/mL; (4) [Na+] ) 4.0 × 103 µg/mL; (0) [scopolamine] ) 1.0 × 102 ng/mL; (]) [harmine] ) 1.0 × 102 ng/mL.

Figure 4. Calibration curve (O) and standard-addition curve (b) for determination of atropine (100 ng/mL) in the synthetic urine.

was observed in the present work. Figure 3 shows the variation in the CL intensities with the dichloromethane-to-cyclohexane volume ratio in the organic phase used for extracting harmine and scopolamine. When present with atropine, both alkaloids could be easily removed by simply shaking the aqueous sample with the Au(III) complex using an organic phase composed of dichloromethane and cyclohexane at a volume ratio of 4 as indicated in Figure 4. On the other hand, nicotine and adrenaline, which can also be protonated and associated with the Au(III) counterion, did not interfere at all. No interference from the metabolite cotinine, which is also found in body fluids, might be expected, as it is chemically close to nicotine, although it was not tested in this work. Cysteine showed a CL inhibition effect when it was present at a weight ratio of 10 or more, and at a weight ratio of 100, no CL signals at all were observed for 100 ng/mL atropine. In the subsequent work, cysteine was successfully masked with 100 µg/mL Cu(II) before CL analysis; hence no interference from the copper(II) ion was observed, as mentioned above (Table 1). Application to the Determination of Atropine. Figure 4 shows both the calibration and standard-addition curves used for

the determination of atropine (100 ng/mL) in synthetic urine samples. The results were better for the standard-addition method, which is hence recommended for analytical work on urine samples. The utility of this method was further tested by applying it to the determination of atropine and scopolamine in standard pharmaceutical products using the optimized conditions and the respective calibration curves. For the quantification of both atropine and scopolamine, the results obtained (Table 3) were in good agreement with the estimated values. CONCLUSIONS An on-line ion-pair formation and solvent extraction method coupled with reversed micellar mediated chemiluminescence detection was applied to the determination of atropine in synthetic urine and of atropine and scopolamine in standard pharmaceuticals. With a simple experimental setup using low-cost instrumentation, the proposed method allows the trace-level quantification of atropine and scopolamine in aqueous samples. The method has several advantages over the manual extraction procedure: simplicity, less human exposure to toxic reagents/solvents, speed, sensitivity, precision, minimum risk of contamination, etc. Since atropine is chemically very similar to cocaine, after improvement/ validation studies, there will be ample opportunities for the proposed method to be confidently applied in the clinical or forensic analysis of cocaine. ACKNOWLEDGMENT This work was partially supported by a Grant-in-Aid for Scientific Research (No. 10440223) from the Ministry of Education, Science, Sports, and Culture of Japan. I.U.M. thanks the Association of International Education of Japan for the award of a Postdoctoral Fellowship and the University of Peshawar for granting leave.

Received for review September 20, 1999. Accepted January 5, 2000. AC991087D

Analytical Chemistry, Vol. 72, No. 7, April 1, 2000

1719